The Innovation Materials,
Journal Year:
2024,
Volume and Issue:
unknown, P. 100086 - 100086
Published: Jan. 1, 2024
<p>In
2005,
Science
magazine
listed
the
��nature
of
a
glassy
substance��
as
one
125
most
challenging
scientific
questions
century.
A
quantitative
understanding
time-temperature
transition
(TTT)
curve
for
critical
nucleation
amorphous
materials
is
crucial
to
answering
this
question.
Despite
extensive
efforts
over
past
70
years,
model
TTT
remains
elusive
due
lack
physical
properties
such
interfacial
energy
at
incubation
time
<i>t</i><sup>*</sup>
nucleation.
In
study,
relationship
between
viscosity
and
function
established
developed.
The
demonstrates
excellent
agreement
with
experimental
data
various
materials.
Most
importantly,
it
allows
accurate
definitive
determination
<i>T</i><sub>0</sub>,
true
minimum
crystallization
temperature
lower
end-point
curve,
well
below
which
liquid-to-solid
state
occurs.
This
offers
an
unambiguous
answer
nature
substances:
Above
liquid
constant
structure
relaxation;
solid
stable
structure.</p>
Journal of Physics Condensed Matter,
Journal Year:
2024,
Volume and Issue:
36(34), P. 343003 - 343003
Published: May 3, 2024
Abstract
Today’s
thermodynamics
is
largely
based
on
the
combined
law
for
equilibrium
systems
and
statistical
mechanics
derived
by
Gibbs
in
1873
1901,
respectively,
while
irreversible
nonequilibrium
resides
essentially
Onsager
Theorem
as
a
separate
branch
of
developed
1930s.
Between
them,
quantum
was
invented
quantitatively
solved
terms
density
functional
theory
(DFT)
1960s.
These
three
scientific
domains
operate
different
principles
are
very
much
separated
from
each
other.
In
analogy
to
parable
blind
men
elephant
articulated
Perdew,
they
individually
represent
portions
complex
system
thus
incomplete
themselves
alone,
resulting
lack
quantitative
agreement
between
their
predictions
experimental
observations.
Over
last
two
decades,
author’s
group
has
multiscale
entropy
approach
(recently
termed
zentropy
theory)
that
integrates
DFT-based
capable
accurately
predicting
free
energy
systems.
Furthermore,
combination
with
presented
Hillert,
author
cross
phenomena
beyond
phenomenological
Theorem.
The
jointly
provide
predictive
theories
electronic
any
observable
scales
reviewed
present
work.
The Journal of Chemical Physics,
Journal Year:
2024,
Volume and Issue:
160(20)
Published: May 28, 2024
The
melting
temperature
is
important
for
materials
design
because
of
its
relationship
with
thermal
stability,
synthesis,
and
processing
conditions.
Current
empirical
computational
point
estimation
techniques
are
limited
in
scope,
feasibility,
or
interpretability.
We
report
the
development
a
machine
learning
methodology
predicting
temperatures
binary
ionic
solid
materials.
evaluated
different
machine-learning
models
trained
on
dataset
points
476
non-metallic
crystalline
compounds
using
embeddings
constructed
from
elemental
properties
density-functional
theory
calculations
as
model
inputs.
A
direct
supervised-learning
approach
yields
mean
absolute
error
around
180
K
but
suffers
low
find
that
fidelity
predictions
can
further
be
improved
by
introducing
an
additional
unsupervised-learning
step
first
classifies
before
melting-point
regression.
Not
only
does
this
two-step
exhibit
accuracy,
also
provides
level
interpretability
insights
into
feature
importance
types
depend
specific
atomic
bonding
inside
material.
Motivated
finding,
we
used
symbolic
to
interpretable
physical
temperature,
which
recovered
best-performing
features
both
prior
provided
Advanced Science,
Journal Year:
2024,
Volume and Issue:
11(24)
Published: April 6, 2024
Abstract
Advanced
ceramic
materials
and
devices
call
for
better
reliability
damage
tolerance.
In
addition
to
their
strong
bonding
nature,
there
are
examples
demonstrating
superior
mechanical
properties
of
nanostructure
ceramics,
such
as
damage‐tolerant
aerogels
that
can
withstand
high
deformation
without
cracking
local
plasticity
in
dense
nanocrystalline
ceramics.
The
recent
progresses
shall
be
reviewed
this
perspective
article.
Three
topics
including
highly
elastic
nano‐fibrous
aerogels,
load‐bearing
nanoceramics
with
improved
properties,
implementing
machine
learning‐assisted
simulations
toolbox
understanding
the
relationship
among
structure,
mechanisms,
microstructure‐properties
discussed.
It
is
hoped
perspectives
present
here
help
discovery,
synthesis,
processing
future
structural
insensitive
flaws
damages
service.
