g-on mean field theory of the t-J model
Abstract
Since the discovery of high-Tc cuprates, there are many attempts to understand this interesting class of materials. It has generally been accepted that the essence of electronic properties is in the CuO2 layers, and that the low-lying excitations of the layers are described by a single-band t-J model. Though the exact results are not known, the slave-boson mean field theory has played important roles in offering explicit predictions to be compared with experiments regarding the doping dependence and effect of fermi surfaces, and the agreement between the theoretical results and what have been seen in the experiments on the spin excitations probed by neutron scattering and nuclear magnetic resonance (NMR), phonons and transport properties are noteworthy. The scheme is based on spin-charge separation and the electron operator is decoupled into spinon and holon, which is assumed to be fermion and boson respectively. The phase diagram derived by the mean field approximation in this scheme is given. The states realized include antiferromagnetic (AF) state, superconducting (SC) state, "spin gap" (SG) phase, anomalous metallic (AM) state, and electron liquid (EL) state. The AM state, where both spinons and holons move coherently and essentially independent of each other, i.e., spin and charge are also separated, is described as the uniform RVB (u-RVB) state. In the SG state, where the spin and the charge are also separated, the spinons make short range singlet pairs, which is described as the singlet RVB (s-RVB) state. On the other hand in the EL state spinons and holons are no longer independent, because of the bose condensation of holons, and couples together to form electrons. Based on this phase diagram various physical quantities, especially the spin gap phenomenon typically seen in the rate of NMR, have been discussed in connection with the SG state in the phase diagram.
