618 Analyst, October, 1967, Vol. 92, $9. 618-621 The Determination of Aluminium with 8-Hydroxy quinoline Part II.* Precipitation in Ainmoniacal Cvanide - EDTA Solution BY A. CLAASSEN, L. BASTINGS AND J. VISSER (Philips Research Laboratories, N . V. Philifis’ Gloeilamfienfabrieken, Eindhoven, The Netherlands) A general procedure is described for the titrimetric determination of 2 to 20 mg of aluminium as the 8-hydroxyquinolinate by precipitation in am- moniacal cyanide - EDTA solution. This method is applicable in the presence of large amounts of a considerable number of elements. Inter- fering elements are beryllium, bismuth, gallium, hafnium, indium, niobium, antimony( 111), scandium, thorium, uranium, vanadium, zirconium and more than 1 mg of fluoride. Interference by chromium and titanium can be prevented by slight changes in the procedure.THE determination of aluminium with 8-hydroxyquinoline in acetate-buff ered solution is very unselective, as it only affords separation from magnesium, beryllium, the alkaline earths and the alkali metals. It can be made more selective by precipitation from ammoniacal cyanide solution, in which many metals are masked as the complex cyanide. The principle of this method is by Lang and Reifer,l and, in particular, by Heczko.2 It has since been applied, in various modifications, for the determination of aluminium in iron and stee1,293Jg6s6 copper al10ys~~~ and zinc al10ys.~ In 1934, we published a photometric methodlo for the determination of aluminium by chloroform extraction of the 8-hydroxyquinoline compound, which was based on a similar procedure ; the selectivity, however, was further enhanced by the addition of EDTA (ethylene- diaminetetra-acetic acid).This principle has since been applied to the macro determination of alumini~m.llsl~s~~ During the course of years an extensive investigation of this method has intermittently been made and has led to the following procedure being recommended for the macro deter- mination of aluminium in complex mixtures. METHOD REAGENTS- 8-Hydroxyqainoline, 2.5 $er cenzt.-Add 25 g of 8-hydroxyquinoline to 29 ml of 6 M hydrochloric acid, dilute with water, filter if necessary, and dilute to 1 litre. 8-Hydroxyqainoline wash solation-Dilute 8 ml of 2.5 per cent. 8-hydroxyquinoline solution to 500 ml, add 3 drops of 0.1 per cent. bromocresol purple (1 g per litre in 20 per cent.ethanol), neutralise with 2 M ammonia solution until the solution just begins to turn purple (pH about 6) and then dilute to 1 litre. Sodium sulphite, anhydrous. Potassium cyanide. Disodium ethylenediaminetetra-acetic acid (Na,EDTA) . Citric acid solution, 20 per cent. w/v. * For details of Part I of this series, see reference list, p. 621.CLAASSEN, BASTINGS AND VISSER 619 RECOMMENDED PROCEDURE- Dilute the aluminium solution containing between 2 and 20 mg of aluminium to about 100 ml, add 5 ml of citric acid solution (Note 1) and then 7 M ammonia solution until present in slight excess. Add 3 g of potassium cyanide and 1 g of anhydrous sodium sulphite (Note 2). Stir until all of the salts have dissolved and dilute to between 150 and 250 ml (Note 2).Heat slowly to a temperature of between 80" and 90" C and maintain the solution at this tem- perature for 2 minutes (Note 3). Then add 1 g of EDTA disodium salt (Note 4) and digest at 80" to 90" C for about 2 minutes. Allow to cool to about 70" C and add, while stirring vigor- ously, the appropriate amount of 2-5 per cent. 8-hydroxyquinoline reagent, add 0.70 ml for each milligram of aluminium, plus an excess of 20 ml. Heat to between 80" and 90" C and maintain the solution at that temperature for about 30 minutes. Allow to cool to about 50" C and filter through a medium filter-paper, retaining as much of the precipitate as possible in the beaker (Note 5). Then wash the precipitate with small portions of warm (50" to 60" C) wash solution, decanting as much as possible; do not use more than about 100 ml.Finally, wash twice with 5 to 10 ml of cold water. Dissolve the precipitate in a hot mixture of 30 ml of concentrated hydrochloric acid (12 M) and 50 ml of water, and titrate with potassium bromate, as described elsewhere (Note 6).14 NOTES- The total amount of citric acid should be 7 to 8 times the weight of tervalent elements, such as iron, rare earths, etc. 2. The amounts of potassium cyanide and sodium sulphite given are the minimum that should always be added. These amounts are adequate for about 0.1 g of elements forming complex cyanides, such as copper, nickel, iron, etc. If these elements are present in larger amounts, the required amounts of cyanide and sulphite should be chosen from Table I, and the volume diluted to that given in the last column.1. Add more citric acid if a precipitate appears when making the solution ammoniacal. TABLE I CONDITIONS FOR PRECIPITATION OF ALUMINIUM Amount of element to be Amount of potassium Amount of sodium complexed by cyanide, cyanide required, sulphite required, Volume after dilution, g g g ml < 0.1 3 1 150 0.1 to 0.25 3 2.5 200 0-25 to 0.5 5 5 200 0.5 to 1.0 10 10 250 If perchlorate is present, potassium cyanide should be replaced by the equivalent amount of sodium cyanide. 3. Do not heat the solution on a hot-plate or flame that is too hot, as local overheating can lead to decomposition of cyanide complexes or formation of organic decomposition products that are carried down with the aluminium 8-hydroxyquinolinate and give rise to increased consumption of bromate.4. The amount of EDTA disodium salt should be a t least 10 times the weight of the elements, cadmium, manganese, lead, zinc and rare earths $us 20 times the weight of alkaline earths and magnesium. 5. The use of quartz beakers is strongly recommended, as glass is attacked by the hot, strongly alkaline solution. Depending on the quality of the glass and the time of heating a t 80" to 90" C, 20 to 80 p g of aluminium can dissolve. 6. A gravimetric finish for this procedure is not advisable, as filter crucibles are heavily attacked by the strongly alkaline solution. A minimum amount of 1 g should always be added. DISCUSSION The reduction of cyanoferrate(II1) can be effected with ~ u l p h i d e ~ , ~ , ~ ~ , ~ ~ or sulphite1~7~*s9 Sulphite proved to be the reducing agent that With the recommended procedure a pH of between 8.5 and 10 is obtained.The aluminium or simply by boiling the alkaline s ~ l u t i o n . ~ ~ ~ ~ ~ gave the best results for cleanness of separation from iron. that escapes precipitation amounts to between 20 and 30pg. NON-INTERFERING ELEMENTS- The determination of 2 to 20mg of aluminium is possible, without interference, in the presence of 0.5 to 1 g of the following elements: silver, arsenic(III), arsenic(V), gold, cadmium,620 CLAASSEN, BASTINGS AND VISSER : DETERMINATION OF [Arta&d, VOl. 92 cerium(III), cerium(IV), cobalt, copper, iron(II), iron(III), germanium, mercury(I), mer- cury( 11) , lanthanum and other rare earths, magnesium, manganese, molybdenum(V1) , nickel, lead, palladium, platinum, antimony(V), selenium(IV), selenium(VI), tin(IV), tellurium(IV), tellurium(VI), thallium(I), thallium(III), tungsten(VI), zinc and the alkaline earths.In these tests, the amount of aluminium found was generally within 20.02 mg of the expected value. Yttrium does not interfere in amounts up to about 50mg; with larger amounts, losses of a few tenths of a milligram of aluminium occur. Chlorides, sulphates, nitrates or perchlorates do not interfere. INTERFERING ELEMENTS- Beryllium, bismuth, gallium, hafnium, indium, niobium, antimony(III), tantalum, thorium, uranium and zirconium should be absent, as they are all precipitated more or less completely as the 8-hydroxyquinolinate.In the presence of scandium the recovery of aluminium is only 80 to 90 per cent. complete. Chromium(V1) precipitates incompletely as an 8-hydroxyquinoline compound, colouring the aluminium 8-hydroxyquinolinate a vivid orange. Up to 20 mg of chromium(V1) scarcely interfere. Interference by chromium(II1) is much stronger, 2 and 10mg of chromium(II1) giving positive errors of about 0.05 and 0.3mg of aluminium, respectively. If the amount of chromium is below 20 mg, interference can therefore be prevented by oxidation to chromium(V1). Larger amounts of chromium can be removed by heating to fumes with perchloric and hydrochloric acids. Interference by chromium can also be prevented by converting it to its EDTA complex by boiling for at least 5 minutes with the necessary amount of EDTA (see Note 4); iron must first be reduced by boiling with sulphur dioxide.Fluoride up to about 1 mg does not interfere; larger amounts give results that are too low, even if a large excess of boric acid is added. It can, however, interfere, if more than about 100 mg of iron are present, by preventing the complete reduction of iron. This results in slight contamination of the precipitate by iron (green colour). Thus 100 mg of iron, with phosphate equivalent to 5 ml of concentrated orthophosphoric acid, showed no interference ; 500 mg of iron with the same amount of orthophosphoric acid, however, caused a positive error of 0.2 mg of aluminium. The positive error caused by 1 mg of vanadium amounts to 0.05 mg of aluminium, and as the vanadium rises to 10 mg, to between 0.2 and 0.3 mg of aluminium.Titanium precipitates completely as the 8-hydroxyquinolinate at a pH of about 9 and below; at a pH above 9, its precipitation is incomplete. Contamination by titanium is shown by the orange discoloration of the precipitate. Determination of aluminium in materials containing amounts of titanium that are not too large can be made by dissolving the contaminated precipitate in hydrochloric acid and removing the titanium by extraction with cupferron - chloroform. Aluminium can be determined in the aqueous phase, as described e1~ewhere.l~ In Table I1 the results are given of the application of the recommended procedure to some standard samples. Orthophosphate does not interfere in principle. Vanadium precipitates incompletely. The interference by vanadium increases below pH 9.TABLE I1 DETERMINATION OF ALUMINIUM IN STANDARD SAMPLES Sample weight, Certified value Sample g and range N.B.S. 93. Borosilicate glass . . .. 1.0 1-03 (1.01 to 1.03) N.B.S. 91. Opal glass . . . . . . 0.5 3.18 (3.16 to 3.31) N.B.S. 94a. Zinc base alloy . . .. 0-25 3.90 (3.88 to 3.95) N.B.S. 62b. Manganese bronze . . . . 1.0 0.97 (0.95 to 0.98) N.B.S. 171. Magnesium base alloy . . 0.5 2.98 (2.96 to 3.00) N.B.S. 106a. Cr - Mo - A1 steel . . . . 1.0 1-08 (1.07 t o 1.12) B.C.S. 179. Manganese brass B . . 0.7, 1.0 1-62 (1.54 to 1.77) B.C.S. 233. Permanent magnet alloy . . 0.15 6.98 (6.93 to 7.08) B.C.S. 312. Permanent magnet alloy . . 0.12 7.87 (7.77 to 7-93) (Ti, 0.79 per cent.) (Ti, 1.21 per cent.; Ta + Nb, 1.35 per cent.; Al, 7.88 per cent.) Aluminium found, per cent.1.02, 1-02 3.19, 3.19 3.91, 3.91 0.97, 0.98, 0.99 2.96, 2.97 1.09. 1.10, 1.10 1.61, 1-81, 1-62 6.95, 6.95 6.96, 6-97 7.85, 7.86October, 19671 ALUMINIUM WITH 8-HYDROXYQUINOLINE. PART I1 REFERENCES 62 1 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. Lang, R., and Reifer, J., 2. analyt. Chem., 1933, 93, 161. Heczko, T., Chemikerzeitung, 1934, 58, 1032. Box, F. W., Analyst, 1946, 71, 317. Klinger, P., Arch. EisenhiittWes., 1939-40, 13, 21. Pigott, E. C., J . Soc. Chem. Ind., Lond., 1939, 58, 139. Steele, S. D., and Russell, L., Iron Steel, Lond., 1942, 16, 182. Edwards, W. T., Analyst, 1948, 73, 556. Weidmann, H., 2. Metallk., 1953, 44, 565. Reutel, C., Metall Erz, 1941, 38, 170. Claassen, A., Bastings, L., and Visser, J., Analytica Chim. Acta, 1954, 10, 373. Detmar, D. A., and van Aller, H. C., R e d . Trav. Chim. Pays-Bas, 1956, 75, 1429. Hoekstra, E., and van Dorp, F. C., Chem. WeeRbl., 1955, 51, 895. Mohr, E., Chem. Tech., Bed., 1959, 11, 598. Claassen, A., and Bastings, L., Analyst, 1967, 92, 614. “1964 Book of A.S.T.M. Standards, Part 32, Chemical Analysis of Metals,” American Society for Methods of Analysis Committee, J . Iron Steel Inst., 1954, 176, 263. NOTE-Reference 14 is to Part I of this series. Testing Materials, Philadelphia, p. 258. Received November 18th, 1966