2026
|
Donato, Katarzyna; Koon, Gavin Kok Wai; Lee, Sarah; Carvalho, Alexandra; Tan, Hui Li; Costa, Mariana; Michalowski, Pawel Piotr; Nemeckova, Zuzana; Ecorchard, Petra; Donato, Ricardo K; Neto, Antonio Castro Engineering Disordered Metallic Carbonaceous Materials: A Protocol for
the Synthesis via Graphene Edge Hydrolysis ACS APPLIED NANO MATERIALS, 9 (10), pp. 4699-4714, 2026, DOI: 10.1021/acsanm.6c00047. Abstract | BibTeX | Endnote @article{WOS:001703246200001,
title = {Engineering Disordered Metallic Carbonaceous Materials: A Protocol for
the Synthesis via Graphene Edge Hydrolysis},
author = {Katarzyna Donato and Gavin Kok Wai Koon and Sarah Lee and Alexandra Carvalho and Hui Li Tan and Mariana Costa and Pawel Piotr Michalowski and Zuzana Nemeckova and Petra Ecorchard and Ricardo K Donato and Antonio Castro Neto},
doi = {10.1021/acsanm.6c00047},
times_cited = {0},
year = {2026},
date = {2026-03-01},
journal = {ACS APPLIED NANO MATERIALS},
volume = {9},
number = {10},
pages = {4699-4714},
publisher = {AMER CHEMICAL SOC},
address = {1155 16TH ST, NW, WASHINGTON, DC 20036 USA},
abstract = {This protocol is a comprehensive account of the intricate processes
involved in the rational design, synthesis, and characterization of
anisotropic metallic carbon materials. The materials were derived
through the hydrolytic oxidation of graphene sheets, followed by
self-assembly and mild annealing. The resulting products are highly
percolated carbon networks that preserve the essential basal area of the
source graphene. Structured into various sections, this document aims to
furnish detailed insights crucial for supporting further investigations
into these carbon materials. In particular, it highlights the key
distinctions from conventional graphite/graphene oxidation protocols,
offering a deeper understanding and ensuring the reproducibility of our
seminal findings. We believe this differentiation is crucial to
preventing the generalization of these materials from the outset, a
limitation widely reported in the graphene oxide family and a major
source of their inconsistencies, particularly in commercial products.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
This protocol is a comprehensive account of the intricate processes
involved in the rational design, synthesis, and characterization of
anisotropic metallic carbon materials. The materials were derived
through the hydrolytic oxidation of graphene sheets, followed by
self-assembly and mild annealing. The resulting products are highly
percolated carbon networks that preserve the essential basal area of the
source graphene. Structured into various sections, this document aims to
furnish detailed insights crucial for supporting further investigations
into these carbon materials. In particular, it highlights the key
distinctions from conventional graphite/graphene oxidation protocols,
offering a deeper understanding and ensuring the reproducibility of our
seminal findings. We believe this differentiation is crucial to
preventing the generalization of these materials from the outset, a
limitation widely reported in the graphene oxide family and a major
source of their inconsistencies, particularly in commercial products. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFKatarzyna Donato
Gavin Kok Wai Koon
Sarah Lee
Alexandra Carvalho
Hui Li Tan
Mariana Costa
Pawel Piotr Michalowski
Zuzana Nemeckova
Petra Ecorchard
Ricardo K Donato
Antonio Castro Neto
- TIEngineering Disordered Metallic Carbonaceous Materials: A Protocol for
the Synthesis via Graphene Edge Hydrolysis - SOACS APPLIED NANO MATERIALS
- DTArticle
- ABThis protocol is a comprehensive account of the intricate processes
involved in the rational design, synthesis, and characterization of
anisotropic metallic carbon materials. The materials were derived
through the hydrolytic oxidation of graphene sheets, followed by
self-assembly and mild annealing. The resulting products are highly
percolated carbon networks that preserve the essential basal area of the
source graphene. Structured into various sections, this document aims to
furnish detailed insights crucial for supporting further investigations
into these carbon materials. In particular, it highlights the key
distinctions from conventional graphite/graphene oxidation protocols,
offering a deeper understanding and ensuring the reproducibility of our
seminal findings. We believe this differentiation is crucial to
preventing the generalization of these materials from the outset, a
limitation widely reported in the graphene oxide family and a major
source of their inconsistencies, particularly in commercial products. - Z90
- PUAMER CHEMICAL SOC
- PA1155 16TH ST, NW, WASHINGTON, DC 20036 USA
- VL9
- BP4699
- EP4714
- DI10.1021/acsanm.6c00047
- UTWOS:001703246200001
- ER
- EF
|
2025
|
Liu, Yuqing; Carvalho, Alexandra; Lai, Wenhui; Pu, Yanhui; Zhang, Zheng; Lim, Sharon Xiaodai; Neto, Antonio Castro H; Gupta, Puneet; Sow, Chorng Haur Enhanced battery performance by fluorescent defects engineering in hard
carbon anodes CHEMICAL ENGINEERING JOURNAL, 514 , 2025, DOI: 10.1016/j.cej.2025.163279. Abstract | BibTeX | Endnote @article{WOS:001490582900001,
title = {Enhanced battery performance by fluorescent defects engineering in hard
carbon anodes},
author = {Yuqing Liu and Alexandra Carvalho and Wenhui Lai and Yanhui Pu and Zheng Zhang and Sharon Xiaodai Lim and Antonio Castro H Neto and Puneet Gupta and Chorng Haur Sow},
doi = {10.1016/j.cej.2025.163279},
times_cited = {4},
issn = {1385-8947},
year = {2025},
date = {2025-06-01},
journal = {CHEMICAL ENGINEERING JOURNAL},
volume = {514},
publisher = {ELSEVIER SCIENCE SA},
address = {PO BOX 564, 1001 LAUSANNE, SWITZERLAND},
abstract = {The availability and accessibility of economical renewable energy
remains a key driving factor towards encouraging the uptake of clean
energy. By incorporating economical hard carbon (HC), recycled from
waste into anodes for lithium-ion batteries (LIBs), and treating the
resulting HC-anode with a focused laser beam, the functionalised
HC-anode exhibits enhanced electrochemical performance with a specific
capacity of 516 mAh g- 1 at 0.1 A g- 1. It achieves over 100% capacity
retention for 700 cycles at 1 A g- 1, and demonstrates super-long
durability for 4000 cycles at 2 A g- 1. The improvements are attributed
to laser-tunable expanded interlayer spacing and fluorescing defects in
the engineered HC-anode. DFT calculations further established that these
fluorescent defects correspond to carbon vacancies (cyan fluorescence),
and their complexes with H heteroatoms (green fluorescence). These
defects lead to the improved electrochemical performance via enhancing
Li+ adsorption energies. Given such correlation, fluorescence studies
are proposed as an interesting mechanism for guiding the development of
carbon materials for energy applications, which serves as a highly
efficient tool for assessing the electrochemical performance,
eliminating the need for costly battery fabrication and testing
processes. The performance achieved and its correlation to the
observable fluorescence will not only contribute towards the effort of
making cheaper batteries with better performance, but also serves as a
rapid and scalable probe for preliminary evaluation of battery
performance.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The availability and accessibility of economical renewable energy
remains a key driving factor towards encouraging the uptake of clean
energy. By incorporating economical hard carbon (HC), recycled from
waste into anodes for lithium-ion batteries (LIBs), and treating the
resulting HC-anode with a focused laser beam, the functionalised
HC-anode exhibits enhanced electrochemical performance with a specific
capacity of 516 mAh g- 1 at 0.1 A g- 1. It achieves over 100% capacity
retention for 700 cycles at 1 A g- 1, and demonstrates super-long
durability for 4000 cycles at 2 A g- 1. The improvements are attributed
to laser-tunable expanded interlayer spacing and fluorescing defects in
the engineered HC-anode. DFT calculations further established that these
fluorescent defects correspond to carbon vacancies (cyan fluorescence),
and their complexes with H heteroatoms (green fluorescence). These
defects lead to the improved electrochemical performance via enhancing
Li+ adsorption energies. Given such correlation, fluorescence studies
are proposed as an interesting mechanism for guiding the development of
carbon materials for energy applications, which serves as a highly
efficient tool for assessing the electrochemical performance,
eliminating the need for costly battery fabrication and testing
processes. The performance achieved and its correlation to the
observable fluorescence will not only contribute towards the effort of
making cheaper batteries with better performance, but also serves as a
rapid and scalable probe for preliminary evaluation of battery
performance. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFYuqing Liu
Alexandra Carvalho
Wenhui Lai
Yanhui Pu
Zheng Zhang
Sharon Xiaodai Lim
Antonio Castro H Neto
Puneet Gupta
Chorng Haur Sow
- TIEnhanced battery performance by fluorescent defects engineering in hard
carbon anodes - SOCHEMICAL ENGINEERING JOURNAL
- DTArticle
- ABThe availability and accessibility of economical renewable energy
remains a key driving factor towards encouraging the uptake of clean
energy. By incorporating economical hard carbon (HC), recycled from
waste into anodes for lithium-ion batteries (LIBs), and treating the
resulting HC-anode with a focused laser beam, the functionalised
HC-anode exhibits enhanced electrochemical performance with a specific
capacity of 516 mAh g- 1 at 0.1 A g- 1. It achieves over 100% capacity
retention for 700 cycles at 1 A g- 1, and demonstrates super-long
durability for 4000 cycles at 2 A g- 1. The improvements are attributed
to laser-tunable expanded interlayer spacing and fluorescing defects in
the engineered HC-anode. DFT calculations further established that these
fluorescent defects correspond to carbon vacancies (cyan fluorescence),
and their complexes with H heteroatoms (green fluorescence). These
defects lead to the improved electrochemical performance via enhancing
Li+ adsorption energies. Given such correlation, fluorescence studies
are proposed as an interesting mechanism for guiding the development of
carbon materials for energy applications, which serves as a highly
efficient tool for assessing the electrochemical performance,
eliminating the need for costly battery fabrication and testing
processes. The performance achieved and its correlation to the
observable fluorescence will not only contribute towards the effort of
making cheaper batteries with better performance, but also serves as a
rapid and scalable probe for preliminary evaluation of battery
performance. - Z94
- PUELSEVIER SCIENCE SA
- PAPO BOX 564, 1001 LAUSANNE, SWITZERLAND
- SN1385-8947
- VL514
- DI10.1016/j.cej.2025.163279
- UTWOS:001490582900001
- ER
- EF
|
Maevskiy, Artem; Carvalho, Alexandra; Sataev, Emil; Turchyna, Volha; Noori, Keian; Rodin, Aleksandr; Neto, Castro A H; Ustyuzhanin, Andrey Predicting ionic conductivity in solids from the machine-learned
potential energy landscape PHYSICAL REVIEW RESEARCH, 7 (2), 2025, DOI: 10.1103/PhysRevResearch.7.023167. Abstract | BibTeX | Endnote @article{WOS:001493752600014,
title = {Predicting ionic conductivity in solids from the machine-learned
potential energy landscape},
author = {Artem Maevskiy and Alexandra Carvalho and Emil Sataev and Volha Turchyna and Keian Noori and Aleksandr Rodin and Castro A H Neto and Andrey Ustyuzhanin},
doi = {10.1103/PhysRevResearch.7.023167},
times_cited = {9},
year = {2025},
date = {2025-05-01},
journal = {PHYSICAL REVIEW RESEARCH},
volume = {7},
number = {2},
publisher = {AMER PHYSICAL SOC},
address = {ONE PHYSICS ELLIPSE, COLLEGE PK, MD 20740-3844 USA},
abstract = {Discovering new superionic materials is essential for advancing
solid-state batteries, which offer improved energy density and safety
compared to traditional lithium-ion batteries with liquid electrolytes.
Conventional computational methods for identifying such materials are
resource-intensive and not easily scalable. Recently, universal
interatomic potential models have been developed using equivariant graph
neural networks. These models are trained on extensive datasets of
first-principles force and energy calculations. One can achieve
significant computational advantages by leveraging them as the
foundation for traditional methods of assessing the ionic conductivity,
such as molecular dynamics or nudged elastic band techniques. However,
the generalization error from model inference on diverse atomic
structures arising in such calculations can compromise the reliability
of the results. In this work, we propose an approach for the quick and
reliable screening of ionic conductors through the analysis of a
universal interatomic potential. Our method incorporates a set of
heuristic structure descriptors that effectively employ the rich
knowledge of the underlying model while requiring minimal generalization
capabilities. Using our descriptors, we rank lithium-containing
materials in the Materials Project database according to their expected
ionic conductivity. Eight out of the ten highest-ranked materials are
confirmed to be superionic at room temperature in first-principles
calculations. Notably, our method achieves a speed-up factor of
approximately 50 compared to molecular dynamics driven by a
machine-learning potential, and it is at least 3000 times faster
compared to first-principles molecular dynamics.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Discovering new superionic materials is essential for advancing
solid-state batteries, which offer improved energy density and safety
compared to traditional lithium-ion batteries with liquid electrolytes.
Conventional computational methods for identifying such materials are
resource-intensive and not easily scalable. Recently, universal
interatomic potential models have been developed using equivariant graph
neural networks. These models are trained on extensive datasets of
first-principles force and energy calculations. One can achieve
significant computational advantages by leveraging them as the
foundation for traditional methods of assessing the ionic conductivity,
such as molecular dynamics or nudged elastic band techniques. However,
the generalization error from model inference on diverse atomic
structures arising in such calculations can compromise the reliability
of the results. In this work, we propose an approach for the quick and
reliable screening of ionic conductors through the analysis of a
universal interatomic potential. Our method incorporates a set of
heuristic structure descriptors that effectively employ the rich
knowledge of the underlying model while requiring minimal generalization
capabilities. Using our descriptors, we rank lithium-containing
materials in the Materials Project database according to their expected
ionic conductivity. Eight out of the ten highest-ranked materials are
confirmed to be superionic at room temperature in first-principles
calculations. Notably, our method achieves a speed-up factor of
approximately 50 compared to molecular dynamics driven by a
machine-learning potential, and it is at least 3000 times faster
compared to first-principles molecular dynamics. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFArtem Maevskiy
Alexandra Carvalho
Emil Sataev
Volha Turchyna
Keian Noori
Aleksandr Rodin
Castro A H Neto
Andrey Ustyuzhanin
- TIPredicting ionic conductivity in solids from the machine-learned
potential energy landscape - SOPHYSICAL REVIEW RESEARCH
- DTArticle
- ABDiscovering new superionic materials is essential for advancing
solid-state batteries, which offer improved energy density and safety
compared to traditional lithium-ion batteries with liquid electrolytes.
Conventional computational methods for identifying such materials are
resource-intensive and not easily scalable. Recently, universal
interatomic potential models have been developed using equivariant graph
neural networks. These models are trained on extensive datasets of
first-principles force and energy calculations. One can achieve
significant computational advantages by leveraging them as the
foundation for traditional methods of assessing the ionic conductivity,
such as molecular dynamics or nudged elastic band techniques. However,
the generalization error from model inference on diverse atomic
structures arising in such calculations can compromise the reliability
of the results. In this work, we propose an approach for the quick and
reliable screening of ionic conductors through the analysis of a
universal interatomic potential. Our method incorporates a set of
heuristic structure descriptors that effectively employ the rich
knowledge of the underlying model while requiring minimal generalization
capabilities. Using our descriptors, we rank lithium-containing
materials in the Materials Project database according to their expected
ionic conductivity. Eight out of the ten highest-ranked materials are
confirmed to be superionic at room temperature in first-principles
calculations. Notably, our method achieves a speed-up factor of
approximately 50 compared to molecular dynamics driven by a
machine-learning potential, and it is at least 3000 times faster
compared to first-principles molecular dynamics. - Z99
- PUAMER PHYSICAL SOC
- PAONE PHYSICS ELLIPSE, COLLEGE PK, MD 20740-3844 USA
- VL7
- DI10.1103/PhysRevResearch.7.023167
- UTWOS:001493752600014
- ER
- EF
|
Ng, Joseph J Q; Tkachev, Sergey; Sim, Glendon C F; de Lima, Luiza Felippi; Koon, Gavin K W; Lima, Alexandre P; Neto, Antonio Castro H Non-Invasive Hydration Monitoring with a Graphene Dual Sweat Sensor APPLIED SCIENCES-BASEL, 15 (9), 2025, DOI: 10.3390/app15094970. Abstract | BibTeX | Endnote @article{WOS:001486021400001,
title = {Non-Invasive Hydration Monitoring with a Graphene Dual Sweat Sensor},
author = {Joseph J Q Ng and Sergey Tkachev and Glendon C F Sim and Luiza Felippi de Lima and Gavin K W Koon and Alexandre P Lima and Antonio Castro H Neto},
doi = {10.3390/app15094970},
times_cited = {3},
year = {2025},
date = {2025-04-01},
journal = {APPLIED SCIENCES-BASEL},
volume = {15},
number = {9},
publisher = {MDPI},
address = {MDPI AG, Grosspeteranlage 5, CH-4052 BASEL, SWITZERLAND},
abstract = {Maintaining optimal hydration is critical for physiological function,
particularly during intense physical activities, in which dehydration or
overhydration can impair performance and recovery. Traditional methods
for monitoring hydration status, such as body weight changes,
bioelectrical impedance, and urine specific gravity, are limited by
inconvenience and lack of real-time capability. This study introduces a
novel graphene-based dual-sensing electrochemical sensor for the rapid
and non-invasive quantification of sodium and potassium concentrations
in human sweat, key biomarkers of hydration status. Leveraging
graphene's exceptional conductivity and functionalization potential, the
sensor employs open-circuit potentiometry (OCP) to achieve high
sensitivity and selectivity in detecting sodium and potassium. The
sensor performance was validated against that of a commercial analyzer
and ICP-OES, demonstrating a near-Nernstian response (61.93 mV/decade
for sodium and 61.21 mV/decade for potassium detection) and a linear
detection range spanning from 0.1 mM to 100 mM for both sodium and
potassium monitoring in sweat. Sweat samples from an athlete during
endurance exercise confirmed the sensor's reliability, with results
closely matching those of ICP-OES and outperforming the commercial
analyzer in regards to accuracy and sample efficiency. This work
represents a cross-validated study of a sweat-based sensor with a second
analytical technique, highlighting its potential as a real-time
hydration monitoring tool for use in sports and beyond.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Maintaining optimal hydration is critical for physiological function,
particularly during intense physical activities, in which dehydration or
overhydration can impair performance and recovery. Traditional methods
for monitoring hydration status, such as body weight changes,
bioelectrical impedance, and urine specific gravity, are limited by
inconvenience and lack of real-time capability. This study introduces a
novel graphene-based dual-sensing electrochemical sensor for the rapid
and non-invasive quantification of sodium and potassium concentrations
in human sweat, key biomarkers of hydration status. Leveraging
graphene's exceptional conductivity and functionalization potential, the
sensor employs open-circuit potentiometry (OCP) to achieve high
sensitivity and selectivity in detecting sodium and potassium. The
sensor performance was validated against that of a commercial analyzer
and ICP-OES, demonstrating a near-Nernstian response (61.93 mV/decade
for sodium and 61.21 mV/decade for potassium detection) and a linear
detection range spanning from 0.1 mM to 100 mM for both sodium and
potassium monitoring in sweat. Sweat samples from an athlete during
endurance exercise confirmed the sensor's reliability, with results
closely matching those of ICP-OES and outperforming the commercial
analyzer in regards to accuracy and sample efficiency. This work
represents a cross-validated study of a sweat-based sensor with a second
analytical technique, highlighting its potential as a real-time
hydration monitoring tool for use in sports and beyond. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFJoseph J Q Ng
Sergey Tkachev
Glendon C F Sim
Luiza Felippi de Lima
Gavin K W Koon
Alexandre P Lima
Antonio Castro H Neto
- TINon-Invasive Hydration Monitoring with a Graphene Dual Sweat Sensor
- SOAPPLIED SCIENCES-BASEL
- DTArticle
- ABMaintaining optimal hydration is critical for physiological function,
particularly during intense physical activities, in which dehydration or
overhydration can impair performance and recovery. Traditional methods
for monitoring hydration status, such as body weight changes,
bioelectrical impedance, and urine specific gravity, are limited by
inconvenience and lack of real-time capability. This study introduces a
novel graphene-based dual-sensing electrochemical sensor for the rapid
and non-invasive quantification of sodium and potassium concentrations
in human sweat, key biomarkers of hydration status. Leveraging
graphene's exceptional conductivity and functionalization potential, the
sensor employs open-circuit potentiometry (OCP) to achieve high
sensitivity and selectivity in detecting sodium and potassium. The
sensor performance was validated against that of a commercial analyzer
and ICP-OES, demonstrating a near-Nernstian response (61.93 mV/decade
for sodium and 61.21 mV/decade for potassium detection) and a linear
detection range spanning from 0.1 mM to 100 mM for both sodium and
potassium monitoring in sweat. Sweat samples from an athlete during
endurance exercise confirmed the sensor's reliability, with results
closely matching those of ICP-OES and outperforming the commercial
analyzer in regards to accuracy and sample efficiency. This work
represents a cross-validated study of a sweat-based sensor with a second
analytical technique, highlighting its potential as a real-time
hydration monitoring tool for use in sports and beyond. - Z93
- PUMDPI
- PAMDPI AG, Grosspeteranlage 5, CH-4052 BASEL, SWITZERLAND
- VL15
- DI10.3390/app15094970
- UTWOS:001486021400001
- ER
- EF
|
Li, Tianbo; Lin, Min; Dale, Stephen G; Shi, Zekun; Neto, Castro A H; Novoselov, Kostya S; Vignale, Giovanni Diagonalization without Diagonalization: A Direct Optimization Approach
for Solid-State Density Functional Theory JOURNAL OF CHEMICAL THEORY AND COMPUTATION, 21 (9), pp. 4730-4741, 2025, DOI: 10.1021/acs.jctc.4c01551. Abstract | BibTeX | Endnote @article{WOS:001478048300001,
title = {Diagonalization without Diagonalization: A Direct Optimization Approach
for Solid-State Density Functional Theory},
author = {Tianbo Li and Min Lin and Stephen G Dale and Zekun Shi and Castro A H Neto and Kostya S Novoselov and Giovanni Vignale},
doi = {10.1021/acs.jctc.4c01551},
times_cited = {2},
issn = {1549-9618},
year = {2025},
date = {2025-04-01},
journal = {JOURNAL OF CHEMICAL THEORY AND COMPUTATION},
volume = {21},
number = {9},
pages = {4730-4741},
publisher = {AMER CHEMICAL SOC},
address = {1155 16TH ST, NW, WASHINGTON, DC 20036 USA},
abstract = {We present a novel approach to address the challenges of variable
occupation numbers in direct optimization of density functional theory
(DFT). By parametrizing both the eigenfunctions and the occupation
matrix, our method minimizes the free energy with respect to these
parameters. As the stationary conditions require the occupation matrix
and the Kohn-Sham Hamiltonian to be simultaneously diagonalizable, this
leads to the concept of ``self-diagonalization,'' where, by assuming a
diagonal occupation matrix without loss of generality, the Hamiltonian
matrix naturally becomes diagonal at stationary points. Our method
incorporates physical constraints on both the eigenfunctions and the
occupations into the parametrization, transforming the constrained
optimization into an fully differentiable unconstrained problem, which
is solvable via gradient descent. Implemented in JAX, our method was
tested on aluminum and silicon, confirming that it achieves efficient
self-diagonalization, produces the correct Fermi-Dirac distribution of
the occupation numbers and yields band structures consistent with those
obtained with SCF eigensolver methods in Quantum Espresso.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
We present a novel approach to address the challenges of variable
occupation numbers in direct optimization of density functional theory
(DFT). By parametrizing both the eigenfunctions and the occupation
matrix, our method minimizes the free energy with respect to these
parameters. As the stationary conditions require the occupation matrix
and the Kohn-Sham Hamiltonian to be simultaneously diagonalizable, this
leads to the concept of ``self-diagonalization,'' where, by assuming a
diagonal occupation matrix without loss of generality, the Hamiltonian
matrix naturally becomes diagonal at stationary points. Our method
incorporates physical constraints on both the eigenfunctions and the
occupations into the parametrization, transforming the constrained
optimization into an fully differentiable unconstrained problem, which
is solvable via gradient descent. Implemented in JAX, our method was
tested on aluminum and silicon, confirming that it achieves efficient
self-diagonalization, produces the correct Fermi-Dirac distribution of
the occupation numbers and yields band structures consistent with those
obtained with SCF eigensolver methods in Quantum Espresso. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFTianbo Li
Min Lin
Stephen G Dale
Zekun Shi
Castro A H Neto
Kostya S Novoselov
Giovanni Vignale
- TIDiagonalization without Diagonalization: A Direct Optimization Approach
for Solid-State Density Functional Theory - SOJOURNAL OF CHEMICAL THEORY AND COMPUTATION
- DTArticle
- ABWe present a novel approach to address the challenges of variable
occupation numbers in direct optimization of density functional theory
(DFT). By parametrizing both the eigenfunctions and the occupation
matrix, our method minimizes the free energy with respect to these
parameters. As the stationary conditions require the occupation matrix
and the Kohn-Sham Hamiltonian to be simultaneously diagonalizable, this
leads to the concept of ``self-diagonalization,'' where, by assuming a
diagonal occupation matrix without loss of generality, the Hamiltonian
matrix naturally becomes diagonal at stationary points. Our method
incorporates physical constraints on both the eigenfunctions and the
occupations into the parametrization, transforming the constrained
optimization into an fully differentiable unconstrained problem, which
is solvable via gradient descent. Implemented in JAX, our method was
tested on aluminum and silicon, confirming that it achieves efficient
self-diagonalization, produces the correct Fermi-Dirac distribution of
the occupation numbers and yields band structures consistent with those
obtained with SCF eigensolver methods in Quantum Espresso. - Z92
- PUAMER CHEMICAL SOC
- PA1155 16TH ST, NW, WASHINGTON, DC 20036 USA
- SN1549-9618
- VL21
- BP4730
- EP4741
- DI10.1021/acs.jctc.4c01551
- UTWOS:001478048300001
- ER
- EF
|
