Computing Related User Projects

• Bio-Inspired Actinides Recognition for Separation Science
  Project Lead: Ping Yang
  Project Lead Institution: Environmental Molecular Sciences Laboratory
  Abstract
  A critically important problem facing the US is the issue of spent nuclear fuel disposal. The grand challenge here lies in the design of an efficient fuel cycle. The development of chelating techniques for capturing and separating actinide species has the potential to address a number of current problems ranging from reprocessing of spent nuclear fuel to environmental remediation. I propose to build a fundamental theoretical understanding of the physico-chemical interactions of binding motifs for actinides and to use that understanding to rationally design chelating reagents with high selectivity to actinides, to allow for better reprocessing of spent nuclear fuels, reducing the need for long-term waste storage, and closing of the fuel cycle. The research will be based on two main thrusts: the thermo- structural investigation of the actinides/bio-ligands complexes, and the tuning of the redox potentials of actinides compounds through interactions with ligands. To unravel the nature of chemical bonding in f- block elements and gain complete control of chemical selectivity in complex environment, the research will focus on characterizing the interactions between actinides and selected biomolecules using first principle methods that include relativistic effects, spin-orbit coupling and multiplet complexity.
• Understanding DNA surface interactions for systems design and engineering
  Project Lead: Bernard Pettitt
  Project Lead Institution: University of Texas Medical Branch
  Abstract
  DNA replication, DNA modification, and single organism engineering are being used in many fields of relevance to the DOE mission. Extensive knowledge of the fundamental reactions involved in these processes will continue to aid the design and engineering processes involved with biofuels and geobacter remediation strategies. This project will examine two fundamental and critical aspects of DNA chemistry and physics; DNA-protein interactions and the ability of proteins to quickly locate their DNA target, and DNA folding in a small volume. First, proteins and DNA interact in many biological functions. How do these two molecules locate the specific site needed for the correct protein-DNA complex formation? Secondly, the mechanism, used in nature, to pack a highly-negatively charge molecule, such as DNA, into a confined, biologically relevant space. When a protein binds to a DNA molecule, necessary for function, the interaction can be a specific or a non-specific interaction. The mechanisms governing protein-DNA recognition and the transition of proteins from their unfolded state to their native state are unanswered. Refolding often occurs in DNA binding. Multiscale simulations including molecular dynamics simulations coupled with Brownian dynamics will be performed to understand the encounter and complex formation. Details of the interfacial properties of the sequence-dependent protein-DNA landscape will also be calculated. Additionally, recent experimental data from one of our collaborators has indicated that specific amino acid residues, in particular lysine and arginine, appear to play a more important role in the protein-DNA dynamic location/recognition process. Using our 2 sets of DNA/protein simulations, this rarely investigated amino acid effect on dynamics will also be analyzed. Understanding the biophysical basis of the biological process which transfers a viral genome to infect a cell is important to the cellular machinery and many bacteria engineering related fields. Predicting the thermodynamic pressures including the osmotic pressure necessary to confine certain sequences of DNA in a specific volume, like a phage capsids with over 250-fold compaction, is a problem with implications in bionanotechnology, and phage delivery of genetic messages. We will resolve questions of the thermodynamic mechanism of DNA ejection by phages by designing coarse-grained models in conjunction with using detailed all-atom calculations and experimental thermodynamic and structural data.
• Computational Simulations of Antimicrobial Polymeric Surfaces and Biological Membranes
  Project Lead: Thereza Soares Da Silva
  Project Lead Institution: Universidade Federal de Pernambuco
  Abstract
  Medical devices can give rise to treatment-resistant infections upon colonization by pathogenous bacteria. A detailed understanding of the interactions between biological cells and the device surface is fundamental for the development of novel materials with anti-bacterial properties. Such understand relies on the accurate description of the physicochemical and structural properties of the cell and the substrate surfaces. We are currently studying these interactions both experimentally and through computational simulations in order to obtain a molecular level understanding of the processes that influence the attachment or repulsion between a cell and a surface. However, a detailed molecular representation of biological membranes, in special bacterial outer membranes, requires long (0.5-0.6 ?s) and large (200.000 atoms) computational simulations only possible through the use of supercomputers such as Chinook at the Environmental Molecular Sciences Laboratory (EMSL). The outcome of the present research can offer important insights into the driving forces underling bacterial attachment to polymeric surfaces and the interactions of cationic antimicrobial peptides with the outer membrane of Gram-negative bacteria, enabling the design of surfaces that repel bacteria but do not harm mammalian cells in their proximity as well as the design of cAMPs with higher pathogen-specificity and decrease host-toxicity.
• Integrating genome-scale metabolic modeling of a microbial community into simulations of subsurface flow and biogeochemical reactive transport.
  Project Lead: Steven Yabusaki
  Project Lead Institution: Pacific Northwest National Laboratory
  Abstract
  The goal of this project is to develop a mechanistic, quantitative, and predictive simulation capability for coupled flow and biogeochemical reactive transport that accounts for • genome-specific networks of metabolic reactions (“in silico” models) for microbial species (Geobacter metallireducens (Sun et al., 2009), Geobacter bemedjiensis, Desulfobacter postgatei) • single reaction TEAPs (Fe(III), U(VI), sulfate) and thermodynamically-constrained Monod-type rate laws for microbial guilds representing specific ecological niches (Yabusaki et al., 2011) • multicomponent geochemical reaction networks for major ion chemistry, surface complexation (H+, Fe(II), U(VI)), ion exchange (Ca++, Mg++, K+, N+), mineral reactions (goethite, FeS, calcite, siderite) (Fang et al., 2009a) • dynamic water table and hydraulic gradients in a physically and geochemically heterogeneous, variably-saturated aquifer (e.g., Yabusaki et al. (2011)). A key component of this project is the use of proteomic data from groundwater samples to assess and refine the magnitude of individual intracellular reaction fluxes predicted by the fundamentally detailed in silico models (> 700 intracellular and exchange reactions per species). Advanced computing is necessary to provide the high performance and large memory to simulate the comprehensively detailed models of coupled processes (e.g., geology, hydrology, biology, chemistry) in the context of three-dimensional multiscale variability in material properties. eSTOMP, a massively parallel processing multifluid flow and biogeochemical reactive transport subsurface simulator is the principal enabling technology that addresses the aforementioned complexity of process and property detail during field experiments in the Rifle (Colorado) IFRC flood plain aquifer. The potential impact of this approach is the engineering of electron donor (e.g., acetate), terminal electron acceptor [e.g., U(VI)], nutrient and/or biogeochemical conditions that enhance metabolic pathways of target microorganism(s) to effect desirable biological transformations, such as reduction in greenhouse gas production and atmospheric release, contaminant destruction and/or immobbilization.
• in Operando Characterization of Contact Catalysts at Mesoscale Distances
  Project Lead: Robert Weber
  Project Lead Institution: Pacific Northwest National Laboratory
  Abstract
  We propose to measure and model the dipole-dipole interactions that shift and quench luminescence spectra of ligand-to-metal charge transfer bands of contact catalysts containing transition metal in order to quantify inter-site geometries and the polarizability of the reaction medium within 1-10 nm of the active sites. The former is needed to refine the syntheses that lead to well defined catalysts and to track their evolution. The latter is needed to understand molecular transport to and from the catalyst surface. We will develop the use of luminescence lifetime spectroscopy to probe the composition and geometry in that regime under conditions that could be employed to characterize catalysts under reaction conditions.The work relies on expertise resident at EMSL (spectroscopy) and PNNL (synthesis, modeling) with EMSL’s advanced facilities for spectroscopy (femtosecond LIF) and quantum chemical modeling (NWChem, CP2k running on large scale, parallel computers).
• Fundamental Studies of Vehicle Emission Control Catalysts
  Project Lead: Charles Peden
  Project Lead Institution: Pacific Northwest National Laboratory
  Abstract
  The abatement of environmentally harmful NOx compounds (NO, NO2, and N2O) emitted from mobile or stationary power sources remains a challenging task for the catalysis community. In particular, conventional three-way catalysts used in the exhaust after treatment technologies of internal combustion engines prove ineffective when the engine is operated under highly oxidizing conditions (to achieve better fuel efficiency). The problem is daunting, since reduction chemistry (NOx to N2) has to be carried out under highly oxidizing conditions. Several approaches have been proposed for lean-NOx abatement, each of them with its own specific sets of problems. The two technologies that seem to have clear advantages among the processes proposed are the selective catalytic reduction either with hydrocarbons (HC-SCR) or with ammonia (NH3-SCR), and NOx storage reduction (NSR). For the NH3-SCR technology, transition metal (in particular Fe and Cu) ion-exchanged zeolite catalysts have shown high activity and N2 selectivity. NSR catalysts work under cyclic operation conditions: oxidation of NO to NO2 followed by storage as nitrates/nitrites on alkali or alkaline oxides under lean conditions, and subsequent reduction of the released NOx under rich conditions. To further advance these important exhaust emission technologies, we are carrying out several research programs, funded by DOE/Office of Energy Efficiency and Renewable Energy (EERE)/Vehicle Technologies Program (VT), involving both programs that include direct collaboration with industry partners (at GM, Ford and Cummins), and a more fundamental program. For all of these programs, we rely on the use of a wide array of state-of-the-art catalyst characterization facilities in the EMSL at PNNL.
• Engineered nano-scale interfaces in scandia stabilized zirconia and samaria doped ceria multi-layer hetero-structures
  Project Lead: Asghar Kayani
  Project Lead Institution: Western Michigan University
  Abstract
  Recent studies showed a colossal enhancement in the oxygen ionic conductivity of multi-layer oxide thin films compared to most commonly used solid oxide fuel cell (SOFC) electrolytes. It has been observed that, the ionic conductivity of multi-layer oxide thin films increases with the increase in number of layers. The enhancement in the ionic conductivity of these multi-layer thin films was attributed to interfacial conductivity due to the lattice strain and extended defects near the interface. In the proposed study, we will investigate the influence of nano-scale interfaces on oxygen ionic conductivity in samaria doped ceria/ scandia stabilized zirconia (SDC/ScSZ) multi-layer thin films. In this study, the epitaxial SDC/ScSZ multi-layer thin films wiil be grown on Al2O3(0001) substrates by oxygen plasma-assisted molecular beam epitaxy (OPA-MBE) capability in EMSL. The growth of SDC/ScSZ multi-layer thin films will be carried out using the optimized growth conditions, dopant concentrations and film properties established for single-layer SDC and ScSZ epitaxial thin films in the previous studies conducted at EMSL. The number of layers in the multi-layer thin films will be varied by keeping the total film thickness constant to understand the influence of individual layer thickness on the ionic conductivity. The structure-property relationship of these multi-layer hetero-structures will be investigated using advanced spectroscopy, diffraction, and microscopy capabilities in EMSL. Oxygen ionic conductivity measurements through the films will be carried out as a function of temperature using four probe surface impedance spectroscopy. Theoretical studies will also be carried out using molecular dynamics simulations in conjunction with the experimental studies.
• The Mechanism of the Hammerhead Ribozyme RNA Cleavage Reaction: Determining Optimized Free Energy Reaction Paths from QM/MM Simulations
  Project Lead: John Weare
  Project Lead Institution: University of California, San Diego
  Abstract
  Ribonucleic acids (RNA) play a central role in the transmission of genetic information. However, the identification of additional non-coding roles of RNA polymers continues to grow. An important class of these behaviors is the discovery of the catalytic function of RNA molecules. A well studied example is the hammerhead ribozyme (HHR), which facilitates the self cleavage of RNA during rolling circle replication of plant viroids. We are requesting support to utilize the Chinook supercomputer to develop a well validated, theoretically based reaction mechanism for the self-cleavage reaction of HHR. Our goal is to produce a minimum free energy path from the reactant (RS) to the product (PS) state that is consistent with all the existing biochemical and structural data for this well studied ribozyme, is based on reliable 1st principle electronic structure calculations, and incorporates as complete an exploration of the possible reaction mechanisms as feasible. Recently the publication of a high resolution X-ray structure of an inactive analog of the S.mansoni HHR reactive state has reconciled many discrepancies between biochemical observations and structural studies of the enzyme-substrate complex. However, many questions concerning the detailed chemical mechanism of this reaction cannot be answered from structural data from an inactive enzyme complex. There are additional experimental probes, pH variation, thio modification rescue experiments, mutagenesis and chemical modification experiments. As yet these diverse data have not been correlated by a consistent reaction mechanism calculated using 1st principle methods that will support prediction for the chemical diversity of the experimental species involved. Theoretical interpretations have been reported. However, these calculations are based on QM/MM models in which the quantum region was treated with semiempirical methods and in addition certain as yet only weakly justified assumptions about the reaction mechanism. For example, a specific general acid/base pair (as yet uncertain from observation) was selected for the calculation and initial proton transfer from the assumed nuclophile to the assumed general base is taken as the initial structure. As arguably the best understood ribozyme (best structural and biochemical information available) this self-cleavage reaction provides an promising target for complete free energy analysis. Our objective is to identify a full reaction path using: high level 1st principle simulation techniques (calculations at the B3LYP level); implementation of sampling methods (metadynamics and temperature-accelerated molecular dynamics) to identify candidate structures and reaction paths: nudged elastic band and string methods to optimize reaction paths; and finite temperature free energy simulations allowing the full structural relaxation of enzyme substrate complex along the reaction path. Since similar residue motifs have been identified in other ribozymes the understanding of the HHR system would support the successful application of theoretical interpretations of this mechanism would lead to a more insight into the chemistry of other ribozyme and to a deeper understanding of other phosphoryl transfer reactions in protein based enzymes. The team assembled to carry out this project contains experts in the development and application of 1st principle methods (Valiev, Elaesser, Weare, and Pirojsirikul (graduate student)) and experts in ribozyme chemistry (Muller, Elaesser).
• Interfacial and Multiphase Characterization of Next Generation Inorganic Resists Synthesized Through Sustainable Materials Chemistries
  Project Lead: Liney Arnadottir
  Project Lead Institution: Oregon State University
  Abstract
  Transformation of small molecular precursors into inorganic solid-state compounds via solution methods has mainly focused on sol-gel techniques involving hydrolysis and condensation reactions of metal-organic compounds in bulk solutions. This method generally produces porous structures because of the difficulties in controlling the hydrolysis and condensation reactions leading to sol-gel formation. As part of the National Science Foundation Center for Sustainable Materials Chemistry (CSMC), we have undertaken a revolutionary approach to forming high-quality inorganic multicomponent films where we trap reactive precursors in an aqueous-based thin film and then induce controlled condensation. Using these approaches we have synthesized smooth, fully dense, crack-free, high-quality films of many compositions with controlled thickness down to 8 Å, including HfO2, ZrO2, HfO2-x(SO4)x, ZrO2-x(SO4)x, and Al4O3(PO4)2. The functionality of the resulting materials has been borne out through fabrication of high-performance thin-film transistors and displays,[1,2] dielectric mirrors and band-pass filters,[3] diffraction and light-trapping gratings,[4,5] and leading-edge directly imaged nanostructures via electron-beam and extreme ultra-violet lithographies.[6,7] The condensation reactions for these films can be initiated through thermal, photon, and electron stimulated processes which has lead to advances in nanolaminate fabrication (with single-digit-nm layer thicknesses) and direct sub-10-nm nanopatterning. The objective of this proposal is to develop a fundamental understanding of interfacial reaction mechanisms for film formation, interdiffusion between nanolaminate layers, and the role of precursor chemistries on nanopatterning fidelity for next generation photoresists. This proposal is well aligned with the EMSL's Science Theme on the “Science of Interfacial Phenomena,” and the EMSL capabilities will allow in-situ integrated approaches that will combine both experimental and theoretical methods to advance sustainable materials chemistries relevant to energy efficient functional materials synthesis.
• First-principles modeling of charge transfer in fullerene organic photovoltaics
  Project Lead: Daniel Neuhauser
  Project Lead Institution: University of California, Los Angeles
  Abstract
  Organic photovoltaics have emerged as an inexpensive alternative to more traditional solar energy. Organic solar cells have long utilized fullerene derivatives as electron acceptors in devices, little is known about what specifically makes particular fullerenes candidates for devices than others. We therefore seek to calculate electron transfer rates in a large number of fullerene derivatives, the majority of which have merely been proposed, to determine which types of fullerenes make ideal devices. One particular difficulty in calculating transfer rates in fullerenes is that the asymmetry in dimer systems typically induces an artificial bias in the system due to the fact that there are only two fragments. To overcome this difficulty, we have developed a method whereby we apply a field to the DFT-generated Fock matrix to eliminate the bias and mimic the bulk. We are then able to calculate the transfer rates using Marcus theory. The main bottleneck in these calculations is generating the Fock and overlap matrices from DFT. Due to the size of the systems, over 200 atoms and several thousand basis functions, DFT calculations are very expensive. Because we seek to scan a very large number of molecules in a high-throughput fashion, using the NWChem on Chinook is essential. We will both to apply our method to get a better understanding of what makes particular solar cells better than others, as well as use it to predict which fullerenes will perform better in actual devices.