Development of Optics and Detectors for Advanced CMB Polarization Measurements

In the four years of my NASA Space Technology Research Fellowship the state of the art in optics and detectors technologies has progressed rapidly. In my first year, ACTPol had just deployed the first of three polarization sensitive kilopixel detector arrays. In the time since, ACTPol has seen full deployment, including the first multichroic detector array of its kind, observing simultaneously at 90 and 150 GHz. Most recently, all three original ACTPol detector arrays have been upgraded to three new multichroic arrays, spanning from 90 GHz to 220 GHz, forming Advanced ACTPol and doubling the number of detectors on the sky from ~3,000 to ~6,000. Starting with the first ACTPol array, I developed a unique optical modeling based polarization calibration. This technique propagates the fabricated detector angles through the optical chain of the telescope to the sky. The lenses in the telescope receiver are off-axis and thus influence the propagation of the polarization vectors through the camera according to their Fresnel coefficient. This additional polarization rotation has been shown to be critical input to the calibration of the polarization angles for the ACTPol detectors for the mapping of the Cosmic Microwave Background (CMB) polarization. This is confirmed by checking that our EB-nulling angle, the global polarization rotation required to zero the EB-cross correlation power spectrum, is consistent with zero. This calibration procedure continues to be used on the second generation of ACTPol, known as Advanced ACTPol. While having an EB-nulling angle consistent with zero is a good check of the calibration procedure, a more direct measure of the detector angles, and possibly the rotation caused by the optics in the ACTPol receiver, is desirable. For this purpose, I developed and performed measurements in the field with a series of polarizers. Taking advantage of hardware developed to smoothly rotate a set of new metamaterial half wave plates on Advanced ACTPol, I rotated these polarizers at a constant rate directly in front of the telescope receiver. This produces a sinusoidal signal on each detector, allowing for the extraction of each detector angle through the determination of the phase of this sine wave. The measure of performance of this method has been the ability to determine if two orthogonal detectors are out of phase. Our best polarization grid measurements so far have achieved a scatter of ±2 degrees. As the next generation of telescopes were being developed, the need to cover several different frequency bands to account for foregrounds such as synchrotron and polarized dust became apparent. This lead to the Advanced ACTPol design containing four multichroic arrays, with science bands at 27, 39, 90, 150, and 220 GHz. These detectors require different physical geometries to observe at each frequency. I worked on selecting those geometries, specifically the geometries of the weak thermal link connecting the transition edge sensors (TES) with the cold based on measurements performed at Cornell on devices with varying leg geometries, fitting to a model, and selecting the geometry that best achieves the desired detector parameters. The 90, 150, and 220 GHz detectors have all been fabricated and are deployed and on the sky now. The 27 and 39 GHz detectors are being fabricated now and will be deployed in early 2018. The field of observational cosmology continues developing at a rapid pace. I have learned about and developed these optics and detector technologies for ACTPol and Advanced ACTPol and am now getting to apply the tools and techniques I have developed to the early design stages for two new telescopes, CCATp and the Simons Observatory, as both begin planning and construction. Further optics and detector technology developments made on CCATp and SO will establish large scale detector arrays and optics technologies for use in future NASA space based missions such as the Inflation Probe.