Hi All,
Don't forget that this week there is a second Theochem seminar, tomorrow
(Wednesday 9/26) at 4pm at MIT. The full ad is below, but Prof.
Hernandez is also coming to visit Harvard on Thursday. If you would
like to meet with him, or join us for lunch, let me know!
Greater Boston Theoretical Chemistry Seminar
Professor Rigoberto Hernandez
Title: Chemical reaction dynamics in complex environments
Location: MIT, Building 4 - Room 163
Time: 4:00 PM, Wednesday ( 9/26/12 )
Abstract:
In the first hour, we will discuss the fundamentals of reaction rate
theory in the gas and
liquid phases.Transition state theory has proven to be a powerful tool
for describing the
leading order terms in reaction rates. At the cutting edge, we and many
others are
working to increase its accuracy particularly in cases involving much higher
dimensionality either in the reacting system, the surrounding solvent or
both.
Meanwhile, the environments themselves may change over time leading to
changes in
the response ---as represented by colored friction kernels in a
generalized Langevin
equation--- or even wholesale changes ---as represented by
nonstationarity. Such
increasingly complex environments are commonplace in present-day
materials and the
theory is expanding to include them.
In the second hour, we will discuss our recent work characterizing the
potential of mean
force along a chosen path of a molecular motion. In particular, we have
developed
adaptive steered molecular dynamics (ASMD) in order to efficiently
obtain the potential
of mean force along a long-distance path. This relies on the convergence
of Jarzynski's
equality, which becomes increasingly computationally expensive as the
contributing
nonequilibrium paths spread father from the initial point. In ASMD, we
break up the path
into a series of segments, with each having a smaller distance along
which to spread.
Equally important, we use the information found in the nonequilibrium
distribution at the
end of each segment to select the initial structures for the subsequent
segment. We
have found that ASMD converges to the same and correct potential of mean
force in the
stretching of decaalanine in vacuum using substantially fewer computer
cycles and at
pulling speeds faster than that required to obtain it reversibly.
Decaalanine has also
been stretched in an aqueous solvent. The intramolecular
hydrogen-bonding in this
latter case appears to be entirely different than that seen in the
vacuum case,
suggesting the importance of including explicit water models at least in
the vicinity of the
internal hydrogen bonds of the protein.
Show replies by date