Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The use of 185 nm radiation for drinking water treatment Furatian, Laith 2017

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

Item Metadata

Download

Media
24-ubc_2017_september_furatian_laith.pdf [ 1.12MB ]
Metadata
JSON: 24-1.0348109.json
JSON-LD: 24-1.0348109-ld.json
RDF/XML (Pretty): 24-1.0348109-rdf.xml
RDF/JSON: 24-1.0348109-rdf.json
Turtle: 24-1.0348109-turtle.txt
N-Triples: 24-1.0348109-rdf-ntriples.txt
Original Record: 24-1.0348109-source.json
Full Text
24-1.0348109-fulltext.txt
Citation
24-1.0348109.ris

Full Text

The Use of 185 nm Radiation forDrinking Water TreatmentInfluence of Major Solutes and Temperature on theDegradation of Trace Organic ContaminantsbyLaith FuratianB.Sc., The University of Alberta, 1999M.Sc., The University of New Hampshire, 2011A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinThe Faculty of Graduate and Postdoctoral Studies(Chemical and Biological Engineering)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)June 2017© Laith Furatian 2017AbstractThe treatment of water via 185 nm radiation allows for the oxidative degra-dation of trace organic contaminants without the need for chemical addition.Critical information required for the practical application of such a processhas been lacking. Carbamazepine was determined to be an ideal probe com-pound for study of the 185 nm regime due to negligible direct photolysis at254 nm. An increase in probe degradation rate due to 185 nm is observedwith increasing temperature when water is the only significant absorber ofphotons. A comparison with the temperature dependence of the 254 nm -H2O2 process is made and a fundamental explanation proposed. Experi-mental evidence reveals that probe degradation rate is strongly influencedby anionic composition at environmentally relevant concentrations, partic-ularly chloride. Evidence for the role of the chlorine radical is obtained bykinetic studies involving select probes, radical scavengers, and ionic strength.Interactions between the major organic and inorganic solutes indicate thatresulting degradation kinetics are highly sensitive to the composition of thewater matrix, a fact that has been neglected from the literature. A methodto quantify molar absorption coefficients is developed that is not prone toerrors due to stray radiant energy or wavelength inaccuracies. A method toquantify the 185:254 nm output of a low pressure mercury lamp is presentedwith results in agreement with values reported in the literature. In additionto the hydroxyl radical ( OH), other radical species such as chlorine (Cl )and sulphate (SO –4 ) are proposed to be involved in oxidative degradationof trace organics in the 185 nm regime. This suggests that the degradationrate of a given target contaminant depends on the composition of the watermatrix, the second-order rate constants with the relevant radicals, and therelative reaction rate constants of the target and the matrix.iiLay SummaryCertain types of chemical impurities are difficult to remove from water. Onemethod that may be useful in destroying certain chemicals is to split watermolecules using ultraviolet radiation to produce an aggressive oxidant thatwill then attack impurities. However, this process is highly dependent onthe types and amounts of naturally occurring substances also dissolved inwater. A complex mixture of oxidants may be produced with potentiallysignificant differences in their abilities to attack a given impurity. Thus,since the composition of water differs between places and over the seasons,the efficiency of this approach may differ substantially between locations andover time. An attempt to understand how the natural composition of waterand its temperature influence the efficiency of this process were the goalsof this research. This information could then be used to optimize the pro-cess and even determine under what conditions its use would be impractical.iiiPrefaceThis dissertation is original, unpublished, independent work by the author,Laith Furatian.ivTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiLay Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . vList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xNomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiiAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . xvDedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Public Drinking Water Supplies . . . . . . . . . . . . . . . . 11.2 A Recent History of Water Treatment . . . . . . . . . . . . . 41.3 UV Disinfection . . . . . . . . . . . . . . . . . . . . . . . . . 101.4 Advanced Oxidation Processes . . . . . . . . . . . . . . . . 131.5 185 nm Advanced Oxidation . . . . . . . . . . . . . . . . . . 181.6 Knowledge Gaps . . . . . . . . . . . . . . . . . . . . . . . . . 211.7 Research Objectives . . . . . . . . . . . . . . . . . . . . . . . 272 Experimental Approach . . . . . . . . . . . . . . . . . . . . . . 292.1 General Approach . . . . . . . . . . . . . . . . . . . . . . . . 29vTable of Contents2.2 The Probe Compound and the Observable k′ . . . . . . . . . 292.3 Expressing the Extent of Reaction . . . . . . . . . . . . . . . 322.4 A 185 nm Kinetic Model . . . . . . . . . . . . . . . . . . . . 362.5 Experimental Apparatus . . . . . . . . . . . . . . . . . . . . 402.6 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432.7 Probe Compound Selection and Characterization . . . . . . . 432.7.1 Molar Absorption Coefficients at 254 nm . . . . . . . 452.7.2 Photolysis Quantum Yields at 254 nm . . . . . . . . 452.7.3 Second-order OH Rate Constants . . . . . . . . . . . 462.8 Analytical Methods . . . . . . . . . . . . . . . . . . . . . . . 472.9 Initial Testing of 185 nm Experimental Methods . . . . . . . 483 Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543.1 Temperature and Water Treatment . . . . . . . . . . . . . . 543.2 Temperature Dependence of UV AOPs . . . . . . . . . . . . 543.3 Experimental Approach . . . . . . . . . . . . . . . . . . . . . 583.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593.4.1 Temperature Effects in the 254 nm - H2O2 Regime . 593.4.2 Temperature Effects in the 185 nm Regime . . . . . . 613.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664 Dissolved Organic Matter . . . . . . . . . . . . . . . . . . . . 674.1 Dissolved Organic Matter in Natural Waters . . . . . . . . . 674.2 The 185 nm AOP and Influence of Dissolved Organic Matter 704.3 Use of Reference Materials . . . . . . . . . . . . . . . . . . . 714.4 Pure Substances as Model Organic Matter . . . . . . . . . . 754.5 Estimation of 185 nm Incident Fluence Rate . . . . . . . . . 824.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845 Chloride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865.1 Chloride in Natural Waters . . . . . . . . . . . . . . . . . . . 865.2 Impact of Chloride on AOPs . . . . . . . . . . . . . . . . . . 875.3 Chloride in the 254 nm - H2O2 Regime . . . . . . . . . . . . 90viTable of Contents5.4 Chloride in the 185 nm Regime . . . . . . . . . . . . . . . . 925.5 Relative Reactivity of OH and Cl . . . . . . . . . . . . . . . 1005.6 Evidence for Cl from Probe-Scavenger Systems . . . . . . . 1015.7 Evidence from Ionic Strength Effects . . . . . . . . . . . . . 1075.8 Product Studies of Phenol Degradation . . . . . . . . . . . . 1125.9 Bleaching of Dissolved Organic Matter in 185 nm Regime . . 1135.10 Molar Absorption Coefficient of Chloride at 185 nm . . . . . 1145.11 Potential Cl to OH Interconversion . . . . . . . . . . . . . . 1185.12 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1196 Sulphate, Bicarbonate and Interactions of Major Solutes 1226.1 The 185 nm AOP and Other Solutes . . . . . . . . . . . . . . 1226.2 Sulphate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1236.2.1 Sulphate in the 254 nm - H2O2 Regime . . . . . . . . 1246.2.2 Sulphate in the 185 nm Regime . . . . . . . . . . . . 1276.3 Bicarbonate . . . . . . . . . . . . . . . . . . . . . . . . . . . 1316.4 Interactions Among Major Solutes . . . . . . . . . . . . . . . 1376.4.1 Sulphate and Bicarbonate . . . . . . . . . . . . . . . 1406.4.2 Chloride and Bicarbonate . . . . . . . . . . . . . . . . 1436.4.3 Chloride, Sulphate, and Bicarbonate . . . . . . . . . 1466.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1517 Conclusions and Recommendations . . . . . . . . . . . . . . 1537.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1537.2 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . 156Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158AppendicesA Experimental Data and Calculations . . . . . . . . . . . . . . 179viiList of Tables1.1 Molar absorption coefficient () and quantum yield (Φ) forCl– and SO 2–4 at 185 nm . . . . . . . . . . . . . . . . . . . . . 221.2 Comparison of rate constants for OH, Cl , and SO –4 withHCO –3 and CO2–3 . . . . . . . . . . . . . . . . . . . . . . . . . 241.3 Important reactions involving Cl , Cl –2 , and SO–4 . . . . . . 252.1 Photochemical reaction parameters for probe compounds at254 nm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443.1 Effect of temperature on removal rate of carbamazepine probein 254 nm-H2O2 regime . . . . . . . . . . . . . . . . . . . . . . 593.2 Effect of temperature on removal rate of carbamazepine probein 185 nm regime . . . . . . . . . . . . . . . . . . . . . . . . . 633.3 Experimental activation energies for carbamazepine degra-dation in the presence of tert-butanol in 254 nm-H2O2 and185 nm regimes . . . . . . . . . . . . . . . . . . . . . . . . . . 633.4 Summary of fundamental activation energies estimated fromthis work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644.1 The removal rate of probe for varying concentration and sourceof DOM in 185 nm regime . . . . . . . . . . . . . . . . . . . . 724.2 Selected pure compounds used as model organic matter instudies of 185 nm regime . . . . . . . . . . . . . . . . . . . . . 754.3 The removal rate of probe for pure scavengers tert-butanol,methanol, and acetone in 185 nm regime . . . . . . . . . . . . 79viiiList of Tables5.1 Comparison of OH and Cl reactivities for pure compoundsused as model organic matter in the 185 nm regime . . . . . . 1025.2 Model evaluation using the degradation of nitrobenzene withincreasing Cl– in the 185 nm regime . . . . . . . . . . . . . . . 1186.1 Numerical values used in calculation of the second-order rateconstant for the reaction of SO –4 with carbamazepine (kSO –4 ,CBZ)with equation 6.2 . . . . . . . . . . . . . . . . . . . . . . . . . 1306.2 Experimental determination of the second-order rate constantfor the reaction of SO –4 with carbamazepine (kSO –4 ,CBZ) withequation 6.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1306.3 Comparison of the calculated and experimental rates due toHCO –3 in the 185 nm regime . . . . . . . . . . . . . . . . . . . 1366.4 Comparison of calculated and experimental rates due to SO 2–4and HCO –3 in the 185 nm regime . . . . . . . . . . . . . . . . 1416.5 Experimental degradation rates of carbamazepine due to Cl–and HCO –3 in the 185 nm regime . . . . . . . . . . . . . . . . 1446.6 Interactions of Cl–, SO 2–4 and HCO–3 , in the presence of Suwan-nee River NOM . . . . . . . . . . . . . . . . . . . . . . . . . . 150ixList of Figures2.1 The 185 nm Collimated Beam Apparatus . . . . . . . . . . . 423.1 Temperature dependence in 254 nm - H2O2 regime. . . . . . . 603.2 Arrhenius plots for 254 nm-H2O2 regime . . . . . . . . . . . . 613.3 Temperature dependence in 185 nm regime. . . . . . . . . . . 623.4 Arrhenius plots for 185 nm regime . . . . . . . . . . . . . . . 634.1 Suwannee River DOM in 185 nm regime. . . . . . . . . . . . 734.2 Nordic DOM in 185 nm regime. . . . . . . . . . . . . . . . . . 744.3 Removal rate of probe with type and concentration of DOM . 744.4 Tert-Butanol in 185 nm regime. . . . . . . . . . . . . . . . . . 764.5 Methanol in 185 nm regime. . . . . . . . . . . . . . . . . . . . 774.6 Acetone in 185 nm regime. . . . . . . . . . . . . . . . . . . . . 774.7 Removal rate of probe with pure compounds as model DOM 785.1 Tert-Butanol in the 254 nm - H2O2 regime and influence ofchloride. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 915.2 Suwannee River NOM in the 254 nm - H2O2 regime and in-fluence of chloride. . . . . . . . . . . . . . . . . . . . . . . . . 925.3 Tert-Butanol in the 185 nm regime and influence of chloride,using carbamazepine as a probe. . . . . . . . . . . . . . . . . 935.4 Suwannee River NOM in the 185 nm regime and influence ofchloride, using carbamazepine as a probe. . . . . . . . . . . . 945.5 Tert-Butanol in the 185 nm regime and influence of chloride,using nitrobenzene as a probe. . . . . . . . . . . . . . . . . . 98xList of Figures5.6 Suwannee River NOM in the 185 nm regime and influence ofchloride, using nitrobenzene as a probe. . . . . . . . . . . . . 995.7 Comparison of second-order rate constants of select organicsolutes with OH and Cl at 25 ◦C . . . . . . . . . . . . . . . 1005.8 Acetate system at 185 nm with both carbamazepine and ni-trobenzene as probes. . . . . . . . . . . . . . . . . . . . . . . 1055.9 Acetone system at 185 nm with both carbamazepine and ni-trobenzene as probes. . . . . . . . . . . . . . . . . . . . . . . 1065.10 Influence of ionic strength in 185 nm regime on the degrada-tion of carbamazepine in the presence of tert-butanol and nochloride. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1095.11 Influence of ionic strength in 185 nm regime on the degra-dation of carbamazepine in the presence of tert-butanol andchloride. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1105.12 Influence of ionic strength on the degradation of carbamazepinewith and without chloride. . . . . . . . . . . . . . . . . . . . 1115.13 Kinetic method of determining 185,Cl− using double cell. . . 1165.14 Calculation of 185,Cl− from kinetic data. . . . . . . . . . . . 1176.1 Tert-Butanol in the 254 nm - H2O2 regime and influence ofsulphate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1256.2 Swannee River NOM in the 254 nm - H2O2 regime and influ-ence of sulphate. . . . . . . . . . . . . . . . . . . . . . . . . . 1266.3 Influence of sulphate in the 185 nm regime with tert-butanol. 1296.4 Influence of bicarbonate in the 185 nm regime with tert-butanol.1356.5 Comparison of the calculated and experimentally observedrates due to HCO –3 influence in the 185 nm regime . . . . . . 1376.6 Interaction of sulphate and bicarbonate in the 185 nm regimewith tert-butanol. . . . . . . . . . . . . . . . . . . . . . . . . 1436.7 Interaction of chloride and bicarbonate in the 185 nm regimewith tert-butanol. . . . . . . . . . . . . . . . . . . . . . . . . 1456.8 Interaction of chloride and sulphate in the 185 nm regimewith Suwannee River NOM and without bicarbonate. . . . . 147xiList of Figures6.9 Interaction of chloride and sulphate in the 185 nm regimewith Suwannee River NOM and with bicarbonate. . . . . . . 148xiiNomenclatureSymbolsαi Absorption coefficient of solute i (cm−1)[C]o Initial concentration of C (M)[C]t Concentration of C after elapsed time t (M)D Absorbed photons per unit volume (J m−3 or M)i Molar absorption coefficient of solute i (M−1 cm−1)E Activation energy (kJ mol−1)F Fluence rate (J m−2)fi Fraction of photons absorbed by species iIa Photon absorption rate (M s−1 or J m−3 s−1)Io Incident photon fluence rate (mol m−2 s−1 or J m−2 s−1)k′ Pseudo-first order rate constant, time based (s−1)k′′ Pseudo-first order rate constant, fluence based (M−1 or m2 J−1)kA,B second order rate constant between species A and B (M−1 s−1)` Optical path length (cm)Φ Quantum yield[ OH]ss Steady-state hydroxyl radical concentration (M)[Cl ]ss Steady-state chlorine radical concentration (M)[SO –4 ]ss Steady-state sulphate radical concentration (M)R Universal gas constant (J K−1 mol−1)t Exposure time (s)V Volume (L or cm3)xiiiNomenclatureAcronymsAOP Advanced Oxidation ProcessAWWA American Water Works AssociationCBZ CarbamazepineDBP Disinfection By-ProductDOC Dissolved Organic CarbonDOM Dissolved Organic MatterGAC Granular Activated CarbonHPLC High-Performance Liquid ChromatographyIHSS International Humic Substances SocietyMIB 2-MethylisoborneolNB NitrobenzeneNOM Natural Organic MatterPAC Powdered Activated CarbonpCBA para-Chlorobenzoic AcidTHM TrihalomethaneTOC Total Organic CarbonxivAcknowledgementsDoctoral studies are a very personal undertaking that can too often becomea lonely pursuit. Yet, I was spared a completely solitary and ascetic exis-tence during this work by the generous support of others.I thank Dr. Madjid Mohseni for providing me an opportunity, his patience,and the autonomy to conduct scientific research at this level. I am alsothankful for the support of my supervisory committee members Dr. BenoitBarbeau, Dr. Xiaotao Bi, and in particular Dr. Domenico Santoro for hisgenuine enthusiasm for the topic. Thanks to Dr. Naoko Ellis, Dr. AllanBertram, and particularly Dr. Ron Hofmann for being motivated examin-ers. And thanks to Dr. Jim Malley for introducing me to the fields of UVdisinfection and advanced oxidation back in New Hampshire, which set meon this path, as well as for his continued mentorship from afar.Scientific research is a messy endeavour. Dialogue with others greatly assistsboth progress and avoiding dead ends. Several individuals have providedmuch valuable discussion and insight. I thank Doug Yuen for his assistancein the design and fabrication of the apparatus, Timothy Ma and PaulaParkinson for their advice regarding chemical analysis, Dr. Mike Thurmanfor sharing his wealth of experience as a researcher, and Dr. Jim Bolton forhis erudite perspective on the topic.As Titus Plautus wrote, “Where there are friends there are riches”, and Ihave been fortunate to have had friends provide moral support, encourage-ment, curiosity in my work, technical advice, and above all companionship.I thank (in no particular order) Lee Hupka, Marv Clark, Ernest McCrank,xvAcknowledgementsChris Lawson, Bruce Haines, Lisa Jensen, Erin Ziegenfuss, Jennifer Cane,Anthony Kennedy, Sean McBeath, Herbert Hartshorn, Gregory Marshall,Madalena Santos, Aaron Dublenko, Cyrus Perron, Rajka Rada Jovic, Kos-mas Panagiotidis, Wilson Wong, Hayder Salem, and Dr. Norman Epstein.Forgive me if I have forgotten you.No greater support is possible than that of my family. The love and kindnessof Anna, our daughter Muna, and Anna’s parents David and Pamela, havegiven me the energy to complete this task. Anna and Muna are my reasonsfor living and have taught me things school never could. Yet, the greatestthanks are to my mother, who has supported me from my first breath, sac-rificed so much, and asked for so little. Thank you.xviDedicationTo Sonia, Anna, and Muna.xviiChapter 1Introduction1.1 Public Drinking Water SuppliesThe provision of safe drinking water supplies has been one of the triumphsof public health. The quality of life enjoyed in developed countries dependson this largely “invisible” service. Where it is absent, an additional burdenis placed upon daily life. Yet, where such service is successful, it is oftenunder appreciated. Success involves substantial technical, economic, and po-litical organization. Criteria for success include adequate safety for humanconsumption, acceptable aesthetics, sufficient capacity to satisfy demand,reliability of service, and reasonable cost.Water in the environment is often unfit to drink without sufficient treatmentdue to the presence of both naturally occurring and anthropogenic contami-nants. Such contaminants include microbiological, chemical and radiologicalhazards that may pose an unacceptable risk to human health if consumed.The primary goal of water treatment is to lower the concentration of haz-ards such that the calculated risk from exposure is reduced to a level deemedacceptable. A secondary but fundamentally important goal of treatment isthat water be rendered aesthetically adequate to consumers. A third goalis the protection of the distribution system integrity by adjustments of suchparameters as pH, alkalinity and hardness, to produce a water that is nei-ther corrosive nor scale forming.Water for which the risk of harm from consumption is acceptable is con-sidered “safe”. However, such risk can never be reduced to zero, and thedetermination of what constitutes acceptable risk, and therefor the extent11.1. Public Drinking Water Suppliesof treatment, is ultimately an economic and political decision. Whether awater is deemed aesthetically acceptable or “palatable” is a determinationmade by consumers, and is generally independent of safety. Aesthetic issueshave been reported as the most common complaint to utilities by consumersand historically have motivated additional treatment beyond that requiredto produce “safe” water. Consumers tend to associate poor aesthetics ofwater with poor quality, regardless of assurances to the contrary. Thus sup-pliers address aesthetics to maintain consumer confidence.Water supply sources are usually selected to be of the best available quality,and thus require the least amount of treatment. Compared to groundwaters,surface waters harbour greater microbial hazards, are subject to greater fluc-tuations in quality and thus usually require greater treatment. Yet, whengroundwater supplies are insufficient to meet demand, surface waters mustbe exploited. Consequently, large population centres tend to be served bytreated surface water. A wide variety of treatment approaches have evolvedin developed countries, depending on site specific challenges as well as cul-tural preferences. In North America, treatment by stages involving chemi-cal coagulation and flocculation, clarification, and filtration is referred to as“conventional treatment”, and has typically been followed with disinfectionby chlorine. The adoption and spread of this process is associated with thedisappearance of the once common waterborne outbreaks of typhoid andcholera (Logsdon and Lippy, 1982; Wolman and Gorman, 1931).In regions where water is scarce, even sources of heavily impaired quality maybe considered to supply drinking water. In some cases, it has been deemedfeasible to recycle wastewater for domestic purposes. The City of WindhoekNamibia has been practicing direct potable reuse since 1958 (Du Pisani,2006). Singapore and Orange County California are other early adoptersthat have found potable reuse expedient, and the numbers of such facilitiesare increasing. In arid yet heavily populated regions, raw sewage representsa reliable source, yet requires extensive processing beyond the capabilities ofconventional treatment. Nevertheless, the alternative is often sea water de-21.1. Public Drinking Water Suppliessalination or importation via aqueducts over great distances, both of whichrequiring substantially higher expenditure of energy than that needed forthe potabalization of municipal wastewater. At the current time, the energyconsumption of reverse osmosis desalination of sea water is approximately3− 6 kW h m−3, compared to about 0.1 kW h m−3 required for conventionaltreatment (Howe et al., 2012). These values serve as useful references whenconsidering the feasibility of alternative treatment technologies required formore challenging sources such as those of water reuse.Water treatment is a constantly evolving field. It is far from being a solvedproblem, and there will likely never exist a unique solution able to addressall situations adequately. With environmental changes, scientific develop-ments, and shifting cultural attitudes, prevailing practices will continuouslybe subject to scrutiny and reevaluation. At the current time, the basic toolsof water suppliers have not fundamentally changed over the past half cen-tury. Turbidity, chlorine residual, and filter backwash frequency are amongthe daily concerns of treatment plant personnel. Turbidity is used as a con-venient measure of treated water quality, yet is essentially a surrogate forparticulate matter. Where particle counters are used, such readings are sig-nificant as surrogates for the presence of microbes that are either planktonicor attached to particulates. Discreet measurements of microbial quality, re-quiring lengthy laboratory procedures, are still largely evaluated using indi-cator organisms as surrogates for the direct detection of pathogens (Gleesonand Gray, 1997). The deficiencies in these nested surrogates are well known.Even if the formidable technological challenge of direct pathogen detectionis eventually achieved, the emergence of previously unknown pathogens willlikely remain a possibility. And regardless of purity, the quality of the finalproduct must satisfy the judgment of the consumer’s palate. The unpre-dictability of health and aesthetic based risk has motivated a precautionaryphilosophy to treatment and the adoption of a multiple barrier approach. Asnew knowledge elucidates known risks and discovers new ones, the practiceof water treatment will continue to adapt for the foreseeable future.31.2. A Recent History of Water Treatment1.2 A Recent History of Water TreatmentThe early twentieth century spread of water treatment practices, filtrationand chlorine disinfection in particular, coincide with decreasing rates of in-fant mortality and longer life expectancies. Yet, with these successes, newchallenges emerged. Environmental pollution strained the capabilities ofwater utilities to provide a product that was not only safe but also aesthet-ically acceptable. Chlorine, the very agent involved in eliminating the oncefrequent typhoid and cholera outbreaks, created nuisances when applied towaters impaired by industrial and urban pollution. Episodic algal bloomssubjected to chlorine resulted in the sudden release of potent “earthy” and“musty” taste and odour compounds. Phenols from industrial discharges,primarily the coke furnace waste streams of iron production, reacted withchlorine to produce chlorophenols, imparting notoriously strong and un-pleasant odour to treated water (Baker and Taras, 1981).During the period spanning the 1920s to 1960s, drinking water suppliersstruggled against the consequences of pollution before the sources them-selves were mitigated. Alternative disinfectants that did not react withphenol were developed. In North America, the chloramination process, in-volving the sequential addition of ammonia followed by chlorine to formchloramines, gained popularity as it was found that chloramines are germi-cidal yet of negligible reactivity with phenols. Chlorine dioxide similarly didnot react with phenols and was deemed an effective disinfectant. Its use,however, was limited due to cost and proprietary equipment requirements.Adsorption of contaminants by powdered activated carbon (PAC), while acostly material, was found to be effective and convenient for managing briefperiodic events attributed to algae. To this day, many North American wa-ter treatment plants are equipped to deploy PAC for this same purpose.In Europe the application of ozone, which had been superseded by inexpen-sive chlorine, experienced a resurgence in interest as it was able to degradephenols while providing disinfection. Granular activated carbon (GAC)41.2. A Recent History of Water Treatmentwas regularly used to improve continuous taste and odour issues as wellas for dechlorination. Europe’s heavily polluted rivers such as the Rhine,the Meuse, and the Seine carried elevated levels of ammonia that requiredsubstantial doses of chlorine for removal by break-point chlorination. A de-sirable alternative was found in the sequential combination of ozone andGAC. Ozonation increased dissolved oxygen and oxidized organics refrac-tory to biodegradation into more biodegradable products. GAC, applied aspacked beds, provided a high surface area that could be colonized by biofilmswhose growth was supported by dissolved oxygen and the more bioavailableorganic products of ozonation. The resulting process provided oxidation ofphenols, biological nitrification of ammonia to nitrate, and the biologicaluptake of organics. The result is the production of a final effluent of loworganic content with a typical chlorine demand on the order of 0.1 mg L−1.Such an effluent is considered biologically stable, as it inhibits the growthof bacterial slimes in the distribution system (Rice and Robson, 1982).By the end of the 1960s, industrial and urban sources of water pollutionwere being addressed directly. In North America, chloramines and PACwere in common use in conjunction with conventional treatment. In Eu-rope, alternatives to chlorine, particularly ozone, and biological treatmentusing GAC were established practices. Waterborne outbreaks, though dras-tically reduced, were not completely eliminated, with many instances ofgastrointestinal illness still being reported and traced to drinking water.One of the main causes for this was believed to be the noncompliance ofa large number of water utilities, particularly smaller systems with limitedresources, to existing regulations and recommended practices. A large frac-tion of the reported outbreaks were of unknown etiology, and some suspectedthe existence of pathogens, most likely viruses, that were neither adequatelyremoved by conventional treatment nor sufficiently susceptible to chlorinedisinfection, as the culprit for the continued waterborne illness rates (Cook-son Jr, 1974; Craun et al., 1976).Meanwhile, technological advances in analytical instrumentation allowed the51.2. A Recent History of Water Treatmentdetection and quantification of chemical contaminants in water at the levelof parts-per-billion or less. Gas-chromatography coupled with the flameionization detector (GC-FID) allowed sensitive detection of hydrocarbons,while the advent of the electron capture detector (GC-ECD) allowed verysensitive detection of halogenated organics (Bellar and Lichtenberg, 1974;Mieure, 1977). Mass spectrometry (GC-MS) facilitated identification via de-termination of molecular structure (Lingg et al., 1977). These instrumentstogether with associated innovations in aqueous sample preparation wouldhave a profound impact on water treatment practice.In 1972, a now famous environmental investigation was conducted in therapidly industrialized Mississippi Delta. Motivated by taste and odour com-plaints in the tap water of New Orleans, and using cutting edge instrumen-tation, 36 organic contaminants were detected and identified in both Mis-sissippi River water and treated water from the Corrollton water treatmentplant (USEPA, 1972). In 1974, J.J. Rook working at the water treatmentplant of Rotterdam in the Netherlands, discovered that the addition of chlo-rine reacted with naturally occurring organic matter (NOM) to form halo-genated organics, primarily trihalomethanes (THMs), mostly chloroform,none of which were detected in the raw water (Rook, 1974). Independentconfirmation of Rook’s findings in the USA by Bellar et al. (1974) furtherexacerbated the trace contaminant crisis as chloroform had recently beendeclared a carcinogen by the National Cancer Institute (Page, 1976). Con-troversy erupted from these discoveries and the belief prevailing at the timethat cancer was primarily caused by environmental exposure to chemicals(i.e. extrinsic factors). In late 1974, President Ford signed the Safe DrinkingWater Act (SDWA) into law, giving the recently formed US EnvironmentalProtection Agency (USEPA) statutory responsibility over the safety of thenation’s public drinking water systems, involving regulatory, monitoring andenforcement powers. Similar legislation was soon passed in other developedcountries (Sayre, 1988).Following the discovery of halogenated disinfection by-products (DBPs), in-61.2. A Recent History of Water Treatmentcluding THMs and others, the USEPA announced a nation wide survey of 80cities to determine the extent of the THM phenomena. Completed in 1975,the National Organics Reconnaissance Survey (NORS) revealed THMs to bewidespread in chlorinated drinking water (Symons et al., 1975). The discov-ery of industrial contaminants in New Orleans drinking water and the reve-lations of the NORS prompted the USEPA to initiate the National OrganicsMonitoring Survey (NOMS) covering 113 community water supplies during1976 to 1977, including a range of source types and treatment processes(USEPA, 1978). The findings of both the NORS and NOMS identified over700 distinct organic contaminants collectively present in US drinking waterat the part-per-billion level or less, with THMs being the most widespread(Cotruvo and Wu, 1978). The USEPA and the drinking water industry inNorth America scrambled to investigate alternative methods of removingthe newly discovered potential hazards from finished drinking water (Brodt-mann Jr and Russo, 1979; LePage, 1981; Norman et al., 1980; Rice et al.,1981; Suffet, 1980).In Europe, the discovery of chlorinated DBPs accelerated the spread of bio-logical treatment using the combination of ozone and GAC and reduction inthe use of chlorine disinfection, while North American interest in Europeanpractices increased (Heilker, 1979; Knoppert et al., 1980; Ku¨hn et al., 1978;Rapinat, 1982; Schalekamp, 1979; Schulhof, 1979; Sontheimer et al., 1978).However, in North America, greater efforts were directed at modifying ex-isting practices to prevent DBP formation. Chlorine disinfection kineticsinvolves a timescale of hours in the treatment plant, while the formation ofhalogenated DBPs occurs over a timescale of days in the distribution sys-tem. Thus, the further discovery that chloramines do not readily form THAsduring distribution timescales prompted some utilities to convert chlorine tochloramines by ammonia addition following disinfection and before distri-bution. Such a conversion involved minimal capital cost, yet neglected thepossibility that chloramines might produce other DBPs that were yet to bediscovered. Another approach later used in North America to prevent theformation of DBPs was the augmentation of conventional treatment to re-71.2. A Recent History of Water Treatmentmove NOM prior to chlorine addition. The process now known as “enhancedcoagulation”, involving higher coagulant doses, results in moderate increasesin NOM removal, but is limited in coagulant dose by available alkalinity andresults in an increase in residual sludge production and disposal cost (Crozeset al., 1995). The USEPA initially proposed a requirement for the use ofGAC as a barrier against synthetic organic contaminants where utilities werevulnerable to industrial pollution. The requirement would have included alarge fraction of the nation’s treatment plants but was never implementeddue in part to concern regarding cost (Symons, 1984). The ability of ozoneto efficiently oxidize many organic compounds stimulated research in whatwould come to be known as advanced oxidation processes (AOPs), discussedbelow.Much debate in the scientific and water treatment community concernedthe significance of trace organic contaminants and their removal. Rice crit-icized the concept of the “alternative disinfection” approach to avoidingTHM formation, arguing a change in disinfectant would simply result information of a different set of by-products, and that efforts should be con-centrated on removing NOM prior to disinfection (Rice and Cotruvo, 1978).Rosen cautioned that the long list of trace organic contaminants that hadbeen detected in drinking water using GC techniques were the most volatilecompounds, perhaps representing less than 10% of the total, and that theuse of methods such as liquid chromatography would reveal the presenceof a much longer list of contaminants (Rosen, 1976). Regulations basedon the health significance of trace contaminants was questioned by Pendy-graft et al. (1979). Stumm et al. (1983) argued for a shift in thinking frommeasuring trace concentrations of contaminants in the environment and thewidespread practice of single organism bioassays of toxicity, to a more ra-tional and holistic ecotoxicology approach and risk analysis. After a longcareer in the drinking water field, Abel Wolman (1892-1989) gave a keynoteaddress to the American Water Works Association (AWWA) annual confer-ence of 1976 regarding the past and future of the field (Wolman, 1976). Hislucid assessment of future challenges remains just as relevant four decades81.2. A Recent History of Water Treatmentlater:The water field, as every other, will be the beneficiary of advanc-ing scientific and technologic discovery. It will also be plaguedby real and pseudo dangers. Managers are slowly becoming ac-customed to the fact that water never was simply the H2O ofthe laboratory, but is the receptacle into which all of the ingre-dients of nature and man has been poured. With ever increasingtools of detection, infinitesimal concentrations of everything willbe recorded. Some of these will be hazardous in truth, othersuseful for television media, and still others grist for the courtsand lawyers. Instead of succumbing to despair, the worker in thefield will have to maintain an equilibrium between the real, thehypothetical and even the hysterical.Wolman continues, urging rationalism and patience in the face of new chal-lenges:Water, as a strong determinant of the health of its users, willbe subjected to increasing scrutiny, simply because it is a uni-versal ingredient in man’s metabolism. The future will focusinevitably upon the infinitesimal concentrations of organic andinorganic materials, if for no other reason than that we can nowdetect them. In the past, our tools of detection were gross. Now,the time frame for possible effects is considered to be decadesrather than days, insofar as carcinogenic, mutagenic and geneticpotentials are concerned. The economic implications are many.Common sense, supported by epidemiologic, rigid scrutiny, willultimately prevail, if we can remain patient.In the case of by-products of chlorination, four decades of research, spanningmore than 60 epidemiological studies, have failed to demonstrate harmful ef-fects at typical exposures (Hrudey and Charrois, 2012). Research continues91.3. UV Disinfectionto discover other halogenated DBPs, with hundreds of new compounds hav-ing been found at increasingly smaller concentrations in chlorinated drinkingwater (Hrudey and Charrois, 2012). Regarding trace organic contaminants,the absence of health significance aside, the changing nature of industrialmanufacturing in North America have made the long list of contaminantsdetected in the 1970s largely an anachronism today, as the majority of thosesubstances are no longer released to North American waters (Tchobanoglouset al., 2015). The development of liquid chromatography coupled with massspectrometry (LC-MS), has indeed revealed the presence of a myriad ofnewly discovered contaminants in water at the part-per-trillion level or less,namely pharmaceuticals and their metabolites (Benotti et al., 2009). Thehuman health significance of exposure to these contaminants at such lowlevels is unclear yet has gained media attention. Meanwhile, real threats tohuman health from waterborne outbreaks persist, due not to newly discov-ered contaminants, but to the banality of known hazards, existing treatmentdeficiencies and human error (Hrudey and Hrudey, 2004).1.3 UV DisinfectionAmong the most dramatic changes in drinking water practice in recent yearshas been the rapid adoption of UV disinfection, now a mature and acceptedtechnology. This has enabled the development of other applications of UVto water treatment, namely UV based AOPs. The rise of UV disinfection isbriefly reviewed.By the 1960s, the germicidal effects of ultraviolet radiation at the molecularlevel where well understood (Jagger, 1967), and while a few water treat-ment facilities used UV in the early days of the twentieth century, virtuallyall such facilities seem to have vanished within a few years as the moreconvenient and inexpensive chlorination process spread. When the Ameri-can Society of Civil Engineers (ASCE) and the AWWA published the firstedition of Water Treatment Plant Design (ASCE, 1969), the technological101.3. UV Disinfectionproblem of disinfection was considered essentially solved. The chapter ondisinfection, while briefly acknowledging the existence of UV and other dis-infectants, was entirely devoted to methods of chlorination. To some extentthis was justified, since the causative organisms of diseases such as typhoidand cholera were spread between humans, originated from sewage contam-ination, involved high infective doses, and were very sensitive to chlorine.The second and third editions acknowledged the need for new disinfectants,adding sections on ozone and chlorine dioxide (ASCE, 1990; AWWA, 1998).By the end of the 1990s, due to a growing body of UV research, drinkingwater disinfection by UV spread to hundreds of towns and cities in NorthAmerica, Europe and elsewhere. Consequently, by the fifth edition of WaterTreatment Plant Design, a chapter devoted entirely to UV disinfection wasdeemed necessary (AWWA, 2012). For a review of UV disinfection practices,the reader is also directed to the AWWA’s Ultraviolet Disinfection Handbook(Bolton and Cotton, 2011).Prior to this rise, UV disinfection of drinking water was an established prac-tice in several European countries by the mid-1980s, yet absent from NorthAmerican utilities. Motivation for use in Europe was driven by a desire foran alternative to chlorine for the disinfection of groundwater and biologicalfiltrates (Kruithof et al., 1992). By the end of the 1990s, the rapid adoptionof UV for drinking water treatment, in North America and elsewhere, canbe understood on the basis of a few now well established facts. Two com-mon pathogens were gradually associated with waterborne outbreaks thatwere not necessarily due to gross contamination of source water, treatmentdeficiencies, or recontamination in distribution. These were identified as theprotozoan parasites Giardia lamblia and Cryptosporidium parvum. Giardiacysts and Cryptosporidium oocysts, the vegetative states of each organism,were found to be ubiquitous in the aquatic environment and common in rawsurface water supplies (LeChevallier and Di Giovanni, 2002; LeChevallierand Norton, 1995; LeChevallier et al., 1991). Infective doses were deter-mined to be low, on the order of ten (oo)cysts or less (DuPont et al., 1995;Rendtorff, 1954). They were found to be highly resistant to chlorine, partic-111.3. UV Disinfectionularly Cryptosporidium (Hibler, 1987; Korich et al., 1990). Contrary to priorunderstanding, Giardia and Cryptosporidium (oo)cysts are very sensitive toUV radiation when using assays to detect infectivity rather than viability(Bukhari et al., 1999; Campbell et al., 1995; Clancy et al., 1998, 2000; Craiket al., 2000; Linden et al., 2002). Lastly, UV irradiation of drinking water attypical disinfection doses does not produce significant by-products of con-cern nor increase formation of regulated DBPs upon subsequent chlorination(Liu et al., 2002; Malley et al., 1996; Reckhow et al., 2010).Much of the research that motivated the discoveries mentioned above wasprompted by the Cryptosporidium outbreak that occurred in Milwaukee WIin the spring of 1993. The largest known outbreak of waterborne illness inUS history involved an estimated 400,000 individuals acquiring the infection(MacKenzie et al., 1994) and resulting in an estimated 100 deaths (Hrudeyand Hrudey, 2004). The cause of the Milwaukee outbreak, though neverfully determined, was associated with treatment plant deficiencies and chal-lenging conditions (Fox and Lytle, 1996). Yet, Cryptosporidium outbreaksin Las Vegas NV (Roefer et al., 1996) and Waterloo ON (Pett et al., 1993)occurred during normal operation in the absence of any known deficienciesor challenging source water conditions, suggesting utilities were more vul-nerable than previously thought. The magnitude of the Cryptosporidiumthreat revealed by the Milwaukee outbreak specifically lead to the USEPApromulgation of the Long-Term 2 Enhanced Surface Water Treatment Rule(LT2ESWTR) and the rapid adoption of UV disinfection in the USA andelsewhere. Large-scale early adopters of UV disinfection include PittsburgPA (2001), Edmonton AB (2002), and Seattle WA (2003) (Hargy, 2002). By2010, a North American survey revealed that 161 utilities in Canada and 148utilities in the USA had installed or were planning to instal UV disinfectionat treatment capacities greater than 2 ML d−1 (Wright et al., 2012). Severalsuch facilities have installed UV treatment with the main objective of dis-infection, yet with the capability of periodic treatment of taste and odourwhen operated in “advanced oxidation” mode. Such plants include those ofthe City of Cornwall and the Region of Peel, both in the Canadian province121.4. Advanced Oxidation Processesof Ontario. The Andijk treatment plant in Holland uses UV for simulta-neous disinfection and destruction of agricultural and industrial pollutantsoriginating from the Rhine River. In Orange County California, UV is usedas a final polishing step in the treatment of wastewater for indirect potablereuse, both as a disinfection barrier and for the destruction of contaminantspoorly retained by reverse osmosis membranes.1.4 Advanced Oxidation ProcessesDuring the 1970s and 1980s, as the inventory of trace organic contaminantsdetectable in water expanded, new treatment technologies were investigatedthat could reduce their concentration. One approach for the removal oftrace organic contaminants is oxidative degradation, with the most promis-ing techniques involving ozone or ultraviolet radiation. Interest in the latterhas recently increased, in part due to the maturity of UV technology devel-oped for disinfection.Both ozone decomposition promoted by H2O2 (O3/H2O2) and UV photoly-sis of H2O2 (UV/H2O2) produce the hydroxyl radical ( OH). Reaction rateconstants of OH with most organic solutes are high, spanning roughly threeorders of magnitude (108 − 1010 M−1 s−1) near the diffusion limit (Buxtonet al., 1988; Haag and Yao, 1992). Elevated rate constants over a narrowrange imply that OH is a highly reactive and nonselective oxidant. Be-cause of these properties, treatment processes based on OH generation arereferred to as Advanced Oxidation Processes (AOPs), a term first used byGlaze et al. (1987). While a variety of other AOPs have been proposed andstudied, only those based on O3 and UV have thus far shown promise infull-scale drinking water applications.Unlike OH, ozone is a very selective oxidant, in that its reaction rate con-stants with aqueous solutes span over ten orders of magnitude. For example,ozone is highly reactive with sulphide (3× 109 M−1 s−1) though practically131.4. Advanced Oxidation Processesinert to tetrachloroethene (< 0.1 M−1 s−1) (Von Gunten, 2003). Decompo-sition of ozone is initiated by OH– and is thus accelerated at elevated pH(Forni et al., 1982). This decomposition generates OH through a chain re-action mechanism with an experimentally observed yield of 0.5 mole OHper mole of O3 (Forni et al., 1982; Staehelin and Hoigne´, 1982). However,HCO –3 and CO2–3 , omnipresent in natural waters, react with OH to producethe much less reactive carbonate radical (CO –3 ), interrupting the chain reac-tion. Furthermore, since the OH rate constant of CO 2–3 (4× 108 M−1 s−1) isapproximately 50 times greater than that of HCO –3 (8.5× 106 M−1 s−1), thescavenging of OH increases significantly above pH 9 (Buxton et al., 1988).In waters of typical pH (6-9) and moderate to high alkalinity (> 2 mM),ozone decomposition is inhibited. It is the relative stability of ozone in suchwaters that provides the required ozone lifetime for effective disinfection.Ozone decomposition, however, occurs rapidly over the usual pH range inthe presence of a small amount of H2O2 (Staehelin and Hoigne´, 1982). Atypical H2O2:O3 (mol/mol) dose ratio of 0.1-0.5 is sufficient to initiate rapidconversion to OH, with the optimum site-specific ratio determined by ex-periment (Acero and Von Gunten, 2001). The O3/H2O2 AOP is a simpleaugmentation in a plant that already uses ozone, with the point of H2O2addition delayed sufficiently so as not to interfere with ozone disinfection.Ozone is commonly followed by biological filtration, which further enhancesremoval of the more biodegradable AOP products and provides quenching ofresidual H2O2. Where ozone is not currently used, considerable infrastruc-ture must be installed for generation, use, and disposal of ozone (Langlaiset al., 1991; Rakness, 2011). A problem with the O3/H2O2 AOP, as with O3disinfection, is the generation of the by-product bromate (BrO3– ) in bromidecontaining waters (Haag and Hoigne´, 1983). A current limit on bromate indrinking water of 10 µg L−1 is suggested or required in North America andEurope (Kristiana et al., 2012). This value is partly based on detectionlimits of existing analytical methods. Methods to avoid bromate formationinvolve additional process complexity and cost (Von Gunten, 2003). Futureimplementation of the O3/H2O2 AOP depends largely on regulatory limitsfor bromate in treated waters.141.4. Advanced Oxidation ProcessesThe UV/H2O2 AOP generates OH by the photolysis of H2O2 that has beenadded upstream of UV reactors. Photolysis of H2O2 to OH involves a quan-tum yield of unity at 254 nm(Baxendale and Wilson, 1957). However, H2O2is a poor absorber of UV with a molar absorption coefficient at 254 nm ofonly 20 M−1 cm−1. Thus higher concentrations of H2O2 must be applied forsufficient photolysis to take place, with typical doses of H2O2 in the rangeof 5 to 15 mg L−1. Also, the dose or fluence of UV energy required is high,approximately 1000 mJ cm−2 or more, compared to the 40 mJ cm−2 typi-cally applied in UV disinfection (Bolton and Cotton, 2011; Dotson et al.,2012). Another significant challenge is the undesirability of residual H2O2in treated effluent and its quenching prior to distribution. Negligible re-duction in H2O2 occurs during passage through UV reactors, thus virtuallythe full amount dosed must be quenched. Benefits of UV include simultane-ous disinfection, absence of by-products of concern, and a compact footprint.The choice between implementing the O3/H2O2 or UV/H2O2 AOP will de-pend on many site-specific conditions. The electrical energy per order orEEO concept (Bolton et al., 1996) has been used to compare the two AOPsin general, with the H2O2 converted to an energy equivalent based on chem-ical and electrical costs (Rosenfeldt et al., 2006). Such a comparison hassuggested that the O3/H2O2 AOP involves a lower operating cost. How-ever, the optimal choice in general depends on such site-specific factors asthe existing treatment process and layout (if any), space availability, waterchemistry, treatment objectives, and electrical, chemical, and capital costs.Both selection of an AOP and its placement in an overall treatment systemmust consider a few features common to both processes.Due to the relatively nonselective reactivity of OH, the oxidation of traceorganic contaminants by AOP generally results in a mixture of daughterproducts, the complete mineralization of contaminants to CO2 requiring animpractical amount of energy. Even if the concentration of the target con-taminant is reduced to below detection, the merit of treatment by AOP151.4. Advanced Oxidation Processesdepends on the desirability of the mixture produced. For example, treat-ment by AOP of the herbicide atrazine creates a mixture of dozens of triazinedaughter products (Acero et al., 2000). The health relevance of human ex-posure to such trace mixtures, relative to that of the parent compound, isgenerally unknown. If the daughter products of such a mixture are morebiodegradable than the parent compound, subsequent biological filtrationmay be suggested. However, the selection of an AOP should be made onlyafter a comparison with alternative treatment processes.UV/H2O2 is the more appropriate AOP in cases where avoiding bromateformation is essential, where the direct photolysis of the target contaminantis significant, and where continuous terminal UV disinfection may be brieflyaugmented to an AOP. This last case has often been motivated by the sea-sonal occurrence of taste and odour events due to the naturally occurringcompounds geosmin and 2-methylisoborneol (MIB).Both geosmin and MIB are produced by cyanobacteria and actinomycetes,causing taste and odours detectable in water by most humans at concentra-tions on the order of 10 ng L−1 or less (Persson, 1980; Young et al., 1996).Geosmin is described as “earthy”, and MIB as “musty”, and together theyhave been a major cause of consumer complaints (McGuire, 1995). Their de-tection often causes public concern regarding the safety of the water supplydespite posing no health risk (Dionigi et al., 1993). Occurrence is typicallyunpredictable and seasonal, and conventional treatment ineffective (Suffet,1995). Both molecules are tertiary aliphatic alcohols. Consequently, theyare poorly sorbed to activated carbon (Chowdhury et al., 2013). They havebeen found to be relatively inert to the conventional oxidants Cl2, ClO2,KMnO4, H2O2, and O3, yet their OH reaction rates are reported to be be-tween 109 − 1010 M−1 s−1 (Glaze et al., 1990; Lalezary et al., 1986; Peterand Von Gunten, 2007). Though the identity of the AOP by-products ofgeosmin and MIB have not been reported, the success of treatment suggeststhat none have a lower odour threshold than the parent compounds. Onepossible daughter product of MIB is camphor, which has an odour thresh-161.4. Advanced Oxidation Processesold 6 orders of magnitude greater than MIB (Amoore and Hautala, 1983).Episodes of geosmin and MIB have typically involved concentrations in therange of 50 to 100 ng L−1, with extreme cases as high as 500 ng L−1 or more(Suffet, 1995). The increased occurrence of eutrophication associated withpopulation growth and climate change may increase the frequency of prob-lems related to geosmin and MIB.Redundancy is a common engineering practice in water treatment, includingUV disinfection. Redundant UV capacity may be fully employed to providethe required UV fluence for AOP treatment when geosmin or MIB are de-tected. The usual disinfection dose of about 40 mJ cm−2 may be instantlyincreased many fold by maximizing lamp outputs, the ignition of additionallamps, and possibly a reduction in flow. This can be done with a relativelysmall footprint, particularly when medium pressure mercury lamps are used.This approach requires the continuous on-site storage of H2O2, often in a50% w/w solution, an additional plant hazard. Lastly, the requirement toquench residual H2O2 presents a major cost to the process.Many North American treatment plants add chlorine, as a secondary disin-fectant, upstream of UV reactors (Dotson et al., 2012). It was discoveredthat this practice resulted in the UV photolysis of HOCl and OCl– with thegeneration of OH (Feng et al., 2007; Watts and Linden, 2007). This UV/Cl2AOP eliminates the need for H2O2 and its quenching, though may requirehigher chlorine doses than used in disinfection. However, this process is pHdependent, due to differences in the relevant properties of the two speciesHOCl and OCl–. Molar absorption coefficients, photolysis quantum yields,and OH rate constants of HOCl and OCl– are such that UV/Cl2 is much lessefficient than UV/H2O2 when treating water above pH 6 (Jin et al., 2011;Wang et al., 2012; Watts et al., 2007). Requirement for pH adjustmentsbefore and after the UV/Cl2, as well as additional dechlorination to trimthe final residual, would negate the benefits of avoiding H2O2 addition.Both ozone and UV based AOPs involve the added complexity of chemical171.5. 185 nm Advanced Oxidationaddition and control. A new AOP approach that is able to generate OHfrom water itself would eliminate such complexity and is the subject of thisstudy.1.5 185 nm Advanced OxidationIn addition to the germicidal radiation emitted at 254 nm, the conventionallow pressure mercury lamp also emits at a wavelength of 185 nm. The 185 nmradiation is capable of generating OH from the photolysis of water. Thisfact forms the basis of using conventional UV lamps, in a suitably designedreactor, as an AOP that does not require chemical addition.Low pressure mercury lamps emit virtually all radiation at two wavelengths,254 and 185 nm. The atomic energy levels of mercury are such that the emit-tance at 254 nm is always greater than that at 185 nm, with a 185 nm to254 nm ratio reported between 0.12 to 0.34 (Barnes, 1960; Johnson, 1971).Increased lamp temperatures and arc currents have been found necessaryfor higher ratios of 185 nm. More recently this ratio has been reported tobe 0.08 (Masschelein and Rice, 2002), without any indication of how it wasmeasured.An envelope encloses the lamp to contain the mercury vapour within whichradiation is generated, while a sleeve with an air gap surrounds the lampand provides isolation from water in a reactor. High purity silica (SiO2) isthe material of choice for lamp envelopes and sleeves. Two general typesare fused quartz, obtained from SiO2 containing minerals, and fused silica,produced from the particulates of high purity SiCl4 vapour combustion withO2 (Koller, 1965; Phillips, 1983). Optical transmission of both materials ishigh at 254 nm, while fused silica is superior at 185 nm, making fused silicathe optimal material for lamp envelopes and sleeves in a 185 nm AOP reac-tor.181.5. 185 nm Advanced OxidationAt the scale of optical path lengths typical of UV reactors, the absorbanceby water at 254 nm is negligible (< 0.01 cm−1) (Quickenden and Irvin, 1980).At 185 nm, absorbance is significant, often cited as αw = 1.8± 0.1 cm−1 at25 ◦C (Weeks et al., 1963). A temperature coefficient for αw was found byWeeks et al. (1963) to be 0.05 cm−1 ◦C−1 over the range 20 to 35 ◦C, in agree-ment with that found between 25 to 50 ◦C by Barrett and Mansell (1960).More recent work has reported αw = 1.60±0.03 cm−1 at 25 ◦C (Kro¨ckel andSchmidt, 2014). Furthermore, the measurements of αw(λ) by Kro¨ckel andSchmidt (2014) over the range of 187 to 181 nm reveals a strong increase inabsorbance with decreasing wavelength, confirming the substantial error inαw at 185 nm due to small wavelength errors.Photons of 185 nm have sufficient energy (6.7 eV) to photolyze water to Hand OH with a quantum yield of approximately 0.3 (Getoff and Schenck,1968). In the presence of dissolved oxygen, the H formed will react with O2at a diffusion limited rate to form the perhydroxyl radical (HO2 ). At typicalwater pH of 6 - 9, HO2 disproportionates to the superoxide radical anionO –2 with pKa = 4.8 (Bielski et al., 1985). Both HO2 and O–2 will reactwith OH at diffusion limited rates, but have low reactivities with organicsolutes. The net result is to provide a source of OH. In pure water, OHrecombination results in the accumulation of H2O2. However, the presenceof major solutes will greatly reduce the likelihood of such a reaction andH2O2 accumulation will be negligible. A summary of key reactions withquantum yield and rate constants (Buxton et al., 1988) resulting in net OHproduction is as follows:H2O185 nm−−−−→ H + OH Φ = 0.3 (1.1)H + O2 −→ HO2 k = 2× 1010 M−1 s−1 (1.2)HO2 −−⇀↽− O –2 + H+ pKa = 4.8 (1.3)O –2 + OH −→ O2 + OH– k = 7× 109 M−1 s−1 (1.4)HO2 + OH −→ H2O + O2 k = 6× 109 M−1 s−1 (1.5)191.5. 185 nm Advanced OxidationBecause of the high absorption coefficient of water, 90% of 185 nm photonsare absorbed within a water layer approximately 5 mm thick. The lifetimeof generated OH is on the order of 1 µs in typical water matrices (Hoigne´,1997), thus confining oxidative degradation to the 185 nm irradiated vol-ume. Achieving sufficient irradiation by 185 nm of the entire volume of flowimposes severe constraints on reactor geometry and mixing. The overall ef-ficiency of the 185 nm AOP therefor depends critically on reactor design.While direct photolysis at 254 nm is significant for some organic molecules,it is likely less so at 185 nm. Molar absorption coefficients of molecules insolution rarely exceed 105 M−1 cm−1 (Wayne, 1988). Trace organic contam-inant concentrations in surface waters are usually much less than 1 µg L−1.For a molecular weight of 100 g mol−1, an upper value for the fraction ofphotons absorbed by the compound in pure water will be on the order of10−3, and much less than this in the presence of major solutes that alsoabsorb. The quantum yield (Φ) of direct photolysis in solution is typicallymuch less than unity due to the influence of the solvent (Wayne, 1988).In a typical drinking water matrix, the contribution of direct photolysis at185 nm to contaminant degradation is thus expected to be negligible relativeto OH oxidation.A recent review of 185 nm water treatment (Zoschke et al., 2014) includesa discussion of research on photochemical reactions of some solutes in the185 nm regime. Based on the work cited therein, several comments can bemade. Removal of NOM, as measured by DOC, using 185 nm radiation isunfeasible based on energy requirements. In general, the biodegradabilityof NOM will be increased by 185 nm irradiation, to an extent dependent ontotal exposure. Nitrate may be converted to nitrite under certain conditions,but the chemistry involved is complex. No other degradation pathways fortrace organic contaminants have been considered other than OH oxidation.Fundamental mechanistic studies are lacking. Most research has involvedqualitative studies or produced quantitative results specific to a particularapparatus and water matrix, often insufficiently described. Few attempts201.6. Knowledge Gapshave been made to measure the extent of reaction, not by time, but on thebasis of photons or energy absorbed by the system. Comparison of resultsfrom different equipment and scales is therefore difficult. A recommendationmade in a review of photochemical water treatment more than two decadesearlier, that researchers adopt an energy based approach to describe theextent of treatment, remains relevant (Legrini et al., 1993). This deficiencymay in part be due to the absence of convenient 185 nm detectors, either re-liable electronic radiometers or chemical actinometers. Unlike the KI-KIO3actinometer for quantification of radiation at 254 nm (Bolton et al., 2011;Rahn, 1997; Rahn et al., 2003), extant actinometric methods at 185 nm aretime consuming, require gas-tight assemblies and sophisticated techniquesto achieve sufficient precision and accuracy (see Kuhn et al. (2004) and ref-erences therein).1.6 Knowledge GapsMany details essential for successful application of the 185 nm AOP are notcurrently available, imprecise, or in dispute. The process relies on the gen-eration of 185 nm radiation, yet the efficiency of this process has not beenreported in the open literature for modern lamps. The absorbance value ofwater below 20 ◦C is not known, nor is any information available regardingthe temperature dependence of the quantum yield for the photolysis of water.Studies of the influence of the major inorganic solutes on the 185 nm AOPare absent. Several studies have investigated nitrate, motivated by con-cern for nitrite generation. However, nitrate is not a major ion in surfacewaters, often present at concentrations less than 1 mg L−1. Chloride, sul-phate, bicarbonate and carbonate are present in some waters from 10 to100 mg L−1(Wetzel, 2001) or more. These solutes are known to be relativelystrong absorbers of 185 nm radiation (Barrett et al., 1965; Fox, 1970; Hayonand McGarvey, 1967; Jortner et al., 1964; Weeks et al., 1963). Upon suchabsorption, an excited state is created in which an electron is shared be-211.6. Knowledge Gapstween the anion and several surrounding water molecules in what is knownas a charge transfer to solvent (CTTS) excited state (Blandamer and Fox,1970). A fraction of these electrons dissociate and escape the solvent cageto leave behind a radical:Cl–185 nm−−−−→ Cl + e –aq (1.6)SO 2–4185 nm−−−−→ SO –4 + e –aq (1.7)The solvated electron (e –aq) produced is efficiently scavenged in the presenceof dissolved oxygen to produce the superoxide anion radical which may thenproceeds as in reaction 1.4 above (Buxton et al., 1988):e –aq + O2 −→ O –2 k = 2× 1010 M−1 s−1 (1.8)Photochemical generation of Cl and SO –4 depends on the molar absorptioncoefficients () and quantum yields (Φ) at 185 nm of their respective anions(see Table 1.1).Table 1.1: Molar absorption coefficient () and quantum yield (Φ) for Cl– andSO 2–4 at 185 nm(M−1 cm−1) Φa Ref.Cl– SO 2–4 Cl– SO 2–43800± 300 260± 30 0.43± 0.02 0.64 Dainton and Fowles (1965)ca. 3500 190 Weeks et al. (1963)200 0.67a Barrett et al. (1965)0.49 Jortner et al. (1964)a Corrected by Fox (1970) using Φ(H2) = 0.4 for 5 M ethanol actinometerThe value of 185(CO2–3 ) is reported to be approximately 103 M−1 cm−1(Hayon and McGarvey, 1967; Weeks et al., 1963). No information is found221.6. Knowledge Gapsregarding 185(HCO–3 ), nor the quantum yield of HCO–3 /CO2–3 (reaction1.9).HCO3 /CO2–3185 nm−−−−→ CO –3 + e –aq (1.9)Only CO –3 is expected under the usual pH conditions, since pKa < 0 forHCO3 /CO–3 (Czapski et al., 1999). If it were to be found that the 185 nmabsorbances of HCO –3 and CO2–3 are similar, then at a typical total carbon-ate concentration of 1 mM, the contribution to absorption by the carbonatesystem would be comparable to that of water itself even at pH < 8, sincepKa = 10.3 for HCO3 /CO2–3 (Butler, 1982). The reactivity of CO–3 isrelatively low, with a few exceptions such as aniline derivatives (Chen et al.,1975; Larson and Zepp, 1988). Hence, HCO3 /CO2–3 is expected to be ascavenger of both OH and 185 nm photons, exhibiting a parasitic effect onprocess efficiency by two distinct phenomena.The second-order reaction rate constants of the three species OH ,Cl , andSO –4 with a few organic and inorganic solutes have been reported, allowinga comparison of their values. In many cases, the reactivity of OH and Clare similar (Buxton et al., 2000), both with 6× 108 M−1 s−1 for 2-methyl-2-propanol (tert-butanol). Available data for SO –4 show a wider range ofsecond order reaction rate constants for substituted benzenes than for OH,suggesting an electron-transfer mechanism (Neta et al., 1977). For smallaliphatic alcohols, carboxylic acids, aldehydes and ketones, SO –4 rate con-stants are typically two orders of magnitude less than those of OH and Cl(Buxton et al., 2000).As mentioned above, the carbonate system is known to be a significant scav-enger of OH. Reaction rate constants with Cl and SO –4 reported in theliterature are displayed in Table 1.2.Unlike OH, no information can be found on the reactivities of Cl and SO –4231.6. Knowledge GapsTable 1.2: Comparison of rate constants for OH, Cl , and SO –4 withHCO –3 and CO2–3k(M−1 s−1) Ref.Radical HCO –3 CO2–3OH 8.5× 106 3.9× 108 Buxton et al. (1988)Cl 2.4× 109 - Buxton et al. (2000)2.2× 108 5.0× 108 Mertens and von Sonntag (1995)SO –4 3.5× 106 - Buxton et al. (2000)9.1× 106 - Ross and Neta (1979)with NOM nor any of its fractions, such as humic and fulvic acids. Further-more, no information is available regarding the contribution by NOM to theabsorption of 185 nm photons. Strong absorbance on a molecular basis isexpected due to the substantial aromaticity of humic and fulvic acids. Yetdue to the colloidal nature of such macromolecules, often with molecularweights of several thousand dalton, the relative contribution of absorptionat 185 nm is likely to be highly dependent on molecular size distribution,pH-dependent spacial conformation, and complexation with other solutes.The photo-generated Cl reacts with Cl– to form Cl –2 , the dichloride radicalanion. This reaction and the reverse decay of Cl –2 result in the equilibriumCl + Cl– ⇀↽ Cl –2 which lies far to the right (Buxton et al., 1998). Both Cland Cl –2 react with water via several postulated equilibria terminating information of OH. At neutral pH with [Cl–] < 0.1 M, an overall equilibriumis reported to lie far to the right (Buxton et al., 1998):Cl /Cl –2 + H2O −−⇀↽− H+ + HOCl – −−⇀↽− Cl– + OH + H+ (1.10)In the presence of solutes reacting with OH at diffusion limited rates, the241.6. Knowledge Gapsabove equilibrium will not be established and the reactions with water will befirst order (Buxton et al., 1998). The resulting Cl –2 exhibits low reactivitywith organic solutes (Hasegawa and Neta, 1978). Photogenerated SO –4 reacts with Cl–, producing Cl (Gilbert et al., 1988). The reverse reactionof Cl and SO 2–4 is competitive resulting in an equilibrium with a constantnear unity (Buxton et al., 1999). The reaction of SO –4 with water mayalso leads to formation of OH. What information exists regarding thesereactions is available from fundamental research in radiation chemistry andmore recent investigations relating to the atmospheric chemistry of clouddroplets. Important reactions and their reported rate constants are listed inTable 1.3. Reactions with tert-butanol are included for comparison.Table 1.3: Important reactions involving Cl , Cl –2 ,and SO –4Reaction Rate ConstantCl + Cl– −→ Cl –2 8.5× 109 M−1 s−1Cl –2 −→ Cl + Cl– 6.0× 104 s−1Cl + H2O −→ 2.5× 105 s−1Cl –2 + H2O −→ 1.3× 103 s−1Cl + SO 2–4 −→ Cl– + SO –4 2.1× 108 M−1 s−1SO –4 + Cl– −→ SO 2–4 + Cl 6.1× 108 M−1 s−1SO –4 + H2O −→ ∼ 700 s−1Cl + t BuOH −→ 6.2× 108 M−1 s−1Cl –2 + t BuOH −→ < 103 M−1 s−1SO –4 + t BuOH −→ 7.8× 105 M−1 s−1OH + t BuOH −→ 6.0× 108 M−1 s−1Rate constants from Buxton et al. (1998, 1999, 2000)It can be seen from the above that the identity and concentrations of in-organic and organic solutes should impart a strong influence on the radicalchemistry of 185 nm irradiated water, and therefore 185 nm AOP efficiency.A fraction of the absorbed 185 nm photons may be lost to those species that251.6. Knowledge Gapsdo not contribute reactive radicals upon absorption. All three radicals ( OH,Cl , and SO –4 ) may be generated at comparable rates depending on the frac-tion of photons absorbed by water and the respective anions. The reactivityof these radicals with scavengers such as NOM and HCO3 /CO2–3 , as wellas reactions among the radicals and anions will determine their steady-stateconcentrations and thus their contribution to the rate of oxidative degrada-tion of specific organic contaminants. The overall rate will also depend onthe rate constants of a specific contaminant with each radical. For exam-ple, if Cl has a higher rate constant with a target contaminant, but is alsomore reactive with the matrix, a smaller steady-state concentration wouldbe available and the contribution to removal would be smaller. Conversely,if SO –4 had a lower rate constant with the target but was much less reactivewith the matrix, a higher steady-state concentration and thus greater con-tribution to removal could result. In general, the observed degradation ratewill depend on the relative distribution of absorbed photons by componentsof the matrix, and the relative reactivites of the generated radicals to thetarget contaminant and the scavengers. In such a system, a small change inone of the components of the matrix may result in a dramatic change in theobserved degradation rate. The dependence of the 185 nm AOP on thesepotential influences requires confirmation and a minimal set of parametersto characterize the process should be identified.While it is a straight forward matter to measure the molar absorbance co-efficients of compounds above 190 nm, below this wavelength commercialspectrophotometers exhibit substantial error due to stray radiant energy.This stems from the use of a broad spectrum deuterium lamp as the UVsource of such instruments, the amount of power emitted below 190 nmbeing very small relative to longer wavelengths extending to 300 nm. Ifimperfections in grating monochrometers result in even a small amount oflonger wavelength stray radiation entering an aqueous sample, such longerwavelength radiation will emerge from the sample with negligible attenua-tion relative to 185 nm, reaching the detector as a significant contributionto the detected signal. Thus, the effect of stray radiant energy results in261.7. Research Objectivesthe detected absorbance signal appearing to have a lower magnitude thanthe actual value. Furthermore, errors due to finite slit widths compoundthe inaccuracies of measurement when the magnitude of absorption variesstrongly with wavelength. Ideally, dedicated instruments would employ alow pressure mercury lamp, which would provide a powerful source of 185 nmradiation that is easily isolated by a monochrometer. No such instrumentis currently commercially available, and unless methods of compensatingfor or reducing the effect of stray radiant energy are explicitly described,any absorbance measurements reported using commercial instruments aresuspect of substantial error. Alternative methods of measuring absorbancecoefficients are needed. For further discussion of this topic, see Burgess andFrost (1999) and Sommer (1989).Lastly, the by-products of aqueous reactions of Cl and SO –4 are virtuallyunknown. In particular, the possibility of Cl reacting by addition to aro-matic rings, such as those that abound in humic and fulvic acids, shouldbe investigated in order to determine whether the 185 nm AOP generateshalogenated organics in the presence of Cl– and DOM.1.7 Research ObjectivesThe design of a practical 185 nm AOP depends critically on proper hy-draulic mixing, optimal use of radiation, and an accurate reaction model.A reliable model of a reacting system requires the identities of the reactingspecies, the elementary reactions in which they participate, and the kineticrate constants of these reactions. The specific objectives of this work fo-cus on fundamental information required for a useful reaction model of the185 nm AOP. These are:Temperature: Investigation of the impact of water temperature on the185 nm AOP. Given the potential effects of temperature over the typicalrange of operation, and the unknown influence on water absorbance at low271.7. Research Objectivestemperatures and photolysis quantum yield, a quantitative observation ofthese effects are desired to assess their significance.Dissolved Organic Matter: Investigation of the influence of DOM, bothas an absorber of 185 nm and as a scavenger of radicals formed. The DOMto be studied includes well characterized NOM reference materials from theInternational Humic Substances Society (IHSS), as well as pure organic sub-stances used as model compounds. Interactions of DOM and Cl–, SO 2–4 , andHCO –3 on the kinetics are included.Inorganic Anions: Evidence of the role of Cl and SO –4 as reactive speciesand significant contributors to oxidative degradation is gathered. The dualinfluence of HCO –3 as an absorber of photons and a radical scavenger isstudied. Interactions between Cl–, SO 2–4 , and HCO–3 are investigated.Molar Absorption Coefficients at 185 nm: Alternative methods ofquantifying the molar absorption coefficients (185) of major solutes are re-quired that do not suffer from errors associated with stray radiant energyor finite monochrometer slit widths. An alternative indirect method is in-vestigated.In order to achieve these goals, a suitable bench-scale apparatus is developedthat will allow photochemical kinetic experiments to be conducted quanti-tatively. An important element of this approach is the identification andcharacterization of appropriate probe compounds and the ability to elimi-nate, compensate or correct for the presence of 254 nm radiation.28Chapter 2Experimental Approach2.1 General ApproachIn order to study the influences of temperature and major solutes systemati-cally, it is expedient to do so at bench-scale with as many variables controlledas possible. This is facilitated by using a quasi-collimated beam apparatusfor the continuous irradiation of a well mixed sample of precisely knowncomposition. The influence of the variables of interest on the kinetics of thesystem may then be measured using a suitably selected probe compound.The precise composition of the irradiated systems is ensured by assemblyof the solutions using ultrapure water and analytical grade reagents, withverification by instrumental analysis. Lastly, before the experiments of pri-mary interest are conducted, key assumptions and experimental conditionsshould be tested. The theoretical approach, experimental techniques, andpreliminary tests are reviewed in this chapter.2.2 The Probe Compound and the Observable k′A probe compound is a compound added to a water matrix being studiedprior to irradiation and in a sufficiently small quantity so as to have neg-ligible impact on the relevant properties of the solution. Upon irradiation,photochemically induced degradation of the probe occurs. Samples of thesolution, after various amounts of irradiation, are obtained and analyzed forthe remaining probe concentration.Under 185 nm irradition, the apparent reduction in probe concentration may292.2. The Probe Compound and the Observable k′occur by several mechanisms. One mechanism involves probe oxidation byradicals generated in solution subsequent to the absorption of 185 nm pho-tons, primarily OH as a product of water photolysis. A second mechanisminvolves the direct photolysis of probe molecules by absorption of 254 nmphotons, which are simultaneously emitted from the low pressure mercurylamp at a much higher proportion relative to 185 nm. Other significantmechanisms may include volatilization and hydrolysis. With regard to di-rect photolysis of trace contaminants at 185 nm, as discussed in Chapter 1,this process is considered negligible.Since, only mechanisms involving 185 nm radiation are of interest, othermechanisms may be avoided by the selection of an appropriate probe com-pound and proper experimental techniques. Criteria for a suitable probeinclude ease of analysis, low limit of detection, sufficient aqueous solubilityand stability, and negligible rate of direct 254 nm photolysis. Probe selectionand characterization is discussed in detail in a subsequent section.Photogenerated OH reacts with the probe compound C, typically produc-ing a mixture of oxidized products:C + OH −→ C ′ox + C ′′ox + C ′′′ox + ... (2.1)The reaction occurs with a second-order rate constant kOH,C , which if un-known may be determined experimentally by competitive kinetics relativeto a reference compound for which the rate constant is known. The degra-dation rate of C is given by the Law of Mass Action as:302.2. The Probe Compound and the Observable k′d[C]dt= −kOH,C [ OH][C] (2.2)Due to the high reactivity of OH with many solutes, it is usually a rea-sonable assumption that once irradiation of the sample has begun, the OHconcentration rapidly increases until the rate of generation equals the rateof consumption by reaction. At this point, a steady-state OH concentrationis achieved, typically very small (< 10−9 M), within a time period very shortrelative to the total time of irradiation (i.e.  1 s). With a constant valueof [ OH]ss in equation 2.2, integration yields the solution:ln([C]t/[C]o) = −kOH,C [ OH]ss t (2.3)where [C]t is the probe concentration remaining after an irradiation time oft, and [C]o is the initial probe concentration. If the steady-state assumptionis indeed valid, then a plot of ln([C]t/[C]o) vs. t will appear first-order andproduce a straight line, with a slope equal to kOH,C [ OH]ss. Equation 2.3may then be simplified to:ln([C]t/[C]o) = −k′ t (2.4)with k′ as the pseudo-first order rate constant with dimensions of reciprocaltime (T−1). If k′ is measured experimentally, and kOH,C is known, then an312.3. Expressing the Extent of Reactionestimate of [ OH]ss may be calculated.The single term k′ contains the dependences of temperature and solutioncomposition. If controlled changes in solution conditions induce an increaseor decrease in k′, such relationships may be investigated quantitatively andfurther insight possibly deduced. The pseudo-first order rate constant k′ isthus the main experimental response used as the observable for this study.2.3 Expressing the Extent of ReactionEquations 2.3 and 2.4 imply that the extent of reaction is measured by theexposure time t, or time of irradiation. However, the extent of photochem-ical reactions are not determined by the exposure time in general, but bythe number or amount of photons absorbed per unit of absorbing materialD. This fact may be considered a consequence of what are often referred toas the first and second laws of photochemistry. Quoting from Calvert andPitts (1966), the first law, formulated by Grotthus and Drapper, states:“Only the light which is absorbed by a molecule can be effective in producingphotochemical change in the molecule. ”The first law implies that before any photochemical reaction can occur, pho-tons must be absorbed.Again, quoting from Calvert and Pitts (1966), the second law, deduced byStark, Einstein, and Bodenstein, states:“The absorption of light by a molecule is a one-quantum process so that thesum of primary process quantum yields Φi must be unity (i.e. ΣΦi = 1). ”The second law applies in the absence of chain reactions, biphotonic pro-cesses, and composite reactions that involve photon absorption in more than322.3. Expressing the Extent of Reactiona single step. During the UV irradiation of aqueous solutions, the lifetimes ofphotogenerated radicals are sufficiently short, and the products of reactionsufficiently nonreactive in general, that no chain reactions are expected tooccur. Furthermore, no evidence of such chain reactions has been reportedin the literature. Biphotonic processes involve the sequential absorption oftwo photons by a single molecule. While this may readily occur in the caseof high intensity radiation provided by a laser, under irradiation via an in-coherent source, such as the low pressure mercury lamp, the population ofmolecules excited by an initial photon is sufficiently small, and their lifetimessufficiently short, that the probability of an excited state molecule absorbinga second photon is assumed to be negligible. Lastly, while the 185 nm baseddegradation of trace organics is certainly a composite reaction, only a singlephoton absorption is assumed involved per molecule degraded.Thus, under continuous incoherent irradiation, the extent of reaction maybe determined if both the exposure time, t, and the rate of photon absorp-tion per unit of absorbing material, Ia, are known. The product given by:D = Ia t (2.5)provides the number or amount of photons absorbed per unit of absorbingmaterial upon a given exposure time. For a given solution, any combinationof Ia and t that produces a given value of D will yield the identical extent ofreaction. The comparison of two photochemical reaction systems, therefore,should be based on D, and not on t alone, since in general Ia will differ.Furthermore, it is the cumulative value of D that dictates reaction extent,regardless of whether irradiation is continuous or intermittent. This detailis important for nearly opaque fluids such as water at 185 nm. Adoptinga Lagrangian description of a fluid, during mixing the microscopic volume332.3. Expressing the Extent of Reactionelements of fluid will absorb photons primarily while passing near irradiatedinterfaces but not while further away. Since, in such a volume element, Iavaries with time, D for a given element will be obtained by an integral,D =∫Ia(t)dt, of the element’s Ia(t) history.For highly absorbing fluids, the local value of Ia at each spatial point inthe fluid will drastically decrease with distance from the irradiated inter-face. This is the case for water under 185 nm and optical path lengths onthe order of a few centimetres. The spatial profile of Ia may thus involve avery high value near the interface, attenuating to vanishingly small valuesfurther inside the fluid. The value of Ia accessible to measurement is theaverage over the irradiated volume. It is shown by Calvert and Pitts (1966)that if a filled sample cell is uniformly irradiated by a parallel beam, andif the diffusion (or lifetime) of radicals can be neglected, the measured andlocal rates will be identical if Ia appears in the rate law to the first power(i.e. d[C]/dt ∝ Ia n, with n = 1). The first-order assumption of Ia in therate expressions of the 185 nm AOP is based on the assumptions that thecomposite reactions involve only a single photon absorbing step, and thatchain reactions and biphotonic processes are absent. Considering equation2.2, this is equivalent to assuming [ OH]ss ∝ Ia.Equation 2.4 can be rewritten in terms of D rather than t:ln([C]D/[C]o) = −k′′ D (2.6)If D is expressed in units of mol L−1, then k′′ must have the reciprocal unitsof L mol−1. Alternatively, D and k′′ may be expressed in energy units usingthe molar photon energy of 6.47× 105 J mol−1 at 185 nm. Then Ia may beexpressed in units of mol L−1 s−1 (or J m−3 s−1). Furthermore, k′ and k′′ are342.3. Expressing the Extent of Reactionsimply related by k′ = k′′Ia, and thus interconversion between equations 2.4and 2.6 requires only that Ia be known.If an irradiated fluid is well mixed, the absorbed photons are evenly dis-tributed throughout the entire fluid volume. If the fluid is nearly opaque(i.e. > 99% photon absorption), the rate of photon absorption per unit vol-ume Ia may be calculated from:Ia =Io SV(2.7)where Io is the incident photon fluence rate (mol m−2 s−1), S is the surfacearea through which the incident radiation enters the fluid (m2), and V is thevolume of absorbing solution under irradiation (m3). Again, using the mo-lar photon energy at 185 nm, the fluence rate may be expressed with unitsof J m−2 s−1. If the incident irradiation is nearly collimated, the incidentfluence rate is approximately equal to the incident irradiance. The incidentphoton fluence, F (mol m−2), is obtained from the product Io t. Alternativeexpressions for D are thus given by:D =IoStV=F`(2.8)where ` is the optical path length or depth of the uniformly irradiated solu-tion. Using chemical actinometry or a calibrated electronic radiometer, Iomay be measured.Note that to make absolute kinetic measurements and quantitative compar-352.4. A 185 nm Kinetic Modelisons between different UV reaction systems, D should be used and the valueof Io, or its equivalent, must be known. If making comparative kinetic mea-surements of different opaque solutions using the identical reaction system,then t may be used, as Io will be identical between cases and Ia approxi-mately so.2.4 A 185 nm Kinetic ModelFor comparative investigations of the temperature and major solute effectson 185 nm kinetics, an equation for the observable, k′, in terms of parame-ters that are either known or can be measured is useful. A model is proposedbased on the relation of [ OH]ss to such parameters, and is obtained by in-voking the steady-state assumption. Implicit in the steady-state assumptionis that the rate of OH generation equals the rate of its consumption. Therate of OH generation by photolysis of water is given by:d[ OH]dt= ΦH2O fH2O Ia (2.9)where ΦH2O is the quantum yield for water photolysis at 185 nm, approx-imately equal to 0.3 (equation 1.1), and fH2O is the fraction of photonsabsorbed by water.In a solution composed of multiple solute species S1, S2, ..., each species Sireacts with OH with a second-order rate constant of kOH,Si . The rate ofOH consumption is given by the Law of Mass Action and may be writtenas a summation of all reactions OH + Si:362.4. A 185 nm Kinetic Modeld[ OH]dt=∑ikOH,Si [ OH]ss [Si] (2.10)Under steady-state conditions d[ OH]/dt = 0, and by equating equations 2.9and 2.10, the expression for [ OH]ss may be written as:[ OH]ss =ΦH2O fH2O Ia∑kOH,Si [Si](2.11)As discussed above, Ia on the right-hand side of equation 2.10 is the averagevalue over the volume. Consequently, the value of [ OH]ss on the left-handside is also the volume averaged value. Yet, it is acknowledged that, likeIa, the local value of [ OH]ss is greatest at the irradiated interface and at-tenuates rapidly deeper into the fluid. As the lifetime of OH is sufficientlyshort, on the order of microseconds, diffusion is negligible and each micro-scopic volume element will receive varying amounts of OH as it travels,depending on the local value of Ia along its trajectory.The combinations of equations 2.3, 2.4, and 2.11 yields the following:k′ =kOH,C ΦH2O fH2O Ia∑kOH,Si [Si](2.12)The right-hand side of equation 2.12 is the observable obtained from exper-iment, calculated from the diminishing concentration of C with increasing372.4. A 185 nm Kinetic Modelt. Note that if the initial concentration of the probe compound [C]o is suf-ficiently small, then it will contribute negligibly to the sum∑kOH,Si [Si].The value of [ OH]ss (equation 2.11) and k′ (equation 2.12) will therefore beindependent of the probe compound concentration [C]o. This condition issatisfied when kOH,C [C]o ∑kOH,Si [Si]. A value for [C]o has been selectedthroughout this work such that kOH,C [C]o is at least ten times less than thevalue of∑kOH,Si [Si] for all other scavengers.If comparative studies are conducted, say between two temperatures or twoconcentrations of a particular solute Si, the relative values of the resultingk′1 and k′2 may be related using 2.12 without the knowledge of Ia, providedit is common to both cases. However, if a solution is obtained for whichall the terms of the left-hand side are known, except Ia, then Ia may bedetermined once k′ is measured. Subsequently, Ia may be used to calculateIo, or convert the abscissa coordinate of kinetic plots from a t-scale to aD-scale. By this approach, a method of measuring Io, as an alternative tochemical actinometry and electronic radiometry, is made available.The discussion thus far is predicated on the assumption that water is anopaque solution at 185 nm, and this is approximately true for optical pathlengths of 1 cm or more. However, for shorter depths water may be consid-ered either semitransparent or transparent, and the calculation of absorbedenergy must be modified accordingly. For such cases, a useful discussion isprovided by Harm (1980).In the case that 185 nm photons are absorbed by other solutes, and generateother radicals able to react with the probe compound C, the observed rateconstant k′ may be expanded to include additional contributions. Basedon the discussion from Chapter 1, it is known a priori that in addition toOH, the radicals Cl and SO –4 may also be significant oxidants of C. Theexpression for k′ may thus be expanded to:382.4. A 185 nm Kinetic Modelk′ = kOH,C [ OH]ss + kCl,C [Cl ]ss + kSO –4 ,C [SO–4 ]ss (2.13)If the generation of each of these radicals is assumed to be independent, withnegligible interconversion between them, the steady-state radical terms maybe expressed in the general form:[R.]ss =ΦRfRIa∑kR,Si [Si](2.14)where ΦR is the quantum yield of the photochemical generation of the rad-ical at 185 nm, and fR is the fraction of 185 nm photons absorbed by theparent species generating R.. The sum of the fR values may not equal unityif other solutes exist that also absorb a significant fraction of 185 nm pho-tons, yet do not contribute to the radicals included in equation 2.14. Thecontribution of each of the terms of k′ will depend on the distribution of ab-sorbed photons (fR), the overall reactivity of the solution with each radical(∑kR,Si [Si]), and the reactivity of the probe compound with each radical(kR,C). The key assumption in using equation 2.14 in the equation 2.13 isthat interconversion between radical species is negligible. This assumption istested experimentally in subsequent chapters. While discrepancies betweenobservations and predictions using equations 2.13 and 2.14 will require mod-ification of the model, the details of the discrepancies themselves may shedmechanistic insights. Lastly, additional terms may be added to the above toincorporate additional radical species other than the three considered thusfar.392.5. Experimental ApparatusThe fractions fR of absorbed photons by each absorbing solute of the solu-tion are given by:fR =αRαtot(2.15)where αR is the absorption coefficient for solute R, and αtot is the absorp-tion coefficient for the entire solution, both with units of cm−1. The valueof αtot = αR1 + αR2 + ... = ΣαRi . For a given solute Si, the value of αi isrelated by αi = i[Si], where i is the molar absorption coefficient for soluteSi, with typical units of M−1 cm−1.2.5 Experimental ApparatusTwo collimated beam apparatus were used for this study and are describedhere. As true collimation requires the absence of divergence of radiation,with the electromagnetic wavefront forming a plane normal to the directionof propagation (i.e. optical path), the apparatus used here are understoodto imply quasi-collimated devices designed to sufficiently approximate truecollimation. One of the collimated beams was dedicated to studies of the185 nm regime, and a second for 254 nm - H2O2 regime (UV/H2O2) studies.The former is described in detail below and illustrated in Figure 2.1. Thelatter is similar but without nitrogen gas purging capabilities or the emissionof 185 nm radiation.The 185 nm collimated beam used a low pressure mercury lamp with anenvelope made of undoped quartz (LightSources Inc, Orange CT USA) anda copper ballast. Such lamps are often referred to as “ozone generating”by lamp manufacturers. The odour of ozone is readily detected when these402.5. Experimental Apparatuslamps are ignited in ambient air. The lamp was mounted in a tubular alu-minum housing with an aperture for fitting of a collimation tube and twoports for purging air from the optical path with nitrogen gas. The interiorof the aluminum housing was polished to increase reflectance of radiationabove the collimation tube. The collimation tube used in all 185 nm exper-iments was composed of black teflon, measured 18 cm in length and 3.2 cminner diameter, and had an inside surface roughened to reduce reflection.The sample cells (i.e reaction vessels) were commercially available precisioncylindrical cells made of fused silica (Starna, Atascadero CA USA), either1.0 cm or 2.0 cm in path length, with teflon stoppers and miniature tefloncoated magnetic stir bars placed inside. These cells had an inner diameter of1.9 cm, ensuring complete cross-sectional illumination of the sample underthe collimation tube. For irradiation, a sample filled cell was placed atop asmall stir plate, mounted upon an x-y stage fastened to a laboratory jackstand, and aligned beneath the collimation tube with the shutter closed.The vertical position of the top of the cell was adjusted with the jack standto a common height used in all experiments. A wide shutter housing wasfastened to the bottom of the collimation tube. A transparent curtain ofplastic, extending approximately 3 cm below the bottom of the shutter hous-ing, and overlapping the sample, was used to ensure an uninterrupted flowof nitrogen gas down the collimation tube and over the sample. Nitrogenflow around the sample was tested using a smoke pen. The nitrogen gasports of the lamp housing were located on opposite ends of the housing toreduce dead zones and used barbed brass fittings to accept plastic tubingcarrying nitrogen gas from a standard cylinder. A regulator on the cylinderwas set to a nominal pressure of 5 psi and a rotameter was used to measureand regulate the flow of nitrogen into the lamp housing to ∼ 5 L min−1.For temperature control, a custom made mount was fabricated (QuantumNorthwest, Liberty Lake WA USA) and is described in more detail in Chap-ter 3. A shutter of thick card was used via a slot in the shutter housing andoperated manually in coordination with a stop-watch to measure exposuretimes. During daily initial use, nitrogen flushing was initiated, followed bythe ignition of the lamp, and the system allowed to stabilize for at least412.5. Experimental Apparatustwenty minutes prior to any sample irradiations. 2 11 1 3 4 4 5 6 7 8 9 10 Figure 2.1: The 185 nm Collimated Beam Apparatus: 1. copper ballastand lamp socket, 2. aluminum lamp housing, 3. low pressure mercury lamp(inside housing), 4. nitrogen gas ports, 5. teflon collimation tube, 6. shutter(in open position), 7. plastic curtain, 8. fused silica sample cell with teflonstir bar and stopper, 9. magnetic stir plate, 10. laboratory jack standThe 254 nm collimated beam used for UV/H2O2 experiments also used alow pressure mercury lamp (LightSources Inc, Orange CT USA), which asmentioned above prevents the emission of 185 nm radiation and is oftendescribed by lamp manufacturers as “germicidal” and “non-ozone generat-ing”. The lamp was cooled using ambient air via fans to push air throughthe lamp housing. A collimation tube, painted black, provided a distancefrom lamp to liquid sample surface of at least 30 cm with an inner diameterof 6 cm. Samples were held in 50 mL dishes of less than 5 cm inner diameter,placed upon a stir plate and used teflon coated stir-bars. The stir plate was422.6. Materialsplaced upon a laboratory jack stand for accurate vertical position control.For temperature controlled experiments, this dish was replaced by a customfabricated water jacketed miniature borosilicate glass beaker (Cansci GlassProducts Ltd, Richmond BC Canada) through which water flowed from anexternal recirculating chiller (Thermo Fischer Scientific, Waltham MA USA)via plastic tubing. Additional details on temperature control are found inChapter 3.2.6 MaterialsAll solutions were made using ultrapure water (18.2 MΩ cm). Analyticalgrade reagents (Sigma-Aldrich, St. Louis MO USA) were used for all chem-icals, other than Natural Organic Matter (NOM). Well characterized NOMreference materials were obtained from the International Humic SubstancesSociety (IHSS) as reverse osmosis isolates in freeze dried powder form. BothSuwannee River and Nordic Lake NOM stock solutions were made by dissolv-ing NOM powder in water acidified with H2SO4, neutralizing the solutionswith NaOH, and filtering by pre-washed 0.45 µm filters into glass bottlesfor storage at 4 ◦C until use. Ultrahigh purity nitrogen gas (Praxair CanadaInc, Mississauga ON Canada) was used for purging air from the optical pathof the 185 nm enabled collimated beam.2.7 Probe Compound Selection andCharacterizationThe two probe compounds used in this work were carbamazepine (CASNumber 298-46-4) and nitrobenzene (CAS Number 98-95-3). The primarycriteria used for selecting the probe compounds were: (1) ease of quantifi-cation by HPLC with UV detection and (2) negligible direct photolysis at254 nm relative to OH oxidation under experimental conditions. The latter432.7. Probe Compound Selection and Characterizationcriteria obviates the need to eliminate 254 nm radiation from the 185 nmcollimated beam, or correct for its effect. An additional criteria was thatcompounds be of practical interest as environmental contaminants. Carba-mazepine, a common and persistent pharmaceutical, is a useful indicator andtracer of wastewater influences in the environment remote from the pointof discharge (Clara et al., 2004; Tixier et al., 2003), while nitrobenzene is awell known industrial contaminant.The direct photolysis rate at 254 nm is proportional to the product of themolar absorption coefficient (254) and the quantum yield (Φ254) at 254 nm.If either of these parameters are sufficiently low, then direct photolysis willbe negligible. To ensure this was the case for both carbamazepine and ni-trobenzene, these parameters were measured. Additionally, the second-orderOH rate constants (kOH,C) of each are required, and may be determined bycompetitive kinetics with a reference compound. The reference compoundused was para-chlorobenzoic acid or pCBA (CAS Number 74-11-3). The ex-perimentally determined values of 254, Φ254, and kOH,C are listed in Table2.1.Table 2.1: Photochemical reaction parameters for probe compoundsat 254 nmCompound 254 Φ254 kOH,C(M−1 cm−1) (M−1 s−1)carbamazepine 6759± 190 0.00067± 0.00002 6.8± 0.6× 109nitrobenzene 6240± 130 0.007± 0.001 3.9× 109 apCBA 3410± 75 0.011± 0.003 5.0× 109 ba Reference rate constants from Buxton et al. (1988)b Reference rate constants from Neta and Dorfman (1968)It can be seen that it is the low value of Φ254 for carbamazepine that ensurenegligible 254 nm photolysis relative to the reference pCBA. Furthermore,both carbamazepine and nitrobenzene probes are superior to pCBA basedon chromatographability, producing larger HPLC peaks with larger dynamic442.7. Probe Compound Selection and Characterizationranges and lower limits of detection due to higher λ values in the UV. Also,their peaks are more symmetric as a result of negligible acidity of eithermolecule. Organic acids, such as pCBA (pKa ∼ 4), are generally moreprone to peak asymmetry and are highly sensitive to mobile phase pH whenwithin approximately two pH units of the pKa. While pCBA has been apopular probe in AOP literature, its replacement by carbamazepine wouldimprove results. Unlike nitrobenzene, carbamazepine is of negligible volatil-ity, simplifying experimental techniques and analysis.2.7.1 Molar Absorption Coefficients at 254 nmFor the determinations of 254, a series of solutions were made by dilution ofa stock of known concentration prepared gravimetrically. The absorbance ofeach solution was then measured at 254 nm in a 1.0 cm path length quartzcuvette using a UVmini-1240 Spectrophotometer (Shimadzu, Kyoto Japan).The slope of the concentration versus the absorbance was calculated by lin-ear regression, then used to calculate 254 based on the Beer-Lambert law.Triplicate determinations were made using three separate gravimetric prepa-rations of stock solutions. The final results are averages of three determina-tions. See Appendix - Table A.2 for data.2.7.2 Photolysis Quantum Yields at 254 nmFor determinations of Φ254, solutions of each probe alone in ultrapure wa-ter at a concentration of approximately 1 µM were irradiated by 254 nmcollimated beam for a range of exposure times, with the remaining concen-trations quantified by HPLC. The incident fluence rate Io at 254 nm, forthese exposures, was determined by KI-KIO3 actinometry. Since such solu-tions are nearly transparent at 254 nm, the incident and volume averagedfluence rates are approximately equal. The measured Io was used to convertthe observed time-based direct photolysis rate constant k′d, with units s−1,to fluence-based rate units of m2 J−1. Then, values of Φ254 were calculated452.7. Probe Compound Selection and Characterizationfrom the fluence-based expression for k′d given by (Bolton and Stefan, 2002):k′d =Φ254 254 ln(10)10 U254(2.16)where 254 has units of M−1 cm−1, and U254 is the molar photon energy4.72× 105 J mol−1, using the more precise wavelength of 253.7 nm. Darkcontrols are essential for nitrobenzene to correct for volatilization, unless asealed irradiation cell is used. See Appendix - Tables A.3 to A.7 for data.2.7.3 Second-order OH Rate ConstantsThe second-order rate constants of the probe carbamazepine with OH (kOH,C)was determined using competitive kinetics in the UV/H2O2 system, withpCBA as the reference compound, using kOH,pCBA ≡ 5.0× 109 M−1 s−1(Neta and Dorfman, 1968). A H2O2 concentration of 5.0 mg L−1 was used.2-methyl-2-propanol (CAS Number 75-65-0), commonly known also as tert-butanol, was used as an OH scavenger at a concentration of approximately10 mg L−1. This scavenger was used to ensure that the value of [ OH]ssremains constant over the time of irradiation and that the contribution ofthe probe compounds to the OH scavenging in solution is negligible (i.e.kOH,C [C]  ΣkOH,S [S] for each). Absorbance of 254 nm by tert-butanol isnegligible (254 < 10 M−1 cm−1). Tert-Butanol has been widely used as ascavenger in pulse radiolysis studies (Schuchmann and Von Sonntag, 1979),often in competitive kinetics determinations of rate constants. See Appendix- Tables A.8 to A.11 for data.462.8. Analytical Methods2.8 Analytical MethodsHPLC Analysis: The quantification of the probe compounds, carbamazepineand nitrobenzene, and the reference compound pCBA, were performed byhigh-performance liquid chromatography (HPLC) using a Dionex UltiMate3000 System (Thermo Fisher Scientific, Waltham MA USA). The autosam-pler withdrew a 100 µL sample volume from each vial, injecting it into a15 mm C18 column maintained at 35 ◦C in isocratic mode. A mobile phaseflow of 1.0 mL min−1 was used and composed of 30% acetonitrile and 70%water acidified to pH 2.5 with approximately 10 mM of phosphoric acid.UV detection was performed using one of three wavelengths 211, 239, and265 nm. Use of 211 nm allowed the greatest dynamic range, while in thepresence of NOM interference required use of longer wavelengths. Stocksolutions were prepared gravimetrically using 5 mL volumetric flasks and asmall amount of methanol as a cosolvent. The stock solutions were theneither diluted using ultrapure water to form working intermediate stock so-lutions or diluted to a set of standards using the HPLC mobile phase asdiluent. The dynamic range was found to span concentrations from 2 µMdown to 0.01 µM with the %RSD always less than 2.5% (see Appendix -Table A.1).DOC Analysis: All dissolved organic carbon (DOC) measurements wereperformed by a GE Sievers M9 TOC Analyzer (GE Analytical Instruments,Boulder CO USA), using the UV-persulfate method. Standard solutions ofpotassium hydrogen phthalate (CAS Number 877-24-7) were used for in-strument calibration, and to verify DOC measurements of gravimetricallyprepared solutions of tert-butanol, methanol, acetate, and acetone, as wellas the reference materials Suwannee River and Nordic Reservoir NOM. Notethat all samples analyzed were either filtered with sub-micron pore size fil-ters or composed of particle free pure solutions. Thus the measurementsmade by TOC are equivalent to the operationally defined DOC.Ion Exchange Chromatography: The anionic composition of all solu-472.9. Initial Testing of 185 nm Experimental Methodstions assembled from analytical grade reagents and ultrapure water wereverified in concentration against prepared standards and checked for con-taminants using ion exchange chromatography, A Dionex Ion Chromatog-raphy system (Thermo Fisher Scientific, Waltham MA USA) was used.KI-KIO3 Actinometry at 254 nm: Quantification of 254 nm fluencerates were performed using the KI-KIO3 actinometer (Bolton et al., 2011;Rahn, 1997; Rahn et al., 2003). Quantification of I –3 was performed spec-trophotometrically at 350 nm, using a UVmini-1240 spectrophotometer (Shi-madzu, Koyoto Japan).Quantification and Quenching of Hydrogen Peroxide: The H2O2concentration applied in UV/H2O2 experiments, both prior to and followingirradiations was performed using the I –3 method (Klassen et al., 1994). Sam-ples were quenched of remaining H2O2 using bovine catalase (Sigma Aldrich)prior to HPLC analysis. Quenching used a 10 µL droplet of 100 µg L−1 ofcatalase, placed at the bottom of an HPLC vial before a 1 mL sample ofsolution was added. Quantification of I –3 was measured spectrophotometri-cally at 350 nm using a UVmini-1240 spectrophotometer (Shimadzu, KoyotoJapan).pH: An Oakton pH meter calibrated at pH 4, 7, and 10 was used for all pHmeasurements using a disposable type gel filled probe.2.9 Initial Testing of 185 nm ExperimentalMethodsThis section describes the preliminary investigations and due diligence per-formed regarding the performance of the 185 nm collimated beam experi-ments.482.9. Initial Testing of 185 nm Experimental MethodsMixing Conditions: Among the requirements of the collimated beam ap-proach is the need for the irradiated sample volume to be well mixed. Toensure that the extent of reaction was not influenced by the mixing speed,the degradation rate of carbamazepine was measured at three speeds of thestir plate (approximately 500, 1000, 1500 rpm) during 10 minute irradia-tions. No effect was found, and circa 1000 rpm was used for all subsequentexperiments, due to greater stability of the stir bar operation at that speed.Dissolved Oxygen: While the presence of dissolved O2 creates oxidativeconditions in the 185 nm regime by scavenging H /e –aq , the absence of dis-solved O2 may induce reductive conditions. In the presence of pure nitrogengas flowing over an irradiated sample, O2 in solution is stripped and ulti-mately depleted if the solution is not covered. Tests using ultrapure watersaturated with air, in both open 10 mL beakers and teflon stoppered fusedsilica cuvettes (i.e. cells), were tested for dissolved O2 following irradiationsin the 185 nm collimated beam. During irradiations, N2 flowed at 5 L min−1over the beakers or cells, mixed with miniature stir bars. A modified iodo-metric method based on Winkler’s titration revealed that approximatelyhalf the dissolved O2 was depleted within ten minutes in the open beakers,whereas depletion was negligible when sealed cuvettes were used instead.Thus, cylindrical cuvettes of 1.0 cm were used for all subsequent experi-ments. A small air bubble was left in the cell next to the teflon stopper andthe portion of the cell arm above painted black. The iodometric test fordissolved O2 miniaturized the quantities of reagents required, and in placeof adding starch indicator and titrating with thiosulfate (S2O2–3 ), the ab-sorbance at 350 nm was measured, with a calibration curve created from airsaturated solutions, N2 saturated solutions, and mixtures of the two.Influence of Phosphate Buffer: Ideally, the pH of irradiated solutionsshould not drift during experiments. Some researchers have used phosphatebuffers (H2PO–4 /HPO2–4 ) composed of sodium salts. However, studying theinfluence of water composition is complicated if the pH buffer itself par-ticipates in photochemical reactions. Thus, the use of phosphate buffers492.9. Initial Testing of 185 nm Experimental Methodswas tested in the 185 nm regime. It was found that orthophosphates doparticipate in the removal of the probe compound carbamazepine, likely byabsorption of photons and formation of phosphate radicals (Maruthamuthuand Neta, 1977). Solutions containing carbamazepine (0.25 µM) and tert-butanol (7.0 mg L−1 as C), with phosphate buffer strength of 10 mM at pH6.0, 7.0 and 8.0, were irradiated. The degradation rate increased with pH,suggesting a greater influence of HPO 2–4 relative to H2PO–4 (See Appendix- Table A.13 for data). When such solutions at pH 7.0 are tested withincreasing buffer strength of 0, 1, 10, and 100 mM, again the degradationrate increases significantly (See Appendix - Table A.13 for data). Thus,phosphate buffers were not used in this work, and the pH of solutions wasmonitored before and after irradiation to detect significant change. Driftingof pH during irradiations was not observed.Influence of Fluoride: Kinetic studies often use ionic strength effectsto elucidate mechanistic aspects of reacting systems in aqueous solutions.The activity of ionic reactants will be more strongly influenced by changesin ionic strength than neutral molecules. However, if neutral reactants (e.g.Cl ) are in equilibrium with ionic species (e.g. Cl + Cl– ⇀↽ Cl –2 ), a so-calledsecondary-salt effect may manifest when ionic strength is varied (Moore andPearson, 1981). Modifying the ionic strength for this purpose requires aspecies that will be photochemically inert, neither absorbing radiation norparticipating in radical reactions. Sodium fluoride (NaF) was considered forthis purpose. The influence of NaF was found to be negligible on the 185 nmsystem, using carbamazepine (0.25 µM) and tert-butanol (7.0 mg L−1 as C),with F– concentrations of 1, 10, and 100 mg L−1 (See Appendix - Table A.14for data).Opaque Assumption of Water at 185 nm: The absorbance coefficientof water at 185 nm is relatively high, and in this work is assumed to be1.8 cm−1 (Weeks et al., 1963). Thus, approximately 98% of the radiationentering pure water will be absorbed within a path length of 1.0 cm, withapproximately 60% absorbed within a path length of 0.2 cm. Unlike the502.9. Initial Testing of 185 nm Experimental Methodsoptical absorption of water at longer wavelengths, the substantial attenua-tion of 185 nm radiation over short path length relative to typical lamp andreactor dimensions implies that water is essentially an opaque fluid at thatwavelength. As discussed earlier, this fact has important implications onhow the absorbed energy dose of a 185 nm AOP is described and imposessevere geometrical constraints on reactor design.To verify the assumption that water is effectively opaque at 185 nm, twoprecision cylindrical fused silica cells of path length 1.0 cm and 2.0 cm werefilled with the same solution of carbamazepine (0.25 µM) and tert-butanol(7.0 mg L−1 as C), irradiated for several exposure times t, and analyzed forthe remaining carbamazepine. The results reveal that when ln([C]t/[C]o) isplotted, not versus t, but versus t normalized by volume V of sample (i.e.t/V ), the slopes of the two curves are in excellent agreement (See Appendix- Table A.15 for data). This supports the description of the extent of re-action using D = IoSt/V (equation 2.8) and the associated assumptions.While the product IoS is the same in both cases above, the same value ofD is delivered to the two systems only when both terms t/V also equal.Since the volume of the smaller cell is half that of the larger, an exposuretime of half as long in the smaller cell, relative to the larger, is required toprovide identical values of D to both cells. This result confirms the opacityassumption of water under the conditions used.KI KIO3 Actinometry and 185 nm: As discussed above, expressingthe extent of reaction in the 185 nm regime requires knowledge of Ia, whichcan be obtained from Io in a collimated beam, and used to calculate Dvia equations 2.5 and 2.7. Measurement of Io by chemical actinometry at185 nm with actinometers described in the literature (Kuhn et al., 2004)is difficult, generally requiring gas-tight assemblies and gas chromatographyanalysis. In contrast, actinometry at 254 nm is a relatively simple determina-tion using the KI KIO3 actinometer (Bolton et al., 2011; Rahn, 1997; Rahnet al., 2003). Given the high absorbance of I– expected at 185 nm (Weekset al., 1963), and the charge-transfer-to-solvent mechanisms involved, it is512.9. Initial Testing of 185 nm Experimental Methodsreasonable to suspect that 185 nm photons may contribute to the responseof the KI KIO3 actinometer. If so, then 185 nm radiation would interferewith a 254 nm measurement, causing a spurious response. However, such acondition would allow 185 nm radiation itself to be measured by KI KIO3actinometry using a difference method. Two irradiations would be required.In the first, the full amount of both 185 and 254 nm radiation reach the acti-nometer. In the second, 185 nm radiation is eliminated before reaching theactinometer, while 254 nm radiation is unchanged from the first case. Undersuch conditions, any difference in response between the two measurementswould be due to the contribution of 185 nm.To selectively block only 185 nm, while transmitting 254 nm radiation un-changed, a properly design interference filter (Heavens, 1991) or fused silicawindow doped with titanium may be applied. However, an adequate alter-native of much lower cost was used, involving a 12 mm thick optical windowof high purity natural quartz (Esco Optics, Oak Ridge NJ USA). It was as-sumed, and supported by manufacturer information, that 185 nm radiationis attenuated by approximately 40%, due to absorption, over a path lengthof 2 mm in natural quartz, while negligible absorption occurs for 254 nmradiation over the same distance. Over the 12 mm thick distance of the nat-ural quartz window, therefore, it is calculated that approximately 0.4% of185 nm radiation entering the window reaches the opposite end. In contrast,full transmission is assumed at 254 nm, minus reflection losses of approxi-mately 4%, estimated using Fresnel’s equation at normal incidence (Bornand Wolf, 1999), and an estimate of the refractive index at 185 nm (Kita-mura et al., 2007).To test this approach, the actinometer solution was placed in a sealed fusedsilica cell of 2.0 cm path length, under both the 254 nm and 185 nm col-limated beam apparatus, with replicate determinations made at the sameexposure time. The resulting apparent incident 254 nm fluence rates Io weresubsequently calculated.522.9. Initial Testing of 185 nm Experimental MethodsIn the case of the 254 nm radiation alone, Io was determined to be 0.501±0.013 mJ cm−2 without the quartz window, and 0.486± 0.008 mJ cm−2 withthe quartz window, each averaged over ten replicate irradiations (see data inAppendix - Table A.15). The ratio of the latter to the former is 0.97± 0.03,which is in agreement, within experimental error, for what would be ex-pected if only reflection losses from the quartz window are significant. In thecase of the 185 nm collimated beam, Io was found to be 0.872±0.026 mJ cm−2without the quartz window, and 0.842±0.017 mJ cm−2 with the quartz win-dow, averaged over five replicate irradiations (See Appendix - Tables A.17for data). The ratio of the latter to the former is 0.97± 0.03, identical withthat obtained using 254 nm radiation alone.The conclusion is that the KI KIO3 actinometer does not respond to the185 nm radiation present. While this actinometer cannot be used to quan-tify 185 nm radiation in the present apparatus, neither does 185 nm interferewith 254 nm measurements. Given the reactions proposed to be involved ingeneration of I –3 by this actinometer (Rahn, 1997), the absorption of 185 nmby I– (Fox, 1970) and subsequent generation of I is expected. The lack ofsignificance may be due to absorption of 185 nm by other components of theactinometer solution, such as IO –3 , without contribution to the measuredresponse. The development of a convenient actinometer for use at 185 nm,that is also insensitive to 254 nm radiation, would be useful.53Chapter 3Temperature3.1 Temperature and Water TreatmentSeasonal fluctuations in surface water temperature span a wide range atlatitudes far from the equator, often from 0 to 20 ◦C or more. Treatmentprocesses in such locations must ensure adequate performance regardlessof temperature. Cold temperatures generally present a greater challenge.The higher viscosity of cold water reduces the particle removal efficiencyof solid-liquid separation processes. Depressed reaction rates of chemicaldisinfection necessitate increased contact time. While UV disinfection isrelatively insensitive to water temperature, the influence of temperature onUV based AOPs is not well documented.3.2 Temperature Dependence of UV AOPsThe OH driven treatment of UV based AOPs involves composite chemicalreactions. Component steps include photolytic generation of OH, reactionwith target contaminants, and competition reactions with major solutes thatact as radical scavengers. These component steps themselves are composedof multiple elementary reactions. Despite this complexity, the net reactionrate of many composite reactions may be represented by a single Arrheniusexpression, involving an overall activation energy.k = Ae−E/RT (3.1)543.2. Temperature Dependence of UV AOPswhere k is an overall reaction rate constant, A is the pre-exponential factor,E is the activation energy, R is the universal gas constant 8.314 46 J K−1 mol−1,and T is the absolute temperature. If a plot of ln(k) vs. 1/T produces astraight line, then E may be extracted from the slope.Component steps may have individual activation energies. Whether OHis generated by 185 nm photolysis of H2O, or 254 nm photolysis of H2O2,the experimentally observed pseudo-first order rate constant k′ for the OHdegradation of a trace organic contaminant C, in a solution containing amuch larger concentration of the scavenger S, may be expressed with amodified version of equation 2.12 as:k′ =kOH,CkOH,SΦ Θ (3.2)where kOH,C and kOH,S are the second order rate constants of OH with atarget C and scavenger S respectively, Φ is either the quantum yield of waterphotolysis at 185 nm or the 254 nm photolysis of H2O2, and Θ contains alltemperature independent terms. This expression can be rewritten in termsof the Arrhenius expressions for the three components as:k′ = Atote−Etot/RTΘ =ACe−EC/RT AΦe−EΦ/RTASe−ES/RTΘ (3.3)from which the overall Arrhenius activation energy may be related to thecomponent activation energies by:553.2. Temperature Dependence of UV AOPsEtot = EC + EΦ − ES (3.4)with Etot found from the experimental Arrhenius plot given byln(k′) = ln(Atot Θ)− Etot/RT (3.5)If the term Θ does in fact posses a significant temperature dependence, thenan Arrhenius plot of ln(k′) vs. 1/T will reveal curvature. The absence ofcurvature allows the estimation of one of the component activation energiesif the others are known. For diffusion limited reactions, such as those in-volving OH, Ea is often in the range of 10 to 20 kJ mol−1 (Buxton et al.,1988). Examples of E for OH reactions include 21 kJ mol−1 for HCO –3 and10 kJ mol−1 for tert-butanol (Buxton et al., 1988). It is important to notethat, in the context of this discussion, an activation energy is not necessarilyassociated with an elementary reaction nor a transient intermediate. Inter-pretation at the molecular level is difficult. Even the reaction rate constants,in particular kOH,S for scavengers, represents an effective reaction rate, asthe scavenger is generally not a pure substance but a complex mixture oforganic matter. The activation energy concept may nevertheless be usefulin gaining mechanistic insights.In the case of 185 nm radiation, the term Θ contains the absorption coeffi-cient of water αH2O, which varies with temperature and may contribute tothe temperature sensitivity of the observed reaction rate. A temperaturecoefficient of 0.05 cm−1 ◦C−1 has been reported between 20 and 50 ◦C (Bar-rett and Mansell, 1960; Weeks et al., 1963). This gives a value for αH2O of1.55 cm−1 at 20 ◦C and 2.30 cm−1 at 35 ◦C. The behaviour at lower tem-563.2. Temperature Dependence of UV AOPsperatures is not known, but a value for αH2O at 5◦C of between 0.80 and1.00 cm−1 is reasonable. The impact of αH2O on the process depends on reac-tor design. For designs that rely on the total absorption of radiation withina short distance (≤ 1 cm) operation at lower temperatures may result in a5 to 10% reduction in absorbed radiation. For reactors that utilize largeroptical path lengths, mixing requirements may become slightly relaxed atlower temperatures, and the overall effect less significant.Once 185 nm photons are absorbed by water, OH are generated with aquantum yield ΦH2O of approximately 0.3 (Getoff and Schenck, 1968). Thetemperature dependence of ΦH2O is not known, but may follow an Arrhe-nius type relation with an activation energy dependent on competing ratesof radical recombination kr and escape ke from the solvent cage.H2O185 nm−−−−⇀↽ −kr[H , OH]aqke−→ H + OH (3.6)The effective activation energy EΦ will depend on kr and ke, which are notdirectly accessible by experimental methods used in this work. The situa-tion is similar for 254 nm photolysis of H2O2, though more information isavailable on the temperature dependence of the quantum yield ΦH2O2, whichis approximately unity at 25 ◦C.H2O2254 nm−−−−⇀↽ −kr[HO , OH]aqke−→ HO + OH (3.7)Based on limited temperature dependence data reported for 254 nm photol-573.3. Experimental Approachysis of H2O2, an activation energy EΦ for equation 3.7 is estimated to bein the range of 11 to 13 kJ mol−1 (Baxendale and Wilson, 1957; Hunt andTaube, 1952; Volman and Chen, 1959).3.3 Experimental ApproachExperimental investigations of temperature dependence employed solutionswith identical compositions for the collimated beam studies of both the185 nm and the 254 nm-H2O2 regimes, except for the addition of H2O2 inthe latter.Solutions were composed of approximately 0.25 µM of the probe compoundcarbamazepine, and 8 mg L−1 as C (0.2 µM) of tert-butanol in ultrapurewater. For 254 nm-H2O2 studies, 3.5 mg L−1 (0.1 µM) was added prior toirradiation.The 185 nm irradiations used a 1.0 cm path length cylindrical fused silica cellwith miniature teflon coated stir bar and stopper, and was placed in a cus-tom made temperature controlled cell holder (Quantum Northwest, LibertyLake, WA USA). The cell holder, composed of a black anodized aluminumbody and black teflon lid, was equipped with a precision Peltier cell, mag-netic stirring motor, water circulation with external ice bath for removalof heat, and nitrogen gas ports for condensation prevention. Temperaturecontrol within 0.1 ◦C was verified using a fine gage thermocouple (Omega,Laval QC Canada) placed in contact with the cell.The 254 nm irradiations used a custom water jacketed borosilicate beaker(Cansci, Richmond BC Canada) with temperature controlled by water flow-ing from a recirculating chiller (Thermo Fisher Scientific, Waltham MAUSA). A miniature teflon stir bar was placed in the beaker and the beakerplaced on a magnetic stir plate mounted to a jack stand beneath the 254 nmcollimated beam described in Chapter 2.583.4. ResultsFor both types of collimated beams, a series of irradiations were performedwith samples exposed for a range of exposure times. All exposures were per-formed in triplicate. Approximately 1 mL samples were taken and placedin 2 mL HPLC vials. In the case of 254 nm-H2O2 tests, a 10 µL droplet of100 µg L−1 bovine catalase solution was placed at the bottom of each HPLCvial to quench residual H2O2 and avoid contributions to probe degradationfrom dark reactions between sampling and HPLC analysis. Dark controlsfor 254 nm-H2O2 tests were also included.3.4 Results3.4.1 Temperature Effects in the 254 nm - H2O2 RegimeThe results for 254 nm - H2O2 tests are displayed in Figure 3.1 and Table3.1. They reveal that the reaction rate observed increases with temperaturefrom 5 ◦C to 35 ◦C.The observed pseudo-first order rate constants k′ are calculated from lin-ear regression of the triplicate measurements for each irradiation time used,with the standard error of the slope used to express uncertainty σk′ . Theuncertainty of ln(k′) is calculated from the approximation σln(k′) ≈ σk′/k′(Harris, 2010). The calculated values are displayed in Table 3.1.Table 3.1: Effect of temperature on removal rate of carbamazepine probein 254 nm-H2O2 regimeT (◦C) 5 20 35k′ × 103 (min−1) 7.3± 0.3 11.4± 0.2 16.1± 0.5ln(k′) −4.92± 0.04 −4.47± 0.02 −4.13± 0.03[H2O2] = 3.5 mg L−1, [tBuOH] = 8 mg L−1 as C, [CBZ]o ' 0.25 µMThe Arrhenius plot for the 254 nm-H2O2 regime is displayed in Figure 3.2using the data tabulated in Table 3.1 and the equation 3.4.593.4. Results0 10 20 30 40 50 60−1−0.8−0.6−0.4−0.205 ◦C20 ◦C35 ◦CTime (min)ln([CBZ] t/[CBZ] o)Figure 3.1: Temperature dependence in 254 nm - H2O2 regime. [H2O2] =3.5 mg L−1. [tBuOH] = 8 mg L−1 as C. [CBZ]o ' 0.25 µM.603.4. Results3.1 3.2 3.3 3.4 3.5 3.6 3.7−5−4.8−4.6−4.4−4.2−41/T × 103 K−1ln(k′ )Figure 3.2: Arrhenius plots for 254 nm-H2O2 regime. [H2O2] = 3.5 mg L−1.[tBuOH] = 8 mg L−1 as C. [CBZ]o ' 0.25 µM.3.4.2 Temperature Effects in the 185 nm RegimeThe results for 185 nm tests are displayed in Figure 3.3. As with the pre-vious case, the reaction rate is observed to increase with temperature from5 ◦C to 35 ◦C.As with the 254 nm-H2O2 regime, the observed pseudo-first order rate con-stants k′ are calculated from linear regression of the triplicate measurementsfor each irradiation time used, with the standard error of the slope used toexpress uncertainty σk′ . As before, the uncertainty of ln(k′) is calculatedfrom the approximation σln(k′) ≈ σk′/k′. The calculated values are displayedin Table 3.2.The Arrhenius plot for the 185 nm regime is displayed in Figure 3.4 usingthe data tabulated in Table 3.2.613.4. Results0 2 4 6 8 10 12−4−3−2−105 ◦C20 ◦C35 ◦CTime (min)ln([CBZ] t/[CBZ] o)Figure 3.3: Temperature dependence in 185 nm regime. [tBuOH] =8 mg L−1 as C. [CBZ]o ' 0.25 µM.The slopes of both Arrhenius plots, as calculated by linear regression, allowthe determination of the experimental activation energy for both the 185 nmand 254 nm-H2O2 regimes. The values are displayed in Table 3.3.623.4. ResultsTable 3.2: Effect of temperature on removal rate of carbamazepine probein 185 nm regimeT (◦C) 5 20 35k′ × 102 (min−1) 24.7± 0.2 30.5± 0.5 35.6± 0.5ln(k′) −1.40± 0.01 −1.19± 0.02 −1.03± 0.01[tBuOH] = 8 mg L−1 as C, [CBZ]o ' 0.25 µMTable 3.3: Experimental (overall) activation energies for carbamazepine degra-dation in the presence of tert-butanol in 254 nm-H2O2 and 185 nm regimes254 nm-H2O2 185 nmEa (kJ mol−1) 18.7± 0.9 8.6± 0.5NB: These values pertain to the composite reactions and not elementary steps.3.1 3.2 3.3 3.4 3.5 3.6 3.7−1.4−1.2−11/T × 103 K−1ln(k′ )Figure 3.4: Arrhenius plots for 185 nm regime. [tBuOH] = 8 mg L−1 as C.[CBZ]o ' 0.25 µM.633.5. Discussion3.5 DiscussionThe results indicate that the 185 nm AOP is less temperature sensitive thanthe UV/H2O2 AOP under the conditions tested. This result is understoodto apply to water matrices for which H2O is the major absorber of 185 nmphotons, where the scavenging term Σki[Si] has a magnitude greater than105 s−1 and were the activation energy of the target contaminant is less thanthe effective activation energy for the scavenging term (i.e. EC < ES). Ifthe last condition is reversed (i.e. EC > ES), the observed rate k′ mayfollow an inverse relationship with temperature.In the present case, the scavenger is the pure substance tert-butanol, withan OH activation energy ES reported as 10 ± 3 kJ mol−1 (Ervens et al.,2003). Using an average of the reported values of the activation energy ofH2O2 photolysis at 254 nm, EΦ = 12 ± 1 kJ mol−1, allows the estimationof the activation energy of OH with carbamazepine, EC , using equation3.4. In this manner, the value EC = 17 ± 5 kJ mol−1 is obtained. Applica-tion of equation 3.4 for the 185 nm regime allows for the estimation of EΦ,the activation energy of the 185 nm photolysis of water itself. A value ofEΦ ≈ 0 kJ mol−1 is obtained. The fundamental activation energies deducedfrom experimental data are listed in Table 3.4.Table 3.4: Summary of fundamental activation energies estimated from thisworkOH+ CBZ H2O185 nm−−−−−→ H + OHEa (kJ mol−1) 17± 5 ≈ 0Explanation for the EΦ ≈ 0 kJ mol−1 value of H2O photolysis at 185 nm isprovided by the proposed structure for H2O in the liquid state. While theprecise structure of liquid H2O remains in dispute (Ball, 2008), the modernconsensus based on evidence from X-ray and neutron diffraction supports theview that virtually all molecules of H2O in the liquid state are dynamically643.5. Discussionhydrogen bonded to an average of four neighbours in an ice-like tetrahedralmotif with distorted bond angles (Frank, 1972; Franks, 2000). The existenceof non-hydrogen bonded interstitial H2O molecules (monomers) is supportedby evidence from Raman and infrared spectroscopy, though the proportionof such molecules is interpreted to be small (< 1%).Evidence from far-UV absorption also supports the existence of interstitialH2O monomers. Extensive measurements by Stevenson (1965) of 185 nmabsorption of ultrapure water in both the vapour and liquid state confirmobservations by others (Barrett and Mansell, 1960; Watanabe and Zelikoff,1953; Weeks et al., 1963) that the molar absorption coefficient of H2O vapouris three orders of magnitude greater than that of the liquid. Stevenson(1965) reported v = 22.1 M−1 cm−1 and ` = 0.0274 M−1 cm−1 at 23.5 ◦Cand made measurements of liquid absorption with increasing temperature.The ratio `/v remained approximately 0.0012 between 23 and 27◦C, risingsharply above 30 ◦C to 0.0090 at 91.8 ◦C. Such observations are explained bythe existence of monomers in the liquid state, representing a fraction of allmolecules on the order of 10−3 in the vicinity of 20 ◦C, with such monomersresponsible for virtually all 185 nm photon absorption. The temperaturedependence reported by Stevenson is interpreted as an increase in monomerpopulation with temperature that nevertheless remains a minority even nearthe boiling point. This is consistent with the considerable degree of hydro-gen bonding remaining at the boiling point and the relatively high criticaltemperature of water.Upon 185 nm excitation of an H2O monomer, it is proposed that only a rel-atively weak van der Waals force must be overcome in order for the photo-products to escape the solvent cage, since no hydrogen bonds are involved.Though the excited H2O molecule may lose energy to the solvent by collisionwith a rate kr proportional to√T , the excess energy of the excited moleculesitself is likely sufficient to overcome a van der Waals energy of ∼ 5 kJ mol−1.The activation energy for the photolysis of an H2O molecule, Eφ ≈ 0, appliesto the excited-state molecule that has absorbed a 185 nm photon. Such pho-653.6. Summarytons possess an energy of 647 kJ mol−1 while the bond-dissociation energybetween HO and H is approximately 494 kJ mol−1 (Darwent, 1970). The ex-cess energy, equivalent to approximately two hydrogen bonds, is more thansufficient to overcome a postulated van der Waals force.3.6 SummaryThe temperature studies conducted indicate that, under the conditions tested,the 185 nm-AOP is relatively insensitive to temperature. An activationenergy for the OH reaction with carbamazepine has been estimated as17 ± 5 kJ mol−1. The activation energy for the 185 nm photolysis of H2Ohas been estimated to be approximately 0 kJ mol−1 and supports the viewthat 185 nm photon absorption occurs in interstitial non-hydrogen bondedH2O monomers present as an approximate 10−3 fraction of all molecules.Additional temperature dependence studies should be investigated in watermatrices for which H2O is not the major absorber of 185 nm photons.66Chapter 4Dissolved Organic Matter4.1 Dissolved Organic Matter in Natural WatersWherever water is found in the environment, it will contain some amountof organic matter. Fractions of organic matter based on size may be catego-rized as particulate, colloidal or dissolved. Dissolved organic matter (DOM)is expressed quantitatively in terms of its carbon content in units of mg L−1as C, and defined operationally as that fraction not retained by a 0.45 µmnominal pore size filter (APHA, 2012). The dissolved organic carbon (DOC),is measured by oxidizing all organic carbon in an aqueous sample to CO2,then measuring the CO2 generated. If unfiltered, the measurement repre-sent the total organic carbon (TOC). The possibility of confusion exist whendiscussing DOM quantitatively, since DOC is measured and not DOM. Thediscrepancy between the two parameters is significant, since the proportionof carbon in DOM is approximately 50% in natural waters, and thus thevalue of DOC is typically half that of DOM when expressed in mass basedconcentration units.In some cases, such as in rainwater, groundwater and the ocean, the DOCis relatively low, often ≤ 1 mg L−1 as C. In other cases, such as swamps,wetlands, and soil it is relatively high, often  10 mg L−1 as C (Thurman,1985). In freshwater rivers and lakes, DOC values between 1 and 10 mg L−1as C are typical. When groundwater is used as a source of drinking water,the impact of organic matter may often be negligible. However, when sur-face waters are used (i.e. rivers and lakes), organic matter has significantconsequences to many practical aspects of treatment, storage and distribu-tion.674.1. Dissolved Organic Matter in Natural WatersSurface waters contain organic matter both of terrestrial origins received viadrainage over the watershed (allochthonous), and of aquatic origins derivedfrom photosynthetic and microbial activity (autochthonous) (Wetzel, 2001).The resulting material is a complex mixture in size and composition that de-fies simple descriptions. In surface waters, the DOC fraction represents themajority by mass, often 90% or more of the TOC (Wetzel, 2001). Yet, sincethe transition between colloidal and dissolved fractions is more a continuumthan a sharply defined point, DOC will contain a colloidal contribution tosome extent.The chemical composition of DOM varies between watersheds and withinthem due to meteorological events and seasonal fluctuations. The major-ity of DOM typically consists of molecules with molecular weights from afew hundred to up to several tens of thousands of daltons. Due to longresidence times in the environment, DOM is often heavily oxidized and oflow biodegradability. The composition of DOM may be divided betweennon-humic and humic components. The non-humic portion includes wellknown types of compounds such as carbohydrates, proteins, amino acids,fatty acids, aldehydes, ketones, alcohols and carboxylic acids. These occupythe lower molecular weight fractions. The majority of naturally occurringDOM is composed of humic substances, a general category of unknown struc-ture, which can be further classified as fulvic and humic acids. Fulvic acidstend to be smaller than humic acids, have higher oxygen content, and aresoluble at all pH values. Humic acids are on average larger and precipitateat pH 2 or less. Based on 13C NMR analysis and UV absorption spectra,the macromolecules of humic substances are known to posses, to varyingextents, carboxyl, hydroxyl, and amine functional groups, as well as sub-stantial conjugation and aromaticity. Such properties are likely to impartstrong influences on the optical and chemical properties of the water matrix,in particular the absorption of photons and oxidant reactivity.In addition to organic matter found in the pristine environment, other con-684.1. Dissolved Organic Matter in Natural Waterstributions to DOM may be significant. Municipal wastewater discharges,and even gross industrial contamination, may contribute measurably to theDOC found in surface waters. Anthropogenic materials such as these differsubstantially in chemical composition to those materials described above.In the case of biological wastewater effluent, for example, DOC will origi-nate largely from the cellular debris of bacteria used in the activated sludgeprocess, and the extracellular polymeric substances they excrete. Negligiblehumic material will be present. In order to distinguish the DOM from thepristine environment with materials from anthropogenic sources, the term“natural organic matter” or NOM is often used by engineers in the drinkingwater field.In water treatment, DOM has several important influences. The modifica-tion of surface chemistry on particulates and larger colloids and the com-plexation of metal coagulants influence the efficiency of solid-liquid sepa-ration processes. Adsorption of DOM to activated carbon reduces removalefficiency of target organic contaminants by occupying adsorption sites. Re-actions with DOM exert a demand on chlorine and other oxidants, usuallyincreasing the dose of oxidant required to produce a residual concentration.Absorption of radiation interferes with UV disinfection, acting as a filterthat reduces exposure of microbes and increases the required UV fluence ordose that must be applied to achieve a given degree of inactivation. Un-intended reactions between oxidants and DOM generate by-products suchas halogenated DBPs or smaller organic molecules that are more availableas substrate for subsequent microbial regrowth in distribution and storage.Components of DOM itself may impart colour, taste and odour to finishedwater, reducing the aesthetic quality of the product.694.2. The 185 nm AOP and Influence of Dissolved Organic Matter4.2 The 185 nm AOP and Influence of DissolvedOrganic MatterThe 185 nm AOP is influenced by DOM in at least two ways. First, photonsthat would otherwise contribute to the generation of OH from water pho-tolysis, may be absorbed by DOM without subsequent radical generation.Aromatic and carbonyl moieties offer sites of strong photon absorption dueto pi-bonds. Second, generated OH will react with DOM molecules in com-petition with target contaminants often many orders of magnitude lower inconcentration. The reactivity of DOM, quantified as an aggregate secondorder radical rate constant, is thus a measure of the strength of competitionand expressed by the product kOH,S [S], where [S] is the DOC in units ofmg L−1 as C and kOH,S has units of L mg−1 s−1. The scavenging product formost waters is typically in the range of 104-105 s−1 (Elovitz and von Gunten,1999; Goldstone et al., 2002; Westerhoff et al., 1999, 2007).As discussed earlier, under steady-state conditions, the removal rate of acompound C may be expressed as ln(C/Co) = −k′t, with the pseudo-firstorder rate constant k′ expressed as:k′ =kOH,C ΦH2O fH2O Ia∑kOH,Si [Si](2.12 revisited)where kOH,C is the second-order rate constants for the reaction of OH witha target contaminant C, Φ is the quantum yield for 185 nm photolysis ofH2O, fH2O is the fraction of absorbed photons that are absorbed by H2O,Ia is the 185 nm absorption rate per unit volume, and∑kOH,S [Si] repre-sents the scavenging term. Scavenging is represented as a sum of individualcontributions from compounds Si, though these individual contributions areinaccessible and only the sum is typically quantified experimentally when704.3. Use of Reference Materialsnaturally occurring DOM is used.In Equation 2.12 , the term fH2O may be significant if a substantial fractionof 185 nm photons are not absorbed by water but by other solutes. Thisterm may be evaluated if the solution obeys the Beer-Lambert law and thecomponent absorption coefficients are known. In this case, if the molar ab-sorbance coefficient associated with DOC is known and the total absorbanceof the solution is known, then the absorbance of water may be determinedfrom αH2O = αtot − αDOC , and fH2O = αH2O/αtot. The value of αDOC maybe determined for any value of DOC, provided the absorption coefficientDOC is known, via the expression αDOC = DOC [DOC]. At the wavelengthof 254 nm, the absorption coefficient is referred to as Specific UV Absorbanceor SUVA. This term is widely used to characterize DOM at 254 nm, withtypical values of 1 to 4 L mg−1 m−1 as C. Comparable values at 185 nm havenot been reported to date.4.3 Use of Reference MaterialsDue to the variability and site-specific characteristics of organic matter,studies of the influence of DOM on treatment benefit from standardizedmaterials. Use of such materials allows researchers to better compare re-sults and more accurately verify the work of others. Such standard refer-ence materials are available from the International Humic Substances Soci-ety (IHSS). This work uses two IHSS reference materials, namely SuwanneeRiver and Nordic Reservoir NOM, obtained from preparative scale separa-tion of the DOM from water, as described in detail by Serkiz and Perdue(1990) and Sun et al. (1995).Stock solutions were made using each of the IHSS reference materials, pre-pared by quantitatively dissolving the freeze dried samples, filtration of so-lutions via pre-washed 0.45 µm filters, and pH neutralization with H2SO4.Determination of the stock solution DOC values was performed by analysis714.3. Use of Reference Materialsof dilutions via TOC Analyzer calibrated with potassium hydrogen phtha-late (KHP) standards. Aliquots of the stock solutions were then used toprepare a set of four solutions with a concentration of DOC in the range of1 to 10 mg L−1 as C, also verified by TOC Analyzer. The final concentra-tion of Cl– and SO 2–4 was calculated to be < 1 mg L−1 and confirmed by ionchromatography. The solutions were then spiked with the probe compoundcarbamazepine (∼ 0.25 µM), transferred to a 1.0 cm path length fused silicacell, and irradiated using the 185 nm collimated beam. Discrete irradiationswere performed at exposure times scaled to produce at least one natural logunit removal of the probe compound in the solution with highest DOC atthe longest exposure time, found to require approximately 5 min. All irra-diations were replicated in triplicate.Experimental results for 185 nm irradiations of NOM solutions are displayedin Figures 4.1 and 4.2, for Suwannee River and Nordic Reservoir NOM re-spectively. Excellent linearity is observed in all cases, and the correspondingpseudo-first order rate constants are tabulated in Table 4.1.Table 4.1: The removal rate of probe for varying concentrationand source of DOM in 185 nm regimeReference DOM [DOC] k′(mg L−1 as C) (min−1)Suwannee River 3.0 0.845± 0.0185.0 0.555± 0.0077.5 0.356± 0.00410.0 0.250± 0.004Nordic Reservoir 1.8 1.09± 0.042.7 0.82± 0.025.2 0.416± 0.0079.8 0.202± 0.001Reference materials from the IHSS.A plot of k′ vs. the reciprocal of the DOC is shown in Figure 4.3, andreveals a linear behaviour as expected from Equation 2.12. The absence of724.3. Use of Reference Materials0 1 2 3 4 5 6−5−4−3−2−103.0 mg L−15.0 mg L−17.5 mg L−110.0 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 4.1: Suwannee River NOM in 185 nm regime. All concentrations arein units of mg L−1 as C. [CBZ]o ∼ 0.25 µM. All solutions at pH 7.significant curvature suggests that the influence of fH2O may be negligibleunder the conditions tested. Furthermore, the ratio of the slopes for theplots 4.3, obtained by linear regression, yields a value of 1.10± 0.07. If theeffect of 185 nm absorbance is indeed negligible, this result suggests that theOH reactivities of the two DOM reference material tested (i.e. kOH,DOC)are comparable despite originating from distinct and disparate sources.734.3. Use of Reference Materials0 1 2 3 4 5 6−4−3−2−101.8 mg L−12.7 mg L−15.2 mg L−19.8 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 4.2: Nordic NOM in 185 nm regime. All concentrations are in unitsof mg L−1 as C. [CBZ]o ∼ 0.25 µM. All solutions at pH 7.0 0.1 0.2 0.3 0.400.20.40.60.811/DOC (L mg−1 as C)k′(min−1)Figure 4.3: Removal rate of carbamazepine probe with DOM, SuwanneeRiver ( ), Nordic Reservoir ( ) NOM.744.4. Pure Substances as Model Organic Matter4.4 Pure Substances as Model Organic MatterDue to the complex composition of naturally occurring organic matter, theuse of pure compounds facilitate the study of the 185 nm AOP. Subsequentstudies involved well characterized compounds used as a radical scavengerand model DOM.Three pure compounds were selected for study. These are listed in theTable 4.2 with the corresponding second-order OH rate constants. Com-pounds were selected based on (1) known OH rate constant, (2) negligibleabsorption of 254 nm radiation, (3) high aqueous solubility, (4) relativelylow vapour pressure. The last two criteria result in low volatility. The threecompounds used are miscible with water.Table 4.2: Selected pure compounds used as model organic matterin studies of 185 nm regimeCompound (S) Formula kOH,S(M−1 s−1)tert-butanol (CH3)3COH 6.0× 108(2-methyl-2-propanol)methanol CH3OH 9.7× 108acetone CH3COCH3 1.1× 108(2-propanone)Rate constants taken from Buxton et al. (1988).Stock solutions of the three compounds, tert-butanol, methanol, and acetonewere prepared gravimetrically from neat HPLC grade reagents and ultrapurewater, with the resulting concentrations measured by TOC Analyzer uponsuitable dilution. Aliquots of the stock solution and spikes of the probe com-pound carbamazepine were then used to assemble solutions for irradiation.Concentrations of the model compounds, [S], were selected to satisfy the754.4. Pure Substances as Model Organic Mattercriteria kOH,CBZ [CBZ]  kOH,S [S]. Discreet irradiations were performedas before for exposure times sufficient to induce at least one natural log unitdecrease in the probe compound concentration (typically 1 to 10 min). Theresults are plotted in Figures 4.4, 4.5, and 4.6.0 2 4 6 8 10 12−4−3−2−105 mg L−110 mg L−115 mg L−120 mg L−125 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 4.4: Tert-Butanol ((CH3)3COH) in 185 nm regime. All concentra-tions are in units of mg L−1 as C.764.4. Pure Substances as Model Organic Matter0 2 4 6 8 10 12−3−2−101.4 mg L−12.1 mg L−13.0 mg L−13.8 mg L−14.7 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 4.5: Methanol (CH3OH) in 185 nm regime. All concentrations arein units of mg L−1 as C.0 2 4 6 8 10−3−2−1060 mg L−1 120 mg L−1240 mg L−1480 mg L−11020 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 4.6: Acetone (CH3COCH3) in 185 nm regime. All concentrations arein units of mg L−1 as C.774.4. Pure Substances as Model Organic Matter0 2 4 6 8 1000.10.20.30.40.50.60.70.80.9k′(min−1)0 5 10 151/[S] (mM−1)0 0.2 0.4 0.6Figure 4.7: Removal rate of probe with pure compounds as model DOM:tert-butanol ( ), methanol ( ), acetone ( ).All plots reveal high linearity (R2 > 0.95) for all values of [S] and the valuesof k′ are displayed in Table 4.3. Plots of k′ vs. 1/[S], in Figure 4.7, revealstrong linearity, supporting the model expressed by Equation 2.12. In thecase of methanol, in dilute solutions the molar absorption coefficient () at185 nm has been reported to be less than 10 M−1 cm−1 (Weeks et al., 1963).For the concentrations of methanol used in these experiments, the fractionfH2O in Equation 2.12 is calculated to be approximately unity. The samecondition is expected for tert-butanol, given structural similarities betweenthe two alcohols. However, the presence of a pi-bond in the carbonyl groupof acetone suggests that absorption at 185 nm may be significant. Further-more, given the lower OH reactivity of acetone, a higher concentration wasrequired in these experiments to ensure the steady-state condition was satis-fied (i.e. kOH,CBZ [CBZ] kOH,S [S]). Thus, a value of fH2O less than unityis expected to yield detectable curvature in the corresponding plot displayedin Figure 4.7. Insufficient resolution is available for the conditions used to784.4. Pure Substances as Model Organic MatterTable 4.3: The removal rate of probe for pure scavengers tert-butanol, methanol, and acetone in 185 nm regimeScavenger Concentration k′(mg L−1 as C) (mM) (min−1)(CH3)3COH 5.0 0.104 0.479± 0.00610.0 0.208 0.269± 0.00315.0 0.312 0.181± 0.00220.0 0.416 0.138± 0.00120.0 0.521 0.120± 0.001CH3OH 1.0 0.083 0.357± 0.0042.0 0.167 0.202± 0.0023.0 0.250 0.145± 0.0024.0 0.333 0.116± 0.0015.0 0.416 0.093± 0.001CH3COCH3 60 1.67 0.814± 0.019120 3.33 0.503± 0.010240 6.66 0.319± 0.003480 13.32 0.189± 0.0011020 28.31 0.118± 0.001Uncertainties represent the standard errors in the slopes calculatedby linear regression.confirm the presence of such curvature. However, a comparison of the ratiosof k′ vs. 1/[S] from the three plots is useful. Modifying Equation 2.12 fora pure substance as scavenger S and incorporating all other terms in theparameter Ω:k′ =kOH,C ΦH2O fH2O IakOH,S [S]=Ω[S](4.1)794.4. Pure Substances as Model Organic MatterIn the case of fH2O ≈ 1, as is expected for tert-butanol and methanol, theratio of the Ω values for two systems using pure scavengers S should ap-proximate the inverse ratio of the corresponding rate constants kOH,S :k′S1k′S2=ΩS1ΩS2≈ kOH,S2kOH,S1(4.2)Linear regression analysis of the data displayed in Table 4.3 and Figure4.7 yield the values of ΩtBuOH = 0.047 ± 0.001 mM min−1 and ΩMeOH =0.027 ± 0.001 mM min−1 for tert-butanol and methanol respectively. Theexperimentally determined ratio is thus:ΩtBuOHΩMeOH≈ kOH,MeOHkOH,tBuOH= 1.7± 0.1 (4.3)The ratio evaluated from rate constants found in the widely cited compila-tion of Buxton et al. (1988) gives:kOH,MeOHkOH,tBuOH=9.7× 108 M−1 s−16.0× 108 M−1 s−1 = 1.6 (4.4)The two values are in relatively good agreement, supporting the validity ofthe model expressed by Equation 2.12 and the assumption that fH2O ≈ 1for solutions of tert-butanol or methanol at the concentrations used.However, in the case of acetone, ΩAcetone = 1.21±0.07 mM min−1 was found,804.4. Pure Substances as Model Organic Matterresulting in ratios between acetone and tert-butanol and methanol that de-viate substantially from those calculated from rate constants. The ratiosΩ(Acetone : MeOH) and Ω(Acetone : tBuOH) are both approximately fivefold greater than the corresponding inverse ratios of the rate constants:ΩAcetoneΩMeOH= 45kOH,MeOHkOH,Acetone= 8.8 (4.5)ΩAcetoneΩtBuOH= 26kOH,tBuOHkOH,Acetone= 5.5 (4.6)It is likely that the assumption of fH2O ≈ 1 is not valid for acetone dueto the possibility of strong absorption at 185 nm due to the pi-bond of thecarbonyl group of acetone. This would result in fH2O < 1, which wouldinduce a curvature to the plot in Figure 4.7 that deviates from the straightline towards the ordinate. This is opposite to the slight curvature that ap-pears towards the abscissa. While it is possible the observation may bedue to an experimental error, an impurity of acetone with a much lowerOH reactivity would tend to produce a similar effect. Such a contaminantmay be oxalic acid or oxalate (pKa = 4.4), with reaction rate constantskOH,S = 1.4× 106 and 7.7× 106 M−1 s−1 respectively (Buxton et al., 1988).Oxalic acid/oxalate is reported to be an oxidative degradation product ofacetone (Stefan and Bolton, 1999; Stefan et al., 1996), which may have beenpresent in the neat solution at the start of irradiations. The irradiated solu-tions were unbuffered and any accidental drop in pH, while not having beendetected, would have amplified the effect of such an impurity. A reexamina-tion of acetone in the 185 nm regime would be useful. Regardless, the likely814.5. Estimation of 185 nm Incident Fluence Ratenon-transparency of acetone make it less ideal as a model scavenger in somecases.4.5 Estimation of 185 nm Incident Fluence RateIn the absence of a convenient chemical actinometer for the determinationof the incident fluence rate at 185 nm, Io, the experimental results obtainedabove for tert-butanol and methanol may be used to estimate its value. Acomparison of its magnitude relative to the incident fluence rate at 254 nm,determined earlier by KI-KIO3 actinometry, will yield an important param-eter of low pressure mercury lamps.Note that for non-absorbing pure scavengers S, a plot of the observed rateconstant k′ vs. 1/[S] yields a slope previously referred to as Ω. From Equa-tion 4.1, the expression for ΩS is given by:ΩS =kOH,CkOH,SΦH2O fH2OIa (4.7)All the terms of the Equation 4.7 are known or measurable except for Ia, therate of 185 nm photon absorption per unit volume. Thus, the experimen-tally determined values of ΩS , known values of kOH,C for carbamazepine,and kOH,S for tert-butanol and methanol, ΦH2O = 0.3, and the assumptionthat fH2O ≈ 1, allow the calculation of Ia in units of M s−1. This valuecan then be converted to a fluence rate using the cross sectional area andvolume of the irradiation cell, and the molar photon energy at 185 nm of647 kJ mol−1.Using an internal diameter of 1.90 cm and path length of 1.00 cm to deter-824.5. Estimation of 185 nm Incident Fluence Ratemine the area and volume of the irradiation cell (neglecting the volume ofthe miniature stir bar), calculations of the 185 nm fluence rate Io,185 yieldthe following values:Io,185 = 0.15± 0.01 mW cm−2 (via tert− butanol) (4.8)Io,185 = 0.14± 0.01 mW cm−2 (via methanol) (4.9)Note the values from the different scavengers are in agreement. The valuefor the incident fluence rate at 254 nm for the same apparatus at the sameposition was determined using the KI-KIO3 actinometer to be:Io,254 = 0.87± 0.03 mW cm−2 (via KI−KIO3) (4.10)as described in Chapter 2 in the section starting on page 48. While thisvalue was obtained using a 2.0 cm path length cell, the liquid surface wasat the same position as that of the 1.0 cm path length cell used in thesedeterminations. The ratio of incident fluence rates at 185 to 254 nm is thus:Io,185 : Io,254 = 0.16± 0.01 (4.11)834.6. SummaryThis value is within the range of 0.12 to 0.34 reported by Barnes (1960)and by Johnson (1971), who used different types of low pressure mercurylamps. Furthermore, according to those studies, the ratio may be increasedby optimization of the diameter, current, temperature of lamps dedicatedfor 185 nm output. Additionally, the replacement of natural quartz withfused silica for the lamp material will further increase output.4.6 SummaryFrom these studies it can be seen that the primary effect of DOM on the185 nm AOP occurs via the scavenging of OH, resulting in an inverse depen-dency of the observed psuedo-first order rate constant with the concentra-tion. The two reference materials used, Suwannee River and Nordic Reser-voir NOM obtained from the IHSS, induce a similar effect on the 185 nmAOP, suggesting that the aggregate reactivity, and potentially the absorp-tion at 185 nm are comparable for the two materials.Furthermore, additional support for a model expressed by Equation 2.12 isprovided by the use of pure compounds tert-butanol and methanol as modelDOM. Indirect experimental evidence supports the assumption that neitherof these compounds are significant photon absorbers at 185 nm. However,the significance of photon absorption to the process in general remains un-clear and requires further investigation.Lastly, the use of a suitable probe compound (carbamazepine) and modelscavenger (tert-butanol or methanol) allowed the estimation of 185 nm flu-ence rate in-lieu of a convenient chemical actinometer. This determinationthus allowed the subsequent estimation of the 185:254 nm ratio of fluencerates emitted by the particular low pressure mercury lamp used in the colli-mated beam. A value of 16± 1% was obtained in agreement with literature844.6. Summaryvalues for low pressure mercury lamps. This technique was possible becauseboth the probe compound and the model scavengers used are effectivelyphotochemically inert to 254 nm radiation, and of negligible absorbance at185 nm at the concentrations used. However, as this method is tedious, aconvenient chemical actinometer remains highly desirable. A comparison ofsuch an actinometer with this kinetic method will be insightful.85Chapter 5Chloride5.1 Chloride in Natural WatersThe major inorganic solutes in surface waters include the cations Ca2+,Mg2+, Na+, and K+, as well as the anions SO 2–4 , Cl–, HCO –3 , and CO2–3 , withother ionic species present at concentrations typically below 1 mg L−1 (Wet-zel, 2001). The typical concentration of Cl– in surface waters ranges from1 to 100 mg L−1, with levels in more pristine waters commonly in the rangeof 5 to 10 mg L−1 (Livingstone, 1963). The actual concentration dependsstrongly on the proximity of the corresponding watershed to the marine en-vironment, as well as its underlying geology (Drever, 1988). Concentrationsof Cl– much higher than 100 mg L−1 may occur in water bodies heavily im-pacted by human activities, such as extensive irrigation, urban runoff, andindustrial pollution. A study of the composition of the Rhine river by Zor-bist and Stumm (1981) showed that the Cl– concentration was 1.1 mg L−1leaving the Swiss Alps, and 178.2 mg L−1 crossing the border from Germanyto the Netherlands. Using statistical techniques and historical data datingback to 1854, it was shown that more than 90% of the Cl– in the lower Rhinewas of anthropogenic origin. A maximum Cl– concentration of 250 mg L−1is recommended by the WHO (2004), and is emulated by Canadian andother national standards. This recommendation is not based on health ef-fects but rather aesthetics of the detectable salty taste of higher Cl– levels.High Cl– levels do not always impart a salty taste, and Cl– levels as highas 1000 mg L−1 may impart no salty taste when the dominant cations areCa2+ and Mg2+ rather than Na+ (APHA, 2012). The results of studies onthe involvement of Cl– in the 185 nm AOP are presented here.865.2. Impact of Chloride on AOPs5.2 Impact of Chloride on AOPsBased on the literature of advanced oxidation processes involving ozone orUV with H2O2, the influence of Cl– on process efficiency is not generally con-sidered and assumed to be negligible (Crittenden et al., 1999; Von Gunten,2003). The radiation chemistry literature, however, shows that the aqueousradical chemistry of chloride is complex, with certain fundamental details indispute.Pulse radiolysis studies of aqueous chloride solutions by Anbar and Thomas(1964) found a transient species with a peak absorbance at 340 nm identifiedas the dichloride radical anion Cl –2 . The formation of Cl–2 was observed atCl– concentrations of 0.2 to 10 mM and pH < 3. At pH 7, Cl –2 was not ob-served unless Cl– concentrations were elevated above 0.1 M. In the presenceof OH scavengers CH3OH or K4[Fe(CN)6], the extent of Cl–2 formation wasfound to be inhibited, but not its rate of decay. From these observationsit was deduced that the formation of Cl –2 involves a two step process, firstthe reaction of OH with Cl– to form Cl , with subsequent reaction of Clwith Cl– to form Cl –2 . Irradiation of neutral solutions of H2O2 and Cl–established that Cl –2 did not form, implying that OH does not react withCl– at neutral pH and that the Cl –2 observed at high Cl– concentrationsoccurred in the spur region where pulse radiolysis energy is deposited andlocal radical concentrations are several orders of magnitude greater as com-pared to the bulk. Thus the reaction of OH with Cl– to form Cl –2 requiresthe presence of H+.Jayson et al. (1973) investigated the pulse radiolysis of solutions containingNaCl at either neutral pH or acidified to pH< 3 using HClO4. Absorptionof a transient at 240 nm was seen to decrease at the same rate as the ab-sorption increased at 340 nm. Experimental evidence supported the identityof the species absorbing at 240 nm to be OH generated from the radiolysisof H2O, while the species absorbing at 340 nm was proposed to be the Cl–2reported by Anbar and Thomas (1964). Their observations were explained875.2. Impact of Chloride on AOPsby the net conversion of OH to Cl –2 by the following mechanism:OH + Cl– −−⇀↽− HOCl – (5.1)HOCl – + H+ −−⇀↽− Cl + H2O (5.2)Cl– + Cl −−⇀↽− Cl –2 (5.3)provided that the forward reaction of 5.3 is fast compared to the genera-tion of Cl by 5.1 and 5.2 and that the equilibrium of 5.1 lies to the left.The overall rate of Cl –2 formation was found to depend on the concentra-tions of Cl–, H+ and ionic strength in a manner consistent with the abovemechanism 5.1 to 5.3. Rate constants for all of the above forward andbackward reactions, as well as equilibrium constants (K) were determined.The equilibrium constants were reported as K1 = 0.70 M−1, K2 = 1.6× 107and K3 = 1.9× 105 M−1 (uncertainties omitted). However, this work didnot take into account the contribution of Cl to the absorption at 340 nm,though it has been reported by others that Cl has a peak absorbance at320 nm and a molar absorption coefficient at 340 nm comparable to thatof Cl –2 (Buxton et al., 1998; Kla¨ning and Wolff, 1985; Nagarajan and Fes-senden, 1985; Treinin and Hayon, 1975).Gilbert et al. (1988) used electron spin resonance spectroscopy to study theproducts of organic molecules exposed to Cl and Cl –2 generated photochem-ically. Despite the equilibrium 5.3, when [Cl–] is sufficiently low, significantCl reaction occurs via addition to unsaturated carbon bonds, H-abstraction,885.2. Impact of Chloride on AOPsor electron transfer at rates near the diffusion limit. It was deduced that Cl –2is of relatively low reactivity compared to Cl , consistent with the findingsof Hasegawa and Neta (1978). Furthermore, it was postulated from theseand previous findings by the same group that Cl is of higher reactivity andlower selectivity relative to OH for the alcohols and organic acids tested.Lastly, while OH reactions tend to favour H-abstraction from the α-carbonposition, reactions with Cl tend to favour attack at the hydroxyl group viaelectron transfer.Buxton et al. (1998) used both pulse radiolysis and laser flash photoly-sis at 193 nm to investigate equilibrium 5.3. They obtained a value of1.4× 105 M−1 in relatively close agreement with that of Jayson et al. (1973),the discrepancy attributed possibly to the previous study not taking Cl intoaccount when measuring Cl –2 . Deviations of reactions 5.3 from equilibriumwere studied by the presence of the organic scavenger tert-butanol. Thereactions of Cl and Cl –2 with water:Cl + H2O −→ H+ + HOCl – (5.4)Cl –2 + H2O −→ H+ + HOCl – + Cl– (5.5)were found to have rate constants of 2.5× 105 s−1 (5.4) and 1.3× 103 s−1(5.5). The reaction 5.4 is equivalent to the reverse reaction of equilibrium5.2 introduced by Jayson et al. (1973). In addition to the scavenging ofCl by Cl– to form Cl –2 , reactions with organic solutes and water itself arepotentially significant sinks of Cl .895.3. Chloride in the 254 nm - H2O2 RegimeGiven the high reactivity of Cl , a subsequent study by Buxton et al. (2000)used laser flash photolysis of chloroacetone at 243 nm to generate Cl in theabsence of Cl– and equilibrium 5.3. With temperature controlled photolysiscells, this approach was used to measure the second-order reaction rate con-stants for Cl with several simple organic compounds and inorganic anionsat 25 ◦C, as well activation energies by varying the temperature from 5 to35 ◦C. Relevant details from this study will be discussed below in relationto findings from this work.As mentioned in Chapter 1, the Cl– anion is transparent to UV radiation atwavelengths above 200 nm at the concentrations found in solution. In thevicinity of 185 nm, a strong absorption band occurs for aqueous Cl– due to acharge-transfer-to-solvent mechanism discussed (Blandamer and Fox, 1970;Fox et al., 1978). As previously mentioned in Table 1.2, the molar absorptioncoefficient (185) for Cl– at 185 nm has been reported as 3800±300 M−1 cm−1,and results in photodissociation to the radical Cl and a solvated electrone –aq with a quantum yield (ΦCl–) of 0.43± 0.02 (Dainton and Fowles, 1965).This implies that at [Cl–] between 15 and 20 mg L−1, more than half the185 nm photons absorbed by the water matrix are absorbed by Cl– ratherthan H2O. Thus, in the 185 nm regime, the presence of Cl– results in theformation of Cl and decreased H2O photolysis, possibly shifting the mainreactive species from OH to Cl . Experimental confirmation of this is oneof the main objectives of the work presented in this chapter.5.3 Chloride in the 254 nm - H2O2 RegimeThe influence of Cl– on the 254 nm photolysis of H2O2 was investigated us-ing the UV/H2O2 collimated beam apparatus. Carbamazepine was used asa probe compound and the DOM used was either tert-butanol or SuwanneeRiver NOM reference material obtained from the IHSS. The Cl– concentra-tion was varied between 0 and 100 mg L−1 using NaCl. Solution pH was905.3. Chloride in the 254 nm - H2O2 Regimeneutralized using NaOH. The dose of H2O2 applied was 7 mg L−1. Darkcontrols were used to verify the absence of significant thermal reactions. Allsolutions were reproduced in triplicate and irradiated. Following irradia-tions, residual H2O2 was quenched with a 10 µL droplet of approximately100 mg L−1 bovine catalase before analysis of carbamazepine by HPLC.Results obtained using tert-butanol at 7 mg L−1 as C are plotted in Figure5.1, and reveal no detectable difference in degradation rate for any of thefour Cl– levels used.0 5 10 15 20 25 30 35−0.8−0.6−0.4−0.20Time (min)ln([CBZ] t/[CBZ] o)Figure 5.1: Tert-Butanol in the 254 nm - H2O2 regime and influence ofchloride. [H2O2] =7 mg L−1. [tBuOH] = 7 mg L−1 as C. [CBZ]o ' 0.25 µM.NaCl used as source of Cl– . [Cl−] < 1 mg L−1 ( ), [Cl−] = 25 mg L−1( ), [Cl−] = 40 mg L−1 ( ), [Cl−] = 100 mg L−1 ( ).Subsequently, Suwannee River DOM at 7 mg L−1 as C was tested, and theresults plotted in Figure 5.2. Again, the observed degradation rates for 0and 100 mg L−1 of Cl– show no statistically significant difference. Thus,under the conditions used and an environmentally meaningful range of Cl–concentration, any influence of Cl– on removal kinetics is undetectable.915.4. Chloride in the 185 nm Regime0 5 10 15 20 25 30 35−0.6−0.4−0.20Time (min)ln([CBZ] t/[CBZ] o)Figure 5.2: Suwannee River NOM in the 254 nm - H2O2 regime and influenceof chloride. [H2O2] =7 mg L−1. [DOC] = 7 mg L−1 as C. [CBZ]o ' 0.25 µM.All solutions prepared at pH 7. NaCl used as source of Cl– . [Cl−] <1 mg L−1 ( ), [Cl−] = 100 mg L−1 ( ).5.4 Chloride in the 185 nm RegimeWhen these same experiments are repeated in the 185 nm regime, a signifi-cant effect is observed, the results of which are plotted in Figure 5.3 whithtert-butanol used as DOM, and in Figure 5.4 with Suwannee River NOM.When tert-butanol is used, an increase in [Cl–] results in a pronounced in-crease in the rate of degradation of the probe carbamazepine. In contrast,when tert-butanol is replaced by Suwannee River NOM, the opposite effectis observed, with an increase in [Cl–] corresponding to a decrease in the rateof probe degradation.These observations may be explained by at least three independent aspects925.4. Chloride in the 185 nm Regime0 2 4 6 8 10−5−4−3−2−10< 1 mg L−110 mg L−125 mg L−140 mg L−1100 mg L−1[Cl– ][Cl−]Time (min)ln([CBZ] t/[CBZ] o)Figure 5.3: Tert-Butanol ((CH3)3COH) in the 185 nm regime and influ-ence of chloride, using carbamazepine (CBZ) as a probe. [(CH3)3COH] =7 mg L−1 as C. [CBZ]o ' 0.25 µM. NaCl used as source of Cl– . Arrow in-dicates trend in observed degradation rate resulting from increased chlorideconcentration.935.4. Chloride in the 185 nm Regime0 1 2 3 4 5 6−2−1.5−1−0.50[Cl−]Time (min)ln([CBZ] t/[CBZ] o)Figure 5.4: Suwannee River NOM in the 185 nm regime and influence ofchloride, using carbamazepine (CBZ) as a probe. [DOC] = 8 mg L−1 as C.[CBZ]o ' 0.25 µM. NaCl used as source of Cl– . [Cl−] < 1 mg L−1 ( ),[Cl−] = 25 mg L−1 ( ), [Cl−] = 40 mg L−1 ( ), [Cl−] = 100 mg L−1( ). Arrow indicates trend in the observed probe degradation rate result-ing from increased chloride concentration.945.4. Chloride in the 185 nm Regimerelated to the increase in [Cl–]. First, increased [Cl–] results in a shift in185 nm photon absorbance away from water and toward Cl–, with corre-sponding changes in the generation rates of OH and Cl . A second effect isthe difference in OH and Cl reactivities towards the matrix, as measuredby the second-order reaction rate constants with the major scavengers. Thegreater the reactivity, the lower the radical concentration should be understeady-state conditions. Lastly, differences in the reactivities of OH andCl with the probe or target compound itself will contribute to changes inthe observed degradation rate. An additional complication is the possibilitythat a portion of the Cl may be converted to OH (discussed below), thoughthe converse is not expected to occur at circumneutral pH based on evidencepresented earlier.In general, using an expression based on Equation 2.13:ln(C/Co) = −k′ t = −(kOH,C [ OH]ss + kCl,C [Cl ]ss) t (5.6)where k′ is the experimentally observed pseudo first-order rate constant.This expression can be rewritten in terms of the quantum yields Φ, thefraction of absorbed photons f , the rate of 185 nm photon absorption perunit volume Ia, the second-order rate constant for the radical-scavenger re-actions and the concentration of scavenger [S]. Grouping common terms,an expression for k′ is given by:k′ =(kOH,CkOH,SΦH2OfH2O +kCl,CkCl,SΦCl–fCl–)Ia[S](5.7)955.4. Chloride in the 185 nm RegimeInspection of this expression reveals that the bracketed term controls thevalue of k′ for a specific compound via four independent second-order rateconstants. Each term in the sum will contribute to k′ if both OH and Clare present, and the contribution by each radical will further depend onthe relative reactivities of the probe and the scavenger. It should be notedthat the two quantum yields are of comparable magnitude, ΦH2O ' 0.3, andΦCl– ' 0.4. Furthermore, the rate constants involving Cl may be viewedas effective rate constants which incorporate the possible equilibrium withCl –2 . The second-order rate constants for OH and Cl with tert-butanolare essentially equal at 6× 108 M−1 s−1 (Buxton et al., 2000). In general,Cl is expected to be more reactive than OH, and increased reactivity ofcarbamazepine with Cl is consistent with the observations.In the absence of pulse radiolysis or laser flash photolysis capabilities, wherebyradical species may be directly detected and distinguished spectrophotomet-rically, indirect studies were conducted using probe compounds. The carefuluse of probe compounds may allow kinetic and mechanistic details to be de-duced. A probe may be selected such that one of the two terms comprisingk′ in the right hand side of equation 5.7 is much smaller than the other.Ideally, two such probes would be available in order to alternately eliminateone of the terms experimentally. Based on the evidence from Figures 5.3and 5.4, carbamazepine appears to be reactive to both OH and Cl , withlikely greater reactivity with the latter than the former.No compound, both soluble in water and easily quantifiable by HPLC, isknown to be highly reactive to Cl while negligibly reactive with OH. How-ever, a probe for the converse situation is suggested by the literature. Dur-ing studies of solar processes in surface waters, it was suggested by Nowelland Hoigne´ (1992) that nitrobenzene would be relatively nonreactive to Clbased on structure-reactivity arguments and evidence from gas-phase reac-tions. Support for this assumption was provided by UV photolysis of aque-ous chlorine at pH 1 containing two probe compounds, 1-chlorobutane and965.4. Chloride in the 185 nm Regimenitrobenzene. Under these conditions, Cl would be formed in the absence ofany Cl–, and it was reported that the degradation of 1-chlorobutane was sub-sequently observed without detectable removal of nitrobenzene. The second-order reaction rate constant of nitrobenzene with OH is 3.0× 109 M−1 s−1(Buxton et al., 1988).Thus, the use of nitrobenzene as a probe in the 185 nm regime should pro-duce results discrepant with those observed in Figures 5.3 and 5.4 obtainedusing carbamazepine. Assuming equation 5.7 to be valid, the use of ni-trobenzene as a probe, with either tert-butanol or Suwannee River NOM asthe scavenger, should result in a decrease of k′ with increasing [Cl–] in bothcases. This is because as [Cl–] increases, the first term of equation 5.7 de-creases due to the decreased absorption of photons by water (fH2O) resultingin a lower [ OH]ss, while a presumed increase in Cl is ineffective due to alow kCl,NB value. Experimental observations confirm the prediction that k′decreases in both cases, as plotted in Figures 5.5 and 5.6.975.4. Chloride in the 185 nm Regime0 2 4 6 8 10 12 14−1.5−1−0.50[Cl−]Time (min)ln([NB] t/[NB] o)Figure 5.5: Tert-Butanol ((CH3)3COH) in the 185 nm regime and influenceof chloride, using nitrobenzene (NB) as a probe. [(CH3)3COH] = 8 mg L−1as C. [NB]o ' 1 µM. NaCl used as a source of Cl–. [Cl–] < 1 mg L−1 ( ),[Cl–] = 25 mg L−1 ( ), [Cl–] = 40 mg L−1 ( ), [Cl–] = 100 mg L−1 ( ).Arrow indicates trend in observed degradation rate resulting from increasedchloride concentration.985.4. Chloride in the 185 nm Regime0 2 4 6 8 10 12 14−1.5−1−0.50[Cl−]Time (min)ln([NB] t/[NB] o)Figure 5.6: Suwannee River NOM in the 185 nm regime and influence ofchloride, using nitrobenzene (NB) as a probe. [DOC] = 8 mg L−1 as C.[NB]o ' 1 µM. NaCl used as a source of Cl–. [Cl–] < 1 mg L−1 ( ),[Cl–] = 25 mg L−1 ( ), [Cl–] = 40 mg L−1 ( ), [Cl–] = 100 mg L−1 ( ).Arrow indicates trend in observed degradation rate resulting from increasedchloride concentration.995.5. Relative Reactivity of OH and Cl5.5 Relative Reactivity of OH and ClOne of the difficulties with indirect kinetic studies of Cl– in the 185 nmregime is the absence of suitable probe compounds that will allow the in-dependent study of the OH and Cl contributions. However, either termof equation 5.7 may be isolated, regardless of the probe used, if a suitablescavenger compound S can be found such that the corresponding valuesof the second-order rate constants kOH,S and kCl,S differ by several ordersof magnitude. Very limited information on kCl,S in the aqueous phase exists.7 8 9 1078910methanoltert-butanolacetoneacetateethanolformateformic acidacetic acidformaldehydeacetaldehydelog k(•OH + S)logk(Cl•+S)Figure 5.7: Comparison of second-order rate constants of select organicsolutes with OH and Cl at 25 ◦C. Key features are the outliers, acetateand acetic acid, from the general trend. Adapted from Buxton et al. (2000)The work by Buxton et al. (2000) reports values of kCl,S for eleven smallorganic compounds. The plot of Figure 5.7 is adapted from that work andplots the values of kOH,S versus kCl,S at 25◦C on log-log scales. It can be1005.6. Evidence for Cl from Probe-Scavenger Systemsseen that fairly good agreement between most of the compounds studiedexists between the two rate constants. However, it is noted by the authorsof that work that the activation energies, that were also measured for bothOH and Cl reactions, differ and that the close correspondence observed at25 ◦C will be weaker at other temperatures. Nevertheless, while compoundssuch as methanol and tert-butanol have virtually the same rate constantsfor both radicals, the two compounds, acetone (CH3COCH3) and acetate(CH3COO–) are prominent outliers from the general trend seen in Figure5.7.5.6 Evidence for Cl from Probe-ScavengerSystemsAn explanation for the observations of the chloride effect made using theprobes carbamazepine (Figures 5.3 and 5.4) and nitrobenzene (Figures 5.5and 5.6) is provided by a model described by Equation 5.7. In the absenceof chloride and any other solutes absorbing at 185 nm (i.e. fH2O ≈ 1), onlyOH is involved and the expression is reduced to include only the first termof Equation 5.7. If the chloride level is sufficiently high (i.e. fH2O  fCl–),the contribution of Cl may dominate. Note, it is assumed thus far that anyconversion of Cl to OH is negligible.For low to moderate levels of chloride, an intermediate situation may prevailand potentially both radicals OH and Cl will contribute to an extent de-pending on relative reactivities to the probe (or target) compound and thescavenger (or matrix). In such a case, both terms of Equation 5.7 may con-tribute to the overall degradation rate. Table 5.1 displays the reaction rateconstants plotted in Figure 5.7 of the four pure substances used as modelscavengers in this work. The reaction rate ratios of the rightmost columncontains the key information used in this study.1015.6. Evidence for Cl from Probe-Scavenger SystemsTable 5.1: Comparison of OH and Cl reactivities for pure compoundsused as model organic matter in the 185 nm regimeScavenger (S) kOH,S kCl,S kOH,S/kCl,S(M−1 s−1)×10−8 (M−1 s−1)×10−8Methanol 9.7 10 0.97CH3OHtert-Butanol 6.0 6.2 0.97(CH3)3COHAcetate 0.75 37 0.02CH3COO–Acetone 1.1 < 0.05 > 22CH3COCH3Rate constants taken from Buxton et al. (1988) and Buxton et al. (2000).Since tert-butanol and methanol have essentially the same reaction rateconstant for both OH and Cl (kOH,S ∼ kCl,S) and the quantum yields arecomparable (ΦH2O ∼ ΦCl–), the effect of chloride depends largely on therelative reactivity of the probe compounds if Equation 5.7 is valid.If acetate is used as the scavenger, the second-order reactivity with Cl isapproximately 50 times greater than that for OH. Thus, when Cl– is presentunder 185 nm irradiation, the formation rate of OH will decrease, while thatof Cl increases, and the Cl formed will be consumed more efficiently by ac-etate relative to OH. This would have the effect of reducing the steady-stateconcentration of both radicals in question. In the case of carbamazepine, thesecond-order rate constants with both OH and Cl are assumed to be of thesame order of magnitude (kOH,C ∼ kCl,C). However, a shift to Cl involvesa shift to greater contribution from the second term in Equation 5.7. Sincewith acetate kCl,C/kCl,S  kOH,C/kOH,S , the observed pseudo-first orderrate constant k′ should decrease as [Cl–] increases. When nitrobenzene is1025.6. Evidence for Cl from Probe-Scavenger Systemsused as a probe, the same trend is expected, with the additional impact dueto kCl,C  kOH,C .The experimental observations are in agreement with the above description.As can be seen in Figure 5.8, the observed rate constants k′ for both probes,carbamazepine and nitrobenzene, decrease upon addition of chloride. Notethat Figure 5.8 shows, that while k′ for nitrobenzene is attenuated by thepresence of Cl–, a residual k′ is measured and attributed to OH entirely.This is expected based on values of 185 for Cl– both reported in the litera-ture and measured in this work (see Section 5.10). At [Cl–] = 100 mg L−1,the fraction of 185 nm photons absorbed by water is approximately 25%.The k′ for nitrobenzene, however, decreases by approximately 50%, suggest-ing the possibility that some Cl does indeed convert to OH. Attempts tomodify the model to account for such a conversion, using the reactions ofCl and Cl –2 with water (Equations 5.4 and 5.5 respectively), have beenunsuccessful and indicate key mechanistic aspects are missing.If acetone is used as the scavenger, the reverse situation is induced comparedto that of acetate. Acetone reactivity with Cl is low relative to OH, withOH being at least 22 times more reactive with acetone. In this case, theeffect of chloride will depend more strongly on the reactivity of the probeswith Cl (kCl,C). When chloride is sufficiently high and the Cl formationrate increases, the lower scavenging of Cl implies that those compoundswith higher reactivities with Cl will degrade faster than when chloride isabsent. For compounds of lower reactivity with Cl , lower degradation ratesare expected. Thus, with acetone as scavenger, the addition of chlorideshould cause the degradation rate of carbamazepine to increase, while thatof nitrobenzene should decrease. This description is consistent with experi-mental observations, as can be seen from Figure 5.9, where k′ increases forcarbamazepine and decreases slightly for nitrobenzene. In the absence ofactual values of kCl,C for these probes, the description must remain semi-quantitative. However, if nitrobenzene is truly nonreactive with Cl , as sug-gested by Nowell and Hoigne´ (1992), then the limited removal observed may1035.6. Evidence for Cl from Probe-Scavenger Systemsbe due to another mechanism, such as the conversion of Cl to OH discussedabove. Since direct photolysis at 254 nm is negligible, and volatilization iseliminated by using a sealed cell, a radical conversion process provides aplausible explanation.Though conducted using pure compounds as scavengers, the results fromacetate and acetone demonstrate that the net effect of chloride dependson several factors. Naturally occurring DOM contains carboxylic acid andketone functional groups that may influence Cl reactivity. While the mech-anisms of Cl are not well understood in the aqueous phase, it has been sug-gested by Gilbert et al. (1988) that Cl oxidation is more likely to occur viaelectron transfer at hydroxyl groups rather than H-abstraction commonlyobserved with OH. Thus, the chemical details of site-specific NOM mayhave a substantial effect on observed degradation rates if the involvement ofCl is significant. Furthermore, while pH effects were not specifically stud-ied in this work, it is worth noting that the radical reactivities of carboxylicacids, with both OH and Cl , are strongly influenced by pH, as can be seenin Figure 5.7 for acetate/acetic acid and formate/formic acid.The low reactivity of acetone with Cl was suggested by Buxton et al. (2000)as the result of the low extent of hydration of acetone. In aqueous solution,ketones react with water to form an equilibrium with the geminal diol (gem-diol). As electron transfer by Cl is preferred at the hydroxyl group, theposition of the equilibrium towards the gem-diol will be related to the re-activity of the compound. In the case of acetone, the equilibrium is far tothe non-hydrated ketone form, with only 0.1% of the molecules present asthe hydrated gem-idol. In the case of formaldehyde, 99.9% is in the formof the gem-diol (McMurry, 1996). Acetaldehyde is approximately 55% hy-drated (Kurz, 1967). The reactivities reported by Buxton et al. (2000) arein the order formaldehyde > acetaldehyde > acetone, matching the orderof hydration.1045.6. Evidence for Cl from Probe-Scavenger Systems0 1 2 3 4 5 6−3−2−10Time (min)ln(Ct/Co)Figure 5.8: Acetate (CH3COO–) system (pH ' 6) at 185 nm with bothcarbamazepine (CBZ) and nitrobenzene (NB) as probes. [CH3COO−]= 8 mg L−1 as C. [CBZ]o ' 0.25 µM. [NB]o ' 1 µM. NaCl used assource of Cl– . [Cl−] < 1 mg L−1-CBZ( ),NB( ); [Cl−] = 100 mg L−1-CBZ( ),NB( ). Arrows indicate shift in degradation rate of the probecompound due to the presence of 100 mg L−1 of Cl–.1055.6. Evidence for Cl from Probe-Scavenger Systems0 2 4 6 8 10−4−3−2−10Time (min)ln(Ct/Co)Figure 5.9: Acetone (CH3COCH3) system at 185 nm with both carba-mazepine (CBZ) and nitrobenzene (NB) as probes. [(CH3COCH3] =1000 mg L−1 as C. [CBZ]o ' 0.25 µM. [NB]o ' 1 µM. NaCl used assource of Cl– . [Cl−] < 1 mg L−1-CBZ( ),NB( ); [Cl−] = 100 mg L−1-CBZ( ),NB( ). Arrows indicate shift in degradation rate of the probecompound due to the presence of 100 mg L−1 of Cl–.1065.7. Evidence from Ionic Strength Effects5.7 Evidence from Ionic Strength EffectsThe dependence of reaction kinetics on ionic strength of the solution mayyield insight on a process. By adjusting the ionic strength, most reactionrates will change. The Debye-Hu¨ckel theory predicts that for dilute so-lutions, the reaction rate k, between ions A and B will depend on ionicstrength (µ), at 25 ◦C, according to:log(k/ko) = 1.02zAzB√µ (5.8)where zA and zB are the charges of the ions A and B, respectively (Laidler,1987). This expression predicts no effect on reaction rate when one or bothof the reactants are uncharged. However, an ionic strength effect is observedfor many reactions involving neutral reactants and an empirical extension tothe theory was made by Hu¨ckel to describe the activity of uncharged reac-tants. This leads to the expression for reactions between neutral molecules:log(k/ko) = bµ (5.9)where b is a constant for which no theory is available to predict either itsmagnitude (Laidler, 1987) nor even its sign (Moore and Pearson, 1981).Ionic strength studies are unable to establish a mechanism but may be usefulin elaborating details regarding reactants. In the absence of charged reac-tants, an ionic strength effect may yet be substantial if one of the reactantsis in equilibrium with charged species. In the case of OH reacting with the1075.7. Evidence from Ionic Strength Effectsprobe carbamazepine and the scavenger tert-butanol, no charged reactantsare involved. However, when Cl– is present and Cl is generated, the equi-librium 5.3 may induce a significant ionic strength effect on the reaction rate.A semi-quantitative study was conducted whereby the ionic strength effectwas investigated for the influence of chloride at 185 nm. Adjusting ionicstrength with NaF, the degradation rate of carbamazepine in the presenceof the scavenger tert-butanol, was studied with and without the addition of100 mg L−1 of chloride. The results in the absence of chloride are plottedin Figure 5.10, while those with chloride are plotted in Figure 5.11. Theobserved pseudo first-order rate constants k′ are normalized to the rate atlowest ionic strength k′o, and log(k′/k′o) is plotted against both µ and√µfor comparison in Figure 5.12.It can be seen from Figures 5.10, 5.11, and 5.12, that the observed rate k′decreases with increasing ionic strength. Furthermore, it is apparent thatthe effect is significantly more pronounced in the presence of chloride. Notethat in Figure 5.12, linear regression of rate data plotted both against µ(upper plot) and√µ (lower plot) indicates better agreement in the absenceof chloride when plotted against µ, while the presence of chloride shows bet-ter agreement with√µ as the abscissa. The data points are limited, butsupports the existence of a secondary salt effect when Cl– is present, likelyinvolving equilibrium 5.3. Additional studies conducted with greater reso-lution may be useful in elucidating the mechanisms involved.1085.7. Evidence from Ionic Strength Effects0 1 2 3 4 5 6 7−2.5−2−1.5−1−0.50µ[Cl–] < 1 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 5.10: Influence of ionic strength (µ) in 185 nm regime on the degrada-tion of carbamazepine (CBZ) in the presence of tert-butanol and no chloride.[CBZ]o ' 0.25 µM, [(CH3)3COH] = 7 mg L−1 as C. Ionic strength adjustedusing NaF. µ = 0.00 M ( ), 0.08 M( ), 0.16 M ( ). Arrow indicatesshift in degradation rate of the probe compound due to increasing µ.1095.7. Evidence from Ionic Strength Effects0 0.5 1 1.5 2 2.5 3−3−2−10µ[Cl–] = 100 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 5.11: Influence of ionic strength (µ) in 185 nm regime on the degrada-tion of carbamazepine (CBZ) in the presence of tert-butanol and chloride.[CBZ]o ' 0.25 µM. [(CH3)3COH] = 7 mg L−1 as C. [Cl−] = 100 mg L−1.NaCl used as source of chloride. Ionic strength adjusted using NaF.µ = 0.00 M ( ), 0.08 M( ), 0.16 M ( ). Arrow indicates shift in degra-dation rate of the probe compound due to increasing µ.1105.7. Evidence from Ionic Strength Effects0 0.05 0.1 0.15 0.2−0.6−0.5−0.4−0.3−0.2−0.100.1R2 = 0.993R2 = 0.963µ (M)log(k′ /k′ o)0 0.1 0.2 0.3 0.4 0.5−0.6−0.5−0.4−0.3−0.2−0.100.1R2 = 0.945R2 = 0.998√µ (M)log(k′ /k′ o)Figure 5.12: Influence of ionic strength (µ) on the degradation of carba-mazepine with and without chloride. Ionic strength adjusted by NaF. Dataobtained from slopes (k′) of Figures 5.10 and 5.11. Upper plot with µ asabscissa. Lower plot with√µ as abscissa. Both plots use identical or-dinates. Chloride obtained from NaCl. Tert-butanol used as a scavengerwith [(CH3)3COH] = 7 mg L−1 as C. Carbamazepine used as probe with[CBZ]o ' 0.25 µM. [Cl–] < 1 mg L−1 ( ). [Cl–] = 100 mg L−1 ( ). Errorbars derive from the standard errors calculated for k′ and the subsequentpropagation of error calculation.1115.8. Product Studies of Phenol Degradation5.8 Product Studies of Phenol DegradationAmong the reaction mechanisms associated with OH is addition to an aro-matic ring or unsaturated alkyl group. With the involvement of Cl , animportant question is whether analogous addition occurs. The potentialformation of chlorinated organics is of practical significance in the field ofdrinking water treatment. Evidence of addition of Cl to alkenes in aqueoussolution was reported by Gilbert et al. (1988) using electron spin resonancespectroscopy.The first generation degradation products of OH with phenol include cat-echol and hydroquinone, which correspond to the addition of an OH groupat the ortho and para positions, respectively. These are formed by an initialOH attack at the ortho or para positions to form a dihydroxycyclohexadi-enyl radical, or by OH attack and H-abstraction from the hydroxyl group toform a phenoxyl radical. The dihydoroxycyclohexanidenyl radicals convertto the resonance-stabilized phenoxyl radical by the elimination of H2O. Thephenoxyl radical then reacts with O2, ultimately forming the final stableproducts (Mvula et al., 2001). In this case, no Cl-adducts should be ob-served, and a product comparison of phenol degradation by OH and Cl issuggested.A series of 185 nm irradiations were conducted using solutions of phenolat approximately 1 µM, with tert-butanol at 7 mg L−1 as C, both with andwithout 100 mg L−1 of chloride. Samples were taken for HPLC analysis atirradiation times less than that required for 50% phenol degradation to pre-vent excessive oxidation of initial daughter products. The procedure usedfor HPLC analysis is identical to that of carbamazepine. Analytical gradestandards were used to identify peaks by retention time and included cate-chol, hydroquinone, 2-chlorophenol, 3-chlorophenol, and 4-chlorophenol.In the absence of chloride, the degradation products were identified as theearlier eluting catechol and hydroquinone. A third peak eluting earliest was1125.9. Bleaching of Dissolved Organic Matter in 185 nm Regimelikely the common degradation product of catechol and hydroquinone, hy-droxyquinol. In the presence of chloride, only catechol and hydroquinonewere detected, with no detectable formation of any of the later elutingchlorophenols.This observation is explained if, as reported by Gilbert et al. (1988), Cl re-actions favour attack at the hydroxyl group to produce the phenoxyl radicaland therefore the identical final products. Additional studies are requiredto confirm whether chlorinated species are formed to any significant extentin the 185 nm regime.5.9 Bleaching of Dissolved Organic Matter in 185nm RegimeThe aromaticity of organic matter is largely responsible for the absorbanceat 254 nm. Treatment by AOP is known to reduce this absorbance by reduc-ing the amount of congugation in the carbon structure and possibly by ringopenings induced by both OH attack and absorption of UV photons. Thisphenomena may be referred to as bleaching. It was desired to determinewhether a significant difference could be observed between the bleaching oc-curring in the 185 nm regime with the presence or absence of chloride.Solutions of Suwannee River NOM with 3.5 mg L−1 as C, at pH 7, wereirradiated with and without 100 mg L−1 of chloride obtained from NaCl. Ir-radiations were conducted for exposure times of 5, 10, and 20 min, all in trip-licate. An exposure time of 20 min, under the conditions used, would havebeen sufficient to induce an approximate 3-log reduction in carbamazepine,had the probe compound been used.Consistent with previous studies, no reduction in DOC was observed, eitherwith or without chloride. The 1 cm absorbance at 254 nm was found to be1135.10. Molar Absorption Coefficient of Chloride at 185 nmreduced by approximately 50%, from 0.12 to 0.06 cm−1 in both cases. Nostatistically significant difference was observed for the bleaching of Suwan-nee River NOM in the 185 nm regime with or without chloride.5.10 Molar Absorption Coefficient of Chloride at185 nmAccurate measurement of the 185 nm molar absorption coefficient of chlorideis required, as it is likely to be the major absorber of photons at this wave-length and may shift the 185 nm regime from OH to Cl as the dominantoxidant.Measurement of the absorbance of aqueous solutes at wavelengths below200 nm is difficult using conventional UV-VIS spectrophotometers. Reasonsfor this include multiple factors that ultimately degrade the quality of themeasurement, and are briefly reviewed here.In a conventional spectrophotometer, the UV radiation source is a deu-terium lamp, which emits a broad spectrum output in the UV, the majorityof which is emitted at wavelengths greater than 200 nm. The small fractionof radiation produced below 200 nm is of relatively low power and highernoise. The absorbance of water itself is high (αH2O > 1 cm−1), resulting insubstantial attenuation of the signal of interest at typical path lengths of1 cm. Thus, to measure the absorbance of a solute involves the subtractionof two weak signals, that of the test solution and the blank. Radiation thatis not of interest reaches the sample due to limitations of the monochrometerslit and imperfections of the gratings. Such radiation is referred to as strayradiant energy and is overwhelmingly of longer wavelength. Such longerwavelength radiation experiences virtually no attenuation in water due tonegligible absorbance, and thus reaches the detector to contribute to thesignal of interest. The net effect of measuring a highly absorbing specimen1145.10. Molar Absorption Coefficient of Chloride at 185 nmwith such a weak signal in the presence of stray radiant energy is an appar-ent absorbance measurement that is lower than the actual absorbance. Thegreater the quantity of stray radiation present, the greater the discrepancy.It is difficult to measure, and therefore correct, for stray radiant energy at185 nm. The influence of stray radiant energy may be reduced by minimiz-ing the path length used for measurement. However, this also reduces themagnitude of a signal already prone to significant noise, and therefore doesnot improve precision. More extensive information regarding stray radiantenergy may be found in Burgess and Frost (1999); Sommer (1989).An alternative method to measuring 185 for aqueous solutes exploits theconvenience of the quasi-collimated beam and commercially available fusedsilica cells of precision thickness.A double cell system was assembled, with an upper and lower cell placedin the quasi-collimated beam and in good optical contact with one another.The lower cell contained a standard reference solution to irradiate, withcarbamazepine as probe and tert-butanol as scavenger. The upper cell wasfilled with either ultrapure water as a reference or the solution to be tested,such as a series of chloride solutions of known concentration. The preciselyknown path lengths are provided by standard commercially available spec-trophotometer cells. The lower cell was of 10.0 mm path length, and theupper cell was of 1.0 mm path length.A series of irradiations produced a set of pseudo first-order degradationcurves, each with a slope of k′ corresponding to the contents of the 1.0 mmcell, with ultrapure water as a reference giving k′o. A plot of log(k′/k′o) vs.[Cl–] for probe compound degradation in the lower cell will have a slopeequal to 0.1× (M−1 cm−1) of the species in the upper cell. It is importantthat the probe be photochemically inert to 254 nm radiation by either a suf-ficiently low 254 or Φ254, a condition satisfied by carbamazepine.Irradiations were performed in triplicate with fresh test solution placed in1155.10. Molar Absorption Coefficient of Chloride at 185 nmthe upper cell for each run. Data related to the calculation of k′ is displayedin Figure 5.13. The series of values of log(k′/k′o) plotted against [Cl–] isdisplayed in Figure 5.14. The last value taken at 100 mg L−1 of chloride wasomitted from the linear regression calculation. While it was found to bereproducible, it was also found to involve significant deviation from Beer-Lambert behaviour. This may be due to ion-pair formation at the higher[Cl–] concentrations (Butler, 1964).0 5 10 15 20−2−1.5−1−0.50Time (min)ln([CBZ] t/[CBZ] o)Figure 5.13: Kinetic method of determining 185,Cl− using double cell.Lower 10.0 mm cell contains [CBZ]o ' 0.25 µM, [(CH3)3COH] = 7 mg L−1as C. Upper 1.0 mm cell contains NaCl solutions or ultrapure water as areference. Solutions of the upper cell: [Cl−] = 0 mg L−1 ( ), [Cl−] =25 mg L−1 ( ), [Cl−] = 50 mg L−1 ( ), [Cl−] = 75 mg L−1 ( ),[Cl−] = 100 mg L−1 ( ).1165.10. Molar Absorption Coefficient of Chloride at 185 nmThe value calculated for 185 of chloride is 3540± 150 M−1 cm−1, in excellentagreement with the value of 3800 ± 300 M−1 cm−1 reported by Dainton andFowles (1965) and the value of ∼ 3500 M−1 cm−1 estimated from a graphby Weeks et al. (1963) (see Table 1.1). It is important to note that theseresults were obtained by different methods. Dainton and Fowles (1965) usedkinetic studies quantified by H2 gas evolution and Weeks et al. (1963) useddirect measurements of 185 nm absorbance by replacing the deuterium lampwith a low pressure mercury lamp in a nitrogen purged spectrophotometer.The current indirect methods using the double cell approach was also usedto measure the value of 185 for sulphate, bicarbonate and Suwannee RiverNOM, the values of which are presented in the following chapter.0 0.5 1 1.5 2 2.5 300.20.40.60.81[Cl–] (mM)−log(k′ /k′ o)Figure 5.14: Calculation of 185,Cl− using double cell from kinetic data.Lower 10.0 mm cell contains [CBZ]o ' 0.25 µM and [tBuOH] = 7 mg L−1as C. Upper 1.0 mm cell contains NaCl solutions or ultrapure water as areference.1175.11. Potential Cl to OH Interconversion5.11 Potential Cl to OH InterconversionThe determination of the molar absorption coefficient of chloride at 185 nmallows the evaluation of the model described by equations 5.6 and 5.7. Thismodel implies that the photogeneration of both OH and Cl are indepen-dent and that any interconversion from one radical to the other is negligible.If nitrobenzene is indeed of negligible reactivity with Cl (i.e. kCl,NB kOH,NB), then the degradation of nitrobenzene observed in Figure 5.5, wheretert-butanol is the organic scavenger, is a result of OH alone. It would thenbe useful to compare the relative changes in k′ and fH2O with increasingCl–. An additional assumption in this approach is that the Beer-Lambertlaw holds. These results are tabulated in Table 5.2.Table 5.2: Model evaluation using the degradation of ni-trobenzene with increasing Cl– in the 185 nm regime[Cl–] k′ k′/k′o fH2O fCl–(mgL−1) (min−1)0 0.127± 0.002 1.00 1.00 0.0025 0.105± 0.002 0.83 0.42 0.5840 0.082± 0.002 0.64 0.31 0.69100 0.068± 0.001 0.53 0.25 0.75NB: Uncertainties are the standard errors calculated by linearregression of the aggregated rate data from triplicate runs.It can be seen from Table 5.2 that fH2O decreases to a much greater extentwith increased Cl– than the relative decrease in the observed rate constantk′/k′o. In fact, for each concentration of Cl–, the resulting reduction in k′ isapproximately half the reduction in fH2O. As discussed earlier, this wouldbe explained if a portion of the photogenerated Cl is converted to OH. At-tempts to account for such a conversion via reactions of Cl and Cl –2 withwater (Equations 5.4 and 5.5) have been unsuccessful. Additional researchis required to understand this phenomena and permit the development of amore accurate kinetic model.1185.12. Summary5.12 SummaryNo previous literature reporting on the 185 nm AOP is known to have men-tioned the role of chloride and of the Cl radical. Experimental evidence hasbeen presented here to support the role of Cl as a major oxidant at environ-mentally relevant levels of both chloride and DOM. The evidence obtainedis in agreement with the work reported from other fields such as radiationand atmospheric chemistry.The chemistry related to Cl– and Cl is complex. Many questions remainunanswered regarding such aspects as the extent to which equilibrium 5.3is maintained, the mechanisms of Cl reactions, the reaction rate constantsof Cl with compounds of interest, and the potential conversion of Cl to OH.Experimental evidence supports the explanation of several observations con-sistent with the work of others. No influence of Cl– is observed for the 254 nm- H2O2 regime using either tert-butanol or Suwannee River NOM at circum-neutral pH. Chloride is expected to exert a scavenging effect below pH 4,which is an unlikely condition in drinking water treatment. The explanationfor this involves an intermediate reversible step between OH and Cl involv-ing H+, which is insufficiently available in the usual pH range of drinkingwater.In contrast, the influence of chloride in the 185 nm regime is substantial.The postulated extent of Cl involvement relative to OH, and the result-ing degradation rate of the target contaminant produced depend on severalfactors. These include the concentration of chloride, the relative overall re-activities of Cl and OH with the mixture of organic matter present (i.e.scavengers), as well as the relative reactivities with target contaminants ofinterest.1195.12. SummaryIonic strength effects observed support the existence of an equilibrium be-tween the uncharged oxidant Cl and the relatively unreactive diradical anionCl –2 via a secondary salt effect. The conditions under which such an equi-librium is maintained require further study, as this will effect the availableoxidant when chloride concentrations are significant.Product studies using phenol did not reveal any chlorinated phenols, exclud-ing addition to an aromatic ring as a significant reaction mechanism in thiscase. The first generation daughter products observed when chloride waspresent were identical to those formed by OH. Mechanistic explanation forthis is provided by the formation of the phenoxyl radical, an intermediatetheorized to form via both OH and Cl .The 185 nm regime with chloride at 100 mg L−1 is interpreted to involve Clas the dominant radical. Yet, under such conditions, no significant effect isobserved on the measured absorbance of Suwannee River NOM solutions.Thus, the replacement of OH by Cl neither enhances nor diminishes therate of bleaching of Suwannee River NOM under 254 nm irradiation.Molar absorption coefficients can be accurately measured using an indirectkinetic technique involving two consecutive cells, with one used as the sam-ple cell and a second the detecting cell used to measure the degradationrate of a probe compound. This method avoids the fundamental limitationsinherent in using conventional spectrophotometers.Further research is required to elucidate the role of Cl in the 185 nm AOP.It is likely that even when chloride is responsible for the majority of 185 nmphoton absorption, the result is not a replacement of OH with Cl , buta mixture of the two. The relative role these two radicals in contaminantdegradation will depend on the concentrations of chloride and DOM, theirrelative reactivities, and the mechanisms involved in any conversion fromCl /Cl –2 to OH. In natural waters the situation is likely complicated furtherby the involvement of yet other species that absorb photons and generate1205.12. Summaryreactive radicals.121Chapter 6Sulphate, Bicarbonate andInteractions of Major Solutes6.1 The 185 nm AOP and Other SolutesAs with the UV/H2O2 and O3/H2O2 AOPs, the 185 nm AOP is subject tothe parasitic effects of OH scavengers such as DOM and bicarbonate. Yet,it has been shown that a unique feature of the 185 nm AOP is the substan-tial sensitivity to the presence of chloride. Experimental evidence reportedin Chapter 5 supports the interpretation of the chloride effect as being dueto the involvement by Cl as a major oxidant, likely simultaneously withOH. Thus, the presence of chloride in the 185 nm regime may result in thepresence of a mixture of OH and Cl , both highly reactive oxidants, withthe proportion of each dependent on their relative reactivities with majorsolutes. This has significant implications on the degradation rate of targetcontaminants, and possibly on the mixture of products formed. Yet, infor-mation regarding Cl reaction rates, reaction mechanisms, and end productsare limited.This situation is complicated further, as typical water matrices contain othermajor solutes that also contribute to the absorption of 185 nm photons andgeneration of radicals. Adequate elucidation of the details involved requiressubstantial experimental work and techniques beyond what is presentedhere. The current objective is to clearly demonstrate the importance ofother solutes influencing the 185 nm AOP, which has not been reported to-date, indicate general trends in relevant reactivities, and provide guidance1226.2. Sulphateto further detailed studies.As mentioned in Chapter 1, the major inorganic ions in natural waters in-clude Ca2+, Mg2+, Na+, and K+ as cations, and Cl–, SO 2–4 , HCO–3 , andCO 2–3 as anions, with other ions occurring at concentrations typically lessthan 1 mg L−1 (Wetzel, 2001).A review of the literature provides no evidence that the major cationicspecies participate directly in any photochemical or radical reactions in the185 nm regime. While it is likely that Mn2+ and Fe3+ may absorb photons,they are minor solutes (i.e.  1 mg L−1). Nevertheless, the involvementof cationic species has not been rigorously tested and their lack of directinvolvement should be experimentally confirmed under relevant conditions.In contrast, all the major anionic species are expected to absorb 185 nmphotons and generate radicals to varying extents. The involvement of Cl–has been established in Chapter 5. In this chapter the focus is on SO 2–4 andHCO –3 and some major interactions. The role of CO2–3 has not been thor-oughly tested. However, at pH 8 and below, its concentration is negligible.Thus, the pH dependent influence of the conjugate acid-base pair HCO –3 /CO2–3 should be included in future studies, particularly where treatedwaters exceed pH 8. As mentioned in Chapter 1, the molar absorption co-efficient (185) for CO2–3 is estimated as approximately 103 M−1 cm−1. Noinformation is available on the value of 185 or of the quantum yield (Φ) forHCO –3 .6.2 SulphateSulphate is a major anion in natural waters present in rivers and lakes typ-ically between 2.5 and 25 mg L−1, though concentrations of 100 mg L−1 orgreater may occur depending on the dominant mineral rock type of thedrainage region (Wetzel, 2001).1236.2. SulphateBefore proceeding to investigate the role of SO 2–4 in the 185 nm regime, in-fluence on the 254 nm-H2O2 regime is first investigated.6.2.1 Sulphate in the 254 nm - H2O2 RegimeExperiments were conducted using the 254 nm collimated beam in the ab-sence of 185 nm radiation. As before, carbamazepine was used as a probecompound. The scavenger used was either tert-butanol or Suwannee RiverNOM, in either case at a concentration of 7 mg L−1 as C. Solutions werebrought to circumneutral pH by addition of NaOH. Though Na2SO4 is ofsufficient aqueous solubility, due to difficulties in completely dissolving thesolid, and in order to avoid colloidal remains even after filtration, a stocksolution of Na2SO4 was prepared by neutralizing analytical grade H2SO4with NaOH and verifying the dilutions by ion chromatography. Two levelsof SO 2–4 were used, 0 and 100 mg L−1.In order to minimize the influence of H2O2 as a radical scavenger, it was de-sired to use a dose of H2O2 less than 10 mg L−1, and a dose of 7 mg L−1 wasfound adequate. All irradiations were performed in triplicate, with samplestaken at regular intervals. Residual H2O2 of samples was quenched witha 10 µL droplet of approximately 100 mg L−1 bovine catalase placed at thebottom of the HPLC vials, as discussed in Chapter 2.Irradiations were conducted with exposure times sufficient to achieve at least20% removal of the probe compound. The results are plotted in Figures 6.2and 6.1 for tert-butanol and Suwannee River NOM as scavengers, respec-tively.The calculated pseudo first-order rate constants k′ for the degradation ofthe probe compound reveal no statistically significant difference betweenthe levels of SO 2–4 used. The scatter in the data is substantial, increasing1246.2. Sulphate0 5 10 15 20 25−0.8−0.6−0.4−0.20Time (min)ln([CBZ] t/[CBZ] o)Figure 6.1: Tert-Butanol in the 254 nm - H2O2 regime and influence ofsulphate. [CBZ]o ' 0.25 µM. [(CH3)3COH] = 7 mg L−1 as C. Solu-tion of Na2SO4 at pH 7 used as source of sulphate. [SO42−] = 0 mg L−1( ),100 mg L−1 ( ).at the longer exposure times. This phenomena appears to be common to254 nm-H2O2 experiments. In contrast, 185 nm experiments exhibit muchgreater repeatability and precision. It is possible that the observed scatteris due to slight differences in the dose of H2O2 delivered for each irradia-tion, despite the fact that the delivered dose of H2O2 was measured to varyby less than ±0.1 mg L−1 between runs. A verification of the quenching ofH2O2 found this procedure to be adequate to reduce H2O2 to below detec-tion (0.01 mg L−1) within a time of less than two minutes.Regardless of the scatter of data and the failure to detect any effect of SO 2–41256.2. Sulphate0 5 10 15 20 25−0.6−0.4−0.20Time (min)ln([CBZ] t/[CBZ] o)Figure 6.2: Swannee River NOM in the 254 nm - H2O2 regime and influenceof sulphate. [H2O2]= 7 mg L−1 as C. [CBZ]o ' 0.25 µM. [DOC] = 7 mg L−1as C. Na2SO4 used as source of sulphate. All solutions prepared at pH 7.[SO42−] = 0 mg L−1 ( ),100 mg L−1 ( ).at the extremes of 0 and 100 mg L−1, any effect that may exist is likely tobe small and practically negligible.At pH 7, the relevant one-electron reduction potentials include E(SO –4 /SO2–4 )' 2.4 V and E( OH/OH–) = 1.9 V (Wardman, 1989). These values suggestthat electron transfer from SO 2–4 to OH is unfavourable. This, and theexperimental evidence support the assumption that SO 2–4 has a negligibleinfluence in the 254 nm-H2O2 regime and therefore upon the UV/H2O2 AOP.Therefore, any effect observed in the 185 nm regime will not be due to pres-ence of 254 nm radiation.1266.2. Sulphate6.2.2 Sulphate in the 185 nm RegimeExperiments conducted under 185 nm irradiation found the impact of sul-phate to be significant. These experiments used tert-butanol as the scav-enger, with four concentrations of SO 2–4 . Results are plotted in Figure 6.3,where it can be seen that the increase in [SO 2–4 ] results in an increase in thedegradation rate of the probe compound carbamazepine.As discussed in Chapter 1, it is expected that SO 2–4 ions will absorb photonsat 185 nm by a charge transfer to solvent mechanism, with electron ejectionforming SO –4 with a quantum yield of∼ 0.65 (see Table 1.1 on page 22). Theresulting SO –4 will react with the main scavenger tert-butanol with a second-order rate constant approximately three orders of magnitude less than therate constant between OH and tert-butanol (Buxton et al., 2000). Thus, ifthe system consists of a mixture of both radicals OH and SO –4 , a relativelysmall contribution of SO –4 is likely sufficient to induce a significant impact,provided the rate constants of both oxidants with the target contaminant(or probe) are similar (i.e kSO –4 ,CBZ ∼ kOH,CBZ), while the rate constantof SO –4 with the matrix is substantially lower (i.e. kSO –4 ,DOM  kOH,DOM ).An expression for the pseudo-first order rate constant k′ of the probe (ortarget) compound, based on a mixture of OH and SO –4 , may be written asthe sum of the two contributions:k′ = k′OH + k′SO –4(6.1)with each term on the right-hand side representing the contribution of theindicated radical. The expression may be further developed using equations2.13 and 2.14 to give the SO –4 analog to equation 5.7:1276.2. Sulphatek′ =(kOH,CkOH,SΦH2OfH2O +kSO –4 ,CkSO –4 ,SΦSO 2–4 fSO 2–4)Ia[S](6.2)where the k values of the right-hand side are the second-order reaction rateconstants between the indicated subscripts, with C representing the probecompound and S the scavenger. ΦH2O is the quantum yield for water pho-tolysis and ΦSO 2–4 is the quantum yield for the ejection of an electron fromSO 2–4 , both at 185 nm. The values of fH2O and fSO 2–4 are the fraction of185 nm photons absorbed by water and SO 2–4 , respectively, calculated usingequation 2.15 (fi = αi/αtot). Ia is the rate at of 185 nm photon absorptionper unit volume, and [S] is the concentration of the scavenger.All the terms of the right-hand side of equation 6.2 are known or can bemeasured, with the exception of kSO –4 ,C , the second-order rate constant ofSO –4 with the probe carbamazepine. With the experimental determina-tions of k′, the value of kSO –4 ,C can be calculated. The quantities measuredseparately or quoted from the literature and used in these calculations arelisted in Table 6.1. The experimentally determined values of k′, the relativeratios of k′OH and k′SO –4, as well as the calculated values of kSO –4 ,C aretabulated in Table 6.2. From the measured value of k′ with no SO 2–4 , theterm Ia was determined to be 2.70× 10−7 M s−1, consistent with the valuesused for the determinations of 185 nm fluence rate presented in Chapter 4.This value of Ia was used here in subsequent calculations of kSO 2–4 ,C . Datausing [SO 2–4 ] = 50 mg L−1 were inconsistent, possibly due to contaminationof samples, and were not used for these calculations but are included withthe raw data in the Appendix.Using the indirect kinetic technique discussed in Chapter 5 for Cl–, the molarabsorption coefficient at 185 nm () of SO 2–4 was determined by a separateexperiment to measure 160± 20 M−1 cm−1 (see Table A.43 on page 216 for1286.2. Sulphate0 1 2 3 4 5−4−3−2−10[SO42−]Time (min)ln([CBZ] t/[CBZ] o)Figure 6.3: Influence of sulphate in the 185 nm regime with tert-butanol.[tBuOH] = 7 mg L−1 as C. [CBZ]o ' 0.25 µM. Na2SO4 used as sourceof sulphate. [SO42−] = 0 mg L−1 ( ), 25 mg L−1 ( ), 75 mg L−1 ( ),100 mg L−1 ( ).1296.2. SulphateTable 6.1: Numerical values used in calculation of the second-order rate constant for the reaction of SO –4 with carbamazepine(kSO –4 ,CBZ) with equation 6.2Value Ref.k OH,C = 6.8× 109 M−1 s−1 This workk OH,S = 6.0× 108 M−1 s−1 Buxton et al. (2000)kSO –4 ,S = 8.4× 105 M−1 s−1 Buxton et al. (2000)αH2O = 1.80 cm−1 Weeks et al. (1963)SO 2–4 = 160 M−1 cm−1 This workΦH2O = 0.3 Getoff and Schenck (1968)ΦSO 2–4 = 0.65 Dainton and Fowles (1965),Barrett et al. (1965)C = carbamazepine, S = tert-butanol, w = water.Table 6.2: Experimental determination of the second-order rate constant for thereaction of SO –4 with carbamazepine (kSO –4 ,CBZ) with equation 6.2[SO 2–4 ] k′ k′OH/k′ k′SO –4 /k′ kSO –4 ,C(mM) (s−1) M−1 s−10.00 6.31± 0.09× 10−3 1.00 0.00 -0.26 1.08± 0.01× 10−2 0.56 0.44 1.43× 1080.78 2.09± 0.07× 10−2 0.30 0.70 1.61× 1081.04 2.26± 0.12× 10−2 0.25 0.76 1.38× 108Average 1.5± 0.1× 108Uncertainty in average the standard deviation from the three determinations.1306.3. Bicarbonatedata)Note from Table 6.2 that the contribution to k′ from SO –4 , relative to OH,is substantial at [SO 2–4 ] values of 25, 75, and 100 mg L−1, despite SO 2–4 con-tributing approximately 2, 6 and 8% of the total absorbance of the solution,respectively. The value of the rate constant for the reaction of sulphate rad-ical and carbamazepine (kSO –4 ,C) was calculated to be 1.5× 108 M−1 s−1.The value 1.9× 109 M−1 s−1 was reported by (Matta et al., 2011) wherebySO –4 was generated by the Fenton-like reaction of persulfate and cobalt(S2O2–8 /Co2+). Since a nitrate salt of cobalt was used by that study, thehigher value obtained may be due to unaccounted for effects of nitrate andthe nitrate radical (NO3 ) or other matrix components. Reevaluation of thisvalue may be accomplished by using the UV/S2O2–3 AOP in a collimatedbeam in a competitive kinetics experiment with a convenient reference com-pound selected from the compilation of Ross and Neta (1979). In any case,the value obtained here is within the range of values expected based on therate constants for other organic compounds (Neta et al., 1977; Ross andNeta, 1979). An important discovery, not previously reported in the lit-erature, is that like chloride, sulphate present at environmentally relevantconcentrations may have a profound influence on the 185 nm process, andthat this AOP involves not only OH, but a mixture of several highly reac-tive radical species.6.3 BicarbonateIn addition to Cl– and SO 2–4 , the major anions in surface waters includethe conjugate acid-base pair HCO –3 /CO2–3 . In surface waters exposed to theatmosphere, these two species form an equilibrium with each other and withdissolved carbon dioxide via its hydrolysis product carbonic acid (H2CO3).The pair HCO –3 /CO2–3 has a pKa of 10.3 at 25◦C, with less than 1% changein this value between 5 and 35 ◦C (Butler, 1982). In the common pH rangefor surface waters of 6 to 8, the concentration of CO 2–3 is less than 1% that1316.3. Bicarbonateof HCO –3 , and thus can usually be considered negligible. The global aver-age concentration of bicarbonate in river waters is approximately 60 mg L−1(∼ 1 mM) with an upper range extending to 300 mg L−1 (Stumm and Mor-gan, 1996; Wetzel, 2001). Depending on the underlying geology, some waterbodies are naturally very low in HCO –3 , with little buffer capacity againstacidification. Such conditions are common in the granitic rock dominatedregions of Eastern Canada and Northeastern United States (Bunce, 1990).Both HCO –3 and CO2–3 react by electron transfer to OH with second-order reaction rate constants of 8.5× 106 and 3.9× 108 M−1 s−1, respectively(Buxton et al., 1988). Due to the carbonate radical being a strong acid(pKa < 0), CO–3 is the common product of both reactions (Czapski et al.,1999). Unlike OH and Cl , the species CO –3 is a highly selective reactant(Chen and Hoffman, 1973; Larson and Zepp, 1988), and with few exceptionswill not significantly contribute to the degradation of trace organic contam-inants. Thus, the pair HCO –3 /CO2–3 are considered OH scavengers, andtogether with DOM act to reduce the efficiency of AOPs. Furthermore, dueto an approximate 50 fold increase in OH reactivity with CO 2–3 relative toHCO –3 , the scavenging effect on AOPs due to the carbonate species increasesdramatically above pH 8.In addition to scavenging OH, electron transfer reactions from HCO –3 /CO2–3to Cl and SO –4 are also expected to occur based on the relative magnitude ofthe one-electron redox potential E(CO –3 /CO2–3 ) = 1.4 V (Huie et al., 1991).The carbonate species therefore also serve as scavengers for the radicals Cland SO –4 , with the available literature values for rate constants tabulatedin Table 1.2 on page 24. From these values, it can be seen that while the re-activities of OH and SO –4 with HCO–3 are comparable, the reactivity of Clwith HCO –3 is larger by one or two orders of magnitude. This may impose asubstantial impact on the 185 nm AOP, depending on the water composition.Photochemical reactions involving HCO –3 /CO2–3 are also expected to occurin the 185 nm regime via charge-transfer-to-solvent mechanisms, though as1326.3. Bicarbonatementioned in Chapter 1, virtually no information is available regarding mo-lar absorption coefficients and quantum yields for electron photodissociation.If these quantities are significant, then HCO –3 /CO2–3 may be considered bothscavengers of radicals as well as scavengers of 185 nm photons (i.e. inner fil-ters), compounding the impact on 185 nm AOP efficiency.Before considering the simultaneous presence of multiple anionic species un-der 185 nm irradiation, the influence of bicarbonate alone was investigatedin a set of solutions using tert-butanol as a scavenger, carbamazepine asthe probe compound, and four levels of HCO –3 concentration. Aliquots ofan NaHCO3 stock were added to each solution. The pH of the solutionswere approximately pH 6.1 in the case of no added HCO –3 and pH 8.1 forsolutions containing 60, 120, and 180 mg L−1 of HCO –3 . This was deemedsufficient to neglect the contribution of CO 2–3 , thus avoiding the depressionof pH using any acids or buffer systems that would impart undesired speciesto the matrix. The results are plotted in Figure 6.4.While the trend observed for the probe compound degradation rate k′ de-creases with increasing [HCO –3 ] as expected, the scavenging of OH by HCO–3does not account for the entirety of the observed effect. Using the indirectkinetic technique discussed in Chapter 5 for Cl–, the molar absorption coef-ficient at 185 nm () of HCO –3 was measured by a separate experiment anddetermined to be 290± 40 M−1 cm−1 (see Table A.43 on page 216 for data).Using the known rate constants and concentrations for the probe carba-mazepine, and the scavengers tert-butanol and HCO –3 , a comparison of thecalculated and observed values of k′ was made to check for consistency ofHCO –3 acting by two independent processes. An expression relating theknown quantities to the observed rate constant k′ is obtained by elaborat-ing Equation 2.11:1336.3. Bicarbonatek′ =kOH,C ΦH2OfH2OIakOH,S [S] + kOH,HCO –3 [HCO–3 ](6.3)where all terms are as previously defined. If the above expression is validand HCO –3 does indeed absorb photons to generate nonreactive CO–3 atthe expense of OH generation, and the Beer-Lambert behaviour holds forHCO –3 at 185 nm, then a correction of fH2O will yield better agreement be-tween calculated and experimental results.The value of Ia is required for calculations and can be obtained from theexperimental results corresponding to the solution with no HCO –3 addition,ΦH2O = 0.3, and fH2O = 1. This yields a value of Ia = 2.5× 10−7 M s−1, inagreement with the values used for the fluence rate determinations reportedin Chapter 4, and the calculations of the previous section. Using the nomi-nal value of HCO –3 = 290 M−1 cm−1 allows the calculation of the absorbancecoefficient αHCO –3 = 185 · [HCO –3 ], which in turn permits the calculation ofαtot = αH2O + αHCO –3 , and fH2O = αH2O/αtot, where αH2O = 1.80 cm−1 istaken from (Weeks et al., 1963). This value of fH2O is then applied to thecalculation of k′calc and compared with experimental values k′exp.The values of the OH rate constants are 6.8× 109 M−1 s−1 for carbamazepine,6.0× 108 M−1 s−1 for tert-butanol, and 8.5× 106 M−1 s−1 for HCO –3 . Theconcentrations of tert-butanol used was 7.0 mg L−1 as C or 1.35× 10−4 M.A comparison of the two processes involving HCO –3 can be made by com-paring the calculated rate values k′calc when absorption of photons by HCO–3is not taken into account (fH2O ≡ 1) and when such absorption is taken intoaccount (fH2O < 1). These are tabulated in Table 6.3.The k′ values of Table 6.3 are plotted in Figure 6.5. An inspection of Figure1346.3. Bicarbonate0 2 4 6 8 10−2.5−2−1.5−1−0.50[HCO3−]Time (min)ln([CBZ] t/[CBZ] o)Figure 6.4: Influence of bicarbonate in the 185 nm regime with tert-butanol.[tBuOH] = 7 mg L−1 as C. [CBZ]o ' 0.25 µM. Solution of NaHCO3 at pH8 used as source of bicarbonate. [HCO3−] = 0 mg L−1 ( ), 60 mg L−1( ), 120 mg L−1 ( ), 180 mg L−1 ( ).1356.3. BicarbonateTable 6.3: Comparison of the calculated and experimental rates due toHCO –3 in the 185 nm regime[HCO –3 ] αHCO –3 fH2O k′calc k′calc k′exp(mgL−1) (cm−1) (min−1) (min−1) (min−1)fw ≡ 1 fw < 10 0.00 1.000 - - 0.367± 0.00560 0.29 0.857 0.332 0.285 0.281± 0.003120 0.58 0.750 0.303 0.227 0.234± 0.004180 0.87 0.667 0.279 0.186 0.186± 0.004NB: Uncertainties are the standard errors calculated by linear regression of theaggregated rate data from triplicate runs.6.5 suggests that the two effects due to HCO –3 in the 185 nm AOP are indeedindependent. The black circles show the values of k′calc corresponding to thecondition where only the OH scavenging effect is considered. The grey cir-cles show the values of k′calc that take into account both OH scavenging andabsorption of 185 nm photons due to HCO –3 . Excellent agreement is foundbetween the experimental values indicated by the asterisk and the lattercalculated values that account for both phenomena. This result also sup-ports the assumption that the reaction between CO –3 and carbamazepine isnegligible. However, it should be noted that the inner filter effect of HCO –3may be negligible in a real water matrix with other solutes such as Cl– andDOM that more strongly absorb at 185 nm.At this point, it is now clear that the 185 nm regime involves not only OHas a reactive radical, but also potentially significant contributions from Cl ,SO –4 , and that the major solutes Cl–, SO 2–4 , HCO–3 , and DOM significantlyinfluence the involvement of these radicals. Aspects of the interactions be-tween these major solutes are next considered.1366.4. Interactions Among Major Solutes0 1 2 3 40.150.20.250.30.350.4[HCO –3 ] (mM)k′(min−1)Figure 6.5: Comparison of calculated and experimentally observed ratesdue to HCO –3 influence in the 185 nm regime, with [CBZ]o ' 0.25 µM and[tBuOH] = 7.0 mg L−1 as C in all solutions. NaHCO3 used as source ofHCO –3 . k′calc with fw ≡ 1 ( ). k′calc with fw < 1 ( ). k′exp ( ) from Figure6.4. Best-fit lines shown for calculated values.6.4 Interactions Among Major SolutesEvidence has now been presented that 185 nm irradiation of aqueous solu-tions will generate the radicals OH, Cl , SO –4 , and CO–3 , which, save for thelast species, will contribute by varying extents to the oxidative degradationof trace organic contaminants. The solutions studied represent simplifiedmodels of natural waters, and the available evidence indicates that thesereactive species cannot be considered independently in general for actualnatural waters.The one-electron redox potentials of E( OH,OH–) = 1.9 V (pH > 2) andE(Cl ,Cl–) = 2.4 − 2.6 V (Wardman, 1989) support the observations thatCl– is not a scavenger of OH. As discussed in the previous chapter, the1376.4. Interactions Among Major Solutes185 nm photogenerated Cl not only reacts with organic solutes, but alsoreacts reversibly with Cl– to form the relatively nonreactive Cl –2 with anequilibrium constant (K = 1.4× 105 M−1) that heavily favours Cl –2 (Bux-ton et al., 1998). Both Cl and Cl –2 react with water itself, and via amechanism not yet fully understood, may ultimately form OH. Thus, un-like the case of SO –4 and OH, the delineation between the contributionsof OH and Cl to the observed degradation rate k′ of a trace contaminantdoes not yield a simple relationship. The relative contributions of OH andCl to k′ is expected to be highly sensitive to the relative rates associatedwith the multiple sinks of Cl , which depends on the matrix composition. Acondition that also depends on the matrix composition is the possibility thatsignificant deviations from equilibrium between Cl and Cl –2 may occur, asobserved by Buxton et al. (1998).As discussed earlier, the value of E(SO –4 , SO2–4 ) ' 2.6 V also supports theobservation that SO 2–4 , like Cl–, is not a scavenger of OH. The 185 nm molarabsorption coefficient of Cl– is over twenty times greater than that of SO 2–4(3540 vs. 160 M−1 cm−1), which alone would suggest that SO 2–4 should havea negligible effect when Cl– is also present at comparable concentrations.Yet, the redox potentials are approximately equal, and a reversible reactionoccurs between these species with an equilibrium constant of approximately1.2 at zero ionic strength (Buxton et al., 1999):SO –4 + Cl– −−⇀↽− Cl + SO 2–4 (6.4)Thus, even when negligible 185 nm absorption occurs by SO 2–4 , resultingin negligible SO –4 photogeneration, SO–4 may still be a significant contrib-utor to trace organic degradation as a result of the oxidation of SO 2–4 by Cl .To illustrate the potential sensitivity of the 185 nm AOP to the composition1386.4. Interactions Among Major Solutesof the water matrix, consider the model solution composed of tert-butanol,chloride, and sulphate at 7 mg L−1 as C, 25 mg L−1, and 50 mg L−1, respec-tively. The fractions of absorbed photons for water, Cl–, and SO 2–4 arecalculated to be 0.41, 0.57, and 0.02, respectively. Inclusion of HCO –3 at100 mg L−1 results in little change to the above values, with the fractionabsorbed by HCO –3 itself approximately 0.01. However, the relative contri-butions of the three radicals, OH, Cl , and SO –4 , to the overall degradationrate of a trace organic contaminant (k′) also depends strongly on the rela-tive scavenging of the matrix and the relative rate constant with the targetcontaminant by the three radicals. Extending equation 6.1, a general ex-pression for k′ may be written as:k′ = k′OH + k′Cl + k′SO –4(6.5)wherek′OH =kOH,C ΦH2OfH2OIakOH,S [S] + kOH,HCO –3 [HCO–3 ](6.6)k′Cl =kCl,C ΦCl–fCl–IakCl,S [S] + kCl,HCO –3 [HCO–3 ](6.7)1396.4. Interactions Among Major Solutesk′SO –4 =kSO –4 ,C ΦSO 2–4 fSO 2–4 IakSO –4 ,S [S] + kSO –4 ,HCO –3 [HCO–3 ](6.8)and where S represents tert-butanol, Ia is the rate of 185 nm photon ab-sorption per unit volume of solution, and all other terms are as describedpreviously. All terms in equations 6.6 to 6.8 are known or can be calculated,except for the second-order rate constant between Cl and carbamazepine(i.e. C) in equation 6.7. However, it has been deduced that this rate con-stant is likely to be comparable to that with OH, if not slightly higher, andestimated to be on the order of 1010 M−1 s−1.Continuing with the model solution from above, including bicarbonate, thesum of the denominators from equations 6.6 to 6.8 represent the scavengingterms for each respective radical and can be calculated using known reactionrate constants with tert-butanol (Buxton et al., 2000) and bicarbonate (seeTable 1.2). The resulting scavenging terms for OH, Cl , and SO –4 are cal-culated as 8.2× 104 s−1, 4.1× 105 s−1, and 1.5× 103 s−1, respectively. Thecontributions to k′ from OH, Cl , and SO –4 are calculated to be 60%, 32%,and 8%, respectively. Note that this result is obtained assuming that inter-conversion between radicals is negligible.6.4.1 Sulphate and BicarbonateThe interaction of SO 2–4 and HCO–3 was investigated with tert-butanol asthe organic scavenger (7 mg L−1 as C), and carbamazepine (' 0.25 µM).Values of 0 to 50 mg L−1 for SO 2–4 , and 0 to 180 mg L−1 for HCO–3 , formedthe points of a 2×2 factorial design, including a centre point, used to detectinteractions and curvature. Triplicate irradiations were performed for eachof the five settings. The experimental values of k′, the observed degradationrate of carbamazepine, were compared to the calculated values obtained by1406.4. Interactions Among Major Solutesusing equations 6.5, 6.6, and 6.8. The experimental value of k′ for no addedsulphate or bicarbonate was used to determine the 185 nm photon absorptionrate Ia, used in subsequent calculations. A value of Ia = 2.5× 10−7 M s−1was found, in excellent agreement with previous determinations. Experi-mental and calculated values of k′ are tabulated in Table 6.4.Table 6.4: Comparison of calculated and experimental rates dueto SO 2–4 and HCO–3 in the 185 nm regime[SO 2–4 ] [HCO–3 ] fH2O k′calc k′exp(mgL−1) (mgL−1) (min−1) (min−1)0 0 1.00 - 0.353± 0.00650 0 0.96 0.870 0.824± 0.0170 180 0.68 0.186 0.167± 0.00225 90 0.79 0.247 0.371± 0.00650 180 0.66 0.182 0.339± 0.007NB: Uncertainties are the standard errors calculated by linear regres-sion of the aggregated rate data from triplicate runs.It can be seen from the plots of Figure 6.6 that, as before, SO 2–4 alone accel-erates degradation, while HCO –3 as a retarding effect. The combination ofSO 2–4 and HCO–3 , at the concentrations used, appears to induce little changein the degradation rate relative to the solution containing only tert-butanolas a major solute. For the solutions containing either SO 2–4 or HCO–3 , a com-parison of the values in Table 6.4 show close agreement between calculatedand experimental k′ values. However, for the case of interactions (bottomtwo rows), the calculated values of k′ underestimate the experimental val-ues. Rather than having a net decrease in the value of k′, the experimentalresults appear as if the two effects cancel each other.Several causes for the discrepancy can be ruled out from previous evidence.Firstly, it is unlikely that CO –3 is responsible for the unaccounted degrada-tion of carbamazepine, since such an effect was not observed during studiesof HCO –3 alone (see Figure 6.5). It is also unlikely that a possible reaction of1416.4. Interactions Among Major SolutesSO –4 with water was sufficient to contribute significantly to the steady-stateconcentration of OH, based on the success of equation 6.2 in estimating thesecond order reaction rate constant between SO –4 and carbamazepine (seeTable 6.2). This phenomena may be due to radical interconversion and re-quires further study.Note, the inclusion of centre points in the concentrations used allows thedetection of curvature in the response, manifested as a significant differencebetween the average value of k′ obtained from the centre points relative tothe average value of k′ of all the factorial points. The presence of curvatureimplies that a linear effects model is inappropriate and that the design mustbe augmented in order to investigate the nonlinearity. The present resultsserve to illustrate that the system is nonlinear using concentrations thatbracket the environmentally relevant values. If one nevertheless proceedswith an Analysis of Variance (ANOVA), the required sum of squares of theerror (SSE) obtained will not represent the random error of pure noise, andany statistical tests that use it may produce spurious conclusions. Instead,SSE will be composed of a contribution of pure noise and a contributionfrom the systematic error resulting from a lack of fit to a linear model. Inthe present case, the greatest contribution to SSE is found to be that re-sulting from a lack of fit. How future investigations should augment theexperimental design will depend on what type of nonlinearity is expected,which will require mechanistic information.1426.4. Interactions Among Major Solutes0 2 4 6 8 10−3−2−10See caption.[SO42−] = 0 mg L−1[HCO3−] = 180 mg L−1[SO42−] = 50 mg L−1[HCO3−] = 0 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 6.6: Interaction of sulphate and bicarbonate in the 185 nmregime with tert-butanol. [tBuOH] = 7 mg L−1 as C and [CBZ]o '0.25 µM. Na2SO4 and NaHCO3 used as sources of sulphate and bicar-bonate. [SO42−] = 0 mg L−1, [HCO3−] = 0 mg L−1( ); [SO42−] =0 mg L−1,[HCO3−] = 180 mg L−1( ); [SO42−] = 25 mg L−1, [HCO3−] =90 mg L−1( ); [SO42−] = 50 mg L−1, [HCO3−] = 180 mg L−1( ).[SO42−] = 50 mg L−1, [HCO3−] = 0 mg L−1( );6.4.2 Chloride and BicarbonateThe interaction between Cl– and HCO –3 was then studied using the samebase system consisting of tert-butanol as organic scavenger and carbamazepineas probe, at 7 mg L−1 as C, and 0.25 µM, respectively. As in the previ-ous case, the concentrations used were 0 to 50 mg L−1 for Cl–, and 0 to180 mg L−1 for HCO –3 . All five matrix conditions were irradiated in trip-licate, from which the experimental values of k′, the degradation rate of1436.4. Interactions Among Major Solutescarbamazepine, were determined. Both k′exp and the calculated fractions ofabsorbed photons by water (fH2O) and Cl– (fCl–) are tabulated in Table 6.5.Table 6.5: Experimental degradation rates of carbamazepinedue to Cl– and HCO –3 in the 185 nm regime[Cl–] [HCO –3 ] fH2O fCl– k′exp(mgL−1) (mgL−1) (min−1)0 0 1.00 - 0.372± 0.00750 0 0.24 0.74 0.999± 0.0240 180 0.68 0.00 0.169± 0.00225 90 0.38 0.53 0.156± 0.00350 180 0.24 0.65 0.107± 0.002NB: Uncertainties are the standard errors calculated by linear re-gression of the aggregated rate data from triplicate runs.As discussed in the previous chapter, the potential conversion of photogen-erated Cl to OH via reactions with water itself, requires a more complexkinetic model than the one that has been used in this work. However, a fewcomments can nevertheless be made regarding the experimental observa-tions. As previously seen, Cl– alone accelerates the degradation rate, whileHCO –3 retards the degradation rate. When both Cl– and HCO –3 , the netinfluence is an even greater decrease in the degradation rate than that ofHCO –3 alone. This is consistent with the reactivity of HCO–3 being greaterwith Cl than with OH, as reported in the literature and summarized inTable 1.2.Again, as in the last section, the experimental data indicates a substantialdiscrepancy between the average value of k′ obtained from the centre pointsrelative to the average of k′ from the other factorial points. This curvatureimplies a lack of fit with a linear model. Further investigation of the nonlin-earity will require an augmented experimental design chosen based on thetype of nonlinearity that may be expected.1446.4. Interactions Among Major Solutes0 2 4 6 8 10−3−2−10See caption.[Cl−] = 0 mg L−1[HCO3−] = 0 mg L−1[Cl−] = 50 mg L−1[HCO3−] = 0 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 6.7: Interaction of chloride and bicarbonate in the 185 nm regimewith text-butanol. [tBuOH] = 7 mg L−1 as C and [CBZ]o ' 0.25 µM.NaCl and NaHCO3 used as sources of chloride and bicarbonate. [Cl−] =0 mg L−1, [HCO3−] = 0 mg L−1( ); [Cl−] = 0 mg L−1,[HCO3−] =180 mg L−1( ); [Cl−] = 25 mg L−1, [HCO3−] = 90 mg L−1( ); [Cl−] =50 mg L−1, [HCO3−] = 0 mg L−1( ); [Cl−] = 50 mg L−1, [HCO3−] =180 mg L−1( ).1456.4. Interactions Among Major Solutes6.4.3 Chloride, Sulphate, and BicarbonateThe interactions of Cl–, SO 2–4 , and HCO–3 are next considered with the pres-ence of naturally occurring DOM, again using Suwannee River NOM fromthe IHSS. The influence on the observed degradation rate k′ of the probecompound carbamazepine is investigated. Yet, before reviewing the resultsof those experiments, the 185 nm absorption of DOM must be considered.Unlike tert-butanol, Suwannee River NOM absorption of 185 nm photonswas suspected to be significant at environmentally relevant concentrations.While the optical properties of DOM have been well studied at 254 nm, noinformation to-date has been found relating to 185 nm. The Specific Ultra-violet Absorbance or SUVA at 254 nm, with units of L mg−1 m−1, is a wellused parameter in the field of water treatment, and typically ranges from1 to 4 L mg−1 m−1 in natural waters (APHA, 2012). The SUVA is essen-tially equivalent to the molar absorption coefficient expressed in differentunits. At 254 nm, SUVA correlates to the degree of unsaturated carbon-carbon bonds, with strong correlation to the amount of aromatic content inthe DOM, the dominant chromophore at 254 nm. At 185 nm, photon ab-sorption also occurs at unsaturated carbon-carbon bonds, as well as oxygencontaining functional groups, and is thus expected to be strong.As previously discussed, an indirect kinetic technique was developed tomeasure the absorptive properties of solutes at 185 nm, thus circumvent-ing the limitations of conventional spectrophotometers operating at thatwavelength. The 185 nm SUVA of Suwannee River NOM was found to be6.2 ± 0.5 L mg−1 m−1. This value is higher than that observed at 254 nmconfirming stronger absorption per unit mass at the shorter wavelength.The experiments here used Suwannee River NOM in all solutions, at a con-centration of 4 mg L−1 as C, which represents an absorption coefficient of0.25 cm−1 at 185 nm. A set of solutions with Cl–, SO 2–4 , and HCO–3 formedat 2 × 2 × 2 matrix with concentrations set at either at 0 or 50, 50, and1466.4. Interactions Among Major Solutes180 mg L−1, respectively. All solutions were irradiated in triplicate, withthe observed degradation rate of the probe carbamazepine calculated fromthe aggregate data. Results obtained in the absence of HCO –3 are plotted inFigure 6.8. Results obtained with 180 mg L−1 of HCO –3 are plotted in Figure6.9. Table 6.6 shows the experimental settings used, calculated distributionof absorbed photons among the major solutes (i.e. fi), experimental ratesk′, and relative rates k′/k′o.0 1 2 3 4 5 6−3−2−10[Cl−] = 50 mg L−1[Cl−] = 0 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 6.8: Interaction of chloride and sulphate in the 185 nm regime withSuwannee River NOM and without bicarbonate. [DOM ] = 4 mg L−1 as Cand [CBZ]o ' 0.25 µM. NaCl and Na2SO4 used as sources of chloride andsulphate. All solutions at pH 7. [Cl−] = 0 mg L−1, [SO42−] = 0 mg L−1( );[Cl−] = 0 mg L−1,[SO42−] = 50 mg L−1( ); [Cl−] = 50 mg L−1, [SO42−] =0 mg L−1( ); [Cl−] = 50 mg L−1, [SO42−] = 50 mg L−1( );1476.4. Interactions Among Major Solutes0 1 2 3 4 5 6−1.5−1−0.50[Cl−] = 50 mg L−1Time (min)ln([CBZ] t/[CBZ] o)Figure 6.9: Interaction of chloride and sulphate in the 185 nm regimewith Suwannee River NOM and with bicarbonate. [DOM ] = 4 mg L−1as C, [HCO3−] = 180 mg L−1 and [CBZ]o ' 0.25 µM. NaCl, Na2SO4,and NaHCO3 used as sources of chloride, sulphate and bicarbonate.All solutions at pH 7. [Cl−] = 0 mg L−1, [SO42−] = 0 mg L−1( );[Cl−] = 0 mg L−1,[SO42−] = 50 mg L−1( ); [Cl−] = 50 mg L−1, [SO42−] =0 mg L−1( ); [Cl−] = 50 mg L−1, [SO42−] = 50 mg L−1( );1486.4. Interactions Among Major SolutesSeveral comments can now be made regarding degradation in the 185 nmregime based on the observations above. In the absence of HCO –3 , the pres-ence of Cl– reduces the degradation rate. This can be explained by a higherreactivity of Cl than OH with DOM. In the absence of Cl–, the presenceof SO 2–4 has little effect on k′, despite the fact that the second-order rateconstant for carbamazepine and SO –4 was found to be lower than that ofOH (i.e. kSO –4 ,C < kOH,C). This can be explained by a lower rate constantfor SO –4 with DOM, than for OH. From this information the relative reac-tivities for DOM may be deduced as follows:kCl,DOM > kOH,DOM > kSO –4 ,DOM (6.9)In the presence of HCO –3 , all degradation rates decrease further, with thelowest associated with the presence of Cl–. In the absence of Cl–, the pres-ence of SO 2–4 induces a substantial increase in degradation rate. Theseobservations are rationalized with the relative radical reactivities of HCO –3 having the following pattern:kCl,HCO –3 > kOH,HCO –3 > kSO –4 ,HCO –3 (6.10)consistent with the kSO –4 ,HCO –3 value of Mertens and von Sonntag (1995) of3.5× 106 M−1 s−1, quoted in Table 1.2, but not with that of Buxton et al.(2000) of 9.1× 106 M−1 s−1, which is comparable to the widely cited value ofkOH,HCO –3 = 8.5× 106 M−1 s−1 from the compilation of Buxton et al. (1988).149Table 6.6: Interactions of Cl–, SO 2–4 and HCO–3 , in the presence of Suwannee River NOM at [DOC] =4 mg L−1 as C, in the 185 nm regime[Cl–] [SO 2–4 ] fi = αi/αtot k′ k′/k′oH2O Cl– SO 2–4 HCO–3 DOM(mgL−1) (mgL−1) (min−1)[HCO –3 ] = 0 mg L−1 0 0 0.88 0 0 0 0.12 0.539± 0.011 1.000 50 0.84 0 0.04 0 0.12 0.546± 0.008 1.0150 0 0.26 0.71 0 0 0.04 0.360± 0.006 0.6750 50 0.25 0.70 0.01 0 0.03 0.374± 0.007 0.69[HCO –3 ] = 180 mg L−1 0 0 0.62 0 0 0.29 0.09 0.206± 0.004 0.380 50 0.60 0 0.03 0.29 0.08 0.290± 0.006 0.5450 0 0.23 0.63 0 0.11 0.03 0.125± 0.004 0.2350 50 0.23 0.63 0.01 0.11 0.03 0.134± 0.004 0.25NB: Uncertainties are the standard errors calculated by linear regression of the aggregated rate data from triplicateruns. The fractions of photons fi absorbed by species i are calculated from the experimentally measured molarabsorption coefficients. Carbamazepine used as the probe compound with [CBZ]o ' 0.25 µM. The observed rate k′ocorresponds to a solution with DOM as the only major solute.1506.5. SummaryRegarding the distribution of absorbed photons, it can be seen that the threeanions and the DOM are significant compared to water, and that all thesespecies should be considered in an actual water matrix. These observationsdemonstrate the sensitivity of the degradation kinetics on the compositionof the water matrix.Other species that may also be significant are nitrate (NO –3 ) and ferric iron(Fe3+), and possibly manganese (Mn2+). While these species are generallypresent at concentrations much less than 1 mg L−1, there are cases wherethe concentration may be elevated significantly above typical levels. In suchcases, these species may induce signifiant photochemical and radical effectsin 185 nm regime and merit study.6.5 SummaryThe UV/H2O2 was found to be insensitive to the presence of SO2–4 . In the185 nm regime, SO 2–4 was found to have a significant impact. Experimentaldata was used to calculate the second-order reaction rate of SO –4 with car-bamazepine, estimated to be 1.5± 0.1× 108 M−1 s−1.The influence of HCO –3 was found to manifest by two separate phenom-ena, including the known scavenging of OH, as well as by the absorptionof 185 nm. The latter inner filter or “photon scavenger” mechanism, previ-ously unreported, should be considered in general as the associated fractionof absorbed photons may be substantial in some waters.When considering the interaction of HCO –3 with either SO2–4 or Cl–, themodel used with success in simpler cases proves inadequate in predictingdegradation rates. It is suspected that the discrepancies observed are due tothe interconversion of radicals not taken into account by the current model.Such interconversions merit further investigation.1516.5. SummaryThe three major anions Cl–, SO 2–4 , and HCO–3 , as well as Suwannee RiverNOM reference material, have been analyzed for their optical absorptionconstants at 185 nm using the indirect kinetic method discussed in Chap-ter 5. The molar absorption coefficients are: Cl– = 3540 ± 150 M−1 cm−1,SO 2–4 = 160± 20 M−1 cm−1, HCO –3 = 290± 40 M−1 cm−1. The SUV A185 ofSuwannee River NOM is 6.2± 0.2 L mg−1 m−1.Experiments investigating the interactions of the three major anions in thepresence of Suwannee River NOM allowed the deduction of relative reac-tivities for the three radical species with both Suwannee River NOM andHCO –3 . In both cases, it was found that reactivity follows the order ofCl > OH > SO –4 .These investigations indicate that the 185 nm removal kinetics are highlysensitive to the presence of the major anions and DOM, and that multipleradicals are likely to be involved in trace organic contaminant degradation.152Chapter 7Conclusions andRecommendations7.1 ConclusionsThe conventional low pressure mercury lamp used in UV disinfection has be-come a mature technology and a method of choice for disinfection of chlorineresistant pathogens. UV technology, developed for disinfection, has allowedthe expanded use of UV treatment to the advanced oxidation process (AOP).The UV AOP, using low pressure lamps, uses OH generated by the 254 nmphotolysis of H2O2 to destroy trace organic contaminants. However, H2O2is a poor absorber of photons, and its addition involves increased processcomplexity that scales poorly to small systems. An alternative method ofgenerating OH exploits the second major emission line of the low pressuremercury lamp at 185 nm to generate OH from the photolysis of water it-self. Fundamental aspects of the 185 nm photochemistry relevant to naturalwaters have been investigated and reported by this work for the first time.Experimental techniques have been developed to investigate the 185 nmAOP on a bench scale. A N2 purged collimated beam was designed foruse with commercially available fused silica cells. Carbamazepine was foundto be an ideal probe compound due to negligible direct photolysis at 254 nmand ease of quantification. Tert-Butanol was found to be an ideal modelof dissolved organic carbon, due to its well characterized radical chemistry,availability in pure form, complete miscibility with water, and lack of sig-1537.1. Conclusionsnificant absorbance at either 254 or 185 nm. Its use allowed better study ofmajor inorganic solutes. Phosphate buffers were found to be inappropriate,as they are not photochemically inert at 185 nm and result in reactive radi-cal formation.Temperature effects were studied quantitatively and by comparison betweenthe 254 nm-H2O2 and 185 nm processes. Of the two processes, it was foundthat the 185 nm AOP is less temperature sensitive under the conditionstested, implying that the change in 185 nm absorbance of water with tem-perature imparts a negligible effect. The calculated activation energy forwater photolysis supports a fundamental model of liquid water that includesa small minority of non-hydrogen bonded interstitial monomers. Major so-lutes, particularly Cl–, were not included in the temperature studies, andmerit additional study.The study of Dissolved Organic Matter (DOM), using pure compounds andhumic reference materials, revealed that DOM acts in two independent waysto impede the efficiency of the 185 nm AOP, acting both as a radical scav-enger and an absorber of photons. Furthermore, the use of transparentpure compounds allowed the estimation of 185 nm fluence rates, which areparticularly useful in the absence of convenient actinometers and reliableradiometers. By using tert-butanol and methanol, it was found that the185 nm fluence rate is approximately 16% of that at 254 nm determined byKI-KIO3 actinometry. This is a useful parameter since the amount of 185 nmradiation produced by a low pressure mercury lamp is a critically importantquantity, yet poorly documented and difficult to measure accurately.The role of chloride in water under 185 nm irradiation is perhaps the mostsignificant of all major inorganic solutes. Using a new measurement tech-nique, the molar absorption coefficient of chloride at 185 nm was determinedto be 185,Cl– = 3540±150 M−1 cm−1. When chloride exceeds approximately20 mg L−1, then chloride, and not water, will be the dominant absorber of185 nm photons. Subsequent to photon absorption, the generated chlorine1547.1. Conclusionsradical, Cl , with a quantum yield of approximately 0.4, is capable of react-ing with many trace organics with second-order rate constants comparableto those of OH. In addition to differences in kinetics as compared to purelyOH driven processes, differences in reaction mechanisms likely result in dif-ferent distributions of products in the resulting mixtures, if not differentproducts themselves. The simultaneous generation of the two radicals Cland OH, and the postulated conversion of the former to the latter via reac-tions with water, result in a mixture of oxidative species. The importanceof each one to the oxidation degradation of trace organic contaminants willdepend on both the composition of the water matrix and the relative reac-tivities of the target contaminant.Sulphate was also found to be a significant contributor to degradation via185 nm irradiation. Its molar absorption coefficient was determined to be185,SO 2–4 = 160 ± 20 M−1 cm−1. Photon absorption followed by generationof the sulphate radical, SO –4 , with a quantum yield of approximately 0.65,is similarly able to oxidize many organic molecules with elevated rate con-stants. However, despite the relatively low absorption of SO 2–4 , the presenceof Cl– allows the generation of SO –4 by electron transfer to Cl and vice-versa. An equilibrium constant close to unity for the reaction Cl + SO 2–4⇀↽ Cl– + SO –4 provides an additional link between SO–4 and Cl , andtherefore OH. The mixture of the three radicals further complicates theresulting degradation kinetics and mechanistic pathways. Even when SO –4may be much smaller than the other two, its significance may be amplifiedif its reactivity with the dominant scavengers is relatively low.Bicarbonate was studied and found to impart not only the usual parasitic ef-fect of radical scavenging, but also via 185 nm photon absorption. Its molarabsorption coefficient was determined to be 185,HCO –3 = 290±40 M−1 cm−1.An investigation of the influence of more than two major solutes at a timereveals that significant interactions exist. The overall kinetics depend there-fore, not only on the concentrations of the major solutes, but also on their1557.2. Recommendationsrelative absorption coefficients, and relative reactivities with the water ma-trix and target contaminants. The kinetic model used successfully for in-dividual matrix components fails to properly predict observed degradationrates when more than one major inorganic solute is present. More accuratepredictions of reaction kinetics may be obtained upon improved understand-ing of radical interconversion mechanisms. The information presented hereshould serve as an aide in planning additional experiments.Lastly, for the reader interested in the practical application of the 185 nmAOP, a few comments can be made. As mentioned, one of the greatest ben-efits of the 185 nm AOP is the elimination of the need for storage, addition,and quenching of H2O2. This fact alone may be sufficiently desirable in somecontexts to favour its use. In other cases, a more detailed comparison of the185 nm and 254 nm-H2O2 AOPs requires information on reactor geometry,water composition, target contaminants, and treatment objectives. Such acomparison could incorporate the figure of merit known as the electrical en-ergy per order (EEO) to express the energy required, in units of kW h m−3,to reduce the concentration of a given contaminant by an order of magnitudeunder the conditions used. Electrical and chemical costs of H2O2 additionand quenching would assist operating cost comparisons. Ultimately, anycalculated estimates need to be confirmed by pilot and full scale testing.Such a rigorous analysis is beyond the scope of this work, yet it is hopedthat fundamental information presented here will assist such an endeavour.7.2 RecommendationsResearch on the 185 nm AOP should report the identity and concentrationsof all major anions present in the studied water matrix when specifying theexperimental conditions. A dedicated instrument for accurate measurementof optical absorption at 185 nm will facilitate routine analysis.While the temperature sensitivity of the 185 nm AOP was found to be small,1567.2. Recommendationsthis was determined in the absence of major inorganic solutes. The ab-sorbances and quantum yields of the charge-transfer-to-solvent absorptionbands are reported to be sensitive to temperature. Thus, temperature sen-sitivity should be reinvestigated in the presence of such solutes.A convenient chemical actinometer should be developed to allow expressionof the extent of treatment specified not by exposure time, but by fluence ordose of photons. Such an actinometer would ideally employ a simple methodof analysis of the photochemical products, such as by UV-VIS spectropho-tometry. Such an actinometer could allow the fluence delivered by a flowthrough reactor to be quantified by a reduction equivalent dose correlatedwith a collimated beam experiment.Laser flash photolysis studies would allow the evaluation of the relative in-volvement of OH, Cl , and SO –4 during oxidative degradation in variouswater matrices. Such studies would also facilitate the determination of rateconstants by competitive kinetics for a list of relevant organic contaminantsand the elucidation of radical interconversion mechanisms.Product studies of 185 nm treatment would facilitate the identification ofproduct ratios, as well as the existence of unique products associated withthe three main reactive radicals. If unique products are found, the parentcompounds could serve as useful probes for determining the contribution ofeach radical degradation pathway.A numerical model could be developed with sufficient kinetic informationthat would allow the exact rate expression to be evaluated. Such a modelwould predict the extent of removal for a particular target contaminant giventhe 185 nm photon absorption rate and water matrix composition. Couplingsuch a kinetic model with radiation and hydrodynamics models would servethe ultimate goal of reactor design.157BibliographyJ L Acero and U Von Gunten. Characterization of oxidation processes:Ozonation and the AOP O3/H2O2. Journal - American Water WorksAssociation, 93(10):90–100, 2001.J L Acero, K Stemmler, and U Von Gunten. Degradation kinetics of atrazineand its degradation products with ozone and OH radicals: A predictivetool for drinking water treatment. Environmental Science and Technology,34(4):591–597, 2000.J E Amoore and E Hautala. Odor as an aid to chemical safety: Odor thresh-olds compared with threshold limit values and volatilities for 214 industrialchemicals in air and water dilution. Journal of Applied Toxicology, 3(6):272–290, 1983.M Anbar and J K Thomas. Pulse radiolysis studies of aqueous sodiumchloride solutions. The Journal of Physical Chemistry, 68(12):3829–3835,1964.APHA. Standard Methods for the Examination of Water and Wastewater.American Public Health Association, American Water Works Association,Water Environment Federation, 22nd edition, 2012.AWWA ASCE. Water Treatment Plant Design. McGraw-Hill, 2nd edition,1990.CSSE ASCE, AWWA. Water Treatment Plant Design. American WaterWorks Association, Inc., New York, N.Y., 1969.ASCE AWWA. Water Treatment Plant Design. McGraw-Hill, 3rd edition,1998.158BibliographyASCE AWWA. Water Treatment Plant Design. McGraw-Hill, 5th edition,2012.M N Baker and M J Taras. The Quest for Pure Water: A History of theTwentieth Century. American Water Works Association, 1981.P Ball. Water: An enduring mystery. Nature, 452(7185):291–292, 2008.B T Barnes. Intensities of λ1850 and λ2537 in low-pressure mercury vaporlamps with rare gas present. Journal of Applied Physics, 31(5):852–854,1960.J Barrett and A L Mansell. Ultraviolet absorption spectra of the moleculesH2O, HDO and D2O. Nature, 187:138, 1960.J Barrett, M F Fox, and A L Mansell. The photochemistry of aqueoussulfate ion. The Journal of Physical Chemistry, 69(9):2996–3000, 1965.J H Baxendale and J A Wilson. The photolysis of hydrogen peroxide at highlight intensities. Transactions of the Faraday Society, 53:344–356, 1957.T A Bellar and J J Lichtenberg. Determining volatile organics at microgram-per-litre levels by gas chromatography. Journal - American Water WorksAssociation, pages 739–744, 1974.T A Bellar, J J Lichtenberg, and R C Kroner. The occurrence oforganohalides in chlorinated drinking waters. Journal - American Wa-ter Works Association, pages 703–706, 1974.M J Benotti, R A Trenholm, B J Vanderford, J C Holady, B D Stanford, andS A Snyder. Pharmaceuticals and endocrine disrupting compounds in USdrinking water. Environmental Science and Technology, 43(3):597–603,2009.B Bielski, D Cabelli, R Arudi, and A B Ross. Reactivity of HO2 /O–2 radicalsin aqueous solution. Journal of Physical and Chemical Reference Data,14(4):1041–1100, 1985.159BibliographyM J Blandamer and M F Fox. Theory and applications of charge-transfer-to-solvent spectra. Chemical Reviews, 70(1):59–93, 1970.J R Bolton and C Cotton. The Ultraviolet Disinfection Handbook. AmericanWater Works Association, 2011.J R Bolton and M I Stefan. Fundamental photochemical approach to theconcepts of fluence (UV dose) and electrical energy efficiency in photo-chemical degradation reactions. Research on Chemical Intermediates, 28(7):857–870, 2002.J R Bolton, K G Bircher, W Tumas, and C A Tolman. Figures-of-meritfor the technical development and application of advanced oxidation pro-cesses. Journal of Advanced Oxidation Technologies, 1:13–17, 1996.J R Bolton, M I Stefan, P Shaw, and K R Lykke. Determination of thequantum yields of the potassium ferrioxalate and potassium iodide - iodateactinometers and a method for the calibration of radiometer detectors.Journal of Photochemistry and Photobiology A: Chemistry, 222(1):166–169, 2011.M Born and E Wolf. Principles of Optics: Electromagnetic Theory of Propa-gation, Interference and Diffraction of Light. Cambridge University Press,7th edition, 1999.N V Brodtmann Jr and P J Russo. The use of chloramine for reduction oftrihalomethanes and disinfection of drinking water. Journal - AmericanWater Works Association, pages 40–42, 1979.Z Bukhari, T M Hargy, J R Bolton, B Dussert, and J L Clancy. Medium-pressure UV for oocyst inactivation. Journal - American Water WorksAssociation, 91(3):86, 1999.N J Bunce. Environmental Chemistry. Wuerz Pub., 1990.C Burgess and T Frost, editors. Standards and Best Practice in AbsorptionSpectrometry. Blackwell Science, 1999.160BibliographyJ Butler. Ionic Equilibrium: A Mathematical Approach. Addison-Wesley,MA, 1964.J Butler. Carbon Dioxide Equilibria and their Applications. Addison-Wesley,1982.G V Buxton, C L Greenstock, W P Helman, and A B Ross. Critical reviewof rate constants for reactions of hydrated electrons, hydrogen atoms andhydroxyl radicals OH and O – in aqueous solution. Journal of Physicaland Chemical Reference Data, 17(2):513–886, 1988.G V Buxton, M Bydder, and G A Salmon. Reactivity of chlorine atoms inaqueous solution: Part 1 The equilibrium Cl + Cl– = Cl –2 . Journal ofthe Chemical Society, Faraday Transactions, 94(5):653–657, 1998.G V Buxton, M Bydder, and G A Salmon. The reactivity of chlorine atomsin aqueous solution: Part II The equilibrium SO –4 + Cl– = Cl + SO 2–4 .Physical Chemistry Chemical Physics, 1(2):269–273, 1999.G V Buxton, M Bydder, G A Salmon, and J E Williams. The reactivityof chlorine atoms in aqueous solution: Part III The reactions of Cl withsolutes. Physical Chemistry Chemical Physics, 2(2):237–245, 2000.J G Calvert and J N Pitts. Photochemistry. Wiley, 1966.A T Campbell, L J Robertson, M R Snowball, and H V Smith. Inactivationof oocysts of Cryptosporidium parvum by ultraviolet irradiation. WaterResearch, 29(11):2583–2586, 1995.S Chen and M Z Hoffman. Rate constants for the reaction of the carbon-ate radical with compounds of biochemical interest in neutral aqueoussolution. Radiation Research, 56(1):40–47, 1973.S Chen, M Z Hoffman, and G H Parsons Jr. Reactivity of the carbonateradical toward aromatic compounds in aqueous solution. The Journal ofPhysical Chemistry, 79(18):1911–1912, 1975.161BibliographyZ K Chowdhury, R S Summers, G P Westerhoff, B J Leto, K O Nowack, andC J Corwin. Activated Carbon: Solutions for Improving Water Quality.American Water Works Association, 2013.J L Clancy, T M Hargy, M M Marshall, and J E Dyksen. UV light inac-tivation of Cryptosporidium oocysts. Journal - American Water WorksAssociation, 90(9):92, 1998.J L Clancy, Z Bukhari, T M Hargy, J R Bolton, B W Dussert, and M MMarshall. Using UV to inactivate Cryptosporidium. Journal - AmericanWater Works Association, 92(9):97, 2000.M Clara, B Strenn, and N Kreuzinger. Carbamazepine as a possible an-thropogenic marker in the aquatic environment: Investigations on thebehaviour of carbamazepine in wastewater treatment and during ground-water infiltration. Water Research, 38(4):947–954, 2004.J T Cookson Jr. Virus and water supply. Journal - American Water WorksAssociation, pages 707–711, 1974.J A Cotruvo and C Wu. Controlling organics: Why Now? Journal-AmericanWater Works Association, 70(11):590–594, 1978.S A Craik, G R Finch, J R Bolton, and M Belosevic. Inactivation of Gi-ardia muris cysts using medium-pressure ultraviolet radiation in filtereddrinking water. Water Research, 34(18):4325–4332, 2000.G F Craun, L J McCabe, and J M Hughes. Waterborne disease outbreaksin the US 1971-1974. Journal - American Water Works Association, 68(8):420–424, 1976.J C Crittenden, S Hu, D W Hand, and S A Green. A kinetic model forH2O2/UV process in a completely mixed batch reactor. Water Research,33(10):2315–2328, 1999.G Crozes, P White, and M Marshall. Enhanced coagulation. Journal -American Water Works Association, 87:78–89, 1995.162BibliographyG Czapski, S V Lymar, and H A Schwarz. Acidity of the carbonate radical.The Journal of Physical Chemistry A, 103(18):3447–3450, 1999.F S Dainton and P Fowles. The photolysis of aqueous systems at 1849 a˚II. solutions containing CI–, Br–, SO 2–4 , or OH– ions. In Proc. R. SOC.London A, volume 287, page 3, 1965.B. Darwent. Bond dissociation energies in simple molecules. NIST SpecialPublication, 1(1), 1970.C P Dionigi, T E Lawlor, J E McFarland, and P B Johnsen. Evaluationof geosmin and 2-methylisoborneol on the histidine dependence of TA98and TA100 Salmonella typhimurium tester strains. Water Research, 27(11):1615–1618, 1993.A D Dotson, C E Rodriguez, and K G Linden. UV disinfection implemen-tation status in US water treatment plants. Journal - American WaterWorks Association, 104(5):77–78, 2012.J I Drever. The Geochemistry of Natural Waters. Prentice Hall EnglewoodCliffs, 1988.P L Du Pisani. Direct reclamation of potable water at Windhoek’s Gore-angab reclamation plant. Desalination, 188(1):79–88, 2006.H L DuPont, C L Chappell, C R Sterling, P C Okhuysen, J B Rose, andW Jakubowski. The infectivity of cryptosporidium parvum in healthyvolunteers. New England Journal of Medicine, 332(13):855–859, 1995.M S Elovitz and U von Gunten. Hydroxyl radical/ozone ratios during ozona-tion processes. I. The Rct concept. Ozone Science and Engineering, 1999.B Ervens, S Gligorovski, and H Herrmann. Temperature-dependent rate con-stants for hydroxyl radical reactions with organic compounds in aqueoussolutions. Physical Chemistry Chemical Physics, 5(9):1811–1824, 2003.163BibliographyY Feng, D W Smith, and J R Bolton. Photolysis of aqueous free chlo-rine species (HOCl and OCl–) with 254 nm ultraviolet light. Journal ofEnvironmental Engineering and Science, 6(3):277–284, 2007.L Forni, D Bahnemann, and E J Hart. Mechanism of the hydroxide ion-initiated decomposition of ozone in aqueous solution. The Journal ofPhysical Chemistry, 86(2):255–259, 1982.K R Fox and D A Lytle. Milwaukee’s Crypto outbreak: Investigation andrecommendations. Journal - American Water Works Association, 88(9):87, 1996.M F Fox. The photolysis of simple inorganic anions in solution. QuarterlyReviews, Chemical Society, 24(4):565–584, 1970.M F Fox, B E Barker, and E Hayon. Far-ultraviolet solution spectroscopyof chloride ion. Journal of the Chemical Society, Faraday Transactions 1:Physical Chemistry in Condensed Phases, 74:1776–1785, 1978.H. S. Frank. Structural models. In F Franks, editor, Water a Comprehen-sive Treatise: Volume 1: The Physics and Physical Chemistry of Water,chapter 14. Plenum Press, 1972.F Franks. Water: A Matrix of Life. Royal Society of Chemistry, 2000.N Getoff and G O Schenck. Primary products of liquid water photolysis at1236, 1470 and 1849 a˚. Photochemistry and Photobiology, 8(3):167–178,1968.B C Gilbert, J K Stell, W J Peet, and K J Radford. Generation and reactionsof the chlorine atom in aqueous solution. Journal of the Chemical Society,Faraday Transactions 1: Physical Chemistry in Condensed Phases, 84(10):3319–3330, 1988.W H Glaze, J Kang, and D H Chapin. The chemistry of water treatmentprocesses involving ozone, hydrogen peroxide and ultraviolet radiation.Ozone Science and Engineering, 9:335–352, 1987.164BibliographyW H Glaze, R Schep, W Chauncey, E C Ruth, J J Zarnoch, E M Aieta,C H Tate, and M J McGuire. Evaluating oxidants for the removal ofmodel taste and odor compounds from a municipal water supply. Journal- American Water Works Association), pages 79–84, 1990.C Gleeson and N Gray. The Coliform Index and Waterborne Disease: Prob-lems of microbial drinking water assessment. CRC Press, 1997.J V Goldstone, M J Pullin, S Bertilsson, and B M Voelker. Reactions ofhydroxyl radical with humic substances: Bleaching, mineralization, andproduction of bioavailable carbon substrates. Environmental Science andTechnology, 36(3):364–372, 2002.W R Haag and J Hoigne´. Ozonation of bromide-containing waters: Kineticsof formation of hypobromous acid and bromate. Environmental Scienceand Technology, 17(5):261–267, 1983.W R Haag and C C Yao. Rate constants for reaction of hydroxyl radicalswith several drinking water contaminants. Environmental Science andTechnology, 26(5):1005–1013, 1992.T Hargy. Status of UV disinfection of municipal drinking water systemsin North America. Water Conditioning and Purification International,pages 30–35, 2002.W Harm. Biological Effects of Ultraviolet Radiation. Cambridge UniversityPress, 1980.D C Harris. Quantitative Chemical Analysis. Macmillan, 2010.K Hasegawa and P Neta. Rate constants and mechanisms of reaction of Cl –2radicals. The Journal of Physical Chemistry, 82(8):854–857, 1978.E Hayon and J J McGarvey. Flash photolysis in the vacuum ultravioletregion of sulfate, carbonate, and hydroxyl ions in aqueous solutions. TheJournal of Physical Chemistry, 71(5):1472–1477, 1967.165BibliographyO S Heavens. Optical Properties of Thin Solid Films. Courier Corporation,1991.E Heilker. The Mu¨lheim process for treating Ruhr River water. Journal-American Water Works Association, 71(11):623–627, 1979.Charles P Hibler. Inactivation of Giardia Cysts with Chlorine at 0.50 to5.00 ◦C. American Water Works Association, 1987.J Hoigne´. Inter-calibration of OH radical sources and water quality param-eters. Water Science and Technology, 35(4):1–8, 1997.K J Howe, D W Hand, J C Crittenden, R R Trussell, and G Tchobanoglous.Principles of Water Treatment. John Wiley and Sons, 2012.S E Hrudey and J W A Charrois. Disinfection By-products and HumanHealth. IWA Publishing, 2012.S E Hrudey and E J Hrudey. Safe Drinking Water: Lessons from recentoutbreaks in affluent nations. IWA publishing, 2004.R E Huie, C L Clifton, and P Neta. Electron transfer reaction rates and equi-libria of the carbonate and sulfate radical anions. International Journal ofRadiation Applications and Instrumentation. Part C. Radiation Physicsand Chemistry, 38(5):477–481, 1991.J P Hunt and H Taube. The photochemical decomposition of hydrogenperoxide. Quantum yields, tracer and fractionation effects. Journal of theAmerican Chemical Society, 74(23):5999–6002, 1952.J Jagger. Introduction to Research in Ultraviolet Photobiology. Prentice-Hall, 1967.G G Jayson, B J Parsons, and A J Swallow. Some simple, highly reactive,inorganic chlorine derivatives in aqueous solution. their formation usingpulses of radiation and their role in the mechanism of the Fricke dosime-ter. Journal of the Chemical Society, Faraday Transactions 1: PhysicalChemistry in Condensed Phases, 69:1597–1607, 1973.166BibliographyJ Jin, M G El-Din, and J R Bolton. Assessment of the UV/chlorine processas an advanced oxidation process. Water Research, 45(4):1890–1896, 2011.P D Johnson. Mercury resonance radiation in the high-current, low-pressuredischarge. Journal of the Optical Society of America, 61(11):1451–1454,1971.J Jortner, M Ottolenghi, and G Stein. On the photochemistry of aqueoussolutions of chloride, bromide, and iodide ions. The Journal of PhysicalChemistry, 68(2):247–255, 1964.R Kitamura, L Pilon, and M Jonasz. Optical constants of silica glass fromextreme ultraviolet to far infrared at near room temperature. Appliedoptics, 46(33):8118–8133, 2007.U K Kla¨ning and T Wolff. Laser flash photolysis of HCIO, CIO–, HBrO,and BrO– in aqueous solution. Reactions of Cl and Br atoms. Berichteder Bunsengesellschaft fu¨r Physikalische Chemie, 89(3):243–245, 1985.N V Klassen, D Marchington, and H C E McGowan. H2O2 determinationby the I –3 method and by KMnO4 titration. Analytical Chemistry, 66(18):2921–2925, 1994.P L Knoppert, G Oskam, and E G H Vreenbenburgh. Overview of Europeanwater treatment practice. Journal - American Water Works Association,72(11):592–9, 1980.L R Koller. Ultraviolet Radiation. J. Wiley, 1965.D G Korich, J R Mead, M S Madore, N A Sinclair, and C R Sterling. Effectsof ozone, chlorine dioxide, chlorine, and monochloramine on Cryptosporid-ium parvum oocyst viability. Applied and Environmental Microbiology, 56(5):1423–8, May 1990.I Kristiana, J Charrois, and S E Hrudey. Research overview, regulatoryhistory and current worldwide status of DBP regulations and guidelines.In S E Hrudey and J Charrois, editors, Disinfection By-Products and167BibliographyHuman Health, chapter 2, pages 11–39. International Water Association,2012.L Kro¨ckel and M A Schmidt. Extinction properties of ultrapure water downto deep ultraviolet wavelengths. Optical Materials Express, 4(9):1932–1942, 2014.J C Kruithof, R Van der Leer, and W Hijnen. Practical experiences withUV disinfection in the Netherlands. Journal of Water Supply: Researchand Technology - Aqua, 41(2):88–94, 1992.H J Kuhn, S E Braslavsky, and R Schmidt. Chemical Actinometry (IUPACTechnical Report). Pure and Applied Chemistry, 76(12):2105–2146, 2004.W Ku¨hn, H Sontheimer, L Steiglitz, D Maier, and R Kurz. Use of ozone andchlorine in water utilities in the Federal Republic of Germany. Journal -American Water Works Association, pages 326–331, 1978.J L Kurz. Hydration of acetaldehyde. I. Equilibrium thermodynamic pa-rameters. Journal of the American Chemical Society, 89(14):3524–3528,1967.K J Laidler. Chemical Kinetics. Harper & Row, New York, 3rd edition,1987.S Lalezary, M Pirbazari, and M J McGuire. Oxidation of five earthy-mustytaste and odor compounds. Journal - American Water Works Associa-tion), pages 62–69, 1986.B Langlais, D A Reckhow, and D R Brink. Ozone in Water Treatment:Application and Engineering. CRC Press, 1991.R A Larson and R G Zepp. Reactivity of the carbonate radical with anilinederivatives. Environmental Toxicology and Chemistry, 7(4):265–274, 1988.M W LeChevallier and G Di Giovanni. Source Water Assessment: Vari-ability of Pathogen Concentrations. American Water Works Association,2002.168BibliographyM W LeChevallier and W D Norton. Giarda and Cryptosporidium in rawand finished water. Journal-American Water Works Association, 87(9):54–68, 1995.M W LeChevallier, William D Norton, and R G Lee. Occurrence of Gi-ardia and Cryptosporidium spp. in surface water supplies. Applied andEnvironmental Microbiology, 57(9):2610–2616, 1991.O Legrini, E Oliveros, and A M Braun. Photochemical processes for watertreatment. Chemical Reviews, 93(2):671–698, 1993.W L LePage. The anatomy of an ozone plant. Journal - American WaterWorks Association, pages 105–111, 1981.K G Linden, G Shin, G Faubert, W Cairns, and M D Sobsey. UV disinfectionof Giardia lamblia cysts in water. Environmental Science and Technology,36(11):2519–2522, 2002.R D Lingg, R G Melton, F C Kopfler, W E Coleman, and D E Mitchell.Quantitative analysis of volatile organic compounds by GC-MS. Journal- American Water Works Association, pages 605–612, 1977.W Liu, S A Andrews, J R Bolton, K G Linden, C Sharpless, and M Stefan.Comparison of disinfection byproduct (DBP) formation from different UVtechnologies at bench scale. Water Science and Technology: Water Supply,2(5-6):515–521, 2002.D A Livingstone. Chemical Composition of Rivers and Lakes. US Govern-ment Printing Office, 1963.G S Logsdon and E C Lippy. The role of filtration in preventing water-borne disease. Journal-American Water Works Association, pages 649–655, 1982.W R MacKenzie, N J Hoxie, M E Proctor, M S Gradus, K A Blair, D EPeterson, J J Kazmierczak, D G Addiss, K R Fox, and J B Rose. Amassive outbreak in Milwaukee of Cryptosporidium infection transmitted169Bibliographythrough the public water supply. New England Journal of Medicine, 331(3):161–167, 1994.J P Malley, J P Shaw, and J R Ropp. Evaluation of By-Products Produced byTreatment of Groundwaters with Ultraviolet Irradiation. AWWAR, 1996.P Maruthamuthu and P Neta. Reactions of phosphate radicals with organiccompounds. The Journal of Physical Chemistry, 81(17):1622–1625, 1977.W J Masschelein and R G Rice. Ultraviolet Light in Water and WastewaterSanitation. Lewis Publishers, 2002.R Matta, S Tlili, S Chiron, and S Barbati. Removal of carbamazepine fromurban wastewater by sulfate radical oxidation. Environmental ChemistryLetters, 9(3):347–353, 2011.M J McGuire. Off-flavor as the consumer’s measure of drinking water safety.Water Science and Technology, 31(11):1–8, 1995.J McMurry. Organic Chemistry. Brooks/Cole, 1996.R Mertens and C von Sonntag. Photolysis (λ= 354 nm) of tetrachloroethenein aqueous solutions. Journal of Photochemistry and Photobiology A:Chemistry, 85(1):1–9, 1995.J P Mieure. A rapid and sensitive method for determining volatileorganohalides in water. Journal - American Water Works Association,pages 60–62, 1977.J W Moore and R G Pearson. Kinetics and Mechanism. John Wiley andSons, 3rd edition, 1981.E Mvula, M N Schuchmann, and C von Sonntag. Reactions of phenol-OH-adduct radicals. Phenoxyl radical formation by water elimination vs. oxi-dation by dioxygen. Journal of the Chemical Society, Perkin Transactions2, (3):264–268, 2001.170BibliographyV Nagarajan and R W Fessenden. Flash photolysis of transient radicals. 1.X –2 with X = Cl, Br, I, and SCN. The Journal of Physical Chemistry, 89(11):2330–2335, 1985.P Neta and L M Dorfman. Pulse radiolysis studies. XIII. Rate constantsfor the reaction of hydroxyl radicals with aromatic compounds in aqueoussolutions. Technical report, Ohio State University, Columbus, 1968.P Neta, V Madhavan, H Zemel, and R W Fessenden. Rate constants andmechanism of reaction of sulfate radical anion with aromatic compounds.Journal of the American Chemical Society, 99(1):163–164, 1977.T S Norman, L L Harms, and R W Looyenga. The use of chloraminesto prevent trihalomethane formation. Journal - American Water WorksAssociation, pages 176–180, 1980.L H Nowell and J Hoigne´. Photolysis of aqueous chlorine at sunlight and ul-traviolet wavelengths—II. Hydroxyl radical production. Water Research,26(5):599–605, 1992.N P Page. Report on carcinogenesis bioassay of chloroform. US Dept. ofHealth, Education, and Welfare, Public Health Service, National Insti-tutes of Health], National Cancer Institute, Division of Cancer Cause andPrevention, Carcinogenesis Program, Carcinogen Bioassay and ProgramResources Branch, 1976.G W Pendygraft, Fred E Schlegel, and Michael J Huston. Organics indrinking water: A health perspective. Journal - American Water WorksAssociation, 71(3):118–126, 1979.P Persson. Sensory properties and analysis of two muddy odour compounds,geosmin and 2-methylisoborneol, in water and fish. Water Research, 14(8):1113–1118, 1980.A Peter and U Von Gunten. Oxidation kinetics of selected taste and odorcompounds during ozonation of drinking water. Environmental Scienceand Technology, 41(2):626–631, 2007.171BibliographyB Pett, F Smith, D Stendahl, and R Welker. Cryptosporidiosis outbreakfrom an operations point of view: Kitchener—Waterloo, Ontario, spring1993. In Proceedings of the Water Quality Technology Conference 1993,Miami, 1993.R Phillips. Sources and Applications of Ultraviolet Radiation. AcademicPress, 1983.T I Quickenden and J A Irvin. The ultraviolet absorption spectrum of liquidwater. The Journal of Chemical Physics, 72(8):4416–4428, 1980.R O Rahn. Potassium iodide as a chemical actinometer for 254 nm radiation:use of iodate as an electron scavenger. Photochemistry and Photobiology,66(4):450–455, 1997.R O Rahn, M I Stefan, J R Bolton, E Goren, P Shaw, and K R Lykke.Quantum yield of the iodide - iodate chemical actinometer: dependenceon wavelength and concentration. Photochemistry and Photobiology, 78(2):146–152, 2003.K L Rakness. Ozone in Drinking Water Treatment: Process design, opera-tion, and optimization. American Water Works Association, 2011.M Rapinat. Recent developments in water treatment in France. Journal -American Water Works Association, 74(12):610–7, 1982.D A Reckhow, K G Linden, J Kim, H Shemer, and G Makdissy. Effectof UV treatment on DBP formation. Journal - American Water WorksAssociation, 102(6):100, 2010.R C Rendtorff. The experimental transmission of human intestinal pro-tozoan parasites. II. Giardia lamblia cysts given in capsules. AmericanJournal of Hygiene, 59(2):209–20, 1954.R G Rice and J A Cotruvo. Introductory Comments. In Rip G Rice andJoseph A Cotruvo, editors, Ozone/Chlorine Dioxide Oxidation Products ofOrganic Materials. Proceedings of a conference held in Cincinnati, Ohio,November 17-19, 1976. Ozone Press International, 1978.172BibliographyR G Rice and C M Robson, editors. Biological Activated Carbon– EnhancedAerobic Biological Activity in GAC Systems. Ann Arbor Science, 1982.R G Rice, C M Robson, G W Miller, and A G Hill. Uses of ozone indrinking water treatment. Journal - American Water Works Association,pages 44–57, 1981.P A Roefer, J T Monscvitz, and D J Rexing. The Las Vegas cryptosporid-iosis outbreak. Journal - American Water Works Association, 88(9):95,1996.J J Rook. Formation of haloforms during chlorination of natural waters.Journal of Water Treatment Examination, 23:234–243, 1974.A A Rosen. Foundations of organic pollutant analysis. In L H Keith, editor,Identification and Analysis of Organic Pollutants in Water, chapter 1.Ann Arbor Science, 1976.E J Rosenfeldt, K G Linden, S Canonica, and U Von Gunten. Comparison ofthe efficiency of OH radical formation during ozonation and the advancedoxidation processes O3/H2O2 and UV/H2O2. Water Research, 40(20):3695–3704, 2006.A B Ross and P Neta. Rate Constants for Reactions of Inorganic Radicalsin Aqueous Solution. US Department of Commerce, National Bureau ofStandards Washington D. C, 1979.I M Sayre. International standards for drinking water. Journal-AmericanWater Works Association, 80(1):53–60, 1988.M Schalekamp. The use of GAC filtration to ensure quality in drinkingwater from surface sources. Journal-American Water Works Association,71(11):638–647, 1979.M N Schuchmann and C Von Sonntag. Hydroxyl radical-induced oxidation of2-methyl-2-propanol in oxygenated aqueous solution. a product and pulseradiolysis study. Journal of Physical Chemistry, 83(7):780–784, 1979.173BibliographyP Schulhof. An evolutionary approach to activated carbon treatment.Journal-American Water Works Association, 71(11):648–659, 1979.S M Serkiz and E M Perdue. Isolation of dissolved organic matter from theSuwannee River using reverse osmosis. Water Research, 24(7):911–916,1990.L Sommer. Analytical Absorption Spectrophotometry in the Visible and Ul-traviolet: The Principles, volume 8. Elsevier, 1989.H Sontheimer, E Heilker, M R Jekel, H Nolte, and F H Vollmer. TheMu¨lheim process. Journal - American Water Works Association, pages393–396, 1978.J Staehelin and J Hoigne´. Decomposition of ozone in water: Rate of initi-ation by hydroxide ions and hydrogen peroxide. Environmental Scienceand Technology, 16(10):676–681, 1982.M I Stefan and J R Bolton. Reinvestigation of the acetone degradation mech-anism in dilute aqueous solution by the UV/H2O2 process. EnvironmentalScience and Technology, 33(6):870–873, 1999.M I Stefan, A R Hoy, and J R Bolton. Kinetics and mechanism of thedegradation and mineralization of acetone in dilute aqueous solution sen-sitized by the UV photolysis of hydrogen peroxide. Environmental Science&amp; Technology, 30(7):2382–2390, 1996.D P Stevenson. On the monomer concentration in liquid water. The Journalof Physical Chemistry, 69(7):2145–2152, 1965.W Stumm and J Morgan. Aquatic Chemistry: Chemical Equilibria andRates in Natural Waters. Wiley-Interscience, 3rd edition, 1996.W Stumm, R Schwarzenbach, and L Sigg. From environmental analyticalchemistry to ecotoxicology—a plea for more concepts and less monitoringand testing. Angewandte Chemie International Edition in English, 22(5):380–389, 1983.174BibliographyI H Suffet. An evaluation of activated carbon for drinking water treatment:a national academy of science report. Journal - American Water WorksAssociation, pages 41–50, 1980.I H Suffet, editor. Advances in Taste-and-Odor Treatment and Control.American Water Works Association, 1995.L Sun, E M Perdue, and J F McCarthy. Using reverse osmosis to obtainorganic matter from surface and ground waters. Water Research, 29(6):1471–1477, 1995.J M Symons. A history of the attempted federal regulation requiring GACadsorption for water treatment. Journal - American Water Works Asso-ciation, pages 34–43, 1984.J M Symons, T A Bellar, J K Carswell, J DeMarco, K L Kropp, G G Robeck,D R Seeger, C J Slocum, B L Smith, and A A Stevens. National organicsreconnaissance survey for halogenated organics. Journal-American WaterWorks Association, pages 634–647, 1975.G Tchobanoglous, J Cotruvo, J Crook, E McDonald, A Olivieri, A Salve-son, S Trussell, J Mattingly, J Mosher, and G Vartanian. Develop-ment of Framework for Direct Potable Reuse Guidelines, Results of theNWRI/WateReuse Association Expert Panel. In Proceedings of the WaterEnvironment Federation, number 5, pages 4943–4946. Water EnvironmentFederation, 2015.E M Thurman. Organic Geochemistry of Natural Waters. Martinus Ni-jhoff/Dr W. Junk Publishers, 1985.C Tixier, H P Singer, S Oellers, and S R Mu¨ller. Occurrence and fateof carbamazepine, clofibric acid, diclofenac, ibuprofen, ketoprofen, andnaproxen in surface waters. Environmental Science and Technology, 37(6):1061–1068, 2003.A Treinin and E Hayon. Charge transfer spectra of halogen atoms in wa-ter - Correlation of the electronic transition energies of iodine, bromine,175Bibliographychlorine, hydroxyl, and hydrogen radicals with their electron affinities.Journal of the American Chemical Society, 97(7):1716–1721, 1975.USEPA. Industrial Pollution of the Lower Mississippi River in Louisiana.(as cited by Symons 1984), 1972.USEPA. National Organics Monitoring Survey (NOMS). Technical report,Technical Support Division, Office of Drinking Water, 1978.D H Volman and J C Chen. The photochemical decomposition of hydrogenperoxide in aqueous solutions of allyl alcohol at 2537 a˚. 1. Journal of theAmerican Chemical Society, 81(16):4141–4144, 1959.U Von Gunten. Ozonation of drinking water: Part I. Oxidation kinetics andproduct formation. Water Research, 37(7):1443–1467, 2003.D Wang, J R Bolton, and R Hofmann. Medium pressure UV combined withchlorine advanced oxidation for trichloroethylene destruction in a modelwater. Water Research, 46(15):4677–4686, 2012.P Wardman. Reduction potentials of one-electron couples involving freeradicals in aqueous solution. Journal of Physical and Chemical ReferenceData, 18(4):1637–1755, 1989.K Watanabe and M Zelikoff. Absorption coefficients of water vapor in thevacuum ultraviolet. Journal of the Optical Society of America, 43(9):753–754, 1953.M J Watts and K G Linden. Chlorine photolysis and subsequent OH radicalproduction during UV treatment of chlorinated water. Water Research,41(13):2871–2878, 2007.M J Watts, E J Rosenfeldt, and K G Linden. Comparative OH radicaloxidation using UV-Cl2 and UV-H2O2 processes. Journal of Water Supply:Research and Technology-AQUA, 56(8):469–477, 2007.R P Wayne. Principles and Applications of Photochemistry. Oxford Uni-versity Press, 1988.176BibliographyJ L Weeks, G Meaburn, and S Gordon. Absorption coefficients of liquidwater and aqueous solutions in the far ultraviolet. Radiation Research, 19(3):559–567, 1963.P Westerhoff, G Aiken, G Amy, and J Debroux. Relationships between thestructure of natural organic matter and its reactivity towards molecularozone and hydroxyl radicals. Water Research, 33(10):2265–2276, 1999.P Westerhoff, S P Mezyk, W J Cooper, and D Minakata. Electron pulse ra-diolysis determination of hydroxyl radical rate constants with SuwanneeRiver fulvic acid and other dissolved organic matter isolates. Environ-mental Science and Technology, 41(13):4640–4646, 2007.R G Wetzel. Limnology: Lake and River Ecosystems. Academic Press, 3rdedition, 2001.WHO. Guidelines for Drinking Water Quality: Recommendations. WorldHealth Organization, 2004.A Wolman. 200 years of water service. Journal-American Water WorksAssociation, 68(8):A13–A18, 1976.A Wolman and A E Gorman. The significance of waterborne typhoid feveroutbreaks, 1920-1930. Journal-American Water Works Association, 23(2):160–201, 1931.H Wright, D Gaithuma, M Heath, C Schulz, T Bogan, A Cabaj, M Schmal-weiser, Aand Schmelzer, and J Finegan-Kelly. UV Disinfection KnowledgeBase. Technical report, Water Research Foundation, 2012.W F Young, H Horth, R Crane, T Ogden, and M Arnott. Taste and odourthreshold concentrations of potential potable water contaminants. WaterResearch, 30(2):331–340, 1996.J Zorbist and W Stumm. Chemical dynamics of the Rhine catchment areain Switzerland - Extrapolation to the ”pristine” Rhine River input to theocean. In Proceedings of Review and Workshop on River Inputs to OceanSystems, Food and Agricultural Organization, Rome., 1981.177K Zoschke, H Bo¨rnick, and E Worch. Vacuum-uv radiation at 185 nm inwater treatment - A review. Water research, 52:131–145, 2014.178Appendix AExperimental Data andCalculationsTable A.1: Raw data and calculations for linearity and precision of Carba-mazepine quantification by HPLC.Standard Peak Area - A(µM) Injection 1 Injection 2 Injection 3 Avg σ %RSD0.01 0.0653 0.0684 0.0677 0.0671 0.0016 2.40.02 0.1215 0.1238 0.1237 0.1230 0.0013 1.10.05 0.3113 0.3276 0.3225 0.3205 0.0083 2.60.10 0.6171 0.6171 0.6138 0.6160 0.0019 0.30.15 0.9172 0.9312 0.9239 0.9241 0.0070 0.80.20 1.2344 1.2354 1.2291 1.2330 0.0034 0.30.40 2.4675 2.4721 2.4677 2.4691 0.0026 0.10.60 3.7176 3.6885 3.6919 3.6993 0.0159 0.40.80 4.8992 4.9044 4.9001 4.9012 0.0028 0.11.00 6.0803 6.0930 6.0672 6.0802 0.0129 0.2179Table A.2: The concentration dependent absorbance of three compounds at 254 nm in 1.00 cm cell.[pCBA] A254 [CBZ] A254 [NB] A254(×10−4M) Run 1 Run 2 Run 3 (×10−4M) Run 1 Run 2 Run 3 (×10−4M) Run 1 Run 2 Run 30.60 0.238 0.247 0.249 0.378 0.248 0.244 0.250 0.465 0.258 0.258 0.2570.90 0.347 0.339 0.356 0.756 0.487 0.490 0.505 0.698 0.434 0.434 0.4341.20 0.452 0.462 0.448 1.130 0.724 0.729 0.739 0.931 0.567 0.565 0.5651.50 0.555 0.571 0.575 1.510 0.995 0.994 0.977 1.160 0.713 0.714 0.7121.80 0.664 0.649 0.666 1.890 1.257 1.301 1.269 1.400 0.865 0.871 0.8681.860 1.134 1.134 1.133180Appendix A. Experimental Data and CalculationsTable A.3: Raw data for direct photolysis of para-chlorobenzoic acid at 254nm.Peak Area AtTime Run 1 Run 2 Run 3(min) Light Dark Light Dark Light Dark0 1.6215 1.5844 1.6416 1.6554 1.7010 1.718110 1.5762 1.5865 1.6542 1.6316 1.704020 1.5455 1.6084 1.5466 1.6528 1.5925 1.709030 1.4900 1.5055 1.5057 1.6626 1.5433 1.717340 1.4459 1.6225 1.4627 1.6597 1.5045 1.721550 1.4078 1.6074 1.4170 1.6752 1.4563 1.717460 1.3704 1.6239 1.4285 1.7246These are some notes for this part about direct photolysis of para-chlorobenzoicacid. The fluence rate for this data is 0.0493 plus or minus 0.0006 mW per cm2.Table A.4: Calculated data for direct photolysis of para-chlorobenzoic acid at 254nm.ln(At/Ao)Time Run 1 Run 2 Run 3(min) Light Dark Light Dark Light Dark0 0.0000 0.0000 0.0000 0.0000 0.0000 0.000010 -0.0283 -0.0341 -0.0007 -0.0417 -0.008220 -0.0480 0.0150 -0.0596 -0.0016 -0.0659 -0.005330 -0.0846 -0.0511 -0.0864 0.0043 -0.0973 -0.000540 -0.1146 0.0238 -0.1154 0.0026 -0.1228 0.002050 -0.1413 0.0144 -0.1471 0.0119 -0.1553 -0.000460 -0.1682 0.0246 -0.1746 0.0038These are some notes for this part about direct photolysis of para-chlorobenzoic acid.181Appendix A. Experimental Data and CalculationsTable A.5: Raw data for direct photolysis of nitrobenzene at 254 nm.Peak Area AtTime Run 1 Run 2 Run 3(min) Light Dark Light Dark Light Dark0 0.6492 0.6492 0.6469 0.6469 0.6567 0.656710 0.5889 0.6114 0.5918 0.6333 0.5883 0.628520 0.5396 0.5969 0.5342 0.6040 0.5258 0.527930 0.4917 0.5765 0.4868 0.5968 0.4791 0.582040 0.4656 0.5602 0.4423 0.5773 0.4341 0.563450 0.4236 0.5410 0.4112 0.5555 0.3863 0.5467Note that nitrobenzene dark reaction is significant due to volatilization from opendish during the course of the irradiations.Table A.6: Calculations for direct photolysis of nitrobenzene at 254 nm.ln(At/Ao)Time Run 1 Run 2 Run 3(min) Light Dark Light Dark Light Dark0 0.0000 0.0000 0.0000 0.0000 0.0000 0.000010 -0.0975 -0.0600 -0.0890 -0.0212 -0.1100 -0.043920 -0.1849 -0.0840 -0.1914 -0.0686 -0.222330 -0.2779 -0.1188 -0.2843 -0.0806 -0.3153 -0.120840 -0.3324 -0.1474 -0.3802 -0.1138 -0.4140 -0.153250 -0.4270 -0.1823 -0.4531 -0.1523 -0.5306 -0.1833A correction for the rate of volatilization is obtained from the dark reaction and appliedto light reaction for determination of the photolysis rate constant.182Appendix A. Experimental Data and CalculationsTable A.7: Raw and calculated data fordirect photolysis of carbamazepine acid at254 nm.Time Fluence At ln(At/Ao)(min) (mJ cm−2)0 0 5.700 0.000060 1031 5.534 -0.029660 1031 5.537 -0.0290120 2062 5.281 -0.0763120 2062 5.453 -0.0443180 3094 5.303 -0.0721180 3094 5.345 -0.0642240 4125 5.197 -0.0923300 5156 4.809 -0.1699420 7218 4.650 -0.2037480 8249 4.848 -0.1618540 9281 4.470 -0.2431600 10312 4.649 -0.2038These are some notes for this part about car-bamazepine photolysis quantum yield at 254nm if needed. For example the fluence ratewas 0.286 +/- 0.001 mW/cm2.183Appendix A. Experimental Data and CalculationsTable A.8: Raw data for indirect photolysis of para-chlorobenzoic acid at254 nm with H2O2.Peak Area AtTime Run 1 Run 2 Run 3(min) Light Dark Light Dark Light Dark0 1.6059 1.6097 1.6496 1.6476 1.6326 1.628310 1.2699 1.6162 1.3245 1.6414 1.2931 1.634920 1.0764 1.6113 1.0726 1.6500 1.0396 1.628030 0.8992 1.6156 0.8729 1.6631 0.8214 1.643440 0.7593 1.6261 0.7134 1.6585 0.6627 1.633350 0.6539 1.6215 0.5788 1.6648 0.5313 1.651660 0.5579 1.6266 0.4685 1.6632 1.6464Table A.9: Calculations for indirect photolysis of para-chlorobenzoic acid at254 nm with H2O2.ln(At/Ao)Time Run 1 Run 2 Run 3(min) Light Dark Light Dark Light Dark0 0.0000 0.0000 0.0000 0.0000 0.0000 0.000010 -0.2347 0.0040 -0.2195 -0.0038 -0.2331 0.004020 -0.4001 0.0010 -0.4304 0.0015 -0.4513 -0.000230 -0.5799 0.0037 -0.6365 0.0094 -0.6869 0.009240 -0.7490 0.0101 -0.8382 0.0066 -0.9016 0.003150 -0.8985 0.0073 -1.0473 0.0104 -1.1226 0.014260 -1.0573 0.0104 -1.2588 0.0094 0.0111184Appendix A. Experimental Data and CalculationsTable A.10: Raw data for indirect photolysis of carbamazepine at 254 nmwith H2O2.Peak Area AtTime Run 1 Run 2 Run 3(min) Light Dark Light Dark Light Dark0 1.6411 1.6496 1.6737 1.6754 1.6461 1.642610 1.2421 1.6419 1.2994 1.6732 1.2646 1.658320 1.0319 1.6622 1.0211 1.6734 0.9876 1.659130 0.8421 1.6665 0.8009 1.6798 0.7608 1.663640 0.6928 1.6663 0.6294 1.6826 0.5825 1.653650 0.5774 1.6557 0.4895 1.6853 0.4423 1.665560 0.4895 1.6719 0.3906 1.6904 1.6710Table A.11: Calculations for indirect photolysis of carbamazepine at 254 nmwith H2O2.ln(At/Ao)Time Run 1 Run 2 Run 3(min) Light Dark Light Dark Light Dark0 0.0000 0.0000 0.0000 0.0000 0.0000 0.000010 -0.2786 -0.0047 -0.2531 -0.0013 -0.2637 0.009520 -0.4640 0.0076 -0.4942 -0.0012 -0.5109 0.010030 -0.6672 0.0102 -0.7371 0.0026 -0.7718 0.012740 -0.8624 0.0101 -0.9780 0.0043 -1.0388 0.006750 -1.0446 0.0037 -1.2294 0.0059 -1.3142 0.013860 -1.2097 0.0134 -1.4551 0.0089 0.0171185Appendix A. Experimental Data and CalculationsTable A.12: Raw data and calculations for phosphate buffer (10 mM) at 185 nmusing Carbamazepine.Phosphate Buffer (10 mM)Time pH 6.0 pH 7.0 pH 8.0(min) At ln (At/Ao) At ln (At/Ao) At ln (At/Ao)0 1.8221 0.0000 1.8263 0.0000 1.8697 0.00002 0.8871 -0.7198 0.8159 -0.7855 0.6999 -0.98264 0.4933 -1.3066 0.4219 -1.5033 0.3161 -1.77756 0.2759 -1.8877 0.2292 -2.1642 0.1650 -2.4276These are some notes for this part.186Table A.13: Influence of phosphate buffer strength at pH 7 at 185 nm using carbamazepine as probe.pH 7.0Time 0 mM* 1 mM 10 mM 100 mM(min) At ln (At/Ao) At ln (At/Ao) At ln (At/Ao) At ln (At/Ao)0 1.9324 0.0000 1.8504 0.0000 1.8263 0.0000 1.8426 0.00002 1.3569 -0.3536 1.2211 -0.4156 0.8159 -0.8058 0.8400 -0.78554 0.9282 -0.7333 0.8140 -0.8212 0.4219 -1.4653 0.4098 -1.50336 0.6764 -1.0497 0.5353 -1.2403 0.2292 -2.0755 0.2116 -2.1642* Ultrapure water pH 5.4.187Appendix A. Experimental Data and CalculationsTable A.14: Influence of fluoride in 185 nm regime[F−] Time At ln(At/Ao)(mg/L) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 31 0 2.8736 2.8631 2.8933 0.000 0.000 0.0002 1.3600 1.3221 1.3349 -0.748 -0.773 -0.7744 0.6643 0.5743 0.6079 -1.465 -1.607 -1.5606 0.3207 0.2788 0.2580 -2.193 -2.329 -2.4178 0.1392 0.1125 0.1179 -3.027 -3.237 -3.20010 0.1038 0.0573 0.0553 -3.321 -3.911 -3.95710 0 2.9734 2.9612 2.9923 0.000 0.000 0.0002 1.5725 1.3576 1.3556 -0.637 -0.780 -0.7924 0.6700 0.5779 0.5935 -1.490 -1.634 -1.6186 0.3383 0.2842 0.2749 -2.174 -2.344 -2.3878 0.1395 0.1193 0.1330 -3.059 -3.212 -3.11310 0.0811 0.0567 0.0567 -3.602 -3.956 -3.966100 0 2.5477 2.5391 2.5281 0.000 0.000 0.0002 1.2020 1.1278 1.1811 -0.751 -0.812 -0.7614 0.5639 0.5299 0.5261 -1.508 -1.567 -1.5706 0.2640 0.2257 0.2268 -2.267 -2.420 -2.4118 0.1094 0.1027 0.1049 -3.148 -3.208 -3.18210 0.0592 0.0508 0.0450 -3.762 -3.912 -4.029NaF used as source of F– . Tert-Butanol used as a radical scavenger(7.0mgL−1 as C) and carbamazepine as probe compound ([CBZ]i = 0.5 µM).188Table A.15: Demonstration of high absorbance of water at 185 nm.1.00 cm cell 2.00 cm cell(Volume = 3.097 mL) (Volume = 5.617 mL)Time Time/Volume At ln (At/Ao) Time/Volume At ln (At/Ao)(min) (min/mL) (min/mL)0 0.00 6.0856 0.000 0.00 6.0856 0.0002 0.65 2.0962 -1.066 0.36 4.8199 -0.2334 1.29 0.8695 -1.946 0.71 2.0839 -1.0726 1.94 0.5578 -2.390 1.07 1.2716 -1.5668 2.58 0.1663 -3.600 1.42 0.8126 -2.01310 3.23 0.1023 -4.086 1.78 0.5520 -2.400This demonstrations involves the assumption that the number of photons absorbed by the sampleis proportional to the exposure time and that both the 1 and 2 cm cells are sufficiently mixed.Thus the quantity Time/Volume is proportional to the fluence. Plotting using the Time/Volumenot Time as the abscissa will thus show agreement between both the 1 and 2 cm cells.189Appendix A. Experimental Data and CalculationsTable A.16: Raw data and calculations of I– - IO3– actinometry study ap-plied to 254 nm.Run A352 Io A352 It(cm−1) (mW cm−2) (cm−1) (mW cm−2)1 0.888 0.518 0.821 0.4782 0.854 0.498 0.844 0.4923 0.879 0.512 0.854 0.4984 0.873 0.509 0.853 0.4975 0.890 0.519 0.818 0.4766 0.830 0.483 0.826 0.4817 0.854 0.498 0.836 0.4878 0.865 0.504 0.828 0.4829 0.836 0.487 0.824 0.48010 0.831 0.484 0.832 0.485Table A.17: Raw data and calculations of I– - IO3– actinometry applied to185 and 254 nm.Run A352 Io A352 It(cm−1) (mW cm−2) (cm−1) (mW cm−2)1 0.856 0.907 0.819 0.86722 0.831 0.880 0.775 0.81993 0.808 0.855 0.794 0.84034 0.827 0.876 0.795 0.84145 0.793 0.839 0.793 0.8392190Appendix A. Experimental Data and CalculationsTable A.18: Influence of temperature in 254 nm - H2O2 regimeT Time At ln(At/Ao)(◦C) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 35 0 0.9022 0.9041 0.8973 0.000 0.000 0.00010 0.8597 0.8760 0.8452 -0.048 -0.032 -0.06020 0.7937 0.8087 0.8147 -0.128 -0.112 -0.09730 0.7667 0.7340 0.7318 -0.163 -0.208 -0.20440 0.7021 0.6999 0.6687 -0.251 -0.256 -0.29450 0.6537 0.6466 0.6030 -0.322 -0.335 -0.39760 0.5858 0.6076 0.5699 -0.432 -0.397 -0.45420 0 0.9259 0.9124 0.9109 0.000 0.000 0.00010 0.8369 0.8130 0.8341 -0.101 -0.115 -0.08820 0.7636 0.7281 0.7244 -0.193 -0.226 -0.22930 0.6517 0.6603 0.6453 -0.351 -0.323 -0.34540 0.5868 0.5824 0.5616 -0.456 -0.449 -0.48450 0.5392 0.5324 0.5132 -0.541 -0.539 -0.57460 0.4705 0.4562 0.4621 -0.677 -0.693 -0.67935 0 0.9208 0.9238 0.9229 0.000 0.000 0.00010 0.7915 0.8014 0.8414 -0.151 -0.142 -0.09220 0.6914 0.7021 0.7142 -0.287 -0.274 -0.25630 0.5999 0.5900 0.6480 -0.428 -0.448 -0.35440 0.5072 0.4870 0.5377 -0.596 -0.640 -0.54050 0.4259 0.3966 0.4420 -0.771 -0.846 -0.73660 0.3552 0.3383 0.3710 -0.953 -1.005 -0.911Notes regarding temperature study at 254 nm. Tert-Butanol used as aradical scavenger (7.0 mg L−1 as C) and carbamazepine as probe com-pound ([CBZ]i = 0.5 µM) with detection at 267 nm.191Appendix A. Experimental Data and CalculationsTable A.19: Influence of temperature in 185 nm regimeT Time At ln(At/Ao)(◦C) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 35 0 0.9385 0.9515 0.9397 0.000 0.000 0.0002 0.5680 0.5578 0.5619 -0.502 -0.534 -0.5144 0.3478 0.3669 0.3483 -0.993 -0.953 -0.9926 0.2057 0.2251 0.2121 -1.518 -1.441 -1.4898 0.1254 0.1266 0.1335 -2.013 -2.017 -1.95110 0.0817 0.0752 0.0816 -2.441 -2.538 -2.44420 0 0.9415 0.9334 0.9272 0.000 0.000 0.0002 0.5149 0.4749 0.5043 -0.604 -0.676 -0.6094 0.3007 0.2901 0.2754 -1.141 -1.169 -1.2146 0.1479 0.1492 0.1698 -1.851 -1.834 -1.6988 0.0745 0.0760 0.0837 -2.537 -2.508 -2.40510 0.0452 -3.02135 0 0.9602 0.9109 0.9364 0.000 0.000 0.0002 0.4755 0.4611 0.4818 -0.703 -0.681 -0.6654 0.2472 0.2194 0.2433 -1.357 -1.424 -1.3486 0.1163 0.1019 0.1224 -2.111 -2.190 -2.0358 0.0585 0.0603 0.0547 -2.798 -2.715 -2.84010 0.0238 0.0124 -3.645 -4.324Notes regarding temperature study. Tert-Butanol used as a radicalscavenger (7.0 mg L−1 as C) and carbamazepine as probe compound([CBZ]i = 0.5 µM) and detection at 267 nm.192Appendix A. Experimental Data and CalculationsTable A.20: Influence of Suwannee River NOM in 185 nm regime[DOC] Time At ln(At/Ao)(mg L−1 as C) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 33.0 0 3.0227 3.0488 3.0278 0.000 0.000 0.0001 1.4541 1.4178 1.4747 -0.732 -0.766 -0.7192 0.7175 0.6777 0.6741 -1.438 -1.504 -1.5023 0.2676 0.3012 0.3204 -2.424 -2.315 -2.2464 0.1173 0.1154 0.1235 -3.249 -3.274 -3.1995 0.0863 0.0401 0.0507 -3.556 -4.331 -4.0905.0 0 1.4844 1.4936 1.4783 0.000 0.000 0.0001 0.8926 0.8895 0.8893 -0.509 -0.518 -0.5082 0.5184 0.5168 0.5139 -1.052 -1.061 -1.0573 0.2914 0.3157 0.2985 -1.628 -1.554 -1.6004 0.1626 0.1793 0.1695 -2.211 -2.120 -2.1665 0.0978 0.0863 0.0992 -2.720 -2.851 -2.7027.5 0 1.4602 1.4846 1.4520 0.000 0.000 0.0001 1.0553 1.0607 1.0361 -0.325 -0.336 -0.3372 0.7594 0.7394 0.7425 -0.654 -0.697 -0.6713 0.4935 0.5143 0.5272 -1.085 -1.060 -1.0134 0.3687 0.3589 0.3522 -1.376 -1.420 -1.4165 0.2528 0.2607 0.2426 -1.754 -1.740 -1.78910.0 0 1.4541 1.4646 1.4499 0.000 0.000 0.0001 1.1266 1.1762 1.1590 -0.255 -0.219 -0.2242 0.8972 0.9049 0.9009 -0.483 -0.482 -0.4763 0.6940 0.6994 0.6964 -0.740 -0.739 -0.7334 0.5599 0.5442 0.5491 -0.954 -0.990 -0.9715 0.4448 0.4175 0.3994 -1.185 -1.255 -1.289Carbamazepine as probe compound ([CBZ]o ' 0.5 µM) with detection at 211 nm.Solutions irradiated at pH 7.0.193Appendix A. Experimental Data and CalculationsTable A.21: Influence of Nordic NOM in 185 nm regime[DOC] Time At ln(At/Ao)(mg L−1 as C) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 31.8 0 2.9128 2.8966 2.9313 0.000 0.000 0.0001 1.1507 1.1362 1.0993 -0.929 -0.936 -0.9812 0.3306 0.3467 0.3768 -2.176 -2.123 -2.0512.7 0 2.8865 2.8914 2.8838 0.000 0.000 0.0001 1.3443 1.3751 1.4256 -0.764 -0.743 -0.7052 0.6755 0.6676 0.6973 -1.452 -1.466 -1.4203 0.2937 0.2620 0.2953 -2.285 -2.401 -2.2794 0.0921 0.1308 0.1198 -3.445 -3.096 -3.1815.2 0 2.9134 2.8901 2.8930 0.000 0.000 0.0001 1.9718 1.9752 1.9533 -0.390 -0.381 -0.3932 1.3019 1.3528 1.3326 -0.805 -0.759 -0.7753 0.8936 0.8839 0.8489 -1.182 -1.185 -1.2264 0.5786 0.5262 0.5534 -1.616 -1.703 -1.6545 0.4180 0.3561 0.3573 -1.942 -2.094 -2.0919.8 0 2.8935 2.8968 2.8878 0.000 0.000 0.0001 2.3711 2.3838 2.3705 -0.199 -0.195 -0.1972 1.9334 1.9556 1.9425 -0.403 -0.393 -0.3973 1.5938 1.5683 1.5770 -0.596 -0.614 -0.6054 1.3123 1.3016 1.3085 -0.791 -0.800 -0.7925 1.0559 1.0663 1.0377 -1.008 -0.999 -1.023Carbamazepine as probe compound ([CBZ]o ' 0.5 µM) with detection at 211 nm.Solutions irradiated at pH 7.0.194Appendix A. Experimental Data and CalculationsTable A.22: Influence of tert-butanol in 185 nm regime[(CH3)3COH] Time At ln(At/Ao)(mg L−1 as C) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 35 0 2.9737 2.9641 2.9671 0.000 0.000 0.0002 1.1674 1.1704 1.2115 -0.935 -0.929 -0.8964 0.4522 0.4366 0.4593 -1.883 -1.915 -1.8666 0.1691 0.1768 0.1877 -2.867 -2.819 -2.7608 0.0621 -3.86610 0 2.9135 2.9171 2.8943 0.000 0.000 0.0002 1.7099 1.7079 1.7197 -0.533 -0.535 -0.5214 1.0165 0.9984 1.0223 -1.053 -1.072 -1.0416 0.5927 0.6228 0.6059 -1.592 -1.544 -1.5648 0.3463 0.3484 0.3443 -2.130 -2.125 -2.12910 0.2007 0.2095 0.1816 -2.675 -2.634 -2.76915 0 2.9266 2.9187 2.9355 0.000 0.000 0.0002 2.0576 1.9928 2.0291 -0.352 -0.382 -0.3694 1.4345 1.4447 1.4563 -0.713 -0.703 -0.7016 0.9931 1.0220 1.0054 -1.081 -1.049 -1.0718 0.6780 0.7022 0.6933 -1.462 -1.425 -1.44310 0.4702 0.4997 0.4690 -1.828 -1.765 -1.83420 0 2.9182 2.9332 2.9485 0.000 0.000 0.0002 2.212 2.2059 2.2089 -0.277 -0.285 -0.2894 1.6493 1.6881 1.7212 -0.571 -0.552 -0.5386 1.2439 1.3 1.2902 -0.853 -0.814 -0.8268 0.9564 0.9898 0.976 -1.116 -1.086 -1.10610 0.7237 0.7366 0.7449 -1.394 -1.382 -1.37625 0 2.8522 2.8601 2.8757 0.000 0.000 0.0002 2.2437 2.2532 2.2763 -0.240 -0.239 -0.2344 1.8091 1.7793 1.7675 -0.455 -0.475 -0.4876 1.4376 1.3782 1.3926 -0.685 -0.730 -0.7258 1.1257 1.1002 1.0885 -0.930 -0.955 -0.97110 0.8590 0.8523 0.8812 -1.200 -1.211 -1.183Carbamazepine as probe compound ([CBZ]o ' 0.5µM) with detection at 211 nm.195Appendix A. Experimental Data and CalculationsTable A.23: Influence of methanol in 185 nm regime[CH3OH] Time At ln(At/Ao)(mg L−1 as C) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 31.0 0 1.9369 1.9463 1.9471 0.000 0.000 0.0002 1.0033 1.0301 0.9827 -0.658 -0.636 -0.6844 0.4701 0.4902 0.4947 -1.416 -1.379 -1.3706 0.2145 0.2297 0.2274 -2.201 -2.137 -2.1478 0.1191 -2.7942.0 0 1.9274 1.9095 1.9278 0.000 0.000 0.0002 1.2889 1.3061 1.2967 -0.402 -0.380 -0.3974 0.8866 0.8736 0.8339 -0.777 -0.782 -0.8386 0.5538 0.5925 0.5821 -1.247 -1.170 -1.1978 0.392 0.3933 0.3618 -1.593 -1.580 -1.67310 0.2591 0.2502 0.2649 -2.007 -2.032 -1.9853.0 0 1.9346 1.9250 1.9408 0.000 0.000 0.0002 1.4608 1.4466 1.4550 -0.281 -0.286 -0.2884 1.1076 1.0918 1.1130 -0.558 -0.567 -0.5566 0.8130 0.8310 0.8273 -0.867 -0.840 -0.8538 0.6295 0.5661 0.5965 -1.123 -1.224 -1.18010 0.4560 0.4713 0.4429 -1.445 -1.407 -1.4784.0 0 1.9229 1.9063 1.9086 0.000 0.000 0.0002 1.5254 1.5283 1.5098 -0.232 -0.221 -0.2344 1.2212 1.2028 1.1973 -0.454 -0.461 -0.4666 0.9683 0.9537 0.9517 -0.686 -0.693 -0.6968 0.7546 0.7415 0.7679 -0.935 -0.944 -0.91010 0.6043 0.6001 0.6010 -1.158 -1.156 -1.1565.0 0 1.9223 1.9189 1.9170 0.000 0.000 0.0002 1.5704 1.5890 1.6170 -0.202 -0.189 -0.1704 1.3257 1.3336 1.3030 -0.372 -0.364 -0.3866 1.0965 1.0871 1.1035 -0.561 -0.568 -0.5528 0.9189 0.9167 0.9311 -0.738 -0.739 -0.72210 0.7528 0.7529 0.7480 -0.937 -0.936 -0.941Carbamazepine as probe compound ([CBZ]o ' 0.3µM) with detection at 211 nm.196Appendix A. Experimental Data and CalculationsTable A.24: Influence of acetone in 185 nm regime[(CH3)2CO] Time At ln(At/Ao)(mg L−1 as C) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 360 0.0 1.5395 1.5135 1.4911 0.000 0.000 0.0000.5 0.9816 0.9826 0.9594 -0.450 -0.432 -0.4411.0 0.6849 0.6570 0.6329 -0.810 -0.834 -0.8571.5 0.4315 0.4342 0.4239 -1.272 -1.249 -1.2582.0 0.2886 0.2881 0.2706 -1.674 -1.659 -1.7072.5 0.1793 0.2061 0.2413 -2.150 -1.994 -1.8213.0 0.1148 0.1247 0.1341 -2.596 -2.496 -2.409125 0.0 1.4909 1.4893 1.5034 0.000 0.000 0.0001.0 0.8842 0.8873 0.8836 -0.522 -0.518 -0.5312.0 0.5236 0.5050 0.5394 -1.046 -1.082 -1.0253.0 0.3155 0.3353 0.3241 -1.553 -1.491 -1.5344.0 0.1942 0.2133 0.1970 -2.038 -1.943 -2.0325.0 0.1070 0.1189 0.1112 -2.634 -2.528 -2.6046.0 0.0595 0.0730 0.0902 -3.221 -3.016 -2.813240 0.0 1.3255 1.3421 1.3374 0.000 0.000 0.0001.0 0.9629 0.9524 0.9473 -0.320 -0.343 -0.3452.0 0.6916 0.6986 0.6785 -0.651 -0.653 -0.6793.0 0.4899 0.5036 0.5066 -0.995 -0.980 -0.9714.0 0.3639 0.3741 0.3804 -1.293 -1.277 -1.2576.0 0.1907 0.1982 0.1890 -1.939 -1.913 -1.9578.0 0.0960 0.1018 0.1102 -2.625 -2.579 -2.496480 0.0 1.4296 1.4203 1.4181 0.000 0.000 0.0001.0 1.1511 1.1365 1.1411 -0.217 -0.223 -0.2172.0 0.9395 0.9423 0.9556 -0.420 -0.410 -0.3953.0 0.7779 0.7883 0.7801 -0.609 -0.589 -0.5984.0 0.6480 0.6469 0.6249 -0.791 -0.786 -0.8196.0 0.4464 0.4412 0.4437 -1.164 -1.169 -1.1628.0 0.3075 0.3093 0.3027 -1.537 -1.524 -1.5441025 0.0 1.3899 1.3977 1.4063 0.000 0.000 0.0001.0 1.2097 1.2155 1.2241 -0.139 -0.140 -0.1392.0 1.0826 1.0570 1.0700 -0.250 -0.279 -0.2733.0 0.9648 0.9341 0.9639 -0.365 -0.403 -0.3784.0 0.8394 0.8585 0.8486 -0.504 -0.487 -0.5056.0 0.6572 0.6715 0.6797 -0.749 -0.733 -0.7278.0 0.5134 0.5367 0.5509 -0.996 -0.957 -0.937Carbamazepine as probe compound ([CBZ]o ' 0.25µM) with detection at 211 nm.197Table A.25: Decrease in 1.0 cm absorbance at 254 nm (A0) for a solution of Suwannee River NOM followingexposure by 185 and 254 nm radiation (At).[Cl−] Time Run 1 Run 2 Run 3(mg L−1) (min) A0 At At/Ao A0 At At/Ao A0 At At/Ao< 1 5 0.123 0.102 0.829 0.123 0.104 0.846 0.126 0.102 0.81010 0.122 0.085 0.697 0.129 0.090 0.698 0.123 0.087 0.70720 0.122 0.057 0.467 0.125 0.060 0.480 0.123 0.062 0.504100 5 0.126 0.105 0.833 0.125 0.106 0.848 0.125 0.107 0.85610 0.124 0.090 0.726 0.124 0.090 0.726 0.126 0.090 0.71420 0.126 0.063 0.500 0.125 0.063 0.504 0.125 0.060 0.480Solution concentration of Suwannee River NOM is 3.5 mg L−1 as C.198Table A.26: Influence of chloride in 254 nm - H2O2 regime with Suwannee River NOM and using carbamazepineas probe[Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 4 Run 5 Run 1 Run 2 Run 3 Run 4 Run 5< 1 0 1.3528 1.3584 1.4248 1.4007 1.4283 0.0000 0.0000 0.0000 0.0000 0.00005 1.2392 1.1675 1.2770 1.2739 1.2716 -0.0877 -0.1095 -0.0949 -0.116210 1.1675 1.1264 1.1541 1.1747 1.1447 -0.1473 -0.1873 -0.2107 -0.1760 -0.221315 1.0600 1.0686 1.0459 1.0701 1.0586 -0.2439 -0.2400 -0.3092 -0.2692 -0.299520 0.9527 0.9829 0.9528 0.9793 0.9747 -0.3506 -0.3236 -0.4024 -0.3579 -0.382125 0.8902 0.9150 0.8496 0.8855 0.8836 -0.4185 -0.3951 -0.5170 -0.4586 -0.480230 0.8403 0.8375 0.7799 0.8095 0.7860 -0.4762 -0.4836 -0.6026 -0.5483 -0.5973100 0 1.4299 1.4221 1.4104 1.3953 1.4404 0.0000 0.0000 0.0000 0.0000 0.00005 1.2935 1.2844 1.2785 1.2809 1.2847 -0.1003 -0.1018 -0.0982 -0.0855 -0.114410 1.1672 1.1723 1.1616 1.1795 1.1670 -0.2030 -0.1932 -0.1941 -0.1680 -0.210515 1.0689 1.0808 1.0590 1.0534 1.0633 -0.2910 -0.2744 -0.2865 -0.2811 -0.303520 0.9871 1.0212 0.9655 0.9900 0.9729 -0.3706 -0.3312 -0.3790 -0.3432 -0.392425 0.8728 0.9588 0.8907 0.8760 0.8856 -0.4937 -0.3942 -0.4596 -0.4655 -0.486430 0.8285 0.9037 0.8057 0.8319 0.8226 -0.5457 -0.4534 -0.5599 -0.5172 -0.5602[CBZ]o ' 0.25 µM with detection at 211 nm. Suwannee River NOM concentration 7.0 mg L−1 as C. H2O2 dose of 3.0 mg L−1.Fluence rate at 254 nm of 1 mW cm−2. Solutions at pH 7.0.199Appendix A. Experimental Data and CalculationsTable A.27: Influence of chloride in 254 nm - H2O2 regime with tert-butanoland using carbamazepine as probe[Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3< 1 0 1.3971 1.4067 1.3986 0.0000 0.0000 0.00005 1.2330 1.2601 1.2075 -0.1249 -0.1101 -0.146910 1.0983 1.1246 1.0795 -0.2406 -0.2238 -0.259015 0.9855 0.9990 0.9376 -0.3490 -0.3422 -0.399920 0.8487 0.9137 0.8351 -0.4984 -0.4315 -0.515725 0.8160 0.7244 -0.5446 -0.657930 0.7214 0.6453 -0.6678 -0.773525 0 1.3983 1.4067 1.3991 0.0000 0.0000 0.00005 1.2376 1.2453 1.2256 -0.1221 -0.1219 -0.132410 1.0952 1.1165 1.0890 -0.2443 -0.2310 -0.250615 0.9672 1.0007 0.9517 -0.3686 -0.3405 -0.385320 0.8414 0.8851 0.8330 -0.5079 -0.4633 -0.518625 0.7507 0.7815 0.7098 -0.6220 -0.5878 -0.678630 0.6440 0.7098 0.6637 -0.7753 -0.6840 -0.745840 0 1.4118 1.4024 1.3926 0.0000 0.0000 0.00005 1.2498 1.2302 1.2420 -0.1219 -0.1310 -0.114410 1.0931 1.0827 1.1089 -0.2558 -0.2587 -0.227815 0.9654 0.9730 0.9695 -0.3801 -0.3656 -0.362120 0.8743 0.8358 0.8564 -0.4792 -0.5176 -0.486225 0.7564 0.7317 0.7625 -0.6241 -0.6506 -0.602330 0.6739 0.6396 0.6751 -0.7395 -0.7851 -0.7241100 0 1.3887 1.3984 1.4060 0.0000 0.0000 0.00005 1.2293 1.2705 1.2274 -0.1219 -0.0959 -0.135910 1.0931 1.1182 1.1007 -0.2394 -0.2236 -0.244815 0.9525 0.9806 0.9757 -0.3770 -0.3549 -0.365320 0.8615 0.9071 0.8713 -0.4774 -0.4328 -0.478525 0.7631 0.7808 0.7622 -0.5987 -0.5828 -0.612330 0.6574 0.6946 0.6759 -0.7478 -0.6997 -0.7325[CBZ]o ' 0.25µM with detection at 211 nm. Tert-butanol concentration 7.0mgL−1 asC. H2O2 dose of 3.0mgL−1. Fluence rate at 254 nm of 1mWcm−2.200Appendix A. Experimental Data and CalculationsTable A.28: Influence of chloride in 185 nm regime with Suwannee RiverNOM and using carbamazepine as probe[Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3< 1 0 1.3568 1.3814 1.3661 0.0000 0.0000 0.00001 1.0033 1.0064 1.0022 -0.3018 -0.3167 -0.30982 0.7448 0.7531 0.7549 -0.5998 -0.6067 -0.59313 0.5720 0.5441 0.5351 -0.8637 -0.9317 -0.93734 0.3857 0.3822 0.3803 -1.2578 -1.2849 -1.27885 0.2871 0.2894 0.2747 -1.5531 -1.5630 -1.604025 0 1.4443 1.4346 1.4460 0.0000 0.0000 0.00000.5 1.2453 1.2355 1.2541 -0.1482 -0.1494 -0.14241 1.0987 1.1041 1.0931 -0.2735 -0.2619 -0.27982 0.8095 0.8057 0.8187 -0.5790 -0.5769 -0.56883 0.6055 0.6036 0.6111 -0.8693 -0.8657 -0.86134 0.4626 0.4518 0.4198 -1.1385 -1.1554 -1.23685 0.3443 0.3213 0.3218 -1.4339 -1.4963 -1.502640 0 1.3006 1.3088 1.2948 0.0000 0.0000 0.00001 1.0267 1.0331 1.0266 -0.2365 -0.2365 -0.23212 0.7877 0.7929 0.8109 -0.5015 -0.5012 -0.46803 0.5959 0.5915 0.6257 -0.7805 -0.7942 -0.72724 0.4423 0.4755 0.4937 -1.0786 -1.0125 -0.96425 0.3480 0.3539 0.3362 -1.3184 -1.3079 -1.3484100 0 1.4500 1.4639 1.4766 0.0000 0.0000 0.00000.5 1.3524 1.3321 1.3545 -0.0697 -0.0943 -0.08631 1.2196 1.2045 1.2079 -0.1730 -0.1950 -0.20092 0.9831 0.9639 1.0144 -0.3886 -0.4179 -0.37543 0.8043 0.7814 0.7772 -0.5893 -0.6278 -0.64184 0.6534 0.5964 0.6122 -0.7971 -0.8979 -0.88045 0.4993 0.4985 0.4860 -1.0661 -1.0773 -1.1113[CBZ]o ' 0.25 µM with detection at 211 nm. Suwannee River NOM concentration7.0mgL−1 as C. Irradiated solutions at pH 7.0.201Appendix A. Experimental Data and CalculationsTable A.29: Influence of chloride in 185 nm regime with tert-butanol andusing carbamazepine as probe[Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3< 1 0 1.4651 1.4607 1.4732 0.000 0.000 0.0001 0.9688 0.9812 0.9809 -0.414 -0.398 -0.4072 0.6681 0.6688 0.6469 -0.785 -0.781 -0.8233 0.4499 0.4571 0.4229 -1.181 -1.162 -1.2484 0.3184 0.3063 0.3060 -1.526 -1.562 -1.5726 0.1349 0.1370 0.1286 -2.385 -2.367 -2.4388 0.0764 0.0439 0.0618 -2.954 -3.505 -3.17110 0 1.4325 1.4437 1.4422 0.000 0.000 0.0001 0.9020 0.8937 0.9182 -0.463 -0.480 -0.4522 0.5912 0.5894 0.5841 -0.885 -0.896 -0.9043 0.3923 0.3868 0.3943 -1.295 -1.317 -1.2974 0.2660 0.2662 0.2600 -1.684 -1.691 -1.7136 0.0941 0.1160 0.1072 -2.723 -2.521 -2.5998 0.0616 0.0534 0.0481 -3.147 -3.297 -3.40125 0 1.1207 1.1258 1.1463 0.000 0.000 0.0001 0.5849 0.5988 0.5944 -0.650 -0.631 -0.6572 0.3262 0.3310 -1.234 -1.2423 0.1971 0.1876 0.1869 -1.738 -1.792 -1.8144 0.1067 0.0972 0.1105 -2.352 -2.449 -2.3396 0.0410 0.0319 0.0333 -3.308 -3.564 -3.53940 0 1.4520 1.4593 1.4658 0.000 0.000 0.0000.5 0.8815 0.8978 0.9045 -0.499 -0.486 -0.4831 0.5806 0.5803 0.5712 -0.917 -0.922 -0.9422 0.2532 0.2574 0.2418 -1.747 -1.735 -1.8023 0.1121 0.1288 0.1102 -2.561 -2.427 -2.5884 0.0466 0.0570 0.0551 -3.439 -3.243 -3.281100 0.0 1.4145 1.4237 1.4271 0.000 0.000 0.0000.5 0.6502 0.6712 0.6859 -0.777 -0.752 -0.7331.0 0.2998 0.3278 0.3354 -1.551 -1.469 -1.4481.5 0.1332 0.1495 0.1406 -2.363 -2.254 -2.3172.0 0.0596 0.0667 -3.167 -3.0612.5 0.0290 0.0236 0.0220 -3.887 -4.100 -4.172[CBZ]o ' 0.25 µM with detection at 211 nm. Tert-butanol concentration 7.0mgL−1as C.202Appendix A. Experimental Data and CalculationsTable A.30: Influence of chloride in 185 nm regime with Suwannee RiverNOM and using nitrobenzene as probe[Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3< 1 0 0.8883 0.8927 0.9061 0.0000 0.0000 0.00002 0.6845 0.7092 0.6993 -0.2606 -0.2301 -0.25914 0.5459 0.5587 0.5249 -0.4869 -0.4686 -0.54596 0.4029 0.4165 0.4082 -0.7906 -0.7624 -0.79748 0.3165 0.3282 0.3170 -1.0320 -1.0006 -1.050210 0.2609 0.2538 0.2476 -1.2252 -1.2577 -1.297312 0.1888 0.1889 0.1711 -1.5486 -1.5530 -1.666925 0 0.8924 0.8907 0.9147 0.0000 0.0000 0.00002 0.7222 0.7230 0.7141 -0.2116 -0.2086 -0.24764 0.5931 0.6039 0.5914 -0.4086 -0.3886 -0.43616 0.4697 0.4686 0.4804 -0.6418 -0.6423 -0.64408 0.3750 0.3794 0.3843 -0.8670 -0.8534 -0.867210 0.3025 0.2949 0.3041 -1.0818 -1.1054 -1.101212 0.2700 0.2491 0.2489 -1.1955 -1.2742 -1.301550 0 0.8850 0.8476 0.8684 0.0000 0.0000 0.00002 0.7161 0.7272 0.7377 -0.2118 -0.1532 -0.16314 0.5538 0.5742 0.5786 -0.4688 -0.3894 -0.40606 0.4447 0.4850 0.4805 -0.6882 -0.5583 -0.59188 0.3677 0.3825 0.3982 -0.8783 -0.7957 -0.779710 0.2907 0.3135 0.3351 -1.1133 -0.9946 -0.952212 0.2527 0.2541 0.2575 -1.2534 -1.2047 -1.2156100 0 0.8902 0.8860 0.8999 0.0000 0.0000 0.00002 0.7913 0.7617 0.7757 -0.1178 -0.1512 -0.14854 0.6665 0.6660 0.6610 -0.2894 -0.2854 -0.30856 0.5771 0.5646 0.5637 -0.4334 -0.4506 -0.46788 0.4780 0.4525 0.4593 -0.6218 -0.6719 -0.672610 0.3760 0.3800 0.3707 -0.8619 -0.8465 -0.886912 0.3343 0.3267 0.3436 -0.9794 -0.9977 -0.9628[NB]o ' 1 µMwith detection at 267 nm. Suwannee River NOM concentration 7.0mgL−1as C. Irradiated solutions at pH 7.0.203Appendix A. Experimental Data and CalculationsTable A.31: Influence of chloride in 185 nm regime with tert-butanol andusing nitrobenzene as probe[Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3< 1 0 0.9068 0.9031 0.9012 0.0000 0.0000 0.00002 0.6777 0.6917 0.7138 -0.2912 -0.2667 -0.23314 0.5296 0.5345 0.5376 -0.5378 -0.5245 -0.51666 0.4329 0.4224 0.4012 -0.7394 -0.7599 -0.80938 0.3268 0.3271 0.3159 -1.0206 -1.0156 -1.048310 0.2560 0.2729 0.2579 -1.2647 -1.1967 -1.251212 0.2022 0.1883 0.1821 -1.5007 -1.5678 -1.599225 0 0.9062 0.8993 0.9191 0.0000 0.0000 0.00002 0.7228 0.7385 0.7396 -0.2261 -0.1970 -0.21734 0.6026 0.6024 0.6214 -0.4080 -0.4007 -0.39146 0.4932 0.4592 0.4812 -0.6083 -0.6721 -0.64718 0.3856 0.3839 0.3821 -0.8545 -0.8512 -0.877710 0.3069 0.3156 0.3124 -1.0827 -1.0471 -1.079112 0.2727 0.2572 0.2570 -1.2009 -1.2518 -1.274350 0 0.9277 0.8991 0.9218 0.0000 0.0000 0.00002 0.7540 0.7767 0.7751 -0.2073 -0.1463 -0.17334 0.6431 0.6574 0.6523 -0.3664 -0.3131 -0.34586 0.5475 0.5507 0.5403 -0.5273 -0.4902 -0.53428 0.4720 0.4898 0.4663 -0.6757 -0.6074 -0.681510 0.3897 0.4106 0.4074 -0.8673 -0.7838 -0.816512 0.3323 0.3348 0.3473 -1.0267 -0.9879 -0.9761100 0 0.9151 0.9154 0.9131 0.0000 0.0000 0.00002 0.8117 0.8008 0.8006 -0.1199 -0.1337 -0.13154 0.6999 0.6887 0.6972 -0.2681 -0.2846 -0.26986 0.6019 0.6084 0.6149 -0.4189 -0.4085 -0.39548 0.5333 0.5232 0.5499 -0.5399 -0.5594 -0.507110 0.4730 0.4488 0.4617 -0.6599 -0.7128 -0.681912 0.3973 0.4262 0.4033 -0.8343 -0.7645 -0.8172[NB]o ' 1 µM with detection at 267 nm. Tert-butanol concentration 7.0mgL−1 as C.204Appendix A. Experimental Data and CalculationsTable A.32: Acetate system in 185 nm regime with carbamazepine andnitrobenzene as probesProbe [Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3CBZ < 1 0 1.4767 1.4757 1.5093 0.000 0.000 0.0000.5 1.1231 1.1300 1.1517 -0.274 -0.267 -0.2701.0 0.8422 0.8602 0.8860 -0.562 -0.540 -0.5332.0 0.5404 0.4970 0.5099 -1.005 -1.088 -1.0853.0 0.2794 0.2716 0.2916 -1.665 -1.693 -1.6444.0 0.1660 0.1524 0.1654 -2.186 -2.270 -2.2115.0 0.0834 0.0937 0.1048 -2.874 -2.757 -2.667100 0 1.5351 1.5085 1.5093 0.000 0.000 0.0000.5 1.1982 1.1782 1.1580 -0.248 -0.247 -0.2651.0 0.9171 0.8941 0.8738 -0.515 -0.523 -0.5472.0 0.5935 0.5787 0.5807 -0.950 -0.958 -0.9553.0 0.3923 0.3808 0.3716 -1.364 -1.377 -1.4024.0 0.2576 0.2662 0.2483 -1.785 -1.735 -1.8055.0 0.1803 0.1648 0.1672 -2.142 -2.214 -2.200NB < 1 0 1.0910 1.0992 1.0665 0.000 0.000 0.0000.5 0.9596 0.9557 1.0037 -0.128 -0.140 -0.0611.0 0.9190 0.9055 0.9271 -0.172 -0.194 -0.1402.0 0.7890 0.7986 0.7798 -0.324 -0.319 -0.3133.0 0.6725 0.6464 0.6524 -0.484 -0.531 -0.4914.0 0.5976 0.5483 0.5818 -0.602 -0.696 -0.6065.0 0.4849 0.5070 0.4591 -0.811 -0.774 -0.843100 0 1.1150 1.0965 1.0718 0.000 0.000 0.0000.5 1.0412 1.0121 1.0231 -0.068 -0.080 -0.0471.0 1.0426 0.9817 1.0007 -0.067 -0.111 -0.0692.0 0.9276 0.9484 0.9306 -0.184 -0.145 -0.1413.0 0.9024 0.8402 0.8741 -0.212 -0.266 -0.2044.0 0.8301 0.8019 0.8016 -0.295 -0.313 -0.2905.0 0.7406 0.7639 0.7482 -0.409 -0.361 -0.359[CBZ]o ' 0.25µM with detection at 211 nm. [NB]o ' 1 µM with detection at 267 nm.Solutions contained 8.2mgL−1 as C of acetate at pH 6.0.205Appendix A. Experimental Data and CalculationsTable A.33: Acetone system in 185 nm regime with carbamazepine andnitrobenzene as probesProbe [Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3CBZ < 1 0 1.3899 1.3977 1.4063 0.0000 0.0000 0.00001.0 1.2097 1.2155 1.2241 -0.1389 -0.1397 -0.13882.0 1.0826 1.0570 1.0700 -0.2499 -0.2794 -0.27333.0 0.9648 0.9341 0.9639 -0.3651 -0.4030 -0.37774.0 0.8394 0.8585 0.8486 -0.5043 -0.4874 -0.50516.0 0.6572 0.6715 0.6797 -0.7490 -0.7331 -0.72718.0 0.5134 0.5367 0.5509 -0.9959 -0.9571 -0.9372100 0 1.4212 1.4263 1.4330 0.0000 0.0000 0.00000.25 0.9466 1.0487 1.0060 -0.4064 -0.3075 -0.35380.50 0.6715 0.6469 0.6611 -0.7497 -0.7906 -0.77360.75 0.4394 0.6326 0.5352 -1.1738 -0.8130 -0.98491.00 0.2725 0.2907 0.2749 -1.6516 -1.5905 -1.65111.50 0.1059 0.1155 0.1307 -2.5968 -2.5136 -2.39462.00 0.0391 0.0566 0.0658 -3.5931 -3.2268 -3.0809NB < 1 0 1.0944 1.0660 1.0588 0.0000 0.0000 0.00001.0 1.0305 1.0388 1.0182 -0.0602 -0.0258 -0.03912.0 1.0143 0.9788 1.0069 -0.0760 -0.0853 -0.05033.0 0.9744 0.9721 0.9657 -0.1161 -0.0922 -0.09204.0 0.9459 0.9552 0.9201 -0.1458 -0.1097 -0.14046.0 0.8692 0.8645 0.8701 -0.2304 -0.2095 -0.19638.0 0.8147 0.8294 0.8163 -0.2951 -0.2510 -0.2601100 0 1.0757 1.0429 1.0658 0.0000 0.0000 0.00000.25 1.0095 1.0530 1.1015 -0.0635 0.0096 0.03290.50 1.0673 1.0625 1.0494 -0.0078 0.0186 -0.01550.75 1.0503 1.0503 1.0529 -0.0239 0.0071 -0.01221.0 1.0432 1.0576 1.0524 -0.0307 0.0140 -0.01271.5 1.0536 1.0303 1.0478 -0.0208 -0.0122 -0.01702.0 1.0345 1.0379 0.9964 -0.0391 -0.0048 -0.06732.5 1.0100 1.0447 1.0452 -0.0630 0.0017 -0.01953.0 1.0075 0.9792 0.9879 -0.0655 -0.0630 -0.07594.0 0.9852 0.9542 0.9476 -0.0879 -0.0889 -0.11756.0 0.9534 0.9492 0.9112 -0.1207 -0.0941 -0.15678.0 0.8619 0.9121 0.8856 -0.2216 -0.1340 -0.1852[CBZ]o ' 0.25 µM with detection at 211 nm. [NB]o ' 1 µM with detection at 267 nm. So-lutions contained 10.2mgL−1 as C of acetone.206Appendix A. Experimental Data and CalculationsTable A.34: Influence of ionic strength and chloride in 185 nm regime withcarbamazepine as probe[Cl−] Ionic Strength Time At ln(At/Ao)(mg L−1) (M) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 30 0.00 0 1.4391 1.4202 1.4200 0.000 0.000 0.0001 0.9875 0.9821 0.9948 -0.377 -0.369 -0.3562 0.6971 0.6753 0.6879 -0.725 -0.743 -0.7253 0.4744 0.4520 0.4780 -1.110 -1.145 -1.0894 0.3135 0.3354 0.3218 -1.524 -1.443 -1.4846 0.1593 0.1276 0.1340 -2.201 -2.410 -2.3610.08 0 1.4589 1.4378 1.4449 0.000 0.000 0.0001 1.0502 1.0295 1.0018 -0.329 -0.334 -0.3662 0.7128 0.7359 0.7315 -0.716 -0.670 -0.6813 0.5257 0.5177 0.5221 -1.021 -1.021 -1.0184 0.3721 0.3843 0.3773 -1.366 -1.319 -1.3436 0.1772 0.1908 0.1933 -2.108 -2.020 -2.0120.16 0 1.4462 1.4352 1.4257 0.000 0.000 0.0001 1.0876 1.0620 1.0963 -0.285 -0.301 -0.2632 0.8202 0.7981 0.8026 -0.567 -0.587 -0.5753 0.6195 0.6180 0.6185 -0.848 -0.843 -0.8354 0.4638 0.4712 0.4533 -1.137 -1.114 -1.1466 0.2870 0.2764 0.2418 -1.617 -1.647 -1.774100 0.00 0.0 1.4250 1.4032 1.4452 0.000 0.000 0.0000.5 0.5726 0.5285 0.5828 -0.912 -0.976 -0.9081.0 0.2234 0.1840 0.2077 -1.853 -2.032 -1.9401.5 0.0952 0.0626 0.0819 -2.706 -3.110 -2.8710.08 0.0 1.4444 1.4404 1.4427 0.000 0.000 0.0000.5 0.9432 0.9254 0.9145 -0.426 -0.442 -0.4561.0 0.6458 0.6516 0.6232 -0.805 -0.793 -0.8391.5 0.4506 0.4270 0.4554 -1.165 -1.216 -1.1532.0 0.3197 0.2970 0.3007 -1.508 -1.579 -1.5682.5 0.2037 0.2110 0.2264 -1.959 -1.921 -1.8520.16 0.0 1.4435 1.4286 1.4230 0.000 0.000 0.0000.5 1.0905 1.0703 1.0859 -0.280 -0.289 -0.2701.0 0.7880 0.7867 0.8267 -0.605 -0.597 -0.5431.5 0.6016 0.5892 0.6573 -0.875 -0.886 -0.7722.0 0.4526 0.4751 0.4766 -1.160 -1.101 -1.0942.5 0.3455 0.3422 0.3214 -1.430 -1.429 -1.488[CBZ]o ' 0.25 µM with detection at 211 nm. Tert-butanol concentration 7.0 mg L−1 as C. Ionicstrength adjusted using NaF. Chloride adjusted using NaCl.207Appendix A. Experimental Data and CalculationsTable A.35: Quantification of molar absorption coefficient of chloride at 185nm using a kinetic method[Cl−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 30 0 1.3595 1.3458 1.3760 0.000 0.000 0.0001 1.1669 1.1589 1.1469 -0.153 -0.150 -0.1822 0.9519 0.9803 -0.356 -0.3174 0.6479 0.6778 0.6701 -0.741 -0.686 -0.7206 0.4705 0.4612 0.4731 -1.061 -1.071 -1.0688 0.3297 0.3410 0.3345 -1.417 -1.373 -1.41410 0.2342 0.2450 0.2158 -1.759 -1.703 -1.85325 0 1.3781 1.3856 0.000 0.0002 1.1393 1.1346 -0.190 -0.2004 0.9570 0.9737 -0.365 -0.3536 0.8049 0.7890 -0.538 -0.5638 0.6381 0.6734 -0.770 -0.72210 0.5317 0.5324 -0.952 -0.95650 0 1.4115 1.3548 0.000 0.0002 1.2229 1.2148 -0.143 -0.1094 1.1131 1.1195 -0.238 -0.1918 0.8759 0.9114 -0.477 -0.39612 0.7465 0.7072 -0.637 -0.65016 0.5962 0.6068 -0.862 -0.80375 0 1.4032 1.3866 0.000 0.0004 1.2412 1.1396 -0.123 -0.1968 1.0955 0.9774 -0.248 -0.35012 0.9173 0.9002 -0.425 -0.43216 0.8019 0.8049 -0.560 -0.54420 0.7225 0.7327 -0.664 -0.638100 0 1.3632 1.4057 0.000 0.0004 1.2598 1.2629 -0.079 -0.1078 1.1402 1.1544 -0.179 -0.19712 1.0397 1.0312 -0.271 -0.31016 0.9424 -0.36920 0.8771 0.8401 -0.441 -0.515Lower 1.0 cm cell contains [CBZ]o ' 0.25µM with detection at 211 nm, andtert-butanol concentration 7.0mgL−1 as C. Upper 1.0mm cell contains varyingchloride concentration adjusted using NaCl.208Appendix A. Experimental Data and CalculationsTable A.36: Influence of sulphate in 254 nm - H2O2 regime with SuwanneeRiver NOM and using carbamazepine as probe[SO42−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 1 Run 2< 1 0 1.3408 1.2895 0.000 0.0005 1.2146 1.1594 -0.099 -0.10610 1.0512 1.0460 -0.243 -0.20915 0.9783 0.8686 -0.315 -0.39520 0.8775 0.8176 -0.424 -0.456100 0 1.3472 1.2546 0.000 0.0005 1.1534 1.1730 -0.155 -0.06710 1.0424 -0.25715 0.9533 0.9554 -0.346 -0.27220 0.8908 0.8643 -0.414 -0.373[CBZ]o ' 0.25µM with detection at 211 nm. Suwannee RiverNOM concentration 7.0 mg L−1 as C. H2O2 dose of 7.0 mg L−1.Fluence rate at 254 nm of 1 mW cm−2. Solutions at pH 7.0.Table A.37: Influence of sulphate in 254 nm - H2O2 regime with tert-butanoland using carbamazepine as probe[SO42−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3< 1 0 1.3630 1.3657 1.3581 0.0000 0.0000 0.0005 1.1872 1.1941 1.1038 -0.1381 -0.1343 -0.20710 1.0513 1.0921 0.9909 -0.2597 -0.2236 -0.31515 0.9148 0.9669 0.8953 -0.3987 -0.3453 -0.41720 0.7770 0.9098 0.7446 -0.5620 -0.4062 -0.60125 0.6928 0.8248 0.6923 -0.6767 -0.5043 -0.674100 0 1.3766 1.4052 1.3469 0.000 0.000 0.0005 1.1920 1.2080 1.1433 -0.144 -0.151 -0.16410 1.0310 1.0748 1.0113 -0.289 -0.268 -0.28715 0.8578 0.9348 0.8529 -0.473 -0.408 -0.45720 0.7530 0.8185 0.7598 -0.603 -0.540 -0.57325 0.6422 0.7367 0.6429 -0.762 -0.646 -0.740[CBZ]o ' 0.25 µM with detection at 211 nm. Tert-butanol concentration 7.0 mg L−1as C. H2O2 dose of 7.0 mg L−1. Fluence rate at 254 nm of 1 mW cm−2.209Appendix A. Experimental Data and CalculationsTable A.38: Influence of sulphate in 185 nm regime with tert-butanol andusing carbamazepine as probe[SO42−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3< 1 0.0 1.4806 1.4633 1.4524 0.000 0.000 0.0001.0 0.9983 0.9992 1.0004 -0.394 -0.381 -0.3732.0 0.7012 0.6988 0.6934 -0.747 -0.739 -0.7393.0 0.5012 0.4767 0.4797 -1.083 -1.122 -1.1084.0 0.3345 0.3337 0.3486 -1.488 -1.478 -1.4275.0 0.2125 0.2249 0.2078 -1.941 -1.873 -1.94425 0.0 1.4396 1.4287 1.4492 0.000 0.000 0.0000.5 1.0617 1.0501 1.0258 -0.304 -0.308 -0.3461.0 0.7772 0.7810 0.7853 -0.616 -0.604 -0.6132.0 0.3949 0.4180 0.4116 -1.293 -1.229 -1.2593.0 0.2185 0.2208 0.2235 -1.885 -1.867 -1.8694.0 0.1002 0.1083 0.1103 -2.665 -2.580 -2.57650 0.0 1.4657 1.4237 1.4360 0.000 0.000 0.0000.5 0.9174 0.9231 0.9157 -0.469 -0.433 -0.4501.0 0.5494 0.5916 0.5995 -0.981 -0.878 -0.8742.0 0.2319 0.2490 0.2480 -1.844 -1.744 -1.7563.0 0.0620 0.0887 0.1071 -3.163 -2.776 -2.5964.0 0.0226 0.0338 0.0445 -4.172 -3.741 -3.47475 0.0 1.4982 1.4928 1.4920 0.000 0.000 0.0000.5 0.8329 0.8126 0.8638 -0.587 -0.608 -0.5471.0 0.4363 0.4458 0.4731 -1.234 -1.209 -1.1491.5 0.2653 0.2671 0.2853 -1.731 -1.721 -1.6542.0 0.1065 0.1227 0.1530 -2.644 -2.499 -2.2772.5 0.0773 0.0897 -2.961 -2.8113.0 0.0271 -4.012100 0.0 1.4475 1.4488 1.4577 0.000 0.000 0.0000.5 0.7200 0.7296 0.7529 -0.698 -0.686 -0.6611.0 0.3322 0.3708 0.3837 -1.472 -1.363 -1.3351.5 0.1941 0.1887 0.2001 -2.009 -2.038 -1.9862.0 0.0721 0.1716 0.1119 -3.000 -2.133 -2.5672.5 0.0571 0.0507 0.0307 -3.233 -3.353 -3.860[CBZ]o ' 0.25 µM with detection at 211 nm. Tert-butanol concentration7.0 mg L−1 as C.210Appendix A. Experimental Data and CalculationsTable A.39: Influence of bicarbonate in 185 nm regime with tert-butanoland using carbamazepine as probe[HCO3−] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 3< 1 0 1.4328 1.4234 1.4295 0.000 0.000 0.0001 1.0150 0.9861 1.0118 -0.345 -0.367 -0.3462 0.6876 0.7015 0.6877 -0.734 -0.708 -0.7323 0.4702 0.4653 0.5051 -1.114 -1.118 -1.0404 0.3222 0.3279 0.3516 -1.492 -1.468 -1.4035 0.2364 0.2376 0.2141 -1.802 -1.790 -1.89960 0 1.5035 1.4866 1.4832 0.000 0.000 0.0001 1.1365 1.1194 1.1512 -0.280 -0.284 -0.2532 0.8692 0.8593 0.8566 -0.548 -0.548 -0.5493 0.6495 0.6525 0.6435 -0.839 -0.823 -0.8354 0.4872 0.5022 0.4938 -1.127 -1.085 -1.1005 0.3687 0.3549 0.3733 -1.406 -1.432 -1.380120 0 1.4084 1.4105 1.4205 0.000 0.000 0.0002 0.9270 0.9071 0.9080 -0.418 -0.441 -0.4484 0.5947 0.5745 0.5930 -0.862 -0.898 -0.8746 0.3831 0.3840 0.3528 -1.302 -1.301 -1.3938 0.2371 0.2262 0.2085 -1.782 -1.830 -1.91910 0.1445 0.1434 0.1331 -2.277 -2.286 -2.368180 0 1.4250 1.4350 1.4150 0.000 0.000 0.0002 1.0082 0.9838 1.0143 -0.346 -0.377 -0.3334 0.6968 0.7082 0.7089 -0.715 -0.706 -0.6916 0.5003 0.4781 0.4773 -1.047 -1.099 -1.0878 0.3506 0.3316 0.3373 -1.402 -1.465 -1.43410 0.2053 0.2153 0.2445 -1.937 -1.897 -1.756[CBZ]o ' 0.25 µM with detection at 211 nm. Tert-butanol concentration7.0mgL−1 as C. Solutions at pH 8.3.211Appendix A. Experimental Data and CalculationsTable A.40: Interaction study for influence of bicarbonate and sulphate in185 nm regime with tert-butanol and using carbamazepine as probe[SO42−] [HCO3−] Time At ln(At/Ao)(mg L−1) (mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 30 0 0 1.4016 1.4000 1.4255 0.000 0.000 0.0001 0.9815 0.9549 1.0026 -0.356 -0.383 -0.3522 0.6844 0.7114 0.6876 -0.717 -0.677 -0.7293 0.4689 0.4880 0.4964 -1.095 -1.054 -1.0554 0.3171 0.3226 0.3697 -1.486 -1.468 -1.3505 0.2566 0.2300 0.2417 -1.698 -1.806 -1.77550 0 0.0 1.3994 1.3967 1.3892 0.000 0.000 0.0000.5 0.9287 0.9136 0.8999 -0.410 -0.424 -0.4341.0 0.5967 0.6036 0.5913 -0.852 -0.839 -0.8541.5 0.4337 0.3967 0.4204 -1.171 -1.259 -1.1952.0 0.2669 0.2498 0.2856 -1.657 -1.721 -1.5822.5 0.1935 0.1594 0.1736 -1.979 -2.170 -2.08025 90 0 1.4129 1.4068 1.4000 0.000 0.000 0.0001 0.9992 0.9767 1.0054 -0.346 -0.365 -0.3312 0.7347 0.7066 0.7535 -0.654 -0.689 -0.6194 0.3521 0.3592 0.3728 -1.389 -1.365 -1.3236 0.1717 0.1604 0.1680 -2.108 -2.171 -2.1208 0.0759 0.0762 0.0672 -2.924 -2.916 -3.0370 180 0 1.4188 1.4540 1.4470 0.000 0.000 0.0001 1.1938 1.2191 1.2023 -0.173 -0.176 -0.1852 1.0138 1.0045 1.0069 -0.336 -0.370 -0.3634 0.7216 0.7501 0.7071 -0.676 -0.662 -0.7166 0.5121 0.5324 0.5157 -1.019 -1.005 -1.0328 0.3825 0.3526 0.3863 -1.311 -1.417 -1.32150 180 0 1.3975 1.4098 1.4129 0.000 0.000 0.0001 1.0407 1.0266 1.0396 -0.295 -0.317 -0.3072 0.7711 0.7680 0.7703 -0.595 -0.607 -0.6074 0.4117 0.4327 0.3982 -1.222 -1.181 -1.2666 0.2106 0.2075 0.2227 -1.892 -1.916 -1.8488 0.0911 0.0961 0.0882 -2.730 -2.686 -2.774[CBZ]o ' 0.25µM with detection at 211 nm. Tert-butanol concentration 7.0mgL−1 as C.Solutions at pH 7-8.212Appendix A. Experimental Data and CalculationsTable A.41: Interaction study for influence of bicarbonate and chloride in185 nm regime with tert-butanol and using carbamazepine as probe[Cl−] [HCO3−] Time At ln(At/Ao)(mg L−1) (mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 30 0 0 1.4467 1.4005 1.4616 0.000 0.000 0.0001 0.9689 1.0193 1.1706 -0.401 -0.318 -0.2222 0.7299 0.6824 0.6949 -0.684 -0.719 -0.7443 0.4906 0.4795 0.5132 -1.081 -1.072 -1.0474 0.3338 0.3532 0.3505 -1.466 -1.378 -1.4285 0.2142 0.2436 -1.878 -1.7926 0.1579 -2.21550 0 0 1.4353 1.4418 1.4099 0.000 0.000 0.0000.5 0.8913 0.8748 0.8394 -0.476 -0.500 -0.5191.0 0.4826 0.4894 0.5046 -1.090 -1.080 -1.0281.5 0.2957 0.2995 0.3427 -1.580 -1.572 -1.4142.0 0.1745 0.1956 0.1940 -2.107 -1.998 -1.9832.5 0.1013 0.1229 0.1287 -2.651 -2.462 -2.39425 90 0 1.4269 1.3671 1.4091 0.000 0.000 0.0001 1.1950 1.1683 1.1803 -0.177 -0.157 -0.1772 1.0311 1.0119 0.9983 -0.325 -0.301 -0.3454 0.7592 0.7641 0.7474 -0.631 -0.582 -0.6346 0.5345 0.5362 0.5629 -0.982 -0.936 -0.9188 0.3836 0.4022 0.4035 -1.314 -1.223 -1.2510 180 0 1.4150 1.3786 1.4369 0.000 0.000 0.0001 1.1908 1.1967 1.2326 -0.173 -0.142 -0.1532 1.0367 0.9668 1.0161 -0.311 -0.355 -0.3474 0.7105 0.7299 0.7418 -0.689 -0.636 -0.6616 0.5411 0.5159 0.5101 -0.961 -0.983 -1.0368 0.3562 0.3606 0.3737 -1.379 -1.341 -1.34750 180 0 1.3932 1.3939 1.4002 0.000 0.000 0.0001 1.2802 1.2473 1.2692 -0.085 -0.111 -0.0982 1.1286 1.1527 1.1396 -0.211 -0.190 -0.2064 0.9136 0.9327 0.8820 -0.422 -0.402 -0.4626 0.7345 0.7661 0.7497 -0.640 -0.599 -0.6258 0.5691 0.6057 0.6068 -0.895 -0.833 -0.836[CBZ]o ' 0.25µM with detection at 211 nm. Tert-butanol concentration 7.0mgL−1 as C.Solutions at pH 7-8.213Appendix A. Experimental Data and CalculationsTable A.42: Interaction study for influence of bicarbonate, chloride and sul-phate in 185 nm regime with Suwannee River NOM and using carbamazepineas probe[HCO3−] [Cl−] [SO42−] Time At ln(At/Ao)(mg L−1) (mg L−1) (mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 30 0 0 0 1.3812 1.3681 1.3993 0.000 0.000 0.0001 0.8061 0.8804 0.8631 -0.539 -0.441 -0.4832 0.5204 0.5021 0.4573 -0.976 -1.002 -1.1183 0.3054 0.3188 0.2830 -1.509 -1.457 -1.5984 0.1579 0.1859 0.1792 -2.169 -1.996 -2.0555 0.0887 0.0967 0.0964 -2.745 -2.650 -2.6750 0 50 0 1.3992 1.4194 1.4406 0.000 0.000 0.0001 0.8619 0.8206 0.8068 -0.485 -0.548 -0.5802 0.4812 0.4769 0.4913 -1.067 -1.091 -1.0763 0.2685 0.2836 0.2709 -1.651 -1.610 -1.6714 0.1671 0.1697 0.1626 -2.125 -2.124 -2.1825 0.0996 0.0826 0.0946 -2.642 -2.844 -2.7230 50 0 0 1.3962 1.3781 1.3605 0.000 0.000 0.0001 0.9809 1.0057 0.9429 -0.353 -0.315 -0.3672 0.7001 0.6611 0.6980 -0.690 -0.735 -0.6673 0.4740 0.5005 0.4858 -1.080 -1.013 -1.0304 0.3109 0.3349 0.3461 -1.502 -1.415 -1.3695 0.2263 0.2337 0.2323 -1.820 -1.774 -1.7680 50 50 0 1.3798 1.3685 1.3661 0.000 0.000 0.0001 1.0042 0.9823 0.9613 -0.318 -0.332 -0.3512 0.6823 0.6832 0.7253 -0.704 -0.695 -0.6333 0.4555 0.4878 0.5072 -1.108 -1.032 -0.9914 0.3049 0.3306 0.3183 -1.510 -1.421 -1.4575 0.2111 0.2211 0.2244 -1.877 -1.823 -1.806180 0 0 0 1.4138 1.3921 1.3842 0.000 0.000 0.0001 1.1360 1.1226 1.1331 -0.219 -0.215 -0.2002 0.9276 0.9269 0.9561 -0.421 -0.407 -0.3703 0.7912 0.7779 0.7417 -0.580 -0.582 -0.6244 0.6498 0.6103 0.6030 -0.777 -0.825 -0.8315 0.4958 0.5144 0.4735 -1.048 -0.996 -1.073180 0 50 0 1.4155 1.3979 1.3581 0.000 0.000 0.0001 1.0175 1.0439 1.0252 -0.330 -0.292 -0.2812 0.7961 0.7628 0.7539 -0.576 -0.606 -0.5893 0.5946 0.6155 0.5768 -0.867 -0.820 -0.8564 0.4268 0.3992 0.4440 -1.199 -1.253 -1.1185 0.3352 0.3375 0.3066 -1.441 -1.421 -1.488180 50 0 0 1.4279 1.4211 1.3797 0.000 0.000 0.0001 1.1874 1.2105 1.2019 -0.184 -0.160 -0.1382 1.0679 1.0828 1.0368 -0.291 -0.272 -0.286Continued on next page214Appendix A. Experimental Data and CalculationsTable A.42: (continued)[HCO3−] [Cl−] [SO42−] Time At ln(At/Ao)(mg L−1) (mg L−1) (mg L−1) (min) Run 1 Run 2 Run 3 Run 1 Run 2 Run 33 0.9606 0.9407 0.9827 -0.396 -0.413 -0.3394 0.8229 0.8503 0.8096 -0.551 -0.514 -0.5335 0.7172 0.7385 0.7522 -0.689 -0.655 -0.607180 50 50 0 1.3946 1.3760 1.4102 0.000 0.000 0.0001 1.1689 1.1928 1.2136 -0.177 -0.143 -0.1502 1.0387 1.0561 1.0678 -0.295 -0.265 -0.2783 0.9170 0.9092 0.8975 -0.419 -0.414 -0.4524 0.7827 0.8124 0.7997 -0.578 -0.527 -0.5675 0.6833 0.7250 0.7105 -0.713 -0.641 -0.686[CBZ]o ' 0.25µM with detection at 211 nm. Suwannee River NOM concentration 7.0 mg L−1 as C. Solu-tions at pH 7-8.215Appendix A. Experimental Data and CalculationsTable A.43: Quantification of molar absorption coefficient of sulphate, bi-carbonate, and Suwannee River NOM at 185 nm using a kinetic method**.Solute [S] Time At ln(At/Ao)(mg L−1) (min) Run 1 Run 2 Run 1 Run 2SO42– 500 0 1.3780 1.3680 0.000 0.0001 1.2120 1.1977 -0.128 -0.1332 1.0454 1.0146 -0.276 -0.2994 0.7702 0.7701 -0.582 -0.5756 0.5832 0.5673 -0.860 -0.8808 0.4203 0.4323 -1.187 -1.152HCO3– 180 0 1.3802 1.3938 0.000 0.0001 1.1850 1.1643 -0.152 -0.1802 1.0362 1.0160 -0.287 -0.3164 0.7571 0.7379 -0.600 -0.6366 0.5657 0.5720 -0.892 -0.8918 0.4290 0.4139 -1.169 -1.21410 0.3217 -1.456SR NOM 43* 0 1.3703 1.3948 0.000 0.0001 1.2706 1.2469 -0.076 -0.1122 1.1719 1.1239 -0.156 -0.2164 0.9502 0.9481 -0.366 -0.3866 0.7869 0.7867 -0.555 -0.5738 0.6599 0.6172 -0.731 -0.815Lower 1.0 cm cell contains [CBZ]o ' 0.25 µM with detection at 211 nm,and tert-butanol concentration 7.0mgL−1 as C. Upper 1.0mm cell con-tains solute under test of known concentration ([S])*SR NOM in units of mg L−1 as C.** See TableA.35 on 208 for data of 1.0mm blank using ultrapure water .216

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0348109/manifest

Comment

Related Items