I'm gradually building a UV microscope for transmission imaging down into the UVB by converting an Olympus BHB. This means swapping out all the original optics for fused silica or similar as glass strongly absorbs the UVB.
The condenser has proved a bit of a sticky thing to deal with. I have a Zeiss Quartz condenser, but that is currently in the US waiting for when I can next visit and pick it up. In the mean time I thought about converting a standard Abbe condenser (simple construction with 2 elements). The lens that goes directly underneath the slide though is quite a complex shape, not something that is 'off the shelf' in UV fused silica. So I tried using a 8mm UV Fused Silica half ball lens instead. As it was a smaller diameter than the original, I fixed it in place with JB Weld. Perhaps not the most elegant solution, but it seems to work. The small size limits me to about 10x and above objectives. Here's how it looks compared to a standard Abbe condenser.
To try it out I took a set of stacks of Gorgonian sea fan spicules (from Diatom Lab) using an Olympus SPlan 20x objective and the standard microscope tungsten light source. Images taken with a Canon 5DSR camera and Olympus 2.5x NFK photoeyepiece. Stacked in Zerene (Pmax) each from about 15 images and retouched in Zerene. The 2 example images have been treated the same way as close as possible, and have not been sharpened.
First with the standard Abbe condenser.
Second with the 8mm UV fused silica half ball lens.
The image with the 8mm half ball UV fused silica lens is not a sharp as the with the standard Abbe condenser, but I'm quite happy with this as a first attempt, as this was just the 8mm half ball lens by itself vs the 2 lenses in the the Abbe condenser. I also have a plano convex UV fused silica lens on order to replace the second lens in the Abbe condenser, and that should be here in a few weeks.
While this simple type of construction will never be classified as an Aplanat Achromatic condenser, it does at least allow me to get down below 300nm and I'll be using narrow bandpass filters for my UV imaging anyway so aberrations should be less of an issue.
