Abstract: Direct experimental measurement of thermal expansion coefficient without substrate effects is a challenging task for two-dimensional (2D) materials, and its accurate estimation with large-scale ab initio molecular dynamics is computationally very expensive. Machine learning-based interatomic potentials trained with ab initio data have been successfully used in molecular dynamics simulations to decrease the computational cost without compromising the accuracy. In this study, we investigated using Gaussian approximation potentials to reproduce the density functional theory-level accuracy for graphene within both lattice dynamical and molecular dynamical methods, and to extend their applicability to larger length and time scales. Two such potentials are considered, GAP17 and GAP20. GAP17, which was trained with pristine graphene structures, is found to give closer results to density functional theory calculations at different scales. Further vibrational and structural analyses verify that the same conclusions can be deduced with density functional theory level in terms of the reasoning of the thermal expansion behavior, and the negative thermal expansion behavior is associated with long-range out-of-plane phonon vibrations. Thus, it is argued that the enabled larger system sizes by machine learning potentials may even enhance the accuracy compared to small-size-limited ab initio molecular dynamics.

Abstract: Immobilization of single metal atoms on a solid host opens numerous possibilities for catalyst designs. If that host is a two-dimensional sheet, sheet curvature becomes a design parameter potentially complementary to host and metal composition. Here, we use a combination of density functional theory calculations and microkinetic modeling to compare the mechanisms and kinetics of formic acid decomposition and formation, chosen for their relevance as a potential hydrogen storage medium, over single Ru atoms anchored to pyridinic nitrogen in a planar graphene flake (RuN4-G) and curved carbon nanotube (RuN4-CNT). Activation barriers are lowered and the predicted turnover frequencies are increased over RuN4-CNT relative to RuN4-CNT. The results highlight the potential of curvature control as a means to achieve high performance and robust catalysts.

Abstract: Semiconductor monolayer transition metal dichalcogenides (TMDs) have brought a new paradigm by introducing optically addressable valley degree of freedom. Concomitantly, their high flexibility constitutes a unique platform that links optics to mechanics via valleytronics. With the intention to expedite the research in this direction, we investigated ten TMDs, namely MoS2, MoSe2, MoTe2, WS2, WSe2, WTe2, MoSSe, MoSeTe, WSSe, and WSeTe, which particularly includes their so-called janus types (JTMDs). First, we obtained their electronic band structures using regular and hybrid density functional theory (DFT) calculations in the presence of the spin-orbit coupling and biaxial or uniaxial strain. Our DFT results indicated that against the expectations based on their reported piezoelectric behavior, JTMDs typically interpolated between the standard band properties of the constituent TMDs without producing a novel feature. Next, by fitting to our DFT data we generated both spinless and spinful k center dot p parameter sets which are quite accurate over the K valley where the optical activity occurs. As an important application of this parametrization, we considered the circular and linear dichroism under strain. Among the studied (J)TMDs, WTe2 stood out with its largest linear dichroism under uniaxial strain because of its narrower band gap and large K valley uniaxial deformation potential. This led us to suggest WTe2 monolayer membranes for optical polarization-based strain measurements, or conversely, as strain tunable optical polarizers.