A major challenge in the production of metabolites by plant cells is the separation and purification of a desired product from a number of impurities. An important application of plant cell culture is the biosynthesis of the anticancer agent paclitaxel. Liquid–liquid extraction plays a critical role in the recovery of paclitaxel and other valuable plant‐derived products from culture broth. In this study, the extraction of paclitaxel and a major unwanted by‐product, cephalomannine, from plant cell culture broth into organic solvents is quantified. Potential solvent mixtures show varying affinity and selectivity for paclitaxel over cephalomannine. The partition coefficient of paclitaxel is highest in ethyl acetate and dichloromethane, with measured values of 28 and 25, respectively; however, selectivity coefficients are less than 1 for paclitaxel over cephalomannine for both solvents. Selectivity coefficient increases to 1.7 with extraction in n‐hexane, but the partition coefficient decreases to 1.9. Altering the pH of the aqueous phase results in an increase in both recovery and selectivity using n‐hexane but does not change the results for other solvents significantly. The addition of extractants trioctylamine (TOA) or tributylphosphate (TBP) to n‐hexane gives significantly higher partition coefficients for paclitaxel (8.6 and 23.7, respectively) but no selectivity. Interestingly, when 20% hexafluorobenzene (HFB) is added to n‐hexane, the partition coefficient remains approximately constant, but the selectivity coefficient for paclitaxel over cephalomannine improves to 4.5. This significant increase in selectivity early in the purification process has the potential to simplify downstream processing steps and significantly reduce overall purification costs. © 2012 American Institute of Chemical Engineers Biotechnol. Prog., 28: 990–997, 2012

We have previously developed a population balance equation (PBE) model for emulsion drop breakage in a high-pressure homogenizer that incorporated multiple-drop formation within two mechanisms of turbulent drop breakage. The model was found to satisfactorily predict the effects of formulation variables on the drop-size distribution, but the model was not extensible to a range of homogenization pressures. The objective of this paper is to determine the additional model elements necessary to obtain acceptable predictions over a wide range of pressures. The most significant improvements were obtained by increasing the number of daughter drops formed upon breakage from 20 to 150 drops and by introducing a maximum stable diameter, below which drops could not break. Smaller improvements were obtained by introducing terms that describe the loss of energy available for drop breakage due to thermal heating of the sample and homogenizer and by extending the model to account for the effects of surfactant adsorption and deficiency on the interfacial tension. The simultaneous implementation of all five enhancements was shown to produce a 62% improvement over the previous model, as measured by a least-squares objective based on the difference between the measured and predicted drop-size distributions over five homogenization passes and five pressures in the range 250-1250 bar. The resulting model was also validated over a range of oil and surfactant concentrations and shown to provide satisfactory predictions.

Batch reactive distillation, which combines the flexibility of batch process with the advantages of reactive distillation, can be an effective alternative to conventional batch processing. For instance, acetate esters such as n-butyl acetate and amyl acetate can be produced using batch reactive distillation. Previous studies have shown that isopropyl acetate can also be produced in batch reactive distillation with the addition of chloroform as an entrainer and using a large reflux ratio and number of stages. This paper compares the vapor-liquid and liquid-liquid features of these acetate ester systems, and shows the similarities and differences among them. We show why the isopropyl system is more difficult than the others and how semibatch reactive distillation without an additional entrainer can result in a feasible and more efficient process for the production of isopropyl acetate than has been previously known. In a semibatch reactive distillation, the loss of isopropyl alcohol can be substantially decreased, the purity of water byproduct can be improved, and the total reflux or recycle of the organic distillate phase can be avoided. The semibatch reactive distillation can provide 20% higher production efficiency than batch reactive distillation.

Most population balance equation (PBE) models of emulsion drop breakage are based on the assumption of binary drop breakage. We previously developed such a PBE model for high-pressure homogenizers with a daughter drop distribution function exhibiting a maximum probability for two equal sized drops. In this paper, we present a PBE model accounting for multiple drop breakage and show that the model provides superior distribution predictions reflected by decrease in least-squares objective function for an oil-in-water emulsion processed in a pilot-scale high-pressure homogenizer. Following our previous work, two distinct rate functions for drop breakage, one due to turbulent eddies and another due to turbulent shear were used to reproduce the measured bimodal distributions. We found that multiple drop breakage was satisfactorily modeled with a uniform daughter drop distribution function if the assumed number of daughter drops formed was chosen to be sufficiently large. The PBE model with multiple drop breakage was shown to provide superior distribution predictions compared to the analogous binary breakage PBE model when adjustable model parameters were determined by nonlinear optimization. The multiple breakage PBE model was shown to be extensible to different emulsion formulations by using these base case model parameters to predict the effects of oil concentration, surfactant concentration, oil-to-surfactant ratio and emulsion premix distribution. Our experiments revealed that substantial breakage of the premix occurred during the first homogenization pass even under zero applied homogenization pressure operation, suggesting an unmodeled pressure independent breakage mechanism.

This Article provides new predictions for selectivity in batch reactive distillation, identifying the reflux or reboil ratio and a Damkohler number (Da) as the key operating parameters. The dimensionless Da incorporates the influence of liquid holdup, vapor rate, and rate of reaction. Example results for a system of serial isomerization reactions and for the synthesis of ethylene glycol are provided. The results show that selectivity improvements in BRD are limited for high values of Da or for high values of the reflux or reboil ratio and that selectivity is enhanced as Da or reflux or reboil ratio is decreased. However, decreasing Da can cause conversion loss, which can be mitigated by increasing the reflux (or reboil) ratio at the expense of selectivity. Consequently, there is an optimum value of reflux or reboil ratio that gives a maximum yield for systems operated at low or moderate Da. For the isomerization in a BRD, the heat released by reaction can improve selectivity at the expense of conversion. For ethylene glycol synthesis at a low reboil ratio, BRD has a negative impact on both conversion and selectivity by causing separation of the reactants. We also show that decreasing the reboil ratio near the end of the BRD can increase the removal rate of EG and thereby improve selectivity. This operating strategy is different from a common operating strategy in distillation of increasing reboil ratio near the end of a batch or cut. We also find that an operation with a constant volumetric liquid flow rate provides lower selectivity than a constant molar liquid flow rate.

Emulsions are usually generated in high-pressure homogenization chambers. The flow field is these units are typically highly turbulent and chaotic, and mechanisms for drop formation under these conditions are not well understood. In this paper, we have applied the PBE modeling approach to droplet break-up in a high pressure homogenizer using mechanistic functions for breakage rate. We have compared our modeling results to experimental data that we have obtained on a model oil-in-water emulsion. In principle, once these functions are known, the PBE approach can be used in a predictive manner to aid in the selection of process and product variables that will lead to the desired drop size distribution. This was verified for a number of test cases by changing product properties and homogenizing conditions. We observed that the population balance model did a reasonably good job of predicting the drop size distribution and therefore look promising. For cases where the model fails, we discuss strategies for improving predictions for these types of systems.

A population balance equation (PBE) model for pure drop breakage processes was developed from homogenization experiments and used to investigate model extensibility over a range of emulsion formulation and homogenizer operating variables. Adjustable parameters in the mechanistic breakage functions were estimated from measured drop volume distributions by constrained nonlinear least-squares optimization. Satisfactory prediction of measured bimodal distributions was achieved by the incorporation of two different breakage functions that accounted for large drop breakage due to turbulent shear and for small drop breakage due to collisions between drops and turbulent eddies. Model extensibility to different emulsion compositions and homogenizer pressures was investigated by comparing model predictions generated with the base case parameters to drop volume distributions measured under different conditions. The PBE model satisfactorily accounted for changes in the dispersed phase volume fraction and the interfacial tension with the base case parameters. By contrast, significantly improved predictions for the continuous phase viscosity or multiple formulation variables were obtained through re-estimation of the model parameters using multiple data sets in which the associated variables were systematically varied. The model was not able to satisfactorily predict drop volume distributions resulting from homogenizer pressure changes, perhaps due to the assumption of a constant pressure throughout the homogenizer. We conclude that PBE models of drop breakage can be used to reasonably predict the effects of emulsion formulation variables on drop volume distributions and have the potential for guiding experimental efforts aimed at the design of novel emulsified products.

We have developed a quick and effective method to predict the feasible products from a double-feed reactive distillation column. The method relies on the prediction of the pinch point for the middle section, using a cross-flow arrangement of vapor-liquid continuously stirred tank reactors (CSTRs). The cross-flow model substantially simplifies the prediction of the compositions at both ends of the middle section of a two-feed column, which are then used to predict the feasible distillate and bottoms compositions, using existing methods. The cross-flow arrangement can also be used to predict minimum and maximum flows for double-feed reactive distillation columns.

We investigate the sensitivity of an inverse population balance equation (PBE) modeling technique for extracting single particle functions from transient size distribution measurements. A dynamic PBE model of a turbulently agitated batch emulsification vessel is used to generate volume size distribution data under the assumption of negligible drop coalescence. The distribution data are subjected to various types of error consistent with available measurement technologies and then introduced as input data to the inverse PBE modeling algorithm, which includes validation of the self-similar assumption. The errors considered include measurement noise, data skewed towards smaller or larger drops, skewed data due to the presence of large dust peaks, and reduced resolution caused by data binning. For each case, the computed functions for the drop breakage rate and the distribution of daughter drops are compared to the actual functions to assess the impact of input data errors on the effectiveness of the inverse PBE modeling approach. The type of measurement errors considered generally lead to underprediction of the breakage rate and, consequently, to overprediction of the number of large drops. Because the estimated and actual breakage rates tend to converge at small drop sizes, the inverse algorithm generates accurate predictions of the drop size distribution at sufficiently long batch times when small drops dominate. Implications for our future work on PBE modeling of drop size distributions in pharmaceutical emulsions prepared with high pressure homogenization are discussed.

Bifurcation studies predict limited ranges of feasibility for products in certain reactive distillations. These are closely related to the bifurcations in the singular points of dynamic models for simple reactive distillation (isobaric open evaporation with liquid phase reaction). A new dynamic model is described with constant vapor rate together with an experimental study for the reactive distillation of acetic acid with isopropanol to produce isopropyl acetate, catalyzed by Amberlyst-15 ion-exchange resin. An experimental apparatus with real-time measurement of liquid compositions based on Fourier transform infrared (FTIR) spectroscopy is described, and used to follow the composition dynamics at several initial conditions and Damkohler numbers (Da). The experimental results match model predictions that show four regions of behavior. For Da approximate to 1, these show a stable node at acetic acid and several other fixed points as saddles. However, near Da 2, both isopropanol and acetic acid are stable nodes and a quaternary singular point appears. The presence of two stable nodes requires the presence of a distillation boundary and, therefore, a limited feasibility for the bottom product compositions from continuous reactive distillation. For the reaction rates studied, the model predictions are closely consistent with the experimental findings, and are robust to variations in the vapor rate. These experiments are among the first to analyze the dynamics and feasibility in a kinetically-controlled reactive distillation and are consistent with previous studies for the reaction equilibrium limit, indicating the formation of a reactive azeotrope. (C) 2005 American Institute of Chemical Engineers.

Necessary conditions for steady-state multiplicity in an isobaric reactive flash with constant split fraction are established. Multiplicity is caused by the interaction between separation and reaction in systems in which the activation energy and the gradient of boiling temperature with composition are sufficiently large. This interaction may cause the net reaction rate to decrease even with an increase in reactant mole fraction. Parametric studies and asymptotic arguments show that the ratio of the activation energy to the heat of vaporization must be greater than one for multiplicity to occur. Even when no vapor is generated, i.e. a CSTR whose liquid contents are at the boiling point, multiplicity can occur. Multiplicity is possible in endothermic systems as well as those with a small heat of reaction, which typically do not display multiplicity in the one phase CSTR. Several examples are considered, including the synthesis of ethylene glycol.