Session S3: Towards Design of Correlated Electron Materials

Spin-Orbit Coupling, Strong Interactions, and Topological Character

In recent years the electronic structure of crystalline solids has come under close scrutiny because of the various types of topological characters that may arise. Most of the work is done at the one-electron (non-interacting) level, and most innovations have arisen from model tight-binding Hamiltonians and their eigenvectors. This talk will focus on a few examples of $discoveries~made~computationally$ through DFT studies of actual materials, thus providing a physical realization as the discovery was made. Competition and partnership between strong interactions and spin-orbit coupling will be emphasized. Examples will include (1) the 'semi-Dirac' point Fermi surface phase in VO$_2$ thin films, the first member of a class now called $multi$~$Weyl$: massive in some direction, massless in other direction; (2) a nodal loop semimetal phase found in computational studies of thin SrVO$_3$ films, realized more recently in NbP etc.; (3) the buckled honeycomb lattice of a (111) bilayer of LaMnO$_3$ encased on LaAlO$_3$, which is a Chern insulator and may be a realization of the Weyl-Mott insulator proposed recently by Morimoto and Nagaosa. Acknowledgments: R. Pentcheva, V. Pardo, K.-W. Lee, S. Gangopadhyay.   

Towards prediction of correlated material properties using quantum Monte Carlo methods

Correlated electron systems offer a richness of physics far beyond noninteracting systems. If we would like to pursue the dream of designer correlated materials, or, even to set a more modest goal, to explain in detail the properties and effective physics of known materials, then accurate simulation methods are required. Using modern computational resources, quantum Monte Carlo (QMC) techniques offer a way to directly simulate electron correlations. I will show some recent results on a few extremely challenging materials including the metal-insulator transition of VO$_2$, the ground state of the doped cuprates, and the pressure dependence of magnetic properties in FeSe. By using a relatively simple implementation of QMC, at least some properties of these materials can be described truly from first principles, without any adjustable parameters. Using the QMC platform, we have developed a way of systematically deriving effective lattice models from the simulation. This procedure is particularly attractive for correlated electron systems because the QMC methods treat the one-body and many-body components of the wave function and Hamiltonian on completely equal footing. I will show some examples of using this downfolding technique and the high accuracy of QMC to connect our intuitive ideas about interacting electron systems with high fidelity simulations. The work in this presentation was supported in part by NSF DMR 1206242, the U.S. Department of Energy, Office of Science, Office of Advanced Scientific Computing Research, Scientific Discovery through Advanced Computing (SciDAC) program under Award Number FG02-12ER46875, and the Center for Emergent Superconductivity, Department of Energy Frontier Research Center under Grant No. DEAC0298CH1088. Computing resources were provided by a Blue Waters Illinois grant and INCITE PhotSuper and SuperMatSim allocations.   

Towards the design of novel cuprate-based superconductors

The rapid maturation of materials databases combined with recent development of theories seeking to quantitatively link chemical properties to superconductivity in the cuprates provide the context to design novel superconductors. In this talk, we describe a framework designed to search for new superconductors, which combines chemical rules-of-thumb, insights of transition temperatures from dynamical mean-field theory, first-principles electronic structure tools, materials databases and structure prediction via evolutionary algorithms. We apply the framework to design a family of copper oxysulfides and evaluate the prospects of superconductivity.   

\textbf{Itinerant magnetism without magnetic elements}

The origin of magnetism in metals has been traditionally discussed in two diametrically opposite limits: itinerant and local moments. Surprisingly, there are very few known examples of materials that are close to the itinerant limit, and their properties are not universally understood. In the case of the two such examples discovered several decades ago, both itinerant~ferromagnets~(IFMs) ZrZn$_{\mathrm{2}}$~and Sc$_{\mathrm{3}}$In, the understanding of their magnetic ground states draws on the existence of 3d electrons subject to strong spin fluctuations. In this talk I will contrast the physical properties of these two IFMs without magnetic elements with those of the recently discovered first itinerant~antiferromagnetic (IAFM)~metal with no magnetic constituents, TiAu. The IFMs have surprisingly different properties, with ZrZn$_{\mathrm{2}}$ showing signatures of mean field, Fermi liquid behavior, while the Sc$_{\mathrm{3}}$In compound is characterized by non-mean field magnetization exponents, and displays non fermi liquid behavior in both the FM and the paramagnetic states. The IAFM TiAu orders below a Neel temperature T$_{\mathrm{N}}$~$\approx $ K, about an order of magnitude smaller than in the IAFM Cr, rendering the spin fluctuations in TiAu more important at low temperatures. Like in the two IFMs, doping induces a quantum phase transition in TiAu, and the quantum critical behavior in all three systems is discussed and compared.   

In situ measurements of high temperature growth of correlated systems: a materials by design scheme

There is great interest in developing new ways to use predictive theory to accelerate materials synthesis. We have previously shown that DFT$+$DMFT electronic structure calculations are successful at predicting gaps and ordered moments, even when correlations are very strong.$^{\mathrm{[1,2]}}$ Building on these results, we set out to explore an even closer integration of theory and synthesis, aiming to discover new routes for doping Mott insulators and producing new superconductors. In situ high temperature high energy X-ray diffraction is used to determine the crystal structures of compounds just as they form from the growths, and the structural information is used as input for DFT$+$DMFT calculations that predict functionality, closing the synthesis loop by suggesting productive new directions. Using this approach, we have investigated the transition metal oxysulfide system Ba-Co-S-O and successfully discovered the new compound BaCoSO, and identified it as an interesting small gap Mott insulator by DFT$+$DMFT calculations even before any traditional crystal growth is attempted in the lab. [1] J. W. Simonson, et al. Proc. Nat. Acad. Sci. 109 (2012) E1815 [2] J. Guo, et al. Nat. Sci. Rep. 3 (2013) 2555