Analyst, December, 1968, Vol. 93, pp. 799-801 799 Determination of Antimony in Titanium Dioxide by Atomic-absorption Spectrophotometry BY J, C. MERANGER AND E. SOMERS (Research Laboratories, Food and Drug Directorate, Department of National Health and Welfare, Ottawa, Canada) A method is described for the determination of the antimony content of titanium dioxide by atomic-absorption spectrophotometry, based on the extraction of antimony with isobutyl methyl ketone. This method is more sensitive than determination by emission-spectrographic and as sensitive as the colorimetric technique, and can detect 25 p.p.m. of antimony in a 0-1-g sample of titanium dioxide. THE titanium dioxide used in pharmaceutical preparations, and as a food colour, may be contaminated with antimony, and there are several official specifications that give limits for the total, or acid-extractable, antimony.Spectrophotometric determination of the antimony in titanium dioxide with Brilliant green has been described by Galliford and Yardley,l and Ratcliffe and Stevens2; the latter workers showed that many of the pharmaceutical samples of titanium dioxide examined exceeded the British Pharmaceutical Codex limits for antimony. Atomic-absorption spectrophotometry was used by Mostyn and Cunningham3 for the determination of antimony in non-ferrous alloys and, in view of the convenience and sensitivity of this technique, we have applied it to the determination of antimony in titanium dioxide. The atomic-absorption analysis has been compared with colorimetric and spectro- graphic methods, and we have determined the effectiveness of dilute hydrochloric acid in extracting antimony from titanium dioxide.EXPERIMENTAL SPECTROGRAPHIC AKALYSIS- A Jarrell-Ash, Model 15000 Lpl, 1*5-m, wide-angle, grating emission spectrograph was used. Titanium dioxide samples (10 mg) were mixed with 30 mg of spectrographic-grade graphite and introduced into 3.2-mm electrodes with 2 x 15-mm craters, with a 3.2-mm counter electrode. The samples were arced for 25 seconds at 4 and 6A to completion; the resulting spectra were recorded on Kodak SA-1 film. The intensity of the antimony 259.8 nm line in the unknown samples was compared with that obtained from mixtures of spectro- scopically pure antimony tetroxide and titanium dioxide. FUSION OF TITANIUM DIOXIDE- The method followed was that of the Food and Agricultural Organisation Specifications for Identity and Purity of Food Additives4 Titanium dioxide (0.1 g) was fused with 5 g of potassium hydrogen sulphate and 10mg of glucose in 30-ml Kjeldahl flasks.The melt was dissolved by heating with 10 ml of sulphuric acid (96 to 98 per cent. w/v) then, when cool, made up to 100 ml with 30 per cent. w/v hydrochloric acid. COLORIMETRIC DETERMINATION- Aliquots of the above solution, containing from 1 to 1Opg of antimony, were oxidised with solid cerium(1V) sulphate, extracted with di-isopropyl ether, and the optical density of the complex formed with Rhodamine B measured at 555 nm.4 0 SAC and the authors.800 MERANGER AND SOMERS: DETERMINATION OF ANTIMONY IN TITANIUM [Artahst, VOl. 93 ATOMIC ABSORPTION- A Perkin-Elmer, Model 303, atomic-absorption spectrophotometer, equipped with a Boling burner and a null recorder read-out accessory coupled to a Westronics, Model S11A/U, 11-inch strip-chart recorder, was used. The operating parameters were : acetylene flow- rate, 6; compressed air flow-rate, 13; solution uptake, 3.2 ml per minute; slit, position 4; wavelength, 217.6 nm; scale expansion, x 3 ; meter response, 4; and antimony hollow- cathode lamp, current 30mA.STANDARD SOLUTIONS- Standards of 1, 10 and 100 pg of antimony per ml were freshly prepared by dissolving analytical-reagent grade potassium antimony tartrate, KSbC,H,O,. 1/2H,O, in hydrochloric acid (30 per cent. w/v). EXTRACTION WITH HYDROCHLORIC ACID- The extraction method is that given in the U.S.Food and Drug Administration Color Additive regulation^.^ Titanium dioxide (10 g) was boiled for 15 minutes in 50 ml of 0.5 N hydrochloric acid, then samples of the solution were filtered (Whatman No. 42) and made up to 100ml. RE s ULTS SOLVENT EXTRACTION OF ANTIMONY- When the acidic solutions from the potassium hydrogen sulphate fusion of titanium dioxide were nebulised in the atomic-absorption spectrophotometer high results were obtained, probably because of light-scattering by salt particles. In addition, this method was not sensitive enough to determine the low levels of antimony present in food-grade titanium dioxide. To overcome these inadequacies, an extraction procedure was devised. The extrac- tion of the ammonium antimony pyrrolidinedithiocarbamate complex* was attempted with isobutyl methyl ketone.However, the chelate was not formed under the conditions of high acidity used in this work. Chelation was subsequently found to be unnecessary as isobutyl methyl ketone quantitatively extracted the antimony( 111) or (V) directly from the aqueous acidic solutions. Consistent extraction of a standard antimony solution was obtained from solutions of hydrochloric and sulphuric acids in the acid concentration range of that of the dissolved, fused titanium dioxide (Table I). TABLE I WITH ISOBUTYL METHYL KETONE (10 ml), BY ATOMIC ABSORPTION EFFECT OF ACID CONCENTRATION ON THE EXTRACTION OF 50p.g OF ANTIMONY Hydrochloric acid, Sulphuric acid, per cent. w/v per cent. w/v & - 10 12 14 6 13 Optical density* .. . . 0.042 0.041 0.041 0.041 0.042 0.043 * Average of duplicate determinations. ATOMIC-ABSORPTION DETERMINATION OF EXTRACTED ANTIMONY- Aliquots (10 ml) of the acidic solution from fused titanium dioxide were transferred, by pipette, into a 125-ml separating funnel, 5 ml of water and 1 O m l of isobutyl methyl ketone added and the funnel shaken vigorously for 1 minute. After the layers had separated, the aqueous layer was drawn off and the organic solvent directly aspirated in the atomic- absorption spectrophotometer. For samples of titanium dioxide with less than 100 pg of antimony per g, the 10-ml aliquot can be increased to 40m1, but 20ml of water, instead of 5m1, must then be added. Two calibration graphs were prepared by following the above procedure with antimony standards over the range 10 to 100pg and 0 to 1Opg.One millilitre of sulphuric acid (98 per cent. w/v) was added to the standards and they were diluted to 10 ml with 30 per cent. w/v hydrochloric acid. Blank values were obtained from the acidic solutions alone. A linear relationship between optical density and antimony concentration was found over the range 0.1 to 10 pg per ml, with a lower detection limit of 0.1 pg of antimony per ml.December, 19681 DIOXIDE BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY 801 When several samples are analysed it is desirable to have a stock solution of antimony (5 pg per ml) extracted into isobutyl methyl ketone. This solution, appropriately diluted, can be nebulised in the spectrophotometer at fixed intervals, and so slight variations in sensitivity can be normalised.The analysis of a series of synthetic mixtures of titanium dioxide and standard antimony solutions (Table 11) showed good recoveries of antimony over the range 250 to 10,OOO p.p.m. TABLE I1 ANTIMONY RECOVERIES FROM 0.1 g OF TITANIUM DIOXIDE Antimony added, pg Antimony recovered, pg* Percentage recovery 25 27 108 60 62 104 100 105 105 400 438 110 1000 905 91 * Average of duplicate determinations. COMPARISON OF ANALYTICAL METHODS- The antimony content of three commercial samples of titanium dioxide (anatase form) was detennined by emission-spectrographic, colorimetric and solvent-extraction atomic- absorption methods. The samples were chosen to represent a range of antimony content, sample A being a high quality, food-additive grade.Table I11 shows that atomic-absorption, combined with extraction, provides a method of analysis that is both consistent with, and as sensitive as, the other two methods. The lower limit of detection of antimony in a 0-1-g sample of titanium dioxide is 25 p.p.m. Extraction with hydrochloric acid removed only a very small proportion of the total antimony. TABLE I11 COMPARISON OF METHODS OF ANALYSIS FOR THE DETERMINATION OF THE ANTIMONY CONTENT OF TITANIUM DIOXIDE, AS P.P.M. Extraction with 0.6 N hydrochloric acid r Atomic Atomii Sample Spectrographic Colorimetric absorption Colorimetric absorption* A < 100 < 25 < 25 (0.1 <Om7 B loot 100 120 <om1 (0.7 C 12,600t 11,300 11,000 4.9 4.6 All values are averages of duplicate determinations.* Direct aspiration of aqueous solution, i.e., not extracted with isobutyl methyl ketone. t &lo per cent. CONCLUSIONS Until recently, the methods available for the determination of low concentrations of antimony in titanium dioxide were erratic and imprecise.lS2 Atomic-absorption spectroscopy is a rapid, accurate and interference-free analytical technique, and the isobutyl methyl ketone extraction method we have adopted combines the advantages of atomic absorption, together with sensitivity equal to the existing colorimetric methods. Acid extraction of antimony is used as the basis for official specifications of titanium dioxide.4~6 Whatever the toxicological significance of the antimony extracted by hydro- chloric acid may be, it is clear that this technique gives no useful information as to the total antimony content of the sample. We are grateful to Mr. R. E. Horton and Mr. C. C. Durham of the Department of Energy, Mines and Resources, Ottawa, for the analyses by emission spectrography. REFERENCES 1. Galliford, D. J. B., and Yardley, J. T., Analyst, 1963, 88, 653. 2. Ratcliffe, R. J. M., and Stevens, S. G. E., Mfg Ckenz. Aerosol News, 1964, 35, (9), 83. 3. Mostyn, R. A., and Cunningham, A. F., Amzlyt. Chem.. 1967, 39, 433. 4. “Specifications for Identity and Purity of Food Additives,” Food and Agricultural Organisation, 5. Fed1 Register, 1966, 31, 1065. Received July 5th, 1968 Rome, 1963, Volume 2, p. 26.