What question did this study set out to answer?

The research aims to explore the optical characteristics and predictive modeling of chalcogenide thin films.

February 28, 2026Open Access

An experimental/theoretical investigation of optical absorption properties for chalcogenide composition in thin-film form: prediction study using an artificial neural network model

Key Points

The research aims to explore the optical characteristics and predictive modeling of chalcogenide thin films.
Prepared binary Se88Bi12 and ternary Se88Bi6Te6 chalcogenide compositions using melt-quenching technique.
Utilized thermal evaporation on heated glass substrates in a high vacuum environment.
Employed X-ray diffraction to assess the amorphous nature of thin films.
Conducted optical measurements using a spectrophotometer to analyze transmittance and reflectance from 500 to 2500 nm.
Applied artificial neural networks to model the optical energy gap and extinction coefficients.
Identified optical energy gap and Urbach tail, revealing disorder in the material.
Determined optical absorption coefficients and related optical parameters, highlighting composition dependence.
3-hidden layer ANN architecture produced the best model performance based on mean squared error comparisons.
Predicted optical parameters across unmeasured wavelength ranges, contributing to material understanding.

Abstract

Abstract The compositional dependency of the optical characteristics of binary Se₈₈Bi₁₂ Se 88 Bi 12 (SB) and ternary Se₈₈Bi₆Te₆ Se 88 Bi 6 Te 6 (SB–Te) chalcogenide compositions, prepared by melt-quenching technique, was investigated. The thermal evaporation approach was used to form these thin films on heated glass substrates under high vacuum conditions (~ 10 −5 Torr). The X-ray diffraction technique was used to test the amorphous nature of the studied films chick the pattern (XRD). Energy-dispersive X-ray (EDX) analysis was employed to investigate the chemical composition of the samples that were prepared. The optical characteristics of the produced thin films through using a spectrophotometer to collect observational spectrum data for transmittance and reflectance ranging from 500 to 2500 nm wavelength. The transmission and absorption measurements were used to calculate the optical absorption coefficient (α) and the extinction coefficient (k k). By analyzing the spectrum change in the dispersion properties, the optical energy gap (E₆) (E g) and Urbach tail (Eₔ E u) were recently identified. This computation offers vital information on the material's electrical structure and disorder. The transition power factor n confirms that the indirect optical transition with (E₆) (E g) decreases with the Te addition, which is in conjunction with the increase in (Eu) and the increase in the randomness of the films studied. The analysis of optical absorption constants (α), and (k k) was used to determine some important optical factors, including the steepness parameter, optical density, skin depth, and metallization parameters, and study their composition dependence as well. Using experimental data, artificial neural networks (ANNs) were used to model the optical energy gap (E₆) E g), and the extinction coefficient (k k) for chalcogenide thin films. One of the soft computing methods that is important for modeling is ANNS. The mean squared error (MSE) values for various ANN configurations will be compared to determine the best ANN model configuration. In most cases, the 3-hidden layers ANN architecture produced the best results. The optical parameters in the range wavelengths that were not experimentally measured were also predicted using this method.

AI에게 질문

Bookmark

View Full Paper