INTRODUCTION Raman Spectroscopy is a widely used and versatile form of analysis used throughout biology, chemistry and solid-state physics (1) (2) (3). Like all spectroscopic techniques it observes how light interacts with a sample in order to understand its properties. This can then be used to, for example, identify an unknown sample, monitor how stress affects a crystal structure or look for impurities. Raman Spectroscopy is not the analysis of light absorbed, emitted, or elastically scattered from a sample, but rather the inelastically scattered light.
When an object elastically scatters, it retains its original energy. Elastic scattering between light and matter is called Rayleigh scattering. It accounts for most scattering events seen between a light source and any given material. However, as discovered by C.V. Raman in 1928 (4), light can undergo inelastic scattering with matter. In this case the photon of light involved has a different energy before and after the scattering event. Having either lost (known as a Stokes shift) or gained (an anti-Stokes shift) energy from the matter it interacted with. This change of energy can be measured as a change in the wavelength (colour) of the light (Figure 1)
Fig. 1. Illustrative example of Raman and Rayleigh scattering events. We can see Stokes shifts give lower energy photons whereas anti-Stokes give higher. Anti-Stokes shifts are rarer as the system needs to already be in an excited state.
The energy for this wavelength shift comes from a change in the energy state of a molecular bond or bonds. This is distinct from an interaction where the photon is absorbed by an atom and then re-emitted at a different wavelength, which is the domain of fluorescence spectroscopy. This means the wavelength shift in the Raman scattered light corresponds directly to the current energy states of the molecular bonds in the sample (5). As these are influenced not just by the atoms involved in those bonds but the total crystal structure and the strain the system is under, it is possible to interpret significant information from a Raman spectrum that can be difficult to obtain by other means.
ANALYSIS A significant advantage of Raman is that it is non-invasive. It does not require any chemical tags or dyes to be inserted into a sample, so often the sample is left unchanged after measurement. Raman can also be done without contacting the sample - helpful when identifying potentially dangerous materials.
There are, however, limitations to Raman spectroscopy. The likelihood of Raman scatter compared to a Rayleigh scatter is approximately 1 in 106 (6). This means the Raman signal tends to be much weaker than the incident light and so measurements must filter out the much brighter Rayleigh scattered light. This is normally achieved with a sufficiently sharp dielectric bandpass, notch or edge filter.
To achieve high signal to noise ratio an intense light source is needed, whereas the maximum spectral resolution of the spectrum is dependent on the bandwidth of the source. As the energy shift caused by two different molecular bond states can be extremely small, these parameters can be critical.
LASERS FOR RAMAN SPECTROSCOPY Choosing the right laser for Raman spectroscopy is critical. Firstly, the intensity of Raman scattered light is proportional to 1 /𝜆^4 (7), which means a much stronger signal will be gained from using shorter wavelengths. Conversely, many materials exhibit strong fluorescence when illuminated by intense light in the visible and UV (8), which can swamp the Raman signal, as well as increasing the risk of sample damage. This means that many wavelengths from the near IR though the visible to the UV are routinely used for Raman, dependent on the sample to be measured.
The spectral purity, measured as the ratio of the intensity of the main emission peak compared to any sidebands or the noise floor, is critical for resolving small Stokes shifts. If Raman scattering from the main emission peak shifts into the wavelength range of a sideband, it can be swamped by the Rayleigh scattering of this wavelength.
Fig. 2. Spectrum analyser trace from UniKLaser DPSS systems, showing an absence of side modes and inherent pure single frequency operation.
Three main categories of lasers are used for Raman Spectroscopy; Distributed Feedback (DFB) Diode lasers, Volume Bragg Grating (VBG) stabilised Diode Lasers, and Diode Pumped Solid State (DPSS) lasers. Whilst diode lasers often have general advantages in terms of cost, efficiency and overall footprint (9), DFB lasers are limited in both power and available wavelengths (10).
VBG stabilised diode lasers can operate at shorter wavelengths and higher powers, however they sacrifice linewidth and spectral purity (11). This makes them unsuitable for high resolution Raman.
DPSS lasers, however, can operate throughout the relevant spectrum with much narrower linewidths and better spectral purity (12), with some compromise on cost and footprint. DPSS systems will operate without side bands (Figure 2), a noise floor 60 dB below the emission peak and a linewidth often significantly below 1 MHz.
Furthermore, DPSS lasers exhibit high power and spectral stability, allowing measurements over a longer period at high resolution, which is critical for Raman imaging applications.
Fig. 3. Monitoring the power and wavelength of one of our 1W 640nm systems over 8 hours. The power remained within ±1% and the wavelength within ±1pm.
SUMMARY DPSS lasers are essential for Raman analysis, in particular high resolution spectroscopy, due to their narrow linewidths, high spectral purity, high power and spectral stability and higher output power.
1. Raman spectroscopy in solid state physics and material science. Theory, techniques and applications. Lucazeau, G. and Abello, L. 11, s.l. : Analusis, 1995, Analusis, Vol. 23, pp. 301-311. 27034645.
2. McCreery, R. L. Raman Spectroscopy for Chemical Analysis. s.l. : John Wiley & Sons, 2005. 9780471231875.
3. Raman Spectroscopy of Biological Tissues. Movasaghi, Z., Rehman, S. and Rehman, I. U. 5, s.l. : Applied Spectroscopy Reviews, Vol. 42. 10.1080/05704920701551530.
4. The Negative Absorption of Radiation. Raman, C. and Krishnan, K. 3062, s.l. : Nature, 1928, Vol. 122. 10.1038/122012b0.
5. Laser-excited Resonance Raman Spectra of Small Molecules and Ions---A Review. Kiefer, W. 2, s.l. : Appl. Spectrosc., 1974, Vol. 28.
6. Koenig, Jack L. Spectroscopy of Polymers (Second Edition). s.l. : Elsevier Science, 1999. "978-0-444-10031-3.
7. Wavelength dependence of the preresonance Raman cross sections of CH3CN, SO42−, ClO4−, and NO3−. Dudik, John M., Johnson, Craig R. and Asher, Sanford A. 4, s.l. : The Journal of Chemical Physics, 1985, Vol. 82. 10.1063/1.448405.
8. Lakowicz, Joseph R. Principles of Fluorescence Spectroscopy. s.l. : Springer US, 1999. 978-1-4757-3061-6.
9. Chirp characteristics of 40-gb/s directly Modulated distributed-feedback laser diodes. Miyamoto, Y., Sato, K. and Kuwahara, S. 11, s.l. : Journal of Lightwave Technology, 2005, Vol. 23. 1558-2213.
10. Design and Realization of High-power DFB Lasers. Wenzel, Hans, et al. s.l. : Proceedings of SPIE - The International Society for Optical Engineering, 2004. 10.1117/12.569039.
11. Spectral Stabilization of High Efficiency Diode Bars by External Bragg Resonator. Glebov, Leonid and Venus, George. 1, s.l. : Journal of Laser Applications, 2005, Vol. 2005. 10.2351/1.5060495.
12. Spectral density contrast in DPSS and ECD lasers for quantum and other narrow-linewidth applications. Szutor, B. and Karpushko, F. s.l. : SPIE, 2020. 10.1117/12.2565011.