Average Absolute Relative Deviation Research Articles

Molecular dynamics simulation of extraction of Curcuma longa L. extract using subcritical water.

Humans have utilized plants for various purposes, including sustenance and medical treatment for millennia. Researchers have extensively investigated medicinal plants' potential in drug development, spurred by their rich array of chemical compounds. Curcumin, a valuable bioactive compound, is extracted from Turmeric, known by the scientific name Curcuma Longa L. Notably, curcumin boasts potent antioxidant and anti-inflammatory properties, making it a promising candidate for treating cancer and other microbial diseases. Therefore, the simulation study of the extraction of this important medicinal compound by water, which is a green solvent, was carried out. This study employed molecular dynamics simulation for the first time to explore the extraction of Curcuma Longa L. extract using subcritical water. The simulations were carried out at constant pressure and different temperatures, using the Compass force field in the Lammps simulation package. The findings revealed an increase in the amount of Curcuma longa extract with rising temperature, indicating a weakening of hydrogen bonds in water molecules. Water lost its polar state with increasing temperature and became a suitable non-polar solvent for extracting non-polar compounds. The average absolute relative deviation (AARD) for calculated and simulated density data was 6.45%.

Modeling and estimation of CO2 capture by porous liquids through machine learning

Porous liquids (PLs) are newly developed porous materials that combine unique fluidity with permanent porosity, which exhibit promising functionalities. They have shown ability to efficiently absorb greenhouse gases such as carbon dioxide (CO2). Experimental measurement is one approach to determining the solubility of various greenhouse gases in PLs, which has drawbacks such as being expensive and time-consuming. Hence, simulation models are valuable to predict the solubility of CO2 in various PLs. This work aims to develop machine learning (ML) modeling methods for accurately estimating CO2 solubility under varying conditions (e.g. PLs, temperature, pressure). Adaptive Neuro-Fuzzy Inference System (ANFIS), Particle Swarm Optimization-ANFIS (PSO-ANFIS), Coupled Simulated Annealing-Least Squares Support Vector Machine (CSA-LSSVM), and Multilayer Perceptron Neural Network (MLP-NN) were established as the state of art algorithms for estimating CO2 solubility. The models demonstrated accurate modeling results with average absolute relative deviation (AARD) of 12.98%, 8.67%, 3.17% and 6.64% for ANFIS, PSO-ANFIS, CSA-LSSVM and MLP-NN, respectively. This work has presented a powerful modeling tool with few parameters that need to be controlled, to precisely estimate CO2 solubility in different PLs of complex structures.

Improvement of the properties of ionic liquids for the absorption refrigeration cycle

Ionic liquids (ILs), recognized as environmentally friendly green reagents, have good application prospects in many fields. However, their high viscosity has hindered their widespread application. Currently, an effective solution is to introduce a solvent as an additive. In this work, ethanol with different molar fractions was chosen to form working pairs with three ILs ([emim]BF4, [bmim]BF4 and [hmim]BF4), to reduce their viscosities. The densities and viscosities of the three working pairs were obtained in a temperature range of 293.15∼338.15 K, thereby determining the thermal expansion coefficients. The effects of ethanol on the volumetric and viscometric characteristics of the three ILs are discussed. To further analyze the interactions between ethanol and the ILs, excess mole volume and viscosity deviations were introduced. The interactions between different molecules were stronger than those between the same molecules in the working pairs. This is the main reason why the addition of ethanol reduces the viscosity of the three ILs. Finally, the hard-sphere model was applied to predict the viscosities of the working pairs. Upon introducing an adjustable argument, the average absolute relative deviation of the prediction for the aforementioned three ILs decreased to 4.4%, 5.9%, and 5.0%, respectively.

Predicting the solubility of CO2 and N2 in ionic liquids based on COSMO-RS and machine learning.

As ionic liquids (ILs) continue to be prepared, there is a growing need to develop theoretical methods for predicting the properties of ILs, such as gas solubility. In this work, different strategies were employed to obtain the solubility of CO2 and N2, where a conductor-like screening model for real solvents (COSMO-RS) was used as the basis. First, experimental data on the solubility of CO2 and N2 in ILs were collected. Then, the solubility of CO2 and N2 in ILs was predicted using COSMO-RS based on the structures of cations, anions, and gases. To further improve the performance of COSMO-RS, two options were used, i.e., the polynomial expression to correct the COSMO-RS results and the combination of COSMO-RS and machine learning algorithms (eXtreme Gradient Boosting, XGBoost) to develop a hybrid model. The results show that the COSMO-RS with correction can significantly improve the prediction of CO2 solubility, and the corresponding average absolute relative deviation (AARD) is decreased from 43.4% to 11.9%. In contrast, such an option cannot improve that of the N2 dataset. Instead, the results obtained from coupling machine learning algorithms with the COSMO-RS model agree well with the experimental results, with an AARD of 0.94% for the solubility of CO2 and an average absolute deviation (AAD) of 0.15% for the solubility of N2.

Thermodynamic Modeling of an Aqueous N,N-Dimethyldipropylenetriamine and Benzylamine Blend for Efficient CO2 Capture.

This study evaluates the carbon dioxide (CO2) capture capabilities of a novel aqueous blend of N,N-dimethyldipropylenetriamine (DMDPTA) and benzylamine (BA). The solvent properties such density, vapor- liquid equilibrium (VLE) of CO2 in the solvent, CO2 absorption enthalpy are evaluated experimentally for solvent composition of (5 mass% DMDPTA + 25 mass% BA), (10 mass% DMDPTA + 20 mass% BA), and (15 mass% DMDPTA + 15 mass% BA). Solvent density were measured in the temperature range of 303.15K-333.15K and correlated using Redlich-Kister excess molar volume model, with a low average absolute relative deviation (AARD) of 0.014. VLE data was measured using a custom-made stirred VLE cell, within CO2 partial pressure range of 2-200 kPa and at temperatures 313.15K, 323.15K and 333.15K. Equilibrium CO2 solubility data were correlated using a modified Kent-Eisenberg model, achieving an AARD of 1.5%. Enthalpy of CO2 absorption was measured at 313.15 K using a Meter Toledo reaction calorimeter. Results indicated that under similar process conditions and solvent composition, (DMDPTA+BA) blends exhibited significantly higher CO2 loading and low absorption enthalpy compared to aqueous BA and monoethanolamine (MEA) solvent alone indicating the potential of (DMDPTA+BA) blend as efficient CO2 capture solvent.

Predicting interfacial tension in brine-hydrogen/cushion gas systems under subsurface conditions: Implications for hydrogen geo-storage

Underground hydrogen storage (UHS) critically relies on cushion gas to maintain pressure balance during injection and withdrawal cycles, prevent excessive water inflow, and expand storage capacity. Interfacial tension (IFT) between brine and hydrogen/cushion gas mixtures is a key factor affecting fluid dynamics in porous media. This study develops four machine learning models— Decision Trees (DT), Random Forests (RF), Support Vector Machines (SVM), and Multi-Layer Perceptrons (MLP)—to predict IFT under geo-storage conditions. These models incorporate variables such as pressure, temperature, molality, overall gas density, and gas composition to evaluate the impact of different cushion gases. A group-based data splitting method enhances the realism of our tests by preventing information leakage between training and testing datasets. Shapley Additive Explanations (SHAP) reveal that while the MLP model prioritizes gas composition, the RF model focuses more on operational parameters like pressure and temperature, showing distinct predictive dynamics. The MLP model excels, achieving coefficients of determination (R2) of 0.96, root mean square error (RMSE) of 2.10 mN/m, and average absolute relative deviation (AARD) of 3.25%. This robustness positions the MLP model as a reliable tool for predicting IFT values between brine and hydrogen/cushion gas (es) mixtures beyond the confines of the studied dataset. The findings of this study present a promising approach to optimizing hydrogen geo-storage through accurate predictions of IFTs, offering significant implications for the advancement of energy storage technologies.

Predictive modeling of membrane reactor efficiency using advanced artificial neural networks for green hydrogen production

The imperative to decarbonize the energy sector has prompted substantial advancements in clean electricity generation, with hydrogen emerging as a promising low-carbon energy carrier. While hydrogen synthesis from renewable sources is crucial, challenges persist, necessitating innovative approaches for efficient and sustainable production. This study leverages diverse artificial neural network (ANN) models to assess and predict system efficiency based on key operational variables in membrane reactor systems. The multilayered perceptron (MLP) and radial basis function (RBF) methodologies are employed, with the MLP models optimized across twelve training algorithms and eight activation functions, exploring up to three hidden layers with variable neuron counts. The MLP model, utilizing the Levenberg-Marquard training algorithm and Tangent-Sigmoid activation function, achieved a high correlation coefficient (R2) of 0.9975 for training and 0.9962 for testing, and a mean squared error (MSE) of 0.00425 for training and 0.23951 for testing, indicating precise and accurate efficiency predictions. The Log-Sigmoid activation function also performed well, with R² values of 0.9971 (training) and 0.9961 (testing), and MSE values of 0.004086 (training) and 0.17694 (testing). Optimization of the RBF network identified the best performance with a spread parameter of 1 and 35 neurons, although the MLP model demonstrated superior accuracy and reduced computational time. Statistical analysis, encompassing correlation coefficient, mean squared error, Root Mean Squared error, absolute average deviation, absolute average relative deviation, and runtime, confirms the network’s consistent and accurate estimation of system efficiency across various input variables. The study highlights that applying tansig and logsig activation functions, configured with neuron counts of 20, 17, 6 and 23, 20, 2 at the first, second and third hidden layers, respectively, offers enhanced accuracy and reliability. The MLP model’s high performance underscores its potential to identify optimal conditions for H2 generation based on system efficiency, thereby advancing membrane reactor technology for hydrogen production.

CO2 gas-liquid equilibrium study and machine learning analysis in MEA-DMEA blended amine solutions

The combination of primary and tertiary amines represents a promising approach to improve sorbent performance in CO2 capture by enhancing absorption efficiency and reducing regeneration energy. This study focuses on investigating the absorption performance of binary mixtures of ethanolamine (MEA) and N,N-dimethylethanolamine (DMEA) at temperatures between 298–323 K and CO2 partial pressures between 5–60 kPa. The species generated during the absorption were analyzed using 13C NMR spectroscopy, to clarify the intricate role of MEA and DMEA in the capture process. A developed excess property model for MEA-DMEA, based on excess CO2 loading, predicted equilibrium CO2 solubility data with an average absolute relative deviation (AARD) of 1.6 %. Additionally, the machine learning models XGBoost, RBFNN, and SVR were applied, providing AARD values between 0.86 % and 1.28 %, demonstrating strong agreement between experimental and predicted outcomes. These comprehensive findings enhance our understanding of mixed amines’ mechanisms and practical applications, contributing to ongoing research development.

Estimation of gross calorific value of coal based on the cubist regression model.

The gross calorific value (GCV) of coal is an important parameter for evaluating coal quality, and regression analysis methods can be used to predict GCV. In this study, we proposed a GCV prediction model based on cubist regression. To develop a good regression model, feature selection of input variables was performed using a correlation analysis and a recursive feature elimination algorithm. Thus, in this study, we determined three sets of variables as the optimal combination for regression models: proximate analysis variables (Set 1: moisture, standard ash, and volatile matter), element analysis variables (Set 2: carbon, sulfur, and oxygen), and comprehensive index variables (Set 3: carbon, volatile matter, standard ash, sulfur, moisture, and hydrogen). Results for comparison with multiple linear regression, random forest regression, and numerous previous prediction models, such as gradient boosting regression tree, support vector regression (SVR), backpropagation neural networks, and particle swarm optimization-artificial neural network (PSO-ANN), indicate that these seven regression models have the best fitting effect on the comprehensive index variables among the three sets of input variables. The cubist model showed higher prediction accuracy and lower error than most other models (R2, mean absolute error, root mean square error, and average absolute relative deviation percentage values are 0.990, 0.476, 0.668, and 0.086% for the proximate analysis variables; 0.992, 0.381, 0.596, and 0.140% for element analysis variables; and 0.999, 0.161, 0.219, and 0.087% for comprehensive index variables, respectively). The cubist model combines the advantages of decision tree and linear regression, which not only enables it to perform well in terms of accuracy but also makes the model highly interpretable because it is based on multiple sublinear equations. In addition, the cubist model shows obvious advantages in terms of running speed, especially compared with SVR and PSO-ANN, which require complex parameter optimization. In summary, the cubist model considers the prediction accuracy, model interpretability, and computational efficiency as well as provides a new and effective method for GCV prediction.

Thermal conductivity measurements of liquid ethane and ethane + propane binary mixtures

Based on the transient hot-wire apparatus, the thermal conductivity of ethane was measured in the temperature range of 146–247 K and pressure < 14 MPa in the liquid phase. The thermal conductivity of ethane + propane binary mixture was measured in the temperature range of 156–288 K and pressure < 25 MPa in the liquid phase. The mole fractions of ethane in binary systems are 0.198, 0.525, and 0.736. The deviation between experimental results and REFPROP calculation falls within 3 % of ethane and mixture. Subsequently, the previous established thermal conductivity model, derived from dimensional analysis, was employed to predict the thermal conductivity of n-alkanes mixture. The mixture critical parameters were obtained first by the mixing rules, and the thermal conductivity of the mixture were calculated. Comparisons were made between the experimental data of ethane and the previously constructed n-alkane model, resulting in an average absolute relative deviation (AARD) of 4.42 %. Similarly, the experimental data of the ethane + propane mixture was compared with the mixture prediction model proposed in this study, yielding an AARD of 1.92 %. The relative deviation between experimental data and model calculations demonstrates good consistency.

Numerical investigation of turbulent mass transfer processes in turbulent fluidized bed by computational mass transfer

Turbulent fluidized bed possesses a distinct advantage over bubbling fluidized bed in high solids contact efficiency and thus exerts great potential in applications to many industrial processes. Simulation for fluidization of fluid catalytic cracking (FCC) particles and the catalytic reaction of ozone decomposition in turbulent fluidized bed is conducted using the Eulerian–Eulerian approach, where the recently developed two-equation turbulent (TET) model is introduced to describe the turbulent mass diffusion. The energy minimization multi-scale (EMMS) drag model and the kinetic theory of granular flow (KTGF) are adopted to describe gas–particles interaction and particle–particle interaction respectively. The TET model features the rigorous closure for the turbulent mass transfer equations and thus enables more reliable simulation. With this model, distributions of ozone concentration and gas–particles two-phase velocity as well as volume fraction are obtained and compared against experimental data. The average absolute relative deviation for the simulated ozone concentration is 9.67% which confirms the validity of the proposed model. Moreover, it is found that the transition velocity from bubbling fluidization to turbulent fluidization for FCC particles is about 0.5 m·s–1 which is consistent with experimental observation.

A quantum-corrected Peng‒Robinson equation of state for helium-4 from 3 K to 50 K considering quantum swelling effects through the Feynman‒Hibbs correction of the EXP-6 potential

Accurate and efficient prediction of the properties of helium-4 (He-4) via an equation of state (EOS) is a prerequisite for evaluating liquid helium-4 (LHe-4) storage technology. However, the performance of the Peng–Robinson (PR) EOS in predicting the density of LHe-4 and vapour helium-4 (VHe-4) deteriorates within the thermodynamic ranges of the LHe-4 tank: 3–50 K and 60–600 kPa. To modify the PR EOS, we establish first-order and second-order Feynman–Hibbs (FH)-corrected EXP-6 potentials and propose a reduced effective particle diameter (REPD) correlation with four parameters to consider the quantum swelling effects of LHe-4. On the basis of the rational function form of the REPD correlation, we introduce a quantum-corrected covolume term to develop a regression model for the FH-corrected Peng–Robinson (FH-PR) EOS. Moreover, to improve the effectiveness of regression near the saturation curve, we propose a hypothetical boundary consisting of the saturation curve from 3 K to the critical temperature and a virtual saturation curve from the critical pressure to 600 kPa. The results indicate that the FH − PR EOS shows satisfactory engineering application performance in predicting the density of He-4 within the studied range. Under verification conditions, the average absolute relative deviation (AARD) of the density determined via the FH − PR EOS ranges from 0.72 % to 1.77 %, and the maximum relative deviation (MRD) ranges from 2.17 % to 5.62 %. Under test conditions, the AARD of the density ranged from 1.06 % to 1.71 %, and the MRD ranged from 3.77 % to 7.38 %.

A theoretical approach based on machine learning for estimation of physical properties of LLDPE in moulding process

This study explores the prediction of mechanical characteristics of linear polyethylene based on oven residence time, employing various regression models and hyper-parameter tuning through the Whale Optimization Algorithm. The dataset comprises one input variable (oven residence time) and three output parameters (Tensile Strength, Impact Strength, and Flexure Strength). The models investigated include Multilayer Perceptron, K-Nearest Neighbors, Support Vector Regression, Polynomial Regression, and Theil–Sen Regression. The results showcased distinct performances across the models for each output parameter. The Polynomial Regression (WOA-PR) method has been identified as the most suitable option for predicting Tensile Strength due to its ability to achieve the lowest errors in terms of Mean Absolute Error, Root Mean Square Error, and Average Absolute Relative Deviation. K-Nearest Neighbors (WOA-KNN) outperforms other models in predicting Impact Strength due to its superior accuracy and reliability. Additionally, Support Vector Regression (WOA-SVR) emerges as the best model for predicting Flexure Strength, showcasing notable performance in minimizing prediction errors. These findings underscore the significance of model selection and optimization techniques in accurately predicting the mechanical properties of polymers, paving the way for enhanced manufacturing processes and material design.

Enhanced Solubility and Miscibility of CO2-Oil Mixture in the Presence of Propane under Reservoir Conditions to Improve Recovery Efficiency

The existence of propane (C3H8) in a CO2-oil mixture has great potential for increasing CO2 solubility and decreasing minimum miscibility pressure (MMP). In this study, the enhanced solubility, reduced viscosity, and lowered MMP of CO2-saturated crude oil in the presence of various amounts of C3H8 have been systematically examined at the reservoir conditions. Experimentally, a piston-equipped pressure/volume/temperature (PVT) cell is first validated by accurately reproducing the bubble-point pressures of the pure component of C3H8 at temperatures of 30, 40, and 50 °C with both continuous and stepwise depressurization methods. The validated cell is well utilized to measure the saturation pressures of the CO2-C3H8-oil systems by identifying the turning point on a P-V diagram at a given temperature. Accordingly, the gas solubilities of a CO2, C3H8, and CO2-C3H8 mixture in crude oil at pressures up to 1600 psi and a temperature range of 25–50 °C are measured. In addition, the viscosity of gas-saturated crude oil in a single liquid phase is measured using an in-line viscometer, where the pressure is maintained to be higher than its saturation pressure. Theoretically, a modified Peng–Robinson equation of state (PR EOS) is utilized as the primary thermodynamic model in this work. The crude oil is characterized as both a single and multiple pseudo-component(s). An exponential distribution function, together with a logarithm-type lumping method, is applied to characterize the crude oil. Two linear binary interaction parameters (BIP) correlations have been developed for CO2-oil binaries and C3H8-oil binaries to accurately reproduce the measured saturation pressures. Moreover, the MMPs of the CO2-oil mixture in the presence and absence of C3H8 have been determined with the assistance of the tie-line method. It has been found that the developed mathematical model can accurately calculate the saturation pressures of C3H8 and/or CO2-oil systems with an absolute average relative deviation (AARD) of 2.39% for 12 feed experiments. Compared to CO2, it is demonstrated that C3H8 is more soluble in the crude oil at the given pressure and temperature. The viscosity of gas-saturated crude oil can decrease from 9.50 cP to 1.89 cP and the averaged MMP from 1490 psi to 1160 psi at 50 °C with the addition of an average 16.02 mol% C3H8 in the CO2-oil mixture.

Establishing rheological models of lignin-based solutions via molecular parameters using machine learning

The rheological models of natural polymer-based solutions are difficult to be established because of the significant non-Newtonian behavior and highly discrete rheological data caused by different molecular parameters including the molecular weights and size of clusters. In this study, a typical natural polymer-lignin was selected and dissolved in polyethylene glycol (PEG) as the lignin-based solutions. The experimental rheological data of different PEG-lignin solutions were trained with machine learning. The rheological models were established considering the molecular parameters including the molecular weights and size of clusters. The models show a high accuracy in predicting the viscosities of different PEG-lignin solutions with the coefficient of determination over 0.9815, mean absolute error less than 0.0132, and average absolute relative deviation less than 6.95 % in both Newtonian and non-Newtonian regimes. The models and relevant methodology can provide scenarios for further application of natural polymer solutions in process industries.

New intelligent models for predicting wax appearance temperature using experimental data – Flow assurance implications

Wax deposition is a major problem during petroleum production causing flow rate reduction and blockage of tubing, pipelines, and surface facilities. Wax deposition occurs when crude oil temperature falls below wax appearance temperature (WAT). WAT is an important parameter for wax deposition modeling and developing, testing, and selecting appropriate wax inhibitors. WAT can be determined in laboratory using several different techniques. Although, each method has its own disadvantages including being time-consuming, expensive, inaccurate, or posing health and safety risks. Experimental WAT data are very scarce in the literature. Reliable and accurate correlations and models for WAT prediction are rare, as well. Hence, development of fast, easy, and reliable models for prediction of WAT is inevitable. The main objective of this research work was to develop and introduce novel intelligent models for precise production of WAT. The artificial intelligent (AI) algorithms used include least-squares support-vector machines (LSSVM), recurrent neural network (RNN), and Adaptive neuro-fuzzy inference system (ANFIS). A high-quality dataset was assembled using experimental WAT data collected from the literature. The dataset consists of 81 experimental which is the largest dataset ever reported. The gathered dataset covers a wide range of density from 0.71 to 0.89 g/cm3, wax content from 0.61 to 25.78 wt%, and Pour point from −36 to 37°C. The selected input parameters include oil density, wax content, and pour point determined using a rigorous feature selection analysis. Statistical analysis showed that pour point has the highest impact on WAT prediction. The modeling results showed that the LSSVM model is the superior model with coefficient of determination (R2) of 0.9893, a root mean squared error (RMSE) of 0.1655, and an average absolute relative deviation (AARD) of 9%. The LSSVM model outperformed the only existing smart model in terms of both performance and accuracy. The proposed smart model can be used for prediction of WAT which is an essential parameter in all wax related research.

Solubility measurement of Ceftriaxone sodium in SC-CO2 and thermodynamic modeling using PR-KM EoS and vdW mixing rules with semi-empirical models

The complete investigation of the solubility of Ceftriaxone sodium drug has not yet been conducted. This communication investigates the solubility of Ceftriaxone sodium drug in supercritical CO2 (SC-CO2) for the first time. The drug solubility measurements were performed using utilizing UV–vis analysis under different operation conditions (at P (bar) = 120–270 and T (K) = 308–338). The results demonstrated that the solubility of Ceftriaxone sodium ranged from 0.90 × 10−6 to 8.01 × 10−6. The mole fraction solubility of Ceftriaxone sodium increases proportionally with an increase in the pressure, while maintaining a constant temperature. However, a crossover was noted. The solubility behavior of Ceftriaxone sodium drug in SC-CO2 was modeled by Peng-Robinson equation of state (PR EoS) with Kwak-Mansoori and van der Waals mixing rules and seven semi-empirical correlations (Bian et al., Kumar-Johnston, Chrastil, Garlapati-Madras, Bartle et al., Sung-Shim and Sodeifian et al.,). The vaporization, total and solvation enthalpies of the Ceftriaxone sodium/SC-carbon dioxide system were obtained by KJ, Chrastil, and Bartle et al. correlations. The evaluation of two methods was performed by absolute average relative deviation (AARD%). Both models exhibited a satisfactory level of agreement with the experimental data. Finally, PR/KM EoS and the Chrastil models, with AARD% values of 8.70 % and 8.10 % respectively, showed superior precision in accurately fitting the solubility data.

Thermal optimization of ivermectin in supercritical carbon dioxide and ethanol to produce nano medicine

The solubility of ivermectin in high pressure carbon dioxide (CO2) at 338, 328, 318, and 308 K temperatures was investigated with and without an ethanol. Several methods were employed to model the data obtained from the experiments. ivermectin exhibited solubility values ranging from 0.130 × 10−5 to 1.760 × 10−4 in the binary system and 1.880 × 10−5 to 1.237 × 10−4 in the ternary system. The outcome indicated that the addition of ethanol significantly increased the mole fraction of ivermectin in CO2. The highest impact of ivermectin solubility was found in the ivermectin -Ethanol- CO2 system at 338 K and 12 MPa, which was about 14.35 times greater than binary system under the same conditions. The models proposed by Jouyban et al. and Sodeifian-Sajadian models showed the best correlation with average absolute relative deviation (AARD%) and Akaike information criterion (AICc) for binary and ternary approaches, respectively. The results confirmed that rapid expansion supercritical solution (RESS) and gas antisolvent (GAS) method can be apply for producing nanoparticles at the suitable ranges of the drug solubility.

Viscosity of deep eutectic solvents: Predictive modeling with experimental validation

Viscosity, the measure of a fluid's resistance to deformation, is a critical parameter in many industries. Being able to accurately predict viscosity is essential for the successful design and optimization of technological processes. In this research, regression models were created to predict the viscosity of deep eutectic solvents (DESs). Machine learning models were trained using a data set of 3440 data points for two component DESs. Different algorithms, such as Multiple Linear Regression, Random Forest, CatBoost, and Transformer CNF, were employed alongside a variety of structural representations like fingerprints, σ-profiles, and molecular descriptors. The effectiveness of the models was assessed for interpolation tasks within the training data and extrapolation outside of it. The results indicate that a rigorous splitting of the dataset into subsets is necessary to accurately evaluate the performance of the models. Two new choline chloride-based DESs were prepared and their viscosities were measured to evaluate the predictive capabilities of the models. The CatBoost algorithm with CDK molecular descriptors was chosen as the recommended model. The average absolute relative deviations (AARD) of this model exhibited fluctuations during 5-fold cross-validation, ranging from 10.8 % when interpolating within the dataset to 88 % when extrapolating to new mixture components. The open access model was presented in this study (http://chem-predictor.isc-ras.ru/ionic/des/).

Zinc (II)-1,4,7,10-tetraazacyclododecane complexes catalyzed tertiary amine-based CO2 capture in a rotating packed bed

A novel tertiary amine-based CO2 absorption process using zinc (II)-1,4,7,10-tetraazacyclododecane complexes (CN) as a catalyst in a rotating packed bed was proposed for energy-efficient CO2 capture. The CN-catalysis reaction mechanism was first studied by density functional theory calculation, and the Gibbs Free Energy profile was established. Furthermore, the effects of various operation conditions for the tertiary amine-CN-CO2 systems on the CO2 removal efficiency (η), the overall gas-phase volumetric mass transfer coefficient (KGa), and the efficiency enhancement factor (Φ) were investigated. It was found that η and KGa with adding 4 wt% CN were 1–3 folds than those uncatalyzed. The absorption performance of CO2 in DEEA solution with CN was superior to that in 10 wt% DEA solution. Finally, an artificial neural network model was established to predict KGa, and the predicted data agreed well with experimental results with the average absolute relative deviation of 5.07 %.