WCIP-based modeling and deep neural network–driven optimization of a terahertz sensor for IoT-enabled green hydrogen monitoring

Reliable hydrogen monitoring is a critical requirement for the safe deployment of emerging green hydrogen infrastructures. In this work, a compact terahertz (THz) hydrogen sensor combining physics-based electromagnetic modeling and data-driven optimization is proposed. The sensing platform consists of three laterally cascaded spiral resonators functionalized with graphene, black phosphorus (BP), and vanadium dioxide (VO \(_2\) ), enabling strong near-field confinement and tunable electromagnetic interaction with the surrounding gas medium. The device is analyzed using an extended Wave Concept Iterative Process (WCIP), which provides an efficient full-wave framework for modeling planar structures incorporating ultrathin conductive and anisotropic materials without volumetric discretization. Hydrogen sensing is evaluated through concentration-dependent resonance-frequency shifts associated with the modeled hydrogen-induced perturbation of the VO \(_2\) sensing region, while the intrinsic refractive-index sensitivity is evaluated separately. The investigated concentrations range from 500 ppm to 40 000 ppm H \(_2\) , covering ppm-level concentrations up to approximately the lower explosive limit (LEL). The optimized configuration achieves sensitivities ranging from 1.22 to 1.39 THz/RIU while maintaining sharp spectral responses and quality factors between 22.2 and 25.8. To accelerate the design exploration process, a deep neural network (DNN) surrogate model is developed to learn the nonlinear relationship between material parameters and sensing performance. Comparative evaluation against several machine-learning algorithms demonstrates the superior predictive accuracy of the proposed DNN, enabling rapid multi-parameter optimization beyond conventional full-wave simulations. Finally, the proposed sensing platform is discussed as a potential component of a 6G-enabled Internet-of-Things (IoT) monitoring architecture for distributed hydrogen-safety monitoring. The proposed framework provides a scalable approach for high-resolution THz hydrogen sensing in next-generation smart energy systems.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2026-11-17
DOI
https://doi.org/10.1038/s41598-026-77548-9
Primary Topic
Microwave and Dielectric Measurement Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

WCIP-based modeling and deep neural network–driven optimization of a terahertz sensor for IoT-enabled green hydrogen monitoring

Turki M. Alanazi, Khalid Faleh Alsirhani, Anis Sahbani, Aymen Hlali
2 links
Scientific Reports
Microwave and Dielectric Measurement Techniques
article

WCIP-based modeling and deep neural network–driven optimization of a terahertz sensor for IoT-enabled green hydrogen monitoring

Turki M. Alanazi, Khalid Faleh Alsirhani, Anis Sahbani, Aymen Hlali
article en

Abstract

Reliable hydrogen monitoring is a critical requirement for the safe deployment of emerging green hydrogen infrastructures. In this work, a compact terahertz (THz) hydrogen sensor combining physics-based electromagnetic modeling and data-driven optimization is proposed. The sensing platform consists of three laterally cascaded spiral resonators functionalized with graphene, black phosphorus (BP), and vanadium dioxide (VO \(_2\) ), enabling strong near-field confinement and tunable electromagnetic interaction with the surrounding gas medium. The device is analyzed using an extended Wave Concept Iterative Process (WCIP), which provides an efficient full-wave framework for modeling planar structures incorporating ultrathin conductive and anisotropic materials without volumetric discretization. Hydrogen sensing is evaluated through concentration-dependent resonance-frequency shifts associated with the modeled hydrogen-induced perturbation of the VO \(_2\) sensing region, while the intrinsic refractive-index sensitivity is evaluated separately. The investigated concentrations range from 500 ppm to 40 000 ppm H \(_2\) , covering ppm-level concentrations up to approximately the lower explosive limit (LEL). The optimized configuration achieves sensitivities ranging from 1.22 to 1.39 THz/RIU while maintaining sharp spectral responses and quality factors between 22.2 and 25.8. To accelerate the design exploration process, a deep neural network (DNN) surrogate model is developed to learn the nonlinear relationship between material parameters and sensing performance. Comparative evaluation against several machine-learning algorithms demonstrates the superior predictive accuracy of the proposed DNN, enabling rapid multi-parameter optimization beyond conventional full-wave simulations. Finally, the proposed sensing platform is discussed as a potential component of a 6G-enabled Internet-of-Things (IoT) monitoring architecture for distributed hydrogen-safety monitoring. The proposed framework provides a scalable approach for high-resolution THz hydrogen sensing in next-generation smart energy systems.

Scientific Reports
Centre National de la Recherche Scientifique (FR), University of Carthage (TN), Jouf University (SA), Sorbonne Université (FR), Université Tunis Carthage (TN), University of Hafr Al-Batin (SA), Institut Systèmes Intelligents et de Robotique (FR)
Microwave and Dielectric Measurement Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon