Simultaneous Electrodialytic Preconcentration and Speciation of

Oct 28, 2015 - Department of Chemistry and Biochemistry, The University of Texas at Arlington, Arlington, Texas 76019-0065, United States. Anal. Chem...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Simultaneous Electrodialytic Preconcentration and Speciation of Chromium(III) and Chromium(VI) Shin-Ichi Ohira,*,† Koretaka Nakamura,† C. Phillip Shelor,‡ Purnendu K. Dasgupta,‡ and Kei Toda† †

Department of Chemistry, Kumamoto University, Kumamoto, 860-8555, Japan Department of Chemistry and Biochemistry, The University of Texas at Arlington, Arlington, Texas 76019-0065, United States

‡

S Supporting Information *

ABSTRACT: Large amounts of chromium (Cr) compounds are used for manufacturing of various products and various chemical processes. Some inevitably find their way into the environment. Environmental Cr is dominantly inorganic and is either in the cationic +3 oxidation state or in the anionic oxochromium +6 oxidation state. The two differ dramatically in their implications; Cr(III) is essential to human nutrition and even sold as a supplement, while Cr(VI) is a potent carcinogen. Drinking water standards for chromium may be based on total Cr or Cr(VI) only. Thus, Cr speciation analysis is very important. Despite their high sensitivity, atomic spectrometric techniques or induction coupled plasma−mass spectrometry (ICP-MS) cannot directly differentiate the oxidation states. We present here a new electrodialytic separation concept. Sample analyte ions are quantitatively transferred via appropriately ionically functionalized dialysis membranes into individual receptors that are introduced into the ICP-MS. There was no significant conversion of Cr(VI) to Cr(III) or vice versa during the very short (6 s) separation process. Effects of salinity (up to ∼20 mM NaCl) can be eliminated with proper membrane functionalization and receptor optimization. With the ICP-MS detector we used, the limits of detection for either form of Cr was 0.1 μg/L without preconcentration. Up to 10-fold preconcentration was readily possible by increasing the donor solution flow rate relative to the acceptor solution flow rates. The proposed approach permits simultaneous matrix isolation, preconcentration, and chromium speciation. hromium (Cr) exists in oxidation states from −2 to +6 and was so named because of the many colors in its different oxidation states. Chromium is an essential engineering metal; a very large amount is used and processed annually worldwide with a considerable amount finding its way into the environment. Historically, annual release amounts may be as large as 1 million Mton Cr.1 Chromium is an essential ingredient of stainless steel; Cr is released into food from stainless cookware.2 Chromium salts are extensively used in the tanning of leather. Residual Cr in leatherwear can be oxidized to Cr(VI) by peroxides present in skin oils.3 In environmental matrices, such as water,4,5 atmospheric particulate matter,6,7 and so on, Cr is largely inorganic, present in the +3 or +6 oxidation states.8,9 In glucose and lipid metabolism, cationic Cr(III) plays a critical role and is an essential nutritional element. In contrast, anionic Cr(VI) (that is present as dinuclear Cr2O72− in strongly acidic solutions and as CrO42− in alkaline solutions) readily crosses cell membranes and is highly toxic due to the high oxidation potential especially at the lower intracellular pH; it is highly carcinogenic.10 Whereas in some cases total Cr is the regulated parameter,11 often it is Cr(VI) that is subject to more stringent regulations; allowable drinking water maximum levels are 10 and 50 μg/L in the State of California12 and Japan, respectively.13 Widespread health problems in a California community stemming from

C

© XXXX American Chemical Society

Cr(VI) in drinking water became the basis of a major Oscarwinning motion picture.14 Environmental conversion of Cr oxidation states is mediated by light,15 solution pH and Eh,16 arsenic and manganese species,17,18 and so on. Cr contamination in natural and drinking water is routinely monitored for public health protection, and there are many ongoing studies that speciate environmental Cr at trace levels. Diphenylcarbazide (DPC) selectively reacts with Cr(VI)19,20 forming a colored product, the reaction mechanism or the final product structure has never been unequivocally established. The high sensitivity of the method has led to wide use, not only with conventional spectrophotometry, but also with solid-phase spectrometry,21,22 liquid core waveguide spectrometry,3 postcolumn reaction detection following chromatography,23 flow injection analysis, and so on.24 Field applicability21 and sub μg/L LODs22 are possible; however, DPC can only measure Cr(VI), oxidizing Cr(III) to Cr(VI) involves multistep chemical23 or radiative25 manipulations. Because of the need to measure low levels, atomic spectrometry and induction coupled plasma-mass spectrometry Received: September 11, 2015 Accepted: October 28, 2015

A

DOI: 10.1021/acs.analchem.5b03464 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

electrode contact and electrolytic gases. The cathode and anode isolator compartments were separated from the respective acceptors by anion and cation exchange membranes, respectively (Selemion DSV and CMV, www.agec.co.jp). The sample channel was separated from the cathode and anode side acceptors by dialysis membranes. If specifically stated, the cathodic and anodic dialysis membranes were respectively functionalized with 25% sodium vinylsulfonate +2 mM H2O241 and N,N-dimethylaminoethyl methacrylate.38 Each channel was 40 mm long, 5 mm wide, and 0.13 mm in depth. Analysis System. A more detailed system diagram is shown in Figure S1. Peristaltic pumps with plasticized polypropylene tubes were used throughout. Isolator solutions (30 mM HNO3) and the sample flows were 850 and 200 μL/min, respectively. The operational flow through the acceptor channels, served by selection valve equipped syringe pumps was 10 mM HNO3 @ 200 μL/min with various DC voltages applied (Vapp), as discussed in detail later. For cleaning between samples, Vapp = 15 VDC and acceptor flows of 100 mM HNO3 @ 0.63 mL/min was used for 45 s. Sample loops (20 μL) of two electrically actuated six-port PEEK valves were filled by the acceptor effluent and periodically injected into the ICP-MS (X series II or Elements, both from www.thermofisher.com) by a 10 mM HNO3 carrier at 200 μL/min, according to the time sequence in Figure S2 in the Supporting Information. The Cr signal was monitored at m/z 52. Monitoring at m/z 53 produced equivalent results. Some experiments were carried out by collecting the acceptor effluent and offline measurements with a flame atomic absorption spectrometer (AAS, AAnalyst 100, www.perkinelmer.com).

(ICP-MS) are often used for trace chromium determination. For oxidation state speciation, chromatographic preseparation is most common.26 Both Cr species have been separated by anion chromatography after chelation of Cr(III) with ethylenediaminetetraacetate.27 The many efforts that exploit the different charge type of Cr(III) and Cr(VI) have been reviewed.28 Cation and anion exchange resins are typically used, but highly selective chelating resins for Cr(III)29 and Cr(VI)30 have been purpose-designed. Biological materials such as cells31 or egg-shell membrane32 have also been used for selective absorption of chromium species. Preconcentration is also possible but sample pH and ionic strength affect recoveries. These pitfalls can be avoided by measuring Cr(VI) after preremoval of Cr(III) by carrier coprecipitation; Dy(OH)333 or Yb(OH)334 has been recommended as the carrier. For samples with specific conductance 1.8 decreased the Cr(III) transfer efficiency in the present system. The acceptor solution pH changes are different when the standard is in pure water compared to an electrolyte solution, because the nature of the ions transported to the acceptor streams change. In addition, in the case of a dilute aqueous

standard as a sample, the sample channel represents the highest resistance in the serially connected channels. At a fixed Vapp, the current increases when this limiting resistance is decreased, increasing the production of OH− in the cathode channel, and possibly increasing the pH locally at the acceptor-side membrane surface separating the cathode side acceptor and the cathode. We explored therefore whether the loss of transfer efficiency for high conductance samples can be compensated for by an internal standard. Adding an isotopic standard is possible; 50Cr has a natural abundance of 4.3% (54Cr has a 2.4% abundance but interference from 54Fe is likely). Another possibility is to use Y(III) as an internal standard because both the solubility product of the trihydroxide and its ionization potential (IP) are very close to that of Cr (Ksp 6.3 × 10−31 and 5.0 × 10−31 for Y(OH)3 and Cr(OH)3, IP 6.77 and 6.38 eV, respectively). The results indicated that the ratio Cr(III)/Y(III) was stable up to ∼10 mM NaCl; fortuitously, this was the same salinity up to which Cr(VI), for which such an internal standard was not explored, was near-quantitatively transferred. As a matter of perspective, according to the USEPA, >75% of 989 drinking water sources examined in the U.S. had a sodium content below 2.2 mM.44 Dealing with High Salinity Samples by Tuning Isolator Composition. The purpose of the isolator channel is to isolate the acceptor from electrolytic gas bubbles and direct redox conversions at the electrode. Using acids as isolator obviously decreases the potential of drastic pH increases in the cathode acceptor. Effects of changing Vapp was studied with 1, 10, and 30 mM HNO3 as isolator using 2 mg/L of Cr(III) and Cr(VI) each in 1 mM NaCl (Figure S4). In the anode acceptor, [Cr(VI)] increased with Vapp and was quantitatively transferred by 7.5 V with 10 and 30 mM HNO3 isolators, while >20 V was required with the 1 mM HNO3 isolator. In the cathode acceptor, [Cr(III)] at first increased and then decreased with increasing Vapp. For 10 and 30 mM HNO3 highest, nearquantitative transfer was observed at Vapp = 7.5 V as well; poor transfer resulted with 1 mM HNO3 at all Vapp values. As previously suspected, the decrease in apparent [Cr(III)] in the cathode acceptor is the dramatic pH increase in this channel at C

DOI: 10.1021/acs.analchem.5b03464 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry respective Vapp transitions of 7.5 → 10 V, 10 → 12.5 V, and 12.5 → 15 V for 1, 10, and 30 mM HNO3 isolator solutions, respectively. Note that cationic Cr(III) can not only precipitate as Cr(OH)3, at high enough pH it can form anionic hydroxo complexes that can then migrate backward. Dealing with High Salinity Samples by Tuning Membrane Functionalization. Effects of sample ionic strength were further studied with 10 and 30 mM HNO3 as isolator solutions (Figure 4a). While the adverse effect of

Figure 5. Response charts with standards (a) Cr(III) only (blue trace), Cr(VI) only (red trace), and mixtures of these (purple trace); (b) output from trace level Cr(III) and Cr(VI) mixtures; (c) rapid Cr(VI) analysis mode, the solution introduced into the sample channel was switched from blank to sample to blank as indicated by the dashed line.

acceptor solutions and separately injected into the ICP-MS carrier stream. As the two injection valves are connected with minimum tubing lengths and injector timing is automated (with a 60 s full cycle time), analyte dispersion was minimized, and both Cr(III) and Cr(VI) results were obtained within a minute. It is also possible to measure Cr(VI) only, either in a continuous mode or with discrete injections 30 s apart (Figure 5c). The peak area and relative standard deviation for 10 μg Cr(VI)/L were 760 ± 29 kcps·s and 3.8% (n = 18), respectively. The limits of detection based on three times the standard deviation for the blank criteria were 0.08 and 0.09 μg/L for Cr(VI) and Cr(III), respectively. The origin of the blank response is not presently clear; residual membrane contamination or polyatomic interference45 is possible. Calibration curves appear in Figure S5, the relevant equations were

Figure 4. Transfer efficiencies for high salt samples with (a) commercial and (b) functionalized cellulose ion transfer membrane. Vapp 7.5 V, sample and acceptor solution flow rates: 0.2 mL/min.

salinity was decidedly lower with the higher [HNO3] isolators, Cr(III) transfer efficiency was still perceptibly subquantitative, even at a salinity of 10 mM, which we set as the benchmark to meet for general applicability to drinking water. As described, the cathode acceptor pH increases from cathodically generated hydroxide. Putting HNO3 in the cathode isolator inhibits this pH increase but nitrate ions present in the cathode isolator and acceptor electromigrate into the sample channel, increasing its specific conductance and lowering transfer efficiency. Instead of exotic internal standards, functionalization of the dialysis membranes that define the sample channel boundaries between the cation and anion acceptor channels was explored to reduce the nitrate transfer extent. When the original commercial dialysis membranes were replaced by the membranes functionalized as described in Materials and Methods, nitrate concentration in the sample channel decreased by a factor of 2. Now the transfer efficiency of Cr(III) for 30 mM HNO3 isolators was quantitative up to 30 mM NaCl in the sample and was near-quantitative even with 100 mM NaCl in the samples (Figure 4b). Performance. The optimized ITD configuration was, thus,

Cr(VI) peak area, kcps·s = (67.2 ± 1.87) × [Cr(VI), μg/L] + (94.5 ± 23.4) r 2 = 0.9953 Cr(III) peak area, kcps·s = (67.2 ± 0.97) × [Cr(III), μg/L] + (74.6 ± 12.2) r 2 = 0.9995

Note identical calibration slopes for the two analytes. A common problem required in speciation analysis is interconversion.46 The residence time of the sample in the ITD is ∼6 s; this minimizes any such opportunity. As accurate quantitation of Cr(VI) is more important, we studied the effect of adding 50 μg/L Cr(III) to 1 μg/L Cr(VI), Figure S6. The peak areas for Cr(VI) with and without Cr(III) addition were, respectively, 168.2 ± 6.1, and 163.0 ± 5.1 (n = 3 ea.), indicating no difference in the [Cr(VI)] measured. Drinking Water Analysis. Several tap and bottled water samples were analyzed with and without standard addition. The concentrations were calculated based on an external calibration method. Recoveries are indicated in Table 1 and ranged from 95.8 to 106%. The total Cr in some of the samples was directly measured by direct introduction into the ICP-MS; this agreed

+|30mMHNO3|CEM|10mM HNO3|anex dialysis membrane|sample| catex dialysis membrane|10mMHNO3|AEM|30mM HNO3| − ,

with Vapp = 7.5 V. Representative system output appears in Figure 5. In chromatographic speciation, various column and eluent parameters govern the details of the separation. With use, retention times shift and MS detection windows must be adjusted. Here, operation is far more robust. The analytes are completely separated and transferred into altogether different D

DOI: 10.1021/acs.analchem.5b03464 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry Table 1. Drinking Water Analysis with the Present System added concn, μg/L tap water A B bottled water A B C D E a

found, μg/L

ICP-MS direct

recovery, %

Cr(III)

Cr(VI)

Cr(III)

Cr(VI)

total Cr, μg/L

Cr(III)

Cr(VI)

0 0 1

0 0 1

0.59 ± 0.02 NDa 0.98 ± 0.02

1.40 ± 0.03 ND 0.96 ± 0.07

2.00 ± 0.39 b 

  97.5

  95.7

0 0 0 1 0 1 0 1

0 0 0 1 0 1 0 1

2.04 0.70 ND 1.06 ND 0.94 ND 0.96

± 0.03 ± 0.02

4.60 ± 0.93 1.68 ± 0.28      

   106  94.1  95.8

   105  97.9  101

± 0.09 ± 0.08

2.73 0.96 ND 1.05 ND 0.98 ND 1.01

± 0.06 ± 0.02 ± 0.03

± 0.04 ± 0.03 ± 0.07

Not detected. bNot available.

consumables like SPE cartridges or chromatographic columns. Unlike SPE based systems, it is easily automated. A comparison of various methods for Cr speciation analysis is summarized in Table S2 in the Supporting Information. The present method required less pretreatment, analysis time and sample volume. Similar approaches should be possible to separate cationic and anionic forms of other elements, for example, V, Mn, and so on. Functionalized membranes used in this work were used continuously for more than a month. Similar membranes have previously been used for the determination of organic acids and lasted for at least 150 wine sample injections. We expect the membranes can be used much longer for water samples.

very well with the sum of the [Cr(III)] and [Cr(VI)] values obtained with the complete system. Potential Interferences. The effects of the other ions were studied. In the presence of 5 mg/L ea. Mn(II), Co(II), Ni(II), Cu(II), Zn(II), Cd(III), Fe(II), and Fe(III) together, Cr(III) was transferred with 99+% efficiency. Similarly there was no effect of 5 mg/L permanganate; Cr(VI) was transferred with 99+% efficiency. Simultaneous Enrichment and Separation. As with SPE approaches, the present ITD method also permits matrix isolation, separation of the two species, and enrichment. Enrichment occurs whenever the flow rates of the respective acceptor solutions are smaller than that of the sample solution. To demonstrate applicability to detection methods other than ICP-MS that may be more affordable, we used a flow injection analysis (FIA) system with DPC chemistry. The Cr(III) effluent from the ITD was oxidized with buffered hydrogen peroxide prior to DPC detection.20 A mixture of 1 μg/L (ppb) each of Cr(VI) and Cr(III) were introduced into the ITD at 200 μL/min while the acceptors, flowed at 20 μL/min, corresponding to 10× enrichment (Figure 6). The acceptor

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b03464. Additional information, as noted in the text (PDF).

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This study was supported partially by the Japan Science and Technology, Development of Systems and Technology for Advanced Measurement and Analysis program (SENTAN) and partially by the Steel Foundation for Environmental Protection Technology. The participation of P.K.D. and C.P.S. was made possible by the Hamish Small Endowment at the University of Texas at Arlington and by the National Science Foundation through CHE-1506572.

Figure 6. Response of ITD-FIA-colorimetric detection system using diphenylcarbazide chemistry with and without preconcentration.

solutions filled the sample loops of a FIA system and were injected separately into a carrier stream of buffered H2O2 (pH 8.5) then mixed with DPC solution (Figure S7). The product was monitored at 520 nm. The calibration slopes (Figure S8) with and without enrichment were 4.1 ± 0.1 and 0.42 ± 0.01 mAU/ppb Cr, respectively, reflecting the expected 10x gain in sensitivity. In summary, electrodialytic separation of Cr(III) and Cr(VI) offers a simple and rapid approach that does not require other

■

REFERENCES

(1) Johnson, J.; Schewel, L.; Graedel, T. E. Environ. Sci. Technol. 2006, 40, 7060−7069. (2) Kamerud, K. L.; Hobbie, K. A.; Anderson, K. A. J. Agric. Food Chem. 2013, 61, 9495−9501. E

DOI: 10.1021/acs.analchem.5b03464 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry (3) Jambunathan, S.; Dasgupta, P. K. J. Soc. Leather Technol. Chem. 2000, 84, 63−73. (4) Zhitkovich, A. Chem. Res. Toxicol. 2011, 24, 1617−1629. (5) Lindsay, D. R.; Farley, K. J.; Carbonaro, R. F. J. Environ. Monit. 2012, 14, 1789−1797. (6) Amouei Torkmahalleh, M.; Lin, L.; Holsen, T. M.; Rasmussen, D. H.; Hopke, P. K. Atmos. Environ. 2013, 71, 92−94. (7) Amouei Torkmahalleh, M.; Lin, L.; Holsen, T. M.; Rasmussen, D. H.; Hopke, P. K. Aerosol Sci. Technol. 2012, 46, 690−696. (8) Kotaś, J.; Stasicka, Z. Environ. Pollut. 2000, 107, 263−283. (9) Gómez, V.; Callao, M. P. TrAC, Trends Anal. Chem. 2006, 25, 1006−1015. (10) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry, 5th ed.; Wiley: New York, 1988. (11) International Agency for Research on Cancer. Monograph 100C. Chromium (VI) Compounds. http://monographs.iarc.fr/ENG/ Monographs/vol100C/mono100C-9.pdf. (12) California Environmental Protection Agency. State Water Resources Control Board. Chromium 6 drinking water MCL. http:// www.waterboards.ca.gov/drinking_water/certlic/drinkingwater/ Chromium6.shtml. (13) Wakayama, H. Revision of Drinking Water Quality Standards in Japan. http://www.nilim.go.jp/lab/bcg/siryou/tnn/tnn0264pdf/ ks0264011.pdf. (14) Soderberg, S., Director. Erin Brockovich; Universal Pictures and Columbia Pictures International: California, 2000. (15) Kaczynski, S. E.; Kieber, R. J. Environ. Sci. Technol. 1993, 27, 1572−1576. (16) Saputro, S.; Yoshimura, K.; Matsuoka, S.; Takehara, K.; Narsito; Aizawa, J.; Tennichi, Y. Chem. Geol. 2014, 364, 33−41. (17) Wang, Z.; Bush, R. T.; Sullivan, L. A.; Liu, J. Environ. Sci. Technol. 2013, 47, 6486−6492. (18) Namgung, S.; Kwon, M. J.; Qafoku, N. P.; Lee, G. Environ. Sci. Technol. 2014, 48, 10760−10768. (19) Moulin, A. Bull. Soc. Chim. 1904, 31, 295−296. (20) Pflaum, R. T.; Howick, L. C. J. Am. Chem. Soc. 1956, 78, 4862− 4866. (21) Saputro, S.; Yoshimura, K.; Matsuoka, S.; Takehara, K.; Narsito. Anal. Sci. 2009, 25, 1445−1450. (22) Sareqiqige; Saputro, S.; Kai, S.; Satoda, M.; Matsuoka, S.; Yoshimura, K. Anal. Sci. 2013, 29, 677−680. (23) Thomas, D. H.; Rohrer, J. S.; Jackson, P. E.; Pak, T.; Scott, J. N. J. Chromatogr. A 2002, 956, 255−259. (24) Samanta, G.; Boring, C. B.; Dasgupta, P. K. Anal. Chem. 2001, 73, 2034−2040. (25) Taguchi, S.; Fujii, A.; Okoshi, T.; Kuramitz, H.; Hata, N. Bunseki Kagaku 2010, 59, 1133−1136. (26) Ščančar, J.; Milačič, R. J. Anal. At. Spectrom. 2014, 29, 427−443. (27) Inoue, Y.; Sakai, T.; Kumagai, H. J. Chromatogr. A 1995, 706, 127−136. (28) Gonzalvez, A.; Cervera, M. L.; Armenta, S.; de la Guardia, M. Anal. Chim. Acta 2009, 636, 129−157. (29) Tokalıoğlu, Ş.; Arsav, S.; Delibaş, A.; Soykan, C. Anal. Chim. Acta 2009, 645, 36−41. (30) Furusho, Y.; Ono, M.; Yamada, M.; Kitade, T.; Motomizu, S. Bunseki Kagaku 2009, 58, 147−152. (31) Zou, A.; Tang, X.; Chen, M.; Wang, J. Spectrochim. Acta, Part B 2008, 63, 607−611. (32) Zou, A.; Chen, X.; Chen, M.; Wang, J. J. Anal. At. Spectrom. 2008, 23, 412−415. (33) Karatepe, A.; Korkmaz, E.; Soylak, M.; Elci, L. J. Hazard. Mater. 2010, 173, 433−437. (34) Duran, A.; Tuzen, M.; Soylak, M. Food Chem. Toxicol. 2011, 49, 1633−1637. (35) He, Y.; Cervera, M. L.; Pastor, A.; de la Guardia, M. Anal. Chim. Acta 2001, 447, 135−142. (36) Safari, M.; Nojavan, S.; Davarani, S. S. H.; Morteza-Najarian, A. Anal. Chim. Acta 2013, 789, 58−64.

(37) Ohira, S.; Kuhara, K.; Kudo, M.; Kodama, Y.; Dasgupta, P. K.; Toda, K. Anal. Chem. 2012, 84, 5421−5426. (38) Ohira, S.; Kuhara, K.; Shigetomi, A.; Yamasaki, T.; Kodama, Y.; Dasgupta, P. K.; Toda, K. J. Chromatogr. A 2014, 1372, 18−24. (39) Ohira, S.; Hiroyama, Y.; Nakamura, K.; Koda, T.; Dasgupta, P. K.; Toda, K. Talanta 2015, 132, 228−233. (40) Nakamura, Y.; Maeda, S.; Nishiyama, H.; Ohira, S.-I.; Dasgupta, P. K.; Toda, K. Anal. Chem. 2015, 87, 6483−6486. (41) Matsumoto, K.; Saitoh, H. Sen'i Gakkaishi 2005, 61, 89−97. (42) Yang, B.; Chen, Y.; Mori, M.; Ohira, S.; Azad, A. K.; Dasgupta, P. K. Anal. Chem. 2010, 82, 951−958. (43) Christian, G. D.; Dasgupta, P. K.; Schug, K. A. Anal. Chem., 7th ed.; Wiley: New York, 2013. (44) United States Environmental Protection Agency. Sodium in drinking water. http://water.epa.gov/scitech/drinkingwater/dws/ccl/ sodium.cfm; Accessed September 1, 2015. (45) Markiewicz, B.; Komorowicz, I.; Sajnóg ; Belter, M.; Barałkiewicz, D. Talanta 2015, 132, 814−828. (46) Chahal, V. K.; Singh, R.; Malik, A. K.; Matysik, F.-M.; Puri, J. K. J. Chromatogr. Sci. 2012, 50, 26−32.

F

DOI: 10.1021/acs.analchem.5b03464 Anal. Chem. XXXX, XXX, XXX−XXX