622 Analyst, October, 1967, Vol. 92, pp. 622-626 The Determination of Aluminium in Wool by Atomic-absorption Spectroscopy BY IT;. R. HARTLEY AND A. S. INGLIS victoria, Australia) (Division of Pvotein Chemistry, CSIRO, Wool Research Laboratories, Parkville N.2 (Melbouvne), Atomic-absorption spectroscopy at 3092-7 A in a nitrous oxide - acetylene flame provides a simple and precise method for determining aluminium in wool. The sample is dissolved in constant-boiling hydrochloric acid, and the solution sprayed directly into the flame. The presence of hydrochloric acid partially suppresses the aluminium absorbance, whereas the amino-acids present in the wool hydrolysate enhance the absorbance. A linear calibration graph over the range 0 to 130,xg per ml is obtained from solutions of alu- minium in constant-boiling hydrochloric acid containing the hydrolysed protein.Although the calibration graph has an accuracy of +Z per cent., the heterogeneous nature of wool limits the accuracy of the analysis to f3 per cent. The method, which has been applied successfully down to 0.02 per cent. of aluminium on wool, is equally suitable for other insoluble protein materials, such as hair and hide. IN connection with a study of the properties of wool after treatment with aluminium salts, it became necessary to determine the effect of various conditions on the uptake of aluminium by wool. As only relatively small amounts of aluminium were involved, a method with a high sensitivity was necessary. It was also desirable that the determination should not be affected by the other metal ions present in wool, or by the brown colour of the solutions obtained on dissolving wool.From these considerations it appeared that a procedure involving atomic-absorption spectroscopy with a nitrous oxide - acetylene flame1 would be superior t o methods based on complexometric titrations,2 gravimetric3 or spectrophotometric analy~is.~ In addition, atomic-absorption spectroscopy offered the possibility that the solution of amino-acids obtained on hydrolysis of wool in constant-boiling hydrochloric acid could be used directly, thus eliminating the necessity for filtration, dilution or the addition of further reagents. EXPERIMENTAL APPARATUS- The equipment used in this work was the Techtron AA-3 atomic-absorption spectro- photometer with an aluminium hollow-cathode lamp supplied by Atomic Spectral Lamps.The operating conditions that gave optimum sensitivity were: slit width, 100 p ; wavelength, 3092.7 A; lamp current, 11 mA. These conditions are in agreement with those suggested by Amos and Wi1lis.l A stainless-steel high temperature burner (type AB40), supplied with the Techtron AA-3 spectrophotometer, was used. Aspiration via the Techtron atomiser was by nitrous oxide (pressure, 151b per sq. inch), and the acetylene pressure used was sufficient to produce a red “feather” in the flame ten times the height of the pale blue cone. When more acetylene was present a deposit of carbon formed around the mouth of the burner, which led to an unstable flame and fluctuations in the absorbance reading.When less acetylene was present the aluminium was less completely atomised. The burner was carefully aligned for direction, horizontal position and height to obtain the maximum sensi- tivity. The first two factors were found to be critical, whereas the height could be varied over a range of about 1 cm without altering the absorbance value. The flame was first ignited with only acetylene present. After increasing the acetylene pressure to about 18 lbHARTLEY AND INGLIS 623 per sq. inch, the nitrous oxide was turned on to give a pressure of 15 lb per sq. inch, and the acetylene pressure then reduced to give the desired flame. The whole operation of lighting and adjusting the flame was carried out as rapidly as possible to prevent the formation of a carbon deposit at the mouth of the burner.It was necessary to light the flame at least half an hour before analysing the samples, as variations in readings occurred during this time. REAGENTS- Constant-boiling hydrochloric acid--Prepare by the method described by VogeL5 Aluminium sulfdzate solution*-Prepare a stock solution about 0.1 M by dissolving 31.52 g of aluminium sulphate [A1,(S0,),.16H20] (analytical-reagent grade) in 1 litre of water. Standardise the solution by using an excess of EDTA and back-titrating against a standard solution of zinc chloride, with Eriochrome T as indicator.2 (The standard zinc chloride solution is prepared by dissolving analytical-reagent grade zinc metal in concentrated hydro- chloric acid, analytical-reagent grade, and diluting with water.) Prepare solutions of aluminium sulphate for the calibration graph by diluting the stock solution and adjusting the concentration of hydrochloric acid in each solution to 6.19 N PREPARATION OF CALIBRATION GRAPH- Transfer by pipette suitable volumes of the standardised aluminium sulphate solution to give transmission values in the region of 40 to 95 per cent.Make up to volume with a mixture of the hydrochloric acid and distilled water, into standard flasks, so that the final solution is 6.19 N with respect to hydrochloric acid. Seal 5-ml aliquots of the solutions, together with 0.3 g of a sample of the wool that has not been treated with aluminium salts, in hard-glass tubes and heat in a forced-draught oven at 110" C for 20 hours. After cooling, shake the tube well and cautiously open.Aspirate the solution directly into the flame and record the transmission value. The calibration graph is linear for aluminium concentrations of up to 130 pg per ml (0.0048 M) and should pass through the origin. PREPARATION OF SAMPLES- Weigh accurately samples of wool (about 0-3 g) that contain sufficient aluminium to give a direct reading without dilution. The wool samples should be equilibrated for at least 24 hours in a constant-humidity room before weighing. Seal each sample in a thick-walled tube, together with 5 ml of constant-boiling hydrochloric acid, accurately measured by pipette, and heat in a forced-draught oven at 110" C for 20 hours. After cooling, shake the tube well and open it cautiously. Aspirate the solution directly into the flame and record the trans- mission. A calibration graph must be prepared each time a batch of samples is analysed to eliminate differences that arise in setting up the instrument.After use, the aspirator must be well washed with distilled water to remove any traces of solvent that could cause corrosion if left on the aspirator. RESULTS AND DISCIJSSION EFFECT OF SOLVENT- A series of calibration graphs was prepared by using solutions of aluminium sulphate in water; aluminium sulphate in 6.19 N hydrochloric acid; and aluminium sulphate and hydrolysed protein in 6.19 N hydrochloric acid. Table I shows that all of the systems studied gave linear plots of absorbance against aluminium concentration of up to 130 pg of aluminium per ml (0.0048 M). At higher concentrations the plot curved away from the absorbance axis, indicating that the aluminium was then being less completely atomised. Typical values for the ratio of the absorbance to aluminium concentration, and the sensitivity in the different media, are given in Table 11.The sensitivity in water is a little less than that observed by Amos and Willis,l who detected 1 pg of aluminium per ml at 1 per cent. transmission. This difference is almost certainly caused by minor instrumental differences. It can be seen that the presence of 6.19 N hvdrochloric acid suppresses the aluminium absorbance by 17.6 per cent. In the presence of 0.2 N hydrochloric acid, Amos and Thomas6 found that the aluminium absorbance is suppressed by 11 per cent. ir; an oxygen - nitrogen - acetylene flame, although Amos and Willisl found later that no effect could be detected with 0.5 N hydrochloric acid in a nitrous oxide - acetylene flame.* A standard solution could be prepared with high purity aluminium metal, but this was not readily available in Australia when this work was undertaken.624 [Analyst, Vol. 92 INTERFERENCES- In the presence of the protein hydrolysates studied the aluminium absorbance is enhanced relative to that observed in 6.19 N hydrochloric acid. This enhancement could be caused by evaporation of water vapour when sealing the samples in glass tubes; caused by the trace metals present in the proteins; or by the amino-acids present in the hydrolysates. HARTLEY AND INGLIS : DETERMINATION OF ALUMINIUM Concentration of aluminium, 12-92 25.85 38.77 51.75 77-60 PP Per ml 103.5 129.2 181.1 258.5 TABLE I CALIBRATION POINTS FOR ALUMINIUM IN DIFFERENT MEDIA Medium Water R s t i o o f absorbance to Absorbance concentration 0.0223 0.0458 0.0680 0-0915 0.137 0.182 0.229 0-310 0.426 1.73 1.77 1-75 1-77 1-77 1.76 1-77 1.71 1.67 6.19 N Hydrochloric acid R a t i o o f absorbance to Absorbance concentration 0.0189 0.0373 0.0555 0.0731 0.115 0.152 0.187 0.252 0.347 Medium 1.46 1.44 1.43 1.41 1.48 1.47 1-45 1.39 1.34 6.19 N Hydrochloric akd + scoured wool r Absorbance 0.0200 0.0398 0.0605 0.0809 0.122 0.158 0.201 0.272 0.387 3 Ratio of absorbance to concentration 1-55 1-54 1.56 1.56 1.57 1-53 1.55 1.50 1.49 Concentration of aluminium, 12.92 25.85 38.77 51-75 77-60 p g Per ml 103.5 129.2 181-1 258.5 6.19 N Hydrochloric acid + bleached wool R a t i o o f absorbance to Absorbance concentration 0~0200 1.55 0.0410 1-58 0.0605 1-56 0.0796 1-54 0.119 1.53 0.158 1.53 0.201 1.55 0.272 1-50 0.382 1-47 6-19 N Hydrochloric acid + bovine hide -Ratiof absorbance to Absorbance concentration 0*0200 1-55 0.0410 1.58 0.0595 1.53 0.0809 1-56 0.122 1.57 0.161 1-56 0.201 1.53 0-276 1.52 0.382 1-47 6.19 N Hydrochloric acid + bovine hair - - - - - i a f absorbance to Absorbance concentration 0.0223 0.0458 0.0680 0-0915 0-135 0.181 0.222 0.305 0.420 1.72 1-77 1.75 1-77 1.74 1-75 1.72 1.68 1.62 TABLE I1 DETERMINATION OF ALUMINIUM BY ATOMIC-ABSORPTION SPECTROSCOPY Ratio of Lower limit of absorbance to Sensitivity,? analysis, $ Medium aluminium concentration* pg of aluminium per ml pg of aluminium per ml Water .. . . . . 1.76 f 0.03 2.50 15-3 6-19 N Hydrochloric acid 1-45 & 0.03 3.04 18-5 + scoured wool . . 1.55 f 0.02 2-84 17.3 + bleached wool . . 1.55 f 0.03 2.84 17-3 + bovine hide . . .. 1.55 & 0.03 2.84 17-3 6.19 N Hydrochloric acid 6.19 N Hydrochloric acid 6.19 N Hydrochloric acid 6.19 N Hydrochloric acid + bovine hair . . . . 1-75 f 0.03 2.52 15.4 * Mean ratio of absorbance to concentration in the aluminium concentration range -f Sensitivity necessary to give 1 per cent. transmission, $ Concentration of aluminium in solution necessary to give 94 per cent. transmission, 0 to 130 pg per ml. which is the highest transmission consistent with an accuracy of f 3 per cent.October, 19671 I N WOOL BY ATOMIC-ABSORPTION SPECTROSCOPY 625 The first possibility was eliminated by preparing a calibration graph from solutions of aluminium in 6.19 N hydrochloric acid that were sealed in glass tubes, heated in a forced- draught oven at 110" C for 20 hours, cooled, cautiously opened and aspirated into the flame.The points on this graph were indistinguishable from points obtained by using solutions that had not been sealed. The exact quantitative concentration of each of the trace metals, calcium, magnesium, iron, sodium and potassium, present in wool,' varies from sample to sample, but as the total ash obtained from wool is about 0.4 per cent.,* the amount of each metal present must not be greater than about 0.1 per cent. When each of these metals was added to standard aluminium solutions in concentrations equivalent to 0.1 per cent.and the absorbance measured, only potassium, which enhanced the absorbance by about 4-5 per cent., had any significant effect. Accordingly the proteins were analysed for potassium by using the atomic- absorption spectrometer with an air-coal gas flame and a technique analogous to that described for aluminium. The results indicated that the amount of potassium present in wool and bovine hide is insignificant (about 0.0015 per cent.). Bovine hair was found to contain a larger amount of potassium (0.019 per cent.), which will account for part of the difference in the enhancement of the aluminium absorbance observed with this protein. The ability of potassium to enhance the absorbance of aluminium corresponds to the enhancement by iron observed by Amos and Wi1lis.l The enhancement of the aluminium absorbance by amino-acids was determined by preparing a mixture of amino-acids in about the same proportions as those present in wool (see Table 111).When the mixture of amino-acids (0.3 g) was added to 5 ml of a solution of aluminium in 6.19 N hydrochloric acid (129.3 pg of aluminium per ml), the absorbance was enhanced from 0.187 to 0.204, which was equal, within experimental error, to the en- hancement by wool. These results suggest that the enhancement of the aluminium absorbance by wool and bovine hide is caused by the amino-acids alone. As the amino-acid contents of bovine hideg and woollo are different, it is interesting that both give similar enhancements of the aluminium absorbance. With bovine hair, the potassium present will also contribute to the enhancement observed.TABLE I11 MIXTURE OF AMINO-ACIDS USED TO REPRESENT WOOL HYDROLYSATE Present in 0.3g of wool, Amino-acid mg* Lysine .. . . . . 9 Arginine . . .. .. 20 Aspartic acid . . . . .. 20 Threonine . . . . .. 18.5 Serine . . . . . . . . 29 Glutamic acid . . . . 34 Proline . . .. . . 18 Glycine . . . . . . 27 Alanine . . . . .. 17 Valine . . . . .. . . 18 Methionine . . . . . . 2 Isoleucine . . . . .. 10 Leucine .. . . .. 24 Histidine . . . . . . 2.5 Half-cystine . . .. . . 34 Tyrosine . . .. .. 12 Phenylalanine . . .. 9 * Data from reference 10. Present in mixture, mi? 9 2 17 22 17-5 28 37 21 27 16 34 17 2 9 24 12 8 As the amount of trace metals present and the amino-acid composition will vary from sample to sample, it is necessary to prepare the calibration graph from standard aluminium solutions that contain the hydrolysed protein under investigation in 6.19 N hydrochloric acid. In this way variations caused by the amino-acid composition and metallic contamination of the protein will be eliminated.There is a 3.4 per cent. increase in volume when samples containing 0.3 g are dissolved in 5 ml of constant-boiling hydrochloric acid. However, this does not affect the analysis as the same volume change occurs in the samples used to prepare the calibration graph.626 HARTLEY AND INGLIS REPROD u CI BILITY- The reproducibility of the method was determined by analysing six samples of scoured and bleached wool that had been treated with aluminium sulphate solutions.The results shown in Table IV indicate that the reproducibility is &3 per cent. It was possible to dissolve up to 0.4 g of wool in 5 ml of constant-boiling hydrochloric acid. Thus, from the lower limit of analysis shown in Table 11, the method can be used down to 0.2 mg of aluminium per g of dry wool (0.02 per cent. w/w), with a reproducibility of f3 per cent. As a reproducibility of +2 per cent. is obtained from the calibration graph, presumably the uptake of aluminium is not completely uniform throughout the constituent proteins of wool, and small sampling errors occur, The large increase in uptake of aluminium by bleached wool is not unexpected and may result from the presence of sulphonic acid groups produced by oxidation of disulphide groups during bleaching with hydrogen peroxide.TABLE Iv REPRODUCIBILITY OF ALUMINIUM DETERMINED IN WOOL SAMPLES Scoured wool Bleached wool v L r \ Dry weight, Dry wool, Dry weight, Dry wool, 0.32790 0.0339 0.364 0.24390 0.155 2.05 0.32360 0.0339 0.369 0.2 7 3 98 0-174 2-06 0-29412 0.0327 0.359 0.22769 0-177 2.06 0.28429 0.0315 0.357 0.30368 0.184 1-96 0-31364 0.0339 0.350 0.26847 0.164 1-97 0.37448 0.0410 0.352 0.31177 0.194 2.01 ---- g Absorbance* mg of aluminium per g g Absorbance* mg of aluminium per g Mean .. . . 0.358 -& 0.011 Mean .. . . 2.02 3 0.06 * Sample hydrolysed in 5 ml of constant-boiling acid. SCOPE OF THE METHOD- Although only wool has been studied in detail, the results in Table I indicate that the method should be equally suitable for other insoluble protein materials, such as hair and hide.The great advantage of being able to use the hydrolysate directly will still apply. It is anticipated that such biological materials treated with different metals could also be analysed with advantage by a similar procedure. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Amos, M. D., and Willis, J. B., Spectrochim. Acta, 1966, 22, 1325. Schwarzenbach, G., “Complexometric Titrations,” Methuen & Co. Ltd., London, 1957, p. 88. Trotman, S. R., and Trotman, E. R., “Textile Analysis,” Second Edition, Charles Griffin and Co. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Third Edition, Interscience Vogel, A. I., “A Textbook of Quantitative Inorganic Analysis including Elementary Instrumental Amos, M. D., and Thomas, P. E., Anzalytica Chirn. Acta, 1965, 32, 139. Dusenbury, J. H., in von Bergen, W., Editov, “Wool Handbook,” Third Edition, Interscience Ward, W. H., “Textile Research Institute Summary Report for October 1948 to October 1952,” Veis, A., ip2 Hall, D. A., Editor, “International Review of Connective Tissue Research,” Volume 3, O’Donnell, I. J., and Thompson, E. 0. P., Aust. J . Biol. Sci., 1962, 15, 751. Ltd., London, 1948, p. 255. Publishers Inc., New York and London, 1959, p. 219. Analysis,” Third Edition, Longmans, Green and Co. Ltd., London, 1961, p. 233. Publishers Inc., New York and London, 1963, p. 219. Textile Research Institute, Princeton, New Jersey, 1953, p. 35, Table XI. Academic Press, London, 1965, p. 115. Received Afiril 17th, 1967