Search

In this study, Non-Equilibrium Molecular Dynamics (NEMD) simulation is employed to investigate the phonon thermal conductivity (PTC) of Sn/hBN van der Waals heterostructures with different vacancy-induced defects. We deliberately introduce three types of vacancies in Sn/hBN bilayer point vacancies, bivacancies, and edge vacancies at various concentrations ranging from 0.25% to 2%, to examine their effects on PTC across temperatures from 100 K to 600 K. The key findings of our work are (i) PTC declines monotonically with increasing vacancy concentration for all types of vacancies, with a maximum reduction of ∼62% observed at room temperature compared to its pristine form. (ii) The position of defects has an impact on PTC, with a larger decrease observed when defects are present in the hBN layer and a smaller decrease when defects are in the Sn layer. (iii) The type of vacancy also influences PTC, with point vacancies causing the most substantial reduction, followed by bivacancies, and edge vacancies having the least effect. A 2% defect concentration results in a ∼62% decrease in PTC for point vacancies, ∼51% for bivacancies, and ∼32% for edge vacancies. (iv) Finally, our results indicate that for a given defect concentration, PTC decreases as temperature increases. The impact of temperature on thermal conductivity is less pronounced compared to the effect of vacancies for the defective Sn/hBN bilayer. The presence of vacancies and elevated temperatures enhance phonon-defect and phonon–phonon scattering, leading to changes in the phonon density of states (PDOS) profile and the distribution of phonons across different frequencies of Sn/hBN bilayer, thus affecting its thermal conductivity. This work offers new insights into the thermal behavior of vacancy-filled Sn/hBN heterostructures, suggesting potential pathways for modulating thermal conductivity in bilayer van der Waals heterostructures for applications in thermoelectric, optoelectronics, and nanoelectronics in future. [ABSTRACT FROM AUTHOR]

Machine learning methods are increasingly used in research to save time and computational expenses. The data-driven approach relies on existing data for predictions based on statistical inference and interpolation. Liquids, in particular, benefit from machine learning due to their higher computational requirements. In this study, we propose a simple data-driven strategy that employs classical molecular dynamics simulations to generate potential energy spatial distribution data. This data is then used to train a machine learning model by using topologically similar scaled-down systems. The trained model is subsequently employed to predict force field behaviour for complex full-scale systems, resulting in time and computational cost savings. During training, the potential energy pseudo field serves as a model descriptor, while the final predictions are generated using multidimensional regression-based learning (Gaussian Process) and neural architecture learning (Artificial Neural Network), which are integrated into the simulation as lookup tables. Comparing the data-driven and conventional approaches reveals a significant acceleration in overall simulation runtime when appropriately trained machine learning models are utilised. Although the initial training phase of the machine learning models is time-consuming, retraining is unnecessary for future simulations with the same setup. Thus, this approach offers a straightforward means of conducting complex simulations in less time. [ABSTRACT FROM AUTHOR]