SpectraFP: a new spectra-based descriptor to aid in cheminformatics, molecular characterization and search algorithm applications.

We have developed an algorithm to generate a new spectra-based descriptor, called SpectraFP, in order to digitalize the chemical shifts of 13C NMR spectra, as well as potentially important data from other spectroscopic techniques. This descriptor is a fingerprint vector with defined sizes and values of 0 and 1, with the ability to correct chemical shift fluctuations. To explore the applicability of SpectraFP, we outlined two application scenarios: (1) the prediction of six functional groups by machine learning (ML) models and (2) the search for structures based on the similarity between the query spectrum and spectra in an experimental database, both in the SpectraFP format. For each functional group, five ML models were built and validated following the OECD principles: internal and external validations, applicability domains, and mechanistic interpretations. All the models resulted in high goodness-of-fit for the training and test sets with MCC respectively between 0.626 and 0.909 and 0.653 and 0.917, and J ranging from 0.812 to 0.957 and 0.825 to 0.961. Using the SHAP (SHapley Additive exPlanations) approach, the mechanistic interpretations of the models were explored; the results indicated that the most important variables for model decision making were coherent with the expected chemical shifts for each functional group. Several metrics, including Tanimoto, geometric, arithmetic, and Tversky, can be used to perform the similarity calculation for the search algorithm. This algorithm can also incorporate additional variables, such as the correction parameter and the difference between the amount of signals in the query spectrum and the database spectra, while preserving its high performance speed. We hope that our descriptor can link information from spectroscopic/spectrometric techniques with ML models to expand the possibilities in understanding the field of cheminformatics. All databases and algorithms developed for this work are open sources and freely accessible.

"Transitivity": A Code for Computing Kinetic and Related Parameters in Chemical Transformations and Transport Phenomena.

The Transitivity function, defined in terms of the reciprocal of the apparent activation energy, measures the propensity for a reaction to proceed and can provide a tool for implementing phenomenological kinetic models. Applications to systems which deviate from the Arrhenius law at low temperature encouraged the development of a user-friendly graphical interface for estimating the kinetic and thermodynamic parameters of physical and chemical processes. Here, we document the Transitivity code, written in Python, a free open-source code compatible with Windows, Linux and macOS platforms. Procedures are made available to evaluate the phenomenology of the temperature dependence of rate constants for processes from the Arrhenius and Transitivity plots. Reaction rate constants can be calculated by the traditional Transition-State Theory using a set of one-dimensional tunneling corrections (Bell (1935), Bell (1958), Skodje and Truhlar and, in particular, the deformed ( d -TST) approach). To account for the solvent effect on reaction rate constant, implementation is given of the Kramers and of Collins-Kimball formulations. An input file generator is provided to run various molecular dynamics approaches in CPMD code. Examples are worked out and made available for testing. The novelty of this code is its general scope and particular exploit of d -formulations to cope with non-Arrhenius behavior at low temperatures, a topic which is the focus of recent intense investigations. We expect that this code serves as a quick and practical tool for data documentation from electronic structure calculations: It presents a very intuitive graphical interface which we believe to provide an excellent working tool for researchers and as courseware to teach statistical thermodynamics, thermochemistry, kinetics, and related areas.

Kinetics of the OH+HCl→H2 O+Cl reaction: Rate determining roles of stereodynamics and roaming and of quantum tunneling.

The OH + HCl → H2 O + Cl reaction is one of the most studied four-body systems, extensively investigated by both experimental and theoretical approaches. Here, as a continuation of our previous work on the OH + HBr and OH + HI reactions, which manifest an anti-Arrhenius behavior that was explained by stereodynamic and roaming effects, we extend the strategy to understand the transition to the sub-Arrhenius behavior occurring for the HCl case. As previously, we perform first-principles on-the-fly Born-Oppenheimer molecular dynamics calculations, thermalized at four temperatures (50, 200, 350, and 500 K), but this time we also apply a high-level transition-state-theory, modified to account for tunneling conditions. We find that the theoretical rate constants calculated with Bell tunneling corrections are in good agreement with extensive experimental data available for this reaction in the ample temperature range: (i) simulations show that the roles of molecular orientation in promoting this reaction and of roaming in finding the favorable path are minor than in the HBr and HI cases, and (ii) dominating is the effect of quantum mechanical penetration through the energy barrier along the reaction path on the potential energy surface. The discussion of these results provides clarification of the origin on different non-Arrhenius mechanisms observed along this series of reactions. © 2018 Wiley Periodicals, Inc.

A novel assessment of the role of the methyl radical and water formation channel in the CH3OH + H reaction.

A number of experimental and theoretical papers accounted almost exclusively for two channels in the reaction of atomic hydrogen with methanol: H-abstraction from the methyl (R1) and hydroxyl (R2) functional groups. Recently, several astrochemical studies claimed the importance of another channel for this reaction, which is crucial for kinetic simulations related to the abundance of molecular constituents in planetary atmospheres: methyl radical and water formation (R3 channel). Here, motivated by the lack of and uncertainties about the experimental and theoretical kinetic rate constants for the third channel, we developed first-principles Car-Parrinello molecular dynamics thermalized at two significant temperatures - 300 and 2500 K. Furthermore, the kinetic rate constant of all three channels was calculated using a high-level deformed-transition state theory (d-TST) at a benchmark electronic structure level. d-TST is shown to be suitable for describing the overall rate constant for the CH3OH + H reaction (an archetype of the moderate tunnelling regime) with the precision required for practical applications. Considering the experimental ratios at 1000 K, kR1/kR2 ≈ 0.84 and kR1/kR3 ≈ 15-40, we provided a better estimate when compared with previous theoretical work: 7.47 and 637, respectively. The combination of these procedures explicitly demonstrates the role of the third channel in a significant range of temperatures and indicates its importance considering the thermodynamic control to estimate methyl radical and water formation. We expect that these results can help to shed new light on the fundamental kinetic rate equations for the CH3OH + H reaction.