Control over the product crystal size distribution (CSD) is very important in industrial crystallization for efficient downstream processing and improved formulation properties1
Sonocrystallization is a prominent technique to reduce the crystal size and to generate uniform crystal size distributions2, 3
. Current state of the art includes the sonocrystallization assisting PAT(Process Analytical Technology) to better understand the sonocrystallization process to better control the crystal size distribution to meet the specific requirements of an active pharmaceutical ingredient (API). Various approaches to use ultrasound for size distribution control in a batch cooling crystallization have been investigated. Solute concentration is monitored by In-situ React IC-IR and FBRM used to identify the metastable zone width and crystal chord length distributions. Crystallization of Piracetam in Isopropyl alcohol (IPA) was studied in a batch crystallizer with a working volume of 500ml. A calibration based approach was used for the solute concentration measurement. Crystal size measurement using light microscopy and
Image-processing is a promising technique to determine the actual CSD rather than measuring with Laser diffraction technique. Width of the crystal size distribution characterized as Span = (D(90)-D(10))/D(50) determines the final product size distribution. It was observed that, conventional cooling crystallization with operating conditions of agitation rate and cooling rate resulted in bigger and broader crystal size distribution. However, reduced Meta stable zone width (MSZW), lower crystal size and lower Span were observed in case of continuously sonicated crystallization. Application of ultrasound in MSZW at low power of ultrasound and lower cooling rate leads to creation of lower nucleation rate and bigger crystals with narrow distribution. The application of ultrasound prior to nucleation leads to higher nucleation rate (which was observed from FBRM count rate) and smaller crystals with a more narrow distribution. Careful supersaturation control after the nucleation step can lead to the formation of uniform product crystals. The only drawback of application of ultrasound in crystallization process is that the heat liberation due to high power ultrasound, which will be difficult to control by cooling systems. The temperature rise because of liberation of heat of crystallization will also lead to partial dissolution of nuclei which also helps in serve as a feedback control system of the nucleation. Total supersaturation was consumed due to ultrasound induced nucleation and concentration followed the solubility curve, where applied intermittent sonication serve as heating cycles rather than creating more nucleation. For optimization, the influence of time and power of ultrasound, and cooling profile were investigated. Crystal size distribution and span were compared between different control strategies including uninsonated, continuous sonicated, sonicated prior to nucleation and sonicated in MSZW followed by intermittent sonication. Additionally continuous crystallization with ultrasound in a two reactor system of Sono-Nucleator (50ml) and subsequent growth of crystals in a 3liters of reactor was studied.
The results of this research provides better understanding of the sonocrystallization process because of the inclusion of in-situ monitoring systems along with ultrasonic probe in a batch crystallizer. An insight is obtained about how to choose process trajectories to produce smaller crystals with narrow CSD and bigger crystals with narrow CSD, and about implementation in a larger scale batch and continuous crystallization.
1. S. Kim, B. Lotz, M. Lindrud, K. Girard, T. Moore, K. Nagarajan, M. Alvarez, T. Lee, F. Nikfar, M. Davidovich, S. Srivastava and S. Kiang, Organic Process Research Development9 (2005), 894â??901.
2. Luque de Castro, M. D.; Priego-Capote, F. Ultrasonics sonochemistry 14 (2007), 717â??24
3. Jieqiong Li,Ying Bao, Jingkang Wang, Chemical engineering technology 36 (2013), 1341â??1346
This work has been supported by Science Foundation Ireland, Grant number: 12/RC/2275.