The figure schematically depicts a proposed achromatic nulling beam combiner. This instrument is intended for use in astronomy - principally, for attenuating light from stars or other bright sources in order to enable detection of fainter objects that lie near the bright sources. In comparison with a prior nulling beam combiner, the proposed instrument would be simpler, made of fewer parts, easier to use, and less sensitive to the details of optical coatings. The proposed design provides for rigorous symmetry of the optical train. Moreover, the simplified design involves a relatively compact, mostly planar, configuration based entirely on flat optics, with fewer reflections than in previous designs. Because of its high degree of symmetry, the instrument would be inherently achromatic (broad-band) and capable of processing dual polarization light.
The impetus for the proposed design was the idea that unlike prior approaches, it should be possible to separate the field-flipping and the beam-combining stages. If a relative field inversion were accomplished first, subsequent superposition of the two input beams in a standard interferometer would yield subtraction rather than addition of the electromagnetic fields of the beams at zero optical path difference. In addition, it was realized that if, unlike in prior designs, the optical train could be made completely symmetric, it would theoretically be possible to subtract two identical input beams perfectly (neglecting such practical limitations as variations of optical coatings and errors of alignment and phasing).
It is assumed that the two input beams would be parallel and collimated. The electromagnetic fields of the two beams would be flipped, relative to each other, by reflection in two mirror-symmetric right-angle periscopes. The two mirrors in each periscope would effect one s-plane and one p-plane reflection, respectively, and together they would reverse the roles of the s-plane and p-plane reflections. Hence, the two polarization states would be affected symmetrically by each periscope and hence, as long as the coatings on the mirrors in both periscopes were identical, no s-p phase delay would be incurred. After passage through these periscopes, the propagating two-polarization-component fields should be identical to the input fields, except for the relative field flip.
The beam-combining stage would be based on a Mach-Zehnder interferometer in which each beam would encounter two beam splitters. With respect to the transmission and reflection coefficients for the two polarization states, the encounter with each beam splitter would be reciprocal to the encounter with the other beam splitter, so that complete symmetry would be ensured.
This work was done by Eugene Serabyn and Mark Colavita of Caltech for NASA's Jet Propulsion Laboratory. NPO-21156