Phosphorene, a single layer of black phosphorus, is a two dimensional material whose direct semiconducting band-gap, high carrier mobility and sensitivity to strain make it promising for applications. The nature of the binding between layers of black phosphorus was probed using the highly accurate many-body diffusion quantum Monte Carlo approach. A large and unexpected reorganization of electronic charge density was discovered, exposing significant weaknesses in current Density Functional Theory (DFT) approximations. By identifying the nature of the binding between the layers, a necessary step has been taken towards understanding the structure and properties of the material. Identifying a clear failure mode of DFT functionals may guide their future improvement. For their study of black phosphorus, the research team used the scalable QMCPACK simulation package. With Mira, the Blue Gene Q at ALCF and Sequoia, the Blue Gene Q at Lawrence Livermore National Laboratory, the researchers were able to obtain highly accurate calculations of binding between layers of black phosphorus both in a bilayer and in the bulk. This demonstrated the accuracy of the method and provided motivation for the calculations of the charge reorganization. “The Nature of the Interlayer Interaction in Bulk and Few-Layer Phosphorus”, L. Shulenburger, A. D. Baczewski, Z. Zhu, J. Guan, D. Tomanek, Nano Letters, 2015, 10.1021/acs.nanolett.5b03615...Read More
Argonne researchers were able to demonstrate the potential of using QMC in their studies of ellipticine, a promising drug for uterine cancer treatment. Research in drug action at the molecular level depends primarily on understanding and defining the physical/chemical interaction between the drug and its receptor, and then understanding how to connect this interaction to the pharmacological response. Over the past couple of decades, the screening of molecules for their active principle has relied greatly on the ability to model candidate molecules before experimental consideration. This task can only be achieved with an ab initio quantum chemistry theory, such as QMC, that takes electron correlation effects into account. For their groundbreaking studies of ellipticine, the Argonne research team used the scalable QMCPACK simulation package. With Mira, the Blue Gene Q at ALCF, the researchers were able to obtain a highly accurate binding energy calculation (at chemical accuracy of ~1 kcal/mol) of ellipticine to DNA (33.6 ±0.9 kcal/mol), while other traditional methods had predicted no binding (DFT: -5 kcal/mol). This case study shows the reliability and efficiency of QMC for characterizing the binding energies for biological systems, providing a critical input for improved drug modeling efforts. “Application of Diffusion Monte Carlo to Materials Dominated by van der Waals Interactions”, Anouar Benali, Luke Shulenberger, Nichols A. Romero, Jeongnim Kim, and O. Anatole von Lilienfeld. Journal of Chemical Theory and Computation (2014). http://pubs.acs.org/doi/abs/10.1021/ct5003225...Read More
Magnetic couplings in a realistic cuprate system have been correctly predicted for the first time with highly accurate Quantum Monte Carlo (QMC) calculations. QMC methods accurately describe many-body systems, but at a high computational cost. High-performance computers, such as those at the Oak Ridge Leadership Computing Facility, allow application of QMC to problems that have remained unsolved for decades, such as magnetic phenomena associated with the remarkable high-temperature superconductivity in cuprate materials. We have used QMC to study magnetic properties of Ca2CuO3, an effectively one-dimensional counterpart of the famous superconducting cuprates. The obtained magnetic superexchange coupling, J, is in very good agreement with experiment, demonstrating the predictive capability. Effective magnetic models of superconductivity can now be derived with confidence from theory, which could lead to better fundamental predictions of superconductor behavior. This development could lead to better predictions of superconductor behavior derived from fundamental laws. The dependence of J on initial approximations is strongly reduced compared to other approaches, due to the variational character of the energy within the QMC approach. K. Foyevtsova, J. T. Krogel, J. Kim, P. R. C. Kent, E. Dagotto, and F. A. Reboredo, “Ab initio quantum Monte Carlo calculations of spin superexchange in cuprates: the benchmarking case of Ca2CuO3,” Physical Review X 4 031003...Read More
QMCPACK is an open-source production level many-body ab initio Quantum Monte Carlo code for computing the electronic structure of atoms, molecules, and solids.