- The SRS signal is detected in transmitted-light mode. Therefore, as a rule of thumb, specimens should be at least semitransparent.
- Thicker or less transparent samples should be sectioned to a suitable thickness, whenever the experiment allows for it. In our experience, 50-100 micrometer thick slices work well. Alternatively, thick samples can be investigated using CARS microscopy which can be detected in both transmitted- and incident-light (epi-illumination) modes.
- Specimens can be fresh or formalin fixed. Methanol fixation is discouraged for lipid related analysis, because this method is known to extract lipids from the specimen. When studying molecules other than lipids, the signal from the lipid content might be disturbing, so for that reason methanol fixation could be beneficial.
- For fresh samples, care has to be taken during long spectroscopy scans as it may be necessary to identify potential motion artifacts.
- Specimens can be sandwiched between a coverslip and a microscope slide. In contrast to spontaneous Raman scattering, signals from out-of-focus regions are highly suppressed by nonlinear optical signal generation. Hence, fluorescence background signals from the substrate and other components in the beam path are virtually nonexistent. Also, there are no special requirements for the substrate material. The position of the object in the sandwich construction should be within the working distance of both the objective and condenser (see below). Thus, ideally, the object should be in direct contact with the surface of the cover glass and microscope slide.
- It is preferred to mount samples in aqueous solution or buffer, e.g., sandwiched using double-sided sticky spacers. When necessary, samples can be embedded (typical fluorescence mounting media like Mowiol or Dako, and hydrogels such as Matrigel or Agarose should work). Paraffin embedded sections can be analyzed with or without removal of the paraffin. However, we strongly recommend removing it.
- Alternatively, specimens in a glass-bottom petri dish can be imaged, e.g., with the TCS SP8 CARS microscope using a S1 oil condenser in a water dipping configuration.
- In many cases, even fluorescently labeled specimens can be probed by SRS, without fluorescence cross-talk into the SRS channel. Therefore, it is possible to correlate fluorescence images with SRS signals. The immunity of SRS to fluorescent signals arises from the lock-in-based detection scheme coupled with the very-much red-shifted detection band. In rare cases when fluorescent contamination signals are present, often they can be separated from true SRS signals in a spectral analysis (fluorescence signals will have a weak spectral dependence, if any, whereas SRS signals will show pronounced spectral peaks).
A suitable choice of objective is crucial for CARS/SRS, as the Pump/Stokes pulses must overlap in 3D space and time to achieve optimal signal strength. Suggested objectives are:
- 10x NA=0.4 Air HC PL