- Email: [email protected]
Determination of arsenic in geological materials by x-ray fluorescence spectrometry after solvent extraction and deposition on a filter
0039-9140/83 $3.00+ 0.00 Pergamon Press Ltd
Talanfo, Vol. 30, No. 12, pp. 967-968, 1983 Printed in Great Britain
SHORT COMMUNICATIONS
DETERMINATION OF ARSENIC IN GEOLOGICAL MATERIALS BY X-RAY FLUORESCENCE SPECTROMETRY AFTER SOLVENT EXTRACTION AND DEPOSITION ON A FILTER A. E. HUBERT U.S. Geological Survey, Box 25046, Denver, CO 80225, U.S.A. (Received 21 February 1983. Accepted 1 June 1983)
Summary-Rock, soil, or sediment samples are decomposed with a mixture of nitric and sulphuric acids. After reduction from arsenic(V) with ammonium thiosulphate, arsenic(M) is extracted as the chlorocomplex into benzene from a sulphuric-hydrochloric acid medium. The benzene solution is transferred onto a filter-paper disc impregnated with a solution of sodium bicarbonate and potassium sodium tartrate, and the benzene allowed to evaporate. The arsenic present is determined by X-ray fluorescence. In a 0.5-g sample, l-1000 ppm of arsenic can be determined. The close proximity of the lead La peak (20 48.739, to the arsenic Ka peak (20 48.83”) does not cause any interference, because lead is not extracted under the experimental conditions. Arsenic values obtained are in agreement with those reported for various reference samples.
Arsenic is considered to be a pathfinder or indicator element in geochemical surveys.’ Because of the relative ease of dispersion by both physical and chemical processes, arsenic forms primary as well as secondary halos around ore deposits. The determination of arsenic at or below the crustal abundance level (1.8 ppm) provides a useful geochemical prospecting guideline for detailed surveys. One of the early methods of arsenic determination was a calorimetric procedure involving the Gutzeit apparatus.* Attempts to replace the calorimetric procedure by direct determination by atomicabsorption spectrometry after sample dissolution have not been successful because of interferences caused by the light-scattering effects of other elements present, regardless of the method of atomization. Hydride generation has been used to isolate arsenic from possible ionic interferences and for preconcentration, but the generation of arsine and the final determination by flame or flameless methods may still be affected by various ions present in the sample solutions3” Direct determination of arsenic by X-ray fluorescence analysis of powder or pelleted samples does not provide satisfactory sensitivity at the level of crustal abundance. The usual matrix problems that arise in the determination of trace amounts of various
elements are compounded by the proximity of the lead La (2&48.73”) and arsenic Ku (28,48.83”) peaks. Secondary arsenic peaks are less sensitive, so provide a lower count-rate and decrease the sensitivity of any direct determination technique. The method proposed here makes use of a nitric-sulphuric acid digestion, after which ammonium thiosulphate, benzene, and hydrochloric acid are added, and arsenic(II1) is extracted as the chloride into the benzene layer according to the procedure of Jewett et al.’ The benzene layer is evaporated in contact with a filter-paper disc impregnated with a solution of sodium bicarbonate and potassium sodium tartrate, and the arsenic present on the disc is determined by X-ray fluorescence analysis. EXPERIMENTAL Apparatus
A computer-controlled Siemens SRS* wavelengthdispersive X-ray spectrometer equipped with a tungsten source (set at 50 kV and 40 mA) and a lithium fluoride 110 crystal was used. Each sample was counted for 100 set at the arsenic Ku peak (28, 48.83”) and the background was counted (2~9,50”) for the same time. Each microgram of arsenic provided a 20-cps count-rate above background. Plastic sample cups (EC16, Chemiplex Industries) were used to hold the benzene extract and the Whatman No. 540 filter-paper discs. Reagents
All chemicals used were of analytical-reagent *Any use of trade names is for descriptive purposes only and does not imply endorsement by the U.S. Geological survey.
grade.
Arsenic standard solution, 100Opg/ml. Dissolve 0.132 g of
As,O, in 2 ml of 1M sodium hydroxide, acidify with 1 ml of 10% (v/v) nitric acid and dilute to lOOm1 with water.
967
968
SHORT
COMMUNICATIONS
Solution of sodium bicarbonate and potassium sodium tartrate. Dissolve 5 g of NaHCO, and 5 g of
KO,C(CHOH),CONa~~H,O Ammonium
in 100ml 02 water.
thiosulphate solution, 1%. Dissolve
1 g of
(NH,),S,O, in 100 ml of water. Calibration graph
Pipette aliquots of the arsenic standard solution containing 0, 1, 2, 5, 10, 20, 50, 100, 200 and 500 pg of arsenic into 50-ml Teflon beakers, each containing 0.5 g of rock or soil known to contain less than 0.1 pg of arsenic. Apply the digestion and extraction procedures as for the test samples. The calibration is linear over this range. The standards prepared in this manner correspond to arsenic concentrations up to 1000 ppm in the samples. Procedure
Weigh a 0.5-g sample of pulverized rock, soil, or sediment into a 50-ml Teflon beaker. Add 2 ml of water and 10 ml or more of concentrated nitric acid, depending on the organic content of the sample. Add 5 ml of sulphuric acid and heat on a hot-plate until copious white fumes of sulphur trioxide are evolved. Allow to cool and transfer the solution to a 16 x 150 mm Pyrex test-tube fitted with a Teflon screw-cap. Add 1 ml of 1% ammonium thiosulphate solution and 5 ml of benzene and allow to cool. Teflon beakers are used because the product from the digestion can be transferred practically completely without rinsing. From a pipette add 0.1 ml of concentrated hydrochloric acid on top of the benzene layer and immediately cap the’tube. Shake it for 10 min and allow the two phases to separate. Add another 0.1 ml of concentrated hydrochloric acid and shake again for lOmin, to obtain consistent results. Centrifuge the test-tube to separate the two phases. Remove the benzene layer with a pipette and place it in an EC16 sample cup. Immediately cover the benzene with a Whatman No. 540 filter-paper disc impregnated with 0.2ml of the sodium bicarbonate-potassium sodium tartrate solution. Allow the benzene to evaporate and the filter-paper disc to dry in air, then cover the cap plug with a layer of Mylar film. Secure the Mylar film to the sample cup with a retaining ring, turn the sample cup over and shake the filter loose so that it comes to rest on the Mylar film, which is used as the support for the determination of arsenic by X-ray fluorescence. RESULTS
Sample
digestion
The procedure
AND DISCUSSION
and solvent extraction
for sample
digestion
1. GXR 1
Jasperoid 2. 3.
4. 5. 6. 7.
GXR 2 Soil GXR 3f Fe-Mn Deposit GXR 4 Copper mill head GXR 5 Soil GXR 6 Soil Mag-1 Marine mud
Range
Mean
PPm
PPm
RSD > “/’ 0
Reported values, ppm
420-472
448
4.0
460 f 30t
34
3.3
31 *st
4005
6.1
4000 * 450t
96112
106
5.5
98 + lot
12-14
13
5.5
12*3t
309
3.8
340 * 30t
7
10.5
65
32-36 383&4360
292-320 6-8
*Relative standard deviation. tReference 8. fO.lO-g sample. BReference 9.
Interferences
Samples were spiked with 1OOO~g of antimony, bismuth, cobalt, copper, nickel, and lead, and analysed by the proposed procedure. Lead was not detected at the 48.73” La lead peak on the filter disc prepared from samples containing 1000 pg of lead. It therefore appears that lead is not extracted under the experimental conditions and does not interfere with the arsenic detection. Under similar conditions, cobalt, copper, and nickel are also not extracted and cannot be detected. Antimony and bismuth are partially extracted, but do not interfere with the arsenic determination. Analysis
and solvent
Table 1. Replicate analyses (n = 6) of arsenic in seven reference
Sample
extraction is adapted from that of Jewett et al.,’ which involves the extraction of arsenic as the chlorocomplex from sulphuric-hydrochloric acid. These authors reported a recovery of W-90% for arsenic in biological materials after digestion and extraction. Discs prepared by spiking a sample and following the procedure were compared with discs prepared by placing an equivalent amount of arsenic on a filter. The results confirmed that S&90% recovery is obtained. If a benzene extract containing arsenic is allowed to evaporate on a filter-paper disc without the sodium bicarbonate-potassium sodium tartrate solution present, the arsenic evaporates along with the benzene. Potassium sodium tartrate prevents this loss, and the sodium bicarbonate neutralizes any sulphuric acid left in the benzene layer and makes it easier to free the filter-paper disc from the bottom of the plastic cup. Any excess of sulphuric acid transferred to the plastic cup would destroy the filter paper and require the analysis to be repeated. It is necessary to carry arsenic-spiked samples through the procedure as standards to compensate for losses in the digestion and extraction steps.
of samples
Replicate analyses of seven geological reference samples with varying matrices gave the arsenic values presented in Table 1. The mean values found by the proposed method are in general agreement with values reported in the literature. REFERENCES
1. R. W. Boyle and I. R. Jonasson, J. Geochem. Explor., 1973, 2, 251. 2. H. Almond, Anal. Chem., 1953, 25, 1766. 3. G. F. Kirkbright and M. Taddia, Anal. Chim. Acta, 1978, 100, 145. 4. G. E. M. Aslin, J. Geochem. Explor., 1976, 6, 321. 5. S. Terashima, Anal. Chim. Acta, 1976, 86, 43. 6. F. D. Pierce and H. R. Brown, Anal. Chem., 1977, 49,
1417. I. G. K. Jewett, R. P. Himes and 0. U. Andens, J. Radioanal. Chem., 1977, 37, 813. 8. E. S. Gladney, D. R. Perrin, J. W. Owens and D. Knab, Anal. Chem., 1979, 51, 1557. 9. E. S. Gladney and W. E. Goode, Geostds. New& 1981, 5, 1.
Talanfo, Vol. 30, No. 12, pp. 967-968, 1983 Printed in Great Britain
SHORT COMMUNICATIONS
DETERMINATION OF ARSENIC IN GEOLOGICAL MATERIALS BY X-RAY FLUORESCENCE SPECTROMETRY AFTER SOLVENT EXTRACTION AND DEPOSITION ON A FILTER A. E. HUBERT U.S. Geological Survey, Box 25046, Denver, CO 80225, U.S.A. (Received 21 February 1983. Accepted 1 June 1983)
Summary-Rock, soil, or sediment samples are decomposed with a mixture of nitric and sulphuric acids. After reduction from arsenic(V) with ammonium thiosulphate, arsenic(M) is extracted as the chlorocomplex into benzene from a sulphuric-hydrochloric acid medium. The benzene solution is transferred onto a filter-paper disc impregnated with a solution of sodium bicarbonate and potassium sodium tartrate, and the benzene allowed to evaporate. The arsenic present is determined by X-ray fluorescence. In a 0.5-g sample, l-1000 ppm of arsenic can be determined. The close proximity of the lead La peak (20 48.739, to the arsenic Ka peak (20 48.83”) does not cause any interference, because lead is not extracted under the experimental conditions. Arsenic values obtained are in agreement with those reported for various reference samples.
Arsenic is considered to be a pathfinder or indicator element in geochemical surveys.’ Because of the relative ease of dispersion by both physical and chemical processes, arsenic forms primary as well as secondary halos around ore deposits. The determination of arsenic at or below the crustal abundance level (1.8 ppm) provides a useful geochemical prospecting guideline for detailed surveys. One of the early methods of arsenic determination was a calorimetric procedure involving the Gutzeit apparatus.* Attempts to replace the calorimetric procedure by direct determination by atomicabsorption spectrometry after sample dissolution have not been successful because of interferences caused by the light-scattering effects of other elements present, regardless of the method of atomization. Hydride generation has been used to isolate arsenic from possible ionic interferences and for preconcentration, but the generation of arsine and the final determination by flame or flameless methods may still be affected by various ions present in the sample solutions3” Direct determination of arsenic by X-ray fluorescence analysis of powder or pelleted samples does not provide satisfactory sensitivity at the level of crustal abundance. The usual matrix problems that arise in the determination of trace amounts of various
elements are compounded by the proximity of the lead La (2&48.73”) and arsenic Ku (28,48.83”) peaks. Secondary arsenic peaks are less sensitive, so provide a lower count-rate and decrease the sensitivity of any direct determination technique. The method proposed here makes use of a nitric-sulphuric acid digestion, after which ammonium thiosulphate, benzene, and hydrochloric acid are added, and arsenic(II1) is extracted as the chloride into the benzene layer according to the procedure of Jewett et al.’ The benzene layer is evaporated in contact with a filter-paper disc impregnated with a solution of sodium bicarbonate and potassium sodium tartrate, and the arsenic present on the disc is determined by X-ray fluorescence analysis. EXPERIMENTAL Apparatus
A computer-controlled Siemens SRS* wavelengthdispersive X-ray spectrometer equipped with a tungsten source (set at 50 kV and 40 mA) and a lithium fluoride 110 crystal was used. Each sample was counted for 100 set at the arsenic Ku peak (28, 48.83”) and the background was counted (2~9,50”) for the same time. Each microgram of arsenic provided a 20-cps count-rate above background. Plastic sample cups (EC16, Chemiplex Industries) were used to hold the benzene extract and the Whatman No. 540 filter-paper discs. Reagents
All chemicals used were of analytical-reagent *Any use of trade names is for descriptive purposes only and does not imply endorsement by the U.S. Geological survey.
grade.
Arsenic standard solution, 100Opg/ml. Dissolve 0.132 g of
As,O, in 2 ml of 1M sodium hydroxide, acidify with 1 ml of 10% (v/v) nitric acid and dilute to lOOm1 with water.
967
968
SHORT
COMMUNICATIONS
Solution of sodium bicarbonate and potassium sodium tartrate. Dissolve 5 g of NaHCO, and 5 g of
KO,C(CHOH),CONa~~H,O Ammonium
in 100ml 02 water.
thiosulphate solution, 1%. Dissolve
1 g of
(NH,),S,O, in 100 ml of water. Calibration graph
Pipette aliquots of the arsenic standard solution containing 0, 1, 2, 5, 10, 20, 50, 100, 200 and 500 pg of arsenic into 50-ml Teflon beakers, each containing 0.5 g of rock or soil known to contain less than 0.1 pg of arsenic. Apply the digestion and extraction procedures as for the test samples. The calibration is linear over this range. The standards prepared in this manner correspond to arsenic concentrations up to 1000 ppm in the samples. Procedure
Weigh a 0.5-g sample of pulverized rock, soil, or sediment into a 50-ml Teflon beaker. Add 2 ml of water and 10 ml or more of concentrated nitric acid, depending on the organic content of the sample. Add 5 ml of sulphuric acid and heat on a hot-plate until copious white fumes of sulphur trioxide are evolved. Allow to cool and transfer the solution to a 16 x 150 mm Pyrex test-tube fitted with a Teflon screw-cap. Add 1 ml of 1% ammonium thiosulphate solution and 5 ml of benzene and allow to cool. Teflon beakers are used because the product from the digestion can be transferred practically completely without rinsing. From a pipette add 0.1 ml of concentrated hydrochloric acid on top of the benzene layer and immediately cap the’tube. Shake it for 10 min and allow the two phases to separate. Add another 0.1 ml of concentrated hydrochloric acid and shake again for lOmin, to obtain consistent results. Centrifuge the test-tube to separate the two phases. Remove the benzene layer with a pipette and place it in an EC16 sample cup. Immediately cover the benzene with a Whatman No. 540 filter-paper disc impregnated with 0.2ml of the sodium bicarbonate-potassium sodium tartrate solution. Allow the benzene to evaporate and the filter-paper disc to dry in air, then cover the cap plug with a layer of Mylar film. Secure the Mylar film to the sample cup with a retaining ring, turn the sample cup over and shake the filter loose so that it comes to rest on the Mylar film, which is used as the support for the determination of arsenic by X-ray fluorescence. RESULTS
Sample
digestion
The procedure
AND DISCUSSION
and solvent extraction
for sample
digestion
1. GXR 1
Jasperoid 2. 3.
4. 5. 6. 7.
GXR 2 Soil GXR 3f Fe-Mn Deposit GXR 4 Copper mill head GXR 5 Soil GXR 6 Soil Mag-1 Marine mud
Range
Mean
PPm
PPm
RSD > “/’ 0
Reported values, ppm
420-472
448
4.0
460 f 30t
34
3.3
31 *st
4005
6.1
4000 * 450t
96112
106
5.5
98 + lot
12-14
13
5.5
12*3t
309
3.8
340 * 30t
7
10.5
65
32-36 383&4360
292-320 6-8
*Relative standard deviation. tReference 8. fO.lO-g sample. BReference 9.
Interferences
Samples were spiked with 1OOO~g of antimony, bismuth, cobalt, copper, nickel, and lead, and analysed by the proposed procedure. Lead was not detected at the 48.73” La lead peak on the filter disc prepared from samples containing 1000 pg of lead. It therefore appears that lead is not extracted under the experimental conditions and does not interfere with the arsenic detection. Under similar conditions, cobalt, copper, and nickel are also not extracted and cannot be detected. Antimony and bismuth are partially extracted, but do not interfere with the arsenic determination. Analysis
and solvent
Table 1. Replicate analyses (n = 6) of arsenic in seven reference
Sample
extraction is adapted from that of Jewett et al.,’ which involves the extraction of arsenic as the chlorocomplex from sulphuric-hydrochloric acid. These authors reported a recovery of W-90% for arsenic in biological materials after digestion and extraction. Discs prepared by spiking a sample and following the procedure were compared with discs prepared by placing an equivalent amount of arsenic on a filter. The results confirmed that S&90% recovery is obtained. If a benzene extract containing arsenic is allowed to evaporate on a filter-paper disc without the sodium bicarbonate-potassium sodium tartrate solution present, the arsenic evaporates along with the benzene. Potassium sodium tartrate prevents this loss, and the sodium bicarbonate neutralizes any sulphuric acid left in the benzene layer and makes it easier to free the filter-paper disc from the bottom of the plastic cup. Any excess of sulphuric acid transferred to the plastic cup would destroy the filter paper and require the analysis to be repeated. It is necessary to carry arsenic-spiked samples through the procedure as standards to compensate for losses in the digestion and extraction steps.
of samples
Replicate analyses of seven geological reference samples with varying matrices gave the arsenic values presented in Table 1. The mean values found by the proposed method are in general agreement with values reported in the literature. REFERENCES
1. R. W. Boyle and I. R. Jonasson, J. Geochem. Explor., 1973, 2, 251. 2. H. Almond, Anal. Chem., 1953, 25, 1766. 3. G. F. Kirkbright and M. Taddia, Anal. Chim. Acta, 1978, 100, 145. 4. G. E. M. Aslin, J. Geochem. Explor., 1976, 6, 321. 5. S. Terashima, Anal. Chim. Acta, 1976, 86, 43. 6. F. D. Pierce and H. R. Brown, Anal. Chem., 1977, 49,
1417. I. G. K. Jewett, R. P. Himes and 0. U. Andens, J. Radioanal. Chem., 1977, 37, 813. 8. E. S. Gladney, D. R. Perrin, J. W. Owens and D. Knab, Anal. Chem., 1979, 51, 1557. 9. E. S. Gladney and W. E. Goode, Geostds. New& 1981, 5, 1.