TY - JOUR
T1 - Model comparison for the prediction of the solubility of green tea catechins in ethanol/water mixtures
AU - Sevillano, David Méndez
AU - Van Der Wielen, Luuk A.M.
AU - Trifunovic, Olivera
AU - Ottens, Marcel
PY - 2013/5/1
Y1 - 2013/5/1
N2 - This article evaluates some of the most used solubility models and their applicability to antioxidants, namely, the family of catechins. From the list of possible catechins, the main four were selected for this study: epicatechin, epigallocatechin, epicatechin gallate, and epigallocatechin gallate. Both the enthalpy of fusion and melting temperature were measured experimentally for these four molecules and compared with the values obtained using predictive methods. Solubility measurements for mixtures of water and ethanol were performed with high-throughput experimentation at temperatures of 293 and 303 K. The experiments show a strong nonideal behavior in both pure and mixed solvents. A comparison of different models was performed, identifying the advantages and disadvantages of each model and their performance in this particular case. Finally, the NRTL-SAC model was found to be the most accurate model for predicting solubility, with an RMSE of 0.550.
AB - This article evaluates some of the most used solubility models and their applicability to antioxidants, namely, the family of catechins. From the list of possible catechins, the main four were selected for this study: epicatechin, epigallocatechin, epicatechin gallate, and epigallocatechin gallate. Both the enthalpy of fusion and melting temperature were measured experimentally for these four molecules and compared with the values obtained using predictive methods. Solubility measurements for mixtures of water and ethanol were performed with high-throughput experimentation at temperatures of 293 and 303 K. The experiments show a strong nonideal behavior in both pure and mixed solvents. A comparison of different models was performed, identifying the advantages and disadvantages of each model and their performance in this particular case. Finally, the NRTL-SAC model was found to be the most accurate model for predicting solubility, with an RMSE of 0.550.
UR - http://www.scopus.com/inward/record.url?scp=84877042841&partnerID=8YFLogxK
U2 - 10.1021/ie400113t
DO - 10.1021/ie400113t
M3 - Article
AN - SCOPUS:84877042841
SN - 0888-5885
VL - 52
SP - 6039
EP - 6048
JO - Industrial and Engineering Chemistry Research
JF - Industrial and Engineering Chemistry Research
IS - 17
ER -