In this interview, Dr. Nimesh Khadka, Senior Product Application Specialist at Thermo Fisher Scientific, explores the role of Raman spectroscopy as a process analytical tool (PAT) in bioprocess monitoring, highlighting its molecular specificity, real-time capabilities, and advantages over traditional analytical methods like ultraviolet-visible (UV-Vis) spectroscopy.
Could you explain the principles and functioning of Raman spectroscopy?
Raman spectroscopy is a technique that examines how light interacts with molecules. When light interacts with vibrating molecules, most of it bounces unchanged – this is called a Rayleigh scattering. However, a small portion of the light shifts in energy because the molecules absorb or release small amounts of energy as they vibrate—this is known as Raman scattering. Although this scattered light is faint, it carries unique molecular fingerprints, helping scientists identify and measure different substances with precision.
What makes Raman spectroscopy particularly suitable for bioprocess monitoring?
Raman spectroscopy offers molecular specificity, which means it can identify molecules based on their distinct spectral fingerprints. Raman also works well in water-based solutions since water has a weak Raman signal. Because it’s an optical and non-destructive technique, it can take frequent, data-rich measurements without affecting the sample. This helps to capture process variations in real-time. Raman can also detect multiple metabolites in a single scan without degrading samples, making it invaluable for dynamic bioprocess environments.
In addition, Raman spectroscopy can detect several key protein features, including amide I, II, and III regions, which reveal details about peptide bonds and secondary structures. Raman also picks up aromatic residues like phenylalanine and tryptophan, as well as disulfide bonds. These insights help scientists monitor protein behavior, detect aggregation, and assess structural configurations.
The speed of Raman analysis is particularly beneficial in bioprocessing. Real-time monitoring enables precise process control, such as initiating UF/DF at specific protein concentrations. It helps maintain consistent product quality, reduces costs, and saves time. Further, Raman can simultaneously monitor multiple parameters at once, enhancing understanding of bioprocess dynamics for better decision-making.
How does Raman spectroscopy differ from UV-Vis spectroscopy in downstream process monitoring?
UV-Vis spectroscopy can be affected by other substances in the sample, such as tryptophan, leading to inaccurate protein concentration measurements. In contrast, Raman spectroscopy is unaffected by these interferences, ensuring more reliable data. It can also quantify proteins and excipients like sucrose and histidine in a single scan, giving a more complete picture of the process.
For example, we performed an experiment focused on monitoring protein concentrations during ultrafiltration/diafiltration (UF/DF) in downstream monoclonal antibody production. Raman's performance was compared with in-line UV-Vis and at-line UV-Vis. The results showed Raman's high accuracy (less than 5% error) and showcased its ability to deliver reliable data even in the presence of interfering compounds like tryptophan.
In addition, Raman spectroscopy is highly specific, detecting only the target molecules without interference from others. In our experiment, UV-Vis overestimated protein concentrations by more than 70% due to tryptophan interference. In contrast, Raman accurately measured concentrations with less than 10% error since tryptophan did not affect the Raman signals in the targeted spectral regions.
How is Raman integrated into bioprocesses?
Raman can be integrated into bioprocesses in three ways: in-line, at-line, or on-line. In-line uses immersible probes placed directly in the process for real-time monitoring and control. At-line requires taking samples for analysis off-line, while on-line involves sampling tubes to transport material to the Raman instrument. In-line Raman is most effective for continuous data collection and process control, ensuring consistent product quality.
There can be challenges with Raman analysis including baseline shifts, path length differences, and spectral overlaps. These issues are managed using data pre-processing techniques such as baseline removal (using filters like Savitzky-Golay) and normalization (e.g., water band normalization). Dimensionality reduction techniques like Partial Least Squares (PLS) regression help build robust chemometric models for accurate predictions.
Can Raman be used for advanced applications beyond basic monitoring?
Yes, Raman goes beyond basic monitoring to support crucial bioprocess quality control, such as deconvoluting protein secondary structures and quantifying protein aggregation. Raman can isolate protein-specific spectra from excipient interference and analyze beta-sheet content or aggregation profiles in real-time, proving deeper insights into process stability and product quality.
Can you summarize the overall value of Raman spectroscopy in bioprocessing?
Yes, process Raman spectroscopy is a versatile tool for real-time monitoring and control in bioprocessing. Unlike traditional methods such as UV-Vis, it provides molecularly specific, non-destructive measurements, even in complex mixtures. Its application ensures high-quality outcomes in both upstream and downstream processes, making it a reliable and cost-effective solution for the biopharmaceutical industry.
About Nimesh Khadka 
Dr. Nimesh Khadka is a Senior Product Application Specialist at Thermo Fisher Scientific and holds a PhD in Biochemistry focusing on bioanalytical techniques.
With extensive expertise in spectroscopy and chemometrics, he is committed to driving innovation through process analytical technology (PAT). Dr. Khadka is particularly passionate about addressing complex challenges in bioprocessing by leveraging advanced tools. Currently, he is at the forefront of promoting Raman spectroscopy as a practical and effective solution for monitoring and controlling bioprocesses within the realm of PAT.