Atomic Weight of Platinum

Several lines of argument lead us to the conclusion that the atomic weight of platinum is approximately 195, and not a multiple or submultiple of this amount. The more important of these may be summarised as follows:

1. The specific heat of platinum between 0° and 100° C. is 0.0323. Assuming a mean atomic heat of 6.4, application of Dulong and Petit's Law leads to the value 198 for the approximate atomic weight of platinum.
2. The chemical properties of platinum and its compounds exhibit interesting resemblances with those of nickel and palladium, on the one hand, and, on the other, they serve to complete the gradation in properties observable as we pass from osmium through iridium to platinum. There can thus be only one position for the metal in the Periodic Scheme, namely, at the close of the third vertical column in Group VIII, and with an atomic weight slightly greater than 193.1 (at. wt. of iridium).
3. The measurements made by Zambonini of the crystalline potassium salts of plumbic, stannic, and platinic acids shows that the salts are isomorphous. By Mitscherlich's Law, therefore, they should have a similar constitution, and assuming the first two salts to be represented by the formulae K₂Pb(OH)₆ and K₂Sn(OH)₆ respectively, it follows that the platinate has the formula K₂Pt(OH)₆. In other words, a molecule of potassium platinate contains one complete atom of platinum, and one only. Similarly potassium chlor-platinate is isomorphous with potassium chlor-stannate, K₂SnCl₆, and may be assumed to have the formula K₂PtCl₆. Analysis of these compounds indicates, therefore, that the atomic weight of platinum is 195.2 and not a multiple or simple fraction of this amount.

The first determination of the atomic weight of platinum was made by Berzelius in 1826. From the analysis of platinous chloride he found Pt = 194.7. Two years later he published his determination of the composition of potassium chlor-platinate; 6.981 parts of the salt lost 2.024 of chlorine when ignited in hydrogen, leaving a residue containing 2.822 of platinum and 2.135 of potassium chloride. Hence:

From K₂PtCl₆: 4Cl; whence Pt = 198.3
From K₂PtCl₆: 2KCl; whence Pt = 196.6
From K₂PtCl₆: Pt; whence Pt = 197.4
From 2KCl: Pt; whence Pt = 197.1

These results are now known to be extremely high, but until 1881 the accepted figure for the atomic weight of platinum rested entirely upon them, and they appeared to be corroborated by some analyses published by Andrews in 1852.

In 1881 Seubert, by his analyses of ammonium and potassium chlor- platinate, showed that the atomic weight of platinum approaches the value Pt = 195, and not 197. His results were confirmed by Halberstadt in 1884, while Dittmar and McArthur three years later arrived at a somewhat higher figure as the result of four rather unsatisfactory analyses of potassium chlor-platinate.

The modern value for the atomic weight of platinum is based upon the extremely careful analyses of the chlor-platinates and brom-platinates of potassium and ammonium published by Archibald in 1909. The utmost precautions which modern experience has shown to be necessary were taken in preparing the salts in a state of purity and dryness, and in effecting the numerous analytical operations.

The analyses of potassium chlor-platinate were carried out in the following manner. The pure dry salt was weighed and heated in a stream of hydrogen:

K₂PtCl₆ + 2H₂ = 2KCl + Pt + 4HCl.

The hydrogen chloride produced was absorbed in water and the hydrogen chloride in the solution determined by titration against a weighed amount of silver, dissolved in nitric acid, the end point being determined nephelometrically. The precise silver equivalent of the hydrogen chloride having been thus determined, the silver chloride produced was collected, washed, dried, and weighed. The residual mixture of potassium chloride and platinum was treated with water to wash out the potassium salt, and the platinum collected, heated in hydrogen, cooled, and weighed. The chlorine in the potassium chloride was determined first by titrating it against silver, and then by collecting and weighing the silver chloride produced in that operation. Thus, corresponding to a certain weight of potassium chlor-platinate, there were obtained two weights of silver, two weights of silver chloride, and one weight of platinum, the six weights being in the following proportions:

K₂PtCl₆, 4Ag, 4AgCl, Pt, 2Ag, 2AgCl.

It will be immediately obvious that such complete analyses, if concordant, afford valuable information regarding the purity of the potassium chlor-platinate. Nine values for the atomic weight of platinum may be deduced from each experiment.

The nine values for the atomic weight should, of course, be highly concordant if the pure dry salt has been accurately analysed. In the event of the original salt not being perfectly dry, all the ratios not involving the weight of the salt should yield the same figure for the atomic weight in question, but the other ratios should give slightly different results.

The analyses of potassium brom-platinate were effected in a manner strictly analogous to those of the chlor-platinate. The analyses of the ammonium salts were simpler; the initial reaction being as follows:

(NH₄)₂PtX₆ + 2H₂ = Pt + 2NH₄X + 4HX,

and the ammonium halide being volatile at the temperature employed, the halogen in the original salt was collected and determined in one operation instead of two.

The results of Archibald thus furnish data for twenty-eight ratios, from which twenty-eight values for the atomic weight of platinum may be deduced. The extreme difference between these values is only 0.083 on a high atomic weight, a very satisfactory result.

The arithmetic mean of the twenty-eight results is Pt = 195.227, the weighted mean, each result being weighted in proportion to the number of experiments made in fixing the ratio, is Pt = 195.229. Eliminating all ratios involving the weights of the chlor-platinates, since the salts may possibly have contained traces of water, the arithmetic mean of the values derived from the other ratios is Pt = 195.221, and the weighted mean is Pt = 195.224. Thus the atomic weight of platinum is most probably Pt = 195.22 with an uncertainty of two or three units in the second decimal place, the uncertainty being partly due to the small errors of experiment, and partly to possible slight errors in the other atomic weights used in the calculations.

In 1912 Schultz published five analyses of ammonium chlor-platinate, giving as a mean result Pt = 194.5. These experiments are clearly of little value.

The International Committee's table for 1918 gives

Pt = 195.2.