Very weak electron–electron exchange interactions in paramagnetic dinuclear tris(pyrazolyl)boratomolybdenum centres with extended bridging ligands: estimation of the exchange coupling constant J by simulation of second-order EPR spectra†

Abstract

Two series of dinuclear complexes have been prepared in which paramagnetic nitrosylmolybdenum(I) or oxomolybdenum(V) units have been attached to either end of very long bis-pyridyl or bis-phenolate bridging ligands respectively. The first series of complexes is [{Mo^I(Tp^Me,Me)(NO)Cl}₂{µ-L}] (L = 4,4′-bis[2-(4-pyridyl)ethen-1-yl]biphenyl; 4,4″-bis[2-(4-pyridyl)ethen-1-yl]terphenyl 2; 4,4′-bis[2-(4-pyridyl)ethen-1-yl]benzophenone 3; 4,4′-bis[2-(4-pyridyl)ethen-1-yl]benzil 4; or 6,6′-bis[2-(4-pyridyl)ethen-1-yl]-2,2′-bipyridine 5). The second series of complexes is [{Mo^V(Tp^Me,Me)(O)Cl}₂(µ-L)] (H₂L = HOC₆H₄OC(S)OC₆H₄OH 6; HOC₆H₄OS(O)OC₆H₄OH 7; or HOC₆H₄OC(O)C₆H₄C(O)OC₆H₄OH 8, with all-para substitution for the C₆H₄ units in each case). The very weak spin exchange interactions between the remote paramagnetic centres result in many cases in second-order EPR spectra, because |J| ≈ A (where J is the exchange coupling constant, and A the electron–nucleus hyperfine coupling). In these cases the appearance of the EPR spectra is complicated and sensitive to small changes in the magnitude of J, which could be exploited to estimate values for |J| by comparing the measured spectra with computer simulations calculated using a range of values of |J|. For the first series of complexes the spin exchange interactions decrease in the order 1 (|J| ≥ 4000), 2 (1000), 3 (150), 4 (43), 5 (|J| ≤ 10 MHz) which is readily explicable in terms of the lengths, conformations and substitution patterns of the bridging ligands. For the second series of complexes, 6 and 7 both gave second-order spectra with |J| = 2000 MHz, whereas 8, with a much longer bridging ligand, has |J| ≤ 10 MHz. Crucially, these spin-exchange interactions are much too weak to be determined by conventional magnetic susceptibility measurements (|J|  1 cm^–1), and therefore simulation of second-order EPR spectra provides a simple route to providing useful information about the relative magnitudes of very weak spin exchange interactions which is not available by any other route.