Tip-enhanced Raman Spectroscopy (TERS)
In this mode, a significant field-enhancement at the tip apex is required to generate enough Raman scattering from the nano-sized sample region directly underneath the tip apex. A few additional constraints are placed on the tool in order to achieve the required field-enhancement:
- the tip needs to be made (or coated) with silver or gold for visible excitation light; ideally the nano-structure of the tip apex is such that the plasmon resonance of the structure matches the excitation wavelength;
- the polarization of the excitation light should be along the tip axis; for top or inverted sources, this is best achieved with radial polarization; for side sources, p-polarization is required;
- the tip needs to be held in close proximity to the sample surface throughout the scanning process, making STM and tuning-fork-based AFM the preferred gap control modalities.
A typical setup for TERS on an inverted microscope. The incident far-field radiation induces a strong plasmonic response in the tip, which in turn creates a field enhancement in the tip-sample junction of several orders of magnitude. The scattered Raman shifted photons are enhanced by the field such that they become detectable. A noble metal tip such as Au or Ag is standard for TERS as they exhibit a strong plasmonic response at optical frequencies. Since the enhancement depends on the tip maintaining close proximity to the sample (<10 nm), SPM modalities that maintain a fixed, close distance between tip and sample are best suited for TERS (e.g. shear-force AFM or STM).
Scattering-type Scanning Near-field Optical Microscopy (s-SNOM)
In s-SNOM, far-field illumination of the tip-sample region results in standing evanescent field, which is scattered into the far field by the AFM tip tapping. Unlike TERS, a plasmonic enhancement is not necessary. The near-field is discerned from the far-field via lock-in detection of the cantilever modulation frequency. By setting up a reference arm to create an interferometer, additional methods such as self-homodyne and pseudoheterodyne can be employed to enhance the near-field contrast.
Schematic of an s-SNOM setup [Al Mohtar et al. Opt. Expr. 22 22232 (2014)].
With conventional s-SNOM using a tapping mode cantilever, it is unavoidable that some portion of the far-field is also modulated in addition to the evanescent field. As a result, the demodulated amplitude and phase from the lock-in do not represent the near-field signal exclusively. Nonlinearity is introduced into the oscillation when the tip strikes the surface, which is also the part of the oscillation responsible for scattering the near-field. As a result, the higher harmonics become populated with the near-field signal. Tapping ‘hard’ with a 20-100 nm amplitude is typically required when doing s-SNOM with a cantilever in order to see any signal at all.
By tapping ‘harder’, the perturbation to a pure sinusoidal oscillation is much more pronounced, and it is that perturbation where the interaction with the evanescent field was strongest. The net effect is that s-SNOM requires somewhat aggressive tapping and high harmonic detection to display any signal. As a result, s-SNOM with a tapping mode cantilever, while it is less invasive than contact mode AFM, it is not necessarily gentle enough for some soft matter samples, as the tapping conditions that are necessary for near-field detection may be too aggressive for the sample.
Note, s-SNOM can be performed with smaller amplitudes (~5 nm) using a quartz tuning fork (QTF) resonator with an etched tip attached. Under this modality, the use of Generalized Lock-in Detection has proven beneficial [Al Mohtar et al. Opt. Expr. 22 22232 (2014)].