Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Potential formation of mutagenicity by low pressure-UV/H2O2 during the treatment of
nitrate-rich source waters
S. Semitsoglou-Tsiapoua,b,† , M. R. Templetona, N. J. D. Grahama, S. Mandalc, L. Hernández
Lealb and J. C. Kruithofb
a. Department of Civil and Environmental Engineering, Imperial College London, South
Kensington Campus, SW7 2AZ, London, UK.
b. Wetsus, European centre of excellence for sustainable water technology, P.O. Box 1113,
8900 CC, Leeuwarden, the Netherlands.
c. Duisburg-Essen University, Universitätsstraße 2, 45117, Essen, Germany.
† Corresponding author: [email protected], Oostergoweg 9, PO Box
1113, 8900 CC, Leeuwarden
1
1
2
3
4
5
6
7
8
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
LP-UV/H2O2 treatment of NOM-containing synthetic waters led to nitrite, nitrophenol and
measurable but not mutagenic Ames responses in the presence of nitrate, where the NOM
type affected the response levels.
Water Impact Statement
Despite the increasing application of UV advanced oxidation processes in water treatment,
few studies have considered the potential formation of nitrated by-products and potential
toxicity issues. This study showed that UV advanced oxidation of waters containing nitrate
and natural organic matter can produce nitrite and nitrophenols, but measurable mutagenicity
formation in Salmonella typhimurium was not significant compared to standard thresholds.
Introduction
Advanced oxidation processes (AOPs) have been increasingly incorporated for drinking
water treatment applications, due to the rise of recalcitrant micro-pollutants in the water, and
because of the effective non-selective degradation achieved1. More specifically, UV-based
AOPs (e.g. UV/O3, UV/H2O2, UV/TiO2, UV/HOCl) have been widely applied in the past
decade due to the combined effect of UV photolysis and hydroxyl radical oxidation for the
degradation of organic micro-pollutants2 (as well as for disinfection purposes). AOPs have
been shown to produce reaction products3, which depend greatly on the water matrix, the
process and the conditions applied. This work has focused specifically on the application of
the Low Pressure (LP)-UV/H2O2 process. UV/H2O2 treatment is becoming popular since it
2
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
has been shown in numerous previous studies to successfully treat various organic
contaminants, e.g. endocrine disrupting compounds (EDCs)4, polycyclic aromatic
hydrocarbons (PAHs)5-6, pharmaceuticals7 and pesticides8-9. It is a promising technology due
to the effective, non-selective degradation of micropollutants via OH-radical-assisted
oxidation with an additional contribution of UV photolysis, while disinfection of the water
takes place at the same time. Specifically, Low Pressure (LP) lamp applications exhibit
advantages over Medium Pressure (MP) applications due to higher energy efficiency, longer
lifetime, and minimum formation of by-products of concern, such as nitrite and bromate10.
The formation of nitrogenous reaction products via the photolysis of nitrate, followed by a
complex series of reactions, has been studied previously only to a limited extent11. In contrast,
the photolysis of nitrate has been studied extensively, especially for medium pressure (MP)-
UV wavelengths (λ < 240 nm) where its absorption is the highest, leading to peroxynitrite
(OONO-) formation as the main intermediate species. Peroxynitrite decomposition leads to
the production of nitro- and nitroso- radicals, with nitrate and nitrite as end products12-13. In
the presence of an organic matrix, the incorporation of inorganic nitrogen into the organic
matrix has been demonstrated14. 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 substances such as benzene and phenol15-18, are hydroxylation, nitration and
nitrosation, the latter being enhanced in the absence of oxygen19. 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 in
NOM-nitrate rich water by UV/H2O2 treatment are few; Martijn et al.17-18 reported the
formation of 2- and 4-nitrophenol and 4-nitrocatechol when irradiating phenol as surrogate
for NOM in the presence of nitrate with MP-UV, whereas Kolkman et al.20 found a variety of
nitrogen containing compounds by MP-UV photolysis and MP-UV/H2O2 treatment of
3
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
synthetic (nitrate-rich, with Pony Lake NOM) and full-scale water samples, respectively,
confirming the following three reaction products: 4-nitrophenol, 4-nitrocatechol, and 2-
methoxy-4,6-dinitrophenol.
Since some nitrated compounds are known to be toxic21 and as a result may pose a health risk
in drinking water, it is critical that their presence and potential contribution to the water’s
toxicity are investigated. For that reason, various bioassays have been developed, such as the
Ames assay, which is widely used as a standard screening method for the detection of
mutagenic compounds22. The Ames assay was selected, since regulators consider that, ‘‘a
substance that is mutagenic in the Salmonella typhimurium bacterium is more likely than not
to be a carcinogen in laboratory animals, and thus, by extension, present a risk of cancer to
humans’’ (U.S. Department of Health and Human Services, 2016)23. The test is advantageous
in terms of its easy application, rapid generation of results and low cost, thus rendering it a
useful tool for assessing potential carcinogenicity. It makes use of strains of the bacterium
Salmonella typhimurium that carry gene mutations that inhibit histidine synthesis; these
mutations can be reversed in the presence of mutagenic chemical compounds. 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 required24. There are two commonly
applied Salmonella typhimurium strains, the frameshift strain TA98 and the base-pair
detecting strain TA100. It has been shown by previous studies that the TA98 strain (and
especially the TA98(-S9) combination, where (-S9) denotes the absence of a rat liver
metabolic activation system) is most responsive in samples treated with oxidation
processes17,25-28, whereas the TA100 strain (both with and without S9 addition) has exhibited
positive responses in cases of chlorinated samples29-32. Therefore, in this study, the strain
TA98, in the absence of a rat liver metabolic activation system (S9), was selected.
4
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
The mutagenicity of MP-UV/H2O2-treated water has been investigated previously and
showed a significant Ames assay response and the important role that the presence of nitrate
played in this response17-18, 20, 27, 33 with only one study giving negative Ames assay results
where natural quartz sleeves with a cut-off of UV light< 240 nm were utilized 34. Martijn et
al.17-18 showed in a preliminary risk assessment, by converting the Ames responses of full-
scale water samples treated by MP-UV and MP-UV/H2O2 into 4-nitroquinoline oxide (4-
NQO) equivalent concentrations, that there is reason for concern and further investigation is
required. Although in theory LP-UV and LP-UV/H2O2 processes are expected to cause little,
if any, mutagenicity formation, very few studies have examined this in detail, and these have
suggested that LP-UV-based treatment did not cause any mutagenicity formation33 or a very
weak response35. In order to provide further clarification of potential effects, this paper
summarises the results of recent research which has evaluated: a) the formation of
mutagenicity, nitrite and nitrophenol in laboratory tests using synthetic water containing
NOM and nitrate, and b) the formation of mutagenicity in samples from a full-scale drinking
water plant where LP-UV/H2O2 treatment was applied.
Methods
Chemicals
Sodium nitrate, hydrogen peroxide (30%), HPLC-grade acetone, ammonia (30%), methanol
and ethyl acetate (99.9%) were supplied by VWR (the Netherlands). Acetonitrile was
supplied by Merck (the Netherlands), formic acid (99%) by Boom (the Netherlands) and
fenoprofen by Sigma-Aldrich (the Netherlands). Laboratory-grade water (LGW) was
produced by a Milli-Q Advantage A10 system (Merck Millipore, Darmstadt, Germany).
Two types of reference natural organic matter were obtained from the IHSS (International
Humic Substances Society) as dry solid extracts: Suwannee River NOM (2R101N) and Pony
5
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
Lake NOM (1R109F). Suwannee River NOM is a widely used reference aquatic NOM, and
Pony Lake NOM is a reference fulvic acid-containing NOM. The nitrogen content for
Suwannee River NOM was 1.27 % (w/w), whereas for Pony Lake NOM it was 6.51 %
(w/w); the highest nitrogen content among the IHSS NOMs available. Their difference is
reflected by the SUVA254 (SUVA254=Abs254/DOC) values obtained experimentally, 3.6 and
1.6 L mg−1 m−1 for Suwannee and Pony Lake, respectively, showing that Suwannee NOM
consists of humic, highly aromatic and hydrophobic matter with high molecular weight
organic fractions, whereas Pony Lake is mostly hydrophilic non-humic (fulvic) matter. The
A254/A203 UV absorbance ratios36 were calculated as 0.42 and 0.32, respectively, suggesting
aromatic rings highly substituted with hydroxyl, carbonyl, ester and carboxyl groups for
Suwannee River NOM, and rings predominantly substituted with aliphatic functional groups
for Pony Lake NOM.
Phenomenex Strata-X Polymeric Solid Phase Extraction (SPE) columns (200 mg Oasis HLB
3 mL glass cartridges) were supplied by Phenomenex (the Netherlands).
Stock solutions of the two types of NOM and NaNO3 were prepared in MilliQ water, and test
solutions were produced from the required amounts of the stock solutions and hydrogen
peroxide solution.
UV collimated beam experiments
UV exposure experiments were carried out with a bench scale collimated beam apparatus,
equipped with a 25-Watt low pressure (LP) mercury arc discharge lamp without a lamp
sleeve. The emission spectrum of the UV lamp, obtained from Trojan UV Technologies
(London, Ontario, Canada) consisted of a strong, almost exclusive emission at 254 nm
(Figure S.1, ESI). The experimental conditions for the LP-UV collimated beam experiments
are given elsewhere37. The exposure time for the desired UV fluences was calculated using a
fluence calculation spreadsheet based on Bolton and Linden38. According to this protocol, the
6
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
UV fluence applied, and consequently the irradiation time, were adjusted for the absorbance
of each solution at 254 nm.
Regarding the study of mutagenicity formation by LP-UV/H2O2 treatment, experiments were
performed in laboratory-grade water spiked with either Suwannee River NOM or Pony Lake
NOM (both at 4 mg/L, which corresponded to 2 and 2.1 mg C/L as DOC for Suwannee River
and Pony Lake NOM, respectively) in the presence of nitrate (50 mg/L). The UV fluences
applied were 0 (no irradiation), 1500 and 2000 mJ/cm2, corresponding to the upper-end of the
fluence range that is commonly applied (1500 mJ/cm2) and worst case (2000 mJ/cm2) UV
fluences, with a peroxide dose of 15 mg/L. All experiments were performed in duplicate and
at room temperature (23-25 ˚C).
Analytical methods
Nitrite and nitrate ions were quantified by ion chromatography (detection limits of 0.05 and
0.1 mg/L, respectively), using a Metrohm IC Compact 761 ion chromatograph (IC) 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.
The analysis for the detection of nitrophenols was performed by a method involving
compound extraction using polymeric SPE tubes (Phenomenex Strata-X Polymeric 200 mg /
3 ml SPE extraction tubes), followed by liquid chromatography tandem mass spectrometry
(LC-MS/MS). For the SPE extraction, acetone/ethyl acetate (1:1), methanol and acidified
blank sample (pH=2) were used consecutively for the activation of the SPE cartridges. The
cartridges were then loaded with 0.5 L of acidified sample. Elution was performed with 7.5
mL of acetone/ethyl acetate (1:1). A volume of 50 µL of the internal standard solution
(fenoprofen in methanol) was added to the eluates, which were further evaporated with a
nitrogen stream until the volume reached 0.2 mL. Finally, 0.8 mL of ultrapure water was
added before analysis.
7
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
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 employed and 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. As internal standard, fenoprofen was used. The compounds were measured with
specific QQQ transitions. For instrument control and data analysis Agilent MassHunter
software was used. The quantitation limit (QL) and detection limit (DL) of the method were
0.02 and 0.007 μg/L, respectively.
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 within the LC-OCD system. The
column 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).
H2O2 concentrations were measured using the triiodide method39.
All experiments were performed in duplicate and statistical t-test comparisons were
conducted with a 95% confidence interval.
Water treatment works sampling
Samples were collected from a selected drinking water treatment works (WTW) in the UK,
incorporating LP-UV/H2O2 technology and treating 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,
8
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
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. Two separate samplings (referred to as (a) and (b)) took
place in February and April 2016, to cover expected differences in water composition and the
maximum design UV/H2O2 conditions that were 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 in both cases). These doses, along with water
quality and operational parameters for the WTW sampling events can be found in Table 1.
Water samples were collected from the inlet (only for sampling (b)), pre-AOP, post-AOP and
post-GAC treatment steps, to assess nitrite formation, NOM fate and potential mutagenicity
in terms of Ames assay response during treatment.
Ames II Mutagenicity assay
Both the synthetic and full-scale water samples were subjected to the Ames II Mutagenicity
assay (the TA98(-S9) combination) in order to evaluate their mutagenic potential to induce
reverse mutations in Salmonella typhimurium. All water samples were extracted by SPE
according to the method described by Heringa et al.27. The extracts were 20,000x
concentrated after SPE extraction, diluted 25x in the Ames II assay (i.e. relative enrichment
factor was 800). The Ames II assay was performed according to the Xenometrix protocol40. A
blank (SPA Reine = spa bottled water) was treated identically as a method control. Solvent
control (100% DMSO), and positive controls (2-nitrofluorene + 4-nitroquinoline N-oxide)
were included. In total, 9 replicates in the control conditions (method and solvent controls)
and 3 replicates for each extract (i.e. sample) were prepared. Per replicate, 48 wells are used,
therefore the mean number of positive wells was expressed as a percentage (%) of the 48.
Both the sample extraction and the Ames assay were performed by VITO laboratories in
Belgium.
9
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
In order to interpret the findings in terms of positive or negative mutagenicity, the number-
fold inductions 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
number-fold inductions 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. The
baseline value represented the mean number of positive wells in solvent control conditions +
1 SD. A 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. The raw data and calculated values are given in Table S.1 (ESI).
Number-fold inductions in revertant numbers over the solvent control are considered as
positive if > 3.0, whereas number-fold inductions in revertant numbers over the baseline
value are positive if > 2.0. A sample (or compound) that shows a clear number-fold increase
> 2.0 (baseline) and significant difference (p<0.05 in the t-test), is classified as mutagenic.
When the number-fold induction is below these values the test substances are not mutagenic
towards the Salmonella typhimurium strain TA98.
Results and Discussion
Ames II assay of synthetic water samples
The Ames II assay 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 a dose of 15 mg/L H2O2. The results for the water samples, including
the method and solvent control, are given in Figure 1 (a-b).
Comparing the combinations of 0 UV+H2O2+NO3 and 2000 UV+H2O2+NO3 the Ames II
assay results for Suwannee River NOM (Figure 1a) were not statistically different. In
10
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
contrast, for Pony Lake NOM, the 2000 UV+H2O2+NO3 combination showed a significant
increase in positive wells levels (Figure 1b). 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), indicating 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)17-18 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 assay response only when nitrate
was present. Overall, significant variations in the Ames II assay 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 of those observed by Martijn
et al. (2014)17 (around 20 positive wells) when MP-UV/H2O2 treatment was applied with a
lower UV fluence and H2O2 dose (560 mJ/cm2 with 6 mg/L H2O2).
In order to interpret the results in this work in terms of mutagenicity with the Ames II assay,
the number-fold inductions over the solvent control and the number-fold inductions over the
baseline value (explained in the Methods Section) were calculated. Figure 2 shows that all
number-fold inductions over the solvent control were below 3 and all number-fold inductions
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 of the total absence of mutagenic potency, as
other mutagenic mechanisms cannot be excluded.
Based on the calculated SUVA values (3.6 and 1.6 L mg−1 m−1 for Suwannee River and Pony
Lake NOM, respectively), Suwannee River NOM is expected to have a more aromatic nature
11
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
compared to Pony Lake NOM, and consequently be more prone to OH-radical attack under
UV/H2O2 treatment at the electron-rich sites41. Nevertheless, this would be the case for low
molecular weight organic compounds42, whereas based on the high SUVA value it is
suggested that Suwannee River NOM contains high molecular weight organic fractions.
According to literature values, the reaction rate constant of these two NOMs with OH-
radicals are reported to be 3.3 108 M-1 s-1 for Suwannee River NOM43 and 2.03 108 M-1 s-1 for
Pony Lake NOM44 (by competition kinetics), so the difference in reactivity is small. If,
however, the degradation of Suwannee River NOM takes place to a greater extent than Pony
Lake NOM, and mutagenic products are formed as intermediates, the degradation of those
intermediates could also take place, therefore lowering the final Ames levels.
The difference between the Ames levels obtained by the two NOMs was not statistically
significant; nevertheless, the role of photolysis as well as the role of the presence of nitrate
were apparent only for Pony Lake NOM under the conditions applied, as already explained
earlier. The incorporation of the nitrate-nitrogen into the organic matrix and the formation on
nitrated/nitrosated compounds by MP-UV treatment has been demonstrated for both
Suwannee River14 and Pony Lake NOM17,20. This could be the reason why mutagenicity levels
(even though low) were measured for both NOMs. It can be suggested that this N-
incorporation is also affected by the nitrogen content of the NOM, especially when it comes
to N-nitrosation45; since this content is 5 times higher for Pony Lake NOM (1.27 % (w/w) and
6.51 % (w/w) for Suwannee River and Pony Lake NOM, respectively), nitrosation could have
been favoured, giving rise to more, potentially mutagenic, compounds. Additionally, the
more aliphatic composition of Pony Lake NOM could have contributed to greater nitrosation,
since, according to Thorn and Cox45, the greater the aliphatic carbon content, the greater the
concentration of activated methylene and methyl carbons that are available for nitrosation and
subsequent rearrangement. Although in the Thorn and Cox study45 the NOMs were treated
12
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
with nitric acid and not a nitrate salt, they state that some detected N-compounds were also
observed when sodium nitrate was used.
Nitrite and nitrophenol formation
While no nitrite was formed in dark samples (i.e. no UV irradiation), the nitrite
concentrations obtained for the irradiated samples (1500/15 and 2000/15, in mJ/cm2 / mg/L)
for both NOMs were 0.08-0.09 mg/L (Figure S.2, ESI), which were very close to the 0.1
mg/L EU regulatory limit for nitrite. However, this limit might be exceeded if the H 2O2 dose
is increased46. It should be noted that conflicting results have been reported regarding the role
of hydrogen peroxide. Thus, Lu et al.47, 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, while in contrast Sharpless et al.48 reported 0.16 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 0.19 mg/L NO2- with the addition of 10 mg/L H2O2, thereby showing that
nitrite formation by MP-UV can be enhanced by H2O2 addition. The nitrite formation with
and without H2O2, under the same UV fluences as in this work, and both in the presence and
absence of NOM, can be found in our previous study46. It was observed that, a) the H2O2
concentration (up to 50 mg/L) and nitrite yield (0.05-0.13 mg/L) were directly proportional
when 50 mg/L nitrate was used in NOM-free water, and b) the presence of NOM (either
Suwannee River, Nordic Lake or Pony Lake NOM) increases the formation of nitrite for all
UV/H2O2 dose combinations and NOM concentrations, compared to UV photolysis and
NOM-free waters. Specifically for Pony Lake NOM (for comparison reasons) the nitrite
concentrations under the fluences of 1500 and 2000 mJ/cm2 without H2O2 were 0.11 and 0.13
mg NO3- /L, respectively, compared to the 0.09 mg NO3
- /L value for both 1500/15 and
2000/15 (mJ/cm2 / mg/L) combinations.
13
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
In addition, the presence of nitrophenols was investigated (only with Pony Lake NOM) since
their formation has been observed previously (as mentioned in the Introduction), due to the
reaction of phenol with UV photolysis intermediates of NO3. Martijn et al. (2014)17 used
Pony Lake NOM in the presence and absence of nitrate, hypothesizing the Ames response
was caused by 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 the tests in this study
stemming from reactions in the NOM-NO3 system, induced by LP-UV photolysis (i.e. 254
nm).
The LC-MS/MS method was used to analyse for the following nitrophenols: 2-nitrophenol, 4-
nitrophenol and a combination of both. All samples yielded trace amounts of mono-
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 3. 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 quantitation limit (0.02
μg/L). For the two UV/H2O2 combinations, a significant nitrophenol concentration (~0.04
μg/L) above the quantitation limit was observed.
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)20 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 two
phenols detected, 2- and 4-nitrophenol, were also found by Martijn et al. (2014)17 along with
14
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
4-nitrocatechol, when irradiating samples containing phenol and NO3 with a MP lamp. In our
experiments with the LP lamp, a significant concentration of nitrophenol was found for both
UV fluences (i.e. 1500 and 2000 mJ/cm2). The measured nitrophenol concentration was less
for the 2000 mJ/cm2 than for the 1500 mJ/cm2 fluence (0.037 and 0.046 μg/L, respectively),
although the difference was not statistically significant (Figure 3). This finding agrees with
Martijn et al. (2014)17 who showed an initial increase in nitrophenol concentration as a
function of the irradiation time, followed by a decrease at longer irradiation times, attributed
to oxidation of the previously produced nitrophenols.
Ames II assay of full-scale water samples
Samples collected from the UK drinking water treatment plant after various treatment steps
on two days in different months (sampling (a) was in February and sampling (b) in April)
were also analysed with the Ames II assay. On both sampling days, duplicate samples were
collected. The water quality and operational parameters, relevant for the LP-UV/H2O2
treatment of the WTW, for the two samplings are given in Table 1.
The number of positive wells from the Ames II assay for the water treatment works samples
are shown in Figure 4. Sampling (a) showed a small but significant decrease in Ames II assay
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 ultrafiltration), 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 assay response
was indeed influenced by the different water characteristics and/or treatment conditions of the
two periods.
15
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
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 co-metabolism of
refractory compounds, have been observed in laboratory models or natural water bodies.
Nevertheless, for drinking water treatment production, the possibility of bio-degradation by-
products has been demonstrated only after O3-BAC treatment49. Regarding the second factor,
according to Camper et al.50, pathogens such as Salmonella can colonize and persist in the
carbon bed, especially on virgin GAC filters. The 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. The likelihood, though, of this speculation is also very
low since the water samples would have a reduced microbial load by having been enriched by
the SPE method as well as having been subjected to high UV fluences before entering the
GAC filters. The most probable explanation is an experimental error during the second
sampling.
Figure 5 depicts the number-fold increases obtained for the WTW samples. All number-fold
increases over the solvent control were < 3.0, and all number-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.
In an attempt to explain the differences in the Ames II assay responses between the two
samplings (Figure 4), 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-
16
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
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 6. 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 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 1) suggested
that 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 pre-AOP (which comprise a major fraction of the
humic content51, 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 by the increase in
LMW acids evident from the LC-OCD fractionation (see Figure 6), which were found to be
statistically significant pre- and post-AOP for both samplings. Nevertheless, these results
only suggest a link between the fate of NOM in the water and the Ames II assay responses
obtained, without proving a direct cause and effect relationship.
MP vs LP-based treatment
The findings from the work of Martijn52 were compared to ours, since the procedure followed
for the Ames II assay was identical in both studies and the concentration factor (cf) of the
samples from the SPE extraction before the Ames II assay (i.e. 20,000 cf) was also the same
17
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
for both studies. Martijn52 measured the Ames II assay response of synthetic water with Pony
Lake NOM (2.5 mg C/L) after MP-UV/H2O2 treatment (600 mJ/cm2 and 6 mg/L H2O2) and
reported a yield of 25 positive wells (Figure 7.252) in the presence of nitrate (12 mg/L); in our
case where Pony Lake NOM (2.1 mg C/L) 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 assay responses obtained in this work were
much lower than the ones obtained by Martijn52 the Ames II assay response in nitrate-rich
water was still significant.
A distinction between significant and positive results (in terms of testing by the Ames assay)
should be kept in mind here; a detectable response is not necessarily considered significant.
For example, an organic compound in levels high enough to be detected may be below the
quantitation limit, therefore it can be considered not significant, and no sound conclusions
can be drawn from this finding. In the example of the Ames responses, the levels obtained
were detectable and statistically significant from one another in many cases; nevertheless,
without reaching the mutagenic level defined by the Ames protocol, the responses are not
considered positive in that respect.
Illustrating the significant role of nitrate for both MP and LP-UV processes, the Ames II
assay response as a function of nitrite formation was shown by Martijn52 (Figure 5.3).
Plotting our nitrite data (0.08-0.09 mg NO3-/L) obtained from the synthetic Pony Lake NOM
in Figure 7, gave an Ames II assay response of 5.5 positive wells, which is within the range
of the Ames II assay response we obtained for Pony Lake NOM for the 1500 mJ/cm 2 (8 ±
3.46 positive wells) (Figure 1b). 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
18
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
observed supports the earlier indication of the impact of UV photolysis of nitrate on both the
nitrite formation and the Ames II assay response (Martijn et al. 2014)17, and is a confirmation
of the clear relation between the two (nitrite and Ames II assay 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.53 performed a
preliminary risk assessment 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) approach54. They found that the 4-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
our findings that the LP-UV/H2O2 treatment gave little, if any, reason for concern from a
health-related aspect.
Conclusions
From the current work, it can be concluded that by LP-UV/H2O2 treatment of nitrate-rich
water, nitrite and mono-nitrophenol formation may be observed. The principal two mono-
nitrophenols, 2- and 4-nitrophenol, reported in previous research, were also detected in this
work. These nitrophenols are known to have higher toxicity to organisms than the parent
compound, phenol, 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, respectively55. However, under the conditions applied in this
research study, the concentrations of these phenols were in the range of trace levels and
would not be expected to cause health-related issues. Moreover, previous studies of 2- and 4-
nitrophenol have shown that no mutagenicity was observed when different variations of the
Ames II assay were applied56.
19
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
These nitrated organics are 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 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 correspondingly enhanced,
compared to the case of 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 are lower in LP-UV applications, giving little reason for concern.
From the present work, it can be concluded that LP-UV/H2O2 treatment is not expected to
produce significant mutagenic activity as shown by the applied Ames II assay results
(although other mutagenic mechanisms cannot be excluded), even when high nitrate
concentrations are present and high UV fluences and 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 our case) and the
conjunctive effect of all factors contributing to health effects is usually complex.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
This work was performed in the cooperation framework of Wetsus, European Centre of
Excellence for Sustainable Water Technology (www.wetsus.eu). Wetsus is co-funded by the
Dutch Ministry of Economic Affairs and Ministry of Infrastructure and Environment, the
Province of Fryslân and the Northern Netherlands Provinces. The authors wish to thank the
participants of the ‘Priority Compounds’ research theme for the fruitful discussions, and
20
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
especially Dr. Bram Martijn for his guidance and advice in the experimental design and
relating the findings of the two studies. The authors would also like to thank Trojan
Technologies Inc. for supplying the low-pressure UV lamp and VITO laboratories for
performing the Ames II assays. The authors also gratefully acknowledge the financial support
of Anglian Water Services Ltd.
References
1 S. Vilhunen and M. Sillanpää, Rev Environ Sci Biotechnol, 2010, 9, 323.2 W. Yang, H. Zhou and N. Cicek, Crit. Rev. Environ. Sci. Technol., 2014, 44, 1443.3 S. D. Richardson, M. J. Plewa, E. D. Wagner, R. Schoeny and D.M. DeMarini, Mutat.
Res., 2007, 636, 178.4 E.J. Rosenfeldt and K.G. Linden, Environ. Sci. Technol., 2004, 38, 5476. 5 F.J. Beltrán, G. Ovejero and J. Rivas, Ind. Eng. Chem. Res., 1996, 35, 883. 6 S. Ledakowicz, J.S. Miller and D. Olejnik, Int. J. Photoenergy, 1999, 1, 1. 7 K. Ikehata, N.J. Naghashkar and M.G. El-Din, Ozone Sci. Eng., 2006, 27, 83. 8 H.D. Burrows, L.M. Canle, J.A. Santaballa and S. Steenken, J. Photochem. Photobiol.
B Biol., 2002, 67, 71. 9 E. Kowalska, M. Janczarek, J. Hupka and M. Grynkiewicz, Water Sci. Technol., 2004,
49, 261.10 U. von Gunten and Y. Oliveras, Environ. Sci. Technol., 1998, 32, 63. 11 T. Bond, M.R. Templeton and N. Graham, J. Hazard. Mater., 2012, 235−236, 1.12 J. Mack and J.R. Bolton, J. Photochem. Photobiol., A, 1999, 128, 1.13 S. Goldstein and J. Rabani, J. Am. Chem. Soc., 2007, 129, 10597.14 K.A. Thorn and L.G. Cox, J. Environ. Qual., 2012, 41, 865-881.15 J. Dzengel, J. Theurich and D.W. Bahnemann. Formation of Nitroaromatic Compounds
in Advanced Oxidation Processes: Photolysis versus Photocatalysis. Environ. Sci. Technol., 1998, 33, 294.
16 D. Vione, V. Maurino, C. Minero, M. Lucchiari and E. Pelizzetti, Chemosphere, 2004, 56, 1049.
17 A.J. Martijn, M.G. Boersma, J.M. Vervoort, I.M.C.M. Rietjens and J.C. Kruithof, Desalin. Water Treat., 2014, 52, 6275–6281.
18 A.J. Martijn, J.C. Kruithof, R.A.M. Hughes, R.A. Mastan, A.R. Van Rompay and J.P. Malley Jr., J. - Am. Water Works Assoc., 2015, 107, 301.
19 F. Machado and P. Boule, J. Photochem. Photobiol., A, 1995, 86, 73.20 Kolkman, B.J. Martijn, D. Vughs, K.A. Baken and A.P. van Wezel, Environ. Sci.
Technol., 2015, 49, 4458.
21
495
496
497
498
499
500
501
502
503
504505506507508509510511512513514515516517518519520521522523524525526527528529530531532
21 H.S. Rosenkranz and R. Mermelstein, J. Environ. Sci. Health, Part C: Environ. Carcinog. Rev., 1985, 3, 221.
22 B.N. Ames, E.G. Gurney, J.A. Miller and H. Bartsch, Proceedings of the National Academy of Sciences of the United States of America, 1972, 69, 3128.
23 National Toxicology Program, http://ntp.niehs.nih.gov/testing/types/genetic/invitro/sa/index.html , (accessed September 2017).
24 S. Flückiger-Isler, M. Baumeister, K. Braun, V. Gervais, N. Hasler-Nguyen, R. Reimann, J. Van Gompel, H.G. Wunderlich and G. Engelhardt, Mutat. Res., Genet. Toxicol. Environ. Mutagen., 2004, 558, 181.
25 F. van Hoof, J.G. Janssens and H. van Dijck, Chemosphere, 1985, 14, 501.26 M.A. van der Gaag, J.C. Kruithof and L.M. Puijker, Sci. Total Environ., 1986, 47, 137.27 M.B. Heringa, D.J.H. Harmsen, E.F. Beerendonk, A.A. Reus, C.A.M. Krul, D.H. Metz
and G.F. IJpelaar, Water Res., 2011, 45, 366.28 A.J. Martijn and J.C. Kruithof, Ozone Sci. Eng., 2012, 34, 92. 29 C.F. van Kreijl, H.J. Kool, M. de Vries, H.J. van Kranen and E. de Greef, Sci. Total
Environ., 1980, 15, 137.30 R.F. Christman, K. Kronberg, R. Singh, L.M. Ball and J.D. Johnson. Identification of
mutagenic by-products from aquatic humic chlorination. Department of Environmental Sciences and Engineering University of North Carolina, 1991.
31 X. Lv, Y. Lu, X. Yang, X. Dong, K. Ma, S. Xiao, Y. Wang and F. Tang, Sci. Rep., 2015, 5, 1.
32 T. Manasfi, M. De Méo, B. Coulomb, C. Di Giorgio and J.-L. Boudenne, presented in part at the 6th International Conference on Swimming Pool and Spa, Amsterdam, March, 2015.
33 R.C. Hofman-Caris, D.J. Harmsen, L. Puijker, K.A. Baken, B.A. Wols, E.F. Beerendonk and L.L. Keltjens, Water Res., 2015, 74, 191.
34 E.J.M. Penders, A.J. Martijn, A. Spenkelink, G.M. Alink, I. Rietjens and W. Hoogenboezem, J. Water Supply: Res. Technol.--AQUA, 2012, 61, 435.
35 G.F. IJpelaar, A.J. van der Veer, G.J. Medema and J.C. Kruithof, J. Environ. Eng. Sci., 2005, 4, S51.
36 G.V. Korshin, C-W Li and M.M. Benjamin, Water Res., 1997, 31, 1787.37 S. Semitsoglou-Tsiapou, M.R. Templeton, N.J.D. Graham, L. Hernández Leal, B. J.
Martijn, A. Royce, J.C. Kruithof, Water Res., 2016(a), 91, 285.38 J.R. Bolton and K.G. Linden, J. Environ. Eng., 2003, 129, 209.39 N.V. Klassen, D. Marchington and H.C.E. McGowan, Anal. Chem., 1994, 66, 2921.40 S. Flückiger-Isler, M. Baumeister, K. Braun, V. Gervais, N. Hasler-Nguyen, R.
Reimann, J. Van Gompel, H.-G. Wunderlich, G. Engelhardt, Mutat. Res., 2012, 558, 181.
41 D. Minakata, K. Li, P. Westerhoff and J. Crittenden, Environ. Sci. Technol., 2009, 43, 6220.
42 G. McKay, J.L. Kleinman, K.M. Johnston, M.M. Dong. F.L. Rosario-Ortiz and S.P. Mezyk, J Soils Sediments, 2014, 14, 298.
43 Y. Ahn, D. Lee, M. Kwon, I. Choi, S-N. Nam and J-W. Kang, Chemosphere, 2017, 184, 960.
22
533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577
44 J. E. Donham, E. J. Rosenfeldt and K. R. Wigginton, Environ Sci Process Impacts, 2014, 16, 764.
45 K.A. Thorn and L.G. Cox, PLoS ONE, 2016, 11, 1/20. 46 S. Semitsoglou-Tsiapou, A. Mous, M.R. Templeton, N.J.D. Graham, L. Hernández Leal
and J.C. Kruithof, Water Res., 2016(b), 106, 312.47 N. Lu, N.Y. Gao, Y. Deng and Q.S. Li, Water Sci. Technol., 2009, 60, 1393. 48 C.M. Sharpless, M.A. Page and K.G. Linden, Water Res., 2003, 37, 4730. 49 P. Jin, X. Jin, X. Wang, Y. Feng and X.C. Wang, in Biomass Now - Cultivation and
Utilization, ed. Dr. Miodrag Darko Matovic, InTech, 2013, ch. 7, pp. 153-192. 50 A.K. Camper, M.W. Lechevallier, S.C. Broadaway and G.A. Mcfeters, Appl. Environ.
Microbiol., 1985, 50, 1378–1382.51 J. Chen, B. Gu, E.J. LeBoeuf, H. Pan and S. Dai, Chemosphere, 2002, 48, 59. 52 A.J. Martijn. Ph.D. Thesis, Wageningen University, the Netherlands, 2015.53 B.J. Martijn, A.R. Van Rompay, E.J.M. Penders, Y. Alharbi, P.K. Baggelaar, J.C.
Kruithof and I.M.C.M. Rietjens, Chemosphere, 2016, 144, 338. 54 EFSA. Guidance of the Scientific Committee on a request from EFSA on the use of the
benchmark dose approach in risk assessment, The EFSA Journal, 2009, 1150, 1.55 WHO. Concise International Chemical Assessment Document 20, Mononitrophenols,
World Health Organization, Geneva, 2000. 56 USEPA. Health Advisories for Drinking Water Contaminants. United States
Environmental Protection Agency. Office of Water Health Advisories. Lewis Publishers, USA, 1993.
Table 1. Water quality and operational parameters for the two WTW sampling events. * denotes ‘’not calculated’’, since the DOC value was below the detection limit (1 mg/L); UVT: Ultraviolet Transmission
Figure 1. Number of positive wells generated by the Ames II assay (TA98 – S9) under different experimental conditions with a) Suwannee River, and
23
Sampling (a)
Sampling (b)
Temperature (ₒC) 5.0 10pH (-) 8.2 8.5
Works inlet DOC (mg/L) 4.05 4.71Pre-AOP DOC (mg/L) 2.68 2.11Post-AOP DOC (mg/L) 1.52 <1.00Inlet SUVA (L/m mg) 4.61 3.74
Pre-AOP SUVA (L/m mg) 4.70 4.18Post-AOP SUVA (L/m mg) 2.90 *
Inlet UVT (%) 65.0 66.7Pre-AOP UVT (%) 74.8 81.7Post-AOP UVT (%) 90.4 92.7UV fluence (mJ/cm2) 2000 1750
Hydrogen peroxide dose (mg/L) 40 40
Works inlet nitrate (mg/L) 30.5 31.7Works effluent nitrate (mg/L) 30.0 30.2
578579580581582583584585586587588589590591592593594595596597598599
600
601
602
603
604605606
607
608
609
610611 612
613614615616617618619620
b) Pony Lake NOM (H2O2 concentration 15 mg/L). The numbers 0, 1500 and 2000 represent the UV fluences applied, in mJ/cm2. The positive control (2-nitrofluorene + 4-nitroquinoline N-oxide) produced 46.9 (±0.782) positive wells. The error bars represent the standard deviation values for the mean number of positive wells, where the mean number of positive wells represents the average of the positive wells for 9 replicates in the control conditions and 3 replicates for each extract.
Figure 2. Number-fold increases over a) solvent control, and b) baseline, calculated from the Ames II assay (TA98 – S9) values for the different experimental conditions with either Pony Lake or Suwannee River NOM (H2O2 concentration 15 mg/L). The numbers 0, 1500 and 2000 represent the UV fluences applied, in mJ/cm2. The error bars represent the standard deviation values for the mean number of positive wells, where the mean number of positive wells represents the average of the positive wells for 9 replicates in the control conditions and 3 replicates for each extract.
24
621622623624625626
627628
629630631632633634635636
637
Figure 3. Concentrations of mono-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 bar representing the value for the 0/15 UV/H2O2
combination is shaded because the nitrophenol concentration is higher than the detection limit of 0.007 μg/L but lower than the quantification limit of 0.02 μg/L. The numbers 0, 1500 and 2000 represent the UV fluences applied, in mJ/cm2, while the 15 represents the H2O2 dosage in mg/L.
Figure 4. Number of positive wells generated by the Ames II assay (TA98 – S9) 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 error bars represent the standard deviation values for the mean number of positive wells, where the mean number of positive wells represents the average of the positive wells for 9 replicates in the control conditions and 3 replicates for each extract.
25
638639640641642643644645
646
647
648649650651652653654
655
Figure 5. Number-fold increases over a) solvent control, and b) baseline, calculated from the Ames II assay (TA98 – S9) values for the full-scale water samples during different stages of treatment for both samplings (a) and (b). The error bars represent the standard deviation values for the mean number of positive wells, where the mean number of positive wells represents the average of the positive wells for 9 replicates in the control conditions and 3 replicates for each extract.
26
656
657658659660661662663
664
Figure 6. LC-OCD fractionation of full-scale water samples: a) sampling (a), and b) sampling (b) (C: Carbon, N: Nitrogen, LMW: Low Molecular Weight).
Figure 7. Ames II 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 LP-UV/H2O2 results superimposed on adapted Figure 5.3 of Martijn, 2015)52.
27
665666667
668
669
670671672673674