Predicting Rate Constants of Alkane Cracking Reactions Using Machine Learning
The Journal of Physical Chemistry A,
Journal Year:
2024,
Volume and Issue:
128(12), P. 2383 - 2392
Published: March 13, 2024
Calculating
the
thermal
rate
constants
of
elementary
combustion
reactions
is
great
importance
in
theoretical
chemistry.
Machine
learning
has
become
a
powerful,
data-driven
method
for
predicting
nowadays.
Recently,
molecular
similarity
combined
with
topological
indices
were
proposed
to
represent
hydrogen
abstraction
alkane
[
Language: Английский
Reconciling Accuracy and Feasibility for Barrierless Reaction Steps by the PCS/DDCI/MC-PDFT Protocol: Methane and Ethylene Dissociations as Case Studies
Journal of Chemical Theory and Computation,
Journal Year:
2024,
Volume and Issue:
unknown
Published: Sept. 17, 2024
Several
enhancements
have
been
introduced
into
state-of-the-art
computational
protocols
for
the
treatment
of
barrierless
reaction
steps
in
framework
variable
coordinate
variational
transition
state
theory.
The
first
step
is
synergistic
integration
Iterative
Difference
Dedicated
Configuration
Interaction
(I-DDCI)
and
Pisa
Composite
Scheme,
which
defines
a
reduced
cost,
yet
very
accurate,
workflow.
This
approach
provides
near
black
box
tool
obtaining
1D
reference
potentials.
Then,
general
strategy
has
devised
tuning
level
theory
used
Monte
Carlo
(MC)
sampling,
employing
Multiconfiguration
Pair
Density
Functional
Theory
(MC-PDFT)
with
dynamically
adjusted
Hartree-Fock
exchange.
Concurrently,
partial
geometry
optimizations
during
MC
simulations
account
coupling
between
coordinates
conserved
modes.
protocol
closely
approaches
full
size
consistency
yields
highly
accurate
results,
several
test
computations
suggesting
rapid
convergence
I-DDCI
correction
basis
set
dimensions.
capabilities
new
platform
are
illustrated
by
two
case
studies
(the
hydrogen
dissociation
from
CH
Language: Английский
Reduced Cost Computation and Exploitation of Accurate Radial Interaction Potentials for Barrier-Less Processes
Journal of Chemical Theory and Computation,
Journal Year:
2024,
Volume and Issue:
20(21), P. 9435 - 9445
Published: Nov. 1, 2024
Barrier-less
steps
are
typical
of
radical
and
ionic
reactions
in
the
gas-phase,
which
often
take
place
extreme
environments
such
as
combustion
reactors
operating
at
very
high
temperatures
or
interstellar
medium,
characterized
by
ultralow
pressures.
The
difficulty
experimental
studies
conditions
mimicking
these
suggests
that
computational
approaches
can
provide
a
valuable
support.
In
this
connection,
most
advanced
treatments
processes
framework
transition
state
theory
able
to
deliver
accurate
kinetic
parameters
provided
underlying
potential
energy
surface
is
sufficiently
accurate.
Since
requires
balanced
treatment
static
dynamic
correlation
(which
play
different
roles
regions),
sophisticated
expensive
quantum
chemical
required.
One
effective
solution
problem
offered
computation
one-dimensional
radial
potentials,
then
used
correct
results
Monte
Carlo
sampling
performed
cheaper
approaches.
paper,
we
will
show
that,
for
large
panel
barrier-less
reaction
steps,
ruled
Language: Английский