Integrated
Tools
for Computational Chemical Dynamics
Overview
The goal of this
research program
is to develop more powerful simulation methods and incorporate them
into a
user-friendly high-throughput integrated software suite for chemical
dynamics.
It is supported by the Office of Naval Research (ONR) for 2005-2009.
This is a joint grant
between the
Chemistry Department and the Supercomputing Institute of the University of Minnesota (PI: Donald. G. Truhlar) and Pacific Northwest National Laboratory
(co-PI: Bruce C.
Garrett).
· The University of Minnesota faculty investigators: Donald. G. Truhlar, J. Ilja Siepmann, Christopher Cramer, Jiali Gao, and Darrin York
· The Pacific Northwest National Laboratory investigators: Bruce C. Garrett, Theresa Windus, Michel Dupuis, Shawn Kathmann, Greg Schenter, and Marat Valiev
This research program consists of three subprojects:
· Computational Electrochemistry
· Computational Photochemistry
Currently,
the project manager is Alek
Marenich.
Background
Photochemical
creation of
excited states offers a means to control chemical transformations
because different
wavelengths of light create different vibronic states, thereby
directing
chemical reactions along different pathways. It is crucial to
understand how
energy deposited into the system is used; this is particularly
complicated in
condensed-phase systems where many channels lead to dissipation of
excess
energy. Similar opportunities and challenges present themselves in the
areas of
electrochemistry and catalysis.
A
deeper insight for these critical problems relies on our accurate
knowledge
about chemical equilibria and kinetics in the complex systems. Thanks
to the
recent advances in computer power and algorithms, we can now calculate
accurately a large variety of chemical properties for many systems.
Nonetheless, applications to complex chemical systems, such as reactive
processes in the condensed phase, remain problematic. The problem is
due to the
lack of a seamless integration of computational methods that allow
modern
quantum electronic structure calculations to be performed with
state-of-the-art
methods for electronic structure, chemical thermodynamics, and reactive
dynamics. The problems is often exacerbated by invalidated methods,
non-modular
and non-portable codes, and inadequate documentation that drastically
limit
software reliability, throughput, and ease of use.
This research
program takes advantage of new computer hardware and algorithms and
develops
computational chemistry software for scientific discovery by modeling
and
simulation. This includes tools for calculating reaction rate and
transport
properties, and for investigating energetic materials, catalysis, and
ultra
fast dynamics. A notable area of emphasis is multi-scale modeling. The
research
includes methods and algorithm development as well as software
integration. An
important program goal is to develop practical solutions and standards
for
interfacing computer programs developed in separate laboratories. We
will
extend software capability, hardware compatibility and parallel
efficiency,
automation and documentation. Computational chemistry codes are
becoming
increasingly complex, not only in the types of computation available,
but also
in the software architectures used in code development. Developing
appropriate
interfaces will enable chemists to leverage current research in
mathematics and
computer science without duplicating the efforts of experts in the
field.
Features
of the Program
Combining Quantum Mechanics
(QM) and Molecular Mechanics (MM)
Conventional force fields are typically not appropriate for treating bond making and breaking processes, which involve electronic structure changes and must be treated by quantum mechanics. A promising strategy to circumvent the computational restrictions of large-scale quantum electronic structure calculations is to apply combined QM and MM methods. In these methods, reactive molecules or reactive site are treated explicitly by QM, while the surrounding subsystem is modeled by MM. A key feature of the hybrid QM/MM approach is that it combines the accuracy of quantum mechanics for the region of primary chemical interest and the computational efficiency of molecular mechanics for the large surrounding subsystem. The QM/MM approach is a practical strategy in treating the complexity in the condensed phase.
Combining Highly Accurate
Potentials with Robust and Reliable Dynamical Algorithms
Prediction of highly accurate potential energy surfaces is the first step in understanding reactions and mechanisms. Robust and reliable dynamical algorithms are the second step. We will integrate accurate and scalable quantum chemistry methods with state-of-the-art dynamics codes to study the dynamics of chemical reactions. An integrated kinetics and electronic structure software will be especially convenient for multi-step kinetics.
Component-Based Software
Engineering
To
date, typical scientific software designs
make rigid assumptions regarding programming languages and data
structures,
frustrating software interoperability and scientific collaboration.
However, component-based
software engineering is an emerging approach to managing the increasing
complexity of scientific software. Component technology facilitates
code
interoperability and reuse, and proves extremely beneficial in
facilitating the
interfacing of the several complex simulation modules. We will adopt
this
methodology embodied in the Common Component Architecture (CCA) Forum,
and we
will take advantage of existing software packages as a starting point,
especially, but not exclusively, NWChem
and software developed at the University of Minnesota. NWChem
is a new generation of high-performance molecular modeling
software for parallel computing systems, and it is described at http://www.emsl.pnl.gov/docs/nwchem/nwchem.html.
Software developed at the University of Minnesota is very diverse, and
much of
it is described at http://comp.chem.umn.edu.
|
Updated:
March 23, 2008 |