Ammonia, like potassium cyanide, has the faculty of easily reacting with platinum dichloride, forming compounds similar to the platinocyanide and cobaltia compounds, which are comparatively stable. But as ammonia does not contain any hydrogen easily replaceable by metals, and as ammonia itself is able to combine with acids, the PtX2 plays, as it were, the part of an acid with reference to the ammonia. Owing to the influence of the ammonia, the X2 in the resultant compound will represent the same character as it has in ammoniacal salts; consequently, the ammoniacal compounds produced from PtX2 will be salts in which X will be replaceable by various other haloids, just as the metal is replaced in the cyanogen salts; such is the nature of the platino-ammonium compounds. PtX2 forms compounds with 2NH3 and with 4NH3, and so also PtX4 gives (not directly from PtX4 and ammonia, but from the compounds of PtX2 by the action of chlorine, &c.) similar compounds with 2NH3 and with 4NH3.[12]
If ammonia acts on a boiling solution of platinous chloride in hydrochloric acid, it produces the green salt of Magnus (1829), PtCl2,2NH3, insoluble in water and hydrochloric acid. But, judging by its reactions, this salt has twice this formula. Thus, Gros (1837), on boiling Magnus's salt with nitric acid, observed that half the chlorine was replaced by the residue of nitric acid and half the platinum was disengaged: 2PtCl2(NH3)2 + 2HNO3 = PtCl2(NO3)2(NH3)4 + 2PtCl2. The Gros's salt thus obtained, PtCl2(NO3)24NH3 (if Magnus's salt belongs to the type PtX2, then Gros's salt belongs to the type PtX4), is soluble in water, and the elements of nitric acid, but not the chlorine, contained in it are capable of easily submitting themselves to double saline decomposition. Thus silver nitrate does not enter into double decomposition with the chlorine of Gros's salt. Most instructive was the circumstance that Gros, by acting on his salt with hydrochloric acid, succeeded in substituting the residue of nitric acid in it by chlorine, and the chlorine thus introduced, easily reacted with silver nitrate. Thus it appeared that Gros's salt contained two varieties of chlorine—one which reacts readily, and the other which reacts with difficulty. The composition of Gros's first salt is PtCl2(NH3)4(NO3)2; it may be converted into PtCl2(NH3)4(SO4), and in general into PtCl2(NH3)4X2.[13]
The salt of Magnus when boiled with a solution of ammonia gives the salt (of Reiset's first base) PtCl2(NH3)4, and this, when treated with bromine, forms the salt PtCl2Br2(NH3)4, which has the same composition and reactions as Gros's salt. To Reiset's salts there corresponds a soluble, colourless, crystalline hydroxide, Pt(OH)2(NH3)4, having the properties of a powerful and very energetic alkali; it attracts carbonic anhydride from the atmosphere, precipitates metallic salts like potash, saturates active acids, even sulphuric, forming colourless (with nitric, carbonic, and hydrochloric acids), or yellow (with sulphuric acid), salts of the type PtX2(NH3)4.[14] The comparative stability (for instance, as compared with AgCl and NH3) of such compounds, and the existence of many other compounds analogous to them, endows them with a particular chemical interest. Thus Kournakoff (1889) obtained a series of corresponding compounds containing thiocarbamide, CSN2H4, in the place of ammonia, PtCl2,4CSN2H4, and others corresponding with Reiset's salts. Hydroxylamine, and other substances corresponding with ammonia, also give similar compounds. The common properties and composition of such compounds show their entire analogy to the cobaltia compounds (especially for ruthenium and iridium) and correspond to the fact that both the platinum metals and cobalt occur in the same, eighth, group.
Footnotes:
[1] Wells and Penfield (1888) have described a mineral sperryllite found in the Canadian gold-bearing quartz and consisting of platinum diarsenide, PtAs2. It is a noticeable fact that this mineral clearly confirms the position of platinum in the same group as iron, because it corresponds in crystalline form (regular octahedron) and chemical composition with iron pyrites, FeS2.
[1 bis] Some light is thrown upon the facility with which the platinum compounds decompose by Thomsen's data, showing that in an excess of water (+ Aq) the formation from platinum, of such a double salt as PtCl2,2KCl, is accompanied by a comparatively small evolution of heat (see Chapter XXI., Note [40]), for instance, Pt + Cl2 + 2KCl + Aq only evolves about 33,000 calories (hence the reaction, Pt + Cl2 + Aq, will evidently disengage still less, because PtCl2 + 2KCl evolves a certain amount of heat), whilst on the other hand, Fe + Cl2 + Aq gives 100,000 calories, and even the reaction with copper (for the formation of the double salt) evolves 63,000 calories.
[2] The largest amount of platinum is extracted in the Urals, about five tons annually. A certain amount of gold is extracted from the washed platinum by means of mercury, which does not dissolve the platinum metals but dissolves the gold accompanying the platinum in its ores. Moreover, the ores of platinum always contain metals of the iron series associated with them. The washed and mechanically sorted ore in the majority of cases contains about 70 to 80 p.c. of platinum, about 5 to 8 p.c. of iridium, and a somewhat smaller quantity of osmium. The other platinum metals—palladium, rhodium, and ruthenium—occur in smaller proportions than the three above named. Sometimes grains of almost pure osmium-iridium, containing only a small quantity of other metals, are found in platinum ores. This osmium-iridium may be easily separated from the other platinum metals, owing to its being nearly insoluble in aqua regia, by which the latter are easily dissolved. There are grains of platinum which are magnetic. The grains of osmium-iridium are very hard and malleable, and are therefore used for certain purposes, for instance, for the tips of gold pens.
[3] In characterising the platinum metals according to their relation to the iron metals, it is very important to add two more very remarkable points. The platinum metals are capable of forming a sort of unstable compound with hydrogen; they absorb it and only part with it when somewhat strongly heated. This faculty is especially developed in platinum and palladium, and it is very characteristic that nickel, which exactly corresponds with platinum and palladium in the periodic system, should exhibit the same faculty for retaining a considerable quantity of hydrogen (Graham's and Raoult's experiments). Another characteristic property of the platinum metals consists in their easily giving (like cobalt which forms the cobaltic salts) stable and characteristic saline compounds with ammonia, and like Fe and Co, double salts with the cyanides of the alkali metals, especially in their lower forms of combination. All the above so clearly brings the elements of the iron series in close relation to the platinum metals, that the eighth group acquires as natural a character as can be required, with a certain originality or individuality for each element.
[3 bis] Platinum was first obtained in the last century from Brazil, where it was called silver (platinus). Watson in 1750 characterised platinum as a separate independent metal. In 1803 Wollaston discovered palladium and rhodium in crude platinum, and at about the same time Tennant distinguished iridium and osmium in it. Professor Claus, of Kazan, in his researches on the platinum metals (about 1840) discovered ruthenium in them, and to him are due many important discoveries with regard to these elements, such as the indication of the remarkable analogy between the series Pd—Rh—Ru and Pt—Ir—Os.