DFT and machine learning investigation of lead-free SiBaX3 (X = Br, I)perovskites for optoelectronic and thermoelectric applications

Liu, Yunlong and Ahmad Khera, Ejaz and Faizan, Muhammad and Shafiee, Saiful Arifin and Nazir, Abrar and Rafiq, Iqra and Peng, Qiong and Wajdan, Muhammad and Almohammedi, Abdullah and Dhahbi, Afef and Sharma, Ramesh (2026) DFT and machine learning investigation of lead-free SiBaX3 (X = Br, I)perovskites for optoelectronic and thermoelectric applications. Physica B: Condensed Matter, 740. p. 419021.

PDF - Published Version Restricted to Repository staff only Download (9MB) | Request a copy

Abstract

In this work, Density Functional Theory (DFT) is used to reveal the structural, optoelectronic and charge transport features of novel SiBaX3 (X = Br and I) perovskites. The Birch-Murnaghan equations of states, tolerance factor, binding energy, formation energy and elastic constants have been used to verify the structural and thermodynamics stability of both examined perovskites. The band structure and total density of states (TDOS) results displays that electronic band gap of 3.89 eV for SiBaBr3 and 3.27 eV for SiBaI3 using TB-mBJ potential. The partial density of states (PDOS) results indicate that the formation of the valence and conduction bands is attributed to the Si-3p, Br-4p, Ba-5d, and I-5p energy states. Regarding optical parameters, the studied perovskites can absorbs electromagnetic radiations of (170–1800 nm) for SiBaBr3 and (165–2490) nm for SiBaI3 covering IR-UV range which makes them suitable materials for optoelectronic applications. The thermoelectric characteristics of the compounds under investigation have been computed using the Boltztrap algorithm, which is patched with the WIEN2K code. The thermoelectric parameters such as Seebeck coefficient, electronic and thermal conductivity, power factor and figure of merit were computed in 100-900K temperature range. SiBaI3 is a viable choice for thermoelectric applications because of its higher figure-of-merit, extraordinary electrical conductivity, power factor and ZT values at room and elevated temperature. This paper integrates machine learning (ML) and DFT to predetermine bandgap, absorption, and Seebeck of SiBaX3 perovskites. The performance of various ML models was compared, in which XGBoost obtained the best accuracy, with the best R and the lowest error measurements. SHAP analysis and feature importance analysis indicate that the optoelectronic, structural, and thermoelectric descriptors are very strong controls of material behavior.

Item Type:	Article (Journal)
Subjects:	Q Science > QD Chemistry
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	UNSPECIFIED
Depositing User:	Dr Saiful 'Arifin Bin Shafiee
Date Deposited:	11 Aug 2026 16:35
Last Update:	11 Aug 2026 16:38
Queue Number:	2026-08-Q4654
URI:	http://irep.iium.edu.my/id/eprint/130696

Actions (login required)

View Item