Raman scattering was discovered in 1928 by Sir Chandrasekhara Venkata Raman who was awarded the Nobel Prize in Physics in 1930 for his work in the field of light scattering. Light can be scattered by a sample with no change in its wavelength (elastic scattering). This effect is called Rayleigh scattering. On the other hand, Raman scattering occurs when a photon interacts with a molecule and the scattered light has a different wavelength than the incident light. If the scattered photon has less energy than the incident photon, and therefore a longer wavelength, it is called Stokes scattering. If the molecule is in a vibrational state to begin with and after scattering is in its ground state then the scattered photon has more energy, and therefore a shorter wavelength. This is called anti-Stokes scattering. The Raman and Rayleigh phenomena are depicted in the energy diagram shown below .


Raman and Infrared (IR) are both vibrational spectroscopy techniques.  They are non–destructive, non–invasive tools that provide information on the molecular vibrations. Vibrational spectroscopy is very useful for analysis of materials. Each molecule possesses a rich spectrum of vibrational transitions, and its Raman and IR spectra give precisely this information; the vibrational spectrum being somewhat like a “fingerprint” of the molecule. For the vibration to be active in infrared, the vibration has to induce a dipole moment change while in Raman, the vibration has to cause a change in polarizability. These two vibrational spectroscopy technique are complementary to each other. The figure below depicts the typical vibrations in a molecule that are probed by IR and Raman spectroscoy.



Raman and IR absorption are thus very powerful techniques for chemical analysis, but the weakness of each is the strength of the other.  Infrared absorption offers a good sensitivity (σR~10-21 cm2) relative to Raman (σR~10-29 cm2), but this efficiency is mitigated by the poor sensitivities of optical detectors in the infrared region.  Raman operates instead in the visible range (400-800 nm) where detectors of the type Si MCCD are very efficient and sensitive (only a few photons are needed for signal detection).  Moreover, the spatial resolution is poor in infrared and excellent in Raman because the resolution limit depends on wavelength (the limit of resolution is ~λ/2 according to the Rayleigh criterion), which is relatively long for infrared (λ~30 µm) and short for Raman (λ=400-600 nm). Applications with Raman would be ideal, but the problem arises with the cross sections of Raman scattering, which are too weak to be useful in optical imaging or molecular marking.



Since the discovery of Raman scattering in the 1920s, technology has progressed such that Raman spectroscopy is now an extremely powerful technique with many applications. Raman imaging can be obtained using th intrinsic Raman scattering of a sample, a method gaining in popularity because it is label-free and non-destructive. It is of great interest since a complete chemical map can be generated. Raman imaging offer many advantages over other imaging techniques such as very narrow bands facilitating multiplexing, no blinking, very little photobleaching and quantitative signal. However its lack of sensitivity remains a challenge especially for imaging trace amounts of chemicals or biomolecules. To address this issue, Raman labels with amplified Raman signals can be used to detect and image target molecules. Raman labels are an emerging class of nanomaterial that have exceptional properties tailored to Raman imaging applications. 


The sensitivity of various imaging techniques can be compared by considering the cross section required to observe light scattering. In Raman spectroscopy, the intensity, /, (in photons/s/cm2) of scattered light for a molecule is proportional to the scattering cross section per molecule and the intensity of the incident light l0, according to the relation / = OR l0. For Raman spectroscopy, OR is between 10-29 and 10-32 cm2, while the equivalent fluorescence and optical absorption cross sections are on the order of 10-19 to 10-18 and 10-29 to 10-32 cm2, respectively. There are therefore approximately more than twelve orders of magnitude difference between the relative efficiencies of the Raman process and those of optical absorption or fluorescence. Raman spectroscopy benefits, however, from higher laser intensities than optical absorption and fluorescence, which compensate for the low efficiency of the scattering process to make this analytical technique more accessible.  Nevertheless, the low sensitivity remains a problem for Raman imaging.  In addition, the use of high intensity excitation lasers can alter the samples being examined due to localized heating.  In such cases, the acquisition of a Raman image is done by sweeping the light-emitting probe point by point with a reduced intensity to avoid the heating, which makes the acquisition time-consuming and inefficient.  Being much more sensitive, fluorescence and absorption/reflection have been heretofore the techniques of choice for optical imaging. Under specific conditions, certain molecules exhibit an increased Raman sensitivity. Nowadays, with the recent advances in lasers, detectors and Raman instrumentation, a cross section of about 10-21 is required to detect single objects. The graph below lists different phenomena and their corresponding Raman sensitivity.  A π-conjugated molecule such as rodhamine 6G has a cross section of 10-24 cm2, which is still too low for single object detection. Carbon nanotubes and SERS nanoparticles are known for their efficient Raman scattering properties and are often used in Raman imaging. The dyes@SWNT Raman labels are novel hybrid nanomaterials that also exhibit amplified Raman scattering with sensitivities that make single object detection possible.



The research team of Professor Richard Martel of Université de Montréal has developed new Raman labels exhibiting an enhanced Raman signal. An assembly of carbon nanotubes (CNT) and dyes (chromophores) allows for an increase in the Raman response. The cross section of sexithiophene dyes within a carbon nanotube nanotracers 6T@CNTs was measured to  be (3 ± 2) × 10-21 cm2 / sr per label which allows detection of individual label (Gaufrès et al, Nat. Photonics 2014, 8, 72). These Raman labels have unique properties and are characterized by an intense optical response, no photobleaching, outstanding capacity for multiplexing, a nanometric size, high chemical stability and the presence of chemical groups or biotin grafts for subsequent functionalization. With these complexes, the Raman signal is now at a level almost comparable to fluorescence and optical detection  allowing quick and unequivocal identification.