A5054786937- Machine Learning in Materials Science
- Advanced Chemical Physics Studies
- Chemical Synthesis and Analysis
- Analytical Chemistry and Chromatography
- Various Chemistry Research Topics
- Catalysis and Oxidation Reactions
- Electrocatalysts for Energy Conversion
- Catalytic Processes in Materials Science
- Surface Chemistry and Catalysis
- Advanced Condensed Matter Physics
- Protein Structure and Dynamics
- Physics of Superconductivity and Magnetism
- Catalysis and Hydrodesulfurization Studies
- Computational Drug Discovery Methods
- Ammonia Synthesis and Nitrogen Reduction
- Catalysts for Methane Reforming
- Advanced battery technologies research
- Business Process Modeling and Analysis
- Thermal Analysis in Power Transmission
- X-ray Diffraction in Crystallography
- Model Reduction and Neural Networks
- Crystallization and Solubility Studies
- Mathematical Control Systems and Analysis
- Power System Optimization and Stability
- Surface and Thin Film Phenomena
Sandia National Laboratories California
2019-2023
Sandia National Laboratories
2019-2023
University of Wisconsin–Madison
2014-2019
University of Minnesota
2012
The atomic simulation environment (ASE) is a software package written in the Python programming language with aim of setting up, steering, and analyzing atomistic simulations. In ASE, tasks are fully scripted Python. powerful syntax combined NumPy array library make it possible to perform very complex tasks. For example, sequence calculations may be performed use simple 'for-loop' construction. Calculations energy, forces, stresses other quantities through interfaces many external electronic...
Specialized computational chemistry packages have permanently reshaped the landscape of chemical and materials science by providing tools to support guide experimental efforts for prediction atomistic electronic properties. In this regard, structure played a special role using first-principle-driven methodologies model complex processes. Over past few decades, rapid development computing technologies tremendous increase in power offered unique chance study transformations sophisticated...
Decentralized on-site production of hydrogen peroxide (H2O2) relies on efficient, robust, and inexpensive electrocatalysts for the selective two-electron (2e–) oxygen reduction reaction (ORR). Here, we combine computations experiments to demonstrate that cobalt pyrite (CoS2), an earth-abundant transition-metal compound, is both active toward 2e– ORR in acidic solution. CoS2 nanomaterials drop-casted rotating ring-disk electrode (RRDE) showed efficient H2O2 formation 0.05 M H2SO4 at high...
We review the GPAW open-source Python package for electronic structure calculations. is based on projector-augmented wave method and can solve self-consistent density functional theory (DFT) equations using three different wave-function representations, namely real-space grids, plane waves, numerical atomic orbitals. The representations are complementary mutually independent be connected by transformations via grid. This multi-basis feature renders highly versatile unique among similar...
The kinetics of polymerization ε-caprolactone (CL) initiated by aluminum-alkoxide complexes supported the dianionic forms N,N-bis[methyl-(2-hydroxy-3-tert-butyl-5-R-phenyl)]-N,N-dimethylethylenediamines, (LR)Al(Oi-Pr) (R = OMe, Br, NO2) were studied. ligands are sterically similar but have variable electron donating characteristics due to differing remote (para) ligand substituents R. Saturation observed using [CL]0 2–2.6 M and [complex]0 7 mM, enabling independent determination substrate...
Identifying transition states-saddle points on the potential energy surface connecting reactant and product minima-is central to predicting kinetic barriers understanding chemical reaction mechanisms. In this work, we train a fully differentiable equivariant neural network potential, NewtonNet, thousands of organic reactions derive analytical Hessians. By reducing computational cost by several orders magnitude relative density functional theory (DFT) ab initio source, can afford use learned...
We present a new algorithm for the optimization of molecular structures to saddle points on potential energy surface using redundant internal coordinate system. This automates procedure defining system, including handling linear bending angles, example, through addition dummy atoms. Additionally, supports constrained null-space sequential quadratic programming formalism. Our determines direction reaction iterative diagonalization Hessian matrix and does not require evaluation full matrix....
Identification and refinement of first order saddle point (FOSP) structures on the potential energy surface (PES) chemical systems is a computational bottleneck in characterization reaction pathways. Leading FOSP strategies for modestly sized molecular require calculation full Hessian matrix, which not feasible larger such as those encountered heterogeneous catalysis. For these systems, standard approach to involves iterative diagonalization Hessian, but this comes at cost longer...
We have developed a flexible, general-purpose microkinetic modeling code, Micki, to analyze complex, heterogeneously catalyzed chemical reactions based upon first-principles calculations. This Python-based code is modular and object oriented, framing the development of models in familiar terms. also present novel approaches, incorporated into describe diffusion limited reactions, multidentate bindings, thermodynamically consistent lateral interactions, Brønsted-Evans-Polanyi estimates...
We examine the role of oxide support in modulating activity and selectivity common oxide-supported transition metal heterogeneous catalysts. Using representative models various supported nanoparticle systems, we use computational studies based on density functional theory to decouple metal–support interaction terms electronic geometric effects, including oxygen vacancies reactivity perimeter sites at metal–oxide boundary. target Fischer–Tropsch conversion silica- titania-supported Rh as a...
Many important industrial processes rely on heterogeneous catalytic systems. However, given all possible catalysts and conditions of interest, it is impractical to optimize most systems experimentally. Automatically generated microkinetic models can be used efficiently consider many conditions. these require accurate estimation thermochemical kinetic parameters. Manually calculating parameters tedious error prone, involving interconnected computations. We present Pynta, a workflow software...
We present an ab initio microkinetic model for the oxidative esterification of 1-propanol to methyl propionate over Pd(111). The fully accounts solvation solution-phase species and added catalytic base provides key insights into factors that limit activity unpromoted Pd aerobic oxidation catalysts. In particular, we find is limited by large steady-state surface H coverage, which destabilizes other adsorbed intermediates via lateral interactions, substantial barriers governing formation O–H...
We present a new geodesic-based method for geometry optimization in basis set of redundant internal coordinates. Our updates the molecular by following geodesic generated displacement vector on coordinate manifold, which dramatically reduces number steps required to converge minimum. can be implemented any existing code, requiring only implementation derivatives Wilson B-matrix and ability numerically solve an ordinary differential equation.
Many important industrial processes rely on heterogeneous catalytic systems. However, given all possible catalysts and conditions of interest it is impractical to optimize most systems experimentally. Automatically generated microkinetic models can be used efficiently consider many conditions. these require accurate estimation thermochemical kinetic parameters. Manually calculating parameters tedious error prone involving interconnected computations. We present Pynta, a workflow software for...
Abstract An efficient and accurate approach for calculating exact exchange other two-electron integrals has been developed periodic electronic structure methods. Traditional approaches used integrating over the Brillouin zone in band calculations, e.g. trapezoidal or Monkhorst-Pack, are not enough integrals. This is because their integrands contain multiple singularities double integration of zone, which with simple methods lead to very inaccurate results. A common this problem replace...
Identification and refinement of first order saddle point (FOSP) structures on the potential energy surface (PES) chemical systems is a computational bottleneck in characterization reaction pathways. Leading FOSP strategies require calculation full Hessian matrix, which not feasible for larger such as those encountered heterogeneous catalysis. For these systems, standard approach to involves iterative diagonalization Hessian, but this comes at cost longer trajectories due lack accurate...
We review the GPAW open-source Python package for electronic structure calculations. is based on projector-augmented wave method and can solve self-consistent density functional theory (DFT) equations using three different wave-function representations, namely real-space grids, plane waves, numerical atomic orbitals. The representations are complementary mutually independent be connected by transformations via grid. This multi-basis feature renders highly versatile unique among similar...
Identifying transition states -- saddle points on the potential energy surface connecting reactant and product minima is central to predicting kinetic barriers understanding chemical reaction mechanisms. In this work, we train an equivariant neural network potential, NewtonNet, ab initio dataset of thousands organic reactions from which derive analytical Hessians fully differentiable machine learning (ML) model. By reducing computational cost by several orders magnitude relative Density...
We present a new geodesic-based method for geometry optimization in basis of redundant internal coordinates. Our updates the molecular by following geodesic generated displacement vector on coordinate manifold, which dramatically reduces number steps required to reach convergence. can be implemented any existing code, requiring only implementation derivatives Wilson B-matrix and ability solve an ordinary differential equation.
We present a new algorithm for the optimization of molecular structures to saddle points on potential energy surface using redundant internal coordinate system. This automates procedure defining system, including handling linear bending angles, e.g. through addition dummy atoms. Additionally, supports constrained null-space sequential quadratic programming formalism. Our determines direction reaction iterative diagonalization Hessian matrix, and does not require evaluation full matrix....
We present a new geodesic-based method for geometry optimization in basis of redundant internal coordinates. Our updates the molecular by following geodesic generated displacement vector on coordinate manifold, which dramatically reduces number steps required to converge minimum. can be implemented any existing code, requiring only implementation derivatives Wilson B-matrix and ability numerically solve an ordinary differential equation.
Identification and refinement of first order saddle point (FOSP) structures on the potential energy surface (PES) chemical systems is a computational bottleneck in characterization reaction pathways. Leading FOSP strategies require calculation full Hessian matrix, which not feasible for larger such as those encountered heterogeneous catalysis. For these systems, standard approach to involves iterative diagonalization Hessian, but this comes at cost longer trajectories due lack accurate...
discrete variable representation and the phase space integration results confirm that motion of H on Cu(111) is highly anharmonic. The were applied to calculate free energy dissociative adsorption H2 resulting Langmuir isotherms at 400, 800, 1200 K in a partial pressure range 0–1 bar. It shows anharmonic effects lead significantly higher predicted surface site fractions hydrogen.
We present a new algorithm for the optimization of molecular structures to saddle points on potential energy surface using redundant internal coordinate system. This automates procedure defining system, including handling linear bending angles, e.g. through addition dummy atoms. Additionally, supports constrained null-space sequential quadratic programming formalism. Our determines direction reaction iterative diagonalization Hessian matrix, and does not require evaluation full matrix....