The problem of deriving from microscopic theory a Ginzburg-Landau free energy functional to describe the Peierls or charge-density-wave transition in quasi-one-dimensional materials is considered. Particular attention is given to how the thermal lattice motion affects the electronic states. Near the transition temperature the thermal lattice motion produces a pseudogap in the density of states at the Fermi level. Perturbation theory diverges and the traditional quasi-particle or Fermi liquid picture breaks down. The pseudogap causes a significant modification of the coefficients in the Ginzburg-Landau functional from their values in the rigid lattice approximation, which neglects the effect of the thermal lattice motion.which is a measure of the fluctuations along a single chain. It was assumed that the coefficients a, b, and c were independent of temperature and the measurable quantities at the transition were determined as a function of the interchain coupling. The transition temperature increases as the interchain coupling increases. The coherence length and specific heat jump depends only on the single chain coherence length, ξ 0 ≡ (c/|a|) 1/2 , and the interchain coupling. The width of the critical region, estimated from the Ginzburg criterion, was virtually parameter independent, being about 5-8 per cent of the transition temperature for a tetragonal crystal. Such a narrow critical region is consistent with experiment, and shows that Ginzburg-Landau theory should be valid over a broad temperature range.This paper uses a simple model to demonstrate some of the difficulties involved in deriving the coefficients a, b, c, and J from a realistic microscopic theory.
C. Microscopic theoryThe basic physics of quasi-one-dimensional CDW materials is believed to be described by a Hamiltonian due to Fröhlich 13 which describes electrons with a linear coupling to phonons. Even in one dimension this is a highly non-trivial many-body system and must treated by some approximation scheme. The simplest treatment 13-15 is a Hence, the temperature dependence is quite different from the dependence κ ∼ T 2 that was assumed in References 10,35,36 and the analysis there needs to be modified. The second and more serious problem is one of self-consistency. The coefficients a, b, and c are calculated neglecting fluctuations in the order parameter which will modify the electronic properties which in turn will modify the coefficients. In this paper a simple model is used to demonstrate that the fluctuations have a significant effect on the single-chain coefficients.An alternative microscopic theory, due to Schulz 16 , and which takes into account fluctuations in only the phase of the order parameter is briefly reviewed in Appendix A.
D. OverviewDiscrepancies between phonon rigid-lattice theory and the observed properties of the Peierls state well below the transition temperature T P were recently resolved 31,32 by taking into account the effect of the zero-point and thermal lattice motion on the electronic properties. It was shown that...