Neural Network Potentials: A Concise Overview of Methods

In the past two decades, machine learning potentials (MLPs) have reached a level of maturity that now enables applications to large-scale atomistic simulations of a wide range of systems in chemistry, physics, and materials science. Different machine learning algorithms have been used with great success in the construction of these MLPs. In this review, we discuss an important group of MLPs relying on artificial neural networks to establish a mapping from the atomic structure to the potential energy. In spite of this common feature, there are important conceptual differences among MLPs, which concern the dimensionality of the systems, the inclusion of long-range electrostatic interactions, global phenomena like nonlocal charge transfer, and the type of descriptor used to represent the atomic structure, which can be either predefined or learnable. A concise overview is given along with a discussion of the open challenges in the field. Expected final online publication date for the Annual Review of Physical Chemistry, Volume 73 is April 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.

Characterizing the Chemical Structure of Ti3C2Tx MXene by Angle-Resolved XPS Combined with Argon Ion Etching

Angle-resolved XPS combined with argon ion etching was used to characterize the surface functional groups and the chemical structure of Ti3C2Tx MXene. Survey scanning obtained on the sample surface showed that the sample mainly contains C, O, Ti and F elements, and a little Al element. Analyzing the angle-resolved narrow scanning of these elements indicated that a layer of C and O atoms was adsorbed on the top surface of the sample, and there were many O or F related Ti bonds except Ti–C bond. XPS results obtained after argon ion etching indicated staggered distribution between C–Ti–C bond and O–Ti–C, F–Ti bond. It is confirmed that Ti atoms and C atoms were at the center layer of Ti3C2Tx MXene, while O atoms and F atoms were located at both the upper and lower surface of Ti3C2 layer acting as surface functional groups. The surface functional groups on the Ti3C2 layer were determined to include O2−, OH−, F− and O−–F−, among which F atoms could also desorb from Ti3C2Tx MXene easily. The schematic atomic structure of Ti3C2Tx MXene was derived from the analysis of XPS results, being consistent with theoretical chemical structure and other experimental reports. The results showed that angle-resolved XPS combing with argon ion etching is a good way to analysis 2D thin layer materials.

Crystal structure of a putative short-chain dehydrogenase/reductase from Paraburkholderia xenovorans

Paraburkholderia xenovorans degrades organic wastes, including polychlorinated biphenyls. The atomic structure of a putative dehydrogenase/reductase (SDR) from P. xenovorans (PxSDR) was determined in space group P21 at a resolution of 1.45 Å. PxSDR shares less than 37% sequence identity with any known structure and assembles as a prototypical SDR tetramer. As expected, there is some conformational flexibility and difference in the substrate-binding cavity, which explains the substrate specificity. Uniquely, the cofactor-binding cavity of PxSDR is not well conserved and differs from those of other SDRs. PxSDR has an additional seven amino acids that form an additional unique loop within the cofactor-binding cavity. Further studies are required to determine how these differences affect the enzymatic functions of the SDR.