Conference Contribution Details
Mandatory Fields
A. G. Ryder, B. Li, A. Calvet, C. Morris, and Y. Casamayou-Boucau.
XV Chemometrics in Analytical Chemistry,
Low content quantification in solid powder matrix using Raman spectroscopy and chemometric methods.
Changsha, Hunan province, China
Keynote Address
Optional Fields
Many active pharmaceutical ingredients (APIs) are solids, which are isolated and purified prior to incorporation into a dosage forms like tablets.  The certification of the purity and quality of solid APIs and excipients is critical in drug manufacture and the level of chemical contamination must, by necessity, be low and precisely specified.  These chemical impurities are in general either solvents, unwanted side-products of the synthesis step, degradation products, unwanted polymorphs, solvates, or hydrates, or foreign material introduced during handling.  For each specific API, a well-defined series of tests for specific contaminants will have been generated as part of the drug licensing process.  The implementation of these tests can involve considerable time and financial costs to the manufacturer. Raman imaging/mapping spectroscopy has been attractively applied for qualitative solid-state analysis in terms of contaminant detection and API/excipient distribution, because the sharper and well-defined Raman spectra often enable easier discrimination in mixtures.[1-3]  The present study developed a robust and accurate analytical methodology for quantifying low-content (<0.25%) components in solid matrices using Raman mapping spectroscopy.[4]  The approach was to employ sub-sampling data collection to generate a large array of heterogeneous Raman spectra, which were, then statistically analysed using chemometric methods to generate a robust quantitative measurement.  Several discrete issues had to be had to be addressed: (1) hydration interference, (2) baseline effects, (3) cosmic ray artefacts, (4) informative variable selection, (5) appropriate data pre-treatment, (6) accurate quantification of low-content analyte, and (7) method validation.  Several discrete high concentration models, by means of partial least-squares (PLS) regression,[5] were used to predict the local concentration of analyte at each pixel.  The combined local concentrations were then statistically analysed to generate the true sample concentration, and piracetam concentration in proline model system could be predicted with a relative accuracy of ~2.4% over the 0.05~1.0% concentration range.  This methodology was competitive with high performance liquid chromatography (HPLC) in terms of limit of detection (0.03% versus 0.041%) and accuracy.  This analytical approach is suitable for high-throughput screening and real-time-release applications, and for quantification of hydrate, polymorph, and solvate contamination, some of which cannot be easily measured using traditional HPLC based methods.   Acknowledgement: This work was undertaken as part of the Synthesis and Solid State Pharmaceutical Centre funded by Science Foundation Ireland and industry partners, and Enterprise Ireland (Grant No:  TC-2012-5106).  We thank Kaiser Optical Systems, Inc. and Mr. H. Owen for the loan of the Raman instrumentation.  References: [1] G. Reich, Advanced Drug Delivery Reviews, 57, 1109 (2005). [2] A.A. Gowen, C.P. O'Donnell, P.J. Cullen, S.E. Bell, Eur J Pharm Biopharm, 69, 10 (2008). [3] S. Šašić, S. Mehrens, Anal. Chem., 84, 1019 (2012). [4] B. Li, A. Calvet, Y. Casamayou-Boucau, C. Morris, and A.G. Ryder,  In revision, Feb. 2015. [5] S. Wold, M. Sjostrom, L. Eriksson, Chemom. Intell. Lab. Syst., 58, 109 (2001).  
Synthesis and Solid State Pharmaceutical Centre funded by Science Foundation Ireland.
Publication Themes