Real-time Raman spectroscopic monitoring of solvent-mediated polymorphic transformations

Yuerong Hu, Purdue University

Abstract

Crystallization processes have been commonly used in the pharmaceutical industry to purify and generate APIs. Although crystallization processes have been widely used, unless the parameters are well defined, various crystallization conditions can produce undesirable solid state forms including polymorphs, solvates, amorphous materials, etc. Raman spectroscopy has greatly improved as a technique over the past twenty years; it has been applied to many different areas of analysis, including monitoring of crystallization processes. Although the use of Raman spectroscopy to monitor crystallization has already been reported in the literature, no study has shown that it can quantitatively determine all the pertinent information from the solute and different polymorphic forms. In the first part of this thesis, it is demonstrated that Raman spectroscopy can be successfully applied to real-time crystallization monitoring, extracting information about both polymorphic transformation processes while simultaneously providing information about the solution concentration. Using the mass balance principle, the amount of the polymorphic form in the solution can also be determined. Although it is well known that particle size can affect Raman signal intensity, the results of published studies and predictions from theoretical principles contradict each other. In addition, there have not been any studies of the influence of particle size on Raman signal intensity for mixtures of polymorphic forms. In this work, the effect of particle size on polymorphic quantitation was probed using Raman spectrometers with different laser beam geometries and it was demonstrated that with a relatively large beam diameter, effects of particle size on the calibration model could be largely eliminated. In the last part of the thesis, it was demonstrated that, through monitoring polymorphic transformations in real-time, quantitative transformation profiles could be generated and correlated to the corresponding temperature. From the aforementioned transformation profiles, thermodynamic and kinetic information about the enantiotropic polymorphic pair could be extracted. For the model compound utilized, flufenamic acid, the thermodynamic transition temperature that was determined from the Raman data was in good agreement with the value reported in the literature.

Degree

Ph.D.

Advisors

Byrn, Purdue University.

Subject Area

Pharmacology|Pharmaceuticals

COinS