Anna Novikova, PhD, is a process analytical technology scientist at Fette Compacting, anovikova@fette-compacting.com.

Anna Novikova

Oral solid-dosage forms are commonly examined by time-consuming, off-line tests of only randomly chosen samples. This article discusses an alternative, fully integrated, at-line spectroscopic quality control approach for oral solid dosage forms and presents the findings of a case study in this field.

	Process analytical technology (PAT) tools allow process monitoring and process control of the critical parameters during production. Furthermore, in many cases, PAT even provides a better process understanding. 

	PAT tools are mainly integrated as in-line, on-line, or at-line solutions. In-line and on-line techniques perform frequent measurements directly in the product stream, allowing fast reaction and even process control. The difference between these two solutions is how the PAT tool is integrated into the process; on-line uses a sampling loop while in-line refers to measurements directly in the process stream. These solutions are the methods of choice for detecting short process disturbances. Nevertheless, integration into the process and, in particular, the requirement to find a representative measurement position without disturbing the process, might be challenging. 

	For at-line PAT tools, the samples are commonly withdrawn out of the process, and the analysis is performed on the defined number of samples in the production environment. Several tests can be performed on exactly the same sample during the manufacturing process, such as measurement of weight and content needed for a content uniformity assay. Therefore, such a system is a good start for the definition of the real-time release (RTR) strategy. Furthermore, as the analysis is performed directly after the manufacturing step (not off-line in the lab), it enables the manufacturer to react immediately to process variations and reduce waste by potentially stopping the process and recapturing the powder.

	Spectroscopic techniques, such as near infrared (NIR) spectroscopy, are non-invasive, non-destructive, and allow fast quality control investigations on higher amounts of samples during the complete manufacturing process. Thus, NIR exhibits great potential as a PAT tool for control of the critical parameters during production and better process understanding. Furthermore, the same NIR measurement delivers information about chemical and physical attributes of the sample. Thus, it is applicable for measurement of API concentration, moisture, density of the sample, and so on. 

	Two major measurement approaches for NIR spectroscopy, transmission and reflection, are possible. Reflection is the method to choose on moving samples (e.g., powder, tablets) due to its ability to measure at high speed and its low sensitivity to changes in sample thicknesses. One limitation is that the penetration depth of the NIR light has to be taken into account, with a special focus on the ratio of measured volume to the whole sample volume. 

	For at-line measurement on a static sample, transmission is a better solution; because the light signal travels all the way through the sample (rather than partway through as in reflection NIR), the information within the sample is acquired. Additionally, measurement on static samples brings an advantage of longer possible measurement time, allowing the usage of slower NIR spectroscopy devices with higher resolution. 

	The combination of measured sample volume, non-moving sample, and higher resolution for transmission NIR allows measurement of lower API concentrations compared to reflection NIR. However, the thickness of the sample might be a limitation for transmission, because the light needs to propagate through the sample. Tablets less than 9 mm thick are usually accessible by transmission NIR.

	To benefit from all the advantages of an at-line spectroscopic PAT solution, the PAT tool should be integrated into the process to allow frequent measurement and control during the process. For example, in an at-line tablet tester, the tablets are automatically sampled after the tableting process and pass individual measurement positions (e.g., tablet weight measured by a scale and API content measured by NIR). Subsequently, these values are transferred back to the process control system. As mentioned, this position is a good choice for the NIR spectroscopy transmission measurement, although there are additional challenges for integration into the process. It is well known that an off-line NIR spectroscopy lab analysis in transmission is a precise and reliable technique, but the positioning has to be as precise as possible, because it influences the signal-to-noise ratio. Additionally, unwanted scatter light, which escapes from the sides of a sample (tablet), also has a high influence on the measurement precision and should be avoided. Therefore, new challenges for an automated at-line PAT solution are precise and repetitive positioning and avoiding scattered light when measuring during the process. 

	Two of the main goals of at-line analysis are to replace the final analysis in the lab (e.g., high-performance liquid chromatography [HPLC] for the API concentration and content uniformity) or the direct release of batches after production (i.e., RTR). At least four release tests (i.e., identity, assay, content uniformity, and even dissolution) performed by NIR in combination with a scale might be a part of the RTR strategy and thus replace these lab tests. The following case study performed by Bayer in collaboration with Kraemer Elektronik and Fette Compacting shows that this goal is achievable. 

	Bayer tested a technique using an automated tablet tester equipped with NIR spectroscopy on a formulation with a low API concentration. In this study, the suitability of an integrated at-line transmission NIR spectroscopy solution to predict the API concentration of the sample tablets with an accuracy comparable to HPLC was investigated.

Tablets were produced on a rotary tablet press (2200i, Fette Compacting) with 6-mm round concave punches. The compression force and compression speed were varied to verify the robustness of the model. Tablets with eight different API concentration ratios (in the range of 0–6% API concentration) were compressed. Thereafter, the tablets were randomly assigned to either a calibration or a test set. 

	Subsequently, the calibration set tablets were measured with NIR Flex N-500 Solids Transmittance (Büchi Labortechnik). The calibration model was calculated with NIRCal Software (Büchi Labortechnik) and then transferred to the NIR Checkmaster (Fette Compacting/Kraemer Elektronik). The NIR Checkmaster, shown in 

, is a fully automated tablet tester with a built-in NIR Flex N-500 Solids Transmittance spectrometer. 

	The tablets are withdrawn out of the process and guided by the sampling gate of the tablet press to the NIR Checkmaster. The NIR Checkmaster positions the tablets by the rotating star wheel in front of individual testing stations: weight, thickness, hardness, diameter, and NIR transmission measurement cell, where the measurements are performed. Thereafter, all measured values (including API content measured by NIR transmission) are transferred to the tablet press. The tablet press can take an action based on the results. Furthermore, these results are included into the batch record.  

	The test set tablets with four different API concentration ratios were analyzed with the NIR Checkmaster, which was operated by the tablet press software. All tablets were measured with HPLC to determine the reference API concentration. Results of the NIR Checkmaster and HPLC were then compared.

shows that the mean API HPLC values and the NIR Checkmaster values have comparable standard deviations for the four concentrations of the test set. The NIR Checkmaster values include manufacturing process variabilities, such as changes of the compression speed and the compression force. An exact positioning of the tablet using a patented format part tablet holder that always positions the tablet in the same way in front of the sensor was crucial for obtaining accurate results using the tablet tester. In addition, the holder avoids the occurrence of unwanted scatter light. 

	Based on the results that are shown in 

, the presented at-line spectroscopic solution exhibits a great potential for a robust RTR strategy, because the NIR Checkmaster is able to predict the same concentration values with a comparable standard deviation as HPLC. Additionally, the tablet tester is fully integrated into the process, enabling process control and documentation. This PAT solution already offers several benefits along the road to RTR: process control and reduction of waste batch production, gaining of process understanding and process knowledge, and finally reduction or elimination of the off-line lab analysis. 

	The author would like to thank Dr. Adrian Funke, Reinhard Gross, and Dr. Albert Tulke from Bayer AG for the performance of the study and provision of the results. 

	Anna Novikova, PhD, is a process analytical technology scientist at Fette Compacting, 

	When referring to this article, please cite it as A. Novikova, "The Road to Real-Time Release: At-line NIR Testing," 

Articles

At-line NIR measurements can replace a laboratory HPLC measurement for API content in oral solid-dosage drug manufacturing.