low pressure uv/h o treatment of the pesticides metaldehyde, … · 2019-11-08 · low pressure...

Low Pressure UV/H2O2 treatment of the pesticides

metaldehyde, clopyralid and mecoprop and matrix

constituents in drinking water

by

Sofia Semitsoglou-Tsiapou

A thesis submitted for the degree of Doctor of Philosophy

Imperial College London

Department of Civil and Environmental Engineering

2016

2

Abstract

Advanced oxidation processes (AOPs) are increasingly being applied in water treatment to

degrade recalcitrant micro-pollutants such as pesticides, pharmaceuticals, and endocrine-

disrupting compounds. This research focused on the combination of low pressure ultraviolet

light with hydrogen peroxide (LP-UV/H2O2) and determined, for the first time, (i) the

degradation, kinetics and formation of reaction products of three pesticides (mecoprop,

clopyralid, metaldehyde) selected because of their current relevance to the water sector and

different chemical structures, (ii) the formation of nitrite and nitrated compounds in waters

containing natural organic matter (NOM) and nitrate, and (iii) the mutagenicity of LP-

UV/H2O2–treated waters, both synthetic and real waters obtained from a UK water treatment

works applying LP-UV/H2O2 at full-scale.

Based on the findings of this research, it is concluded that effective degradation of the

selected recalcitrant pesticides can be achieved using practical UV fluences and H2O2 doses.

The degradation and kinetic order was mecoprop > metaldehyde > clopyralid, with the latter

being the most affected by the presence of a background matrix (NOM, nitrate, bromide), and

their main reaction products were identified, with reaction mechanisms proposed. Nitrite

formed from nitrate photolysis was directly proportional to nitrate concentration, H2O2 dose,

pH and NOM presence. Three different types of NOM were considered, but the chemical

composition of the NOM was not a significant factor in nitrite formation by LP-UV/H2O2

under the conditions tested. The regulatory limit (i.e. 0.1 µg/L NO2- in the EU) was exceeded

in the presence of NOM only at the highest fluence tested (2000 mJ/cm2). Nitrophenol

formation was observed and attributed to NOM-nitrate interactions. Mutagenic activity below

the accepted level of concern was detected by the Ames II test in both synthetic and full-scale

LP-UV/H2O2-treated waters. Overall therefore, the findings of this research support the use of

LP-UV/H2O2 in water treatment applications.

3

Statement of work

All contents in this thesis are an original work, and any previous work of others has been

appropriately referenced and acknowledged.

Copyright

The copyright of this thesis rests with the author and is made available under a Creative

Commons Attribution Non-Commercial No Derivatives licence. Researchers are free to copy,

distribute or transmit the thesis on the condition that they attribute it, that they do not use it

for commercial purposes and that they do not alter, transform or build upon it. For any reuse

or redistribution, researchers must make clear to others the licence terms of this work.

4

Acknowledgements

Firstly, I would like to express my sincere gratitude to my supervisors from Imperial College

London, Dr. Michael Templeton and Professor Nigel Graham, for guiding me all these years

through the rollercoaster called PhD! For sharing their knowledge with me, for being patient,

encouraging me all the way, for their endless corrections and recommendations and full

support from day one. I am also grateful to my Wetsus supervisor, Lucía Hernández Leal, for

her guidance and support, especially in the first year where I needed it the most.

I would like to acknowledge the financial support of Wetsus, and technical support and co-

operation of Anglian Water and TrojanUV that made this project possible. I am thankful to

all theme members for their fruitful discussions, always providing me with suggestions to

better myself, helping me out with practical matters and for always being responsive and

available.

I would also like to thank the technical team of Wetsus, Marianne, Mieke, Ton and anybody

else for helping me and carrying out analyses for me all these years, which would have

otherwise taken me ages to finish.

A big thank you to all my colleagues, and especially my office mates in both buildings;

Judita, Florian, Sławek, Johannes (old building) for welcoming me and helping me out while

I was struggling to fit in and adjust to the new country and work environment; Tim, Louis,

Jorrit, Vinu, Ilse, Tania, Natalia, Casper, Antoine, Yin (new building) for sharing fun

moments but also common troubles, lifting each other’s spirit, texting me to come back from

London and stay with them forever! Among my colleagues, I need to thank my Wetsus band

members; Pau, Victor, Ricardo, Jorrit, Prashanth, Louis, Fabian for an awesome, musical last

year, sharing our common love for music which was the best way to de-stress after long and

busy days.

I couldn’t forget my students, Astrid, Giorgos and Suyash, for contributing to my research in

so many ways, from their endless hours in the lab to writing reports, always responsible and

understanding; I honestly couldn’t have asked for better students.

One special person to whom I owe a big thank you is Ourania; a Wetsus colleague that took

me by the hand (literally) and into her house in my very first month, taught me about life

5

abroad and how to overcome obstacles, she kept me company and guided me; I cannot think

of a better start for someone who takes their first steps outside their comfort zone for the first

time.

To my non-Wetsus, Greek friends, Elena, Evangelos, Efi and Varvara: thank you for keeping

me sane in the Dutch madness, for celebrating Greek traditions together and for enjoying

complaining about the weather (and boy do we love to do that!).

I could not omit my friends from London, Christina, Vasilis, Tom, Yiran, Chrissa, Karl that

kept me company and made it so much easier to deal with the stress of relocating to London

and staying there for months at a time. They took me out, showed me London, included me in

everything!

A special thank you goes to my family, parents, brother, grandparents, uncles, all included

(real Greek family!) for their endless love and support. They were the ones that ‘’pushed’’ me

to apply and really go for it while I was being stubborn, and turns out they did well! Special

thanks to my friends in Greece that every time I went back they made me feel like nothing

changed since I left and were so proud of me, especially Andreas, Natasa F., Natasa K.,

Georgia, Tania and Eirini.

A PhD is not just being a student, is not just being an employee and presenting for

companies, is not just having fun with colleagues and friends, it is all these things combined.

One main ‘’ingredient’’ for all this to be possible is love and respect for each other, since

cooperation is an essential part of being a PhD researcher. And that is why my boyfriend,

Jimmy, deserves a big thank you; for loving me and supporting me through the last, most

demanding years of my PhD, cheering me up, putting up with my moods and schedules, for

making me feel worth of being things and aspiring me to do even more.

6

Table of Contents

1. Introduction .................................................................................................................... 20

2. Project background ........................................................................................................ 24

2.1 Basic photochemistry concepts .............................................................................. 24

2.2 UV/H2O2 Advanced Oxidation Process ................................................................. 25

2.3 Pesticide degradation by UV/H2O2 ........................................................................ 28

2.3.1 Kinetic modelling for compound degradation by UV/H2O2 ............................................................... 29

2.4 Nitrate/Nitrite issues by UV/H2O2 ......................................................................... 32

2.4.1 Nitrite formation by nitrate UV photolysis ..................................................................................... 32

2.4.2 Photochemistry of nitrate UV photolysis ........................................................................................ 34

2.4.3 Hydrogen peroxide effect on nitrite formation ............................................................................... 36

2.4.4 Natural organic matter effect on nitrite formation .......................................................................... 37

2.5 Nitrated organic product formation by UV/H2O2 ................................................ 39

2.6 Mutagenicity issues related to UV/H2O2 treatment ............................................. 40

3. Aims and Objectives ....................................................................................................... 43

3.1 Thesis Outline .......................................................................................................... 44

4. Materials and Methods .................................................................................................. 46

4.1 Chemicals ................................................................................................................. 46

4.2 Analytical methods .................................................................................................. 46

4.3 Collimated beam experiments ................................................................................ 48

4.3.1 Pesticide degradation and (in)organic product formation by LP-UV/H2O2 .................................... 49

4.3.2 Nitrite formation by LP-UV/H2O2 .................................................................................................. 51

4.3.3 Nitrated product formation ............................................................................................................. 51

4.4 Water treatment plant sampling ............................................................................ 52

4.5 Ames II Mutagenicity assay ................................................................................... 52

4.6 Quality Assurance / Quality control (QA/QC) .......................................................... 54

5. Results .............................................................................................................................. 55

5.1 Pesticide degradation and product formation by LP-UV/H2O2 ......................... 55

5.1.1 Degradation by LP-UV photolysis and LP-UV/H2O2 treatment ..................................................... 55

7

5.1.2 Kinetic study by LP-UV photolysis and LP-UV/H2O2 treatment ................................................... 59

4.1.3 Importance of OH-radicals ............................................................................................................. 63

4.1.4 Background water matrix effect ...................................................................................................... 68

4.1.5 Reaction product identification ....................................................................................................... 70

5.2 Nitrate related issues by LP-UV/H2O2 .................................................................. 77

5.2.1 Effect of initial NO3- concentration and H2O2 dose ........................................................................ 77

5.2.2 Effect of NOM presence on nitrite formation ................................................................................. 79

5.2.3 Effect of pH in the presence of NOM ............................................................................................. 83

5.2.4 Effect of NOM type on nitrite formation ........................................................................................ 85

5.2.5 Mutagenicity from NOM-NO3- interactions in synthetic water ...................................................... 90

5.2.6 Reaction product identification from NOM-NO3- interactions ....................................................... 93

5.2.7 Mutagenicity from NOM-NO3- interactions in full-scale water ...................................................... 95

6. Discussion and Conclusions ......................................................................................... 101

6.1 Pesticide degradation and product formation by LP-UV/H2O2 ....................... 101

6.2 Nitrate related issues by LP-UV/H2O2 ................................................................ 102

6.3 Mutagenicity .......................................................................................................... 103

6.4 MP versus LP AOP treatment ............................................................................. 105

6.5 DBP formation issues ............................................................................................ 107

6.6 Costs and regulations ............................................................................................ 109

6.7 Recommendations for future work ...................................................................... 111

6.8 Conclusions ............................................................................................................ 113

References ............................................................................................................................. 115

Appendix ............................................................................................................................... 132

A. Figures .......................................................................................................................... 132

B. Elemental Composition of NOMs (IHSS) .................................................................. 138

C. Quality Assurance / Quality Control (QA/QC) ........................................................ 139

D. Calculations .................................................................................................................. 155

E. Photographs ................................................................................................................. 156

8

List of Figures

Figure 2.1. Collimated beam apparatus used in this research. ................................................. 25

Figure 2.2. Reactions occurring during the UV/H2O2 process (adapted from Legrini et al.,

1993). ............................................................................................................................... 28

Figure 2.3. ONO2H/HONO2- equilibrium versus pH (adapted from KWR (2004)). ............... 35

Figure 5.1. Chemical structure of the studied pesticides. ........................................................ 56

Figure 5.2. Degradation of a) clopyralid, b) metaldehyde and c) mecoprop by LP-UV

photolysis (no H2O2) and LP-UV/H2O2 treatment. .......................................................... 58

Figure 5.3. Degradation kinetics (used for the determination of the pseudo first-order rate

constants) of (a) clopyralid, (b) metaldehyde and (c) mecoprop, by LP-UV photolysis

(no H2O2) and LP-UV/H2O2 treatment. ........................................................................... 60

Figure 5.4. Enhancement of reaction rate constants for the three pesticides with the addition

of either 5 or 15 mg/L compared to LP-UV photolysis (no H2O2). ................................. 63

Figure 5.5. Comparison of the kinetics of a) mecoprop, b) metaldehyde and c) clopyralid,

with and without the presence of a matrix, with different initial concentrations and

varying hydrogen peroxide doses. ................................................................................... 70

Figure 5.6. Mecoprop (10 mg/L) degradation and reaction product formation as a function of

the UV fluence (30 mg/L H2O2; secondary vertical axis refers to the concentration of

mecoprop (dashed line)). ................................................................................................. 72

Figure 5.7. Proposed pathways for mecoprop degradation by LP-UV/H2O2 treatment. ........ 73

Figure 5.8. Clopyralid (20 mg/L) degradation and reaction product formation as a function of


clopyralid (dashed line)). ................................................................................................. 74

Figure 5.9. Proposed pathways for clopyralid degradation by LP-UV/H2O2 treatment. ......... 75

9

Figure 5.10. Metaldehyde (5 mg/L) degradation and reaction product formation as a function

of the UV fluence (30 mg/L H2O2; secondary vertical axis refers to the concentration of

metaldehyde (dashed line)). ............................................................................................. 76

Figure 5.11. Proposed pathways for metaldehyde degradation by LP-UV/H2O2 treatment. ... 77

Figure 5.12. Effect of initial nitrate concentration in NOM-free water on nitrite formation

after LP-UV/H2O2 treatment, for an initial concentration of a) 25 mg/L, and b) 50 mg/L

NO3- as a function of the UV fluence (no bar shown for specific conditions signifies

concentrations below the method detection limit (MDL), shown with a bold line at 0.05

mg/L). .............................................................................................................................. 78

Figure 5.13. Nitrite formation after LP-UV/H2O2 treatment of nitrate (50 mg/L) in the

presence of Suwannee River NOM (4 mg/L) (no bar shown for specific conditions

signifies concentrations below the method detection limit (MDL), shown with a bold

line at 0.05 mg/L). ............................................................................................................ 80

Figure 5.14. Nitrite formation after LP-UV/H2O2 treatment of nitrate (50 mg/L) in the a)

absence and b) presence of Suwannee River NOM (4 mg/L) (no bar shown for specific

conditions signifies concentrations below the method detection limit (MDL), shown

with a bold line at 0.05 mg/L). ......................................................................................... 81

Figure 5.15. Nitrite formation for two different pH values (6 and 8) after LP-UV/H2O2

treatment of nitrate (50 mg/L) in the presence of Nordic Lake NOM (4 mg/L). LP-UV

fluence was 2100 mJ/cm2 for all H2O2 doses applied. The detection limit is shown with a

bold line (0.05 mg/L). ...................................................................................................... 84

Figure 5.16. LC-OCD fractionation: a) Suwannee River NOM, b) Nordic Lake NOM, c)

Pony Lake NOM, before and after LP-UV/H2O2 treatment (UV fluence: 2100 mJ/cm2,

H2O2 dose: 50 mg/L). Initial DOC concentrations of 2.1, 1.7 and 1.8 mg C/L for

Suwannee NOM, Nordic NOM and Pony Lake NOM, respectively. C: Carbon, N:

Nitrogen, LMW: Low Molecular Weight, DOC: Dissolved Organic Carbon, HOC:

Hydrophobic Organic Carbon. ......................................................................................... 88


presence of a) Suwannee NOM, b) Nordic Lake NOM, and c) Pony Lake NOM (pH=8)

10

(no bar shown for specific conditions signifies concentrations below the detection limit,

shown with a bold line at 0.05 mg/L). ............................................................................. 89

Figure 5.18. Number of positive wells generated by the Ames II test under different

experimental conditions with a) Suwannee River, and b) Pony Lake NOM (H2O2

concentration 15 mg/L). The positive control (2-nitrofluorene + 4-nitroquinoline N-

oxide) produced 46.9 (±0.782) positive wells. ................................................................ 90

Figure 5.19. Fold increases over a) solvent control and b) baseline calculated from the Ames

II test values for the different experimental conditions with either Pony Lake or

Suwannee River NOM (H2O2 concentration 15 mg/L). ................................................... 92

Figure 5.20. Nitrite concentrations produced by the treatment of either Pony Lake or

Suwannee River NOM (4 mg/L) in the presence of nitrate (50 mg/L) (detection limit

0.05 mg/L). ...................................................................................................................... 94

Figure 5.21. Concentrations of nitrophenol (2-nitrophenol, 4-nitrophenol or a combination of

both) produced by the LP-UV/H2O2 treatment of synthetic water samples containing

Pony Lake NOM and nitrate. The quantification and detection limits were 0.02 and

0.007 μg/L, respectively. ................................................................................................. 95

Figure 5.22. Number of positive wells generated by the Ames II test for the full-scale water

samples during different stages of treatment for the two sampling dates. The positive

control (2-nitrofluorene + 4-nitroquinoline N-oxide) produced 46.9 (±0.782) positive

wells. ................................................................................................................................ 97


II test values for the full-scale water samples during different stages of treatment for

both samplings (a) and (b). .............................................................................................. 98

Figure 5.24. LC-OCD fractionation of full-scale water samples from a) sampling (a) and b)

sampling (b). C: Carbon, N: Nitrogen, LMW: Low Molecular Weight, DOC: Dissolved

Organic Carbon. ............................................................................................................. 100

11

Figure 6.1. Ames test response in water samples (20,000 concentration factor) as a function

of the nitrite formation by MP-UV treatment at WTP Andijk and in CB experiments with

IHSS Pony Lake NOM (current results superimposed on adapted Figure 5.3 of Martijn

(2015)).................................................................................................................................... 107

Figure A.1. Normalized spectrum of the Low Pressure lamp used for the collimated beam

experiments (peak at 254 nm) (obtained by Trojan UV Technologies). ....................... 132

Figure A.2. Absorbance spectra of solutions (10 mg/L) of all six compounds in Milli-Q water

at neutral pH. .................................................................................................................. 133

Figure A.3. The main photoproduct (2-(4-hydroxy-2-methylphenoxy)propanoic acid)

produced by LP-UV photolysis and LP-UV/H2O2 oxidation treatment of mecoprop (10

mg/L). Error bars represent the standard deviation for duplicate measurements. ......... 133

Figure A.4. Conversion (%) of organic chlorine to chloride ion in aqueous solutions under

various LP-UV/H2O2 combinations for a) mecoprop (10 mg/L) and b) clopyralid (20

mg/L). Error bars represent the standard deviation for duplicate measurements. ......... 134

Figure A.5. Production of the m/z =210 product from UV/H2O2 treatment of metaldehyde (5

mg/L) as a function of UV fluence for both H2O2 doses applied. Error bars represent the

standard deviation for duplicate measurements. ............................................................ 135

Figure A.6. Acetic acid formation during LP-UV/H2O2 treatment of metaldehyde (5 mg/L).

Error bars represent the standard deviation for duplicate measurements. ..................... 135

Figure A.7. Absorption spectra of the Suwannee River, Nordic Lake and Pony Lake NOM

(concentration of 4 mg/L for all three NOMs, UV wavelength range 190-350 nm). .... 136

Figure A.8. Scheme of the drinking water treatment plant with the sampling points where

samplings (a) and (b) took place, for Ames testing purposes. ....................................... 137

Figure C.1. Clopyralid: an example of a calibration curve (on the left) denoting with an arrow

the data for which the two fragmentations were considered (on the right). .................. 140

12

Figure C.2. Metaldehyde: an example of a calibration curve (on the left) denoting with an

arrow the data for which the two fragmentations were considered (on the right). ........ 141

Figure C.3. Mecoprop: an example of a calibration curve (on the left) denoting with an arrow

the data for which the two fragmentations were considered (on the right). .................. 142

Figure C.4. Nitrophenol: an example of a calibration curve (on the left) denoting with an

arrow the data for which two fragmentations were considered. .................................... 147

Figure C.5. The LC-MS/MS chromatogram of the nitrophenols detected for the 0/15, 1500/15

and 2100/15 mJ/cm2 UV/H2O2 combinations for both duplicates (R1 and R2) performed.

....................................................................................................................................... 148

Figure C.6. Calibration curve for the nitrate ion (ion chromatography analysis). ................. 150

Figure C.7. Calibration curve for the nitrite ion (ion chromatography analysis). ................. 151

Figure C.8. IC calibration for TOC analysis. ......................................................................... 152

Figure C.9. NPOC calibration for TOC analysis. .................................................................. 153

Figure C.10. TC calibration for TOC analysis. ...................................................................... 154

Figure D.1. Calculations of maximum UV fluence required to exceed the EU limit of 0.1 mg

NO2-/L. ........................................................................................................................... 155

13

List of Tables

Table 4.1. Overview of experiments for the study of background water matrix on the kinetics

of the degradation of the three pesticides. ....................................................................... 50

Table 5.1. Physicochemical characteristics of the pesticides in this study. ............................. 56

Table 5.2. Molar absorption coefficient (εC,254) and quantum yield (φC,254) values derived

from this study and reported in literature (values from duplicates given in brackets in

case they differed). ........................................................................................................... 61

Table 5.3. Time-based pseudo first-order and second order rate constants for LP-UV

photolysis and LP-UV/H2O2 oxidation (kp, kΤ, s-1) for the three pesticides (values from

duplicates given in brackets where they differed). .......................................................... 65

Table 5.4. Second-order rate constants of the pesticides with the OH-radicals k•OH/C (M-1 s-1)

obtained from this study versus literature. ....................................................................... 67

Table 5.5. Physicochemical properties of the three NOMs used in this study. All parameters

were measured in this work, apart from the N-content (% w/w) that was obtained from

the IHSS (values from duplicates for DOC are given in brackets). ................................. 85

Table 5.6. Water treatment-related parameters for the two sampling dates. ........................... 96

Table C.1. Standard concentrations for pesticides. ................................................................ 143

Table C.2. Standard and blanks responses for internal standard (FNPF). ............................. 143

Table C.3. Standard and blanks responses for clopyralid. ..................................................... 144

Table C.4. Standard and blanks responses for metaldehyde. ................................................. 145

Table C.5. Standard and blanks responses for mecoprop. ..................................................... 146

14

Table C.6. Standard concentrations for nitrophenol (NP). .................................................... 149

Table C.7. Standard and blanks responses for nitrophenol (NP). .......................................... 149

15

Nomenclature

Kinetic study symbols

Symbol Quantity name Dimension

[OH]ss

steady state concentration of

OH-radicals formed via

H2O2 photolysis

M

E254 incident photon irradiance in

the centre of the dish Ein/cm2 s

k∙OH,H2O2

reaction rate constant

between hydrogen peroxide

and OH-radicals

M-1 s-1

k•OH/C


between a compound and

OH-radicals

M-1 s-1

k•OH/pCBA


between pCBA and OH-

radicals

M-1 s-1

kH2O2,254

specific rate of light

absorption of hydrogen

peroxide at 254 nm

Ein/mol s

kox OH-radical oxidation rate

constant s-1

kp photolysis rate constant s-1

ks,c,254

specific rate of light

absorption of a compound at

254 nm

Ein/mol s

rc reaction rate of the organic

compound M-1 s-1

rpCBA reaction rate of pCBA M-1 s-1

16

Uλ,254 molar photon energy at 254

nm J/Ein

z solution depth cm

Kinetic study Greek symbols

Symbol Quantity name Dimension

α254 absorbance of solution at

254 nm cm-1

εC,254 molar absorption coefficient

of a compound at 254 nm M-1 cm-1

εH2O2,254 molar absorption coefficient

of H2O2 at 254 nm M-1 cm-1

φC,254 quantum yield of a

compound at 254 nm mol/Ein

φH2O2,254 quantum yield of hydrogen

peroxide at 254 nm mol/Ein

Abbreviations

Abbreviation Meaning

AOP Advanced Oxidation process

BAC Biological Activated Carbon

CSF Coagulation-Sedimentation-

Filtration

DBE Double Bond Equivalent

17

DBPs Disinfection By-Products

(M)DL (Method) Detection Limit

DOC Dissolved Organic Carbon

DON Dissolved Organic Nitrogen

EEM Excitation-Emission Matrix

EDC Endocrine Disrupting Compound

EDI Estimated Daily Intake

EOC Emerging Organic Contaminant

EPA Environmental Protection Agency

FP Formation Potential

GAC Granular Activated Carbon

HAAs Haloacetic Acids

HOC Hydrophobic Organic Carbon

HS Humic Substance

IARC International Agency for

Research on Cancer

IC Ion Chromatography

IR Infrared

LC-MS/MS Liquid Chromatography - Mass

Spectrometry

18

LC-OCD Liquid Chromatography -

Organic Carbon Detection

LMW Low Molecular Weight

LP Low Pressure

MLD Million Litres per Day

MP Medium Pressure

NMR Nuclear Magnetic Resonance

NOM Natural Organic Matter

PAH Polycyclic Aromatic

Hydrocarbon

PCB Polychlorinated biphenyls

pCBA Para-chlorobenzoic acid

PCDDs Polychlorinated dibenzo-p-

dioxins

PCDF Polychlorinated dibenzofurans

PCP Personal Care Products

PFCs Perfluorinated chemicals

QL Quantification Limit

QqQ Triple Quadrupole

QSAR Quantitative Structure–Activity

Relationship

SEC Size Exclusion Chromatography

19

SPE Solid Phase Extraction

SUVA Specific Ultraviolet Absorbance

THMs Trihalomethanes

TOC Total Organic Carbon

TOF Time-Of-Flight

UHPLC Ultra-High Pressure Liquid

Chromatography

UV Ultraviolet

UVT Ultraviolet Transmittance

WHO World Health Organization

WTW Water Treatment Works

20

1. Introduction

Emerging organic contaminants (EOCs) consist of a wide variety of organic compounds, as

well as metabolites and transformation products, including pesticides, pharmaceuticals,

personal care products (PCPs), endocrine disrupting compounds (EDCs) and industrial

compounds (Lapworth and Gooddy 2006, Houtman 2010). Conventional water treatment

technologies (e.g. coagulation-sedimentation-filtration, CSF) may not provide sufficient

removal of EOCs in all cases, even when followed by additional treatment steps, such as

granular activated carbon (GAC) filtration, which is ineffective at removing polar

compounds. Ozonation, another widely used oxidation treatment, leads to selective

degradation of some compounds but also results in bromate formation in bromide-rich waters

(von Gunten and Hoigne 1994, von Gunten and Oliveras 1998, Ratpukdi et al. 2011).

In view of ozone-recalcitrant compounds (e.g. pesticides such as metaldehyde and clopyralid)

and the potential for inorganic reaction products, advanced oxidation processes (AOPs) are

being applied as alternatives to the conventional treatment methods. AOPs are based mainly

(but not exclusively) on the production of highly reactive, short-lived ·OH radicals, which

react rapidly and non-selectively with most of the target compounds (Haag and Yao 1992,

Rosenfeldt et al. 2006, Comninellis et al. 2008) with high reaction rate constants, usually in

the 109-1010 M-1 s-1 range (Andreozzi et al., 1999).

The growing interest in AOPs is reflected through the increasing research on their application

as alternative treatment for industrial and hazardous effluents, removal of organic

micropollutants and pathogens. In particular, they have been proven to act very effectively

towards micro-pollutants that are recalcitrant to conventional treatment methods (Swaim et

al. 2008, Vilhunen and Sillanpaa 2010). Nevertheless, high capital and operational costs

related to AOP applications due to costly chemicals (e.g. H2O2), energy consumption, severe

treatment conditions related to heavily contaminated influent water etc. render AOPs in many

cases the last resort. Another important consideration in AOP applications should be the DBP

formation (e.g. trihalomethanes (THMs) and haloacetic acids (HAAs)) upon subsequent

chlorination; the hydrophobic part of NOM is found to be the mostly contributing one to that

direction, when it is not effectively mineralized by the AOP (Lin and Wang, 2011). Even so,

the imperative for companies to meet regulation standards, for example for recalcitrant

pesticides, as well as the unsustainable pest control strategies due to reasons like ignorance or

21

heavy initial costs to switching to more sustainable systems (Wilson and Tisdell, 2001), along

with the benefits of the AOPs compared to conventional treatment methods, justify their

growing use for (waste) water applications.

The combination of monochromatic low pressure (LP) ultraviolet (UV) light and hydrogen

peroxide (H2O2) (UV/H2O2) is one of the AOPs based on the ·OH-radical production and was

the AOP studied in this research. This AOP has been shown in numerous previous studies to

successfully treat various organic contaminants. Studies on UV photolysis and UV/H2O2

treatment have been carried out for selected compounds widely detected in surface waters –

e.g. endocrine disrupting compounds (EDCs) (Rosenfeldt and Linden, 2004), polycyclic

aromatic hydrocarbons (Beltrán et al. 1996, Ledakowicz et al. 1999), pharmaceuticals

(Ikehata et al., 2006) and pesticides (Burrows et al. 2002, Kowalska et al. 2004). In

comparison with polychromatic medium pressure (MP) lamps, the advantages of using LP-

UV include the higher energy efficiency and longer lifetime, lower energy consumption and

minimum formation of by-products of concern, such as nitrite (potentially formed in some

MP applications). Bromate formation, a by-product of ozonation, is also not expected to be

formed with the UV/H2O2 application.

In practice, the UV/H2O2 process is already being successfully used in drinking water

treatment applications. One example is Hall Water Treatment Works of Anglian Water, one

of the largest suppliers of drinking water and wastewater services in the United Kingdom,

which includes a low pressure (LP)-UV/H2O2 process as part of a multi-barrier treatment

approach, since mid-2015. The site draws water from the River Trent and produces 20

million litres per day (MLD) of drinking water. Another example is the Andijk III plant from

PWN Technologies in the Netherlands, which draws water from IJssel Lake and utilizes MP-

UV/H2O2 as its oxidation step to produce 120 MLD of drinking water.

Regarding pesticides, although water treatment processes such as GAC filtration and/or

ozonation are effective barriers for their removal and degradation, this is not the case for all

(e.g. metaldehyde, clopyralid) because of their polarity and chemical structure (Cooper,

2011). Many pesticides are chemically stable, toxic, and non-biodegradable and may also be

resistant to direct decomposition by sunlight (Gill and Garg, 2014). Pesticide residues persist

in the environment and may pose a risk to both ecosystems and human health. This research

focused on three particular pesticides, metaldehyde, clopyralid and mecoprop, because of

their differences in susceptibility to degradation by LP-UV photolysis and hydroxyl radical

22

oxidation, their presence in European water bodies and the scarce information on their

degradation by LP-UV/H2O2 AOP in the literature, especially for clopyralid and

metaldehyde. On a risk basis, the mere presence of these pesticides does not compromise the

quality of the water and does not pose a health threat; nevertheless, in the European Union,

the regulation that applies for pesticides in drinking water is non-selective and sets the limit

to 0.1 μg/L and 0.5 μg/L, for each individual pesticide and for the total concentration of

pesticides, respectively. Therefore, drinking water suppliers are obligated to comply with

these directives.

The importance of the water matrix for competition of UV light and hydroxyl radical

scavenging on influencing the performance of the AOP process regarding the degradation of

the target micro-pollutants has been studied previously (Rosenfeldt et al. 2006, Autin et al.

2013). In relation to the formation of metabolites from the break-down of micropollutants,

such as pesticides, both their potential toxicity compared to the parent compound, as well as

the mechanism behind the formation are important aspects to be considered, since such

findings could contribute to the basic understanding and knowledge required in treating other

micro-pollutants that might arise in the future with properties and structure similar to the ones

of the compounds studied.

Nevertheless, a less examined issue, the formation of reaction products from the background

water matrix, stemming either from the organic part, the inorganic part or combination of the

two, requires consideration as a potential public health concern. Two main inorganic products

related to health concerns are bromate and nitrite (WHO, 2011). Nitrite is the main DBP

(disinfection by-product) detected in nitrate-containing waters by (MP-)UV-based AOPs; its

formation by nitrate UV photolysis has been studied (Mark et al. 1996, Mack and Bolton

1999) but parameters affecting this complicated process have had either conflicting results

(presence of hydrogen peroxide) or have been studied to a lesser extent (presence and

composition of Natural Organic Matter, NOM). Furthermore, the intermediate radicals

generated from the nitrate/nitrite photolysis can react with the organic matrix of the water and

result in the formation of potentially harmful nitrated compounds, an issue that has only

recently been investigated (Martijn et al. 2014, Kolkman et al. 2015).

The first part of the research investigated the suitability of the LP-UV/H2O2 AOP for the

degradation of the three selected pesticides (metaldehyde, clopyralid and mecoprop) by

evaluating the comparative degradation kinetics for LP-UV photolysis and hydroxyl radical

23

oxidation, the formation of the major reaction products and the possible reaction pathways for

their formation. The second part investigated issues associated with treating nitrate-rich

waters, such as the River Trent (containing typically 40-50 mg/L of nitrate). The effect of

three different types of natural organic matter (NOM) on nitrite formation from nitrate-rich

water by LP-UV photolysis and LP-UV/H2O2 treatment and the relevance of the nitrite

formation under standard and more extreme conditions of LP-UV photolysis and LP-

UV/H2O2 treatment were evaluated. Reaction product formation by the interactions of NOM

and nitrate by LP-UV/H2O2 was investigated and the mutagenicity of synthetic and real water

samples treated by LP-UV/H2O2 was evaluated.

24

2. Project background

2.1 Basic photochemistry concepts

The two laws of photochemistry are the foundation for understanding photochemical

transformations under the influence of visible and/or ultraviolet light. The first law of

photochemistry, the Grotthus-Draper Law, states that a photochemical reaction occurs only

when light is absorbed by a molecule. The absorbed light must have enough energy to

dissociate the weakest bond of the molecule. The molar absorption coefficient (ε) is a

measure of a molecule’s ability to absorb a photon of a specific wavelength, as described in

Lambert-Beer’s law (equation 1):

Abs = (Iout/Iin) = ε c l (1)

where Iout is the intensity of light transmitted, Iin the incident intensity, ε the molar absorption

coefficient (L/mol cm), c the concentration of the absorbing species (mol/L) and l the

pathlength (cm) (Wardle, 2009).

The second law of photochemistry, the Stark-Einstein Law, states that for each photon of

light absorbed during a photochemical reaction, only one molecule is excited. However, this

does not mean that every photon absorbed by a molecule leads to a photochemical reaction.

Quantum yield is the term used to define the efficiency with which absorbed light produces

some effect. It is defined as the number of molecules of reactant consumed per photon of

light absorbed (Wardle, 2009).

Another important concept of photochemistry is irradiance. According to Bolton and Linden

(2003) it is defined as the “total radiant power incident from all upward directions on an

infinitesimal element of surface of area dA containing the point under consideration divided

by dA” and it is the “appropriate term when a surface is being irradiated by UV light coming

from all directions above the surface”.

In the studies involving the degradation of organic compounds, it is necessary to calculate the

UV fluence necessary for a specific removal percentage of a compound, as well as the

irradiation time needed to achieve that. UV fluence-response curves are constructed, with this

response usually determined in a bench scale apparatus (a scheme of the apparatus used in

this work is given in Figure 2.1) often referred to as a ‘collimated beam’, where the UV light

25

beam is directed on the surface of the sample via a cylindrical tube called ‘collimator’. The

beam is never truly collimated, since there is always some dispersion, which should be

considered when long water path lengths are used (Bolton and Linden, 2003).

Figure 2.1. Collimated beam apparatus used in this research.

2.2 UV/H2O2 Advanced Oxidation Process

In general, oxidation processes which are based on the generation of hydroxyl radical

intermediates are termed Advanced Oxidation Processes (AOPs). AOPs are an attractive

technology because the reaction rate of target compounds with hydroxyl radicals is often

several orders of magnitude higher than with any conventional oxidant. The reason for this is

the second largest redox potential of the OH-radical (2.80V at 25◦C) after fluorine (3.03V at

25◦C) and over the second choice for oxidation, O3 (2.07V at 25

◦C). AOPs make use of

different reacting systems, including photochemical degradation processes (UV/O3,

UV/H2O2, and H2O2/O3/UV), photocatalysis (TiO2/UV, photo-Fenton reagent), and chemical

oxidation processes (O3, O3/H2O2, H2O2 /Fe2+) (Legrini et al., 1993).

UV-based AOPs have been suggested as an effective way of treating water with organic

micro-pollutants and have recently been introduced in drinking water treatment practice. The

application of this technology is evaluated by examining the degradation efficacy for a wide

range of pollutants. Nowadays, two types of UV light sources are utilized (Schalk et al.,

2006):

26

1. Low Pressure (LP-UV) Lamps – They may be either low intensity or high intensity

lamps and they emit UV light only at a wavelength of 253.7 nm (usually referred to

as 254 nm).

2. Medium Pressure (MP-UV) Lamps – High intensity lamps that emit energy within

the 200-400 nm range.

The degradation of organic compounds takes place via two parallel mechanisms, photolysis

from the UV light and oxidation from the OH-radicals. Regarding the photodegradation of

pollutants in water by UV light three main mechanisms are followed:

i. Direct photodegradation (photolysis): direct excitation of the organic compound by

UV light.

ii. Photooxidation: light-driven oxidation processes principally initiated by hydroxyl

radicals.

iii. Photoreduction: light-driven reduction processes principally initiated by hydrated

electrons (Helz et al., 1994).

When UV light is coupled with hydrogen peroxide, the cleavage of the H2O2 molecule takes

place via UV photolysis with a quantum yield of two OH-radicals formed per quantum of

radiation absorbed (reaction r1). Photons with wavelengths less than or equal to 560 nm have

enough energy to split the O-O bond in H2O2 (Bolton, 2004). However, H2O2 absorbs

negligibly at wavelengths above 300 nm and even at 254 nm (the main emission of a Low

Pressure UV lamp) the molar absorption coefficient of H2O2 is 19.6 L/mol cm (Baxendale

and Wilson, 1956).

H2O2 + hv ( λ ≤ 300 nm) → 2 · OH (r1)

The rate of photolysis of aqueous H2O2 has been found to be pH dependent and increases

when more alkaline conditions are used. This might be primarily due to the higher molar

absorption coefficient of the peroxide anion (HO2-) at 254 nm (ε254=240 M-1 cm-1), produced

via the reactions (r2)-(r4) and photolyzed according to reaction (r5) (Pera-Titus et al., 2004).

H2O2 + · OH → H2O + HO2 · (r2)

HO2 · → H+ + O2

· − (r3)

HO2 · + O2· − → HO2

− + O2 (r4)

27

HO2− + hv →· OH + O· − (r5)

As stated previously, the OH-radicals generated are extremely potent oxidizing agents that

attack the organic compounds relatively non-selectively. Reactions of hydroxyl radicals with

organic substrates fall into the three following categories: hydrogen abstraction (r6),

electrophilic addition (r7) and electron-transfer reaction (r8), with the first one being the

major mechanism in most cases (Ikehata and El-Din, 2006).

RH + · OH → H2O + R · (r6)

R2C = CR2 + · OH → R2C. − CR2OH (r7)

· OH + RX → OH− + RX· + (r8)

The sequence of reactions occurring during the UV/H2O2 process used for the oxidation of

organic substrates is shown in Figure 2.2. Starting from the hydrogen peroxide photolysis

leading to the generation of hydroxyl radicals (a) that react with organic compounds (HRH)

mainly by hydrogen abstraction (b), various reactions take place subsequently, including

organic radical formation (c), generation of organic cations as well as superoxide anion (d)

and disproportionation of the superoxide anion to yield H2O2 (i). The importance of oxygen

saturation becomes obvious via these reactions depicted, since without oxygen “organic

radicals will initiate polymerization of unsaturated organic substrate present in the reaction

system or generated by dismutation”. This chain of reactions under specific conditions can

eventually lead to mineralization of organics into carbon dioxide, water, and inorganic ions

(Legrini et al., 1993).

28

Figure 2.2. Reactions occurring during the UV/H2O2 process (adapted from Legrini et al.,

1993).

2.3 Pesticide degradation by UV/H2O2

Various pesticides and herbicides have been previously studied by AOPs; e.g. chlorophenoxy

acid herbicides and aromatic pesticides by direct photolysis and Low Pressure-UV/H2O2

(Aaron and Oturan, 2001), acetamide herbicides (Benitez et al., 2004), phenyl-urea herbicides

(Benitez et al., 2006). An extensive review about the treatment of pesticides and herbicides,

among other aspects, is given by Buchanan et al. (2009).

This research focused on the study of three recalcitrant pesticides, mecoprop, clopyralid and

metaldehyde. Mecoprop ((R,S) 2-(2-methyl-4-chlorophenoxy)-propionic acid), a

chlorophenoxy herbicide, developed circa 1956, is commonly applied to control a variety of

weeds and is found in groundwater wells and abstractions in many areas around Europe

(University of Hertfordshire, 2015). Clopyralid (3,6-dichloro-2-pyridine-carboxylic acid) is

used to control broadleaf weeds in certain crops and turf. Its chemical stability along with its

mobility enables penetration through the soil, causing a long term contamination of

groundwater as well as surface water supplies (Tizaoui et al., 2011). Both mecoprop and

clopyralid are frequently detected in drinking water (Donald et al., 2007). Metaldehyde

29

(2,4,6,8-tetramethyl-1,3,5,7-tetraoxocane) is a contact and systemic molluscicide bait for

controlling slugs and snails. In 2009, the UK Drinking Water Inspectorate (DWI) Annual

Reports for drinking water quality in England and Wales reported that metaldehyde was

responsible for one third of the 1103 water quality failures, since it is not sufficiently

removed by GAC (Granular Activated Carbon) filtration or degraded by ozonation (Drinking

Water Inspectorate, 2015).

Specifically for the pesticides studied in this work, their treatability by various AOPs has

been investigated previously but no studies have been reported so far concerning LP-

UV/H2O2 treatment, apart from one study on metaldehyde degradation by Autin et al. (2012).

Meunier and Boule (2000) and Boule et al. (2002) studied the photo-transformation of

aromatic pesticides, including mecoprop. They reported that the photo-transformation of

mecoprop yielded a number of photo-products, mainly by heterolytic photo-hydrolysis, was

pH-dependent and was not influenced by oxygen or UV light in the wavelength range of 254-

310 nm. Topalov et al. (1999) studied mecoprop degradation by MP-UV/TiO2 treatment and

proposed radical reactions resulting into a hydroxylated/dechlorinated aromatic moiety and

acetic acid as the main products. Šojić et al. (2009) proposed pathways of clopyralid

degradation by medium pressure (MP)-UV/TiO2 treatment suggesting radical reactions and

hydroxylation of the ring, whereas Xu et al. (2013) applied MP-UV/H2O2 treatment resulting

in dechlorination and formation of further oxidation products. Autin et al. (2012) reported on

the degradation of metaldehyde by LP-UV/H2O2 and LP-UV/TiO2 treatment but did not

include any details of reaction product formation. Moriarty et al. (2003) proposed

mechanisms for the reaction of cyclic ethers with OH-radicals that could apply to

metaldehyde as well.

2.3.1 Kinetic modelling for compound degradation by UV/H2O2

The overall degradation of a contaminant in a UV/H2O2 process is described by equation (2),

based on the kinetic model given below (Sharpless and Linden 2003, Lester et al. 2010,

Baeza and Knappe 2011). It involves the sum of the direct photolysis rate constant (kp, s-1)

and the OH-radical oxidation rate constant (kox, s-1):

d[C]

dt= −(kp + kox)[C] (2)

The direct photolysis rate constant is described by:

30

kp = ks,c,254 x φc,254 (3)

where φC,254 (mol/Ein) is the quantum yield of the compound at 254 nm and ks,C,254 (Ein/mol

s) is the specific rate of light absorption of the target compound at 254 nm, given by the

following equation:

ks,C,254 =1000 E254 εC,254 [1−10

−a254 z]

a254 z (4)

where εC,254 (M-1 cm-1) is the molar absorption coefficient of the compound at 254 nm, z is

the solution depth (equal to 1.96 cm for the experiments in this study), 𝑎254 (cm-1) the

absorbance of the solution (comprised of the compound, the hydrogen peroxide and any

water background present) given by equation (5):

α254nm = awater backgr + εH2O2,254[H2O2] + εC,254[C] (5)

and E254 (Ein/cm2 s) is the incident photon irradiance calculated from the irradiance in the

centre of the dish given as a radiometer reading (E, in mW/cm2) and the Uλ,254 factor

representing the energy of 1 Ein of photons at 254 nm and is equal to 471528 J/Ein (Bolton

and Stefan, 2002):

E254 =E

Uλ,254 (6)

The indirect photolysis rate constant is described by:

kox = kOH/C [OH]ss (7)

where k∙OH/C is the second-order rate constant for the reaction between the OH-radicals and

the compound and [OH]ss the steady state concentration of OH-radicals formed via H2O2

photolysis calculated by:

[∙ OH]ss =KH2O2,254 φH2O2,254[H2O2]

k∙OH,H2O2 [H2O2] (8)

kH2O2,254 =1000 E254 εH2O2,254[1−10

−a254 z]

a254 z (9)

where kH2O2,254 (Ein/mol s) is the specific rate constant of light absorbed by the hydrogen

peroxide at 254nm, φH2O2,254 is the quantum yield of hydrogen peroxide at 254nm (1

31

mol/Ein) and εH2O2,254 is the molar absorption coefficient of the hydrogen peroxide at 254

nm (εH2O2,254 = 19.6 M-1 cm-1). The scavenging term in this study includes only the second-

order rate constant for the reaction of hydrogen peroxide and OH-radicals (kOH,H2O2=2.7x107

M-1 s-1) (Buxton et al., 1988).

Combining equations (7)-(9) and solving for k•OH/C we obtain equation (10):

k∙OH/C =kox k∙OH,H2O2

kH2O2,254 𝜑H2O2,254 (10)

The second-order rate constants between the pesticides and the OH-radicals (k∙OH/C) can also

be determined experimentally, via competition kinetics experiments with a •OH-probe

compound. Para-chlorobenzoic acid (pCBA) was selected, since it reacts only with the OH-

radicals, is unsusceptible to UV photolysis and has been widely used as a probe compound

(Elovitz and Gunten 1999, Pi et al. 2005). By measuring the change in concentration of

pCBA, an indirect measurement of the •OH exposure can be provided. The reaction of the

organic compound and the pCBA with •OH can be described by the following kinetic

equations (Autin et al., 2012):

Compound + •OH → Products

rC = k•OH/C[C][• OH] = k’C [C] (11)

pCBA + •OH → Products

rpCBA = k•OH/pCBA[pCBA][• OH] = k’pCBA [pCBA] (12)

Where rC and rpCBA are the reactions rates of the organic compound and pCBA

respectively, in M-1 s-1, k•OH represents the second-order rate constant in M-1 s-1 and k′ the

pseudo-first-order rate constant in s-1.

The oxidation of all three compounds followed the pseudo-first-order kinetic model and from

the slopes of the ln (C/C0) – t graphs the pseudo-first-order rate constants (k’) were obtained.

ln(C

Co) = −k′ t (13)

32

The second-order rate constants between each compound and •OH (k•OH,c ) were obtained

combining the equations (11) and (12):

k•OH/c = k•OH/ pCBA k’C

k’pCBA (14)

2.4 Nitrate/Nitrite issues by UV/H2O2

2.4.1 Nitrite formation by nitrate UV photolysis

Nitrate is a common constituent of natural surface and groundwater. Nitrite is usually present

when either microbial or chemical reduction takes place. As an example of the latter, nitrite is

reported to be formed when MP lamps are applied for either disinfection or oxidation

processes due to the strong absorption of UV light by nitrate in the wavelength emission

range of these lamps (200-300 nm) (Mack and Bolton, 1999). After human exposure,

although in the blood most absorbed nitrite is oxidized to nitrate, a residual amount can react

with haemoglobin and cause a disorder known as methaemoglobinaemia which can lead to

cyanosis; on the other hand nitrite can react in the stomach with secondary and tertiary

amines or amides to form N-nitroso compounds. To that extent, both nitrate and nitrite are

regulated. The regulation limit for nitrate in drinking water is 50 mg/L as NO3- (or 11 mg/L

as NO3--N) to protect against methaemoglobinaemia , whereas the regulation limit for nitrite

varies between countries and organizations, from the guideline value of the WHO (2011) of 3

mg/L as NO2- (or 0.9 mg/L as NO2

--N) (short-term exposure) to the stricter European

Communities Drinking Water Regulations (2007) value of 0.5 mg/L for tap water and 0.1

mg/L ex-water treatment works, both as NO2- (EU, 2007).

The absorption spectra of NO2- and NO3

- are dominated by intense π π* bands at 205 nm

(ε=5500 M−1 cm−1) and 200 nm (ε=9900 M−1 cm−1), respectively. These anionic species can

hinder the efficiency of the UV-H2O2 AOP by reducing the amount of incident UV light for

H2O2 photolysis (in the case of UV/H2O2 treatment) by causing an ‘inner filter’ effect. On the

other hand, their photolysis results in the formation of ˑOH radicals which are beneficial for

the efficiency of AOPs (Mack and Bolton, 1999).

The photolysis of nitrate applying MP-UV lamps has been extensively studied (Mack and

Bolton 1999, Sharpless and Linden 2001, Sharpless et al. 2003). Martijn and Kruithof (2012)

demonstrated that for NO3- concentrations of either 2.1 or 13.6 mg/L, nitrite did not exceed

33

the 0.1 mg/L EU limit for MP-UV photolysis with UV fluences up to 90 mJ/cm2. However

the nitrite formation exceeded this limit when MP-UV/H2O2 was applied.

In some studies nitrate photolysis utilizing low pressure (LP) lamps was investigated (Plumb

and Edwards 1992, Mark et al. 1996, IJpelaar et al. 2005, Goldstein and Rabani 2007, Lu et

al. 2009) studying either the mechanistic reactions or the effect of parameters, such as pH and

H2O2 addition. IJpelaar et al. (2005) observed that no or low amounts of nitrite were

produced by LP-UV photolysis. von Sonntag and Schuchmann (1992) demonstrated that

during irradiation of 50 mg/L NO3− with λ=254 nm (LP lamp) and a UV fluence of 40

mJ/cm2, the NO2− concentration did not exceed the maximum allowed concentration of 0.1

mg/L. Sörensen and Frimmel (1997) found that irradiation at 254 nm (LP) of nitrate in

concentrations of 10-50 mg/L produced nitrite in the 0.1–0.5 mg/L range.

Regarding the role of hydrogen peroxide, conflicting results have been reported; Lu et al.

(2009) utilizing a LP lamp at pH=9.5 for a concentration of 10 mg/L NO3--N (44 mg/L NO3

-)

found that 0.8 mg/L NO2- was produced both in the absence of H2O2 and with 10 mg/L H2O2.

Contrary to Lu et al. (2009), Sharpless et al. (2003) showed that nitrite formation by MP-UV

can be enhanced by H2O2 addition; they reported 0.17 mg/L NO2- formed by MP-UV

photolysis of water containing 10 mg/L of nitrate with a UV fluence of 150 mJ/cm2 and 0.20

mg/L NO2- with the addition of 10 mg/L H2O2 at pH=8.3, with this difference being more

enhanced at pH=6.5, where 0.18 and 0.13 mg/L NO2- was formed with and without the

presence of H2O2, respectively.

The presence of natural organic matter (NOM) can affect nitrite formation in a more

complicated way than the abovementioned parameters, because of the complex composition

of NOM and its behaviour under photolysis or advanced oxidation. The role of scavengers on

nitrite formation has been studied by several authors mainly describing the effect of specific

organic compounds on the NO2- formation during UV photolysis of NO3

- under different

wavelengths (Daniels et al. 1968, Shuali et al. 1969, Bayliss and Bucat 1975, Mark et al.

1996, Goldstein and Rabani 2007). According to Mark et al. (1996), the nitrite yield at low

pH (pH=4-7) increases significantly when more hydroxyl radical scavengers are present,

whereas this is not observed at more alkaline pH (pH>8). Martijn et al. (2014) proposed the

reaction of nitrogenous radical intermediates from nitrate photolysis with Pony Lake NOM

for the production of nitrated products by MP-UV photolysis, reporting the formation of

nitrophenols when phenol was used as an organic model compound.

34

From the current literature there appears to be only one study that has examined the effect of

NOM (only Suwannee River NOM) on the nitrite yield by MP-UV and LP-UV photolysis

(Sharpless and Linden, 2001), which concluded that the presence of NOM resulted in higher

nitrite formation compared to the yields in the absence of NOM. They reported nitrite

concentrations of 0.23 and 0.39 mg NO2-/L by MP photolysis with a UV fluence of 400

mJ/cm2 with an initial nitrate concentration of around 44 mg/L at pH=8, in the absence and in

the presence of 5 mg/L DOC, respectively.

So far, there has been no reported investigation of the effect of different types of NOM, nor

of the combined effect of NOM and H2O2, on nitrite formation. Also, the similar previous

studies focused on MP-UV (polychromatic) advanced oxidation rather than LP-UV

(monochromatic, mainly at 254 nm) advanced oxidation. NOM is not only an important

water quality constituent from the standpoint of UV light absorption and ˑOH radical

scavenging, theoretically decreasing the nitrite yield, but because UV photolysis of NOM

causes formation of radicals that can act as reducing agents, contributing to the production of

nitrite via conversion from nitrate.

2.4.2 Photochemistry of nitrate UV photolysis

The photochemistry of the nitrite formation from nitrate consists of a complex set of

reactions. The photolysis of nitrate is initiated by reactions (r9-r11) (Lu et al., 2009):

NO3− + hv → 𝐍𝐎𝟐

− + O(3P) φ < 0.001 (r9)

NO3− + H+ + hv → NO2 · + · OH φ = 0.09 (r10)

NO3− + hv → ONOO− φ = 0.1 (r11)

The quantum yields shown (φ) are for the wavelength of 254 nm (Mark et al., 1996).

Reaction (r9) is dominant at very low pH, but with increasing pH the reaction loses its

significance to the nitrite production. Reaction (r10) is mainly significant for wavelengths λ >

280 nm (Goldstein and Rabani, 2007). Plumb and Edwards (1992) stated that at a wavelength

of 254 nm direct photolysis of nitrate will account for about 9% of the nitrite formed. The rest

is produced via the intermediate production of the peroxynitrite ion (ONOO-), see reaction

(r11).

35

Peroxynitrite is a strong, relatively long-lived intermediate (Alvarez et al., 1995). A KWR

report (KWR, 2004) stated that peroxynitrite formation is essential for the nitrite formation

by MP photolysis. At pH ≥ 8, it is the dominant species compared to its conjugate base

(ONOOH) (Figure 2.3 from KWR (2004)). Various reactions originate from peroxynitrite,

leading to formation of intermediates with nitrite as end product.

Figure 2.3. ONO2H/HONO2- equilibrium versus pH (adapted from KWR (2004)).

ONOO−hv→ 𝐍𝐎𝟐

− + O(3P) (r12)

ONOO− ↔ NO · +O2· − k = 0.017 s-1 (r13)

1

ONOO− + H+ ↔ ONOOH pKa = 6.6 (r14)

ONOOH → H+ + NO3− k = 1.4 s-1 (r15)

ONOOH + ONOO− → O2 + H+ + 2𝐍𝐎𝟐

− (r16)

Reaction (r12) shows the photolysis of peroxynitrite. Mark et al. (1996) argued that this

reaction, at best, only plays a minor role in the nitrite formation at 254 nm. At high pH the

peroxynitrite is predominantly present because of the pKa value of 6.6 (Sharpless et al.,

2003) (reaction r14), therefore hardly any ONOOH is available and thus at high pH the

reactions (r15) and (r16) are not expected to account for a big part of the nitrite formation or

1 (Lu et al., 2009)

0

20

40

60

80

100

2 3 4 5 6 7 8 9 10 11 12

ON

O2H

/HO

NO

2-ra

tio

(%

)

pH

ONO2H ONO2-

36

the decomposition to nitrate (Mack and Bolton, 1999). Coddington et al. (1999) showed that

nitrite formation through reaction (r16) occurred for a pH value up to 8.5, at which it reached

a plateau. Taking this into account and the fact that reaction (r13) and (r14) will not result

into the formation of nitrite on their own, low formation of nitrite is expected under LP

photolysis of nitrate.

2.4.3 Hydrogen peroxide effect on nitrite formation

Nitrite formation is enhanced when hydrogen peroxide is added to the nitrate solution. The

photolysis of hydrogen peroxide results into the production of hydroxyl radicals, as seen in

reaction (r17).

H2O2hv→ 2 · OH (r17)

These hydroxyl radicals, along with the peroxynitrite, take part in a series of reactions

producing a variety of radicals as intermediates, subsequently leading to increased formation

of nitrite (reactions r18-r31) (Lu et al., 2009).

ONOO− +· OH → ONOO · +OH− k = 5 109 M-1s-1 (r18) 2

NO2 · +ONOO− → ONOO · + 𝐍𝐎𝟐

− (r19)

ONOO · → NO · +O2 (r20)

NO · + · OH → H+ + 𝐍𝐎𝟐− (r21)

· OH + NO2− → NO2 · +OH

− k = 1 1010 M-1s-1 (r22) 2

2NO2 · → N2O4 k = 4.5 108 M-1s-1 (r23) 2

NO2 · + NO · → N2O3 k = 1.1 109 M-1s-1 (r24) 2

2 NO2 · +H2O → 𝐍𝐎𝟐− + NO3

− + 2H+ (r25)

NO2 · + NO · + H2O → 2 𝐍𝐎𝟐− + 2H+ (r26)

NO2 · + · OH → ONOOH k = 1.3 109 M-1s-1 (r27) 2

NO2 · +ONOO · + H2O → 2NO3− + 2H+ (r28)

2 (Lu et al., 2009)

37

N2O3 + H2O → 2 H+ + 2𝐍𝐎𝟐

− k = 5.3 102 s-1 (r29) 3

N2O4 + H2O → 𝐍𝐎𝟐− + NO3

− + 2H+ k = 1 103 s-1 (r30) 3

N2O3 + ONOO− → 𝐍𝐎𝟐

− + 2 NO2 · k = 3 108 M-1s-1 (r31) 3

The reactions show that the addition of hydrogen peroxide will increase the formation of

nitrite due to an increase in the number of pathways leading to this formation (reactions r19,

r21, r25, r26, r29-r31).

Hydrogen peroxide itself can also react with the peroxynitrite (reactions r32-r34) (Alvarez et

al., 1995). The reaction of the superoxide ion (O2∙-) with the NO2

∙ radical (reaction r34) is

faster by three orders of magnitude than the reaction with the HO2∙ radical (reaction r33),

resulting in nitrite formation.

ONOO− + H2O2 → · O2− + H2O + NO2 · (r32)

HO2 · + O2· − + H2O → H2O2 + O2 + OH

− k = 1 105 M-1s-1 (r33) 4

NO2 · + O2· − → 𝐍𝐎𝟐

− + O2 k = 1 108 M-1s-1 (r34) 4

Hydrogen peroxide can also scavenge the hydroxyl radicals that are produced (reaction r35).

H2O2 +· OH → H2O + H+ + O2

· − k = 2.7 107 M-1s-1 (r35) 4

2.4.4 Natural organic matter effect on nitrite formation

The presence of a background water matrix is expected to lead partly to consumption of a

part of the UV light and the hydroxyl radicals, reducing the contribution of pathways for

nitrite formation via nitrate photolysis. The UV light consumption and OH-radical

scavenging by the NOM are well-known processes (Crittenden et al., 2012) but the reactions

involved in the UV photolysis of NOM itself affect the pathways leading to nitrite formation.

UV photolysis of NOM leads to species, such as 1O2, O2∙-,

∙OH, other peroxyl radicals and

solvated electrons (Cooper et al. 1989, Frimmel 1994, Bems et al. 1999). The production of

those species is summarized by the reactions (r36)-(r44) given below.

3 (Lu et al., 2009) 4 (Alvarez et al., 1995)

38

The excited states of NOM (HS in the reactions stands for Humic Substances) lead to the

production of singlet oxygen (1O2) via reaction (r38) (which is quenched by water, producing

molecular oxygen, 3O2). Superoxide radicals are produced via reactions (r37), (r41) and (r42)

and solvated electrons via reactions (r39)-(r40). Hydroperoxyl (HO2∙) and superoxyl radicals

(O2∙-) form hydrogen peroxide which in turn is photolyzed into hydroxyl radicals (reactions

r43-r44).

HS1 → H1 S∗ → H3 S∗ (r36)

H3 S∗ + O3 2 → HS+ . + O2

.− (r37)

H3 S∗ + O3 2 → H1 S∗ + O1 2 (r38)

HS∗ or H3 S∗ → [ HS+. + e− ] (r39)

[HS+. + e−] → HS+. + es− (r40)

O2 + es− → O2

.− k= 1.9-2.2 1010 M-1s-1 (r41) 5

HS1 + O1 2 → HS+ . + O2

.− (r42)

HO2. + O2

.− → O2 + HO2− → O2 + H2O2 + OH

− (r43)

H2O2 → 2 HO. (r44)

The solvated electrons generated from the excitation of NOM can act as reducing agents and

lead to the formation of nitrite via reactions (r45-r47).

es− + NO3,s

− → NO3,s2− (r45)

NO3,s2− + H2O → NO2,s + 2OHs

− (r46)

2NO2,s + H2O → NO3,s− + 2H+ + NO2,s

− (r47)

5 (Cooper et al., 1989)

39

It should be noted that, apart from the photolysis of NOM, partial oxidation from the H2O2

present can be expected. Although H2O2 has a high redox potential (1.77V at 25◦C; acidic

conditions), its direct reaction with organic compounds comprising NOM is expected to be

insignificant due to the slow kinetics of these reactions. On the other hand, some inorganic

compounds, such as transition metals and inorganic anions, are known to react readily with

H2O2, initiating reactions involving free radicals (e.g. O2∙-,

∙OH, HO2

∙ etc.) (Siegrist et al.,

2011). This fact could be of significance, since, as shown above, these radicals play a role in

some of the reactions leading to the formation of nitrite.

2.5 Nitrated organic product formation by UV/H2O2

AOPs have been shown to produce reaction and disinfection by-products (DBPs) (e.g.

trihalomethanes (THMs) and haloacetic acids (HAAs)) (Richardson et al. 2007, Gonsior et al.

2014) that greatly depend on the water matrix, the process and the conditions applied. Toxic

by-products of UV irradiation of NOM include low molecular weight (LMW) compounds

such as oxalic and formic acid and acetaldehyde, as well as metals released from the NOM

structure, e.g. copper (Parkinson et al., 2001). The nitrogenous DBPs have been studied only

to a small extent (Shah et al. 2011, Bond et al. 2011 and 2012). Specifically, studies

investigating the reaction product formation in NOM-nitrate rich water by UV/H2O2

treatment are few. Since the nitro and/or nitroso-derivatives produced from these interactions

are known carcinogens (e.g. nitrated polycyclic aromatic hydrocarbons, PAHs, especially

dinitropyrenes, biphenyls, naphtalenes etc) (Rosenkranz and Mermelstein 1985, Kovacic and

Somanathan 2014) this suggests a concern from a public health standpoint; as a result, it is

critical that their presence, nature and potential contribution to the water’s toxicity is

investigated.

As shown in the photochemistry sections (2.4.2-2.4.4), the photolysis of nitrate, especially in

the MP-UV range, leads to peroxynitrite (HOONO) formation as the main intermediate

species. Peroxynitrite’s decomposition leads to the production of nitro and nitroso radicals

with nitrate and nitrite as end products (Mack and Bolton 1999, Goldstein and Rabani 2007).

In the presence of an organic matrix, the incorporation of inorganic nitrogen into the organic

matrix has been demonstrated (Thorn and Cox, 2012). The main mechanisms shown to take

place during the photolysis of nitrate or nitrite in the presence of low molecular weight

organic molecules, mainly aromatic ones, such as benzene, biphenyl, and phenol (Niessen et

40

al. 1988, Machado and Boule 1995, Dzengel et al. 1999, Vione et al. 2001, Vione et al. 2004,

Martijn et al. 2014, Martijn et al., 2015) are hydroxylation, nitration and nitrosation, with the

latter being enhanced in the absence of oxygen (Machado and Boule, 1995).

In these studies, the formation of a variety of nitrated compounds was observed (e.g. 2-

nitrophenol, 4-nitrophenol, 4-nitropyrocatechol, 4-nitrosophenol). Studies investigating the

reaction product formation from the NOM-nitrate interactions are few; Martijn et al. (2014)

reported the formation of 2- and 4-nitrophenol and 4-nitrocatechol when irradiating phenol in

the presence of nitrate with MP-UV, whereas Kolkman et al. (2015) found a variety of

nitrogen containing compounds in synthetic (nitrate-rich, with Pony lake NOM) and full-

scale water samples with MP-UV and MP-UV/H2O2, respectively, confirming the following

three: 4-nitrophenol, 4-nitrocatechol, and 2-methoxy-4,6-dinitrophenol.

2.6 Mutagenicity issues related to UV/H2O2 treatment

The contamination of the drinking water sources with a wide range of anthropogenic micro-

pollutants arising every year (Houtman, 2010) calls for their treatment with various methods,

including AOPs, which might lead to the production of compounds more harmful than the

parent ones. The physicochemical analyses available along with the recent broad screening

techniques based on high resolution mass spectrometry (e.g. Time-of-flight (TOF) and

Orbitrap analyzers) have contributed in target analysis, but their limitations have encouraged

the development of short-term bioassays.

During the last decade, various bioassays have been developed for the drinking and waste

water applications. The growing development and use of bioassays is a result of strict quality

requirements that are set in relation to public health, especially when it comes to drinking

water. In bioassays, two testing strategies can be distinguished: a) mutagenicity testing

(endpoint mutation), and b) genotoxicity testing (different endpoints representing primary

DNA damage). Some examples of mutagenicity tests are the Ames test, the E.coli test

(bacterial tests) and the Micronucleus Assay (eukaryotic test). To study genotoxicity effects,

tests like the umuC test (bacterial test) and the Comet Assay (eukaryotic test) can be applied

(OSPAR Commission, 2002).

In this study, the formation of nitrophenols was investigated as a part of the nitrate-related

issues from LP-UV/H2O2 treatment. Nitrated compounds are known to be toxic (Rosenkranz

41

and Mermelstein, 1985) and as a result may pose a health risk in drinking water. Therefore it

is critical that their presence and potential contribution to the water’s toxicity are

investigated. For this purpose, the Ames assay was selected, because this test is used as a

standard screening method for the detection of mutagenic compounds in water. ‘‘A substance

that is mutagenic in the Salmonella typhimurium bacterium is more likely than not to be a

carcinogen in laboratory animals, and thus, presents a risk of cancer to humans’’ (U.S.

Department of Health and Human Services, 2016). The Ames test is advantageous in terms of

easy application, rapid generation of results and low cost, thus rendering it a useful tool for

assessing potential carcinogenicity. As all tests, the Ames test also presents some limitations;

since Salmonella is a prokaryote, it is not a perfect model for humans. In mammalian

organisms, chemical compounds are often metabolized in the liver, which can lead to the

formation of mutagenic metabolites. For that reason, a liver extract (S9) (e.g. rat or human)

can be added (Hakura et al., 1999). Moreover, nitrate-containing molecules that can

potentially generate nitric oxide (NO) as well as protein-containing mixtures that involve

aminoacids, have been shown to give false positives in the Ames test (Khandoudi et al.,

2009).

Martijn et al. (2015) showed that reaction products from the water matrix may have a

stronger health impact than reaction products from priority compounds. Specifically the

nitrate photolysis role becomes obvious, when synthetic quartz sleeves that allow higher UV

transmittance at the low UV range (λ< 240 nm) are used, leading to a significant Ames

response (Martijn and Kruithof, 2012). Similarly, Penders et al. (2012) stated negative Ames

(and Comet assay) test results when natural quartz sleeves with a cut-off of UV light at λ<

240 nm were utilized.

MP-UV and MP-UV/H2O2 treatment have been found to cause significant Ames responses,

(Heringa et al. 2011, Hughes 2013, Hofman-Caris et al. 2015) with key parameter being the

nitrate photolysis in the presence of an organic matrix, yielding nitrated compounds (as stated

in section 2.5) and posing a health concern (Kolkman et al. 2015, Martijn et al. 2016).

Martijn et al. (2014 and 2015) stated that a preliminary risk assessment through the

conversion of the Ames responses of full-scale water samples treated by MP-UV and MP-

UV/H2O2 to 4-NQO equivalents showed that there is reason for concern and that further

investigation is required. Although in theory LP-UV and LP-UV/H2O2 are expected to give

little, if any, toxicity, very few studies have actually put that theory to the test, concluding

42

that LP-UV-based treatment did not cause any genotoxicity formation (Hofman-Caris et al.,

2015) or a very weak response (Haider et al. 2002, IJpelaar et al. 2005).

43

3. Aims and Objectives

The overall aim of this project was to assess the suitability of the LP-UV/H2O2, efficiency-

and safety-wise, when it comes to its application to drinking water production. On the one

hand, the concentration of pesticides in drinking water is regulated, displaying the importance

for compliance, and on the other hand, knowledge on their kinetic parameters and

degradation mechanisms are essential to elucidate their behaviour under oxidation and

photolysis processes. This knowledge is the basis for the optimization of the AOP, the

application of which cannot be avoided if prevention (partly) fails (for instance sustainable

use of pesticides from the farmers’ side, promotion of farming methods with low input of

pesticides, such as organic farming).

The AOP application, via the degradation/photolysis of pesticides along with other

constituents of the background water matrix, results in the formation of metabolites, either

reaction or disinfection by-products, introducing the need to ensure that the final water is as

safe as before treatment, if not safer. Moreover, important factors, such as the presence of

H2O2, with a controversial effect, as well as the effect of NOM, the biggest organic part in

water, on nitrite formation, which has been overlooked, drove this research. To this extent,

the fate of mecoprop, metaldehyde and clopyralid as well as the formation of nitrite and water

mutagenicity after LP-UV/H2O2 treatment were the main issues examined, and the following

specific research objectives were formulated:

Determine the ability of LP-UV/H2O2 to degrade the pesticides and elucidate the reaction

kinetics and the degradation mechanisms.

Determine the impact of the presence of water matrix constituents (e.g. natural organic

matter, inorganic ions) on the degradation of the pesticides by LP-UV/H2O2.

Identify reaction products arising from the degradation of the pesticides by LP-UV/H2O2.

Investigate the effect of NOM on the formation of nitrite by LP-UV/H2O2 in nitrate-

containing waters.

Identify nitrogen-containing reaction products formed by the interactions of natural

organic matter (NOM) and nitrate by LP-UV/H2O2.

Evaluate the mutagenicity of synthetic and real water samples treated by LP-UV/H2O2.

44

3.1 Thesis Outline

This thesis includes chapters for Introduction, Literature Review and Materials and Methods,

followed by a chapter summarising the experimental results and discussing them. In the

penultimate chapter a discussion focussed on the overall water treatment implications of the

thesis findings is presented, concluding with the final chapter summarising the contributions

to knowledge of this research.

Chapter 1 is an Introduction to the current problem of micro-pollutants in drinking water

treatment and to advanced oxidation processes (AOPs) as alternatives to conventional

treatment methods, specifically the LP-UV/H2O2 process. Reasoning is provided to explain

the choice of this process and the specific pesticides to be considered, followed by a

statement of the specific research objectives.

The first two sub-chapters of Chapter 2 consist of a comprehensive literature review, dealing

with the photochemistry concepts necessary to interpret the findings of this work, followed

by a short description of the UV-based AOPs and the chemistry involved in the (LP)-

UV/H2O2 process. A review of the few previous studies referring to the specific pesticides

that are considered in this research is given in the third sub-chapter, and the kinetic model

used to describe their degradation kinetics is explained. Nitrate photolysis concepts as well as

the parameters affecting this photolysis, especially natural organic matter (NOM), leading to

nitrated products, are presented in a fourth sub-chapter, with the fifth sub-chapter

highlighting the lack of mutagenicity information regarding the UV/H2O2 process.

Chapter 3 describes the chemicals and analytical methods used and gives a detailed

description of the UV collimated beam experiments performed on water samples. A separate

sub-chapter is included for the experiments involving samples from a drinking water

treatment plant, followed by the mutagenicity testing performed on both synthetic and water

treatment plant samples.

Chapter 4 summarises the experimental results, divided into two sub-chapters; the first one

presents the results of the pesticide degradation kinetics study and reaction product formation

and the second one reports the formation of nitrite, the effect of NOM, and the formation of

nitrated products from the incorporation of inorganic nitrogen into the NOM and potential

mutagenicity issues when the LP-UV/H2O2 process is applied.

45

All the findings mentioned in Chapter 4 are discussed in the frame of real water treatment

applications, concluding that there are no evident concerns arising from LP-UV/H2O2 under

the conditions applied in this research. The chapter concludes with a comparison between this

research (where LP-UV/H2O2 was applied) and similar research conducted previously in

which the medium pressure (MP-UV/H2O2) process was considered.

The contribution of this research to scientific knowledge, highlighting the successful

addressing of the objectives that are presented in the Introduction, is presented in Chapter 6,

and the thesis is then concluded, proposing future research opportunities.

At the end of this thesis, the references as well as an Appendix can be found, consisting of

supporting information and data.

Two publications have arisen from this work:

- Semitsoglou-Tsiapou, S., Templeton, M.R., Graham, N.J.D., Hernández Leal, L.,

Martijn, B.J., Royce, A., Kruithof, J.C., 2016. Low pressure UV/H2O2 treatment for

the degradation of the pesticides metaldehyde, clopyralid and mecoprop - kinetics and

reaction product formation. Water Research, 91, 285-294.

- Semitsoglou-Tsiapou, S., Mous, A., Templeton, M.R., Graham, N.J.D., Hernández

Leal, L., Kruithof, J.C., 2016. The role of natural organic matter in nitrite formation

by LP-UV/H2O2 treatment of nitrate-rich water. Water Research, 106, 312-319.

A third article which included the mutagenicity work has been submitted for publication at

the time that this thesis was submitted:

- Semitsoglou-Tsiapou, S., Mous, A., Templeton, M.R., Graham, N.J.D., Mandal, S.,

Hernández Leal, L., Kruithof, J.C., 2016. Mutagenicity formation during drinking

water treatment from LP-UV/H2O2 treated nitrate-rich waters. Submitted.

46

4. Materials and Methods

4.1 Chemicals

Metaldehyde, clopyralid, mecoprop, 4-chlorobenzoic acid (pCBA), sodium bromide (NaBr)

and bovine catalase were purchased from Sigma-Aldrich (the Netherlands). Monosodium

dihydrogen phosphate (NaH2PO4), disodium monohydrogen phosphate (Na2HPO4), sodium

nitrate (NaNO3), hydrogen peroxide (H2O2) (30%), HPLC-grade acetone ((CH3)2CO),

methanol (CH3OH) and ethyl acetate (CH3COOC2H5) (99.9%) were supplied by VWR (the

Netherlands).

Three types of reference organic matter were obtained from the IHSS (International Humic

Substances Society) as dry solid extracts (Photo 2, Appendix E): Suwannee River NOM

(2R101N), Nordic Lake NOM (1R108N) and Pony Lake NOM (1R109F). The first two types

were described by the IHSS as ‘aquatic NOM’, and the last type as a ‘Reference fulvic acid’.

The elemental composition of the NOMs is given in Appendix B.

Laboratory-grade water (LGW) was produced by a Milli-Q Advantage A10 system (Merck

Millipore, Germany). Stock and working solutions for all experiments were prepared in Milli-

Q water. The stock solutions for each pesticide were prepared by adding the pesticide,

followed by moderate heating and sonication if needed, to achieve complete dissolution.

Working solutions were prepared with the required amounts of the stock solutions and, where

appropriate, hydrogen peroxide solution.

4.2 Analytical methods

The pesticides (metaldehyde, clopyralid, mecoprop) and p-chlorobenzoic acid were detected

and quantified in each sample by liquid chromatography tandem mass spectrometry (LC-

MS/MS) using an Agilent 6410 Triple Quadrupole (QqQ) Mass Analyzer with electrospray

ion source. Metaldehyde was detected in the positive mode, whereas clopyralid, mecoprop

and p-chlorobenzoic acid were detected in the negative mode. A Phenomenex Kinetex

Phenyl-Hexyl column (100mm*2.1mm, 2.6µm particle size) was used, equipped with an

appropriate guard column. The mobile phases used for the positive mode were A: 2.5 L Milli-

Q water with 2 mL formic acid (99%) and 1 mL ammonia (30%), and B: 2.5 L acetonitrile

with 0.1% formic acid, and for the negative mode A: 2.5 L Milli-Q water with 0.75 mL

formic acid (99%) and 1.5 mL ammonia (30%), and B: 2.5 L acetonitrile. The flow rate was

47

0.35 mL/min. As internal standards, fenoprofen for the negative mode and

dihydrocarbamazepine for the positive method were used. For instrument control and data

analysis Agilent Masshunter Quant software was used. The same method was used for the

detection of the reaction products from each pesticide. Before every set of experiments,

calibration curves (0.5-1000 μg/L) for the pesticides were generated with good linearity

(R2>0.99). An example of calibration curves for the pesticides as well as standard, blank and

pesticide responses are given in Figures C.1-C.3 and Tables C.1-C.5, Appendix C.

The analysis for the detection of nitrophenols (2-nitrophenol + 4-nitrophenol (NP), 2,4-

dinitrophenol (DNP), 2-methyl-4,6-dinitrophenol (DNOC) and Dinoseb (DINO)) was

performed by extraction using polymeric Solid Phase Extraction (SPE) tubes (Photo 1,

Appendix E), followed by liquid chromatography tandem mass spectrometry (LC-MS/MS).

For the SPE extraction, acetone/ethyl acetate (1:1), methanol and acidified blank sample were

used consecutively for the activation of the SPE cartridges. The cartridges were then eluted

with 0.5 L of acidified sample. Elution was performed with 7.5 mL of acetone/ethyl acetate

(1:1). 50 µL of the internal standard solution (fenoprofen in methanol) was added to the

eluates, which were further evaporated with a nitrogen stream until 0.2 mL. Finally, 0.8 mL

of ultrapure water was added before analysis. For the negative ion electrospray LC-MS/MS

analysis, an Agilent 6410 QqQ Mass Analyzer with electrospray ion source was used. A

Phenomenex Gemini Phenyl-hexyl column (150mm*3mm, 3µm particle size) was used,

equipped with an appropriate guard column for separation. The mobile phases used were A:

2.5 L Milli-Q water with 0.75 mL formic acid (99%) and 1.75 mL ammonia (30%), and B:

2.5 L acetonitrile. The flow rate was 0.6 mL/min. The compounds were measured with

specific QqQ transitions. The recoveries for the spiked blanks for NP, DNP, DNOC and

DINO were 109%, 95%, 98% and 97%, respectively. For instrument control and data

analysis Agilent MassHunter software was used. The quantification limit (QL) and detection

limit (DL) of the method were 0.02 and 0.007 μg/L, respectively. An example of a calibration

curve for nitrophenol as well as standard and blank responses are given in Figure C.4 and

Tables C.6-C.7, Appendix C.

NOM characterisation was performed by Liquid Chromatography-Organic Carbon Detection

(LC-OCD) (Model 8, DOC-LABOR, Karlsruhe, Germany), equipped with both DOC and

DON detection, an organic carbon detector (NDIR), an organic nitrogen detector (UV 220

nm), as well as a UV detector (254 nm), all integrated in the LC-OCD system. The column

48

used was a Toyopearl HW-50S (30μm, 250 mL) and a phosphate buffer was used as eluent.

Data analysis was performed with DOC-LABOR software (ChromLog version).

The chloride, nitrate, nitrite, bromate and bromide ions were quantified by ion

chromatography (detection limits of 0.05 mg/L for nitrite and 0.1 mg/L for the rest of the

ions). A Metrohm IC Compact 761 ion chromatograph (IC) was used, equipped with a

Metrohm Metrosep A Supp 5 (150/4.0 mm) column, a Metrohm Metrosep A Supp 4/5 Guard

pre-column and a conductivity detector. Examples of calibrations curves for nitrate and nitrite

ions are given in Figures C.6-C.7, Appendix C. Low molecular weight organic acids were

detected and quantified by ultra-high pressure liquid chromatography (UHPLC), consisting of

a Phenomenex Rezex Organic Acid H+ (300x7.8 mm) column, an Ultimate 3000 RS Column

Compartiment column oven and an Ultimate 3000 RS Variable Wavelength Detector.

Total organic carbon (TOC) content was measured by a TOC-LCPH analyser equipped with

an ASI-L autosampler. Calibration curves of TOC analysis are shown in Figures C.8-C.10,

Appendix C. H2O2 concentrations were measured using the triiodide method (Klassen et al.,

1994). The pH measurements were performed with a Metrohm 827 pH meter (Applikon

analytical, the Netherlands). KCl was used as electrolytic solution and the meter was

calibrated daily with buffer solutions of pH 4, 7 and 10 before the sample measurements took

place.

4.3 Collimated beam experiments

UV exposure experiments were carried out with a bench scale collimated beam apparatus,

equipped with a 25 Watt low pressure mercury arc discharge lamp without a lamp sleeve

(Photo 4, Appendix E). The emission spectrum of the UV lamp, obtained from Trojan UV

Technologies (London, Ontario, Canada), mainly consisted of a strong, almost exclusive

emission at 254 nm (Figure A.1, Appendix A). The type of quartz used for the lamp did not

permit transmission at 185 nm. A warm-up time of at least 20 min was allowed to ensure a

constant light output before irradiating the solution. A sample volume of 55 mL was placed in

a Petri dish and H2O2 was added to obtain the desired concentration. The distance between

the lamp and the surface of the sample was 29.5 cm. The fluence rate in the centre of the

sample was 0.2 mW/cm2 and the path length through the sample solution was 1.96 cm. As

soon as the sample was placed under the collimating tube, mixing was started, and the

irradiation time was measured as soon as the shutter was opened (Photo 3, Appendix E).

49

Immediately after irradiation, H2O2 was quenched with the addition of bovine catalase, the

samples were filtered (0.45μm pore size) and stored in the dark at 4◦C until analysis.

4.3.1 Pesticide degradation and (in)organic product formation by LP-UV/H2O2

For the kinetic study, single-solute experiments were performed. The initial pesticide

concentration was 0.3 mg/L, in order to ensure that the reduced concentrations during the

experiments exceeded the analytical detection limit, while being at the same time as close as

possible to realistic concentrations of these compounds in surface water. UV fluences in the

range of 0-1000 mJ/cm2, in steps of 200 mJ/cm2, and H2O2 doses of 0, 5 and 15 mg/L were

applied. Irradiation times ranged from 30 to 90 min, with larger times corresponding to

higher UV fluences. The solutions were not buffered but pH was monitored and variations

were within 1 pH unit (6.5-7.2).

The effect of a background water matrix on the degradation kinetics of the three pesticides

was studied via experiments where the water samples contained a mixture of the three

pesticides along with three pharmaceuticals (sulfamethoxazole, diclofenac and

carbamazepine), (Suwannee River) NOM, nitrate and bromide in laboratory-grade water. The

three sub-sections of these experiments are given in Table 4.1. In part (a) an extended UV-

H2O2 dose matrix was applied in order to observe any trends in degradation and kinetics of

the pesticides using practical conditions. In part (b) a matrix containing NOM, bromide and

nitrate was introduced in order to study the OH-scavenging effect of that matrix on the

kinetics of the pesticides. The NOM concentration in raw water is found to be in the range of

2-15 mg/L (Hepplewhite et al., 2004) which drops to around 2-5 mg C/L as DOC after pre-

treatment (e.g. filtration), therefore the NOM concentration for the experiments was chosen at

4 mg/L which corresponded to around 2 mg C/L as DOC, in order to simulate conditions of

water to be treated by the AOP step. Bromide and nitrate concentrations were kept to 0.3 and

50 mg/L, respectively, which represent high levels of these inorganic compounds in surface

water, as a worst case scenario. Finally, the experimental conditions were expanded in part

(c) to observe the behaviour of the system under high UV fluences and hydrogen peroxide

doses that could be used in drinking water treatment plants, if required, depending on the

composition of the influent water. The initial pesticide concentration was also lowered, in

order to study its effect along with the rest of the parameters (background water matrix and

50

UV/H2O2 conditions) on the degradation kinetics. Nitrite and bromate formation were

monitored in cases (b) and (c).

Table 4.1. Overview of experiments for the study of background water matrix on the kinetics

of the degradation of the three pesticides.

Condition

Pesticide

concentration

(μg/L) *

Br-

(mg/L)

NO3-

(mg/L)

NOM

(mg/L)

UV fluence

(mJ/cm2)

H2O2

(mg/L)

a) 300 - - - 0-1000

(step 200)

0-15

(step 5)

b) 300 0.30 50 4 0, 400, 600,

800

0-15

(step 5)

c) 100 0.30 50 4 0-1000

(step 500)

0-50

(step 25)

* The pharmaceutical concentration was the same as the pesticide concentration in every

case.

Competition kinetics experiments for the determination of the second-order rate constants

between the pesticides and the OH-radicals were conducted in laboratory grade water spiked

with 1 mg/L of each compound and 0.5 mg/L of pCBA as a hydroxyl radical probe

compound. The samples were irradiated with a range of UV fluences 200-500 mJ/cm2 in

steps of 100 mJ/cm2 and 5 mg/L of H2O2 as the source of the OH-radicals.

For the reaction product formation experiments, the initial concentrations, UV fluences and

H2O2 doses were increased compared to those for the kinetics experiments, in order to

achieve formation of reaction products at quantifiable levels and determine their profiles with

UV fluence (reaction time). The initial concentrations were 5 mg/L for metaldehyde, 10 mg/L

for mecoprop and 20 mg/L for clopyralid. The UV fluences were 0-1500 mJ/cm2, applied in

steps of 500 mJ/cm2 for all three pesticides. The H2O2 doses were 10 and 30 mg/L for

metaldehyde and mecoprop, and 10, 30 and 60 mg/L for clopyralid. For mecoprop, LP-UV

photolysis experiments (no H2O2) were performed as well, since it is the only photo-labile

compound of the three. The de-chlorination of mecoprop and clopyralid was detected by

chloride ion measurements. All samples were buffered at pH=8 with a phosphate buffer

solution. Before analysis, the hydrogen peroxide was quenched with the addition of bovine

catalase (except TOC analysis), the samples were filtered (0.45μm) and stored in the dark at

4◦C until analysis. All experiments were performed in duplicate.

51

4.3.2 Nitrite formation by LP-UV/H2O2

For the experiments considering the effect of the initial nitrate concentration and hydrogen

peroxide addition on nitrite formation, two nitrate concentrations, 25 and 50 mg/L NO3-, were

used, in order to represent nitrate-rich waters. The role of the NOM presence was studied

using Suwannee River NOM (4 mg/L corresponding to 2.1 mg C/L), known for its high

aromaticity (Leenheer J.A., 1994), with a nitrate concentration of 50 mg/L. For both sets of

experiments the same range of UV fluences and H2O2 doses was applied; UV fluences were:

0, 500, 1000, 1500 and 2100 mJ/cm2, and the following hydrogen peroxide doses: 0 (no

hydrogen peroxide), 10, 25 and 50 mg/L. The theoretically determined UV fluence required

to exceed the EU regulatory limit (i.e. 0.1 µg/L NO2-) was calculated to be 2083 mJ/cm2

(Appendix D) and as such an approximation 2100 mJ/cm2 was used as the maximum UV

fluence applied for all experiments.

The effect of pH on nitrite formation in the presence of NOM was studied via experiments at

both pH=6 and pH=8 with Nordic Lake NOM (4 mg/L corresponding to 1.7 mg C/L), a

nitrate concentration of 50 mg/L, and a UV fluence of 2100 mJ/cm2 in combination with 0,

10, 25 and 50 mg/L of H2O2.

The effect of the type of NOM, in terms of humic content, UV absorbance, aromaticity and

nitrogen content, on the nitrite formation was studied using Suwannee River NOM, Nordic

Lake NOM and Pony Lake NOM, in solutions buffered at pH=8. The initial concentration for

all three NOMs was 4 mg/L, which resulted in an initial DOC concentration of approximately

2 mg C/L. The UV fluences applied were: 0, 1500 and 2100 mJ/cm2, and the hydrogen

peroxide doses: 0 (no hydrogen peroxide), 10, 25 and 50 mg/L.

4.3.3 Nitrated product formation

Regarding the reaction product formation by LP-UV/H2O2 treatment from the reaction

between NOM and nitrate, experiments were performed in laboratory-grade water spiked

with either Suwannee River NOM or Pony Lake NOM (4 mg/L) in the presence of nitrate (50

mg/L). The UV fluences applied were 0 (no irradiation), 1500 and 2100 mJ/cm2 with a

hydrogen peroxide dose of 15 mg/L. All experiments were performed in duplicate and the

samples were analysed for the following nitrophenols: 2-nitrophenol + 4-nitrophenol (NP),

2,4-dinitrophenol (DNP), 2-methyl-4,6-dinitrophenol (DNOC) and Dinoseb (DINO).

52

4.4 Water treatment plant sampling

Samples were collected from a drinking water treatment works (WTW) in the UK, which

treats surface water from a lowland river known for its poor water quality caused by the

presence of industrial discharges, pesticides and high nitrate concentrations. The WTW

comprises river bank side storage where surface water is collected from the adjacent river,

followed in sequence by roughing granular activated carbon (GAC) filtration, submerged

ultrafiltration, LP-UV/H2O2 oxidation, polishing post GAC filtration, UV disinfection and

residual chlorine addition. The LP-UV/H2O2 treatment process makes use of four UV

reactors, each containing 96 UV 1kW lamps, and a H2O2 dosing system. Samples were

collected during two different periods, February (referred to as ‘‘sampling (a)’’) and April

(referred to as ‘‘sampling (b)’’) 2016, due to expected differences in water composition and

the maximum designed UV/H2O2 conditions applied as part of testing the efficiency of the

system in a worst-case scenario (UV fluences of 2000 and 1750 mJ/cm2 for sampling (a) and

(b), respectively, and a H2O2 dose of 40 mg/L). The samples were collected from the inlet

(only for April), pre-AOP, post-AOP and post-GAC treatment steps, to assess nitrite

formation, NOM fate and potential mutagenicity in terms of Ames test response during

treatment. All samples were subjected to the Ames II Mutagenicity assay (described in

section 4.5).

4.5 Ames II Mutagenicity assay

The Ames assay makes use of strains of the bacterium Salmonella typhimurium that carry

gene mutations that inhibit histidine synthesis but at the same time they are auxotrophic

mutants, i.e. they require histidine for growth; these mutations can be reversed in the

presence of mutagenic chemical compounds. The bacteria are spread on an agar plate with a

small amount of histidine which allows the bacteria to grow for an initial time and have the

opportunity to mutate. When the histidine is depleted, only bacteria that have mutated due to

mutagenic chemical compounds present and gained the ability to produce their own histidine

(a back mutation) will survive. The plate is incubated for 48 hours. The mutagenicity of a

substance is proportional to the number of revertants (i.e. colonies) observed. The Ames II

assay used in this study is a modification of the standard Ames test; it uses a liquid

microplate format and much smaller sample volumes are required (Gee et al. 1998,

Flückiger-Isler et al. 2004).

53

Both the synthetic (only with Pony Lake NOM and nitrate, described in section 4.3.3) and

full-scale water samples (described in section 4.4) were subjected to the Ames II

Mutagenicity assay (Ames et al., 1972) in order to evaluate their mutagenic potential to

induce reverse mutations in Salmonella typhimurium. There are two commonly used strains,

the frameshift strain TA98 and the base-pair detecting strain TA100. In this study, the strain

TA98, in the absence of a rat liver metabolic activation system (S9), was selected. It has been

shown by previous studies that the TA98 strain (and especially the TA98(-S9) combination)

is most responsive in samples treated with oxidation processes (e.g. van Hoof et al. 1985, van

der Gaag et al. 1986, Heringa et al. 2011, Martijn and Kruithof 2012, Martijn et al. 2014)

whereas the TA100 strain (both with and without S9 addition) has exhibited positive

responses in cases of DBPs formation (e.g. van Kreijl et al. 1980, Christman et al. 1991, Lv et

al. 2015, Manasfi et al., 2015). The synthetic water samples were prepared and treated with

LP-UV/H2O2; the H2O2 was not quenched, in order to avoid introducing agents that would

interfere with the Ames II testing process. All water samples were extracted by SPE

according to the method described by Heringa et al. (2011). The extracts were 20000x

concentrated and the Ames test was performed according to the Xenometrix ‘‘Ames II

Microplate Format Mutagenicity Assay’’ protocol (Xenometrix Inc., 2015). Solvent control

(100% DMSO), and positive controls (2-nitrofluorene + 4-nitroquinoline N-oxide) were

included. Both the sample extraction and the Ames test were performed by VITO laboratories

in Belgium.

Regarding the results obtained, in order to interpret them in terms of positive or negative

mutagenicity, the fold increases over the solvent control (ratio of the mean number of

positive wells for the test item divided by the mean number of positive wells for the solvent

control) and the fold increases over the baseline value (ratio of the mean number of positive

wells for the test item divided by the baseline value for the solvent control) were calculated.

Student’s t-test (1-sided, unpaired) was also carried out to determine significance at the p ≤

0.05 level between the range of data for the solvent control and the data for the test

substances.

A sample (or compound) that shows a clear fold increase > 2.0 (baseline) and significant

difference (p<0.05 in the t-test), is classified as a mutagen. Fold increases in revertant

numbers over the solvent control are considered as positive if > 3.0, whereas fold increases in

revertant numbers over the baseline value are positive if > 2.0. When the fold increases are

54

below these values the test substances are not mutagenic towards the Salmonella typhimurium

strain TA98 (Xenometrix Inc., 2015).

4.6 Quality Assurance / Quality control (QA/QC)

All experiments as well as analyses were performed in a controlled lab. Instrumental analysis

(LC-MS/MS, LC-OCD, IC, TOC) was performed by trained analytical staff. Any technical

challenges throughout the project were dealt with by the technical staff. Protocol use and/or

development were based on the requirements of the research and instrumental limitations

were taken into account.

Experiments were duplicated in order to conduct t-test comparisons with a 95% confidence

interval (Lyman Ott and Longnecker, 2010). Triplicates were performed in cases duplicates

differed significantly.

Before every set of experiments involving the pesticides, calibration curves (0.5-1000 μg/L)

were generated with good linearity (R2>0.99). A constant output of the UV lamp was ensured

before the start of the experiments. The exposure time for the desired UV fluences for all

irradiation experiments was calculated using the absorbance of the water samples at 254 nm

in a fluence calculation spreadsheet based on Bolton and Linden (2003). The Petri Factor was

measured before every batch of daily experiments by measuring the fluence rate across the x–

y surface using a radiometer and was equal to 0.96 (ILT1700 Radiometer, LOT-

QuantumDesign GmbH, USA). Information on the calibration of the radiometer is given in

Appendix C.

Ames testing was performed at a certified lab, VITO laboratories in Belgium. The analysis

procedure followed was agreed upon the requirements of the study. The Xenometrix ‘‘Ames

II Microplate Format Mutagenicity Assay’’ protocol (Xenometrix Inc., 2015) was applied

according to the OECD Guideline 471 for the Testing of Chemicals (OECD, 1997).

55

5. Results

5.1 Pesticide degradation and product formation by LP-UV/H2O2

5.1.1 Degradation by LP-UV photolysis and LP-UV/H2O2 treatment

The single-solute experiments are described in section 4.3.1 and Table 4.1 (part a). For the

overall degradation of a contaminant by LP-UV/H2O2 treatment a kinetic model described by

Sharpless et al. (2003), Lester et al. (2010) and Baeza and Knappe (2011) was applied and is

given in section 2.3.1. The structures of these pesticides are shown in Figure 5.1 and their

physicochemical characteristics are given in Table 5.1. The absorption spectra of the three

pesticides in the UV region are given in Figure A.2 (Appendix A).

Although the pH is expected to affect oxidation processes such as UV/H2O2 for the

degradation of micro-pollutants by changing their state in the solution, it was decided not to

buffer for the kinetics-related experiments. Taking into account the pKa values for mecoprop

(pKa=3.78, Boule et al., 2002) and clopyralid (pKa1=1.4 and pKa2=4.4, Corredor et al., 2006)

(Table 5.1) and the monitored pH range over which the experiments took place (pH=6.5-7.2),

both pesticides are expected to be present in their anionic states. The pH effect would be

significant if the pH varied beyond the range below and above the pKa values; within the

range considered in these experiments they are not expected to shift between neutral and

ionic state, therefore the kinetics are not expected to be affected. Metaldehyde does not have

a pKa value and does not dissociate in water, therefore the pH is not expected to affect its

degradation by this process. The photolysis of H2O2 is also expected to be favoured at neutral

pH, although the greatest enhancement is expected at alkaline pH (due to the higher molar

absorption coefficient of the peroxide anion (HO2-) at 253.7 nm) (Legrini et al., 1993).

Additionally, the absence of a buffer eliminates any chance of interference of the buffer on

the kinetics.

56

Figure 5.1. Chemical structure of the studied pesticides.

Table 5.1. Physicochemical characteristics of the pesticides in this study.

Name CAS # Molecular

Formula

Molecular

Weight

Water

solubility pKa logKow

(g/mol) (mg/L)

Metaldehyde 9002-91-9 C8H16O4 176.21 222 - 0.12

Clopyralid 1702-17-6 C6H3Cl2NO2 192.00 1000 1.4, 4.4 1.07

Mecoprop 7085-19-0 C10H11ClO3 214.65 620 3.78 3.21

Figure 5.2 shows the degradation profiles of the three pesticides by LP-UV photolysis and

LP-UV/H2O2 treatment for all UV fluence/H2O2 dose combinations. Mecoprop exhibited the

highest degradation ranging from 17% to 60% (0.08 to 0.4-log), since its aromatic structure

makes it susceptible to LP-UV photolysis. The degradation was directly proportional to UV

fluence. No degradation of clopyralid was observed by LP-UV photolysis, despite the

presence of a heteroatom (nitrogen) and an aromatic system in its structure. This behaviour

could be attributed to the photochemical dissociation mechanism of the pyridine ring;

irradiation at 254 nm is thought to cause an n → p* excitation leading to a bicyclic valence

isomer, Dewar pyridine, which re-aromatizes completely to pyridine within 15 min at room

temperature (Wilzbach and Rausch, 1970). Degradation of metaldehyde by LP-UV photolysis

was negligible (0.005-log or 1%). This behaviour was expected due to its low molar

http://www.chemicalbook.com/ProductChemicalPropertiesCB0438854_EN.htm



57

absorption coefficient (42 M-1 cm-1) and the absence of aromaticity, unsaturated sites or

heteroatoms in the molecule.

Addition of H2O2 caused its photolysis and subsequent production of OH-radicals, which are

non-selective oxidants, thereby enhancing the degradation of all three pesticides. LP-

UV/H2O2 treatment of mecoprop, with a UV fluence as low as 200 mJ/cm2 and 5 mg/L of

H2O2, led to a degradation of almost 80% (0.68-log). Under the maximum treatment

conditions applied in this research (1000 mJ/cm2 and 15 mg/L H2O2) the achieved

degradation was 99.6% (2.4-log). This high reactivity towards OH-attack could be attributed

to the presence of the benzene ring substituted with activating groups (-OCHCH3COOH and

–CH3).

For clopyralid, addition of hydrogen peroxide enhanced the degradation compared to LP-UV

photolysis, causing 56% (0.35-log) degradation by a UV fluence of 1000 mJ/cm2 and 5 mg/L

H2O2. When the H2O2 dose was tripled from 5 mg/L to 15 mg/L, a degradation of 84% (0.8-

log) was achieved. Of the three pesticides, clopyralid was the least susceptible compound to

LP-UV/H2O2 treatment. This can be explained by the pyridine ring of the molecule, where

electrophilic attack is hindered by the low energy of the orbitals of the ring’s π-system. In

addition, the lone electron pair of the nitrogen atom is not delocalized and destabilizes the

cationic ‘would-be’ intermediate from the electrophilic attack (Clayden et al., 2012).

The combination of UV light with hydrogen peroxide was essential for the degradation of

metaldehyde due to its non-susceptibility to LP-UV photolysis. LP-UV/H2O2 treatment

caused a gradual increase of degradation, reaching approximately 97% (1.5-log) degradation

for a UV fluence of 1000 mJ/cm2 and a H2O2 dose of 15 mg/L. Metaldehyde is a

cyclic tetramer of acetaldehyde and an ether derivative. It is proposed that the reaction of

OH-radicals with cyclic ethers occurs by direct H atom transfer, in which the hydrogen-

bonded adduct formed between the OH-radical and the ether is sterically restricted, leading to

a much lower reactivity (Moriarty et al., 2003). Furthermore, the possibility of an H-atom

transfer becomes less likely as the ring size increases, due to entropy restrictions (eight-

membered-ring in the case of metaldehyde).

58

Figure 5.2. Degradation of a) clopyralid, b) metaldehyde and c) mecoprop by LP-UV

photolysis (no H2O2) and LP-UV/H2O2 treatment.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 200 400 600 800 1000

log (Co/C)

UV fluence (mJ/cm2)

a

No H2O2

5mg/L H2O2

15mg/L H2O2

No H2O2

5mg/L H2O2

15mg/L H2O2

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 200 400 600 800 1000

log (Co/C)

UV fluence (mJ/cm2)

b

No H2O2

5mg/L H2O2

15mg/L H2O2

No H2O2

5mg/L H2O2

15mg/L H2O2

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0 200 400 600 800 1000

log (Co/C)

UV fluence (mJ/cm2)

c

No peroxide

5mg/L peroxide

15mg/L peroxide

5mg/L H2O2

15mg/L H2O2

No H2O2

59

5.1.2 Kinetic study by LP-UV photolysis and LP-UV/H2O2 treatment

The degradation of the compounds was found to follow pseudo first-order kinetics for all

combinations of UV fluences and H2O2 doses. The LP-UV photolysis rate constants (kp, s-1)

were derived from the slopes of the regression curves corresponding to LP-UV photolysis

and the LP-UV photolysis-hydroxyl radical oxidation combined rate constants (kT) from

those corresponding to the LP-UV/H2O2 treatment (Figure 5.3). The fastest degradation

kinetics by LP-UV photolysis were exhibited by mecoprop (1.5x10-4 s-1) followed by

metaldehyde (4.3x10-6 s-1) and clopyralid (9.4x10-7 s-1) (Table 5.3).

The molar absorption coefficients (εC,254, M-1 cm-1) were measured experimentally and

together with the photolysis rate constants (kp, s-1) were used to calculate the quantum yields

(φC,254, mol/Ein), following the photochemical approach given by Bolton and Stefan (2002)

and using equation (15):

kp =φC,254 εC,254 ln10

1000 Uλ,254 (15)

where Uλ,254 is the molar photon energy and represents the energy of 1 Ein of photons at 254

nm and is equal to 471528 J/Ein (Bolton and Stefan, 2002).

Both the molar absorption coefficient and quantum yield are important parameters,

determining the degree of compound degradation by LP-UV photolysis. Although

experimentally the highest molar absorption coefficient was obtained for clopyralid (1044 M-

1 cm-1) compared to mecoprop and metaldehyde (211 and 42 M-1 cm-1, respectively),

mecoprop was the compound most effectively degraded by LP-UV photolysis at 254 nm,

reflecting its greater quantum yield. The quantum yield values that were determined were: for

mecoprop 0.8810 mol/Ein, for metaldehyde 0.2014 mol/Ein and for clopyralid 0.0047

mol/Ein (Table 5.2).

60

Figure 5.3. Degradation kinetics (used for the determination of the pseudo first-order rate

constants) of (a) clopyralid, (b) metaldehyde and (c) mecoprop, by LP-UV photolysis (no

H2O2) and LP-UV/H2O2 treatment.

R² = 0.9307

R² = 0.9891

-2

-1.6

-1.2

-0.8

-0.4

0

0 1000 2000 3000 4000 5000 6000

ln(C/Co)

Time (sec)

a

No H2O2

5mg/L

H2O2

15mg/L

H2O2

No H2O2

5mg/L

H2O2

15mg/L

H2O2

R² = 0.9967

R² = 0.971

-4

-3

-2

-1

0

0 1000 2000 3000 4000 5000 6000

ln(C/Co)

Time (sec)

b

No

H2O2

5mg/L

H2O2

15mg/L

H2O2

No H2O2

15mg/L

H2O2

No H2O2

5mg/L

H2O2

R² = 0.9978

R² = 0.9995

-5

-4

-3

-2

-1

0

0 1000 2000 3000 4000 5000 6000

ln(C/Co)

Time (sec)

c

No

H2O2

5mg/L

H2O2

15mg/L

H2O2

No H2O2

5mg/L

H2O2

15mg/L

H2O2

No H2O2

5mg/L

H2O2

61

Table 5.2. Molar absorption coefficient (εC,254) and quantum yield (φC,254) values derived

from this study and reported in literature (values from duplicates given in brackets in case

they differed).

This study Literature

Compound

εC,254

(M-1 cm-1)

φC,254

(mol/Ein)

εC,254

(M-1 cm-1)

φC,254

(mol/Ein)

Metaldehyde

42.11

(42.06-42.17)

0.2014

(0.1938 – 0.2090)

21.4 6 , 19.6 7

0.000017 6,

0.00020 7

Clopyralid

1044

(998.1-1090)

0.0047

(0.0036 – 0.0059)

845 8 0.068 8

Mecoprop

210.7

(203.4-218.1)

0.8810

(0.8474 – 0.9145)

244 9

0.75 10 , 0.34 11

0.23 12

Regarding the very low quantum yield of metaldehyde and its difference to the one literature

value reported by Autin et al. (2012) (Table 5.2), this was attributed to the behaviour of

metaldehyde under LP-UV photolysis, i.e. for the first two UV fluences applied (200 and 400

mJ/cm2) the degradation increases to approximately 1% and with higher fluences it

subsequently decreases; this phenomenon was observed only for UV photolysis (not for

UV/H2O2) and also in cases where all pesticides were present in the solution (mixture of the

three pesticides). For this reason only the first two UV fluences were used for the derivation

of the photolysis rate constant (kp) for metaldehyde that was further used for the

determination of the quantum yield. Such small amount of data points as well as the quite

significant variation of the kp values from the duplicates (Table 5.3), an indication of the

variability observed by photolysis, are expected to introduce an error in the value obtained,

6 Autin et al. (2012) 7 Autin et al. (2013) 8 Orellana-Garcia et al. (2014) 9 Estimated from Fig.1 of Meunier and Boule (2000) 10 Meunier and Boule (2000) (λ=280nm, pH 5.5, +/- O2) 11 Meunier and Boule (2000) (λ=280nm, pH 2.15, O2) 12 Meunier and Boule (2000) (λ=280nm, pH 2.15, -O2)

62

which might also account for the differences observed between this study and the one from

Autin et al. (2012).

When hydrogen peroxide was present the same order of reactivity as with the LP-UV

photolysis was observed; for a hydrogen peroxide dose of 15 mg/L, mecoprop exhibited the

fastest kinetics (1.9x10-3 s-1), followed by metaldehyde (5.8 x10-4 s-1) and clopyralid (3.2x10-4

s-1) (Table 5.3). It should be noted that due to the fact that all three pesticides were studied

under the same conditions, the rate constant for mecoprop for the 15 mg/L H2O2 dose was

derived on a single datum, since the concentrations of mecoprop for the other UV fluences

applied were below the detection limit, thus decreasing the confidence in the value obtained.

The addition of hydrogen peroxide in two different concentrations (5 mg/L and 15 mg/L) had

a different impact on the degradation of each pesticide (Figure 5.4). Mecoprop kinetics were

the least affected; compared to LP-UV photolysis a 9-fold and 13-fold increase in kinetic rate

constants were observed when 5 mg/L and 15 mg/L of hydrogen peroxide were added,

respectively. On the other hand, the rate constants for metaldehyde and clopyralid were

strongly enhanced when hydrogen peroxide was added. For metaldehyde, the increase was

75-fold and 137-fold for 5 mg/L and 15 mg/L of hydrogen peroxide, respectively. Clopyralid

exhibited the largest enhancement in terms of kinetic rate constants when hydrogen peroxide

was added with a 161-fold and 343-fold increase, respectively. Triplication of the hydrogen

peroxide concentration from 5 to 15 mg/L resulted in a 1.4-fold, 1.8-fold and 2-fold increase

of the rate constants for mecoprop, metaldehyde and clopyralid, respectively.

63

Figure 5.4. Enhancement of reaction rate constants for the three pesticides with the addition

of either 5 or 15 mg/L compared to LP-UV photolysis (no H2O2).

From a practical point of view, the choice of tripling the H2O2 depends on the initial influent

concentration of the pesticide in a WTW and the level of pesticide degradation desired. In the

case of mecoprop, the reaction is fast even at the lowest H2O2 concentration, with the

degradation being enhanced only by around 2% when the H2O2 concentration is tripled. On

the other hand, for metaldehyde and clopyralid, the degradation is enhanced by 11-12%. As

an example, an influent concentration of 0.8 μg/L metaldehyde treated by a practical

UV/H2O2 combination of 200 mJ/cm2 / 5 mg/L would result in 70% degradation, reducing the

concentration to 0.24 μg/L, while the permitted level of an individual pesticide in drinking

water according to the EU legislation is 0.1 μg/L (Council Directive 98/83/EC, 1998).

Nevertheless, this estimation relies on data obtained with pure water experiments, which

indicates that in real water samples including the background water matrix, the degradation is

expected to be less than 70%.

4.1.3 Importance of OH-radicals

In order to assess the reactivity of each pesticide towards hydroxyl radicals, the parameters

obtained from the single-solute experiments (i.e. the molar absorption coefficients, quantum

yields and pseudo first-order rate constants) were incorporated into the UV/H2O2 model to

9

75

161

13

137

343

0

50

100

150

200

250

300

350

400

Mecoprop Metaldehyde Clopyralid

Fold increase in k

constant

Compound

5mg/L peroxide

15mg/L peroxide

64

obtain the second-order rate constants between each compound and •OH (k•OH/c). By

combining the equations from the LP-UV/H2O2 kinetic model and solving for k•OH/C, the

equation below is obtained (section 2.3.1, equation 10):

k∙OH/C =kox k∙OH,H2O2

kH2O2,254 𝜑H2O2,254

where kox (s-1) is the OH-radical oxidation rate constant obtained from the rate constant

determined for the UV/H2O2 process and corrected for direct photolysis contribution,

k∙OH,H2O2 (M-1 s-1) is the second-order rate constant for the reaction between hydrogen

peroxide and OH-radicals which was assumed to be equal to 2.7x107 M-1 s-1 (Buxton et al.,

1988), kH2O2,254 (Ein/mol s) is the specific rate of light absorption by hydrogen peroxide at

254 nm, and φH2O2,254 is the quantum yield of hydrogen peroxide at 254 nm (1 mol/Ein).

The values obtained were 3.59x108, 1.96x108 and 1.09x109 M-1 s-1 for metaldehyde,

clopyralid and mecoprop, respectively (Table 5.3). From these values it is evident that

mecoprop exhibited a substantially greater reactivity towards OH-radicals compared to the

other two pesticides (an increase by a factor of 3 and 5.6 compared to metaldehyde and

clopyralid, respectively).

65

Table 5.3. Time-based pseudo first-order and second order rate constants for LP-UV photolysis and LP-UV/H2O2 oxidation (kp, kΤ, s-1) for the

three pesticides (values from duplicates given in brackets where they differed).

Compound

kp

(photolysis)

kΤ

(photolysis + •OH)

k•OH/C

(kinetic modeling)

k•OH/C

(pCBA kinetics)

(s-1) (s-1) (M-1 s-1) (M-1 s-1)

No H2O2

UV +

5 mg/L H2O2

UV +

15 mg/L H2O2

- -

Metaldehyde

4.3x10-6

[7.5x10-6 – 1.0x10-6]

3.2x10-4

[2.9x10-4– 3.5x10-4]

5.8x10-4

[5.6x10-4– 6.0x10-4]

3.6x108

[3.3x108- 3.9x108]

8.3x108

[7.0x108-9.8x108]

Clopyralid 9.4x10-7 1.5x10-4

3.2x10-4

[3.4 x10-4 – 3.1x10-4]

2.0x108

6.3x108

[5.0x108-7.5x108]

Mecoprop 1.5x10-4 1.4x10-3 1.9x10-3

1.1x109

[1.11x109- 1.08x109]

8.9x109

[8.6x109-9.3x109]

66

The second-order rate constants were also calculated via the competition kinetics with pCBA

(kinetic equations given in section 2.3.1 and experiments described in section 4.3.1) and the

values obtained were 8.3x108, 6.3x108 and 8.9x109 M-1 s-1 1 for metaldehyde, clopyralid and

mecoprop, respectively. The variability for these values was larger as it is shown in Table

5.3, and also the values themselves were greater for all three pesticides compared to the ones

obtained by the kinetic model. This could be attributed to the fact that the method relies on

the application of the rate constant of a probe compound, which could be inaccurate, but also

to the fact that pCBA is not completely recalcitrant to photolysis (Pereira et al., 2007).

In the literature, two second-order rate constants are available for metaldehyde and one for

clopyralid, whereas for mecoprop more values are available, however with a large variation

between them (Table 5.4). Specifically for mecoprop, even though the kox rate constant for

the 15 mg/L H2O2 was obtained from a single datum, as explained above, increasing the

uncertainty of the value obtained, the second-order rate constant (k•OH/C) (the calculation of

which relies on the kox value) was in the same order of magnitude as the ones reported

already in literature. The differences observed between the values obtained in the present

study and the ones stated in literature can be attributed to a number of reasons; apart from the

competition kinetics with pCBA, Autin et al. (2012) also utilized the model of Minakata et al.

(2009) where the calculations are based on the molecule geometry and structure. Armbrust

(2000) used acetophenone as a reference compound and Beltran et al. (1994) used O3 or

O3/H2O2 as oxidation processes.

67

Table 5.4. Second-order rate constants of the pesticides with the OH-radicals k•OH/C (M-1 s-1) obtained from this study versus literature.

Metaldehyde Clopyralid Mecoprop

k•OH/C

(M-1 s-1)

from this study

Kinetic model

3.59x108

(3.33x108- 3.86x108)

1.96x108

(1.98x108- 1.94x108)

1.09x109

(1.11x109- 1.08x109)

k•OH/C

(M-1 s-1)

from literature

Autin thesis

(2012)

(Minakata model)

4.36x108 1.15x107 5.11x109

Autin et al.

(2012)

1.30x109 - -

Armbrust et al.

(2000) - -

2.50x109

(9.0x1012 M-1 h-1)

Beltran et al.

(1994) - - 1.90x109

Shu et al.

(2013)

- - 6.46x109± 0.04

68

4.1.4 Background water matrix effect

The effect of the presence of a matrix (pharmaceuticals, NOM, nitrate and bromide) on the

degradation kinetics of the three pesticides by LP-UV/H2O2 treatment was studied via the

experiments described in section 4.3.1 and Table 4.1 (part b and c). Figure 5.5 shows the

regression curves obtained from these experiments, compared to the single-solute

experiments, exhibiting the combined, complex effect of different parameters (presence of

matrix, initial pesticide concentration and hydrogen peroxide dose).

The presence of a matrix hindered the kinetics of mecoprop, compared to when no matrix

was present (squares versus triangles). This was attributed to two factors: a) the consumption

of the biggest part of UV light by the background matrix, resulting in reduced photolysis of

mecoprop, as well as smaller hydroxyl radical production from H2O2, and b) the scavenging

of the produced OH-radicals by the background matrix. Clopyralid and metaldehyde are not

susceptible to photolysis, so their degradation depends entirely on OH radical oxidation.

Clopyralid degradation was slowed down by the presence of the matrix, indicating that the

OH-scavenging by the matrix hindered the kinetics. On the other hand, metaldehyde was not

significantly affected by the matrix, which might be due to the molecule’s attack by other

oxidative species formed by the photolysis of NOM. Another possibility is that the effect of

the specific matrix was too small in this case to impact the kinetics significantly.

Degradation was faster for metaldehyde and clopyralid as the hydrogen peroxide dose

increased (rhombuses and circles) even though the same matrix was present and the

concentration of the compound was 3 times lower. This may indicate that the presence of this

specific matrix may not be affecting the kinetics as much as the radical production that plays

the key role for the oxidation of these compounds. Specifically for mecoprop, the higher

H2O2 doses (25 and 50 mg/L) actually slowed down the kinetics (rhombuses and circles

which gave identical kinetic profile) compared to 15 mg/L H2O2 with and without a matrix

(triangles and squares, respectively) (as shown in Figure 5.5), which is not surprising, given

the fact that mecoprop was the least enhanced by H2O2, and could point to OH-radical

scavenging from H2O2 itself. It should be noted that only two data points were available for

the regression lines of the 25 and 50 mg/L H2O2 dosages (rhombuses and circles) (Figure

5.5a-c), thus increasing the uncertainty of the results. Moreover, the data for mecoprop for

these regression lines did not follow a linear pattern, therefore a constant slope could not be

defined.

69

Nitrite and bromate ions were monitored for all experiments. Most combinations gave no

detectable presence of nitrite, with only the experiments with the 25 and 50 mg H2O2/L

(Table 4.1, part c) producing 0.06-0.08 mg NO2-/L. No detectable bromate formation was

observed under any conditions applied. This is due to the hydrogen peroxide present which is

expected to hinder the further oxidation of the intermediate HOBr/-OBr (formed by the OH-

attack of Br-) to bromate (BrO3-).

R² = 0.9861

R² = 0.9929

R² = 0.7

R² = 0.7

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

0 200 400 600 800 1000 1200

ln(C/Co)

UV fluence (mJ/cm2)

a

300μg/L mecoprop

matrix-free 15mg/L

H2O2300μg/L mecoprop

matrix-present 15mg/L





H2O2

300μg/L mecoprop matrix-

free (15mg/L H2O2)


present (15mg/L H2O2)





R² = 0.9888

R² = 0.9377

R² = 0.9686

R² = 0.9941

0.0

0.5

1.0

1.5

2.0

2.5

0 200 400 600 800 1000 1200

ln(C/Co)

UV fluence (mJ/cm2)

b

300μg/L metaldehyde

matrix-free 15mg/L

H2O2300μg/L metaldehyde






H2O2

300μg/L metaldehyde matrix-

free (15mg/L H2O2)







70

Figure 5.5. Comparison of the kinetics of a) mecoprop, b) metaldehyde and c) clopyralid,

with and without the presence of a matrix, with different initial concentrations and varying

hydrogen peroxide doses.

4.1.5 Reaction product identification

A mechanism for the degradation of each pesticide under LP-UV photolysis combined with

OH-attack (and also of LP-UV photolysis in the case of mecoprop) was proposed. The

identification of the reaction products was mainly based on the structural information

obtained from the fragmentation patterns of the mass spectrometry applied, in combination

with the hydroxyl radical chemistry. OH-radical attack followed two main mechanisms; 1) H-

atom abstraction from a C-H bond by •OH reactions, and 2) electrophilic substitution by the

•OH, in cases of double bonds and aromatic systems. Groups attached to the aromatic rings

can either activate or deactivate the aromatic ring towards electrophilic substitution (Clayden

et al., 2012).

The double-bond-equivalents (DBE) parameter was also used to determine unknown

structures. DBE is a well-established tool in mass spectrometry that determines the

summation of the double bonds plus rings in a molecule, calculated via a specific molecular

formula (Nassar and Talaat, 2004). The verification of the final structure is possible in

combination with the use of Nuclear Magnetic Resonance (NMR), Infrared (IR) spectroscopy

and mass spectrometry (only the latter applied in this study). Starting from the parent

compound, the DBE was calculated for all the possible molecular structures of the products

R² = 0.9903

R² = 0.9823

R² = 0.9842

R² = 0.9649

0.0

0.4

0.8

1.2

1.6

2.0

0 200 400 600 800 1000 1200

ln(C/Co)

UV fluence (mJ/cm2)

c

300μg/L clopyralid

matrix-free 15mg/L

H2O2300μg/L clopyralid






H2O2

300μg/L clopyralid matrix-

free (15mg/L H2O2)







71

detected, and the most likely structure for each product was suggested. In cases where the

molecular structures resulted either in a negative or fraction DBE value (which is not possible

since DBE has to be a positive integer) or the respective structures would not match the

oxidation pathways expected from the OH-attack, the molecular structure was rejected.

Following this process for all products detected for all three pesticides, the structure

elucidation process was facilitated.

The main photo-product of LP-UV photolysis of mecoprop, 2-(4-hydroxy-2-

methylphenoxy)propanoic acid (Product I, m/z=196), was produced via photo-hydrolysis of

the chlorine atom on the benzene ring (Figure A.3, Appendix A). This product had the

highest peak area in the chromatograms and increased with increasing UV fluence (no H2O2

present). It was also produced by LP-UV/H2O2 treatment but with a smaller area, decreasing

at UV fluences higher than 1000 mJ/cm2. This product has also been reported as the main

photo-product by UV photolysis at 254 nm, by Meunier and Boule (2000) and Boule et al.

(2002). Subsequent OH-attack on the tertiary carbon atom of the propanoic moiety led to H-

abstraction and carbon-centered radical formation, which reacted with oxygen and formed 2-

methylhydroquinone (Product II, m/z=124) and pyruvic acid (Product III, m/z=88). Direct

OH-attack on the tertiary carbon atom of the propanoic moiety of mecoprop led to 2-methyl-

4-chlorophenol (Product IV, m/z=142) and pyruvic acid. A product with an m/z of 108

suggests formation of 2-cresol (Product V), although radical-radical reactions in aquatic

solution are very unlikely. Another possibility would be the reduction by the hydroperoxyl

radical (HO2•) formed by photolysis of hydrogen peroxide. The formation of this product in

small concentrations by direct LP-UV photolysis has been reported before, while the rest of

the products mentioned have also been observed after either UV photolysis or UV/TiO2

oxidation (Topalov et al. 1999, Meunier and Boule 2000, Boule et al., 2002). De-chlorination

of the molecule was confirmed by the formation of chloride ion; 94% of the organic chlorine

was converted into chloride ion after three hours of irradiation (UV fluence 1500 mJ/cm2)

with 30 mg/L H2O2 (Figure A.4a, Appendix A) with a chloride ion concentration in the

solution up to 1.5 mg/L. Products from electrophilic substitution of the aromatic ring were

not observed, suggesting that this mechanism is not favoured. Figure 5.6 shows the

degradation of mecoprop (10 mg/L) and the reaction product formation as a function of the

UV fluence for a H2O2 dose of 30 mg/L. The proposed degradation pathways for mecoprop

are shown in Figure 5.7.

72

Figure 5.6. Mecoprop (10 mg/L) degradation and reaction product formation as a function of


mecoprop (dashed line)).

0

2000

4000

6000

8000

10000

12000

0

2000

4000

6000

8000

10000

0 500 1000 1500

Mecoprop

concentration

(μg/L)

Peak area

UV fluence (mJ/cm2)

pyruvic acid

2-cresol

2-(4-hydroxy-2-

methylphenoxy)

propanoic acid

mecoprop

2-

methylhydroqui

none

73

Figure 5.7. Proposed pathways for mecoprop degradation by LP-UV/H2O2 treatment.

The degradation of clopyralid by LP-UV/H2O2 treatment (Figure 5.8) yielded two main

reaction products, 3,6-dichloro-4-hydroxypyridine-2-carboxylic acid and 3,6-dichloro-5-

hydroxypyridine-2-carboxylic acid (Products I, m/z=208 for both structures) by OH-attack

onto one of the two available carbon atoms of the pyridine ring, causing H-abstraction and

carbon-centered radical formation, which after reacting with oxygen yielded the mentioned

product. 3,6-dichloro-5-hydroxypyridine-2-carboxylic acid was the main product, in

agreement with the directing and activating effects of the substituents of the ring (Clayden et

al., 2012). These two products have been reported as well for clopyralid oxidation by MP-

UV/TiO2 oxidation (Šojić et al., 2009). A peak of m/z=164, probably 3,6-dichloro-pyridin-2-

ol (Product II), was observed at a UV fluence of 500 mJ/cm2, slowly decreasing with

increasing UV fluence; this compound was identified also by Šojić et al. (2009). However,

formation of this compound is theoretically unlikely, since it suggests radical-radical

reactions. Finally, a chloro-hydroxypyridine carboxylic acid with m/z=174 was observed

(Product III) with the largest peak area at a UV fluence of 500 mJ/cm2, followed by

74

degradation at increasing UV fluence. This product may have been formed by de-chlorination

and OH-radical substitution on one of the two chlorine-bearing carbon atoms of the pyridine

ring. Figure 5.9 shows the degradation of clopyralid (20 mg/L) and the reaction product

formation as a function of the UV fluence for a H2O2 dose of 60 mg/L. The de-chlorination of

the molecule was supported by the increasing chloride concentration in the solution. The

chloride concentration increased with increasing UV and H2O2 doses and the conversion of

organic chlorine to chloride ion reached 64% (4.7 mg/L Cl- released) at a fluence of 1500

mJ/cm2 with 60 mg/L H2O2 (Figure A.4b, Appendix A). De-chlorination products were also

found in the case of clopyralid treatment by MP-UV/H2O2 treatment (Xu et al., 2013). The

proposed degradation pathways for clopyralid are shown in Figure 5.9.

Figure 5.8. Clopyralid (20 mg/L) degradation and reaction product formation as a function of


clopyralid (dashed line)).

0

6000

12000

18000

24000

0

2000

4000

6000

8000

10000

12000

0 500 1000 1500

Clopyralid

Concentration

(μg/L)

Peak area

UV dose (mJ/cm2)

3,6-dichloro-5-

hydroxypyridine

-2-carboxylic

acidclopyralid

3,6-dichloro-4-

hydroxypyridine

-2-carboxylic

acid3,6-dichloro-

pyridin-2-ol

chloro-

hydroxypyridine

carboxylic acid

75

Figure 5.9. Proposed pathways for clopyralid degradation by LP-UV/H2O2 treatment.

The degradation of metaldehyde by LP-UV/H2O2 treatment yielded quite a different set of

reaction products compared to the other two pesticides, since the molecule is a cyclic

polyether lacking aromaticity. The two main reaction products with m/z 210 and 226 were

detected twice, suggesting that two isomers with the same molecular weight were formed for

each m/z observed. The formation profile for the m/z 210 product can be seen in Figure A.5

(Appendix A), and the same profile was found for the m/z 226 product. The main products

observed for the degradation of metaldehyde (Figure 5.10) suggest successive reactions that

could be supported by the following mechanism: metaldehyde becomes hydroxylated once

(m/z=210) or twice (m/z=226) via OH-attack on the tertiary carbon atoms of the ring. Due to

the analytical detection method (positive mode detection with NH4+) formation of a complex

via hydrogen bonding between the oxygen atoms of the hydroxylated molecule and the NH4+

group, similarly to a crown ether (Hurtado et al., 2012), is thought to have taken place

(Figure 5.11). These products exhibited specific fragmentation with the production of a

76

fragment with m/z=62 identified as the [CH3CHO---H-NH3] complex. This complex

suggests the formation of acetaldehyde, which was the main expected product. Since

carbonyl compounds could not be detected with the method used, acetaldehyde was not

detected directly. Acetaldehyde formation was also supported by the detection of acetic acid

as an end product at UV fluences higher than 500 mJ/cm2 (Figure A.6, Appendix A).

Another possible mechanism could proceed through de-protonation of the CH3-bearing

carbon atom followed by OO addition (Auzmendi-Murua and Bozzelli, 2014) forming a

molecule with a keto-ether on the one end and an oxyl radical on the other, whereas de-

protonation of the methyl groups (less likely) would lead to demethylation.

Little mineralization was observed during the LP-UV/H2O2 treatment of the studied

pesticides. The TOC levels for mecoprop and metaldehyde exhibited a small decrease under

the conditions applied; the maximum TOC reductions observed under the maximum UV

fluence and H2O2 doses were 21% and 17%, respectively. For clopyralid the TOC content in

the solution showed a gradual decrease with increasing UV fluence and H2O2 dose, and

reached a maximum TOC reduction of 34% under the highest UV/H2O2 doses, applied in this

research study.

Figure 5.10. Metaldehyde (5 mg/L) degradation and reaction product formation as a function

of the UV fluence (30 mg/L H2O2; secondary vertical axis refers to the concentration of

metaldehyde (dashed line)).

0

1000

2000

3000

4000

5000

6000

0

50000

100000

150000

200000

250000

300000

0 500 1000 1500

Metaldehyde

concentration

(μg/L)

Peak area

UV fluence (mJ/cm2)

210(1)

210(2)

226(1)

226(2)

metaldehyde

77

Figure 5.11. Proposed pathways for metaldehyde degradation by LP-UV/H2O2 treatment.

A more complete investigation of the reaction products would ideally include their

quantification. Nevertheless, due to time restrictions of the project, this study was limited to

be mainly observational (i.e. seeking to identify the main likely reaction products) rather than

a quantitative determination of reaction product yields. Furthermore, the lack of standards

available for all the products was also a limiting factor towards that direction; standards were

available only for the low molecular weight products for mecoprop but not for the main

photolysis product, and no standards were available for the products of clopyralid and

metaldehyde (for metaldehyde the products were also formed partly due to the NH4+ present

in the analytical method used).

5.2 Nitrate related issues by LP-UV/H2O2

5.2.1 Effect of initial NO3- concentration and H2O2 dose

Preliminary experiments were carried out in NOM-free water, to determine the effect of the

initial nitrate concentration and hydrogen peroxide dose on nitrite formation. The literature

contains conflicting suggestions regarding the effect of H2O2 concentration in such

78

experiments; Lu et al. (2009) showed that under LP-UV/H2O2 the nitrite yield was reduced

by the addition of H2O2, agreeing with the earlier findings of KWR (2004), while Sharpless et

al. (2003) obtained opposite results by MP-UV treatment, and Derks (2010) found that nitrite

formation under LP-UV photolysis was not influenced by the addition of H2O2.

Two different concentrations of nitrate were applied; 25 mg/L and 50 mg/L (see Figure

5.12a and b). With 25 mg/L of nitrate and for the range of UV fluences and hydrogen

peroxide doses applied, nitrite was formed in the range of 0.06-0.09 mg NO2-/L. When the

initial nitrate concentration was doubled to 50 mg/L, LP-UV photolysis with a UV fluence of

2100 mJ/cm2 formed 0.09 mg/L NO2-.

In Figure 5.12b, the effect of hydrogen peroxide becomes evident. It can be seen that, when

H2O2 was absent, a UV fluence above 1500 mJ/cm2 was required to produce a nitrite

concentration above 0.05 mg/L; when H2O2 was present, a UV fluence of 1000 mJ/cm2 was

sufficient to produce a nitrite concentration in the range of 0.06-0.08 mJ/cm2 for the range of

H2O2 doses applied, suggesting that the presence of H2O2 enhanced nitrite formation. The

increasing trend reached a maximum of 0.13 mg NO2-/L with a UV fluence of 2100 mJ/cm2

and 50 mg/L H2O2. The decrease observed in nitrite formation with the addition of 10 mg/L

H2O2 was not significant.

Figure 5.12. Effect of initial nitrate concentration in NOM-free water on nitrite formation

after LP-UV/H2O2 treatment, for an initial concentration of a) 25 mg/L, and b) 50 mg/L NO3-

as a function of the UV fluence (no bar shown for specific conditions signifies concentrations

below the method detection limit (MDL), shown with a bold line at 0.05 mg/L).

0.00

0.05

0.10

0.15

0 10 25 50

NO

2-(m

g/L

)

H2O2 dose (mg/L)

a

0.00

0.05

0.10

0.15

0 10 25 50

H2O2 dose (mg/L)

b

0

500

1000

1500

2100

No UV

500 mJ/cm2

1000 mJ/cm2

1500 mJ/cm2

2100 mJ/cm2

79

The effect of hydrogen peroxide can be explained by the chain of reactions in which the

hydroxyl radical is a key component; these can be seen from the reactions starting with (r18)

and (r21) in section 2.4.3. Additionally to the nitrite formation by hydroxyl radicals,

hydrogen peroxide can also initiate nitrite production via reactions (r32-r34). It should also be

noted that experiments with hydrogen peroxide in the absence of UV light (i.e. molecular

hydrogen peroxide only, without hydroxyl radicals) did not have any impact on nitrite

formation.

5.2.2 Effect of NOM presence on nitrite formation

The addition of Suwannee River NOM (4 mg/L) to the nitrate solutions resulted in the nitrite

levels after LP-UV/H2O2 treatment indicated in Figure 5.13. Formation of nitrite was

observed for UV fluences of 1000 mJ/cm2 and above, with or without hydrogen peroxide.

Increase in UV fluence resulted in an increase in the NO2– formation, for UV fluences above

500 mJ/cm2. For the UV fluences of 1500 and 2100 mJ/cm2, LP-UV/H2O2 treatment for the

highest hydrogen peroxide dose (50 mg/L) produced the same NO2– levels as UV photolysis

(0.15 mg/L), whereas the intermediate hydrogen peroxide doses (10 and 25 mg/L) led to

lower NO2– levels.

80


presence of Suwannee River NOM (4 mg/L) (no bar shown for specific conditions signifies

concentrations below the method detection limit (MDL), shown with a bold line at 0.05

mg/L).

In comparison with the results in the absence of NOM, the presence of NOM (4 mg/L)

increased the nitrite formation (Figure 5.14). The highest NO2– concentration in the presence

of NOM was found to be 0.15 mg/L, either by photolysis alone at the highest UV fluence

(2100 mJ/cm2) or by LP-UV/H2O2 treatment at the highest fluence and hydrogen peroxide

dose (50 mg/L). The presence of hydrogen peroxide was found to have an effect on nitrite

formation when NOM was present; for a UV fluence of 1500 mJ/cm2, either 10 or 25 mg/L

H2O2 compared to its absence produced significant changes in the nitrite concentration

formed; a significant nitrite increase was observed as well when the hydrogen peroxide dose

was increased from 10 to 50 mg/L. For a UV fluence of 2100 mJ/cm2 only the difference in

nitrite yield between the hydrogen peroxide doses of 25 and 50 mg/L was significant.

0.000

0.050

0.100

0.150

0 500 1000 1500 2100

NO

2-(m

g/L

)

UV fluence (mJ/cm2)

No peroxide

10ppm peroxide

25ppm peroxide

50ppm peroxide

No H2O2

10mg/L H2O2

25mg/L H2O2

50mg/L H2O2

81

Figure 5.14. Nitrite formation after LP-UV/H2O2 treatment of nitrate (50 mg/L) in the a)

absence and b) presence of Suwannee River NOM (4 mg/L) (no bar shown for specific

conditions signifies concentrations below the method detection limit (MDL), shown with a

bold line at 0.05 mg/L).

Sharpless and Linden (2001) also studied the effect of NOM on the nitrite yield by LP-UV

photolysis; they used Suwannee River NOM at an equivalent concentration of either 5 or 15

mg/L DOC. In both cases photolysis at 254 nm resulted in increased nitrite quantum yield

compared to the tests in the absence of DOC. More specifically, they observed a decreasing

trend with no DOC present (0.028-0.016 mol/Es) for an increasing amount of absorbed UV

light, whereas a fairly steady NO2– production rate was observed for the 5 mg/L DOC (0.053-

0.056 mol/Es). Tripling the DOC concentration to 15 mg/L resulted in an increasing

production rate (0.056-0.072 mol/Es) for an increasing amount of absorbed UV light. Their

observed effect of the presence of NOM on nitrite formation agrees with the findings with

Suwannee River NOM. Two major differences exist between the work of Sharpless and

Linden (2001) and this one: (i) in the present study hydrogen peroxide was utilized, rendering

the reaction pathways of nitrite formation more complicated than by just UV photolysis; and

(ii) the nitrate concentration they used of around 0.5 mg/L was low compared to the 50 mg/L

used here. The latter could explain why the nitrite formation trend was clearly increasing with

increasing UV fluence, while they observed a slightly decreasing trend.

Regarding the effect of the presence of OH scavengers on nitrite formation, the explanation

presented by Mark et al. (1996) is based on the hindrance of the recombination reaction

between OH and NO2 (reaction r27; section 2.4.3) by the OH scavengers, that would

otherwise lead to dissociation of the intermediate HOONO to NO3- (reaction r27), leaving the

0.00

0.05

0.10

0.15

0 10 25 50

NO

2-(m

g/L

)

H2O2 dose (mg/L)

a

0.00

0.05

0.10

0.15

0 10 25 50

H2O2 dose (mg/L)

b

0

500

1000

1500

2100

No UV

500 mJ/cm2

1000 mJ/cm2

1500 mJ/cm2

2100 mJ/cm2

82

excess NO2 to form additional NO2− through the N2O4 intermediate (reactions r23 and r30 in

the section 2.4.3). This effect is reduced at more alkaline pH since the relatively stable

ONOO− greatly slows the rate of the dissociation reaction. Enhanced nitrite formation can

also be attributed to the inhibition of reaction (r18) by OH scavengers, resulting in protection

of the ONOO− and NO2− formed during NO3

− photolysis from attack by OH.

There may be two more mechanisms arising from the photolysis of NOM and contributing to

nitrite formation: (i) via O2- and OH radicals, and (ii) via solvated electrons. The O2

∙- and

∙OH radicals and the solvated electrons are already known photoproducts of NOM (Cooper et

al., 1989) and reactions (r36)-(r47) show their formation.

Concerning the first mechanism, O2∙- and ∙OH radicals are also produced by nitrate photolysis

in the absence of NOM and as explained previously they play an important role in the

reactions leading to nitrite formation. The ∙OH radicals contribute to the production of

ONOO∙ (reaction r18), which in turn produces NO∙ (reaction r20) which, via reaction with the

∙OH radicals, leads to the production of NO2- (reaction r21). The ∙OH radicals lead to NO2

∙

formation via reaction (r22) producing NO2- directly via reactions (r25)-(r26) but also

indirectly via reactions (r29)-(r31) by producing the intermediates N2O3 and N2O4 (reactions

r23-r24). Compared to reactions (r29) and (r30), reaction (r31) is faster by six and five orders

of magnitude, respectively. Additionally, O2- radicals can enhance NO2

- production via

reaction (r34).

The second mechanism involves the solvated electrons (e-s) formed via NOM photolysis

(reaction r40) as reducing agents towards nitrate. Zepp et al. (1987) and Cook et al. (2001)

both stated that nitrate, besides oxygen, is one of the most effective scavengers of electrons.

Chen and Freeman (1995) and Tarr (2003) refer to the basic reactions, and Cook et al. (2001)

gave an overview of the reactions taking place between nitrate and the solvated electrons,

where NO3,s- (s stands for solvated) produces the NO3,s

2- as intermediate and finally NO2

-,s

(reactions r45-r47).

It should be noted that experiments with NOM in the absence of nitrate did not lead to nitrite

formation by either LP-UV photolysis or LP-UV/H2O2 treatment under any dose

combination, indicating that the nitrogen content of NOMs is not a source leading to nitrite

formation. Also, no nitrite formation was observed in dark experiments (i.e. no UV light

applied). Considering the reactions involved in nitrate photolysis in the presence of H2O2 and

83

OH-radical scavengers, and also of the photolysis of NOM itself, it can be hypothesized that

the presence of NOM leads to an increased nitrite formation, when compared to the same

water matrices without NOM. The results of this study support this hypothesis, showing that

the effect of NOM on the process studied here is more complex than just scavenging OH-

radicals.

On a related note, the molar absorption coefficient (ε) of nitrite (εNO2,254=7 L/mol cm) is

nearly double the molar absorption coefficient of nitrate (εNO3,254=4 L/mol cm).

Consequently, the kinetics of the nitrite photolysis will be slowed proportionally more than

the photolysis of nitrate when NOM is present (compared to no NOM), and hence nitrite will

be ‘‘preserved’’, assuming that only photolysis takes place and no oxidation is involved.

5.2.3 Effect of pH in the presence of NOM

Alkaline pH conditions favour the formation of NO2- by NO3

- photolysis (Mark et al. 1996,

Lu et al. 2009). Thus, experiments involving Nordic Lake NOM resulted in greater NO2–

levels under pH=8 than pH=6 (Figure 5.15), and similar findings have been shown in the

presence of NOM (Sharpless and Linden, 2001). For the LP-UV/H2O2 treatment, nitrite

formation was also greater at pH=8 compared to pH=6, except at the lowest hydrogen

peroxide dose of 10 mg/L, when the formation was the same (~0.11 mg/L). For the hydrogen

peroxide doses of 25 and 50 mg/L, the nitrite concentrations between the two pH values

varied significantly.

84

Figure 5.15. Nitrite formation for two different pH values (6 and 8) after LP-UV/H2O2

treatment of nitrate (50 mg/L) in the presence of Nordic Lake NOM (4 mg/L). LP-UV

fluence was 2100 mJ/cm2 for all H2O2 doses applied. The detection limit is shown with a bold

line (0.05 mg/L).

The explanation behind this observation lies in the peroxynitrite formation (reactions (r11),

(r14) in section 2.4.2) which is essential for nitrite formation and its conjugate base, the

peroxynitrous acid (ONOOH) with a pKa of 6.6 (Sharpless et al. 2003, Goldstein et al. 2005).

At alkaline pH values the ONOOH/ONOO- equilibrium shifts to ONOO- which is the main

intermediate for NO2– production (KWR, 2004). At pH=8 only 5% is present as ONOOH

(ONOO-/ONOOH ratio 95/5 %) (Figure 2.3) and thus recombination to nitrate via reaction

(r15) is limited. At pH=6 ONOOH is dominant, with the ONOO-/ONOOH ratio equal to

15/85 (%), favouring the recombination to nitrate, and thereby reducing the nitrite yield.

It should be noted that at pH=8 the initial bicarbonate concentration in the solution would be

higher than at pH=6 due to higher CO2 dissolution (as H2CO3) from the atmosphere.

According to the carbonate species equilibrium (H2CO3⇄HCO3-⇄CO3

2-) (Greenwood and

Earnshaw, 1997), at pH=6 in pure water 70% is expected to be H2CO3 and 30% HCO3-,

whereas at pH=8 the HCO3- reaches 97%. Nevertheless, inorganic carbon analysis revealed

that the concentration of inorganic carbon in the samples was elevated only for the highest

UV fluence (2000 mJ/cm2 corresponding to around 5 hours of irradiation time). Figure 5.15

shows that, despite this, at pH=8 the nitrite concentration produced was still higher than at

0

0.05

0.1

0.15

pH 6 pH 8

NO

2-(m

g/L

) No H2O2

10 ppm

H2O225 ppm

H2O250 ppm

H2O2

No H2O2

10mg/L H2O2

25mg/L H2O2

50mg/L H2O2

85

pH=6. Therefore, taking into account that some OH scavenging is expected for pH=8, an

even higher nitrite formation than the one we have reported should be expected.

5.2.4 Effect of NOM type on nitrite formation

The effect of the type of NOM on nitrite formation was studied with the use of Suwannee

River NOM, Nordic Lake NOM and Pony Lake NOM. As stated previously, these NOMs

differ in terms of composition and aromaticity that are reflected by the following parameters:

UV absorbance, Specific Ultraviolet Absorbance (SUVA254 = A254/DOC), the A254/A203 ratio

and nitrogen content. These parameters can play a role in nitrite formation by LP-UV/H2O2

since most of them are related to the UV light absorbed and the scavenging of the OH

radicals involved in the nitrate photolysis (Weishaar et al. 2003, Sarathy and Mohseni 2007);

the values obtained for these parameters for the three NOMs are shown in Table 5.5.

Table 5.5. Physicochemical properties of the three NOMs used in this study. All parameters

were measured in this work, apart from the N-content (% w/w) that was obtained from the

IHSS (values from duplicates for DOC are given in brackets).

NOM type A254

(cm-1)

DOC

(mg/L)

SUVA254

(L/mg m)

𝐀𝟐𝟓𝟒𝐀𝟐𝟎𝟑

N-content

(%w/w)

Suwannee

River NOM 0.075

2.12

(2.07-2.18)

3.6 0.42 1.27

Nordic Lake

NOM 0.041

1.69

(1.52-1.86)

2.4 0.27 1.10

Pony Lake

NOM 0.029

1.75

(1.73-1.77)

1.6 0.32 6.51

The UV absorption spectra of the three NOMs (range 190-350 nm) are shown in Figure A.7

(Appendix A). At 254 nm, the UV absorbance for Suwannee River (0.075 cm-1) is 1.8-fold

and 2.6-fold higher than that for Nordic and Pony Lake, which were 0.041 and 0.029 cm-1,

respectively. The SUVA254, which provides a quantitative measure of the aromatic content

per unit of organic carbon content, also gives information about the nature and reactivity of

86

the NOM (Karanfil et al. 2002, EPA 2012). The SUVA254 values obtained, 3.6, 2.4 and 1.6 L/

mg m for Suwannee, Nordic and Pony Lake, respectively, show that both Suwannee NOM

and Nordic NOM are transphilic in nature, probably consisting of a mixture of hydrophobic

humic and hydrophilic non-humic matter, with intermediate molecular weight DOM

fractions, whereas Pony Lake is mostly hydrophilic non-humic (fulvic) matter. According to

Korshin et al. (1997), a high A254/A203 ratio (> 0.40), where A254 and A203 correspond to the

electronic transfer (ET) and the benzenoid band (BZ) of the NOM, suggests aromatic rings

highly substituted with hydroxyl, carbonyl, ester and carboxyl groups, whereas a low ratio

suggests that the rings are predominantly substituted with aliphatic functional groups. The

A254/A203 ratios for Suwannee, Nordic and Pony Lake NOM were determined as 0.42, 0.27

and 0.32, respectively.

A high UV absorbance of the NOM at 254 nm leads to an increased formation of photo-

excited species that in turn generate other radical species through reactions (r36)-(r44)

(section 2.4.4). These radicals can enhance nitrite formation, as explained previously. A high

SUVA value suggests that Suwannee NOM is more prone to OH-radical attack than the other

two NOMs, producing oxidation products and oxidative radical species, which in turn favour

nitrite formation. At the same time, scavenging of OH-radicals is also expected to be higher

for Suwannee NOM, since high SUVA values and unsaturated bonds are known to correlate

positively with reactions involving OH-radical or other oxidants (Westerhoff et al., 1998); as

shown in reactions (r18), (r21) and (r22) in section 2.4.3, the OH-radical is enabling the

pathways for nitrite formation, therefore this scavenging can impede nitrite formation. The

A254/A203 ratio was higher for Suwannee River NOM than for the other two NOMs,

suggesting once again susceptibility to OH-radical attack. It is clear therefore that these

reactions which take place in parallel can have either a negative or positive effect on nitrite

formation, making it difficult to predict what the overall effect will be.

The differences in NOM composition were confirmed by LC-OCD fractionation for the three

NOMs before and after treatment with 2100 mJ/cm2 and 50 mg/L H2O2 (Figure 5.16). An

attempt was made to relate the composition of the NOMs with the nitrite levels produced but

no firm conclusions could be drawn. In Figure 5.16 it can be seen that the humic acids

content, which is mainly responsible for the aromaticity of the NOM, was for Suwannee

River NOM 1.6-fold and 1.4-fold the content of Nordic and Pony Lake NOM, respectively.

The initial hydrophobic part (HOC), which hydroxyl radicals preferentially react with due to

87

its higher aromaticity (Sarathy and Mohseni, 2009), was almost three times higher for

Suwannee NOM compared to Nordic Lake, while in Pony Lake NOM it was non-existent.

After LP-UV/H2O2 treatment, the humic acid content exhibited a 30% and 25% decrease for

Suwannee and Nordic Lake NOM, respectively; on the other hand, almost no decrease was

observed for Pony Lake NOM (2%). This finding can be attributed to the higher aromaticity

of the first two NOMs and consequently the increased reactivity of the humics towards

photolysis and oxidation during the treatment, compared to the less reactive and less

photolabile hydrophilic, non-humic (fulvic) matter of Pony Lake NOM.

After LP-UV/H2O2 treatment, the highest nitrite concentration (0.15 mg NO2-/L) for

Suwannee River NOM, as stated previously, corresponded to the highest hydrogen peroxide

dose (50 mg/L). For the Nordic Lake and Pony Lake experiments the nitrite levels were only

slightly higher for the 2100 mJ/cm2 compared to the 1500 mJ/cm2. Comparing the results

overall for the three NOMs after LP-UV/H2O2 treatment with 2100 mJ/cm2 and 50 mg/L

H2O2, no significant variations were exhibited and the same range of nitrite concentrations

(between 0.08 and 0.15 mg/L) was obtained (Figure 5.17a-c) despite their different

characteristics. This suggests that the radical scavenging from the NOMs plays a bigger role

during the UV/H2O2 process than the ‘inner filter’ effect caused by the UV light absorption.

On the other hand, nitrite formation is influenced to a larger extent by the UV light available

and less by the OH-radicals formed during the process. Since the same amount of light is

available in all experiments (i.e. the irradiation time was adjusted according to the UV

absorbance of the sample to achieve the desired UV fluence), it was hypothesized that the

highest radical species production was expected for the case of Suwannee River NOM

compared to the other NOMs due to its higher aromatic content. Therefore, an enhancement

in nitrite production was expected; however, this effect could have been counteracted by

other factors, e.g. increased OH-scavenging, which, as explained previously, leads to the

hindering of nitrite formation. Also, the LC-OCD analysis suggested that there were no

important changes in the fractions of the three NOMs. Therefore it can be concluded that

although the presence of NOM leads to increased nitrite formation via pathways caused by

the radical species formed, the structural composition of the NOM, its aromaticity and

nitrogen content did not play a significant role in the levels of nitrite produced under the

conditions applied in this study.

88

Figure 5.16. LC-OCD fractionation: a) Suwannee River NOM, b) Nordic Lake NOM, c) Pony Lake NOM, before and after LP-UV/H2O2

treatment (UV fluence: 2100 mJ/cm2, H2O2 dose: 50 mg/L). Initial DOC concentrations of 2.1, 1.7 and 1.8 mg C/L for Suwannee NOM, Nordic

NOM and Pony Lake NOM, respectively. C: Carbon, N: Nitrogen, LMW: Low Molecular Weight, DOC: Dissolved Organic Carbon, HOC:

Hydrophobic Organic Carbon.

0.00

0.50

1.00

1.50

2.00

2.50

Co

nce

ntr

ati

on

(m

g/L

)a

0.00

0.50

1.00

1.50

2.00

2.50

b

0.00

0.50

1.00

1.50

2.00

2.50

c

Before

treatment

After

treatment

89

Figure 5.17. Nitrite formation after LP-UV/H2O2 treatment of nitrate (50 mg/L) in the presence of a) Suwannee NOM, b) Nordic Lake NOM,

and c) Pony Lake NOM (pH=8) (no bar shown for specific conditions signifies concentrations below the detection limit, shown with a bold line

at 0.05 mg/L).

0.00

0.05

0.10

0.15

0.20

0 1500 2100

NO

2-(m

g/L

)

UV fluence (mJ/cm2)

a

0.00

0.05

0.10

0.15

0.20

0 1500 2100

UV fluence (mJ/cm2)

b

0.00

0.05

0.10

0.15

0.20

0 1500 2100

UV fluence (mJ/cm2)

c

No peroxide

10 mg/L

peroxide25 mg/L

peroxide50 mg/L

peroxide

No H2O2

10mg/L H2O2

25mg/L H2O2

50mg/L H2O2

< MDL < MDL < MDL

90

5.2.5 Mutagenicity from NOM-NO3- interactions in synthetic water

Ames II testing was performed to assess the mutagenicity of the treated water, using two

different types of NOM, Pony Lake and Suwannee River NOM, in the absence and presence

of nitrate (50 mg/L) under the following conditions: UV fluences of 0 (no irradiation), 1500

and 2000 mJ/cm2 and 15 mg/L H2O2. The results for the water samples, including the method

and solvent control, are given in Figure 5.18 (a-b).

Figure 5.18. Number of positive wells generated by the Ames II test under different

experimental conditions with a) Suwannee River, and b) Pony Lake NOM (H2O2

concentration 15 mg/L). The positive control (2-nitrofluorene + 4-nitroquinoline N-oxide)

produced 46.9 (±0.782) positive wells.

0

2

4

6

8

10

12

14

Method

control

Solvent

control

0 UV+ NO3 2000 UV 1500 UV +

NO3

2000 UV +

NO3

Number of

positive wells

[-]

a

0 UV+H2O2

+NO3

1500 UV

+H2O2

+NO3

2000 UV

+H2O2

+NO3

2000 UV

+H2O

0

2

4

6

8

10

12

14

Method

control

Solvent

control

0 UV+ NO3 2000 UV 1500 UV +

NO3

2000 UV +

NO3

Number of

positive wells

[-]

b

0 UV+H2O2

+NO3

1500 UV

+H2O2

+NO3

2000 UV

+H2O2

+NO3

2000 UV

+H2O

91

Comparing 0 UV+H2O2+NO3 and 2000 UV+H2O2+NO3 the Ames II test results for

Suwannee River NOM (Figure 5.18a) were not statistically different. In contrast, for Pony

Lake NOM, the 2000 UV+ H2O2+NO3 combination showed a significant increase in positive

wells levels (Figure 5.18b). Therefore it can be concluded that LP-UV photolysis and/or

hydroxyl radical oxidation was responsible for the increase in the number of positive wells. A

statistically significant increase was also observed for the 2000 UV+H2O2+NO3 combination

compared to the 2000 UV+H2O2 (no NO3), suggesting that NO3 photolysis played a major

role in the increased number of positive wells. This finding agrees with the findings from

Martijn et al. (2014, 2015) treating samples containing Pony Lake NOM with and without the

presence of nitrate with either MP-UV photolysis or MP-UV/H2O2 treatment, producing an

increase in the Ames II test response only when nitrate was present.

Overall, significant variations in the Ames II test response were observed in Pony Lake NOM

samples in the presence of nitrate after LP-UV/H2O2 treatment. Nevertheless, the levels

produced (around 10 positive wells) with a high nitrate concentration (50 mg/L) and high UV

fluence (2000 mJ/cm2) were only half the ones observed by Martijn et al. (2014) (around 20

positive wells) when MP-UV/H2O2 treatment was applied (560 mJ/cm2 with 6 mg/L H2O2).

In order to interpret the results in this aspect of the study in terms of mutagenicity with the

Ames II test, the fold increases over the solvent control and the fold increases over the

baseline value (explained in section 4.5) were calculated. Figure 5.19 shows that all fold

increases over the solvent control were below 3 and all fold increases over the baseline value

were below 2, indicating that none of the SPE extracts of the samples could be classified as

mutagenic towards the Salmonella typhimurium strain TA98. It should be noted that negative

results are not conclusive on the total absence of mutagenic potency, as other mutagenic

mechanisms are not excluded.

92


II test values for the different experimental conditions with either Pony Lake or Suwannee

River NOM (H2O2 concentration 15 mg/L).

Relating the LC-OCD fractionation results of the two NOMs treated with a UV fluence of

2100 mJ/cm2 and a H2O2 dose of 50 mg/L with the Ames response of the NOMs for the 2000

UV+H2O2 combination, the response for the Suwannee River NOM was double the response

for Pony Lake NOM (i.e. 6 and 3.3 number of positive wells, respectively, Figure 5.18),

which might be caused by the fact that the humic acid content decrease was observed for

Suwannee River but not for Pony Lake NOM (Figure 5.16). When humics are degraded by

LP-UV/H2O2, they produce lower molecular weight compounds that may give a mutagenic

0.00

0.50

1.00

1.50

2.00

2.50

3.00

0 UV+ NO3 2000 UV 1500 UV +

NO3

2000 UV +

NO3

Fold increase over

solvent control

a

Pony Lake NOM

Suwannee River

NOM

0 UV + NO30 UV+H2O2

+NO3

2000 UV

+H2O2

+NO3

0 UV+H2O2

+NO3

2000 UV

+H2O2

1500 UV

+H2O2

+NO3

0.00

0.50

1.00

1.50

2.00

0 UV+ NO3 2000 UV 1500 UV +

NO3

2000 UV +

NO3

Fold increase over

baseline

b

Pony Lake NOM

Suwannee River

NOM

2000 UV

+H2O2

+NO3

1500 UV

+H2O2

+NO3

0 UV+H2O2

+NO3

2000 UV

+H2O2

93

response. The same difference in response between Suwannee River and Pony Lake NOM

was expected for the addition of nitrate, the incorporation of which in the organic matrix is

known to produce nitrogenated (often toxic) compounds, but this was not observed; in fact

significant responses were found for Pony Lake NOM but not for Suwannee River NOM.

Thorn and Cox (2012) have shown that nitrate-nitrogen incorporation in Suwannee River

NOM indeed takes place, but nitro groups could undergo subsequent transformation to other

functional groups, producing aromatic amine, amide, lactam and nitrile compounds

(Pennington et al., 2007). These compounds might be non- or less mutagenic than the nitro-

compounds and as a consequence not enhance the Ames response compared to the responses

in the absence of nitrate.

5.2.6 Reaction product identification from NOM-NO3- interactions

While no nitrite was formed in dark samples (i.e. no UV irradiation), the range obtained for

the irradiated samples (1500/15 and 2000/15) for both NOMs was 0.08-0.09 mg/L (Figure

5.20), which was very close to the 0.1 mg/L EU regulatory limit for nitrite. In addition, the

presence of nitrophenols was investigated since their formation has been observed previously

(as mentioned in the Introduction), due to the reaction of phenol with NO3 photolysis

intermediates. Martijn et al. (2014) used Pony Lake NOM in the presence and absence of

nitrate, hypothesizing the Ames response as a result of the incorporation of inorganic nitrogen

into the organic matrix by MP-UV and MP-UV/H2O2 treatment, and subsequently observed

the formation of nitrophenols by MP-UV photolysis when phenol was used as a model

compound for NOM. Therefore, it was considered worthwhile to investigate any nitrophenol

formation in these tests stemming from reactions in the NOM-NO3 system, induced in the LP

range (i.e. 254 nm).

94

Figure 5.20. Nitrite concentrations produced by the treatment of either Pony Lake or

Suwannee River NOM (4 mg/L) in the presence of nitrate (50 mg/L) (detection limit 0.05

mg/L).

The LC-MS/MS method was used to analyse for the following nitrophenols: 2-nitrophenol +

4-nitrophenol (NP), 2,4-dinitrophenol (DNP), 2-methyl-4,6-dinitrophenol (DNOC) and

Dinoseb (DINO). All samples yielded trace amounts of nitrophenol (0.014-0.046 μg/L),

without the possibility to distinguish between 2-nitrophenol and 4-nitrophenol, since the

analytical method used only determines the total mono-nitrophenol concentration. The

experiments were performed in duplicate and the results are shown in Figure 5.21. With no

UV irradiation, 0.01 μg/L of nitrophenol was found, which was very close to the detection

limit (0.007 μg/L) and lower than the quantification limit (0.02 μg/L). In contrast, for the two

UV/H2O2 combinations, there was a significant increase in the nitrophenol concentration

(~0.04 μg/L) that was detectable above the quantification limit. The LC-MS/MS

chromatogram for the nitrophenols detected for the 2100/15 and 1500/15 combinations can

be found in Figure C.5 (Appendix C).

Nitrophenol production is expected, since phenolic groups are one of the aromatic structures

present either in the NOM or from the degradation products of NOM molecules by oxidation.

The results agree with Kolkman et al. (2015) who found a variety of nitrogen containing

compounds, including 4-nitrophenol, when MP-UV photolysis was applied in samples with

Pony Lake NOM, as well as in full-scale water samples treated with MP-UV/H2O2. The 2

0.00

0.02

0.04

0.06

0.08

0.10

1500/15 2000/15

Nitrite

(mg/L)

UV (mJ/cm2) / H2O2 (mg/L)

Pony Lake

NOM

Suwannee

River

NOM

95

phenols detected, 2- and 4-nitrophenol, were also found by Martijn et al. (2014) along with 4-

nitrocatechol, when irradiating samples containing phenol and NO3 with a MP lamp. The

measured nitrophenol concentration was less for the 2000 mJ/cm2 than for the 1500 mJ/cm2

(0.037 and 0.046 μg/L, respectively) (Figure 5.21). This finding agrees with Martijn et al.

(2014) who showed an initial increase in nitrophenol concentration as a function of the

irradiation time, followed by a decrease at longer irradiation times, that was attributed to

oxidation of the previously produced nitrophenols.

Figure 5.21. Concentrations of nitrophenol (2-nitrophenol, 4-nitrophenol or a combination of

both) produced by the LP-UV/H2O2 treatment of synthetic water samples containing Pony

Lake NOM and nitrate. The quantification and detection limits were 0.02 and 0.007 μg/L,

respectively.

5.2.7 Mutagenicity from NOM-NO3- interactions in full-scale water

Samples collected from the drinking water treatment plant after various treatment steps and at

two dates in two different periods (sampling (a) was in February and sampling (b) in April)

were also analysed with the Ames II test. In each sampling, 2 samples were collected for each

sampling point (i.e. inlet, pre-AOP, post-AOP and post-GAC). A scheme of the treatment

plant along with the 4 sampling points is given in Figure A.8 (Appendix A). The water

parameters, relevant for the LP-UV/H2O2 treatment of the WTW, for the two samplings are

given in Table 5.6.

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0/15 1500/15 2000/15

Nitrophenol

(μg/L)

UV (mJ/cm2) / H2O2 (mg/L)

96

Table 5.6. Water treatment-related parameters for the two sampling dates.

Sampling

(a)

Sampling

(b)

Temperature (ₒC) 5.0 10

pH (-) 8.2 8.5

Works inlet DOC (mg/L) 4.05 4.71

Pre-AOP DOC (mg/L) 2.68 2.11

Post-AOP DOC (mg/L) 1.52 <1.00

Inlet SUVA (L/m mg) 4.61 3.74

Pre-AOP SUVA (L/m mg) 4.70 4.18

Post-AOP SUVA (L/m mg) 2.90 *

Inlet UVT (%) 65.0 66.7

Pre-AOP UVT (%) 74.8 81.7

Post-AOP UVT (%) 90.4 92.7

UV fluence (mJ/cm2) 2000 1750

Hydrogen peroxide dose (mg/L) 40 40

Works inlet nitrate (mg/L) 30.5 31.7

Works effluent nitrate (mg/L) 30.0 30.2

* Not calculated, since the DOC value was below the detection limit (1 mg/L).

The number of positive wells from the Ames II test for the water treatment works samples are

shown in Figure 5.22. Sampling (a) showed a small but significant decrease in Ames II test

response as the water passed through the AOP step. However, in sampling (b) (when the inlet

water was also analysed) the number of positive wells significantly decreased from the inlet

to the pre-AOP step (the submerged filtration), but remained unchanged for the AOP step and

significantly increased after passing the GAC filters. Although all samples produced low

levels of positive wells compared to the positive control, the two sampling sets differed

significantly for the pre-AOP and post-AOP steps, indicating that the Ames II test response

was indeed influenced by the different water characteristics and/or treatment conditions of the

two periods.

97

Figure 5.22. Number of positive wells generated by the Ames II test for the full-scale water

samples during different stages of treatment for the two sampling dates. The positive control

(2-nitrofluorene + 4-nitroquinoline N-oxide) produced 46.9 (±0.782) positive wells.

The increased response after the post-GAC filtration found during the second sampling could

be attributed to either: a) toxic chemical compound formation on the GAC filter via

biological activity, or b) pathogen colonization of the GAC filter, e.g. by Salmonella

bacterium, both of which would induce undesired biological activity on the GAC filters.

Regarding the first factor, examples of microbiologically assisted processes like methylation

of mercury, hydroxylamine formation, pesticide-related molecule conversion to toxic

metabolites, nitrosamine formation with nitrite as a precursor and cometabolism of refractory

compounds, have been observed in laboratory models or natural water bodies. Nevertheless,

such microbiologically produced toxicants have not been identified on GAC used for

drinking water treatment production (National Academy of Sciences, 1980). According to

Camper et al. (1985), pathogens such as Salmonella can colonize and persist in the carbon

bed, especially on virgin GAC filters. Presence of Salmonella could increase the response of

the Ames test, which makes use of strains of the bacterium Salmonella typhimurium, giving a

false positive. This possibility, though, is also small since the water has reduced microbial

load, being subjected to high UV fluences before entering the GAC filters. Alternatively, an

experimental error during the second sampling can also not be ruled out.

0

2

4

6

8

10

12

14

16

Number of

positive wells

[-]

98

The results given in Figure 5.23 depict the fold increases obtained for the WTW samples. All

fold increases over the solvent control were < 3.0, and all fold increases over the baseline

value were < 2.0, indicating that none of the SPE extracts of the samples could be identified

as mutagenic towards the Salmonella typhimurium strain TA98.


II test values for the full-scale water samples during different stages of treatment for both

samplings (a) and (b).

0.00

0.50

1.00

1.50

2.00

2.50

3.00

Sampling (a) Sampling (b)

Fold increase over

solvent control

a

Inlet

Pre-AOP

Post-AOP

Post-GAC

0.00

0.50

1.00

1.50

2.00

Sampling (a) Sampling (b)

Fold increase over

baseline

b

Inlet

Pre-AOP

Post-AOP

Post-GAC

99

In an attempt to explain the differences in the Ames II test responses between the two

samplings (Figure 5.22), all samples were analysed for nitrite concentrations and LC-OCD

fractionation was performed to assess any changes in the NOM content due to the pre-

treatment (up to UV/H2O2 treatment), the UV/H2O2 treatment itself and post-GAC filtration.

The nitrite concentrations for all samples were at the detection limit (0.05 mg/L). The results

from the LC-OCD fractionation are given in Figure 5.24. For both samplings, the decrease in

the humics-C and humics-N content through all the treatment stages (inlet, pre-AOP, post-

AOP and post-GAC steps) was statistically significant (except for the post-AOP/post-GAC

for the humics-N in sampling (a)). Comparison for either humics-C or humics-N content

between the two samplings showed that the differences were significant for all the treatment

steps, except for only the GAC filtration step (for the humics-N), suggesting that the NOM

composition variations were period-dependent.

The SUVA254 values obtained (SUVA254 > 4 L/m mg) for both samplings (Table 5.6)

suggested that initially (i.e. pre-AOP) the water contained humic, aromatic and hydrophobic

matter with high molecular weight DOM fractions. The SUVA values decreased post-AOP,

suggesting that the aromatic compounds present in the water initially (which comprise a

major fraction of the humic content (Chen et al., 2002), the NOM fraction most susceptible to

photolysis and oxidation via UV/H2O2), underwent degradation to lower molecular weight

(LMW) compounds which are more likely to have sustained loss of aromaticity (e.g. loss of

cyclicity, conjugated π-system). This hypothesis was supported by the decrease in humic

content, as well as the increase in LMW acids by the LC-OCD fractionation (seen in Figure

5.24), which were found to be statistically significant pre- and post-AOP for both samplings.

The important role of GAC filtration as a pre-treatment step becomes evident from the

changes in the LC-OCD fractions (Figure 5.24) as well as the water quality parameters given

in Table 5.6. The main advantages are the decrease of the humics and the DOC content,

causing an increase of the UVT of the water, as the water passes through the roughing GAC

filters (and the ultrafiltration step) prior to the AOP treatment step. In the two samplings (a)

and (b), an increase of 13% and 18% in UVT, respectively, was observed. A higher UV

transmittance of the water entering the UV reactor results in a reduced UV light requirement

(or even fewer lamps) to achieve the same effect. A decrease of 38% and 39% in humic

content, respectively, is shown in Figure 5.24; this indicates that a big part of the aromatic

fraction in the intake water (which are susceptible to photolysis and to oxidation by the OH

100

radicals formed at the AOP step) is adsorbed by the GAC filters, rendering both more UV

light and OH radicals available for the photolysis/degradation of other organic compounds,

like the micropollutants of interest (e.g. pesticides). Finally, a significant reduction in DOC

was measured for both samplings, (a) and (b), 34% and 55%, respectively, indicating a

smaller organic load for the AOP to deal with, in addition to a reduced ‘inner filter’ effect.

Figure 5.24. LC-OCD fractionation of full-scale water samples from a) sampling (a) and b)

sampling (b). C: Carbon, N: Nitrogen, LMW: Low Molecular Weight, DOC: Dissolved

Organic Carbon.

0

1

2

3

4

5

Inlet Pre-AOP Post-AOP Post-GAC

Concentration

(mg/L)

Treatment step

a

Humic acids-C

Humic acids-N

LMW neutrals

LMW acids

0

1

2

3

4

5

Inlet Pre-AOP Post-AOP Post-GAC

Concentration

(mg/L)

Treatment step

b

Humic acids-C

Humic acids-N

LMW neutrals

LMW acids

101

6. Discussion and Conclusions

6.1 Pesticide degradation and product formation by LP-UV/H2O2

Sufficient degradation of the selected pesticides was achieved under practical UV/H2O2

doses; this finding suggests that LP-UV/H2O2 can be considered a suitable process for the

treatment of these recalcitrant pesticides in practice, despite their differences in structure and

properties, which resulted in different susceptibility in photolysis and OH-radical attack.

Assuming that their influent concentrations are in the typical hundreds ng/L to a few μg/L,

the average LP-UV/H2O2 conditions can degrade them to below permitted levels; in cases of

higher influent levels, the desired degradation can be achieved by an increase of the doses

whilst still remaining within practical ranges (UV fluences up to 1000 mJ/cm2 and H2O2 up to

15 mg/L). Furthermore, the findings of this work add to the existing database of photolysis

and oxidation-related parameters and contribute to the knowledge required in treating other

micro-pollutants with similar properties and structure with the pesticides studied that might

arise in the future.

Regarding the significance of the reaction products identified, toxicity information is

available only for 2-cresol. The EPA (IRIS) has classified 2-cresol as a possible human

carcinogen (Classification C) based on an increased incidence of skin papillomas in mice in

an initiation-promotion study (U.S. EPA, 2014). However, in water treatment the low

molecular weight compounds usually produced by oxidation processes are expected to be

biodegraded and/or adsorbed by post treatment processes (e.g. granular activated carbon

filtration).

Introduction of a matrix containing bromide, among other constituents, did not lead to

bromate formation, as expected, due to the presence of hydrogen peroxide. Bromate is a

suspected human carcinogen and is classified in Group 2B as possibly carcinogenic to

humans by the IARC (International Agency for the Research on Cancer) (International

Agency for Research on Cancer, 1999) with the provisional guideline value set by the WHO

(2011) as well as the US Environmental Protection Agency (United States Environmental

Protection Agency, 2000) of 10 μg/L. This gives an advantage of the LP-UV/H2O2 process

compared to the widely-used ozonation and ozone-related AOPs (such as O3/H2O2) which are

known to lead to bromate formation in bromide-containing waters.

102

In relation to previous studies, only the work of Autin et al. (2012 and 2013) on metaldehyde

degradation by LP-UV/H2O2 can be compared to the present work, with conflicting results on

kinetics as well as on the background water matrix impact. The discrepancy between the

kinetic rate constants was attributed to the small amount of data points and the significant

variation of the kp values from the duplicates acquired in this work, introducing an

uncertainty in the value obtained. With respect to the matrix-induced scavenging reported by

Autin et al. (2013) that hindered the degradation kinetics of metaldehyde to a small extent,

specific amino-acids were used as NOM surrogates and alkalinity was introduced, a

background that differed from the one used in this work, i.e. Suwannee River NOM plus

inorganics. The impact of the two matrices on metaldehyde degradation most possibly

differed, which is expected since alkalinity, a major factor in OH-radical scavenging, was not

considered in the present work.

The European Drinking Water Directive standards of 0.1 μg/L for each individual pesticide

and a standard of 0.5 μg/L as Total Pesticides for drinking water are not health-based

standards but were mainly set as an environmental policy to generally limit pesticides. These

standards first appeared in Directive 80/778/EEC in 1980, when little was known about the

actual risks and effects of pesticides on human health. Despite the advancement of scientific

methods and the increasing toxicological and technical information available for some of the

compounds of concern since then, the 0.1 μg/L standard is still applied to all pesticides in EU

indiscriminately; on the occasion of this fact, a discussion has started recently about a new

revision of the drinking water quality standards and regulations in EU (Dolan et al., 2013).

6.2 Nitrate related issues by LP-UV/H2O2

From the current work it was concluded that nitrite formation takes place by the LP-UV/H2O2

application, although to a lesser extent than MP-UV and MP-UV/H2O2 applications.

Parameters such as initial nitrate concentration, H2O2 dose and pH all play a role in the

complex photochemical system that leads ultimately to the formation of nitrite. The nitrite

concentration produced was found to be directly proportional to the parameters mentioned

above, as well as to the presence of a background water matrix (NOM), suggesting the

involvement of NOM-induced mechanisms as well. Concerning the presence of NOM, the

results presented in this study agree with former studies reporting an increase in nitrite

formation when specific ∙OH scavengers are present (e.g. Mark et al. 1996, Goldstein and

103

Rabani 2007), and with Sharpless and Linden (2001) concluding that the presence of NOM

resulted in higher nitrite formation compared to the yields in the absence of NOM.

Nitrite concentrations above the EU limit of 0.1 mg NO2-/L were observed from nitrate

photolysis in the presence of NOM only under extreme UV/H2O2 conditions (initial nitrate

concentration of 50 mg/L when UV fluences above 1500 mJ/cm2 and H2O2 doses higher than

10 mg/L were applied), reducing the likelihood of related problems. However, this EU limit

applies to final treated water and not specifically to AOP-treated water. In practice, granular

activated carbon (GAC) filtration is commonly applied after AOP treatment. It has been

shown that GAC filters with biological activity present are expected to re-oxidize the formed

nitrite back to nitrate as part of the nitrification process (Martijn et al. 2007), addressing in

this way any problem that may arise concerning nitrite formation by the AOP.

Extrapolating from this study’s results to real water applications is not straightforward; it

should be noted that use of reference NOMs may not be indicative of the NOM in various

natural water samples. Few reports have dealt with the photochemical formation of nitrite

directly from NOM, with conflicting results. In the present study (with the LP-UV/H2O2), as

well as for Sharpless and Linden (2001) (utilizing the MP-UV range), irradiation experiments

with NOM in the absence of nitrate did not lead to nitrite formation, indicating that the

nitrogen content of NOMs is not a source of nitrite; however, Kieber et al. (1999) extracted

humic substances for various water samples and reported significant photo-production of

nitrite (irradiation range of 250-1000 nm), proportionally related both to the concentration of

humic substances and irradiation exposure time. This finding suggests that the humic acid-

bound nitrogen may be more photolabile in some cases and NOMs with high DON content

might exhibit a higher nitrite production.

6.3 Mutagenicity

There are only a few studies of pesticide treatment by LP-UV/H2O2 treatment that discuss

toxicity effects (Linden et al. 2004, Choi et al. 2013, Lekkerkerker-Teunissen et al. 2013,

Mariani et al. 2015), as well as a few considering toxicity of medium pressure UV treatment

(Kruithof et al. 2007, Heringa et al. 2011, Martijn and Kruithof 2012). These have generally

indicated less genotoxicity formation by low pressure compared to medium pressure UV

applications. Nevertheless, toxicity testing after LP-UV/H2O2 treatment is scarce and recent

research suggesting that background water components (e.g. nitrate and natural organic

104

matter) may play a more important role in toxicity formation by UV/H2O2 treatment than the

target micro-pollutants (Parkinson et al. 2001, Martijn and Kruithof 2012, Martijn et al. 2015)

pointed to the conclusion that more attention should be given to this matter.

From the findings of this study it was observed that the presence of nitrate in NOM-nitrate

water samples treated by LP-UV/H2O2 showed mutagenic responses with the Ames II test;

these responses were higher compared to the cases where nitrate was absent. This finding

suggests the incorporation of nitrogen in the organic matter, producing mutagenic nitrated

organics, as shown by the trace amounts of nitrophenol (2- and 4-nitrophenol) measured.

These nitrophenols have also been reported previously (Martijn et al., 2014). They are known

to have higher toxicity to organisms than the parent compound, phenol, itself, i.e. the oral

LD50 values in mice are 0.3, 1.30-2.08 and 0.38-0.47 g/kg for phenol, 2- and 4-nitrophenol,

respectively (WHO, 2000). However, under the conditions applied in this research, the

concentrations of these phenols were in the range of trace levels and are not expected to cause

health-related issues. Moreover, previous studies on 2- and 4-nitrophenol have shown that no

mutagenicity was observed when different variations of the Ames II test were applied

(USEPA, 1993).

The decrease in humic content and aromaticity observed after the AOP step for the real water

samples may be linked to the reduction of the number of positive wells observed from the

Ames II testing, at least for sampling (a) (Figure 5.22). Aromaticity, in conjunction with

other properties (such as hydrophobicity, electron deficiency, planar orientation of the

molecule), may partially contribute to the mutagenic potency of some organic compounds.

This is achieved via the facilitation of pi-driven (π–π interaction) intercalation, i.e. the

insertion of molecules between the planar bases of DNA (Bruce et al., 2008). π–π interactions

require π-systems (aromatic). Some organic compounds, such as polycyclic aromatic

hydrocarbons (PAHs), have been shown to act as efficient intercalators, therefore exhibit

mutagenic properties, as suggested by Quantitative Structure–Activity Relationship (QSAR)

modelling (Bruce et al., 2008). Bruce et al. (2008) reported a very good agreement between

the predicted QSAR responses and the actual responses obtained by the Ames bioassay.

Benzo-derivatives of PAHs, such as benzo(a)pyrene, a human carcinogen (Group 1)

according to IARC (International Agency for Research on Cancer, 2012), have been detected

previously in the raw water of the treatment plant where the samplings took place.

105

Nevertheless, these results only suggest a link between the fate of NOM in the water and the

Ames II test responses obtained, without proving a direct cause and effect relationship.

More importantly, all the Ames responses found for the wide-range experimental as well as

Ames-related conditions were below the level of concern for all samples tested, both

synthetic and full-scale ones. On this basis, LP-UV/H2O2 can be regarded as a safe process

with little reason for concern regarding human health implications. In cases where an

unexpectedly high Ames response were to occur, subsequent biological GAC filtration has

been shown to decrease it to acceptable levels (Guzzella et al. 2002, Martijn and Kruithof

2012).

6.4 MP versus LP AOP treatment

With respect to nitrite formation in MP and LP AOP applications, two recent examples from

the water industry support a lower level of concern when LP lamps are used, in contrast to

MP ones with a wider UV absorbance range for nitrate. In the case of Hall WTW of Anglian

Water, where the AOP treatment consists of the LP-UV/H2O2 combination, nitrite does not

pose a problem under the practical conditions applied, despite the heavy organic load and

high nitrate levels (40-50 mg NO3-/L)13, whereas at the Andijk III plant (PWN Technologies)

in the Netherlands where MP-UV/H2O2 is utilized, nitrite formation is a known problem,

tackled by back-end GAC filtration.

It should be noted that the role of H2O2 when nitrate photolysis takes place is complicated;

although both NO3- and H2O2 are known to absorb more in the MP-UV range, studies have

shown that OH-radical formation under LP-UV/H2O2 is actually more efficient than under

MP-UV/H2O2. Bolton (2001) calculated that H2O2 absorbs 21.4% of the UV-light (total

‘photon flow’) from a LP lamp, compared to 15.5% from an MP lamp. A KWR report

(KWR, 2011) verified that ‘‘as a result of the low UV absorption of the water matrix at a

wavelength of 253.7 nm, the production of radicals from H2O2 will be higher than at lower

wavelengths, although the absorption by H2O2 will be lower at a wavelength of 253.7 nm’’.

The potential formation of undesirable nitrated organics mentioned earlier is the result of the

reaction of intermediate radicals (formed by LP-UV photolysis of nitrate) with the organic

matrix of the water samples. Since the nitrate photolysis is much more enhanced in the MP

13 Barrie Holden, Anglian Water. Personal communication, 2015.

106

region due to its absorption spectrum, especially at wavelengths below 240 nm, the formation

of radical species as well as nitrite (the end product of a complex series of photolysis

reactions) is enhanced, compared to the LP-UV (254 nm). Therefore, as expected, the

incorporation of inorganic nitrogen into the organic matrix and consequently the reaction

product formation and mutagenicity response is lower in LP-UV applications, giving little

reason for concern.

Martijn (2015) measured the Ames response of synthetic water with Pony Lake NOM after

MP-UV/H2O2 treatment (600 mJ/cm2 and 6 mg/L H2O2) and reported a yield of 25 positive

wells (Figure 7.2, Martijn (2015)) in the presence of nitrate (12 mg/L); in this reasearch

where Pony Lake was treated by LP-UV/H2O2 (1500 mJ/cm2 and 15 mg/L H2O2) the number

of positive wells was 3 times lower (8 positive wells), even though the initial nitrate

concentration was 4 times higher (50 mg/L). Applying a maximum fluence of 2000 mJ/cm2, 4

and 10 positive wells were obtained in the absence and presence of nitrate, respectively.

Therefore, even though the Ames II test responses obtained in this work were much lower

than the ones obtained by Martijn (2015) (which was expected due to the limited photolysis

of nitrate at 254 nm), the Ames II test response in nitrate-rich water was still significant.

Illustrating the significant role of nitrate for both MP and LP-UV processes, the Ames II test

response as a function of nitrite formation was shown by Martijn (2015). The obtained nitrite

data (0.08-0.09 mg NO3-/L) from the synthetic Pony Lake NOM samples were overlaid on

Figure 5.3 (from Martijn (2015)) and around 5.5 positive wells were estimated (Figure 6.1),

which is within the range of the Ames II test response that was obtained for Pony Lake NOM

for the 1500 mJ/cm2 (8 ± 3.46 positive wells) (Figure 5.18b). Although different treatment

processes (MP-UV vs LP-UV/H2O2) and different nitrate contents (12 mg/L vs 50 mg/L)

were involved, the overlap observed supports the earlier indication of the impact of UV

photolysis of nitrate on both the nitrite formation and the Ames II test response (Martijn et

al., 2014), and is a confirmation of the clear relation between the two (nitrite and Ames II

response).

Based on the observations of MP-UV processes, it became clear that due to the significant

Ames responses observed, a risk assessment was required. Martijn et al. (2016) performed a

preliminary risk assessment for MP-UV processes, by converting the Ames test responses

into 4-nitroquinoline oxide (4-NQO) equivalent concentrations, in order to obtain a risk

indication via the Margin of Exposure (MOE) approach (EFSA, 2005). They found that the 4-

107

NQO equivalent concentrations exceeded the Estimated Daily Intake (EDI), associated with a

negligible risk, indicating a concern of the water quality, should it be distributed “as drinking

water without further post treatment”. Such an approach was not required for this work, since

it became evident from the findings of this study that the LP-UV/H2O2 treatment gave little, if

any, reason for concern from a health-related aspect.

Figure 6.1. Ames test response in water samples (20,000 concentration factor) as a function

of the nitrite formation by MP-UV treatment at WTP Andijk and in CB experiments with

IHSS Pony Lake NOM (current results superimposed on adapted Figure 5.3 of Martijn

(2015)).

The present work concludes that LP-UV/H2O2 treatment is not expected to produce

significant mutagenic activity as shown by the applied Ames II test (although other

mutagenic mechanisms cannot be excluded), even when high nitrate concentrations are

present and high UV/H2O2 doses are applied. Nevertheless, case-specific studies should be

conducted since the nitrite levels produced by the LP-UV/H2O2 treatment are not always

negligible (0.08-0.09 mg NO2-/L in this study) and the conjunctive effect of all factors

involved in health issues is usually complex.

6.5 DBP formation issues

The application of an advanced oxidation process is, in most cases, part of a multi-barrier

approach of waste or drinking water treatment, followed by a filtration or biological process

y = 31.025x + 2.6477

0

25

50

0 0.2 0.4 0.6 0.8 1

nu

mb

er o

f p

osi

tiv

e w

ells

in T

A9

8-S

9

(-)

nitrite (mg NO2 / L)

WTP Andijk MP UV AOP

MP UV CB exp. IHSS Pony Lake NOM plus NO3

LP-UV/H2O2 exp. Pony Lake NOM plus NO3

(current work)

WTP Andijk MP UV AOP

MP UV CB exp. IHSS Pony Lake NOM plus NO3

LP-UV/H2O2 exp. Pony Lake NOM plus NO3

(From this PhD research)

108

and ultimately disinfection. The formation of DBPs has been widely studied the last decades,

stemming from the reaction of the NOM in the water with the disinfectant used (chlorine,

ozone, chlorine dioxide, or chloramines). The most widely used disinfectant is chlorine,

leading to the production of chlorinated DBPs of high importance, like THMs and HAAs.

The nature of NOM seems to influence the BDP-FP (disinfection by-product formation

potential); more specifically the hydrophobic fraction is found to be proportionally related to

the THM-FP. Lin and Wang (2011) compared the THM-FP after HP-UV/H2O2 treatment of

commercially obtained humic acid extract to the ones from raw waters from three different

water treatment plants and reported that the bigger the hydrophobic fraction of the NOM, the

higher the THM-FP. The humic acid comprised of hydrophobic material by 91% and

exhibited the highest THM-FP, compared to the rest of the waters, whose hydrophobic

fraction was in the 40-60% range.

When an AOP is applied, mineralization of NOM is not a realistic expectation, because such

high doses would result in prohibitive costs. Usually, partial oxidation of NOM is observed,

which changes the nature and characteristics of NOM. The extent of this change defines the

reactivity of the NOM fractions with chlorine. Specifically for the UV/H2O2 process, Toor

and Mohseni (2007) showed that under practical conditions, the biodegradable organic

carbon content increased after treatment, whereas the DBP-FP remained practically the same

as before treatment. Increased UV/H2O2 doses (UV fluence of 1000 mJ/cm2 and initial H2O2

concentration of 23 mg/L and above) were found to reduce DBPs effectively. Sarathy et al.

(2011) stated that the reduction in aromatic structures by the UV/H2O2 did not change the

THM-FP; only when UV/H2O2 was coupled with BAC (biological activated carbon) was the

DBP-FP reduction significant. Haloacetamide and other N-DBPs formation after chlorination

was found to be reduced when UV/H2O2 was used as pre-treatment (Chu et al., 2014).

Therefore, it becomes evident that both the applied AOP conditions as well as the post-

treatment play ultimately a role in DBP formation. Inorganic precursors involvement (e.g.

bromide, iodide) should also be taken into account, since the brominated DBPs are both more

genotoxic and carcinogenic than their chlorinated equivalents, and iodinated DBP exhibit

high genotoxicity, but carcinogenicity information is lacking (Richardson et al., 2007).

109

6.6 Costs and regulations

Opting for an AOP system and selecting the most suitable one ultimately depends on the

water matrix to be treated and the treatment goals, along with the international or regional

differences in terms of energy costs, investments and regulations. Therefore, a carefully

calculated decision will have to be made in each case, taking into account the suitability and

feasibility of the available processes.

Two important parameters to consider regarding the application of AOPs is the fact that they

are non-selective processes, with the possibility of interference from other water constituents,

as well as the desirable effluent concentrations of regulated micropollutants that need to be

maintained. To compensate for the limitations and achieve the treatment goals in each case,

more energy, higher chemical dosages and contact times may be required, raising the

operational and capital costs. These costs are strongly dependent on the source water quality,

mainly the types and concentrations, as well as the EEO value of the contaminants of interest;

the higher the EEO value of the contaminant the more difficult and costly the removal

(Kommineni et al., 2000).

Technical factors contributing to operational and capital costs vary between the AOPs. In the

O3-based AOPs (O3/UV, O3/H2O2), the major operating cost is the cost of electricity for

ozone generation. Ozone removal by thermal destruction, catalytic reduction or a

combination of the two, results in additional costs. An air permit for ozone emissions is also

required. In UV oxidation technologies, UV lamps are often equipped with quartz sleeve

cleaning devices, especially when the influent water has high concentrations of fouling

agents. The UV/H2O2 and O3/H2O2 processes are more energetically efficient than the O3/UV

process for the generation of OH radicals in large quantities, due to the higher solubility of

H2O2 in water compared to that of ozone, therefore the operational costs are expected to be

lower than for O3/UV. The energy usage for cavitation systems (sonication, pulse plasma

cavitation and hydrodynamic cavitation) is comparable to AOPs using UV lamps; the main

energy cost comes from the use of pumps, but supplemental oxidants such as O3 and H2O2 in

order to achieve sufficient removal of the target micropollutants can raise the costs

significantly. For Fenton systems, removal of the residual iron in the effluent water, as well

as pH adjustments, increase the costs of the system (Kommineni et al., 2000).

110

Specifically for the UV/H2O2 process, the influent water quality plays a great role in the AOP

efficiency; by improving the water quality, higher UV transmission will lead to lower UV

doses required, therefore a smaller energy demand. At the same time, radical scavengers can

be removed, resulting in a more effective AOP step. In order to achieve this, a pre-treatment

step is essential; nevertheless, the costs of such pre-treatment should not be higher than the

cost reductions obtained in the advanced oxidation process. On the other hand, optimization

of the UV reactor may result in a substantial decrease in energy costs for the process. The

choice between the two commercially available lamps, the low pressure and medium pressure

one, mainly depends on the water matrix treated and the energy requirements. LP lamps

exhibit lower energy consumption, higher energy efficiency (25-30% as opposed to 15% for

MP lamps) and double the expected life span of MP lamps; however they have smaller

electrical power, and thus represent a larger footprint (KWR, 2011).

Companies also need to abide to the legislative body of each country. Drinking water

standards vary around the world, which is evident for a wide list of regulated organic

contaminants, including pesticides, PCBs, PAHs, as well as inorganic ones, such as nitrite,

heavy metals etc. Specifically for pesticides, Australia has set standards for the biggest

number of pesticides; however the EU, Germany, and the UK have the most stringent

pesticide standards. Instead of regulating individual pesticides and their metabolites, the

maximum limit is set to 0.1 μg/L for each pesticide and to 0.5 μg/L for the total pesticide

concentration. The difference between the maximum permitted levels can vary by much, for

example the regulated level for metaldehyde in EU is 0.1 μg/L, whereas in Australia it is 20

μg/L and in the USA it is not regulated. Nitrite, which was included in this study, also

exhibits a variation in the regulated levels around the world, with the WHO guideline at 3

mg/L as NO2- due to methaemoglobinaemia risk in bottle-fed babies (World Health

Organization, 2011), Australia and Canada at 3 mg NO2-/L and 3.2 mg NO2

-/L, respectively,

the EU regulated concentrations of 0.5 mg NO2-/L and 0.10 mg NO2

-/L, at consumer’s tap

and ex water treatment works, respectively (European Communities Drinking Water

Regulations, 2007) and the EPA Maximum Contaminant Level (MCL) at 1 mg/L as nitrogen,

equal to 3.3 mg/L as NO2- (USEPA, 2009). In the case of this research, the EU regulations

were taken into account, but it becomes evident that these differences complex the

comparison between studies worldwide.

111

6.7 Recommendations for future work

This research has examined a number of issues concerning the LP-UV/H2O2 process, though

some additional aspects that are recommended for future research are highlighted below.

Regarding the reaction product formation from the pesticides studied, the objective of

this research was the detection of the main products formed under the studied

conditions in order to come up with a tentative degradation pathway for all three

pesticides. Nevertheless, quantification of these products along with a calculation of

mass balance are essential in order to obtain a more complete investigation of the

degradation and reaction mechanisms. A different detection system could also provide

more accurate results; in this case a Triple Quadrupole Mass Spectrometer (QqQ) was

used, whereas a TOF (Time-Of-Flight) analyser with a higher resolution could be

helpful for structure elucidation.

Since LP-UV/H2O2 was found to be an effective process for degrading the studied

recalcitrant pesticides, future studies could consider broadening the research to other

pesticides and micro-pollutants of concern, with regards to their degradation potential

and product formation. More specifically, compounds with similar structure to the

ones studied here, are expected to follow a similar degradation profile. Compounds

containing a pyridine ring resembling clopyralid’s structure (e.g. herbicides like

picloram, triclopyr etc.) would probably exhibit similar behaviour under LP-UV/H2O2

degradation, while phenoxypropionic compounds (e.g. the herbicide dichlorprop) are

expected to be fairly easily degraded and in ways similar to mecoprop degradation.

Not many compounds share similarities in structure with metaldehyde, a heterocyclic,

non-aromatic polyether, except some oxolane and oxane structures; various cyclic

ethers have been used extensively in industry (e.g. 1,4-dioxane, 1,3-dioxolane) and

studying their removal via oxidation processes is of high environmental importance.

Other compounds worthy of attention are widely applied antibiotics with more

complicated structures (e.g. containing the oxacyclotetradecane-2,10-dione structure),

such as erythromycin and clarithromycin. Types of chemicals with high resistance to

conventional treatment due to their unique structures, such as polychlorinated

dibenzo-p-dioxins (PCDDs), perfluorinated chemicals (PFCs), polychlorinated

dibenzofurans (PCDF) and polychlorinated biphenyls (PCB) are also worth

investigating.

112

The assessment of the mutagenicity of the reaction products formed from the

degradation of the pesticides or other micro-pollutants is recommended, considering

that metabolites may be more toxic than the parent compounds. QSAR modelling

could be a useful initial screening tool for prioritising which products are especially

worthy of experimental mutagenicity testing.

The kinetic model for the photolysis/oxidation of nitrate is well established, and a

comparison between the experimental data and a quantitative model of the system, is

suggested. The modelling of chemical processes entails the prediction of the evolution

of the concentrations of the reaction species as a function of time, and kinetic

calculations via computational tools that are nowadays available (e.g. Matlab

programs) serve to this end. This modelling work requires the knowledge of the

mechanisms involved, as well as the reaction rate constants. In this study, not all

kinetic rate constants were known, an issue that prevented this research from

providing quantitative predictions. The determination of rate constants is possible

either via chemical/physical methods (e.g. directly, by measuring concentrations of

reactants and/or products, or indirectly, by measuring physical properties, like

absorption of a solution) or via computational chemistry, utilizing computer codes to

predict parameters such as energy bonds, electronic structures and transition states. In

cases where radicals are involved, such as in the nitrate photolysis system, these can

be produced by techniques like pulse radiolysis or flash/laser photolysis, creating

intermediates which react with appropriate solutes producing secondary radicals that

exhibit a measurable property. Once all reaction constants are known, modelling of

the photochemical system is possible.

This study, along with previous ones, indicate the significant role that the background

water matrix can play in the health-related issues of UV/H2O2-treated water,

compared to the contribution of individual organic micro-pollutants. Although the LP

application is proven to be of minimum concern, issues such as the effect of NOM

nitrite formation, and nitrogen incorporation in the organic matrix, yielding harmful

nitrated compounds – both researched only to a limited extent - could arise and need

to be addressed. The background water matrix is a site- and conditions-specific

parameter, with complex composition and properties, indicating the importance of

shifting the attention towards other constituents that may prove to be problematic

under AOP treatment. An important example of such challenges is the post-

113

disinfection formation of compounds of concern, mainly brominated organic

compounds from bromide-rich water sources, as well as the formation of nitrogen-

containing DBPs, due to their known toxicity.

Due to time and resource constraints, only two samplings took place from the

drinking water treatment plant for mutagenicity testing. Although the sampling

periods were carefully chosen and the Ames results confirmed the expectation of low

concern for the LP-UV/H2O2 treatment regarding mutagenicity issues, sampling

throughout the year is recommended, since it would yield more information about the

variability in water quality pre- and post-AOP.

6.8 Conclusions

The first part of this research dealt with the degradation of the pesticides mecoprop,

metaldehyde and clopyralid, due to their presence in surface waters and their recalcitrance

towards other treatment options, especially for metaldehyde and clopyralid. The advanced

oxidation process of LP-UV/H2O2 was applied and a wide range of UV/H2O2 doses was

utilised, in order to study the degradation and kinetics for each pesticide, taking into account

their degradability under OH-radical oxidation and UV photolysis. The observed degradation

and kinetic order was mecoprop > metaldehyde > clopyralid; both oxidation and photolysis

played a role for the degradation of mecoprop, whereas for metaldehyde and clopyralid the

OH-radical attack was key to their break-down. The kinetic rate constants and quantum yields

related to UV photolysis as well as the kinetic rate constants of the reaction of the pesticides

with the OH-radicals were calculated.

The impact of the presence of water matrix (i.e. a mixture of natural organic matter, inorganic

ions and pharmaceuticals) on the degradation kinetics was investigated in combination with

other parameters (initial pesticide concentration and hydrogen peroxide dose). The kinetics

were hindered for mecoprop and clopyralid when the background matrix was present, but not

for metaldehyde. However, the increase in hydrogen peroxide dose enhanced the kinetics in

the presence of the background matrix, demonstrating the important role of the hydroxyl

radicals (and the negative impact of their scavenging) in the degradation of these pesticides.

The main reaction products formed from the degradation of the pesticides by LP-UV/H2O2

were identified. The structure and active sites available for OH-attack determined the

mechanism involved in each case. Mecoprop underwent photo-hydrolysis and OH-attack on

114

the side chain, with further degradation to smaller molecules (e.g. pyruvic acid). De-

chlorination and hydroxylation where observed for clopyralid, whereas metaldehyde

produced a hydroxylated NH4+ complex (a product resulting from the detection method used)

that led to subsequent formation of lower molecular weight compounds (acetaldehyde and

acetic acid).

The second part of the research focused on the nitrite-related issues under LP-UV/H2O2, the

formation of nitrite, nitrogenated by-products and the Ames mutagenicity response of both

synthetic and full scale samples. The little studied effect of the presence of NOM on nitrite

formation by LP-UV/H2O2 treatment of NOM-NO3- containing samples, in conjunction with

parameters like H2O2 dose and pH, was investigated. The presence of NOM led to an increase

in nitrite formation compared to its absence, suggesting the effect of NOM on the complex

reaction steps of nitrate photolysis, through the production of oxidative species and reducing

agents, along with the scavenging of OH-radicals. Moreover, it was concluded that variations

in the NOM composition (humic content, UV absorbance, aromaticity and nitrogen content),

under the experimental conditions used in this study, did not play a role in the concentration

of nitrite formed (range of 0.08-0.15 mg NO2-/L). Regarding the nitrogenated compound

formation, only specific nitrophenolic compounds were considered. Trace amounts of

nitrophenol were detected, confirming the incorporation of intermediate radicals (formed by

LP-UV photolysis of nitrate) into the organic matrix that has already been shown for MP-UV

applications.

Finally, synthetic water samples containing NOM (two different NOMs tested) and nitrate

were treated with LP-UV/H2O2 and tested for nitrite concentrations formed as well as

mutagenicity with the Ames II test. The results revealed significant nitrite formation and a

small variation of responses (with the one of the two NOMs) between the conditions applied,

indicating a contribution of the presence of NO3- to the mutagenicity response. This result

suggested the formation of nitrated compounds (e.g. nitrophenols, as discussed above), which

are already known for their toxicity. However, both synthetic and real water samples showed

no mutagenicity towards the Salmonella typhimurium strain TA98 used, even under extreme

UV/H2O2 conditions.

115

References

Aaron, J.J., Oturan, M.A., 2001. New photochemical and electrochemical methods for the

degradation of pesticides in aqueous media. Environmental applications. Turkish J.

Chem. 25, 509–520.

Alvarez, B., Denicola, A., Radi, R., 1995. Reaction between peroxynitrite and hydrogen

peroxide: Formation of oxygen and slowing of peroxynitrite decomposition. Chem. Res.

Toxicol. 8, 859–864. doi:10.1021/tx00048a006

Ames, B.N., Gurney, E.G., Miller, J.A., Bartsch, H., 1972. Carcinogens as frameshift

mutagens: metabolites and derivatives of 2-acetylaminofluorene and other aromatic

amine carcinogens. Proc. Natl. Acad. Sci. U. S. A. 69, 3128–3132.

doi:10.1073/pnas.69.11.3128

Andreozzi, R., Caprio, V., Insola, A., Marotta, R., 1999. Advanced oxidation processes

(AOP) for water purification and recovery. Catal. Today 53, 51–59. doi:10.1016/S0920-

5861(99)00102-9

Armbrust, K.L., 2000. Pesticide Hydroxyl Radical Rate Constants: Measurements and

Estimates of Their Importance in Aquatic Environments. Environ. Toxicol. Chem. 19,

2175–2180. doi:10.1002/etc.5620190905

Autin, O., Hart, J., Jarvis, P., MacAdam, J., Parsons, S.A., Jefferson, B., 2013. The impact of

background organic matter and alkalinity on the degradation of the pesticide

metaldehyde by two advanced oxidation processes: UV/H2O2 and UV/TiO2. Water Res.

47, 2041–2049. doi:10.1016/j.watres.2013.01.022

Autin, O., Hart, J., Jarvis, P., MacAdam, J., Parsons, S.A., Jefferson, B., 2012. Comparison

of UV/H2O2 and UV/TiO2 for the degradation of metaldehyde: Kinetics and the impact

of background organics. Water Res. 46, 5655–5662. doi:10.1016/j.watres.2012.07.057

Auzmendi-Murua, I., Bozzelli, J.W., 2014. Thermochemical properties and bond dissociation

enthalpies of 3- to 5-member ring cyclic ether hydroperoxides, alcohols, and peroxy

radicals: Cyclic ether radical + 3O2 reaction thermochemistry. J. Phys. Chem. A 118,

3147–3167. doi:10.1021/jp412590g

Baeza, C., Knappe, D.R.U., 2011. Transformation kinetics of biochemically active

compounds in low-pressure UV Photolysis and UV/H2O2 advanced oxidation processes.

Water Res. 45, 4531–4543. doi:10.1016/j.watres.2011.05.039

116

Baxendale, J.H., Wilson, J.A., 1956. The Photolysis of Hydrogen Peroxide At High Light

Intensities. Trans. Faraday Soc. 53, 344–356.

Bayliss, N.S., Bucat, R.B., 1975. The photolysis of aqueous nitrate solutions. Aust.J.Chem.

28, 1865–1878. doi:10.1071/CH9751865

Beltrán, F.J., González, M., Rivas, J., Marín, M., 1994. Oxidation of mecoprop in water with

ozone and ozone combined with hydrogen peroxide. Ind. Eng. Chem. Res. 33, 125–136.

doi:10.1021/ie00025a017

Beltrán, F.J., Ovejero, G., Rivas, J., 1996. Oxidation of Polynuclear Aromatic Hydrocarbons

in Water. 3. UV Radiation Combined with Hydrogen Peroxide. Ind. Eng. Chem. Res.

35, 883–890. doi:10.1021/ie950363l

Bems, B., Jentoft, F.C., Schlögl, R., 1999. Photoinduced decomposition of nitrate in drinking

water in the presence of titania and humic acids. Appl. Catal. B Environ. 20, 155–163.

doi:10.1016/S0926-3373(98)00105-2

Benitez, F.J., Acero, J.L., Real, F.J., Roman, S., 2004. Oxidation of MCPA and 2,4- d by UV

Radiation, Ozone, and the Combinations UV/H2O2 and O3/H2O2. J. Environ. Sci. Heal.

Part B 39, 393–409. doi:10.1081/PFC-120035925

Benitez, F.J., Real, F.J., Acero, J.L., Garcia, C., 2006. Photochemical oxidation processes for

the elimination of phenyl-urea herbicides in waters. J. Hazard. Mater. 138, 278–287.

doi:10.1016/j.jhazmat.2006.05.077

Bolton, J.R., 2004. Fundamentals of Ultraviolet Treatment – Terms and Definitions do

Matter ! 27–31.

Bolton, J.R., 2001. Ultraviolet Applications Handbook, second edition, Bolton Photosciences

Inc.

Bolton, J.R., Linden, K.G., 2003. Standardization of Methods for Fluence (UV Dose)

Determination in Bench-Scale UV Experiments. J. Environ. Eng. 129, 209–215.

doi:10.1061/(ASCE)0733-9372(2003)129:3(209)

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. Res. Chem. Intermed. 28, 857–870. doi:10.1163/15685670260469474

Bond, T., Huang, J., Templeton, M.R., Graham, N., 2011. Occurrence and control of

nitrogenous disinfection by-products in drinking water - A review. Water Res. 45, 4341–

117

4354. doi:10.1016/j.watres.2011.05.034

Bond, T., Templeton, M.R., Graham, N., 2012. Precursors of nitrogenous disinfection by-

products in drinking water--A critical review and analysis. J. Hazard. Mater. 235-236,

1–16. doi:10.1016/j.jhazmat.2012.07.017

Boule, P., Meunier, L., Bonnemoy, F., Boulkamh, A., Zertal, A., Lavedrine, B., 2002. Direct

phototransformation of aromatic pesticides in aqueous solution. Int. J. Photoenergy 4,

69–78. doi:10.1155/S1110662X02000119

Bruce, E.D., Autenrieth, R.L., Burghardt, R.C., Donnelly, K.C., McDonald, T.J., 2008. Using

quantitative structure-activity relationships (QSAR) to predict toxic endpoints for

polycyclic aromatic hydrocarbons (PAH). J. Toxicol. Environ. Health. A 71, 1073–84.

doi:10.1080/15287390802114337

Buchanan, I., Liang, H.C., Khan, W., Liu, Z., Singh, R., Ikehata, K., Chelme-Ayala, P., 2009.

Pesticides and Herbicides, Water Environment Research.

doi:10.2175/106143009X12445568400331

Burrows, H.D., Canle L, M., Santaballa, J.A., Steenken, S., 2002. Reaction pathways and

mechanisms of photodegradation of pesticides. J. Photochem. Photobiol. B Biol. 67, 71–

108. doi:10.1016/S1011-1344(02)00277-4

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. J. Phys. Chem. Ref. Data 17, 513–886.

doi:10.1063/1.555805

Camper, A.K., Lechevallier, M.W., Broadaway, S.C., Mcfeters, G.A., 1985. Growth and

persistence of pathogens on granular activated carbon filters. Appl. Environ. Microbiol.

50, 1378–1382.

Chen, J., Gu, B., LeBoeuf, E.J., Pan, H., Dai, S., 2002. Spectroscopic characterization of the

structural and functional properties of natural organic matter fractions. Chemosphere 48,

59–68. doi:10.1016/S0045-6535(02)00041-3

Chen, R., Freeman, G.R., 1995. Solvent Effects on Reactivity of Solvated Electrons with

Nitrate Ion and on Electrolyte Conductivity in C1 to C10 n-Alcohols. J. Phys. Chem 99,

4970–4975.

Choi, H.-J., Kim, D., Lee, T.-J., 2013. Photochemical degradation of atrazine in UV and

118

UV/H2O2 process: pathways and toxic effects of products. J. Environ. Sci. Heal. 48,

927–934. doi:10.1080/03601234.2013.816587

Christman, R.F., Kronberg, K., Singh, R., Ball, L.M., Johnson, J.D., 1991. Identification of

mutagenic by-products from aquatic humic chlorination. Department of Environmental

Sciences and Engineering University of North Carolina.

Chu, W., Gao, N., Yin, D., Krasner, S.W., Mitch, W.A., 2014. Impact of UV/H2O2 Pre-

Oxidation on the Formation of Haloacetamides and Other Nitrogenous Disinfection

Byproducts during Chlorination. Environ. Sci. Technol. 48, 12190–12198.

Clayden, J., Greeves, N., Warren, S.G., 2012. Organic Chemistry 1006 – 1010.

Coddington, J.W., Hurst, J.K., Lymar, S. V, 1999. Hydroxyl radical formation during

peroxynitous acid decomposition. J. Am. Chem. Soc. 121, 2438–2443.

Comninellis, C., Kapalka, A., Malato, S., Parsons, S.A., Poulios, I., Mantzavinos, D., 2008.

Perspective Advanced oxidation processes for water treatment: advances and trends for

R&D. J. Chem. Technol. Biotechnol. 83, 769–776. doi:10.1002/jctb

Cook, A.R., Dimitrijevic, N., Dreyfus, B.W., Meisel, D., Curtiss, L.A., Camaioni, D.M.,

2001. Reducing Radicals in Nitrate Solutions . The NO32- System Revisited 1, 3658–

3666.

Cooper, B., 2011. Working with farmers to reduce pesticide movement to water. Asp. Appl.

Biol. 106, 289–294.

Cooper, W.J., Zika, R.G., Petasne, R.G., Fischer, A.M., 1989. Sunlight-induced

photochemistry of humic substances in natural-waters - Major reactive species. Acs

Symp. Ser. 219, 333–362.

Corredor, M.C., Mellado, J.M.R., Montoya, M.R., 2006. EC(EE) process in the reduction of

the herbicide clopyralid on mercury electrodes. Electrochim. Acta 51, 4302–4308.

doi:10.1016/j.electacta.2005.12.010

Council Directive 98/83/EC, 1998. 3 November 1998. On the quality of water intended for

human consumption. Off. J. Eur. Communities L 330/32 – 330/54.

Crittenden, J.C., Trussell, R.R., Hand, D.W., Howe, K.J., Tchobanoglous, G., 2012. MWH’s

Water Treatment: Principles and Design, Third Edition. John Wiley & Sons, Inc.

Daniels, M., Meyers, R. V., Belardo, E. V., 1968. Photochemistry of the Aqueous Nitrate

System. I. Excitation in the 300-mμ Band. Symp. Ionorganic Photochem. 72, 389–399.

119

Derks, J.G., 2010. Degradation of organic micropollutants by advanced oxidation through

UV/H2O2. Master Thesis. Delft Univ. Technol. Delft, Netherlands.

Dolan, T., Howsam, P., Parsons, D.J., Whelan, M.J., 2013. Is the EU drinking water directive

standard for pesticides in drinking water consistent with the precautionary principle?

Environ. Sci. Technol. 47, 4999–5006. doi:10.1021/es304955g

Donald, D.B., Cessna, A.J., Sverko, E., Glozier, N.E., 2007. Pesticides in surface drinking-

water supplies of the northern Great Plains. Environ. Health Perspect. 115, 1183–1191.

doi:10.1289/ehp.9435

Drinking Water Inspectorate. Annual report., 2015.

Dzengel, J., Theurich, J., Bahnemann, D.W., 1999. Formation of nitroaromatic compounds in

advanced oxidation processes: Photolysis versus photocatalysis. Environ. Sci. Technol.

33, 294–300. doi:10.1021/es980358j

EFSA, 2005. Conclusion regarding the peer review of the pesticide risk assessment of the

active substance clopyralid. Scientific Report.

Elovitz, M.S., Gunten, U. von, 1999. Hydroxyl Radical/Ozone Ratios During Ozonation

Processes. I. The Rct Concept. Ozone Sci. Eng. 21, 239–260.

doi:10.1080/01919519908547239

EPA, 2012. EPA Drinking Water Guidance on Disinfection By-Products Advice Note No. 4.

Version 2. Disinfection By-Products in Drinking Water 1–27.

European Communities (Drinking Water) (No. 2) Regulations, 2007.

Flückiger-Isler, S., Baumeister, M., Braun, K., Gervais, V., Hasler-Nguyen, N., Reimann, R.,

Van Gompel, J., Wunderlich, H.G., Engelhardt, G., 2004. Assessment of the

performance of the Ames IITM assay: A collaborative study with 19 coded compounds.

Mutat. Res. - Genet. Toxicol. Environ. Mutagen. 558, 181–197.

doi:10.1016/j.mrgentox.2003.12.001

Frimmel, F.H., 1994. Photochemical aspects related to humic substances. Environ. Int. 20,

373–385. doi:10.1016/0160-4120(94)90123-6

Gee, P., Sommers, C.H., Melick, A.S., Gidrol, X.M., Todd, M.D., Burris, R.B., Nelson, M.E.,

Klemm, R.C., Zeiger, E., 1998. Comparison of responses of base-specific salmonella

tester strains with the traditional strains for identifying mutagens: The results of a

validation study. Mutat. Res. - Genet. Toxicol. Environ. Mutagen. 412, 115–130.

120

doi:10.1016/S1383-5718(97)00172-1

Gill, H.K., Garg, H., 2014. Pesticides: Environmental Impacts and Management Strategies.

Pestic. - Toxic Asp. 187–230. doi:10.5772/57399

Goldstein, S., Lind, J., Merényi, G., 2005. Chemistry of peroxynitrites as compared to

peroxynitrates. Chem. Rev. 105, 2457–2470. doi:10.1021/cr0307087

Goldstein, S., Rabani, J., 2007. Mech Nitrite Formation by Nitrate Photolysis in Aq Solns:

Role of Peroxynitrite, Nitrogen Dioxide NO2, Hydroxyl Radical. J. Am. Chem. Soc.

129, 10597–10601. doi:10.1021/ja073609+

Gonsior, M., Schmitt-Kopplin, P., Stavklint, H., Richardson, S.D., Hertkorn, N., Bastviken,

D., 2014. Dissolved Organic Matter Changes along Treatments of a Drinking Water

Plant in Sweden and the Formation of Previously Unknown DBPs. Environ. Sci.

Technol. 48, 12714–22. doi:10.1021/es504349p

Greenwood, N.N., Earnshaw, A., 1997. Chemistry of the Elements, 2nd Edition.

Guzzella, L., Feretti, D., Monarca, S., 2002. Advanced oxidation and adsorption technologies

for organic micropollutant removal from lake water used as drinking-water supply.

Water Res. 36, 4307–4318. doi:10.1016/S0043-1354(02)00145-8

Haag, W.R., Yao, C.C.D., 1992. Rate Constants for Reaction of Hydroxyl Radicals With

Several Drinking-Water Contaminants. Environ. Sci. Technol. 26, 1005–1013.

doi:10.1021/es00029a021

Haider, T., Sommer, R., Kansmüller, S., Eckl, P., Pribil, W., Cabaj, A., Kundi, M., 2002.

Genotoxic response of Austrian groundwater samples treated under standardized UV

(254nm)-disinfection conditions in a combination of three different bioassays. Water

Res. 36, 25–32. doi:10.1016/S0043-1354(01)00188-9

Hakura, A., Suzuki, S., Satoh, T., 1999. Advantage of the use of human liver S9 in the Ames

test. Mutat. Res. - Genet. Toxicol. Environ. Mutagen. 438, 29–36. doi:10.1016/S1383-

5718(98)00159-4

Hepplewhite, C., Newcombe, G., Knappe, D.R.U., 2004. NOM and MIB , who wins in the

competition for activated carbon adsorption sites ? Water Sci. Technol. 49, 257–265.

Heringa, M.B., Harmsen, D.J.H., Beerendonk, E.F., Reus, A.A., Krul, C.A.M., Metz, D.H.,

Ijpelaar, G.F., 2011. Formation and removal of genotoxic activity during UV/H2O2-

GAC treatment of drinking water. Water Res. 45, 366–374.

121

doi:10.1016/j.watres.2010.08.008

Hofman-Caris, R.C.H.M., Harmsen, D.J.H., Puijker, L., Baken, K.A., Wols, B.A.,

Beerendonk, E.F., Keltjens, L.L.M., 2015. Influence of process conditions and water

quality on the formation of mutagenic byproducts in UV/H2O2 processes. Water Res. 74,

191–202. doi:10.1016/j.watres.2015.01.035

Houtman, C.J., 2010. Emerging contaminants in surface waters and their relevance for the

production of drinking water in Europe. J. Integr. Environ. Sci. 7, 271–295.

doi:10.1080/1943815x.2010.511648

Hughes, R.M., 2013. Formation of genotoxic compounds by medium pressure ultraviolet

treatment of pre-treated surface water and groundwater. University of New Hampshire.

Hurtado, P., Gámez, F., Hamad, S., Martínez-Haya, B., Steill, J.D., Oomens, J., 2012.

Multipodal coordination of a tetracarboxylic crown ether with NH4(+): a vibrational

spectroscopy and computational study. J. Chem. Phys. 136, 114301.

doi:10.1063/1.3693518

IJpelaar, G.F., van der Veer, B., Medema, G.J., Kruithof, J.C., 2005. By-product formation

during ultraviolet disinfection of a pretreated surface water. J. Environ. Eng. Sci. 4,

S51–S56. doi:10.1017/CBO9781107415324.004

Ikehata, K., El-Din, M.G., 2006. Aqueous pesticide degradation by hydrogen

peroxide/ultraviolet irradiation and Fenton-type advanced oxidation processes: a review.

J. Environ. Eng. Sci. 5, 81–135. doi:10.1139/s05-046

Ikehata, K., Naghashkar, N.J., El-Din, M.G., 2006. Degradation of Aqueous Pharmaceuticals

by Ozonation and Advanced Oxidation Processes: A Review. Ozone Sci. Eng. 27, 83–

114. doi:10.1080/01919510600985937

International Agency for Research on Cancer, 2012. IARC Monographs on the Evaluation of

Carcinogenic Risks to Humans.Benzo[a]pyrene. Volume 100F. IARC Monogr.

International Agency for Research on Cancer, 1999. IARC Monographs on the Evaluation of

Carcinogenic Risks to Humans.Some Chemicals that Cause Tumours of the Kidney or

Urinary Bladder in Rodents and Some Other Substances. Volume 73, World Health

Organization. Lyon, France.

Karanfil, B.Y.T., Schlautman, M. a, Erdogan, I., 2002. Survey of DOC and UV measurement

practices with implications for SUVA determination. J. Am. Water Works Assoc. 94,

122

68–80.

Khandoudi, N., Porte, P., Chtourou, S., Nesslany, F., Marzin, D., Le Curieux, F., 2009. The

presence of arginine may be a source of false positive results in the Ames test. Mutat.

Res. - Genet. Toxicol. Environ. Mutagen. 679, 65–71.

doi:10.1016/j.mrgentox.2009.03.010

Kieber, R.J., Li, A., Seaton, P.J., 1999. Production of nitrite from the photodegradation of

dissolved organic matter in natural waters. Environ. Sci. Technol. 33, 993–998.

doi:10.1021/es980188a

Klassen, N.V., Marchington, D., McGowan, H.C.E., 1994. H2O2 Determination by the I3-

Method and by KMnO4 Titration. Anal. Chem. 66, 2921–2925.

doi:10.1017/CBO9781107415324.004

Kolkman, A., Martijn, B.J., Vughs, D., Baken, K.A., Van Wezel, A.P., 2015. Tracing

nitrogenous disinfection byproducts after medium pressure UV water treatment by stable

isotope labeling and high resolution mass spectrometry. Environ. Sci. Technol. 49,

4458–4465. doi:10.1021/es506063h

Kommineni, S., Zoeckler, J., Stocking, A., Liang, S., Flores, A., Kavanaugh, M., 2000.

Treatment Technologies for Removal of Methyl Tertiary Butyl Ether (MTBE) from

Drinking Water: Chapter 3, Advanced Oxidation Processes., National Water Research

Institute.

Korshin, G. V, Li, C., Benjamin, M.M., 1997. Monitoring the properties of natural organic

matter through UV spectroscopy: A consistent theory 31, 1787–1795.

Kovacic, P., Somanathan, R., 2014. Nitroaromatic compounds: Environmental toxicity,

carcinogenicity, mutagenicity, therapy and mechanism. J. Appl. Toxicol. 34, 810–824.

doi:10.1002/jat.2980

Kowalska, E., Janczarek, M., Hupka, J., Grynkiewicz, M., 2004. H2O2/UV enhanced

degradation of pesticides in wastewater E. Water Sci. Technol. 49, 261–266.

Kruithof, J.C., Kamp, P.C., Martijn, B.J., 2007. UV/H2O2 treatment: A practical solution for

organic contaminant control and primary disinfection. Ozone-Science Eng. 29, 273–280.

doi:10.1080/01919510701459311

KWR, 2011. New concepts of UV/H2O2 oxidation. KWR report for WaterRF.

KWR, 2004. Effect of hydrogen peroxide and water quality on nitrite formation during nitrate

123

photolysis. KWR report for KIWA N.V.

Lapworth, D.J., Gooddy, D.C., 2006. Source and persistence of pesticides in a semi-confined

chalk aquifer of southeast England. Environ. Pollut. 144, 1031–1044.

doi:10.1016/j.envpol.2005.12.055

Ledakowicz, S., Miller, J.S., Olejnik, D., 1999. Oxidation of PAHs in water solutions by

ultraviolet radiation combined with hydrogen peroxide. Int. J. Photoenergy 1, 1–6.

doi:10.1155/S1110662X99000100

Leenheer, J.A., 1994. Environmental chemistry of lakes and reservoirs, in: Chapter 7 in

Environmental Chemistry of Lakes and Reservoirs. Advances in Chemistry. pp. 195–

221.

Legrini, O., Oliveros, E., Braun, A.M., 1993. Photochemical processes for water treatment.

Chem. Rev. 93, 671–698. doi:10.1021/cr00018a003

Lekkerkerker-Teunissen, K., Knol, A.H., Derks, J.G., Heringa, M.B., Houtman, C.J.,

Hofman-Caris, C.H.M., Beerendonk, E.F., Reus, A., Verberk, J.Q.J.C., van Dijk, J.C.,

2013. Pilot Plant Results with Three Different Types of UV Lamps for Advanced

Oxidation. Ozone Sci. Eng. 35, 38–48. doi:10.1080/01919512.2013.721317

Lester, Y., Avisar, D., Mamane, H., 2010. Photodegradation of the antibiotic

sulphamethoxazole in water with UV/H2O2 advanced oxidation process. Environ.

Technol. 31, 175–183. doi:10.1080/09593330903414238

Lin, H., Wang, G., 2011. Effects of UV/H2O2 on NOM fractionation and corresponding

DBPs formation. Desalination 270, 221–226. doi:10.1016/j.desal.2010.11.049

Linden, K.G., Rosenfeldt, E.J., Chen, P.J., Kullman, S.W., 2004. UV and UV/H2O2

degradation and subsequent toxicity of endocrine disrupting chemicals in water, in:

European Conference on UV Radiation. Karlsruhe, Germany.

Lu, N., Gao, N.Y., Deng, Y., Li, Q.S., 2009. Nitrite formation during low pressure ultraviolet

lamp irradiation of nitrate. Water Sci. Technol. 60, 1393–1400.

doi:10.2166/wst.2009.475

Lv, X., Lu, Y., Yang, X., Dong, X., Ma, K., Xiao, S., Wang, Y., Tang, F., 2015. Mutagenicity

of drinking water sampled from the Yangtze River and Hanshui River (Wuhan section)

and correlations with water quality parameters. Sci. Rep. 5, 1–8. doi:10.1038/srep09572

Lyman Ott, R., Longnecker, M.T., 2010. roduction to Statistical Methods and Data Analysis,

124

6th Edition. Belmont, CA: Brooks/Cole, Cengage Learning.

Machado, F., Boule, P., 1995. Photonitration and photonitrosation of phenolic derivatives

induced in aqueous solution by excitation of nitrite and nitrate ions. J. Photochem.

Photobiol. A Chem. 86, 73–80. doi:10.1016/1010-6030(94)03946-R

Mack, J., Bolton, J.R., 1999. Photochemistry of nitrite and nitrate in aqueous solution: a

review. J. Photochem. Photobiol. A Chem. 128, 1–13. doi:10.1016/S1010-

6030(99)00155-0

Manasfi, T., De Méo, M., Coulomb, B., Di Giorgio, C., Boudenne, J.-L., 2015. A

Comparison Between Freshwater and Seawater Swimming Pools : From Disinfection

By-products Profile to Genotoxicity.

Mariani, M.L., Romero, R.L., Zalazar, C.S., 2015. Modeling of degradation kinetic and

toxicity evaluation of herbicides mixtures in water using the UV/H2O2 process.

Photochem. Photobiol. Sci. 14, 608–617. doi:10.1039/C4PP00269E

Mark, G., Korth, H.G., Schuchmann, H.P., Von Sonntag, C., 1996. The photochemistry of

aqueous nitrate ion revisited. J. Photochem. Photobiol. A Chem. 101, 89–103.

doi:10.1016/S1010-6030(96)04391-2

Martijn, A.J., 2015. Impact of the water matrix on the effect and the side effect of MP

UV/H2O2 treatment for the removal of organic micropollutants in drinking water

production. Wageningen University.

Martijn, A.J., Boersma, M.G., Vervoort, J.M., Rietjens, I.M.C.M., Kruithof, J.C., 2014.

Formation of genotoxic compounds by medium pressure ultraviolet treatment of nitrate-

rich water. Desalin. Water Treat. 52, 6275–6281. doi:10.1080/19443994.2014.925654

Martijn, A.J., Kruithof, J.C., 2012. Treatment: The Silver Bullet for By-product and

Genotoxicity Formation in Water Production. Ozone Sci. Eng. 34, 92–100.

doi:10.1080/01919512.2012.649596

Martijn, B.J., Kruithof, J.C., Hughes, R.M., Mastan, R.A., Van Rompay, A.R., Malley, J.P.,

2015. Induced genotoxicity in nitrate-rich water treated with medium-pressure

ultraviolet processes. J. Am. Water Works Assoc. 107, E301–E312.

doi:10.5942/jawwa.2015.107.0079

Martijn, B.J., Van Rompay, A.R., Penders, E.J.M., Alharbi, Y., Baggelaar, P.K., Kruithof,

J.C., Rietjens, I.M.C.M., 2016. Development of a 4-NQO toxic equivalency factor

125

(TEF) approach to enable a preliminary risk assessment of unknown genotoxic

compounds detected by the Ames II test in UV/H2O2 water treatment samples.

Chemosphere 144, 338–345. doi:10.1016/j.chemosphere.2015.08.070

Meunier, L., Boule, P., 2000. Direct and induced phototransformation of mecoprop [2-(4-

chloro-2-methylphenoxy)- propionic acid] in in aqueous solution 1085, 1077–1085.

Minakata, D., Li, K., Westerhoff, P., Crittenden, J., 2009. Development of a Group

Contribution Method To Predict Aqueous Phase Hydroxyl Radical (HO center dot)

Reaction Rate Constants. Environ. Sci. Technol. 43, 6220–6227. doi:10.1021/es900956c

Moriarty, J., Sidebottom, H., Wenger, J., Mellouki, A., Le Bras, G., 2003. Kinetic Studies on

the Reactions of Hydroxyl Radicals with Cyclic Ethers and Aliphatic Diethers. J. Phys.

Chem. 107, 1499–1505. doi:10.1021/jp061536v

Nassar, A.-E.F., Talaat, R.E., 2004. Strategies for dealing with metabolite elucidation in drug

discovery and development. Curr. Opin. Drug Discov. Devel. 7, 317–327.

National Academy of Sciences., 1980. Drinking Water and Health, Volume 2. Safe Drinking

Water Committee. Board on Toxicology and Environmental Health Hazards. Assembly

of Life Sciences. National Research Council.

Niessen R., Lenoir D., Boule P., 1988. Phototransformation of phenol induced by excitation

of nitrate ions. Chemosphere 17, 1977–1984.

OECD, 1997. The OECD Guideline 471 for the Testing of Chemicals - Bacterial Reverse

Mutation Test. Organisation for Economic Co-operation and Development publishing.

doi:10.1787/9789264071247-en

Orellana-Garcia, F., Alvarez, M.A., Lopez-Ramon, V., Rivera-Utrilla, J., Sanchez-Polo, M.,

Mota, A.J., 2014. Photodegradation of herbicides with different chemical natures in

aqueous solution by ultraviolet radiation. Effects of operational variables and solution

chemistry. Chem. Eng. J. 255, 307–315. doi:10.1016/j.cej.2014.06.047

OSPAR Commission, 2002. Survey on Genotoxicity Test Methods for the Evaluation of

Waste Water within Whole Effluent Assessment, Assessment.

Parkinson, A., Barry, M.J., Roddick, F.A., Hobday, M.D., 2001. Preliminary toxicity

assessment of water after treatment with uv-irradiation and UVC/H2O2. Water Res. 35,

3656–3664. doi:10.1016/S0043-1354(01)00096-3

Penders, E.J.M., Martijn, A.J., Spenkelink, A., Alink, G.M., Rietjens, I.M.C.M.,

126

Hoogenboezem, W., 2012. Genotoxicity testing of samples generated during UV/H2O2

treatment of surface water for the production of drinking water using the Ames test in

vitro and the Comet assay and the SCE test in vivo. J. Water Supply Res. Technol. -

AQUA 61, 435–445. doi:10.2166/aqua.2012.069

Pennington, J.C., Thorn, K.A., Cox, L.G., MacMillan, D.K., Yost, S., Laubscher, R.D., 2007.

Photochemical Degradation of Composition B and Its Components. ERDC/EL TR-07-

16. Strategic Environmental Research and Development Program. US Army Corps of

Engineers.

Pera-Titus, M., Garcia-Molina, V., Banos, M.A., Gimenez, J., Esplugas, S., 2004.

Degradation of chlorophenols by means of advanced oxidation processes: A general

review. Appl. Catal. B Environ. 47, 219–256. doi:10.1016/j.apcatb.2003.09.010

Pereira, V.J., Weinberg, H.S., Linden, K.G., Singer, P.C., 2007. UV Degradation Kinetics

and Modeling of Pharmaceutical Compounds in Laboratory Grade and Surface Water

via Direct and Indirect Photolysis at 254 nm 41, 1682–1688.

Pi, Y., Schumacher, J., Jekel, M., 2005. The Use of para-Chlorobenzoic Acid (pCBA) as an

Ozone/Hydroxyl Radical Probe Compound. Ozone Sci. Eng. 27, 431–436.

doi:10.1080/01919510500349309

Plumb, R., Edwards, J., 1992. Color centers in UV-irradiated nitrates. J. Phys. Chem. 3245–

3247. doi:10.1021/j100187a014

Ratpukdi, T., Casey, F., DeSutter, T., Khan, E., 2011. Bromate Formation by Ozone-VUV in

Comparison with Ozone and Ozone-UV: Effects of pH, Ozone Dose, and VUV Power.

J. Environ. Eng. 137, 187. doi:10.1061/(ASCE)EE.1943-7870.0000313

Richardson, S.D., Plewa, M.J., Wagner, E.D., Schoeny, R., DeMarini, D.M., 2007.

Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection

by-products in drinking water: A review and roadmap for research. Mutat. Res. - Rev.

Mutat. Res. 636, 178–242. doi:10.1016/j.mrrev.2007.09.001

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. Environ. Sci. Technol. 38, 5476–5483. doi:10.1021/es035413p

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

127

processes O3/H2O2 and UV/H2O2. Water Res. 40, 3695–704.

doi:10.1016/j.watres.2006.09.008

Rosenkranz, H.S., Mermelstein, R., 1985. The genotoxicity, metabolism and carcinogenicity

of nitrated polycyclic aromatic hydrocarbons. J. Environ. Sci. Heal. Part C Environ.

Carcinog. Rev. 3, 221–272. doi:10.1080/10590508509373334

Sarathy, S., Mohseni, M., 2009. The fate of natural organic matter during UV/H2O2 advanced

oxidation of drinking water. Can. J. Civ. Eng. 36, 160–169. doi:10.1139/S08-045

Sarathy, S.R., Mohseni, M., 2007. The Impact of UV/H2O2 Advanced Oxidation on

Molecular Size Distribution of Chromophoric Natural Organic Matter. Environ. Sci.

Technol. 41, 8315–8320. doi:10.1021/es071602m

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. Environ.

Technol. 32, 1709–1718. doi:10.1080/09593330.2011.553843

Schalk, S., Adam, V., Arnold, E., Brieden, K., Voronov, A., Witzke, H.-D., 2006. UV-Lamps

for Disinfection and Advanced Oxidation - Lamp Types, Technologies and applications.

IUVA news 8, 32–37.

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. Environ. Sci. Technol. 45, 3657–3664.

doi:10.1021/es104240v

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. Environ. Sci. Technol. 37,

1933–1940.

Sharpless, C.M., Linden, K.G., 2001. UV photolysis of nitrate: Effects of natural organic

matter and dissolved inorganic carbon and implications for UV water disinfection.

Environ. Sci. Technol. 35, 2949–2955. doi:10.1021/es002043l

Sharpless, C.M., Page, M.A., Linden, K.G., 2003. Impact of hydrogen peroxide on nitrite

formation during UV disinfection. Water Res. 37, 4730–4736. doi:10.1016/S0043-

1354(03)00442-1

128

Shuali, U., Ottolenghi, M., Rabani, J., Yelin, Z., 1969. On the Photochemistry of Aqueous

Nitrate Solutions Excited in the 195-nm Band. Photochem. Aqueous Nitrate Solut. 73,

3445–3451.

Siegrist, R.L., Crimi, M., Simpkin, T.J., 2011. In Situ Chemical Oxidation for Groundwater

Remediation. Chapter 2: Fundamentals of ISCO using hydrogen peroxide, SERDP

ESTCP Environmental Remediation Technology Ser. doi:10.1007/978-1-4419-7826-4

Šojić, D. V., Anderluh, V.B., Orčić, D.Z., Abramović, B.F., 2009. Photodegradation of

clopyralid in TiO2 suspensions: Identification of intermediates and reaction pathways. J.

Hazard. Mater. 168, 94–101. doi:10.1016/j.jhazmat.2009.01.134

Sörensen, M., Frimmel, F.H., 1997. Photochemical degradation of hydrophilic xenobiotics in

the UV/H2O2-process: Influence of nitrate on the degradation rate of EDTA, 2-amino-1-

naphthalenesulfonate, diphenyl-4-sulfonate and 4,4’-diaminostilbene-2,2'-disulfonate

31, 2885–2891.

Swaim, P., Royce, A., Smith, T., Maloney, T., Ehlen, D., Carter, B., 2008. Effectiveness of

UV Advanced Oxidation for Destruction of Micro-Pollutants. Ozone Sci. Eng. 30, 34–

42. doi:10.1080/01919510701753390

Takanashi, H., Kishida, M., Nakajima, T., Ohki, A., Akiba, M., Aizawa, T., 2009. Surveying

the mutagenicity of tap water to elicit the effects of purification processes on Japanese

tap water. Chemosphere 77, 434–439. doi:10.1016/j.chemosphere.2009.06.038

Tarr M.A., 2003. Chemical Degradation Methods for Wastes and Pollutants., Environmental

and Industrial Applications. CRC Press.

Thorn, K.A., Cox, L.G., 2012. Ultraviolet Irradiation Effects Incorporation of Nitrate and

Nitrite Nitrogen into Aquatic Natural Organic Matter. J. Environ. Qual. 41, 865.

doi:10.2134/jeq2011.0335

Tizaoui, C., Mezughi, K., Bickley, R., 2011. Heterogeneous photocatalytic removal of the

herbicide clopyralid and its comparison with UV/H2O2 and ozone oxidation techniques.

Desalination 273, 197–204. doi:10.1016/j.desal.2010.11.036

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,

2087–2095. doi:10.1016/j.chemosphere.2006.09.043

Topalov, A., Molnár-Gábor, D., Kosanić, M., Abramović, B., 1999. Photomineralization of

129

the herbicide mecoprop dissolved in water sensitized by TiO2. Water Res. 34, 1473–

1478. doi:10.1016/S0043-1354(99)00304-8

U.S. Department of Health and Human Services., 2016. The Salmonella/E. coli Mutagenicity

Test or Ames Test. National Toxicology Program. [WWW Document]. URL

http://ntp.niehs.nih.gov/testing/types/genetic/invitro/sa/index.html

U.S. EPA, 2014. Integrated Risk Information System, 2-Methylphenol (CASRN 95-48- 7).

[WWW Document]. URL http://www.epa.gov/iris/subst/

United States Environmental Protection Agency, 2000. EPA Drinking Water Standards and

Health Advisories. Off. Water. Washington, D.C.

University of Hertfordshire, 2015. PPDB: Pesticide Properties DataBase, mecoprop. [WWW

Document]. URL http://sitem.herts.ac.uk/aeru/ppdb/en/Reports/430.htm (accessed

7.30.16).

USEPA, 2009. National Primary Drinking Water Regulations. United States Environmental

Protection Agency., Drinking Water Contaminants.

USEPA, 1993. Health Advisories for Drinking Water Contaminants. United States

Environmental Protection Agency. Office of Water Health Advisories. Lewis Publishers,

USA.

van der Gaag, M.A., Kruithof, J.C., Puijker, L.M., 1986. The influence of water treatment

processes on the presence of organic surrogates and mutagenic compounds in water. Sci.

Total Environ. 47, 137–153.

van Hoof, F., Janssens, J.G., van Dijck, H., 1985. Formation of mutagenic activity during

surface water preozonization and its removal in drinking water treatment. Chemosphere

14, 501–509.

van Kreijl, C.F., Kool, H.J., de Vries, M., van Kranen, H.J., de Greef, E., 1980. Mutagenic

activity in the rivers Rhine and Meuse in The Netherlands. Sci. Total Environ. 15, 137–

147.

Vilhunen, S., Sillanpää, M., 2010. Recent developments in photochemical and chemical

AOPs in water treatment: A mini-review. Rev. Environ. Sci. Biotechnol. 9, 323–330.

doi:10.1007/s11157-010-9216-5

Vione, D., Maurino, V., Minero, C., Lucchiari, M., Pelizzetti, E., 2004. Nitration and

hydroxylation of benzene in the presence of nitrite/nitrous acid in aqueous solution.

130

Chemosphere 56, 1049–1059. doi:10.1016/j.chemosphere.2004.05.027

Vione, D., Maurino, V., Minero, C., Vincenti, M., Pelizzetti, E., 2001. Formation of

nitrophenols upon UV irradiation of phenol and nitrate in aqueous solutions and in TiO2

aqueous suspensions. Chemosphere 44, 237–248. doi:10.1016/S0045-6535(00)00170-3

von Gunten, U., Hoigne, J., 1994. Bromate formation during ozonation of bromide-

containing waters: Interaction of ozone and hydroxyl radical reactions. Environ. Sci.

Technol. 28, 1234–1242. doi:10.1021/es00056a009

von Gunten, U., Oliveras, Y., 1998. Advanced Oxidation of Bromide-Containing Waters :

Bromate Formation Mechanisms. Environ. Sci. Technol. 32, 63–70.

doi:10.1021/es970477j

von Sonntag, C., Schuchmann, H.P., 1992. UV Disinfection of Drinking Water and By-

Product Formation – Some basic considerations. Aqua 41, 67.

Wardle, B., 2009. Principles and Applications of Photochemistry, Journal of Modern Optics.

doi:10.1080/09500348914551321

Weishaar, J., Aiken, G., Bergamaschi, B., Fram, M., Fujii, R., Mopper, K., 2003. Evaluation

of specific ultra-violet absorbance as an indicator of the chemical content of dissolved

organic carbon. Environ. Sci. Technol. 37, 4702–4708. doi:10.1021/es030360x

Westerhoff, P., Aiken, G., G.Amy, Debroux, J., 1998. Relationships between the structure of

natural organic matter and its reactivity rowards molecular ozone and hydroxyl radicals.

Wat. Res. 33, 2265–2276.

WHO, 2000. Concise International Chemical Assessment Document 20, Mononitrophenols,

World Health Organization, Geneva.

Wilson, C., Tisdell, C., 2001. Why farmers continue to use pesticides despite environmental,

health and sustainability costs. Ecol. Econ. 39, 449–462. doi:10.1016/S0921-

8009(01)00238-5

Wilzbach, K.E., Rausch, D.J., 1970. Photochemistry of Nitrogen Heterocycles. Dewar

Pyridine and Its Intermediacy in Photoreduction and Photohydration of Pyridine. J. Am.

Chem. Soc. 92, 2178–2179.

World Health Organization, 2011. Guidelines for Drinking-water Quality. Fourth Ed. 155–

201.

Xenometrix Inc., 2015. Ames II Microplate Format Mutagenicity Assay.

131

Xu, G., Liu, N., Wu, M., Bu, T., Zheng, M., 2013. The Photodegradation of Clopyralid in

Aqueous Solutions: Effects of Light Sources and Water Constituents BT - Industrial &

Engineering Chemistry Research. doi:10.1021/ie302844v

Zepp, R.G., Braun, A.M., Hoigne, J., Leenheer, J.A., 1987. Photoproduction of hydrated

electrons from natural organic solutes in aquatic environments. Environ. Sci. Technol.

21, 485–90. doi:10.1021/es00159a010

132

Appendix

A. Figures

Figure A.1. Normalized spectrum of the Low Pressure lamp used for the collimated beam

experiments (peak at 254 nm) (obtained by Trojan UV Technologies).

133

Figure A.2. Absorbance spectra of solutions (10 mg/L) of all six compounds in Milli-Q water

at neutral pH.

Figure A.3. The main photoproduct (2-(4-hydroxy-2-methylphenoxy)propanoic acid)

produced by LP-UV photolysis and LP-UV/H2O2 oxidation treatment of mecoprop (10

mg/L). Error bars represent the standard deviation for duplicate measurements.

0

5000

10000

15000

20000

25000

30000

35000

0 500 1000 1500

Pea

k A

rea

UV fluence (mJ/cm2)

No peroxide

10mg/L peroxide

30mg/L peroxide

10mg/L H2O2

30mg/L H2O2

No H2O2

134

Figure A.4. Conversion (%) of organic chlorine to chloride ion in aqueous solutions under

various LP-UV/H2O2 combinations for a) mecoprop (10 mg/L) and b) clopyralid (20 mg/L).

Error bars represent the standard deviation for duplicate measurements.

0

20

40

60

80

100

0 500 1000 1500

Conversion to Cl-

(%)

UV fluence (mJ/cm2)

a

No peroxide

10mg/L peroxide

30mg/L peroxide

No H2O2

10mg/L H2O2

30mg/L H2O2

0

20

40

60

80

100

0 500 1000 1500

Conversion to Cl-

(%)

UV fluence (mJ/cm2)

b

10mg/L peroxide

30mg/L peroxide

60mg/L peroxide

10mg/L H2O2

30mg/L H2O2

60mg/L H2O2

135

Figure A.5. Production of the m/z =210 product from UV/H2O2 treatment of metaldehyde (5

mg/L) as a function of UV fluence for both H2O2 doses applied. Error bars represent the

standard deviation for duplicate measurements.

Figure A.6. Acetic acid formation during LP-UV/H2O2 treatment of metaldehyde (5 mg/L).

Error bars represent the standard deviation for duplicate measurements.

0

50000

100000

150000

200000

250000

300000

0 500 1000 1500

Pea

k A

rea

UV fluence (mJ/cm2)

210 (1) - 10mg/L

210 (1) - 30mg/L

10mg/L H2O2

30mg/L H2O2

0

1

2

3

4

10 30

Acetic acid

(mg/L)

Hydrogen peroxide (mg/L)

500 uv

1000 uv

1500 uv

500 mJ/cm2

1000 mJ/cm2

1500 mJ/cm2

136

Figure A.7. Absorption spectra of the Suwannee River, Nordic Lake and Pony Lake NOM

(concentration of 4 mg/L for all three NOMs, UV wavelength range 190-350 nm).

137

Figure A.8. Scheme of the drinking water treatment plant with the sampling points where samplings (a) and (b) took place, for Ames testing

purposes.

138

B. Elemental Composition of NOMs (IHSS)

NOM Cat. No. H2O Ash C H O N S P

Suwannee

River 2R101N 5.69 4.01 50.70 3.97 41.48 1.27 1.78 nd

Nordic

Lake 1R108N nd 41.4 53.17 5.67 nd 1.10 nd nd

Pony Lake 1R109F 4.32 1.25 52.47 5.39 31.38 6.51 3.03 0.55

* H2O content is the % (w/w) of H2O in the air-equilibrated sample (a function of relative

humidity). Ash is the % (w/w) of inorganic residue in a dry sample. C, H, O, N, S, and P are

the elemental composition in % (w/w) of a dry, ash-free sample. ‘‘nd’’ means that an item

was not determined.

139

C. Quality Assurance / Quality Control (QA/QC)

Radiometer calibration

The radiometer (ILT1700) calibration took place at the International Light (ILT) office in the

USA, via LOT-QuantumDesign GmbH, and included the following:

- Electrical accuracy and linearity check

- Recalibration to NIST standards

- I/O check

- PIR (Peak Irradiance Response): Irradiance response, (amps)(cm2)(watt-1), at peak

sensitivity of an IL detector type: SED240/NS254/W

140

LC-MS/MS calibration curves and fragmentations for pesticide detection

Figure C.1. Clopyralid: an example of a calibration curve (on the left) denoting with an arrow the data for which the two

fragmentations were considered (on the right).

141

Figure C.2. Metaldehyde: an example of a calibration curve (on the left) denoting with an arrow the data for which the two


142

Figure C.3. Mecoprop: an example of a calibration curve (on the left) denoting with an arrow the data for which the two


143

LC-MS/MS standard, blank and pesticide responses

Table C.1. Standard concentrations for pesticides.

std CLOP METALDH MCPP

ng/ml ng/ml ng/ml

std-0 0.0 0.0 0.0

std-1 41.8 47.6 37.4

std-2 83.6 95.1 74.8

std-3 167.2 190.3 149.7

std-4 334.4 380.6 299.4

std-5 668.7 761.1 598.8

Table C.2. Standard and blanks responses for internal standard (FNPF).

Sample

FNPF (ISTD)

Name RT Height Area

std-0 8.06 186034 902562

std-1 8.06 189861 922049

std-2 8.07 193500 943850

std-3 8.07 188719 924258

std-4 8.07 193991 945593

std-5 8.08 188351 918427

BLANK 8.06 178607 869583

BLANK 8.06 171144 827564

BLANK 8.07 171858 827995

BLANK 8.07 163094 796670

BLANK 8.06 163994 796130

BLANK 8.07 167128 816510

std-0 8.07 160695 778293

std-1 8.07 161320 787898

std-2 8.07 161244 783202

std-3 8.05 150607 733497

std-4 8.06 149323 728416

std-5 8.06 149377 723381

144

Table C.3. Standard and blanks responses for clopyralid.

Sample CLOPYRALID Qualifier (192.0 -> 147.9)

Results

Name RT Height Area RT Height Area Ratio Calc.

Conc.

std-0 1.31 28 130 1.36 25 197 0.4

std-1 1.36 2058 13165 1.36 1414 9189 70 41.0

std-2 1.36 4212 28016 1.35 2753 18426 66 83.6

std-3 1.36 9033 57322 1.37 5893 37951 66 168.8

std-4 1.35 19541 122188 1.36 12421 79253 65 330.4

std-5 1.35 44150 273157 1.35 28076 176382 65 673.6

BLANK 1.34 37 241 1.25 25 107 0.8

BLANK 1.36 47 326 1.35 16 24 1.2

BLANK 1.32 22 88 1.33 25 119 0.3

BLANK 1.36 44 250 1.19 47 392 0.9

BLANK 1.34 15 47 1.25 65 759 0.2

BLANK 1.30 17 66 1.33 67 801 0.2

std-0 1.31 18 58 1.17 55 711 0.2

std-1 1.35 1616 10219 1.34 1174 7394 72 37.3

std-2 1.35 3678 23476 1.34 2491 16016 68 84.4

std-3 1.35 7625 48668 1.34 4757 31749 65 179.8

std-4 1.35 14701 93829 1.35 9555 61534 66 329.5

std-5 1.36 35070 211784 1.35 20999 132768 63 665.0

145

Table C.4. Standard and blanks responses for metaldehyde.

Sample METALDEHYDE Qualifier (194.1 -> 106.0)

Results


Conc.

std-0 5.48 3 18 5.48 2 8 0.0

std-1 5.44 69121 421717 5.44 5426 34339 8.1 42.5

std-2 5.45 143688 883587 5.45 10880 68482 7.8 86.5

std-3 5.45 307763 1905891 5.45 23622 147275 7.7 188.3

std-4 5.44 636019 3972525 5.45 50235 310938 7.8 375.6

std-5 5.44 1285855 8107334 5.45 106671 658809 8.1 756.9

BLANK 5.50 14 52 5.43 7 37 0.0

BLANK 5.43 20 95 5.50 1 4 0.0

BLANK 5.40 1 1 5.37 1 4 0.0

BLANK 5.59 16 87 5.65 2 12 0.0

BLANK 5.39 0 0 5.24 1 3 0.0

BLANK 5.40 2 9 5.72 1 4 0.0

std-0 5.45 3 14 5.49 3 13 0.0

std-1 5.42 59027 360256 5.43 4445 28123 7.8 42.5

std-2 5.43 128729 784715 5.44 9529 59000 7.5 92.5

std-3 5.44 260731 1600174 5.44 19492 120505 7.5 199.0

std-4 5.44 510096 3161639 5.45 39561 242843 7.7 387.5

std-5 5.44 1033901 6457149 5.44 82924 511955 7.9 764.7

146

Table C.5. Standard and blanks responses for mecoprop.

Sample MECOPROP Qualifier (215.0 -> 143.0)

Results


Conc.

std-0 6.63 22 106 6.72 44 262 245.9 0.0

std-1 6.69 41190 201712 6.69 11934 61093 30.3 30.1

std-2 6.69 90454 450615 6.69 25618 127894 28.4 64.5

std-3 6.71 224931 1096709 6.71 63800 313258 28.6 154.0

std-4 6.69 466386 2310216 6.70 139020 688002 29.8 298.0

std-5 6.73 1001647 5034046 6.73 300061 1489499 29.6 594.7

BLANK 6.67 170 805 6.73 113 607 75.5 0.1

BLANK 6.74 98 339 6.71 95 617 182.3 0.1

BLANK 6.67 115 439 6.68 24 109 24.7 0.1

BLANK 6.67 69 276 6.66 83 341 123.3 0.0

BLANK 6.75 135 651 6.69 74 502 77.2 0.1

BLANK 6.64 48 114 6.66 52 379 332.3 0.0

std-0 6.71 69 239 6.66 24 90 37.5 0.0

std-1 6.68 32628 161888 6.68 9573 49211 30.4 28.2

std-2 6.68 77135 378942 6.68 22673 111691 29.5 65.4

std-3 6.68 172302 843430 6.68 49799 245556 29.1 149.5

std-4 6.69 375154 1848752 6.69 111960 549218 29.7 308.3

std-5 6.68 812412 4018435 6.68 246207 1206843 30.0 601.3

147

LC-MS/MS calibration curve for nitrophenol detection

Figure C.4. Nitrophenol: an example of a calibration curve (on the left) denoting with an

arrow the data for which two fragmentations were considered.

148

Figure C.5. The LC-MS/MS chromatogram of the nitrophenols detected for the 0/15,

1500/15 and 2100/15 mJ/cm2 UV/H2O2 combinations for both duplicates (R1 and R2)

performed.

Transition 138>108

2000/15 R1 1500/15 R1 0/15 R1 BLANK

2000/15 R2 1500/15 R2 0/15 R2 BLANK

Transition 138>108

149

Table C.6. Standard concentrations for nitrophenol (NP).

NP

ng/ml

std-0 0.0

std-1 80.0

std-2 160.0

std-3 320.0

std-4 640.0

Table C.7. Standard and blanks responses for nitrophenol (NP).

Sample NP (138 -> 108) Qualifier (138 -> 92)


Conc.

std-0 4.27 55 248 4.27 35 164 4.0

std-1 4.25 20903 103458 4.26 1752 8759 8.5 78.5

std-2 4.25 41727 209773 4.25 3451 17290 8.2 155.3

std-3 4.27 89126 439138 4.27 7060 35441 8.1 321.0

std-4 4.27 175803 871227 4.27 13970 69873 8.0 633.1

blank 4.23 148 1430 4.28 14 46 4.8

blank 4.23 86 412 4.27 24 70 4.1

blank 4.30 158 527 4.28 23 92 4.2

blank 4.28 259 1107 4.26 24 97 4.6

blank 4.38 143 804 4.29 4 15 4.4

std-0 4.33 123 771 4.37 3 14 4.3

std-1 4.29 19740 99292 4.28 1695 8515 8.6 75.5

std-2 4.29 42011 209359 4.28 3299 16703 8.0 155.0

std-3 4.29 90337 448129 4.29 7019 35427 7.9 327.5

std-4 4.29 178005 888875 4.29 13582 68257 7.7 645.8

150

Ion Chromatography (IC) calibration curves for NO3- and NO2

- ions

Figure C.6. Calibration curve for the nitrate ion (ion chromatography analysis).

151

Figure C.7. Calibration curve for the nitrite ion (ion chromatography analysis).

152

Figure C.8. IC calibration for TOC analysis.

153

Figure C.9. NPOC calibration for TOC analysis.

154

Figure C.10. TC calibration for TOC analysis.

155

D. Calculations

Figure D.1. Calculations of maximum UV fluence required to exceed the EU limit of 0.1 mg NO2-/L.

156

E. Photographs

Photo 1. Sample pre-concentration with Solid Phase Extraction for the subsequent detection

of nitrophenols.

157

Photo 2. Dry solid extracts of the Natural Organic Matters (NOMs) used.

158

Photo 3. Irradiation of a sample with UV light (254 nm).

159

Photo 4. Bench-scale collimated beam apparatus equipped with a LP-lamp and a 10-sample

carousel.

low pressure uv/h o treatment of the pesticides metaldehyde, … · 2019-11-08 · low pressure...

Documents