Complex compounds of platinum(II) have played an important part in the formation and development of co-ordination theory, and this has given a strong impetus to the growth of inorganic chemistry. The existence of two isomeric forms for the PtX₂A₂ compounds (where A is amine, and X is anion) allowed Werner to conclude that platinum(II) complexes have square-planar structures (1). These isomeric forms are a result of the cis- and trans- locations of ligands A and X.

Later, the conclusion about the square-planar structure of platinum(II) complexes was repeatedly confirmed by various physical and chemical methods of investigation.

Isomeric compounds of platinum(II) differ significantly in their reactivities, and many of the chemical methods used for the identification of individual forms are based on this (2). Different rates of substitution of ligands often result in a substantial difference in the catalytic properties of two isomeric complexes. A remarkable discovery was made twenty years ago when it was found that os-[PtCl ₂(NH₃)₂] had clearly pronounced anti-cancer properties, while its trans- isomer was completely inactive (3, 4). The same picture could be observed for other diaminedihalogen complexes of platinum(II). This discovery was used as a basis for a perceived application in medicine; namely the use of co-ordination compounds for the chemical treatment of cancer.

Isomeric compounds are often different in their thermodynamic stabilities. Thus, Chemyaev and his co-workers established that the total energy of trans -[PtCl₂(NH₃)₂] is about 13 kJ/mol less than that of the cis -isomer (5). Thus, under certain conditions, it can be expected that cis -[PtCl₂(NH₃)₂] will undergo isomerisation into trans-[ PtCl₂(NH₃)₂]. The existence of isolable isomeric pairs is caused by the inertness of platinum(II) complexes. Square-planar complexes of nickel(II) are more labile than platinum(II) complexes; for this reason, cis- and trans -isomeric pairs are almost unknown for nickel(II) complexes. Palladium(II) complexes have intermediate lability. Therefore, isomeric cis- and trans -pairs of palladium(II) are fewer in number than for platinum(II), but are more numerous than for nickel(II).

Experimental evidence shows that the isomeric complexes of platinum(II) and palladium(II) interconvert, often spontaneously or in the presence of catalytic agents. Sometimes, trans -configuration complexes transform to their respective cis -configuration complexes, but more frequently this occurs the other way round, that is the cis -form being less stable converts to the trans-form. These conversions can be observed both in the solid phase and in solution. For example, the complex [PtBr₂(NH₃)₂] isomerises into trans -[PtBr₂(NH₃)₂] at 250°C (6), while with trans -[PtCl₂py(Et₂SO)] (where py is pyridine and Et is ethyl), the reverse takes place, and it isomerises into cis -[PtCl₂py(Et₂SO)] at 134°C (7).

Predicting the direction of isomerisation is very important for chemists. If such a prediction can be made, then the relative thermodynamic stabilities can be predicted simultaneously for individual isomeric complexes.

Different approaches can be used to explain the direction of isomerisation. Many facts about the isomerisation of platinum(II) and palladium(II) complexes have been considered by the author with regard to the trans -effect of the ligands (2). As a result, a substantial amount of experimental data has been explained and the direction predicted for yet unknown isomerisation processes (8, 9).

The trans -effect phenomenon has kinetic and static aspects. In a study of the kinetic and static aspects of the phenomenon, two respective series of ligands are obtained.

Since isomerisation is connected with the dynamic side of the said phenomenon, then for the most common ligands the trans -effect series can be written as follows: H₂0 < NH₃, amines < CI^– < Br^– < NO^–₂ < PR₃, R₂SO < CO < CR^–₃. It should be noted that ligands are arranged in order of increasing trans -effect. The more to the right a ligand is in the trans -effects row, the greater its lability-producing action will be.

In order to predict the direction of geometric isomerisation for platinum(II) and palladium(II) complexes, it is necessary to select a neutral ligand with the lowest trans -effect to use as an indicator. The isomerisation goes in the direction which results in the formation of a complex where that ligand will be located opposite another ligand with the smallest trans -effect. If, in the complex in question, the indicator ligand is located opposite a ligand with the next smallest trans -effect, then such a compound should not isomerise.

As was noted above, the following process takes place in the solid phase at 250°C:

In the intial compound, the ammonia molecule has the lowest trans -effect and the bromide ion has the greatest. Thus, in die initial compound, both of the ammonia molecules will be prone to isomerisation. Isomerisation takes place in the direction of formation of the trans- [PtBr₂(NH₃)₂] complex wherein the ammonia molecules are located opposite the ligand having the lowest trans -effect, (that is trans to themselves).

In contrast, trans -[PtCl₂py(Et₂SO)] isomerises into the cis -configuration compound:

In the intial complex the pyridine molecule has the lowest trans -effect, while the greatest trans -effect is exhibited by the diethyl sulphoxide molecule co-ordinated to platinum(II) by its sulphur atom. Thus, in the initial complex, pyridine is labilised by the trans diethyl sulphoxide ligand with its strong trans -effect and so isomerisation takes place. The molecule of pyridine in cis -[PtCl₂py(Et₂SO)] is less labile, and this complex happens to be more stable thermodynamically.

Two more examples from numerous experimental results will now be considered. Thus, it has been established diat the following process takes place in the solid phase (10):

Since in these complexes the methyl ligand has the greatest trans -effect, the phosphine molecule is labilised and hence isomerisation occurs. The process occurs in the given direction despite the fact that the other phosphine molecule is moving from the position trans to the chloride ligand into a trans -position relative to the phosphine ligand, that is from a ligand with relatively low trans -effect towards a ligand with a somewhat greater trans -effect. The important thing is that the phosphine molecule moves away from the ligand having the greatest trans -effect.

It has been established that the use of sulphamic acid to remove two nitro-ligands from the [Pt(NO₂)₄]^2– complex in a water medium leads to the formation of cis -[Pt(NO₂)₂(H₂O)₂], which would then isomerise into trans-[Pt(NO₂)₂(H₂O)₂] (11). Both of these processes are caused by the high trans -effect of the nitro-ligand, with isomerisation proceeding in accordance with the rule.

Isomerisation of complexes in the solid state can be clearly demonstrated using differential thermal analysis (DTA) curves. In many cases, isomerisation is accompanied by an exothermic peak, as for example, for ds-[PtBr₂(NH₃)₂]. Frequently, isomerisation shows on the DTA curve as coupled endo- and exothermic peaks, as for example, for trans -[PtCl₂py(Et₂SO)]. The endothermic peak corresponds to the melting of the initial complex, while the following exothermic peak reflects two processes: isomerisation of the complex and crystallisation of the isomerisation products from the melt. Identical coupled endo-and exothermic peaks can also be observed during isomerisation of cis -[PtMeCl(PMe₂Ph)₂] (10). Thus, the presence of an exothermic peak and, especially, of characteristic coupled endothermic and exothermic peaks (without any change in mass in both cases) can point to the isomerisation process.

Since in many cases, platinum(II) and palladium(II) complexes isomerise without any change in mass, it may be concluded that, in the absence of a solvent, isomerisation proceeds without loss of ligands, that is in an intramolecular manner. Probably, a complex of tetrahedral type is an intermediate in the isomerisation process (2).

Thus, using the rule described, it becomes possible to explain comprehensively the isomerisation processes which take place both in the solid phase and in solution.

A Russian proverb states that: “There are no rules without exceptions”, and this rule also has exceptions. In particular, the rule does not describe all of the experimental facts for the isomerisation of [PtX₂(PR₃)₂]-type phosphine-halogenide complexes. On the other hand, it is quite possible that this rule holds true not only for platinum(II) and palladium(II) complexes, but also for the square-planar complexes of nickel(II), rhodium(I), iridium(I) and gold(III), that is for complexes with a d⁸-electron configuration of the central atom. As was noted previously, the square-planar complexes of nickel(II) are rather labile, and for this reason both cis- and trans -isomeric forms are extremely unlikely to be encountered for [NiX₂L₂]-type compounds. However, if we want to see which of the two possible forms exists, then it will be the particular form in whose direction the isomerisation process should be taking place.

No doubt there are different approaches which could be used to explain the direction of isomerisation processes in square-planar complexes. Thus, Jørgensen has formulated the ligand anti-symbiosis principle which enables a judgement to be made about the relative thermodynamic stability of isomeric complexes (12). However, this principle does not describe all of the known experimental facts, either. For example, in a recent paper it was shown that [ PtCl₂LY] -type complexes (where L is pyridine and its derivatives, and Y is a ligand with a donor atom of oxygen), always have the form of a trans- isomer (13). Such a situation cannot be explained using the anti-symbiosis principles (13). It is worth mentioning that the same experimental facts can, however, be explained by the rule described in this paper.

In conclusion, we must note that in addition to the trans -effect of inner-sphere ligands, the nature of the solvent and the steric difficulties created by large ligands can be important factors which sometimes influence isomerisation. Naturally, these could complicate the predictions for the direction of the isomerisation process.