Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Application of the UV/Chlorine Advanced Oxidation Process
for Drinking Water Treatment
by
Ding Wang
A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy
Graduate Department of Civil Engineering
University of Toronto
© Copyright by Ding Wang 2015
ii
Application of the UV/Chlorine Advanced Oxidation Process
for Drinking Water Treatment Ding Wang
Doctor of Philosophy, 2015
Graduate Department of Civil Engineering
University of Toronto
ABSTRACT
This research investigated the feasibility of a novel advanced oxidation process (AOP)
using ultraviolet light combined with free chlorine (UV/chlorine) in drinking water treatment.
A bench-scale study using a medium pressure UV collimated beam apparatus showed that
the UV/chlorine process was more efficient than the UV combined with hydrogen peroxide
(UV/H2O2) AOP for the destruction of trichloroethylene (TCE) at pH 5 in a laboratory prepared
water, but was less efficient than the latter at pH 7.5 and 10. A Matlab® mathematical model
made accurate predictions of the observed experimental rates of TCE decay. The model
predicted that increasing concentrations of hydroxyl radical scavengers in the treated water
would tend to raise the pH at which UV/chlorine would remain competitive relative to
UV/H2O2.
Full-scale experiments at the City of Cornwall Water Purification Plant (Ontario, Canada)
and pilot-scale tests in a Rayox® batch UV reactor using water from the Keswick Water
Treatment Plant (Ontario, Canada) demonstrated comparable performance of UV/chlorine AOP
to UV/H2O2 for geosmin, 2-methylisoborneol (MIB), and caffeine destruction.
iii
Organic and inorganic disinfection by-products (DBPs) were also monitored in the full-
and pilot-scale tests. Minimal trihalomethane and haloacetic acid formation was observed across
the UV reactor, while dichloroacetonitrile and bromochloroacetonitrile were produced rapidly,
although overall concentrations were below 6 µg L–1. Adsorbable organic halide was formed
rapidly (up to 70 µg Cl L–1) in water that had not been prechlorinated, while little formation was
observed in previously chlorinated water. Chlorate and bromate were formed, equivalent to
approximately 2–17% and 0.01–0.05% of the photolyzed chlorine, respectively, while no
perchlorate or chlorite formation was observed. In addition, the 24 h organic DBP formation
potential was increased by UV/chlorine pretreatment to an extent that was similar to that
observed when the water was pretreated with UV/H2O2.
iv
ACKOWLEDGEMENTS
I would like to express my sincere gratitude to my supervisor, Professor Ron Hofmann, for
his suggestions, guidance, encouragement, understanding, patience, and support throughout my
research. I would like to thank Prof. James R. Bolton for his continuously invaluable help and
advice on UV technology since my Master’s study. I also appreciate Prof. Susan Andrews for
being on my supervisory committee and providing excellent suggestions in my research. I am
grateful to Prof. Robert Andrews for his encouragement.
I also appreciate Leigh McDermott from Stantec Consulting Ltd., Owen O'Keefe, Daniel
Drouin, and Morris McCormick from Cornwall Water Purification Plant, Dr. Vasile Furdui from
the Ontario Ministry of Environment and Climate Change, Dr. Hong Zhang, Dr. A.H.M. Anwar
Sadmani, Zhen (Jim) Wang, and Jiafan (Kevin) Yang from Drinking Water Research Group for
their great help in my research.
Last but not least, I would like to give my special thanks to my wife and parents for their
endless support and encouragement.
This work was financially supported by Natural Sciences and Engineering Research
Council of Canada (NSERC) through the Engage Grant program and the Industrial Research
Chair program, and by Stantec Consulting Ltd. and the Centre for Control of Emerging
Contaminants (CCEC).
v
TABLE OF CONTENTS
ABSTRACT.............................................................................................................................. ii
ACKOWLEDGEMENTS ....................................................................................................... iv
TABLE OF CONTENTS ..........................................................................................................v
LIST OF TABLES .................................................................................................................. ix
LIST OF FIGURES ..................................................................................................................x
LIST OF ACRONYMS .......................................................................................................... xii
1. INTRODUCTION .............................................................................................................1
References ..............................................................................................................................2
2. LITERATURE REVIEW .................................................................................................5
2.1 Advanced Oxidation Processes (AOPs) ..........................................................................5
2.1.1 Theory of Advanced Oxidation Processes (AOPs) ......................................................5
2.1.2 Miscellaneous Methods for AOPs ...............................................................................7
2.2 UV Combined with Chlorine as an AOP ........................................................................7
2.2.1 Fundamental Chemistry of Aqueous Free Chlorine .....................................................7
2.2.2 Fundamental Chemistry of the UV/Chlorine AOP..................................................... 10
2.3 Formation of Disinfection By-Products (DBPs) in UV/Chlorine ................................. 17
2.3.1 Chlorinated DBPs ..................................................................................................... 17
2.3.2 DBP Formation Kinetics ........................................................................................... 19
2.3.3 DBP Formation by UV and AOPs ............................................................................. 22
2.4 Summary of Literature Review .................................................................................... 25
References ............................................................................................................................ 26
3. MEDIUM PRESSURE UV COMBINED WITH CHLORINE ADVANCED
OXIDATION FOR TRICHLOROETHYLENE DESTRUCTION IN A MODEL
WATER ................................................................................................................................... 36
Abstract ............................................................................................................................... 36
3.1 Introduction ................................................................................................................... 36
3.2 Materials and Methods .................................................................................................. 40
3.2.1 Reagents and Materials ............................................................................................. 40
3.2.2 UV Exposure and Irradiance Measurements .............................................................. 40
vi
3.2.3 Analytical Methods ................................................................................................... 41
3.3 Results and Discussion .................................................................................................. 41
3.3.1 Molar Absorption Coefficients of TCE, Active Chlorine, Peroxide and Hydroxide
Species .............................................................................................................................. 41
3.3.2 Quantum Yields of Active Chlorine and Hydrogen Peroxide Photolysis.................... 42
3.3.3 TCE Decay Rates by UV Alone, and the UV/Chlorine and the UV/H2O2 AOPs ........ 44
3.3.4 Mathematical Modeling of the TCE Decay ............................................................... 51
3.3.5 Comment on Active Chlorine Reaction with •OH ..................................................... 52
3.4 Conclusions .................................................................................................................... 54
Acknowledgements .............................................................................................................. 55
References ............................................................................................................................ 55
4. FULL-SCALE COMPARISON OF ULTRAVIOLET/CHLORINE ADVANCED
OXIDATION TO ULTRAVIOLET/HYDROGEN PEROXIDE FOR TASTE AND
ODOUR CONTROL IN DRINKING WATER TREATMENT .......................................... 60
Abstract ............................................................................................................................... 60
4.1 Introduction ................................................................................................................... 60
4.2 Material and Methods ................................................................................................... 62
4.2.1 Reagents and Materials ............................................................................................. 62
4.2.2 Experimental Facilities and Procedures ..................................................................... 62
4.2.3 Sample Analysis ....................................................................................................... 64
4.3 Results and Discussion .................................................................................................. 65
4.3.2 Free Chlorine Decay ................................................................................................. 65
4.3.3 Geosmin and MIB Decay .......................................................................................... 66
4.3.4 Caffeine Decay ......................................................................................................... 70
4.3.5 Electrical Energy per Order (EEO) ............................................................................. 71
4.3.6 Comment on Chlorine Radical (•Cl) and Disinfection By-Product (DBP)
Formation during Chlorine Photolysis ............................................................................... 73
4.4 Conclusions .................................................................................................................... 74
Acknowledgements .............................................................................................................. 74
References ............................................................................................................................ 74
vii
5. FORMATION OF DISINFECTION BY-PRODUCTS IN THE
ULTRAVIOLET/CHLORINE ADVANCED OXIDATION PROCESS .............................. 78
Abstract ............................................................................................................................... 78
5.1 Introduction ................................................................................................................... 78
5.2 Material and Methods ................................................................................................... 80
5.2.1 Experimental Procedures .......................................................................................... 80
5.2.2 Analytical Methods ................................................................................................... 83
5.3 Results ............................................................................................................................ 84
5.3.1 THMs ....................................................................................................................... 89
5.3.2 HAAs ....................................................................................................................... 89
5.3.3 HANs, HKs, and CP ................................................................................................. 90
5.3.4 AOX ......................................................................................................................... 90
5.3.5 Inorganic DBPs: ClO4–, ClO3
–, ClO2–, and BrO3
– ...................................................... 92
5.4 Discussion....................................................................................................................... 94
5.4.1 Rapid DBP Formation within the UV/Chlorine Reactor ............................................ 94
5.4.2 Impact of UV/Chlorine on 24 Hour DBP-FP ............................................................. 95
5.4.3 Role of the Chlorine Radical (•Cl) ............................................................................ 95
5.5 Conclusions .................................................................................................................... 96
Acknowledgements .............................................................................................................. 96
References ............................................................................................................................ 97
6. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS ................................. 102
6.1 Summary and Conclusions .......................................................................................... 102
6.2 Recommendations for Future Work ........................................................................... 103
APPENDICES ....................................................................................................................... 105
A. Pulse Radiolysis Analysis for Determination of Rate Constant of Free Chlorine
with Hydroxyl Radical ...................................................................................................... 105
B. Determination of the Fluence Rate of the MP Lamp in the Collimated Beam
Apparatus .......................................................................................................................... 108
C. Example of Matlab® Codes ......................................................................................... 110
C.1 Matlab® Codes for Simulation of Trichloroethylene Decay by the UV/Chlorine
AOP at 11 mg L–1 and pH 5 ............................................................................................. 110
viii
C.2 Matlab® Codes for Simulation of Trichloroethylene Decay by the UV/Chlorine
AOP at 11 mg L–1 and pH from 5 to 10............................................................................ 112
D. Estimation of OH Radical Concentration Using an Excel Spreadsheet ..................... 113
E. UV Dose Estimation Using UVCalc® Version 2B ........................................................ 115
F. Absorption Spectra of Geosmin, MIB, and Caffeine ................................................... 116
G. Sample Analysis ............................................................................................................ 118
G.1 Geosmin and MIB..................................................................................................... 118
G.2 Caffeine .................................................................................................................... 120
G.3 Trihalomethanes (THMs), Haloacetonitriles (HANs), Haloketones (HKs),
Chloropicrin (CP), and Trichloroethylene (TCE) ............................................................. 125
G.4 Haloacetic Acids (HAAs) ......................................................................................... 129
G.5 Chlorate .................................................................................................................... 134
H. Quality Assurance / Quality Control (QA/QC) ........................................................... 135
I. Experimental Data ......................................................................................................... 140
ix
LIST OF TABLES
Table 2.1 Application of AOPs in different areas and for different contaminants ....................... 6
Table 2.2 Quantum yields of OCl– decay and photochemical product formation .......................16
Table 2.3 Halogenated by-products in drinking water treatment ...............................................20
Table 3.1 Fluence-based rate constants (10–6 m2 J–1) for active chlorine and peroxide
photolysis .................................................................................................................44
Table 3.2 Comparison of reported quantum yields of active chlorine photolysis .......................45
Table 3.3 TCE Fluence-based decay rate constants (10–4 cm2 mJ– 1) (excluding evaporation)
by UV alone, and the UV/chlorine and the UV/H2O2 AOPs from experimental
and model results .....................................................................................................48
Table 3.4 Calculated hydroxyl radical concentrations (10–13 M) in TCE solutions treated by
the UV/chlorine and the UV/H2O2 AOPs at various pH values .................................50
Table 3.5 Reaction mechanisms of TCE decay by UV alone, the UV/chlorine and the
UV/H2O2 AOPs ........................................................................................................53
Table 4.1 Post-filtration water quality parameters for full- and pilot-scale tests ........................63
Table 4.2 Full- and pilot-scale EEO values (kWh m–3 order–1) for geosmin, MIB, and
caffeine removal .......................................................................................................72
Table 5.1 Full- and pilot-scale post-filtration water quality parameters .....................................80
Table 5.2 Monitored organic and inorganic DBPs ....................................................................83
x
LIST OF FIGURES
Figure 2.1 Percentage distribution of HOCl as a function of pH ................................................ 9
Figure 2.2 Molar absorption coefficients of HOCl, OCl–, and NH2Cl .......................................11
Figure 3.1 Absorption spectra of TCE, chlorine, peroxide and hydroxide species .....................38
Figure 3.2 Relative spectral emittance of the MP lamp in this research .....................................38
Figure 3.3 Rates of active chlorine (a) and hydrogen peroxide (b) photolysis at various pH
values. Error bars represent the standard deviations of triplicate runs. Straight
lines represent the linear regression. ........................................................................43
Figure 3.4 Experimental TCE decay rates by UV alone, the UV/chlorine and the UV/H2O2
AOPs at various pH values. Error bars represent the standard deviations of
triplicate runs. Straight lines represent the linear regression.....................................47
Figure 3.5 Solution pH at which the UV/chlorine and the UV/H2O2 AOPs are equally
efficient as a function of TOC concentration ...........................................................54
Figure 4.1 Full-scale Trojan UVSwift reactor (left) and pilot-scale Rayox® reactor (right) .......63
Figure 4.2 Percentage of free chlorine photolysis by UV exposure. Error bars represent the
values of experimental duplicates. ...........................................................................66
Figure 4.3 Geosmin (top plot) and MIB (bottom plot) decay in the 1st full-scale test. Error
bars represent the values of experimental duplicates. ...............................................67
Figure 4.4 Geosmin decay in the 2nd full-scale test. Error bars represent the values of
experimental duplicates. ..........................................................................................69
Figure 4.5 Caffeine decay in the 1st full-scale (top plot) and pilot-scale (bottom plot) tests.
Error bars represent the values of experimental duplicates.......................................71
Figure 5.1 Full-scale Trojan UVSwift reactor (left) and pilot-scale Rayox® reactor (right) .......81
Figure 5.2 THM formation in full- and pilot-scale tests. Plots on the left show THMs after
various treatment processes for short reaction time (30–60 s contact time). Plots
on the right show THM formation potentials (free chlorine dose: 6.5 mg L–1 for
24 h) in the water pretreated by selected processes shown on the x-axis. Error
bars represent the values of the duplicates measured. ..............................................85
xi
Figure 5.3 HAA formation in full- and pilot-scale tests. Plots on the left show HAAs after
various treatment processes for short reaction time (30–60 s contact time). Plots
on the right show HAA formation potentials (free chlorine dose: 6.5 mg L–1 for
24 h) in the water pretreated by selected processes shown on the x-axis. Error
bars represent the values of the duplicates measured. ..............................................86
Figure 5.4 HAN formation in full- and pilot-scale tests. Plots on the left show HANs after
various treatment processes for short reaction time (30–60 s contact time). Plots
on the right show HAN formation potentials (free chlorine dose: 6.5 mg L–1 for
24 h) in the water pretreated by selected processes shown on the x-axis. Error
bars represent the values of the duplicates measured. ..............................................87
Figure 5.5 AOX formation in full- and pilot-scale tests. Plots on the left show AOX after
various treatment processes for short reaction time (30–60 s contact time). Plots
on the right show AOX formation potentials (free chlorine dose: 6.5 mg L–1 for
24 h) in the water pretreated by selected processes shown on the x-axis. Error
bars represent the values of the duplicates measured. ..............................................88
Figure 5.6 Formation of ClO3– relative to free chlorine photodecomposition in the full- and
pilot-scale experiments. Error bars represent the values of the duplicates
measured. Low, medium, and high represent free chlorine doses of 2, 6, and 10
mg L–1, respectively. ...............................................................................................93
xii
LIST OF ACRONYMS
AOP Advanced oxidation process
BCAA Bromochloroacetic acid
BCAN Bromochloroacetonitrile
BDCAA Bromodichloroacetic acid
BDCM Bromodichloromethane
Br– Bromide ion
BrO3– Bromate ion
Ca(OCl)2 Calcium hypochlorite
CDBAA Chlorodibromoacetic acid
CDBM Chlorodibromomethane
CI Chemical ionization
•Cl Chlorine radical
Cl– Chloride ion
Cl2 Chlorine molecule
ClO2– Chlorite ion
ClO3– Chlorate ion
ClO4– Perchlorate ion
CO32– Carbonate ion
CP Chloropicrin
DBAA Dibromoaceic acid
DBAN Dibromoacetonitrile
DBP Disinfection by-product
DCAA Dichloroacetic acid
DCAN Dichloroacetonitrile
DCP 1,1-Dichloropropanone
EEO Electrical energy per order
EI Electron ionization
FP Formation potential
xiii
GC-ECD Gas chromatography-electron capture detector
GC-MS Gas chromatography-ion-trap mass spectrometry
HAA Haloacetic acid
HAN Haloacetonitrile
HCO3– Bicarbonate ion
H2CO3 Carbonic acid
HK Haloketone
H2O2 Hydrogen peroxide
HOBr Hypobromous acid
HOCl Hypochlorous acid
HOI Hypoiodous acid
H2SO4 Sulphuric acid
HS-SPME Headspace solid phase micro-extraction
IC Ion chromatograph
IC-MS/MS Ion chromatograph tandem mass spectrometer
LP Low pressure
MAC Maximum acceptable concentration
MBAA Monobromoacetic acid
MCAA Monochloroacetic acid
MCL Maximum contaminant level
MDL Method detection limit
MIB 2-Methylisoborneol
MP Medium pressure
MTBE Methyl tertiary-butyl ether
NaOCl Sodium hypochlorite
NaOH Sodium hydroxide
Na2SO3 Sodium sulphite
NB Nitrobenzene
NDMA N-nitrosodimethylamine
NDRL Notre Dame Radiation Laboratory
NH4Cl Ammonium chloride
xiv
(NH4)2SO4 Ammonium sulphate
OCl– Hypochlorite ion
•OH Hydroxyl radical
OH– Hydroxide ion
pCBA para-Chlorobenzoic acid
SPE Solid-phase extraction
TBAA Tribromoacetic acid
TBAN Tribromoacetonitrile
TBM Bromoform/tribromomethane
TCAA Trichloroacetic acid
TCAN Trichloroacetonitrile
TCE Trichloroethylene
TCM Chloroform/trichloromethane
TCP 1,1,1-Tichloropropanone
THM Trihalomethane
TOC Total organic carbon
TOX Total organic halides
UV Ultraviolet
VOC Volatile organic compound
VUV Vacuum-UV
WHO World Health Organization
1
1. INTRODUCTION
Free chlorine (hypochlorous acid, HOCl, and hypochlorite ion, OCl–) exposed to UV light
at wavelengths ranging from 200 to 400 nm (UV/chlorine) is a potential advanced oxidation
process (AOP). An AOP refers to a water or wastewater treatment process that produces and
utilizes hydroxyl radicals (•OH) to destroy contaminants that are not readily oxidized by
conventional oxidants (Nowell and Hoigné, 1992a; Watts and Linden, 2007; Parsons, 2004;
Gültekin and Ince, 2007; Suty et al., 2004). •OH has been proven to be a very strong oxidizing
agent that reacts with many reductants extremely rapidly (Svrcek and Smith, 2004;
Tchobanoglous et al., 2003). Since HOCl is a weak acid, with a pKa of 7.54 at 25°C (Deborde
and von Gunten, 2008), free chlorine contains both HOCl and OCl– in the normal pH range of
drinking water (Health Canada, 2012). Although both HOCl and OCl– photolysis can produce
•OH, the concentrations of •OH generated by HOCl and OCl– at the same molar concentration
are expected to be different based on their different molar absorption coefficients, quantum
yields of •OH formation, and •OH scavenging efficiencies (Feng et al., 2007; Watts and Linden,
2007). This in turn leads to a UV/chlorine efficiency that is sensitive to pH (Watts et al., 2007).
Compared to other commonly used AOPs, such as UV combined with hydrogen peroxide
(UV/H2O2), the UV/chlorine AOP is still novel and not fully explored. Previous studies
indicated that it was potentially an alternative to the UV/H2O2 AOP, since HOCl/OCl– absorbs
UV photons more efficiently than H2O2 when using typical low pressure (LP) or medium
pressure (MP) UV lamps, and produces •OH relatively efficiently under some conditions (Watts
and Linden, 2007; Watts et al., 2007; Feng et al., 2007; Nowell and Hoigné, 1992b; Chan et al.,
2012). However, the mechanisms of chlorine photolysis that are probably associated with a
series of chain reactions with the formation of many intermediates and products (Buxton and
Subhani, 1972a, 1972b, and 1972c) are not completely understood. The quantum yields of
chlorine photolysis and hydroxyl radical formation, which are decisive factors in UV/chlorine
efficiency, are inconsistent among published research papers (Buxton and Subhani, 1972b; Feng
et al., 2007; Watts and Linden, 2007; Jin et al., 2011; Chan et al., 2012). The effectiveness of
this process in drinking water treatment has been only investigated preliminarily and
incompletely.
2
In addition, the application of chlorine can lead to disinfection by-products (DBPs).
Chlorine doses used for the UV/chlorine process (5–10 mg L–1 as free chlorine) are generally
higher than doses normally used for disinfection (0.2–2 mg L–1), but the chlorine contact time in
a practical UV/chlorine AOP can be much shorter than that in disinfection (e.g. seconds/minutes
compared to hours) (Watts et al., 2007; Wang et al., 2011; AWWA, 1999; USEPA, 1999; Sadiq
and Rodriguez, 2004). Under such unique chlorination conditions, DBP formation during the
UV/chlorine treatment could be different from that in chlorine disinfection, but has not been
investigated.
This research therefore (1) explored the mechanisms of chlorine photolysis, (2)
systematically evaluated the UV/chlorine efficiency in the removal of selected contaminants
(trichloroethylene, taste and odour compounds, and caffeine) under different conditions
(chlorine doses and pH) and at different experimental scales (bench-, pilot-, and full-scales), and
(3) investigated the DBP formation during the UV/chlorine process. At each step, UV/chlorine
was compared to UV/H2O2 as a reference. The relevant results are shown in Chapters 3 to 5.
The analytical details and raw data have been included in Appendices G to I.
References American Water Works Association (AWWA), 1999. Water Quality and Treatment: A
Handbook of Community Water Supplies, 5th Edition. McGraw-Hill, Inc., New York, USA.
Buxton, G.V., Subhani, M.S., 1972a. Radiation chemistry and photochemistry of oxychlorine
ions. Part 1.–Radiolysis of aqueous solutions of hypochlorite and chlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
947–957.
Buxton, G.V., Subhani, M.S., 1972b. Radiation chemistry and photochemistry of oxychlorine
ions. Part 2.–Photodecomposition of aqueous solutions of hypochlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
958–969.
Buxton, G.V., Subhani, M.S., 1972c. Radiation chemistry and photochemistry of oxychlorine
ions. Part 3.–Photodecomposition of aqueous solutions of chlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
970–977.
3
Chan, P.Y., Gamal El-Din, M., Bolton, J.R., 2012. A solar-driven UV/chlorine advanced
oxidation process. Water Research, 46(17), 5672–5682.
Deborde, M., von Gunten, U., 2008. Reactions of chlorine with inorganic and organic
compounds during water treatment–kinetics and mechanisms: a critical review. Water
Research, 42(1–2), 13–51.
Feng, Y., Smith, D.W., Bolton, J.R., 2007. Photolysis of aqueous free chlorine species (HOCl
and OCl–) with 254 nm ultraviolet light. Journal of Environmental Engineering and Science,
6(3), 277–284.
Gültekin, I., Ince, N.H., 2007. Synthetic endocrine disruptors in the environment and water
remediation by advanced oxidation processes. Journal of Environmental Management, 85(4),
816–832.
Health Canada, 2012. Guidelines for Canadian Drinking Water Quality – Summery Table.
Water, Air and Climate Change Bureau, Healthy Environments and Consumer Safety Branch,
Health Canada, Ottawa, Ontario, Canada.
Jin, J., Gamal El-Din, M., Bolton, J.R., 2011. Assessment of the UV/chlorine process as an
advanced oxidation process. Water Research, 45(4), 1890–1896.
Parsons, S., 2004. Advanced Oxidation Processes for Water and Wastewater treatment. IWA
Publishing, London, UK.
Nowell, L.H., Hoigné, J., 1992a. Photolysis of aqueous chlorine at sunlight and ultraviolet
wavelengths–I. Degradation rates. Water Research, 26(5), 593–598.
Nowell, L.H., Hoigné, J., 1992b. Photolysis of aqueous chlorine at sunlight and ultraviolet
wavelengths–II. Hydroxyl radical production. Water Research, 26(5), 599–605.
Sadiq, R., Rodriguez, M.J., 2004. Disinfection by-products (DBPs) in drinking water and
predictive models for their occurrence: a review. Science of the Total Environment, 321(1–3),
21–46.
Suty, H., De Traversay, C., Cost, M., 2004. Applications of advanced oxidation processes:
present and future. Water Science and Technology, 49(4), 227–233.
Svrcek, C., Smith, D.W., 2004. Cyanobacteria toxins and the current state of knowledge on
water treatment options: a review. Journal of Environmental Engineering and Science, 3(3),
155–185.
4
Tchobanoglous, G., Burton, F.L., Stensel, H.D., 2003. Wastewater Engineering, Treatment and
Reuse, 4th edition. The McGraw-Hill Companies, Inc., New York.
United States Environmental Protection Agency (USEPA), 1999. Alternative disinfectants and
oxidants guidance manual. Available from <http://www.epa.gov/ogwdw000/mdbp/alternative
_disinfectants_guidance.pdf > (accessed 07.31.14).
Wang, D., Walton, T., McDermott, L., Hofmann, R., 2011. Control of TCE using UV combined
with hydrogen peroxide or chlorine. Proceedings of IOA-IUVA 2011 World Congress &
Exhibition, Paris, France.
Watts, M.J., Linden, K.G., 2007. Chlorine photolysis and subsequent OH radical production
during UV treatment of chlorinated water. Water Research, 41(13), 2871–2878.
Watts, M.J., Rosenfeldt, E.J., Linden, K.G., 2007. Comparative OH radical oxidation using UV-
Cl2 and UV-H2O2 processes. Journal of Water Supply: Research and Technology–AQUA,
56(8), 469–477.
5
2. LITERATURE REVIEW
2.1 Advanced Oxidation Processes (AOPs)
2.1.1 Theory of Advanced Oxidation Processes (AOPs)
Advanced oxidation processes (AOPs) involve the generation and application of hydroxyl
radicals (•OH), which have a very high oxidation potential of 2.80 V (Parsons, 2004). AOPs are
used in water and wastewater treatment to destroy refractory compounds that cannot be oxidized
by conventional oxidants such as oxygen, ozone, and chlorine (AWWA, 1999; Bolton, 2010;
Gültekin and Ince, 2007; Parsons, 2004; Pera-Titus et al., 2004; Svrcek and Smith, 2004;
Tchobanoglous et al., 2003). AOPs have been proven to be very effective in treating a variety of
dissolved organic contaminants in water, since the short-lived •OH can oxidize almost all
reducing materials including a large group of organic chemicals at very high reaction rates (in
the range of 108 to 1010 M–1 s–1) at normal pressure and temperature through non-selective
pathways (Rosenfeldt et al., 2006; Bolton, 2010; Gültekin and Ince, 2007; Parsons, 2004;
Tchobanoglous et al., 2003). AOPs can be applied in many areas to treat a broad range of
aqueous contaminants, as shown in Table 2.1. The reaction types between hydroxyl radicals and
target compounds are commonly radical addition, hydrogen abstraction, and/or electron transfer
(Tchobanoglous et al., 2003, Wang et al., 2006). After a series of oxidation reactions, the target
compounds convert to less complex intermediate products, or are eventually mineralized to CO2
with sufficient contact time and an enough hydroxyl radical concentration (Bolton, 2010;
Gültekin and Ince, 2007). However, since the life of hydroxyl radicals is very short (in the order
of nanoseconds) and their concentration in water is fairly low (for example, approximately 10–13
to 10–12 M generated by hydrogen peroxide (H2O2) photolysis at concentrations of 2–10 mg L–1)
(Mamane et al., 2007; Roots and Okada, 1975), it is not practicable to use AOPs for the removal
of a large amount of contaminants from water. AOPs are thus usually used to treat trace
amounts of contaminants (Tchobanoglous et al., 2003). One important advantage of AOPs over
other oxidants is that AOPs are environmental-friendly without transfer of pollutants from one
phase to another or production of a large amount of sludge (Gültekin and Ince, 2007;
Tchobanoglous et al., 2003).
6
Table 2.1 Application of AOPs in different areas and for different contaminants
(modified from Parsons, 2004)
Different areas for AOPs
Groundwater Surface water
Industrial wastewater Municipal wastewater
Odour and volatile organic compounds (VOCs) Industrial sludge
Municipal sludge Water recycling
Leachates Swimming pool
Disinfection Ultra purification
Different contaminants treated by AOPs
Amino acids Methyl tertiary-butyl ether (MTBE)
Antibiotics Arsenic
Chromium Pesticides
Coliforms Escherichia coli
Cryptosporidium Disinfection by-products (DBPs)
Paper mill effluent Distillery wastewater
Landfill leachate Drug residues
Glass fibre wastewater Taste and odour causing compounds
Hospital wastewater Grey water
Insecticides Rubber process wastewater
Kraft leaching wastewater Chemical specialities wastewater
Natural organic matter (NOM) Humic materials
Nickel plating wastewater Oilfield wastewater
Cyanide Olive mill wastewater
Parasites Municipal wastewater treatment plant effluent
Phenolic wastewater Urine
Printing wastewater Seed corn wastes
Endocrine disruptors Spent caustic
Pharmaceutical compounds Cyanotoxins
N-nitrosodimethylamine (NDMA) Textile industry wastewater
Chlorophenols Nitrophenols
7
2.1.2 Miscellaneous Methods for AOPs
AOPs can be realized through many methods, some of which are exemplified as shown
below (Gültekin and Ince, 2007; Parsons, 2004; Pera-Titus et al., 2004; Tchobanoglous et al.,
2003). The most common AOPs are the combinations of hydrogen peroxide, ozone, and/or UV
light (Tezcanli-Güyer and Ince, 2004).
1. Photochemical processes
• Vacuum-UV (VUV) photolysis
• UV + H2O2
• UV + O3
• UV + H2O2 + O3
• UV + ultrasound
• Photo-Fenton (H2O2 + Fe2+/ Fe3+ + UV/visible light)
• Photocatalysis (UV + TiO2)
2. Non-photochemical processes
• Ozonation (O3)
• O3 + H2O2
• Fenton (H2O2 + Fe2+)
• O3 + Fe2+/ Fe3+
• Ultrasound
• O3 + ultrasound
2.2 UV Combined with Chlorine as an AOP
2.2.1 Fundamental Chemistry of Aqueous Free Chlorine
Chlorine is a traditional and widely used oxidant and disinfectant in water and wastewater
treatment. It is available in gaseous form (Cl2), concentrated aqueous solution (sodium
hypochlorite, NaOCl) and solid form [calcium hypochlorite, Ca(OCl)2] (AWWA, 1999;
Koivunen and Heinonen-Tanski, 2005; Kuo and Smith, 1996; Sadiq and Rodriguez, 2004). The
use of chlorine is relatively economical compared to other methods (Sadiq and Rodriguez,
2004). However, since chlorination can lead to the formation of regulated and/or harmful
8
disinfection by-products (DBPs), such as trihalomethanes (THMs) and haloacetic acids (HAAs),
and chlorine has also been proven to be challenged by some resistant pathogens (e.g. Giardia
lamblia cysts and Cryptosporidium parvum oocysts), alternative oxidation and disinfection
methods (e.g. chlorine dioxide, ozone, UV, and their combinations) have been investigated and
applied (Benabbou et al., 2007; Hijnen et al., 2006; Jung et al., 2008; Koivunen and Heinonen-
Tanski, 2005).
When Cl2 gas is added to water, a hydrolysis reaction takes place to form hypochlorous
acid (HOCl), which may further dissociate to hypochlorite ion (OCl–), shown in Equations [2.1]
and [2.2] (AWWA, 1999; Tchobanoglous et al., 2003).
Cl2 (aq) + H2O ↔ HOCl + H+ + Cl– [2.1]
Equilibrium constant at 25°C: K = 2
[HOCl][H ][Cl ][Cl (aq)]
+ −
= 4.5 × 10–4 M2
Forward reaction rate constant at 20°C: kforward = 11.0 s–1 (Eigen and Kustin, 1962)
Reverse reaction rate constant at 20°C: kreverse = 1.80 × 104 M–1 s–1 (Eigen and Kustin, 1962)
HOCl ↔ H+ + OCl– [2.2]
Equilibrium constant at 25°C: K = Ka = [H ][OCl ][HOCl]
+ −
= 2.9 × 10–8 M, thus pKa = 7.54 (Deborde
and von Gunten, 2008; Morris, 1966).
According to K in Equation [2.1], a large fraction of molecular chlorine can dissolve into
water to produce HOCl/OCl– except at a very low pH and/or a high chloride concentration.
When NaOCl and Ca(OCl)2 are added to water, HOCl/OCl– may be produced following the
reactions [2.3], [2.4], and [2.2].
NaOCl → Na+ + OCl– [2.3]
Ca(OCl)2 → Ca2+ + 2 OCl– [2.4]
9
The total quantity of HOCl and OCl– is referred to as free available chlorine. The relative
proportions of these chlorine species are dependent on the solution temperature and pH. Figure
2.1 illustrates the pH effect on the percentage distribution of HOCl concentration at 25°C in a
closed aqueous system containing 0.05 M free chlorine, determined using Equations [2.1] and
[2.2]. Since the acceptable range of pH in drinking water is approximately 6.5 to 8.5, both HOCl
and OCl– may be predominant (USEPA, 2009). However, the oxidation and disinfection
efficiencies of HOCl can be about 100 times higher than OCl– (AWWA, 1999; Edstrom
Industries, 2003). As a result, low pH is preferred for chlorine disinfection (USEPA, 1999).
However, pH values should not be kept very low, since at pH lower than 4, HOCl can convert to
dissolved Cl2 gas, which is harmful to structures (Edstrom Industries, 2003). In addition, the
auto-decomposition of free chlorine, with a maximum decomposition rate at pH of
approximately 6.7, results in the formation of chlorite (ClO2–) and chlorate (ClO3
–) (Adam et al.,
1992; Chapin, 1934). These undesirable species, usually as the main disinfection DBPs of
chlorine dioxide, induce oxidative damage in human red blood cells. The World Health
Organization (WHO) provisional guideline values for both are 0.7 mg L–1 (WHO, 2005),
compared to the maximum acceptable concentrations (MACs) of 1 mg L–1 for both established
by Health Canada (2012). In the Stage 1 Disinfectants/Disinfection By-products (D/DBP) Rule
issued by USEPA, the maximum contaminant level (MCL) of chlorite is 1.0 mg L–1, with no
information available for chlorate (USEPA, 1999 and 2009).
Figure 2.1 Percentage distribution of HOCl as a function of pH
10
2.2.2 Fundamental Chemistry of the UV/Chlorine AOP
Since conventional chlorine disinfection has been found to be challenged by resistant
protozoa, especially Cryptosporidium parvum, alternative disinfectants, such as ozone, chlorine
dioxide, and UV, and different disinfection strategies, such as sequential treatment with two or
more disinfectants, have been applied (Betancourt and Rose, 2004; Venczel et al., 1997). The
incorporation of UV and chlorine is a multiple-barrier disinfection scenario where chlorination
can be applied before, during, and/or after UV disinfection (Watts and Linden, 2007). This
process enhances the inactivation of Cryptosporidium due to the efficacy of UV and also
maintains residual disinfection effect in distribution systems (Betancourt and Rose, 2004; Feng
et al., 2007). However, free and combined chlorine were observed to degrade when exposed to
UV light (Feng et al., 2007; Watts and Linden, 2007). For example, free chlorine is not stable in
the presence of sunlight in treated water and swimming pool. Nowell and Hoigné (1992a) found
that HOCl and OCl– exposed to sunlight followed the first-order degradation kinetics with the
decay rate constants of 2 × 10–4 s–1 and 1.2 × 10–3 s –1, respectively. A higher chlorine dose
should be added to swimming pool to maintain an appropriate chlorine residual during sunny
days (Nowell and Hoigné, 1992a).
Many studies focused on the photolysis of gaseous HOCl in the stratosphere, since the
photochemical products of HOCl in the atmosphere may play a role in catalyzed destruction of
stratospheric ozone layer (Guo, 1993; Molina and Molina, 1978; Molina et al., 1980; Tanaka et
al., 1998; Vogt and Schindler, 1992). Photolysis of aqueous chlorine may be very different from
that in the atmosphere, but has been explored in very few studies. Chlorine photolysis by UV
light is revealed to be complex, which involves a series of chain reactions with the formation of
many intermediates and products, including •OH and chlorine radicals (•Cl) (Nowell and
Hoigné, 1992a; Watts and Linden, 2007). The molar absorption coefficients of HOCl, OCl–, and
NH2Cl (Figure 2.2) are wavelength dependent, according to the results of Watts and Linden
(2007). This is consistent with Morris (1966) and Feng et al. (2007), who also indicated that the
absorption spectrum peaked at 236 nm for HOCl and 292 nm for OCl–.
Buxton and Subhani (1972a, 1972b and 1972c) explored the photochemical chain reactions
of aqueous hypochlorite and chlorite ions. In alkaline solution, in which OCl– is predominant, a
series of OCl– photodecomposition reactions at 254, 313, and 365 nm were proposed by Buxton
and Subhani (1972b), shown in Equations [2.10] to [2.33]. The final products at 365 nm are O2,
11
Cl–, ClO2–, and ClO3
–, while the yield of ClO2– is not observed at the other two lower
wavelengths, since ClO2– is completely depleted at wavelengths lower than 320 nm shown in
Equations [2.31] and [2.33] (Buxton and Subhani, 1972b). The production of ClO2– and ClO3
–
may need to be considered due to their adverse effect on human health.
Figure 2.2 Molar absorption coefficients of HOCl, OCl–, and NH2Cl
OCl– + hν → Cl– + O(3P) (ground state triplet oxygen atom) [2.10]
OCl– + hν → •Cl + O– [2.11]
OCl– + hν → Cl– + O(1D) (excited singlet oxygen atom) at λ < 320 nm [2.12]
O(3P) + OCl– → ClO2– [2.13]
O(3P) + OCl– → Cl– + O2 [2.14]
O(3P) + ClO2– → ClO3
– [2.15]
O(3P) + ClO2– → Cl– + O2 + O(3P) [2.16]
O– + H2O ↔ •OH + OH– [2.17]
Forward reaction rate constant at 20–25°C for [2.17]: kforward ≈ 108 s–1 (Buxton, 1969)
Reverse reaction rate constant at 20–25°C for [2.17]: kreverse ≈ 1010 M–1 s–1 (Buxton, 1969)
•OH + OCl– → ClO• + OH– [2.18]
0
50
100
150
200
250
300
350
400
450
200 225 250 275 300 325 350 375 400
Mol
ar a
bsor
ptio
n co
effic
ient
(M
–1cm
–1)
Wavelength (nm)
HOCl
OCl-
NH2ClOCl–
NH2Cl
12
Reaction rate constant: k = (9.0 ± 0.5) × 109 M–1 s–1 (Buxton and Subhani, 1972a) or
8 × 109 M–1 s–1 (Nowell and Hoigné, 1992a; Watts and Linden, 2007)
•OH + ClO2– → ClO2 + OH– [2.19]
Reaction rate constant: k = (6.3 ± 0.5) × 109 M–1 s–1 (Buxton and Subhani, 1972a)
O– + OCl– → ClO• + O2– [2.20]
O– + ClO2– → ClO2 + O2– [2.21]
•Cl + OCl– → ClO• + Cl– [2.22]
Reaction rate constant for [2.22]: k = 8.2 × 109 M–1 s–1 (NIST Database, 2002)
2 ClO• ↔ Cl2O2 [2.23]
Cl2O2 + H2O → ClO2– + OCl– + 2 H+ [2.24]
Cl2O2 + H2O → O2 + Cl– + OCl– + 2 H+ [2.25]
Cl2O2 + ClO2– + H2O → ClO3
– + 2 OCl– + 2 H+ [2.26]
O(1D) + H2O → H2O2 [2.27]
H2O2 + OCl– → O2 + Cl– + H2O [2.28]
If the ClO2– concentration is high, such as in a ClO2
– solution, this species can also absorb
photons and produce photochemical products (Buxton and Subhani, 1972c).
ClO2– + hν → OCl– + O(3P) at λ > 320 nm [2.29]
ClO2– + hν → ClO• + O– [2.30]
ClO2– + hν → OCl– + O(1D) at λ < 320 nm [2.31]
ClO2– + hν → (ClO2
–)* [2.32]
(ClO2–)* + ClO2
– → ClO2 + OCl– + O– [2.33]
13
Oppenländer (2003) indicated that Cl– reacts with hydroxyl radicals described in the following
equations:
•OH + Cl– → •ClOH– [2.34]
Reaction rate constant for [2.34]: k = 4.3 × 109 M–1 s–1 (Oppenländer, 2003)
•ClOH– + H+ → •Cl + H2O [2.35]
Meanwhile, Cl– also reacts with the generated •Cl very rapidly, shown in Equation [2.36], to
produce •Cl2–, which is much less reactive than •Cl and •OH (Buxton et al., 1998). The further
reactions of •Cl2– with target contaminants are negligible (Buxton et al., 1998).
•Cl + Cl– ↔ •Cl2– [2.36]
Forward reaction rate constant: kforward = 8.5 × 109 M–1 s–1 (Buxton et al., 1998)
Reverse reaction rate constant: kreverse = 6.0 × 104 M–1 s–1 (Buxton et al., 1998)
No past study systematically elaborated the photodecomposition process of HOCl in acidic
chlorine solutions. The primary photodecomposition of HOCl exposed to UV light is shown in
Equation [2.37] (Feng et al., 2007; Oliver and Carey, 1977; Thomsen et al., 2001; Watts and
Linden, 2007). Thomsen et al. (2001) indicated that the dissociation of HOCl molecule is very
fast, less than 1 picosecond, while the recombination is slower, taking approximately 50
picoseconds.
HOCl + hv → •OH + •Cl [2.37]
The proof of this reaction is that some organic compounds that do not react with HOCl
efficiently in dark solution, such as ethanol, n-butanol, and benzoic acid, can degrade largely in
the presence of sunlight probably via two chain processes shown in Equations [2.38] to [2.41]
(Feng et al., 2007; Kobayashi and Okuda, 1972; Nowell and Hoigné, 1992b; Oliver and Carey,
14
1977). However, it should be noted that chlorinated products may be produced, which may
generate undesirable chlorinated DBPs. For example, based on the results of Oliver and Carey
(1977), organic chlorine accounts for 16.5% by weight of total photochemical products of an n-
butanol solution containing 0.01 M HOCl at pH 4 after 30 min exposure of 350 nm UV light.
(1) •OH chain reactions:
•OH + RH → R• + H2O [2.38]
R• + HOCl → RCl + •OH [2.39]
(2) •Cl chain reactions:
•Cl + RH → R• + HCl [2.40]
R• + HOCl → ROH + •Cl [2.41]
Since HOCl is the conjugate acid of OCl–, photochemical chain reactions of HOCl are
probably analogous to those of OCl–, which have been shown above. Feng et al. (2007) have
shown some possible chain reactions related to HOCl in Equations [2.42] and [2.43]. However,
based on their results, Feng et al. (2007) also proposed that the chain reactions of OCl–
photolysis may not be as important as those for HOCl.
•OH + HOCl → H2O + ClO• [2.42]
Reaction rate constant: k = 8.46 × 104 M–1 s–1 (Watts and Linden, 2007)
•Cl + HOCl → HCl + ClO• [2.43]
Reaction rate constant: k = 3 × 109 M–1 s–1 (NIST Database, 2002)
Feng et al. (2007) also indicated that the termination of HOCl chain reactions may involve:
15
•Cl + ClO• + H2O → 2 HOCl [2.44]
2 ClO• + H2O → HCl + HClO3 [2.45]
The quantum yields of HOCl/OCl– decay and photochemical product formation have been
evaluated. The quantum yield (Φ) is used to measure the photon efficiency in a photochemical
reaction. It is defined as the number of moles of reactant consumed or product produced per
einstein (1 einstein = one mole of photons = 6.022 x 1023 photons) of photons absorbed (Bolton,
2010). Buxton and Subhani (1972b) determined the quantum yields of OCl– decay and its
photochemical product formation at 365, 313, and 254 nm in OCl– solutions at concentrations of
10–4 to 10–3 M, which are shown in Table 2.2. Feng et al. (2007) reported the quantum yields of
1.0 ± 0.1 and 0.9 ± 0.1 for HOCl and OCl– decay, respectively, at 254 nm at initial
concentrations of less than 70 mg L–1 as free chlorine, while these values as determined by
Watts and Linden (2007) were 1.5 and 1.3 for HOCl and OCl–, respectively, at 254 nm at initial
concentrations of approximately 1–4 mg L–1 as free chlorine. The quantum yields higher than 1
are probably because the chain reactions that are initiated by the generated •OH and •Cl lead to
the further destruction of HOCl and OCl– (Feng et al., 2007; Watts and Linden, 2007). For
example, Feng et al. (2007) proposed that the linear increase of quantum yield of HOCl
photolysis with the increase of HOCl concentration from 71 to 1,350 mg L–1 as free chlorine
was caused by the promotion of the HOCl chain reaction at higher concentrations. However, the
detailed process of HOCl chain reactions has not yet been identified. In addition, Feng et al.
(2007) also found that the OCl– quantum yield was independent of its concentration ranging
from 3.5 to 640 mg L–1 as free chlorine. They therefore postulated no chain reaction initiated by
OCl– photoproducts, which conflicts with Buxton and Subhani (1972a, 1972b and 1972c).
The photolysis of free chlorine solutions is influenced by multiple factors. Based on Table
2.2, the OCl– quantum yield increases with the reduction of the wavelength. A higher production
ratio of •OH generated by HOCl at 254 nm was observed when compared to sunlight irradiation
at higher wavelengths (Nowell and Hoigné, 1992b). Watts and Linden (2007) found a much
higher quantum yield for HOCl decay (3.7) when exposed to a medium pressure (MP) mercury
UV lamp (a polychromatic UV source emitting 200 to 400 nm UV light), than the value of 1.5
when exposed to a low pressure (LP) mercury UV lamp (a monochromatic UV source only
emitting 254 nm UV light in the range of 200 to 400 nm). In addition, free chlorine photolysis is
16
also impacted by pH and different water matrices. Nowell and Hoigné (1992a) found that the
half-life of free chlorine species was longer at lower pH. Both Kobayashi and Okuda (1972) and
Watts and Linden (2007) reported that the water matrix was an impact factor for free chlorine
photolysis. This may be because the different compounds in the water play various roles in the
promotion or inhibition of chlorine photolysis chain reactions, and/or the different photon
absorption efficiencies of these compounds results in attenuation of photon irradiation on free
chlorine molecules to different extents (Kobayashi and Okuda, 1972; Watts and Linden, 2007).
Table 2.2 Quantum yields of OCl– decay and photochemical product formation
365 nm 313 nm 254 nm
OCl– 0.60 ± 0.02 0.39 ± 0.01 0.85 ± 0.02 ClO2
– 0.160 ± 0.005 0 0
O2 0.04 ± 0.02 0.069 ± 0.005 0.200 ± 0.005 ClO3
– 0.08 ± 0.02 0.08 ± 0.02 0.15 ± 0.02
Cl– 0.36 ± 0.03 0.27 ± 0.02 0.70 ± 0.03
The photolysis of OCl– was observed to be less efficient in the destruction of nitrobenzene
(NB) than HOCl photolysis under parallel conditions (Watts et al., 2007), which implies that the
concentration of •OH produced by UV/OCl– is lower than UV/HOCl. Nowell and Hoigné
(1992b) found that the •OH yield factor for OCl– photolysis was only 0.1 (i.e., 1 mole of OCl–
consumed can produce 0.1 mole of •OH), while this value for HOCl was 0.7 in sunlight and 0.9
in 254 nm UV light. However, less difference in the quantum yields of •OH formation by HOCl
and OCl– photolysis was reported by other studies, such as 1.40 and 0.28, respectively, reported
by Watts and Linden (2007) and Watts et al. (2007), or 0.46 and 0.61, respectively, reported by
Jin et al. (2011) and Chan et al.(2012). As a result, the lower efficiency of UV/OCl– relative to
UV/HOCl may be more attributed to the much stronger •OH scavenging efficiency of OCl– than
that of HOCl (shown in Equations [2.18] and [2.42]). For example, Watts et al. (2007) found
that almost 100% of the available •OH was consumed by NB at pH 5, while this value was only
less than 10% at pH 7. The other fraction of available •OH was consumed by OCl–. This,
therefore, implies that UV/HOCl is probably a more promising AOP than UV/OCl– (Watts and
Linden, 2007). Unfortunately, few studies have focused on this topic. Using •OH probe
17
compounds, i.e., para-chlorobenzoic acid (pCBA) and NB, Watts and Linden (2007) found that
the steady-state •OH concentrations were 5.7 × 10–13 M and 1.7 × 10–12 M generated by 1 and 3
mg L–1 free chlorine solutions, respectively, under LP UV exposure at pH 4. This •OH
concentration level is comparable to that generated by UV/H2O2 (Mamane et al., 2007). Watts
and Linden (2007) also proposed the potential use of UV/HOCl under mildly acidic conditions
as an alternative to UV/H2O2 AOP, based on their findings that HOCl produces •OH more
efficiently than does H2O2, but the •OH scavenging rate is lower than the latter. Watts et al.
(2007) evaluated the efficiency of UV/HOCl for treatment of NB, and found that the first-order
decay of NB increased following this order of treatment: UV/chlorine (at pH 7) < UV/H2O2 (at
pH 7) < UV/chlorine (at pH 6) < UV/chlorine (at pH 5). Additionally, in terms of the energy
demand, the efficiency of UV/HOCl AOP at pH 5 is also higher than UV/H2O2 and UV/chlorine
at higher pH values (Watts et al., 2007).
The photolysis of monochloramine exposed to LP and MP lamps was also investigated by
Watts and Linden (2007) and Zhang et al. (2015). Although NH2Cl has much higher molar
absorption coefficients at low UV wavelengths than HOCl and OCl– (based on Figure 2.2), the
quantum yields of NH2Cl photolysis was observed to be minimal, implying that it is not a
promising chemical that can be used as an AOP.
2.3 Formation of Disinfection By-Products (DBPs) in UV/Chlorine
2.3.1 Chlorinated DBPs
Free chlorine is the most commonly used primary disinfectant in North America to
inactivate pathogens and limit waterborne diseases, but it also leads to the formation of
chlorination DBPs, which can be regulated and/or may be suspected of being associated with
adverse health effects such as toxicity, carcinogenicity, mutagenicity, and/or genotoxicity
(USEPA, 1999; AWWA, 1999; Nieuwenhuijsen et al., 2000; Matilainen and Sillanpää, 2010).
THMs were the first group of DBPs discovered as a result of drinking water chlorination in
1974 (USEPA, 1999). At present, more than 700 halogenated DBPs have been reported,
whereas in many cases more than 50% of total organic halides (TOX) formed by chlorination is
still unknown (USEPA, 1999; Richardson, 2003; USEPA, 2002; Hansen et al., 2012; Matilainen
and Sillanpää, 2010). Most of the identified DBPs can be produced by the reactions between
18
free chlorine, hypobromous acid (HOBr) (from the oxidation of bromide by HOCl or ozone,
shown in Equations [2.46] and [2.47]) or hypoiodous acid (HOI) (mainly present in
chloramination, because of instability in the presence of free chlorine) and natural organic
matter (NOM), such as humic and fulvic acids (Glezer et al., 1999; Nieuwenhuijsen et al., 2000;
Kristiana et al., 2009; Richardson, 2003; Westerhoff et al., 2004). The reaction types between
halogens and NOM are mainly oxidation (i.e., breakage of carbon-carbon double bonds of NOM
molecules) and substitution (i.e., replacement of functional groups by halogen molecules)
(Westerhoff et al., 2004). Therefore, there are generally two approaches to reduce DBP
formation in water treatment: (1) using a non-chlorine-based disinfectant or minimizing chlorine
doses for primary disinfection, (2) reducing the amount of NOM before chlorination using
chemical and physical processes, such as coagulation, flocculation, and filtration (Chin and
Bérubé, 2005).
Br– + HOCl → HOBr + Cl– [2.46]
Reaction rate constant: k = 2950 M–1 s–1 (Westerhoff et al., 2004)
Br– + O3 → OBr– + O2 [2.47]
Reaction rate constant: k = 160 M–1 s–1 (Westerhoff et al., 2004)
Amongst many halogenated DBPs, THMs and HAAs occur most consistently and are
frequently associated with the highest concentrations during chlorination (AWWA, 1999;
Glezer et al., 1999; Nieuwenhuijsen et al., 2000; Tchobanoglous et al., 2003). THMs include
chloroform, bromodichloromethane, chlorodibromomethane, and bromoform (when not
considering the iodinated species of THMs). Compared with the other components, chloroform
is most prevalent (AWWA, 1999). HAAs, which are the second most common DBPs, consist of
9 chlorinated/brominated compounds: monochloro-, monobromo-, dichloro-, dibromo-,
bromochloro-, trichloro-, tribromo-, bromodichloro-, chlorodibromoacetic acids. Dichloro- and
trichloroacetic acids are the most prevalent HAAs (AWWA, 1999; Boorman et al., 1999;
USEPA, 2002). In addition, haloacetonitriles (HANs), haloketones (HKs), and chloropicrin (CP)
19
are also produced during chlorination with the amounts typically less than THMs and HAAs
(Glezer et al., 1999; Boorman et al., 1999; USEPA, 1999; Yang et al., 2007). HANs, HKs, and
CP have not been regulated yet; however, they have been included in the USEPA Information
Collection Rule for periodically monitoring requirements because of their potentially
carcinogenic and/or mutagenic effects on human health (Yang et al., 2007; USEPA, 1996;
Plewa et al., 2008; Muellner et al., 2007). Futhermore, inorganic DBPs, including perchlorate
(ClO4–), chlorate (ClO3
–), chlorite (ClO2–), and bromate (BrO3
–), may be also produced during
free chlorine photolysis (Buxton and Subhani, 1972a, 1972b, and 1972c; von Gunten and
Hoigné, 1994; Kang et al., 2006). These compounds have been considered for their health
concerns (Richardson et al., 2007; Korn et al., 2002; York et al., 2001). Table 2.3 enumerates
some typical halogenated by-produces, their median concentrations in treated drinking water in
the United States, and the USEPA’s MCLs, WHO’s guideline values, and/or Health Canada’s
MACs (Boorman et al., 1999; USEPA, 2009; WHO, 2011; Health Canada, 2012).
2.3.2 DBP Formation Kinetics
DBP formation by chlorination depends on a number of factors, including the type and
concentration of organic precursors, free chlorine concentration, bromide concentration, pH,
temperature, etc. (Tchobanoglous et al., 2003; USEPA, 1999; Yang et al., 2007). NOM is the
primary precursor of organic DBPs, which principally consists of humic material (hydrophobic)
and fulvic material (hydrophilic), including proteins, lipids, carbohydrates, carboxylic acids,
amino acids, and hydrocarbons (USEPA, 1999; Westerhoff et al., 2004). However, accurate
determination of NOM molecular propensities is very difficult because NOM is a heterogeneous
mixture of compounds. For example, the molecular weights of humic acid can vary from one
thousand to several hundred thousand grams per mole (Cameron et al., 1972).
In general, increasing pH promotes the formation of THMs but reduces the formation of
HAAs, HANs and HKs, while pH effects on different components might be different
(Pourmoghaddas and Stevens, 1995; Yang et al., 2007). For example, the formation of the four
components of THMs is enhanced at higher pH but to different extents (Peters et al., 1980;
Hansen et al., 2012; Liang and Singer, 2003); increasing pH has a minimal effect on the
formation of monohaloacetic acid and dihaloacetic acid, but may significant decrease
trihaloacetic acid formation (Liang and Singer, 2003) or may not (Hansen et al., 2012); and
20
dichloroacetonitrile, trichloroacetonitrile, and 1,1,1-trichloro-2-propanone formation decreases
with the increase of pH (Yang et al., 2007; Glezer et al., 1999; Hansen et al., 2012).
Table 2.3 Halogenated by-products in drinking water treatment
By-products U.S. Median
concentrations (μg L–1)
USEPA MCLs (mg L–1)
WHO (Provisional)
Guideline values (mg L–1)
Health Canada MACs (mg L–1)
THMs 0.08 (total THMs) 0.1 (total THMs) Chloroform 25 0.3 Bromodichloromethane 9.5 0.06 Chlorodibromomethane 1.6 0.1 Bromoform < 0.2 0.1 HAAs 0.06 (HAA5) 0.08 (total HAAs) Dichloroacetic acid 15 0.05 Trichloroacetic acid 11 0.2 Bromochloroacetic acid 2.69 Monochloroacetic acid 1.3 0.02 Dibromoaceic acid < 0.5 Monobromoacetic acid < 0.5 0.02 Tribromoacetic acid Bromodichloroacetic acid Chlorodibromoacetic acid Haloacetonitriles (HANs) Dichloroacetonitrile 2.1 0.02 Bromoacetonitrile 0.7 Bromochloroacetonitrile 0.6 Dibromoacetonitrile < 0.5 0.07 Trichloroacetonitrile < 0.02 Tribromoacetonitrile Haloketones (HKs) 1,1,1-Tichloropropanone 1.0 1,1-Dichloropropanone 0.4 1,3-Dichloropropanone Others Chlorate 161 0.7 1 Chlorite 1.0 0.7 1 Bromate 0.01 0.01 0.01 Perchlorate Chloral hydrate 2.1 Chloropicrin 0.4 MX 0.005 Cyanogen chloride 0.62 Cyanogen bromide Halonitriles 0.4 N-nitrosodimethylamine (NDMA) 0.0001 0.00004 2,4,6-Trichlorophenol 0.2 0.005 Formaldehyde 2-Chlorophenol
21
The mechanisms of DBP formation from NOM are still uncertain because of the high
heterogeneity and complexity of NOM components (Iriarte-Velasco et al., 2006). Among
various types of DBPs, THMs have been investigated most thoroughly (Gallard and von
Gunten, 2002). THM formation can generally be divided into two stages: a rapid initial reaction
stage followed by a slow reaction stage (Boccelli et al., 2003; Gallard and von Gunten, 2002;
Westerhoff et al., 2004). The THMs formed in the first stage account for approximately 15–30%
of the total THMs (Gallard and von Gunten, 2002). The two reaction stages probably result from
different reaction rates between chlorine and THM precursors, or different sites in the NOM
structure with different reactivities (Gallard and von Gunten, 2002; Greca and Fabbricino,
2008). For example, Gallard and von Gunten (2002) indicated that resorcinol-type structures,
which present 15–30% of THM precursors, and readily enolizable compounds, such as β-
diketones and β-ketoacids, might play a role in the fast reacting THM precursors, while other
phenolic compounds might be the slowly reacting THM precursors. Adin et al. (1991) proposed
that the two stage THM formation arose from a multi-step process, consisting of rapid reactions
between chlorine and precursors to produce chlorinated intermediates and the subsequently
slower reactions to produce THMs and other products.
The intersection point between the rapid and slow reaction stages is not well-defined. Sohn
et al. (2004) indicated that this point occurred at about 5 h of reaction time, which is generally
consistent with Iriarte-Velasco et al. (2006) and Gallard and von Gunten (2002), who both found
that the fast formation of THM from chlorination occurred in the first 3 h of reaction. In
contrast, Westerhoff et al. (2004) reported that the rapid initial reaction only lasted 1–5 min. The
reason for the discrepancy is unknown.
Gallard and von Gunten (2002) indicated that THM formation at the slow stage exhibited
second-order reaction kinetics with rate constants between 0.01 and 0.03 M–1 s–1, which was
controlled by the reactions between free chlorine and NOM. Iriarte-Velasco et al. (2006)
reported that the second-order rate constants of chloroform formation at the fast stage from the
chlorination of humic and fulvic acid were 1.47 and 1.43 M–1 s–1, respectively, at 20°C and pH
7.0, while the corresponding values for slow reaction were 0.24 and 0.129 M–1 s–1. Westerhoff et
al. (2004) investigated free chlorine consumption by NOM, and reported that the rate constant of
the initial rapid chlorine consumption was 50–500 M–1 s–1 within the first 1–5 min followed by
0.7–5 M–1 s–1 in the subsequent slower chlorine consumption stage. However, no past study
22
focused on DBP formation kinetics within the first one minute after addition of free chlorine,
which may be the scenario when the UV/chlorine AOP is applied.
Clark (1998) proposed that the formation rate of THMs was proportional to the
consumption rate of chlorine during the slow reaction stage. This is in agreement with Gang et
al. (2003), who pointed out that 31 to 42 μg THM was formed per mg Cl2 consumed. Gallard
and von Gunten (2002) also confirmed this linear relationship, and found 0.029 mol CHCl3
formed per mol Cl2 consumed (equivalent to 48.8 μg CHCl3 formed per mg Cl2 consumed). This
implies that during the slow reaction stage only a small fraction (less than 10%) of the
consumed chlorine is involved in the final formation of THMs.
2.3.3 DBP Formation by UV and AOPs
UV exposure alone cannot produce halogenated DBPs in the water without halogenating
agents such as chlorine and chloramine, while disinfection doses of UV (e.g., ≤ 200 mJ cm–2 for
4-log inactivation of viruses, based on USEPA Long Term 2 Enhanced Surface Water
Treatment Rule, LT2ESWTR) can even slightly decrease some DBPs, such as CDBM, TBM,
DBAA, DBAN (Lyon et al., 2012; Malley Jr. et al., 1995; USEPA, 2006). When extended UV
exposure (such as hours) with high UV doses (e.g., thousands of mJ cm–2) is applied, most of
common DBPs, such as THMs, HAAs, HANs, HKs, CP, as well as adsorbable organic halides
(AOX) and NOM, undergo slow photodecomposition processes (Hansen et al., 2013; Li and
Blatchley III, 2007; Weng et al., 2012; Kulovaara et al., 1996; Corin et al., 1996; Deng et al.,
2014; Höfl et al., 1997). Generally, bromated DBPs are more susceptible to UV exposure than
their chlorinated counterparts (Jo, 2008; Hansen et al., 2013).
Variable effects of UV treatment on DBP formation during secondary chlorination have
been reported in the literature. Certain studies have shown that THM, HAA, and AOX
formation during subsequent chlorination was not significantly affected by UV exposure
(Malley Jr. et al., 1995; Lyon et al., 2010; Lyon et al., 2012; Kashinkunti et al., 2004; Reckhow
et al., 2010; Liu et al., 2002a), while some other studies have shown contradictions or different
UV impacts on other DBPs. For example, Liu et al. (2006) and Liu et al. (2012) observed
significant increases of THM, HAA, and AOX formation during chlorination for 1 h, and 1, 3,
and 7 days, because of LP or MP UV pretreatment at a dose of 60 mJ cm–2. Lyon et al. (2012)
found that UV irradiation at disinfection doses did not significantly affect THM and HAA
23
formation during subsequent chlorination, but did increase CP when using MP UV rather than
LP UV. It was explained that MP UV, not LP UV, promotes nitrate photolysis to generate
reactive nitrogen species, which can react with NOM to produce CP, as also reported by
Reckhow et al. (2010) and Shah et al. (2011).
The enhancement effect of UV at an AOP dose on DBP formation during subsequent
chlorination is likely to be greater than when UV is applied at a disinfection dose. Lyon et al.
(2012) observed that MP UV pretreatment at a dose of 1,000 mJ cm–2 could increase TCM
formation by 30–40% during subsequent 24 h chlorination, while the impact of LP UV at such
doses on THM and HAA formation was low. MP UV did not increase HAA formation. This is
partially consistent with research carried out by Dotson et al. (2010), who found that MP UV
pretreatment increased THM formation to an extent much higher than that for LP UV, while
HAA formation was decreased by LP UV, but increased by MP UV. This UV effect was not
able to be reproduced by the similar work performed by Liu et al. (2002a), who found both
THM and HAA formation during secondary chlorination was not significantly increased by UV
pretreatment at doses of up to 5,000 mJ cm–2, while THM formation was slightly decreased.
Toor and Mohseni (2007) did not detect any impact of UV at a UV does of 2,500 mJ cm–2 on
THM formation during subsequent chlorination. The reason for this discrepancy is uncertain,
but reflects the complicated mechanisms of DBP formation by chlorination in water pretreated
by UV exposure. UV exposure is proposed to affect NOM structures and properties, and
subsequently to change its reactivity toward chlorine (Lyon et al., 2012; Liu et al., 2012).
During UV exposure, high-molecular-weight NOM components probably follow complicated
photochemical reactions to decompose to low-molecular-weight compounds with more carboxyl
and carbonyl carbon atoms (Kulovaara et al., 1996; Corin et al., 1996).
When UV-based AOPs are applied, •OH can eliminate undesired organic pollutants, as
well as various DBPs (Glauner et al., 2005; Höfl et al., 1997; Tang and Tassos, 1997; NIST
Database, 2002; Lifongo et al., 2004; Jo, 2008). •OH can also simultaneously react with NOM
resulting in the mineralization of NOM under extreme conditions (e.g., a UV/H2O2 AOP at a
UV dose of 3,500–5,000 mJ cm–2 with a H2O2 dose of 20–100 mg L–1), or changes in NOM
structure and reactivity in mild AOP treatment (e.g., a UV/H2O2 AOP at a UV dose in the order
of 1,000 mJ cm–2). Changes to NOM can lead to DBP reduction or enhancement depending on
the specific characteristics of the changes during subsequent chlorination (Dotson et al., 2010;
24
Toor and Mohseni, 2007; Pisarenko et al., 2013; Glauner et al., 2005; Goslan et al., 2006;
Sarathy et al., 2011; Rezaee et al., 2014; Matilainen and Sillanpää, 2010). For example, Toor
and Mohseni (2007) found that after the treatment of UV/H2O2 AOP at an H2O2 concentration
of 23 mg L–1 and a UV dose of ~3,500 mJ cm–2, THM and HAA formation was decreased by
~70% and ~40%, respectively, during secondary chlorination for 3 days, which is well
consistent with the work carried out by Lamsal et al. (2011), who showed a similar reduction in
THM and HAA formation during 24 h chlorination preceded by a UV/H2O2 AOP at a UV dose
of ~1,100 mJ cm–2 and 23 mg L–1 H2O2 concentration. Liu et al. (2002b) also reported reduction
in THM and HAA formation by 20–50% and 40–60%, respectively, during secondary
chlorination for 24 h after UV/H2O2 pretreatment at 100 mg L–1 initial H2O2 dose and 5,000 mJ
cm–2 UV dose. In contrast, Dotson et al. (2010) observed an approximately 10–20% increase of
THM and HAA formation during 24 h chlorination attributed to the UV/H2O2 pretreatment at a
LP or MP UV dose of 1,000 mJ cm–2 and a H2O2 concentration of 10 mg L–1, whereas AOX
formation either increased or decreased under different conditions. Likewise, Kleiser and
Frimmel (2000) reported that 50 min UV/H2O2 pretreatment at an initial H2O2 dose of 8 mg L–1
(UV dose unavailable) led to increases in THM and AOX formation by 20% and 5%,
respectively, during subsequent 48 h chlorination, while extended treatment of UV/H2O2 for up
to 1050 min resulted in the reduction of THM and AOX formation by 75% and 71%,
respectively. Bond et al. (2009) indicated that incomplete mineralization of NOM by the
UV/H2O2 process increased DCAA formation in subsequent chlorination, which was verified by
Toor and Mohseni (2007).
DBP formation in the UV/chlorine AOP is probably more complicated than that in other
AOPs, because the formation of Cl radicals (•Cl) by free chlorine photolysis may react with
NOM to produce DBPs, and inorganic DBPs, such as ClO4–, ClO3
–, ClO2–, and BrO3
–
(Pisarenko et al., 2013; Deng et al., 2014; Buxton and Subhani, 1972a, 1972b, and 1972c; von
Gunten and Hoigné, 1994; Kang et al., 2006). The role of •Cl in the UV/chlorine AOP is not
well established. It has been considered to be negligible in some of studies (Watts and Linden,
2007; Nowell and Hoigné, 1992b). In contrast, Fang et al. (2014) reported evidence that •Cl
reacts with benzoic acid, suggesting that if NOM were to contain structurally similar
components, chlorinated DBPs could possibly be formed.
25
To date, only a limited number of studies have investigated DBP formation by the
UV/chlorine AOP. Pisarenko et al. (2013) found HAA concentration was increased by ~50% by
UV/chlorine at 10 mg L–1 free chlorine and 3,900 mJ cm–2 LP UV dose (2 h exposure),
compared to a dark control with only chlorination, while THM concentration did not change,
and AOX was either increased or decreased under different conditions. Weng et al. (2012)
observed the promotion of DCAN formation by UV/chlorine at 3 mg L–1 chlorine and 120–360
mJ cm–2 LP UV dose (10–30 min exposure). Deng et al. (2014) reported the significant
promotion of CP formation by UV/chlorine for a 5 min exposure (chlorine doses: 4–10 mg L–1,
LP UV doses: 600–1,600 mJ cm–2). UV/chlorine has also been shown to affect subsequent DBP
formation upon chlorination. Liu et al. (2006) found UV/chlorine at 7 mg L–1 chlorine
concentration and 60 mJ cm–2 LP or MP UV dose led to increases in THMs and HAAs during
the subsequent 3 day chlorination, compared with those without pretreatment. Shah et al. (2011)
also found increased CP and DCAN formation in subsequent chlorination when pretreated with
UV/chlorine at UV doses ranging from 60 to 1,500 mJ cm–2 and a chlorine dose of 7 mg L–1.
2.4 Summary of Literature Review The literature review in this chapter discussed the following key points:
• AOPs are very effective in treating a variety of dissolved organic contaminants at trace levels.
• UV/chlorine is a potential AOP, since the photolysis of free chlorine in the UV range
produces hydroxyl radicals. However, chlorine photolysis may involve a complicated series
of chain reactions, which is not completely understood. Furthermore, UV/chlorine efficiency
is probably pH dependent, since the quantum yields of •OH formation and the •OH
scavenging efficiencies for both HOCl and OCl– were observed to be different.
• Organic and inorganic DBPs may be produced in the UV/chlorine process. Since higher
chlorine doses but much shorter contact times are probably applied in the UV/chlorine AOP
compared to conventional chlorine disinfection, DBP formation by UV/chlorine may be
different from that observed when applying chlorine conventionally, and has not been
explored.
26
References Adam, L.C., Fábián, I., Suzuki, K., Gordon, G., 1992. Hypochlorous acid decomposition in the
pH 5–8 region. Inorganic Chemistry, 31(17), 3534–3541.
Adin, A., Katzhendler, J., Alkaslassy, D., Rav-Acha, C., 1991. Trihalomethane formation in
chlorinated drinking water: a kinetic model. Water Research, 25(7), 797–805.
American Water Works Association (AWWA), 1999. Water Quality and Treatment: A
Handbook of Community Water Supplies, 5th Edition. McGraw-Hill, Inc., New York.
Benabbou, A.K., Derriche, Z., Felix, C., Lejeune, P., Guillard, C., 2007. Photocatalytic
inactivation of Escherischia coli Effect of concentration of TiO2 and microorganism, nature,
and intensity of UV irradiation. Applied Catalysis B: Environmental, 76(3–4), 257–263.
Betancourt, W.Q., Rose, J. B., 2004. Drinking water treatment processes for removal of
Cryptosporidium and Giardia. Veterinary Parasitology, 126(1–2), 219–234.
Boccelli, D.L., Tryby, M.E., Uber, J.G., Summers, R.S., 2003. A reactive species model for
chlorine decay and THM formation under dechlorination conditions. Water Research, 37(11),
2654–2666.
Bolton, J. R., 2010. Ultraviolet Applications Handbook, 3rd edition. Bolton Photosciences Inc.
Published by ICC Lifelong Learn Inc., Edmonton, AB, Canada.
Boorman, G.A., Dellarco, V., Dunnick, J.K., Chapin, R.E., Hunter, S., Hauchman, F., Gardner,
H., Cox, M., Sills, R.C., 1999. Drinking water disinfection byproducts: review and approach
to toxicity evaluation. Environmental Health Perspective, 107(Supplement 1), 207–217.
Bond, T., Goslan, E.H., Jefferson, B., Roddick, F., Fan, L., Parsons, S.A., 2009. Chemical and
biological oxidation of NOM surrogates and effect on HAA formation. Water Research,
43(10), 2615–2622.
Buxton, G.V., 1969. Pulse radiolysis of aqueous solutions. Some rates of reaction of OH and O–
and pH dependence of the yield of O3–. Transactions of the Faraday Society, 65, 2150–2158.
Buxton, G.V., Bydder, M., Salmon G.A., 1998. Reactivity of chlorine atoms in aqueous solution.
Part 1. The equilibrium Cl• + Cl– ↔ Cl2•–. Journal of the Chemical Society, Faraday
Transactions, 94(5), 653–657.
Buxton, G.V., Subhani, M.S., 1972a. Radiation chemistry and photochemistry of oxychlorine
ions. Part 1.–Radiolysis of aqueous solutions of hypochlorite and chlorite ions. Journal of the
27
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
947–957.
Buxton, G.V., Subhani, M.S., 1972b. Radiation chemistry and photochemistry of oxychlorine
ions. Part 2.–Photodecomposition of aqueous solutions of hypochlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
958–969.
Buxton, G.V., Subhani, M.S., 1972c. Radiation chemistry and photochemistry of oxychlorine
ions. Part 3.–Photodecomposition of aqueous solutions of chlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
970–977.
Cameron, R.S., Thornton, B.K., Swift, R.S., Posner, A.M., 1972. Molecular weight and shape of
humic acid from sedimentation and diffusion measurements on fractionated extracts. Journal
of Soil Science, 23(4), 394–408.
Chin, A., Bérubé, P.R., 2005. Removal of disinfection by-products precursors with ozone-UV
advanced oxidation process. Water Research, 39(10), 2136–2144.
Chapin, R.M., 1934. The effect of hydrogen-ion concentration on the decomposition of
hypohalites. Journal of the American Chemical Society, 56(11), 2211–2215.
Corin, N., Backlund, P., Kulovaara, M., 1996. Degradation products formed during UV-
irradiation of humic waters. Chemosphere, 33(2), 245–255.
Deborde, M., von Gunten, U., 2008. Reactions of chlorine with inorganic and organic
compounds during water treatment–kinetics and mechanisms: a critical review. Water
Research, 42(1–2), 13–51.
Deng, L., Huang, C.-H., Wang, Y.-L., 2014. Effects of combined UV and chlorine treatment on
the formation of trichloronitromethane from amine precursors. Environmental Science &
Technology, 48(5), 2697–2705.
Dotson, A.D., Keen, V.S., Metz, D., Linden, K.G., 2010. UV/H2O2 treatment of drinking water
increases post-filtration DBP formation. Water Research, 44(12), 3703–3713.
Edstrom Industries, Inc., 2003. Forms of chlorine in water. Available from
<http://www.edstrom.com/Resources.cfm?doc_id=164> (accessed 07.10.2009).
Eigen, M., Kustin, K., 1962. The kinetics of halogen hydrolysis. Journal of the American
Chemical Society, 84(8), 1355–1361.
28
Fang, J., Fu, Y., Shang, C., 2014. The roles of reactive species in micropollutant degradation in
the UV/free chlorine system. Environmental Science & Technology, 48(3), 1859–1868.
Feng, Y., Smith, D.W., Bolton, J.R., 2007. Photolysis of aqueous free chlorine species (HOCl
and OCl–) with 254 nm ultraviolet light. Journal of Environmental Engineering and Science,
6(3), 277–284.
Gallard, H., von Gunten, U., 2002. Chlorination of natural organic matter: kinetics of
chlorination and of THM formation. Water Research, 36(1), 65–74.
Gang, D.C., Clevenger, T.E., Banerji, S.K., 2003. Relationship of chlorine decay and THMs
formation to NOM size. Journal of Hazardous Materials, 96(1), 1–12.
Glauner, T., Kunz, F., Zwiener, C., Frimmel, F.H., 2005. Elimination of swimming pool water
disinfection by-products with advanced oxidation processes (AOPs). Acta Hydrochimica et
Hydrobiologica, 33(6), 585–594.
Glezer, V., Harris, B., Tal, N., Iosefzon, B., Lev, O., 1999. Hydrolysis of haloacetonitriles:
linear free energy relationship, kinetics and products. Water Research, 33(8), 1938–1948.
Goslan, E.H., Gurses, F., Banks, J., Parsons, S.A., 2006. An investigation into reservoir NOM
reduction by UV photolysis and advanced oxidation processes. Chemosphere, 65(7), 1113–
1119.
Greca, G.D., Fabbricino, M., 2008. DBP formation in drinking water: kinetics and linear
modelling. Water Science & Technology: Water Supply, 8(2), 161–166.
Guo, H., 1993. Time-dependent quantum dynamical study of the photodissociation of HOCl.
The Journal of Physical Chemistry, 97(11), 2602–2608.
Gültekin, I., Ince, N.H., 2007. Synthetic endocrine disruptors in the environment and water
remediation by advanced oxidation processes. Journal of Environmental Management, 85(4),
816–832.
Hansen, K.M.S., Willach, S., Antoniou, M.G., Mosbæk, H., Albrechtsen, H.-J., Andersen, H.R.,
2012. Effect of pH on the formation of disinfection byproducts in swimming pool water – Is
less THM better? Water Research, 46(19), 6399–6409.
Hansen, K.M.S., Zortea, R., Piketty, A., Vega, S.R., Andersen, H.R., 2013. Photolytic removal
of DBPs by medium pressure UV in swimming pool water. Science of the Total Environment
443, 850–856.
29
Health Canada, 2012. Guidelines for Canadian Drinking Water Quality – Summery Table.
Water, Air and Climate Change Bureau, Healthy Environments and Consumer Safety Branch,
Health Canada, Ottawa, Ontario, Canada.
Hijnen, W.A.M., Beerendonk, E.F., Medema, G.J., 2006. Inactivation credit of UV radiation for
viruses, bacteria and protozoan (oo)cysts in water: A review. Water Research, 40(1), 3–22.
Höfl, C., Sigl, G., Specht, O., Wurdack, I., Wabner, D., 1997. Oxidative degradation of AOX
and COD by different advanced oxidation processes: a comparative study with two samples
of a pharmaceutical wastewater. Water Science and Technology, 35(4), 257–264.
Iriarte-Velasco, U., Álvarez-Uriarte, J.I., González-Velasco, J.R., 2006. Kinetics of chloroform
formation from humic and fulvic acid chlorination. Journal of Environmental Science and
Health, Part A: Toxic/Hazardous Substances & Environmental Engineering, 41(8), 1495–
1508.
Jo, C.H., 2008. Oxidation of disinfection byproducts and algae-related odorants by UV/H2O2,
Ph.D. dissertation, Virginia Polytechnic Institute and State University, Blacksburg, Virginia,
USA.
Jung, Y.J., Oh, B.S., Kang, J-W., 2008. Synergistic effect of sequential or combined use of
ozone and UV radiation for the disinfection of Bacillus subtilis spores. Water Research, 42(6–
7), 1613–1621.
Kang, N., Anderson, T.A., Jackson, W.A., 2006. Photochemical formation of perchlorate from
aqueous oxychlorine anions. Analytica Chimica Acta, 567(1), 48–56.
Kashinkunti, R.D., Linden, K.G., Shin, G.-A., Metz, D.H., Sobsey, M.D., Moran, M.C.,
Samuelson, A.M., 2004. Investigating multibarrier inactivation for Cincinnati – UV, by-
products, and biostability. Journal of American Water Works Association, 96(6), 114–127.
Kleiser, G., Frimmel, F.H., 2000. Removal of precursors for disinfection by-products (DBPs) –
differences between ozone- and OH-radical-induced oxidation. The Science of the Total
Environment, 256(1), 1–9.
Kobayashi, T., Okuda, T., 1972. A continuous method of monitoring water quality by chlorine
consumption under ultraviolet radiation. Water Research, 6(2), 197–209.
Koivunen, J., Heinonen-Tanski, H., 2005. Inactivation of enteric microorganisms with chemical
disinfection, UV irradiation and combined chemical/UV treatments. Water Research, 39(8),
1519–1526.
30
Korn, C., Andrews, R.C., Escobar, M.D., 2002. Development of chlorine dioxide-related by-
product models for drinking water treatment. Water Research, 36(1), 330–342.
Kristiana, I., Gallard, H., Joll, C., Croué, J.-P., 2009. The formation of halogen-specific TOX
from chlorination and chloramination of natural organic matter isolates. Water Research,
43(17), 4177–4186.
Kulovaara, M., Corin, N., Backlund, P., Tervo, J., 1996. Impact of UV254-radiation on aquatic
humic substances. Chemosphere, 33(5), 783–790.
Kuo, J.-F., Smith, S.O., 1996. Disinfection. Water Environment Research, 68(4), 503–510.
Lamsal, R., Walsh, M.E., Gagnon, G.A., 2011. Comparison of advanced oxidation processes for
the removal of natural organic matter. Water Research, 45(10), 3263–3269.
Li, J., Blatchley III, E.R., 2007. Combined application of UV radiation and chlorine:
implications with respect to DBP formation and destruction in recreational water applications.
Proceedings of the Water Environment Federation Annual Technical Exhibition and
Conference 2007, Disinfection, San Diego, California, USA, 128–133.
Liang, L., Singer, P.C., 2003. Factors influencing the formation and relative distribution of
haloacetic acids and trihalomethanes in drinking water. Environmental Science & Technology,
37(13), 2920–2928.
Lifongo, L.L., Bowden, D.J., Brimblecombe, P., 2004. Photodegradation of haloacetic acids in
water. Chemosphere, 55(3), 467–476.
Liu, W., Andrews, S.A., Bolton, J.R., Linden, K.G., Sharpless, C., Stefan, M., 2002a.
Comparison of disinfection byproduct (DBP) formation from different UV technologies at
bench scale. Water Science and Technology: Water Supply, 2(5–6), 515–521.
Liu, W., Andrews, S.A., Sharpless, C., Stefan, M., Linden, K.G., Bolton, J.R., 2002b. Bench-
scale investigation into comparative evaluation of DBP formation from different UV/H2O2
technologies. Proceedings of 2002 AWWA Water Quality Technology Conference, Seattle,
WA, USA.
Liu, W., Cheung, L.-M., Yang, X., Shang, C., 2006. THM, HAA and CNCl formation from UV
irradiation and chlor(am)ination of selected organic waters. Water Research, 40(10), 2033–
2043.
31
Liu, W., Zhang, Z., Yang, X., Xu, Y., Liang, Y., 2012. Effects of UV irradiation and
UV/chlorine co-exposure on natural organic matter in water. Science of the Total
Environment, 414, 576–584.
Lyon, B., Weinberg, H., Dotson, A., Linden, K., 2010. Impact of UV-chlorine/chloramine
drinking water treatment on DBP formation in bromide- and nitrate-containing waters.
Proceedings of 2010 Water Quality Technology Conference and Exposition, Savannah,
Georgia, USA.
Lyon, B.A., Dotson, A.D., Linden, K.G., Weinberg, H.S., 2012. The effect of inorganic
precursors on disinfection byproduct formation during UV-chlorine/chloramine drinking
water treatment. Water Research, 46(15), 4653–4664.
Malley Jr., J.P., Shaw, J.P., Ropp, J.R., 1995. Evaluation of by-products produced by treatment
of groundwaters with ultraviolet irradiation. American Water Works Association Research
Foundation, Denver, CO., USA
Mamane, H., Shemer, H., Linden, K.G., 2007. Inactivation of E. coli, B. subtilis spores, and
MS2, T4, and T7 phage using UV/H2O2 advanced oxidation. Journal of Hazardous Materials,
146(3), 479–486.
Matilainen, A., Sillanpää, M., 2010. Removal of natural organic matter from drinking water by
advanced oxidation processes. Chemosphere, 80(4), 351–365.
Molina, M.J., Ishiwata, T., Molina, L.T., 1980. Production of OH from photolysis of HOCl at
307–309 nm. The Journal of Physical Chemistry, 84(8), 821–826.
Molina, L.T., Molina, M.J., 1978. Ultraviolet spectrum of HOCl. The Journal of Physical
Chemistry, 82(22), 2410–2414.
Morris, J.C., 1966. The acid ionization constant of HOCl from 5 to 35°. The Journal of Physical
Chemistry, 70(12), 3798–3805.
Muellner, M.G., Wagner, E.D., McCalla, K., Richardson, S.D., Woo, Y.-T., Plewa, M.J., 2007.
Haloacetonitriles vs. regulated haloacetic acids: are nitrogen-containing DBPs more toxic?
Environmental Science & Technology, 41(2), 645–651.
Nieuwenhuijsen, M.J., Toledano, M.B., Eaton, N.E., Fawell, J., Elliott, P., 2000. Chlorination
disinfection byproducts in water and their association with adverse reproductive outcomes: a
review. Occupational and Environmental Medicine, 57(2), 73–85.
32
NIST Database, 2002. NDRL/NIST chemical kinetics database on the web. Available from
<http://kinetics.nist.gov/solution> (accessed 07.10.2009).
Nowell, L.H., Hoigné, J., 1992a. Photolysis of aqueous chlorine at sunlight and ultraviolet
wavelengths–I. Degradation rates. Water Research, 26(5), 593–598.
Nowell, L.H., Hoigné, J., 1992b. Photolysis of aqueous chlorine at sunlight and ultraviolet
wavelengths–II. Hydroxyl radical production. Water Research, 26(5), 599–605.
Oliver, B.G., Carey, J.H., 1977. Photochemical production of chlorinated organic in aqueous
solutions containing chlorine. Environmental Science & Technology, 11(9), 893–895.
Oppenländer, T., 2003. Photochemical Purification of Water and Air Advanced Oxidation
Processes (AOPs): Principles, Reaction Mechanisms, Reactor Concepts. Wiley–VCH Verlag
GmbH & Co. KGaA, Weinheim.
Parsons, S., 2004. Advanced Oxidation Processes for Water and Wastewater Treatment. IWA
Publishing, London, UK.
Pera-Titus, M., García-Molina, V., Baños, M.A., Giménez, J., Esplugas, S., 2004. Degradation
of chlorophenols by means of advanced oxidation processes: a general review. Applied
Catalysis B: Environmental, 47(4), 219–256.
Peters, C.J., Young, R.J., Perry, R., 1980. Factors influencing the formation of haloforms in the
chlorination of humic materials. Environmental Science & Technology, 14(11), 1391–1395.
Pisarenko, A.N., Stanford, B.D., Snyder, S.A., Rivera, S.B., Boal, A.K., 2013.Investigation of
the use of chlorine based advanced oxidation in surface water: oxidation of natural organic
matter and formation of disinfection byproducts. Journal of Advanced Oxidation
Technologies, 16(1), 137–150.
Plewa, M.J., Wagner, E.D., Muellner, M.G., Hsu, K.-M., Richardson, S.D., 2008. Comparative
mammalian cell toxicity of N-DBPs and C-DBPs. In: Disinfection By-Products in Drinking
Water: Occurrence, Formation, Health Effects, and Control. American Chemistry Society,
Washington, D.C., USA, 36–50.
Pourmoghaddas, H., Stevens, A.A., 1995. Relationship between trihalomethanes and haloacetic
acids with total organic halogen during chlorination. Water Research, 29(9), 2059–2062.
Reckhow, D.A., Linden, K.G., Kim, J., Shemer, H., Makdissy, G., 2010. Effect of UV treatment
on DBP formation. Journal of American Water Works Association, 102(6), 100–113.
33
Rezaee, R., Maleki, A., Jafari, A., Mazloomi, S., Zandsalimi, Y., Mahvi, A.H., 2014.
Application of response surface methodology for optimization of natural organic matter
degradation by UV/H2O2 advanced oxidation process. Journal of Environmental Health
Science & Engineering, 12(67), 1–8.
Richardson, S.D., 2003. Disinfection by-products and other emerging contaminants in drinking
water. Trends in Analytical Chemistry, 22(10), 666–684.
Roots, R., Okada, M., 1975. Estimation of life times and differences of radicals involved in X-
ray-induced DNA strand breaks or killing of mammalian cells. Radiation Research, 64(2),
306–320.
Rosenfeldt, E.J., Linden, K.G., Canonica, S., von Gunten, U., 2006. Comparison of the
efficiency of •OH radical formation during ozonation and the advanced oxidation processes
O3/H2O2 and UV/H2O2. Water Research, 40(19), 3695–3704.
Sadiq, R., Rodriguez, M.J., 2004. Disinfection by-products (DBPs) in drinking water and
predictive models for their occurrence: a review. Science of the Total Environment, 321(1–3),
21–46.
Sarathy, S.R., Stefan, M.I., Royce, A., Mohseni, M., 2011. Pilot-scale UV/H2O2 advanced
oxidation process for surface water treatment and downstream biological treatment: effects on
natural organic matter characteristics and DBP formation potential. Environmental
Technology, 32(15), 1709–1718.
Shah, A.D., Dotson, A.D., Linden, K.G., Mitch, W.A., 2011. Impact of UV disinfection
combined with chlorination/chloramination on the formation of halonitromethanes and
haloacetonitriles in drinking water. Environmental Science & Technology, 45(8), 3657–3664.
Sohn, J., Amy, G., Cho, J., Lee, Y., Yoon, Y., 2004. Disinfection decay and disinfection by-
products formation model development: chlorination and ozonation by-products. Water
Research, 38(10), 2461–2478.
Svrcek, C., Smith, D.W., 2004. Cyanobacteria toxins and the current state of knowledge on
water treatment options: a review. Journal of Environmental Engineering and Science, 3(3),
155–185.
Tanaka, Y., Kawasaki, M., Matsumi, Y., Fujiwara, H., Ishiwata, T., Rogers, L.J., Dixon, R.N.,
Ashfold, M.N.R., 1998. The ultraviolet photodissociation of Cl2O at 235 nm and of HOCl at
235 and 266 nm. Journal of Chemical Physics, 109(4), 1315–1323.
34
Tang, W.Z., Tassos, S., 1997. Oxidation kinetics and mechanisms of trihalomethanes by
Fenton’s reagent. Water Research, 31(5), 1117–1125.
Tchobanoglous, G., Burton, F.L., Stensel, H.D., 2003. Wastewater Engineering, Treatment and
Reuse, 4th edition. The McGraw-Hill Companies, Inc., New York.
Tezcanli-Güyer, G., Ince, N.H., 2004. Individual and combined effects of ultrasound, ozone and
UV irradiation: a case study with textile dyes. Ultrasonics, 42(1–9), 603–609.
Thomsen, C.L., Madsen, D., Poulsen, J.A., Thøgersen, J., Jensen, S.J.K., Keiding, S.R., 2001.
Femtosecond photolysis of aqueous HOCl. Journal of Chemical Physics, 115(20), 9361–9369.
Toor, R., Mohseni, M., 2007. UV-H2O2 based AOP and its integration with biological activated
carbon treatment for DBP reduction in drinking water. Chemosphere, 66(11), 2087–2095.
United States Environmental Protection Agency (USEPA), 1996. ICR manual for bench- and
pilot-scale treatment studies. Available from
<http://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey= 20001VOG.txt> (accessed 08.20.2014).
United States Environmental Protection Agency (USEPA), 1999. Alternative Disinfectants and
Oxidants Guidance Manual. Available from
<http://www.epa.gov/safewater/mdbp/mdbptg.html#disinfect> (accessed 07.10.2009).
United States Environmental Protection Agency (USEPA), 2002. The occurrence of disinfection
by-products (DBPs) of health concern in drinking water: results of a nationwide DBP
occurrence study. Available from
<http://epa.gov/athens/publications/reports/EPA_600_R02_068.pdf> (accessed 07.10.2009).
United States Environmental Protection Agency (USEPA), 2006. Ultraviolet Disinfection
Guidance Manual for the Final Long Term 2 Enhanced Surface Water Treatment Rule.
Washington, DC.
United States Environmental Protection Agency (USEPA), 2009. National Primary Drinking
Water Regulations. Available from <http://water.epa.gov/drink/contaminants/index.cfm#List>
(accessed 07.10.2009).
Venczel, L.V., Arrowood, M., Hurd, M., Sobsey, M., 1997. Inactivation of Cryptosporidium
parvum oosysts and Clostridium perfringens spores by a mixed-oxidant disinfectant and by
free chlorine. Applied and Environmental Microbiology, 63(4), 1598–1601.
Vogt, R., Schindler, R.N., 1992. Product channels in the photolysis of HOCl. Journal of
Photochemistry and Photobiology A: Chemistry, 66(2), 133–140.
35
von Gunten, U., Hoigné, J., 1994. Bromate formation during ozonation of bromide-containing
waters: interaction of ozone and hydroxyl radical reactions. Environmental Science &
Technology, 28(7), 1234–1242.
Wang, L.K., Hung, Y.-T., Shammas, N.K., 2006. Handbook of Environmental Engineering,
Volume 4: Advanced Physicochemical Treatment Processes. The Humana Press Inc., Totowa,
NJ, USA.
Watts, M.J., Linden, K.G., 2007. Chlorine photolysis and subsequent OH radical production
during UV treatment of chlorinated water. Water Research, 41(13), 2871–2878.
Watts, M.J., Rosenfeldt, E.J., Linden, K.G., 2007. Comparative OH radical oxidation using UV-
Cl2 and UV-H2O2 processes. Journal of Water Supply: Research and Technology–AQUA,
56(8), 469–477.
Weng, S., Li, J., Blatchley III, E.R., 2012. Effects of UV254 irradiation on residual chlorine and
DBPs in chlorination of model organic-N precursors in swimming pools. Water Research,
46(8), 2674–2682.
Westerhoff, P., Chao, P., Mash, H., 2004. Reactivity of natural organic matter with aqueous
chlorine and bromine. Water Research, 38(6), 1502–1513.
Zhang, X., Li, W., Blatchley III, E.R., Wang, X., Ren, P., 2015. UV/chlorine process for
ammonia removal and disinfection by-product reduction: Comparison with chlorination.
Water Research, 68(1), 804–811.
36
3. MEDIUM PRESSURE UV COMBINED WITH CHLORINE
ADVANCED OXIDATION FOR TRICHLOROETHYLENE
DESTRUCTION IN A MODEL WATER
Most of this chapter has been previously published as:
Wang, D., Bolton J.R., Hofmann, R., 2012. Medium pressure UV combined with chlorine advanced oxidation
for trichloroethylene destruction in a model water. Water Research, 46(15), 4677–4686.
Reproduction of the paper has been permitted by the publisher.
Abstract The effectiveness of ultraviolet (UV) combined with chlorine as a novel advanced
oxidation process (AOP) for drinking water treatment was evaluated in a bench scale study by
comparing the rate of trichloroethylene (TCE) decay when using UV/chlorine to the rates of
decay by UV alone and UV/hydrogen peroxide (H2O2) at various pH values. The UV/chlorine
process was more efficient than the UV/H2O2 process at pH 5, but in the neutral and alkaline pH
range, the UV/H2O2 process became more efficient. The pH effect was probably controlled by
the increasing concentration of OCl– at higher pH values. A mechanistic kinetic model of the
UV/chlorine treatment of TCE showed good agreement with the experimental data.
3.1 Introduction Advanced oxidation processes (AOPs) involve the generation of the hydroxyl radical
(•OH), a very strong and non-selective chemical oxidant, that can destroy organic contaminants
in water and wastewater which are not readily oxidized by conventional oxidants (Parsons,
2004; Gültekin and Ince, 2007; Suty et al., 2004). Ultraviolet (UV)-based AOPs are becoming
more common in water treatment, and typically involve the addition of hydrogen peroxide
(H2O2) prior to the UV reactor (the UV/H2O2 AOP), with the photolysis of H2O2 forming •OH
(Parsons, 2004).
Ideally, a compound selected for photolysis in a UV-based AOP to generate •OH should be
able to absorb UV light efficiently and then produce a high yield of •OH. Although H2O2
photolysis gives an •OH formation quantum yield (Φ) of 1.11 ± 0.07 (Goldstein et al., 2007),
37
H2O2 absorbs UV only weakly. For example, the molar absorption coefficient of H2O2 at 254
nm, the emission peak of a low-pressure (LP) mercury UV lamp as a monochromatic light
source, is only ~19 M–1 cm–1 (Stefan et al., 1996; Baxendale and Wilson, 1957).
It may be possible to use the photolysis of aqueous active chlorine (i.e., hypochlorous acid,
HOCl, and/or hypochlorite ion, OCl–) as an alternative UV-based AOP, since active chlorine
species absorb UV photons and produce •OH relatively efficiently under some conditions
(Watts and Linden, 2007; Watts et al., 2007). HOCl and OCl– absorb UV light more efficiently
than H2O2 when using typical LP lamps or medium-pressure (MP) UV lamps (a polychromatic
light source emitting UV from 200 to 400 nm), according to their absorption spectra shown in
Figure 3.1. In general, when exposed to a LP lamp, HOCl and OCl– both absorb ~3.3 times
more UV photons than H2O2 at the same molar concentration. In the case of the MP lamp used
in this study, with the relative spectral emittance shown in Figure 3.2, HOCl and OCl– absorb
UV light 2.3 and 10.7 times more efficiently than H2O2, respectively, under equal molar
conditions, because of the high absorbance of OCl– at 290–320 nm. Compared to the well-
established mechanisms of H2O2 photolysis, however, the mechanisms of aqueous HOCl/OCl–
photolysis have not been fully explored. Evidence suggests that active chlorine
photodecomposition involves a series of chain reactions with the formation of many
intermediates and products (Buxton and Subhani, 1972a, 1972b and 1972c). Both •OH and
chlorine radicals (•Cl) are produced from photolysis at wavelengths less than 400 nm via
Equations [3.1] – [3.3] (Nowell and Hoigné, 1992a, 1992b; Oliver and Carey, 1977; Feng et al.,
2007; Watts and Linden, 2007).
HOCl + hv → •OH + •Cl [3.1]
OCl– + hν → •Cl + •O– [3.2]
•O– + H2O ↔ •OH + OH– [3.3]
Compared to •OH, •Cl was observed to play a negligible role in the oxidation of organic probes
(e.g., nitrobenzene and 1-chlorobutane), based on work performed by Nowell and Hoigné
(1992b) and Watts and Linden (2007).
The values reported in the literature for the quantum yields of •OH formation by HOCl and
OCl– photolysis vary. Watts and Linden (2007) and Watts et al. (2007) reported quantum yields
38
of 1.4 and 0.28 for HOCl and OCl–, respectively, while Jin et al., (2011) and Chan et al. (2012)
reported values of 0.46 and 0.61. The reasons for these discrepancies are not clear.
Figure 3.1 Absorption spectra of TCE, chlorine, peroxide and hydroxide species
Figure 3.2 Relative spectral emittance of the MP lamp in this research
0
200
400
600
800
1000
200 250 300 350 400
Wavelength (nm)
Mol
ar a
bsor
ptio
n co
effic
ient
s of
H
OC
l, O
Cl– , H
2O2,
HO
2– and
OH
–
(M–1
cm
–1)
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
Mol
ar a
bsor
ptio
n co
effic
ient
of
TCE
(M–1
cm
–1)
HO2–
OH–
HOCl
OCl–
H2O2
TCE
0
5
10
15
20
200 250 300 350 400
Wavelength (nm)
Rel
ativ
e sp
ectra
l em
ittan
ce
39
While HOCl and OCl– photolysis can produce •OH, they are also •OH scavengers as
shown in Equations [3.4] and [3.5], with reaction rate constants of 8.46 × 104 (Watts and
Linden, 2007) and 9.0 × 109 M–1 s–1 (Buxton and Subhani, 1972a), respectively.
•OH + HOCl → H2O + ClO• k = 8.46 × 104 M–1 s–1 [3.4]
•OH + OCl– → ClO• + OH– k = 9.0 × 109 M–1 s–1 [3.5]
Although both HOCl and OCl– photolysis can produce •OH, the concentrations of •OH
generated by the photolysis of HOCl and OCl– for equimolar concentrations are different,
because of their different molar absorption coefficients (shown in Figure 3.1), the different
quantum yields of •OH formation, and the different •OH scavenging efficiency for HOCl and
OCl–. Therefore, the efficiencies of HOCl and OCl– photolysis used as AOPs may not be the
same. Since the components of aqueous active chlorine are pH dependent [(pKa for HOCl at
25°C is 7.54 according to Deborde and von Gunten (2008)], the effectiveness of the
UV/chlorine AOP is therefore predicted to be sensitive to pH.
While there are still many unknowns about the specifics of active chlorine photolysis,
enough information is available to suggest that it is relatively efficient at absorbing UV and
produces •OH on photolysis. The goal of this study was to confirm the theoretical effectiveness
of the UV/chlorine AOP in the context of drinking water treatment, using the destruction of
trichloroethylene (TCE) as an example.
TCE (ClHC=CCl2, molecular weight: 131.39 g mol–1) is susceptible to •OH oxidation with
a reaction rate constant of 2.4 × 109 M–1 s–1 (Li et al., 2007; Li et al., 2004). The Middleton
Water Supply System in Waterloo (Ontario, Canada) has already explored the application of the
UV/chlorine process at full scale to remove TCE from groundwater (Wang et al., 2011). This
study is intended to verify further its feasibility and to explore the UV/chlorine treatment from a
fundamental perspective under controlled laboratory conditions. Bench-scale experiments using
a MP collimated beam apparatus and a mathematical modeling analysis were both performed to
evaluate the effectiveness of the UV/chlorine AOP, comparing it to the more conventional
UV/H2O2 AOP under parallel conditions.
40
3.2 Materials and Methods
3.2.1 Reagents and Materials
An appropriate volume of TCE (≥99.5%, A.C.S. grade, Sigma-Aldrich) was diluted in
Milli-Q® water to prepare the working solutions to be exposed at a TCE concentration of
approximately 145 μg L–1 (1.10 × 10–6 M). Active chlorine and hydrogen peroxide solutions
were prepared from sodium hypochlorite (NaOCl) solution (10–15 wt. %, reagent grade, Sigma-
Aldrich) and H2O2 solutions (50 wt. %, Sigma-Aldrich), respectively. The concentrations of
active chlorine and H2O2 in the working solutions were both approximately 0.15 mM. Other
compounds used in this research were all analytical reagent grade. Milli-Q® water was used in
all experiments and analytical determinations. The Milli-Q® water contained a low
concentration of total organic carbon (TOC) (approximately 0.1 mg L–1 as C), which was
included in the modeling studies.
3.2.2 UV Exposure and Irradiance Measurements
A 1 kW MP mercury UV lamp (Heraeus Noblelight GmbH, Germany) installed in a
collimated beam apparatus (Model: PS1-1-120, Calgon Carbon Corporation) was used to expose
15 mL TCE samples with/without the addition of active chlorine or H2O2 contained in Pyrex®
Petri dishes (inner diameter: 4.9 cm). The samples were buffered at pH 5, 7.5 or 10 using 5 mM
phosphate and/or borate buffers. The exact pH was adjusted by adding NaOH and/or H2SO4.
According to the pKa of HOCl at 25°C, at pH 5, 7.5 and 10, the active chlorine components
consisted of 99.7% HOCl, 52.3% HOCl + 47.7% OCl–, and 99.6% OCl–, respectively. An
appropriate length of exposure time was performed to deliver the desired fluence into each
sample. The incident irradiance (mW cm–2) of samples from 200 to 345 nm was calculated from
the difference measured by ferrioxalate actinometry with and without a 3 mm-thick 345 nm
long-pass filter placed in the light path, according to the procedures described by Bolton et al.
(2009) and Sharpless and Linden (2003). The different quantum yields of Fe2+ formation in the
ferrioxalate system at the various wavelengths were also considered (Goldstein and Rabani,
2008). After obtaining the relative spectral emittance of the MP lamp (shown in Figure 3.2), the
fluence rate from 200 to 400 nm was calculated proportionally (shown in Appendix B). The
41
correction factors, such as Petri factor, water factor, divergence factor and reflection factor,
were also applied as per the procedures described by Bolton and Linden (2003).
3.2.3 Analytical Methods
The active chlorine (HOCl and/or OCl–) concentrations in working solutions were
determined by the DPD colorimetric method (APHA et al., 2005), using a HACH®
spectrophotometer (Model: DR/2500, HACH). H2O2 concentrations were analyzed with the
triiodide method (Klassen et al., 1994). Gas chromatography-electron capture detector (GC-
ECD) (Model: HP 5890 Series II, Hewlett-Packard) was applied for measurement of TCE,
following the USEPA method 551.1 (USEPA, 2008). The method detection limit of TCE was 1
μg L–1. Details for the analytical method are shown in Appendix G. Chlorine and H2O2 residuals
were quenched by sodium sulphite (Na2SO3) before TCE analysis. The relative spectral
emittance of the MP lamp was measured using a calibrated spectroradiometer (Model:
USB4000-UV-VIS, Ocean Optics) with a fiber-optic cable (Model: QP200-2-SR-BX, Ocean
Optics) and a cosine corrector (Model: CC-3-UV, Ocean Optics). In addition, a Cecil UV/vis
spectrophotometer (Model: CE3055, Cecil Instruments) and a PerkinElmer UV/vis
spectrophotometer (Model: Lambda 25, PerkinElmer) were applied for measuring solution
absorbances at a single wavelength and in the entire UV band, respectively.
3.3 Results and Discussion
3.3.1 Molar Absorption Coefficients of TCE, Active Chlorine, Peroxide and Hydroxide
Species
As part of this work, mathematical models of the UV/chlorine and the UV/H2O2 AOPs
were built to interpret experimental results. These models required measurement of the molar
absorption coefficients of TCE, HOCl (at pH 5.0), OCl– (at pH 10.0) and H2O2 (at pH ~5.5) at
wavelengths ranging from 200 to 400 nm, shown in Figure 3.1. The maximum molar absorption
coefficients of HOCl and OCl– were 98 and 359 M–1 cm–1, peaking at 235 and 292 nm,
respectively. These values are very close to those measured by Feng et al. (2007), who found
HOCl and OCl– peak molar absorption coefficients of 101 and 365 M–1 cm–1 at 236 and 292 nm,
respectively. The molar absorption coefficient of HO2– was also required because H2O2
disassociates in strong alkaline solutions (Equation [3.6]) (Czapski and Bielski, 1963).
42
H2O2 ↔ H+ + HO2– pKa = 11.8 [3.6]
Unlike other compounds in this study, however, the absorbance of HO2– could not be measured
directly, because at pH near 13.8, where ~100% H2O2 converts to HO2–, the absorbance of OH–
at the wavelengths <220 nm is too high to allow measurement of the HO2– absorbance
accurately (e.g., OH– absorbance at 200 nm >1.9, according to Figure 3.1). As a result, in order
to determine the molar absorption coefficient of HO2–, the absorbance of a mixture of H2O2 and
HO2– containing a low concentration of OH– (1 mM) was first measured, followed by accurately
measuring the total concentration of peroxide species and the pH. Using the pKa of H2O2, the
respective concentrations of H2O2 and HO2– were subsequently calculated. After the molar
absorption coefficients of H2O2 and OH– were obtained, the corresponding values (also shown
in Figure 3.1) for HO2– were calculated. Although the values may not be as accurate as those
measured directly, they are generally reliable, compared with the data at several selected
wavelengths reported by Baxendale and Wilson (1957). For example, the molar absorption
coefficients of HO2– at 254, 270, and 290 nm were calculated to be 269, 145, 31 M–1 cm–1,
respectively, compared with 229, 122 and 45 M–1 cm–1 at the corresponding wavelengths
directly measured by Baxendale and Wilson (1957).
3.3.2 Quantum Yields of Active Chlorine and Hydrogen Peroxide Photolysis
When performing the exposure experiments for TCE decay by the UV/chlorine and the
UV/H2O2 AOPs using the collimated beam apparatus, the fluence-based decay rate s for active
chlorine and H2O2 (Figure 3.3 and Table 3.1) first had to be determined to estimate the quantum
yields of active chlorine and H2O2 photolysis, following methods reported by Bolton and Stefan
(2002) and Stefan and Bolton (2005). Details for chlorine and H2O2 photolysis are included in
Appendix I. It is noted that these observed quantum yields are specific to the water matrix, and
in particular the presence of •OH scavengers. The •OH scavengers may either inhibit the
photodegradation of active chorine or hydrogen peroxide, since the scavengers consume •OH
and diminish the reactions between •OH and active chlorine components or hydrogen peroxide;
or in contrast, may enhance their photolysis, since the products of scavengers may further react
with active chlorine components and/or hydrogen peroxide. For example, Feng et al. (2010) and
43
Jin et al., (2011) found that in the presence of TOC and methanol, the photodegradation of
HOCl is enhanced and thus increases the observed photolysis quantum yield.
Figure 3.3 Rates of active chlorine (a) and hydrogen peroxide (b) photolysis at various pH
values. Error bars represent the standard deviations of triplicate runs. Straight lines
represent the linear regression.
The observed quantum yields determined here for HOCl, OCl– and H2O2 photolysis in the
water matrices studied are 1.06 ± 0.01, 0.89 ± 0.02 and 0.76 ± 0.01, respectively. The chlorine
quantum yields match closely the values reported by Feng et al. (2007), but are considerably
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0 5000 10000 15000 20000
Fluence (J m–2)
ln([a
ctiv
e C
l]/[a
ctiv
e C
l] 0) pH 5 pH 7.5 pH 10
(a)
-0.12
-0.1
-0.08
-0.06
-0.04
-0.02
0
0 5000 10000 15000 20000
Fluence (J m–2)
ln([H
2O2]
/[H2O
2]0)
pH 5 pH 7.5 pH 10
(b)
44
different from some of the other reported quantum yields (Table 3.2). Likewise, the quantum
yield for hydrogen peroxide photolysis in this study (0.76) is lower than the published value of
~1.0 (Hunt and Taube, 1952; Baxendale and Wilson, 1957; Volman and Chen, 1959; Goldstein
et al. 2007). The difference of the H2O2 photolysis quantum yields between this study and the
literature may arise from the presence of TCE in this study, compared to organic-free water used
in the other studies. Like chlorine species, the photolysis of H2O2 may involve chain reactions
(Crowell et al., 2004; Luňák and Sedlák, 1992). The •OH formed by H2O2 photolysis may react
with H2O2, resulting in an increase in the observed quantum yield of H2O2 photolysis.
Therefore, the presence of •OH scavengers (such as TCE) may induce a lower observed
photolysis quantum yield. Indeed, results from other trials in this study (not reported) indicated
that the quantum yield of H2O2 photolysis increased by ~10% in the absence of TCE. In
addition, the previous studies reporting the quantum yield of H2O2 photolysis were also
conducted exclusively using monochromatic LP lamps, as opposed to the polychromatic MP
lamp used in this study. The calculated quantum yield of photolysis would therefore differ if
they were to vary with wavelength over the emission spectrum of the MP lamp. Further research
is needed under controlled conditions (e.g. well-defined •OH scavenging, single wavelengths,
etc.) to confirm the quantum yields of active chlorine or hydrogen peroxide photolysis in the
absence of water matrix and lamp effects.
Table 3.1 Fluence-based rate constants (10–6 m2 J–1) for active chlorine and peroxide
photolysis
pH 5 pH 7.5 pH 10
Chlorine species 15.0 ± 0.1 33.7 ± 0.4 58.7 ± 1.5 H2O2 species 4.74 ± 0.09 4.77 ± 0.04 5.67 ± 0.02
3.3.3 TCE Decay Rates by UV Alone, and the UV/Chlorine and the UV/H2O2 AOPs
TCE decay rates were measured in the presence of UV alone, as well as in the UV/chlorine
and the UV/H2O2 AOPs. TCE is sensitive to both UV exposure and •OH oxidation, and is also
volatile (Li et al., 2004; Li et al., 2007; Peng and Wan, 1997). The measured decrease of TCE
concentration therefore arises from the sum of evaporative loss, direct photolysis by UV
45
exposure, and •OH oxidation loss (if AOPs are applied). In a dilute aqueous solution, the change
of a solute concentration arising from evaporation is first-order (Tinsley, 2004). Since the rate of
TCE photolysis and the rate of TCE decay by •OH oxidation also follow first-order kinetics
(assuming steady-state •OH concentration), the total rate of TCE loss follows first-order
kinetics, as described in Equation [3.7] (Schwarzenbach et al., 2003; Rosenfeldt and Linden,
2004; Rosenfeldt, 2005).
Table 3.2 Comparison of reported quantum yields of active chlorine photolysis
Chlorine species
Quantum yield
Experimental conditions Reference Initial chlorine
concentration wavelengt
h (nm) pH
OCl– 0.60 ± 0.02 7 mM 365 12.1 Buxton and Subhani (1972b)
OCl– 0.39 ± 0.01 1 mM 313 12.0 Buxton and Subhani (1972b)
OCl– 0.85 ± 0.02 1 mM 254 11.5 Buxton and Subhani (1972b)
HOCl 1.0 ± 0.1 < 2mM 254 5 Feng et al. (2007)
OCl– 0.9 ± 0.1 < 2 mM 254 10 Feng et al. (2007)
HOCl 1.5 0.014 – 0.056 mM 254 4 Watts and Linden (2007)
OCl– 1.3 0.014 – 0.056 mM 254 10 Watts and Linden (2007)
HOCl 3.7 0.014 – 0.056 mM MP lamp 4 Watts and Linden (2007)
OCl– 1.7 0.014 – 0.056 mM MP lamp 10 Watts and Linden (2007)
HOCl 1.0 ± 0.1 1.41 mM LP lamp 5 Jin et al. (2011)
OCl– 1.15 ± 0.08 1.41 mM LP lamp 10 Jin et al. (2011)
OCl– 0.87 ± 0.01 0 – 4.23 mM 303 10 Chan et al. (2012)
HOCl 1.06 ± 0.01 0.15 mM MP lamp 5 This work
OCl– 0.89 ± 0.02 0.15 mM MP lamp 10 This work
evaporation UV OH0
ln ( )C k k k tC •= − + + [3.7]
where, C is the final TCE concentration (M) after the exposure, C0 is the initial TCE
concentration (M) before the exposure, kevaporation is the first-order evaporative rate constant (s–1)
of TCE, kUV is the first-order photolysis rate constant (s–1) of TCE due to UV exposure, k•OH is
the first-order decay rate constant (s–1) of TCE due to •OH oxidation, and t is the exposure time
46
(s). kevaporation was determined from the reduction of TCE concentration of a 15 mL sample (with
no addition of active chlorine or H2O2) that was placed in the same Petri dish used for
experiments and stirred for 5 min (the longest time in exposure experiments). According to 10
replicates, 25.0 ± 0.51% of initial TCE concentration was lost due to evaporation. Therefore, the
evaporative rate was 9.6 × 10–4 s–1. After subtracting the evaporation loss from the total loss of
TCE, the TCE destruction rate arising from photolysis and •OH oxidation is calculated using
Equation [3.8]. Details for TCE decay after subtracting the evaporation loss are shown in
Appendix I.
evaporation UV OH0
ln ( )C k t k k tC •+ = − + [3.8]
Since the fluence (mJ cm–2) delivered into the sample is the product of the exposure time (s) and
the fluence rate (mW cm–2) in the same wavelength range, Equation [3.8] is equivalent to:
evaporation UV OH0
ln ( )C k t k k FC •′ ′+ = − + [3.9]
where F is the fluence (mJ cm–2) from 200 to 400 nm, and UVk′ and OHk•′ are fluence-based rate
constants (cm2 mJ–1) (Bolton and Stefan, 2002; Stefan and Bolton, 2005), while the unit for
kevaporation is still s–1. The fluence rate is considered to be constant during the exposure, since in
low absorbance solutions the decrease of solute concentrations impacts the fluence rate only
slightly. Therefore, evaporation0
ln C k tC
+ is a linear function of F with a slope of ( )UV OHk k•′ ′− + . The
experimental results are shown in Figure 3.4. The TCE decay rate constants with the standard
deviations of triplicates after subtracting out evaporation under different conditions are
summarized in Table 3.3. The corresponding values calculated from numerical models are also
shown in Table 3.3, and will be discussed later.
47
Figure 3.4 Experimental TCE decay rates by UV alone, the UV/chlorine and the UV/H2O2
AOPs at various pH values. Error bars represent the standard deviations of triplicate
runs. Straight lines represent the linear regression.
-3
-2
-1
0
0 500 1000 1500 2000
Fluence (mJ cm–2)
ln(C
/C0)
+ k
evap
orat
ion
t
UV alone UV+Chlorine UV+H2O2
pH 5
UV+H2O2
-3
-2
-1
0
0 500 1000 1500 2000
Fluence (mJ cm–2)
ln(C
/C0)
+ k
evap
orat
ion
t
UV alone UV+chlorine UV+H2O2
pH 7.5
UV+H2O2
-3
-2
-1
0
0 500 1000 1500 2000
Fluence (mJ cm–2)
ln(C
/C0)
+ k
evap
orat
iont
UV alone UV+Chlorine UV+H2O2
pH 10
UV+H2O2
48
Table 3.3 TCE Fluence-based decay rate constants (10–4 cm2 mJ– 1) (excluding evaporation)
by UV alone, and the UV/chlorine and the UV/H2O2 AOPs from experimental and model
results
pH 5 pH 7.5 pH 10
Experimental Model Experimental Model Experimental Model
UV alone 7.23 ± 0.25 5.30 6.03 ± 0.05 5.30 5.25 ± 0.03 5.30
UV/chlorine AOP 80.9 ± 1.0 54.4 8.15 ± 0.03 7.89 7.33 ± 0.20 7.68
UV/H2O2 AOP 34.7 ± 1.1 35.3 32.3 ± 0.8 35.2 9.69 ± 0.09 11.8
According to Figure 3.4 and Table 3.3, TCE is photolyzed by UV exposure at a fairly high
photolysis rate (i.e., a fairly high kUV). For example, more than 60% of TCE is destroyed at a
fluence of approximately 1,900 mJ cm–2 from 200 to 400 nm. Based on the rates of TCE
photolysis (in the case of UV alone) in Table 3.3, the observed quantum yields of TCE
photolysis at pH 5, 7.5 and 10 were calculated to be 0.64 ± 0.02, 0.53 ± 0.04 and 0.46 ± 0.03,
respectively, following the methods discussed by Bolton and Stefan (2002) and Stefan and
Bolton (2005). This is in general agreement with Li et al. (2004), who indicated that TCE is
fairly sensitive to UV exposure, comprising several photochemical pathways with an overall
quantum yield of 0.354 at a neutral pH. Interestingly, however, as the pH increases, the TCE
photolysis rate constant declines slightly. The reason is not clear. It may be because in an
alkaline solution more OH– is available to scavenge •Cl as shown in Equation [3.10], which is
an intermediate produced by TCE photolysis and promotes the destruction of TCE (Kläning and
Wolff, 1985; Li et al., 2004).
•Cl + OH– → •ClOH– k = 1.8 × 1010 M–1 s–1 [3.10]
From Figure 3.4 and Table 3.3, the UV/chlorine and the UV/H2O2 AOPs were found to be
more efficient at destroying TCE than UV alone because of the additional loss of TCE by •OH
oxidation. When comparing the TCE decay rates by the UV/chlorine and the UV/H2O2 AOPs at
the same pH, the UV/chlorine AOP is observed to be 2.3 times more efficient for destruction of
TCE than the UV/H2O2 AOP at pH 5, while it is less efficient than the latter at pH 7.5 and 10.
49
Moreover, it is obvious that the efficiencies of both AOPs become lower at a higher pH because
the TCE decay rates by both AOPs decrease with increasing pH. In particular, the efficiency of
the UV/chlorine AOP decreases significantly faster than that of the UV/H2O2 AOP.
The different efficiencies of the UV/chorine versus UV/H2O2 processes are largely a
reflection of different steady-state concentrations of hydroxyl radicals, [•OH]ss, with a higher
[•OH]ss giving a faster rate of TCE decay. [•OH]ss can be calculated using Equation [3.11]
(Rosenfeldt and Linden, 2004).
k•OH = k•OH-TCE × [•OH]ss [3.11]
where, k•OH is the first-order decay rate constant (s–1) of TCE due to •OH oxidation, k•OH-TCE is
the second-order rate constant (M–1 s–1) between •OH and TCE, k•OH-TCE = 2.4 × 109 M–1 s–1, and
[•OH]ss is the steady-state concentration (M) of •OH. k•OH can be determined experimentally
using values from Table 3.3 by subtracting the reaction rate constants of TCE photolysis, kUV,
from the overall reaction rate constants of TCE decay observed in the UV/chorine or UV/H2O2
experiments. It is noted that the k•OH calculated this way has units of cm2 mJ–1, which is different
from the units (s–1) used in Equation [3.11]. Therefore, in order to fit Equation [3.11], k•OH must
be converted from cm2 mJ–1 to s–1 by multiplying the value of k•OH by the average fluence rate
(mW cm–2) in the solution. Table 3.4 presents the calculated [•OH]ss in the UV/chlorine and the
UV/H2O2 AOP tests at different pHs, based on Table 3.3 and Equation [3.11]. In addition to
Equation [3.11], assuming that the quantum yield of •OH formation is independent of
wavelength, [•OH]ss can be theoretically determined from Equation [3.12], as described by
Schwarzenbach et al. (2003).
( )OH
ssOH-
( ) [1 10 ][Ox]
[ OH][ ]
a zp
jj
Ea z
k j
λλ
λ λ
λ εΦ
−
•
•
−⋅ ⋅
• =⋅
∑∑
[3.12]
where, Φ•OH is the quantum yield of •OH formation by chlorine or hydrogen peroxide
photolysis, Ep(λ) is the incident photon irradiance (10–3 einstein cm–2 s–1) at wavelength λ, ελ is
50
the molar absorption coefficient (M–1 cm–1) of the chlorine or hydrogen peroxide species at
wavelength λ, aλ is the decadic absorption coefficient (cm–1) of the TCE solution at wavelength
λ, z is the solution depth (cm), [Ox] is the concentration (M) of the oxidant, i.e., chlorine or
hydrogen peroxide species, k•OH-j is the second-order reaction rate constants (M– 1 s– 1) between
•OH and a scavenger, and [j] is the concentration (M) of the corresponding •OH scavenger j.
Table 3.4 Calculated hydroxyl radical concentrations (10–13 M) in TCE solutions treated
by the UV/chlorine and the UV/H2O2 AOPs at various pH values
pH 5 pH 7.5 pH 10
UV/chlorine AOP 197 ± 2.8 5.58 ± 0.16 5.48 ± 0.54 UV/H2O2 AOP 73.3 ± 3.0 70.1 ± 2.1 11.8 ± 0.2
According to Equation [3.12] and the calculated [•OH]ss for the UV/chlorine AOP at pH 5
and 10 and the UV/H2O2 AOP at pH 5 shown in Table 3.4, the observed quantum yields of •OH
formation by the photolysis of HOCl, OCl– and H2O2 are determined to be 0.79 ± 0.01, 1.18 ±
0.12 and 1.15 ± 0.05, respectively. The value for H2O2 (1.15 ± 0.05) determined in this study is
very close to the literature (1.11 ± 0.07) (Goldstein et al., 2007), while the values for HOCl and
OCl– (0.79 ± 0.01 and 1.18 ± 0.12, respectively) are quite different from other published data:
e.g., 1.40 and 0.28, respectively (Watts and Linden, 2007; Watts et al. 2007), or 0.46 (Jin et al.,
2011) and 0.61 (Chan et al., 2012). This difference could arise from the different experimental
conditions in these studies, such as active chlorine concentrations, •OH scavenging potentials,
UV lamp types, etc.
According to Equation [3.12], the variation of [•OH]ss with pH arises theoretically from the
difference in Φ•OH, ελ and/or k•OH-j[j]. For the UV/chlorine AOP, the components of chlorine
species at pH 5, 7.5 and 10 are very different (pKa = 7.54). Although OCl– absorbs more photons
than HOCl and the Φ•OH for OCl– is higher than that for HOCl, the much higher reported
reaction rate of OCl– with •OH than that for HOCl (reaction rate coefficient: 9.0 × 109 M–1s–1 for
OCl– vs. 8.46 × 104 M–1s–1 for HOCl) makes the UV/chlorine AOP less efficient with increasing
pH. For the UV/H2O2 AOP, although the quantum yield of •OH formation from HO2–
photolysis is unknown, it is likely similar to that of H2O2. Therefore, the lower efficiency of the
UV/H2O2 AOP at a higher pH may also arise principally from the much higher (~278 times)
51
•OH scavenging efficiency of HO2– than H2O2 as shown in Equations [3.13] and [3.14] (Stefan
et al., 1996).
H2O2 + •OH → HO2• + H2O k = 2.7 × 107 M–1 s–1 [3.13]
HO2– + •OH → •O2
– + H2O k = 7.5 × 109 M–1 s–1 [3.14]
In addition, because of a relatively high pKa for H2O2 disassociation equilibrium constant
compared to that of HOCl (pKa = 11.8 for H2O2 vs. 7.54 for HOCl), only a small fraction (1.6%)
of H2O2 is converted to HO2– at pH 10. The result of these factors is that the efficiency of the
UV/chlorine AOP decreases more quickly with increasing pH than that of the UV/H2O2 AOP.
3.3.4 Mathematical Modeling of the TCE Decay
Numerical models were built with Matlab® to simulate TCE decay by UV alone, as well as
by the UV/chlorine and the UV/H2O2 AOPs. The TCE decay rate arising from UV exposure was
first predicted based on the quantum yields of TCE photolysis through different pathways
discussed by Li et al. (2004). The TCE destruction rates by the UV/chlorine and the UV/H2O2
AOPs were subsequently simulated according to Equations [3.8], [3.11] and [3.12]. The
parameters required in the models (shown in Table 3.5), such as the reaction rate constants of
•OH with the scavengers, the quantum yield of H2O2 photolysis and the quantum yield of •OH
formation by H2O2 photolysis, were obtained from the published literature. However, because of
the discrepancy in the literature (Watts and Linden, 2007; Watts et al. 2007; Jin et al., 2011;
Chan et al., 2012) for the quantum yields of HOCl and OCl– photolysis, and the quantum yields
of •OH formation by their photolysis, the parameters determined by this study were used in the
models. The •OH scavenging of total organic carbon (TOC) was also included, since
approximately 0.1 mg L–1 as carbon of TOC is present in the Milli-Q® water. Radical
scavenging by inorganic carbon and buffers was ignored, as it was calculated that they would
contribute to less than 2% of the •OH scavenging potential. Examples of Matlab® codes are
shown in Appendix C. A simpler and more approximate estimation can be made using a
Microsoft Excel spreadsheet, shown in Appendix D.
The TCE decay rate constants by UV alone, UV/chlorine, and UV/H2O2 simulated by the
models are shown in Table 3.3. Compared to the experimental results in Table 3.3, the models
52
generally give good predictions, suggesting that the reactions with their parameters described in
Table 3.5 are generally accurate.
One useful application of the model is to predict the solution pH at which the UV/chlorine
and the UV/H2O2 AOPs are equally efficient: that is, the same [•OH]ss exists in both scenarios.
Based on Equation [3.12], [•OH]ss depends on the scavenging potential in solutions. Assuming
that the TOC is the only •OH scavenger other than 0.15 mM active chlorine and the hydrogen
peroxide species, the impact of pH and TOC on the relative efficiency of the UV/chlorine AOP
versus the UV/H2O2 AOP is shown in Figure 3.5. It is evident that in pure waters, such as those
used for these experiments, with minimal background •OH scavenging, the UV/H2O2 AOP is
more efficient than the UV/chlorine AOP at all but very low pH values (i.e. less than 5.3). If the
water contains more •OH scavengers, the UV/chlorine AOP becomes more competitive at a
higher pH. For example, it is predicted that the UV/chlorine AOP will be equally efficient as the
UV/H2O2 AOP at pH values higher than 7.0 provided that the TOC is approximately 5.0 mg/L
or above.
3.3.5 Comment on Active Chlorine Reaction with •OH
An important element in modeling the UV/chlorine AOP is the reported reaction rate
constant between •OH and HOCl or OCl–, with the reaction reportedly 5 orders of magnitude
faster with OCl– than with HOCl. This was used to explain the observed reduced UV/chlorine
AOP efficiency at higher pH: the OCl– became a dominant •OH scavenger. Feng et al. (2007),
however, reported that the quantum yield of HOCl photolysis increased with increasing HOCl
concentration. This suggests a chain reaction that is initiated by the reaction between •OH and
HOCl, which would likely only occur if there were a fast reaction between •OH and HOCl.
Feng et al. (2007) also reported that the quantum yield of OCl– photolysis was not a function of
OCl– concentration, implying a possible slow reaction between •OH and OCl–. These results
contradict the earlier research. It is therefore clear that more fundamental research is needed to
clarify the mechanism of active chlorine photolysis. Appendix A shows a preliminary trial using
pulse radiolysis analysis to determine the rate constant between HOCl/OCl– and •OH. The
results successfully verified the rate constant of OCl– with •OH reported by Buxton and Subhani
(1972a), but cannot be used to determine the rate constant between HOCl and •OH.
53
Table 3.5 Reaction mechanisms of TCE decay by UV alone, the UV/chlorine and the
UV/H2O2 AOPs
No. Reaction Rate constant Reference
UV alone
3.15 TCE + hν → ClHC=C•Cl + Cl• ΦTCE,1 = 0.13 Li et al. (2004)
3.16 TCE + hν → ClHC(OH)CHCl2 ΦTCE,2 = 0.1 Li et al. (2004)
3.17 TCE + hν → HC≡CCl + Cl2 ΦTCE,3 = 0.032 Li et al. (2004)
3.18 TCE + hν → ClC≡CCl + HCl ΦTCE,4 = 0.092 Li et al. (2004)
3.19 TCE + Cl• → Cl2HC-C•Cl2 4.88 × 1010 M–1 s–1 Li et al. (2004)
UV/chlorine AOP Reactions [3.15] – [3.19], and 3.20 OCl– + H2O ↔ HOCl + OH– kforward = 1.8 × 103 s–1
kreverse = 3 × 109 M–1s–1 Fogelman et al. (1989)
3.1 HOCl + hv → •OH + •Cl ΦHOCla = 1.06
Φ•OHb = 0.79
This work
3.21 OCl– + hν → •OH + other products ΦOCl–c = 0.89 Φ•OH
b = 1.18 This work
3.4 •OH + HOCl → H2O + ClO• 8.46 × 104 M–1 s–1 Watts and Linden (2007)
3.5 •OH + OCl– → ClO• + OH– 9.0 × 109 M–1 s–1 Buxton and Subhani (1972a)
3.22 TOC + •OH → products 3 × 108 M–1 s–1 Westerhoff et al. (1999)
3.23 TCE + •OH → ClCH(OH)-C•Cl2 2.4 × 109 M–1 s–1 Li et al. (2007)
UV/H2O2 AOP Reactions [3.15] – [3.19], [3.22] – [3.23], and 3.6 H2O2 ↔ H+ + HO2
– kforward = 0.126 s–1
kreverse = 5 × 1010 M–1s–1 Song (1996)
3.24 H2O2 + hv → 2•OH ΦH2O2d = 1.0
Φ•OHb = 1.11
Stefan et al. (1996) Goldstein et al. (2007)
3.13 H2O2 + •OH → HO2• + H2O 2.7 × 107 M–1 s–1 Stefan et al. (1996)
3.14 HO2– + •OH → •O2
– + H2O 7.5 × 109 M–1 s–1 Stefan et al. (1996)
aΦHOCl is the quantum yield of HOCl photolysis bΦ•OH is the quantum yield of •OH formation cΦOCl
– is the quantum yield of OCl– photolysis dΦH2O2 is the quantum yield of H2O2 photolysis
54
Figure 3.5 Solution pH at which the UV/chlorine and the UV/H2O2 AOPs are equally
efficient as a function of TOC concentration
3.4 Conclusions TCE can be eliminated by direct UV photolysis; however, the decay rate increases when
using UV as part of an advanced oxidation process, since TCE is rapidly oxidized by •OH.
According to our experiments, the UV/chlorine AOP is more efficient than the UV/H2O2 AOP
at pH 5 in the model systems explored in this study (relatively free of other •OH scavengers),
but it loses efficiency with increasing pH. The numerical modeling analysis is generally
consistent with the experimental results in terms of the TCE decay rates by UV alone and the
UV/chlorine and the UV/H2O2 AOPs. Interestingly, in water containing more •OH scavengers
(i.e. more ‘realistic’ waters), the model predicts that the UV/chlorine AOP will become more
competitive relative to the UV/H2O2 AOP, with all other factors being equal, as the scavenger
concentration increases. However, this is admittedly a simplification. Factors such as the
reaction of chlorine with organic matter (with hydrogen peroxide typically being much less
reactive) would complicate such predictions. More research is required.
The authors also wish to emphasize that very little work has been done to assess potential
chlorination by-product formation in the UV/chlorine AOP. Some limited research suggests that
approximately 17–30% (by mass) of chlorine that is photolyzed by UV might yield chlorate
(Buxton and Subhani, 1972b; Feng et al., 2010). The formation of organochlorine species under
5.0
5.5
6.0
6.5
7.0
7.5
0 1 2 3 4 5
TOC concentration (mg L–1)
Equa
lly e
ffici
ent p
H s
55
the unique conditions of the UV/chlorine AOP system (i.e., high chlorine dose but very low
contact time) is also largely unexplored. Research on this issue is recommended.
Acknowledgements This work was partially funded by the NSERC Engage Grant program, and by Stantec
Consulting Ltd. The technical and logistical assistance of Leigh McDermott of Stantec
Consulting Ltd., and Tim Walton of the Region of Waterloo, is also gratefully acknowledged.
References APHA, AWWA, WEF, 2005. Standard Methods for the Examination of Water & Wastewater,
21st edition. 4500-Cl G. DPD colorimetric method, 4-67. American Public Health Association,
Washington, D.C., USA.
Baxendale, J.H., Wilson, J.A., 1957. The photolysis of hydrogen peroxide at high light
intensities. Transactions of the Faraday Society, 53, 344–356.
Bolton, J.R., Linden, K.G., 2003. Standardization of methods for fluence (UV dose)
determination in bench-scale UV experiments. Journal of Environmental Engineering, 129(3),
209–215.
Bolton, J.R., Stefan, M.I., 2002. Fundamental photochemical approach to the concepts of
fluence (UV dose) and electrical energy efficiency in photochemical degradation reactions.
Research on Chemical Intermediates, 28(7–9), 857–870.
Bolton, J.R., Stefan, M.I., Shaw, P.-S., Lykke, K.R., 2009. Determination of the quantum yield
of the ferrioxalate and KI/KIO3 actinometers and a method for the calibration of radiometer
detectors. CDROM Proceedings 5th UV World Congress, Amsterdam, The Netherlands.
Buxton, G.V., Subhani, M.S., 1972a. Radiation chemistry and photochemistry of oxychlorine
ions. Part 1.–Radiolysis of aqueous solutions of hypochlorite and chlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
947–957.
Buxton, G.V., Subhani, M.S., 1972b. Radiation chemistry and photochemistry of oxychlorine
ions. Part 2.–Photodecomposition of aqueous solutions of hypochlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
958–969.
56
Buxton, G.V., Subhani, M.S., 1972c. Radiation chemistry and photochemistry of oxychlorine
ions. Part 3.–Photodecomposition of aqueous solutions of chlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
970–977.
Chan, P.Y., Gamal El-Din, M., Bolton, J.R., 2012. A solar-driven UV/chlorine advanced
oxidation process. Water Research, 46(17), 5672–5682.
Crowell, R.A, Lian, R., Sauer Jr., M.C., Oulianov, D.A., Shkrob, I.A., 2004. Geminate
recombination of hydroxyl radicals generated in 200 nm photodisassociation of aqueous
hydrogen peroxide. Chemical Physics Letters, 383(5–6), 481–485.
Czapski, G., Bielski, B.H.J., 1963. The formation and decay of H2O3 and HO2 in electron-
irradiated aqueous solutions. The Journal of Physical Chemistry, 67(10), 2180–2184.
Deborde, M., von Gunten, U., 2008. Reactions of chlorine with inorganic and organic
compounds during water treatment–kinetics and mechanisms: a critical review. Water
Research, 42(1–2), 13–51.
Feng, Y., Smith, D.W., Bolton, J.R., 2007. Photolysis of aqueous free chlorine species (HOCl
and OCl–) with 254 nm ultraviolet light. Journal of Environmental Engineering and Science,
6(3), 277–284.
Feng, Y., Smith, D.W., Bolton, J.R., 2010. A potential new method for determination of the
fluence (UV dose) delivered in UV reactors involving the photodegradation of free chlorine.
Water Environment Research, 82(4), 328–334.
Fogelman, K.D., Walker, D.M., Margerum, D.W., 1989. Non-metal redox kinetics: hypochlorite
and hypochlorous acid reactions with sulfite. Inorganic Chemistry, 28(6), 986–993.
Goldstein, S., Aschengrau, D., Diamant, Y., Rabani, J., 2007. Photolysis of aqueous H2O2:
quantum yield and applications for polychromatic UV actinometry in photoreactions.
Environmental Science & Technology, 41(21), 7486–7490.
Goldstein, S., Rabani, J., 2008. The ferrioxalate and iodide-iodate actinometers in the UV region.
Journal of Photochemistry and Photobiology A: Chemistry, 193(1), 50–55.
Gültekin, I., Ince, N.H., 2007. Synthetic endocrine disruptors in the environment and water
remediation by advanced oxidation processes. Journal of Environmental Management, 85(4),
816–832.
57
Hunt, J.P., Taube, H., 1952. The photochemical decomposition of hydrogen peroxide. Quantum
yields, tracer and fractionation effects. Journal of the American Chemical Society, 74(23),
5999–6002.
Jin, J., Gamal El-Din, M., Bolton, J.R., 2011. Assessment of the UV/chlorine process as an
advanced oxidation process. Water Research, 45(4), 1890–1896.
Kläning, U.K., Wolff, T., 1985. Laser flash photolysis of HClO, ClO–, HBrO, and BrO– in
aqueous solution. Reaction of Cl- and Br-atoms. Berichte der Bunsen-Gesellschaft für
Physikalische Chemie, 89(3), 243–245.
Klassen, N.V., Marchington, D., McGowan H.C.E., 1994. H2O2 determination by the I3– method
and by KMnO4 titration. Analytical Chemistry, 66(18), 2921–2925.
Li, K., Stefan, M.I., Crittenden, J.C., 2004. UV photolysis of trichloroethylene: product study
and kinetic modelling. Environmental Science & Technology, 38(24), 6685–6693.
Li, K., Stefan M.I., Crittenden, J.C., 2007. Trichloroethylene degradation by UV/H2O2 advanced
oxidation process: product study and kinetic modelling. Environmental Science &
Technology, 41(5), 1696–1703.
Luňák, S., Sedlák, P., 1992. Photoinitiated reactions of hydrogen peroxide in the liquid phase.
Journal of Photochemistry and Photobiology A: Chemistry, 68(1), 1–33.
Nowell, L.H., Hoigné, J., 1992a. Photolysis of aqueous chlorine at sunlight and ultraviolet
wavelengths–I. Degradation rates. Water Research, 26(5), 593–598.
Nowell, L.H., Hoigné, J., 1992b. Photolysis of aqueous chlorine at sunlight and ultraviolet
wavelengths–II. Hydroxyl radical production. Water Research, 26(5), 599–605.
Oliver, B.G., Carey, J.H., 1977. Photochemical production of chlorinated organics in aqueous
solutions containing chlorine. Environmental Science & Technology, 11(9), 893–895.
Parsons, S., 2004. Advanced Oxidation Processes for Water and Wastewater Treatment, 1–6.
IWA Publishing, London, UK.
Peng, J., Wan, A., 1997. Measurement of Henry’s constants of high-volatility organic
compounds using a headspace autosampler. Environmental Science & Technology, 31(10),
2998–3003.
Rosenfeldt, E.J., Linden, K.G., 2004. Degradation of endocrine disrupting chemicals bisphenol
A, ethinyl estradiol, and estradiol during UV photolysis and advanced oxidation processes.
Environmental Science & Technology, 38(20), 5476–5483.
58
Rosenfeldt, E.J., Melcher, B., Linden, K.G., 2005. UV and UV/H2O2 treatment of
methylisoborneol (MIB) and geosmin in water. Journal of Water Supply: Research and
Technology–AQUA, 54(7), 423–434.
Schwarzenbach, R.P., Gschwend, P.M., Imboden, D. M., 2003. Environmental Organic
Chemistry, 2nd edition, 611–686. John Wiley & Sons, Inc., Hoboken, New Jersey, USA.
Sharpless, C.M., Linden, K.G., 2003. Experimental and model comparisons of low- and
medium-pressure Hg lamps for the direct and H2O2 assisted UV photodegradation of N-
nitrosodimethylamine in simulated drinking water. Environmental Science & Technology,
37(9), 1933–1940.
Song, R., 1996. Ozone-bromide-NOM interactions in water treatment, Ph.D. thesis. University
of Illinois at Urbana-Champaign, IL, USA.
Stefan, M.I., Bolton J.R., 2005. Fundamental approach to the fluence-based kinetic and
electrical energy efficiency parameters in photochemical degradation reactions:
polychromatic light. Journal of Environmental Engineering and Science, 4(S1), S13–S18.
Stefan, M.I., Hoy, A.R., Bolton, J.R., 1996. Kinetics and mechanism of the degradation and
mineralization of acetone in dilute aqueous solution sensitized by the UV photolysis of
hydrogen peroxide. Environmental Science & Technology, 30(7), 2382–2390.
Suty, H., De Traversay, C., Cost, M., 2004. Applications of advanced oxidation processes:
present and future. Water Science and Technology, 49(4), 227–233.
Tinsley, I.J., 2004. Chemical Concepts in Pollutant Behavior, 2nd edition, 123–148. John Wiley
& Sons, Inc. Hoboken, New Jersey, USA.
United States Environmental Protection Agency (USEPA), 2008. Method 551.1. Determination
of chlorination disinfection byproducts, chlorinated solvents, and halogenated
pesticides/herbicides in drinking water by liquid-liquid extraction and gas chromatography
with electron-capture detection, Revision, 1.0. Available from
<http://www.epa.gov/sam/pdfs/EPA-551.1.pdf> (accessed 11.09.11).
Volman, D.H., Chen, J.C., 1959. The photochemical decomposition of hydrogen peroxide in
aqueous solutions of allyl alcohol at 2537 Å. Journal of the American Chemical Society,
81(16), 4141–4144.
59
Wang, D., Walton, T., McDermott, L., Hofmann, R., 2011. The effectiveness of UV+chlorine
treatment of TCE in drinking water. Proceedings of the 2011 IOA and IUVA World Congress
& Exhibition, Paris, France.
Watts, M.J., Linden, K.G., 2007. Chlorine photolysis and subsequent OH radical production
during UV treatment of chlorinated water. Water Research, 41(13), 2871–2878.
Watts, M.J., Rosenfeldt, E.J., Linden, K.G., 2007. Comparative OH radical oxidation using UV-
Cl2 and UV-H2O2 processes. Journal of Water Supply: Research and Technology–AQUA,
56(8), 469–477.
Westerhoff, P., Aiken, G., Amy, G, Debroux, J., 1999. Relationships between the structure of
natural organic matter and its reactivity towards molecular ozone and hydroxyl radicals.
Water Research, 33(10), 2265–2276.
60
4. FULL-SCALE COMPARISON OF
ULTRAVIOLET/CHLORINE ADVANCED OXIDATION TO
ULTRAVIOLET/HYDROGEN PEROXIDE FOR TASTE AND
ODOUR CONTROL IN DRINKING WATER
TREATMENT
This chapter has been submitted for publication as follows:
Wang, D., Bolton J.R., Andrews, S.A., Hofmann, R. UV/chlorine control of drinking water taste and odour at
pilot and full-scale and the potential for caffeine as an experimental surrogate. Water Research.
Abstract Advanced oxidation processes (AOPs) can be used to destroy taste and odour-causing
compounds in drinking water. This work investigated pilot- and full-scale performance of the
ultraviolet/chlorine AOP for the destruction of geosmin, 2-methylisoborneol (MIB) and caffeine
(as a surrogate) in two different surface waters. The efficiency of UV/chlorine at pH 7.5 and 8.5
was comparable to that of UV/hydrogen peroxide (UV/H2O2) under parallel conditions, and was
superior at pH 6.5. Caffeine was found to be a suitable surrogate for geosmin and MIB, and
could be used in future research as a more economical alternative to geosmin or MIB spiking.
4.1 Introduction The previous chapter discussed UV/chlorine efficiency for TCE decay using a bench-scale
UV collimated beam apparatus. It was found that UV/chlorine is a promising process that could
be an alternative to UV/H2O2. A mathematical reaction kinetic model demonstrated similar
promising results for UV/chlorine. To ultimately implement UV/chlorine at full-scale, however,
a greater degree of confidence would be required. To this end, the performance of UV/chlorine
was assessed using pilot-scale and full-scale testing.
Geosmin and 2-methylisoborneol (MIB) can be destroyed during drinking water treatment
by advanced oxidation processes (AOPs), which generate hydroxyl radicals (•OH). The
reactions of geosmin and MIB with •OH are rapid, with rate constants of 7.8 × 109 and 5.1 × 109
M–1 s–1, respectively (Peter and von Gunten, 2007; Rosenfeldt et al., 2005). While ozone-based
61
AOPs tend to be the most common in the drinking water industry, UV-based AOPs—usually
UV/H2O2—are becoming more popular. UV/H2O2 is used at the City of Cornwall Water
Purification Plant (Ontario, Canada) for the control of seasonal taste and odour events that
usually occur in late summer (McCormick et al., 2013). UV/H2O2, while effective, can create
operational problems. Most of the applied H2O2 survives UV exposure, and the residual
therefore needs to be quenched because it will otherwise create a strong chlorine demand that
would interfere with secondary disinfection. At Cornwall, the residual H2O2 is quenched with a
stoichiometric excess of chlorine, but operationally this is challenging because of an inability to
measure or accurately predict the H2O2 residual, issues with different residual H2O2 in the
effluent from multiple UV reactors, H2O2 handling challenges, and other factors.
The UV/chlorine AOP may be an alternative to UV/H2O2 for water and wastewater
treatment. It has been investigated to treat contaminants such as para-chlorobenzoic acid,
benzoic acid, nitrobenzene, phenol, maleic acid, and trichloroethylene in lab prepared water
(Watts et al., 2007; Jin et al., 2011; Fang et al., 2014; Zhao et al., 2011; Wang et al., 2012),
emerging contaminants and taste and odour compounds in drinking water (Zhang et al., 2014;
Sichel et al., 2011; Watts et al., 2012), and naphthenic acids and fluorophore organic
compounds in oil sands wastewater (Chan et al., 2012; Shu et al., 2014). For a drinking water
treatment plant, operations could be relatively simple and cost-effective compared to UV/H2O2,
especially if the chlorine dose is selected to provide adequate photolysis for the control of taste
and odour compounds, with remaining chlorine serving as a secondary disinfectant (Watts et al.,
2012). Research on UV/chlorine, however, is still largely at the theoretical level or laboratory-
scale. In this study, the effectiveness of a medium pressure (MP) UV/chlorine AOP for taste and
odour control was investigated at the Cornwall Water Purification Plant through full-scale trials,
with a comparison of UV/chlorine to UV/H2O2 under parallel conditions. Caffeine as a potential
surrogate for geosmin and MIB was also evaluated. A pilot-scale study using caffeine
destruction in a Rayox® batch reactor was conducted to further investigate UV/chlorine
efficiency in water from a second source.
62
4.2 Material and Methods
4.2.1 Reagents and Materials
Geosmin and MIB (>95% purity) from Dalton Chemical Laboratories were dissolved
together in Milli-Q® water to make a stock solution at a concentration of approximately 80 mg
L–1 for full-scale tests. Caffeine stock solutions at concentrations of 10 g L–1 and 1 g L–1 for the
full- and pilot-scale tests, respectively, were prepared from purchased caffeine (ReagentPlus®
grade, Sigma-Aldrich). Deuterated d3-geosmin (99 atom % D, Sigma-Aldrich) and d3-caffeine
(99 atom % D, CDN Isotopes) were used as internal standards for quantification of
geosmin/MIB and caffeine, respectively. Sodium hypochlorite solution (NaOCl) (10–15 wt. %,
reagent grade, Sigma-Aldrich) and H2O2 solution (50 wt. %, Sigma-Aldrich) were used in the
pilot-scale tests. Industrial grade NaOCl solution (12.5 wt. %, NSF 60 certified, Olin Chlor
Alkali) and H2O2 solution (35 wt. %, NSF 60 certified, Arkema Inc.) were used in the Cornwall
full-scale tests. Sulphuric acid (H2SO4, 95–98%, A.C.S. grade, Sigma-Aldrich) and sodium
hydroxide (NaOH, ≥97.0%, A.C.S. grade, Sigma-Aldrich) were freshly diluted to appropriate
concentrations for pH adjustment in both full- and pilot-scale tests. Other compounds used in
experiments and sample analyses were all analytical reagent grade or higher. Milli-Q® water
was used in all experiments and analytical determinations.
4.2.2 Experimental Facilities and Procedures
Cornwall Full-Scale Tests
Full-scale experiments were carried out at the Cornwall Water Purification Plant in early
summer (May) and late summer (September, when taste and odour events typically occur), and
are referred to as the 1st and 2nd full-scale tests in the following text. One of the MP UV reactors
(Model: UVSwift 8L24, TrojanUV) was isolated for the experiments (shown in Figure 4.1). UV
power was set at maximum output to perform the UV/chlorine and UV/H2O2 AOPs, which
delivered a UV dose of 2,000 ± 150 mJ cm–2 from 200 to 400 nm for an exposure of 7.2 seconds
at a water flowrate of 100 L s–1, as estimated based on the free chlorine photolysis (Wang et al.,
2012). The water was drawn from the St. Lawrence River and had been treated by conventional
treatment (prechlorination, alum coagulation, flocculation, settling, and conventional
sand/anthracite filtration). Water quality parameters are summarized in Table 4.1.
63
Figure 4.1 Full-scale Trojan UVSwift reactor (left) and pilot-scale Rayox® reactor (right)
Table 4.1 Post-filtration water quality parameters for full- and pilot-scale tests
Test Source pH Turbidity
(NTU)
Alkalinity (mg CaCO3
L–1)
Total organic carbon (TOC)
(mg C L–1)
Nitrate (NO3
–) (mg L–1)
Spiked target chemicals
Cornwall 1st full-scale
St. Lawrence
River 7.9 0.02 92 1.5 1.2
400 ng L–1 geosmin 400 ng L–1 MIB
20 μg L–1 caffeine
Cornwall 2nd full-
scale
St. Lawrence
River 8.1 0.03 88 1.8 – Ambient geosmin
and MIB
Rayox® Pilot-scale
Lake Simcoe 7.5 0.2 123 3.5 0.67 20 μg L–1 caffeine
In the 1st full-scale test, approximately 400 ng L–1 geosmin and MIB, and 20 μg L–1
caffeine were spiked into the water flow upstream of the UV reactor, along with a chlorine dose
of 2, 6, or 10 mg L–1 as free chlorine, or an H2O2 dose of 1.0, 2.9, or 4.8 mg L–1 (equimolar
concentrations as chlorine) and pH adjustment to 6.5, 7.5, or 8.5. Additional trials in the absence
of chlorine and H2O2 were carried out to evaluate the stability of geosmin, MIB, and caffeine to
UV exposure alone. Preliminary tests (not shown) indicated that caffeine, geosmin and MIB
64
were all stable in the presence of 40 mg L–1 free chlorine or 10 mg L–1 H2O2 for at least 20
minutes, with the decay less than 0.5%.
The 2nd full-scale test of UV/chlorine at Cornwall was more limited than the first, with no
geosmin, MIB, or caffeine spiked, with the performance of UV/chlorine (only) monitored for
controlling the existing 18 ng L–1 of geosmin in the incoming water (MIB was below detection
limits). The main purpose of this second test was to monitor by-product formation (reported
elsewhere), but it was also used as another opportunity to validate the UV/chlorine performance
under more limited conditions. Chlorine doses of 2, 6, or 10 mg L–1 at pH 6.5, 7.5, and 8.5 were
applied.
Rayox® Pilot-Scale Test
A 40 L Rayox® completely-mixed batch reactor (Model: PS1-1-120, Calgon Carbon
Corporation), shown in Figure 4.1, equipped with a 1 kW MP UV lamp (Heraeus Noblelight
GmbH, Germany) was used in the pilot-scale experiments to evaluate UV/chlorine efficiency,
using spiked caffeine as a performance indicator. Water was collected post-filtration from the
Keswick Water Treatment Plant (Ontario, Canada), which draws water from Lake Simcoe.
Treatment at the Keswick plant is similar to that at Cornwall, except that no prechlorination is
used. The water contained approximately twice the total organic carbon (TOC) as the St.
Lawrence River water (Table 4.1). The experimental conditions in the Rayox® reactor were
similar to those applied in the full-scale tests. Caffeine at 20 μg L–1 was spiked into 40 L water
and treated by UV alone, UV/chlorine (2, 6, or 10 mg L–1) or UV/H2O2 (1.0, 2.9, or 4.8 mg L–1)
at pH 6.5, 7.5, or 8.5. The UV exposure time in the Rayox® reactor was 40 s, which was
predicted to deliver a UV dose (200–400 nm) of 1,820 ± 110 mJ cm– 2, using UVCalc®
software version 2B (from Bolton Photosciences Inc.). The UV dose varied slightly with
different chlorine/H2O2 doses and pH values.
4.2.3 Sample Analysis
Free chlorine was determined using a HACH® spectrophotometer (Model: DR/2500,
HACH), according to the Standard Methods 4500-Cl G (APHA et al., 2005). A Cecil UV/vis
spectrophotometer (Model: CE3055, Cecil Instruments) was used to determine H2O2
concentrations based on the triiodide method described by Klassen et al. (1994). An Agilent
8453 UV/vis photodiode array spectrophotometer (Model: G1103A, Angilent Technologies)
65
was used to determine solution absorbances in the entire UV range. Geosmin and MIB samples
spiked with 100 ng L–1 d3-geosmin were extracted and concentrated using the headspace solid
phase micro-extraction (HS-SPME) method and quantified using a Varian 3800 gas
chromatography coupled with a Varian 4000 ion-trap mass spectrometry (GC-MS) in the
electron ionization (EI) mode, according to Standard Methods 6040D (APHA et al., 2012).
Caffeine samples with 1 μg L–1 d3-caffeine added were extracted and concentrated using solid-
phase extraction (SPE) cartridges and analyzed using the same GC-MS, but in the positive ion
chemical ionization (CI) mode, based on the method described by Verenitch et al. (2006).
Method detection limits (MDLs) for geosmin, MIB, and caffeine were 2, 9, and 31 ng L–1,
respectively. Details for the analytical methods are shown in Appendix G.
4.3 Results and Discussion
4.3.2 Free Chlorine Decay
One of the potential advantages of UV/chlorine over UV/H2O2 is the predicted greater
photolysis of chlorine across the UV reactor than H2O2, minimizing the need to quench excess
oxidant. For both the full-scale and pilot-scale tests, the chlorine concentrations decreased by
approximately 40–80% across the UV reactors at doses (200 nm to 400 nm) of 1,800–2,000 mJ
cm–2 (Figure 4.2). Raw data are summarized in Appendix I. H2O2 decay in parallel tests was at
most approximately 5% (data not shown). Chlorine was observed to undergo greater photolysis
at a higher pH, with about 1.5–2 times the total decay at pH 8.5 compared to pH 6.5. This can be
explained by the higher OCl– concentration relative to HOCl at the higher pH (pKa of HOCl =
7.54 at 25 °C). OCl– absorbs MP UV light about 4.5 times more than HOCl (Wang et al., 2012),
and photolyzes at approximately the same rate as HOCl (quantum yield of 0.9, compared to 1.0
for HOCl photolysis) (Feng et al., 2007). It was also observed that chlorine photodecomposition
was faster at a lower initial dose, especially at pH 8.5. It is assumed that this is due to a higher
average UV dose at a lower chlorine concentration, because the UV dose is a function of the
chlorine concentration (chlorine blocks transmission of the UV light into the water). For
example, using UVCalc® 2B software, the UV dose for 40 s exposure in the Rayox® reactor at
an initial chlorine dose of 2 mg L–1 at pH 8.5 was predicted to be approximately 13% higher
than that at a dose of 10 mg L–1. It should be noted that the UV reactors used at the Cornwall
66
plant were an early version of an advanced oxidation system, and are likely undersized relative
to a modern installation. Similarly, the UV dose applied in the pilot-scale reactor was also
relatively “low”, to remain consistent with the Cornwall tests. As such, it is likely that a much
greater amount of chlorine photolysis would occur in a current UV-AOP reactor. The authors
have observed greater than 90% chlorine photolysis across a UV reactor used for
trichloroethylene destruction in a groundwater treatment system (Wang et al., 2011).
Figure 4.2 Percentage of free chlorine photolysis by UV exposure. Error bars represent the
values of experimental duplicates.
4.3.3 Geosmin and MIB Decay
The destruction of geosmin and MIB due to UV photolysis alone, UV/chlorine, or
UV/H2O2 for the 1st full-scale test is shown in Figure 4.3. Approximately 20% and 10% of
spiked geosmin and MIB, respectively, were destroyed by UV alone (~2,000 mJ cm–2). This is
generally consistent with Rosenfeldt et al. (2005), who reported that MP UV exposure at 2,000
mJ cm–2 led to about 35–40% destruction of geosmin and MIB. When an AOP was applied,
geosmin and MIB destruction was increased due to the additional •OH oxidation. UV/chlorine
and UV/H2O2 led to similar amounts (less than 10% in most cases) of geosmin and MIB
destruction at pH 7.5 and 8.5. At pH 6.5, however, UV/chlorine was substantially superior to
UV/H2O2, resulting in 10–25% more destruction for all applied doses.
67
Figure 4.3 Geosmin (top plot) and MIB (bottom plot) decay in the 1st full-scale test.
Error bars represent the values of experimental duplicates.
The superior UV/chlorine performance at lower pH has been reported by Watts and Linden
(2007), Watts et al. (2007), and Wang et al. (2012), and it is probably because of the conversion
of OCl– to HOCl at the lower pH. HOCl absorbs MP UV light about 2.3 times more efficiently
than H2O2 (Wang et al. 2012) and produces •OH at a similar efficiency (quantum yield of •OH
formation for HOCl: 0.85 vs. 1.11 for H2O2) (Nowell and Hoigné, J., 1992; Goldstein et al.,
2007), but also reacts with •OH (i.e. scavenges) more slowly than H2O2 (rate constant with •OH
for HOCl: 8.46 × 104 M–1 s–1 vs. 2.7 × 107 M–1 s–1 for H2O2) (Watts and Linden, 2007; Stefan et
al., 1996). This all leads to a higher efficiency of UV/HOCl than that of UV/H2O2. In contrast,
68
the much higher reaction (scavenging) rate of OCl– with •OH than either HOCl or H2O2 (rate
constant: 9.0 × 109 M–1 s–1) (Buxton and Subhani, 1972) more than offsets the benefit of OCl–′s
stronger UV absorption. In previous studies it has been proposed that UV/chlorine would be
superior to UV/H2O2 at pH <6 in pure water (i.e. laboratory grade water, containing no •OH
scavengers). It was also proposed that the pH at which UV/chlorine remained competitive
relative to UV/H2O2 would increase with increasing •OH scavenger concentration. The
Cornwall water had an •OH scavenging potential 2–17 times higher than the three previous
studies as estimated using the total organic carbon (TOC) and alkalinity concentrations reported
in Table 4.1. As such, this result tends to corroborate the theory that UV/chlorine remains
competitive with UV/H2O2 at higher pH—up to pH 8.5 in this case—with sufficient •OH
scavengers present.
Similar to UV/chlorine, a pH effect on UV/H2O2 efficiency was also observed, shown in
Figure 4.3. However, unlike UV/chlorine, whose efficiency is largely dependent on the
differential •OH scavenging of HOCl and OCl–, UV/H2O2 efficiency is probably changed by the
bicarbonate/carbonate equilibrium. With the increase of pH from 6.5 to 8.5, carbonic acid
(H2CO3) and bicarbonate (HCO3–) present in the water convert to carbonate (CO3
2–) (pKa of
H2CO3 = 6.37, pKa of HCO3– = 10.36) (Fanghänel et al, 1996; Kolthoff and Bosch, 1928). CO3
2–
is a much stronger •OH scavenger than H2CO3 and HCO3–. The rate constant of •OH with CO3
2–
is 3.9 × 108 M–1 s–1, compared to 8.5 × 106 M–1 s–1 for HCO3– and negligible for H2CO3 (Buxton
et al., 1988; Liao et al., 2001). In the presence of an alkalinity of 92 mg CaCO3 L–1 in the
Cornwall water, the H2CO3/HCO3–/CO3
2– species was calculated to increase the •OH
scavenging potential of the H2O2 solution by approximately 12%. However, this is still less than
the 40% decrease of the rates of geosmin and MIB decay due to UV/H2O2 when pH increased
from 6.5 to 8.5, as shown in Figure 4.3. This implies that there might be other •OH scavengers
present in the water that could also increase the •OH scavenging potential of the solution at a
higher pH. In addition, the effect of carbonate species was also present in the UV/chlorine
treated water. However, since OCl– is a much stronger •OH scavenger than carbonate species,
the contribution of carbonate species to the increase of the chlorine solution scavenging
potential was estimated to be minimal, compared to OCl– (theoretically less than 1%).
Since the geosmin and MIB decay in the UV/chlorine and UV/H2O2 processes consisted of
the direct photolysis by UV exposure and the oxidation by •OH, the difference in their
69
concentrations after treatment of UV and an AOP reflects the net destruction by •OH only. In
theory, the rate of decay of a compound reacting with •OH is a function of the second-order rate
constant with •OH and the concentration of •OH. The ratio of the rates of geosmin and MIB
decay by •OH in the same solution is thus directly proportional to the respective reaction rate
constants with •OH. As shown in Figure 4.3, the rate of MIB decay by •OH was observed to be
approximately 50–90% of that for geosmin under parallel conditions. This is generally
consistent with the difference in their rate constants with •OH—the rate constant for MIB is
approximately 65% of that for geosmin (Peter and von Gunten, 2007).
In the 2nd full-scale test, only UV/chlorine performance was evaluated, and only in terms of
the destruction of the 18 ng L–1 geosmin already present in the water. The trend in geosmin
destruction (Figure 4.4) was similar to the first testing campaign. UV/chlorine was observed to
be more efficient at lower pH, but substantial geosmin destruction was still accomplished at pH
7.5 and 8.5. The natural pH at Cornwall is approximately 8, and the data suggest that chlorine
doses in the order of 6 mg L-1 could reduce the incoming geosmin of 18 ng L–1 to below the
common detection threshold of 7 to 10 ng/L at that pH.
Figure 4.4 Geosmin decay in the 2nd full-scale test. Error bars represent the values of
experimental duplicates.
70
4.3.4 Caffeine Decay
The feasibility of using caffeine as a surrogate for geosmin and MIB to estimate the
efficiency of UV/chlorine for taste and odour control in future research was investigated in the
1st full-scale test. Spiking caffeine simultaneously with geosmin and MIB also gave more
confidence in the performance of UV/chlorine treatment. Caffeine was subsequently used in the
pilot-scale test to further evaluate the UV/chlorine efficiency in a second water matrix (Lake
Simcoe) that contained approximately twice the TOC as at Cornwall, and therefore a presumed
higher scavenging potential. The full-scale results (top plot of Figure 4.5) show that caffeine is
similarly photosensitive as geosmin and MIB, with a UV dose (alone) of ~2,000 mJ cm–2
leading to a 10–15% decrease, which was approximately the same as the destruction of geosmin
(20%) and MIB (10%).
Past studies have reported the rapid reaction between caffeine and •OH, with an average
rate constant of 5.0 × 109 M–1 s–1 (Kesavan and Powers, 1985; Shi et al., 1991; Devasagayam et
al., 1996; Brezová et al., 2009; and León-Carmona and Galano, 2011). The rate constant is in
the same order of magnitude as those for geosmin (7.8 × 109 M–1 s–1), and is almost identical to
MIB (5.1 × 109 M–1 s–1). This means in theory that the rate of caffeine decay by •OH should be
approximately the same as that of MIB, but 36% less than that of geosmin. The results shown in
Figure 4.5 generally reflected the theory, which illustrated that the experimental rate of caffeine
decay by •OH was approximately 90% and 70% (on average) those observed for MIB and
geosmin, respectively. After including UV photolysis, caffeine destruction by UV/chlorine and
UV/H2O2 was on average equivalent to 95% and 67% of MIB and geosmin decay, respectively.
Caffeine, therefore, is an almost perfect surrogate to estimate the MIB destruction by
UV/chlorine or UV/H2O2 treatment, and that geosmin decay can be expected to be
approximately 35% greater. Based on the caffeine destruction in the Lake Simcoe water in the
pilot-scale test (bottom plot of Figure 4.5), which was found to be similar to that in the Cornwall
water, UV/chlorine was predicted to have similar performance for geosmin and MIB destruction
in both waters.
71
Figure 4.5 Caffeine decay in the 1st full-scale (top plot) and pilot-scale (bottom plot) tests.
Error bars represent the values of experimental duplicates.
4.3.5 Electrical Energy per Order (EEO)
Electrical energy per order (EEO) (kWh m–3 order–1) is the electrical energy in kilowatt
hours (kWh) required to degrade a target contaminant by one order of magnitude (90%) in 1 m3
of water (Bolton et al., 2001). The equations for full-scale and pilot-scale EEO are given in
Equations [4.1] and [4.2]:
72
EO lg( / )i f
PEF C C
= for full-scale reactor [4.1]
EO1000lg( / )i f
PtEV C C
= for pilot-scale Rayox® batch reactor [4.2]
where P is the electrical power input (kW) into the UV reactor, F is the flowrate (m3 h–1), Ci and
Cf are the initial and final concentrations (M) of the target contaminant before and after the
AOP, t is the UV exposure time (h) in the batch reactor, and V is the volume of treated water (L)
in the batch reactor. In this study, P was 83.5 kW for the full-scale test and 1.8 kW for the pilot-
scale test; F was 360 m3 h–1 converted from 100 L s–1; t was 0.011 h converted from 40 s; and V
was 40 L. The calculated EEO values for geosmin, MIB, and caffeine are given in Table 4.2.
Table 4.2 Full- and pilot-scale EEO values (kWh m–3 order–1) for geosmin, MIB, and
caffeine removal
Treatment pH 6.5 pH 7.5 pH 8.5
Geosmin MIB Caffeine Geosmin MIB Caffeine Geosmin MIB Caffeine
1st full-scale test
UV/chlorine 2 mg L–1 0.26 0.35 0.43 0.49 0.68 0.65 0.65 1.0 0.87
UV/chlorine 6 mg L–1 0.22 0.29 0.26 0.39 0.51 0.47 0.59 0.81 0.67
UV/chlorine 10 mg L–1 0.16 0.22 0.21 0.28 0.34 0.42 0.40 0.58 0.54
UV/H2O2 1.0 mg L–1 0.47 0.66 0.94 0.60 0.90 1.2 1.0 2.2 1.9
UV/H2O2 2.9 mg L–1 0.30 0.46 0.49 0.37 0.53 0.69 0.51 0.73 0.87
UV/H2O2 4.8 mg L–1 0.23 0.31 0.36 0.29 0.34 0.41 0.36 0.52 0.52
Pilot-scale test
UV/chlorine 2 mg L–1 0.98 1.5 2.1
UV/chlorine 6 mg L–1 0.53 1.0 1.4
UV/chlorine 10 mg L–1 0.39 0.88 1.1
UV/H2O2 1.0 mg L–1 2.1 2.7 3.1
UV/H2O2 2.9 mg L–1 1.2 1.3 1.7
UV/H2O2 4.8 mg L–1 0.80 0.91 1.1
EEO reflects the relative efficiency of an AOP. Comparing the EEO values for each
compound in the full-scale test summarized in Table 4.2, it was found that EEO values for
73
UV/chlorine were lower than those for UV/H2O2 in most cases, which demonstrates the superior
efficiency of UV/chlorine. The only case where UV/chlorine was less efficient (higher EEO) than
UV/ H2O2 occurred at the highest pH (8.5) with the highest oxidant dose (chlorine: 10 mg L–1
and H2O2: 4.8 mg L–1). Similarly, in the pilot-scale test using a different water matrix, the EEO
values for UV/chlorine were all lower than or the same as those for UV/H2O2. This tends to
verify the theory proposed by Wang et al. (2012) that UV/chlorine becomes more competitive
relative to UV/H2O2 with more •OH scavengers, because the levels of two major scavengers,
TOC and alkalinity, in the Lake Simcoe water were approximately 2 and 1.3 times higher than
those of the Cornwall water.
4.3.6 Comment on Chlorine Radical (•Cl) and Disinfection By-Product (DBP) Formation
during Chlorine Photolysis
Chlorine radicals (•Cl) are known to be generated during chlorine photolysis (Oliver and
Carey, 1977; Nowell and Hoigné, J., 1992), however, their role in the destruction of a target
compound is uncertain. Nowell and Hoigné (1992) and Watts and Linden (2007) considered •Cl
effectiveness to be negligible compared to •OH for the destruction of 1-chlorobutane and para-
chlorobenzoic acid, while Fang et al. (2014) reported that •Cl contributed to the majority of
benzoic acid decay during UV/chlorine treatment. This may be due to a reported high selectivity
of •Cl to different compounds. For example, •Cl reacts with benzoic acid faster than •OH (rate
constant with •Cl: 1.8 × 1010 M–1 s–1 vs. 5.9 × 109 M–1 s–1 with •OH), while it reacts with 1-
chlorobutane more slowly than •OH (rate constant with •Cl: 1.3 × 108 M–1 s–1 vs. 3.0 × 109 M–1
s–1 with •OH) and the reaction with nitrobenzene is negligible (Fang et al., 2014, Nowell and
Hoigné, J., 1992; Bell et al., 1981). The competition between •Cl and •OH dictates the impact of
•Cl in the destruction of the compound. In this study, the role of •Cl in geosmin, MIB, and
caffeine destruction is unknown due to the lack of information in the literature for the reaction
rates between •Cl and these compounds.
The concentration of chlorine used in practice for the UV/chlorine AOP may be in the
order of 5–10 mg L–1, compared to the more traditional 0.2–2 mg L–1 used for chlorine
disinfection. This might cause an increased formation of chlorination (disinfection) by-products.
Tending to mitigate this effect, however, is the very short (i.e. seconds) chlorine contact time
during the AOP process during which much of the higher chlorine dose undergoes photolysis.
74
Nevertheless, the formation of by-products during UV/chlorine treatment is an issue that
warrants careful study, but is beyond the scope of this paper.
4.4 Conclusions The UV/chlorine AOP was found to be equally or more efficient than the UV/H2O2 AOP in
the destruction of geosmin, MIB, and caffeine for similar oxidant (molar) doses, except at the
highest pH (8.5) and the highest oxidant dose in the Cornwall water. Caffeine was found to be
an appropriate surrogate for geosmin and MIB. These results are likely to be water-specific, but
were generally consistent for the two waters tested in this study, containing different amounts of
TOC and carbonate alkalinity (TOC: 1.5 vs. 3.5 mg C L–1, alkalinity: 92 vs. 123 mg CaCO3
L–1). Given the presumed operational advantages of UV/chlorine over UV/H2O2, this suggests
merit in further investigating this treatment technology. One key issue to be explored before
full-scale implementation is the potential for by-product formation.
Acknowledgements This work was funded by the Natural Sciences and Engineering Research Council of
Canada through the Industrial Research Chair program. The authors express their appreciation
to the staff at the Cornwall Water Purification Plant, including Owen O'Keefe, Daniel Drouin,
and Morris McCormick, for their help with the full-scale experiments. The help of Zhen (Jim)
Wang, Jiafan (Kevin) Yang, Hong Zhang and A.H.M. Anwar Sadmani is also gratefully
acknowledged.
References APHA, AWWA, WEF, 2005. Standard Methods for the Examination of Water & Wastewater,
21st edition. American Public Health Association, American Water Works Association, and
Water Environment Federation, Washington, D.C., USA.
APHA, AWWA, WEF, 2012. Standard Methods for the Examination of Water & Wastewater,
22nd edition. American Public Health Association, American Water Works Association, and
Water Environment Federation, Washington, D.C., USA.
75
Bell, T.N., Perkins, K.A., Perkins, P.G., 1981. Chlorination of paraffin hydrocarbons. 3.
chlorination of n-butane, 1-chlorobutane, and 2-chlorobutane. The Journal of Physical
Chemistry, 82(2), 160–165.
Bolton, J.R., Bircher, K.G., Tumas, W., Tolman, C.A., 2001. Figures-of-merit for the technical
development and application of advanced oxidation technologies for both electric- and solar-
driven systems. Pure and Applied Chemistry, 73(4), 627–637.
Brezová, V., Šlebodová, A., Staško, A., 2009. Coffee as a source of antioxidants: an EPR study.
Food Chemistry, 114(3), 859–868.
Buxton, G.V., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988. Critical review of rate
constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals (•OH/•O–)
in aqueous solution. Journal of Physical and Chemical Reference Data, 17(2), 513–886.
Buxton, G.V., Subhani, M.S., 1972. Radiation chemistry and photochemistry of oxychlorine
ions. Part 1.–Radiolysis of aqueous solutions of hypochlorite and chlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
947–957.
Chan, P.Y., Gamal El-Din, M., Bolton, J.R., 2012. A solar-driven UV/chlorine advanced
oxidation process. Water Research, 46(17), 5672–5682.
Devasagayam, T.P.A., Kamat, J.P., Mohan, H., Kesavan, P.C., 1996. Caffeine as an antioxidant:
inhibition of lipid peroxidation induced by reactive oxygen species. Biochimica et Biophysica
Acta, 1282(1), 63–70.
Fanghänel, T., Neck, V., Kim, J.I., 1996. The ion product of H2O, dissociation constants of
H2CO3 and Pitzer parameters in the system Na+/H+/OH–/HCO3–/CO3
2–/ClO4–/H2O at 25°C.
Journal of Solution Chemistry, 25(4), 327–343.
Fang, J., Fu, Y., Shang, C., 2014. The roles of reactive species in micropollutant degradation in
the UV/free chlorine system. Environmental Science & Technology, 48(3), 1859–1868.
Feng, Y., Smith, D.W., Bolton, J.R., 2007. Photolysis of aqueous free chlorine species (HOCl
and OCl–) with 254 nm ultraviolet light. Journal of Environmental Engineering and Science,
6(3), 277–284.
Jin, J., Gamal El-Din, M., Bolton, J.R., 2011. Assessment of the UV/chlorine process as an
advanced oxidation process. Water Research, 45(4), 1890–1896.
76
Kesavan, P.C., Powers, E.L., 1985. Differential modification of oxic and anoxic components of
radiation damage in Bacillus megaterium spores by caffeine. International Journal of
Radiation Biology, 48(2), 223–233.
Klassen, N.V., Marchington, D., McGowan H.C.E., 1994. H2O2 determination by the I3– method
and by KMnO4 titration. Analytical Chemistry, 66(18), 2921–2925.
Kolthoff, I.M., and Bosch, W., 1928. The influence of neutral salts on acid-salt equilibria: III.
The second dissociation constant of carbonic acid and the influence of salts on the activity of
the hydrogen ions in a bicarbonate-carbonate mixture. Recueil des Travaux Chimiques des
Pays-Bas, 47(10), 819–825.
León-Carmona, J.R., Galano, A., 2011. Is caffeine a good scavenger of oxygenated free radicals?
The Journal of Physical Chemistry B, 115(15), 4538–4546.
Liao, C.-H., Kang, S.-F., Wu, F.-A., 2001. Hydroxyl radical scavenging role of chloride and
bicarbonate ions in the H2O2/UV process. Chemosphere, 44(5), 1193–1200.
McCormick, M., O’Keefe, O., Drouin, D., 2013. The transition from GAC to UV/H2O2: taste &
odour control at the Cornwall Water Purification Plant. In: 2013 OWWA Treatment
Committee’s Spring Seminar, Vaughan, Ontario, Canada.
Nowell, L.H., Hoigné, J., 1992. Photolysis of aqueous chlorine at sunlight and ultraviolet
wavelengths–II. Hydroxyl radical production. Water Research, 26(5), 599–605.
Oliver, B.G., Carey, J.H., 1977. Photochemical production of chlorinated organic in aqueous
solutions containing chlorine. Environmental Science & Technology, 11(9), 893–895.
Peter, A., von Gunten, U., 2007. Oxidation kinetics of selected taste and odor compounds during
ozonation of drinking water. Environmental Science & Technology, 41(2), 626–631.
Rodenfeldt, E.J., Melcher, B., Linden, K.G., 2005. UV and UV/H2O2 treatment of
methylisoborneol (MIB) and geosmin in water. Journal of Water Supply: Research and
Technology– AQUA, 54(7), 423–434.
Shi, X., Dalal, N.S., Jain, A.C., 1991. Antioxidant behavious of caffeine: efficient scavenging of
hydroxul radicals. Food and Chemical Toxicology, 29(1), 1–6.
Shu, Z., Li, C., Belosevic, M., Bolton, J.R., Gamal El-Din, M., 2014. Application of a solar
UV/chlorine advanced oxidation process to oil sands process-affected water remediation.
Environmental Science & Technology, 48(16), 9692–9701.
77
Sichel., C., Garcia, C., Andre, K., 2011. Feasibility studies: UV/chlorine advanced oxidation
treatment for the removal of emerging contaminants. Water Research, 45(19), 6371–6380.
Verenitch, S.S., Lowe, C.J., Mazumder, A., 2006. Determination of acidic drugs and caffeine in
municipal wastewaters and receiving waters by gas chromatography-ion trap tandem mass
spectrometry. Journal of Chromatography A, 1116(1–2), 193–203.
Wang, D., Bolton J.R., Hofmann, R., 2012. Medium pressure UV combined with chlorine
advanced oxidation for trichloroethylene destruction in a model water. Water Research,
46(15), 4677–4686.
Wang, D., Walton, T., McDermott, L., Hofmann, R., 2011. Control of TCE using UV combined
with hydrogen peroxide or chlorine. Proceedings of IOA-IUVA 2011 World Congress &
Exhibition, Paris, France.
Watts, M.J., Hofmann, R., Rosenfeldt, E., 2012. Low-pressure UV/Cl2 for advanced oxidation
of taste and odor. Journal of American Water Works Association, 104(1), E58–E65.
Watts, M.J., Linden, K.G., 2007. Chlorine photolysis and subsequent OH radical production
during UV treatment of chlorinated water. Water Research, 41(13), 2871–2878.
Watts, M.J., Rosenfeldt, E.J., Linden, K.G., 2007. Comparative OH radical oxidation using UV-
Cl2 and UV-H2O2 processes. Journal of Water Supply: Research and Technology–AQUA,
56(8), 469–477.
Zhang, X., Li, W., Gong, X., Fan, W., Ren, P., 2014. Optimization of the UV/chlorine process
for ammonia removal and disinfection by-products reduction. Desalination and Water
Treatment, 52, published online.
Zhao, Q., Shang, C., Zhang, X., Ding, G., Yang, X., 2011. Formation of halogenated organic
byproducts during medium-pressure UV and chlorine coexposure of model compounds,
NOM and bromide. Water Research, 45(19), 6545–6554.
78
5. FORMATION OF DISINFECTION BY-PRODUCTS IN THE
ULTRAVIOLET/CHLORINE ADVANCED OXIDATION
PROCESS
This chapter has been submitted for publication as follows:
Wang, D., Bolton J.R., Andrews, S.A., Hofmann, R. Formation of disinfection by-products in the
ultraviolet/chlorine advanced oxidation process. Science of the Total Environment.
Abstract Disinfection by-product (DBP) formation may be a concern when applying ultraviolet light
and free chlorine (UV/chlorine) as an advanced oxidation process (AOP) for drinking water
treatment because of the typically large chlorine doses (e.g. 5–10 mg L–1 as free chlorine)
relative to normal chlorination. A potential mitigating factor is the low chlorine contact times
for AOP treatment (e.g. seconds). Full-scale and pilot-scale test results showed minimal
trihalomethane (THM) and haloacetic acid (HAA) formation during UV/chlorine treatment,
while dichloroacetonitrile (DCAN) and bromochloroacetonitrile (BCAN) were produced
rapidly. Adsorbable organic halide (AOX) formation was significant in application of the
UV/chlorine process in water that had not been previously chlorinated, while little additional
formation was observed in prechlorinated water. Chlorine photolysis led to chlorate and bromate
formation, equivalent to approximately 2–17% and 0.01–0.05% of the photolyzed chlorine,
respectively. No perchlorate or chlorite formation was observed. During simulated secondary
disinfection of the AOP-treated water, the DBP formation potential for THMs, HAAs, HANs,
and AOX was observed to increase approximately to the same extent as was observed for
pretreatment using UV combined with hydrogen peroxide (UV/H2O2) AOP.
5.1 Introduction In Chapters 3 and 4, it was demonstrated that the UV/chlorine AOP has comparable
efficiency to UV/H2O2 for the destruction of organic contaminants in water treatment under
certain conditions. One of the concerns associated with the UV/chlorine treatment is the need
for relatively high chlorine doses relative to conventional drinking water chlorination (e.g. 5–10
79
mg L–1 vs. 0.2–2 mg L–1 as free chlorine for disinfection) (Wang et al., 2012; Watts et al., 2007).
This might tend to promote the formation of chlorinated by-products (conventionally referred to
as disinfection by-products, DBPs). A possible mitigating factor is that chlorine contact times
during UV/chlorine treatment might be in the order of seconds because of the rapid and almost
complete chlorine photolysis expected in the UV reactor (Wang et al., 2011), in contrast to the
hours of contact time that are normally associated with DBP formation in chlorine disinfection.
The formation of DBPs associated with chlorination has been widely studied, but little
information is available on the unique high dose/short contact time conditions that would occur
during UV/chlorine treatment. UV/chlorine photolysis also leads to the formation of the chlorine
radical (•Cl), which may react with natural organic matter (NOM) to form chlorinated DBPs.
This has not been experimentally confirmed, although a study reported by Fang et al. (2014)
indicated that •Cl played an important role in the destruction of benzoic acid in a UV/chlorine
AOP under laboratory conditions. Chlorine photolysis may also directly produce inorganic
DBPs of possible health concern, such as perchlorate (ClO4–), chlorate (ClO3
–), and chlorite
(ClO2–) (Buxton and Subhani, 1972; Feng et al., 2010; Kang et al., 2006), as well as bromate
(BrO3–) in the presence of bromide (Br–) (von Gunten and Hoigné, 1994).
The potential impact of UV/chlorine treatment on DBPs includes not only DBPs that may
form during the treatment, but also the possible effect on organic precursors that may
subsequently react with chlorine on secondary chlorination. Previous studies on the impact of
UV/H2O2 pretreatment on subsequent THM and HAA formation have suggested that such an
effect is AOP-dose specific, with low and moderate AOP doses (e.g. UV in the order of 1,000
mJ cm–2) leading to enhanced THM/HAA formation on subsequent chlorination (Dotson et al.,
2010; Kleiser and Frimmel, 2000), while THM and HAA formation following higher AOP
doses (UV of 3,500–5,000 mJ cm–2) have been observed to be decreased (Toor and Mohseni,
2007; Liu et al., 2002).
There have been only a limited number of studies to date on DBP formation following
UV/chlorine treatment, most of which have reported some formation of organic chlorinated
DBPs during the AOP and/or in subsequent secondary chlorination (Liu et al., 2006; Pisarenko
et al., 2013; Weng et al., 2012; Deng et al., 2014; Shah et al., 2011). None of these studies have
simulated the conditions that would be expected to occur in a plant, that is high free chlorine
doses (5–10 mg L–1 as free chlorine), high UV doses (>1,000 mJ cm–2), but short reaction times
80
(<1 minute). This research addresses such conditions, as well as the impact of the AOP
treatment on subsequent secondary chlorination DBP formation. The work was conducted at
full-scale using a medium pressure (MP) UV reactor with a treatment capacity of 8.6 MLD, as
well as using a pilot-scale batch MP UV reactor with a volume of 40 L. The impact of
UV/chlorine treatment on DBPs was compared to parallel treatment using the more
conventional UV/H2O2 AOP.
5.2 Material and Methods
5.2.1 Experimental Procedures
Full-Scale Experiments
Full-scale experiments were conducted at the City of Cornwall Water Purification Plant
(Ontario, Canada). The plant draws water from the St. Lawrence River, which is treated by
prechlorination, alum coagulation, flocculation, settling, conventional sand/anthracite filtration,
and UV disinfection. The water quality after the post-filtration stage is summarized in Table 5.1.
Following filtration, Trojan MP UVSwift 8L24 UV reactors are used for primary disinfection as
well as for UV/H2O2 advanced oxidation during periods of taste and odour problems, which
historically occur in late summer. One UV unit was isolated for this study, and operated with its
100 L s–1 flow going to waste (shown in Figure 5.1).
Table 5.1 Full- and pilot-scale post-filtration water quality parameters
Test Cornwall - early April Cornwall - late May Pilot-scale
Source St. Lawrence River St. Lawrence River Lake Simcoe
Temperature (°C) 3 12 7–11
pH 8.1 7.9 7.5
Turbidity (NTU) 0.04 0.02 0.2 Alkalinity (mg CaCO3 L–1) 85 92 123
TOC (mg C L–1) 1.6 1.5 3.5 Absorbance at 254 nm (cm–1) 0.02 0.02 0.04
Chlorine residual (mg L–1 as free chlorine)
0.3 0.1 0
81
Figure 5.1 Full-scale Trojan UVSwift reactor (left) and pilot-scale Rayox® reactor (right)
The full-scale study involved treating the flow with chlorine alone at three doses (2, 6, and
10 mg L–1 as free chlorine), UV alone, UV/chlorine (2, 6, and 10 mg L–1), UV/H2O2 (1.0, 2.9,
and 4.8 mg L–1 H2O2, the same molar concentrations as chlorine), and at three pH levels (6.5,
7.5, and 8.5). Sodium hypochlorite (NaOCl) stock solution (12.5 wt. %, NSF 60 certified, Olin
Chlor Alkali) or H2O2 stock solution (35 wt. %, NSF 60 certified, Arkema Inc.), provided by the
Cornwall plant, was injected between the filters and the UV reactors. The travel time between
chemical injection immediately downstream of the filter and the UV reactor effluent was
determined by tracer tests to be approximately 30 s at the flowrate of 100 L s–1. The UV dose
from 200 to 400 nm for all treatment conditions was estimated to be approximately 1,800 ± 100
mJ cm–2, based on the method of Wang et al. (2012). DBP formation by UV/chlorine was
compared to that by UV/H2O2 under parallel conditions. pH adjustment was achieved by the
addition of 10% (w/w) sulphuric acid (H2SO4) or 1.5% (w/w) sodium hydroxide (NaOH)
through injection ports beside the chlorine or H2O2 injection port. Once the desired pH value
was achieved, it was found to be very stable in all trials.
The DBPs that were monitored are listed in Table 5.2. DBP samples were collected
downstream of the UV reactor with simultaneous quenching of residual free chlorine (i) using
200 mg L–1 sodium sulphite for THMs (Farré et al., 2011), HAAs (Plummer and Edzwald,
82
2001), and AOX (followed by acidification to pH <2) (Crebelli et al., 2005), (ii) using 50 mg
L–1 H2O2 for HANs, HKs, and CP (Shams El Din and Mohammed, 1998) with addition of
phosphate buffer to pH 4.8–5.5 (USEPA, 1995), and (iii) using 50 mg L–1 ethylenediamine for
inorganic DBPs (USEPA, 1997). In addition to the measurement of these DBPs immediately
after the UV reactor representing 30 s of chlorine contact time, the water downstream of the
reactor that had been treated with UV alone and UV/chlorine at the maximum chlorine dose (i.e.
10 mg L–1) was dosed with chlorine to reach 6.5 mg L–1 as free chlorine (an arbitrarily ‘high’
amount), and then subjected to DBP formation potential (DBP-FP) tests for 24 h at room
temperature, according to a modified uniform formation condition test (Summers et al., 1996).
DBP-FP following UV/H2O2 treatment at 4.8 mg L–1 H2O2 was also tested, but 0.2 mg L–1
catalase from bovine liver (powder, 2,000–5,000 units/mg protein, Sigma-Aldrich) was added to
samples to quench the residual H2O2 before dosing with 6.5 mg L–1 chlorine (Liu et al., 2003).
The purpose of these formation potential tests was to observe the impacts of the UV,
UV/chlorine, and UV/H2O2 pretreatments on subsequent chlorination DBP formation.
Pilot-Scale Experiments
Pilot-scale experiments were carried out in a 40 L Rayox® completely-mixed batch reactor
(Model: PS1-1-120, Calgon Carbon Corporation) equipped with a 1 kW MP lamp (Heraeus
Noblelight GmbH, Germany) (shown in Figure 5.1) to simulate the same UV and chemical
oxidant doses as applied at Cornwall, and the same pH conditions. In this test, however, water
was collected post-filter from the Keswick Water Treatment Plant (Ontario, Canada). This plant
treats water from Lake Simcoe, a water with approximately twice the total organic carbon
(TOC) as the St. Lawrence River (Table 5.1), by conventional treatment processes similar to
those used at Cornwall except that prechlorination is not employed. All experimental and
analytical methods were similar to those used at Cornwall, except for tests of DBP formation by
chlorine alone, which were conducted by chlorinating 500 mL water for 60 s in a beaker with
the same chlorine doses and pH values as those used in the Rayox® reactor. The UV exposure
time in the Rayox® reactor was 40 s, which was found to result in approximately the same
percentage of chlorine decay as was observed at Cornwall. UVCalc® software version 2B (from
Bolton Photosciences Inc.) predicted that the UV dose for 40 s in the Rayox reactor was 1,820 ±
110 mJ cm– 2 (200–400 nm), varying slightly with different chlorine or H2O2 doses and pH
values. DBP-FPs using an initial chlorine concentration of 6.5 mg L–1 were tested after the
83
treatment of UV alone, UV/chlorine at 10 mg L–1, and UV/H2O2 at 4.8 mg L–1, as well as with
no treatment, at the same three pH levels (6.5, 7.5, and 8.5) employed in the full-scale tests at
Cornwall.
Table 5.2 Monitored organic and inorganic DBPs
Group Components Trihalomethanes (THMs) Chloroform/trichloromethane (TCM) Bromodichloromethane (BDCM) Chlorodibromomethane (CDBM) Bromoform/tribromomethane (TBM) Haloacetic acids (HAAs) Monochloroacetic acid (MCAA) Dichloroacetic acid (DCAA) Trichloroacetic acid (TCAA) Monobromoacetic acid (MBAA) Dibromoacetic acid (DBAA) Bromochloroacetic acid (BCAA) Bromodichloroacetic acid (BDCAA) Chlorodibromoacetic acid (CDBAA) Tribromoacetic acid (TBAA) Haloacetonitriles (HANs) Bromochloroacetonitrile (BCAN) Dibromoacetonitrile (DBAN) Dichloroacetonitrile (DCAN) Trichloroacetonitrile (TCAN) Haloketones (HKs) 1,1-Dichloro-2-propanone (DCP) 1,1,1-Trichloro-2-propanone (TCP) Chloropicrin (CP) Adsorbable organic halides (AOX) Inorganic DBPs Chlorite (ClO2
–) Chlorate (ClO3
–) Perchlorate (ClO4
–) Bromate (BrO3
–)
5.2.2 Analytical Methods
Free chlorine was measured using a HACH® spectrophotometer (Model: DR/2500, HACH)
and the DPD method (APHA et al., 2005). A Cecil UV/vis spectrophotometer (Model: CE3055,
Cecil Instruments) was used to determine H2O2 concentrations based on the triiodide method
(Klassen et al., 1994). THMs, HANs, HKs, and CP were extracted and analyzed using gas
chromatography-electron capture detection (GC-ECD, Model: HP 5890 Series II, Hewlett-
Packard), according to USEPA Method 551.1 (USEPA, 1995). HAAs samples were extracted
and methylated using APHA Method 6251 B (APHA et al., 2005) and analyzed using the same
84
GC-ECD equipment. AOX was determined using an AOX analyzer (Model: Xplorer, TE
Instruments) based on APHA Method 5320 (APHA et al., 2005). ClO3– samples at expected
concentrations higher than 50 μg L–1 were analyzed using an ion chromatograph (IC, Model:
Dionex ICS-5000+ analytical system, Thermo Scientific) based on USEPA Method 300.1
(USEPA, 1997). ClO2–, ClO4
–, and BrO3–, as well as ClO3
– at concentrations lower than 50 μg
L–1 were analyzed by the Ministry of Environment and Climate Change of Ontario, Canada,
using an ion chromatograph tandem mass spectrometer (IC-MS/MS, Model: Dionex 2500,
Thermo Scientific) following the method described by Furdui and Tomassini (2010). The
method detection limits (MDLs) for THMs, HAAs, HANs, HKs, and CP ranged from 0.2 to 1.2
μg L–1. The MDL for ClO3– using IC was 7 μg L–1, while the MDLs of ClO2
–, ClO3–, ClO4
–, and
BrO3– using IC-MS/MS ranged from 0.03 to 0.05 μg L–1. Details for the analytical methods are
shown in Appendix G.
5.3 Results The data are presented in the same format for all classes of organic DBPs, so Figure 5.2
(THMs) will be explained in detail to assist the reader. The figure is divided into two principal
sections: short-term THM formation during the 30–60 s of travel time across the UV reactor (the
left side), and THM formation potential (THM-FP) following UV/AOP after being dosed with
6.5 mg L–1 chlorine for 24 h (the right side). The treatment conditions (from left to right) include
reporting any initial DBPs prior to UV/AOP treatment (from prechlorinated water, for example),
the application of chlorine alone in the absence of UV, the application of UV alone at a dose of
approximately 1,800 mJ cm– 2, UV/chlorine treatment at three chlorine concentrations, and then
UV/H2O2 at the same three molar H2O2 concentrations as for chlorine in the UV/chlorine tests.
The right section reports the formation potentials in the water without any pretreatment as a
control, followed by those in the water pretreated by UV, UV/chlorine, and UV/H2O2 at the
highest chemical oxidant dose (i.e. UV with 10 mg L–1 chlorine or UV with 4.8 mg L–1 H2O2).
Each treatment scenario was repeated at three pH values, as shown by the triplets of shaded
bars. Results from all waters tested are shown in the same figure to facilitate comparisons. All
DBPs were investigated in the full-scale test carried out in early April at Cornwall and the pilot-
scale test. HAAs, AOX, and inorganic DBPs were also investigated in the other full-scale test
carried out in late May. All raw data are summarized in Appendix I.
85
Figure 5.2 THM formation in full- and pilot-scale tests. Plots on the left show THMs after various treatment processes for
short reaction time (30–60 s contact time). Plots on the right show THM formation potentials (free chlorine dose: 6.5 mg L–1
for 24 h) in the water pretreated by selected processes shown on the x-axis. Error bars represent the values of the duplicates
measured.
.
THM
s (μ
g L–1
) TH
M-F
P (μ
g L–1
)
pH 6.5 pH 7.5 pH 8.5
Lake Simcoe (pilot-scale)
St. Lawrence River (full-scale test in April)
86
HA
As
(μg
L–1)
HA
A-F
P (μ
g L–1
)
pH 6.5 pH 7.5 pH 8.5
Figure 5.3 HAA formation in full- and pilot-scale tests. Plots on the left show HAAs after various treatment processes for short
reaction time (30–60 s contact time). Plots on the right show HAA formation potentials (free chlorine dose: 6.5 mg L–1 for 24 h) in
the water pretreated by selected processes shown on the x-axis. Error bars represent the values of the duplicates measured.
St. Lawrence River (full-scale test in April)
St. Lawrence River (full-scale test in May)
Lake Simcoe (pilot-scale)
87
HA
Ns
(μg
L–1)
HA
N-F
P (μ
g L–1
)
pH 6.5 pH 7.5 pH 8.5
Figure 5.4 HAN formation in full- and pilot-scale tests. Plots on the left show HANs after various treatment processes for
short reaction time (30–60 s contact time). Plots on the right show HAN formation potentials (free chlorine dose: 6.5 mg L–1
for 24 h) in the water pretreated by selected processes shown on the x-axis. Error bars represent the values of the duplicates
measured.
Lake Simcoe (pilot-scale)
St. Lawrence River (full-scale test in April)
88
AO
X (μ
g C
l L–1
) AO
X-FP
(μg
Cl L
–1)
pH 6.5 pH 7.5 pH 8.5
Figure 5.5 AOX formation in full- and pilot-scale tests. Plots on the left show AOX after various treatment processes for short
reaction time (30–60 s contact time). Plots on the right show AOX formation potentials (free chlorine dose: 6.5 mg L–1 for 24 h) in
the water pretreated by selected processes shown on the x-axis. Error bars represent the values of the duplicates measured.
St. Lawrence River (full-scale test in April)
St. Lawrence River (full-scale test in May)
Lake Simcoe (pilot-scale)
89
5.3.1 THMs
There was no observed short-term THM formation across the UV reactor (30 s contact
time) during the full-scale UV/chlorine or chlorine alone trial at Cornwall, with reactor effluent
concentrations similar to the influent concentrations of about 18 µg L–1 (Figure 5.2). For the
Lake Simcoe pilot system, there was a small formation by chlorine alone of approximately 8 µg
L–1 THMs during the 60 s of contact time at the highest 10 mg L–1 chlorine dose, but the
presence of UV had no observable impact. THMs consisted primarily of TCM (~50%) and
BDCM (~40%) due to the low Br– in these waters (2–3 µg L–1). It is unknown whether the more
brominated species would form more quickly during these short reaction times.
UV and both AOP pretreatments did appear to have a significant impact on THM-FP. UV
alone at a dose of approximately 1,800 mJ cm –2 increased the 24 h THM formation by 20–30%
for both Cornwall and Lake Simcoe waters compared to a non-pretreated control, suggesting
that UV exposure at that dose can create THM precursors. Under AOP conditions, THM-FP
increased even more by 30–110% compared to the controls. Both UV/chlorine and UV/H2O2
generally had the same effect, except at pH 6.5 where UV/chlorine led to more THMs than
UV/H2O2 (increased by 90–110%, compared to the controls). Previous modelling by the authors
suggests that UV/chlorine at pH 6.5 is more effective at producing •OH than UV/H2O2 in a
similar water matrix at that pH (Wang et al., 2012).
5.3.2 HAAs
The predominant HAAs were DCAA and TCAA, contributing approximately 60% of the
total detected HAAs. As shown in Figure 5.3, only a small amount of HAAs was formed by
chlorine alone during the 30–60 s of contact time in either of the full- or pilot-scale water.
However, the UV/chlorine AOP led to HAA formation of up to 13 µg L–1 at the highest chlorine
dose (10 mg L–1) at pH 6.5, with the result more pronounced for Lake Simcoe water than for
Cornwall water (which had been prechlorinated).
Pretreatment by UV alone was observed to increase the 24 h HAA formation potential by
10–25% for Cornwall water compared to a non-pretreated control, but led to a decrease in HAA-
FP by 10–30% for Lake Simcoe water. Similar to THMs, pretreatment by both UV/chlorine and
UV/H2O2 AOPs resulted in a higher 24 h HAA formation when compared to the controls (by
90
40–110% in Cornwall water and 20–90% in Lake Simcoe water). As with the THMs, the effect
was associated with the higher anticipated AOP efficiency, with UV/chlorine at pH 6.5 leading
to the greatest increase in HAA-FP in all waters.
5.3.3 HANs, HKs, and CP
No short-term HK or CP formation was observed across the UV reactor under any of the
conditions. Of the four HANs monitored, only DCAN and BCAN were detected following the
30–60 s contact time, with significantly more formed during UV/chlorine treatment (0.4–2.9 µg
L–1 in Cornwall water and 1.6–5.2 µg L–1 in Lake Simcoe water) than when using chlorine alone
(0–0.3 µg L–1 in Cornwall water and 0.6–0.8 µg L–1 in Lake Simcoe water), as shown in Figure
5.4. The total concentrations of DCAN and BCAN after 30–60 s UV/chlorine treatment could
reach more than 50% and 100% of the 24 h HAN formation potentials in the non-pretreated
controls for Lake Simcoe and Cornwall waters, respectively, although overall concentrations
were still quite low at below 6 µg L–1. The highest concentrations were observed at pH 6.5
relative to the other pH values, which is consistent with a possible role of radicals in their
formation during UV/chlorine treatment, since the higher formation coincided with greater
radical generation at lower pH. This effect, however, may be confounded with other evidence
that HAN formation is promoted at lower pH even in the absence of AOP treatment (Yang et al.,
2007; Glezer et al., 1999; Hansen et al., 2012).
UV pretreatment increased the 24 h HAN formation by 80–100% in Cornwall water and
20–30% in Lake Simcoe water, compared to a non-pretreated control. The formation was
enhanced more significantly by UV/chlorine and UV/H2O2 AOPs, which led to increases of
110–260% and 50–220% in Cornwall and Lake Simcoe waters, respectively. Similar to THMs
and HAAs, the highest formation occurred with UV/chlorine pretreatment at pH 6.5.
5.3.4 AOX
AOX (unit: µg Cl L–1) is a collective parameter to indicate the total halogenated DBPs that
can be adsorbed by activated carbon. Prechlorination at the Cornwall plant resulted in AOX
concentrations in the order of 40 µg Cl L–1 to 80 µg Cl L–1 (April and May, respectively) in the
post-filtration water, as shown in Figure 5.5. Subsequent addition of up to 10 mg L–1 free
chlorine for AOP treatment led to negligible increases in AOX during the 30 s travel across the
91
UV reactor. The situation was quite different for the Lake Simcoe water treated at pilot scale.
This water had not been prechlorinated, so presumably the organic matter would be expected to
be more reactive to chlorine. This was not readily apparent when chlorine alone was added, with
only a small AOX formation of approximately 15 µg Cl L–1 measured during the 60 s of
chlorine contact time. However, when UV/chlorine was applied, AOX formation was increased
in most cases to approximately 20 to 70 µg Cl L–1. Previous research suggests that advanced
oxidation can increase the THM, HAA, and HAN formation potentials, presumably by altering
the organic matter to make it more readily reactive with chlorine to form such by-products (Toor
and Mohseni, 2007; Pisarenko et al., 2013; Glauner et al., 2005). The AOX results suggest that
this alteration can produce organic precursors that can react with chlorine very quickly—within
60 s—to form AOX species, but that the species formed during this time do not include
appreciable THMs or HAAs (but some DCAN and BCAN). Again, the maximum AOX
formation during the 60 s of UV/chlorine contact time occurred at pH 6.5 (70 µg Cl L–1), where
it is predicted that maximum •OH formation occurs.
Pretreatment by AOPs generally led to much smaller differences in 24 h AOX formation
potential (AOX-FP) relative to the non-pretreated controls than were observed for the THMs,
HAAs, or HANs under parallel conditions. For example, the largest increase in 24 h AOX-FP
was when pretreating with UV/chlorine at pH 6.5, with a 30–60% increase compared to the non-
pretreated controls in all waters. This is compared to 90–110% for THMs and HAAs, and 220–
260% for HANs. This also suggests that UV/chlorine treatment changes the nature of the AOX
precursors, such that they react more quickly (as observed during the 60 s chlorine contact time
in the pilot-scale test discussed earlier), but that there is only a relatively small increase in the
total precursor concentration when given a long enough contact time (24 h) to form AOX. In
other words, AOP pretreatment makes the AOX precursors react more quickly but leads to a
smaller increase in the total precursor material relative to THMs and HAAs, under the
conditions tested.
Of the total AOX-FP without any pre-UV or AOP treatment, the unidentified fraction
consistently made up about 40–50% for both Cornwall and Lake Simcoe waters (data not
shown). This is consistent with much previous research (e.g. Pourmoghaddas and Stevens, 1995;
Dotson et al., 2010). When the samples were then exposed to UV photolysis, the unidentified
fraction of the AOX-FP was reduced by 2–15%. Pretreatment by UV/chlorine or UV/H2O2 led
92
to an additional 2–15% reduction in unidentified DBPs. This suggests that UV and AOPs
convert the high-molecular-weight AOX precursors to low-molecular-weight precursors that
lead to the formation of identifiable DBPs. Based on the data shown previously, the increase in
identifiable DBPs exhibits a disproportionately higher increase in HAN-FP (100–260% increase
relative to no pretreatment), compared to THM-FP and HAA-FP (30–110% increases relative to
no pretreatment for both).
5.3.5 Inorganic DBPs: ClO4–, ClO3
–, ClO2–, and BrO3
–
No ClO4–, ClO3
–, ClO2–, or BrO3
– was formed by the application of H2O2 alone, or by UV
alone or UV/H2O2. The free chlorine solution contained a trace level of ClO2– approximately
equivalent to 0.2% of the final free chlorine concentration, as well as lower amounts of ClO4–,
and BrO3– (0.001–0.01% and 0.01–0.03%, respectively). ClO3
– was present in the free chlorine
at a much higher concentration, ranging from 1–15% depending on the chlorine source. ClO3– in
hypochlorite solutions arises from auto-decomposition of free chlorine during manufacture,
shipment, and storage (Stanford et al., 2011).
The pre-existing ClO2– concentration decreased by more than 99% on exposure to
UV/chlorine treatment at 10 mg L–1. However, the UV/chlorine AOP was observed to have no
effect on the measured ClO4–, with the pre-existing ClO4
– in the free chlorine solution remaining
unchanged. The minimal formation of ClO4– and ClO2
– is consistent with reports by Buxton and
Subhani (1972) and Feng et al. (2010).
BrO3– formation in the order of 0.1 µg L–1 to 2 µg L–1 was observed on UV/chlorine
treatment, with the higher formation occurring at the lower pH. Chlorine photolysis was
reported to produce •OH and ozone (Forsyth et al., 2013), which both play important roles in
BrO3– formation from Br– (Hofmann and Andrews, 2006). BrO3
– normally is formed in an
ozonated water preferentially at higher pH (Song et al. 1996), so this trend likely reflects the
greater •OH formation by UV/chlorine at lower pH. While this bromate formation is reasonably
small, the pre-existing bromate in the hypochlorite solution available at the Cornwall plant (full-
scale test in April) added approximately 3 µg L–1 to the water when the chlorine was dosed at 10
mg L–1. Therefore, the possible formation of BrO3– during the application of UV/chlorine needs
to be considered in view of bromate limits that are often in the order of 10 µg L–1 (e.g. Health
Canada, 2012).
93
ClO3– was found to be a major product of chlorine photolysis. Approximately 2–17% (by
mass) of the photolyzed free chlorine was converted to ClO3–. There is a possible trend that can
be observed in Figure 5.6 whereby the percentage formation of chlorate from photolyzed
chlorine may increase with both increasing chlorine dose and pH; however, there was not
sufficient data generated to determine this trend with statistical confidence. Buxton and Subhani
(1972) observed that the percentage of photolyzed free chlorine being converting to ClO3– was
9% when exposed to monochromatic UV light at 365 or 313 nm. They also found that the
percentage increased to 17% at 254 nm, a result that was substantiated by Feng et al. (2010),
who found 20–30% of the photolyzed free chlorine became ClO3– at 254 nm. It is noted that MP
UV was used in this research, different from the studies carried out by Buxton and Subhani
(1972) and Feng et al. (2010).
Figure 5.6 Formation of ClO3
– relative to free chlorine photodecomposition in the full- and
pilot-scale experiments. Error bars represent the values of the duplicates measured. Low,
medium, and high represent free chlorine doses of 2, 6, and 10 mg L–1, respectively.
Some jurisdictions have limits on allowable ClO3– in drinking water. In Canada, a national
guideline of 1 mg L–1 exists (Health Canada, 2012). Assuming 17% conversion of chlorine to
chlorate on photolysis, the maximum chlorine concentration that can undergo photolysis for a
UV/chlorine AOP would be 5.8 mg L–1, assuming that no additional chlorate is added through
94
free chlorine solution contamination. In this work, not all chlorine that was applied was
photolyzed. For example, at the highest chlorine dose of 10 mg L–1, approximately up to 6 mg
L–1 of free chlorine underwent photolysis, however the UV reactors used for the full-scale
testing in this study were early AOP models and may not apply as high a UV dose as would be
used today. As such, it would be expected that more complete chlorine photolysis would be
experienced across modern reactors.
5.4 Discussion
5.4.1 Rapid DBP Formation within the UV/Chlorine Reactor
To the authors’ knowledge, this is the first work to explore DBP formation during the very
short chlorine contact times (30–60 s) but relatively high chlorine doses (2–10 mg L–1)
associated with a UV/chlorine AOP. Previous work that explored DBP formation from
UV/chlorine used longer contact times. For example, Pisarenko et al. (2013) compared THM
and HAA formation when exposing a natural water to either UV/chlorine or chlorine alone, for
2 h. Exposure to UV/chlorine led to increases in both THM and HAA formation of up to 27 µg
L–1 relative to chlorination alone. Similarly, Weng et al. (2012) observed significantly greater
formation of DCAN from its precursors, L-histidine and L-arginine, when applying UV/chlorine
over 10–30 min of exposure compared to chlorine alone.
In this work, THM and HAA formation by UV/chlorine during the 30–60 s contact time
were always observed to be below 14 µg L–1, and usually much lower. DCAN and BCAN
formation was quick, with up to 6 µg L–1 formation within 30–60 s, approximately equivalent to
the 24 h formation potential. AOX formation in water that had previously been chlorinated was
minimal. For the water that had not been prechlorinated, little AOX was formed because of the
chlorine exposure alone (15 µg Cl L–1), but UV/chlorine produced up to 70 µg Cl L–1. This
suggests that the formation of AOX within the short contact time of a UV reactor is enhanced in
part by UV photolysis and/or radical oxidation.
While the formation of organic DBPs during the short exposure time was relatively small,
chlorate formation was significant, at up to 17% (by mass) of the free chlorine that underwent
photolysis. These results suggest that chlorate may be a limiting DBP associated with
UV/chlorine treatment for waters similar to those tested in this study. The formation of chlorate
95
is likely to be manageable through careful chlorine dosing. More research is needed, however,
to identify factors that might increase the conversion of chlorine to chlorate during UV/chlorine
treatment.
5.4.2 Impact of UV/Chlorine on 24 Hour DBP-FP
Previous studies have reported the impact of AOP pretreatment on subsequent DBP-FP, but
usually associated with the UV/H2O2 AOP. In the previous work, low AOP doses (e.g., UV in
the order of 1,000 mJ cm–2) often caused increases in THM/HAA-FP (Dotson et al., 2010;
Kleiser and Frimmel, 2000; Liu et al., 2006; Shah et al., 2011; Liu et al., 2012), but higher AOP
doses (UV of 3,500–5,000 mJ cm–2) led to reductions in THM/HAA-FP (Toor and Mohseni,
2007; Liu et al., 2002). Our tests applied AOP doses that were in the ‘low to moderate’ range
where often THM/HAA-FP reportedly increased. Consistent with this previous work, the results
showed that THM, HAA, HAN, and AOX formation potentials were all increased by the
UV/chlorine and UV/H2O2 AOPs. In general, UV/chlorine and UV/H2O2 when applied at molar
equivalent concentrations were found to have similar effects on the subsequent 24 h DBP-FP,
except that the greatest increase in FP was consistently associated with UV/chlorine at pH 6.5,
where the greatest formation of •OH is predicted to occur (Wang et al., 2012).
In terms of the individual components of THMs and HAAs, increases of TCM-FP and
BDCM-FP by UV/chlorine and UV/H2O2 were observed to be higher than the increase of
CDBM-FP. For example, at pH 7.5 with the highest chlorine and H2O2 doses (10 and 4.8 mg
L–1, respectively) TCM-FP and BDCM-FP increased by 30–60% and 60–80%, respectively,
while less than a 10% increase in CDBM-FP was found. Difference in DCAA-FP from TCAA-
FP and BCAA-FP were also observed. Under the same condition of the previous example,
DCAA-FP increasd by approximately 50–100% because of the AOP pretreatment, while
TCAA-FP and BCAA-FP increased by less than 20% in most cases. Further investigation
should be conducted to explore the impact on individual DBP species in more detail.
5.4.3 Role of the Chlorine Radical (•Cl)
•Cl and •OH are simultaneously produced on chlorine photolysis, but it is difficult to
distinguish the role of each one in terms of DBP formation. In general, •Cl is a strong but
selective oxidant. It is scavenged quickly (8.5 × 109 M–1 s–1) by chloride (Cl–) that is usually
96
present in free chlorine solutions to form •Cl2–, which in turn has negligible reactivity with
NOM (Buxton et al., 2000; Jayson et al., 1973; Nagarajan and Fessenden, 1985). The role of •Cl
in forming organic DBPs might therefore depend on its competitive reactivity with NOM
compared to chloride. Unfortunately, no reaction rate with NOM has been reported. As a
hypothetical example, if it is assumed that •Cl reacts with NOM at the same rate as with
benzene (1.8 × 1010 M–1 s–1; Fang et al., 2014), then, with the same amount of NOM and
chloride present in the Cornwall and Lake Simcoe waters, it might be predicted that about 13-
14% of the •Cl might react with the NOM, suggesting to some that the role of •Cl in organic
DBP formation would be non-negligible. More research in this area is needed.
5.5 Conclusions The UV/chlorine AOP process is attractive because of its operational simplicity and its
comparable efficiency to the UV/H2O2 process under some conditions. The most significant
potential drawback, however, is likely to be concerns about DBP formation. This research
suggests that under the conditions tested, the formation of organic DBPs within the UV reactor
could be characterized as ‘low’ relative to regulatory limits and current DBP concentrations at
the sites from which water was collected. There was evidence, however, that AOP treatment
formed DBP precursors, leading to higher organic DBP concentrations in formation potential
tests. This effect was similar for both UV/chlorine and UV/H2O2 treatment. However, since the
formation potential tests are designed to accentuate DBP formation by applying unrealistically
high chlorine concentrations, the magnitude of any increases in DBP formation in the
distribution system following AOP treatment is therefore likely to be much lower than was
observed in the formation potential tests in this study.
The most likely limiting factor for UV/chlorine treatment would be the formation of
chlorate, for waters similar to those studied in this work. The maximum conversion of chlorine
to chlorate that was observed was 17% (by mass).
Acknowledgements The authors express their appreciation to the staff at the Cornwall Water Purification Plant,
including Owen O'Keefe, Daniel Drouin, and Morris McCormick, for their help with the full-
scale experiments. The authors also would like to thank Dr. Vasile Furdui from the Ontario
97
Ministry of Environment and Climate Change for his generous help in analyzing inorganic DBP
samples.
This work was funded by the Natural Sciences and Engineering Research Council of
Canada through the Industrial Research Chair program.
References APHA, AWWA, WEF, 2005. Standard Methods for the Examination of Water & Wastewater,
21st edition. American Public Health Association, American Water Works Association, and
Water Environment Federation, Washington, D.C., USA.
Buxton, G.V., Bydder, M., Salmon, G.A., Williams, J.E., 2000. The reactivity of chlorine atoms
in aqueous solution. Part III. The reaction of Cl• with solutes. Physical Chemistry Chemical
Physics, 2, 237–245.
Buxton, G.V., Subhani, M.S., 1972. Radiation chemistry and photochemistry of oxychlorine
ions. Part 2.–Photodecomposition of aqueous solutions of hypochlorite ions. Journal of the
Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68,
958–969.
Crebelli, R., Conti, L., Monarca, S., Feretti, D., Zerbini, I., Zani, C., Veschetti, E., Cutilli, D.,
Ottaviani, M., 2005. Genotoxicity of the disinfection by-products resulting from peracetic
acid- or hypochlorite-disinfected sewage wastewater. Water Research, 39(6), 1105–1113.
Deng, L., Huang, C.-H., Wang, Y.-L., 2014. Effects of combined UV and chlorine treatment on
the formation of trichloronitromethane from amine precursors. Environmental Science &
Technology, 48(5), 2697–2705.
Dotson, A.D., Keen, V.S., Metz, D., Linden, K.G., 2010. UV/H2O2 treatment of drinking water
increases post-filtration DBP formation. Water Research, 44(12), 3703–3713.
Fang, J., Fu, Y., Shang, C., 2014. The roles of reactive species in micropollutant degradation in
the UV/free chlorine system. Environmental Science & Technology, 48(3), 1859–1868.
Farré, M.J., Reungoat, J., Argaud, F.X., Rattier, M., Keller, J., Gernjak, W., 2011. Fate of N-
nitrosodimethylamine, trihalomethane and haloacetic acid precursors in tertiary treatment
including biofiltration. Water Research, 45(17), 5695–5704.
98
Feng, Y., Simith, D.W., Bolton, J.R., 2010. A potential new method for determination of the
fluence (UV dose) delivered in UV reactors involving the photodegradation of free chlorine.
Water Environment Research, 82(4), 328–334.
Forsyth, J.E., Zhou, P., Mao, Q., Asato, S.S., Meschke, J.S., Dodd, M.C., 2013. Enhanced
inactivation of Bacillus subtilis spores during solar photolysis of free available chlorine.
Environmental Science & Technology, 47(22), 12976–12984.
Furdui, V.I., Tomassini, F., 2010. Trends and sources of perchlorate in arctic snow.
Environmental Science & Technology, 44(2), 588–592.
Glauner, T., Kunz, F., Zwiener, C., Frimmel, F.H., 2005. Elimination of swimming pool water
disinfection by-products with advanced oxidation processes (AOPs). Acta Hydrochimica et
Hydrobiologica, 33(6), 585–594.
Glezer, V., Harris, B., Tal, N., Iosefzon, B., Lev, O., 1999. Hydrolysis of haloacetonitriles:
linear free energy relationship, kinetics and products. Water Research, 33(8), 1938–1948.
Hansen, K.M.S., Willach, S., Antoniou, M.G., Mosbæk, H., Albrechtsen, H.-J., Andersen, H.R.,
2012. Effect of pH on the formation of disinfection byproducts in swimming pool water – Is
less THM better? Water Research, 46(19), 6399–6409.
Health Canada, 2012. Guidelines for Canadian Drinking Water Quality – Summery Table.
Water, Air and Climate Change Bureau, Healthy Environments and Consumer Safety Branch,
Health Canada, Ottawa, Ontario, Canada.
Hofmann, R., Andrews, R.C., 2006. Impact of H2O2 and (bi)carbonate alkalinity on ammonia’s
inhibition of bromate formation. Water Research, 40(18), 3343–3348.
Jayson, G.G., Parsons, B.J., Swallow, A.J., 1973. Some simple, highly reactive, inorganic
chlorine derivatives in aqueous solution. Their formation using pulse of radiation and their
role in the mechanism of the Fricke dosimeter. Journal of the Chemistry Society, Faraday
Transaction 1: Physical Chemistry in Condensed Phases, 69, 1597–1607.
Kang, N., Anderson, T.A., Jackson, W.A., 2006. Photochemical formation of perchlorate from
aqueous oxychlorine anions. Analytica Chimica Acta, 567(1), 48–56.
Klassen, N.V., Marchington, D., McGowan H.C.E., 1994. H2O2 determination by the I3– method
and by KMnO4 titration. Analytical Chemistry, 66(18), 2921–2925.
99
Kleiser, G., Frimmel, F.H., 2000. Removal of precursors for disinfection by-products (DBPs) –
differences between ozone- and OH-radical-induced oxidation. The Science of the Total
Environment, 256(1), 1–9.
Liu, W., Andrews, S.A., Bolton, J.R., Linden, K.G., Sharpless, C., Stefan, M., 2002.
Comparison of disinfection byproduct (DBP) formation from different UV technologies at
bench scale. Water Science and Technology: Water Supply, 2(5–6), 515–521.
Liu, W., Andrews, S.A., Stefan, M.I., Bolton, J.R., 2003. Optimal methods for quenching H2O2
residuals prior to UFC testing. Water Research, 37(15), 3697–3703.
Liu, W., Cheung, L.-M., Yang, X., Shang, C., 2006. THM, HAA and CNCl formation from UV
irradiation and chlor(am)ination of selected organic waters. Water Research, 40(10), 2033–
2043.
Liu, W., Zhang, Z., Yang, X., Xu, Y., Liang, Y., 2012. Effects of UV irradiation and
UV/chlorine co-exposure on natural organic matter in water. Science of the Total
Environment, 414, 576–584.
Nagarajan, V., Fessenden, R.W., 1985. Flash photolysis of transient radicals. 1. X2– with X = Cl,
Br, I., and SCN. The Journal of Physical Chemistry, 89(11), 2330–2335.
Pisarenko, A.N., Stanford, B.D., Snyder, S.A., Rivera, S.B., Boal, A.K., 2013.Investigation of
the use of chlorine based advanced oxidation in surface water: oxidation of natural organic
matter and formation of disinfection byproducts. Journal of Advanced Oxidation
Technologies, 16(1), 137–150.
Plummer, J.D., Edzwald, J.K., 2001. Effect of ozone on algae as precursors for trihalomethane
and haloacetic acid production. Environmental Science & Technology, 35(18), 3661–3668.
Pourmoghaddas, H., Stevens, A.A., 1995. Relationship between trihalomethanes and haloacetic
acids with total organic halogen during chlorination. Water Research 29 (9), 2059–2062.
Shah, A.D., Dotson, A.D., Linden, K.G., Mitch, W.A., 2011. Impact of UV disinfection
combined with chlorination/chloramination on the formation of halonitromethanes and
haloacetonitriles in drinking water. Environmental Science & Technology, 45(8), 3657–3664.
Shams El Din, A.M., Mohammed, R.A., 1998. Kinetics of the reaction between hydrogen
peroxide and hypochlorite. Desalination, 115(2), 145–153.
100
Song, R., Donohoe, C., Minear, R., Westerhoff, P., Ozekin, K., Amy, G., 1996. Empirical
modeling of bromate formation during ozonation of bromide-containing waters. Water
Research, 30(5), 1161–1168.
Stanford, B.D., Pisarenko, A.N., Snyder, S.A., Gordon, G., 2011. Perchlorate, bromate, and
chlorate in hypochlorite solutions: guidelines for utilities. Journal of American Water Works
Association, 103(6), 71–83.
Summers, R.S., Hooper, S.M., Shukairy, H.M., Solarik, G., Owen, D., 1996. Assessing DBP
yield: uniform formation conditions. Journal of American Water Works Association, 88(6),
80–93.
Toor, R., Mohseni, M., 2007. UV-H2O2 based AOP and its integration with biological activated
carbon treatment for DBP reduction in drinking water. Chemosphere, 66(11), 2087–2095.
United States Environmental Protection Agency (USEPA), 1995. Method 551.1. Determination
of chlorination disinfection byproducts, chlorinated solvents, and halogenated
pesticides/herbicides in drinking water by liquid-liquid extraction and gas chromatography
with electron-capture detection, Revision, 1.0. Available from
<http://www.epa.gov/sam/pdfs/EPA-551.1.pdf> (accessed 11.09.11).
United States Environmental Protection Agency (USEPA), 1997. Determination of inorganic
anions in drinking water by ion chromatography. Available from
<http://water.epa.gov/scitech/drinkingwater/labcert/upload/met300.pdf > (accessed 09.09.14).
Von Gunten, U., Hoigné, J., 1994. Bromate formation during ozonation of bromide-containing
waters: interaction of ozone and hydroxyl radical reactions. Environmental Science &
Technology, 28(7), 1234–1242.
Wang, D., Bolton J.R., Hofmann, R., 2012. Medium pressure UV combined with chlorine
advanced oxidation for trichloroethylene destruction in a model water. Water Research,
46(15), 4677–4686.
Wang, D., Walton, T., McDermott, L., Hofmann, R., 2011. Control of TCE using UV combined
with hydrogen peroxide or chlorine. Proceedings of IOA-IUVA 2011 World Congress &
Exhibition, Paris, France.
Watts, M.J., Linden, K.G., 2007. Chlorine photolysis and subsequent OH radical production
during UV treatment of chlorinated water. Water Research, 41(13), 2871–2878.
101
Watts, M.J., Rosenfeldt, E.J., Linden, K.G., 2007. Comparative OH radical oxidation using UV-
Cl2 and UV-H2O2 processes. Journal of Water Supply: Research and Technology–AQUA,
56(8), 469–477.
Weng, S., Li, J., Blatchley III, E.R., 2012. Effects of UV254 irradiation on residual chlorine and
DBPs in chlorination of model organic-N precursors in swimming pools. Water Research,
46(8), 2674–2682.
Yang, X., Chang, C., Westerhoff, P., 2007. Factors affecting formation of haloacetonitriles,
haloketones, chloropicrin and cyanogen halides during chloramination. Water Research, 41(6),
1193–1200.
102
6. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS
6.1 Summary and Conclusions Chlorine photolysis by UV exposure is a novel AOP, which probably involves simpler
operation than UV/H2O2, but has not been widely investigated in the literature. This work has
explored the mechanisms of chlorine photolysis through bench-scale experiments and numerical
simulation (Chapter 3). Based on the results for HOCl and OCl– photolysis at the same initial
molar concentration (0.15 mM), although OCl– absorbed MP UV light approximately 4.5 times
higher than HOCl, and the quantum yields of photolysis and •OH formation for both HOCl and
OCl– were quite close to each other, the •OH concentration generated by HOCl was
approximately 40 times higher than that generated by OCl–. This implies that OCl– is a much
stronger •OH scavenger than HOCl.
Bench-scale TCE destruction (Chapter 3), pilot-scale caffeine destruction, and full-scale
geosmin, MIB, and caffeine destruction (Chapter 4) all showed that the efficiency of the
UV/chlorine AOP was generally comparable to that of the UV/H2O2 AOP at pH 7.5 and 8.5, but
was superior at pH 6.5. This reflects the impact of the prevalence of HOCl relative to OCl– at
the lower pH, with its reported lower •OH scavenger potential.
The formation of various DBPs by the UV/chlorine AOP was investigated in full- and
pilot-scale experiments in two post-filtration waters (Chapter 5) under practical operating
conditions (chlorine doses: 2–10 mg L–1, contact time: <1 minute). THM and HAA formation
was observed to be minimal, while fast formation of DCAN and BCAN was found. Fast
formation of AOX was observed in the water without prechlorination, while insignificant
additional formation was found in the prechlorinated water. Among the monitored inorganic
DBPs, including chlorate, bromate, perchlorate, and chlorite, only chlorate and bromate were
formed during chlorine photolysis, equivalent to approximately 2–17% and 0.01–0.05% of
photolyzed chlorine, respectively. Additionally, DBP formation potential for THMs, HAAs,
HANs, and AOX was observed to increase following UV/chlorine pretreatment, approximately
to the same extent as was observed following UV/H2O2 pretreatment.
In summary, UV/chlorine is a promising AOP that may be an alternative to the UV/H2O2
AOP for the destruction of organic contaminants and the control of taste and odour issues in
103
drinking water treatment. Under the tested conditions, UV/chlorine produces low concentrations
of organic DBPs within the UV reactors relative to the regulatory limits. The process does
appear to form DBP precursors that would consequently increase DBP formation during
secondary chlorination, but the effect is similar to that of UV/H2O2 treatment. Chlorate is a
significant photoproduct of chlorine photolysis. This is the single factor that may possibly limit
UV/chlorine applications.
6.2 Recommendations for Future Work A number of areas were identified that deserve further investigation, including:
• Investigation of rate constant of free chlorine with •OH. The reaction rate constants
between HOCl/OCl– and •OH were obtained from limited studies, and need to be verified.
Pulse radiolysis analysis is often used to determine reaction rate constants involving •OH,
and it was successfully used in this work to verify the rate constant between OCl– and •OH
by measuring the change of the •OCl absorbance (shown in Appendix A). •OCl is a product
of the reaction between OCl– and •OH. However, the value between HOCl and •OH could
not be measured, because there was no significant increase of •OCl absorbance. A novel
approach is therefore required.
• Investigation of quantum yields of chlorine photolysis and •OH formation without the
impact of chain reactions. Chlorine photolysis has been found to involve chain reactions
that may influence the apparent quantum yields of chlorine photolysis and •OH formation.
It is thus important to determine the quantum yields accurately in the absence of the chain
reactions, such that accurate kinetic models of chlorine photolysis can be built. A chain
reaction inhibitor may be applied for this purpose. However, the inhibitor needs to be
selected carefully. For example, it should be a strong •OH scavenger, but stable in the
presence of UV or chlorine individually. It should not absorb UV light significantly, and its
products with •OH must not react with free chlorine.
• Investigation of UV/chlorine efficiency in a wide range of water quality. UV/chlorine
efficiency was evaluated in 3 types of water in this study, which could all be considered to
have low- to moderately-low •OH scavenging potential (TOC levels less than 3.5 mg L–1
and alkalinity less than 150 mg L–1 as CaCO3). Since UV/chlorine is probably more
competitive with UV/H2O2 in a water with a higher •OH scavenging potential, it is
104
important to investigate UV/chlorine performance over a wide range of water quality, and
in particular in waters with a higher TOC level where not only higher scavenging potential
exists but where the TOC may exert a significant chlorine demand.
• Further investigation of DBP formation. Potential DBP formation is likely the most
significant concern with UV/chlorine treatment, and while the results of this study are
promising in this area, only two natural water matrices were evaluated. More waters should
be tested, such as a water containing more DBP precursors and/or bromide (which converts
to hypobromous acid in a chlorinated water to produce brominated organic DBPs). The
formation of chlorate may possibly be a limiting factor for UV/chlorine treatment, and the
influence of water quality and treatment variables on its formation are unknown.
• Investigation of UV/chlorine using low-pressure lamps. This research was carried out
exclusively using MP UV lamps. Since LP UV lamps are also commonly used in practice,
and have a very different spectral emittance from that of MP lamps, it is necessary to
investigate the UV/chlorine mechanisms, efficiency, and DBP formation with the
application of LP lamps.
105
APPENDICES
A. Pulse Radiolysis Analysis for Determination of Rate Constant of Free
Chlorine with Hydroxyl Radical Pulse radiolysis was performed at the Notre Dame Radiation Laboratory (NDRL) for
determination of the rate constants between free chlorine (HOCl and OCl–) and •OH, according
to the methods described by Buxton and Subhani (1972)1, Ulanski and von Sonntag (2000)2,
and Westerhoff et al. (2007)3. pH values were adjusted to 5 and 10, respectively, to obtain
generally pure HOCl and OCl– solutions. A single electron pulse was applied to impinge a free
chlorine solution and to make the solution excited and ionized instantly. Three transients
were produced during this process (with the relative abundance): •OH (45%), H (10%), and eaq–
(45%). Since the solutions were saturated with N2O, eaq– was converted to •OH promptly.
Therefore, it is considered that after electron pulse within several nanoseconds 90% of transients
were •OH and 10% were H atoms that are less reactive than •OH. The produced •OH then
reacted with chlorine and produced •OCl, which was monitored by a spectrophotometer at 280
nm in very short time intervals (2.5 × 10–7 s). By measuring the absorbance changes versus time
at different initial chlorine concentrations, the second order rate constant between chlorine and
•OH can be determined.
Figure A.1 shows the raw data and the regression for pulse radiolysis analysis for rate
constant between OCl– and •OH at pH 10. The y axis is the solution absorbance at 280 nm,
which was primarily contributed by •OCl. Therefore, it can be considered to be the absorbance
of •OCl solely. The x axis shows the time scale after the pulse was performed.
The main reaction between OCl– and •OH is: OCl– + •OH → •OCl + OH–. This is an
irreversible reaction, thus
1 Buxton, G.V., Subhani, M.S., 1972. Radiation chemistry and photochemistry of oxychlorine ions. Part 1.–Radiolysis of aqueous solutions of hypochlorite and chlorite ions. Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases, 68, 947–957. 2 Ulanski, P., von Sonntag, C., 2000. OH-radical-induced chain scission of chitosan in the absence and presence of dioxygen. Journal of the Chemical Society, Perkin Transactions 2, 2000, 2022–2028. 3 Westerhoff, P., Mezyk, S.P., Cooper, W.J., Minakata, D., 2007. Electron pulse radiolysis determination of hydroxyl radical rate constants with Suwannee River fulvic acid and other dissolved organic matter isolates. Environmental Science & Technology, 41(13), 4640–4646.
106
Figure A.1 Pulse radiolysis results for rate constant between OCl– with •OH
[ OCl] [ OH][OCl ]d kdt
−•= • [A.1]
Since the concentration of OCl– was much higher than that of •OH, the change of OCl–
concentration was minimal. Therefore, [OCl–] is considered to be constant. •OH was generated
immediately (at t = 0) after the pulse radiolysis was applied, and it was gradually consumed by
OCl– until depleted completely when the curves reached plateaus shown in Figure A.1. When
•OH was just exhausted, •OCl reached the maximum concentration. Then it started to decrease,
because •OCl could not be generated any more, but underwent decay. Assuming ' [OCl ]k k −= ,
[ OCl] '[ OH]d kdt•
= • [A.2]
On the other hand, [ OH] [ OH][OCl ]d kdt
−•= − • , or
[ OH] '[ OH]d kdt•
= − • [A.3]
Integrating Equation [A.3] gives
0
[ OH]ln '[ OH]
k t•= −
• [A.4]
If k’ values at different [OCl–] are known, k can be determined by plotting k’ versus [OCl–].
Therefore, the purpose was to determine k’ from the raw data shown in Figure A.1. k’ can be
calculated using the data fitting method discussed below.
107
Supposing •OH was consumed by OCl– only, besides the Equations [A.2] and [A.3], there
was a relationship between [•OCl] and [•OH]:
[•OH] = [•OH]0 – [•OCl] [A.5]
which means that the concentration of •OH at time t was equal to the initial concentration of
•OH at time 0 minus the concentration of •OCl at time t, because 1 mole of •OCl was produced
by 1 mole of •OH consumed.
Substituting Equation [A.5] to Equation [A.2] yields 0[ OCl] '([ OH] [ OCl])d k
dt•
= • − • . After
integration, '
0[ OCl] [ OH] (1 )k te−• = • − [A.6]
The absorbance of •OCl (shown in y-axis of Figure A.1) also had the same relationship
with t as [•OCl]. Since the initial absorbance of the solution before pulse radiolysis was not
zero, the relationship between •OCl absorbance and t can be expressed as: '
0k ty y Ae−= + [A.7]
where, y is the absorbance of •OCl at time t, k’ is the observed first order rate constant in
Equations [A.2] and [A.3], y0 and A are constants.
According to the least square regression method, k’ as well as y0 and A was solved using
the “Solver” function in Excel. Since ' [OCl ]k k −= , k was then determined, which is the slope of
k’ as a function of [OCl–]. k was calculated to be 7.17 × 109 M–1 s–1, which is close to 9.0 × 109
M–1 s–1 determined by Buxton and Subhani (1972).
The raw data for pulse radiolysis analysis for the rate constant between HOCl and •OH at
pH 5 are shown in Figure A.2. It shows that there was no significant trend of •OCl absorbance
increase, which means the formation of •OCl was not obvious. There are several probable
reasons:
• The product of HOCl with •OH was not •OCl, thus the product cannot be detected at 280 nm.
However, since HOCl is the conjugate acid of OCl–, •OCl is likely the main product.
• The reaction rate between HOCl and •OH is slow compared with that between OCl– with
•OH. Therefore, the principal portion of •OH was consumed by Cl–, according to Equation
[2.34], and, thus, •OCl concentration was too low to detect.
A different approach to measure the rate constant between HOCl and •OH is therefore required.
108
Figure A.2 Pulse radiolysis results for rate constant between HOCl with •OH
B. Determination of the Fluence Rate of the MP Lamp in the Collimated
Beam Apparatus A ferrioxalate actinometer method was used to determine the fluence rate of the MP lamp
in the collimated beam apparatus, following the procedures described by Bolton et al. (2009)4.
The actinometer with a volume of 10 mL was contained in a Petri dish (diameter = 4.9 cm),
with/without the coverage of a 345 nm long-pass filter. An opaque aluminum cap (diameter =
10 cm) with a circular hole (diameter = 1.50 cm, area = 1.77 cm2) was also placed at the centre
of the Petri dish, so that the determined fluence rate reflected the value at the centre of the Petri
dish without the impact of a Petri factor. After the lamp exposure for 3 min in a dark room, the
rate of Fe2+ generation (d[Fe2+]/dt) due to 200–345 nm UV exposure can be determined by
subtracting the d[Fe2+]/dt in the presence of the filter from the d[Fe2+]/dt in the absence of the
filter, according to Sharpless and Linden (2003)5. The fluence rate (200–400 nm) was then
calculated using the spreadsheet shown in Table B.1. The determined value was then used to
calculate the average fluence rate in a specific trichloroethylene solution, using the Bolton
spreadsheet (Bolton, 2002)6.
4 Bolton, J.R., Stefan, M.I., Shaw, P.-S., Lykke, K.R., 2009. Determination of the quantum yield of the ferrioxalate and KI/KIO3 actinometers and a method for the calibration of radiometer detectors. CDROM Proceedings 5th UV World Congress, Amsterdam, The Netherlands. 5 Sharpless, C.M., Linden, K.G., 2003. Experimental and model comparisons of low- and medium-pressure Hg lamps for the direct and H2O2 assisted UV photodegradation of N-nitrosodimethylamine in simulated drinking water. Environmental Science & Technology, 37(9), 1933–1940. 6 Bolton, J.R., 2002. Germicidal fluence (UV dose) calculation for a medium pressure UV lamp. Available at: http://www.iuva.org or from J.R. Bolton ([email protected]). Accessed on Sep. 24, 2009.
109
Table B.1 Calculation of 200–400 nm fluence rate from the actinometry and spectroradiometer results A B C D E F G H I J K L
Wavelength (nm)
Photon energya
(J einstein–1)
Assumed spectral irradiance from the spectroradiometerb
(µW cm–2 nm–1)
Assumed spectral photon irradiancec
(einstein cm–2 s–1 nm–1)
∆λd (nm)
Reflection factor
from air to watere
Quantum yield of
Fe2+ formationf
Assumed d[Fe2+]/dt
200–345 nmg
Actual d[Fe2+]/dt
200–345 nmh
Correction factori
Actual spectral photon
irradiancej (einstein cm–2 s–1)
Spectral irradiancek (mW cm–2)
199.81 200.03 5.98E+05 0 0 0.22 0.969 1.49 0 2.60E-06 2.22E-06 0 0.00E+00 200.24 5.97E+05 0 0 0.21 0.970 1.49 0 0 0.00E+00 200.46 5.97E+05 0 0 0.22 0.970 1.49 0 0 0.00E+00 200.67 5.96E+05 0 0 0.21 0.970 1.49 0 0 0.00E+00 200.89 5.95E+05 0 0 0.22 0.970 1.49 0 0 0.00E+00
⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ 253.84 4.71E+05 19.0 4.03E-05 0.22 0.975 1.40 2.14E-03 1.97E-11 9.29E-03 254.05 4.71E+05 15.8 3.37E-05 0.21 0.975 1.40 1.70E-03 1.57E-11 7.39E-03 254.26 4.70E+05 13.3 2.82E-05 0.21 0.975 1.40 1.43E-03 1.32E-11 6.19E-03 254.47 4.70E+05 11.0 2.34E-05 0.21 0.975 1.40 1.19E-03 1.09E-11 5.14E-03 254.69 4.70E+05 9.74 2.07E-05 0.22 0.975 1.40 1.10E-03 1.01E-11 4.76E-03 254.9 4.69E+05 9.80 2.09E-05 0.21 0.975 1.40 1.06E-03 9.74E-12 4.57E-03
⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ 398.57 3.00E+05 2.43 8.11E-06 0.20 0.979 1.25 2.08E-03 3.60E-12 1.08E-03 398.78 3.00E+05 2.47 8.22E-06 0.21 0.979 1.25 2.21E-03 3.83E-12 1.15E-03 398.98 3.00E+05 2.41 8.02E-06 0.20 0.979 1.25 2.06E-03 3.56E-12 1.07E-03 399.19 3.00E+05 2.41 8.03E-06 0.21 0.979 1.25 2.16E-03 3.74E-12 1.12E-03 399.40 3.00E+05 2.38 7.95E-06 0.21 0.979 1.25 2.14E-03 3.71E-12 1.11E-03 399.60 2.99E+05 2.43 8.10E-06 0.20 0.979 1.25 2.08E-03 3.60E-12 1.08E-03 399.81 2.99E+05 2.43 8.12E-06 0.21 0.979 1.25 2.18E-03 3.79E-12 1.13E-03
Total 1.17E+00 (200– 345 nm) 7.36
(200–400 nm)
aPhoton energy (J einstein–1) = 6.6216 × 10–34 ×2.9979 × 108 × 6.02214 × 1023 / (Column A × 10–9), according to Bolton (2001)7 bColumn C was from the data determined by the spectroradiometer (Model: USB4000-UV-VIS, Ocean Optics) cColumn D = Column C / Column B d∆λ represents the difference of the two adjacent wavelengths in Column A eColumn F is calculated using the Schiebener’s model described by Huiber (1997)8 fData were obtained from Goldstein and Rabani (2008)9 gColumn H = Column G × Column D × Column E × 1.77 × Column F / (10 / 1000) hThis result was from the ferrioxalate actinometry measurements iCorrection factor = Actual d[Fe2+]/dt from 200–345 nm / the sum of Column H from 200–345 nm jColumn K = correction factor × Column D × Column E kColumn L = Column K × Column B × 1000
7 Bolton, J. R., 2001. Ultraviolet Applications Handbook. Bolton Photosciences Inc. 628 Cheriton Cres. NW, Edmonton, AB, Canada. 8 Huibers, P.D.T., 1997. Models for the wavelength dependent of the index of refraction of water. Applied Optics, 36(16), 3785–3787. 9 Goldstein, S., Rabani, J., 2008. The ferrioxalate and iodide-iodate actinometers in the UV region. Journal of Photochemistry and Photobiology A: Chemistry, 193(1), 50–55.
110
C. Example of Matlab® Codes
C.1 Matlab® Codes for Simulation of Trichloroethylene Decay by the UV/Chlorine AOP at
11 mg L–1 and pH 5
function dydt = TCE_UV_Cl2(t,y); % y(1): OH radical, y(2): HOCl, y(3): OCl–, y(4): DOC, y(5): TCE, y(6): HCO3
–, y(7):CO32–, y(8): Cl radical formed by TCE photolysis
phiHOCl_OH=0.79; phiOCl_OH=1.18; %data from my TCE experiments % phiHOCl_OH and phiOCl_OH mean quantum yields of OH radical formation by HOCl and OCl– photolysis, respectively phiHOCl=1.06, phiOCl=0.89; %data from my TCE experiments % phiHOCl and phiOCl mean quantum yields of HOCl and OCl– photolysis, respectively Ep0=[0.00E+00 0.00E+00 0.00E+00 0.00E+00 0.00E+00 0.00E+00 4.74E-12 4.43E-12 2.87E-12 3.88E-12 3.33E-12 3.17E-12 4.02E-12 3.48E-12 3.45E-12 4.61E-12 4.89E-12 4.33E-12 5.74E-12 6.38E-12 5.66E-12 7.72E-12 8.44E-12 7.23E-12 9.48E-12 1.03E-11 8.47E-12 1.04E-11 1.04E-11 1.10E-11 8.92E-12 9.96E-12 9.23E-12 7.40E-12 1.03E-11 9.63E-12 5.89E-12 9.10E-12 9.14E-12 7.28E-12 8.54E-12 4.80E-12 3.31E-12 2.77E-12 4.26E-12 4.86E-12 5.23E-12 1.59E-11 2.14E-11 8.87E-12 5.22E-12 6.80E-12 1.47E-11 1.22E-11 8.10E-12 1.17E-11 1.73E-11 2.84E-11 2.90E-11 2.53E-11 1.73E-11 1.73E-11 1.43E-11 1.24E-11 2.90E-11 4.54E-11 2.40E-11 7.70E-12 9.40E-12 1.49E-11 1.17E-11 8.78E-12 5.88E-12 5.39E-12 6.42E-12 1.23E-11 8.62E-12 4.06E-12 5.20E-12 1.49E-11 2.93E-11 1.52E-11 7.04E-12 4.34E-12 3.94E-12 3.40E-12 4.11E-12 3.67E-12 9.11E-12 1.42E-11 1.09E-11 4.85E-12 7.86E-12 4.93E-12 3.98E-12 7.26E-12 4.11E-11 2.85E-11 1.45E-11 5.72E-12 7.14E-12 2.84E-11 7.53E-11 3.77E-11 8.42E-12 3.91E-12 4.09E-12 3.71E-12 3.59E-12 2.95E-12 4.07E-12 1.41E-11 1.07E-10 1.01E-10 5.88E-11 1.45E-11 6.71E-12 3.95E-12 4.22E-12 3.72E-12 3.42E-12 3.27E-12 2.46E-12 2.93E-12 2.89E-12 2.80E-12 2.15E-12 2.65E-12 2.52E-12 2.54E-12 2.62E-12 2.15E-12 2.79E-12 1.42E-11 2.74E-11 1.05E-11 2.76E-12 2.35E-12 2.39E-12 2.51E-12 1.77E-12 2.05E-12 2.07E-12 2.28E-12 2.65E-12 2.12E-12 2.51E-12 2.43E-12 2.22E-12 2.12E-12 2.32E-12 1.89E-12 2.17E-12 2.46E-12 3.57E-12 3.04E-12 2.55E-12 3.07E-12 2.96E-12 3.11E-12 3.11E-12 2.62E-12 4.05E-12 8.74E-12 1.70E-10 3.77E-10 1.99E-10 4.68E-11 1.79E-11 8.85E-12 7.77E-12 6.04E-12 4.49E-12 3.09E-12 3.58E-12 3.64E-12 3.13E-12 3.04E-12 3.34E-12 2.79E-12 3.00E-12 2.85E-12 2.95E-12 2.55E-12 2.40E-12 2.02E-12 2.66E-12 2.38E-12 2.43E-12 3.45E-12 7.50E-12 5.28E-12 2.29E-12 2.33E-12 2.31E-12 2.32E-12 2.36E-12 2.33E-12 2.48E-12 2.02E-12]; %Ep0 is a matrix, which represents the spectral photon irradiance at each wavelength (200–400 nm) at the centre of the Petri dish, and which corresponds to a total irradiance of 1 mW cm–2 from 200 to 400 nm Ep1=7.10*Ep0; %Ep1 is the actual photon irradiance at the centre of the Petri dish used in TCE experiments E1=[8022 7769 7440 7109 6805 6541 6306 6089 5885 5675 5449 5179 4876 4546 4212 3892 3613 3374 3165 2972 2785 2601 2412 2223 2040 1858 1682 1512 1352 1198 1055 922 801 689 588 501 423 354 293 242 198 160 128 102 80 61 47 36 27 19 13 8 3 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02]; %E1 is the spectral molar absorption coefficient of TCE at each wavelength (200 to 400 nm) E2=[91 77 66 58 53 49 46 44 43 42 42 43 44 46 48 50 52 55 58 61 64 68 71 74 77 80 83 86 89 91 93 95 96 97 98 98 98 98 97 96 95 94 92 90 87 85 82 79 76 73 70 67 64 61 58 55 53 50 47 45 43 41 39 37 36 35 33 32 31 31 30 30 29 29 29 28 28 28 28 28 29 29 29 29 29 29 29 30 30 30 30 30 30 30 29 29 29 29 28 28 28 27 27 26 26 25 25 24 24 23 22 22 21 20 20 19 18 17 17 16 15 15 14 13 13 12 12 11 10 10 9 9 8 8 7 7 7 6 6 5 5 5 4 4 4 4 3 3 3 3 3 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02 1.00E-02]; %E2 is the spectral molar absorption coefficient of HOCl at each wavelength (200 to 400 nm) E3=[362 325 294 267 242 220 199 180 162 144 128 113 99 86 74 63 53 45 38 32 26 22 18 15 13 11 10 9 8 7 7 7 7 7 8 8 9 10 11 12 14 15 17 19 21 24 26 30 33 37 41 45 50 55 60 67 73 80 87 95 103 111 120 130 139 149 159 169 180 190 201 212 223 234 245 255 266 276 286 295 304 312 320 327 334 339 345 349 352 355 357 358 359 359 357 355 353 350 346 342 337 331 325 318 311 304 296 288 280 271 262 253 244 235 226 217 208 199 190 181 173 164 156 148 140 133 126 118 112 105 99 93 87 82 77 72 67 62 58 54 50 47 43 40 37 34 32 29 27 25 23 21 20 18 17 15 14 13 12 11 10 9 8 8 7 6 6 5 5 5 4 4 4 3 3 3 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1.00E-02]; %E3 is the spectral molar absorption coefficient of OCl– at each wavelength (200 to 400 nm) A=E1*y(5)+E2*y(2)+E3*y(3);
111
%A is the solution absorption coefficient at each wavelength (200 to 400 nm) RF=[0.969457168 0.969647411 0.969831759 0.970010483 0.970183836 0.970352058 0.970515374 0.970673997 0.970828125 0.970977949 0.971123644 0.971265381 0.971403318 0.971537605 0.971668386 0.971795796 0.971919963 0.972041009 0.972159049 0.972274195 0.97238655 0.972496214 0.972603282 0.972707845 0.972809989 0.972909796 0.973007345 0.97310271 0.973195965 0.973287177 0.973376412 0.973463732 0.973549199 0.97363287 0.973714799 0.973795041 0.973873646 0.973950662 0.974026138 0.974100117 0.974172643 0.974243758 0.974313502 0.974381913 0.974449029 0.974514885 0.974579516 0.974642954 0.974705233 0.974766382 0.974826432 0.974885411 0.974943347 0.975000267 0.975056196 0.97511116 0.975165183 0.975218288 0.975270497 0.975321834 0.975372318 0.97542197 0.97547081 0.975518857 0.975566131 0.975612648 0.975658426 0.975703483 0.975747834 0.975791496 0.975834484 0.975876813 0.975918497 0.975959552 0.975999989 0.976039824 0.976079068 0.976117735 0.976155836 0.976193383 0.976230389 0.976266864 0.976302819 0.976338264 0.976373211 0.976407669 0.976441648 0.976475158 0.976508208 0.976540806 0.976572963 0.976604685 0.976635983 0.976666863 0.976697334 0.976727404 0.97675708 0.97678637 0.976815281 0.976843819 0.976871992 0.976899806 0.976927268 0.976954385 0.976981162 0.977007605 0.977033721 0.977059515 0.977084993 0.977110161 0.977135024 0.977159586 0.977183854 0.977207833 0.977231527 0.97725494 0.977278079 0.977300947 0.97732355 0.97734589 0.977367973 0.977389804 0.977411385 0.977432722 0.977453817 0.977474676 0.977495301 0.977515697 0.977535867 0.977555815 0.977575544 0.977595057 0.977614359 0.977633452 0.977652339 0.977671024 0.97768951 0.9777078 0.977725896 0.977743802 0.97776152 0.977779054 0.977796406 0.977813579 0.977830575 0.977847397 0.977864048 0.97788053 0.977896845 0.977912996 0.977928985 0.977944815 0.977960488 0.977976006 0.977991371 0.978006586 0.978021652 0.978036571 0.978051347 0.978065979 0.978080472 0.978094826 0.978109044 0.978123126 0.978137076 0.978150895 0.978164584 0.978178146 0.978191582 0.978204894 0.978218083 0.978231151 0.9782441 0.978256931 0.978269646 0.978282246 0.978294733 0.978307108 0.978319373 0.978331528 0.978343577 0.978355519 0.978367356 0.97837909 0.978390721 0.978402252 0.978413683 0.978425015 0.978436251 0.97844739 0.978458434 0.978469385 0.978480244 0.978491011 0.978501688 0.978512275 0.978522775 0.978533188 0.978543514 0.978553756]; %RF is the reflection factor at each wavelength (200 to 400 nm) WF=(1-10.^(-A*0.80))./(2.3026*A*0.8); %WF is the water factor at each wavelength (200 to 400 nm) Ep=Ep1.*RF.*WF*0.9814*0.945; %Ep is the average spectral photon irradiance at each wavelength (200 to 400 nm) through the TCE solution. 0.9814 is the divergence factor, 0.945 is the Petri factor sumEp=sum(sum(Ep)); %sumEp is the total photon irradiance at 200–400 nm through the TCE solution R1=((Ep./WF).*E1.*(1-10.^(-A*0.80)))./(A*0.80)*1000; %R1 is the specific rate of photon absorption by TCE at each wavelength (200 to 400 nm) R2=((Ep./WF).*E2.*(1-10.^(-A*0.80)))./(A*0.80)*1000; %R2 is the specific rate of photon absorption by HOCl at each wavelength (200 to 400 nm) R3=((Ep./WF).*E3.*(1-10.^(-A*0.80)))./(A*0.80)*1000; %R3 is the specific rate of photon absorption by OCl– at each wavelength (200 to 400 nm) sumR1=sum(sum(R1)); sumR2=sum(sum(R2)); sumR3=sum(sum(R3)); %sumR means the summation of specific rate of photon absorption from 200 to 400 nm OHFormation=phiHOCl_OH*sumR2*y(2)+phiOCl_OH*sumR3*y(3); % OHFormation means OH radical formation rate HOClPhotolysis=phiHOCl*sumR2*y(2); OClPhotolysis=phiOCl*sumR3*y(3); % HOClPhotolysis and OClPhotolysis mean HOCl and OCl– photolysis rates, respectively % Assume HOClPhotolysis and OClPhotolysis are positive numbers kforward=1.8e3, kbackward=3e9; % for HOCl hydrolysis kw=1.9055e-14; %[OH–] = kw/10^-pH kforward2=2.2, kbackward2=5e10; % for HCO3
– hydrolysis kHOCl_OH=8.46e4; kOCl_OH=9e9; kDOC_OH=3e8; kTCE_OH=2.4e9; kHCO3_OH=8.5e6; kCO3_OH=3.9e8; phiTCE=0.354; phiTCE1=0.13; %data from Li et al. (2004) % phiTCE means the total quantum yield of TCE photolysis; phiTCE1 means the quantum yield of TCE photolysis in Reaction 1 (see Table 3.5) TCEPhotolysis=phiTCE*sumR1*y(5); ClFormation = phiTCE1*sumR1*y(5); % TCEPhotolysis means TCE photolysis rate, ClFormation means the Cl radical formation rate % Assume TCEPhotolysis is a positive number kTCE_Cl=4.88e10; dydt=[OHFormation-kHOCl_OH*y(2)*y(1)-kOCl_OH*y(3)*y(1)-kDOC_OH*y(1)*y(4)-kTCE_OH*y(1)*y(5)-kHCO3_OH*y(1)*y(6)-kCO3_OH*y(1)*y(7); kforward*y(3)-kbackward*y(2)*kw/(10^-pH)-HOClPhotolysis-kHOCl_OH*y(2)*y(1); -kforward*y(3)+kbackward*y(2)*kw/(10^-pH)-OClPhotolysis-kOCl_OH*y(3)*y(1); -kDOC_OH*y(4)*y(1); -kTCE_OH*y(5)*y(1)-TCEPhotolysis-kTCE_Cl*y(5)*y(8); -kforward2*y(6)+kbackward2*y(7)*10^-pH-kHCO3_OH*y(6)*y(1); kforward2*y(6)-kbackward2*y(7)*10^-pH-kCO3_OH*y(7)*y(1); ClFormation-kTCE_Cl*y(5)*y(8)]; end close all clear all
112
clc pH=5; y0 = [0,1.55e-4/(1+10^(pH-7.52)),1.55e-4-1.55e-4/(1+10^(pH-7.52)),8.33e-6,1.10e-6,0,0,0]; tspan = [0:1:40]; [t,y]=ode15s('TCE_UV_Cl2',tspan,y0);
C.2 Matlab® Codes for Simulation of Trichloroethylene Decay by the UV/Chlorine AOP at
11 mg L–1 and pH from 5 to 10
function dydt = TCE_MP_Cl2_pHX(t,y,pH); % y(1): OH radical, y(2): HOCl, y(3): OCl–, y(4): DOC, y(5): TCE, y(6): HCO3
–, y(7):CO32–, y(8): Cl radical formed by TCE photolysis
phiHOCl_OH=1.4; phiOCl_OH=0.28; %data from Watts and Linden (2007) % phiHOCl_OH and phiOCl_OH mean quantum yields of OH radical formation by HOCl and OCl– photolysis, respectively phiHOCl=3.7, phiOCl=1.7; %data from Watts and Linden (2007) % phiHOCl and phiOCl mean quantum yields of HOCl and OCl– photolysis, respectively Ep0=[1.10E-11 1.11E-11 8.89E-12 1.08E-11 1.09E-11 8.86E-12 1.13E-11 1.10E-11 8.74E-12 1.10E-11 1.09E-11 8.87E-12 1.11E-11 1.10E-11 8.78E-12 1.13E-11 1.10E-11 8.92E-12 1.12E-11 1.11E-11 8.99E-12 1.12E-11 1.18E-11 9.12E-12 1.15E-11 1.17E-11 9.66E-12 1.18E-11 1.19E-11 1.18E-11 9.68E-12 1.20E-11 1.20E-11 9.56E-12 1.22E-11 1.23E-11 9.49E-12 1.26E-11 1.22E-11 9.92E-12 1.23E-11 1.18E-11 1.16E-11 9.50E-12 1.19E-11 1.20E-11 9.80E-12 1.41E-11 1.65E-11 1.05E-11 1.20E-11 1.22E-11 1.48E-11 1.23E-11 1.23E-11 1.34E-11 1.32E-11 1.90E-11 1.93E-11 1.87E-11 1.40E-11 1.61E-11 1.52E-11 1.29E-11 1.95E-11 2.93E-11 1.83E-11 1.13E-11 1.40E-11 1.66E-11 1.31E-11 1.39E-11 1.28E-11 1.29E-11 1.08E-11 1.67E-11 1.43E-11 1.02E-11 1.28E-11 1.66E-11 2.62E-11 1.51E-11 1.38E-11 1.27E-11 1.24E-11 1.04E-11 1.29E-11 1.25E-11 1.51E-11 1.78E-11 1.61E-11 1.29E-11 1.60E-11 1.13E-11 1.30E-11 1.35E-11 3.77E-11 2.72E-11 1.84E-11 1.40E-11 1.52E-11 2.41E-11 6.76E-11 3.26E-11 1.54E-11 1.10E-11 1.33E-11 1.31E-11 1.33E-11 1.06E-11 1.35E-11 1.78E-11 9.68E-11 9.51E-11 5.13E-11 2.05E-11 1.60E-11 1.17E-11 1.41E-11 1.36E-11 1.38E-11 1.34E-11 1.05E-11 1.32E-11 1.32E-11 1.31E-11 1.03E-11 1.32E-11 1.29E-11 1.31E-11 1.31E-11 1.06E-11 1.34E-11 2.38E-11 4.06E-11 2.02E-11 1.09E-11 1.32E-11 1.29E-11 1.35E-11 1.04E-11 1.29E-11 1.28E-11 1.27E-11 1.30E-11 1.04E-11 1.28E-11 1.29E-11 1.27E-11 1.27E-11 1.24E-11]; %Ep0 is a matrix, which represents the spectral photon irradiance at each wavelength (200–400 nm) at the centre of the Petri dish, and which corresponds to a total irradiance of 1 mW cm–2 from 200 to 400 nm Ep=5.11*Ep0; %Ep1 is the actual photon irradiance at the centre of the Petri dish used for the model E1=[7.20E+03 6.90E+03 6.60E+03 6.30E+03 6.10E+03 5.80E+03 5.70E+03 5.50E+03 5.30E+03 5.10E+03 4.80E+03 4.60E+03 4.30E+03 4.00E+03 3.70E+03 3.50E+03 3.20E+03 3.00E+03 2.80E+03 2.70E+03 2.50E+03 2.30E+03 2.10E+03 2.00E+03 1.85E+03 1.70E+03 1.55E+03 1.35E+03 1.20E+03 1.10E+03 1.00E+03 9.00E+02 8.00E+02 7.00E+02 6.00E+02 5.00E+02 4.00E+02 3.50E+02 3.00E+02 2.50E+02 2.00E+02 1.75E+02 1.50E+02 1.25E+02 1.00E+02 8.00E+01 6.00E+01 4.00E+01 2.00E+01 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100 1.00E-100]; %E1 is the spectral molar absorption coefficient of TCE at each wavelength (200 to 400 nm) E2=[91 77 66 58 53 49 46 44 43 42 42 43 44 46 48 50 52 55 58 61 64 68 71 74 77 80 83 86 89 91 93 95 96 97 98 98 98 98 97 96 95 94 92 90 87 85 82 79 76 73 70 67 64 61 58 55 53 50 47 45 43 41 39 37 36 35 33 32 31 31 30 30 29 29 29 28 28 28 28 28 29 29 29 29 29 29 29 30 30 30 30 30 30 30 29 29 29 29 28 28 28 27 27 26 26 25 25 24 24 23 22 22 21 20 20 19 18 17 17 16 15 15 14 13 13 12 12 11 10 10 9 9 8 8 7 7 7 6 6 5 5 5 4 4 4 4 3 3 3 3 3]; %E2 is the spectral molar absorption coefficient of HOCl at each wavelength (200 to 400 nm) E3=[362 325 294 267 242 220 199 180 162 144 128 113 99 86 74 63 53 45 38 32 26 22 18 15 13 11 10 9 8 7 7 7 7 7 8 8 9 10 11 12 14 15 17 19 21 24 26 30 33 37 41 45 50 55 60 67 73 80 87 95 103 111 120 130 139 149 159 169 180 190 201 212 223 234 245 255 266 276 286 295 304 312 320 327 334 339 345 349 352 355 357 358 359 359 357 355 353 350 346 342 337 331 325 318 311 304 296 288 280 271 262 253 244 235 226 217 208 199 190 181 173 164 156 148 140 133 126 118 112 105 99 93 87 82 77 72 67 62 58 54 50 47 43 40 37 34 32 29 27 25 23]; %E3 is the spectral molar absorption coefficient of OCl– at each wavelength (200 to 400 nm) A=E1*y(5)+E2*y(2)+E3*y(3); %A is the solution absorption coefficient at each wavelength (200 to 400 nm) R1=(Ep.*E1.*(1-10.^(-A*3.45)))./(A*3.45)*1000; %R1 is the specific rate of photon absorption by TCE at each wavelength (200 to 400 nm) R2=(Ep.*E2.*(1-10.^(-A*3.45)))./(A*3.45)*1000; %R2 is the specific rate of photon absorption by HOCl at each wavelength (200 to 400 nm) R3=(Ep.*E3.*(1-10.^(-A*3.45)))./(A*3.45)*1000;
113
%R3 is the specific rate of photon absorption by OCl– at each wavelength (200 to 400 nm) sumR1=sum(sum(R1)); sumR2=sum(sum(R2)); sumR3=sum(sum(R3)); %sumR means the summation of specific rate of photon absorption from 200 to 400 nm OHFormation=phiHOCl_OH*sumR2*y(2)+phiOCl_OH*sumR3*y(3); % OHFormation means OH radical formation rate HOClPhotolysis=phiHOCl*sumR2*y(2); OClPhotolysis=phiOCl*sumR3*y(3); % HOClPhotolysis and OClPhotolysis mean HOCl and OCl– photolysis rates, respectively % Assume HOClPhotolysis and OClPhotolysis are positive numbers kforward=1.8e3, kbackward=3e9; % for HOCl hydrolysis kw=1.9055e-14; ; %[OH–] = kw/10^-pH kforward2=2.2, kbackward2=5e10; % for HCO3
– hydrolysis kHOCl_OH=8.46e4; kOCl_OH=8e9; kDOC_OH=3e8; kTCE_OH=2.4e9; kHCO3_OH=8.5e6; kCO3_OH=3.9e8; phiTCE=0.354; phiTCE1=0.13; %data from Li et al. (2004) % phiTCE means the total quantum yield of TCE photolysis; phiTCE1 means the quantum yield of TCE photolysis in Reaction 1 (see Table 3.5) TCEPhotolysis=phiTCE*sumR1*y(5); ClFormation = phiTCE1*sumR1*y(5); % TCEPhotolysis means TCE photolysis rate, ClFormation means the Cl radical formation rate % Assume TCEPhotolysis is a positive number kTCE_Cl=4.88e10; dydt=[OHFormation-kHOCl_OH*y(2)*y(1)-kOCl_OH*y(3)*y(1)-kDOC_OH*y(1)*y(4)-kTCE_OH*y(1)*y(5)-kHCO3_OH*y(1)*y(6)-kCO3_OH*y(1)*y(7); kforward*y(3)-kbackward*y(2)*kw/(10^-pH)-HOClPhotolysis-kHOCl_OH*y(2)*y(1); -kforward*y(3)+kbackward*y(2)*kw/(10^-pH)-OClPhotolysis-kOCl_OH*y(3)*y(1); -kDOC_OH*y(4)*y(1); -kTCE_OH*y(5)*y(1)-TCEPhotolysis-kTCE_Cl*y(5)*y(8); -kforward2*y(6)+kbackward2*y(7)*10^-pH-kHCO3_OH*y(6)*y(1); kforward2*y(6)-kbackward2*y(7)*10^-pH-kCO3_OH*y(7)*y(1); ClFormation-kTCE_Cl*y(5)*y(8)]; end close all clear all clc tspan = [0:.1:50]; pH = 4.9; for i=1:51, %Set up a loop for different pH. i changes from 1 to 51, which makes pH changes from 5 to 10 pH=pH+0.1; y0 = [0,1.27e-4/(1+10^(pH-7.52)),1.27e-4-1.27e-4/(1+10^(pH-7.52)),5.42e-5,3.81e-8,2.88e-3/(1+10^(pH-10.36)),2.88e-3-2.88e-3/(1+10^(pH-10.36)),0]; s(i)=pH; %pH is represented by s(i) [t,y]=ode15s(@(t,y)TCE_MP_Cl2_pHX(t,y,pH),tspan,y0); %@(t,y) means ode15s solves the function dydt only considering t and y as variables, other values, like pH, are considered as parameters. Z(i)=y(20,1); %In the matrix of [y1(t1), y2(t1), y3(t1), y4(t1) % y1(t2), y2(t2), y3(t2), y4(t2) % y1(t3), y2(t3), y3(t3), y4(t3) % ...........................................] % y(20,1) means y(1) (i.e. OH radical) at 20th time (0.1s *20 = 2s) end plot(s,Z) %plot the relationship between pH and y(1) at 20th time
D. Estimation of OH Radical Concentration Using an Excel Spreadsheet Besides the Matlab® calculation, a Microsoft Excel spreadsheet can be used to estimate the
•OH concentration in the UV/chlorine and UV/H2O2 AOPs. An example for the spreadsheet is
briefly shown in Table D.1. The steady-state •OH concentration was assumed in the calculation.
114
Table D.1 Estimation of the •OH concentration generated by the UV/chlorine AOP at 11 mg L–1 and pH 5
Wavelength λ (nm)
Incident photon irradiance Ep(λ)a
(einstein cm–2 s–1)
Reflection Factorb
Water Factorc
Incident photon irradiance (adjusted by factors) (einstein cm–2 s–1)d
Cl2 molar absorption coefficient
ε (M–1 cm–1)
Solution absorption coefficient a (cm–1)e
Specific rate of photon absorption
(einstein mol–1 s–1)
[•OH]ss (M)
HOCl OCl– HOClf OCl–g Spectralh Totali A B C D E F G H I J K L
200 0 0.969 0.979 0 91 362 0.0230 0.00E+00 0.00E+00 0.00E+00 9.3E-13 201 0 0.970 0.981 0 77 325 0.0206 0.00E+00 0.00E+00 0.00E+00 202 0 0.970 0.983 0 66 294 0.0186 0.00E+00 0.00E+00 0.00E+00 203 0 0.970 0.985 0 58 267 0.0170 0.00E+00 0.00E+00 0.00E+00 204 0 0.970 0.986 0 53 242 0.0158 0.00E+00 0.00E+00 0.00E+00 205 0 0.970 0.987 0 49 220 0.0148 0.00E+00 0.00E+00 0.00E+00 206 3.36E-11 0.971 0.987 2.99E-11 46 199 0.0141 3.15E-06 1.37E-05 3.21E-15 207 3.15E-11 0.971 0.988 2.80E-11 44 180 0.0136 2.83E-06 1.16E-05 2.88E-15 208 2.04E-11 0.971 0.988 1.82E-11 43 162 0.0132 1.79E-06 6.77E-06 1.82E-15
⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ 254 5.75E-11 0.975 0.992 5.16E-11 58 60 0.0090 6.92E-06 7.18E-06 6.95E-15 255 8.28E-11 0.975 0.992 7.43E-11 55 67 0.0086 9.47E-06 1.14E-05 9.52E-15 256 1.23E-10 0.975 0.993 1.10E-10 53 73 0.0082 1.33E-05 1.85E-05 1.34E-14 257 2.02E-10 0.975 0.993 1.81E-10 50 80 0.0078 2.09E-05 3.33E-05 2.10E-14
⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ ⁞ 395 1.65E-11 0.979 1.000 1.50E-11 0.01 0.63 1.85E-06 3.45E-10 2.19E-08 4.43E-19 396 1.68E-11 0.979 1.000 1.52E-11 0.01 0.62 1.84E-06 3.50E-10 2.16E-08 4.48E-19 397 1.65E-11 0.979 1.000 1.50E-11 0.01 0.55 1.81E-06 3.46E-10 1.89E-08 4.31E-19 398 1.76E-11 0.979 1.000 1.60E-11 0.01 0.53 1.80E-06 3.68E-10 1.94E-08 4.56E-19 399 1.43E-11 0.979 1.000 1.30E-11 0.01 0.01 1.56E-06 2.99E-10 2.99E-10 3.00E-19
[HOCl] 1.54E-04 M [OCl–] 4.66E-07 M
ΦOH, HOCl 0.79 mole einstein–1 ΦOH, OCl- 1.18 mole einstein–1 kHOCl, ·OH 8.46E+04 M–1 s–1 kOCl-, ·OH 9.00E+09 M–1 s–1
aData of Column B are from Column K of Table B.1 bData of Column C are from Column F of Table B.1 cWater factor = (1 – 10–al) / [ln(10)al], based on Bolton and Linden (2003)10, where a is the decadic absorption coefficient of the solution (Column G); l is the solution depth, l = 0.795 cm for 15 mL solution contained in a Petri dish with a diameter of 4.9 cm. dColumn E = Column B × Column C × Column D × 0.9814 (divergence factor) × 0.945 (Petri factor) eColumn H = [HOCl] × Column F + [OCl–] × Column G fColumn I = Column E / Column D × Column F × (1 – 10–al) / (al) × 1000 gColumn J = Column E / Column D × Column G × (1 – 10–al) / (al) × 1000 hColumn K = (ΦOH, HOCl × [HOCl] × Column I + ΦOH, OCl- × [OCl–] × Column J) / (kHOCl, ·OH × [HOCl] + kOCl-, ·OH × [OCl–]) iThe total [[•OH]ss from 200 to 400 nm is the sum of Column K
10 Bolton, J.R., Linden, K.G., 2002. Standardization of methods for fluence (UVdose) determination in bench-scale UV experiments. Journal of Environmental Engineering, 129(3), 209–215.
115
E. UV Dose Estimation Using UVCalc® Version 2B UVClac® version 2B purchased from Bolton Photosciences Inc. was used to calculate the
UV doses delivered by the 1 kW MP UV lamp to the 40 L Lake Simcoe post-filtration water in
the Rayox® reactor for 40 s exposure. An example of the software interface associated with the
input of parameters is shown in the following screen images (Figure E.1). Since various
chlorine/H2O2 doses and pH values were applied, which changed the solution absorbances,
calculated UV doses were different in different water matires, which are summerized in Table
E.1.
116
Figure E.1 UVcalc® 2B interface and parameter selection
Table E.1 Calculated UV doses (mJ cm–2) in the Rayox® reactor
pH 6.5 pH 7.5 pH 8.5 UV/chlorine 2 mg L–1 1885 1847 1774 UV/chlorine 6 mg L–1 1813 1751 1681 UV/chlorine 10 mg L–1 1795 1663 1572 UV/H2O2 1.0 mg L–1 1881 1862 1839 UV/H2O2 2.9 mg L–1 1862 1855 1840 UV/H2O2 4.8 mg L–1 2071 1887 1835
F. Absorption Spectra of Geosmin, MIB, and Caffeine The molar absorption coefficients of geosmin, MIB, and caffeine at wavelengths from 200
to 400 nm are shown in Figure F.1. UV absorption is an important parameter that allows the
photodecomposition to take place. However, the determined molar absorption coefficients of
117
geosmin and MIB are different from those reported by Jo et al. (2011)11, Rosenfeldt et al.
(2005)12, and Kutschera et al. (2009)13, which also conflict with each other. The reason is
unknown, but may be due to the interference of methanol in the UV absorption of the water
sample (shown in Figure F.1), which was the solvent used for purchased geosmin and MIB
solutions in these previous studies. For example, the typically commercial geosmin and MIB
standard solutions are at a concentration of 100 mg L–1 in methanol. This means the methanol
molar concentration is at least 40,000 times higher than that of geosmin or MIB. Although
geosmin and MIB molar absorption are approximately 5,000 times higher than methanol, the
much higher difference in their concentrations leads to the methanol absorbance being a
significant interference in the measurement for geosmin and MIB. In this work, geosmin and
MIB solutions prepared by dissolving 50 mg pure geosmin and MIB in 500 mL Milli-Q® water
were used to simplify the solution matrices.
Figure F.1 Absorption spectra of geosmin, MIB, and methanol (amplified by 1000 times)
(left y-axis), and absorption spectrum of caffeine (right y-axis), with typical emission
spectrum of MP UV lamp
11 Jo, C.H., Dietrich, A.M., Tanko, J.M., 2011. Simultaneous degradation of disinfection byproducts and earthy-musty odorants by the UV/H2O2 advanced oxidation process. Water Research, 45(8), 2507–2516. 12 Rosenfeldt, E.J., Melcher, B., Linden, K.G., 2005. UV and UV/H2O2 treatment of methylisoborneol (MIB) and geosmin in water. Journal of Water Supply: Research and Technology– AQUA, 54(7), 423–434. 13 Kutschera, K., Börnick, H., Worch, E., 2009. Photoinitiated oxidation of geosmin and 2-methylisoborneol by irradiation with 254 nm and 185 nm UV light. Water Research, 43(8), 2224–2232.
118
G. Sample Analysis
G.1 Geosmin and MIB
Geosmin and 2-methylisoborneol (MIB) were extracted and concentrated from aqueous
samples by headspace solid phase micro-extraction (HS-SPME) and quantified using gas
chromatography-mass spectrometry (GC-MS), according to the Standard Methods 6040D
(APHA et al., 2012)14. An internal standard, d3-Geosmin, at a concentration of 100 ng L–1 was
spiked in each sample. The internal standards were used to monitor retention time, relative
response, and quantify analytes in the sample. The analysis was carried out with a Varian® 3800
Gas Chromatograph with a Varian® Ion-trap Mass Spectrometer Detector, using electron impact
(EI) ionization and an autosampler.
Samples were collected in 23 mL amber vials with Teflon®-lined septa screw caps
(headspace-free), followed by the addition of 166 μL of 25g L–1 sodium azide (NaN3) (Reagent
Plus® grade, ≥99.5%) as a preservative. Sample preparation involved first adding by pipette 10
mL Milli-Q® water for a calibration standard or a blank, or 10 mL of a sample into a 20 mL
clear vial (Supelco, Bellefonte, PA) that contained 3.5 g of reagent grade sodium chloride
(NaCl). For the calibration standard, an appropriate volume of geosmin and MIB stock solution
at 100 μg L–1 was then spiked. Internal standard (d3-Geosmin) was then added to the sample by
dispensing 100 μL of a stock solution at 10 μg L–1 in methanol to achieve a 100 ng L–1
concentration in the sample. The vial was capped with a Teflon®-lined septum magnetic crimp
cap (Supelco, Bellefonte, PA), which is manufactured for use with the autosampler. The vial
was then placed into a sample tray, where the autosampler took the sample vial and delivered it
to the spinning box. The temperature of the spinning box was preset to 65°C ± 1°C at a rotation
speed of 500 rpm. The vial was placed in the spinning box for 5 minutes to dissolve the salt. The
needle (23 gauge) containing a 1 cm long SPME fiber (Supelco, Bellefonte, PA) was then
inserted into the vial through the septum, and the fiber was extended into the vial’s headspace
for exactly 30 minutes. At the end of the contact time, the fiber was then retracted back into the
SPME holder. The needle was then inserted directly into the GC-MS, the fiber was extended
14 APHA, AWWA, WEF, 2012. Standard Methods for the Examination of Water & Wastewater, 22nd edition. American Public Health Association, American Water Works Association, and Water Environment Federation, Washington, D.C., USA.
119
into the GC/MS injection port, and the GC-MS run was started. After five minutes of
desorption, the fiber was retracted back into the SPME holder, a new sample was put into the
spinning box, and a new sample extraction process was started.
The GC-MS operating conditions are shown in Table G.1. The quantitative ions (m/z) for
geosmin, MIB, and d3-Geosmin are 112, 95, and 115, with retention times of 11.1, 9.0, and 11.1
min, respectively.
Calibration and Method Detection Limit
An example for geosmin and MIB calibration is shown in Table G.2. and Figure G.1.
Method detection limits (MDLs) for geosmin and MIB were determined to be 1.7 and 9.0 ng
L–1, respectively, based on the 8 replications at spiked concentrations of 10 ng L–1 for geosmin,
and 50 ng L–1 for MIB (shown in Table G.2).
Table G.1 GC-MS operating conditions for geosmin and MIB analysis
Parameter Description Column VF-5MS capillary column (30 m × 0.25 mm ID, 0.25
µm film thickness) Carrier gas Helium at 1 mL min–1 at 25°C Injection method Temperature: 250°C
Desorbing time: 5 min Mode: Splitless for first 2 minutes, split after 2 minutes with split ratio of 50 Split Valve: Open after 2 min, flow at 50 mL min–1 Injection Volume: 1 µL at normal speed
Autosampler method Syringe: SPME Fiber Supelco Divinylbenzene/Carboxen/Polydimethylsiloxane (DVB/CAR/PDMS), df 50/30 μm, needle size 24 ga Agitator Temperature: 65.0°C Pre-incubation time: 5 min Extraction agitation speed: 400 rpm Extraction time: 30 min
GC temperature program Initial: start from 40°C, hold for 2 min Ramp: increase to 250°C at 15°C min–1 Equilibration: hold at 250°C for 7 min
MS conditions Scan Mode: SIS (Single Ion Selection) Ionization Type: EI Emission current: 30 µAmps Scan average: 3 microscans (0.89 s per scan) Multiplier Offset: 150 volts
Total run time 23.00 min per run
120
Table G.2 Typical calibration standards for geosmin and MIB MIB Geosmin
Concentration (ng L–1) Peak area d3-Geosmin
area Ratio Concentration (ng L–1) Peak area d3-Geosmin
area Ratio
0 0 2.81E+06 0.00E+00 0 159560 2.36E+06 6.77E-02 10 40888 3.18E+06 1.29E-02 30 563147 2.78E+06 2.02E-01 50 260845 3.51E+06 7.44E-02 50 672080 2.15E+06 3.12E-01
100 615533 3.46E+06 1.78E-01 80 993562 2.08E+06 4.77E-01 100 586764 3.54E+06 1.66E-01 100 1.20E+06 2.34E+06 5.14E-01 300 2.06E+06 3.60E+06 5.73E-01 300 4.28E+06 2.71E+06 1.58E+00 500 3.39E+06 3.43E+06 9.88E-01 500 4.89E+06 1.93E+06 2.53E+00 800 5.24E+06 3.15E+06 1.67E+00 700 8.72E+06 2.53E+06 3.44E+00 1000 7.51E+06 3.33E+06 2.26E+00 1000 6.92E+06 3.27E+06 2.12E+00 1200 8.89E+06 3.40E+06 2.62E+00
MDLs MDLs 50 256476 3.34E+06 7.67E-02 10 158478 1.31E+06 1.21E-01 50 230848 3.43E+06 6.73E-02 10 176778 1.46E+06 1.21E-01 50 241160 3.71E+06 6.51E-02 10 125828 1.07E+06 1.17E-01 50 254573 3.46E+06 7.35E-02 10 187364 1.64E+06 1.15E-01 50 258391 3.41E+06 7.58E-02 10 201311 1.74E+06 1.16E-01 50 239846 3.37E+06 7.13E-02 10 198185 1.74E+06 1.14E-01 50 218141 3.42E+06 6.38E-02 10 185367 1.64E+06 1.13E-01 50 277794 3.44E+06 8.07E-02 10 221789 1.85E+06 1.20E-01
Recovery 89% Recovery 124% MDL (ng L–1) 9.0 MDL (ng L–1) 1.7
Figure G.1 Example for MIB and geosmin calibration curves
(internal standard: 100 ng L–1)
G.2 Caffeine
Caffeine samples were analyzed using a Varian 3800 gas chromatography (equipped with a
DP 1701 column, 30 m × 0.25 mm × 0.25 µm) coupled with a Varian 4000 ion-trap mass
spectrometry (GC-MS) in the positive ion chemical ionization (CI) mode, based on the method
121
described by Verenitch et al. (2006)15. The quantitative ions (m/z) for caffeine and d3-caffeine
are 195, and 198, respectively, with retention time of 17.1 min for both.
Table G.3 GC-MS operating conditions for caffeine analysis
Parameter Description Column DP 1701 capillary column (30 m × 0.25 mm × 0.25 µm) Carrier gas Helium at 1 mL min–1 at 25°C Injection method Temperature: 250°C
Mode: Splitless for first 0.5 min, split after 0.5 min with split ratio of 20 Split Valve: Open after 0.5 min, flow at 20 mL min–1 Injection Volume: 2 µL at normal speed
GC temperature program 100°C for 4 min 25°C min–1 temperature ramp to 180°C 180°C for 5 min 10°C min–1 temperature ramp to 240°C 240°C for 4 min
MS conditions Filament delay: 12 min Scan Mode: full Ionization Type: CI, reagent: methanol Emission current: 10 µAmps Scan average: 3 microscans Multiplier Offset: 300 volts
Total run time 22.20 min per run
Samples were collected in 1 L pre-cleaned amber bottles with TeflonTM-lined caps, stored
in the refrigerator (4° C), extracted within 28 days of the sampling date, and analyzed within 7
days thereafter. Samples or Milli-Q® water (for calibration standards) with volumes of 500 mL
were spiked with 1 µg L–1 d3-caffeine (as the internal standard) by adding 0.5 mL of the internal
standard stock solution at a concentration of 1 mg L–1. Samples were then extracted using
Waters Oasis hydrophilic-lipophilic balance (HLB) solid phase extraction (SPE) cartridges
(WAT106202) placed on a vacuum manifold (Visiprep, Supelco Inc.). Two blank samples
consisting of 500 mL of the water matrix examined in the experiments were spiked with internal
standard and processed alongside samples. Each of extracted samples into the cartridges was
then eluted by 6 mL methanol/MTBE mixture (1:9 v./v.). After that, 1.5 mL of each elute was 15 Verenitch, S.S., Lowe, C.J., Mazumder, A., 2006. Determination of acidic drugs and caffeine in municipal wastewaters and receiving waters by gas chromatography-ion trap tandem mass spectrometry. Journal of Chromatography A, 1116(1–2), 193–203.
122
blown-down to dryness using nitrogen, reconstituted to a volume of 150 µL in chloroform, and
then loaded onto the autosampler of the GC-MS with the operating conditions summarized in
Table G.3.
Sample extraction
• Samples with volumes of at least 600 mL are collected in 1 L amber glass bottles with
TeflonTM-lined caps. Prior to extraction, samples are refrigerated (4°C) and stored in the
dark to avoid photo-decomposition.
• Using a volumetric flask, transfer 500 mL samples to clean 1 L bottles. For calibration
standards, transfer 500 mL Milli-Q® water.
• Spike 1 µg L–1 d3-caffeine to each sample by adding 0.5 mL of the 1 mg L–1 stock solution.
• Place the appropriate number of SPE cartridges on the Visiprep manifold.
• Condition the SPE cartridges on the manifold using the following procedure. Flow should
be set at approximately 1 drop every two seconds.
• Slowly aspirate approximately 3 mL MTBE through each SPE cartridge and do not allow
the cartridges to go dry.
• Slowly aspirate approximately 3 mL methanol through each SPE cartridge and do not allow
the cartridges to go dry.
• Slowly aspirate approximately 3 mL Milli-Q® water through each SPE cartridge and do not
allow the cartridges to go dry.
• Close the valve on each cartridge.
• Attach a pre-rinsed Teflon adaptor/Teflon tube to each SPE cartridge.
• Place the free end of each Teflon adaptor/Teflon tube in a separate bottle, making sure the
tube reaches the bottom of the sample bottle. Label each SPE cartridge.
• Open the valve on each cartridge and apply vacuum (approximately -10 to -15 inHg) to the
Visiprep manifold. Flow rates through the SPE cartridges should be approximately 5–10
mL min–1. Close the value when the bottle is empty. Make sure that cartridge is not dry.
• After all samples passed completely through the cartridges, rinse each sample bottle with 30
mL of Milli-Q® water. Rinses are allowed to flow through the cartridges at approximately
5–10 mL min–1. Close the value when the bottle is empty. Make sure that cartridge is not
dry.
• Wash the cartridges with 2 ml of 25% methanol in Milli-Q® water (v./v.).
123
• Vacuum dry each cartridge by applying vacuum (approximately -10 to -15 inHg) for 2
minutes. Once all the water has been aspirated, turn OFF the manifold vacuum and remove
the Teflon adaptor/Teflon tubes from the SPE cartridges.
• Place the Teflon adaptor/Teflon tubes on Kimwipes. Remove the SPE cartridges from the
Visiprep manifold and place them on clean Kimwipes. Make sure the cartridges are labelled
properly.
• Using cotton swabs and/or Kimwipes, dry the inside of the SPE cartridges.
• Remove the Visiprep manifold cover. Dry any excess water from the underside of the
manifold cover and place the cover on clean Kimwipes.
• Allocate one 15 mL polypropylene centrifuge tube for each SPE cartridge.
• Label each 15 mL centrifuge tube with the appropriate sample identification number and
place the tube in the proper slots of the Visiprep collection rack.
• Place the Visiprep collection rack in the vacuum manifold and reseat the vacuum manifold
cover with the SPE cartridges. Check to ensure that the manifold cover exhaust tubes are
aligned with the 15 mL polypropylene centrifuge tubes.
• Add 1 mL of methanol to each SPE cartridge. Turn ON the manifold vacuum and let
methanol slowly reach the bottom of the cartridge. Turn OFF vacuum and close the valves,
soak the cartridges with methanol for 2 minutes.
• Add 6 mL methanol/MTBE mixture (1:9 v./v.) to each SPE cartridge.
• Open manifold valves and elute by gravitational force until dry. Collect the eluent in the
corresponding centrifuge tube. NOTE: Some cartridges may require slight vacuum when
beginning the elution to initiate a constant flow of methanol.
• Once all solvent appears to be eluted, turn ON the manifold vacuum for approximately 2
minutes to aspirate remaining solvent in the cartridge.
• Turn OFF the manifold vacuum. Lift the manifold cover and remove the collection rack
containing the polypropylene centrifuge tubes with the sample extracts from the manifold.
• Using a calibrated pipettor, transfer 1.5 mL of each final extract to separate 1.8 mL clean
GC vials.
• Evaporate extracts in vials to dryness using a gentle stream of nitrogen (pressure < 2 psi).
Low heat at 40°C can be applied to expidite evaporation.
124
• Reconstitute each vial with 150 μL of chloroform and cap the vial with a TeflonTM lined
septum screw cap. Rinse the walls of the vial thoroughly by gently tumbling the vial.
Transfer the concentrated extract into a vial insert and recap.
• Store in freezer (-15°C) until required for GC-MS analysis.
Calibration and Method Detection Limit
An example for caffeine calibration is shown in Table G.4 and Figure G.2. Method
detection limit was determined to be 31 ng L–1, based on the 8 replications at a spiked
concentration of 200 ng L–1.
Figure G.2 Example for caffeine calibration curves (internal standard: 1 µg L–1)
Table G.4 Typical calibration standards for caffeine Concentration (µg L–1) Peak area d3-Caffeine area Ratio
0 75771 503662 0.150 0.2 1.27E+06 4.17E+06 0.303 0.4 2.27E+06 4.83E+06 0.470 1 1.84E+06 1.71E+06 1.074 5 1.20E+07 2.26E+06 5.292 10 4.78E+06 471955 10.124 10 2.64E+07 2.64E+06 10.004 10 4.34E+07 4.18E+06 10.385 30 1.42E+08 4.79E+06 29.522 50 6.05E+06 108358 55.815
MDLs 0.2 230625 790957 0.292 0.2 248069 787763 0.315 0.2 565086 1.90E+06 0.297 0.2 269516 887596 0.304 0.2 915212 3.16E+06 0.290 0.2 625972 2.24E+06 0.279 0.2 1.04E+06 3.46E+06 0.301 0.2 64686 213594 0.303
Recovery 101% MDL (ng L–1) 31
125
G.3 Trihalomethanes (THMs), Haloacetonitriles (HANs), Haloketones (HKs),
Chloropicrin (CP), and Trichloroethylene (TCE)
Trihalomethanes (THMs), haloacetonitriles (HANs), haloketones (HKs), chloropicrin (CP),
and trichloroethylene (TCE) were analyzed by liquid-liquid extraction and gas chromatography
with electron-capture detection (GC-ECD), according to USEPA Method 551.1 (USEPA,
1995)16. A 25 mL sample or Milli-Q® water for calibration, expect TCE, was spiked with ~50
μg L–1 1,2-dibromopropane (1,2-DBP, as the internal standard) in methanol and then extracted
with 4 mL of MTBE. For the TCE samples, 10 mL aliquot of each sample was spiked with ~10
μg L–1 1,2-DBP and extracted followed the same way as that for the other compounds. One μL
of the extract was then injected into a Hewlett Packard 5890 Series II Plus GC equipped with a
DB 5.625 fused silica capillary column (30 m × 0.25 mm ID, with 0.25 µm film thickness), and
subsequently separated and detected by an ECD. The operating conditions are shown in Table
G.5.
Table G.5 GC operating conditions for THMs, HANs, HKs, CP, and TCE
Parameter Description System HP 5890 Series II Plus Column DB 5.625 capillary column Injector temperature Splitless 200°C
15 sec purge activation time Detector temperature 300°C Temperature program 35°C for 10.0 min
4°C min–1 temperature ramp to 60°C 60°C for 1.0 min 20°C min–1 temperature ramp to 110°C 110°C for 4.5 min 30°C min–1 temperature ramp to 240°C 110°C for 2.0 min
Carrier gas Helium Flow rate 24.8 cm sec–1 at 150°C Total run time 27.6 min
16 United States Environmental Protection Agency (USEPA), 1995. Method 551.1. Determination of chlorination disinfection byproducts, chlorinated solvents, and halogenated pesticides/herbicides in drinking water by liquid-liquid extraction and gas chromatography with electron-capture detection, Revision, 1.0. Available from <http://www.epa.gov/sam/pdfs/EPA-551.1.pdf> (accessed 11.09.11).
126
Sample storage
• TCE samples collected in the bench-scale experiments and DBP samples collected in the
pilot-scale experiments were not required to store, since immediately extraction to the
MTBE phase after sample collection was carried out. Sample storage was conducted in the
full-scale tests. Samples were collected in 40 mL vials (headspace-free) preceded by the
addition of 0.8 g dry salt of the buffer. Chlorine quenching agent, hydrogen peroxide
(H2O2) solution, at a concentration of 100 mg mL–1 with a volume of 20 μL was added to
the sample immediately after the collection. H2O2 concentration in the samples was
approximately 50 mg L–1. Prepare the buffer by adding 1% dibasic sodium phosphate
(Na2HPO4) to 99% monobasic potassium phosphate (KH2PO4) by weight (example: 2 g
Na2HPO4 and 198 g KH2 PO4 to yield a total weight of 200 g). The phosphate buffer is used
to lower the sample matrix pH to 4.8 to 5.5 in order to inhibit base catalyzed degradation of
the haloacetonitriles.
• Store samples in the dark at 4ºC for up to 14 days.
• Before extracting samples, remove samples from refrigerator and place them at room
temperature for 30 min.
Sample extraction
• Transfer 25 mL of a sample (or Milli-Q® water for a calibration standard or a blank) into a
clean 40 mL vial. For the calibration standard, subsequently spike an appropriate volume of
a stock solution at 10 μg mL–1 (in methanol for THMs and TCE, or in acetone for HANs,
HKs, and CP).
• Add 24 µL of the internal standard (1,2-DBP) stock solution at 50 µg mL–1 in methanol.
• Add 2 tsp of sodium sulphate (Na2SO4) using scoop to increase extraction efficiency.
• Add 4 mL of MTBE extraction solvent using a bottle-top dispenser and cap the vial using a
cap with a Teflon®-lined silicon septa.
• Shake the sample vial for approximately 30 seconds and place it on counter on its side.
Repeat and complete for all samples, blanks and standards before proceeding.
• Place all the vials upright in a rack. Let samples stand for 20 minutes for phase separation.
• Remove 2 mL from the MTBE layer, without disturbing the water layer, using a Pasteur
pipette and place it in a 1.8 mL clean GC vial without headspace. The GC vial was
127
previously added with oven baked Na2SO4 with the amount that just covers the bottom of
the GC vial. Use a clean pipette for each sample.
• Place samples in the freezer. If not analyzing immediately, samples may be kept in the
freezer for up to 21 days. Before loading onto the GC, check the appearance of Na2SO4 at
the bottom of each GC vial. If the salt lumps, this means water is present in the sample. In
this case, transfer the top portion of the sample into a new GC vial with oven baked
Na2SO4.
Table G.6 Typical calibration standards for trichloroethylene Concentration (µg L–1) Peak area 1,2-DBP area Ratio
0 0 6.0804 0 6 1.434 5.5508 0.259 10 2.220 5.4167 0.410 20 3.8856 5.0337 0.772 30 5.7235 5.2746 1.085 50 9.4544 4.772 1.981 70 9.3315 3.5521 2.627 100 17.9739 4.0345 4.456 149 32.2459 5.2264 6.170 198 43.54 5.1045 8.530
MDLs 6 1.4345 5.5508 0.258 6 1.4615 5.2567 0.278 6 1.5296 5.4308 0.282 6 1.4266 5.3629 0.266 6 1.4026 5.3185 0.264 6 1.2492 5.0891 0.245 6 1.4199 5.4948 0.258 6 1.5341 5.4693 0.280
Recovery 113% MDL (µg L–1) 1.1
Table G.7 Typical calibration standards for THMs, HANs, HKs, and CP Conc.
(µg L–1) Peak area
1,2-DBP TCM BDCM CDBM TBM TCAN DCAN DCP CP BCAN TCP DBAN 0 47.976 2.693 0 0 0 0 0 0 0 0 n.a. n.a. 2 46.809 3.002 3.951 4.218 2.768 2.342 12.156 1.785 17.353 17.868 20.956 21.467 4 51.616 4.649 8.543 8.977 5.460 5.614 26.566 4.012 17.337 17.860 20.946 21.446 6 51.843 6.087 14.262 14.724 8.420 10.961 41.971 7.122 17.323 17.855 20.937 21.435 8 50.924 7.668 21.803 21.784 11.729 14.079 52.950 9.862 17.328 17.866 20.952 21.447 10 50.237 9.226 29.167 29.020 15.152 19.022 64.208 12.496 17.322 17.861 20.945 21.437 20 51.344 16.330 57.109 58.870 29.460 47.274 146.325 29.592 17.303 17.854 20.939 21.425 30 55.145 23.078 90.602 90.236 42.685 77.828 212.409 47.322 17.297 17.857 20.943 21.424
MDLs 2 47.046 3.359 4.175 4.209 2.639 2.497 12.126 1.737 17.35 17.861 20.951 21.454 2 47.151 3.163 4.488 4.581 2.844 2.590 12.473 2.002 17.337 17.852 20.938 21.441 2 45.857 3.035 4.333 4.467 2.852 2.493 12.419 1.9004 17.339 17.854 20.938 21.444 2 48.282 3.185 4.654 4.750 2.954 2.754 12.836 2.044 17.343 17.856 20.939 21.448 2 42.078 3.401 4.326 4.371 2.706 2.891 13.054 1.994 17.353 17.868 20.954 21.463 2 47.311 3.196 4.547 4.712 2.925 2.632 12.881 1.985 17.341 17.854 20.938 21.442 2 54.280 3.498 5.135 5.272 3.252 2.959 14.049 2.296 17.348 17.867 20.958 21.458 2 48.620 3.509 5.002 5.131 3.190 2.851 13.758 2.205 17.335 17.852 20.932 21.439
Recovery 95% 91% 92% 106% 81% 105% 89% 82% 95% 82% 77% MDL
(µg L–1) 1.2 0.3 0.3 0.3 0.4 0.4 0.4 0.5 0.5 0.4 0.5
128
Calibration and Method Detection Limit
Examples for the calibration of the above-mentioned compounds are shown in Tables G.6
and G.7, and illustrated in Figure G.3. Method detection limits are also shown in these two
tables.
129
Figure G.3 Example for calibration curves for TCE, THMs, HANs, HKs, and CP
(internal standards: 10 µg L–1 for TCE, and 50 µg L–1 for others)
G.4 Haloacetic Acids (HAAs)
Nine haloacetic acids (HAA9) were analyzed by liquid-liquid extraction and gas
chromatography with electron-capture detection (GC-ECD), according to Standard Method
130
6251 B (APHA et al., 2005)17. A 25 mL sample (or Milli-Q® water for calibration) spiked with
1600 μg L–1 2,3,4,5-tetrafluorobenzoic acid (TFBA, as the internal standard) in MTBE was
acidified with 2.8 mL concentrated sulfuric acid (95–98% v./v.) and extracted with 5 mL of
MTBE. A Hewlett Packard 5890 Series II Plus gas chromatography (GC) equipped with a DB
5.625 fused silica capillary column (30 m × 0.25 mm ID, with 0.25 µm film thickness) was then
used to separate the analytes, followed by a electron-capture detector (ECD). The GC operating
conditions are shown in Table G.8.
Table G.8 GC operating conditions for HAAs
Parameter Description System HP 5890 Series II Plus Column DB 5.625 capillary column Injector temperature 200°C Detector temperature 300°C Temperature program 35°C for 10.0 min
2.5°C min–1 temperature ramp to 65°C 10°C min–1 temperature ramp to 85°C 20°C min–1 temperature ramp to 205°C 205°C for 7.0 min
Carrier gas Helium 30cm s–1 Makeup gas: 5% CH4 + 95% Ar at a flowrate of 23.1 mL min–1 Total run time 37 min
Sample storage
• Collect samples in 40 mL amber vials quenched with 8 mg sodium sulfite powder. Ensure
that samples are headspace free.
• Store dechlorinated sample at 4°C, but for no more than 9 days. Sample extracts can be held
in a freezer at -11°C for 21 days.
• Before extracting samples, remove samples from refrigerator and place them at room
temperature for 30 min.
Sample extraction
17 APHA, AWWA, WEF, 2005. Standard Methods for the Examination of Water & Wastewater, 21st edition. American Public Health Association, American Water Works Association, and Water Environment Federation, Washington, D.C., USA.
131
• Transfer 25 mL of a sample (or Milli-Q® water for a calibration standard or a blank) into a
clean 40 mL vial. For the calibration standard, subsequently spike an appropriate volume of
a stock solution at 20 μg mL–1 in MTBE.
• Add 20 µL of the internal standard (TFBA) stock solution at 2000 µg mL–1 in MTBE.
• Add 2 tsp of sodium sulphate (Na2SO4) using scoop to increase extraction efficiency.
• Use bottle-top dispensers to add 2.8 mL concentrated sulphuric acid and 5 mL of MTBE
extraction solvent in turn, and cap the vial using a cap with a Teflon®-lined silicon septa.
• Shake the sample vial for approximately 30 seconds and place it in a rack for 20 minutes for
phase separation. Repeat and complete for all samples, blanks and standards before
proceeding.
• Transfer ~1.5 mL of the extract from MTBE layer to a 1.8 mL clean GC vial, without
disturbing the water layer, using a Pasteur pipette. The GC vial was previously added with
oven baked Na2SO4 with the amount that just covers the bottom of the GC vial. Use a clean
pipette for each sample.
• Place samples in the freezer. If not analyzing immediately, samples may be kept in the
freezer for up to 21 days. Sample need to be derivatized overnight using 150 µL
diazomethane before loading onto the GC. When loading samples on the GC, check the
appearance of Na2SO4 at the bottom of each GC vial. If the salt lumps, this means water is
present in the sample. In this case, transfer the top portion of the sample into a new GC vial
with oven baked Na2SO4.
Diazomethane Production
• Set up MNNG diazomethane generation apparatus (shown in Figure G.4) in a beaker filled
with crushed ice and water.
• Add ~0.5 cm of Diazald (N-methyl-N-nitroso-p-toluenesulfonamide) to the inner tube of the
apparatus using the large end of a Pasteur pipette.
• Add methanol to the inner tube until the bulb of the inner tube is half full. Secure cap and
septum
• Add 2.6 mL of MTBE to the outer tube of the apparatus and place the apparatus in the ice
bath.
132
• Place O-ring in glass joint, position inside tube firmly on top and secure clamp. Ensure that
vapour exit hole is located on opposite side of clamp and rest clamp on spout of beaker. The
seal must be very tight to ensure maximum CH2N2 generation and recovery.
• Add 600 µL of 20% NaOH solution (100 g of NaOH in 500 mL Milli-Q® water) dropwise
to inner tube with gas tight syringe (1 drop per 5 seconds). When NaOH is initially being
added, there is a slight delay before the Diazald reacts violently so be sure to add dropwise.
Aim drops straight down into Diazald in bottom of inner tube, avoiding tube surface and
vapour exit hole. Leave syringe in place after all NaOH has been added. Removal of the
syringe will leave a hole in the septum from where CH2N2 may escape.
• Allow CH2N2 to form for 30–45 min in ice bath. MTBE will become yellow when CH2N2 is
formed.
• Transfer CH2N2 in MTBE to 15 mL vials using specially flamed Pasteur pipette and store
vials in explosion-proof freezer. The solution can be used within 2–4 weeks if possible.
• Rinse inner tube and NaOH syringe several times with Milli-Q® water.
• Rinse inner and outer tube with methanol and MTBE until glassware is clean. Put glassware
in oven at ~100°C until dry.
Figure G.4 MNNG diazomethane generation apparatus
Calibration and Method Detection Limit
An example for the calibration of HAAs is shown in Table G.9 and Figure G.5. Method
detection limits are also shown in the table.
133
Table G.9 Typical calibration standards for HAAs Concentration
(µg L–1) Peak area
TFBA MCAA MBAA DCAA TCAA BCAA DBAA BDCAA CDBAA TBAA 0 35.8296 0 0 0 0 0 0 0 0 0 2 36.5602 0.2383 2.0356 3.6594 7.98 5.7279 5.6102 1.587 0.7932 0.426 4 36.2235 0.488 3.8329 7.0241 14.1269 10.7051 10.7964 3.4509 1.2331 0.8273 6 35.4795 0.6946 5.6331 9.7997 20.5488 15.7549 16.1369 6.6554 2.556 1.33 8 35.2702 0.9442 7.4464 12.7828 28.4983 21.6905 22.5879 11.2472 4.7431 2.0227 10 36.6199 1.3107 10.017 16.4279 39.962 28.8607 30.3848 14.6404 7.5273 3.4555 20 35.4867 2.657 19.1228 29.9972 70.7059 54.2047 57.0819 39.5848 19.3598 8.7111 40 35.6578 5.1763 35.8799 54.7879 102.1244 53.0131 18.8538
MDLs 2 35.4688 0.2424 2.4333 3.9949 9.563 6.7062 6.7666 3.2167 1.469 0.2951 2 34.9655 0.2325 2.4021 4.0265 9.3444 6.5562 6.5297 3.1634 1.4594 0.3062 2 35.1744 0.2296 2.4171 3.9915 10.0212 7.4291 6.5622 3.2067 1.519 0.4432 2 35.576 0.2413 2.3725 4.0296 9.4425 6.6518 6.5133 3.2298 1.5012 0.4943 2 36.9077 0.2492 2.4766 4.1631 9.8079 6.7175 6.7302 4.2271 1.6079 0.4639 2 37.0982 0.2385 2.5654 4.1934 9.9635 6.7559 6.7187 4.2566 1.624 0.4707 2 36.6272 0.2408 2.4111 4.1871 9.7929 6.8322 6.7494 4.2643 1.6394 0.4965 2 37.8648 0.2315 2.2217 3.8122 8.5484 6.1489 5.9556 2.6197 0.9235 0.5516
Recovery 96% 124% 123% 127% 121% 113% 131% 119% 75% MDL (µg L–1) 0.2 0.4 0.3 0.5 0.5 0.4 1.3 1.2 0.9
134
Figure G.5 Example for calibration curves for HAAs (internal standards: 1600 μg L–1)
G.5 Chlorate
Chlorate (ClO3–) samples at expected concentrations higher than 50 μg L–1 were analyzed
using an ion chromatograph (IC, Model: Dionex ICS-5000+ analytical system, Thermo
135
Scientific) based on USEPA Method 300.1 (USEPA, 1997) 18 . The calibration of HAAs is
shown in Table G.10 and Figure G.6. Method detection limit was 7.2 μg L–1.
Table G.10 Calibration standards for chlorate NaClO3 concentration
(μg L–1) ClO3
– concentration (μg L–1)
Retention time (min) Peak area Regression
(μg L–1) 0 0 0 15.61 20 15.7 10.813 0.0075 32.23 50 39.2 10.92 0.0155 49.95 100 78.4 10.957 0.0335 89.82 200 156.8 10.983 0.0676 165.36 500 392.0 10.977 0.1696 391.30
1000 783.9 10.973 0.3278 741.73 2000 1567.8 10.943 0.6798 1521.45 5000 3919.6 10.887 1.7743 3945.89 MDL
50 39.2 11.003 0.0115 27.20 50 39.2 11.003 0.0117 27.66 50 39.2 11 0.0116 27.43 50 39.2 11.013 0.0139 32.74 50 39.2 11.007 0.0132 31.12 50 39.2 11.007 0.013 30.66 50 39.2 11.003 0.0145 34.12 50 39.2 11.017 0.013 30.66 50 39.2 11.017 0.0137 32.28
Recovery 78% MDL (µg L–1) 7.2
Figure G.6 Calibration curve for chlorate
H. Quality Assurance / Quality Control (QA/QC) Various QA/QC measures were undertaken to ensure analytical precision and accuracy for
all analytes, including:
• All chemicals used were analytical or high grade.
18 United States Environmental Protection Agency (USEPA), 1997. Determination of inorganic anions in drinking water by ion chromatography. Available from <http://water.epa.gov/scitech/drinkingwater/labcert/upload/met300.pdf> (accessed 09.09.14).
136
• All glassware to come in contact with samples were cleaned, by first rinsing them
thoroughly three times with hot tap water, twice with demineralized water, and twice with
distilled water, and drying them in the oven overnight at 250°C.
• The glassware to come in contact with organic solvents were cleaned, by first rinsing them
thoroughly three times with acetone, three times with MTBE, three times with methanol,
and three times with Milli-Q® water, followed by heating them in the oven overnight at
250°C. Volumetric flasks were not dried in the oven.
• Seal and store glassware in clean drawers free of all potential contamination. The lids of
glass containers were also washed with hot tap water three times and with distilled water
twice, and subsequently dried inverted at the room temperature.
• The bottles used in disinfection by-product formation potential test were prepared chlorine-
demand-free, by washing them with hot tap water and soaking them in a concentrated
sodium hypochlorite solution (~1000 mg L–1) for at least 24 h. The bottles were thereafter
rinsed thoroughly with distilled water three times and dried in the oven.
• Internal standards were used for analyses of geosmin, MIB, caffeine, and organic DBPs.
• Calibration curves for all analytes were made, by spiking known concentrations of the
analytes to Milli-Q® water. Calibration standards were prepared and run with each set of
samples. At least 7 standards were prepared, with the range of concentrations covering the
expected concentrations in the samples. The typical calibration curves are shown in
Appendix G.
• Along with each set of samples, blanks were prepared using Milli-Q® water and spiked with
internal standards to verify the interference was absent.
• Check standards at certain concentrations in the range of expected sample concentrations
were prepared at the same time and in the same way as the samples, and were analysed
every 10 samples. These check standards were then plotted on quality control charts. Figure
H.1 is an example shown the quality control charts of DBPs. The quality control charts are
used to validate and assess the accuracy of instrumental analysis. Recalibration is required
if any of the following occurs: (where M is the historical mean and SD is the historical
standard deviation). The historical mean and historical standard deviation are based on 8
standards prepared individually and analyzed consecutively.
(1) 2 consecutive measurements are outside the control limits of M ± 3 SD
137
(2) 3 out of 4 consecutive measurements are outside the warning limits of M ± 2 SD
(3) 5 out of 6 consecutive measurements are outside of M ± SD
(4) 5 out of 6 consecutive measurements follow an increasing or decreasing trend
(5) 7 consecutive measurements are above M or 7 consecutive measurements are below M
• Method detection limit (MDL) of each analyte was determined by multiplying the standard
deviation of 8 replicates near the anticipated detection limit by the Student t value (i.e.,
2.998).
• All the experiments were conducted at least in duplicate.
Quality control charts for organic DBPs are shown below.
138
139
Figure H.1 Quality control charts for organic DBPs. Lines above and below the mean (M)
represent ±1, 2 and 3 standard deviations (SD) of the samples from the mean.
140
I. Experimental Data Table I.1 Chlorine and H2O2 photolysis measured in the bench-scale trichloroethylene study (Chapter 3)
UV/chlorine pH 5 UV/chlorine pH 10 UV/chlorine pH 7.5
Fluence (J m–2)
ln(C/C0) for Cl2
Std dev Fluence
(J m–2) ln(C/C0) for Cl2
Std dev Fluence
(J m–2) ln(C/C0) for Cl2
Std dev
0 0 0 0 0 0 0 0 0 3848 -0.04 0.0056 3835 -0.22 0.008 3836 -0.10 0.020 7696 -0.11 0.0060 7669 -0.44 0.010 7672 -0.25 0.023
11543 -0.15 0.0062 11504 -0.65 0.015 9590 -0.31 0.014 15391 -0.21 0.0133 15338 -0.89 0.011 11507 -0.38 0.029 19239 -0.30 0.0072 19173 -1.14 0.041 15343 -0.52 0.025
19179 -0.63 0.009 UV/H2O2 pH 5 UV/H2O2 pH 10 UV/H2O2 pH 7.5
Fluence (J m–2)
ln(C/C0) for H2O2
Std dev Fluence
(J m–2) ln(C/C0) for H2O2
Std dev Fluence
(J m–2) ln(C/C0) for H2O2
Std dev
0 0 0 0 0.000 0 0 0.0000 0 1536 -0.0082 0.0051 3839 -0.025 0.0033 1537 -0.0070 0.0016 3073 -0.0139 0.0039 7679 -0.046 0.0034 3074 -0.0151 0.0026 4609 -0.0214 0.0040 11518 -0.065 0.0041 4612 -0.0236 0.0026 6146 -0.0279 0.0032 15358 -0.086 0.0021 6149 -0.0322 0.0048 7682 -0.0397 0.0025 19197 -0.112 0.0012 7686 -0.0366 0.0032
13444 -0.0608 0.0045 13450 -0.0615 0.0028 19206 -0.0925 0.0034 19215 -0.0935 0.0007
141
Table I.2 Trichloroethylene decay by the bench-scale UV/chlorine and UV/H2O2 AOPs
UV only pH 5 UV only pH 10 UV only pH 7.5 Fluence
(mJ cm–2) ln(C/C0) for TCE Std dev
Fluence (mJ cm–2)
ln(C/C0) for TCE Std dev
Fluence (mJ cm–2)
ln(C/C0) for TCE Std dev
0 0 0 0 0 0 0 0 0 385 -0.30 0.057 385 -0.21 0.045 385 -0.26 0.097 770 -0.56 0.053 770 -0.43 0.083 770 -0.49 0.053 1155 -0.88 0.071 1155 -0.64 0.028 1154 -0.70 0.063 1540 -1.13 0.022 1540 -0.82 0.045 1539 -0.95 0.023 1925 -1.39 0.047 1925 -1.00 0.053 1924 -1.17 0.023
UV/chlorine pH 5 UV/chlorine pH 10 UV/chlorine pH 7.5
Fluence (mJ cm–2)
ln(C/C0) for TCE Std dev
Fluence (mJ cm–2)
ln(C/C0) for TCE Std dev
Fluence (mJ cm–2)
ln(C/C0) for TCE Std dev
0 0 0 0 0.00 0 0 -0.02 0.034 51 -0.42 0.062 383 -0.30 0.056 384 -0.32 0.027 103 -0.75 0.066 767 -0.54 0.075 767 -0.63 0.056 154 -1.17 0.084 1150 -0.84 0.078 959 -0.77 0.031 205 -1.62 0.052 1534 -1.16 0.059 1151 -0.92 0.083 257 -2.10 0.034 1917 -1.39 0.032 1534 -1.31 0.045
1918 -1.55 0.036 UV/H2O2 pH 5 UV/H2O2 pH 10 UV/H2O2 pH 7.5
Fluence (mJ cm–2)
ln(C/C0) for TCE Std dev
Fluence (mJ cm–2)
ln(C/C0) for TCE Std dev
Fluence (mJ cm–2)
ln(C/C0) for TCE Std dev
0 0.00 0 0 0 0 0 0 0 154 -0.49 0.029 384 -0.44 0.037 154 -0.54 0.158 307 -1.03 0.015 768 -0.78 0.060 307 -1.01 0.120 461 -1.52 0.092 1152 -1.18 0.050 461 -1.52 0.084 615 -2.15 0.064 1536 -1.52 0.066 615 -2.06 0.079 768 -2.64 0.079 1920 -1.87 0.027 769 -2.46 0.055
142
Table I.3 Raw data of chlorine, H2O2, geosmin, MIB, and caffeine destruction in the full- and pilot-scale studies Process UV alone UV/chlorine UV/H2O2 Dose 1800–2100 mJ cm–2 2 mg L–1 6 mg L–1 10 mg L–1 1.0 mg L–1 2.9 mg L–1 4.8 mg L–1 pH 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 Cornwall full-scale in late May, 2013 Cl2/H2O2 initial (mg L–1) 2.0 6.2 10.4 0.89 2.81 4.75 Cl2/H2O2 final (mg L–1) 1.2 0.87 0.40 4.0 3.4 1.9 6.4 5.5 4.4 0.88 0.67 0.78 2.58 2.57 2.48 4.45 4.57 5.05 Geosmin initial (ng L–1) 352 Geosmin final (ng L–1) 260 273 277 48 118 155 30 89 141 13 52 93 114 144 211 59 84 123 35 56 80 MIB initial (ng L–1) 406 MIB final (ng L–1) 357 367 363 88 185 241 65 144 210 35 83 162 180 255 320 127 148 195 71 85 146 Caffeine initial (µg L–1) 22.09 Caffeine final (µg L–1) 20.21 19.29 18.53 6.43 9.66 11.96 2.88 7.15 9.94 1.66 6.21 8.19 12.49 14.24 16.76 7.38 10.23 11.98 5.04 6.09 7.91 Cornwall full-scale in early September, 2013 Cl2/H2O2 initial (mg L–1) 2.0 6.3 10.7 Cl2/H2O2 final (mg L–1) 1.0 0.70 0.34 3.5 2.1 1.5 6.6 5.3 4.7 Geosmin initial (ng L–1) 18 Geosmin final (ng L–1) 16 15 15 5 7 11 2 6 10 0 5 9 Cornwall full-scale in early April, 2014 Cl2/H2O2 initial (mg L–1) 2.2 6.0 9.8 0.91 2.76 5.07 Cl2/H2O2 final (mg L–1) 1.3 1.1 0.73 4.2 3.5 2.4 6.9 5.4 4.9 0.87 1.15 0.83 2.65 2.81 2.45 4.29 4.90 5.05 Rayox® pilot-scale Cl2/H2O2 initial (mg L–1) 2.0 2.1 2.2 6.5 6.4 6.5 10.8 10.8 11.1 0.87 0.91 0.92 2.56 2.74 2.77 4.52 4.36 4.65 Cl2/H2O2 final (mg L–1) 0.88 0.53 0.36 3.8 2.6 2.0 6.8 5.3 4.4 0.84 0.87 0.90 2.48 2.65 2.62 4.33 4.13 4.42 Caffeine initial (µg L–1) 18.35 Caffeine final (µg L–1) 15.78 15.95 15.92 5.65 8.53 10.46 2.13 5.78 8.26 0.99 4.96 6.61 10.70 12.00 12.69 7.20 7.63 9.16 4.39 5.19 6.69
143
Table I.4 Raw data of organic DBP formation (µg L–1) in the full-scale test in late May, 2013 MCAA MBAA DCAA TCAA BCAA DBAA BDCAA CDBAA TBAA TCAN DCAN BCAN DBAN AOX Initial 1.3 0 7.4 3.3 1.4 0 1.8 0 0 0 1.2 1.2 0 83 pH 7.5 Cl2 10 mg L–1 1.0 0 6.3 3.2 1.3 0 1.8 0 0 0 2.4 1.5 0 79 pH 7.5 Cl2 6 mg L–1 0.9 0 7.3 3.6 1.4 0 1.9 0 0 0 2.0 1.6 0 71 pH 7.5 Cl2 2 mg L–1 1.2 0 6.7 3.3 1.3 0 1.7 0 0 0 1.7 1.6 0 71 pH 7.5 UV 1.1 0 6.6 3.2 1.1 0 1.1 0 0 0 1.3 1.3 0 65 pH 7.5 UV/Cl2 10 mg L–1 1.7 0 9.3 3.9 1.3 0 1.4 0 0 0 2.9 3.2 0 77 pH 7.5 UV/Cl2 6 mg L–1 1.6 0 8.8 3.7 1.2 0 1.2 0 0 0 2.7 3.3 0 74 pH 7.5 UV/Cl2 2 mg L–1 1.4 0 7.7 3.4 1.1 0 1.3 0 0 0 2.3 2.7 0 76 pH 7.5 UV/H2O2 4.8 mg L–1 1.3 0 7.1 3.8 1.1 0 1.5 0 0 0 1.4 1.5 0 63 pH 7.5 UV/H2O2 2.9 mg L–1 1.8 0 9.6 3.8 1.3 0 0.7 0 0 0 1.4 1.5 0 69 pH 7.5 UV/H2O2 1.0 mg L–1 1.7 0 8.4 3.7 1.3 0 0.8 0 0 0 1.6 1.5 0 75 pH 6.5 Cl2 10 mg L–1 0.5 0 8.8 3.6 1.3 0 1.4 0 0 0 2.3 1.7 0 78 pH 6.5 Cl2 6 mg L–1 0.6 0 7.5 3.3 1.3 0 1.4 0 0 0 2.0 1.6 0 75 pH 6.5 Cl2 2 mg L–1 0.4 0 7.0 3.1 1.2 0 1.4 0 0 0 1.6 1.5 0 72 pH 6.5 UV 0.6 0 7.4 3.0 1.0 0 0.9 0 0 0 1.6 1.4 0 80 pH 6.5 UV/Cl2 10 mg L–1 2.0 0 11.7 4.0 1.1 0 1.2 0 0 0 4.4 5.3 0 89 pH 6.5 UV/Cl2 6 mg L–1 1.4 0 10.3 3.7 1.1 0 0.9 0 0 0 3.6 4.2 0 85 pH 6.5 UV/Cl2 2 mg L–1 1.4 0 10.0 3.5 1.1 0 1.0 0 0 0 2.9 3.2 0 82 pH 6.5 UV/H2O2 4.8 mg L–1 1.7 0 8.1 3.7 1.3 0 1.6 0 0 0 1.6 1.5 0 69 pH 6.5 UV/H2O2 2.9 mg L–1 1.1 0 8.0 3.5 1.1 0 1.3 0 0 0 1.5 1.5 0 69 pH 6.5 UV/H2O2 1.0 mg L–1 1.0 0 8.4 3.3 1.1 0 1.1 0 0 0 1.6 1.5 0 81 pH 8.5 Cl2 10 mg L–1 0.7 0 6.3 3.8 1.3 0 1.5 0 0 0 2.3 1.6 0 62 pH 8.5 Cl2 6 mg L–1 0.9 0 5.7 3.3 1.3 0 1.4 0 0 0 2.0 1.5 0 62 pH 8.5 Cl2 2 mg L–1 1.1 0 5.6 3.2 1.3 0 1.5 0 0 0 1.6 1.5 0 64 pH 8.5 UV 1.1 0 5.9 3.0 1.1 0 1.0 0 0 0 1.6 1.4 0 69 pH 8.5 UV/Cl2 10 mg L–1 1.4 0 6.4 3.3 1.2 0 1.2 0 0 0 2.1 2.4 0 66 pH 8.5 UV/Cl2 6 mg L–1 1.3 0 6.5 3.5 1.2 0 0.9 0 0 0 1.9 2.3 0 65 pH 8.5 UV/Cl2 2 mg L–1 1.5 0 7.1 3.1 1.1 0 0.5 0 0 0 1.3 1.8 0 70 pH 8.5 UV/H2O2 4.8 mg L–1 1.4 0 6.4 3.6 1.1 0 1.3 0 0 0 1.3 1.3 0 56 pH 8.5 UV/H2O2 2.9 mg L–1 1.3 0 6.2 3.6 1.1 0 1.2 0 0 0 1.4 1.4 0 62 pH 8.5 UV/H2O2 1.0 mg L–1 1.1 0 5.9 3.4 1.1 0 1.1 0 0 0 1.6 1.5 0 64 24 h formation potential pH 7.5 Initial 2.1 2.6 16.5 11.1 3.3 1.9 5.1 0 0 0 1.9 2.3 0 265 pH 7.5 H2O2 4.8 mg L–1 2.1 2.0 13.2 8.6 2.4 1.5 3.8 0 0 0 1.6 2.0 0 242 pH 7.5 UV 4.3 2.8 19.2 10.9 2.9 2.2 4.3 0 0 0 2.4 5.0 0 289 pH 7.5 UV/Cl2 10 mg L–1 3.3 3.2 29.5 13.6 3.5 2.6 6.4 0 0 0 2.7 7.9 0 305 pH 7.5 UV/H2O2 4.8 mg L–1 4.3 3.0 32.2 15.6 3.1 2.8 6.5 0 0 0 2.6 8.5 0 303 pH 6.5 Initial 2.0 2.9 15.8 12.4 3.7 1.8 5.7 0 0 0 2.0 2.3 0 278 pH 6.5 H2O2 4.8 mg L–1 1.9 1.8 11.6 6.3 2.4 1.3 3.4 0 0 0 1.5 1.6 0 214 pH 6.5 UV 5.0 3.5 21.9 12.7 3.1 2.1 6.2 0 0 0 3.2 4.2 0 284 pH 6.5 UV/Cl2 10 mg L–1 4.2 3.4 46.2 20.4 4.2 2.7 10.0 0 0 0 4.3 11.0 0 364 pH 6.5 UV/H2O2 4.8 mg L–1 3.5 3.2 34.4 17.5 3.8 2.6 9.4 0 0 0 3.4 8.4 0 319 pH 8.5 Initial 2.1 3.1 15.2 10.3 3.4 2.4 4.7 0 0 0 1.7 2.2 0 287 pH 8.5 H2O2 4.8 mg L–1 1.9 2.3 14.2 10.4 2.6 1.8 3.4 0 0 0 1.7 2.0 0 260 pH 8.5 UV 4.9 3.6 20.9 11.2 3.0 3.4 4.0 0 0 0 2.0 4.2 0 286 pH 8.5 UV/Cl2 10 mg L–1 4.1 3.9 26.2 13.5 3.6 3.6 6.6 0 0 0 2.0 4.9 0 287 pH 8.5 UV/H2O2 4.8 mg L–1 4.4 3.3 27.3 16.8 3.5 3.7 6.5 0 0 0 2.2 6.2 0 335
144
Table I.5 Raw data of organic DBP formation (µg L–1) in the full-scale test in early April, 2014 TCM BDCM CDBM TBM MCAA MBAA DCAA TCAA BCAA DBAA BDCAA CDBAA TBAA TCAN DCAN BCAN DBAN AOX Initial 9.2 6.5 2.0 0 0 0 4.1 2.8 1.6 0.4 0.9 0 0 0 0.6 0.7 0 46 pH 7.5 Cl2 10 mg L–1 9.2 6.6 2.0 0 0 0 4.7 3.0 1.6 0.4 1.2 0 0 0 0.6 0.7 0 41 pH 6.5 Cl2 10 mg L–1 8.7 6.8 2.2 0 0 0 5.0 3.0 1.6 0.4 0.9 0 0 0 0.7 0.8 0 44 pH 8.5 Cl2 10 mg L–1 10.1 5.1 2.1 0 0 0 4.0 2.7 1.5 0.4 1.0 0 0 0 0.6 0.8 0 46 pH 7.5 Cl2 6 mg L–1 9.2 6.2 1.9 0 0 0 4.8 2.8 1.6 0.4 0.7 0 0 0 0.5 0.7 0 46 pH 6.5 Cl2 6 mg L–1 8.7 6.6 2.0 0 0 0 5.2 2.9 1.6 0.4 1.4 0 0 0 0.6 0.7 0 41 pH 8.5 Cl2 6 mg L–1 10.2 6.7 2.1 0 0 0 4.2 2.9 1.6 0.4 1.4 0 0 0 0.5 0.8 0 43 pH 7.5 Cl2 2 mg L–1 9.3 6.3 2.0 0 0 0 4.4 2.9 1.5 0.4 1.4 0 0 0 0.7 0.8 0 47 pH 6.5 Cl2 2 mg L–1 8.8 6.5 2.1 0 0 0 5.5 2.9 1.6 0.4 1.5 0 0 0 0.6 0.7 0 50 pH 8.5 Cl2 2 mg L–1 10.4 6.8 2.1 0 0 0 4.1 2.8 1.6 0.4 1.2 0 0 0 0.6 0.7 0 45 pH 7.5 UV 9.2 6.7 1.4 0 0 0 5.3 3.4 1.5 0.3 1.0 0 0 0 0.8 1.0 0 45 pH 6.5 UV 7.8 6.6 1.5 0 0 0 5.5 3.3 1.4 0.3 0.9 0 0 0 0.9 1.0 0 45 pH 8.5 UV 9.2 6.5 1.5 0 0 0 4.5 3.0 1.4 0.3 1.0 0 0 0 0.7 0.9 0 39 pH 7.5 UV/Cl2 10 mg L–1 8.9 6.5 1.5 0 0 0 6.1 3.6 1.5 0.3 1.4 0 0 0 1.2 1.5 0 50 pH 6.5 UV/Cl2 10 mg L–1 8.1 6.1 1.5 0 0 0 7.6 3.9 1.4 0.3 1.2 0 0 0 1.7 2.4 0 57 pH 8.5 UV/Cl2 10 mg L–1 8.8 4.6 1.4 0 0 0 4.6 3.2 1.5 0.3 1.2 0 0 0 0.7 1.0 0 33 pH 7.5 UV/Cl2 6 mg L–1 8.6 6.3 1.4 0 0 0 6.5 3.6 1.5 0.3 1.5 0 0 0 1.1 1.4 0 48 pH 6.5 UV/Cl2 6 mg L–1 7.3 5.9 1.3 0 0 0 7.6 3.9 1.6 0.3 1.4 0 0 0 1.5 2.0 0 52 pH 8.5 UV/Cl2 6 mg L–1 9.0 5.8 1.5 0 0 0 5.1 3.4 1.6 0.3 1.1 0 0 0 0.7 0.9 0 39 pH 7.5 UV/Cl2 2 mg L–1 8.4 6.1 1.4 0 0 0 5.6 3.3 1.5 0.3 1.0 0 0 0 1.0 1.2 0 45 pH 6.5 UV/Cl2 2 mg L–1 7.8 6.1 1.4 0 0 0 6.3 3.4 1.5 0.3 0.8 0 0 0 1.2 1.4 0 49 pH 8.5 UV/Cl2 2 mg L–1 9.2 6.1 1.5 0 0 0 4.5 3.0 1.5 0.3 0.9 0 0 0 0.7 0.9 0 37 pH 7.5 UV/H2O2 4.8 mg L–1 8.1 5.8 1.3 0 0 0 4.8 2.9 1.4 0.3 1.0 0 0 0 0.7 0.7 0 38 pH 6.5 UV/H2O2 4.8 mg L–1 7.9 5.9 1.4 0 0 0 5.4 3.0 1.4 0.3 0.8 0 0 0 0.9 0.7 0 37 pH 8.5 UV/H2O2 4.8 mg L–1 9.3 6.2 1.7 0 0 0 4.4 3.0 1.4 0.3 1.0 0 0 0 0.6 0.7 0 34 pH 7.5 UV/H2O2 2.9 mg L–1 7.9 6.0 1.5 0 0 0 4.8 2.9 1.4 0.3 0.9 0 0 0 1.1 0.7 0 37 pH 6.5 UV/H2O2 2.9 mg L–1 7.6 6.0 1.5 0 0 0 5.5 2.9 1.4 0.3 0.9 0 0 0 1.0 0.7 0 43 pH 8.5 UV/H2O2 2.9 mg L–1 9.0 6.1 1.4 0 0 0 4.4 3.0 1.4 0.3 1.2 0 0 0 0.8 0.7 0 36 pH 7.5 UV/H2O2 1.0 mg L–1 8.6 5.9 1.3 0 0 0 4.8 2.9 1.4 0.3 0.8 0 0 0 0.9 0.8 0 42 pH 6.5 UV/H2O2 1.0 mg L–1 7.7 5.5 1.1 0 0 0 5.0 2.8 1.3 0.2 0.7 0 0 0 1.0 0.8 0 44 pH 8.5 UV/H2O2 1.0 mg L–1 9.4 6.1 1.4 0 0 0 4.1 2.8 1.4 0.3 0.9 0 0 0 0.7 0.7 0 34 24 h formation potential pH 7.5 Initial 16.1 13.2 3.4 0 0.5 0.1 12.2 12.5 3.6 0.8 3.7 0.7 0 0 1.6 2.0 0 100 pH 6.5 Initial 13.5 11.5 3.0 0 0.7 0.2 14.5 13.2 4.4 0.9 4.6 0.8 0 0 1.9 1.9 0 99 pH 8.5 Initial 19.8 16.0 4.0 0 0 0.2 11.2 10.4 3.3 0.8 3.1 0.9 0 0 1.3 1.8 0 94 pH 7.5 UV 19.2 17.7 2.6 0 1.0 0.3 18.6 14.0 3.9 0.7 3.1 1.2 0 0 2.1 4.4 0 94 pH 6.5 UV 17.1 17.0 2.3 0 0.9 0.2 18.1 14.4 4.3 0.9 3.7 0 0 0 2.8 3.8 0 87 pH 8.5 UV 24.3 21.4 3.8 0 0.8 0.3 13.2 11.5 3.6 0.6 3.2 1.7 0 0 1.9 4.4 0 88 pH 7.5 UV/Cl2 10 mg L–1 21.1 22.8 3.3 0 0.8 0.3 25.3 14.6 4.4 1.0 6.5 1.3 0 0 2.6 6.4 0 100 pH 6.5 UV/Cl2 10 mg L–1 23.4 24.4 4.1 0 0.9 0.5 35.8 24.5 5.6 1.7 9.0 0 0 0 4.4 9.3 0 101 pH 8.5 UV/Cl2 10 mg L–1 24.4 23.2 4.9 0 0.7 0.3 15.1 13.8 4.0 0.8 4.5 3.0 0 0 2.0 4.7 0 98 pH 7.5 H2O2 4.8 mg L–1 15.1 12.3 2.7 0 0.5 0.2 10.4 12.0 3.3 0.6 3.2 0.6 0 0 1.3 1.8 0 119 pH 6.5 H2O2 4.8 mg L–1 13.0 11.5 2.5 0 0.6 0.3 11.1 13.5 3.6 0.6 4.6 0.8 0 0 1.6 1.9 0 156 pH 8.5 H2O2 4.8 mg L–1 17.0 13.7 3.8 0 0 0.2 9.9 11.1 2.9 0.5 2.9 0.5 0 0 1.1 1.6 0 103 pH 7.5 UV/H2O2 4.8 mg L–1 25.6 23.4 3.2 0 0.6 0.4 30.0 19.0 4.4 1.3 6.8 1.7 0 0 2.0 7.9 0 127 pH 6.5 UV/H2O2 4.8 mg L–1 22.1 20.7 3.0 0 0.8 0.3 32.5 22.5 5.0 1.4 7.4 0.9 0 0 2.9 7.8 0 135 pH 8.5 UV/H2O2 4.8 mg L–1 31.1 22.6 4.9 0 0.7 0.4 20.2 17.0 4.6 1.1 6.2 1.8 0 0 2.2 5.2 0 119
145
Table I.6 Raw data of organic DBP formation (µg L–1) in the pilot-scale test TCM BDCM CDBM TBM MCAA MBAA DCAA TCAA BCAA DBAA BDCAA CDBAA TBAA TCAN DCAN BCAN DBAN AOX Initial for UV/H2O2 2.4 0.8 0.3 0 0 0 0.4 0.2 0.1 0 0 0 0 0 0 0 0 21 Initial for UV only 2.2 0.7 0.4 0 0 0 0.4 0.2 0.1 0 0 0 0 0 0 0 0 24 Initial for UV/Cl2 1.9 0.7 0.2 0 0 0 0.5 0.2 0.2 0 0 0 0 0 0 0 0 24 Initial for Cl2 only 2.1 0.8 0.3 0 0 0 0.4 0.3 0.1 0 0 0 0 0 0 0 0 33 pH 6.5 Cl2 10 mg L–1 5.4 3.2 1.1 0 0 0 1.4 1.3 0.2 0 0 0 0 0 0.8 0 0 47 pH 7.5 Cl2 10 mg L–1 5.7 4.5 1.3 0 0 0 1.3 1.2 0.2 0 0 0 0 0 0.7 0 0 38 pH 8.5 Cl2 10 mg L–1 4.8 3.7 1.0 0 0 0 1.1 0.9 0.2 0 0 0 0 0 0.6 0 0 38 pH 6.5 Cl2 6 mg L–1 4.3 2.9 1.0 0 0 0 1.2 1.0 0.2 0 0 0 0 0 0.6 0 0 36 pH 7.5 Cl2 6 mg L–1 5.2 4.3 1.3 0 0 0 1.2 1.1 0.3 0 0 0 0 0 0.6 0 0 38 pH 8.5 Cl2 6 mg L–1 4.5 3.4 1.0 0 0 0 1.1 0.8 0.3 0 0 0 0 0 0 0 0 37 pH 6.5 Cl2 2 mg L–1 2.3 1.8 0.7 0 0 0 0.9 0.7 0.2 0 0 0 0 0 0 0 0 35 pH 7.5 Cl2 2 mg L–1 3.0 2.4 0.8 0 0 0 1.0 0.8 0.2 0 0 0 0 0 0 0 0 42 pH 8.5 Cl2 2 mg L–1 2.8 1.6 0.4 0 0 0 0.7 0.5 0.2 0 0 0 0 0 0 0 0 34 pH 6.5 UV 2.5 0.6 0.3 0 0 0 0.4 0.2 0.1 0 0 0 0 0 0 0 0 18 pH 7.5 UV 2.3 0.7 0.3 0 0 0 0.4 0.2 0.1 0 0 0 0 0 0 0 0 19 pH 8.5 UV 2.2 0.8 0.4 0 0 0 0.4 0.2 0.1 0 0 0 0 0 0 0 0 17 pH 6.5 UV/Cl2 10 mg L–1 4.8 3.8 0.7 0 0 0.5 7.9 3.9 0.7 0.0 1.6 0 0 0 2.2 3.1 0 92 pH 7.5 UV/Cl2 10 mg L–1 4.7 3.7 1.0 0 0 0.2 5.0 2.2 0.6 0.0 0.9 0 0 0 1.3 1.9 0 71 pH 8.5 UV/Cl2 10 mg L–1 4.7 2.1 0.4 0 0 0.0 3.6 1.2 0.3 0.0 0.0 0 0 0 0.8 1.1 0 42 pH 6.5 UV/Cl2 6 mg L–1 4.4 3.6 0.8 0 0 0.6 6.8 3.1 0.8 0.0 1.4 0 0 0 1.7 2.4 0 66 pH 7.5 UV/Cl2 6 mg L–1 4.5 3.4 0.9 0 0 0.3 5.4 2.2 0.7 0.0 0.9 0 0 0 1.2 1.8 0 48 pH 8.5 UV/Cl2 6 mg L–1 3.8 1.9 0.4 0 0 0.0 3.4 1.1 0.4 0.0 0.0 0 0 0 0.8 0.9 0 34 pH 6.5 UV/Cl2 2 mg L–1 3.6 2.5 0.7 0 0 0.4 4.6 2.0 0.6 0.0 0.9 0 0 0 1.2 1.5 0 52 pH 7.5 UV/Cl2 2 mg L–1 4.5 3.2 0.9 0 0 0.3 4.8 1.7 0.6 0.0 0.0 0 0 0 1.1 1.5 0 38 pH 8.5 UV/Cl2 2 mg L–1 3.6 1.5 0.4 0 0 0.0 3.0 0.8 0.3 0.0 0.0 0 0 0 0.8 0.8 0 26 pH 6.5 UV/H2O2 4.8 mg L–1 2.4 0.7 0.3 0 0 0.0 0.5 0.2 0.1 0.0 0.0 0 0 0 0 0 0 17 pH 7.5 UV/H2O2 4.8 mg L–1 2.3 0.7 0.4 0 0 0.0 0.4 0.2 0.1 0.0 0.0 0 0 0 0 0 0 18 pH 8.5 UV/H2O2 4.8 mg L–1 2.2 0.7 0.3 0 0 0.0 0.5 0.2 0.1 0.0 0.0 0 0 0 0 0 0 21 pH 6.5 UV/H2O2 2.9 mg L–1 2.2 0.7 0.3 0 0 0.0 0.5 0.2 0.1 0.0 0.0 0 0 0 0 0 0 20 pH 7.5 UV/H2O2 2.9 mg L–1 2.2 0.7 0.3 0 0 0.0 0.4 0.2 0.1 0.0 0.0 0 0 0 0 0 0 16 pH 8.5 UV/H2O2 2.9 mg L–1 2.2 0.6 0.2 0 0 0.0 0.4 0.2 0.1 0.0 0.0 0 0 0 0 0 0 12 pH 6.5 UV/H2O2 1.0 mg L–1 2.4 0.7 0.3 0 0 0.0 0.4 0.2 0.1 0.0 0.0 0 0 0 0 0 0 14 pH 7.5 UV/H2O2 1.0 mg L–1 2.1 0.7 0.3 0 0 0.0 0.4 0.2 0.1 0.0 0.0 0 0 0 0 0 0 14 pH 8.5 UV/H2O2 1.0 mg L–1 2.4 0.8 0.4 0 0 0.0 0.5 0.2 0.1 0.0 0.0 0 0 0 0 0 0 14 24 h formation potential pH 6.5 Initial 24.0 21.7 2.6 0 0 0.7 25.4 26.1 4.0 0.4 12.4 3.2 0 0 3.7 5.1 0.2 193 pH 7.5 Initial 32.7 29.2 3.8 0 0 0.5 24.3 26.6 3.6 0.3 10.4 2.1 0 0 3.1 4.8 0.2 190 pH 8.5 Initial 45.7 37.1 6.1 0 0 0.7 19.1 19.1 4.4 0.8 7.4 3.4 0 0 1.7 0.7 0.3 190 pH 6.5 UV 25.2 28.6 2.9 0 0 0.6 28.7 21.0 3.4 0.4 8.4 2.0 0 0 3.4 8.4 0.2 196 pH 7.5 UV 37.4 37.2 4.3 0 0 0.7 26.9 19.1 3.2 0.4 8.6 2.8 0 0 2.8 6.9 0.3 19 pH 8.5 UV 52.4 46.0 6.8 0 0 0.7 17.4 14.8 3.4 0.7 5.3 2.7 0 0 1.8 1.1 0.4 204 pH 6.5 UV/Cl2 10 mg L–1 40.7 55.6 3.0 0 0 1.1 74.9 40.5 5.6 0.3 13.1 2.5 0 0 5.7 23.4 0.2 288 pH 7.5 UV/Cl2 10 mg L–1 42.3 54.1 4.2 0 0 0.7 52.3 28.3 5.4 0.5 11.8 3.2 0 0 3.5 13.2 0.2 227 pH 8.5 UV/Cl2 10 mg L–1 54.0 55.3 8.1 0 0 0.5 28.7 20.9 5.0 1.0 7.2 3.5 0 0 1.9 1.9 0.4 214 pH 6.5 UV/H2O2 4.8 mg L–1 31.8 41.2 2.8 0 0 0.5 50.1 29.8 4.8 0.4 11.4 2.4 0 0 3.6 15.4 0.2 249 pH 7.5 UV/H2O2 4.8 mg L–1 43.0 48.0 4.0 0 0 0.7 46.3 27.6 4.4 0.5 11.1 2.9 0 0 2.8 12.1 0.2 233 pH 8.5 UV/H2O2 4.8 mg L–1 58.7 54.7 6.1 0 0 0.8 29.7 22.3 4.6 0.7 5.0 2.2 0 0 1.9 2.0 0.3 231
146
Table I.7 Raw data of inorganic DBP formation (µg L–1) in the full- and pilot-scale tests Process UV alone UV/chlorine UV/H2O2 Dose 1800–2100 mJ cm–2 2 mg L–1 6 mg L–1 10 mg L–1 1.0 mg L–1 2.9 mg L–1 4.8 mg L–1 pH 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 6.5 7.5 8.5 Cornwall full-scale in late May, 2013 Chlorate initial 1.87 Chlorate final 1.71 2.44 2.20 1.65 1.90 1.65 1.65 1.76 1.68 1.96 1.84 1.66 Chlorite initial 0 Chlorite final 0 0.15 0.29 0.12 0.26 0.41 0.16 0.29 0.34 0.16 0.22 0.23 Perchlorate initial 0.05 Perchlorate final 0.04 0.05 0.06 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 Bromate initial 0.05 Bromate final 0 0.05 0 0 0 0 0 0 0 0 0.05 0 Cornwall full-scale in early September, 2013 Chlorate initial 2.63 44.0 43.5 40.2 132 124 126 219 214 231 Chlorate final 2.37 2.32 2.74 90.1 132 175 330 608 730 759 1153 1323 Chlorite initial 0 2.44 4.09 3.67 10.9 11.5 11.9 20.1 21.4 19.1 Chlorite final 0 0.14 0.28 0.79 1.30 1.90 0.25 0.12 0.26 0.15 0.15 0.16 Perchlorate initial 0.06 0.08 0.08 0.07 0.10 0.10 0.10 0.13 0.13 0.15 Perchlorate final 0.06 0.06 0.06 0.08 0.08 0.07 0.10 0.10 0.10 0.14 0.13 0.13 Bromate initial 0 0.30 0.28 0.28 0.96 0.87 0.93 1.67 1.52 1.56 Bromate final 0 0 0 0.48 0.46 0.40 2.00 1.84 1.33 3.88 2.88 2.05 Cornwall full-scale in early April, 2014 Chlorate initial 0.06 276 283 270 872 857 876 1542 1532 1506 0.06 Chlorate final 0.18 00.49 0.37 291 309 337 1045 1122 1183 1731 1865 2145 0.04 0.06 0.09 0.04 0.07 0.08 0.04 0.20 0.08 Chlorite initial 0 2.93 4.30 4.24 7.73 11.9 13.7 11.2 16.8 21.3 0 Chlorite final 0 0.67 2.11 0.93 1.19 2.47 0.36 0 0 0 0 0 0 0.30 1.04 0 0.36 0.87 0 0.18 0.32 Perchlorate initial 0.08 0.20 0.21 0.20 0.50 0.50 0.52 0.79 0.77 1.11 0.08 Perchlorate final 0.07 0.07 0.07 0.21 0.21 0.20 0.50 0.51 0.50 0.80 0.87 0.84 0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.06 Bromate initial 0.02 0.57 0.64 0.60 1.91 1.87 1.97 3.17 3.07 2.96 0.02 Bromate final 0.04 0.05 0.03 0.72 0.70 0.67 2.74 2.58 2.22 4.94 4.70 3.78 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.01 Rayox® pilot-scale Chlorate initial 0.02 284 899 1577 0.02 Chlorate final 0.03 0 0 416 450 504 1233 1403 1555 2169 2517 2718 0.04 0.01 0.01 0.01 00.01 0 0.03 0 0 Chlorite initial 0 3.92 9.92 16.8 0 Chlorite final 0 0 0.04 0.61 0.79 1.78 0.07 0.06 0.22 0.07 0.06 0.07 0 0.04 0.09 0 0 0.04 0 0 0.09 Perchlorate initial 0.01 0.09 0.31 0.54 0.01 Perchlorate final 0.01 0.01 0.02 0.13 0.13 0.13 0.36 0.37 0.36 0.68 0.60 0.60 0.01 0.01 0.02 0.02 0.02 0.03 0.01 0.01 0.02 Bromate initial 0.04 0.24 0.65 1.04 0.04 Bromate final 0.05 0.05 0.04 0.38 0.44 0.33 1.69 1.77 1.04 3.08 3.05 1.61 0.04 0.05 0.04 0.05 0.05 0.05 0.05 0.04 0.05