ANALYTICAL SCIENCES FEBRUARY 2007, VOL. 23 135 2007 © The Japan Society for Analytical Chemistry

Rapid Communications

A Rapid Field Detection Method for Arsenic in Drinking Water

Anuradha BAGHEL, Beer SINGH,† Pratibha PANDEY, and K. SEKHAR

Defence Research & Development Establishment, Jhansi Road, Gwalior-474 002, India

In this study, a rapid colorimetric method for arsenic detection was developed. Different reagents containing magnesium turnings in combination with a series of acids were tested for arsine generation. The arsine was then allowed to react with auric chloride on Whatman filter paper No. 3, which in turn changed color. The detection time and detection limit were measured for each acid. Oxalic acid was found to be the most appropriate acid among all the acids used for detection in this study. It took 10 min to detect 10 ppb arsenic concentration and only 1 min to detect concentrations higher than 50 ppb. This method thus reduced the detection time for arsenic and has the potential to develop better field kit.

(Received November 8, 2006; Accepted December 12, 2006; Published February 10, 2007)

Arsenic contamination of drinking water has been prevalent Moreover, both of these methods use toxic mercuric salt as a around the world.1 In India and Bangladesh between forty and color-generating reagent. To alleviate these problems we have eighty million people are at risk of consuming too much arsenic developed an inexpensive, rapid as well as more sensitive from well water, which might have already caused one hundred method for the detection of arsenic. The detection limit for the thousand cancer cases and thousands of deaths. Many millions proposed method falls under the safe WHO guideline value of elsewhere in South-East Asia and South America may suffer 0.01 mg/L. The method uses magnesium turnings along with similarily. Arsenic comes through natural as well as oxalic acid as reagents for arsine generation and also the agricultural and industrial sources into the soil.2,3 There is an reduction of auric chloride to metallic gold for color urgent need for the detection and thorough testing of arsenic in generation.18 High stability of gold chloride compared with water at the time of war or in the highly contaminated area as silver halide favored our choice of the reagent. well as areas where relatively elevated levels of arsenic are present in the ground water.4–7 The arsenic contamination of Experimental ground water has been found to adversely affect the human body at a level as low as 0.01 mg/L.8 The maximum General test procedure permissible limit in the drinking water contaminant level for A stock solution of 100 mg/L of arsenic was prepared by 9 arsenic as per WHO is 0.01 mg/L. Arsenic is mainly present in dissolving sodium meta arsenate (Na2HAsO4) in 100 mL of 3– natural water in inorganic trivalent (AsO3 ) and penta-valent water and diluting further for sub-standards. A 1% auric 3– (AsO4 ) oxidation states. There are different analytical chloride (AuCl3) solution was prepared by dissolving 1.0 g of techniques for inorganic arsenic detection, such as atomic auric chloride into 10 mL of aqua regia and then further diluting absorption spectrophotometry (AAS), hydride generation-AAS, it with double-distilled water. Then a 100 mL contaminated inductively coupled plasma, neutron activation analysis, an water sample was taken in a plastic bottle and one of the electrochemical method, colorimetry etc.10 Most of these different acids mentioned in Table 1 was added to it, followed techniques are expensive and require some training to handle by a further addition of one scoopful of magnesium turnings. them efficiently. The most widely used method is the hydride The quality of magnesium turnings and other chemicals was generation based-AAS method because it is highly sensitive and regulated by blank test analysis. A drop of auric chloride selective. solution was placed on Whatman filter paper No. 3, and was A number of field kits for arsenic detection based on the then kept on a plastic bottle so as to quickly expose the side of generation of volatile arsine to separate arsenic from other filter paper treated with auric chloride to the arsine gas possible interferents present in the sample are available.11 One generated from the above-mentioned solution, which in turn such example is the Gutzheit Method, in which arsine is changes color. generated using zinc and hydrochloric acid.12–14 This method The commonly proposed reaction mechanism is based on the has a lower detection limit of 0.1 mg/L for arsenic. A generation of volatile arsine gas during the reaction of As(III) or ubiquitous presence of arsenic in zinc and sometimes even in As(V) present in contaminated water with magnesium and acid. hydrochloric acid, may lead to a false positive test of arsenic. This arsine gas subsequently came in contact with auric chloride The handling and transportation of hydrochloric acid is also soaked filter paper and reduced the auric chloride to metallic problematic. The original first generation Merck kit (based on gold.18 This reaction resulted in color generation ranging from Gutzheit Method) widely used for the routine analysis of dull pink to violet on the filter paper. The intensity of color drinking water in Bangladesh had a poor detection level of 0.1 depends on concentration of arsenic in water sample according to mg/L. The Hach method is another such method, which uses 15,16 3– 3– 0 solid sulfamic acid in place of HCl. The acidity of sulfamic AsO3 or AsO4 + Mg + (COOH)2 ⎯→ AsH3 (g) 17 0 acid is less and the detection time is longer than HCl. AsH3 (g) + AuCl3 + H2O ⎯→ Au + HCl + H3AsO3 Metallic gold (pink-violet) † To whom correspondence should be addressed. 136 ANALYTICAL SCIENCES FEBRUARY 2007, VOL. 23

Table 1 Screening of acids for arsenic detection at room temperature

AcidDetection time/min Arsenic concentration/mg L–1 Blank test Color

Boric acid 1.0 mg 15 – 20 0.2 No Color No color L-Tartaric acid 1.0 mg 20.0 0.2 No Color No color Nitric acid (90 – 100˚C) 2.0 mL 5.0 0.2 No Color Dull pink Hydrochloric acid 2.0 mL 5.0 0.05 No Color Dull pink Sulfuric acids 2.0 mL 10.0 0.05 No Color Pink-violet Chloroacetic acid 1.0 mg 15.0 0.2 No Color Dull violet Perchloric acid 2.0 mL 15.0 0.2 No Color Violet color Formic acid 2.0 mL 15.0 0.2 No Color Violet color Sulfamic acid 1.0 mg 7.0 0.2 No Color Violet color Oxalic acid 1.0 mg 1.0 > 0.05 No Color Sharp pink-violet Oxalic acid 1.0 mg 5.0 > 0.01 to 0.05 No Color Pink-violet Oxalic acid 1.0 mg 10.0 0.01 No Color Pink-violet

Fig. 1 EDX of AuCl3 on filter paper. Fig. 2 EDX of gold precipitate obtained from positive test in 1% AuCl3 solution.

SEM-EDX analysis To substantiate the proposed reaction scheme, energy give a positive test. Though sulfuric acid gives a detectable dispersive X-ray analysis (Quanta 400-ESEM with EDAX-FEI, pink-violet color, it is a corrosive acid in the liquid state and is Netherlands), is carried out. The EDX spectra of a dried 1% therefore unsuitable for handling and designing a field kit.

AuCl3 solution as well as the precipitate formed by the reaction Hydrochloric acid is also good for this test, but is not suitable of a AuCl3 solution with arsine were collected and used for this for field purposes due to the poor sensitivity (dull color study. For obtaining the gold precipitate, arsine gas generated generation). Moreover, it shares shortcomings, like a liquid from the reaction mixture (arsenic, magnesium turnings and state and corrosiveness with sulfuric acid. Boric acid and L- oxalic acid) in a plastic bottle was passed through a 1% AuCl3 tartaric acid (weak acid) are not able to produce color even in 15 solution, which was kept in another vial. Figure 1 shows the – 20 min with a 0.2 mg/L contaminant concentration. Although

EDX analysis spectra of AuCl3 used in the present study. The formic acid produces intense color, it is hazardous due to health inset table shows the percentage ratio of gold to chlorine (Au because of being highly corrosive to the skin as well as 25.9% and Cl 74.1%), which is very close to the expected ratio unsuitable for transport. The present improvised detection of 1:3. Figure 2 shows the EDX analysis spectra of the method involves oxalic acid as the main reagents to produce precipitate. The inset table shows the composition of the arsine along with magnesium turnings. Oxalic acid is one of the precipitate that is 93.6% gold and 6.4% chlorine. This is strongest organic acid that occurs in the solid state. Oxalic acid probably due to the reduction of AuCl3 to metallic gold based can be converted in the form of a tablet (1.0 mg), suitable for a on the above-mentioned reaction scheme. The presence of practical field kit to apply tests to drinking water in rural and 6.4% chlorine in the precipitate is apparently a contribution remote areas. Oxalic acid acts as a chelating agent/bidentet form the unreacted AuCl3. ligand for magnesium, which gives a facile reaction with magnesium, turning to generate hydrogen. An oxalic acid tablet Results and Discussion and magnesium turnings-based field kit will avoid the wastage of reagents, and will be easy to handle. Other benefits of this Table 1 shows the results of experiments carried out in the test are the use of auric chloride as a color-producing reagent, presence of different acids. The color of AuCl3 soaked filter unlike other methods that use silver or mercury salts. Because paper changed the original light yellow to dull pink (nitric and gold chloride is environmentally more stable compared with hydrochloric acid) or pink-violet (sulfuric and oxalic acid) or silver halides (sensitive to light) and unlike mercury salts it is violet (perchloric, formic and sulfamic acid), depending upon nontoxic. the acid used in the reaction mixture. The detection time, When compared with the kits given in Table 2, the current detection limit and color generated for various acids are given in oxalic acid based system appears to be better in terms of Table 1. Table 1 also indicates that the test with nitric acid sensitivity, detection limit, ease of handling and stability/non- requires heating to a temperature in the range of 90 – 100˚C to toxicity of color-generating reagent (gold chloride vs. silver ANALYTICAL SCIENCES FEBRUARY 2007, VOL. 23 137

Table 2 Comparison of the present method with available kits for arsenic detection Detection Detection Method Reagent limit/mg L–1 time/min

Merck kit Zinc, hydrochloric acid, 0.1 10 – 30a mercury salt of chloride or bromide Merck Zinc, hydrochloric acid, 0.01 10 – 30a (improved) mercury salt of chloride or bromide Hach Zinc, sulfamic acid, 0.01 10 – 30a mercury salt Oxalic acid Magnesium turnings, > 0.05 1.0 Fig. 3 AuCl3 soaked filter paper (blank test). oxalic acid, gold chloride Oxalic acid Magnesium turnings, 0.01 – 0.05 5.0 oxalic acid, gold chloride Oxalic acid Magnesium turnings, 0.01 10.0 oxalic acid, gold chloride a. See Ref. 19.

halide and mercuric halides). The present method gives a very sharp pink-violet color in 1 min at arsenic concentration higher than 0.05 mg/L, pink-violet color in 5 min in the concentration range of arsenic >0.01 – 0.05 mg/L and pink-violet color in 10 min in the range of 0.01 mg/L. The intensity of color generated on filter paper in this reaction is found to be dependent on the amount of metallic gold generated in the reaction and in turn on the arsenic concentration in the water sample. Figures 3 and 4 Fig. 4 Pink-violet filter paper (positive test for detection of As). indicate a colored photograph of AuCl3 soaked filter paper (blank test) and pink-violet filter paper (positive test for detection of As), respectively. Hence, it is clear from present study that oxalic acid-based Exposure and Health Effects”, 1997, Chapman & Hall, arsine generation in combination with auric chloride reduction New York, 112. to metallic gold is a better suited method for the preparation of a 7. B. K. Biswas, R. K. Dhar, G. Samanta, B. K. Mandal, D. field kit when compared with available kits. This method gives Chakraborti, S. Roy, A. Jafar, A. Islam, G. Ara, S. Kabir, a sharp color without any device, such as an arsenator. An A. W. Khan, A. Ahmed, and S. A. Hadi, Curr. Sci., 1997, added advantage of this test is that it is more cost effective than 73, 48. the other colorimetric methods. This method is relatively fast 8. P. L. Smedley and D. G. Kinniburg, Appl. Geochem., 2002, and easy and can work under harsh climatic conditions, such as 17, 517. –20 to +50˚C. Further work is going on in our lab for the 9. WHO, “Guidelines for Drinking-Water Quality: evaluation of other carboxylic acids to establish their suitability Recommendations”, 2nd ed., 1993, Vol. 1, World Health for a practical arsenic detection field kit. Organization, Geneva. 10. a) D. Q. Hung, O. Nekrassova, and R. G. Compton, References Talanta, 2004, 64, 269. b) R. K. Dhar, Y. Zheng, J. Rubenstone, and A. van Green, Anal. Chim. Acta, 2004, 1. National Research Council, “Arsenic in Drinking Water”, 526, 203. c) K. Anezaki, I. Nukatsuka, and K. Ohzeki, 1999, National Academy Press, Washington, D. C. Anal. Sci., 1999, 15, 829. d) S. Hirata, H. Toshimitsu, and 2. A. Brandstetter, E. Lombi, W. W. Wenzel, and D. C. M. Aihara, Anal. Sci., 2006, 2, 39. Adriano, “Arsenic Contaminated Soils: I. Risk Assessment”, 11. K. J. Irgolic, in “Arsenic: Exposure and Health, Science in “Remediation Engineering of Contaminated Soils”, ed. and Technology Letters”, ed. W. R. Chapell, C. O. Abernathy, D. L. Wise, D. J. Trantolo, E. J. Cichon, H. I. Inyang, and and C. R. Cothern, 1994, Northwood, UK, 51 – 60. U. Stottmeister, 2000, Chap. 33, Dekker, New York, 715. 12. H. Gutzeit, Pharmaz. Zeitung, 1879, 24, 263. 3. M. Leist, R. J. Casey, and D. Caridi, J. Hazard. Mater. B, 13. P. Lohmann, Pharmaz. Zeitung, 1891, 36, 756. 2000, 76, 125. 14. C. B. Sanger and O. F. Black, J. Soc. Chem. Ind., 1907, 26, 4. D. Das, G. Samanta, B. K. Mandal, C. R. Chowdhury, P. P. 1115. Chowdhury, G. K. Basu, and D. Chakraborti, Environ. 15. D. Kroll, Paper presented at the 3rd NSF International Geochem. Health, 1996, 18, 5. Symposium on Small Drinking Water and Wastewater 5. T. R. Chowdhury, B. K. Mondal, G. Samanta, G. K. Basu, Systems, Washington, D. C., 2001. P. P. Chowdhury, C. R. Chanda, N. K. Karan, D. Lodh, D. 16. Application Note 124, HACH Company, USA. Das, K. C. Saha, and D. Chakroborti, in “Arsenic Exposure 17. D. G. Kinniburg and W. Kosmus, Talanta, 2002, 58, 165. and Health Effects”, 1997, Chapman & Hall, New York, 93. 18. F. Feigl, “Spot Test Inorganic Applications”, 1954, Vol. 1, 6. D. N. G. Mazumder, J. Dasgupta, A. Santra, A. Pal, A. Elsevier Publishing Company, Amsterdam, 97. Ghose, N. Chattopadhaya, and D. Chakraborti, in “Arsenic 19. I. Jaunakais, Int. Environ. Technol., 2002, 12, 2.