PDF《Supporting Information》

8 S-4 Bimetallic fragment geometries, energies and expectation values for (1). Fragments of the polymeric structure of

1 were selected which reproduced the local coordination environment around the metal and the bridging ligands. Initial guesses of the broken symmetry singlet (ioss =1) and triplet configurations were defined using the &

atomic command within Jaguar to define the formal charge (formal) and spin multiplicity (2spin) on each metal center and ligand. The LACV3P*+ basis set was employed which implemented triple-zeta 6-311G*+ for all non-metal atoms and an effective core potential LACV3P*+ for the CuII centers. The strength of the exchange interaction was determined using the method of Yamaguchi. However, particularly for the case of the paddlewheel dimer, the strength of magnetic communication was found to be particularly sensitive to the functional employed and the functional chosen;

B3LYP over-estimated the value of J by two orders of magnitude, whereas PBE0 over-estimated J by one order of magnitude. Whilst the Minnesota functional M06-L also over-estimated J, the value was the correct order of magnitude and used for subsequent discussion. Whilst there is considerable uncertainty in the computed energies of the exchange interactions we see that exchange within the paddlewheel is significantly larger than exchange within the tetrameric unit. Moreover at all levels of theory implemented the exchange via the acetato-/chloro-bridge is ferromagnetic whereas it is antiferromagnetic through the bipy ligand. Paddlewheel fragment (symmetry C1 ,748 basis functions) B3LYP/LACV3P*+ E(T)/Hartree -1835.1319727978 E(BSS)/Hartree -1835.19090899742 2.020 0.985 J/k -17,856 K PBE0/LACV3P*+ E(T)/Hartree -1833.31178125393 E(BSS)/Hartree -1833.32094955132 2.028 1.018 J/k -2,........