|Abstract:||For more than fifty years, observations of the cosmic microwave background (CMB) have provided fundamental insights into the universe we inhabit. Through a combination of ground-based and balloon- and satellite-borne experiments, we have measured the CMB's temperature anisotropy and power spectrum to high precision (Barrow and Coles, 1991; de Bernardis et al., 2000; Planck Collaboration, 2016a). Most recently, experiments are attempting to measure the CMB's polarization anisotropy. Measurements of polarization anisotropy will provide insight into the epoch of reoinization, information about the energy scale of inflation, and characterization of galactic dust, for instance (Barkana and Loeb, 2001; Hu, 2003; Planck Collaboration, 2016b).
These are difficult measurements to make, however. They require observations at both large and small angular scales, and sensitive instruments capable of observing at multiple wavelengths in order to constrain foreground contamination. One of the experiments contributing to this effort is the third generation survey camera on board the South Pole Telescope (SPT-3G; Benson et al., 2014}). SPT is located at the geographic South Pole (elevation ~2800 m), which is an exceptionally arid environment (Radford and Holdaway, 1998; Radford, 2011). A fifty year study of precipitable water vapor (PWV) at the South Pole found the average PWV is far less than 1 mm, making the South Pole one of the driest regions on Earth (Chamberlin and Grossman, 2012). This is important for telescopes observing at mm wavelengths, where atmospheric water vapor can severely attenuate the interesting radiation. Other perks of observing from the Pole are that the SPT can easily avoid the galactic plane and its dust-contaminated signal, enabling it to integrate on a patch of extragalactic sky for long periods. In addition, the low temperature variability of the six-month long polar night keeps the atmosphere stable for long stretches. The SPT optics and structure benefit from the low temperature variability as well, since thermal gradients across the SPT are small.
SPT's 10 m primary aperture and diffraction-limited optics uniquely position it to map the CMB at arcminute-scale angular resolution. SPT-3G---the new survey camera installed during the 2016/2017 austral summer---is polarization sensitive, and observes in three band passes that correspond to atmospheric transmission windows. Although there are other polarization-sensitive experiments that observe at multiple wavelengths, there are relatively few that do so with multichroic, polarization-sensitive pixels. SPT-3G fits all three colors (hence multichroic) and both polarizations within a single ~5 mm diameter pixel footprint---that's six detectors per pixel footprint. To top it off, the camera contains more than 2700 of these pixels.
As it happens, building a multichroic, polarization-sensitive mm wave camera with exceptionally high detector density is an engineering challenge---or, seemingly, a countably infinite set of challenges. This dissertation attempts to address one of them: antireflection (AR) coatings for high-refractive index optics. I say "attempts to address" because the problem is far from solved. Rather, what follows is the first iteration of a process that can complement the growing need for high detector density CMB experiments. With the upcoming CMB-S4 project intending to field hundreds of thousands of detectors, the days of bespoke quasi-optical coupling elements are drawing to a close. What we will need going forward is a scalable, high-throughput manufacturing process for quasi-optical coatings. That is to say, something like the process detailed in this work. I also address the similar challenge that is AR coating large-scale reimaging optics. Again, this isn't a solved problem, but we've developed a method that works well enough. The materials needed to manufacture both the small-scale (i.e., quasi-optical) and large-scale AR coatings are inexpensive: about $1 per pixel for the quasi-optical coating; about $1500 per SPT-3G-sized lens. Both coatings increase the amount of radiation transmitted through the optics by >25% across all three observing bands.
In addition to the manufacturing methods described above, this work introduces an open-source Python package that will be beneficial to those who design and model mm wave AR coatings, as well as laboratory measurements of SPT-3G AR coatings and their constituent materials. These measurements include optical measurements (in the form of Fourier-transform spectroscopy, vector network analysis, and reflectometry), x-ray spectroscopy, and scanning electron microscopy. There is also a discussion of the end-to-end optical efficiency of the SPT-3G survey camera and some of the factors that impact it.