Prediction of partition and distribution coefficients in various solvent pairs with COSMO-RS

Comparison, optimization and antioxidant activity of ultrasound-assisted natural deep eutectic solvents extraction and traditional method: A greener route for extraction of flavonoid from Moringa oleifera Lam. leaves

To develop an environmentally sustainable and efficient extraction method for flavonoids from Moringa oleifera Lam. (M. oleifera) leaves, natural deep eutectic solvents (NADES) with ultrasound-assisted extraction was utilized in this study. After optimization of extraction parameters of NADES, including ultrasonic power, ultrasonic time, and liquid-solid ratio, the extraction yield of ultrasound-assisted NADES (UAN) composed of betaine and urea (Bet-Urea) reached 54.69 ± 0.19 mg RE/g DW, which made a 1.7-fold increase compared to traditional ultrasound-assisted traditional solvent (UATS). UPLC-Q Exactive/MS analysis revealed that M. oleifera leaves flavonoids (MOLF) was mainly composed of Quercetin 3-β-D-glucoside, Rutin, Kaempferol-3-O-glucoside, Vitexin and Quercetin. Furthermore, the COSMO-RS model was employed to verify the optimal compatibility of solubility and activity coefficient between Bet-Urea and the five primary flavonoids in MOLF. In vitro antioxidant assays verified that MOLF extracted by UAN exhibited superior antioxidant activity compared to MOLF extracted by UATS. Overall, the devised process not only augmented the extraction yield of MOLF but also effectively preserved the bioactive compounds, thus promoting the utilization of green extraction solvents in the food industry.

Designing solvent systems using self-evolving solubility databases and graph neural networks

Designing solvent systems is key to achieving the facile synthesis and separation of desired products from chemical processes, so many machine learning models have been developed to predict solubilities. However, breakthroughs are needed to address deficiencies in the model's predictive accuracy and generalizability; this can be addressed by expanding and integrating experimental and computational solubility databases. To maximize predictive accuracy, these two databases should not be trained separately, and they should not be simply combined without reconciling the discrepancies from different magnitudes of errors and uncertainties. Here, we introduce self-evolving solubility databases and graph neural networks developed through semi-supervised self-training approaches. Solubilities from quantum-mechanical calculations are referred to during semi-supervised learning, but they are not directly added to the experimental database. Dataset augmentation is performed from 11 637 experimental solubilities to >900 000 data points in the integrated database, while correcting for the discrepancies between experiment and computation. Our model was successfully applied to study solvent selection in organic reactions and separation processes. The accuracy (mean absolute error around 0.2 kcal mol-1 for the test set) is quantitatively useful in exploring Linear Free Energy Relationships between reaction rates and solvation free energies for 11 organic reactions. Our model also accurately predicted the partition coefficients of lignin-derived monomers and drug-like molecules. While there is room for expanding solubility predictions to transition states, radicals, charged species, and organometallic complexes, this approach will be attractive to predictive chemistry areas where experimental, computational, and other heterogeneous data should be combined.

Applications of biomass-derived solvents in biomass pretreatment - Strategies, challenges, and prospects

Biomass pretreatment is considered a key step in the 2nd generation biofuel production from lignocellulosic biomass. Research on conventional biomass pretreatment solvents has mainly been focused on carbohydrate conversion efficiency, while their hazardousness and/or carbon intensity were not comprehensively considered. Recent sustainability issues request further consideration for eco-friendly and sustainable alternatives like biomass-derived solvents. Carbohydrate and lignin-derived solvents have been proposed and investigated as green alternatives in many biomass processes. In this review, the applications of different types of biomass pretreatment solvents, including organic, ionic liquid, and deep eutectic solvents, are thoroughly discussed. The role of water as a co-solvent in these pretreatment processes is also reviewed. Finally, current research challenges and prospects of utilizing biomass-derived pretreatment solvents for pretreatment are discussed. Given bioethanol's market potential and increasing public awareness about environmental concerns, it will be a priority adopting sustainable and green biomass pretreatment solvents in biorefinery.

The separation regularity of the three-phase solvent system of counter-current chromatography based on polarity parameter modeling

The three-phase solvent system of counter-current chromatography can separate compounds with a wide range of polarity, but there is no study of its separation regularity. Therefore, in this work, the separation regularity of the three-phase solvent system was initially investigated from the perspective of solvent polarities and compound polarities. The standard compounds covering a wide polarity range were selected, and three-phase solvent systems, n-hexane/methyl acetate/acetonitrile/water, and n-hexane/methyl tert-butyl ether/acetonitrile/water were used for modeling. The results showed that in the three-phase solvent system, the partition coefficient for the middle and lower phases (lgK_M/L) increased with increasing logP values in three intervals logP < 0, 0 < logP < 4, and logP > 4. In addition, the partition coefficient for the upper and middle phases (lgK_U/M) between the upper and middle phases of the small polarity compounds increases with increasing logP values. LogP vs lgK_M/L of 7 solvent systems were employed for the smoothing spline fit through a predictive model design of the curve fitting toolbox in MATLAB software, and good results were achieved. LogP versus lgK_M/L for n-hexane/methyl tert-butyl ether/acetonitrile/water solvent systems were used for the second-order power fit, and satisfactory results were obtained. The relationship between polarity parameters and separation case parameters was explored using a heat map approach. The separation regularity of the three-phase solvent system was preliminarily investigated. This regularity study gives hope of assistance to the chemists studying three-phase solvents and counter-current chromatography.

New Screening Protocol for Effective Green Solvents Selection of Benzamide, Salicylamide and Ethenzamide

New protocol for screening efficient and environmentally friendly solvents was proposed and experimentally verified. The guidance for solvent selection comes from computed solubility via COSMO-RS approach. Furthermore, solute-solvent affinities computed using advanced quantum chemistry level were used as a rationale for observed solvents ranking. The screening protocol pointed out that 4-formylomorpholine (4FM) is an attractive solubilizer compared to commonly used aprotic solvents such as DMSO and DMF. This was tested experimentally by measuring the solubility of the title compounds in aqueous binary mixtures in the temperature range between 298.15 K and 313.15 K. Additional measurements were also performed for aqueous binary mixtures of DMSO and DMF. It has been found that the solubility of studied aromatic amides is very high and quite similar in all three aprotic solvents. For most aqueous binary mixtures, a significant decrease in solubility with a decrease in the organic fraction is observed, indicating that all systems can be regarded as efficient solvent-anti-solvent pairs. In the case of salicylamide dissolved in aqueous-4FM binary mixtures, a strong synergistic effect has been found leading to the highest solubility for 0.6 mole fraction of 4-FM.

LogP determination for highly lipophilic hydrogen-bonding anion receptor molecules

Lipophilicity, usually expressed as octanol-water partition coefficient (logP_o/w), is an important property in biomedical research, drug design and technology. However, high logP_o/w values of complex hydrogen-bonding molecules are not easy to measure or calculate. Exemplary problematic molecules are prospective active components (ionophores) of polymeric sensor membranes - the working elements of ion-selective electrodes. High lipophilicities of the membrane components are crucial for the sensor lifetime. In this work, lipophilicities of a wide range of urea-, carbazole- and indolocarbazole-based anion receptor molecules (some newly synthesized) and two common plasticizers were determined using a chromatography-based approach and/or the COSMO-RS method. Very high logP_o/w values, up to around 20, i.e. far beyond directly experimentally accessible range, were obtained. The agreement between the two approaches ranged from very good to satisfactory. Based on these results, simple fragment-based equations were developed for quick lipophilicity estimation without any specialized software. Membrane-water partition coefficients for the studied compounds were modeled. Limitations and biases of the used methods are discussed.

Ecotoxicity of ammonium chlorophenoxyacetate derivatives towards aquatic organisms: Unexpected enhanced toxicity upon oxygen by sulfur replacement

Phenoxyacetate herbicides, such as 2,4-D and MCPA, having a high toxicity to non-target organisms are commonly used for controlling broadleaf weeds in agriculture. However, novel and environmentally friendly analogs are constantly sought after. For this purpose, various substituents at the phenyl group have been tested to find the optimal balance between the potent herbicidal activity and safety for non-target species. In this work, we investigated the influence of the oxygen by sulfur replacement in the phenoxy moiety of ammonium chlorophenoxyacetates on the toxicity towards aquatic organisms, such as bacteria (Vibrio fischeri), water flea (Daphnia magna) and freshwater fish (Pimephales promelas) by determining experimental (Microtox® test - V. fischeri) and predicted (ACD Lab Percepta software - D. magna, P. promelas) EC₅₀/LC₅₀ values. The achieved results showed that in contrary to the literature observations, where O-compounds were more toxic than their S-analogs (urea/thiourea), the O/S replacement in chlorophenoxyacetate significantly increased ecotoxicity of the S-analogs (up to 11 times). Moreover, one- and two-substituted phenoxyacetates in the form of ammonium salts were less toxic to V. fischeri than the commercially available phenoxy herbicides in the acid form. The logP/logD values were also calculated to understand hydro/lipophilic nature of the investigated compounds and differences in their toxicity.