UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Electrochemical oxidation of phenol for waste water treatment Sucre, Vivian Smith de 1979-03-06

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-UBC_1979_A7 S97.pdf [ 8.14MB ]
Metadata
JSON: 831-1.0058801.json
JSON-LD: 831-1.0058801-ld.json
RDF/XML (Pretty): 831-1.0058801-rdf.xml
RDF/JSON: 831-1.0058801-rdf.json
Turtle: 831-1.0058801-turtle.txt
N-Triples: 831-1.0058801-rdf-ntriples.txt
Original Record: 831-1.0058801-source.json
Full Text
831-1.0058801-fulltext.txt
Citation
831-1.0058801.ris

Full Text

ELECTROCHEMICAL OXIDATION OF PHENOL FOR WASTE WATER TREATMENT by VIVIAN SMITH de^SUCRE B.Sc. Universidad Simon Bolivar, 1975 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Chemical Engineering We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August, 1979 0 Vivian Smith de Sucre, 1979 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of ^UxCcAL &0&IM£IUA)G The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 DE-6 BP 75-51 1 E ABSTRACT The electrochemical oxidation of phenol for waste treatment appli cations was investigated on lead dioxide packed-bed anodes. The electro lytic cell was operated in both batch and continuous modes with feed streams up to 1100 mg/1 phenol dissolved in aqueous solutions of Na2S0i+ and H2SL\ or NaOH. Electrodeposited lead dioxide was found to be a better anode for phenol oxidation, than the lead dioxide obtained by anodizing lead shot. Results showed that all the phenol in solution could be readily oxidized but complete total organic carbon (T.O.C.) removal was more difficult. Rates of phenol oxidation were similar in divided and undivided cells. The oxidation of phenol was favoured by an acidic pH, but an alkaline pH improved the further oxidation of intermediate products. In divided cells, an anionic membrane, which allowed migration of hydroxyl ions, proved to be superior than a cationic membrane for T.O.C. removal. The percent of phenol oxidized increased with increasing current density, and decreased as initial phenol concentration, electrolyte flow rate, and anode particle size were increased. Comparisons of the experimental results with a mass transfer model are presented for the batch experiments, and a simplified model is pro posed to interpret the results from continuous experiments in terms of relative mass transfer and electrochemical reaction resistances. ii TABLE OF CONTENTS ABSTRACT ii LIST OF TABLES v LIST OF FIGURES vACKNOWLEDGMENTS viii Chapter 1 INTRODUCTION 1 1.1 Phenols as pollutants 1 1.2 Methods of treatment of phenolic wastes 2 2 BASES OF THE ELECTROCHEMICAL PROCESS 6 2.1 General concepts 6 2.2 Literature review on the electrochemical oxidation of phenol 12 2.2.1 Reaction products 12.2.2 Proposed reaction mechanisms 13 2.2.3 Electrode materials tested 8 2.2.4 Effect of current density 20 2.2.5 Effect of nature of the electrolyte .... 21 2.2.6 Effect of pH 24 2.3 The lead dioxide electrode 25 3 OBJECTIVES 29 4 EXPERIMENTAL APPARATUS AND METHODS 31 4.1 Apparatus 34.1.1 Cell design 31 4.1.2 Flow diagram of the apparatus 39 4.2 Experimental methods 43 4.2.1 Batch experiments4.2.2 Continuous experiments 45 4.3 Analytic techniques 46 4.3.1 Phenol analysis4.3.2 Total organic carbon analysis 47 4.3.3 Lead analysis 48 iii 5 RESULTS AND DISCUSSION 49 5.1 Electrode materials . 49 5.2 Effect of pH using the divided cell 55 5.3 Effect of current using the divided cell 64 5.4 Comparisons of membrane performances 65 5.5 Effect of pH using the undivided cell 8 5.6 Effect of current using the undivided cell .... 69 5.7 Comparisons of divided and undivided cells .... 75 5.8 Effect of conductivity of the electrolyte 76 5.9 Effect of initial phenol concentration 78 5.10 Effect of electrolyte flow rate 83 5.11 Effect of particle size 86 5.12 Comparisons of experimental results with mathematical models 8 5.12.1 Batch experiments 85.12.2 Continuous experiments 89 5.13 Current efficiencies, energy requirements and energy costs for phenol oxidation 95 5.13.1 Batch experiments 95.13.2 Continuous experiments 96 5.13.3 Cost comparisons6 CONCLUSIONS 99 7 RECOMMENDATIONS 101 NOMENCLATURE 104 BIBLIOGRAPHY 7 APPENDIX 1 Specification of auxiliary equipment and materials . . . Ill 2 Experimental data 116 3 Mathematical models 158 4 Calculations 171 5 Relevant physical data 18iv LIST OF TABLES Table 1 Rates of phenol oxidation on different electrode materials 19 2 Effect of current density and type of electrolyte on C.O.D. removal 21 3 Effect of type of electrolyte on phenol oxidation .... 23 4 Fundamental specifications of the electrolytic cell ... 37 5 Comparisons of divided and undivided cells 76 6 Typical current efficiencies, energy requirements and energy costs in batch experiments with undivided cell . 95 7 Typical current efficiencies, energy requirements and energy costs in continuous experiments with undivided cell 96 8 Operating costs of various treatment methods, estimated for 1974 for a catalytic cracker effluent containing 700 mg/1 phenol 7 Appendix 1 A-l Summary of typical properties of IONAC membranes .... 113 Appendix 2 Experimental data tables for: Run 1-1 to Run 1-9: Divided cell, batch experiments with anodized lead 119-12Run 2-1 to Run 2-11: Divided cell, batch experiments with electrodeposited Pb02 124-134 Run 3-1 to Run 3-15: Undivided cell, batch experiments with electrodeposited Pb02 135-149 Run 4-1 to Run 4-8: Undivided cell, continuous experiments with electrodeposited Pb02 150-157 Appendix 4 A-2 Theoretical phenol fractional conversion vs time for a mass transfer—controlled batch system 173 A-3 Calculation of experimental, mass transfer and reaction rate constants from experiments 4-1, 4-2, 4-3 175 A-4 Calculation of experimental, mass transfer and reaction rate constants for experiment 4-4 176 A-5 Calculation of experimental, mass transfer and reaction rate constants for experiment 4-8 177 A-6 pH of solutions of NaOH and H^O^ at 20°C 181 A-7 Conductivities of aqueous solutions of NaOH, H^O^ and N32S01+ at 20°C 18A-7 % phenol ionized vs pH 2 v LIST OF FIGURES Figure 1 Voltage components in a divided electrolytic cell .... 9 2 Reaction products 12 3 Half-wave potential vs pH, for the oxidation of 4 x IO-4 phenol 24 4 Side view of the general divided-cell arrangement .... 32 5 Front and side views of the anode chamber for the < anodized lead electrode 34 6 Front and side views of the anode chamber for the electrodeposited Pb02 electrode 35 7 Detail of the inlet or outlet connection adapted on the electrodeposited Pb02 on graphite anode .... 36 8 Detail of the mechanism used to hold the cell 38 9 Flow diagram of the apparatus 40 10 Effect of type of lead dioxide electrode at 10 A and initial pH - 9.4 with IONAC MC-3470 membrane 52 11 Scanning electron-micrographs of the electrodeposited Pb02 particles after use 53 12 Scanning electron-micrographs of the anodized lead particles, after use 4 13 Effect of current on pH and % T.O.C. oxidation with IONAC MC-3470 membrane 56 14 Effect of current on % phenol oxidation at initial pH = 9.4 with IONAC MC-3470 membrane 59 15 Effect of current on pH, % T.O.C. oxidation and % phenol oxidation with NAFION-127 membrane 60 16 Effect of pH on % T.O.C. and % phenol oxidation at 20 A with NAFION-127 membrane 62 17 Type of membrane-pH effect on % T.O.C. and % phenol oxidation at 20 A .................. 63 18 Current effect on % T.O.C. and % phenol oxidation at pH = 2.5 with IONAC MC-3470 66 19 Type of cationic membrane effect on % T.O.C. and % phenol oxidation at 20 A and pH = 2.5 67 20 pH effect on % T.O.C. and % phenol oxidation at 10 A in an undivided cell 70 21 pH effect on % T.O.C. and % phenol oxidation at 20 A in an undivided cell 1 22 Effect of pH on % T.O.C. and % phenol oxidation at 30 A in an undivided cell • . . 72 23 Current effect on % T.O.C. and % phenol oxidation at pH - 2.5 in. an undivided cell 73 24 Effect of current on % T.O.C. and % phenol oxidation at initial pH = 12, in an undivided cell 74 vi 25 Effect of electrolyte conductivity at 20 A (in alkaline and acid media) 77 26 Effect of electrolyte conductivity at 10 A and initial pH = 12 79 27 Effect of electrolyte conductivity at 10 A and initial pH - 2.5 80 28 % phenol oxidized vs time for various initial phenol concentrations at 10 A and pH = 2.5 81 29 Phenol concentration effect on phenol concentration vs time at 10 A and 2.5 pH 82 30 Effect of flow rate on the single pass % phenol oxidation at (a) 10 A, (b) 20 A 85 31 Effect of anode surface area-particle size on the % phenol oxidized in a single-pass vs flow rate ... 87 32 -Jln(l - X) vs (u)--*- for the calculation of experimental rate constants in single pass experiments 91 Appendix 3 A-l Packed bed reactor in plug flow 159 A-2 Potential distribution in a particulate electrode . . . 161 A-3 Schematic representation of eq. A-14 164 A-4 Schematic representation of eq. A-15 6 A-5 Schematic representation of a batch recirculation system 168 vii ACKNOWLEDGEMENTS I would like to thank my supervisor Prof. Paul Watkinson for his advice and encouragement throughout the whole of this work. I am also grateful to Prof. Colin Oloman for his sincere interest in the project and for the many useful discussions we had together. Thanks are also due to my husband, Gustavo Sucre for his sug gestions, patience, and understanding. I wish to express my appreciation to the Chemical Engineering staff for their cooperation and assistance and to the personnel of the Environmental Engineering Laboratory in the Civil Engineering Department for their generous help in the operation of analytical apparatus. Also acknowledged are Mrs. Rima Kaplan for her translations from Russian papers, Mrs. Monica Gutierrez for the drafting of figures, and Mrs. Nina Thurston for typing the manuscript. Financial support from the Venezuelan Government through FONINVES (Fondo para la Investigacion en Materia de Hidrocarburos) is gratefully appreciated. viii CHAPTER 1 INTRODUCTION 1.1 Phenols as pollutants "Phenols" in waste water treatment terminology includes not only phenol (CgHsOH), but all those derivatives of the aromatic ring that con tain one or more hydroxyl groups. Phenols are constituents of many industrial waste water streams. The major sources of phenolic wastes are oil refineries and coke plants. Phenols are finding increasing use in coatings, stripping agents, sol vents, paint vehicles, plastics, explosives, rubber substitutes, fertilizers, wood preservatives, and drugs. Given the usefulness of phenolic compounds, they will undoubtedly continue to be a major product of the chemical industry. Unfortunately, some of the chemical charac teristics that make phenols so useful, are also responsible for their high pollution potential. Chlorine used in drinking water combines with phenols to form chlorophenols which are persistent pollutants, since they are not easily degradable in the environment. Concentrations as low as 5 ug/1 of phenols will impart objectionable tastes and odours to drinking waters when phenols are combined with chlorine (1). For this reason the U.S. Public Health Service has set the allowable concentration of phenols in drinking waters at 1 yg/1 (2). Phenols are toxic to fish at levels above 2 mg/1, but can cause taste in fish flesh at concentrations far 1 2 below the toxic level (3). The chemical oxygen demand, C.O.D. of phenols is relatively high (theoretically 2.4 mg 02/mg phenol) and in sufficient concentration can deplete the oxygen of a receiving body of water causing the death of vegetable and animal species. Permissible levels of phenols have been established by the U.S. Environmental Protection Agency (E.P.A.) for different industrial wastes. These guidelines generally limited phenolic concentrations to 0.1 mg/1 in 1977 and project a standard of 0.02 mg/1 for 1983 (3). 1.2 Methods of treatment of phenolic wastes The concentration of phenol in industrial effluents varies widely as do the effluent flow rates (4,5). Generally recovery is only applicable for wastes of at least 2000 mg/1 of phenol and flows in excess of about 50 G.P.M. (2). Phenols may be recovered by liquid-liquid extraction processes using organic solvents such as benzene, butyl acetate, or butyl alcohol. These methods show efficiencies of recovery up to 99.7%, but the concen trations remaining in the aqueous phase after recovery are still signif icant from the pollution control point of view. Therefore the waste stream requires further treatment before being discharged. The choice between recovery or destruction of the phenolic content of a given stream is made on the basis of economics. Solvent extrac tion, for example, has been found to be an extremely expensive alter native in many cases (1). There are several conventional methods for treating phenolic wastes that cannot be economically recovered. These include adsorption, 3 incineration, biological treatment, and chemical oxidation. Carbon adsorption is applicable for relatively low (100-200 mg/1) phenolic concentrations. Thus it may be necessary to pretreat or dilute the waste stream before it is applied to the carbon beds (3). The main disadvantage of the activated carbon process is that the carbon has a finite capacity for removing phenols (0.09-0.4 g phenol/g activated carbon) and eventually the bed becomes fully loaded. To make the cost of the operation reasonable the carbon must be re activated and re-used. Chemical and thermal regenerations are possible. The first produces a more concentrated phenol stream and the second destroys the adsorbed phenol completely. Very high temperatures are required for this purpose (900°C) and carbon losses of 5-10% can result from the operation. Operating costs of the activated carbon process have been compared with those of other proven treatment methods in a recent study (1) and it was found that activated carbon was the most expensive. Incineration techniques are only applicable to concentrated wastes. In the case of dilute phenol solutions, the cost of energy to evaporate large amounts of water would be prohibitive. Typical operating tempera tures for combustion of phenol to carbon dioxide and water are as high as 800°C. Biological treatments for the degradation of phenolic wastes are applicable for concentrations up to several thousands mg/1. Many bio logical plants report treated effluents in the range of 0.1 mg/1 for influent loads of about 1000 mg/1 (6). Different biological treatment flow schemes, such as alternating activated sludge, trickling filter, or lagoon can be used, but the 4 activated sludge system is the most common. A very critical aspect in the success of biological treatment is the control of shock loads to the system, because the microorganisms are only adaptable to a certain range of phenol concentration and stable conditions of pH and temperature. Therefore, in many cases it has been necessary to provide an equaliza tion basin before the biological treatment. Operating costs of bio logical treatment are relatively low, but generally large land areas are required which may result in substantial capital costs. Chemical oxidation of phenols includes treatment by hydrogen perox ide, potassium permanganate, ozone, and chlorine dioxide. Depending on the dose of oxidizing agent the phenol can be completely oxidized to carbon dioxide and water or only partly converted to certain intermed iate, less harmful organic compounds. In the latter case, additional treatment may be required to reduce the total organic carbon (T.O.C.) or the chemical oxygen demand (C.O.D.) of the waste to acceptable levels. Oxidation by hydrogen peroxide can provide 99% phenol removal and about 40% C.O.D. removal when using a ratio of 2 g H£02/g phenol (3), but when substituted phenols are present the amount of peroxide neces sary can increase to about 4 g H202/g substituted-phenol. The presence, of a metal catalyst is required in the oxidation reaction which increases significantly the operating costs, and for this reason potas sium permanganate is a more desirable oxidizing agent. Potassium permanganate requires a higher oxidant to phenol weight ratio for the oxidation reaction—theoretically 15.7 g KMnO^/g phenol. The main disadvantage is that the reaction produces a precipitate of manganese dioxide that has to be removed, complicating the operation and increasing its cost (1). 5 Ozonization can be very effective in the destruction of phenols. For example, starting at 2500 mg/1, 99% removal can ba achieved in 60 min, when using a ratio of 1.7 g ozone/g phenol. Ozone can oxidize the phenol completely to CO2 and water but the usual practice is to partially oxidize the phenol to organic compounds more easily biodegradable and then use a biological treatment. This is done since at low phenol con centrations the necessary ozone to phenol ratio is too high to be econ omical. Initial costs are relatively high because the ozone generating system has to be installed. The products of the oxidation of phenol by chlorine dioxide are very dependent on pH. At near neutral pHs, in the range 7-8, phenol is oxi dized to benzoquinone with a theoretical requirement of 1.5 g Cl02/g phenol and above pH 10 the products are maleic and oxalic acid requiring a weight ratio of 3.3. Chlorophenols are not produced by this process because the benzene ring is completely destroyed. An economic study made on the oxidation of phenolic coking wastes by CIO2 indicated that the process was excessively expensive unless the oxidizing agent was already being produced on site (3). New methods for treating phenolic wastes are being nought, because of the importance of the pollution problem and the highly restrictive future pollution control standard. Increasing interest is being shown in methods such as Gamma Irradia tion (7), wet air, and catalytic oxidation (8), ultraviolet oxidation (9), and electrochemical oxidation, which is the subject of the present study. In the following chapter the fundamental bases of the electrochemical process are presented along with a literature review on previous attempts at electrochemical oxidation cf phenol. CHAPTER 2 BASES OF THE ELECTROCHEMICAL PROCESS 2.1 General concepts For any electrochemical reaction j of the form B + ze" Z A the reversible equilibrium potential is written as Tj - *j" I '» where the activity coefficient of each species (i.e., f^ = a^/C^) is equal to unity (10). The electrode potential is defined as the difference between the potential of the metal of the electrode and the potential of the solution adjacent to the electrode (Fig. 1). Thus, Anode potential = v = d> - cb [2] r a ma sa Cathode potential = V = <j> - <j> [3] * c mc sc The rate of an electrochemical reaction is a function of the over-potential, or difference between the electrode potential and the equili brium potential for reaction j. The anodic and cathodic overpotentials for reaction j, are respectively, n . = v * - v. [4] aj a 3 The current density (i) is defined as the amount of current passing per unit area of the electrode, and may be related to the overpotential, either linearly at low overpotentials (n a i) or through the Tafel equa tion at high overpotentials, H. - a. + b. log i [6] 3 3 3 In a given electrolyte, the Tafel constants a and b have specific values for each electrochemical reaction j occurring on a given electrode at determined conditions of pH and temperature. These have been reported for common electrode reactions on different electrodes (10,11). A side reaction will occur if the potential of the electrode is equal to the total potential required to drive the side reaction (equilibrium poten tial plus overpotential). Major side reactions associated with electro lytic processes in aqueous solutions are the reactions of water electro lysis, that is, the anodic formation of oxygen and the cathodic formation of hydrogen. Depending on the pH of the electrolyte and potential of the electrod different water electrolysis reactions, may occur, (Standard reduction Oxygen evolution reactions: potentials) 2 0H~ t % 02 + H20 + 2e~ V° = 0.4010 [Rl] H20 t % 02 + 2H+ + 2e" V° = 1.2290 [R2] Hydrogen evolution reactions: H20 + e~ t hE2 + OH"' V° =-0.8277 [R3] H+ + e~ t h H2 V° =0.0000 [R4] Side reactions will compete with the desired electrochemical reac tion for current so that the applied current density will be the sum of the partial current densities supporting each reaction. Considering 8 the electrochemical reaction (A -* B + ze ), the current efficiency for the oxidation of A is defined as the ratio between the theoretical amount of electricity needed to oxidize one equivalent of A, and the actual amount of electricity passed per equivalent of A oxidized. Thus, an expression for the percent of current efficiency is: % C.E. = x 100 [7] where, m = number of moles of A oxidized. If the oxidation of A is the desired electrochemical reaction, sometimes it is necessary to suppress the reverse reduction reaction (depending on the relative reaction rates). In order to avoid contact between the oxidized species and the cathode, or to prevent mixing of anolyte and catholyte with possible reaction an ion exchange membrane or a diaphragm can be used to separate the anode and cathode chambers in a divided cell. A diaphragm allows the transport of ions and molecules. An ion exchange membrane can be either anion selective or cation selective, in which case only anions or cations respectively will be transported through the membrane. Due to the selective properties of ion exchange membranes, they can also be used to control the pH of the anolyte and catholyte, since the transport of [OH ] or [H+] ions will be determined by the type of membrane used. For the general case of plane electrodes and two separate chambers, the different potential drops through the cell are illustrated in Fig. 1. 9 ig. 1. Voltage components in a divided electrolytic cell. If K , K , and K are the electrxcal conductivities of the anolyte, e e, e . • J a d c diaphragm, and catholyte, and are uniform, the Ohm's law can be written as follows: K I A<p I K I A<j> , I K I A<p I e 1 sa1 e ' d' ° 1 r<=^i i = e ' sc1 c S. a 'd c Therefore, the total ohmic drop is given by: S . S - . £ AV . . = A(J> + A<}), + A<(> = i(-^ + —S-+ —-). ohmic Ysa Yd Ysc K K K ' e e, e a d c When the anolyte and catholyte have appreciably different compositions, an extra potential drop may exist, called "liquid junction potential" ; 10 (10) but usually it is relatively small. The total electrolysing voltage for a current density i would be, AV = Va*+ |Vc*| + i + [8] e e., e a d c The total electrolysing voltage is of importance since the operating cost of the electrolytic process will depend on its power requirements which is directly related to the total voltage drop through the cell at a given current density. Practically, the total electrolysing voltage is easier to measure than the particular electrode potentials because to measure or Vc a reference electrode has to be connected at the surface of the anode or cathode to detect the difference in potential between the metal and the solution. The rate at which an electrochemical reaction occurs depends strongly on the electrode potential as will be shown below. Consider a single reversible reaction that occurs on an electrode (anode or cathode), 1 A X B + ze 1: oxidation reaction 2 2: reduction reaction The net rate of electrochemical reaction over the electrode is the modulus of the difference between the rates of oxidation and reduction of the reversible reaction m2 mi -=|Kr2CB -KriCA | [9] s s where C. and C„ are the concentrations of A and B at the surface of A B s s the electrode, mi and m2 are the orders of the oxidation and reduction reactions respectively, and Kri and Kr2 are the electrochemical reaction rate constants, which can be expressed in terms of the electrode poten tial by using an Arrhenius type of rate constant-activation energy relationship, Kr 1 = Kr° exp 1— Kr = Kr! exp 2 2 RT J f(l-a) zF |v*r RT [10] [11] Here Kr£ and Kr° are rate constants referred to a particular electrode potential at standard conditions, and a is a constant known as the charge transfer coefficient. These equations imply that a fraction of the • *• electrode potential ct|V | drives the forward reaction and the remainder (1-a) |V I drives the reverse reaction. The fact that the electrochemical rate constant depends exponen tially on the electrode potential and not just on the temperature as in the case of a pure chemical reaction, illustrates that the reaction rate can be varied by orders of magnitude by simply adjusting the potential., At a given reaction rate, the concentration of reactant at the surface of-the electrode will be related to the rate of mass transfer. .. from the bulk of the solution to the electrode surface, Mass transfer flux = K (C. - C. ) [12] . m -\ As where K is the. mass transfer coefficient, characteristic of the partic-m ular electrode configuration and fluid dynamics. Empirical and theor etical expressions for transfer coefficients, suitable for design purposes are available in standard texts (12). 12 2.2 Literature review on the electrochemical oxidation of phenol A substantial literature exists related to the electrochemical oxidation of phenol. However, owing to the complexity of the oxidation reactions and the variety of operating conditions used in each study, mechanisms have been proposed, some of which appear highly speculative. In order to consider possible rate determining factors, a review of the literature is presented here and some of the contradictions are dis cussed. 2.2.1 Reaction products The anodic oxidation of phenol was extensively studied by Fichter and. co-workers (13-15) during the early part of this century. They reported that when phenol is oxidized at a lead dioxide electrode in sulphuric acid media the products shown in Fig. 2 are involved. reported findings are sometimes contradictory. Many different reaction OH QH 0 I! 0 CH-COOH CH-COOK rt OH phenol hydroquinone p-benzoquinone maleic acid OH OH ether of pyrocatechol 2>4' dihydroxy diphenyi 4,4' dihydroxy diphenyi Fig. 2. Reaction products. 13 It was also found that if benzene (13) is electrochemically oxidized at a platinum electrode in sulphuric acid solution some of the same products were encountered, e.g., hydroquinone, p-benzoquinone, catechol, maleic, and oxalic acids. It was suggested (11,13) that probably phenol was first produced as an intermediate in the oxidation of benzene, even though phenol had not been isolated from the reaction mixture. Thus, information concerning benzene electrooxidation can be useful for the , present study. 2,2.2 Proposed reaction mechanisms a) Hydroxylation The formation of hydroquinone and catechol was attributed to the introduction of hydroxyl groups into the aromatic ring, by the action of anodically generated oxygen. By the same mechanism phenol would be produced if benzene was the starting substrate (13). However, this assumption is contradicted in a more recent paper (16) where the oxida tion of benzene to p-benzoquinone is reported at 100% current efficiency at potentials below those at which oxygen evolution occurs. But a clear explanation of the mechanism is not given in the paper. b) Nuclear linkage Fichter explained the formation of diphenyi derivatives by supposing that a linkage of two aromatic nuclei was brought about by a bond of oxygen, [R5] A similar mechanism would produce all the diphenyi compounds indicated in Fig. 2. Those compounds were found when lead peroxide anodes were 14 employed at a relatively low current density (25A/m2). Fichter also reported that the nuclear linkage,.is even more pronounced with phenol homologues like o-Cresol due to the presence of an electron donating group. The diphenols were found susceptible to further oxidation upon continuing the electrolysis or upon increasing the current density. In more recent publications (17-22) the presence of d.iphenyl derivatives has not been reported among the products of oxidation of pure phenol solutions. Generally the reported products are hydroquinone, benzo-quinone, maleic acid, and carbon dioxide. It is possible that under the reported conditions the diphenyl derivatives either are not formed, or are further oxidized to the above mentioned end-products. However, in the case of substituted phenols, the presence of coupled products is always reported in recent studies (23,24) which is in agreement with Fichter's findings. c)• Primary electron, transfer mechanism In the oxidation of phenol, two different, first-step mechanisms of electron transfer have been proposed (23,25), due to the ability of phenols to exist in the ionized or unionized form depending on the pH of the solution, -r OH 0 4- H+ unionized form phenoxide ion [R6] At low pH in aqueous solutions phenols will tend to be in the unionized form, and at high pH values will tend to be as a phenoxide ion. The following mechanisms have been proposed, for the first step of the oxidation: In acidic solutions the initial step involves two electrons where the 15 electrophilic attack of the aromatic nucleus produces the "phenoxonium ion", - 2e H" 0 -y U - ^ phenoxonium ion (mesomeric) [R7] In alkaline solutions the primary anodic reaction of phenoxide ions is a one-electron transfer with the formation of a phenoxy free radical that is very reactive. • • - Ie. phenoxi radical [R8] In appendix 5, the percentage phenol ionized as a function of pH is calculated from the dissociation constant for phenol at 20°C (K, = 1.28 x 10~10). d d) The divided or undivided cell and the reaction mechanism-A great deal of information regarding the electrochemical oxidation of phenol exists because of commercial interest in the production of hydroquinone or" p-benzoquinone (.19-22). Covitz studied the electro chemical oxidation of phenol for hydroquinone production at lead dioxide, anodes in an undivided cell in acid media. He showed that the reaction can be controlled to produce hydroquinone at over 90% yield. The simplified mechanism for the electrolytic process proposed is (26, p. .157): OH It is of interest to note that in the anodic reaction, water is utilized to introduce an oxygen atom into the starting phenol molecule. In the undivided cell p-benzoquinone is reduced at the cathode to produce hydro quinone. From this reaction scheme it is obvious that if the process is carried out in a divided cell, by using a membrane or a diaphragrc. p-benzoquinone would not contact the cathode and therefore would not be reduced back to hydroquinone. Covitz reported (19) that when using a semipermeable membrane, the only measurable product in the anolyte was p-benzoquinone. Another possible reaction in an undivided cell is the oxidation cf hydroquinone back to p-benzoquinone which would compete with the phenol for oxidation at the anode, thus lowering the current efficiency for phenol oxidation. e) Electrolytic action of lead dioxide Some authors (16,23,26) support the hypothesis of electrocatalytic oxidation of phenol on lead dioxide. In other words, phenol is oxidized chemically by lead dioxide and the reduced lead species so formed are rapidly oxidized back to Pb02 by a charge transfer or electrolytic step. 1) Pb02 + ORGANIC + PRODUCT + Pb+2 [R12] This mechanism is suggested as an alternative explanation to electron transfer from the organic molecule. Experiments have been carried out to determine the oxidative ability of Pb02 in the absence of current. Clarke and co-workers (16) made studies with granular Pb02 in stirred benzene emulsions. The analysis of the products showed that benzoquinone and maleic acid were rapidly formed. However, .numerical results were not provided. As discussed previously, the same paper contradicts the hypothesis of hydroxylation as a mechanism of introduction of oxygen into the organic molecule. Thus if Pb02 is supposed to be the oxygen carrier it would be necessary to replace the oxygen lost in reaction [R12] by reaction [R13]. The ultimate source of oxygen, is of course, the water, whatever is the prevailing mechanism. f) Further oxidation of intermediates The available information concerning the last stages of the electro lytic oxidation of phenol to open chain organic compounds or eventually to carbon dioxide is very limited, probably because only a few investiga tions have been concerned with the total destruction of the organic substrate for waste treatment applications (i.e., 17,18,25). A mechanism for the last-stage reactions has not ever been proposed. However, in the early investigations (14) it was shown that intermedi ates were susceptible to further disintegration. Catechol, for example, was well known as an easily oxidizable substrate, and p-benzoquinone, which offered a high resistance to chemical oxidation, was shown to be readily broken down by electrochemical means (27). Fichter established that the decomposition process occurs faster at high current densities, when the aromatic nucleus is saturated with elec trolytic oxygen. He reported less characteristic final products such as oxalic acid, formic acid, and carbon monoxide, which is in agreement with some of the products reported by Gladisheva (17). A particularly interesting controlled potential study is presented for the electrolytic oxidation of benzene (16) where it is shown that as the potential is increased above that of oxygen evolution, fragmentation of the aromatic ring occurs and the current efficiency for maleic acid and carbon dioxide production increases. Anodically generated oxygen is known as one of the most powerful oxidizing agents, and seems to be responsible for the further oxidation of the intermediates, even though the exact reaction mechanism has not been determined. 2.2.3 Electrode materials tested Lead dioxide has been the most commonly used electrode in the electrochemical oxidation of phenol and in several cases is recommended as the electrode material of choice. However, it is of interest to review and compare the performances of different electrode materials. In the paper by Gladisheva and Lavrenchuk (17) several anode mater ials were tested: nickel, smooth platinum, graphite, and lead dioxide electrodeposited on a nickel base. The experiments showed that under the same operating conditions, the highest oxidation rate occurred on the lead dioxide electrode. The results are shown in Table 1 where the rate of oxidation is given at two phenol concentrations and two current den sities. The chemical stability of the different electrodes tested was TABLE 1 RATES OF PHENOL OXIDATION ON DIFFERENT ELECTRODE MATERIALS (17) Electrode Initial phenol Rate of Oxidation (mg/min] material cone, (mg/1) i = 50A/ m2 i = 1000 A/m2 Electro deposited 200 1.0 3.7 lead dioxide 1000 9.2 21.6 Graphite 200 0.7 2.4 1000 6.3 17.0 Smooth platinum 200 0.4 2.4 Nickel 200 0.2 2.9 also discussed. The-graphite anode was found relatively stable at cur rent densities between 50-250 A/m2 but at higher current densities the graphite started to break down, forming small particles that were diffi cult to remove by filtration. Nickel electrodes were unsuitable since at pH = 10 nickel dissolution occurred parallel to phenol oxidation, consuming a significant amount of current, and destroying the electrode. The smooth platinum electrode was, of course, electrochemically stable but the rates of oxidation of phenol were much lower than expected, considering that platinum has a high oxygen overpotential. This fact was explained by the formation of a tar film on the surface 20 of the anode which did not dissolve in alkaline or acid solution. How ever, the presence of such a film was not mentioned in the case of the lead dioxide electrode. In the study of electrochemical oxidation of phenol to quinone by Fioshin et al (22) the same result was obtained when comparing the plat inum and lead dioxide electrodes. The chemical yield to quinone was 33% on the lead dioxide anode, whereas it was only 5% on platinum, although it is known that the overpotential of these electrodes in acidic solutions are practically the same. The reason suggested was the differ ent adsorptive powers, of the two electrodes towards the same organic substrate. However, the lower quinone yield on platinum could also have been caused by further disintegration of the quinone. This possibility was not suggested and other products analyses were not performed. 2.2.4 Effect of current density Studies have been carried out over a wide range of current densities. For example, in reference (17) the current effect on phenol oxidation was studied in the range of 50-2000 A/m2. It was concluded that among the variables studied, the current density was the strongest determining factor in the rate of electrochemical oxidation of phenol. The results are given in Tables 1 and 2. In Table 1 it can be observed that on the Pb02 electrode, at the initial concentration of 1000 mg/1 of phenol, the rate of phenol oxidation approximately doubled when the current density was increased 20 times. Table 2 shows that in the sodium sulphate electrolyte, when starting at 466 mg/1 of chemical oxygen demand (C.O.D.) at a current density of 50 A/m2, the final C.O.D. was 420 mg/1 after 5 h, whereas at 2000 A/m2 the C.O.D. dropped to 30 mg/T in only 1 h. 21 TABLE 2 EFFECT OF CURRENT DENSITY AND TYPE OF ELECTROLYTE ON C.O.D. REMOVAL (17) Time of Current density Electrolyte Electrolysis Final C.O.D. A/m2 type (h) (mg/1 of 02) 50 I 5 307 II 5 420 500 I 3 90 II 5 120 1000 I 1 30 II 2 75 2000 I 0.5 0 II 1.0 30 Notes. Initial phenol concentration = 200 mg/1 Initial C.O.D. concentration = 466 mg/1 of 02 Electrolyte I - 1 g/1 NaCl, 1.5 g/1 Na2S0it Electrolyte II - 3 g/1 Na2SO^ 2.2.5 Effect of nature of the electrolyte In several studies on the electroxidation of phenol for waste treatment, chloride salts were used as electrolytes (17,18,25,28-30). In such media the oxidation of phenol follows totally different reaction paths. When using NaCl or CaCl2 as electrolytes, the following reactions have been proposed (17) 1. Evolution of chloride at the anode, 2 Cl" - 2e~ -» Cl2 [R14] 22 2. a) Formation of hypochlorite followed by chemical reaction with phenol, Cl2 + H20 HCl + HCl© [R15] 8 HClO + C6H5OH •> CH - CO-OH II + 2 C02 + 8 HCl + H20 CH - CO-OH [R16] b) Chlorination of-phenol by molecular Cl2 producing 2,4 dichlorophenol and 2,4,6 trichlorophenol As can be seen, this process does not represent pure electrochemical oxidation of phenol, but instead it is equivalent to the electrolytic production of chlorine and hypochlorite followed by a chemical oxidation of phenol. It also gives rise to undesirable chlorination products. Although these are claimed to be capable of further oxidation, to give products of the quinone type, the removal is never 100%. This was shown in refer ence (25) where experiments were carried out starting with p-chlorophenol, 2,4 dichlorophenol and 2,4,6 trichlorophenol and the percent removal of chemical oxygen demand (C.O.D.) were 82%, 79%, and 58% respectively. Knowing that chlorinated phenols are more objectionable than phenol itself, the addition of chloride salts to the electrolyte does not appear to be a good solution to the pollution problem even if the rate of phenol oxidation is higher than when using an inert support electrolyte such as sodium sulphate. The performance of a mixture of NaCl and Na2S0it and pure Na2S0i+ electrolytes have been compared (17) in terms of final C.O.D. after treatment. The results are also shown in Table 2 where it is observed that at the four current densities used, the final C.O.D. was always lower when the electrolyte contained NaCl than when the electro lyte was pure Na2S0tf. However, an analysis of chlorinated phenols which 23 may have been produced was not provided, thus it is difficult to decide which electrolyte is more suitable. Other electrolytes, such as Na2Bit0y, NH3 and H^SO^ were tested using a packed bed graphite electrode (18). The results are shown in Table 3. TABLE 3 EFFECT OF TYPE OF ELECTROLYTE ON PHENOL OXIDATION (18) Electrolysis Final phenol Electrolyte pH time (h) cone, (mg/1) 0.1 M sodium borate 9.7 3.5 420 0.1 M ammonia 10.0 5.5 400 0.1 M sulphuric acid 1.1 3.7 400 5% sodium chloride 5.7 20-0 10 Initial phenol cone. = 1000 mg/1 Volume of electrolyte = 700 ml Current = 0.6 A (on graphite anode) With the first three electrolytes phenol was removed from 1000 mg/1 to about 400-420 mg/1 in electrolysis times from 3 to 5 h. The final phenol concentration reached 10 mg/1 only when using a sodium chloride electrolyte and at the excessively large time of 20 h. The different electrolysis times allowed makes the comparison between electrolyte performances difficult. In this case the current applied was relatively low and an estimation of the electrode area was not provided. 24 2.2.6 Effect of pH The effect of pH on the oxidation potential of phenol has been reported (23). The results of this polarographic study are represented in Fig. 3. > r 1 1 1 1 r 1 < 02 1 1 1 1 1 1 : 1 1 X 0 A 8 12 pH Fig. 3. Half-wave potential vs pfi, for the oxidation of 4 x 10"4 M phenol (23). The half wave potential is defined as the potential on a polaro graphic curve when the current is equal to one half the mass transfer limiting current (31). It can be observed that the half wave potential decreases when going from acid to basic solutions and eventually stabilizes at a constant value for a pH equal to the pK^ (logarithmic of the dissociation constant of phenol). At pH = pK^ all the phenol will be in the ionized form, or in other words, protonation will be negligible. This means that a high pH makes the phenol more easily oxidizable as far as potential require ments are concerned. The effect of pH on the rate of phenol oxidation is not well docu mented. In reference (17), it was concluded that the velocity of oxidation was practically independent of pH of the solution in the range 25 of pH 6 to 9, in absence of chloride ions. However, when the pH was changed to 11.9 a rapid increase in the optical density of the solution was reported, which was explained by an increase in the concentration of hydroquinone. However, phenol or C.O.D. analyses were not reported in this case. In the study by Tarjanyi et al (18) the effect of pH can not be isolated from the data (Table 3). No definitive results could be found in the reviewed literature about the effect of pH on the further oxidation of intermediate products. Fichter (13) suggested that in alkaline solu tions the primary products of the oxidation would probably be the same as in acid media, but that p-benzoquinone would be unstable at high pH and more easily oxidizable by the atomic oxygen. 2.3 The lead dioxide electrode Two types of lead dioxide commonly exist. These differ according to the crystal structure. a-Pb02 is orthorombic and B-Pb02 is tetragonal. Each variety can be prepared substantially free from the other under carefully controlled conditions (32). Lead dioxide never conforms exactly to the stoichiometric Pb02 formula, an oxygen deficiency is always detected (33), with a-Pb02 tend ing to show a lower oxygen content. Contrary to most metal oxides, lead dioxide is a good electronic conductor, and is better in fact than lead itself (26). It is believed that the high electrical conductivity may be connected with the oxygen deficiency in the lead dioxide structure (32). Potential-pH diagrams (or Pourbaix diagrams) show the regions of thermodynamic stability of lead and lead compounds (34). From thermo dynamic predictions lead can be used as an anode at high electrode 26 potentials for pH values between 0 and 12 without appreciable corrosion. Under such conditions the metal will be covered with a layer of Pb02. Delahay et al (35) have constructed the potential-pH diagram for lead in the presence of sulfate ions (1 g-ion/1). Many electrode reactions are thermodynamically possible at different potentials and pHs. From these, the most studied reactions are those that form the basis of the universally used lead acid storage battery, Pb02 + 2e~ + 4H+ + SOu~2 t PbSOi* + 2 H20 V° = 1.685 [R17] Pb + SOLT2 t PbSOtt + 2e~ V° = -0.356 [R18] The reactions occur in the indicated direction during discharge and in the opposite direction during charging (23). It is well known that the mech anism of discharge of Pb02 in the presence of excess sulfate involves blocking of the Pb02 surface with a deposit of PbSOLv (32) . These reac tions may also be of importance in the oxidation of phenol in sulphuric acid media. The suitability of lead dioxide as an anode material has been known for many years. It is clear that lead dioxide is able to withstand prolonged high anodic potentials more effectively than graphite (which undergoes degredation). Also, lead dioxide possesses a relatively high oxygen overvoltage (10,36) of the same order of magnitude as platinum and is much cheaper. Preparation. Lead dioxide can be prepared as a coating on lead by anodization, or deposited onto other metals by electrodeposition. a) Anodization. The most conventional anodization method is to put the lead in contact with an aqueous H2S0t| electrolyte and provide a flow of current until oxygen evolution is plainly visible, and the grey lead has acquired the characteristic black deposit of lead dioxide (26,33). It is well known that oxygen is evolved from a lead anode only when a layer of lead dioxide has been laid down (32). After anodization, the electrode should not be left in contact with sulfate ions to avoid losses by reductive processes (Reaction R17) and should be used as soon as possible (26). Some authors (37,38) have postulated that a shortcoming of the preparation of Pb02 by anodization of Pb is that a solid phase reaction occurs between the Pb02 and the underlying Pb to produce the less con ductive PbO, Pb02 + Pb t 2 PbO [R19] However, another'study (39) reports that the only products of the anodi zation of lead in H2SO4 observed by x-ray diffraction, were ct-Pb02, B-Pb02 and PbSOt+j but no PbO was detected, even after several weeks of storage of the electrode in the dry state. In the same study a mechanism is proposed for the anodization of lead. First, PbSO^ is formed from Pb at the Pb/PbSO^ potential, and later when the potential rises to the oxygen overpotential value, the PbSOij film transforms to 8-Pb02 and the underlying grid metal is converted directly to a-Pb02-It has been reported (36) that anodized lead can not tolerate the presence of chloride ions which cause it to disintegrate. b) Electrodeposition. There are several methods for the electro-deposition of Pb02 on inert metals from electrolytes containing lead. Some of the methods are summarized in reference (32). Pb02 has been deposited on nickel, tantalum, platinum, carbon, or graphite. Most other metals are unsuitable because of their inherently easy oxidation (26). Several types of electrolytes have been used for the deposition of Pb02, and of these, lead nitrate has been found to give 28 the.best deposits (40). The largest producers of commercial electrodeposited lead dioxide anodes in the world are Pacific Engineering and Production Co. of Nevada and Sanwa Chemical Co. Ltd., of Tokyo, Japan. Since the breakthrough by Pacific with a lead dioxide coated graphite anode, the greatest interest for Pb02 formation has been shown in the electrodeposition process (36). The electrodeposition on graphite uses a lead nitrate electrolyte in acid media as described by Gibson (41). The tetragonal 3-Pb02 is the form found in the commercial anodes produced from acid lead nitrate baths. The a-form is less common and can be deposited from alkaline solutions (26,36). Unlike the anodized lead, the electrodeposited Pb02 can operate effectively in chloride concentrations close to saturation. In fact the lead dioxide anodes produced by Pacific and Sanwa Co. are in use for perchlorate manufacture. ' CHAPTER 3 OBJECTIVES The aim of this work was to study the electrochemical oxidation of phenol for waste treatment applications. A packed bed anode was selected because it provides larger electrode surface areas per unit cell volume compared to a simple flat plate electrode. This is partic ularly important where dilute solutions are to be treated. The research reported here includes the design and construction of equipment to carry out the process and an experimental study of the effect of important operating variables. These variables include type of lead dioxide anode (anodized lead versus electrodeposited lead dioxide), cell configuration (divided or undivided cell), type of ion-selective membrane (anionic or cationic), current applied, pH of the electrolyte, conductivity of the electrolyte, phenol concentration, flow-rate, and particle size. Lead dioxide was selected as the anode material to carry out this study for the reasons given in Chapter 2. The performance of anodized lead and electrodeposited lead dioxide is compared in terms of phenol oxidation but also some tests were made to compare them in terms of corrosion resistance. The use of a divided or undivided cell is of importance, since it may completely change the reaction mechanism for the further oxidation of intermediates. There is little information in the literature on such effects. 29 The electrolytes to be used for the oxidation of phenol consist of mixtures of Na2S0i+ and H^SO^ or Na2S0L; and NaOH, to be able to vary independently pH and conductivity of the electrolytes. From the point of view of waste treatment, it is of interest to determine what fraction of the phenol is converted to carbon dioxide, or how much organic carbon remains in solution after the electrochemical treatment. Therefore, the effect of the important variables is reported not only in terms of phenol oxidation but also on the total organic carbon (T.O.C.) oxidation. In the reviewed literature no T.O.C. analyses have been reported. In some cases chemical oxygen demand (C.O.D.) have been reported. But this is not an adequate technique of analysis in the case of aromatic compounds (42). Relatively low phenol concentrations are used in this study (up to 1100 mg/1) in order to investigate the process until practically total phenol oxidation is achieved for the range of operating conditions of the experiments. A batch-recirculation system was selected as the operating mode to study the effect of some of the variables. Once the system was better understood and controlled, some experiments were performed in the contin uous mode. Finally, the experimental fractional conversions of phenol are com pared with the mass transfer model for a batch recirculation operation, and a simplified model including electrochemical reaction control is presented in order to analyze the data from the continuous experiments and compare the mass transfer and electrochemical reaction resistances. CHAPTER 4 EXPERIMENTAL APPARATUS AND METHODS 4.1 Apparatus 4.1.1 Cell design The electrolytic cell consists of a stack of elements arranged in series and compressed by a clamp mechanism. This flexible design is used because it permits the assembly of different arrangements (divided or undivided cell) and simplifies work with electrode materials, a) Divided cell A side view of the divided cell arrangement is shown in Fig. 4. Basically, the cell consists of two flat plates, the anode and cathode current feeders, which are in contact with the anodic and cathodic packings. Both packings are contained in 3 mm thick slotted neoprene gaskets and are separated from each other by an ion-selective membrane which prevents the mixing of the anolyte and catholyte. The cathodic packing is used to prevent the membrane from sagging due to the weight of the anodic packing. The anolyte and catholyte inlets are located at the bottom of the cell and the outlets at the top, to facilitate the exit of the gases that will be produced during the electrolysis. Two different kinds of lead dioxide anode plates are used in this study: anodized lead sheet and electrodeposited lead dioxide on graphite plate. In the case of the anodized lead electrode the current feeder plate 31 Fig. 4. Side view of the general divided-cell Legend a = 1.6 mm thick neoprene insulator b = 1.6 mm thick cathodic feeder plate (s.s. 316 plate) c = 3 mm thick slotted neoprene gasket containing cathodic packing d = ion selective membrane against protective plastic screen (variable thickness) e = 3 mm thick slotted neoprene gasket containing anodic packing f = anodic current feeder: 3 mm thick lead plate, or lead dioxide coated graphite plate 3 cm thick gement. (no scale) 33 was cut from a 0.3 cm thick lead sheet. The detailed front and side views of the anode chamber when using such electrode are shown in Fig. 5. A neoprene gasket determines the cross sectional area of the feeder plate that will be transporting the current. The side view shows that an extra stainless steel plate is used mainly to facilitate the welding of the inlet and outlet connectors to the cell and also to give more strength to the lead sheet. The front and side views of the cathode chamber are the same as the anode chamber, except that the thickness of the s.s. 316 cathode feeder was 0.16 cm. The electrodeposited lead dioxide on graphite plate was obtained from Pacific Engineering Co. The total thickness of the plate is 3 cm and the thickness of the lead dioxide coating on each side of the graphite is 0.2 cm. Some modifications had to be made to the original commercial electrode to adapt it to the cell design being used. The final front and side views of the electrodedeposited Pb02 anode are shotvn in Fig. 6. To avoid possible cracking of the Pb02 coating, the electrode was left with its original width of 15 cm. A modification had to be made in order to introduce the electrolyte flow through the graphite coated plate. A detail of the connection adapted is shown in Fig. 7. The electrolyte never comes in contact with the graphite base plate because the nylon connection was insulated by means of a neoprene washer. This type of connection prevents corrosion .. of the graphite base and eventual deterioration of the lead dioxide layer. Some fundamental specificationsof the different elements of the cell are given in Table 4. The. dimensions of the anode and cathode chambers were never changed, but different sizes of anodic packings were used. The:cathodic packing consisted of several stainless steel-304 screens Legend a = lead sheet anode feeder b = slotted neoprene gasket c = electrolyte outlet d = anode packing (lead shot) e = electrolyte inlet f = neoprene insulator g = stainless steel 316 plate (where connectors are welded) Fig. 5. Front and side views of the anode chamber for the anodized lead electrode. (no scale). Legend a = uncoated graphite section of the anode feeder b = electrolyte outlet c = anode packing d = slotted neoprene gasket e = electrodeposited Pb02 section of the feeder plate f = electrolyte inlet g = neoprene insulator Fig. 6. Front and side views of the anode chamber for the electro deposited PbC>2 electrode, (no scale) 2.54 cm A 6.3 mm Legend a = nylon connection b = stainless steel mesh c = neoprene washer d = lead dioxide layer e = graphite base plate f = compressing nut (threaded) Fig. 7. Detail of the inlet or outlet connection adapted on the electrodeposited Pb02 on graphite anode. (no scale) u> ON TABLE 4 FUNDAMENTAL SPECIFICATIONS OF THE ELECTROLYTIC CELL Dimensions of the anode and cathode chambers: Length = 38 cm Width = 5 cm Thickness = 3 mm Anodic packings: Particle size (mm) lead shot (to anodize) 2 electrodeposited lead dioxide 1.7-2.00 0.7-1.1 Cathodic packing: stainless steel-304 screens (20 x 20 mesh) cross sectional area of the screens (38 x 5)cm Membranes: cationic: IONAC MC 3142 IONAC MC 3470 NAFION 127 anionic: IONAC MA 3475 Protective screens: saran polypropylene 2 mm 2.54 cm Jl E o O OJ C: -6 cm E o CM . 6 E o •3-m c\j G: CLAMP WELDED TO SQUARE TUBES 0 i G= -5 cm-la) PLAN VIEW . Fig. 8. Detail of the mechanism used to hold the cell. (b) FRONT VIEW (no scale) 39 (20 mesh) cut to the size of the cathode chamber (5 x 38 cm2) and joined so that the total thickness of the packing was 0.3 cm. The different ion selective membranes tested in this study are listed in Table 4, along with the protective plastic screens. The properties and charac teristics of these materials as supplied by the manufacturers are given in Appendix 1. A detail of the mechanism used to compress the various parts of the cell is given in Fig. 8. It consists of four mild steel square tubes which are welded to six C-clamps (Jorgensen, Style 81). The cell is introduced through the upper part of the press mechanism, and once the C-clamp screws are tightened, the four square tubes compress the neoprene gaskets, providing an effective seal for the cell. This versatile press design permits variations in the thickness and width of the cell materials within a certain range, and also can be rapidly opened and closed. b) Undivided cell A side view of the undivided cell arrangement is similar to that represented in Fig. 4, except that the ion selective membrane and the cathodic packing are eliminated and the inlet and outlet of the cathode side are closed by using a cathode feeder plate without holes. In this case, only a plastic screen (saran or polypropylene) is placed between the anodic bed and the cathodic feeder plate. 4.1.2 Flow diagram of the apparatus Figure 9 is the schematic flow diagram. Equipment specifications are given in Appendix 1. Two main flow circuits exist. At the right hand side of the electro-J lytic cell is the anolyte flow circuit. Pump PU-1 delivers the anolyte Fig. 9. Flow diagram of the apparatus. o 41 Legend for Fig. 9. P.S.) Power supply (D.C.) V) Voltmeter E.C.) Electrolytic cell T-l) Anolyte tank (contains phenol solution) T-2) Anodization tank T-3) Catholyte tank T-4) Washing tank PU-1) Anolyte pump PU-2) Catholyte pump R-l) Anolyte rotameter R-2) Catholyte rotameter P-l) Anolyte pressure and temperature gauges P-2) Catholyte pressure and temperature gauges F-l) Anolyte filter F-2) Catholyte filter GL-1) Gas-liquid separator for the anolyte GL-2) Gas-liquid separator for the catholyte V-1) Anolyte tank shut-off valve V-2) Anodization tank shut-off valve V-3) Catholyte tank shut-off valve V-4) Washing tank shut-off valve V-5) Anolyte flow control valve V-6) Catholyte flow control valve V-7) Cathode chamber pressure-control valve V-8) Liquid sample valve V-9) Liquid level control valve in GL-1 V-10) Liquid level control valve in GL-2 42 from tanks T-l or T-2 to the anode chamber. The liquid flow rate is controlled by adjusting valve V-5 and is measured with rotameter R-l. Pressure and temperature of the anolyte at the entrance of the anolyte chamber are measured in P-l. Filter F-l is located at the outlet of the anode chamber, to collect small particles that might be withdrawn from the cell, thus protecting the pump from damage. This glass-wool filter serves also to agglomerate small gas bubbles produced in the electrolysis into bigger ones to facilitate the gas liquid separation in GL-1. The gas-liquid mixture enters at the bottom of the gas-liquid separator GL-1 in which a bed of glass beads provides extra agglomeration surface for the gas bubbles. Valve V-9 controls the liquid level at the outlet of GL-1, to ensure that gas bubbles are not carried out with the liquid flow. This would result in a progressive accumulation of gas in the anolyte line which may affect the results of the experiments. The gas is then released at the top of GL-1 and the liquid flows towards the feed tanks (T-l or T-2) to be recycled to the cell. The dotted line represents the recycle line when the anodization tank T-2 is in use. Valve V-8 serves to collect liquid samples after passage through the cell. The flow diagram corresponding to the catholyte circuit is basically analogous to the anolyte circuit above described. Valve V-7 serves to control the pressure in the catholyte chamber, providing pressure equal ization at both sides of the membrane, thus avoiding too high pressure differences between the anolyte and catholyte chambers that may result in membrane breaking. When the cell was assembled with only one chamber, only the anolyte circuit was used. The catholyte circuit was eliminated by closing the 43 catholyte inlet and outlet to the cell. The cell was powered by a 1 KVA D.C. power supply (Appendix 1). The cell current was read from the power-supply meter, and the voltage drop across the electrodes was measured independently. 4.2 Experimental methods 4.2.1 Batch experiments The experimental procedure is described for the more complicated two chambers-cell operation, since the one-chamber cell operation can be considered as a particular case of the first, a) Anodization process Before each experiment, the lead electrode was anodized by electro lysis in 20% l^SO^ (43), to ensure that the anode was equally active before every run. Valves V-1 and V-4 were shut off and tanks T-2 and T-3 were filled with a 20% H2S01+ solution. The D.C. power supply was turned on. Valves V-2 and V-3 were then opened and both pumps, PU-1 and PU-2, were activated at the same time. About 2 & of solution coming out from the gas-liquid separators was withdrawn at each side to purge the system before the anolyte and catholyte flows were recycled to tanks T-2 and T-3 respectively. Five l of r^SOtj solution remained in each tank for the anodization process. Immediately the current was adjusted to 10 A (c.d. = 526.3 A/m2) and the liquid pressures in both chambers were equalized by adjusting valve V-7. When lead was to be anodized for the first time, a 12 h anodization time was allowed, but for successive experiments the standard anodization time was 1 h. (Choice of the anodization time was justified experimen tally, as shown in Chapter 5.) After anodization, both pumps were 44 simultaneously turned off and valves V-2 and V-3 were closed. Then, tanks T-l and T-4 were filled with distilled water and connected to pumps PU-1 and PU-2 respectively. The cell was washed until the current dropped practically to zero and the potential difference through the cell increased indicating that essentially no electrolyte was contained in the cell. b) Phenol electrochemical oxidation process After the cell was thoroughly washed, 8 I of anolyte solution were prepared in tank T-l. The concentration of phenol, the pH and the con ductivity of the anolyte were set to the desired levels by adding the necessary volumes of stock solutions of phenol, NaOH or ^SO^, and N32S01+, which had been previously prepared. The tank was well agitated before the initial sample was taken, and pH and conductivity were measured and readjusted if necessary. An equal volume of catholyte solution was then prepared in tank T-3 and conductivity and pH were measured. Anolyte and catholyte data are recorded for each experiment in Appendix 2. Tanks T-l and T-3 were connected to the corresponding pumps and flow rates of anolyte and catholyte were set by adjusting valves V-5 and V-6 respectively. Pressure equalization was provided adjusting valve V-7. Immediately the current was set at the desired value. As the operating conditions were being set 3 i of the solutions coming out from the gas liquid separators were discarded in order to purge the system and also to provide some time for flows and current stabiliza tion. At the moment the liquids were recycled to tanks T-l and T-3, 5 H of electrolyte remained in each tank. The electrolysis time was measured from the moment when the anolyte was recycled to tank T-l. Thirty ml samples were taken in intervals of 15 min for phenol analysis by opening valve V-8. When the usual electrolysis time of 2 h was com pleted, the cell was washed with distilled water while the current was still flowing to avoid reduction of the Pb02 anode. The samples were first analyzed for phenol and T.O.G. and later pH and conductivity were measured, to avoid possible contamination of the samples when introducing the pH and conductivity probes. 4.2.2 Continuous experiments Some experiments were carried out in the continuous mode (without recycle), to test the effect of varying flow on phenol oxidation in a single pass through the undivided cell. The electrode pretreatment or anodization was carried out by the .standard method of electrolysis with 20% H2SO4 at 10 A for 1 h. In this case the phenol solution to be treated was prepared in the same manner described previously but the total volume of the electrolyte in tank T-l was 20 L After the initial electrolyte sample was taken, and conductivity and pH were measured and adjusted to the desired values, the D.C. power supply was turned on, and the electrolyte was fed to the cell. A liquid flow rate was then fixed by adjusting valve V-5, and the desired current was set. Four I of electrolyte were withdrawn at the outlet of GL-1 before the first sample was taken, to ensure stabilization of the process. A different flow rate was then set up and the same procedure repeated, until the liquid flow rate range provided by rotameter R-l was covered. In Appendix 2 the experiments are divided by groups according to cell assembly and operating mode. 46 4.3 Analytic techniques The samples taken at the outlet of the cell were analyzed for con centrations of phenol, total organic carbon, and in some cases, lead. 4.3.1 Phenol analysis Phenol concentrations in the samples were determined by gas chroma tography using a flame ionization detector. The analytic equipment specifications and operating conditions used are given in Appendix 1. Standard phenol solutions ranging from 0-116 mg/1 were prepared by pipetting from the same phenol solution used to prepare the electrolyte for the experiments. Copper sulfate was added to the standards to an approximate concentration of 1 g/1 to preserve them from possible bio logical degradation (42). Phenol peaks were thin enough so that the estimation of the area below the peak was not necessary, and peak heights could be used. Before the analysis of the samples from each run, the phenol standards were always injected and the calibration curve of peak heights vs phenol concentration was constructed. Standards and samples were injected until the variation in peak height was not more than 2-3%. This means that when the recorder full scale corresponded to about 100 mg/1 the maximum allowed variation repre-sented ± 2 mg/1, and the recorder detectability was 1 mg/l/division. No peaks other than the solvent and phenol peaks were observed, in other words no interference of the phenol oxidation products was detected in the G.C. analysis. Phenol detention time was approximately 30 seconds. All the samples were analyzed on the day of the experiment, even though it was shown that the concentration of phenol in the sample did 47 not vary after one week of storage. For those experiments performed at phenol concentrations higher than 100 mg/1 the samples were diluted by the necessary factor in order to work with the same standards and detectability. 4.3.2 Total organic carbon analysis All total organic carbon analyses were carried out on a Beckman analyzer in the Civil Engineering Department. Basically, the total organic carbon analyzer consists of two furnaces: the total carbon furnace and the inorganic carbon furnace. The total carbon furnace operates at a temperature of 1000°C to convert all the carbon contained in the sample to carbon dioxide. The inorganic carbon furnace operates at 150°C to convert only the inorganic carbon contained in the sample to carbon dioxide. The amount of carbon dioxide thus produced is detected in an infrared analyzer (44). The total organic carbon (T.O.C.) in the sample is calculated by subtracting the inorganic carbon from the total carbon, therefore two different calibrations are required, one for each channel. The stan dards used for the total carbon analysis were the same phenol standards used to calibrate the chromatograph. Carbon content ranged between 0-90 mg/1. Before the analysis of the samples from each run, the total carbon standards were injected and the calibration curve peak height vs mg/1 carbon constructed. Variations in peak height in successive injec tions of a certain sample never exceeded 2%. The inorganic carbon standards were prepared from a stock solution of Na2C03 and NaHC03 containing 1000 mg/1 of carbon, so that the inorgan ic carbon in the standards also ranged between 0-90 mg/1. After the calibration curve of mg/1 inorganic carbon vs peak height was constructed, 48 the samples were injected in the inorganic channel. Again maximum vari ations in peak height for a same sample were of the order of 2%. Knowing that phenol contains 0.7657 g C/g phenol, it is possible to compare the phenol concentration obtained by gas chromatography with the concentration calculated from the T.O.C. analysis, for the initial sample. Deviations between both analysis were usually in the range of 2-3% which gives confidence in the analytical results. All the specifications and operating conditions of the T.O.C. anal ysis are given in Appendix 1. The phenol standards for T.C. analysis showed no alteration after one month of storage when compared with fresh standards, and the samples from one experiment did not show variation in the T.O.C. analysis after one week. 4.3.3 Lead analysis To test electrode corrosion, the samples from some of the experiments were analyzed for lead using atomic absorption spectrophotometry. Lead standards were prepared by diluting a stock solution containing 1000 mg/1 of lead. The range of concentrations of the standards was chosen according to the lead content of the samples. The specifications of the atomic absorption apparatus used and its operating conditions are given in Appendix 1. CHAPTER 5 RESULTS AND DISCUSSION 5.1 Electrode materials Some preliminary experiments were performed to test possible electrode materials before starting to investigate other variables affecting the process. Results are given in the table for each run in Appendix 2. The anodized lead electrode was tested in experiments group No. 1. Runs 1-1 and 1-2 were carried out to test the effect of anodization time on T.O.C. removal. The results showed that 1 h and 12 h anodization times produced approximately the same % T.O.C. removed vs time. This indicates that 1 h anodization time is probably sufficient to coat the lead shot with a layer of Pb02- After each of these runs the electrode showed a fairly uniform brown coating. It is probable that the coating is thicker for the longer anodization time. After the anodization period and while the cell was being washed with distilled water, the liquid at the outlet took on a brown colour as if some fine particles from the electrode surface were entering the water. Therefore it was considered necessary to test for lead concentration in solution. A series of tests were carried out to investigate what governed the presence of lead in the anolyte. Experiments 1-4 to 1-8 were performed under different electrolyte concentrations and pHs. The results are summarized on p. 122. in all 49 50 these runs the concentration of lead in the anolyte tended to decrease with time, indicating that the corrosion rate is probably higher when the electrode first conies in contact with the solution. Also, the lead ions are able to pass to the catholyte chamber through the cation-membrane therefore reducing the lead content of the anolyte. (The pH changes in those runs occurred spontaneously as will be explained in the next sec tion.) In Run 1-4 the electrolyte (5 g/1 NaOH) was recirculated through the cell without an applied current and it produced the highest maximum con centration of lead in solution (140 mg/1). Later this run was repeated, except that a potential difference was applied before the electrolyte was fed.to the cell. The maximum amount of lead in solution was 2.7 mg/1. This observation is in agreement with the potential-pH diagram for lead which indicates that Pb02 is tbermodynamically stable at positive or anodic electrode potentials with respoct to the normal hydrogen electrode. Runs 1-5 and 1-6 showed that at the same initial pH of 9.8 the maximum lead concentration was 4.2 mg/1 when working at 30 g/1 Na2S0i+, and it was 1.7 mg/1 when working with a 5 g/1 Na2S0it electrolyte. At the same concentration of Na2S0i| (30 g/1) the amount of lead in solution was higher at higher initial pH-(Runs 1-6, 1-7 and 1-8 respec tively). An attempt was made to study the response of the anodized lead electrode under a 5 g/1 NaCl electrolyte (Run 1-3). It was not possible to finish the experiment because the electrode coating began to dissolve rapidly, and as a result the anolyte pressure increased and the voltage drop decreased. When the cell was opened the electrode had lost its brown coating, showing the underlying grey lead, and the membrane was covered with a deposit of some lead compound. 51 The performances of anodized lead and electrodeposited Pb02 can be compared from Runs 1-9 and 2-2 respectively. Both experiments were per formed under the same experimental conditions, the same cationic membrane (IONAC MC-3470) and approximately equal voidage fraction (0.6). The results are represented in Fig. 10. The curve for % phenol oxidized vs time when using the electrodeposited Pb02 electrode is above that curve corresponding to the anodized lead. The % T.O.C. removed vs time curves are practically coincidental for both electrodes, but the maximum con centration of lead in solution was about ten times higher in the case of the anodized lead. Scanning electronmicrographs of the electrodeposited Pb02 particles are shown in Fig. 11. Lead analyses are reported in many other experi ments, (experiments Groups 2 and 3) where electrodeposited Pb02 was used. In each case the maximum lead concentration was lower than 0.4 mg/1, and tended to zero towards the end of the run. The only exception was found in Run 2-3 where no current was applied and the lead concentration built up, reaching 2 mg/1 after 90 minutes. Scanning electron micrographs of the anodized lead particles (Fig. 12) show that flaking of the Pb02 film occurred. Once the electrode flakes are carried out from the cell they would dissolve more easily in the absence of the applied anodic potential. The anodized lead electrode could become more resistant to corro sion after successive anodizations or after long periods of use. Also it is possible that its resistance to corrosion could be improved by anodizing the lead very slowly using more dilute H2SO4 solutions, lower current densities and longer anodization times. .This might prevent the formation of cracks (where corrosion probably starts) that result from a 52 d 20 d H io 0 100 80 Q LU 60 X o _J Z 40 UJ X CL 8" =8* -A" -o-20 0 -A -O KEY RUN N2 ELECTRODE O 1-9 ANODIZED LEAD A 2-2 ELECTRO DEPOSITED Pb02 I i l . I 30 60 00 TIME (min) Fig. 10. Effect of type of lead dioxide electrode at 10 A and initial pH = 9.4 with IONAC MC-3470 membrane. Fig. 11. Scanning electron-micrographs of the electrodeposited PbC>2 particles after use. (particle sizes between 1.7-2.00 mm) Fig. 12. Scanning electron-micrographs of the anodized lead particles, after use. (prepared from 2 mm lead shot) 55 perhaps too strong anodizing action. An attempt was made to test a packed bed nickel electrode at pH = 12 and 10 A, but in two hours the voltage drop through the cell increased considerably and after opening the cell, it was observed that a precip itate was plugging the bed and covering the membrane. Tungsten carbide (WC) particles were also tested at pH = 12 and 10 A. The anolyte took a grey colour indicating dissolution of the elec trode and the total carbon test revealed increasing amount of carbon in solution. From these preliminary tests it was concluded that the electro deposited Pb02 was the most convenient choice to carry out the remainder of the experiments because of its better corrosion resistance and its higher % phenol oxidation when compared to the anodized lead. 5.2 Effect of pH using the divided cell Because of the lack of information in the literature about phenol electrooxidation in alkaline electrolytes, the effect of a basic pH range was investigated first using the divided cell. Before discussing the effect of pH on the electrooxidation of phenol, it is necessary to consider some important observations about the pH behaviour in the divided cell. Using the cation membrane MC-3470 which is suitable for alkaline electrolytes, the anolyte phenol concentration was set at 100 mg/1 and the pH was adjusted to 9.4. It was observed that after 15 min the pH of the sataple at the outlet of the cell had dropped to about 3, when working at. 10 A (Run 2-2, Fig. 13). Later, the same experimental con ditions as in Run 2-2 were repeated but without phenol being present and the same pH drop was observed. Thus, the pH drop appears to be due to side reactions of oxygen evolution, and not to production of an acid from phenol oxidation. What probably happens is that the hydroxyl ions are discharged first at the electrode, and then the mechanism of oxygen evolution changes (Reactions RI or R2) producing hydrogen ions. If the rate at which the hydrogen ions are produced is higher than the rate of trans port through the membrane, the net effect is a progressive increase in the concentration of hydrogen ions, which would result in the observed pH drop. This would also explain why the electrolyte conductivity increases towards the end of the run and as a result the voltage drop through the cell decreases (Run 2-2). If the side reactions of oxygen evolution are responsible for the observed pH behaviour, it follows that the pH response must be related to the current because the rate at which the side reactions occur will ultimately be determined by the current applied. The pH response was then compared at different currents (0, 3, 6, and 10 A) with all the other experimental conditions held constant. The pH vs time curves are also represented in Fig. 13. When working at 6 A (Run 2-5) the pH dropped from 9.4 to 3.7 in 15 min, showing the same tendency described previously for Run 2-2. A lower current of 3 A (Run 2-4) produced an unexpected result. In 15 min the pH of the sample at the outlet of the cell had increased from 9.4 to 11.7. The same behaviour was observed when the solution was recirculated without any current applied (Run 2-3). Two possible explanations are proposed for this effect. Since the membrane used in these experiments was the cationic MC-3470, theoretic ally it only allows the transport of positive ions but as the ionic 58 selectivity is not 100%, it is possible that a certain amount of hydroxyl ions passed to the anolyte from the alkaline catholyte, therefore raising the pH. Another possibility is that the pH increase is associated with a lead dioxide reaction. This would explain why a potential difference was detected without an applied current (Run Table 2-3) and the lead con centration was building up during the run. In other words, the electrode may have behaved as in the lead battery where the Pb02 reduces in the presence of sulphate ions generating a flow of current (Reaction R17). In order to ensure that the different pH changes observed were associated with the applied current, during Run 2-4 the current was suddenly changed from 3 to 10 A, after 90 min and again the pH dropped from 11.8 to 3.8 during the next 15 min, following the same behaviour observed in Runs 2-2 and 2-5 (Fig. 13). The effect of pH-current changes on the rate of phenol oxidation is shown in Fig. 14. When comparing the % T.O.C. removed in Runs 2-2, 2-4, and 2-5, it is observed that the three curves tended to a 17% T.O.C. removed in 90 min, but when the current was changed in Run 2-4 from 3 to 10 A, a higher percent of the carbon was oxidized in the interval from 90 to 120 min (Fig. 13). The most probable reason for this difference was the higher pH observed at 90 min in Run 2-4. This was the first indication of an enhancement of T.O.C. removal at high pH. To investigate if there was any difference in the pH response when a different cation membrane was used, experiment 2-8 was carried out. Using the NAFION-127 membrane, starting at pH = 12 and 10 A the pH dropped even faster than in Run 2-2 (Fig. 15). After 15 min the pH had dropped from 12 to 2. In Run 2-10 the pH was kept in the basic range for a longer period 59 Fig. 14. Effect of current on % phenol oxidation at initial pH = with IONAC MC-3470 membrane. 9.4 60 61 of time, by starting at a pH of 12.8. As shown in Fig. 16 the pH started to drop after 45 min, and went from 12.2 to a value of 2.2 in the next 15 minutes. In the same figure, it can also be observed that the oxidation of phenol is severely limited by a high pH. At 15 min and pH = 12.7 (Run 2-10) only a 10% of the phenol had been oxidized, whereas at pH - 2, 65% of the phenol had already been oxidized after the same time (Run 2-9). On the other hand, the curve for % T.O.C. removed vs time corresponding to the high pH run is above the % T.O.C. curve corresponding to the low pH run. It should be noted that the pH is measured at the outlet of the cell, and when the pH drop is recorded there must be a pH profile within the cell (higher pH at the entrance), and a variable pH in the recircu lation tank. This makes the interpretation of the results more diffi cult, but still there is a definite enhancement in T.O.C. removal at high pH. These experiments indicated that an optimum pH sequence would be a low pH' at the beginning of the run, favouring phenol oxidation, and a high pH at the end, favouring T.O.C. removal. This idea suggested that an anion selective membrane could provide precisely such an optimum pH sequence. If the transport of hydroxyl ions from the NaOH catholyte, and through the membrane were sufficient to produce a pH increase from about 2 to 12, during the course of a run, at a given current, an optimum type of run could be possible. With this objective, several preliminary tests were performed with the anionic membrane IONAC MA-3475. In Fig. 17, Run 2-11, it is observed that the desired increase in pH indeed occurs, and at 30 minutes 90% of 62 63 64 the phenol had been oxidized, before the pH started to increase. When the pH increase was produced, the T.O.C. started to be oxidized at a higher rate than in Run 2-10. It should be noted that in the high pH range the oxidized organic carbon remains in solution in the form of inorganic carbon (e.g., Run 2-11) probably carbonates, due to the higher solubility of carbon dioxide in alkaline solutions. Particularly interesting colour changes were observed during these experiments. The electrolyte at the outlet of the cell showed a brown-reddish colour when the pH was alkaline (higher than 9.4), and in those runs where the pH dropped spontaneously, the colour changed to light yellow. For example, in Pom 2-10 the electrolyte took a deep brown-reddish colour until 60 min and when the pH drop was produced, the outlet showed the light yellow colour while tank was still brown. These colour reactions, when the pH was changed, were also observed in a pure benzo-quinone solution, prepared for comparison. 5.3 Effect of current using the divided cell In experiments 2-3, 2-4, and 2-5, current and pH were dependent variables and therefore the unique effect of the current can not be analyzed separately. However, in Fig. 13 it can be seen that the runs at 6 and 10 A showed approximately similar pH drops and therefore the current effect can be compared. The % T.O.C. removed vs^ time curves for those two runs are practically coincidental. From Fig. 14 it is observed that at 10 A and 15 min, 12% more of the phenol had been oxidized than at 6 A. At 10 A, total phenol conversion was achieved in 90 min, but at 6 A, 120 min were necessary to oxidize the 65 phenol completely. Two runs were carried out with the cationic membrane MC-3470 to study the effect of the current without significant pH changes (Runs 2-6, 2-7). The results are plotted in Fig. 18, where it is observed that the % phenol oxidized vs time curves are very close and total phenol conver sion is achieved at 75 min for both currents. However, the T.O.C. analyses revealed that at 20 A, 47% of the carbon had been oxidized to carbon dioxide in 120 min, whereas at 10 A only a 12% had been oxidized during the same time. 5.4 Comparisons of membrane performances The performances of cationic membranes IONAC MC-3470 and NAFION 127 can be compared from Runs 2-6 and 2-9, both at 20 A and pH = 2. The results are plotted in Fig. 19. The % phenol oxidized vs time curves showed that when using the IONAC MC-3470, about 77% of the phenol had been oxidized in 15 min versus 65% when using NAFI0N-127. But at 75 min total phenol conversion had been achieved with both membranes. The % T.O.C. removed vs time curves are very close but not exactly coincidental. At 120 min, 10% more carbon was oxidized in the run with IONAC MC-3470. The anionic membrane MA-3475 gave the best performance as far as T.O.C. removal is concerned, because it provides the favourable pH increase as described previously (Fig. 17). In 15 min, phenol oxidation was 70%, and total phenol oxidation was also achieved in 75 min. How ever, the membrane was found to be changed in colour from yellow to brown, when the cell was opened after Run 2-11. The actual reason for this is unknown, but it is possible that the polymer structure of the membrane experienced oxidation, or that it interacted in some way with the phenol 66 Fig. 18. Current effect on % T.O.C. and % phenol oxidation at pH = 2.5 with IONAC MC-3470. 67 Fig. 19. Type of cationic membrane effect on % T.O.C. and % phenol oxidation at 20 A and pH = 2.5. oxidation products. These changes may affect the lifetime of the mem brane. The cationic membrane MC-3142 was only tested during the preliminary experiments with anodized lead. On several occasions the membrane was perforated by protruding particles, which resulted in short circuits. For this reason, a plastic screen was introduced between the membrane and the particles. Another disadvantage of the IONAC MC-3142 membrane is that its nylon support will deteriorate with use in NaOH solutions. In terms of mechanical resistance, the IONAC membranes MC3470 and MA3475 were stronger than the NAFION-127 and IONAC MC3142. Problems of breaking never occurred with the first two owing to their different sup port materials and thicknesses (Appendix 1). The polypropylene screen, used in Run 2-1, produces much higher potential drops than the saran cloth used in Run 2-2, probably due to its closely packed screen structure. These experiments show that the use of the different plastic screens does not affect the results as far as phenol and T.O.C. removal is concerned. 5.5 Effect of pH using the undivided cell From the pH vs time curve in Fig. 20, it can be seen that when working with only one chamber the pH does not drop as fast as in the divided cell. In Run 3-2 the pH only dropped from 12 to 11.5 in two hours whereas in corresponding Run 2-8 (Fig. 15 ) the pH dropped from 12 to 2 in 15 min. The reason for this different pH behaviour is that in the case of the undivided cell, the reaction of hydrogen evolution taking place at the cathode changes the hydroxyl ion balance in the electrolyte. When starting at pH = 9.5 the pH dropped to 3.8 which also repre sents a smaller drop than in the case of the divided cell, and when starting at pH = 2.5, the pH was held practically constant. As was found in experiments with the divided cell, the %•phenol oxidized vs time curves at 10 A for the undivided cell (Fig. 20), show that phenol is oxidized much faster at a low pH than at a high pH. In 15 min, 70% of the phenol was oxidized when the pH was 2.5, compared with only a 33% when the pH was 12. On the other hand, more T.O.C. was oxidized at the high pH. At 120 min, 19% of the carbon was oxidized at pH = 2.5, whereas at pH = 12, a 32% of the carbon had been oxidized but remained in solution as inorganic carbon (e.g., Run 3-2). The pH effect was also studied when working at 20 and 30 A (Figs. 21 and 22) and the results showed the same tendency. Phenol oxidation is favoured by a low pH whereas the further oxidation of intermediates is improved at a high pH. When comparing Figs. 20, 21 and 22 it is also observed that as the current increases, the pH has a greater effect on T.O.C. oxidation than on phenol oxidation. 5.6 Effect of current using the undivided cell Figure 23 describes the effect of 10, 20, and 30 A currents at con stant low pH (2.5), and Fig. 24 at a high pH (12). It should be noted that during the runs at 10 A, the temperature of the electrolyte remains constant at room temperature. At 20 A slight heating occurs, and a 4°C temperature increase can be detected in the recirculation tank after 120 min operation. When working at 30 A, approximately a 12°C tempera ture increase is registered (e.g., Runs 3-6 and 3-7). These temperature increases are small and the side effects due to temperature variations 70 Fig. 20. pH effect on % T.O.C. and % phenol oxidation at 10 A in an undivided cell. Fig 21. pH effect on % T.O.C. and % phenol oxidation at 20 A in an undivided cell. I JL 100 r-20 KEY RUN N2 PH O X 3-6 3-7 2.5 12.0 MASS TRANSFER -CONTROLLED REGION 1 1 1 i 30 60 90 120 TIME (min) Fig. 22. Effect of pH on % T.O.C. and % phenol oxidation at 30 A in an undivided cell. 74 100 100 O 80 UJ N Q X O 60 UJ JO. CL 40 20 0 0 120 —a KEY RUN N2 I (A) • 3-2 10 3-5 20 O 3-7 30 _L JL 30 24. 60 90 TIME (min) Effect of current on % T.O.C. and % phenol oxidation at initial pH = 12, in an undivided cell. 120 75 can probably be neglected. In Fig. 23 and Fig. 24 the effect of increasing current is to sub stantially raise the initial rate of phenol oxidation, thus decreasing the time to complete oxidation and to increase the % of T.O.C. removed. These effects are not proportional to the current since there is a bigger increase in T.O.C, removal and in the rate of phenol oxidation in going from 10 A to 20 A than in going from 20 A to 30 A (i.e., Fig. 24). It can also be observed that in Fig. 23 the % phenol oxidized vs time curves are closer together than in Fig. 24, which indicates that at high pH the current has a greater effect on phenol oxidation than at low pH. The maximum % T.O.C* oxidation is encountered in Run 3-7, for the highest current and alkaline pH, where 92% of the carbon was oxidized in 120 minutes, 5.7 Comparisons of divided and undivided cells It is possible to evaluate the performances of divided and undivided cells under similar conditions. Table 5 shows the operating condition and results of three comparisons. Cationic membranes IONAC MC-3470 and NAFION 127 gave similar per formances to the undivided cells in terms of phenol oxidation. Con sidering that 2% error arises in the determinations of.phenol and T.O.C. .... it can be said that the results seem to favour the undivided cell slightly in terms of T.O.C. removal. Run 2-11, with anionic membrane MA-3475, can be either compared with Run 3-4 or with Run 3-5. Because in Run 2-11 the pH increased from 2.4 to 12, higher % T.O.C. oxidation was achieved than in Run 3-4, but 76 TABLE 5 COMPARISONS OF DIVIDED AND UNDIVIDED CELLS Run No. % Phenol ox. % T.O.C. ox. Conditions (cell configuration) at 15 min at 120 min 1(A) pH Run 2-7 70 11 10 2.5 (IONAC MC-3470) Run 3-3 70 19 10 2.5 (undivided) Run 2-9 77 47 20 2.5 (NAFION 127) Run 3-4 75 53 20 2.5 (undivided) Run 2-11 72 67 20 2.4 (IONAC MA 3475) to 12 Run 3-5 53 73 20 12 (undivided) similar % phenol oxidation. When comparing Run 2-11 with Run 3-5 (at a constant pH of 12) it is observed that higher T.O.C. oxidation was achieved in Run 3-5, even if the % phenol oxidized at 15 min was lower due to the higher pH. On the basis of these results and because the undivided cell is simpler to operate, it was decided to perform the rest of the experi ments with the undivided cell. 5.8 Effect of conductivity of the electrolyte Figure 25 shows the effect of electrolyte conductivity when working at 20 A. At a pH of 12, in Runs 3-5 and 3-11, the electrolyte conduc--3 -3 -1 tivity was changed from 8 * 10 to 32 x 10 (ft.cm) , by adding 5 g/1 and 30 g/1 of Na2S0i+, respectively. The results showed that the increase in conductivity did not affect the phenol oxidation, nor the T.O.C. removal. Both experiments gave almost perfectly coincidental curves of % phenol oxidized vs time and of % T.O.C. vs time. The same result was obtained when the conductivity was changed, at pH = 2.5 in Runs 3-4 and 3-10. However, when working at 10 A, and pH = 12, the same increase in conductivity produced an extra 20% carbon oxidized after 120 min, but the % phenol oxidized vs time curve remained practically unchanged (Fig. 26). Similarly, at 10 A, and pH = 2.5 (Fig. 27) about 8% more carbon was -3 oxidized after 120 min when conductivity was changed from 8.4 x 10 to -3 -1 30 x 10 (fi.cm) , but again, the % phenol oxidized vs time curve showed no variation. An explanation can be proposed for these observations. An increase in conductivity of the electrolyte produces a lower solution potential, and therefore a higher electrode potential. At 10 A, such an increase in electrode potential results in an appreciable change in the rate of further oxidation of intermediates to CO2. But at 20 A, the metal potential is much higher so that the decrease in solution potential did not change the net electrode potential appreciably, and therefore no change is detected in the % of T.O.C. removed. 5.9 Effect of initial phenol concentration Figure 28 represents the % phenol oxidized vs time for three differ ent initial concentrations of phenol (Runs 3-3, 3-12, and 3-13). As the initial phenol concentration was increased from 93 mg/1 to 525 mg/1 to 1100 mg/1, the % phenol oxidized after 15 min decreased from 70% to 38% and to 27%. However, in Fig. 29 it is possible to see that a given 79 Fig. 26. Effect of electrolyte conductivity at 10 A and initial pH = 12. 80 Fig. 27. Effect of electrolyte conductivity at 10 A and initial pH - 2.5 81 100 Q UJ N X o UJ X 0_ 60 TIME (min) Fig. 28. % phenol oxidized vs time for various initial phenol concentration at 10 A and pH = 2.5. N5 VO PHENOL CONCENTRATION (mg/L ) o — "2. m 3 CO ho time the net amount of phenol oxidized is higher as the concentration increases. Thus, 350 mg/1 of phenol were oxidized in 15 min in Run 3-13 compared to 200 mg/1 in Run 3-12 and only 75 mg/1 in Run 3-3. In this figure it is easier to observe the parallelism between the three curves. For example, if the curve for Run 3-12 was displaced along the time axis until it matched the curve corresponding to Run 3-13, both curves would be practically coincidental. In other words, after' 30 min of Run 3-13, the same results would be observed as in Run 3-12 from time zero. As these runs were performed at 10 A and pH = 2.5, the T.O.C. removal is relatively low. When the initial concentration was 93 mg/1, 25% of the carbon was removed in 2 h versus 11% when the initial concen tration was 525 mg/1. But for the initial concentration of 1100 mg/1 of phenol, the net change in T.O.C. was practically undetectable due to the higher concentration of carbon present in solution. 5.10 Effect of electrolyte flow rate Run 3-15 was performed at a flow rate of 0.55 £/min. A comparison of the results with those from Run 3-4 where the flow, was 1.12 £/min, under otherwise equal operating conditions, shows that at both flow rates practically the same % phenol vs time and % T.O.C. oxidized vs time were obtained. Thus, phenol was 75% oxidized after 15 min in Run 3-4 and about 73% in Run 3-15. The % T.O.C. oxidized after 120 min were 53% and 54% respectively. (Refer to Run tables in Appendix 2.) The effect of the flow rate is masked by the presence of the rel atively large recirculation tank. In a single pass experiment, a higher conversion would be expected at a lower flow rate. With re circulation, a lower flow rate means it takes more time for a given 84 effect to be detected in the recirculation tank analysis. The effect of both inlet concentration and flow rate, is more easily examined in single pass experiments where the cell operates in steady state. Thus some single pass experiments were carried out. Figure 30 represents the % phenol oxidized in a single pass through the cell vs flow rate, at different initial concentrations, when operat ing at 10 A and pH = 2.5. Runs 4-1, 4-2, and 4-3 were carried out at essentially the same phenol concentration of 100 ± 5 mg/1 to check reproducibility. They produced practically coincidental % phenol oxi dized vs flow rate curves. About 90% of the phenol was oxidized at a flow rate of 0.11 £/min, and as the flow rate was increased the % of phenol oxidized dropped, reaching a 20% at a flow rate of 1.1 2,/.mini Run 4-4 was performed under the same conditions (pH = 2.5, I = 10 A) but starting at the higher phenol concentration of 580 mg/1. The phenol analysis showed that about 70% was oxidized when the liquid flow rate was 0.11 Ji/min and that the % phenol oxidized decreased with increasing flow rate, to about 8% when the flow rate reached 1.1 2,/min. Similar effects were observed when operating at a current of 20 A, for initial concentrations of 110 and 510 mg/1. The table for Run 4-1 shows that when the flow rate was 0.11 £/min at 10 A, the temperature of the electrolyte was raised from 24°C at the inlet to 32°C at the outlet of the cell. But when the flow was increased to 0.25 £/min the outlet temperature increased only 4°C above the inlet temperature. The operation can be considered practically isothermal for all flows equal or higher than 0.25 il/min. At 20 A (i.e., Run 4-5) the temperature at the outlet increased more than at 10 A, as could be expected, but the operation can be 100 80 Q UJ U O 60 X o o Z LU X 0_ 40 20 0 J. (a) KEY RUN N2 PHENOL C0 (mg/l) O 4 - 1 105 + 4-2 95 4-3 100 • 4-4 580 0.2 0.4 0.6 0.8 1.0 FLOWRATE (L/min) 1.2 1.4 100 80 Q LU N 9 60 X o LU 40 X 0_ 20 0 T 02 1 KEY RUN N2 PHENOL C0 (mg/L) 4-5 110 • 4-6 510 0.4 30. 06 0.8 1.0 FLOWRATE (L/min) Effect of flow rate on the single pass % phenol oxidation at (a) 10 A, (b) 20 A. 1.2 1.4 86 considered practically isothermal for flows above or equal to 0.55 £/min. In retrospect, it would have been better to report all the results under constant room temperature to eliminate any possible side effect due to temperature variations. However, this would require the design of a totally different cell with a built-in heat exchanger to remove the heat released by the current. 5.11 Effect of particle size Figure 31 shows the % of phenol oxidized vs flow using three differ ent anode surface areas. The lower curve corresponds to experiment 4-8, where no particles were present and the anode was just the lead dioxide on graphite plate. The intermediate curve corresponds to the average % phenol oxidized from Runs 4-1, 4-2, and 4-3 where a particle size between 1.7 and 2.0 mm was used and the upper curve was obtained when working with particle sizes between 0.7 and 1.1 mm. Taking as a reference, the flow of 0.4 A/min, where the operation is practically isothermal, it is observed that when working without particles the specific area was 3.3 cm ^ and only 20% of the phenol was oxidized, whereas for a specific area of 20.6 cm ^ using the larger particles, about a 59% oxidation was achieved. With the smaller particles (sizes = 0.7-1.1 mm) about a 67% of the phenol was oxidized and the specific area was 41.5 cm ^. (Calculations of the specific electrode areas are found in Appendix 4.) These effects will be discussed in the next section based on mathe matical models for the process. 87 0 0.2 0.4 06 0.8 1.0 1.2 1.4 FLOWRATE (I/min) Fig. 31. Effect of anode surface area-particle size on the % phenol oxidized in a single-pass vs flow rate. 88 5.12 Comparisons of experimental results with mathematical models 5.12.1 Batch experiments If the rate of phenol disappearance is controlled by mass transfer of phenol from the bulk electrolyte to the surface of the electrode, then for a batch recirculation system, the following equation relates the phenol fractional conversion with dimensionless time and mass transfer groups (Appendix 3). X = 1 - exp (exp -K a L m K a L - 1)— - m t u m Here, t is the time the electrolyte spends in the mixing tank, a the specific surface area of the electrode, L the electrode height, and u the superficial electrolyte velocity. The model calculations are represented as a band or region of con version vs time, taking into account that a 10% error can be expected from the empirical equation for the mass transfer coefficient. As well, the correlation chosen may underestimate the mass transfer coefficient for this particular case, as discussed in Appendix 3, page 160. Calcu lations of X vs t are presented in Appendix 4, Table A-2. Figure 28 shows the deviations of the experimental results from the mass transfer model when operating at 10 A. As expected, the curve cor responding to the lowest initial phenol concentration (93 mg/1) is closest to the mass transfer-controlled region with a 15 min conversion approximately 15% lower than predicted by the mass transfer model. Where the initial concentration of phenol was 1100 mg/1 (Run 3-13) the experimental % phenol oxidized after 15 min is about 55% below that predicted by the mass transfer model. This implies that the electro chemical reaction kinetics controls the rate of phenol oxidation under 89 such conditions. However, as the phenol concentration drops during the course of the experiments, the curves approach that of the mass transfer model. In other words, the process is controlled by reaction kinetics at the beginning and as phenol depletes, mass transfer becomes the con trolling mechanism. Figures 21 and 22 show that as the current is increased at an initial phenol concentration of the order of 100 mg/1 (Runs 3-4, 3-6) the experimental % phenol oxidized vs time curve moves closer to the mass transfer region as might be expected. At 20 A, the experimental curve is still below the mass transfer region until about 75 min when the con version was complete. But at 30 A the mass transfer band encloses the experimental curve from Run 3-6, at 30 min time. For both 20 and 30 A currents, the curve at low pH is closer to the mass transfer model than the curve corresponding to the high pH run. It should be noted that the experimental curves obtained with either the divided or the undivided cell are never above the mass transfer region. The closest experimental curve to the mass transfer model is that of Run 3-6 at 30 A and pH = 2.5. 5.12.2 Continuous experiments For a.single pass through the packed bed reactor where phenol dis appears, by a rate process first order in phenol concentration, the assumption of plug flow yields (Appendix 3) - Jtn(l - X) = K a L/u where X is the fraction of phenol oxidized and K is the overall rate constant, which can be related to the mass transfer and electrochemical rate constants by using the concept of additive mass transfer and electro chemical resistances as, 90 Using the data from those experiments performed in the continuous mode, it is possible to evaluate an experimental rate constant. When - &n(l - X) is plotted vs -jj a straight line is obtained (Fig. 32) and the experimental rate constant K can be determined from the slope. Thus, using available correlations for K , K can be determined. m r The nature of the electrochemical rate constant K was alluded to in r Chapter 2. Unlike a chemical reaction rate constant, it is dependent on electrode potential. As discussed in Appendix 3, when operating at a fixed current rather than at a fixed potential, changes in phenol concen tration, flow rate, surface area, etc. will result.in changes in potential and hence in K . Therefore, the analysis of results in terms of K to be r 3 r presented here, is of limited usefulness except to indicate what resistance (mass transfer or electrochemical kinetics) is more important under deter mined conditions. However, considerations of the effects of the process variables on K does show qualitative agreement with what is expected from the simple model, as will be shown with the following examples. Taking the average fractional conversion from experiments 4-1, 4-2, 4-3, performed at 10 A and initial concentrations of phenol of 100 ± 5 mg/1, with particle sizes between 1.7 and 2.00 mm, the experimental:rate con stant is obtained from the corresponding straight line represented in Fig. _3 32. The resulting value of K is 3.1 x 10 cm/s. Using the correlation by Pickett and Stanmore (45) for mass transfer coefficient in electro chemical packed bed reactors K and K are calculated at various flows m r (Table A-3). 91 92 For example, at 0.25 Vmin : K = 3.9 x 10~3 cm/s, K = 15.7 x io"3 cm/s, and m ' r ' at 1.10 A/min : K = 8.8 x io"3 cm/s, K = 4.8 x io"3 cm/s. m r This implies that at the low flow rate, mass transfer controls the phenol oxidation because the mass transfer resistance (~~) represents about an m 80% of the overall resistance. On the other hand, at the high flow rate, the resistance to reaction (~) represents a 65% of the overall resis-tance (tr). The same calculation procedure was applied to the data from Run 4-4 when the initial concentration was 580 mg/1, working under otherwise equal conditions as in the previous example. The resulting experimental rate _3 constant was 1.4 x 10 cm/s. In this case the electrochemical reaction resistance was higher than the mass transfer resistance at all flows (Table A-4) -3 -3 i.e., at 0.25 A/min : K = 3.8 x 10 cm/s, K = 2.21 x 10 cm/s m r at 1.10 l/mln : K = 8.8 x io"3 cm/s, K = 1.67 x 10~3 cm/s. mThe results from both examples are in qualitative agreement with the pre dictions of the model, since at a constant initial concentration, K ' r should decrease when the flow rate increases, and at a constant flow rate, K should also decrease when the initial concentration increases as r described in Appendix 3. The experimental rate constant was also evaluated for the smaller particle size (0.7-1.1 mm). Using the results from Run 4-8, the experi--3 mental rate constant is K = 1.8 x 10 cm/sec (Table A-5), and the mass transfer and reaction coefficients at the extreme flows are, at 0.25 £/min : K = 5.3 x 10~3 cm/s, K = 2.68 x 10~3 cm/s m ' r 1.10 A/min : K = 12.1 x io"3 cm/s, K = 2.09 x io"3 cm/s. m ' r 93 These data indicate that the reaction offered a higher resistance than mass transfer. In this case, the initial concentration of phenol was 100 mg/1 and all the experimental conditions except for particle size, were equal to those in the first example. The mass transfer coefficients K are higher at a given flow rate than in the first m example, due to the dependance of the empirical mass transfer coefficient on particle size, but lower reaction coefficients Kr are obtained with the smaller particles. This result is also in qualitative agreement with the theoretical model, since for a higher specific surface area of the bed (smaller particles), the electrode potential will be lower for the same input current applied to the cell, and thus lower values of Kr are expected. Even though a lower value of the overall rate constant K is obtained with the smaller particles, the net effect is that higher conversions are achieved at a given flow rate than with the larger particles (Fig. 31), due to the increase of the specific surface area of the bed, from 20.6 to 41.5 cm \ This is expected from the plug flow equation: K a X = 1 - exp r u It should be emphasized that the values reported here for the reac tion rate constants are particular for the set of conditions used in each experiment, and therefore should not be used in a scale-up, since for a larger cell even with the same particle sizes than used in this study, the average electrode potential (or the average value) will be different for the same net current input, according to the model. Using the data from Run 4-7 where the anode was the electrodeposited Pb02 on graphite feeder plate (without particles), the experimental rate constant obtained from the slope of the straight line on Fig. 32 is _3 K = 5 x 10 cm/sec. For the ranges of flows used in the experiments the Re numbers referred to the hydraulic diameter, de = 2SW/(S + W) are lower than 700. For Re^e < 2000, an applicable correlation for the mass transfer coefficient in a parallel plate reactor is (page 133, Ref. 10) m = 1.47 (Re, Sc^ %)1/3 D v de L S For liquid flow rates between 0.25 and 0.11 Jl/min, this correlation -4 results in mass transfer coefficient values between 4 x 10 and -4 7 x 10 cm/sec which are about ten times lower than the calculated experimental constant. Since the overall rate constant can never be higher than the mass transfer coefficient, it means that the values of the mass transfer coefficient obtained by the correlation are an under estimation of the real situation. It could be suggested that in prac tice there is an enhancement in the mass transfer coefficient due to gas evolution, which is not taken into account by the correlation. Most of the correlations for mass transfer coefficients in gas evolving parallel plate electrodes have been developed for the case of stationary solution where mixing of the electrolyte is only provided by the gas bubbles. It is known that mass transfer coefficients under gas evolution are between four or five times greater than the mass transfer coefficients from conventional free-convection correlations in flat plates (10). But for the case of forced convection and gas evolving electrodes, an empirical correlation could not be found in the literature. However, the results seem to indicate that mass transfer controls the rate of phenol oxidation on the flat plate, since even if the mass transfer coef ficients were increased by a factor of five, the resulting values 95 would still be lower than the experimental rate constant. 5.13 Current efficiencies, energy requirements and energy costs for  phenol oxidation 5.13.1 Batch experiments Typical current efficiencies, energy requirements, and costs are estimated for batch experiments nos. 3-3, 3-12, 3-13, after recirculation times of 15 and 90 min. A sample calculation is given in Appendix 4, page 178. For the estimation of the % C.E. it was assumed that four electrons are transferred from the phenol molecule as proposed by Covitz (Reaction R9). Energy per g mol phenol oxidized is calculated taking $0.02/Kw-h as a basis. Table 6 shows that for lower initial concentrations of phenol or higher recirculation times, the % C.E. are relatively lower and the energy costs are higher. TABLE 6 TYPICAL CURRENT EFFICIENCIES, ENERGY REQUIREMENTS AND ENERGY COSTS IN BATCH EXPERIMENTS WITH UNDIVIDED CELL (Q=1.12 £/min) Run No. \ (mg/1) AV (volts) t (min) % (mg/1) X (%) C.E. (%) Energy rKw-h s ^g moi; Electrical costs ($/g mol) 3-3 93 8.6 15 90 28 0 70 100 11 3 6.7 22.6 0.14 0.45 3-12 525 8.3 15 90 320 5 39 99 36 19 2.5 4.8 0.05 0.10 3-13 1100 8.5 15 90 792 22 28 98 54 35 1.7 2.4 0.03 0.05 96 Run 3-3 at C. =93 mg/1, yields the maximum energy cost of A0 $0.45/g mol oxidized for the 90 min operation, where practically no phenol was present after treatment. 5.13.2 Continuous experiments Using the data from Runs 4-3 and 4-4, it is possible to estimate % C.E. and energy costs for various flows in a single pass through the cell. A sample calculation is given in Appendix 4, page 180, and the results are presented in Table 7. TABLE 7 TYPICAL CURRENT EFFICIENCIES, ENERGY REQUIREMENTS AND ENERGY COSTS IN CONTINUOUS EXPERIMENTS WITH UNDIVIDED CELL C C Energy Electrical Run 1 Q AV 2 X C.E. rKw-h . costs No. (mg/1) (A/min) (V) (mg/1) (%) (%) ^g mol'' ($/g mol) 4-3 95 0.25 8.7 39 59 9.6 9.7 0.195 0.55 8.3 63 34 12.2 7.4 0.148 1.1 7.4 76 20 14.3 5.6 0.110 4-4 580 0.25 8.6 365 37 36.7 2.5 0.050 0.55 7.9 470 19 41.5 2.0 0.041 1.10 7.0 526 9 39.3 1.9 0.037 The electrical cost of treatment decreases as the initial phenol concentration and electrolyte flows are increased. Comparing the results from Run 3-12 after 15 min (Table 6) with those from Run 4-4 at 0.25 5,/min, it can be seen that in terms of current efficiences and energy requirements, both situations are practically equivalent. 5.13.3 Cost comparisons It is of interest to compare the estimated electrical costs with the operating costs for other treatments given by Katzer (8). Table 8 97 TABLE 8 OPERATING COSTS OF VARIOUS TREATMENT METHODS, ESTIMATED FOR 1974 FOR A CATALYTIC CRACKER EFFLUENT CONTAINING 700 mg/1 PHENOL (8) Treatment costs Process ($/1000 gal) ($/g mol)(5) Oxidation pond^ 0.14-0.51 0.005-0.018 Activated sludge^ 0.24 0.009 Activated sludge, with dilution { } 2.4 0.090 (3) Carbon adsorption 0.86 0.031 (4) Catalytic oxidation ' 0.57 0.020 Electrochemical (JJ) oxidation 2.00U/) 0.08^ ' (1) $0.38/103 gal in 1967; operating costs of biological oxidation ponds at Billing Montana Oil Refinery; extrapolation to 1974 gives $0.51/103 gal. Capital investments not included. (2) Extrapolated from 1968 to 1974 using a factor of 1.34; if dilution of effluent is required cost of treatment will increase; 10:1 dilution factor results in $2.4/103 gal. 106 gal/day design for treating unfiltered activated sludge plant effluent to produce water with 8 mg/1 C.O.D. (4) Calculated for a 99% phenol removal by catalytic oxidation. ^"^Calculated from the original paper (Ref. 8) assuming that total phenol conversion was achieved in all the examples. ^Calculated in this study for treating a 51 volume with 99% removal (Appendix 4). Calculated in this study by interpolating from Table 6 for 700 mg/1 phenol initial cone, using a factor of $0.02/Kw-h for electrical energy cost. 98 shows the 1974 costs for treating a catalytic cracker effluent containing 700 mg/1 phenol. For the electrochemical process, electrical energy costs are interpolated from Table 6 using the data from Run 3-12 (525 mg/1 phenol) and Run 3-13 (1100 mg/1 phenol) at 90 min, when phenol con versions are 98-99%. To treat an effluent with 700 mg/1 phenol, approximately $0.08/g mol phenol would be necessary. In Appendix 4 is estimated the electrical cost per volume of waste in about $2.00/103 gal, to compare with the data by Katzer. Electrical energy cost is about ten times higher than the activated sludge cost in normal conditions, but of the order of the activated sludge cost if dilution is required. The cost of other processes appear to be lower than the estimated electrical cost; for example, carbon adsorption operating costs would be less than a half ($0.03/g mol). It should be noted that the electrical cost ($0.08/g mol) has been estimated for the arbitrary cell characteristics and operating conditions used in the experiments. To draw firm conclusions about feasibility of the process would require an optimization of operating costs and a capital cost estimate of this and the other processes. Some recommen dations for the former are given in Chapter 7. CHAPTER 6 CONCLUSIONS An investigation was made of the electrochemical oxidation of phenol on lead dioxide packed bed anodes. In all the experiments performed, phenol oxidation occurred more easily than the further oxidation of inter mediate organics to carbon dioxide or carbonates. The conclusions of this study are summarized below. 1. Electrodeposited lead dioxide was found to be a better anode than the anodized lead shot in terms of phenol oxidation and corrosion resistance. 2. The oxidation of phenol was more rapid under acidic conditions but the removal of oxidized products (measured by the total organic carbon) was favoured by alkaline conditions. In divided cells, operated with recirculation of solution, the electrolyte pH depended on the type of ion selective membrane used. An anionic membrane which provided a pH increase from acidic to alkaline proved to be superior to a cationic membrane in terms of T.O.C. removal. 3. The rates of phenol oxidation in divided and undivided cells were similar. In terms of T.O.C. removal no improvement was obtained with the divided cell even under optimum pH controlled conditions provided by the anionic membrane. 4. The extents of phenol and T.O.C. oxidation increased with applied current density at high or low pH. But at high pH, current density 99 changes affected the rate of phenol oxidation more strongly than at low pH. -3 -3 5. An increase in electrolyte conductivity from 8 x 10 to 32 x 10 (ft.cm) ^ had no effect on the rates of phenol or T.O.C. oxidation at high or low pH at c.d. = 1052.6 A/m2, but at 526.3 A/m2 c.d., the same increase in conductivity produced higher T.O.C. oxidation rates, even though the phenol oxidation rates remained constant. 6. The effect of increasing the initial phenol concentration in a single pass was to reduce the % phenol oxidized at a given time or in a single pass. However, the current efficiency for phenol oxidation increased. 7. Increasing electrolyte flow rate reduced the single pass phenol conversion at a given inlet phenol concentration in continuous experi ments. 8. Increasing the specific surface area of the anode in the range of 3.3 to 41.5 cm2/cm3 produced higher single-pass phenol conversions at a given flow rate and inlet phenol concentration of the electrolyte. 9. Comparisons of the experimental results from batch experiments with the mass transfer model indicated that the oxidation of phenol is . ..-• controlled by the electrochemical reaction at the beginning and as phenol depletes, the experimental % phenol oxidized vs time approach that of the mass transfer model. Continuous experiments showed that the electrochemical reaction resistance becomes more important at higher electrolyte flow rates, higher inlet phenol concentrations and smaller particle sizes. 10. Electric energy costs estimated were generally higher than operating costs of other processes. However, power costs can likely be reduced with further work. CHAPTER 7 RECOMMENDATIONS Investigation of other factors regarding the electrochemical oxida tion of phenol could possibly raise the prospective feasibility of the process. Improvements to the model would lend confidence to the feas ibility calculations. On the basis of this study, the following recommendations are made, to yield a more complete knowledge of the process. 1. Analysis of the phenol oxidation products: Routine analysis of the phenol oxidation products, (benzoquinone, hydroquinone and possibly maleic acid) should be attempted, to draw definite conclusions about effluent quality under different operating conditions. The T.O.C. analysis would permit one to determine what fraction of the carbon is in the form of each compound. 2. Measurements of electrode potential: Reference electrodes should be located at certain points on the bed, to determine electrode potential variations. It would be of inter est to determine the effect of electrode potential on the kind of oxidation products under different operating conditions. Also electrode potential measurements are of interest for the process modelling. 3. Control of the electrolyte temperature: A heat exchanger built in the electrolytic cell would permit 101 102 temperature control within the cell in single pass experiments. However, the construction of a cell equipped with such a heat exchanger might be difficult. 4. Analysis of the gases produced during electrolysis: A gas analysis technique could be assessed for complete monitoring of the process. It would permit a total carbon balance. Also a measurement of the rate of gas production would be of interest to include the effect of gas evolution in the process modeling. After these experimental improvements were assessed, the effect of some of the other important factors on the electrochemical oxidation of phenol might be studied, e.g., the effect of different anode materials on the kind of oxidation products, and the effect of temperature and elec trode potential on the rates of oxidation of phenol and intermediates. The following recommendations would allow improvements to the mathematical modeling of the process in addition to furthering general under s t and ing. 5. The effect of gas evolution on the mass transfer coefficient should be studied. Another electrochemical reaction that is purely mass transfer controlled could be used for this purpose (45). 6. Cell length, width, and thickness of the bed should be varied. Measurements of electrode potential at different points on the bed would permit one to determine under which conditions the assumption of a uniform or average electrode potential is reasonable. 7. Experiments could be set up to determine the relationship between electrode potential and electrochemical rate constant under different operating conditions. 8. The effect of gas evolution and electrolyte conductivity on the 103 electrode potential could be studied to include these effects in the mathematical models. 9. Computer methods could be used to correlate all the variables affecting the process, and determine costs of energy under different operating conditions to optimize operating costs. 104 NOMENCLATURE Typical units a specific surface area of the bed cm2/cm2 a.,b. constants of the Tafel equation for the J ~* reaction j V cd. current density referred to the surface area of the feeder plate A/m2 C.E. current efficiency C. initial phenol concentration in batch experiments mg/1 C^ inlet phenol concentration mg/C. outlet phenol concentration mg/1 A2 C. phenol concentration in the bulk of solution mg/1 % C^ phenol concentration at the surface of the s electrode mg/C concentration associated with a water w /-i electrolysis reaction mg/1 dp average particle diameter mm D diffusivity of phenol in water cm2/s F Faraday's constant coul/g equiv. i^ local current density for phenol oxidation A/m ig local current density carried by the solution A/m i local current density carried by the metal A/ m m 2 i^ average current density for phenol oxidation A/m iw average current density for water electrolysis A/m2 i total average current density A/m2 I applied current A electrical conductivity of the anolyte electrical conductivity of the catholyte electrical conductivity of the metal electrical conductivity of the solution electrochemical reaction rate constant mass transfer coefficient overall or experimental rate constant length of the cell pressure electrolyte flow rate universal gas constant Reynolds number Schmidt number thickness of the bed (in the direction of current) temperature time of electrolysis dimensionless time residence time in the mixing tank superficial velocity anode potential cathode potential reversible equilibrium potential for the reaction j standard reduction potential for the reaction j total voltage drop through the cell volume of the mixing tank 106 W width of the bed cm X phenol fractional conversion y variable length of the bed cm z number of electrons associated with phenol oxidation Greek letters a transfer coefficient E voidage of the bed rij overpotential for the reaction j E, shape factor for the particles 9 K a L/u dimensionless mass transfer group m v sa mc sc kinematic viscosity of water (cm2/s) cf> metal potential of the anode V ma solution potential of the anolyte V metal potential of the cathodesolution potential of the catholyte V 107 BIBLIOGRAPHY 1. Throop, M. W. "Alternative methods of phenol waste water control." Journal of Hazardous Materials 1, 319 (1975/77). 2. Siegerman, H. "Applications of electrochemistry to environmental problems." Chem. Tech. 675 (Nov. 1971). 3. Lanouette, K. "Treatment of phenolic wastes." Chem. Eng. 84_, 99 (Oct. 1977). 4. Hardsity, D. M., and Bishop, H. E. "Waste water treatment experi ence at organic chemical plants using a pure oxygen system." A.I. Ch. E. Symposium Series 73., No. 167, 140 (1976). 5. Dunlap, R., and McMichael, F. C. "Reducing coke plant effluent." Env. Sci. & Tech. 10, No. 7, 654 (1976). 6. Kostenbader, P. D., and Flecksteiner, J. W. "Biological oxidation of coke plant weak ammonia liquor." J. Water Pollution Control Federation 41, No. 2, (Feb. 1969). 7. Kazuo, S., Kazuto, T., and Satoru, T. "Degradation of aqueous phenol solutions by gamma irradiation." Am. Chem. Soc. 12_, No. 9, 1043 (1978). 8. Katzer, J., Sadana, A., and Ficke, H. "Aqueous phase catalytic oxidation as a waste water treatment technique." Eng. Ext. Series 145, 29 (1974). 9. Zeff, J. "Uv. ox. process for the effective removal of organics in waste water." A.I. Ch. E. Symposium Series _73» No. 167, 206 (1976). 10. Pickett, D. J. Electrochemical reactor design. Elsevier, Amsterdam (1977). 11. Allen, M. J. Organic electrode processes. Reinhold, New York, (1958). 12. Sherwood, T. K., Pigford, R. L., and Wilke, C. R. Mass Transfer. McGraw Hill, New York (1975). 13. Fichter, F. "Electrochemical oxidation of aromatic hydrocarbons." Trans. Amer. Electrochem. Soc. 45_, 107 (1924). 14. Fichter, F., and Stocker, R. "Electrochemical oxidation of aromatic hydrocarbons and phenols." Chem. Abstr. j?, 3037 (1914). 108 15. Fichter, F. , and Brunner, E. "New products of the electrochemical oxidation of phenol." Chem. Abstr. 10, 2873 (1916). 16. Clarke, J. S., Ehigamusoe, R. E., and Kuhn, A. T. "The anodic oxidation of benzene, toluene, and anisole." Electroanalytical Chem. and Interfacial Electrochem. TQ, No. 3, 336 (1976). 17. Gladisheva, A. I., and Lavrenchuck, V. I. "Electrochemical oxidation of phenol," (original title in Russian). Uch. Zap. Tsent. Nauch. Issled. Inst. Olovyan. Prom, No. 1, 68 (1966); Chem. Abstr. 67, 28639X (1967). 18. Tarjanji, M., et al. "Decreasing the phenolic content of liquids by an electrochemical technique." U.S. Pat. 3,730,864. (May 1, 1973). 19. Covitz, F. "Electrochemical oxidation of phenol." U.S. Pat. 3,509,031. (Apr. 28, 1970). 20. Jones, G. C., et al. "Electrolytic oxidation of phenol at lead thalium anodes." U.S. Pat. 4,035,253.(July 12, 1977). 21. Jones, G. C., and Payne, D. A. "Electrochemical oxidation of phenol." U.S. Pat. 3,994,788. (Nov. 30, 1976). 22. Fioshin, M. Y., et al. "Electrochemical oxidation of phenol to quinone." Electrokhimiya 13, No. 3, 381 (1977). 23. Ronlan, A. "Phenols." Encyclopedia of electrochemistry of the elements. M. Dekker, New York, Vol. XI, 242 (1978). 24. Nilsson, A., Ronlan, A., and Parker, V. "Anodic oxidation of phenolic compounds. Journal Chem. Soc. Perkin trans. 1 20, 2337 (1973; Chem. Abstr. 80 115431k (1974). 25. Surfleet, B. "Electrolytic destruction of industrial effluents." The Electricity Council Research Centre. Report No. 204, (Oct. 1969). 26. Rifi, M. R., and Covitz, F. H. Introduction to organic electro chemistry. M. Dekker, New York (1974). 27. Kempf, R. "Electrolytic oxidation of P.Benzoquinone." Chem. Abstr. _5, 2815 (1911). 28. Lure, Y. Y., and Genkim, V. E. "The use of anodic oxidation and cathodic reduction methods for the purification of industrial waste waters." Chem. Abstr. 57, 8371c (1962). 29. Shakarnov, A. V. "Dephenolization of wastes by electrochemical oxidation." Chem. Abstr. 56,13978b (1962). 109 30. Mieluch, J. , et al. "Electrochemical oxidation of phenol compounds in an aqueous solution." Chem. Abstr. 84^, 10275a (1976). 31. Antropov, L. Theoretical electrochemistry. Mir, Moscow (1972), p. 396. 32. Carr, J. P., and Hampson, N. A. "The lead dioxide electrode." Chem. Rev. 72, No. 6, 679 (1972). 33. Ruetschi, P., and Cahan, B. "Electrochemical properties of Pb02 and the anodic corrosion of lead and lead alloys." J. of Electrochem. Soc. 105, No. 7, 369 (1958). 34. Pourbaix, M. "Atlas of electrochemical equilibria in aqueous solutions." Houston, Texas. National Association of Corrosion Engineers, (1974), p. 489. 35. Delahay, P., Pourbaix, M., and Van Rysselberghe, P. "Potential pH diagram of lead and its applications to the study of lead corrosion and to the lead storage battery." J. Electrochem. Soc. 9j5, No. 2, 57 (1951). 36. Kuhn, A. T., and Wright, P. M. Industrial electrochemical processes. North Holland Publ. Co. (1971), p. 525. 37. Burbank, J. J. "Anodization of lead and lead alloys in sulfuric acid." J. of Electrochem. Soc. 104, 693 (1957). 38. Lander, J. J. "Further studies on the anodic corrosion of lead in H2SO4 solutions." J. of Electrochem. Soc. 103, 1 (1956). 39. Ruetschi, P., and Cahan, B. D. "Anodic corrosion and hydrogen and oxygen overvoltage on lead and lead antimony alloys." J. of Electrochem. Soc. 104, No. 7, 406 (1957). 40. Grigger, J. C, Miller, H. C, and Loomis, F. D. "Lead dioxide anode for commercial use." J. Electrochem. Soc. 105, 100 (1958). 41. Gibson, F. D. "Inert lead dioxide anode and process of production." U.S. Pat. 893,823. (April 11, 1962). 42. Standard methods for the examination of water and waste water. American Public Health Association, 12th ed. (1955) pp. 516, 511. 43. Shields, J. R., and Coull, J. "Rate studies in the electrochemical oxidation of phenol." Trans. Am. Electrochem. Soc. 8Q, 113 (1941). 44. Wesley, W. Water quality engineering for practicing engineers. Barnes & Noble, New York (1970), p. 25. 110 45. Pickett, D. J., and Stanmore, B. R. "An experimental study of a single layer packed bed cathode in an electrochemical flow reac tor." J. App. Electrochem. _5, 95 (1957). 46. Grot, W. G. F., Munn, G. E., and Walmsley, P. N. "Perfluorinated ion exchange membranes." 141st meeting of the Electrochemical Soc. Houston, Texas, May 7, 1972. 47. Sedahmed, G. H. "Mass transfer behaviour of gas evolving particu-late-bed electrode." J. of Appl. Electrochem. % 37 (1979). 48. Alkire, R. "Two dimensional current distribution within a packed bed electrochemical flow reactor." J. Electrochem. Soc. 121, No. 1, 95 (1974). 49. Perry, R. H., and Chilton, C. H. Chemical engineers handbook. Fifth edition, McGraw Hill, New York (1973). 50. Bird, R. B., Stewart, W. E., and Lightfoot, E. N. Transport phenomena. John Wiley & Sons, Inc., New York (1960) p. 515. 51. Weast, R. Handbook of chemistry and physics. The Chemical Rubber Co. 53rd edition (1972/73). Ill APPENDIX 1 Specification of Auxiliary Equipment and Materials Power supply Sorenson DCR 40-25B Voltmeter range: 0-40V Am meter range: 0-25A Voltmeter Central Scientific Co., D.C. Voltmeter Scales: 0-1.5 volts 0-15 volts Rotameters a) Anolyte: Brooks, full view indicating rotameter Type: 7-1110 Tube No.: R-7M-25-1 Float: 316 stainless steel Max. flow: 1400 cc/min (s.g. = 1) Scale: 0-100% linear b) Catholyte: Brooks, full view indicating rotameter Type: 8-1110 Tube No.: R-8M-25-2 Float type: 8-RS-8 Max. flow: 1 U.S. G.P.M. (s.g. = 1) Scale: 0-250 mm linear Gas liquid separators 2.5 cm I.D. and 60 cm long glass tube. Liquid outlet located at 40 cm from the bottom. Packed to approximately 20 cm depth with 2 mm diameter glass beads. Filters 3.0 cm I.D. and 15 cm long glass tube filled with glass wool (Merck). Pressure gauges Marsh-type 3-100-SS with 316 stainless steel tube Scale: 0-30 psi (1/4 psi/div). 112 Pumps Barrish Pumps Co., N.Y. Model type: 12A-60-316 Flow data: 21 6.P.H. at 40 psid, 29 6.P.H. at 0 psid. (psid indi cates outlet pressure minus inlet pressure.) Pumps are preset at 45 psid but are adjustable to 65 psid max. pH meter Corning, Model 101 (accuracy ± 0.001 pH) Electrode: Fisher, plastic body-protected bulb type Model No. 13-639-97 Conductivity meter Seibold, Model LTA. Provided with adjustable maximum ranges (from 1 mmho cm--'- to 100 mmho cm--'- maximum) Conductivity cell constant = 0.88 cm--'-Tubing Imperial Eastman "Poly Flo" 66-P-3/8" Valves Whitey, forged body regulating 316 stainless steel. 3/8" connections Fittings Swagelok compression tube fittings 316 stainless steel 3/8" Membranes a) IONAC membranes The IONAC membranes are supplied by Ionac Chemical Sybron Corporation, Birmingham, N.J. Table A-l includes some of the specification sent by the manufacturer. b) NAFI0N 127 membrane Supplied by duPont de Nemours & Co., Delaware. NAFION 127 consists of an homogeneous film of 1200 equivalent weight polymer, 7 mils thick (perfluorosulfonic acid polymer). Supports are made from teflon. Details of membrane properties such as strength, ionic transport, water permeability etc., are available in a publication supplied by the manufacturer (46). 113 TABLE A-l SUMMARY OF TYPICAL PROPERTIES OF IONAC MEMBRANES Anion Exchange Cation Exchange Membranes Membranes Property MC-3142 MC-3470 MA-3475 Electrical Resistance (ohm-cm2, A.C. measure ment) O.IN NaCl l.ON NaCl 14 5 12 6 17 8 % Perselectivity (0.5N NaCl/l.ON NaCl) (0.2N NaCl/O.lN NaCl) 94.1 99.0 96.2 99.0 Water Permeability (ml/hr/ft.2/5 psi) negligible (less than 30) (less than 30) Mullen Burst Strength (minimum psi.) 175 200 200 Membrane Thickness (mils) 7 15 15 Approx. Density Net as shipped oz./yd2 g/m2 6 202 12 405 12 405 Capacity meq/g 1.08 1.22 0.70 Dimensional Stability (ability to rewet after drying) good good good Chemical Stability H2S01+ HCl NaOH Salt up to 60°C up to 5% up to 4% NR* OK at all cone. up to 125°C up to 35% cone. HCl 50% NaOH OK at all cone. up to 125°C Superior to MC-3470 in low and high pH media Size Available nominal sheet sizes (inches 40 x 120 40 x 30 x 120 96 40 x 120 *Not recommended 114 Plastic screens Supplied by Chicopee Manufacturing Co., Georgia Saran type Max. operating temperature = 125°F Chemical resistance: good resistance to acids and most alkalis Style: 6100900 Weight/sq.yd. = 7 oz. Polypropylene type Max. operating temperature: 180°F Chemical resistance: excellent resistance to most acids and alkalis with exception of chlorosulfonic acids and oxidizing agents Style: 60070XX Weight/sq.yd. = 8.7 oz. Cathode chamber packing material Stainless steel 304 - 20 x 20 mesh Analytic equipment and operating conditions specifications a) Gas chromatography specifications  Gas chromatograph Manufacturer: Varian Aerograph Model: 1440 series, single column model Detector: H2 flame ionization detector Chromatographic column Supplier: Western Chromatography Supplies, New Westminster, B.C. Material: glass Dimensions: 2mm I.D., 6.4mm0.D., 6 ft long Packing: 10% SP-2100 ON 100/120 Supelcoport (details of the packing are given in Bulletin 742D by Supelco, Inc.) Operating conditions Injector port temperature 150°C Column temperature 120°C Detector temperature 175°C Carrier gas N2 Carrier gas flow 30 ml/min Air flow 300 ml/min H2 flow 30 ml/min Attenuation setting 4 x 10 Recorder Model: Corning 840 series Response 1 mv full scale Chart speed: 1 cm/min Syringe Supplier: Unimetrics Sample size: 1 uil 115 b) T.O.C. analysis specifications Model: Beckman 915 total organic carron Analyzer: Beckman 865 infrared analyzer Operating conditions Temperature of the total carbon channel 1000°C Temperature of the inorganic carbon channel 150°C Oxygen flow in each channel 250 ml/min Syringe Hamilton with automatic plunger Sample size 50 y£ Recorder Model: Hewlett Packard 7127A Response: 1 mv full scale Chart speed: 1 cm/min c) Atomic absorption specifications Manufacturer: Jarrel Ash, Division of Fisher Sci. Co. Model: 810 Atomic absorption spectrophotometer Lead lamp: Westinghouse hollow lead cathode Wave length: 2170 A Flame: rich (air-acetylene) Recorder Hewlett Packard 7127A Speed: variable Range: 1 or 2 mv full scale Reagents Phenol—liquified reagent Matheson Coleman & Bell Manuf. Chem. Code: PX 511-CB-1040 Assay: min 88% phenol, max 12% H2O NaOH—pellets, reagent American Scientific Chemical Code: SS 350 Assay: NaOH min 98% H2S04—reagent A.C.S. Allied Chemical Code: 001180-009580 Assay: 95.5-96.5% Na2S0it—anhydrous reagent grade American Scientific and Chemical SS 530 Buffers: Fisher Scientific (2 ± 0.02) American Scientific and Chemical (10 ± 0.01) Lead standard solution for atomic absorption: Fisher Scientific, 1000 ppm Water for solutions: single distilled water 116 APPENDIX 2 Experimental Data 1. Characteristics of each group of experiments Group No. 1 a) Mode of operation: Batch experiments using the divided cell and the anodized lead electrode b) Cell description: Cathode: Stainless steel 316 plate and mesh Anode: Fresh lead feeder plate and particles Particles: lead spheres Size: dp = 2 mm Weight: 250 g Volume: "° f— = 22 cm3 (ref 49) 11.337 g/cc Void fraction: e = (57 - 22)/57 = 0.61 Membranes tested: IONAC MC-3142 or IONAC MC-3470 protected by saran screen Group No. 2 a) Mode of operation: Batch experiments using the divided cell and the electrodeposited Pb02 electrode b) Cell description: Cathode: Stainless steel 316 feeder plate and mesh Anode: Pb02 electrodeposited on graphite feeder plate Particles: electrodeposited Pb02 crushed and sized Size: 1.7 < dp < 2.0 mm Weight: 215 g Volume: 23 cm3 Void fraction (e) =0.6 Membranes tested: Cationic membranes: IONAC MC-3470 NAFION 127 Anionic membrane: IONAC MA 3475 In all runs, a saran screen was located between particles and membrane to prevent membrane breaking and short circuit. The only exception was Run 2-1 where a polypropylene screen was used. Group No. 3 a) Mode of operation: Batch experiments using the undivided cell (only one chamber) and the electrodeposited Pb02 electrode 117 b) Cell description: Cathode: stainless steel 316 plate (no packing mesh used to avoid by-passing of the liquid) Anode: feeder and particles are the same as described in group No. 2. Separator: two pieces of saran screen between cathodic plate and Pb02 particles Group No. 4 a) Mode of operation: Continuous experiments (single pass) using the undivided cell and electrodeposited Pb02 anode b) Cell description: The same as in group No. 3 with exceptions of Runs 4-7 and 4-8. In Run 4-7: the anode was the feeder plate only (without particles) In Run 4-8: the particles had the following characteristics: Size: 0.7 < dp < 1.1 mm Weight: 230 g Volume: 24.5 cm Void fraction (e) :(57 - 24.5)7,57 = 0.57 2. Volume, flow pressure and temperature of the electrolytes For groups No. 1 and No. 2 (unless otherwise stated) Electrolyte Volume (£) Flows U/min) P+ (kPa) Anolyte 5 1.12 2.40 Catholyte 5 1.54 2.40 +Pressures may vary ± 20 kPa due to gas evolution. Note: The symbol (*) which appears in each Run table (Groups 1 and 2) indicates that the catholyte pH and electrical conductivity reported were measured on sample diluted by a factor of ten. For group No. 3 the electrolyte volume flow and pressure are the same given for the anolyte in groups No. 1 and 2 with exception of Run 3-15 where data are recorded. In experiments group No. 4 the electrolyte flow was varied and there fore the pressure, too. Data are then reported in each particular Run table. Temperatures All the experiments were set up at room temperature (22 to 24°C). During the batch experiments (groups 1,2,3) at currents below or equal to 10 A the electrolyte temperatures remained constant. But for all those experiments performed at 20 A (i.e., Runs 3-4, 3-5) a temperature increase of 4°C was always detected in the recirculation tank (after the 120 min run). Also, when working at 30 A the electrolyte temper ature rose about 12°C (i.e., Runs 3-6 and 3-7). 118 For the continuous experiments (group No. 4) the temperature variations at a given current are a function of the flow and will be reported in each experiment table. 3. Anodization Before each experiment the anode was treated with 20% H2SO4 at 10 A for 1 hour. (Refer to experimental method section.) Any exception to the standard procedure is given in the particular run table. 119 RUN 1-1 Membrane: IONAC MC-3142 Anodization: Starting from the fresh lead particles for 1 h (at the specified conditions) Electrolytes Approx. Cone. PH Anolyte 1% H2S0i+ • 1.5 Catholyte 10% H2S0i+ 1.0* I = 20 A c .d. = 1052.6 A/m2 t (min) AV (V) T.O.C. (mg/1) % T.O.C. 0 7.3 72 0 15 6.8 68 6 30 6.5 62 14 , 60 5.7 59 17 90 5.7 57 21 120 5.6 52 28 Comments: Particles showed an homogeneous brown colour after the 1 h anodization. *Indicates that the pH was measured on a sample diluted by a factor of 10. 120 RUN 1-2 Membrane: MC-3142 Anodization: Starting from fresh lead particles for 12 h (at the specified conditions) Electrolytes Approx. Cone. pH Anolyte 1% H2SO4 1.5 Catholyte 10% H2S0tj 1.1* I = 20 A cd. = 1052.6 A/m t AV (min) (V) . T.O.C. (mg/1) % T.O.C. (Pb) (mg/1) 0 7.3 80 0 0 15 6.9 75 6 30 6.4 72 11 0.8 60 5.8 66 18 0.7 90 5.6 61 24 0.6 120 5.5 5 30 0.4 Comments: Particles showed homogeneous brown coating (with the same appearance than after Run 2-1) 121 RUN 1-3 Membrane: IONAC MC-3142 Anodization: The bed was originally anodized for 12 h and then was reanodized for 1 h before the run Electrolyt e Cone. pH Anolyte 5 g/1 NaCl 5.4 Catholyte 10% H2S0i+ 1.0* I = 10 A c. d. = 526.3 A/m2 t AV T.O.C. (min) (V) mg/1 (%) 0 10 73 0 15 5 55 24 30 3 52 28 Comments: It was observed that the glass wool filter collected small fragments that had flaked off the electrode, and the solution took a dark grey colour, indicating that the electrode was rapidly dissolving. When the cell was opened the electrode had lost the brown oxide ^ coating showing the underlaying grey lead. The membrane was fouled with deposits. RUNS 1-4 to 1-8 (Corrosion studies on anodized lead) Membrane: IONAC MC3142 Anodization time: 12 h starting from fresh lead, and 1 h prior to each run I = 10 A c.d. = 526.3 A/m2 Catholyte: 25 g/1 NaOH (pH* =12.7 K * = 14 x 10 (ft cm)"1) c Run No. 1-4 1--5 1--6 1--7 1--8 Anolyte 5 g/1 NaOH 5 g/1 Na2S0it 30 g/1 NaaSOtt 30 g/1 Na2S0it 30 g/1 Na2S0it NaOH to adj. pH NaOH to adj. pH NaOH to adj. pH NaOH to adj. pH t (Pb)1 (Pb)2 (Pb) (Pb) (Pb) (Pb) (min) pH (mg/1) (mg/1) pH (mg/1) pH (mg/1) PH (mg/1) pH (mg/1) 0 12.7 140 0.0 9.8 0.0 9.8 0.0 12.0 0.0 7.0 0.0 15 12.6 36 2.7 2.9 1.7 3.2 4.2 3.4 5.0 2.5 3.4 30 12.5 24 1.7 2.7 1.6 2.9 3.5 2.9 5.0 2.2- 3.0 45 12.4 13 1.7 2.6 1.4 2.8 2.7 2.6 4.3 2.0 2.3 60 11.8 3 0.7 2.5 1.3 2.7 2.0 2.5 3.3 1.9 2.2 75 2.4 1.1 2.6 1.7 2.4 3.0 1.8 2.0 90 2.3 0.8 2.5 1.4 2.3 2.3 :The electrolytes were recirculated without a potential applied. 2A potential was applied before the electrolytes entered the cell. *Measured on a sample diluted by a factor of 10. RUN 1-9 Membrane: IONAC MC-3470 I = 10 A cd. = 526.3 A/m2 Electrolytes Concentration K x 103  e _i (n cm) pH Anolyte 5 g/1 N32S01+ NaOH to adjust pH 6.4 9.41 Catholyte 25 g/1 NaOH 16* 12.6* Ke Xl°3 Phenol T.O.C. (Pb) t AV a (min) (V) (fi cm) 1 P a (mg/1) (%) (mg/1) (%) (mg/1) 0 6.0 6.4 15 6.0 7.9 30 5.7 9.0 45 5.4 10.5 60 5.2 11.5 75 5.0 12.0 90 5.0 12.7 9.41 104 3.00 55 47 2.80 42 60 2.70 33 68 2.63 27 74 2.58 19 82 2.54 11 89 80 74 8 2.2 72 11 1.7 71 13 1.5 70 14 0.9 68 15 0.3 67 16 0.3 RUN 2-1 Membrane: IONAC MC-3470 against polypropylene screen I = 10 A c.d. = 526.3 A/m2 K x 103 e Electrolyte Concentration (ft cm) PH Anolyte 5 g/1 Na2S0i+ 6.4 9.44 NaOH to adjust pH Catholyte 25 g/1 NaOH 14* 12.7* K xlO3 Phenol T.O.C. (Pb) t AV e a _1 pH r a (min) (V) (ft cm) (mg/1) (%) (mg/1) (%) (mg/1) 0 12.9 6.4 9.44 102.0 78 0 0.0 15 15.0 8.1 3.04 51.0 50 0.2 30 16.0 9.4 2.80 29.0 71 72 8 0.1 45 16.0 11.0 2.67 9.0 91 60 15.0 12.0 2.58 1.0 99 67 14 0.1 75 14.0 12.5 2.52 0.0 100 90 14.0 13.0 65 16 0.1 105 13.5 13.5 2.48 120 13.5 14.0 2.42 62 21 0.0 Comments: The polypropylene screen produces a higher potential drop compared with the saran screen. RUN 2-2 Membrane: IONAC MC-3470 against saran screen I = 10 A cd. = 526.3 A/m2 K x 103 Electrolyte Concentration e -i (fi cm) pH Anolyte 5 g/1 Na2S0it 6.5 9.42 NaOH to adjust pH Catholyte 25 g/1 NaOH 14* 12.6* K xlO3 Phenol T.O.C. (Pb) t AV ea (min) (V) ($2 cm) 1 P a (mg/1) (%) (mg/1) (%) (mg/1) 0 8.0 6.5 9.42 100 - 77 0.0 15 8.0 7.8 3.12 45 55 72 6 0.2 30 8.0 8.8 2.9 25 75 70 9 0.1 45 7.9 9.5 2.82 11 89 68 13 0.1 60 7.5 10.5 2.75 5 95 66 14 0.0 75 7.3 11.0 2.68 2 98 65 16 0.0 90 7.1 11.5 2.60 0 100 64 17 0.0 105 7.0 12.0 2.58 63 18 120 7.0 12.0 2.56 1 21 RUN 2-3 Membrane: IONAC MC-3470 1=0 (no current applied) - K x 103 Electrolyte Concentration (ft cm) pH Anolyte 5 g/1 N32S01+ 6.5 9.47 NaOH to adjust pH Catholyte 25 g/1 NaOH 14* 12.7* K xlO3 Phenol T.O.C. (Pb) t AV e a PHa (min) (V) (ft cm) (mg/1) (mg/1) (mg/1) 0 0.8 6.5 9.47 95 73 0.0 15 0.75 5.7 10.74 95 73 0.5 30 0.7 6.1 11.22 95 73 0.7 45 0.01 6.3 11.47 95 73 1.0 60 6.6 11.64 95 73 1.2 75 6.6 11.8 95 73 1.5 90 6.6 11.9 2.0 Comments: No appreciable change in phenol or T.O.C. concentration was detected, but while the first 31 were withdrawn (to purge the system), the solution showed the brownish colour characteristic of the oxidation of phenol. RUN 2-4 Membrance: IONAC MC-3470 I = 3 A cd. = 157.9 A/m2 Electrolyte Concentration K x 103 e -i (ft cm) pH Anolyte 5 g/1 Na2S0i+ NaOH to adjust pH 6.2 9.44 Catholyte 25 g/1 NaOH 15* 12.6* t [min) AV (V) K xio3 e 3 -1 (ft cm) pHa Phenol (mg/1) (%) (mg/1) T.O.C. (%) (Pb) (mg/i; 0 4.7 6.2 9.44 100 _ 77 0.1 15 - 4.7 11.70 83 17 76 1 0.4 30 4.7 7.0 11.77 68 32 73 5 0.4 45 4.5 11.82 58 42 70 9 0.4 60 4.5 7.2 11.83 53 47 66 14 0.4 75 4.5 11.84 49 51 64 17 0.4 90 4.5 7.2 11.88 43 57 63 18 0.4 105 7.9 3.38 18 82 58 25 . 0.4 120 7.9 8.0 3.00 6 94 44 43 0.3 135 7.7 2.80 2 98 40 48 0.2 150 7.5 9.0 2.73 0 100 37 52 0.2 Note: Current was changed to 10 A at 90 min to observe pH response. During the high pH interval the solution had a brown-reddish colour and after the pH drop the colour changed to light yellow. RUN 2-5 Membrane: IONAC MC-3470 I = 6 A c.d. = 315.8 A/m2 K x 103 ,6 N-l Electrolyte Concentration (ft cm) pH Anolyte 5 g/1 Na2SOi+ 6.2 9.42 NaOH to adjust pH Catholyte 25 g/1 NaOH 15* 12.6* K xlO3 Phenol T.O.C. t AV ea (min) (V) (ft cm) L pHa (mg/1) (%) (mg/1) 0 5.8 6.2 9.42 85 15 3.74 49 42 65 4 30 5.7 6.5 3.50 28 67 62 7 45 3.21 19 78 59 10 60 5.6 7.1 3.05 12 86 56 14 75 2.88 7 92 54 17 90 5.5 8.3 2.80 5 94 53 18 105 2.78 2 98 53 18 120 5.5 8.5 2.76 0 100 52 20 KJ OO RUN 2-6 Membrane: IONAC MC-3470 I = 20 A cd. = 1052.6 A/m2 K x 103 Electrolyte Concentration (ft cm) PH Anolyte 5 g/1 Na2S0it 8.5 2.45 0.44 g/1 H2SO4 Catholyte 25 g/1 NaOH 14* 12.7* K xlO3 ... e Phenol T.O.C. AV a 1 (min) (V) (ft cm)_1 P a . (mg/1) (%) (mg/1) 0 12.7 8.5 2.45 105 81 15 12.4 24 77 81 0 30 9.7 12.5 1.80 5 95 80 1 45 8.9 2 98 78 4 60 8.7 15.5 1.67 1 99 72 11 75 3.4 0 100 66 18 90 8.2 17.0 1.58 59 27 105 7.9 52 36 120 7.8 17.5 1.50 42 48 RUN 2-7 Membrane: IONAC MC-3470 I = 10 A c.d. = 526.3 A/m2 Electrolyte Concentration K x 103 e -i (ft cm) pH Anolyte 5 g/1 Na2S0i+ 0.44 g/1 H2S0it 8.3 2.46 Catholyte 25 g/1 NaOH 14* 12.7* t (min) AV (V) K xlO3 e a -1 (ft cm) PRa (mg/1) Phenol (%) (mg/1) T.O.C. <%: 0 7.5 8.3 2.46 100 0 75 0 15 30 70 30 6.7 10.5 1.96 8 92 74 1 45 4 96 60 6.3 12.5 1.81 2 98 74 1 75 0 100 90 6.0 14.0 1.72 72 4 105 120 5.9 15.0 1.66 67 11 RUN 2-8 Membrane: NAFION 127 I = 10 A cd. = 526.3 A/m2 K x 103 e Electrolyte Concentration (ft cm) PH Anolyte 5 g/1 Na2S0ix 8.0 12.03 0.44 g/1 NaOH Catholyte 25 g/1 NaOH 14* 12.6* t (min) AV (V) K xlO3 e a -1 (ft cm) pHa Phenol (mg/1) (%) T.O.C. (mg/1) (%) 0 5.3 8.0 12.03 95 76 15 2.1 53 44 75 1 30 5.5 9.0 1.8 20 79 74 3 45 7 93 60 5.2 11.0 1.54 3 97 73 4 75 1 99 90 5.1 13.0 1.38 0 100 70 8 120 5.0 14.5 1.29 67 12 150 4.9 15.0 1.26 66 13 Note: Colour change observed from brown reddish to yellow when the pH dropped. RUN 2-9 Membrane: NAFION 127 I = 20 A c.d. = 1052.6 A/m2 K x 103 Electrolyte Concentration (ft cm) PH Anolyte 5 g/1 Na2S0LL 8.2 2.48 0.44 g/1 H2S0tt Catholyte 25 g/1 NaOH 15* 12.7* K xio^ Phenol T.O.C. t AV ea _ (min) (V) (ft cm) 1 P a (mg/1) (%) (mg/1) (%) 0 8.5 8.2 2.48 15 7.5 30 6.8 12.5 1.7 45 6.4 60 5.9 15.0 1.67 75 5.5 90 5.5 16.5 1.64 105 5.5 120 5.5 17.5 1.62 102 85 37 64 84 1 4 91 82 3 1 98 77 9 0 100 72 15 0 100 67 21 63 26 58 32 53 38 to RUN 2-10 Membrane: NAFION 127 I = 20 A c.d. = 1052.6 A/m2 K x IO3 ,e -l Electrolyte Concentration (ft cm) pH Anolyte 5 g/1 Na2S0it 30 12.84 5 g/1 NaOH Catholyte 25 g/1 NaOH 15* 12.7 K xlO3 Total Inorg. t AV 6a _1 Phenol Carbon Carbon T.O.C. (min) (V) (ft cm) 1 pHa (mg/1) (%) (mg/1) (mg/1) (mg/1) (%) 0 3.9 30.0 12.84 102 0 82 0 82 15 3.9 23.5 12.66 92 10 82 2 80 2 30 4.9 17.0 12.54 69 32 82 4 78 5 45 5.0 11.0 12.24 36 65 82 8 74 10 60 7.0 2.15 15 85 72 5 67 18 75 5.0 9.5 1.56 3 97 62 2 60 26 90 5.0 11.0 1.40 0 100 55 3 52 37 105 5.0 120 5.0 13.5 1.28 43 0 43 48 135 5.0 150 5.0 14.0 1.26 39 0 39 52 Comments: The colour of the electrolyte changed from brown reddish to yellow when the pH drop was produced. RUN 2-11 Membrane: IONAC MA-3475 (anionic) I = 20 A c.d. = 1052.6 A/m2 K x 103 e Electrolyte Concentration (ft .cm) pH Anolyte 5 g/1 Na2S0ii 0.44 g/1 H^SOtt 8 2.4 Catholyte 25 g/1 NaOH 15* 12.7* K x10^ t AV ea . _ Phenol Carbon (mg/1) (min) (V) (ft cm) P a (mg/1) (%) Total Inorg. T.O.C. 0 7.7 8.0 2.40 94 15 2.80 26 30 7.6 8.1 3.15 9 45 8.33 4 60 7.4 8.2 11.20 2 75 11.75 0 90 7.1 8.3 12.08 120 7.0 9.5 12.38 150 6.6 10.0 12.53 0 78 0 78 0 72 2 76 3 90 78 3 75 4 96 7 71 9 98 78 18 60 23 100 33 45 42 78 45 33 58 78 52 26 67 78 60 18 77 Comment: The anionic membrane showed a change in colour (from yellow to brown) after the run. RUN 3-1 Electrolyte 1(A) c.d. (A/m 2) 5 g/1 Na2S0it 10 526.3 NaOH to adjust pH t (min) AV (V) ' K e (ft x 103 cm) PH (mg/D Phenol (%) (mg/1) T.O.C. (%: 0 9.7 6.2 9.46 110 84 15 10.5 6.5 3.78 60 45 80 5 30 9.7 6.7 3.62 40 64 78 7 45 9.1 6.8 3.52 19 83 77 8 60 8.9 6.9 3.46 9 92 75 11 75 8.7 6.9 3.42 3 97 71 15 90 8.5 7.0 3.39 1 99 67 20 120 8.5 7.0 3.36 0 100 60 29 RUN 3-2 Electrolyte 1(A) cd. (A/m 2) 5 g/1 Na2SOtx 0.44 g/1 NaOH 10 526.3 t AV K x 103  6 -1 Phenol Carbon (mg/1) (min) (V) (ft cm) pH (mg/1) (%) Total Inorg. T .O.C. % T.O.C. 0 6.5 8.0 11.98 106 81 0 81 15 6.4 7.5 11.88 71 33 81 5 76 7 30 6.3 7.2 11.81 52 51 81 7 74 9 45 6.3 7.0 11.72 34 68 81 8 73 10 60 6.3 6.9 11.67 18 83 81 10 70 13 75 6.4 6.8 11.58 12 89 81 13 68 16 90 6.5 6.8 11.57 9 92 81 18 63 22 120 6.6 6.7 11.54 5 95 81 26 55 32 RUN 3-3 Electrolyte 1(A) cd. (A/m ) 5 g/1 Na2S0i+ 10 526.3 0.44 g/1 H2S0LL AT7 K x 103 Phenol T.O.C. t AV e (min) (V) (ft cm)"1 pH (mg/1) (%) (mg/1) (%) 0 7.8 8.4 2.50 93 75 15 7.3 8.6 28 70 75 0 30 7.2 8.8 2.45 12 86 74 1 45 6.9 8.9 4 95 71 5 60 6.7 8.9 2.43 2 98 69 7 75 6.5 8.9 1 99 68 9 90 6.4 8.9 2.41 0 100 65 13 120 6.4 8.9 2.39 61 19 150 6.4 8.9 2.38 56 25 RUN 3-4 Electrolyte 1(A) cd. (A/m ) 5 g/1 Na2S0ix 20 1052.6 0.44 g/1 H2SO11 t AV K x 103 T Phenol T.O.C. (min) (V) (ft cm)-1 pH (°c) (mg/D (%) (mg/1) (%: 0 9.6 8.5 2.47 23 95 73 15 9.7 9.1 2.28 23 75 71 3 30 9.2 9.2 2.28 25 10 89 68 7 45 8.4 9.2 2.27 3 97 64 12 60 8.2 9.2 2.25 26 2 98 59 19 75 8.2 9.3 2.24 1 99 54 26 90 8.2 9.4 2.22 27 0 100 47 36 120 9.3 9.5 2.20 28 34 53 150 10.7 9.6 2.19 28 23 68 (Pb)" = undetectable. RUN 3-5 Electrolyte 1(A) cd. (A/m ) 5 g/1 Na2S0i+ 20 1052.3 0.44 g/1 NaOH t (min) AV (V) K e (a x io3 cm) pH T (°C) (mg/1) Phenol (%) Carbon (mg/1) Total Inorg. T, .O.C. (%) (Pb) (mg/1] 0 9.2 8.2 12.04 24 102 86 0 86 0 0.0 15 9.1 7.9 11.91 48 53 84 6 78 9 0.2 30 8.9 7.7 11.76 25 26 75 84 13 71 17 0.2 45 8.9 7.5 11.56 12 88 84 22 62 28 0.0 60 9.0 7.2 11.26 26 3 97 84 30 54 37 0.0 75 9.4 7.1 10.76 0 100 84 37 46 46 0.0 90 10.4 7.1 10.26 27 84 46 38 56 0.0 120 11.4 7.1 9.56 28 83 60 23 73 0.0 150 12.0 7.0 8.84 28 81 62 19 78 0.0 RUN 3-6 _2 _Electrolyte 1(A) c.d. (A/m ) 5 g/1 NaaSOit 30 0.44 g/1 H2SO4 1578.9 t (min) AV (V) K e (ft x 103 cm) pH T (°C) (mg/1) Phenol (%) (mg/1) T.O.C. (%) 0 14.0 8.6 2.48 24 96 76 15 11.9 8.9 2.40 26 20 79 71 7 30 11.6 9.1 2.36 28 4 96 68 11 45 11.4 9.3 2.32 29 1 99 61 20 60 13.0 9.5 2.30 31 0 100 47 38 75 14.0 9.6 2.28 33 0 100 37 51 90 15.0 9.7 2.26 34 28 63 120 15.5 9.8 2.25 36 24 69 (Pb): undetectable RUN 3-7 Electrolyte 1(A) c.d. (A/m ) 5 g/1 Na2SOLf 30 1578.9 0.44 g/1 NaOH t AV K e x 103 T Phenol Carbon (mg/1) (min) (V) (ft cm) pH (°C) (mg/1) (%) Total Inorg. T.O.C. (%: 0 11.9 8.1 12.08 23 110 0 86 0 86 0 15 10.7 ' 11.97 25 36 67 74 10' 74 14 30 10.4 7.6 11.80 26 12 89 61 22 61 29 45 10.5 11.47 28 1 99 48 35 48 44 60 12.9 7.2 10.80 29 0 100 32 51 32 63 75 13.4 10.43 31 25 58 23 73 90 13.5 7.3 10.18 32 16 64 16 81 120 13.6 7.0 9.5 35 7 73 7 92 [Pb]: undetectable RUN 3-8 Electrolyte 1(A) cd. (A/m ) 5 g/1 Na2SOi+ 10 526.3 0.44 g/1 H2S0LL t (min) AV (V) K x 103 -1 (ft cm) pH (mg/1) Phenol (%) (mg/1) T.O.C. (%: 0 4.5 30.5 2.5 108 0 83 0 15 33 70 80 4 30 4.2 31.0 2.44 13 88 79 5 45 5 95 77 7 60 4.1 31.5 2.42 1 99 72 13 75 0 100 68 18 90 4.1 31.5 2.40 67 19 120 4.1 31.5 2.38 61 27 RUN 3-9 Electrolyte 1(A) c.d. (A/m ) 5 g/1 Na2S0i+ 10 526.3 0.44 g/1 NaOH AV Ke X 103 Phenol Carbon (mg/1)  (min) (V) (ft cm)"1 pH (mg/1) (%) Total Inorg. T.O.C. (%) 0 4.6 32 12.06 105 80 0 80 0 15 4.6 32 11.88 72 31 80 2 78 3 30 4.6 32 11.81 48 54 80 5 75 6 45 4.6 32 11.64 38 64 80 8 72 10 60 4.6 32 11.41 18 83 80 14 66 18 75 4.6 32 10.96 7 93 80 20 60 25 90 4.6 32 10.36 1 99 80 27 53 34 120 4.6 32 9.24 0 100 80 40 40 50 150 4.6 32 7.7 72 45 27 66 RUN 3-10 Electrolyte 1(A) c.d. (A/m 30 g/1 N32S01+ 20 1052.6 0.44 g/1 t (min) AV (V) K x 103  e -i (ft cm) PH T (°C) (mg/1) Phenol (%) T.O.C. (mg/1) (%: o' 6.6 32 2.46 25 98 0 80 0 15 20 80 77 4 30 6.1 32 2.42 25 7 93 75 6 45 3 97 70 13 60 6.0 32 2.40 26 1 99 65 19 75 0 100 58 28 90 6.2 32 2.38 28 54 33 120 6.4 32 40 50 150 6.7 32 2.36 28 28 65 RUN 3-11 Electrolyte 1(A) cd. (A/m ) 30 g/1 N32S01, 20 1052.6 0.44 g/1 NaOH K x 103 Phenol Carbon (mg/1) t AV e s  (min) (V) (ft cm) pH (mg/1) (%) Total Inorg. T.O.C. (%) 0 4.4 32.0 12.03 113 0 87 0 87 15 54 52 87 7 80 8 30 4.4 31.5 11.93 29 74 87 13 74 15 45 14 88 87 22 65 25 60 4.5 31.0 11.70 3 98 87 32 55 37 75 0 100 87 40 47 45 90 4.6 30.5 11.16 87 51 36 59 120 4.7 30.0 10.42 87 65 22 75 150 87 70 17 80 RUN 3-12 -2. Electrolyte KA) c.d. (A/m ) 5 g/1 Na2S04 10 526.3 0.44 g/1 H2S0I+ t AV K e x 103 s-1 cm) Phenol T.O.C. (min) (V) (ft pH (mg/1) (%) (mg/1) (%: 0 8.7 8.0 2.43 525 0 395 15 320 39 30 8.6 8.7 2.25 175 67 380 4 45 75 86 60 8.5 9.3 2.18 25 95 375 5 75 15 97 90 8.2 9.3 2.16 5 99 370 8 120 7.8 9.3 2.14 0 100 350 11 RUN 3-13 Electrolyte KA) _2 c.d. (A/m ) 5 g/1 Na2S0Lj 10 526.3 0.44 g/1 H2S01+ K x 103 Phenol t AV e (min) (V) (ft cm) 1 pH (mg/1) (%) 0 9.3 8.7 2.42 1100 0 15 792 28 30 8.3 8.9 2.32 506 54 45 341 69 60 8.0 8.9 2.28 187 83 75 88 92 90 7.9 9.1 2.27 22 98 120 7.8 9.2 2.26 5 100 Comment: The net change in T.O.C. was practically undetectable due to the high amount of carbon present in solution. RUN 3-14 Electrolyte KA) _2 c.d. (A/m ) 5 g/1 Na2SOi+ 10 526.3 2.2 g/1 H2S01+ t (min) AV (V) K x 103 e -i (ft cm) pH (mg/1) Phenol <%) (mg/1) T.O.C. (%: 0 5.4 13.5 1.8 90 77 15 5.2 13.7 30 67 74 4 30 5.1 14.0 1.78 9 90 73 5 45 5.0 14.2 3 97 72 6 60 4.9 14.3 1.75 1 99 69 10 75 4.9 14.5 0 100 68 12 90 4.9 14.5 1.75 66 14 120 4.9 14.5 1.75 64 17 RUN 3-15 Electrolyte 1(A) c.d. (A/m ) Q (£/min) P(Kpa) 5 g/1 NaaSOi, 20 1052.6 0.55 145 0.44 g/1 H2S0i+ .„ K x 103 Phenol T.O.C. t AV e (min) (V) (ft cm) 1 pH (mg/1) (%) (mg/1) (%) 0 12.1 8.3 2.5 105 0 79 0 15 28 73 74 6 30 10.5 8.5 2.43 9 91 71 10 45  98 69 13 60 10.9 8.7 2.40 1 99 62 22 75 0 100 55 30 90 11.1 8.8 2.38 48 39 120 12.0 8.9 2.37 36 54 RUN 4-1 K x 103 T. Phenol 6 in Electrolyte 1(A) c.d. (A/m~2) (ft cm)"1 pH (°C) Cin(rag/:L) 5 g/1 Na2S0i4 10 526.3 7.9 2.5 24 100 0.44 g/1 H2SO4 Q AV P T Phenol (Jo/min) (V) (kPa) °™ C „ (mg/1) (%) 0.11 8.7 108 32 9 91 0.25 8.6 120 28 42 58 0.40 8.3 129 26 55 45 0.55 7.9 143 25 67 33 0.85 7.4 184 24 75 25 1.10 7.1 232 24 79 21 RUN 4-2 K x IO3 T. Phenol 6 in Electrolyte 1(A) cd. (A/m~2) (ft. cm)"1 pH (°C) Cin(mg/ 5 g/1 N32S04 10 526.3 8.1 2.5 24 105 0.44 g/1 H2S0it Q AV P T Phenol (A/min) (V) (kPa) C (mg/1) (%)' 0.25 8.6 119 27 43 60 0.40 8.5 129 26 60 44 0.55 8.3 143 25 71 34 0.85 7.8 184 25 80 26 1.10 7.3 232 24 85 21 1.30 7.0 280 24 90 17 RUN 4-3 Kg x IO3 T Phenol Electrolyte 1(A) cd. (A/m"2) (ft cm)"1) pH (°C) Cin(nig/1) 5 g/1 UazSOn 10 526.3 8.0 2.5 23 95 0.44 g/1 H2S0i+ Q AV P U/min) (V) (kPa) 0.250 8.7 108 0.400 8.6 120 0.550 8.3 129 0.850 7.9 143 1.100 7.4 184 1.300 7.1 232 T Phenol (°C) Cout(mg/1) (%) 27 39 59 25 53 44 24 63 34 24 72 24 23 76 20 23 81 17 RUN 4-4 Kg x 103 T Phenol Electrolyte 1(A) c.d. (A/m"2) (ft cm)"1 pH (°C) c±a^lAs/^) 5 g/1 Na2S04 10 526.3 8 2.45 23 580 0.44 g/1 H2S0i+ Flow AV P U/min) (V) (kPa) 0.11 8.7 108 0.25 8.6 120 0.40 8.3 129 0.55 7.9 143 0.85 7.4 184 1.10 7.0 234 T Phenol (°C) Cout(m8/1> (%) 31 175 70 27 365 37 26 435 25 25 470 19 23 510 12 23 526 9 RUN 4-5 Kg x 103 T Phenol Electrolyte 1(A) cd. (A/m"2) (ft cm)"1 pH (°C) Cin(mg^ 5 g/1 Na2S0k 20 1052.6 7.9 2.43 24 110 0.44 g/1 H2S0it Q AV P T Phenol (4/min) (V) (kPa) ™\ C Amg/1) (%) .11 13.4 110 48 7 94 .25 13.4 122 38 40 63 .40 13.3 129 32 62 44 .55 12.6 145 27 73 34 .85 11.5 185 26 80 27 1.10 10.8 234 25 90 19 RUN 4-6 K x 103 T. Phenol Electrolyte KA) _2 c.d. (A/m ) e -i (ft cm) in pH (°C) C (mg/1) in 5 g/1 Na2S0i+ 20 1052.6 7.9 2.45 24 515 0.44 g/1 NaOH Q AV P T Phenol (£/min) (V) (kPa) out (°C) C (mg/1) out (%) 0.11 12.2 110 48 56 89 0.25 11.8 121 38 260 49 0.40 11.3 129 30 335 35 0.55 10.8 145 27 390 24 0.85 10.1 185 25 430 16 1.10 9.7 234 24 450 12 H1 RUN 4-7 Kg x IO3 Tin Phenol Electrolyte 1(A) cd. (A/m~2) (ft cm)"1 pH (°C) Cin(mg/ 5 g/1 N32S04 10 523.6 8.1 2.43 25 110 0.44 g/1 H2S01+ Q AV P T Phenol (£/min) (V) (kPa) C (mg/1) (%) 0.11 5.9 110 30 68 38 0.25 5.9 115 28 87 21 0.40 5.9 115 27 98 11 0.55 5.9 115 26 100 9 0.85 5.9 115 25 103 6 1.10 5.9 115 25 106 4 Note: The anode used was the feeder plate only (refer to general specifications). Therefore, the pressure was practically constant at 115 Kpa. RUN 4-8 Electrolyte KA) -2 cd. (A/m ) K x 103  e -i (ft cm) T. Phenol PH (°C) Cin(mg/1) 5 g/1 Na2SOi+ 0.44 g/1 H2SOi+ 10 526.3 8.0 2.43 24 100 Q (Jo/min) AV (V) P (kPa) T «. out (°c) Phenol C (mg/1) (%) out 0.11 6.3 112 31 5 95 0.25 6.0 128 28 33 67 0.40 5.9 143 27 47 53 0.55 5.8 162 26 55 45 0.85 5.7 204 25 66 34 1.10 5.6 271 25 71 29 Note: This run corresponds to the smaller particle size (0.7 < dp < 1.1 mm) For bed data refer to general specifications. 158 APPENDIX 3 Mathematical Models The electrochemical oxidation of phenol is a heterogeneous process that takes place on the surface of the anode. A simplified picture of the process is that the disappearance of phenol is the result of two steps which occur in series: a) the transfer of phenol molecules or phenoxonium ions from the bulk of the solution to the surface cf the electrode b) an electrochemical reaction by which the phenol is converted into some oxidation products as discussed in Chapter 2. Two limiting cases are considered in which either of the two steps are so slow that they control the overall rate of the process. Such models assume that adsorption phenomena and transfer of oxidation products from the electrode surface are not rate-limiting. A third case is also presented where both the resistance to mass transfer and to electro chemical reactions are comparable in magnitude. 1. Mass transfer controlled model The model presented here deals with a packed bed electrochemical reactor operating continuously in plug flow. Assumptions - The resistance to the electrochemical reaction is negligible compared with the resistance to mass transfer. In other words, the concen tration of phenol at the surface of the electrode is negligible 159 compared with the concentration of phenol in the bulk of the solution. Variations in phenol concentration in the directions perpendicular to the flow are neglected compared to the variations of concentration in the direction of flow. All the bed is active for phenol oxidation. Q,C Mass balance for a differential height. Q dCA = - K a(S W dy)(C - ) [A.l] b 11 D " s The superficial velocity referred to the cross sectional area of the bed is giver. by: u = Q/W S Q,C Neglecting , (Eq. A.l) can be expressed s Fig.A-1 Plug-flow packed bed reactor K a dy dc. = _ .3 Ci u A. integrating, cA(y) dC Al K a dv m u In C.(y) K a v A ' m u [A. 2] which can also be expressed in terms of the fractional conversion X, for and Aj A2 A2 ____ - 1 - — [A. 3] 160 then, K a L Jon(l - X) = - -5L [A.4] The specific surface area of the bed is given by the sum of the specific surface area of the feeder plate plus the specific surface area of the particles, as where 5 is a shape factor for the particles, and e is the fraction of voids (49). The average mass transfer coefficient can be estimated using the correlation by Pickett and Stanmore (45) as follows, K dp -2=— = 0.83 (u dp 0 . 56 V 1/3 I v J [A.6] for 23 < Re < 520 This equation was developed using a single layer of spheres and correlated the experimental data within ± 10%. It should be noted that the effect of gas evolution is not included in equation A.6. Studies of mass transfer on gas-evolving packed bed electrodes have been made recently (47), but for the case of a stationary solution where the motion of the electrolyte in the cell is only provided by gas bubbles. It was found that the rate of mass trans fer was increased by gas evolution. For the present case of forced convection, a correlation for mass transfer in gas evolving particulate electrodes could not be found in the literature. However, it is reasonable to expect that enhancement of the mass transfer coefficient due to gas bubbling is less significant than in free convection, since the electrolyte is mainly moved by mechan ical circulation. 161 2. Electrochemical reaction controlled model Assumptions - The resistance to electrochemical reaction is so high that the concen tration of phenol at the surface of the particles is equal to the con centration at the bulk of the solution. - The oxidation reaction is assumed to be first order in phenol concen tration. - The electrode potential is assumed to be uniform all over the cell. This assumption represents a significant simplification because in reality an electrode potential distribution exists within and along the electrode (10). In a differential length of the anode, the potential distribution would be as indicated in Fig. A-2. CURRENT PACKED FEEDER BED.ANODE ' k 1 d<f> o i =-K -r^ s e dx s 7? UJ O 0. -• OO ) ma j 0 L K — m e dx m i = i + i . s m x = 0 The potential distribution in the x direction origin ates because the total charge at a central x is carried by X=SQ both the metal and the solution. Fig. A-2. Potential distribution in a particulate electrode. 162 The total charge at x = 0 is entirely carried by the metal of the feeder plate but at x = S the current is entirely carried by the 3. solution. This implies that the metal potential presents its maximum gradient at x = 0 and the solution potential at x = S , so that the shape a of the curves <j> vs x and cj> vs x are as indicated in the figure, m s In the ideal case of a metal of infinite conductivity, the poten tial drop through the metal is zero and is represented by the horizontal dotted line. Generally, the conductivity of the metal is much greater than that of the electrolyte, therefore the solution potential drop would be greater than the metal potential drop. Thus, the electrode poten tial V* = <j> - cb tends to increase towards x = S . This means that the Tm s a side reactions are more likely to occur at the edge of the packed bed opposite to the feeder electrode. A two dimensional model for potential, current, and concentration distributions has been developed (48) for a concentric cylindrical elec trode, but the mathematical solution involved is extremely complicated. The model to be presented here assumes that some average elec trode potential exists within the cell. This permits a simple correla tion of all the variables affecting the process. For the case where the electrochemical reaction is the rate limiting step, the mass balance in a differential length of cell is: Q dC = - K a S W dy C [A. 7] % rA -The electrochemical reaction coefficient K is related to the electrode r potential by definition °AZAVa F 163 Under the assumption of uniform electrode potential and C = C , s % eq. A.7 can be integrated (with u = Q/W S), cA(y) dC Al K a dy rA An CA(y) Kay rA [A.9] for y = L cA(y) = CA2 and x = 1 -Ai then, £n(l - X) = -K a L J^A u [A.10] a) Single reaction If the only reaction occurring at the electrode is the electro chemical oxidation of phenol, a local reaction rate (referred to the true electrode area) would be given by: iA(y) TT = Kr CA(^ [A.11] Solving for C.(y) from eq. A.9 and substituting in equation A.11 iA(y) ~TY~ " Kr CAl eXp K ay rA [A.12] the average true current density through the cell XA = a" is, iA(y) XA = dy [A.13] dy 164 Substituting eq. A.12 into eq. A.13 and integrating, an average reaction rate can be obtained as: XA . A* z F a L Equation A.14 implies that if the initial concentration of phenol is increased holding all the other parameters constant, the value of will be lower to sustain the same current flowing. To illustrate this, eq. A.8 is combined with eq. A.14, to find the relation between the average reaction rate, and the electrode potential, as represented in Fig. A-3. (V*)' V* Fig A-3 Schematic representation of eq. A.14 K a u [A.14] If the initial concentration of phenol is increased from C. to A] (C. )1 at a constant current input, the value of the electrode potential Al (V*)' will be lower, and therefore the value of K will drop too. Analogous reasoning shows that Kr will also decrease if u is increased, or if a is increased at a constant applied current. Equation A.14 also implies that when the electrode potential, tends to infinity, the 165 rate of reaction i^/z F will tend to a maximum given by CA u a L b) More than one electrochemical reaction occurs at the electrode If, for example, a side reaction of water electrolysis occurs in parallel to the phenol oxidation, the average current density will be given by the sum of the average current densities driving each reaction. If water electrolysis is represented with the subscript w, the current balance would be expressed by, i = i. + i where i = I/a A w Using eq. A.14 for both partial average currents, 1 = ZA F C, u A Ai 1 - exp K a L rA u z F C u w w a L' 1 - exp K a L r w u where K = K exp rA r A A A and K = K exp r r w w a. z V* F A A a R T a z V* F w w a [A.15] [A.8] R T Therefore, the average current density, referred to the true area of the bed, can be related to the electrode potential as represented in Fig. A-4, where it is assumed that both reactions take place at all potentials greater than zero. Figure A-4 shows that there is only one electrode potential at which i. + i = i for each set of parameters a, L, u, C . Aw A]^ If the initial concentration of phenol is increased holding all other parameters constant, a lower value of V* will sustain the same total average current density i. v* Fig.A-4 Schematic representation of eq. A.15 Analogously» if u or a is increased V* should also decrease, which means that the values of the reaction rate constants K and K would be A w lower. 3. Mass transfer and electrochemical reaction controlled model i The most general case to.consider is when both resistances to mass transfer and to electrochemical reaction are comparable. Under steady-state (when the concentration profiles have been developed), the rate of electrochemical reaction and the rate of mass transfer per unit area of the electrode, will be equal in a differential of cell length (dy), - Q dCA K a S-W dy (C. - CA ) [A.l] \ m \ As — Q dC. = K a S W dy C. [A.7] *b r As Therefore, under steady-state, K (C. - CA ) = K C. m A, A r A b s s which permits to express the concentration at the surface as a function of the concentration in the bulk of solution, 167 K C. m *b X (K + K ) s m r [A.16] Substituting eq. A.16 into eq. A.l with Q = u/W S, and integrating, y CA(y) dC. CA ^ K K a dy m r  (K + K ) u 0 m r An cA(y) K K m r a y JAi (K + K ) u r for y = L,C.(y) = CA and with X = 1 - — J ' A J hz C [A.17] An(l - X) = -K K m r a L (K + K ) u m r where K K K m r K + K m r [A.18] [A.19] is the overall rate constant, which could also be found applying the con cept of resistances in series as: K K K m r [A.20] Analogous to eq. A.15, in the general case when a side reaction of water electrolysis occurs in parallel to phenol oxidation, the total average current density would be given by, z. F C. u l = a L 1 - exp K a L u z F C u w w a L 1 - exp K a L r w u [A.21] 168 Note that K is used for the water electrolysis reaction, since it w it is an activation controlled process and would not be affected by mass transfer. In this case, an increase in the phenol inlet concentration C. or in the specific surface area of the bed, would result in lower Al values of (or V*) for a same average current density, as discussed previously. However, an increase in the superficial velocity u would result in increased mass transfer coefficients (eq. A.6) but at the same time, lower reaction rate constants K , K should be expected. r ' r A w 4. Mass transfer model for a batch recirculation system If the cell is operated in a batch recirculation system, as shown in Fig. A-5 both inlet and outlet concentrations will be a function of time. + CA (t) Fig. A-5 Schematic repre sentation of a batch sys tem It is possible to find an expression to correlate the inlet and outlet concentrations with time, utilizing the approach of Pickett (32, p. 325) From equations A.3 and A.4, JA2 exp K a L m u [A.22] An instantaneous material balance over the recirculation tank would be written as, dC. Ct - C, = t A2 Aj m dt [A.23] 169 with t = V /Q m m Combining equation A. 22 with A.23 would give, - cA = t Ai m at C. exp Al K a L m [A.24] integrating, yields, rt f > K a L m exp u - 1 dt t m JAi dC, JAi JA0 In Ai :A0 exp ( K a L m u - 1 t t m C. = C. exp Al A0 f K a L] * m -1 t exp u t m^ [A.25] Substituting C. from eq. A.25 into eq. A. 22, Al C. = C. exp A2 A0 f f K a ll exp m u - i -K a L t m t u m [A.26] defining, t* = = dimensionless time m K a L m = dimensionless mass transfer group A0 A2 X = = fractional conversion Equation A.26 can be written as, X = 1 - exp [ (exp (-0) - 1) t* - 0] [A.27] Note that if the reaction controls the process, an analogous mathematical solution for a constant applied current operation is not possible, because the value of will change as the concentration changes with time, as discussed from eq. A.14. 171 APPENDIX 4 Calculations 1. Batch experiments. Calculation of theoretical phenol fractional conversion if mass transfer controls. This estimate will be suitable to compare all those experiments in groups No. 2 and No. 3, performed under the following experimental condi tions, Flow rate 1.12 £./min Particle size 1.7 < dp < 2.00 mm Particle weight 250 g 250 e o Particle volume = n . " & ,—r = 23 cm3 *density of Pb02 from 9.375* g/cm3 ref. (49) Void fraction (57-23)/57 = 0.60 Specific surface area of the bed (eq. A.5) a=l+6 *•(!- e) S % x dp taking £ = 0.75 for A and, from ref. (49), and dp =0.185 cm. 1 . 6 x (l - Q.57) on , -1 a = -T—Z 1— r» "7c ^ n -ice = /U.o cm 0.3 cm 0.75 x 0.185 Determination of the mass transfer coefficient from eq. A.6,. . . K = 0.83 f (Re)°-56(Sc)1/3 m dp Using an equation for diffusivity of liquids given by Wilke (50). (Original nomenclature.) D = = y VA°-6 D = 7.4 x IP'8 (2.6x18)^x 297 = ^ x 1Q-6 cm2/s 1 x (105)°-6 Superficial velocity of the liquid, u = 1,12 £/mln x (I03cm3/Jo) x (1 min/60 s) = 12.4 cm/s 5 x 0.3 cm2 u dp 12.4 cm/s x 0.185 cm oor, . Re - c-= = 229.4 V 0.01 cm2/s Sc = V 0.01 cm2/s = 10g6 9.2 x 10 cm2/s Q 9 x m"^ —^ K = 0.83 x * r\ i QC (229.4) (1086) = 8.7 x 10 J cm/s ±10% m U.lo5 Extreme values of K , taking into account the 10% error in correlation A.6, K~ = 0.078 cm/s K+ = 0.0096 cm/s m m The dimensionless mass transfer group is given by K a L K 20.6 cm"1 38 cm e = _m = _ni u 12.4 cm/s Therefore, the extreme values of 0 will be e~ = 0.49 6+ = 0.60 The phenol fractional conversion for mass transfer control is given by, X= 1 - exp[(exp(-6) - 1) t* - 6] [A.27] t 5 I with t* = — and t = , , „ . , — = 4.464 min. t m 1.12 £/mm m 173 Using the extreme values of 6, equation A.27 is used to calculate the range of theoretical fractional conversion for mass transfer control at a given time, as shown in Table A-2. TABLE A-2 THEORETICAL PHENOL FRACTIONAL CONVERSION VS TIME FOR A MASS TRANSFER- -CONTROLLED BATCH SYSTEM t (min) + t* X~ X 0 .00 0.39 . 0.45 15 3.36 0.83 0.88 30 6.72 0.95 0.97 45 10.08 0.99 0.99 60 13.44 1.00 1.00 A theoretical mass transfer controlled region is obtained in this manner which is represented in Fig. 28 for comparison with the experimental results. 2. Continuous experiments. Determination of experimental, mass transfer, and reaction rates constants. Procedure a) According to the plug flow equation, An(l - X) = - [A. 18] If - Jin/ (1 - X) vs is plotted from the experimental data, a straight line should be obtained, with a slope equal to K a L. The experimental rate constant can be determined from the slope. The specific surface area of the electrode is calculated from eq. A.5. 174 b) Using the empirical equation for the mass transfer coefficient, is estimated: K = 0.83 (Re)°'56(Sc)1/3 22 < Re < 520 m dp c) The reaction rate constant is then calculated from the equation of resistances in series: 1 K _1 K r m [A.20] where K K K m r K - K m TABLE A-3 CALCULATION OF EXPERIMENTAL, MASS TRANSFER AND REACTION RATE CONSTANTS FROM EXPERIMENTS 4-1, 4-2, 4-3* (USING AVERAGE PHENOL FRACTIONAL CONVERSION) Q ,/min) u (cm/s) -1 U -1 (cm/s) X - £n(l - X) Re K m (cm/s) K r (cm/s) 0.25 2.78 0.360 0. 59 0.89 51.4 3. 85 x 10"3 15.65 x 10' 0.40 4.44 0.225 0. 44 0.58 82.1 5. 01 x 10~3 8.06 x 10' 0.55 6.11 0.163 0. 34 0.42 113.0 5. 98 x 10~3 6.39 x 10' 0.85 9.44 0.106 0. 25 0.29 174.6 7. 64 x 10~3 5.19 x 10' 1.10 12.22 0.082 0. 21 0.24 226.0 8. 82 x 10"3 4.76 x 10' 1.30 14.44 0.069 0. 17 0.19 267.0 9. 69 x 10-3 4.54 x 10' *For these experiments: C = 100 ± 5 mg/1 1.7 < dp < 2.00 mm a=20.6 cm" (from page 171) Ao From Fig. 32: Slope = 2.42 cm/sec = K = —2.42 cm/s = 3>Q9 x 1Q-3 cm/g 20.6 cm x 38 cm u dp u x 0.185 cm .n r Re = 11 = n j-.— = 18.5 u v 0.01 cmz/s K = 0.83 -f (Re)0'56 (Sc)1/3 = 0.83 x 9-\\lf x Re°'56 (1086)1/3 = 4.24 x 10" m dp U.LoD TABLE A-4 CALCULATION OF EXPERIMENTAL, MASS TRANSFER AND REACTION RATE CONSTANTS FOR EXPERIMENT (4-4)* u 1 K K Q u m r (Jl/min) (cm/s) (cm/s) X - £n(l - X) Re (cm/s) (cm/s) 0.25 2.78 0.36 0.37 0.46 51.4 3.85 X io"3 2.21 X 10' 0.40 4.44 0.225 0.25 0.29 82.1 5.01 X io"3 1.95 X 10' 0.55 6.11 0.163 0.19 0.21 113.0 5.98 X io"3 1.84 X 10" 0.85 9.44 0.106 0.12 0.13 174.6 7.64 X IO"3 1.72 X 10 1.10 12.22 0.082 0.09 0.09 226.0 8.82 X io"3 1.67 X 10 *In experiment (4-4): C. = 580 mg/1 1.7 < dp < 2.00 mm a = 20.6 cm"1 "(from page 171) A0 From Fig. 32: Slope = 1.1 cm/s = K = !•! cm/s = 1<405 x 10~3 cm/s 20.6 cm x 38 cm with, Re = 18.5 u and K = 4.24 x 10~4 Re°"56 (as in Table A-3) TABLE A-5 CALCULATION OF EXPERIMENTAL, MASS TRANSFER AND REACTION RATE CONSTANT FOR EXPERIMENT (4-8)* Q (il/min) u (cm/s) -1 u (cm/s) 1 X An(l - X) Re K m (cm/s) K r (cm/s) 0.25 2.78 0.36 0.67 1. 11 25 5.3 x 10~3 2.68 x 10~3 0.40 4.44 0.225 0.53 0. 76 40 6.9 x 10"3 2.40 x 10"3 0.55 6.11 0.163 0.45 0. 60 55 8.2 x 10~3 2.27 x 10-3 0.85 9.44 0.106 0.34 0. 42 85 10.5 x 10"3 2.14 x lo"3 1.10 12.22 0.082 0.29 0. 34 110 12.1 x 10~3 2.09 x 10-3 *In experiment (4-8) C. = 100 mg/1 0.7 < dp < 1.1 e = 0.57 A0 Specific surface area of the bed (eq. A.5): 1 (1 - 0.57) x 6 , -1 a * 0.3 cm (0.75)(0.09) cm " 10 cm From Fig. 32: Slope - 2.81 cm/s = K : 2,81 fm/s - 1.78 x 10~3 cm/s (41.5cm )(38cm) Re =«= u dp/v = u 0.09/0.01 = 9 u K . 0.83 x 9-2 * 10"6 cm2/s (Re°-56)(1086)1/3 = 8.72 x lo"4 (Re)0'56 m U.uy cm 178 3. Estimates of typical % current efficiency a) Batch experiments % C.E. = F Z (m^S °*idized> x 100 [Eq. 7] Assuming a 4-electron transfer from the phenol molecule as proposed by Covitz (Reaction R.ll), the % C.E. can be calculated. Sample calculation: Run 3-13, recirculation time = 15 min. For the calculation, it is supposed that after 15 min, recirculation is stopped and the volume remaining in the tank is treated continuously. The steady state final concentration at the outlet is assumed to be equal to the recorded at 15 min time. total electrolysis time = 15 min + 5 A/1.12 A/min = 19.5 min moles phenol oxidized l = (1100-792)mg/l x 5A x 1 g mol/94 g x 1 g/103 mg = 16.4x10 g mol _3 _ 96500 coul/eq x 4 eq/mol x 16.4x10 g mol .. nn n 10 A x 19.5 mm x 60 sec/mm This is a sample calculation for Table 6, where % C.E. is given after 15 and 90 min recirculation times, b) Continuous experiments F z Q X C. Al % C.E. = - x 100 Therefore, current efficiencies for phenol oxidation in a single pass, are a function of flow, conversion, initial phenol concentration, and current applied. Sample calculation: Run 4-4, at Q = 0.25 A/min, % = 96500 coul/eq x 4 eq/mol x 0.25 A/min x 0.37 x 580 mg/1 x 10Q 10 A x 103 mg/g x 94 g/g mol x 60 sec/1 min % C.E. = 36.7. 179 4. Estimate of typical electrical energy requirements and costs a) Batch experiments Energy requirements _ I AV t Moles of phenol oxidized V (C. - C. ) K m AQ A2 Sample calculation: Run 3-13, recirculation time = 15 min. As it was assumed in the previous section, after the batch operation the tank volume is treated continuously. Total electrolysis time = 15 min + 5 1/1.12 A/min = 19.5 min Average AV = 8.5 V Energy = 10 A * 8.5 V x 1 kw/103 w * 19.5 min x 1 h/60min mol 5A x (1100-792) mg/1 x i g/l03mg x 1 g mol/94 g g mol Power costs are taken as 2^/kw-h for illustration, Cost =1.69 kw-h/g mol x 0.02 $/kw-h =0.03 $/g mol Energy requirements and costs are presented in Table 6 for the batch experiments Nos. 3-3, 3-12, 3-12 at 10 A, after 15 and 90 min recircula tion time. In Table 8 the energy requirements per volume of waste is estimated as follows: in Runs 3-12 and 3-13, 99% and 98% phenol removal is achieved at 90 min recirculation time. Total electrolysis time = 90 mm + 5A/1.12 A/min = 94.5 min 10 A x 8.5 V x 94.5 min x lp3 kw/w Energy = g0 y . /n , = 0.027 kw-h/A OJ 5A x 60 mm/1 h Energy = 0.027 kw-h/A x 3.785 A/1 gal = 0.1022 kw-h/gal 180 The electrical cost for treating 10 gal of waste would be: 102.2 kw-h/1000 gal x 0.02 $/kw-h = 2 $/1000 gal This would be the approximate cost for treating a 700 mg/1 effluent with a 99% phenol removal, under the operating conditions of experiments 3-12 or 3-13. b) Continuous experiments energy required I AV mol phenol oxidized Q (C. - C. ) Ai A2 Sample calculation: Run 4-4 at Q - 0.25 £/min, v 10 A x 8.5 V x 103 kw/w x 103 mg/g x 94 g/g mol . . . , . En6rgy " 0.25 £/mln * (580-365)mg/l x 60 min/1 h = 2'51 kw~h/g mo1 Cost of electrical energy per mol of phenol oxidized: 0.02 $/kw-h x 2.51 kw-h/g mol = 0.05 $/g mol 181 APPENDIX 5 Relevant Physical Data TABLE A-6 pH OF SOLUTIONS OF NaOH AND H^O^ AT 20°C (51) N Aq. NaOH solutions Aq. H^O^ solutions g/1 pH g/1 pH .1 40.0 14 49.0 0.3 0.1 4.0 13 4.9 1.2 0.01 0.4 12 0.49 2.1 TABLE A-7 CONDUCTIVITIES OF AQUEOUS SOLUTIONS OF NaOH, H2S0i+ AND Na2SOl+ AT 20°C (51) NaOH H2S0it Na2SOi+ K x 103 K x 103 K x 103 cone. e cone. e cone. e (g/1) (ft cm) 1 (g/1) (ft cm) 1 (g/1) (ft cm) 5 24.8 5.0 24.3 5.0 5.9 10.1 48.6 10.0 47.8 10.1 11.2 15.2" 71.3 15.1 70.3 15.2 15.7 20.4 93.1 20.2 92.0 20.3 19.8 25.5 23.9 30.8 27.9 182 % phenol ionized vs pH at 20°C dissociation constant " 1.28 x 10 ^ (51) Kd = [if] [C6H50~] / [C6H5OH] Ionized = r l [c R] fraction J D 0 Ionized _ d = d = 1.28 x 10 xu fraction " [R+] 1Q-pH = 1Q-pH % ionized = fraction/(fraction + 1) x ioo TABLE A-8 % PHENOL IONIZED VS pH pH 2 4 6 8 10 12 ionized to unionized _fi _/- _/ _o fraction 1.28 x 10 1.28 x 10 1.28 x 10 1.28 x 10 1.28 128 % ionized 1.28 x io"6 1.28 x io"4 1.28 x io"2 1.16 x iff1 56 99 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0058801/manifest

Comment

Related Items