Label-free nanometer-resolution imaging of biological architectures through surface enhanced raman scattering

Label free imaging of the chemical environment of biological specimens would readily bridge the supramolecular and the cellular scales, if a chemical fingerprint technique such as Raman scattering can be coupled with super resolution imaging. We demonstrate the possibility of label-free super-resolution Raman imaging, by applying stochastic reconstruction to temporal fluctuations of the surface enhanced Raman scattering (SERS) signal which originate from biomolecular layers on large-area plasmonic surfaces with a high and uniform hot-spot density (> 10(11)/cm(2), 20 to 35 nm spacing). A resolution of 20 nm is demonstrated in reconstructed images of self-assembled peptide network and fibrilated lamellipodia of cardiomyocytes. Blink rate density is observed to be proportional to the excitation intensity and at high excitation densities (> 10 kW/cm(2)) blinking is accompanied by molecular breakdown. However, at low powers, simultaneous Raman measurements show that SERS can provide sufficient blink rates required for image reconstruction without completely damaging the chemical structure.