Abstract
Carbon
dioxide
(CO
2
)
utilization
technology
is
of
great
significance
for
achieving
carbon
neutrality,
in
which
the
catalytic
materials
play
crucial
roles,
and
among
them,
single‐atom
alloys
(SAAs)
are
particular
interests.
In
this
study,
density
functional
theory
(DFT)
calculations
machine
learning
employed
to
assess
effectiveness
Cu‐,
Ag‐,
Ni‐host
SAAs
as
catalysts
electrochemical
CO
reduction
CH
3
OH.
The
Gibbs
free
energies
477
elementary
reactions
across
35
involved
calculated,
by
utilizing
dataset,
a
trained
gradient
boosting
regression
model
established
with
an
excellent
accuracy.
Subsequently,
properties
46
unknown
predicted,
including
their
pathways,
products,
potential‐determining
steps
(PDS),
corresponding
PDS
(
G
).
Three
promising
candidates,
ZnCu,
AuAg
MoNi,
stand
out
due
lowest
Ag‐
Ni‐
hosted
SAAs,
respectively.
Growing
global
population,
escalating
energy
consumption,
and
climate
change
threaten
future
security.
Fossil
fuel
combustion,
primarily
coal,
oil,
natural
gas,
exacerbates
the
greenhouse
effect
driving
warming
through
CO
Langmuir,
Год журнала:
2025,
Номер
41(5), С. 3490 - 3502
Опубликована: Янв. 31, 2025
The
applications
of
machine
learning
(ML)
in
complex
interfacial
interactions
are
hindered
by
the
time-consuming
process
manual
feature
selection
and
model
construction.
An
automated
ML
program
was
implemented
with
four
subsequent
steps:
data
distribution
analysis,
dimensionality
reduction
clustering,
selection,
optimization.
Without
need
intervention,
descriptors
metal
charge
variance
(ΔQCT)
electronegativity
substrate
(χsub)
(δχM)
were
raised
up
good
performance
predicting
electrochemical
reaction
energies
for
both
nitrogen
(NRR)
CO2
(CO2RR)
on
metal-zeolites
MoS2
surfaces.
important
role
tuning
catalytic
reactivity
NRR
CO2RR
highlighted
from
SHAP
analysis.
It
proposed
that
Fe-,
Cr-,
Zn-,
Nb-,
Ta-zeolites
favorable
catalysts
NRR,
while
Ni-zeolite
showed
a
preference
CO2RR.
elongated
bond
N2
or
bent
configuration
shown
V-,
Co-,
Mo-zeolites,
indicating
molecule
could
be
activated
after
adsorption
pathways.
generalizability
automatically
built
is
demonstrated
to
other
systems
such
as
metal-organic
frameworks
SiO2
useful
tool
accelerate
data-driven
exploration
relationship
between
structures
material
properties
without
selection.
Abstract
The
combination
of
single‐atom
catalysts
and
vacancy
engineering
is
an
emerging
hotspot
in
the
field
catalysis.
existence
vacancies
provides
additional
adsorption
sites,
which
can
increase
capacity
diffusion
rate
reactant
molecules.
Single
atoms
have
single‐height
atom
utilization
steric
hindrance
effects.
two
synergistically
show
different
activation
synergistic
processes
on
defect
materials.
Under
effect
catalysts,
they
affect
orientation
reactants
catalyst
surface
formation
intermediate
configurations,
thereby
regulating
reaction
path
product
selectivity,
greatly
enhancing
catalytic
performance,
maximizing
material
itself.
Therefore,
between
pervacancies
realize
more
efficient,
sustainable,
economical
systems,
possessing
high‐efficiency
performance
photocatalysis,
electrocatalysis
thermocatalysis,
setting
off
a
new
wave
fields
energy
conversion,
environmental
protection,
organic
synthesis.
constructed
by
single
broad
bright
application
prospects.
ACS Catalysis,
Год журнала:
2024,
Номер
14(17), С. 12947 - 12955
Опубликована: Авг. 14, 2024
Metal–organic
framework-supported
single-atom
catalysts
(SACs@MOF)
show
considerable
promise
in
CO2
reduction
reactions
(CO2RR).
However,
efficiently
screening
and
designing
optimal
is
hindered
by
the
lack
of
effective
descriptors
for
encoding
complex
chemical
microenvironments
SAC@MOF
systems.
Herein,
through
combining
an
intuition-guided
dimensionality
strategy
with
machine
learning
(ML),
we
identified
critical
based
on
atomic
features
SAC's
constrained
coordination
geometry,
which
capture
effects
electrochemical
CO2RR
activity
selectivity
UiO-66-supported
SACs.
With
these
descriptors,
accurate
ML
models
were
developed
to
predict
limiting
potentials
producing
HCOOH,
CO,
CH4/CH3OH
48
SACs@UiO-66-X
(X
=
H,
NH2,
Br).
Moreover,
transferability
was
demonstrated
additional
systems
X
CH3,
OH,
NO2.
The
accuracy
predicted
trends
specific
SACs
combined
different
linker
groups
top-performing
validated
DFT
calculations.
This
study
provides
framework
understanding
modulating
microenvironments,
enhancing
design
development
MOF-supported
CO2RR.
The Journal of Physical Chemistry Letters,
Год журнала:
2025,
Номер
16(6), С. 1424 - 1431
Опубликована: Янв. 31, 2025
Dual-metal
site
catalysts
(DMSCs)
supported
on
nitrogen-doped
graphene
have
shown
great
potential
in
heterogeneous
catalysis
due
to
their
unique
properties
and
enhanced
efficiency.
However,
the
precise
control
stabilization
of
metal
dimers,
particularly
oxygen
activation
reactions,
present
significant
challenges
practical
applications.
In
this
study,
we
integrate
high-throughput
density
functional
theory
calculations
with
machine
learning
techniques
predict
optimize
catalytic
DMSCs.
Transfer
is
employed
enhance
model's
generalization
capability,
successfully
predicting
performance
across
new
combinations.
Additionally,
application
SISSO
method
enables
derivation
interpretable
symbolic
regression
models,
revealing
critical
correlations
between
electronic
structure
features
This
approach
not
only
advances
understanding
dual-metal
but
also
provides
a
novel
framework
for
systematic
design
optimization
highly
efficient
catalysts,
broad
applicability
science.
The Journal of Physical Chemistry Letters,
Год журнала:
2025,
Номер
16(9), С. 2357 - 2368
Опубликована: Фев. 26, 2025
Accurately
controlling
the
interactions
and
dynamic
changes
between
multiple
active
sites
(e.g.,
metals,
vacancies,
lone
pairs
of
heteroatoms)
to
achieve
efficient
catalytic
performance
is
a
key
issue
challenge
in
design
complex
reactions
involving
2D
metal-supported
catalysts,
metal-zeolites,
metal–organic
metalloenzymes.
With
aid
machine
learning
(ML),
descriptors
play
central
role
optimizing
electrochemical
elucidating
essence
activity,
predicting
more
thereby
avoiding
time-consuming
trial-and-error
processes.
Three
kinds
descriptors─active
center
descriptors,
interfacial
reaction
pathway
descriptors─are
crucial
for
understanding
designing
catalysts.
Specifically,
as
sites,
synergize
with
metals
significantly
promote
reduction
energy-relevant
small
molecules.
By
combining
some
physical
interpretable
can
be
constructed
evaluate
performance.
Future
development
ML
models
faces
constructing
vacancies
multicatalysis
systems
rationally
selectivity,
stability
Utilization
generative
artificial
intelligence
multimodal
automatically
extract
would
accelerate
exploration
mechanisms.
The
transferable
from
catalysts
metalloenzymes
provide
innovative
solutions
energy
conversion
environmental
protection.
