Atomic Weight of Iron

That the atomic weight of iron is approximately 56, and not a multiple or submultiple of this amount, is indicated by various considerations.

1. The mean specific heat of iron is approximately 0.110. Assuming a mean atomic heat of 6.4, the atomic weight of iron, according to Dulong and Petit's Law, is approximately 58.
2. As already indicated, iron closely resembles manganese in many of its chemical properties, and the three elements iron, cobalt, and nickel constitute very fitting intermediaries between manganese and copper. Reference to the Periodic Table shows that the only manner in which this relationship can be harmonised with the Periodic Law is to assume that the atomic weights of these three metals lie between 54.93 (the atomic weight of manganese) on the one hand and 63.57 (the atomic weight of copper) on the other.
3. Ferric sulphate yields, with the sulphates of the alkali metals, a series of well-defined crystalline salts which are isomorphous with similar series yielded by aluminium sulphate. By the application of Mitscherlich's Law, therefore, analogous formulae are to be anticipated, so that the general formula for these iron alums is
   
   M₂SO₄.Fe₂(SO₄)₃.24H₂O.
   
   Analyses of these compounds indicate that the atomic weight of iron is 56.

The early determinations of the atomic weight of iron are of no present value, and little need be said concerning them. Most investigators chose to determine the composition of ferric oxide; Wackenroder, Svanberg and Norlin, Erdmann and Marchand, and Rivot worked by reducing the weighed oxide to metal in a stream of hydrogen, while Berzelius, Maumene, and also Svanberg and Norlin converted a known weight of iron into ferric oxide by oxidation.

The only other early determinations to be mentioned are Dumas' analyses of anhydrous ferrous and ferric chloride, in which the amount of silver required to combine with the chlorine was determined. and a few experiments by Winkler, in which a weighed amount of iron was dissolved in a solution of iodine in potassium iodide, the excess of iodine being determined by titration with sodium thiosulphate. The results were as follow (Cl = 35.457, I = 126.92, Ag = 107.880): -

2Ag:FeCl₂:: 100.000: 58.866; Fe = 56.09
Ag:FeCl₃:: 100.000: 50.244; Fe = 56.23
I₂: Fe:: 100.00: 22.145; Fe = 56.21

Modern work on the atomic weight of iron begins in 1900 with Richards and Baxter's analyses of ferric oxide by reduction to the metal in a stream of hydrogen. Richards and Baxter found that the exact determination of this ratio is a matter of extreme difficulty, and regarded their experiments as preliminary in character. The mean result was as follows (O = 16.000): -

Fe₂O₃: 2Fe:: 100.000: 69.9576; Fe = 55.887

Three years later Baxter made another determination of the atomic weight of iron, analysing anhydrous ferrous bromide for the purpose. The salt was sublimed in a porcelain tube, which introduced a little sodium bromide into it; due allowance was made for this source of error, and the following results obtained (Ag = 107.880, Br = 79.916): -

FeBr₂: 2AgBr:: 57.4195: 100.000; Fe = 55.833
FeBr₂: 2Ag:: 99.960: 100.000; Fe = 55.842

At the time these experiments were made an erroneous value for the atomic weight of silver was in use, in consequence of which the value Fe = 55.87, in confirmation of that deduced from the oxide analyses, was deduced from the bromide analyses. With the establishment of the modern value for the atomic weight of silver, it accordingly became desirable to repeat the preceding work in order to determine the source and magnitude of the errors involved in it. Hence, in 1911, Baxter, Thorvaldson, and Cobb repeated the analyses of ferrous bromide, which they were able to obtain quite free from sodium bromide by utilising fused quartz apparatus in its preparation. Their preliminary analyses gave the following results (Ag = 107.880, Br = 79.916): -

FeBr₂: 2Ag:: 99.9593: 100.0000; Fe = 55.840
FeBr₂: 2AgBr:: 57.4221: 100.0000; Fe = 55.840

and the final experiments gave almost identical results: -

FeBr₂: 2Ag:: 99.9583: 100.0000; Fe = 55.838
FeBr₂: 2AgBr:: 57.4214: 100.0000; Fe = 55.838

The essential accuracy of Baxter's earlier work on the bromide was thus confirmed.

Baxter and Thorvaldson extended the preceding investigation by making a number of analyses of ferrous bromide prepared from meteoric iron. The results were as follow: -

FeBr₂: 2Ag:: 99.9561: 100.0000; Fe = 55.834
FeBr₂: 2AgBr:: 57.4191: 100.0000; Fe = 50.829

These results are a trifle lower than the preceding, but in each series of five experiments two are of doubtful value as the result of a modification of the method of analysis; excluding the doubtful analyses, the above two results become Fe = 55.837 and Fe = 55.835 respectively.

The preceding results leaving little doubt that Richards and Baxter's analyses of ferric oxide are affected by a slight, but, nevertheless, appreciable error, Baxter and Hoover undertook a thorough revision of this process. For full details of their work, which, though apparently quite simple, was, in reality, extremely difficult, the reader must be referred to the original memoir. The result was as follows (O = 16.000): -

Fe₂O₃: 2Fe:: 100.0000: 69.9427; Fe = 55.847

The preceding result requires slight correction. Baxter and Hoover found that at 1050° to 1100° C., the temperature at which they prepared and reduced their ferric oxide, there was a slight doubt as to the stability of the ferric oxide. Five grams of the oxide when ignited to constant weight in oxygen, lost in weight when ignited in air, the loss being one- fifth of a milligram. The subsequent researches of Sosman and Hostetter have shown that this loss in weight is due to the dissociation of ferric oxide into magnetic oxide and oxygen. Thus the higher weight is the correct weight of ferric oxide. Baxter and Hoover, however, chose the lower, regarding the higher weight as due to occluded oxygen. The necessary correction to the atomic weight of iron is — 0.007, and so the corrected value is Fe = 58.840. Of the twelve analyses on which this result is based, seven were made with terrestrial and five with meteoric iron, and no difference in the results was observed. The International Committee for 1920 give the value

Fe = 55.84.