- Email: [email protected]
Determination of trace arsenic, antimony, selenium and tellurium in various oxidation states in water by hydride generation and atomic-absorption spectrophotometry after enrichment and separation with thiol cotton
Talam, Printed
0039-9140/83/040265-06503.00/O Copyright 0 1983 Pergamon Press Ltd
Vol. 30, No. 4, pp. 265-270, 1983 III Great Bntain. All tights reserved
DETERMINATION OF TRACE ARSENIC, ANTIMONY, SELENIUM AND TELLURIUM IN VARIOUS OXIDATION STATES IN WATER BY HYDRIDE GENERATION AND ATOMIC-ABSORPTION SPECTROPHOTOMETRY AFTER ENRICHMENT AND SEPARATION WITH THIOL COTTON MU-QING Yu and GUI-QIN LIU Changchun Institute of Geography, Chinese Academy of Science
and Department
QINHAN JIN* of Chemistry, Jilin University, Changchun, People’s Republic of China
(Received 4 March 1982. Revised 9 October 1982. Accepted 20 October 1982) Summary-A novel procedure for determination of trace As(II1) and As(V), Sb(II1) and Sb(V), Se(IV) and Se(VI), Te(IV) and Te(V1) in water by atomic-absorption spectrophotometry after separation and enrichment with “thiol cotton” and hydride generation has been established. The sorption behaviour of various oxidation states of arsenic, antimony, selenium and tellurium, and the conditions of quantitative sorption and desorption of these species were studied. The procedures for reducing species from higher oxidation states were optimized. Interferences from other species and their elimination were investigated. The selectivity of the procedure for the determination of species in higher and lower oxidation states was examined. The procedure has been successfully used to determine arsenic, antimony, selenium and tellurium in water, in the range from pg/ml to ng/ml. The recoveries for added spikes were in the range 9lrllO%, with coefficients of variation in the range 3-8%.
It has been shown by more and more facts that the transport and conversion processes as well as the biological toxic effects of a variety of trace elements, such as mercury, chromium, arsenic, antimony, selenium and tellurium, in the environment are closely related to the oxidation states and compounds present. It is obvious that ignoring this fact can lead to inadequate or even totally wrong evaluation of environmental quality. For this reason the investigation of methods for specification of trace elements has been of interest for several years. Various methods for determining chromium,‘-’ arsenic,“” selenium’~*’ antimony,‘2-‘4 and tellurium22-24 in various oxidation states, after separation by co-precipitation, solvent extraction, ionexchange, hydride generation and stepwise reduction, have been established, but most of them are limited in their application. It has been shown that the thiol group has very strong affinity for heavy-metal ions and that cotton impregnated with thioglycollic acid (“thiol cotton”) quantitatively adsorbs a variety of trace elements from water,25-27 but its adsorptive power depends significantly on the oxidation state of the element concerned. In general, the lower oxidation states are readily adsorbed whereas the higher states are not. It
*To whom correspondence
should be addressed. 265
should therefore be possible to adsorb the lower states selectively, and then to reduce the higher states and adsorb the products. On the basis of investigations on the enrichment in this way of more than twenty trace species, including Pt(II), Pd(II), Au(III), Se(W), Te(IV), As(III), Hg(II), Sb(III), Bi(III), Sn(II), Ag(I), Cu(II), In(III), Pb(II), Cd(II), Zn(II), Co(I1) and Ni(II),2~28 we have established this procedure for determining trace As(III), As(V), Sb(III), Sb(V), Se(IV), Se(VI), Te(IV) and Te(V1) in water by hydride generation and atomic-absorption spectrophotometry after separation and enrichment with thiol cotton. EXPERIMENTAL Apparatus A Pye-Unicam SP 1900 atomic-absorption spectrophotometer and a home-made- hydride-generator (Fig. 1) were used, under the conditions listed in Table 1. A 250-ml (or 500-ml) Pyrex separatory funnel fitted with a thiol-cotton tube was used for separation and enrichment of various trace species. The thiol-cotton tube was a glass tube with about 0.1 g of thiol cotton in it. It was 8 mm in bore and 100 mm long, and one of the ends was narrowed. Reagents Suprapur-grade hydrochloric acid and nitric acid were used. All the other reagents were analytical-reagent grade. All water used was redistilled. Standard antimony(III) solution (loo0 ppm). Prepared from antimony potassium tartrate.
MU-QINGYu et al.
266
I
I
I
4
I
--
--
6 i_-J
Fig. 1. Eqmpment for hydride-generation determmation. 1, N, cylinder; 2, flowmeter; 3, hydride generator (25 mm bore, 80 mm long); 4, light durce; 5, quartz-tube atomizer (8 mm bore, 150 mm long); 6, detector; I, syringe.
Standard antimonyp)
oxidizing the standard manganate.
solution (10 ppm). Prepared by antimony(II1) solution with per-
Standard arsenic(III) solution (1000 ppm). Prepared from
arsenious oxide. Standard arsenic(V) solution (la00 ppm). Prepared from
sodium arsenate. Standard selenium(W) solution (1 OOOppm).Prepared from
selenium dioxide. Standard
tellurium(IV)
solution (1000 ppm).
Prepared
solution (1000 ppm).
Prepared
from tellurium powder. Standard
tellurium(W)
from telluric acid. Potassium iodide (20x)-thiourea (2%) solution. Potassium iodide (20%)~ascorbic acid (4%) solution. Potassium borohydride solution (2%). Prepared in 0.5%
sodium hydroxide solution. Titanium(III) chloride, 15-20x solution. Hydrogen peroxide, 30% solution. Thiol cotton. Thioglycollic acid (50 ml), acetic anhydride
(35 ml), glacial acetic acid (16 ml), concentrated sulphuric acid (0.15 ml) and water (5 ml) were put in that order into a bottle and mixed well, and then 15 g of absorbent cotton were added. The bottle was closed, then left in an oven at 4&50” for 4-5 days. The cotton was collected on a sinteredglass funnel and washed with water till the washings were neutral, then dried at 40-50” and kept in a brown bottle. Procedures Separation and enrichment. Put a selected volume of water sample into the separation assembly, adjust its acidity to the
value indicated in Table 2 and then pass it through the thiol-cotton tube at a suitable flow-rate to extract the Sb(III), As(III), Se(IV) or Te(IV). Treat the effluent solution with the appropriate reagents (as indicated in Table 3) to reduce Sb(V), As(V), Se(V1) or Te(V1) to the next lower oxidation state, and collect the reduced species in another thiol-cotton tube. Desorb the As(III), Sb(III), Se(IV) or Te(IV) as follows. Arsenic. Elute slowly with 3 ml of hot concentrated hydrochloric acid. Collect the eluate in a lo-ml graduated tube. Add 0.2 ml of potassium iodide solution to the eluate and dilute to 10 ml with water. Antimony. Elute slowly with 3 ml of 5M hydrochloric acid into a lo-ml graduated tube. Add 0.2 ml of potassium iodide solution to the eluate and dilute to 10 ml with water. Selenium. Put the cotton into a lo-ml graduated tube. Add 2.5 ml of concentrated hydrochloric acid and 1 drop of concentrated nitric acid. Heat in a boiling water-bath for 3 min. Cool to room temperature and dilute to 10 ml with water. Tellurium. Place the cotton in a lo-ml graduated tube. Add 2.5 ml of concentrated hydrochloric acid. Heat in a boiling water-bath for 3 min. Cool to room temperature and dilute to 10 ml with water.
Use a separate sample for each element, but if desired, antimony(II1) can be selectively eluted first. Determination. Set the AAS instrument according to Table 1. Place a known amount of sample solution in the
Table 1. Conditions for hydride generation and AAS determination Parameter Light-source Wavelength, nm Lamp current, mA Slit-width setting Temperature of quartz tube, “C Flow-rate of carrier-gas (NJ, l./min Acidity of solution analysed, N
Sb
As
Se
Te
Pye hollow-cathode lamps 217.6 193.7 196.0 214.2 12 10 6 10 0.2 0.2 0.2 0.2 900 900 900 900 0.6
0.6
0.6
0.6
1.5
3
3
3
Determination
of trace arsenic etc.
Table 2. Acidity for enriching As(III), Sb(III), Se(W) and Te(IV) Species
Acidity of water sample
Sb(II1) As(III) Se(IV) Te(IV)
pH 6-8 1M HCl 0.3M HCl 0.3M HCl
hydride generator and make up to 5 ml with hydrochloric acid of the same concentration as the sample. Close the generator and with a syringe rapidly add 1.5 ml of potassium borohydride solution to the lower part of the solution. From the recorded absorbance calculate the content of the species in the sample.
261
reduced by the thiol group and then adsorbed by the thiol cotton. The degree of reduction depends on the species concerned and the acidity but is never quantitative. For example, Sb(V) is partly reduced to Sb(II1) in dilute hydrochloric acid. Therefore, this species can be adsorbed significantly (- 50% maximum) at pH < 5, but Se(VI), Te(V1) and As(V) can only be slightly reduced and adsorbed at acidities below SM hydrochloric acid. The range of acidity for quantitative adsorption shown in Figs. 2 and 3 will change slightly with the composition of the solution, the concentration of the trace element and the quality of the thiol cotton. The conditions shown in Table 2 are recommended as optimal.
RESULTS AND DISCUSSION
Adsorption behaviour of thiol cotton towards arsenic, antimony, selenium and tellurium
It is clear from Figs. 2 and 3 that the adsorption behaviour of those elements depends greatly on the acidity of the solution and the oxidation state of the element. Thiol cotton efficiently adsorbs only As(III), Sb(III), Se(IV) and Te(IV), the ranges of acidity for quantitative adsorption being Sb(III), pH 8-2M HCl; As(III), 0.557M HCl; Te(IV) pH 8-9M HCl; Se(IV), pH 2-10M HCl. It was also noticed that within certain ranges of hydrochloric acid concentration, some species in higher oxidation states could be
Choice of desorption procedures
The adsorptive power of thiol cotton for Sb(II1) is rather poor, and this species can be desorbed quantitatively with SM hydrochloric acid. To desorb As(III), the thiol cotton should be dipped into concentrated hydrochloric acid for several minutes. Se(IV) and Te(IV) are adsorbed very firmly by thiol cotton, so it is necessary to heat the cotton with concentrated hydrochloric acid in a boiling waterbath to desorb these species quantitatively; this process also attacks the cotton itself.
Table 3. Conditions for reducing As(V), Sb(V), Se(W) and Te(VI) Species
SW’) As(V) Se(W) Te(VI)
Amount of reductant used, and reduction procedure
Acidity
0.3M HCl Add KI to 0.1% and ascorbic acid to 0.02%; let stand for 10 min at room temperature 1M HCl Add KI to 0.2% and thiourea to 0.02%; heat for 3-5 min in a boiling water-bath (the temperature of the solution reaches 80”) 0.3M HCl Add TiCl, to 0.1%; after 30 min add H,O, to 0.1% and heat for 5 min in a boiling water-bath 0.3M HCl Add TiCl, to 0.1%; after 10 min add H,O, to 0.1% and let stand for 5 min at room temperature
IO
6
6
4
2
OIN
I Fig 2.
The
Acldlty 3
5
7
PH
effects of acidity on the adsorption of As(M), As(V), Sb(III) and Sb(V) (all 10 ng/ml concentration) on thiol cotton. (0) As(M); (0) As(V); (A) Sb(III); (A) Sb(V).
268
MU-QING
IO
8
6
4
Yu ef al.
2
OIN
Acldlty
Fig. 3. The effects of acidity on the adsorption of Se(N), %@I), Te(IV) and Te(V1) (all 10 ng/ml concentration) on thiol cotton. (0) Se(N); (0) Se(V1);(A) Te(IV); (A) Te(VI). The procedure for determination of arsenic, antimony, selenium and tellurium by hydride-generation atomic-absorption spectrophotometry is also related to the oxidation state. Therefore, the oxidation state of the species to be determined in the desorbed solution should be consistent with the r~uirements of this technique. For this reason, a small amount of nitric acid is added when selenium is to be desorbed, to ensure the selenium is in the quadrivalent state. Choice of reduction procedures
Since As(V), Sb(V), Se(VI) and TefVI) must be reduced to As(III), Sb(III), Se(IV) and Te(IV) before they can be quantitatively adsorbed by thiol cotton, the reduction procedure should be compatible with the enrichment technique. Those selected are as follows. SbfV) can be rapidly reduced by adding potassium iodide to the dilute hydrochloric acid sample solution. As(V) is more difficult to reduce than Sb(V). It is necessary to add potassium iodide and thiourea to a sample solution that is 1M in hydrochloric acid and heat in a boiling water-bath for several minutes. Se(VI) and Te(V1) are diflicult to reduce selectively to Se(IV) and Te(IV) because they are very easily reduced to the elements in hydrochloric acid medium. We have found that if enough titanium trichloride to reduce Te(VI) to Te(IV) and Se(V1) to Se is added to a dilute hydrochloric acid solution followed by enough hydrogen peroxide to oxidize the excess of titanium trichloride and to form a complex with it, a stable Te(IV) solution is formed and the elemental selenium is oxidized to Se(N). The concentration of hydrochloric acid, the amounts of titanium trichloride and hydrogen peroxide used and the temperature and time of reduction all affect the efficiency of reduction. Table 3 shows the recommended reduo tion conditions. This reduction procedure is simple, rapid and easy to operate, and Se(V1) and Te(V1) can be converted into Se(IV) and Te(IV) simultaneously and quantitatively.
Amount of thiol cotton to use andflow-rate of sample solution The adsorption capacity of thiol cotton for the species concerned and the flow-rates for quantitative adsorption are shown in Table 4. The operational adsorption capacities are between a third and a half of the corresponding saturation capacities. On the basis of the concentration ranges of these species that can be adsorbed by thiol cotton and those existing in practical environmental water samples, about 0.1 g of thiol cotton is adequate for enrichment. Selectivity for higher or lower oxidation states The selectivity for enriching and determining species in their higher or lower oxidation states is quite good. The maximum ratios allowed between higher and lower oxidation-state species, for an error of lo%, are shown in Table 5. To obtain selective determination the following measures should be taken. 1. The thiol-cotton tube should first be rinsed with water or with hydrochloric acid of appropriate concentration to bring it to the optimum state for the adsorption. 2. If the concentration ratio of lower to higher oxidation state species in the water sample is very high, to adsorb the lower-valence species the amount of thiol cotton used should be increased accordingly and the flow-rate should be reduced. 3. If the higher to lower oxidation state concentration ratio is high, the lower-valence species ad-
Table 4. Saturation adsorption capacity of thiol cotton and Row-rate of water sample As(W)
SbfIII)
k-5
5-l
Saturation capacity, rng/g of thiol cotton Flow-rate, ml/min* *For 0.1 g of thiol cotton.
G2
Se(IV) Te(IV) 7-9 65
8-11
Determination
of trace arsenic etc.
Table 5. Selectivity for determining As, Sb, Se and Te at higher or lower valence resnectivelv.* with X+lo”/, error Maximum ratio of higher valence/lower valence species for determination of lower valence species Maximum ratio of lower valence/higher valence species for determination of higher valence species
As
Sb
Se
Te
30
20
40
50
20
20
20
30
*The volume of water sample enriched was 100-200 ml and the concentration of As, Sb. Se or Te was 24 ng/ml. sorbed on the thiol cotton should he washed with water or the appropriate concentration of hydrochloric acid to remove any residual higher oxidation state species on it. Interference
studies
It has been shown
that
except
for some
slight
269
interference in adsorption of Sb(II1) from higher concentrations of humic acid, the species usually found in environmental water, such as Cl-, Brr, SO:-, SO;-, CO:-, HCO;, NO;, NO;, HPO:-, K+, Na+, Ca*+, Md+, Fe’+, A13+ and various heavy metal ions, as well as fulvic acid, glycine, cysteine, citric acid and the like, do not interfere in the quantitative adsorption of As(III), Sb(III), Se(IV) and Te(IV) by thiol cotton. It has also been shown that the direct determination of As, Sb, Se and Te by hydride-generation atomic-absorption spectrophotometry would suffer interferences from Co*+, Ni*+, Fe3+, Cu*+, Pb*+, Sn4+, Bi3+ and Ag+ ions, and between the four elements themselves, but after enrichment with thiol cotton and stepwise desorption these interferences are effectively eliminated. Co’+, Ni*+, Cu*+, Sn4+, Bi3+, Pb2+ and Ag+ adsorbed by thiol cotton can be removed quantitatively by eluting with l-2M hydrochloric acid.
Table 6. Analysis of real water samples and recovery of the procedure Found Lower valency Standard added, ng/ml Element Arsenic
Spring water Natural river water Polluted river water
Antimony
Natural river water Polluted lake water
Polluted river water
Selenium
Well water Natural river water Polluted river water
Tellurium
Natural river water Sea-water Polluted river water
*Average of three replicates.
Lower valency
Higher valency
Total, nglml
0 4.00 0 2.00 0 2.00 6.00 10.00
0 4.00 0 2.00 0 10.00 6.00 2.00
3.05 6.95 0.40 2.54 6.58 8.80 12.3 15.7
0 1.oo 2.00 0 1.oo 3.00 0 1.00 1.00 2.00
0 2.00 1.oo 0 3.00 1.00 0 2.00 1.00 1.00
0 0.50 0 1.00 0 1.50 1.00 0.50
0 0.50 0 1.oo 0 0.50 1.00 1.50
0.04 0.58 0.10 1.18 0.28 1.66 1.27 0.77
0 2.00 1.00 0 2.00 1.00 0 0.50 1.00 2.00
0 1.oo 2.00 0 1.oo 2.00 0 2.00 1.oo 0.50
Recovery of added standard,* 0’ /a 98 107 115 97 92 95 88 90 93 87 95 88
Higher valency
Total, nglml 5.44 9.90 0.76 2.60 135 22.0 19.4 15.4 0.26 2.30 1.30 1.10 4.25 2.00 0.65 2.90 1.68 1.73
Recovery of added standard,* 9/, 117 92 85 98 95 102 104 98 90 112 103 108
0.64
108 108 92 99 98 98 102 96 112 110 101 99
1.15
102 95 110 93 86 98 105 100 105 96 87 104
MU-QINGYu et al.
270
Table 7. Sensitivity and precision of the procedure Sensitivity (1% absorbance),* ng/ml Detection limit,? ng/ml Lower valency Coefficient of Higher valencv variation.6 ‘/,
As
Sb
Se
Te
0.11 0.006 3.5 5.4
0.09 0.005 7.9 5.0
0.20 0.010 2.8 3.2
0.16 0.008 2.7 3.4
*The volume analysed was 5 ml.
tThe volume of water sample enriched was 200 ml. §Based on 11 samples.
To isolate antimony, the thiol cotton is eluted with acid. As, Se and Te are retained and
5M hydrochloric
the Sb desorbed. When selenium is being determined, any coexisting arsenic is oxidized to As(V) and so does not interfere. When arsenic is being determined, the only interference is from selenium, but this can be eliminated by adding a small amount of potassium iodide. The interference from selenium and arsenic in determination of tellurium can be largely eliminated by enrichment with thiol cotton and desorption. Sensitivity
and precision
As, Sb, Se and Te in various oxidation states in various environmental water samples were determined by this procedure. The recoveries were in the range 90-110x, as shown in Table 6. The sensitivity and the coefficient of variation (11 replicates) for these elements in the range 2-4 ng/ml are shown in Table 7. Conclusions
The procedure has the following characteristics. 1. The enrichment efficiency is high, the enrichment factor ranging up to several hundred. The detection limit can be as low as a few rig/l.. 2. The separation procedure is simple and the higher and lower oxidation-state species can be separated and enriched separately. Therefore the error introduced by the subtraction technique adopted in some other methods is avoided. 3. The selectivity is good. There is no mutual interference between higher and lower valence species when their concentration ratio is not more than 20-50, depending on the particular system. 4. The accuracy is good. Interference from various concomitant species is eliminated and the method can be used to analyse various environmental water samples.
REFERENCES
1. T. Goto and S. Ginba, Bunseki Kaguku, 1974, 23, 517. 2. T. Matsuo, J. Shida, M. Abiko and K. Konno, ibid., 1975, 24, 723. 3. K. Hiiro, T. Owa, M. Takaoka, T. Tanaka and A. Kauahara, ibid., 1976, 25, 122. 4. M. Q. Yu and Y. J. Zhao, J. Environ. Sci. (Chinese), 1979, No. I, 20. 5. S. R. Wang, F. Z. Xu, H. F. Zhou and X. L. Jin, ibid., 1980, No. 3, 11. 6. T. Kamada, Talanta, 1976, 23, 835. 7. J. Aggett and A. C. Aspell, Analyst, 1976, 101, 341. 8. D. J. Myers and J. Osteryoung, Anal. Chem., 1973, 45, 267. 9. F. T. Henry, T. 0. Kirch and T. M. Thorpe, ibid., 1979, 51, 215. 10. Y. Talmi and D. T. Bostic, ibid., 1975, 47, 2145. 11. M. 0. Andreae, ibid., 1977, 49, 820. 12. T. Kamada and Y. Yamamoto, TaZanta, 1977,24, 330. 13. I. Valente and H. J. M. Bowen. Analvst. 1977. 102.842. Anal. ?him. 14. K. S. Subramanian and J. C.‘Mer&gx, Acta, 1981, 124, 131. 15. 0. Yoshil, K. Hiraki, Y. Nishikawa and T. Shigematsu, Bunseki Kagagu, 1971, 26, 91. 16. Y. Shimoshi and K. Toei, Anal. Chim. Acta, 1978, 100, 15. 17. T. Kamada, T. Shiraishi and Y. Yamamoto, Talanta, 1978, 25, 15. 18. G. A. Cutter, Anal. Chim. Acta, 1978, 98, 59. 19. Y. Sugimura and Y. Suzuki. J. Oceanogr. Sot. Japan, 1977, 33, 23. 20. C. I. Measures and J. D. Burton, Anal. Chim. Acta, 1980, 120, 177. 21. H. J. Robberecht and R. E. Van Grieken, Anal. Chem., 1980, 52, 449. 22. T. Kamada, N. Sugita and N. Yamamoto, Talanta, 1979, 26, 337. 23. K. Jin, M. Taga, H. Yoshida and S. Hikime, BUN. Chem. Sot. Japan, 1979, 52, 2276. 24. X. Q. Shan and Z. M. Ni, Acta Scientiae Circumstantiae (Chinese), 1981, 1, 80. 25. S. Nishi, Y. Horimoto and R. Kobayashi, Intern. Symp. Identification and Measurement of Environmental Pollutants, p. 201. NRCC, Ottawa, 1971. 26. M. Q. Yu, G. Q. Liu and S. H. Wang, J. Enorron. SCL (Chinese), 1979, No. 5, 46. 27. M. Q. Yu and G. Q. Liu, Acta Scientiae Circumstantiae (Chinese), 1981, 1, 180. 28. Idem, J. Environ. Chem. (Chinese), 1982, No. 1, 12.
0039-9140/83/040265-06503.00/O Copyright 0 1983 Pergamon Press Ltd
Vol. 30, No. 4, pp. 265-270, 1983 III Great Bntain. All tights reserved
DETERMINATION OF TRACE ARSENIC, ANTIMONY, SELENIUM AND TELLURIUM IN VARIOUS OXIDATION STATES IN WATER BY HYDRIDE GENERATION AND ATOMIC-ABSORPTION SPECTROPHOTOMETRY AFTER ENRICHMENT AND SEPARATION WITH THIOL COTTON MU-QING Yu and GUI-QIN LIU Changchun Institute of Geography, Chinese Academy of Science
and Department
QINHAN JIN* of Chemistry, Jilin University, Changchun, People’s Republic of China
(Received 4 March 1982. Revised 9 October 1982. Accepted 20 October 1982) Summary-A novel procedure for determination of trace As(II1) and As(V), Sb(II1) and Sb(V), Se(IV) and Se(VI), Te(IV) and Te(V1) in water by atomic-absorption spectrophotometry after separation and enrichment with “thiol cotton” and hydride generation has been established. The sorption behaviour of various oxidation states of arsenic, antimony, selenium and tellurium, and the conditions of quantitative sorption and desorption of these species were studied. The procedures for reducing species from higher oxidation states were optimized. Interferences from other species and their elimination were investigated. The selectivity of the procedure for the determination of species in higher and lower oxidation states was examined. The procedure has been successfully used to determine arsenic, antimony, selenium and tellurium in water, in the range from pg/ml to ng/ml. The recoveries for added spikes were in the range 9lrllO%, with coefficients of variation in the range 3-8%.
It has been shown by more and more facts that the transport and conversion processes as well as the biological toxic effects of a variety of trace elements, such as mercury, chromium, arsenic, antimony, selenium and tellurium, in the environment are closely related to the oxidation states and compounds present. It is obvious that ignoring this fact can lead to inadequate or even totally wrong evaluation of environmental quality. For this reason the investigation of methods for specification of trace elements has been of interest for several years. Various methods for determining chromium,‘-’ arsenic,“” selenium’~*’ antimony,‘2-‘4 and tellurium22-24 in various oxidation states, after separation by co-precipitation, solvent extraction, ionexchange, hydride generation and stepwise reduction, have been established, but most of them are limited in their application. It has been shown that the thiol group has very strong affinity for heavy-metal ions and that cotton impregnated with thioglycollic acid (“thiol cotton”) quantitatively adsorbs a variety of trace elements from water,25-27 but its adsorptive power depends significantly on the oxidation state of the element concerned. In general, the lower oxidation states are readily adsorbed whereas the higher states are not. It
*To whom correspondence
should be addressed. 265
should therefore be possible to adsorb the lower states selectively, and then to reduce the higher states and adsorb the products. On the basis of investigations on the enrichment in this way of more than twenty trace species, including Pt(II), Pd(II), Au(III), Se(W), Te(IV), As(III), Hg(II), Sb(III), Bi(III), Sn(II), Ag(I), Cu(II), In(III), Pb(II), Cd(II), Zn(II), Co(I1) and Ni(II),2~28 we have established this procedure for determining trace As(III), As(V), Sb(III), Sb(V), Se(IV), Se(VI), Te(IV) and Te(V1) in water by hydride generation and atomic-absorption spectrophotometry after separation and enrichment with thiol cotton. EXPERIMENTAL Apparatus A Pye-Unicam SP 1900 atomic-absorption spectrophotometer and a home-made- hydride-generator (Fig. 1) were used, under the conditions listed in Table 1. A 250-ml (or 500-ml) Pyrex separatory funnel fitted with a thiol-cotton tube was used for separation and enrichment of various trace species. The thiol-cotton tube was a glass tube with about 0.1 g of thiol cotton in it. It was 8 mm in bore and 100 mm long, and one of the ends was narrowed. Reagents Suprapur-grade hydrochloric acid and nitric acid were used. All the other reagents were analytical-reagent grade. All water used was redistilled. Standard antimony(III) solution (loo0 ppm). Prepared from antimony potassium tartrate.
MU-QINGYu et al.
266
I
I
I
4
I
--
--
6 i_-J
Fig. 1. Eqmpment for hydride-generation determmation. 1, N, cylinder; 2, flowmeter; 3, hydride generator (25 mm bore, 80 mm long); 4, light durce; 5, quartz-tube atomizer (8 mm bore, 150 mm long); 6, detector; I, syringe.
Standard antimonyp)
oxidizing the standard manganate.
solution (10 ppm). Prepared by antimony(II1) solution with per-
Standard arsenic(III) solution (1000 ppm). Prepared from
arsenious oxide. Standard arsenic(V) solution (la00 ppm). Prepared from
sodium arsenate. Standard selenium(W) solution (1 OOOppm).Prepared from
selenium dioxide. Standard
tellurium(IV)
solution (1000 ppm).
Prepared
solution (1000 ppm).
Prepared
from tellurium powder. Standard
tellurium(W)
from telluric acid. Potassium iodide (20x)-thiourea (2%) solution. Potassium iodide (20%)~ascorbic acid (4%) solution. Potassium borohydride solution (2%). Prepared in 0.5%
sodium hydroxide solution. Titanium(III) chloride, 15-20x solution. Hydrogen peroxide, 30% solution. Thiol cotton. Thioglycollic acid (50 ml), acetic anhydride
(35 ml), glacial acetic acid (16 ml), concentrated sulphuric acid (0.15 ml) and water (5 ml) were put in that order into a bottle and mixed well, and then 15 g of absorbent cotton were added. The bottle was closed, then left in an oven at 4&50” for 4-5 days. The cotton was collected on a sinteredglass funnel and washed with water till the washings were neutral, then dried at 40-50” and kept in a brown bottle. Procedures Separation and enrichment. Put a selected volume of water sample into the separation assembly, adjust its acidity to the
value indicated in Table 2 and then pass it through the thiol-cotton tube at a suitable flow-rate to extract the Sb(III), As(III), Se(IV) or Te(IV). Treat the effluent solution with the appropriate reagents (as indicated in Table 3) to reduce Sb(V), As(V), Se(V1) or Te(V1) to the next lower oxidation state, and collect the reduced species in another thiol-cotton tube. Desorb the As(III), Sb(III), Se(IV) or Te(IV) as follows. Arsenic. Elute slowly with 3 ml of hot concentrated hydrochloric acid. Collect the eluate in a lo-ml graduated tube. Add 0.2 ml of potassium iodide solution to the eluate and dilute to 10 ml with water. Antimony. Elute slowly with 3 ml of 5M hydrochloric acid into a lo-ml graduated tube. Add 0.2 ml of potassium iodide solution to the eluate and dilute to 10 ml with water. Selenium. Put the cotton into a lo-ml graduated tube. Add 2.5 ml of concentrated hydrochloric acid and 1 drop of concentrated nitric acid. Heat in a boiling water-bath for 3 min. Cool to room temperature and dilute to 10 ml with water. Tellurium. Place the cotton in a lo-ml graduated tube. Add 2.5 ml of concentrated hydrochloric acid. Heat in a boiling water-bath for 3 min. Cool to room temperature and dilute to 10 ml with water.
Use a separate sample for each element, but if desired, antimony(II1) can be selectively eluted first. Determination. Set the AAS instrument according to Table 1. Place a known amount of sample solution in the
Table 1. Conditions for hydride generation and AAS determination Parameter Light-source Wavelength, nm Lamp current, mA Slit-width setting Temperature of quartz tube, “C Flow-rate of carrier-gas (NJ, l./min Acidity of solution analysed, N
Sb
As
Se
Te
Pye hollow-cathode lamps 217.6 193.7 196.0 214.2 12 10 6 10 0.2 0.2 0.2 0.2 900 900 900 900 0.6
0.6
0.6
0.6
1.5
3
3
3
Determination
of trace arsenic etc.
Table 2. Acidity for enriching As(III), Sb(III), Se(W) and Te(IV) Species
Acidity of water sample
Sb(II1) As(III) Se(IV) Te(IV)
pH 6-8 1M HCl 0.3M HCl 0.3M HCl
hydride generator and make up to 5 ml with hydrochloric acid of the same concentration as the sample. Close the generator and with a syringe rapidly add 1.5 ml of potassium borohydride solution to the lower part of the solution. From the recorded absorbance calculate the content of the species in the sample.
261
reduced by the thiol group and then adsorbed by the thiol cotton. The degree of reduction depends on the species concerned and the acidity but is never quantitative. For example, Sb(V) is partly reduced to Sb(II1) in dilute hydrochloric acid. Therefore, this species can be adsorbed significantly (- 50% maximum) at pH < 5, but Se(VI), Te(V1) and As(V) can only be slightly reduced and adsorbed at acidities below SM hydrochloric acid. The range of acidity for quantitative adsorption shown in Figs. 2 and 3 will change slightly with the composition of the solution, the concentration of the trace element and the quality of the thiol cotton. The conditions shown in Table 2 are recommended as optimal.
RESULTS AND DISCUSSION
Adsorption behaviour of thiol cotton towards arsenic, antimony, selenium and tellurium
It is clear from Figs. 2 and 3 that the adsorption behaviour of those elements depends greatly on the acidity of the solution and the oxidation state of the element. Thiol cotton efficiently adsorbs only As(III), Sb(III), Se(IV) and Te(IV), the ranges of acidity for quantitative adsorption being Sb(III), pH 8-2M HCl; As(III), 0.557M HCl; Te(IV) pH 8-9M HCl; Se(IV), pH 2-10M HCl. It was also noticed that within certain ranges of hydrochloric acid concentration, some species in higher oxidation states could be
Choice of desorption procedures
The adsorptive power of thiol cotton for Sb(II1) is rather poor, and this species can be desorbed quantitatively with SM hydrochloric acid. To desorb As(III), the thiol cotton should be dipped into concentrated hydrochloric acid for several minutes. Se(IV) and Te(IV) are adsorbed very firmly by thiol cotton, so it is necessary to heat the cotton with concentrated hydrochloric acid in a boiling waterbath to desorb these species quantitatively; this process also attacks the cotton itself.
Table 3. Conditions for reducing As(V), Sb(V), Se(W) and Te(VI) Species
SW’) As(V) Se(W) Te(VI)
Amount of reductant used, and reduction procedure
Acidity
0.3M HCl Add KI to 0.1% and ascorbic acid to 0.02%; let stand for 10 min at room temperature 1M HCl Add KI to 0.2% and thiourea to 0.02%; heat for 3-5 min in a boiling water-bath (the temperature of the solution reaches 80”) 0.3M HCl Add TiCl, to 0.1%; after 30 min add H,O, to 0.1% and heat for 5 min in a boiling water-bath 0.3M HCl Add TiCl, to 0.1%; after 10 min add H,O, to 0.1% and let stand for 5 min at room temperature
IO
6
6
4
2
OIN
I Fig 2.
The
Acldlty 3
5
7
PH
effects of acidity on the adsorption of As(M), As(V), Sb(III) and Sb(V) (all 10 ng/ml concentration) on thiol cotton. (0) As(M); (0) As(V); (A) Sb(III); (A) Sb(V).
268
MU-QING
IO
8
6
4
Yu ef al.
2
OIN
Acldlty
Fig. 3. The effects of acidity on the adsorption of Se(N), %@I), Te(IV) and Te(V1) (all 10 ng/ml concentration) on thiol cotton. (0) Se(N); (0) Se(V1);(A) Te(IV); (A) Te(VI). The procedure for determination of arsenic, antimony, selenium and tellurium by hydride-generation atomic-absorption spectrophotometry is also related to the oxidation state. Therefore, the oxidation state of the species to be determined in the desorbed solution should be consistent with the r~uirements of this technique. For this reason, a small amount of nitric acid is added when selenium is to be desorbed, to ensure the selenium is in the quadrivalent state. Choice of reduction procedures
Since As(V), Sb(V), Se(VI) and TefVI) must be reduced to As(III), Sb(III), Se(IV) and Te(IV) before they can be quantitatively adsorbed by thiol cotton, the reduction procedure should be compatible with the enrichment technique. Those selected are as follows. SbfV) can be rapidly reduced by adding potassium iodide to the dilute hydrochloric acid sample solution. As(V) is more difficult to reduce than Sb(V). It is necessary to add potassium iodide and thiourea to a sample solution that is 1M in hydrochloric acid and heat in a boiling water-bath for several minutes. Se(VI) and Te(V1) are diflicult to reduce selectively to Se(IV) and Te(IV) because they are very easily reduced to the elements in hydrochloric acid medium. We have found that if enough titanium trichloride to reduce Te(VI) to Te(IV) and Se(V1) to Se is added to a dilute hydrochloric acid solution followed by enough hydrogen peroxide to oxidize the excess of titanium trichloride and to form a complex with it, a stable Te(IV) solution is formed and the elemental selenium is oxidized to Se(N). The concentration of hydrochloric acid, the amounts of titanium trichloride and hydrogen peroxide used and the temperature and time of reduction all affect the efficiency of reduction. Table 3 shows the recommended reduo tion conditions. This reduction procedure is simple, rapid and easy to operate, and Se(V1) and Te(V1) can be converted into Se(IV) and Te(IV) simultaneously and quantitatively.
Amount of thiol cotton to use andflow-rate of sample solution The adsorption capacity of thiol cotton for the species concerned and the flow-rates for quantitative adsorption are shown in Table 4. The operational adsorption capacities are between a third and a half of the corresponding saturation capacities. On the basis of the concentration ranges of these species that can be adsorbed by thiol cotton and those existing in practical environmental water samples, about 0.1 g of thiol cotton is adequate for enrichment. Selectivity for higher or lower oxidation states The selectivity for enriching and determining species in their higher or lower oxidation states is quite good. The maximum ratios allowed between higher and lower oxidation-state species, for an error of lo%, are shown in Table 5. To obtain selective determination the following measures should be taken. 1. The thiol-cotton tube should first be rinsed with water or with hydrochloric acid of appropriate concentration to bring it to the optimum state for the adsorption. 2. If the concentration ratio of lower to higher oxidation state species in the water sample is very high, to adsorb the lower-valence species the amount of thiol cotton used should be increased accordingly and the flow-rate should be reduced. 3. If the higher to lower oxidation state concentration ratio is high, the lower-valence species ad-
Table 4. Saturation adsorption capacity of thiol cotton and Row-rate of water sample As(W)
SbfIII)
k-5
5-l
Saturation capacity, rng/g of thiol cotton Flow-rate, ml/min* *For 0.1 g of thiol cotton.
G2
Se(IV) Te(IV) 7-9 65
8-11
Determination
of trace arsenic etc.
Table 5. Selectivity for determining As, Sb, Se and Te at higher or lower valence resnectivelv.* with X+lo”/, error Maximum ratio of higher valence/lower valence species for determination of lower valence species Maximum ratio of lower valence/higher valence species for determination of higher valence species
As
Sb
Se
Te
30
20
40
50
20
20
20
30
*The volume of water sample enriched was 100-200 ml and the concentration of As, Sb. Se or Te was 24 ng/ml. sorbed on the thiol cotton should he washed with water or the appropriate concentration of hydrochloric acid to remove any residual higher oxidation state species on it. Interference
studies
It has been shown
that
except
for some
slight
269
interference in adsorption of Sb(II1) from higher concentrations of humic acid, the species usually found in environmental water, such as Cl-, Brr, SO:-, SO;-, CO:-, HCO;, NO;, NO;, HPO:-, K+, Na+, Ca*+, Md+, Fe’+, A13+ and various heavy metal ions, as well as fulvic acid, glycine, cysteine, citric acid and the like, do not interfere in the quantitative adsorption of As(III), Sb(III), Se(IV) and Te(IV) by thiol cotton. It has also been shown that the direct determination of As, Sb, Se and Te by hydride-generation atomic-absorption spectrophotometry would suffer interferences from Co*+, Ni*+, Fe3+, Cu*+, Pb*+, Sn4+, Bi3+ and Ag+ ions, and between the four elements themselves, but after enrichment with thiol cotton and stepwise desorption these interferences are effectively eliminated. Co’+, Ni*+, Cu*+, Sn4+, Bi3+, Pb2+ and Ag+ adsorbed by thiol cotton can be removed quantitatively by eluting with l-2M hydrochloric acid.
Table 6. Analysis of real water samples and recovery of the procedure Found Lower valency Standard added, ng/ml Element Arsenic
Spring water Natural river water Polluted river water
Antimony
Natural river water Polluted lake water
Polluted river water
Selenium
Well water Natural river water Polluted river water
Tellurium
Natural river water Sea-water Polluted river water
*Average of three replicates.
Lower valency
Higher valency
Total, nglml
0 4.00 0 2.00 0 2.00 6.00 10.00
0 4.00 0 2.00 0 10.00 6.00 2.00
3.05 6.95 0.40 2.54 6.58 8.80 12.3 15.7
0 1.oo 2.00 0 1.oo 3.00 0 1.00 1.00 2.00
0 2.00 1.oo 0 3.00 1.00 0 2.00 1.00 1.00
0 0.50 0 1.00 0 1.50 1.00 0.50
0 0.50 0 1.oo 0 0.50 1.00 1.50
0.04 0.58 0.10 1.18 0.28 1.66 1.27 0.77
0 2.00 1.00 0 2.00 1.00 0 0.50 1.00 2.00
0 1.oo 2.00 0 1.oo 2.00 0 2.00 1.oo 0.50
Recovery of added standard,* 0’ /a 98 107 115 97 92 95 88 90 93 87 95 88
Higher valency
Total, nglml 5.44 9.90 0.76 2.60 135 22.0 19.4 15.4 0.26 2.30 1.30 1.10 4.25 2.00 0.65 2.90 1.68 1.73
Recovery of added standard,* 9/, 117 92 85 98 95 102 104 98 90 112 103 108
0.64
108 108 92 99 98 98 102 96 112 110 101 99
1.15
102 95 110 93 86 98 105 100 105 96 87 104
MU-QINGYu et al.
270
Table 7. Sensitivity and precision of the procedure Sensitivity (1% absorbance),* ng/ml Detection limit,? ng/ml Lower valency Coefficient of Higher valencv variation.6 ‘/,
As
Sb
Se
Te
0.11 0.006 3.5 5.4
0.09 0.005 7.9 5.0
0.20 0.010 2.8 3.2
0.16 0.008 2.7 3.4
*The volume analysed was 5 ml.
tThe volume of water sample enriched was 200 ml. §Based on 11 samples.
To isolate antimony, the thiol cotton is eluted with acid. As, Se and Te are retained and
5M hydrochloric
the Sb desorbed. When selenium is being determined, any coexisting arsenic is oxidized to As(V) and so does not interfere. When arsenic is being determined, the only interference is from selenium, but this can be eliminated by adding a small amount of potassium iodide. The interference from selenium and arsenic in determination of tellurium can be largely eliminated by enrichment with thiol cotton and desorption. Sensitivity
and precision
As, Sb, Se and Te in various oxidation states in various environmental water samples were determined by this procedure. The recoveries were in the range 90-110x, as shown in Table 6. The sensitivity and the coefficient of variation (11 replicates) for these elements in the range 2-4 ng/ml are shown in Table 7. Conclusions
The procedure has the following characteristics. 1. The enrichment efficiency is high, the enrichment factor ranging up to several hundred. The detection limit can be as low as a few rig/l.. 2. The separation procedure is simple and the higher and lower oxidation-state species can be separated and enriched separately. Therefore the error introduced by the subtraction technique adopted in some other methods is avoided. 3. The selectivity is good. There is no mutual interference between higher and lower valence species when their concentration ratio is not more than 20-50, depending on the particular system. 4. The accuracy is good. Interference from various concomitant species is eliminated and the method can be used to analyse various environmental water samples.
REFERENCES
1. T. Goto and S. Ginba, Bunseki Kaguku, 1974, 23, 517. 2. T. Matsuo, J. Shida, M. Abiko and K. Konno, ibid., 1975, 24, 723. 3. K. Hiiro, T. Owa, M. Takaoka, T. Tanaka and A. Kauahara, ibid., 1976, 25, 122. 4. M. Q. Yu and Y. J. Zhao, J. Environ. Sci. (Chinese), 1979, No. I, 20. 5. S. R. Wang, F. Z. Xu, H. F. Zhou and X. L. Jin, ibid., 1980, No. 3, 11. 6. T. Kamada, Talanta, 1976, 23, 835. 7. J. Aggett and A. C. Aspell, Analyst, 1976, 101, 341. 8. D. J. Myers and J. Osteryoung, Anal. Chem., 1973, 45, 267. 9. F. T. Henry, T. 0. Kirch and T. M. Thorpe, ibid., 1979, 51, 215. 10. Y. Talmi and D. T. Bostic, ibid., 1975, 47, 2145. 11. M. 0. Andreae, ibid., 1977, 49, 820. 12. T. Kamada and Y. Yamamoto, TaZanta, 1977,24, 330. 13. I. Valente and H. J. M. Bowen. Analvst. 1977. 102.842. Anal. ?him. 14. K. S. Subramanian and J. C.‘Mer&gx, Acta, 1981, 124, 131. 15. 0. Yoshil, K. Hiraki, Y. Nishikawa and T. Shigematsu, Bunseki Kagagu, 1971, 26, 91. 16. Y. Shimoshi and K. Toei, Anal. Chim. Acta, 1978, 100, 15. 17. T. Kamada, T. Shiraishi and Y. Yamamoto, Talanta, 1978, 25, 15. 18. G. A. Cutter, Anal. Chim. Acta, 1978, 98, 59. 19. Y. Sugimura and Y. Suzuki. J. Oceanogr. Sot. Japan, 1977, 33, 23. 20. C. I. Measures and J. D. Burton, Anal. Chim. Acta, 1980, 120, 177. 21. H. J. Robberecht and R. E. Van Grieken, Anal. Chem., 1980, 52, 449. 22. T. Kamada, N. Sugita and N. Yamamoto, Talanta, 1979, 26, 337. 23. K. Jin, M. Taga, H. Yoshida and S. Hikime, BUN. Chem. Sot. Japan, 1979, 52, 2276. 24. X. Q. Shan and Z. M. Ni, Acta Scientiae Circumstantiae (Chinese), 1981, 1, 80. 25. S. Nishi, Y. Horimoto and R. Kobayashi, Intern. Symp. Identification and Measurement of Environmental Pollutants, p. 201. NRCC, Ottawa, 1971. 26. M. Q. Yu, G. Q. Liu and S. H. Wang, J. Enorron. SCL (Chinese), 1979, No. 5, 46. 27. M. Q. Yu and G. Q. Liu, Acta Scientiae Circumstantiae (Chinese), 1981, 1, 180. 28. Idem, J. Environ. Chem. (Chinese), 1982, No. 1, 12.