THE ARYL EFFECT IN DISUBSTITUTED DIMANGANESE AND DICOBALT CARBONYLS

The energies of the intense bands in the near-UV-vis spectra of a numb er of axially substituted Mn-2(CO)(8)L-2 and Co-2(CO)(6)L-2 complexes containing non-pi-acid phosphine substituents have been measured in so lution. When combined with data reported previously, the trends show u nambiguously that a major contributor to the energies is Giering and F rock's aryl effect (quantifiable by E-ar values that are almost propor tional to the number of aryl groups attached to the donor P atom). The equation hv = alpha + beta chi + gamma theta + omega E-ar gives an ex cellent empirical description of the data (100(1 - R-2) = 0.5% and 2.6 % for the Mn-2 and Co-2 complexes, respectively) where chi is the Tolm an/Bartik measure of the sigma-basicity and theta is the Tolman cone a ngle of the ligands. The values of hv increase with increasing sigma-b asicity and decrease with increasing ligand size and increasing number s of aryl groups. The maximum increase of 1650 cm(-1) in the hv values of the Mn-2 complexes is found when L = P(p-F3CC6H4)(3) is replaced b y P(n-Bu)(3), and this is made up of 49 +/- 4% contribution by the les s unfavorable aryl effect, 33 +/- 5% due to the greater basicity, and 18 +/- 1% due to the smaller size of the P(n-Bu)(3) ligands. In the Co -2 complexes the aryl and steric effects contribute 82 +/- 11% and 16 +/- 3%, respectively, but the electronic effect is small and not well defined. The dominant contributions of the aryl effect are striking. I ndeed, the correlation of the data with the aryl effect alone leads to R-2 values of 0.827 and 0.845 for the Mn-2 and Co-2 complexes, respec tively. The effect of sigma-donicity is ascribed to the greater concen tration of electron density in the region of the M-P bond when the ele ctron is in the sigma orbital, and the effect of ligand size is ascri bed to a lengthening of the M-M and M-PI bonds, the former decreasing sigma-orbital overlap and the latter decreasing the effective sigma-do nicity. No explanation is yet available for the origins of the empiric ally well established aryl effect even though it shows up in a wide ra nge of physicochemical measurements.