Comparative studies on parametric sensitivity analyses using conventional and factorial design methodologies: mathematical modelling of clavulanic acid adsorption on zeolites

Abstract

BACKGROUND: The main objective of this work was to evaluate the parametric sensitivity of model parameters of clavulanic acid (CA) adsorption in a stirred‐tank reactor using zeolites. The system sensitivity was evaluated in terms of mass transfer and hydrodynamic parameters using two methodologies. First, the sensitivity factors of input variables were evaluated according to a disturbance of ± 20% in the nominal parameter values; second, the factorial design methodology was employed. In both cases, the responses were obtained by means of model simulations. RESULTS: From the sensitivity analysis by experimental design, the best operational conditions for adsorption of CA on zeolites were solid to liquid percentage 7.0%, initial CA concentration 300 mg L−1 and particle diameter 0.100 mm, which led to a reduction of 25 min in the adsorption time and 9% in the adsorption capacity, increasing the process productivity. CONCLUSION: The use of the factorial design had advantages compared with the conventional method because it provided information regarding the possibility of simultaneous changes in the factors investigated, enabling analyses of system sensitivity regarding operational parameters, as well as the choice of more suitable parameters. Copyright © 2012 Society of Chemical Industry

Estudo da adsorção de ácido clavulânico em hidróxidos duplos lamelares

“…The system was evaluated in terms of equilibrium and mass-transfer parameters, namely, linear model adsorption constant (klin), film coefficient (kf) and effective diffusivity (Def) from changes of ± 50 %. Simulations were carried out using the validated model to perform the parametric sensitivity through an alternative methodology previously suggested (Forte et al, 2012b): the factorial design methodology was used to evaluate the effects of input changes on the process sensitivity. In this case, a central composite rotatable design (CCRD) was proposed evaluating the effects of klin, kf and Def on the required time to the adsorption equilibrium (t * ), which was considered as the related time to a dimensionless concentration C * /C0 = 0.1.…”