Discussion of Itamar

1
Discussion of Itamars talk, August 2008
T. Witten, University of Chicago
2
Responding to Itamars picture of glassy state
Itamars talk raises basic issues, challenges
conventional wisdom
What do we think of these challenges?
Summarize the distinctive aspects of Itamars
model
Points we probably agree with
Points we may not agree with
3
Basic strategy identified local defects
Main assertion simple motions of readily
identified local structure account for the
macroscopic relaxation
This strategy is accepted and successful in
dislocation plasticity in metal crystals
topological defect annealing in liquid crystals...
Itamars defects Voronoi cells labeled by their
number of edges and content
5, 6, 7 sides, big or small particle inside
Every real configuration (that counts) is
associated with a cell assignment
Itamars dynamics T_1 flips, as in foam cells
Every continuous motion of the particles
corresponds to a specific sequence of flips
ASSERTION simple equilibrium kinetics of these
cells by these flips ? glassy relaxation
4
Itamars Minimal model of how cells evolve
Obtain mean energy and volume of each cell type
from a real simulation
Evolve cell population using T_1 moves using
equilibrium kinetics
Tractable by conventional stat. mech.
Conclude certain liquidlike (7s, 5L) disappear
at low temperature (as observed)
Density quantitatively predicted
Connection to relaxation
observe that fluidity requires these liquidlike
cells
7s 5L? 0
Dynamic relaxation annihilation reaction
Activation energy to annihilate is argued
proportional to their separation
Thus low density ? super-Arrhenius slow
relaxation times
as observed in the simulation.
5
Meaning of Itamars strategy whats at stake?
Do locally determined objects and motions explain
glassy relaxation as they explain relaxation of
other systems?
Why not? It seems to work.
6
Things we probably agree with
Itamars choice of defects and moves is plausible
(though other choices seem possible)
Ergodicity Itamar argues that after sufficient
time even a glass of finite size is ergodic.
Not in dispute (argument places no limit on the
time required).
Vogel-Fulcher temperature is not a real freezing
of all relaxation
Entropy does not vanish at a nonzero temperature.
Itamars minimal model if executed numerically
could well give long, relaxation like the
microscopic simulation.
7
Things we may not agree with
That Itamars choice of defects has microscopic
significance
Cf. dislocation lines T_1 flips in a foam
....
Other choices are also plausible eg. cells with
extra large or small volume.
Such cells would also go away at low temperature.
Do the cells have a robust identity?
Are the variations in energy within one type
smaller than the difference between types?
Are T1 flips the rate-limiting local steps as in
a foam?
Activation barrier? Many attempts before
successful T1?
Do they explain spatial heterogeneity?
(This feature offers hope to overcome the
ambiguity of global relaxation measurements)
Do liquidlike cells proliferate in regions
observed to be mobile?
8
(No Transcript)
9
??????????????????????????????? ??????????????????
???????? ?????? ?????????????????????????? ???????
???????? ????????????????????? ???? ?????????????
?????????? ??????? ??????????????????????????