Process analytical technology and compensating for nonlinear effects in process spectroscopic data for improved...

Area of Science:

• Analytical Chemistry
• Biotechnology
• Process Analytical Technology (PAT)

Background:

• On-line/in-line quantitative monitoring of bio-chemicals and pharmaceuticals relies on spectroscopic instruments.
• Spectroscopic measurements are susceptible to variations in measurement conditions and sample properties, affecting accuracy.
• Traditional calibration models may fail due to non-linear relationships caused by these variations.

Purpose of the Study:

• To discuss the impact of variations in measurement conditions and sample properties on spectroscopic measurements.
• To present an overview of recent methodologies for modeling and correcting these detrimental effects.
• To demonstrate the applicability of discussed methods using complex datasets and industrial applications.

Main Methods:

• Development of robust multivariate calibration models.
• Modeling and correction of variations in measurement conditions (e.g., temperature).
• Modeling and correction of variations in sample physical properties (e.g., cell density, particle size).

Main Results:

• Demonstrated methodologies for addressing variations in spectroscopic data.
• Successful application of algorithms to complex datasets.
• Validation of techniques in an industrial plant setting.

Conclusions:

• Robust calibration models are crucial for reliable on-line/in-line spectroscopic monitoring.
• Addressing variations in measurement conditions and sample properties is key to maintaining model accuracy.
• The presented approaches offer effective solutions for real-time quantitative analysis in biopharmaceutical processes.