Determination of total tin in geological materials by electrothermal atomic-absorption spectrophotometry using a tungsten-impregnated graphite furnace

Talanra, Vol. 31, No. 1, pp. 73-76, 1984 Printed in Great Britain

0039-9140/84 $3.00 + 0.00 Pergamon Press Ltd

SHORT COMMUNICATIONS DETERMINATION OF TOTAL TIN IN GEOLOGICAL MATERIALS BY ELECTROTHERMAL ATOMIC-ABSORPTION SPECTROPHOTOMETRY USING A TUNGSTEN-IMPREGNATED GRAPHITE FURNACE LIYI ZHOU,* T. T. CHAot and A. L. MEIER U.S. Geological Survey, Box 25046, Federal Center, Denver, CO 80225, U.S.A. (Received 23 June 1983. Accepted 16 August 1983)

Summary-An electrothermal atomic-absorption spectrophotometric method is described for the determination of total tin in geological materials, with use of a tungsten-impregnated graphite furnace. The sample is decomposed by fusion with lithium metaborate and the melt is dissolved in 10% hydrochloric acid. Tin is then extracted into trioctylphosphine oxide-methyl isobutyl ketone prior to atomization. Impregnation of the furnace with a sodium tungstate solution increases the sensitivity of the determination and improves the precision of the results. The limits of determination are 0.5-20 ppm of tin in the sample. Higher tin values can be determined by dilution of the extract. Replicate analyses of eighteen geological reference samples with diverse matrices gave relative standard deviations ranging from 2.0 to 10.8% with an average of 4.6”. Average tin values for reference samples were in general agreement with, but more precise than, those reported by others. Apparent recoveries of tin added to various samples ranged from 95 to 111% with an average of 102%.

Tin occurs in geological materials either as a constituent of the silicate lattice,lm3 or in the form of cassiterite.‘,’ Knowledge of the relative distribution of tin in the two forms is important in studies related to the geochemistry of tin and in geochemical prospecting for tin.‘z3 Whereas cassiterite in geological samples can be converted into tin(IV) iodide and volatilized by heating with ammonium iodide,‘,“’ total tin can only be released by fusion with an alkaline flux such as lithium metaborate.8-‘0 The tin released may be determined by atomic-absorption spectrophotometry, as ti.e hydride”” or through aspiration of an organic tin extract into a nitrous oxide-acetylene flame,4,5 or by ICP atomic emission spectroscopy following hydride generation.6 The difference between the total tin determined by fusion with lithium metaborate and the cassiterite tin would give an approximate value for the tin held in the lattice of silicates. The proposed method involves fusion of the sample with lithium metaborate to release the total tin, which is extracted with trioctylphosphine

An Instrumentation Laboratory (IL) Model 951 atomicabsorption spectrophotometer was used for this study.3 The spectrophotometer was equipped with the following accessories: controlled-temperature furnace atomizer (IL 555 CTF), auto-sampling device (IL 254 FASTAC), background corrector, and tin hollow-cathode lamp. Settings for the spectrophotometer were: lamp current 10 mA; wavelength 286.3 nm; band-width, 0.5 nm; integration time, 8 set; read-out mode, peak height; photomultiplier voltage (HV), adjusted by the HV control until the log intensity meter reads between 0.2 and 0.8 V. Settings for the controlled-temperature furnace atomizer were: purge gas, nitrogen at 6.9 l./min; auto operation

*Geochemist, on leave from the Institute of Geophysical and Geochemical Prospecting, Beijing, China. tTo whom correspondence should be addressed. $Use of brand names in this paper is for descriptive purposes only and does not constitute endorsement by the U.S. Geological Survey.

Step Temp., “C Time, x 5 set

oxide-methyl isobutyl ketone (TOPSMIBK), by the procedure of Burke.” The tin in the organic extract is atomized in a tungsten-impregnated graphite furnace. The sensitivity, precision, and accuracy of the method are desirable features for the determination of tin in geological samples of diverse chemical composition.

EXPERIMENTAL Apparatus

mode; temperature-feedback programme as follows:

on; auto-clean

Drying

73

1 0 0

2 150 1

Charring 3 750 4

off; atomization

Atomizing

4 5 6 1000 2500 2500 4 0 2

14

SHORT

COMMUNICATIONS

Settings for the auto-sampling device were: “door calibration”, 150”; sample deposit time, 2 set; delay time, 5 sec. The FASTAC sample-delivery system aspirates the sample through a pneumatic nebulizer and converts it into an aerosol in the graphite furnace. Impregnated graphite tubes. The impregnation of pyrolytic graphite tubes was done following the general guidelines of Fritzsche et al.‘* The tubes were soaked overnight in a 7.8% solution of sodium tungstate dihydrate and dried at 120” in an electric oven for 4 hr. Before use, each impregnated tube was conditioned by application of the atomization programme five times.

furnace/atomic-absorption spectrophotometer system. Correct the results for the blank and for the tin present as a contaminant in the lithium metaborate (as described under “sample decomposition” below). RESULTS AND

Sample decomposition Fusion with lithium metaborate as a flux has been shown to decompose geological materials, bringing both lattice-bound tin and cassiterite tin into solution for the determination of total tin in the sample.8-‘0 The lithium metaborate used in this study was analysed by the proposed procedure for possible contamination by tin, and found to contain 0.74 + 0.017 ppm Sn (ten replicates). It follows that a correction must be applied for tin in the flux used. There are two ways of doing this: (a) running a complete blank (including the fusion) for each batch of samples, taking care to use virtually identical weights of flux for all tests, or (b) determining a reagent blank (excluding the flux), analysing the lithium borate for tin (and correcting for the reagent blank), and correcting the results for the samples by subtraction of both the reagent blank and the tin contained in the flux (which must be weighed accurately for each sample). The result is then calculated by means of the formula

Reagents Ascorbic acid. Medium fine crystals (3&80 mesh). TOP0 Ctrioctvlvhosvhine oxidetMIBK Cmethvl isobutvl ketone). Dissolve-4.0 g of TOPO’in 100 ml of MIBK. . Lithium metaborate. Anhydrous powder (G. Frederick

Smith Chemical Company). Stock and standard solutions. Prepare the 1000~yg/ml stock tin solution by dissolving 0.5000 g of reagent-grade tin metal in 250 ml of 50% v/v hydrochloric acid and diluting to 500 ml with water. Prepare the I-pg/ml standard tin solution in 10% v/v hydrochloric acid by serial dilution of the stock tin solution. The dilute standard tin solution is stable for at least 6 months in a Pyrex glass bottle. Organic tin standard solutions in TOPO-MIBK (Sn 0, 0.025, 0.050, 0.075, 0.100, 0.200, 0.300, 0.400 and 0.500 pgglml). Using Eppendorf micropipettes, transfer 0, 0.25,

0.50, 0.75, 1.00, 2.00, 3.00, 4.00 and 5.00 of the 1 pg/ml standard tin solution to individual 25 x 200 mm screw-cap tubes and make up to a total volume of 40 ml with 10% v/v hydrochloric acid. Scoop 0.75 g of ascorbic acid into each tube and use a vortex mixer to dissolve the solid. Add exactly 10 ml of the TOPO-MIBK solution to each tube, shake for 1 min, and centrifuge the capped tubes to separate the organic phase. Transfer the organic tin extracts to 16 x 150 mm test-tubes and cap them to prevent evaporation. The tin in the extract is stable for at least 1 week if kept in a refrigerator.

ppm Sn = [(tin in extract o( g/ml) - tin in fluxless blank (p g/ml)) x MIBK volume (ml)] - tin in flux kg) sample weight (g)

Procedure

Weigh 0.750 g of lithium metaborate and 0.250 g of
DISCUSSION

Sensitivity The characteristic mass (weight of analyte giving 1% absorption) of the method described above has been found to be 7.0 pg. With a 0.250-g sample, the range of tin concentrations that can be determined is 1.00-20.0 ppm in the sample. If the volume of TOPO-MIBK used for extraction is reduced to 5 ml, the lowest tin concentration that can be accurately determined is 0.50 ,ppm, which is considered to he below the crustal abundance.” The calibration graph for tin is linear up to .0.200 pg/ml, corresponding to an absorbance of 0.689 (Table l), above which curvature occurs. When the concentration of tin in the sample exceeds 20.0 pptu,

Table 1. Effect of tungsten-impregnation on absorbance readings of tin in standard solutions W-impregnated Standard, fi.cnlml 0.025 0.050 0.075 0.100 0.200

Absorbance 0.098 0.193 0.277 0.363 0.689

*Calculated from three injections.

Untreated

R.S.D..* “/, 0.8 1.1 1.6 2.3 2.3

Absorbance 0.046 0.108 0.143 0.200 0.417

R.S.D. .* “/” 6.5 10.1 5.1 9.1 8.9

SHORT COMMUNICATIONS

Table 2. Replicate determinations

(n = 5) of tin in various U&Geological

15

Survey reference samples

This work R.S.D., % Rock reference samples 1.74 f 0.08 4.6 1.90 f 0.12 1.90 * 0.07 2.98 f 0.19 2.34 f 0.22 El 2.14 f 0.17 3.65 f 0.12 3:3 3.78 f 0.29 3.22 + 0.07 2.2 3.02 f 0.18 2.83 f 0.09 3.2 2.68 f 0.16 2.31 f 0.07 3.0 1.38 f 0.18 9.02 * 0.22 2.4 6.70 + 0.34 Geochemical exploration reference samples Mean, ppm

Sample BHVG-1, basalt MAG-1, marine mud QLO-1, quarts latite RGM- 1, rhyolite SCo-1, Cody shale SDC-1, mica shist SGR-1, oil shale STM- 1, nepheline syenite GXR- 1, jasperoid GXR-2, soil GXR-3, Fe-Mn deposit GXR-4, Cu mill head GXR-5, soil GXR-6. soil

52.4 + 1.93 1.98f 0.06 0.94 f 0.07 4.79 f 0.25 2.84 f 0.09 0.86 f 0.06

GSB, glass standard GSC, glass standard GSD, glass standard GSE, glass standard

0.74 f 0.08 3.62 + 0.15

10.8 4.1

62.2 + 1.27 433 i21.9.

2.0 5.1

Reported values, ppm 1.8’O 3.2 2.4 4.4 3.9 3.2 6.7

(6.3)*” (6.4) (7.9) (7.7) (6.0) (6.0) (13.0) (5.1)

2.15’s 5.0 2.3-4.2 3.9 4.1 3.0 1.58 10+3

3.7 3.0 7.4 5.2 3.2 7.0 Glass reference standards 0.616 5

43 440

*R.S.D. (%) in parentheses.

the sample extract should be diluted with MIBK so that the absorbance is within the linear range of the calibration graph, preferably around 0.500, to reduce experimental error. Effect of tungsten -impregnation on the performance of the graphite tube Impregnation of the graphite tube, with sodium tungstate solution nearly doubles the absorbance readings for tin standards (Table 1) thereby enhancing the sensitivity, and also improves the precision of the absorbance readings (Table 1). The tungstenimpregnated graphite tube can be used for at least 500 firings. Interferences The chelation and extraction of tin from the sample solution with TOPG-MIBK is adapted from

Table 3. Recovery of known amounts of tin added to various reference samples (averages of duplicate analysis of 0.250-a- samdes) -,

Present, Sample

Icg

Added, N

Found, fig

RGM- 1

0.93 0.93 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.18 0.18

1.00 2.00 1.00 2.00 3.00 5.00 1.00 2.00 3.00 5.00 0.20 0.40

1.92 2.89 1.74 2.85 3.69 5.95 1.89 2.73 3.74 5.83 0.41 0.55

SDC- 1

GXR-5

GSB

Recovery, % 100 99 102 106 100 104 111 101 101 102 107 95

the procedure of Burke” for the determination of tin in ahnninium, iron and nickel-base alloys. The same extraction has been used for determining tin in geological materialsS and in mineralized rocks and ores.’ Welsch and Chaos found no interferences from 1000 pg of Cu, Pb, Zn, Mn, Hg, MO, V, and the equivalent of 20% Fe in a l-g sample in the flame atomic-absorption determination of tin. It is inferred that interferences in the graphite-tube atomicabsorption determination of tin would be minimal. This is supported by the data of Tables 2 and 3 showing the general agreement of the tin values found for the reference samples with those reported by others, and the very good recoveries of tin added to various samples of diverse chemical composition. Results for geological reference samples The proposed method was applied to three sets of U.S. Geological Survey reference samples: (1) eight rock standardal (2) six geochemical exploration reference samples,i5 and (3) four glass reference standards.16 The average tin values obtained for the eight USGS reference rocks are in most cases in general agreement with, but more precise than, the values reported by Terashima,” as shown in Table 2. Replicate analyses of eighteen reference samples with various matrices gave relative standard deviations ranging from 2.0 to 10.8x, with an average of 4.6% (Table 2). Recoveries of tin added to various samples ranged from 95 to 111% with an average of 102% (Table 3). Thus, the proposed method can be applied to the determination of total tin in a wide range of geological materials of diverse chemical composition.

16

SHORT

COMMUNlCATlONS

REFERENCES

1. L. H. Ahrens and W. R. Liebenberg, Am. Mineral., 1970, 35, 571. 2. J. Agterdenbos and J. Vlogtman, Talanta, 1972, 19, 1295. 3. A. M. R. Neiva, Min. Mug., 1976, 40, 453. 4. J. D. Mensik and H. J. Seidemann, Jr., At. Absorp. Newsl., 1974, 13, 8. 5. E. P. Welsch and T. T. Chao, Anal. Chim. Acta, 1976, 82, 337. 6. B. Pahlavanpour, M. Thompson and S. J. Walton, J. Geochem. Explor., 1979, 12, 45. 7. D. Gladwell, M. Thompson and S. J. Wood, .I. Geothem. Explor., 1981, 16, 41. 8. A. Hall, Chem. Geol., 1980, 30, 135. 9. K. S. Subramanian, Talanta, 1980, 21, 469.

10. C. Y. Chan and M. W. A. Baig, Anal. Chim. Acta, 1982, 136, 413. 11. K. E. Burke, Analyst, 1972, 97, 19. 12. H. Fritzche, W. Wegscheider, G. Knapp and H. M. Ortner, Talanta, 1979, 26, 219. 13. Tan Lee and Chi-lung Yao, Intern. Geol. Rev., 1970, 12,

778. VS. Geol. Survey Prof. Paper, 840, 1976. 15. G. H. Allcott and H. W. Lakin, Geochemical Exploration 1974, p. 659. Elsevier, Amsterdam, 1975.

14. F. J. Flanagan,

16. A. T. Mvers. R. G. Havens. J. J. Connor. H. M. Conklin and H. J. Rose, U.S. Geol. Survey Proj Paper, 1013, 1976.

17. S. Terashima, Geostd. New& 1982, 6, 77. 18. E. S. Gladney and W. G. Goode, ibid., 1981, 5, 3 1.