Application of machine learning approach to estimate the solubility of some solid drugs in supercritical CO2

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Фев. 12, 2025

Accurate estimation of the solubility solid drugs (SDs) in supercritical carbon dioxide (SC-CO2) plays an essential role related technologies. In this study, artificial intelligence models (AIMs) by gene expression programming (GEP) and adaptive neuro-fuzzy inference system (ANFIS) methods were applied to estimate SDs SC-CO2. Hence, a comprehensive database (1816 datasets) comprising operational conditions (T, P) wide ranges (308–348.2 K 80–400 bar), SD's molecular weight (MWSDs), melting point (MPSDs) gathered. Investigation analysis models' strength showed that model developed ANFIS exhibited more satisfactory approximation than GEP model. According optimized model, statistical parameters R2, RMSE, MAE, AARD% obtained, equivalent 0.991, 0.260, 0.167, 13.890% for training 0.990, 0.256, 0.157, 15.273% validation, order. Sensitivity highest effect independent variables on calculating SC-CO2 belong MWSDs, P, MPSDs, T, respectively. Therefore, MWSDs is key factor modeling various Comparing estimated results obtained from AIM with previous semi-empirical AIMs could be accurate

Язык: Английский

---

Improvement of the dissolution of the antineoplastic drug regorafenib through impregnation into pullulan polysaccharide using supercritical fluid technology: Optimization of the process

Journal of CO2 Utilization, Год журнала: 2025, Номер 93, С. 103040 - 103040

Опубликована: Фев. 9, 2025

Язык: Английский

---

Determination of Gefitinib hydrochloride anti-cancer drug solubility in supercritical CO2: Evaluation of sPC-SAFT EoS and semi-empirical models

Journal of the Taiwan Institute of Chemical Engineers, Год журнала: 2024, Номер 161, С. 105569 - 105569

Опубликована: Май 27, 2024

Язык: Английский

---

Micro and nanosizing of Tamsulosin drug via supercritical CO2 antisolvent (SAS) process

Journal of CO2 Utilization, Год журнала: 2024, Номер 84, С. 102847 - 102847

Опубликована: Июнь 1, 2024

Supercritical anti-solvent (SAS) method is a solvent-free approach where supercritical carbon dioxide (SC-CO2) used as and one can control the output particle size without disturbing nature of particles. The primary objective this study to develop SAS for production Tamsulosin (TML) nanoparticles investigate parameters affecting resultant size. In order optimize nano-sized particles, response surface methodology (RSM) based on Box-Behnken design (BBD (was used. Using dynamic light scattering (DLS) scanning electron microscopy (SEM) analyses, morphologies mean diameter ranges were examined. To look into how process affects TML's chemical physical characteristics, Fourier-transform infrared spectroscopy (FT–IR), X-ray diffraction analysis (XRD) differential calorimetry (DSC) further performed. acquired results demonstrated that pressure imposes largest impact TML production, with exhibiting an inverse correlation pressure. Observations indicated increase in increasing temperature from 35 50 °C. By flow rate drug solution, average exhibited initial decrease followed by trend. Results optimization showed at pressure, temperature, injection 25 MPa, 41.3 °C, 2.9 mL/min, respectively, particles shrink 600 nm.

Язык: Английский

---

Determination of 5-fluorouracil anticancer drug solubility in supercritical CO2using semi-empirical and machine learning models

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Фев. 7, 2025

In order to provide the facilities design supercritical fluid (SCF) processes for micro or nanosizing of solid solute compounds such as drugs, it is essential obtain their solubility in green solvents like pressurized CO2. This important role first stage assessing each SCF technology. A statistical method was developed time and employed determine 5-fluorouracil (5-Fu) solubility. The measurements were performed at different pressures (120-270 bar) temperatures (308-338 K) through UV-vis spectrophotometry, time. obtained between 0.0024 0.0176 g/L. 5-Fu mole fraction constant temperature, increases with an increase pressure. Whereas, a crossover point has been seen. Three models approaches applied correlate model experimental data set: (i) seven density-based models, (ii) PR equations state (vdW2 mixing rule), (iii) machine learning-based namely non-linear regressions, Random Forest, Gradient Boosting, Decision Tree, Kernel Ridge. All tested successfully within acceptable accuracy. Meanwhile, empirical suggested by Sodeifian 2, superior lowest AARD% (AARD = 4.12%). Finally, total, solvation, vaporization enthalpies drug/Sc-CO2 binary system determined using semi-empirical correlations,

Язык: Английский

---

Study of Baclofen Solubility in Supercritical CO₂ with and without Cosolvents: Experimental Analysis, Thermodynamic Evaluation, and Machine Learning Methods

Journal of Chemical & Engineering Data, Год журнала: 2025, Номер unknown

Опубликована: Янв. 15, 2025

This study investigates the solubility of Baclofen in supercritical CO2, which is essential for developing an efficient drug delivery system using processes. Solubility measurements were carried out both with and without presence various cosolvents (ethanol dimethyl sulfoxide (DMSO)), across a temperature range 308–338 K pressure 12–30 MPa. exhibited ranging from 1.62 × 10–5 to 2.30 mole fractions pure CO2. In cosolvents, increased 5.76 12.79 ethanol 3.50 7.02 DMSO. Indeed, addition DMSO by approximately 3.55–5.56 times 2.15–3.05 times, respectively. Several density-based empirical models thermodynamic (Soave–Redlich–Kwong (SRK) Peng–Robinson (PR) equations state) used correlate data. Jafari Nejad's model CO2–Baclofen Jouyban's CO2–ethanol/DMSO–Baclofen systems displayed higher consistency. Also, PR showed better accuracy correlating while SRK outperformed CO2–ethanol/DMSO cosolvents. Moreover, machine learning exceptional accuracy, over 99% predictions closely matching experimental data, emphasizing its outstanding performance.

Язык: Английский

---

Machine learning analysis of rivaroxaban solubility in mixed solvents for application in pharmaceutical crystallization

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Янв. 17, 2025

This study investigates the use of machine learning models to predict solubility rivaroxaban in binary solvents based on temperature (T), mass fraction (w), and solvent type. Using a dataset with over 250 data points including encoded one-hot encoding, four were compared: Gradient Boosting (GB), Light (LGB), Extra Trees (ET), Random Forest (RF). The Jellyfish Optimizer (JO) algorithm was applied tune hyperparameters, enhancing model performance. LGB achieved best results, an R2 0.988 test set low error rates (RMSE 9.1284E-05 MAE 5.85322E-05), surpassing other predictive accuracy generalizability. Parity plots confirmed model's close alignment between predicted actual values, highlighting its robust Furthermore, 3D surface partial effect demonstrated LGB's capacity across different systems, capturing complex interactions T, w, effects. Finally, maximum at 305.76 K 0.753 dichloromethane + methanol mixture, providing valuable insights for optimization selection. work underscores effectiveness prediction, potential applications formulation experimental planning.

Язык: Английский

---

Solubility of gemcitabine (an anticancer drug) in supercritical carbon dioxide green solvent: Experimental measurement and thermodynamic modeling

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Фев. 6, 2025

In order to develop supercritical fluid (SCF) processes for the micro or nanosizing of solid solute components, such as pharmaceuticals, it is essential assess their solubility in solvents including carbon dioxide (SC–CO2). This crucial step fundamental choosing and evaluating SCF technology. A statistical method was developed used determine gemcitabine SC-CO2. The measurements at various pressures temperatures were conducted using UV–vis analysis. Two model types correlate data: semi-empirical (with 3–6 parameters) equation state (EoS) models. SC-CO2 a static 308–338 K 120–270 bar measured modeled first time. Solubility ranged from 0.1274 × 10− 5 1.8128 mole fraction 0.00295 0.08489 kg/L. PR EoS performed best 308 with an AARD 12.58%, SRK excelled 318–338 AARDs between 12.93 15.68%. Application Joback (AARD 12.04–27.13%) COSMO-RS 18.68–31.28%) models also compared. Bian et al. showed fit among density-based 16.62%.

Язык: Английский

---

Application of theoretical and semi-empirical equations for modeling of Maprotiline hydrochloride (anti-depressant) drug solubility in SC-CO2

Measurement, Год журнала: 2025, Номер unknown, С. 117508 - 117508

Опубликована: Апрель 1, 2025

Язык: Английский

---

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

Case Studies in Thermal Engineering, Год журнала: 2024, Номер 61, С. 105074 - 105074

Опубликована: Сен. 1, 2024

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

Язык: Английский

---