Differential CARS microscopy with chirped femtosecond laser pulses

Optical microscopy is an indispensable tool that is driving progress in cell biology, and is still the only practical means of obtaining spatial and temporal resolution within living cells and tissues. Much effort is being devoted recently to achieve intrinsic three-dimensional (3D) spatial resolution by exploiting optical nonlinear effects which can only take place in the small focal volume where high photon densities are reached. One of the most utilised multiphoton (ie nonlinear) microscopy techniques is two-photon fluorescence where the biomolecules of interest are labelled with fluorophores, which are optically excited via simultaneous absorption of two photons. However, these modified biomolecules raise questions if their behaviour is real or artefactual. Furthermore, all organic fluorophores are prone to photo-bleaching which severely limits time-course observations and is accompanied by toxicity effects and consequent cell damage. Coherent Antistokes Raman Scattering (CARS) microscopy has recently emerged as a new multiphoton microscopy technique which overcomes the need of fluorescent labelling and yet retains biomolecular specificity and intrinsic 3D resolution [1]. We have developed in our laboratory a fully home-built CARS microscope featuring innovative CARS excitation/detection schemes. In particular, we have demonstrated differential-CARS (D-CARS) with strongly suppressed non-resonant background and improved chemical sensitivity [2], and single-laser CARS [3] utilising femtosecond laser pulses linearly chirped by glass dispersion [4]. In D-CARS we replicate the exciting pump- Stokes pulse pairs to create a pulse train at twice the laser repetition rate. By controlling the instantaneous frequency difference of each pair using glass dispersion, we adjust the Raman frequency probed by each pair in an intrinsically stable and cost-effective way. The resulting CARS intensities are detected simultaneously by a single photomultiplier as sum and difference using phase-sensitive frequency filtering, ensuring fast D-CARS acquisition without motion artifacts crucial for living cell microscopy. The method can be extended to more than two pairs by cascading the setup, and we have demonstrated four-pair D-CARS. I will present our latest progress with D-CARS and its application to cell imaging