Analyst, February, 1967, Vol. 92, $9. 103-106 103 Spectrsphotometric Determination of Aluminium in Soil Extracts with Xylenol Orange BY D. T. PRITCHARD (Soil Survey of England & Wales, Rothanzsted Experimental Station, Havpenden, Herts.) A spectrophotometric method for determining aluminium in soil extracts with xylenol orange a t pH 3.8 is described. Various extraction reagents can be used. Interference of iron(II1) and iron(I1) is eliminated with EDTA. Rectilinear calibration covers a range of 0 to 60pg of aluminium. Typical results are shown for an iron - humus podzol. Recovery of added aluminium and levels of interference of anions and cations relevant to soil analysis are shown. The practicability of the method in contrast with “lake” methods is emphasised. MOST conventional spectrophotometric methods for the determination of aluminium depend on the adsorption of coloured dyes on colloidal aluminium hydroxide.Sandelll commented on the unsatisfactory nature of these “lake”-forming methods (exemplified by aluminon) and deplored the necessity for their use. Korbl and Pi-ibi12 prepared xylenol orange as a metallo- chromic indicator that forms a soluble coloured complex with aluminium. Otomo3 examined this reaction in detail, and Budeshinsky4 used the reagent to determine aluminium in uranium. Although xylenol orange is not specific for aluminium, only iron is likely to cause serious interference in soil extracts, and this is eliminated in the proposed method by the use of EDTA. METHOD REAGENTS- All reagents, except xylenol orange, were of AnalaR quality.Tamm’s reagent-Dissolve 24.9 g of hydrated ammonium oxalate and 12-6 g of hydrated Bupeer solation, pH 3.8-Dissolve 136 g of hydrated sodium acetate in water, adjust to oxalic acid in water and dilute to 1 litre. pH 3.8 with hydrochloric acid and dilute to 1 litre. APPARATUS- measurements. A Rausch and Lomb colorimeter (Spectronic 20) was used for spectrophotometric PROCEDURE- Preparation of soil extracts-Shake overnight 2-00 g of soil (crushed to pass a 2-mm sieve) with 100.0ml of Tamm’s reagent and spin it in a centrifuge. Take a 25-ml aliquot and destroy organic matter with 40-volume hydrogen peroxide. Add 10 ml of 9 N sulphuric acid and evaporate to fuming. Cool the solution, dilute it, adjust to about pH 2 and dilute t o 100ml.S$ectrophotornetry-Transfer an aliquot of the soil extract containing 0 to 60 pg of a h - minium into a 100-ml calibrated flask. Add 25 ml of buffer solution (pH 3-8) and 10.0 ml of 0.15 per cent. aqueous xylenol orange. Place the flask in a water-bath at 40” C for 18 hours. Cool the solution, add 5 ml of 0.05 M EDTA (disodium salt), dilute to 100 ml and allow it to stand at room temperature for 1 hour. Measure the optical density at 550mp. RESULTS AND DISCUSSION In the absence of EDTA, the colour produced by the aluminium - xylenol orange complex is stable for several days. The colour develops slowly at room temperature and the rate of development is further retarded by the presence of iron. The temperature chosen (40” C) ensures that a stable, coloured product is obtained in 18 hours.Higher temperatures cause a more intense colouration, but reduce its stability, even in the absence of EDTA.104 PRITCHARD : SPECTROPHOTOMETRIC DETERMINATION OF [A ?Zdy.St, VOl. 92 I I 60 I20 Time. minutes Fig. 1. Rates of colour fading clue to EDTA: curve A, 50 pg of aluminiumplus 1000 pgof iron(II1) plus xylenol orange; curve B, 50 pg of aluminium plus xylenol orange Both iron(II1) and iron(I1) react with xylenol orange to give coloured solutions that react similarly in all respects, and absorb strongly in the region of maximum absorption of the aluminium - xylenol orange complex. The method depends on the difference caused by EDTA in the rates of fading of the colours produced by iron and aluminium with xylenol orange (Fig. 1).It can be seen that the colour produced by 1000 pg of iron fades completely within 1 hour, whereas that produced by the aluminium fades slowly enough to allow repro- ducible readings to be made. Theoretical considerations suggested that ascorbic acid or mercury(I1) - EDTA might mask the iron - xylenol orange reactions without affecting the aluminium - xylenol orange colour. However, under various conditions, neither reagent was wholly satisfactory. Fig. 2 shows the absorption spectra of a solution containing aluminium and xylenol orange, and of xylenol orange alone, which indicate a wavelength of maximum absorption of 550 mp for the aluminium - xylenol orange complex. Wavelength, mp Fig. 2. Absorption spectra: curve A, 100 p g of aluminium #Zus 2 ml of 0.15 per cent.xylenol orange; curve B, xylenol orange aloneFebruary, 19671 ALUMINIUM IN SOIL EXTRACTS WITH XYLENOL ORANGE 105 Fig. 3 illustrates the variation of optical density at 550 mp of the aluminium - xylenol orange complex, and of xylenol orange alone, as a function of pH. The graph is horizontal at pH 3.2 and 3.8, and both points are therefore suitable for measuring the optical density. At pH 3.8, there is some loss of sensitivity, but this pH value was chosen for use because Beer's law is obeyed over a wider range of aluminium concentration than it is at pH 3.2. A rectilinear calibration is obtained over the range 0 to 60 pg of aluminium. Molar absorp- tivity is 11,000 under the conditions described. It is not necessary to include standards with each batch of samples, nor to re-calibrate when using fresh solutions of xylenol orange.2.0 3.0 4.0 PH Fig. 3. Variation of optical density as a func- tion of pH: curve A, 50 pg of aluminium plus 10 ml of 0.15 per cent. xylenol orange; curve B, xylenol orange alone Organic anions such as citrate, oxalate and acetate are common constituents of extraction reagents for soils, and are readily destroyed in the normal course of treatment to oxidise soil organic matter. Hydrofluoric acid, which is used as a soil extractant, particularly for aluminium, can be easily removed. Pyrophosphate extracts can also be analysed after hydrolysis to orthophosphate. TABLE I IRON AND ALUMINIUM VALUES IN ALIQUOTS OF ORIGINAL SAMPLE, RECOVERY OF ADDED ALUMINIUM AND RECOVERY OF ALUMINIUM AFTER ADDING IRON Original aliquot Experiment I Experiment I1 v Iron Aluminium r-- e- Clg pg mg per cent.* Pg t-G Iron Aluminium found, Aluminium Aluminium present, & added, found, added, found, 52 7.5 15 20 26-5 800 8.0 12 5.4 11 20 24.5 so0 5.7 24 7.5 16 20 26.5 800 8.0 40 26.0 52 20 46.5 800 26.0 268 34.0 136 20 54-7 so0 34.0 180 26.2 105 20 46.7 800 26.2 * Concentration of aluminium expressed as a percentage of the soil. A typical extraction procedure with Tamm's reagent was carried out with samples taken from six horizons of an iron - humus podzol profile. Separate aliquots were taken for the determination of iron with thioglycollic acid, and for the determination of aluminium by the proposed method. The results obtained are shown in Table I, which also shows the recovery of added aluminium (Experiment I), and the aluminium found after adding 800pg of iron106 PRITCHARD (Experiment 11). Table I1 shows the order of interference of anions and cations relevant to soil analysis.The results shown in Tables I and I1 indicate that the method can be successfully applied to the determination of aluminium in the amounts usually occurring in soil extracts. TABLE I1 LEVELS OF INTERFERENCE OF ANIONS AND CATIONS RELEVANT TO SOIL ANALYSIS Aluminium present, 50 pg Cation or anion added, pg Aluminium found, pg Iron (Fe) 1000 .. .. 50.0 Titanium (Ti) 50 . . .. 55.0 Calcium (Ca) 5000 . . Magnesium (Mg) 5000 . . :} 50.3 Silicate (SiO,) 500 . . . . 50.3 Citrate 1000 .. . . 47.0 Oxalate 1000 .. .. 11.7 Pyrophosphate (P) 1000 . . .. 15.8 Manganese (Mn) 1000 . . . . Fluoride (F) 50 . . . . 47.3 Phosphate (P) 1000 .. .. 49.7 Phosphate (P) 10,000 . . .. 49.3 REFERENCES 1. 2. 3. 4. Sandell, E. B., “Colorimetric Determination of Traces of Metals,” Third Edition, Interscience Korbl, J., and Pfibil, R., Chemist-Analyst, 1956, 45, 162. Otomo, M., Bull. Chem. SOC. Japan, 1963, 36, 809. Budeshinsky, B., Zh. Analit. Khim., 1963, 18, 1071. Publishers Ltd., London, 1959, p. 227. Received July 7th, 1966