Development of an Experimental Setup for Measurements of Photon Statistics in both Time and Frequency Domains

Abstract

In quantum optical experiments, the detected signal always contains noise due to the particle nature of photons as well as due to thermal fluctuations. Although these noises may appear as an unwanted disturbance that masks the desired signal, it also carries important physical information about the quantum properties of light. Measurement of light intensity fluctuations using high efficiency photodetectors can reflect the statistical nature of photon emission from the light source and can reveal whether the light source is thermal (super Poissonian), perfectly coherent (Poissonian), or (single photon-like) sub Poissonian, etc. This MS Thesis project focuses mainly on the instrumental development of two complementary experimental setups from scratch and then uses these for investigations of the photon correlation and photon statistics of various light sources, including a UV laser, a He-Ne laser, a low-noise current-driven laser diode, and white light. In time-domain measurements, we developed the experimental technique of Hanbury-Brown-Twiss (HBT) measurement, which requires very low optical intensities to avoid detector saturation and initial dead-time effects. Dead time errors during initialisation in coincidence detection can significantly distort experimental results and lead to deviations from theoretical predictions. Experimentally reducing these dead times of detectors and time-correlators by applying appropriate experimental corrections improves agreement between measurements and theory. We have executed a few such experimental troubleshooting procedures. Photon statistics can also be investigated in the frequency domain through noise spectrum analysis. In frequency-domain measurements, higher optical intensities can be used because the analysis is performed using continuous photodetection and spectral noise measurements rather than single-photon coincidence counting. Therefore, in addition to time-domain HBT-based g²(τ) measurements, noise levels of various light sources were also photo-detected and measured using a GHz spectrum analyser. These allowed us to do a comparative study of photon statistics in both the time domain and the frequency domain. This, in a way, helped us to gather complementary insights into the nature of photon statistics of various light sources.