Benchmarking Density Functionals, Basis Sets, and Solvent Models in Predicting Thermodynamic Hydricities of Organic Hydrides

Many renewable energy technologies, such as hydrogen gas synthesis and carbon dioxide reduction, rely on chemical reactions involving hydride anions. When selecting molecules to be used in such applications, an important quantity to consider is the thermodynamic hydricity, which is the free energy required for a species to donate a hydride anion. Theoretical calculations of thermodynamic hydricity depend on several parameters, mainly the density functional, basis set, and solvent model. In order to assess the effects of the above three parameters, we carry out hydricity calculations for a set of molecules with known experimental hydricity values, generate linear �fits, and compare the R-squared, root-mean-squared error, and Akaike Information Criterion across different combinations of density functionals, basis sets, and solvent models. Based on these results we are able to quantify the accuracy of theoretical predictions of hydricity and recommend the parameters with the best compromise between accuracy and computational cost.

Beyond Continuum Solvent Models in Computational Homogeneous Catalysis

In homogeneous catalysis solvent is an inherent part of the catalytic system. As such, it must be considered in the computational modeling. The most common approach to include solvent effects in quantum mechanical calculations is by means of continuum solvent models. When they are properly used, average solvent effects are efficiently captured, mainly those related with solvent polarity. However, neglecting atomistic description of solvent molecules has its limitations, and continuum solvent models all alone cannot be applied to whatever situation. In many cases, inclusion of explicit solvent molecules in the quantum mechanical description of the system is mandatory. The purpose of this article is to highlight through selected examples what are the reasons that urge to go beyond the continuum models to the employment of micro-solvated (cluster-continuum) of fully explicit solvent models, in this way setting the limits of continuum solvent models in computational homogeneous catalysis. These examples showcase that inclusion of solvent molecules in the calculation not only can improve the description of already known mechanisms but can yield new mechanistic views of a reaction. With the aim of systematizing the use of explicit solvent models, after discussing the success and limitations of continuum solvent models, issues related with solvent coordination and solvent dynamics, solvent effects in reactions involving small, charged species, as well as reactions in protic solvents and the role of solvent as reagent itself are successively considered.

Implicit solvent models fail to reproduce secondary structures of de novo designed peptides

We test a range of standard implicit solvent models and protein forcefields for a set of 5 experimentally characterized, designed α-helical peptides. 65 combinations of forcefield and implicit solvent models are evaluated in >800 µs of molecular dynamics simulations. The data show that implicit solvent models generally fail to reproduce the experimentally observed secondary structure content, and none performs well for all 5 peptides. The results show that these models are not usefully predictive.

Bridging the experiment-calculation divide: machine learning corrections to redox potential calculations in implicit and explicit solvent models

Prediction of redox potentials is essential for catalysis and energy storage. Although density functional theory (DFT) calculations have enabled rapid redox potential predictions for numerous compounds, prominent errors persist compared to experimental measurements. In this work, we develop machine learning (ML) models to reduce the errors of redox potential calculations in both implicit and explicit solvent models. Training and testing of the ML correction models are based on the diverse ROP313 dataset with experimental redox potentials measured for organic and organometallic compounds in a variety of solvents. For the implicit solvent approach, our ML models can reduce both the systematic bias and the number of outliers. ML corrected redox potentials also demonstrate less sensitivity to DFT functional choice. For the explicit solvent approach, we significantly reduce the computational costs by embedding the microsolvated cluster in implicit bulk solvent, obtaining converged redox potential results with a smaller solvation shell. This combined implicit-explicit solvent model, together with GPU-accelerated quantum chemistry methods, enabled rapid generation of a large dataset of explicit-solvent-calculated redox potentials for 165 organic compounds, allowing detailed investigation of the error sources in explicit solvent redox potential calculations.

Bridging the experiment-calculation divide: machine learning corrections to redox potential calculations in implicit and explicit solvent models

Prediction of redox potentials is essential for catalysis and energy storage. Although density functional theory (DFT) calculations have enabled rapid redox potential predictions for numerous compounds, prominent errors persist compared to experimental measurements. In this work, we develop machine learning (ML) models to reduce the errors of redox potential calculations in both implicit and explicit solvent models. Training and testing of the ML correction models are based on the diverse ROP313 dataset with experimental redox potentials measured for organic and organometallic compounds in a variety of solvents. For the implicit solvent approach, our ML models can reduce both the systematic bias and the number of outliers. ML corrected redox potentials also demonstrate less sensitivity to DFT functional choice. For the explicit solvent approach, we significantly reduce the computational costs by embedding the microsolvated cluster in implicit bulk solvent, obtaining converged redox potential results with a smaller solvation shell. This combined implicit-explicit solvent model, together with GPU-accelerated quantum chemistry methods, enabled rapid generation of a large dataset of explicit-solvent-calculated redox potentials for 165 organic compounds, allowing detailed investigation of the error sources in explicit solvent redox potential calculations.

Ensemble completeness in conformer sampling: the case of small macrocycles

In this study we compare the three algorithms for the generation of conformer ensembles Biovia BEST, Schrödinger Prime macrocycle sampling (PMM) and Conformator (CONF) form the University of Hamburg, with ensembles derived for exhaustive molecular dynamics simulations applied to a dataset of 7 small macrocycles in two charge states and three solvents. Ensemble completeness is a prerequisite to allow for the selection of relevant diverse conformers for many applications in computational chemistry. We apply conformation maps using principal component analysis based on ring torsions. Our major finding critical for all applications of conformer ensembles in any computational study is that maps derived from MD with explicit solvent are significantly distinct between macrocycles, charge states and solvents, whereas the maps for post-optimized conformers using implicit solvent models from all generator algorithms are very similar independent of the solvent. We apply three metrics for the quantification of the relative covered ensemble space, namely cluster overlap, variance statistics, and a novel metric, Mahalanobis distance, showing that post-optimized MD ensembles cover a significantly larger conformational space than the generator ensembles, with the ranking PMM > BEST >> CONF. Furthermore, we find that the distributions of 3D polar surface areas are very similar for all macrocycles independent of charge state and solvent, except for the smaller and more strained compound 7, and that there is also no obvious correlation between 3D PSA and intramolecular hydrogen bond count distributions.

Mechanisms of a Cyclobutane-Fused Lactone Hydrolysis in Alkaline and Acidic Conditions

Searching for functional polyesters with stability and degradability is important due to their potential applications in biomedical supplies, biomass fuel, and environmental protection. Recently, a cyclobutane-fused lactone (CBL) polymer was experimentally found to have superior stability and controllable degradability through hydrolysis reactions after activation by mechanical force. In order to provide a theoretical basis for developing new functional degradable polyesters, in this work, we performed a detailed quantum chemical study of the alkaline and acidic hydrolysis of CBL using dispersion-corrected density functional theory (DFT-D3) and mixed implicit/explicit solvent models. Various possible hydrolysis mechanisms were found: BAC2 and BAL2 in the alkaline condition and AAC2, AAL2, and AAL1 in the acidic condition. Our calculations indicated that CBL favors the BAC2 and AAC2 mechanisms in alkaline and acidic conditions, respectively. In addition, we found that incorporating explicit water solvent molecules is highly necessary because of their strong hydrogen-bonding with reactant/intermediate/product molecules.

TD-DFT/DFT study of Toluidine blue O in aqueous solution: vibronic transitions and electronic properties

The vibronic absorption spectrum of Toluidine blue O (TBO) dye in an aqueous solution was calculated using the time-dependent density functional theory (TD-DFT). The calculations were performed using all hybrid functionals supported by Gaussian16 software and 6-31++G(d,p) basis set with IEFPCM and SMD solvent models. The IEFPCM gave underestimated values of λmax in comparison with the experiment, what is a manifestation of the TD-DFT “cyanine failure”. However, the SMD made it possible to obtain good agreement between calculated and experimental spectra. The best fit was achieved using the X3LYP functional. The dipole moments and atomic charges of the ground and excited states of the TBO molecule were calculated. Photoexcitation leads to an increase in the dipole moment of the dye molecule. An insignificant photoinduced electron transfer was found in the central ring of the chromophore of the TBO molecule. Vibronic transitions play a significant role in the absorption spectrum of the dye.

Thermodynamic Cyclic Voltammograms: Peak Positions and Shapes

Based on a mean-field description of thermodynamic cyclic voltammograms (CVs), we analyse here in full generality, how CV peak positions and shapes are related to the underlying interface energetics, in particular when also including electrostatic double layer (DL) effects. We show in particular, how non-Nernstian behaviour is related to capacitive DL charging, and how this relates to common adsorbate-centered interpretations such as a changed adsorption energetics due to dipole-field interactions and the electrosorption valency -- the number of exchanged electrons upon electrosorption per adsorbate. Using Ag(111) in halide-containing solutions as test case, we demonstrate that DL effects can introduce peak shifts that are already explained by rationalizing the interaction of isolated adsorbates with the interfacial fields, while alterations of the peak shape are mainly driven by the coverage-dependence of the adsorbate dipoles. In addition, we analyse in detail how changing the experimental conditions such as the ion concentrations in the solvent but also of the background electrolyte can affect the CV peaks via their impact on the potential drop in the DL and the DL capacitance, respectively. These results suggest new routes to analyse experimental CVs and use of those for a detailed assessment of the accuracy of atomistic models of electrified interfaces e.g. with and without explicitly treated interfacial solvent and/or approximate implicit solvent models.