Ruthenium and osmium, obtained by the ignition or reduction of their compounds in the form of powder, have a density considerably less than in the fused form, and differ in this condition in their capacity for reaction; they are much more difficultly fused than platinum and iridium, although ruthenium is more fusible than osmium. Ruthenium in powder has a specific gravity of 8·5, the fused metal of 12·2; osmium in powder has a specific gravity of 20·0, and when semi-fused—or, more strictly speaking, agglomerated—in the oxyhydrogen flame, of 21·4, and fused 22·5. The powder of slightly-heated osmium oxidises very easily in the air, and when ignited burns like tinder, directly forming the odoriferous osmic anhydride (hence its name, from the Greek word signifying odour); ruthenium also oxidises when heated in air, but with more difficulty, forming the oxide RuO2. The oxides of the types RO, R2O3, and RO2 (and their hydrates) obtained by reduction from the higher oxides, and also from the chlorides, are analogous to those given by the other platinum metals, in which respect osmium and ruthenium closely resemble them. We may also remark that ruthenium has been found in the platinum deposits of Borneo in the form of laurite, Ru2S3, in grey octahedra of sp. gr. 7·0.
For osmium, Moraht and Wischin (1893) obtained free osmic acid, H2OsO4, by decomposing K2OsO4 with water, and precipitating with alcohol in a current of hydrogen (because in air volatile OsO4 is formed); with H2S, osmic acid gives OsO3(HS)2 at the ordinary temperature.
Debray and Joly showed that ruthenic anhydride, RuO4, fuses at 25°, boils at 100°, and evolves oxygen when dissolved in potash, forming the salt KRuO4 (not isomorphous with potassium permanganate).
Joly (1891), who studied the ruthenium compounds in greater detail, showed that the easily-formed KRuO4 gives RuKO4RuO3 when ignited, but it resembles KMnO4 in many respects. In general, Ru has much in common with Mn. Joly (1889) also showed that if KNO3 be added to a solution of RuCl3 containing HCl, the solution becomes hot, and a salt, RuCl3NO2KCl, is formed, which enters into double decomposition and is very stable. Moreover, if RuCl3 be treated with an excess of nitric acid, it forms a salt, RuCl3NOH2O, after being heated (to boiling) and the addition of HCl. The vapour density of RuO4, determined by Debray and Joly, corresponds to that formula.
[10] Although palladium gives the same types of combination (with chlorine) as platinum, its reduction to RX2 is incomparably easier than that of platinic chloride, and in the case of iridium it is also very easy. Iridic chloride, IrCl4, acts as an oxidising agent, readily parts with a fourth of its chlorine to a number of substances, readily evolves chlorine when heated, and it is only at low temperatures that chlorine and aqua regia convert iridium into iridic chloride. In disengaging chlorine iridium more often and easily gives the very stable iridious chloride, IrCl3 (perhaps this substance is Ir2Cl6 = IrCl2,IrCl4, insoluble in water, but soluble in potassium chloride, because it forms the double salt K3IrCl6), than the dichloride, IrCl_2. This compound, corresponding to IrX2, is very stable, and corresponds with the basic oxide, Ir2O3, resembling the oxides Fe2O3, Co2O3. To this form there correspond ammoniacal compounds similar to those given by cobaltic oxide. Although iridium also gives an acid in the form of the salt K2Ir2O7, it does not, like iron (and chromium), form the corresponding chloride, IrCl6. In general, in this as in the other elements, it is impossible to predict the chlorine compounds from those of oxygen. Just as there is no chloride SCl6, but only SCl2, so also, although IrO3 exists, IrCl6 is wanting, the only chloride being IrCl4, and this is unstable, like SCl2, and easily parts with its chlorine. In this respect rhodium is very much like iridium (as platinum is like palladium). For RhCl4 decomposes with extreme ease, whilst rhodium chloride, RhCl3, is very stable, like many of the salts of the type RhX3, although like the platinum elements these salts are easily reduced to metal by the action of heat and powerful reagents. There is as close a resemblance between osmium and ruthenium. Osmium when submitted to the action of dry chlorine gives osmic chloride, OsCl4, but the latter is converted by water (as is osmium by moist chlorine) into osmic anhydride, although the greater portion is then decomposed into Os(HO)4 and 4HCl, like a chloranhydride of an acid. In general this acid character is more developed in osmium than in platinum and iridium. Having parted with chlorine, osmic chloride, OsCl4, gives the unstable trichloride, OsCl3, and the stable soluble dichloride, OsCl2, which corresponds with platinous chloride in its properties and reactions. The relation of ruthenium to the halogens is of the same nature.
[11] This acid character is explained by the influence of the platinum on the hydrogen, and by the attachment of the cyanogen groups. Thus cyanuric acid, H3(CN)3O3, is an energetic acid compared with cyanic acid, HCNO. And the formation of a compound with five molecules of water of crystallisation, (PtH2(CN)4,5H2O), confirms the opinion that platinum is able to form compounds of still higher types than that expressed in its saline compounds, and, moreover, the combination of hydroplatinocyanic acid with water does not reach the limit of the compounds which appears in PtCl4,2HCl,6H2O.
A whole series of platinocyanides of the common type PtR2(CN)4nH2O is obtained by means of double decomposition with the potassium or hydrogen or silver salts. For example, the salts of sodium and lithium contain, like the potassium salt, three molecules of water. The sodium salt is soluble in water and alcohol. The ammonium salt has the composition Pt(NH4)2(CN)4,2H2O and gives crystals which reflect blue and rose-coloured light. This ammonium salt decomposes at 300°, with evolution of water and ammonium cyanide, leaving a greenish platinum dicyanide, Pt(CN)2, which is insoluble in water and acid but dissolves in potassium cyanide, hydrocyanic acid, and other cyanides. The same platinous cyanide is obtained by the action of sulphuric acid on the potassium salts in the form of a reddish-brown amorphous precipitate. The most characteristic of the platinocyanides are those of the alkaline earths. The magnesium salt PtMg(CN)4,7H2O crystallises in regular prisms, whose side faces are of a metallic green colour and terminal planes dark blue. It shows a carmine-red colour along the main axis, and dark red along the lateral axes; it easily loses water, (2H2O), at 40°, and then turns blue (it then contains 5H2O, which is frequently the case with the platinocyanides). Its aqueous solution is colourless, and an alcoholic solution deposits yellow crystals. The remainder of the water is given off at 230°. It is obtained by saturating platinocyanic acid with magnesia, or else by double decomposition between the barium salt and magnesium sulphate. The strontium salt, SrPt(CN)4,4H2O crystallises in milk-white plates having a violet and green iridescence. When it effloresces in a desiccator, its surfaces have a violet and metallic green iridescence. A colourless solution of the barium salt PtBa(CN)4,4H2O is obtained by saturating a solution of hydroplatinocyanic acid with baryta, or by boiling the insoluble copper platinocyanide in baryta water. It crystallises in monoclinic prisms of a yellow colour, with blue and green iridescence; it loses half its water at 100°, and the whole at 150°. The ethyl salt, Pt(C2H5)2(CN)4,2H20, is also very characteristic; its crystals are isomorphous with those of the potassium salt, and are obtained by passing hydrochloric acid into an alcoholic solution of hydroplatinocyanic acid. The facility with which they crystallise, the regularity of their forms, and their remarkable play of colours, renders the preparation of the platinocyanides one of the most attractive lessons of the laboratory.
By the action of chlorine or dilute nitric acid, the platinocyanides are converted into salts of the composition PtM2(CN)5, which corresponds with Pt(CN)3,2KCN—that is, they express the type of a non-existent form of oxidation of platinum, PtX3 (i.e. oxide Pt2O3), just as potassium ferricyanide (FeCy3,3KCy) corresponds with ferric oxide, and the ferrocyanide corresponds with the ferrous oxide. The potassium salt of this series contains PtK2(CN)5,3H2O, and forms brown regular prisms with a metallic lustre, and is soluble in water but insoluble in alcohol. Alkalis re-convert this compound into the ordinary platinocyanide K2Pt(CN)4, taking up the excess of cyanogen. It is remarkable that the salts of the type PtM2Cy5 contain the same amount of water of crystallisation as those of the type PtM2Cy4. Thus the salts of potassium and lithium contain three, and the salt of magnesium seven, molecules of water, like the corresponding salts of the type of platinous oxide. Moreover, neither platinum nor any of its associates gives any cyanogen compound corresponding with the oxide, i.e. having the composition PtK2Cy6, just as there are no compounds higher than those which correspond to RCy3nMCy3 for cobalt or iron. This would appear to indicate the absence of any such cyanides, and indeed, for no element are there yet known any poly-cyanides containing more than three equivalents of cyanogen for one equivalent of the element. The phenomenon is perhaps connected with the faculty of cyanogen of giving tricyanogen polymerides, such as cyanuric acid, solid cyanogen chloride, &c. Under the action of an excess of chlorine, a solution of PtK2(CN)4 gives (besides PtK2Cy5) a product PtK2Cy4Cl2, which evidently contains the form PtX4, but at first the action of the chlorine (or the electrolysis of, or addition of dilute peroxide of hydrogen to, a solution of PtK2Cy4, acidulated with hydrochloric acid) produces an easily soluble intermediate salt which crystallises in thin copper-red needles (Wilm, Hadow, 1889). It only contains a small amount of chlorine, and apparently corresponds to a compound 5PtK2Cy4 + PtK2Cy4Cl2 + 24H2O. Under the action of an excess of ammonia both these chlorine products are converted either completely or in part (according to Wilm ammonia does not act upon PtK2Cy4) into PtCy2,2NH3, i.e. a platino-ammonia compound (see further on). It is also necessary to pay attention to the fact that ruthenium and osmium—which, as we know, give higher forms of oxidation than platinum—are also able to combine with a larger proportion of potassium cyanide (but not of cyanogen) than platinum. Thus ruthenium forms a crystalline hydroruthenocyanic acid, RuH4(CN)6, which is soluble in water and alcohol, and corresponds with the salts M4Ru(CN)6. There are exactly similar osmic compounds—for example, K4Os(CN)6,3H2O. The latter is obtained in the form of colourless, sparingly-soluble regular tablets on evaporating the solution obtained from a fused mixture of potassium osmiochloride, K2OsCl6, and potassium cyanide. These osmic and ruthenic compounds fully correspond with potassium ferrocyanide, K4Fe(CN)6,3H2O, not only in their composition but also in their crystalline form and reactions, which again demonstrates the close analogy between iron, ruthenium, and osmium, which we have shown by giving these three elements a similar position (in the eighth group) in the periodic system. For rhodium and iridium only salts of the same type as the ferricyanides, M3RCy6, are known, and for palladium only of the type M2PdCy4, which are analogous to the platinum salts. In all these examples a constancy of the types of the double cyanides is apparent. In the eighth group we have iron, cobalt, nickel, copper, and their analogues ruthenium, rhodium, palladium, silver, and also osmium, iridium, platinum, gold. The double cyanides of iron, ruthenium, osmium have the type K4R(CN)6; of cobalt, rhodium, iridium, the type K3R(CN)6; of nickel, palladium, platinum the type K2R(CN)4 and K2R(CN)5; and for copper, silver, gold there are known KR(CN)2, so that the presence of 4, 3, 2, and 1 atoms of potassium corresponds with the order of the elements in the periodic system. Those types which we have seen in the ferrocyanides and ferricyanides of iron repeat themselves in all the platinoid metals, and this naturally leads to the conclusion that the formation of similar so-called double salts is of exactly the same nature as that of the ordinary salts. If, in expressing the union of the elements in the oxygen salts, the existence of an aqueous residue (hydroxyl group) be admitted, in which the hydrogen is replaced by a metal, we have then only to apply this mode of expression to the double salts and the analogy will be obvious, if only we remember that Cl2, (CN)2, SO4, &c., are equivalent to O, as we see in RO, RCl2, RSO4, &c. They all = X2, and, therefore, in point of fact, wherever X (= Cl or OH, &c.) can be placed, there (Cl2H), (SO4H), &c., can also stand. And as Cl2H = Cl + HCl and SO4H = OH + SO3, &c., it follows that molecules HCl or SO3, or, in general, whole molecules—for instance, NH3, H2O, salts, &c., can annex themselves to a compound containing X. (This is an indirect consequence of the law of substitution which explains the origin of double salts, ammonia compounds, compounds with water of crystallisation, &c., by one general method.) Thus the double salt MgSO4,K2SO4, according to this reasoning, may be considered as a substance of the same type as MgCl2, namely, as = Mg(SO4K)2, and the alums as derived from Al(OH)(SO4), namely, as Al(SO4K)(SO4). Without stopping to pursue this digression further, we will apply these considerations to the type of the ferrocyanides and ferricyanides and their platinum analogues. Such a salt as K2PtCy4 may accordingly be regarded as Pt(Cy2K)2, like Pt(OH)2; and such a salt as PtK2Cy5 as PtCy(Cy2K)2, the analogue of PtX(OH)2, or AlX(OH)2, and other compounds of the type RX3. Potassium ferricyanide and the analogous compounds of cobalt, iridium, and rhodium, belong to the same type, with the same difference as there is between RX(OH)2 and R(OH)3, since FeK3Cy6 = Fe(Cy2K)3. Limiting myself to these considerations, which may partially elucidate the nature of double salts, I will now pass again to the complex saline compounds known for platinum.
(A) On mixing a solution of potassium thiocyanate with a solution of potassium platinosochloride, K2PtCl4, they form a double thiocyanate, PtK2(CNS)4, which is easily soluble in water and alcohol, crystallises in red prisms, and gives an orange-coloured solution, which precipitates salts of the heavy metals. The action of sulphuric acid on the lead salt of the same type gives the acid itself, PtH2(SCN)4, which corresponds with these salts. The type of these compounds is evidently the same as that of the cyanides.
(B) Platinous chloride, PtCl2, which is insoluble in water, forms double salts with the metallic chlorides. These double chlorides are soluble in water, and capable of crystallising. Hence when a hydrochloric acid solution of platinous chloride is mixed with solutions of metallic salts and evaporated it forms crystalline salts of a red or yellow colour. Thus, for example, the potassium salt, PtK2Cl4, is red, and easily soluble in water; the sodium salt is also soluble in alcohol; the barium salt, PtBaCl4,3H2O, is soluble in water, but the silver salt, PtAg2Cl4, is insoluble in water, and may be used for obtaining the remaining salts by means of double decomposition with their chlorides.