Physics 
Towards nanometer resolution in light microscopy …
 

Common knowledge: We all learned at school that due to the wave nature of light, a light microscope can never separate objects that are closer than about a third of the wavelength. In practice this means a resolution of 200 nm in the focal plane and about 500 nm along the optical axis. First established by Ernst Abbe in 1873, this limit has been paradigmatic ever since. Common belief is that the exploration of the nanoworld requires electron, force or scanning tunneling microscopes. These microscopes have had a tremendous impact in the natural sciences, indeed, including in biology. However, that these modern microscopes have their shortcomings, as well. They are either confined to specimen surfaces or are not compatible with live cells. As a matter of fact, focused light is the only means by which it is possible to explore a whole cell in 3D non-invasively.

The gold standard and beyond: The best standard 3D-light microscopes are the scanning confocal and the multiphoton fluorescence microscope. Both employ sharply focused lasers, sensitive detection and above all, their effective size of the focal spot depends quadratically on the illumination intensity. Nevertheless, their best resolution is still in the micron range (> 0.2 µm). Light microscopy with spatial resolution on the nanoscale ( < 5 - 90 nm) has remained a fascinating dream for biologists and physicists alike. For biologists, optical nanoscopy would be a powerful tool to visualise and quantify protein distributions within and around organelles and for a physicist, breaking the diffraction barrier is an exciting physics problem… Is it possible? And if, how?
Fluorescence is more than glimmer: We rely on fluorescence labeling of the parts to be studied. Why? Firstly, fluorescence is by far the most popular read-out mode of cellular information. Fluorescence tags can be specifically attached to proteins, DNA, and even produced by the cell itself (e.g. the green fluorescent protein, GFP). Secondly, to a physicist, there is something very attractive about fluorescence. Image formation theory is much simpler and there are many options of poking around with it. In fact, well thought-out modfications of the fluorescence process have opened a back-door for circumventing Abbe's paradigm [1, 2].
Point-Spread-Function Engineering as a philosophy: In a scanning microscope, improving the resolution is equivalent to decreasing the spatial extent of the fluorescent spot that is scanned through the sample. Hence, we are scrutinizing physical phenomena that bear the potential of cutting down the size of the fluorescence spot. The focal spot is termed 'the point-spread-function'. So we dubbed this down-sizing approach of the spot as Point-Spread-Function Engineering [2]. But also non-scanning microscopes have Point-Spread-Functions. In fact virtually all of the physical ingredients that we use in a scanning microscope can be employed in a non-scanning CCD camera based system as well, with some modifications and advantages and disadvantages.
Physical ingredients: What are the physical ingredients for Point-Spread-Function Engineering? Potential candidates are all processes between the ground and the fluorescence state that can be saturated [1, 2, 3]. Such processes are saturated depletion of the fluorescent state or saturated depletion of the ground state of the fluorescent label. Depletion of the fluorescent state can be accomplished through stimulated emission [1], whereas depletion of the ground state [3] can be best attained through shelving the fluorescent molecule in a long-lived non-fluorescence state, such as the molecule's triplet state. Other ingredients are coherent superposition of focal fields that lead to a synthetic increase of the aperture [4, 5, 6], and even more exotic ideas such as excitation and emission events with causal links [7, 8]. Our tools are sophisticated femto- and picosecond laser systems, scanning microscopes, time-correlated photon counting, higher harmonics generation, active and adaptive optics, and other modern optical equipment to modify visible light in frequency, space and time.
Achievements: Our ideas have lead, among others, to the invention of the 4Pi-confocal and the STED-fluorescence microscope. The latter was the first microscope to break the diffraction barrier, both in theory [1] and experiment [9]. The combination of both microscopes, the STED-4Pi-microscope was the first to provide evidence for a spatial resolution of 30-40 nm with visible light and regular lenses [10], heralding "optical nanoscopy". We are striving to do even better…. Join us!
[1]
Hell, S. W. and J. Wichmann (1994). "Breaking the diffraction resolution limit by stimulated emission." Opt. Lett. 19(11): 780-782.
[2]
Hell, S. W. (1997). "Increasing the Resolution of Far-Field Fluorescence Microscopy by Point-Spread-Function Engineering." Topics In Fluorescence Spectroscopy; 5: Nonlinear and Two-Photon-Induced Fluorescence, edited by J. Lakowicz. Plenum Press, New York: 361-426.
[3]
Hell, S. W. and M. Kroug (1995). "Ground-state depletion fluorescence microscopy, a concept for breaking the diffraction resolution limit." Appl. Phys. B 60: 495-497.
[4]
Hell, S. W. and E. H. K. Stelzer (1992). "Fundamental improvement of resolution with a 4Pi-confocal fluorescence microscope using two-photon excitation." Opt. Commun. 93: 277-282.
[5]
Schrader, M. and S. W. Hell (1996). "4Pi-confocal images with axial superresolution." J. Microsc. 183: 189-193.
[6]
Egner, A., S. Jakobs and S. W. Hell (2002). "Fast 100-nm resolution 3D-microscope reveals structural plasticity of mitochondria in live yeast." Proc. Natl. Acad. Sci. USA 99: 3370-3375.
[7]
Schönle, A., P. E. Hänninen and S. W. Hell (1999). "Nonlinear fluorescence through intermolecular energy transfer and resolution increase in fluorescence microscopy." Ann. Phys. (Leipzig) 8(2): 115-133.
[8]
Schönle, A. and S. W. Hell (1999). "Far-field fluorescence microscopy with repetetive excitation." Eur. Phys. J. D 6: 283-290.
[9]
Klar, T. A., S. Jakobs, M. Dyba, A. Egner and S. W. Hell (2000). "Fluorescence microscopy with diffraction resolution limit broken by stimulated emission." Proc. Natl. Acad. Sci. USA 97(15): 8206-8210.
[10]
Dyba, M. and S. W. Hell (2002). "Focal spots of size λ/23 open up far-field fluorescence microscopy at 33 nm axial resolution." Phys. Rev. Lett. 88: 163901.