The selective one-step preparation of trans-Fe(CO)3(PPh3)2, 1, in the absence of Fe(CO)4-(PPh3) has made it possible to obtain 1 pure and in crystalline form. trans-Fe(CO)3(PPh3)2 crystallizes in the orthorhombic space group Pbca with cell parameters a = 18.216(5) Å, b = 17.0380(10) Å, c = 21.804(3) Å, and Z = 8; R = 0.042 and Rw = 0.056 for 3891 independent reflections with I > 2.0σ(I). The crystal structure of 1 is compared to Ru(CO)3(PPh3)2, Os(CO)3(PPh3)2) Fe(CO)2(CS)(PPh3)2, and a selection of complexes of the types trans-Fe(CO)3(PR3)2 (R = Ph, NMe2, OMe) and trans-Fe(CO)4PR3 (R = Ph, NMe2, CMe3, SiMe3). Effects of equatorial CO substitution in 1 by the diazonium ion and nitrile ligands are discussed. Ab initio calculations that employ effective core potential basis sets at the Hartree-Fock level and which include perturbational corrections for electron correlation at the MP2 level are reported for trans-Fe(CO)3(PH3)2. Good agreement between theory and experiment can only be obtained at correlated levels of theory in conjunction with double-ζ quality basis sets on iron and all ligands. Population analyses are employed to examine the effects of electron correlation and to delineate the iron-ligand bonding. Electron correlation serves overall to reduce the molecular quadrupole moment. Comparison between the crystal structure of 1 and the theoretical structure of the model system allow differentiation between intrinsic structural preferences and packing effects. In particular, the comparison points out that the Fe-P bond length differences and the nonlinearity of the P-Fe-P backbone are not intrinsic bonding features but are caused by the packing.
Available at: http://works.bepress.com/rainer-glaser/45/