Znorg. Chem. zyxwvu 1986, zyxwvu 25, zyxwvu 1830-1841 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ binds with either a sulfoxide or thioether donor and whether five-coordinate intermediate zyxwvutsr I1 or 111 reacts with ligand zyxwvuts 3. Inspection of Figure 5 and the scheme in the supplementary material reveals that five-coordinate intermediate I1 with the trans-dihalo configuration cannot give the observed products but would give only the desired all-trans complex. Since this isomer is never observed, this pathway would not be operative. This means that the initial trigonal-bipyramidal intermediate, trans-RuBr2- (S(CH2)4)3, would undergo an isomerization (pseudorotation) to yield a cis-dibromo five-coordinate intermediate (structure 111 of Figure 5) and finally complexation with ligand 3 would yield the observed products. An analysis of all the possible pathways available for coordination of ligand 3 to five-coordinate inter- mediate 111 of Figure 5 is presented in the supplementary material. Nevertheless, observed geometric isomers C (10) and E(9) are indeed predicted from the complexation pathways that utilize five-coordinate intermediate 111 (Figure 5). Thus, it appears that the use of a linear tridentate ligand to generate the all-trans 2,2,2 complex will not succeed. This ap- parently arises because the trans-dihalo geometry of the five- coordinate intermediate is in a very unfavorable equilibrium with the five-coordinate intermediate with the cis-dihalo geometry (111 of Figure 5). Consequently, the cis-dihalo geometry will be more favorable in the products of the substitution reactions, thus limiting their usefulness for the synthesis of the all-trans complexes. In the catalytic situation in which monodentate ligands are present, this problem can apparently be overcome. The subtle electronic factors favoring the cis geometries can be overridden by steric factors that favor the all-trans geometry. When bulkier mono- dentate substrates are used, the faster rates that are observed3 are consistent with the formation of a higher concentration of the desired all-trans complex, although very bulky thioether ligands (t-Bu2S) afford no catalytic activity since complexes of the stoichiometry R U X ~ ( R ~ S ) ~ ( R ~ S O ) ~ do not form in this case. Acknowledgment is made to L. C. Strickland (of these labo- ratories) for his assistance with the crystal structure investigation, to R. E. Shumate for his assistance with the syntheses, and to Professor Devon Meek of The Ohio State University for useful discussions. Registry No. 2, 101653-80-9; 3, 101835-53-4; 5, 101810-60-0; 6, 101835-55-6; 7, 101835-54-5; 8, 101835-56-7; 9, 101835-57-8; 10, 101915-10-0. Supplementary Material Available: Tables of H atom parameters, anisotropic thermal parameters, bond lengths and bond angles, and the numerical description of the pseudomirror plane and a figure illustrating the possible substitution pathways for the reaction between trans- RUB~~(S(CH~),CH~)~ and ligand 3 (5 pages). Ordering information is given on any current masthead page. According to policy instituted Jan I, 1986, the tables of calculated and observed structure factors (17 pages) are being retained in the editorial office for a period of 1 year following the appearance of this work in print. Inquiries for copies of these ma- terials should be directed to the Editor. . Contribution from the Laboratoire de Dynamique des Cristaux Molkulaires, ERA No. 465 du CNRS, Universit; des Sciences et Techniques de Lille, 59655 Villeneuve d’Ascq Cidex, France, and ER No. 139 du CNRS, Laboratoire de Chimie Quantique, Universiti Louis Pasteur, 67000 Strasbourg, France Dynamic, Static, and Theoretical Electron Deformation Density for Binuclear Transition-Metal Complexes: Dicobalt Hexacarbonyl Acetylene F. Baert,Ia A. Guelzim,Ia J. M. Poblet,Ib R. Wiest,lc J. Demuynck,lC and M. B5nard*lc Received July zyxwvutsrqpon 31, 1985 The electron deformation density distribution in C O ~ ( C O ) ~ R ~ C ~ (R = C(CH,),) has been obtained from low-temperature X-ray and neutron data and from theoretical wave functions at the Hartree-Fock and configuration interaction levels. The experimental determination of the deformation density of ( C - B U ~ C ~ ) C O ~ ( C O ) ~ at 122 K (space group zyxwv Pi, a = 8.289 (2) A, b = 8.400 (2) A, c = 13.552 (2) A, (Y = 88.67 (2)O, 0 = 94.45 (2)’, y = 106.75 (4)O, 2 = 2) does not show accumulation of density along the Cc-Co line (CG-Co = 2.462 A). The static deformation density maps obtained from multipolar refinement display two electron depopulation regions around each Co atom, separated by two peaks. An axis of minimum deformation density can be defined in the vicinity of each cobalt. This axis is collinear with the direction of the bent metal-metal bond defined by quantum-chemical LCAO calculations. The theoretical maps at S C F and CI levels display similar depopulation regions, but four distinct peaks are found around each metal. A “bent-bond peak” is obtained between the acetylene carbon atoms in both the experimental and the theoretical maps. The position of the peak, which is shifted away from the metal-metal line, results from the displacement of the r overlap, in turn caused by the distortion of the acetylenic system. These experimental and theoretical results are compared with the density maps previously published for the isolobal complex Ni2(C5HJ2C2H2. Much similarity is obtained between the distributions computed for both complexes and with the experimental maps of the dicobalt system. It seems that the disagreement with the experimental maps of Ni2(C5HS)2C2H2 is a consequence of the crystalline acentric and disordered structure of the dinickel complex. Fragment deformation densities have been computed for C O ~ ( C O ) ~ C ~ H ~ . The corresponding maps display an important accumulation of density in the T* orbital of the acetylene carbons, directed toward the Co atoms. The Mulliken population analysis shows that back-donation corresponds to a charge transfer of 0.88 e, balanced by a a-donation transfer of the same magnitude. A significant peak is also obtained at the center of the metal-metal bent bond. The results illuminate the synergic bonding characteristic of these complexes, where metal-metal interaction, the a donation, and the T back-donation are involved. Introduction The relatively small deformation of the valence part of the electron cloud is known to be the key for bonding energy and molecular structure. The more or less accurate analysis of the electron density distribution in molecules has been for a long time the privilege of theoretical chemistry, which can obtain such a (1) (a) Laboratoire de Dynamique des Cristaux Mol6culaire-s. (b) ER No. 139 du CNRS. Permanent address: Department de Quimica Fisica, Facultat de Quimica, Pz. Imperial Tarraco 1, Tarragona 43005, Spain. (c) ER No. 139 du CNRS. distribution from a computed wave function. Over the last 15 years, however, increasingly accurate X-ray or neutron diffraction experiments have allowed us, according to the term coined by Coppens, to “see the electrons”.2 Another condition for ensuring the chemical interest of these experimental density distributions is to focus the discussion of the results on this part of the electron cloud, which has been reorg- anized by chemical bonding. This is obtained by substracting from the observed density distribution of the system another distribution (2) Coppens, P. J. Chem. Educ. 1984, 61, 761. 0020-1669/86/1325-1830$01.50/0 0 1986 American Chemical Society