Atomic Weight of Antimony

That the atomic weight of antimony is approximately 122, and not a multiple or submultiple of this amount, is indicated by several considerations:

The specific heat of antimony between 0 and 100° C. averages 0.05 calorie. Assuming a mean atomic heat of 6.4, the atomic weight, according to Dulong and Petit's Law, is approximately 128.

The properties of antimony indicate that the most appropriate position in the Periodic Table for this element lies immediately below arsenic, in the fifth group. This places it between tin (At. wt. 118.7) and tellurium (At. wt. 127.61), so that its atomic weight should lie between these values.

The atomic number of antimony, namely 51, confirms its position between tin (At. No. 50) and tellurium (At. No. 52).

Application of Avogadro's hypothesis to the results of vapour density determinations of volatile antimony compounds indicates that the atomic weight of antimony is approximately 122.

The mass spectrum of antimony consists of two strong lines, 121 and 123 respectively, so that the atomic weight of the element must lie between these two values.

The dissatisfaction expressed by Berzelius in the words "Ich habe niemals mit einer Materie, wo es so ausserordentlich schwer gewesen ist, konstante Resultate zu erhalten, gearbeitet" has been experienced by many workers on the atomic weight of antimony, and few elements have proved so troublesome in this respect. It is only during the last few years that consistent values have been obtained; therefore, in the accompanying table, it will suffice merely to mention in most cases the mean results of the earlier researches. All the atomic weight values have been recalculated from the ratios given, using in addition to the antecedent data quoted in the Introduction, p. x, the following values: Cu, 63.57; Ba, 137.6.

Of the researches prior to 1921, those of Cooke alone need be mentioned.

Cooke attempted to exercise the same care in his work as Stas had done in his classical determinations. He worked, however, with small quantities, whereas Stas sometimes used more than 100 grams. Cooke therefore reduced the error due to occlusion of solution by the precipitate by employing much more dilute solutions. In his initial work he employed four methods.

(a) Synthesis of Antimony Trisulphide. - Balls of antimony were treated with hydrochloric acid containing a little nitric acid, and the solution boiled until it became colourless. The balls were then removed, washed, dried and weighed; the loss in weight gave the amount of antimony dissolved. The solution was diluted with aqueous tartaric acid, and antimony trisulphide precipitated by means of water saturated with hydrogen sulphide. The washed precipitate was dried at 130° C. The analysis of this gave 2Sb:3S =71.4269:28.5731, whence Sb = 120.22. On heating to 210° the red trisulphide changed into the black variety. This gave the ratio 2Sb:3S =71.4818:28.5182; whence Sb = 120.56.

(b) Analysis of Antimony Trichloride. - The material was purified by distillation and by crystallisation from carbon disulphide. The analysis was carried out by dissolving the trichloride in aqueous tartaric acid and adding silver nitrate solution. The ratio SbCl₃:3AgCl = 53.066:100 gives a value for antimony, 121.82, almost identical with that obtained by Dumas, using the same method. Cooke, however, was not satisfied with the result since it did not agree with certain of his other determinations. He advanced the suggestion that the high result was due to the presence of some oxychloride in the trichloride; this, however, is not supported by the evidence of his own work.

(c) Analysis of Antimony Tribromide. - The tribromide was prepared by the action of powdered metallic antimony upon bromine in carbon disulphide solution. It was purified by distillation over finely powdered antimony and crystallised from carbon disulphide. The ratio SbBr₃:3AgBr = 63.830:100 gave Sb =119.863.

(d) Analysis of Antimony Triiodide. - The analysis was carried out in the same way as those of the tribromide and trichloride. The ratio SbI₃:3AgI = 71.060:100 gave Sb =119.786.

In the years 1880 and 1881 Cooke carried out his final determinations using antimony tribromide. The material was repeatedly distilled from metallic antimony, recrystallised several times from carbon disulphide, subjected to repeated fractional distillation, and finally twice sublimed in a current of carbon dioxide. It is probable that Cooke would have obtained better results had his process of purification been less prolonged. Working without modern refinements in the handling of highly hygroscopic materials, the introduction of a trace of moisture was inevitable; thus the carbon dioxide, though described as absolutely dry, was only subjected to the action of calcium chloride and sulphuric acid. It is clear, therefore, that the resublimed product was much more likely to contain hydrogen bromide than antimony oxybromide, the impurity which Cooke feared. Willard and McAlpine, as a result of a critical study of Cooke's papers, consider that his material may have contained as much as 1 per cent, of hydrogen bromide.

On the basis of the work of Kessler, Dexter and Dumas the value 122 was adopted, although Schneider's results pointed to the lower value. After the laborious investigation carried out by Cooke, which gave results of such striking concordance, the number 120 was immediately adopted. The electrochemical studies of Pfeifer and Popper indicated the higher value once more, but so great was the prejudice in favour of Cooke's work that no alteration was made; moreover the electrochemical work was adversely criticised by Cohen, Collins and Strengers, on the ground that the method did not give constant results. The work of Friend and Smith, however, indicated that Cooke's results were somewhat too low, so that after 1902 the number 120.2 was adopted; an unjustifiable compromise which was obviously unsatisfactory. The controversy continued; certain workers on antimony appeared to find the value 120.2 satisfactory. Others, however, obtained results pointing to the higher value, and expressed the opinion that the older value, 122, was the more correct.

The insecurity of the basis for the atomic weight led Willard and McAlpine in 1921 to reinvestigate the whole question. They prepared pure antimony tribromide with careful exclusion of moisture. In an all-glass apparatus, three different preparations of antimony were combined with bromine, the product twice distilled at a pressure of 5 to 10 mm., then distilled a third time at less than 1 mm. into a series of small bulbs which were sealed off from each other as individual samples. From the time the pure dry materials were placed in the apparatus until the bulbs were broken under tartaric acid solution, only inert gases came into contact with the preparation. The resulting product was analysed for bromine in two ways: first, volumetrieally, by finding the amount of silver, dissolved in nitric acid, equivalent to the sample, using a nephelometric end point; second, gravimetrically, by adding excess of silver nitrate, then filtering out and weighing the silver bromide. The precautions taken and corrections applied included all those which had been described within recent years on similar work. In eight of the best volumetric analyses, a total of 35.69757 grams of antimony bromide formed 55.63121 grams of silver bromide, from which the atomic weight of antimony is 121.768. By taking into consideration the three slightly less satisfactory volumetric analyses, and eight gravimetric analyses, Willard and McAlpine gave the mean value 121.773.

Knop obtained an appreciably higher value. He treated pure antimony with nitric acid and converted the product into the tetroxide by ignition at 850° to 900° C., at which temperature the pentoxide is fully reduced to the tetroxide, but the latter is not further reduced. The purity of the product was established by the iodine-thiosulphate method. The results gave a mean value Sb = 122.06, or 121.96 when reduced to vacuum.

Honigschmid, Zintl and Linhard hydrolysed chloroantimonic acid, HSbCl₆, 4.5H₂O (prepared from antimony pentasulphide) and reduced the resulting antimonic acid in hydrogen at 500° C. The metal was converted into the chloride or bromide by heating in a current of the halogen, and the halides fractionally distilled, first in pure nitrogen and then in a vacuum. The silver equivalent of each halide was determined by gravimetric titration and weighing the silver halide formed. The mean of thirty-two very concordant results gave Sb =121.76.

Weatherill applied Willard and McAlpine's method to the trichloride. Kahlbaum's purest antimony was twice fused in hydrogen, combined with pure chlorine, and the product repeatedly distilled in an evacuated glass apparatus, considerable head and tail fractions being rejected in each distillation. The mean of 8 analyses gave the ratio SbCl₃:3Ag as 0.704864, from which the atomic weight of antimony is 121.748. This is slightly lower than the values obtained by Willard and McAlpine, and by Honigschmid, but agrees remarkably well with that of Krishnaswami.

The discovery of non-radioactive isotopes of certain elements has taught that elements from different localities may conceivably possess their constituent isotopes in different proportions, so that their atomic weights may vary. A review of the earlier work on the atomic weight of antimony led Muzaffar to inquire whether or not such might be the case with this element. Stibnite was obtained from Peru, Bolivia, Borneo and Hungary. After purifying all samples by the same method, the antimony was converted into trichloride and the ratio between antimony trichloride and potassium bromate determined by titration:

3SbCl₃ + KBrO₃ + 6HCl = 3SbCl₅ + KBr + 3H₂O

The results were as follows:

Source of Stibnite	Ratio 3Sb:KBrO3	No. of Experiments.	Atomic Weight of Antimony.
Hungary	2 17592	7	121.138
Borneo	2.1836	7	121.565
Peru	2.1862	7	121.710
Bolivia	2.1986	11	122.400

Excellent concordance was obtained in the first set of results using Hungarian material, and the low value for the atomic weight is remarkable.

More recently specimens of stibnite have been obtained from the same sources as those used by Muzaffar. The metal was extracted, and the densities compared with that of a specimen of Kahlbaum's antimony. In addition, solutions were titrated with solutions of potassium bromate and the ratio KBrO₃/3Sb was determined. It was found that the densities of the specimens were all within 0.1 per cent, of each other; and that the ratios KBrO₃/3Sb agreed to 0.05 per cent. The evidence of variation in the atomic weight of antimony from different sources does not therefore appear to have been confirmed.

Krishnaswami directed attention to certain disadvantages occurring in practice when Muzaffar's method is adopted, and gave the results of determining the atomic weight of antimony from four ores of Indian and Burmese origin, using Willard and McAlpine's method in its entirety. His results were as follows:

Source of Material.	Ratio SbBr3:3AgBr.	No. of Experiments.	Atomic Weight of Antimony.
Kahlbaum's Sb2O3	0.641664	6	121.758
Mysore stibnite	0.641647	4	121.748
Mysore cervantite	0.641647	3	121.748
Amherst stibnite	0.641652	5	121.751
Shan States stibnite	0.64159	3	121.716

A close agreement between the values from the Mysore stibnite and cervantite was of course to be expected since the latter is an alteration product of the former. The results do not indicate any appreciable difference between the samples.

The International Committee on Atomic Weights for 1936 has adopted the value

Sb = 121.76

The Council of the Chemical Society had, in 1919, recommended this value and it has been retained until the present time (1936).