A “Star” Antiferromagnet: A Polymeric Iron(III) Acetate That Exhibits Both Spin Frustration and Long‐Range Magnetic Ordering

Abstract

The preparation of new geometrically spin-frustrated magnetic materials that approximate theoretical models is a challenge. Although theMermin–Wagner theorem indicates that long-range magnetic order can exist in two dimensions at zero Kelvin, order can be destroyed either by quantum fluctuations or geometric frustration even at this temperature. Theoretical studies indicate that the ground state of a spin-1/2 Heisenberg antiferromagnet is most likely to be semiclassically ordered. However, the interplay of geometric frustration and quantum fluctuations has been found to give rise to a paramagnetic ground state without semi-classical long-range order in two types of lattice. The first of these lattices is the famous Kagom+ lattice (T8) and the second is the so-called “star” lattice (T9; Scheme 1), which may serve as a new example of a quantum paramagnet. 5] The triangles are corner-sharing in the Kagom+ lattice whereas they are separated by a bridge in the star lattice, which means that their next-nearest-neighbor exchange interactions are different. 5] The magnetic J exchange pathways in the Kagom+ lattice are all equivalent, whereas the intra-triangular JT pathway in the star lattice is weaker than the inter-triangular JD pathway. In contrast to the rapid development of Kagom+-type antiferromagetic lattices 7] and related, geometrically spin-frustrated lattices, there appears to date to be no report of a compound with a genuine star lattice. Triangular clusters with superexchange pathways, such as the widely employedM3(m3-O) clusters, whereMmay be Fe , Fe, Co, Ni, Cu, V, or Cr, can be used to generate frustrated lattices, including the desired magnetically frustrated star lattice. This star lattice can be described in vertex notation as 3.12 (see Scheme S1 in the Supporting Information), a lattice that is a uniform, three-connected twodimensional net with large voids. Three-connected node subunits that prefer to bond in a planar fashion, such as the basic cationic iron(III) carboxylate cluster [Fe3(m3-O)(mO2CR)6L3] , where L may be water, methanol, or pyridine, must be used to avoid three-dimensional connections. These carboxylate clusters are good potential building blocks because they are easily prepared, prefer planar bonding, and the R groups and L ligands can easily be varied. The cationic [Fe3(m3-O)(m-O2CR)6L3] + moiety has previously served as a sixor three-connected node (see Scheme S2 in the Supporting Information) to form either threeor zero-dimensional porous frameworks depending upon the nature of the carboxylate, which may be either fully or partially substituted by dicarboxylates; the L ligands are usually retained as terminal ligands. Although no example is known to date, it should be possible to substitute the L ligands located in the triangular [Fe3(m3-O)(m-O2CR)6L3] + cation plane with other bridging bidentate ligands that are better at both mediating antiferromagnetic interactions and producing a two-dimensional star lattice. Herein, we report the use of bidentate acetate bridging ligands to link [Fe3(m3-O)(mOAc)6] + cations together to form [Fe3(m3-O)(m-OAc)6(H2O)3][Fe3(m3-O)(m-OAc)7.5]2·7H2O (1), a new compound with the desired star lattice. Single-crystal X-ray diffraction studies of 1 at 293 and 90 K revealed that isolated [Fe3(m3-O)(m-OAc)6(H2O)3] + cations (Figure 1) occupy the dodecagonal channels formed by the stacking of acetate-bridged [Fe3(m3-O)(m-OAc)7.5] 1/2 anionic layers; the dihedral angle between the triangular [Fe3(m3-O)(m-OAc)6(H2O)3] + cations and the [Fe3(m3-O)(mScheme 1. A comparison of the Kagom (T8, left) and star (T9, right) lattices with indication of the magnetic J exchange pathways.