Abstract: The added value of supervised Machine Learning (ML) methods to determine the Absolute Configuration (AC) of compounds from their Vibrational Circular Dichroism (VCD) spectra was explored. Among all ML methods considered, Random Forest (RF) and Feedforward Neural Network (FNN) yield the best performance for identification of the AC. At its best, FNN allows near-perfect AC determination, with accuracy of prediction up to 0.995, while RF combines good predictive accuracy (up to 0.940) with the ability to identify the spectral areas important for the identification of the AC. No loss in performance of either model is observed as long as the spectral sampling interval used does not exceed the spectral bandwidth. Increasing the sampling interval proves to be the best method to lower the dimensionality of the input data, thereby decreasing the computational cost associated with the training of the models.

Abstract: Using first-principles calculations, we study the effect of embedding various atoms into the nanopore sites of both C6N6 and C6N8 monolayers. Our results indicate that the embedded atoms significantly affect the electronic and magnetic properties of C6N6 and C6N8 monolayers and lead to extraordinary and multifarious electronic properties, such as metallic, half-metallic, spin-glass semiconductor and dilute-magnetic semiconductor behaviour. Our results reveal that the H atom concentration dramatically affects the C6N6 monolayer. On increasing the H coverage, the impurity states also increase due to H atoms around the Fermi-level. C6N6 shows metallic character when the H atom concentration reaches 6.25%. Moreover, the effect of charge on the electronic properties of both Cr@C6N6 and C@C6N8 is also studied. Cr@C6N6 is a ferromagnetic metal with a magnetic moment of 2.40 mu(B), and when 0.2 electrons are added and removed, it remains a ferromagnetic metal with a magnetic moment of 2.57 and 2.77 mu(B), respectively. Interestingly, one can observe a semi-metal, in which the VBM and CBM in both spin channels touch each other near the Fermi-level. C@C6N8 is a semiconductor with a nontrivial band gap. When 0.2 electrons are removed, it remains metallic, and under excess electronic charge, it exhibits half-metallic behaviour.

Abstract: Two-dimensional Transition Metal Dichalcogenides (TMDs) possessing extraordinary physical properties at reduced dimensionality have attracted interest due to their promise in electronic and optical device applications. However, TMD monolayers can show a broad range of different properties depending on their crystal phase; for example, H phases are usually semiconductors, while the T phases are metallic. Thus, controlling phase transitions has become critical for device applications. In this study, the energetically low-lying crystal structures of pristine and Janus TMDs are investigated by using ab initio Nudged Elastic Band and molecular dynamics simulations to provide a general explanation for their phase stability and transition properties. Across all materials investigated, the T phase is found to be the least stable and the H phase is the most stable except for WTe2, while the T' and T '' phases change places according to the TMD material. The transition energy barriers are found to be large enough to hint that even the higher energy phases are unlikely to undergo a phase transition to a more stable phase if they can be achieved except for the least stable T phase, which has zero barrier towards the T ' phase. Indeed, in molecular dynamics simulations the thermodynamically least stable T phase transformed into the T ' phase spontaneously while in general no other phase transition was observed up to 2100 K for the other three phases. Thus, the examined T ', T '' and H phases were shown to be mostly stable and do not readily transform into another phase. Furthermore, so-called mixed phase calculations considered in our study explain the experimentally observed lateral hybrid structures and point out that the coexistence of different phases is strongly stable against phase transitions. Indeed, stable complex structures such as metal-semiconductor-metal architectures, which have immense potential to be used in future device applications, are also possible based on our investigation.