Abstract

The cosmic microwave background radiation contains great amounts of information that allow for studying the physics of the early universe through constraining cosmological parameters in the standard ΛCDM model. The CMB temperature signal has been measured to high precision, but measuring the CMB polarization signal is still in its early stages.The theoretically small primordial CMB polarization B-mode signal has not yet been measured, but has principle importance in that its existence would be strong evidence of inflation. This measurement allows one to probe the earliest state of the universe at energy scales of 10^16 GeV thought to be near the Grand Unified Theory scale. The B-mode signal arising from weak gravitational lensing by large scale structures provides information about the matter composition of the universe and puts strong constraints on the sum of the neutrino masses.This dissertation discusses the optical design, instrumentation, data analysis, and first season science results of the POLARBEAR experiment, a CMB polarization telescope aimed to measure the B-mode signal. The results show the first evidence of non-zero lensing B-modes at sub-degree angular scales on the sky. The development and measurement results of the Fourier transform spectrometer calibration instrument used to characterize the spectral response of the POLARBEAR detectors are also described. The optical design development and systematic studies for the Simons Array, the next generation installment of the experiment, are described as well. The cross polarization effect of Mizuguchi-Dragone breaking due to a prime focus half-wave plate, and the optical redesign of the Simons Array re-imaging optics for increased optical performance at higher frequencies were studied in detail. The Simons Array is planned to fully deploy in 2018 to further study the CMB with enhanced sensitivity.