Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters: A Computer Vision-Based Transfer Learning Approach

Can Polat; Hasan Kurban; Erchin Serpedin; Mustafa Kurban

doi:10.1109/CPE-POWERENG63314.2025.11027214

Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters: A Computer Vision-Based Transfer Learning Approach

Can Polat
, Hasan Kurban^*
, Erchin Serpedin
, Mustafa Kurban^*

^*Corresponding author for this work

HBKU College of Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

This study investigates the size-dependent hydrogen (H₂) storage capacity of magic-sized copper nanoclusters (Cu NCs) by analyzing their structural and electronic properties along with formation energy. Density functional tight binding (DFTB) calculations were performed to determine the HOMO and LUMO energy levels, energy gap (E_g), Fermi level (E_f), and formation energy (E_F) for different Cu NC sizes interacting with H₂. The results reveal a strong correlation between cluster size and H₂ storage potential, with smaller clusters exhibiting more favorable adsorption characteristics. However, as the system size increases, the computational cost of DFT calculations rises significantly. To address this, an machine learning approach was employed using image-based representations of Cu NCs with transfer learning. This method enabled rapid predictions of electronic properties with DFT-level accuracy while significantly reducing computational time. The model converged in just a few epochs, capturing the size-dependent trends in H₂ adsorption. These findings highlight the potential of AI-driven techniques for accelerating material discovery and optimizing nanoscale hydrogen storage systems.

Original language	English
Title of host publication	2025 Ieee 19th International Conference On Compatibility, Power Electronics And Power Engineering, Cpe-powereng
Publisher	Institute of Electrical and Electronics Engineers Inc.
Number of pages	6
ISBN (Electronic)	9798331515171
ISBN (Print)	979-8-3315-1518-8
DOIs	https://doi.org/10.1109/CPE-POWERENG63314.2025.11027214
Publication status	Published - 22 May 2025
Event	19th IEEE International Conference on Compatibility, Power Electronics and Power Engineering, CPE-POWERENG 2025 - Antalya, Turkey Duration: 20 May 2025 → 22 May 2025

Publication series

Name	Compatibility Power Electronics And Power Engineering

Conference

Conference	19th IEEE International Conference on Compatibility, Power Electronics and Power Engineering, CPE-POWERENG 2025
Country/Territory	Turkey
City	Antalya
Period	20/05/25 → 22/05/25

Keywords

Computational methods
Computer vision
Copper Nanoclusters
Hydrogen storage
Size-dependent
Transfer learning

Access to Document

10.1109/CPE-POWERENG63314.2025.11027214

Cite this

Polat, C., Kurban, H., Serpedin, E., & Kurban, M. (2025). Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters: A Computer Vision-Based Transfer Learning Approach. In 2025 Ieee 19th International Conference On Compatibility, Power Electronics And Power Engineering, Cpe-powereng (Compatibility Power Electronics And Power Engineering). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CPE-POWERENG63314.2025.11027214

Polat, Can ; Kurban, Hasan ; Serpedin, Erchin et al. / Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters : A Computer Vision-Based Transfer Learning Approach. 2025 Ieee 19th International Conference On Compatibility, Power Electronics And Power Engineering, Cpe-powereng. Institute of Electrical and Electronics Engineers Inc., 2025. (Compatibility Power Electronics And Power Engineering).

@inproceedings{0648caf090e64d20b8d73af6b59fcefe,

title = "Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters: A Computer Vision-Based Transfer Learning Approach",

abstract = "This study investigates the size-dependent hydrogen (H2) storage capacity of magic-sized copper nanoclusters (Cu NCs) by analyzing their structural and electronic properties along with formation energy. Density functional tight binding (DFTB) calculations were performed to determine the HOMO and LUMO energy levels, energy gap (Eg), Fermi level (Ef), and formation energy (EF) for different Cu NC sizes interacting with H2. The results reveal a strong correlation between cluster size and H2 storage potential, with smaller clusters exhibiting more favorable adsorption characteristics. However, as the system size increases, the computational cost of DFT calculations rises significantly. To address this, an machine learning approach was employed using image-based representations of Cu NCs with transfer learning. This method enabled rapid predictions of electronic properties with DFT-level accuracy while significantly reducing computational time. The model converged in just a few epochs, capturing the size-dependent trends in H2 adsorption. These findings highlight the potential of AI-driven techniques for accelerating material discovery and optimizing nanoscale hydrogen storage systems.",

keywords = "Computational methods, Computer vision, Copper Nanoclusters, Hydrogen storage, Size-dependent, Transfer learning",

author = "Can Polat and Hasan Kurban and Erchin Serpedin and Mustafa Kurban",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 19th IEEE International Conference on Compatibility, Power Electronics and Power Engineering, CPE-POWERENG 2025 ; Conference date: 20-05-2025 Through 22-05-2025",

year = "2025",

month = may,

day = "22",

doi = "10.1109/CPE-POWERENG63314.2025.11027214",

language = "English",

isbn = "979-8-3315-1518-8",

series = "Compatibility Power Electronics And Power Engineering",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2025 Ieee 19th International Conference On Compatibility, Power Electronics And Power Engineering, Cpe-powereng",

address = "United States",

}

Polat, C, Kurban, H, Serpedin, E & Kurban, M 2025, Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters: A Computer Vision-Based Transfer Learning Approach. in 2025 Ieee 19th International Conference On Compatibility, Power Electronics And Power Engineering, Cpe-powereng. Compatibility Power Electronics And Power Engineering, Institute of Electrical and Electronics Engineers Inc., 19th IEEE International Conference on Compatibility, Power Electronics and Power Engineering, CPE-POWERENG 2025, Antalya, Turkey, 20/05/25. https://doi.org/10.1109/CPE-POWERENG63314.2025.11027214

Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters: A Computer Vision-Based Transfer Learning Approach. / Polat, Can; Kurban, Hasan; Serpedin, Erchin et al.
2025 Ieee 19th International Conference On Compatibility, Power Electronics And Power Engineering, Cpe-powereng. Institute of Electrical and Electronics Engineers Inc., 2025. (Compatibility Power Electronics And Power Engineering).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters

T2 - 19th IEEE International Conference on Compatibility, Power Electronics and Power Engineering, CPE-POWERENG 2025

AU - Polat, Can

AU - Kurban, Hasan

AU - Serpedin, Erchin

AU - Kurban, Mustafa

PY - 2025/5/22

Y1 - 2025/5/22

N2 - This study investigates the size-dependent hydrogen (H2) storage capacity of magic-sized copper nanoclusters (Cu NCs) by analyzing their structural and electronic properties along with formation energy. Density functional tight binding (DFTB) calculations were performed to determine the HOMO and LUMO energy levels, energy gap (Eg), Fermi level (Ef), and formation energy (EF) for different Cu NC sizes interacting with H2. The results reveal a strong correlation between cluster size and H2 storage potential, with smaller clusters exhibiting more favorable adsorption characteristics. However, as the system size increases, the computational cost of DFT calculations rises significantly. To address this, an machine learning approach was employed using image-based representations of Cu NCs with transfer learning. This method enabled rapid predictions of electronic properties with DFT-level accuracy while significantly reducing computational time. The model converged in just a few epochs, capturing the size-dependent trends in H2 adsorption. These findings highlight the potential of AI-driven techniques for accelerating material discovery and optimizing nanoscale hydrogen storage systems.

AB - This study investigates the size-dependent hydrogen (H2) storage capacity of magic-sized copper nanoclusters (Cu NCs) by analyzing their structural and electronic properties along with formation energy. Density functional tight binding (DFTB) calculations were performed to determine the HOMO and LUMO energy levels, energy gap (Eg), Fermi level (Ef), and formation energy (EF) for different Cu NC sizes interacting with H2. The results reveal a strong correlation between cluster size and H2 storage potential, with smaller clusters exhibiting more favorable adsorption characteristics. However, as the system size increases, the computational cost of DFT calculations rises significantly. To address this, an machine learning approach was employed using image-based representations of Cu NCs with transfer learning. This method enabled rapid predictions of electronic properties with DFT-level accuracy while significantly reducing computational time. The model converged in just a few epochs, capturing the size-dependent trends in H2 adsorption. These findings highlight the potential of AI-driven techniques for accelerating material discovery and optimizing nanoscale hydrogen storage systems.

KW - Computational methods

KW - Computer vision

KW - Copper Nanoclusters

KW - Hydrogen storage

KW - Size-dependent

KW - Transfer learning

UR - https://www.scopus.com/pages/publications/105009399622

U2 - 10.1109/CPE-POWERENG63314.2025.11027214

DO - 10.1109/CPE-POWERENG63314.2025.11027214

M3 - Conference contribution

AN - SCOPUS:105009399622

SN - 979-8-3315-1518-8

T3 - Compatibility Power Electronics And Power Engineering

BT - 2025 Ieee 19th International Conference On Compatibility, Power Electronics And Power Engineering, Cpe-powereng

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 20 May 2025 through 22 May 2025

ER -

Polat C, Kurban H, Serpedin E, Kurban M. Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters: A Computer Vision-Based Transfer Learning Approach. In 2025 Ieee 19th International Conference On Compatibility, Power Electronics And Power Engineering, Cpe-powereng. Institute of Electrical and Electronics Engineers Inc. 2025. (Compatibility Power Electronics And Power Engineering). doi: 10.1109/CPE-POWERENG63314.2025.11027214

Data-Efficient Hydrogen Adsorption Prediction in Copper Nanoclusters: A Computer Vision-Based Transfer Learning Approach

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this