Modelagem da adsorção de diclofenaco sódico em leito fixo com carvão ativado: estimativa de parâmetros por monte carlo via cadeia de markov e análise de aumento de escala

Abstract

The presence of emerging pollutants in aquatic ecosystems has led to adverse effects on species life cycles and the food chain. Among the mitigation strategies for this environmental issue, adsorption stands out due to its effectiveness and economic viability. In this context, the present study employed Bayesian modeling to evaluate the dynamic behavior of fixed-bed columns used for the removal of sodium diclofenac by activated carbon. The approach coupled various adsorption isotherms—Langmuir, Freundlich, Sips, and Redlich-Peterson—with a differential equation derived from the liquid-phase mass balance. For the solid phase, a driving force model was adopted, with the exponent gamma varying within the range 0<γ≤2. Parameter estimation was performed using the Markov Chain Monte Carlo (MCMC) method, enabling predictive and reliable analyses even for data not included in the likelihood. The Bayesian Information Criterion (BIC) indicated that the Sips isotherm provided the best fit to the observed behavior. The modeling was validated against the experimental dataset demonstrated a strong agreement between the model estimates and the observed data, highlighting the potential of this modeling approach for future applications and scale-up strategies.