2025
|
Lukianov, M Y; Maevskiy, A; Kazeev, N; Mylnikov, D; Svintsov, D A; Novoselov, K S; Ustyuzhanin, A; Bandurin, D A Inverse design of broadband antennas for terahertz devices based on
two-dimensional materials PHYSICAL REVIEW APPLIED, 24 (5), 2025, DOI: 10.1103/gr2z-3qjp. Abstract | BibTeX | Endnote @article{WOS:001633477300003,
title = {Inverse design of broadband antennas for terahertz devices based on
two-dimensional materials},
author = {M Y Lukianov and A Maevskiy and N Kazeev and D Mylnikov and D A Svintsov and K S Novoselov and A Ustyuzhanin and D A Bandurin},
doi = {10.1103/gr2z-3qjp},
times_cited = {0},
issn = {2331-7019},
year = {2025},
date = {2025-11-01},
journal = {PHYSICAL REVIEW APPLIED},
volume = {24},
number = {5},
publisher = {AMER PHYSICAL SOC},
address = {ONE PHYSICS ELLIPSE, COLLEGE PK, MD 20740-3844 USA},
abstract = {Terahertz technology, a cornerstone of next-generation high-speed
communication and sensing, has long been hindered by impedance mismatch
challenges that limit device performance and applicability. These
challenges become particularly pronounced when ultrasensitive
two-dimensional (2D) materials are employed as the detecting element in
the terahertz range, further complicating their integration in realworld
applications. Furthermore, conventional antenna designs often fail to
provide adequate matching across the broad terahertz spectrum. In this
work, we tackle these challenges using a procedural generation algorithm
to design terahertz broadband antennas that satisfy specific performance
criteria. Namely, the developed inverse design methodology enables
customization for the target impedance value, bandwidth, and contact
topology requirements. The proposed antenna achieves an improvement of
up to 40% in power transfer efficiency, compared with traditional
bow-tie antennas, under realistic operating conditions. High-fidelity
electromagnetic simulations validate these results, confirming the
design's practicality for terahertz applications. This work addresses
critical limitations of existing antenna designs and advances the
feasibility of high-frequency applications in both communication and
sensing.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Terahertz technology, a cornerstone of next-generation high-speed
communication and sensing, has long been hindered by impedance mismatch
challenges that limit device performance and applicability. These
challenges become particularly pronounced when ultrasensitive
two-dimensional (2D) materials are employed as the detecting element in
the terahertz range, further complicating their integration in realworld
applications. Furthermore, conventional antenna designs often fail to
provide adequate matching across the broad terahertz spectrum. In this
work, we tackle these challenges using a procedural generation algorithm
to design terahertz broadband antennas that satisfy specific performance
criteria. Namely, the developed inverse design methodology enables
customization for the target impedance value, bandwidth, and contact
topology requirements. The proposed antenna achieves an improvement of
up to 40% in power transfer efficiency, compared with traditional
bow-tie antennas, under realistic operating conditions. High-fidelity
electromagnetic simulations validate these results, confirming the
design's practicality for terahertz applications. This work addresses
critical limitations of existing antenna designs and advances the
feasibility of high-frequency applications in both communication and
sensing. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFM Y Lukianov
A Maevskiy
N Kazeev
D Mylnikov
D A Svintsov
K S Novoselov
A Ustyuzhanin
D A Bandurin
- TIInverse design of broadband antennas for terahertz devices based on
two-dimensional materials - SOPHYSICAL REVIEW APPLIED
- DTArticle
- ABTerahertz technology, a cornerstone of next-generation high-speed
communication and sensing, has long been hindered by impedance mismatch
challenges that limit device performance and applicability. These
challenges become particularly pronounced when ultrasensitive
two-dimensional (2D) materials are employed as the detecting element in
the terahertz range, further complicating their integration in realworld
applications. Furthermore, conventional antenna designs often fail to
provide adequate matching across the broad terahertz spectrum. In this
work, we tackle these challenges using a procedural generation algorithm
to design terahertz broadband antennas that satisfy specific performance
criteria. Namely, the developed inverse design methodology enables
customization for the target impedance value, bandwidth, and contact
topology requirements. The proposed antenna achieves an improvement of
up to 40% in power transfer efficiency, compared with traditional
bow-tie antennas, under realistic operating conditions. High-fidelity
electromagnetic simulations validate these results, confirming the
design's practicality for terahertz applications. This work addresses
critical limitations of existing antenna designs and advances the
feasibility of high-frequency applications in both communication and
sensing. - Z90
- PUAMER PHYSICAL SOC
- PAONE PHYSICS ELLIPSE, COLLEGE PK, MD 20740-3844 USA
- SN2331-7019
- VL24
- DI10.1103/gr2z-3qjp
- UTWOS:001633477300003
- ER
- EF
|
Kazeev, Nikita; Nong, Wei; Romanov, Ignat; Zhu, Ruiming; Ustyuzhanin, Andrey; Yamazaki, Shuya; Hippalgaonkar, Kedar Wyckoff Transformer: Generation of Symmetric Crystals Singh, A; Fazel, M; Hsu, D; Lacoste-Julien, S; Berkenkamp, F; Maharaj, T; Wagstaff, K; Zhu, J (Ed.): INTERNATIONAL CONFERENCE ON MACHINE LEARNING, pp. 29495-29526, JMLR-JOURNAL MACHINE LEARNING RESEARCH, 1269 LAW ST, SAN DIEGO, CA, UNITED STATES, 2025, (42nd International Conference on Machine Learning-ICML-Annual,
Vancouver, CANADA, JUL 13-19, 2025). Abstract | BibTeX | Endnote @inproceedings{WOS:001693126000256,
title = {Wyckoff Transformer: Generation of Symmetric Crystals},
author = {Nikita Kazeev and Wei Nong and Ignat Romanov and Ruiming Zhu and Andrey Ustyuzhanin and Shuya Yamazaki and Kedar Hippalgaonkar},
editor = {A Singh and M Fazel and D Hsu and S Lacoste-Julien and F Berkenkamp and T Maharaj and K Wagstaff and J Zhu},
times_cited = {0},
issn = {2640-3498},
year = {2025},
date = {2025-01-01},
booktitle = {INTERNATIONAL CONFERENCE ON MACHINE LEARNING},
volume = {267},
pages = {29495-29526},
publisher = {JMLR-JOURNAL MACHINE LEARNING RESEARCH},
address = {1269 LAW ST, SAN DIEGO, CA, UNITED STATES},
series = {Proceedings of Machine Learning Research},
abstract = {Crystal symmetry plays a fundamental role in determining its physical,
chemical, and electronic properties such as electrical and thermal
conductivity, optical and polarization behavior, and mechanical
strength. Almost all known crystalline materials have internal symmetry.
However, this is often inadequately addressed by existing generative
models, making the consistent generation of stable and symmetrically
valid crystal structures a significant challenge. We introduce WyFormer,
a generative model that directly tackles this by formally conditioning
on space group symmetry. It achieves this by using Wyckoff positions as
the basis for an elegant, compressed, and discrete structure
representation. To model the distribution, we develop a
permutation-invariant autoregressive model based on the Transformer
encoder and an absence of positional encoding. Extensive experimentation
demonstrates WyFormer's compelling combination of attributes: it
achieves best-in-class symmetry-conditioned generation, incorporates a
physics-motivated inductive bias, produces structures with competitive
stability, predicts material properties with competitive accuracy even
without atomic coordinates, and exhibits unparalleled inference speed.
https://github.com/SymmetryAdvantage/WyckoffTransformer},
note = {42nd International Conference on Machine Learning-ICML-Annual,
Vancouver, CANADA, JUL 13-19, 2025},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Crystal symmetry plays a fundamental role in determining its physical,
chemical, and electronic properties such as electrical and thermal
conductivity, optical and polarization behavior, and mechanical
strength. Almost all known crystalline materials have internal symmetry.
However, this is often inadequately addressed by existing generative
models, making the consistent generation of stable and symmetrically
valid crystal structures a significant challenge. We introduce WyFormer,
a generative model that directly tackles this by formally conditioning
on space group symmetry. It achieves this by using Wyckoff positions as
the basis for an elegant, compressed, and discrete structure
representation. To model the distribution, we develop a
permutation-invariant autoregressive model based on the Transformer
encoder and an absence of positional encoding. Extensive experimentation
demonstrates WyFormer's compelling combination of attributes: it
achieves best-in-class symmetry-conditioned generation, incorporates a
physics-motivated inductive bias, produces structures with competitive
stability, predicts material properties with competitive accuracy even
without atomic coordinates, and exhibits unparalleled inference speed.
https://github.com/SymmetryAdvantage/WyckoffTransformer - FNClarivate Analytics Web of Science
- VR1.0
- PTMisc
- AFNikita Kazeev
Wei Nong
Ignat Romanov
Ruiming Zhu
Andrey Ustyuzhanin
Shuya Yamazaki
Kedar Hippalgaonkar
- TIWyckoff Transformer: Generation of Symmetric Crystals
- DTInproceedings
- ABCrystal symmetry plays a fundamental role in determining its physical,
chemical, and electronic properties such as electrical and thermal
conductivity, optical and polarization behavior, and mechanical
strength. Almost all known crystalline materials have internal symmetry.
However, this is often inadequately addressed by existing generative
models, making the consistent generation of stable and symmetrically
valid crystal structures a significant challenge. We introduce WyFormer,
a generative model that directly tackles this by formally conditioning
on space group symmetry. It achieves this by using Wyckoff positions as
the basis for an elegant, compressed, and discrete structure
representation. To model the distribution, we develop a
permutation-invariant autoregressive model based on the Transformer
encoder and an absence of positional encoding. Extensive experimentation
demonstrates WyFormer's compelling combination of attributes: it
achieves best-in-class symmetry-conditioned generation, incorporates a
physics-motivated inductive bias, produces structures with competitive
stability, predicts material properties with competitive accuracy even
without atomic coordinates, and exhibits unparalleled inference speed.
https://github.com/SymmetryAdvantage/WyckoffTransformer - Z90
- PUJMLR-JOURNAL MACHINE LEARNING RESEARCH
- PA1269 LAW ST, SAN DIEGO, CA, UNITED STATES
- SN2640-3498
- VL267
- BP29495
- EP29526
- UTWOS:001693126000256
- ER
- EF
|
2024
|
Al-Maeeni, Abdalaziz; Lazarev, Mikhail; Kazeev, Nikita; Novoselov, Kostya S; Ustyuzhanin, Andrey Review on automated 2D material design 2D MATERIALS, 11 (3), 2024, DOI: 10.1088/2053-1583/ad4661. Abstract | BibTeX | Endnote @article{WOS:001248928500001,
title = {Review on automated 2D material design},
author = {Abdalaziz Al-Maeeni and Mikhail Lazarev and Nikita Kazeev and Kostya S Novoselov and Andrey Ustyuzhanin},
doi = {10.1088/2053-1583/ad4661},
times_cited = {8},
issn = {2053-1583},
year = {2024},
date = {2024-07-01},
journal = {2D MATERIALS},
volume = {11},
number = {3},
publisher = {IOP Publishing Ltd},
address = {TEMPLE CIRCUS, TEMPLE WAY, BRISTOL BS1 6BE, ENGLAND},
abstract = {Deep learning (DL) methodologies have led to significant advancements in
various domains, facilitating intricate data analysis and enhancing
predictive accuracy and data generation quality through complex
algorithms. In materials science, the extensive computational demands
associated with high-throughput screening techniques such as density
functional theory, coupled with limitations in laboratory production,
present substantial challenges for material research. DL techniques are
poised to alleviate these challenges by reducing the computational costs
of simulating material properties and by generating novel materials with
desired attributes. This comprehensive review document explores the
current state of DL applications in materials design, with a particular
emphasis on two-dimensional materials. The article encompasses an
in-depth exploration of data-driven approaches in both forward and
inverse design within the realm of materials science.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Deep learning (DL) methodologies have led to significant advancements in
various domains, facilitating intricate data analysis and enhancing
predictive accuracy and data generation quality through complex
algorithms. In materials science, the extensive computational demands
associated with high-throughput screening techniques such as density
functional theory, coupled with limitations in laboratory production,
present substantial challenges for material research. DL techniques are
poised to alleviate these challenges by reducing the computational costs
of simulating material properties and by generating novel materials with
desired attributes. This comprehensive review document explores the
current state of DL applications in materials design, with a particular
emphasis on two-dimensional materials. The article encompasses an
in-depth exploration of data-driven approaches in both forward and
inverse design within the realm of materials science. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFAbdalaziz Al-Maeeni
Mikhail Lazarev
Nikita Kazeev
Kostya S Novoselov
Andrey Ustyuzhanin
- TIReview on automated 2D material design
- SO2D MATERIALS
- DTArticle
- ABDeep learning (DL) methodologies have led to significant advancements in
various domains, facilitating intricate data analysis and enhancing
predictive accuracy and data generation quality through complex
algorithms. In materials science, the extensive computational demands
associated with high-throughput screening techniques such as density
functional theory, coupled with limitations in laboratory production,
present substantial challenges for material research. DL techniques are
poised to alleviate these challenges by reducing the computational costs
of simulating material properties and by generating novel materials with
desired attributes. This comprehensive review document explores the
current state of DL applications in materials design, with a particular
emphasis on two-dimensional materials. The article encompasses an
in-depth exploration of data-driven approaches in both forward and
inverse design within the realm of materials science. - Z98
- PUIOP Publishing Ltd
- PATEMPLE CIRCUS, TEMPLE WAY, BRISTOL BS1 6BE, ENGLAND
- SN2053-1583
- VL11
- DI10.1088/2053-1583/ad4661
- UTWOS:001248928500001
- ER
- EF
|
2023
|
Kazeev, Nikita; Al-Maeeni, Abdalaziz Rashid; Romanov, Ignat; Faleev, Maxim; Lukin, Ruslan; Tormasov, Alexander; Neto, Castro A H; Novoselov, Kostya S; Huang, Pengru; Ustyuzhanin, Andrey Sparse representation for machine learning the properties of defects in
2D materials 35 NPJ COMPUTATIONAL MATERIALS, 9 (1), 2023, DOI: 10.1038/s41524-023-01062-z. Abstract | BibTeX | Endnote @article{WOS:001016773900002,
title = {Sparse representation for machine learning the properties of defects in
2D materials},
author = {Nikita Kazeev and Abdalaziz Rashid Al-Maeeni and Ignat Romanov and Maxim Faleev and Ruslan Lukin and Alexander Tormasov and Castro A H Neto and Kostya S Novoselov and Pengru Huang and Andrey Ustyuzhanin},
doi = {10.1038/s41524-023-01062-z},
times_cited = {35},
year = {2023},
date = {2023-06-01},
journal = {NPJ COMPUTATIONAL MATERIALS},
volume = {9},
number = {1},
publisher = {NATURE PORTFOLIO},
address = {HEIDELBERGER PLATZ 3, BERLIN, 14197, GERMANY},
abstract = {Two-dimensional materials offer a promising platform for the next
generation of (opto-) electronic devices and other high technology
applications. One of the most exciting characteristics of 2D crystals is
the ability to tune their properties via controllable introduction of
defects. However, the search space for such structures is enormous, and
ab-initio computations prohibitively expensive. We propose a machine
learning approach for rapid estimation of the properties of 2D material
given the lattice structure and defect configuration. The method
suggests a way to represent configuration of 2D materials with defects
that allows a neural network to train quickly and accurately. We compare
our methodology with the state-of-the-art approaches and demonstrate at
least 3.7 times energy prediction error drop. Also, our approach is an
order of magnitude more resource-efficient than its contenders both for
the training and inference part.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Two-dimensional materials offer a promising platform for the next
generation of (opto-) electronic devices and other high technology
applications. One of the most exciting characteristics of 2D crystals is
the ability to tune their properties via controllable introduction of
defects. However, the search space for such structures is enormous, and
ab-initio computations prohibitively expensive. We propose a machine
learning approach for rapid estimation of the properties of 2D material
given the lattice structure and defect configuration. The method
suggests a way to represent configuration of 2D materials with defects
that allows a neural network to train quickly and accurately. We compare
our methodology with the state-of-the-art approaches and demonstrate at
least 3.7 times energy prediction error drop. Also, our approach is an
order of magnitude more resource-efficient than its contenders both for
the training and inference part. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFNikita Kazeev
Abdalaziz Rashid Al-Maeeni
Ignat Romanov
Maxim Faleev
Ruslan Lukin
Alexander Tormasov
Castro A H Neto
Kostya S Novoselov
Pengru Huang
Andrey Ustyuzhanin
- TISparse representation for machine learning the properties of defects in
2D materials - SONPJ COMPUTATIONAL MATERIALS
- DTArticle
- ABTwo-dimensional materials offer a promising platform for the next
generation of (opto-) electronic devices and other high technology
applications. One of the most exciting characteristics of 2D crystals is
the ability to tune their properties via controllable introduction of
defects. However, the search space for such structures is enormous, and
ab-initio computations prohibitively expensive. We propose a machine
learning approach for rapid estimation of the properties of 2D material
given the lattice structure and defect configuration. The method
suggests a way to represent configuration of 2D materials with defects
that allows a neural network to train quickly and accurately. We compare
our methodology with the state-of-the-art approaches and demonstrate at
least 3.7 times energy prediction error drop. Also, our approach is an
order of magnitude more resource-efficient than its contenders both for
the training and inference part. - Z935
- PUNATURE PORTFOLIO
- PAHEIDELBERGER PLATZ 3, BERLIN, 14197, GERMANY
- VL9
- DI10.1038/s41524-023-01062-z
- UTWOS:001016773900002
- ER
- EF
|
Huang, Pengru; Lukin, Ruslan; Faleev, Maxim; Kazeev, Nikita; Al-Maeeni, Abdalaziz Rashid; Andreeva, Daria V; Ustyuzhanin, Andrey; Tormasov, Alexander; Neto, Castro A H; Novoselov, Kostya S Unveiling the complex structure-property correlation of defects in 2D
materials based on high throughput datasets 47 NPJ 2D MATERIALS AND APPLICATIONS, 7 (1), 2023, DOI: 10.1038/s41699-023-00369-1. Abstract | BibTeX | Endnote @article{WOS:000924124300002,
title = {Unveiling the complex structure-property correlation of defects in 2D
materials based on high throughput datasets},
author = {Pengru Huang and Ruslan Lukin and Maxim Faleev and Nikita Kazeev and Abdalaziz Rashid Al-Maeeni and Daria V Andreeva and Andrey Ustyuzhanin and Alexander Tormasov and A H Castro Neto and Kostya S Novoselov},
doi = {10.1038/s41699-023-00369-1},
times_cited = {47},
year = {2023},
date = {2023-02-01},
journal = {NPJ 2D MATERIALS AND APPLICATIONS},
volume = {7},
number = {1},
publisher = {NATURE PORTFOLIO},
address = {HEIDELBERGER PLATZ 3, BERLIN, 14197, GERMANY},
abstract = {Modification of physical properties of materials and design of materials
with on-demand characteristics is at the heart of modern technology.
Rare application relies on pure materials-most devices and technologies
require careful design of materials properties through alloying,
creating heterostructures of composites, or controllable introduction of
defects. At the same time, such designer materials are notoriously
difficult to model. Thus, it is very tempting to apply machine learning
methods to such systems. Unfortunately, there is only a handful of
machine learning-friendly material databases available these days. We
develop a platform for easy implementation of machine learning
techniques to materials design and populate it with datasets on pristine
and defected materials. Here we introduce the 2D Material Defect (2DMD)
datasets that include defect properties of represented 2D materials such
as MoS2, WSe2, hBN, GaSe, InSe, and black phosphorous, calculated using
DFT. Our study provides a data-driven physical understanding of complex
behaviors of defect properties in 2D materials, holding promise for a
guide to the development of efficient machine learning models. In
addition, with the increasing enrollment of datasets, our database could
provide a platform for designing materials with predetermined
properties.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Modification of physical properties of materials and design of materials
with on-demand characteristics is at the heart of modern technology.
Rare application relies on pure materials-most devices and technologies
require careful design of materials properties through alloying,
creating heterostructures of composites, or controllable introduction of
defects. At the same time, such designer materials are notoriously
difficult to model. Thus, it is very tempting to apply machine learning
methods to such systems. Unfortunately, there is only a handful of
machine learning-friendly material databases available these days. We
develop a platform for easy implementation of machine learning
techniques to materials design and populate it with datasets on pristine
and defected materials. Here we introduce the 2D Material Defect (2DMD)
datasets that include defect properties of represented 2D materials such
as MoS2, WSe2, hBN, GaSe, InSe, and black phosphorous, calculated using
DFT. Our study provides a data-driven physical understanding of complex
behaviors of defect properties in 2D materials, holding promise for a
guide to the development of efficient machine learning models. In
addition, with the increasing enrollment of datasets, our database could
provide a platform for designing materials with predetermined
properties. - FNClarivate Analytics Web of Science
- VR1.0
- PTJ
- AFPengru Huang
Ruslan Lukin
Maxim Faleev
Nikita Kazeev
Abdalaziz Rashid Al-Maeeni
Daria V Andreeva
Andrey Ustyuzhanin
Alexander Tormasov
A H Castro Neto
Kostya S Novoselov
- TIUnveiling the complex structure-property correlation of defects in 2D
materials based on high throughput datasets - SONPJ 2D MATERIALS AND APPLICATIONS
- DTArticle
- ABModification of physical properties of materials and design of materials
with on-demand characteristics is at the heart of modern technology.
Rare application relies on pure materials-most devices and technologies
require careful design of materials properties through alloying,
creating heterostructures of composites, or controllable introduction of
defects. At the same time, such designer materials are notoriously
difficult to model. Thus, it is very tempting to apply machine learning
methods to such systems. Unfortunately, there is only a handful of
machine learning-friendly material databases available these days. We
develop a platform for easy implementation of machine learning
techniques to materials design and populate it with datasets on pristine
and defected materials. Here we introduce the 2D Material Defect (2DMD)
datasets that include defect properties of represented 2D materials such
as MoS2, WSe2, hBN, GaSe, InSe, and black phosphorous, calculated using
DFT. Our study provides a data-driven physical understanding of complex
behaviors of defect properties in 2D materials, holding promise for a
guide to the development of efficient machine learning models. In
addition, with the increasing enrollment of datasets, our database could
provide a platform for designing materials with predetermined
properties. - Z947
- PUNATURE PORTFOLIO
- PAHEIDELBERGER PLATZ 3, BERLIN, 14197, GERMANY
- VL7
- DI10.1038/s41699-023-00369-1
- UTWOS:000924124300002
- ER
- EF
|
