so BOOTH AND EVETT : TH:E ABSORPTIOMETRIC [Vol. 83 The Absorptiometric Determination of Traces of Iron in E!ismuth BY E. BOOTH AND T. W. E:VETT (U.K.A.E.A ., Research Group, WooZwich Outstation, Woolwich, S.E. 18) The development of an accurate method for the determination of traces of iron in bismuth is described. Of the many reagents considered, 4 : 7-di- phenyl-1 : 10-phenanthroline was found to be the most suitable, both in specificity and sensitivity. In a hydrochloric acid solution of the sample, iron111 is reduced to iron11 by stannous chloride. The 4 : 7-diphenyl-l : 10-phenanthroline is then added, followed by a mixture of disodium ethyl enediaminetetra-acetate and sodium citrate. This serves the dual purpose of adjusting the pH and complexing the bismuth, hence preventing its precipitation.The iron complex is extracted with n-hexyl alcohol and the optical density of the solution is measured at 633mp. AN analytical method was required that woulcl permit the iron content of bismuth to be determined accurately in the range 2 to 100 p.p.m., the range 2 to 10 p.p.m. being particularly important. Of the many reagents available,l potassium thiocyanate appeared to be an obvious choice and was found to be satisfactory at levels a.bove about 10 pg, but, below this, interference from bismuth and the instability of the coloured complex caused us to abandon its use. Smith, McCurdy and DiehP describe a new reagent, 4: 7-diphenyl-l : 10-phenanthroline (bathophenanthroline) for the determination of very small amounts of iron. This reagent is almost specific for iron11 and is twice as sensitive as 1 : 10-phenanthroline, and the ferrous complex can easily be extracted with isoamyl alcohol or rt-hexyl alcohol.The general behaviour of bathophenanthroline is similar to that of other phenanthrolines, and, since the method of Smith, McCurdy and Diehl, as described, was inapplicable in the presence of bismuth, attention was directed to existing methods applicable to bismuth that make use of 1 : lO-phenanthr~line?,~ In general, they appear to suffer from poor reproduci- bility, together with other undesirable features, such as slow formation of colour, fairly critical adjustment of pH and occasional precipitation of basic bismuth salts. There are also con- siderable differences of opinion on the choice of reductant for iron111 and on the order of addition of the reagents.Since the substitution of bathophenanthroline for 1 : 10-phenan- throline could not be expected to overcome these difficulties, our attention was turned to the factors affecting the formation of the ferrous plhenanthrolines. At pH 4, in non-complexing media, the rate of formation of ferroin (ferrous 1 : 10-phenan- throline) from iron11 and the usually employed 100 to 200-fold excess of reagent is im- measurably fast. A slow rate of colour formation indicates, therefore, a deficiency of some or all of the reacting ionic species. In the presence of bismuth, the iron complex must be finally formed in the presence of appreciable arnounts of reagents such as ethylenediamine- tetra-acetic acid (EDTA), tartrate or citrate, the last named being the most effective complexing agent.Both iron111 and iron11 are complexed by citrate5 and the greater stability of the iron111 complexes makes the rcduction of iron111 more difficult. Further, there will be competition for iron11 between the citrate and the phenanthroline. Hence it may be assumed that attempts to form the ferrous complex in the presence of citrate a t pH 4, with reducing agents such as hydroxylamine or hydroquinone, are not likely to be entirely satisfactory . EXPERIM:ENTAL From the above-mentioned consideration, the reduction stage was carried out in 2 M acid with tin11 as the reductant. The order of addition of the other reagents was then investigated. It was immediately apparent that the bathophenanthroline must be added before the citrate, otherwise little or no colour was produced and the rate of formation of the colour was far too slow.It was further observed that, when iron11 was added to a solution already containing bathophenan- throline and citrate at pH 4, no trace of the coloured complex was visible, even after a long period of time.February, 19581 DETERMINATION OF TRACES OF IRON IN BISMUTH 81 From the work of Lee, Kolthoff and LeussingJ6y7 and our observations on the lack of reactivity of iron11 in the presence of citrate, it appeared likely that the intermediate ionic species occurring in the presumably stepwise formation of Fe(bathophenanthroline)2+ are present in relatively concentrated acid solutions. When the pH is subsequently raised to 4, the final complex should be formed rapidly and quantitatively.The rate of formation of the ferrous complex was, however, still slow, and this could only be attributed to too small a concentration of bathophenanthroline. Owing to the low solubility of bathophenanthroline in water, further addition of the reagent was impracticable, but the addition of 10ml of ethanol before the addition of the citrate ensured that all the added reagent remained in solu- tion. This produced a striking improvement-the red colour now appearing instantaneously. This marked dependence of the rate of formation of the complex on the concentration of bathophenanthroline agrees qualitatively with the findings of Lee, Kolthoff and Leussing6 who have shown that the law governing the rate of formation of ferroin involves a third-power dependence on the concentration of phenanthrolinium ions.Recoveries of added iron from the proposed base solutions were, however, not very satisfactory, since they are dependent on the amount of citrate present. The results are shown in Table I. TABLE I RECOVERY OF IRON FROM VARIOUS BASE SOLUTIONS Iron Iron Optical density at Base solution added, p g recovered, pg 533 mp in 4-cm cell Acetate . . .. .. .. .. . . 11.2 11.2 0-710 10 ml of a 50 per cent. solution of sodium citrate 11-2 10-3 0.650 25 ml of a 50 per cent. solution of sodium citrate 11.2 8.2 0.520 Trials with other complexing agents were carried out, but finally the most satisfactory was found to be a mixture of 20 ml of 0.1 M EDTA and 10 ml of a 50 per cent.solution of sodium citrate. Recoveries of added iron were then carried out at various levels, both in the presence and absence of bismuth. The results are shown in Table 11. TABLE I1 RECOVERY OF ADDED IRON BY THE PROPOSED METHOD Bismuth present, Iron added, Iron recovered, g Pg Pg 11.2 11.2 11.3 11.2 { ;;:; 11-6 5.8 5-6 { 4.7 4-9 1.4 1.1 1.4 1.4 1.1 i 1.1 1 1.1 Optical density a t 533 mp 0.653 0.658 0.654 0.628 0.683 0.342 0.276 0.286 0.085 0.064 0.082 0.086 0.062 0-066 in 4-cm cell It can be seen that the optical density equivalent to 11.2 pg of iron is lower in Table I1 than it is for the experiment in an acetate base solution in Table I. This is because recoveries from the citrate base solution are reproducibly lower than those from the acetate base solution.If the calibration is prepared in the citrate solution, 100 per cent. relative recoveries will be obtained. The above-mentioned differences in optical densities would no doubt disappear if the concentration of bathophenanthroline were increased to the point where a vanishingly small amount of iron11 remained uncomplexed in acid solution. Since the free acidity of the bismuth solution must be 1 to 2 M to prevent precipitation of oxy salts, and since the complexing of iron11 is dependent on the ratio of the concentration of reagent to hydrogen82 BOOTH AND EVETT [Vol. 83 ions, it is impracticable to attempt to achieve the conc:entration of reagent likely to be necessary. The proposed method gives excellent results at all levels down to 1 pg of iron.Levels below this can be determined, the limits being dependent on the magnitude and reproducibility of the blank value. METHOD REAGENTS- water should be used throughout. throline in 175 ml of ethanol. All reagents should be of recognised anal.ytica1 grade. Redistilled or demineralised Bathophenanthroline solution, 0.2 per cent.-Dissolve 0.5 g of 4 : 7-diphenyl-l : 10-phenan- Sodium citrate solution, 50 $er cent. w/v. EDTA solution-A 4 per cent. w/v solution of the disodium salt of ethylenediaminetetra- Stannous chloride solution-A 10 per cent. w/v solution of stannous chloride dihydrate. Hydrochloric acid, concentrated, 6 M and 2 M. Nitric acid, 16 M. n-Hexyl alcohol. Ethanol. Dilute to 250 rrtl with wa.ter and store in polythene. acetic acid. PROCEDURE- Dissolve the bismuth in a mixture of 6 M hydrochloric acid and 16 M nitric acid (10 ml of hydrochloric acid and 0-5 ml of nitric acid per g of bismuth). Evaporate the solution almost to dryness under an infra-red lamp.Re-dissolve the solid in concentrated hydro- chloric acid and repeat the evaporation. Dissol.ve the residue in concentrated hydrochloric acid and dilute with water so that the solution contains 0.2g of bismuth per ml and is approximately 2 M with respect to hydrochloric acid. With a pipette, place 5 ml of the bismuth solution in a 10-ml beaker, add 0-2 ml of stannous chloride solution and boil. Coo1 to room temperature and transfer the solution to a 100-ml separating funnel. Rinse out the beaker with two 5-ml portions of ethanol, and finally with 2 ml of 2 M hydrochloric acid, transferring the washings to the funnel.Add 4 ml of bathophenanthroline solution and mix thoroughly. Mix 20 ml of EDTA solution with 10 ml of sodium citrate solution, and add it to the solution in the funnel, swirling to dissolve any precipitate formed. Set the solution aside for 5 minutes and then extract the ferrous complex with 10 ml of n-hexyl alcohol. Discard the aqueous phase and transfer t.he alcohol extract to a 25-ml cali- brated flask, rinsing the funnel with ethanol. Ililute almost to the mark with ethanol, add 0.2 ml of 2 M hydrochloric acid and shake until a clear solution is obtained. Dilute to the mark with ethanol, filter the solution into a 4-cm cell and measure the optical density at 533 mp. Carry out a blank exactly as described above, but omitting the sample. PROCEDURE FOR RECOVERING THE BATHOPHENA.NTHROLINE- The present price of bathophenanthroline makes it desirable to recover it for further use, Evaporate the n-hexyl alcohol extract to dryness under Warm the residue with 1OIV sodium hydroxide and then extract with and this can be done as follows. reduced pressure. hot benzene. Evaporate the benzene; the residue should be fit to use again. REFERENCES 1. 2. 3. Holmes, D. G., Atomic Energy Research Establishment Report CE/R 2025, Harwell, 1956; 4. 5. 6. 7. Sandell, E. B. , “Colorimetric Determination of Traces of Metals,” Second Edition, Interscience Smith, G. F., McCurdy, W. H., jun., and Diehl, H., Analyst, 1952, 77, 418. Austing, C. E., personal communication. Hamm, R. E., Shull, C. M., and Grant, D. M., J . Amer. Chern. Soc., 1954, 76, 2111. Lee, T. S., Kolthoff, I. M., and Leussing, D. L., Ibid., 1948, 70, 2348. -,-,- , Ibid., 1948, 70, 3596. Publishers Inc., New York and London, 1950. Analyst, 1957, 82, 628. Received July 2nd) 1967