The bond order wave (BOW) phase of half-filled linear Hubbard-type models is narrow and difficult to characterize aside from a few ground state properties. The BOW phase of a frustrated Heisenberg spin chain is wide and tractable. It has broken inversion symmetry C i in a regular array and finite gap E m to the lowest triplet state. The spin-BOW is exact in finite systems at a special point. Its elementary excitations are spin-1/2 solitons that connect BOWs with opposite phase. The same patterns of spin densities and bond orders appear in the BOW phase of Hubbard-type models. Infrared (IR) active lattice phonons or molecular vibrations are derivatives of P, the polarization along the stack. Molecular vibrations that are forbidden in regular arrays become IR active when C i symmetry is broken. 1:1 alkali-TCNQ salts contain half-filled regular TCNQ -stacks at high temperature, down to 100 K in the Rb-TCNQ(II) polymorph whose magnetic susceptibility and polarized IR spectra indicate a BOW phase. More complete modeling will require explicit electronic coupling to phonons and molecular vibrations.
IntroductionFace-to-face stacks of organic π-radicals suggest a one-dimensional (1D) electronic structure [1,2]. It was realized [1] early on that 1D Hubbard models rationalize the striking magnetic, optical and electronic properties of π-radical solids in terms of the LUMO of π-acceptors A or the HOMO of π-donors D. Segregated stacks of A or D can have different degree of filling by electrons or holes and different unit cells along the stack. Half-filled systems with one electron per site are Mott insulators. Good conductors deviate from half filling in a regular array. Triplet spin excitons signal dimerized stacks. Mixed DA stacks have alternating site energy ±ε and undergo a neutral-ionic transition (NIT) in charge-transfer (CT) salts of special interest. The bandwidth of π-stacks is 4t ~ 1 eV, an order of magnitude smaller than in conjugated polymers with π-bonds along the backbone. 1D quantum cell or Hubbard-type models for π-radical organic solids or conjugated polymers have been extensively developed and applied for decades, sometimes almost quantitatively [3][4][5][6][7]. Both electron-phonon coupling and electron correlation are important.