Why Continuous Real-Time Raman PAT Outperforms Discrete PAT in Bioprocessing

Overview

Continuous, in-line, real-time process Raman spectroscopy significantly outperforms discrete PAT approaches in bioprocessing. By enabling deeper process understanding and proactive control, Raman spectroscopy can monitor upstream monoclonal antibody production through real-time measurements of glucose and lactate, which enables automated feedback control that significantly increased titer and reduced glycation (as high as a 10% increase in titer and a 60% reduction in glycation); such outcomes would be difficult to achieve with intermittent sampling. In downstream UF/DF operations, real-time Raman monitoring of sucrose and other components allows direct observation of volume exclusion effects, improved endpoint determination, and reduced inefficiencies and batch risk. Broadly speaking, continuous Raman-based PAT shifts bioprocessing from reactive, assumption-based decision-making to a data-driven, feedback-controlled process.

Continuous data PAT vs discrete PAT differences

In biomanufacturing, process analytical technology (PAT), particularly process Raman spectroscopy, is not just a regulatory expectation—it is a practical necessity. Every process scientist has experienced the challenge of making decisions based on limited data while a dynamic biological system continues to evolve.

Over the course of our experimental studies and the application work referenced here, we evaluated both discrete and continuous PAT strategies, with a focus on process Raman technology, in real manufacturing-relevant environments. Based on the data we generated, one conclusion became clear: Continuous, in-line, real-time PAT provides a fundamentally deeper level of process understanding and control than discrete sampling approaches.

Discrete PAT is comprised of at-line, online, or offline measurements where samples are periodically withdrawn and analyzed. These approaches provide snapshots of process performance.

Continuous data PAT is the term for in-line technologies that acquire high-frequency, near real-time measurements directly from the process stream, generating data-rich profiles that reflect true process dynamics. An example of continuous data PAT is a process Raman analyzer.

The difference between the two is not merely analytical speed. It is the difference between observing a process occasionally and truly understanding it as it unfolds.

Bioprocessing PAT case studies

Case I: Improving bioreactor efficiency by using continuous data PAT

In upstream monoclonal antibody (mAb) production, glucose and lactate metabolism are highly dynamic. During early cell growth, excess glucose leads to lactate accumulation. Later, under glucose-limited conditions, cells can consume lactate as a secondary carbon source. Literature evidence and our internal experimental data consistently show that elevated lactate negatively impacts productivity, while excess glucose increases unwanted glycation.

At the same time, complete glucose depletion is not viable; a basal concentration must be maintained to preserve cell health. The real challenge is maintaining that delicate balance between glucose and lactate in real time as the process is happening.

To address this, we deployed in-line process Raman spectroscopy to simultaneously monitor glucose and lactate throughout the bioreactor run. Rather than reacting to delayed laboratory results, we implemented a feedback control strategy that maintained the combined glucose and lactate concentration at 2 g/L.

When lactate accumulated and total carbon exceeded this threshold, glucose feeding was paused. This encouraged cells to metabolize lactate while maintaining sufficient glucose for viability.

The outcome was measurable and significant:
– 10% increase in titer
– 60% reduction in glycation

These results align with the findings described in the referenced application note on automated multi-component feedback control. They were only achievable because continuous data enabled proactive adjustments during the run.

Attempting this strategy using discrete sampling would have required impractically high sampling frequency, introduced contamination risk, increased analytical burden, and still failed to capture rapid metabolic shifts. Continuous data transformed what would have been reactive decision-making into active process control.

Reference: Enhancing Monoclonal Antibody Yield and Quality Through Automated Multi-Component Feedback Control

Case II: Real-time monitoring of sucrose volume exclusion during UF/DF

Downstream processing presents equally dynamic challenges. Ultrafiltration/Diafiltration (UF/DF) is used to concentrate product and exchange buffers into final formulations containing stabilizing excipients such as sucrose.

While often considered a straightforward buffer exchange step, UF/DF involves complex phenomena including Donnan effects, concentration polarization, membrane–protein interactions, and sucrose volume exclusion.

During high-concentration processing, sucrose does not always distribute ideally within the retentate volume. Volume exclusion effects can lead to deviations from target formulation concentrations, directly affecting product stability and quality.

In our downstream evaluation studies, we implemented in-line process Raman spectroscopy to continuously monitor sucrose concentration during UF/DF, along with protein, histidine and arginine. Instead of relying on theoretical diavolume calculations or intermittent HPLC testing, we were able to observe sucrose behavior as the process evolved.

As highlighted in the referenced downstream application note, in-line Raman serves as a continuous data PAT that allowed us to monitor the decrease in sucrose concentration in real time as the protein was concentrated to high concentration. Interestingly, histidine and arginine remained unchanged.

This use of process Raman spectroscopy allowed us to:

• Directly observe sucrose volume exclusion
• Determine process endpoints based on measured data
• Detect deviations immediately
• Reduce over-diafiltration and associated inefficiencies

In fast-paced UF/DF operations, delayed discrete measurements can prevent timely corrective action, leading to batch failure. Continuous monitoring provided visibility and confidence, significantly reducing the risk of costly deviations.

Reference: Process Raman as a comprehensive solution for downstream buffer workflow

From data points to process intelligence

Discrete PAT provides isolated data points.

Continuous data PAT provides a living, evolving process profile.

Discrete PAT	Continuous data PAT
Periodic sampling	Real-time in-line monitoring
Limited data density	High-frequency data streams
Reactive adjustments	Proactive, feedback-based control
Dependence on theoretical assumptions where data are missing	True data-driven decision making

Conclusion

Continuous data PAT is a foundation for advanced bioprocess control.

Across both upstream and downstream case studies, the data generated in our experiments as referenced application notes demonstrate a consistent outcome: Continuous data PAT improves yield, enhances product quality, reduces variability, strengthens overall process robustness, and allows the implementation of advanced multi-component feedback control strategies.

As biomanufacturing continues to move toward higher titers, intensified processes, and greater automation, continuous in-line technologies such as Process Raman spectroscopy are not simply analytical upgrades — they are foundational tools for next-generation process control.

Ultimately, the ability to see the process clearly in real time — and act on that insight with confidence — is what defines modern, data-driven, artificial intelligence (AI) powered biomanufacturing.

---

Frequently Asked Questions