Group Contribution and Machine Learning Approaches to Predict Abraham Solute Parameters, Solvation Free Energy, and Solvation Enthalpy

We present a group contribution method (SoluteGC) and a machine learning model (SoluteML) to predict the Abraham solute parameters, as well as a machine learning model (DirectML) to predict solvation free energy and enthalpy at 298 K. The proposed group contribution method uses atom-centered functional groups with corrections for ring and polycyclic strain while the machine learning models adopt a directed message passing neural network. The solute parameters predicted from SoluteGC and SoluteML are used to calculate solvation energy and enthalpy via linear free energy relationships. Extensive data sets containing 8366 solute parameters, 20,253 solvation free energies, and 6322 solvation enthalpies are compiled in this work to train the models. The three models are each evaluated on the same test sets using both random and substructure-based solute splits for solvation energy and enthalpy predictions. The results show that the DirectML model is superior to the SoluteML and SoluteGC models for both predictions and can provide accuracy comparable to that of advanced quantum chemistry methods. Yet, even though the DirectML model performs better in general, all three models are useful for various purposes. Uncertain predicted values can be identified by comparing the three models, and when the 3 models are combined together, they can provide even more accurate predictions than any one of them individually. Finally, we present our compiled solute parameter, solvation energy, and solvation enthalpy databases (SoluteDB, dGsolvDBx, dHsolvDB) and provide public access to our final prediction models through a simple web-based tool, software packages, and source code.

Residual solvent analysis with hyper-fast gas chromatography-mass spectrometry and a liquid carbon dioxide cryofocusing in less than 90 s

A new hyper-fast gas chromatography method with less than 90 s runtime including the column cool down was developed for the analysis of four gases and 16 residual solvents, combining a CO₂ cryofocusing with a flow-field thermal gradient gas chromatograph (FF-TG-GC) and ToF-MS. The extremely low analysis time can be achieved by combining the new FF-TG-GC and a very short Rxi-624 Sil MS separation column with a small inner diameter and small film thickness (2.05 m × 0.1 mm × 1.0 µm). The column is inserted into a low thermal mass, resistively heated stainless steel capillary. This enables fast temperature programs with heating rates up to 3000 °C/min and a column cool down within a few seconds. In addition to temporal temperature gradients, the FF-TG-GC can generate a spatial temperature gradient that leads to an improved peak shape. Further, an external liquid CO₂ cryo-trap was designed in order to reduce the injection bandwidths of analytes and to take full advantage of the resolving power of the separation column. No modifications are required to the FF-TG-GC for the use of the cryogenic trap, as the cooled spot is heated by the resistively heated stainless steel capillary during the temperature program. With cryofocusing, analyzed residual solvents are baseline separated. R² values over 0.99 for calibration curves and low relative standard deviations (mainly < 3%) for repeatability tests were obtained.

Generalized molecular solvation in non-aqueous solutions by a single parameter implicit solvation scheme

In computer simulations of solvation effects on chemical reactions, continuum modeling techniques regain popularity as a way to efficiently circumvent an otherwise costly sampling of solvent degrees of freedom. As effective techniques, such implicit solvation models always depend on a number of parameters that need to be determined earlier. In the past, the focus lay mostly on an accurate parametrization of water models. Yet, non-aqueous solvents have recently attracted increasing attention, in particular, for the design of battery materials. To this end, we present a systematic parametrization protocol for the Self-Consistent Continuum Solvation (SCCS) model resulting in optimized parameters for 67 non-aqueous solvents. Our parametrization is based on a collection of ≈6000 experimentally measured partition coefficients, which we collected in the Solv@TUM database presented here. The accuracy of our optimized SCCS model is comparable to the well-known universal continuum solvation model (SMx) family of methods, while relying on only a single fit parameter and thereby largely reducing statistical noise. Furthermore, slightly modifying the non-electrostatic terms of the model, we present the SCCS-P solvation model as a more accurate alternative, in particular, for aromatic solutes. Finally, we show that SCCS parameters can, to a good degree of accuracy, also be predicted for solvents outside the database using merely the dielectric bulk permittivity of the solvent of choice.

Estimating system parameters for solvent-water and plant cuticle-water using quantum chemically estimated Abraham solute parameters

Polyparameter Linear Free Energy Relationships (pp-LFERs) using Abraham system parameters have many useful applications. However, developing the Abraham system parameters depends on the availability and quality of the Abraham solute parameters. Using Quantum Chemically estimated Abraham solute Parameters (QCAP) is shown to produce pp-LFERs that have lower root mean square errors (RMSEs) of predictions for solvent-water partition coefficients than parameters that are estimated using other presently available methods. pp-LFERs system parameters are estimated for solvent-water, plant cuticle-water systems, and for novel compounds using QCAP solute parameters and experimental partition coefficients. Refitting the system parameter improves the calculation accuracy and eliminates the bias. Refitted models for solvent-water partition coefficients using QCAP solute parameters give better results (RMSE = 0.278 to 0.506 log units for 24 systems) than those based on ABSOLV (0.326 to 0.618) and QSPR (0.294 to 0.700) solute parameters. For munition constituents and munition-like compounds not included in the calibration of the refitted model, QCAP solute parameters produce pp-LFER models with much lower RMSEs for solvent-water partition coefficients (RMSE = 0.734 and 0.664 for original and refitted model, respectively) than ABSOLV (4.46 and 5.98) and QSPR (2.838 and 2.723). Refitting plant cuticle-water pp-LFER including munition constituents using QCAP solute parameters also results in lower RMSE (RMSE = 0.386) than that using ABSOLV (0.778) and QSPR (0.512) solute parameters. Therefore, for fitting a model in situations for which experimental data exist and system parameters can be re-estimated, or for which system parameters do not exist and need to be developed, QCAP is the quantum chemical method of choice.

Quantum Chemically Estimated Abraham Solute Parameters Using Multiple Solvent-Water Partition Coefficients and Molecular Polarizability

Polyparameter Linear Free Energy Relationships (pp-LFERs), also called Linear Solvation Energy Relationships (LSERs), are used to predict many environmentally significant properties of chemicals. A method is presented for computing the necessary chemical parameters, the Abraham parameters (AP), used by many pp-LFERs. It employs quantum chemical calculations and uses only the chemical's molecular structure. The method computes the Abraham E parameter using density functional theory computed molecular polarizability and the Clausius-Mossotti equation relating the index refraction to the molecular polarizability, estimates the Abraham V as the COSMO calculated molecular volume, and computes the remaining AP S, A, and B jointly with a multiple linear regression using sixty-five solvent-water partition coefficients computed using the quantum mechanical COSMO-SAC solvation model. These solute parameters, referred to as Quantum Chemically estimated Abraham Parameters (QCAP), are further adjusted by fitting to experimentally based APs using QCAP parameters as the independent variables so that they are compatible with existing Abraham pp-LFERs. QCAP and adjusted QCAP for 1827 neutral chemicals are included. For 24 solvent-water systems including octanol-water, predicted log solvent-water partition coefficients using adjusted QCAP have the smallest root-mean-square errors (RMSEs, 0.314-0.602) compared to predictions made using APs estimated using the molecular fragment based method ABSOLV (0.45-0.716). For munition and munition-like compounds, adjusted QCAP has much lower RMSE (0.860) than does ABSOLV (4.45) which essentially fails for these compounds.

Predicting Abraham model solvent coefficients

Background

The Abraham general solvation model can be used in a broad set of scenarios involving partitioning and solubility, yet is limited to a set of solvents with measured Abraham coefficients. Here we extend the range of applicability of Abraham’s model by creating open models that can be used to predict the solvent coefficients for all organic solvents.

Results

We created open random forest models for the solvent coefficients e, s, a, b, and v that had out-of-bag R² values of 0.31, 0.77, 0.92, 0.47, and 0.63 respectively. The models were used to suggest sustainable solvent replacements for commonly used solvents. For example, our models predict that propylene glycol may be used as a general sustainable solvent replacement for methanol.

Conclusion

The solvent coefficient models extend the range of applicability of the Abraham general solvation equations to all organic solvents. The models were developed under Open Notebook Science conditions which makes them open, reproducible, and as useful as possible.

Graphical Abstract

Understanding and modeling removal of anionic organic contaminants (AOCs) by anion exchange resins

Ionic organic contaminants (OCs) are a growing concern for water treatment and the environment and are removed inefficiently by many existing technologies. This study examined removal of anionic OCs by anion exchange resins (AXRs) as a promising alternative. Results indicate that two polystyrene AXRs (IRA910 and IRA96) have higher sorption capacities and selectivity than a polyacrylate resin (A860). For the polystyrene resins, selectivity follows: phenolates ≥ aromatic dicarboxylates > aromatic monocarboxylates > benzenesulfonate > aliphatic carboxylates. This trend can be explained based on hydration energy, the number of exchange groups, and aromaticity and hydrophobicity of the nonpolar moiety (NPM) of the anions. For A860, selectivity only varies within a narrow range (0.13-1.64). Despite the importance of the NPM of the anions, neutral solutes were sorbed much less, indicating synergistic combinations of electrostatic and nonelectrostatic interactions in the overall sorption. By conducting multiple linear regression between Abraham's descriptors and nature log of selectivity, induced dipole-related interactions and electrostatic interactions were found to be the most important interaction forces for sorption of the anions, while solute H-bond basicity has a negative effect. A predictive model was then developed for carboxylates and phenolates based on the poly parameter linear free energy relationships established for a diverse range of 16 anions and 5 neutral solutes, and was validated by accurate prediction of sorption of five test solutes within a wide range of equilibrium concentrations and that of benzoate at different pH.