How Coherent Anti-Stokes Raman Spectroscopy (CARS) Works

By Samudrapom DamReviewed by Susha Cheriyedath, M.Sc.Oct 29 2026

Importance of CARS
An Overview of CARS
Working of CARS
Suppressing FWM NRB
Enhancement of the Pure Coherent Raman Spectra
Applications
The Way Forward
References and Further Reading

Coherent anti-Stokes Raman spectroscopy (CARS) is a type of Raman spectroscopy that relies on the nonlinear conversion of two laser beams into a coherent, high-intensity anti-Stokes Raman beam. This nonlinear optical method is extensively used for molecular characterization and vibrational imaging in fields like medicine, biology, materials science, and chemistry.^1-4

Image Credit: suyu2025/Shutterstock

Importance of CARS

Spontaneous Raman (SR) spectroscopy, a commonly used vibrational spectroscopic method, analyzes the inelastic scattering of light to provide detailed insights into molecular vibrations.²

However, the inherently low scattering cross-section in SR spectroscopy leads to weak signal intensities, necessitating long acquisition times. Additionally, the signal's incoherence reduces its sensitivity, and Raman signals can be obscured by fluorescence interference.²

SR also has limited depth penetration, which further limits its usage in probing thicker samples. To address these issues, coherent Raman scattering (CRS) methods have been developed. Among them, CARS has gained prominence owing to its effectiveness in imaging and spectroscopic applications.^2,3

An Overview of CARS

In CARS, signal generation is achieved via four-wave mixing (FWM), in which photons from a Stokes beam, a probe beam, and a pump beam interact with the sample. This interaction generates coherent anti-Stokes photons. However, the probe beam has been a part of the pump beam in several instances.²

The CARS signal intensity shows high sensitivity to the vibrational modes of the sample and reaches its maximum when the frequency difference between the pump and Stokes beams matches the resonant frequency of the probed vibrational mode.²

In broadband CARS (BCARS), all vibrational modes within the fingerprint and CH band are probed simultaneously using a broad Stokes pulse. Thus, a comprehensive BCARS spectrum can be acquired, making BCARS an effective tool for chemical analysis. The vibrational specificity and improved signal strength of BCARS make it useful across diverse fields.²

Working of CARS

CARS utilizes a broadband excitation in combination with a narrowband delayed probe pulse. The use of ultrafast laser pulses is crucial in CARS. Typically, Stokes and pump pulses excite molecular Raman vibrations, and then a third pulse, or probe pulse, scatters off these molecules as they vibrate coherently.⁴

The scattered photon energy decreases or increases by an amount of the Raman frequency. In coherent Stokes Raman scattering, the scattered light is red-shifted, while in coherent anti-Stokes Raman scattering, the scattered light wavelength is blue-shifted.⁴

Simultaneously, both Stokes and pump pulses prepare and excite molecules in a coherent state. With or without delay, the probe pulse arrives, generating an anti-Stokes/Stokes signal.⁴

However, FWM contamination remains inevitable. FWM is a third-order nonlinear process that does not rely on molecular Raman vibrations. As non-resonant background (NRB) noise, FWM is added to and mixed with the overall measured spectral data. Thus, extracting pure resonant CARS information from the overall FWM-contaminated data is crucial.⁴

Suppressing FWM NRB

While several methods have been developed to suppress FWM NRB, few methods that use broadband pump and Stokes excitation pulses with a narrowband-shaped and delayed probe pulse have been successfully demonstrated.⁴

Broadband Stokes and pump pulses excite multiple Raman vibrations simultaneously, providing fingerprint signatures recorded at the single-laser-shot level, which is advantageous for detecting rapidly moving bacterial spores.⁴

Download the PDF of this page here

The use of a narrowband probe provides sufficient spectral resolution for the target material’s Raman spectra. Most Raman linewidths are less than 10 wavenumbers. Highly suppressed FWM can be achieved through delayed probing relative to the excitation.⁴

Frequency- and time-domain-specific probe shapes have been adapted, including Gaussian-Gaussian, exponential-Lorentzian, square-sinc, and sinc-square forms.⁴

Although these studies displayed sufficient suppression of unwanted FWM to some degree only at fixed time delays, extracting the pure CARS contribution across the entire spectrogram remains difficult for arbitrary probe delays.⁴

Computational advances played a significant role in NRB removal. For instance, developments in the Kramers–Kronig (KK) approach have made it an effective framework for modern deep learning (DL) algorithms, including generative adversarial networks (GANs), long short-term memory networks (LSTMs), and convolutional neural networks (CNNs).²

These algorithms have demonstrated significant potential to enhance the efficiency, robustness, and accuracy of NRB removal and phase retrieval. While these advances address several limitations of conventional methods, challenges such as interpretability, model generalizability, and the need for large training datasets still persist.²

Enhancement of the Pure Coherent Raman Spectra

At a positive delay, the CARS signal is maximized, and this effect becomes more pronounced with a longer probe pulse. Thus, the threshold delay at which the CARS signal is maximized relies on the molecular decoherence time and the duration of the probe pulse.⁴

The threshold delay time is measured as the stationary point of the Faddeeva function. Positions of CARS maxima shift away from zero delay as the probe width decreases, whereas FWM maxima remain at zero delay.⁴

Applications

CARS was initially developed to investigate reaction flows and gas dynamics in combustion systems owing to its high sensitivity to molecular vibrations. Over time, advancements in sensitivity, spectral resolution, signal-to-noise ratio (SNR), detection methods, instrumentation, and signal processing have expanded its applications.²

In materials science, CARS enables the characterization of molecular composition and structure, supporting the development of novel materials. In chemical analysis, BCARS enables detailed investigation of molecular vibrations and chemical bonds, allowing accurate identification and quantification of substances.²

Moreover, BCARS has gained prominence in biomedical imaging, providing noninvasive, high-contrast, high-resolution visualization of biological tissues and cells, thereby supporting advanced diagnostic techniques and biomedical research applications.²

The Way Forward

CARS has emerged as a powerful technique for molecular characterization, chemical analysis, and biomedical imaging, offering high sensitivity, rapid acquisition, and detailed vibrational information. However, NRB interference remains a major challenge in obtaining pure Raman spectra.

Future research must focus on improving background suppression by developing robust computational approaches, including DL algorithms. Addressing challenges related to interpretability, generalizability, and training data will further improve the accuracy, reliability, and applicability of CARS across diverse scientific fields.

References and Further Reading

1. Tolles, W. M., Nibler, J. W., McDonald, J. R., & Harvey, A. B. (1977). A review of the theory and application of coherent anti-Stokes Raman spectroscopy (CARS). Applied Spectroscopy, 31(4), 253-271. DOI:10.1366/000370277774463625, https://www.researchgate.net/publication/237654885_A_Review_of_the_Theory_and_Application_of_Coherent_Anti-Stokes_Raman_Spectroscopy_CARS
2. Junjuri, R., & Bocklitz, T. (2025). Review of Coherent Anti-Stokes Raman Scattering Nonresonant Background Removal and Phase Retrieval Approaches: From Experimental Methods to Deep Learning Algorithms. Advanced Photonics Research, 6(9), 2500035. DOI: 10.1002/adpr.202500035, https://advanced.onlinelibrary.wiley.com/doi/full/10.1002/adpr.202500035
3. Inoue, K., & Okuno, M. (2025). Coherent Anti-Stokes Hyper-Raman Spectroscopy. Nature Communications, 16(1), 306. DOI: 10.1038/s41467-024-55507-0, https://www.nature.com/articles/s41467-024-55507-0
4. Ariunbold, G. O., & Altangerel, N. (2016). Coherent anti-Stokes Raman spectroscopy: Understanding the essentials. Coherent Phenomena, 3(1), 6-17. DOI:10.1515/coph-2016-0002, https://www.researchgate.net/publication/313534558_Coherent_anti-Stokes_Raman_spectroscopy_Understanding_the_essentials

Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.