Superresolving Optical Microscopy

Light microscopy is a widely used analytical tool because it provides non-destructive, real-time, three-dimensional imaging with chemical and material specific contrast. Despite microscopy advances in detection, identification, and manipulation, today's demands on chemical imaging have grown beyond current capabilities. The principal focus of this project is to establish the scientific and metrology underpinnings necessary for the realization and operation of superresolution light microscopy in real world applications. We have designed and fabricated a flexible superresolution optical microscopy platform that combined with vector point spread function engineering, represents the state-of-the-art in optical diagnostics for in-situ characterization of organic and biological materials.

We have developed a superresolving coherent anti-Stokes Raman spectroscopy (CARS) microscope. CARS is a third-order nonlinear optical process which in our application mixes two laser beams together to greatly enhance the sensitivity of Raman spectroscopy and imaging, typically by a factor of 10⁴ to 10⁵.  The technique is useful and popular for this reason alone, but the two color nature of its excitation can be used to greatly improve the microscope's spatial resolution by using two separate pupil phase masks. Pupil-phase masks were first discussed in 1952 by theoretician Toraldo di Francia. The focal spot is produced by a converging spherical wave, and a pupil phase mask introduces a relative delay to a portion of that wavefront. This delay causes a portion of the wavefront to destructively interfere in some regions of the focal spot, and constructively interfere elsewhere. By adjusting the position and magnitude of the delay, these interference effects can be controlled to narrow the focal spot.  Figure 1 shows how superresolving CARS can be implemented using two spatial light modulators (SLM's). The resulting point-spread functions for a simple design were measured using two photon photoluminescence from gold nanoparticles. The engineered Pump beam focus has been split into an hour-glass figure by the phase mask, with a narrow region designed to intersect with the Stokes beam focus. It is at this intersection/overlap where the CARS signal will be generated.

 

 

Figure 1. The experimental schematic shows how two spatial light modulator phase masks are used to independently reshape the focal field distribution of the Pump beam and the Stokes Beam. Shown on the right are the resulting measured point-spread functions (taken through the y = 0 plane) for each beam using gold nanoparticle emitters.

                                       (A)                               (B)

 

Figure 2. (A) Conventional CARS image of 200 nm polystyrene beads, taken at a frequency of 1005 cm-1 (c-c stretch). (B) Superresolving CARS image of the same area, showing the improved spatial resolution.

The superresolved CARS imaging was demonstrated by imaging 200 nm polystyrene beads. The laser system uses two synchronized, mode-locked Ti:Sapphire picosecond lasers for CARS excitation. The Pump laser wavelength was set at 785 nm, and the Stokes to 852 nm, giving a relative energy separation of 1005 cm^-1, which corresponds to the in plane carbon-carbon vibrational band. Figure 2(A) shows an 80 micrometer scan of a film of the beads using a conventional focus. Figure 2(B) shows the same region of the sample, which has been reimaged using a superresolving focus. The superresolving image demonstrates the ability to resolve individual beads that were not resolved in 2(A). 

• Demonstrate SLM based superresolving fluorescence microscope achieving true 78 nm resolution, with anticipated key applications being bioscience and heath.
• Demonstrate SLM based superresolution non-linear phase filters with via vibrationally-resolved chemical (CARS) and two-photon fluorescence microscopies.