We know that platinum and palladium give compounds of lower types than iridium and rhodium, whilst ruthenium and osmium give the highest forms of oxidation; this shows itself in this case also. We have purposely cited the same compounds with 4NH3 for osmium and ruthenium as we have for platinum and palladium, and it is then seen that Ru and Os are capable of retaining 2H2O and 3H2O, besides Cl2 and NH3, which the compounds of platinum and palladium are unable to do. The same ideas which were developed in Note [35], Chapter XXII. respecting the cobaltia compounds are perfectly applicable to the present case, i.e. to the platinia compounds or ammonia compounds of the platinum metals, among which Rh and Ir give compounds which are perfectly analogous to the cobaltia compounds.
Iridium and rhodium, which easily give compounds of the type RX3, give compounds (Claus) of the type IrX3,5NH3, of a rose colour, and RhX3,5NH3, of a yellow colour. Jörgensen, in his researches on these compounds, showed their entire analogy with the cobalt compounds, as was to be expected from the periodic system.
[13] Subsequently, a whole series of such compounds was obtained with various elements in the place of the (non-reacting) chlorine, and nevertheless they, like the chlorine, reacted with difficulty, whilst the second portion of the X's introduced into such salts easily underwent reaction. This formed the most important reason for the interest which the study of the composition and structure of the platino-ammonium salts subsequently presented to many chemists, such as Reiset, Blomstrand, Peyrone, Raeffski, Gerhardt, Buckton, Clève, Thomsen, Jörgensen, Kournakoff, Verner, and others. The salts PtX4,2NH3, discovered by Gerhardt, also exhibited several different properties in the two pairs of X's. In the remaining platino-ammonium salts all the X's appear to react alike.
The quality of the X's, retainable in the platino-ammonium salts, may be considerably modified, and they may frequently be wholly or partially replaced by hydroxyl. For example, the action of ammonia on the nitrate of Gerhardt's base, Pt(NO3)4,2NH3, in a boiling solution, gradually produces a yellow crystalline precipitate which is nothing else than a basic hydrate or alkali, Pt(OH)4,2NH3. It is sparingly soluble in water, but gives directly soluble salts PtX4,2NH3 with acids. The stability of this hydroxide is such that potash does not expel ammonia from it, even on boiling, and it does not change below 130°. Similar properties are shown by the hydroxide Pt(OH)2,2NH3 and the oxide PtO,2NH3 of Reiset's second base. But the hydroxides of the compounds containing 4NH3 are particularly remarkable. The presence of ammonia renders them soluble and energetic. The brevity of this work does not permit us, however, to mention many interesting particulars in connection with this subject.
[14] Hydroxides are known corresponding with Gros's salts, which contain one hydroxyl group in the place of that chlorine or haloid which in Gros's salts reacts with difficulty, and these hydroxides do not at once show the properties of alkalis, just as the chlorine which stands in the same place does not react distinctly; but still, after the prolonged action of acids, this hydroxyl group is also replaced by acids. Thus, for example, the action of nitric acid on Pt(NO3)2Cl2,4NH3 causes the non-active chlorine to react, but in the product all the chlorine is not replaced by NO3, but only half, and the other half is replaced by the hydroxyl group: Pt(NO3)2Cl2,4NH3 + HNO3 + H2O = Pt(NO3)3(OH),4NH3 + 2HCl; and this is particularly characteristic, because here the hydroxyl group has not reacted with the acid—an evident sign of the non-alkaline character of this residue. I think it may be well to call attention to the fact that the composition of the ammonio-metallosalts very often exhibits a correspondence between the amount of X's and the amount of NH3, of such a nature that we find they contain either XNH3 or the grouping X2NH3; for example, Pt(XNH3)2 and Pt(X2NH3)2, Co(X2NH3)3, Pt(XNH3)4, &c. Judging from this, the view of the constitution of the double cyanides of platinum given in Note [11] finds some confirmation here, but, in my opinion, all questions respecting the composition (and structure) of the ammoniacal, double, complex, and crystallisation compounds stand connected with the solution of questions respecting the formation of compounds of various degrees of stability, among which a theory of solutions must be included, and therefore I think that the time has not yet come for a complete generalisation of the data which exist for these compounds; and here I again refer the reader to Prof. Kournakoff's work cited in Chapter XXII., Note [35]. However, we may add a few individual remarks concerning the platinia compounds.
To the common properties of the platino-ammonium salts, we must add not only their stability (feeble acids and alkalis do not decompose them, the ammonia is not evolved by heating, &c.), but also the fact that the ordinary reactions of platinum are concealed in them to as great an extent as those of iron in the ferricyanides. Thus neither alkalis nor hydrogen sulphide will separate the platinum from them. For example, sulphuretted hydrogen in acting on Gros's salts gives sulphur, removes half the chlorine by means of its hydrogen, and forms salts of Reiset's first base. This may be understood or explained by considering the platinum in the molecule as covered, walled up by the ammonia, or situated in the centre of the molecule, and therefore inaccessible to reagents. On this assumption, however, we should expect to find clearly-expressed ammoniacal properties, and this is not the case. Thus ammonia is easily decomposed by chlorine, whilst in acting on the platino-ammonium salts containing PtX2 and 2NH3 or 4NH3, chlorine combines and does not destroy the ammonia; it converts Reiset's salts into those of Gros and Gerhardt. Thus from PtX2,2NH3 there is formed PtX2Cl2,2NH3, and from PtX2,4NH3 the salt of Gros's base PtX2Cl2,4NH3. This shows that the amount of chlorine which combines is not dependent on the amount of ammonia present, but is due to the basic properties of platinum. Owing to this some chemists suppose the ammonia to be inactive or passive in certain compounds. It appears to me that these relations, these modifications, in the usual properties of ammonia and platinum are explained directly by their mutual combination. Sulphur, in sulphurous anhydride, SO2, and hydrogen sulphide, SH2, is naturally one and the same, but if we only knew of it in the form of hydrogen sulphide, then, having obtained it in the form of sulphurous anhydride, we should consider its properties as hidden. The oxygen in magnesia, MgO, and in nitric peroxide, NO2, is so different that there is no resemblance. Arsenic no longer reacts in its compounds with hydrogen as it reacts in its compounds with chlorine, and in their compounds with nitrogen all metals modify both their reactions and their physical properties. We are accustomed to judge the metals by their saline compounds with haloid groups, and ammonia by its compounds with acid substances, and here, in the platino-compounds, if we assume the platinum to be bound to the entire mass of the ammonia—to its hydrogen and nitrogen—we shall understand that both the platinum and ammonia modify their characters. Far more complicated is the question why a portion of the chlorine (and other haloid simple and complex groups) in Gros's salts acts in a different manner from the other portion, and why only half of it acts in the usual way. But this also is not an exclusive case. The chlorine in potassium chlorate or in carbon tetrachloride does not react with the same ease with metals as the chlorine in the salts corresponding with hydrochloric acid. In this case it is united to oxygen and carbon, whilst in the platino-ammonium compounds it is united partly to platinum and partly to the platino-ammonium group. Many chemists, moreover, suppose that a part of the chlorine is united directly to the platinum and the other part to the nitrogen of the ammonia, and thus explain the difference of the reactions; but chlorine united to platinum reacts as well with a silver salt as the chlorine of ammonium chloride, NH4Cl, or nitrosyl chloride, NOCl, although there is no doubt that in this case there is a union between the chlorine and nitrogen. Hence it is necessary to explain the absence of a facile reactive capacity in a portion of the chlorine by the conjoint influence of the platinum and ammonia on it, whilst the other portion may be admitted as being under the influence of the platinum only, and therefore as reacting as in other salts. By admitting a certain kind of stable union in the platino-ammonium grouping, it is possible to imagine that the chlorine does not react with its customary facility, because access to a portion of the atoms of chlorine in this complex grouping is difficult, and the chlorine union is not the same as we usually meet in the saline compounds of chlorine. These are the grounds on which we, in refuting the now accepted explanations of the reactions and formation of the platino-compounds, pronounce the following opinion as to their structure.
In characterising the platino-ammonium compounds, it is necessary to bear in mind that compounds which already contain PtX4 do not combine directly with NH3, and that such compounds as PtX4,4NH3 only proceed from PtX2, and therefore it is natural to conclude that those affinities and forces which cause PtX2 to combine with X2 also cause it to combine with 2NH3. And having the compound PtX2,2NH3, and supposing that in subsequently combining with Cl2 it reacts with those affinities which produce the compounds of platinic chloride, PtCl4, with water, potassium chloride, potassium cyanide, hydrochloric acid, and the like, we explain not only the fact of combination, but also many of the reactions occurring in the transition of one kind of platino-ammonium salts into another. Thus by this means we explain the fact that (1) PtX2,2NH3 combines with 2NH3, forming salts of Reiset's first base; (2) and the fact that this compound (represented as follows for distinctness), PtX2,2NH3,2NH3, when heated, or even when boiled in solution, again passes into PtX2,2NH3 (which resembles the easy disengagement of water of crystallisation, &c.); (3) the fact that PtX2,2NH3 is capable of absorbing, under the action of the same forces, a molecule of chlorine, PtX2,2NH3,Cl2, which it then retains with energy, because it is attracted, not only by the platinum, but also by the hydrogen of the ammonia; (4) the fact that this chlorine held in this compound (of Gerhardt) will have a position unusual in salts, which will explain a certain (although very feebly-marked) difficulty of reaction; (5) the fact that this does not exhaust the faculty of platinum for further combination (we need only recall the compound PtCl4,2HCl,16H2O), and that therefore both PtX2,2NH3,Cl2 and PtX2,2NH3,2NH3 are still capable of combination, whence the latter, with chlorine, gives PtX2,2NH3,2NH3,Cl2, after the type of PtX4Y4 (and perhaps higher); (6) the fact that Gros's compounds thus formed are readily reconverted into the salts of Reiset's first base when acted on by reducing agents; (7) the fact that in Gros's salts, PtX2,2NH3(NH3X)2, the newly-attached chlorine or haloid will react with difficulty with salts of silver, &c., because it is attached both to the platinum and to the ammonia, for both of which it has an attraction; (8) the fact that the faculty for further combination is not even yet exhausted in the type of Gros's salts, and that we actually have a compound of Gros's chlorine salt with platinous chloride and with platinic chloride; the salt PtSO4,2NH3,2NH3,SO4 combines further also with H2O; (9) the fact that such a faculty of combination with new molecules is naturally more developed in the lower forms of combination than in the higher. Hence the salts of Reiset's first base—for example, PtCl2,2NH3,2NH3—both combine with water and give precipitates (soluble in water but not in hydrochloric acid) of double salts with many salts of the heavy metals—for example, with lead chloride, cupric chloride, and also with platinic and platinous chlorides (Buckton's salts). The latter compounds will have the composition PtCl2,2NH3,2NH3,PtCl2—that is, the same composition as the salts of Reiset's second base, but it cannot be identical with it. Such an interesting case does actually exist. The first salt, PtCl2,4NH3,PtCl2, is green, insoluble in water and in hydrochloric acid, and is known as Magnus's salt, and the second, PtCl2,2NH3, is Reiset's yellow, sparingly soluble (in water). They are polymeric, namely, the first contains twice the number of elements held in the second, and at the same time they easily pass into each other. If ammonia be added to a hot hydrochloric acid solution of platinous chloride, it forms the salt PtCl2,4NH3, but in the presence of an excess of platinous chloride it gives Magnus's salt. On boiling the latter in ammonia it gives a colourless soluble salt of Reiset's first base, PtCl2,4NH3, and if this be boiled with water, ammonia is disengaged, and a salt of Reiset's second base, PtCl2,2NH3, is obtained.
A class of platino-ammonium isomerides (obtained by Millon and Thomsen) are also known. Buckton's salts—for example, the copper salt—were obtained by them from the salts of Reiset's first base, PtCl2,4NH3, by treatment with a solution of cupric chloride, &c., and therefore, according to our method of expression, Buckton's copper salt will be PtCl2,4NH3,CuCl2. This salt is soluble in water, but not in hydrochloric acid. In it the ammonia must be considered as united to the platinum. But if cupric chloride be dissolved in ammonia, and a solution of platinous chloride in ammonium chloride is added to it, a violet precipitate is obtained of the same composition as Buckton's salt, which, however, is insoluble in water, but soluble in hydrochloric acid. In this a portion, if not all, of the ammonia must be regarded as united to the copper, and it must therefore be represented as CuCl2,4NH3,PtCl2. This form is identical in composition but different in properties (is isomeric) with the preceding salt (Buckton's). The salt of Magnus is intermediate between them, PtCl2,4NH3,PtCl2; it is insoluble in water and hydrochloric acid. These and certain other instances of isomeric compounds in the series of the platino-ammonium salts throw a light on the nature of the compounds in question, just as the study of the isomerides of the carbon compounds has served and still serves as the chief cause of the rapid progress of organic chemistry. In conclusion, we may add that (according to the law of substitution) we must necessarily expect all kinds of intermediate compounds between the platino and analogous ammonia derivatives on the one hand, and the complex compounds of nitrous acid on the other. Perhaps the instance of the reaction of ammonia upon osmic anhydride, OsO4, observed by Fritsche, Frémy, and others, and more fully studied by Joly (1891), belongs to this class. The latter showed that when ammonia acts upon an alkaline solution of OsO4 the reaction proceeds according to the equation: OsO4 + KHO + NH3 = OsNKO3 + 2H2O. It might be imagined that in this case the ammonia is oxidised, probably forming the residue of nitrous acid (NO), while the type OsO4 is deoxidised into OsO2, and a salt, OsO(NO)(KO), of the type OsX4 is formed. This salt crystallises well in light yellow octahedra. It corresponds to osmiamic acid, OsO(ON)(HO), whose anhydride [OsO(NO)]2, has the composition Os2N2O5, which equals 2Os + N2O5 to the same extent as the above-mentioned compound PtCO2 equals Pt + CO2 (see Note [11]).