As a first step our aim is not to perform imaging, the excitation-depletion processes are induced in a uniform distribution of fluorescence emitters, i.e. a Coumarin 490 dye solution and the beam is not scanned. The thickness of the light sheet can be obtained by simply measuring the width of the fluorescence traces. The following setup
is used with the four quadrant chromatic waveplate inserted before the focusing lens. We examined the fluorescence traces induced by the UV laser beam with and without the overlapping donut-shaped 532 nm beam.
the figure above shows typical images of the fluorescence traces induced either by the excitation beam alone and by the excitation beam combined with the donut-shaped STED beam. An important remark is to be made. The incident converging beam has to cross interfaces between air and the spectrometer cell, which is made of glass, and into the ethanol solution. Due to the refractive index mismatch between these media, a substantial amount of spherical aberration is present in the beams. This is responsible for the asymmetric aspect of the fluorescence excited by the 355 nm beam along the propagation axis. Also, the FWHM waist of the UV trace is close to 5 µm, which is somewhat higher than the value expected for a 355 nm beam focused by a 0.1 NA lens. However, the presence of spherical aberration is not necessarily a disadvantage in STED-SPIM microscopy.
The beam sections as a function of the position along the propagation axis are plotted on the figure below.
The respective theoretical beam profiles, assuming a gaussian approximation with identical beam waist are also plotted.
For the excitation beam alone the 100 µm pseudo-Rayleigh length is more extended than the theoretical Rayleigh length (23 µm), this is certainly due to the the spherical aberration. With stimulated depletion the difference is much more important : The pseudo-Rayleigh length is now 160 µm while the theoretical Rayleigh length is 7 µm. This result is very interesting because the stimulated emission depletion with a doughnut mask reduce noticeably the size of the fluorescent beam, ie the thickness of the light sheet, and stimultaneously strongly extends the thickness uniformity over a large field of view. This is exactly the opposite behaviour of light sheets obtained with gaussian beams.