ALPS 2 Tutorials:DMFT-02 Hybridization

Revision as of 22:03, 13 May 2010 by Troyer (talk | contribs) (New page: =Tutorial 02: Hybridization Expansion CT-HYB / Tutorial 03: Interaction Expansion CT-INT= We will now reproduce the same result with a continuous-time Quantum Monte Carlo code: the Hybrid...)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Tutorial 02: Hybridization Expansion CT-HYB / Tutorial 03: Interaction Expansion CT-INT

We will now reproduce the same result with a continuous-time Quantum Monte Carlo code: the Hybridization or CT-HYB code by Werner et al.. The parameters are the same, apart from the command for the solver:


You can find the parameter files (called *tsqrt2) in the directory tutorials/dmft-02-hybridization in the examples.

After running these simulations compare the output to the Hirsch Fye results. To rerun a simulation, you can specify a starting solution by defining G0OMEGA_INPUT, e.g. copy G0omga_output to G0_omega_input, specify G0OMEGA_INPUT = G0_omega_input in the parameter file and rerun the code.

You will notice that the results are relatively noisy. The reason for that is that the expansion order at such high temperatures is very small, which renders the measurement procedure inefficient. You can improve statistics by increasing the total run time (MAX_TIME) or by running it on more than one CPU. For running it with MPI, try mpirun -np procs dmft parameter_file or consult the man page of your mpi installation.

It is also instructive to run these calculations with a CT-INT code. This code performs an expansion in the interaction (instead of the hybridization). The corresponding parameter files are very similar, you can find them in the directory tutorials/dmft-03-interaction.

Tutorial 04: Mott Transition

Mott transitions are metal insulator transitions that occur in many materials, e.g. transition metal compounds, as a function of pressure or doping. The review by Imada et al. gives an excellent introduction to the subject and mentions V2O3 and the organics as typical examples.

MIT are easily investigated by DMFT as the relevant physics is essentially local (or k-independent): At half filling the MIT can by modeled by a self energy with a pole at $\omega=0$ which splits the noninteracting band into an upper and a lower Hubbard band. In this context it is instructive to suppress antiferromagnetic long range order and enforce a paramagnetic solution in the DMFT simulation, to mimic the paramagnetic insulating phase. For this the up and down spin of the Green's functions are symmetrized (parameter SYMMETRIZATION = 1;).

We investigate the Mott transition in single-site DMFT, as a function of interaction at fixed temperature \beta t=20 (see e.g. Fig. 2 in this paper). Starting from a non-interacting solution we see in the imaginary time Green's function that the solution is metallic for U/t \leq 4.5, and insulating for U/t\geq 5. A coexistence region could be found by starting from an insulating (or atomic) solution and trying to convert it for smaller $U$.

Imaginary time Green's functions are not easy to interpret, and therefore many authors employ analytic continuation methods. There are however two clear features: the value at \beta corresponds to -n, the negative value of the density (per spin). And the value at \beta/2 goes for decreasing temperature to A(\omega=0), the spectral function at the Fermi energy: \beta G(\beta/2 \rightarrow A(0). From a temperature dependence of the imaginary time Green's function we can therefore immediately see if the system is metallic or insulating.

The beta20_U* parameter files in the directory tutorials/dmft-04-mott should show you how to go from a metallic (at small $U$) to an insulating (at large $U$) solution, at fixed $\beta$. The largest value of $U$ is deep within the insulating phase.

Tutorial 05: Orbitally Selective Mott Transition

An interesting phenomenon in multi-orbital models is the orbitally selective Mott transition, first examined by Anisimov et al. A variant of this, a momentum-selective Mott transition, has recently been discussed in cluster calculations as a cluster representation of pseudogap physics.

In an orbitally selective Mott transition some of the orbitals involved become Mott insulating as a function of doping or interactions, while others stay metallic.

As a minimal model we consider two bands: a wide band and a narrow band. In addition to the intra-orbital Coulomb repulsion $U$ we consider interactions $U'$, and $J$, with $U' = U-2J$. We limit ourselves to Ising-like interactions - a simplification that is often problematic for real compounds.

We choose here a case with two bandwidth $t1=0.5$ and $t2=1$ and density-density like interactions of $U'=U/2$, $J=U/4$, and $U$ between $1.6$ and $2.8$, where the first case shows a Fermi liquid-like behavior in both orbitals, the $U=2.2$ is orbitally selective, and $U=2.8$ is insulating in both orbitals.

The parameter files can be found in tutorials/dmft-05-osmt in the directories beta30_U1.8_2orbital, beta30_U2.2_2orbital and beta30_U2.8_2orbital. A paper using the same sample parameters can be found here.

Tutorial by Emanuel - Please don't hesitate to ask!