In order to test that we can actually produce a focused illumination pattern on a fluorescent microscope sample, we added a fairly simple static projection system into the epifluorescence illumination pathway of a Leica DM IL infinity tube microscope. A transmitted-light resolution target was used as the object for this system, and the light from it collimated using a plano-convex lens of the same 200mm focal length as the microscope's internal tube lens. The demagnification at the sample for any objective lens should therefore be identical to the objective's own nominal magnification.

Although a housing was made to allow the system to be used safely with the microscope's mercury lamp attachment, the combination of reduced accessibility and the operating constraints on the lamp made the process rather unwieldy and these initial tests were done instead in the open with an ordinary halogen source. In the event this provided adequate excitation light for our highly fluorescent sample, a layer of sodium fluorescein solution under a plain glass cover slip.

The following photograph is of the illumination pattern imaged onto the sample by a 10x objective. The illumination is notably uneven and the camera's auto-exposure somewhat misjudged, but the focused pattern can be clearly seen.

This next is a detail from the same image at the original capture resolution, and we can see that the 18/mm lines -- the smallest on the target -- are reasonably clearly resolved. At 10x demagnification this will be 180/mm at the sample, for a line pitch of about 5.5 microns. Most of the blurring here is likely due to the image being overexposed, and also to inadequacies in the projection setup, but there is probably also some contribution from out-of-focus emissions, given that there is no real object here, only the pattern itself exciting a more-or-less uniformly fluorescent medium.

The final image is the same projection via a 40x objective, and we can see that the illumination pattern is considerably less well resolved in this case. Innaccuries in the projection system and exposure are both evident, and I suspect that the out-of-focus emissions contribute to a greater degree also, since the light cone will be significantly broader.

A few further tests are planned, including using a different tube lens, a more powerful objective, more even illumination and (ahem) manual exposure control, but in the meantime this already provides some useful input into the project.