Optimizing coherent population trapping-based vapor cell atomic clocks at low temperatures

Remote sensing in extreme environments like the undersea Arctic can yield critical climate data but relies on low-power, long-term stable timing that is unavailable undersea. Current chip-scale atomic clocks (CSACs) based on the coherent population trapping (CPT) resonance in alkali vapor cells can meet stringent power requirements but have not been optimized for long-term stability at undersea temperatures. Here, I detail my progress toward a CPT cesium clock in the -4 to 10 Celsius range with $<10^{-10}$ uncertainty at year-long averaging time. On one front, I develop a model of the CPT lineshape to optimize the CPT linewidth:contrast ratio, a key figure-of-merit for short-term stability---linewidth minimization in particular is equivalent to maximizing the coherence lifetime of the hyperfine ground state superposition in alkali vapor, which has utility in many other vapor cell technologies. On a second front, I study the first-order temperature coefficient of the cesium ground-state hyperfine transition frequency in neon buffer gas, a coefficient with limited experimental validation in our temperature range of interest and one that is integral to the long-term stability of our clock.