Reduced Order Machine Learning Models for Accurate Prediction of CO2 Capture in Physical Solvents.

CO2 sorption in physical solvents is one of the promising approaches for carbon capture from highly concentrated CO2 streams at high pressures. Identifying an efficient solvent and evaluating its solubility data at different operating conditions are highly essential for effective capture, which generally involves expensive and time-consuming experimental procedures. This work presents a machine learning based ultrafast alternative for accurate prediction of CO2 solubility in physical solvents using their physical, thermodynamic, and structural properties data. First, a database is established with which several linear, nonlinear, and ensemble models were trained through a systematic cross-validation and grid search method and found that kernel ridge regression (KRR) is the optimum model. Second, the descriptors are ranked based on their complete decomposition contributions derived using principal component analysis. Further, optimum key descriptors (KDs) are evaluated through an iterative sequential addition method with the objective of maximizing the prediction accuracy of the reduced order KRR (r-KRR) model. Finally, the study resulted in the r-KRR model with nine KDs exhibiting the highest prediction accuracy with a minimum root-mean-square error (0.0023), mean absolute error (0.0016), and maximum R2 (0.999). Also, the validity of the database created and ML models developed is ensured through detailed statistical analysis.

Chemoinformatics and Machine Learning Approaches for Identifying Antiviral Compounds.

Current pandemics propelled research efforts in unprecedented fashion, primarily triggering computational efforts towards new vaccine and drug development as well as drug repurposing. There is an urgent need to design novel drugs with targeted biological activity and minimum adverse reactions that may be useful to manage viral outbreaks. Hence an attempt has been made to develop Machine Learning based predictive models that can be used to assess whether a compound has the potency to be antiviral or not. To this end, a set of 2358 antiviral compounds were compiled from the CAS COVID-19 antiviral SAR dataset whose activity was reported based on IC50 value. A total 1157 two-dimensional molecular descriptors were computed among which, the most highly correlated descriptors were selected using Tree-based, Correlation-based and Mutual information-based feature selection methods. Seven Machine Learning algorithms i. e., Random Forest, XGBoost, Support Vector Machine, KNN, Decision Tree, MLP Classifier and Logistic Regression were benchmarked. The best performance was achieved by the models developed using Random Forest and XGBoost algorithms in all the feature selection methods. The maximum predictive accuracy of both these models was 88 % with internal validation. Whereas, with an external dataset, a maximum accuracy of 93.10 % for XGBoost and 100 % for Random Forest based model was achievable. Furthermore, the study demonstrated scaffold analysis of the molecules as a pragmatic approach to explore the importance of structurally diverse compounds in data driven studies.

Combined Experimental and Computational Study of the Gelation of Cyclohexane-Based Bis(acyl-semicarbazides) and the Multi-Stimuli-Responsive Properties of Their Gels.

The current study reports the one-step synthesis and gelation properties of cyclohexane-based bis(acyl-semicarbazide) gelators with an additional -NH group incorporated into urea moieties and carrying hydrophobic chains of varying length (C8-C18). The gels exhibited thermoreversibility and could be tuned in the presence of anions at different concentrations in addition their the ultrasound-responsive nature, thus making them multi-stimuli-responsive. The combined experimental and computational study on these gels reveals that the balance between two noncovalent interactions, viz., hydrogen bonding between the amide groups in acyl-semicarbazide moieties and van der Waals forces between long hydrocarbon tails, is found to be the determining factor in the process of organogelation. A systematic increase in alkyl chain length leads to equilibrium between these two types of noncovalent forces that is manifested in the spectral and thermal properties of the gels. The H-bonding interactions dominated up to a certain chain length, and further increases in the alkyl chain length led to increased van der Waals interactions as observed by IR, XRD, and thermal studies. Computational calculations were carried out on dimer structures of C8-C18 to understand the variation in noncovalent forces responsible for aggregate formation in the gel state as a function of the alkyl chain length. The results indicate that both intermolecular and intramolecular hydrogen bonding stabilize the aggregate structures. Supramolecular aggregation in the gel state led to the viscoelastic nature of the gels, and the addition of anions led to the disruption of self-assembly, which was studied by rheology.

Benzimidazole-functionalized ancillary ligands for heteroleptic Ru(II) complexes: synthesis, characterization and dye-sensitized solar cell applications.

We have designed and synthesized heteroleptic Ru(ii) complexes with a pyridine-benzimidazole ligand (PYBI) for dye-sensitized solar cell (DSSC) applications. The PYBI ligand has major advantages by having the flexibility to introduce appropriate substituents at the four readily available positions through molecular engineering () compared to other ancillary bipyridyl-based ligands. We have substituted position A of the PYBI ligand with either electron-releasing triphenylamine () or pyrene (). We have also introduced 2-hexylthiophene at position A and 3,5-di tert-butyl phenyl group at position B of the PYBI ligand (). All three heteroleptic Ru(ii) complexes have been characterized by mass spectrometry, (1)H NMR, and absorption and emission spectroscopies as well as electrochemical methods. The absorption spectrum of complex is red-shifted and the emission spectrum is blue-shifted, when compared to the standard sensitizer. Testing of these newly designed heteroleptic Ru(ii) sensitizers has revealed that complex exhibits an efficiency of 7.88% using an I(-)/I3(-) redox electrolyte. Experimental observations corroborated by computational calculations have elucidated the high efficiency of complex , primarily due to the fact that the substituents at position A are more influential than those at position of B of the PYBI ligand.

Evaluating the cation binding strength and selectivity of calix[4]pyrroles: a computational and ESI-MS/MS study.

The cation binding strength of calix[4]pyrroles in the gas phase has been evaluated by computational studies and further substantiated by ESI mass spectrometry experiments. The DFT optimized geometries of [CP + X](+) complexes are found to be stable in a 1,3-alternate conformation through cation-π interactions and interestingly CPs are found to be better cation receptor than calix[4]arenes. The binding energy values of [CP + X](+) complexes computed at B2PLYP/TZVP//M05-2X/TZVP follows the binding order, Li(+) > Na(+) > K(+) > Rb(+) > Cs(+). The diameter of Li(+) matches very well with the cavity size of CP and thus is optimally disposed to interact simultaneously with all four pyrrole rings through multiple cation-π interactions. However, other cations, due to the increase in their size, drift away from the cavity center towards the rim of the cavity exhibiting weak cation-π interactions. Energy decomposition analysis (EDA) reveals that the electrostatic and polarization effects act as the major driving force in these interactions. The important outcome of the current study is that the stability of precursor and product ions is found to be crucial in the experimental evaluation of binding affinity of Li(+) and Na(+) complexes of CP. The ESI-MS/MS experiments on the cation complexes of different substituted CPs revealed that the binding strength of CPs towards cations is also dependant on the substituents at the meso-position.

Evaluating the efficacy of amino acids as CO2 capturing agents: a first principles investigation.

Comprehension of the basic concepts for the design of systems for CO2 adsorption is imperative for increasing interest in technology for CO2 capture from the effluents. The efficacy of 20 naturally occurring amino acids (AAs) is demonstrated as the most potent CO2 capturing agents in the process of chemical absorption and physisorption through a systematic computational study using highly parametrized M05-2X/6-311+G(d,p) method. The ability of AAs to bind CO2 both in the noncovalent and covalent fashion and presence of multiple adsorption sites with varying magnitude of binding strengths in all 20 AAs makes them as most promising materials in the process of physisorption. The binding energies (BEs) estimating the strength of noncovalent interaction of AAs and CO2 are calculated and results are interpreted in terms of the nature and strength of the various types of cooperative interactions which are present. The study underlines the possibility to engineer the porous solid materials with extended networks by judiciously employing AA chains as linkers which can substantially augment their efficacy. Results show that a significant increase in the CO2···AA affinity is achieved in the case of AAs with polar neutral side chains. Furthermore, the study proposes AAs as effective alternatives to alkanolamines in chemical dissolution of CO2.

A new family of heteroleptic ruthenium(II) polypyridyl complexes for sensitization of nanocrystalline TiO2 films.

Two new heteroleptic ruthenium(II) photosensitizers that contains 2,2';6,2''-terpyridine with extended π-conjugation with donor groups, a 4,4'-dicarboxylic acid-2,2'-bipyridine anchoring ligand and a thiocyanate ligand have been designed, synthesized and fully characterized by CHN, mass spectrometry, UV-vis and fluorescence spectroscopies and cyclic voltammetry. The new sensitizers have either 3,5-di-tert-butyl phenyl (m-BL-5) or triphenylamine (m-BL-6) groups, where the molar extinction coefficient of both the sensitizers is higher than the analogous ruthenium dyes. Both the sensitizers were tested in dye-sensitized solar cells using two different redox electrolytes.

Further shortening of the C-C single bond in substituted tetrahedranyl tetrahedrane systems: an energy decomposition analysis.

The computational study explores the electronic fine tuning of the exocyclic C-C single bond length in tetrahedranyl tetrahedrane as a function of various substituents. The factors which determine the bond lengths and bond strengths are examined by using the EDA method.