Complex Energy Landscapes and Rare Events
The dynamics of complex systems is ofter driven by rare but important events. These are the events that
happen infrequently compared with the relaxation time scale of the system. Chemical reactions,
conformational changes of biomolecules, nucleation events in phase transitions, and in some
cases extreme events that lead to material failure are all examples of rare events.
The long time scale associated with these rare events is a consequence of the disparity between
the effective thermal energy and typical energy barrier of the systems. The dynamics proceeds
by long waiting periods around metastable states followed by sudden jumps from one state to another.
The main questions of interest are the mechanism of such transition and the transition rates.
I and collaborators developed
theoretical framework and numerical methods (the zero and finitetemperature string method) for
dealing with such rare events in complex systems.
Smooth energy landscapes and the zerotemperature string method.
For systems with simple energy landscape in which the metastable states are separated by a few
isolated barriers, the key objects are the transition states, which are saddle points on the potential
energy landscape that separate the metastable states.
These saddle points act as bottlenecks in a particular transition.
The relevant notion for the transition pathways
is that of minimum energy paths (MEPs). MEPs are paths
in the configuration space that connects the metastable
states along which the potential force is parallel to the tangent vector.
The MEP allows us to identify the relevant saddle points,
as well as the unstable directions at these points
which are needed for the calculation of the prefactor in the transition rates.
The zerotemperature string method (ZTS) is an effective tool for computing MEPs.
Starting with an initial string, which is a curve in the configuration space
connecting two metastable states,
the method finds the MEP by repeating a twostep procedure (steepest descent and reparametrization)
until convergence. See the following references for more details.


Fig 1. An example of the smooth energy landscape and the minimum energy path.

 A climbing string method for saddle point search,
W. Ren and E. VandenEijnden,
J. Chem. Phys. 138, 134105 (2013)
 Simplified and improved string method for computing the minimum energy paths
in barriercrossing events,
W. E, W. Ren and E. VandenEijnden,
J. Chem. Phys. 126, 164103 (2007)
 Higher order numerical scheme
in the string method for finding minimum energy paths and saddle points,
W. Ren, Commun. Math. Sci. 1, 377 (2003)
 String method for the study of rare events,
W. E, W Ren, and E. VandenEijnden, Phys. Rev. B 66, 052301 (2002)
Click HERE to download the fortran code for the original string method.
Rough energy landscapes and the finitetemperature string method.
The problem of identifying transition pathways becomes much more challenging for
systems with rough energy landscapes, as are the cases for typical chemical reactions of
solvated systems and conformational changes of macromolecules.
In these situations,
the traditional notion of transition states becomes inappropriate since
the potential energy landscape typically contains numerous saddle points,
most of which are separated by barriers that are less than or comparable to k _{B}T,
and therefore do not act as barriers.
There may not exist specific microscopic configurations that identify the bottleneck of the transition.
The most probable transition path is not unique; instead, a collection of paths are important.
The finitetemperature string method (FTS)
aims at solving problems of this type.
The central objects here are the transition tube and the transition state ensemble ,
which are generalizations of the concepts of the MEP and the transition state.
The FTS is designed to effectively solve the backward Kolmogorov equation for the committor function
in the highdimensional configuration space.
Specifically, it determines a tube (the transition tube) by which transitions occur with high probability,
and a family of hyperplanes within the tube which are approximations to the isocommittor surfaces.
These are done by evolving a smooth curve using an averaged potential force in the configuration space.
This curve converges to the center of the tube; the hyperplanes perpendicular to the converged
curve are approximations to the isocommittor surfaces.
The isocommittor surface with the value 1/2 and the Gibbs density function restricted to this
surface define the transition state ensemble.


Fig 2. An example of the rough energy landscape (upper) and the transition tube (lower)
identified by the string method.

 Transition pathways in complex systems: Application of the
finitetemperature string method to the alanine dipeptide,
W. Ren, E. VandenEijnden, P. Maragakis, and W. E,
J. Chem. Phys. 123, 134109 (2005)
 Transition pathways in complex systems: Reaction coordinates,
isocommittor surfaces, and transition Tubes
W. E, W. Ren, and E. VandenEijnden,
Chem. Phys. Lett. 413, 242 (2005)
 Finite temperature string method for the study of rare events,
W. E, W. Ren, and E. VandenEijnden,
J. Phys. Chem. B 109, 6688 (2005)
Transition pathways in nongradient systems and finitetime switching.
The WentzellFreidlin theory of large deviations provides a theoretical foundation
for identifying the transition pathways associated with a dynamical system perturbed by
a small noise. In particular,
it gives an estimate on the probability of the transition paths in terms of an action functional.
This estimate shows that the most probable transition pathway can be obtained by the
(constrained) minimization of the action.
This path is called the minimum action path (MAP).
The minimum action method is an effective tool for computing the MAP for a given switching time.
The MAP reduces to the MEP for gradient systems and infinite switching time.
 Noiseinduced transition in barotropic flow over topography and application to Kuroshio,
W. Yao and W. Ren, J. Comput. Phys. 300, 352 (2015)
 Minimum action method for the KardarParisiZhang equation,
H. C. Fogedby and W. Ren, Phys. Rev. E 80, 041116 (2009)
 Adaptive minimum action method for the study of rare events,
X. Zhou, W. Ren and W. E, J. Chem. Phys. 128, 104111 (2008)
 Minimal action method for the study of rare events,
W. E, W. Ren and E. VandenEijnden,
Commun. Pure Appl. Math. 57, 637 (2004)
Some applications of the string method.
The string method has been applied to a variety of problems arising
from different disciplines. These include the study of the energy landscape of ferromagnetic
thin films, conformational changes of biomolecules,
current dissipation in thin superconducting wires,
dislocation motion in crystalline solids, and
quantum metastability
induced by tunneling.
 Numerical study of the effects of surface topography and chemistry on the wetting transition using the string method",
Y. Zhang and W. Ren,
J. Chem. Phys., 141, 244705 (2014)
 Numerical study of vapor condensation on patterned hydrophobic surfaces using the string method,
Y. Li and W. Ren,
Langmuir, 30, 9567 (2014)
 Wetting transition on patterned surfaces: transition states and energy barriers,
W. Ren,
Langmuir, 30, 2879 (2014)
 Computing transition rates of thermally activated events in dislocation
dynamics,
C. Jin, W. Ren and Y. Xiang,
Scripta Materialia, 62, 206 (2010)
 Application of the string method to the study of critical nulei in capillary condensation,
C. Qiu, T. Qian and W. Ren,
J. Chem. Phys., 129, 154711 (2008)
 Phase slips in superconducting wires with nonuniform cross section: A numerical evaluation using the string method,
C. Qiu, T. Qian and W. Ren,
Phys. Rev. B, 77, 104516 (2008)
 Numerical study of metastability due to tunneling: The quantum string method,
T. Qian, W. Ren, J. Shi, W. E and P. Sheng,
Physica A, 379, 491 (2007)
 Current dissipation in thin superconducting wires:
Accurate numerical evaluation using the string method,
T. Qian, W. Ren and P. Sheng,
Phys. Rev. B, 72, 014512 (2005)
 Energy landscape and thermally activated switching of
submicronsized ferromagnetic elements,
W. E, W. Ren and E. VandenEijnden, J. Appl. Phys. 93, 2275 (2003)
Numerical examples (to be available soon!)
 Thermally activated switching of magnetization in ferromagnetic thin films
 Martensitic phase transformation
 Finitetime switching of a GinzburgLandau system
WEIQING REN
Last modified: Thur Oct 15 21:17:06 EST 2015
