Literature Cited (1) Junge, C. E., Ryan. T . G., Quart. J . Roy. Meteorol. SOC.,84, 46 (1958). (2) Foster, P. M., Atmos. Environ.. 3, 157 (1969). (3) Thomas, M . D., “Air Pollution,” W.H.O. Monograph Series 46, p p 233-75, Columbia University Press, New York, N.Y., 1961. (4) Burdick, L. R., U.S. Bureau of Mines, Inform. Cir. 7064. 1939. (5) Amdur, M . O., h t . J. Air Poilut., 1,170 (1959). (6) Urone, P., Enriron. Sei. Technol., 3,436 (1967). (7) Junge, C. E., J. Geophys. Res., 65,229 (1960). (8) Neytzell de Wilde, F. G., Taverner, L.. Second UK Int. Conf. on Peaceful Uses of Atomic Energy, Proc., 3,303 (1958). (9) Johnstone, H. F., Coughanowr, D. R., Ind. Enp. Chem., 50, 1169 (1958). (10) Johnstone, H . F., Moll, A. .J., ibid., 52,861 (1960). (11) Karraker, D . G., J . Phys. Chem., 67,871 (1963). (12) Schuck, E. A.. Doyle. G. J., Air Pollut. Found. Rep., p 29, Los Angeles, 1959. (13) Prager. M . .J., Stephens, E . R., Scott, W. E., Ind. Enp. Chem., 6,521 (1960). (14) Wilson, W. E.. Levy. A. J., J. Air Pollut. Contr. Ass., 20, 385 (1970).
(15) Harkins, J., Nicksic, S . W., ibid., 15,218 (1965). (16) Groblicki, P. J., Nebel, G . J., “Chemical Reactions in Urban Atmospheres,’’ C. S. Tuesday, Ed., p 263, American Elsevier Co., 1971. (17) Hoather, R. C., Goodeve. C. F., Trans. Faraday SOC.,30, 626,630,1149 (1934). (18) Freiberg, J., “The Catalytic Oxidation of SO? to Acid Mist in Dispersing Plumes,” PhD thesis. The Johns Hopkins University, 1972. (19) Johnstone, H. F . , Leppla, P. W., J. Amer. Chem. Soc. 56, 2233 (1934). (20) Junge, C., Schleich, G., Atmos. Enuiron., 5 , 165 (1971). (21) Cadle, R. D.. Robbins, R. C . , J . Phys. Chem.. 62,469 (1938). (22) Gartrell, F . E.. Thomas F. W., Carpenter, S. B., Amer. Ind. Hyg. Assoc. J . , 24, 113 (1963). (23) Stephens, E. T., McCaldin, R., Enciron. Sei. Technol.. 5, 615 (1971). (24) Berger, A. W , , Billings, C. E., e t al.. “Study of Reactions of Sulfur in Stack Plumes,” GCA Corp., First Annual Report GCA-TR-68-19G. 1969.
Received for revieu, October 12, 1973. Accepted March 28, 1974
Arsenic in Water by Flameless Atomic Absorption Spectrophotometry Kai C. Tam Freshwater Institute, Department of the Environment, 501 University Crescent, Winnipeg, Man., Canada R3T 2N6
.Arsenic in water is extracted with diethylammonium diethyldithiocarbamate in carbon tetrachloride and determined by atomic absorption spectrophotometry using the carbon rod atomizer. The method will determine arsenate, arsenite, and any organoarsenic compounds soluble in carbon tetrachloride. By using ultraviolet photooxidation to decompose organoarsenic compounds, the method determines total arsenic. No matrix interference is observed. Precision is f 0 . 4 gg/l. at 3.1 gg/l. and the detection limit is 1gg/l.
the rather short wavelength (1937 A) of the most sensitive arsenic resonance line. Although this can partially be overcome by using other types of flame, there seems to be more advantage with flameless techniques, such as thermal decomposition of arsine (12) or graphite furnace (23). Some preliminary work here confirmed the latter authors’ observation that matrix effects could be troublesome, requiring calibration by the standard additions method, for analysis of natural water. However, if arsenic is separated by a chelate-extraction method, interferences are avoided and a useful increase of sensitivity can be obtained.
Instruments Chronic poisoning has been reported to be caused by the utilization of drinking water containing 0.21-10.0 mg/l. of arsenic (As) (1-3). Standards for maximum allowable arsenic concentrations have been established by many agencies. The World Health Organization in 1958 ( 4 ) set a permissible limit of 0.2 mg/l. In 1963 the limit was revised to 0.05 mg/l. T h e United States Public Health Service ( 5 ) has a recommended acceptable concentration of 0.01 mg/l. and a maximum permissible limit of 0.05 mg/l. The Canadian Joint Committee on Drinking Water Standard (6) recommends the same levels. Inorganic arsenic in water is usually analyzed by evolution of arsine (AsH3) and colorimetric determination with silver diethyldithiocarbamate (7, 8) or by molybdenum blue methods, such as that of Johnson ( 9 ) , which require correction for any phosphate present. Kopp and Kroner (10) used emission spectrography. Neutron activation was used by Smales and Pate (11) for seawater. Other techniques such as X-ray fluorescence, polarography, and atomic absorption spectrophotometry (AAS) are potentially useful. There is difficulty with AAS methods if the air-acetylene flame is used because of flame absorption a t 734
Environmental Science & Technology
Means of irradiating samples in fused silica tubes (110ml capacity) a t 3-4 cm from a 450- or 550-W mercury arc lamp A Varian Techtron atomic absorption spectrophotometer (Model AA-5) equipped with carbon rod atomizer (Model 61 or Model 63) and strip chart recorder (Model A-25) using the following settings: AA-5
Wavelength, 1937 A Lamp current, 7 m A
Slit width, 300 p Slit height, 3 mm
C R A 6 1 or 63 Nitrogen flow, 1 I./min (CRA 61) or 4 I./rnin (CRA 63) Water flow. 0.5 I.lmin
Voltage setting Time, sec Range, 5 mV
Dry
Ash
3.5 20
6.25 15
Recorder
Atomize (step mode)
6.75
2
Chart speed, 25 in./hr
Reagents (All chemicals used were Analytical Grade): Nitric acid (1M). Dilute 61 ml concd. nitric acid (16.4M) to one liter with distilled water. Hydrogen peroxide (30%), arsenic free. Hydrochloric acid (4.5M). Dilute 388 ml concd. hydrochloric acid (11.6M) to 1 liter with distilled water. Sodium hydrogen sulfite solution (1.48M). Dissolve 18 grams of sodium bisulfite (NaHS03) or 14 grams of sodium metabisulfite (Na2S205) in distilled water and dilute to 100 ml. Sodium thiosulfate solution (0.056M). Dissolve 1 gram of sodium thiosulfate (Na2S203) or 1.4 gram of sodium thiosulfate pentahydrate (NazS203.5H20) in distilled water and dilute to 100 ml. Reducing reagent. Add 20 ml of hydrochloric acid (4.5M) slowly to 40 ml of sodium hydrogen sulfite solution (1.48M) and then add 40 ml of sodium thiosulfate solution (0.056M); prepare reagent fresh when required; keep reagent refrigerated; discard all remaining reagent a t the end of the work day. Diethylammonium diethyldithiocarbamate (DDDC) solution A (0.09M). Dissolve 2.0 grams of diethylammonium diethyldithiocarbamate [(C2H5)2N.CS.SNH2(C2H5)21 in carbon tetrachloride (cc14)and dilute to 100 ml; keep reagent refrigerated. DDDC solution B (0.0045M). Dilute 5 ml of DDDC solution A to 100 ml with cc14; prepare fresh daily. Sodium hydroxide solution (1M). Dissolve 4 grams of sodium hydroxide (NaOH) in distilled water and dilute to 100 ml. Arsenite standard A (1 mg As/ml). Dissolve 0.1320 gram of arsenic trioxide (As2O3) in 2 ml of sodium hydroxide (1M);add 25 ml of deionized water; make slightly acid by adding 0.5 ml of hydrochloric acid (4.5M) and dilute to 100 ml with distilled water. Arsenite standard B (10 pg As/ml). Dilute 1 ml of arsenite standard A to 100 ml with distilled water; prepare fresh daily.
Procedure Place 100-ml portions of water samples (or their dilute aliquots) into fused silica tubes (110-ml capacity). Acidify each sample with 1 ml of nitric acid (1M) to approximately pH 2. Add 5 drops of hydrogen peroxide (30%) to each tube and mix well. Irradiate the samples a t 3-4 cm from a 450- or 550-W mercury arc lamp for 3 hr. Remove the tubes from the reactor, add another 5 drops of hydrogen peroxide, mix well, and give three more hours of
/-
23-
4
E
133-
I
C
G
63-
.C-
r&
C R A 61 C R A 63
I
c5
3
I5 Lpq A%
2c
25
:3YCEYTRA-lSU 5vI Figure 1 . Calibration curves for the determination of arsenic using
carbon rod atomizers
radiation. Transfer the irradiated samples quantitatively into individual separatory funnels (125-ml capacity) by rinsing each tube with three 1-ml portions of distilled water. Pipet 10 ml of reducing reagent into each funnel and mix. Pipet 5 ml of DDDC solution B (0.0045M) into each funnel. Shake vigorously for 5 min and allow 15 min for phases to separate. Drain the CC14 layer into a calibrated centrifuge tube (12-ml capacity). Repeat extraction with another 5 ml of DDDC solution B and combine both extracts. Place the centrifuge tube into a hot water bath and allow cc14 to evaporate by gently blowing air into the tube. Adjust final volume to 0.5 ml. Inject 1 p1 of the final solution (or an aliquot diluted with a 0.09M of DDDC solution A) into the carbon rod atomizer (CRA). Prepare dilute aliquots of arsenite standard B (10 pg As/ ml) ranging from 0-2 pg As and treat them exactly the same as the samples.
Results and Discussion Little is known of the chemical form of arsenic in natural waters. It is possible (by analogy with phosphorus) that organoarsenic compounds may exist, and these should be decomposed before analysis. Evaporation and acid digestion may lead to losses of volatile arsenic compounds. The ultraviolet photooxidation procedure of Armstrong et al. (14) avoids this and was adopted. This treatment should leave arsenic in the quinquevalent state [As(V)], and reduction is necessary if trivalent arsenic [As(III)] is to be extracted as a chelate. The use of thiosulfate and bisulfite or metabisulfite in acid solution (15) proved satisfactory, giving complete recovery of added arsenate, with no interference from the nitric acid present. Various chelating agents have been used for extraction of arsenic (16). We tried ammonium pyrrolidindithiocarbamate (APDC) with chloroform but found on evaporation t h a t water separated, and the reagent (probably in the form of pyrrolidindithiocarbamic acid) crystallized from the solution. Diethylammonium diethyldithiocarbamate (DDDC) with carbon tetrachloride (cc14) proved satisfactory a t concentrations between 0.0022V and 0.0066M, and a concentration of 0.0045M was used. The DDDC concentration in the final solution should not exceed 0.13M to avoid matrix effects. Since the concentration in our final solution is 0.09M, it is advisable, if dilution is necessary, to use a 0.09M DDDC solution. Calibration curves are shown in Figure 1, for the CRA 61 and CRA 63 carbon rod atomizers. The linear portion of each curve is from zero absorbance to about 0.25 (measured as a chart reading of 120 mm a t 5 mV range of the recorder) corresponding to 2.5 pg of arsenic in 0.5 ml of extract. Above this level the curves are concave toward the concentration axis. The CRA 61 atomizer gives slightly better sensitivity b u t shows a higher noise level, and a measurable “blank” in the absence of arsenic that is most likely due to the incandescence of the rod (as shown by the absorption using a hydrogen continuum lamp). Twelve injections (1 p1 each) of a 2-pg As/O.5 ml standard gave relative standard deviations of 4.3% (CRA 61) and 5.6% (CRA 63). In analyzing water samples, the precision is &0.4 pg/l. a t 3.1 pg/l., and the detection limit is 1 pg/l. The noise level prohibits the detection of arsenic in water samples containing less than 0.4 pg/l. Table I summarizes the recovery of various arsenicals added to 100-ml samples of potable water. When the spiked samples are acidified with 2 ml of hydrochloric acid (4.5M) and then extracted with DDDC solution, only added arsenite is recovered. After the addition of reducing agent, added arsenite and arsenate are completely recovered while added organic arsenicals are only partially recovered. After ultraviolet photooxidation and h e n the adVolume 8 , Number 8 , August 1974 735
Table I. Recovery of Various Arsenicals (1 pg As) Added to 100-MI Samples of Potable Water (3.1 0.4 pg As/l.)
*
Table II. Intermethodological and lnterlaboratory Analytical Results
% Recovered
No
Arsenicals
oxidation; no reduction
Arsenic trioxide
100‘:
1100
(ASiOs)
Arsenic pentoxide
