An open and extensible sigma‐profile database for COSMO‐based models

Ferrarini, Fabrício; Flôres, Guilherme Braganholo; Muniz, André R.; Soares, Rafael de Pelegrini

doi:10.1002/aic.16194

Cited by 61 publications

(38 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Ferrarini et al . 18 distributed a sigma-profile database for a wide range of molecules using the GAMESS software. They also tested different quantum chemistry theories for the calculation of the electronic structure.…”

Section: Introductionmentioning

confidence: 99%

Investigation of COSMO-SAC model for solubility and cocrystal formation of pharmaceutical compounds

Mahmoudabadi

Pazuki

2020

Sci Rep

View full text Add to dashboard Cite

In this study, a predictive model named COSMO-SAC was investigated in solid/liquid equilibria for pharmaceutical compounds. The examined properties were the solubility of drug in the pure and mixed solvents, octanol/water partition coefficient, and cocrystal formation. The results of the original COSMO-SAC model (COSMO-SAC (2002)) was compared with a semi-predictive model named Flory–Huggins model and a revised version of the COSMO-SAC (COSMO-SAC (2010)). The results indicated the acceptable accuracy of the COSMO-SAC (2002) in the considered scope. The results emphasized on the suitability of the COSMO-SAC model for simple molecules containing C, H, and O by covalent and hydrogen bonding interactions. Applicability of the COSMO-SAC for more complicated molecules made of various functional groups such as COO and COOH doubly requires more modification in the COSMO-SAC.

show abstract

Section: Introductionmentioning

confidence: 99%

Investigation of COSMO-SAC model for solubility and cocrystal formation of pharmaceutical compounds

Mahmoudabadi

Pazuki

2020

Sci Rep

View full text Add to dashboard Cite

show abstract

“…COSMO cavity surface areas and apparent surface charge densities were not optimized. COSMO cavity volumes calculated as described by Ferrarini et al [28] were used as initial guesses in the estimation problem.…”

Section: Cavity Volumes and Areasmentioning

confidence: 99%

“…The area of one hole was assumed to be a universal constant, corresponding to the area of 1 Å radius sphere (12.57 Å 2 ). A much smaller hole surface area would be inconsistent with the underlying COSMO-SAC contact surface area [2,28,29]. The hole cavity volume b h (representing free volume) was optimized individually for each substance.…”

Section: Cavity Volumes and Areasmentioning

confidence: 99%

“…where s m is the apparent surface charge density of segment m; a 0 is a constant; E HB m;n is the hydrogen bond (HB) formation energy; and E Disp m;n is the dispersion contribution. The required COSMO apparent surface charge densities were taken from the open source LVPP-sigma profile database [28], freely available at https://github.com/lvpp/sigma. The COSMO-SAC universal constants and HB energies were taken from the GMHB1808 parametrization, also freely available on that github repository.…”

Section: Segment Interaction Energiesmentioning

confidence: 99%

See 1 more Smart Citation

A pairwise surface contact equation of state: COSMO-SAC-Phi

Soares

Baladão

Staudt

2019

Fluid Phase Equilibria

Self Cite

View full text Add to dashboard Cite

In this work a new method for inclusion of pressure effects in COSMO-type activity coefficient models is proposed. The extension consists in the direct combination of COSMO-SAC and lattice-fluid ideas by the inclusion of free volume in form of holes. The effort when computing pressure (given temperature, volume, and mole numbers) with the proposed model is similar to the cost for computing activity coefficients with any COSMO-type implementation. For given pressure, computational cost increases since an iterative method is needed. This concept was tested for representative substances and mixtures, ranging from light gases to molecules with up to 10 carbons. The proposed model was able to correlate experimental data of saturation pressure and saturated liquid volume of pure substances with deviations of 1.16% and 1.59%, respectively. In mixture vapor-liquid equilibria predictions, the resulting model was superior to Soave-Redlich-Kwong with Mathias-Copeman a-function and the classic van der Waals mixing rule in almost all cases tested and similar to PSRK method, from low pressures to over 100 bar. Good predictions of liquid-liquid equilibrium were also observed, performing similarly to UNIFAC-LLE, with improved responses at high temperatures and pressures.

show abstract

“…Database search is often applied for solvent screening when using COSMO‐based models, 10 where the time‐consuming DFT (Density Functional Theory) calculation (shown in Figure A1a in the supplementary information) is performed once, and the results are saved for future reuse. However, this screening method is limited to only a set of known solvents existing in the database, 11 while, with the help of CAMD techniques, promising solvents can be designed in a much larger search space with respect to the target properties through efficient mathematical optimization algorithms.…”

Section: Introductionmentioning

confidence: 99%

Machine learning‐based atom contribution method for the prediction of surface charge density profiles and solvent design

et al. 2020

View full text Add to dashboard Cite

Solvents are widely used in chemical processes. The use of efficient model‐based solvent selection techniques is an option worth considering for rapid identification of candidates with better economic, environment and human health properties. In this paper, an optimization‐based MLAC‐CAMD framework is established for solvent design, where a novel machine learning‐based atom contribution method is developed to predict molecular surface charge density profiles (σ‐profiles). In this method, weighted atom‐centered symmetry functions are associated with atomic σ‐profiles using a high‐dimensional neural network model, successfully leading to a higher prediction accuracy in molecular σ‐profiles and better isomer identifications compared with group contribution methods. The new method is integrated with the computer‐aided molecular design technique by formulating and solving a mixed‐integer nonlinear programming model, where model complexities are managed with a decomposition‐based strategy. Finally, two case studies involving crystallization and reaction are presented to highlight the wide applicability and effectiveness of the MLAC‐CAMD framework.

show abstract

An open and extensible sigma‐profile database for COSMO‐based models

Cited by 61 publications

References 42 publications

Investigation of COSMO-SAC model for solubility and cocrystal formation of pharmaceutical compounds

Investigation of COSMO-SAC model for solubility and cocrystal formation of pharmaceutical compounds

A pairwise surface contact equation of state: COSMO-SAC-Phi

Machine learning‐based atom contribution method for the prediction of surface charge density profiles and solvent design

Contact Info

Product

Resources

About