|  Image slicers

Each fibre of PEPSI will be equipped with a so-called "wave-guide image-slicer" assembly.
The slicer stack consists of a stack of n single glass plates (slicers) that are adhesively bonded together (figure above). The overall dimensions of the stack are n times (30 µm to 100 µm thickness of the single slices). The end of every slice is polished under an angle of 45 degrees and coated with silver. The entire image-slicer assembly contains the slicer stack itself, a scrambler in front of its entrance window, the fiber that couples the light from the telescope, a ball lens to focus the light into the image-slicer and a beam-splitting cube to monitor the adjusted state of the focused spot and a tank that contains an immersion liquid. The assembly, except of the beam-splitter, is shown in the figure below. 

Our design approach is based on the availability of pre-thinned 30μm thick plane-parallel glass plates of high throughput for the entire wavelength range. Up to 12 of these glass plates are then glued on top of each other where the "glue" acts as the cladding for the waveguide. Its thickness is crucial for the efficiency of the device. New technologies for achieving glue thicknesses of 3μm (three times the maximum wavelength of PEPSI) were already achieved at IOF. As a comparison, classical techniques allow only 6-8μm spacing which limits the device efficiency to 80% for a 200μm fibre output. The IOF and the AIP are working now to realize a 100μm slicer based on fore optics that enlarge the beam diameter that is to be sliced. This would further increase the instrument efficiency. 

For the given three fibre cores of PEPSI, the UHR mode can only use the smallest of the three, the 100μm fibre. When slicing this into 12 slices we reach the practical limit due to the available space on the CCD detector and due to the quality of the optics. Most likely, UHR-PEPSI could reach even a resolution twice as large, approx. 600,000, but then only with a "one-pixel resolution" on the CCD, and accordingly large light losses at the entrance slit. 
Last updated:16:56 22/01 2008, by JB
|  |