We propose a variational wavefunction containing parameters to tune the probabilities of all the possible onsite configurations for the periodic Anderson model. We call it the full onsite-correlation wavefunction (FOWF). This is a simple extension of the Gutzwiller wavefunction (GWF), in which one parameter is included to tune the double occupancy of the f electrons at the same site. We compare the energy of the GWF and the FOWF evaluated by the variational Monte Carlo method and that obtained with the density-matrix renormalization group method. We find that the energy is considerably improved in the FOWF. On the other hand, the physical quantities do not change significantly between these two wavefunctions as long as they describe the same phase, such as the paramagnetic phase. From these results, we not only demonstrate the improvement by the FOWF, but we also gain insights on the applicability and limitation of the GWF to the periodic Anderson model.We propose a variational wavefunction containing parameters to tune the probabilities of all the possible onsite configurations for the periodic Anderson model. We call it the full onsite-correlation wavefunction (FOWF). This is a simple extension of the Gutzwiller wavefunction (GWF), in which one parameter is included to tune the double occupancy of the f electrons at the same site. We compare the energy of the GWF and the FOWF evaluated by the variational Monte Carlo method and that obtained with the density-matrix renormalization group method. We find that the energy is considerably improved in the FOWF. On the other hand, the physical quantities do not change significantly between these two wavefunctions as long as they describe the same phase, such as the paramagnetic phase. From these results, we not only demonstrate the improvement by the FOWF, but we also gain insights on the applicability and limitation of the GWF to the periodic Anderson model.

arXiv:1611.05174Journal of the Physical Society of JapanNov 30, 20164 citations

Variational study of the one-dimensional t-J-V model

Gang Su

Physics Letters ADec 1, 20000 citations

Variational Wavefunction for Quantum Antiferromagnets

Franjo Franjic, Sandro Sorella

We present here a new approach to determine an accurate variational wavefunction for general quantum antiferromagnets, completely defined by the requirement to reproduce the simple and well known spin-wave expansion. By this wavefunction, it is possible to obtain the correct behavior of the long distance correlation functions for the one dimensional $S=1/2$ antiferromagnet, i.e. {\em a system without long range order}. The qualitative difference between the integer and half integer case is also easily understood with this variational approach. Finally we present numerical results for the 2D XY model, showing that the present wavefunction has an overlap $> 0.99 $ to the exact ground state of the $S=1/2$ model for finite system up to $6 \times 6$ clusters.We present here a new approach to determine an accurate variational wavefunction for general quantum antiferromagnets, completely defined by the requirement to reproduce the simple and well known spin-wave expansion. By this wavefunction, it is possible to obtain the correct behavior of the long distance correlation functions for the one dimensional $S=1/2$ antiferromagnet, i.e. {\em a system without long range order}. The qualitative difference between the integer and half integer case is also easily understood with this variational approach. Finally we present numerical results for the 2D XY model, showing that the present wavefunction has an overlap $> 0.99 $ to the exact ground state of the $S=1/2$ model for finite system up to $6 \times 6$ clusters.

arXiv:cond-mat/9601070Jan 18, 19960 citations

S=1/2

antiferromagnet, i.e. {\em a system without long range order}. The qualitative difference between the integer and half integer case is also easily understood with this variational approach. Finally we present numerical results for the 2D XY model, showing that the present wavefunction has an overlap $> 0.99 $ to the exact ground state of the

S=1/2

model for finite system up to

6 \times 6

clusters.