Atomic Weight of Palladium

That the atomic weight of palladium is approximately 107, and not a multiple or fraction of this amount, is clearly indicated by several considerations. Chief amongst these are:

1. The specific heat of palladium is 0.059; assuming, therefore, a mean atomic heat of 6.4, the atomic weight of the metal, according to Dulong and Petit's Law, is approximately 108.
2. A study of the chemical properties of palladium and its compounds reveals a close analogy, between this element and nickel and platinum. Furthermore, a more or less regular gradation in properties can be traced as we compare ruthenium with rhodium, and rhodium with palladium. These facts suggest that palladium should be placed at the end of one of the triads of Group VIII in the same vertical line as nickel and platinum. At present there is only one such vacancy, namely, that next to rhodium, and befitting an element of atomic weight lying between 102.9 (atomic weight of rhodium) and 107.88 (atomic weight of silver).
3. Alkali chlor-palladites and chlor-palladates are isomorphous with the corresponding chlor-platinites and chlor-platinates. Application of Mitscherlich's Law of Isomorphism leads to the conclusion that these salts have a similar constitution, and are therefore represented by the general formulae M₂PdCl₄ and M₂PdCl₆, assuming those for the platinum derivatives to be M₂PtCl₄ and M₂PtCl₆ respectively.

Analyses of these compounds indicate that the atomic weight of palladium is 106.7.

The first determinations of the atomic weight of palladium were made by Berzelius. His analysis of palladium monosulphide is of no present value; and the same remark applies to his analyses of potassium chlor-palladite, K₂PdCl₄, although until the year 1889 the accepted value for the atomic weight of palladium was based upon the results. Berzelius2 decomposed the salt by ignition in a stream of hydrogen, determining the loss in weight and the amounts of palladium and potassium chloride in the residue. His results lead to values for the atomic weight of palladium varying from 104.9 to 111.1, probably because the double salt was incompletely dried; his ratio Pd:2KCl, which is independent of this source of error, leads to the value Pd = 106.2.

The first modern determination of the atomic weight of palladium was published by Reiser in 1889. This chemist considered that palladosammine chloride, Pd(NH₃)₂Cl₂, was the most suitable palladium compound for atomic weight work, and most subsequent experimenters have utilised this compound or the corresponding bromide. Keiser analysed the compound for palladium by heating it in a stream of pure, dry hydrogen, and this method has also been employed by Bailey and Lamb, Keller and Smith, Keiser and Breed, Krell, Woernle, and Kemmerer. On the other hand, Keller and Smith, who consider that traces of palladium are lost by this method, determined the palladium content of the compound electrolytically, as also did Amberg in two series of experiments, and Woernle. Moreover, Amberg made a series of determinations in which the reduction to metal was accomplished in solution with hydrazine sulphate, while Shinn effected his reductions with ammonium formate. The results of these investigations are summarised:

Pd(NH₃)₂Cl₂:Pd::100.000:x

where x within the range 50.171-50.550 and atomic weight of palladium within the range 105.70-107.31, respectively.

Two of the preceding investigators determined the chlorine in the compound by weighing it as silver chloride: first, Bailey and Lamb, who collected the ammonium chloride evolved when palladosammine chloride was heated in hydrogen; and, secondly, Amberg, who precipitated the chloride in his filtrates after removing the palladium (electrolytically in series (i) and with hydrazine sulphate in series (iii) above):

2AgCl:Pd(NH₃)₂Cl₂::100.000:x

where x within the range 73.807-73.818 and atomic weight of palladium within the range 106.61-106.64, respectively.

Haas, Gebhardt, and Gutbier analysed palladosammine bromide by reduction in hydrogen, with the following results:

Pd(NH₃)₂Br₂:Pd::100.000:x

where x within the range 35.491-35.495 and atomic weight of palladium within the range 106.67-106.69, respectively.

A few investigations have been made with other palladium compounds. Joly and Leidie used potassium chlor-palladite, K₂PdCl₄. The dried salt was electrolysed in a solution acidulated with hydrochloric acid, the deposit of palladium weighed, the potassium chloride recovered from the liquid by evaporation and also weighed. Three experiments were carried out with a salt containing a little water, and a second series of four experiments made with a salt that had been dried in vacuo at 100° C. The results were not very concordant. dFinally, in two experiments the salt was reduced by heating it in hydrogen, and the palladium and potassium chloride determined in the residue:

1st series	2KCl:Pd::100.000:70.979	Pd = 105.84
2rd series	K2PdCl4:Pd::100.000:32.450	Pd = 105.70
	K2PdCl4:2KCl::100.000:45.528	Pd = 107.50
	2KCl:Pd::100.000:71.257	Pd = 106.25
3rd series	K2PdCl4:Pd::100.000:32.480	Pd = 105.84
	K2PdCl4:2KCl::100.000:45.700	Pd = 106.26
	2KCl:Pd::100.000:71.055	Pd = 105.95

Hardin employed three compounds, diphenylpalladodiammonium chloride and bromide, and palladium ammonium bromide. Each compound was analysed for palladium by reduction at a red heat in hydrogen, with the following results:

7 expts.	Pd(C6H5.NH2Cl)2:Pd::100.000:29.390	Pd = 107.00
5 expts.	Pd(C6H5.NH2Br)2:Pd::100.000:23.622	Pd = 107.01
4 expts.	(NH4)2PdBr4:Pd::100.000:23.1135	Pd = 106.94

Kemmerer, in addition to studying palladosammine chloride, reduced the corresponding cyanide to the metal in a stream of hydrogen:

6 expts. Pd(NH₃)₂(CN)₂:Pd::1000.000:55.291 whence Pd = 106.46

There is a wide variation in the preceding results, which point strongly, however, to a value for the atomic weight of palladium about half-way between Pd = 106 and Pd = 107. Shinn expresses the opinion that palladosammine chloride, which has been used in most of the determinations, has probably never been prepared perfectly free from traces of palladium ammonium chloride, in which case most workers have not analysed a truly homogeneous substance. Should this criticism be substantiated, it is probable that a similar one would hold against the corresponding bromide and cyanide and the organic derivatives employed by Hardin.

The International Committee's table for 1918 gives

Pd = 106.7.

This value is probably not far from the truth, since with the exception of Kemmerer's results and the undoubtedly defective series (i) of Amberg's electrolytic determinations, all the modern values for the atomic weight of palladium fall between the limits 106.6 and 106.7.