This is an open access article published under an ACS AuthorChoice License, which permits

copying and redistribution of the article or any adaptations for non-commercial purposes.

Letter

pubs.acs.org/journal/estlcu

Occurrence of Neonicotinoid Insecticides in Finished Drinking Water

and Fate during Drinking Water Treatment
Kathryn L. Klarich,†,‡ Nicholas C. Pflug,†,‡,§ Eden M. DeWald,† Michelle L. Hladik,∥ Dana W. Kolpin,⊥
David M. Cwiertny,*,†,‡ and Gregory H. LeFevre*,†,‡
†
Department of Civil & Environmental Engineering, University of Iowa, Iowa City, Iowa 52242, United States
‡
IIHR-Hydroscience and Engineering, University of Iowa, Iowa City, Iowa 52242, United States
§
Department of Chemistry, University of Iowa, Iowa City, Iowa 52242, United States
∥
California Water Science Center, U.S. Geological Survey, 6000 J Street, Placer Hall, Sacramento, California 95819, United States
⊥
Illinois-Iowa Water Science Center, U.S. Geological Survey, 400 South Clinton Street, Iowa City, Iowa 52240, United States
*
S Supporting Information

ABSTRACT: Neonicotinoid insecticides are widespread in sur-

face waters across the agriculturally intensive Midwestern United
States. We report for the first time the presence of three
neonicotinoids in finished drinking water and demonstrate their
general persistence during conventional water treatment. Periodic
tap water grab samples were collected at the University of Iowa
over 7 weeks in 2016 (May−July) after maize/soy planting.
Clothianidin, imidacloprid, and thiamethoxam were ubiquitously
detected in finished water samples at concentrations ranging from
0.24 to 57.3 ng/L. Samples collected along the University of Iowa
treatment train indicate no apparent removal of clothianidin or
imidacloprid, with modest thiamethoxam removal (∼50%). In contrast, the concentrations of all neonicotinoids were
substantially lower in the Iowa City treatment facility finished water using granular activated carbon (GAC) filtration. Batch
experiments investigated potential losses. Thiamethoxam losses are due to base-catalyzed hydrolysis under high-pH conditions
during lime softening. GAC rapidly and nearly completely removed all three neonicotinoids. Clothianidin is susceptible to
reaction with free chlorine and may undergo at least partial transformation during chlorination. Our work provides new insights
into the persistence of neonicotinoids and their potential for transformation during water treatment and distribution, while also
identifying GAC as a potentially effective management tool for decreasing neonicotinoid concentrations in finished drinking
water.

■ INTRODUCTION
Neonicotinoid pesticides have become the most widely used
High use and chemical properties have resulted in
proliferation of neonicotinoids in surface waters.20−23 In a
nationwide study of streams in the United States, at least one
insecticides in the world.1,2 Neonicotinoids are systemic, insect-
neonicotinoid compound was detected in 63% of the 48
targeting,3−5 potent neurotoxins that are often applied as seed
streams measured.21 Neonicotinoids were ubiquitously de-
treatments to crops in the United States and in urban pest tected at all streams sampled that drain intensively row-cropped
control applications. 1,6 Neonicotinoids have also been areas of the Midwestern United States,20 with maximal
implicated in a variety of ecosystem effects,7 including declines concentrations of 260, 43, and 190 ng/L for clothianidin,
in populations of pollinators8,9 (e.g., honeybees) and effects on imidacloprid, and thiamethoxam, respectively, which represent
nontarget organisms.10−15 They are substantially more toxic to the most widely used and commonly observed compounds in
insects than vertebrates;6 however, most vertebrate toxicity this class of insecticides. Neonicotinoids are water-soluble6
research has focused on acute exposure, and chronic exposure (340, 610, and 4100 mg/L for clothianidin, imidacloprid, and
remains a concern.13 Several studies report associations thiamethoxam, respectively) and polar20 (log Kow = 0.91, 0.57,
between chronic exposure to neonicotinoids and adverse and −0.13 for clothianidin, imidacloprid, and thiamethoxam,
developmental or neurological outcomes.16 Other studies respectively). Research to date suggests general neonicotinoid
highlight potential concerns, including inflammation of the
liver and central nervous system due to chronic exposure to Received: March 8, 2017
neonicotinoids,17 loss of insect selectivity in transformation Revised: March 21, 2017
products,4,18,19 and negative effects on nontarget species in Accepted: March 21, 2017
aquatic ecosystems.10

© XXXX American Chemical Society A DOI: 10.1021/acs.estlett.7b00081

Environ. Sci. Technol. Lett. XXXX, XXX, XXX−XXX
Environmental Science & Technology Letters Letter

Figure 1. Concentrations of clothianidin, imidacloprid, and thiamethoxam in samples collected from University of Iowa tap water in 2016.
Concurrent streamflow in the Iowa River in Iowa City, IA, is shown. Iowa River flow is regulated by a reservoir, generating the hydrograph pictured.

persistence in the environment24 (e.g., imidacloprid and the University of Iowa. Samples of the City drinking water were
clothianidin were documented to have conservative transport collected from three residential taps at separate locations in
through a study stream reach21), although photolysis can occur Iowa City. To assess neonicotinoid fate during treatment, the
to various extents among the different neonicotinoids.19,25 raw source water, sedimentation basin effluent, recarbonation
On the basis of limited data, neonicotinoids appear to be effluent (prechlorination), recarbonation effluent (postchlori-
poorly removed via treatment systems, with insignificant or nation), filtration effluent, and finished water were sampled at
very marginal removal observed during conventional waste- the UI DWTP, and the source and finished water were sampled
water treatment and no removal in a constructed treatment at the City DWTP (Figure S1). Water samples were enriched
wetland.26,27 To date, no known research has examined the via solid phase extraction (SPE), analyzed using liquid
presence of neonicotinoids in finished drinking water, chromatography with tandem mass spectrometry (LC−MS/
particularly for communities relying on agriculturally impacted MS), and quantified according to established U.S. Geological
surface water sources. Here, we present results of field analyses Survey methods.30 Fate during unit processes was tested in
and laboratory experiments measuring the fate of neonicoti- laboratory batch systems using free chlorine, GAC, and pH
noids during drinking water treatment. Our objectives were (1) adjustment, with neonicotinoid concentrations measured by LC
to quantify neonicotinoid residues in two public drinking water with a diode array detector and mass spectrometry (LC−DAD/
facilities that derive their water from agriculturally impacted MS). Field and laboratory QA/QC samples were analyzed
sources and (2) to determine the efficacy of drinking water throughout the study (described in the Supporting Informa-
treatment operations to remove neonicotinoids. tion). Experimental details and analytical methods are provided

■
in the Supporting Information.
MATERIALS AND METHODS
Between May and July 2016 following maize/soy planting,
finished drinking water samples were collected from taps at the
■ RESULTS AND DISCUSSION
Occurrence of Neonicotinoids in Drinking Water.
University of Iowa and at three locations in Iowa City, IA. The Clothianidin, imidacloprid, and thiamethoxam were ubiqui-
University of Iowa drinking water treatment plant (UI DWTP) tously present (i.e., 100%) in all samples (n = 16) collected
serves the University of Iowa (UI), while the Iowa City water from UI tap water, with concentrations ranging between 3.89
treatment plant (City DWTP) serves Iowa City (City). The UI and 57.3 ng/L, between 1.22 and 39.5 ng/L, and between 0.24
DWTP (Figure S1) uses the Iowa River for source water and and 4.15 ng/L, respectively (Table S4). Maximal concen-
uses screening, chemical pretreatment, sedimentation, lime trations of clothianidin and imidacloprid occurred a few days
softening, recarbonation, chlorination, and sand filtration for after peak flow in the Iowa River (Figure 1), indicating a
treatment. The City DWTP (Figure S1) uses water from possible relationship between neonicotinoid concentration and
alluvial wells fed by the Iowa River (i.e., groundwater influenced river flow. The delay between maximal river flow and maximal
by surface water) and provides treatment via aeration, lime tap water concentration may be due to the residence time in
softening, recarbonation, granular activated carbon (GAC) the distribution system, which is typically <1−3 days but can be
filtration, and chlorination. The Iowa River drains a watershed in some locations up to 6 days.31 Samples of City finished tap
that is 8150 km2 in a heavily row-cropped agroecosystem,28,29 water collected at private residences (Table S5) contained up to
where prior work has demonstrated frequent detection of 0.52 ng/L thiamethoxam; however, clothianidin and imidaclo-
neonicotinoid pesticides.20 The river flow is composed of prid were not present above detection limits.
overland flow and tile drainage (from rainfall, no snowmelt The concentrations of clothianidin, imidacloprid, and
during the study period) and groundwater. The City alluvial thiamethoxam measured in UI tap water are consistent with
wells and UI DWTP intakes are located approximately 10 and documented environmental concentrations.20,30−32 In a nation-
15 km downstream of the Coralville reservoir, respectively. wide study, at least one neonicotinoid was detected in 63% of
University drinking water samples were collected periodically the 48 streams monitored.21 Similarly, in a study of streams in
from a tap in the laboratory located in the Seamans Center at Iowa, at least one neonicotinoid compound was detected in all
B DOI: 10.1021/acs.estlett.7b00081
Environ. Sci. Technol. Lett. XXXX, XXX, XXX−XXX
Environmental Science & Technology Letters Letter

samples.20 These detections include clothianidin (3.5−79 ng/ thiamethoxam concentrations exhibited a clear drop of ∼40−
L), imidacloprid (not detected to 15 ng/L), and thiamethoxam 60% after lime softening and recarbonation but were essentially
(not detected to 43 ng/L) as measured in the Iowa River in stable thereafter.
Wapello, IA (approximately 45 miles downstream of Iowa We also collected samples from the City and UI DWTP to
City),20 and imidacloprid measured in Old Man’s Creek near compare source water and finished water concentrations of
Iowa City (4.5−35 ng/L).20 In other studies, imidacloprid was clothianidin, imidacloprid, and thiamethoxam (Figure 3 and
measured in a stream (3.4−10 ng/L) in Georgia,30 as well as in
other small streams32 throughout the Midwest (not detected to
2900 ng/L; measured via grab and passive sampling).
Fate of Neonicotinoids during Drinking Water Treat-
ment. Samples collected from the UI DWTP (Figure 2)

Figure 3. Concentrations of the three neonicotinoids measured in City

and UI DWTP source and finished drinking waters (August 9, 2016).
The City DWTP uses granular activated carbon (GAC) filtration
compared to rapid sand filtration at the UI DWTP. The asterisk
indicates no detection. Error bars represent the standard error of
regression associated with the composite enrichment sample extraction
and analysis (1 L enriched to 1 mL).

Tables S9−S11) between the two treatment plants. Samples

from the UI DWTP were collected within 3 h of City DWTP
samples. Source water concentrations of the three compounds
were within 30% between sites for a given compound, despite
the fact that UI DWTP water originates from the Iowa River
and the City DWTP water originates from the shallow alluvial
aquifer under the influence of the Iowa River.
Decreases in neonicotinoid concentrations appeared to be
greater at the City DWTP (∼100, 94, and 85% for clothianidin,
imidacloprid, and thiamethoxam, respectively) than at the UI
DWTP (∼1, 8, and 44%, respectively). A notable distinction is
that the City DWTP uses GAC filtration compared to rapid
sand filtration at the UI DWTP; the latter process removes only
Figure 2. Concentrations of clothianidin, imidacloprid, and particles. These analyses were consistent with earlier UI DWTP
thiamethoxam measured at different unit operations at the UI process train results that indicated no discernible changes in
DWTP on the two indicated sampling dates (additional data in concentration for clothianidin or imidacloprid and a modest
Tables S6−S8). Neonicotinoid concentrations differed on the two loss of thiamethoxam. Additionally, finished water concen-
sampling dates, but overall trends across the treatment train were trations of clothianidin, imidacloprid, and thiamethoxam from
consistent. Error bars represent the standard error of regression each treatment plant were similar to the corresponding
associated with the composite enrichment sample extraction and measurements from tap water samples.
analysis (1 L enriched to 1 mL). Hydrolysis of Thiamethoxam. We attribute thiamethox-
am removal to base-catalyzed hydrolysis. Base-catalyzed
suggest that clothianidin and imidacloprid persist throughout hydrolysis of thiamethoxam has been reported with half-lives
conventional water treatment processes, while thiamethoxam is (t1/2 values) ranging from 2.1 days33 at pH 9.2 and 28 °C
partially removed. Neonicotinoid concentrations on the two (corresponding to a pseudo-first-order rate constant, kobs, value
different sampling dates (Figure 2) varied, but trends across the of 0.33 day−1) to 6.1 days34 at pH 9.0 and 25 °C (kobs = 0.11
treatment train were consistent. Raw source water (i.e., Iowa day−1). Furthermore, the stability of thiamethoxam is known to
River) concentrations ranged from 10.7 to 25.9 ng/L for decrease with increasingly alkaline conditions.19,33,35
clothianidin, from 2.15 to 13.3 ng/L for imidacloprid, and from Batch tests confirmed that thiamethoxam hydrolysis is likely
1.93 to 8.23 ng/L for thiamethoxam, whereas finished water to occur over time scales relevant to treatment and distribution
concentrations ranged from 10.6 to 31.2 ng/L for clothianidin, (Figures S2−S4 and Table S1). Using a UI DWTP softening
from 1.97 to 13.6 ng/L for imidacloprid, and from 1.07 to 3.11 basin water sample spiked with 100 μM thiamethoxam, we
ng/L for thiamethoxam. Although we did not attempt to follow measured a t1/2 of 0.75 day (kobs = 0.9 day−1) at pH 10.4 (the
a single parcel of water through the treatment process (i.e., all softening basin pH) and 20 °C. During the lime softening
samples were collected at approximately the same time in a process at the UI DWTP, the pH is increased to ≥10.3 with a
given sampling round), little to no concentration change for residence time of 1.5−3.2 h. Accordingly, thiamethoxam
clothianidin and imidacloprid was measured. In contrast, removal observed in Figures 2 and 3 reflects degradation
C DOI: 10.1021/acs.estlett.7b00081
Environ. Sci. Technol. Lett. XXXX, XXX, XXX−XXX
Environmental Science & Technology Letters Letter

Figure 4. Neonicotinoid batch kinetic tests. The left panel shows the change in aqueous neonicotinoid concentration (C0 = 100 μg/L) in
suspensions of granular activated carbon (5 g/L GAC in pH 7 phosphate buffer). Data fitted to an exponential decay model (Table S12 and Figure
S5). The right panel shows chlorination loss kinetics. Cl2/neonicotinoid values reported as molar ratio (M/M). Titrations with FAS revealed
chlorine concentrations (10, 50, and 100 mg/L as Cl2) that were constant during the experiment, allowing calculation of kobs from the slopes of linear
regressions.

from hydrolysis during treatment and distribution (finished S7)] at rates greater than those expected from estimated k2
water pH of ∼9.9), as well as during the handling time between values. We suspect that differences in the clothianidin
sample collection and processing (typically 24 h). Thiamethox- transformation rate across a range of chlorine concentrations
am hydrolysis is also expected to occur in the City DWTP, reflect the formation of highly reactive intermediates that
which also employs lime softening (finished water pH of ∼9.2). contribute to chlorine demand, which in turn influences the
Removal of Neonicotinoids via Sorption onto Gran- extent of clothianidin degradation (Figure S9).
ular Activated Carbon. All three neonicotinoids studied Environmental Implications. To the best of our knowl-
exhibited relatively rapid removal via sorption onto GAC, with edge, this is the first peer-reviewed study to document the
>80% removal in suspensions after 1 h of contact time (Figure presence of neonicotinoids in finished tap water samples.
4). Initial sorption was rapid, followed by stabilized aqueous Conventional water treatment results in no measurable removal
concentrations consistent with equilibrium by 30 min. Some of clothianidin or imidacloprid, although the alkaline conditions
heterocyclic aromatic nitrogen compounds and protonated of lime softening result in the partial transformation of
bases, such as the neonicotinoids studied herein, have been thiamethoxam via base-catalyzed hydrolysis. Because of their
reported36 to exhibit greater removal by GAC than would be pervasiveness in source waters20,21,30,32 and persistence through
predicted by Kow values alone. Neonicotinoid removal by GAC treatment systems,27 neonicotinoids are likely present in other
is likely attributable to specific binding interactions between drinking water systems across the United States. Trans-
surface sites on GAC and specific structural moieties in the formation products formed by chlorination or hydrolysis
neonicotinoids, although additional experimental studies are warrant great consideration because of the potential to form
recommended to evaluate adsorption mechanisms, long-term toxic transformation products (Figures S3 and S10). For
effectiveness, optimal dosing, and overflow rates. example, the metabolite desnitro-imidacloprid exhibits a
Transformation of Neonicotinoids during Chemical mammalian receptor binding affinity 300 times greater than
Disinfection with Free Chlorine. Both treatment plants that of imidacloprid because of the loss of the nitro group that
employ chlorination, with typical contact times of 3−4 h (City confers insect specificity.4 For management, GAC filtration
DWTP) and 20 min to 3 h (UI DWTP), and with residuals of presents a treatment option for removal of neonicotinoids in
1.8 mg/L Cl2 (City DWTP) and 2.5 mg/L Cl2 (UI DWTP). resource-constrained communities that rely of agriculturally
Laboratory batch studies revealed a range of reactivities of impacted surface waters or point-of-use systems that is
neonicotinoids toward free chlorine [HOCl (Figure 4)]. substantially more economical than reverse osmosis or
Thiamethoxam was generally recalcitrant, exhibiting no advanced oxidation processes.38
significant loss (p > 0.50) at even the greatest free chlorine
concentrations tested (Cl2:thiamethoxam molar ratio of 12500)
over a prolonged reaction time. In contrast, imidacloprid and
■ ASSOCIATED CONTENT
* Supporting Information
S
clothianidin exhibited greater reactivity, with clothianidin being
The Supporting Information is available free of charge on the
most reactive. Second-order rate coefficients for the reaction of
ACS Publications website at DOI: 10.1021/acs.estlett.7b00081.
HOCl with clothianidin (4.7 × 10−2 M−1 s−1) and imidacloprid
(1.6 × 10−3 M−1 s−1) were calculated from measured pseudo- Additional method details, statistical analysis, quality
first-order rate constants (Figure 4) assuming a constant HOCl assurance/control, and additional detailed data, results,
concentration (k2 = kobs/[HOCl]). At chlorine concentrations and analysis in figures and tables (PDF)

■
more typical for disinfection (i.e., 5 mg/L as Cl2) and assuming
a constant residual, half-lives for clothianidin and imidacloprid
AUTHOR INFORMATION
would be ∼2.5 and ∼70 days, respectively. Although
imidacloprid is practically resistant to transformation, a modest Corresponding Authors
degree of clothianidin decay may be expected during chemical *Address: 4105 Seamans Center, University of Iowa, Iowa City,
disinfection, particularly in distribution systems with longer IA 52242. E-mail: gregory-lefevre@uiowa.edu. Phone: 319-335-
residence times.37 We note that using conditions more 5655.
representative of drinking water treatment (C0 = 5 mg/L *Address: 4105 Seamans Center, University of Iowa, Iowa City,
HOCl as Cl2; 0.10−1.25 mg/L clothianidin), extensive IA 52242. E-mail: david-cwiertny@uiowa.edu. Phone: 319-335-
transformation of clothianidin occurred [>80% in 1.5 h (Figure 1401.
D DOI: 10.1021/acs.estlett.7b00081
Environ. Sci. Technol. Lett. XXXX, XXX, XXX−XXX
Environmental Science & Technology Letters Letter

ORCID associated plant defense responses. Proc. Natl. Acad. Sci. U. S. A.

Michelle L. Hladik: 0000-0002-0891-2712 2010, 107, 17527−17532.
(15) Ford, K. A.; Casida, J. E. Comparative metabolism and
Gregory H. LeFevre: 0000-0002-7746-0297
pharmacokinetics of seven neonicotinoid insecticides in spinach. J.
Notes Agric. Food Chem. 2008, 56, 10168−10175.
The authors declare no competing financial interest. (16) Cimino, A. M.; Boyles, A. L.; Thayer, K. A.; Perry, M. J. Effects

■
of Neonicotinoid Pesticide Exposure on Human Health: A Systematic
ACKNOWLEDGMENTS Review. Environ. Health Perspect. 2017.12515516210.1289/EHP515
(17) Duzguner, V.; Erdogan, S. Chronic exposure to imidacloprid
We acknowledge funding from the University of Iowa Center induces inflammation and oxidative stress in the liver & central
for Health Effects of Environmental Contamination (Grant nervous system of rats. Pestic. Biochem. Physiol. 2012, 104, 58−64.
18018213 BR05). Scott Slee at the UI DWTP and Jonathan (18) Tomizawa, M. Neonicotinoids and Derivatives: Effects in
Durst at City DWTP provided plant access for sampling. K.L.K. Mammalian Cells and Mice. J. Pestic. Sci. 2004, 29, 177−183.
was supported by a National Science Foundation Graduate (19) de Voogt, P., Ed. Reviews of Environmental Contamination and
Research Fellowship. U.S. Geological Survey (USGS) con- Toxicology; Springer: Cham, Switzerland, 2017; Vol. 242.
tributions were provided by the USGS Toxic Substances (20) Hladik, M. L.; Kolpin, D. W.; Kuivila, K. M. Widespread
Hydrology Program. Any use of trade, product, or firm names is occurrence of neonicotinoid insecticides in streams in a high corn and
for descriptive purposes only and does not imply endorsement soybean producing region, USA. Environ. Pollut. 2014, 193, 189−196.
by the U.S. Government. (21) Hladik, M. L.; Kolpin, D. W. First national-scale reconnaissance

■
of neonicotinoid insecticides in streams across the USA. Environ.
Chem. 2016, 13, 12−20.
REFERENCES (22) Main, A. R.; Michel, N. L.; Headley, J. V.; Peru, K. M.;
(1) Simon-Delso, N.; Amaral-Rogers, V.; Belzunces, L. P.; Bonmatin, Morrissey, C. A. Ecological and Landscape Drivers of Neonicotinoid
J. M.; Chagnon, M.; Downs, C.; Furlan, L.; Gibbons, D. W.; Giorio, Insecticide Detections and Concentrations in Canada’s Prairie
C.; Girolami, V.; et al. Systemic insecticides (neonicotinoids and Wetlands. Environ. Sci. Technol. 2015, 49, 8367−8376.
fipronil): trends, uses, mode of action and metabolites. Environ. Sci. (23) Main, A. R.; Headley, J. V.; Peru, K. M.; Michel, N. L.; Cessna,
Pollut. Res. 2015, 22, 5−34. A. J.; Morrissey, C. A. Widespread Use and Frequent Detection of
(2) Jeschke, P.; Nauen, R.; Schindler, M.; Elbert, A. Overview of the Neonicotinoid Insecticides in Wetlands of Canada’s Prairie Pothole
Status and Global Strategy for Neonicotinoids. J. Agric. Food Chem.
Region. PLoS One 2014, 9, e92821.
2011, 59, 2897−2908.
(24) Karmakar, R.; Singh, S. B.; Kulshrestha, G. Persistence and
(3) Tomizawa, M.; Lee, D. L.; Casida, J. E. Neonicotinoid
transformation of thiamethoxam, a neonicotinoid insecticide, in soil of
Insecticides: Molecular Features Conferring Selectivity for Insect
versus Mammalian Nicotinic Receptors. J. Agric. Food Chem. 2000, 48, different agroclimatic zones of India. Bull. Environ. Contam. Toxicol.
6016−6024. 2006, 76, 400−406.
(4) Tomizawa, M.; Casida, J. E. Neonicotinoid Insecticide (25) Lu, Z.; Challis, J. K.; Wong, C. S. Quantum Yields for Direct
Toxicology: Mechanisms of Selective Action. Annu. Rev. Pharmacol. Photolysis of Neonicotinoid Insecticides in Water: Implications for
Toxicol. 2005, 45, 247−268. Exposure to Nontarget Aquatic Organisms. Environ. Sci. Technol. Lett.
(5) Tomizawa, M.; Casida, J. E. Unique Neonicotinoid Binding 2015, 2, 188−192.
Conformations Conferring Selective Receptor Interactions. J. Agric. (26) Sadaria, A. M.; Sutton, R.; Moran, K. D.; Teerlink, J.; Brown, J.
Food Chem. 2011, 59, 2825−2828. V.; Halden, R. U. Passage of fiproles and imidacloprid from urban pest
(6) Bonmatin, J.-M.-M.; Giorio, C.; Girolami, V.; Goulson, D.; control uses through wastewater treatment plants in northern
Kreutzweiser, D. P.; Krupke, C.; Liess, M.; Long, E.; Marzaro, M.; California. Environ. Toxicol. Chem. 2016, 3098, 2555−2563.
Mitchell, E. A. D.; et al. Environmental fate and exposure; (27) Sadaria, A. M.; Supowit, S. D.; Halden, R. U. Mass Balance
neonicotinoids and fipronil. Environ. Sci. Pollut. Res. 2015, 22, 35−67. Assessment for Six Neonicotinoid Insecticides During Conventional
(7) Goulson, D. REVIEW: An overview of the environmental risks Wastewater and Wetland Treatment: Nationwide Reconnaissance in
posed by neonicotinoid insecticides. J. Appl. Ecol. 2013, 50, 977−987. U.S. Wastewater. Environ. Sci. Technol. 2016, 50, 6199−6206.
(8) Christen, V.; Mittner, F.; Fent, K. Molecular effects of (28) Iowa Department of Natural Resources. Mapping and GIS.
neonicotinoids in honey bees (Apis mellifera). Environ. Sci. Technol. http://www.iowadnr.gov/Conservation/Mapping-GIS (accessed No-
2016, 50, 4071−4081. vember 11, 2016).
(9) Henry, M.; Béguin, M.; Requier, F.; Rollin, O.; Odoux, J.-F.; (29) IIHR Hydroscience and Engineering, U.S. Geological Survey.
Aupinel, P.; Aptel, J.; Tchamitchian, S.; Decourtye, A. A common Iowa Water Quality Information System | IWQIS. https://iwqis.
pesticide decreases foraging success and survival in honey bees. Science iowawis.org/ (accessed November 10, 2016).
2012, 336, 348−350. (30) Hladik, M. L.; Calhoun, D. L. Analysis of the herbicide diuron,
(10) Pisa, L. W.; Amaral-Rogers, V.; Belzunces, L. P.; Bonmatin, J. three diuron degradates, and six neonicotinoid insecticides in water:
M.; Downs, C. A.; Goulson, D.; Kreutzweiser, D. P.; Krupke, C.; Liess, Method details and application to two Georgia streams. U.S.
M.; McField, M.; et al. Effects of neonicotinoids and fipronil on non-
Geological Survey Scientific Investigations Report 2012-5206; U.S.
target invertebrates. Environ. Sci. Pollut. Res. 2015, 22, 68−102.
(11) Van Dijk, T. C.; Van Staalduinen, M. A.; Van der Sluijs, J. P. Geological Survey: Reston, VA, 2012 (https://pubs.er.usgs.gov/
Macro-Invertebrate Decline in Surface Water Polluted with publication/sir20125206).
Imidacloprid. PLoS One 2013, 8, e62374. (31) Personal communication with D. McClain, Drinking Water
(12) Hallmann, C. A.; Foppen, R. P. B.; van Turnhout, C. A. M.; de Plant (Facilities Management), University of Iowa, Iowa City, IA,
Kroon, H.; Jongejans, E. Declines in insectivorous birds are associated 2016.
with high neonicotinoid concentrations. Nature 2014, 511, 341−343. (32) Van Metre, P. C.; Alvarez, D. A.; Mahler, B. J.; Nowell, L.;
(13) Gibbons, D.; Morrissey, C.; Mineau, P. A review of the direct Sandstrom, M.; Moran, P. Complex mixtures of Pesticides in Midwest
and indirect effects of neonicotinoids and fipronil on vertebrate U.S. streams indicated by POCIS time-integrating samplers. Environ.
wildlife. Environ. Sci. Pollut. Res. 2015, 22, 103−118. Pollut. 2017, 220, 431−440.
(14) Ford, K. A.; Casida, J. E.; Chandran, D.; Gulevich, A. G.; (33) Karmakar, R.; Singh, S. B.; Kulshrestha, G. Kinetics and
Okrent, R. A.; Durkin, K. A.; Sarpong, R.; Bunnelle, E. M.; mechanism of the hydrolysis of thiamethoxam. J. Environ. Sci. Health,
Wildermuth, M. C. Neonicotinoid insecticides induce salicylate- Part B 2009, 44, 435−441.

E DOI: 10.1021/acs.estlett.7b00081
Environ. Sci. Technol. Lett. XXXX, XXX, XXX−XXX
Environmental Science & Technology Letters Letter

(34) Maienfisch, P. Synthesis and Properties of Thiamethoxam and

Related Compounds. Z. Naturforsch., B: J. Chem. Sci. 2006, 61, 353−
359.
(35) Gusvany, V.; Csanadi, J.; Gaal, F. NMR Study of the Influence
of pH on the Persistence of Some Neonicotinoids in Water. Acta
Chim. Slov. 2006, 53, 52−27.
(36) Westerhoff, P.; Yoon, Y.; Snyder, S.; Wert, E. Fate of endocrine-
disruptor, pharmaceutical, and personal care product chemicals during
simulated drinking water treatment processes. Environ. Sci. Technol.
2005, 39, 6649−6663.
(37) Effects of Water Age on Distribution System Water Quality.
Distribution System Issue Paper; Office of Water (4601M), Office of
Ground Water and Drinking Water, U.S. Environmental Protection
Agency: Washington, DC, 2002 (https://www.epa.gov/sites/
production/files/2015-09/documents/2007_05_18_disinfection_tcr_
whitepaper_tcr_waterdistribution.pdf).
(38) Comninellis, C.; Kapalka, A.; Malato, S.; Parsons, S. A.; Poulios,
I.; Mantzavinos, D. Advanced oxidation processes for water treatment:
advances and trends for R&D. J. Chem. Technol. Biotechnol. 2008, 83,
769−776.

F DOI: 10.1021/acs.estlett.7b00081
Environ. Sci. Technol. Lett. XXXX, XXX, XXX−XXX
 Supporting Information

Occurrence of Neonicotinoid Insecticides in

Finished Drinking Water and Fate during

Drinking Water Treatment

Kathryn L. Klarich,§,† Nicholas C. Pflug,§,†,‡ Eden M. DeWald§, Michelle L. Hladik,# Dana W.

Kolpin,∆ David M. Cwiertny,§,†, * Gregory H. LeFevre§,†,*

§
Department of Civil & Environmental Engineering, University of Iowa, Iowa City, IA, 52242,

United States; †IIHR-Hydroscience and Engineering, University of Iowa, Iowa City, IA, 52242,

United States; ‡Department of Chemistry, University of Iowa, Iowa City, IA, 52242 United

States; #U.S. Geological Survey, California Water Science Center, 6000 J Street, Placer Hall,

Sacramento, CA 95819, United States; ∆U.S. Geological Survey, Illinois-Iowa Water Science

Center, 400 S. Clinton Street, Iowa City, IA 52240, United States

*Corresponding Authors:
GHL: gregory-lefevre@uiowa.edu; Phone: 319-335-5655; 4105 Seamans Center, University of
Iowa, Iowa City IA, United States

DMC: david-cwiertny@uiowa.edu; Phone: 319-335-1401; 4105 Seamans Center, University of

Iowa, Iowa City IA, United States

Supporting information includes: chemicals, water treatment plant schematics,

QA/QC procedures, analytical methods, and an additional 12 tables and 10 figures.
CHEMICALS.
Important chemicals used in the experiments include: clothianidin (99.9%, CAS 210880-92-
5), imidacloprid (99.9%, CAS 138261-41-3), imidacloprid-d4 (99.9%, CAS 1015855-75-0), and
thiamethoxam (99.6%, CAS 153719-23-4). All neonicotinoids were manufactured by Fluka and
used as received. All solvents used for LC-MS analysis were of LC-MS grade.

Solvents: Acetonitrile (optima grade, HPLC grade). Acetone (optima grade). Dichloromethane
(>99%).

Other Chemicals: Sodium hypochlorite solution 5.65-6% (Fisher Scientific). Granular

Activated Carbon (Calgon Centaur 12X40). 5 mM potassium phosphate buffer (made in lab).
Sodium Sulfite (Fisher).

METHOD DETAILS.

Disinfection
(a) (Chlorine gas)

1 2 3 4 5 6
Screening Flocculation Sedimentation Lime Recarbonation Filtration Chlorine
Softening (Rapid Sand) Contact Tap

Disinfection
(b) (Chlorine gas)

1 2
Aeration Lime Softening Recarbonation Filtration Chlorine
(GAC) Contact

Figure S1: Schematic of sampling locations (circled) at the two drinking water treatment plant
(DWTP) systems studied. a. University of Iowa DWTP schematic. Samples: (1) Raw source
water, (2) Sedimentation basin effluent (3) Recarbonation effluent – pre-chlorination, (4)
Recarbonation effluent – post chlorination (5) Filtration effluent (6) Finished water. b. Iowa City
DWTP Schematic. Samples (1) Source water (2) finished water.
Table S1: Hydraulic residence times for the University of Iowa Water Treatment Plant unit
operations. Ranges based on minimum expected flow (2.0 mgd) and maximum expected flow
(4.25 mgd).
Operation Residence time
(h)
Flocculation and Sedimentation 2.7-5.7
Softening 1.5-3.2
Filtration 1.1-2.3
Total 5.3-11.2

QA/QC Procedure: Deionized water (5 mL) was spiked with clothianidin (1 µM). A sample of
the 1 µM solution was run on the LC-MS/MS as a control. Three jars were filled with 1 L of
deionized water, and each jar was spiked with 1 mL of the 1 µM clothianidin solution (the
concentration in each 1 L jar is 1 nM). The 1 nM samples were each run through the entire SPE
process, concentrating the 1 L samples down to 1 mL. The concentrated samples were analyzed
by the LC/MS/MS, and peak areas were compared to the control to estimate recovery. Recovery
of clothianidin was 95%, 95% and 96% percent (average = 95%, SD=0.4%) for the three
samples.
All water samples (i.e., tap water and those from the DWTP process trains) were
collected directly into clean 1 L amber glass jars (pre-baked at 550 °C) with minimal headspace.
DWTP samples were collected from each unit operation and analyzed within 48 h of collection.
For tap water samples, the faucet was flushed for at least two minutes prior to sample collection,
and samples were stored for a maximum of 30 d at 11 °C to analysis.
A five-point internal standard normalized external calibration curve was used to account
for surrogate recovery and matrix effects during ionization. A calibration curve was run with
each set of samples. The instrument response was linear throughout the calibration range.
Multiple blanks were run with each set of samples, and no contamination was observed in the
blanks. Lab blanks only were generated (i.e., no “field blanks”) because neonicotinoids are non-
volatile making cross-contamination unlikely and residential samples were all collected by the
authors in their private residences where neonicotinoids were not used.

Sorption of Neonicotinoids to Granular Activated Carbon: Batch experiments measured the

extent and timescale of neonicotinoid sorption onto granular active carbon (GAC). Reactors were
assembled in clear, crimp-top glass vials (10-40 mL) and contained 5 g/L of GAC (Calgon) and
100 µg/L of an individual neonicotinoid (clothianidin, imidacloprid, or thiamethoxam) in
deionized water. A second set of experiments was conducted in pH 7-phosphate buffer (Figure
S.2). Once assembled, reactors were mixed by an end-over-end rotator for up to 4 h. Periodically,
samples (0.5 mL) of the suspension supernatant were collected at specified time intervals for LC-
DAD/MS analysis.

Chlorination of Neonicotinoids: Bench scale chlorination experiments were conducted to

assess the potential for neonicotinoid transformation during chemical disinfection and
distribution in the presence of residual disinfectant. To initiate reaction, hypochlorous acid
(HOCl) was added to a closed reactor (10 – 50 mL) containing either clothianidin, imidacloprid,
or thiamethoxam in 5 mM phosphate buffer at pH 7. A range of neonicotinoid (from 0.34 – 10
µM or 0.10 – 2.9 µg/L) and HOCl (0.0014-1.41 mM or 0.1-100 mg/L as Cl2) concentrations
were tested. Samples (0.5 – 1.0 mL) were collected at defined intervals and transferred to amber
glass vials for immediate analysis via high performance liquid chromatography coupled with a
diode array detector and single quadrupole mass spectrometer (LC-DAD/MS). Measurements of
solution pH and chlorine concentration (via titration of ferrous ammonium sulfate or FAS1) were
conducted immediately after chlorine addition and at the conclusion of each experiment. We note
that for experiments with clothianidin, which was most reactive toward free chlorine, residual
chlorine in samples was quenched with 1.8 mg sodium sulfite (Na2SO3) per mg of chlorine2 (as
Cl2) prior to LC-DAD/MS analysis. Sodium sulfite was not used for reaction samples with
imidacloprid and thiamethoxam; both reacted sufficiently slowly such that samples could be
immediately analyzed without altering the extent of decay.

Analytical Methods: Water samples collected from the taps and treatment plants were enriched
by solid phase extraction (SPE) methods adapted from the USGS.3 Briefly, DWTP samples were
filtered using a 0.7 µm glass filter (GF/F, Whatman) prior to SPE. Tap water samples were not
filtered. Samples were then spiked with imidacloprid-d4 as an internal standard before being
loaded onto an Oasis SPE cartridge (500 mg HLB; Waters). Prior to use, cartridges were
conditioned with 5 mL of dichloromethane (DCM), 5 mL of acetone, and 10 mL of deionized
water. One liter of sample (containing imidacloprid-d4) was loaded onto the cartridge using
negative pressure at a flow rate of ~10 mL/min or less. Sample bottles were washed with 100 mL
of DI and the rinsate was also loaded onto the cartridge. Following extraction, the cartridge was
dried under vacuum until visibly dry. The sample was then eluted into an acid-washed glass vial
using 10 mL of 50/50 DCM:acetone. The solvent was evaporated until just dry using a gentle
stream of nitrogen. The sample was then reconstituted into 1 mL of 50/50 acetonitrile: DI water
and stored at -20 °C until analysis via LC-MS/MS (Tables S.1 and S.2). Clean water controls
indicated a method recovery of 95 ± 0.4% (average ± SD, n = 3). Additional details are included
in the quality assurance and control (QA/QC).
Neonicotinoid samples were analyzed via high performance liquid chromatography
(Agilent 1260) coupled to a MS/MS spectrometer (LC-MS/MS; Agilent 6460 Triple Quadrupole
MS with MassHunter, version B.07.00) for tap water samples or DAD/MS (Agilent 6140
Quadrupole LC/MS and diode array detector with OpenLab ChemStation C.07.00) for
chlorination or GAC experiments. The chromatography column was a C18 Zorbax Eclipse Plus
(4.6 mm x 150 mm, 5 µm) held at 50 °C for LC-MS/MS and ambient temperature for LC-
DAD/MS. An injection volume of 20 µL was used, and the mobile phases were acetonitrile and
water with 0.1% formic acid at 0.8 mL/min. The mobile phase gradient is described in Table S.1.
Samples were quantified using the DAD at a wavelength of 260 nm (clothianidin and
thiamethoxam) and 280 nm (imidacloprid) and by mass spectrometry where possible. For
detection with mass spectrometer, samples were analyzed on electrospray ionization positive
mode, gas temperature 300 °C, gas flow 5 L/min, nebulizer 45 psi, sheath gas temp 250 °C,
sheath gas flow 11 L/min, capillary voltage 3500 V. Data were collected in multiple-reaction-
monitoring (MRM) mode using two transition ions (quantitation and verification). Optimum
MRM parameters were determined using Agilent Optimizer software (version B.07.00) by
injecting a 1 mg/L solution of each compound (clothianidin, imidacloprid, thiamethoxam) onto
the LC-MS/MS. MRM parameters are provided in Table S.2.
A five-point internal standard normalized external calibration curve was used to account
for surrogate recovery and matrix effects during ionization, and was run with each set of
samples. Multiple blanks were run with each set of samples, and no contamination was observed
in the blanks. The lower level of detection (LLD) on the LC-MS/MS without sample enrichment
for clothianidin, imidacloprid and thiamethoxam were 167, 99.7 and 204 ng/L, respectively. The
LLD following sample enrichment for clothianidin, imidacloprid and thiamethoxam were 0.167,
0.010 and 0.204 ng/L respectively.

Table S2: HPLC mobile phase gradient.

Time (min) % Acetonitrile % Deionized Water
0 15 85
11 25 75
13 25 75
15 95 5
15.5 15 85
21 15 85

Table S3: Multiple Reaction Monitoring (MRM) Parameters

Compound Precursor Quantitation Qualitative Fragmentor Quantitation Qualitative Retention
Ion (m/z) Ion (m/z) Ion (m/z) (V) ion collision ion collision time
energy (V) energy (V) (min)
Imidacloprid 256.06 209 175.1 59 12 12 11.2
Clothianidin 250.02 169.1 131.9 67 8 12 10.0
Thiamethox 292.03 211 181 63 8 20 7.8
am
Imidacloprid 260.09 213 179.1 59 12 16 11.1
-d4

FAS titration method:

Reagents (see full description in standard methods):
Phosphate buffer solution (169 mM as PO4)
N,N-Diethyl-p-phenylenediamine (DPD) indicator solution (5.72 mM)
Ferrous ammonium sulfate (FAS) titrant (2.8 mM as FeII)

1. Measure 100 mL of DI water using a volumetric flask

2. Pour DI water into a beaker
3. Add 1 mL of sample to the 100 mL of DI water
4. Add 5 mL of phosphate buffer solution and 5 mL of N,N-Diethyl-p-phenylenediamine
(DPD) indicator solution
5. Titrate with Standard ferrous ammonium sulfate (FAS) until the red color is gone
6. Calculate free chlorine concentration: (volume of FAS added)*100=Free chlorine (mg/L
as Cl2)
Lower Level of Detection Calculation:
Based on Standard Methods 1030 E Method Detection Level1. Method overview:
1. A standard containing 0.1 uM of clothianidin, imidacloprid and thiamethoxam was
injected seven times in a row on the LC-MS/MS.
2. The standard deviation (s) of the concentration measured was calculated for each
compound
3. To reduce the probability of a type I error, the standard devation was multiplied by two
times 1.645 from a cumulative normal probability table: LLD = 2*1.645*s.
Table S4: University of Iowa tap water sample results. Samples collected from the same tap in
the laboratory at Seaman’s Center for Engineering.

Thiamethoxam Imidacloprid Clothianidin

Date Location (ng/L) (ng/L) (ng/L)
5/31 SC 4249 0.65 2.32 5.73
6/2 SC 4249 0.42 1.22 3.89
6/3 SC 4249 0.61 1.38 4.24
6/6 SC 4249 1.04 3.26 10.19
6/7 SC 4249 2.04 2.26 5.73
6/10 SC 4249 1.22 2.33 7.02
6/15 SC 4249 2.61 5.53 13.57
6/20 SC 4249 4.15 3.38 10.29
6/21 SC 4249 1.13 4.24 13.88
6/23 SC 4249 0.84 5.05 12.58
6/27 SC 4249 1.19 26.36 27.27
6/28 SC 4249 0.85 16.30 33.46
6/29 SC 4249 1.19 16.13 30.97
7/1 SC 4249 0.26 10.20 20.51
7/7 SC 4249 0.49 5.27 11.19
7/18 SC 4249 0.77 3.69 13.30

Table S5: Iowa City tap water results summary from samples collected from three residential
locations in Iowa City.
Imidacloprid Thiamethoxam
Date Location Clothianidin (ng/L) (ng/L) (ng/L)
7/18/16 1 ND ND 0.34
7/18/16 2 ND ND <0.20
7/18/16 3 ND ND 0.37
7/27/16 1 ND <0.10 0.47
**ND indicates non-detect, <LLD indicates that compound was detected at concentrations below
the LLD
Table S6: Clothianidin concentrations in samples from the University of Iowa water treatment
plant (concentrations in nanograms per liter)
Recarbonation
Source Recarbonation (Post-
Water Sedimentation (Pre- chlorination) Filtration Finished
Date (1) Basin (2) chlorination)(3) (4) effluent (5) Water (6)
6/29/16 26.0 26.6 26.0 24.3 26.2 31.2
7/18/16 7.82 10.9 11.0 8.75 7.76 9.50

Table S7: Imidacloprid concentrations in samples from the University of Iowa water treatment
plant (concentrations in nanograms per liter)
Recarbonation Recarbonation
(Pre- (Post- Filtration
Source Sedimentation chlorination) chlorination) effluent Finished
Date Water (1) Basin (2) (3) (4) (5) Water (6)
6/29/16 13.3 11.6 13.1 11.6 0.72 13.6
7/16/16 4.00 4.48 4.78 3.58 3.30 4.14

Table S8: Thiamethoxam concentrations in samples from the University of Iowa water treatment
plant (concentrations in nanograms per liter)
Recarbonation Recarbonation
(Pre- (Post- Filtration
Source Sedimentation chlorination) chlorination) effluent Finished
Date Water (1) Basin (2) (3) (4) (5) Water (6)
6/29/16 8.23 10.7 2.12 2.43 2.40 3.11
7/18/16 2.81 4.01 1.78 1.84 1.55 1.71

Table S9: Clothianidin concentrations in the University of Iowa and Iowa City Source and
Finished waters (August 9, 2016)
WTP Source Finished Water
Water (ng/L)
(ng/L)
UI 10.7 10.6
City 7.53 ND

Table S10: Imidacloprid concentrations in the University of Iowa and Iowa City Source and
Finished Waters (August 9, 2016).
WTP Source Water Finished Water
(ng/L) (ng/L)
UI 2.15 1.97

City 1.53 0.09

Table S11: Thiamethoxam concentrations in the University of Iowa and Iowa City Source and
Finished Waters (August 9, 2016).

WTP Source Water Finished Water

(ng/L) (ng/L)
UI 1.93 1.07
City 2.50 0.37

1.20

1.00
Fraction Remaining

0.80

0.60 100 uM Thia, pH 7

100 uM Thia, pH 10.4
0.40

0.20

0.00
0 10 20 30 40 50 60 70 80
Time Elapsed (hr)
Figure S2: Thiamethoxam hydrolysis in ambient pH University DWTP softening basin water
(pH 10.4) compared to University DWTP softening basin water adjusted to pH 7

1.00 1000
0.90 900
0.80 800
Fraction Remaining

0.70 700 Product Peak Area

0.60 600
100 uM Thiamethoxam,
0.50 500
pH 10.4
0.40 400
0.30 300
0.20 200
0.10 100
0.00 0
0 5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
Time (hr)
Figure S3: Product formation during thiamethoxam hydrolysis in University DWTP softening
basin water
 0.00
0 10 20 30 40 50 60 70 80
-0.50

-1.00
Ln(C/Co)

-1.50
y = -0.0379x - 0.1078
-2.00 R² = 0.9859

-2.50

-3.00

-3.50
Time (hr)

Figure S4: Ln(C/Co) versus time for thiamethoxam hydrolysis in University DWTP softening
basin water. Kobs = 0.0379 h-1, t1/2 = ln(0.5)/Kobs = 18.3 h.

1.00
0.90
0.80
0.70
fraction remaining

Clothianidin
0.60
Imidacloprid
0.50
0.40 Thiamethoxam

0.30
0.20
0.10
0.00
0 20 40 80 60 100 120 140
Time (min)
Figure S5: Adsorption of clothianidin, imidacloprid, thiamethoxam to GAC. Experimental
conditions: GAC 5 g/L, chemical concentration 100 ug/L (clothianidin, imidacloprid and
thiamethoxam), experiment conducted in pH 7 phosphate buffer.
 1.00
0.90
0.80
fraction remaining

0.70
0.60
0.50 Clothianidin
0.40 Imidacloprid
0.30
Thiamethoxam
0.20
0.10
0.00
0 20 40 60 80 100 120 140
Time (min)

Figure S6: Replication experiment of adsorption of clothianidin, imidacloprid, thiamethoxam to

GAC. Experimental conditions: GAC 5 g/L, chemical concentration 100 ug/L (clothianidin,
imidacloprid and thiamethoxam), experiment conducted in DI water.

Table S12: Exponential decay parameters for GAC adsorption (Figure 4)

Compound Clothianidin Imidacloprid Thiamethoxam

K (h-1) 0.3098 0.1469 0.1812
R2 0.9910 0.9974 0.9784
 1 14
0.9
12
0.8
Fraction remaining

Product Peak Area

0.7 10
Clothianidin 0.4 uM
0.6 8
0.5 Product 7.2
0.4 6
Product 8.6
0.3 4
0.2 Product 10.0
2
0.1
Product 13.1
0 0
0 5 10 15 20 25
Time Elapsed (h)

Figure S7: Product formation during chlorination of clothianidin. Experimental conditions:

Chlorine 10 mg/L as Cl2, clothianidin 0.4 µM, pH 7. The formation of intermediates (shown in
this figure) may explain why we observe initial fast reaction rates followed by slow decay of
clothianidin. We hypothesize that the intermediates are more reactive and may outcompete
clothianidin for chlorine causing the decay of clothianidin to slow after a fast initial reaction.

Figure S8: Chromatogram of chlorination reaction shown in Figure S7. Clothianidin

concentration 100 µg/L, just prior to adding chlorine (0 mg/L Cl2, t=0). Clothianidin residence
time = 11.13 min, wavelength = 260 nm.
Figure S9: Chromatogram of chlorination reaction shown in Figure S7. Clothianidin
concentration 100 µg/L, chlorine concentration 10 mg/L Cl2, time = 3 h. Clothianidin residence
time = 11.13 min, wavelength = 260 nm.

1.1 12
1
0.9 10

Product Peak Area

0.8
% Remaining

0.7 8
0.6 6
0.5 Clothianidin
0.4 4
0.3 Product 1
0.2 2
0.1
0 0
0 20 40 60 80 100
Time (hr)

Figure S10: Product formation during chlorination of clothianidin. Experimental conditions:

Chlorine 5 mg/L as Cl2, clothianidin 5 µM, pH 7. The formation of intermediates (shown in this
figure) may explain why we observe initial fast reaction rates followed by slow decay of
clothianidin. We hypothesize that the intermediates are more reactive and may outcompete
clothianidin for chlorine causing the decay of clothianidin to slow after a fast initial reaction.

REFERENCES.

(1) American Public Health Association; American Water Works Association; Water
Pollution Control Federation. Standard Methods for the Examination of Water and
Wastewater; American Public Health Association: Washington DC, 1992.
(2) Tikkanen, M. W.; Schroeter, J. H.; Leong, L. Y. C.; Ganesh, R. Guidance manual for the
disposal of chlorinated water; AWWA Research Foundation, American Water Works
Association: Denver, 2001.
(3) Hladik, M.L. and Calhoun, D.L. Analysis of the herbicide diuron, three diuron degradates,
and six neonicotinoid insecticides in water—Method details and application to two
Georgia streams: U.S. Geological Survey Scientific Investigations Report 2012–5206, 10
p. 2012.