Enabling Direct Photoelectrochemical H2 Production using Alternative Oxidation Reactions on WO3.

The efficient and inexpensive conversion of solar energy into chemical bonds, such as in H2 via the photoelectrochemical splitting of H2O, is a promising route to produce green industrial feedstocks and renewable fuels, which is a key goal of the NCCR Catalysis. However, the oxidation product of the water splitting reaction, O2, has little economic or industrial value. Thus, upgrading key chemical species using alternative oxidation reactions is an emerging trend. WO3 has been identified as a unique photoanode material for this purpose since it performs poorly in the oxygen evolution reaction in H2O. Herein we highlight a collaboration in the NCCR Catalysis that has gained insights at the atomic level of the WO3 surface with ab initio computational methods that help to explain its unique catalytic activity. These computational efforts give new context to experimental results employing WO3 photoanodes for the direct photoelectrochemical oxidation of biomass-derived 5-(hydroxymethyl) furfural. While yield for the desired product, 2,5-furandicarboxylic acid is low, insights into the reaction rate constants using kinetic modelling and an electrochemical technique called derivative voltammetry, give indications on how to improve the system.

DeepCV: A Deep Learning Framework for Blind Search of Collective Variables in Expanded Configurational Space.

We present Deep learning for Collective Variables (DeepCV), a computer code that provides an efficient and customizable implementation of the deep autoencoder neural network (DAENN) algorithm that has been developed in our group for computing collective variables (CVs) and can be used with enhanced sampling methods to reconstruct free energy surfaces of chemical reactions. DeepCV can be used to conveniently calculate molecular features, train models, generate CVs, validate rare events from sampling, and analyze a trajectory for chemical reactions of interest. We use DeepCV in an example study of the conformational transition of cyclohexene, where metadynamics simulations are performed using DAENN-generated CVs. The results show that the adopted CVs give free energies in line with those obtained by previously developed CVs and experimental results. DeepCV is open-source software written in Python/C++ object-oriented languages, based on the TensorFlow framework and distributed free of charge for noncommercial purposes, which can be incorporated into general molecular dynamics software. DeepCV also comes with several additional tools, i.e., an application program interface (API), documentation, and tutorials.

Two Novel Dinuclear Cobalt Polypyridyl Complexes in Electro- and Photocatalysis for Hydrogen Production: Cooperativity Increases Performance.

Syntheses and mechanisms of two dinuclear Co-polypyridyl catalysts for the H2 evolution reaction (HER) were reported and compared to their mononuclear analogue (R1). In both catalysts, two di-(2,2'-bipyridin-6-yl)-methanone units were linked by either 2,2'-bipyridin-6,6'-yl or pyrazin-2,5-yl. Complexation with CoII gave dinuclear compounds bridged by pyrazine (C2) or bipyridine (C1). Photocatalytic HER gave turnover numbers (TONs) of up to 20000 (C2) and 7000 (C1) in water. Electrochemically, C1 was similar to the R1, whereas C2 showed electronic coupling between the two Co centers. The E(CoII/I ) split by 360 mV into two separate waves. Proton reduction in DMF was investigated for R1 with [HNEt3 ](BF4 ) by simulation, foot of the wave analysis, and linear sweep voltammetry (LSV) with in-line detection of H2 . All methods agreed well with an (E)ECEC mechanism and the first protonation being rate limiting (≈104  m-1 s-1 ). The second reduction was more anodic than the first one. pKa values of around 10 and 7.5 were found for the two protonations. LSV analysis with H2 detection for all catalysts and acids with different pKa values [HBF4 , pKa (DMF)≈3.4], intermediate {[HNEt3 ](BF4 ), pKa (DMF)≈9.2} to weak [AcOH, pKa (DMF)≈13.5] confirmed electrochemical H2 production, distinctly dependent on the pKa values. Only HBF4 protonated CoI intermediates. The two metals in the dualcore C2 cooperated with an increase in rate to a competitive 105  m-1 s-1 with [HNEt3 ](BF4 ). The overpotential decreased compared to R1 by 100 mV. Chronoamperometry established high stabilities for all catalysts with TONlim of 100 for R1 and 320 for C1 and C2.

A Concise Review on Recent Developments of Machine Learning for the Prediction of Vibrational Spectra.

Machine learning has become more and more popular in computational chemistry, as well as in the important field of spectroscopy. In this concise review, we walk the reader through a short summary of machine learning algorithms and a comprehensive discussion on the connection between machine learning methods and vibrational spectroscopy, particularly for the case of infrared and Raman spectroscopy. We also briefly discuss state-of-the-art molecular representations which serve as meaningful inputs for machine learning to predict vibrational spectra. In addition, this review provides an overview of the transferability and best practices of machine learning in the prediction of vibrational spectra as well as possible future research directions.

Closer Look at Inverse Electron Demand Diels-Alder and Nucleophilic Addition Reactions on s-Tetrazines Using Enhanced Sampling Methods.

Inverse electron demand [4+2] Diels-Alder (iEDDA) reactions as well as unprecedented nucleophilic (azaphilic) additions of R-substituted silyl-enol ethers (where R is Phenyl, Methyl, and Hydrogen) to 1,2,4,5-tetrazine (s-tetrazine) catalyzed by BF 3 have recently been discovered (Simon et al. in Org Lett 23(7):2426-2430, 2021), where static calculations were employed for calculation of activation energies. In order to have a more realistic dynamic description of these reactions in explicit solution at ambient conditions, in this work we use a semiempirical tight-binding method combined with enhanced sampling techniques to calculate free energy surfaces (FESs) of the iEDDA and azaphilic addition reactions. Relevant products of not only s-tetrazine but also its derivatives such as BF 3 -mediated s-tetrazine adducts are investigated. We reconstruct the FESs of the iEDDA and azaphilic addition reactions using metadynamics and blue moon ensemble, and compare the ability of different collective variables (CVs) including bond distances, Social PeRmutation INvarianT (SPRINT) coordinates, and path-CV to describe the reaction pathway. We find that when a bulky Phenyl is used as a substituent at the dienophile the azaphilic addition is preferred over the iEDDA reaction. In addition, we also investigate the effect of BF 3 in the diene and steric hindrance in the dienophile on the competition between the iEDDA and azaphilic addition reactions, providing chemical insight for reaction design. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11244-021-01516-y.

Machine Learning-Assisted Discovery of Hidden States in Expanded Free Energy Space.

Collective variables (CVs) are crucial parameters in enhanced sampling calculations and strongly impact the quality of the obtained free energy surface. However, many existing CVs are unique to and dependent on the system they are constructed with, making the developed CV non-transferable to other systems. Herein, we develop a non-instructor-led deep autoencoder neural network (DAENN) for discovering general-purpose CVs. The DAENN is used to train a model by learning molecular representations upon unbiased trajectories that contain only the reactant conformers. The prior knowledge of nonconstraint reactants coupled with the here-introduced topology variable and loss-like penalty function are only required to make the biasing method able to expand its configurational (phase) space to unexplored energy basins. Our developed autoencoder is efficient and relatively inexpensive to use in terms of a priori knowledge, enabling one to automatically search for hidden CVs of the reaction of interest.

How ab initio Molecular Dynamics Can Change the Understanding on Transition Metal Catalysed Water Oxidation.

Artificial water splitting is a promising technology that allows the storage of renewable energy in the form of energy-rich compounds. This mini-review showcases how theoretical studies contribute to the under-standing of existing water oxidation catalysts (WOCs) as well as inspiring the development of novel WOCs. In order to understand the chemical complexity of transition metal complexes and their interaction with the solvent environment, the use of sophisticated simulation protocols is necessary. As an illustration, a family of ruthe- nium-based WOCs is presented which were investigated employing a wide range of forefront computational methods with emphasis on ab initiomolecular dynamic based approaches. In those studies a base assisted oxygen-oxygen bond formation was identified as the energetically most favourable reaction mechanism. By examining the role of local environmental effects at ambient temperature and the effect of modifications in the ligand framework, a comprehensible picture of the WOCs can be given, where the latter can serve as a guideline for further experimental and computational studies. In this mini-review, we provide a description of the methods, and the findings of our previous computational studies in compacted form, aimed at scientists with a theoretical as well as experimental background.

OctaDist: a tool for calculating distortion parameters in spin crossover and coordination complexes.

OctaDist is an interactive and visual program for determination of structural distortion in octahedral coordination complexes such as spin crossover complexes, single-ion magnets, perovskites or metal-organic frameworks. OctaDist computes the octahedral distortion parameters initially designed in the context of the spin-crossover phenomenon and denoted ζ, Σ, and Θ from standard structural files. The program also provides additional tools for molecular analyses and visualization. It emphasizes performance, flexibility, ease of use, application programming interface (API) consistency, and clear documentation. The modules and classes in OctaDist can be easily customized to include new algorithms or analytical tools. OctaDist is cross-platform supported for modern operating systems and is available as open-source distributed under the GNU General Public License version 3.

Solvatomorphism and anion effects in predominantly low spin iron(iii) Schiff base complexes.

A series of iron(iii) complexes [Fe(naphEen)2]X·sol (naphEen = 1-{[2-(ethylamino)-ethylimino]methyl}-2-naphtholate; X = F, sol = 0.5CH2Cl2·H2O 1; sol = H2O, X = Cl, 2 and X = Br 3) and [Fe(naphEen)2]I 4 has been prepared. The UV-Vis spectra reveal clear differences for 1 which DFT/TDDFT calculations suggest are due to an equilibrium between [Fe(naphEen)2]F and [Fe(naphEen)2F], the latter having a coordinated F ligand. The X-ray crystal structures of 2-4 show LS Fe(iii) centres in all cases and extensive aryl interactions that link the Fe centres into supramolecular squares. In 3 at room temperature the compound loses half an equivalent of water resulting in a change in space group from Monoclinic P21/n to C2/c. Magnetic studies indicate that 1 is trapped in a mixed spin state being ca. 40% HS while 2-4 are effectively low spin up to 350 K. In contrast, Mössbauer spectroscopic studies of 1 indicate a gradual but incomplete spin crossover. The magnetic properties of 2-4 contrast with the related [Fe(salEen-X)2]anion derivatives which are often spin crossover active.