Efficiency enhancement of a thin-cell near-field thermophotovoltaic device via Bayesian optimization

Kose Y. M., ERTÜRK H.

SOLAR ENERGY MATERIALS AND SOLAR CELLS, cilt.301, 2026 (SCI-Expanded, Scopus)

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 301
  • Basım Tarihi: 2026
  • Doi Numarası: 10.1016/j.solmat.2026.114310
  • Dergi Adı: SOLAR ENERGY MATERIALS AND SOLAR CELLS
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Chemical Abstracts Core, Chimica, Compendex, Environment Index, Greenfile, INSPEC, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO)
  • Orta Doğu Teknik Üniversitesi Adresli: Hayır

Özet

This study presents the design of an efficiency-optimized near-field thermophotovoltaic (TPV) device based on thin cells introducing a novel material selection method. Three different cell materials, germanium (Ge), indium arsenide (InAs), and indium gallium antimonide (In1-xGaxSb) are considered in separate optimization studies to identify the optimal dielectric functions of the emitter and the cell substrate metals. Bayesian optimization, which employs a surrogate model to approximate the computationally expensive objective function and an acquisition function to explore and exploit the input space, is used for the first time in a near-field TPV study. The results show the crucial role of the emitter's plasmonic resonant frequency. Additionally, a highly reflective cell substrate, exhibiting a low damping coefficient, is desired to reflect waves that are not absorbed by the thin cell while avoiding a matching with the plasma frequency of the emitter for suppressed undesired coupling. Tungsten exhibits the desired emitter behavior at elevated temperatures due to its properties that change significantly with temperature. Furthermore, copper performs better compared to gold or silver due to its lower attenuation. Using a 1000 K tungsten emitter separated by 100 nm from a thin InAs cell on a copper substrate, the estimated efficiency and power density values reach 31.1% and 18.3 kW/m2, respectively.