AmericanChemicalSociety.com

The dream of accelerating the discovery of materials with useful properties using computation and theory is quite old, but actual implementations of this idea are recent. Successes in material design using weakly correlated materials, are to a large degree due to implementations of first principle methods, such as Local Density Approximation (LDA) which are relatively accurate and computationally very efficient.

Since a large number of interesting physical phenomena, such as high temperature superconductivity and large Seebeck coefficients, are realized in strongly correlated electron systems, there is a great interest in the possibility of carrying out rational material design with correlated materials.

The advent of Dynamical Mean Field Theory (DMFT) removed many difficulties of the traditional electronic structure methods.  DMFT describes Mott insulators, as well as correlated metals.  DMFT combines ideas of quantum chemistry, such performing a full configuration interaction calculation (at a local level to avoid size consistency problems), and physics, such as carrying out a diagrammatic expansion around the band limit. DMFT treats quasiparticle bands and Hubbard bands on the same footing, and, unlike simpler approaches such as LDA+U, is able to describe the multiplet structure of correlated solids. The latter is being inherited from open shell atoms and ions.

A combination of DMFT with electronic structure methods LDA+DMFT is quite recent[1], and the most accurate implementation of this method was developed in part by support of this grant [2]. Just like LDA, this tools connect the atomic positions with the physical observables using very little information from experiment, and therefore it has the potential to accelerate material discovery.

The rational material design being developed with support of this grant, begins with a qualitative idea, which is then tested by a calculation. One of the major advances of realistic DMFT implementations is that now this calculation can be made material specific, resulting in a set of predictions that can be tested experimentally . The experimental results can either rule out the qualitative idea, in which case the process stops, or reinforce and refine the idea. Experiments also help calibrate the computational methods, which in turn lead to an improved material specific prediction in the next iteration.

The Dynamical Mean Field theory builds on the principle that the high order Feynman diagrams in perturbation theory are very local in space, if the problem of the solid is formulated in suitable localized basis set. It is important to build such localized basis by suitable projectors defined on Kohn-Sham states of LDA. The optimal projectors for DMFT were developed in Ref. [2]. The projector affects the valence of the ion. Since the DMFT solution is very sensitive to the valence of the correlated ion, it is crucial to build good projectors, in which the valence of the ion conicides with the valence measured in experiment. This is the crucial step of testing and calibrating the computational method. To this end, we build a module to compute the nuclear magnetic form factor [4], which is very sensitive measure of the valence of the magnetic ion. We computed the magnetic form factor for a series of actinide compounds, such as NpCoGa_5, PuCoGa_5, PuSb, PuTe, for which a high quality experimental data exists, and the form factor was not previously explained by other theories.  The agreement between LDA+DMFT theory and experiment is excellent, thus verifying that our projectors are very good.

We recently developed a module to compute thermoelectric properties of correlated materials within LDA+DMFT [4] and we tested the theory on two promissing materials FeAs2 and FeSb2 [5]. These two materials are isostructural, and differ only in the type of the pnictogen ion. However, the seebeck coefficient of FeSb2 is almost five times larger than in FeAs2, although FeSb2 has seven times smaller gap. Our theoretical method explains all essential features of experimental thermopower of FeAs2, but it does not explain measured thermopower of FeSb2. In Ref. [5] we derived the upper limit for electronic Seebeck coefficient of correlated semiconductor, and found that FeSb2 violates the limit, hence the thermopower of FeSb2 is not of electronic origin, and is most likely due to phonon drag, which unfortunately eliminates FeSb2 as a promising candidate for large electronic figure of merit.

We recently adapted our LDA+DMFT method to describe heterostructures and nanoscopic conductors [6], such as nanocontacts. This molecular DMFT approach will now allow us to describe super and nano-structures where the lattice contribution to the thermal conductivity is significantly reduced resulting in enhanced figure of merit.

[1] Electronic structure calculations with dynamical mean-field theory, G. Kotliar, S. Y. Savrasov, K. Haule, V. S. Oudovenko, O. Parcollet, and C. A. Marianetti Rev. Mod. Phys. 78, 865 (2006).

[2] Dynamical Mean-Field Theory within the Full-Potential Methods: Electronic structure of Ce-115 materials, Kristjan Haule, Chuck-Hou Yee, Kyoo Kim Phys. Rev. B 81, 195107 (2010).

[3] Neutron magnetic form factor in strongly correlated materials, Maria Elisabetta Pezzoli, Kristjan Haule, Gabriel Kotliar, arXiv:1007.3997.

[4] Thermoelectrics Near the Mott Localization—Delocalization Transition, K. Haule and G. Kotliar, Appeared in "Properties and Applications of Thermoelectric Materials", NATO Science for Peace and Security Series B: Physics and Biophysics, 2009, 119-131, DOI: 10.1007/978-90-481-2892-1_7.

[5] Thermopower of correlated semiconductors : application to FeAs2 and FeSb2, Jan M. Tomczak, K. Haule, T. Miyake, A. Georges, G. Kotliar, Phys. Rev. B 82, 085104 (2010).

[6] Dynamical Mean-Field Theory for Molecular Electronics: Electronic Structure and Transport Properties, D. Jacob, K. Haule, G. Kotliar, arXiv:1009.0523.