Resolution of Organic Polymorphic Crystals by Raman Spectroscopy

Abstract

Depending on crystallization conditions, many organic compounds can form crystals of different structure. Their proper characterization is important, for example, in the pharmaceutical industry. While the X-ray diffractometry established as a standard method, alternative techniques are desirable for broader application flexibility and economic reasons. In the present study, Raman spectroscopy combined with the density functional calculations is suggested as a complementary method to the X-ray and other higher resolution techniques. The potential to discriminate structural differences in polymorphic crystalline forms is documented on three model compounds of industrial importance. Methacrylamide, piracetam, and 2-thiobarbituric acid were crystallized under various conditions, and their Raman spectra were recorded using 532 and 1064 nm laser excitations. X-ray diffractometry and nuclear magnetic resonance spectroscopy were used as complementary techniques to verify sample composition and structure. To interpret the observed differences in Raman frequencies and intensities, three computational strategies were explored based on single molecule, a cluster model, and a plane-wave periodic boundary conditions calculation. The single-molecule modeling was found inadequate, whereas the plane-wave approach provides the most realistic spectra. For all compounds, the differences in the Raman spectra of polymorphic forms could be unambiguously assigned to the simulations. The modeling revealed that the spectral differences were caused by the molecular structure itself as well as by crystal packing. The relative importance of these factors significantly varied across the investigated samples. Owing to its simplicity, Raman spectroscopy appears to be a promising technique capable of reliable discriminating between organic crystal polymorphic states.

Effects of Quantum Nuclear Delocalisation on NMR Parameters from Path Integral Molecular Dynamics

“…In contrast, similar comparison of calculated and experimental infrared or Raman spectra of organic crystals is much less satisfactory and the use of vibrational spectroscopy for crystal structure determination is, therefore, limited. [37] The PIMD approach enables predictions of isotope effects in larger systems such as organic molecular crystals in excellent agreement with experiment. Convergence errors, which would otherwise dominate the calculated results, can be avoided by convoluting probability distributions of selected geometrical parameters with the calculated shielding or J coupling surface scans.…”

Substrate‐Induced Phase of a Benzothiophene Derivative Detected by Mid‐Infrared and Lattice Phonon Raman Spectroscopy

“…Raman spectroscopy in the low‐wavenumber regime probes lattice phonons instead, that is, those modes depending on intermolecular interactions, and therefore, is sensitive to even slight modifications of the molecular packing motif. Thus, each crystal structure displays typical spectral features in this energy range and should be easily distinguishable . A comparison of experimentally obtained Raman spectra with state‐of‐the‐art DFT calculations, including van der Waals interactions, recently published confirms the expected crystal structures …”

Vibrational averaging of the chemical shift in crystalline α‐glycine

“…As observed before for clusters of a solute and solvent molecules,[24] the temperature variations of the geometry cause a relatively large dispersion of the shielding values, and many clusters are required for a converged average. In our case (33 geometries), we could also average over the four glycine molecules in the elementary cell, which resulted in 132 quasi‐independent values and provided a reasonably converged average shielding (Supporting Information, Fig.…”

“…[16, 17] They are also increasingly popular for the solid state[18–23] and solutions. [24, 25] Several modeling and simulation techniques have been proposed to describe the influence of the intermolecular interactions on chemical shifts. In the cluster model,[26] neighboring molecules or fragments are considered explicitly.…”