Optical Superresolution Microscopy (Nanoscopy)

光谱学延伸到“衍射极限”以下（约三分之一）

The extension of optical spectroscopy below the "diffraction limit" (about a third of the wavelength of light; e.g.,230nm) has been realized in recent years by two different classes of microscope: "PALM/STORM" and "RESOLFT/STED". The former recreates a biological scene in a 'pontillist' manner; centers of individual fluorescent 'paint' dots are located with 20nm precision on the scene, a few at a time, until the full picture emerges. It is precise but painstakingly slow. The second method, STED (STimulated Emission Depletion) superposes the normal spot illuminating the scene with another diffuse "donut" beam whose job is to erase fluorescence around the edges. This leaves a smaller spot at the center of the donut to sweep across the image, revealing it in 50nm detail. Both sorts of microscope are commercially available. The PALM version is inexpensive but slow, best for acquiring still images. The STED version (over $1.1M) has the potential for video nanoscopy but applies large laser powers (in the "erase" donut beam) that damages living cells. Most of our nanoscopy effort is devoted to STED and STED-like methods. We have constructed our own STED microscope around existing CARS lasers and FCS detection electronics (from other prior projects). We have designed (and provisionally patented) an 'azicon' (azimuthal polarizer axicon) to make the central spot of the donut beam very dark (preserving central brightness in the image, allowing for stronger erase beams and hence finer resolution). For widespread STED use, we developed a general calibration scheme for STED dyes that enables nanoscopists to compensate (during data processing) for the quirks in their individual optics or lasers. The "Saturation Intensity" calibration manuscript was published previously. In recent years we have designed, patented, and begun testing a new class of fluorescent dyes that provide two key features: 1. lower power requirements for erase beam. This allows finer resolution and longer observations in living cells, making video nanoscopy more practical. 2. Simultaneous multicolor erase beam. STED had previously been limited to two colors, but the mechanism inherent in our dyes expands the available palette. This is important in providing biological context to the image of macromolecules one will paint. Multicolor tubulin fibrils and beads have been imaged in the same, single-frame image. We also began exploiting the nanosecond nature of our dyes to design a microscope using inexpensive diode lasers to achieve our STAQ nanoscopy, both CW and ns-pulsed, and we worked toward adding external "TAQ" antennae to the popular GFP family of fluorescent protein "paint" molecules. The latter are popular because they can be genetically connected to the structure of interest in cells. Finally, combining STAQ or STED with pulse-modulated donut time profiles has been theoretically examined to suggest even lower powers for encoding, and we have designed a "STEN" ("SpatioTemporal Encoding Nanoscopy") microscopy scheme for the new STED microscopes. The software to accomplish this is in early development stages. A further development based on photobleaching/shelving with multiple photons similar to the "PIM" methods of others is in process, requiring our building a multibeam timed illumination instrument, which has been delayed by limited on-site presence. We have also begun building a lock-in amplifier baser SE (Stimulated Emission) microscope whose spectra should help us refine our STAQ method and probes. .