Principles of Astrophotonics

by Simon Ellis, Joss Bland-Hawthorn, and Sergio Leon-Saval
World Scientific, 2023, 260 pages
reviewed by T. Nelson

strophotonics, say the authors, is a whole new approach to astronomical instrumentation. It uses waveguides and optical fibers instead of bulk optical components like lenses, mirrors, and filters. The book starts with a gentle introduction to basic concepts like seeing, collimators, CCDs, spectrographs, interferometers, and conservation of étendue. Then we get several chapters on waveguides, i.e. fiber optics. A big challenge is how to couple a telescope image efficiently into a fiber without losing signal. The challenge is all the more acute because most nanophotonics devices require single-mode fibers. For these fibers, coupling efficiency decreases nearly to zero as seeing gets over 0.2 arcsec. Thus an additional optical device is needed to convert multimode to single mode. Photonic lanterns are the authors' choice here.

Another challenge is to eliminate the OH lines, which consist of a bright forest of lines between 1 and 1.8 μm. These lines are caused by emission from atmospheric OH molecules at an altitude of 87 km, and they increase the sky brightness in near IR by over 250-fold. There are so many of them that conventional dichroic filters are impractical. Fiber Bragg gratings, which are optical fibers inscribed with evenly-spaced lines of a different refractive index to create an interference pattern, have been very successful so far, but the authors propose ring resonators as an even better alternative. These have to be fabricated by lithography, as the resonator is only around 10 μm in diameter. Both types also require single-mode fibers, so a device would still need a photonic lantern on the input and output side.

Another photonic technology is the arrayed waveguide grating, which uses a large number of fibers of different lengths to impart a phase difference, potentially achieving a higher resolving power than any conventional grating.

Ideally, we would like to do away with lenses and gratings altogether and create a perfectly flat telescope. The last chapter is a little vague about how this could be done, and omits some of the more interesting developments like optical sieves and diffractive lenses. But this book is a great introduction on the practical aspects of the new technology and avoids, as much as possible, bogging down the reader with quantum mechanics.

It's frustrating, but I can't think of a single snarky comment to make about this book. Hopefully the authors will make their next book a bit easier to snark on.

jun 23, 2023

Diffractive Lens Design

by Andy Wood and James Babington
IOP, 2023, 238 pages
reviewed by T. Nelson

iffractive lenses, also known as zone plates, are a new type of lens that is perfectly flat. A DOE [diffractive optical element] uses diffraction instead of refraction to focus light. Even though the image quality tends to be lower than conventional refractive lenses, DOEs are potentially useful for focusing X-rays, which conventional optics can't handle.

In essence, diffractive lenses are circular diffraction gratings. The authors say they can be manufactured using diamond turning and propose them for mid- to long-infrared (3–5 and 8–12 μm). They say this has some advantages because traditional IR lens materials such as ZnS, ZnSe, and germanium are expensive. They provide Zemax simulations and ray-tracing equations for optimizing the blazing of mid-IR Fresnel zone plates. Another chapter discusses tolerancing, an important aspect of manufacture.

However, I suspect this book will disappoint readers hoping to learn about DOEs. Much of the discussion of zone plates actually describes Fresnel lenses, which though indeed flat are actually stepped refractive designs. Fresnel lenses are easy to manufacture out of PMMA (for visible wavelengths) and unlike diffractive zone plates, they have fairly low chromatic aberration (CA). Thus, the real topic of this book is not diffractive lens design, but ‘hybrid’ designs, where a diffractive surface modification is applied to a conventional refractive lens. The hope is that the diffractive element will counteract the CA from the refractive part and eliminate the need for additional refractive elements.

This is definitely a niche application, and potentially an important one if it works, but it would not work in the EUV or in X-rays. ZnS, Ge and similar materials have very high refractive indexes in the IR (Ge has an RI of 4.005 at 8 μm, compared with 1.40–1.45 for glass and fused silica in the visible), so refractive lenses only need a modest curvature in the IR. So it was not clear to me whether adding a diffractive surface treatment to such a lens would be worth the cost, as the optical aberrations are already pretty low. The real value of DOEs in my opinion would be as a replacement for refractive elements at the extreme other end of the spectrum, but that application isn't discussed.

In the last chapter, the authors show the benefits of the technology. They show how a hybrid lens of GASIR1 (a chalcogenide IR-transmitting glass) is almost as good in LWIR as Ge but has the advantage that it can be molded. A hybrid design may also be useful for athermalization. Another advantage is that aspher­iz­ation can be done at the same time as the surface treatment with no extra cost. This is often needed, as a diffractive element can be a trade-off between reducing CA at the expense of increasing off-axis aberrations such as field curvature and astigmatism.

Diffractive optics are also subject to a new type of aberration whereby higher diffraction orders generate undesired ghost images, which are copies of the original appearing at different magnifications. This is not to be confused with the technique of ghost imaging using correlated photons.

jul 24, 2023