Optical Coherence Tomography
An Alternative Methodology to Monitor Pharmaceutical Coating Processes
- Fig. 1: OCT images of film-coated tablets. Left part shows OCT images of several commercially-available tablets (optical image dimensions (in air): 3.208 x 1.332 mm2). Right part shows OCT images of Thrombo ASS tablets obtained at four different relative speed differences between the OCT sensor head and the tablets (optical image dimensions (in air): lateral image size for 0.05 m/s: 3.44 mm, 0.3 m/s: 11.17 mm, 0.5 m/s: 18.19; 0.7 m/s: 25.31 mm; vertical image size corresponds to a depth of 2.5 mm).
- Fig. 2: Schematic of the thickness determination of film-coated pellets. The OCT sensor head was integrated in a fluid-bed coater to monitor the coating growth. Samples were drawn from the process to compare in-line and off-line OCT measurements. The results of both off-line and in-line measurements are illustrated depending on the elapsed process time (total process time of 32 minutes).
Optical coherence tomography (OCT) features a contact-free non-destructive technique, offering high resolution images of internal structures of materials. It represents an alternative process monitoring technology for coating processes, which does not depend on calibration models as are required for spectroscopic methods. Due to its high acquisition rate-OCT enables the measurement of local parameters of film-coated pellets or tablets even for fast moving samples and therefore, features a perfect in-line process monitoring technology.
OCT is an emerging, non-invasive optical imaging modality, which enables cross-sectional tomographic visualization of the internal microstructure of samples. Since its invention in the late 1980s and early 1990s, the first applications were related to the evaluation of biological tissue, i.e. the real-time, in vitro imaging of retinal structure as well as of an arterial wall with fatty-calcified plaque regions . Due to the non-invasive and contact-free character, the biomedical field pushed the development of this method towards in vivo diagnostics focusing the evaluation of ocular structures. Not surprisingly, in the last twenty years OCT was primarily applied in the biomedicine. However, the unique characteristics, namely to provide high resolution depth-resolved information also in strongly scattering media, make it an attractive method for a broad spectrum of research topics and applications beyond biomedicine . These applications include off-line characterization of paper , silicon integrated-circuits , food , fiber composites  and pharmaceutical tablets . Moreover, it was recently demonstrated as an in-line quality control tool for monitoring printed electronics , in-line characterization of multi-layered foils  and in-line monitoring of a pharmaceutical fluid-bed coating process . These publications reveal the immense potential of the method. Specifically, quality control of pharmaceutical materials could act as a key application to respond to specific demands and open questions in this field.
Optical Coherence Tomography
OCT enables the acquisition of cross-sectional or 3-D reconstructions of semi-transparent or turbid materials by measuring the echo time delay and magnitude of back-scattered and back-reflected light.
The light is back-reflected from different sample structures (e.g., interfaces, impurities, pores and cells) defined by surfaces separating two media with different refractive indices. In some sense it can be compared to ultrasound, although in OCT distances between interfaces within the sample are measured using light instead of sound waves. The application of light enables resolutions down to several microns (typically 1 - 15 µm). However, the high speed of light makes it impossible to use the direct measurement of the "optical echo" to determine such short distances. This method thus uses an interferometric technique together with a low-coherent light source (i.e., high spatial and low temporal coherence), which allows the measurement of axial-(depth) profiles of a sample . A sequence of such 1-D axial measurements performed at different transversal positions enables the synthesis of two- (2-D) and three-dimensional (3-D) images. A transverse displacement between successive axial measurements can either be carried out with the aid of some scanning means (causes a deflection of the optical beam), such as a galvanometer mirror, or by a relative transverse movement of the sample to the sensor head.
In the case of employing it as an in-line sensor, where the samples under investigation are moving, no internal scanning procedure is required. On the other hand, for off-line quality control applications, a galvanometer mirror is employed to scan the beam across the sample. In the following pharmaceutical applications a galvanometer mirror was used to generate cross-section images.
In pharmaceutical manufacturing numerous quality tests were introduced to guarantee the quality and safety of pharmaceutical products. Regulatory facilities stimulated the pharmaceutical industry by several approaches, including Quality by Design (QbD) and Process Analytical Technology (PAT), to raise their interest in an in-depth understanding of process and product characteristics. Implementing a suitable control strategy based on proper in-line monitoring techniques is essential for process improvement, optimization and quality assurance. In-line monitoring strategies require the assessment of critical process parameters such as coating thickness and its homogeneity for a coating process. It is specifically important to control theses parameters in case of functional coatings, which typically control the rate of drug release and therefore, directly affect efficacy and patient safety.
Recent studies show that OCT exactly meets those needs. Figure 1 depicts images of several commercially-available film-coated tablets of different shapes, formulations and coating thicknesses . These images were acquired by OCT systems which use a light source operating at a central wavelength of 830 nm and a bandwidth of 62 nm. Such a light source provides an axial resolution (limited by the specifications of the light source) of 7.5 µm. Penetration depth and image contrast strongly depend on the optical properties of the tablets (i.e., refractive index, scattering and absorption properties).
Figure 1 also illustrates images acquired during the movement of the sensor head of the OCT system across a static tablet bed. This experimental setup was used to gain a basic understanding of the relationship between tablet speed and motion effects, which occurs during in-line measurements . Evaluating the homogeneity of the coating turns more difficult with increasing speeds, whereat the determination of the coating thickness is still highly accurate at speeds up to 0.7 m/s.
Beside the investigation of tablets, the method even allows the evaluation of the coating of pellets. Figure 2 depicts the experimental setup for in-line monitoring of a fluid-bed coating process . The sensor head was protected from the pellets and the coating by a thin plastic foil. This foil can be seen by two horizontal lines at the top of the in-line OCT images illustrated in figure 2. The gathered images were post-processed to provide not only the mean coating thickness but also information about the inter- and intra-pellet coating variability.
The coating thickness could be determined directly, without a chemometric calibration model required for the quantification. The direct integration of the OCT sensor head into the fluid-bed systems allowed continuous monitoring of the coating growth. Moreover, the in-line investigation of the intra- and inter-pellet coating uniformity was possible due to OCT‘s high acquisition rate. Since OCT provides cross-section images of the pellets, it facilitates the determination of several thickness measurements per pellet. Such data is used to analyze local coating properties (i.e., intra-pellet coating variability) in dependence on process time. The used experimental setup provided up to 130 images of pellets per minute, which allowed the analysis of coating thickness variations between pellets, denoted as inter-tablet coating variability. This parameter is quantified by the standard deviation of the mean thickness measurements and is marked in figure 2 (bottom) by the error bars.
The in-line measurements were validated by samples drawn every 6 minutes. Images of several pellets for each sample were acquired and used for further analysis. In the absence of motion effects and vibrations, the contrast and image quality was higher compared to the in-line measurements. However, the off-line OCT measurements are in high agreement with the in-line measurements.
Three batches were produced under the same process conditions demonstrating the reproducibility of the results. Therefore, the OCT technology allows the operator to directly monitor the coating thickness and uniformity in sub-micron resolution, which makes it a promising in-line PAT method.
Material characterization and quality control is a necessary and crucial step in pharmaceutical industry. The results of the first studies on pharmaceutical application show that OCT reveals a promising potential for the quality control of solid dosage forms. The direct thickness measurement of film-coatings with high resolution and without prior calibration by means of this novel PAT tool is an alternative for the currently established technologies such as near-infrared (NIR) and Raman spectroscopy. However, applying this method as an in-line quality control tool requires beside a suitable mounting position in a coater, an automatic evaluation algorithm. Such an algorithm needs to detect the samples and to determine the coating thickness in real-time. The implementation of such an algorithm is a significant challenge, but is essential to make OCT working for in-process applications. Besides the development of an OCT system suitable for pharmaceutical application, the Research Center Pharmaceutical Engineering together with the Research Center for Non-Destructive Testing works on the development of algorithms suitable to compute representative parameters in real time. This will support monitoring and control of production processes in batch and continuous mode.
 D. Huang et al.: Optical coherence tomography, Science (80-.). 254 1178-1181 (1991)
 Stifter D.: Appl. Phys. B. 88 337-357 (2007)
 Prykäri T. et al.: Opt. Rev. 17 (2010) 218-222.
 Serrels K.A. et al.: Microelectron. Eng. 87 1785-1791 (2010)
 Verboven P. et al.: Postharvest Biol. Technol. 78 123-132 (2013)
 Stifter D. et al.: Meas. Sci. Technol. 19 074011 (2008)
 Koller D.M. et al.: Eur. J. Pharm. Sci. 44 142-148 (2011)
 Alarousu E. et al.: Sci. Rep. 3 1562 (2013)
 Hanneschläger G. A. et al.: International Workshop of NDT Experts, Prague, p. Paper 7 (2011)
 Markl D. et al.: Chem. Eng. Sci. (2014)
 Wojtkowski M.: Appl. Opt. 49 30-61 (2010)
 Markl D. et al.: Eur. J. Pharm. Sci. 55 58-67 (2014)
 Markl D. et al.: Opt. Lasers Eng. 59 1-10 (2014)
Daniel Markl1, Stephan Sacher1, Johannes G. Khinast1,2
1Research Center Pharmaceutical Engineering GmbH,
2Institute for Process and Particle Engineering
Graz University of Technology