Advanced Electrochemical Reforming of Methanol for Hydrogen Production by Caroline R. Cloutier M.A.Sc., Metals and Materials Engineering, University of British Columbia, 2006 B.A.Sc., Chemical Engineering, Environmental Option, Co-op, University of Ottawa, 1999. A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Chemical & Biological Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December, 2011 @ Caroline R. Cloutier, 2011 ii Abstract The issue of efficient, low-cost, sustainable hydrogen (H2) production is one of the barriers to the adoption of a H2 economy. In this thesis, the electrochemical production of H2 from liquid methanol (CH3OH) in acidic aqueous media was studied in a proton exchange membrane (PEM) electrolyser in the static mode at low temperatures. A baseline study showing the influence of CH3OH concentration, catalyst, catalyst support, operating temperature and operating mode was established. A theoretical thermodynamic analysis of the system was carried out as a function of temperature, and the limiting current densities, kinetic parameters, including the Tafel slopes and current exchange density, and apparent activation energies were determined. The effect of electrochemical promotion (EP) was investigated to see if it can increase the efficiency and performance of H2 production through electrochemical processes. The electrochemical promotion of electrocatalysis (EPOE) was investigated by carrying out the electrolysis in triode and tetrode operation. It was shown to improve the PEM electrolysis in the galvanostatic and potentiostatic modes. A decrease in electrolysis voltage or an increase in electrolysis current proportional to the current or potential imposed in the auxiliary circuit was observed when the auxiliary current or potential was opposite to the electrolyser circuit current or potential. The effect was observed using catalytic and non-catalytic non-precious electrolyser electrode materials. It was postulated that triode and tetrode operation enhanced the electro-oxidation rate through electrochemical pumping and spillover of protons. With this novel electrolysis configuration, electrolysis cost reduction may be achieved through the use of non-precious electrolyser anode materials and/or improving electrolyser performance. The electrochemical promotion of catalysis (EPOC) was also investigated for the catalytic reforming of CH3OH at low temperature with Pt-Ru/C and Pt-Ru/TiO2. The synthesized Pt-Ru/TiO2 was characterized physico-chemically and electrochemically. Powder catalytic CH3OH reforming tests showed that both catalysts can be used to generate H2. EPOC experiments were conducted on gas diffusion electrodes (GDEs) in galvanostatic control. Under the experimental conditions, only supplying H+ to the catalyst iii working electrode surface resulted in only in a Faradaic enhancement of the catalytic activity for the low temperature reforming of CH3OH, which appears to be a purely electrophilic behaviour. iv Preface The literature review, design of the experimental set-up, experimental data collection and analysis was conducted by Caroline R. Cloutier under the supervision of Dr. David P. Wilkinson. The material covered in Chapter 2 was published: • Cloutier, C. R., and Wilkinson, D. P., “Electrolytic Production of Hydrogen from Aqueous Acidic Methanol Solutions”, Int. J. H2 Energy, 35, (2010) 3967-3984. Co-op student Jason Gao co-produced the data used in Fig. 2.6 (e) in Chapter 2 under the supervision of Caroline R. Cloutier. Some of the material covered in Chapter 3 was published in a conference proceeding and another paper has been submitted for publication: • Cloutier, C. R., and Wilkinson, D. P., "Triode Operation of a Proton Exchange Membrane (PEM) Electrolyser", ECS Transactions, 25 (23), (2010) 47-57. • Cloutier, C. R., and Wilkinson, D. P., “Electrochemical Promotion of Aqueous Acidic Methanol PEM Electrolysis in Triode and Tetrode Operation”, (submitted). A manuscript based on the content of Chapter 4 is in preparation. The initial and final drafts of this thesis were prepared by Caroline R. Cloutier, with revisions edited and approved by Dr. David P. Wilkinson. The thesis proposal and the thesis manuscript were reviewed by the Ph.D. Committee, which included Dr. David Wilkinson, Dr. Elöd Gyenge, Dr. Walter Mérida, and Dr. David Dreisinger. v Table of Contents Abstract………………………………………………………………………............ii Preface………………………………………………………………………………..iv Table of Contents……………………………………………………………………v List of Tables…………………………………………………………………………x List of Figures………………………………………………………………………xii Nomenclature…………………………………………………………………...xvii List of Acronyms……………………………………………………………...…xxi Acknowledgements…………………………………………………………..…xxiv Dedication…………………………………….…………………………………xxvi Chapter 1: Introduction……………………..……………………………………1 1.1 Background……………………………………………...…………………………..…1 1.1.1 Alternative fuels..……….………………………………………………………………..……1 1.1.2 Technology gap ……………………………………..……………………………………..…3 1.1.3 Proposed solution………………………..………………………………………………….3 1.1.4 A sustainable closed loop system………….….…………………………………..............4 1.2 Thesis overview……………………………………………………..…………………6 1.2.1 Justification…………………………………..………………………………………….……6 1.2.2 Research objectives……………………………………………………………………..…7 1.2.3 Layout…………………………………………………………...……………………………8 1.3 Literature review………………..………………….…………………………………10 1.3.1 Heterogeneous catalysis……………..…………………………………………………….10 1.3.1.1 Steam reforming……………………...…………………………………………………11 1.3.1.2 Liquid phase reforming…………………………………………………………………12 1.3.1.3 Hydrogen spillover in heterogeneous catalysis………………………………………13 1.3.1.4 Catalyst local heating…………………………………………………………………14 1.3.2 Proton exchange membrane fuel cell………...………………………………………….15 1.3.2.1 Membrane electrode assembly……………………………………………………….16 1.3.2.2 Overvoltage………………………………………………………………………………17 1.3.2.3 Hydrogen proton exchange membrane fuel cell…………………………………….18 Hydrogen purity requirements……………….……..…………………………….…………19 1.3.2.4 Direct methanol fuel cell………………………………………………………………..20 Anodic methanol oxidation……………………………………………………………….....21 Methanol crossover…………………………………….………………………………..…22 1.3.2.5 Indirect methanol fuel cell………………………………………………………………23 1.3.3 Proton exchange membrane electrolysis………………………….……………………24 1.3.3.1 Proton exchange membrane water electrolysis……………………………………24 1.3.3.2 Depolarized proton exchange membrane water electrolysis……………………….26 Methanol depolarized electrolysis…………………………………………………………27 Applications of depolarized methanol electrolysis………………………………………32 vi 1.3.4 Electrochemical promotion of heterogeneous catalysis………………………………34 1.3.4.1 Theory……………………………………………………………………………………35 1.3.4.2 Parameters………………………………………………………………………………39 1.3.4.3 Electrochemical promotion of methanol oxidation…………………………………41 1.3.4.4 Electrochemical promotion using protonic electrolytes……………….……………42 1.3.4.5 Applications for the electrochemical promotion of catalysis………………………44 1.3.5 Electrochemical promotion of electrocatalysis………………………………….………..47 Chapter 2: Electrocatalysis Baseline Study ………………………………….49 2.1 Synopsis………………………………………………………………………………49 2.2 Experimental………………………………………………………………………..…50 2.2.1 Materials……………………………………………………………………………………...50 2.2.2 Equipment……………………………………………………………………………………51 2.2.3 Electrochemical measurements……………………………………………………………54 2.2.4 Characterization…………………..………………………………………………………....54 2.3 Results and discussion……………………………………………………………57 2.3.1 Thermodynamic evaluation of methanol reforming at low temperatures……………...57 2.3.2 Electrolysis polarizations…………………………………………………………………62 2.3.2.1 Cell voltage stability……………………………………………………………………62 2.3.2.2 Effect of methanol concentration and anode catalyst………………………………64 2.3.2.3 Effect of dry purged cathode…………………………………………………………71 2.3.2.4 Effect of temperature……………………………………………………………………73 2.3.2.5 Anode Tafel kinetics……………………………………………………………………76 2.3.2.6 Activation energy………………………………………………………………………85 2.4 Summary……………………………………………………………………………….87 Chapter 3: Electrochemical Promotion of Electrocatalysis (EPOE).…….88 3.1 Synopsis………………………………………………………………………………88 3.2 Overview………………………………………………………………………………89 3.3 Experimental…………………………………………………………………………93 3.3.1 Materials and membrane electrode assembly…………………………………………...93 3.3.2 Glass cell for triode and tetrode operation……………..…………………………………95 3.3.3 Electrochemical measurements……………………………………………………………97 3.3.4 Stability of carbon containing electrolyser ring electrodes………………….…………101 3.4 Triode or tetrode operation evaluation expressions…………………………103 3.4.1 Triode or tetrode operation evaluation in the galvanostatic mode………………….103 3.4.2 Triode or tetrode operation evaluation in the potentiostatic mode……………………105 3.5 Results and discussion…………………………………………………………....106 3.5.1 Galvanostatic characterization…………………………………………………………106 3.5.1.1 Triode and tetrode operation in galvanostatic control…………………………….106 3.5.1.2 Parallel and reverse operation in galvanostatic control…….……………..………107 3.5.1.3 Electrolyser ring electrode material in galvanostatic control………………………109 3.5.1.4 Electrolyser voltage ratio, power enhancement ratio and power gain ratio in galvanostatic control……………………………………………………………………………111 3.5.1.5 Benefits of starting in triode operation in galvanostatic control…………………114 3.5.2 Potentiostatic characterization……………………..……………………………………116 3.5.2.1 Triode configuration in potentiostatic control……….…...………………………….116 3.5.2.2 Tetrode configuration in potentiostatic control……..……………………...............118 3.5.2.3 Counter electrode potential measurements in potentiostatic control……………119 3.5.2.4 Electrolyser ring electrode material in potentiostatic control..………..…………121 3.5.2.5 Electrolyser rate enhancement ratio, power gain ratio and Faradaic efficiency in potentiostatic control………………………………………………………..………………123 3.5.3 Durability investigation………………………..…………………………………………126 3.5.3.1 Long-term triode and tetrode operation in galvanostatic control…………………126 3.5.3.2 Methanol concentration change over time in long-term triode or tetrode vii galvanostatic control….....................................................................................................128 3.5.3.3 Long-term tetrode operation in mixed potentiostatic and galvanostatic control...130 3.5.3.4 Methanol concentration change over time in long-term tetrode operation in mixed potentiostatic and galvanostatic control……………………………………………………132 3.5.4 Mechanism investigation……………………………………….....................................134 3.5.4.1 Electrolytic contact of auxiliary electrodes in potentiostatic control ..……………134 3.5.4.2 Effect of acid concentration and conductivity on triode operation in potentiostatic control………………………………………………………………………………………….136 3.5.4.3 Effect of triode operation on membrane conductivity in potentiostatic control….138 3.5.4.4 Effect of electrolyser working electrode ring geometry in potentiostatic control..139 3.5.4.5 Effect of proton flux lines direction in potentiostatic control….……………………140 3.5.4.6 Effect of the electrochemical surface area of the carbon fibre paper electrolyser working electrode in potentiostatic control….……………………….………………...........143 3.5.4.7 Effect of triode operation using non-noble non-carbon containing working electrode materials in potentiostatic control…………….…………………………………...144 3.5.4.8 Proposed mechanism…………………………………………………………………153 3.6 Summary………………………………………………….…………………………156 Chapter 4: Catalysis and Electrochemical Promotion of Catalysis (EPOC)……………………………………………………………………………157 4.1 Synopsis………………………………………………………………………..…….157 4.2 Experimental…………………………………………………………………………160 4.2.1 Materials……..……………………………………………………………………………160 4.2.2 Catalyst synthesis…….……………………………………………………………………160 4.2.3 Pt-Ru/TiO2 physicochemical characterization…………….…………………………162 4.2.3.1 Scanning electron microscopy……………………………………………………162 4.2.3.2 Energy-dispersive X-ray spectroscopy………………………………………………162 4.2.3.3 Transmission-electron microscopy…….……………………………………………163 4.2.3.4 X-ray diffraction…………………………………………………………………...……163 4.2.3.5 X-ray photoelectron spectrometry……………………………………………………164 4.2.3.6 Inductively coupled plasma optical emission spectrometry……………………164 4.2.3.7 Brunauer, Emmett and Teller surface area and Barett-Joyner-Halenda pore size and volume analysis……………………………..…………………………………………164 4.2.4 Pt-Ru/TiO2 electrochemical characterization……………..…………………………165 4.2.5 Apparatus………………………………………………………………………………...…166 4.2.5.1 Powder catalysis………………………………………………………………………166 4.2.5.2 GDE catalysis and electrochemical promotion of catalysis……………………….171 4.3 Results and discussion……………………………………………………………173 4.3.1 Pt-Ru/TiO2 physicochemical characterization…………….…………………….………173 4.3.1.1 SEM images………..………….……………………………………………….………173 4.3.1.2 EDX measurements……………...……………………………………………....…...174 4.3.1.3 TEM images……….……..………………………………………………………..…...175 4.3.1.4 XRD measurements……………...……………………………………………….…..177 4.3.1.5 XPS measurements…………………………………………………………………...180 4.3.1.6 ICP-OES measurements…………..………………………………………………….181 4.3.1.7 BET surface area and BJH pore size and volume analysis……………………….182 4.3.2 Pt-Ru/TiO2 electrochemical characterization……………...……………………………184 4.3.3 Powder catalysis…………………………….……………………………………………..192 4.3.4 GDE catalysis and electrochemical promotion of catalysis………………………..….199 4.4 Summary……………………………………………………………………………214 Chapter 5: Conclusions…………………………………………………………216 5.1 Electrocatalysis baseline study……………………………………….…………216 5.2 Electrochemical promotion of electrocatalysis (EPOE).……………………218 5.2.1 Galvanostatic and potentiostatic characterization……………………………………218 viii 5.2.2 Durability investigations…………………………………………………………………219 5.2.3 Mechanism investigations………………………………………………………………220 5.3 Electrochemical promotion of catalysis (EPOC).……………………………..221 5.3.1 Catalyst synthesis and characterization…………………………………………………222 5.3.2 Powder catalysis…………………………………………………………………………...222 5.3.3 GDE catalysis and electrochemical promotion of catalysis…………………………...223 5.4 Potential application, significance and impact of research findings……...224 5.4.1 Potential applications……………………………………………………………………...224 5.4.2 Significance and impact…………………………………………………………………..228 5.5 Recommendations for further research………………...………………………230 5.5.1 Electrocatalysis baseline study…………………………………………………………..230 5.5.2 Electrochemical promotion of electrocatalysis (EPOE)….…………………………….231 5.5.3 Electrochemical promotion of catalysis (EPOC).……………………………………….234 References……………………………….…………………………………….….236 Appendix A: Publications, presentations, and posters…………………………..250 A.1 Publications………………………………………...…………………………..…………….250 A.2 Presentations………………………………………………………...………………………250 A.3 Posters……………………………………………………………...………………………...250 Appendix B: Methanol electro-oxidation mechanism………………………….…251 B.1 Methanol electro-oxidation parallel pathways………………………………………..…..251 B.2 Methanol electro-oxidation mechanism on Pt……………………………………….……251 B.3 Methanol electro-oxidation mechanism on Pt-Ru…………………………………..……252 B.4 References……………………………………………………………………………….…..252 Appendix C: Efficiency and economic comparison of H2 production…………254 C.1 Objective…………………………………………………………………………..………….254 C.2 System definition………………………………………………………………………...…..254 C.3 Block diagrams………………………………………………………………………………254 C.4 Feedstock processing………………………………………………………………...…….256 C.4.1 Methanol production…………………………………………………………………….256 C.4.2 Sodium borohydride production………………………………………………………..257 C.5 Hydrogen production………………………………………………………………...……...257 C.5.1 Electrolysis……………………………………………………………………………….258 C.5.2 Catalytic steam reforming………………………………………………………………259 C.5.3 Sodium borohydride hydrolyser………………………………………………………..260 C.6 Efficiency analysis…………………………………………………………..……………….260 C.7 Economic analysis………………………………………………………...………………...262 C.7.1 Capital cost……………………………………………………………………………….264 C.7.2 Feedstock cost…………………………………………………………………………..265 C.7.3 Energy cost………………………………………………………………………………265 C.7.4 Carbon cost………………………………………………………………………………266 C.7.5 Economic analysis summary…………………………………………………………...267 C.8 Conclusion…………………………………………………………………………..………..268 C.9 References…………………………………………………………………………………...269 Appendix D: Supplemental EPOC and EPOE theory…………………….…...…..271 D.1 EPOC rules…………………………………………………………..……………...……….271 D.2 Chemisorption on a transition metal……………………………………………………….273 D.3 Kirchhoff’s first law for triode and tetrode operation………………….………………….275 D.3.1 Triode fuel cell……………………………………………………………………………275 D.3.2 Triode electrolyser……………………………………………………………………….276 D.3.3 Tetrode electrolyser……………….…………………………………………………….276 D.4 References…………………………………………………………………………………...277 ix Appendix E: Experimental procedures…………………………………………...…278 E.1 Catalyst electrode preparation…………………………………...………..…….………...278 E.1.1 Cathode carbon sublayer composition calculation…………………………………..278 E.1.2 Anode and cathode catalyst ink composition calculation…………………………...279 E.1.3 Catalyst ink preparation procedure…………………………………………………….281 E.1.4 Electrode spraying procedure………………………………………………………….282 E.2 Platinization procedure…………………………………………………………..………….283 E.3 Silicon ring preparation………………………………….…………………….…………….284 E.4 Nafion membrane conditioning………………………………..…………………………...285 E.5 Determination of CH3OH concentration……………………..…………………………….286 E.6 RDE catalyst ink preparation………………………………...……………………………..289 E.7 Pt-Ru/TiO2 preparation…………………………………………….………………………..290 E.8 References…………………………………………………………...………………………294 Appendix F: Electrochemical techniques………………………………………..…295 F.1 Cyclic voltammetry………………………………………………………………….……….295 F.1.1 Pt-C catalyst electrochemical surface area determination…………………………..296 F.1.2 Pt-Ru/C catalyst electrochemical surface area determination………………………300 F.2 Electrochemical impedance spectroscopy………………………………………………303 F.3 Determination of the open circuit voltage…………………………………………………305 F.4 Current - potential transients…………………….………………………………………....306 F.5 Rotating disk electrode……………………………..…………………………………..…...309 F.6 References……………………………………………………………………………………311 Appendix G: Thermodynamic data and sample calculations…………..……….313 G.1 Thermodynamic data………………………………………………………………………..313 G.2 Thermodynamic sample calculations…………………………………………………..….313 G.2.1 Liquid phase……………………………………………………………………………..313 G.2.2 Gas phase……………………………...………………………………………………..315 G.2.3 System efficiency calculation…………………………………………………………..320 G.2.4 Effect of pressure in the gas phase……………………………………………………321 G.2.5 Effect of anode and cathode pressure in the liquid phase………………………….322 G.3 References…………………………………………………………………………………...323 Appendix H: Electrochemical data and sample calculations..………………….324 H.1 Power consumption………………………………………………………………………….324 H.2 CO2 emissions and H2 production rate……………………………………………...…….324 H.3 Proton diffusion limited H2 production – back envelope calculation……..…………….326 H.4 Overall system efficiency…….……………………………………………………...……...328 H.5 References…………………………………………………………………………………330 x List of Tables Table 1.1 Methanol properties………………………………………………………………………..2 Table 1.2 Relevant methanol electrochemical reforming membrane reactors..……………….28 Table 1.3 Vapour phase methanol oxidation EPOC studies with oxide conductors……………42 Table 1.4 Vapour/gaseous phase EPOC studies using protonic conducting electrolytes……..43 Table 1.5 EPOC studies in aqueous systems…………………………………………………….44 Table 2.1 Electrochemical characterization of supported and unsupported catalysts studied…………………………………………………………………………………….56 Table 2.2 Literature Tafel kinetic parameters for the electro-oxidation of methanol and water on Pt and Pt-Ru catalysts……………………………….……………………………………78 Table 2.3 Tafel kinetic parameters from current work for the electro-oxidation of methanol and water on Pt and Pt-Ru catalysts………………………………….………………………80 Table 2.4 Activation energy data for the electro-oxidation of methanol on Pt and Pt-Ru catalysts…………………………………………………………………………………….86 Table 3.1 List of possible half-cell reactions and their standard potentials..……………………92 Table 3.2 IR corrected on-set potentials for CFP samples with and without Pt black at varying CH3OH concentrations………………………………………..…………………………102 Table 3.3 Effect of auxiliary WEaux potential (Uaux, we) on the rate enhancement ratio (ρe) (Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, Uelec, we = 0.52 V vs. SHE)…………… ……….123 Table 3.4 IR corrected on-set potential for different electrolyser and auxiliary WEaux materials at varying CH3OH concentrations………………………………………….…………...148 Table 3.5 Effect of auxiliary WEaux potential (Uaux, we) on the rate enhancement ratio (ρe) using Au plated 316 SS components (2 mg/cm2 Pt/C CE, Uelec, we = 0.52 V vs. SHE)…150 Table 4.1 Scherrer’s crystallite size for selected peaks from the carbon, Pt-Ru/C, TiO2 and Pt- Ru/TiO2 XRD patterns.……………………………………………………….………179 Table 4.2 ICP-OES results for H2 reduced Pt-Ru/TiO2 before and after ball milling and Pt- Ru/C………………………………………………………………………………………182 Table 4.3 BET surface areas of the Pt-Ru/TiO2 materials synthesized, as well as commercial TiO2 and Pt-Ru/C materials……………………………………………………………183 Table 4.4 On-set potentials for the forward and reverse CH3OH oxidation on different catalytic surfaces (2 M CH3OH in 0.5 M H2SO4).…………………….………………………..187 Table 4.5 Peak potentials for the forward and reverse CH3OH oxidation on different catalytic surfaces (2 M CH3OH in 0.5 M H2SO4).…....………..………………………………...187 Table 4.6 List of CV studies on unsupported and supported Pt, Pt-Ru and TiO2 available in the literature………………………………………………………………………………….188 Table 4.7 Literature powder catalysis CH3OH reforming production rates at low temperature (0.5 g Pt-Ru/TiO2, 10 wt% metal, Pt:Ru 1:1 a/o, 60 ml of 2 M CH3OH)……...…….196 Table 4.8 Powder catalysis CH3OH reforming production rates (0.5 g catalyst, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4, 75 ±1oC)…..………………....196 Table 4.9 Imposed currents or potentials on selected chemical reactions subjected to EPOC……………………………………………………………………………………..200 Table 4.10 Average H2 unpromoted production rates for GDE catalysis of CH3OH at low temperature (2.01 cm2, 4 mg/cm2 Pt-Ru/C or Pt-Ru/TiO2, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75±1oC)……………………………………………………….204 Table 4.11 Faradaic H2 production rate, measured H2 production rates, Faradaic efficiencies and EPOC efficiencies for the GDE catalytic reforming of CH3OH at low temperature (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4, 75±1oC)…………………………………………………………………...210 Table C.1 High level comparison of four H2 generation process schemes………………….…258 Table C.2 Efficiency comparison of four H2 generation process schemes…………………….261 xi Table C.3 Cost comparison of four H2 generation process schemes…………………………..268 Table E.1 Cathode carbon sublayer calculation spreadsheet…………………………………...279 Table E.2 Anode catalyst ink composition calculation spreadsheet………….…………………280 Table E.3 Cathode catalyst ink composition calculation spreadsheet……………………….…281 Table E.4 Precursor salt addition calculation spreadsheet for the impregnation method….…291 Table F.1 Total metal area results comparison for electrode degradation evaluation………..300 Table G.1 Thermodynamic data………………………………………………………………….…313 Table G.2 Empirical coefficients for Shomate equations…………………………………………317 Table H.1 Direct methanol reformer electrochemical active surface area as a function of proton diffusion coefficient in Nafion, based on H2 production for a 100 kW PEMFC at a voltage of 0.6 V………………………………………………………………………….328 xii List of Figures Figure 1.1 (a) Schematic diagram of conventional methanol synthesis combined with a DMR and a PEMFC (b) Schematic diagram of a future sustainable system comprising a DMR and PEMFC…………………………………………………………………………………………...5 Figure 1.2 Schematic diagram of a methanol electrochemical reforming membrane cell (Modified from patent DE 197007384 [112])…………………………………………………………...29 Figure 1.3 Schematic diagram of an alcohol electrolyser (Modified from Hu et al. [7] with permission from Elsevier)………………………………….……………………………........30 Figure 1.4 Schematic diagram of a metal catalyst-electrode deposited on (a) an oxide-conducting solid electrolyte, and (b) a proton-conducting solid electrolyte (Modified from Vayenas et al. [18] with permission from Springer)……………………………………………………...37 Figure 1.5 Schematic diagram of a triode fuel cell electrical circuit…...……………………………...47 Figure 2.1 Schematic diagram of (a) the electrochemical glass cell, (b) the MEA components (c) the MEA holder.………… ………………………………………………….…………….......52 Figure 2.2 Picture of (a) the electrochemical glass cell, (b) the MEA holder………………………...53 Figure 2.3 Theoretical thermodynamic efficiency and theoretical cell voltage as a function of temperature (1 atm) for a DMR and PEM water electrolyser……………………………..61 Figure 2.4 Electrolysis cell voltage stability with respect to current density (16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC)………………………………………………...62 Figure 2.5 IR corrected electrolysis cell voltage repeatability as a function of geometric current density in the stable and unstable regions (2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 2 mg/cm2 Pt/C anode and cathode, 23 ±2oC)………………………………………………………………………………………......63 Figure 2.6 Effect of methanol concentration on the IR corrected electrolysis cell voltage as a function of geometric current density for different anode catalysts (a) 4 mg/cm2 Pt-Ru/C, (b) 2 mg/cm2 Pt/C, (c) 4 mg/cm2 Pt-Ru black, (d) 2 mg/cm2 Pt black, (e) well-defined Pt disk (0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 23 ±2oC).……………………………………………….......………..66 Figure 2.7 Dependence of limiting geometric current densities for the oxidation of methanol and water on the CH3OH concentration for various anode catalysts (0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 23 ±2oC)………………………………………………………………………………………........67 Figure 2.8 Effect of CH3OH concentration on the individual electrode potential (0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC)………..……………………..68 Figure 2.9 Effect of the IR corrected voltage on electrolysis for Pt/C and Pt black anode catalysts as a function of real electrochemical current density (0, 2, or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 23 ±2oC)………….....70 Figure 2.10 Effect of dry N2 purged cathode on the IR corrected electrolysis cell voltage as a function of geometric current density (0 or 1 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, wet electrolyser catholyte: 0.5 M H2SO4, dry electrolyser catholyte: N2 purge electrolyser, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC)……………..71 Figure 2.11 Effect of dry N2 purged cathode and a wet cathode on the (a) anode potential and (b) cathode potential, as a function of geometric current density (0 or 1 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, wet electrolyser catholyte: 0.5 M H2SO4, dry electrolyser catholyte: N2 purge, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC)…….72 Figure 2.12 Effect of CH3OH concentration on the IR corrected electrolysis cell voltage as a function of geometric current density for different temperatures (a) 23 ±2oC (b) 50 ±1oC (c) 75 ±1oC (0, 2 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode)……………75 Figure 2.13 Tafel plot of the IR corrected anodic overpotential as a function of the log of the geometric current density (0, 2, or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 xiii M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt black anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC)…………………………………………………………………………………….......….76 Figure 3.1 Schematic diagram of the triode electrolyser electrical circuit…………………………....89 Figure 3.2 Side view of membrane electrode assemblies in different configurations: (a) triode, normal operation, common CE (b) like (a) with auxiliary WEaux away from surface (c) like (a) in inverted operation, and (d) tetrode, normal operation, independent CEs……………………………………………………………………………………….……..90 Figure 3.3 Schematic drawing of auxiliary TeflonTM cap with apertures………………………….…..94 Figure 3.4 Front view of the triode or tetrode operation working electrode components…………..94 Figure 3.5 Apparatus for triode and tetrode electrolysis (a) electrochemical glass cell, (b) auxiliary electrode ……………………………………………………………….……………………....96 Figure 3.6 Picture of the electrochemical glass cell set-up in triode operation……….……………..97 Figure 3.7 Potentiostatic working electrode potential experimental settings relative to the standard half-cell electrochemical potentials……………….……………………………………….…98 Figure 3.8 Schematic diagram of EPOE testing system connections for (a) galvanostatic triode operation, (b) potentiostatic triode operation (c) galvanostatic tetrode operation, and (d) potentiostatic tetrode operation……………………………………………………...…...….99 Figure 3.9 IR-corrected forward anodic potential sweep (5 mV/s, 0, 2 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, CFP with and without 2 mg/cm2 Pt black WE, Pt flag CE, 23 ±1oC)…………......…….……………………….…101 Figure 3.10 Triode effect in galvanostatic control (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2, iaux, cell = -50 mA/cm2)……………….…107 Figure 3.11 Effect of varying Iaux, cell on the electrolyser circuit voltage in galvanostatic control (a) parallel operation (b) reverse operation (4 mg/cm2 Pt-Ru black electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2, iaux, cell = -50 mA/cm2)…...108 Figure 3.12 Effect of varying Iaux, cell on the electrolyser circuit voltage in galvanostatic control (a) 4 mg/cm2 Pt-Ru black electrolyser ring WEelec, (b) CFP electrolyser ring WEelec (Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2)………………………………....110 Figure 3.13 Effect of iaux, cell on Uelec, cell and on the triode voltage ratio (R) (a) 4 mg/cm2 Pt-Ru black electrolyser ring WEelec (b) CFP electrolyser ring WEelec (Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2)…………………………………………………………112 Figure 3.14 Power enhancement ratio (ρelec, cell) and power gain ratio (γelec, cell) as a function of the auxiliary circuit current density (a) 4 mg/cm2 Pt-Ru black electrolyser ring WEelec (b) CFP electrolyser ring WEelec (Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2)……………………………………………………………………………………..113 Figure 3.15 Benefits of starting in triode operation in galvanostatic control (CFP ring electrolyser WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 50, 75 mA/cm2, iaux, cell = -5 A/cm2)……………………………………………………………………………………........115 Figure 3.16 Triode effect in potentiostatic control (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)……......……………………………….116 Figure 3.17 Triode and tetrode effect in potentiostatic control (4 mg/cm2 Pt-Ru black electrolyser ring WEelec, 2 mg/cm2 Pt/C electrolyser CEelec, Pt auxiliary WEaux and CEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE).……………………………………………………………………………………119 Figure 3.18 Triode effect in potentiostatic control (a) geometric current density (b) counter electrode potential (4 mg/cm2 Pt-Ru black electrolyser ring WEelec, 2 mg/cm2 Pt/C electrolyser CE, Pt auxiliary WEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)……..........……………………………120 Figure 3.19 Tetrode effect in potentiostatic control (a) geometric current density (b) counter electrode potential (4 mg/cm2 Pt-Ru black electrolyser ring WEelec, 2 mg/cm2 Pt/C xiv electrolyser CEelec, Pt auxiliary WEaux and CEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)...120 Figure 3.20 Effect of varying Uaux, we on different electrolyser ring WEelec materials in potentiostatic control (a) 0 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte and (b) 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte (2 mg/cm2 Pt black, 4 mg/cm2 Pt-Ru black, and CFP electrolyser WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, Uelec, we = 0.52 V vs. SHE)…................122 Figure 3.21 Power gain ratio (γelec, cell) as a function of Uaux, we (CFP electrolyser ring WEelec and 4 mg/cm2 Pt-Ru black electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)…………………………………………………………………………125 Figure 3.22 Durability test in normal, triode and tetrode electrolysis in galvanostatic control (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2, iaux, cell = -50 mA/cm2)………………………………………………………………………..127 Figure 3.23 Methanol concentration as a function of time as predicted by Faraday’s law, and during normal, triode, and tetrode electrolysis in galvanostatic control (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2, iaux, cell = -50 mA/cm2)….……………………………...…………………...........................................129 Figure 3.24 Potentiostatic electrolyser circuit control and galvanostatic auxiliary circuit control durability test in tetrode electrolysis (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CEelec, Pt auxiliary CEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 1 V vs. SHE, iaux, cell = -20 mA/cm2)…...131 Figure 3.25 Methanol concentration as a function of time as predicted by Faraday’s law and as measured during potentiostatic electrolyser circuit control and galvanostatic auxiliary circuit control in tetrode electrolysis (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CEelec, and Pt auxiliary CEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 1 V vs. SHE, iaux, cell = -20 mA/cm2)…………………………………………………………………………………….....133 Figure 3.26 Effect of varying Uaux, we on the electrolyser current density in tetrode operation (a) Pt auxiliary WEaux 1 cm away from surface (b) Pt auxiliary CEaux 1 cm away from surface (CFP electrolyser ring WEelec, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)…………………………..........135 Figure 3.27 Nafion® N-117 membrane conductivity and solution conductivity as a function of sulphuric acid concentration………………………………………………………………136 Figure 3.28 Effect of varying Uaux, we on the electrolyser current density (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 0 M CH3OH in 0.05 or 5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)...137 Figure 3.29 Resistance measurement over time in normal electrolysis and in triode electrolysis (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE, Uaux, we = -0.59 V vs. SHE)……………………………………………………………139 Figure 3.30 Effect of varying Uaux, we on the ielec, cell for different electrolyser ring WEelec gap distances (CFP electrolyser ring WEelec, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)……………...….....................140 Figure 3.31 Effect of varying Uaux, we on the current density in tetrode operation (a) inverted CEs (CFP ring WEelec, Pt rod WEaux, Pt rod CEelec, Pt/C CEaux) (b) inverted WEs (Pt rod WEelec, CFP ring WEaux, Pt/C ring CEelec, Pt rod CEaux), (2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)...142 Figure 3.32 Effect of electrolyser ring CFP WEelec carbon SA on ielec, cell (2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE)...144 Figure 3.33 IR corrected forward anodic potential sweeps (a) 316 SS auxiliary WEaux (b) 316 SS mesh electrolyser ring WEelec (c) Au plated 316 SS auxiliary WEaux (d) Au plated 316 SS mesh electrolyser ring WEelec (e) Pt auxiliary WEaux (5 mV/s, 0, 2, or 16 M CH3OH in 0.5 M H2SO4, Pt flag CE, 23 ±1oC)……….…………….………………………………………147 Figure 3.34 Effect of varying Uaux, we on (a) Au plated 316 SS electrolyser ring WEelec, Pt auxiliary WEaux, 0 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, (b) Au plated 316 SS electrolyser ring WEelec, Pt auxiliary WEaux, 2 M CH3OH xv in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, (c) Au plated 316 SS electrolyser ring WEelec, Au plated 316 SS auxiliary WEaux, 0 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, (d) Au plated 316 SS electrolyser ring WEelec, Au plated 316 SS auxiliary WEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte (triode, Pt/C CE, Uelec, we = 0.52 V vs. SHE)…………………………………………………………………………………..149 Figure 3.35 Power gain ratio (γelec, cell) as a function of Uaux, we (a) 0 M CH3OH, (b) 2 M CH3OH (Au plated 316 SS components, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE).……………...152 Figure 3.36 Schematic diagram of the proposed triode/tetrode mechanism in reverse electrolysis operation………………………………………………...................………………………...154 Figure 4.1 Schematic diagram of the experimental set-up used for the catalytic reforming of liquid CH3OH using catalyst powders……………………………………………………............167 Figure 4.2 Schematic diagram of experimental set-up used for the catalytic CH3OH reforming and the electrochemical promotion of catalytic CH3OH reforming using GDEs…………....171 Figure 4.3 Picture of the experimental set-up used for the catalytic CH3OH reforming and the electrochemical promotion of catalytic CH3OH reforming using GDEs (a) auxiliary equipment (b) glass cell configuration……………………………………………………..172 Figure 4.4 SEM image of 20 wt% metal Pt-Ru/TiO2 (WD 15.2 mm, 20.0 kV, x2.0 k)……….…….173 Figure 4.5 EDX spectra for Pt-Ru/TiO2 (a) 10 wt% metal, (b) 20 wt% metal……………………....174 Figure 4.6 TEM images of 20 wt% Pt-Ru/TiO2 (a) direct mag. 200000x and (b) direct mag. 600000x (20 wt% Pt-Ru, H2 reduction method, ball milled 60 min., HV = 100.0 kV)……………………………………………………………………………………….. …...176 Figure 4.7 X-ray diffraction data for (a) Carbon and 20 wt% Pt-Ru/C, and (b) TiO2 and 20 wt% Pt- Ru/TiO2……………………………………………………………………………...………...177 Figure 4.8 Survey XPS spectrum of 20 wt% Pt-Ru/TiO2…………………..…………………………181 Figure 4.9 Pt-Ru/TiO2 BJH desorption pore volume distribution………………………………….....184 Figure 4.10 Cyclic voltammograms of (a) Pt, (b) Pt-Ru/C on GC, and (c) Pt-Ru/TiO2 on GC (2 M CH3OH in 0.5 M H2SO4, 23 ±2oC, 50 and 75 ±1oC, 50 mV/s, no rotation).……………186 Figure 4.11 H2 production rate as a function of time (ambient to 100 ±1oC, 0.5 g catalyst, 20 wt% Pt- Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4)………………………………..193 Figure 4.12 H2 production rate as a function of time (ambient to 75 ±1oC, 0.5 g catalyst, 20 wt% Pt- Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4)………………………………..194 Figure 4.13 H2 production as a function of temperature (0.5 g catalyst, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4)……………………………………………………..195 Figure 4.14 CH3OH partial pressure as a function of time (a) Pt-Ru/C, (b) Pt-Ru/TiO2 (ambient to 75 ±1oC, 0.5 g catalyst, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4)…………………………………………………………………………………………198 Figure 4.15 H2 production rate and electrolysis cell voltage as a function of time during unpromoted and promoted catalysis (2.01 cm2, 4 mg/cm2 Pt-Ru/TiO2, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC)…………………………………………………………..202 Figure 4.16 H2 production rate and electrolysis cell voltage as a function of time during unpromoted and promoted catalysis (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC)…………….......………………………………………….203 Figure 4.17 Schematic diagram of the reverse electrolyser electrical circuit and of the proposed EPOC mechanism for the catalytic reforming of methanol………………………………205 Figure 4.18 H2 production rate and electrolyser current as a function of time during unpromoted and promoted catalysis (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC ).……..………………………………………………………………......207 Figure 4.19 H2 production rate and electrolyser current as a function of time during unpromoted and promoted catalysis (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC)………………………………………………………………………...208 Figure 4.20 Absolute Faradaic efficiency as a function of (a) average cell voltage, and (b) average xvi working electrode potential (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC)…………………………………………………………...211 Figure 4.21 Catalytic H2 production rate as a function of average working electrode potential (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC).…212 Figure 4.22 Electrophilic behaviour of the catalytic reaction rate as a function catalyst WE potential (Modified from Vayenas et al. [18] with permission from Springer)…………………….213 Figure 5.1 Example of working electrode scale-up design for triode or tetrode operation (a) single auxiliary working electrode, (b) multiple auxiliary working electrodes………………….226 Figure 5.2 Diagram showing possible electrolysis circuit current reversal under triode or tetrode operation.……………………………………………………………………………………...227 Figure B.1 Methanol electro-oxidation parallel pathways………………………………………….....251 Figure B.2 Methanol electro-oxidation bi-functional mechanism on Pt-Ru…………………………252 Figure C.1 Block diagrams of four H2 generation process schemes (a) PEM methanol electrolyser, (b) PEM water electrolyser, (c) methanol steam reforming, and (d) sodium borohydride hydrolysis.……………………………………………………………………………………..255 Figure D.1 CO chemisorption on a transition metal. Molecular orbitals and density of states before (a, b) and after (c, d). Effect of varying Φ and EF on electron back donation (c) and donation (d) (Modified from Vayenas et al. [1] with permission from Springer).………274 Figure D.2 Application of Kirchhoff’s first law for fuel cell triode operation………………………….275 Figure D.3 Application of Kirchhoff’s first law for triode electrolysis in parallel operation…………276 Figure D.4 Application of Kirchhoff’s first law for triode electrolysis in reverse operation…………276 Figure D.5 Application of Kirchhoff’s first law for tetrode electrolysis in reverse operation…….....277 Figure E.1 Electrolysis in dilute sulphuric acid…………………………………………………………283 Figure E.2 Lead-free platinization…………………………………………………………………….....284 Figure F.1 Picture of a three-port glass cell for cyclic voltammetry…………………………….……295 Figure F.2 Cyclic voltammogram (2.01 cm2, 2 mg/cm2 Pt/C, 5 mV/s, 0.5 M H2SO4, Pt flag CE, 23 ± 2oC)…………………………………………………………………………………...............298 Figure F.3 Cyclic voltammograms for a fresh catalyst sample, a used catalyst sample, and an oxidized catalyst sample (2.01 cm2, 2 mg/cm2 Pt/C, 5 mV/s, 0.5 M H2SO4, Pt flag CE, 23 ± 2oC)………………………………………………………………………………………299 Figure F.4 Cyclic voltammogram (4 mg/cm2 Pt-Ru, 5 mV/s, 0.5 M H2SO4, Pt flag CE, 23 ± 2oC).302 Figure F.5 Typical Nyquist plots for a fresh MEA at varying methanol concentration (initial frequency of 1x105 Hz to final frequency of 0.01 Hz, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte)……………............………………………………………...304 Figure F.6 MEA resistance as a function of methanol concentration for various MEA conditions (4 mg/cm2 Pt/Ru anode, 2 mg/cm2 Pt cathode, N-117, 0.5 M H2SO4 electrolyte and catholyte)…………........................................................................................................304 Figure F.7 H-cell set-up for OCV measurements……………………………………………………...306 Figure H.1 Power consumption as a function of cell voltage for different CH3OH concentrations (0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru anode, 2 mg/cm2 Pt cathode, 23 ±2oC)……………...…….324 Figure H.2 Rate of CO2 emission and volumetric rate of H2 evolution rate in function of the cell Voltage (2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru or 2 mg/cm2 Pt anode, 2 mg/cm2 Pt cathode, 23 ±2oC)….326 xvii Nomenclature i) Arabic symbols Symbol Description Units a Activity dimensionless a Tafel constant V A Absorbance dimensionless AGeom Active cell geometric area cm2 Ar Pre-exponential factor 1/s b Tafel slope V/dec B Levich parameter mA.s1/2/cm2 c Charge C C Concentration M Co Bulk concentration M Cal Calibration number dimensionless calf Calibration factor dimensionless CC Carbon tax $/tonC CCAP Capital cost $-y/GJ CE Energy cost $/GJ CF Feed cost $/GJ CH2 Hydrogen cost $/GJ COM Operation and maintenance cost $/GJ Cpo Standard heat capacity J/molK D Diffusion coefficient m/s2 e Charge of an electron C E Operating cell voltage V Eo Standard reversible cell voltage V Ea Anodic potential V EA Activation energy kJ/mol EoA Open circuit activation energy eV Eoa Standard anodic potential V Ec Cathodic potential V Eoc Standard cathodic potential V Ee Thermodynamic equilibrium (reversible) cell voltage V Eff Efficiency GJ produced/GJ feedstock Ep Peak potential V fNafion Fraction of Nafion in solution dimensionless fPTFE Fraction of PTFE in solution dimensionless F Faraday constant C/mol FCR Fixed charge rate on capital 1/y Fr Molar flow rate mmol/min. FWHM Full width half max radians i Current density mA/cm2 io Exchange current density mA/cm2 iaux, cell Auxiliary current density in triode or tetrode operation mA/cm2 ielec, cell Electrolyser current density in triode or tetrode operation mA/cm2 ik Kinetic current density mA/cm2 IL Limiting current density mA/cm2 I Total current A Io Exchange current of the metal-solid electrolyte interface A xviii Symbol Description Units Iaux Auxiliary circuit current A Iaux, cell Auxiliary current in triode or tetrode operation A Ioaux, cell Auxiliary current in normal operation A Ielec Electrolyser circuit current A Ielec, cell Electrolyser current in triode or tetrode operation A Ioelec, cell Electrolyser current in normal electrolysis A I’elec, cell Calculated electrolyser current based on measured CH3OH concentration changes in triode or tetrode operation A IFar Net Faradaic fuel consuming current A Ifc Fuel cell current A Ir Reversible current A l Path length of light in solution cm k Shape factor dimensionless kA Reaction rate constant 1/s Lcat Catalyst loading mg/cm2 LCsub Carbon sublayer loading mg/cm2 m Mass g Mi Mass rate of species i g/s n Number of electrons electrons NH+ Protonic migration flux mol/m2.s Ni Molar rate of species i mol/cm2.s NG Number of moles of catalyst on the metal surface moles of metal Ns Total number of squares dimensionless P Pressure atm Paux, cell Auxiliary power in triode or tetrode operation W PC Carbon emissions TonC/GJ Pe Electric power W Pelec, cell Electrolyser power in triode or tetrode operation W Poelec, cell Electrolyser power in normal operation W Pi Promotion index dimensionless Pp Partial pressure atm Qb Background charge C Qcv H2 adsorption charge C Qexp Total charge measured C QUDP Underpotential deposition charge C r Electrochemically promoted catalytic reaction rate mol/s ro Open circuit unpromoted catalytic reaction rate before EPOC experiment mol/s r' Open circuit unpromoted catalytic reaction rate after EPOC experiment mol/s re Electrochemical reaction rate mol/s relec, cell Electrochemically promoted electrochemical reaction rate mol/s roelec, cell Unpromoted electrochemical reaction rate mol/s rF Faradaic reaction rate mol/s reH+ Electrochemical proton reaction rate mol/s R Universal gas constant J/mol·K R Triode or tetrode voltage ratio dimensionless Re Reynolds number dimensionless Rs Ohmic resistance ohm s Number of moles of species moles Sd Standard deviation dimensionless t Time (s) T Temperature oC or K Uaux, cell Auxiliary cell voltage in triode or tetrode operation V xix Symbol Description Units Uaux, we Auxiliary WE potential in triode or tetrode operation V Uoaux, cell Auxiliary cell voltage in normal operation V Uelec, cell Electrolyser cell voltage in triode or tetrode operation V Uelec, we Electrolyser WE potential in triode or tetrode operation V Uoelec, cell Electrolyser cell voltage in normal operation V Ufc Fuel cell voltage V Urev, cell Reversible cell voltage V UWR Working electrode potential with respect to a reference electrode V UWC Working electrode potential with respect to a counter electrode V V Voltage V Vi Volumetric rate of species i m3/h Vr Volumetric flow rate ml/h W X-ray wavelength nm Wo Instrumentation broadening radians Y Molar fraction dimensionless Z Absolute impedance ohm Z’ Real portion of the impedance ohm Z” Imaginary portion of the impedance ohm ii) Greek symbols Symbol Description Units αa Anodic transfer coefficient dimensionless αH Enthalpic parameter dimensionless β Symmetry factor dimensionless ΔCp Change in heat capacity J/molK ΔG Gibbs free energy of reaction kJ/mol ΔGo Standard Gibbs free energy of reaction kJ/mol ΔGof Standard Gibbs free energy of formation kJ/mol ΔHo Standard enthalpy of reaction kJ/mol ΔHof Standard enthalpy of formation kJ/mol ΔHTf Change in enthalpy of combustion at temperature T kJ/mol ΔHj Binding energy or enthalpy of adsorption kJ/mol ΔNmol,g Number of mole of gaseous species in the products minus in the reactants mol ΔPelec, cell Change in electrolyser power output in triode or tetrode operation W ΔQD Ru oxide charge gain μC ΔQH H2 adsorption charge loss μC Δr Electrochemically induced change in catalytic reaction rate mol/s or g equivalents/s ΔSo Standard entropy of reaction kJ/molK ΔSof Standard entropy of formation kJ/molK ΔΦw Change in the work function of the working electrode eV xx Symbol Description Units ΔUWR Overpotential between the working electrode catalyst and a reference electrode V ΔVg Change in volume of gaseous components l ε Molar extinction coefficient l/mol.cm γelec, cell Electrolyser cell power gain ratio dimensionless η Overpotential V ηa Anodic overpotential V ηc Cathodic overpotential V ηconc Concentration overpotential V ηE Voltage efficiency dimensionless ηEPOC EPOC production rate efficiency dimensionless ηfuel Fuel efficiency dimensionless ηI Current efficiency dimensionless ηmax Thermodynamic efficiency % ηmax, system System thermodynamic efficiency % ηoverall Overall system efficiency dimensionless ηohm Ohmic overvoltage V ηs Charge transfer or activation overpotential V λ Fuel utilization or stoichiometry % Λ Rate enhancement factor or Faradaic efficiency dimensionless τ Mean size of the crystalline domains nm θB Bragg angle radians θi Coverage of the promoting ion on the catalyst surface dimensionless μe electrochemical potential of an electron at infinite separation from a metal KJ ρe Electrolyser rate enhancement ratio dimensionless ρelec, cell Electrolyser power enhancement ratio dimensionless ν Scan rate V/s Φ Work function eV Ψ Outer potential V ω Angular velocity 1/s xxi List of Acronyms Acronyms Definition AC Alternating current ACS American Chemical Society AES Atomic emission spectroscopy APR Aqueous phase reforming ATR Autothermal reforming BET Brunauer-Emmett-Teller BEV Buttler-Erdey-Gruz-Volmer equation BJH Barrett-Joyner-Halenda BPG Barium-praseodymium-gadolinium oxide BSE Back-scattering electron BZY Barium-zirconium-yttrium oxide C Carbon CE Counter electrode CEaux Auxiliary counter electrode CEelec Electrolyser counter electrode CFP Carbon fibre paper CNT Carbon nanotube CPO Catalytic partial oxidation CV Cyclic voltammetry D Thermal decomposition DC Direct current DI Deionised water DHE Dynamic hydrogen electrode DHRFC Direct hydrogen redox fuel cell DMFC Direct methanol fuel cell DMR Direct methanol reformer DOE Department of energy DOS Density of state EATR Electrochemical autothermal reforming ECSA Electrochemical surface area EDS Energy dispersive X-ray spectroscopy EDX Energy dispersive X-ray spectroscopy EIS Electrochemical impedance spectroscopy EMR Electrochemical membrane reactor EP Electrochemical promotion EPOC Electrochemical promotion of catalysis EPOE Electrochemical promotion of electrocatalysis ESCA Electron spectroscopy for chemical analysis F Fuel FC Fuel cell FCV Fuel cell vehicle FWHM Full width half max GB Global behaviour GC Glassy carbon GDE Gas diffusion electrode GDL Gas diffusion layer GHG Green house gas xxii Acronyms Definition HER Hydrogen evolution reaction HHV Higher heating value HP High performance HPLC High performance liquid chromatography ICE Internal combustion engine ICP Induced couple plasma ICPMS Induced couple plasma mass spectrometry ID Inner diameter IMFC Indirect methanol fuel cell IR Current resistance I-R Infra-red ITO Indium tin oxide IV Inverted volcano LCR Inductance capacitance resistance LDH Layered double hydroxide LHV Lower heating value MCFC Molten carbonate fuel cell MEA Membrane electrode assembly MEPR Membrane electrochemically promoted reactor MPL Micro-porous layer MS Mass spectrometer MSE Mercury-mercurous sulphate electrode NEMCA Non-Faradaic electrochemical modification of catalytic activity OCV Open circuit voltage OD Outer diameter OES Optical emission spectrometry ORR Oxygen reduction reaction OSR Oxidative stream reforming PBI Polybenzimidazole PE+ Purely electrophilic PE- Purely electrophobic PEM Proton exchange membrane PEMFC Proton exchange membrane fuel cell PGM Precious group metal PSA Pressure swing adsorption PTFE Polytetrafluoro-ethylene PZC Point of zero charge RDE Rotating disk electrode RE Reference electrode RMFC Reformed methanol fuel cell SA Surface area SCY Strontia-ceria-ytterbia SMSI Strong metal-support interactions SE Secondary electron SEM Scanning electron microscopy SHE Standard hydrogen electrode SOFC Solid oxide fuel cell SR Steam reforming SS Stainless steel SSE Silver-silver chloride electrode STP Standard temperature and pressure xxiii Acronyms Definition TEM Transmission-electron microscopy TMA Total platinum area TNT Titanium dioxide nanotube TPB Triple phase boundary UPD Under potential deposition UV Ultra-violet V Volcano WE Working electrode WEaux Auxiliary working electrode WEelec Electrolyser working electrode WGS Water gas shift XPS X-ray photoelectron spectroscopy XRD X-ray diffraction YSZ Yttria-stabilized zirconia xxiv Acknowledgements I wish to thank my supervisor Professor Dr. David Wilkinson for his mentorship, guidance, and support, without which this work would not have been possible. Our discussions helped my progression as a scientific researcher through the strengthening and development of skills that will last a lifetime. I am also very grateful for his encouragement to participate in various learning opportunities throughout the completion of this degree, which contributed to my development at many different levels. I would like to extend my sincere appreciation to my Ph.D. Committee members, Professors Dr. Elöd Gyenge, Dr. Walter Mérida, and Dr. David Dreisinger, for their insightful comments. Additionally, I am grateful to Dr. Ken-Ichiro Ota, from Yokohama National University, for allowing the study of the low temperature contact glow discharge of water during the summer of 2008, Dr. Phillippe Vernoux from the Université Claude Bernard Lyon 1, for hosting the 3rd international training school on the electrochemical promotion of catalysis, Dr. Elena Baranova, from the University of Ottawa, as well as Dr. Constantinos Vayenas, from Patras University, for interesting discussions on electrochemical promotion. Many thanks to the National Sciences and Engineering Research Council of Canada, the Marie Curie Action Project and the European Union, the Japan Society for the Promotion of Science, the Pacific Century, Scott Paper, the University of British Columbia, my supervisor for scholarships and research assistantships, as well as my family for financial aid in times of need. I would like to address a special thank you to Dr. Kevin Smith for allowing me to use the mass spectrometer in the Catalysis Laboratory, and to Dr. Farnaz Sotoodeh for providing assistance with it. I would like to acknowledge the technical staff at the Bio-imaging laboratories, at the Materials Engineering Department’s Electron Microscope Laboratory, at the Chemistry Department’s Structural Chemistry Facility, at the National Research Council for Fuel Cell Innovations, and at Cansci Glass Inc. I wish to thank the staff in the Chemical and Biological Engineering Department admistration office, store and workshop for their valuable help. I am very grateful to my friends and research colleagues at the Chemical and Biological Engineering Department, at the Clean Energy Research Center, as well as at xxv the National Research Council for Fuel Cell Innovations for their aid in this endeavour, and for an enjoyable and supportive work environment. I would also like to acknowledge my friends at the West Coast Symphony Orchestra and all my musically inclined friends for helping me maintain harmony and balance in my life. Last but not least, I wish to express my deepest gratitude to my father, Michel Cloutier, for his unconditional love, continuous encouragement and support over the years, and to my beloved mother, Colette C. Cloutier and grandmother, Jeanne H. Crégheur, whose passing affected me beyond words. Warm acknowledgements to the loved ones who helped me while completing this thesis, especially Rick Harkness, and to all the people who encouraged me in difficult times. Without you, I could not have done this. xxvi Dedication To whom I own all things. To my father, and to the memory of my mother (12/15/2005) and grandmother (12/12/2010). You gave me the opportunity to study science and music, you taught me to be strong, and you inspired me to achieve academic excellence. This work is the fruit of the years of support and encouragement you provided me, and for which I will never be able to thank you enough. I complete this thesis in your honour. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ À ceux et celles à qui je dois tout. À mon père, ainsi qu'à la mémoire de ma mère (12/15/2005) et de ma grand-mère (12/12/2010). Vous m'avez incité à poursuivre mes études en sciences et en musique, vous m'avez appris à être forte et persévérante, et vous avez été mon inspiration dans la poursuite de l'excellence et de la réussite. Ce travail est le fruit de votre soutien et de vos encouragements. Je vous en suis infiniment reconnaissante. Je vous rends hommage en complétant cette thèse. 1 Chapter 1: Introduction 1.1 Background 1.1.1 Alternative Fuels The increased global energy demand and environmental challenges we are facing drive efforts in clean energy research. A peak in conventional non-renewable oil production before 2030 appears imminent [1]. The burning of fossil fuels generates green house gases (GHG), which are considered to be the major contribution to climate change. The escalating need to reduce our fossil fuel dependency for increased energy security and to reduce GHG emissions for a cleaner environment, elevate pure hydrogen (H2) as a promising energy fuel for the future. Switching to H2 as an energy carrier for energy production could alleviate many environmental concerns related to the combustion of fossil fuels [2]. Fuel cells (FCs) are electrochemical devices capable of continuously converting the chemical energy of a fuel into electricity. Hydrogen and fuel cells could provide an efficient sustainable alternative to fossil-fuel based technologies. Hydrogen can be stored or readily used in various applications. It can serve as a chemical reagent in hydrogenation processes and in the production of fertilizers, it can be directly combusted in an internal combustion engine (ICE), or used to feed fuel cells. Fuel cell systems can achieve 2 to 3 times greater overall energy efficiencies than conventional gasoline ICE when fuel production is not taken into account [3]. There is no doubt that fuel processing plays an important role in fuel cell development. The strong interest in H2 has led to the search for various methods to deliver it in a convenient way. However, the issues of H2 production, distribution, compression, safety, and public acceptance, limit the large scale adoption of H2 as an energy carrier. The scientific development of practical fuelling technologies is required to close these existing gaps and facilitate the introduction of a H2 economy. Although H2 is the most abundant element on earth, it is not available in its elemental form in sufficient amounts, thus, it needs to be extracted from other sources. It can be produced from primary fossil energy sources which, even if combined with efficient carbon sequestration, emit various pollutants. 2 In addition, the anticipated decrease in oil reserves will result in crude oil price increases. Therefore, ways to produce H2 from non-fossil sources need to be developed. Until then, H2 applications will mostly depend on hydrocarbon fuel processing for H2 generation. Alternative fuels, such as liquid alcohols, could serve as a bridge between gasoline and gaseous H2. These alternative fuels possess high energy densities and can be easily stored and distributed through the existing gasoline infrastructure. It has also been argued that alcohols are the next liquid fuels to use after the depletion of petroleum resources [4]. Methanol (CH3OH) is a readily-available electroactive alcohol, which can be economically mass-produced catalytically from non-renewable resources, such as natural gas, and coal, and from renewable resources, such as gasified biomass. It may also be electro- catalytically generated from CO or CO2 [5, 6]. Table 1.1 summarizes some important properties of CH3OH. Table 1.1: Methanol properties. Properties Methanol Energy density [kWh/kg]/[kWh/l] 6.4/4.6 Boiling point [oC] 64.7 Freezing point [oC] -97 Flash point [oC] 11 Toxicity – Oral [mg/kg] 14.3 In addition, methanol possesses a higher hydrogen to carbon ratio than gasoline (octane, C8H18), i.e., 4 vs. 2.25, respectively. Although CH3OH can be used directly in a fuel cell or combustion process to generate electricity, it will be explained that its electrolysis could be an efficient and economical H2 production method to ease H2 transportation and storage issues with minimum infrastructure changes. 3 1.1.2 Technology Gap There exist four categories of H2 production technologies: biological, chemical, electrochemical and thermal [7]. Most H2 is currently produced by the well-established catalytic steam reforming (SR) process of hydrocarbon based fuels, which requires temperatures ranging from 250 to 1545oC, depending on the fuel [8, 9]. Reforming can be carried out externally, integrated with a fuel cell, or done directly inside a fuel cell such as a molten-carbonate fuel cell (MCFC, 600–700oC) or a solid oxide fuel cell (SOFC, 600–1000oC). On-board gasoline SR was originally believed to be the best way of generating H2 for transportation, but it was determined that it did not offer clear advantages over other available technologies, such as gasoline ICE and battery hybrids [10]. The practical issues of durability, size and weight, resistance to vibration, cold-start, transient response, and H2 purity concerns, complicates the application of SR for H2 generation in transportation. Combining SR with low temperature H2 proton exchange membrane (PEMFC) technology is also restricted by the low tolerance of the H2 PEMFC anode catalyst to the residual CO present in the reformed fuel feed. 1.1.3 Proposed solution As described later in this thesis, there may be various applications for a low temperature H2 production processes. One of them is the production of H2 for direct combustion or for a PEMFC inside a vehicle. An approach to achieving the ultimate performance targets, including start-up time and energy, may be to conduct on-board reforming at temperatures lower than that of conventional steam reforming (SR) processes. Electrocatalysis can be used to extract H2 from CH3OH at low temperatures (20-80oC) [11- 12]. Methanol catalytic reforming can also be conducted in the liquid phase at lower temperatures (72– 200oC) [13–16]. It will be seen that, using new electrochemical approaches, the electrocatalytic or catalytic CH3OH reforming process may result in reduced energy consumption over comparable existing technologies. To the author’s knowledge, no studies have verified if these low temperature alcohol electrocatalytic or catalytic reforming processes could meet the on-board fuel processing targets and if their adoption could facilitate the implementation of a H2 economy. This thesis focuses on the catalytic and electrocatalytic oxidation of CH3OH at low temperatures, i.e., below 200oC. 4 1.1.4 A sustainable closed loop system Low temperature CH3OH oxidation technologies may meet the short-to-medium term demand for H2, until it can be effectively produced without releasing significant CO2. Figure 1.1 (a) shows an example of how conventional methanol synthesis processes can be combined with a direct methanol reformer (DMR) and a PEMFC. For example, without going into the details of the reaction mechanism, syngas can be produced by methane steam reforming (SR): CH4 + H2O ↔ CO + 3H2 (1.1) The water-gas shift (WGS) reaction is used to adjust the CO to H2 ratio for the synthesis of CH3OH. CO + H2O ↔ CO2 + H2 (1.2) Methanol is synthesized from the highly selective conversion of the synthesis gas generated. CO + 2H2 ↔ CH3OH (1.3) However, as CO is used in the CH3OH synthesis, the WGS reaction reverses, producing more CO. Therefore, the overall reactions, which produces CH3OH from syngas can be summarised in the following overall reaction: CO2 + CO + 5H2 → 2CH3OH + H2O (1.4) Biomass gasification can also be used to produce CH3OH. For example, wood biomass can be gasified to syngas, which can in turn be synthesized to CH3OH. 2C16H23O11 + 19H2O + O2 → 42H2 + 21CO + 11CO2 → 21CH3OH + 11CO2 (1.5) 5 O2 PEM FC H2 Electricity (a) DMR CH3OH CO2 H2O Stream reforming, or gasification HeatSteam PEM FC DMR Electro- synthesis O2 H2 CH3OH Solar Electricity CO2 H2O CO2, H2O CO2 (b) Natural gas or biomass H2O, CO2 In this case the process is carbon neutral as the CO2 generated is used to produce biomass by photosynthesis. Ideally, the system would exclusively depend on renewable resources: the methanol could be originating from biomass, while the electricity required for direct current (DC) voltage supply could originate from solar, wind, hydro, tidal, or geothermal energy. Figure 1.1 (b) demonstrates how, in the future, the system could be made sustainable when paired with solar energy and an electrochemical synthesis process. The CH3OH would be synthesized from CO2, generated by the DMR, i.e., methanol electrolyser, and H2O, generated by a PEMFC. The current efficiency for the electro-synthesis of CH3OH using RuO2/TiO2 nanotube composite electrodes was reported to be about 60% [17]. It may appear to be simpler to produce H2 and O2 directly from water using solar energy, however, the scheme presented in Fig. 1.1 could also use CO2 from external sources (dashed line), such as fossil fuel combustion, and contribute to the reduction of GHG from other existing processes. Figure 1.1: (a) Schematic diagram of conventional methanol synthesis combined with a DMR and a PEMFC (b) Schematic diagram of a future sustainable system comprising a DMR and PEMFC. 6 In this thesis, it will be seen that electrochemical promotion can be applied to improve the current or voltage efficiency of electrochemical processes. Over the long term, this research may benefit the creation of a sustainable energy system for the production of H2 from renewable methanol originating from biomass. Over the short term, the findings of this research are more likely to impact the development of transitional low temperature CH3OH based H2 generation technologies, utilizing energy sources currently available. 1.2 Thesis overview 1.2.1 Justification Ideally, the CH3OH and water catalytic or electrocatalytic H2 production process should possess a higher efficiency than other currently available technologies. The non-Faradaic enhancement of catalytic activity (NEMCA), also referred to as the electrochemical promotion (EP) of catalysis (EPOC), was identified as a novel approach to improve the overall performance of catalytic chemical reactions. This phenomenon has been shown to enhance the rate of a variety of catalytic reactions [18]. A new EP approach, known as triode operation, has been shown to affect the overpotential of electrodes in fuel cells [19]. This approach is analogous to EPOC, except that an electrochemical reaction is promoted instead of a catalytic one. We denote it as the electrochemical promotion of electrocatalysis (EPOE). In this thesis, both of these approaches, i.e., EPOC of the catalytic reforming of CH3OH, and EPOE of the electrocatalytic reforming of CH3OH and water, were investigated at low temperature. The application of EP, through the synergy of solid-state electrochemistry and catalysis, could be a key to efficient fuel processing technologies for H2 production. As EP is a fairly new topic, many cutting-edge research opportunities still remain. This thesis provides valuable information on the effect of EP on the electrocatalytic and catalytic reforming of CH3OH and water at low temperature, i.e., ambient to 75oC, and gives insights for further improvements. Applying this new electrochemical engineering technique might eventually lead to the development of a ground-breaking H2 production technology, which could have a tremendous market impact on the energy and transportation sectors, as well as on the environment. 7 1.2.2 Research objectives The main objective of this research was to improve the state-of-the-art methanol and water electrochemical and chemical reforming processes by rendering them more thermodynamically and kinetically attractive. Investigating the thermodynamic and kinetic characteristics of H2 production via catalytic and electrocatalytic methanol liquid phase reforming involved the following goals: 1) Evaluate the influence of different operating conditions, such as CH3OH concentration (0 to 16 M), and low temperatures (25 to 75oC), as well as the performance of different electro-catalysts (Pt, Pt- Ru) on the electrochemical reforming of CH3OH in acidic media; 2) Determine the effect of EPOE on the electrolysis of CH3OH and water using different triode and tetrode configurations and materials in the galvanostatic and potentiostatic modes; 3) Synthesize, characterize, and evaluate the catalytic and electrocatalytic capabilities of Pt-Ru/TiO2; 4) Determine the effect of EPOC on the low temperature catalytic reforming of CH3OH in acidic media. This applied research project comprises several innovative contributions, which, to the author’s knowledge, have not been reported in the literature to date. For example, there have been no scientific papers on: 1) The passive electrochemical reforming or electrolysis of CH3OH and water in the liquid phase; 2) The use of the triode and tetrode configuration in the electrolysis mode; 3) The application of triode and tetrode operation in the potentiostatic control; 4) The operation of a tetrode in the mixed galvanostatic and potentiostatic control; 5) The low temperature catalytic reforming of CH3OH and water using Pt-Ru/C; 6) The application of EP on the low temperature catalytic reforming of CH3OH and water using proton- conducting electrolytes. This thesis provides a better understanding of the chemical and electrochemical CH3OH reforming processes at low temperature, as well as valuable information on possible ways to improve their effectiveness, particularly through electrochemical promotion. 8 1.2.3 Layout This study was achieved by drawing from knowledge in electrochemical, chemical, and material engineering. Six key topics were identified: catalytic reforming, electrochemical membrane reactors, H2 pumping, direct fuel cells, proton-conducting electrolytes, and electrochemical promotion. Only the most essential information is discussed in the literature review, which completes Chapter 1 and starts in Section 1.3. Chapter 2 constitutes a baseline study of the electrochemical reforming of methanol and water using a proton exchange membrane (PEM) electrolyser and extends the existing knowledge of this system, which is presently limited as will be seen in Section 1.3.3. A theoretical thermodynamic evaluation demonstrating that the electrochemical reforming systems could become thermodynamically favourable under certain operating conditions is presented. The investigation of the kinetic aspect of alcohol electrochemical reforming systems was valuable in determining if the electrochemical reaction occurs at a rate of practical interest. The characteristics of the alcohol electrochemical reformer were determined under various conditions and using different membrane electrode assembly (MEA) compositions. The content of this Chapter is based on a publication and is reprinted from the International Journal of H2 Energy, 35, C. R. Cloutier, and D. P. Wilkinson, “Electrolytic Production of Hydrogen from Aqueous Acidic Methanol Solutions”, 3967-3984, Copyright (2010), with permission from Elsevier. Chapter 3 describes the electrochemical promotion of electrocatalysis in triode and tetrode operation in the galvanostatic and potentiostatic mode. This new application was developed based on fuel cell triode operation knowledge presented in Section 1.3.5. The effects of electrochemical promotion on the electrolysis of methanol and water are evaluated under various conditions and different design configurations. The results obtained are used to compare the energetics of triode and tetrode electrolysis versus conventional electrolysis. It will be demonstrated that the application of electrochemical promotion on the electrochemical reforming of alcohols can lead to the elimination of the need for anodic noble metal catalysts. Part of the work presented in this Chapter was published and reprinted from the ECS Transactions, 25 (23), C. R. Cloutier, and D. P. Wilkinson, "Triode Operation of a 9 Proton Exchange Membrane (PEM) Electrolyser", 47-57, Copyright (2010), with permission from the Electrochemical Society. Another more complete article was also recently submitted for publication. Chapter 4 contains a baseline low temperature chemical reforming study. The electrochemical promotion of low temperature liquid phase catalytic reforming of methanol and water requires that the chosen catalyst, in addition to being a chemical catalyst, is also an electrocatalyst for the methanol and water electrochemical reforming reaction. To achieve this requirement, Pt-Ru/TiO2 was synthesized, characterized and tested for its catalytic reforming capabilities. Also in this Chapter, the effect of EP on the catalytic reforming of CH3OH and water was evaluated using the commercial Pt-Ru/C CH3OH electro-oxidation catalyst and the synthesized Pt-Ru/TiO2. The results obtained are used to compare the energetics and viability of the electrochemically enhanced low temperature CH3OH and water reforming systems versus conventional low temperature catalytic reforming, CH3OH and water PEM electrolysis. The content of Chapter 4 is the basis for a journal submission currently in preparation. Chapter 5 summarizes the research outcomes of Chapters 2 to 4. Possible applications are described and the significance and impact of the work is discussed. Finally, recommendations for future work are given. Appendices A to J contain detailed information which is supporting but not essential to the thesis. Appendix A contains a list of publications, presentations, and posters which resulted from this research. The mechanism for the electro-oxidation of methanol on Pt and Pt-Ru is shown in Appendix B. Appendix C contains an efficiency and economic comparison of H2 production methods. Some EPOC rules and triode equation derivations are provided in Appendix D. Appendix E contains experimental procedures used to carry out this work while Appendix F contains background information on the various electrochemical techniques employed. Appendices G and H contain thermodynamic and electrochemical data, and sample calculations. 10 1.3 Literature Review 1.3.1 Heterogeneous catalysis In heterogeneous catalysis, the reactant(s) phase differs from that of the catalyst. Most of the time, the reactants are in the liquid or gaseous phase, and the catalyst is in the solid phase. The main stages of heterogeneous catalysis includes (1) the diffusion of the reactant(s) to the catalyst surface (2) the adsorption of the reactants on the catalyst active site through the formation of chemical bonds, (3) after chemical reaction, desorption of the products formed from the catalyst surface, and (4) product(s) diffusion away from the catalyst surface. Heterogeneous catalysis is often the most important step in the synthesis of petrochemicals, pharmaceuticals, chemicals, and in environmental applications. A catalyst is a substance which is not consumed in a chemical reaction, but accelerates its rate. In heterogeneous catalysis, the diffusion and adsorption of reactants to the catalyst surface is crucial for the chemical reaction to occur. Catalysts can reduce the activation energy required for a chemical reaction to take place. Most metal catalysts used in heterogeneous catalysis are transition metals. Their electronic structure impacts the interactions between the reactants and the catalyst as electronically unsaturated metal atoms allow electrons to be accepted by available d orbitals. Most hydrogen is produced by the heterogeneous catalytic reforming of fossil fuels [20]. Three main vapour phase production paths exist for the generation of H2 from methanol: thermal decomposition (D), steam reforming (SR), and catalytic partial oxidation (CPO). Autothermal reforming (ATR), also referred to as oxidative stream reforming (OSR), combines the SR with the CPO reactions. Of these processes, only SR is discussed, as it produces the highest H2 concentration. Methanol can be converted to H2 at lower temperatures than other hydrocarbons, making it possible to catalytically reform it by aqueous phase reforming (APR) or liquid phase reforming. 11 1.3.1.1 Steam reforming The catalytic SR of CH3OH is an endothermic reaction, which is conducted at temperatures ranging from 250 to 350oC [6]. Heat is needed to attain a reasonable CH3OH conversion. The SR process starts with the splitting of CH3OH into CO and H2, followed by the exothermic water gas shift (WGS) reaction (Eq. 1.9), as per the following chemical reactions, which do not describe the details of the reaction mechanism: Step 1: CH3OH ↔ CO + 2H2 (1.6) Step 2 (WGS): H2O ↔ 1/2O2 + H2 (1.7) CO + 1/2O2 ↔ CO2 (1.8) Overall Step 2: CO + H2O ↔ CO2 + H2 (1.9) Overall: CH3OH + H2O ↔ CO2 + 3H2 (1.10) The enthalpy of the overall reaction at standard conditions is ΔHo = 49.2 kJ/mol. It will be later explained that the overall CH3OH chemical reforming reaction is the same as the overall electrochemical reaction for the electrolysis of CH3OH in a proton exchange membrane (PEM) electrolyser. Catalysts commonly used for the steam reforming of CH3OH are composed of copper supported on zinc oxides, such as CuO/ZnO and CuO/ZnO/Al2O3, in which Al2O3 can be added for thermal stability [21]. Different Cu-Al catalysts of varying Cu contents have been investigated for the production of H2 from CH3OH. A maximum H2 production efficiency of 78 mol% was obtained for the SR reaction at 250oC using a catalyst containing 27.8 wt% Cu and calcined at 700oC [22]. Since the WGS reaction (Eq. 1.9) is reversible, CO can be generated. The activity of layered double hydroxide (LDH) catalysts containing various metal combinations was evaluated for the SR of CH3OH at 150–400 °C and atmospheric pressure [23]. Again, the most favourable LDH for the SR of CH3OH was Cu-Al based. A WO3/CeO2/YSZ nanocomposite catalyst was used for the reforming of CH3OH, with and without H2O, at temperatures ranging from 100 to 300oC [24]. In both cases, the only product was CO2, which formed at 12 room temperature when H2O was present. The SR CO2 emissions need to be suppressed by concentrating and sequestering CO2, which results in energy losses. 1.3.1.2 Liquid phase reforming The terms aqueous phase reforming (APR) are used to designate the reforming of aqueous solutions composed of direct biomass products, such as methanol, ethylene glycol, glycerol, sugars and sugar alcohols (sorbitol, etc.). Biomass is a renewable carbon-containing fuel, which can be obtained from a variety of sources such as animal, agricultural, and municipal wastes. In this process, a feed of oxygenated hydrocarbons, having a limited volatility and a C:O ratio of 1:1, is catalytically converted with liquid H2O over a heterogeneous catalyst at temperatures around 230oC to produce H2 and CO2 [25]. This single-step catalytic process is achievable at conditions where the WGS reaction is favourable, making it possible to generate H2 with low amounts of CO. The APR process consumes less energy than the conventional SR process, as it does not require the vaporization of the oxygenated hydrocarbon and water feeds. Platinum metal has a high activity for the dehydrogenation of CH3OH, so typically, Pt/Al2O3 is used for APR, while other promising materials include Pt and Ni-Sn based catalysts [26-28]. Bimetallic catalysts were shown to result in higher activities for the APR reaction than monometallic catalysts [25]. It was noted that the oxygenated hydrocarbons have various aqueous phase reforming reaction pathways, resulting in catalyst selectivity challenges. The process selectivity depends on many factors, including the catalytic metal, the support, the solution pH, the feed and the operating conditions [25]. The CO species adsorbed at the catalyst surface result in low catalytic activity and are partially removed by the WGS to form CO2. The development of heterogeneous catalysts comprising metals, metal alloys, support and WGS reaction promoters, which are stable under APR conditions, is still under way. There are only a limited number of studies focusing on the aqueous phase catalytic reforming of CH3OH, which do not originate from biomass sources. Hence, these processes are not referred too as APR in the literature. This liquid phase reforming reaction was first studied on a Ru/C catalyst with external heating at 100-200oC [13]. Compared to Pt-based catalysts, commercial Cu/ZnO reforming catalysts were found to have no activity under liquid phase reaction conditions, although their activity is high for 13 the steam reforming reaction. Silica-supported Pt-Ru catalysts were also investigated at 77 to 84oC at ambient pressure, and CH3OH dehydrogenation was determined to be the rate limiting step of the reaction [14]. The formation of CO2 over Pt-Ru/SiO2 did not proceed via HCHO decomposition, and partly-dehydrogenated CH3OH (CH2OH*) was determined to be the initial reaction intermediate, from which H2 and CO2 were formed through HCOOCH3 and HCOOH as successive reaction intermediates. The catalytic activity and selectivity towards CO2 increased with basic oxide catalyst supports and decreased with acidic catalyst supports [29]. Platinum was most active for liquid phase H2 production reforming when supported on TiO2. In addition, Pt-Ru/SiO2 resulted in an even greater activity than Pt/SiO2. The addition of Ru to Pt/SiO2 accelerated the reaction of the formaldehyde (H2CO) intermediate to form methyl formate (HCOOCH3), as the product formation rate was not restricted in the presence of CO. The dehydrogenation of CH3OH was reported as the rate determining reaction. Ir-Re/SiO2 catalysts were also studied at 105oC and resulted in an activity comparable to Pt-Ru/SiO2 [15]. The highest activity for the catalytic reforming of liquid CH3OH was obtained with Pt-Ru/TiO2 between 77 and 84oC [16]. It proceeded through the partially dehydrogenated HCOOCH3 and HCOOH intermediates through a mechanism similar to that obtained using Pt-Ru/SiO2. 1.3.1.3 Hydrogen spillover in heterogeneous catalysis Spillover is defined as a phenomenon in which species activated on one phase are transported, usually across a surface, to another phase where they may then participate directly or indirectly in catalysis [30]. The second surface should not normally adsorb or form the active species which are adsorbed or formed on the first surface under the same conditions. The species spilled on the second surface may (1) diffuse on the surface, (2) diffuse or react in the bulk, (3) react on or with the surface, (4) create sites capable of adsorption or catalysis [30]. Spillover can extend from the initial accepting surface to adjacent surfaces in direct or indirect contact with the source of spillover, or the surface that initially accepted the spillover species. This means that spillover species can be transferred over long distances, i.e., over millimetre and centimetre distances [30]. Spillover constitutes an important mechanistic step which can occur during adsorption in heterogeneous catalysis and was found to be involved in many surface reactions and transport phenomena. 14 H2 spillover from a metal to an oxide or carbon surface was extensively studied because most heterogeneous catalysts comprise metal particles supported on high surface area carbon or oxides, and because H2 is involved in many catalytic reactions. It is the fastest spillover process and may occur at room temperature [30]. H2 spillover has been demonstrated on several supported metals, which are known to adsorb H2 dissociatively. A model assuming the dynamic equilibrium of two coexisting activated H2 spillover species (H atoms and H+ ions) was proposed to explain the synergy between the components of bifunctional catalysts [31]. It was concluded that the nature of the activated hydrogen species depended on the physico-chemical properties of the catalyst [32]. For the promotion of acidic catalysis, the spillover species is proposed to be part of the active site as H+ [28]. The principles of H2 spillover have been applied to the design of highly selective and active catalysts from physical mixtures [33]. The spillover of H2 over macroscopic distances was exploited in the design of a dual-bed reactor comprising two separate zones of metal and bimetallic catalysts. Using this design, activities 2.7 times greater than that of the noble metal alone were achieved, demonstrating that a catalytic reaction may occur at different reactive sites as the activating species moved via surface diffusion [34]. Indirect catalyzed hydrogenation was shown to occur via a mechanism in which H2 is activated on a metal catalyst, desorbed in an activated form, which may then react with a second reactant not in contact with the catalyst [35]. The deactivation kinetics of H2 spillover in the gas phase was found to be a first-order reaction, which is accelerated by glass surfaces [36]. The effect of H2 spillover through the gas phase on the hydrogenation of methane was studied on graphite and activated carbon using various independent methods [37]. Studies of the H2 spillover to the carbon support of metal catalysts demonstrated that the uptake and spillover of H2 was greater for Ru containing catalysts than for Pt/C catalysts [38]. 1.3.1.4 Catalyst local heating Catalysts can be thermally heated locally by using heating elements or by the passage of an electric current. The selective oxidation of CH3OH was investigated at temperatures ranging from 100 to 350oC on indium tin oxide (ITO). The study was carried out by (1) thermally heating the catalyst with a ceramic heating element, (2) by heating the catalyst though ohmic heating due to the passage of an electric 15 current, and (3) by thermal heating followed by ohmic heating [39]. It was found that the electrically activated and thermally heated catalyst was much more active than the non-electrically activated but thermally heated catalyst. The electrically activated catalyst was active even after the current was cut-off. This phenomenon was thought to be analogous to the EPOC effect, which is described in more details in Section 1.3.4. However, it was determined that the mechanism was different in this case, as reversing the catalyst electrical leads did not affect the reaction kinetics or selectivity, and the catalyst was still active after the current was cut-off, and the reaction took place on ITO deposited on a porous insulating surface and not on a metal deposited on a solid ion conducting electrolyte. An electrically heatable device using electrically conductive non-metallic materials [40] and a process for the electrically activated transformation of chemical and material compositions have been proposed [41]. Electrically activated catalysis was also studied for the steam reforming of methane at 600oC [42]. A conversion of 76% H2 was obtained upon passing a current through the catalyst, while a conversion of 66% was obtained upon thermal heating only. More detailed investigations are required to elucidate the reasons behind the catalyst activity enhancement observed. 1.3.2 Proton exchange membrane fuel cell The electrochemical operating principles of FCs are similar to those of batteries as neither requires combustion. However, fuel cells can effectively convert chemical energy into electrical energy as long as a fuel is supplied, so they do not require recharging like batteries. They are quiet and flexible devices, which convert a fuel to electricity and heat through a spontaneous electrochemical reaction. H2 fuel cells were initially developed for the space program for providing electricity and drinking water to astronauts. As they can produce electrical energy efficiently and can generate less emissions than combustion engines, various types of FCs are being considered for many applications. Most FC research is focused on the development of direct H2 and direct liquid fuel cells. Only low temperature proton exchange (or polymer electrolyte) membrane (PEM) fuel cells are discussed in this thesis. 16 1.3.2.1 Membrane electrode assembly The membrane electrode assembly (MEA) forms the core of a PEMFC. It is usually inserted between flow field plates, which allow current flow and reactant and product distribution. The MEA comprises a PEM, which separates an anodic electrode from a cathodic electrode. The most common PEM is a perfluorinated ion exchange membrane from DuPont, which is referred to as Nafion®. The mobile ion of this sulphonated fluoro-polymer is H+. This acidic polymer is an excellent gas separator and electrical insulator, and has good chemical resistance and mechanical strength. However, it dehydrates at temperatures exceeding 100oC and loses its proton conductivity. Therefore, Nafion® membranes must be kept humidified to maintain their high proton conductivity. The presence of cationic contaminants also affects the mechanical properties and the protonic conductivity of Nafion® [43, 44]. Due to these shortcomings, work has been done on the development of other proton conducting electrolyte membranes for PEMFC applications, such as ceramics [45-50], composite [51-54], other polymers [55- 57], and metal coated polymers [58, 59]. The anode and cathode of fuel cells are usually composed of electroactive Pt group metal (PGM) based catalysts. As platinum has a high cost and a low abundance, precious metal nanoparticles are often dispersed or supported on larger carbon particles to reduce the amount of Pt needed while increasing the electrochemically active surface area (ECSA) of the catalyst. ECSA losses deteriorate the PEMFC performance. Improving the catalyst activity, utilization and stability are important topics in PEMFC research. The main approaches taken to reduce or replace Pt while maintaining PEMFC performance and efficiency are to reduce the noble metal loading or use less expensive noble metals [60], or explore using non-Pt electrocatalysts [61, 62]. Changing the carbon catalyst support is also looked at as it may affect the catalyst activity and lead to the partial reduction of the Pt loading [63]. A triple phase boundary (TPB) between the electrolyte, reactants and catalyst is required for an electrochemical reaction to occur. Different methods can be used to create this three-dimensional electrochemical reaction zone. The most common is the application of an ink comprised of a polymer electrolyte solution and electrocatalytic particles. Thin films of catalyst inks usually contain an electrolyte ionomer binder (typically Nafion®), and supported or unsupported nanocatalyst Pt particles. A three 17 phase contact between the H2, protons and solid catalyst is obtained by applying this electronically connected ink coating onto a porous conducting carbon substrate or a membrane. The three main methods consist of applying an ink onto (1) a Teflon film which is transferred by decal to the membrane through a hot pressing process [64] (2) a porous conducting carbon substrate (gas diffusion layer, GDL) to form a gas diffusion electrode (GDE) [65], or (3) onto each side of the membrane to form a catalyst- coated membrane (CCM) [66]. 1.3.2.2 Overvoltage In fuel cells, the operating voltage, E (V), is always greater than the standard reversible cell voltage, Eo (V): nF G E o o Δ−= (1.11) where ΔGo (kJ/mol) is the change in Gibbs free energy of reaction, n is the number of electrons participating in the reaction, and F is Faraday’s constant (96485 C/mol). When not operating at standard conditions, the thermodynamic equilibrium (reversible) cell voltage at non-standard temperature and pressure, Ee (V), can be estimated with the Nernst equation: ⎟⎟ ⎟ ⎠ ⎞ ⎜⎜ ⎜ ⎝ ⎛ Π Π −= jO Jr s jOxj s jdjo e a a Log nF RTEE , , , ,Re303.2 (1.12) where R is the ideal gas constant (8.314 J/molK), T is the temperature (K), and a is the activity (dimensionless). The excess amount of electrical energy required for the electrocatalytic process to occur is known as overvoltage, η (V). It is the difference between the operating cell voltage, E, and the equilibrium cell voltage, Ee: 18 eEE −=η (1.13) The overvoltage represents the sum of the cell efficiency losses which are mainly due to the activation overvoltage or charge transfer overvoltage, ηs, the concentration overvoltage, ηconc, and the ohmic overvoltage, ηohm. ohmconcseEE ηηη −−−= (1.14) The activation overvoltage and the concentration overvoltage in Eq. 1.14 can be separated in their anodic and cathodic components. If so, the absolute values in Eq. 1.14 are then removed, and the anodic component is subtracted as it is a positive value, while the cathodic component is added, as it is a negative value. The activation overvoltage is related to the activity of the electrode surface which affects the charge transfer process for a particular electrochemical reaction. The concentration overvoltage is linked to the change in the mass transport of reactants to the electrode surface as the electrochemical reaction proceeds. The ohmic resistance is caused by the PEM/electrocatalyst interface resistance, as well as the bulk resistance and the ionic conductor’s resistance to the transport of electrons. Other losses include fuel crossover and internal currents [67]. The power output and thermodynamic efficiency of electrochemical systems depend on the minimization of overpotential losses at the anode and/or cathode. Significant overpotentials will result in inefficient electrocatalysis and restrict the commercialization of electrochemical devices as they are often not sufficiently cost-effective. For this reason, research efforts are concentrated towards the minimization of overvoltage losses, which is crucial to the development of effective electrocatalytic processes. 1.3.2.3 Hydrogen proton exchange membrane fuel cell The H2 PEMFC uses H2 as the fuel and O2 as the oxidant. The oxidation occurs at the anode and the reduction occurs at the cathode. The electrons pass through an external electrical circuit while the protons diffuse through the membrane. The overall reaction is the sum of the anodic and cathodic reactions. The PEMFC electrochemical reactions are as follows: 19 Anode half-cell reaction: H2 → 2H+ + 2e- Eoa = 0 V vs. SHE (1.15) Cathode half-cell reaction: 1/2O2 + 2H+ + 2e- → H2O Eoc = 1.23 V vs. SHE (1.16) Overall reaction: H2 + 1/2O2 → H2O Eo = 1.23 V (1.17) Note that throughout this thesis, the convention of reversing the Eoa values reported for the half-cell reactions written as electro-oxidation was adopted. The subscript “a” designate the standard potential of the half-cell reaction occurring on the anode, while the subscript “c” designate the standard potential of the half-cell reaction occurring on the cathode. The standard reversible cell voltage, Eo, based on thermodynamics is 1.23 V. The rate limiting process is the cathode oxygen reduction reaction (ORR), which suffers from kinetic limitations. For this reason, a larger quantity of catalyst is needed for the slow ORR cathodic reaction compared than for the fast H2 anodic reaction. The H2 PEMFC has significant water management issues, which need to be continuously addressed to reach stable operation, and for this reason, their operation is usually limited to temperatures around 60-90oC. Hydrogen purity requirements The PEMFC electrodes usually contain a Pt catalyst, which promotes the reactions occurring at the electrodes. The electrocatalytic properties of Pt are greatly affected by contaminants, which adsorb to the surface of the catalyst and prevent H2 adsorption onto the catalyst surface [68]. At temperatures below 150oC, CO strongly adsorbs onto Pt and poisons the H2 reaction. This CO poisoning effect occurs at concentrations of CO as low as 5-100 ppm (0.0005-0.01%) in the H2 inlet stream, depending on the catalyst loading [69]. A tolerance to CO levels of 100 ppm or more would greatly simplify the PEMFC 20 fuelling system when H2 is produced by catalytic reforming processes. Typically, the H2 fuel stream needs to contain less than 10 ppm of CO in order to be fed directly to a PEMFC. Different options exist to overcome electrocatalyst CO poisoning in the PEMFC [70]. These include increasing the cell operating temperature, which weakens the CO bond to the Pt active sites and reduces the catalyst’s sensitivity to CO poisoning. It is also possible to remove CO from the reformate using advanced fuelling system design or membranes for CO separation. Preferential oxidation reactors or H2 separation membranes containing Pd-based alloys or proton-conducting ceramics can be used to remove CO. For example, Pt-alloy catalysts can preferentially oxidize CO and selectively remove it [71]. Other methods are the introduction of oxidant in the H2 fuel feed stream, referred to as O2 bleeding, or the use of cyclic operation to clean the surface from adsorbed species. 1.3.2.4 Direct methanol fuel cell The direct methanol fuel cell (DMFC) is the most advanced type of direct liquid fuel cell. In a DMFC, a diluted aqueous solution of CH3OH is fuelled directly to the anode of a PEM fuel cell, while air or O2 is supplied to the cathode, as per the following electrochemical reactions: Anode half-cell reaction: CH3OH + H2O → CO2 + 6H+ + 6e- Eoa = -0.016 V vs. SHE (1.18) Cathode half-cell reaction: 3/2 O2 + 6H+ + 6e- → 3H2O Eoc = 1.23 V vs. SHE (1.19) Overall reaction: CH3OH + 3/2O2 → 2H2O + CO2 Eo = 1.21 V (1.20) 21 The DMFC has a standard reversible cell voltage of 1.21 V. The MEA arrangement for the DMFC is similar to that for the H2 PEMFC. When liquid CH3OH mixtures are used as the fuel, the membrane can maintain high humidification levels, but cross-over of the liquid CH3OH can be a problem. While the cathode reaction of the PEMFC suffers from poor kinetics, in the DMFC, the anode and the cathode reactions both result in kinetic losses. Thus, the two main issues which restrict the performance of DMFCs are the poor anodic CH3OH oxidation kinetics, and CH3OH crossover. Anodic methanol oxidation The slow CH3OH oxidation kinetics at the electrode/electrolyte interface in acidic media limits the electrochemical reaction rate. The high activation overpotentials observed are mostly caused by the complex anodic electrochemical reaction mechanism [4]. Possible CH3OH electro-oxidation pathways are illustrated in Appendix B. The direct electrochemical oxidation of CH3OH on Pt occurs in multiple stages, which include various adsorbed intermediates. Strongly adsorbed CO remains un-oxidized at the surface and blocks the electrode active sites from adsorbing reaction intermediates which are formed during the oxidation process [4, 72]. The poisoning species which deactivates the electrode surface is CO. It is the rate limiting component for the electro-oxidation of CH3OH. Typical methods used to prevent CO poisoning in DMFCs include modifying the anode electrocatalyst structure/composition (e.g., Pt-Ru) to enhance the reaction rate, adding oxygen in the fuel (O2 bleeding), using cyclic operation, and increasing the operating temperature. It was found that adding a second component to Pt prevented the formation of strongly absorbed CO by accelerating its oxidation kinetic at lower potentials. To date, the preferred anode material for the electrochemical oxidation of CH3OH is an optimized Pt-Ru binary electrocatalyst [73]. The presence of Ru results in the oxidation of CO via the formation of adsorbed hydroxide ions at low potential and the lowering of the oxidation potential of CO [72]. The bi-functional mechanism for CH3OH electro-oxidation on Pt-Ru is described in Appendix B. Binary metal-oxide Pt catalysts, which are poor electrical conductors, were also studied to promote the electrochemical oxidation of CH3OH [74]. It was reported that Pt-WO3 electrodes are more active and resistant to poisoning than Pt or Pt-Ru alloy catalysts [75]. In this work, Pt and Pt-Ru supported and unsupported catalysts will be employed. 22 Methanol crossover Methanol is fully miscible, i.e., a highly soluble molecule which readily mix with water. Nafion® membranes are permeable to CH3OH. In DMFCs, water and CH3OH molecules are transferred through the membrane by diffusion due to the driving force of concentration gradients between the anode and the cathode. They are also transferred through the PEM by electro-osmotic drag, as induced by the movement of protons, and by hydraulic pressure gradients which is usually small compared to the two other crossover mechanisms. In a flowing-electrolyte DMFC, diffusion is replaced by convection- diffusion [76]. Water diffusion can be neglected in a well-humidified membrane. However, CH3OH diffusion cannot be neglected because a CH3OH concentration gradient develops between the anode and the cathode. The diffusive flux and electro-osmotic drag of CH3OH are both directed from the anode to the cathode. While anodic gaseous CO2 results from the electro-oxidation of CH3OH, cathodic gaseous CO2 is usually attributed to the combustion of the crossed-over CH3OH with O2. However, it was shown that anodic permeation CO2 to the cathode by diffusion and convection due to electro- osmosis also occurs in DMFC, and is not negligible [77]. The anodic CO2 permeation was determined to be 20% of the total cathodic flux in a DMFC at room temperature. Methanol crossover from the anode to the cathode results not only in fuel waste, but also reduces the voltage and Faradaic efficiency of DMFCs [78, 79]. This crossover phenomenon lowers the DMFC performance at higher fuel concentrations [80]. For this reason, there is an optimum CH3OH concentration to use at specific DMFC operating conditions [81]. With Nafion®, the crossover typically decreases with decreasing operating temperatures and concentrations, increasing cathodic pressure, increasing membrane thickness and equivalent weight, as well as increasing current density. Typical measures to minimize CH3OH crossover and its effects include the optimization of the proton- conducting membrane structure and composition, the use of oxygen reduction electrocatalysts insensitive to CH3OH at the cathode, and the improvement of electrochemical oxidation catalyst utilization and optimization of the cell operating conditions. Modifying the CH3OH membrane permeability without decreasing its ionic and/or electrical conductivity is not a simple process [79]. High 23 temperature (120-200oC) acid-doped polybenzimidazole (PBI) membranes have high conductivities [82] and improve DMFC cell performance by decreasing CH3OH crossover [83]. 1.3.2.5 Indirect methanol fuel cell The indirect methanol fuel cell (IMFC), also referred to as the reformed methanol fuel cell (RMFC), is the combination of a H2 PEMFC with an upstream catalytic CH3OH SR. This approach has the benefits of both the high electrochemical activity of H2, as well as the convenience of liquid CH3OH storage and distribution. However, as mentioned earlier, the SR product stream needs to be purified before H2 can be fed to the PEMFC. Any remaining un-recoverable H2 is combusted with air. As the IMFC system typically operates at higher temperatures than the DMFC, heat losses have to be properly managed. Due to the energetic inefficiencies associated with the catalytic CH3OH SR, the IMFC tends to be larger, heavier, and have longer start-up times than the DMFC [84]. For this reason, most research is directed towards the DMFC instead of the more complex IMFC system. Feeding liquid CH3OH directly to the anode of a DMFC eliminates the need for a fuel processor, but, as discussed earlier, the full potential of the DMFC is hindered by the slow CH3OH electro-oxidation kinetics, and CH3OH crossover. To reduce the effect of crossover, the operation of DMFCs is limited to low CH3OH concentrations, which renders their performance more suitable for low power density applications. As the H2 PEMFC can attain higher power densities than the DMFC, the DMFC’s low power density would require improvements to be competitive with the IMFC. The IMFC process could become a more interesting option if a low temperature CH3OH fuel reforming process could be developed. The drive cycle dynamic response and efficiency of an IMFC’s CH3OH SR has been investigated [85, 86]. A study from the Los Alamos National Laboratory compared the trade-offs of using a PEMFC, DMFC or an IMFC for portable power applications [87]. Ishihara et al. compared the energy of an IMFC with that of a DMFC and determined that, at a current density of 600 mA/cm2, the DMFC needed to operate at a minimum voltage of 0.5 V to have a greater energy efficiency than the IMFC [88]. Brown compared seven fuels, including CH3OH, gasoline and H2, as possible H2 sources for PEMFC for use in fuel cell 24 vehicles (FCVs) [89]. Lattner et al. compared CH3OH based fuel processors using commercial catalysts for PEMFC systems [90]. A high-level efficiency and economic comparison of some H2 production methods is provided in Appendix C. 1.3.3 Proton exchange membrane electrolysis The principle of electrolysis has been known since the early 19th century. It consists in the separation of chemically-bonded elements and substances by the passage of an electric current. The electrochemical decomposition of water is the most practical electrolytic process for the production of pure H2 and O2. Existing H2 electrolytic production methods include alkaline electrolysis, PEM electrolysis, ceramic oxide electrolysis, and photoelectrolysis. PEM water electrolysis was developed over the last decades [91]. It is an acidic process which possesses many benefits over traditional alkaline electrolysis: it results in higher energy efficiencies and power densities, and it is a simple system that can operate at low temperature and high pressure [92, 93]. It has been used in the industrial gas markets for H2 production, in the military and aerospace markets for O2 production, and is now close to entering new emerging energy markets, such as H2 vehicle refuelling [94]. It represents an attractive technology for portable applications due to its compactness and its ability to effectively produce pure H2 on demand. 1.3.3.1 Proton exchange membrane water electrolysis PEM water electrolysers are the reverse of H2 PEMFCs. Water is split into oxygen and protons at the anode. The protons travel through the proton conducting membrane and combine with electrons at the cathode to form H2 via the hydrogen evolution reaction (HER). The water electrolysis electrochemical reactions are as follows: Anodic half-cell reaction: H2O → 1/2O2 + 2H+ + 2e- Eoa = -1.23 V vs. SHE (1.21) 25 Cathodic half-cell reaction: 2H+ + 2e- → H2 Eoc = 0 V vs. SHE (1.22) Overall reaction: H2O → 1/2O2 + H2 Eo = -1.23 V (1.23) Water electrolysis is an endothermic reaction with a positive Gibbs free energy, which results in a negative standard equilibrium cell voltage of -1.23 V. It is a power sink which requires electrical energy input to extract H2 from water. The amount of potential/current applied controls the electrocatalytic reaction rate as per Faraday’s law. Commercial PEM water electrolysers operate between 60-80oC. Although there exists significant opportunities for the development of H2O electrolysis systems, they remain economically limited by the high overall cost required to split H2O. While the H2O electrolyser operating costs may be more affordable where electricity prices are low, the cost of electricity greatly fluctuates with geographical location. When combined with renewable sources, PEM H2O electrolysis becomes a possible pathway to zero pollution fuelling [95]. To render the electrochemical generation of H2 as economic as the production of hydrocarbon fuels, significant research advancements are needed to reduce the high energetic cost of H2O splitting, increase its operating efficiency, and reduce its capital cost. The core structure of PEM water electrolysers is similar to that of H2 PEMFC and they are subject to similar overpotential losses. Fortunately, some of the challenges encountered in PEMFCs, such as cooling and water management, do not apply to PEM electrolysers as there is a constant contact with H2O. The anode, where the electro-oxidation of water to oxygen occurs, has a high activation overpotential and represents the largest portion of the PEM water electrolyser losses. Like in the case of the PEMFC, various PEM water electrolysis research approaches are taken to address the concerns related to catalytic activity, electronic conductivity and stability. Non-noble metal catalysts corrode and Pt forms a conducting oxide film when in contact with the acidic Nafion® membranes. Pt has a low 26 overpotential at low current densities, but its resistance must be reduced at high current densities. The performance of the membrane and electrocatalysts is crucial to maintain high electrolyser efficiencies at high current densities. Since the development of dimensionally-stable anodes (DSA), the anode of PEM electrolysers usually comprises noble metal oxides (Ir, Ru) combined with oxides of non-noble metal (Sn, Ti) as electrocatalysts. These anodes usually possess large active surface areas. However, it was reported that there may be a poor contact between the catalyst layer and the PEM electrolyte because of high open porosity, and that the thin-film electrocatalyst layer technique, also used in PEMFCs, may be a preferred option [96]. Fujishima et al. photo-electrochemically produced H2 using n-type semiconductor TiO2 at the electrolyser anode [97]. Composite membranes based on Nafion® and TiO2 or Nafion® and SiO2 were also developed to improve the membrane conductivity for high temperature PEM electrolysis [98, 99]. Some PEM water electrolysers are commercially available. Hydrogenics Corporation’s HyLYZERTM consists of 7 PEM cells which can generate up to 1 Nm3/h of 99.99% H2 at a stack efficiency of 4.9 kWh/Nm3 and ambient temperature (5 to 35oC) [100]. Its maximum power consumption is 7.2 kW, when all equipment is included. The H-Tec EL 30 electrolyser has a maximum H2 production rate between 0.3 m3/h (13 cells) and 3.6 m3/h (144 cells) at ambient temperatures ranging from 4 to 50 oC [101]. Their rated power range is from 1.8 to 20 kW, but no information is given about the EL 30 bar electrolysers efficiencies. It is not clear if these commercial electrolysers operate at lower currents to attain greater efficiencies. While reduction in the electrical energy required for the electrolysis may be obtained by reducing the overpotential losses, it is not possible to operate the electrolysis system at voltage more positive than the H2O electrolysis reversible cell voltage at standard conditions. Further energy cost reductions might be achieved through depolarized electrolysis methods as described below. 1.3.3.2 Depolarized proton exchange membrane water electrolysis Increasing the electrolysis efficiency and reducing the reversible cell voltage can be accomplished by substituting the O2 evolution reaction (Eq. 1.16) with another anodic reaction taking place at a lower potential than that of O2 evolution. This technique, called anodic depolarization, consists of depolarizing the H2O electrolysis anodic electro-oxidation reaction by oxidizing a fuel, which forms oxidized species along with H+ ions and electrons [102]. The electro-oxidation of the chosen depolarization fuel must have more favourable thermodynamics than the O2 evolution reaction, i.e., a smaller reversible standard 27 cell voltage, Eo. In addition, Eo and the operating electrolyser cell voltage in normal operation, Uoelec, cell, would both be further reduced, if the chosen anodic electro-oxidation involved the transfer of more than two electrons. Therefore, depolarized electrolysis may result in reduced energy consumption, hence cost reductions, over conventional water electrolysis. Depolarization fuels studied in the past include coal [103], glucose (C6H12O6) [104], sulphur dioxide (SO2) [105, 106], methanol (CH3OH) [107, 108], and ferrous ions (Fe2+) [109]. Only the case of methanol depolarization is discussed here. Methanol depolarized PEM electrolysis Methanol is a particularly attractive water electrolysis depolarizer for the generation of H2. The DMR or CH3OH PEM electrolyser is analogous to a DMFC as the anodic reaction is the electrochemical oxidation of CH3OH in both cases. Consequently, most issues taking place at the electrodes during CH3OH depolarized water PEM electrolysis are encountered at the DMFC anode. The electrochemical reactions are as follows: Anode half-cell reaction: CH3OH + H2O → CO2 + 6H+ + 6e- Eoa = -0.016 V vs. SHE (1.24) Cathode half-cell reaction: 6H+ + 6e- → 3H2 Eoc = 0 V vs. SHE (1.25) Overall reaction: CH3OH + H2O → CO2 + 3H2 Eo = -0.016 V (1.26) The overall reaction for CH3OH catalytic reforming process (Eq. 1.10) is the same as for the electrochemical CH3OH reforming process (Eq. 1.26). The standard reversible cell voltage required to 28 drive the CH3OH depolarized PEM electrolysis (Eq. 1.26), is much less (minimum 0.016 V) than for water PEM electrolysis (minimum of 1.23 V). Like in the PEM water electrolysis case, the amount of electrical energy required mainly depends on the anodic activation overvoltage. This is because the H2 evolution reaction is very rapid, and the CH3OH oxidation is rate limiting. The electrolyzer electrode where the electro-oxidation of CH3OH occurs causes most of the overpotential in this electrochemical system. As in the case of the DMFC anode, the anodic activation overvoltage of the CH3OH depolarized water electrolyser results from the poor alcohol electro-oxidation kinetics of the complex reaction mechanism involving adsorbed intermediates. In CH3OH electrolysis, CH3OH crossing over to the cathode will not be oxidized, but will result in a fuel loss, which should be minimized. It appears that no scientific papers were published on the CH3OH depolarization of PEM water electrolysis in acidic media until 2007. However, there was some prior activity on this topic in the patent literature. Some electrochemical reforming membrane reactors for the electrochemical reforming of CH3OH investigated are described in Table 1.2. Table 1.2: Relevant methanol electrochemical reforming membrane reactors. Reactant Product Solid Electrolyte Porous Electrode Temperature Reference Electron Donor Electron Acceptor [ oC] CH3OH (l) H+ H2, CO2 Aqueous NaOH or KOH [6-12 M] Pt/C 23-60 [110] CH3OH H+ H2, CO2 - - - [111] CH3OH (g) H+ H2, CO2 Nafion® Pt, Pt/Ru anode, Pd-Ag foil cathode 50-100 [112] The electrochemical production of H2 from CH3OH in the presence of a base, such as NaOH or KOH, was carried out at Pt/C electrodes [110]. The inclusion of a base permitted the generation of H2 without emission of greenhouse gas and without the need to separate H2 from the other gaseous products. However, the resulting carbonate ion by-product needs to be dealt with. An electrochemical reformer and fuel cell system was developed to reform organic fuels into H2 and CO2 [111]. It uses electricity 29 and/or thermal energy to supply the necessary reaction energy to convert the fuel to H2. There is no mention of the operating temperature, and of the type of electrode or electrolyte used. An electrochemical process to reform a mixture of CH3OH and water at temperatures between 50 and 100oC was proposed [112]. As shown in Fig. 1.2, the methanol electrochemical reforming membrane cell operating in the gas phase resembles a DMFC. Figure 1.2: Schematic diagram of a methanol electrochemical reforming membrane cell (Modified from patent DE 197007384 [112]). A polymeric proton-conducting membrane and Pt or Pt/Ru catalysts were used. A Pd-Ag foil was hot- pressed on the cathode polymer side of the membrane to prevent water and CH3OH crossover. A cell voltage between 300 and 500 mV was sufficient to generate H2. There are three Japanese patents pertaining to the electrolysis of alcohols: two of them relate to the electrolysis of alcohols in conjunction with a PEMFC, which is similar to an entirely electrochemical version of an IMFC [113. 114], and the other describes an organic electrolytic synthesis in the presence of a base [115]. The first American patent on the generation of H2 production by electrolysis of organic solutions dates from 2001 [116]. It was later followed by a patent on the electrolytic production of H2 in Nafion® e- e - H+ H2 CO2 CH3OH H2O Anode Cathode Pd Pd-Ag alloy Pt, Pt/Ru e-e - DC Power Source 30 the presence of a base [110]. Narayanan et al. last re-amended their original 2001 US patent in 2006 [117]. An example of a PEM CH3OH electrolyser in transient operation is shown in Fig. 1.3. The anode chamber is continuously fed with un-acidified CH3OH aqueous solution, which is recycled using a pump. The cathode chamber is purged with argon. The anode catalyst is Pt or Pt-Ru/C and, typically, the cathode catalyst is Pt/C, but the performance of a Pt-WC/C catalyst has also been investigated. The electrolysis of CH3OH resulted in an absolute voltage of 0.4 V to 1 V at the same current density depending on the cathode catalyst used [7, 118]. The current collectors and diffusion layers were made of carbon paper. As Nafion® 117 was the electrolyte, the CH3OH electrolyser was operational at temperatures up to 120oC. Figure 1.3: Schematic diagram of an alcohol electrolyser (Modified from Hu et al. [7] with permission from Elsevier). At low current densities, the voltage reached a steady-state approximatively an hour after electrolysis started. At higher current densities, a rapid drop in negative voltage was observed. The drop happened because the applied current could not be sustained with only the oxidation of CH3OH and drifted to a CH3OH + H2O Anode Proton Conductor Electrolyte Anode Reaction: CH3OH + H2O → CO2 + 6H+ + 6e- CO2 Fluid Pump H+ DC Power Source Cathode Reaction: 6H+ + 6e- → 3H2 Cathode Ar, H2 Ar 31 value where O2 evolution could also occur [118]. Hu et al. reported that the electrolysis voltage did not depend on the CH3OH concentration at concentrations greater than 2 M and at temperatures below 60oC [7]. The explanation given was that at CH3OH concentrations exceeding 2 M, CH3OH crossover to the cathode resulted in a mixed potential which reduced the activity of the electrocatalyst. Shen et al. reported that the electrolyser voltage was about the same at CH3OH concentrations between 2 and 4 M, and that the voltage was greater outside this concentration range [11]. Again, this was explained by mass transport limitations at low CH3OH concentrations, and by CH3OH crossover to the cathode, which supposedly created a mixed potential, at higher CH3OH concentrations. Yet, Take et al. confirmed that the voltage did not depend on the CH3OH concentration, but only on the current density [118]. However, the limiting current density at which the voltage rapidly decreases varies with CH3OH concentration. It attained a maximum value when the methanol-water ratio was equal to 1 [118]. In the same study, it was also established that the oxidation of crossover methanol in the cathode compartment does not occur. The CO2 in the cathode exhaust is not produced at the cathode because no O2 is supplied at the cathode. In fact, any cathodic CO2 permeated from the anode in a similar manner as to that previously discussed for DMFCs. The CO2 permeation rate diminished with increasing anode CH3OH concentrations and increased with increasing current densities. Also like for the DMFC, the water and CH3OH would also cross over to the cathode at permeation rate which decreased as the current density increased. Take et al. verified that the CH3OH crossover rate escalated with increasing CH3OH concentrations, while the water permeation rate was not affected [118]. The water and CH3OH at the cathode can be separated from the H2 produced by molecular sieves. However, since it was found that the permeated water-methanol solution concentration was almost the same as the solution supplied to the anode, the permeated solution was recycled back to the anode. The flow rate of the H2 produced increased proportionally to the current density and came close to the theoretical H2 production rates [118]. The electrolysis voltage became more positive with increasing temperature as the mass transport and the kinetics are more favourable and the activation and concentration polarizations are reduced. For example, the voltage at 80oC was reduced by half compared to the voltage at 20oC [11]. 32 Applications of depolarized methanol PEM electrolysis The CH3OH PEM electrolysis H2 production has advantages over conventional CH3OH steam reforming as it may result in reduced energy consumption. First, the single-reactor CH3OH electrochemical reforming process has the potential to be conducted in the liquid phase at a lower operating temperature and at higher overall system efficiencies compared to a conventional high-temperature multi-reactor SR system. Secondly, since the production of H2 from depolarized electrolysis should result in lower power consumption than H2 production from water electrolysis. Therefore, its large scale commercialization might not be as significantly energetically restrained as that of water electrolysis. Narayanan et al. estimated that H2 production from the electrolysis of CH3OH would cost about 50% less compared to that of water, even when the cost of CH3OH is taken into account [117]. An efficiency, energy and economical comparison of various H2 production methods is presented in Appendix C. Although CO2 is produced whether CH3OH is thermally or electrochemically reformed, it is more localized and concentrated in the electrochemical reforming process than in the chemical reforming process. For this reason, electrochemical reforming may result in more effective CO2 capture for sequestration and in lower CO2 disposal cost in many applications. The cost of CO2 capture and disposal for various coal-fired generation technologies ranges from 1.1 to 4.9 cent/kWhe and 0.6 to 6.7 cent/kWhe, respectively [119]. In some cases, it may be desirable to combine the CO2 produced with the H2 generated at high temperature and pressure to obtain methane, as per the Sabatier reaction: CO2 + 4H2 → CH4 + 2H2O (1.27) Alternatively, other depolarized water PEM electrolysis processes using compounds which do not produce CO2, such as the ferrous ions (Fe2+), could be evaluated for H2 production. Methanol electrolysis could effectively be used as part of systems cogenerating chemical energy and electrical energy, which may include a H2 PEMFC, as long as the amount of energy required to generate H2 does not exceed the energy resulting from using the H2 produced. A system composed of an efficient electrochemical CH3OH electrolyser providing H2 to a PEMFC could be more advantageous than a 33 DMFC because it might be possible to use higher concentration fuel mixtures in a system combining a CH3OH electrolyser with a H2 PEMFC than in a DMFC. Without taking the storage and infrastructure issues into account, the overall performance of a system composed of a CH3OH electrolyser and a H2 PEMFC is expected to be less than a system in which gaseous H2 tanks directly feeds a PEMFC, but greater than that for a IMFC. The system combining the principles of H2 pumping and autothermal reforming referred to as electrochemical autothermal reforming (EATR) [120, 121] could be modified to include an electrochemical reformer in order to simultaneously produce H2 and useful heat. The resulting higher temperatures would render the thermodynamic and kinetic aspects of the electrochemical CH3OH oxidation more attractive. The prompt start-up and shut-down time of PEM electrolysers and their ease of scalability render them appropriate to generate pure H2 for portable and stationary applications. Having similar scalability limits as PEMFCs, multiple PEM electrolysis cells can be put into stacks. Currently, some of the largest PEM electrolysis stacks commercially available have a rated power of 14 to 20 kW [101]. The combination of an electrochemical alcohol reformer and a fuel cell system could potentially be made compact, light and simple enough for use in small low power applications, which usually do not justify the cost of a separate fuel processor. It could be developed into a portable fuelling option for high and low power output PEMFCs, as well as for micro-fuel cell power sources, and used to power laptops, cell phones and other small portable devices. Under thermodynamically favourable operating conditions, the direct electrochemical CH3OH reformer could be used as a stand-alone H2 generation unit for H2 combustion engine vehicles or FCVs requiring fuelling at local electrochemical stations. Distributed H2 production at a H2 refuelling station constitutes an attractive option to supply H2 to FC vehicles, allowing the supply to match the demand as more H2 vehicles are driven. As it has clear advantages over high temperature SR, the low temperature CH3OH electrochemical reformer may meet the DOE on-board fuel processing targets for automotive applications. It may represent an attractive on-board H2 storage alternative to high pressure H2 storage in FCVs, as the H2 could be produced at the PEMFC’s demand. A CH3OH fuel processor and fuel cell system could be optimized to give maximum power density as an off-grid generator system for back-up and remote power applications for the telecommunications, utilities and military sectors. As it requires less energy to electrolyser CH3OH than water, and depending on the source of CH3OH, CH3OH electrolysis may be a preferred option to produce H2 during off-peak hours so 34 electricity can be provided by H2 PEMFCs during high grid loads or blackouts. Based on these examples, it can be seen that an efficient CH3OH electrochemical reformer has a significant market potential. 1.3.4 Electrochemical promotion of heterogeneous catalysis Catalysts typically consist of an active metal or metal oxide phase, a support, and one or more promoters. The promoters affect the properties of the supports or of the active phase [122]. The metal-support interactions between catalytic nanoparticles and their support is a well studied heterogeneous catalysis concept. A special type of heterogeneous catalysis, in which the catalyst properties are modified by the application of an electric field, was discovered in 1980, and was reported as a new electrochemically induced catalytic effect in 1988 [123, 124]. It was found that solid electrolytes or active catalyst supports may be used to modify the catalytic properties of metals and metal oxides. Applying polarization controls the surface work function by creating promoters and modifying their quantity on the catalyst surface, which in turns affects the adsorption, amount, and surface diffusion properties of reaction intermediates [125]. The multidisciplinary principles of this phenomenon combine features of heterogeneous catalysis and electrocatalysis as the working electrode is used to catalyze a chemical reaction and an electrochemical reaction simultaneously. Two types of processes occur at electrodes: Faradaic and non-Faradaic [126]. In Faradaic processes, the electrons transferred across the metal-solution interface cause an oxidation or reduction reaction to occur at a rate proportional to the current. Under unfavourable thermodynamic and kinetic conditions, charge transfer reactions may not occur. However, non-Faradaic processes like adsorption or desorption may occur. The structure of the electrode-solution interface may then be altered by changing the current or potential, or the solution composition [126]. In the electrical promotion of catalysis, most of the yield is obtained through the heterogeneous catalytic reaction, and is non- Faradaic. The superimposed electrochemical reaction used to control and improve the catalytic properties of the electrode surface result in the Faradaic part of the yield [125]. 35 1.3.4.1 Theory The non-Faradaic electrochemical modification of catalytic activity (NEMCA), also called the electrochemical promotion of catalysis (EPOC), is ascribed to the modification of the catalytic activity of metals on ionic conducting substrates. It is obtained upon variation of the potential on the working electrode, which is used as an electrode and as a catalyst for the heterogeneous catalytic reaction under study [127]. The application of a current (± 5-50 mA/cm2) or voltage (± 1-2 V) between the catalyst working electrode WE and a counter electrode CE, also deposited on the solid electrolyte, polarizes the catalyst and changes its work function. This property correlates the electrocatalytic and catalytic reaction rates with the properties of the catalyst involved. The catalyst work function is directly proportional to the catalyst electrode overpotential, which is the extent of polarization at the metal catalyst [128]. The work function, Φ (eV), is the minimum energy required to remove an electron from the interior of a solid to a position just outside. It is defined as: Ψ−=Φ− eeμ (1.28) where μe is the electrochemical potential of an electron at infinite separation from a metal, i.e., at the Fermi level (KJ), e is the charge of an electron (C), and Ψ is the outer potential (V), which represents the work required to bring a charge from infinity to a point just outside a charged phase and is equal to zero if there is no net charge on the metal surface. It was determine that the changes in the catalyst surface work function closely follow the changes in potential difference under galvanostatic or potentiostatic transients for a large number of systems [129]. The change in the work function of the catalyst surface, ΔΦW (eV), can be controlled in situ by an applied potential, UWR (V) as per the following relationship: WWRUe ΔΦ=Δ (1.29) where ΔUWR (V) is the overpotential between the working electrode catalyst and a reference electrode [130]. The catalyst potential is represented by UWR and is defined as the catalyst WE potential with 36 respect to a reference electrode (RE) while UWC represents the cell voltage and is defined as the catalyst-working electrode potential with respect to a counter CE. Vayenas et al. qualitatively explained this effect by the scheme represented in Fig. 1.4 for a metal- electrode deposited on an oxide-conducting and proton-conducting solid electrolyte. For a system where the solid electrolyte mobile ions are H+, the reaction at the electrode surface would be: H+ + M + e- → Mδ- - Hδ+ (1.30) The migration of ions to or from the ion conducting solid electrolyte from or to the catalyst surface is called backspillover, in analogy with the term spillover, which refers to migration from a metal to a support [129]. This backspillover of ions forms neutral spillover dipoles which establish an effective chemical double layer over the entire catalyst surface which is exposed to the reactants. Hence, EPOC is catalysis in the presence of an electrochemically controllable double layer at the catalyst/reactant interface [18]. The charged double layers are accumulations of charges on a surface, which are either formed due to diffusion effects, reactions between electrons in the electrodes and ions in the electrolyte, or results from applied currents or voltages. The classical metal-solid electrolyte double layer and the effective double layer resulting from the current or potential controlled ion migration are indicated in Fig. 1.4 (a) and (b) for an oxide conductor and for a proton conductor, respectively. All EPOC studies to date have been carried out using gaseous reactants. The electrocatalytic reaction takes place at the catalyst-solid electrolyte-gas three phase boundaries (TPB) while the catalytic reaction with no charge transfer takes place at the metal-gas interface [128]. Here, the term catalytic refers to a chemical reaction and the term electrocatalytic refers to an electrochemical reaction which includes charge transfer. The electrochemically-induced protonic current supplies promoting H+ ions to the catalytically-active metal-gas interface but do not directly affect the catalytic reaction rate [131]. It is the effective chemical double layer which interacts electrochemically with covalent chemisorbed reactants and reaction intermediates by lowering their energy and modifying their binding strength. 37 Figure 1.4: Schematic diagram of a metal catalyst-electrode deposited on (a) an oxide-conducting solid electrolyte, and (b) a proton-conducting solid electrolyte (Modified from Vayenas et al. [18] with permission from Springer). Oxide conducting solid electrolyte (ex. YSZ) _ Adsorbed Species (ex. CO) Effective Double Layer Metal-Reactant Interface (catalytic reaction) Working catalyst electrode Reactants (ex. H2O, CH3OH) W Metal (ex. Pt) O2- ++ O2- ++ 2-O ++ δ+ δ+ Oδ- δ+ Oδ- Oδ- Oδ- Oδ- δ+ δ+ Oδ- δ + C R Double Layer G/P UWR UWCA 1-2 V Triple Phase Boundary (TPB) (electrocatalytic reaction) Spillover ions C O (a) Proton conducting solid electrolyte (ex. Nafion) _ Adsorbed Species (ex. CO) Effective Double Layer Metal-Reactant Interface (catalytic reaction) Working catalyst electrode Reactants (ex. H2O, CH3OH) W Metal (ex. Pt) H+ _ H+ +H _ δ- δ- Hδ+ δ- Hδ+ Hδ+ Hδ+ Hδ+ δ- δ- Hδ+ δ- C R Double Layer G/P UWR UWCA 1-2 V Triple Phase Boundary (TPB) (electrocatalytic reaction) Spillover ions C cO (b) 38 A change in the strength of the chemisorptive bonds can be induced upon polarization of the metal-solid electrolyte interface. The corresponding decrease in overpotential can enhance the catalytic reaction rate at the gas-metal interface in a very pronounced and reversible manner [18]. This induced exponential catalytic rate non-Faradaic enhancement is analogous to the Buttler-Erdey-Gruz-Volmer (BEV) equation for high electrode overpotentials, highlighting the resemblances between electrochemistry and heterogeneous catalysis [127]. Some of the rules of EPOC are listed in Appendix D. Supplying H+ with a proton conductor (positive voltage application) is equivalent to removing O2- with an oxide conductor (negative voltage application) as they both decrease the catalyst work function. An electron donor is defined as a compound that gives electrons during its oxidation, while an electron acceptor is defined a compound which accepts electrons during its reduction. As per the first rule of EPOC, adding electropositive promoters, such as H+, weaken the chemisorptive bond of electron donor (negatively charged) adsorbates and strengthen the chemisorptive bond of electron acceptor (positively charged) adsorbates. This first principle has been experimentally confirmed and rationalized on an electrostatic and quantum mechanic basis as per Eq. 1.26. The binding energy or enthalpy of adsorption, ∆Hj, is related linearly to the change in work function, WΔΦ , as follows: WHj J H ΔΦ⋅≈ΔΔ α (1.31) where αH,j is a parameter which is positive for electropositive (electron donor) adsorbates, and negative for electronegative (electron acceptor) adsorbates [18]. Although not a general fundamental equation, Vayenas et al. are not aware of any exceptions to the physical meaning which Eq. 1.26 conveys, i.e., that ∆Hj, and hence, the coverage of an electron acceptor/donor adsorbate decreases/increases with increasing work function, and thus, decreasing Fermi level, EF. More details on this can be found in Appendix D. It is known that CH3OH will be an electron donor (O bonding) on surfaces with high work functions and will be an electron acceptor (CO bonding) on surfaces with low work functions [18]. It was also established that the strength of the CO metal bond was reduced by the addition of electronegative modifiers (e.g., O2-) as it decreases the activation energy of CO desorption [18]. 39 Varying the potential and work function of the catalyst affects the heterogeneous reaction catalytic rate and the activation energy of the reaction. There is a linear variation in the activation energy, EA (kJ/mol): WH o AA EE ΔΦ⋅+= α (1.32) where αH is the enthalpic parameter, a constant, which is usually negative for electrophobic reactions (donor reactions), and oAE is the open circuit activation energy value (eV). Hence, the activation energy of an electron acceptor (positively charged) decreases linearly with decreasing work function [18]. Reactions for which the activation energy decreases with increases in the catalyst work function are termed electrophobic, and when the opposite occurs, the reactions are termed electrophilic. 1.3.4.2 Parameters The magnitude of EPOC is described by some important parameters: the rate enhancement factor, the rate enhancement ratio, and the promotion index. Normally, the electrochemical reaction rate, re (mol/s), is given by Faraday’s law: nF I nF Ai r Geome =⋅= , (1.33) where i is the applied current density (A/cm2), A is the active geometric surface area (cm2), I is the applied current (A), n is the number of electrons transferred, and F is the Faraday constant (C/mol). Following the established EPOC terminology conventions, in the non-Faradaic enhancement case, this equation becomes: nF Ir ⋅Λ=Δ , (1.34) 40 where ∆r is the non-Faradaic electrochemically induced change in the catalytic reaction rate (mol/s) [132], and is equal to the difference between the promoted catalytic reaction rate, r and the unpromoted catalytic reaction rate, ro. Here, n represents the number of electrons promoting the electrochemical reaction at the triple phase boundary. Hence, it is equal to 2 in the case of an oxide conductor and equal to 1 in the case of a proton conductor, as per the following electrochemical reactions: For an oxide conductor: ½ O2 + 2e- → O2- (1.35) For a protonic conductor: ½ H2 + 1e- → H+ (1.36) Rearranging equation 1.29, and knowing that for a proton conductor n = 1, the dimensionless rate enhancement factor, also referred to as the Faradaic efficiency, Λ, is defined as: I rnFΔ=Λ , (1.37) For catalytic oxidation reactions, Λ > 1 implies an electrophobic (donor reaction) behaviour for which the rate increases with increasing catalyst work function, while Λ < -1 implies an electrophilic (acceptor reaction) behaviour for which the rate increases with decreasing catalyst work function. For a pure electrocatalyst, Λ is always unity [18]. There are four types of global catalytic reaction r vs. Φ behaviours over the entire experimentally accessible Φ range: purely electrophobic (∂r/∂Φ>0), purely electrophilic (∂r/∂Φ<0), volcano type (∂r/ ∂Φ>0 followed by ∂r/∂Φ<0), and inverted volcano-type (∂r/∂Φ<0 followed by ∂r/∂Φ>0). All purely electrophobic reactions are positive order in electron donor and zero or negative order in electron acceptor. All purely electrophilic reactions are positive order in electron acceptor and zero or negative order in electron donor. Volcano-type reactions are always in positive order for one reactant and purely negative order in the other. Inverted volcano-type reactions are positive in both reactants [18]. The EPOC global behaviour rules are listed in Appendix D. 41 A reaction is electrochemically promoted when ΙΛΙ > 1 and it is electrocatalyzed when ΙΛΙ ≤ 1. The exchange current of the metal-solid electrolyte interface, Io (A), can be extracted from Tafel plots and used to estimate the expected magnitude of the absolute value of the rate enhancement factor as per: 0 0 I nFr≈Λ (1.38) where ro is the open circuit unpromoted catalytic reaction rate (mol/s) [133]. This expression defines the applicability limit of EPOC: when the absolute value of the rate enhancement factor well exceeds unity, EP is observed and the kinetic efficiency is enhanced. For a reaction to be electrochemically-promoted, the open circuit catalytic rate, ro, must be greater than Io/nF and the catalytic reaction must be faster than the electrocatalytic one [127]. The fact that Io increases exponentially with temperature in conjunction with the fact that Λ is inversely proportional to Io explains why most EPOC studies are restricted to lower temperatures [134]. The dimensionless rate enhancement ratio, ρ, is defined as the ratio of the promoted to unpromoted catalytic rate: 0r r=ρ , (1.39) where r is the electrochemically promoted catalytic reaction rate (mol/s). 1.3.4.3 Electrochemical promotion of methanol oxidation The EPOC effect has been demonstrated for a variety of catalytic reactions, including the CH3OH oxidation on oxide-conducting ceramics. Table 1.3 summarizes the rate enhancement factors and rate enhancement ratios reported for CH3OH oxidation EPOC studies. All studies listed in Table 1.3 were conducted using gaseous reactants. The most pronounced rate enhancements were typically observed at the lowest temperature examined [128]. It is important to remark that most EPOC catalytic oxidation reaction studies were conducted via oxygen pumping using solid oxide ion conductors. 42 Table 1.3: Vapour phase methanol oxidation EPOC studies with oxide conductors. Reactant Electron Donor Electron Acceptor Product Electrolyte Catalyst T [oC] Λ ρ Global Behaviour (GB)* Ref. CH3OH O2 H2CO, CO2 YSZ Pt 250- 320 100 - PE- [136] CH3OH O2 H2CO, CO2 YSZ Ag 500 -95.5 2 PE+ [137] CH3OH O2 H2CO, CO2 YSZ Pt 250- 500 - 2.5 - [132] CH3OH O2 H2CO, CO, CH4 YSZ Ag 550- 750 -25 6 PE+ [138] CH3OH O2 H2CO, CO2 YSZ Pt 300- 500 1x104 4, 15 IV [18] CH3OH O2 H2CO, CO2 YSZ Pt 400- 500 -10 3 PE+ [18] *GB: global behaviour, PE-: purely electrophobic, PE+: purely electrophilic, IV: inverted volcano, V: volcano. Note: Although EPOC was mostly studied on oxidation reactions, it has been applied to many hydrogenation and dehydrogenation reactions. 1.3.4.4 Electrochemical promotion using protonic electrolytes The use of the electrochemical promotion of heterogeneous catalyst surfaces has been investigated on various reactions with proton-conducting electrolytes. The EPOC effect was observed at high temperatures using proton-conducting ceramics, and also at low temperatures using Nafion® as a protonic polymer electrolyte. Table 1.4 summarizes the findings of EPOC studies using protonic conductors. All studies in Table 1.4 were conducted using gaseous reactants. No EPOC studies on the electrochemical reforming of CH3OH using proton-conducting electrolytes have been reported in the literature. By comparing the rate enhancement factors and rate enhancement ratios reported in Table 1.3 and Table 1.4, it can be generally concluded that, to date, it seems that there are less reactions having a high rate enhancement factor with proton conductors than with solid oxide ion conductors. Table 1.5 lists the few EPOC studies which have been carried out in alkaline aqueous media. 43 Table 1.4: Vapour/gaseous phase EPOC studies using protonic conducting electrolytes. Reactant Electron Donor Electron Acceptor Product Electrolyte Catalyst T [oC] Λ ρ GB* Ref. C2H4 O2 CO2 Ba3Ca1.18 Nb1.82O9-a Pt 250 - 350 -1340 4 PE+ [139] H2 N2 NH3 CaIn0.1Zr0. 9O3-a-a Fe 440 6 ∞ PE+ [13, 140] C2H4 O2 CO2 CaIn0.1Zr0. 9O3-a Pt 385 - 470 -3x104 5 PE+ [18] NH3 N2, H2 CaIn0.1Zr0. 9O3-a Fe 530 - 600 150 3.6 PE- [18] H2 N2 NH3 SrCe0.95Yb 0.05O3-a (SCY) Pd 550 - 750 ~ 1-2 - - [141] CH4 - C2H6, C2H4 SCY Ag 750 - 8 PE- [18, 128] H2 CO2 H2O, CO SrZr0.90Y0.1 0O3-a Cu 550 - 750 < 1 - - [142] C2H4 O2 CO2 Gd-doped BaPrO3 (BPG) and Y-doped BaZrO3 (BZY) Pt 400 - 600 ~ 1 1.3 PE+ [143] H2 C2H4 C2H6 CsHSO4 Ni 150 - 170 300 2 PE+ [18, 144] C2H4 O2 CO2 Nafion® 117 Pd/C - 6 1230 - [145] H2 O2 H2O Nafion® 117 Pt 25 20 6 V [18, 146] 1-C4H8 - C4H10, 2-C4H8 Nafion® 117 Pd 70 -28 40 PE+ [18, 147, 148] *GB: global behaviour, PE-: purely electrophobic, PE+: purely electrophilic, IV: inverted volcano, V: volcano. 44 Table 1.5: EPOC studies in aqueous systems. Reactant Electron Donor Electron Acceptor Product Electrolyte Catalyst T [oC] Λ ρ GB* Ref. H2 O2 H2O 0.1 M KOH Pt 25-50 20 6 PE- [18] H2 O2 H2O 0.01-0.5 M KOH / LiOH Pt - 7.2 - PE- [135] *GB: global behaviour, PE-: purely electrophobic, PE+: purely electrophilic, IV: inverted volcano, V: volcano. It is important to note that, for H2 oxidation using Pt as a catalyst in aqueous media, non-Faradaic rate changes were only obtained in alkaline aqueous solutions. Only Faradaic changes were observed in acidic solutions as the catalytic rates were much higher than those measured in the alkaline aqueous solutions [135]. This was attributed to the larger electrocatalytic activity measured in acidic aqueous solutions. In aqueous media, the concentration of electrolyte affects the magnitude of EPOC [135]. It was also reported that the low temperatures of aqueous electrochemistry may have limited the number of reactions where non-Faradaic enhancements may have been obtained, as the open circuit catalytic activity, ro, must be measurable to obtain an EPOC effect [132]. No EPOC studies on the electrochemical reforming of liquid CH3OH in acidic media have been reported in the literature. 1.3.4.5 Applications for the electrochemical promotion of catalysis Currently, there is a strong industrial interest in developing commercial EP applications with most efforts focused on lab-scale EPOC research. The acceptance and understanding of EPOC is still limited [125]. The use of EPOC in commercial applications depends on technical and economical aspects: material cost minimization, ease of electrical connection, efficient and compact reactor design [149]. Although applied research efforts are needed to develop industrial EPOC applications, fundamental research is still required to improve commercialization possibilities. For example, low temperature EPOC with aqueous electrolytes has a tremendous potential for many applications but has not been thoroughly studied in the literature. Suitable chemical processes for industrial EPOC applications are slow processes, which can be activated through the use of EPOC, and processes requiring low investments, 45 as it would reduce the market entry risks for the first commercial EPOC application. EPOC has been recognized as a mean to facilitate the commercialization of fuel cells [125, 149]. H2 production and utilization was identified as a potential niche market where EP could find an open field opportunity for industrialization [149]. A compact and flexible EPOC reactor design will minimize heat losses, maximize efficiencies, and facilitate the practical utilization of EPOC in industrial settings. To reduce application implementation costs, an efficient EPOC reactor design based on established technologies which have simple electrical connections, and use inexpensive materials would need to be developed. Durability, lifetime, and scalability are other aspects which will need to be addressed to develop a practical EPOC reactor [149]. To date, most EPOC reactor designs have been based on heterogeneous catalysis reactors and fuel cells. Simplified EPOC reactors may be obtained through single chamber reactors and bipolar systems. Scalable liquid electrolyte EPOC reactors designs may be obtained based on scrubber technologies [127]. Most EPOC studies have been conducted in electrochemical membrane reactors (EMR) using solid electrolytes [150]. Two main types of catalytic-electrocatalytic reactors have been used: double-chamber reactors and single-chamber reactors. In the double-chamber reactor, the catalyst WE is exposed to the catalytic reactants and products, while the CE and RE are exposed to a reference gas in a separate compartment. This design possesses an accurate RE but is difficult to scale-up. The EPOC effect has also been studied in continuous flow single-chamber reactors containing three electrodes (WE, CE and RE), which are all exposed to the reactants and products in order to promote the rate of catalytic reactions. The CE and RE are both made of catalytically-inert materials so that no catalytic reaction occurs on their surface. The single-chamber reactor requires the use of a quasi-reference electrode, which may cause inaccuracies in measuring the catalyst potential, but is easier to scale-up. It was shown that direct electrical contact to the catalyst electrode is not necessary to induce EP in a bipolar monolithic reactor [151]. EPOC can be induced without external potential application by using the potential difference developed between the catalyst WE and a catalytically-inert CE. This single pellet wireless configuration was used to demonstrate the effect of EPOC on the oxidation of CH3OH at 46 250oC with an oxide ion-conducting electrolyte [132]. A change in the CH3OH consumption rate was obtained by using the RE as a CE, and short-circuiting it with the WE. A potential difference developed between the CE and the WE, and resulted in a current flow between two electrodes [136]. The wireless EPOC for the oxidation of CH3OH in a mixed-reactant single-chamber alkaline fuel cell has been patented [152, 153]. The design of EPOC reactors based on this wireless EPOC concept may lead to simpler designs. Yiokari et al. evaluated the effect of EPOC on the synthesis of NH3 via the Haber-Bosh process, using proton conducting ceramic catalyst pellets [154]. This study was the first to use an industrial catalyst, to be carried out at high pressure (50 atm), and it was the first attempt at an EPOC scale-up. St-Pierre et al. employed EPOC in an electrolytic cell to purify a reformate stream H2 feeding a PEMFC [155]. EPOC can efficiently be carried out on thin sputtered metal films which possess stability and endurance. Balomenou et al. conducted development work on a high temperature membrane electrochemically promoted reactor (MEPR), which is a hybrid between a monolithic honeycomb catalytic reactor and a planar flat or ribbed plate SOFC for the EPOC of the catalytic oxidation of hydrocarbons [156, 157]. In this scalable reactor design, two external electrical connections are needed to dynamically control the applied current/potential required to induce EPOC to the catalyst film plates and only one gas stream containing all reactants and products was used. The benefits of EPOC were also applied to the treatment of automotive exhaust in catalytic converters to reduce CO emissions [158-161]. EPOC studies were extended to non-redox systems and Salazar et al. electrochemically enhanced the isomerization reaction of olefin by over a thousand fold in a PEMFC using Pd/C catalyst at the cathode [162]. Sapountzi et al. have shown that EPOC could efficiently purify the H2 stream from the CO formed during the H2 production from hydrocarbons or alcohols, by electrochemically enhancing the water-gas-shift reaction or by improving the selectivity towards CO oxidation [163]. EPOC was used in the development of a catalytic system for the after-treatment of diesel exhaust. The unit combustion, tested on a commercial diesel engine, was found to be satisfactory for soot but not for NOx, and its Faradaic efficiency was estimated to be 66 [149]. More recently, a method using EPOC in a tubular reactor at temperatures from 150 to 600oC was developed to improve the reaction kinetics of biofuel production from biomass [164]. 47 1.3.5 Electrochemical promotion of electrocatalysis Recently, a new engineering solution analogous to EPOC was developed to enhance electrochemical reactions instead of chemical ones. Hence we refer to it as the electrochemical promotion of electrocatalysis (EPOE). This approach, involving a third electrode, is referred to in the literature as triode operation. Balomenou et al. have shown that it can enhance the power output of fuel cells [19, 165]. A dual-chamber triode fuel cell arrangement was designed to study the application of EP in fuel cells using gaseous reactants. It comprised three electrodes: the fuel cell anode WE, the fuel cell cathode CE, and an auxiliary CE. The fuel cell anode WE was also the auxiliary WE. The electrodes were in electrolytic contact and form two electrical circuits as shown on Fig. 1.5. The fuel cell circuit consisted of the fuel cell anode WE and the fuel cell cathode CE. The auxiliary circuit consisted of the auxiliary WE, which was also the anode WE of the fuel cell circuit, and the auxiliary CE. Figure 1.5: Schematic diagram of a triode fuel cell electrical circuit. The net Faradaic fuel-consuming current, Ifar, is the difference between the fuel cell current, Ifc, and the auxiliary circuit current, Iaux [14]. When the auxiliary circuit is not used, the fuel cell operates normally. When a current is supplied in the auxiliary circuit, the fuel cell anode and cathode are forced to operate under a controlled potential [14, 166]. It was reported that the triode improvement resulted from using Fuel Cell Ring CEfc Polymer Electrolyte Membrane Common Current Collector Fuel Cell Current Collector Auxiliary Circuit Fuel Cell Circuit Common WE WEaux = WEfc Ifc Ufc Iaux Uaux Auxiliary CEaux G/P 48 the auxiliary circuit potential to keep the fuel cell anode (or cathode) at a corrosion-type potential inaccessible during normal fuel cell operation. At the new WE operating potential resulting from triode operation, the surface coverage of the promoting species and the adsorption of adsorbates is modified. In some cases, the reaction at the WE surface is enhanced in triode operation as it is no longer limited by the supply and adsorption of the fuel. Triode operation is advantageous under high anodic and cathodic overpotentials [14]. The application of EPOE resulted in gains similar to that obtained in classical EPOC experiments. The only results available in the literature pertain to triode operation of an SOFC and a PEMFC using gaseous reactants. Balamenou et al. obtained power enhancement ratios up 8, i.e., efficiencies over 700% for a SOFC operating on dry H2, ethane or methane at 400-750oC in triode operation [14]. In this case, the low anodic potential created in triode operation decreased the surface coverage of O2-, increased the adsorption of H2, CO, or CH4, and enhanced the rate of the anodic oxidation. In SOFC’s, the solid electrolyte ohmic resistance contribution typically exceeds that of the electrodes, and as the use of the auxiliary circuit diminishes the ohmic losses between the fuel cell anode and cathode, triode operation can enhance the fuel cell performance [14]. Triode operation was also shown to improve the efficiency of a PEMFC running on pure H2 or H2 poisoned with CO, using an anode feed of humidified H2/CO/He or H2/CO2/CO/N2 at 30oC [165]. Using 4 mg/cm2 of a Pt-Ru/C anode catalyst and 4 mg/cm2 of a Pt black cathode catalyst, the application of a current of 0.09-0.5 mA to the auxiliary circuit enhanced the power enhancement ratio of a CO-poisoned PEMFC by a factor of 5 [165]. The enhancement was attributed to the electrolytic H+ supplied by the auxiliary circuit, which decreased the coverage of adsorbed CO and enhanced the membrane conductivity due to additional H+ pumping. Vayenas et al. [166] developed the triode gas phase fuel cell and battery concept as it is an advantageous alternative method to conduct electrochemical reactions in non-liquid systems. It appears that no EPOE studies have been carried out using liquid phase reactants and that triode operation has not been evaluated in the electrolysis mode to date. In this thesis, the EPOE effect was evaluated in triode and tetrode operation using liquid phase reactants and in the electrolysis mode for the first time. 49 Chapter 2: Electrocatalysis Baseline Study 2. 1 Synopsis As discussed in Section 1.3.3, all published work relevant to CH3OH electrolysis or DMR was carried out in a recycle operation mode or in a flowing mode, in which an un-acidified CH3OH aqueous solution is pumped to the anode, while the H2 produced at the cathode was continuously purged with Ar. In this Chapter, the electrochemical production of H2 from the PEM electrolysis of liquid CH3OH in acidic aqueous media was investigated in the static mode (non-flowing, stirred). Experiments were carried out in a two-compartment glass cell with a MEA composed of a Nafion® 117 membrane and gas diffusion electrodes (GDE). This glass cell configuration allowed for the separate measurements of the anodic and cathodic potential contributions to the overall cell voltage. Most tests were conducted with 0.5 M H2SO4 in the anode and cathode compartment, but the effect of having a dry N2 purge cathode was also investigated. Methanol electrolysis was studied at concentrations ranging from 0 to 16 M, where 0 M corresponds to water electrolysis. The characteristics of the CH3OH - H2O electrolysis were described and compared to that of the H2O electrolysis under the same acidic conditions. The influence of supported (carbon) and unsupported (black) conventional fuel cell catalysts (Pt and Pt-Ru), operating temperatures (23, 50 and 75oC) and operating modes (dry and wet cathode) on the electrocatalytic reforming of CH3OH and water were evaluated. A theoretical thermodynamic analysis of the system was conducted and the limiting current densities, kinetic parameters, including the Tafel slopes and current exchange density, and apparent activation energies were determined. Additional electrochemical sample calculations can be found in Appendix H. The work presented in this section was published and is reprinted from the International Journal of H2 Energy, 35, C. R. Cloutier, and D. P. Wilkinson, “Electrolytic Production of Hydrogen from Aqueous Acidic Methanol Solutions”, 3967-3984, Copyright (2010), with permission from Elsevier. 50 2.2 Experimental 2.2.1 Materials Certified electronic grade CH3OH, and American Chemical Society (ACS) certified plus H2SO4, both from Fisher Scientific, were used with deionised (DI) H2O having a resistivity of 18 MΩcm to make up the solutions. The H2O associated with H2SO4 was taken into account when making the CH3OH solutions. Behmann et al. [167] stated that the addition of CH3OH to H2SO4 and H2O mixtures containing less than 18 M H2SO4 had no influence on the dissociation behaviour of H2SO4 in the ternary mixture, and had dissociation behaviour is similar to that of the H2SO4 and H2O binary system. Hence the dissociation behaviour of H2SO4 will be the same for all the aqueous acidic CH3OH solution concentration used in this study. Ultra high-purity H2 and N2 gases from Praxair were used in some tests. Potassium dichromate salt (K2Cr2O4) reagent plus ≥ 99.5 % from Sigma-Aldrich was used for the spectrophotometric studies with a Shimadzu UV Vis spectrophotometer (UV mini 1240). A supporting electrolyte concentration of 0.5 M H2SO4 was used in the anode and cathode cell compartments in most tests, unless other wise indicated. A volume of 80 ml of solution was used in the anode and cathode cell compartments in most tests, unless otherwise indicated. Nafion® 117 (Ion Power Inc.), was selected as the solid polymer proton exchange membrane. Conditioning was conducted by first boiling the membrane in a solution of 3 wt% H2O2 (certified, Fisher Scientific), for 30 minutes, then boiling it in DI H2O for 30 minutes and finally, boiling it in 0.5 M H2SO4 for 30 minutes. It was rinsed with DI between each boiling step and then cut to shape and stored in DI H2O until used. TorayTM carbon paper, TGPH-060, with 20 wt% Teflon® (PTFE) wet-proofing from BASF Fuel Cell Inc., was used as the gas diffusion layer (GDL). A micro-porous layer (MPL) consisting of a 1 mg/cm2 coating composed of carbon black and 20 wt% PTFE was sprayed on top of the cathode GDL surface, to form a double-layer gas diffusion layer. No MPL was used on the anode side. Conventional CH3OH fuel cell noble metal electrocatalysts were employed: 20 wt% high performance (HP) Pt supported on Vulcan XC-72 and 20 wt% HP Pt:Ru alloy (1:1 atomic ration (a/o)) supported on 51 Vulcan XC-72, both from E-TEK, Pt black (99.9+%, fuel cell grade) and Ru black (99.9%) both from Sigma Aldrich, and Pt:Ru alloy (1:1 a/o) from Alfa Aesar. The anode Pt-Ru catalyst loading was 4 mg/cm2, while the cathode Pt loading was only 2 mg/cm2, as the cathode half-cell reaction is simple and straightforward compared to the anodic half-cell reaction. The catalyst inks were composed of the supported catalyst powder, 30 wt% of a proton conductive ionomer in solution (Nafion®, 5 wt% perffluorosulfonic acid-PTFE copolymer, Alfa Aesar), H2O and isopropanol (IPA, 2-propanol certified ACS plus from Fisher Scientific). Catalyst inks were prepared and sprayed with an Accuspray® air gun (Model 07HS, ISAAC Series HVLP Spray), on one side of the GDL, forming a gas diffusion electrode (GDE). More details of the ink preparation and spraying procedure are given in Appendix E. A planar Pt disk (0.25 mm, 99.99%+ Pt foil, Goodfellow Cambridge Limited) of a 2.01 cm2 area was used as an anode catalyst in some experiments for comparison purposes. The current collectors were made from an annealed 0.25 mm thick Nb foil, 99.9% pure from Sigma-Aldrich. 2.2.2 Equipment All experiments were conducted in a dual-chamber borosilicate glass (Pyrex) cell composed of two chambers as shown in Fig. 2.1 and 2.2, which was built by CanSci Glass Products Ltd. Various ports permitted the inclusion of reference electrodes, thermocouples, and gas spargers or bubblers in the anode or cathode compartments. Slight internal bevels were provided in the upper part of the anode and cathode compartments to facilitate the gas exit. The CO2 gas forming at the anode and H2 gas forming at the cathode were vented to atmosphere in all experiments. The MEA holder contained different layers: the membrane electrode assembly, the current collectors, two silicon sealing rings and a PTFE spacer ring, as shown in Fig. 2.1 (b). The MEA holder was machined in polyetherimide (ULTEM 1000) by Core Tools Ltd. and is shown in Fig. 2.1 (c). Its central opening had a circular active geometric area of 2 cm2. The silicon rings were made in house using a procedure described in Appendix E. A 25 mm diameter membrane sample was sandwiched between the anodic and cathodic 16 mm diameter GDE samples. Two niobium current collectors (25 mm OD, 16 mm ID), having a thin current collecting metal cross through their circular openings were placed on each side of the MEA. Silicone seal rings (25 mm OD, 16 mm ID, 1 mm thickness) made in house were 52 placed on each side of the current collectors. Finally, a PTFE spacer ring (54273 AMG06, 25 mm OD, 16 mm ID, 1 mm thickness) was placed as the last layer on the cathode side to reduce friction when turning the cap to close the MEA holder. Figure 2.1: Schematic diagram of (a) the electrochemical glass cell, (b) the MEA components (c) the MEA holder. Current Collectors Gas Diffusion Electrodes Membrane Spacer Seals (b) Multistat and PC Mercury Sulphate Reference electrode Anode Current Collector Cathode Current Collector Magnetic Stirrer Magnetic Stirrer Anolyte Reservoir Catholyte Reservoir Flow Heater / Circulator Heater / Circulator Connection to/from Temperature Controlled Water Bath CO2 Outlet H2 Outlet O-rings MEA Holder Temperature Indicator Thermocouple Threaded Caps with Plugs (a) (c) 53 The glass cell, MEA holder and O-rings were joined and held together using a screw clamping apparatus. It was designed to press the thick flat ends of the glass cell compartments together and ensured alignment of the glass compartments. This permitted the application of a uniformly distributed force to seal the glass cell flanges to the MEA holder O-rings. All experiments carried out at room temperature were conducted in ambient air (23oC ± 2oC). For experiments at 50 and 75oC ± 1oC, H2O was circulated within the double wall of the glass cell and the temperature was controlled by a Haake DC-30 immersion circulator. Thermocouples were used to monitor the temperature in each of the glass cell compartments. Figure 2.2: Picture of (a) the electrochemical glass cell, (b) the MEA holder. (b) Current Collectors O-ring Gas Diffusion Electrode (a) Magnetic stirrers MEA holder Glass cell holder Multistat connectionsReference electrode Anodic compartment Cathodic compartment 54 2.2.3 Electrochemical measurements All electrochemical measurements were carried out at steady-state using a multistat (Solartron Analytical, Model 1470E) connected to a computer, using CoreWare software. A single junction mercury-mercurous sulphate (MSE, Hg/Hg2SO4, Radiometer Analytical) reference electrode located in the anode compartment was used for all experiments at ambient temperature. At 50 and 75oC, a double junction silver/silver chloride (SSE, Ag/AgCl, Radiometer Analytical) reference electrode was used, with 0.5 M H2SO4 in its outer compartment, as it is more stable than the MSE at higher temperatures. This is because the MSE has an outer porous plug tip while the SSE electrode outer tip is a porous glass frit, which limits the ion migration to increase with temperature. The reference electrode was placed in the anode compartment. All voltages are reported versus the standard hydrogen electrode (SHE). All electrolysis experiments were conducted in the galvanostatic, current-controlled mode. More information on the electrochemical technique used to obtain current/potential transients is given in Appendix F. AC impedance measurements were carried out with an impedance/gain phase analyzer (Solartron Analytical, Model 1260A) to obtain the MEA and solution resistance. Measurements were performed with an AC amplitude of 10 mV, in the frequency range of 10-3 to 107 Hz. More details on this electrochemical technique can be found in Appendix F. As the measured potential values were affected by the ohmic drop, they were corrected for the current resistance (IR) before kinetic parameter information was extracted. Using this technique, it was confirmed that the MEA resistance increased with increasing CH3OH concentration, MEA usage and electrode degradation due to carbon oxidation. 2.2.4 Characterization Cyclic voltammetric techniques were used to determine the true active electrochemical surface area (ECSA) of the catalysts in a conventional three electrode single chamber electrochemical cell, using a platinized Pt counter electrode (CE) (see Appendix E for platinization procedure), an Ag/AgCl double junction reference electrode (RE), and various working electrodes (WE). The platinization procedure employed can be found in Appendix E. The ECSA of the various Pt catalyst working electrodes (WEs) was estimated from the H2 adsorption charge on the cyclic voltammograms [168], while that of Pt-Ru catalysts was estimated by the copper under potential deposition (UPD) method [169, 170]. More details 55 on the CV technique employed can be found in Appendix F. The average total Pt area (TPA) and ECSA values obtained by H2 adsorption or Cu UPD for the different catalysts are summarized in Table 2.1. The values reported represent the average of three tests conducted on three different catalyst samples of the same batch of GDE. The Brunauer, Emmett and Teller (BET) surface area (SA) from the catalyst powder certificate of analysis were provided for comparison purposes only. The BET estimated ECSA exceeded the measured ECSA in all cases. The experimentally determined values were later used in the estimation of kinetic parameters and in Fig. 2.9. For the pure Pt disk, a surface roughness factor of 4 was assumed, giving an ECSA of 8 cm2/g Pt. These techniques were also used to confirm that the catalyst electrochemical area decreases with increasing utilization and carbon oxidation of the catalyst support. 56 Table 2.1: Electrochemical characterization of supported and unsupported catalysts studied. Total metal area (TMA) by Different Methods Electrochemical Surface Area (ECSA) by Different Methods BET SA* Typical BET SA Range Actual Loading Geometric Area Roughness Factor of 4 H2 adsorption Cu UPD Roughness Factor of 4 H2 adsorption Cu UPD Catalyst [m2/g] [m2/g] [mg/cm2] [cm2] [cm2 Pt] [cm2 Pt] [cm2 Pt and Ru] [cm2/g Pt] [cm2/g Pt] [cm2/g Pt and Ru] Pt-Ru/C 130 120-140 4.01 2.01 - - 562 - - 140 Pt-Ru/C 130 120-140 3.99 2.01 - - 562 - - 140 Pt/C 180 160-200 2.04 2.01 - 3929 - - 1926 - Pt/C 180 160-200 2.01 2.01 - 3746 - - 1863 - Pt-Ru black 86.65 80-90 3.98 2.01 - - 69 - - 17 Pt black 29 25 to 34 2.06 2.01 - 248 - - 120 - Pt disk - - - 2.01 8 - - 8 - - * From catalyst powder vendor certificate of analysis. 57 2.3 Results and Discussion 2.3.1 Thermodynamic evaluation of methanol reforming at low temperatures A theoretical thermodynamic analysis was carried out for a DMR in acidic media and a water PEM electrolyser as a function of temperature. The effect of pressure on the theoretical cell voltage and thermodynamic efficiency was found to be negligible compared to the effect of temperature, and hence is not discussed here. The thermodynamic information for the species involved, including the standard Gibbs free energy of formation (ΔGof), the standard enthalpy of formation (ΔHof), the standard entropy of formation (ΔSof), and the heat capacity (Cpo), are available in the literature [171, 172] and are listed in Appendix G. These values were used to establish the theoretical cell voltage (Ee) and theoretical thermodynamic efficiency (ηmax) as a function of temperature (25 to 150oC) assuming a constant pressure (1 atm). Examples of the thermodynamic calculations for the DMR electrochemical reactions in the liquid phase and in the gas phase are provided in Appendix G. The standard Gibbs free energy associated with the electrochemical reaction is given by: ( )reactantsproductsGsG o fiiio −Δ∑=Δ , (2.1) The standard enthalpy of the electrochemical reaction can be calculated by: o fiii o HsH ,Δ∑=Δ (2.2) Similarly, assuming that the entropy is independent of pressure, the standard entropy of formation of the electrochemical reaction can be determined by: o fiii o SsS ,Δ∑=Δ (2.3) 58 The standard reversible cell voltage can then be calculated by Eq. 1.6. Various Shomate equations of the form ∑ = += N n b noP nTaaC 1 (2.4) based on empirical data, were used to evaluate the heat capacity of the various species as a function of temperature. Over the temperature range of 25 to 50oC, the average change in the Cp values was about 8 % for methanol in the liquid phase, and over the temperature range of 100 to 150oC, the change in the Cp value was about 2 % for water in the gas phase, respectively. As demonstrated in Appendix G, the heat capacities hence obtained were used to evaluate the molar enthalpy and entropy of formation at temperature T, assuming a constant pressure, are given by: dTCHH p To T 298∫+Δ=Δ (2.5) dTC T SS p To T 1 298∫+Δ=Δ (2.6) For CH3OH electrolysis, over the temperature range of 25 to 50oC, the temperature correction was about 1 % for the ΔST value in the liquid phase and over the temperature range of 76.7 to 150oC, and the temperature correction was about 2 % for the ΔST value in the gas phase. The reversible cell voltage approximated as a function of temperature by: ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ Δ⋅−+= nF STEE Toe )298( (2.7) And the corresponding Gibbs free energy of formation was calculated: eT nFEG −=Δ (2.8) 59 Finally, the maximum thermodynamic efficiency, ηmax, can be calculated, at a temperature T and 1 atm, as per: 100max ⋅⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ Δ Δ= T T H Gη (2.9) These theoretical calculations assume that there is no gas forming before the fuel boiling point is reached and that there is no liquid present after the fuel boiling point is attained. The boiling point of the CH3OH-H2O mixture will change depending on the concentration. For the DMR thermodynamic calculations an equimolar mixture of H2O and CH3OH, which has a boiling point of 76.7oC, was assumed. The effect of the electrolyte is also ignored. The CO2 released at the anode of the DMR was assumed to be in the gas phase at all temperatures. It is important to note that the sign of the thermodynamic efficiency for the electrolysis systems can be positive or negative, depending on the temperature at which the system is evaluated. For both electrolysis systems, the enthalpy of formation was positive for the temperature range studied, indicating that both overall reactions are endothermic. The Gibbs free energy of formation for the water electrolysis system is positive (> 0) under the entire temperature range studied, indicating that the electrolysis of water is non-spontaneous. Hence, the thermodynamic efficiency of water electrolysis was positive at all temperature studied. Although, the methanol electrolysis Gibbs free energy of formation was positive at 25oC, indicating that the reaction is non-spontaneous at this temperature, the Gibbs free energy of formation was negative (< 0) at higher temperatures, indicating that the reaction became spontaneous. Under these conditions, the thermodynamic efficiency was a negative value. As shown in Fig. 2.3, the PEM H2O electrolyser resulted in a positive theoretical thermodynamic efficiency over the entire temperature range studied, while that of the DMR was positive at 25oC and became negative at higher temperatures. The theoretical thermodynamic efficiencies determined for the PEM H2O electrolyser indicate that work is required on the system in order to generate H2 under these conditions. The theoretical thermodynamic efficiencies determined for the DMR indicate that the system 60 requires work input to generate H2 at 25oC, but that work is generated by the system at higher temperatures. As shown in Fig. 2.3, for both systems, the theoretical cell voltage became less negative as temperature increased. In the case of the DMR, the theoretical cell voltage even became positive as temperature increased. It was determined that the DMR Gibbs free energy of formation and theoretical cell voltage of methanol electrolysis become positive at a temperature of about 41oC. Hence, according to thermodynamics at temperatures greater than 50oC, the DMR should be a spontaneous source of electricity, as well as H2. However, favourable reaction thermodynamics do not imply that the reaction is kinetically favourable. The Gibbs free energy of formation and theoretical cell voltage of the PEM H2O electrolyser, on the other hand, do not become positive over the temperature range investigated, i.e., up to 150oC. The thermodynamic analysis demonstrates which electrochemical reactions can occur spontaneously at specific conditions and equilibrium, but does not provide any kinetic information about the electrochemical reactions. The actual amount of electrical energy which will be required to generate H2 is equal to the Gibbs free energy of the electrochemical reaction plus the losses in the system. 61 Figure 2.3: Theoretical thermodynamic efficiency and theoretical cell voltage as a function of temperature (1 atm) for a DMR and PEM water electrolyser. Liquid Gas -100 -80 -60 -40 -20 0 20 40 60 80 100 0 20 40 60 80 100 120 140 160 Temperature (oC) T h e r m o d y n a m i c E f f i c i e n c y ( % ) -1.4 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 T h e o r e t i c a l C e l l V o l t a g e ( V ) Thermodynamic efficiency, Methanol Electrolyser Thermodynamic efficiency, water electrolyser Theoretical cell voltage, methanol electrolyser Theoretical cell voltage, water electrolyser Power Sink Power Source Boiling Point of Equimolar Methanol/Water Mixture (76oC) Boiling Point of Pure Water (100oC) GasLiquid Boiling Point of Pure Methanol (64.7oC) T h e r m o d y n a m i c E f f i c i e n c y ( % ) T h e o r e t i c a l C e l l V o l t a g e ( V ) 62 2.3.2 Electrolysis Polarizations 2.3.2.1 Cell voltage stability Typical responses obtained for different electrolysis current densities are shown in Fig. 2.4. The open circuit voltage (OCV) for the reaction was measured in an H-cell as described in Appendix F. At low current densities, steady-state was rapidly reached and the CH3OH electrolysis voltage stabilized. At high current densities, the CH3OH oxidation current alone was insufficient to sustain the high current density. The electrolysis voltage dropped off until water oxidation and carbon oxidation started to occur. Figure 2.4: Electrolysis cell voltage stability with respect to current density (16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC). Water electrolysis starts to occur at voltages below -1.23 V vs. SHE, and carbon oxidation, e.g., the anode oxidation catalyst support starts to oxidize at a cell voltage of about -1.8 V vs. SHE. Hence, a limiting current density is reached for the electro-oxidation of CH3OH beyond which no stable cell voltage measurements could be obtained. Although the experiments were done in the static mode, this effect has also been observed for CH3OH electrolysis in the active mode [118]. Voltage measurements were -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 -50 0 50 100 150 200 250 300 350 400 450 500 Time (s) C el l V ol ta ge (V ) 0.02 mA/cm2 200 mA/cm25 mA/cm 2 OCV 63 recorded at current densities where steady-state was reached, i.e., when the change in cell voltage was less than 20 mV over five minutes. The region of cell voltage stability varied as a function of current density, CH3OH concentration, temperature and the type of anode electrocatalyst. In the limiting current density region, steady-state could not be reached. Figure 2.5 shows an example of the electrolysis test repeatability in the different regions. In the stable region, the electrolysis polarization curve is very reproducible. The experiments conducted in the forward direction, from small current densities to large current densities, in the backward direction, from large current densities to small current densities, and in a random order of current densities all resulted in very similar voltages at a particular current density. When the test was started backward from a point in the unstable region, the results were still reproducible, but the error for the first data point in the stable region was slightly larger. When the test was started at a current density where the safety limit of -4 V was reached, irreversible damage was observed as the catalyst was oxidized and precious metal was lost from its surface. This was also verified by conducting cyclic voltammetric experiments on fresh, used, and oxidized supported catalyst coupons. Figure 2.5: IR corrected electrolysis cell voltage repeatability as a function of geometric current density in the stable and unstable regions (2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 2 mg/cm2 Pt/C anode and cathode, 23 ±2oC). -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 5 10 15 20 25 30 35 40 45 50 Geometric Current Density (mA/cm2) IR C or re ct ed C el l V ol ta ge (V ) Forward Backward Random Backward, from unstable region Backward, from unstable region Backward, after C oxidation 64 The data points shown in the figures represent the arithmetic average of values obtained for a minimum of three different samples. The calculated standard deviation, Sd, representing the upper and lower limits, was typically in the range of 10 mV but closer to 200 mV in the limiting current region. In the limiting current density region, voltage values were unstable and were empirically obtained by linear interpolation between the last stable data points and the safety limit, which was set at -4 V. 2.3.2.2 Effect of methanol concentration and anode catalyst In order to investigate the effect of CH3OH concentration on the polarization curve, the anode compartment CH3OH concentration was varied from 0 to 16 M (i.e., 0, 1, 2, 6, 16 M), with 0 M representing H2O electrolysis. Figure 2.6 shows the dependence of the cell voltage corrected for the ohmic losses on the current density for different CH3OH concentrations and represent typical electrolysis polarization curves. They show open-circuit voltages (OCV) less than the equilibrium cell voltage likely due to fuel cross-over; activation control at lower current densities resulting in a sharp initial decrease in cell voltage as the current density increases; a subsequent linear slow voltage decrease with current density due to ohmic control followed by a rapid decrease in cell voltage due to mass transport limitations. The limiting current densities, iL, are reached in the transport control region when the fuel is used at a rate equal to its maximum transport supply rate. At the limiting current density, voltages where other competing reactions can occur, such as H2O electrolysis and/or carbon oxidation, were reached. As can be seen in Fig. 2.6, the iL value varied with CH3OH concentration. At low current densities (<12 mA/cm2), the CH3OH concentration did not impact the cell voltage significantly, while at higher current densities, lower overpotentials were obtained at higher CH3OH concentrations. As expected, for CH3OH oxidation the required electrolytic voltages were significantly less than for H2O electrolysis at the same current density. In contrast with active CH3OH electrolysis studies where the anode reactants are continuously flowing, the static system electrolytic polarizations are more sensitive to CH3OH concentration. Hu et al. [7] reported that the electrolysis performance was not dependent on the CH3OH concentration at concentrations greater than 2 M at 60oC, but that the concentration polarization was obvious at high current densities for concentrations lower than 1 M. In another study, by Shen et al. [11], the voltage losses for CH3OH electrolysis were reported to be almost the same in the 2-4 M range, while the voltage losses were greater at concentrations lower or higher than this range. At low CH3OH 65 concentrations, high current densities could not be sustained. At high CH3OH concentrations, less voltage was required to hold the same current densities and higher current densities could be sustained. Methanol permeation from the anode compartment to the cathode compartment will increase with increasing CH3OH concentrations and will decrease fuel efficiency. However, in the system under study, CH3OH present in the cathode compartment can not be oxidized at the reducing cathode potentials observed during electrolysis. In their study, Hu et al. attributed the poor electrolysis performance at high CH3OH concentration to the creation of a mixed potential at the cathode, which in turn reduced its electrocatalytic activity. Other phenomena such as adsorption and catalyst site blockage were not considered but are possible. As the concentration of CH3OH increases, the solution conductivity will decrease, and the mass transfer limitations will decrease, which will affect rate of CH3OH electrolysis. As the data presented in Fig. 2.6 is for catalysts having different ECSAs, it is not possible to make a direct comparison of the performance of the three-dimensional catalysts studied for the electrolysis of CH3OH or H2O on a geometric current density basis. Nevertheless, Fig. 2.6 (e) clearly demonstrates that the well-defined low electrochemical surface area non-porous Pt disk did not perform as well as the dispersed three-dimensional electrode catalysts. Furthermore, it can be seen that the Pt disk electrode, for which no carbon is present, was able to sustain greater current densities for H2O electrolysis than for CH3OH oxidation. Interestingly, for the 1 M CH3OH case with the well-defined Pt disk at low current densities, two different potential values were obtained for the same current density. In this system, proton conductivity of the MEA was provided by the electrolyte in the cathodic compartment. The bi- stability was observed in multiple tests which were performed separately at the same conditions. This phenomenon was also observed for a PEM fuel cell running at 37oC using H2 partial pressure less than 1.5 kPa, and 4 mg/cm2 Pt-Ru black anode and a 4 mg/cm2 Pt black cathode [173]. It was attributed to the total resistance of Nafion®, which is split in an ohmic component, due to proton migration in the membrane’s aqueous phase, and a non-ohmic one, due to proton tunnelling in the membrane. Figure 2.6: Effect of methanol concentration on the IR corrected electrolysis cell voltage as a function of geometric current density for different anode catalysts (a) 4 mg/cm2 Pt-Ru/C, (b) 2 mg/cm2 Pt/C, (c) 4 mg/cm2 Pt-Ru black, (d) 2 mg/cm2 Pt black, (e) well-defined Pt disk (0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 23 ±2oC). 66 (a) (b) (c) (d) (e) -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Geometric Current Density (mA/cm2) I R C o r r e c t e d C e l l V o l t a g e ( V ) 0 M CH OH 1 M CH OH 2 M CH OH 6 M CH OH 16 M CH OH 3 3 3 3 3 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Geometric Current Density (mA/cm2) I R C o r r e c t e d C e l l V o l t a g e ( V ) 0 M CH OH 2 M CH OH 16 M CH OH3 3 3 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Geometric Current Density (mA/cm2) I R C o r r e c t e d C e l l V o l t a g e ( V ) 0 M CH OH 2 M CH OH 16 M CH OH3 3 3 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Geometric Current Density (mA/cm2) I R C o r r e c t e d C e l l V o l t a g e ( V ) 0 M CH OH 2 M CH OH 16 M CH OH3 3 3 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 Geometric Current Density (mA/cm2) I R C o r r e c t e d C e l l V o l t a g e ( V ) 0 M CH OH 1 M CH OH (low) 1 M CH OH (high) 2 M CH OH 6 M CH OH 16 M CH OH 3 3 3 3 3 3 I R C o r r e c t e d C e l l V o l t a g e ( V ) I R C o r r e c t e d C e l l V o l t a g e ( V ) I R C o r r e c t e d C e l l V o l t a g e ( V ) I R C o r r e c t e d C e l l V o l t a g e ( V ) I R C o r r e c t e d C e l l V o l t a g e ( V ) 67 Figure 2.7 shows that the iL values varied with the CH3OH concentration and the type of anode catalyst used. Figure 2.7: Dependence of limiting geometric current densities for the oxidation of methanol and water on the CH3OH concentration for various anode catalysts (0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 23 ±2oC). For CH3OH oxidation, the largest limiting current densities were reached with the Pt-Ru/C catalyst. For H2O electrolysis, the largest limiting current density was reached with the well-defined Pt disk. However, the iL values obtained for this catalyst are lower in the case of CH3OH oxidation than for H2O oxidation. Although this well-defined Pt catalyst is not affected by carbon oxidation thus allowing higher current densities to be achieved, its low electrochemical surface area was strongly affected by CO adsorption in the case of CH3OH oxidation. The iL values for the Pt/C catalyst were lower than the one for the unsupported catalyst at low concentrations, while the iL values were similar to the unsupported catalyst ones at high concentration. The iL values for the 4 mg/cm2 Pt-Ru black catalyst and for the 2 mg/cm2 Pt black catalyst were within the same range. It would be expected that the supported catalysts, which possess higher active areas, would result in higher limiting current densities. Yet, this was not observed for Pt/C and Pt black. Take et al. [118] determined limiting current densities for an active CH3OH electrolysis electrolytic cell using 5.4 cm diameter Pt catalyst samples (real surface area of 23 cm2) at the anode and cathode, of 130 mA/cm2 for 1 M CH3OH, 170 mA/cm2 for 2 M CH3OH, 210 mA/cm2 for 6 M 0 20 40 60 80 100 120 140 160 0 2 4 6 8 10 12 14 16 18 Methanol Concentration (M) Li m itin g G eo m et ric C ur re nt D en si ty (m A /c m 2 ) 4 mg/cm Pt-Ru/C 2 mg/cm Pt/C 4 mg/cm Pt-Ru black 2 mg/cm Pt black Well-defined Horizontal Pt Disk 2 2 2 2 68 CH3OH, and 260 mA/cm2 for 16 M CH3OH. These reported values are greater than the limiting geometric current densities shown in Fig. 2.7 for the same CH3OH concentrations using the 2 mg/cm2 supported or unsupported Pt catalyst. In the passive mode, the transport of the fuel to the electrode is slower and its access to the electrode is more restrained by the CO2 production. This resulted in lower limiting current densities compared to the active mode. Figure 2.8 shows the individual IR corrected anode potential and cathode potential, respectively, obtained for stable voltage values before the limiting current density region. The change in cathode potential is insignificant compared to the change in anode potential. As the cathode potential remained more or less the same, it can be deduced that the increase in cell voltage required is mainly due to the anode. It can also be seen that higher CH3OH concentrations reduce the anode overpotential. Figure 2.8: Effect of CH3OH concentration on the individual electrode potential (0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC). As shown in Fig. 2.8, the electrolysis of H2O in 0.5 M H2SO4 commences at the anode at a potential of about 1.3 V vs. SHE (Eq. 1.21). At anode potentials equal to and greater than this value for the CH3OH oxidation reaction, the current efficiency is not 100 %, as there is a current loss due to the H2 evolution -0.4 0.0 0.4 0.8 1.2 1.6 0 25 50 75 100 125 150 Geometric Current Density (mA/cm2) A no de a nd C at ho de P ot en tia l ( V vs . S H E) 0 M CH OH 1 M CH OH 2 M CH OH 6 M CH OH 16 M CH OH Anode Potential Cathode Potential 3 3 3 3 3 69 reaction. Beyond this point, the H2O electrolysis current density should be subtracted from the CH3OH electro-oxidation current density. At higher cell voltages, the current density is also increasingly sustained by carbon corrosion. According to thermodynamics, carbon can be oxidized on the anode at a standard potential of -0.21 V vs. SHE: C + 2H2O → CO2 + 4H+ + 4e- Eoa = -0.21 V vs. SHE (2.7) C + H2O → CO + 2H+ + 2e- Eoa = -0.52 V vs. SHE (2.8) However, the carbon oxidation is kinetically-inhibited and it does not usually begin until a potential of about 1.8 V vs. SHE is reached at the anode [174]. Beyond this potential, the cell current density is increasingly sustained by carbon corrosion. This was confirmed by the observation of a decrease in cell voltage after re-utilization of catalysts which were subjected to anode potentials greater than this value. The degradation of the catalyst carbon support and the consequent loss in electrochemical surface area was also confirmed by cyclic voltammetry (refer to Appendix F for more details). Furthermore, Ru can dissolve at anodic potentials greater than the standard potential of -0.25 V vs. SHE, as per the following electrochemical reactions: Ru → Ru2+ + 2e- Eoa = -0.46 V vs. SHE (2.9) Ru2+ → Ru3+ + e- Eoa = -0.25 V vs. SHE (2.10) Hence, as the absolute current density increases and the anode potential become more positive, a decrease in the CH3OH oxidation performance due to Ru loss can also be anticipated. However, this is likely kinetically limited, similar to the case of carbon oxidation. The impact of Ru dissolution and crossover on the performance of a PEMFC and on the functionality of the anode was studied in the literature [175]. 70 Figure 2.9 compares the performance of Pt/C and Pt black catalysts on a real electrochemical current density basis for different CH3OH concentrations. Both catalysts had the same catalyst loading of 2 mg Pt/cm2 and a Nafion® content of 30 wt%, even though the optimized amount of ionomer required for the unsupported catalyst is likely different than for the supported catalyst. Figure 2.9: Effect of the IR corrected voltage on electrolysis for Pt/C and Pt black anode catalysts as a function of real electrochemical current density (0, 2, or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 23 ±2oC). The Pt black catalyst had a much smaller ECSA than the Pt/C catalyst. It is possible that the non- optimized Nafion® content partly contributed to a decrease in the surface area of the unsupported catalyst. However, the required applied voltage to oxidize H2O or CH3OH was less for the Pt black catalyst than for the Pt/C catalyst at the same current density. Thus, the Nafion® loading used did not restrain the access to the active surface sites of the black catalyst and did not render it less active than the Pt/C catalyst. Moreover, the Pt black catalyst was able to sustain higher real current densities than the Pt/C catalyst at the same CH3OH concentration. It appears that carbon oxidation started to occur at lower real current densities for the Pt/C electrode than for the Pt black electrode. This suggests that the carbon of the supported catalyst oxidizes before the carbon of the carbon fibre paper on which the both catalysts were sprayed start to oxidize. -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 Real Current Density (mA/cm2) IR C or re ct ed C el l V ol ta ge (V ) Pt/C, 0 M CH OH Pt/C, 2 M CH OH Pt/C, 16 M CH OH Pt Black, 0 M CH OH Pt Black, 2 M CH OH Pt Black, 16 M CH OH 3 3 3 3 3 3 71 2.3.2.3 Effect of dry purged cathode Experiments were also conducted with a dry cathode, which was purged with N2 gas, to confirm the feasibility of this possible design simplification. Figure 2.10 compares the polarization curves for a DMR running with the cathode compartment filled with 0.5 M H2SO4 and for one with a dry cathode compartment purged with N2. The cell voltage became more negative for the DMR operating with a dry purged cathode compartment compared to the DMR with a cathode compartment containing 0.5 M H2SO4. Figure 2.10: Effect of dry N2 purged cathode on the IR corrected electrolysis cell voltage as a function of geometric current density (0 or 1 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, wet electrolyser catholyte: 0.5 M H2SO4, dry electrolyser catholyte: N2 purge electrolyser, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC). In the case of CH3OH electrolysis, slightly more voltage is required to drive the electrochemical reaction when the cathode does not contain any electrolytic solution. In both cases, the required voltage for the dry and wet cases were within 10 % at high current densities and similar limiting current densities were attained in the dry and wet cathode case for 0 and 1 M CH3OH. However, the AC impedance (Appendix F) results indicated that the ohmic resistance across the MEA can be up to about 55 % higher in the dry cathode case compared to the wet cathode one, i.e., 0.22 Ω versus 0.1 Ω for 1 M CH3OH. This indicates -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 Geometric Current Density (mA/cm2) IR C or re ct ed C el l V ol ta ge (V ) 0 M CH OH, wet cathode 0 M CH OH, dry cathode 1 M CH OH, wet cathode 1 M CH OH, dry cathode 3 3 3 3 72 humidification of the membrane by diffusion of H2O from the liquid anode side to the dry cathode side in this case might be insufficient to maintain the membrane conductivity. Of course, membrane humidification could be improved by decreasing the membrane thickness and pressurizing the anode side, for example. Fig. 2.11 shows the corresponding anode and cathode potentials respectively, for the wet and dry cathode cases. The cathode potential remained largely unaffected by the wet or dry cathode operation compared to the anode potential. No tests were carried out using humidified N2. Figure 2.11: Effect of dry N2 purged cathode and a wet cathode on the (a) anode potential and (b) cathode potential, as a function of geometric current density (0 or 1 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, wet electrolyser catholyte: 0.5 M H2SO4, dry electrolyser catholyte: N2 purge, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC). Fig. 2.11 shows that for CH3OH electrolysis the anode potential was significantly larger with the dry purged cathode than with the wet cathode. The CH3OH cross-over and MEA resistance are likely greater in the dry N2 case. It is not clear why the anode potential is larger in the dry cathode case. At high enough current densities in both modes, the anode potential is high enough that H2O electrolysis will also occur in parallel. Using a dried N2 purged cathode compartment seems to represent a possible design simplification for the liquid CH3OH electrochemical reformer. -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 0 5 10 15 20 25 30 35 40 45 50 Geometric Current Density (mA/cm2) A no de a nd C at ho de P ot en tia l (V v s. S H E) 0 M CH OH, wet cathode 0 M CH OH, dry cathode 1 M CH OH, wet cathode 1 M CH OH, dry cathode (a) Anode Potential (b) Cathode Potential 3 3 3 3 73 2.3.2.4 Effect of temperature The effect of temperature was studied for CH3OH and H2O electrolysis using the 4 mg/cm2 Pt-Ru/C catalyst. As can be seen in Fig. 2.12, increasing the operating temperature significantly improved the Pt- Ru/C electrode kinetics of the CH3OH electrolysis. There was a large improvement in the electrolyser performance for all catalyst types with temperature. For example, for the 16 M CH3OH solution and a current density of 150 mA/cm2, the cell voltage was -1.55 V, -1.26 V and -0.87 V corresponding to temperatures of 23, 50 and 75oC, respectively. Hence, there was a reduction in the electrolyser cell voltage of 31 % between 50 and 75oC. The reduced electrolytic overpotentials resulted from an improvement in the mass transport and in the electrode kinetics at the higher temperature. Compared to the limiting current density at 23oC, at 50oC, the limiting current density was about 25 mA/cm2 higher in the case of the 2 M CH3OH solution, and 50 mA/cm2 higher in the case of the 16 M CH3OH solution. However, the increase in limiting current densities between 50 and 75oC was not significant, and a similar limiting current was obtained at both temperatures in the case of water and CH3OH electrolysis. This is not in agreement with what has been previously observed for active CH3OH electrolysis systems [7, 11]. For example, Shen et al. [11] using a 2.5 mg/cm2 Pt-Ru/C anode and a 1.5 mg/cm2 Pt/C cathode in a 2 M CH3OH active system shown that over the temperature range of 50 to 80oC, the electrolysis cell voltage at 80oC was reduced to half of the electrolysis cell voltage at 50oC. There is no indication in this paper that measures were taken to prevent CH3OH loss to the vapour phase. However, as a closed-loop CH3OH-H2O anode feeding system, it can be assumed that any CH3OH evaporating would be condensing back in the CH3OH-H2O storage tank, and that the concentration of CH3OH fed to the anode is constant. In our static system set-up, the anolyte is not constantly replenished and any CH3OH evaporating is loss. This may explain why the electrolyser cell voltage did not decrease as significantly as for active systems between 50 and 75oC in our experiments. As the 16 M CH3OH solution has a boiling point of 76.7oC, and our study uses a non-flowing system, a partial loss of solution to the vapour phase was possible. Using a K2Cr2O7 reduction technique combined with spectophometric measurements, which is described in Appendix E, it was verified that, the CH3OH concentration change for a 16 M CH3OH at 75oC was less than 2 M over the period of time for which polarization measurements were taken. This small concentration change might partly explain why the 74 limiting current density was similar at 50 and 75oC in the case of the 16 M CH3OH solution. Another possible explanation may be that there is a greater crossover rate of CH3OH at 16 M CH3OH, which may affect the kinetics of the H2 production at the cathode. However, in the case of H2O electrolysis, while minor improvements in the anode kinetics were observed with increasing temperature, the limiting current also remained the same over the temperature range of 23 to 75oC studied. Therefore, CH3OH crossover can not explain why similar limiting current densities were obtained for the electrolysis of 2 and 16 M CH3OH at 50 and 75oC. Nevertheless, it may be preferable to operate the CH3OH electrolyser at higher temperatures as less current would be wasted for secondary reactions and performance is better under these conditions. 75 Figure 2.12: Effect of CH3OH concentration on the IR corrected electrolysis cell voltage as a function of geometric current density for different temperatures (a) 23 ±2oC (b) 50 ±1oC (c) 75 ±1oC (0, 2 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode). -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 Geometric Current Density (mA/cm2) IR C or re ct ed C el l V ol ta ge (V ) 0 M CH OH 2 M CH OH 16 M CH OH 3 3 3 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 Geometric Current Density (mA/cm2) IR C or re ct ed C el l V ol ta ge (V ) 0 M CH OH 2 M CH OH 16 M CH OH 3 3 3 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 Geometric Current Density (mA/cm2) IR C or re ct ed C el l V ol ta ge (V ) 0 M CH OH 2 M CH OH 16 M CH OH 3 3 3 (a) (b) (c) IR C or re ct ed C el l V ol ta ge (V ) IR C or re ct ed C el l V ol ta ge (V ) IR C or re ct ed C el l V ol ta ge (V ) 76 2.3.2.5 Anode Tafel kinetics In the case CH3OH and H2O electrolysis, the anodic reaction kinetics is slow, resulting in a high overpotential. The overvoltage at the cathode is negligible compared to the overvoltage at the anode. The Buttler-Erdey-Grutz-Volmer (BEV) equation can be simplified to the Tafel equation as per a well- known treatment described elsewhere [126, 176] and in more details in Appendix F. The anodic form of the Tafel equation was used to make Tafel plots (log(i) vs. ηa) from which the Tafel slope and the Tafel constant were extracted to obtain kinetic information on the CH3OH and H2O electrolysis. Fig. 2.13 shows an example of the IR corrected anode overpotential vs. log of the geometric current density for different CH3OH concentrations, demonstrating the linearity of the Tafel lines obtained. Figure 2.13: Tafel plot of the IR corrected anodic overpotential as a function of the log of the geometric current density (0, 2, or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt black anode, 2 mg/cm2 Pt/C cathode, 23 ±2oC). The slope of the anodic Tafel curves for the CH3OH electrolysis changed at about 0.6 V for the Pt and Pt-Ru catalysts, indicating two distinct kinetic control regions. Wu et al. [177] have reported that these two regions represent two different reaction mechanisms. In the first region, where ηa < ~0.6 V vs. SHE, the splitting of the first C-H bond with the first electron charge transfer reaction is the rate-determining 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8 5.2 -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 Log(i) (mA/cm2) IR C or re ct ed A no di c O ve rp ot en tia l ( V v s S H E ) 0 M CH3OH 2 M CH3OH 16 M CH3OH Limiting current Limiting current Limiting current 77 step, while in the second region, where ηa > ~0.6 V vs. SHE, the reaction between the adsorbed CO on Pt sites is the rate-determining step [178]. For H2O electrolysis on the other hand, only one apparent slope was obtained over the entire potential range for the different catalysts studied. Hence, it is likely that there is only one reaction mechanism over the entire region. However, Kinoshita [179] reported that two distinct Tafel slopes were obtained for most metal oxide based catalysts studied in 1 N H3PO4 at 25oC. They reported a low Tafel slope value at low current densities and a high Tafel slope value at high current densities. Tafel parameters reported in the literature for Pt and Pt-Ru catalysts for CH3OH and H2O electrolysis, along with the values obtained in this study, are summarized in Table 2.2 and 2.3, respectively. For the case of the 4 mg/cm2 Pt-Ru anode, the exchange current density values, io, were approximated by using twice the ECSA determined for the 2 mg/cm2 Pt catalyst, as well as the ECSA determined through Cu UPD, for comparison purposes. This is indicated by the superscript 1 and 3, respectively, as indicated in the legend provided at the bottom of Table 2.3. 78 Table 2.2: Literature Tafel kinetic parameters for the electro-oxidation of methanol and water on Pt and Pt-Ru catalysts. Experimental Conditions Anode Overpotential Tafel Slope Exchange Current Density Electrode [H2SO4] [CH3OH] T ηa [mV vs. SHE] b [mV/dec] io [A/cm2] Ref. Method Catalyst Description [M] [M] [oC] Zone I Zone II Zone I Zone II Zone I Zone II IR, P. Pt (111) 0.2 cm OD spheres 0.1 0.2 20 <422 - 119 - 1.00x10 -2 - [180] IR, P. Pt (110) 0.2 cm OD spheres 0.1 0.2 20 <422 - 123 - 1.00x10 -2 - [180] IR, P. Pt (100) 0.2 cm OD spheres 0.1 0.2 20 <422 - 63 - 1.00x10 -2 - [180] IR, P. Pt (111) 0.2 cm OD spheres 0.1 0.2 20 <422 - 106 - 6.31x10 -6 - [181] IR, P. Pt (110) 0.2 cm OD spheres 0.1 0.2 20 <422 - 123 - 6.31x10 -6 - [181] M, P. Pt/C 14 ug/cm2, Nafion® film on GC. 0.5 0.5 60 <550 - 80 - - - [182] P. Pt black 5.6 and 4.2 mg/cm2, on N- 315. 0.5 1 16-20 400-700 400-700 100-140 100-140 - - 181] Rs, P. Pt black 5.6 and 4.2 mg/cm2, on N- 315. 0.5 1 16-20 <500 >600 60 120 4.00x10-8 1.55x10-6 [183] Rs, P. Pt/C 40 wt%, on GC. 0.5 0.1-2 30-80 <650 - 80-94 - 2.00x10-8 - [184] IR, G. Pt-Ru/C 40 wt%, 1:1, 3.2 mg/cm2, on GC. 0.5 0.5 40-70 <550 >550 108-128 300-700 5.24x10 -5 8.95x10-4 [177] IR, G. Pt-Ru/C 40 wt%, 1:1, 3.2 mg/cm2, on GC. 0.5 0.5 40 <550 >550 128 525 2.51x10 -4 7.41x10-4 [177] IR, G. Pt-Ru/C 40 wt%, 1:1, 3.2 mg/cm2, on GC. 0.5 0.5 40 <550 >550 123 647 3.51x10 -5 8.19x10-4 [177] IR, G. Pt-Ru/C 40 wt%, 1:1, 3.2 mg/cm2, on GC. 0.5 0.5 40 <550 >550 116 465 1.48x10 -6 2.24x10-4 [177] IR, G. Pt-Ru/C 40 wt%, 1:1, 3.2 mg/cm2, on GC. 0.5 0.5 70 <550 >550 114 650 1.64x10 -5 2.27x10-3 [177] IR, G. Pt-Ru/C 40 wt%, 1:1, 3.2 mg/cm2, on GC. 0.5 0.5 70 <550 >550 124 777 1.00x10 -5 1.00x10-3 [177] IR, G. Pt-Ru/C 40 wt%, 1:1, 3.2 mg/cm2, on GC. 0.5 0.5 70 <550 >550 108 537 1.93x10 -7 3.16x10-4 [177] M, P. Pt-Ru/C 14 ug/cm2, Nafion® film on GC. 0.5 0.5 60 <550 - 195 - - - [182] Rs, P. Pt-Ru black 5.6 and 4.2 mg/cm2, on N- 315. 0.5 1 16-20 <500 >600 70 163 2.88x10-7 2.88x10-6 [183] Rs, P. Pt-Ru/C 30 wt%, 1:1, on GC 0.5 0.1-2 30-80 <650 - 140-192 - 4.50x10 -12 - [184] 79 Experimental Conditions Anode Overpotential Tafel Slope Exchange Current Density Electrode [H2SO4] [CH3OH] T ηa [mV vs. SHE] b [mV/dec] io [A/cm2] Ref. Method Catalyst Description [M] [M] [oC] Zone I Zone II Zone I Zone II Zone I Zone II Rs, IR, V. Pt-Ru/C 40 wt%, 1:1, 1 mg/cm2, on TGPH-60. 1 1 20-80 <750 - 159-169 - - - [185] V. Pt-Ru/C 1.5 mg/cm2, carbon cloth with GDL, N-117. 0.5 0.1, 1 20-45 - - 60 - - - [186] V. Pt/C 1.5 mg/cm2, carbon cloth with GDL, N-117. 0.5 0.1, 1 25 - - 80 - - - [186] V. Pt/C 1.5 mg/cm2, carbon cloth with GDL, N-117. 0.5 0.1, 1 45 - - 90 - - - [186] R. Pt 0.5 1 Room T - - 125 - - - [187] F, P. Pt/C 20 wt%, 50-100 ug/cm2, on GC. 0.5 0.1 100 >330 - 83 - - - [188] R Pt (oxide free) - 0.1 N - - - - 70 - 2.00x10 -11 - [179] G Pt - 1 N - 25 - - 65 - 1.00x10-9 - [179] G Pt black - 20 wt% - 70 - - 65 - 4.00x10-10 - [179] R Pt - 0.05 - 25 - - 145 - 3.00x10-6 - [179] - Ru - 1 N - 25 - - 80 - 1.00x10-12 - [179] - Pt - 1 N 20 - - 107 - 7.60x10-10 - [179] - Pt (111) - 1 N - 20 - - 120 - 3.70x10-10 - [179] - Pt (100) - 1 N - 20 - - 114 - 6.40x10-10 - [179] - Pt - 1 N - 80 - - 90 - 1.30x10-11 - [179] - Pt-Ru 50 atom % Ru 1 N - 80 - - 120 1.80 x10-8 - [179] - Ru - 1 N - 80 - - 41 - 5.10 x10-9 - [179] P RuO2 - 1 - - - - 30-40 50-65 5-15 x10-9 - [179] P, IR RuOx - 1 N - 25 - - 32 42 7.00 x10-13 1.00x10-11 [179] Legend: all italic values are average values estimated from graphs, F = flow through, G = galvanostatic mode (constant current applied vs. time, monitor V vs. time), IR = IR corrected current density, M = mass specific current density, P = potentiostatic mode (constant potential applied vs. time, monitor I vs. time), R = Rotating disk electrode using the Levich method, RS = real surface area current density, V = voltammetry. 80 Table 2.3: Tafel kinetic parameters from current work for the electro-oxidation of methanol and water on Pt and Pt-Ru catalysts. Experimental Conditions Anode Overpotential Tafel Slope Exchange Current Density Electrode [H2SO4] [CH3OH] T ηa [mV vs. SHE] b [mV/dec] io [A/cm2] Method Catalyst Description [M] [M] [oC] Zone I Zone II Zone I Zone II Zone I Zone II G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 0 23 <600 >600 125.90 125.90 7.33x10-16 7.33x10-16 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 1 23 <600 >600 134.50 557.9 9.77x10-8 1.27x10-4 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 2 23 <600 >600 159.70 649.20 3.46x10-7 1.50x10-4 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 6 23 <600 >600 191.6 1407.8 2.89x10-6 1.60x10-3 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 16 23 <600 >600 186.40 274.60 8.22 x10-7 1.03x10-5 G, IR. Pt/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 0 23 <600 >600 176.60 176.60 1.94x10-12 1.94x10-12 G, IR. Pt/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 2 23 <600 >600 113.40 447.80 1.42x10-8 7.53x10-5 G, IR. Pt/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 16 23 <600 >600 104.40 284.90 5.40x10-9 2.19x10-5 G, IR. Pt-Ru black 50:50, 4 mg/cm2, on TGPH-60, N-117.3 0.5 0 23 <600 >600 244.80 244.80 4.40x10-9 4.40x10-9 G, IR. Pt-Ru black 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 2 23 <600 >600 144.80 701.60 5.45x10-6 1.16x10-2 G, IR. Pt-Ru black 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 16 23 <600 >600 156.10 527.90 1.20x10-5 7.13x10-3 G, IR. Pt black 2 mg/cm2.4 0.5 0 23 <600 >600 150.30 150.30 1.19x10-12 1.19x10-12 G, IR. Pt black 2 mg/cm2.4 0.5 2 23 <600 >600 147.20 640.60 3.10x10-6 5.03x10-3 G, IR. Pt black 2 mg/cm2.4 0.5 16 23 <600 >600 142.00 522.30 2.11x10-6 2.74x10-3 G, IR. Pt Disk 0.25 mm thick, 99.99%+ Pt. 2 0.5 0 23 <600 >600 120.60 120.6 6.54x10-77 6.54x10-27 G, IR. Pt Disk 0.25 mm thick, 99.99%+ Pt. 2 0.5 1 23 <600 >600 400.30 400.30 7.75x10-7 7.75x10-7 G, IR. Pt Disk 0.25 mm thick, 99.99%+ Pt. 2 0.5 1 23 <600 >600 49.30 408.70 1.19x10-15 6.60x10-5 G, IR. Pt Disk 0.25 mm thick, 99.99%+ Pt. 2 0.5 2 23 <600 >600 143.90 325.20 1.29x10-5 2.37x10-3 G, IR. Pt Disk 0.25 mm thick, 99.99%+ Pt. 2 0.5 6 23 <600 >600 145.50 393.90 3.70x10-5 8.68x10-3 G, IR. Pt Disk 0.25 mm thick, 99.99%+ Pt. 2 0.5 16 23 <600 >600 155.30 533.20 1.24x10-4 4.49x10-2 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 0 50 <600 >600 359.60 359.60 5.83x10-12 5.83 x10-12 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 2 50 <600 >600 145.40 140.00 4.52x10-10 3.08 x10-4 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 16 50 <600 >600 126.20 216.10 3.10x10-10 3.34x10-8 81 Experimental Conditions Anode Overpotential Tafel Slope Exchange Current Density Electrode [H2SO4] [CH3OH] T ηa [mV vs. SHE] b [mV/dec] io [A/cm2] Method Catalyst Description [M] [M] [oC] Zone I Zone II Zone I Zone II Zone I Zone II G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 0 75 <600 >600 325.50 325.50 1.43x10-11 1.43x10-11 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 2 75 <600 >600 206.10 285.90 2.98x10-7 1.59x10-6 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117.1 0.5 16 75 <600 >600 215.10 222.30 2.93x10-7 2.98x10-7 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 0 23 <600 >600 - - 3.01x10-24 3.01x10-24 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 2 23 <600 >600 - - 5.77x10-6 8.24x10-4 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 16 23 <600 >600 - - 1.33x10-5 2.22x10-2 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 0 50 <600 >600 - - 7.90x10-11 7.90x10-11 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 2 50 <600 >600 - - 6.12x10-9 4.17x10-3 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 16 50 <600 >600 - - 4.20x10-9 4.53x10-7 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 0 75 <600 >600 - - 1.93x10-10 1.93x10-10 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 2 75 <600 >600 - - 4.03x10-6 2.15x10-5 G, IR. Pt-Ru/C 50:50, 4 mg/cm2, on TGPH-60, N-117. 3 0.5 16 75 <600 >600 - - 3.98x10-6 4.03x10-6 Legend: all italic values are average values estimated from graphs, F = flow through, G = galvanostatic mode (constant current applied vs. time, monitor V vs. time), IR = IR corrected current density, M = mass specific current density, P = potentiostatic mode (constant potential applied vs. time, monitor I vs. time), R = Rotating disk electrode using the Levich method, RS = real surface area current density, V = voltammetry. 1Based on the electrochemical surface area of a 2 mg/cm2 Pt/C catalyst determined by H2 adsorption, 2 Geometric surface area assuming a roughness factor of 4, 3Based on the electrochemical surface area of a 4 mg/cm2 Pt-Ru black catalyst determined by Cu UPD, 4Based on the electrochemical surface area of a 2 mg/cm2 Pt black catalyst determined by H2 adsorption. 82 A proper comparison with the literature is difficult due to differences in the experimental conditions and whether geometric or real surface area was used. The type of electrolyte, the electrolyte and CH3OH concentration, as well as the catalyst loading used varied among the studies and most Tafel parameters available were mainly obtained using single-compartment cells without membranes. In general, the reported Tafel slope using Pt-Ru catalysts ranged from 70 to 128 mV/dec at low overpotentials, and ranged from 75 to 777 mV/dec at high overpotentials. This is comparable to the 4 mg/cm2 Pt-Ru/C catalyst Tafel slopes obtained at low CH3OH concentrations, but smaller than the ones obtained at high CH3OH concentrations in our work. Amongst different systems for which Tafel slopes were reported, the one used by Aramat et al. [183] seems to be the closest to our system. Tafel data was obtained using a two-compartment cell for unsupported catalyst coated with Nafion® 315 at 16-20oC in 1 M CH3OH and 0.5 M H2SO4. The Tafel slopes were in the range of 100-140 mV/dec for Pt black- Nafion® 315 in the overvoltage range of 0.4 to 0.7 V vs. SHE, 70 mV/dec for Pt-Ru black-N315 at ηa < 0.5 V vs. SHE and 163 mV/dec for the same catalyst at ηa > 0.6 V vs. SHE. The Tafel slopes obtained in our work using a 2 mg/cm2 Pt black GDE anode and N117 membrane, were in agreement with the Pt-N315 reported values for ηa < 0.6 V vs. SHE and CH3OH concentrations equal to or lower than 2 M. However, the Tafel slope values obtained in this study for CH3OH electrolysis with the Pt black catalyst at greater CH3OH concentrations and at ηa > 0.6 V vs. SHE are up to four times higher than the ones reported in the literature. This may be linked to a change in the rate determining step at higher potentials, as it is known that CO stripping from an unsupported Pt-Ru catalyst surface normally occurs at potentials ranging from about 0.5 to 0.7 V vs. SHE at 25oC [189, 190]. Tafel slopes were also evaluated as a function of temperature using the 4 mg/cm2 Pt-Ru/C anode at three different CH3OH concentrations. Zhang et al. [184] reported that at low temperatures, non-linearity is due to CO poisoning, while at high temperature, it is due to concentration polarization. They also noticed that Tafel slope data reported in the literature conflictingly increase or decrease with increasing temperature. This was attributed to different rate determining steps. In our study, the Tafel slopes values obtained for Pt-Ru/C were larger at ambient temperature than at 50oC, and the values obtained for the same catalyst at 75oC were the largest of all. Although it might be possible that the catalyst went through degradation at higher temperature, it is also possible that operation at 50oC is more desirable for 83 the Pt-Ru/C catalyst as it improves its activity for the electrolysis of CH3OH. From an efficiency standpoint, faster oxidation kinetics might allow the use of higher CH3OH concentrations. In general, reported Tafel slope values for Pt-Ru/C varied between 108 to 195 mV/dec in the temperature range of 60 to 80oC [177, 182, 184, 185], compared to experimental values obtained in this study which ranged from 126 to 215 mV/dec between 50 to 75oC . For ηa < 0.6 V, the Tafel slopes obtained at 75oC and 2 to 16 M CH3OH averaged 211 mV/dec. This seems to be in reasonable agreement with the 192 mV/dec reported on 30 wt% Pt-Ru/C on glassy carbon at 80oC and low CH3OH concentration in a single compartment cell [184]. The effect of CH3OH concentration on the kinetic parameters seems to be complex and no clear trend could be established for the change in b and io values as a function of increasing CH3OH concentration. However, it can be observed that at ambient conditions and low overpotentials, the lowest Tafel slopes were obtained using the Pt/C anode, while the greatest Tafel slopes were obtained using the Pt-Ru/C anode. It was suggested by Zhang et al. [184] that high b values on Pt-Ru/C are due to the initial dissociative dehydrogenation of CH3OH while low b values for Pt/C are due to the oxidation of CO rather than an adsorption process. Based on the real surface area of a 2 mg/cm2 catalyst determined by H2 adsorption, at 2 M CH3OH and for ηa < 0.6 V, the b value obtained in this study was about 30% larger for Pt-Ru/C than for Pt/C, and the io value was about 96 % smaller for Pt/C than for Pt-Ru/C. Under those conditions, the oxidation of CH3OH may occur faster on Pt-Ru/C than on Pt/C. For the same conditions, but at ηa > 0.6 V, the b value obtained was also about 30 % larger for Pt-Ru/C than for Pt/C. Similarly, the value of io was about still 96 % smaller for Pt/C than for Pt-Ru/C. The io value determined based on the ECSA of the 4 mg/cm2 Pt-Ru/C catalyst obtained by Cu UPD was even larger. Based on this, it is likely that Pt-Ru/C possesses a superior catalytic activity for the electrolysis of CH3OH than Pt/C. This difference was not observed between the Pt-Ru black and Pt black anodes at ambient conditions, which had similar Tafel slopes at all concentrations. For pure Pt, at CH3OH concentrations of 1 M or less, reported Tafel slopes ranged from 63 to 125 mV/dec below 0.6 V, and no Tafel data was reported for Pt above 0.6 V. As previously mentioned in Section 2.3.2.2, bi-stability was observed for the well-defined Pt case at 1 M CH3OH and two Tafel slopes were calculated. In the same potential range, it seems that 84 the Tafel value obtained for the more positive curve (49 mV/dec) is in much better agreement with the literature than the one obtained for the more negative curves (400 mV/dec). The Tafel slope value of 121 mV/dec obtained for H2O electrolysis using the well-defined Pt disk is in agreement with the reported value range of 65 to 145 mV/dec at 20-25oC, however the corresponding io value is 100 % smaller than the ones reported in the literature. For H2O electrolysis, the Tafel slope was about 60 % larger for the Pt-Ru black catalyst than for the Pt black catalyst, which would indicate that Pt black is more favourable to H2O electrolysis than Pt-Ru black. However, the io value obtained was three orders of magnitude larger for Pt-Ru black than for Pt black. It was the largest io value obtained compared to all other catalysts tested for H2O electrolysis. Based on this, it seems like Pt-Ru black is a more favourable catalyst than Pt black for H2O electrolysis. The latter is in agreement with Miles et al. [191], who reported that Pt-Ru black was a better catalyst than Pt black for the O2 evolution reaction in 0.1 M H2SO4 at 80oC. They attributed this to the formation of an oxide film which decreased the Pt surface catalytic activity towards O2 evolution. For H2O electrolysis at ambient conditions, the Pt/C catalyst had a b 40 % greater than that for the Pt-Ru/C catalyst, and its io was 4 orders of magnitude smaller. This indicates that, under these conditions, Pt-Ru/C is more active for H2O electrolysis than Pt/C. The Tafel slopes for H2O electrolysis with the Pt-Ru/C at 50 and 75oC, were about 60 % higher than when using the same catalyst at ambient conditions. The kinetic parameters reported for electrolysis are varied and often contradictory. It is important to note that at IR corrected anodic overpotentials greater than 0.6 V, mass transport effects are not negligible. Hence, the kinetic parameters extracted in this region do not likely represent true kinetic behaviour and have to be considered with caution as mass transport effects could explain some of the variability in reported results. A low overpotential, which is desired to use energy efficiently and to operate as close as possible to equilibrium, can be obtained by choosing a catalyst material and system resulting in a b value as small as possible and an io value as large as possible for the desired electrochemical reaction. 85 2.3.2.6 Activation energy The activation energy was determined for CH3OH electrolysis using the Pt-Ru/C catalyst. It can be determined from the Arrhenius equation: RT Eii Aoo −= )ln()ln( 0 (2.11) where 0oi is the absolute exchange current density of the reaction, which represents the maximum io at infinite temperature, i.e., )( ∞→Toi . Using the exchange current densities based on the ECSA estimated from the intercept of the Tafel Plots, Arrhenius plots (log(io) vs. 1/T) were constructed for various CH3OH concentrations in the linear kinetic Tafel region (<0.6 V) and in the region just above 0.6 V, where the effect of mass transfer starts to appear. The apparent activation energy, EA, was determined from the slope (-EA/R) of the fitted linear trends. The values obtained, as well as some values reported in the literature, are included in Table 2.4. In agreement with the literature [184], over the concentration range studied, the larger the CH3OH concentration, the larger was the resulting Ea value. The Ea values obtained with the Pt-Ru/C anode in the kinetic region below 0.6 V are significantly lower than the smallest values reported in the literature, although the Ea values reported in the literature were obtained in single compartment cells and not using MEAs as in the present study. Zhang et al. [184] have reported that small Ea values (20-30 kJ/mol) obtained at low CH3OH concentrations (0.1-0.5 M), indicate that CO surface diffusion is the rate determining step. These researchers obtained larger Ea values (51- 84 kJ/mol), at higher CH3OH concentrations (1, 2 M) representing a change in the rate determining step. At Ea values of about 60 kJ/mol, the dissociative adsorption of CH3OH is the rate determining step, while at Ea values around 70 kJ/mol, CO oxidation is the rate determining step. In this study, the Ea values obtained in the region below 0.6 V vs. SHE for 2 and 16 M CH3OH are both below 30 kJ/mol. This might signify that, under those conditions, CO surface diffusion is the rate determining step. The Ea values obtained in the region above 0.6 V vs. SHE for 2 and 16 M CH3OH are 50 and 154 kJ/mol, respectively. Although there were no changes in the rate determining step for the electrolysis of CH3OH on Pt-Ru/C as 86 a function of CH3OH concentration in the low potential region, there might be one in the high potential region. Table 2.4: Activation energy data for the electro-oxidation of methanol on Pt and Pt-Ru catalysts. Concentration Temperature Anode Overpotential Activation Energy Anode Anode [H2SO4] [CH3OH] Range ηa [V vs. SHE] Ea [kJ/mol] Ref. Catalyst Description [M] [M] [oC] Zone I Zone II Zone I Zone II Pt-Ru 50:50 0.5 1 17-50 0.50 - 60 - [186] Pt - 0.5 1 17-50 0.50 - 30 - [192] Pt(100)- Ru - 0.5 0.6 0-65 0.25 - 61 ± 11 - [192] Pt(100)- Ru - 0.5 0.6 0-65 0.31 - 53 ± 7 - [192] Pt(100)- Ru - 0.5 0.6 0-65 0.37 - 65 ± 4 - [192] Pt(110)- Ru - 0.5 0.6 0-65 0.25 - 38 ± 4 - [192] Pt(110)- Ru - 0.5 0.6 0-65 0.31 - 50 ± 5 - [192] Pt(110)- Ru - 0.5 0.6 0-65 0.37 - 47 ± 8 - [192] Pt(111)/R u Deposited Ru from 2x10-4 M Ru(NO)(NO3)3 in 0.5 M H2SO4 0.5 0.6 0-65 0.25 - 84 ± 10 - [192] Pt(111)/R u Deposited Ru from 2x10-4 M Ru(NO)(NO3)3 in 0.5 M H2SO4 0.5 0.6 0-65 0.31 - 66 ± 7 - [192] Pt(111)/R u Deposited Ru from 2x10-4 M Ru(NO)(NO3)3 in 0.5 M H2SO4 0.5 0.6 0-65 0.37 - 56 ± 5 - [192] Pt(111)/R u 5X10-5 M RuCl3 in 0.1 M HClO4 0.5 0.6 0-65 0.25 - 70 ± 7 - [192] Pt(111)/R u 5X10-5 M RuCl3 in 0.1 M HClO4 0.5 0.6 0-65 0.31 - 56 ± 5 - [192] Pt(111)/R u 5X10-5 M RuCl3 in 0.1 M HClO4 0.5 0.6 0-65 0.37 - 54 ± 4 - [192] Pt-Ru/C 30 wt%, 1:1 0.5 0.1, 0.5, 1, 2 30-80 - - 54-84 - [184] Pt/C 40 wt% 0.5 0.1, 0.5, 1, 2 30-80 - - 23-78 - [184] Pt-Ru/C 40 wt%, 1:1 1 1 20-80 - - 33.6 - [185] Pt-Ru/C 3.2 mg/cm2 0.5 0.5 23-70 - - 62.5 - [174] Pt-Ru/C 50:50, 4 mg/cm2 0.5 2 23-75 <0.60 >0.60 12.1 59.72 Current work Pt-Ru/C 50:50, 4 mg/cm2 0.5 16 23-75 <0.60 >0.60 27.46 154.12 Current work 87 2.4 Summary It was thermodynamically and experimentally demonstrated that H2 production from CH3OH electrolysis in aqueous acidic media is significantly more effective than the H2 production from acidic H2O electrolysis. The theoretical thermodynamic evaluation proved that alcohol electrochemical reforming systems could become thermodynamically-favourable under certain operating conditions. It was estimated that in the liquid phase CH3OH electrolysis becomes thermodynamically favourable at about 41oC. However, it was demonstrated that at 50 and 75oC, the kinetics were not favourable and energy was still required to electrochemically generate H2 from the acidic aqueous CH3OH solution with all catalysts. The kinetics of the PEM electrolyser were investigated for the first time under static conditions, i.e., stirred solutions, but no flow-through or recycling systems were investigated. For all catalysts studied, it was possible to operate the CH3OH electrolyser at much greater current densities than the H2O electrolyser. The electrolysis has to be carried-out at currents (< iL) and voltages (> -1.8 V) that avoid carbon oxidation or more stable materials comprising no carbon should be used. In acidic aqueous media and for all operating conditions, carbon oxidation and catalyst degradation prevented the DMR or PEM electrolyser from sustaining higher currents in the vicinity of the limiting current. Operating the CH3OH electrolyser with a dry cathode, purged with N2 resulted in similar cell voltage requirements for electrolysis, particularly at high current densities. Very limited kinetic data was found to describe similar electrochemical systems in the literature. Only two studies were found to report kinetic data on catalyzed membranes or electrodes with membranes in the literature. All data available focusing on the electro- oxidation of CH3OH was determined for low concentrations of CH3OH (< 2 M). The Tafel kinetic analysis carried out revealed that Pt-Ru binary catalysts enhance the kinetics of the CH3OH oxidation. The activation energy obtained with the MEA was found to be in the same range as previously reported values for similar systems. A comprehensive baseline study was established for future comparison with other PEM electrolysers. 88 Chapter 3: Electrochemical Promotion of Electrocatalysis (EPOE) 3.1 Synopsis In this Chapter, the use of electrochemical promotion in triode and tetrode operation was investigated as a method to improve the kinetics of the direct aqueous acidic CH3OH PEM electrolyser or reformer studied in Chapter 2. To the author’s knowledge, the application of electrochemical promotion of electrocatalysis concept described in Section 1.3.5 to the electrolysis mode remains an unexplored area which requires more research. This Chapter consists of a comprehensive study of the triode and tetrode electrolysis concept, and addresses aspects of this knowledge gap. Triode and tetrode operation was shown to have an electrochemical promotion effect on the PEM electrolysis of liquid CH3OH and water for H2 production in a non-flowing system at ambient conditions in the galvanostatic mode and in the potentiostatic mode. A decrease in electrolysis voltage requirements proportional to the current or potential imposed in the auxiliary circuit was observed only when the auxiliary current or potential was opposite to the electrolyser circuit current or potential. The effect was observed with catalytic and non- catalytic electrolyser ring working electrode and auxiliary working electrode materials. It was shown that the improvement obtained was not caused by intrinsic membrane or solution conductivity enhancements during triode operation and was not due to carbon oxidation of the carbon fibre paper at the electrolyser ring electrode. Some design aspects of the triode and tetrode electrolyser were investigated in order to further elucidate the mechanism causing the enhancements observed. The need for electrode contact with the electrolyte was investigated, as well as the effect of the electrolyser working electrode ring geometry, the orientation of the proton flux lines, and the use of a high-surface area carbon electrolyser working electrode were studied. It appears that triode and tetrode operation enhanced the electro- oxidation through the electrochemical pumping of protons and proton spillover. Triode and tetrode operation may lead to electrolysis cost reduction through the use of smaller amounts or no precious metal catalyst at the working electrodes while maintaining an adequate performance. Part of the work presented in this section was published and reprinted from the ECS Transactions, 25 (23), C. R. Cloutier, 89 and D. P. Wilkinson, "Triode Operation of a Proton Exchange Membrane (PEM) Electrolyser", 47-57, Copyright (2010), with permission from the Electrochemical Society. 3.2 Overview Similar to the triode fuel cell, the triode electrolyser is composed of three electrodes: the electrolyser electrode, an auxiliary electrode, and a counter electrode (CE). Three electrodes form two electrical circuits: (a) the electrolyser circuit, which includes the electrolyser ring electrode and the CE, and (b) the auxiliary circuit, which includes the auxiliary electrode and the CE. The electrolyser ring electrode and the auxiliary electrode serve as the working electrodes (WE) of each respective circuit while the CE is the same for both circuits. A typical PEM triode electrolyser arrangement is shown in Fig. 3.1. The electrolyser WEelec and the auxiliary WEaux, are not physically or conductively connected to one another, i.e., electrons are not traveling between the WEs of both circuits. Therefore, the anodic or cathodic current of an electrolyser circuit electrode can not be added to the anodic or cathodic current of an auxiliary circuit electrode to form a total anodic or cathodic current. Both circuits are power sinks as under the studied conditions, their respective electrochemical reactions are not spontaneous, and they require the input of current or voltage to occur. Figure 3.1: Schematic diagram of the triode electrolyser electrical circuit. Electrolyser Ring WEelec Polymer Electrolyte Membrane Common Current Collector Electrolyser Current Collector Auxiliary Circuit Electrolyser Circuit Common CE CEaux = CEelec Ielec Uelec Iaux Uaux G/P G/P 90 Examples of MEA side views of some of the different system configurations studied are illustrated in Fig. 3.2. The location of the auxiliary WEaux, the electrolyser ring WEelec, and the CE is shown. In normal operation, the electrolyser circuit uses the ring electrode while the auxiliary circuit uses the centered electrode. In inverted operation, the auxiliary circuit uses the ring electrode while the electrolysis circuit uses the centered electrode. Tetrode operation is similar to the triode concept, except that it uses a fourth auxiliary counter electrode. In tetrode operation, the electrolyser circuit and the auxiliary circuit are completely independent from one another. Triode tests were reproduced in the tetrode mode to confirm that the triode circuits were independent and not interacting. Figure 3.2: Side view of membrane electrode assemblies in different configurations: (a) triode, normal operation, common CE (b) like (a) with auxiliary WEaux away from surface (c) like (a) in inverted operation, and (d) tetrode, normal operation, independent CEs. The operation of the triode or tetrode configuration consists of promoting the electrolysis reaction through the use of an auxiliary circuit to impose electrolytic currents or potentials. When the auxiliary circuit is not used, the electrolyser operates in the conventional electrolysis mode. However, when it is used in (a) (c) Electrolyser Ring Working Electrode (WEelec) Auxiliary Working Electrode (WEaux) Common Current Collector Electrolyser Current Collector Counter Electrode (CE) Electrolyser Working Electrode (WEelec) Auxiliary Ring Working Electrode (WEaux) Auxiliary Working Electrode (WEaux) (b) M em brane (d) Auxiliary Counter Electrode (CEaux) Electrolyser Ring Counter Electrode (CEelec) Counter Electrode (CE) Counter Electrode (CE) Electrolyser Ring Working Electrode (WEelec) Electrolyser Ring Working Electrode (WEelec) Auxiliary Working Electrode (WEaux) Common Current Collector Common Current Collector Electrolyser Current Collector Electrolyser Current Collector Electrolyser Current Collector M em brane 91 the galvanostatic mode, the auxiliary circuit affects the potential of the anode and the cathode of the primary electrolyser circuit. Balomenou et al. [19] showed that by applying Kirchhoff's first law, to the WE of a fuel cell triode, a simple equation can be derived to explain the circuit current relationships in triode operation. Details on how this equation was derived can be found in Appendix D. For triode or tetrode electrolysis, this relationship becomes: auxFarelec III += (3.1) where Ielec is the electrolyser circuit current, IFar is the net Faradaic fuel-consuming current, and Iaux is the auxiliary circuit current. According to this formula, when the auxiliary circuit is not used, then Iaux = 0, and Ielec = IFar. Therefore, in triode or tetrode operation, IFar corresponds to the current remaining after subtracting the auxiliary circuit current. In the galvanostatic mode, the fixed value of the imposed electrolyser circuit current is always positive, while the varying value of the imposed auxiliary current can be positive or negative. Therefore, in triode or tetrode operation, the magnitude of Iaux will affect IFar. The auxiliary circuit current can operate in the same direction as the electrolyser (parallel) or in a direction opposite to the electrolyser (reverse). In other words, the polarity of the auxiliary WE can be changed to be in anodic or in cathodic operation. The polarity of the auxiliary WE may affect the potential of the electrolyser circuit WE. In either cases, (Eq. 3.1) remains valid, as the sign of Iaux will be positive or negative for parallel or reverse operation, respectively. When the auxiliary circuit polarity is the same as the electrolyser circuit polarity, the triode electrolyser is in parallel operation. The electrolyser and the auxiliary circuits are operated like two electrolysers in parallel and the electrolysis anodic electrochemical reaction (Eq. 1.24) occurs on both electrodes in the anodic compartment of the cell in conjunction with H2O oxidation or O2 evolution reaction if the potential is large enough (Eq. 1.21). That is, when the auxiliary circuit current polarity is opposite to the polarity of the electrolyser current, i.e., in the direction opposite to the electrolyser. Hence, this configuration is similar to having two electrolysis circuits in parallel, but working in directions opposite from one another. In this case, the normal electrolysis anodic electrochemical reactions (Eq. 1.24) and (Eq. 1.21) still occur at the electrolyser 92 electrode in the anodic cell compartment, but the electrochemical reaction occurring at the auxiliary WEaux in the anodic cell compartment will now be the H2 evolution reaction (HER): 2H+ + 2e- → H2 Eoaux, we = 0 V vs. SHE (3.2) The H2 evolution reaction independently occurring on the electrolyser CE (Eq. 1.25) and on the auxiliary WEaux (Eq. 3.2), has a low overvoltage compared to the electrochemical oxidation of CH3OH (Eq. 1.24) or H2O (Eq. 1.21) occurring on the electrolyser ring WEelec. The rate limiting step of the electro-oxidation reaction, in this case CH3OH electro-oxidation on the electrolyzer ring WEelec, will be the system’s most sensitive electrochemical reaction to changes in current or potential, thus it will be most sensitive to electrochemical promotion. Table 3.1 summarizes all the possible half-cell reactions (all written in the reduction sense) which may occur at the various electrodes of the electrolyser and auxiliary circuits in triode or tetrode operation, along with their respective standard potential. Table 3.1: List of possible half-cell reactions and their standard potentials. Half-cell reaction Standard cell potential Eo (V vs. SHE) H2O → 1/2O2 + 2H+ + 2e- -1.23 C + H2O → CO + 2H+ + 2e- -0.52 Ru → Ru2+ + 2e- -0.46 Ru2+ → Ru3+ + e- -0.25 C + 2H2O → CO2 + 4H+ + 4e- -0.21 CH3OH + H2O → CO2 + 6H+ + 6e- -0.016 H2 ↔ 2H+ + 2e- 0 The experiments in this research were mainly carried out in normal operation (no electrochemical promotion) and in reverse operation (with electrochemical promotion). In normal operation, the electrolyser circuit or the auxiliary circuit were studied separately and independently, without triode or tetrode operation. Normal operation is indicated with the superscript “o”, as previously done in the triode 93 fuel cell literature, and to be consistent with EPOC studies. Tests were conducted in the triode operating mode unless otherwise indicated. The electrolyser and auxiliary WEs located in the glass cell compartment in which an acidified water solution with or without CH3OH resides (Fig. 3.5) are referred to as the electrolyser ring WEelec and as the auxiliary WEaux, respectively. In this study, the triode and tetrode operation of a PEM electrolyser was evaluated in the liquid phase, at low temperature and in the static mode. Results were obtained galvanostatically and potentiostatically, and in some cases, with independent CE measurements. The triode concept was evaluated with respect to the electrolysis of CH3OH and H2O in acidic media with various anodic materials and confirmed with tetrode operation. The possibility of carrying out the triode or tetrode electrolysis with non-precious materials was demonstrated for the first time in a liquid system. The use of the tetrode configuration is also reported for the first time. Finally, a preliminary mechanism is suggested for the electrochemical promotion of a low temperature PEM electrolyser. 3.3 Experimental 3.3.1 Materials and membrane electrode assembly The materials used are the same as the ones previously described in Section 2.2.1. The CE had a MPL consisting of a 1 mg/cm2 coating composed of carbon black and 20 wt% PTFE was sprayed on top of the CE GDL surface, bonded to the membrane. No MPL was used for the GDE used in the WE compartment. The same electrocatalysts were used and the loadings were the same as in the previous Chapter. The electrolyser ring WEelec was a 4 mg/cm2 Pt-Ru black GDE or a 2 mg/cm2 Pt black GDE, while the electrolyser CE was a 2 mg/cm2 Pt/C GDE. For the non-precious electrolyser ring WEelec experiments, a simple CFP was employed as the GDE. Those experiments were conducted to demonstrate that triode and tetrode effects can be obtain without noble metal catalysts on the electrolyser ring WEelec. The MEA was similar to the one previously used in Chapter 2, except that the GDE geometries were different. For triode operation, the electrolyser WEelec was cut in a ring using die cutters (Arkwel Industries Ltd.) with a 16 mm OD. Different ID sizes were tested for the electrolyser ring 94 WEelec, but unless otherwise indicated, a ring with a 6.35 mm ID was employed. For tetrode operation, the electrolyser CE was also cut into a ring of the same dimensions. The auxiliary WEaux (for triode operation) or auxiliary CEaux (also needed for tetrode operation) consisted of a 15 cm long 0.5 mm OD Pt wire (99.99%, Goodfellow Cambridge Limited) which was sealed in a cylindrical glass tube, and which was welded at the end to the center of a circular Pt surface of 3.18 mm OD. The alignment of the auxiliary electrodes, i.e. WEaux (and CEaux) with the center of the MEA was guided by the aperture inner diameter of the tip of the TeflonTM cap as indicated on Fig. 3.3. Figure 3.3: Schematic drawing of auxiliary TeflonTM cap with apertures. Marks were made on the side of the rod to help ensure that the length and positioning was consistent for each experiment. Tightening of the caps ensured consistent contact pressure between the auxiliary and the MEA. The resistance of the electrolyser circuit (standard deviation of about 0.2 ohm) and of the auxiliary circuit (standard deviation of about 2 ohm) was measured by AC impedance for each experiment. A front view of the WEs used in triode operation is shown in Fig. 3.4. Figure 3.4: Front view of the triode or tetrode operation working electrode components. 3.18 mm OD Auxiliary WEaux Electrolyser WEelec ring N-117 24 mm OD 16 mm OD, 6.35 mm ID Threads Aperture fitting auxiliary electrode Threaded aperture Fitting glass cell 95 3.3.2 Glass cell for triode and tetrode operation The design of the dual-chamber glass cell described in Section 2.2.2 was modified by CanSci Glass Products Ltd. for triode and tetrode electrolysis, as shown in Fig. 3.5 and 3.6. Ports located at the center-end of the glass cell compartments were added for the auxiliary electrodes used in triode or tetrode operation. The auxiliary electrodes were centered with the electrolyser MEA through the use of Teflon caps with apertures as was shown in Fig. 3.3. A new screw clamping apparatus was designed to hold the glass cells, MEA holder and O-rings compressed together to allow the insertion of the auxiliary electrode(s). In this Chapter, all experiments were conducted at room temperature in ambient air (23oC ± 1oC). 96 Figure 3.5: Apparatus for triode and tetrode electrolysis (a) electrochemical glass cell, (b) auxiliary electrode. Pt wire 0.5 mm OD Thin Wall Glass Tube Pt flat tip (b) (a) Multistat and PC Reference electrode Anode Current Collector Cathode Current Collector Magnetic Stirrer Magnetic Stirrer Anolyte Reservoir Catholyte Reservoir Gas Outlet Gas Outlet O-rings Temperature Indicator Thermocouple Threaded Caps with Plugs O-ring clamp Auxiliary Electrode Port for Triode and Tetrode Operation Alignment Groove for Glass Cell Holder MEA Holder Auxiliary Electrode Port for Tetrode Operation 97 Figure 3.6: Picture of the electrochemical glass cell set-up in triode operation. 3.3.3 Electrochemical measurements Electrochemical measurements were conducted using the same equipment as described in Section 2.2.3. Each circuit for triode or tetrode was independently controlled with different channels of the Multistat, i.e., the current or potential at one of the circuits was fixed while the current or potential of the other circuit was varied in increments. The experiments were either conducted in the galvanostatic, current- controlled mode (0.01 to 1700 mA/cm2) or in the potentiostatic, potential-controlled mode (-0.6 to 1.6 V vs. SHE). It is the first time potentiostatic control is used for triode operation. A schematic representation of the electrochemical potentials is provided in Fig. 3.7. The WEelec and WEaux potential settings (dashed lines) for the potentiostatic control experiments are showed along with the electrochemical half-cell reaction standard potentials. As the electrochemical half-cell reactions are kinetically inhibited, they will not start to occur until larger potentials are applied. It will be seen that, in the potentiostatic mode, triode or tetrode enhancements were only obtained when the WEaux was in the potential range of -0.6 to 0.52 V vs. SHE. For the galvanostatic experiments, Uelec, cell is measured, and not Uelec, we, however, it was verified that Uelec, we is approximately equal to Uelec, cell, as Uelec, ce is close to 0 V vs. SHE. Magnetic stirrers MEA holder Multistat connections Auxiliary WEaux WE compartment CE compartment Glass cell holder 98 Figure 3.7: Potentiostatic working electrode potential experimental settings relative to the standard half- cell electrochemical potentials. A schematic diagram representing the various independently controlled Multistat connections used in the galvanostatic and potentiostatic modes is shown in Fig. 3.8. In some cases, a digital multimeter (Fluke 179) was used to independently monitor the common CE potential in triode operation, or the CEelec potential in tetrode operation. As per its instruction manual, the input resistance of this digital multimeter is listed as > 10 MΩ, which is greater than the resistance of the circuit monitored. The digital multimeter impedance was high enough to avoid polarization of the reference electrode, i.e., it was sufficient to ensure that the reference electrode was subjected to a very limited measuring current. Furthermore, it did not appear to affect the accuracy of the voltage measurements, as the circuit operation was not disturbed when the multimeter was turned on or off during normal, triode or tetrode electrolysis operation. Eo [V vs. SHE] 0 1.23 0.52 -0.6 Oxidation (+) Reduction (-) 0.016 O2 + 4H+ + 4e- ← 2H2O CO2 + 6H+ + 6e- ← CH3OH + H2O WEelec (fixed) WEaux (varying) 2H+ + 2e- ↔ 2H2 1.6 99 Figure 3.8: Schematic diagram of EPOE testing system connections for (a) galvanostatic triode operation, (b) potentiostatic triode operation (c) galvanostatic tetrode operation, and (d) potentiostatic tetrode operation. In galvanostatic and potentiostatic triode operation, line interactions were avoided by grounding the auxiliary circuit through the grounding of the electrolyzer circuit. Galvanostatic experiments were conducted in three steps: (a) the electrolyser circuit was turned on first and held at a current where electrolysis occurs for 500 s, (b) the auxiliary circuit was turned on for 500 s, and (c) the auxiliary circuit was turned off for 500 s for system relaxation during which only the electrolyser circuit was still kept running. These three steps were repeated for each auxiliary current increment tested. All potentiostatic experiments were also conducted in three testing steps: (a) the electrolyser circuit was turned on by itself (Uoelec, cell) (b) the electrolyser circuit was turned off, and the auxiliary circuit was turned on by itself (Uoaux, cell), and (c) both circuits, the electrolyser and the auxiliary circuits, were turned on simultaneously, i.e., the system was in triode operation (Uelec, cell and Uaux, cell). The system relaxation step was to re-start (a) for the next experiment. Segment (a) repeatedly resulted in the same values between each PC MSE Auxiliary Circuit Electrolyser Circuit Multimeter PC MSE Auxiliary Circuit Electrolyser Circuit Multimeter Ielec,cell = constant Iaux,cell = varying Uaux, ce = Uelec, ce Uaux, ce = Uelec, ce Uaux, we = varying Uelec, we = constant Solartron Multistat Solartron Multistat CEaux CEelec WEaux WEelec CEaux WEaux CEelec WEelec REaux REelec (a) (b) PC MSE Auxiliary Circuit Electrolyser Circuit Multimeter PC MSE Auxiliary Circuit Electrolyser Circuit Multimeter Ielec,cell = constant Iaux,cell = varying Uaux, ce = Uelec, ce Uaux, ce = Uelec, ce Uaux, we = varying Uelec, we = constant Solartron Multistat Solartron Multistat CEaux CEelec WEaux WEelec CEaux WEaux CEelec WEelec REaux REelec (c) (d) 100 experiment. Segment (b) was carried out to evaluate if there was a difference between Ioaux, cell when the auxiliary circuit was operated alone, and Iaux,cell, when the system was in triode or tetrode operation. Each potentiostatic testing segment ran for 200 s instead of 500 s to speed-up the experiments, as this time was found to be sufficient for the current to attain a stable value in normal, triode or tetrode operation (±0.002 A). In tetrode operation, (Fig. 3.2 (d)) the electrolyser circuit and the auxiliary circuit used separate CEs. This allowed grounding of both circuits separately. Clearly, in tetrode operation, both circuits were independent and without electronic connection. The possibility of running a circuit in the galvanostatic mode while running the other in the potentiostatic mode was also evaluated in tetrode operation. As enhancements were obtained in both the triode and tetrode cases, it was further confirmed that the effects observed in the triode case were not an artefact of the experimental setup. Most AC impedance measurements were conducted in the same manner as described in Section 2.2.3 and in Appendix F. The measurements were obtained after the system reached equilibrium, i.e., at relaxation when no current or potential were applied to the electrolyser or auxiliary circuits. In general, the resistance in the auxiliary circuit was greater than that of the electrolyser circuit. In Section 3.5.4.3, a high-precision inductance capacitance resistance (LCR) meter (GW Instek, LCR-821) was used at a frequency of 100 kHz to record the MEA AC impedance while the multistat controlled the electrolyser in the potentiostatic mode. During these 1000 s tests, current measured by the multistat remained constant, even if the LCR meter was turned on/off to take a resistance measurement every 100 s. The perturbation of the LCR meter on the measured current was 1.5 mA at the most. In the galvanostatic mode, after the electrolyzer reached steady-state, i.e., when the change in cell potential was less than 20 mV over several minutes, AC impedance measurements were recorded every 100 s over a period of 1000 s. In most tests, AC impedance measurements were obtained with the Solartron impedance/gain phase analyzer after the system reached equilibrium, i.e., at relaxation when no current or potential were applied to the electrolyzer or auxiliary circuits. Cyclic voltammetry (CV) experiments were conducted at ambient temperature (23 ±2oC) on electrodes and substrates in a single compartment 3-electrode glass cell, which was cleaned with a concentrated H2SO4-HNO3 (vol/vol, 3:1) mixture and rinsed with DI water between tests. A platinized Pt flag CE and a MSE RE equipped with a Luggin capillary were employed. N2 was bubbled using a sparger to de-aerate 101 the 0, 2 and 16 M CH3OH solutions for a few minutes. Ten conditioning potential sweep scans were carried out at 100 mV/s to ensure stability, after which two characterization potential sweep scans were conducted at 5 mV/s, all over the potential range from H2 evolution to carbon oxidation. Double-sided conductive carbon tape (SPI Supplies) was used to ensure a strong electrical connection between the GDE WE and the alligator clip. More information on the CV technique can be found in Appendix F. 3.3.4 Stability of carbon containing electrolyser ring electrodes Cyclic voltammetric measurements over a wide potential range were conducted in a three electrode system in order to verify at which potential carbon oxidation starts to occur on different electrolyser ring electrode materials in various electrolysis solutions. Figure 3.9 shows the forward anodic potential sweeps only for Pt black and CFP 2.01 cm2 circular samples. The cyclic voltammograms were corrected for ohmic resistance. A high on-set potential towards the oxidation of carbon is desired for an optimized electrolyser ring electrode. Figure 3.9: IR-corrected forward anodic potential sweep (5 mV/s, 0, 2 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, CFP with and without 2 mg/cm2 Pt black WE, Pt flag CE, 23 ±1oC). 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75 3 IR Corrected V vs. SHE C ur re nt (A ) 16 M CH3OH 2 M CH3OH 2 M CH3OH 0 M CH3OH 16 M CH3OH Pt black/CFP CFP C ur re nt (A ) 102 The on-set potential values obtained are summarized in Table 3.2. Table 3.2: IR corrected on-set potentials for CFP samples with and without Pt black at varying CH3OH concentrations. [CH3OH] [M] 2 mg/cm2 Pt black/CFP [V vs. SHE] CFP [V vs. SHE] 0 1.70 1.86 2 0.60 1.96 16 0.65 2.00 The on-set potential values obtained on the CFP correspond to the potential at which carbon oxidation starts to occur in the different solutions according to the electrochemical anodic half-cell oxidation reaction (Eq. 2.7). It is known that carbon oxidation is kinetically-inhibited and begins to oxidize at a WE voltage around 1.8 V vs. SHE in the presence of Pt [174]. Water is the source of oxygen for the formation of CO2 in carbon corrosion. Therefore, it is expected that carbon corrosion will be slower at higher CH3OH concentrations, since less H2O molecules are available for carbon corrosion. Lim et al. [193] have reported that the electrochemical corrosion of carbon in the gas phase is dependent on humidity and cell temperature, and that the presence of O2 had little effect at a H2 PEMFC temperature of 90oC. It is also known that Pt can accelerate the corrosion rate of the electrocatalyst carbon support, and cause permanent carbon loss, loss of catalytic activity, as well as structural collapse of a H2 PEMFC cathode [194]. Hence, it is not expected that carbon corrosion will start at the same potential for the Pt black and CFP electrolyser ring WEs. In our system, the kinetics of carbon corrosion in the absence of an electrocatalyst was slow. It did not start until WE potentials of about 1.86 V vs. SHE or greater were reached. In most galvanostatic experiments, the electrolyser cell current density was maintained at 1.25 mA/cm2. At this current density, Uelec, cell was sufficient for the electro-oxidation reactions of interest to occur on the electrolyser WE while avoiding the carbon corrosion potential region. Also, the electrolyser circuit was operated in normal electrolysis between every triode case and, as Uelec, cell was stable, this provided further evidence that no degradation of the CFP electrode occurred. In all potentiostatic experiments, the electrolyser ring WE potential was maintained at 0.52 V vs. SHE. At this 103 potential, carbon oxidation is not an issue for H2O electrolysis or CH3OH electrolysis on Pt black or on CFP. The auxiliary WE potential was varied from -0.6 to 1.6 V vs. SHE, but since it is made of pure Pt, carbon oxidation was not a concern. No SEM microscopic visual corrosion of the CFP electrodes was observed after the triode and tetrode experiments were run in the galvanostatic or potentiostatic mode. 3.4 Triode or Tetrode operation evaluation expressions 3.4.1 Triode or tetrode operation evaluation in the galvanostatic mode The electrolyser triode or tetrode voltage ratio, R, is defined as cellrev cellelec U U R , ,= (3.3) where Uelec, cell is the potential resulting from operating the electrolyser in galvanostatic triode or tetrode operation and Urev, cell is the reversible cell voltage, i.e., -1.23 V for the liquid phase electrolysis of H2O and -0.016 V for the liquid phase electrolysis of CH3OH at 25oC, which can be thermodynamically calculated from nF GU cellrev Δ−=, (3.4) where ∆G is the Gibbs free energy of reaction [J/mol], n is the number of electrons and F is Faraday’s constant [C/mol]. This equation is equivalent to (Eq. 1.11). Balomenou and Vayenas [19] have shown that in triode fuel cell operation, it is possible for the fuel cell to produce power with a potential difference between the anode and cathode, Uelec, cell, at least 50 % greater than Urev, cell. This is not possible for normal fuel cell operation, where the fuel cell voltage, Ufc is always less than Urev, cell. Equations for the power enhancement ratio and the power gain ratio have also been developed for fuel cell triode operation [19, 164]. These types of equations are not available in the literature for electrolysis operation. 104 Similar equations were derived in this thesis work for triode or tetrode electrolysis operation. The electrolyser power enhancement ratio, ρelec, cell, can be defined as o cellelec o cellelec celleleccellelec o cellelec cellelec cellelec UI UI P P ,, ,, , , , ⋅ ⋅==ρ (3.5) where Pelec, cell is the power of the electrolyser in triode or tetrode operation and Poelec, cell is the power of the electrolyser in normal operation. For fuel cell triode operation, all ρelec, cell values reported were positive [19, 165]. In normal electrolysis, Uoelec, cell is always negative, however, depending on the operating conditions, in triode or tetrode electrolysis operation, Uelec,cell can be negative or positive. To observe this sign change, the sign of the potential values in Eq. 3.5 were taken into account. In order to compare the power consumption of the electrolyser to that of the auxiliary circuit during triode or tetrode operation, the electrolyser power gain ratio, γelec, cell, can be defined as cellauxcellaux o cellelec o celleleccelleleccellelec cellaux o celleleccellelec cellaux cellelec celle IU IUIU P PP P P ,, ,,,, , ,, , , , ⋅ ⋅−⋅=−=Δ=γ (3.6) where ∆Pelec, cell is the change in the power output of the electrolyser in triode or tetrode operation due to the operation of the auxiliary circuit and Paux ,cell is the power consumed in the auxiliary circuit. The power gain ratio indicates when it is advantageous to use the auxiliary circuit for triode or tetrode operation. In normal electrolysis, the main electrolyser current is positive and the electrolysis cell voltage is negative. In reverse triode or tetrode operation, the auxiliary circuit is also in electrolysis, but in the opposite direction to the main electrolyser. When the auxiliary circuit is operating independently by itself, the auxiliary current is negative and the auxiliary cell voltage is positive. Triode or tetrode galvanostatic operation may affect the sign of the electrolyser cell voltage, but the auxiliary cell voltage should remain positive. Therefore, absolute values are required to make interpretation of the results more straightforward. When γelec, cell < -1, the power gained by operating the electrolyser in triode or tetrode exceeds the power consumed in the auxiliary circuit. Even if γelec, cell > -1, resulting in a net power loss, there may still be an advantage if a non-precious surface is used at the electrolyser WEelec. 105 3.4.2 Triode or tetrode operation evaluation in the potentiostatic mode Prior to our studies, potentiostatic triode or tetrode operation had not been reported in the literature. The electrolyser power enhancement ratio and the electrolyser power gain ratio, γelec, cell as previously defined (Eq. 3.5 and 3.6), can be used in the potentiostatic operating mode as well. Based on the EPOC theory, the rate enhancement ratio is defined as the ratio of the promoted catalytic rate, rp, over the unpromoted catalytic rate, ro. Similarly, for the electrochemical promotion (EP) of an electrochemical reaction, the electrolyser rate enhancement ratio, ρe, can be defined as o cellelec cellelec o cellelec cellelec e I I r r , , , , ==ρ (3.7) which is the same as the current improvement ratio. In EPOC studies, promotional rate enhancement ratio values in the range of 10 to 100 are common [18]. Based on EPOC theory, (Eq. 1.37) the Faradaic efficiency, Λ, of the triode or tetrode electrolyser H2 production using a protonic solid electrolyte for which n = 1, is a ratio which can be defined as ( ) ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ −=Λ nF I rr cellaux o celleleccellelec , ,, (3.8) If the absolute value of Λ exceeds unity, the change in the electrochemical rate is non-Faradaic. In this case, the H2 production rate when the electrolyser is in triode or tetrode operation would exceed the H2 production rate when the electrolyser is operating normally. This could be verified by measuring the rates or currents experimentally. Normally, pure electrocatalysis is limited to an absolute value of Λ ≤ 1. 106 3.5 Results and discussion Triode operation experiments were conducted in the galvanostatic and potentiostatic mode, with some testing in tetrode operation. Analysis of the triode and tetrode effects on the electrolysis of H2O and CH3OH in the galvanostatic and potentiostatic modes follows. 3.5.1 Galvanostatic characterization 3.5.1.1 Triode and tetrode operation in galvanostatic control Figure 3.10 shows an example of a triode experiment conducted in the galvanostatic mode in reverse operation. The experiment started by imposing a normal electrolyser current Ioelec, cell on the electrolysis circuit. At this time, the auxiliary circuit was not active and Iaux, cell was equal to zero. The initial values of the electrolyser and the auxiliary circuit potentials, Uoelec, cell and Uoaux, cell, were then recorded. Then the auxiliary circuit was set to a pre-determined value Iaux, cell not equal to zero, and triode operation started. Triode operation was held for 500 s. During this time, the values of the electrolyser and auxiliary circuit potentials changed to Uelec, cell and Uaux, cell respectively, while the imposed electrolyser circuit current Ielec, cell, remained the same as in normal electrolysis, i.e., Ielec, cell = Ioelec, cell. When the auxiliary circuit was turned off, triode operation stopped and both circuit potentials returned to their original normal electrolysis values, indicating that the triode effect is reversible. In this galvanostatic triode electrolysis experiment, the electrolyzer circuit current was fixed so that the electro-oxidation of CH3OH occurs on the electrolyzer ring WEelec, while the auxiliary WEaux current was varied in increments. A current of 1.25 mA/cm2 was chosen to operate the electrolyser above the carbon oxidation potential region, and to avoid the limiting current in 2M CH3OH for various WEelec materials studied for CH3OH electro-oxidation in Chapter 2 (Fig. 2.6). In the case presented in Fig. 3.10, the triode electrolyzer was operated in reverse operation, i.e., the auxiliary circuit applied current was opposite to the electrolyzer circuit applied current. As mentioned in the Section 3.3.3, a wide range of iaux,cell values was studied. In general, the value of ielec, cell was less than that of iaux.cell because the geometric surface 107 area of the electrolyser ring WEelec (1.69 cm2) is much greater than the geometric surface area of the auxiliary WEaux (0.08 cm2), and because the resistance in the auxiliary circuit is greater than that in the electrolyser circuit. The tetrode configuration is similar to the triode configuration, except that the auxiliary CE is independent from the electrolyser CE (see Fig. 3.2 (d)). Four distinct electrodes are used to form two completely independent circuits. Investigations of the tetrode configuration in the potentiostatic and galvanostatic modes revealed that the effect obtained in tetrode operation was similar to that observed in triode operation. Figure 3.10: Triode effect in galvanostatic control (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2, iaux, cell = -50 mA/cm2). 3.5.1.2 Parallel and reverse operation in galvanostatic control Figure 3.11 shows an example of triode electrolysis operation in the parallel and reverse direction. In parallel operation, Fig. 3.11 (a), the electrolyser potential remained the same whether the auxiliary circuit was operating or not. In reverse operation, Fig. 3.11 (b), the electrolyser potential was enhanced. Using -2.2 -1.8 -1.4 -1.0 -0.6 -0.2 0.2 0.6 1.0 0 250 500 750 1000 1250 1500 Time (s) IR C or re ct ed C irc ui t V ol ta ge (V ) -0.008 -0.007 -0.006 -0.005 -0.004 -0.003 -0.002 -0.001 0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 C ur re nt (A ) Triode effect Uoaux, cell Uoelec, cell Ioelec, cell Ioaux, cell Ioelec, cell Ioaux, cell Iaux, cell Ielec, cell Uaux, cell Uelec, cell Aux. Circuit Off (500s), Triode Operation Stops Auxiliary Circuit On (500s) Triode Operation Starts IR C or re ct ed C irc ui t V ol ta ge (V ) C ur re nt (A ) 108 the auxiliary circuit to apply a current opposite to the electrolyser circuit current reduced the electrolyser voltage requirements. The enhancement in the electrolyser potential was proportional to the auxiliary current applied. Under certain conditions, the electrolyser potential became positive as a result of triode operation. Triode electrolysis appears to be only beneficial in when the auxiliary circuit is operated in reverse compared to the electrolyser circuit. All subsequent galvanostatic experiments were carried out in the reverse electrolysis mode. Figure 3.11: Effect of varying Iaux, cell on the electrolyser circuit voltage in galvanostatic control (a) parallel operation (b) reverse operation (4 mg/cm2 Pt-Ru black electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2, iaux, cell = -50 mA/cm2). -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0 500 1000 1500 Time (s) IR C or re ct ed C irc ui t V ol ta ge (V ) -0.0025 -0.0015 -0.0005 0.0005 0.0015 0.0025 0.0035 Cu rr en t ( A) -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0 500 1000 1500 Time (s) IR C or re ct ed C irc ui t V ol ta ge (V ) -0.005 -0.004 -0.003 -0.002 -0.001 0 0.001 0.002 0.003 C ur re nt (A ) (a) Ielec, cell Iaux, cell Uaux, cell Uelec, cell Iaux, cell Ielec, cell Uelec, cell Uaux, cell (b) Triode operation Triode operation IR C or re ct ed C irc ui t V ol ta ge (V ) Cu rr en t ( A) IR C or re ct ed C irc ui t V ol ta ge (V ) C ur re nt (A ) 109 3.5.1.3 Electrolyser ring electrode material in galvanostatic control Fig. 3.12 demonstrates the effect of triode operation an unsupported noble metal electrocatalyst and on a CFP. In both cases, the Uelec, cell increased as the applied Iaux, cell was increased, but the Uelec, cell enhancement was not linearly proportional to the increase in the applied Iaux, cell. At sufficiently high auxiliary currents, the sign of the electrolyzer potential reversed and became positive as a result of triode operation. At this point, it may be possible that the electrolyser is driven in the opposite direction because the auxiliary current is driving the electrolysis circuit. The auxiliary cell voltage, Uaux ,cell, followed the electrolyser cell voltage, Uelec, cell , in all cases, even when the auxiliary circuit was at open circuit, i.e., while the electrolyser was in normal operation. When the auxiliary circuit was run by itself, a positive Uoaux, cell value was obtained. In inverted triode operation, i.e., when the auxiliary WEaux is the ring as shown in Fig. 3.2 (c), Uaux ,cell was close to zero while at open circuit in normal electrolysis, and did not equal Uelec, cell while in triode operation. This suggests that in normal triode operation, Uoaux, cell and Uaux, cell were affected by the current line distribution on the WE plane, as the CE potential is the same in both cases. The half-cell reaction occurring on the auxiliary WEaux during enhanced triode operation is the production of H2 (Eq. 3.2). In order to determine the effect of this H2 production on the triode improvement, the auxiliary WEaux was moved 1 cm away from the membrane surface, while still producing H2 at the same rate. In triode operation, at the same electrolyser and auxiliary currents, there was a 67 % loss in the electrolyser circuit potential when the auxiliary WEaux was pulled 1 cm away from the membrane surface. Further information on this potentiostatic test is provided in Section 3.5.4.1. In addition, in galvanostatic normal electrolysis, 500s of continuous H2 saturation of the WE glass cell compartment solution only resulted in a 3 % improvement in the electrolyser WEelec circuit potential compared to when no H2 was provided. In both cases, the electrolysis improvement obtained via direct H2 production is much less significant compared to when the auxiliary WEaux is in physical contact with the membrane in triode operation. A greater resistance was measured in the auxiliary circuit when it was pulled away, but not in the electrolyser circuit. This increase in the auxiliary circuit resistance is due to the solution resistance between the auxiliary WEaux and the membrane surface. Since auxiliary WEaux contact with the solid 110 electrolyte is required to observe any significant triode effects, the triode improvement must be resulting from an electrochemical pumping of protons rather than from the generation of H2. Further confirmation of this is provided in Section 3.5.4.1. Figure 3.12: Effect of varying Iaux, cell on the electrolyser circuit voltage in galvanostatic control (a) 4 mg/cm2 Pt-Ru black electrolyser ring WEelec, (b) CFP electrolyser ring WEelec (Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2). As expected, when a CFP was used as the electrolyser ring WEelec, as shown in Fig. 3.12 (b), the electrolyser current, without triode operation, resulted in a more negative electrolyser potential than when a noble metal electrocatalyst was present on the electrolyser ring WEelec as shown in Fig. 3.12 (a) However, the beneficial triode effect was also observed on CFP and the electrolyser potential -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 0 500 1000 1500 2000 2500 3000 3500 Time (s) IR C or re ct ed C ir cu it Vo lta ge (V ) -0.005 -0.004 -0.003 -0.002 -0.001 0 0.001 0.002 0.003 0.004 0.005 0.006 Cu rr en t ( A) Auxiliary Potential Electrolyser Potential Auxiliary Current Electrolyser Current -2.2 -1.8 -1.4 -1.0 -0.6 -0.2 0.2 0.6 1.0 0 500 1000 1500 2000 2500 3000 3500 Time (s) IR C or re ct ed C ir cu it Vo lta ge (V ) -0.005 -0.004 -0.003 -0.002 -0.001 0 0.001 0.002 0.003 0.004 0.005 0.006 C ur re nt (A ) Auxiliary Potential Electrolyser Potential Auxiliary Current Electrolyser Current (a) Ielec, cell Iaux, cell Uaux, cell Uelec, cell Iaux, cell Ielec, cell Uelec, cell Uaux, cell (b) IR C or re ct ed C ir cu it Vo lta ge (V ) Cu rr en t ( A) IR C or re ct ed C ir cu it Vo lta ge (V ) C ur re nt (A ) 111 enhancement observed (Uelec, cell - Uoelec, cell), i.e., the electrolyser voltage in triode less the baseline electrolyser voltage, was similar to the one observed on Pt-Ru black. No CFP anode ring damage was observed before and after multiple triode tests. The auxiliary current was turned on and off numerous times, and the electrolyser potential without triode operation returned to the same original value, indicating the absence of carbon oxidation in the process. This confirms that the application of an auxiliary current activated a process which made the electrolyser CFP ring WEelec surface electrocatalytically active. On CFP, no apparent capacitive effect was observed indicating that the change in surface charge concentration was completely reversible. This was not the case for Pt-Ru/C, for which the WEelec surface charge concentration after triode operation did not return to its original value in normal operation. It appears that, triode operation promotes the electrolysis by electrochemically modifying the localised H+ species surface concentration. It was also verified that a comparable triode effect can be obtained using an auxiliary WEaux made of a non-catalytic material, such as Au plated SS in Section 3.5.4.7. In this case, protons are transported through the membrane, but no H2 is generated at the auxiliary WEaux surface. Additional experimental details on this are provided in Chapter 4. 3.5.1.4 Electrolyser voltage ratio, power enhancement ratio and power gain ratio in galvanostatic control The electrolyser potential in triode operation, Uelec, cell, and the electrolyser triode voltage ratio, R (dimensionless), were determined for different auxiliary circuit current densities. The results obtained for the Pt-Ru black ring and the CFP ring are shown in Fig. 3.13 (a) and (b), respectively. For both anode materials, the value of Uelec, cell went from being negative to being positive as the auxiliary circuit current density was increased to more negative values. This means that, in triode operation, the electrolyser can become a power source when the auxiliary circuit current reaches a certain value. The triode voltage trend was similar to the one of Uelec, cell but the signs were reversed as Urev, cell is negative. At iaux, cell = -0.01 A/cm2, the value of R obtained for the CFP ring was about 2.8 times greater than the one obtained for the Pt-Ru black ring WEelec. At iaux ,cell = -0.05 A/cm2, the R value was two times greater for the electrolyser using a Pt-Ru black ring WEelec than for the electrolyser using a CFP ring WEelec. Triode operation reduces some of the overvoltage losses and allows operating at voltages lower than 112 Urev, cell, which is not possible in normal electrolysis. Balomenou et al. [165] have shown for an SOFC fuel cell operating on 2 % dry CH4, an increase in the fuel cell voltage, Ufc, of 33 % greater than Urev, cell at a fixed auxiliary current of -0.5 mA and a fuel cell current of 0.2 mA at 700oC using Pt electrodes. The improvements obtained at low temperature in the current electrolysis study are of a similar magnitude. It was possible to operate the electrolyser in triode mode at 27 % and 13 % less voltage than Urev, cell on a Pt-Ru black ring WEelec and a CFP ring WEelec, respectively. It appears that this is the first time such overvoltage reductions have been reported in a low temperature triode electrolysis system. Figure 3.13: Effect of iaux, cell on Uelec, cell and on the triode voltage ratio (R) (a) 4 mg/cm2 Pt-Ru black electrolyser ring WEelec (b) CFP electrolyser ring WEelec (Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2). (a) (b) -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0 Auxiliary Circuit Geometric Current Density (A/cm2) IR C or re ct ed E le ct ro ly se r C irc ui t V ol ta ge (V ) -40 -30 -20 -10 0 10 20 30 40 50 Tr io de V ol ta ge R at io (d im en si on le ss ) -2.5 -2 -1.5 -1 -0.5 0 0.5 -0.06 -0.05 -0.04 -0.03 -0.02 -0.01 0 Auxiliary Circuit Geometric Current Density (A/cm2) IR C or re ct ed E le ct ro ly se r C irc ui t V ol ta ge (V ) -20 0 20 40 60 80 100 120 140 Tr io de V ol ta ge R at io (d im en si on le ss ) IR C or re ct ed E le ct ro ly se r C irc ui t V ol ta ge (V ) Tr io de V ol ta ge R at io (d im en si on le ss ) IR C or re ct ed E le ct ro ly se r C irc ui t V ol ta ge (V ) Tr io de V ol ta ge R at io (d im en si on le ss ) 113 The electrolyser power enhancement ratio, ρelec, cell, (dimensionless) and the electrolyser power gain ratio, γelec, cell, (dimensionless), are shown in Fig. 3.14. Figure 3.14: Power enhancement ratio (ρelec, cell) and power gain ratio (γelec, cell) as a function of the auxiliary circuit current density (a) 4 mg/cm2 Pt-Ru black electrolyser ring WEelec (b) CFP electrolyser ring WEelec (Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2). Under the conditions studied, the power enhancement ratio, ρelec, cell , did not exceed unity in any case, indicating that less power was needed in triode electrolysis than in normal electrolysis. The reduction in the triode power requirements became more significant as the applied auxiliary current density was -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 -6 .0 E -0 2 -5 .0 E -0 2 -4 .0 E -0 2 -3 .0 E -0 2 -2 .0 E -0 2 -1 .0 E -0 2 0. 0E +0 0 Auxiliary Circuit Geometric Current Density (A/cm2) P ow er E nh an ce m en t R at io (d im en si on le ss ) -4000 -3500 -3000 -2500 -2000 -1500 -1000 -500 0 500 P ow er G ai n R at io (d im en si on le ss ) -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 -6 .0 E -0 2 -5 .0 E -0 2 -4 .0 E -0 2 -3 .0 E -0 2 -2 .0 E -0 2 -1 .0 E -0 2 0. 0E +0 0 Auxiliary Circuit Geometric Current Density (A/cm2) P ow er E nh an ce m en t R at io (d im en si on le ss ) -5 -4.5 -4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 P ow er G ai n R at io (d im en si on le ss ) (a) (b) (-4.4) (-3.3) (-2.5) -6 .0 E -0 2 -5 .0 E -0 2 -4 .0 E -0 2 -3 .0 E -0 2 -2 .0 E -0 2 -1 .0 E -0 2 0. 0E +0 0 P ow er E nh an ce m en t R at io (d im en si on le ss ) P ow er G ai n R at io (d im en si on le ss ) -6 .0 E -0 2 -5 .0 E -0 2 -4 .0 E -0 2 -3 .0 E -0 2 -2 .0 E -0 2 -1 .0 E -0 2 0. 0E +0 0 P ow er E nh an ce m en t R at io (d im en si on le ss ) P ow er G ai n R at io (d im en si on le ss ) 114 increased. While the normal electrolysis cell voltage is always negative, the triode electrolysis cell voltage became positive once a sufficiently large auxiliary current density was applied. This explains the ρelec cell sign changes observed. At large auxiliary current densities, ρelec, cell was more favourable for the Pt-Ru black electrolyser ring WEelec than for the CFP ring WEelec. The catalytic surface responded better than the non-catalytic one to the changes imposed by the auxiliary circuit. The power gain ratio did not reach positive values under the conditions studied, indicating that, in all cases, the power required to operate the electrolyser in triode mode was less than the power required to operate under normal electrolysis. In some cases, γelec, cell reached values less than -1, signifying that the reduction in the power requirements in triode electrolysis was greater than that consumed in the auxiliary circuit. This was the case at low auxiliary current densities for the Pt-Ru black electrolyser ring WEelec and for all CFP electrolyser ring WEelec cases studied. For the Pt-Ru black electrolyser ring WEelec, a maximum γelec, cell was observed at iaux, cell = -0.012 A/cm2 and for the CFP electrolyser ring WEelec, a maximum γelec, cell was observed at iaux, cell = -0.037 A/cm2. The maximum decrease in the power requirements was three orders of magnitude larger for the CFP electrolyser ring WEelec than for the Pt-Ru black electrolyser ring WEelec. Thus, it is clear that for the non-noble CFP WEelec case, the decrease in power requirements caused by triode operation significantly exceeded the power sacrificed in the auxiliary circuit. Such high power gain ratios have not been reported previously. 3.5.1.5 Benefits of starting in triode or tetrode operation in galvanostatic control At a sufficiently high electrolysis current, the electrolyser cell may reach voltages where CFP oxidation can occur (~1.8 V vs. SHE) on the electrolyser ring WEelec. Figure 3.15 illustrates that triode operation may be a possible way to prevent electrolyser ring WEelec electrocatalyst degradation through CFP or any other material corrosion. 115 Figure 3.15: Benefits of starting in triode operation in galvanostatic control (CFP ring electrolyser WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 50, 75 mA/cm2, iaux, cell = -5 A/cm2). As can be observed in Fig. 3.15, when the electrolyser operates at Ielec, cell = 50 mA/cm2 and the auxiliary circuit operates at a high current density of -5000 mA/cm2, the electrolyser cell voltage is about 2.7 V. When the auxiliary circuit is shut-off, the electrolyser cell voltage drops to about -3 V and the CFP electrolyser ring WEelec is irreversibly damaged. Similarly, when the electrolyser operates at Ielec ,cell = 75 mA/cm2 and the auxiliary circuit operates at the same current density, the electrolyser cell voltage is around 1 V. When the auxiliary circuit is shut-off, the electrolyser voltage drops to a value less than -4 V, i.e., reaching the chosen instrument set voltage safety limit, and the CFP electrolyser ring WEelec is irreversibly destroyed. In Chapter 2, it was shown that, in normal PEM electrolysis, there exists a limiting current density for the electro-oxidation of CH3OH beyond which no stable cell voltage measurements can be obtained as the electrocatalyst carbon support is oxidized and precious metal is lost from the electrode surface. In both triode operation cases here, the imposition of a high auxiliary current density, where the triode effect is more significant, prevented the electrolyser ring WEelec from damage through carbon oxidation. At a lower auxiliary current density, where the triode effect is not as significant, no beneficial effect was observed, and the electrolyser ring WEelec was irreversibly damaged at higher electrolyser current densities. -4 -3 -2 -1 0 1 2 3 4 0 100 200 300 400 500 600 700 800 900 1000 Time (s) IR C or re ct ed E le ct ro ly se r Ci rc ui t V ol ta ge (V ) 50 mA/cm 75 mA/cm Triode Operation Electrolyser Circuit Only Triode Operation Stops Electrode Damage 2 2 No Electrode Damage 50 mA/cm2 75 mA/cm2 50 mA/cm2 75 mA/cm2 116 3.5.2 Potentiostatic characterization 3.5.2.1 Triode configuration in potentiostatic control The electrolyser triode configuration with a common CE was also studied in potentiostatic control. Fig. 3.7 showed the working electrode potential settings and the electrochemical half-cell reactions occurring on the working electrodes. An example of a triode experiment carried out in the potentiostatic mode and demonstrating the effect of step changes in the auxiliary WEaux potential varying from -0.6 to 1.6 V vs. SHE is provided in Fig. 3.16. As no triode operation measurements have been previously conducted in the potentiostatic mode according to the previously existing literature, this new graphical representation of the results requires explanation. Figure 3.16: Triode effect in potentiostatic control (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). The electrolyser ring WEelec potential, Uelec, we, was maintained at 0.52 V vs. SHE (0 electrolyser current at this potential) as indicated by the horizontal dashed line. This potential value was chosen based on the fundamental study of the PEM electrolysis system discussed in Chapter 2 because it is far removed -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs SHE) Au xi lia ry G eo m et ric C ur re nt D en si ty (A /c m 2 ) -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 El ec tr ol ys er G eo m et ric C ur re nt D en si ty (A /c m 2 ) iaux, cell ielec, cell Oxidation on WEaux Reduction on WEelec Uelec,we= 0.52 V Reduction on WEaux Oxidation on WEelec Au xi lia ry G eo m et ric C ur re nt D en si ty (A /c m 2 ) El ec tr ol ys er G eo m et ric C ur re nt D en si ty (A /c m 2 ) Au xi lia ry G eo m et ric C ur re nt D en si ty (A /c m 2 ) El ec tr ol ys er G eo m et ric C ur re nt D en si ty (A /c m 2 ) 117 from the potential for carbon oxidation and it is below the onset potential for CH3OH oxidation. Cyclic voltammetry results shown in Fig. 3.9 also supported this decision. The electrolyser circuit was operated alone between each set of tests and the potential always returned to the same value between tests. This confirms that there was no carbon degradation between the tests. Furthermore, no apparent damage to the electrode was visually observed after the entire test sequence. The same potential value was used for CH3OH and H2O electrolysis, although larger electrolyser ring WEelec potential values (Uelec,we up to ~1.5 V vs. SHE) could be used in the case of H2O electrolysis. The auxiliary WEaux potential, Uaux, we, was varied and is represented on the x-axis. A vertical dashed line divides the reverse triode operation data for the WEs into two potential regions. On the right hand side, CH3OH oxidation occurs on the auxiliary WEaux, and reduction (H2 production) occurs on the electrolyser ring WEelec. On the left hand side, CH3OH oxidation occurs on the electrolyser ring WEelec and reduction (H2 production) occurs on the auxiliary WEaux. The electrolyser geometric current density, ielec, cell and the auxiliary current density, iaux, cell, are indicated. As can be seen on Fig. 3.16, at the set voltage of 0.52 V vs. SHE, in normal electrolysis operation, ioelec ,cell is close to zero. In triode operation, when Uaux, we = Uelec, we, ielec, cell is close to zero. On the right hand side of the vertical dashed line, triode operation does not have a positive effect on the desired electrolyser current compared to normal operation. When electro-oxidation occurs on both WEs, no ielec, cell enhancements are observed, which is consistent with the parallel operation findings in the galvanostatic mode described earlier in Section 3.3.1. On the left hand side of the vertical dashed line, triode operation has a beneficial effect on the desired electrolyser current. When Uaux, we reaches the reduction region, i.e., < 0.52 V vs. SHE, ielec, cell becomes positive. Under these conditions, the electro-reduction reaction occurring on the auxiliary WEaux enhances the electro-oxidation reaction occurring on the electrolyser ring WEelec. Therefore, focus is made on the region where triode operation appears to be beneficial to the electrolyser. At Uaux, we lower than -0.8 V vs. SHE, it may be possible that, eventually, the driving force of the electrolyser electrochemical oxidation, ielec,cell, becomes so large as to create excessive CO2 evolution, which would block the active catalyst surface WEelec from the solution and decrease the overall PEM electrolyser performance. 118 During some triode tests, the common CE potential was measured against a second MSE reference electrode located in the glass cell CE compartment. The common CE potential is generally not reported during triode operation in the literature. However, it is important to know how the potential of both individual electrodes is influenced under these test conditions, as per the diagram of the electrochemical half-cell potential presented in Fig. 3.7. During triode operation, the CE potential is forced to a common value for the electrolyser and auxiliary circuits. For some tests, it was verified that the potential of the CE when the electrolyser or the auxiliary circuit were operated separately and alone, i.e., Uoelec, ,ce or Uoaux, ce, was constant between each Uaux, we increments throughout the entire course of the triode experiment, as was shown in Fig. 3.7 (b) and 3.8 (b). This confirms that the carbon of the common CE was not oxidized by the auxiliary circuit during the course of the triode experiment. It was also verified that Uoaux,ce, and Uaux,ce, as well as Uoelec, ce and Uelec,ce did not change much during triode operation. However, small discrepancies were observed when Uoaux, we was less or equal to -0.14 V vs. SHE, i.e., when the auxiliary WEaux was in the H2 production region, where the mixed potential effect at the common CE becomes more apparent. At these auxiliary WE potentials, H2 evolution starts to occur at the auxiliary WEaux and the mixed potential at the common CE becomes more apparent. 3.5.2.2 Tetrode configuration in potentiostatic control The tetrode configuration is similar to the triode configuration, except that the auxiliary CEaux is independent from the electrolyser CEelec (Fig. 3.2 (d)). Four distinct electrodes are used to form two independent circuits. Investigations of the tetrode configuration in potentiostatic control using a Pt-Ru black electrolyser WEelec, as shown in Fig. 3.17, revealed that the effect obtained in tetrode operation was less but similar to that observed in triode operation using the same WEelec material. This can be also observed by comparing Fig. 3.17 (Pt-Ru black electrolyser WEelec) and Fig. 3.16 (CFP electrolyser WEelec). Thus, the EPOC effect is independent of the electrolyser WEelec material, which does not need to be catalytic. 119 Figure 3.17: Triode and tetrode effect in potentiostatic control (4 mg/cm2 Pt-Ru black electrolyser ring WEelec, 2 mg/cm2 Pt/C electrolyser CEelec, Pt auxiliary WEaux and CEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). The difference in the tetrode results compared to the triode results may be attributed to the fact that the tetrode CEs have different surface areas, and that the resistance in the auxiliary circuit is greater in the tetrode case than in the triode case. 3.5.2.3 Counter electrode potential measurements in potentiostatic control For the tests presented in Fig. 3.18, the potential of the counter electrode(s) was measured independently with a multimeter. For the triode case shown in Fig. 3.18, the common counter electrode potential was first measured when the electrolyser was operated normally, then the common counter electrode potential was measured when only the auxiliary circuit was operating, and finally, it was measured when both circuits were in operation, i.e., triode electrolysis. Similarly, for the tetrode case -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) iaux, cell, tetrode iaux, cell, triode ielec, cell, tetrode ielec, cell, triode Uelec,we= 0.52 V vs. SHE A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) 120 shown in Fig. 3.19, the potential of both counter electrodes, CEelec and CEaux, was measured in normal electrolysis, then when only the auxiliary circuit was operating, and then during tetrode operation. Figure 3.18: Triode effect in potentiostatic control (a) geometric current density (b) counter electrode potential (4 mg/cm2 Pt-Ru black electrolyser ring WEelec, 2 mg/cm2 Pt/C electrolyser CE, Pt auxiliary WEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). Figure 3.19: Tetrode effect in potentiostatic control (a) geometric current density (b) counter electrode potential (4 mg/cm2 Pt-Ru black electrolyser ring WEelec, 2 mg/cm2 Pt/C electrolyser CEelec, Pt auxiliary WEaux and CEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). (a) (b) Uelec,we= 0.52 V vs. SHE -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs SHE) C ou nt er E le ct ro de P ot en tia l (V v s S H E ) U elec, ce U aux, ce Uelec, ce = Uaux, ce o o iaux, cellielec, cell A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) C ou nt er E le ct ro de P ot en tia l (V v s S H E ) -5 -4 -3 -2 -1 0 1 2 3 4 5 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) -0.15 -0.1 -0.05 0 0.05 0.1 0.15 E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) (a) (b) Uelec,we= 0.52 V vs. SHE iaux, cellielec, cell -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs SHE) C ou nt er E le ct ro de P ot en tia l (V v s S H E ) U aux, ce U elec, ce Uaux, ce Uelec, ce o o A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) C ou nt er E le ct ro de P ot en tia l (V v s S H E ) 121 As shown on Fig. 3.18 (b), in the triode case, the common counter electrode potential was close to zero in normal electrolysis. When the auxiliary circuit was operated alone, the common counter electrode potential was positive, as the system operates in reverse electrolysis, and the common counter electrode then behaves as an anode. In triode electrolysis at an electrolyser WEelec potential of 0.52 V vs. SHE and an auxiliary WEaux potential of -0.6 V vs. SHE, the common CE reached an intermediate potential of 0.5 V vs. SHE. Therefore, when operating the triode at this particular auxiliary WEaux potential, the proton flux likely travels from the positive WEelec to the negative WEaux in the WE compartment, and possibly from the common CE to the WEaux through the membrane. As shown on Fig. 3.19 (b), in the tetrode case, when each circuit was operated separately, the respective potential of the independent CEs potentials was similar as the ones measured in the triode case. However, in tetrode operation at an electrolyser WEelec potential of 0.52 V vs. SHE and an auxiliary WEaux potential of -0.6 V vs. SHE, the electrolyser CEelec potential decreased to about -0.5 V vs. SHE, while the auxiliary CEaux potential increased to about 2.75 V vs. SHE. When operating the tetrode at this particular auxiliary WEaux potential, the proton flux may still travel from the positive WEelec to the negative WEaux in the WE compartment, but also from the positive WEelec to the negative CEelec through the membrane, and from the positive CEaux to the negative WEaux through the membrane. 3.5.2.4 Electrolyser ring electrode material in potentiostatic control Different electrolyser ring electrode materials were compared in triode operation for H2O electrolysis and CH3OH electrolysis. Figure 3.20 shows that, in triode operation, Pt was a superior catalyst for H2O electrolysis while Pt-Ru was a superior catalyst for CH3OH electrolysis. The enhancement in the electrolyser geometric current density resulting from triode operation was more significant for the electrocatalysts than for the CFP at a given auxiliary WEaux potential and was greater for CH3OH electrolysis than for H2O electrolysis. The current enhancement trend was comparable for all catalysts and for both H2O and CH3OH electrolysis, indicating that the mechanism behind the improvement observed in triode operation is similar for different electrolyser ring WEelec materials and electro-oxidation reaction mechanisms. Like the galvanostatic experiments indicated, it appears that triode operation promoted the electro-oxidation by influencing the H+ concentration at the WEelec surface. 122 Figure 3.20: Effect of varying Uaux, we on different electrolyser ring WEelec materials in potentiostatic control (a) 0 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte and (b) 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte (2 mg/cm2 Pt black, 4 mg/cm2 Pt-Ru black, and CFP electrolyser WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, Uelec, we = 0.52 V vs. SHE). -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) El ec tro ly se r G eo m et ri c Cu rr en t D en si ty (A /c m 2 ) Pt Black Pt-Ru Black CFP (a) (b) -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) E le ct ro ly se r G eo m et ri c Cu rr en t D en si ty (A /c m 2 ) Pt Black Pt-Ru Black CFP El ec tro ly se r G eo m et ri c Cu rr en t D en si ty (A /c m 2 ) E le ct ro ly se r G eo m et ri c Cu rr en t D en si ty (A /c m 2 ) 123 3.5.2.5 Electrolyser rate enhancement ratio, power gain ratio and Faradaic efficiency in potentiostatic control Table 3.3 list some of the dimensionless rate enhancement ratios, ρe, obtained for H2O and CH3OH electrolysis on different electrolyser ring WEelec materials under the selected experimental conditions in triode operation and potentiostatic control. In most cases, ρe (Eq. 3.7) exceeded the typical catalytic rate enhancement ratios reported for EPOC studies, which are usually in the range of 10 to 100. Once the auxiliary WEaux potential reached values in the H2 production region, a peak ρe was eventually reached in most cases. It occurred at different auxiliary WEaux potentials for the different electrolyser ring WEelec surfaces studied. The maximum ρe was less significant on Pt-Ru black and Pt black than on the CFP, for which it was an order of magnitude greater than on the catalytic WEelec surfaces. Table 3.3: Effect of auxiliary WEaux potential (Uaux, we) on the rate enhancement ratio (ρe) (Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, Uelec, we = 0.52 V vs. SHE). Electrolyser ring WEelec 4 mg/cm2 Pt-Ru black 2 mg/cm2 Pt black CFP [CH3OH] Uaux, we ρe [M] [V vs. SHE] [Dimensionless] 0 0.42 692 31 1621 0 -0.05 5242 164 10072 0 -0.14 1646 305 13278 0 -0.25 1903 422 2750 0 -0.59 675 485 2950 2 0.42 22 42 2655 2 -0.05 94 288 17270 2 -0.14 71 1014 25518 2 -0.25 66 1005 5900 2 -0.59 89 620 3053 124 In triode H2O electrolysis, Pt-Ru black resulted in a greater ρe than Pt black, while in triode CH3OH electrolysis, Pt black resulted in a greater ρe than Pt-Ru black. This is contrary to that observed in Chapter 2 under normal electrolysis conditions, for which Pt black was a better catalyst for H2O electrolysis than Pt-Ru black, while Pt-Ru black was a better catalyst for CH3OH electrolysis than Pt black. It is possible that the CO adsorption is affected during triode or tetrode operation and that the effectiveness of the bi-functional mechanism of the Pt-Ru black electrolyser WEelec catalyst is surpassed by the triode or tetrode effect. The CFP electrolyser ring WEelec maximum ρe attained was about two times more pronounced for CH3OH electrolysis than for H2O electrolysis. For the CFP electrolyser ring WEelec, after reaching a maximum, ρe dropped sharply as Uaux, we was further increased. This was not the case for the Pt black and Pt-Ru black electrolyser ring WEelec, as the ρe decreased more gradually as Uaux, we increased after it reached its peak. This may indicate that the peak ρe was not linked to carbon oxidation. In fact, stability tests were conducted for an hour at the peak ρe for the CFP electrolyser ring WEelec in 2 M CH3OH at Uaux, we = -0.14 V vs. SHE and within 20 min. and ρe stabilized to 10 % of its peak value. It seems that the H+ supplied through triode or tetrode operation affected the electrochemical oxidation occurring at the WEelec surface, independently of the material used for the WEelec. It also appears that the beneficial effect obtained in triode or tetrode operation is independent of the electrochemical reaction occurring at the WEelec, as rate enhancements were observed for both, H2O electrolysis as well as CH3OH electrolysis, which follow two different reaction mechanisms. Although it is possible that the work function of the WEelec surface was affected in triode or tetrode operation in the galvanostatic more, and may have affected the adsorption strength of species at the WEelec surface, and reduce the activation energy required for the electrochemical oxidation of H2O or CH3OH to occur, this is not possible in the potentiostatic mode, as the WEelec potential was fixed to a constant value throughout the entire duration of the triode or tetrode experiment. Figure 3.21 shows the triode electrolyser power gain ratio, γelec, cell, which is also dimensionless. The values obtained for γelec, cell were calculated for the cases where the common CE potential was monitored separately, because the cell voltages were required for the analysis. Most γelec, cell were negative indicating that the power required to operate in triode electrolysis was less than that required to operate 125 in normal electrolysis. In some cases, γelec, cell reached values less than -1, suggesting that the decrease in the electrolyser power requirement in triode operation exceeded the power consumed in the auxiliary circuit. Figure 3.21: Power gain ratio (γelec, cell) as a function of Uaux, we (CFP electrolyser ring WEelec and 4 mg/cm2 Pt-Ru black electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). The maximum γelec, cell obtained was -30.78 for the Pt-Ru black electrolyser ring WE at Uaux, we = -0.05 V vs. SHE and -8.15 for the CFP electrolyser ring WEelec at Uaux, we 0.42 V vs. SHE. The range of Uaux, we for which γelec, cell < -1 was larger for the Pt-Ru black than for the CFP electrolyser ring WEelec. The power gain ratios obtained with the CFP electrolyser ring WEelec in the potentiostatic control were not as significant as the ones obtained in the galvanostatic mode as the auxiliary circuit current densities studied were larger in the potentiostatic mode than in the galvanostatic control. This indicates that larger triode enhancements can be obtained when the imposed auxiliary currents/potentials are very small. Finally, it was verified that the triode Faradaic efficiency of H2 production, Λ, averaged from 0.90 to 1.001 for all electrolyser ring WEelec materials potentiostatically evaluated for the electrolysis of 0 and 2 M CH3OH. The largest Λ obtained using Pt-Ru black electrolyser ring WEelec was 1.18 in H2O electrolysis at Uaux, we = -0.25 V vs. SHE and 1.07 in CH3OH electrolysis at Uaux, we = -0.59 V vs. SHE. Although Λ -35 -30 -25 -20 -15 -10 -5 0 5 10 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 Auxiliary Working Electrode Potential (V vs. SHE) P ow er G ai n R at io (d im en si on le ss ) CFP electrolyser ring Pt-Ru black electrolyser ring 126 exceeded unity, it can be argued that these values fall within experimental error. According to the Faradaic efficiency definition, the overall triode electrolysis process is Faradaic in nature and the behaviour observed is related to electrocatalysis. However, when comparing the electrolyser circuit in normal electrolysis to the electrolyser circuit in triode electrolysis, it can be said that the current (or potential) of the electrolyser circuit in triode electrolysis exceeds what would be expected from Faraday’s law in normal electrolysis. 3.5.3 Durability investigation 3.5.3.1 Long-term triode and tetrode operation in galvanostatic control Durability experiments were conducted in triode and tetrode operation in the galvanostatic mode using a CFP electrolyser ring WEelec at the same operating conditions as for Fig. 3.10. Similar to the prior galvanostatic tests, the electrolyser circuit was first operated by itself for 500 s. The auxiliary circuit was turned on and triode or tetrode operation was held for 3 or 5 hours. After this, the auxiliary circuit was stopped and normal electrolysis was carried-out for 500 s to confirm that the electrolyser circuit came back to its original potential values. The triode or tetrode effect attained a stable maximum about 800 s after the auxiliary circuit was turned on. Triode or tetrode operation was stable for the rest of the experiment, and normal electrolysis before and after the triode or tetrode operation was consistent in all cases. Figure 3.22 shows the durability test results obtained for 3 hours of normal, triode or tetrode operation under galvanostatic control. 127 Figure 3.22: Durability test in normal, triode and tetrode electrolysis in galvanostatic control (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2, iaux, cell = -50 mA/cm2). These tests were also repeated for 5 hours of operation. They demonstrated that triode and tetrode electrolysis is reliable over extended periods of operation without any signs of degradation, even when no precious metal catalysts are used on the electrolyser ring WEelec. As Fig. 3.22 indicates, the measured Uaux, cell, triode differed from the Uaux, cell, tetrode because the auxiliary CEaux, and the electrolyser CEelec have different surface areas in the triode than in the tetrode case, as described on Fig. 3.2 (d). The IR corrected Uelec, cell, triode and Uelec, cell, tetrode measured in triode and tetrode operation were almost the same. At the resulting triode and tetrode enhanced Uelec, cell values, the oxidation of CH3OH on the electrolyser ring WEelec is promoted, while carbon corrosion is avoided. Although the potential of the common CE in triode operation, or of the CEaux and CEelec in tetrode operation was not monitored for the galvanostatic control experiments presented in Fig. 3.22, it is known from other triode and tetrode experiments that the potential of the counter electrode(s) does not vary significantly during triode or tetrode electrolysis. Thus, the working electrode(s) potential is close to the cell voltage. -2.2 -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 1800 3600 5400 7200 9000 10800 Time (s) IR C or re ct ed C el l V ol ta ge (V ) Uelec, cell tetrode Uaux, cell triode Uelec, cell normal electrolysis Uaux, cell tetrode Uelec, cell triode Triode and tetrode Uelec, cell improvement IR C or re ct ed C el l V ol ta ge (V ) 128 3.5.3.2 Methanol concentration change over time in long-term triode or tetrode galvanostatic control During the normal, triode or tetrode electrolysis durability tests, samples of the anolyte solution were taken at 30 min. intervals. The samples were diluted for analysis by a K2Cr2O7 reduction technique used in combination with spectrophotometric measurements as in Chapter 2, and described in Appendix E in order to determine the change in CH3OH concentration over time. The dilution effect of sample extractions over time was taken into account in the CH3OH concentration calculations and in the Faradaic CH3OH concentration predictions. The methodology was validated by measuring the CH3OH concentration change over time in normal electrolysis in the galvanostatic control. The CH3OH concentration did not change significantly over the course of the normal electrolysis, and agreed with that predicted by Faraday’s law under the same conditions. Faraday’s law predicts that, at the normal electrolysis current employed, the change in CH3OH concentration would be about 0.001 % over 3 hours. The concentration results also confirm that there was no significant CH3OH evaporation over time due to reduced volume of reactant as solution samples were drawn out at ambient conditions. It was also verified that, when running the auxiliary circuit by itself in tetrode configuration under the same experimental conditions and for the same time period, the CH3OH concentration did not change over the duration of the experiment. Results obtained for the three hour normal, triode and tetrode electrolysis are shown on Figure 3.23. The K2Cr2O7 technique did not detect any significant CH3OH concentration changes over the first hour of the experiment. However, after the first hour, the triode and tetrode CH3OH concentration measurement trends indicate a decrease in the CH3OH concentration over time during triode and tetrode electrolysis, even though no catalyst was present on the electrolyser ring WEelec. 129 Figure 3.23: Methanol concentration as a function of time as predicted by Faraday’s law, and during normal, triode, and tetrode electrolysis in galvanostatic control (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, ielec, cell = 1.25 mA/cm2, iaux, cell = -50 mA/cm2). The slope of the concentration plots revealed that the measured CH3OH consumption was within 10 % for the triode and the tetrode operation cases. The CH3OH concentration measurements indicate that, in triode or tetrode operation, CH3OH was consumed at a rate exceeding the imposed electrolysis current. That is, the CH3OH consumption rate attained in triode and tetrode electrolysis exceeded what was expected from Faraday’s Law for normal electrolysis. It is possible to calculate the electrolysis circuit current density equivalent to the CH3OH consumption rate, i’elec, cell. For the triode operation case, i'elec, cell was 2130 mA/cm2, and for the tetrode operation case, i’elec, cell was 1911 mA/cm2. Both are three orders of magnitude greater than the imposed ielec,cell, of 1.25 mA/cm2 (2.11 mA). This calculated current density even exceeds the sum of the imposed ielec,cell and the imposed iaux,cell of -50 mA/cm2 (4 mA). The difference between the calculated CH3OH consumption rates and the one expected from the current density imposed may be explained by an increase in the low temperature catalytic reforming reaction rate which may result from an EPOC effect, as the WEelec potential, hence its work function, was affected. 0.0 0.5 1.0 1.5 2.0 2.5 0 2000 4000 6000 8000 10000 12000 Time (s) M et ha no l C on ce nt ra tio n (M ) Triode electrolysis Tetrode electrolysis Normal electrolysis Faraday's law 130 This non-Faradaic effect would imply that the catalytic reaction rate exceeds the rate of H+ transport, which is equal to the rate of the electrochemical reaction. The EPOC effect will be discussed further in Chapter 5. Another possible explanation is that the auxiliary current is indirectly contributing, fully or partly, to the electrochemical oxidation of methanol by pulling protons away from the WEelec surface. In the case of H2O electrolysis, the absorption strength of the H2O molecules at the WEelec surface may have decreased, rendering it easier to be electro-oxidized. In the case of CH3OH electrolysis, this in-situ WEelec surface potential change may help the removal of WEelec surface adsorbed CO, which normally constitutes the rate limiting step of CH3OH oxidation. The effect would be similar if the species absorbed to the WEelec surface were contaminants or poison, catalytic or not in nature. Surface contaminants reduce the Faradaic efficiency and hinder the WEelec’s performance. The WEelec surface adsorption limits mass transfer, affects the reaction mechanism, and impacts the double-layer capacitance. When CO is removed from the surface, a recovery in performance is expected, like it was seen in the case of triode or tetrode electrolysis. 3.5.3.3 Long-term tetrode operation in mixed potentiostatic and galvanostatic control Long-term durability electrolysis experiments were also conducted in tetrode operation with a CFP electrolyser ring WEelec in mixed control. For these tests, the tetrode circuits were controlled differently in order to demonstrate the advantage of the complete independence of the electrolyser and of the auxiliary circuits. The electrolyzer circuit was controlled potentiostatically, while the auxiliary circuit was independently controlled galvanostatically. This is not possible in triode operation, for which both circuits have to be controlled in the same operating control, i.e., both in galvanostatic or both in potentiostatic control, as there would be grounding interferences for mixed control as a common CE is used. The testing methodology employed was similar to the one used in the tests where both circuits were controlled galvanostatically. The electrolyzer circuit was operated by itself in normal electrolysis for 500 s before the auxiliary circuit was turned on. In these tests, tetrode operation was held for 30 minutes instead of 500 s. The auxiliary circuit was then turned off and re-started again for another 30 minutes. The shut-down times, which did not exceed two minutes, are indicated by the vertical dotted lines in Fig. 131 3.24. After a series of consecutive shut-down and start-ups, the auxiliary circuit was turned-off and normal electrolysis was carried out for another 500 s to confirm that the electrolyzer circuit came back to its original current/potential value. Figure 3.24: Potentiostatic electrolyser circuit control and galvanostatic auxiliary circuit control durability test in tetrode electrolysis (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CEelec, Pt auxiliary CEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 1 V vs. SHE, iaux, cell = -20 mA/cm2). As seen on Fig. 3.24, the tetrode effect attained a stable maximum about 50 s after the auxiliary circuit was turned on. Tetrode electrolysis was stable for the rest of the experiment, and it was determined that normal electrolysis results before and after tetrode operation were consistent. Therefore, tetrode electrolysis was reliable for 5 hours without any significant losses or degradation, even with auxiliary current interruptions and when no precious metal catalysts were used on the electrolyzer ring WEelec. In normal operation, the set CFP electrolyzer ring WEelec potential chosen (1 V vs. SHE) resulted in an electrolyzer circuit current, Ioelec,cell of 2x10-4 A. Nevertheless, with the application of an appropriate auxiliary circuit current in the opposite direction to the electrolyzer current (-20 mA/cm2 in this case), it was possible to raise the electrolyzer circuit current to about 0.4 A. This is orders of magnitude greater than what can be obtained on CFP during normal electrolysis. The tetrode enhancement was stable and reproducible, as the same electrolyzer circuit current was repeatedly obtained for each 30 min. segment -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 Time (s) I e le c, ce ll (A ) -0.01 -0.005 0 0.005 0.01 I a ux ,c el l ( A) start-up and shut-down Ielec, cell Tetrode Ielec, cell improvement Iaux, cell Ioelec, cell Ioaux, cell = 0 v I e le c, ce ll (A ) I a ux ,c el l ( A) 132 of tetrode operation. Consequently, it was demonstrated that, in tetrode operation, it is possible to control the electrolyzer circuit potentiostatically while independently controlling the auxiliary circuit galvanostatically for extended periods of time, and with multiple start-ups and shut-downs of the auxiliary circuit. 3.5.3.4 Methanol concentration change over time in long-term tetrode operation in mixed potentiostatic and galvanostatic control The regular shut-down times of the long-term durability tetrode electrolysis tests carried out in mixed control permitted the withdrawal of anolyte solution from the Wes compartment for CH3OH concentration analysis. The extracted samples were diluted and analyzed using the same K2Cr2O7/spectrophometric technique as before. The CH3OH concentration measured at different times during the experiments carried-out in Fig. 3.24 are shown in Fig. 3.25. Although not shown here, the tetrode electrolysis mixed control durability test carried out in Fig. 3.24 was also repeated at Uelec,we = 0.52 V vs. SHE and iaux,cell = -5 m/A/cm2, and the trend obtained for the CH3OH concentration change over time was very similar. The CH3OH concentration changes measured during mixed control tetrode operation significantly exceeded that measured in normal electrolysis. The calculated electrolyzer circuit current equivalent to the measured CH3OH consumption rate obtained in tetrode operation, i’elec, cell, was calculated to be 189 mA/cm2, which is within 20 % of the average measured Ielec, cell obtained in tetrode operation, which is 237 mA/cm2. Note that the power enhancement ratio, ρelec, cell, (Eq. 3.5) and electrolyzer power gain ratio, γelec, cell, (Eq. 3.6) could have been calculated with Ielec, cell and by replacing Ielec, cell by I’elec, cell if the electrolyzer CEelec potential would have been monitored during the experiments. This would have allowed Uelec,cell to be calculated, as Uelec,we was fixed. Under the tetrode mixed control conditions, the electrolyzer rate enhancement ratio, ρe (Eq. 3.7, dimensionless), can be calculated since it is the ratio of the electrolyzer current in tetrode operation to the electrolyzer current in normal operation. For the tetrode test carried out in at Uelec, we = 1 V vs. SHE, and iaux, cell = -20 mA/cm2, it was calculated to be 1600 using I’elec, cell and 2011 using Ielec, cell. In EPOC studies, promotional rate enhancement ratio values in the range of 10 to 100 are common [18]. 133 Figure 3.25: Methanol concentration as a function of time as predicted by Faraday’s law and as measured during potentiostatic electrolyser circuit control and galvanostatic auxiliary circuit control in tetrode electrolysis (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CEelec, and Pt auxiliary CEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 1 V vs. SHE, iaux, cell = -20 mA/cm2). The Faradaic efficiency, Λ (Eq. 3.8, dimensionless), is the ratio of the current gain attained in tetrode electrolysis vs. normal electrolysis over the auxiliary current. For the tetrode test carried out at Uelec, we = 1 V vs. SHE, iaux, cell = -20 mA/cm2, it was 160 using I’elec, cell and 192 using Ielec, cell. As Λ exceeded unity, the change in the electrochemical reaction rate is non-Faradaic as the H2 production rate in tetrode electrolysis exceeded the attainable H2 production rate when the electrolyzer was operating normally. These results provided further evidence that the CH3OH consumption rate obtained in triode and tetrode electrolysis exceed what can be obtained from normal electrolysis. The change in CH3OH concentration measured over time in tetrode operation was much more significant for the case for which the electrolyser circuit was controlled galvanostatically, then for the case for which the electrolyser circuit was controlled potentiostatically, i.e., by fixing the WEelec potential. In the galvanostatic control case, the electrolyser current density was fixed at 1.25 mA/cm2, while in the mixed control case, the electrolyser WEelec potential was fixed at 1 V vs. SHE, which corresponded to a current density of 237 mA/cm2 in tetrode operation. Based on this, it would be expected that the CH3OH Tetrode operation starts 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 Time (s) M et ha no l C on ce nt ra tio n (M ) Measured in tetrode operation Measured in normal operation (no auxiliary) Faraday's Law prediction in normal operation M et ha no l C on ce nt ra tio n (M ) 134 consumption rate in mixed control would exceed the CH3OH consumption rate in galvanostatic control. However, it was not the case. As it was mentioned earlier, the significant enhancement in the CH3OH consumption rate obtained in galvanostatic control may be linked to an EPOC effect which lowered the activation energy for the catalytic CH3OH reforming reaction at the electrolyser ring WEelec potential attained during the application of EPOE. Thus, combining these phenomena, it may be possible to carry out the catalytic reforming of CH3OH at ambient conditions using non-precious metal at the electrolyser ring WEelec. The EPOC phenomenon will be discussed in more details in Chapter 5. 3.5.4 Mechanism investigation 3.5.4.1 Electrolytic contact of auxiliary electrodes in potentiostatic control As explained earlier, when the auxiliary WEaux is in reduction, H2 is generated, and when the auxiliary WEaux is in oxidation, O2 and/or CO2 are generated. It was noticed that, when the contact between the auxiliary WEaux and the membrane was poor, an enhancement was still observed, but to a lesser extent. For this reason, the effect of generating the same amounts of H2 or O2, but 1 cm away from the membrane surface was examined in triode and tetrode operation. Results for tetrode operation are shown in Fig. 3.26. As shown in Fig. 3.26 (a), when the auxiliary WEaux located in the CH3OH or water electro-oxidation compartment was pulled 1 cm away from the membrane surface, no effect on the electrolyser geometric current density could be observed, even when the area in front of the auxiliary WEaux was saturated with H2. The auxiliary WEaux needs to be in ionic contact with the membrane in order to observe the beneficial triode effect. Thus, triode or tetrode enhancements can not be achieved with molecular H2 and therefore appear to be protonic in nature. When bubbling an equivalent amount of H2 away from the electrolyser ring WEelec surface, no similar enhancement effect was obtained. The increased availability of H+ may contribute to enhance the protonic conductivity of the membrane in triode operation. This will be looked at in Section 3.5.4.2. 135 Figure 3.26: Effect of varying Uaux, we on the electrolyser current density in tetrode operation (a) Pt auxiliary WEaux 1 cm away from surface (b) Pt auxiliary CEaux 1 cm away from surface (CFP electrolyser ring WEelec, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). (a) (b) -1.00 -0.75 -0.50 -0.25 0.00 0.25 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) -3.5E-02 -2.5E-02 -1.5E-02 -5.0E-03 5.0E-03 1.5E-02 2.5E-02 3.5E-02 E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) iaux, cell ielec, cell -1.00 -0.75 -0.50 -0.25 0.00 0.25 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) -3.0E-02 -2.0E-02 -1.0E-02 0.0E+00 1.0E-02 2.0E-02 3.0E-02 E le ct ro ly se r G eo m et ric C ur re nt D en si ty (m A /c m2 ) iaux, cell ielec, cell A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (m A /c m2 ) 136 In tetrode operation, the independent auxiliary CEaux was pulled 1 cm away from the surface. As shown on Fig. 3.26 (b), a slight enhancement was observed but it was not as significant as when the CEaux was in electrolytic contact with the membrane, probably due to an increase in the circuit resistance. Hence, it appears that the auxiliary CEaux does not need to be in electrolytic contact with the membrane. 3.5.4.2 Effect of acid concentration and conductivity on triode operation in potentiostatic control Water electrolysis was conducted on a CFP electrolyser ring WEelec using 0.05, 0.5 and 5 M H2SO4 as the electrolyte in the WE compartment in order to evaluate the effect of H+ concentration with and without the auxiliary circuit. Figure 3.27 shows the effect of acid concentration on membrane and solution conductivity. The enhancements obtained in conductivity are in a similar range to other data reported in the literature [195-197]. The intrinsic increases in conductivity with increasing electrolyte concentration are not significant enough to account for the triode operation enhancement observed. Figure 3.27: Nafion® N-117 membrane conductivity and solution conductivity as a function of sulphuric acid concentration. In general it is assumed that there is sufficient supporting electrolyte to transfer the ions from the bulk of the solution to the electrode surface by diffusion. Potentiostatic mode results for water electrolysis with different acid concentrations are shown in Fig. 3.28. They were obtained under different conditions than 0 10 20 30 40 50 60 70 80 0 1 2 3 4 5 6 Sulfuric Acid Concentration (M) N -1 17 M em br an e C on du ct iv ity (m S/ cm ) 0 100 200 300 400 500 600 700 800 S ol ut io n C on du ct iv ity (m S /c m ) Fully Hydrated N117 in H O at 25 C Partially Hydrated N117 in Air at 21 C Solution Conductivity at 25 C ° ° ° 2 137 the tests carried out in the galvanostatic mode in Fig. 3.11 b). As explained earlier, in the potentiostatic mode, Uelec, we was set to 0.52 V vs. SHE to avoid carbon corrosion of the electrolyser CFP ring WEelec. It was verified that changing this value to 0.6 or 0.7 V vs. SHE did not affect the results significantly. Figure 3.28: Effect of varying Uaux, we on the electrolyser current density (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 0 M CH3OH in 0.05 or 5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). Fig. 3.28 indicates that, in the potentiostatic mode, the relative difference between Uelec, we and Uaux, we had a direct influence on the magnitude of the triode enhancement obtained. This is in agreement with the galvanostatic results obtained in (Fig. 3.24 b)). In the galvanostatic mode or in the potentiostatic mode, the relative difference between the current/potential of the electrolyser circuit to the current/potential of the auxiliary circuit has an impact on the triode enhancement obtained. Most importantly, Fig. 3.28 demonstrates that increasing the electrolyte concentration or solution pH had no significant effect when the triode was not in use and resulted in only slight enhancements in triode operation. Consequently, the triode enhancement can not be primarily attributed to an increase in the -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) 0.05 M H SO , triode operation 5 M H SO , triode operation 0.05 M H SO , without triode operation 5 M H SO , without triode operation 2 4 2 4 2 4 2 4 138 bulk H+ concentration of the electrolyte in the entire WE compartment. This indicates that the largest part of the performance enhancement is related to localized effects at the electrodes rather than the electrolyte. 3.5.4.3 Effect of triode operation on membrane conductivity in potentiostatic control Electrochemical impedance spectroscopy (EIS) measurements were performed while operating the electrolyser in normal potentiostatic control for triode operation. These measurements were taken to determine if the resistance of the MEA decreased in triode electrolysis compared to normal electrolysis. It was previously suggested that during triode fuel cell operation, the conductivity of the Nafion® polymer improved by lowering the overpotential for proton tunnelling between the membrane sulphonated groups (proton tunnelling was reported to increase exponentially with potential), which would improve proton migration through the membrane [173, 198]. The data points shown in Fig. 3.29 represent the average of three different tests for normal and triode electrolysis. The measurements taken before 200 s indicate that the MEA resistance is less in triode electrolysis than in normal electrolysis. However, after 200 s, the resistance stabilizes and the MEA resistance is on average about 0.03 ohms larger for triode electrolysis compared to normal operation. It confirms that the membrane conductivity does not improve due to the H+ flux produced by the auxiliary circuit. Therefore, the beneficial effect observed for triode operation appears to be limited to the working electrode surfaces only. 139 Figure 3.29: Resistance measurement over time in normal electrolysis and in triode electrolysis (CFP electrolyser ring WEelec, Pt auxiliary WEaux, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE, Uaux, we = -0.59 V vs. SHE). 3.5.4.4 Effect of electrolyser working electrode ring geometry in potentiostatic control Various triode arrangements geometries could be considered, but the discussion is limited to the cell configuration described in Section 3.1.2. The impact of the electrolyser ring WEelec dimension was evaluated by changing the gap distance between the auxiliary WEaux Pt disk and electrolyser ring WEelec. In Fig. 3.30, the effect of a small gap distance (3.18 mm) was compared to a larger gap distance (6.73 mm). More significant enhancements were observed when the gap distance between the electrolyser ring WEelec and the auxiliary WEaux Pt disk was smaller. A larger distance might negatively affect the spill- over of proton charges from the auxiliary WEaux to the electrolyser ring WEelec. The gap size effect was more important in the case of 2 M CH3OH than in the case of H2O electrolysis. Hence, the spillover of protons appears to enhance the complex electro-oxidation mechanism of CH3OH depolarized H2O electrolysis and have less effect on the mechanism of H2O electrolysis. 0.30 0.31 0.32 0.33 0.34 0.35 0.36 0.37 0.38 0.39 0.40 0 200 400 600 800 1000 Time (s) Re si st an ce (o hm ) Normal electrolysis Triode operation 140 Figure 3.30: Effect of varying Uaux, we on the ielec, cell for different electrolyser ring WEelec gap distances (CFP electrolyser ring WEelec, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). 3.5.4.5 Effect of proton flux line direction in potentiostatic control The effect of changing the proton flux lines direction through the MEA was investigated in tetrode operation in the potentiostatic mode. For a normal electrolyser with a single electrolyser ring WEelec and electrolyser CEelec, the current lines are straight. The electrolyser current lines will be different when an auxiliary WEaux is added and used in the reverse direction. The local H+ concentration around the auxiliary WEaux may be higher, and a concentration gradient of protons might result over the entire surface, with the peak in H+ concentration on the auxiliary WEaux. In addition, CH3OH crossover from the electrolyser ring WEelec to the electrolyser CEelec might be hindered by the proton flux from the CEaux to the auxiliary WEaux. Fig. 3.31 shows the effect of inverting the CEs or inverting the WEs in tetrode operation. In Fig. 3.31 (a), the auxiliary CEaux is the ring and the electrolyser CEelec is the Pt rod. In Fig. 3.31 (b) the auxiliary WEaux is the ring and the electrolyser WEelec is the Pt rod. -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) Large gap, 2 M CH OH Small gap, 2 M CH OH Large gap, 0 M CH OH Small gap, 0 M CH OH 3 3 3 3 141 In general, a weak to no effect was observed when the current lines were intersecting each other. A slight effect was observed when the CEs were inverted; however, it is so small that it is not significant. The beneficial tetrode effect was not obtained when the WEs were inverted. It is likely that, when the large CFP WE ring is the WEaux, less H+ is generated per surface area than when the WEaux is the smaller Pt rod tip surface area. Consequently, it is possible that there is less H+ spilling over from the CFP ring WEaux to the Pt WEelec than when it is spilling over from the Pt WEaux to the CFP ring WEelec. Also, it is possible that the misalignment of the electric fields generated when the WEs or CEs are inverted creates dynamic transport issues through the solid electrolyte (crossover, H2O transport, etc.), while the alignment of the electric fields improves the transport and kinetics of the electron transfer reaction at the WEelec surface. 142 Figure 3.31: Effect of varying Uaux, we on the current density in tetrode operation (a) inverted CEs (CFP ring WEelec, Pt rod WEaux, Pt rod CEelec, Pt/C CEaux) (b) inverted WEs (Pt rod WEelec, CFP ring WEaux, Pt/C ring CEelec, Pt rod CEaux), (2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). (a) (b) -0.75 -0.50 -0.25 0.00 0.25 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) -1.2E-02 -1.0E-02 -8.0E-03 -6.0E-03 -4.0E-03 -2.0E-03 0.0E+00 2.0E-03 4.0E-03 6.0E-03 8.0E-03 1.0E-02 1.2E-02 E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) iaux, cell ielec, cell -0.75 -0.50 -0.25 0.00 0.25 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) -1.2E-02 -1.0E-02 -8.0E-03 -6.0E-03 -4.0E-03 -2.0E-03 0.0E+00 2.0E-03 4.0E-03 6.0E-03 8.0E-03 1.0E-02 1.2E-02 E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) iaux, cell ielec, cell A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) 143 3.5.4.6 Effect of the electrochemical surface area of the carbon fibre paper electrolyser working electrode in potentiostatic control The use of triode and tetrode electrolysis in the absence of PGM catalyst at the electrolyser ring WEelec was demonstrated in Sections 3.5.1 and 3.5.2. It was also confirmed that no CFP corrosion was occurring over the operating potential region used during the galvanostatic and potentiostatic tests. In all experiments so far, a simple plane carbon fibre paper was employed as the electrolyser ring WEelec. In this section, an MPL at a loading of 1 mg/cm2 (composed of carbon black and 20 wt% PTFE) was sprayed on the CFP surface and was tested as the electrolyser ring WEelec to investigate the influence of the surface area and CFP surface morphology. Figure 3.32 compares the results obtained with this three-dimensional high surface area CFP electrolyser ring WEelec to the ones obtained with the planar CFP fibre paper electrolyser ring WEelec. Using a high surface area 3-dimensional CFP based electrolyser ring WEelec did not improve the operation of the triode electrolyser significantly compared to when a lower surface area carbon based electrolyser ring WEelec was used. The maximum improvement in electrolyser geometric current density obtained was about 16 % at the maximum Uaux, we value of -0.6 V vs. SHE in the H2 production region. The pore structure of the carbon based surface might have affected the chemisorption strength of adsorbed species. It would be expected that a structure with more pores and channels would lead to a larger triple phase boundary (TPB) which would in turn improve its activity. Microscopic surface analysis measurement techniques would be required to confirm exactly how H+ catalyzed the carbon based surfaces. 144 Figure 3.32: Effect of electrolyser ring CFP WEelec carbon SA on ielec, cell (2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). 3.5.4.7 Effect of triode operation using non-noble non-carbon containing working electrode materials in potentiostatic control Triode and tetrode operation were shown to promote the electrocatalytic properties of the CFP electrolyser ring WEelec. Carbon oxidation, which could affect the CFP electrolyser ring WEelec, was avoided by restricting the electrolyser operating voltage range. However, in the potentiostatic mode, it was possible to subject the auxiliary WEaux material (Pt) to a wider range of potentials than the electrolyser ring WEelec material (CFP or noble metal catalyst GDE). It may be possible to operate the electrolyser over a wider range of currents/potentials if a suitable carbon-free electrolyser ring WEelec material can be used. The possibility of conducting the electro-oxidation of CH3OH on a non-precious carbon-free electrolyser ring WEelec was investigated in triode operation. -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m2 ) -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 E le ct ro ly se r G eo m et ric C ur re nt D en si ty (A /c m2 ) iaux, cell, high SA CFP ring iaux, cell, CFP ring ielec, cell, high SA CFP ring ielec, cell, CFP ring 145 In order to achieve this, appropriate materials needed to be selected. Ideally, the electrolyser ring WEelec material would exhibit corrosion resistance over a wide range of electrolyser ring WEelec currents/potentials in 0.5 M H2SO4. In addition to conducting electricity and possessing a good stability, the electrolyser ring WEelec material should allow for the migration and spreading of the H+ ions generated at the auxiliary WEaux surface. In addition, material corrosion or degradation may negatively impact the PEM conductivity due to ion contamination. Pourbaix diagrams demonstrate in which state a material will be at equilibrium at specific pH values and potentials, but do not provide information about the material surface kinetics for a particular electrochemical reaction. Ideally, based on the potentiostatic studies carried out in Section 3.5.2, the electrolyser material would be immune or form a passive film at a pH of 0 (0.5 M H2SO4) and at anodic potential of ≥ 0.5 V vs. SHE at ambient conditions. Similarly, in reverse operation, the ideal auxiliary WEaux material would be immune or form a passive film at a pH of 0 in a potential range ≤ 0.5 V vs. SHE at ambient conditions. Grade 316 SS is an austenitic SS, which contains Mo for higher corrosion resistance. According to its Pourbaix diagram, at pH = 0 and potentials greater than -1 V vs. SHE, the 16 to 18 % of Cr contained in the 316 SS will eventually start leaching out, and the 316 SS will corrode [199]. However, gold is stable at pH = 0, up to potentials of about 1.4 V vs. SHE [200]. Different materials can be used for the triode electrolyser ring WEelec than for the auxiliary WEaux. For example, the electrolyser ring WEelec could be an electrochemically etched fine titanium mesh, as Ti is passive at potentials greater than 0 V vs. SHE [201]. In the same triode experiment, the auxiliary WEaux rod could be made of niobium, which does not exhibit corrosion at potentials less than -0.3 V vs. SHE [202], if the auxiliary WEaux, is operated in the H2 production region at potentials lower than this value. All experiments were conducted in a 0 or 2 M CH3OH acidic anolyte and a 0.5 M H2SO4 catholyte, for which more information was provided in Chapter 2. For carbon free non-precious electrolyser ring WEelec experiments a corrosion resistant 316 SS woven wire cloth (McMaster-Carr) with a mesh size of 400 x 400, a wire diameter of 0.025 mm, individual opening dimensions of 0.038 mm x 0.038 mm and an open area of 36 % was used. In order to further protect the 316 SS mesh from corrosion and evaluate a different metal, for some tests, it was electrochemically plated with a 2.5 μm Ni barrier layer followed by a 5 μm layer of 24K Au (Acme Plating & Silver Shop Ltd.). For the non-precious metal carbon free auxiliary WEaux experiments, a corrosion-resistant 3.18 mm OD 316 SS rod (passivated and annealed, 146 McMaster-Carr). In some cases, it was plated with Ni and Au as for the 316 SS mesh electrolyser ring WEelec. The auxiliary rod was enclosed in a transparent PTFE heat shrink tubing (McMaster-Carr), leaving only the tip exposed. Like in Section 3.3.4, CV measurements were conducted in a single compartment cell comprising a three electrode system to confirm carbon-free electrode material stability in water and CH3OH acidic solutions and obtain their on-set potentials. The electrodes tested comprised Pt, 316 SS and Au plated 316 SS electrolyser ring WEelec and auxiliary WEaux rod materials. Fresh components were used for every different solution, since conducting CVs over a wide potential range may have been destructive to the 316 SS and to the Au plated 316 SS surfaces. Figure 3.33 shows the forward anodic potential sweeps only. The CVs were all corrected for ohmic resistance from impedance measurements taken after the CV was performed. In some cases, this resulted in a slight overcorrection of the cyclic voltammogram at high potentials. The current was more stable for the rods, than for the electrolyser ring WEelec meshes, which have a greater surface area exposed to the solution. For the 316 SS mesh, it seems that two different electrochemical processes are occurring. This was not the case for the Au plated 316 SS mesh. Also, in H2O electrolysis, the current was noticeably smaller on the Au plated 316 SS mesh than on the non- plated one. This seems to indicate that there is significant corrosion happening for the non-plated 316 SS mesh. For convenience, the IR corrected on-set potential values obtained for the selected non- carbon containing materials, i.e., 316 SS and Au plated 316 SS electrolyser ring WEelec components, as well as the Pt auxiliary WEaux are summarized in Table 3.4, along with the IR corrected on-set potential previously obtained for Pt black and CFP electrolyser ring WEelec at varying CH3OH concentrations. 147 Figure 3.33: IR corrected forward anodic potential sweeps (a) 316 SS auxiliary WEaux (b) 316 SS mesh electrolyser ring WEelec (c) Au plated 316 SS auxiliary WEaux (d) Au plated 316 SS mesh electrolyser ring WEelec (e) Pt auxiliary WEaux (5 mV/s, 0, 2, or 16 M CH3OH in 0.5 M H2SO4, Pt flag CE, 23 ±1oC). 0 0.05 0.1 0.15 0.2 0.25 0.3 1 1.25 1.5 1.75 2 2.25 IR Corrected V vs SHE C ur re nt (A ) 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 1 1.25 1.5 1.75 2 2.25 IR Corrected V vs SHE C ur re nt (A ) 0 0.01 0.02 0.03 0.04 0.05 0.06 1 1.5 2 2.5 IR Corrected V vs SHE C ur re nt (A ) 0 0.05 0.1 0.15 0.2 0.25 0.3 1 1.25 1.5 1.75 2 2.25 2.5 IR Corrected V vs SHE C ur re nt (A ) 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 1 1.25 1.5 1.75 2 2.25 IR Corrected V vs SHE C ur re nt (A ) (a) (b) (c) (d) 0 M CH3OH 0 M CH3OH 0 M CH3OH 0 M CH3OH 2 M CH3OH 2 M CH3OH 16 M CH3OH 16 M CH3OH 2 M CH3OH (e) 0 M CH3OH 2 M CH3OH 16 M CH3OH 16 M CH3OH 2 M CH3OH 16 M CH3OH C ur re nt (A ) C ur re nt (A ) C ur re nt (A ) C ur re nt (A ) C ur re nt (A ) 148 Table 3.4: IR corrected on-set potential for different electrolyser and auxiliary WEaux materials at varying CH3OH concentrations. Electrolyser ring WEelec Auxiliary WEaux [CH3OH] Pt black* CFP* 316 SS mesh Au plated 316 SS mesh Pt rod 316 SS rod Au plated 316 SS rod [M] [V vs. SHE] 0 1.70 1.86 1.34 1.87 2.17 1.93 2.06 2 0.60 1.96 1.26 1.83 1.90 1.90 1.81 16 0.65 2.00 1.23 2.05 1.86 1.90 1.88 *From Table 3.2. By comparing the values listed, it can be seen that only the Au plated 316 SS mesh had a greater on-set potential than the Pt black but was close to that of the CFP for H2O electrolysis. In CH3OH solutions, the on-set potential of the Au plated 316 SS mesh was about 1.3 V greater than that of Pt black but close to that of the CFP. In H2O electrolysis, the on-set potential of the 316 SS rod was 0.24 V less than that of the Pt rod, but 0.59 V greater than for the 316 SS mesh. It appears that corrosion is occurring on the 316 SS surface (particularly the mesh), probably preferentially at the ring edges, which were cut-off. This was visually confirmed after carrying the experiments. During the CVs of 316 SS components, the solution gradually became more yellow over time, indicating that Cr was leaching out. It is also possible that any of the components may contain impurities which may be electrochemically active contaminants. The contaminants leaching out likely negatively impact the kinetics of the electro-oxidation, electrolyte conductivity, and in turn, hinder the electrochemical performance of the system. Therefore, it is important to ensure components are stable and relatively free of contaminants. The addition of an Au coating on the 316 SS mesh moved its on-set potential value close to that of the CFP. Thus, the Au plated 316 SS components are more stable than 316 SS components. All potentiostatic triode experiments were conducted at potentials below the on-set potential values listed in Table 3.4. The tests carried out using 316 SS components are not reported as they were inconclusive due to corrosion issues. Therefore, only the results obtained using Au plated 316 SS components are discussed in this Section. 149 Various Au plated 316 SS combinations with the CFP electrolyser ring WEelec and the Pt auxiliary WEaux were studied in the potentiostatic mode as shown in Fig. 3.34. The operating range of the triode electrolyser was limited to auxiliary WEaux potentials from -0.2 to 1.02 V vs. SHE. Figure 3.34: Effect of varying Uaux, we on (a) Au plated 316 SS electrolyser ring WEelec, Pt auxiliary WEaux, 0 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, (b) Au plated 316 SS electrolyser ring WEelec, Pt auxiliary WEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, (c) Au plated 316 SS electrolyser ring WEelec, Au plated 316 SS auxiliary WEaux, 0 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, (d) Au plated 316 SS electrolyser ring WEelec, Au plated 316 SS auxiliary WEaux, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte (triode, Pt/C CE, Uelec, we = 0.52 V vs. SHE). By comparing Fig. 3.34 (a) and (c), it can be observed that, at a Uaux, we of -0.25 V vs. SHE, the difference in the auxiliary geometric current density and in the electrolyser geometric current density obtained in the case of the Au plated electrolyser ring WEelec with the Pt auxiliary WEaux or with the Au plated 316 SS auxiliary WEaux was within experimental error, i.e., 12 %. Thus, in 0 M CH3OH, there were no major (a) (b) (c) (d) -40 -30 -20 -10 0 10 20 30 40 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m 2 ) -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 El ec tro ly se r G eo m et ric C ur re nt D en si ty (A /c m 2 ) Iaux, cell Ielec, cell -140 -120 -100 -80 -60 -40 -20 0 20 40 60 80 100 120 140 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m 2 ) -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 El ec tro ly se r G eo m et ric C ur re nt D en si ty (A /c m 2 ) iaux, cell ielec, cell -40 -30 -20 -10 0 10 20 30 40 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m 2 ) -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 El ec tro ly se r G eo m et ric C ur re nt D en si ty (A /c m 2 ) iaux, cell ielec, cell -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Auxiliary Working Electrode Potential (V vs. SHE) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m 2 ) -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 El ec tro ly se r G eo m et ric C ur re nt D en si ty (A /c m 2 ) iaux, cell ielec, cell A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m 2 ) El ec tro ly se r G eo m et ric C ur re nt D en si ty (A /c m 2 ) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m 2 ) El ec tro ly se r G eo m et ric C ur re nt D en si ty (A /c m 2 ) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m 2 ) El ec tro ly se r G eo m et ric C ur re nt D en si ty (A /c m 2 ) A ux ili ar y G eo m et ric C ur re nt D en si ty (A /c m 2 ) El ec tro ly se r G eo m et ric C ur re nt D en si ty (A /c m 2 ) 150 differences observed when the auxiliary WEaux was made of Pt or Au plated 316 SS. However, there is a difference in electrochemical activity on these materials for CH3OH oxidation and for H2O oxidation, i.e., the on-set potentials are different. The greatest triode effects were obtained at Uaux, we in the reduction potential region when using an Au plated 316 SS electrolyzer ring WEelec and a Pt auxiliary WEaux in 2 M CH3OH, as shown in Fig. 3.34 (b). Thus, the auxiliary WEaux and the electrolyser ring WEelec do not need to be catalytic materials in order to observe the triode effect on the electrolyser ring WEelec. Table 3.5 summarizes the ρe values obtained in triode electrolysis as the Uaux, we was varied when using different combinations of Au plated 316 SS WEs components. In general, the ρe values increases as Uaux, we increased. All peak ρe were observed when Uaux, we was -0.14 V vs. SHE or lesser, i.e., in the H2 evolution region. Table 3.5: Effect of auxiliary WEaux potential (Uaux, we) on the rate enhancement ratio (ρe) using Au plated 316 SS components (2 mg/cm2 Pt/C CE, Uelec, we = 0.52 V vs. SHE). WEelec and WEaux Material Au plated 316 SS WEelec, Pt WEaux Au plated 316 SS WEelec, Au plated 316 SS WEaux CFP WEelec, Au plated 316 SS WEaux [CH3OH] Uaux, we ρe [M] [V vs. SHE] [Dimensionless] 0 0.42 42 265 23 0 0.34 189 4294 53 0 -0.05 818 17405 157 0 -0.14 869 18519 255 0 -0.25 756 21624 491 2 0.42 489 222 79 2 0.34 929 475 28 2 -0.05 3201 1674 954 2 -0.14 3819 1781 1603 2 -0.25 5638 2040 360 151 The ρe for the Au plated 316 SS electrolyser ring WEelec and Pt auxiliary WEaux combination were the same order of magnitude as the ones reported in Table 3.2 for various electrolyser ring WEelec materials and the same WEaux. It appears that most of the fluctuation in ρe values reported are caused by differences in the material surface properties, conductivities and structures. The Au plated 316 SS mesh electrolyser ring WEelec possesses large pores and likely has a smaller ECSA compared to the electrolyser CFP gas diffusion ring WEelec. Therefore, it is probable that the protons generated by the Pt auxiliary WEaux are only reaching the inner edge of the WEelec ring and do not spread out evenly over the entire Au plated 316 SS electrolyser mesh ring WEelec surface during triode electrolysis, explaining why the ρe values obtained were greater for the CFP electrolyser ring WEelec than for the Au plated 316 SS electrolyser ring WEelec. While the entire surface of the mesh ring WEelec is considered for normal electrolysis, the promoted electro-oxidation of water or CH3OH may be limited to the inner edge of the electrolyser mesh ring WEelec. Therefore, H+ spillover to the electrochemical oxidation sites may be more restricted for the Au plated 316 SS electrolyser ring WEelec mesh than for the CFP, which has a denser and more connected structure. The Au plated 316 SS WEs material combination for which the largest ρe was obtained occurred when both, the electrolyser and the auxiliary were made of Au plated 316 SS in 0 M CH3OH. According to this, the Au plated 316 SS ring WEelec and WEaux combination results in greater triode enhancements for H2O electrolysis than for CH3OH depolarized electrolysis. It was shown in Table 3.3, for the same Pt auxiliary WEaux material, the electrolyser ring WEelec with precious metals, which are known to be catalytic to the electro-oxidation of H2O or CH3OH, did not result in larger triode enhancements than that obtained for the CFP, which is not catalytic during normal electrolysis. Therefore, larger ρe values are expected when the electrolyser ring WEelec surface is non-catalytic to the electro-oxidation of H2O or CH3OH, which is the case for Au plated 316 SS components. In 2 M CH3OH, using a Au plated 316 SS electrolyser ring WEelec with the Pt auxiliary WEaux resulted in a ρe about two times greater than when using a Au plated 316 SS electrolyser ring WEaux. Similarly, also in 2 M CH3OH and with the Pt auxiliary WEaux, using a CFP electrolyser ring WEelec resulted in a ρe 26 times greater than when using the Au plated 316 SS electrolyser ring WEelec at Uaux, we = -0.14 V vs. SHE. This demonstrates that it may be possible to use a WEelec and WEaux combination made of non-precious, non-carbon containing materials to gain the triode electrochemical promotion effect, as it is the proton pumping which is important. 152 Since the common CE potential was independently measured, it was possible to calculate the power gain ratio, γelec, cell (Eq. 3.6). Figure 3.35 shows γelec, cell variation with Uaux, we for the various WE material combinations studied in 0 and 2 M CH3OH. Figure 3.35: Power gain ratio (γelec, cell) as a function of Uaux, we (a) 0 M CH3OH, (b) 2 M CH3OH (Au plated 316 SS components, 2 mg/cm2 Pt/C CE, 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, Uelec, we = 0.52 V vs. SHE). All γelec, cell were less than unity and some γelec, cell were negative, indicating that the power required to operate in triode electrolysis was less than that required to operate in normal electrolysis. Negative γelec, cell values were obtained for the Au plated 316 SS electrolyser ring WEelec and Pt auxiliary WEaux combination. Peak γelec, cell values less than -1 were obtained which demonstrate that it is possible that (a) (b) -35 -30 -25 -20 -15 -10 -5 0 5 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Auxiliary Working Electrode Potential (V vs. SHE) P ow er G ai n R at io (d im en si on le ss ) Au plated 316 SS electrolyzer ring, Pt auxiliary rod CFP electrolyzer ring, Au plated 316 SS auxiliary rod Au plated 316 SS electrolyzer ring, Au plated 316 SS auxiliary rod -35 -30 -25 -20 -15 -10 -5 0 5 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Auxiliary Working Electrode Potential (V vs. SHE) P ow er G ai n R at io (d im en si on le ss ) Au plated 316 SS electrolyzer ring, Pt auxiliary rod CFP electrolyzer ring, Au plated 316 SS auxiliary rod Au plated 316 SS electrolyzer ring, Au plated 316 SS auxiliary rod P ow er G ai n R at io (d im en si on le ss ) P ow er G ai n R at io (d im en si on le ss ) 153 the decrease in the electrolyser power requirement during triode operation exceeds the power consumed in the auxiliary circuit. When both WEs were made of Au plated 316 SS, the peak γelec, cell obtained was - 35 in 0 M CH3OH and -12 in 2 M CH3OH at a Uaux, we of -0.05 V vs. SHE, i.e., when the auxiliary WEaux is in the reduction region. In all cases, including the results obtained on Fig. 3.21 and 3.35, a maximum γelec, cell seems to be located between a Uaux, we of -0.14 and 0.42 V vs. SHE. The average absolute value of the Faradaic efficiency of H2 production, Λ, (Eq. 3.8) were in the range of 0.8 to 1.02 for the different WEs materials combinations evaluated in the reverse potentiostatic mode for the CH3OH depolarized H2O electrolysis and normal H2O electrolysis. According to this, the triode electrolysis process using carbon-free non-precious WEs materials would be Faradaic and electrocatalytic in nature. However, in normal H2O electrolysis, Λ exceeded unity when the auxiliary WEaux potential was -0.05 and -0.14 V vs. SHE for the CFP electrolyser ring WEelec and Au plated 316 SS WEaux combination indicating some non-Faradaic activity. 3.5.4.8 Proposed mechanism A number of tests were done to determine what triggers the performance enhancement obtained in triode and tetrode electrolysis when in reverse operation, and its effect on the electrolyser ring WEelec surface. A general triode mechanism was proposed as shown in the schematic of Fig. 3.36 based on the information acquired from the tests conducted in reverse operation to date. Additional experiments to further clarify the mechanism are recommended Section 5.5.2. 154 Figure 3.36: Schematic diagram of the proposed triode/tetrode mechanism in reverse electrolysis operation. The tests conducted in this study were meant to elucidate some of the interactions occurring between the auxiliary WEaux and the electrolyser ring WEelec surfaces located in the CH3OH or H2O oxidation compartment. The electrochemical triode or tetrode enhancement was only observed when the operation of the auxiliary was opposite to that of the electrolyser circuit. It was proportional to the current/potential of the auxiliary circuit and was reproducible. The effect was beneficial to the electrolyzer potential only when the auxiliary circuit was operating in reduction. Similar results were obtained for both the triode and the tetrode configuration. In the tetrode configuration, the auxiliary and electrolyzer circuits were completely independent with no electronic contact or interaction. The electrolyzer ring electrode did not require a precious group metal (PGM) electrocatalyst to achieve an electrolyzer current/potential enhancement in triode or tetrode operation. Although the magnitude of the (N-117) Porous & Conductive Electrolyzer WE Ring Conductive Auxiliary WE H+ H+ H+ H+ H+ H+ H+ H+ 6H+ 6H+ 2H+ CH3OH + H2O CO2 H2 (Pt/C) W E e le c W E a ux 6H+ + 6e-→ 3H2 CH3OH + H2O → CO2 + 6H+ + 6e- H2 → 2H+ + 2e- H+ H+ H+ H+ H+ H+ H+ H+ H2O → 1/2O2 + 2H+ + 2e- 2H+ + 2e-→ H2 W E e le c C E a ux Conductive Auxiliary CE C E e le c C E e le c Porous & Conductive Electrolyzer CE Ring H+ H+ W E e le c W E a ux W E e le c C E a ux C E e le c C E e le c 155 electrocatalytic rate enhancement ratio obtained depended on the electrolyzer ring WEelec material, neither carbon oxidation nor hydrogen oxidation was responsible for the rate enhancement. The same triode or tetrode effect was obtained using non-catalytic materials for the auxiliary WEaux (e.g., Au plated SS, etc.), which just forces protons through the membrane but with no surface electrochemical reaction, i.e., no H2 is produced. It was observed that the auxiliary WEaux needed to be in physical contact with the membrane for a significant triode or tetrode effect to occur. Pulling the auxiliary WEaux away from the membrane surface, further confirmed that the triode improvement is mainly caused by the electrochemical pumping of protons rather than by the generation of molecular H2 from a catalytic auxiliary WEaux. Finally, it was determined that the performance enhancement obtained in triode and tetrode operation was linked to the working electrode surfaces rather than a change in the intrinsic membrane or solution conductivity. All these observations appear to agree with a triode or tetrode mechanism involving localized H+ species rather than molecular H2. In this study, it was shown that the H+ flux to the auxiliary WEaux surface in cathodic operation had a promotional effect which proportionally affected the properties of the electrolyzer ring electrode surface. It appears that the localized in-situ migration of mobile promoting H+ species, either from the auxiliary WEaux to the anodic active electrocatalyst surface of the electrolyzer or vice-versa, is responsible for the effect observed. Surface charge differences are likely the driving force which makes the H+ ions migrate to/from the auxiliary WEaux surface from/to the electrolyzer ring electrode surface. However, in the case of water or methanol electrolysis, the triode or tetrode improvement was only observed when the proton flux was directed from the WEelec to the WEaux. The local removal of H+ from the WEelec surface appears to drive the electrochemical oxidation reaction consistent with Le Châtelier’s principle and a change in work function for galvanostatic control, thus improving the electrolysis charge transfer rate. It also explains why the gap between the WEaux and the WEelec has to be kept as small as possible. This is analogous to the EPOC effect where there is a backspillover of promoting ions occurring from the support to the metal surface. The main difference here is that the promoted reaction is an electrochemical reaction and not a chemical reaction. 156 3.6 Summary The triode and tetrode operation of a low temperature H2O or CH3OH PEM electrolyser, running under non-flowing conditions was observed under various operating modes. The electrolyser and the auxiliary circuits were both galvanostatically, or both potentiostatically controlled. In tetrode operation, it was even possible to use potentiostatic control for the electrolyser circuit while using galvanostatic control for the independent auxiliary circuit. Under some electrochemical promotion conditions obtained via triode or tetrode operation, the electrolysis performance improved with a decrease in energy required, i.e., less current/potential was needed in the electrolysis circuit when the auxiliary circuit was in operation. It was demonstrated that the effect was durable, reproducible, stable and reliable over periods of times up to 5 hours and upon multiple start-up and shut-down cycles of the auxiliary circuit. It was proposed that triode and tetrode operation enhanced the CH3OH electro-oxidation rate through electrochemical pumping and backspillover of protons. Therefore, the effect observed is the electrochemical promotion of electrocatalysis (EPOE) and not the electrochemical promotion of catalysis (EPOC). The only case where the latter might have been observed was when both circuits were controlled galvanostatically. Nevertheless, triode or tetrode operation appear to be a potential way to reduce the cost of electrolytic H2 production processes. By operating the auxiliary circuit electrolyser in the opposite direction to the main circuit electrolyser, it is possible to use a non-precious anodic electrode materials and to reduce the energy requirement for the main electrolysis process. Hence, with this new electrolysis configuration, capital cost reductions may be achievable as the working electrodes can be made non-precious materials, while maintaining or reducing the operating cost by improving the performance of the H2 production process. The electrochemical promotion of catalysis and the electrochemical promotion of electrocatalysis are both emerging areas which offer a number of potential advantages for H2 production. The promotional effect obtained under triode and tetrode operation represents an alternative technological approach to other research efforts aiming to maximize the electrochemical surface area and develop new catalytic materials with improved properties. 157 Chapter 4: Catalysis and Electrochemical Promotion of Catalysis (EPOC) 4.1 Synopsis As seen in the prior Chapters, the catalyst carbon support has a high surface area and good electronic conductivity, but its corrosion limits the operating conditions of the electrochemical system under study. Various metal oxide catalyst supports were found to possess a higher oxidation resistance than C, and a better Pt adhesion [203]. Titanium dioxide (TiO2) is known for its high catalytic activity, good mechanical properties, stability in acidic solutions, non-toxic properties, and low cost. It can be used as a catalyst or support in electrocatalysis or catalysis. It can also be used as a photocatalyst under UV light in photocatalysis. Furthermore, in reducing atmospheres, TiO2 tends to loose oxygen and becomes an n- type semi-conducting material. To study the electrochemical promotion of catalysis (EPOC) phenomenon, the catalyst chosen not only needs to be a catalyst to the chemical reaction, but also needs to be an electrocatalyst to the electrochemical reaction used to promote the catalytic reaction. As discussed in Section 1.3.1 in Chapter 1, a high activity towards liquid CH3OH catalytic reforming at low temperature has been reported with Pt-Ru/TiO2 [16]. However, to the author’s knowledge, no low temperature CH3OH catalytic reforming studies have been conducted with Pt-Ru/C. The catalyst support may play a significant role in the activity of noble metals toward a certain catalytic reaction. Strong metal-support interactions (SMSI), which decrease the H2 and CO chemisorption capacity of the TiO2-supported metals, may exist when a TiO2 supported catalyst is reduced at a temperature between 200 and 500oC [204]. It was noted that SMSI interactions only occurred when H2 adsorption on the supported metal was dissociative and that they were not due to change in the metal particle sizes. An accepted explanation for this is that at temperatures greater than 300oC, the titania is partially reduced and a suboxide phase migrates onto the metal particles, blocking the metal surface partially covered by TiOx, which decreases the chemisorption capacity of the metal and suppresses structure-sensitive reactions [204]. At higher reduction temperatures, partial or complete encapsulation of the metal particles in titania may occur, and above 158 800oC, the formation of intermetallic compounds and alloys was reported [204]. These interactions may improve the performance of precious metal catalysts by surface improvements and/or limitation of poisoning effects. Typical electrocatalyst requirements include being electrically conductive, high surface area, high electrocatalytic activity, and low cost. Proton-conducting hydrophilic TiO2 was shown to improve the stability of PEMFC electrodes when present in the catalyst layer [61] or as the catalyst support [205]. Using Pt-TiO2/C was shown to result in a greater electrochemical active surface area and fuel cell performance than using Pt/C [206]. Although the TiO2 support is not sufficiently electronically conductive at low precious metal loadings (10-5 S/cm at <10 % Pt), it was shown that increasing the weight percent of precious metal on this support may increase the electronic conductivity of the material to levels sufficient to sustain an electrochemical reaction (1.26 S/cm at 60 % Pt) [207]. Also, the addition of Nb metal to the TiO2 support can increase its conductivity enough to make it suitable for electrochemistry. It was reported that the activity of Pt-Ru supported on Nb-TiO2 increased the electrochemical activity by 83 % compared to the same electrocatalysts supported on carbon Vulcan XC-72R and by 64 % compared to the same metal catalyst but supported on TiO2 only, for the electro-oxidation of CH3OH [207]. Nevertheless, as explained in Section 1.3.4.1, only low currents or potentials need to be applied to study the effect of EPOC. Therefore, the addition of Nb to the TiO2 support is not necessary to study the effect of electrochemical promotion. The preparation of most TiO2-supported metal catalyst starts by depositing precursor metal salts followed by a reduction of the metal ions. The catalytic performance is often affected by the average metal particle size and by the homogeneity of the distribution of the two metals. The metal particle size and structure is affected by both, the pre-reduction and reduction treatments [204]. Many techniques exist for the preparation of TiO2 supported bimetallic catalysts, such as impregnation [208], sputtering [209], sol- gel [210], and ion exchange [211]. In this study, an impregnation synthesis method was developed based on previous work available in the literature [14, 16, 29, 210, 212]. The ratio of Pt to Ru, the solution pH and composition, the heat treatment, the hydration level of the precursor salts, and the recrystallization conditions and pressure, are all factors which could influence the formation of the Pt-Ru/TiO2 catalyst. However, they were not investigated in this study. 159 The catalyst powder synthesized was characterized by Scanning Electron Microscopy with Energy dispersive X-ray Spectroscopy (SEM-EDX), Transmission Electron Microscopy (TEM), X-ray Diffraction (XRD), X-ray Photoelectron Spectroscopy (XPS), Inductively Coupled Plasma Mass Spectrometry (ICPMS), and Brunauer-Emmett-Teller (BET) surface area, before and after ball-milling. Once the targeted composition for the Pt-Ru/TiO2 was obtained, cyclic voltammetry was employed to determine the CH3OH oxidation activity of the Pt-Ru/TiO2 catalyst, and compare it to that of Pt and commercial Pt- Ru/C catalyst. In Chapter 2, it was shown that the theoretical thermodynamic efficiency and theoretical cell voltage of CH3OH electrochemical or chemical reforming become positive at a temperature of about 41oC. Catalytic CH3OH reforming experiments reported in the literature were carried out at temperatures ranging from 72 to 105oC with Pt-Ru/TiO2. Catalysis was first conducted using catalyst powders in a three port round-bottom glass cell for comparison with low temperature CH3OH reforming data available in the literature. It was the first time the low temperature catalytic reforming of CH3OH was conducted using a commercial fuel cell catalyst, i.e., Pt-Ru/C. Catalysis was also carried out on gas diffusion electrodes prepared using the method described in Chapter 2. The effect of EPOC was investigated in the galvanostatic mode, in order to find out if a non-Faradaic effect could enhance the catalytic activity and selectivity towards H2 production. If successful, the application of EPOC to the catalytic reforming of CH3OH has the potential of reducing the operating temperature and energy requirements, as well as shortening the start-up time of the catalytic CH3OH reformer. 160 4.2 Experimental 4.2.1 Materials The catalyst support used was a TiO2 nanopowder (P25, Degussa Co., 99.995%, average primary particle size of 20 nm, BET surface area of 50 ± 15 m2/g, 70% anatase, 30% rutile). The platinum and ruthenium precursor salts were chloroplatinic acid hexahydrate (H2PtCl6.xH2O, ACS reagent, 99.999%, ≥ 37.5% metal), and ruthenium (III) chloride hydrate, (RuCl3.xH2O, ReagentPlus®, 99.99%, 40 to 49% metal) both from Sigma Aldrich. In some experiments, ultra high purity (5.0) helium from Praxair was used as a carrier gas. All other materials used are the same as the ones described in prior Chapters. 4.2.2 Catalyst synthesis Initially, two methods were employed to synthesize the Pt-Ru/TiO2: an impregnation procedure followed by calcination and reduction in H2, and a chemical reduction procedure followed by calcination in Ar. Since the first procedure resulted in a more homogeneous distribution of precious metal nanoparticles, it was preferred over the chemical reduction method. Information on both procedures can be found in Section E.7 of Appendix E. However, only results obtained by the impregnation method are discussed. The first step of the impregnation is to suspend the TiO2 in an aqueous solution of precursor salts. A quantity of TiO2 was added to DI water and sonicated for an hour. The Ru metal precursor salt was then added to the aqueous TiO2 suspension, and the mixture was sonicated for another hour before the Pt metal precursor salt was added, followed by another hour of sonication. The slurry formed was then heated to 80oC in an oil bath under continuous stirring until all water was evaporated. After this, the slurry was evaporated to dryness in a vacuum oven in air at 110oC overnight. The hardened particles were then crushed using a pestle and mortar. The loading of Pt and Ru metal was adjusted to the desired wt% with a molar ratio of Pt/Ru equal to unity. For comparison with the catalytic CH3OH reforming literature, the targeted catalyst precious metal 161 loading was initially 10 wt% for the preliminary catalytic reforming tests carried out with Pt-Ru/TiO2 powder. However, difficulties were encountered for the cyclic voltammetry tests as the catalyst’s substrate electrical conductivity was too low. Therefore, in order to increase the catalyst conductivity, and for ease of comparison with the commercial Pt-Ru/C catalyst used in prior thesis Chapters, the catalyst precious metal loading was increased to 20 wt%. An example of the catalyst composition is located in Table E.4 in Appendix E. The catalyst synthesized was calcined and reduced in a H2 atmosphere in 20 ml alumina (99.6%) combustion boats (AdValue Technology, LLC) using a programmable Barnstead International tube furnace (model F21135, 120 V, 11.3 A, 1350 W). The tube furnace was first flushed by passing H2 and Ar for a minimum of 2 hours to remove air/O2. The gas flow rates were adjusted to obtain 10% H2 in Ar (10 sccm H2 and 90 sccm Ar). The temperature was ramped up at a rate of 5oC/min to a dwell temperature of 500oC for 2 hours. This time period was selected to avoid extensive particle growth. Following this, the furnace was cooled down to room temperature at a rate of 5oC/min. The total time for the reduction was about 7 hours and 30 minutes. Although the calcination temperature has an influence on the structure, morphology, crystallinity and particle size [213], the effect of the heat treatment on the microstructure of the bi-metallic material catalyst particles was not investigated. At calcination temperatures closer to 600oC, the TiO2 support conductivity would increase, however its SA would decrease, as the transition from the anatase phase to the rutile phase starts to occur near 550oC. The reduced and calcined catalyst powder was ball milled for an hour and 30 min. at a time, using a 0.25” OD tungsten carbide milling ball (Spex 8004A) in a ball milling barrel inserted in a ball mill (Spex Sample Prep 8000-D). For the catalysis and EPOC experiments, inks of the Pt-Ru/TiO2 and Pt-Ru/C catalyst powder were sprayed on TGPH-60 using the same composition and technique as described in Chapter 2. The catalyst loading was 4 mg/cm2 of Pt-Ru/TiO2 or Pt-Ru/C at the electrolyser anode, and 2 mg/cm2 of Pt/C at the electrolyser cathode. The same ink composition and catalyst loading resulted in thicker electrodes for the Pt-Ru/TiO2 catalyst (0.34 mm), than for the Pt-Ru/C catalyst (0.23 mm). 162 4.2.3 Pt-Ru/TiO2 physicochemical characterization Specimens of the different Pt-Ru/TiO2 powders produced were physicochemically characterized using various techniques: SEM, EDX or EDS, XPS or ESCA, TEM, XRD and BET. In some cases, commercial 20 wt% Pt-Ru/C, Vulcan X72-R carbon, and TiO2 were used as standards for comparison with the Pt- Ru/TiO2 catalyst powder specimens. The characterization techniques employed are briefly described in this section. 4.2.3.1 Scanning electron microscopy Scanning Electron Microscopy (SEM) images were taken using an Hitachi model S-3000N system to obtain information about the catalyst’s surface morphology. Most precipitate crystallographic morphologies were observed at low scanning speed in the low vacuum mode. The vacuum pressure was 20 Pa and the accelerating voltage was either 20, 80 or 100 kV. The high resolution surface images, which are produced by scanning the sample with a high energy beam of electrons, reveals information about the surface topography, composition and electrical conductivity. Aluminum alloy mounts (SPI Supplies) covered with a double sided carbon tape coupon was used for sample mounting. 4.2.3.2 Energy-dispersive X-ray spectroscopy In addition to the secondary electron (SE) and the back-scattering electron (BSE) detectors of the Hitachi model S-3000N system, an energy-dispersive X-ray spectroscopy (EDX or EDS) detector was also employed. The SEM EDX capability permits the detection of light elements having atomic numbers greater than 4. The visualization of the material surface with this technique generates a spectrum which allows one to obtain information on the elemental analysis or chemical characterization of the specimen. 163 4.2.3.3 Transmission-electron microscopy Transmission-electron microscopy (TEM) was employed to estimate the average particle size and visualize the shape of the Pt and Ru deposits. In this type of microscopy, the electron beam interacts with the sample as it passes through it. The Hitachi H800 TEM used was operated at an acceleration voltage of 200 kV. A suspension of 20 mg of catalyst powder in 5 ml of DI water was prepared. After sonication, about 2 μl of the suspension was extracted and deposited on polymer plated Cu grids which was then dried under a lamp. 4.2.3.4 X-ray diffraction X-ray diffraction (XRD) measurements were taken with a Bruker AEX D8 advanced powder X-ray diffractometer using Cu radiation to obtain information about the catalyst’s structure, composition and particle size. The diffraction patterns were recorded at 40 kV and 40 mA in the angular range of θ=5-90o, with a continuous scan speed of 1.2 to 2.8°/min. and a sampling width of 0.02 or 0.04o. The intensities observed were corrected for background noise, and peaks were determined using the EVA 10.0 (2004) Differac Plus basic evaluation package which has access to the Cambridge Structural Database and the Inorganic Crystal Structure Database. A qualitative XRD analysis required about 1 g of catalyst powder evenly spread out in a thin layer over a 1x1 cm2 surface. This technique permits the identification of phases in a sample by comparing them with standard patterns available in the literature. It is also used to estimate the relative proportions of the different phases present in a multi-phase material sample by comparing the peak intensities associated with the phases identified. A quantitative characterization permits the determination of the quantity of the different phases in a multi-phase sample, but requires about 4 to 5 g of catalyst powder (2 g for a semi-quantitative test), and was not carried out for this study because the quantities of catalyst prepared were too small. 164 4.2.3.5 X-ray photoelectron spectrometry X-ray photoelectron spectrometry (XPS), also referred to as electron spectroscopy for chemical analysis (ESCA), was used to analyze the material’s elemental surface chemistry for atomic numbers of 3 and greater. Measurements were obtained with a Leybold model MAX 200 to determine the surface atomic ratio of the synthesized Pt-Ru/TiO2. 4.2.3.6 Inductively coupled plasma optical emission spectrometry Inductively coupled plasma (ICP) optical emission spectrometry (OES) or atomic emission spectroscopy (AES) is a sensitive and robust technique for the determination of trace elements in a solution. ICP-OES analysis was carried out by Exova Canada Inc. using a Thermo Fisher IRIS Intrepid II. Two different analyses were conducted and required different liquid sample preparation procedures. The total metal content was determined by alkaline fusion for oxide dissolution, while the physical, and the aggregate properties were determined by acid digestion. The liquid samples were sprayed into an ICP source obtained by inductively heating a flow of Ar gas to about 10000oC. Under the plasma conditions, the sample’s elements become ionized and excited, and their electrons emit electromagnetic radiation at characteristic wavelengths. The light emission intensity is measured by OES and is indicative of a particular element’s concentration in the sample. A minimum of 0.4 g of sample was required to perform the two analysis. 4.2.3.7 Brunauer, Emmett and Teller surface area, and Barrett-Joyner-Halenda pore size and volume analysis Brunauer, Emmett and Teller (BET) developed a theory pertaining to the physical adsorption of gaseous molecules on a solid surface and which allowed for the measurement of the specific surface area of a material [214]. Some characterization tests were carried out with a single point Micromeritics Flow Sorb II (model 2300FC) with Ar at liquid N2 temperatures. Before each surface area measurement, the apparatus was calibrated twice by injecting a known volume of N2 gas. The catalyst sample to be characterized was degassed and dried at 250oC overnight. A multi-point Belkman Coulter surface area 165 analyzer (model SA3100) was also used to obtain the Pt-Ru/TiO2 nanopowder BET surface area, as well as to perform a Barrett-Joyner-Halenda (BJH) analysis to obtain surface porosity and particle size information. In this case, degassing was performed at 120oC for 15 minutes. About 0.5 to 1 g of catalyst powder was sufficient to carry out the tests and obtain reliable results. 4.2.4 Pt-Ru/TiO2 electrochemical characterization Cyclic Voltammetry (CVs) were used to evaluate electrochemical activity of the noble metal oxide catalysts synthesized. A five-port single compartment glass cell was used to conduct CVs with a Pine Instrument Company bi-potentiostat (model AFCBP1). The CVs were corrected for resistance with measurements obtained using an impedance meter (LCR-821, GW Instek) at a frequency of 200 kHz. Experiments were conducted in 0, 2 and 16 M CH3OH in 0.5 M H2SO4 at 23 ±2 oC, 50 and 75 ±1oC. The glass cell was submerged in an oil bath on a magnetic heating plate. The reference electrodes and Pt flag counter electrode were the same as the ones used in Chapter 2 and 3. All potential scans are reported vs. SHE. The working electrodes tested were a Pt disk electrode (AFE1XFG030PTR, 3 mm disk OD, Pine Research Instrumentation), and Pt-Ru/C or Pt-Ru/TiO2 inks deposited on a glassy carbon disk electrode (GC, AFE1XFP030GCR, 3 mm disk OD, Pine Research Instrumentation) which was not rotated. A quantity of catalyst powder was soaked by 0.5 ml of DI water first, then 9.5 ml of IPA was added as the solvent. After 30 min. of sonication, a quantity of 5 wt% Nafion® solution was added, followed by another 30 min. of sonication, to ensure a uniform dispersion. Using a micropipette, an aliquot of the dispersion was drop coated onto a clean and polished GC disk electrode, which was dried in air for solvent evaporation. Preliminary ink composition optimization work was required as the Pt-Ru/TiO2 catalyst particles had a tendency to come off the disk electrode. In an attempt to overcome this, the mass of catalyst in the ink was varied between 10 and 40 mg, different Nafion® contents ranging between 5 to 40 wt% were tested, and the volume of catalyst ink deposited on the disk electrode surface was in the range of 2 to 10 μl. The final ink composition used contained 40 mg of catalyst with 20 wt% Nafion® for Pt-Ru/C and 5 wt% Nafion for Pt-Ru/TiO2, and 2 μl was sufficient to cover the GC surface uniformly, resulting in a thin film 166 loading of 28.5 mg/cm2. More details on the ink preparation procedure can be found in Section E.6 of Appendix E. The solution in the glass cell was degassed with N2 for a few minutes and the working electrode surface was conditioned prior to carrying out the CVs by running ten scans at 100 mV/s between 0.2 and 1.2 V vs. SHE. As no significant differences in the CVs were noted for the Pt-Ru/TiO2 tests conducted in ambient light vs. dark conditions, all remaining tests were conducted in ambient light. 4.2.5 Apparatus 4.2.5.1 Powder catalysis Preliminary heterogeneous catalytic CH3OH reforming tests were conducted in the closed semi-batch reactor system shown in Fig. 4.1. 167 Figure 4.1: Schematic diagram of the experimental set-up used for the catalytic reforming of liquid CH3OH using catalyst powders. A three-port single compartment 50 ml round bottom borosilicate glass flask (14/20 connections, Kontes, batan-ware) acted as the reaction vessel. One side-port was occupied by a Viton septum enclosing a thermometer; the other side-port was used for carrier gas injection using a glass frit gas sparger, while the center port was connected to a double walled condenser (14/20 connections). Vacuum grease (Apiezon AP-101) was spread on all glass joint connections to ensure gas tightness. The solution and the catalyst powder were both added at ambient temperature before the carrier gas flow was turned on, and the total liquid volume of each test was kept the same. The condensing water temperature was kept at 2oC ± 1oC using an ice water circulation bath (Haake DC-30). The condenser gas exit was connected to two gas bubblers containing DI water at room temperature, which allowed for the condensation of H2O and CH3OH entrained vapours in the He flow. Before entering the MS capillary, the gas also went through a desiccant to remove any entrained H2O vapours and a line heater, maintained at 110oC. A 2 M CH3OH 0.5 M H2SO4 He Gas to measurement Circulating Ice H2O Reflux Condenser Stirring and heating mantle 50 ml round bttm flask Closed gas circulation system Thermometer Frit Gas Bubbler 1 H2O Pt-Ru/C or Pt-Ru/TiO2 Gas Bubbler 2 H2O Dessicant Line Heater Gas to fumehood 168 stirring and heating mantle (EMA0050/CEBX1, Thermoscientific) was used to uniformly stir the reacting solution and maintain the single-compartment glass cell reactor at the desired reaction temperature. Preliminary powder catalysis tests were carried out to confirm the validity of the methodology employed. For the preparations tests, the volume of gas generated was monitored by using an inverted burette, located on the reflux condenser exit line after one gas bubbler, by water displacement, and in some cases, N2 was used as a carrier gas. The catalyst powder sample and/or the CH3OH solution were either added separately or together at room temperature or at the reacting temperature. It was assumed that the relative quantity of H2 and CO2 gas generated was stoichiometric. As mentioned in Chapter 2, the boiling point of a CH3OH equimolar solution is 76.7oC. To verify the CH3OH evaporation rate during the course of the powder catalysis experiment, tests were carried out under the same experimental conditions as for the powder catalysis experiments, but without any catalyst. At the end of the experiment, an aliquot of the reacting solution was analyzed for CH3OH concentration using the K2Cr2O7 - spectrophotometric method described in Chapter 2, 3 and Appendix E. It was confirmed that, using a cold water condenser, the CH3OH concentration of the reacting solution did not change by less than 5% over an hour. It confirmed that most of the CH3OH vapours were effectively condensed by the condenser. Preliminary tests were also carried out to confirm that no H2SO4 would be entrained in the gas to the MS capillary. During these tests, the pH of the first gas bubbler water was monitored over time using a calibrated pH meter and electrode (Oakton Ph 110 series, ±0.01 pH, 35811-98), while passing a N2 carrier gas at a rate of 150-175 bubbles/min. under powder catalytic reforming conditions similar to the ones used in the final powder catalysis experiments. Experiments involved one hour of N2 purging, followed by a 30 min. ramp-up to 75oC, where conditions were held for 5 hours. In the first experiment, the pH of 40 ml of DI water in a single bubbler increased by 1.8% from 4.98 to 5.07 over the course of the experiment. In the second experiment, two 200 ml DI water bubblers were used. The pH of the second bubbler decreased by 3% from 5.93 to 5.75 during the experiment. The pH fluctuations measured fall within experimental error, and indicate that entrained acid should not cause issues to the MS and its capillary. These preliminary tests confirmed the validity of the experimental method. 169 For all final experiments, the exit gas composition was continuously monitored using a quadrupole mass spectrometer (MS, RGA-200, Jerome & Francis Co. Ltd.) which was controlled using RGA 3 software. In these cases, the carrier gas was He, which was controlled with a mass flow controller (model 5878, Brooks Instrument Division, Emerson Electric Co) to generate a flow rate of 170 ml/min. at standard temperature and pressure (STP) to continuously remove the produced gases and to reduce the lag time as much as possible. Before entering the MS capillary, the gas stream was split into two flows: (i) a portion of the gas was sent to analysis to the mass spectrometer capillary, and (ii) the other portion of the gas was vented to a fumehood. This is because the total gas flow can not be directed into the MS due to its vacuum limit. Therefore, only a small portion of the exit total flow goes to the MS and the rest is vented off to a fumehood. Although the partial pressures of CH3OH, H2O, CO2, H2, CO, CH4 and air were monitored, the MS was only calibrated for various partial pressures of H2 relative to the total flow, making it possible to determine the flow rate of H2 but not of the other gases monitored. Since the H2 gas is quantified using a calibration curve obtained from the partial pressures related to the total flow, the portion of the total flow which goes to the MS is not used in the calculations. Knowing the He flow rate and the partial pressures of He and H2, it was possible to determine the production rate of H2 using the following expression: 22 , , Hp Hep H He P P Cal Fr Fr ⋅= (4.1) Where Fr is the molar flow rate in mmol/min., Cal is a dimensionless calibration number, and Pp is the gas partial pressure in torr. The MS has a partial pressure precision of 5x10-14 torr. The calibration number comprises a factor for the conversion of the He volumetric flow rate, Vr, to a molar flow rate, Fr, using the ideal gas law at STP, as well as a dimensionless calibration factor, calf, to convert the ratio of the partial pressures to a molar ratio. The calibration factor, calf, was determined experimentally to be 10.714 from the slope of the calibration curve as per the following equation: He H f HeP HP Y Y cal P P 22 , , ⋅= (4.2) 170 Knowing calf and Cal, it is possible to determine the generation rate of H2 using Eq. 4.1, as the He molar flow rate is known and the respective gas Pp are measured by the MS. The hydrogen production rate calculated from Eq. 4.1 is equivalent to the unpromoted or promoted catalytic reaction rate, r or ro in mol/s in Eq. 1.33, 1.36 and 1.37. The solution and catalyst were introduced into the reaction vessel at ambient conditions. Prior to starting an experiment, the air was flushed out of the system and the solution de-aerated by flowing He gas for 5 min. under stirring. The filament was then turned on and the MS valve was opened. It usually took about half an hour for the air partial pressure to drop below 1x10-7 torr. Once this low air level was reached, the experiment was started by heating the reaction vessel under a He atmosphere until the desired reaction temperature was reached. Once reflux started, it was checked that no droplets were forming in the upper part of the condenser, but only in the lower part. Even though the lines between the gas sampling point and the mass spectrometer injection point were as short as possible (about 125 cm), there was a delay of about 2 min. between the an electrochemical current was imposed at the MEA and a response was seen in the MS measurement. The MS measurements were taken every 10 s, and the data collected was corrected to take the retention time into account. 171 4.2.5.2 GDE catalysis and electrochemical promotion of catalysis The experimental set-up used for the GDE catalysis and EP of catalysis tests is similar to the one used for the powder catalytic reforming tests, except that as shown in Fig. 4.2 and 4.3, the double-wall 2-compartment glass cell used in Chapter 3, was employed as the reactor instead of a single compartment round bottom flask. Figure 4.2: Schematic diagram of the experimental set-up used for the catalytic CH3OH reforming and the electrochemical promotion of catalytic CH3OH reforming using GDEs. Heating Water Bath and Pump Multistat PC TI TI WE CEGas to Fumehood Gas toatm Cooling Water Bath and Pump N2 Mass Spectrometer 172 Figure 4.3: Picture of the experimental set-up used for the catalytic CH3OH reforming and the electrochemical promotion of catalytic CH3OH reforming using GDEs (a) auxiliary equipment (b) glass cell configuration. Gas to fumehood Hot water T controlled circulation bath Temperature indicator Thermocouples Ice water circulation bath Condenser He flow controller Capillary to MS Double-wall two compartment glass cell and holder He sparger PC for MS control Gas to bubblers, desiccant and line heater Gas to capillary He and produced Gas flow (a) Thermocouples He sparger Magnetic stirrers Condenser Hot water circulation MEA holder He flow controller Anodic compartment(b) 173 A sparger permitted the distribution of He bubbles in the anolyte. The cap port of the anodic glass cell compartment which is closest to the MEA was connected to the condenser. The cathodic glass cell compartment gas was vented to the atmosphere. Both glass cell compartments had ports for thermocouples. A water circulation bath (Haake DC-10) was provided to heat and maintain the two- compartment glass cell at 75oC ± 0.5oC, while an ice water circulation bath (Haake DC-30) cooled the condenser to 2oC ± 1oC. The MEA composition and assembly was similar to the one used in Chapter 2. The WE GDE was 4 mg/cm2 Pt-Ru/C or Pt-Ru/TiO2 and the CE GDE was 2 mg/cm2 Pt/C. 4.3 Results and Discussion 4.3.1 Pt-Ru/TiO2 physicochemical characterization 4.3.1.1 SEM images SEM images permitted the visual observation of the synthesized Pt-Ru/TiO2 particles shape, structure, composition and surface morphology. Figure 4.4 shows an image obtained for the 20 wt% metal Pt- Ru/TiO2 catalyst. Figure 4.4: SEM image of 20 wt% metal Pt-Ru/TiO2 (WD 15.2 mm, 20.0 kV, x2.0 k). 20 μm 174 The Pt-Ru/TiO2 catalyst powder is composed of scattered clusters or agglomerations constituted of small individual particles. It seems like the catalyst particle’s surface had a low porosity. The Pt and Ru precious metal particles are spread out at the surface, and are represented by conductive white spots. However, the SEM images do not allow to distinguish between the Pt and Ru particles. 4.3.1.2 EDX measurements The EDX spectrum gave information about the elemental composition of the catalyst samples. An example of a typical EDX spectra conducted on a 10 wt% metal Pt-Ru/TiO2 catalyst sample and a 20 wt% metal Pt-Ru/TiO2 catalyst sample are presented in Fig. 4.5. The spectra exhibited peaks at 2 keV and 2.5 keV, which were assigned to Pt and Ru respectively. There were two peaks assigned to Ti at 4.5 and 5 keV. These binding energies are in close agreement with the ones reported in the literature for similar catalysts [215]. Hence, the EDX measurements confirmed the presence of Pt, Ru and Ti. Figure 4.5: EDX spectra for Pt-Ru/TiO2 (a) 10 wt% metal, (b) 20 wt% metal. For the 10 wt% metal spectrum shown in Fig. 4.5 (a), the Ru content was 3 wt% and the Pt content was 4 wt% which translates into 7 wt% metal. This is about 30 % lower than the targeted metal loading of 10 wt%. Various EDX measurements were taken on 10 wt% metal Pt-Ru/TiO2 and in average the Pt and Ru metal content was about 8 wt%, i.e., 4-6 wt% of Pt, and 2-3 wt% of Ru. In general, the Ru wt% was In te ns ity (C ou nt s) Binding Energy (keV) Binding Energy (keV) In te ns ity (C ou nt s) (a) (b) In te ns ity (C ou nt s) In te ns ity (C ou nt s) 175 about 47% lower than the Pt wt%. It is possible that the Ru precursor salts impregnation is more difficult than the Pt precursor salt impregnation, or that the Ru particles attached to the substrate more easily detach than the Pt ones. In order to compensate for this, an excess of Ru precursor salt was used to synthesize the 20 wt% metal Pt-Ru/TiO2 catalyst samples. For the 20 wt% metal spectrum shown in Fig. 4.5 (b), the Ru content was 5 wt% and the Pt content was 12 wt%, according to which the total metal content was about 15% lower than the targeted loading of 20 wt%. For the commercial 20 wt% metal Pt-Ru/C, which has a Pt:Ru 1:1 a/o, EDX results on an atomic % basis were 0.34 for Pt and 0.29 for Ru, which resulted in a Pt to Ru atomic ratio of about 1.2 instead of 1. The accuracy of the EDX spectrum may have been reduced by raising the over-voltage on the SEM, overlapping peaks, and the nature of the sample. For inhomogeneous and rough samples, the results tend to be less precise. Therefore, it appears that this material characterization technique was not sufficiently accurate to precisely determine the metal content of the catalyst sample. 4.3.1.3 TEM images TEM images were taken to allow for the estimation of the nanoparticles size and the observation of the metal catalyst dispersion over the TiO2 substrate surface. Two Pt-Ru/TiO2 TEM images of different magnifications are shown in Fig. 4.6. 176 Figure 4.6: TEM images of 20 wt% Pt-Ru/TiO2 (a) direct mag. 200000x and (b) direct mag. 600000x (20 wt% Pt-Ru, H2 reduction method, ball milled 60 min., HV = 100.0 kV). The Pt-Ru/TiO2 catalyst formed multilayer agglomerations of small irregular flakes. The variation in contrast seen on the TEM images result from overlapping of the TiO2 particles and their different thicknesses. High agglomerations have also been obtained for Ru/TiO2 formed using a similar wetness impregnation technique [216]. The TiO2 catalyst support was homogeneously decorated with Pt and Ru nanoparticles. The TiO2 flakes were about 150 nm and larger, which is at least 7.5 times larger than the original average primary particle size of 20 nm. Small particle sizes can cause melting point depression compared to bulk materials, therefore, the TiO2 nanoparticles may depress the melting point in such a way that the TiO2 nanoparticles melted together during the H2 reduction step. However, since the melting point of TiO2 is 1640oC, a depression of the melting point below 500oC (the temperature used for the H2 reduction step) is unlikely. Therefore, the adherence of the TiO2 particles most likely results from sintering during the H2 reduction process. From the visual observation of various TEM images, the metal nanoparticles size seems to range from 5 to 10 nm, which falls into the reported range size for metallic nanoparticles for other similar materials (3 to 10 nm) [215, 217], but which is slightly greater than the average particle size reported for 20 wt% Pt/C (E-TEK) which is 2.6 nm [218]. Usually, smaller metal particles result in larger exposed surface area, (a) (b) 100 nm 20 nm 177 and in a more uniform distribution on the substrate. For photocatalysis, it would be desirable to have the TiO2 surface exposed with smaller metal particles. However, for the electrochemical oxidation of CH3OH, it may be preferable to have a high portion of the TiO2 surface area covered by the metals particles in order to enhance the electrocatalyst’s conductivity. Even if the 20 wt% Pt and Ru metal particles seem to be homogeneously distributed, it does not look like they will have sufficient connectivity to permit the conduction of current with a non-conductive support. The nanoparticles size of the metal will affect CH3OH adsorption and the activity of the catalyst towards CH3OH oxidation as the reaction is structure sensitive. 4.3.1.4 XRD measurements While electron microscope techniques allowed the examination of the Pt-Ru/TiO2 particles, XRD allows the identification of the Pt-Ru/TiO2 crystal structure and the proportion of the different crystal phases. In addition to identifying the microstructure of the catalyst synthesized, the average particle size can also be calculated. Figure 4.7 shows representative diffraction patterns for commercial Vulcan X72-R carbon, Pt-Ru/C, TiO2 powders and the 20 wt% metal Pt-Ru/TiO2 prepared by the H2 reduction method after an hour of ball milling. Figure 4.7: X-ray diffraction data for (a) Carbon and 20 wt% Pt-Ru/C, and (b) TiO2 and 20 wt% Pt- Ru/TiO2. 0 100 200 300 400 500 600 700 5 10 15 20 25 30 35 40 45 50 55 60 65 70 2 - Theta In te ns ity (c ps ) 0 100 200 300 400 500 600 700 800 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 2 - Theta In te ns ity (c ps ) (a) (b) Carbon, Vulcan X-72-R Pt-Ru/C Pt-Ru/TiO2 TiO2, P25An at as e R ut ile Carbon Carbon Pt, Ru Pt, Ru In te ns ity (c ps ) In te ns ity (c ps ) In te ns ity (c ps ) In te ns ity (c ps ) In te ns ity (c ps ) In te ns ity (c ps ) A na ta se R ut ile 178 The carbon peak was located at the same 2θ value of 25 for the carbon sample and the Pt-Ru/C sample. The Pt and Ru peaks located at a 2θ of about 40 were present in the Pt-Ru/C sample and in the Pt- Ru/TiO2 sample. The Pt and Ru peaks slightly overlap making them more difficult to distinguish. The peak position and intensity also indicate that the crystalline structure of the anatase and rutile form were both present in the Pt-Ru/TiO2 sample, and that the crystalline anatase phase was predominant (~60 %), although its content was lower than in the original TiO2 P25 sample. The original anatase phase of the TiO2 P25 sample was determined to be about 79%, which is almost 10% greater than the manufacturer’s specification. The sintering temperature was low enough to retain much of the anatase structure of the original TiO2 powder. The intensity of the rutile phase did not change considerably between the Pt- Ru/TiO2 sample and the original TiO2 P25 sample. In catalysis, the anatase phase is mostly used as there is a predominant molecular adsorption of alcohols and water on the anatase surface and a predominant dissociative adsorption on the rutile surface [2]. The anatase crystal structure is also known to be more active for electrochemical oxidation reactions than the rutile crystal structure [207]. Scherrer’s equation for nanoparticles, for grains less than 0.1 μm, correlates the size of particles to the broadening of a peak in the diffraction pattern. It defines τ as the mean size of the crystalline domains in nm, which may be smaller or equal to the grain size: BoWFWHM Wk θτ cos . 2 ⋅− = (4.3) Where k is the shape factor, a dimensionless constant which depends on the crystal size and is assumed to be 0.89 for spherical crystals with a cubic crystallographic structure, W is the X-ray wavelength, which was 0.154 nm, FWHM is the line broadening at the maximum intensity, i.e., full width half max in radians, and θB is the Bragg angle in radians. The Scherrer equation typically leads to results within an accuracy of 20 to 30 %. Correcting the FWHM value for the instrumentation broadening, Wo, which was of 0.05o 2θ did not increase the τ values by more than 5%. Table 4.1 lists selected peak 2θ 179 values, corresponding FWHM values, and average size of the crystallites based on Scherrer’s equation for the XRD patterns shown in Fig. 4.3. Table 4.1: Scherrer’s crystallite size for selected peaks from the carbon, Pt-Ru/C, TiO2 and Pt-Ru/TiO2 XRD patterns. XRD pattern Peak ID 2θ [o] ± 0.06 FWHM [o] τ [nm] VulcanX-72-R carbon 25 4.152 19 20 wt% Pt-Ru/C carbon 25 3.419 23 20 wt% Pt-Ru/C Pt 40.5 2.395 32 20 wt% Pt-Ru/C Ru 41.3 2.440 32 P25 TiO2 TiO2 anatase 25.25 0.364 216 P25 TiO2 TiO2 rutile 27.5 0.208 392 20 wt% Pt-Ru/TiO2 TiO2 anatase 25.4 0.496 158 20 wt% Pt-Ru/TiO2 TiO2 rutile 27.6 0.278 287 20 wt% Pt-Ru/TiO2 Pt3Ti 40.4 0.763 102 20 wt% Pt-Ru/TiO2 RuO2 54.4 0.563 139 As can be seen in Table 4.1, smaller crystals results in broader XRD peaks. The crystallites of the commercial Pt-Ru/C powder were an order of magnitude smaller than the ones of the synthesized Pt- Ru/TiO2 powder. The carbon crystallite thickness of the Vulcan X-72 sample was 37.77% smaller than the average particle size reported from the manufacturer, which is 30 nm. However, it was similar to that obtained for the carbon crystallites in the commercial Pt-Ru/C sample. The Pt and Ru crystallites of the commercial Pt-Ru/C sample had about the same size (32 nm). The P25 TiO2 support anatase and rutile crystals were about 10 to 20 times larger than the Vulcan X-72-R carbon support ones. As expected, the anatase TiO2 had larger crystals than the rutile TiO2 for both, the commercial P25 TiO2 and the synthesized Pt-Ru/TiO2. The commercial TiO2 P25 anatase and rutile crystallites were about 25 to 27 % larger than the ones in the synthesized Pt-Ru/TiO2 sample. Thus, it seems like the TiO2 crystallites did not grow upon the H2 reduction at 500oC. It seems like the TiO2 anatase crystallite size estimated in 180 Table 4.1 (158 nm) corresponded to the TiO2 particle size observed on the TEM picture (about 150 nm and greater) shown in Fig. 4.6. However, the Pt and Ru crystallite size estimated via Scherrer’s equation for the Pt-Ru/C catalyst (32 nm) is at least two times greater than the Pt and Ru particle size observed for Pt-Ru/TiO2 in Fig. 4.6 (5-10 nm). Therefore, it is possible that the agglomeration of crystals resulted in broader XRD peaks than expected. In the Pt-Ru/TiO2 sample, the precious metal particles and the titanium may form intermetallic compounds. For example, the formation of TiPt3 is thermodynamically favourable at 500oC [217]: TiO2 + 2H2 + 3Pt → TiPt3 + 2H2O (4.4) As most of the Ru present was in the oxide form, it seems that the chosen reduction temperature was unfavourable to the segregation of Ru or its alloying with TiO2. 4.3.1.5 XPS measurements XPS measurements are used to generate the valence-band spectra at the material’s conduction band. This gives information about the elemental composition of the surface, the empirical formula of pure materials, and the chemical and electronic states of elements in the surface. Elements associated with the XPS peaks can be identified with binding energies available in handbooks or databases. The XPS surface quantitative measurements (± 5%) along with a chemical identification of the main peaks is presented in Fig. 4.8. 181 Figure 4.8: Survey XPS spectrum of 20 wt% Pt-Ru/TiO2. The main elements identified were Ti, O, C, Pt and Ru. The C peak, resulting from CO from the air, interfered with the Ru peak, which rendered the identification of the Ru peak difficult. From the literature, the Ru peak is at 280 eV, RuO2 is at 281 eV and RuO3 is at 282.5 eV. The Pt-Ru/TiO2 spectra yielded a Ru peak at 281.28 eV, indicating that most Ru is present in the oxide form. From the literature, the Pt peak is at about 71 eV, while the ones for Pt oxides vary between 72 and 75 eV. The Pt-Ru/TiO2 spectra yielded Pt peaks at 71.5 and 75 eV, indicating that Pt is present in the metallic and oxide form. Based on the XPS surface analysis, which gives the % distribution amongst the different compounds detected at the sample surface, the precious metal loading was 16 wt%, which is lower than the targeted metal loading of 20 wt%. However, it was confirmed that the atomic ratio of Pt/Ru was 0.95, which is very close to the targeted atomic ratio of 1. 4.3.1.6 ICP-OES measurements Among the different methods taken to determine the metal content of the catalyst powder prepared, ICP- OES was the most accurate one (± 0.01 wt%). This quantitative technique was used for the 020040060080010001200 0 50000 100000 150000 200000 250000 300000 C P S Binding energy (eV) C C Ti Ti Ti Pt Ti O Ru O C P S 182 determination of the catalyst Pt, Ru and Ti metal content, as well as the PtOx, RuOx and TiO2 content. The ICP-OES analysis results were converted to weight percentage in terms of total Pt, Ru and Ti content for ease of comparison. Selected results are shown in Table 4.2. Table 4.2: ICP-OES results for H2 reduced Pt-Ru/TiO2 before and after ball milling and Pt-Ru/C. Sample description Excess Pt salt [%] Excess Ru salt [%] Ball milling time [min.] Total Pt [wt%] Total Ru [wt%] 20 wt% Pt-Ru/TiO2 30 60 0 13.77 13.37 20 wt % Pt-Ru/TiO2 30 60 30 11.68 10.77 20 wt % Pt-Ru/TiO2 15 50 60 10.85 8.12 20 wt% Pt-Ru/C - - 0 8.45 8.79 An excess of precursor salt was needed to obtain the desired metal loading. It is possible that some experimental error is caused by a loss of the precursor salt’s crystallization water due to evaporation, or that the Ru salt did not dissolve as well as the Pt salt, and for this reason, the excess if Pt was less than the excess of Ru. Ball milling gradually resulted in a larger partial loss of metal loading from the support. The composition of the Pt-Ru/TiO2 sample which was ball milled for 60 min. resulted in a metal loading 5% within that of the desired metal loading target (20 wt%). The commercial Pt-Ru/C catalyst ICP-OES total metal loading was 17.3% lower than the supplier quoted 20 wt%. Based on these results, the composition of the Pt-Ru/TiO2 catalyst prepared by the H2 reduction method was judged to be adequate for the purposes of this study. 4.3.1.7 BET surface area and BJH pore size and volume analysis The surface area (SA) is a key factor in catalysis and electrocatalysis as it has a strong effect on the overall rate of reaction. Therefore, it is desired to obtain a catalyst with a high specific SA. Also, the desorption area resulting from the N2 evaporation from the sample pores was always slightly greater than the prior adsorption area resulting from the liquefaction of N2 in the sample pores. Three measurements 183 taken on a same sample resulted in an error of 3%. The BET SA experimental results are listed in Table 4.3 along with the TiO2 and Pt-Ru/C BET SA values given in the manufacturer’s certificate of analysis. Table 4.3: BET surface areas of the Pt-Ru/TiO2 materials synthesized, as well as commercial TiO2 and Pt-Ru/C materials. Sample description BET SA [m2/g] Pt-Ru/TiO2 (NaBH4 reduction, 10 wt%) 30.23 Pt-Ru/TiO2 (H2 reduction, after 60 min. ball milling, 20 wt%, 1 point method) 19.69 Pt-Ru/TiO2 (H2 reduction, after 60 min. ball milling, 20 wt%, multi-point method) 15.74 P 25 TiO2 50.00* carbon VulcanX-72-R 240.00* 20 wt% Pt-Ru/C 130.00* * From certificate of analysis As expected, increasing the metal loading decreased the catalyst’s SA. Also, the one point method resulted in a surface area 20% greater than the multi-point method. This may be because a longer sample degassing period and higher drying temperature were used in the one point method. On average, after an hour of ball milling, the 20 wt% Pt-Ru/TiO2 synthesized by the H2 reduction method had an average SA about 60% lower than that of the commercial TiO2 and about 86% lower than the commercial Pt-Ru/C catalyst. This reduction in SA may partly be attributed to incorporation of metal particles in the pores of the TiO2 support and the fact that the carbon support has a SA 4.8 times greater than the TiO2 support. The BJH pore size and pore volume is a widely accepted model, which characterizes the pore size distribution independently from external area due to the particle size of the sample. The multi-point BET method allowed for the measurement of the porosity. Micropores have a diameter less than 2 nm, 184 mesopores, a diameter between 2 and 50 nm, and macropores, a diameter greater than 50 nm. Figure 4.9 shows the BJH desorption pore size distribution for the 20 wt% Pt-Ru/TiO2 nanopowder. Figure 4.9: Pt-Ru/TiO2 BJH desorption pore volume distribution. From the diagram, it can be seen that the average pore volume of the Pt-Ru/TiO2 sample with the greatest cumulative volume is 44 nm, indicating that the prominent pores are mesopores. From the TEM images shown in Fig. 4.6, it was determined that the surface metal nanoparticles size ranged between 5 to 10 nm. Thus, it may be possible that smaller metal nanoparticles were inside the TiO2 pores, reducing the surface area. The catalyst activity towards the species involved in the electrochemical or chemical reaction will be greater for a material which has a high porosity. 4.3.2 Pt-Ru/TiO2 electrochemical characterization The electrochemical response was obtained for a Pt RDE and Pt-Ru/C or Pt-Ru/TiO2 thin films coated on a GC RDE in 0, 2 or 16 M CH3OH acidic solutions. Cyclic voltammetry was carried out in the absence of O2 at ambient temperature, 50 and 75oC over the potential range of 0.2 to 1.2 V vs. SHE, at a scan rate 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 A ve ra ge v ol um e (m l/g ) 148 114 77.7 44 27 19.5 14.8 11.6 9.54 8.08 6.99 6.14 4.89 4.2 3.48 Average diameter (nm) A ve ra ge v ol um e (m l/g ) 185 of 50 mV/s, without rotation. From the thermodynamic analysis carried out in Chapter 2, it was determined that the catalytic CH3OH reforming reaction becomes spontaneous at 41oC. Thus, at 75oC, the catalytic CH3OH reforming reaction is thermodynamically spontaneous making it possible to study the electrochemical promotion of catalysis (EPOC). Results obtained for the oxidation of 2 M CH3OH in 0.5 M H2SO4 are illustrated in Fig. 4.10 and discussed in this section. 186 Figure 4.10: Cyclic voltammograms of (a) Pt, (b) Pt-Ru/C on GC, and (c) Pt-Ru/TiO2 on GC (2 M CH3OH in 0.5 M H2SO4, 23 ±2oC, 50 and 75 ±1oC, 50 mV/s, no rotation). The CH3OH oxidation on-set potentials and peak potentials, Ep, for the forward and reverse scans are summarized in Table 4.4 and 4.5 respectively. Table 4.6 lists the conditions of relevant CV studies 0.00E+00 2.00E-04 4.00E-04 6.00E-04 8.00E-04 1.00E-03 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 IR Corrected V vs. SHE C ur re nt (A ) ambient 50oC 75oC (a) (b) (c) 0.00E+00 1.00E-02 2.00E-02 3.00E-02 4.00E-02 5.00E-02 6.00E-02 7.00E-02 8.00E-02 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 IR Corrected V vs. SHE C ur re nt (A ) ambient 50oC 75oC 0.00E+00 5.00E-04 1.00E-03 1.50E-03 2.00E-03 2.50E-03 3.00E-03 3.50E-03 4.00E-03 4.50E-03 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 IR Corrected V vs. SHE C ur re nt (A ) ambient 50oC 75oC C ur re nt (A ) C ur re nt (A ) C ur re nt (A ) C ur re nt (A ) C ur re nt (A ) 187 which were carried-out in the literature using similar electrocatalysts and conditions. Direct comparison with data available in the literature was difficult due to the varying CV test conditions. Table 4.4: On-set potentials for the forward and reverse CH3OH oxidation on different catalytic surfaces (2 M CH3OH in 0.5 M H2SO4). Temperature 23 ±2oC 50 ±1oC 75 ±1oC Electrode Forward [V] Reverse [V] Forward [V] Reverse [V] Forward [V] Reverse [V] Pt 0.63 0.94 0.65 0.93 0.75 0.90 Pt-Ru/C - GC, 20 wt% Nafion 0.6 1 0.5 0.45 0.4 0.4 Pt-Ru/TiO2 - GC, 5 wt% Nafion 0.73 - 0.68 0.82 0.82 0.82 Table 4.5: Peak potentials for the forward and reverse CH3OH oxidation on different catalytic surfaces (2 M CH3OH in 0.5 M H2SO4). Temperature 23 ±2oC 50 ±1oC 75 ±1oC Electrode Forward [V] Reverse [V] Forward [V] Reverse [V] Forward [V] Reverse [V] Pt 0.90 0.74 0.88 0.80 0.87 0.81 Pt-Ru/C - GC, 20 wt% Nafion 0.93 0.79 1.2 - 1.2 - Pt-Ru/TiO2 - GC, 5 wt% Nafion 0.87 - 1.2 - 1.2 - As can be seen in Fig. 4.10 (a), on Pt, a CH3OH oxidation peak is present on the positive (anodic) potential scan. At the onset potential of about 0.6 V vs. SHE, the CH3OH adsorbed on Pt starts to oxidize and dissociate to CO, increasing the anodic current: CH3OHads → COads + 4 H+ + 4 e- (Eq. 4.5) The complete CH3OH oxidation can also occur, further increasing the current: 188 COads + OHads → CO2 + H+ + e- (Eq. 4.6) Table 4.6: List of CV studies on unsupported and supported Pt, Pt-Ru and TiO2 available in the literature. Electrode Description Concentration [M] Scan Rate Potential Range [V vs. SHE] T Ref. Material H2SO4 CH3OH [mV/s] Low High [oC] Pt (110) 0.1 0.2 50 -0.1 1 20 [179] Pt(110 and 111) 0.1 0.2 50 0 1 20 [181] Pt 0.5 0.1 12.5 0.1 1.5 [219] Pt 0.1(HClO4) 0.1 50 0 1.5 25 [220] Pt 1 (HClO4) 0.1 100 0 1.8 [221] Pt/C 40 wt% Pt 0.5 0.1, 0.5, 1, 2 50 0 1 30, 60, 80 [184] Pt/C 45.9 wt% 25 - 80 [222] Pt/C 20 wt%, Pt:Ru 1:1 a/o 0.5 0 - 3 20 0.1 0.8 [223] Pt/C 0.2 mg/cm2 0.3 0 50 -0.1 1 [224] Pt/C 0.3 0.2 0 1 [224] Pt/C 0.1 - 0.8 mg/cm2 0.1 - 1.2 5 - 50 0.7 - 1.2 25 - 65 [224] Pt/C 47.5 wt% Pt 15-20 μg/cm2 0.5 0.5 10 0.1 1.1 22, 60 [225] Pt(100)/Ru, Pt(110)/Ru and Pt(111)/Ru 0.1 (HClO4) 0.6 50 0 0.6 [192] PtRu Pt:Ru 7/3 a/o 1 (HClO4) 0.1 100 0 1.8 [221] Pt-Ru/C 40 wt% Pt-Ru, Pt:Ru 1:1 a/o 0.5 0.5 20 0 1 23 [177] Pt-Ru/C 27.87 mg/cm2 0.5 0 - 2 20 0 1 [222] Pt-Ru/C 29.8 wt% Pt, 23.1 wt% Ru, Pt:Ru 2:3 a/o 0.5 0.5 0.2 1.4 [226] Pt-Ru/C 54 wt% Pt-Ru Pt:Ru 2:3 a/o 15-20 μg/cm2 0.1 0.5 10 0.1 1.1 22, 60 [225] Pt (111) Ru modified Pt(111) 0.1 (HClO4) 0.5 50 0.1 0.9 25 [227] Pt-Ru/TiO2 0.08 (HClO4) 4.67 200 -0.2 0.8 20 [228] Pt-Ru/C-TiO2 30 μg/cm2 1 0.5 50 0 1.4 25 [229] TiO2 nanotubes (TNT)/Pt/C - Au TiO2:Pt 1:1 a/o 1 (HClO4) 1 (C2H5OH) 50 -0.2 1 25 [230] TiO2/carbon nanotubes (CNT)/Pt 0.042 to 1.15 mg/cm2 0.5 1 100 0 1 Room T [231] Pt-TiO2/CNT 20 wt% Pt Pt:TiO2 1:1 a/o 1 (HClO4) 1 (C2H5OH) 50 -0.2 1 25 [232] 189 The formation of OHads on Pt occurs between 0.5 to 0.8 V vs. SHE [233]. The electro-oxidation of CO occurs at about 0.2 V lower on Pt-Ru [234]. At potentials greater than about 0.9 V vs. SHE, the rate of CH3OH oxidation decreases as the formation of Pt oxides starts to occur. The formation of PtO film, which deactivates the catalyst’s surface is: Pt + H2O → PtO + 2H+ + 2 e- Eao = 0.88 V vs. SHE (Eq. 4.7) The scan potential was reversed at 1.2 V vs. SHE to avoid oxygen evolution which starts to occur at 1.23 V vs. SHE. A reduction peak appeared on the negative (cathodic) potential scan due to the dissolution of the Pt oxide film to Pt: PtO + 2H+ + 2e- → Pt + H2O Eco = 0.88 V vs. SHE (Eq. 4.8) After the PtO is reduced, CH3OH is oxidized on the clean Pt particles surface and another CH3OH oxidation peak appears. This reverse oxidation peak current depends on the amount of poisoning species removed from the Pt surface. The CH3OH oxidation decreases due to an increase in adsorbed CO species. The CH3OH oxidation current is always greater on the forward scan than on the reverse scan because of the large irreversibilities involved in the surface oxidation and oxide reduction processes [181]. These forward and reverse scan observations are in agreement with similar studies reported for CH3OH oxidation on Pt/C in acidic CH3OH solutions [224, 226]. The CH3OH oxidation peak current depends on the catalyst loading, the acid concentration, scan rate, temperature, and the amount of poisoning species formed at the catalyst surface [224]. In this study, only the effect of temperature was looked at. Although there may have been a shift in the MSE reference electrode potential of about -0.04 V at 75oC, based on a temperature coefficient of -8.1x10-4 V/oC, it was clear that the kinetics of CH3OH oxidation improved with temperature. The forward CH3OH oxidation peak current was an order of magnitude greater at 75oC than at 23oC, indicating that the electrocatalytic activity increases at higher temperature as the removal of the CO species poisoning the Pt surface becomes easier. 190 The CVs obtained for CH3OH oxidation on 20 wt% metal Pt-Ru/C-GC with 20 wt% Nafion are shown on Fig. 4.10 (b). At potentials greater than 0.8 V vs. SHE, Ru dissolution (oxidation) may start to occur [235], however, conditioning has shown that the CVs obtained were stable and that no working electrode degradation occurred. On Pt-Ru/C, the forward and reverse CH3OH oxidation on-set potentials were shifted to lower potentials compared to that observed on Pt. Although Ru itself is inactive towards CH3OH oxidation, its presence results in a bi-functional CH3OH oxidation mechanism, which lowers the oxidation potential of CO oxidation via the formation of adsorbed hydroxide ions [226]. It was reported that OH adsorption on Ru starts at 0.2 V vs. SHE [236], and that it shifts the CH3OH oxidation onset potential on Pt-Ru/C negatively compared to Pt/C. Nevertheless, at ambient temperature, the activity of Pt for CH3OH oxidation was comparable to that of Pt-Ru/C. The forward CH3OH oxidation peaks current obtained on Pt-Ru/C was the same as the forward CH3OH oxidation peak current obtained on Pt under similar conditions (6.45x10-4 A). The reverse oxidation peak current obtained on Pt-Ru/C (3.27x10-4 A) was only slightly smaller than the reverse peak current obtained on Pt (7.53x10-4 A). This agrees with studies according to which Ru is not very active for CH3OH adsorption and dehydrogenation at ambient temperatures [225]. The difference in the peak potentials obtained at ambient temperature for the Pt-Ru/C and Pt catalyst is due to the formation of Pt oxide on the forward scan, which occurs at a potential slightly higher on Pt-Ru/C than on Pt, and on the reverse scan, the reduction of Pt oxide, which occurs at a potential slightly higher on Pt-Ru/C than on Pt. Temperature had a strong effect on the activity of Pt-Ru/C for CH3OH oxidation. As expected, at higher temperatures, the CH3OH oxidation rate was faster, and the onset potential for CH3OH oxidation was shifted to lower values on Pt-Ru/C, as can be seen in Table 4.4. In accordance with the literature, increasing the temperature activated the Ru sites and resulted in improvements in the adsorption and dehydrogenation steps of CH3OH oxidation, which exceeded that achievable on Pt [225]. At 50oC, the forward CH3OH oxidation peak on Pt-Ru/C - GC is at a much greater current than that observed on Pt at the same temperature due to improvements in the CO oxidation kinetics (Eq. 4.6). Initially, on the forward scan, the Pt-Ru/C surface is covered by CO. Once the onset potential for CO oxidative removal is reached, most CO is stripped from the surface. It seems like an insignificant amount of inactive Pt oxides were formed in parallel as the CO oxidation effect of Ru was not significantly hindered. Also, on 191 the reverse scan, CH3OH re-oxidation occurred right away as the Pt surface was already cleared of Pt oxides. The CO eventually re-accumulated on the surface, and the CH3OH oxidation gradually diminished. At 75oC however, the Pt-Ru/C - GC CH3OH oxidation and re-oxidation peaks completely overlapped. At this temperature, the Ru surface very effectively oxidized CO from the Pt sites, and was able to sustain the decomposition of CH3OH, after the CH3OH oxidation onset potential was reached. The adsorption of OH on Ru occurred at an even lower potential, but the adsorbed OH also desorbs faster. As the CO poison does not occupy the Pt active sites, the Pt surface is clean to oxidize CH3OH. Again, it seems like the CO removal effect of Ru surpassed the formation and reduction of the oxides, which should also be accelerated at higher temperature. Forward and reverse scans almost overlapped for a Pt/C catalyst in 0.5 M CH3OH in 0.5 M H2SO4 at 80oC due to very fast reaction kinetics [226]. According to Tremiliosi-Filho et al., overlapping indicated that the catalyst surface did not experience considerable CO poisoning in both scan directions, and that no other surface effects were involved [192]. The CVs obtained for 20 wt% metal Pt-Ru/TiO2 - GC with 5 wt% Nafion are shown in Fig. 4.10 (c). The ECSA of the Pt electrode is close to its geometric surface area, and is much smaller than the ECSA of the GC Pt-Ru/C or Pt-Ru/TiO2 ink coated electrodes, and therefore it could be expected that the currents would be greater for both ink coated GCs. However, the polarization of the Pt-Ru/TiO2 - GC was poor, likely because of the low conductivity of the TiO2 support. Forward and reverse oxidation peaks have been observed on conductive TiO2 catalyst containing carbon and carbon nanotubes, such as for TiO2/Pt/C, and Pt-Ru/C-TiO2, and TiO2/CNT/PtRu in acidic CH3OH solutions [229-231]. For Pt/Nb-TiO2, (Nb:Ti 0.1:0.9), Pt loadings less than 10 wt% resulted in conductivity of about 10-5 S/cm, while with Pt loading of 60 wt% the conductivity increased to 1.26 S/cm [207]. The poor response of synthesized Pt-Ru/TiO2 may also be due to the larger particles compared to that of the commercial Pt-Ru/C, as observed in the material characterization Section 4.3.1. Generally, the activity is greater on smaller particles. Therefore the activity of the synthesized Pt-Ru/TiO2 may be improved by lowering the calcination temperature to reduce particle growth. It was also reported that the predominant face present in small supported catalyst particles, Pt(111), is more active towards CH3OH oxidation [237]. Hepel et al. [228], observed forward and reverse CH3OH oxidation peaks on Pt-Ru supported on basal TiO2 films formed by a voltammetric procedures which prevented the formation of compact non-conductive TiO2 films. To obtain CVs showing a large current loop extending the CH3OH anodic oxidation currents to 192 very low potentials, they first applied an anodization potential greater than the COad oxidative desorption potential to clean and activate the catalyst surface. The current Pt-Ru/TiO2 ink composition was based on an ink composition used for Pt-Ru/C inks and further ink optimization may be required as TiO2 is hydrophilic. Furthermore, the hydrophilic nature of TiO2 may affect the CH3OH bonding ability on the Pt surface. CH3OH may not absorb as easily to the Pt surface as extra energy may be required to displace water molecules which are in the way, making the electrochemical process slower and less efficient. Strong metal-support interactions (SMSI) can occur in catalysts systems comprising metal particles dispersed on the surface of inorganic oxides after reduction at certain temperatures [217]. Therefore, it is possible that the interactions between the Pt and Ru metallic phase and the TiO2 support affected the intrinsic activity of the Pt-Ru/TiO2 electrocatalyst. The chemisorption ability of the precious metals may have been altered by interactions with the TiO2 surface and the TiO2 support may have shifted the CH3OH oxidation onset potential to a more positive value. In addition to Ru, the TiO2 support may also provide hydroxide ions to remove CO poisoning from the Pt surface. Based on the CVs obtained, it is not possible to conclude that the CH3OH electrochemical oxidation mechanism observed on Pt-Ru/TiO2 - GC is similar to the one observed on Pt-Ru/C. Nevertheless, catalytic reforming tests were successful using Pt-Ru/TiO2 in the literature, and a low electrocatalytic activity (low electrode polarization) is actually preferred to carry out EPOC experiments. This is because the currents or potentials involved in EPOC experiments are usually very small, and because the rate enhancement factor, Λ, is inversely proportional to the electrode exchange current, Io as per Eq. 1.38. 4.3.3 Powder catalysis The mass spectrometer provided a real-time analysis of catalytic CH3OH reforming H2 production rate using either 0.5 g of Pt-Ru/C or 0.5 g Pt-Ru/TiO2 catalyst powders. In preliminary tests, the H2 formation rate was measured as a function of time as the temperature was continuously increased from ambient temperature to 100 ±1oC. The H2 formation rates hence obtained were transient values and not steady- state equilibrium values. An example is shown in Fig. 4.11. 193 Figure 4.11: H2 produced as a function of time (ambient to 100 ±1oC, 0.5 g catalyst, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4). It was found that, for both catalysts, some H2 was produced at temperatures below 41oC, the temperature at which the catalytic reforming of CH3OH in the liquid phase becomes thermodynamically spontaneous for an equimolar CH3OH - water solution. However, the system studied here is not at thermodynamic equilibrium. To be at thermodynamic equilibrium, the catalytic system studied would need to be at chemical and thermal equilibrium. For example, reaction rates are not taken into account in equilibrium thermodynamics calculations. The powder catalysis system evaluated in Fig. 4.11 is not at thermal equilibrium as the temperature is constantly changing over time. Nevertheless, a steeper increase in the quantity of H2 produced was observed from 80 to 85oC with Pt-Ru/C and from 80 to 90oC with Pt-Ru/TiO2. Improvements in the H2 production rate are expected as temperature increases as the catalytic reaction rate constant depends exponentially on the temperature as per Arrhenius’s law: ⎟⎠ ⎞⎜⎝ ⎛ −⋅= RT EAk ArA exp (4.9) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 Time (min.) H 2 pr od uc ed (m m ol ) 0 10 20 30 40 50 60 70 80 90 100 Te m pe ra tu re (o C ) Pt-Ru/C Pt-Ru/TiO Temperature, Pt-Ru/C Temperature, Pt-Ru/TiO2 2 Pt-Ru/C Pt-Ru/TiO2 Pt-Ru/TiO2 Pt-Ru/C H 2 pr od uc ed (m m ol ) Te m pe ra tu re (o C ) 194 where kA, the reaction rate constant and Ar, the pre-exponential factor, both have units of 1/s for a first order reaction, and EA is the activation energy. For the purpose of this study, the temperature region between 80 to 90oC was avoided in all subsequent tests in order to keep the conversion rate low and increase the sensitivity to the effect of EPOC. A temperature of 75oC was chosen to carry out the powder catalysis and GDE catalysis experiments, as well as to study the effect of EP. Results obtained for the baseline powder catalysis tests carried out from ambient temperature to 75 ±1oC, until a steady- state H2 production rate was reached at that temperature are presented as shown in Fig. 4.12. It took about 30 to 40 minutes to reach 75 ±1oC, and the catalytic reforming experiment was held for about 70 minutes at this temperature. For both catalysts, the H2 production rate did not increase significantly from ambient temperature to 75 ±1oC. For a semi-batch reactor, it is expected that the longer the reactant is in the reactor, the more reactant is converted until either equilibrium is reached, or the reactant is exhausted. Therefore, the CH3OH catalytic conversion to H2 will depend on the time spent in the reactor. It may be possible that equilibrium was not reached over the duration of the powder catalysis experiment. To evaluate this, the rate of H2 production was plotted as a function of temperature in Fig. 4.13. Figure 4.12: H2 produced as a function of time (ambient to 75 ±1oC, 0.5 g catalyst, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4). 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 0 10 20 30 40 50 60 70 80 90 100 Time (min.) H 2 pr od uc ed (m m ol ) 0 10 20 30 40 50 60 70 80 90 100 Te m pe ra tu re (o C ) Pt-Ru/C Pt-Ru/TiO Temperature, Pt-Ru/C Temperature, Pt-Ru/TiO Pt-Ru/C Pt-Ru/TiO2 Pt-Ru/TiO2 Pt-Ru/C 2 2 75oC H 2 pr od uc ed (m m ol ) Te m pe ra tu re (o C ) 195 Figure 4.13: H2 production rate as a function of temperature (0.5 g catalyst, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4). Once 75 ±1oC was reached, the H2 production rate kept increasing slowly at this temperature with increasing run time. The Pt-Ru/C test was kept at 75 ±1oC for an hour and 30 min. and the Pt-Ru/TiO2 test was kept at 75 ±1oC for an hour. The average H2 production rate at 75 ±1oC on Pt-Ru/TiO2 (0.017 mmol/min.) was 36.6% greater than on Pt-Ru/C (0.013 mmol/min.) under the same experimental conditions. The average H2 formation rates obtained at 75 ±1oC were converted to μmol/h.g catalyst and x106 ml/h.g catalyst for comparison with experimental results available in the literature, which are summarized in Table 4.7. 0.000 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.010 0.011 0.012 0.013 0.014 0.015 0.016 0.017 0.018 0.019 0.020 0 10 20 30 40 50 60 70 80 90 100 Temperature (oC) H 2 pr od uc tio n ra te (m m ol /m in .) Pt-Ru/C Pt-Ru/TiO 196 Table 4.7: Literature powder catalysis CH3OH reforming production rates at low temperature (0.5 g Pt- Ru/TiO2, 10 wt% metal, Pt:Ru 1:1 a/o, 60 ml of 2 M CH3OH). H2 reduction Reaction Molar flow rate (Fr) Volumetric flow rate (Vr) Ref T T H2 CO CO2 H2 CO CO2 Gas [oC] [oC] [μmol/h.g catalyst] [x106ml/h.g catalyst] - 72 102.9 - - 2.91 2.91 [29] 400 77-84 105.6 4.2 30.6 3.06 0.12 0.89 4.07 [14] 200 105 59.8 12 4 1.86 0.37 0.12 2.35 [16] 200 105 158.6 21.2 29.8 4.92 0.66 0.92 6.50 [16] 400 105 205.8 6.6 66 5.96 0.19 1.91 8.07 [16] 450 105 321.6 9.6 106.8 9.98 0.30 3.31 13.59 [16] All catalytic CH3OH reforming experiments reported in the literature used powder catalysts with a precious metal content of 10 wt%, and a metal catalyst mass (0.5 g) to solution volume (60 ml) ratio of 8x10-4. The experiments carried out in this study were conducted using catalysts with double the precious metal content (20 wt%), and a metal catalyst mass (0.5 g) to solution volume (20 ml) ratio of 0.005. Results obtained in the current study are summarized in Table 4.8. In the last two columns of Table 4.8, the experimental results were normalized to the metal catalyst mass to solution volume ratio of 8x10-4 for better comparison with the results available in the literature. Table 4.8: Powder catalysis CH3OH reforming production rates (0.5 g catalyst, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4, 75 ±1oC). Catalyst H2 reduction Reaction Molar flow rate (Fr) Volumetric flow rate (Vr) Normalized Fr Normalized Vr T T H2 H2 H2 H2 [ oC] [oC] [μmol/h.g catalyst] [x106ml/h.g catalyst] [μmol/h.g catalyst] [x106ml/h.g catalyst] Pt-Ru/C (commercial) Unknown 75 ±1 1531 44 255 7 Pt-Ru/TiO2 (synthesized) 500 75 ±1 2090 60 348 10 197 The H2 production rate or formation volume was 36.5 % greater for the Pt-Ru/TiO2 catalyst than for the Pt-Ru/C catalyst under the same reaction conditions. Thus, the synthesized Pt-Ru/TiO2 had a greater activity for the catalytic reforming of CH3OH than the commercial Pt-Ru/C. As the precious metal content was the same for both catalysts studied, it may be possible that the larger H2 production rate observed with the Pt-Ru/TiO2 catalyst is linked to strong metal-support interactions (SMSI). The normalized Pt- Ru/TiO2 H2 formation rate and the H2 formation volume were about three times greater than the results reported by Miyao et al. for the same catalyst, but reduced at 400oC and at reaction temperatures of 77- 84oC [14]. The normalized Pt-Ru/TiO2 H2 formation rate was 69 % greater and the H2 formation volume was 22.8 % greater than the results reported by Miyao et al. for the same catalyst, but reduced at 400oC and a reaction temperature of 105oC [16]. However, when comparing with the same catalyst, but reduced at 450oC and at a reaction temperature of 105oC, the H2 formation rate was 8% greater and the H2 formation volume was 3.7% smaller in this study. The differences in the rates may arise from the fact that their catalyst reduction temperature was slightly lower (400-450oC) and their reaction temperature, slightly greater (77-105oC) than the one used in this study (500oC, 75 ±1oC). Greater reaction temperatures would improve the CH3OH conversion and result in greater H2 production. Also, the experiments reported in the literature were conducted for 5 hours, and a longer run time will result in more H2 production. Nevertheless, since the powder catalysis production rates obtained in this study are good and generally in a similar range as the ones reported in the literature under similar conditions, the H2 rate measurement method was determined to be adequate to conduct further catalytic studies and evaluate the effect of EPOC. Further confirmation of the validity of the experimental H2 monitoring method was obtained by looking at the CH3OH partial pressure monitored by the MS over time. Based on the preliminary CH3OH evaporation test described in Section 4.2.5, it can be estimated that, over a period of 2 hours at 75oC, the loss of CH3OH concentration due to evaporation could be up to 10%. Lower H2 production rates would result if there would be a significant loss of unreacted CH3OH by evaporation. Figure 4.14 shows the CH3OH partial pressure measured during the catalytic reforming experiments shown on Fig. 4.12 and 4.13. 198 Figure 4.14: CH3OH partial pressure as a function of time (a) Pt-Ru/C, (b) Pt-Ru/TiO2 (ambient to 75 ±1oC, 0.5 g catalyst, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 20 ml of 2 M CH3OH in 0.5 M H2SO4). The CH3OH partial pressure was very low and did not increase significantly during the course of the catalytic reforming experiment, indicating that the gas bubbler was effective at preventing CH3OH vapours from entering the MS. The MS also recorded the partial pressure of CO2 during the catalytic reforming experiment. It would be expected that the partial pressure of CO2 would increase proportionally to the partial pressure of H2. However, this was not the case and the partial pressure of CO2 detected by the MS did not change significantly during the course of the experiment. This is may be due to the large solubility of CO2 in water which is 3.4 g/l at 0oC, i.e., 0.578 g of CO2 in the 170 ml of H2O contained in the two gas bubblers. Assuming the molar flow rate of CO2 is 1/3 of the H2 molar flow rate, when using 0.5 g of Pt-Ru/TiO2, it can be calculated that 0.015 g of CO2 is generated per hour, which is much less than the solubility limit of the water contained in the chilled gas bubblers. The DI water in the gas bubblers was changed between each experiment. Knowing the catalytic CH3OH reforming H2 formation rate at open-circuit and the exchange current, it is possible to estimate the magnitude of the absolute Faradaic efficiency on Pt-Ru/C at 75 ±1oC using Eq. 1.38. As per Table 2.3 in Chapter 2, the exchange current for 2M CH3OH at 75 ±1oC was estimated to 0.0E +00 5.0E ‐09 1.0E ‐08 1.5E ‐08 2.0E ‐08 2.5E ‐08 3.0E ‐08 3.5E ‐08 4.0E ‐08 0 10 20 30 40 50 60 70 80 90 100 Time (min.) M et ha no l P ar tia l P re ss ur e (k Pa ) 0 10 20 30 40 50 60 70 80 90 100 Te m pe ra tu re ( o C ) Methanol partial pressure, Pt-Ru/TiO Methanol partial pressure, Pt-Ru/C Temperature, Pt-Ru/TiO Temperature, Pt-Ru/C 2 2 199 be 8.1x10-6A using a GDE, and as per Table 4.8, the unpromoted catalytic rate is 4.25x10-7 mol/s using catalyst powder. Hence, |Λ| is estimated to be approximately 5067, which is two orders of magnitude larger than that reported for most low temperature EPOC studies using Nafion, which range from 20 to 28. The large |Λ| calculated may have been caused by discrepancies in the exchange current density estimation, and the use of an unpromoted catalytic rate obtained for a catalyst powder reforming experiment rather than one using a GDE. 4.3.4 GDE catalysis and electrochemical promotion of catalysis The EPOC tests were carried out in the two-compartment glass cell and required the use of gas diffusion electrodes. The anodic compartment of the electrolyser served also as the catalytic reaction chamber. Therefore, in EPOC experiments, the anodic GDE acted as a catalyst and as an electrocatalyst at the same time. The H2 produced in the cathodic compartment during the EPOC tests was vented to atmosphere and was not included in the H2 generation rate values reported. Only the gases evolving from the anodic compartment were sent to the MS for measurements. The anode GDE was a 2.01 cm2 disk coupon of a 4 mg/cm2, 20 wt% metal Pt-Ru/C or Pt-Ru/TiO2. The anodic compartment was filled with 80 ml of 2 M CH3OH acidic solution. The resulting metal catalyst mass to volume ratio was of 2x10-5, which is many orders of magnitude smaller than that used in the prior powder catalysis experiments (0.005). Nevertheless, this should keep the conversion rates sufficiently low (below 5 %) and improve the sensitivity of the EPOC effect on the catalytic reaction rate. The helium carrier gas flow rate was kept at the same rate as for the powder catalysis experiments. The range of imposed currents or potentials evaluated for selected chemical reactions for which EPOC was studied in the literature are listed in Table 4.9. As can be seen, the range of current imposed in the literature varied widely depending on the type of conducting electrolyte used, the chemical reaction studied, and the reaction temperature. Values were only reported on a current basis in the literature, and as not all geometric surface areas were available, they could not be converted to current densities. All chemical reactions using cationic conductors studied for the EPOC effect have been using gaseous reactants, even at low reaction temperature. Based on prior Pt-Ru/C polarization curves shown in Fig. 2.12 c) of Chapter 2, at 75oC for 2 M CH3OH, it can be expected that electrocatalysis should start at 200 currents between 100 and 500 μA, i.e., current densities of 0.05 and 0.25 mA/cm2. When currents smaller than this were imposed, the electrolyser cell voltage, UWC, remained positive for the entire experiment. Larger currents were also evaluated for the electrochemical reforming of 2 M CH3OH using Pt-Ru/C, but they did not surpass the limiting current density of 75 mA/cm2, i.e., 0.15 A. Table 4.9: Imposed currents or potentials on selected chemical reactions subjected to EPOC. Electrolyte Chemical Catalyst and T Imposed Imposed potential Ref type Reaction protonic electrolyte [oC] current [μA] [V vs. SHE] Oxide Methanol oxidation Pt-YSZ 350-650 - 0 to 12.5, and 0 to - 12.5 [238] Methanol oxidation Ag-YSZ 500 0 to -2000 - [138] Methanol dehydrogenation Ag-YSZ 600-680 -2000 - [139] Methane reforming Ni-YSZ 750-950 300000 - [239] C2H4 oxidation Pt-YSZ 370 1 - [240] C2H4 oxidation IrO2/YSZ 380 10 - [241] C2H4 oxidation Rh-YSZ 350 400 - [242] Protonic Hydrogen oxidation Pt black-N117 room T 5000 - [147] C2H4 oxidation Pt-CaZr0.9In0.1O3-α 450 3 - [243] CO oxidation Pt-β"-Al2O3 350 -20 [244] 1-butene isomerization Pt black/N- 117/Pd/C 70 - -0.05 to 0.38 [245] Liquid Hydrogen oxidation Pt in 0.1 M KOH 25-50 -10000 to 15000 - [246] Hydrogenation of maleic acid Platinized Pt in 0.5 M HClO4 26 - 0.05 to 1.8 [247] 201 Each EPOC test was conducted using a different MEA of similar composition. For each test, three different EP currents were sequentially applied. In general, it took about 30 min. to attain the desired temperature of 75 ±1oC, as indicated by the first vertical dash line of the graph. Once at the desired reaction temperature, the catalytic reaction rate was measured for about 2.5 hours under OCV before the first galvanostatic current was applied to obtain the unpromoted catalytic reaction H2 production rate value, ro. The galvanostatic currents were applied for 30 min. and ranged from -0.15 to 0.01 A (-75.63 to 5 mA/cm2). In EPOC studies with oxide conductors, the current I is positive when O2- is pumped to the catalyst surface and negative when O2- is pumped away from the catalyst surface [238]. Based on this convention, for a protonic conductor, I was taken as positive when H+ is removed from the surface and as negative when H+ is supplied to the catalyst surface. The system was brought back to OCV for 30 min. between each galvanostatic current application. Vertical dashed lines indicate the time at which the galvanostatic current application periods started and stopped. The promoted catalytic reaction H2 production rate obtained during the EPOC portion of the test were averaged as r values. The unpromoted catalytic reaction rate during the other OCV portions of the test were averaged as r’ values. These r’ values should be the same as the ro values, if 30 min. was sufficient for the chemical reaction system to attain equilibrium again. An EPOC test conducted with Pt-Ru/TiO2 is shown in Fig. 4.15 and one conducted on Pt-Ru/C is shown in Fig. 4.16. 202 Figure 4.15: H2 production rate and electrolysis cell voltage as a function of time during unpromoted and promoted catalysis (2.01 cm2, 4 mg/cm2 Pt-Ru/TiO2, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC). It appeared that 300 min. (5 hours) was sufficient for the catalytic reaction to reach steady-state under open circuit conditions. The catalysis conditioning profiles were similar for other MEAs evaluated in our research with Pt-Ru/TiO2 or Pt-Ru/C. It was calculated that a steady state H2 production rate of 0.006 mmol/min corresponded to an electrolysis current density of about 30 mA/cm2. At this current density, CH3OH would be electrolysed on the WE of the electrolyser as per earlier 2 M CH3OH electrolysis results on 4 mg/cm2 Pt-Ru/C at 75oC (Fig. 2.12). This calculated current density is unlikely at open circuit voltage. It is possible that the MS has a poor sensitivity at detecting low flow rates of H2 compared to the sensitivity of the Multistat at controlling currents or potentials. 0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016 0.018 0.020 0 30 60 90 12 0 15 0 18 0 21 0 24 0 27 0 30 0 33 0 36 0 39 0 42 0 45 0 48 0 51 0 54 0 Time (min.) H 2 P ro du ct io n R at e (m m ol /m in .) -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 C el l V ol ta ge (V ) a nd W E P ot en tia l ( V v s. S H E ) Pt-Ru/TiO H production rate Cell voltage WE potential ro r’r OC V OC V OC V OC V OC V 1 µA -1 µ A 10 00 0 µA -1 00 00 µ A 22 0 30 60 90 12 0 15 0 18 0 21 0 24 0 27 0 30 0 33 0 36 0 39 0 42 0 45 0 48 0 51 0 54 0 H 2 P ro du ct io n R at e (m m ol /m in .) C el l V ol ta ge (V ) a nd W E P ot en tia l ( V v s. S H E ) 203 The experiments carried out in this section were conducted using a geometric area of 2.01 cm2 with a loading of 4 mg/cm2 catalysts with a precious metal content (20 wt%), and, hence, a metal catalyst mass (0.0016 g) to solution volume (80 ml) ratio of 2x10-5. The H2 formation rates for the unpromoted catalytic reaction, ro, obtained with GDEs were normalized to a metal catalyst mass to solution volume ratio of 8x10-4 for comparison with the powder catalysis tests results discussed in Section 4.3.4. The average unpromoted catalytic H2 production rates obtained with the GDEs are listed in Table 4.10 along with the normalized rates for comparison with the powder catalysis test results. Similar to the powder catalysis cases, the unpromoted catalytic H2 production rate using Pt-Ru/TiO2 was about 27 % greater than the average unpromoted catalytic H2 production rate obtained using various Pt-Ru/C based MEAs. Figure 4.16: H2 production rate and electrolysis cell voltage as a function of time during unpromoted and promoted catalysis (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC). 0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016 0.018 0.020 0 30 60 90 12 0 15 0 18 0 21 0 24 0 27 0 30 0 33 0 36 0 39 0 42 0 45 0 48 0 51 0 54 0 Time (min.) H 2 P ro du ct io n R at e (m m ol /m in .) -0.50 -0.45 -0.40 -0.35 -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 C el l V ol ta ge (V ) a nd W E P ot en tia l ( V v s. S H E ) Pt-Ru/C H production rate Cell voltage WE potential 10 00 0 µ A -1 00 00 µ A ro r’r OC V OC V OC V OC V OC V 1 µA -1 µ A 2 0 30 60 90 12 0 15 0 18 0 21 0 24 0 27 0 30 0 33 0 36 0 39 0 42 0 45 0 48 0 51 0 54 0 H 2 P ro du ct io n R at e (m m ol /m in .) C el l V ol ta ge (V ) a nd W E P ot en tia l ( V v s. S H E ) 204 Table 4.10: Average H2 unpromoted production rates for GDE catalysis of CH3OH at low temperature (2.01 cm2, 4 mg/cm2 Pt-Ru/C or Pt-Ru/TiO2, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC). Catalyst Molar flow rate (Fr) H2 Normalized Fr H2 [μmol/h.g catalyst] ± 10184 μmol/h.g catalyst ro [μmol/h.g catalyst] ± 407360 μmol/h.g catalyst ro Pt-Ru/C 36582 1463280 Pt-Ru/TiO2 42163 1686520 The standard deviation values were calculated based on tests conducted with three different MEAs with Pt-Ru/C GDEs. The un-normalized molar flow rates measured were adjusted to 1 g of catalyst, while the normalized results were adjusted for the metal catalyst mass to solution volume ratio. It was verified that, without the catalyst mass adjustment and without the metal catalyst mass to solution volume ratio, more H2 was evolved in the powder catalysis cases than for the GDE catalysis cases, as expected. However, it can be seen that the normalized H2 formation rates for the unpromoted catalytic CH3OH reforming reaction obtained with GDEs are several orders of magnitude greater than the ones obtained for powder catalysis in Table 4.8. Different reactor configurations were used to carry the powder catalysis tests (single-compartment glass cell reactor) and the GDE tests (2-compartment glass cell reactor). It is possible that EPOC was induced at the GDE in the 2-compartment reactor without the application of external current to stimulate the catalytic reaction. It has been reported that, for certain configurations, a current or potential difference may developed between the catalyst WE surface and CE at open circuit due to the occurrence of the catalytic reaction on the catalyst WE surface [132]. This may be possible as the H2 production rate measurements have shown that the GDE Pt-Ru/C catalyst WE was not completely inert to catalytic activity even at ambient temperature where there is no thermodynamic driving force for H2 production. This could be further verified by measuring the quantity of H2 produced during a test carried out in the single-compartment reactor using a non-conductive connection to a GDE catalyst sample immersed in the reacting solution. This value would results in the true ro value, i.e., without the wireless application of a small current or potential due to the 2-compartment cell MEA. If the H2 production rate obtained in the single-compartment reactor case, ro, is indeed lower than that obtained 205 for the equivalent test carried out in the double-compartment reactor, r, then a wireless EPOC effect exist when using the 2-compartment reactor. Knowing the electrochemical cell internal current at open circuit voltage in the 2-compartment cell would then allow the determination of the Faradaic efficiency for this self-driven wireless EPOC case. As shown in Fig. 4.15 and 4.16, during the application of 1 or -1 μA (0.5 or -0.5 μA/cm2), the electrolyser cell voltage, UWC remained approximately zero and no effect on the H2 production rate were observed. It was found that larger negative currents, i.e. -0.01 A (-5 mA/cm2) with a positive cell voltage, were required to observe an effect. For Pt-Ru/TiO2 or Pt-Ru/C, only the application of a negative current resulted in an increase in the H2 production rate. The proposed electrochemical enhancement mechanism for the catalytic reforming of CH3OH is described in Fig. 4.17. Figure 4.17: Schematic diagram of the reverse electrolyser electrical circuit and of the proposed EPOC mechanism for the catalytic reforming of methanol. When a negative current is applied to the electrolyser circuit, H+ ions are supplied to the WE surface. Water is oxidized in the catholyte compartment of the glass cell and the electrolyser operated in reverse. WE Polymer Electrolyte Membrane Current Collector Electrolyser Current Collector Electrolyser Circuit CE Ielec Uelec G/P H+ Catalytic reforming: CH3OH + H2O → CO2 + 3H2 2 M CH3OH 0.5 M H2SO4 0.5 M H2SO4 (N-117) (Pt/C)(Pt-Ru/C or Pt-Ru/TiO2) H2O → 1/2O2 + 2H+ + 2e-2H+ + 2e-→ H2 206 No CH3OH or water is electro-oxidized in the catalytic reaction compartment of the glass cell. However, the reverse electrolyser may generate H2 in the catalytic reaction compartment. Under the experimental conditions studied, only providing H+ to the catalyst surface was beneficial to the catalytic CH3OH reforming reaction rate, and removing H+ from the catalyst surface did not increase the catalytic CH3OH reforming reaction rate. At an applied current of -0.01 A (-5 mA/cm2), the electrochemical promotion effect was about 20 % lower when using Pt-Ru/TiO2 than when using Pt-Ru/C. This is likely due to the low electronic conductivity the Pt-Ru/TiO2 for CH3OH electro-oxidation compared to that of Pt-Ru/C, as was discussed in Section 4.3.2. Based on the powder catalysis tests, it was shown that Pt-Ru/C did possess a lower catalytic activity than Pt-Ru/TiO2 for the electrochemical oxidation of CH3OH. However, using a GDE, to obtain a same enhancement as for the Pt-Ru/C catalyst, more current may need to be applied with the Pt-Ru/TiO2 catalyst to obtain a similar effect on the work function of the working electrode. Consequently, all further EPOC experiments were conducted with commercial Pt-Ru/C catalyst GDEs. Figure 4.18 demonstrates the reproducibility of the catalytic rate enhancement was investigated by applying multiple negative current consecutively. Again, it can be seen that application of negative currents smaller than -0.01 A (-5 mA/cm2) did not impact the catalytic H2 production rate significantly. The consecutive application of a current of -0.01 A following OCV showed that the effect was reversible, i.e., the catalytic H2 production rate always returned to the unpromoted rate upon interruption of the current. The average promoted H2 production rate, r, for the four consecutive current applications of -0.01 A with Pt-Ru/C as the working electrode was of 46469 ±456 μmol/h.g catalyst and the average unpromoted H2 production rate, r’, was of 35131 ±1172 μmol/h.g catalyst. The average r’ did not change significantly (6 %) from the initial ro value of 33057 μmol/h.g catalyst, indicating a reproducible process. 207 Figure 4.18: H2 production rate and electrolyser current as a function of time during unpromoted and promoted catalysis (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC). The application of consecutively larger negative currents was also investigated with Pt-Ru/C and the results are shown in Figure 4.19. The application of greater negative currents resulted in greater increases in the catalytic H2 production rate. As shown in Fig. 2.12 c), the limiting current for normal water electrolysis with 4 mg/cm2 Pt-Ru/C at the anode was 25 mA/cm2 (0.05 A) when the electrolyser cell voltage reached values greater than -2 V in normal electrolysis with 2 mg/cm2 Pt/C at the cathode at 75 ±1oC. Although currents of -0.1 A (-0.05 A/cm2) and -0.15 A (-0.075 A/cm2) were imposed, the reverse electrolyser cell voltage during the EPOC tests carried out in Fig. 4.19 did not reach values greater than 2 V. At reverse electrolysis cell voltages greater than 2 V, carbon oxidation would start to 0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 0.016 0.018 0.020 0 40 80 12 0 16 0 20 0 24 0 28 0 32 0 36 0 40 0 44 0 48 0 52 0 56 0 60 0 Time (min.) H 2 P ro du ct io n R at e (m m ol /m in .) -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 C el l V ol ta ge (V ) a nd W E P ot en tia l ( V v s S H E ) Pt-Ru/C H production rate Cell voltage WE potential ro r’r -0 .0 1 A -0 .0 1 A OC V OC V OC V OC V OC V -0 .0 1 A -0 .0 1 A 2 OC V -0 .0 01 A 0 40 80 12 0 16 0 20 0 24 0 28 0 32 0 36 0 40 0 44 0 48 0 52 0 56 0 60 0 H 2 P ro du ct io n R at e (m m ol /m in .) C el l V ol ta ge (V ) a nd W E P ot en tia l ( V v s S H E ) 208 occur on the counter electrode, where the electrochemical oxidation of water occurs. It was checked that the reverse electrolyser returned to cell voltages close to its original OCV value between each test, indicating that there was no electrode degradation. Figure 4.19: H2 production rate and electrolyser current as a function of time during unpromoted and promoted catalysis (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC). The average H2 formation rates for each EPOC test segment carried out in Fig. 4.18 and 4.19 are reported in Table 4.11. The calculated theoretical Faradaic H2 production rate based on the imposed current, rF, along with the Faradaic efficiencies, Λ, and the EPOC production rate efficiency, ηEPOC. The H2 production Faradaic efficiencies (Eq. 1.37) were calculated using n = 2 and the initial test ro values obtained at steady-state for each respective test. The ro value was taken as the average of the H2 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0 40 80 12 0 16 0 20 0 24 0 28 0 32 0 36 0 40 0 44 0 48 0 52 0 56 0 60 0 Time (min.) H 2 P ro du ct io n R at e (m m ol /m in .) -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 C el l V ol ta ge (V ) a nd W E P ot en tia l ( V v s. S H E ) Pt-Ru/C H production rate Cell voltage WE potential ro r’r 2 -0 .0 5 A -0 .1 A OC V OC V OC V OC V OC V -0 .1 5 A -0 .0 1 A OC V -0 .0 1 A 0 40 80 12 0 16 0 20 0 24 0 28 0 32 0 36 0 40 0 44 0 48 0 52 0 56 0 60 0 H 2 P ro du ct io n R at e (m m ol /m in .) C el l V ol ta ge (V ) a nd W E P ot en tia l ( V v s. S H E ) 209 formation rate over the 30 min. period before the first EPOC galvanostatic current was imposed. Equation 1.32 can be rewritten in terms of production rates as follows: F o r rr −=Λ (4.10) where rF is the Faradaic production rate. The dimensionless EPOC production rate efficiency was evaluated by comparing the promoted catalytic reaction rate to the electrochemical Faradaic production rate as follows: F EPOC r r=η (4.11) As the ro for the different MEAs is not exactly the same, it makes it difficult to compare the promoted H2 production rates, r, for the different test directly. However, as the relative differences between the ro and r, for a particular EPOC test are similar and the efficiencies are calculated on the basis of ro and r values, it is possible to compare the Faradaic efficiencies obtained for the different cases. For all EPOC tests for which negative currents were applied, r was greater than ro, and the difference between r and ro was greater than the theoretical Faradaic H2 production rate calculated based on the current imposed. Typically, r’ exceeded ro, especially for the EPOC tests carried out at currents > -0.01 A (5 mA/cm2), for which the unpromoted H2 production rate, r’, did not have time to return to its original unpromoted H2 production rate, ro, over the 30 min. relaxation period. 210 Table 4.11: Faradaic H2 production rate, measured H2 production rates, Faradaic efficiencies and EPOC efficiencies for the GDE catalytic reforming of CH3OH at low temperature (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC). Applied current Applied current density Faradaic H2 molar flow rate [μmol/s.g catalyst] Measured H2 molar flow rate (Fr) [μmol/s.g catalyst] Faradaic efficiency (Λ) EPOC efficiency (ηEPOC) [μA] [μA/cm2] rF ro r r’ (r-ro)/rF r/rF -1000 -497.51 0.65 9.18 9.24 9.26 -0.08 -14.24 -10000 -4975.12 6.49 9.18 12.73 9.71 -0.55 -1.96 -10000 -4975.12 6.49 9.18 13.01 9.56 -0.59 -2.01 -10000 -4975.12 6.49 9.18 12.91 9.80 -0.58 -1.99 -10000 -4975.12 6.49 9.18 12.99 10.16 -0.59 -2.00 -10000 -4975.12 6.49 7.95 11.22 8.40 -0.50 -1.73 -15000 -7462.69 32.43 7.95 31.42 10.85 -0.72 -0.97 -100000 -49751.24 64.66 7.95 60.37 14.18 -0.81 -0.93 -150000 -74626.87 97.29 7.95 88.45 17.76 -0.83 -0.91 -10000 -4975.12 6.49 7.95 11.77 8.86 -0.59 -1.81 Applying a negative current, i.e., supplying H+ to the WE, resulted in a negative Λ. For all cases listed in Table 4.11, the low temperature catalytic reforming of CH3OH was electrochemically promoted, as the absolute value of Λ was less than unity. The average Faradaic efficiency for all the EPOC tests carried out at -0.01 A (5 mA/cm2) was and -0.56 ±0.03. It is important to note that the Faradaic efficiency became closer to 1, i.e. indicating a Faradaic process, as the imposed current increased. Also at a current of -0.01 A (5 mA/cm2), the average EPOC efficiency was -1.92 ±0.12. The largest EPOC efficiency (-14.24) was obtained at the smallest current applied (-0.001 A or 0.5 mA/cm2). Vayenas and Neophytides [238] have shown that, in general, dehydrogenation or decomposition reactions exhibit negative Λ values for EPOC. The Λ values obtained for the tests carried out in Fig. 4.19 were plotted as 211 a function of the reverse electrolyser cell voltage and the working electrode potential measured during the EPOC experiment as shown in Fig. 4.20. Figure 4.20: Absolute Faradaic efficiency as a function of (a) average cell voltage, and (b) average working electrode potential (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC). As shown in Fig. 4.20 (a), the greatest change in the absolute Faradaic efficiency appears to be at cell voltages < 0.5 V. At greater UWC values, the change in the absolute Λ was not as significant. Non carbon-containing GDE at the reverse electrolyser anode (CE) would be required to evaluate greater EPOC currents (with greater cell voltages, i.e., > 1.23 V) to avoid carbon oxidation. As Fig. 4.20 (b) indicates, decreasing the potential of the working electrode, UWR, enhanced the absolute Faradaic efficiency. (a) (b) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 Average cell voltage, UWC (V) A bs ol ut e Fa ra da ic e ffi ci en cy (d im en si on le ss ) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 -0.3 -0.2 -0.1 0.0 0.1 Average working electrode potential, UWR (V vs. SHE) A bs ol ut e Fa ra da ic e ffi ci en cy (d im en si on le ss ) A bs ol ut e Fa ra da ic e ffi ci en cy (d im en si on le ss ) A bs ol ut e Fa ra da ic e ffi ci en cy (d im en si on le ss ) 212 Figure 4.21 shows that, as is expected, the catalytic H2 production rate, r, depended exponentially on the working electrode potential [238]. The measured promoted catalytic reaction rate increased exponentially with decreasing working electrode potential. When small currents were applied, the measured promoted catalytic H2 production tented to get closer to the unpromoted catalytic H2 production rate. However, promoted catalytic H2 production rate matched closely the absolute Faradaic H2 production rate calculated. It seems like the effect observed is mostly Faradaic and that the increase in the promoted H2 production rate observed can mainly be attributed to the electrochemical reaction. Therefore, applying currents resulting in working electrode potentials more negative than the working electrode potential at OCV only caused a Faradaic enhancement in the H2 production rate. Figure 4.21: Catalytic H2 production rate as a function of average working electrode potential (2.01 cm2, 4 mg/cm2 Pt-Ru/C, 20 wt% Pt-Ru, Pt:Ru 1:1 a/o, 80 ml of 2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 80 ml of 0.5 M H2SO4 electrolyser catholyte, 75 ±1oC). As per Eq. 1.29, the working electrode potential, UWR, is proportional to the catalyst work function, ΦW. Under promoted conditions over the working electrode potential range of Fig. 4.21, the H2 evolution reaction is occurring at the working electrode catalyst surface as shown in Fig. 4.17. It is known that supplying H+, an electropositive adsorbate, to the catalyst surface with a protonic conductor is equivalent to removing an electronegative promoter (O2-) with an oxide conductor, and both decrease the catalyst 0 10 20 30 40 50 60 70 80 90 100 110 -1.0 -0.9 -0.8 -0.7 -0.6 -0.5 Average working electrode potential, UWR (V vs. SHE) H 2 P ro du ct io n R at e (μ m ol /s .g ca ta ly st ) Measured - Unpromoted Measured - Promoted Absolute Faradaic (OCV) H 2 P ro du ct io n R at e (μ m ol /s .g ca ta ly st ) 213 work function [18]. Under small negative catalyst working electrode potentials, the catalytic reforming of CH3OH appears to have an electrophilic behaviour, for which the catalytic reaction rate increases with decreasing work function (∂r/∂Φ<0). This behaviour was reported for the dehydrogenation of CH3OH on Ag with an oxide conductor at high temperature (> 400oC) and a partial pressure of CH3OH up to 7 kPa, which had a Faradaic efficiency of -25 [18, 138]. It was also reported for the non-redox isomeration of 1-butene (1-C4H8) on Pd using NafionTM as the protonic conductor at 70oC, which had a Λ of -28 [18, 147]. Large Λ values for the removal of O2- from the catalyst surface (supply of H+ for a proton conductor), rather than for the supply of O2-, were reported for EPOC studies on the oxidation of CH3OH on Pt/ZrO at 350-650oC [248]. Hence, the EPOC behaviour observed for the low temperature catalytic reforming of CH3OH appears to be as illustrated in Fig. 4.22. Figure 4.22: Electrophilic behaviour of the catalytic reaction rate as a function catalyst WE potential (Modified from Vayenas et al. [18] with permission from Springer). The possible wireless enhancement in the catalytic activity at OCV could result from an H+ spill-over mechanism between the supported precious metal nanoparticles and the electrolyte similar to the one illustrated in Fig. 1.4. Decreasing the catalyst working electrode work function for electrophilic reactions was reported to strengthen the metal-adsorbate chemisorptive bond of methoxy intermediates and thus weaken the intra-adsorbate C=O bonds, the cleavage of which is usually rate-limiting [137]. In their study, Neophytides and Vayenas reported that the resulting weakening in the intra-adsorbate C-H bonds facilitated the dehydrogenation of CH3OH on Ag using an oxide-conducting electrolyte [138]. They also r Electrophilic behaviour UWR 214 mentioned that, while the application of EPOC enhanced the rate of the surface dehydrogenation and decomposition reaction, it did not affect the rate of CH3OH adsorption. Therefore, a plausible explanation for the wireless improvement in the catalytic activity observed in the 2-compartment reactor is an enhancement in the formation of CO2, which is likely the rate-limiting step of the low temperature CH3OH catalytic reforming mechanism. 4.4 Summary A Pt-Ru/TiO2 catalyst was synthesized and a commercially available Pt-Ru/C catalyst was used, and both were characterized physico-chemically and electrochemically. The Pt-Ru/TiO2 electrocatalytic activity for the electrochemical reforming of CH3OH was lower than that of a commercial Pt-Ru/C catalyst of the same composition, because of the TiO2 support poor conductivity and large particle size. Its catalytic activity for the low temperature catalytic reforming of CH3OH was compared to that of Pt-Ru/C in the powder form, using a MS to record the partial pressure of H2, which was converted to a catalytic H2 production rate. For the same quantity of catalyst and precious metal content, the Pt-Ru/TiO2 catalyst had a greater activity towards the CH3OH catalytic reforming reaction than the Pt-Ru/C catalyst. This was attributed to the different properties of the catalyst support. The normalized results obtained for the Pt-Ru/TiO2 catalyst were comparable to those reported in the literature under similar conditions. This appears to be the first time that low temperature catalytic CH3OH reforming has been conducted on Pt- Ru/C. Catalytic CH3OH reforming tests were also carried out using GDEs under similar conditions, but in a two-compartment cell including a MEA. The normalized H2 catalytic reaction rates obtained in this manner were much greater than the ones obtained for powder catalysis in the single-compartment cell. It appears that the small current or potential which developed through the MEA may have been sufficient to induce a wireless EPOC effect to the working electrode surface. The effect of EPOC was also evaluated while applying an electrochemical stimulus in the galvanostatic mode using a Multistat, while recording the transient effect on the catalytic H2 production rate with the MS. It was determined that the reaction’s kinetic regime influenced the extent of the EPOC enhancement caused by the addition or removal of H+ ions. Under the EPOC conditions studied, the 215 catalytic reforming of CH3OH was Faradaic in nature as the absolute values of the Faradaic efficiencies were close to unity. However, it appears that supplying H+ to the working electrode surface at OCV improved the H2 production rate of the low temperature catalytic CH3OH reforming reaction. This appears to be the first time EPOC experiments have been conducted for a catalytic reaction using liquid reactants. This Chapter has demonstrated that the effect of EPOC could play an important role in improving the catalyst performance and the efficiency for the low temperature catalytic reforming of CH3OH for H2 production. 216 Chapter 5: Conclusions The main goal of the research presented in this thesis was to contribute to the advancement of electrochemical reforming of CH3OH (and H2O) for H2 production. Two different approaches for low temperature H2 production were investigated: the electrochemical promotion of electrocatalysis and the electrochemical promotion of catalysis. There approaches are different from other research efforts which aim at maximizing the catalyst surface area and performance, and at developing new catalytic materials with improved properties. The following conclusions were drawn from the work carried-out in Chapters 2 to 4 and summarize the major contributions of the thesis. In addition, potential applications are discussed, the significance and impact of the research work reported are highlighted, and recommendations for future work are provided. 5.1 Electrocatalysis baseline study The first thesis objective was to evaluate the performance of a PEM CH3OH electrolyser and compare it with a PEM water electrolyser under the same conditions. This work was carried out in Chapter 2. It provided a baseline for the subsequent thesis Chapters, and will serve as a ground work for future organic fuel depolarized PEM water electrolysis research. • A theoretical thermodynamic evaluation demonstrated that the CH3OH depolarized electrolysis or electrochemical reforming becomes thermodynamically-favourable at a temperature of 41oC, while water electrolysis does not become thermodynamically favourable over the temperature range studied. • GDE characterization by H2 adsorption and UPD showed that the catalyst ESCA decreased with increasing usage and carbon oxidation. • IR corrected polarization curves were obtained for various CH3OH concentrations and different anode electrocatalysts. At high geometric current densities, lower overpotentials were obtained for 217 greater CH3OH concentrations. The largest limiting geometric current density for CH3OH electrolysis was obtained with Pt-Ru/C, while that for water electrolysis was obtained with a well-defined Pt disk. • AC impedance measurements confirmed that the MEA resistance increased with increasing CH3OH concentration, MEA usage, and MEA oxidation. • The anode and cathode potentials were measured and reported separately for acidic aqueous CH3OH electrolysis for the first time. The change in cathode potential was insignificant compared to the change in the anode potential. • The performance of Pt/C and Pt black catalyst was compared on an ECSA basis. Carbon oxidation started at a lower real current density on Pt/C than for Pt black, suggesting that the carbon of the supported catalyst start to oxidize before the carbon of the CFP. • It was demonstrated that the CH3OH electro-oxidation should be conducted at conditions that avoid water electrolysis and other undesirable secondary electrochemical reactions. In acidic aqueous media and at ambient conditions, carbon oxidation can cause a premature catalyst degradation which prevents the DMR from sustaining high current densities. • Operating the PEM electrolyser with a dry cathode, purged with N2 resulted in lower cell voltage requirements and may be a practical method of operation provided there are no durability issues. • Increasing the temperature improved the electrolyser performance for all electrocatalysts studied. There was no significant change in the limiting current density between 50 and 75oC. • Tafel parameters were determined for various CH3OH concentrations, catalysts and temperature. In general, Pt-Ru/C is more favourable for the electrolysis of CH3OH and water than Pt/C. However, as Tafel kinetic analysis indicated, this may depend on the electrolyser operating conditions. The Tafel slopes obtained at ηa < 0.6 V vs. SHE and CH3OH ≤ 2 M on 2 mg/cm2 Pt black were in agreement with values reported in the literature. Other comparisons with values available in the literature were difficult due to differences in the experimental conditions and whether geometric or real surface area was used. • The apparent activation energy was estimated for Pt-Ru/C. It increased with increasing CH3OH concentration and indicated that CO surface diffusion was the rate determining step for the electrolysis of 2 and 16 M CH3OH. 218 Based on these results, CH3OH depolarized PEM water electrolysis represents an energy efficient alternative to PEM water electrolysis. The remaining part of the thesis focused on the application of electrochemical promotion. 5.2 Electrochemical promotion of electrocatalysis (EPOE) The second thesis objective was to study the effect of EPOE on the PEM water electrolyser, and on the CH3OH depolarized PEM water electrolyser or electrochemical CH3OH reformer. The effect of EPOE was examined as an advancement of the PEM electrolysis systems evaluated in Chapter 2. This work is reported in Chapter 3. The research presented led to a variety of findings regarding the application of EPOE via triode and tetrode operation and the mechanism behind the enhancement. 5.2.1 Galvanostatic and potentiostatic characterization • It was determined that triode and tetrode effects are only obtained when the electrolyser and the auxiliary circuits were operating in reverse and not when they were operating in parallel. In reverse operation, the system comprises two electrolysers working in different directions. • It was confirmed that carbon corrosion of the electrolyser working electrode was not an issue under the system’s operating conditions, and was not responsible for the enhancement observed. • Triode and tetrode operation effects were observed in the galvanostatic and potentiostatic mode with working electrode rings made of noble electrocatalysts, as well as non-catalytic materials like carbon. • A triode and tetrode effect was obtained independently of the electrolyser working electrode material tested, and of the electrolysis process being tested, i.e., water electrolysis or CH3OH electrolysis. • Normal and inverted operation galvanostatic tests indicated that the current line distribution on the WE plane affected the auxiliary circuit cell potential. • There is a capacitive effect when using a noble electrocatalyst at the electrolyser working electrode. • Triode or tetrode operation may be use as a measure to prevent electrode damage from carbon oxidation or other material corrosion. 219 • The electrolyser triode or tetrode voltage ratio, power enhancement ratio and power gain ratio were determined in galvanostatic control. Triode or tetrode operation allowed electrolysis operation at voltages lower than the reversible cell voltage, which is not possible in normal electrolysis. The power enhancement ratio and the power gain ratio showed that the power required to operate in triode or tetrode was less than the power required to operate in normal electrolysis. Under certain operating conditions, the reduction in the power requirements in triode or tetrode electrolysis can exceed that consumed in the auxiliary circuit. • Electrolyser triode or tetrode current gain ratio, rate enhancement ratio, power gain ratio and Faradaic efficiency were measured in potentiostatic control. The rate enhancement ratios were large and reached a maximum which occurred at different auxiliary working electrode potential depending on the electrolyser working electrode surface. Operation at the maximum rate enhancement ratio was stable and durable. The power gain ratios indicated that the power required to operate in triode or tetrode electrolysis was less than that required to operate in normal electrolysis. The Faradaic efficiencies were close to unity indicating that the enhancement is not non-Faradaic in nature, but electrocatalytic. • The current or potential of the electrolyser circuit in triode or tetrode operation exceeded what is expected according to Faraday’s law for normal electrolysis. 5.2.2 Durability investigations • Tetrode operation allows for mixed independent control of the electrolyser and auxiliary circuits, i.e., one circuit in galvanostatic mode and the other in the potentiostatic mode. • Long-term triode and tetrode tests shown that the enhancement effect is stable, durable, and reversible. • The change in CH3OH concentration over time during normal electrolysis was consistent with Faraday’s law. The change in CH3OH concentration over time in triode and tetrode operation exceeded what was predicted from Faraday’s law for normal electrolysis and was consistent with the electrochemical parameters measured. 220 5.2.3 Mechanism investigations • The auxiliary WEaux needs to be in direct physical contact with the membrane to observe a triode or tetrode effect. In tetrode operation, a weaker effect can be obtained even if the independent auxiliary CEaux is not in direct contact with the membrane surface. The triode or tetrode effect is related to the generation of H+ rather than H2 generation. • An increase in the intrinsic membrane conductivity or an increase in the localized H+ concentration of the electrolyte in the vicinity of the WE electrode provide no effect or a small effect, and does not account the triode or tetrode effects observed. • The membrane conductivity does not improve due to the H+ flux produced by the auxiliary circuit and the beneficial effect observed for triode or tetrode operation is limited to the electrode surface. • There appear to be a linear relationship between the auxiliary current/potential and the electrolyser current/potential, as more electrolyser current/potential requires more auxiliary current/potential stimulation to attain an EPOE enhancement. • Larger triode and tetrode effects are observed when the gap between the electrolyser ring WEelec and the auxiliary WEaux is reduced. • Intersecting proton flux lines by inverting the CEs or inverting the WEs in tetrode operation resulted in an insignificant triode or tetrode effect. • Increasing the surface area of the electrolyser ring CFP did not significantly improve the triode effect obtained. • The on-set potential of non-catalytic and non-carbon containing materials, e.g., Au plated 316 SS WEelec and WEaux components were determined and compared with that of noble metal and carbon- containing materials studied in triode and tetrode operation. • It was demonstrated that the effect of triode or tetrode operation can be successfully obtained using non-precious and carbon-free electrolyser and auxiliary WEs material combinations. In triode or tetrode operation, the only catalyst required to carry out the triode electrolysis is then the 2 mg/cm2 Pt/C at the common electrolyser CE. 221 • With a Pt WEaux, the rate enhancement ratio was greater when using a CFP WEelec ring than when using an Au plated 316 SS mesh, due to the differences in the WEelec material surface properties, conductivity, porosity, and geometry. • The power gain ratio values obtained indicated that, under certain operating conditions, less power is needed to operate in triode or tetrode electrolysis than in normal electrolysis. • The Faradaic efficiency ratio indicated that, under the operating conditions studied, triode or tetrode electrolysis with precious or non-precious WEs is Faradaic and electrocatalytic in nature. • Like for the case of carbon containing precious metal WEs, the current or potential of the electrolyser circuit in triode or tetrode operation for the carbon-free non-precious WEs exceeded what is expected according to Faraday’s law for normal electrolysis. • It was proposed that the triode and tetrode effects observed are due to a mechanism which influences the electrolysis charge transfer rate at the WEelec, by the in-situ migration of mobile promoting H+ ions from the WEaux to the WEelec which is driven by surface charge differences. • The triode and tetrode EPOE effect is analogous to the EPOC effect which relies on the backspillover of promoting ions from the support to the metal catalyst surface, except that in this case, the promoted reaction is an electrochemical reaction instead of a chemical one. These findings add to the current knowledge of the mechanism behind EPOE. These contributions may serve as a basis for the development of new circuit designs and WE configurations for controlling the potential or current of an electrode surface, and improving the efficiency of electrocatalytic processes. 5.3 Electrochemical promotion of catalysis (EPOC) The third and last objective of this thesis was to determine the effect of EPOC on the catalytic reforming of CH3OH using a commercial fuel cell catalyst, Pt-Ru/C, and a synthesized Pt-Ru/TiO2 catalyst. The overall reaction of CH3OH catalytic reforming is the same as the overall CH3OH electrolysis reaction. To study EPOC, the catalyst/electrocatalyst working electrode material had to be both, a catalyst to the CH3OH catalytic reforming reaction and an electrocatalyst to the CH3OH electrolysis reaction. The application of EPOC was studied for the oxidation of CH3OH at low temperature, using a protonic 222 conductor, and liquid phase reactants for the first time. The work function of the catalyst working electrode surface was affected by applying currents in galvanostatic mode through the use of a PEM electrolyser configuration. The ground work carried out in Chapter 4 is valuable in the evaluation of the application of EPOC in low temperature systems using liquid reactants. 5.3.1 Catalyst synthesis and characterization • A Pt-Ru/TiO2 catalyst was synthesized by an impregnation method with calcination and reduction in H2 because it was not available commercially. • The synthesized Pt-Ru/TiO2 catalyst was characterized physicochemically by SEM, EDX or EDS, XPS or ESCA, TEM, XRD and BET, and electrochemically by cyclic voltammetry, and compared to commercially available Pt-Ru/C catalyst. 5.3.2 Powder catalysis • Hydrogen was catalytically produced by the low temperature reforming of 2 M CH3OH using Pt-Ru/C or Pt-Ru/TiO2 catalysts powders. • H2 was produced at low levels even at temperatures below 41oC, temperature at which the catalytic reforming of CH3OH in the liquid phase becomes thermodynamically spontaneous for an equimolar CH3OH - water solution. • The H2 production rate did not increase significantly from ambient temperature to 75oC. Once 75oC was reached, the H2 production rate increased slightly as run time increased. • At 75oC, the average H2 production rate for Pt-Ru/TiO2 was about 36% greater for Pt-Ru/C under the same experimental conditions. The larger H2 production rate observed with the Pt-Ru/TiO2 catalyst may be linked to strong metal-support interactions (SMSI). • Due to the greater precious metal content and larger catalyst to solution volume ratio, the H2 generation rates obtained in this study were normalized for comparison with values reported in the literature, which were the similar to the ones reported in this thesis. 223 5.3.3 GDE catalysis and electrochemical promotion of catalysis • The application of EPOC was studied for the oxidation of CH3OH at low temperature for the first time. The work function of the catalyst working electrode surface was affected by applying currents in galvanostatic mode through the use of a PEM electrolyser double-compartment cell configuration with an MEA. • Under the chemical reaction conditions studied and using a GDE catalyst working electrode, it took about 300 min. for the catalytic reaction to reach steady-state. • The normalized catalytic H2 production rate measured at open circuit with the GDE exceeded that obtained for the powder catalysis test in the single-compartment cell. It appears that the small current/potential gradient which developed in the MEA may have been sufficient to induce a wireless EPOC effect on the catalytic H2 production rate. Similar effects have been previously observed in the literature [132]. • The absolute Faradaic efficiency exceeded unity for all currents applied, indicating that the process was non-Faradaic. Supplying H+ to the catalyst working electrode surface resulted in a non-Faradaic catalytic H2 rate enhancement for the low temperature catalytic reforming of CH3OH. It was demonstrated that EPOC can be applied to heterogeneous reactions using liquid reactants at low temperature for the first time. • Catalytic rate enhancement was increasingly significant as the catalyst work function was decreasing. The low temperature catalytic reforming of CH3OH on Pt-Ru/C appears to exhibit an electrophilic behaviour. • The predominant mechanistic effect explaining the rate enhancement when H+ is supplied to the catalyst working electrode is likely due to an enhancement in the formation of CO2, which is likely the rate-limiting step of the low temperature CH3OH catalytic reforming mechanism. 224 5.4 Potential applications, significance and impact 5.4.1 Potential applications Some applications for CH3OH depolarized PEM electrolysers were discussed in Section 1.3.3.2 and the development of practical reactors for the application of EP for various catalytic reactions was reviewed in Section 1.3.4.5. The applications of electrochemically promoted CH3OH depolarized PEM electrolysers will be similar to the ones discussed for normal CH3OH depolarized PEM electrolysers. Likewise, the applications for electrochemically promoted or normal low temperature CH3OH catalytic reformers will also be similar to the applications discussed for electrochemically promoted or normal CH3OH depolarized PEM electrolysers. Just like for the normal CH3OH depolarized PEM electrolyser, triode or tetrode CH3OH depolarized PEM electrolysis, and EP low temperature CH3OH catalytic reforming would be adaptable to small or large H2 production processes. They could be developed into portable fuelling modules producing H2 for combustion vehicles or fuel cell vehicles (FCV), or for PEMFCs and micro-fuel cells (e.g., Angstrom Power Inc.) powering small electronic devices. They could also be developed into a stationary large- scale H2 fuelling station, to be part of off-grid electricity generator systems for back-up and remote power applications, or into off-peak H2 storage devices used for electricity production during high grid loads or blackouts. Nevertheless, based on the work carried-out in this thesis, it is still too early to establish which of these advanced electrochemical H2 production options would be the most advantageous for each of these practical applications. Both low temperature H2 production processes studied, i.e., the CH3OH depolarized water PEM electrolysis and the catalytic reforming of CH3OH, have potential advantages over conventional H2 production methods such as steam reforming, as they both can result in reduced energy consumption. Furthermore, it was shown that the efficiency of these low temperature H2 production processes may be improved by applying the principles of EP. Usually, improving efficiency directly results in reduced energy losses, hence, greater environmental benefits. It was also demonstrated that the application of 225 EPOE and EPOC may result in cost reductions, such as the use of non-precious metals, which may lead to more affordable H2 production. EPOE can reduce the cost and operating energy requirements, as well as increase the efficiency of electrolytic H2 production processes. It was shown that under certain EPOE conditions attained via triode or tetrode operation, the electrolyzer performance improved with a decrease in energy required, i.e., less current/potential was needed in the electrolysis circuit when the auxiliary circuit was in operation. Triode and tetrode operation permitted the use of non-precious and carbon-free electrolyzer and auxiliary working electrode material combinations, limiting the need for precious metal electro-catalyst to the counter electrode(s) only. While material degradation in acidic media imposed some limitations on the current/potential operating range of the triode and tetrode PEM electrolyser, it may be possible to operate the triode or tetrode electrolyser over a wider range of currents/potentials with suitable non- precious carbon-free working electrode materials. The fact that EPOE can be obtained using a variety of WEelec materials, and that the auxiliary WEaux H2 production can be purely ionic and does not require being catalytic adds flexibility to the overall system design. Scaling-up the triode or tetrode operation approach to meet the requirements of large commercial electrolycatalytic applications could potentially be achieved by new electrolyser and auxiliary WEs designs. These new WE designs and configurations could be easily interchanged with the WEs currently used in commercial PEM water electrolysers, without requiring significant modifications. Figure 5.1 shows an example of a working electrode design that could be scaled-up for triode or tetrode operation. 226 Figure 5.1: Example of working electrode scale-up design for triode or tetrode operation (a) single auxiliary working electrode, (b) multiple auxiliary working electrodes. Moreover, the tetrode configuration may offer additional advantages over the triode configuration. It allows for the independent mixed control of the electrolyser and of the auxiliary circuits, results in more concentrated auxiliary current lines, and does not require the auxiliary counter electrode to be in contact with the membrane. This configuration may open the door to the development of a novel electrolyser or fuel cell architectures, which have not been considered before. It may also allow more opportunities for the synergetic juxtaposition of various electrochemical systems, such as the one presented in Fig. 1.1. In some cases, triode and tetrode electrolysis could effectively cogenerate chemical energy and electrical energy by itself, or in various system combinations, which may include a H2 PEMFC, as long as the amount of energy required for H2 generation does not exceed the energy resulting from using the H2 produced. An example of this is shown in Fig. 5.2: Auxiliary WEaux Electrolyser WEelec ring N-117 (a) (b) 227 Figure 5.2: Diagram showing possible electrolysis circuit current reversal under triode or tetrode operation. Electrochemical promotion has been applied to an aqueous phase catalytic reaction using liquid phase reactants for the first time. Much work remains to be accomplished to confirm that the application of EPOC has benefits on the low temperature catalytic reforming of CH3OH. However, the ability to use a commercially available catalyst such as Pt-Ru/C, instead of a synthesized Pt-Ru/TiO2 catalyst to carry out the low temperature catalytic CH3OH reforming reaction will reduce the process’s capital cost. Generating H2 from the catalytic reforming of CH3OH at low temperature requires only a very small electric-power consumption rate to electrochemically promote the catalytic reaction. EPOC may improve the low temperature CH3OH catalytic reforming system start-up time and result in a better response to H2 requirements. As seen in Section 1.3.4.5, EPOC reactors with simple electrical connections and using inexpensive materials are being developed and attempts to scale-up attempts are being undertaken. A scalable high temperature single chamber membrane electrochemically promoted reactor (MEPR) design is underway. However, the EP catalytic reaction studied in this thesis was carried out in a low temperature dual chamber reactor, which may be more difficult to scale-up than the proposed single chamber MEPR. The successful development and manufacturing of a single or dual chamber MEPR may make the application of EPOC to the low temperature catalytic reforming of CH3OH commercially possible. Auxiliary Circuit Electrolyser Circuit e- e - H2 228 The application of EPOC may be transferrable to other electrochemical and chemical reactions. To the author’s knowledge, there are currently no commercial applications using the benefits of electrochemical promotion for electrocatalysis or catalysis. 5.4.2 Significance and impact Sustainable energy is of growing importance due to a combination of environmental and socio- economical issues. As the demand for energy keeps increasing, finding ways to provide a sustainable supply of clean energy to guarantee ecological safety without resulting in a negative impact on society is becoming more and more crucial. Efficient H2 generation from renewable sources is a key energy path to the future. A H2 economy is a compelling solution to lower pollution GHG climate change, increase energy security, and sustainability. It encompasses efficient electrochemical methods for the transformation of renewable sources to H2, as well as efficient energy production through fuel cells. Even though many political, societal, and technological issues still remain to be resolved before H2 becomes the sustainable fuel of the future, it is anticipated that there will be a growing demand for H2 production in the interim. This will gradually become more evident as gasoline prices continue to rise, and society starts to recognize the long term benefits of a reduced fossil fuel dependency for a cleaner environment. Research efforts on the practical application of new fuelling technologies in the market place and closure of existing technology gaps, such as H2 production, transportation and storage, will help to overcome the remaining issues and facilitate the introduction of alternative fuels for fuel cells and other applications to the general public. This thesis brought together core competencies from various areas of chemical engineering, which can be directly involved in alternative fuel, H2 production, and fuel cell research. The experiments conducted required the use of a variety of chemical, electrochemical and material characterization techniques. As the knowledge in the field of EPOC is still limited, the work carried out in this thesis has significantly increased the understanding of the application of this new approach to low temperature electrochemical and chemical systems for H2 production. The technological development of efficient low temperature H2 production may have a significant impact on the energy sector including the transportation sector. The EP of low temperature electrocatalytic or 229 catalytic reforming of CH3OH may provide a way to surpass the Department of Energy (DOE) 2012 technical target of 69% electrolyser efficiency, and 66.2% H2 generation system efficiency for a 1500kg/day water electrolysis refuelling station, while reducing the capital cost and operating cost of the electrolyser. Efficient low temperature H2 production processes could provide H2 on-demand and eliminate the need to transport and store H2 in the liquid or gaseous phase. This research complements the existing high level of hydrogen and fuel cell research in Canada and may help the development of a broader research outlook in automotive fuel cells and other fuel cell applications. Although the operating costs of electrochemical processes are affordable where electricity prices are low, the capital cost needs to be reduced and the operating efficiency needs to be increased. If the cost of precious metals keeps increasing, the precious metal loadings presently required to meet the needed catalytic activity for electrochemical reactions may eventually become impractical for commercialization. In this regards, electrochemical promotion may represent a viable option to achieve material cost reduction while maintaining or enhancing the performance of conventional electrochemical or chemical systems. The principles of EP are versatile since they can be applied to a variety of electrochemical and chemical processes. Thus, there may well be many future opportunities for electrochemical promotion technology transfer. In addition, the electrochemical promotion processes studied in this thesis can be applied to different fuels. This fuel flexibility is a socio-economic advantage as there are a wide variety of fuels available throughout the world. Hence, the field of electrochemical promotion could play a role in technology development in rural and urban areas, as well as in developing and industrialized countries. Fuel Cell Energy Inc., Idatech LLC, Methanex Corporation, and the General Hydrogen Corporation are only a few examples of existing businesses which might be interested in the commercialization of technologies using electrochemical promotion in North America and abroad. This thesis has demonstrated some of the potential benefits that advanced electrochemical processes may have in a H2 or CH3OH economy. It was shown that H2 production from renewable fuels, such as CH3OH, may be a viable option in some sustainable energy systems of the future. More importantly, this research work suggests that EPOE and EPOC may represent a cost-effective solution to tap into the 230 hidden capacity of existing electrochemical or chemical processes, respectively. Therefore this approach may play an important role in the development of a new energy era. 5.5 Recommendations for further research Collectively, this thesis’s contributions will be extremely valuable for further studies on PEM electrolysis, as well as on the application of EP to electrocatalysis and catalysis at low temperature. During the completion of this work, different aspects of the research were identified as requiring more investigation. The proposed research directions mainly concern the optimization of the approach taken and supplemental experiments, or refinement of the experimental methods. These recommendations will help address remaining technical challenges. They may open the door to new research areas, as well as bring the application of the principles of EP of electrocatalysis or catalysis closer to commercialization. 5.5.1 Electrocatalysis baseline study • To mitigate carbon corrosion issues, anode catalysts supported on a non-carbon containing material, such as TiO2, should be investigated. However, the electrical conductivity of the TiO2 support needs to be increased and a possible way of doing this is by adding a metal, such as Nb, to the TiO2 support. • The polarization curves for CH3OH and water electrolysis using Pt-Ru/TiO2 and Pt-Ru/Nb-TiO2 GDEs, as well as using fine porous Pt or Nb meshes of known surface areas, should be established. When the support contains TiO2, the effect of applying UV light during electrolysis could be evaluated. • The polarization effect of using a different carbon-containing PEM water electrolysis depolarization fuel, such as ethanol, could be investigated. Similarly, the effect of using no supporting electrolyte, or of using a different acid than H2SO4 could be evaluated. • Although acidic PEM are generally more robust, it may also be useful to evaluate depolarized PEM electrolysis in alkaline media as no CO2 would be generated, and cheaper materials could be used. While water electrolysis in alkaline media is a well established process, other processes, such as 231 CH3OH [110] or ammonia electrolysis, could also be evaluated in an alkaline environment. The application of EPOE could then be evaluated in alkaline media for the first time. • Experiments with humidified N2 in the cathode compartment should be carried out in the static and flowing mode. Long-term stability tests using a dried or humidified N2 purged cathode compartment should also be carried out. This would further confirm the practicality of the proposed PEM electrolyser design simplification. • The effect of temperature and activation energy should be determined for all electrocatalyst investigated to find out which catalyst results in the lowest activation energy for the electro-oxidation of CH3OH. • Different polymeric membrane materials, e.g. PBI, could be used to evaluate the PEM electrolyser at higher temperatures. The CH3OH and CO2 crossover rates should be measured and the effect of polymeric membrane thickness and composition should be evaluated as a measure to decrease fuel losses. • The CH3OH depolarized PEM water electrolysis should be evaluated in a flowing system using a carrier gas in the anolyte and/or catholyte cell compartment to allow for the determination of the CO2 and/or H2 gas evolution rates. 5.5.2 Electrochemical promotion of electrocatalysis (EPOE) • The EP study should be expanded to evaluate the effect of CH3OH concentration (16 M CH3OH) and temperature (50 and 75oC) on the electro-oxidation of CH3OH in triode and tetrode operation. Triode or tetrode operation could be evaluated using ring WEelec composed of TiO2, Nb-TiO2, Pt-Ru/TiO2 or Pt-Ru/Nb-TiO2. The effect of applying UV light on the electrolyser working electrode containing TiO2 could be evaluated during triode or tetrode operation. • It may be insightful to evaluate the Tafel kinetics data from polarization curves obtained in triode and tetrode operation, and compare them with that obtained for normal electrolysis operation as shown in Chapter 2. Similarly, kinetic data as a function of temperature could be used to determine the activation energy in triode or tetrode operation, and compare with that obtained for the same electrolyser working electrode catalyst material in normal electrolysis operation as in Chapter 2. 232 • The effect of the auxiliary ECSA should be investigated by platinising the Pt WEaux tip or using a Pt black or Pt/C GDE. The use of WE materials resistant to degradation, such as electrochemically etched Nb or Ti meshes for the WEelec rings and an Nb rod for the WEaux, should be also be evaluated to widen the triode or tetrode electrolyser operating potential or current range. • In triode or tetrode electrolysis, the effect of activating the auxiliary circuit only should be evaluated on the electrolyser circuit, when a non-catalytic material is used as the WEelec. To do this, the electrolyser circuit would be left at OCV while the current or potential of the auxiliary circuit would be varied in increments, and its effect on WEelec potential would be monitored as well as the CH3OH concentration in the electrolyser anolyte. This test could clarify the direction of the spillover H+ between the WEs. • It would be helpful to use a reference electrode in galvanostatic control as well as in potentiostatic control to track the individual electrolyser and auxiliary electrode potentials in triode or tetrode operation. In some cases, this may require the use of two reference electrodes, one in each cell compartments. For this, the glass cell caps may need to be modified. Similarly, tracking the local pH in the electrolyser working electrode compartment would be useful. • The effect of EPOE in triode or tetrode operation could be applied to the electrolyser cathode rather than to the electrolyser anode could be investigated in triode or tetrode operation. It should also be more thoroughly verified if the tetrode independent circuit configuration offers operating advantages over the triode configuration, such has having concentrated current lines (tetrode) vs. spread out current lines (triode). • Other materials which just forces protons through the membrane but with no surface electrochemical reaction, i.e., no H2 is produced, should be evaluated for triode and tetrode operation. • The MEA holder and WEelec current collector design should be modified to allow for multiple concentric electrolyser WE rings connections and/or multiple auxiliary WEaux connections. • A carrier gas could be used to continuously monitor the H2 gas being produced with an MS or micro- GC in the WE or CE compartment during triode or tetrode operation. The gas production rates (especially H2) obtained in triode or tetrode operation in the WE compartment as well as in the CE compartment should be compared to that obtained in normal electrolysis, and with that predicted by 233 Faraday’s law under the same normal electrolysis operating conditions. This may help further clarify the proposed enhancement mechanism. • In tetrode electrolysis, the PEM should be replaced by a non-ion conducting material, such as capton, to further confirm if the H+ generated at the auxiliary WEaux are going through the PEM or are just produced at the WEaux surface. Similarly, both circuits could have their own membranes: the electrolyser circuit PEM would be a ring and the auxiliary PEM a disk having the same dimensions as the auxiliary electrode tip. The respective circuit membranes could then be alternatively replaced with capton to determine the path of the H+ species. • The electrolyser WE ring surface could be partially masked by ring of capton of different sizes to evaluate the extent of H+ spillover species when using a non-catalytic material as the WEelec in triode or tetrode operation. • A WEelec current or potential mapping techniques could be developed to evaluate the distribution of the WEaux spillover H+ on the WEelec surface. This may be useful in the evaluation of localized gradients around the WE surfaces during triode or tetrode operation and in the evaluation of new designs. For example, modeling of equipotential lines could be conducted using the COMSOL Inc. multiphysics software. • Different electrolyser WEelec and auxiliary WEaux geometries and alternative configurations should be investigated. The fact that the electrolyser cathode does not need to be in contact with the membrane in tetrode operation could be exploited in new designs. • A new WE scaled-up configuration with multiple auxiliary electrodes, such as the one suggested in Fig. 5.1 should be developed. Ideally, it should be easy to manufacture and be interchangeable with the WE currently used in commercial PEM electrolysers. • A PEM electrolyser WE scaled-up design should be evaluated in a state-of-the-art commercial PEM electrolyser in the active mode. Its response to varying H2 demand should be evaluated in normal electrolysis, as well as in triode and tetrode electrolysis operation. Tests should be performed under extended time periods under different operating cycles to understand the long-term effects of using EPOE on the scaled-up PEM electrolyser. Assess if the CH3OH depolarized PEM water electrolyser and PEM water electrolyser would meet or exceed the 2011 DOE electrolyser technical targets in normal electrolysis, and in triode or tetrode operation. 234 • Active systems comprising combinations of PEMFCs, triode or tetrode PEM electrolysers, and triode or tetrode electrochemical synthesis, such as the one described in Fig. 1.1, should be evaluated though modeling studies. 5.5.3 Electrochemical promotion of catalysis (EPOC) • A catalytic reforming test should be carried out by using a non-conductive material to hang a GDE catalyst coupon in the solution in the single-compartment catalytic glass cell reactor used for the powder catalysis test to determine and verify the true ro value. The resulting H2 production rate should then be compared to that obtained for the test carried out in the 2-compartment reactor using a MEA when no external current or potential is applied. This will confirm if a wireless EPOC effect is induced by a gradient developing between the catalyst working electrode and the counter electrode due to the catalytic reaction occurring at the catalyst working electrode. If this is the case, the catalytic rate determined when the 2-compartment cell at OCV is in fact a promoted r value, and the internal current developing in the 2-compartment cell at open circuit voltage should be determined to allow for the calculation of the Faradaic efficiency for this wireless EPOC case. • Longer relaxation time periods between each promoted tests segments should be evaluated to ensure the H2 production rate eventually returns to its equilibrium value between each EPOC test. • The final CH3OH concentration of the un-reacted solution should be measured after the GDE catalysis experiments and compared to that after long-term EPOC experiments tests. It should also be compared with the CH3OH consumption rate calculated based on the measured H2 production rates during the same experiments. • The effect of H2SO4 concentration (pH) on the powder catalysis, and unpromoted and promoted GDE catalysis reaction rate should be evaluated. Similarly, various concentrations of CH3OH should be evaluated. • The catalytic reaction rate should be evaluated at different temperatures under unpromoted (powder and GDE) and promoted conditions (GDE). This would allow for the determination and comparison of the activation energy under various unpromoted and promoted conditions. • The effect of applying UV light could be explored when using a Pt-Ru/TiO2 GDE. 235 • It should be determined if the application EPOC had an effect on the catalytic reaction order. For this, the catalytic reaction order, the reaction rate constant and the rate determining step would need to be established. • An investigation should be carried out to see if the low temperature CH3OH reforming EPOC experiment could be scaled up to a single or dual chamber membrane electrochemically promoted reactor (MEPR). • The synergetic effects of EPOE (triode or tetrode operation) and EPOC should be further investigated. As mentioned in Section 3.5.3, in triode or tetrode operation, it may be possible to carry out the low temperature catalytic reforming of CH3OH using non-precious materials at the electrolyser ring WEelec. The high level H2 production process efficiency and economic comparison carried out in Appendix C should be expanded to include the low temperature electrocatalytic and catalytic reforming of CH3OH and the effect of electrochemical promotion on the different processes. 236 References [1] S. Sorrella, J. Speirsb, R. Bentleyc, A. Brandt, and R. Millere, “Global oil depletion: a review of the evidence, energy policy”, 38(9), (2010) 5290-5295. [2] D. Hart, “Hydrogen - a truly sustainable transport fuel?”, Front. Ecol. Environ., 1(3), (2003)138-145. [3] D. Sperling, and N. Lutsey, “Energy efficiency in passenger transportation”, The Bridge, 39(2), (2009) 22-30. [4] C. Lamy, J.-M. Leger, and S. Srinivasan, “Direct methanol fuel cells: from a twentieth century electrochemist’s dream to a twenty-first century emerging technology”, Modern aspects of electrochemistry, John O’M. Bockris J. O’M., Conway B. E., White R. E., editors. New York, NY: Kluwer Academic Publishers, 34 (2001) 53-117. [5] B. I. Podlovchenko, and T. D. Gladysheva, “Copper ion-catalysed electrochemical synthesis of methanol from carbon monoxide on palladium electrodes”, Mendeleev Commun., 16(4), (2006) 232-233. [6] J. Li, and G. Prentice, “Electrochemical synthesis of methanol from CO2 in high-pressure electrolyte”, J. Electrochem. Soc., 144(12), (1997) 4284-4288. [7] Z. Hu, M. Wu, Z. Wei, S. Song, and P. K. Shen, “Pt-WC/C as a cathode electrocatalyst for hydrogen production by methanol electrolysis“, J. Power Sources, 166, (2007) 458-461. [8] J. W. Phair, and S. P. S. Badwal, “Review of proton conductors for hydrogen separation”, Ionics, 12, (2006) 103-115. [9] S. Ishikura, and C. F. Maciel, “Steam reforming of renewable feedstock’s for the production of hydrogen”, http://www.ecodesarrollo.cl/descargas/Caio_Ishikura_Diploma_hidrogeno.pdf , (2007) 55 pp. [10] Department of Energy (DOE) decision team committee report. “On-board fuel processing go/no-go decision”, http://www1.eere.energy.gov/hydrogenandfuelcells/news_fuel_processor.html, (2004) 1-7. [11] P.-K. Shen, S.-L. Wang, Z.-L. Hu, Y.-L. Li, R. Zeng, and Y.-Q. Li, “Hydrogen production by alcohol electrolysis“, Acta Phys. - Chim. Sin., 23(1), (2007) 107-110. [12] T.Tetsuo, T. Kazuhiko, and U. Minoru, “Hydrogen production by methanol-water solution electrolysis”, J. Power Sources, 164, (2007) 9–16. [13] N. Meng, T. Yamakawa, and S. Shinoda, “Methanol dehydrogenation in the liquid phase with Ru/activated carbon catalyst”, React. Kinet. Catal. Lett., 58(2), (1996) 341-348. [14] T. Miyao, M. Yamauchi, and S. Naito, “Liquid phase methanol reforming with water over silica supported Pt-Ru catalysts”, Catalysis Today, 87, (2003) 227-235. [15] T. Miyao, Y. Watababe, M. Teramoto, and S. Naito, “Catalytic activity of various supported Ir-Re catalysts for liquid phase methanol reforming with water”, Catalysis Communications, 6, (2003) 113-117. [16] T. Miyao, M. Yamauchi, H. Narita, and S. Naito, “Remarkable support effect for liquid phase methanol reforming with water over supported Pt-Ru catalysts”, Appl. Catal. A: Gen., 299, (2006) 285- 291. [17] J. Qu, X. Zhang, Y. Wang, and C. Xie, “Electrochemical reduction of CO2 on RuO/TiO2 nanotubes composite modified Pt electrode”, Electrochim. Acta, 50(16–17), (2005) 3576-3580. 237 [18] C. G. Vayenas, S. Bebelis, C. Pliangos, S. Brosda, and D. Tsiplakides, “Electrochemical activation of catalysis - promotion, electrochemical promotion and metal-support interactions”, Kluwer Academic / Plenum Publishers, (2001) 574 pp. [19] S. P. Balomenou, and C. G. Vayenas, “Triode fuel cells and batteries”, J. Electrochem. Soc., 151(11), (2004) A1874-A1877. [20] D. P. Wilkinson, “Fuel cells”, The Electrochemical Society Interface, 10(1), (2001) 22-25. [21] S. Leclerc, R. F. Mann, and B. A. Peppley, “Evaluation of the catalytic ethanol-steam reforming process as a source of hydrogen-rich gas for fuel cells”, CANMET Energy Technology Center (CETC), (1998) 37 pp. [22] R. O. Idem, and N. N. Bakhshi, “Production of hydrogen from methanol. 2. Experimental studies”, Industrial Engineering Chemistry Research, 33(9), (1994) 2056-2065. [23] S. R. Segal, K. B. Anderson, K. A. Carrado, and C. L. Marshall, “Low temperature steam reforming of methanol over layered double hydroxide-derived catalysts”, Appl. Cat. A, 231, (2002) 215-226. [24] M. M. Natile and A. Glisenti, “WO3/CeO2/YSZ nanocomposite as a potential catalyst for methanol reforming”, J. Power Sources, 145, (2005) 644-651. [25] R. R. Davda, J. W. Shabaker, G. W. Huber, R. D. Cortright, and J. A. Dumesic, “A review of catalytic issues and process conditions for renewable hydrogen and alkanes by aqueous-phase reforming of oxygenated hydrocarbons over supported metal catalysts”, Applied Catalysis B: Environmental, 56, (2005) 171-186. [26] G. W. Huber, J. W. Shabaker, and J. A. Dumesic, “Raney Ni-Sn catalyst for H2 production from biomass-derived hydrocarbons”, Science, 300, (2003) 2075-2077. [27] J. W. Shabaker, R. R. Davda, G. W. Huber, R. D. Cortright, and J. A. Dumesic, “Aqueous-phase reforming of methanol and ethylene glycol over alumina-supported platinum catalysts”, Journal of Catalysis, 215, (2003) 344-352. [28] G. W. Huber, and J. A. Dumesic, “An overview of aqueous-phase catalytic processes for production of hydrogen and alkanes in a biorefinery”, Catalysis Today, 111, (2006) 119-132. [29] S. Naito, “Market effect of various supports and additives upon liquid phase methanol reforming with water over supported group 8-10 metal catalysts”, Catalysis Surveys from Asia, 9(4), (2006) 243-258. [30] W. Curtis Conner Jr., and J. L. Falconer, “Spillover in heterogeneous catalysis”, Chem. Rev., 95, (1995) 759-788. [31] F. Roessner, and U. Roland, “Hydrogen spillover in bifunctional catalysis”, J. of Molecular Catalysis A: Chemical, 112, (1996) 401-412. [32] U. Roland, T. Braunschweig, and F. Roessner, “On the nature of spilt-over hydrogen”, J. of Molecular Catalysis A: Chemical, 127, (1997) 61-84. [33] H. Chang, and J. Phillips, “Catalytic synergism in physical mixtures on supported iron-cerium and supported noble metal for hydro-isomerization of 1, 3-butadiene”, Langmuir, 13, (1997) 477-482. [34] J. C. Weigle, and J. Phillips, “Novel dual-bed reactors: utilization of hydrogen spillover in reactor design”, Langmuir, 20, (2004) 1189-1193. [35] E. Baumgarten, and I. Niemeyer, “On the role of surfaces in hydrogenation reactions with gas phase spillover hydrogen”, React. Kinet. Catal. Lett., 70(2), (2000) 371-377. 238 [36] E. Baumgarten, and G. Meyer, “Hydrogen spillover through the gas phase. Some kinetic aspects”, React. Kinet. Catal. Lett., 71(2), (2000) 325-333. [37] E. Baumgarten, and L. Maschke, “Hydrogen spillover through the gas phase reaction with graphite and activated carbon”, Applied catalysis A: General, 202, (2000) 171-177. [38] P. C. H. Mitchell, A. J. Ramirez-Cuesta, S. F. Parker, J. Tomkinson, and D. Thompsett, “Hydrogen spillover on carbon-supported metal catalysts studied by inelastic neutron scattering. Surface vibrational states and hydrogen riding modes”, J. Phys. Chem. B, 107, (2003) 6838-6845. [39] B. Miremadi, T. Yadav, J. Z. Zhang, and J. L. Falconer, “Low-temperature hydrogen production using electrically activated catalysts”, Chem. Comm., (2000) 1875-1876. [40] C. K. Narula, A. Arbor, J. H. Visser, and A. A. Adamczyk, Jr., “Electrically heatable catalyst device using electrically conductive non-metallic materials”, US 5480622, (1996). [41] Y. Tapesh, and M. Bijan, “Processes for electrically activated transformation of chemical and material compositions”, WO 02046130, (2002). [42] J. Z. Zhang, M. Still, T. Yadav, and G. T. Gaudet, “Low-temperature hydrogen production from steam reforming of methane using electrically activated catalyst”, ECS Transactions, 5(1), (2007) 673- 676. [43] S. Kundu, L. C. Simon, M. Fowler, and S. Grot, “Mechanical properties of Nafion electrolyte membranes under hydrated conditions”, Polymer, 46(25), (2005) 11707-11715. [44] M. J. Kelly, G. Fafilek, J. O. Besenhard, H. Kronberger, and G. E. Nauer, “Contaminant absorption and conductivity in polymer electrolyte membranes”, J. Power Sources, 145(2), (2005) 249-252. [45] T. Higuchi, T. Tsukamoto, N. Sata, M. Ishigame, K. Kobayashi, S. Yamaguchi, and S. Shin, “Electronic structures of protonic conductors SrTiO3 and SrCeO3 by O 1s x-ray absorption spectroscopy”, Solid State Ionics, 145-155, (2002) 735-739. [46] T. Schober, and H. G. Bohn, “Water vapour solubility and electrochemical characterization of the high temperature proton conductor BaZr0.9Y0.1O2.95”, Solid State Ionics, 127(3-4), (2001) 351-360. [47] N. Bonanos, and M. Mogensen, “H2 oxidation at the interface Ni/Sr0.995Ce0.95Y0.05O2.975”, Solid State Ionics, 97(1-4), (1997) 483-488. [48] H. Iwahara, K. Uchida, K. Ono, and K. Ogaki, “Proton conduction in sintered oxides based on BaCeO3” , J. Electrochem. Soc., 135(2), (1988) 529-533. [49] K. Katahira, Y. Kohchi, T. Shimura, and H. Iwahara, “Protonic conduction in Zr-substituted BaCeO3”, Solid State Ionics, 138(1-2), (2000) 91-98. [50] A. S. owick, and Y. Du, “High-temperature protonic conductors with perovskite-related structures”, Solid State Ionics, 77, (1995) 137-146. [51] J. A. Kosek, G. Jose, and A. B. Laconti, “Method and system for producing high-pressure hydrogen”, WO 2005014468, (2005). [52] R. Lawson, C. Wang, J. Hong, J. Ma, B. Fang, and D. Chu, “Nafion-bimevox composite membrane for fuel cell applications”, J. Electrochem. Soc., 154(1), (2007) B48-B52. [53] I. Honma, S. Nomura, and H. Nakajima, “Protonic conducting organic inorganic nanocomposites for polymer electrolyte membrane”, Journal of Membrane Science, 185, (2001) 83-94. 239 [54] V. Antonucci, A. Di Blasi, V. Baglio, R. Ornelas, F. Matteucci, J. Ledesma-Garcia, L. G. Arriaga, and A. S. Arico, “High temperature operation of a composite membrane-based solid polymer electrolyte water electrolyser”, Electrochim. Acta, 53, (2008) 7350-7356. [55] L. Xiao, H. Zhang, E. Scanlon, L. S. Ramanathan, E.-W. Choe, D. Rogers, T. Apple, and B. C. Benicewicz, “High-temperature polybenzimidazole fuel cell membranes via a sol-gel process”, Chem. Mater., 17, (2005) 5328-5333. [56] L. Quingfeng, H. A., Hjuler, and N. J. Bjerrum, “Phosphoric acid doped polybenzimidazole membranes: physicochemical characterization and fuel cell applications”, J. of Appl. Electrochem., 31, (2001) 773-779. [57] X. Glipa, M. E. Haddad, D. J. Jones, and J. Roziere, “Synthesis and characterization of sulphonated polybenzimidazole: a highly conducting proton exchange polymer”, Solid State Ionics, 97, (1997) 323- 331. [58] N. Ito, S. Aoyama, T. Nakata, M. Lijima, and H. Sato, “Hydrogen-permeable membrane and manufacturing method of the same”, US 7049008, (2006). [59] R. E. Buxbaum, “Hydrogen permeable membrane and hydride composition”, US 6576350, (2003). [60] S. A. Grigoriev, P. Millet, and V. N. Fateev, “Evaluation of carbon-supported Pt and Pd nanoparticles for the hydrogen evolution reaction in PEM water electrolysers”, J. Power Sources, 177, (2008) 281-285. [61] J. Tian, G. Sun, M. Cai, Q. Mao, and Q. Xin, “PtTiOx/C electrocatalysts with improved durability in H2/O2 PEMFCs without external humidification”, J. Electrochem. Soc., 155 (2), (2008) B187-B193. [62] M. Goetz, and H, Wendt, “Composite electrocatalysts for anodic methanol and methanol-reformate oxidation”, J. of Appl. Electrochem., 31, (2001) 811-817. [63] P. Paunovic, I. Radev, A. T. Dimitrov, O. Popovski, E. Lefterova, E. Slavcheva, and S. H. Jordanov, “New nanostructure and interactive supported composite electrocatalysts for hydrogen evolution with partially replaced platinum loading”, Int. J. H2 Energy, 34, (2009) 2866-2873. [64] M. Wilson, and S. Gottesfeld, “High performance catalyzed membranes of ultra-low Pt loadings for polymer electrolyte fuel cells”, J. Electrochem. Soc., 139, (1992) L28-L30. [65] T. Frey, and M. Linardi, “Effects of membrane electrode assembly preparation on the polymer electrolyte membrane fuel cell performance”, Electrochim. Acta, 50, (2004) 99–105. [66] Y. J. Zhang, C. Wang, N. F. Wan, Z. X. Liu, and Z. Q. Mao, “Study on a novel manufacturing process of membrane electrode assemblies for solid polymer electrolyte water electrolysis”, Electrochem. Commun., 9, (2007) 667–670. [67] J. Larminie, and A. Dikes, “Fuel cell system explained”, John Wiley & Sons Ltd., England, (2003) 406pp. [68] S. Gilman, “The mechanism of electrochemical oxidation of carbon monoxide and methanol on platinum. 1. Carbon monoxide adsorption and desorption and simultaneous oxidation of the platinum surface at constant potential”, 67, (1963) 1898-1905. [69] S. Gottesfeld, “Preventing CO poisoning in fuel cells”, US 4910099, (1990). [70] X. Yan, S. Wang, X. Li, M. Hou, Z. Yuan, D. Li, L. Pan, C. Zhang, J. Liu, P. Ming, and B. Yi, “A 75 kW methanol reforming fuel cell system”, J. Power Sources, 162(2), (2006) 1265-1269. [71] J. Zhang, and R. Datta, “Electrochemical preferential oxidation of CO in reformate”, J. Electrochem. Soc., 152(6), (2005) A1180-1187. 240 [72] T. Iwasita, “Fuel cells: spectroscopic studies in the electrocatalysis of alcohol oxidation”, Journal of the Brazilian Chemical Society, 13(4), (2002) 1-19. [73] N. Munichandraiah, K. McGrath, G. K. S. Prakash, R. Aniszfeld, and O. G. A. Robert, “A potentiometric method of monitoring methanol crossover through polymer electrolyte membranes of direct methanol fuel cells”, J. Power Sources, 117(1-2), (2003) 98-101. [74] A. Hamnett, B. J. Kennedy, and S. A. Weeks, “Base metal oxides as promoters for the electrochemical oxidation of methanol”, J. Electroanal. Chem., 240, (1988) 349-353. [75] P. K. Shen, and A. C. C. Tseung, “Anodic oxidation of methanol on Pt/WO3 in acidic media”, J. Electrochem. Soc., 141(11), (1994) 3082-3090. [76] E. Kjeang, J. Goldak, M.R. Golriz, J. Gub, D. James, and K. Kordesch, “A parametric study of methanol crossover in a flowing electrolyte-direct methanol fuel cell”, J. Power Sources, 153, (2006) 89- 99. [77] J. A. Drake, W. Wilson and K. Killeen, “Evaluation of the Experimental Model for Methanol Crossover in DMFCs”, J. Electrochem. Soc., 151(3), (2004) A413-A417. [78] R. Jiang, C. Rong, and D. Chu, “Fuel cross-over and energy conversion in lifetime operation of direct methanol fuel cells”, J. Electrochem. Soc., 154(1), (2007) B13-B19. [79] F. J. Rodriguez-Varela, and O. Savadogo, “Ethanol-tolerant oxygen reduction reaction (ORR) cathodes for direct ethanol fuel cell applications”, The Electrochemical Society Interface, Conference Proceedings, 208th ECS Meeting, 1(6), (2005) 331-338. [80] M. K. Ravikumar, and A. K. Shukla, “Effect of methanol crossover in a liquid-feed polymer-electrolyte direct methanol fuel cell”, J. Electrochem. Soc., 143(8), (1996) 2601-2606. [81] A. Heinzel, and V. M. Barragan, “A review of the state-of-the-art of the methanol crossover in direct methanol fuel cells”, J. Power Sources, 84(1), (1999) 70-74. [82] Y.-L. Ma, J. S. Wainright, M. H. Litt, and R. F. Savinell, “Conductivity of PBI membranes for high- temperature polymer electrolyte fuel cells”, J. Electrochem. Soc., 151(1), (2004) A8-A16. [83] J.-T. Wang, J. S. Wainright, R. F. Savinell, and M. Litt, “A direct methanol fuel cell using acid doped polybenzimidazole as polymer electrolyte”, J. Appl. Electrochem., 26, (1996) 751-756. [84] S. R., Narayan, and T. K. Valdez, “High-energy portable fuel cell power sources”, The Electrochemical Society, Interface, 17(4), (2008) 40-45. [85] S. Ramaswamy, M. Sundaresan, A. Eggert, and R. M. Moore, “System dynamics and efficiency of the fuel processor for an indirect methanol fuel cell vehicle”, 35th Intersociety Energy Conversion Engineering Conference, 24-28, (2000) 2000-3048. [86] K.-H. Hauer, “Dynamic interactions between the electric drive train and fuel cell system for the case of an indirect methanol fuel cell vehicle”, AIAA-2000-3039, American Institute of Aeronautics and Astronautics Inc., (2000) 1317-1325. [87] S. A. Fateen, “Fuel cell systems for personal and portable power applications”, Symposium 2001: Highlighting Student and Postdoctoral Research, ESA-EPE, Los Alamos National Laboratory, (2001) 5 pp. [88] A. Ishihara, S. Mitsushima, N. Kamiya, and K.-I. Ota, “Exergy analysis of polymer electrolyte fuel cell systems using methanol”, J. Power Sources, 126, (2004) 34-40. [89] L. F. Brown, “A comparative study of fuels for on-board hydrogen production for fuel-cell-powered automobiles”, Int. J. H2 Energy, 26, (2001) 381-397. 241 [90] J. R. Lattner, and M. P. Harold, “Comparison of methanol-based fuel processors for PEM fuel cell systems”, Applied Catalysis B: Environmentla 56, (2005) 149-169. [91] S. A. Grigoriev, V. I. Porembsky, and V. N. Fateev, “Pure hydrogen production by PEM electrolysis for hydrogen energy”, Int. J. H2 Energy, 31, (2006) 171-175. [92] E. Rasten, G. Hagen, and R. Tunold, “Electrocatalysis in water electrolysis with solid polymer electrolyte”, Electrochim. Acta, 48, (2003) 3945-3852. [93] D. Stolten (Ed.), “Hydrogen and fuel cells, fundamentals, technologies and applications”, Wiley-VCH Verlag GMBh & Co., Weinbeim, (2010) 908 pp. [94] T. M., Maloney, “An electrolysis-based pathway towards hydrogen fuelling”, Vehicle Power and Propulsion, 2005 IEEE Conference, (2005) 652 - 656. [95] T. M. Maloney, “Vehicle power and propulsion”, IEEE Conference, (2005) 487-491. [96] S. Song, H. Zhang, B. Liu, P. Zhao, Y. Zhang, and B. Yi, “An improved catalyst-coated membrane structure for PEM water electrolyzer”, Electrochemical and Solid-State Letters, 10(8), (2007) B122-B125. [97] A. Fujishima, and K. Honda, “Electrochemical photolysis of water at a semiconductor electrode”, Nature, 238, (1972) 37-38. [98] V. Baglio, R. Ornelas, F. Matteucci, F. Martina, G. Ciccarella, I. Zama, L. G. Arriaga, V. Antonucci, and A. S. Arico, “Solid polymer electrolyte water electrolyser based on Nafion-TiO2 composite membrane for high temperature operation”, Fuel Cells, 9(3), (2009) 247-252. [99] V. Antonucci, A. Di Blasi, V. Baglio, R. Ornelas, F. Matteucci, J. Ledesma-Garcia, L. G. Arriaga, and A. S. Arico, “High temperature operation of a composite membrane-based solid polymer electrolyte water electrolyzer”, Electrochim. Acta, 53, (2008) 7350-7356. [100] Hydrogenics Corporation, HyLYZERTM Annex, “HyLYZERTM 1Nm3/h PEM hydrogen generation technology”, http://www.hydrogenics.com/assets/pdfs/ANNEX%201%20-%20HyLYZER-1.pdf, 7 pp. [101] H-tec Wasserstoff-Energie-Systeme GmbH, “EL 30 electrolysers for distributed hydrogen production”, http://h-tec-electrolyser.com/_downloads/htec_EL30_Electrolyser.pdf, 1 p. [102] A. H. Shaaban, “Pulsed DC and anode depolarization in water electrolysis for hydrogen generation”, Engineering Research Division, HQ Air Force Civil Engineering Support Agency, Final Report, ESL-TR- 92-55, (1994) 69 pp. [103] R.W. Coughlin, and M. Farooque, “Electrochemical gasification of coal – simultaneous production of hydrogen and carbon dioxide by a single reaction involving coal, water and electrons”, Ind. Eng. Chem. Process Des. Dev., 19, (1980) 211-219. [104] M. R. St. John, A. J. Furgala, and A. F. Sammells, “Hydrogen generation from glucose depolarized water electrolysis”, J. Electrochem. Soc., 129, (1986)1174-1981. [105] A. J. Appleby, and B. Pinchon, “Electrochemical aspects of the H2SO4-SO2 thermo-electrochemical cycle for hydrogen production”, Int. J. H2 Energy, 5, (1980) 253-267. [106] J. A. O’Brien, J. T. Hinkley, and S. W. Donne, “The electrochemical oxidation of aqueous sulphur dioxide. I. Experimental parameter influences on electrode behaviour”, J. Electrochem. Soc., 157(9), (2010) F111-F115. [107] C. R. Cloutier, and D. P. Wilkinson, “Electrolytic production of hydrogen from aqueous acidic methanol solutions”, Int J. H2 Energ., 35, (2010) 3967-3984. 242 [108] R. R. Adzic, W. E. O’Grady, and S. Srinivasan, “Oxidation of HCOOH and CH3OH on platinum modified by foreign metal adatoms in 85% phosphoric acid”, J. Electrochem. Soc., 128(9), (1981)1913- 1919. [109] L. Soler, J. Macanas, M. Munoz, and J. Casado, “Electrocatalytic production of hydrogen boosted by organic pollutants and visible light”, Int. J. H2 Energy, 31, (2006) 129-139. [110] B. Reichman, and W. Mays, “Electrolytic production of hydrogen”, US 6890419, (2005). [111] R. Woods, I. Root, R. H. Lyer, and B. F. Porter, “Electrochemical reformer and fuel cell system”, WO 2004046408, (2004). [112] I. Busenbender, V. Peinecke, H. Dohle, and T., Kels, “Process for reforming methanol”, DE 197007384, (1998). [113] K. Teruo, K. Yuichi, H. Tatsuo, and T. Seiji, “Hydrogen production”, JP 63219593, (1988). [114] S. Takayuki, and A. Yoshikawa, “Fuel cell”, JP 2003308869, (2003). [115] T. Toshiki, and F. Toshio, “Organic electrolytic synthesis method and organic electrolytic synthesis device”, JP 2005256033, (2005). [116] S. R. Narayanan, and W. Chun, B. Jeffries-Nakamura and T. I. Valdez, “Hydrogen generation by electrolysis of aqueous organic solutions”, US 6299744, (2001). [117] S. R. Narayanan, W. Chun, B. Jeffries-Nakamura, and T. I. Valdez, “Hydrogen generation by electrolysis of aqueous organic solutions”, US 7056428, (2006). [118] T. Take, K. Tsurutani, and M. Umeda, “Hydrogen production by methanol-water solution electrolysis”, J. Power Sources, 164, (2007) 9-16. [119] J. W. Tester, E. M. Drake, M. W. Golay, and A. P. William, “Sustainable energy choosing among options”. Cambridge, MA: MIT Press, (2005). [120] D. P. Bloomfield, and A. N. Rabe, “Electrochemical autothermal reformer“, US 6143159, (2000). [121] D. P. Bloomfield, “Electrochemical hydrogen compressor with electrochemical autothermal reformer”, US 6068673, (2000). [122] C. G. Vayenas, University of Patras, Department of Chemical Engineering, (2006) http://www.chemeng.upatras.gr/index.htm. [123] M. Stoukides, and C. G. Vayenas, “Solid electrolyte aided study of the ethylene oxide oxidation on Silver”, Journal of Catalysis, 64, (1980) 18-28. [124] C. G. Vayenas, S. Bebelis, and S. Neophytides, “Non-Faradaic electrochemical modification of catalytic activity”, J. Phys. Chem., 92(18), (1988) 5083-5085. [125] N. A. Anastasijevic, “NEMCA - From discovery to technology”, Catalysis Today, 146, (2009) 308- 311. [126] A. J. Bard, and L. R. Faulkner, “Electrochemical Methods Fundamentals and Applications”, Second Edition, John Wiley & Sons, Inc., USA, (2001) 833 pp. [127] C. G. Vayenas, M. M., Jaksic, S. I., Bebelis, and S., G. Neophytides, “The electrochemical activation of catalytic reactions”, Modern Aspects of Electrochemistry, Edited by John O’M. Bockris et al., 29, (1996) 57-202. 243 [128] P. H. Chiang, and M. Stoukides, “Solid electrolyte aided direct coupling of methane”, Journal of Catalysis, 139(2), (1993/2) 683-687. [129] C. Sanchez, and E. Leiva, “Theoretical aspects of some prototypical fuel cell reactions”, Handbook of fuel cells - fundamentals, technology and applications, Volume 2: Electrocatalysis, W. Vielstich, H. A. Gasteiger, and A. Lamm (Ed), John Wiley & Sons, Ltd., England, Chapter 11, (2003) 93-131. [130] E. P. M. Leiva, and C. G. Sanchez, “NEMCA effect: why are the work function changes of the gas exposed catalyst-electrode surface one-to-one related to the changes in the catalyst working electrode potential?”, Solid State Electrochem., 7, (2003) 588-592. [131] H. Matsumoto, T. Shimura, T. Higushi, H. Tanaka, K. Katahira, T. Otake, T. Kudo, K. Yashiro, A. Kaimai, T. Kawada, and J. Mizusaki, “Protonic-electronic mixed conduction and hydrogen permeation in BaCe0.9-xY0.1RuxO3-alpha”, J. Electrochem. Soc.,152(3), (2005) A488-A492. [132] M. Marwood, S. Balomenou, A. Tsiliras, C. A. Cavalca, C. Pliangos, and C. G. Vayenas, “Electrochemical promotion of electronically isolated and dispersed Pt catalysts”, Ionics, 4, (1998) 207- 214. [133] P. E. Tsiakaras, S. L. Douvartzides, A. K. Demin, and V. A. Sobyanin, “The oxidation of ethanol over Pt catalyst-electrodes deposited on ZrO2 (8 mol% Y2O3)”, Solid State Ionics, 152-153, (2002) 721- 726. [134] S. L. Douvartzides, and P. E. Tsiakaras, “Electrochemically promoted catalysis: the case of ethanol oxidation over Pt”, Journal of Catalysis, 211(2), (2002) 521-529. [135] S. G. Neophytides, D. Tsiplakides, P. Stonehart, M. Jasksic, and C. G. Vayenas, “Non-Faradaic electrochemical modification of the catalytic activity of Pt for H2 oxidation in aqueous alkaline media”, J. Phys. Chem., 100, (1996) 14803-14814. [136] C. A. Cavalca, G. Larsen, C. G. Vayenas, and G. L. Haller, “Electrochemical modification of CH3OH oxidation selectivity and activity on a Pt single-pellet catalytic reactor”, J. Phys. Chem., 97, (1993) 6115- 6119. [137] J. K. Hong, I.-H. Oh, S.-A. Hong, and W. Y. Lee, “Electrochemical oxidation of methanol over a silver electrode deposited on Yttria-stabilized zirconia electrolyte”, Journal of Catalysis, 163(1), (1996) 95-105. [138] S. Neophytides, and C. G. Vayenas, “Non-Faradaic electrochemical modification of catalytic activity 2. The case of methanol dehydration and decomposition on Ag”, Journal of Catalysis, 118, (1989) 147- 163. [139] A. Thursfield, S. Brosda, C. Pliangos, T. Schober, and C. G. Vayenas, “Electrochemical promotion of an oxidation reaction using a proton conductor”, Electrochim. Acta, 48(25-26), (2003) 3779-3788. [140] C. G. Yiokari, G. E. Pitselis, D. G. Pollydoros, A. D. Katsanounis, and C. G. Vayenas, “High- pressure electrochemical promotion of ammonia synthesis over an industrial iron catalyst”, J. Phys. Chem. A, 104, (2000) 10600-10602. [141] G. Marnellos, S. Zisekas, and M. Stoukides, “Synthesis of Ammonia at Atmospheric Pressure with the Use of Solid State Proton Conductors”, Journal of Catalysis, 193(1), (2000), 80-87. [142] G. Karagiannakis, S. Zisekas, and M. Stoukides, “Hydrogenation of carbon dioxide on copper in an H+ conducting membrane-reactor“, Solid State Ionics, 162-163, (2003) 313-318. [143] D. Poulidi, M. A. Castillo-del-Rio, R. Salar, and I. S. Metcalfe, “Electrochemical promotion of catalysis using solid-state proton-conducting membranes”, Solid State Ionics, 162-163, (2003) 305-311. 244 [144] I. Politova, V. A. Sobyanin, and V. D. Belyaev, “Ethylene hydrogenation in electrochemical cell with solid proton-conducting electrolyte”, React. Kinet. Catal. Lett., 41(2), (1989) 321-326. [145] Salazar, and E. S. Smotkin, “Electrochemically promoted olefin isomerization reactions at polymer electrolyte fuel cell membrane electrode assemblies”, J. Appl. Electrochem., 36, (2006) 1237-1240. [146] D. Tsiplakides, S. G. Neophytides, O. Enea, M. Jaksic, and C. G. Vayenas, “Non-Faradaic electrochemical modification of the catalytic activity of Pt-black electrodes deposited on Nafion 117 solid polymer electrolytes”, J. Electrochem. Soc., 144(6), (1997) 2072-2078. [147] L. Ploense, M. Salazar, B. Gureau, and E. S. Smotkin, “Proton spillover promoted isomerization of n-butylenes on Pd-black cathodes/Nafion 117”, J. Am. Chem. Soc., 119, (1997) 11550-11551. [148] L. Ploense, M. Salazar, B. Gureau, and E. S. Smotkin, “Spectroscopic study of NEMCA promoted alkene isomerizations at PEM fuel cell Pd-Nafion cathodes”, Solid State Ionics, 136-137, (2000) 713-720. [149] D. Tsiplakides, and S. Balomenou, “Electrochemical promoted catalysis: towards practical utilization”, Chemical Industry & Chemical Engineering Quarterly, 14(2), (2008) 97-105. [150] G. Marnellos, and M. Stoukides, “Catalytic studies in electrochemical membrane reactors”, Solid State Ionics, 175, (2004), 597-603. [151] M. Marwood, and C. G. Vayenas, “Electrochemical promotion of electronically isolated Pt catalysts on stabilized zirconia”, Journal of Catalysis, 168, (1997) 538–542. [152] C. G. Vayenas, S. Bebelis, I. V. Yentekakis, and P. E. Tsiakaras, “Metal-solid electrolyte catalysts”, EP 0480116, (1992). [153] M. A. Priestnall, M. J. Evans, and M. S. P. Shaffer, “Mixed reactant fuel cells”, US 20040058203, (2004). [154] C. G. Yiokari, G. E. Pitselis, D. G. Polydoros, A. D. Katsaounis, and C. G. Vayenas, “High-pressure electrochemical promotion of ammonia synthesis over an industrial iron catalyst”, J. Phys. Chem. A, 104, (2000) 10600-10602. [155] J. St-Pierre, D. P. Wilkinson, and S. A. Campbell, “Method and device for improved catalytic activity in the purification of fluids”, US 20030010629, (2003). [156] S. P. Balomenou, D. Tsiplakides, A. Katsaounis, S. Brosda, A. Hammad, G. Foti, C. Comninellis, S. Thiemann-Handler, B. Cramer, and C. G. Vayenas, “Novel monolithic electrochemically promoted catalytic reactor for environmentally important reactions”, Applied Catalysis, 52, (2004) 181-196. [157] S. P. Balomenou, D. Tsiplakides, A. Katsaounis, S. Brosda, A. Hammad, G. Foti, C. Comninellis, S. Thiemann-Handler, B. Cramer, and C. G. Vayenas, “Monolithic electrochemically promoted reactors: a step for the practical utilization of electrochemical promotion”, Solid State Ionics, 177(26-32), (2006) 2201-2204. [158] L. Kriksunov, and C. G. Vayenas, “Method and device for treating automotive exhaust”, US 2006177365, (2006) [159] C. A. Cacalca, “Method for low temperature, high activity and selectivity catalytic conversion using electrochemical (NEMCA) cells”, US 20060131179, (2006). [160] Y. Ding, and J. C. Burba, “Fuel cell power plant with electrochemical enhanced carbon monoxide removal”, US 6733909, (2004). [161] J. M. Naber, “Fuel cell power plant with electrochemical enhanced carbon monoxide removal”, US 20020164507, (2002). 245 [162] M. Salazar, and E. S. Smotkin, “Electrochemically promoted olefin isomerization reactions at polymer electrolyte fuel cell membrane electrode assemblies”, J. Appl. Electrochem., 36, (2006) 1237- 1240. [163] F.M. Sapountzi, M.N. Tsampas, and C.G. Vayenas, “Methanol reformate treatment in a PEM fuel cell-reactor”, Catalysis Today, 127, (2007) 295–303. [164] Q. Fan, “Biofuel production by high-temperature non-Faradaic electrochemical modification of catalysis”, US 20100258447 (2010). [165] S. P. Balomenou, F. Sapountzi, D. Presvytes, M. Tsampas, and C. G. Vayenas, “Triode fuel cells”, Solid State Ionics, 177(19-25), (2006) 2023-2027. [166] C. G., Vayenas, and S. Balomenou, “Triode fuel cell and battery and method for conducting exothermic chemical reactions”, WO 2005/008820 (2005). [167] M. Behmann, S.-H. Kim, H.-J. Kob, and K. Lucas, “Experimental and theoretical investigations at the system CH3OH + H2SO4 + H2O”, Thermodynamic Properties of Complex Fluid Mixture, (2004) 3-25. [168] J. M. Dona Rodriguez, J. A. Herrera Melian, and J. Perez Pena, “Determination of the real surface area of Pt electrodes by hydrogen adsorption using cyclic voltammetry”, J. Chem. Educ., 77(9), (2000)1195-1197. [169] T. Ralph, G. A. Hards, J. E. Keating, S. A. Campbell, D. P. Wilkinson, M. Davis, J. St-Pierre, and M. C. Johnson. “Low cost electrodes for proton exchange membrane fuel cells”, J. Electrochem. Soc., 144(11), (1997) 3845-3857. [170] C. L. Green., and A. Kucernak, “Determination of the platinum and ruthenium surface areas in platinum-ruthenium alloy electrocatalysts by underpotential deposition of copper, I. Unsupported Catalysts”, J. Phys. Chem. B, 106, (2002) 1036-1047. [171] J. A. Dean, “Lange’s handbook of chemistry”, New York, NY, McGraw Hill, (1999) 8.2, 8.7. [172] D. R. Lide, “CRC handbook of chemistry and physics”, Boca Raton, FL, CRC Press LLC, (2006- 2007). [173] A. Katsaounis, S. Balomenou, D. Tsiplakides, S. Brosda, S. Neophytides, and C. G. Vayenas, “Proton tunnelling-induced bistability, oscillations and enhanced performance of PEM fuel cells”, Appl. Catal. B Environ., 56, (2006) 251-258. [174] T.R. Ralph, S. Hudson, and D. P. Wilkinson, “Electrocatalyst stability in PEMFCs and the role of fuel starvation and cell reversal tolerant anodes”, ECS Transactions, 1(8), (2006) 67-84. [175] T. T. H. Cheng, V. Colbow, S. Wessel, C. Chuy, and P. He, “Impacts of Ru dissolution and crossover on polymer electrolyte membrane fuel cell performance and anode functionality”, ECS Transactions, 28 (23), (2010) 243-251. [176] E. Gileadi, “Electrode kinetics for chemists, chemical engineers, and materials scientists”, VCH Publishers Inc., USA, (1993) 597 pp. [177] G. Wu, L. Li, and B.-Q. Xu, “Effect of electrochemical polarization of PtRu/C catalysts on methanol electrooxidation”, Electrochim. Acta, 50, (2004) 1-10. [178] W. F. Gale, and T. C. Totemeier., “Smithells metals reference book”, 8th edition, Oxford, Elsevier, (2004) 2080 pp. [179] K. Kinoshita, “Electrochemical oxygen technologies”, The electrochemical society series. New York, NY, John Wiley & Sons Inc., (1992) 431 pp. 246 [180] E. Herrero, K. Franaszczuk, and A. Wieckowski, “Electrochemistry of methanol at low index crystal planes of platinum: an integrated voltammetric and chronoamperometric study”, J. Phys. Chem., 98 (1994) 5074-5083. [181] K. Franaszczuk, E. Herrero, P. Zelenay, and A. Wieckowski, “A comparison of electrochemical and gas-phase decomposition of methanol on platinum surfaces”, J. Phys. Chem., 96, (1992) 8509-8516. [182] T. J. Schmidt, H. A. Gasteiger, and R. J. Behm., “Methanol electrooxidation on a colloidal PtRu- alloy fuel cell catalyst”, Electrochem. Comm., 1, (1999) 1-4. [183] A. Aramata, T. Kodera, and M. Masuda, “Electrooxidation of methanol on platinum bonded to solid polymer electrolyte, Nafion”, J. Appl. Electrochem., 18, (1998) 577-582. [184] L. Zhang, X. Cheng, H. Tu, M. You, Y. Zhang, and Q. Fan, “Kinetic analysis of methanol oxidation on Pt/C and PtRu/C catalysts”, ECS Transactions, 11(1), (2007) 1347-1357. [185] R. Baker, Z. Xie, J. Zhang, and D. P. Wilkinson., “Kinetic study for methanol oxidation at different temperatures using a new and improved electrochemical cell design”, Proceedings of the International Symposium on Fuel Cell and Hydrogen Technologies.1st, Calgary, AB, Canada, (2005) 3. [186] L. Dubau, C. Coutanceau, E. Garnier, J.-M. Leger, and C. Lamy, “Electrooxidation of methanol at platinum–ruthenium catalysts prepared from colloidal precursors: atomic composition and temperature effects”, J. Appl. Electrochem. 33(5), (2003) 419-429. [187] H. Guoyan, and J. Prakash, “Rotating rate dependency of methanol oxidation on a smooth polycrystalline platinum surface”, ECS Transactions, 1(14), (2006) 27-33. [188] T. H. Madden, N. Arvindan, and E. M. Stuve, “Development of an electrochemical flow-cell technique for studying methanol electro-oxidation at elevated temperatures”, J. Electrochem. Soc., 150(1), (2003) E1-E10. [189] H. N. Dinh, X. Ren, F. H. Garzon, P. Zelenay, and S. Gottesfeld, “Electrocatalysis in direct methanol fuel cells: Ii-situ probing of PtRu anode catalyst surfaces”, J. Electroanal. Chem., 491(1), (2000) 222-233. [190] P. Waszczuk, A. Wieckowski, P. Zelenay, S. Gottesfeld, C. Coutanceau, J.-M. Leger, and C. Lamy, “Adsorption of CO poison on fuel cell nanoparticle electrodes from methanol solutions: a radioactive labelling study”, J. Electroanal. Chem., 511(1-2), (2001) 55-64. [191] M. H. Miles, E. A. Klaus, B. P. Gunn, J. R. Locker, and W. E. Serafin, “The oxygen evolution reaction on platinum, Iridium, ruthenium and their alloys at 80oC in acid solutions”, Electrochim. Acta, 23, (1978) 521-526. [192] G. Tremiliosi-Filho, H. Kim, W. Chrzanowski, A. Wieckowski, B. Grzybowska, and P. Kulesza, “Reactivity and Activation parameters in methanol oxidation on platinum single crystal electrodes “decorated” by ruthenium adlayers”, J. Electroanal. Chem., 467, (1999) 143-156. [193] K. H. Lim, H.-S. Oh, S.-E. Jang, Y.-J. Ko, H.-J. Kim, and H. Kim, "Effect of operating conditions on carbon corrosion in polymer electrolyte membrane fuel cells”, J. Power Source, 193, (2009) 575-579. [194] L. M. Roen, C. H. Paik, and T. D. Jarvi, “Electrocatalytic corrosion of carbon support in PEMFC cathodes”, Electrochem. and Solid-State Lett, 7(1), (2004) A19-A22. [195] A. Zawodzinski, M. Neeman, L. O. Sillerud, and S. Gottesfeld, “Determination of water diffusion coefficients in perfluorosulfonate ionomeric membranes”, J. Phys. Chem., 95, (1991) 6040-6044. [196] M. M. Mench, “Fuel Cell Engines”, John Wiley & Sons Inc., (2008) 528 pp. [197] H. E. Darling, “Conductivity of sulphuric acid solutions”, J. Chem. Eng. Data, 9(3) (1964) 421-426. 247 [198] M. N. Tsampas, A. Pikos, S. Brosda, A. Katsaounis, and C. G. Vayenas, "The effect of membrane thickness on the conductivity of Nafion”, Electrochim. Acta, 51, (2006) 2743-2755. [199] P. R. Roberge, Handbook of Corrosion Engineering, McGraw-Hill Companies, USA, Appendix F, (2000) 1101-1123. [200] N. P. Finkelstein, and R. D. Hancock, “A New Approach to the Chemistry of Gold”, Gold bulletin, 7(3), (1974) 72-77. [201] P. Tengvall, and I. Lundstrom, “Physico-chemical considerations of titanium as a biomaterial”, Clin. Mater, 9, (1992)115-134. [202] E. Asselin, T. M. Ahmed, and A. Alfantazi, “Corrosion of niobium in sulphuric acid and hydrochloric acid solutions at 75 and 95oC”, Corrosion Sci., 49, (2007) 649-710. [203] Wesselmark, M., Wickman, B., Lagergren, C., andLindbergh, G., “Electrochemical performance and stability of thin film electrodes with metal oxides in polymer electrolyte fuel cells”, Electrochim. Acta, 55, (2010) 7590-7596. [204] K. I. Hadjiivanov, and D. G. Klissurski, “Surface chemistry of titania (anatase) and titania-supported catalysts”, Chemical Society Reviews, (1996) 61-69. [205] T. Ioroi, H. Senoh, S.-I. Yamazaki, Z. Siroma, N. Fujiwara, and K. Yasuda, “Stability of corrosion- resistant magneli-phase Ti4O7-supported PEMFC catalysts at high potentials”, J. Electrochem. Soc., 155(4), (2008) B321-B326. [206] J. Shim, C.-R. Lee, H.-K. Lee, J.-S. Lee, and E. J. Cairns, “Electrochemical characteristics of Pt- WO3/C and Pt-TiO2/C electrocatalysts in a polymer electrolyte fuel cell”, J Power Sources, 102, (2001) 172-177. [207] R. E. Fuentes, B. L. Garcia, and J. W. Weidner, “Effect of Titanium Dioxide Supports on the Activity of Pt-Ru toward Electrochemical Oxidation of Methanol”, J. Electrochem. Soc., 158(5), (2011) B461- B466. [208] G. R. Bamwenda, S. Tsubota, T. Nakamura, M. Haruta, "Photoassisted Hydrogen Production from a Water-Ethanol Solution: A Comparison of Activities of Au-TiO2 and Pt-TiO2", Journal of Photochemistry and Photobiology A: Chemistry, 89, (1995) 177-189. [209] K. Hirakawaa, M. Inoueb, and T. Abea, “Methanol oxidation on carbon-supported Pt–Ru and TiO2 (Pt–Ru/TiO2/C) electrocatalyst prepared using polygonal barrel-sputtering method”, Electrochim. Acta, 55 (20), (2010) 5874-5880. [210] S.Lj. Gojkovića, B.M. Babićb, V.R. Radmilovićc and N.V. Krstajića, “Nb-doped TiO2 as a support of Pt and Pt–Ru anode catalyst for PEMFCs”, Journal of Electroanalytical Chemistry, 639 (1-2), (2010) 161- 166. [211] P. Claus, S. Schimpfb, R. Schödelc, P. Kraakc, W. Mörked, and D. Hönickeb, “Hydrogenation of crotonaldehyde on Pt/TiO2 catalysts: Influence of the phase composition of titania on activity and intramolecular selectivity”, Applied Catalysis A: General, 165 (1-2), (1997) 429-441. [212] P. Pnangiotopoulou, and D. I. Kondarides, "Effect of Morphological Characteristics of TiO2- supported noble metal catalysts on their activity for the water-gas shift reaction", Journal of Catalysis, 225, (2004) 327-336. [213] F. L. S. Purgato, L. A. Montoro, J. Riberiro, K. B. Kokoh, and P. Olivi, “The effect of heat treatment on the preparation of Pt-RuO2/C electrocatalysts.”, Electrocatal, (2010). [214] S. Brunauer, P.H. Emmett, and E. Teller, J. Am. Chem. Soc., 1938, 60, 309. 248 [215] A. S. Polo, M. C. Santos, R. F. B. De Souza, W. A. Alves, “Pt-Ru-TiO2 photoelectrocatalysts for methanol oxidation”, J. Power Sources, 196, (2011) 872-876. [216] N. Perkas, Z. Zhong, L. Chen, M. Besson, and A. Gedanken, “Sonochemically prepared high dispersed Ru/TiO2 mesoporous catalyst for partial oxidation of methane to syngas”, Catalysis Letters, 103(1-2), (2005) 9-14. [217] S. J. Tauster, S. C. Fung, and R. L. Garten, “Strong metal-support interactions, group 8 noble metals supported on TiO2”, Journal of the American Chemical Society, 100(1), (1978) 170-175. [218] E. Antolini, L. Giorgi, F. Cardellini, and E. Passalacqua, “physical and morphological characteristics and electrochemical behaviour in PEM fuel cells of Pt-Ru/C catalysts”, J. Solid State Electrochem., 5, (2001) 131-140. [219] J. Willsau, and J. Heitbaum, “Analysis of adsorbed intermediates and determination of surface potential shifts by DEMS”, Electrochim. Acta, 31(8), (1986) 943-948. [220] C. Lamy, E. M. Belgsir, and J.-M. Leger, “Electrocatalytic oxidation of aliphatic alcohols: application to the direct alcohol fuel cell (DAFC).”, J. of Appl. Electrochem., (31), (2001) 799-809. [221] M. Metikos-Hukovic, R. Babic, and Y. Piljac, “Kinetics and electrocatalysis of methanol oxidation on electrodeposited Pt and Pt70Ru30 catalysts.”, J. of New Materials for Electrochemical Systems, 7, (2004) 179-190. [222] N. Karthikeyan, V. V. Giridhar, and D. Vasudevan, “Surfactant effects on methanol oxidation at Pt– Ru/C coated glassy carbon electrode”, J. Solid State Electrochem, 14(5), (2010) 877-881. [223] W. H. lizcano-Valbuena, V. A. Paganin, and E. R. Gonzalez, “Methanol electro-oxidation on gas diffusion electrodes prepared with Pt/Ru/C catalysts”, Electrochim. Acta, 47(22-23), (2002) 3715-3722. [224] A. N. Golikand, M. G. Maragheh, S. S. Sherehjini, K. M. Taghi-Ganji, M. Yari, „Carbon-supported Pt particles as a catalyst for electrooxidation of methanol and cyclic voltammetry studies under acidic conditions, Electroanalysis, 18(9), (2006) 911-917. [225] A. V. Tripkovic, K. D. Popovic, B. N. Grgur, B. Blizanac, P. N. Ross, and N. M. Markovic, “Methanol electrooxidation on supported Pt and PtRu catalysts in acid and alkaline solutions.”, Electrochim. Acta, 47, (2002) 3707-3714. [226] C.-G. Lee, M. Umeda, and I. Uchida, “Cyclic voltammetric analysis of C1-C4 alcohol electrooxidations with Pt/C and Pt-Ru/C microporous electrodes”, J. Power Sources, 160, (2006) 78-89. [227] T. Iwasita, H. Hoster, A. John-Anacker, W. F. Lin, and W. Vielstich, “Methanol oxidation on PtRu electrodes. Influence of surface structure and Pt-Ru atom distribution”, Langmuir, 16, (2000) 522-529. [228] M. Hepel, I. Kumarihamy, and C. J. Zhong, “Nanoporous TiO2-supported bimetallic catalysts for methanol oxidation in acidic media.”, Electrochem. Comm., 8, (2006) 1439-1444. [229] Z. Wang, G. Chen, D. Xia, and L. Zhang, “Studies on the electrocatalytic properties of PtRu/C-TiO2 toward the oxidation of methanol.”, J. of Alloys and Compounds, 450, (2008) 148-151. [230] H. Song, X. Qui, X. Li, F. Li, W. Zhu, and L. Chen, “TiO2 nanotubes promoting Pt/C catalysts for ethanol electro-oxidation in acidic media.”, J. Power Sources, 170, (2007) 50-54. [231] D. He, L. Yang, S. Kuang, and Q. Cai, “Fabrication and catalytic properties of Pt and Ru decorated TiO2/CNTs catalyst for methanol electrooxidation”, Electrochem. Comm., 9, (2007) 2467-2472. [232] H. Song, Z. Qiu, F. Li, W. Zhu, and L. Chen, “Ethanol electro-oxidation on catalysts with TiO2 coated carbon nanotubes as support”, Electrochem. Comm., 9, (2007) 1416-1421. 249 [233] B. N. Grgur, N. M. Markov, C. A. Lucas, and P. N. Ross, “Electrochemical oxidation of carbon monoxide: from platinum single crystals to low temperature fuel cells catalysts. Part I: Carbon monoxide oxidation onto low index platinum single crystals”, J. Serb. Chem. Soc., 66(11–12), (2001) 785-797. [234] J. Zhang, “PEM fuel cell electrocatalysts and catalyst layers, Fundamentals and Applications”, Springer-Verlag London Limited, (2008) 1137 pp. [235] l. D. Santos, F. Colmati, E. R. Gonzales, “Preparation and characterization of supported Pt-Ru catalysts with a high Ru content.”, J. Power Sources, 159 (2006) 869-877. [236] E. Ticanelli, J. G. Beery, M. T. Paffett, S. Gottesfeld, “An electrochemical, ellipsometric, and surface science investigation of the Pt-Ru bulk alloy surface”, J. Electroanal. Chem., 258(1), (1989) 61- 77. [237] P. Stonehart, “Development of alloy electrocatalysts for phosphoric acid fuel cells (PAFC)”, J. Applied Electrochem, 22(11), (1992) 995-1001. [238] C. G. Vayenas, and S. Neophytides, "Non-Faradaic electrochemical modification of catalytic activity: 3. the case of methanol oxidation on Pt”, J. Catal., 127, (1991) 645-664. [239] I. V. Yentekakis, Y. Jiang, S. Neophytides, S. Bebelis, and C. G. Vayenas, "Catalysis, electrocatalysis and electrochemical promotion of the steam reforming of methane over Ni film and Ni- YSZ cermet anodes, Ionics, 1, (1995) 491-498. [240] S. Bebelis, and C. G. Vayenas, Non-Faradaic electrochemical modification of catalytic activity: 1. the case of ethylene oxidation on Pt", J. Catal., 118, (1989) 125-146. [241] D. Tsiplakides, J. Nicole, C. G. Vayenas, and C. Comninellis, "Work function and catalytic activity measurements of an IrO2 film deposited on YSZ subjected to in situ electrochemical promotion", J. Electrochem. Soc., 145(3), (1998) 905-908. [242] C. Pliangos, I. V. Yentekakis, X. E. Verykios, and C. G. Vayenas, "Non-Faradaic electrochemical modification of catalytic activity: 8. Rh-catalyzed C2H4 oxidation, J. Catal., 154, (1995) 124-136. [243] M. Makri, A. Buekenhoudt, J. Luyten, and C. G. Vayenas, "Non-Faradaic electrochemical modification of the catalytic activity of Pt using a CaZr0.9In0.1O3-α proton conductor", Ionics 2, (1996) 282-288. [244] I. V. Yentekakis, G. Moggridge, C. G. Vayenas, and R. M. Lambert, "In-situ controlled promotion of catalyst surfaces via NEMCA: the effect of Na on the Pt-catalyzed CO oxidation, J. Catal., 146, (1994) 292-305. [245] L. Ploense, M. Salazar, B. Gurau, and E. S. Smotkin, "Spectroscopic study of NEMCA promoted alkalene isomerizations at PEM fuel cell Pd-Nafion cathodes, Solid State Ionics, 136-137, (2000) 713- 720. [246] S. Neophytides, D. Tsiplakides, P. Stonehart, M. Jaksic, and C. G. Vayenas, "Electrochemical enhancement of a catalytic reaction in aqueous solutions, Nature, 370, (1994), 292-294. [247] E. Lamy-Pitawa, S. E. Mouahid, and J. Barbier, "Effect of anions on catalytic and electrocatalytic hydrogenations and on the electrocatalytic oxidation and evolution of hydrogen on platinum", Electrochim. Acta, 45, (2000), 4299-4308. [248] C. G. Vayenas, S. Bebelis, I. V. Yentekakis, P. Tsiakaras, and H. Karasali, “Non-Faradaic electrochemical modification of catalytic activity, Reversible promotion of platinum metals catalysts.” Platinum Metals Rev., 34(3), (1990) 122-130. 250 Appendix A: Publications, Presentations, and Posters A.1 Publications - C. R. Cloutier, and D. P. Wilkinson, “Electrochemical Promotion of Aqueous Acidic Methanol PEM Electrolysis in Triode and Tetrode Operation”, (submitted). - C. R. Cloutier, and D. P. Wilkinson, "Triode Operation of a Proton Exchange Membrane (PEM) Electrolyser", ECS Transactions, 25 (23), (2010) 47-57. - C. R. Cloutier, and D. P. Wilkinson, “Electrolytic Production of Hydrogen from Aqueous Acidic Methanol Solutions”, Int. J. H2 Energy, 35, (2010) 3967-3984. A.2 Presentations - Hydrogen and Fuel Cells Conference, Hydrogen production, storage and utilisation – from nanomaterials to demonstrators, Presentation, (2011), Cancun, Mexico. - Catalysis Society of South Africa (CATSA), Invited Keynote Presentation, (2011), Gauteng, South Africa. - 3rd International Training School on the Electrochemical Promotion of Catalysis, Short Communication, (2009), Université Claude Bernard Lyon 1, Villeurbanne, Lyon, France. - 216th ECS Meeting, Electrochemistry: Symposium on Interfacial Electrochemistry in Honour of Brian - E. Conway, Presentation, (2009), Vienna, Austria. - T. Jang, C. R. Cloutier, K. Matsuzawa, S. Mitsushima, and K.-I. Ota, Industrial Electrolysis Symposium, Presentation, (2008), Kobe, Japan. - Department of Energy and Safety Engineering, Chemical Energy Laboratory Conference, Yokohama - National University, Presentation, (2008), Yokohama, Kansai, Japan. A.3 Posters - The Graduate University of Advanced Studies, JSPS Summer Program, Poster, June 20 (2008), Zushi, Kansai, Japan. - 58th Annual Meeting of the International Society of Electrochemistry, Poster, September 9-14 (2007), Banff, AB, Canada. 251 Appendix B: Methanol Electro-oxidation Mechanism B.1 Methanol electro-oxidation parallel pathways The schematic representation of the parallel pathways for methanol electro-oxidation is described in Fig. B.1. Figure B.1: Methanol electro-oxidation parallel pathways. B.2 Methanol electro-oxidation mechanism on Pt The generally accepted reaction sequence of the mechanism of the electro-oxidation of methanol on Pt- based catalysts is described as follows [1]. Pt + CH3OH → Pt -(CH3OH)ad (B.1) Pt-(CH3OH)ad → Pt-(CH2OH)ad + H+ + e- (B.2) Pt-(CH2OH)ad → Pt-(CHOH)ad + H+ + e- (B.3) Pt-(CHOH)ad → Pt-(COH)ad + H+ + e- (B.4) Pt(COH)ad → Pt-(CO)ad + H+ + e- (B.5) Pt + H2O → Pt(H2O)ad (B.6) CO2 HCOOH COOH CO COH CHO CH2O CH2OH CHOH CH3OH 252 Pt-(H2O)ad + Pt(CO)ad → 2Pt + CO2 + 2H+ + 2e- (B.7) Pt-(H2O)ad → Pt-(OH)ad + H+ + e- (B.8) Pt-(OH)ad + Pt-(CO)ad → 2Pt + CO2 + H+ + e- (B.9) B.3 Methanol electro-oxidation mechanism on Pt-Ru The bi-functional mechanism illustrating the enhancement effect of Pt-Ru on the electro-oxidation of methanol is described as follows [2]. CH3OH + Pt(H2O) → Pt(CH3OH)ad + H2O (B.10) Pt(CH3OH)ad → Pt(CO)ad + 4H+ + 4e- (B.11) Ru + H2O → Ru(OH)ad + H+ + e- (B.12) Pt(CO)ad + Ru(OH)ad → Pt + Ru + CO2 + H+ + e- (B.13) Ru(CO)ad + Ru(OH)ad → 2Ru + CO2 + H+ + e- (B.14) The schematic representation of the bi-functional mechanism of Pt-Ru on the electro-oxidation of methanol is shown on Fig. B.2 [3]. Figure B.2: Methanol electro-oxidation bi-functional mechanism on Pt-Ru. B.4 References [1] N. A. Tapan, W. E. Mustain, B. Gurau, G. Sandi and J. Prakash, “Investigation of Methanol Oxidation Electrokinetics on Pt Using the Asymmetric Electrode Technique”, Journal of New Materials for Electrochemical Systems, 7 (2004) pp. 281-286. e- - H+- H+ Pt C OH e- - H+ Pt C O H H e- - H+ e- - H+ e- Pt C HOHH OH H C Pt CO2 + Pt Pt ~ 0.9 V Ru ~ 0.3 V O H2 O C H OH H H Pt 253 [2] T. Iwassita, H. Hoster, A. John-Anacker, W. F. Lin and W. Vielstich, “Methanol Oxidation on PtRu Electrodes. Influence of Surface Structure and Pt-Ru Atom Distribution”, Langmuir, 16 (2000) p. 522-529. [3] Presentation: “Nanostructured Architectures for Solar Energy and Beyond”, Radiation Laboratory, Dept. of Chemical and Biomolecular Engineering, University of Notre Dame, Nore Dame, Indiana, 46556-0579. 254 Appendix C: Efficiency and Economic Comparison of H2 Production C.1 Objective Determine the merits of the overall direct methanol electrochemical reforming system approach compared to other H2 generation approaches. This will be accomplished by conducting a high level analysis composed of an efficiency and economic comparison of different methods of H2 production. C.2 System definition Hydrogen can be produced in various ways. In this study, three production process schemes for the manufacturing of H2 are compared: electrolysis, catalytic steam reforming, and chemical hydride hydrolysis. Three fuels were considered: methanol, water, and sodium borohydride. These processes are only ecologically feasible if the fuels are originating from renewable resources. Hence, only production from renewable raw material sources was examined: biomass (methanol), water, and borax mineral ore (sodium borohydride). The four process combinations evaluated are: methanol PEM electrolysis or direct methanol electrochemical reforming (DMR), water PEM electrolysis, methanol catalytic steam reforming, and sodium borohydride hydrolysis. Other H2 production processes and similar comparative studies have been provided elsewhere [1, 2]. C.3 Block Diagrams Block Diagrams of the H2 production processes compared are shown in Fig. C.1. For the purpose of this high level evaluation, all processes were compared on a stationary basis and only the most important processing steps were considered. Purification processes such as water filtration and deionization, biosyngas purification and methanol sulphur removal, as well as passive devices such as liquid/gas separators were all ignored. In all cases, the end product is pure H2 (99.9%). The produced H2 255 compression, storage and application requirements were not included in the analysis. The same applied for the pipelines and infrastructure required for the transportation of materials. Figure C.1: Block diagrams of four H2 generation process schemes (a) PEM methanol electrolyser, (b) PEM water electrolyser, (c) methanol steam reforming, and (d) sodium borohydride hydrolysis. PEM Methanol Electrolyser Pump H2 CO2 Electricity CH3OH H2O Biomass Gasification Heat Steam Fuel Tank PEM Water Electrolyser Water Tank Pump H2 H2O Electricity Pump CH3OH H2O Biomass Gasification Heat Steam Fuel Tank Steam Reforming Heat CO2Steam Shift Reactor CO2 Purification (PSA) H2 Fuel Tank Catalytic Reactor Pump NaBH4 H2O Schlesinger and Brown Process Borate Ore Refining Electricity and Heat NaCl CH4 Wastes Borax Na2B4O7.10H2O Trimethoxyborane (CH3O)3B H2 NaBO2 (a) (b) (c) (d) CO2Steam 256 C.4 Feedstock processing C.4.1 Methanol production Biomass is renewable organic matter available through natural processes or as a by-product of processes using renewable resources [3]. It represents a large potential feedstock for sustainable processes generating chemicals or electricity. The exact source of biomass was not specified. Pre- treatment of biomass feedstock typically includes processes such as drying and chipping which were not taken into account in this evaluation. After pre-treatment, the biomass is converted to biosyngas through sequential steps, which usually include a pressurized gasification, a gas conditioning, a catalytic reaction, and a distillation step. The gasification is the most crucial step and is the only process considered in this study. It is a thermochemical process usually carried at high temperature and pressure, during which all the biomass is converted to biosyngas. Indirect gasification is of interest for methanol synthesis as it does not require an oxygen plant [4]. The biosyngas produced requires cleaning as it contains contaminants. This purification process was ignored in the evaluation. Methanol is exothermically synthesised from the biosyngas as per: CO + 2H2 ↔ CH3OH (C.1) CO2 + 3H2 ↔ CH3OH + H2O (C.2) The overall production of methanol by the gasification of biomass can be described by the following general reaction: Biomass → Biosyngas (CO, CO2, H2) → CH3OH (C.3) The methanol generated contains water, and, as some side reactions are possible, it also contains minor amounts of by-products. For this reason, the methanol produced is usually purified to the required final purity by multistage distillation. Again, this purification step was not taken into account. 257 C.4.2 Sodium borohydride production Most NaBH4 is produced by the Schlesinger and Brown process, which is based on the reduction of trimethoxyborane ((CH3O)3B) obtained from borate ore refining. The borate ore refining process is relatively simple and typically includes four basic steps to transform the raw ore into refined borates: dissolution, purification, crystallization and drying. The Schlesinger and Brown process is quite energy extensive as it includes the electrolysis of NaCl to produce Na and the catalytic steam reforming of methane to generate H2. The overall chemical reaction is: NaBO2 + 4Na + 2H2 + 1/2H2SO4 → NaBH4 + 3NaOH + 1/2Na2SO4 (C.4) The electrolysis of NaCl is the reaction step having the highest activation energy requirements and is the most energy consuming of the entire process [5]. There is also a large amount of by-products (Na2SO4), wastes (Cl2) and greenhouse gases (CO2) generated by the overall process. C.5 Hydrogen production Table C.1 shows a high level comparison of the H2 generation processes discussed below. 258 Table C.1: High level comparison of four H2 generation process schemes. System Description DMR PEM Water Electrolyser Methanol Steam Reformer Sodium Borohydride System Type Electrochemical Electrochemical Chemical Others Fuel Methanol and water mixture Pure Water Pure Methanol NaBH4 and water mixture Fuel Safety Flammable Non-flammable Flammable Non-flammable H2 Production Reaction CH3OH + H2O → CO2 + 3H2 2H2O → 2H2 + O2 CH3OH + H2O → CO2 + 3H2 NaBH4 + 4H2O → 4H2 + NaBO2.2H2O Operating T (oC) 80 80 270 90 Start-up time minutes minutes 30 min. less than 30 seconds Footprint Small Small Large Small CO2:H2 1:3 - 1:3 - Type of CO2 emissions Localized - Non-localized Non-localized Major Processing steps 2 1 4 3 C.5.1 Electrolysis In general, PEM water electrolysers are cleaner, safer and result in a higher H2 purity than traditional water-alkaline electrolysers. The PEM electrolysis of methanol will consume less electricity than the PEM electrolysis of water. This is based on the fact that the standard potential for the electro-oxidation of alcohol in the liquid phase is only 0.02 V vs. SHE for methanol, compared to the 1.23 V vs. SHE for the electro-oxidation of water. It was reported that the cost of methanol electrolysis is about half of the cost of water electrolysis, even when the cost of methanol is taken into account [6]. However, no details were provided on this calculation and the source used for the production of the methanol was not specified. A drawback of methanol electrolysis over water electrolysis is that CO2, a greenhouse gas, is emitted while in water electrolysis useful O2 is released. However, it is important to note that the CO2 emissions 259 from the electrochemical reformer are localized and can be more easily controllable and captured than the non-localized emissions resulting from other processes. C.5.2 Catalytic steam reforming The catalytic SR of methanol is an endothermic reaction carried out at temperatures between 250 and 350 oC which produces H2 [7]. Ideally, the process starts by the splitting of methanol into CO and H2 in the reformer. It is followed by the splitting of the water into H2 and O2 through the exothermal water gas shift (WGS) reactor, where CO is combined with O2 to form CO2. The chemical reactions are: Reformer: CH3OH ↔ CO + 2H2 (C.5) Shift Reactor: CO + H2O ↔ CO2 + H2 (C.6) Overall: CH3OH + H2O ↔ CO2 + 3H2 (C.7) The maximum H2 production efficiency of 78 mol % was obtained for the SR reaction at 250 oC using a catalyst containing 27.8 wt% Cu and calcined at 700oC [8]. A gas clean-up sub-system is required to obtain pure H2. Pressure swing adsorption (PSA) is a cost- effective and reliable industrial process based on selective adsorbent beds, which is commonly used for the purification of H2 from catalytic reformer off-gas. The gas mixture is introduced in the bed at an elevated pressure and the solid adsorbent selectively adsorbs certain components of the gas mixture. The un-adsorbed components pass through the bed as purified produced gas. Multiple beds are cycled in the process, allowing for the adsorbed pollutants to be periodically desorbed, cleaning the beds for the next cycle [9]. PSA results in pure H2 (>99.9%) and has recovery rates of 90% or higher [4]. 260 C.5.3 Sodium borohydride hydrolyser The overall reaction is exothermic and liberates 75 kJ/mole of H2 formed at 25oC [10]. Hence, the reaction is very energetically favourable and no energy input is required to produce H2. The hydrolysis of NaBH4 liberates a determined amount of H2 at ambient temperature and pressure with no side reactions or by-product. The H2 generation system comprises two storage tanks for NaBH4 and NaBO2 respectively, a pump, a catalyst chamber reactor, a liquid/gas separator, and a heat exchanger. The last two units were ignored in this evaluation. The release of H2 from the NaBH4 alkaline solution occurs by pumping the fluid mixture through a reactor containing a catalyst. The heat supplied by the exothermic reaction partially humidifies the H2 gas by evaporating water. The H2 produced contains less than 0.5 ppm of CO, and no NOx or SOx [5]. The H2 is usually cooled down in a heat exchanger before being used. C.6 Efficiency analysis The efficiency was estimated for each major processing step. The overall efficiency of the H2 production process was obtained by multiplying the efficiency of the individual processing steps. It was assumed that the stoichiometry was one in all case, i.e. no feed was provided in an excess of the required stoichiometric amounts. The electrolyser current efficiencies were assumed to be equal to unity. Compression efficiencies were ignored. The efficiencies obtained from the literature represent various process conditions. For example, the range of efficiencies reported for water electrolysis varied from 30 to 90%. The results are summarized in Table C.2. 261 Table C.2: Efficiency comparison of four H2 generation process schemes. Process Major Steps Efficiency (%) Notes Ref. Overall Efficiency (%) Methanol Electrolyser Gasification from biomass 57 Upper average of reported efficiencies [11, 12] Liquid Pump 90 [13] Electrolysis 70 Assumed methanol electrolyser would have an efficiency 10% higher than the water electrolyser due to higher energy efficiency. 36 Water Electrolyser Liquid Pump 90 [13] Electrolysis 60 Vary between 30 to 90% [14, 15- 17] 54 Methanol Steam Reforming Gasification from biomass 57 Upper average of reported efficiencies [11, 12] Gas Pump 75 Steam reformer and shift reactor 78 78% reported maximum, 77% generally accepted [15, 8] PSA Purification 90 [4] 30 Sodium Borohydride Hydrolysis Borate refining 55 [19] Schlesinger and Brown process 47 Na manufacture energy efficiency of 50%, borohydride generation reaction yield of 94% [20] Liquid Pump 90 [13] Catalytic reactor 93 Fuel input volume efficiency of 95% and yield of H2 of 98% [13] 22 The efficiency calculated indicates that the sodium borohydride hydrolysis process is the least efficient of the processes reviewed while water electrolysis is the most efficient of the process reviewed. The results give the general idea that the electrolysis processes are more efficient. However, due to the low efficiency of the gasification of biomass to generate methanol, the methanol electrolysis process fall behind the water electrolysis process. 262 If the storage and dispensing efficiency are taken into account the results would be significantly altered. The storage and dispensing efficiency was reported to be 65 % for H2 and 99.8 % for methanol [15]. This would result in an overall efficiency of 35 % for H2 electrolysis and of 36 % for methanol electrolysis. Hence, it might be appropriate to have the methanol production facility at a remote central location, and the methanol electrolyser at distributed locations. The water electrolysis would have to be at distributed locations, as the transport of hydrogen has a low efficiency. C.7 Economic analysis The economic analysis was high level. No complex statistical or modelling methods were used to determine the individual contributions to the cost of H2. The cost of hydrogen production was simply broken down in the feedstock cost, the capital cost, the operation and maintenance cost and the carbon cost as per this basic formula: CH2 = CCAP * FCR + COM + CF / Eff + CC * PC (C.8) where CH2 is the hydrogen cost ($/GJ), CCAP is the capital cost ($-y/GJ), FCR is the fixed charge rate on capital (1/y), COM is the operation and maintenance cost ($/GJ), CF is the feedstock cost ($/GJ), Eff is the efficiency (GJ of Hydrogen Produced / GJ of Potential Energy in Feedstock), CC is the carbon tax ($/ton C), and PC is the carbon emissions (tons of C / GJ of hydrogen produced) [21]. The CCAP includes only the cost of the essential equipment of the facility. The practical standard in capital cost evaluation in the energy industry is to include an interest rate of 8 %, depreciation period of 15 years, i.e., about 50 % of the expected technical lifetime of the plant. However, for the purpose of this high level analysis, the FCR, which encompasses the investment return rate, income taxes, insurance charges, interest during construction, start-up expenses, working capital and other costs, was assumed to be equal to one. Little if any information pertaining to the COM, i.e., operating and maintenance costs, could be found in the literature, and these costs were not considered. The efficiency effect on the 263 feedstock costs was neglected and the Eff value was assumed to be unity. The resulting simplified formula is as follows: CH2 = CCAP * FCR + CF + CE + CC * PC (C.9) where CE is the energy cost ($/GJ). A cost should also be added to take the cost of waste disposal into account as well as the credits for any useful by-products resulting from the process. However, as these costs are difficult to establish, they were neglected in this simple cost analysis. Cost estimates for each component were found from the literature. There is a wide range of costs reported in the literature as they are all based on different assumptions. Major factors affecting the H2 production cost include the feedstock and the energy source from which the H2 is produced, the size of the H2 production facility and the transportation requirements for delivery it to the customer, the state of the technology used, and whether or not the CO2 by-product is sequestered. In this simplified cost analysis, it is assumed that all electricity is generated from hydro-electric power plants and does not include any costs associated with the various hydro-electric power plant capacities and unit operation sizes. The H2 production plant cost is proportional to its capacity, and plants with smaller footprints should result in lower real-estate, installation and maintenance costs. However, these issues were not taken into account in this analysis. The cost estimate values found from the different studies were assumed to be all on the same plant capacity basis. Furthermore, no costs associated with transportation were taken into account. Lastly, the cost estimates were not adjusted to current dollar amounts using inflation rates and the costs were not normalized to the current year. They were left in the year of publication of the data. Historical correction of the currency conversion between US$ and CAN$ was not conducted and it was assumed that 1$US was equal to 1$CAN. Units were converted from $/kg H2 produced to $/GJ or vice-versa using the energy density of H2, which is about 120000 KJ/kg H2. 264 C.7.1 Capital cost The capital cost of an electrolyser depends on its H2 generation capacity and its rated current density [22]. The cost of commercial PEM water electrolysers vary from 13 to 159 $/GJ depending on the size of the unit [23]. For a distributed plant, the cost of water electrolysis was reported as being between 7.36 and 6.82 $/kg H2 [16]. This leads to an average of 59.08 $-y/GJ. It was assumed that the capital cost of the PEM methanol electrolyser would be very similar to that of the PEM water electrolyser. The capital cost of methanol production from biomass by thermal gasification and by supercritical water gasification are 1.5 and 4.5x103 US$/kW, respectively [11]. This results in an average capital cost of 95.13 $/GJ. The capital cost of H2 production by methanol reforming is reported as 4.65x103 US$/kW [11]. This is equivalent to 28.54 $/GJ. For simplicity, it will be assumed that the steam reformer, the shift reactor and the PSA unit are all included in this cost. As described earlier, the extraction and production of borates is quite simple and the mining costs only vary by the depth of the pit and the amount of overburden to be removed to expose the ore [24]. However, one of the most difficult things to determine in the borate market is the cost of production of each supplier. About 75 % of the borate market is controlled by its two major producers, and there are strong indications that capacity is restricted and there is collusion on pricing [25]. For this reason, the capital cost associated with borate mining was ignored in this high level cost estimate. The capital cost needed for the production of NaBH4 through the Schlesinger and Brown process could not be found in the literature. However, it is expected to be very high as it includes an electrolyser, a reformer and a reactor. For estimation purposes, the electrolyser was assumed to be the same cost as the PEM water electrolyser, the reformer and the reactor the same as the methanol steam reformer. This would result in a capital cost estimate of 116.16 $/GJ. The cost of the hydrolysis reactor could not be found in the literature either as it is a new technology still under development. For estimation purposes, it was assumed as being a third of the cost of the PEM water electrolyser as it is a one chamber reactor, does not contain any membrane and likely uses non-noble catalyst. This would give an estimated capital cost of 19.19 $/GJ. 265 C.7.2 Feedstock cost The cost of methanol made from biomass was reported as being 75 US$/MWh or 20.83 US$/GJ [26]. In all cases, it was assumed that the cost of water usage and water demineralization was negligible. The listed price of NaBH4 is about $50US/kg for industrial quantities. Based on this feed price, it would cost $230US to produce 1 kg of H2 from NaBH4. This is equivalent to 19166.67 $/GJ. C.7.3 Energy cost All electricity was assumed to originate from hydroelectric power plants. The cost of electricity at BC Hydro is about 0.05 $/kWh [27]. The energy requirements for water electrolysis was reported to be between 53.4 and 70.1 kWh/kg of H2 produced [16]. Taking the average and using the assumed electricity cost, this results in an energy cost of 25.67 $/GJ. Based on the preliminary experiments conducted, there is about 1 V difference between the voltage that needs to be supplied for the electrolysis of methanol compared to the voltage required for the electrolysis of water at the same conditions. Hence, for estimation purposes, the methanol electrolyser was assumed to cost about 1/3 of the energy cost of the water electrolyser, i.e. 8.56 $/GJ. The energy requirement of the gasification process largely depends on the type of biomass used. In some cases, the net electrical output is null or positive [12]. This is because the gross electricity produced by gas turbine and/or steam turbine can be equal or greater than the internal electricity use from pumps, compressors, oxygen separator, and other units operations. An average value of 28.78 kWh/GJ was calculated based on the electricity requirements of four different types of gasifiers [28]. Using the assumed electricity cost, this results in an energy cost of 1.44 $/GJ. No values could be found for the energy requirements of the methanol steam reforming process. Hence, a generic energy requirement value of 40 kWh/kg H2 produced, based on hydrocarbon steam reforming, was used. This corresponds to a cost of 166.66 $/GJ. The most energy consuming process of the manufacture of NaBH4 is the electrolysis of Na The electrolysis of Na requires 9.7 kWh per kg of Na produced [29]. As it takes 4 moles of Na to produce one 266 mole of NaBH4, which in turn can produce 4 moles of H2, this results in 111.55 kWh/kg H2. Based on the electricity cost, and converting using the energy density of H2, this can be converted to an energy cost of 464.79$/GJ. It was assumed that the catalytic reactor does not require any energy input as the sodium borohydride hydrolysis reaction is exothermic. C.7.4 Carbon cost The emissions from manufacturing the H2 production facility become of interest when renewable resources are used. Cost of the renewable H2 production paths becomes more comparable when the cost of environmental damage is taken into account. This cost can be evaluated by assuming a carbon tax or by estimating the capture and sequestration cost of CO2. Capture and sequestration only makes economical sense if it costs less than the imposed cost of releasing carbon to the atmosphere. The cost of CO2 sequestration includes the costs for drying and compressing the CO2 to the pressure required for CO2 transport and injection, the costs of pipelines for transporting the CO2 to the sequestering sites and the costs for wells and surface facilities at the storage sites [30]. For simplicity, a carbon tax, CC, of 20 $/tonC was assumed in this high level cost analysis. As the quantity of CO2 emitted from the electric power generation plants was not taken into account, the water electrolyser system results in no CO2 emissions. The total emissions resulting from the production of methanol from biomass, i.e. emission from fuel extraction, transportation and conversion was reported as 30 gC/kWh or 0.08 tonC/GJ [27]. Using the carbon tax, this can be converted to a carbon cost of 1.66 $/GJ. As the methanol electrolyser and the methanol steam reforming have the same overall reaction, the total carbon emissions can be assumed to be similar in both cases if the efficiencies are not taken into account. Based on a feed stoichiometry of 1, there will be one mol of CO2 produced for every three moles of H2 generated. This represents an emission rate of 7333 g CO2 per kg H2 or 0.61 tonC/GJ, which results in a carbon cost of 12.22 $/tonC. In the Schlesinger and Brown process, four moles of H2 are needed to generate one mole of NaBH4. It can be that the H2 used originates from the steam reforming of bio-methanol. In this case, for the four moles of H2 needed to generate a mole of NaBH4, there will be 1.33 mol of CO2 produced. This still results in a CO2 emission rate of 0.61 tonC/GJ or a carbon cost of 12.22 $/ton C. 267 C.7.5 Economic analysis summary The results of the simplified cost analysis are shown in Table C.3. It gives a general idea of the relative cost ranking of the different processes. As expected, the most expensive H2 production process is the sodium borohydride hydrolysis, followed by the methanol steam reforming process, the methanol electrolysis and the water electrolysis, which is the cheapest. The H2 production cost could play a key role in deciding which of the process evaluated has an overall advantage. It appears that the cost of the methanol electrolysis process would cost about twice the cost of water electrolysis. This cost difference is largely due to the fact that producing H2 from a sustainable biomass based methanol is substantially higher than producing H2 from the electrolysis of water. The most interesting outcome of this high level economic analysis is that it revealed that the methanol and water electrolysis H2 production cost could meet the 2008 US DOE H2 pre-tax target at refuel pump station is $US3/kg H2 or $US250/GJ [31]. 268 Table C.3: Cost comparison of four H2 generation process schemes. Process Major Steps CCAP.FCR ($/GJ) CF ($/GJ) CE ($/GJ) CC * PC ($/GJ) CH2 ($/GJ) Methanol from biomass 95 1 2 Methanol Electrolyser Electrolysis 59 21 9 12 199 Water Electrolyser Electrolysis 59 - 26 - 85 Methanol from biomass 95 1 2 Methanol Steam Reforming Steam reformer, shift reactor, PSA 29 21 167 12 326 Schlesinger and Brown process 116 465 Sodium Borohydride Hydrolysis Catalytic reactor 19 19167 - 12 19779 C.8 Conclusion The electrolysis of methanol or the direct methanol electrochemical reformer might represent an economical way to produce H2 for use in PEM fuel cells and other energy-consuming systems under certain conditions and might be more suitable for portable applications rather than stationary ones. However, considering the high level of this comparison and the large number of assumptions made to carry out this efficiency and economic analysis, there is some uncertainty in the estimated values, and one should be cautious in ranking the concepts compared based on this evaluation. A more precise evaluation would include mass and heat balance modelling of each H2 generation process flow sheets, using Hysis or Aspen. Based on the resulting dimensions of the unit operations, a proper statistical economic analysis could be carried out. However, this type of detailed analysis lies outside of the scope of the proposed doctorate thesis. 269 C.9 References [1] J. M., Ogden, M. M. Steinbugler, T. G. Kreutz, A comparison of hydrogen, methanol, and gasoline as fuels for fuel cell vehicles: implications for vehicle design and infrastructure development, J. Power Sources, 19, (1999) 143-168. [2] R. G. Lemus, and J. M. M. Duart, Updated hydrogen production costs and parities for conventional and renewable technologies, International J. of H2 energy, in press, (2010) 1-8. [3] D. A. Bowen, F. Lau, R. Zabransky, R. Remick, R. Slimane, S. Doong, and M. Paster, “Techno- Economic Analysis of Hydrogen Production by Gasification of Biomass”, Hydrogen, Fuel Cells, and Infrastructure Technologies, Progress Report, (2003) 4 pp. [4] D. C. C. Barranon, "Methanol and Hydrogen Production, Energy and Cost Analysis", Master's Thesis, Department of Applied Physics and Mechanical Engineering, Division of Energy Engineering, (2006) 58 pp. [5] R. M. Mohring, and Y. Wu, American Institute of Physics Conference Proceedings, 671, (2003) 90- 100. [6] Z. Hu, M. Wu, Z. Wei, S. Song and P. K. Shen, “Pt-WC/C as a Cathode Electrocatalyst for Hydrogen Production by Methanol electrolysis“, J. Power Sources, 166, (2007) 458-461. [7] J. W. Phair and S. P. S. Badwal, “Review of Proton Conductors for Hydrogen Separation”, Ionics, 12, (2006) 103-115. [8] R. O. Idem and N. N. Bakhshi, “Primary: Production of Hydrogen from Methanol. 2. Experimental Studies”, Industrial Engineering Chemistry Research, 33(9), (1994) 2056-2065. [9] D. B. Myers, G. D., Ariff, B. D. James, J. S. Lettow, C. E. Thomas and R. C. Kuhn, “Cost and Comparison of Stationary Hydrogen Fuelling Appliances”, Proceedings of the 2002 US DOE Hydrogen Program Review, NREL/CP-610-32405, (2002) 8 pp. [10] S. C. Amendola, S. L.Sharp-Goldman, M. Saleem Janjua, N. C. Spencer, M. T. Kelly, P. J. Petillo, and M. Binder, “A Safe, Portable, Hydrogen Gas Generator using Aqueous Borohydride Solution and Ru Catalyst”, Int. J. H2 Energy, 25, (2000) 969-975. [11] Y. Yoshisa, K, Diwaki, Y. Matsumura, R. Matsuhashi, D. Li, H. Ishitani, and H. Komiyama, "Comprehensive comparison of efficiency and CO2 emissions between biomass energy conversion technologies - position of supercritical water gasification in biomass technologies", Biomass and Energy, 25, (2003) 257-272. [12] C. N. Hamelinck, and A. P. C. Faaij, "Future Prospects for Production of Methanol and Hydrogen from Biomass", J. Power Sources, 111, (2002) 1-22. [13] D. Gervasio, "Fuel Cell for Hand-Carried Portable Power", Presentation, Arizona State University, Incheon, Korea (2005). [14] J. Larminie, and A. Dicks, “Fuel Cell Systems Explained”, John Wiley & Sons Ltd., UK, Second Edition, (2003), 406 pp. [15] K. A. Adamson, P. Pearson, "Hydrogen and Methanol: A Comparison of Safety, Economics, Efficiencies and Emissions", J. Power Sources, 86, (2000) 548-555. [16] A. Haryanto, S. Fernando, N. Murali, and S. Adhikari, "Current Status of Hydrogen Production Techniques by Steam Reforming of Ethanol: A Review", Energy and Fuels, 19 (2005). 270 [18] M. A. Rosen, "Thermodynamic Investigation and Comparison of Selected Production Processes for Hydrogen and Hydrogen-Derived Fuels", Energy, 21(12), (1996) 1079-1094. [19] BCS, Incorporated, U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, “Mining, Industry of the Future, Energy and Environmental Profile of the US Mining Industry", Chapter 3, Potash, Soda Ash and Borates, (2002) 3-1. [20] Y. Wu, M. T. Kelly, and J. V. Ortega, "Review of Chemical Processes for the Synthesis of Sodium Borohydride", Millennium Cell Inc., Under DOE Cooperative Agreement DE-FC36-04GO14008, (2004) 21 pp. [21] G. Brinkman, “Economics and Environmental Effects of Hydrogen Production Methods”, (2003) 29 pp. [22] T. Oi, and K. Wada, “Feasibility Study on Hydrogen Refuelling Infrastructure for Fuel Cell Vehicles Using the Off-peak Power in Japan”, Int. J. H2 Energy, 29, (2004) 347-354. [23] C.E.G. Padro, and V. Putsche, "Survey of the Economics of Hydrogen Technologies", National Renewable Energy Laboratory, NREL/TP-570-27079, (1999). [24] V.C.Y. Kong, F. R. Foulkes, D.W.Kirk, J.T. Hinatsu, Int. J. H2 Energy, 24, (1999) 665-675. [25] P.S.E. Cave, “Why the borate industry must respond to the market in order to avoid substitution”, Editörün Görüşü, (2007), http://www.boraxtr.com/boraxtr/Etihol/cave2.htm. [26] B. Johansson, "The Economy of Alternative Fuels when Including the Cost of Air Pollution", Transportation Research, Part D4, (1999) 91-108. [27] BC Hydro for Generations - Involvement in Education, http://www.bchydro.com/education/8-12/8- 12_2676.html. [28] R. H. Williams, E. D. Larson, R. E. Katofsky, and J. Chen, “Methanol and Hydrogen from Biomass for Transportation”, Energy for Sustainable Development, 1, (1995) 18-34. [29] Y. Wu, “Hydrogen Storage via Sodium Borohydride”, Presentation, GCEP Hydrogen Workshop, Stanford University, CA, USA, (2003), http://gcep.stanford.edu/pdfs/hydrogen_workshop/Wu.pdf. [30] R. U. Ayres, and P. M. Weaver, “Eco-Restructuring: Implications for Sustainable Development”, United Nations University Press, (1998), http://www.unu.edu/unupress/unupbooks/uu24ee/uu24ee0f.htm. [31] Department of Energy (DOE), “Controlled H2 Fleet and Infrastructure Demonstration and Validation Project”, Draft Revision 3, (2003) A1-A25. 271 Appendix D: Supplemental EPOC and EPOE Theory D.1 EPOC rules Promoters are usually not consumed during a catalytic reaction. If a promoter is consumed, it is referred too as a sacrificial promoter. An electropositive promoter is an electron donor and causes a decrease in the catalyst work function, while an electronegative promoter is an electron acceptor and causes an increase in the catalyst work function. Some of the most important EPOC rules are listed below [1]: Rule 1: Electropositive (electron donor) promoters enhance the chemisorption of electron acceptor adsorbates and weaken the chemisorption of electron donor adsorbates. Rule 2: Electronegative (electron acceptor) promoters enhance the chemisorption of electron donor adsorbates and weaken the chemisorption of electron acceptor adsorbates. Rule 3: In the presence of a strong electron donor (electropositive) adsorbate (e.g. K, Na) a weaker electron donor (e.g. NO on Pt(111)) behaves as an electron acceptor. Rule 4: In the presence of a strong electron acceptor (electronegative) adsorbate (e.g. O) a weaker electron acceptor (e.g. CO on Pt(111)) behaves as an electron donor. Rule P1: If a catalyst surface is predominantly covered by an electron acceptor reactant (e.g. O) then an electron acceptor promoter (e.g. O2-) is recommended. Rule P2: If a catalyst surface is covered predominantly by an electron donor reactant (e.g. C6H6, C2H4), than an electron donor promoter (e.g. Na+) is recommended. 272 Rule P3: If both reactants are weakly adsorbed on the catalyst surface, then both electron acceptor and electron donor promoters can enhance the rate. Rule F1: Increasing work function (via the addition of electronegative promoters) strengthens the chemisorptive bond of electron donor adsorbates and weakens the chemisorptive bond of electron acceptor adsorbates. Rule F2: Decreasing work function (via the addition of electropositive promoters) weakens the chemisorptive bond of electron donor adsorbates and strengthens the chemisorptive bond of electron acceptor adsorbates. Rule G1: A reaction exhibits purely electrophobic behaviour when the kinetics are positive order in the electron donor reactant and negative or zero order in the electron acceptor reactant. Rule G1’: A reaction exhibits purely electrophobic behaviour when the electron acceptor reactant is strongly adsorbed and much more strongly adsorbed on the catalyst surface than the electron donor reactant. Rule G2: A reaction exhibits purely electrophilic behaviour when the kinetics are positive order in the electron acceptor reactant and negative or zero order in the electron donor reactant. Rule G2’: A reaction exhibits purely electrophilic behaviour when the electron donor reactant is strongly adsorbed and much more strongly adsorbed on the catalyst surface than the electron acceptor reactant. Rule G3: A reaction exhibits volcano-type behaviour when both the electron donor and electron acceptor are strongly adsorbed on the catalyst surface. Rule G4’: A reaction exhibits inverted volcano-type behaviour when the kinetics are positive order in both the electron acceptor and the electron donor reactant. 273 Rule G5: The above rules G1-G4 apply also when the electron donor and the electron acceptor are both electron acceptors or electron donors. In this case the electron donor is always the stronger electron donor or weaker electron acceptor and the electron acceptor is always the weaker electron donor or stronger electron acceptor. Rule G6: A mononuclear reaction is electrophobic for an electron donor adsorbate and electrophilic for an electron acceptor adsorbate. Rule G7: The maximum rate modification obtained under electrochemical promotion conditions increases for every fixed overpotential with increasing difference in the electron acceptor-electron donor character of the two reactants. D.2 Chemisorption on a transition metal When adsorbed on a transition metal, CO can exhibit electron acceptor or electron donor characteristics. The example illustrated in Fig. D.1 is based on quantum mechanical calculations [2]. Fig. D.1 (a) shows the sp and d bands of a transition metal, such as Pt, i.e., the density of states (DOS) as a function of electron energy, E. The outer orbital energy levels of a gaseous CO molecule is also shown: orbitals 4σ, 1π and 5σ are occupied as indicated by the arrows, while orbital 2π* is empty [2]. The geometry of these molecular orbitals is shown in Fig. D.1 (b). Fig. D.1 (c) and (d) show the energy and density of states of the resonances (adsorbed molecular orbitals) formed upon CO adsorption due to the interaction of the 2π* orbitals and 5σ orbitals with the metal surface [2]. 274 Figure D.1: CO chemisorption on a transition metal. Molecular orbitals and density of states before (a, b) and after (c, d). Effect of varying Φ and EF on electron back donation (c) and donation (d) (Modified from Vayenas et al. [1] with permission from Springer). Fig. D.1 (c) shows the electron backdonation interaction, i.e., electrons are transferred from the Fermi level of the metal to the hydridized 2π* molecular orbital which was originally [2]. Fig. D.1 (d) shows the electron donation interactions, i.e., electrons are transferred from the initially fully occupied 5σ molecular orbitals to the Fermi level of the metal [2]. Fig. D.1 (c), demonstrates the electron acceptor character for which EF decreases and Φ increases. Lowering the work function or increasing the Fermi level enhances electron backdonation to the 2π* orbitals and diminishes electron donation from the 5σ orbitals. Fig. D.1 (d) demonstrates the electron donor character for which EF increases and Φ decreases. It is now experimentally and theoretically established that CO behaves as an electron acceptor on most (a) (b) Energy E density of states transition metals sp - band Electron acceptor coadsorbate Electron donor coadsorbate EF d - band ∆UWR < 0 ∆Φ < 0 ∆UWR > 0 ∆Φ > 0 CO in vacuum 4σ 1π 5σ 2π* Energy E density of states EF e- (c) (d) Energy E density of states EF e- 2π* 5σ 2π* 2π* 2π*2π* 5σ 1π 1π 275 transition metal surfaces [2]. The enhance backdonation of electrons to the 2π* orbitals results in a strengthening of the Pt=CO bond, which eventually may lead to the weakening in the C=O bond [2]. D.3 Kirchhoff’s first law for triode and tetrode operation Equation 3.1 originates from the one used in triode fuel cell operation [3]. For the case of parallel electrolysis operation, the sign of the electrolyser current would be the same as the sign of the auxiliary circuit current (+). In triode operation, the electrolyser current would be greater than in normal electrolysis. In the case of reverse electrolysis operation, the sign of the electrolyser circuit current (+) would be opposite to the sign of the auxiliary circuit current (-). According to this, in triode operation, the electrolyser circuit current would be less than that in normal electrolysis. Please note that in Eq. 3.1, the Faradaic current was defined as the total fuel-consuming current and not a cathodic current as the reviewer is suggesting. It is stated in the paper that this equation is based on Kirchhoff’s first law. According to this law, the current entering a point must be equal to the current leaving the point. The current always flow from the positive terminal to the negative one. The triode and tetrode diagrams below (Fig. D.3.2 to D.3.5) illustrate how Eq. 3.4 was derived for the fuel cell and electrolyser cases. D.3.1 Triode fuel cell In a fuel cell, the anode is negative, the cathode is positive and current is generated at the WEfc (out). Figure D.2: Application of Kirchhoff’s first law for fuel cell triode operation. WEfc anode (-) CEaux Cathode (+) CEfc IFar Iaux Ifc = Iaux + IFar Iaux IFar IfcWEfc IFar 276 D.3.2 Triode electrolyser In electrolysis, the anode is positive, the cathode is negative and current is needed at the WEelec (in). Taking into account the signs, for parallel operation, the auxiliary current and the electrolyser current are both positive. Figure D.3: Application of Kirchhoff’s first law for triode electrolysis in parallel operation. In reverse operation, the electrolyser current is positive and the auxiliary current is opposite. Figure D.4: Application of Kirchhoff’s first law for triode electrolysis in reverse operation. D.3.3 Tetrode electrolyser In reverse tetrode operation, the electrolyser current is positive and the auxiliary current is opposite, like in the reverse triode case. The only difference is that two separate counter electrodes are used and that the electrolyser and the auxiliary circuits are completely independent. Common CE WEaux WEelec IFar Iaux Iaux + Ielec = IFar Ielec = IFar - Iaux or Ielec = IFar + Iaux, with Iaux as a negative value for reverse operation. Iaux IFar Ielec WEelec WE IFar CE cathode (-) WEaux WEelec anode (+) WEelec IFarIaux Ielec = Iaux + IFar Iaux IFar Ielec WEelec IFar 277 Figure D.5: Application of Kirchhoff’s first law for tetrode electrolysis in reverse operation. D.4 References [1] C. G. Vayenas, S. Bebelis, C. Pliangos, S. Brosda and D. Tsiplakides, “Electrochemical Activation of Catalysis - Promotion, Electrochemical Promotion and Metal-Support Interactions”, Kluwer Academic / Plenum Publishers, (2001) 574 pp. [2] M. T. M. Koper, and R. A. van Santen, “Electric field effects on CO and NO adsorption at the Pt(111) surface”, J. Electroanal. Chem., 476, (1999) 64-70. [3] S. P. Balomenou, and C. G. Vayenas, Triode Fuel Cells and Batteries, Journal of The Electrochemical Society, 151 (11), (2004) A1874-A1877. WEaux WE WEelec IFar Iaux Iaux + Ielec = IFar Ielec = IFar - Iaux or Ielec = IFar + Iaux, with Iaux as a negative value for reverse operation. Iaux IFar Ielec WEelec CEaux CEelec CE IFar 278 Appendix E: Experimental procedures E.1 Catalyst electrode preparation E.1.1 Cathode carbon sublayer composition calculation 1. Choose a desired carbon sublayer loading (LCsub). It is typically 1 mg/cm2 2. Choose the fraction of PTFE required (fPTFE). It is typically 20 wt% (i.e. 0.2) 3. Calculate the amount of carbon (mcarbon) required. For the automated sprayer allow for an excess of 2x, for hand spraying allow for an excess of 3x ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛= mg gALgm GeomCsubCarbon 1000 12)( (E.1) 4. Calculate the amount of PTFE solution required. PTFE solution typically supplied at 60% ⎟⎠ ⎞⎜⎝ ⎛ −= 6.0 1 1 )(_ PTFE CarbonPTFE SolPTFE f mfgm (E.2) 5. Clean an appropriate-sized beaker with isopropanol 6. Using an analytical balance, weigh the amount of carbon and PTFE solution calculated in step 3. & 4. and add to the beaker 7. Add water and isopropanol to the beaker 8. Follow the procedure for the automated sprayer 9. Oven dry for 40 min. at 350oC 279 Table E.1: Cathode carbon sublayer calculation spreadsheet. Carbon in Sublayer Required Cathode Desired Loading [mg/cm2] 1 Area [cm2] 121 Carbon [wt%] 1 pure carbon powder Excess 3 hand spraying excess m carbon [g] 0.363 PTFE in Sublayer Required Cathode Desired f PTFE 0.2 % PTFE in Solution 0.6 bottle specification m PTFE [g] 0.15125 Final Mass of GDL w/ Sublayer Cathode m GDE (initial) [g] 2.12196 weight m GDE (final) [g] 2.27321 target mass m GDE (Actual) [g] 2.25556 mass after drying Actual Loading [mg/cm2] 0.88331 E.1.2 Anode and cathode catalyst ink composition calculation 1. Choose a desired catalyst loading (Lcat = mg/cm2) and Nafion® wt% (fNafion) 2. Calculate the amount of supported catalyst (mcat-VulcanXCR) based on the total weight percentage of the purchased catalyst. For the automated sprayer allow for an excess of 2x, for hand spraying allow for an excess of 3x ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛=− mg g Wt ALgm cat GeomcatVulcanXCRCat 1000 1 % 12)( (E.3) 3. Calculate the amount of Nafion® solution required. Nafion® solution typically supplied as 5 wt% ⎟⎠ ⎞⎜⎝ ⎛ −= + 05.0 1 1 )(_ Nafion VulcanXCRCatNafion SolNaf f mf gm (E.4) 4. Clean an appropriate sized beaker with isopropanol 280 5. Using an analytical balance, weigh the amount of catalyst calculated in step 2. and add to the beaker 6. Add enough DI water to cover the supported catalyst powder 7. Using an analytical balance, weigh the amount of Nafion® solution calculated in step 3. and add to the beaker 8. Add isopropanol to the beaker 9. Sonicate the catalyst ink solution for 90 minutes 10. Follow the procedure for the automated sprayer 11. Mass of substrate with loading can be expressed by the following: iGDE Nafioncat Geomcat fGDE mfWtmg ALgm ,, )1(%1000 )( +−×× ×= (E.5) Table E.2: Anode catalyst ink composition calculation spreadsheet. Catalyst Required Anode Nafion Required Anode Desired Loading [mg/cm2] 4 Desired f Nafion 0.3 Area [cm2] 121 m Nafion® soln [g] 62.23 Catalyst [wt%] 0.2 bottle specification Excess 3 hand spraying excess m cat + vulcan [g] 7.26 Final Mass of GDE Anode m GDE (initial) [g] 2.19 weight m GDE (final) [g] 5.64 target mass m GDE (Actual) [g] 5.65 mass after drying Actual Loading [mg/cm2] 4 1 4 8 Total m cat + vulcan 1.82 7.26 14.52 9.08 m Nafion® 15.56 62.23 124.46 77.786 281 Table E.3: Cathode catalyst ink composition calculation spreadsheet. Catalyst Required Cathode Nafion Required Cathode Desired Loading [mg/cm2] 2 Desired f Nafion 0.3 Area [cm2] 121 m Nafion® soln [g] 31.11429 Catalyst [wt%] 0.2 bottle specification Excess 3 hand spraying excess mM cat + vulcan [g] 3.63 Final Mass of GDE Cathode m GDE (initial) [g] 2.25556 (GDE w/ Sublayer) m GDE (final) [g] 3.98413 target mass m GDE (Actual) [g] 3.99312 mass after drying Actual Loading [mg/cm2] 2.0104 E.1.3 Catalyst ink preparation procedure 1. On the Pt-loading spread sheet a. Insert desired loading (1, 2, 4 mg/cm2) b. Insert desired area (121 cm2) (11 x 11 cm = max. plate area) c. Insert mass or required supported catalyst (20%) d. Insert desired Nafion® wt% (30%) 2. Mix in a beaker a. Add catalyst powder as required by the spreadsheet (M cat + Vulcan (g)) b. Add water just enough to cover the catalyst powder and protect the carbon from isopropyl alcohol attack c. Add Nafion® solution as required by the spreadsheet (M Nafion® soln (g)) d. Add isopropyl alcohol (IPA, CH3CHOHCH3) (enough so it does not stick) 3. Parafilm beaker and sonicate for 90 minutes 4. Bring the balance to spraying area in the fumehood 282 E.1.4 Electrode spraying procedure 1. Weight CFP (TGPH-060) & record as MGDE(initial) 2. Clean out spray gun, plate and ink reservoir with IPA 3. Place CFP between metal frame and tape to the metal frame with masking tape to ensure it does not move 4. Turn on hot plate to ~80oC for one hour if spraying catalyst or at 100oC for one hour if spraying a carbon sub-layer 5. Attach gun to air line, open air fully and set pressure to ~ 15 psi (14.696 psi or 1,1013 bar) adjust spraying while holding sprayer air trigger while adjusting spraying pressure 6. Ensure nozzle is parallel to the handle or to the spraying direction 7. Test spray gun adjustment on paper towel 8. Spray catalyst Ink a. Adjust first back knob clockwise to make spray finer (adjust spray to ~ 2 cm) b. First stage of trigger is air. Second stage of trigger is catalyst ink. c. Load reservoir with about 2/3 of the solution d. Spray evenly (turn sheet around for even coating distribution) e. Use air pass between each spraying passes for better drying f. Change metal frame and GDL orientation every two pass 9. Weight GDL & repeat step 9-10 as needed 10. Dry for one hour at 80oC in an oven (or at 350oC for 40 min. for sub-layer preparation) 11. Turn off air and bleed air gun 12. Clean air gun tip, remove with wrench, sonicate tip in IPA for one hour 13. 17. Determine final GDL weight 283 E.2 Platinization procedure This platinization procedure was extracted from Janz and Ives “Reference Electrodes” [1]. 1. Clean substrate with warm concentrated nitric acid. It is also recommended to briefly immerse the substrate in dilute aqua regia: 1) 4 volumes of DI water 2) 1 volume of 16 M (~70%) HNO3 3) 3 volumes of 12 M (~40%) HCl So for 24 ml of solution, one needs 12 ml of DI water, 3 ml of 16 M HNO3, and 9 ml of 12 M HCl. If dilute aqua regia is used, it must be followed by treatment with concentrated HNO3. Wash the substrate with DI water prior to platinization. 2. Cathodic cleaning prior to platinization is also recommended to remove the platinum oxides (PtOx). This is performed by using the substrate as the cathode for water electrolysis in dilute sulphuric acid using a 9 volt battery for a period of ten minutes. Use a Pt wire as the cathode and a Pt mesh as the anode. Hydrogen should be evolved in small bubbles over the cathode’s whole surface. If not, the surface is not clean enough. Figure E.1: Electrolysis in dilute sulphuric acid. A, + C, - 9 V 0.5 M H2SO4 H2 O2 PtO 284 3. Lead-free platinization can be carried out in a solution composed of 0.05 M H2PtCl6(H2O)6 (1.41 g per 50 ml) and 2 M HCl (8.36 ml 36.5% HCl per 50 ml) at a current density of about -10 to -20 mA/cm2 (-0.01 to -0.02 A/cm2) for 10-20 min using the Multistat. The cathode is the working electrode. The deposit is a hardly visible grey or golden film and should remain sufficiently active for several weeks. Store in DI water. Never transfer a dry platinized electrode directly from air to hydrogen. It will deactivate completely. Figure E.2: Lead-free platinization. E.3 Silicon ring preparation Air was removed from the mixture before it was poured in the grooves of the seal plate moulds, which was then cured by heating at 100oC for an hour or drying overnight. The following points summarize the steps that have to be followed in order to make silicone seals: 1. Get the JRTV silicone moulding rubber and curing agent “J” 2. The ratio between the silicone and the curing agent is 10:1. Use 10:0.6 to make softer seals. A C, WE 0.05 M H2PtCl6(H2O)6 2 M HCl Pt + - H2 Multistat Galvanostatic Mode -0.02 A/cm2 285 3. Grab a small plastic container and weight it on the balance. “Zero” de balance once the container has been weighted. 4. Pour some curing agent (around 1.6 grams) and weigh it on the balance. 5. Once the curing agent is weighed, “zero” the balance and put the corresponding amount of silicone inside the plastic container (i.e. around 16 grams). Make sure that the ratio is met (10:1) as accurate as possible by weighing the total content. Mix the two substances. 6. Put the mix in the oven/vacuum, turn the vacuum pump on and pressurize the chamber to around 30inHg. At this point the silicone will expand and return to its normal size. This is done so that no air bubbles are in the silicone. Make sure that you do not put a lot of silicone and curing agent inside the container to avoid any spills when they expand in the vacuum. 7. Repeat steps 3 to 5 (using other containers) based on how much silicone is necessary. For the seal 4cm2 fuel cell around 17-20 grams of the final mixture are needed. 8. Take all the seal plates (cleaned with acetone) and pour the silicone onto the grooves (a small spatula or/and a syringe could be used). Make sure that the silicone covers all the grooves. 9. If any bubbles are seen put more silicone beside them (not on top of them). Once all the silicone has covered all the corresponding surfaces and no bubbles are evident, use a straight edge (or spatula) and slide it over the plates in order to remove all the silicone excess. 10. The silicone can cured overnight or if it is necessary the plates can be heated up to 100ºC for around 1 hour. 11. Once the silicone is cured, remove the seals slowly without breaking them. 12. Clean the mold plates with Acetone. E.4 Nafion membrane conditioning 1. Cut membrane in squares 2. Boil in 3% H2O2 for 30 min. 3. (3.4% H2O2, bp. of pure H2O2 = 151.4oC 4. Rinse with DI H2O 5. Boil in DI H2O for 30 min. 286 6. Rinse with DI H2O 7. Boil in 0.5 M H2SO4 for 30 min. 8. Store in DI H2O 9. Cut membranes in circles with the die-cutter E.5 Determination of CH3OH concentration The concentration of alcohol crossover to the cathode in a DMFC can be determined using various techniques. First, the amount of alcohol crossover at the cathode can be established by determining the steady-state concentration of CO2 in the cathode exhaust by gas chromatography [2]. From the amount of CO2 produced at the cathode exhaust, it is possible to determine the oxidation current density equivalent to the alcohol crossover. Another common way of determining the alcohol crossover is to use an optical infra-red CO2 sensor to monitor the flow of CO2 in the cathode exit gas. However, this method requires precise calibrations, and overestimation of the alcohol crossover is possible at high current density [3, 4]. This is because, in a DMFC, the total cathodic CO2 includes a non-negligible amount of permeated CO2 from the anodic electro-oxidation [5, 6]. In the liquid phase, it is possible to measure alcohol partitioning and transport in the proton-conducting membrane under the electrochemical system operating conditions by using voltammetric methods while correcting for electro-osmotic drag effects [4]. In this method, the diffusion coefficient and CH3OH concentration in the membrane were determined by measuring the transient limiting current density following a potential step. The rate of CH3OH transport induced by electro-osmotic drag can also be estimated by using a limiting-current technique [7]. A potentiometric method was also developed to determine the alcohol crossover [3]. In this method, the potential of the catalyst working electrode is recorded during the CH3OH crossover. It was shown that the slope of the plot of the potential vs. time is proportional to the crossover rate. Crossover determinations were also carried out by using a dynamic hydrogen electrode (DHE) on the cathode for single electrode potential measurements [8]. The DHE allowed for the measurement of the mixed potential due to a simultaneous CH3OH oxidation and O2 reduction in a DMFC. 287 Methanol permeability through a membrane can be determined in a two-compartment diffusion cell by using a refractive index infra-red (I-R) detector [7, 9]. One compartment is filled with a mixture of alcohol and DI water while the other is filled with DI water. It is connected to a computer with software which calculates methanol permeability. A Waters IR detector model 2414 and a Shimadza liquid chromatograph pump are available. However, the internals are all made of SS and hence, it cannot be used with an acid electrolyte, which would be present in the systems presently studied. In the gas phase, methanol crossover could be determined by purging the cathode compartment with N2 or Ar and analyzing the gas composition through the use of a previously calibrated gas chromatograph, if column contamination from H2SO4 is not a concern. If it is, the cathode compartment gas would have to be condensed and analysed in the liquid phase. High-performance liquid chromatography (HPLC) could be used to determine the CH3OH concentration in the cathode compartment as a function of time and for different electrode potentials [10]. More conveniently, there exists a method to determine the methanol content of aqueous solutions based the reduction of potassium dichromate (K2Cr2O7) and using a UV spectrophotometer. The ionic reduction of potassium disulphate is: 3CH3OH + 2Cr2O72- + 16H+ → 3HCOOH + 4Cr3+ + 11H2O (E.6) A sample is added to excess K2Cr2O7 in acid and incubated at 65oC for 30 minutes. As the reaction proceeds, the orange color of Cr2O72- decreases while the blue-violet color of Cr3+ increases. A spectrophotometer is used to measure either the disappearance of Cr2O72- at a wavelength of 430 nm or the appearance of Cr3+ at a wavelength of 580 nm. The concentration of CH3OH in the solution is determined by using Beer’s law: ClA ε= , (E.7) 288 where A is the dimensionless absorbance, ε is the molar extinction coefficient (l/mol.cm), C is the concentration (M) and l is the path length of the light in the solution (cm). The value of A is directly proportional to the aqueous CH3OH concentration up to 1 wt%, after which A remains about constant for all CH3OH concentrations. Hence, if the CH3OH concentration is too high, the solution will need to be diluted to preferably close to 0.5 wt% in order to use this method effectively. In addition to being used for monitoring crossover in the electrochemical reformer, this method could also be used to determine the change in CH3OH concentration in an aqueous acidic CH3OH solution over time, at a temperature close or greater than the pure alcohol’s boiling point. The detailed procedure is as follows: 1. Prepare 500 ml of a stock solution of 0.04 M K2Cr2O7 in 3 M H2SO4 a) Weigh 5.8838 g of the K2Cr2O7 salt and dissolve in a 500 ml volumetric flask b) Slowly add 80 ml concentrated H2SO4 and dilute close to the mark with de-ionized water. Be careful not to allow the flask to get hot c) After allowing the solution to cool to room temperature, dilute with the de-ionized water to the mark 2. Dilute the sample(s) into a 50 ml or 100 ml volumetric flask with de-ionized water so as to ensure that the resulting methanol concentration lies between 0 and 1 wt%, preferably near 0.5 wt% 3. Pipette 2.5 ml into a 25 ml volumetric flask and dilute to the mark with the prepared 0.04 M K2Cr2O7 in 3 M H2SO4 4. Prepare a reagent blank in the same manner using 2.5 ml de-ionized water instead of the diluted samples 5. Place both solutions into a water bath at 65oC for at least 30 minutes but not more than an hour, and then allow to cool to room temperature 289 A UV Vis spectrophotometer model 1240 by Shimadza was used, and the K2Cr2O7 was a reagent plus grade from Sigma-Aldrich. It was calculated that 29.78 ml H2O needed to be added to a 2.5 ml aliquot sample of a 2 M CH3OH solution in 0.5 M H2SO4 in order to get a concentration of 0.5 wt% CH3OH. Using solutions of known CH3OH concentrations (0 to 2 M) and H2SO4 concentrations (0 to 0.5 M), it was established that this CH3OH concentration determination technique gave results within 10%. The random error was within 9% on many multiple dilutions of a same sample. E.6 RDE catalyst ink preparation 1. RDE preparation: a. Clean RDE b. Polish the RDE with 0.05 mm Al2O5. If electrode is rough start polishing with the 3 mm and go down to 1 mm, 0.3 mm and 0.05 mm. c. Dry 2. Materials: a. 40 mg of nanocatalyst powder b. Nafion ionomer (~200 mg 5% Nafion solution). Use 10 mg of to obtain a Nafion content of 20 w%. c. IPA 3. Procedure: a. Weigh nanocatalyst powder in a disposable beaker b. Disperse the nanocatalyst in a volume of solvent i. Add DI water to wet the nanocatalyst powder ii. Add 20 to 40% IPA and sonicate to disperse the nanocatalyst powder (Typically 500 to 1000 μg nanocatalyst/ml solvent) c. Add 200 mg 5% Nafion ionomer solution and IPA to reach a total volume of 5 ml (Typically 32ml of Nafion ionomer for 8 ml of catalyst solution) d. Sonicate 60 min. to create an uniform dispersion 290 e. Drop coat the RDE using a micro-pipette (5 μL of ink) and dry in air for 30 min. in a casting enclosure (Typically 20 μl of dispersion for a 5 mm OD disk electrode) f. Heat dry in air for an hour or in an oven at 50-80oC for 10 min. After solvent evaporation, a thin Nafion film (~0.1 μm) should remain on the disk. Estimate the amount of catalyst in the film from the concentration of the catalyst in the dispersion and the volume dispensed. Thick films (> 0.5 μm) should be avoided to prevent interference with ideal solution flow at a RDE 4. Cleaning: a. wipe with a Kimwipe dipped with acetone b. Polish the RDE with 0.05 mm Al2O5 E.7 Pt-Ru/TiO2 preparation 1. Impregnation procedure for 0.25 g Pt-Ru/TiO2 20 wt% (g metal or catalyst) 1:1 a/o 1. Add TiO2 to 20 ml aqueous solution and sonicate for an hour 2. Add the Ru precursor salt, adjusted to the desired metal loading and molar ratio 3. Sonicate the mixture for an hour 4. Add the Pt precursor salt, adjusted to the desired metal loading and molar ratio 5. Sonicate the mixture for an hour 6. Heat the impregnated slurry to 80oC under continuous stirring until complete water evaporation 7. Leave sample in the beaker, cover with Al foil, and dry in over overnight at 110oC 8. Place catalyst in a ceramic boat. 9. Follow-up with reduction heat treatment (See procedure for reduction in 10% H2 in Ar) 2. Chemical reduction procedure for 0.25 g Pt-Ru/TiO2 20 wt% (g metal or catalyst) 1:1 a/o 1. Weigh 0.2 g TiO2 2. Sonicate the TiO2 in 100 ml Millipore DI water for an hour 291 3. Add 0.08591 g of H2PtCl6.6H2O (38.34%Pt) and 0.04265 g of RuCl3.nH2O (40.01%Ru) precursors 4. Sonicate the mixture for an hour 5. Heat the suspension to 80oC (225oC on hot plate) in an oil bath under constant stirring and bubbling of Ar (or N2) with reflux condenser 6. Prepare the reducing agent by dissolving 95 mg NaBH4 and 200 mg NaOH in 25 ml Millipore DI water 7. After cooling to room T, add excess reducing agent drop-wise over a period of 2h while sonicating the suspension with N2 bubbling 8. After the reduction process, filter and wash the catalyst several times with pure water and dry at 110oC overnight 9. Place the catalyst in a ceramic boat 10. Follow-up with heat treatment (See procedure for calcination in Ar) Table E.4: Precursor salt addition calculation spreadsheet for the impregnation method. Description Quantity Units Material Dry Pt content 47.6052 wt% H2PtCl6 Dry Ru content 48.7249 wt% RuCl3 Wet Pt content 38.05 wt% H2PtCl6.6H2O Wet Ru content 39 wt% RuCl3.nH2O Final mass of catalyst 0.25 g Pt-Ru/TiO2 Final metal content 0.2 wt (%) Pt-Ru Mass of metal 0.05 g Pt-Ru Mass of support 0.2 g TiO2 Mol of metal 1.69E-04 mol Pt-Ru Pt:Ru a/o ratio 1 Pt:Ru Mol of Pt 1.69E-04 mol Pt Mol of Ru 1.69E-04 mol Ru Mass of Pt 3.29E-02 g Pt Mass of Ru 1.71E-02 g Ru Mass of dry salt 0.06919 g H2PtCl6 Mass of dry salt 0.03502 g RuCl3 Mass of wet salt 0.08656 g H2PtCl6.6H2O Mass of wet salt 0.04375 G RuCl3.nH2O 292 3. Calcination in Ar 1. Insert sample from the left hand side of the tube on top of thermocouple. Center sample boat with wire and flashlight. Insert isolating wool and close tube furnace ends 2. Turn on Ar gas tank. Set Ar flow rate to full flow for purging. Check if exhaust gas is flowing and the pressure drop across tube furnace. Check for leaks with snoop, especially at the inlet. 3. Purge with Ar for a min. of three air changes (1 hour at room T) 4. Program safety s/d temperature. Program ramp-up rate (5oC/min) and dwell time (500oC for 2 hrs) 5. Reduce the Ar flow rate a set rate of 10 ml/min and start the temperature operation of the furnace 6. Once program is finished, let cool down to room T 7. Turn off Ar gas tank, open tube, retrieve sample 8. Crush particles with mortar and pestle, ball mill for one hour 4. Reduction in 10% H2 in Ar 1. Insert sample from the left hand side of the tube on top of thermocouple. Center sample boat with wire and flashlight. Insert isolating wool and close tube furnace ends 2. Turn on Ar gas tank. Set Ar flow rate to full flow for purging. Check if exhaust gas is flowing and the pressure drop across tube furnace. Check for leaks with snoop, especially at the inlet. 3. Purge with Ar for a min. of three air changes (1 hour at room T) 4. Turn on 10% H2 in Ar gas tank. Set to 10% H2 in Ar flow rate. Turn off Ar gas tank. Check for leaks with H2 sensor, especially at the inlet. 5. Purge with 10% H2 in Ar for a min. of three air changes 6. Program safety s/d temperature. Program ramp-up rate (5oC/min) and dwell time (500oC for 2 hrs). 7. Reduce the flow rate a set rate of 100 cm3 STP/min and Start the temperature operation of the furnace. 8. Once program is finished, turn on Ar gas tank and turn off 10% H2 in Ar gas tank, let cool down to room T 9. Turn off Ar gas tank, open tube, retrieve sample 293 10. Crush particles with mortar and pestle, ball mill for one hour 5. Tube furnace programming procedure: ramp to a set-point, hold for a fixed time, then turn off 1. Setup the program a. Press SCROLL button to choose ‘m – A’, use ↑↓ button to choose ‘MAN’ b. Press PAGE button until ‘SP’ is shown c. Press the SCROLL button so ‘SPrr’ is shown, then using the ↑↓ button set your ramp rate. Ramp rate units are in degrees per minute. d. Press SCROLL so that ‘tm.OP’ is displayed, then by using the ↑↓ button select Opt.1 e. Press the SCROLL button so ‘dwEll’ is shown, then using the ↑↓ button select your desired dwell time. Dwell time units are in minutes. f. Press the PAGE button until the actual temperature is displayed g. Use the ↑↓ button to set the ‘dwEll’ temperature (oC) 2. Run the program a. Press the SCROLL button until ‘StAt’ is displayed and then using the ↑↓ button choose ‘run’ b. Press the PAGE button to the show the actual temperature. Temperature should start rising now. c. Stopping the program d. Press the SCROLL button until ‘StAt’ is displayed and then using the ↑↓ button choose ‘oFF’ 294 E.8 References [1] D. J. G. Ives and G. J. Janz, “Reference Electrodes, Theory and Practice”, Academic Press Inc, New York and London, (1961) 65 pp. [2] H. Uchida, Y. Mizuno and M. Watanabe, “Suppression of Methanol Crossover in Pt-Dispersed Polymer Electrolyte Membrane for Direct Methanol Fuel Cells”, Chemical Letters, (2000) p. 1268-1269. [3] N. Munichandraiah, K. McGrath, G. K. S. Prakash, R. Aniszfeld and O. G. A. Robert, “A Potentionmetric Method of Monitoring Methanol Crossover through Polymer Electrolyte Membranes of Direct Methanol Fuel Cells”, Journal of Power Sources, 117(1-2) (2003) p. 98-101. [4] X. Ren, T. E. Springer, T. A. Zawodzinski and S. Gottesfeld, “Methanol Transport through Nafion Membranes - Electro-osmotic Drag Effects on Potential Step Measurements”, Journal of the Electrochemical Society, 147(2) (2000) p. 466-474. [5] J. A. Drake, W. Wilson and K. Killeen, “Evaluation of the Experimental Model for Methanol Crossover in DMFCs”, Journal of the Electrochemical Society, 151(3) (2004) p. A413-A417. [6] H. Dohle, J. Divisek, J. Mergel, H. F. Oetjen, C. Zingler and D. Stolten, “Recent Developments of the Measurement of the Methanol Permeation in a Direct Methanol Fuel Cell“, Journal of Power Sources, 105 (2002) p. 274-282. [7] B. Bae, H. Y. Ha and D. Kim, “Preparation and Characterization of Nafion/Poly(1-vinylimidazole) Composite Membrane for Direct Methanol Fuel Cell Application”, Journal of the Electrochemical Society, 152(7) (2005) p. A1366-A1372. [8] V. A. Paganin, E. Sitta, T. Iwasita and W. Vielstich, “Methanol Crossover Effect on the Cathode Potential of a Direct PEM Fuel Cell“, Journal of Applied Electrochemistry, 35 (2005) p. 1239-1243. [9] K. Y. Cho, J.-Y. Eom, H.-Y. Jung, N.-S. Choi, Y. M. Lee, J.-L. Park, J.-H. Choi, K.-W. Park and Y.-E. Sung, “Characteristics of PVdF Copolymer/Nafion Blend Membrane for Direct Methanol Fuel Cell (DMFC)”, Electrochim. Acta, 50 (2004) p. 583-588. [10] C. Lamy, J.-M. Leger and S. Srinivasan, “Direct Methanol Fuel Cells: From a Twentieth Century Electrochemist’s Dream to a Twenty-first Century Emerging Technology”, Modern Aspects of Electrochemistry, Edited by John O’M. Bockris et al., 34 (2001), p. 53-117. 295 Appendix F: Electrochemical Techniques F.1 Cyclic voltammetry Cyclic voltammetry (CV) is a controlled potential technique in which the potential of the working electrode is slowly and linearly cycled or scanned between a high and a low voltage limit while the resulting current is measured. A cyclic voltammogram is obtained by plotting the current measured at the working electrode versus the voltage. A three-electrode cell, like the one shown in Fig. F.1, is typically used to conduct cyclic voltammetric, studies. Figure F.1: Picture of a three-port glass cell for cyclic voltammetry. A potentiostat is used to control and apply the voltage between the WE and CE. This drives a current to flow between the WE and CE. As the voltage is scanned in the positive direction, the reduced compound is oxidized at the electrode surface, then, at a certain voltage value, the scan is reversed and the material that was oxidized is then reduced. Hence, the forward scan represents the electro-oxidation reaction and the reverse scan represents the electro-reduction reaction. At the beginning of the scan, the current is dominated by the electron kinetics. As the potential increases, the limiting current Reference electrode Counter electrodeWorking electrode 296 decreases and diffusion begins to affect the current. At the peak, both influences are equal. At higher potentials, diffusion control dominates: the concentration of species near the electrode surface is depleted, and the diffusion is too slow to replenish the species near the electrode surface. The current measured at a given potential during the anodic scan is often different from the current during the cathodic scan. Cyclic voltammetry can be used to investigate the redox behaviour of electrode reactions such as potentials and diffusion vs. adsorption behaviour, as well as to elucidate kinetics and mechanisms of electrode reactions, such as reversibility and electron transfer rates. CV can be used to determine the electrochemical surface area (ECSA) of supported or unsupported precious metal electrocatalysts. The ECSA available for the electrochemical reaction is larger than the total noble metal particle area. The three ports glass cell were first cleaned with aquea regia, a concentrated acid solution of HCl:HNO3 3:1, to eliminate possible contaminants, then rinsed with DI water, followed by a final rinse with the electrolytic solution. A catalyst coupon, acting as the WE, a platinized Pt flag CE and an MSE were immersed in the unstirred solution. The platinized Pt flag electrodes were cleaned using a concentrated solution of 30 wt% H2O2:H2SO4 1:1 while the catalyst coupon was cleaned in a 3 wt% H2O2 solution. Nitrogen gas was bubbled in the solution for 15 minutes prior to the measurements and a N2 blanket was maintained throughout the scans to prevent O2 ingress into the solution. Details on the voltammetric procedure used to determine the ECSA of the Pt/C catalyst and on the copper underpotential deposition (UPD) method used to determine the ECSA of the Pt-Ru/C catalyst follows. F.1.1 Pt/C catalyst electrochemical surface area determination The Pt/C and Pt black catalyst ECSAs were determined by recording CV’s in the presence of deoxygenated 0.5 M H2SO4 at ambient conditions [1, 2]. The circular 2 cm2 GDE sample was held in the cell by a Cu alligator clip model BU-34C from Mueller Electric Company Inc., which was gold plated by Acme Plating Inc. with a 2.5 micron sub-layer of nickel and a 5 micron layer of 24K gold. The electrode was first activated with an extended sweep to more positive values to eliminate impurities by cycling 20 times at a fast scan rate of 250 mV/s. The potential was then scanned at a potential step rate of 5 mV/s 297 between 0 and 1.2 V vs. SHE to obtain a stable CV. It is assumed that each Pt surface atom absorbs about one H2 atom as per: Pt + H+ + e- ↔ Pt – H. (F.1) Hence, the charged associated with H2 adsorption indicates the number of surface Pt atoms and hence, the real electrode surface area. It was established that the charge associated with a monolayer of H2 atoms formed on a polycrystalline Pt surface is 210 μCPt/cm2 [2]. The area under the cathodic part of the H2 adsorption region was estimated in two different ways: by using the integration feature available in CorrView, and by exporting the data to Excel, re-plotting it and dividing the area in the H2 adsorption region with squares. In the latter method, the time division, t (s), for one square was given by: v Vt = , (F.2) where ν is the scan rate (mV/s) and V is the voltage of one square (mV). The charge of one square, c (C) is given by: tIc ⋅= , (F.3) where I represents the current division of one square (A). The H2 adsorption charge, Qcv (C) is given by: SCV NcQ ⋅= , (F.4) where Ns is the total number of squares. Finally, the following relationship was used to determine the total metal area (TMA) (cm2Pt) and the electrochemical surface area (ECSA) (cm2Pt/gPt): 210 CVQTMA = (F.5) 298 cat CV L QECSA ⋅= 210 (F.6) where, QCV is the H2 adsorption charge (μC), and where Lcat is the catalyst electrode loading (g Pt/cm2). Figure F.2 shows a CV example for the determination of the ECSA of a 2.01 cm2 Pt/C catalyst sample. -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 IR Corrected Potential (V vs SHE) C ur re nt (A ) Figure F.2: Cyclic voltammogram (2.01 cm2, 2 mg/cm2 Pt/C, 5 mV/s, 0.5 M H2SO4, Pt flag CE, 23 ± 2oC). The charge associated with the formation of a monolayer of H2 was first estimated using the CorrView software by integrating the cathodic part of the H2 adsorption region. The total charge Qcv was estimated to be 0.764 C. This resulted in a TMA of 3638 cm2 and an ECSA of 1819 cm2/gPt. The graphical evaluation of the area under the curve was also carried out by importing the data in Excel and dividing the cathodic region of the H2 adsorption region in squares of 0.01 A and 0.002 V. At the scan rate of 0.5 mV/s, each square represented 2 s and 0.004 C. It was estimated that Ns was 184, which results in a charge Qcv of 0.736 C. This resulted in a TMA of 3503 cm2 and an ECSA of 1751 cm2/gPt. There was only an error of about 4% between the CorrView and Excel ECSA estimation methods. For each catalyst material evaluated (Pt/C and Pt black), an average ECSA was obtained using different catalyst GDE samples of the same material. 299 The same technique was used to evaluate electrode degradation. Different catalyst samples were compared for surface area: a fresh unused catalyst sample, a used catalyst sample, which was previously employed to conduct galvanostatic measurements in CH3OH acidic solutions, and an oxidized catalyst sample, which carbon was oxidized at low voltages while conducting galvanostatic measurements in the same solutions, at high current densities. Figure F.3 shows the cyclic voltammograms obtained for the various catalyst samples. Figure F.3: Cyclic voltammograms for a fresh catalyst sample, a used catalyst sample, and an oxidized catalyst sample (2.01 cm2, 2 mg/cm2 Pt/C, 5 mV/s, 0.5 M H2SO4, Pt flag CE, 23 ± 2oC). As indicated on the cyclic voltammograms, the H2 adsorption region for the fresh GDE was the greatest and decreased in size as the catalyst utilisation increased. Conversely, the Pt oxide peak height increased as the catalyst usage increased. It was at its highest value when the carbon in the catalyst was oxidized. Similarly, the double layer region became less defined and the H2 adsorption region became smaller as catalyst usage increased. The coulombic charge, QCV, was calculated by estimating the area under the curve in the region illustrated on the scans. It was used to calculate the active Pt surface of the electrode. The estimated area under the curve determined in Excel was within 5% of the one determined in CorrView in all cases. Table F.1 compares the ECA results obtained for various Pt/C catalyst samples. -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0 0.25 0.5 0.75 1 1.25 IR Corrected Potential (V vs. SHE) C ur re nt (A /c m 2 ) Fresh Used Oxidized 300 Table F.1: Total metal area results comparison for electrode degradation evaluation. Sample Description TMA (cm2 Pt) Fresh 3503 Used 2392 Oxidized 1433 The used GDE sample lost 40% of its TMA when compared to the fresh sample. This loss in Pt from the catalyst surface might be due to its prolonged exposure to CH3OH acidic solution. The oxidized catalyst sample lost about 60% of its TMA. When the carbon paper of the GDE was oxidized, the carbon leached away with some of the Pt on its surface. This shows the importance of the effect of catalyst degradation on the TMA and ECSA and emphasizes the necessity to use fresh catalyst samples to ensure good reproducibility of electrochemical experiments which will be carried out in this study. F.1.2 Pt-Ru/C catalyst electrochemical surface area determination The H2 adsorption and stripping technique described above cannot be used to determine the ECA of the bi-metallic Pt-Ru/C electrocatalyst as the H2 and Ru oxidation currents overlap. Hence, a CV technique based on the underpotential deposition (UPD) of copper was used to determine the ECSA of the Pt-Ru/C electrocatalyst [3, 4]. The underpotential deposition of a monolayer of metal atoms on an electrode surface is carried out at a potential more positive than that required for bulk deposition [4]. The potential is first pre-cycled between -0.68 V and 0.27 V vs. MSE by immersing the clean 4 mg/cm2 Pt-Ru/C 2 cm2 circular catalyst coupon in a 0.5 M H2SO4 solution at a scan rate of 5 mV/s to obtain a background scan. The electrode surface oxides are reduced at a potential of -0.48 V vs. MSE for 300 s in the same electrolytic solution. Next, the three electrode cell is emptied and a solution composed of 2 mM CuSO4 in 0.5 M H2SO4 was transferred to it. A monolayer of Cu is then deposited on the catalyst coupon surface by a UPD adsorption transient at -0.38 V vs. MSE for 300 s after which a linear oxidative 301 sweep from the hold potential of -0.38 V to 0.27 V vs. MSE at a scan rate of 5 mV/s was applied to strip the Cu from the catalyst surface. The Cu deposited is hence oxidatively removed from the electrode surface in the scan. Other potential scans after this one will not detect any Cu on the surface. All of these steps were carried out at ambient conditions. It is assumed that one Cu atom absorbs to each surface metal atom as per: CuUPD → Cu2+ + 2e- (F.7) The total charged measured, Qexp (C), is given by: Qexp = QUPD - Qb, (F.8) where is QUPD is the copper UPD layer stripping charge (μC), and Qb is the charged obtained from the background scan in the absence of Cu (μC) [201]. Two peaks should appear on the Cu stripping voltammogram: the first one, at about -0.3 V vs. MSE represents the Cu stripping mostly from Ru and the second one, at about -0.08 V vs. MSE, represents the Cu stripping from Pt only. The Ru oxide charge gained (ΔQD) and the H2 adsorption charge lost (ΔQH) of the catalyst working electrode can be obtained by comparing the background scan to the copper stripping voltammograms. It was established that the charge associated with a monolayer of Cu atoms formed on a polycrystalline Pt surface is 420 μC/cm2 [3]. Hence, the ECSA is given by: catL Q ECSA ⋅= 420 exp (F.9) This test is non-destructive as the electrode surface area can be fully recovered by cycling between - 0.68 and 0.37 V vs. MSE [5]. 302 Figure F.4 shows a CV example for the determination of the ECSA of a 2.01 cm2 unsupported Pt-Ru catalyst sample. Figure F.4: Cyclic voltammogram (4 mg/cm2 Pt-Ru, 5 mV/s, 0.5 M H2SO4, Pt flag CE, 23 ± 2oC). For the Cu UPD method, the charge associated with the formation of a monolayer of Cu could only be estimated in Excel. The graphical evaluation of the area under the curve was also carried out by importing the data in Excel and dividing the graph in squares of 1x10-4 A and 2x10-2 V. At the scan rate of 0.5 mV/s, each square represented 2 s and 0.0002 C. It was estimated that NUPD – Nb was 143, which results in a charge in the Cu desorption region Qexp of 0.029 C. This results in a TMA of 68 cm2 Pt-Ru and an ECSA of 17 cm2/gPt-Ru. For each catalyst material evaluated (Pt-Ru/C and Pt-Ru black), an average ECSA was obtained using different catalyst GDE samples of the same material. -3.00E-03 -2.50E-03 -2.00E-03 -1.50E-03 -1.00E-03 -5.00E-04 0.00E+00 5.00E-04 1.00E-03 1.50E-03 2.00E-03 2.50E-03 3.00E-03 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 IR Corrected Potential (V vs SHE) C ur re nt (A ) Stripping Scan Background Scan 303 F.2 Electrochemical impedance spectroscopy AC impedance or electrochemical impedance spectroscopy (EIS) is a powerful technique for characterizing and evaluating fuel cells [6]. Impedance is the resistance or resistivity in the alternating current (AC) version of Ohm’s law and measures the inability of a material to carry an electrical current. A small sinusoidal voltage or current is applied in addition to a conventional DC voltage or current. The AC component of the resulting current or voltage is measured and used to calculate, by Fourier analysis, the resulting impedance, which is frequency-dependent. Hence, in this analog experiment, a wide spectrum of time responses is examined at one potential value. When plotted as a function of frequency, a semi-circle is obtained. In this plot, called the Nyquist plot, the x axis is the real part (Z’) and the y-axis is the imaginary part (Z”) of the impedance. The absolute value of the impedance is expressed as a complex number represented by Z (ohm), which is given by: Z = (( Z’)2 +(Z“)2)1/2 (F.10) At low frequencies, the impedance is dominated by the resistor and the diameter of the semi circle is equal to the ohmic resistance, RS (ohm). Hence, RS is the real part of the impedance when the imaginary impedance is zero. Half-cell EIS measurements will be employed to determine RS of the electrolyte, solution or MEAs utilized in the cells after various electrochemical measurements are carried out. Typically, in order to obtain a clear half circle intercept, the initial frequency scan must be in the range of 50000 Hz. The ohmic overvoltage, ηohm (V), given by: ηohm=I.Rs (F.11) where I is the current (A), will be used to correct the voltage measurements for ohmic losses in order to pursue further data treatment to obtain kinetic information on the system studied. In this thesis, the MEA and solution resistance measurements were obtained from AC impedance measurements performed at different final frequencies (1x104 to 1x107 Hz) and at different initial 304 frequencies (0.001, 0.1 and 100 Hz) and at an AC amplitude of 10 mV, depending on the set-up for which the resistance was measured. The MEA or solution resistance was found by reading the real axis value at the high frequency intercept, near the origin of a Nyquist plot, a typical example of which is shown in Fig. F.5. 0 1 2 3 4 5 -2.0 -1.5 -1.0 -0.5 0 Z' Z' ' Figure F.5: Typical Nyquist plots for a fresh MEA at varying methanol concentration (initial frequency of 1x105 Hz to final frequency of 0.01 Hz, 4 mg/cm2 Pt-Ru/C anode, 2 mg/cm2 Pt/C cathode, 0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte). The intercept was similar for all MEAs except when in the equimolar CH3OH:H2O solutions. The average resistance obtained for fresh, used and oxidized MEAs are shown in Fig. F.6 for different CH3OH concentration. Each point represents the average of three measurements. Figure F.6: MEA resistance as a function of methanol concentration for various MEA conditions (4 mg/cm2 Pt/Ru anode, 2 mg/cm2 Pt cathode, N-117, 0.5 M H2SO4 electrolyte and catholyte). 16 M 0, 1, 2, and 6 M 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Methanol Concentration (M) M EA R es is ta nc e (o hm ) Fresh MEA Used MEA Oxidized MEA 305 As can be seen, the MEA resistance increased slightly with the CH3OH concentration and MEA usage. Also, the polymeric membrane material became less flexible after usage in high CH3OH concentration. As there is less water in the solution as the CH3OH concentration increases, it is possible that the membrane lost some of its conductivity due to partial dehydration. For an oxidized sample in 16 M CH3OH at the highest current density studied, the ohmic drop will be of: Eohm = 0.3 ohm (150x10-3 A/cm2) (2.02 cm2) = 0.09 mV (F.12) This is slightly higher than what can be estimated for a N117 membrane, having an area resistance of 0.26 ohm.cm2, at 20oC and at the same current density: Eohm = 0.26 ohm.cm2 (150x10-3 A/cm2) = 0.04 mV (F.13) The values of the measured potentials were corrected for the IR drop before kinetic parameter information was extracted. For example, an average value of 0.2 ohm was used to correct for the ohmic losses of an MEA having a 4 mg/cm2 Pt-Ru/C anode in 0 to 6 M CH3OH in 0.5 M H2SO4. An average value of 0.25 ohm was used to correct for the ohmic losses for the same MEA but in 16 M CH3OH in 0.5 M H2SO4, while an average value of 0.07 ohm was used to correct for the ohmic losses of MEA’s having a 2 mg/cm2 anode in 0 and 2 M CH3OH in 0.5 M H2SO4. F.3 Determination of the open circuit voltage An H-cell, as shown on Fig. F.7, was used to conduct open circuit voltage (OCV) measurements to determine the voltage under no-load conditions. To accomplish this, the electrical potential of an electrode is measured against a reference electrode when no current flows through the electrode. 306 Figure F.7: H-cell set-up for OCV measurements. A reference electrode, either MSE or SSE, was placed one of the H-cell compartment, where N2 was bubbled. A Pt flag electrode was platinized and placed in the other H-cell compartment, where H2 was bubbled, to act as an isothermal standard H2 reference electrode (SHE). The glass frit plugs were pre- wetted in 0.5 M H2SO4. The H-cell was first filled with 0.5 M H2SO4. The glass frit plugs were then put in place. The extra 0.5 M H2SO4 solution above the glass frit in the WE compartment was removed. The compartment was rinsed with a 1 M CH3OH solution in 0.5 M H2SO4 and refilled with the same solution. All solutions were deoxygenated by bubbling N2 for 20 minutes and then degassed in a sonicator for 20 minutes prior to the measurements, as residual O2 could cause errors. The potential of the WE was recorded against the potential of SHE. For measuring the OCV at higher temperatures, a controlled temperature water bath will be used to circulate water at the appropriate temperature in the double wall of the H-cell. F.4 Current - potential transients The determination of the current-potential relationship under a certain set of fixed conditions, such as concentration and temperature, is a common electrochemical measuring technique [7]. It is used for electrochemical analysis or the determination of the kinetics and mechanism of electrode reactions. Multiple measurements, based on the control of the current flowing through the system, are made to obtain polarization curves, i.e. current vs. potential plots, over a range of changing conditions. These experiments can be carried out galvanostatically, by controlling the current and measuring the resulting potential, or potentiostatically, by controlling the potential and measuring the resulting current [7]. 1 M CH3OH, 0.5 M H2SO4 0.5 M H2SO4 0.5 M H2SO4 H2 bubbler WE SHE Glass frit Platinized Pt electrode Hg/Hg2SO4 or Ag/AgCl N2 bubbler 307 In a galvanostatic experiment, the potentiostat controls the current flowing between the working and counter electrode and externally controls the reaction rate. The potential of the working electrode is allowed to assume a value appropriate for the rate and is measured against a reference electrode at a constant current as a function of time. The potential measured must be corrected for the resistance, which can be determined by EIS. To conduct these experiments, two lines are required on the multistat, one for the galvanostatic loop between the anode and the cathode and one for an OCV loop between the reference electrode and the cathode. The first loop measures the cell voltage while the other measures the cathode potential vs. the reference electrode. The working electrodes of both loops are connected together. The reference electrode loop was always started prior to the galvanostatic loop so that interactions between the two lines are minimized. The OCV was also checked with an independent multimeter and the error was not significant. In the alcohol electrolysis or electrochemical reforming mode, the current imposed on the galvanostatic loop will be negative and the resulting overall cell voltage measured will also be negative. It is the minimum cell voltage needed to be applied to drive the reaction at a given current. The cathode potential, which forces the electrons into reduction, is negative while the anode potential, which extracts the electrons from oxidation, is positive. Polarization curves obtained under various conditions will be corrected for the ohmic resistance and used to extract kinetic information on the electrochemical system being investigated. In the case of pure kinetic control, the reaction is only limited by the rate of electron transfer between the electrode and active species. It is assuming the concentration overpotential due to diffusion, i.e. mass transfer limitations, is negligible. Hence, the overpotential is limited by charge transfer and is the surface activation or charge transfer overvoltage, ηs. It is the difference between the actual electrode potential and the equilibrium electrode potential and is the electrochemical reaction’s driving force (from Eq. 1.13). es EE −=η (F.14) 308 According to the Butler-Erdey-Gruz-Volmer (BEV) equation, the current is exponentially related to the overpotential: ( ) ⎥⎦ ⎤⎢⎣ ⎡ ⎟⎠ ⎞⎜⎝ ⎛−⎟⎠ ⎞⎜⎝ ⎛ −= sso RT nF RT nFii ηβηβ exp1exp , (F.15) where β is the symmetry coefficient, and io is the exchange current density. The first term of the equation describes the anodic contribution and the second term describes the cathodic contribution. The exchange current density is the absolute value of oxidizing and reducing current density at null potential or equilibrium electrode potential. Its value usually depends on the electrode electrocatalytic properties, temperature and concentration as well as on the reaction mechanism and is related to the exchange current, Io, as follows: AiI oo ⋅= Geom (F.16) where AGeom is the exposed catalyst geometric surface area. For large activation overpotentials, i.e. ηs > 50 mV, the anodic reaction rate will be very high compared to that of the cathodic reaction rate, resulting in the high field approximation of the BEV equation: ⎟⎠ ⎞⎜⎝ ⎛= aao RT Fii ηαexp (F.17) The Tafel equation is: aibs +⋅= logη (F.18) where in this case, |ηs| = ηa and |i| = ia. The Tafel parameters are defined as: 309 nF RTb aα 303.2= , (F.19) oiba log−= , (F.20) where αa is the anodic transfer coefficient which is a measure of the symmetry of the energy barrier, b is the Tafel slope, and a is the intercept. By plotting the ηa vs. logi, the slope and the intercept can be determined. The experimental extrapolation of the Tafel line to obtain io is often challenging and as the result can vary widely on the region of the data used for the straight line fit. F.5 Rotating disk electrode The rotating disk electrode (RDE) is a specialized hydrodynamic electrode which creates controllable flow conditions and a nearly uniform density distribution along the electrode, making it a powerful tool to investigate electrochemical reactions. It is used in the investigation of the kinetics of electro-oxidation and electro-reduction reactions, in the mechanistic analysis of electrode reactions and in electro- analytical applications to ensure a known and controllable flow of solution over the electrode. The RDE creates additional convection which enhances the rate of mass transport to the electrode and increases the current and sensitivity compared to voltammetric measurements in stagnant solution. Flow control is achieved by bringing fresh reactant to the surface by a flat PTFE disc electrode which is rotated in the solution. The working electrode, Pt for example, is embedded in the top face of the PTFE shield. This permits a controlled variation in the electrochemical reaction rate by changing the convection rate in the solution. As the rotation speed is increased, the distance that a species must diffuse from the surface before it is removed by convection decreases. Hence, under the effect of rotation in the presence of a highly-concentrated electrolyte, the mass transport is dominated by diffusion and convection. A mass transport-limited current will arise when the system reaches a steady-state, once the equilibrium at the surface is driven to the product side. 310 A typical measurement used with the RDE is linear sweep voltammetry. The rotating disk electrode rotation speed is controlled so that there is a well-defined uniform transport of the solution phase reactant to the electrode surface resulting in a laminar flow of solution in the cell. This is the case when the dimensionless Reynolds number, Re, given by: ν ω 2Re r= (F.21) where ω is the angular velocity (1/s), r is the disk radius (cm) and ν is the kinematic fluid viscosity (cm2/s), is less than 1x105 [7]. The angular velocity can also be expressed in revolutions per minute (rpm): 60 )(2 rpmπω = (F.22) When the flow is laminar in the cell, the fluid dynamics can be described by the Navier-Stokes equation, which represents the conservation of momentum in the system. This equation relates the mass transport limited current and the rotation speed for a reversible electron transfer reaction. It was solved by Levich for the RDE geometry. The properties of the limiting current, iL, at the RDE can be described by the Levich equation: o L CnFDi 2 1 6 1 3 2 62.0 ων −= , or (F.23) 2 1ωBiL = (F.24) where Co is the bulk concentration (M), D is the diffusion coefficient (cm2/s) and B is the Levich parameter of Levich slope constant [8]. Typically, a set of current voltage curves are recorded for a reversible electron transfer reaction at different rotation speeds. The total current flowing depends on the rotation speed and as the rotation speed increases, the distance that the reactant can diffuse from the surface before being removed by 311 convection decreases. From polarization curves, iL can be obtained for various rotation rates. By plotting iL vs. ω1/2 for a set of experimental data, a straight line will be observed if the reaction is reversible. From the slope, B, the diffusion coefficient can be obtained. The mass transfer limiting current density, iL, represents the condition when the surface concentration of the reactant is zero while the electrode activation kinetic current density, ik, is the theoretical current that would be expected to flow in the absence of mass transfer limitations. The total current density, i, in the mixed control region, where activation and mass control are present, is a function ik and iL: lk iii 111 += . (F.25) The Koutecky-Levich equation shows that as ω goes to infinity, the current approaches ik: 2 1 111 ωBii k += . (F.26) A Koutecky-Levich plot can be obtained by choosing values of the overpotentials in the mixed activation and mass-transport control region. At every overpotential, the current density is found at each rotation rates. If each overpotential point is extrapolated to the y axis, the y intercepts represent the kinetic currents at every overpotential. Each line has the same slope of 1/B. The plot of 1/i vs. 1/ω1/2 should be a straight line intercept of 1/ik [7]. F.6 References [1] J. M. D. Rodriguez, J. A. H. Melian and J. P. Pena, “Determination of the Real Surface Area of Pt Electrodes by Hydrogen Adsorption Using Cyclic Voltammetry”, Journal of Chemical Education, 77(9), (2000) 1195-1197. [2] T. R. Ralph, G. A. Hards and J. E. Keating, “Low Cost Electrodes for Proton Exchange Membrane Fuel Cells”, Journal of the Electrochemical Society, 144(11), (1997) 3845-3857. 312 [3] C. L. Green and A. Kucernak, “Determination of the Platinum and Ruthenium Surface Areas in Platinum-Ruthenium Alloy Electrocatalysts by Underpotential Deposition of Copper. 1. Unsupported Catalysts”, Journal of Physical Chemistry B, (106), (2002) 1036-1047. [4] C. L. Green and A. Kucernak, “Determination of the Platinum and Ruthenium Surface Areas in Platinum-Ruthenium Electrocatalysts by Underpotential Deposition of Copper. 2. Effect of Surface Composition on Activity”, Journal of Physical Chemistry B, (106), (2002) 11446-11456. [5] S.-R. Wang and P. S. Fedkiw, “Pulsed-potential Oxidation of Methanol, 1. Smooth Platinum Electrode With and Without Tin Surface Modification”, J. Electrochemical Society, 139(9), (1992) 2519-2525. [6] A. J. Bard and L. R. Faulkner, “Electrochemical Methods Fundamentals and Applications”, Second Edition, John Wiley & Sons, Inc., USA (2001) 833 pp. [7] E. Gileadi, “Electrode Kinetics for Chemists, Chemical Engineers, and Materials Scientists”, VCH Publishers Inc., USA (1993) 597 pp. [8] V. A. Paganin, E. Sitta, T. Iwasita and W. Vielstich, “Methanol Crossover Effect on the Cathode Potential of a Direct PEM Fuel Cell“, Journal of Applied Electrochemistry, 35, (2005) 1239-1243. 313 Appendix G: Thermodynamic data and sample calculations G.1 Thermodynamic data Table G.1: Thermodynamic data. Species State ΔGof ΔHof ΔSof Cpo [kJ/mol] [kJ/mol] [J/molK] [J/molK] CH3OH g -162.30 -201.00 39.90 44.10 lq -166.60 -239.10 126.80 81.20 H2O g -228.61 -241.83 188.80 33.60 lq -237.14 -285.83 70.00 75.35 H2 g 0.00 0.00 130.70 28.84 CO2 g -394.40 -393.51 213.80 37.10 O2 g 0.00 0.00 205.20 29.40 References [1, 2] G.2 Thermodynamic sample calculations G.2.1 Liquid phase The electrochemical reactions for the direct methanol reformer (DMR) in the liquid phase written in the reduction sense are as follows: Anode: 6H+ (aq)+ 6e- + CO2 (g) → CH3OH (l) + H2O (l) Cathode: 6H+ (aq) + 6e- → 3H2 (g) Overall: CH3OH (l) + H2O (l) → CO2 (g) + 3H2 (g) The Gibbs free energy associated with the electrochemical reaction at standard condition is: 314 ( )reactantsproductsGsG o fiiio −Δ∑=Δ , (G.1) Anode: ΔGoa = ((-166.6) + (-237.14) - (-394.4)) · 1000 = -9340 J/mol Cathode: ΔGoc = ((3·0) - 0) · 1000 = 0 J/mol Overall: ΔGo = ΔGoc - ΔGoa = 0 - (-9340) = 9340 J/mol Similarly, the enthalpy of formation associated with the electrochemical reaction at standard condition is: ( )reactantsproductsHsH o fiiio −Δ∑=Δ , Anode: ΔHoa = ((-239.1) + (-285.83) - (-393.51)) · 1000 = -131420 J/mol Cathode: ΔHoc = ((3·0) - 0) · 1000 = 0 J/mol Overall: ΔHo = ΔHoc - ΔHoa = 0 - (-131420) = 131420 J/mol The standard cell voltage or reversible cell voltage at standard conditions is: nF GE o o Δ−= (G.2) Anode: Eoa = - (-9340) / 6 · 96485 = 0.016 V Cathode: Eoc = - (0) / 6 · 96485 = 0 V Overall: Eo = Eoc - Eoa = 0 - 0.016 = -0.016 V The maximum thermodynamic efficiency at standard condition is determined using the higher heating value (HHV) for the enthalpy of formation of liquid CH3OH at 25oC: 100max ⋅⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ Δ Δ= o o H Gη % (G.3) 315 =maxη (9340/131420) · 100 = 7.11% The thermodynamic efficiency is positive as work is required to generate H2 under these conditions since the overall reaction is endothermic (∆Ho > 0), and non-spontaneous (∆Go > 0). G.2.2 Gas phase The electrochemical reactions for the direct methanol reformer (DMR) in the gas phase written in the reduction sense are as follows: Anode: 6H+ (aq)+ 6e- + CO2 (g) → CH3OH (g) + H2O (g) Cathode: 6H+ (aq) + 6e- → 3H2 (g) Overall: CH3OH (g) + H2O (g) → CO2 (g) + 3H2 (g) Standard Gibbs free energy of formation: ( )reactantsproductsGsG o fiiio −Δ∑=Δ , Anode: ΔGoa = ((-162.3) + (-228.61) - (-394.4)) · 1000 = 3490 J/mol Cathode: ΔGoc = ((3·0) – 0) · 1000 = 0 J/mol Overall: ΔGo = ΔGoc - ΔGoa = 0 - (3490) = -3490 J/mol Standard cell theoretical cell voltage or reversible cell voltage: nF GE o o Δ−= Anode: Eoa = - (3490) / 6 · 96485 = -0.006 V Cathode: Eoc = - (0) / 6 · 96485 = 0 V 316 Overall: Eo = Eoc - Eoa = 0 - (-0.006) = 0.006 V Standard enthalpy of formation: o fiii o HsH ,Δ∑=Δ (G.4) Anode: ΔHoa = ((-201) + (-241.83) - (-393.51)) · 1000 = -49316 J/mol Cathode: ΔHoc = ((3 · 0) - 0) · 1000 = 0 J/mol Overall: ΔHo = ΔHoc - ΔHoa = 0 - (-49316) = 49316 J/mol Standard entropy of reaction: o fiii o SsS ,Δ∑=Δ (G.5) Anode: ΔSoa = ((39.9) + (188.8) - (-213.8)) · 1000 = 15 J/mol.K Cathode: ΔSoc = ((3 · 130.7) - 0) · 1000 = 392 J/mol.K Overall: ΔSo = ΔSoc - ΔSoa = 392 - (15) = 377 J/mol.K Temperature correction: Assuming constant pressure, the molar enthalphy and entropy at T are given by: dTCHH p To T 298∫+Δ=Δ p o T CTHH )298( −+Δ=Δ (G.6) dTC T SS p To T 1 298∫+Δ=Δ p o T CTSS )298ln()ln( −+Δ=Δ (G.7) 317 The Shomate equations and empirical coefficients for the heat capacity for the relevant species are available in the literature [3-6] and are summarized in Table G.2. Table G.2: Empirical coefficients for Shomate equations. Species State A B C D E T Range Eq. Ref [oK] Form CH3OH g 39250 87900 1917 53650 897 200- 1500 1 [4] lq 86.2 323 [6] H2O g 30.092 6.832514 6.793435 -2.53448 0.082139 500- 1700 2 [3] g 4.0700 -0.6160/103 1.2810/106 -0.508/109 0.0769/1012 300- 1000 3 [5] lq -203.606 1523.29 -3196.413 2474.455 3.855326 298- 500 2 [3] H2 g 33.06618 -11.363417 11.432816 -2.772874 -0.158558 298- 1000 2 [3] g 3.0570 1.4870/103 -1.793/106 0.9470/109 -0.1726/1012 300- 1000 3 [5] CO2 g 24.99735 55.18696 -33.69137 7.948387 -0.136638 298- 1200 2 [3] g 2.4010 4.8530/103 -2.039/106 0.3430/109 0.0000 300- 1000 3 [5] O2 g 3.6260 -1.0430/103 2.1780/106 -1.160/109 0.2053/1012 300- 1000 3 [5] Shomate equation form 1 [4]: Cp/K = A + B*((C/T)/sinh(C/T))2 + D*((E/T)/cosh(E/T)2 Shomate equation form 2 [3]: Cp/K =(A + B*T + C*T2 + D*T3 + E*T4)*8.314 Shomate equation form 3 [5]: Cp/R = A + BT + CT2+ DT3 + ET12 As noted in Section 2.3.1, the heat capacity in a specific phase did not vary significantly over the temperature range studied (> 8 %). However, there is a more significant change in the heat capacity when the phase changes from liquid to gas. 318 Using the heat capacities determined as a function of temperature, it is possible to evaluate the enthalpy of formation and the entropy of formation at temperature T using Eq. G.6 and G. 7, respectively. For example, for methanol at 100oC in the gas phase: ∆H373 = (-201*1000) + ((100+273.15)-298) · 49.5315 = -197285.13 J/mol ∆S373= 39.9 + (ln((100+273.15)-ln(298)) · 49.5315 = 51.02 J/mol It was determine that using the average heat capacity over the temperature range from 298 K to T did not have a significant impact (> 2 %) in the resulting enthalpy of formation and entropy of formation values. Like in the case of standard conditions, it is possible to determine the overall electrolysis enthalpy and entropy of reaction using the temperature corrected values. An example calculation is given for the CH3OH electrolyser at 100oC. Enthalpy of formation: fiii HsH ,,373373 Δ∑=Δ (G.8) Anode: ΔHa, 373 = ((-19785.13) + (-2397.1) - (-390489.54)) = -46067.7 J/mol Cathode: ΔHc, 373 = ((3 · 181.59) - 0) = 6544.78 J/mol Overall: ΔH373 = ΔHc, 373 - ΔHa, 373 = 6544.78 - (-46067.7) = 5612.48 J/mol Entropy of reaction: fiii SsS ,,373373 Δ∑=Δ (G.9) Anode: ΔSa, 373 = ((51.02) + (196.44) - (222.84)) · 1000 = 24.62 J/mol.K Cathode: ΔSc, 373 = ((3 · 137.23) - 0) = 411.69 J/mol.K Overall: ΔS373 = ΔSc, 373 - ΔSa, 373 = 411.69 - (24.62) = 387.06 J/mol.K 319 The effect of temperature on the reversible electrode potential comes from the Gibbs-Helmholtz equation: d∆G/dT = -∆S nFdE/dT = ∆S nFdEo/dT = ∆So Thus, assuming a small effect of temperature on S over the temperature range studied, ⎟⎠ ⎞⎜⎝ ⎛ Δ⋅−+= nF STEE Toe )298( (G.10) Typical values of ∆Eo/∆T are of the order of ± 1e-3 V/K. For the overall methanol electrolysis reaction, at 100oC and 1 atm: Overall: Ee, 373 = (0.006) + (373-298) · ((387.06)/(6 · 96485)) = 0.0562 V From this, the Gibbs free energy associated with the electrochemical reaction at temperature T is: TeT nFEG ,−=Δ (G.11) Overall: ∆G373 = -6 (96485) 0.0562 = -32520.01J/mol Maximum thermodynamic efficiency: It was shown that the ΔH and ΔG values were corrected for the temperature using the heat capacity for the species involved. For the methanol electrolyser at 100oC and 1 atm: 320 100max ⋅⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ Δ Δ= T T H Gη % (G.12) =373max,η -32520.01/5612.48 · 100 = -61.81 % Note that the thermodynamic efficiency for the electrolyser is negative as, under these conditions, work is produced by the system. For a PEMFC operating at similar conditions, (100oC, 1 atm), the thermodynamic efficiency would be: =373max,η -225104.25/-242575.51· 100 = 92.80% G.2.3 System efficiency calculation ( ) 100...3max,2max,1max,max, ⋅⋅⋅= ηηηη system (G.13) For a PEMFC in the liquid phase at 100oC and 1 atm combined with a DMR combination in the gas phase at 100oC and 1 atm: PEMFC thermodynamic efficiency = =1max,η 92.80 % Absolute DMR thermodynamic efficiency = 61.81 % System overall efficiency: =systemmax,η (0.9280 · 0.6181) · 100 = 57.36 % 321 G.2.4 Effect of pressure in the gas phase From the Maxwell relation and based on the ideal gas law, the volume change of gaseous components is related to the change in the number of moles of gaseous components based on the reaction stoichiometry: ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ Δ=Δ=∂ Δ∂ tot gmol g P RTN V P G , (G.14) Assuming that the total pressure of the gaseous species on the cathode and the anode side is PH2 + PCO2 + PCH3OH + PH2O = P1 =1 atm and that the total pressure is increased to P2 = 2.5 atm: ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛⋅⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ ⋅Δ−= 1 2, ln P P nF RTN EE gmoloe (G.15) ΔNmol,g anode = 1+1-1= 1 ΔNmol,g cathode = 3-0 = 3 ΔNmol,g overall = 3+1-1-1 = 2 Anode: E a = (-0.0596) – ((1 · 8.314 · 473)/(6 · 96485)) · ln(2.5/1) = -0.07 V Cathode: Ec = 0 – ((3 · 8.314 · 473)/(6 · 96485)) · ln(2.5/1) = -0.02 V Overall: Ee = Ec – Ea = (-0.02) – (-0.07) = 0.05 V 322 G.2.5 Effect of anode and cathode pressure in the liquid phase Anode: 6H+ (aq)+ 6e- + CO2 (g) → CH3OH (l) + H2O (l) Cathode: 6H+ (aq) + 6e- → 3H2 (g) Overall: CH3OH (l) + H2O (l) → CO2 (g) + 3H2 (g) Applying Nernst equation: Anode: ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ ⋅ ⋅−= + 2 6 23 ][ ][][ ln CO o a PH OHOHCH nF RTEE Cathode: ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛−= + 6 3 ][ ln 2 H P nF RTEE Hoc Overall: ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ ⋅ ⋅−⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛⋅−=−= ++ 2 2 6 23 6 3 ][ ][][ ln ][ ln CO Ho ace PH OHOHCH nF RT H P nF RTEEEE Knowing that ln a – ln b = ln a/b, ⎟⎟⎠ ⎞ ⎜⎜⎝ ⎛ ⋅⋅−= ][ ln 3 3 22 OHCH PP nF RTEE COHoe Hence, as the CO2 pressure on the anode increases, Ee becomes more negative. Similarly, as the H2 pressure on the cathode increases, Ee becomes more negative. Therefore, increasing the anode or cathode compartment pressure is not beneficial to the cell voltage required for CH3OH electrolysis. 323 G.3 References [1] Dean, J. A., “Lange’s Handbook of Chemistry”, 15th Edition, McGraw Hill, New York, (1999) Section 8.2 and 8.7. [2] D. R. Lide, “CRC Handbook of Chemistry and Physics”, 87th Edition, CRC Press LLC, (2006-2007) 2608 pp. [3] M. W. Chase, “NIST-JANAF Thermodynamical Tables”, Fourth edition, J. Phys. Chem. Ref. Data, Monograph 9, (1998) 1-1951. [4] Perry, R.H.; and Green, D.W., “Perry's Chemical Engineers' Handbook”, 7th edition, McGraw-Hill, (1997) 2581 pp., Table 2-198. [5] M. J. Moran and H. N. Shapiro, “Fundamentals of engineering thermodynamics”, Third Edition, John Wiley & Sons, Inc., (1995), 859pp. [6] Hough, E.W.; Mason, D.M.; and Sage, B.H., “Heat capacities of several organic liquids”, J. Am. Chem. Soc., 72, (1950) 5775-5777. 324 Appendix H: Electrochemical Sample Calculations H.1 Power consumption The electric power consumption of the electrochemical reformer, Pe (W), is given by: GeomPTe AiEP ⋅⋅= , , (H.1) where E T,P is the operating cell voltage, i is the current density (A/cm2) and AGeom is the cell geometric area which is 2 cm2. Figure H.1: Power consumption as a function of cell voltage for different CH3OH concentrations (0, 1, 2, 6 or 16 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt- Ru anode, 2 mg/cm2 Pt cathode, 23 ±2oC). H.2 CO2 emissions and H2 production rate For the CH3OH electrochemical reformer, 1 mol of CO2 is emitted for every 3 moles of H2 produced. According to Faraday’s law, the rate of CO2 emissions, 2CO N (mol/cm2s), is: 0.00 0.01 0.02 0.03 0.04 0.05 0.06 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 Cell Voltage (V) Po w er C on su m pt io n (W ) 0 M CH3OH 1 M CH3OH 2 M CH3OH 6 M CH3OH 16 M CH3OH 325 F i F iNN HCO 1863 1 3 1 22 =⋅=⋅= . (H.2) This rate can be converted in the total mass rate of CO2 emissions, 2CO M (g/day), as per the following equation where t = 3600 s/h x 24 h = 86400 s: GeomcoCOCO AtMwNM ⋅⋅⋅= 222 (H.3) The rate of H2 evolution, 2H N (mol/m2s), is given by Faraday’s law at STP is: F iN H 62 = . (H.4) By combining this expression with the ideal gas law, RTANPV GeomHH ⋅⋅= 22 , (H.5) where A is the electrode area and 2H V is the volumetric H2 evolution rate (m3/h), we can obtain that: P RT F AiV GeomH ⋅⎟⎠ ⎞⎜⎝ ⎛ ⋅= 62 (H.6) 326 Figure H.2: Rate of CO2 emission and volumetric rate of H2 evolution rate in function of the cell Voltage (2 M CH3OH in 0.5 M H2SO4 electrolyser anolyte, 0.5 M H2SO4 electrolyser catholyte, 4 mg/cm2 Pt-Ru or 2 mg/cm2 Pt anode, 2 mg/cm2 Pt cathode, 23 ±2oC). H.3 Proton diffusion limited H2 production – back envelope calculation A PEMFC automotive system having a peak electric power, Pe, of 100 kW and a voltage of 0.6 V at peak power will have a current of I = Pe/E = 166.67 A or C/s. Generally, the diffusion coefficients of most ions in dilute aqueous solutions are similar and have values that at room temperature are in the range of 0.6x10-9 to 2x10-9 m2/s. Various values were found for the diffusion coefficient of protons. In a dilute aqueous solution, it is equal to 9.31x10-9 m2/s [1]. In a fully hydrated N-117 membrane, the proton diffusion coefficient was reported as 1.92x10-9 m2/s [2], in a fully hydrated Nafion® membrane, it was reported as 8.36x10-10 m2/s, and in N-117 at 80oC, it was reported as 4.50x10-9 m2/s [3]. These examples of proton diffusion coefficients show that there is a wide variability in the values reported. A DMR operating temperature of 80oC (or 353.15 K) was assumed. The membrane thickness of 161 mm (or 1.61e-4 m) for a hydrated N-115 membrane was used in the calculation [4]. The DMR voltage loss through MEA resistance. ηohm, was taken to be equal to 0.21 V, based on the experimental values 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 -1.3 -1.0 -0.8 -0.5 -0.3 0.0 Cell Voltage (V) R at e of C O 2 E m is si on (g /d ay ) 0.E+00 2.E-07 4.E-07 6.E-07 8.E-07 1.E-06 1.E-06 1.E-06 2.E-06 R at e of H 2 P ro du ct io n (c m 3 /h ) 4 mg/cm2 Pt-Ru, Rate of CO2 Emission (g/day) 2 mg/cm2 Pt, Rate of CO2 Emission (g/day) 2 mg/cm2 Pt, Rate of H2 Produciton (cm3/h) 4 mg/cm2 Pt-Ru, Rate of H2 Production (cm3/h) 327 obtained for a N-117, 4 mg/cm2 Pt-Ru/C anode, a 2 mg/cm2 Pt/C cathode, a 1 to 16 M CH3OH anode, and a 0.5 M H2SO4 electrolyte. The DMR proton concentration, CH+, was assumed to be equal to 1 M (1000 mol/m3) based on 0.5 M H2SO4. Assuming that 3H2 is equivalent to 6H+ and that the number of electrons transferred is equal to 6, the DMR proton rate, reH+, can be calculated from Faraday’s law for the PEMFC current determined above: reH+ = 6I / 6F = 166.67/96485 = 1.73x10-3 mol/s (H.7) The proton flux by migration, NH+, is equal to: NH+ = (F.CH+.ηohm.DH+)/(t.RT) (H.8) = (96485.1000.0.21. 9.31x10-9) / (1.61x10-4.8.314.353.15) = 3.99x10-4 mol/m2.s Therefore, the DMR active geometric surface area can be obtained: AGeom = reH+ / NH+ . (H.9) = 1.73x10-3 / 3.99x10-4 = 4.33 m2 Table H.1 summarizes the results obtained for various proton diffusion coefficients. As can be seen, the DMR active surface area greatly depends on the proton diffusion in the membrane. 328 Table H.1: Direct methanol reformer electrochemical active surface area as a function of proton diffusion coefficient in Nafion, based on H2 production for a 100 kW PEMFC at a voltage of 0.6 V. DH+ [m2/s] NH+ [mol/m2.s] AGeom [m2] 9.31x10-9 3.99x10-4 4.33 1.92x10-9 8.23x10-5 20.99 1.92x10-9 3.58x10-5 48.21 4.50x10-9 1.93x10-4 8.96 H.4 Overall system efficiency The calculation for the thermodynamic efficiency ηmax, and the system thermodynamic efficiency, ηmax, system, was demonstrated in Appendix G. A thermodynamic efficiency ηmax of 7.11 % was obtained for the liquid DMR system at 25oC. The voltage efficiency, ηE, is obtained by dividing the cell voltage by the thermodynamic reversible cell voltage. e E E E=η (H.10) For the DMR in the liquid phase at 25oC, the reversible cell voltage, Ee, is equal to -0.016 V. At a current density of 50 mA/cm2 and ambient conditions, the IR-corrected experimental liquid CH3OH electrolyser voltage was -1.22 V, when using N-117, 2 M CH3OH, a 4 mg/cm2 Pt/Ru anode, and a 2 mg/cm2 Pt cathode. Under these conditions, the voltage efficiency, ηE, can be calculated as: 25.76 016.0 22.1 =− −=Eη , (H.11) assuming that the fuel utilization coefficient is one, i.e. that the mass of fuel reacted in the electrolyser equals the mass of fuel input in the electrolyser. 329 The Faradaic current efficiency,ηI, of an electrolyser is the fraction of the current (or charge) passed through the cell that is used for converting starting materials to products. It is obtained by dividing the current, I, by the reversible current, Ir, which is composed of the current and the cross-over current. In terms of current densities, this is given by: c I ii i +=η (H.12) where ic is the CH3OH cross-over current density. The concentration of CH3OH crossing over to the cathode was not determined in the preliminary experiments. From this value, it would be possible to obtain ic using Faraday’s law. For a DMFC, a realistic value for the Faradaic current efficiency is 0.8 [5]. The reactants are usually supplied in an amount exceeding the reaction stoichiometry. The fuel stoichiometry, λ, refers to the ratio of reactant supplied to the electrolyser over reactant consumed by the electrolyser. However, at a given concentration of CH3OH, λ will vary with the current. The fuel utilization efficiency, ηfuel, would be given by: λη 1=fuel (H.13) Assuming a fuel efficiency of 2, the overall system efficiency, ηoverall, can be calculated by multiplying all the above efficiencies: fuelIEoverall ηηηηη ⋅⋅⋅= max (H.14) 57.128.025.76013.0 −=⋅⋅⋅−=overallη The operating overall system efficiency will vary widely based on the electrolyser operating conditions. 330 H.5 References [1] D. R. Lide, “CRC Handbook of Chemistry and Physics”, 87th Edition, CRC Press LLC, (2006-2007) 2608 pp. [2] S. J. Paddison, and S. Paul, "The Nature of Proton Transport in fully Hydrated Nafion", Phys. Chem. Chem. Phys., 4 (2002) p. 1158-1163. [3] S. J. Paddison, S. Paul, and T. A. Jr. Zawodzinski, "A Statistical Mechanical Model of Proton and Water Transport in a Proton Exchange Membrane", Journal of the Electrochemical Society, 147(2) (2000) p. 617-626. [4] S. Slade, S. A. Campbell, T. R. Ralph and F. C. Walsh, “Ionic Conductivity of an Extruded Nafion 1100 EW Series of Membranes”, Journal of the Electrochemical Society, 149(12) (2002) pp. A1556- A1564. [5] J. Larminie, and A. Dicks, “Fuel Cell Systems Explained”, John Wiley & Sons Ltd., UK, Second Edition, (2003), 406 pp.