614 Analyst, October, 1967, Vol. 92, $9. 614-617 The Determination of Aluminium with 8-Hydroxy quinoline Part I. Precipitation in Acetate-buffered Solution BY A. CLAASSEN AND L. BASTINGS (Philips Research Laboratories, N . V . Philips' Gloeilampenfubrieken, Eindhoven, The Netherlands) The conditions for an accurate gravimetric or titrimetric determination of aluminium with 8-hydroxyquinoline in acetate-buffered solution have been established. Precipitation is carried out at a pH higher than 5-2 with an excess of 200 to 300 mg of 8-hydroxyquinoline above the stoicheiometric amount in a final volume of 150 to 200 ml, and the precipitate washed with a neutral 0.02 per cent. solution of 8-hydroxyquinoline. In pure solutions a standard deviation of about 0.01 mg of aluminium has been obtained.ALTHOUGH precipitation of aluminium with 8-hydroxyquinoline has been described in many papers, the conditions given in many textbooks for the determination of aluminium by this method differ. In particular, the instructions as to the amount of reagent required for complete precipitation vary between a slight ~ x c ~ s s , ~ ~ ~ , ~ 15 to 25 per cent. excess415 and an excess of 10 ml of 2.5 per cent. reagent solution.6 As preliminary experiments had shown that the use of a 25 per cent. excess of reagent gave rise to slightly low results it was decided to investigate this method more closely. The procedure chosen was that in which a solution of 8-hydroxyquinoline in dilute acid was added to a slightly acidic aluminium solution, and precipitation effected by the slow addition of ammonium acetate.This method was selected as it furnished a rather coarse, crystalline precipitate, in contrast to that given by the procedure in which the order of addition of reagent and ammonium acetate was reversed, an effect also observed by Stumpf.7 Completeness of precipitation under various conditions of pH and excess of reagent was examined either by titration of the precipitate with bromate or by determination of the amount of unprecipitated aluminium. The latter was carried out by repeated extraction of the filtrate with chloroform and photometric determination of the aluminium as 8-hydroxy- quinolinate, as previously described.8 With a 25 per cent. excess of reagent results were low by 0.15 to 0-20mg in the range from 15 to 30 mg of aluminium, and 1 mg too low with 5 mg of aluminium in a final volume of 150 to 200ml.Complete precipitation was obtained with an excess of S to 15ml of 2.5 per cent. reagent solution; the filtrate then contained 20 to 25 pg of aluminium for amounts of aluminium between 5 and 30 mg. In the final procedure an excess of 10 ml has therefore been prescribed ; much larger amounts lead to crystallisation of 8-hydroxyquinoline during filtration. Miller and Chalmersg also prescribe an excess of 10 ml of 2.5 per cent. reagent for complete precipitation. We cannot, however, confirm their observation that errors of 0.2 per cent. occur when the excess is either 8 or 12 ml. The solubility of the precipitate as a function of pH is given in Table I. TABLE I UNPRECIPITATED ALUMINIUM AS A FUNCTION OF pH IN 150ml OF SOLUTION AT ABOUT 50" c, AFTER PRECIPITATION WITH A 10 ml EXCESS OF 2.5 PER CENT.S-HYDROXYQUINOLINE I N ACETATE-BUFFERED SOLUTION pH . . . . 3-8 4.0 4.3 4.5 4.7 4.8 4.9 5.0 5.3 to 7.0 Aluminium, pg .. >lo00 75 50 41 33 30 28 26 22 to 20 (The results given are the mean of closely agreeing values.) Precipitation can be considered complete at pH values of 5.2 and above. The amount of aluminium escaping precipitation in the pH range of 5.2 to 7 remained at the level of about 20 pg for precipitation at temperatures between 20" and 70" C.CLAASSEN AND BASTINGS 615 Tests performed under the above conditions for complete precipitation were still too low, when the titration with bromate was used, by amounts of 0.05 to 0.08 mg of aluminium, which were much greater than the amounts of unprecipitated aluminium in the filtrate.The cause of these low results was suspected to be the washing procedure of the precipitate. In our experiments the precipitate had been washed with 100 to 150ml of water at about 70" C. Washing with hot water,6 warm water4 and cold water2 have been recommended. We therefore determined the loss of aluminium when washing a precipitate of aluminium 8-hydroxyquinolinate with water at various temperatures. The results are shown in Table 11. TABLE I1 ALUMINIUM IN 100 ml OF FILTRATE WHEN ABOUT 200 mg OF ALUMINIUM 8-HYDROXYQUINOLINATE ARE WASHED WITH WATER AT VARIOUS TEMPERATURES Temperature, "C . . . . . 100 80 60 25 Aluminium, pg . . . . 27 15 7 6 The loss of aluminium is slight and does not account for the heavier loss mentioned above.As it was thought probable that hydrolysis occurred during washing, the amount of 8-hydroxy- quinoline in the wash liquid was also determined photometrically; the results are given in Table 111, together with the equivalent amount of aluminium. TABLE I11 8-HYDROXYQUINOLINE I N 100 ml OF FILTRATE WHEN ABOUT 200 mg OF ALUMINIUM 8-HYDROXYQUINOLINATE ARE WASHED WITH WATER AT VARIOUS TEMPERATURES Temperature, "C . . . . . . 100 80 60 25 8-Hydroxyquinoline, pg . . 1300 700 375 175 Equivalent aluminium, pg . . 81 44 24 11 Comparing Tables I1 and I11 we see that the aluminium 8-hydroxyquinolinate hydrolyses, leaving most of the aluminium as hydroxide on the filter. As the titrimetric determination is based on the determination of the 8-hydroxyquinoline content of the precipitate, it is clear from Table I11 that washing with hot or even warm water leads to considerable loss.This does not occur when washing with a small volume of cold water. We prefer to wash with a neutral 0.02 per cent. solution of 8-hydroxyquinoline. With this solution the amount of aluminium found in 100ml of filtrate is about 5 p g , irrespective of the temperature in the range 20" to 100" C. That the hydrolysis is completely prevented is shown by the accurate results obtained with the recommended procedure. Solutions of 8-hydroxyquinoline in dilute acetic acid or ethanol have been described for the reagent. The latter solvent is unsuitable as the solubility of the precipitate is increased considerably by the presence of ethanol.Thus, in solutions containing 5 per cent v/v of ethanol the solubility increases to about 50 pg and, in solutions containing 10 per cent. V/V, to about 85 pg of aluminium per 150 ml. Most textbooks prescribe a 2-5 or 5 per cent. solution of 8-hydroxyquinoline in 1 to 2 M acetic acid. One disadvantage of these solutions is that, after evaporation to dryness, the residue that has spilt on the wall of the beaker is difficult to dissolve in the test solution. A second disadvantage is that they contain much free acid and it is therefore difficult to reach a sufficiently high pH value. We therefore prefer to use a 2-5 per cent. solution of 8-hydroxy- quinoline in dilute hydrochloric acid. This solution contains practically no free acid (pH about 3) and the dried salt is readily soluble in water and in the test solution.The following procedure is recommended for the gravimetric or titrimetric determination. REAGENTS- S-HydroxyquinoZine, 2.5 per cent.-Dissolve 25 g of 8-hydroxyquinoline in 29 ml of 6 M hydrochloric acid, dilute with water, filter if necessary, and dilute to 1 litre. Ammonium acetate, 20 per cent. w/v-Dilute this solution with water (1 + 9) when it should have a pH of 6-5 to 7.0. If necessary, adjust the pH with ammonia solution or acetic acid. S-Hydroxyquinoline wash solzdion-Dilute 8 ml of 2.5 per cent. 8-hydroxyquinoline solu- tion with water to 500 ml, add 3 drops of bromocresol purple (1 g per litre in 20 per cent. METHOD616 CLAASSEN AND BASTINGS DETERMINATION OF [Analyst, Vol.92 ethanol) and ammonia solution (2 M) until the solution just begins to turn purple (pH about 6), and then dilute to 1 litre. SAMPLE SOLUTION- and 2 to 30mg of aluminium for the gravimetric procedure. The sample solution may contain 2 to 20 mg of aluminium for the titrimetric procedure RECOMMENDED PROCEDURE- Transfer the sample solution into a 400-ml beaker and dilute to between 100 and 125 ml. Add to the cold solution, dropwise, with stirring, 2 M ammonia solution until a slight pre- cipitate of aluminium hydroxide persists. Then add, dropwise, with stirring, 2 M hydrochloric acid until the solution is clear again, followed by an excess of 5 to 10 drops. If the precipitate of aluminium hydroxide cannot be seen clearly, as with small amounts of aluminium, adjust to a pH between 3 and 4, by using pH paper, and add 5 to 10 drops of 2 M hydrochloric acid.Add 0.70ml of 2.5 per cent. 8-hydroxyquinoline for each milligram of aluminium present $,!us an excess of 10 ml. Heat to about 70" C and add slowly, with stirring, 20 ml of ammonium acetate solution. Heat just to boiling, stand the solution on a steam-bath for 30 minutes, and then allow it to cool to about 50" C. (a) Gravimetric determinatiofi-Filter the solution through a weighed No. 4 sintered- glass or quartz crucible, decanting as much as possible. Transfer and wash the precipitate with small portions of warm (50" to 60" C) wash solution; do not use more than about 100 ml in total. Avoid sucking the precipitate dry during washing, otherwise a compact mass is formed that is difficult to wash free from salts. The gravimetric factor for aluminium is 0.05873 and that for aluminium oxide, 0.1110.(b) Titrimetric determination-Filter the solution and wash, as under (a), but use a medium filter-paper. Dissolve the precipitate in a hot mixture of 30ml of concentrated hydrochloric acid (12 M) and 50 ml of water, taken in small portions. Wash the filter with water. Collect the solution in a 300-ml conical, glass-stoppered flask, having a mark at the volume of 150 ml. Dilute to 150 ml and cool to room temperature. Now titrate the solution slowly with 0.1 or 0-2 M potassium bromate solution containing about 20 g of potassium bromide per litre. Shortly before the end-point, add a few drops of 0-1 per cent.methyl red and continue the titration slowly until the colour becomes quite yellow; then add 2 ml of bromate solution in excess. Stopper the flask, mix and wait 2 to 3 minutes. Then add about 2 g of potassium iodide, dissolve it by swirling, and titrate the liberated iodine with 0.1 N sodium thiosulphate solution, adding starch as indicator near the end-point. Subtract the value of the back-titration with thiosulphate from the potassium bromate added. One millilitre of 0.1000 N potassium bromate is equivalent to 2.248 mg of aluminium or 4.248 mg of aluminium oxide. Proceed according to (a) or (b) below. Finally, wash the precipitate twice with 5ml of cold water. Dry to constant weight at 140" to 150°C. NOTE- With this procedure the pH after precipitation lies between 5-2 and 5.8.TEST OF ACCURACY- To test the recommended procedure a standard aluminium solution containing about 1 mg of aluminium per ml was made by dissolving an accurately known weight of high purity aluminium in hydrochloric acid and diluting this solution to a known weight. Weighed aliquots of this solution were used so that the amount of aluminium taken for analysis was known with an accuracy to within 0.001 mg. The results for the titrimetric procedure are given in Table IV, and for the gravimetric procedure, in Table V. The bromate solution was standardised with high purity arsenic(II1) oxide. TABLE IV RESULTS SHOWING THE ACCURACY OF THE TITRIMETRIC PROCEDURE Approximate amount of Number of Main difference from Standard deviation of aluminium taken, determinations the theoretical value, single determination, mg mg mg 2 9 - 0.01, 0.007 10 12 * 0-00 0416 20 10 + 0.00, 0.009October, 19671 ALUMINIUM WITH 8-HYDROXYQUINOLINE. PART I TABLE V RESULTS SHOWING THE ACCURACY OF THE GRAVIMETRIC PROCEDURE 617 Approximate amount of Number of Mean difference from Standard deviation of aluminium taken, determinations the theoretical value, single determination, mg mg mg 2 6 - 0.013 0.005 10* 17 f 0.000 0.007 20 5 + 0.004 0.012 * The results for 10 mg were obtained by four different operators.It follows from these tables that with both procedures the standard deviation for a single determination is about 0.01 mg of aluminium. Excepting the tests with 2 mg, the mean does not differ significantly from the theoretical value.Strictly speaking, results somewhat too low should have been expected, as the solubility of the precipitate corresponds to between 20 and 25 pg of aluminium. EFFECT OF SOME ANIONS- It is known from the literature that the usual inorganic anions, such as chloride, sulphate, nitrate and perchlorate, do not interfere in this determination. As with Wassiljew,lo it was found that small amounts of fluoride (greater than 1 mg) interfere by giving low results. Addition of 1 to 5 g of boric acid decreases the error but is only effective in masking up to 2 mg of fluoride. As in the presence of tartrate no precipitate of aluminium hydroxide appears when neutralising the solution, the neutralisation has to be carried out by using indicator paper at a pH of up to about 3.An extended study of the accuracy in the presence of tartrate has not been made. Citrate and oxalate interfere, as in their presence precipitation is only complete at a pH above 7 to 8. Tartrate does not interfere but gives a much finer precipitate. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. REFERENCES Kolthoff, I. M., and Sandell, E. B., “Textbook of Quantitative Inorganic Analysis,” Third Edition, Charlot, G., “Les Me‘thodes de la Chimie Analytique. Analyse Quantitative Mintfrale,” Fifth Erdy, L., “Theorie und Praxis dev gravimetrischen Analyse, ” Volume 11, Akaddmiai Kiado, Buda- Vogel, A. I., “A Text-Book of Quantitative Inorganic Analysis including Elementary Instrumental Kolthoff, I. M., Belcher, R., Stenger, V. A., and Matsuyama, G., “Volumetric Analysis,” Volume Wilson, C. L., and Wilson, D. W., Editors, “Comprehensive Analytical Chemistry,” Elsevier Pub- Stumpf, K. E., 2. analyt. Chem., 1953, 138, 30. Claassen, A., Bastings, L., and Visser, J., Analytica Chim. Acta, 1954, 10, 373. Miller, C. C., and Chalmers, R. A., Analyst, 1953, 78, 686. Wassiljew, K. A., Zav. Lab., 1937, 6, 432. Macmillan & Co. Ltd., New York and London, 1952, p. 320. Edition, Masson et Cie, Paris, 1966, p. 584. pest, 1964, pp. 317 and 322. Analysis,” Third Edition, Longmans, Green & Co. Ltd., London, 1961, p. 387. 111, Interscience Publishers Inc., New York and London, 1957, p. 554. lishing Co., Amsterdam, London, New York and Princeton, 1962, Volume I C, p. 106. Received November 18th, 1966