(604g) A Multidimensional Population Balance Model for Growth and Dissolution Identified From a Designed Temperature-Cycling Experiment

A
large proportion of pharmaceutical crystallizations produce rod-like crystals,
which can cause problems in process operations downstream from the
crystallizer. Many recent efforts have been directed towards the construction
of multidimensional population balance models suitable for the design of
process operations to optimize crystal shape [1-11]. These models require the
estimation of growth kinetics along multiple axes, with most methods based on
sampling the slurry during crystallization, or on employing imaging technology
that only applies for low crystal number densities or is not easy to implement
in an industrial environment [12-21]. While it is not possible to construct an
arbitrary two-dimensional crystal size distribution (2D CSD) from measurements
of the chord length distribution (CLD) obtained by Focused Beam Reflectance
Measurement (FBRM, which is a probe for in-situ laser backscattering
measurements of particles), application to a specific solute-solvent system is
simpler due to limited range of 2D CSDs that can occur for the allowed
variation in seed mass and temperature profiles.

Previously,
we have reported the application of the commonly available FBRM for the in-situ simultaneous measurement of
crystal shape and size for rod-like organic
crystals during a temperature cycling experiment, which resulted
in substantial continual increases in aspect ratio, mean crystal length, and
mean crystal width with cycle number. This
presentation will describe the construction of a multidimensional population
balance model for growth and dissolution rates along the two axes. The
experimental observations motivated the
incorporation of size-dependent dissolution into the multidimensional population
balance model

                                                                     (1)

where f is
the 2D population density, t is time,
and the growth and dissolution rates in the length and width directions are 2D
generalizations of the 1D equations [22], i.e.,

                                                                                (2)

and the d_ij and γ
define size dependencies. The size dependency broadens the crystal size distribution,
which result in more crystals becoming dissolved in each cycle, as required to
explain the experimentally observed systematic increase in both the average
length and width of crystals with cycle number. A parameter estimation code (the
numerical algorithms and software are similar to that used in [23] and [24])
employed a successive quadratic programming algorithm to minimize the
sum-of-squared-deviations between the model predictions and experiments,

              (3)

where θ
is the vector of model parameters and α
weighs the confidence in each of the mean length and width measurements
relative to the solute concentration measurements and takes into account the
differing units and number of experimental data points. A Pareto-optimality
plot of squared-error values was constructed with each point being associated
with a value of α, which ensured
that the value of α of 0.00008 reasonably
weighed the sets of solute concentration and mean length/width data.

Compared to previous multidimensional population
balance models (PBMs), this PBM includes the effects of dissolution rates that
are estimated from experimental data. Although the PBM's modeling of the size
dependency of dissolution is rather simplistic, the model fits the experimental
data with limited number of parameters with a relatively small level of
uncertainty in the parameters that was quantified by first-order multivariate
statistical analysis. In addition, the model simulation at optimal kinetic parameters
was able to predict the crystal size and shape distribution in a different
temperature-cycling experiment
reasonably well. These results suggest
that growth and dissolution rates estimated in a single temperature-cycling
experiment may be suitable for designing a temperature-cycling protocol for
optimizing the mean length, mean width, mean aspect ratio, or some weighted
combination of the three characteristics of the product crystals. These results
also suggest that an automated protocol may be developed that can assess the
potential usefulness of temperature cycling for removing or reducing
undesirable crystal size and shape characteristics produced by more standard
crystallization operations.

References

[1] D. B. Patience and J. B. Rawlings, "Particle-shape
monitoring and control in crystallization processes," AIChE
J., vol. 47, pp. 2125-2130, 2001.

[2] R. Gunawan, D. L. Ma, M. Fujiwara and R. D. Braatz,
"Identification of kinetic parameters in multidimensional crystallization
processes," Int. J. Mod. Phys. B, vol. 16,
pp. 367-374, 2002.

[3] K. Lee, J. H. Lee, M. Fujiwara, D. L. Ma and R. D. Braatz,
"Run-to-run control of multidimensional crystal size distribution in a
batch crystallizer," in Proceedings of the American Control Conference, pp. 1013-1018, 2002.

[4] F. Puel, G. Févotte and J. P. Klein, "Simulation and analysis of
industrial crystallization processes through multidimensional population
balance equations. Part 1: A
resolution algorithm based on the method of classes," Chemical Engineering
Science, vol. 58, pp. 3715-3727, 2003.

[5] H. Briesen,
"Simulation of crystal size and shape by means of a reduced
two-dimensional population balance model," Chemical Engineering
Science, vol. 61, pp. 104-112, 2006.

[6] C. Y. Ma, X. Z. Wang and K. J. Roberts, "Multi-dimensional
population balance modeling of the growth of rod-like L-glutamic acid crystals
using growth rates estimated from in-process imaging," Advanced Powder
Technology, vol. 18, pp. 707-723, 2007.

[7] C. Y. Ma, X. Z. Wang and K. J. Roberts, "Morphological population
balance for modeling crystal growth in face directions," AIChE J., vol. 54, pp. 209-222, 2008.

[8] C. Y. Ma and X. Z. Wang, "Crystal growth rate dispersion modeling
using morphological population balance," AIChE
J., vol. 54, pp. 2321-2334, 2008.

[9] J. Wan, X. Z. Wang and C. Y. Ma, "Particle shape
manipulation and optimization in cooling crystallization involving multiple
crystal morphological forms," AIChE
J., vol. 55, pp. 2049-2061, 2009.

[10] J. J. Liu, C. Y. Ma, Y. D. Hu and X. Z. Wang, "Effect of seed
loading and cooling rate on crystal size and shape distributions in protein
crystallization-A study using morphological population balance
simulation," Computers and Chemical Engineering, vol. 34, pp.
1945-1952, 2010.

[11] J. J. Liu, C. Y. Ma, Y. D. Hu
and X. Z. Wang, "Modelling protein crystallisation using morphological population balance
models," Chem. Eng. Res. Design, vol. 88, pp. 437-446, 2010.

[12] F. Puel, P. Marchal and J. Klein, "Habit transient analysis in
industrial crystallization using two dimensional crystal sizing
technique," Chem. Eng. Res. Design, vol. 75, pp. 193-205, 1997.

[13] D. L. Ma and R. D. Braatz, "Worst-case analysis of finite-time
control policies," IEEE Trans. Control Syst. Technol., vol. 9, pp.
766-774, 2001.

[14] J. Calderon De Anda, X. Z. Wang, X. Lai and
K. J. Roberts, "Classifying organic crystals via in-process image analysis
and the use of monitoring charts to follow polymorphic and morphological
changes," J. Process Control, vol. 15, pp. 785-797, 2005.

[15] J. De Calderon Anda, X. Z. Wang, X. Lai, K. J. Roberts, K. H. Jennings, M.
J. Wilkinson, D. Watson and D. Roberts, "Real-time product morphology
monitoring in crystallization using imaging technique," AIChE J., vol. 51, pp. 1406-1414, 2005.

[16] J. De Calderon Anda, X. Z. Wang and K. J. Roberts, "Multi-scale
segmentation image analysis for the in-process monitoring of particle shape
with batch crystallisers," Chemical
Engineering Science, vol. 60, pp. 1053-1065, 2005.

[17] P. A. Larsen, J. B. Rawlings and N. J. Ferrier, "An algorithm for
analyzing noisy, in situ images of high-aspect-ratio crystals to monitor
particle size distribution," Chemical Engineering Science, vol. 61,
pp. 5236-5248, 2006.

[18] M. Oullion, F. Puel, G. Févotte, S. Righini and P. Carvin,
"Industrial batch crystallization of a plate-like organic product. In situ monitoring and 2D-CSD modelling: Part 1: Experimental study," Chemical
Engineering Science, vol. 62, pp. 820-832, 2007.

[19] J. Eggers, M. Kempkes and M. Mazzotti,
"Measurement of size and shape distributions of particles through image
analysis," Chemical Engineering Science, vol. 63, pp. 5513-5521,
2008.

[20] X. Z. Wang, J. Calderon De Anda and K. J.
Roberts, "Real-time measurement of the growth rates of individual crystal
facets using imaging and image analysis: A feasibility study on needle-shaped
crystals of L-glutamic acid," Chem. Eng. Res. Design, vol. 85, pp.
921-927, 2007.

[21] X. Z. Wang, K. J. Roberts and
C. Ma, "Crystal growth measurement using 2D and 3D imaging and the
perspectives for shape control," Chemical Engineering Science, vol.
63, pp. 1173-1184, 2008.

[22] M. Shoji, T. Eto
and H. Takiyama, "A kinetic study of the
influence of modulated undersaturation operation on
crystal size distribution in cooling-type batch crystallization," J.
Chem. Eng. Japan, vol. 44, pp. 191-196, 2011.

[23] R. Gunawan, I. Fusman and
R. D. Braatz, "High resolution algorithms for multidimensional population
balance equations," AIChE J., vol.
50, pp. 2738-2749, 11, 2004.

[24] T. Togkalidou, H. -. Tung, Y. Sun, A. T.
Andrews and R. D. Braatz, "Parameter estimation and optimization of a
loosely bound aggregating pharmaceutical crystallization using in situ infrared
and laser backscattering measurements," Industrial and Engineering
Chemistry Research, vol. 43, pp. 6168-6181, 2004.