Tablet uniformity has to be checked during formulation and manufacturing development as well as during ongoing production. The requirements of unit dose potency and stable dissolution profile require that the tablet has uniform distribution of its ingredients. A plurality of tablets can be scanned to establish the uniform pattern of the API (active pharmaceutical ingredient) and excipients. Automated SWIR hyperspectral measurement is ingredient-specific and obtains high resolution images of each tablet for comparison.
Pharmaceutical Research and Quality Control with Hyperspectral Imaging
Chemical imaging is the application of hyperspectral imaging to identify and quantify chemical components of a sample or product and its dispersion or homogeneity. Chemical imaging typically uses the near-infrared (NIR) or short-wave (SWIR) infrared ranges, which contain information about chemical bonds, in contrast to hyperspectral imaging, which uses any wavelength range, from visible to long-wave infrared. Organic chemicals that comprise most pharmaceutical products have unique spectra in the NIR and SWIR ranges. Spectral information allows identification and quantification of chemical components within a sample. Chemical imaging is used in many areas of pharmaceutical research and for quality control in pharmaceutical industry, including blending, tablet production monitoring, and counterfeit product identification. Application examples are described below.
Chemical imaging is used in the pharmaceutical industry for quality control and process validation. FDA regulations require control of the drug content at various stages of the manufacturing process, including granulation and final blending. Statistical sampling for content uniformity assessment is straightforward, but in practice, there is a potential for sampling bias when only a small sample volume is extracted (Berman, 2005). Pharmaceutical companies are discovering that chemical imaging is an effective and efficient way to detect problems in blending and tableting (El-Hagrasy, 2001, Lyon, 2002).
For several years, blending has been monitored using the near-infrared region to non-invasively determine the concentration of an active ingredient in a sample (Ma, 2007). The challenge is that by taking one single-point measurement at a time (for example during one rotation of the blender), the distributions of the ingredients are not revealed. The image (below left) illustrates ideal blending. Multiple components can be predicted from the same hypercube of data. The images (below right) are calculated concentration maps of three different, poorly mixed components. They demonstrate the rich detail available from hyperspectral imaging.
One hyperspectral image datacube can detect areas of poor blending, the location of high concentration pockets of particular constituents, as well as the progression of the blending. The predicted images for the first few turns of an experimental blend, taken after each rotation of the blender, are shown below (Kemeny, 2010).
Pharmaceutical tablets can be monitored on the production line with hyperspectral imaging. To establish the possible optical coverage at real production speeds, the relationship between the speed of tablets measured and the resolution of the image is plotted below in a SWIR camera example (320 pixels per sample width). For this calculation, it is assumed that the tablets are of uniform diameter and arranged in a square pattern. The hyperspectral camera operates with the push-broom method, collecting one line of sample data in each frame. At 100 frames per second, for example, it is possible to image 200 tablets per second with a spatial resolution of 125 pixels per tablet. An approximation of the image points per tablet is shown below. With fewer measurement points on each tablet, the number of tablets covered can be increased even further. As a screening tool, uncoated tablets can be measured to detect uneven distribution of ingredients, whereas coated tablets can be measured to check for potentially uncoated or unevenly coated areas.
Near-infrared hyperspectral imaging is a fast, nondestructive tool which can provide quantitative analyses of pharmaceutical transdermal films on the manufacturing line. From our article in Transdermal Magazine,
A complete measurement system consists of an illumination module, a hyperspectral camera with its controller module, power supplies, and a data collection and processing system. [The images below] show the two basic optical arrangements.
When choosing placement and arrangement for measurement, the manufacturer must consider the spectral characteristics of the transdermal product. For example, the spectrum of a 0.075 mg-per-day estradiol patch shows approximately 0.4 to 0.5 peak-to-valley absorbance in both transmission and reflectance,which is a reasonable intensity for quantitative work in either mode.
In contrast, other products may limit measurement to either transmission or reflectance only. For transdermal products in general, the transmission-measurement mode is preferable because the light usually passes through the entire cross section of the product. It is generally a good practice to choose a measurement point physically closest to the location where the manufacturing process deposits the API-containing-material and where the process has not yet applied the additional layers.
Manufacturers can monitor transdermal and other roll-to-roll production using high-speed, hyperspectral instrumentation if the analytical task lends itself to spectroscopic monitoring using reflectance or transmission, near-infrared measurements. Pull samples in manufacturing are disruptive and costly, and manufacturers can replace them with process monitoring that constitutes 100% inspection using a hyperspectral camera system. Themanufacturer can perform feasibility measurements in the lab and then can move the hyperspectral device to the production floor and integrate it into the process monitoring and process-control system.
CFR 21 Part 11 Compliant Online Software
For more details on the software for the real-time monitoring of film thickness and composition please visit the Web Monitor page.
Identifying counterfeit products is extremely important in the pharmaceutical industry, as the number of counterfeit drug products has grown exponentially in recent years. Counterfeit drugs have serious potential health risks to the consumer, and can also jeopardize the reputations of pharmaceutical companies (Puchert, 2009). Hyperspectral imaging combined with chemometric data analysis is a useful method for identifying off-specification products. Hyperspectral cameras can be placed over a moving line of products to efficiently scan and report counterfeits in real-time, or products can be measured individually off-line. Chemical imaging provides several advantages over older methods of identifying counterfeit pharmaceuticals. In contrast to chromatographic assays and dissolution testing, chemical imaging is fast and non-destructive (Dubois, 2007). Researchers at the Institute of Pharmacy and Molecular Biotechnology in Germany demonstrated that near-infrared chemical imaging is an innovative, effective and non-destructive way to detect counterfeit tablets by discovering differences between amount and spatial distribution of ingredients within a genuine or counterfeit tablet (Puchert, 2010).
- Berman, J., & Planchard, J .A. (2005). Blend uniformity and unit dose sampling. Drug Development and Industrial Pharmacy, 21 (11), 1257-1283.
- Dubois, J., et al. (2007). NIR Chemical Imaging for counterfeit pharmaceutical products analysis. Spectroscopy, 22 (2), 40.
- El-Hagrasy, et al. (2001). Near-Infrared spectroscopy and imaging for the monitoring of powder blend homogeneity, Journal of Pharmaceutical Sciences, 90 (9), 1298-1307.
- Kemeny, G., Stuessy, G., & Crothers, N. (2010). Pharmaceutical blend homogeneity. AAPS, New Orleans.
- Lyon, R. C., et al. (2002). Near-infrared spectral imaging for quality assurance of pharmaceutical products. AAPS PharmSciTech 3 (3),1-15.
- Ma, H., & Anderson, C. A. (2007). Optimisation of magnification levels for near infrared chemical imaging of blending of pharmaceutical powders. Journal of Near-Infrared Spectroscopy, 15 (3), 137–151.
- Puchert, T., et al. (2010) Near-infrared chemical imaging (NIR-CI) for counterfeit drug identification – A four-stage concept with a novel approach of data processing (Linear Image Signature). Journal of Pharmaceutical and Biomedical Analysis, 51 (2010), 138-145.