Molecular Foundry Seminar

"What Happens to Crystals When Thermodynamics Breaks Down?"

Dr. Jim DeYoreo, Deputy Director for Research, Molecular Foundry,

Tuesday, June 22nd at 1:30 pm, Bldg. 67-Room 3111

Abstract:

Growth of crystals from solution is a key process in biological,
geochemical and technological settings. The classical
“terrace-ledge-kink” model of crystal growth is widely used to
interpret growth in these settings, particularly near ambient
conditions. A key underlying assumption of this model is that the
system satisfies the Fluctuation Dissipation Theorem. That is,
thermal fluctuations of atomic steps on surfaces are sufficiently
rapid to produce an abundance of kink sites for attachment of growth
units. High-resolution in situ AFM studies and kinetic Monte Carlo
simulations of step-edge structure and dynamics show this physical
picture to be invalid for some common crystal systems whose steps
exhibit low kink density and weak fluctuations. As a consequence,
system behavior cannot be interpreted with traditional thermodynamic
models based on minimization of the Gibbs free energy. Instead step
morphology, growth dynamics and impurity interactions follow a
different mechanism determined by the kinetics of attachment and
detachment. For example, growth inhibition by impurities that bind to
the step is no longer the result of step curvature between pinning
sites, which is an ensemble characteristic described by the
Gibbs-Thomson effect. Rather it is due to the competition between
rates of kink creation through solute attachment to steps and impurity
attachment to newly created kinks. This kink-limited model explains
many anomalous features of crystal growth that have recently been
reported and offers a plausible explanation for reports of ‘kinetic
disequilibrium’ of trace element signatures. Moreover, because kink
density is tied to crystal solubility, these findings argue for a
theory based on weak fluctuations to interpret growth of many common
crystalline phases of importance in geochemical and biological
settings. More generally, except in cases where a step is atomically
rough, one can always define a system size that falls below that
average spacing between fluctuations for a given time scale. This
statement should also apply to other types of fluctuations in
nanoscale systems, including those that eliminate thermodynamic
imbalances caused by voltage differences, stresses, or temperature
gradients. The implication is that nanoscale systems in general, not
just those involving crystallization, may require a theory of weak
fluctuations to describe their behavior.