Spontaneous Raman scattering generally yields weak signals, resulting in limited sensitivity. The same behavior is observed with TERS. Even with a great signal amplification that can be up to 106, the cross-section remains low and the signal to noise remains below 100 with a maximum of thousands of counts per second at the detector.
A signal improvement by many orders of magnitude can be obtained by Coherent Raman Scattering (CRS), which requires two synchronized pulses, the Pump (frequency ωp) and Stokes (frequency ωs), with their frequency difference tuned to a Raman-active molecular vibration Ω =ωs−ωp in order to drive it coherently27. In Coherent Antistokes Raman scattering (CARS)28-30, one detects the Anti-Stokes signal at ωas=2ωp−ωs, while in Stimulated Raman Scattering (SRS) one measures amplification of the Stokes pulse (and simultaneous depletion of the Pump) via the stimulated emission from a virtual state to the probed vibrational state. CARS and SRS, each with its advantages and drawbacks, have allowed vibrational imaging with unprecedented speed, up to video rate, but so far only in the far field, with diffraction-limited spatial resolution of ≈ 200-300 nm.
Developments of TE-CRS31 (Tip Enhanced Coherent Raman Scattering) is particularly interesting since the signal enhancement provided by the TERS tip is expected to scale approximately with the local field-enhancement factor. Stimulated emission in Tip Enhanced Raman Spectroscopy has been demonstrated recently; thanks to the achieved billion fold stimulated gain over conventional TERS, the STERS (Stimulated TERS)32 technique opens venues for ultra fast imaging applications in the TERS field through an enormous improvement of image contrast and detection sensitivity.