Session D10: Invited Session: Advancements in Computational Physics using NSF's TeraGrid/XSEDE Resources

Benchmark Calculations of Atomic Collision Processes

The rapid development of computational resources has resulted in enormous improvements in the accuracy of numerical calculations of atomic collision processes. This talk will concentrate on recent advances in the computational treatment of charged-particle and intense short-pulse laser interactions with atoms, ions, and small molecules. Examples include electron collisions with heavy complex targets that are of significant importance in many modelling applications in plasma and astrophysics, fundamental studies of highly correlated 4-body Coulomb processes such as simultaneous ionization with excitation, and the accurate solution of the time-dependent Schr\"odinger equation in the presence of intense femto/attosecond laser fields, which paves the way for quantum dynamic imaging and coherent control.   

Prediction and Design of Materials from Crystal Structures to Nanocrystal Morphology and Assembly

Predictions of structure formation by computational methods have the potential to accelerate materials discovery and design. Here we present two computational approaches for the prediction of crystal structures and the morphology of nanoparticles. Many materials properties are controlled by composition and crystal structure. We show that evolutionary algorithms coupled to ab-initio relaxations can accurately predict the crystal structure and composition of compounds without any prior information about the system. We will discuss results for various systems including the prediction of unexpected quasi-1D and 2D electronic structures in Li-Be compounds under pressure [1] and of the crystal structure of the superconducting high-pressure phase of Eu [2]. The self-assembly of nanocrystals into mesoscale superlattices provides a path to the design of materials with tunable electronic, physical and chemical properties for various applications. The self-assembly is controlled by the nanocrystal shape and by ligand-mediated interactions between them. To understand this, it is necessary to know the effect of the ligands on the surface energies (which tune the nanocrystal shape), as well as the relative coverage of the different facets (which control the interactions). Density functional calculations for the binding energy of oleic acid-based ligands on PbSe nanocrystals determine the surface energies as a function of ligand coverage. The Wulff construction predicts the thermodynamic equilibrium shape of the PbSe nanocrystals as a function of the ligand coverage. We show that the different ligand binding energies on the {100} and {111} facets results in different ligand coverages on the facets and predict a transition in the equilibrium shape from octahedral to cubic when increasing the ligand concentration during synthesis. Our results furthermore suggest that the experimentally observed transformation of the nanocrystal superlattice structure from fcc to bcc is caused by the preferential detachment of ligands from particular facets, leading to anisotropic ligand coverage [3]. \\[4pt] [1] J. Feng, R. G. Hennig, N. W. Ashcroft and Roald Hoffmann. Nature 451, 445 (2008). \\[0pt] [2] W. Bi, Y. Meng, R. S. Kumar, A. L. Cornelius, W. W. Tipton, R. G. Hennig, Y. Zhang, C. Chen, and J. S. Schilling. Phys. Rev. B 83, 104106 (2011). \\[0pt] [3] J. J. Choi, C. R. Bealing, K. Bian, K. J. Hughes, W. Zhang, D.-M. Smilgies, R. G. Hennig, James R. Engstrom, and Tobias Hanrath. J. Am. Chem. Soc. 133, 3131 (2011).   

Scalable DAG-Based PDE Frameworks for Multi-Scale Multi-Physics Problems

The task-based approach to software and parallelism is well-known and has been proposed as a potential candidate, named the silver model, for exascale software. This approach is not yet widely used in the large-scale multi-core parallel computing of complex systems of partial differential equations. The central idea is to use a Directed Acyclic Graph (DAG) based approach to express the structure of the underlying software. The aim of this talk is to explore the usefulness of DAG based approaches, using recent developments in the parallel Uintah software framework for partial differential equations to assess how well the DAG type approach works on present-day large-scale architectures for complex multi-physics multiscale applications up to 200K cores. As a result of these investigations, a preliminary and tentative evaluation of the DAG type approach for PDE software infrastructures will be given. The conclusion is that these approaches show great promise for petascale but that considerable algorithmic challenges remain.   

Simulations of Strongly Correlated Systems

The field of strongly correlated systems is one of the most active areas in condensed matter physics. This interest is motivated, in part, by a variety of complex emergent phenomena, including high-temperature superconductivity, quantum criticality and complex phases induced by electron-phonon couplings. The recent very rapid development of high performance heterogeneous computer platforms together with a similar emergence of highly accurate many-body algorithms allow the treatment and modeling of complex correlated material systems which were intractable just a few years ago. Important progress has been made by the development of finite size methods, including exact diagonalization and Quantum Monte Carlo techniques. However, due to the minus sign problem, these methods are limited to small lattice sizes. Another successful approach is the dynamical mean field approximation and its cluster extensions, which treat the local or short-ranged dynamical correlations exactly and non-local or long-ranged correlations in a mean field approximation. Due to the effective medium, the Fermion minus sign problem is much milder than that found in finite sized simulations. However, it is still the primary limitation of these methods. To address this problem, multiscale approaches are used which treat only the correlations at the shortest length scales with exact cluster solvers, intermediate length scales are treated using a diagrammatic approach, such as the parquet equations or the dual-fermion formalism and, the longest length scales are captured by the mean field. I will discuss how these new algorithms impact a few model systems including our understanding of quantum criticality in the Hubbard and Anderson model, new phases in ultracold quantum gases, spintronics materials and the role of electron-phonon interaction. I will conclude discussing recent algorithm redesign motivated by the evolution towards hybrid multicore architectures employing graphical processing unit (GPU).   

Accessing correlated electron motion on the attosecond timescale

In the last decade, there have been tremendous advances in the production of coherent ultrashort light pulses as short as 80 attoseconds (1 as = $10^{-18}$ s). The availability of these pulses has led to the development of the field of attosecond physics, which aims to follow and control electron motion on its natural timescale (1 atomic unit of time is about 24 attoseconds). One of the major goals of attosecond physics is to access correlated electron dynamics. This requires a description of the target system that goes beyond the commonly used single-active-electron approximation. The large bandwidth of ultrashort pulses and many-photon absorption in strong infrared fields make such a description extremely challenging. I will discuss our work on the full numerical solution of the two-electron Schr\"odinger equation for helium, which already displays rich correlation effects. I will focus on two applications: The first is attosecond streaking, in which temporal information about the photoionization process in an attosecond pulse is mapped into a momentum shift by a synchronized infrared pulse. This promises to give access to the Eisenbud-Wigner-Smith time delay of photoionization. I will discuss the additional effects that are induced by the infrared field, and how these have to be taken into account for attosecond streaking to fulfill its promise. I will then discuss the possibility of accessing two-electron wave packet dynamics in doubly excited states of helium by an attosecond pump-attosecond probe setup. Such experiments have been called the ``holy grail'' of attosecond physics and should come within reach in the near future. I will discuss our recent proposal of using two-photon absorption from a single pulse as a coherent reference wave, which can be used to increase the experimental signal by almost two orders of magnitude. This provides direct access to time-dependent observables (e.g., the distance between the two electrons) of the two-electron wave packet.