Hochdimensionale neuronale Netze für Potentialhyperflächen großer molekularer und kondensierter Systeme

Integrated Micro Space Electrostatic Field in Aqueous Zn-Ion Battery: Scalable Electrospray Fabrication of Porous Crystalline Anode Coating

The inhomogeneous consumption of anions and direct contact between electrolyte and anode during the Zn-deposition process generate Zn-dendrites and side reactions that can aggravate the space-charge effect to hinder the practical implementation of zinc-metal batteries (ZMBs). Herein, electrospray has been applied for the scalable fabrication (>10 000 cm² in a batch-experiment) of hetero-metallic cluster covalent-organic-frameworks (MCOF-Ti₆ Cu₃ ) nanosheet-coating (MNC) with integrated micro space electrostatic field for ZMBs anode protection. The MNC@Zn symmetric cell presents ultralow overpotential (≈72.8 mV) over 10 000 cycles at 1 mAh cm^-2 with 20 mA cm^-2 , which is superior to bare Zn and state-of-the-art porous crystalline materials. Theoretical calculations reveal that MNC with integrated micro space electrostatic field can facilitate the deposition-kinetic and homogenize the electric field of anode to significantly promote the lifespan of ZMBs.

Towards Data-driven Design of Asymmetric Hydrogenation of Olefins: Database and Hierarchical Learning

Asymmetric hydrogenation of olefins is one of the most powerful asymmetric transformations in molecular synthesis. Although several privileged catalyst scaffolds are available, the catalyst development for asymmetric hydrogenation is still a time- and resource-consuming process due to the lack of predictive catalyst design strategy. Targeting the data-driven design of asymmetric catalysis, we herein report the development of a standardized database that contains the detailed information of over 12000 literature asymmetric hydrogenations of olefins. This database provides a valuable platform for the machine learning applications in asymmetric catalysis. Based on this database, we developed a hierarchical learning approach to achieve predictive machine leaning model using only dozens of enantioselectivity data with the target olefin, which offers a useful solution for the few-shot learning problem and will facilitate the reaction optimization with new olefin substrate in catalysis screening.

Hierarchically Structured Allotropes of Phosphorus from Data-Driven Exploration

The discovery of materials is increasingly guided by quantum-mechanical crystal-structure prediction, but the structural complexity in bulk and nanoscale materials remains a bottleneck. Here we demonstrate how data-driven approaches can vastly accelerate the search for complex structures, combining a machine-learning (ML) model for the potential-energy surface with efficient, fragment-based searching. We use the characteristic building units observed in Hittorf's and fibrous phosphorus to seed stochastic ("random") structure searches over hundreds of thousands of runs. Our study identifies a family of hierarchically structured allotropes based on a P8 cage as principal building unit, including one-dimensional (1D) single and double helix structures, nanowires, and two-dimensional (2D) phosphorene allotropes with square-lattice and kagome topologies. These findings yield new insight into the intriguingly diverse structural chemistry of phosphorus, and they provide an example for how ML methods may, in the long run, be expected to accelerate the discovery of hierarchical nanostructures.

Role of Water in the Reaction Mechanism and endo/exo Selectivity of 1,3-Dipolar Cycloadditions Elucidated by Quantum Chemistry and Machine Learning

Asymmetric 1,3-dipolar cycloadditions of azomethine ylides with activated olefins are among the most important and versatile methods for the synthesis of enantioenriched pyrroline and pyrrolidine derivatives. Despite both theoretical and practical importance, the role of water molecules in the reactivity and endo/exo selectivity remains unclear. To explore how water accelerates the reactions and improves the endo/exo selectivity of the cycloadditions of 1,3-dipole phthalazinium-2-dicyanomethanide (1) and two dipolarophiles, an ab initio-quality neural network potential that overcomes the computational bottleneck of explicitly considering water molecules was used. It is demonstrated that not only the nature of both the dipolarophile and the 1,3-dipole, but also the solvent medium, can perturb or even alter the reaction mechanism. An extreme case was found for the reaction of 1,3-dipole 1 with methyl vinyl ketone, in which the reaction mechanism changes from a concerted to a stepwise mode on going from MeCN to H₂ O as solvent, with formation of a zwitterionic intermediate that is a very shallow minimum on the energy surface. Thus, high stereocontrol can still be expected despite the stepwise nature of the mechanism. The results indicate that water can induce global polarization along the reaction coordinate and highlight the role of microsolvation effects and bulk-phase effects in reproducing the experimentally observed aqueous acceleration and enhanced endo/exo selectivity.

Quantifying Chemical Structure and Machine-Learned Atomic Energies in Amorphous and Liquid Silicon

Amorphous materials are being described by increasingly powerful computer simulations, but new approaches are still needed to fully understand their intricate atomic structures. Here, we show how machine-learning-based techniques can give new, quantitative chemical insight into the atomic-scale structure of amorphous silicon (a-Si). We combine a quantitative description of the nearest- and next-nearest-neighbor structure with a quantitative description of local stability. The analysis is applied to an ensemble of a-Si networks in which we tailor the degree of ordering by varying the quench rates down to 1010  K s-1 . Our approach associates coordination defects in a-Si with distinct stability regions and it has also been applied to liquid Si, where it traces a clear-cut transition in local energies during vitrification. The method is straightforward and inexpensive to apply, and therefore expected to have more general significance for developing a quantitative understanding of liquid and amorphous states of matter.

Prediction of Carbon Dioxide Adsorption via Deep Learning

Porous carbons with different textural properties exhibit great differences in CO₂ adsorption capacity. It is generally known that narrow micropores contribute to higher CO₂ adsorption capacity. However, it is still unclear what role each variable in the textural properties plays in CO₂ adsorption. Herein, a deep neural network is trained as a generative model to direct the relationship between CO₂ adsorption of porous carbons and corresponding textural properties. The trained neural network is further employed as an implicit model to estimate its ability to predict the CO₂ adsorption capacity of unknown porous carbons. Interestingly, the practical CO₂ adsorption amounts are in good agreement with predicted values using surface area, micropore and mesopore volumes as the input values simultaneously. This unprecedented deep learning neural network (DNN) approach, a type of machine learning algorithm, exhibits great potential to predict gas adsorption and guide the development of next-generation carbons.

Molecular Modelling for Reactor Design

Chemical reactor modelling based on insights and data on a molecular level has become reality over the last few years. Multiscale models describing elementary reaction steps and full microkinetic schemes, pore structures, multicomponent adsorption and diffusion inside pores, and entire reactors have been presented. Quantum mechanical (QM) approaches, molecular simulations (Monte Carlo and molecular dynamics), and continuum equations have been employed for this purpose. Some recent developments in these approaches are presented, in particular time-dependent QM methods, calculation of van der Waals forces, new approaches for force field generation, automatic setup of reaction schemes, and pore modelling. Multiscale simulations are discussed. Applications of these approaches to heterogeneous catalysis are demonstrated for examples that have found growing interest over the last few years, such as metal-support interactions, influence of pore geometry on reactions, noncovalent bonding, reaction dynamics, dynamic changes in catalyst nanoparticle structure, electrocatalysis, solvent effects in catalysis, and multiscale modelling.