Lezhan Chen

Cone angles: Tolman's and Plato's

Abstract The successful application of Tolman cone angles for P-donor and other ligands in accounting quantitatively for steric effects in a wide variety of physicochemical processes is contrasted with the variability of cone angles obtained from crystallographic studies. It is maintained that the latter are not relevant in describing steric effects for reactions in solution. Problems with cone angles for ligands with conformational uncertainties are best dealt with by systematic measurement of deviations of data for those ligands from trends defined by ligands with less ambiguous cone angles. In fact a body of cone angles for all ligands could be obtained by adjusting cone angles to give perfect fits to individual steric profiles and then averaging the values obtained from a large number of such studies. In this way a set of cone angles could be obtained that are divorced from their origins in Tolmans models and justified solely by their successful quantitative application. Although it is tempting to relate these cone angles to some Platonically perfect set, consideration given here to even a small number of sets of data of this sort suggests that there may always be some ligands with uncertain cone angles. Thus, a combination of eight sets of data that include P(OMe) 3 as a ligand suggests that it should have its Tolman cone angle increased by ca. 10°. A much larger proposed increase is disputed. Another set of data suggests that its Tolman cone angle is appropriate when it is a nucleophile but that P(O i -Pr) 3 as a nucleophile should have its Tolman cone angle reduced by ca. 5°. Thus the deviant behaviour of conformationally ambiguous ligands may be systematic, and could depend on whether the P-donors are acting as ligands or nucleophiles. However, the concept of a Platonically perfect set of cone angles is probably justified for many ligands that are conformationally unambiguous and these ideal cone angles may be essentially identical with Tolmans values. In the course of these analyses a new set of π-acidity parameters was developed and values are tabulated.

Associative reactions of metal carbonyl clusters: systematic kinetic studies of some ruthenium and other clusters

Abstract Rate constants for associative reactions of metal carbonyl clusters with a range of P-donor nucleophiles of various basicities and sizes can always be fitted successfully to the equation log k2=∞+ β (pK′a+4)+γ(σ-σth))λ. The evolution of this equation, and the use of the electronic parameter pK′a and steric parameter σ assigned to particular P-donor nucleophiles, is outlined. λ is a switching function which is zero when B 5 σth (the steric threshold) and becomes unity when σ > σth. The data can be fitted to the equation by a least-squares programme developed initially by Professor W.P. Giering. The significance of the parameters a (defined as the standard reactivity of the carbonyl), β (the susceptibility of the carbonyl to the δ basicity of the nucleophiles), θth (the Tolman cone angle of the nucleophile where steric effects begin to be apparent, i.e. the steric threshold) and y (the sensitivity of the rates to the cone angles of the nucleophiles when θ > θth) are discussed in detail and their interrelationships considered. In particular the sharpness of the steric threshold is taken to imply that a substrate carbonyl undergoes an isomerization reaction when approached by the nucleophile. This produces a so-called transition state isomer (TSI) that contains a roughly conical opening into which all nucleophiles with θ ⩽ θth can fit without steric difficulty. When θ > θth the nucleophiles can attain the same metal-P-donor bond distance in the transition state if the TSI opens up further, so that y is a measure of the flexibility of the TSI. Alternatively the flexibility of the TSI may be so low that it cannot accommodate large nucleophiles unless they penetrate less deeply into the space in the TSI. The application of this systematic kinetic approach to the delineation of the dynamic character of metal carbonyl clusters is illustrated by examples, many of them unpublished, that involve Ru3(CO)11L clusters (where the effect of varying the P-donor substituent is probed), some Rh4 and Ir4 clusters, some high nuclearity Ru and Fe clusters, and a variety of osmium clusters. Not only are the clusters characterizable in this way but also the actual mechanistic paths followed, and the products formed, can undergo major changes which depend mainly on the size of the nucleophiles.

Use of QALE to generate potential energy surfaces in transition states: associative reactions of Ru3(CO)12 with P-donor nucleophiles

Previous studies of associative reactions of Ru3(CO)12 have been greatly extended to include a total of 23 P-donor nucleophiles with widely differing electronic and steric properties. Application of standard QALE methodology (QALE = quantitative analysis of ligand effects) enabled the rate constants to be analyzed according to the electronic and steric properties of the nucleophiles. It was unexpectedly found necessary to include what has become known as the aryl effect in this analysis, together with a positive contribution to the rates due to the π-acidity of phosphite nucleophiles. Also the analysis was found to be quite sensitive to the cone angle taken for P(n-Bu)3 and the value 136°, favoured by Giering and Prock, was found to be superior to Tolmans value of 132°. The individual contributions of the various effects to the rate constants for each ligand are presented as absolute contributions to ΔG‡ or fractional contributions to the rate constants. σ-Basicity and steric effects are represented graphically by a three-dimensional ‘logk2 surface’ to which the aryl and π-acidity effects can be added and appear as peaks. This logk2 surface can be converted into a free energy or potential energy surface by a simple scale change for the y-axis. However, a free energy surface would be difficult to represent, unless minus ΔG‡ was plotted on the y-axis, because ‘negative spikes’ would represent a graphical problem.

Substitution kinetics of tetracarbonyl(η2-alkene)ruthenium complexes: alkene = ethene and methyl acrylate

Abstract The kinetics in heptane of displacement of the alkene ligands ethene and methyl acrylate from Ru(CO)4(η2-alkene) by P(OEt)3 have been measured. The reactions occur by reversible dissociation of the alkenes, and activation parameters are compared with those for dissociation of CO from Ru(CO)5 and for reactions of the corresponding Os complexes. A linear free energy relationship for ligand dissociation from Ru(CO)5, Ru(CO)4(C2H4) and Ru(CO)4(MA) has a gradient close to unity, indicating virtually complete bond breaking in the transition states. Competition parameters for reactions of what is probably a solvated Ru(CO)4S intermediate have been measured for the alkenes and P(OEt)3, and for eleven other P-donor nucleophiles. Correlations with the electronic and steric properties of the P-donors show negligible dependence on the electron donicity of the nucleophiles and a small but significant dependence on their sizes. The sizes were quantified by Tolman cone angles or by ‘cone angle equivalents’ derived directly from Browns ligand repulsion energies (Er). These correlations compared with those, reported elsewhere, for reactions of the probably solvated intermediates Co2(CO)5(μ2-C2Ph2) and H3Re3(CO)11 formed by ligand dissociative processes. In all cases the discrimination between nucleophiles by the intermediates is weak confirming their high reactivity and the borderline nature of the mechanisms of these bimolecular reactions between Id and Ia.

QALE analysis of CO dissociative kinetics of Ru(CO)4L (L = P-donor ligands): accelerating effects of hydrogen in PHnR3 − n ligands (n = 1–2)

Studies of CO-dissociative substitution reactions of the complexes Ru(CO)4L (L = a wide variety of P-donor ligands) have been extended and analysis of the results by the QALE methodology has been refined (QALE = quantitative analysis of ligand effects). Rates increase substantially with increasing size of L, mainly as a consequence of increasingly favourable activation entropies. These can be associated with increasing Ru-CO bond breaking that is compensated enthalpically by increasing Ru-P bond making allowed by release of steric strain. Explicit allowance for pi-acidity shows that these effects are just significant while sigma-donor and aryl effects are negligible. However, pendent hydrogen atoms, attached directly to the phosphorus atoms, have a pronounced and unique positive effect on the rates, with significant kinetic isotope effects (KIE). This is associated with the novel occurrence of direct Ru-H or incipient Ru-(eta2-P-H) agostic bond making as the CO ligand departs.

Systematic Kinetic Studies of Associative and Dissociative Reactions of Substituted Metal Carbonyl Clusters: The Intimate Mechanisms

Systematic kinetic studies1 of associative reactions2 of metal carbonyl clusters can provide reactivity profiles that are as usefully characteristic of the clusters as their crystal structures. A given cluster can be quantitatively characterized by its sensitivity to the electronic and steric nature of a series of P-donor nucleophiles,2 by is standard or “intrinsic” reactivity (I.R.) towards nucleophilic attack,2 and by the distribution between substitution and fragmentation products formed as a consequence of nucleophilic attack.3 Thus the electronic and steric profiles for associative reactions of Ru3(CO)11L show that bond-making in the transition states decreases along the series L = P(OEt)3 > CO > P-n-Bu3 and this correlates with increasing intrinsic reactivities of the clusters. This must result from a balance between opposing steric and electronic effects but the exact way in which these operate is not yet clear.