(T1530-09-57) An Automated Process Analytical Interface to Estimate Tablet Porosity and Height Based on Terahertz Time-Domain Spectroscopy Reflection Measurements

Purpose: The physical quality attributes of tablets are commonly assessed using tablet weight and breaking force testers, which are destructive and sample a small number of tablets.¹ While the industry standard is to use an automated at- or on-line setup, it is still found insufficient to meet the industry and regulatory authorities aim of implementing process analytical technologies (PAT) to improve process understanding and facilitate the transition to continuous manufacturing.² To address the gap, this study proposes a strategy for automating tablet porosity and height measurements using terahertz time-domain spectroscopy (THz-TDS) in reflection mode with the potential of replacing the conventional tablet weight and breaking force measurements. The study includes the development of a prototype measurement setup and an algorithm for selecting relevant measurements, with the ultimate aim of achieving a measurement speed of one tablet per second.

Methods: Microcrystalline cellulose (MCC) tablets were compressed (XP 1, Korsch AG, Berlin, Germany) at 230, 240, and 250 mg each with 10, 15, 20, 25, and 30 % porosity as replicates of six corresponding to 90 tablets in total. Flat-faced tooling with a diameter of 10 mm was used. Tablets were measured with THz-TDS (Terapulse LX, Teraview Ltd., Cambridge, UK) in reflection focused approximately at the average tablet height. Reference measurements were carried out with a gold mirror at positions from ‑500 – 500 µm relative to the focus position. THz-TDS measurements were carried out continuously at a 50 Hz acquisition rate while tablets were presented to the probe for 0.5 s using a newly designed prototype setup. Tablets were manually dropped onto the chute through a funnel (see Figure 1) of the prototype to simulate an in-line setup.

Results: Figure 1 displays the automated measurement setup, featuring a wheel catching and directing tablets from position A to B for THz-TDS acquisition and subsequently releasing them at position C. The settings utilized in this study allowed for a 0.5 s time window for tablet presentation to the probe and an additional 0.5 s to release the tablet and prepare for the next. However, the actual waiting time was longer due to manual tablet placement. Measurements were continuously taken without interference from the measurement setup, and recordings of signal reflection from the chute or wheel (Figure 1) were taken between tablet presentations. Figure 2 depicts the THz-TDS response for these different scenarios, but only displaying measurements taken from the tablet surface exhibiting a strong reflection. A signal intensity threshold of a maximum intensity of 0.5 for data cleaning was utilized to select measurements reflected from the tablet surface. Further, moving a tablet from position A to B will always result in a time window during which the chute is measured, allowing to group measurements of a single tablet. The set acquisition rate of 50 Hz and the hold time of 0.5 s resulted in around 30 measurements for each tablet. This includes measurements where the tablet is not held still yet but moved into position B. To avoid including these measurements, 20 measurements were taken from the center of this time window and subsequently averaged and analyzed. For each signal, two parameters were extracted: the time of flight of the reflection of the signal based on the signal maximum (t_max) and the amplitude after Fast Fourier Transformation at 0.3 –0.55 THz. Tablet height: Figure 3 shows the signal’s time of flight as a function of tablet height. As the distance of the reflection head is not adjusted once the measurements are started, the time of flight of the signal solely depends on the distance between the reflecting surface, i.e., the tablet face and the detector. Hence, a clear correlation (2^nd order polynomial, adjusted R² = 0.964) between tablet height and t_max was identified (Figure 3), which after an initial calibration with tablets of known height, can be used for tablet height determination. Tablet porosity: Figure 3 shows the ratio between the amplitudes reflected from the tablet surface and the reference mirror at similarly recorded time-delays to compare at same distance to the detector. The intensity of the reflected amplitude at an air/solid interface highly depends on the refractive index of the solid material, which is well known to correlate linearly with the porosity of pharmaceutical tablets. Hence, a linear trend (R² = 0.9637) can also be observed for amplitude ratio and tablet porosity in Figure 3. Similar to tablet height, a regression function can be used to estimate tablet porosity once established.

Conclusion: The correlation between t_max and tablet height and amplitude ratio and tablet porosity can determine these two parameters once calibration functions are established for a particular tablet product. Since both parameters can be obtained from the same measurement and within an overall measurement time of one second, the setup shows potential for on-line PAT-based control of these important physical properties of tablets.

References: 1. Singh, R., et al., "Implementation of an advanced hybrid MPC–PID control system using PAT tools into a direct compaction continuous pharmaceutical tablet manufacturing pilot plant." International Journal of Pharmaceutics, 2014. 473(1-2).
2. D. Markl, J. Sauerwein, D. J. Goodwin, S. van den Ban and J. A. Zeitler. 2017. "Non-destructive Determination of Disintegration Time and Dissolution in Immediate Release Tablets by Terahertz Transmission Measurements." Pharm. Res. 34 (5): 1012-1022. doi: 10.1007/s11095-017-2108-4

Acknowledgements: Novo Nordisk A/S (Bagsværd, Denmark) has financed the PhD project of Moritz Anuschek (STAR program). Thomas Kvistgaard Vilhelmsen is employed at Novo Nordisk A/S. Jukka Rantanen and J. Axel Zeitler have not received any consulting fees from Novo Nordisk A/S.

Automated process analytical interface for tablet presentation to the THz-TDS probe. The tablet gets caught by the turning wheel at position A, presented at position B to the THz-TDS probe for the measurement, and released at position C.

Signal response of tablets measured close to the focal position, the chute, and the wheel. For the tablet examples in the signal response, different heights and different porosities are exemplified. Please note that the signal response of the tablet is only for illustrative purposes and is not based on real measurements.

Time of maximum (tmax) as a function of tablet height and the ratio of the amplitude between tablet and mirror (amplitude ratio) as a function of tablet porosity.