UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Effect of water matrix on Vacuum UV process for the removal of organic micropollutants in surface water Duca, Clara 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2015_may_duca_clara.pdf [ 3.87MB ]
Metadata
JSON: 24-1.0167680.json
JSON-LD: 24-1.0167680-ld.json
RDF/XML (Pretty): 24-1.0167680-rdf.xml
RDF/JSON: 24-1.0167680-rdf.json
Turtle: 24-1.0167680-turtle.txt
N-Triples: 24-1.0167680-rdf-ntriples.txt
Original Record: 24-1.0167680-source.json
Full Text
24-1.0167680-fulltext.txt
Citation
24-1.0167680.ris

Full Text

Effect of water matrix on Vacuum UV process for the removal oforganic micropollutants in surface waterbyClara DucaM.A.Sc., University of Torino, Italy 2009B.Sc., University of Torino, Italy 2007A THESIS SUBMITTED IN PARTIAL FULFILLMENTOF THE REQUIREMENTS FOR THE DEGREE OFDoctor of PhilosophyinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Chemical and Biological Engineering)The University of British Columbia(Vancouver)February 2015c© Clara Duca, 2015AbstractUV-based advanced oxidation processes (UV-AOPs) have been demonstrated as effective technolo-gies for the removal of micropollutants in water. One promising UV-AOP is Vacuum UV (VUV),which relies on the formation of hydroxyl radicals (HO•) by the photolysis of water induced byVUV photons. TiO2/UV photocatalysis is another promising AOP. Both VUV and UV photo-catalysis are greatly affected by the water matrix, inorganic ions and natural organic matter (NOM).Water constituents can absorb the UV and VUV radiations, they can act as HO• radicals scavengers,and can produce radicals when photolyzed.The main objective of this research was to study the effects of water matrix on the efficiencyof VUV for the degradation of micropollutants (with atrazine as a model contaminant). First, theabsorbance of radiation at 254 nm and 185 nm was measured in the presence of different ionsand NOM. All the inorganic ions showed a molar absorption coefficient equal to zero at 254 nmexcept nitrate with a ε= 3.51 M−1 cm−1. On the other hand, at 185 nm all the ions absorbed185 nm radiation, with chloride showing the highest absorption coefficient ε=2791 M−1 cm−1.NOM showed a high absorption coefficient at both 254 and 185 nm ranging from 116 to 638 M−1cm−1 at 254 and from 1137 to 1537 M−1 cm−1 at 185 nm). Second, the HO• scavenging effectsof different components were evaluated; nitrate showed a detrimental effect both with UV/H2O2and with VUV. The presence of 50 ppm of bicarbonate reduced the degradation rate of atrazineconsiderably. Sulfate seemed to photolyze at 185 nm to form HO•. NOM was found to be a strongHO• scavenger: in the presence of 9 ppm (DOC) NOM, less than 1% of the HO radicals wereavailable to react with atrazine. The next component of this work involved developing a method forthe measurement of quantum yield of atrazine at 185 nm. This allowed to measure the degradationof atrazine due to photolysis only. Finally, this research investigated the combination of VUV withTiO2/UV. The results showed that incorporating photocatalysis cannot improve significantly theefficacy of VUV.iiPrefaceSeveral manuscripts in journals and in conference proceedings have been published or are underconsideration for publication pertaining to the results of this thesis. The following is a list of suchmanuscripts which have been published, presented in conferences or as posters.Journals:• Synthesis, characterization, and comparison of Sol Gel TiO2 immobilized photocatalysts.Clara Duca, Gustavo Imoberdorf, and Madjid Mohnseni. Int. J. Chem. react. Eng., 2013,Volume 11, Issue 2, Pages 633-639.• Novel Collimated Beam Setup to Study the Kinetics of VUV-Induced Reactions. Clara Duca,Gustavo Imoberdorf, and Madjid Mohnseni. Photochem. Photobiol., 2014, Volume 90, Issue1, Pages 238-240.• In addition chapter 6,7 and 8 are currently under preparation.Conference proceedings:• Composite titania photocatalytist coating for Vacuum UV photoreactor. Clara Duca, GustavoImoberdorf, and Madjid Mohnseni. 2nd North American Conference on Ozone, UltraVioletand advanced Oxidation technologies Toronto, On September 19-20, 2011.• Degradation of Micropollutants using Vacuum UV (VUV) Advanced Oxidation. GustavoImoberdorf, Clara Duca, and Madjid Mohseni. The17th International Conference on Ad-vanced Oxidation Technologies for Treatment of Water, Air and Soil (AOTs-17) San Diego,Ca November 7-10 2011.• Simultaneous inactivation of microorganisms and removal of micropollutants in VUV reac-tors. Laith Furatian, Clara Duca, Mehdi Bagheri, Gustavo Imoberdorf, and Madjid MohseniBC water and waste water association annual conference and trade show Penticton, BC April21-25 2012.iii• Effects of Inorganic Ions and NOM on the Degradation of Micropollutants during UV-H2O2AOP. Clara Duca, Gustavo Imoberdorf, and Madjid Mohseni. IUVA Moving forward: sus-tainable UV solutions to meet evolving regulatory challenges Washington DC, August 12-14,2012.• Study of the Kinetics of VUV-induced Degradation of Micropollutants. Clara Duca, GustavoImoberdorf, Madjid Mohseni 62nd Canadian Chemical Engineering Conference VancouverBC October 14-17, 2012.• Effect of NOM and Inorganic Ions on the Degradation of Micropollutants with Vacuum-UV Radiation. Clara Duca, Gustavo Imoberdorf, Madjid Mohseni 15th Canadian nationalconference and 6th policy forum on drinking water Kelowna BC October 21-24 2012.• New methodology for the quantification of the quantum yield of micropollutants at 185 nm.Clara Duca, Gustavo Imoberdorf, Madjid Mohseni. IUVA World congress and exhibition.Las Vegas, NV September 22-26 2013.• Treating fracturing flowback by biological and advanced oxidation processes. Yaal Lester,Tesfayohanes Yacob, Clara Duca, Kurby Sitterley, Julie Korak, James Rosenblum, and KarlLinden. 247th ACS meeting and exposition. Dallas TX March 16-20 2014.The last work was produced during my time I have spent working in the University of Boulder,Colorado. During that time I have worked on the flowback treatment and characterization.Poster presentations:• Advanced oxidation using vacuum UV, catalyst development. Clara Duca, Madjid Mohseni(Part of UBC Advanced Oxidation and Small Systems Research Group Poster Presentation).The Consensus Conference on Small Water Systems Management for the Promotion of In-digenous Health. Victoria, BC, March 21-22, 2010.• Composite photocatalytic coatings for vacuum-UV reactor. Clara Duca Res’Eau Waternet:3rd roundtable Vancouver BC March 12th, 2010.• Effect of water matrix on UV based processes. Clara Duca, Gustavo Imoberdorf, MadjidMohnseni. Big Value in Small Systems Vancouver BC, October 5-6, 2011. Behaviour ofmicropollutants and water matrix constituents exposed to 185 nm radiation. Clara DucaRes’Eau Waternet roundtable Toronto ON November 3th, 2012.• Degradation of micropollutants with Vacuum UV process (VUV). Clara Duca Res’Eau Im-pact Vancouver BC October 2-3, 2013.iv• Influence of Water Matrix on UV based Processes for the Detoxification of Water Clara DucaRes’Eau meeting Whistler BC May 28-29, 2014.vTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiGlossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Problem of micropollutants in drinking water . . . . . . . . . . . . . . . . . . . . 11.2 Presence of natural organic matter (NOM) in raw surface water . . . . . . . . . . . 21.3 Conventional water treatments for drinking water . . . . . . . . . . . . . . . . . . 31.4 Advanced oxidation processes (AOPs) . . . . . . . . . . . . . . . . . . . . . . . . 41.5 Thesis layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72.1 Hydrogen peroxide/UV process . . . . . . . . . . . . . . . . . . . . . . . . . . . 72.2 Photocatalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122.3 Vacuum UV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4 Potential of combined VUV and UV/TiO2 for micropollutant degradation . . . . . 192.5 Effect of water matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192.5.1 Effect of inorganic ions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202.5.2 Effect of natural organic matter (NOM) . . . . . . . . . . . . . . . . . . . 242.6 Knowledge gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263 Research objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28vi4 Experimental setups and procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . 304.1 Setups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304.1.1 Photoreactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304.1.2 UV and VUV reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324.1.3 VUV collimated beam set-up . . . . . . . . . . . . . . . . . . . . . . . . . 354.2 Experimental procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394.2.1 Actinometry at 254 nm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394.2.2 Actinometry at 185 nm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404.2.3 Hydrogen peroxide measurement . . . . . . . . . . . . . . . . . . . . . . 404.2.4 Absorption coefficient measurements at 185 and 254 nm . . . . . . . . . . 414.2.5 pH and Dissolved oxygen (DO) measurements . . . . . . . . . . . . . . . 414.2.6 Reproducibility and data accuracy . . . . . . . . . . . . . . . . . . . . . . 425 TiO2 photocatalyst development and evaluation of photocatalysis coupled with VUV 435.1 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445.1.1 Synthesis of the photocatalysts . . . . . . . . . . . . . . . . . . . . . . . . 445.1.2 Physical characterization of the photocatalysts . . . . . . . . . . . . . . . 455.1.3 Photocatalytic activity assessment . . . . . . . . . . . . . . . . . . . . . . 455.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465.2.1 Physical characterization of the photocatalysts . . . . . . . . . . . . . . . 465.2.2 Photocatalytic efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515.2.3 Evaluation of the deactivation of the photocatalysts . . . . . . . . . . . . . 535.2.4 Evaluation of the adherence of the photocatalysts to the support . . . . . . 545.2.5 Performance of the photocatalysts in a flow-through reactor . . . . . . . . 555.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576 Effects of inorganics on the degradation of micropollutants with VUV advancedoxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606.1 Material and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 616.1.1 Water samples and chemicals . . . . . . . . . . . . . . . . . . . . . . . . 616.1.2 UV/H2O2 and VUV irradiations . . . . . . . . . . . . . . . . . . . . . . . 616.1.3 Analytical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626.2.1 Molar absorption coefficients at 185 nm and 254 nm . . . . . . . . . . . . 626.2.2 Methodology for the determination of the observed kinetic constant . . . . 656.2.3 Degradation of atrazine in millipore water . . . . . . . . . . . . . . . . . . 656.2.4 Degradation of atrazine in the presence of different inorganic ions . . . . . 756.2.5 Effect of Na+ on VUV and H2O2/UV processes . . . . . . . . . . . . . . . 75vii6.2.6 Effect of nitrate on VUV and H2O2/UV processes . . . . . . . . . . . . . . 756.2.7 Effect of bicarbonate on VUV and H2O2/UV processes . . . . . . . . . . . 796.2.8 Effect of sulfate on VUV and H2O2/UV processes . . . . . . . . . . . . . 816.3 Effect of different inorganic ions on the degradation of atrazine . . . . . . . . . . . 836.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857 Quantitative effect of natural organic matter on the efficacy of Vacuum UV oxida-tion of atrazine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 877.1 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887.1.1 Water samples and chemicals . . . . . . . . . . . . . . . . . . . . . . . . 887.1.2 UV/H2O2 and VUV irradiations . . . . . . . . . . . . . . . . . . . . . . . 887.1.3 Analytical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 887.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897.2.1 Molar absorption coefficients at 185 nm and 254 nm . . . . . . . . . . . . 897.2.2 UV/H2O2 irradiations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 927.2.3 VUV irradiations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 987.2.4 NOM sensitization effect during VUV treatment . . . . . . . . . . . . . . 1027.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1038 The effect of chloride on Vacuum-UV and UV/H2O2 photo-induced degradation ofphenol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1058.1 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068.1.1 Water samples and chemicals . . . . . . . . . . . . . . . . . . . . . . . . 1068.1.2 UV/H2O2 and VUV irradiations . . . . . . . . . . . . . . . . . . . . . . . 1068.1.3 Analytical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1068.2 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1078.2.1 Molar absorptions coefficients at 185 nm and 254 nm . . . . . . . . . . . . 1078.2.2 Degradation of phenol with UV/H2O2 . . . . . . . . . . . . . . . . . . . . 1078.2.3 Degradation of phenol with VUV . . . . . . . . . . . . . . . . . . . . . . 1148.2.4 Effect of pH on the equilibrium of chloride . . . . . . . . . . . . . . . . . 1178.2.5 Effect of real water matrix on the degradation of atrazine in real water samples1198.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1219 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1239.1 Overall conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1239.2 Significance of the research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1269.3 Recommendations for future research . . . . . . . . . . . . . . . . . . . . . . . . 127viiiBibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128Appendix A Supplementary data for the photocatalysts deactivation tests and evalua-tion in a flow through reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139Appendix B Supplementary data for the UV/H2O2 and VUV irradiation in the pres-ence of different ions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144Appendix C Supplementary data for the UV/H2O2 and VUV irradiation in the pres-ence of different concentration an type of NOM. . . . . . . . . . . . . . . . . . . . . 151Appendix D Supplementary data for the measurement of the absorbance of inorganicions and NOM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154Appendix E Quantum yield of micropollutants . . . . . . . . . . . . . . . . . . . . . . 163Appendix F Degradation of micropollutants with ozone-generating Hg lamps (185 and254 nm) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165ixList of TablesTable 5.1 Percentage of the 4 polymorphous in the 5 photocatalysts. . . . . . . . . . . . . 48Table 5.2 Particle size of the 5 different sols. . . . . . . . . . . . . . . . . . . . . . . . . 48Table 6.1 Absorbtion coefficient of different compounds at 254 nm. . . . . . . . . . . . . 63Table 6.2 Absorbtion coefficient of different compounds at 185 nm. . . . . . . . . . . . . 64Table 6.3 Effect of different inorganic ions on the distribution of 254, 185 nm photons andon the HO radicals concentration. . . . . . . . . . . . . . . . . . . . . . . . . . 84Table 7.1 Absorption coefficient of various compounds at 254 nm and 185 nm. . . . . . . 90Table 7.2 distribution of HO radicals and 254 nm photons during UV/H2O2 with NordicNOM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97Table 7.3 distribution of HO radicals and 254 nm photons during UV/H2O2 with Suwan-nee NOM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97Table 7.4 distribution of HO radicals and 185 nm photons during VUV process with NordicNOM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100Table 7.5 distribution of HO radicals and 185 nm photons during VUV process with Suwan-nee NOM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100Table 8.1 Percentage of photons at 185 nm absorbed by water and by various concentra-tions of chloride ions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108Table 8.2 Effect of Bowen Island water matrix on the degradation of atrazine. . . . . . . . 120Table 8.3 Effect of peachland water matrix on the degradation of atrazine. . . . . . . . . . 121Table A.1 2,4-D conversion for photocatalysis/UV with millipore water and a flowrate of1L/min . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141Table A.2 2,4-D conversion for P25 slurry/UV with millipore water and a flowrate of 1L/min.141Table A.3 2,4-D conversion for VUV process with millipore water and a flowrate of 1L/min. 142Table A.4 2,4-D conversion for photocatalysis/UV with surface water and a flowrate of0.25 L/min. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142xTable A.5 2,4-D conversion for VUV with surface water and a flowrate of 0.25 L/min. . . 143Table D.1 Absorbance at 185 nm for different concentration of NaCl . . . . . . . . . . . . 154Table D.2 Absorbance at 185 nm for different concentration of NaHCO3 . . . . . . . . . . 160Table D.3 Absorbance at 185 nm for different concentration of NaNO3 . . . . . . . . . . . 160Table D.4 Absorbance at 254 nm for different concentration of NaNO3 . . . . . . . . . . . 160Table D.5 Absorbance at 185 nm for different concentration of Nordic NOM . . . . . . . 161Table D.6 Absorbance at 254 nm for different concentration of Nordic NOM . . . . . . . 162Table D.7 Absorbance at 185 nm for different concentration of HSO4− . . . . . . . . . . . 162xiList of FiguresFigure 4.1 Batch experimental setup a) and differential reactor b). The batch reactor con-sisted of a sparging beaker, a pump, a flowmeter, the differential reactor andUV lamps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31Figure 4.2 Flow through reactor experimental setup; the reactor consisted of a annularreactor (a), a pump (b), a feed storage tank (c), and a treated water tank d). . . . 32Figure 4.3 UV collimated beam set-up (a) and cross section of the UV collimated beam(b); the collimated beam consisted of a UV lamp, a reactor chamber and a stirrer. 34Figure 4.4 Diagram of the collimated beam and of the top of the PVC enclosure. (1) motor,(2) reaction vessel, (3) stirrer (4) head of the enclosure, (5) enclosure, (6) VUVlamp, (7) Teflon cylinder (8) Orings, (9) quartz sleeve, (10) PVC head of theenclosure, (11) optical filter, and (12) Suprasil quartz. . . . . . . . . . . . . . . 37Figure 4.5 Radial distribution of radiation on the surface of the reaction vessel. . . . . . . 37Figure 4.6 Angular distribution of radiation on the surface of the reaction vessel. . . . . . 38Figure 5.1 SEM micrograph of photocatalyst D. The micrograph shows important frac-tures on the surface of the catalyst. . . . . . . . . . . . . . . . . . . . . . . . . 48Figure 5.2 SEM micrograph of photocatalyst E. The micrograph shows an higher homo-geneity compared to the one of catalyst D. . . . . . . . . . . . . . . . . . . . 49Figure 5.3 XRD of the five different coatings. The figure shows hat catalyst B is amor-phous and catalyst A, B, D and E are a a mixture of rutile and anatase. . . . . . 49Figure 5.4 UV-VIS spectra of quartz, photocatalysts A, B and C. . . . . . . . . . . . . . . 50Figure 5.5 Photocatalytic activity of the five different photocatalysts for the degradation of0.1 ppm 2,4-D in millipore water. Error bars represent the standard deviationsof three replicates samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51Figure 5.6 Attrition tests conducted with the photocatalyst D. The degradation of 0.1 ppmof 2,4 D in millipore water was followed. The curves show the photocatalyticactivity of the fresh photocatalyst and after 24 h of H2O recycling. Error barsrepresent the standard deviations of three replicates samples. . . . . . . . . . . 53xiiFigure 5.7 Attrition tests conducted with the photocatalyst E. The degradation of 0.1 ppmof 2,4 D in millipore water was followed. The curves show the photocatalyticactivity of the fresh photocatalyst and after 24 h of H2O recycling.Error barsrepresent the standard deviations of three replicates samples. . . . . . . . . . . 54Figure 5.8 Apparent first-order rate constants for the degradation of 2,4-D obtained withphotocatalyst D and photocatalyst E after repeated photocatalytic experiments.Error bars represent the standard deviations of three replicates samples. . . . . 55Figure 5.9 2,4 D conversion for photocatalysis/UV, VUV and P25 slurry/UV with milli-pore water and a flowrate of 1 L/min. . . . . . . . . . . . . . . . . . . . . . . 57Figure 5.10 2,4 D conversion for photocatalysis/UV, and VUV with surface water and aflowrate of 0.25 L/min. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Figure 5.11 2,4 D conversion for VUV with surface water and millipore water. . . . . . . . 58Figure 6.1 Water absorbance for different cell path length . . . . . . . . . . . . . . . . . 64Figure 6.2 Apparent first order kinetic constant for the degradation of atrazine obtainedwith H2O2/UV and VUV.The fuence rates were 0.33mW/cm2 for 254 nm and0.06 mW/cm2 for 185 nm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70Figure 6.3 Photolysis of atrazine with VUV and UV. The fuence rates were 0.03mW/cm2and 0.29mW/cm2, respectively. Error bars represent the standard deviations ofthree replicates samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72Figure 6.4 First order rate constant for the degradation of atrazine using UV/H2O2 in thewith different concentrations of NaF. The fuence rate was 0.29mW/cm2. Errorbars represent the standard deviations of three replicates samples. . . . . . . . 76Figure 6.5 First order rate constant for the degradation of atrazine using VUV with dif-ferent concentrations of NaF. The fuence rates was 0.03mW/cm2. Error barsrepresent the standard deviations of three replicates samples. . . . . . . . . . . 77Figure 6.6 First order constant for the degradation of atrazine using UV/H2O2 in the pres-ence of different concentration of nitrate. The fuence rate was 0.31mW/cm2.Error bars represent the standard deviations of three replicates samples. . . . . 78Figure 6.7 First order constant for the degradation of atrazine using UV in the presenceof different concentration of nitrate. The fuence rate was 0.29mW/cm2. Errorbars represent the standard deviations of three replicates samples. . . . . . . . 79Figure 6.8 First order rate constant for the degradation of atrazine using VUV with dif-ferent concentrations of nitrate. The fuence rate was 0.03mW/cm2. Error barsrepresent the standard deviations of three replicates samples. . . . . . . . . . . 80xiiiFigure 6.9 First order constant for the degradation of atrazine using UV/H2O2 in the pres-ence of different concentration of bicarbonate. The fuence rate was 0.29mW/cm2.Error bars represent the standard deviations of three replicates samples. . . . . 81Figure 6.10 First order constant for the degradation of atrazine using VUV in the presenceof different concentration of bicarbonate. The fuence rate was 0.03mW/cm2.Error bars represent the standard deviations of three replicates samples. . . . . 82Figure 6.11 First order constant for the degradation of atrazine using UV/H2O2 in the pres-ence of different concentration of sulfate. The fuence rate was 0.29mW/cm2.Error bars represent the standard deviations of three replicates samples. . . . . 83Figure 6.12 First order constant for the degradation of atrazine using VUV in the presenceof different concentration of sulfate. The fuence rate was 0.05mW/cm2 Errorbars represent the standard deviations of three replicates samples. . . . . . . . 84Figure 7.1 Absorption spectra of 9 ppm of Suwannee river NOM and of Nordic NOM. . . 92Figure 7.2 First order rate constant for the degradation of atrazine using UV/H2O2 withvarious concentrations of NOM Suwannee river, and Nordic reservoir. Thefluence rate was 0.29mW/cm2. Error bars represent the standard deviations ofthree replicates samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Figure 7.3 First order rate constant for the degradation of atrazine using VUV with variousconcentrations of NOM (Suwannee river, and Nordic reservoir) and Methanol.The fluence rate was 0.03mW/cm2. Error bars represent the standard deviationsof three replicates samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99Figure 8.1 First order rate constant for the degradation of phenol using UV/H2O2 with var-ious concentrations of chloride ion. The fuence rate was 0.25mW/cm2. Errorbars represent the standard deviations of three replicates samples. . . . . . . . 109Figure 8.2 Schematic decomposition path of aqueous phenol [1]. . . . . . . . . . . . . . . 110Figure 8.3 Phenol degradation (0.1 ppm) and byproducts formation during the UV/H2O2oxidation process. Error bars represent the standard deviations of three repli-cates samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111Figure 8.4 Distribution of chlorinated by-products formed during UV/H2O2 at variouschloride concentrations, from 0 to 1.5 mM. Error bars represent the standarddeviations of three replicates samples. . . . . . . . . . . . . . . . . . . . . . . 113Figure 8.5 First order rate constant for the degradation of phenol using VUV with variousconcentrations of chloride ion. The fuence rate was 0.06mW/cm2. Error barsrepresent the standard deviations of three replicates samples. . . . . . . . . . . 115Figure 8.6 Phenol degradation (0.1 ppm) and byproducts formation during the VUV pro-cess. Error bars represent the standard deviations of three replicates samples. . 116xivFigure 8.7 Distribution of chlorinated by-products formed during the VUV process at var-ious chloride concentrations, from 0 to 1.5 mM. . . . . . . . . . . . . . . . . . 117Figure A.1 Degradation of 2,4-D in the presence of photocatalyst E after repeated photo-catalytic experiments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140Figure A.2 Degradation of 2,4-D in the presence of photocatalyst D after repeated photo-catalytic experiments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140Figure B.1 Degradation of atrazine in the presence of different concentration of NaF duringUV/ceH2O2 treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145Figure B.2 Degradation of atrazine in the presence of different concentration of NaF duringVUV treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145Figure B.3 Degradation of atrazine in the presence of different concentration of NaNO3during UV treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146Figure B.4 Degradation of atrazine in the presence of different concentration of NaNO3during UV/H2O2 treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . 146Figure B.5 Degradation of atrazine in the presence of different concentration of NaNO3during VUV treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147Figure B.6 Degradation of atrazine in the presence of different concentration of NaHCO3during UV/H2O2 treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . 147Figure B.7 Degradation of atrazine in the presence of different concentration of NaHCO3during VUV treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148Figure B.8 Degradation of atrazine in the presence of different concentration of NaHSO3during VUV treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148Figure B.9 Degradation of atrazine in the presence of different concentration of NaHSO3during VUV treatment. In the experiments with NaHSO3 1 ppm of methanolwas added in order to slow the degradation rate and have more stable measure-ments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149Figure B.10 Degradation of atrazine in the presence of different concentration of Cl– duringUV/H2O2 treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150Figure B.11 Degradation of atrazine in the presence of different concentration of Cl– duringVUV treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150Figure C.1 Degradation of atrazine in the presence of different concentration of NordicNOM during UV/H2O2 treatment. . . . . . . . . . . . . . . . . . . . . . . . . 152Figure C.2 Degradation of atrazine in the presence of different concentration of NordicNOM during VUV treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . 152xvFigure C.3 Degradation of atrazine in the presence of different concentration of Suwanneeriver NOM during UVH2O2 treatment. . . . . . . . . . . . . . . . . . . . . . . 153Figure C.4 Degradation of atrazine in the presence of different concentration of Suwanneeriver during VUV treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . 153Figure D.1 Absorbance of sodium chloride at 185 nm at different sodium chloride concen-trations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155Figure D.2 Absorbance of sodium bicarbonate at 185 nm at different sodium bicarbonateconcentrations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155Figure D.3 Absorbance of sodium nitrate at 185 nm at different sodium nitrate concentrations.156Figure D.4 Absorbance of sodium nitrate at 185 nm at different sodium nitrate concentrations.156Figure D.5 Absorbance of sodium sulfate at 185 nm at different sodium sulfate concentra-tions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157Figure D.6 Absorbance of Nordic NOM at 185 nm at different TOC concentrations. . . . . 157Figure D.7 Absorbance of Nordic NOM at 185 nm at different TOC concentrations. . . . . 158Figure D.8 Absorbance of Suwannee NOM at 185 nm at different TOC concentrations. . . 158Figure D.9 Absorbance of Suwannee NOM at 185 nm at different TOC concentrations. . . 159xviGlossary2,4-D 2,4-Dichlorophenoxyacetic acid4-CP 4-chlorophenol4-NP 4-nitrophenolA Surface of the reaction vessel (cm2)C Concentration (M−1)CBZ CarbazepineDFC DiflonacDOC Dissolved organic carbonF Fluence mJ cm−2G Incident radiation mW cm−2HPLC High Performance Liquid CromatographyI Specific spectral radiation intensity mWcm−2s−1IC Ionic ChromatographyICP/OES Inductively Coupled Plasma Optical Emission SpectrometryK Apparent kinetic constant M−1s−1LP Low pressure amalgam lampMP Medium pressure mercury lampMTBE Methyl-tert-butil etherMW Molecular weightxviiNDMA N-nitrosodimethylamineNOM Natural Organic MatterSMX sulfamethoxazoloneSUVA Specific ultraviolet absorbanceTHMFP Trihalomethane formation potentialTOC Total organic carbonVUV Vacuum UVV Volume of the reaction vessel (L)Z Path lenght[Greek alphabets]α Absorbance coefficient of the propagating mediumε Molar absorption Coefficientζ Fraction of photons absorbedφ Quantum yield[Superscripts and subscripts]0 Initial condition254 254 nm185 185 nmAQ AqueousATZ AtrazineCON ConsumedGEN GeneratedH2O WaterI i specieMEHO MethanolxviiiPHOT PhotolysisRV Reaction vesselxixChapter 1BackgroundMost population, who lives in middle size and large urban centers, take drinking water for granted.However, a large number of small rural communities have difficulties to access drinkable water ondaily basis, putting their health at risk for water born diseases. Given the remote settings, conven-tional water treatments are not suitable, since they require intense energy, high cost and experiencedoperators. For this reason, often, these communities rely on outdated treatment technologies. Theresult is that many communities live in a situation with frequent boil water advisories. Boiling wa-ter can solve the problem related to microbial contamination, but it cannot remove heavy metals,organics and taste and odor compounds. For these reasons, there is the need to develop effectiveand simple technologies that can be installed in remote locations and that can provide safe water,removing micropollutants and taste and odor compounds.1.1 Problem of micropollutants in drinking waterMicropollutants are contaminants which exist in trace amounts (ng/L, pg/L) but they pose poten-tially adverse health influences on humans. These emerging or new unregulated contaminants inwater supplies have become an environmental problem. Micropollutants can have severals originsand they can cause various problems [2]. For example, they can be industrial products such as ph-thalates and polychlorinated biphenyls: in this case the problem associated with those compoundsis the biomagnifications and the long range transportation [3]. These chemicals are slowly degraded1and therefore, they can be transported by water or air to locations far from their source. Micropol-lutants can also be products used in everyday life such as detergents, pharmaceuticals, hormones:these chemicals could be responsible for bacterial resistance or feminization of fish [4]. With mi-cropollutants another class is biocides: some examples are atrazine and Dichlorvos which havebeen proven to bring toxic effects and create persistent metabolites [5]. In addition, micropollutantscan have a geogenic source and they can be natural chemicals. Examples are heavy metals, cyan-otoxins, human hormones: all these chemicals can cause drinking water quality problems due totheir potentials negative effects on human health [6]. Another category of micropollutants are thedisinfection byproducts such as trihalomethanes and haloacetic acid [7].The source, behavior, and treatments of few macropollutants such as natural organic matter(NOM) and salts occurring at µg/L and mg/L is well understood [8]. However, it is far moredifficult to assess the influence of the thousands of micropollutants that may be present at lowconcentrations. These chemicals have been found not only in industrialized areas, but also in moreremote environments because of their low degradability.Because of their physico-chemical properties (high water solubility, often poor biodegradabil-ity, and low concentrations, usually in part per billion or trillion), most current wastewater treatmentplants are not designed to treat micropollutants [9]. For this reason, so-called tertiary treatmentssuch as carbon adsorption, and reverse osmosis are added to the system. These treatments, how-ever, are expensive and not able to degrade the contaminants; they are able only to transfer thecontaminants from one phase to another leading to waste generation and disposal issues. Hence,mitigation technologies to reduce the impacts of micropollutants require further development. Thewater supply industry is faced with a particularly difficult problem: low level (sub ppb to ppt level)of micropollutants must be inexpensively removed from large volumes of raw water.1.2 Presence of natural organic matter (NOM) in raw surface waterNOM consists of humic substances, humic and fulvic acid, and non-humic substances such as pro-teins, amino acids, sugars and polysaccharides. The chemical characteristics of NOM are influencednot only by the source materials, but also by the biogeochemical processes involved in carbon cy-2cling within the terrestrial and aquatic system. The heterogeneity of humic substances derives fromtheir various molecular sizes and from their various chemical structures. The dissolved organiccarbon (DOC) is an operational classification and it consists of compounds below 0.45 microme-ters. In water it typically consists of 90% fulvic acid and 10% humic acid [10]. Aquatic humicsubstances have a moderate aromatic character, and contain primarily carboxyl groups, phenolicgroups, alcohol and metoxyl groups, aldehydes and ketones. Characterization of the functionalgroups is important to understand their reactivity with oxidants and other species dissolved in thewater.The presence of NOM in water could be troublesome for water treatment processes. NOMcauses membrane fouling, absorbs ultraviolet (UV) irradiation, scavenges oxidants, among others.Further, NOM is a precursor of chlorination disinfection byproducts such as trihalomethanes andhaloacetic acids [11] and may induce bacteria regrowth in the distribution system. Therefore, it isimportant to remove NOM from the source water before applying various treatment options and/orsending the water into the distribution system. Several processes have been used to remove NOMsuch as coagulation, which is the most well-established method of removing NOM.More recently some studies have been carried out on the coupling of various processes. Forexample, UV or VacuumUV (VUV) radiation photooxidation followed by biological treatments hasbeen an option under investigation, because pre-oxidation promotes the formation of biodegradableintermediates which can allow microbes to mineralize these oxidation products to CO2 [12].1.3 Conventional water treatments for drinking waterThe combination of coagulation, sedimentation, filtration and disinfection is used to provide cleanand safe drinking water to the public [13] and is considered the so-called conventional water treat-ments that could reduce the concentration of NOM substantially. Coagulants, which may be alu-minum or iron based, are chemicals that can be added to water to induce dissolved and colloidalspecies to agglomerate into larger particles known as flocs. These flocs are removed in a clarifica-tion step, which may be based on gravity or buoyancy. The effectiveness of coagulation to removeNOM depends primarily on the many variables of coagulation processes including the coagulant3type, pH, coagulant dose, flocculation time, and NOM characteristics. Other technologies used forthe removal of NOM are membrane filtration, adsorption processes (granulated activated carbon),ion exchange resins, and advanced oxidations processes such as ozonation (O3), UV/H2O2, UV/O3,and Vacuum UV radiation ( VUV) [14].Conventional methods of water disinfection and treatment can address most problems relatedto water quality including NOM. However, these treatments are often chemically, energetically,and operationally intensive, focused on large scale applications. In addition, such treatments canproduce residuals resulting from the process (sludge, toxic wastes) that can add to the problem ofcontamination and salting of freshwater sources [15]. Furthermore, conventional water treatmentsare not suitable for the removal of micropollutants [16]. To address this emerging problems, ter-tiary treatments are added to the process: examples of which being carbon adsorption and reverseosmosis [17]. These technologies are however expensive and bring additional problems, associatedwith the transfer of contaminants from one phase to another or generate some concentrated wasterequiring further treatments.1.4 Advanced oxidation processes (AOPs)Most organic compounds are resistant to conventional chemical and biological water treatments.For this reason, in the past years, new technologies have been developed to achieve the require-ments of stringent standards for those substances exerting potentially toxic effects. Among thesetechnologies, AOPs will probably constitute the most promising options and have been viewed fa-vorably by industries and municipalities. AOPs, although using several reacting systems, are allcharacterized by the same chemical feature: production of hydroxyl radicals (HO•). HO radicalsare extremely reactive species, they react with most organic molecules at rate constants in the orderof 106-109 M−1s−1 and their steady state concentrations are between 10−10 and 10−12 M. Hydroxylradicals are also characterized by their little selectivity, a useful attribute for an oxidant used inwater treatment. The principal reaction pathways of HO• with organic compounds include hydro-gen abstraction from aliphatic carbon, addition to double bonds and aromatic rings, and electrontransfer. These reactions generate organic radicals as transient intermediates, which then undergo4further reactions, resulting in final products corresponding to the net oxidative degradation of thestarting molecule. AOPs can offer various possible ways for HO radicals production thus allowinga better versatility with the specific treatment requirements. Hydroxyl radicals can be generatedby various approaches such as: Fenton reaction Fe2+/ H2O2), Photo-Fenton reaction Fe3+/H2O2/ hν), UV-hydrogen peroxide (UV/ H2O2), UV-ozone (O3/UV), ultrasound, Vacuum UV (VUV)radiation, and photocatalysis (UV/TiO2).Another advantage is that AOPs usually operate at or near ambient temperature and pressure,making the process easier to control. One disadvantage of these AOPs is that, ideally the oxidationcan cause total mineralization, but in reality partly oxidized byproducts which can be more toxicsthan the initial compounds are formed. Another problem lies in the high cost of reagents such asozone, hydrogen peroxide or energy-light sources like ultraviolet light. However, for some AOPs,using solar radiation as an energy source can reduce costs.AOPs have found applications as diverse as groundwater treatment, soil remediation, munici-pal wastewater sludge conditioning, production of ultrapure water and volatile organic compoundstreatment, and odor control. Depending on the properties of the waste stream to be treated and thegoal of the treatment, AOPs can be employed either alone or coupled with other physicochemicaland biological processes.While AOPs (e.g. UV/H2O2) have found many applications in the water treatment industry forthe removal of micropollutants, there are still challenges and issues with them. One such challengeis the effect of water characteristics on their efficiency. This research has tried to address thesechallenges by focusing on two emerging AOPs: VUV radiation and UV/TiO2 and studying theeffects of water matrix (inorganic ions and NOM) on their efficiency in degrading micropollutants.1.5 Thesis layoutThis thesis is organized in several chapters:Chapter One gives a general background on the challenge of water treatments and water con-tamination.Chapter Two presents a literature review of the advanced oxidation processes and the effects5of water characteristic on these AOPs. The literature on this topic is analyzed and the knowledgegaps are highlighted.Chapter Three presents the overall and the specific objectives of this study.Chapter Four describes the experimental setups and methodologies that are commonly usedfor achieving different objectives of the research. Setups and procedures specific to a particularobjective or activity are presented in the respective chapters.Chapter Five contributes to meeting objectives of developing a TiO2 heterogeneous catalystand of assessing the effects of the combination of photocatalysis with VUV.Chapters Six, Seven and Eight contribute primarily to meeting the objective of the effect ofwater matrix on the efficacy of the VUV process. In particular, in chapter six the effects of inorganicions are studied, in chapter seven the ones of NOM and eight presents the effects of chloride.Chapter Nine presents the overall conclusions and the recommendations for future work.6Chapter 2Literature review2.1 Hydrogen peroxide/UV processUV/H2O2 is one of the most widely studied AOP which has also been commercialized for thedegradation of micropollutants. This process includes H2O2 injection to the water and irradiating itwith UV light (200 to 280 nm). During this process, ultraviolet radiation is used to cleave the O-Obond in the hydrogen peroxide and generate hydroxyl radicals. The overall quantum yield of thisreaction is 1.0 for UV below 300 nm [18].H2O2+hν → 2HO• (2.1)The HO radicals produced are then able to react with the substrate leading to short chain acidsbyproducts or ideally to mineralization:HO•+Substrate→ Products (2.2)In addition, the HO radicals produced can react again with H2O2:HO•+H2O2 → HO2•+H2O (2.3)7High concentrations of H2O2 are needed for the process because H2O2 has a low molar absorp-tion coefficient at 254 nm (19.9 M−1 cm−1) [19] and for this reason to produce enough HO radicals,relatively high doses are required. On the other hand, H2O2 itself can scavenge HO radicals so highconcentration of hydrogen peroxide can reduce the effectiveness of the process. In addition, highconcentration of H2O2 can reduce the penetration of photons reducing the direct photolysis of thetarget contaminant. For all these reasons, the ideal concentration of H2O2 is in the range from 10 to50 mg/L [20]. In addition, during the H2O2/UV process other reactive radicals are formed, such assuperoxide radicals (HO2•). Superoxide radicals are generated with dissolved oxygen according tothe reaction [21]:2HO•+O2 → 2HO2• (2.4)The radicals formed can participate in the following range of reactions leading to the scavengingof HO radicals and to the formation of H2O2 [21]:HO•+HO2• → H2O+O2 (k = 6.6×109M−1s−1) (2.5)HO•+O2−•+→ HO−+O2 (k = 7.0×109M−1s−1) (2.6)HO•+HO• → H2O2 (k = 4.0×109M−1s−1) (2.7)HO2•+HO2• → H2O2+O2 (k = 2.6×109M−1s−1) (2.8)O2−•+HO2−•+H2O+→ H2O2+O2+HO− (k = 9.7×107M−1s−1) (2.9)The application of UV/H2O2 for commercial drinking water treatment involves a continuousflow system including one or more UV reactors and a H2O2 dosing components. The two keyparameters in the process are the UV fluence and the H2O2 concentration. UV fluence is the totalradiant energy of all wavelengths received by an infinitesimally small sphere. it is the product ofthe fluence rate and the exposure time [22]. The lamp technologies commonly applied in drinkingwater UV/H2O2 applications are the next: low-pressure (LP) amalgam lamps and medium-pressure8mercury (Hg) lamps (MP). LP Hg and LP amalgam lamps emit two wavelengths 185 nm and 254nm. LP lamps are constructed of a quartz envelope that permits transmission of only photons of254 nm and above [23]. Medium Pressure Lamps have higher electrical power input compared withLPs: it is important to point out that UV radiation, in the range of 200-300 nm, is absorbed by DNAdisrupting its structure and leading to the deactivation of cell. For this reason, these types of lampsare used intensively for the disinfection of water, substituting chlorine methods [24]. The rangeof operating parameters for commercial UV/H2O2 installation for the treatment of trace organicpollutants are the following: an initial H2O2 concentration up to 15 mgL−1 and fluences up to 1500mJ cm−2.The commercial application of UV/H2O2 in drinking water treatment processes dates to theearly 1990s. Recently, research has focused on the efficacy of UV/H2O2 on the removal of specificmicropollutants. Kruithof et al. [25] studied the degradation of some organic contaminants in realsurface water with a UV dose of 540 mJ/cm2 (about 0.5 kWh/m3) and 6 mg/L H2O2. Under thoseconditions, pesticides (atrazine), N-Nitrosodimethylamine (NDMA), methyl-t-butil etere (MTBE),dioxane, endocrine disruptors (bisphenol A), microcystine and pharmaceuticals (diclofenac, ibupro-fen) could be removed by up to the required 80%. In addition, the effectiveness of H2O2/UV wasstudied for the degradation of other contaminants such as phenol [26], taste and odors compoundsand methyl tert- butyl ether [27]. Some research focused also on the treatment of taste and odorcausing compounds such as methyisoborneol (MIB) in water. UV/H2O2 oxidized more than 70%of this compound at a UV fluence of 1000 mJcm2. The study was conducted with both MP UVand LP UV showing that MP UV consistently performed better than LP UV for MIB oxidation[28]. Zhang et al. [29] studied the photodegradation of 4-nitrophenol (4-NP) with UV/H2O2 usinga LP UV lamp placed in a double quartz sleeves annular photo reactor. The photon flux enteringthe reactor was 3.14 x10−6 Einstein−1. With these conditions 98% of 4-NP was removed within 12minutes, and thus UV/H2O2, in this case, was suggested suitable as tertiary treatments.There are numerous researches on the degradation of different organic compounds with UV/H2O2such as [30], [31], [32], [33]. It is difficult to compare the results among these studies since the con-ditions used such as source of irradiations, photo reactor geometry, and initial concentration of the9target compounds were different. However, some considerations can be made regarding the con-centration of H2O2 that has been used. Vogna et al. [30] and Lekkerkerker et al. [33] in particular,obtained good removal of the substrate with concentration of H2O2 among 5-10 mg/L which isfeasible for industrial applications. In particular, Vogna et al. [30] studied the degradation of car-bamazepine with UV/H2O2. The oxidation treatment caused an effective removal of the drug. Thesubstrate (aqueous solution of 2.0 mM) was completely removed after 4 min treatment, and 35%removal of organic carbon (TOC) was obtained. Carbamazepine or benzoic acid solutions in waterwere irradiated with a nominal 17 W low-pressure mercury monochromatic lamp emitting at 254nm (Helios Italquartz) in a 0.420 L photoreactor. The power output of the lamp was 2.7x10−6 Es−1.Lekkerkerker et al. [33] studied the transformation of atrazine (ATZ), carbazepine (CBZ), di-flonac (DFC) and sulfamethoxazole (SMX) by UV/H2O2 treatment. Irradiations were carried outin a collimated beam. The treatments employed a range of doses between 300 to 700 mJ/cm2 and aconcentration of H2O2 from 0 to 10 mg L−1. The results showed that DFC and SMX were removedby 100% after an irradiation of 230 mJ/cm2 with a low pressure lamp. The addition of H2O2, in thiscase, had little effect to the direct photodegradation. For ATZ, the addition of H2O2 had an effect onits removal. The addition of 5-10 mg/L of H2O2 increased the removal efficiency by 10-15%. CBZwas poorly removed with each doses and also with the addition of different concentration (from 0to 10 mg/L) of H2O2. In general, it is important to point out that the H2O2 dose will depend on theDOC load of water.On the other hand Johnson et al. [31] and Li et al. [32], although they showed a very appealingremoval of the substrate, used concentration of H2O2 too high for any industrial applications ( 25and 100 mg/L respectively). Johnson et al. [31] studied the metronidazole degradation by UV/H2O2process. The irradiation experiments were performed in a UV collimated beam equipped with a LPUV lamp emitting at 253.7 nm. The initial concentration of metronidazole was 1 mg L−1 and theone of H2O2 was 25 mg L−1. The rate constant was found to be 1.98 x 109 M−1 s−1, and thus theefficiency was found to be high, however, the concentration of H2O2 was too high for any industrialapplications (the concentration used is usually 10 ppm).10Li et al. [32] explored the photochemical degradation of typical herbicide simazine by UV/H2O2in an aqueous solution. The experiments were performed in a 20 L double cylindrical stainless reac-tor. The annular reactor was equipped with an immersible low pressure lamp with an output at 254nm. The concentration of simazine was 100 µg L−1 and the light intensity was 45x10−6 einsteinL−1 s−1. With a concentration of H2O2 of 100 mg L−1 the removal was 100%. The study showedcomplete removal, however the concentration of H2O2 was too high for any industrial application.These studies showed that UV/H2O2 can successfully remove micropollutants from water howeveroften the H2O2 concentration used in these studies is not feasible for any industrial applications.In addition all these researches studied the removal of micropollutants in Millipore water, and theeffect of real raw surface water on the efficacy of this AOP was not analyzed.UV/H2O2 can be also used for the removal of NOM: some studies were conducted in order toassess the capability of UV/H2O2 to degrade NOM. Kleiser and Frimmel [34] exposed river water,with DOC of 2.3 mg L−1 and to a LP lamp in the presence of 4, 8, and 16 mg L−1 H2O2. Underthese irradiation times DOC removal was minimal. The trihalomethane formation potential (THM-FP) was observed after irradiation times, of 100 min. However, THM-FP dropped at an irradiationtime longer than 100 min and that was attributed to the observed mineralization of DOC. Fluenceswere not reported, but it is likely that the oxidation conditions at 100 min were similar to thosetypically found in drinking water applications.Wang et al. [35] evaluated LP UV/H2O2 oxidation for the degradation of NOM in water. About90% of humic acid was removed after 30 minutes of irradiation in the 10 L batch reactor using a450 W high-pressure Hg vapor lamp in the presence of 0.01% H2O2. Specific UV fluences werenot reported but due to the observation of TOC destruction it was likely fluences were very high(i.e., far greater than those in commercial applications).UV/H2O2 AOP has proved to be an effective treatment for the removal of micropollutants, butit has several disadvantages. First H2O2 has to be added and removed downstream or it needs to becoupled with BAC (Biological activated carbon) treatment: this increases the cost of treatment andit can make the technology complicated. Second, complete mineralization is rarely achieved, andsometimes the by-products created are more toxic than the initial contaminants. In addition, many11organics have a high absorption coefficient in the UV (254 nm) and for this reason they can act asinner filter therefore fewer H2O2 molecules are photolyzed. On the other hand, other technologiesshould be applied for the removal of NOM, since THMFP was observed and the TOC removal tookplace at very high fluences.2.2 PhotocatalysisPhotocatalysis with a semiconductor is another AOP which has shown great potentials for the re-moval of pollutants [36]. Photocatalysis with semiconductor is a process where a semiconductorcatalyst is activated by UV radiation. When a semiconductor absorbs photon with energy higherthan its band gap, an electron/hole couple (e−/h+) is generated [37]. The electron/hole couple caneither be used to generate electricity in photovoltaic cells or to drive a chemical reaction (photocat-alytic process). One of the most widely studied photocatalysts in environmental decontaminationis titanium dioxide, TiO2. It is used to degrade a large variety of organics, which can be totallydegraded and mineralized to CO2, H2O, and harmless inorganic anions. This performance is at-tributed to titania’s highly oxidizing holes that can react directly with the adsorbed substrates oroxidize water to give HO•. TiO2 can exist in different polymorphous: brookite, rutile and anatase.Anatase is the crystalline structure that shows greater photocatalytic activity for most reactions [38].It has been suggested that this increased photoreactivity is because of anatase’s slightly higher fermilevel, lower capacity to adsorb oxygen and higher hydroxylation [39].When TiO2 absorbs a photon with an energy equal to or higher than its band gap (Eg), anelectron and hole pair are generated:hν ≥ Eg+TiO2 → e−+h+ (2.10)The hole formed (h+) can then generate HO•:h++H2O→ HO•+H+ (2.11)12Studies on the degradation rate of various pesticides with various TiO2 catalysts have beenperformed. The degradation rates of dichlorovoros and phosphamidon in a Degussa p25 slurry (1gL−1) at 0.5 mM and 0.25 mM, respectively, found to be 0.020 and 0.02 x10−3 L−1 min−1 [40].Gelover et al. [41] studied the degradation of chlorophenol using degussa P25 as catalyst. Afterthree hours of irradiation the final dissolved concentration of 4-chlorophenol was 24% of its initialvalue.In addition, laboratory scale studies with artificial irradiation source and small volumes of waterhave shown complete contaminant disappearance for a range of pesticides such as triazines [42],lindane [43], methyl paration [44] at various concentrations.Some authors have studied the degradation of taste and odors compounds, in particular geosmin,with suspensions of Degussa P25 [45]. The suspension was 1% P25 and the samples were irradi-ated with a xenon lamp with a spectral output from 330-500 nm. The photonic output was found2.15x10−5 einstein min−1. Geosmin was degraded with more than 99% decomposition achieved inone hour.Another group [42] studied the degradation of atrazine by using Degussa P25 TiO2 as photocatalyst under simulated solar light. The irradiation was carried out by 1500 W xenon lamp andthe flux in the range of 340-545 nm, was 5.8x10−5 einstein min−1 cm−2.The results showed thatin the part-per-million domain and beside TiO2 slurries and illumination, s triazine herbicides wererapidly degraded: only a few minutes were needed to reduce an initial concentration of 2 ppb downto less than 0.1 ppb. Other studies showed that the photocatalytic activity was dependent to the lightintensity [46], [47].Dalrymple et al. [48] published a review on the removal of pharmaceutical and endocrinedisrupting compounds from wastewater by Degussa P25 TiO2 photocatalysis. The contaminantsincluded were approximately 30. The rate constants were between 1 to 11.2 x10−6 M−1 s−1. Fur-thermore it was found that the addition of catalyst has shown to increase the degradation rate withina range of concentration, from 0.1 to 4 gL−1.All these studies reported the degradation of micropollutants obtained in the presence of TiO2P25 Degussa. The results were promising, meaning that TiO2 can efficiently remove organics,13however, the TiO2 was in the form of a concentrate suspension which needs to be removed down-stream and this an important disadvantage. For this reason the process is not easily manageable forindustrial application, since the filtration of slurry can be challenging and expensive.Other researchers worked with immobilized photocatalysts which can be used in real applica-tion since it does not require to be removed downstream. For example, Carbonaro et al [49] andZabar et al [50] studied the degradation of different organics. The conditions used in the studies,such as the type of photocatalysts, the source of irradiation, the type of photoreactor and the initialconcentration of the micropollutants were different. However in both cases, the removal rate of theorganics was lower compared to the one obtained with the Degussa P25 suspension (46% and 99%removal respectively). Carbonaro’ s group in particular studied the continuous-flow photocatalytictreatment of pharmaceutical micropollutants. The activity of immobilized photocatalyst thin filmswas studied in a serpentine-pattern plug flow reactor. The reactor was designed with five channels,each with the width of 1 standard microscope slide (25 mm) and length of 5 slides (375 mm), fora total of 25 film-coated slides placed along the entire reactor flow path. Five 15 W 18 inches longUV-A lamps with spectrum centered at 365 nm were suspended and centered over each of the fivechannels. The degradation of four pharmaceutical micropollutants (iopromide, acetaminophen, sul-famethoxazole, and carbamazepine) was monitored. The kobs values were found to be 0.97 h−1 foracetaminophen, 0.50 h−1 for carbamazepine, 0.49 h−1for iopromide and 0.79 h−1 for sulfamethox-azole.Zabar et al [50], on the other hand, describes the photocatalytic degradation of 6-chloronicotinicacid (6CNA). Photocatalytic experiments were performed using immobilized titanium dioxide onsix glass slides in a spinning basket inside a photocatalytic quartz cell. Three low-pressure mercuryfluorescent lamps were used as a UVA (315 to 400 nm) radiation source (CLEO 20 W, 438 mm x 26mm, Phillips; broad maximum at 355 nm). The photon flux in the cell was determined to be 2.3 x10−5 Einstein L−1 s−1. The degradation within 120 min, obeyed first-order kinetics. The observeddisappearance rate constant was k = 0.011 min−1. Furthermore 46% mineralization was achievedwithin 120 min of irradiation.Another promising application of UV/TiO2 can be the removal of cyanotoxins [51]. The hy-14droxyl radicals, produced during TiO2/UV process reacts in the order microcystins (1.1 x 1010 M−1s−1) > cylindrospermopsin (5.5 x 109 M−1 s−1) > anatoxin (3.0 x 109 M−1 s−1) [52]. Becausethe rate constant with HO• is several orders of magnitude larger than that for other oxidants (rateconstant for ozone is 104 M−1 s−1 and for chlorine 1 M−1 s−1 or less depending on the conditions),TiO2 could be used succesfully for the inactivation of cyanotoxins, as alternative of UV/O3 andchlorinated processes.2.3 Vacuum UVVacuum UV is another AOP which was found to be a highly effective process for the degradationof micropollutants. It is an oxidant free, UV-based AOP, which relies on the formation of reactivespecies such as HO•, H•, e−aq, HO2•, O2•− through the photolysis of water by irradiating it withVUV (photons of less than 200 nm). VUV photons can be generated by several sources, the mostcommon being excimer lamps and ozone-generating low-pressure Hg lamps (VUV-Hg lamps), eachpresenting some advantages and disadvantages. Excimer lamps can emit a high powered quasi-monochromatic radiation at various wavelengths (depending on the gas in the lamp) [53]. Theelectrical efficiency is in the order of 5-40%. For water treatments, Xe2-excimer lamps, which emitat 172 nm have been proposed [54]. Because water absorptivity at 172 nm is high (550 cm−1) [55],photons are mostly absorbed in a 10 micron layer close to the lamp, with the generation of importantmass transfer resistance that lowers the efficacy of the process. VUV-Hg lamps, on the other hand,emit about 10% of the radiation at 185 nm and 80-90% radiation at 254 nm, and a low percentagein the visible range. The absorptivity of water at 185 nm is reported to be 1.80 cm−1 [19]: in thiscase, photons are absorbed in about 0.3 cm layer close to the lamp. As a result, diffusive resistancesare less significant and could potentially be minimized or eliminated by increasing the turbulencein the reactor. The problem of these lamps, however, is that only 10% of all radiation is in the VUVrange.Two of the most important reactions occuring during the VUV radiation/water systems are [56]:15• The photochemical homolysis of water:H2O+hν<200nm → HO•+H• (2.12)• The photochemical ionization of water:H2O+hν<200nm → HO•+ e−+H+ (2.13)The quantum yields at 185 nm of the reactions are 0.33 and 0.045, respectively. HO• can beinvolved in a number of reactions among them being [56] :O2+H• → HO2• (k = 2.1×1010) (2.14)HO•+HO• → H2O2 (k = 4.0×109M−1s−1) (2.15)HO•+H2O2 → HO2•+H2O (k = 2.7×107M−1s−1) (2.16)H2O2+HO2• → HO•+O2+H2O (k = 5.3×102M−1s−1) (2.17)HO2•+H2O→ H3O+O2•− (pK = 4.8) (2.18)During the VUV process the degradation of pollutants can occur through several mechanisms,with the most significant of those being the reactions between HO• and pollutants:Pollutant+HO• → intermediates→ CO2+H2O+mineral acids (2.19)Another possible reaction occurring in the system is the direct photolysis of the pollutant:Pollutant+hν254nm → products (2.20)Pollutant+hν185nm → products (2.21)Different researches have focused on the degradation of organics with VUV process. Oppen-lander et al. [57]. studied the VUV-induced oxidation of organic micropollutant (C1-C8) in ho-16mogeneous aqueous solution using a xenon-excimer (with an electrical input power of 155W) flowthrough reactor. The degradation efficiency depended on the structure of the molecules: after 80minutes of irradiation the concentration decreased from 48 to 13 mg/L in DOC. However, due tothe complexity of the reactions, further investigation concerning the TOC diminution, kinetics, themechanism of intermediary product formation and of the subsequent mineralization would be valu-able. In addition, the electrical energy dosage used in this study is not appealing (since it is toohigh) for industrial applications.Baum et al. [58] investigated the application of VUV to the degradation of chloroorganic com-pounds in a flow through reactor. The initial concentrations of trichloroethane, dichloroethene andtetrachloroethene were 103 mg/L, 25 mg/L and 0.4 mg/L respectively. After 60 minutes of VUVradiation, the concentration of chloroorganic water pollutants decreased by more than 97%. In an-other study involving citric acid and gallic acid as model pollutants, the apparent first-order rateconstants of the degradation in aqueous solution were 5.45 × 104 s−1 and 8.98 × 104 s−1, respec-tively [59]. Many of these studies were performed in a batch and with a small reactor. Hence,there was no consideration given to the amount of energy used to degrade a specific concentrationof contaminant. In addition the degradation of organics was performed preparing the samples withMillipore water hence without taking into account the effects that any compounds present in thewater matrix may have.The removal of NOM, as already mentioned, is a major objective of drinking water treatments.Several studies on the application of VUV irradiation for NOM removal have been conducted andgenerally it could be shown that VUV light increases the biodegradability and eventually mineralizeNOM: VUV was found the most effective AOPs among UV/H2O2 and UV in the degradation ofNOM. It was reported that VUV irradiation emitted by a low pressure mercury vapor lamp achieveda five times greater NOM removal, measured by the reduction of DOC concentration than UVirradiation at 254 nm [60], [12]. Imoberdorf et al. [60] used VUV radiation to degrade naturalorganic matter (NOM) in an annular reactor. It was found that after 180 min of irradiation theTOC of raw water decreased from 4.95 mg L−1 to 0.3 mg/L−1. In addition, it was found that theefficiency of the VUV process depends on initial water quality because the inorganic ions present17in water can absorb photons or scavenge HO radicals. This study showed a good removal of TOC,however the energy efficiency of the treatment was not discussed.Ratpukdi et al. [61] studied the removal of DOC with a stainless steel reactor. The reactor hada diameter of 30 cm and a height of 25 cm, and was filled with 16 L of the filtered water sampleand it was equipped with 4 VUV lamps which had have a power input of 30 W per lamp, Mixing(60 rpm) was provided by a magnetic stirring system. The initial concentration of DOC was 4.18mg/L and after 60 minutes of irrafiation the DOC was removed up to 30%.The consequence of the sequential NOM transformation under VUV irradiation for the disinfec-tion byproducts formation potential were studied by Buchanan et al. [12]. Buchanan [12] studiedthe formation of hazardous by-products resulting from the irradiation of natural organic matter withVUV. It was found that after an initial increase the trihalomethane formation potential (THMFP)was reduced and the reduction correlated well with DOC mineralization. The study showed thatduring the VUV treatment the THMFP increased with the increment of the mineralization. Thework did not provide any comparison with other AOPs regarding THMFP. On the other hand, theenergy used to observe mineralization in this study (20000 kJ/m3) was the same order as the energyrequired for reverse osmosis of salt water [62] and about 3 order of magnitude more than that forsome commercial AOPs. For this reason the use of VUV for the removal of NOM or micropollu-tants, with the current technology, cannot be use for any industrial application due to its extremelyhigh energetic cost.Other recent studies were focused on disinfection obtained with VUV radiation. The efficacyof UV and VUV disinfection of Bacillus (B.) subtilis spores in aqueous suspension at 172 nm and254 nm was evaluated [63]. A Xe2 excimer lamp and a low pressure lamp were used as irradiationsources for these two wavelengths. The first order inactivation rate constant at 172 and 254 nm were0.0023 and 0.069 cm2 mJ−1 respectively. A 2 log reduction of B. subtilis spores was reached withfluences (UV doses) of 870 and 40.4 mJ cm2 at these individual wavelengths, respectively. There-fore, for the inactivation of B. subtilis spores, VUV exposure at 172 nm is much less efficient thanexposure at 254 nm, since the deactivation of microorganism is due to the interaction of DNA with254 nm photons and not due to the attack of HO radicals. This research indicated quantitatively18that VUV at 172 nm is not practical for microorganism inactivation in water and wastewater treat-ment. On the other hand, as already mentioned, if the source of VUV radiation is a low pressure Hglamps, two wavelengths are emitted: 185 and 254 nm. In this case, the removal of micropollutantsand disinfection (obtained with 254 nm ) are achieved simultaneously.2.4 Potential of combined VUV and UV/TiO2 for micropollutantdegradationWith the mercury VUV lamps emitting 10% of radiation at 185 nm (VUV radiation) and 90% at 254nm, there is a potential to combine the VUV process with photocatalysis. The radiation at 254 nmcan be used either for water disinfection purpose or for the activation of the photocatalyst while theone at 185 nm can photolyse the water molecules for HO• formation. This approach could enhancethe overall efficiency and could use all the photons emitted by the lamps. Han et al. [64] chosepCBA as a model compound and investigated the difference between water photolysis at less than185 nm radiation and photocatalysis with irradiation at 254 nm. The pseudo first order reactionrate constant was found to be 0.036 min−1 for TiO2/UV, 0.21 min−1 for VUV and 0.23 min−1for TiO2/VUV. As shown, the rate constant was higher for the combination of TiO2 with VUV. Inanother work, Han et al. [65] measured the photocatalytic decomposition and mineralization of 4-chlorophenol (4-CP), hydroquinone and 4-nitrophenol (4-NP) in aqueous solution using two kindsof low-pressure mercury lamps: one was UV lamp emitting at 254 nm and the other was VUV lampemitting at both 254 and 185 nm. It was demonstrated that VUV irradiation led to the most efficientdegradation of the organics. No significant difference of degradation rate was observed due to TiO2catalyst under UV or VUV irradiation except the UV irradiated 4-chlorophenol solutions, whichindicated that the main degradation reaction occurred on the catalyst surface.2.5 Effect of water matrixWater matrix makes a significant contribution to the efficiency of every AOP. Two of the primaryinfluencing constituents within the natural water matrix are inorganic ions and NOM which canreact with HO radicals leaving less HO radicals available for the degradation of the target micropol-19lutants. In addition these constituents can absorb photons, acting as inner filter or being photolyzed.In the following section, the detailed effects of inorganic and NOM on AOPs are presented.2.5.1 Effect of inorganic ionsIn natural water, there are several inorganic ions which could act in various ways towards VUV andTiO2/UV. In general terms, inorganics can:• Absorb the incident radiation making the process less efficient because less photons will beavailable for the photolysis of water, and/or H2O2, or for the activation of TiO2.• Scavenge HO• reducing the concentration of HO• available for the reaction with target con-taminants.• Produce radicals when photolyzed.There are some literature reports on the effect of several inorganic ions on the efficacy of variousUV based AOPs. Usually a decrease in the efficiency has been reported with some ions [66]. It isunclear whether the efficiency is decreased because of the absorbtion of photons by those ions orbecause of the reactivity of ions toward HO• or the combination of both. HNO2, NO2− and NO3−are known to absorb radiation in the UV region of the electromagnetic spectrum. The absorbtionspectra of HNO2, NO2− and NO3− show intense pi → pi* bands in the UVC region (εHNO2,371nm∼ 2900 M−1 cm−1 , εNO2,205nm ∼ 5500 M−1 cm−1, εNO3,201nm ∼ 9900 M−1 cm−1) and weak n→ pi* at longer wavelengths [67]. Weeks et al. [19] reported the extinction coefficient of differentcompounds at 185 nm: for ethyl acetate at 0.03 mol L−1 was 158 M−1 cm−1, for H2O2 at 0.007 molL−1 was 289 M−1 cm−1 and for sulfuric acid at 0.001 mol L−1 was 186 M−1 cm−1. These resultsshow that the absorptivity of inorganic ions at 185 and 254 nm could be important and should beconsidered. At the same time, it is well established that many inorganic ions, naturally present inwater, can react with HO•. The rate constant of most of the inorganic ions with HO• is quite high(108 - 109 L mol−1 s−1) [68].On the other hand, there are some inorganic ions able to increase the efficiency of the AOPsbecause they can generate HO•. Examples are: nitrous acid/nitrite, nitrate, and bicarbonate/carbon-20ate systems. The mechanisms of photolysis of nitrous acid, nitrite, and nitrate involve photolyticpathways that result in the formation of HO• and different nitrogen species such as NO•, NO2•and ONOO−t, as primary photoproducts. The primary photoprocesses and the main subsequentreactions leading to the production of HO• may be represented as follows [69]:Nitrous acid/nitrite systems:HNO2+hν → HO•+NO• φ = 0.01 (2.22)NO2−+hν → O•−+NO• φ = 0.01 (2.23)O•−+NO•+H+ → HO•+NO• (2.24)In the nitrate system, during irradiation at wavelength< 280 nm, peroxynitrite anion (ONOO−)is formed via isomerization of NO3∗ (the excited specie of NO3)NO−3 +hν → NO3∗ (2.25)NO3∗ → ONOO−+H+ → ONOOH pK = 6.6 (2.26)→ (NO3−+H+)72%+(HO•+NO2•)28% (k = 0.9s−1) (2.27)NO3−+hν → O•−+NO2•+H2O→ 2HO•+NO2• φ = 0.099 (2.28)Quantum yields reported for nitrous acid and nitrite photoinduced production of HO• are higherthan the ones reported for nitrate [70]. In general, the quantum yields increase with decreasingwavelength, this dependence being stronger for nitrate than for nitrite and still less significant forHNO2 [71].21While NO3− exhibits little reactivity towards hydroxyl radicals, the kinetic role of HNO2 andNO2− as HO• scavengers has to be taken into account [72].NO2−+HO• → NO3H•− (k = 2.5×109M−1s−1) (2.29)NO2−+HO• → HO−+NO2• (k1 = 1×1010M−1s−1) (2.30)Another ion that can have a large impact on water photochemical process is bicarbonate. Theradical CO3•− can be produced from HO• and CO32−, HCO3• or 3DOM∗ (excited triplet state).HO•+CO32− → HO−+CO3−• (k = 3.9×108M−1s−1) (2.31)HO•+HCO3− → H2O+CO3−• (k = 8.5×106M−1s−1) (2.32)3DOM∗+CO32− → DOM−•+CO3−• (k = 1×105M−1s−1) (2.33)The radical CO3−• can react with organic contaminants in water and provide some degree ofdecontamination. For instance, the reaction rate constants between CO3−• and some pesticides suchas fenthion, atrazine, malathion and phorate are, respectively, 2.0×107 M−1 s−1 , 4.0×106 M−1 s−1,8.9×105 M−1 s−1, and 1.2×107 M−1 s−1 [73]. Furthermore, the rate constant for benzene is 3×103M−1 s−1 [74]. The radical CO3−• is less reactive than HO• toward the degradation of organiccompounds. At the same time, it is scavanged to a lesser extent by DOM in surface water. For thisreason, CO3−• can reach a higher steady-state concentration than HO•, which may compensate forits lower reactivity [75].Another ion that plays a significant role is chloride, which can absorb radiation at 185 nm and254 nm and can be photolyzed forming HO radicals [76]. Also, it can scavenge photons and HOradicals making the AOP less efficient [77]. In addition, since Cl• is formed it can react with theorganic substrate forming toxic chlorinated byproducts. The reactions in which chloride is involvedare:Cl−+HO• → Cl•+HO− (2.34)22Chloride can react with HO radicals and form chloride radicals which can react with the sub-strate and create chlorinated byproducts. On the other hand, the chloride radicals in these wayformed can be scavenged according to:Cl•+H2O→ HOCl•−+H+ (k = 1.6×107M−1s−1) (2.35)The hypochlorous radical ions can be involved in the following reactions [77]:HO•+Cl− → HOCl•− (k = 4.3×109M−1s−1) (2.36)HOCl•− → HO•+Cl− (k = 6.1×109M−1s−1) (2.37)HOCl•−+H+ → Cl•+H2O (k = 2.1×1010M−1s−1) (2.38)Cl•+H2O→ HOCl•−+H+ (k = 1.3×103M−1s−1) (2.39)As this manifold of reactions show, pH makes an important contribution to the production ofchlorine radicals and it is clear that hydroxyl radicals and chloride ions are in equilibrium with thehypochlorous radicals. It is clear that the role of chloride is complicated; on one hand it can createHO radicals, on the other it can scavenge HO radicals. One research in the literature [78] studiedthe photooxidation with UV/H2O2 of reactive orange dye with and without NaCl. The photoreactorused in the study consisted of 8 medium pressure mercury vapor lamps set in parallel emitting 365nm wavelength. With 20 mM of H2O2, 98.3% of the dye was degraded after 150 min of irradiation.With the addition of 1g/L of NaCl a small decrease was observed (5%). This behavior was explainedbased on the scavenging effect of chloride towards HO radicals.Another study [79] investigated the decolorization and degradation of commercial reactive azodyes in the presence of different inorganic ions. It was deduced that all the anions examined affectthe decolorization rate adversely but to a varying degree. The trend was somewhat different at loweranion concentration (0.01 M) when compared to a higher one (0.1 M). Specifically, at 0.01 M, theorder of inhibition measured was as follows: H2PO4 > CO32− > HCO3> NO3> Cl–> SO42−while at the higher concentration of 0.1 M, the order of inhibition measured was: H2PO4 > Cl–>23HCO3 > CO32− >SO42−> NO3. Unfortunately, the researches did not do a systematic study onthe different effect that each ion could have on the process, such as the absorption coefficient andtheir capability to scavenge HO radicals. Yuan et al. [80] investigated the effects of chloride ionon the degradation of Acid orange with sulfate radical-based advanced oxidation. It was found thatchloride at concentration higher than 5mM can increase the efficacy of the process since chlorideradicals are formed and they can assist HO radicals during the degradation of the organic.2.5.2 Effect of natural organic matter (NOM)NOM plays an important role in the degradation of micropollutants. At high concentration, NOMcan decrease the overall process efficiency. It is known that NOM absorbs radiation and attenuatesUV. Hence, the photons absorbed by NOM are no longer available for photochemical reactions.Furthermore, NOM can scavenge HO• with a very high reaction rate constant of about 1.25×108M−1 s−1 [81]. The reactions of HO•, produced by different AOPs, with NOM proceeds in threedifferent ways: i) addition of HO• to double bonds, ii) abstraction of an H-atom, which yieldscarbon centered radicals, and iii) reactions whereby HO• gets an electron for an organic substituent.The carbon centred radicals then react very rapidly with oxygen to form organic peroxyl radicals.The reactions of peroxyl radicals among themselves can lead to the production of ketones andaldehydes and/or carbon dioxide [82]. Another problem specific to heterogenous TiO2/UV is thatNOM could adsorb on the catalyst, leading to its deactivation [82].Low concentration of NOM, on the other hand, could actually accelerate the degradation oftarget contaminants due to the formation of reactive species [83]. NOM can act as a sensitizer andthe energy transfer, in the presence of oxygen, can lead to singlet oxygen formation, which is an ef-ficient oxidant for a variety of unsaturated organic compounds. Photosensitized transformations oforganic chemicals in surface water are mostly initiated through light absorption by chromophorespresent in dissolved organic material (DOM). The occurrence of several reactive photooxidants,including the hydroxyl radical, the carbonate radical, singlet molecular oxygen, and solvated elec-trons, has been documented, and photo stationary steady state concentrations of these species havebeen calculated [84]. Recently, details of the fast photooxidation of organics by short lived triplet24states of DOM have been reported. It was postulated that 3DOM∗ reacts with organics by electronabstraction and/or hydrogen transfer [85]. The reactions presented here are the ones that have beenproposed:DOM+hν → DOM∗ (2.40)DOM∗+O2 →1 O2 (2.41)DOM+hν → 1DOM∗ (2.42)1DOM∗ → 3DOM∗ (2.43)3DOM∗+P→ 3P∗ (2.44)3DOM∗+P→ POx+DOMRed (2.45)where DOM∗ is the dissolved organic matter in an excited state, 1DOM∗ and 3DOM∗ are DOM∗ ina singulet and triplet state, respectively, Pox is the oxidized substrate, 3P∗ is the substrate in a tripletexcited state, and DOMred is the DOM reduced.The impact of NOM was studied by different researchers [86], [85]. Depending on the NOMinitial concentration and on the type of irradiations, different results were obtained: NOM couldact as HO• scavenger or as a sensitizer. In particular, Wu et al. [86] studied the degradation ofmetolachlor with UV and UV/H2O2. UV irradiations were carried out using a 15 W germicidallow pressure UV lamp. The concentration of metolachlor was 2 µM and the concentration of H2O2varied from 0 to 50 mgL−1. An increase of the NOM concentration from 1.6 to 4.7 mg L−1 resultedin a decrease of the destruction rate of metolachlor. With 25 mgL−1 of H2O2 and 1.57 mgL−1 DOC,the first order destruction rate of metalochlor was 6.3 x 103 s−1, and with 4.71 mgL−1 decreased to2.7 x 103 s−1. This effect was explained with the scavenging effect of NOM towards HO radicals.On the other hand, Canonica et al. [85] studied the transformation of phenols in water andtheir photosensitization by dissolved natural organic material in water. Different dissolved naturalorganic materials photo- sensitized the transformation of a series of methyl and methoxy phenols atpH 8 with a very similar high selectivity (reactivity range 50). It was found that the photooxidationat pH 8 was not controlled by singlet oxygen. Deuterium isotope effects suggested an electron25transfer mechanism. The reactive triplet state concentration was estimated to be 10−14 M in thetop meter of Lake Greifensee under summer noon sunlight, and this leads to a half-life of 7 h for2,4,6- trimethylphenol. Further, it was found that still uncharacterized photooxidants derived fromthe dissolved organic material are also involved in the phototransformation of the phenolsThe relative contribution of inorganic ions and NOM on the scavenging effect and inner filtereffect they may have is not very well understood for VUV process. Most literature studies havefocused on the reaction mechanisms of different micropollutants [87] , [59], and the degradation ofNOM and the formation of byproducts [12]. Only few researchers conducted a study on the efficacyof VUV in the presence of NOM or inorganic ions [88], [89], [90]. A research group [88] conducteda study on the degradation of 2,4 dichlorophenoxyacetic acid (2,4-D) in Milli-Q water and in thepresence of raw water with 5.66 mg/L TOC. The 2,4-D degradation rate decreased significantly(72%) when raw surface was used. The results showed that water matrix can play an importantrole on the efficacy of VUV. However, in this study there was not any discussion on the mechanismresponsible to the decrement of VUV efficiency in the presence of NOM.Other studies [89], [90] were focused on the effect of nitrate on VUV. All these studies havereported a decelerating effect of the degradation in the presence of nitrate due to its capability toabsorb 185 nm photons. For this reason it can act as inner filter, and thus less photons are availablefor the photolysis of water. However, the mechanism responsible to this decrement of efficiencywas not studied in details and the research analyzed only the effect of nitrate and not any other ion.2.6 Knowledge gapsVUV oxidation and photocatalysis with TiO2 can be used in drinking water treatment. VUV has notbeen studied thoroughly and in details; the research around it is limited. The chemical mechanismof the process has always been a challenge because nearly every species absorbs 185 nm photons,and many are photolyzed. For this reason it is difficult to separate the effect of photons from the oneof HO radicals on the overall degradation. Understanding the effect of water matrix, in particular,the effect of NOM and inorganic ions, will be valuable for real and commercial applications ofVUV process because they make an important contribution in the efficiency of process. Therefore,26research is necessary to enrich existing knowledge of the scavenging effect of the water matrixtoward 185 nm radiation and HO• radicals. In addition, to the best of my knowledge, no priorresearch has investigated the process with just 185 nm radiation, because the lamps mostly usedemit at 185 nm and 254 nm. In this case the total output of the lamp is a mixture of two wavelengths(254 and 185 nm). Proper understanding of the process with just one type of radiation and of thewater matrix would be valuable for future applications since it will possible to quantify the impactof different water matrices on the VUV process. Therefore, there is an opportunity to develop sometools (absorption coefficients and scavenging effects of different components of water matrix) thatcan be used to predict the efficiency of VUV in different conditions.Furthermore, as pointed out before, the VUV lamps emit at 254 nm and 185 nm. Hence, it isvaluable to study the combination of UV/TiO2 and VUV in a reactor, allowing for all the photonsto be used: those at 185 nm for the photolysis of water, leading to the formation of HO radicals,and those at 254 nm for the activation of the semiconductor. To the best of my knowledge no priorresearch has explored this possibility.27Chapter 3Research objectivesThe main objective of this research was to study the effect of VUV process at degrading targetmicropollutants and to assess the effects of water matrix on the VUV treatment. These overallobjectives were achieved through the following specific objectives:• Study the absorbance of different ions at 254 and 185 nm and their scavenging effect towardsHO radicals.• Study the absorbance of different sources of NOM at 254 and 185 nm and their scavengingeffect towards HO radicals. The secondary goal was to study the combination of VUV withTiO2/UV process. This objective was achieved through the following specific objectives:• Develop a heterogeneous TiO2 photocatalyst through sol-gel technique with the followingcharacteristics: robust, high efficiency, easy to produce and stable on the support.• Assess the synergistic and beneficial effect or both of incorporating heterogeneous TiO2 pho-tocatalyst in VUV reactors.In order to study in detail the effect of water matrix, first the absorption coefficient of each com-pounds was determinated. Secondly, the scavenging effect was explored following the degradationof a target micropollutant in the presence of different compounds with UV/H2O2. To study the syn-ergistic effects of incorporating heterogenous TiO2 photocatalyst in VUV reactor, the internal walls28of the annular reactor were coated with the previously developed catalyst that showed the highestefficiency and the highest stability.29Chapter 4Experimental setups and proceduresIn this chapter, the general and common experimental setup and procedures that have been used inthis work will be described and explained. Specific experimental procedure associated with eachtask or objective will be explained in the chapter dedicated to that task.4.1 Setups4.1.1 PhotoreactorsFor the photocatalytic experiments two reactors were used, a batch reactor and a flow throughreactor.4.1.1.1 Batch reactorThe batch reactor was used in order to evaluate the photo efficiency of the synthesized TiO2 coat-ings. The reactor was a differential reactor (Figure 4.1 (b)) which consisted of a 63-mm-wide alu-minum reactor designed to allow water to flow through a 225 mm long passage of 25 mm (width)and 3 mm (height), and over the coated glass plates. The reactor was coupled with a pump, a stor-age tank (a 4 L flask), a flow meter, and connecting tubing (Figure 4.1 (a)). A UV lamp emittingat 254 nm (Light Source Inc. GPHVA357T5L with an output power of 11 W) was used as thesource of radiation, and the volume of water utilized in the experiments was 1 L. Microscope slides(25x75 mm) coated with the TiO2 catalysts were placed in the reactor channel, which was closed30Figure 4.1: Batch experimental setup a) and differential reactor b). The batch reactor con-sisted of a sparging beaker, a pump, a flowmeter, the differential reactor and UV lamps.with a quartz plate. Two UV lamps were placed overhead mounted under an aluminum reflector.The lamp centerlines were located 45 mm above the quartz plate and 20 mm away from each other.The photocatalytic reactor was operated under differential mode with complete recirculation of thereagents.4.1.1.2 Flow through reactorThe flow through reactor was used to test the coupling of UV/TiO2 with VUV. The interior ofthe reactor was coated with the photocatalyst that showed the best efficiency and adherence to thesupport in the batch reactor. The photoreactor (4.2) was made of a suprasil quartz envelope with anannular configuration. A VUV-Hg lamp (Light Sources, Inc. G10T51-2-VH), which emits radiationat 254 nm (90%) and 185 nm (10%) or a UV lamp (Light Source Inc. GPHVA357T5L), was placed31Figure 4.2: Flow through reactor experimental setup; the reactor consisted of a annular reactor(a), a pump (b), a feed storage tank (c), and a treated water tank d).longitudinally at the axial center of the reactor in such a way that the envelope of the lamp formedthe internal wall of the reactor. The external and internal diameters of the annular reactor were 2.5and 1.5 cm, respectively, giving an annulus thickness of 0.5 cm and a total reactor volume of 85cm3. The reactor was coupled with a peristaltic pump, a feed storage tank, a treated water tank andconnecting tubing.4.1.2 UV and VUV reactorsUV and VUV irradiations were performed in two collimated beams, one involving only 254 nm (UVcollimated beam) and the other one involving only 185 nm (VUV collimated beam). The kineticsof the photochemical reactions associated with UV/H2O2 were studied in details with collimatedbeams. When properly designed, collimated beam can provide quasi-collimated and uniform radi-ation, which allows obtaining valuable kinetic data that can be easily analyzed. In addition, a smallvolume of the reacting solution is required for each test and, if the solution is thoroughly mixed,the concentration of reagents in the reaction vessel is uniform. Conventional UV-collimated beamsetups, however, cannot be used to study VUV process since oxygen in air partially absorbs photons32at wavelengths shorter than 190 nm. In spite of the advantages of collimated beams, there had notbeen any setup specifically built to study the kinetics of VUV-induced reactions. To the best of myknowledge, VUV process had only been studied in batch and flow-through lab scale reactors. Insuch cases, the concentration of micropollutants and the local incident radiation may not be uniformin the reactor volume, which represent a clear obstacle to interpret the kinetic data obtained. For allthese reasons a VUV collimated beam was designed and built.4.1.2.1 UV collimated beam set-upThe experiments were conducted with a collimated beam bench scale reactor equipped with a lowpressure amalgam lamp emitting light at 254 (Light sources Inc. GPHVA357T5L) positioned 28cm above a circular stirred reactor chamber. The reactor chamber was a petri dish of 5 cm indiameter, and the water path length was 3.33 cm (Figure 4.3). The samples were positioned underthe collimated beam on top of a magnetic stirrer. The flux of the lamp will be measured with anactinometry. The method of choice was the potassium-ferrioxalate actinometry .33Figure 4.3: UV collimated beam set-up (a) and cross section of the UV collimated beam (b);the collimated beam consisted of a UV lamp, a reactor chamber and a stirrer.344.1.3 VUV collimated beam set-upThe VUV collimated beam (Figure 4.4) is comprised of an ozone generating amalgam Hg lamp(Light Sources GPHVA357T5VH/4W) placed in a T-shape PVC enclosure, which is continuouslypurged with nitrogen to remove oxygen present in air. Two Teflon cylinders were placed around thequartz envelope to cover the ends of the lamp in order to improve the collimation of the radiation.The top of the T-shaped enclosure is sealed by a PVC head with a Suprasil quartz and/or an opticalfilter, which allows studying of a specific wavelength during the batch kinetics studies. The quartzand filter are held by three Teflon Orings, which seal the enclosure. Two different combinationsof quartz windows and filters are used. To conduct experiments with only 185 nm radiation, aspecial optical filter is used to block 254 nm. This filter is formed by a MgF2 flat quartz coatedwith a metal-dielectric-metal (MDM) aluminum thin film (eSource Optics). Its peak transmissionwavelength is at 184.9 nm and its diameter and thickness are 50.8 mm and 4.2 mm, respectively.An analytical quality flat UV grade Fused silica window was placed under the 185 nm filter toprotect the metallic coating from ozone that could be formed inside the enclosure. To exposesamples to both 254 nm and 185 nm photons, a head with an analytical quality flat UV gradefused silica can be used. When only 254 nm is required, a regular germicidal Hg lamp (LightSources 35 GPHVA357T5L/4W) is used instead of the ozone generating Hg lamp. To irradiatewater samples, two special cylindrical reaction vessels were built. The first vessel is an open topcontainer made with regular quartz, except the bottom part made of Suprasil quartz to allow 185and 254 nm radiation to be transmitted. The second vessel is similar as the first one, but with thetop part closed, to make it possible to work with volatile compounds. The diameter and heightof both vessels are 4.8 cm and 1.5 cm, respectively. A stirrer is mounted on top of the setup tomix the solution during the irradiation. Ideally, a collimated beam setup should be able to provideuniform and collimated radiation to the solution. That is, the incident radiation at the windowof the reaction vessel should be uniform and the direction all the photons should be normal tothe window of the vessel. However, in real setups, these two requirements cannot be completelysatisfied, and quasi-collimated radiation with small changes in the incident radiation is achieved.To analyze the uniformity and extent of collimation of the radiation reaching the reaction vessel,35a radiation model based on the Monte Carlo technique was developed in collaboration with Dr.Gustavo Imoberdorf (research associate in the department of chemical and biological engineering,UBC). The approach and hypotheses considered in the model were the same as those presentedelsewhere [91]. Figure (4.5) shows the normalized incident radiation at the bottom of the vessel. Afairly uniform distribution of the radiation was obtained and the difference between the maximumincident radiation at the center of the vessel and the minimum at the borders was less than 5%.Similarly, modeling results shows a small divergence of the radiation figure 4.6 and 90% of thebeams are deviated less than 20 degrees from the normal. The low dispersion of radiation wasreduced significantly by the two Teflon tubes surrounding the quartz sleeve (Figure 4.4), which cutthe beams with a high divergence of radiation. Modeling results show that the developed collimatedbeam can provide uniform and quasi-collimated radiation.36Figure 4.4: Diagram of the collimated beam and of the top of the PVC enclosure. (1) motor,(2) reaction vessel, (3) stirrer (4) head of the enclosure, (5) enclosure, (6) VUV lamp,(7) Teflon cylinder (8) Orings, (9) quartz sleeve, (10) PVC head of the enclosure, (11)optical filter, and (12) Suprasil quartz.Figure 4.5: Radial distribution of radiation on the surface of the reaction vessel.37Figure 4.6: Angular distribution of radiation on the surface of the reaction vessel.384.2 Experimental proceduresThe followings are common and general procedures used in this project:4.2.1 Actinometry at 254 nmThe protocol that was used for measuring the flux at 254 nm was an iodide/iodate actinometry[92]. An iodide-iodate-borate solution was prepared and irradiated with UV desired geometry. Theiodideiodate chemical actinometer on exposure to UV forms triiodide, the proposed reaction havingthe following stoichiometry according to [92].8KI+KIO3+3H2O+hν → 3I−+6HO−+9K+ (4.1)The absorbance of triiodide can be determined spectrophotometrically and used to calculate the UVfluence. After that the absorbance was measured at 450 nm. In order to calculate the the quantumyield, the following equation was used:φ = 0.75× [1+0.02× (T −20.7)]× [1+0.23× (C−0.577)] (4.2)where φ is the quantum yield, T is the solution temperature in Celsius degrees and C is the molarconcentration of iodide. Once the quantum yield was calculated, the fluence was quantified:F = 4.72×105×∆A450nm×V ×1000ε450×Φ×A(4.3)where F is the fluence (mJ/cm2), ∆ A450nm is the change in absorbance at 450 nm (cm−1), V isthe solution volume (L), ε450nm is the molar absorption coefficient of triodide at 450 nm (1600M−1 cm−1, φ is the quantum yield (mol/einstein), A is the area irradiated (cm2), 4.72x 105 is theconverting factor from einstein to Joules and 1000 is the converting factor from joule to millijoule.The fluence that was obtained with the geometry used during the UV experiments was found to be0.29 mW/cm2.394.2.2 Actinometry at 185 nmThe protocol that was used for measuring the flux at 185 nm was a methanol actinometry [93].The degradation of methanol during VUV was used to calculate the incident radiation, since theparameters affecting primarily the rate of methanol degradation, i.e., the incident photon rate, theconcentration of dissolved molecular oxygen, the initial methanol concentration were determinedand optimized for a general actinometric procedure.The incident radiation was calculated following:P0 = 1.0575×d[MeOH]dt× VφH2O×ζH2O+φMeOH×ζMeOH(4.4)where P0 is the incident radiation, φh2o is the quantum yield of water (0.375), φMeOH is the quantumyield of methanol (1) , ζH2O is the fraction of photon absorbed by water (0.99), ζMeOH is the fractionof photons absorbed by methanol (0.0038), and d[MeOH]dt is the apparent measured rate of methanolphotolysis (mol L−1 s−1). The concentration of methanol was measured following an enzymaticprocedures. The methanol was determinated through the oxidation of methanol to formaldehydewith an alcohol oxidase enzyme (Pichia Pastoris), followed by condensation with 2,4-pentanedioneto yield the colored product 3,5-diacetyl-1,4-dihydro-2,6-dimethylpyridin [94].4.2.3 Hydrogen peroxide measurementThe residual concentration of H2O2 for the UV/H2O2 and the formation of H2O2 during the VUVprocess were measured with the protocol described by klassen et al.[95]. The procedure was asfollows: first reagent A, which consisted of 10 gr of Potassium hydrogen phtalated dissolved in500 mL of Distilled water (DW) was prepared. After that Reagent B which consisted of 33 gr ofpotassium iodide, 1 gr of NaOH and 0.1 g of Ammonium molybdate tetrahydrate in 500 mL ofdistilled water was prepared. The analysis consisted in the measurement of the absorbance (A) at351 nm of a solution prepared with 2.5 mL of reagent A with 2.5 mL of reagent B with 0.5 mL ofsample diluted with distilled water in a 10 mL volumetric flask. The blank (A0) was prepared in thesame way without the addition of the sample. In order to calculate the concentration of H2O2 the40following equation was applied:H2O2(ppm) =(A−A0)×D0.7776×S (4.5)where D is the additional dilution (1 if none) and S is the sample volume (0.5 mL).4.2.4 Absorption coefficient measurements at 185 and 254 nmThe absorbtion coefficient of different compounds was measured at different concentrations with anAgilent Cary 4000 spectrophotometer. The spectrophotometer worked in a double beam mode un-der nitrogen purging (20 L/min). The spectra were collected from 300 nm to 180nm. The scan ratewas 80 nm, the temperature was set at 25 C and the spectral bandwidth was set to 2. The baselinewas a cuvette with Millipore water. Each sample was placed in the measurement compartment andpurged with N2 for 15 minutes in order to remove all the oxygen before the measurement started.Measuring the absorbance at 185 nm is difficult for the typical spectrophotometers because it is atthe edge of the vacuum UV band where air and, in particular, oxygen absorbs the bulk of thesehigh energy photons. For this reason, it is necessary to purge the optical compartments of the spec-trophotometer with dry, pre-purified N2 gas. Another problem is that the transmission of the quartzsuprasil cuvette falls significantly at shorter wavelength. In addition, the filters, lenses, and gratingsin the spectrophotometer may perform poorly at these high energies as they are optimized for theUV and IR regions. For all these reasons the absorbance of different compounds was measured atdifferent concentrations in order to assess if the absorbance was linear and followed Lambert Beerlaw. If a straight line was obtained the measurements were precise and the absorption coefficientwas the slope of this straight line.4.2.5 pH and Dissolved oxygen (DO) measurementsThe pH was monitored during the irradiations using a pH meter (Thermo Orion PerpHecT LogR1330 meter, 9206BN electrode).The concentration of dissolved oxygen was monitored with a dis-solved oxygen meter (YSI 52 meter, YSI 5909 probe). The pH was found to be 5.2 a the beginningof the irradiations and 5.6 at the end. The effect of pH on the absorbance of ions and its effect on41the degradation of the substrate was not explored. The effect of pH on the percentage of HO rad-icals formed in the system was studied only in the presence of different concentration of chloride.The dissolved oxygen was monitored at the beginning of the irradiations and it was found to be 5.5mg/L.4.2.6 Reproducibility and data accuracyFor the absorption coefficient measurements, for each ion the absorbance was measured at threedifferent concentrations in order and since the absorbance was found to be linear the validity of themethodology was proven. H2O2, TOC, HPLC analysis were all done in duplicate and the valueconsidered was the average of the two runs. In other words, the respective analytical methods werecarried out twice for each sample and the average reported. In order to consider the reproducibilityof the UV/H2O2 and VUV treatment experiments, experimental runs were done in triplicates soeach point that is plotted is the average of three runs with its own standard deviation. Thus the errorbars on figures represent the standard deviation for the three replicate of each sample. Thus, errorbars are presented to indicate the statistical reproducibility of the tests.42Chapter 5TiO2 photocatalyst development andevaluation of photocatalysis coupledwith VUVThis chapter presents the results regarding the secondary objectives of this thesis. As discussed inthe literature review chapter, one emerging and promising UV-AOP is Vacuum UV (VUV) irradi-ation which relies on the formation of hydroxyl radicals (HO•) by the photolysis of water inducedby VUV radiation (Gonzalez et al., 2004). Low pressure Hg or amalgam Hg VUV lamps, whichemit about 10% at 185 nm (VUV) and 90% at 254 nm (UV), can be used as the source of radia-tion. Among the advantages of Hg-VUV lamps are that they are easily available in different sizesand powers and are very affordable. Since only 185 nm photons are capable of promoting waterphotolysis, 90% of the energy (254 nm photons) is not utilized. A potential option to take advan-tage of the 254 nm radiation is the integration of VUV process with photocatalysis, whereby 254nm photons can activate the photocatalyst to promote redox reactions leading to the degradation oforganic contaminants. In this work, five different TiO2 coatings were synthesized and their pho-toactivities, physical characteristics as well as their potentials for integration with VUV photolysiswere evaluated.435.1 Materials and methods5.1.1 Synthesis of the photocatalystsTiO2 coatings were synthesized with different sol-gel techniques, utilizing titanium tetraisopropox-ide (TIP) as the precursor. TiO2 colloids were prepared and deposited on microscope slides (75mmx 25mm). In total, five different photocatalysts were synthesized.• Photocatalyst A was prepared based on the method proposed by Yip et al. [96]. TiO2 colloidswere synthesized by adding 25mL of TIP to 5mL of glacial CH3COOH (Fisher ScientificCanada, 99.7%) at 4 ◦C. To this solution, 250 mL of 1M concentrated HNO3 (Fisher ScientificCanada, 70%) was added, and the sol was stirred for 12h.• Photocatalyst B was prepared with the colloidal suspension obtained by mixing 5 mL ofTIP, 1mL of polyethylene glycol (PEG) (SigmaAldrich Mn400), and 0.5mL of CF3COOH(SigmaAldrich 99%). Ethanol was added at the end until the volume reached 50mL.• Photocatalyst C was prepared following the recipe developed by Yamazaki-Nishida et al.[97]. A suspension of TiO2 was synthesized by mixing 415mL of distilled water with 4.5mLof HNO3 (Fisher Scientific Canada 65%) and 35 mL of TIP. The sol obtained was stirred at80C for 10 h.• Photocatalyst D was prepared according to the method proposed by Keshmiri et al. [98].Ninty-six millilitres of denatured ethanol, 6.4mL of H2O, 16mL of concentrated hydrochloricacid (HCl) (Fisher Scientific Canada, 99%), and 120mL of TIP were mixed, and stirred for 2h. After that, 38 g of Degussa P25 (sigma Aldrich) was added to the sol and the suspensionwas stirred for 12 h.• Photocatalyst E was prepared in the same manner as photocatalyst A, except here 50 g ofDegussa P25 TiO2 was added to the sol and the suspension was stirred for 12 h.The sols obtained were deposited on glass microscope slides with the dip coating technique witha deposition speed of 1 cm min−1. The obtained coatings were calcinated at 400 ◦C for 3 h. The44ramping up temperature rate was 5 ◦C every 5 minutes. At the end of the calcination process, thefurnace was turned off, and the coatings were left inside the furnace to cool down for 2 h.5.1.2 Physical characterization of the photocatalystsThe surface and morphology of the coatings were examined using a scanning electron micro-scope/energy dispersive X-ray spectrometer (SEM/EDX PHILIPS XL30 with Bruker Quantax 200Microanalysis system and light element detector Silicon DriftDetector XFLASH 4010). Solid sam-ples were ground into fine powder with a corundum mortar and then smeared them onto a glassslide with ethanol. Step-scan X-ray powder-diffraction data were collected over a range of 3802qwith CoKa radiation on a Bruker D8 Focus Bragg-Brentano diffractometer equipped with an Femonochromator foil, 0.6-mm (0.3) divergence slit, incident- and diffracted-beam Soller slits, anda LynxEye detector. The long fine-focus Co X-ray tube was operated at 35 kV and 40 mA, usinga take-off angle of 6 ◦C. Phase identification was done using the International Centre for Diffrac-tion Database PDF-4 and Search-Match software by Siemens (Bruker). To determine the UV-Visspectral transmittance of the coatings, the sols were immobilized on quartz slides instead of on themicroscope slides since glass absorbs UV radiation. Spectra were registered with a UV-Vis-NirCary spectrophotometer. The transmittance of the coating was registered from 400 to 200 nm. Thesizes of the TiO2 particles of the colloidal suspensions were measured with light scattering tech-nique (Mastersizer 2000). The reflection index was set at 2.493. The size range of the instrumentwas from 0.020 to 2,000 µm. The light source was a helium neon laser.5.1.3 Photocatalytic activity assessmentTests with a lab-scale glass plate photocatalytic described in Section 4.1.1.1 were performed inorder to measure the photocatalytic activity for the degradation of a target micropollutant, 2,4-D.Contaminated water samples were prepared by adding 2,4-D (10 ppm) to Millipore water. Con-taminated water was placed in the tank, and it was recycled through the reactor at 0.25 L min−1.Each run lasted 90 min. In the first 30 min, the UV lamp was off to assess the adsorption of themicropollutant on the photocatalyst under dark condition. Then, the lamp was turned on, and sam-ples were taken every 10 min for 60 min. The concentration of the micropollutant was quantified45using a high-performance liquid chromatograph (HPLC, Dionex 2695) equipped with C-18 column(4-micronmeter particle diameter) and a UV detector. Methanol/water/acetic acid (58%:40%:2%v/v) were used as mobile phase. The injected sample volume was 100 µ L. The flow rate of themobile phase was 1 mL min−1 and analyses were conducted at λ = 280 nm via UV detection. Threereplicate tests were conducted for each of the five photocatalysts.In order to test the stability of the TiO2 films, attrition tests were performed. Initially, the photo-catalytic efficiency of the fresh films was assessed. Then, water was recycled over the photocatalystsurface at a high flow rate (1 L min−1) in order to determine the amount of photocatalyst lost undera high turbulence flow for 24 h. Finally, the photocatalytic efficacy was evaluated again. Differ-ences in the photocatalytic activity of the fresh photocatalyst and that after 24 h could representthe potential loss of photocatalyst from the surface due to continuous exposure to a turbulent flowduring an extended period of time. As for the deactivation test, consecutive experiments (sevenruns) were performed using the same photocatalyst in order to determine the potential loss of activ-ity due to deactivation that may be caused by the deposition of photooxidation by-products on thephotocatalyst surface.5.2 Results and discussion5.2.1 Physical characterization of the photocatalystsTiO2 films in photocatalysts A and C were not observed in SEM micrographs, possibly becausethose films were too thin. SEM micrographs of composite photocatalysts D and E are shown inFigure 5.1 and Figure 5.2, respectively. Photocatalyst D was not homogeneous and showed signifi-cant fragmentation and cracking, probably due to the evaporation of the solvent during calcination.Photocatalyst E, on the other hand, was more homogeneous and better attached to the support.These early results gave an indication that photocatalyst E would provide a better attachment to thesupport, providing higher attrition resistance.The percentages of crystallinity and different polymorphs in TiO2 play a major role in the ef-ficiency of the photocatalyst. It is known from the literature that rutile has a lower photocatalytic46activity than anatase [99]. However, the activity of Degussa P25, which consists of anatase andrutile (4/1 w/w), exceeds that of pure anatase. For this reason, it is widely believed that photocata-lysts with mixtures of rutile and anatase show a higher degree of activity, and hence, this ratio wasconsidered important in this investigation. Figure 5.3 shows the XRD spectra obtained for the fivephotocatalyst coatings, and Table 5.1 shows the percentage of crystallinity and the percentages ofanatase, rutile, and brookite in each coating. Photocatalyst A was 100% crystalline, but the ratio ofanatase to rutile was not close to the one of Degussa P25. Photocatalyst B was completely amor-phous, while catalyst C was 100% crystalline, even though it was a mixture of anatase and brookite,which does not show any photocatalytic activity. Photocatalyst D was not completely crystalline,but it had a high ratio of anatase to rutile: the percentages of anatase and rutile were 57.8 and 12%,respectively. Photocatalyst E was 100% crystalline with a composition close to that of DegussaP25.Light scattering technique was used to determine the particle size of the colloidal suspensionused to synthesize the immobilized photocatalysts. As shown in Table 5.2, the particle size variesfrom 0.07 to 5.50 µm. The optimum particle size is the result of competing impacts of the effectiveparticle size on light absorption and scattering efficiency, charge-carrier dynamics, and surface area.Small-size particles have a larger surface area but a higher rate of electron and hole recombinationat the particle surface because of the proximity of the charges. Large-size particles have slowerelectron/hole recombination rates, but they possess smaller surface area [100]. Figure 5.4 showsthe UV-Vis spectra of photocatalysts A, B and C. As expected, all the TiO2 films have very lowtransmittance at wavelengths below 300 nm, indicating that they are able to absorb all the photonsreceived at 254 nm, which is the main wavelength emitted by low-pressure UV lamps. Photocat-alysts D and E showed a zero transmittance for the entire range of wavelengths. This is probablydue to the scattering produced by the P25 powder incorporated in the composite coating and due totheir thickness.47Figure 5.1: SEM micrograph of photocatalyst D. The micrograph shows important fractureson the surface of the catalyst.Table 5.1: Percentage of the 4 polymorphous in the 5 photocatalysts.Photocatalyst % Crystallinity Anatase % Rutile % BrookiteA 100 61 38 0B 0 0 0 0C 100 73 0 26D 70 82 17 0E 100 83 16 0Table 5.2: Particle size of the 5 different sols.Colloid Particle size (micron)A 0.07B 5.50C 0.88D 0.75E 0.548Figure 5.2: SEM micrograph of photocatalyst E. The micrograph shows an higher homogene-ity compared to the one of catalyst D.Figure 5.3: XRD of the five different coatings. The figure shows hat catalyst B is amorphousand catalyst A, B, D and E are a a mixture of rutile and anatase.49Figure 5.4: UV-VIS spectra of quartz, photocatalysts A, B and C.505.2.2 Photocatalytic efficacyFigure 5.5 shows the photocatalytic activities of the five photocatalysts. Photocatalysts A, B and Cprovided relatively low degradation of the model contaminant 2,4-D: photocatalysts A and C gave20% degradation, while photocatalyst B showed 10% degradation. The composite photocatalystsD and E provided much higher degradation, about 70% after 1 h of treatment.Figure 5.5: Photocatalytic activity of the five different photocatalysts for the degradation of0.1 ppm 2,4-D in millipore water. Error bars represent the standard deviations of threereplicates samples.Photocatalyst B showed the lowest photocatalytic activity, likely because of its low crystallinity(as shown by the XRD results). It is known that amorphous TiO2 has very little photocatalyticactivity [101]. It has been documented that an increase in the crystallization degree of anatase TiO2can usually lead to an enhancement in its catalytic activity for the degradation of organics. Thisenhancement is often ascribed to the fact that highly crystallized anatase TiO2 would have less51defects acting as recombination centers for photogenerated electrons and holes [102].The fact thatphotocatalysts D and E had higher photocatalytic efficiencies probably was related to the presenceof particles of Degussa P25 and to its ratio of rutile and anatase. The superiority of TiO2 P25 maybe attributed to the morphology of crystallites, which was proposed to be one of the most criticalproperties for the photocatalytic efficiency of P25. Crystallography studies showed that it consistedof multiphases of amorphous, anatase and rutile forms. The close proximity of these phases hasbeen documented to be the reason for long lasting excitation of electrons in the valence bands,allowing for efficient and effective degradation of organic compounds. In the case of TiO2 P25 atransfer of the photogenerated electrons from rutile to anatase takes place leading to the stabilizationof the charge separation and therefore lowering the recombination of the photogenerated carries,which determine the efficiency of most photocatalyst [103]. For all these reasons, Degussa P25is more active than both the pure crystalline phases [37],[91]. Furthermore, it can be stated thatphotocatalysts A, B and C are thin film as shown in the UV-VIS spectra, and so for this reason thephotocatalysic activity is lower.Thickness is an important characteristic that has an impact on the photocatalytic activity. Infact, if the photocatalyst is too thin, the recombination of electron hole can be faster than the oneobtained with thicker film and also not all the photons can be absorbed, and consequently not allthe radiation received on the TiO2 is able to activate the photocatalyst. From the UV-VIS spectrait can be noted that photocatalysts A, B and C are thinner than D and E, since photocatalysts Dand E showed a transmittance equal to 0 for the whole range of wavelengths (800 to 200 nm).Photocatalysts A, B and C are thinner than D and E, as already mentioned, and their thickness ishigher than 150 nm. In fact, if their thickness was equal or less than 150 nm, the UV-VIS spectra,should have shown interference fringes (maximum and minimum due to interference effects) whereTiO2 does not absorb (400 nm) [104]. For this reason it can be stated that probably photocatalystsA, B and C are too thin and they have not the optimal thickness thus the recombination of electronand hole is too fast making the photo degradation less efficient than the one of photocatalysts D andE.525.2.3 Evaluation of the deactivation of the photocatalystsAttrition tests, which allow to assess the adherence of the photocatalytic films to the support, wereperformed only on photocatalyst coatings D and E, as these formulations provided the best pho-tocatalytic efficiencies. The photocatalytic activity of the fresh film was assessed and then waterwas recycled in the reactor over the photocatalyst for 24 h, and another photocatalytic test wasperformed. The results of the two tests are shown in Figures 5.6 and 5.7, the raw data used forplotting the data are reported in Appendix A. Photocatalyst E showed lower loss of photocatalyticactivity, suggesting that it had a better adherence to the support compared to that of photocatalystD. The fact that photocatalyst E showed a similar degradation of 2,4-D in the two tests (i.e. beforeand after the extensive exposure to the high turbulence during 24 h) indicates that there was littlechange in the amount of TiO2 on the surface of the support. For photocatalyst D, on the other hand,the conversion reduced from 70 to 50%, which was believed to be associated with the loss of TiO2from the surface due to attrition caused by the shear forces of the aqueous stream.Figure 5.6: Attrition tests conducted with the photocatalyst D. The degradation of 0.1 ppm of2,4 D in millipore water was followed. The curves show the photocatalytic activity ofthe fresh photocatalyst and after 24 h of H2O recycling. Error bars represent the standarddeviations of three replicates samples.53Figure 5.7: Attrition tests conducted with the photocatalyst E. The degradation of 0.1 ppm of2,4 D in millipore water was followed. The curves show the photocatalytic activity ofthe fresh photocatalyst and after 24 h of H2O recycling.Error bars represent the standarddeviations of three replicates samples.5.2.4 Evaluation of the adherence of the photocatalysts to the supportTo determine the potential deactivation of photocatalysts D and E with time, they were tested re-peatedly for seven runs. Figure 5.8 shows the apparent first-order rate constants obtained in eachtest performed. Photocatalyst D gradually lost its photocatalytic activity, partially due to attrition(discussed above), and also very likely due to the accumulation of oxidation intermediates on thephotocatalyst surface. The latter hypothesis is plausible, because the apparent first-order rate con-stant decreased by about 50%, which is much greater than the anticipated loss of activity due toattrition (shown in Figures 5.6, and 5.7). The raw data used for plotting the data are reported inAppendix A.The oxidation intermediates could adsorb on the photocatalyst surface, reducing thenumber of active sites available for the degradation of 2,4-D. For photocatalyst E, on the otherhand, the activity remained relatively constant after several repeated runs with the same photocat-alyst. The different behavior of the two photocatalysts was likely due to the morphology of thesurface (refer to SEM images). Photocatalyst coating E had much more homogeneous surface and54less cracking. Also, dark adsorption tests indicated that coating D provided more adsorption of2,4-D on the surface (10% of the 2,4-D was removed via adsorption for coating D, while only 2%of the contaminant was adsorbed on coating E). The greater adsorption capacity of photocatalyst Dcould be extended to its oxidation products/intermediates on the surface. That is, as coating D hasa greater tendency to adsorb 2,4-D, it may also retain the oxidation intermediates on its surface.Figure 5.8: Apparent first-order rate constants for the degradation of 2,4-D obtained with pho-tocatalyst D and photocatalyst E after repeated photocatalytic experiments. Error barsrepresent the standard deviations of three replicates samples.5.2.5 Performance of the photocatalysts in a flow-through reactorWith colloid E showing high photocatalytic activity and high adherence on the support, it wasimmobilized and then evaluated in a flow-throught reactor operating in a continuos mode understeady-state. Tests were performed with VUV involving 185 and 254 nm radiation and no TiO2and with UV radiation with TiO2. Also, control experiments were carried out with UV/slurry TiO2(Degussa P25), 1 gr/L. The flow rate used was 1 L min−1 in the case of Millipore water and 0.25L/min in the case of surface water and the initial concentration of 2,4 D was 0.1 ppm. The data55used to calculate the conversion of 2,4-D during the various UV based processes are reported inAppendix A.The results (Figure 5.9) obtained from the flow through reactor showed that photocatalysis hasa very low degradation rate. The heterogeneous TiO2 coating in a flow through reactor with a highflow rate shows lower contact time with the water and probably this is the reason for the very lowdegradation rate achieved. A test was carried out in the presence of oxygen purging but the effi-ciency increased slightly and the conversion of 2,4-D was found to be only 4%: when the turbulence(air bubbles) was increased the efficiency increased, not enough to compete with the VUV processthough. Also, the test with P25 shows a much slower degradation rate than that achieved with VUV.This proves that the main problem of the low efficiency of photocatalysis is the contact time betweenthe photocatalyst and the pollutant. Heterogenous photocatalysis involves two kinds of reactions,classified according to two mechanism: one is the direct photogenerated electron reduction andphotogenerated hole oxidation: the other is the radical-mediated reaction, which involves reactivespecies such as hydroxyl radicals, superoxide radicals generated during the process. The formermechanism occurs only at the surface of the photocatalyst and is therefore a heterogenous reac-tion: the latter can occur very close to the surface and thus can be a homogeneous reaction. Sinceheterogenous reaction is a multistep process, it is more complex than a homogeneous reaction andmass transfer is usually the rate-limiting step in the heterogeneous photocatalysis. A homogeneousradical reaction in liquid is fast, with rate constants usually in the order of 106-109M−1 s−1 [105].On the other hand, in a heterogeneous process, the radicals are generated on the photocatalyst sur-face and in order to react with the substrate the substrate has to migrate towards the surface ofthe photocatalyst. The photocatalytic reaction rate can be increased if the substrate diffuses moreefficiently from the solution to the photocatalyst. For this reason the addition of air bubbles couldimprove the turbolence and hence the photocatalytic process. In addition, this could explain thereason why in the presence of P25 TiO2 the efficiency was higher: the photocatalyst was in theform of a slurry and for thus the contact time was higher.Figure 5.9 and Figure 5.10 show the degradation rate of TiO2/UV, VUV and P25 slurry withmilli pure water and with surface water containing NOM (natural organic matter). In both cases,56VUV showed a greater photoefficiency. With surface water (Figure 5.10), the degradation rate ofVUV decreases because NOM can act as HO• scavenger and it can adsorb photons. In the case ofTiO2/UV the efficiency in the presence of organic matter increased. The reason could be the lowerflow rate. In this case the contact time between the micropollutant and the photocatalyst could beimproved since with lower flow rate the substrate can interact longer with the photocatalyst.In general, the VUV process was found to be more efficient since the photolysis of water has ahigher quantum yield than the one of TiO2 (0.375 and 0.040 respectively). In addition, VUV is anhomogeneous process and therefore is not limited by mass transfer. On the other hand, Figure 5.11shows the 2,4-D conversion for VUV with surface water and millipore water. It is evident that theefficiency decreased in the presence of surface H2O. This is due to the presence of inorganic ionsand NOM that can either absorb photons or scavenge HO radicals.Figure 5.9: 2,4 D conversion for photocatalysis/UV, VUV and P25 slurry/UV with milliporewater and a flowrate of 1 L/min.5.3 ConclusionsFive different photocatalysts were synthesized through the sol-gel technique and coated on glasssupports. Three coatings were prepared using aqueous sols of TiO2 nanoparticles with different57Figure 5.10: 2,4 D conversion for photocatalysis/UV, and VUV with surface water and aflowrate of 0.25 L/min.Figure 5.11: 2,4 D conversion for VUV with surface water and millipore water.58amounts of TIP, and different quantities and types of acids. The other two photocatalysts werecomposite photocatalysts which were prepared by incorporating Degussa P25 into the TiO2 matrixsynthesized through the sol-gel technique. The composite photocatalyst showed much higher pho-tocatalytic activities, about 70% 2,4-D degradation, compared to 12% 2,4-D degradation resultedfrom the photocatalysts without Degussa P25. The two composite TiO2 photocatalysts were alsoevaluated in terms of attrition resistance and photocatalytic stability. The composite coating withmore homogeneous morphology and less cracking presented lower attrition and no deactivation.The conversions obtained for the VUV, UV/immobilized TiO2, and UV/slurry TiO2 were 100%,2.5% and 38% with millipore water and 63% and 7% with surface water respectively. The resultsshowed that photocatalysis cannot improve significantly the efficacy of VUV in a flow through re-actor for the degradation of micropollutants. When the samples were prepared with surface waterthe efficiency of VUV process decreased. This decrement is due to compounds (inorganic ions andNOM) dissolved in surface water which can interfere in the process. In particular, NOM and inor-ganic ions can absorb 254 and 185 nm photons, and/or scavenge HO radicals. On the other hand,the efficiency of TiO2/UV increased because the flow rate was lower and thus the mass transfer wasimproved.59Chapter 6Effects of inorganics on the degradationof micropollutants with VUV advancedoxidationThis chapter and the following ones address the main objective of the thesis which is to evaluate theimpact of water matrix on the efficiency of VUV process.In this chapter, the effect of various inorganic ions on the degradation of atrazine is presented.Inorganic ions can absorb photons, scavenge HO radicals and can produce reactive species whenphotolyzed. To understand how profoundly inorganic ions can affect the degradation of atrazineduring AOPs, various experiments were performed. First, the molar absorption coefficient of eachion at 254 and 185 nm was evaluated: in this way it was possible to quantify the photons distributionduring the process with inorganic ions. Second, the quantification of atrazine’s quantum yieldat 185 nm and 254 nm was determinate. The value of atrazine’s quantum yield at 185 nm is afundamental parameter for the development of a kinetic model. To the best of my knowledgethere is no methodology developed for the quantification of the quantum yield of organics at 185nm. Furthermore, the ability of inorganic ions to scavenge HO radicals was studied in UV/H2O2.UV/H2O2 was used for these experiments because, during this process, the degradation of atrazine60occurs only through HO radicals attack: on the other hand, in VUV the overall degradation is thecombination of HO radicals attack and 185 nm photolysis. Last, experiments were conducted inVUV process, to explore the overall effect of each ion on the degradation of atrazine during VUVirradiations.6.1 Material and methods6.1.1 Water samples and chemicalsContaminated water samples were prepared by adding atrazine (Sigma Aldrich) intoMillipore waterto have a final concentration of atrazine of 0.1 ppm. To study the effects of inorganic ions, variousconcentrations of NaNO3 (Fisher, Canada), NaHCO3 (Fisher, Canada), KF (Fisher, Canada), NaCl(Fischer Canada) and Na2SO4 (Fisher, Canada) were added to the water samples. For UV/H2O2experiments, H2O2 (Fisher, Canada) was added and the final concentration was 10 ppm. For quan-tum yield experiments methanol (Fisher, Canada) was added. Elemental analysis was performed onsurface natural water samples from Bowen island and Peachland.Elemental analysis was performed on surface natural water samples from Bowen island andPeachland BC. The elemental analysis of two different natural and surface waters was performed.The concentration of each metals was carried out with ICP/OES, the one of inorganics was mea-sured with an IC and the one of organic and inorganic carbon was performed with a using a TOC-VCPH Shimadzu analyzer. pH was measured using a pH meter (Thermo Orion PerpHecT LogR1330 meter, 9206BN electrode). The concentration of dissolved oxygen was monitored with adissolved oxygen meter (YSI 52 meter, YSI 5909 probe).6.1.2 UV/H2O2 and VUV irradiationsThe samples prepared were spiked to 10 ppm H2O2 and irradiated first with the conventional colli-mated beam described in (Section 4.1.2.1). The volume of the samples was 25 mL. Aliquots werecollected at 5 minutes intervals and the irradiation lasted 25 minutes. The experiments involvingVUV radiation were conducted in the VUV collimated beam described in Section 4.1.3 emitting at185 nm and in this case no H2O2 was added. The reactor vessel was a beaker with an area of 3861cm2 and the sample volume was 25 mL. In this case, every two minutes a sample was taken and theirradiation lasted for 10 minutes. In the case where the inorganic ions would not show any scav-enging effects, methanol was added at 1 ppm as scavenger of HO radicals to slow the degradationrate.6.1.3 Analytical methodsThe degradation of atrazine was quantified using a high-performance liquid chromatograph (HPLC,Dionex 2695) equipped with C-18 column (4-micronmeter particle diameter) and a UV detector.Methanol/water/acetic acid (58%:40%:2% v/v) were used as mobile phase. The injected samplevolume was 100 µ L. The flow rate of the mobile phase was 1 mL min−1 and analyses were con-ducted at λ = 280 nm via UV detection. The concentration of H2O2 was measured using thetriiodide method described in Section 4.2.3. The molar absorption coefficient of various inorganicswas measured following the procedure described in Section 4.2.4. The elemental analysis of twodifferent natural and surface waters (Bowen island BC and Peachland BC) was performed. Theconcentration of each metals was carried out with ICP/OES, the one of inorganics was measuredwith an IC and the one of organic and inorganic carbon was performed with a using a TOC-VCPHShimadzu analyzer. pH was measured using a pH meter (Thermo Orion PerpHecT LogR 1330meter, 9206BN electrode). The concentration of dissolved oxygen was monitored with a dissolvedoxygen meter (YSI 52 meter, YSI 5909 probe).6.2 Results and discussion6.2.1 Molar absorption coefficients at 185 nm and 254 nmVarious solutions of inorganic ions were prepared with salts that have Na+ or F− as counter ion.It is known that either Na+ or F− [106] does not absorb at 185 nm and hence, it was possible toisolate the absorbance of the anion and cation.These experiments allowed to determine the molar absorption coefficients at 185 nm. The dataof the absorbance measurements are reported in Appendix D. The absorption coefficient of waterwas first measured to be 1 M−1 cm−1. In previous work the absorption coefficient was found to62be 1.8 M−1 cm−1 [19]. In this work, however, the methodology followed was accurate and thusthe value is trustable. In this case, the absorbance at 185 nm was measured in three cells withvarious path lengths (0.2, cm, 0.5 cm and 1 cm). The absorbance was linear, following the LambertBeer law, (Figure 6.1) and this proves the fact that the methodology followed was accurate and theresults trustable. In addition all the measurements were performed under N2 purging in order toremove all the O2 which is known to absorb 185 nm ration. The same approach was followed forvarious inorganic ions, atrazine (the model micropollutant), methanol and H2O2. The absorbanceof each ion was measured at 4 concentrations to demonstrate the linearity of the absorbance. Table6.1 shows the absorbance of various ions at 254 nm. As it is shown only nitrate, atrazine andH2O2 could impact the photons distribution at 254 nm. This mean that during VUV 254 nm canpenetrate in the water and potentially can disinfect the water. On the other hand, all the compoundsshowed a high absorption coefficient at 185 nm (Table 6.2) indicating their significant impact onthe VUV process. They can act as inner filter during the VUV process absorbing photons, whichare not available anymore for the photolysis of water and for this reason fewer HO radicals can beproduced.Table 6.1: Absorbtion coefficient of different compounds at 254 nm.CompoundsAbsorption coefficient(M−1cm−1)Na+ < 0.01K+ < 0.01HSO4− <0.01Ca2+ < 0.01HCO3− < 0.01H2O2 19NO3− 3Atrazine 377463Figure 6.1: Water absorbance for different cell path lengthTable 6.2: Absorbtion coefficient of different compounds at 185 nm.CompoundsAbsorption coefficient(M−1cm−1)Na+ < 0.01K+ 841HSO4− 148Ca2+ 109HCO3− 269H2O2 341NO3− 5568.Atrazine 25370646.2.2 Methodology for the determination of the observed kinetic constantDuring this work the observed kinetic constant was evaluated experimentally. In the followingparagraph the methodology is presented for the case of atrazine in Millipore water. The degradationof atrazine in millipore water , in the presence of HO radicals, takes place according to:d[Atrz]dt=−kobs[HO][Atrz] (6.1)As it is shown the disappearance of atrazine theoretically follows a second order law, in whichthe concentration of two reactants must be followed simultaneously. However, in the case of thereaction with HO radicals, a pseudo-first order approximation is used. In this case the concentrationof atrazine depends only on the initial concentration of atrazine and not on the concentration of HOradicals which is assumed to be constant during the process. Experimentally the kobs is obtainedplotting CtC0 versus time of irradiation, where Ct is the concentration measured after time t of irradi-ation and C0 is the initial concentration. The fit of the points obtained with an exponential functiongives the kobs according to:CtC0= exp−kobst (6.2)6.2.3 Degradation of atrazine in millipore waterAs already mentioned the VUV-induced degradation of micropollutants is based on the formationof the highly reactive radicals that are generated when water molecules photolyze after absorbing185 nm radiation:H2O+hv185nm → HO•+H• (6.3)H2O+hv185nm → HO•+H++ e−aq (6.4)The quantum yields for the photochemichal homolysis and ionization of water at 185 nm are0.33 and 0.045, respectively. In solution saturated with air H• and e−aq react almost exclusively with65dissolved O2 to produce superoxide radicals:H•+O2 → HO2• (k = 2.1×1010M−1s−1) (6.5)e−aq+O2 → O2•− (k = 2.0×1010M−1s−1) (6.6)HO• can react with superoxide radicals (HO2•/O2•−) to produce water and O2:HO•+HO2• → H2O+O2 (k = 6.6×109M−1s−1) (6.7)HO•+O2•− → O2+HO− (k = 7.0×109M−1s−1) (6.8)Hydroxyl radicals and superoxide radicals can react among themselves to produce H2O2:HO•+HO• → H2O2 (k = 4.0×109M−1s−1) (6.9)HO2•+HO2• → H2O2+O2 (k = 2.9×106M−1s−1) (6.10)O2•−+HO2• → O2+HO2− (k = 9.7×107M−1s−1) (6.11)In addition, 185 and 254 nm photolyzes H2O2:H2O2+hv185nm → 2HO• (φ = 0.5) (6.12)H2O2+hv254nm → 2HO• (φ = 0.5) (6.13)In the presence of organics, HO radicals react with them to produce by-products:Organics+HO• → byproducts (6.14)In addition, depending on the nature of organics, they can photolyze when they absorb 254 or185 nm radiation:66Organics+hv185nm → byproducts (6.15)Organics+hv254nm → byproducts (6.16)The byproducts generated after the partial oxidation or photolysis of organics can undergo fur-ther degradation through reactions with HO radicals and other radicals, potentially leading to com-plete mineralization.To summarize, the degradation of atrazine in millipore water was examined prior to investigat-ing the effects of inorganics. This was because of the need to understand the fundamental mecha-nisms responsible for the degradation of micropollutants.In millipore water, the degradation of atrazine can occur through three mechanisms: reaction withHO radicals, interaction with 185 and 254 nm radiation.Atrazine+HO→ Products (6.17)Atrazine+hν254nm → Products (6.18)Atrazine+hν185nm → Products (6.19)Since the formation of HO radicals and the direct photolysis of atrazine are activated by eitherVUV or UV radiation, the spatial distribution of photons should be evaluated. This is done bytaking into account the spectral local incident radiation defined as:G=∫ 0ωIλ (x,ω)dω (6.20)Where Iλ ,ω is the specific spectral radiation intensity that reaches a given point located at positionx, and ω is the solid angle. When the experiments are conducted in a a collimated beam , theLambert-Beer law can be applied to describe the radiation propagation. The incident radiation was67calculated using the Lambert-Beer law:Gλ (x) = G0λ ×10αλ x (6.21)where x is the distance of a given point from the bottom of the VUV reaction vessel (or from thetop in the case of the UV collimated beam), G0 is the incident radiation at the entrance of the vessel(determinated by chemical actinometry), and αλ is the absorbance coefficient of the propagatingmedium. The absorption coefficient of the solution depends on the concentration of the speciesabsorbing radiation at a given wavelength and their corresponding molar absorption coefficient.αλ =n∑i=1εiCiln(10) (6.22)where εi is the molar absorption coefficient of the specie i, and Ci is the concentration of specie i.In order to quantify and separate the rates of direct photolysis at 254 nm and 185 nm, thequantum yield of micropollutants at this wavelength is needed. In literature the quantum yield oforganics at 254 nm can be found, but, to the best of my knowledge, there is not any methodologyfor the quantification at 185 nm. In the following paragraph a methodology for the quantificationof the quantum yield is presented.6.2.3.1 Quantum yield of atrazine at 254 nmThe quantum yield of atrazine at 254 nm was obtained experimentally by irradiating solutions ofatrazine with 254 nm radiation in the UV collimated beam setup. The quantum yield was evaluatedas the ratio between the rate of atrazine degradation due to the photolysis only and the volumetricrate of photon absorption: (the detailed explanation of the procedure is reported in Appendix E.φAtz,254nm =kphot254,AtzVrv ·C0AtzG0254(1−10−εAtz,254C0AtzVrvArv )(6.23)where kphot254,atr is the apparent kinetic constant obtained from the photolysis of atrazine, Vrv isthe volume of the reaction vessel, Avr is the surface of the reaction vessel, C0,Atr is the initial68concentration of atrazine, and G0254 is the incident radiation at 254 nm obtained with a chemicalactinometry. The kinetic constant was obtained from experiments conducted in the UV collimatedbeam in which samples made with only millipore water and atrazine were irradiated under UV at254 nm. The k obtained in this way was the one corresponding to the photolysis. The experimentalvalue of the atrazine quantum yield at 254 nm was 5.6x10−2 which was consistent with thosepreviously reported [107].6.2.3.2 Quantum yield of atrazine at 185The measurement of the quantum yield of organics at 185 nm is more difficult since water moleculesphotolyze and the HO• formed reacts with atrazine (Katz−HO = 2.54x109 M−1 s−1). This challengecould be addressed by adding an organic compound to scavenge HO radicals formed and suppressthe degradation of atrazine via HO radicals. In this research, methanol was used because of its highreactivity with HO radicals (kMeHO = 9.7 x 108 M−1 s−1) and its relatively low absorption coeffi-cient at 185 nm (ε185,MeHO= 6.3 M−1 cm−1). In order to prove that methanol can scavenge HO•effectively, some experiments were conducted with H2O2 and UV. In the absence of methanol, thedegradation of atrazine was mainly due to the attack of HO•. Increasing methanol concentration,however, reduced the concentration of HO radicals available for the reaction with atrazine resultingin a decrease in the apparent first order rate constant. With methanol concentrations greater than20 ppm, the kinetic constant reached a plateau at 0.024 min−1, which is equal to the value corre-sponding to the rate constant of the photolysis of atrazine in millipore water (Figure 6.2). Underthese conditions, methanol reacted with all the HO radicals and so the degradation of atrazine wasonly due to photolysis at 254 nm and the value was consistent with the one obtained (5.6x10−2).The same experiments were conducted with 185 nm radiation. Since the amount of HO radicalsgenerated in the VUV process was higher than that in the H2O2/UV process, higher concentrationsof methanol were required to scavenge the HO radicals. In this case, the plateau was reached withmethanol concentrations higher than 50 ppm. The apparent kinetic constant corresponding to theplateau could be associated to that of the photolysis of atrazine at 185 nm and was used for thecalculation of the quantum yield.69Figure 6.2: Apparent first order kinetic constant for the degradation of atrazine obtained withH2O2/UV and VUV.The fuence rates were 0.33mW/cm2 for 254 nm and 0.06 mW/cm2for 185 nm.The quantum yield of atrazine at 185 nm was evaluated as the ratio between the rate of atrazinedegradation and the volumetric rate of photon absorption. For the calculation of the quantum yieldat 185 nm, the absorbance of water, atrazine and methanol should be considered:φAtz,185nm =kphot185,AtzVrv(εH2O185CH2O+ εAtz,185C0Atz+ εMeOH,185C0MeOH)G0185AVrεAtz,185(1−10(εH2O,185+εAtz,185C0Atz+εMeOH,185C0MeOH)VrvArv(6.24)where k(phot185,atr) is the apparent kinetic constant corresponding to the degradation rate of atrazinedue only to 185 nm photolysis, ε(H2O,185) is the absorption coefficient of water at 185 nm CH2Othe concentration of water, ε(atr,185) the absorption coefficient of atrazine at 185 nm, ε(MeHO,185)the absorption coefficient of methanol at 185 nm, C0(MeHO) the concentration of methanol (varies70through the experiments), G0(185) the incident radiation at 185 nm. The calculated quantum yieldfor atrazine at 185 nm was found to be 6.9x10−2, indicating that the photolysis of atrazine at 185nm is not important and its degradation occurs primarily through reaction with HO radicals. Themethodology developed in this study can be applied to any organics and it can be used as a valuabletool for the study of the VUV process, since the quantum yield is a fundamental value for thestudy of the VUV reactions. To the best of our knowledge there is no such simple yet effectivemethodology reported in the literature. Since the quantum yield of atrazine at 185 nm was found tobe low, atrazine was chosen to be the micropollutant used for studying the effect of water matrix onVUV process. 2,4 D, the micropollutants used for the work related on photocatalysis, was not usedanymore since it is slightly acid. This means that during VUV irradiation its dissociation could playa role in the VUV system mechanism.6.2.3.3 Degradation of Atrazine in millipore water during VUV and UV processThe first set of experiments involved examining the degradation rates of atrazine in Millipore waterwith either 185 nm or 254 nm with a fluence rate of 0.03 mW/cm2 and 0.29 mW/cm2 respectively.Figure 6.3 shows that the degradation of atrazine with VUV was much faster than the one with 254nm. The photolysis of water irradiated with 185 nm photons and consequently formation of HO•led to the degradation of atrazine through HO• attack and through direct photolysis with 185 nm.The quantification of HO radicals produced during the VUV process is a key value for the studyof the process and the evaluation of the scavenging effects of water matrix. The rate of generationof HO radicals was given by:rHO,gen = G185nm(φ1,185nm+φ2,185nm)εH2O,185nmln(10)CH2O (6.25)The contribution of H2O2 photolysis is negligible since the concentration of H2O2 formed inthe system is low (less than 1 ppm). For this reason HO radicals are generated mainly due to thewater photolysis. The rate of consumption of HO radicals can be estimated according to:rHO,con = ∑Ni=1 ki−HO ·Ci ·CHO (6.26)71Figure 6.3: Photolysis of atrazine with VUV and UV. The fuence rates were 0.03mW/cm2and 0.29mW/cm2, respectively. Error bars represent the standard deviations of threereplicates samples.where equation includes all those stable species present in the sample that can react with HO radi-cals. On the other hand, by applying the local steady state concentration assumption to HO radicals(rHO,gen = rHO,con) the concentration of HO• in the can be calculated as:CHO(x) =G185(x)(φ1,185+φ2,185) · εH2O,185 · logCH2O∑Ni=1 ki−HO ·Ci(6.27)where CHO(x) is the local HO• concentration, G185(x) is the incident radiation, φ1,185 is the quan-tum yield of the photochemical homolysis of water (0.33), and φ2,185 is the quantum yield of thephotochemical ionization of water (0.045). Equation 6.27 gives an approximate value of the HO•concentration and involves the following assumptions:• The local steady-state approximation was assumed for radical species. This assumption canbe justified taking into account that radicals are extremely reactive, and therefore, the accu-72mulation and the diffusive terms of the differential local mass balance for a given radical arenegligible when compared to the generation and degradation rates in chemical reactions. Forexample, the diffusion coefficient of HO radicals is about 2.3x10−5 cm2 s−1 and its averagelife time in water is in the order of 10−9-10−6 s−1. Therefore, HO radicals are expected toreact close to the place where they are generated. This assumption implies that the rate ofgeneration of radicals is equal to their rate of degradation,• HO radicals are only formed due to the photolysis of water at 185 nm. There is a smallgeneration by the photolysis of H2O2 that is generated in the process, but it does not play asignificant role,• The reactions of HO radicals with other radicals are neglected,• The total absorption coefficient of the solution is constant during the irradiation. The validityof this hypothesis was verified experimentally by measuring the absorption coefficient ofwater after and before VUV irradiations,• Other radicals are either at very low concentration (H+, e−) or have a low reactivity and theydon’t play a significant role,• The formation of H2O2 is not significant; this assumption was confirmed experimentally,showing maximum 0.1 ppm of H2O2 formation.In such system the steady state concentration of HO was found to be 1.46x10−11 M. The samecalculation was conducted without any compounds that can scavenge HO radicals. In this case thesteady state concentration of HO radicals was found 2.08x10−10 M. This proved the fact that even atlow concentration the presence of organics affected the amount of HO radicals in the system. Sinceduring the VUV process the degradation of atrazine can occur through HO attacks and through 185nm radiation, the contribution of 185 nm photolysis to the overall degradation was calculated. Theobserved rate of disappearance of atrazine (Figure 6.3) is the sum of the degradation rate due to the73photolysis and the degradation rate due to HO radicals:r = φ ×G× ε × [C]+KHO× [C]HO× [C] (6.28)where r is the apparent kinetic constant, φ is the quantum yield of the micropollutant, G the radiationflux, ε the absorbtion coefficient of the micropollutant, [C] the concentration of the micropollutant,kHO is the kinetic constant between the micropollutant and the HO radicals and [C]HO is the concen-tration of HO radicals in the system. The first term of the equation 6.28. represent the degradationdue to the photolysis, the second the one due to the reaction between the micropollutant and theHO radicals. r is measured experimentally, kHO is known from UV/H2O2 experiments, [C]HO iscalculated according to equation 6.27, and φ is the quantum yield.Knowing all these parameters, it was possible to calculate the contribution to the degradationrate due to photolysis only. The photolysis of atrazine at 185 nm contributes only to 0.14% of itsdegradation and for this reason the degradation of atrazine was considered to be driven primarily bythe HO• attack. Knowing the absorbtion coefficient of water and atrazine it was possible to evaluatethe photons distribution in the system. The percentage of photons absorbed by a specific compoundduring VUV process was calculated according to the formula:%185nm= Ciεi∑Ni=1Ciεi(6.29)where Ci and εi are the concentration and the absorption coefficient of specie i, respectively. DuringVUV process in the presence of 0.1 ppm of atrazine in millipore water, about 95% of the photonsare absorbed by water, 2.7% by atrazine, and the remaining 2.3% transmitted without absorption.This shows that most of the photons were absorbed by water leading to the formation of HO radi-cals. Without any other compounds in the system, the degradation of atrazine occurred through HOradicals attack. In addition, atrazine showed to be a good micropollutant for the VUV experimentssince degrades only a little due to the 185 nm photolysis (0.14%). The same calculation was con-ducted in the absence of any compounds that are able to scavenge HO radicals. In this case, the74steady state concentration of HO radicals was found to be 2.08x10−10 M, a value two order of mag-nitude higher than the one obtained in the presence of 0.1 ppm of atrazine. This showed that evenat low concentration, the presence of organics affected the amount of HO radicals in the system.6.2.4 Degradation of atrazine in the presence of different inorganic ions6.2.5 Effect of Na+ on VUV and H2O2/UV processesThe first set of experiment was conducted with UV/H2O2 in the presence of different concentrationof NaF since from the absorption coefficient experiments it was found that NaF does not absorb anyphotons at 254 (Table 6.1). As it it shown in Figure 6.4, the first order constant for the degradationof atrazine did not change with different concentration of NaF, indicating that Na+ and F− do noscavenge any HO radicals and do not absorb 254 nm photons. The second set of experiment wasperformed with VUV. As it it shown in figure 6.5, also in this case, the first order constant for thedegradation of atrazine did not change with different concentration of NaF. The results show thatNa and F do not absorb 185 nm photons as well and this is in line with the results obtained for theabsorption coefficients (Table 6.2).Since Na+ did not show any effect on the distribution of 254 nm and 185 nm photons and notto have any scavenging effect, it was chosen to be the counter ion for the study of different anions.6.2.6 Effect of nitrate on VUV and H2O2/UV processesThe effect of nitrate is complex since it can:• Absorb 254 nm photons• Be photolyze and produce HO radicals• Absorb 185 nm photonsThe first set of experiments were performed following the degradation of atrazine in the pres-ence of different concentrations of nitrate during UV/H2O2 AOP. As shown in Figure 6.6. thepresence of nitrate decreased the degradation rate of atrazine by 75%. This trend can be explainedby the fact that nitrate acts as inner filter for the 254 nm radiation. This means that less photons75Figure 6.4: First order rate constant for the degradation of atrazine using UV/H2O2 in the withdifferent concentrations of NaF. The fuence rate was 0.29mW/cm2. Error bars representthe standard deviations of three replicates samples.are available for the photolysis of H2O2 with the consequence of less HO radicals being generatedin the system. In addition, fewer photons are available for direct photolysis of atrazine. As alreadypresented, the absorption coefficient of nitrate at 254 nm is 3.51 M−1cm−1.In this system the percentage of photons at 254 nm absorbed by each compound was calculatedaccording to the equation:%254nm= Ciεi∑Ni=1Ciεi(6.30)In this system, 98% of the photons are absorbed by nitrate, 1.54% by H2O2 and 0.2% byatrazine. On the other hand, it is known that nitrate is able to photolyzes when exposed to 254nm forming HO radicals [108]. For this reason experiments were performed in the same conditionbut without the addition of H2O2. As shown in Figure 6.7, degradation rate of atrazine increasedwith the increases in NO3− concentrations. This confirms the fact that nitrate is able to photolyzeand to produce HO radicals at concentration higher than 20 ppm..76Figure 6.5: First order rate constant for the degradation of atrazine using VUV with differentconcentrations of NaF. The fuence rates was 0.03mW/cm2. Error bars represent thestandard deviations of three replicates samples.The steady state concentration of HO radicals was calculated during UV photolysis of nitrate:Ktotexp = Katr254+KatzHO[HO] (6.31)where Ktotexp is the total observed kinetic constant obtained experimentally with the UV/H2O2experiments performed, Katr254 is the kinetic constant for the photolysis of atrazine at 254 nmobtained experimentally with the quantum yield experiments, and katrHO is the kinetic constantbetween HO radicals and atrazine (1.3 x108 L mol−1 s−1) [109]. In such system the concentrationof HO radicals in the presence of 100 ppm of nitrate was found to be 7.6 x 10−12 M.Although nitrateis able to photolyze and to produce HO radicals, in H2O2/UV the degradation rate decreases. InH2O2/UV HO radicals are generated through the photolysis of H2O2. The formation of HO radicalsin this process is much more effective than through the photolysis of nitrate since the quantum yieldfor the formation of HO radicals is 0.09 for nitrate and 1 for H2O2 [110]. For this reason even at77Figure 6.6: First order constant for the degradation of atrazine using UV/H2O2 in the presenceof different concentration of nitrate. The fuence rate was 0.31mW/cm2. Error barsrepresent the standard deviations of three replicates samples.the highest concentration (100 ppm) nitrate produces 10% less HO radicals than 10 ppm of H2O2.For this reason the in the case of UV/H2O2 the nitrate plays the role of inner filter making thedegradation rate of atrazine slower.The last set of experiments was performed with VUV. In the case of VUV process the k de-creased (Figure 6.8) with the increase of the nitrate concentration. In this case, nitrate again actsas inner filter for the photon absorption at 185 nm. At 50 ppm nitrate is able to absorb 81% of 185nm photons, leaving just 16% for the photolysis of water and consequently HO formation. Sincethe decrement of the first order rate constant is 50%, also in the case of VUV nitrate can photolyzeand produce HO radicals but, in addition in the case of VUV, the effect of inner filter is compensatewith the ability of producing HO radicals.78Figure 6.7: First order constant for the degradation of atrazine using UV in the presence ofdifferent concentration of nitrate. The fuence rate was 0.29mW/cm2. Error bars repre-sent the standard deviations of three replicates samples.6.2.7 Effect of bicarbonate on VUV and H2O2/UV processesThe first set of experiments were performed following the degradation of atrazine in the presence ofdifferent concentrations of bicarbonate during the UV/H2O2 treatment. As shown in Figure 6.9 thepresence of bicarbonate decreased the degradation rate of atrazine. Given that bicarbonate does notabsorb photons at 254 nm (Table 7.1) it cannot act as inner filter and for this reason the decrementobserved in the kinetic constant is due to the scavenging effect. The percentage of HO scavengedin the system was calculated according to the following equation:%HO• = CiKHO,i∑Ni=1CiKHO,i(6.32)where Ci is the concentration of specie i, KHO,i is the kinetic constant between specie i and HOradicals. In the case of atrazine, the effective kinetic constant corresponding to the HO attack wascalculated (the detailed methodology is presented in Appendix F). During the UV/H2O2 process79Figure 6.8: First order rate constant for the degradation of atrazine using VUV with differentconcentrations of nitrate. The fuence rate was 0.03mW/cm2. Error bars represent thestandard deviations of three replicates samples.with 0.1 ppm of atrazine, 50 ppm of bicarbonate, and 10 ppm of H2O2, the percentage of HOradicals scavenged by each component was 68%, 14% and 17%, respectively. Experimentally thefirst order rate constant decreased by 25% when 50 ppm of bicarbonate was in the system comparedto the case where there was no bicarbonate.The second set of experiments were performed with VUV irradiation and as figure 6.10 showsthe degradation rate decrease with higher concentration of bicarbonate. As far as the absorptionat 185 nm is concerned, bicarbonate has a low absorption coefficient, which is one of the lowestamong all the ions. Hence, the effect of inner filter at 185 nm is not the major effect responsibleto the reduction of the degradation of atrazine. This can be proven by the fact that at 50 ppmbicarbonate, only 16% of the photons are absorbed by bicarbonate and 76% by water. On the otherhand, at the same concentration, 70% of HO radicals are scavenged by bicarbonate and only 30% areavailable for the degradation of atrazine. The amount of HO radicals at steady state was calculatedfrom equation 6.27 in the case of no bicarbonate in the system and in the presence of 50 ppm of80bicarbonate. Under such conditions, the steady state concentration of HO radicals were 1.46x10−11M and 1.30x12−12 M respectively. This shows that the concentration of HO decrease dramaticallyin the presence of only bicarbonate. Overall, the data presented indicate that the presence of 50ppm of bicarbonate reduces the degradation rate of atrazine by an order of magnitude.Figure 6.9: First order constant for the degradation of atrazine using UV/H2O2 in the presenceof different concentration of bicarbonate. The fuence rate was 0.29mW/cm2. Error barsrepresent the standard deviations of three replicates samples.6.2.8 Effect of sulfate on VUV and H2O2/UV processesSulfate does not react with HO radicals and for this reason it cannot interfere with the degradationrate of atrazine as HO scavenger. In addition the ε is negligible at 254 nm and this is the reasonwhy the addition of sulfate does not have any effects on the UV/H2O2 because it cannot act asHO• scavenger or as inner filter (Table 6.1, Figure 6.11). The degradation of atrazine however,showed some increases with increasing concentration of sulfate during VUV process. Figure 6.12shows the degradation of atrazine during VUV: in this case a slightly increment is observed. ForVUV experiments 1 ppm of methanol was added in the system since sulfate does not react withHO radicals and for this reason there was no any scavenger in the system. One ppm of methanol81Figure 6.10: First order constant for the degradation of atrazine using VUV in the presence ofdifferent concentration of bicarbonate. The fuence rate was 0.03mW/cm2. Error barsrepresent the standard deviations of three replicates samples.was added in order to decrease the rate of the degradation rate. As it is shown in Table 6.2),sulfate absorbs photons at 185 nm. At this wavelength sulfate can photolyze and can form HOradicals [111]. When sulfate absorbs photons with a wavelength shorter than 200 nm, the followingmanifold of reactions take place [111]:SO42−+hν<200nm → SO42−+ e−aq (6.33)O2+ e−aq → O2− (6.34)2H2O+O2−+O2 → H2O2+O2+2HO− (6.35)SO42−+H2O → SO2−+HO•+HO2 (6.36)The increment of the apparent kinetic constant in the presence of the highest concentration ofsulfate (50 ppm) was not dramatic (33%) because the absorption coefficient of sulfate at 185 nm is82Figure 6.11: First order constant for the degradation of atrazine using UV/H2O2 in the pres-ence of different concentration of sulfate. The fuence rate was 0.29mW/cm2. Errorbars represent the standard deviations of three replicates samples.one of the lowest among other inorganics (Table 6.2). For this reason, in the presence of 50 ppm ofsulfate only 9% of the photons are absorbed by sulfate.6.3 Effect of different inorganic ions on the degradation of atrazineIn the following section a comparison between the effects of different ions is presented. Having thevalue of ε at 254 and 185 nm of each ion and the kinetic constant rate between each ion and HOradicals, it was possible to evaluate the percentage of photons absorbed and HO radicals scavengedby each ions. In the following Table 6.3 different ions at the same concentration (1 ppm) areconsidered. For each of them the percentage of 254 and 185 nm photons absorbed was calculatedas well as the percentage of HO scavenged by each ions. In the calculation the effects of anymicropollutants was not taken into account.As it is shown the only ion that shows a scavenging effects is bicarbonate. In the calculationideally 100% of HO radicals are scavenged by HCO–3 but in the presence of any micropollutants83Figure 6.12: First order constant for the degradation of atrazine using VUV in the presenceof different concentration of sulfate. The fuence rate was 0.05mW/cm2 Error barsrepresent the standard deviations of three replicates samples.Table 6.3: Effect of different inorganic ions on the distribution of 254, 185 nm photons andon the HO radicals concentration.Compounds ppm% of 254 % of 185 % of HOnm photons nm photons radicals scavenedNO3− 1 0.01 6.6 0.0001HCO3− 1 0.0001 0.3 100SO42− 1 0.001 0.11 0.0001K+ 1 0.0001 6.9 0.0001Ca2+ 1 0.22 0.20 0.0001Na+ 1 0.0001 0 0.0001H2O2 10 1.3 0.0001H2O 0.0001 75.11 0.000184there will be a competition for HO radicals. Inorganic ions do not affect the distribution of 254 nmphotons, and for this reason all the photons reaching the sample are available for the photolysis ofH2O2. On the other hand, nitrate and potassium are able to act as inner filter for 185 nm photons,however the effect is not high since more than 75% of photons are available for the photolysis ofwater. In conclusion, the inorganic ions that can have a detrimental impact on the degradation ofmicropollutant are bicarbonate (since it can scavenge HO radicals), potassium and nitrate (sincethey can absorb 185 nm photons). In the case of bicarbonate, the relative effect will depend on theconcentration of this ion and on the type of micropollutant that needs to be removed. If the kineticconstant between the micropollutant and HO radicals is higher than the one between bicarbonateand HO radicals the micropollutant can compete more efficiently with the bicarbonate for the reac-tion with HO radicals. In addition, the concentration play an important role. If the concentration ofbicarbonate is much higher than the one of the micropollutant it will have a greater impact on theprocess because it will be able to scavenge more HO radicals. In the case of potassium and nitrate,the relative effect it will depend on the concentration of this ion and on the ε of the micropollutant.In general, all the organic micropollutants have a very high molar absorption coefficient at 185 nm,however they are in very low concentration. For this reason the relative concentration of potassiumand nitrate will play the key role in determining the amount of photons that are scavenged and thusthe amount of photons available for the photolysis of water.6.4 ConclusionsThe quantum yield of atrazine at 185 and 254 nmwas experimentally determined to be 5.6x10−2 and6.9x10−2 respectively. The molar absorption coefficient at 185 and 254 nm of atrazine and differentinorganic ions was measured. The results showed that atrazine has a high absorption coefficient forboth 185 nm and 254 nm radiation. All of the other inorganic ions, showed an absorption coefficientequal to zero for 254 nm except nitrate. This proves that during the UV/H2O2 process only nitrateplays a role on the distribution of photons in the system, thus it is the only ion that can act as aninner filter. On the other hand, at 185 nm radiation most of the ions were able to absorb 185 nmradiation. Amongst them, nitrate was the one that showed the highest capability to absorb photons85with a molar absorption coefficient equal to 5568 M−1 cm−1. The scavenging effects of differentions were then tested first during the UV/H2O2 process, and then with VUV. Nitrate showed adetrimental effect both with UV/H2O2 and with VUV. The reason is the high absorption coefficientthat nitrate has at 254 and 185 nm. In both cases Nitrate acts as an inner filter scavenging thephotons. On the other hand, the degradation rate of atrazine increases in the presence of UV only:nitrate is able to photolyze to produce HO radicals. The concentration though is very low (4.85 x10−16 mol L−1) and for this reason the increment of the degradation rate is minimal. Bicarbonatealso has detrimental effects. The presence of 50 ppm of bicarbonate reduces the degradation rateof atrazine by one order of magnitude. Sulfate again is able to photolyze at 185 nm to form HOradicals. For this reason the degradation rate of atrazine in the presence of 30 ppm of sulfate slightlyincreases. In addition, the ions that will have a higher impact on the degradation of micropollutantduring the VUV process are bicarbonate, potassium and nitrate. For this reason, in some caseswhen the concentration of these ions is relatively high compared to the one of the micropollutants,a pretreament can be added before the VUV, such as ion exchange. In this way the process willbe more efficient, since all the photons will be available for the photolysis of water and all the HOradicals will be available for the degradation of the micropollutant.86Chapter 7Quantitative effect of natural organicmatter on the efficacy of Vacuum UVoxidation of atrazineThis chapter presents the results of the work on the effect of NOM on the degradation of atrazineduring VUV and UV/H2O2 processes. First, the absorption coefficients at 185 nm and 254 nm ofNOM from two sources (Suwannee river and Nordic reservoir) were determined. Second, the scav-enging effects of NOM were determined using UV/H2O2 experiments in the 254 nm collimatedbeam setup and afterward with the VUV collimated beam described previously. The UV/H2O2 ex-periments were used as control and for the purpose of calculating the reaction rate between NOMand HO radicals: this was necessary because in the VUV process the degradation of the microp-ollutants occur due to the reaction with HO radicals and the photolysis with 185 nm radiationsimultaneously. Further, some experiments were conducted with various concentration of methanolto assess if NOM during the VUV process act as photosensitizer. Finally, a simple model was de-veloped to calculate the steady state concentration of HO radicals in the system, with the percentageof photons and radicals scavenged by NOM.877.1 Materials and methods7.1.1 Water samples and chemicalsContaminated water samples were prepared with millipore water and 0.1 ppm of atrazine (SigmaAldrich) with various concentrations of NOM from Suwannee river and Nordic reservoir (RO isola-tion, International humic substances society). The concentrations of TOC used were 3, 7, 10 ppm.For the experiments conducted to assess the capability of NOM to act as sensitizer, contaminatedwater was prepared adding different concentrations (3, 7 and 10 ppm) of methanol (Fischer scien-tific) in millipore water. For the UV/H2O2 experiments, 10 ppm of H2O2 (Fischer 30%) was addedin the samples.7.1.2 UV/H2O2 and VUV irradiationsFor UV/H2O2 experiments the samples were irradiated with a conventional collimated beam emit-ting at 254 nm (described in Section Section 4.1.2.1) and 10 ppm of H2O2 were added in the con-taminated water. Samples were irradiated for different times (5, 10, 15, 20 and 25 minutes). Thesamples were exposed to a UV fluence from 0 to 435 mJ/cm2, which was calculated multiplyingthe fluence rate (obtained with an actinometry) by the time of exposure. The experiments involvingVUV radiation were conducted in a VUV collimated beam emitting at 185 nm (described in SectionSection 4.1.3). The sample volume was 25 mL. Samples were irradiated for different times (5, 10,15, 20 25 and 30 minutes). For the VUV irradiation, the VUV fluence utilized was from 0 to 45mJ/cm2 which was calculated multiplying the fluence rate (obtained with an actinometry) by thetime of exposure.7.1.3 Analytical methodsThe degradation of atrazine was quantified using a high-performance liquid chromatograph (HPLC,Dionex 2695) equipped with C-18 column (4-micronmeter particle diameter) and a UV detector.Methanol/water/acetic acid (58%:40%:2% v/v) were used as mobile phase. The injected samplevolume was 100 µL. The flow rate of the mobile phase was 1 mL min−1 and analyses were con-ducted at λ = 280 nm by a UV detector.88The absorbance of atrazine, H2O2, NOM, methanol and atrazine was measured at various con-centrations following the procedures described in section Section 4.2.4. The concentration of H2O2was measured following the procedure described in section Section 4.2.3. The concentration of to-tal organic carbon (TOC) was measured using a TOC-VCPH Shimadzu analyzer. pH was measuredusing a pH meter (Thermo Orion PerpHecT LogR 1330 meter, 9206BN electrode). The concentra-tion of dissolved oxygen was monitored with a dissolved oxygen meter (YSI 52 meter, YSI 5909probe).7.2 Results and discussion7.2.1 Molar absorption coefficients at 185 nm and 254 nmTable 7.1 shows the absorption coefficients at 254 and 185 nm for the two sources of NOM,Atrazine, H2O2 and methanol. The raw data of the absorbance measurements are reported in Ap-pendix D. The absorption coefficient of NOM is higher than all the inorganic ions, and H2O2. It isimportant to note that the absorption coefficient of H2O2 at 254 nm is considerably lower than theone of NOM. For this reason it is expected that the presence of NOM will reduce the efficiency ofthe process since it will scavenge 254 nm photons and consequently less photons will be availablefor the photolysis of H2O2. The same trend is observed for the absorption coefficient at 185 nm:the data indicate that absorption coefficients at 185 nm are noticeably high, and NOM could theo-retically compete with water for the absorption of 185 nm photons and thus, less HO radicals willbe formed. The absorbance of both types of NOM showed a linear response following the LambertBeer law. This behavior indicates that even at high concentrations up to 10 ppm, NOM does notform aggregates that show varying optical properties. Aqueous humic substances exist primarilyas soluble species, but they form colloids at high concentrations. Myneni et al. [112] observedthat Suwannee river does not exhibit any measurable organized structure below a concentration of1.0 gr C L−1. As the concentration increases Suwannee river NOM forms aggregates of differentshapes and sizes such as globular, ringlike and sheetlike structures. These aggregates show differentchemical and physical properties including different absorbance from the homogeneous NOM. The89results obtained in this research are in accordance with the one of the literature. At concentration of10 ppm the absorption coefficient was still linear meaning that no aggregates/colloids were formed.If colloids had been formed, the absorbance would have not been linear since the absorbance ofaggregates and colloids do not follow Lambert Beer law due to interference and scattering phenom-ena. For this reason, the results obtained in this research can theoretically apply with an extensiverange of NOM concentration typically found in surface water.Table 7.1: Absorption coefficient of various compounds at 254 nm and 185 nm.CompoundsAbsorption coefficient 254 nm Absorption coefficient 185 nm(M−1 cm−1) (M−1 cm−1)Suwannee river 116 1537Nordic reservoir 638 1137H2O2 19.9 341Atrazine 3774 25370Methanol 0 237Figure 7.1 shows the absorption spectra of the Suwannee river and Nordic reservoir NOM iso-lates. Evidently the NOM from these two sources show various spectra/responses. This differencecan be attributed to various structures and compositions of the NOM from these two sources. Nordicreservoir NOM shows a higher absorption at wavelengths greater than 210 nm. The Suwannee riverNOM, on the other hand, shows a noticeably higher absorption coefficient at wavelengths lowerthan 190 nm. It is possible to make some considerations on the NOM structure from the absorptionspectra obtained. Korshin et al. [113] proposed that the electronic transfer (ET) and the benzenoidband (BZ) of the NOM could be extracted from NOM’s absorbance of UV at 254 and 203 nm,respectively. Electrons in the ground state (designated 1A1g) may absorb quanta of light and bepromoted to the pi antibonding orbital corresponding to energy levels denoted 1B1u and 1B2u. Thetransition from 1A1g to 1B1u generates an absorption band centered at 203 nm, this band is referred asBZ. Another band corresponding to a transition from 1A1g to 1B2u is centered at 253 nm and has lowintensity because of very strong quantum yield prohibition. The intensity of the ET band is greatlyaffected by the presence on the ring of polar functional groups such as hydroxyl, carbonyl, carboxyl90and ester groups. By contrast, non-polar aliphatic groups attached to the ring do not increase theintensity of the ET transition significantly. The BZ band is also sensitive to substitution on the ring,but less so than the ET band. Based on the ratio of the absorbance of ET band to the absorbance ofBZ band (A254nm/A203nm) the degree of activation of the aromatic rings can be interpreted. A lowA254nm/A203nm value was said to be representative of compounds in which the aromatic rings weresubstituted predominately with aliphatic functional groups. Increasing A254nm/A203nm was said tobe indication of higher substitution of aromatic rings in NOM with hydroxyl, carbonyl ester andcarboxyl groups.For Suwannee river the ratio of A254nm/A203nm was found to be 0.5 and for Nordic was found tobe 0.83, this shows that Nordic NOM has a higher degree of substitution of aromatic rings. Fromthese values it is possible to make some considerations about the composition of Suwannee riverNOM and Nordic Reservoir NOM. In general humic acid is hydrophobic and aromatic; fulvic acids,on the other hand, are more soluble and contain more acidic functional groups such as carboxylicacids that increase the overall charge density. From the spectra obtained , and from the value of(A254nm/A203nm) it can be stated that Suwannee river has a higher concentration of hydrocarbonsand aromatics with low substitutions, Nordic NOM has a higher concentration of fulvic acid with ahigh degrees of functional groups such as carbonyl, hydroxyl [114].Another important property that can be extrapolated from the spectra is the Specific ultravioletabsorbance (SUVA). SUVA254 is defined as the UV absorbance at 254 nm measured in inverse me-ters (m−1) divided by the TOC concentration in mg L−1. SUVA is an average absorptivity for all themolecules that comprise the TOC in a water sample and has been used as a surrogate measurementfor DOC aromaticity. With the results obtained in this research the SUVA obtained for Suwanneeriver and for Nordic reservoir NOM was found to be 3.4 and 4.2 L mg−1 m−1 respectively. The oneof Suwannee was in accordance with the literature [115]. As Nordic, to the best of my knowledge,there is no study on the physical properties and the reactivity of this type of NOM. Different lit-erature studies [116] showed that low SUVA (2 or lower) is correlated with low molecular weight(MW) hydrophylic acid hydrocarbons, SUVA values on the order of 3 are correlated with fulvicacids, while values on the order of 4 to 6 are correlated with humic acids components. From the91results obtained it can be stated that Suwannee river is mainly made by fulvic acids. On the otherhand, Nordic NOM is mainly made by humic acids with acid functional groups.Figure 7.1: Absorption spectra of 9 ppm of Suwannee river NOM and of Nordic NOM.7.2.2 UV/H2O2 irradiationsThe scavenging effect of NOM toward HO radicals was measured and Figure 7.2 shows the resultsobtained. The obtained results were different for Nordic and Suwannee NOM, meaning that thereactivity toward HO radicals depends on the NOM structure. During the degradation of atrazinewith the UV/H2O2 process, in the presence of NOM, the following reactions take place:92H2O2+hν → 2HO• (φ = 1) (7.1)NOM+HO• → byproducts (kHO,TOC = determinate in this work) (7.2)Atrazine+HO• → byproducts (k = 2.6×109M−1s−1) (7.3)NOM+hν → byproducts (7.4)Figure 7.2: First order rate constant for the degradation of atrazine using UV/H2O2 with vari-ous concentrations of NOM Suwannee river, and Nordic reservoir. The fluence rate was0.29mW/cm2. Error bars represent the standard deviations of three replicates samples.The degradation rate of atrazine in the samples prepared with Nordic NOM was lower, indica-tion that this type of NOM has a higher reactivity toward HO radicals. That is, HO• reacts prefer-entially with the NOM from the Nordic reservoir and there will be less HO• for the reaction with93atrazine. It is known from the literature that the HO radical rate constants are positively correlatedwith aromatic carbon content, and inversely correlated with aliphatic carbon content [116]. In otherwords, HO radicals were found to react preferentially with the aromatic constituent of NOM andspecifically electron enriched aromatics. Furthermore, it was observed a slight dependence of HO•reactivity upon SUVA [117] and this can represent the dependence of the reactivity between HO•and NOM on the molecular weight of the NOM. Higher SUVA results in an higher reactivity ofNOM towards HO radicals. The correlation may be dependent on the electron rich carbon-carbondouble bonds on HO radicals reactions. Unsaturated bonds are highly reactive in hydrophobic andhydrophilic organic acids. This is in line with the results obtained because Nordic showed a higherreactivity with HO radicals and a higher SUVA (Figure 7.1). With UV/H2O2 experiments it waspossible to calculate the kinetic constant between HO• and Suwannee river and between HO andNordic.7.2.2.1 Determination of the kinetic constant between NOM and HO radicalsControlled experiments were conducted to gather data for estimating the two unknown reactionrate constants, kHO,Nordic and kHO,Suwannee. In an advanced oxidation system, atrazine is primarilydegraded by HO radicals, thus the reaction with other species was neglected in this calculation. Inaddition, the NOM was assumed to be degraded by HO radicals only. Direct photolysis of NOMwas neglected because prior experiments demonstrated that, when treatment was conducted withoutH2O2, there was no NOM mineralization.For this reason, this calculation assumes that TOC was constant over the duration of the treatment.The degradation of atrazine was followed at different concentrations of TOC. In these experi-ments atrazine served as HO• probe, thus by monitoring the concentration of atrazine, competitionkinetics was used to estimate the reaction rate constant for the reaction between HO• and TOC.The rate of depletion of atrazine could be described by a first order kinetic with respect to theconcentration of atrazine [109]:d[Atz]dt=−kobs[Atz]−φ [Atz]εAtz,254nmG254nm[Atz] (7.5)94where kobs is equal to:kobs = kHO,atz[HO•]ss (7.6)where kHO,atz is the kinetic rate constant between HO• and atrazine, [HO•]ss is the steady stateconcentration of HO radicals, φAtz is the quantum yield of atrazine, [Atz] is the concentration ofatrazine and εAtz,254nm is the absorption coefficient of atrazine at 254 nm and G254nm is the flux at254 nm. In this way it was possible to calculate the steady state concentration of HO radicals whichwas found to be 6.95x10−13 M in the case of the samples prepared with Suwannee river NOM and6.03x10−13 M in the case of the samples prepared by Nordic reservoir NOM.Knowing all these parameters it was possible to calculate the k between HO and NOM (Suwan-nee and Nordic) (kHO,TOC) with the following equation:dHO•dt= 2kH2O2,254nmφH2O2 − kHO,TOC[TOC][HO•]− kHO,atraz[HO•][Atz]− kHO,H2O2 [H2O2][HO] (7.7)where kH2O2,254nm is the specific rate of light absorption of H2O2 and is equal to:kH2O2,254nm =EpεH2O2 [1−10−αtot,254nmz]×1000αtot,254nmz(7.8)and α is defined according to:αtot,254 = [H2O2]εH2O2 +[TOC]εTOC (7.9)where kHO,TOC is the kinetic constant between HO radicals and NOM, [TOC] is the initial con-centration of NOM, kHO,atraz is the kinetic constant between HO radicals and atrazine, [HO] is thesteady state concentration of HO radicals (previously calculated), [Atz] is the initial concentrationof atrazine, εH2O2 is the molar absorption coefficient of H2O2 at 254 nm, φH2O2 is the quantum yieldof H2O2 (0.5), [H2O2] is the H2O2 initial concentration, kHO,H2O2 is the kinetic constant betweenHO and H2O2, Ep is the incident photon fluence rate, z the path-lenght, αtot,254nm is the total water95absorption coefficient, εH2O2 is the absorption coefficient of H2O2 at 254 nm, and εTOC is the oneof NOM. Knowing all these values and making the approximation the concentration of HO radicalswas steady it was possible to calculate the k for Suwannee river and Nordic river. For Suwanneeriver the kinetic constant was found to be 2.7x104 L mg−1 s−1. This value was in a good agreementwith other values obtained from other researches [118], [119]. The value for the kinetic constantfor Nordic NOM was found to be 3.7 x 104 Lmg−1 s−1, slightly higher than the one of Suwannee.Figure 7.2 shows the apparent first order constant of atrazine degradation with various concen-trations of TOC. As it is shown in the presence of different concentration of NOM, the rate constantdecrease. With 9 ppm of NOM the apparent kinetic constant of atrazine decreased by 77%. Thissignificant reduction of the degradation efficiency was because of the scavenging effects and the254 nm photons absorption. To understand throughly the mechanism of the process, the % of 254nm photons absorbed was calculated using the following equation:%254nmphotons= CNOMεNOM,254CNOMεNOM,254+CAtrzεAtrz,254+CH2O2εH2O2,254(7.10)where CNOM is the concentration of NOM, εNOM,254 is the absorption coefficient of NOM at 254nm, CAtrz is the concentration of atrazine, εAtrz,254 is the absorption coefficient of atrazine at 254nm, CH2O2 is the concentration of H2O2 and εH2O2,254 is the absorption coefficient of H2O2 at 254nm.Also, the % of HO radicals scavenged was calculated using the following equation:%HOscavenged = CNOMkNOM,HOCNOMkNOM,HO+CAtrzkAtrz,HO+CH2O2kH2O2,HO(7.11)where kNOM,HO is the rate constant between NOM and HO radicals, kAtrz,HO is the rate constantbetween atrazine and HO radicals and kH2O2HO is the rate constant between HO and H2O2 [109].Table 7.2 shows the distribution of 254 nm photons, and HO radicals in UV/H2O2 system duringthe degradation of atrazine with various concentrations of Nordic NOM.The same calculation was conducted for Suwannee NOM (Table 7.3)As it is shown in Tables 7.2 and 7.3, Suwannee river absorbs less 254 nm photons than Nordic96Table 7.2: distribution of HO radicals and 254 nm photons during UV/H2O2 with NordicNOM.Compounds % 254 nm photons % of HO•NOM 3 ppm 18 90Atrazine 0.25 1.47H2O2 1.05 7.98Compounds % 254 nm photons % of HO•NOM 6 ppm 32 95Atrazine 0.23 0.77H2O2 1 4.19Compounds % 254 nm photons % of HO•NOM 9 ppm 45 96.6Atrazine 0.21 0.52H2O2 0.40 2.84Table 7.3: distribution of HO radicals and 254 nm photons during UV/H2O2 with SuwanneeNOM.Compounds % 254 nm photons % of HO•NOM 3 ppm 6.41 87Atrazine 0.27 1.97H2O2 1.21 10Compounds % 254 nm photons % of HO•NOM 6 ppm 12 93Atrazine 0.26 1H2O2 1.17 5.69Compounds % 254 nm photons % of HO•NOM 9 ppm 18 95Atrazine 0.25 0.71H2O2 1.13 4.597NOM. This will have a direct effect on the UV/H2O2 process because for Nordic NOM fewer pho-tons are available for the photolysis of H2O2 and, consequently, fewer HO radicals are formed.This is in line with the results obtained during UV/H2O2 irradiation (Figure 7.2) where NordicNOM had a greater impact on the degradation of atrazine. In addition, the kinetic constant betweenHO radicals and NOM was found to be higher than the one of Suwannee and for this reason NordicNOM scavenge more HO radicals than Suwannee NOM. On the other hand, since NOM fromNordic reservoir absorbs three times more photons at 254 nm than NOM from Suwannee river, it isexpected than NOM from Nordic reservoir would have a considerably higher impact on the degra-dation of micropollutants during the UV/H2O2 than the one Suwannee river has. The differencebetween the two NOM, however, was found to be minimal. This can be explained by the fact thatthe capability to scavenge HO radicals plays a major role during the UV/H2O2 than the inner filtereffect. Since the kinetic constant between the two types of NOM and HO radicals was found to bevery similar, the effect of both type of NOM was found to be very similar, despite the importantdifference between the two NOM in terms of capability to absorb photons.7.2.3 VUV irradiationsVUV irradiations at 185 nm of atrazine were conducted with various concentrations of NOM(Nordic and Suwannee) and methanol. The purpose of monitoring the degradation of atrazine withmethanol and comparing the trend with the one obtained with NOM was to establish if NOM canbe a sensitizer.Figure 7.3 shows the degradation of atrazine with VUV at various NOM concentrations. Aswith the UV/H2O2, the degradation of atrazine decreased with increasing NOM concentration. Thepercentage of HO scavenged was again calculated according to:%HOscavenged = CNOMkNOM,HOCNOMkNOM,HO+CAtrzkAtrz,HO(7.12)where CNOM and CAtrz are the initial concentrations of NOM and atrazine, respectively, kNOM,HO isthe rate constant between NOM and HO radicals, kAtrz,HO is the rate constant between atrazine andHO radicals [109].98Figure 7.3: First order rate constant for the degradation of atrazine using VUV with vari-ous concentrations of NOM (Suwannee river, and Nordic reservoir) and Methanol. Thefluence rate was 0.03mW/cm2. Error bars represent the standard deviations of threereplicates samples.The percentage of 185 nm photons was calculated according to:%185nmphotons= CNOMεNOM,185CNOMεNOM,185+CAtrzεAtrz,185+CH2OεH2O,185(7.13)where CNOM is the concentration of NOM, εNOM,185 is the absorption coefficient of NOM at 185nm, CAtrz is the concentration of atrazine, εAtrz,185 is the absorption coefficient of atrazine at 185 nm,CH2O2 is the concentration of H2O and εH2O,185nm is the absorption coefficient of H2O at 185nm.Table 7.4 and Table 7.5 show the 185 nm photons distribution and the % of HO scavenged byNOM, Nordic and Suwannee respectively during VUV.The major effect of this decrease is again the scavenging effect of HO radicals because even99with 9 ppm of NOM 46.12% of 185 nm photons are available for the photolysis of water and,consequently, for the production of HO radicals. On the other hand, 99.25% of HO radicals arescavenged by NOM and only 0.74% are available for the degradation of atrazine.Table 7.4: distribution of HO radicals and 185 nm photons during VUV process with NordicNOM.Compounds % 185 nm photons % of HO•NOM 3 ppm 26 97.78Atrazine 0.83 2.21H2O 69Compounds % 185 nm photons % of HO•NOM 6 ppm 42.21 98.8Atrazine 0.67 1.11H2O 55Compounds % 185 nm photons % of HO•NOM 9 ppm 52.64 99.25Atrazine 0.55 0.74H2O 46.16Table 7.5: distribution of HO radicals and 185 nm photons during VUV process with Suwan-nee NOM.Compounds % 185 nm photons % of HO•NOM 3 ppm 21 96.45Atrazine 0.85 3.54H2O 73.46Compounds % 185 nm photons % of HO•NOM 6 ppm 34 98.18Atrazine 1.11 1.80H2O 61Compounds % 185 nm photons % of HO•NOM 9 ppm 44.8 98.79Atrazine 0.64 1.20H2O 53.15100NOM has a higher influence on VUV because during UV/H2O2 some HO are scavenged byH2O2 which is able to produce other oxidizing species (such as O–2•) that can react back to formH2O2 which is the species that can photolyze and produce HO radicals. In the case of VUV the pro-duction of H2O2 is minimal (less than 1 ppm) and for this reason H2O2 cannot scavenge effectivelyHO radicals. In addition during VUV process Nordic has a higher impact on the degradation ofatrazine, and this can be explained again with the fact that the kinetic constant between HO radicalsand Nordic is higher than the one between HO and Suwannee. This means that Nordic NOM has ahigher affinity with HO than Suwannee has. This trend can be observed only when NOM is at lowconcentration (1 ppm) since already when the concentration is higher (2 ppm) the first rate constantof atrazine goes promptly to zero for both type of NOM. In addition Nordic NOM is able to absorbmore photons than Suwannee river, and thus, less photons are available for the photolysis of water.The concentration of HO radicals in VUV system with only water and atrazine was calculatedwith the following equation:CHO(x) =G185(x)(φ1,185+φ2,185) · εH2O,185 · logCH2O∑Ni=1 ki−HO ·Ci(7.14)where CHO(x) is the local HO• concentration, G185(x) is the incident radiation, φ1,185 is the quantumyield of the photochemical homolysis of water (0.33), and φ2,185 is the quantum yield of the photo-chemical ionization of water (0.045) and εH2O,185 is the absorbtion coefficient of water at 185 nm(1 M−1 cm−1). For samples with only atrazine and milli-Q water the concentration of HO radicalswas found 1.46x10−11 M. For samples with 6 ppm of NOM, because of its pronounced scavengingeffect, the steady state concentration of HO reduced to 1.47x 10−13 M. During VUV process, there-fore, the presence of NOM affects the efficiency due its ability to scavenge HO radicals. BecauseNOM is able to scavenge HO radicals so promptly, the change of TOC during the process wasmonitored to assess if mineralization can occur. No change was observed because during a shortirradiation time (25 min) and with low doses the mineralization of NOM does not occur.1017.2.4 NOM sensitization effect during VUV treatmentLiterature studies have shown that the presence of low concentration of dissolved organic matter canpromote the phototransformation of various organics [120]. These results were obtained with solarirradiation without the presence oxidants. In this study some tests were carried out with the sameconcentration of dissolved organic carbon (DOC) deriving from dissolved methanol. The purposeof these tests was to assess if NOM can be a sensitizer also with VUV radiation. It is interesting thatmethanol has the same influence as NOM on the degradation of atrazine (Figure 7.3). This indicatesthat NOM can be a sensitizer. Methanol has a lower absorption coefficient at 185 nm than NOM(237 M−1 cm−1 ) and for this reason with methanol more HO are formed than with NOM sincemore photons are available for the photolysis of water. Also the rate constant of methanol with HOradicals, is greater than the one of NOM (kNOM−HO= 2.6x108 M−1 s−1and kMeOH−HO=9.7x108 M−1s−1). In the case of methanol the absorption coefficient is approximately 5 times lower than the oneof NOM, so for this reason it is expected to find a considerably higher concentration of HO radicalsin the system, since more photons are available for the water photolysis. Thus, the degradationof atrazine should be faster in the presence of MeOH than in the presence of NOM (when theyare in the same concentration). The distribution of 185 nm photons in the samples prepared withMillipore water, atrazine, methanol and NOM was calculate. In the case of the samples preparedwith methanol, 0.94% of photons were absorbed by methanol, 1% by atrazine and 90% by water.This demonstrates that methanol does not scavenge 185 nm photons and for this reason all thephotons are absorbed by water. On the other hand hand, in the case of NOM 55% of 185 nmphotons were absorbed by NOM, 43% by water and 1% by atrazine. In this case NOM has a higherimpact on the distribution of photons, and for this reason less 185 nm photons are available for thephotolysis of water. In order to understand if methanol is a greater HO scavenger the distribution ofHO radicals was calculated in the presence of NOM and in the presence of NOM. In this case, 10ppm of MeOH scavenged 99.5% HO radicals, and NOM 99.3%. The results showed that, althoughmethanol has an higher rate constant with HO radicals, the scavenging effects that can create is verysimilar to the one of NOM. For this reason it is expected that the degradation of atrazine would befaster in the presence of Methanol than in the presence of NOM, since more photons are available102for the water photolysis and for this reason more HO radicals are formed.The fact that NOM has the same influence as methanol is because it can be a sensitizer. Pho-tosensitized transformations of organics chemicals are mostly started through light absorption ofchromophores in NOM and the formation of reactive species. The formation of these species suchas HO radicals, singlet molecular oxygen and solvated electrons has been studied and reported[121]. In addition a short-lived triplet states of DOM (3DOM∗) have also been reported [122]. Allthese species can have an influence on the degradation of organics.In VUV process, NOM can be involved in the following reactions:NOM+hν → byproducts (7.15)NOM+hν → 3NOM∗ (7.16)3NOM∗+3 O2 → 1O2 (7.17)3NOM∗+S → 3S∗ (7.18)NOM+HO• → byproducts (7.19)where 3NOM∗ is the excited triplet form of NOM, 3O2 is the triplet molecular oxygen,1O2 is thesinglet molecular oxygen, S is the compound of interest (in our case atrazine), 3S∗ is the excitedtriplet form of S, and Sox is the oxidized product of S. According to this study the principal reactionis the reaction (7.19). The sensitization path is the less important one because NOM in complexcan be a scavenger and has an harmful effect on the degradation of micropollutant. This studyshows that NOM, compared to other organics, can act as sensitizer and decrease the degradationof atrazine less than expected. However, in complex, during VUV process, NOM has an overalldetrimental effect on the degradation of atrazine acting as HO scavenger and inner filter for 185 nmphotons.7.3 ConclusionsThe presence of NOM negatively affects the degradation rate of atrazine by consuming HO radicalsand partially absorbing VUV photons. HO radicals predominantly react with, and are scavenged103by, 9 ppm of NOM, and only 0.64% of the generated HO radicals will remain to react with atrazine.Because of the pronounced scavenging effect of NOM, the steady state concentration of HO radicalsdecreases to 1.47x10−13 M. In the presence of 9 ppm of Nordic NOM 96.6% of 254 nm photonsare absorbed by Nordic NOM, 0.52%by atrazine, and 2.84% by H2O2; in the case of Suwanneeriver NOM, 95% of 254 nm photons are absorbed by NOM, 0.71% by atrazine and 4.5% by H2O2.About the interaction with VUV photons with 9 ppm of Nordic NOM, 46.16% of the photons areabsorbed by water, 52.64% by NOM, and 0.55% by atrazine. For Suwannee river, 44.8% photonsare absorbed by NOM, 0.64% by atrazine and 53.15% by water. The overall result is that thedegradation rate of atrazine reduces by 99.1% with 6 ppm of NOM and this reduction is primarilydue to the the scavenging effects of NOM rather than its ability to act as an inner filter for VUVradiation. In addition it has been proved than NOM can act as a sensitizer (compared to otherorganics), but in complex, this effect is lower and not comparable to its ability to react with HOradicals. With VUV, NOM mostly photolyzes to form by-products rather than forming triplet stateDOM species, and the amount of reactive oxidants that are formed by its interaction with radiationare lower than those that comes from the photolysis of water. The results of these studies can provideinformation about the possibility of simple pre treatments before the VUV process depending onthe water source.104Chapter 8The effect of chloride on Vacuum-UVand UV/H2O2 photo-induceddegradation of phenolThe effect of chloride was studied separately from other inorganic ions since its behavior was foundto be the combination of different effects. As discussed previously, chloride can form chloride rad-icals (Cl•) upon reaction with HO radicals. These radicals can potentially react with the substrateleading to the formation of chlorinated byproducts and contribute to the degradation of the modelmicropollutant. In this chapter phenol was used as target micropollutant since its byproduct dis-tribution after the oxidation with HO radicals is well known. Hence, the distribution of phenolbyproducts was monitored in order to delineate the mechanisms and reaction pathways of phenoloxidation in the presence of chloride ion. The presence of chlorinated byproducts would prove thefact that chloride radicals are formed and can react with the substrate.1058.1 Materials and methods8.1.1 Water samples and chemicalsContaminated water samples were prepared with Millipore water and 1 ppm of phenol (SigmaAldrich) with various concentrations of sodium chloride (Fischer 99%). The concentrations ofchloride ions used were 0, 0.25, 0.50, 0.75, 1 and 1.50 mM. For UV/H2O2 experiments, 10 ppm ofH2O2 (Fischer 30%) was added.8.1.2 UV/H2O2 and VUV irradiationsFor the UV/H2O2 experiments the samples prepared were spiked with 10 ppm of H2O2 and irradi-ated with a conventional collimated beam emitting at 254 nm (described in section Section 4.1.2.1).Aliquots were collected at 10 minutes intervals and the irradiation lasted 60 minutes. The exper-iments involving VUV radiation were conducted in a VUV collimated beam emitting at 185 nm(described in Section 4.1.3). The sample volume was 25 mL. In this case, at 0, 10, 20, 30, and 40minutes a sample was taken.8.1.3 Analytical methodsThe degradation and the by-products of phenol were quantified using a high-performance liquidchromatograph (HPLC, Dionex 2695) equipped with C-18 column (4-micronmeter particle diame-ter) and a UV detector. The analysis was carried out using a gradient, by which the first 8 minuteswere operated with an eluent composition of 90% of water and 10% acetonitrile, followed by a 4-minute period with a composition of 50% of acetonitrile and water, stabilized for another 2 minutesbefore ramping back down to 90% water and 10% acetonitrile. The total runtime was 16 minutes.The absorbance of chloride and phenol was measured following the procedure described in Sec-tion 4.2.4. The concentration of H2O2 was measured using the triiodide method ( Section 4.2.3).The pH of the system was measured using a pH meter (Thermo Orion PerpHecT LogR 1330 meter,9206BN electrode) over the degradation time with a sample made by 1 ppm of phenol with variousconcentrations of chloride. The concentration of dissolved oxygen was monitored with a dissolvedoxygen meter (YSI 52 meter, YSI 5909 probe).1068.2 Results and discussion8.2.1 Molar absorptions coefficients at 185 nm and 254 nmThe data of the absorbance measurements are reported in Appendix D. The absorption coefficientsof chloride at 254 and 185 nm were 0 and 2791 M−1 cm−1, respectively. This result showed thatchloride can absorb photons at 185 nm and it can act as inner filter leaving fewer photons availablefor the photolysis of water. Also, chloride photolyze to form reactive chloride radicals [77].Cl−+hν185nm → Cl•+ e−acq (8.1)Cl−+hν185nm → Cl∗ (8.2)where Cl∗ is the excited state of Cl− which can return to the ground state after release of thermicenergy. Because chlorine has such a high absorption coefficient, it can greatly influence the distri-bution of 185 nm photons, the percentage of photons absorbed by various concentration of chloridewas calculated as:%185nm= εClCClεH2OCH2O+ εClCCl(8.3)where εCl is the absorption coefficient of chloride at 185 nm, CCl is the concentration of chloride,εH2O is the absorption coefficient of water and CH2O is the concentration of water. Table 8.1 showsthat at 0.50 mM of chloride, water absorbs 60% of the photons and the remaining 40% are absorbedby chloride. On the other hand, the absorption coefficient of phenol at 254 and 185 nm was foundto be 516 and 320000 M−1 cm−1 respectively.8.2.2 Degradation of phenol with UV/H2O2The irradiations conducted in the UV/H2O2 system showed that the degradation of phenol does notchange with chloride as it is shown in Figure 8.1. Chloride does not absorb photons at 254 nm thusit does not act as inner filter. On the other hand, some studies have reported that in a system with107Table 8.1: Percentage of photons at 185 nm absorbed by water and by various concentrationsof chloride ions.Chloride concentration mM% of photons % of photonsabsorbed by water absorbed by chloride0 100 00.25 60 400.50 40 600.75 30 701.00 25 751.50 20 80Cl− and HO • the following reaction can occur:Cl−+HO• ⇀↽ ClOH−• (k = 0.7M−1s−1) (8.4)ClOH−•+H+ ⇀↽ Cl•+H2O (k = 1.6×107M−1s−1) (8.5)Cl•+Cl− ⇀↽ Cl2•− (k = 1.9×105M−1s−1) (8.6)Because the degradation rate did not change with various concentration of chloride, it is spec-ulated that chloride reacts with HO• so, the concentration of HO radicals decreases while chlorideradicals are formed. So, the net effect could be the same overall rate constant. In fact, if chlorineradicals would had been formed without scavenging HO radicals the degradation rate would haveincreased since chloride radicals can assist HO• for the degradation of organics. To prove this con-cept phenol by-products were monitored. The byproducts that are formed when HO radicals inter-act with phenol are hydroquinone, catechol, resorcinol, hydroxyhydroquinon and p-benzochinone[123]. In the phenol molecule, the HO radicals disturb the electronic distribution density on thearomatic ring and attack in ortho and para positions: for this reason intermediate are formed andthe ring of phenol eventually can be broken. If chloride radicals are present in the system witha concentration comparable to the one of HO radicals, they could react with phenol and chlori-nated byproducts should be formed. The formation of byproducts with Cl radicals follows the samemechanism of the one with HO radicals. Therefore, the end products include 2-chlorophenol, 3-108Figure 8.1: First order rate constant for the degradation of phenol using UV/H2O2 with vari-ous concentrations of chloride ion. The fuence rate was 0.25mW/cm2. Error bars repre-sent the standard deviations of three replicates samples.chlorophenol and 4-chlorophenol [1]. Figure 8.2 shows the schematic degradation pathway andformation of byproducts during the oxidation of phenol with HO radicals oxidation.First, the byproducts formed during the degradation of phenol during UV/H2O2 were moni-tored without the addition of any chloride (Figure 8.3). This first set of experiments showed thedistribution of byproducts obtained during the oxidation of phenol with HO radicals.The primary byproducts were predominately catechol, p-benzoquinone with trace quantitiesof hydroquinone and resorcinol. The phenol concentration decreases within the 60 minute period,from an initial concentration of 1.075 ppm to a final concentration of 0.068 ppm. Catechol increasedfrom a concentration of 0 ppm to a maximum of 0.215 ppm after 30 minutes, and then decreasedagain as it is mineralized at a faster rate than it is produced until reaching a final concentration of0.059 ppm. Similarly, p-benzoquinone increased to a maximum of 0.254 ppm after 20 min before109Figure 8.2: Schematic decomposition path of aqueous phenol [1].decreasing to a final value of 0.009 ppm at 60 min. Hydroquinone and resorcinol each reached amaximum of 0.030 ppm and 0.003 ppm, respectively. All the byproducts reached a peak in concen-tration and after that they decreased because of being further oxidized to short chain organic acids.As it is shown in Figure 8.3, catechol is one of the primary oxidation product, indicating that hy-droxilation takes place predominately in the ortho-position. It can be concluded that for the phenolhydroxylation the attack of HO radicals is regioselective (favorable in orthoposition, followed byparaposition). The formation of resorcinol, on the other hand, is the result of different pathwayless favorable than the simple HO attack which is the results of phenol pure photolysis. Duringthe UV/H2O2 treatment, phenol can be degraded by simple photolysis. The ε of phenol is 516110Figure 8.3: Phenol degradation (0.1 ppm) and byproducts formation during the UV/H2O2 ox-idation process. Error bars represent the standard deviations of three replicates samples.M−1cm−1 [123] and the quantum yield of phenol in aerated solution has been found to vary inthe interval 0.12-0.02 at pH 1.6-13.2 [123]. When phenol absorbs 254 nm radiation the followingreaction takes place:Phenol+hν →1 phenol (8.7)where 1phenol is its first excited singlet state. 1phenol can then decay by fluorescence to the groundstate:1phenol→ phenol+hv (8.8)111or through intersystem crossing:1phenol→3 phenol (8.9)or through chemical dissociation into phenoxy radical (ArO•) and H•.1phenol→ ArO•+H• (8.10)or photoionization1phenol→ ArO•+ e−aqu (8.11)where 3phenol is the triplet state. The e−aqu formed during phenol photolysis can be scavenged by aproton or by dissolved oxygen.H++ e−aq → H• (8.12)H•+O2 → HO2• (8.13)O2+ e−aq → O2•− (8.14)O2•−+H+ → HO2• (8.15)The phenoxy radical finally can react with HO2• leading to the phenoxy-OOH adduct whichcan yield to dihydroxybenzene which can be transformed to resorcinol with the reaction of twoprotons or electrons [124]. This photolysis pathway would lead to the formation of resorcinol, butis less favorable than the reaction with HO radicals. For this reason, the most favorable pathway isthe direct hydroxylation rather than the photolysis one and this explains the low concentration thatwas found for resorcinol compered to the other byproducts.Another set of experiments was performed monitoring the formation of phenol byproducts inthe presence of difference concentration of chloride. Figure 8.4 shows the results obtained. As itcan be noted no chlorinated byproducts were formed.The formation of chlorophenol is likely to be hydroxyl mediated which is initiated by an HOattack [125] on an aromatic ring to yield to the dihydroxycyclohexadienyl radicals (HO-(H)-Ph•-1120.60.40.20.0Concentration (mM)Catechol p-BQ HQ 2-CP 3-CP 4-CP 0.00 mM 0.25 mM 0.50 mM 0.75 mM 1.00 mM 1.50 mMFigure 8.4: Distribution of chlorinated by-products formed during UV/H2O2 at various chlo-ride concentrations, from 0 to 1.5 mM. Error bars represent the standard deviations ofthree replicates samples.OH). In the presence of oxygen the radicals can evolve in dihydroxybenzene:HO-(H)-Ph•-OH+O2 → HO-Ph-OH+HO2• (8.16)But an alternative could be the reaction with Cl2•−.HO-(H)-Ph•-OH+Cl2•− → HO-Ph-OH+H2O+Cl− (8.17)With this pathway the ortho/para chlorinated byproducts are formed since HO attacks prefer-able in ortho and para positions. The results obtained in this research showed that no chlorinatedbyproducts are formed during UV/H2O2: this could be explain by the fact that in this system H2O2is at higher concentration compared to the one of phenol. In the presence of chloride the followingreactions take place:113H2O2+hv→ 2HO• (8.18)H2O2+HO• → HO2•+H2O (k = 2.7×107M−1s−1) (8.19)H2O2+Cl2•− → HO2•+H++2Cl− (k = 1.4×105M−1s−1) (8.20)HO•+Phenol→ Byproducts (8.21)As it is shown the presence of H2O2 and the reaction induced by its photolysis can results in thedepletion of the reactive species for the phenol photochlorination such as Cl2•−, phenoxylradicaland phenol itself.8.2.3 Degradation of phenol with VUVVUV irradiations were conducted and Figure 8.5 shows the degradation rate for the VUV oxida-tion of phenol with various concentrations of chloride. The results followed a trend by which thedegradation rate of phenol decreases by 37% at 1mM. As previously discussed, chloride has a highabsorption coefficient at 185 nm (Table 8.1), which is the reason for the decrease in the degradationrate. On the other hand, chloride when absorbing a photon can create chloride radicals, a reactivespecies that can help the hydroxyl radicals degradation rate of phenol. This would explain why thedegradation rate did not decrease dramatically.First, the byproducts formed during the degradation of phenol during the VUV process weremonitored without the addition of any chloride. This first set of experiments showed the distributionof byproducts obtained during the oxidation of phenol with HO radicals (figure 8.6).The primary degradation byproducts were catechol, hydroquinone and p-benzoquinone andtrace quantities of resorcinol. Phenol was degraded form an initial concentration of 0.8 ppm to0.0035 ppm in 20 minutes. Catechol increased to a maximum concentration of 0.2 ppm after 4 min-utes. Similarly, p-benzoquinone increased to a maximum in concentration of 0.074 ppm after 4 min-utes before decreasing until the end of the irradiation. The peak concentration of hydroquinone wasabout half of catechol, whereas resorcinol was in trace amounts. The formation of byproducts with-114Figure 8.5: First order rate constant for the degradation of phenol using VUV with variousconcentrations of chloride ion. The fuence rate was 0.06mW/cm2. Error bars representthe standard deviations of three replicates samples.out chloride during the VUV process was similar to the one discussed for UV/H2O2 which is dueto HO attack. Also, in this case the major byproducts formed were catechol and p-benzoquinone,and this another proof that HO attack is more favorable in meta and para position.As it can be noted the trend of byproducts formation in VUV process was different from the onein UV/H2O2. In the case of VUV the highest concentration of byproducts were after just few min-utes, and after that the byproducts formed went to zero concentration. This means that the degrada-tion is faster than in UV/H2O2 and that the byproducts are degraded quickly and transformed to thefinal short chain organic acids such as maleic, malic and tartaric acids. VUV degradation was fasterthan that in UV/H2O2 since the concentration of HO radicals is higher and also 185 nm radiationcould contribute to the photolysis of phenol more than 254 nm radiation: the absorption coefficientfor 254 nm is 516 mol−1 cm−1 and for 185 nm is 320000 M−1 cm−1. In the presence of chlorine115Figure 8.6: Phenol degradation (0.1 ppm) and byproducts formation during the VUV process.Error bars represent the standard deviations of three replicates samples.with VUV, 4-CP is formed although at very low concentration. In this case the scavenging effect ofH2O2 (8.19) is not very important since its formation during VUV is very low (in the range of 0.1ppm), and for this reason the specie Cl2•− is not scavenged 8.19. In addition, it is speculated thatunder 185 nm the following reaction can occur:Cl−+hv185nm → Cl•+ e−aq (8.22)In conclusion, Cl2•− is not scavenged and some Cl• is formed due to the photolysis of Cl− at185 nm and both of them can react with phenol. 4-CP is the only byproducts formed since the attackof chloride radical is favorable in para position. This would explain the fact that the degradationrate of phenol did not decrease dramatically in the presence of chloride. In general the chlorinated116byproducts found were minimum, since the overall favorable mechanism was found to be the onedriven by HO radicals attack.Figure 8.7 shows the chlorinated by-products formed during the VUV process. As it is shown,for high concentration of chloride, some chlorinated by-products are formed, confirming the hy-pothesis that some chloride radicals are formed.0.80.60.40.20.0Concentration (mM)Catechol p-BQ HQ 2-CP 3-CP 4-CP 0.00 mM 0.25 mM 0.50 mM 0.75 mM 1.00 mMFigure 8.7: Distribution of chlorinated by-products formed during the VUV process at variouschloride concentrations, from 0 to 1.5 mM.8.2.4 Effect of pH on the equilibrium of chlorideBecause the reaction with chloride forms hypochloride radicals, it is important to assess the roleof pH on the equilibrium of chloride on the VUV process. The following interactions betweenhydroxyl radicals and chloride ions have been reported [76]:117HO•+Cl− → HOCl•− (k5 = 4.3×109M−1s−1) (8.23)HOCl•− → HO•+Cl−1 (k6 = 6.1×109M−1s−1) (8.24)HOCl•−+H+ → Cl•+H2O (k7 = 2.1×1010M−1s−1) (8.25)Cl•+H2O→ HOCl•−+H+ (k7 = 1.3×103M−1s−1) (8.26)As this manifold of equations shows, pH makes an important contribution to the production ofchloride radicals. At low pH, Eq. 7 is favorable leading to a higher production of chloride radicals.It is clear that hydroxyl radicals and chloride ion are in equilibrium with the hypochlorous radicals.In addition the hypochlorous radicals with a hydrogen ion are in equilibrium with chloride radicalsand water. In this sequence of events, if the dissociation of water is considered, the net equilibriumequation becomes:HO•+Cl− → Cl•+HO−Keq = 1.13×10−7M−1s−1 (8.27)It is possible to calculate the % of chloride radicals in the system for various values of pH using:[ [Cl•][HO•] ] =Keq[Cl−][HO−] = R (8.28)where [Cl•] and [HO•] are the concentrations of Cl and HO radicals, respectively, Keq is the equilib-rium constant, [Cl−] the concentration of chloride ions and [HO−] is the concentration of hydroxylions. The percentage of chloride radicals was calculated:%Cl• = [R]1+[R] (8.29)Given the pH of the system it was possible to calculate the concentration of HO− and thereforeit was possible to calculate the percentage of Cl radicals at various levels of pH. The pH was mon-itored during the irradiations and it was found 5.2, slightly acidic. This was expected because the118samples were prepared with millipore water (slightly acid) and with phenol, which is a weak acid.For a pH of 5.2 and with a concentration of chloride equal to 1.5 mM, 13.2% chloride radicals and86.8% of hydroxyl radicals were formed. As the pH increased to 5.6 during the irradiations, theconcentration of chloride radicals decreased to 4.1%. For this reason some chlorinated by-productswere observed with high concentration of chloride, but the concentration was low because the con-centration of hydroxyl radicals was 8 times higher and for this reason the dominant mechanism ofdegradation was the one driven by the attack of HO•.8.2.5 Effect of real water matrix on the degradation of atrazine in real watersamplesThe effects of each ions on the distribution of photons and on the scavenging of HO radicals wasstudied. The outcomes was used in order to assess the effect of real water matrices on the degra-dation of atrazine with VUV. The elemental composition of two natural waters, Bowen Island andPeachland, was assessed.In the following tables 8.2, 8.3 the elemental composition of two different water matrices(Bowen Island and Peachland) is showed. In the samples, 0.1 ppm of atrazine and 10 ppm ofH2O2 were added. For each components the percentage of 254, 185 nm photons absorbed and thepercentage of HO scavenged was calculated.In the case of Bowen Island water, more than 50% of the 185 nm photons are available forthe photolysis of water and consequently for the production of HO radicals. Consequently, in thiswater matrix 17% of 185 nm photons are scavenged by NOM, 4.6% by chloride and 12% by K+. Inthe case of 254 nm photons 0.96% of the photons that penetrate the water sample are absorbed byH2O2: Ca+2 and NOM in the system have an impact on the distribution of 254 nm photons absorbing0.7% and 43% of the photons, respectively. At last 0.5% of HO radicals are available and react withatrazine leading to its degradation. Only 51% of HO radicals are scavenged by NOM and 47% bychloride. In the case of chloride the percentage of HO radicals scavenged was calculate taking intoaccount the kinetic rate constant between chloride and HO. This value is an high overestimationsince the role of chloride was explained in detail and it was pointed out that once chloride reactswith HO radicals, it can react back to form chloride again and HO radicals. In reality , thus, the119Table 8.2: Effect of Bowen Island water matrix on the degradation of atrazine.Compounds ppm% of 254 % of 185 % of HOnm photons nm photons radicals scavenedNO3− 0.2 0 1 0Cl– 1.1 0 4.6 47HCO3− 0.5 0.28 0.12 0.02SO42− 0.5 2.7 0.05 0K+ 1.8 0 12.32 0Ca2+ 4.3 0.7 0.61 0Na+ 5.7 0 0 0Atrazine 0.1 0.21 0.7 0.5NOM 6 43 17 51H2O2 10 0.96 4.13 0H2O 0 59 0effects of chloride on the concentration of HO radicals is negligible. These results show that in thecase of Bowen Island water matrix, NOM plays a major role and VUV can be used for the removalof atrazine with a pretreatments such as ion exchange in order to make the process more efficient.Another water matrix was studied in order to assess the different affect of different water matrixcomposition.Table (8.3) shows the elemental composition of Peachland water. In the model, 0.1 ppm ofatrazine and 10 ppm of H2O2 were added. For each components the percentage of 254, 185 nmphotons absorbed and the percentage of HO scavenged was calculated.As it is shown peachland 63% of 185 nm photons are available for the photolysis of water, andthe other 37% are absorbed by bicarbonate, potassium and NOM. About 1.06% 254 nm photons areabsorbed by H2O2 and 5.29 are absorbed by calcium which is in high concentration in Peachlandwater (24 ppm). At last, in this case, only 1.8% of HO radicals are available to react with atrazine:80.8% are scavenged by NOM, 2.2% by bicarbonate and 15% by chloride. In this case VUV canbe used for the removal of micropollutants: in order to achieve an higher efficiency, a pretreatmentcan be added in order to remove NOM which id the compound that shows the higher impact. Thevalues obtained during all those experiments were applied in the case of real raw water samples120Table 8.3: Effect of peachland water matrix on the degradation of atrazine.Compounds ppm% of 254 % of 185 % of HOnm photons nm photons radicals scavenedNO3− 0.1 0 0.56 0Cl– 0.1 0 0.37 15HCO3− 13 0 3.6 2.20SO42− 3.84 0 0.4 0K+ 1.7 0 11 0Ca2+ 24.42 5.29 4.6 0Na+ 9.64 0 0 0Atrazine 0.1 0.27 0.84 1.82NOM 3 24 9.40 80.8H2O2 10 1.06 4.42 0H2O 0 63.31 0(Bowen Island and Peachland). The results showed that the major impact that water matrix has onboth UV/H2O2 and VUV is the capability to scavenge HO radicals rather than act as inner filter.For this reason, VUV have the same challenges as other AOPs may have in the presence of real rawsurface waters: for this reason a pretreatment can be added in the system in order to remove theions that shows the highest capability to scavenge HO radicals, such as bicarbonate and NOM.8.3 ConclusionsThe effects of chloride ions on the degradation of phenol during VUV and UV/H2O2 are com-plex. The absorption coefficient of chloride was measured at 254 and 185 nm and were 0 and2791 M−1cm−1, respectively. During the VUV process, the degradation rate of phenol decreasedby 37% in the presence of different concentrations of chloride. Chloride ion acts as an inner filterof 185 nm photons during VUV due to its high absorption coefficient at this wavelength. On theother hand, chloride was found to produce Cl radicals when exposed to 185 nm radiation. Thiswas demonstrated by monitoring the chlorinated byproducts of phenol. With 1.5 mM of chloride,after 30 minutes of irradiations, 0.2 ppm of 4-chlorophenol was found in the system. The over-all mechanism though was found to be the one driven by HO radicals attack: in fact at pH 5.2121only 13.2% of the radicals were chloride radicals, the other 86.8% were HO radicals. During theUV/H2O2 process no change in degradation rate was observed, since chloride does not absorb at254 nm and consequently no chloride radicals were formed due to light photolysis. In addition, nochanges were observed since chloride reacting with HO radicals can form HOCl−•, which then canreform HO radicals. During VUV the degradation rate of phenol decreases by 37% at 1mM. Aspreviously noted, chloride has a high absorption coefficient at 185 nm which is the reason for thedecrease in the degradation rate. On the other hand, chloride when absorbing a photon can createchloride radicals, a reactive species that can help the hydroxyl radicals degradation rate of phenol.This would explain why the degradation rate did not decrease dramatically.122Chapter 9Conclusions9.1 Overall conclusionsThis research studied the effects of water matrix on UV based processes. In addition the synergisticpossibility, of incorporating heterogeneous TiO2 photocatalyst in VUV reactors was investigated.The overall conclusions of this research are the following:• The absorption coefficients at 185 and 254 nm play a central role in the photon distribution.Atrazine has a high absorption coefficient for both 185 nm and 254 nm radiation. All the otherinorganic ions, showed an absorption coefficient equal to zero for 254 nm except nitrate. Thisshows evidence that during UV/H2O2 process only nitrate plays a role on the distribution ofphotons in the system, thus it is the only ion that can act as inner filter. On the other hand, at185 nm radiation of all the ions showed to absorb 185 nm radiation. Among them chloridewas the one that showed the highest capability to absorb photons at 185 nm. The scavengingeffects of different ions was then tested first during the UV/H2O2 process and after that withVUV. Nitrate showed a detrimental effect both with UV/H2O2 and with VUV. The reason isthe high absorption coefficient that nitrate has at 254 and 185 nm. In both case Nitrate acts asinner filter scavenging the photons. Bicarbonate also showed to have detrimental effects dueto its capability to scavenge HO radicals .On the other hand, the degradation of atrazine in the presence of sulfate slightly increases.123Sulfate is able to photolyze at 185 nm and to form HO radicals.At last the values obtained during all those experiments were applied in the case of real rawwater samples (Bowen Island and Peachland). The results showed that the major impact thatinorganic ions have on both UV/H2O2 and VUV is the capability to scavenge HO radicalsrather than act as inner filter. For this reason, VUV have the same challenges that other AOPsmay have in the presence of real raw surface waters: for this reason a pretreatment, such asion exchange resins and flocculations, can be added in the system in order to remove the ionsthat shows the highest capability to scavenge HO radicals, such as bicarbonate.• The presence of NOM negatively affects the degradation rate of atrazine by consuming HOradicals and partly absorbing VUV photons. HO radicals predominantly react with NOM,and thus less HO radicals generated will remain to react with atrazine. Because of the pro-nounced scavenging effect of NOM, the steady state concentration of HO radicals reducessignificantly. About the interaction with VUV photons, the results showed that the absorp-tion coefficient depends on the type and thus on the structure of NOM. The percentage of HOscavenged by Nordic Reservoir NOM and by Suwannee river NOM was slightly differentdue to their different structures. Nordic reservoir was found to have an higher absorption co-efficient at 254 nm but lower at 185 nm than the ones of Suwannee river. In addition NordicNOM showed an higher reactivity towards HO radicals.The overall result is that the degradation rate of atrazine reduces significantly in the presenceof NOM and this reduction is primarily because of the scavenging effects of NOM rather thanits ability of acting as inner filter for VUV radiation. In addition the results from this worksuggest that NOM can be a sensitizer, but this effect is lower and not comparable to its abilityto react with HO radicals. With VUV NOM mostly photolyze forming by-products ratherthan forming triplet state DOM species and the amount of reactive oxidants that is formed byits interaction with radiation is lower to the one that comes from the photolysis of water. Theresults of the studies can provide information about the possibility of simple pre treatmentsbefore the VUV process according to the water source.124• The effects of chloride ions on the degradation of phenol during VUV and UV/H2O2 arecomplex. The absorption coefficient of chloride was measured at 254 and 185 nm and it wasfound to be 0 and 2791 M−1cm−1, respectively. During VUV, the degradation rate of phenoldecreased by 37% in the presence of different concentrations of chloride. Chloride ions actas inner filter of 185 nm photons during the VUV treatment due to its high absorption coeffi-cient at this wavelength. On the other hand chloride was found to produce Cl radicals whenexposed to 185 nm radiation at high concentration (1.5 mM). This was proven by monitoringthe chlorinated byproducts of phenol. With 1.5 mM of chloride 0.2 ppm of 4-chlorophenolwas found in the system. The overall mechanism though was found to be the one driven byHO radicals attack: in fact, at pH 5.2 only 13.2% of the radicals were chloride radicals, theother 86.8% were HO radicals. During UV/H2O2 treatment, no change in the degradationrate was observed, since chloride does not absorb at 254 nm and consequently no chlorideradicals were formed due to photolysis. In addition, no changes were observed since chloridewhen reacts with HO radicals can form HOCl−•, which can reform HO radicals. Overall, theconcentration of chloride found in surface natural water will not show an important impacton VUV process.• Among the five different catalysts, the composite catalysts showed much higher photocat-alytic activities, about 70% 2,4-D degradation, compared to 10% 2,4-D degradation resultedfrom the catalysts without Degussa P25. The two composite TiO2 catalysts were also evalu-ated in terms of attrition resistance and photocatalytic stability. The composite coating withmore homogeneous morphology and less cracking presented lower attrition and no deacti-vation. The conversions obtained for the VUV, UV/immobilized TiO2, and UV/slurry TiO2were 100%, 2.5% and 38% with millipure water and 63% and 7% with surface water, re-spectively. The results show that photocatalysis cannot improve significantly the efficacy ofthe VUV in a flow through reactor for the degradation of micropollutants. When the sampleswere prepared with surface water the efficiency decreased. This decrement was due to com-pounds (inorganic ions and NOM) dissolved in surface water which can interfere with theprocess. In particular, NOM and inorganic ions can absorb 254 and 185 nm photons, and/or125scavenge HO radicals.9.2 Significance of the researchAOPs are emerging as viable solutions for the degradation of micropollutants in water. Among dif-ferent AOPs, Vacuum UV is one of the most promising because it does not require any chemicalsand the process efficiency is high. VUV is a technique not thoroughly studied so its potentials arenot completely exploited. In particular, photochemical mechanisms involved during the process arenot completely understood. Given that, the constituents of the water matrix play important roleon the degradation rate of target contaminants and the overall process efficiency, it is essential toasses the role played by the water matrix e.g. inorganic ions and natural organic matter on the per-formance of VUV process. The efficiency of VUV for the degradation of a target micropollutantdecreases dramatically from millipore H2O to raw surface water. Knowing the photochemical pro-cesses taking place in real water (with all its constituents) allows to calculate the kinetics constantsbetween t the constituents and the micropollutant and allows to predict the photochemical behaviorof micropollutants in different water samples. Apply this knowledge, in turn, helps with the properdesign of the process with conditions to achieve the most efficient treatment.In addition, in this project the photochemical processes involved during the VUV treatment wasstudied under just one wavelength (185 nm). To the best of our knowledge, there has not beenany study isolating the 185 nm and 254 nm wavelengths and studying the fundamental processesassociated with each wavelength independently. To achieve this objective a novel collimated beamwas designed and built and it can be used as a tool for future researches that involve 185 nmradiation. With this setup, furthermore, it was possible to developed a method in order to calculatethe quantum yield of any organic at 185 nm, which is a fundamental parameter for any kinetic studyat 185 nm. To the best of my knowledge there is no such simple yet effective methodology reportedin the literature.Finally, VUV can be a very powerful technology as tertiary treatments for the removal of mi-cropollutant since with VUV treatment both degradation of organic compounds and disinfectioncould be achieved. This work would contribute to the development of an additional tool in the arse-126nal of drinking water treatment technology available for dealing with micropollutants in small andrural contexts, while maintaining disinfection capabilities.9.3 Recommendations for future research• In this research the molar absorption coefficient of water was measured. The methodologydeveloped was accurate at 25 ◦C: however, it is strongly recommended to measure the molarabsorption coefficient at different temperatures. The temperature dependence reported sug-gests a 20% difference in absorbance between 20 ◦C and 30 ◦C. VUV treatment can be usedin environment with higher or lower temperature of 25 ◦C in some case, water temperaturesduring treatment may descend to just above 0 ◦C. For this reason it is important to quantifythe absorbance of water for a wide range of temperature.• The effect of NOM was studied in details; however, it is important to understand what func-tionality in humics is most sensitive to the oxidation and following the change that occurswithin the NOM structure during VUV process.• In addition for VUV treatment the lamps emit 90% of photons at 254 nm and 10% at 185 nm.It is valuable to explore if it is possible to achieve disinfection and removal of micropollutantsat the same time during VUV.• Finally this research studied in details the efficacy of VUV and the impact that water matrixmay have on the process in a collimated beam. The next step would be to apply the toolsobtained in this research in a small scale VUV pilot plan. In this way it would be possible toevaluate the efficacy of VUV for real cases.127Bibliography[1] Herve´ Gallard and Urs von Gunten. Chlorination of phenols: Kinetics and formation ofchloroform. Environmental Science & Technology, 36(5):884–890, 2002. → pages xiv,109, 110[2] Rene P Schwarzenbach, Beate I Escher, Kathrin Fenner, Thomas B Hofstetter, C AnnetteJohnson, Urs Von Gunten, and Bernhard Wehrli. The challenge of micropollutants inaquatic systems. Science, 313(5790):1072–1077, 2006. → pages 1[3] RW Macdonald, LA Barrie, TF Bidleman, ML Diamond, DJ Gregor, RG Semkin, WMJStrachan, YF Li, F Wania, M Alaee, et al. Contaminants in the canadian arctic: 5 years ofprogress in understanding sources, occurrence and pathways. Science of the TotalEnvironment, 254(2):93–234, 2000. → pages 1[4] Dana W Kolpin, Edward T Furlong, Michael T Meyer, E Michael Thurman, Steven DZaugg, Larry B Barber, and Herbert T Buxton. Pharmaceuticals, hormones, and otherorganic wastewater contaminants in us streams, 1999-2000: A national reconnaissance.Environmental science & technology, 36(6):1202–1211, 2002. → pages 2[5] Hisato Iwata, Shinsuke Tanabe, Norio Sakai, Atsushi Nishimura, and Ryo Tatsukawa.Geographical distribution of persistent organochlorines in air, water and sediments fromasia and oceania, and their implications for global redistribution from lower latitudes.Environmental Pollution, 85(1):15–33, 1994. → pages 2[6] Kristina M Zierold, Lynda Knobeloch, and Henry Anderson. Prevalence of chronic diseasesin adults exposed to arsenic-contaminated drinking water. American journal of publichealth, 94(11):1936–1937, 2004. → pages 2[7] Jane Ellen Simmons, Susan D Richardson, Thomas F Speth, Richard J Miltner, Glenn Rice,Kathleen M Schenck, E Sidney Hunter, et al. Development of a research strategy forintegrated technology-based toxicological and chemical evaluation of complex mixtures ofdrinking water disinfection byproducts. Environmental Health Perspectives, 110(Suppl6):1013, 2002. → pages 2[8] Martin Mengis, Rene Gachter, Bernhard Wehrli, and Stefano Bernasconi. Nitrogenelimination in two deep eutrophic lakes. Limnology and Oceanography, pages 1530–1543,1997. → pages 2128[9] Mark A Shannon, Paul W Bohn, Menachem Elimelech, John G Georgiadis, Benito JMarinas, and Anne M Mayes. Science and technology for water purification in the comingdecades. Nature, 452(7185):301–310, 2008. → pages 2[10] F Martin, Francisco Javier Gonza´lez-Vila, JC Del Rio, and T Verdejo. Pyrolysisderivatization of humic substances 1. pyrolysis of fulvic acids in the presence oftetramethylammonium hydroxide. Journal of Analytical and Applied Pyrolysis,28(1):71–80, 1994. → pages 3[11] Anastasia D Nikolaou and Themistokles D Lekkas. The role of natural organic matterduring formation of chlorination by-products: A review. Acta hydrochimica ethydrobiologica, 29(2-3):63–77, 2001. → pages 3[12] W. Buchanan, F. Roddick, and N. Porter. Formation of hazardous by-products resultingfrom the irradiation of natural organic matter: Comparison between uv and vuv irradiation.Chemosphere, 63(7):1130–1141, 2006. → pages 3, 17, 18, 26[13] JK Edzwald. Coagulation in drinking water treatment: Particles, organics and coagulants.Water Science and Technology, 27(11):21–35, 1993. → pages 3[14] Sudha Goel, Raymond M Hozalski, and Edward J Bouwer. Biodegradation of nom: effectof nom source and ozone dose: Natural organic matter. Journal-American Water WorksAssociation, 87(1):90–105, 1995. → pages 4[15] HA Elliott, BA Dempsey, and PJ Maille. Content and fractionation of heavy metals in watertreatment sludges. Journal of environmental quality, 19(2):330–334, 1990. → pages 4[16] M Gomez, F Plaza, G Garralon, J Perez, and M A Gomez. A comparative study of tertiarywastewater treatment by physico-chemical-uv process and macrofiltration–ultrafiltrationtechnologies. Desalination, 202(1):369–376, 2007. → pages 4[17] JAMH Hofman, EF Beerendonk, HC Folmer, and JC Kruithof. Removal of pesticides andother micropollutants with cellulose-acetate, polyamide and ultra-low pressure reverseosmosis membranes. Desalination, 113(2):209–214, 1997. → pages 4[18] JH Baxendale and JA Wilson. The photolysis of hydrogen peroxide at high light intensities.Transactions of the Faraday Society, 53:344–356, 1957. → pages 7[19] JL Weeks, G. Meaburn, and S. Gordon. Absorption coefficients of liquid water and aqueoussolutions in the far ultraviolet. Radiation Research, pages 559–567, 1963. → pages 8, 15,20, 63[20] Y Lester, D Avisar, and H Mamane. Photodegradation of the antibiotic sulphamethoxazolein water with uv/h2o2 advanced oxidation process. Environmental technology,31(2):175–183, 2010. → pages 8[21] MC Gonzalez, AM Braun, A Bianco Prevot, and E Pelizzetti. Vacuum-ultraviolet (vuv)photolysis of water: mineralization of atrazine. Chemosphere, 28(12):2121–2127, 1994. →pages 8129[22] James R Bolton. Calculation of ultraviolet fluence rate distributions in an annular reactor:significance of refraction and reflection. Water Research, 34(13):3315–3324, 2000. →pages 8[23] Guus F IJpelaar, Danny JH Harmsen, Erwin F Beerendonk, Robin C van Leerdam,Debbie H Metz, Antonie H Knol, Alice Fulmer, and Simon Krijnen. Comparison of lowpressure and medium pressure uv lamps for uv/h2o2 treatment of natural waters containingmicro pollutants. Ozone: Science & Engineering, 32(5):329–337, 2010. → pages 9[24] Karl G Linden, Gwy-Am Shin, Gaetan Faubert, William Cairns, and Mark D Sobsey. Uvdisinfection of giardia lamblia cysts in water. Environmental science & technology,36(11):2519–2522, 2002. → pages 9[25] Joop C Kruithof, Peer C Kamp, and Bram J Martijn. Uv/h2o2 treatment: A practicalsolution for organic contaminant control and primary disinfection. Ozone: Science andEngineering, 29(4):273–280, 2007. → pages 9[26] Santiago Esplugas, Jaime Gimenez, Sandra Contreras, Esther Pascual, and MiguelRodriguez. Comparison of different advanced oxidation processes for phenol degradation.Water Research, 36(4):1034–1042, 2002. → pages 9[27] Stephen R Cater, Mihaela I Stefan, James R Bolton, and Ali Safarzadeh-Amiri. Uv/h2o2treatment of methyl tert-butyl ether in contaminated waters. Environmental science &technology, 34(4):659–662, 2000. → pages 9[28] Erik J Rosenfeldt, Barbara Melcher, and Karl G Linden. Uv and uv/h 2 o 2 treatment ofmethylisoborneol (mib) and geosmin in water. Journal of Water Supply: Research &Technology-AQUA, 54(7), 2005. → pages 9[29] Wenbing Zhang, Xianming Xiao, Taicheng An, Zhiguang Song, Jiamo Fu, Guoying Sheng,and Mingchao Cui. Kinetics, degradation pathway and reaction mechanism of advancedoxidation of 4-nitrophenol in water by a uv/h2o2 process. Journal of Chemical Technologyand Biotechnology, 78(7):788–794, 2003. → pages 9[30] Davide Vogna, Raffaele Marotta, Roberto Andreozzi, Alessandra Napolitano, and MarcodIschia. Kinetic and chemical assessment of the uv/h2o2 treatment of antiepileptic drugcarbamazepine. Chemosphere, 54(4):497–505, 2004. → pages 9, 10[31] Melody Blythe Johnson and Mehrab Mehrvar. Aqueous metronidazole degradation byuv/h2o2 process in single-and multi-lamp tubular photoreactors: kinetics and reactor design.Industrial & Engineering Chemistry Research, 47(17):6525–6537, 2008. → pages 9, 10[32] Cong Li, Nai-yun Gao, and Wang Li. Photochemical degradation of typical herbicidessimazine by uv/h2o2 in aqueous solution. Desalination and Water Treatment,36(1-3):197–202, 2011. → pages 9, 10, 11[33] Karin Lekkerkerker-Teunissen, Mark J Benotti, Shane A Snyder, and Hans C van Dijk.Transformation of atrazine, carbamazepine, diclofenac and sulfamethoxazole by low andmedium pressure uv and uv/h2o2 treatment. Separation and Purification Technology,96:33–43, 2012. → pages 9, 10130[34] G Kleiser and FH Frimmel. Removal of precursors for disinfection by-products(dbps)differences between ozone-and oh-radical-induced oxidation. Science of the totalenvironment, 256(1):1–9, 2000. → pages 11[35] GS Wang, C-H Liao, H-W Chen, and HC Yang. Characteristics of natural organic matterdegradation in water by uv/h2o2 treatment. Environmental technology, 27(3):277–287,2006. → pages 11[36] Hugo de Lasa, Benito Serrano, and Miguel Salaices. Kinetic modeling of the photocatalyticreaction network: The parallel-series approximation. In Photocatalytic ReactionEngineering, pages 101–118. Springer, 2005. → pages 12[37] O. Carp, CL Huisman, and A. Reller. Photoinduced reactivity of titanium dioxide. Progressin Solid State Chemistry, 32(1-2):33–177, 2004. → pages 12, 52[38] A. Yamakata, T. Ishibashi, and H. Onishi. Pressure dependence of electron-andhole-consuming reactions in photocatalytic water splitting on Pt/TiO2 studied bytime-resolved IR absorption spectroscopy. Vacuum, 10:3. → pages 12[39] K. Tanaka, M.F.V. Capule, and T. Hisanaga. Effect of crystallinity of TiO2 on itsphotocatalytic action. Chemical Physics Letters, 187(1-2):73–76, 1991. → pages 12[40] M.A. Rahman and M. Muneer. Photocatalysed degradation of two selected pesticidederivatives, dichlorvos and phosphamidon, in aqueous suspensions of titanium dioxide.Desalination, 181(1-3):161–172, 2005. → pages 13[41] S. Gelover, P. Mondragon, and A. Jimenez. Titanium dioxide sol-gel deposited over glassand its application as a photocatalyst for water decontamination. Journal of Photochemistryand Photobiology A: Chemistry, 165(1-3):241–246, 2004. → pages 13[42] E. Pelizzetti, V. Maurino, C. Minero, V. Carlin, M.L. Tosato, E. Pramauro, and O. Zerbinati.Photocatalytic degradation of atrazine and other s-triazine herbicides. EnvironmentalScience & Technology, 24(10):1559–1565, 1990. → pages 13[43] C. Guillard, P. Pichat, G. Huber, and C. Hoang-Van. The GC-MS Analysis of OrganicIntermediates from the TiO2 Photocatalytic Treatment of Water Contaminated by Lindane.J. Adv. Oxid. Technol, 1(1):53–60, 1996. → pages 13[44] E. Evgenidou, I. Konstantinou, K. Fytianos, I. Poulios, and T. Albanis. Photocatalyticoxidation of methyl parathion over TiO2 and ZnO suspensions. Catalysis today,124(3-4):156–162, 2007. → pages 13[45] Linda A Lawton, Peter KJ Robertson, Russell F Robertson, and Fiona G Bruce. Thedestruction of 2-methylisoborneol and geosmin using titanium dioxide photocatalysis.Applied Catalysis B: Environmental, 44(1):9–13, 2003. → pages 13[46] Yiming Xu and Cooper H Langford. Variation of langmuir adsorption constant determinedfor tio2-photocatalyzed degradation of acetophenone under different light intensity. Journalof Photochemistry and Photobiology A: Chemistry, 133(1):67–71, 2000. → pages 13131[47] Andrew Mills, Jishun Wang, and David F Ollis. Dependence of the kinetics of liquid-phasephotocatalyzed reactions on oxygen concentration and light intensity. Journal of Catalysis,243(1):1–6, 2006. → pages 13[48] Omatoyo K Dalrymple, Daniel H Yeh, and Maya A Trotz. Removing pharmaceuticals andendocrine-disrupting compounds from wastewater by photocatalysis. Journal of ChemicalTechnology and Biotechnology, 82(2):121–134, 2007. → pages 13[49] Sean Carbonaro, Matthew N Sugihara, and Timothy J Strathmann. Continuous-flowphotocatalytic treatment of pharmaceutical micropollutants: Activity, inhibition, anddeactivation of tio2 photocatalysts in wastewater effluent. Applied Catalysis B:Environmental, 129:1–12, 2013. → pages 14[50] Romina zabar, Darko Dolenc, Tina Jerman, Mladen Franko, and Polonca Trebse. Photolyticand photocatalytic degradation of 6-chloronicotinic acid. Chemosphere, 85(5):861–868,2011. → pages 14[51] G.D. Onstad, S. Strauch, J. Meriluoto, G.A. Codd, and U. Von Gunten. Selective oxidationof key functional groups in cyanotoxins during drinking water ozonation. Environ. Sci.Technol, 41(12):4397–4404, 2007. → pages 14[52] J.A. Westrick, D.C. Szlag, B.J. Southwell, and J. Sinclair. A review of cyanobacteria andcyanotoxins removal/inactivation in drinking water treatment. Analytical and BioanalyticalChemistry, pages 1–10. → pages 15[53] T. Oppenlander and S. Gliese. Mineralization of organic micropollutants (homologousalcohols and phenols) in water by vacuum-UV-oxidation (H2O-VUV) with an incoherentxenon-excimer lamp at 172 nm. Chemosphere, 40(1):15–21, 2000. → pages 15[54] T. Oppenlander. Mercury-free sources of VUV/UV radiation: application of modernexcimer lamps (excilamps) for water and air treatment. Journal of EnvironmentalEngineering and Science, 6(3):253–264, 2007. → pages 15[55] E.A. Sosnin, T. Oppenlander, and V.F. Tarasenko. Applications of capacitive and barrierdischarge excilamps in photoscience. Journal of Photochemistry and Photobiology C:Photochemistry Reviews, 7(4):145–163, 2006. → pages 15[56] M.G. Gonzalez, E. Oliveros, M. Worner, and A.M. Braun. Vacuum-ultraviolet photolysis ofaqueous reaction systems. Journal of Photochemistry and Photobiology C: PhotochemistryReviews, 5(3):225–246, 2004. → pages 15, 16[57] Thomas Oppenlander, Carsten Walddorfer, Jens Burgbacher, Martin Kiermeier, KlausLachner, and Helga Weinschrott. Improved vacuum-uv (vuv)-initiated photomineralizationof organic compounds in water with a xenon excimer flow-through photoreactor (lamp,172nm) containing an axially centered ceramic oxygenator. Chemosphere, 60(3):302–309,2005. → pages 16[58] G. Baum and T. Oppenlander. Vacuum-uv-oxidation of chloroorganic compounds in anexcimer flow through photoreactor. Chemosphere, 30(9):1781–1790, 1995. → pages 17132[59] N. Quici, M.I. Litter, A.M. Braun, and E. Oliveros. Vacuum-uv-photolysis of aqueoussolutions of citric and gallic acids. Journal of Photochemistry and Photobiology A:Chemistry, 197(2-3):306–312, 2008. → pages 17, 26[60] Gustavo Imoberdorf and Madjid Mohseni. Degradation of natural organic matter in surfacewater using vacuum-uv irradiation. Journal of hazardous materials, 186(1):240–246, 2011.→ pages 17[61] Thunyalux Ratpukdi, Sumana Siripattanakul, and Eakalak Khan. Mineralization andbiodegradability enhancement of natural organic matter by ozone–vuv in comparison withozone, vuv, ozone–uv, and uv: Effects of ph and ozone dose. Water research,44(11):3531–3543, 2010. → pages 18[62] Peter H Gleick. Water and energy. Annual Review of Energy and the Environment,19(1):267–299, 1994. → pages 18[63] Ding Wang, Thomas Oppenla¨nder, Mohamed Gamal El-Din, and James R Bolton.Comparison of the disinfection effects of vacuum-uv (vuv) and uv light on bacillus subtilisspores in aqueous suspensions at 172, 222 and 254 nm. Photochemistry and photobiology,86(1):176–181, 2010. → pages 18[64] W. Han, P. Zhang, W. Zhu, J. Yin, and L. Li. Photocatalysis of p-chlorobenzoic acid inaqueous solution under irradiation of 254 nm and 185 nm uv light. Water research,38(19):4197–4203, 2004. → pages 19[65] Wenya Han, Wanpeng Zhu, Pengyi Zhang, Ying Zhang, and Laisheng Li. Photocatalyticdegradation of phenols in aqueous solution under irradiation of 254 and 185nm uv light.Catalysis Today, 90(3):319–324, 2004. → pages 19[66] J. Soler, A. Garcia-Ripoll, N. Hayek, P. Miro, R. Vicente, A. Arques, and AM Amat. Effectof inorganic ions on the solar detoxification of water polluted with pesticides. Waterresearch, 43(18):4441–4450, 2009. → pages 20[67] J. Mack and J.R. Bolton. Photochemistry of nitrite and nitrate in aqueous solution: areview. Journal of Photochemistry and Photobiology A: Chemistry, 128(1-3):1–13, 1999.→ pages 20[68] G.V. Buxton, C.L. Greenstock, W.P. Helman, and A.B. Ross. Critical review of rateconstants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals. Phys.Chem. Ref. Data, 17:513–886, 1988. → pages 20[69] F.S.G. Einschlag, J.I. Felice, and J.M. Triszcz. Kinetics of nitrobenzene and 4-nitrophenoldegradation by UV irradiation in the presence of nitrate and nitrite ions. Photochemical andPhotobiological Sciences, 8(7):953–960, 2009. → pages 21[70] G. Mark, H.G. Korth, H.P. Schuchmann, and C. von Sonntag. The photochemistry ofaqueous nitrate ion revisited. Journal of Photochemistry and Photobiology A: Chemistry,101(2-3):89–103, 1996. → pages 21133[71] M. Fischer and P. Warneck. Photodecomposition of nitrite and undissociated nitrous acid inaqueous solution. J. Phys. Chem, 100(48):18749–18756, 1996. → pages 21[72] D. Vione, V. Maurino, C. Minero, and E. Pelizzetti. Phenol photonitration upon UVirradiation of nitrite in aqueous solution I: Effects of oxygen and 2-propanol. Chemosphere,45(6-7):893–902, 2001. → pages 22[73] J. Huang and S.A. Mabury. A new method for measuring carbonate radical reactivity towardpesticides. Environmental toxicology and chemistry, 19(6):1501–1507, 2000. → pages 22[74] S.N. Chen, M.Z. Hoffman, and G.H. Parsons Jr. Reactivity of the carbonate radical towardaromatic compounds in aqueous solution. The Journal of Physical Chemistry,79(18):1911–1912, 1975. → pages 22[75] D. Vione, V. Maurino, C. Minero, M.E. Carlotti, S. Chiron, and S. Barbati. Modelling theoccurrence and reactivity of the carbonate radical in surface freshwater. Comptes RendusChimie, 12(8):865–871, 2009. → pages 22[76] Chih-Hsiang Liao, Shyh-Fang Kang, and Fu-An Wu. Hydroxyl radical scavenging role ofchloride and bicarbonate ions in the h2o2 uv process. Chemosphere, 44(5):1193–1200,2001. → pages 22, 117[77] Naeem Kashif and Feng Ouyang. Parameters effect on heterogeneous photocatalyseddegradation of phenol in aqueous dispersion of tio2. Journal of Environmental Sciences,21(4):527–533, 2009. → pages 22, 23, 107[78] M Muruganandham and M Swaminathan. Photochemical oxidation of reactive azo dye withuv h2o2 process. Dyes and pigments, 62(3):269–275, 2004. → pages 23[79] A Riga, K Soutsas, K Ntampegliotis, V Karayannis, and G Papapolymerou. Effect ofsystem parameters and of inorganic salts on the decolorization and degradation of procionh-exl dyes. comparison of h2o2/uv, fenton, uv/fenton, tio2/uv and tio2 uv/h2o2 processes.Desalination, 211(1):72–86, 2007. → pages 23[80] Ruixia Yuan, Sadiqua N Ramjaun, Zhaohui Wang, and Jianshe Liu. Effects of chloride ionon degradation of acid orange 7 by sulfate radical-based advanced oxidation process:Implications for formation of chlorinated aromatic compounds. Journal of hazardousmaterials, 196:173–179, 2011. → pages 24[81] C.M. Sharpless, D.A. Seibold, and K.G. Linden. Nitrate photosensitized degradation ofatrazine during UV water treatment. Aquatic Sciences-Research Across Boundaries,65(4):359–366, 2003. → pages 24[82] S.R. Sarathy and M. Mohseni. The impact of UV/H2O2 advanced oxidation on molecularsize distribution of chromophoric natural organic matter. Environmental science andtechnology, 41(24):8315–8320, 2007. → pages 24[83] A.J. Schindelin and F.H. Frimmel. Nitrate and natural organic matter in aqueous solutionsirradiated by simulated sunlight. Environmental Science and Pollution Research,7(4):205–210, 2000. → pages 24134[84] Celine Tixier, Heinz P Singer, Silvio Canonica, and Stephan R Mu¨ller. Phototransformationof triclosan in surface waters: A relevant elimination process for this widely used biocidelaboratory studies, field measurements, and modeling. Environmental science &technology, 36(16):3482–3489, 2002. → pages 24[85] Andreas C Gerecke, Silvio Canonica, Stephan R Mu¨ller, Michael Scha¨rer, and Rene´ PSchwarzenbach. Quantification of dissolved natural organic matter (dom) mediatedphototransformation of phenylurea herbicides in lakes. Environmental science &technology, 35(19):3915–3923, 2001. → pages 25[86] Changlong Wu, Hilla Shemer, and Karl G Linden. Photodegradation of metolachlorapplying uv and uv/h2o2. Journal of agricultural and food chemistry, 55(10):4059–4065,2007. → pages 25[87] Hajime Shirayama, Yoshimitsu Tohezo, and Shigeru Taguchi. Photodegradation ofchlorinated hydrocarbons in the presence and absence of dissolved oxygen in water. Waterresearch, 35(8):1941–1950, 2001. → pages 26[88] GE Imoberdorf and M Mohseni. Experimental study of the degradation of 2, 4-d inducedby vacuum-uv radiation. Water Science & Technology, 63(7), 2011. → pages 26[89] S Echigo, K Kosaka, H Yamada, S Matsui, and K Shishida. The effect of water qualitycharacteristics on the performance of the ozone/vacuum ultraviolet process. InPROCEEDINGS-JAPAN SOCIETY OF CIVIL ENGINEERS, pages 51–60. DOTOKUGAKKAI, 1999. → pages 26[90] Harumi Yamada and Hiroshi Tsuno. Characteristics of advanced oxidation processes for thedecomposition of organic compounds. Journal of the Chinese Institute of Environmentalengineering, 10(1):61–68, 2000. → pages 26[91] J Esteban Duran, Madjid Mohseni, and Fariborz Taghipour. Computational fluid dynamicsmodeling of immobilized photocatalytic reactors for water treatment. AIChE Journal,57(7):1860–1872, 2011. → pages 36, 52[92] Ronald O Rahn. Potassium iodide as a chemical actinometer for 254 nm radiation: use oflodate as an electron scavenger. Photochemistry and Photobiology, 66(4):450–455, 1997.→ pages 39[93] G. Heit, A. Neuner, P.Y. Saugy, and A.M. Braun. Vacuum-UV (172 nm) actinometry. Thequantum yield of the photolysis of water. J. Phys. Chem. A, 102(28):5551–5561, 1998. →pages 40[94] Jerome A Klavons and Raymond D Bennett. Determination of methanol using alcoholoxidase and its application to methyl ester content of pectins. Journal of Agricultural andFood Chemistry, 34(4):597–599, 1986. → pages 40[95] Normal V Klassen, David Marchington, and Heather CE McGowan. H2o2 determinationby the i3-method and by kmno4 titration. Analytical Chemistry, 66(18):2921–2925, 1994.→ pages 40135[96] Chan-Hoe Yip, Yet-Ming Chiang, and Chee-Cheong Wong. Dielectric band edgeenhancement of energy conversion efficiency in photonic crystal dye-sensitized solar cell.The Journal of Physical Chemistry C, 112(23):8735–8740, 2008. → pages 44[97] S Yamazaki-Nishida, KJ Nagano, LA Phillips, S Cervera-March, and MA Anderson.Photocatalytic degradation of trichloroethylene in the gas phase using titanium dioxidepellets. Journal of Photochemistry and Photobiology A: Chemistry, 70(1):95–99, 1993. →pages 44[98] Mehrdad Keshmiri, Tom Troczynski, and Madjid Mohseni. Oxidation of gas phasetrichloroethylene and toluene using composite sol–gel tio2 photocatalytic coatings. Journalof hazardous materials, 128(2):130–137, 2006. → pages 44[99] A. Sclafani and JM Herrmann. Comparison of the photoelectronic and photocatalyticactivities of various anatase and rutile forms of titania in pure liquid organic phases and inaqueous solutions. The Journal of Physical Chemistry, 100(32):13655–13661, 1996. →pages 47[100] C.B. Almquist and P. Biswas. Role of synthesis method and particle size of nanostructuredTiO2 on its photoactivity. Journal of Catalysis, 212(2):145–156, 2002. → pages 47[101] Karin Eufinger, Dirk Poelman, Hilde Poelman, Roger De Gryse, and GB Marin.Photocatalytic activity of dc magnetron sputter deposited amorphous tio¡ sub¿ 2¡/sub¿ thinfilms. Applied surface science, 254(1):148–152, 2007. → pages 51[102] Hai Wang, Yu Wu, and Bo-Qing Xu. Preparation and characterization of nanosized anatasetio2 cuboids for photocatalysis. Applied Catalysis B: Environmental, 59(3):139–146, 2005.→ pages 52[103] A Bojinova and C Dushkin. Photodegradation of malachite green in water solutions bymeans of thin films of tio2/wo3 under visible light. Reaction Kinetics, Mechanisms andCatalysis, 103(1):239–250, 2011. → pages 52[104] R Swanepoel. Determining refractive index and thickness of thin films from wavelengthmeasurements only. JOSA A, 2(8):1339–1343, 1985. → pages 52[105] Dong Li, Kui Xiong, Zhuhong Yang, Chang Liu, Xin Feng, and Xiaohua Lu. Processintensification of heterogeneous photocatalysis with static mixer: Enhanced mass transfer ofreactive species. Catalysis Today, 175(1):322–327, 2011. → pages 56[106] R Stephen Berry and Curt W Reimann. Absorption spectrum of gaseous f-and electronaffinities of the halogen atoms. The Journal of Chemical Physics, 38(7):1540–1543, 1963.→ pages 62[107] J De Laat, P Berger, T Poinot, N Karpel vel Leitner, and M Dore. Modeling the oxidation ofatrazine by h2o2/uv. estimation of kinetic parameters. Ozone: science & engineering,19(5):395–408, 1997. → pages 69[108] Oliver C Zafiriou and Mary B True. Nitrate photolysis in seawater by sunlight. MarineChemistry, 8(1):33–42, 1979. → pages 76136[109] H Gallard and J De Laat. Kinetic modelling of fe (iii)/h2o2 oxidation reactions in diluteaqueous solution using atrazine as a model organic compound. Water Research,34(12):3107–3116, 2000. → pages 77, 94, 96, 98[110] Liang Chu and Cort Anastasio. Quantum yields of hydroxyl radical and nitrogen dioxidefrom the photolysis of nitrate on ice. The Journal of Physical Chemistry A,107(45):9594–9602, 2003. → pages 77[111] JH Baxendale and AL Mansell. Flash photolysis of ferrous sulphate solutions. Nature,190:622–623, 1961. → pages 82[112] SCB Myneni, JT Brown, GA Martinez, and W Meyer-Ilse. Imaging of humic substancemacromolecular structures in water and soils. Science, 286(5443):1335–1337, 1999. →pages 89[113] Gregory V Korshin, Chi-Wang Li, and Mark M Benjamin. Monitoring the properties ofnatural organic matter through uv spectroscopy: a consistent theory. Water Research,31(7):1787–1795, 1997. → pages 90[114] Jerry A Leenheer, Robert L Wershaw, and Michael M Reddy. Strong-acid, carboxyl-groupstructures in fulvic acid from the suwannee river, georgia. 2. major structures.Environmental science & technology, 29(2):399–405, 1995. → pages 91[115] Marc Benedetti, James F Ranville, Marie Ponthieu, and Jose Paulo Pinheiro. Field-flowfractionation characterization and binding properties of particulate and colloidal organicmatter from the rio amazon and rio negro. Organic Geochemistry, 33(3):269–279, 2002. →pages 91[116] P. Westerhoff, G. Aiken, G. Amy, and J. Debroux. Relationships between the structure ofnatural organic matter and its reactivity towards molecular ozone and hydroxyl radicals.Water Research, 33(10):2265–2276, 1999. → pages 91, 94[117] Lionel Ho, Gayle Newcombe, and Jean-Philippe Croue´. Influence of the character of nomon the ozonation of mib and geosmin. Water research, 36(3):511–518, 2002. → pages 94[118] Siva R Sarathy, Mohammad M Bazri, and Madjid Mohseni. Modeling the transformation ofchromophoric natural organic matter during uv/h 2 o 2 advanced oxidation. Journal ofEnvironmental Engineering, 137(10):903–912, 2011. → pages 96[119] Paul Westerhoff, Stephen P Mezyk, William J Cooper, and Daisuke Minakata. Electronpulse radiolysis determination of hydroxyl radical rate constants with suwannee river fulvicacid and other dissolved organic matter isolates. Environmental science & technology,41(13):4640–4646, 2007. → pages 96[120] Matthew Grandbois, Douglas E Latch, and Kristopher McNeill. Microheterogeneousconcentrations of singlet oxygen in natural organic matter isolate solutions. Environmentalscience & technology, 42(24):9184–9190, 2008. → pages 102137[121] Silvio Canonica and Ju¨rg Hoigne´. Enhanced oxidation of methoxy phenols at micromolarconcentration photosensitized by dissolved natural organic material. Chemosphere,30(12):2365–2374, 1995. → pages 103[122] Silvio Canonica, Urs Jans, Konrad Stemmler, and Jurg Hoigne. Transformation kinetics ofphenols in water: photosensitization by dissolved natural organic material and aromaticketones. Environmental science & technology, 29(7):1822–1831, 1995. → pages 103[123] Hrvoje Kusˇic´, Natalija Koprivanac, Ana Loncˇaric Bozˇic´, and Iva Selanec. Photo-assistedfenton type processes for the degradation of phenol: a kinetic study. Journal of hazardousmaterials, 136(3):632–644, 2006. → pages 108, 111[124] Davide Vione, Valter Maurino, Claudio Minero, Paola Calza, and Ezio Pelizzetti. Phenolchlorination and photochlorination in the presence of chloride ions in homogeneous aqueoussolution. Environmental science & technology, 39(13):5066–5075, 2005. → pages 112[125] Swapan Khanra, Claudio Minero, Valter Maurino, Ezio Pelizzetti, Binay K Dutta, andDavide Vione. Phenol transformation induced by uva photolysis of the complex fecl2+.Environmental Chemistry Letters, 6(1):29–34, 2008. → pages 112138Appendix ASupplementary data for thephotocatalysts deactivation tests andevaluation in a flow through reactorIn Figure A.1 and A.2 the data for the deactivation testa are reported. The ratio between the con-centration of 2,4-D at a certain time to the initial concentration was plotted versus the time ofirradiation. Each run was performed three times and the errors bars are the standard deviation ofeach point to the average.In the following tables the efficiency of the photocatalyst in different conditions in the flowthrough reactor is reported. The tables show the inlet and outlet concentration for different times ofsampling. The sampling took place for different time in order to assure that the photoreactor wastable and the efficiency did not depend on absorbtion phenomena on the photocatalyst. The 2,4-Dconversion was calculated according to:Concentrationin−ConcentrationoutConcentrationout×100 (A.1)139Figure A.1: Degradation of 2,4-D in the presence of photocatalyst E after repeated photocat-alytic experiments.Figure A.2: Degradation of 2,4-D in the presence of photocatalyst D after repeated photocat-alytic experiments.140Table A.1: 2,4-D conversion for photocatalysis/UV with millipore water and a flowrate of1L/minTime(min)Concentration (ppm) Concentration (ppm)inlet outlet)0 0.115 0.1131 0.119 0.1122 0.118 0.1143 0.118 0.1144 0.118 0.11295 0.118 0.1136 0.119 0.1127 0.118 0.1148 0.118 0.1149 0.118 23710 0.116 0.11211 0.116 0.10312 0.1171 0.11113 0.118 0.11114 0.117 0.1115 0.1165 0.118Table A.2: 2,4-D conversion for P25 slurry/UV with millipore water and a flowrate of 1L/min.Time(min)Concentration (ppm) Concentration (ppm)inlet outlet)0 0.5 0.481 0.51 0.432 0.49 0.363 0.45 0.344 0.45 0.365 0.44 0.386 0.44 0.337 0.42 0.318 0.5 0.49 0.43 0.310 0.5 0.38141Table A.3: 2,4-D conversion for VUV process with millipore water and a flowrate of 1L/min.Time(min)Concentration (ppm) Concentration (ppm)inlet outlet)0 0.1044 01 0.0861 02 0.0874 03 0.0874 04 0.0873 05 0.0881 06 0.0882 07 0.0873 08 0.0863 09 0.0878 010 0.0880 0Table A.4: 2,4-D conversion for photocatalysis/UV with surface water and a flowrate of 0.25L/min.Time(min)Concentration (ppm) Concentration (ppm)inlet outlet)0 0.1216 0.1081 0.1088 0.10512 0.1077 0.10363 0.1083 0.10284 0.1081 0.10225 0.106 0.10186 0.1081 0.10177 0.1075 0.10128 0.1084 0.1049 0.1082 0.101210 0.1072 0.099142Table A.5: 2,4-D conversion for VUV with surface water and a flowrate of 0.25 L/min.Time(min)Concentration (ppm) Concentration (ppm)inlet outlet)0 0.11 0.11 0.1 0.072 0.08 0.063 0.08 0.064 0.09 0.065 0.09 0.056 0.09 0.057 0.09 0.058 0.09 0.049 0.09 0.0410 0.09 0.05143Appendix BSupplementary data for the UV/H2O2and VUV irradiation in the presence ofdifferent ions.The degradation of atrazine is plotted versus the time of irradiation in the presence of different ionsat different concentrations. Each experiment was conducted in triplicate and the error bars representthe standard deviation.144Figure B.1: Degradation of atrazine in the presence of different concentration of NaF duringUV/ceH2O2 treatment.Figure B.2: Degradation of atrazine in the presence of different concentration of NaF duringVUV treatment.145Figure B.3: Degradation of atrazine in the presence of different concentration of NaNO3 dur-ing UV treatment.Figure B.4: Degradation of atrazine in the presence of different concentration of NaNO3 dur-ing UV/H2O2 treatment.146Figure B.5: Degradation of atrazine in the presence of different concentration of NaNO3 dur-ing VUV treatment.Figure B.6: Degradation of atrazine in the presence of different concentration of NaHCO3during UV/H2O2 treatment.147Figure B.7: Degradation of atrazine in the presence of different concentration of NaHCO3during VUV treatment.Figure B.8: Degradation of atrazine in the presence of different concentration of NaHSO3during VUV treatment.148Figure B.9: Degradation of atrazine in the presence of different concentration of NaHSO3during VUV treatment. In the experiments with NaHSO3 1 ppm of methanol was addedin order to slow the degradation rate and have more stable measurements.149Figure B.10: Degradation of atrazine in the presence of different concentration of Cl– duringUV/H2O2 treatment.Figure B.11: Degradation of atrazine in the presence of different concentration of Cl– duringVUV treatment.150Appendix CSupplementary data for the UV/H2O2and VUV irradiation in the presence ofdifferent concentration an type of NOM.151Figure C.1: Degradation of atrazine in the presence of different concentration of Nordic NOMduring UV/H2O2 treatment.Figure C.2: Degradation of atrazine in the presence of different concentration of Nordic NOMduring VUV treatment.152Figure C.3: Degradation of atrazine in the presence of different concentration of Suwanneeriver NOM during UVH2O2 treatment.Figure C.4: Degradation of atrazine in the presence of different concentration of Suwanneeriver during VUV treatment.153Appendix DSupplementary data for themeasurement of the absorbance ofinorganic ions and NOMIn the following section the graphs and the data for the measurement of the absorbance are reported.It is important to note that in some cases the plot does not pass though zero. This can be explainedby the reflection and scattering phenomena of the quartz cell.Table D.1: Absorbance at 185 nm for different concentration of NaClAbsorbanceConcentration(mol/L)0.9 3.4 x 10−40.7 2.56 x 10−40.5 1.71 x 10−40.4 1.28 x 10−40.2 8.5 x 10−50.1 3.42 x 10−5154Figure D.1: Absorbance of sodium chloride at 185 nm at different sodium chloride concen-trations.Figure D.2: Absorbance of sodium bicarbonate at 185 nm at different sodium bicarbonateconcentrations.155Figure D.3: Absorbance of sodium nitrate at 185 nm at different sodium nitrate concentra-tions.Figure D.4: Absorbance of sodium nitrate at 185 nm at different sodium nitrate concentra-tions.156Figure D.5: Absorbance of sodium sulfate at 185 nm at different sodium sulfate concentra-tions.Figure D.6: Absorbance of Nordic NOM at 185 nm at different TOC concentrations.157Figure D.7: Absorbance of Nordic NOM at 185 nm at different TOC concentrations.Figure D.8: Absorbance of Suwannee NOM at 185 nm at different TOC concentrations.158Figure D.9: Absorbance of Suwannee NOM at 185 nm at different TOC concentrations.159Table D.2: Absorbance at 185 nm for different concentration of NaHCO3AbsorbanceConcentration(mol/L)0.03 4.7 x 10−40.36 1.19 x 10−30.5 2.38 x 10−30.4 4.76 x 10−3Table D.3: Absorbance at 185 nm for different concentration of NaNO3AbsorbanceConcentration(mol/L)6.60 1.17 x 10−34.7 8.82 x 10−43.4 5.88 x 10−41.7 2.94 x 10−40.9 1.17 x 10−40.5 5.88 x 10−50.0.9 1.17 x 10−5Table D.4: Absorbance at 254 nm for different concentration of NaNO3AbsorbanceConcentration(mol/L)0.11 5.88 x 10−40.05 2.94 x 10−40.02 1.17 x 10−40.01 5.88 x 10−4160Table D.5: Absorbance at 185 nm for different concentration of Nordic NOMAbsorbanceConcentration(mg/L)0.8 90.53 4.50.33 2.20.14 0.5161Table D.6: Absorbance at 254 nm for different concentration of Nordic NOMAbsorbanceConcentration(mg/L)0.12 90.03 1.128 x 10−3 2.22 x 10−3 0.5Table D.7: Absorbance at 185 nm for different concentration of HSO4−AbsorbanceConcentration(mg/L)2.81 x 10−3 0.432.11 x 10−3 1.121.40 x 10−3 2.27.04 x 10−4 0.13.52 x 10−4 0.07162Appendix EQuantum yield of micropollutantsThe quantum yield of organics is defined as the ratio between the degradation of atrazine and therate of photons absorption according to:φAtz,254nm =∫V rratz,254nmdV∫V rrphotonsdV(E.1)Where the numerator represents the rate of atrazine degradation and the denominator the rate of thephotons absorption.∫Vrratz,254nmdV =−∫VrkAtzC0AtzdV −∫VrkAtzC0AtzdV = kAtzC0AtzVrv (E.2)On the other hand,∫VrrphotonsdV = G−G0 (E.3)where G is the rate of photons on the surface of the reactor vessel, and G0 is the rate of photonsout. According to the Lambert beer law:dIdL= εatzCatzI (E.4)For collimated radiation:163dGdL= εatzCatzGL=VrvArvG= G010−εatzC0atzVrvArv∫VrrphotonsdV = G0(1−10εatzCatzVrvArv(E.5)thus the quantum yield is equal to:φAtz,254nm =kphot254,AtzVrv ·C0AtzG0254(1−10−εAtz,254C0AtzVrvArv )(E.6)164Appendix FDegradation of micropollutants withozone-generating Hg lamps (185 and 254nm)With relatively low concentration of H2O2 (less than 1 ppm), HO radicals are generated mainly forwater photolysis and the contribution of H2O2 photolysis is negligible.rHO,gen(x) = G185(x)(φ1,185+φ2,185)εH2O,185ln(10)CH2O (F.1)where G185(x) is the local incident radiation, φ1,185 and φ2,185 are the quantum yield of water at 185nm, εH2O,185 is the absorption coefficient of water at 185 nm, andCH2O is the concentration of water.In the presence of organics and inorganics HO radicals react almost exclusively with them. Underthose conditions, the rate of consumption of HO radicals can be approximately:rHO,con = ∑Ni=1 ki−HO ·Ci ·CHO (F.2)where the summation includes all those stable species present in the sample that can react withHO radicals, Ci is the initial concentration of i specie and ki−HO is the reaction rate between165specie i and HO radical. Applying the local steady state concentration assumption to HO radicals(rHO,gen=rHO,con) the concentration of HO radicals can be calculated:CHO(x) =G185(x)(φ1,185+φ2,185) · εH2O,185 · logCH2O∑Ni=1 ki−HO ·Ci(F.3)where CHO(x) is the local HO• concentration, G185(x) is the incident radiation, ε1,185 is the quantumyield of the photochemical homolysis of water (0.33), and ε2,185 is the quantum yield of the photo-chemical ionization of water (0.045). By assuming that micropollutants are degraded via HO attackand photolysis, their reaction can be obtained as:rMP = kAtz−HOCMPCHO(x)+G185(x)φMP,185εMP,185ln(10)CMP+G254(x)φMP,254εMP,254ln(10)CMP(F.4)where kMP−HO is the kinetic constant between the micropollutant and HO radicals, CMP is the con-centration of micropollutant, CHO(x) is the local concentration of HO radicals, G185(x) is the localincident radiation at 185 nm, φMP,185 is the quantum yield of micropollutant at 185 nm, εMP,185 is theabsorption coefficient of micropollutant at 185 nm, G254(x) is the local incident radiation at 254 nm,φMP,254 is the quantum yield of micropollutant at 254 nm and εMP,254 is the absorption coefficient ofmicropollutant at 254 nm. Taking into account that the concentration of micropollutants is uniformin the irradiation vessel (perfect mixing), the volume averaged degradation rate can be calculatedas:[rMP(x)]vol = (KMP−HO+KMP−185+KMP−254)CMP (F.5)whereKMP−HO =ArvVrv∫ArvVrv0kMP−HOCHO(x)dx (F.6)where Arv is the area of the reaction vessel and Vrv is the volume of the reaction vessel.KMP−185 =ArvVrv∫ArvVrv0G185(x)φMP,185εMP,185ln(10)CMPdx (F.7)166KMP−254 =ArvVrv∫ArvVrv0G254(x)φMP,254εMP,254ln(10)CMPdx (F.8)On the other hand, the local incident radiation can be obtained by using the Beer-Lambert law:G185(x) = G0185exp(−(n∑i=1εi,185Ci)ln(10)x)dx (F.9)G254(x) = G0254exp(−(n∑i=1εi,254Ci)ln(10)x)dx (F.10)by replacing equations F.2 and F.9 into equation F.6KMP−HO =ArvVrvCHO(x)G185(x)(φ1,185+φ2,185) · εH2O,185 · logCH2O∑Ni=1 ki−HO ·Ci∫ArvVrv0exp(−(n∑i=1εi,185Ci)ln(10)x)dx(F.11)As the concentration of micropollutants in raw water is low, they do not contribute significantly tothe total absorbtion coefficient at 185 nm. Therefore, the previous equation can be solved to obtain:KMP−HO =kMP−HO(G0185−Gout185)(φH2O,185+φH2O,185)εH2O,185CH2OArv(∑ni=1 ki−HOCi)(∑ni=1 εi,185Ci)Vrv(F.12)In a similar way equation F.7 and can be operated to obtain:KMP−HO=(G0185−Gout185)(φMP,185εMP,185Arv(∑ni=1 εi,185Ci)Vrv(F.13)KMP−HO=(G0254−Gout185)(φMP,254εMP,254Arv(∑ni=1 εi,254Ci)Vrv(F.14)Finally, by applying a mass balance to the reaction vessel operating under batch mode:CMP =C0MPexp(−(KMP−HO+KMP−185+KMP−254)t) (F.15)167

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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0167680/manifest

Comment

Related Items