Programme

O.6 -- Oral, VCTP-51

Date: Monday, 3 August 2026

Time: 15:20 - 15:40

High-performance sensor design based on co-doped 2D materials using first-principles and machine learning approaches

Huynh Anh Huy (1), Nguyen Huu Hung (2), Vo Duy Dat (3,4), Duy Khanh Nguyen (3,4,*)

(1) School of Education, Can Tho University, Can Tho City, Vietnam (2) Faculty of Natural Sciences, Hung Vuong University, Phu Tho Province, Vietnam (3) Laboratory for Computational Physics, Institute for Computational Science and Artificial Intelligence, Van Lang University, Ho Chi Minh City, Vietnam (4) Faculty of Mechanical, Electrical, and Computer Engineering, Van Lang School of Technology, Van Lang University, Ho Chi Minh City, Vietnam *Email: khanh.nguyenduy@vlu.edu.vn or nguyenkhanhphysics2015@gmail.com

Air pollution caused by toxic gases has become a significant global issue through human transportation and industrial activities. Moreover, the potential risks of leakage and diffusion of CO, CO2 and NH3 in case of accidents are extremely important. Thereby, the development of gas sensor flatform with high sensitivity, selectivity and other essential characteristics have been and are being vigorously promoted. In this work, first-principles calculations with effective machine learning approach were conducted to explore the adsorption of CO, CO2, and NH3 molecules on B–O and B–N co-doped MoS2 monolayer surfaces at atomic scale. The B–O and B–N co-doping creates active sites, improves the adsorption capacity and interaction with gas molecules of the MoS2 monolayer, thereby improving its gas selectivity. The co-doping reveals good selectivity toward CO and outstanding selectivity toward NH3, however, manifesting limited response to CO2. Most prominently, NH3 exhibits outstanding electronic sensitivity, gas selectivity, and recovery behavior on the B–O and B–N co-doped MoS2 monolayer surfaces. The electronic sensitivity of NH3 increases by approximately 400% for B–O–MoS2 and 900% for B–N–MoS2 compared with the non-adsorbed state, while the work function change reaches –1.347 eV and –1.269 eV, respectively. Furthermore, the recovery time for NH3 decreases monotonically with increasing temperature and converges to about 31 s (B–O–MoS2) and 25 s (B–N–MoS2). These results indicate that hetero-co-doping with B–N/B–O significantly enhances the electronic properties and key sensing metrics of the MoS2 monolayer toward CO and NH3 molecules; importantly, showing strong potential for practical gas-sensing applications under high-temperature operating conditions in the case of NH3 adsorption.

Presenter: Nguyen Duy Khanh