G. Févotte

Simulation and analysis of industrial crystallization processes through multidimensional population balance equations. Part 1: a resolution algorithm based on the method of classes

In order to obtain constant solid properties with particles exhibiting a low order of symmetry, it is necessary to monitor and to control several distributed parameters characterising the crystal shape and size. A bi-dimensional population balance model was developed to simulate the time variations of two characteristic sizes of crystals. The nonlinear population balance equations were solved numerically over the bi-dimensional size domain using the so-called method of classes. An effort was made to improve usual simulation studies through the introduction of physical knowledge in the kinetic laws involved during nucleation and growth phenomena of complex organic products. The performances of the simulation algorithm were successfully assessed through the reproduction of two well-known theoretical and experimental features of ideal continuous crystallization processes: the computation of size-independent growth rates from the plot of the steady-state crystal size distribution and the possibility for MSMPR crystallizers to exhibit low-frequency oscillatory behaviours in the case of insufficient secondary nucleation.

On-line ATR FTIR measurement of supersaturation during solution crystallization processes. Calibration and applications on three solute/solvent systems

Abstract Recently, fourier transform infraRed (FTIR) spectroscopy was reported as a promising technique for in situ measurement of supersaturation during solution crystallization processes. The attenuated total reflection (ATR) probes appeared to be particularly suited to the chemical and physical complexity of industrial suspensions. However, to be used in an industrial context, the technique must be easily and quickly adaptable to different systems. In order to achieve such an aim, a calibration procedure to monitor supersaturation from FTIR measurements is presented. General comments and recommendations about important technical aspects of the technique are also given. Then the technique is used to develop efficient procedures for the determination of solubility and metastability curves. The crystallization of three fine chemical products is studied, thanks to on-line FTIR measurements of supersaturation and off-line CSD determinations. It is shown that the monitoring of supersaturation is a valuable tool for an improved analysis of key phenomena involved during crystallization processes (primary and secondary nucleation, agglomeration, phase transition, seeding, etc.).

New perspectives for the on-line monitoring of pharmaceutical crystallization processes using in situ infrared spectroscopy.

Chemists and engineers involved in the industrial production of solid drugs have to deal with difficult new challenges, including the on-line mastery of the crystal habits and size distribution, the control of polymorphic transitions or the improvement of the chemical purity. A major limitation to improving the control of industrial crystallizers lies in the lack of versatile, accurate and reliable on-line sensors. It is shown that supersaturation measurements can be performed using in situ ATR mid-infrared spectroscopy thus providing valuable real-time information about the crystallization process. Several case studies are presented to illustrate new potential applications of the technique. The reported experimental results outline recent advances in the acquisition of key data characterizing the solute/solvent system in question (i.e. solubility, metastability, phase transformations...), the design of on-line control strategies capable of improving both the crystal size distribution (CSD) and the reproducibility of the quality of the final product, the assessment of improved operating strategies (e.g. seeding batch crystallizers), and the monitoring of polymorphic transitions during cooling crystallization operations. The possibility of evaluating on-line the process impurities, which could allow the reduction of batch-to-batch variations of the quality of the solid product, is also briefly envisaged.

Simulation and analysis of industrial crystallization processes through multidimensional population balance equations. Part 2: a study of semi-batch crystallization

A bi-dimensional population balance model was presented in the previous part of this series of papers to simulate the time variations of two characteristic sizes of hydroquinone particles during crystallization. The multidimensional population balance equations combined with kinetic models and mass balance equations were shown to allow the simulation of the solution crystallization of hydroquinone characterized by a rod-like habit. Semi-continuous isothermal operations were performed at the lab-scale in the presence of various additive concentrations. Both the experimental solute concentration trajectory and the final bi-dimensional crystals size distribution were correctly predicted by the model. The simulated elongation shape factor characterizing the crystal shape was therefore in agreement with the experimental one. Due to the use of tailor-made additive, inhibition effects were observed to affect both primary nucleation and growth kinetics in the length direction. For secondary nucleation, indirect effects were assumed to occur which allowed satisfactory predictions of the final number of fine particles. The representation of the kinetics involved required the evaluation of a set of nine parameters. As a result it was observed that the elongation ratio characterizing the shape of the rod-like particles increases with the length in a nonlinear way. A major interest of the two-dimensional model lies in its ability to relate the time variations of the crystal habit: the particles lengthen in the first moments of their growth and then progressively get thicker until the end of the process.

An Integrated Approach to Polymer Reaction Engineering: A Review of Calorimetry and State Estimation

ABSTRACT In order to completely master a polymerisation process, it is necessary to take a pluri-disciplinary approach that incorporates the use of kinetic and polymer property models (e.g. of Rp, the rate of polymerisation and of copolymer composition distribution, CCD), hardware and software sensors, non-linear observers for the data interpretation, and easily tuneable, robust controllers for the optimal and safe operation of reactors that produce polymer of a pre-specified quality. An overview of the results obtained in this area show that such an approach allows not only to follow and control production rates and the evolution of the CCD, but to obtain other useful process information on the evolution of such properties as the number of particles per litre of emulsion (Np) and the number of radicals per particle () in emulsion polymerisation, or the overall termination rate constant in solution reactions. It is also shown that this method can be used not only for reactor control, but also for fault detection and process optimisation.

On‐line monitoring of methyl methacrylate–vinyl acetate emulsion copolymerization

An adaptive calorimetric method, coupled with state estimators for emulsion copolymerization, is shown to provide accurate, on-line information on the evolution of the composition and kinetics of an emulsion copolymerization. This method was evaluated for the emulsion copolymerization of methyl methacrylate–vinyl acetate (MMA–VAc) under nonisothermal conditions. In addition to providing on-line estimates of the number of moles of each polymerizing species in the reactor, the state estimator provides a value for a lumped kinetic parameter proportional to the product of nNp. This information can be combined with off-line measurements to study the evolution of polymerization kinetics and to explain the trends observed for the molecular weight distribution and glass transition temperatures. Values of n were found to vary from 0.5 to 30 for the homopolymerization of MMA. However, the presence of VAc in the copolymerizing system drastically reduces n. This can lead to a dramatic increase in the average molecular weight of the copolymer since it alters the ratio of propagation to termination in the polymerizing particles.

Real-time evaluation of the concentration of impurities during organic solution crystallization

ATR FTIR spectroscopy was previously presented as a promising technique for measuring supersaturation on-line during crystallization operations, without requiring complex hardware developments. The present paper reports new experimental results where, in addition to the determination of the concentration of the main dissolved product, the real-time measurement of the concentration of impurities is shown to be feasible before the onset of primary nucleation. Using spectral data obtained with both clear solutions and suspensions, the calibration of the in situ infrared sensor was performed to relate the time variations of the concentrations of the main product and of its main impurity, in the presence of both pH and temperature variations. As far as the on-line monitoring of crystallizing slurries is concerned, it is shown that measuring the concentration of impurities is very difficult to perform. However, the quantitative evaluation of impurities present in the clear solution feeding the crystallizer is valuable and expected to allow further design of monitoring and control policies leading to the reduction of batch-to-batch variations of the final quality of the solid produced.

ON-LINE MONITORING OF EMULSION TERPOLYMERIZATION PROCESSES

A new estimator for the on-line monitoring of emulsion terpolymerization is presented. A high gain observer is used to estimate the individual monomer conversions, the instantaneous and cumulative terpolymer composition, and the total number of moles of radicals in the polymer particles on-line. The estimator requires the input of the overall conversion. This input is obtained by calorimetry in the current case, but can be provided by any other type of in-line probe. The estimator has been experimentally validated (on- and offline) for hydrophobic and slightly water soluble monomers, in 3 liter and 250 liter pilot scale reactors. The results obtained demonstrate the usefulness of the high gain approach.

On-Line Monitoring of Emulsion Copolymerisations Using Hardware Sensors and Calorimetry

ABSTRACT We have used a simplified model of emulsion polymerisation, along with an adaptive calorimetric approach and a set of non-linear state estimators to monitor the individual monomer conversions in a pilot scale polymerisation reactor for the production of methyl methacrylate/butyl acrylate/carboxylic acid polymers. The usefulness of reaction calorimetry for the monitoring of such systems is demonstrated. It is also shown that non-linear, high gain estimators can be used to monitor such systems even if they are based on simplified models that ignore polymerisation of the two principal monomers in the aqueous phase, and of the carboxylic acid. The advantages and disadvantages of several different types of on-line sensors for use in industrial situations are also reviewed.

On-Line Monitoring and Modelling of Free Radical Copolymerisations: Butyl Acrylate/Vinyl Acetate

ABSTRACTAfter demonstrating the efficiency of a previously-developed method for the estimation of the rate of heat generation in a stirred tank polymerisation reactor, the use of a non-linear state observer that exploits this rate of heat generation for the evaluation of the copolymer composition, overall conversion, as well as of certain kinetic parameters associated with the reaction was tested. The results of the estimator were compared with both theoretical modelling studies and experimental results. It was also demonstrated that even in dilute solution polymerisation it is necessary to account for variation of the homopolymerisation rate constants with the concentration in the reaction medium. Standard kinetic models, when used with variable parameters, predict the heat of reaction much better than the same models with fixed parameters. This suggests that calorimetry could provide a valuable tool for the estimation of the kinetic parameters in a multiple monomer system.