Chemometric Quantitation of the Active Substance (Containing C≡N) in a Pharmaceutical Tablet Using Near-Infrared (NIR) Transmittance and NIR FT-Raman Spectra

Abstract

In this study, near-infrared (NIR) transmittance and Raman spectroscopy chemometric calibrations of the active substance content of a pharmaceutical tablet were developed using partial least-squares regression (PLS). Although the active substance contained the strongly Raman active C≡N functional group, the best results were obtained with NIR transmittance, which highlights the difference between (microscopic) surface sampling and whole tablet diffuse transmittance sampling. The tablets exist in four dosages with only two different concentrations of active substance (5 mg (5.6% w/w), and 10, 15, and 20 mg (8.0% w/w) active substance per tablet). A calibration on all four dosages resulted in a prediction error expressed as the root mean squared error of cross-validation (RMSECV) of 0.30% w/w for the NIR transmittance calibration. The corresponding error when using Raman spectra was 0.56% w/w. Specially prepared calibration batches covering the range 85–115% of the nominal content for each dosage were added to the first sample set, and NIR transmittance calibrations on this set—containing coated as well as uncoated tablets—gave a further reduction in prediction errors to 0.21–0.289% w/w. This corresponds to relative prediction errors (RMSECV/ynom) of 2.6–3.7%. This is a reasonably low error when compared to the error of the chromatographic reference method, which was estimated to 3.5%.

Bayesian variable selection for Gaussian process regression: Application to chemometric calibration of spectrometers

“…4 is that those variables close to wavenumber 8830 cm −1 have a relatively high probability of being selected. According to the fundamental spectroscopic analysis, in the NIR spectrum for the tablets studied, this is the only visually identifiable band, and is believed to be informative with regard to the prediction of the active substance in the tablets [40]. This confirms that the proposed strategies are capable of selecting informative covariates for the calibration model.…”

“…The transmittance spectra were generated from the analysis of Escitalopram R tablets, manufactured by a pharmaceutical company. The aim of the study was to determine the weight percentage of the active substance in the tablets, based on the NIR spectra over the wavenumber range of 7400 − 10500 cm −1 (corresponding to 404 covariates) [40]. The tablet samples corresponding to the "preliminary calibration set" defined in the original paper [40] were used, where the data set comprised 120 samples.…”

“…In the current study, a subjective prior distribution is adopted which ensures that the MCMC algorithms converge to the region where approximately fewer than 10 covariates are selected. This reflects the belief that a priori the number of the most relevant variables to the prediction should not significantly exceed 10, because the spectrum appears to have only one "visually" identifiable band (one band can contain several variables) that is affected by the active substance [40]. Using this criterion, λ was selected and resulted in reasonable acceptance rate of approximately 0.25.…”

Variation in the Transmission Near-Infrared Signal with Depth in Turbid Media

“…3,6,8 In addition, calibration models built using transmission mode data were shown to be more robust than those built using reflectance data when there was considerable variability in the chemical composition and physical properties of the tablets. 5 Spectra measured in reflectance mode were also found to be dominated by tablet coatings 10,11 and were more sensitive to tablet engraving or embossing. 4,5 Transmission measurements have been shown to be less affected by sample inhomogeneity.…”

The impact of temperature variations on spectroscopic calibration modelling: a comparative study

“…As powerful analytical tools, spectroscopic techniques in combination with calibration models have seen increasing implementation in sectors as diverse as food, pharmaceuticals and petrochemical [1][2][3]. However, as a consequence of the large number of spectral wavelengths, multivariate calibration methods, such as partial least squares (PLS) [4] [5] and other approaches including neural networks [1] [6] and Gaussian process models [7], are required for the development of robust calibration models.…”