BNL Home

research

Overview

Why strong correlations? The focus of all of Comscope activities are centered around attempts to better understand strongly correlated materials. Strongly correlated materials have exceptional properties ranging from metal to insulator transitions, colossal magnetoresistance, high temperature superconductivity, heavy fermion behavior, and huge volume collapses to name but a few. Discoveries in this field of research have been made mostly by serendipity and it is clear by the discoveries carried out so far that these classes of materials hold great potential for applications, while still containing fundamental problems in basic science.

To overcome this problem and to ultimately advance the analysis and design of functional strongly correlated materials, the overarching objective of this Center is to develop Comsuite, a computational platform to perform first principles calculations for strongly correlated materials using extensions and simplified versions of dynamical mean field theory (DMFT) in combination with realistic electronic structure methods.

While developing Comsuite is Comscope’s primary mission, its scientists are also engaged in complementary basic research that advances the state of the art of dynamical mean field theory. This research takes a number of different forms, experimental, theoretical, and computational. These are described below:

Theoretical Spectroscopy

An important component of Comscope’s research efforts is to develop tools by which various experimental spectroscopies can be computed. We are developing software modules by which the response as measured by ARPES, the optical conductivity, the Seebeck coefficient, and RIXS can be computed. By post-processing numerical results obtained from Comsuite and incorporating how probe-particles interact with electrons or nuclei in materials (the so-called matrix element effect), we can calculate the response of materials numerically. Learn more

Materials Design

It is one of the foci of Comscope to produce tools and workflows that enable material design for strongly correlated systems. We eventually want to be able to take some desirable material property (high thermopower, a superconductor with a high Tc) and work backwards finding a material that has such properties. For weakly correlated materials, this possibility has largely been realized. The tools there, density functional theory based, are now sufficiently sophisticated and agile enough to enable the solution of such inverse problems. Correspondingly, we have seen the crucial role that density functional theory has played in the search for new topological insulators. However, because strongly correlated materials are under much poorer theoretical control, we are not yet at a point where the inverse problem can be immediately and reliably solved.

To make progress in this direction, Comscope is focusing on using its codes to produce better thermoelectrics. Learn more

Experimental Validation

Directly complementing our theoretical spectroscopy efforts, we have a number of experimental efforts based at both BNL and Rutgers. These range from synthesis of new strongly correlated semiconductors with a marcasite or pyrite structure (such as FeSb2) with superior thermoelectric properties, photoemission and optics on such materials to an independent RIXS effort bringing experiment and theory together to study materials with hidden order such as URu2Si2. Learn more

Multiscale Modeling

Comscope is actively engaged in connecting DMFT technologies to efforts to use these as inputs into multiscale models. These currently take two forms: (i) adapting quantum molecular dynamic simulations by taking into account strong electronic correlations using the Gutzwiller approach; and (ii) deriving Landau-Ginzburg theories using input from DMFT simulations. Learn more

High Performance Computation

Comscope is making an active effort to make its codes ready for high performance computation. It is doing so hand in hand with the computational science initiative (CSI) at BNL. The primary focus of these efforts has been the optimization of those parts of the Comsuite codes that can sustain the largest benefit from accelerated computing, in particular that involving GPUs. Two computational intensive parts of our codes are being targeted: GW and continuous time quantum Monte Carlo (CTQMC). Learn more

NSLS-II

Comscope is making active efforts to tie into activities at NSLS-II. It is already working closely with users at the SIX (a resonant inelastic x-ray beamline), and plans to extend this to CSX (a coherent soft x-ray beamline), ESM (a photoemission beamline), XPD (a powder x-ray diffraction beamline), and FIS (an infrared optics beamline). The theoretical spectroscopies post-processing modules that we are developing are a key part of this effort.