intensity of the main line) and is lost in the base line. If a hidden higher binding energy peak were comparable in intensity to that of the lower binding energy peak, then the Ru 3d512 intensity derived from the deconvoluted spectrum shown in Figure 1 would be anomalously low compared with that of other complexes while the C 1s line intensity would be anomalously high. Furthermore, the large Ru 3d splitting required to hide the high-binding-energy Ru 3d peak under the C 1s line ought to be observable even in the Ru 2p3/2 spectrum of the complex. We examined these possibilities by comparing Ru 3d5/2/C1 2p and Ru 2p312/CI 2p intensity ratios recorded for 7 with those of all the other ruthenium chloride complexes. The Ru 3d5/?/C1 2p and Ru 2~~,~/ C l 2p intensity ratios for 7 are both within 20% (albeit on the low side of all but one) of all other measured ratios. The Ru 2p3p line width for 7 is normal. Given experimental errors, we find then that at least 80% of the Ru 3d512 intensity expected from two ruthenium centers in 7 is accounted for in the deconvoluted spectrum. Consequently, the metal XP spectrum of 7 is consistent with that of a strongly delocalized class I11 complex. For such a complex a large portion of the total metal core level photoemission intensity resides in a single, low-binding-energy line. From the quality of the spectrum we estimate, using eq 2, that the electron exchange parameter a for 7 must be greater than 0.9.
ConclusionsThe diruthenium complexes examined have provided the opportunity to systematically assess the effects of ligand coordination environment, metal oxidation state, and metal-metal interaction on the metal XP spectra. One complex is of particular interest in this series; the symmetric, mixed-valence [Ru(II),Ru(III)]
complex [As(p-tol)3]2C1RuC13RuC1[As(p-tol)3]2. Depending onthe strength of the metal-metal interaction, the metal XP spectrum from this complex could give any results along the continuum from two equally intense metal lines separated by 1 eV or more, to unequal intensities from the two lines with the lower binding energy line "stealing" intensity from the higher binding energy line, to a single, narrow metal line (indistinguishable metal centers). The spectrum observed is a function of the extent of unpaired electron delocalization during photoemission. It was found that, to the degree of resolution possible, this complex gives a metal spectrum in which at least 80% of the metal signal is present in a single, low-binding-energy final state. This experimental result reflects more extensive unpaired electron delocalization in this complex compared with that in the Creutz-Taube complex [(NH3)5Ru(pz)Ru(NH3)5]5+, which gives two well-separated metal lines of equal intensity. We estimate the value of a in 7 is greater than 0.9, whereas in the Creutz-Taube complex it is less than 0.10. This work, combined with Hush's theory, suggests that very unusual metal XP spectra ought to be obtained from class I11 mixed-valence complexes having metal-metal spacings between 3.5 an...