Process variants

Process variants

One main focus of the crystallization-related research of our group is the development of advanced and alternative process variants for separation problems. This concerns separation tasks where one specific target component has to be isolated from a complex multi-component mixture (such as a drug molecule out of a plant extract) and also cases where just two components have to be separated which are hardly resolvable by crystallization (such as enantiomers and other very similar isomers).

Process variants studied refer to continuous, multistage and batch operation modes, advanced process strategies for enantiomer separation (coupled reactor concepts (fluidized bed and MSMPR reactors), application of “eutectic shifts”, innovative strategies of preferential crystallization), separation concepts of solid solutions and the recovery of solid lignin from organosolv pulping liquors.

In the related projects feasibility, yield, productivity and product properties achievable with the different process options are examined and the potential applicability of the concepts at industrial scale is evaluated. The application of improved measurement techniques and rationalization of process design are also considered.


Table of content:

Modelling and application of crystallization concepts for separation of enantiomers

Methods and rational approach for batch enantioselective crystallization processes

Continuous crystallization strategies

Counter-current crystallization for the separation of solid solutions

 

Modelling and application of crystallization concepts for separation of enantiomers

Each enantiomer of a pair presents different pharmacological activity and there is an increasing demand for producing pure chiral molecules. Enantiomer separation can be achieved by two main approaches: asymmetric synthesis or nonselective synthesis followed by chiral resolution [1]. A great effort has been done to develop separation techniques to obtain pure enantiomers. Crystallization is an attractive process since it provides a solid product, as frequently desired form for pharmaceuticals.

Population balance model (PBM) (Equation 1) describes particulate processes [2] and it has been extensively applied to model and optimize Preferential Crystallization (PC) [3]. PBM incorporates the various mechanisms involved in crystallization, e.g. growth, nucleation, breakage, agglomeration. Simplified models to describe PC based on lumped crystallization mechanisms are developed to enable more rapid analysis of important process design parameters [4]. 

Equation 1: Population balance equation for crystal growth.

References

[1] Lorenz, H.; Seidel-Morgenstern, A. (2014) Processes to separate enantiomers. Angew. Chem. Int. Ed., 53, pp. 1218−1250.
[2] Ramkrishna, D. (2000) Population balances. Academic Press, London.
[3] Elsner, M. P., Menéndez, D. F., Muslera, E. A., Seidel-Morgenstern, A. (2005) Experimental study and simplified mathematical description of preferential crystallization. Chirality, 17(S1), S183-S195.
[4] Bhandari S., Carneiro T., Temmel, E.,  Qamar, S., Lorenz, H., Seidel-Morgenstern, A. (2018) Modeling Batch Preferential Crystallization for Conglomerates and Racemic Compounds. In Proceedings of BIWIC 2018; Univ Rouen Havre:  Rouen, France, 2018; pp. 121-125.

Methods and rational approach for batch enantioselective crystallization processes

Preferential Crystallization of conglomerate forming systems is a selective method that can applied as separation technique in the area of the pharmaceutical, agricultural and food industries. The classical single batch approach, in which one crystallizer is seeded with homochiral crystals is an inherently low-robust process and suffers from strong yield limitations. The challenge is to overcome this limitations not only by using a continuous approach [1] but also keeping the process in batch configuration. One of the configurations that can be used in order to increase the performance of the process consists in Coupling Preferential Crystallization with selective Dissolution (CPCD) using two coupled stirred tanks operating at different temperatures (Figure1). This method is capable of providing robustly both pure enantiomers from racemic mixtures introducing only a small amount of seeds of one of the enantiomers [2-3].

Figure 1: Scheme of a CPCD separation: start (top) and end of the process. Adapted from [2].

At the beginning, both tanks are filled with a saturated solution of the racemic material to separate at a certain saturation temperature (Tsat). The crystallization tank is then cooled to a crystallization temperature (Tcryst) while the dissolution tank remains at Tsat. A given mass of racemic solid is added to the dissolution tank creating a suspension. After seeding the crystallization tank with one of the enantiomers, the preferential crystallization takes place leading to a decrease of the concentration of the seeded enantiomer in the mother liquor. Because of the liquid phase exchange between the two tanks, a transient undersaturation of the seeded enantiomer is generated in the dissolution tank leading to its selective dissolution, while the unseeded enantiomer remains in the solid phase.  Crystallization and dissolution continue simultaneously in both tanks until a quantitative dissolution of the seeded enantiomer from the racemic solid in the dissolution tank suspension is achieved.  The application of on line / in line measurements techniques to monitor the progress of the resolution with time allows to control the process parameters in both tanks.

References

[1] Lorenz, H.; Seidel-Morgenstern, A. (2014) Processes to separate enantiomers. Angew. Chem. Int. Ed., 53, pp. 1218−1250.
[2] Ramkrishna, D. (2000) Population balances. Academic Press, London.
[3] Elsner, M. P., Menéndez, D. F., Muslera, E. A., Seidel-Morgenstern, A. (2005) Experimental study and simplified mathematical description of preferential crystallization. Chirality, 17(S1), S183-S195.
[4] Bhandari S., Carneiro T., Temmel, E.,  Qamar, S., Lorenz, H., Seidel-Morgenstern, A. (2018) Modeling Batch Preferential Crystallization for Conglomerates and Racemic Compounds. In Proceedings of BIWIC 2018; Univ Rouen Havre:  Rouen, France, 2018; pp. 121-125.
[5] F. Cascella, E. Temmel, H. Lorenz, A. Seidel-Morgenstern (2018) Fundamental studies for Continuous Preferential Crystallization of Guaifenesin, BIWIC, Rouen (FR), pp. 64-65.
[6] G. Levilain, M. J. Eicke, A. Seidel-Morgenstern (2012) Efficient Resolution of Enantiomers by Coupling Preferential Cystallization and Dissolution. Part 1: Experimental Proof of Principle, Cryst. Growth Des. 12, pp. 5396-5401.
[7] M. J. Eicke, G. Levilain, A. Seidel-Morgenstern (2013) Efficient Resolution of Enantiomers by Coupling Preferential Crystallization and Dissolution. Part 2: A Parametric Simulation Study to Identify Suitable Process Conditions, Cryst. Growth Des. 13, pp. 1638-1648.

Continuous crystallization strategies

Continuous strategies can often improve the process performance in terms of productivity, product quality as well as in terms of reproducibility. The choice of an appropriate concept of continuous crystallization depends, for example, on the product specific aims and requirements. Current work is directed to investigations of two different strategies of continuous crystallization on examples of several life-science products. The first strategy is the Mixed Suspension Mixed Product Removal (MSMPR) process, which is applied for the mass crystallization of a food additive. The second strategy is the application of Coupled Fluidized Bed Crystallizers (CFBC) for the separation of enantiomers. Main focus is here a high product purity, why it is attractive for the production of active pharmaceutical ingredients (APIs).

Mixed Suspension Mixed Product Removal (MSMPR)

Achieving a consistent crystal size distribution is a key element in continuous crystallization. To design and develop a continuous crystallization, preliminary experiments are required, which include solvent screening, solubility and metastable zone measurements, as well as seed preparation studies. The crystallization kinetics e. g. nucleation and growth are dominantly important in determining the crystals shape, size and size distribution as well as to provide information for designing the related crystallization processes, and thus, have to be determined [1-2]. A mathematical model has been developed to analyze the effects of main operating parameters (e. g. mean residence time, initial supersaturation ratio) on the final products properties and productivity. The simulation studies guide the design and development of the continuous crystallization process (Figure 1). Experimental works were also performed in a 25 L draft tube baffle (DTB) crystallizer (Figure 2) while the food additive calcium propionate was applied as the model compound [3]. The DTB crystallizer provides perfect mixing of the suspension to avoid classification of crystals and temperature fluctuations.

Figure 1: Sketch of a single stage MSMPR crystallizer for continuous crystallizations.

Figure 2: Picture of the 25 L tank for continuous crystallizations

Coupled Fluidized Bed Crystallization (CFBC)

The aim of the CFBC is the continuous separation of enantiomers from a 50:50 mixture of both, a so-called racemic mixture. It combines the principle of fluidized bed crystallization with the technique of preferential crystallization (Figure 3). Therefore the CFBC uses two conical shaped tubular crystallizers, which both are seeded with one of the pure enantiomers. The solid racemic mixture to be separated is dispersed in a corresponding saturated solution within a shared feed tank. The crystal-free saturated solution is pumped continuously from the bottom through the tubular crystallizers, while cooling is provided via double-wall jackets. To ensure continuous selective seeding, grown enantiopure crystals are withdrawn at the bottom of the crystallizers, fragmented in a bypass and fed back as seeds to the crystallizers [4-5].

Figure 3: Sketch of utilized CFBC pilot plant (adapted from [5])

The conical shape of the tubular crystallizer causes a varying fluid velocity over the crystallizer height, which results in a classifying effect on the crystals. Hence, the application of CFBC shows particular beneficial features. Currently, the performance of this strategy is studied on the example of two different chiral systems [6-9].

References

[1] Temmel, E., Eisenschmidt, H., Lorenz, H., Sundmacher, K., Seidel-Morgenstern, A. (2016) A Short-Cut Method for the Quantification of Crystallization Kinetics. 1. Method Development. Cryst. Growth Des. 16, pp. 6743-6755.
[2] Temmel, E., Eicke, M., Lorenz, H., Seidel-Morgenstern, A. (2016) A Short-Cut Method for the Quantification of Crystallization Kinetics. 2. Experimental Application. Cryst. Growth Des. 16, pp. 6756-6768.
[3] Li, T., Lorenz, H., Seidel-Morgenstern, A. (2017) Solubility study and thermal stability analysis of calcium propionate. Chem. Eng. Technol. 40, pp. 1221-1230.
[4] Binev, D., Seidel-Morgenstern, A., Lorenz, H. (2015) Study of crystal size distributions in a fluidized bed crystallizer. Chem. Eng. Sci. 133, pp. 116-124.
[5] Binev, D., Seidel-Morgenstern, A., Lorenz, H. (2016) Continuous separation of isomers in fluidized bed crystallizers. Cryst. Growth Des. 16, pp. 1409-1419.
[6] Mangold, M., Khlopov, D., Temmel, E., Lorenz, H., Seidel-Morgenstern, A. (2017) Modelling geometrical and fluid-dynamic aspects of a continuous fluidized bed crystallizer for separation of enantiomers. Chem. Eng. Sci. 160, pp. 281-290.
[7] Kerst, K., Roloff, C., Medeiros de Souza, L.G., Bartz, A., Seidel-Morgenstern, A., Thévenin, D., Janiga, G. (2017) CFD-DEM simulations of a fluidized bed crystallizer. Chem. Eng. Sci. 165, pp. 1-13.
[8] Cascella, F., Temmel, E., Seidel-Morgenstern, A., Lorenz, H. (2018) Fundamental studies for continuous preferential crystallization of guaifenesin. BIWIC, pp. 64-69.
[9] Temmel, E., Gänsch, J., Lorenz, H., Seidel-Morgenstern, A. (2018) Measurement and evaluation of the crystallization kinetics of L‑asparagine monohydrate in the ternary L/D asparagine/water system. Cryst. Growth Des. DOI 10.1021/acs.cgd.8b01322.

Counter-current crystallization for the separation of solid solutions

For several reasons, a single crystallization step might not be sufficient to purify a mixture of two or more components. Possible reasons can be the inclusion of impurities in the crystal lattice by kinetic effects, residual mother liquor attached to crystals after solid-liquid separation or miscibility in the solid state [1]. Especially the latter phenomena, also known as formation of solid solutions or mixed crystals, is a challenging separation problem as the impurity results of the systems thermodynamics and, usually cannot be overcome by seeding.

Classically, fractional crystallization has to be applied to resolve solid solutions. Fig. 1 (left) shows the process sequence of a single fractional crystallization step to separate a mixture of molecules A and B using a solvent in the ternary phase diagram. First, an impure feed is dissolved in a solvent, then supersaturation is generated by solvent removal and therefore the molecules A and B crystallize as an A enriched solid solution and B enriched mother liquor, which is withdrawn as waste fraction. Multiple repetitions of this procedure with the crystallized solid as new feed is capable to purify the mixture up to a highly pure crystalline component A on the one hand, but produces huge amounts of liquid waste, resulting in low process yields.

Similar to rectification or counter-current extraction, this drawback can be resolved when using a cascade of separation stages (crystallizers), in which solid and liquid phases are transported counter-currently as shown in Fig. 2. The transportation of the solid phase in a stage j-1 along this cascade is guaranteed by dissolving it in the mother liquor of stage j+1. Then this solution is transported to stage j, where a new crystallization run is performed. By this sequence, known as HAPILA concept, the typically challenging solid phase transport can be realized easily via its dissolved form [2], producing, in an ideal case, no waste but 2 pure products A and B. Figure 1 (right) illustrates the described sequence in more detail for a crystallizer j in the ternary phase diagram.

Figure 1: Procedure to resolve a solid solution by fractional crystallization (left) and by multistage counter-current crystallization (right), reproduced from [3]. (1): Mixing solid crystals from j-1 with solid feed; (2): mixing solid phase with mother liquor from j+1; (3): Dissolution by solvent addition (transport solution to crystallizer j); (4): Solvent evaporation; (5): Crystallization along characteristic tie line.

Figure 2: Scheme of a counter-current crystallization cascade according to [2], reproduced from [3].

First experimental works were directed to fundamental studies of the solid-liquid equilibrium of interesting inorganic and organic solid solution forming systems in water [4-7], and provided simple empirical expressions to describe the phase diagram mathematically.

A flexible equilibrium mass balance model, using additional solid-liquid equilibrium information, was developed to study and optimize this complicated sequentially operated counter-current crystallization process by simulation [6]. Moreover, the model is used to highlight analogies to more common counter-current processes (e.g. rectification, counter-current extraction or simulated moving bed chromatography) [3].  

Based on this simulation studies a pilot plant using a modified HAPILA concept was constructed to demonstrate the counter-current crystallization principle experimentally.

References

[1] Lorenz, H. and Beckmann, W. (2013) Purification by Crystallization, in Crystallization, Wiley-VCH Verlag GmbH & Co. KGaA, pp. 129-148.
[2] Grawe, D., Eilers, R., and Gliesing, S., US9095783B2, 2009.
[3] Münzberg, S., Vu, T.G., and Seidel-Morgenstern, A. (2018) Chemie-Ingenieur-Technik. 90(11), pp. 1769-1781.
[4] Temmel, E., Wloch, S., Müller, U., Grawe, D., Eilers, R., Lorenz, H., and Seidel-Morgenstern, A. (2013), Chem. Eng. Sci. 114, pp. 662–673.
[5] Isakov, A. I., Kotelnikova, E. N., Muenzberg, S., Bocharov, S. N., and Lorenz, H. (2016), Cryst. Growth & Des. 16(5), pp. 2653-2661.
[6] Münzberg, S., Lorenz, H., and Seidel-Morgenstern, A. (2016), Chem. Eng. Technol. 39(7), pp. 1242-1250.
[7] Münzberg, S., Lorenz, H., and Seidel-Morgenstern, A. (2017), Poster presented at the 20th International Symposium on Industrial Crystallization, Dublin, Ireland.
Go to Editor View