HEXANUCLEAR FERRIC COMPLEXES POSSESSING DIFFERENT DEGREES OF SPIN FRUSTRATION

The synthesis, single-crystal X-ray structures, and magnetochemical properties are reported for two new hexanuclear ferric complexes, [Fe6(mu3-O)2(mu2-OH)2(mu2-O2CCH3)10(C7H11N2O)2]-4CH2Cl2 (9.4CH2Cl2) and [Fe6(mu3-O)2(C6H6NO)8Cl4](ClO4)2.4MeCN (10.4MeCN). The ligand C7H11N2O- is the anion of 2-(N-methylimidazol-2-yl)-2-hydroxypropane and C6H6NO- is the anion of 2-pyrididylcarbinol. The reaction of [Fe3O(OAc)6(PY)3]ClO4 in acetonitrile with 2-(N-methylimidazol-2-yl)-2-hydroxypropane gives a brown oil which can be crystallized via vapor diffusion of CH2Cl2 With hexanes to give complex 9.4CH2Cl2. The reaction of FeCl3.6H2O and 2-pyridylcarbinol in acetonitrile gives, upon addition of NaClO4, Complex 10.4MeCN. Complex 9 has six high-spin Fe(III) ions and can be viewed as two trinuclear mu3-oxo-bridged subunits bridged by two mu2-hydroxo and mu2-acetato ligands. Complex 10 can also be viewed as two asymmetric triangular Fe3-mu3-oxo subunits bridged together by alkoxo groups. However, in complex 9 there is a planar array of six Fe(III) ions, whereas in complex 10 the six high-spin Fe(III) ions are arranged in a chair conformation. Variable temperature DC magnetic susceptibility data measured at 10.0 kG are presented for both complexes. For complex 9.3/2CH2CI2 mu(eff)/molecule was found to be 9.10 mu(B) at 300.0 K, and as the temperature is decreased, this value increases to a maximum of 10.55 mu(B) at 30.0 K, whereupon there is a decrease to 9.77 mu(B) at 5.00 K. Complex 10.MeCN in a 10.0 kG field gives mu(eff)/molecule = 8.82 mu(B) at 320.0 K. In contrast to 9, the mu(eff)/molecule for complex 10-MeCN decreases with decreasing temperature to 6.08 mu(B) at 5.01 K. Least-squares fitting of the reduced magnetization (M/Nmu(B)) versus H/T data for paraffin-embedded complex 9.3/2CH2CI2 in the range of 5.00-50.0 kG external field and 2.0-30.0 K clearly shows that complex 9 has a well isolated S(T) = 5 ground state. Reduced magnetization versus H/T data are also presented for a parafilm-embedded sample of complex 10.MeCN in external fields of 0.50-50 kG at temperatures of 2-30 K. Fitting the high-field data suggests the presence of a S(T) = 3 ground state. However, the fit of the low-field data is not good for just an isolated S(T) = 3 ground state. Theoretical calculations were carried out for two of the known Fe(III)6 complexes. The energies of all of the 4332 different spin states of a Fe(III)6 complex were calculated, taking into account the pairwise magnetic exchange interactions within mu3-oxobridged Fe(III)3 triangular subunits (parameters J1, J2, and J3) and the interaction (J4) between iron ions in two Fe3O triangular subunits. The 20-320 K data measured at 10.0 kG for complex 5 (isostructural to 9) could be fit well with a theoretical calculation where J1 = -5.6(5) cm-1, J2 = J3 = -38(1) cm-1, and J4 = -7.5(l) cm-1. In agreement with experimental data, a well-isolated S(T) = 5 ground state is predicted. The theoretical fit of 20-320 K data for complex 10.MeCN in a 10.0 kG field gives fitting parameters of J1 = J2 = -18(1) cm-1, J3 = -52(2) cm-1, and J4 = -3(2) cm-1. Even though all pairwise exchange interactions are antiferromagnetic, a Fe(I16 complex can have a ground state with a spin ranging from S(T) = 0 to 5. The ground state found for a given Fe(III)6 complex is dependent upon which pairwise antiferromagnetic interactions are strongest.