Spectra-Tech Reflachromat 15x objective
While the large majority of microscope objectives use refractive optics, a few use reflective optics, either alone or combined with refractive elements. An interesting property of reflective optics is that they are immune to chromatic aberrations, since reflection angles do not vary with wavelength (refraction angles do). However, reflective optics are affected, for example, by spheric aberration, just like refractive objectives. Another interesting property is that purely reflective optics can typically be used in a much broader spectrum of radiation (depending on the materials used to coat the mirrors, for aluminium about 200-20,000 nm) than typical refractive optics.
Some of the earliest reflective microscope objectives (e.g., made by Leitz) were designed at a time when some of the aberrations introduced by microscope objectives, including chromatic aberrations, were corrected in the eyepiece. Thus, at least some of these objectives were designed to be compatible with refractive objectives and eyepieces of the same time and manufacturer. Refractive elements, for instance, were added to reflective elements with the purpose - in part - of adding the aberrations expected by eyepieces of the same manufacturer. In addition, spherical aberration is easier and cheaper to compensate with refractive elements. While reflecting objectives with more than two reflecting elements involve significant design problems (see e.g. an article by Steel from the 1950s), large numbers of refractive elements can easily be stacked to correct for numerous aberrations.
Refractive-reflective elements (in practice, second-surface mirrors with both glass surfaces ground into curved surfaces) were also used in reflecting objectives (one example is the Olympus OM 500 mm f/8). One of their advantages is that the silvered or aluminium-coated mirror surfaces in this way are protected by the glass instead of being directly exposed to potential degradation by air and contaminants.
This page discusses a type of reflecting objective that uses exclusively reflective optics. The Spectra-Tech Reflachromat 15x discussed on this page uses two reflecting surfaces, and corrects spherical aberration by using aspherical surfaces. It also corrects spherical aberration introduced by cover glass of varying thickness between the subject and objective by varying the reciprocal distance of the reflecting elements.
Multiple Reflachromat series were made. Earlier ones were designed for use in finite systems with a tube length of 142 (or sometimes 160) mm. Subsequent ones are infinity corrected and require a tube lens. 10x, 15x and 32x magnifications were available. Different models were also made with different correction ranges for the thickness of cover glass (from 0 up to several mm). Most Reflachromat objectives were apparently used in microscopes that split VIS and IR, and sends VIS to a conventional binocular or trinocular microscope head and IR to a spectroscope for simultaneous chemical analysis of the small imaged subject area. Because of the good UV reflecting properties of aluminium mirror coatings, the Reflachromat objectives are usable also in UVA and UVB, although the mirror coatings do not appear to be optimized for UV.
Matching condensers, apparently using the same optics of the corresponding objectives, were also available. While the Reflachromat objectives seem no longer to be made, comparable optics are still available from Edmund Optics and Thorlabs. At one point or another in time, reflecting objectives were also made by Ealing, Sigma Koki, Leitz, Melles Griot, Beck, Newport and Vickers.
The Reflachromat 15x has an NA of 0.58, which is high for an objective of this magnification. Working distance at this magnification is shallow (14 mm from the black front baffle), and removing the front baffle would increase it by only a couple of mm. Given the very large diameter of the barrel (52 mm at the front baffle, almost 65 mm in its widest portion), this objective is primarily of interest for transmitted illumination, or axial illumination via a beam splitter inserted at the rear of the objective.
The optical scheme is simple, with a large primary mirror with a central hole mounted in the barrel and facing forward, and a small secondary, convex mirror mounted in a three-legged spider and facing toward the eyepiece. Aside from the legs of the spider, there is nothing but air in the optical path of the objective. Fouling of the mirrors by dust, chemical fumes, and even hairs and cobwebs is a real possibility. The spider is recessed within the aperture of the front baffle by about 5.5 mm, so the working distance could in principle be increased by this amount if the front of the barrel is removed. Doing so, however, exposes the spider to deformation by accidental impacts, which would make the objective unusable. In fact, some of the Reflachromats on eBay do display a visibly deformed spider and are therefore worthless. Removing the front baffle also invites stray light and flare.
Some of these lenses have a narrow, projecting conical "lens shade" with the shape of a truncated cone with its narrower end toward the subject, placed at the center of the black front baffle and integral with this baffle. Mine has a flat baffle. This type of baffle is too wide and admits lots of stray light directly from the subject plane (as seen by looking directly through the microscope tube without an eyepiece). I have also seen some of these lenses equipped with a front glass cone, with the tip almost touching the subject.
My specimen of the lens has a correction ring and scale, but the scale has no unit markings except for a single 0 (presumably indicating no cover glass). The correction ring rotates in both directions past the 0 marking, but only a short distance in one direction. I assume that the opposite direction is the right one for use with a cover glass. In this direction, however, the correction ring rotates more than one turn, so the 0 marking is met at least twice while rotating the ring. The short ticks of the correction scale are still slightly visible after one whole turn of the ring (see pictures below). This is puzzling, but by similarity with pictures of Reflachromat 15x specimens with shorter correction scales I assume that the first 0 (with the shorter scale ticks fully visible) is the correct one. I also assume that the longer ticks correspond to 1 mm intervals in cover glass thickness.
The objective weighs a remarkable 604 g and requires a solid microscope stand. On ordinary RMS revolving objective turrets, this objective prevents the use of the two adjacent objective positions. The 50 mm length of the objective barrel also means that it cannot be made parfocal with many other RMS objectives. The stage should preferably be lowered (or the objective lifted) with the coarse focus knob before rotating the turret. The very shallow depth of field (2-2.5 µm) makes it necessary to use a focusing rack with a fine focus precision of at least 200 µm per turn.
The above picture shows the test rig I used for the initial tests of this lens. The camera mounted on the rig is a full-frame 42 Mpixel Sony Alpha 7R II. The distance between objective flange and sensor can be varied by adding M42 extension tubes. Unlike extension tubes with bayonet mounts, multiple screw-mount extension tubes can be stacked without losing rigidity.
The main problem of this lens is that, initially, stray light caused a very large amount of flare that almost completely washed out the subject. Switching to a black background only provided a slight improvement. Adding a flat circular baffle with an opening diameter of 15 mm at the front of the objective (the built-in baffle of the objective has a diameter of 26 mm) caused another small improvement, but the image circle reduced to about 22 mm.
In addition, the magnification at the tube length of 142 mm was too high (about 20x), and I had to reduce the flange-to-sensor distance to 114 mm to provide the nominal 15x magnification. At this magnification, even without any cover glass, I had to set the compensating ring of the objective to 2 mm (20 minor ticks of the scale) to get rid of a pronounced curvature of field. This suggests that at 114 mm tube length the objective is working substantially outside its design parameters. Since some curvature of field is still present at 142 mm tube length, I decided to use instead a 160 mm tube length, which gives a flat field when the correction ring is set to 0. This, however, provides a magnification of ??x, and a slightly shorter working distance than at lower magnifications.
[work in progress]