Stimulated Raman scattering (SRS) is a light-based process in which two laser beams interact with molecules, causing one beam to gain energy and the other to lose it, with the difference going directly into making the molecules vibrate. That energy transfer only happens when the frequency difference between the two beams matches a specific vibration in the target molecule, which makes SRS an extraordinarily selective way to identify and image chemicals without needing dyes or labels. Since its development into a practical microscopy tool in the late 2000s, SRS has found a surprisingly wide range of uses, from spotting tumor boundaries during brain surgery to detecting nanoplastics in bottled water.
How the Two Beams Work Together
Every molecule vibrates at characteristic frequencies determined by the masses of its atoms and the strengths of the bonds between them. In conventional (spontaneous) Raman scattering, a single laser beam hits a sample and a tiny fraction of the light shifts in frequency by an amount that corresponds to one of those vibrations. The signal is real but extremely weak, which limits how fast you can image anything.
SRS gets around that weakness by using two beams at the same time, called the pump beam and the Stokes beam. When the frequency difference between them matches a molecular vibration, the interaction stimulates the molecule to scatter light far more efficiently than it would on its own. What you see in practice is that the Stokes beam picks up extra photons (called stimulated Raman gain) while the pump beam loses photons (stimulated Raman loss). Measuring either of those tiny changes in intensity gives you a signal that is directly proportional to the concentration of the vibrating molecules in the sample.1PubMed Central. Fast vibrational imaging of single cells and tissues by stimulated Raman scattering microscopy
This proportionality is one of the key practical advantages. Because the signal scales linearly with how many target molecules are present, you can use SRS to do quantitative chemistry on a sample, not just detect that something is there but measure how much of it is there. That distinction matters a great deal in applications like drug uptake studies and tissue diagnostics.
Why SRS Rather Than Other Raman Techniques
The most common alternative coherent Raman method is called coherent anti-Stokes Raman scattering, or CARS. Both techniques use two laser beams tuned to the same vibrational frequency, and in fact both processes happen simultaneously in the sample. The critical difference is that CARS generates a new beam at a shifted frequency, while SRS modifies the existing beams. That new CARS beam carries a so-called nonresonant background, a signal that arises even when no target molecule is present, which distorts images and makes quantitative measurements unreliable.
SRS does not have this problem. When the beams are not tuned to a vibration, there is no stimulated gain or loss to detect. Comparative spectral studies of lipid-rich samples have confirmed that SRS retrieves clean molecular information, whereas CARS images are generally biased by the nonresonant contribution.2PubMed. Molecular orientational order probed by coherent anti-Stokes Raman scattering (CARS) and stimulated Raman scattering (SRS) microscopy: a spectral comparative study The practical upshot is that SRS microscopy images look much closer to what you would see with traditional staining methods, which makes them easier for pathologists and biologists to interpret.
Compared with spontaneous Raman microscopy, SRS is dramatically faster. One demonstration showed that mapping microplastic particles by SRS was roughly a thousand times faster than conventional Raman, simply because the stimulated process generates a much stronger signal per unit time.3Journal of Raman Spectroscopy. Fast microplastics identification with stimulated Raman scattering microscopy Speed gains of that magnitude turn what would be an overnight mapping experiment into something you can do in seconds or minutes.
Detecting Brain Tumors Without Staining
One of the most compelling biomedical applications of SRS microscopy is in neurosurgery. When a surgeon removes a brain tumor, the central challenge is knowing where the tumor ends and healthy brain begins. The standard method involves removing a tissue sample, freezing it, slicing it, staining it, and having a pathologist examine it under a microscope. That process takes time during an operation where every minute matters.
SRS offers a label-free alternative. Because tumor tissue and healthy brain tissue differ in their protein-to-lipid ratios, cellularity, and axonal density, SRS microscopy can distinguish between them based on their vibrational signatures alone. Early work demonstrated this principle in mouse models of human glioblastoma, showing that SRS could differentiate tumor-infiltrated brain from normal tissue without any staining or sample preparation.4PubMed Central. Rapid, label-free detection of brain tumors with stimulated Raman scattering microscopy
A follow-up study moved this into human tissue. Researchers imaged fresh, unprocessed surgical specimens from 22 neurosurgery patients and found that SRS detection of tumor infiltration agreed nearly perfectly with traditional stained-slide histology. By building a classifier based on the chemical information in SRS images, they achieved better than 97% sensitivity and 98% specificity for detecting tumor infiltration.5PubMed Central. Detection of human brain tumor infiltration with quantitative stimulated Raman scattering microscopy Those numbers are high enough to be clinically useful, and the technique works on tissue that has not been frozen, sliced, or stained, which means it could potentially be done during the surgery itself.
Watching Metabolism in Living Cells
Traditional methods for tracking what cells are doing metabolically often require fluorescent labels, which can be bulky enough to alter the behavior of the molecules they are attached to. SRS opens a different route: you can feed cells amino acids that contain deuterium (a heavier form of hydrogen) and then image where those amino acids end up. Carbon-deuterium bonds vibrate at a frequency that nothing else in a living cell matches, creating a signal in what researchers call the “cell-silent” region of the Raman spectrum, a frequency range where biological tissue produces essentially no background.6PubMed Central. Vibrational imaging of newly synthesized proteins in live cells by stimulated Raman scattering microscopy
This approach has been extended to whole organisms. By feeding deuterium-labeled nutrients to living animals, researchers can track where newly synthesized proteins accumulate in different tissues, providing a window into metabolism that does not require killing or fixing the sample.7PubMed Central. Imaging Complex Protein Metabolism in Live Organisms by Stimulated Raman Scattering Microscopy with Isotope Labeling More targeted versions of this strategy focus on individual amino acids. Deuterium-labeled methionine, for example, has been imaged as it enters living HeLa cells, with the deuterium label proving less disruptive to normal uptake than some of the chemical analogues previously used to track the same amino acid.8PubMed. Probing Methionine Uptake in Live Cells by Deuterium Labeling and Stimulated Raman Scattering
Drug Uptake and Resistance in Tumor Models
Understanding why some cancer cells resist drugs is one of the hardest problems in oncology. Part of the answer may lie in how cells physically take up the drug, and SRS provides a way to see that happening. By tuning to the carbon-deuterium or other distinctive vibrations in a drug molecule, researchers can image exactly how much of the drug ends up inside cells and where within the cell it concentrates.
A recent study used SRS to compare drug uptake in flat, single-layer cell cultures (the standard laboratory setup) with uptake in three-dimensional spheroid cultures that more closely mimic real tumors. The goal was to understand whether differences in how much drug actually gets inside cells could explain why 3D cultures often appear more drug-resistant than 2D ones.9PubMed Central. Assessing Drug Uptake and Response Differences in 2D and 3D Cellular Environments Using Stimulated Raman Scattering Microscopy This kind of spatially resolved, label-free pharmacology would be essentially impossible with conventional absorption-based assays, which average over millions of cells and tell you nothing about distribution.
Seeing 24 Colors at Once
Fluorescence microscopy, the workhorse of modern cell biology, runs into a hard limit: fluorescent dyes have broad emission spectra that overlap, making it difficult to image more than about four or five targets simultaneously. Raman peaks are far narrower, which means more labels can fit into the same spectral space without stepping on each other.
Researchers have exploited this by creating palettes of small-molecule dyes that each carry a distinctive bond vibrating in the cell-silent Raman window. A landmark study demonstrated 24 simultaneously resolvable colors by combining custom Raman dye probes with standard fluorescent labels, using SRS under conditions that dramatically boosted the sensitivity of each probe down to concentrations around 250 nanomolar.10Nature. Super-multiplex vibrational imaging Applied to neuronal co-cultures and brain tissue, the approach revealed cell-type-specific differences in DNA and protein metabolism that would have been invisible with conventional fluorescence.
Parallel efforts have pushed multiplexing into thick tissue. One method combined an expanded set of Raman dyes with a tissue-clearing protocol, enabling imaging of up to 11 protein targets through millimeter-thick brain slices, extending the usable imaging depth by ten- to a hundredfold compared to earlier multiplexed protein imaging.11PubMed Central. Highly-multiplexed volumetric mapping with Raman dye imaging and tissue clearing Another group developed a 14-probe panel read out by confocal Raman micro-spectroscopy at a throughput of about 3,600 cells per hour, fast enough for practical single-cell profiling experiments.12Nature Communications. Multiplexed live-cell profiling with Raman probes
Finding Micro- and Nanoplastics
Plastic pollution research has a measurement problem. Particles smaller than a few micrometers are very hard to identify chemically using standard Raman or infrared spectroscopy because the signal is weak and the scan is slow. SRS microscopy has emerged as a powerful solution, leveraging its speed and chemical specificity to identify common plastic polymers particle by particle.
An early proof of concept identified five high-production-volume polymer types in microplastics extracted from environmental and consumer-product samples, including polyethylene terephthalate particles from a commercial personal care product, at roughly a thousand times the mapping speed of conventional Raman.3Journal of Raman Spectroscopy. Fast microplastics identification with stimulated Raman scattering microscopy More recent work has pushed the size limit below 100 nanometers, a regime where particles are too small for most optical identification methods. Using a hyperspectral SRS platform with an automated identification algorithm, one group successfully detected and chemically identified nanoplastics from major polymer types in bottled water.13PubMed Central. Rapid single-particle chemical imaging of nanoplastics by SRS microscopy
Three-dimensional imaging adds another dimension, literally. By stepping through a sample along the vertical axis, SRS can reconstruct the spatial distribution and surface topography of mixed microplastic samples, distinguishing different polymer types by their vibrational spectra at each depth slice.14Cell Reports Physical Science. Fast detection and 3D imaging of nanoplastics and microplastics by stimulated Raman scattering microscopy For environmental scientists trying to understand where plastics accumulate and what they are made of, this represents a major step up from the painstaking particle-by-particle analysis that older methods required.
Capturing Chemical Reactions on Femtosecond Timescales
SRS is not limited to microscopy. A variant called femtosecond stimulated Raman spectroscopy, or FSRS, captures vibrational snapshots of molecules as they undergo ultrafast chemical transformations. The method achieves time resolution down to about 50 femtoseconds (a femtosecond is a millionth of a billionth of a second) while maintaining enough spectral resolution to tell different vibrations apart.15PubMed. Femtosecond stimulated Raman spectroscopy Unlike purely electronic spectroscopies that tell you about energy levels, FSRS tells you about bond lengths and angles, revealing how a molecule’s structure actually changes during a reaction.
One striking demonstration tracked the structural evolution of a light-driven molecular rotary motor, a synthetic molecule designed to spin when hit with light. FSRS revealed that after the initial absorption of light, the molecule relaxes into an excited state where the central bond weakens and lengthens, lowering the barrier to rotation. Product formation was observed in real time.16PubMed. Ultrafast Dynamics in Light-Driven Molecular Rotary Motors Probed by Femtosecond Stimulated Raman Spectroscopy This level of structural detail during a reaction that finishes in trillionths of a second would be inaccessible to essentially any other laboratory technique.
FSRS does have its own artifacts. Cross-phase modulation, a distortion that arises when the intense pump pulse alters the refractive index of the sample and warps the probe beam, can contaminate spectra. While the technique is free of the nonresonant background that plagues CARS, managing cross-phase modulation artifacts requires careful experimental design and post-processing.17Journal of Raman Spectroscopy. Mitigating cross‐phase modulation artifacts in femtosecond stimulated Raman scattering
Fiber Amplifiers and Telecommunications
Outside the microscopy world, SRS has a long history in fiber optics. When a powerful signal travels through an optical fiber, SRS can transfer energy from shorter-wavelength light to longer-wavelength light, amplifying one channel at the expense of another. This was originally seen as a nuisance in long-haul telecommunications, but engineers turned it into a tool: fiber Raman amplifiers use the effect deliberately to boost signals over wavelength ranges that traditional erbium-doped amplifiers cannot cover. Fiber Raman lasers exploit the same principle to generate laser light at wavelengths that are otherwise difficult to reach.18PubMed Central. Fiber Amplifiers and Fiber Lasers Based on Stimulated Raman Scattering: A Review In both cases, the underlying physics is identical to what happens under a microscope: a strong pump field drives vibrational excitation in the fiber material, and the Stokes-shifted light gets amplified as a result.
Mapping Plant Cell Walls
Bioenergy researchers need to understand the chemical architecture of plant cell walls because the arrangement of cellulose, hemicellulose, and lignin determines how easily the plant material can be broken down into fermentable sugars. SRS microscopy can resolve these components in situ without staining. Work on corn stover (the stalks and leaves left after harvest) showed that specific Raman bands are sensitive to xylan, a major type of hemicellulose, and that SRS could map xylan distribution in both two and three dimensions with surprising detail when combined with enzymatic digestion steps.19Springer Nature (Biotechnology for Biofuels). In situ label-free imaging of hemicellulose in plant cell walls using stimulated Raman scattering microscopy Understanding where xylan sits relative to cellulose and lignin could help guide efforts to engineer crops that are easier to convert into biofuel.
Bringing SRS Into the Operating Room
All the biomedical promise of SRS means little if the instrument has to stay in a specialized optics lab. Two engineering developments are pushing it toward clinical use. The first is a handheld SRS microscope that delivers laser light through an optical fiber, enabling the kind of in situ chemical imaging that would be needed during surgery or at a patient’s bedside.20ACS Photonics. In Vivo and in Situ Spectroscopic Imaging by a Handheld Stimulated Raman Scattering Microscope The second is a flexible coherent Raman endoscope, just 4.2 mm in outer diameter, that can potentially image tissue inside the body without requiring an open surgical field.21Light: Science & Applications. High-resolution multimodal flexible coherent Raman endoscope Neither device matches the image quality of a benchtop research microscope yet, but they represent real steps toward the goal of real-time, label-free histopathology at the point of care.
Deeper Imaging and Longer Wavelengths
Standard SRS microscopes typically use wavelengths around 800 and 1,064 nanometers. At those wavelengths, scattering by tissue limits useful imaging depth to roughly a hundred micrometers or so in dense samples. Switching to longer wavelengths above 1,500 nanometers reduces scattering and allows deeper penetration, though at the cost of greater engineering complexity. A purpose-built system using the idler output of an optical parametric oscillator demonstrated improved depth penetration in both plastic test materials and biological tissue, imaging lipid structures at depths that the standard wavelength configuration could not reach.22Journal of Raman Spectroscopy. Stimulated Raman scattering microscopy with long wavelengths for improved imaging depth For applications like skin imaging or intraoperative brain scanning, where you need to see below the surface without cutting, that extra depth is valuable.
Machine Learning Meets Vibrational Imaging
Raw SRS data can be noisy, especially when speed is prioritized or when the target molecules are present at low concentrations. Deep learning has become an increasingly important companion to SRS microscopy, addressing several of its practical limitations at once. Neural networks trained on SRS image data can suppress noise to recover clean images from low-power scans, push effective resolution beyond the optical diffraction limit through computational deconvolution, and untangle overlapping spectral contributions in hyperspectral data sets so that closely related molecules can be distinguished more reliably.23Advanced Intelligent Discovery. Deep Learning‐Assisted Coherent Raman Scattering Microscopy In brain tumor detection, for instance, the classifier that achieved better than 97% sensitivity was built on quantitative features extracted from SRS images.5PubMed Central. Detection of human brain tumor infiltration with quantitative stimulated Raman scattering microscopy As these computational tools mature, they may lower the barrier to deploying SRS in settings where trained spectroscopists are not available to interpret the raw data, which is most clinical environments.