Determination of Caffeine and Its Metabolites in

ANALYTICAL SCIENCES MARCH 2018, VOL. 34 349

Introduction

Caffeine is a common purine alkaloid found in many popular drinks, especially tea and coffee. After human consumption, caffeine enters wastewater treatment systems,1–4 and have been detected in many surface water sample.5–7 Furthermore, caffeine generates demethylated products because of biodegradation by human metabolism and microbial biodegradation.8,9 Biodegradation has so far been the main process involved in the removal of caffeine during activated sludge treatment,10 and demethylation is expected to be the main pathway. Firstly, theophylline, paraxanthine and theobromine are formed by oxidative demethylation of caffeine. Secondly, 1-methylxanthine, 3-methylxanthine, 7-methylxanthine and xanthine are formed from theophylline, paraxanthine and theobromine by demethylases (N-1 demethylase, N-7 demethylase and N-3 demethylase, respectively).11 Caffeine metabolites were well studied in previous studies on human metabolism, and a few studies reported several caffeine metabolites such as theophylline and paraxanthine in wastewater influents and effluents.12–17 In addition, the occurrence of theobromine was reported in natural water.18 In Japan, caffeine and theophylline have been widely detected in wastewater,19,20 but no study has evaluated the occurrence of other caffeine metabolites in the water environment.

In previous studies, liquid chromatography–tandem mass spectrometry (LC-MS/MS) was used for the detection of caffeine and a limited number of its metabolites (i.e., paraxanthine, theophylline, and theobromine).21–24 However, this technique was applicable to urine and blood samples and not measurable for all demethylated metabolites in environmental samples (i.e., wastewater and river water). For biodegradation of organic compounds, such as nonylphenol, structural isomers of the metabolites were expected to occur in wastewater samples as the treatments involve complicated degradation processes.25 A similar situation was expected in the biodegradation of caffeine,11 resulting in isomeric metabolites with a similar fragment in mass spectrometry, and separation of structural isomers by LC could be feasible for the detection of caffeine metabolites in wastewater. In environmental samples, the complexity of co-existing materials and trace level concentrations of target analytes were expected, so solid phase extraction (SPE) is important to extract target compounds from samples.26,27 In this study, a simultaneous detection method by LC-MS/MS and SPE process was applied for caffeine and its demethylated metabolites in environmental samples.

A concentration of caffeine and its metabolites in raw municipal wastewater reflects their consumption in the communities served by the target wastewater treatment plant (WWTP), which can be used to track the lifestyle in the area.28,29 In addition, caffeine in water was considered as an indicator of wastewater contamination30 and a biomarker of population size according to the consumption habits.29,31 However, the occurrence of caffeine in wastewater influents alone cannot

2018 © The Japan Society for Analytical Chemistry

† To whom correspondence should be addressed.E-mail: [email protected]

Determination of Caffeine and Its Metabolites in Wastewater Treatment Plants Using Solid-Phase Extraction and Liquid Chromatography–Tandem Mass Spectrometry

Kai HE,*† Shinya ECHIGO,** Yasuhiro ASADA,* and Sadahiko ITOH*

* Department of Environmental Engineering, Graduate School of Engineering, Kyoto University, Nishikyo, Kyoto 615–8540, Japan

**Department of Environmental Health, National Institute of Public Health, Wako, Saitama 351–0197, Japan

For caffeine and its seven major metabolites (i.e., theobromine, theophylline, paraxanthine, 1-methylxanthine, 3-methylxanthine, 7-methylxanthine, and xanthine), an optimized analytical method using liquid chromatography–tandem mass spectrometry (LC-MS/MS) for their detection in wastewater samples was developed in this study. Extraction of these compounds (recoveries ranged from 60.3 to 83.2%) was made possible by combining universal polymeric reversed-phase cartridge and polymeric strong cation exchange cartridge. This method was applied to the determination of caffeine and its metabolites in the influent and effluent of an anaerobic–anoxic–oxic (A2O) process. In the A2O influent, caffeine and its metabolites (except xanthine) ranged from 1.39 to 5.45 μg/L, and their concentrations in the A2O effluent ranged from 10.2 to 171.3 ng/L. The mass load of caffeine was 14.9 g/day/1000 inhabitants, considering the population served by the target wastewater treatment plant (WWTP). The concentration of caffeine derivatives in wastewater influent is a tool for estimating the population size in the area served by WWTPs.

Keywords Caffeine, metabolites, LC-MS/MS, wastewater

(Received September 3, 2017; Accepted November 7, 2017; Published March 10, 2018)

350 ANALYTICAL SCIENCES MARCH 2018, VOL. 34

provide comprehensive information on population estimation with respect to consumption because of possible transformation by biodegradation. Thus, the concentration of caffeine metabolites in water samples should be evaluated to augment the role of caffeine as an environmental biomarker of the population by coffee consumption.

In this study, our goal was to develop an LC-MS/MS method for quantifying caffeine and its demethylated metabolites in wastewater samples. Furthermore, this study addressed the occurrence of caffeine metabolites including paraxanthine, theobromine, methylxanthines and xanthine in wastewater influent and effluent for the first time in Japan. Consumption of caffeine in the studied area was estimated based on concentrations of caffeine and its metabolites in wastewater influents, and removal characteristics of caffeine and its metabolites in the WWTP were evaluated.

Experimental

Considering the demethylation of caffeine in biodegradation, paraxanthine, theobromine, theophylline, 1-methylxanthine, 3-methylxanthine, 7-methylxanthine, and xanthine were chosen as target metabolites in this study.

Reagents and chemicalsCaffeine was purchased from Cosmo Bio, Japan. We obtained

3-methylxanthine, theophylline, theobromine and xanthine from Wako Pure Chemical Industries, Japan. A supply of 1-methylxanthine was purchased from Tokyo Chemical Industry, Japan, while 7-methylxanthine was purchased from Sigma-Aldrich, USA. Paraxanthine was purchased from Toronto Research Chemicals, Canada. Characteristics (i.e., chemical structure, molecular formula, pKa, and log P) of caffeine and its metabolites are summarized in Table S1 (Supporting Information). Ultrapure water was produced by a Milli-Q Advantage A10 System.

Target WWTPIn this study, influent and effluent water from an actual

WWTP adopting the anaerobic–anoxic/oxic (A2O) process in Kyoto City (Japan) were sampled for analysis of caffeine and its metabolites. The daily flow rate of the WWTP was approximately 500000 m3/day in the sampling period. The

population of people served by the target WWTP was 784, 300. A2O effluent was sampled in the effluent port of the oxic tank, and A2O influent was sampled in the sedimentation tank. Samples were collected from July 27th to August 16th, 2016.

Solid phase extraction (SPE)HLB cartridge (Oasis, Waters, USA) or other C18 cartridge

was used for the concentration treatment for caffeine and theophylline.4,32 However, the pKa values of caffeine derivatives range from 7.53 to 10.76, indicating the caffeine derivatives to be weakly basic compounds. A strong cation exchange (SCX) cartridge could also be used to extract weakly basic compounds possessing one or more positive charges. In this study, a combination of HLB cartridge and Strata X-C cartridge (500 mg, Phenomenex, USA) was used for extracting caffeine derivatives from wastewater samples. Samples were filtered by glass fiber membrane filters (GF-75, 0.3 μm, 90 mm, ADVANTEC, Toyo Roshi Kaisha, Japan), and the pH of the filtrates was adjusted to 3 by adding hydrochloric acid. The acidified filtrates were concentrated by SPE. The protocol of the SPE treatment is summarized in Fig. 1. The HLB Plus cartridge (225 mg, Waters, USA) was conditioned with 5 mL of methanol and 5 mL ultrapure water at pH 3, and the Strata-X-C cartridge was conditioned with 5 mL of methanol, 5 mL of ultrapure water, and 5 mL of acetate buffer (pH 4.5, 0.5 M). Different concentrations of caffeine and its metabolites were expected in A2O influent and effluent water, so samples were pumped through the cartridge at 5 mL/min for 40 min (200 mL, A2O effluent) and at 2 mL/min for 25 min (50 mL, A2O influent) with a concentrator pump (Sep-Pak, Waters), respectively. The HLB cartridge was flushed with 5 mL of ultrapure water at pH of 3, and The Strata-X-C cartridge was flushed with 5 mL of acetate buffer (pH 4.5, 0.5 M). After drying the SPE cartridge with nitrogen gas, the HLB cartridge was rinsed with 5 mL of methanol, and the Strata-X-C cartridge was rinsed with 5 mL of a mixture of methanol and ammonia water (v/v of 95/5, and the initial concentration of ammonia water was 28%). The eluted fractions from each SPE cartridge were divided into two, respectively; one fraction was used to evaluate the recovery ratio of each SPE process, and the other was mixed together for simultaneous quantification of all components of caffeine and its metabolites.

The eluted solution from these two cartridges was evaporated until dry under a gentle stream of nitrogen gas with a dry thermo

Fig. 1　Protocol of SPE for caffeine and its metabolites.


bath at 35°C (MG-2000, EYELA, Japan). The residues of the A2O process influents and the effluents were re-dissolved in 2 mL of a mixture of ultrapure water and methanol (92/8, v/v, respectively).

Instrumentation and conditionsIn this study, chromatographic separations were achieved on a

Shimadzu prominence 20A HPLC system (Shimadzu, Japan) with selected LC columns. VP-ODS column (4.6 × 150 mm, 4.6 μm, Shimadzu, Japan), XR-ODS (2.0 × 75 mm, 2.2 μm, Shimadzu, Japan), and kinetex C18 (2.1 × 75 mm, 2.6 μm, Phenomenex, USA) were selected, as they were widely used in previous studies on analyzing pharmaceuticals as well as other C18 columns.33,34 VP-ODS was chosen on the basis of separation of caffeine metabolites followed by optimization of LC conditions including the flow rate and the organic solvent.

Method validationThe LC system was interfaced to a triple quadrupole mass

spectrometer (4000 QTRAP, SCIEX, USA) with an electrospray ionization (ESI) source (Turbo V Ion Source, SCIEX). The operating conditions of LC and MS/MS are shown in Table S2 (Supporting Information). Standard solutions of caffeine derivatives were spiked into the ultrapure water for validation. Furthermore, the matrix effect of wastewater from the target WWTP mentioned above was evaluated for caffeine and its metabolites, referring to previous works on matrix effects of pharmaceuticals including caffeine in wastewater.35 The SPE extracts obtained for wastewater influent and wastewater effluent from five different batches were spiked at 100 and 10 μg/L for each caffeine derivative. The formula of signal suppression was expressed as Eq. (1):

Signal suppression (%)= s+x x

s1 100− −⎛

⎝⎜⎞⎠⎟

×I II

(1)

where Is+x is the peak area of caffeine derivative spiked in the wastewater extract, Ix is the peak area of that in the unspiked extract, and Is is the peak area of caffeine derivative in ultrapure water spiked at the same level.

Results and Discussion

Optimization of MS/MS and LC parametersCaffeine metabolites are derivatives of xanthine with methyl

functional groups at different positions, and they have similar Q1/Q3 transitions in the multiple reaction monitoring (MRM) method (Table S3, Supporting Information). For example, the Q1/Q3 transitions of theophylline and paraxanthine were 181/124 and the Q1/Q3 transitions of 3-methylxanthine and 7-methylxanthine were 167/124. Optimized LC method is necessary for the quantitation of caffeine derivatives.

To improve the separation of caffeine and its metabolites, the LC method was optimized as described in Tables S4 and S5 (Supporting Information), and their chromatograms were shown in Figs. S1 – S4 (Supporting Information). The resolutions of peaks under selected conditions are shown in Table S6 (Supporting Information). Resolution above 1.5 is highly desirable for quantitative analysis.36 VP-ODS column was selected as target LC column for the detection of caffeine metabolites, and a flow rate of 0.8 mL/min using a gradient of 10% methanol was considered as an optimized gradient condition for detecting caffeine and its metabolites in this study (Fig. 2). The retention times of caffeine derivatives were 3.53,

5.31, 6.24, 7.20, 10.04, 15.48, 16.80, and 19.86 min for xanthine, 7-methylxanthine, 3-methylxanthine, 1-methylxanthine, theobromine, paraxanthine, theophylline, and caffeine, respectively. The quantitation of caffeine derivatives will be discussed in the following section.

As presented above, the influence of the amount of formic acid in the ultrapure water on the intensity of caffeine derivatives was evaluated (Fig. S5, Supporting Information). The intensities of xanthine, 7-methylxanthine, and caffeine in ultrapure water were higher than those in ultrapure water containing 0.3 and 0.1% of formic acid as solvent A. Though the presence of volatile formic acid enhances ionization efficiencies of compounds in positive mode,37 caffeine derivatives are weak basic compounds with pKa ranging from 7.53 to 10.76, indicating the acid in the mobile phase enhances their dissociation in the LC column. In summary, ultrapure water and methanol were used as mobile phases in this study.

SPE procedure and recoveries of caffeine and its metabolitesThe Strata-X-C cartridge is packed with sorbent with a

benzenesulfonic acid group uniformly bonded on the polymeric surface,15 and they were used after HLB cartridge for extracting caffeine derivatives from wastewater samples. Caffeine and its metabolites were spiked into wastewater influent and wastewater effluent, respectively. In Table 1, the recoveries of caffeine and its metabolites in samples of the A2O influent ranged from 60.3 to 78.4%, and the recoveries of caffeine derivatives in samples of the A2O effluent ranged from 66.3 to 83.2% (n = 7). The recoveries of most caffeine derivatives in the A2O effluent were higher than that in the A2O influent. This could be attributed to the large interference of co-extracted compounds.37

Fig. 2　Chromatogram of caffeine and its metabolites (50 μg/L).

Table 1　Recoveries of caffeine and its metabolites in the A2O influent and effluent (n = 7)

CompoundRecoveries in A2O

influent, %Recoveries in A2O

effluent, %

Caffeine 64.3 ± 8.0 83.2 ± 5.0Theophylline 72.2 ± 4.0 81.1 ± 6.3Theobromine 74.1 ± 10.1 69.2 ± 6.5Paraxanthine 67.1 ± 4.5 82.2 ± 4.71-Methylxanthine 60.3 ± 5.8 79.0 ± 6.23-Methylxanthine 75.6 ± 7.6 80.3 ± 4.57-Methylxanthine 67.1 ± 8.3 69.6 ± 9.9Xanthine 78.3 ± 6.0 66.3 ± 5.2


Among caffeine derivatives, the recoveries of theobromine, 7-methylxanthine, and xanthine were relatively lower than those of caffeine, theophylline, paraxanthine, 1-methylxanthine, and 3-methylxanthine. The complexity of the matrix in environmental samples are likely responsible for the lower recoveries of pharmaceuticals.38 However, the standard deviations of caffeine derivatives in the A2O influent and the A2O effluent were less than 10%, except for theobromine in the A2O influent (10.1%). Referring to previous studies on pharmaceuticals in wastewater,20,38 the recoveries and standard deviations of caffeine derivatives in this study indicated that this SPE method was reliable for extracting caffeine derivatives. In addition, a recovery experiment is recommended together with a wastewater application experiment to compensate the matrix difference at the different sampling time.

In Fig. 3, the contributions of Strata-X-C cartridge and HLB cartridge to the recoveries of caffeine and its metabolites in wastewater influent and effluent were evaluated by measuring the concentration of caffeine derivatives in the elution from these two cartridges. The HLB cartridge contributed to the recoveries of caffeine, theophylline, theobromine, and paraxanthine, while the Strata-X-C cartridge contributed to the recoveries of xanthine in wastewater influent and effluent. Furthermore, both HLB and Strata-X-C cartridges contributed to the recoveries of 1-methylxanthine, 3-methylxanthine, and 7-methylxanthine, and Strata-X-C cartridge contributed to the dominant extraction of xanthine in wastewater samples (73.3% and 72.1% in influent and effluent of the A2O process, respectively). Several previous studies reported caffeine and theophylline were also extracted from samples by Strata-X-C cartridge, and higher recoveries of caffeine and theophylline were reported.39,40 This study expanded this idea with not only the Strata X-C cartridge, but also the HLB cartridge to establish a reliable SPE method for all caffeine demethylated metabolites in wastewater samples for the first time.

Method validationThe linearity of this method was confirmed by plotting the

response of peak area versus spiked concentrations from 0.05 to 500 μg/L (0.05, 0.1, 0.5, 1, 5, 10, 50, 100, 250, 500 μg/L). The correlation coefficient (R) of each caffeine derivative was at or above 0.99 (Table S7, Supporting Information). These results indicated that this method was reliable for the quantitative analysis of caffeine derivatives in wastewater.

Chromatograms of matrix samples of the A2O influents and effluents are shown in Fig. S6 (Supporting Information), and the clear separation of caffeine and its metabolites was observed for wastewater samples in the chromatograms. In the chromatogram of xanthine in the A2O influent, several impurities with a transition of 153/109 were observed. Suppression by matrix effects from A2O influent and effluent in two months (July and August) for the security of homogeneity of matrix solution used was evaluated for caffeine and its metabolites (Fig. S7, Supporting Information), and signal suppression was observed for caffeine and its metabolites in the A2O samples, except for 3-methylxanthine in A2O effluent. Compared with methylxanthines and xanthine (ranging from 27.4 to 36.3% in the A2O influent and from 7.2 to 23.5% in the A2O effluent), signal suppressions were observed for caffeine and dimethylxanthines (ranging from 36.4 to 43.8% in the A2O influent and from 24.0 to 38.3% in the A2O effluent). As expected, greater signal suppressions were observed for caffeine and its metabolites in the A2O influent than the A2O effluent, which could be attributed to the complexity of the matrix in the wastewater influent.3 For example, matrix compounds entering the ionization at the same time with caffeine derivative.41 In addition, other pharmaceuticals could exist in the samples after SPE treatment.4 Referring to the work of Sui et al.,2,35 the matrix suppression of caffeine in wastewater was 49%, and a calibration curve was applied for the quantitation of caffeine by a standard solution. As previous studies recommended,42,43 surrogate and matrix match calibration should be applied for the quantitation of pharmaceuticals or target compounds in complicated matrix samples. However, it is difficult to get surrogates for all caffeine derivatives. In addition, in wastewater samples, matrix match calibration was not practical due to the wide existence of caffeine derivatives in wastewater samples. Thus, a calibration curve by standard solution was used for the quantitative analysis of caffeine derivatives, but the concentration of caffeine derivatives in wastewater samples was corrected by considering recoveries of these compounds.

Application to wastewater samplesThe developed method was applied to determine caffeine

derivatives in the A2O influent and effluent. The concentrations of caffeine derivatives (caffeine, paraxanthine, theobromine, theophylline, 1-methylxanthine, 3-methylxanthine, 7-methylxanthine and xanthine) in A2O influent and A2O effluent are

Fig. 3　The contribution of Strata-X-C cartridge and HLB cartridge on the recoveries of caffeine and its metabolites in the A2O influent and effluent.


shown in Fig. 4.In the A2O influent, caffeine, theobromine, theophylline,

paraxanthine, 1-methylxanthine, 3-methylxanthine, 7-methylxanthine, and xanthine were detected at 5.45 ± 0.87, 3.03 ± 0.47, 3.60 ± 0.83, 2.73 ± 0.67, 3.75 ± 0.91, 1.40 ± 0.31, 1.39 ± 0.25 and 0.106 ± 0.017 μg/L, respectively. Similar concentrations of several caffeine derivatives were detected in other studies: mean concentrations in Italian wastewaters were 5.0 – 61.9 and 7.3 – 84.1 μg/L for 1-methylxanthine and 7-methylxanthine, respectively.31 Concentrations of 1.6 – 238 μg/L of paraxanthine were observed in wastewaters in northeastern Spain,13 and paraxanthine was detected at 11.8 μg/L in wastewater influent.16 These results showed that the concentration level of caffeine derivatives in Japan was similar to that in Italy and Spain. The concentration fluctuation of caffeine derivatives could be attributed to water quality in the influent.44 The total concentration of caffeine metabolites was 16.01 μg/L in wastewater influent, which was 2.93 times the caffeine concentration, their parent compound. From the perspective of hazard assessment, caffeine metabolites were expected to be similar to their parent compound, caffeine.45 Thus, the concentrations of caffeine metabolites in wastewater should not be ignored. This study addressed the concentrations of caffeine derivatives in wastewater in Japan for the first time, with concentrations of 3-methylxanthine and xanthine being reported for the first time in wastewater samples.

Mass loads of caffeine and of its metabolites were calculated by multiplying their concentration (μg/L) by the flow rate of the selected WWTP (500000 m3/day) and divided by the population served by the target WWTP (784300). The mass load of caffeine derivatives was 13.68 g/day/1000 inhabitants, indicating that the equivalent mass load of caffeine was 14.90 g/day/1000 inhabitants. This was comparable with loads in Swiss WWTPs (15.8 ± 3.8 g/day/1000 inhabitants)46 and Italian WWTPs (14 ± 5.2 g/day/1000 inhabitants).31 Coffee consumption in Japan was 3.54 kg/person/year in 2014; 100 g of coffee contains 40 mg of caffeine. These results indicated that the mass load profile of caffeine derivatives was 3.75 times the coffee consumption (caffeine of 3.97 g/day/1000 inhabitants), possibly due to consumption of other food and drink containing caffeine (i.e., chocolate products and tea).47 Tea is also a popular drink

in the Kyoto area. The consumption of green tea in Kyoto was 755 g/person/year in 2015. Considering caffeine in green tea (approximately 20 mg caffeine in 10 g of green tea), the consumption of green tea contributed to caffeine of 4.13 g/day/1000 inhabitants. Similarly, the consumption of black tea in Kyoto contributed to caffeine of 1.75 g/day/1000 inhabitants (the consumption of black tea in Kyoto was 32.6 bags/person/year; caffeine in one bag of black tea was approximately 20 mg), and the consumption of chocolate in Kyoto contributed to caffeine of 1.54 g/day/1000 inhabitants (chocolate consumption in Kyoto was 1309 g/person/year; 100 g chocolate contains 43 mg caffeine).48 In summary, total caffeine consumption was 11.39 g/day/1000 inhabitants in Kyoto based on a calculation of the consumption of caffeine-rich drink and food, which was 76% of estimated caffeine consumption by measuring the concentration of caffeine derivatives. Though in-sewer degradation of caffeine and its metabolites could occur,31 the concentration of caffeine derivatives in wastewater influent is a tool to estimate the caffeine consumption of population in the area served by WWTPs.

In the A2O effluent (Fig. 4), the concentrations of caffeine derivatives (caffeine, paraxanthine, theobromine, theophylline, 1-methylxanthine, 3-methylxanthine, 7-methylxanthine and xanthine) were 115.2 ± 18.9, 47.8 ± 21.4, 109.1 ± 23.8, 66.9 ± 21.3, 171.3 ± 67.9, 10.2 ± 7.1, 47.7 ± 27.7, 113.2 ± 36.6 ng/L, respectively. Besides the concentration fluctuation of caffeine derivatives in influent as presented above, the performance of activated sludge could result in the fluctuation of caffeine derivatives in effluents. The concentration of 1-methylxanthine was higher than that of caffeine in the A2O effluent. Similar concentrations of caffeine metabolites were observed in previous studies: theophylline was detected in high concentration (230 ng/L on average) in effluents of WWTPs.14 Theobromine was detected at above 100 ng/L (n = 4) in effluents of several WWTPs,49 and 1-methylxanthine was detected at 61 ng/L in river samples 50. During the A2O process, removal efficiencies of caffeine derivatives were effective (above 95%) except for xanthine (–5.4%). Effective removals of caffeine and theophylline in activated sludge were previously reported by other studies, which could be attributed to sorption and biodegradation in activated sludge.51 Xanthine was formed through the biodegradation of methylxanthines,11 which could contribute to the increase in xanthine after A2O treatment. In order to clarify the transformation trend of caffeine in activated sludge, the distribution of caffeine derivatives in water phase and solid phase (activated sludge) should be further evaluated.

Conclusions

A simultaneous detection method by LC-MS/MS was provided for quantitative analysis of caffeine and its demethylated metabolites in wastewater samples. Caffeine, theobromine, theophylline, and paraxanthine were extracted by the universal polymeric reversed-phase cartridge. Meanwhile, 1-methylxanthine, 3-methylxanthine, 7-methylxanthine, and xanthine were extracted by polymeric strong cation exchange cartridge. The combination of these two cartridges was suitable for extracting caffeine and its metabolites from wastewater samples (recoveries ranged from 60.3 to 78.3% in A2O influent and from 69.2 to 83.2% in A2O effluent).

In the A2O influent, caffeine, theobromine, theophylline, paraxanthine, 1-methylxanthine, 3-methylxanthine, 7-methylxanthine, and xanthine ranged between 0.106 – 5.45 μg/L. The mass load of caffeine derivatives was 13.68 g/day/1000 inhabitants;

Fig. 4　The concentration of caffeine and its metabolites in the A2O influent and A2O effluent (error bar: standard deviation, n = 4).


the concentration of caffeine derivatives is a tool to estimate the population in the area served by WWTPs. In the A2O effluent, concentrations of caffeine derivatives ranged from 10.2 to 171.3 ng/L, indicating the removal efficiencies of caffeine derivatives were effective (above 95%) except in the case of xanthine.

Acknowledgements

This study was supported by the Environment Research and Technology Development Fund (ERTDF) (5-1653) of the Ministry of the Environment, Japan. We also thank Kyoto City Waterworks Bureau for providing the feed water and the technical support required for this project.

Supporting Information

Optimization of selected conditions and overall LC-MS/MS conditions. This material is available free of charge on the Web at http://www.jsac.or.jp/analsci/.

References

1. C. Vatovec, P. Phillips, E. Van Wagoner, T. M. Scott, and E. Furlong, Sci. Total Environ., 2016, 572, 906.

2. Q. Sui, J. Huang, S. Deng, G. Yu, and Q. Fan, Water Res., 2010, 44, 417.

3. L. Bijlsma, J. V. Sancho, E. Pitarch, M. Ibáñez, and F. Hernández, J. Chromatogr. A, 2009, 1216, 3078.

4. K. He, S. Echigo, and S. Itoh, Sci. Total Environ., 2016, 565, 672.

5. T. A. Ternes, M. Bonerz, and T. Schimdt, J. Chromatogr. A, 2001, 938, 175.

6. D. W. Kolpin, E. T. Furlong, M. T. Meyer, E. M. Thurman, S. D. Zaugg, L. B. Barber, and H. T. Buxton, Environ. Sci. Technol., 2002, 36, 1202.

7. S. Franke, S. Hildebrandt, J. Schwarzbauer, M. Link, and W. Francke, Fresenius, J. Anal. Chem., 1995, 353, 39.

8. S. Ibrahim, M. Y. Shukor, M. A. Syed, A. Arina, N. Rahman, K. Abdul Khalil, and S. A. Ahmad, Asian J. Plant Sci., 2014, 2, 18.

9. S. S. Dash and S. N. Gummadi, Biotechnol. Lett., 2006, 28, 1993. 10. M. Majewsky, T. Gallé, V. Yargeau, and K. Fischer,

Bioresour. Technol., 2011, 102, 7415. 11. S. Gokulakrishnan, K. Chandraraj, and S. N. Gummadi,

Enzyme Microb. Technol., 2005, 37, 225. 12. A. Z. Tong, A. J. Ghoshdastidar, and S. Fox, Environ. Sci.

Pollut. Res., 2015, 22, 689. 13. M. Huerta-Fontela, M. T. Galceran, J. Martin-Alonso, and

F. Ventura, Sci. Total Environ., 2008, 397, 31. 14. J. W. Kim, S. M. Yoon, S. J. Lee, M. Narumiya, N. Nakada,

I. S. Han, and H. Tanaka, in 2nd International Conference on Environment and Industrial Innovation, 2012, Singapore, 57.

15. P. Segura, A. García-Ac, A. Lajeunesse, D. Ghosh, C. Gagnon, and S. Sauvé, J. Environ. Monit., 2007, 9, 307.

16. A. M. Urtiaga, G. Pérez, R. Ibáñez, and I. Ortiz, Desalination, 2013, 331, 26.

17. M. Kim, P. Guerra, A. Shah, M. Parsa, M. Alaee, and S. A. Smyth, Water Sci. Technol., 2014, 69, 2221.

18. O. Hillebrand, K. Nödler, T. Licha, M. Sauter, and T. Geyer, Water Res., 2012, 46, 395.

19. N. Nakada, K. Komori, Y. Suzuki, C. Konishi, I. Houwa,

and H. Tanaka, Water Sci. Technol., 2007, 56, 133. 20. T. Okuda, N. Yamashita, H. Tanaka, H. Matsukawa, and K.

Tanabe, Environ. Int., 2009, 35, 815. 21. A. S. Ptolemy, E. Tzioumis, A. Thomke, S. Rifai, and M.

Kellogg, J. Chromatogr. B, 2010, 878, 409. 22. T. Ogawa, H. Hattori, M. Iwai, K. Ito, K. Iwakura, A. Ishii,

O. Suzuki, and H. Seno, Forensic Toxicol., 2012, 30, 142. 23. P. M. M. De Kesel, W. E. Lambert, and C. P. Stove, Talanta,

2015, 144, 62. 24. J. Chae, D. Kim, B. Lee, E. J. Kim, and K. Kwon, J.

Chromatogr. B, 2012, 889–890, 44. 25. R. Hao, J. Li, Y. Zhou, S. Cheng, and Y. Zhang, Chemosphere,

2009, 75, 987. 26. E. Morita and E. Nakamura, Anal. Sci., 2011, 27, 489. 27. C. Y. Cheng, Y. C. Wang, and W. H. Ding, Anal. Sci., 2011,

27, 197. 28. S. Castiglioni, K. V. Thomas, B. Kasprzyk-Hordern, L.

Vandam, and P. Griffiths, Sci. Total Environ., 2014, 487, 613. 29. K. V. Thomas and M. J. Reid, Environ. Sci. Technol., 2011,

45, 7611. 30. M. G. Cantwell, D. R. Katz, J. C. Sullivan, T. Borci, and R.

F. Chen, Mar. Pollut. Bull., 2016, 108, 321. 31. I. Senta, E. Gracia-Lor, A. Borsotti, E. Zuccato, and S.

Castiglioni, Water Res., 2015, 74, 23. 32. U. Kotowska, K. Bieganska, and V. Isidorov, Polish J.

Environ., 2012, 21, 129. 33. K. Arai, H. Terashima, S. Aizawa, A. Taga, A. Yamamoto,

K. Tsutsumiuchi, and S. Kodama, Anal. Sci., 2015, 31, 831. 34. H. Nishi and K. Nagamatsu, Anal. Sci., 2014, 30, 205. 35. Q. Sui, J. Huang, S. B. Deng, and G. Yu, Chin. Sci. Bull.,

2010, 54, 4633. 36. D. C. Harris, “Quantitative Chemical Analysis”, 2010, W.

H. Freeman, New York. 37. L. Tong, P. Li, Y. Wang, and K. Zhu, Chemosphere, 2009,

74, 1090. 38. S. L. Rice and S. Mitra, Anal. Chim. Acta, 2007, 589, 125. 39. W. M. A. W. M. Khalik, M. P. Abdullah, F. K. Baharudin,

and S. A. Zulkepli, J. Mater. Environ. Sci., 2016, 7, 720. 40. M. E. Dasenaki and N. S. Thomaidis, Anal. Bioanal. Chem.,

2015, 407, 4229. 41. F. Gosetti, E. Mazzucco, D. Zampieri, and M. C. Gennaro,

J. Chromatogr. A, 2010, 1217, 3929. 42. K. Hájková, J. Pulkrabová, J. Schurek, J. Hajšlová, J.

Poustka, M. Nápravníková, and V. Kocourek, Anal. Bioanal. Chem., 2007, 387, 1351.

43. J. Radjenovi, A. Jeli, M. Petrovi, and D. Barceló, Anal. Bioanal. Chem., 2009, 393, 1685.

44. B. Li and T. Zhang, Chemosphere, 2011, 83, 1284. 45. L. Dimitra and L. Nollet, “Transformation Products of

Emerging Contaminants in the Environment: Analysis, Processes, Occurrence, Effects and Risks”, 2014, John Wiley and Sons, New York.

46. I. J. Buerge, T. Poiger, H. Buser, and C. Wa, Environ. Sci. Technol., 2003, 37, 691.

47. B. Srdjenovic, V. Djordjevic-Milic, N. Grujic, R. Injac, and Z. Lepojevic, J. Chromatogr. Sci., 2008, 46, 144.

48. Region-case.com, http://www.region-case.com/category/ranking/consume/consumption/, (in Japanese).

49. S. Nicolardi, S. Herrera, M. J. Martinez Bueno, and A. R. Fernandez-Alba, Anal. Methods, 2012, 4, 3364.

50. K. Nödler, T. Licha, K. Bester, and M. Sauter, J. Chromatogr. A, 2010, 1217, 6511.

51. A. Y. C. Lin, C. A. Lin, H. H. Tung, and N. S. Chary, J. Hazard. Mater., 2010, 183, 242.

Documents

Determination of Caffeine and Its Metabolites in