Transmission fourier transform Raman spectroscopy of pharmaceutical tablet cores

Transmission Fourier transform (FT) Raman spectroscopy of pharmaceutical tablet cores is demonstrated using traditional, unmodified commercial instrumentation. The benefits of improved precision over backscattering Raman spectroscopy due to increased sample volume are demonstrated. Self-absorption effects on analyte band ratios and sample probe volume are apparent, however. A survey of near-infrared (NIR) absorption spectra in the FT-Raman spectral range (approximately 0 to 3500 wavenumber shift from 1064 nm, or 1064 to 1700 nm) of molecules with a wide range of NIR-active functional groups shows that although absorption at the laser wavelength (1064 nm) is relatively small, some regions of the Raman spectrum coincide with NIR absorbances of 0.5 per cm or greater. Fortunately, the pharmaceutically important regions of the Raman shift spectrum from 0 to 600 cm(-1) and from 1400 to 1900 cm(-1) exhibit low self-absorption for most organic materials. A statistical analysis of transmission FT-Raman noise in spectra collected from different regions of a pharmaceutical tablet provides insight into both spectral distortion and reduced sampling volume caused by self-absorption.     

An in-depth analysis of Raman and near-infrared chemical images of common pharmaceutical tablets

This study reports on the application of Raman and near-infrared (NIR) imaging techniques for determining the spatial distribution of all (five) components in a common type of pharmaceutical tablet manufactured in two different ways. Multivariate chemical images were produced as principal component (PC) scores, while univariate images were produced by using the most unique spectra selected by the orthogonal projection approach (OPA), a searching algorithm. Multivariate Raman images were obtained for all five components in both tablets, while only two or three components could be imaged with the NIR instrument. Very interesting PC results are reported that in effect cast doubt on the effectiveness of the established criteria for determining signal-related PCs in the Raman data. PCA has been found to be indispensable for imaging the minor components using the Raman data. Significant similarity between the multivariate and univariate chemical images has been noted despite there being considerable spectral overlap within the Raman and, especially, within the NIR mapping data sets. Gray-scale images are carefully thresholded, which allowed for quantitative comparison of the obtained binarized images. A thorough discussion is given on the problems and approximations needed for producing composite images.     

Bulk Raman analysis of pharmaceutical tablets

We compare and contrast two Raman collection geometries, backscattering and transmission, to identify their potential for monitoring the bulk chemical composition of turbid media. The experiments performed on pharmaceutical tablets confirm the expected strong bias of the backscattering Raman collection towards surface layers of the probed sample. However, this bias is largely absent with the transmission geometry, exhibiting gross insensitivity to the depth of impurities within the sample. The results are supported by Monte-Carlo simulations. The applicability of transmission geometry to tablets without any thinning is possible because of long migration times of Raman photons in non-absorbing powder media. The absolute measured intensity of the Raman signal was only 12 times lower in transmission geometry compared with backscattering geometry for a standard paracetamol tablet with a thickness of 3.9 mm. This makes detection relatively straightforward, and detectable Raman signals were observed even after propagation through three paracetamol tablets. Given its properties and instrumental simplicity, the transmission method is particularly well suited to the on-line analysis of bulk content of tablets in pharmaceutical applications.     

Multivariate statistical analysis of concatenated time-of-flight secondary ion mass spectrometry spectral images. Complete description of the sample with one analysis

Time-of-flight secondary ion mass spectrometry (TOF-SIMS) instruments are capable of saving an entire mass spectrum at each pixel of an image, allowing for retrospective analysis of masses that were not selected for analysis during data collection. These TOF-SIMS spectral images contain a wealth of information, but few tools are available to assist the analyst in visualizing the entire raw data set and as a result, most of the data are not analyzed. Automated, nonbiased, multivariate statistical analysis (MVSA) techniques are useful for converting the massive amount of data into a smaller number of chemical components (spectra and images) that are needed to fully describe the TOF-SIMS measurement. Many samples require two back-to-back TOF-SIMS measurements in order to fully characterize the sample, one measurement of the fraction of positively charged secondary ions (positive ion fraction) and one measurement of the fraction of negatively charged secondary ions (negative ion fraction). Each measurement then needs to be individually evaluated. In this paper, we report the first MVSA analysis of a concatenated TOF-SIMS data set comprising positive ion and negative ion spectral images collected on the same region of a sample. MVSA of concatenated data sets provides results that are intuitive and fully describe the sample. The analytical insight provided by MVSA of the concatenated data set was not obtained when either polarity data set was analyzed separately.     

Multivariate statistical analysis of three-spatial-dimension TOF-SIMS raw data sets

Three-spatial-dimension (3D) time-of-flight-secondary ion mass spectrometry (TOF-SIMS) analysis can be performed if an X-Y image is saved at each depth of a depth profile. In this paper, we will show how images reconstructed from specified depths, depth profiles generated from specific X-Y coordinates, as well as three-spatial-dimensional rendering provide for a better understanding of the sample than traditional depth profiling where only a single spectrum is collected at each depth. We will also demonstrate, for the first time, that multivariate statistical analysis (MVSA) tools can be used to perform a rapid, unbiased analysis of the entire 3D data set. In the example shown here, retrospective analysis and MVSA revealed a more complete picture of the 3D chemical distribution of the sample than did the as-measured depth profiling alone. Color overlays of the MVSA components as well as animated movies allowing for visualization (in 3D) from various angles will be provided.     

Red-excitation dispersive Raman spectroscopy is a suitable technique for solid-state analysis of respirable pharmaceutical powders

Dispersive Raman spectroscopy with excitation by a red diode laser is suitable for quantitative crystallinity measurements in powders for pulmonary drug delivery. In spray-dried mixtures of salmon calcitonin and mannitol, all three crystalline polymorphs of mannitol and amorphous mannitol were unambiguously identified and their mass fractions were measured with a limit of quantification of about 5%. The instrument design offered high sensitivity and adequate background suppression, resulting in a low limit of detection in the range of 0.01% to 1%. This spectroscopy method has significant advantages over established techniques regarding specificity, sensitivity, and sample requirements.     

Multivariate statistical process control with artificial contrasts

A multivariate control region can be considered to be a pattern that represents the normal operating conditions of a process. Reference data can then be generated and used to learn the difference between this region and random noise. Then multivariate statistical process control can be converted to a supervised learning task. This can dramatically reshape the control region and open the control problem to a rich collection of supervised learning tools. Such tools provide generalization error estimates that can be used to specify error rates. The effectiveness of such an approach is shown here. Such a computational approach is now easily accomplished with modern computing resources. Examples use random forests and a regularized least squares classifier as the learners.     

Multivariate statistical process control for autocorrelated processes

Multivariate statistical process control is often used in chemical and process industries where autocorrelation is most prevalent. We present a realistic model that generates autocorrelation and crosscorrelation and provides a useful approach to characterizing process data. We show how our model relates to the widely-used method of principal component analysis, distinguish between types of assignable causes, and present a useful control statistic based on a principal component decomposition that is not autocorrelated. The control chart for this statistic can be developed by a routine analysis even when the input data is autocorrelated. Furthermore, to characterize our results, we show that any linear combination of the input data that is not autocorrelated is related to our control statistic.     

Multivariate methods in the development of a new tablet formulation: optimization and validation

In a previous study of the development of a tablet formulation approximately 100 excipients were characterized in screening experiments using multivariate design. Acceptable values for important responses were obtained with some of the formulations. The relationships between the properties of the excipients and the responses were evaluated using PLS. In this study additional experiments were performed in order to validate models obtained from the screening study and to find a formulation of suitable composition with desired tablet properties. A formulation with the desired disintegration time was found with the additional experiments and the agreement between observed and predicted values was fair for the tablets that did disintegrate. A limitation of this study was that tablets from four experiments did not disintegrate within the set time limit. The lack of agreement between observed and predicted values of these four experiments was probably due to the nature of one of the factors in the design. Considering the reduced experimental design the results are still encouraging.     

Multivariate hyperspectral Raman imaging using compressive detection

A multivariate hyperspectral imaging (MHI) instrument has been designed and constructed to achieve greatly increased Raman imaging speeds by utilizing a compressive spectral detection strategy. The instrument may be viewed as a generalized spectrometer, which can function either as a conventional monochromator or in a wide variety of other hyperspectral modalities. The MHI utilizes a spatial light modulator (SLM) to produce programmable optical filters to rapidly detect and map particular sample components. A sequence of Hadamard-transform or random filter functions may be used to regenerate full Raman spectra. Compressive detection is achieved either using multivariate signal processing filter functions or the actual component spectra. Compressive detection is shown to be capable of achieving sampling speeds exceeding 1 ms per image pixel and the collection of chemical images in less than a minute.     

Multivariate methods in developing an evolutionary strategy for tablet formulation

The aim of this study was to develop a new strategy for choosing excipients in tablet formulation. Multivariate techniques such as principal component analysis (PCA) and experimental design were combined in a multivariate design for screening experiments. Of a total 87 investigated excipients, the initial screening experiments contained 5 lubricants, 9 binders, and 5 disintegrants, and 35 experiments were carried out. Considering a reduced factorial design was used, the resulting PCA and partial least squares (PLS) models offered good insight into the possibilities of tablet formulation. It also offered solutions to the problems and clearly gave directions for optimum formulations. Further, it offered several alternatives for achieving quality formulations. Additional experiments conducted to validate and verify the usefulness of the model were successful, resulting in several tablets of good quality. The conclusion is that a multivariate strategy in tablet formulation is efficient and can be used to reduce the number of experiments drastically. Combining multivariate characterization, physicochemical properties, experimental design, multivariate design, and PLS would lead to an evolutionary strategy for tablet formulation. Since it includes a learning strategy that continuously incorporates data for new compounds and from conducted experiments, this would be an even more powerful tool than expert systems.     

Multivariate statistical process control charts: an overview

In this paper we discuss the basic procedures for the implementation of multivariate statistical process control via control charting. Furthermore, we review multivariate extensions for all kinds of univariate control charts, such as multivariate Shewhart‐type control charts, multivariate CUSUM control charts and multivariate EWMA control charts. In addition, we review unique procedures for the construction of multivariate control charts, based on multivariate statistical techniques such as principal components analysis (PCA) and partial least squares (PLS). Finally, we describe the most significant methods for the interpretation of an out‐of‐control signal. Copyright © 2006 John Wiley & Sons, Ltd.     

Thermal images generated by a line-scanning technique: statistical properties

The analysis of statistical parameters of computer-generated therfial images by a line-scanning technique indicates an approach to efficient image transmission.     

Multivariate statistical analysis of the bibliographic output from a research institution,in relation to the measures of scientific policy

The publications produced in a medical research institute in a 16 year interval were classified into five categories (scientific papers in the journals covered byCurrent Contents orScience Citation Index, scientific papers in other journals, books and monographs, technical papers, congress and symposia communications) and counted for each year separately. The number of researchers and yearly budgets were also recorded. The data were analysed by contingency table, correlation and factor-analytical methods. It was shown that, upon introducing quantitative minimal criteria for job promotions, the proportion of scientific papers increased. Principal component analysis indicated that the data can be approximately represented as linear combinations of three mutually independent factors. The approach used is recommended for evaluating the production of scientific information in research institutions and for assessing the effects of the measures of scientific policy.     

Multivariate methods in the development of a new tablet formulation: excipient mixtures and principal properties

A tablet formulation for direct compression has previously been studied using multivariate design. An optimization study of one of the most important tablet properties, disintegration time, revealed that excipients with Principal Properties (PP's) that were predicted as suitable by the model were not represented within the studied material. The feasibility of using mixtures of excipients in the multivariate approach to tablet formulation to solve this problem has been investigated in the present study. By mixing different excipients of the same excipient class, it should be possible to obtain mixtures with the predicted PP's, which in turn should give a formulation with the desired properties. In order to investigate the utility of this approach, separate mixture designs were applied to both binders and fillers (diluents). As reported here, the Partial Least Squares Projections to Latent Structures (PLS) model developed in the previously published screening study has been validated in the sense that the interesting region of the PP space identified in it has been shown to contain excipients, pure or mixed, that give the formulation suitable properties. Formulations with suitable properties were found with the mixture experiments. The local models also offer several alternatives for the composition of the formulation that yield the desired disintegration time.     

Quantitative transmission Raman spectroscopy of pharmaceutical tablets and capsules

Quantitative analysis of pharmaceutical formulations using the new approach of transmission Raman spectroscopy has been investigated. For comparison, measurements were also made in conventional backscatter mode. The experimental setup consisted of a Raman probe-based spectrometer with 785 nm excitation for measurements in backscatter mode. In transmission mode the same system was used to detect the Raman scattered light, while an external diode laser of the same type was used as excitation source. Quantitative partial least squares models were developed for both measurement modes. The results for tablets show that the prediction error for an independent test set was lower for the transmission measurements with a relative root mean square error of about 2.2% as compared with 2.9% for the backscatter mode. Furthermore, the models were simpler in the transmission case, for which only a single partial least squares (PLS) component was required to explain the variation. The main reason for the improvement using the transmission mode is a more representative sampling of the tablets compared with the backscatter mode. Capsules containing mixtures of pharmaceutical powders were also assessed by transmission only. The quantitative results for the capsules' contents were good, with a prediction error of 3.6% w/w for an independent test set. The advantage of transmission Raman over backscatter Raman spectroscopy has been demonstrated for quantitative analysis of pharmaceutical formulations, and the prospects for reliable, lean calibrations for pharmaceutical analysis is discussed.     

Bhattacharyya distance as a contrast parameter for statistical processing of noisy optical images

In many imaging applications, the measured optical images are perturbed by strong fluctuations or boise. This can be the case, for example, for coherent-active or low-flux imagery. In such cases, the noise is not Gaussian additive and the definition of a contrast parameter between two regions in the image is not always a straightforward task. We show that for noncorrelated noise, the Bhattacharyya distance can be an efficient candidate for contrast definition when one uses statistical algorithms for detection, location, or segmentation. We demonstrate with numerical simulations that different images with the same Bhattacharyya distance lead to equivalent values of the performance criterion for a large number of probability laws. The Bhattacharyya distance can thus be used to compare different noisy situations and to simplify the analysis and the specification of optical imaging systems.     

Analyzing Raman maps of pharmaceutical products by sample-sample two-dimensional correlation

Sample-sample (SS) two-dimensional (2D) correlation spectroscopy is applied in this study as a spectral selection tool to produce chemical images of real-world pharmaceutical samples consisting of two, three, and four components. The most unique spectra in a Raman mapping spectral matrix are found after analysis of the covariance matrix. (This is obtained by multiplying the original mapping data matrix by itself.) These spectra are identified by analyzing the slices of the covariance matrix at the positions where covariance values are at maxima. Chemical images are subsequently produced in a univariate fashion by visually selecting the wavenumbers in the extracted spectra that are least overlapped. The performance of SS 2D correlation is compared with principal component analysis in terms of highlighting the most prominent spectral differences across the whole data set (which typically comprises several thousand spectra) and determining the total number of species present. In addition, the selection of the unique spectra by SS 2D correlation is compared with the selection obtained by the orthogonal projection approach (OPA). Both comparisons are found to be satisfactory and demonstrate that a quite simple SS 2D correlation routine can be used for producing reliable images of unknown samples. The main benefit of using SS 2D correlation is that it is based on a few data processing commands that can be executed separately and produce results that are closely related to the chemical features of the system.     

Multivariate analysis applied to the study of spatial distributions found in drug-eluting stent coatings by confocal Raman microscopy

Multivariate data analysis was applied to confocal Raman measurements on stents coated with the polymers and drug used in the CYPHER Sirolimus-eluting Coronary Stents. Partial least-squares (PLS) regression was used to establish three independent calibration curves for the coating constituents: sirolimus, poly(n-butyl methacrylate) [PBMA], and poly(ethylene-co-vinyl acetate) [PEVA]. The PLS calibrations were based on average spectra generated from each spatial location profiled. The PLS models were tested on six unknown stent samples to assess accuracy and precision. The wt % difference between PLS predictions and laboratory assay values for sirolimus was less than 1 wt % for the composite of the six unknowns, while the polymer models were estimated to be less than 0.5 wt % difference for the combined samples. The linearity and specificity of the three PLS models were also demonstrated with the three PLS models. In contrast to earlier univariate models, the PLS models achieved mass balance with better accuracy. This analysis was extended to evaluate the spatial distribution of the three constituents. Quantitative bitmap images of drug-eluting stent coatings are presented for the first time to assess the local distribution of components.