Desulfurization of a commercial diesel fuel by different layered adsorbents and regeneration of the latter were studied in a fixed-bed unit operated at ambient temperature and pressure. In general, a layered bed consisting of 12 wt % activated carbon, 22 wt % of activated alumina (Selexsorb CDX), followed by Cu(I)-Y, activated carbon/Selexsorb CDX/Cu(I)-Y, is capable of producing 41 cm 3 of desulfurized diesel fuel/g of adsorbent. The matrix is capable of processing 27 cm 3 of deep-desulfurized diesel with a weighted average content of 76 ppbw S. These lowsulfur fuels are suitable for fuel cell applications. For layered-bed regeneration, it was determined that calcination of the adsorbed sulfur moieties with air at 350 °C followed by autoreduction of the copper species recovered all of the original desulfurization capacity when activated aluminas are used as a guard layer. Solvent elution experiments indicate that carbon tetrachloride and N,N-dimethylformamide are suitable solvents to recover all of the adsorbed organosulfur species.

Recent advances in statistical procedures, coupled with the availability of high performance computational resources and the large mass of data generated from high throughput screening, have enabled a new paradigm for building mathematical models of the kinetic behavior of catalytic reactions. A Bayesian approach is used to formulate the model building problem, estimate model parameters by Monte Carlo based methods, discriminate rival models, and design new experiments to improve the discrimination and fidelity of the parameter estimates. The methodology is illustrated with a typical, model building problem involving three proposed Langmuir−Hinshelwood rate expressions. The Bayesian approach gives improved discrimination of the three models and higher quality model parameters for the best model selected as compared to the traditional methods that employ linearized statistical tools. This paper describes the methodology and its capabilities in sufficient detail to allow kinetic model builders to evaluate and implement its improved model discrimination and parameter estimation features.

No abstract

The value of reaction kinetic models for manufacturing APIs (active pharmaceutical ingredient) has been well established in the Quality by Design (QbD) paradigm. Creating such models during the early phase...

The use of particle size distribution (PSD) similarity metrics and the development and incorporation of drug release predictions based on PSD properties into PBPK models for various drug administration routes may provide a holistic approach for evaluating the effect of PSD differences on in vitro drug release and bioavailability of disperse systems. The objectives of this study were to provide a rational approach for evaluating the utility of in vitro PSD comparators for predicting bioequivalence for subcutaneously administered test and reference drug emulsions. Two types of in vitro comparators for test and reference emulsion products were evaluated: PSD characterization comparators (overlap metrics, median, and span ratios) and release profile comparators (f and various fractional time ratios). A subcutaneous-input PBPK disposition model was developed to simulate blood concentration-time profiles of reference and test emulsion products and pharmacokinetic responses (e.g., AUC, C, and T) were used to determine bioequivalence. A pool of 10,440 pairs of test and reference products was simulated using Monte Carlo experiments. The PSD and release profile comparators were correlated to pass/fail bioequivalence metrics using logistical regression. Based on the use of single in vitro comparators, the f method was the best predictor of bioequivalence prediction. The use of combinations of f and PSD overlap comparators (e.g., OVL or PROB) improved bioequivalence prediction to about 90%. Simulation procedures used in this study demonstrated a process for developing reliable in vitro BE predictors.