Field-induced spin-sensity-wave phases in quasi-one-dimensional conductors: theory versus experiments.

We show that the "quantized nesting" model misses important features of the magnetic-field-induced spin-density-wave (FISDW) phase diagram. Among them are the following: (1) the FISDW wave vector is not strictly quantized; (2) in some compounds, the FISDW diagram consists of two regions: (a) At low temperatures, there are jumps of the wave vector (i.e., the first order transitions between FISDW phases); (b) at high temperatures the jumps and the first order transitions disappear, but the wave vector is still a nontrivial function of a magnetic field. These are in agreement with the experiments on (TMTSF)2PF6.