Raman spectroscopy has a long-held association with biophotonics, and the same qualities that make the technique indispensable for biomedical imaging and sensing are now driving real-time process insights across industrial fields. Advancements in the sensitivity, speed, and ruggedization of Raman technology are enabling its deeper integration into materials science, failure analysis, chemistry, pharmacology, forensics, and more. Raman spectroscopy systems are finding increased adoption for the characterization of hydrocarbons and contaminants in the oil and gas industry. Courtesy of Thermo Fisher Scientific. As a result, Raman systems are finding expanded use for many nonroutine applications. Raman’s noninvasive chemical insights make it ideally suited for these applications and the evolving demands of modern process monitoring and control. Material insights The advent of Raman monitoring in manufacturing stems, at least in part, from the wave of automation spreading through industry. Increasingly automated and regulated processes require efficient and real-time process monitoring solutions. Technology providers in sectors ranging from biotechnology to oil and gas are focused on finding and implementing innovative analytical tools. As manufacturing processes grow more automated and regulated, technology providers recognize the continued need for efficient, real-time process monitoring. The focus is on developing innovative analytical tools that support industries ranging from biotechnology to oil. Courtesy of Thermo Fisher Scientific. According to Mark Kemper, director of business development at Bruker Optics, obtaining a complete understanding of a set of processes is vital to achieve maximum control over them. “A technique like Raman can be extraordinarily valuable in this regard. Having a tool that can help diagnose what is happening, from a molecular standpoint, can be invaluable, [thereby] adding a piece to the overall puzzle of process understanding and control,” he said. Improving performance is not the only motivation for end users to adopt Raman systems; the portability and affordability are also compelling. Together, these qualities make Raman an inherently attractive technology. A Raman fiber optic probe is inserted into a bioreactor to monitor cell growth under sterile conditions. Courtesy of Metrohm. During the past two decades, though, Raman’s popularity has also surged because of component innovations, such as those underway in light sources, optics, and sensors. Advancements are especially evident in compact and rugged lasers, filters, lenses, and detectors. And, alongside these hardware improvements, expanded spectral libraries and more sophisticated matching algorithms have also extended Raman’s capabilities. These areas of progress have converged into a trend that is enabling increasingly advanced analytical tasks. In parallel, shifting industry needs are pulling Raman further onto the production floor. Faster, more accurate analytics are no longer merely desirable — they have become essential for an increasing number of critical applications, from detecting illicit fentanyl analogs and chemical warfare agents to real-time monitoring of downstream bioprocessing. As it translates to system performance, adopters are turning to Raman solutions for what they can detect and for how quickly and clearly they tell a complete story. “Process people need to pay attention to their processes,” Kemper said. “Their process analytical technology [PAT] tools simply need to work. The minute [manufacturers] have to worry more about their analytical tools than their processes, the instrumentation will be mothballed.” Automation in pharma manufacturing Raman spectroscopy is widely used for reaction monitoring and crystallization of small molecules. As pharmaceutical manufacturing increasingly demands continuous process monitoring, Raman’s role is expanding too, in both upstream and downstream processes. Specifically, manufacturers are increasingly implementing Raman-based solutions for small-molecule manufacturing operations such as blending, mixing, tablet coating, granulation, and extrusion. A first responder uses a hand-held Raman analyzer to identify the compound xylene during a chemical spill response. A fiber optic immersion probe accessory provides high chemical resistance and can be used directly with the sample. Courtesy of Metrohm. “The integration of the automation of biomanufacturing is a cutting-edge aspiration in the biopharmaceutical industry,” said Nimesh Khadka, senior application PAT specialist at Thermo Fisher Scientific. “Reliable, accurate, real-time, and frequent data from in-line Raman process monitoring is key to achieving the automation goals.” Raman spectra have highly specific and attributable bands that can be traced to the molecular bonds in a material. This quality enables drugmakers, for example, to make qualitative and quantitative measurements — even from several hundred meters away, when necessary. This is critical for continuous process monitoring in the pharmaceuticals industry. Conventional analytical methods typically require offline analysis, which is often accomplished only via a time-consuming process that can lead to delayed feedback and potential product variability. An example of this is raw material identification in manufacturing. “Traditionally, a package would be opened, a sample taken, delivered to a lab, and meanwhile, the material is quarantined,” said Richard Crocombe, principal at Crocombe Spectroscopic Consulting. “With a portable spectrometer [Raman or NIR], the material can be scanned through a polyethylene liner or with an immersion probe and verified immediately and put into production.” Raman can also be used to make univariate measurements; these measurements focus on areas or intensities of single peaks and are easier to understand and use than multivariate methods. Further, nondestructive fiber optic-coupled probe measurements make in situ measurement much easier and simpler, contributing to better data than that which is typically obtained from offline measurements. For extrusion in the high-volume pharmaceutical and polymer industries, process Raman spectrometers integrated with a fiber optic Raman probe are helping manufacturers to monitor the chemical composition of the extrudate and ensure uniformity and accurate concentration in real time. “This is crucial for consistent drug release profiles and therapeutic efficacy,” said Tom Dearing, senior scientist at Thermo Fisher Scientific. “Process Raman spectrometers can also monitor polymorphic transformations, the formation of solid dispersions, and the distribution and chemical composition of combination products containing multiple active pharmaceutical ingredients,” Dearing said. The optical fiber probes used in these deployments are themselves the subject of ongoing design innovation. These components are now specifically designed to fit into the ports on the die or midbarrel. Tracking chemical changes with speed Raman plate readers that are currently on the market enable the live monitoring of reactions in up to 96 well plates. This capability is a boon to drug developers, and it marks a notable improvement compared to the first Raman systems to be used in and for PAT (see sidebar). It also positions Raman to surpass the NIR technologies that, according to Michael Allen, vice president of products and marketing at Metrohm Spectro Inc., enabled manufacturers to pioneer PAT solutions. “NIR has very broad peaks that must be integrated and deconvolved to construct meaningful analysis models,” Allen said. “In contrast, Raman spectra exhibit much sharper features, facilitating easier method development, more accurate analysis, and higher information density.” Examples of monitoring chemical changes exist, though Fran Adar, HORIBA Scientific’s principal applications scientist, believes there is room to further develop this set of applications. She points to recent software developments that decipher chemical changes even in complex spectra, such as those that are typically associated with chemical reactions. Adar cites the 2D-correlation spectroscopy software developed by Isao Noda at Procter & Gamble as a solution for users to analyze the evolution of the spectra. “Even when spectral features are heavily overlapped, this software enables one to follow the sequence of chemical changes as revealed in the spectra,” Adar said. North Carolina-based Wasatch Photonics makes available open-access software, software development kits, and plug-ins, all of which aim to streamline and simplify the development of software for application-specific instrument developers. The company also offers spectral calibration and spectral response corrections to ensure that users obtain consistent results — whether from a single instrument or a thousand — with high reliability. “We have one goal: to allow our OEM customers in PAT and other fields to focus on their application instead of the Raman hardware at its core, and to give them the raw data and software tools to have complete control over their analysis, to develop their own methodologies,” said Cicely Rathmell, the company’s vice president of marketing. From vaccines to therapeutics The monitoring and automation of feedback control of in vitro transcription reactions in messenger ribonucleic acid (mRNA) manufacturing real-time monitoring is an application that gained prominence during the COVID-19 pandemic. mRNA manufacturing hit the headlines for enabling the rapid development of Moderna’s COVID-19 vaccine. mRNAs also find use in the pursuit of personalized cancer vaccines. By packing mRNAs within lipid nanoparticles, the immune system receives “instructions” to target specific cancer cells. “Raman spectroscopy provides valuable information on individual lipid quality, their distribution within [lipid nanoparticles], and encapsulation efficiency by measuring the lipid-to-mRNA ratio. This ensures high-quality and effective drug delivery,” Thermo Fisher Scientific’s Khadka said. Similar biopharmaceutical applications include monitoring antibody-drug conjugate reactions for cancer therapy, as well as the manufacturing of chimeric antigen receptor (CAR) T-cells or viral vectors. For antibody-drug conjugates, clinicians use Raman spectroscopy to monitor the conjugation process and ensure the precise attachment of cytotoxic drugs to antibodies. Energy: Production, storage, and recycling Raman spectroscopy is proving to be invaluable in the pursuit of high-performance energy storage at industrial scales, and, specifically, in the development and production of battery materials. The technique delivers insight into the composition and uniformity of anode materials, such as graphite and silicon; cathode materials, such as lithium cobalt oxide and nickel manganese cobalt oxide; and certain electrolytes. According to HORIBA scientist Peng Miao, the Raman spectrum of hard carbons used in lithium and sodium batteries can be effectively analyzed spectroscopically because the approach can monitor and predict whether organic materials can be graphitized. In particular, the grain size, which predicts electronic conductivity and ion mobility, can be derived from the Raman spectra. A hand-held Raman analyzer is used to measure through an opaque container to reliably identify gasoline (top). A liquid sample is placed into a vial holder accessory of a Raman analyzer for identification (middle). A Raman analyzer is used to determine the presence or absence of narcotics in a plastic bag (bottom). A right-angle sampling accessory allows the operator to contact the sample without exposure. Courtesy of Metrohm. For this application, according to Thermo Fisher Scientific’s Dearing, a Raman-based approach can help detect impurities, analyze the formation of the solid-electrolyte interphase layer, and provide information on molecular interactions and phase changes that occur during battery cycling. “This aids in optimizing the manufacturing process, ultimately leading to more reliable and efficient batteries for various applications, from consumer electronics to electric vehicles,” he said. Raman spectroscopy also plays a critical role in end-of-life battery recycling. It helps to identify and characterize recovered anode and cathode materials, as well as electrolytes, to ensure their purity and readiness for reuse. But Raman’s utility extends further across the broader clean energy landscape; solutions are emerging in support of the modern demands of energy production. Among other end uses, Raman offers the necessary versatility to analyze materials and processes in solar cells and characterize hydrocarbons and contaminants in the oil and gas industry. For example, in carbon capture, use, and storage, Raman spectroscopy is crucial for analyzing the chemical composition of captured carbon dioxide, monitoring the efficiency of capture processes, and ensuring the integrity of stored carbon dioxide. Configuration trade-offs End users determine the degree to which a system-level advancement represents an improvement in comparison with prior iterations of a given solution. While an upgraded system must offer tangible benefits compared to the earlier version, it must also retain the most useful attributes for a given process monitoring task. As it relates to Raman instrumentation, the development of the high-throughput virtual slit has led to improved outcomes across application areas. This technology builds on traditional fiber-coupled dispersive Raman spectrometers. These devices exhibit a trade-off between sensitivity (signal strength) and spectral resolution (the instrument’s ability to distinguish closely spaced peaks). Narrowing the spectrometer slit improves resolution. At the same time, however, the system design restricts the amount of light entering the system, which reduces sensitivity. Conversely, widening the slit boosts throughput. But this serves to degrade resolution. Bruker’s Tornado High Throughput Virtual Slit (HTVS) technology is one of the commercial offerings that decouples this trade-off. “Instead of relying solely on a physical slit to define resolution, HTVS maximizes the amount of Raman light collected while maintaining high spectral resolution,” Bruker’s Kemper said. Specifications play a role in gauging the extent to which systems designers must weigh such a trade-off. According to Metrohm Spectro’s Allen, it is common for trade-offs to be made when configuring a system. At the same time, chasing top parameters is often unnecessary, he said. “[For example,] achieving fast, high-resolution measurements is more challenging and often unnecessary for monitoring a process. Overall, the miniaturization and low power requirements of all components allow a ‘more than good enough’ Raman spectrum to be collected.” Some Raman peaks are broad, Allen said — so increasing the system’s resolving power will not improve performance. “Designing a portable Raman system involves balancing competing demands, guided by the principle of achieving performance that is ‘good enough’ for the specific application,” Wasatch Photonics’ Rathmell said. Lingering limitations Despite its impressive growth, constraints remain for Raman technologies — particularly Raman spectroscopy. Fluorescence remains the most stubborn bottleneck. It is a noticeable obstacle in complex biologicals, crude mixtures, and/or heavily pigmented materials. Citing examples, Kemper said, “Applications such as crude oil analysis and analyses of certain foodstuffs and coffee are generally unsuitable for Raman. Fluorescence is simply too much of a barrier in some cases.” Advancements such as 1064-nm excitation, gated detection, and shifted-excitation Raman help to reduce this effect, though a silver bullet still eludes the industry. Cost is another consideration. While Raman instrumentation often appears more expensive than alternatives, its ability to measure multiple analytes simultaneously can help offset cost concerns. But this is not a simple one-to-one trade-off. Device costs remain prohibitive in some cases. Continued research, collaboration, and regulatory support across pharmaceutical and industrial markets are likely to promote the wider adoption of Raman spectroscopy as an integral PAT. In addition, the introduction and eventual integration of next-generation technologies, such as AI, will further automate monitoring and control processes. Built for the Factory Floor Decades ago, Raman instruments were massive systems — slow, delicate, and best suited for controlled laboratory conditions. “I started making measurements in the 1960s, when a Raman instrument was a very large double monochromator and had a single channel detector,” said Fran Adar, principal applications scientist at HORIBA Scientific. “It took something like an hour to produce a spectrum, and the output was a strip chart recorder.” Digital cameras, single-monochromator systems, and edge filters triggered major turning points. “The first development that changed things was the introduction of a microscope as a sampling device and then the digital camera, which was used as a multichannel (multiwavelength) detector,” Adar said. According to Adar, compact filters, benchtop computing, and multivariate analysis techniques accelerated the transition. “Finally, the development of chemometrics, also known as multivariate analysis, enabled extraction of information from complicated spectra with multiple components.” Contemporary Raman systems benefit from high-throughput optical engines — such as virtual slit architectures — along with stable diode lasers and ultrasensitive detectors. These advancements support measurements on faster time- scales and across more complex media, which is critical in the continuous and downstream operations necessary for process monitoring.