SPECTROSCOPIC AND ELECTROCHEMICAL PROBES OF ELECTRONIC COUPLING IN SOME CYANIDE-BRIDGED TRANSITION-METAL DONOR ACCEPTOR COMPLEXES/

The effects of donor-acceptor (D/A) electronic coupling, H-DA, on the spectroscopic and electrochemical properties of several series of CN-- bridged transition metal complexes have been examined. The complexes e mployed were formed by ruthenation of M(L)(CN)(2)(n+) parent complexes (for n = 0, M = Ru(II) or Fe(II), and L = bpy or phen; for n = 1, M = Cr(III), Rh(III), or Co(III), and L = bpy, phen, or a tetraazamacrocy clic ligand). The observed half-wave potentials of the resulting CN--b ridged D/A complexes spanned a 300-350 mV range in contrast to the ran ge of about 80 mV expected on the basis of the oscillator strength, h( DA), of the D/A charge-transfer MM'CT absorption band and the geometri cal distance between donor and acceptor, r(DA). Different series of co mplexes exhibit different correlations between E-1/2 and h(DA). Severa l factors have been found to contribute to these differences: (a) symm etry effects; (b) solvational differences that arise when nonbridging ligands are changed; (c) solvational effects arising from differences in overall electrical charges; (d) partial delocalization of electron density along the D/A axis in such a way as to reduce the effective di stance between centers of charge, r(ge)(c). To take account of the eff ects of the solvational factors, systematic examination has been made of (a) the metal independent shifts of E-1/2 which occur when nonbridg ing ligands are changed; (b) the differences in E-/12 that occur in cl osely related Ru(III)/Ru(II) couples which differ in charge; and (c) s olvent peturbations of E-1/2(Ru(NH3)(5)(3+,2+)) and solvatochromic shi fts of the central metal-to-ligand charge transfer (MLCT) and MM'CT ab sorbancies of (bpy)(2)(CN)Ru(CNRu(NH3)(5))(3+) and (bpy)(2)Ru(CNRu(NH3 )(5))(2)(6+). The experimental observations indicate that changes in t he nonbridging ligand of the central metal can result in a range of ab out 90 mV variation in E-1/2(Ru(NH3)(5)(3+,2+)), the effect of a one u nit increase in charge of the central metal is to increase E-1/2 by ap proximately 65 +/- 15 mV, solvent perturbations of E-1/2 and the elect ron-transfer reorganizational energy, lambda(r), are approximately equ al in magnitude, solvational corrections can be treated linearly, and the solvational contributions to E-1/2 that arise from charge delocali zation are less than about 10 mV in these complexes. The complexes hav e a very rich charge-transfer spectroscopy, and in some complexes as m any as seven different CT transitions can be identified which depend o n the oxidation state of the Ru(NH3)(5) moiety. There is evidence for considerable mixing between these transitions. The mixed valence (Ru(N H3)(5)(2+)/Ru(NH3)(5)(3+)), bisruthenates exhibit a unique Ru(NH3)(5)/ M MM'CT component in addition to the expected Ru(NH3)(5)(2+) --> Ru(NH 3)(5)(3+) CT; this relatively weak absorption tracks the dominant Ru(N H3)(5)/central metal MM'CT absorption, and it is attributable to the d ifferent effects of local M-c(CN-)Ru(NH3)(5) electronic coupling in th e mixed valence complex. Values of E-1/2(obsd), corrected for solvatio nal effects implied by the experimental observations, correlate with h (DA), corrected for symmetry effects, E-1/2(corr) = E-1/2(ref) +/- (4. 2 x 10(-4)) h(DA)/r(DA), only if the ''solvational correction'' for Fe (II)- and Ru(II)-centered complexes is about 70% larger than suggested by the experimental observations. This may imply greater charge deloc alization onto (or from) the bridging ligand for these two metal cente rs. For either interpretation, the correlation between E-1/2(obsd) and h(DA) implies that r(ge)(c) less than or equal to 0.62r(DA). This rel atively small value of r(ge)(c) can be interpreted in terms of charge delocalization onto (or from) the bridging ligand, and it can be quali tatively described in terms of perturbational mixing of the ground and excited electron-transfer states with higher energy CT states. This m ixing is described in terms of a multicenter (M-c-C-N-Ru-t) vibronic c oupling model which was previously (Inorg. Chem. 1996, 34, 3463) used to account for the anamolous shifts of the CN- stretch in CN--bridged D/A complexes.