Monitoring and Controlling Powder Blending Online at AstraZeneca

Significance of the topic

Ensuring blend uniformity in powder blends for tablets and capsules is critical for consistent dose delivery and product quality. Traditional thief sampling is labor-intensive, potentially hazardous and introduces sampling variability that can bias quality decisions. Online, non-contact monitoring using near-infrared (NIR) spectroscopy offers a route to continuous, representative measurements that can reduce or eliminate the need for destructive post-production sampling and improve process understanding and control.

Objectives and overview of the study

This study evaluated a microelectromechanical systems (MEMS)-based NIR spectrometer (Thermo Electron/Antaris Target Series Blend Monitor) as an online tool to monitor powder blend uniformity in a bin blender. Goals included demonstrating feasibility for detecting blend endpoint, characterizing measurement repeatability under different process conditions (fill level, rotation), and identifying practical considerations for implementation across lab, pilot and production scales.

Instrumentation used

The key instrument evaluated was the Antaris Target Series MEMS NIR blend monitor. Principal specifications and features:

• Spectral range: 1350–1800 nm.
• Light source: semiconductor-based NIR tunable laser.
• Wavelength selection: high-resolution Fabry–Pérot tunable filter (4 or 8 cm−1 resolution).
• Detector: single-element InGaAs photodiode.
• Optical interface: measurements through a sapphire window integrated into the blender lid or vessel.
• Mechanical/electronic: MEMS accelerometer for rotation-triggered acquisition, battery operation, hermetically sealed bench under dry nitrogen, internal referencing for wavelength and absorbance stability, no moving macro parts.
• Performance: ~10 scans per second; 5 scans averaged per revolution in the study; approximate sampling spot ~40 mm (~600 mg dosage equivalent).

Methodology

Lab-scale evaluations used a 20 L Bohle bin blender and a model formulation: acetaminophen (API), microcrystalline cellulose, spray-dried lactose monohydrate, crospovidone and magnesium stearate. The analyzer was mounted in a modified lid and configured to collect five scans per blender revolution, triggered by the accelerometer at a preset rotation angle. Spectral preprocessing employed second derivative transformation to suppress baseline variation and emphasize subtle spectral features. Blend homogeneity was quantified using a moving-block approach: for each block of rotations (block size seven in the study), the standard deviation at each wavelength was computed and then summed across wavelengths to yield a single summed spectral standard deviation metric per block. Plotting this metric versus number of rotations produced blend curves used to define the spectral endpoint.

Main results and discussion

- Raw and second-derivative spectra showed distinct contributions from API and excipients; early blending produced large spectral variation that decreased as blending progressed, with major reduction in variability by approximately rotation 15 for the tested conditions.
- Blend endpoint detection via summed-moving-block spectral standard deviation correlated with visual and spectral convergence: as summed deviation reached a low plateau the blend was considered homogeneous.
- Fill level strongly affected blend time: a 90% filled vessel required approximately three times more rotations to reach the same spectral homogeneity as a 60% filled vessel, demonstrating the limitation of time-based controls and the importance of spectral endpoint monitoring.
- The approach is sensitive to material attributes (bulk density, moisture content, particle surface area) and lot-to-lot excipient variability, which can alter blend kinetics and endpoint timing.
- The MEMS-based instrument provided rapid (sub-second per revolution) stable measurements that were insensitive to blender position and vibration, facilitating robust online acquisition.

Contributions and practical applications

This study demonstrates that a compact MEMS NIR spectrometer can:

• Provide representative, continuous monitoring of powder blends without contact or sample withdrawal, reducing operator exposure and labor compared with thief sampling.
• Deliver rapid, reproducible spectra suitable for endpoint detection across lab-to-manufacturing scales when integrated into blender lids or vessels.
• Enable spectral endpoint algorithms to replace fixed-time blending rules, accommodating process variations such as fill level and excipient lot changes.
• Support lifecycle PAT (Process Analytical Technology) use cases, including in-line blend acceptance and potentially downstream monitoring (e.g., before compression and tablet content uniformity checks), reducing offline testing burden.

Limitations, implementation considerations and challenges

Key practical and regulatory considerations include:

• Instrument qualification and software/hardware validation per GMP requirements.
• Appropriate hazard classification and cleaning ratings for use in production environments, and ensuring the optical window remains representative and clean.
• Battery operation, wireless communications and physical mounting must meet site safety and process integration constraints.
• Algorithm robustness: endpoint detection algorithms must be validated across formulations, scales and expected variabilities (e.g., excipient lots, fill levels, blender types).

Future trends and potential uses

Anticipated developments and applications:

• Refinement and validation of automated endpoint-detection algorithms (multivariate metrics, moving-window statistics, chemometric models) to improve reliability across diverse formulations and process conditions.
• Scale-up studies to confirm translatability from lab to pilot and manufacturing blenders, including adjustment of optical sampling geometry and modeling for sample representativeness.
• Integration into broader PAT frameworks to follow material from blending through compression and coating, enabling real-time release strategies and reduced end-product testing.
• Adoption of MEMS-based spectrometers for other unit operations (drying, granulation, raw material ID) thanks to small footprint, low power and sealed optical benches.

Conclusion

The evaluated MEMS-based NIR spectrometer is a viable tool for online monitoring of powder blend uniformity. It delivers rapid, reproducible spectral data compatible with derivative preprocessing and block-standard-deviation endpoint algorithms. Spectral endpoint monitoring addresses shortcomings of thief sampling by offering representative, non-destructive, and continuous assessment of blend homogeneity, with clear potential to reduce laboratory workload and improve process control. Successful deployment requires attention to instrument qualification, algorithm validation and practical integration into production environments.

Reference

1. FDA. Current Good Manufacturing Practice: Amendment of Certain Requirements for Finished Pharmaceuticals; Proposed Rule (61 FR 20103), May 1996.
2. Boehm G., Clark J., Dietrick J., et al. The Use of Stratified Sampling of Blend and Dosage Units to Demonstrate Adequacy of Mix for Powder Blends. PDS J Pharm Sci Tech 57:59–74, 2003.
3. FDA. Guidance for Industry: Powder Blends and Finished Dosage Units – Stratified In-Process Dosage Unit Sampling and Assessment, October 2003.
4. Hwang R.-C., Wu S.-J. Challenges of Blend Uniformity Testing for Tablet Formulation. American Pharmaceutical Review 7:101–103, Jan/Feb 2004.
5. FDA. Guidance for Industry: PAT – A Framework for Innovative Pharmaceutical Development, Manufacturing and Quality Assurance, September 2004.
6. Sekulic S., Ward H., Brannegan D., et al. On-Line Monitoring of Powder Blend Homogeneity by Near-Infrared Spectroscopy. Analytical Chemistry 68:509–513, 1996.
7. Berntsson O., Danielsson L.-G., Lagerholm B., Folestad S. Quantitative In-line Monitoring of Powder Blending by Near Infrared Reflection Spectroscopy. Powder Technology 123:185–193, 2002.
8. Cogdill R., Anderson C., Delgado-Lopez M., et al. Process Analytical Technology Case Study Part I: Feasibility Studies for Quantitative Near-Infrared Method Development. AAPS PharmSciTech 6(2):E262–E272, 2005.
9. Crocombe R. MEMS Technology Moves Process Spectroscopy into a New Dimension. Spectroscopy Europe, July 2004.
10. Parris J., Airiau C., Escott R., Rydzak J., Crocombe R. Monitoring API Drying Operations with NIR. Spectroscopy 20(2):34–42, Feb 2005.
11. Sullivan M. The Use of NIR as a PAT Tool for Measuring Blend Uniformity. Spectroscopy Supplement: Process Analytical Technologies, Feb 2006.
12. Sekulic S.S., Wakeman J., Doherty P., Haily P.A. Automated System for the On-line Monitoring of Powder Blend Processes Using Near-Infrared Spectroscopy Part II: Qualitative Approaches to Blend Evaluation. J Pharm Biomed Anal. 17:1285–1309, 1998.