YQI Talk - Kevin Smith - U. Washington

In analogy to the way localized electronic states of atoms hybridize to form delocalized states of molecules, confined photonic excitations of two or more optical cavities can electromagnetically interact to form supermodes of a resulting “photonic molecule”. Despite their broad importance for a wide variety of applications such as optical sensing, study of PT symmetry, nonclassical light generation and quantum simulation, photonic molecules have largely been modeled by heuristic coupled-mode theories which are afflicted by serious drawbacks, including (i) over-reliance on phenomenological parameters and numerical fitting and (ii) inaccuracy beyond the weak coupling limit. As a result, these heuristic models are incapable of providing the freedom of analytic exploration and predictivity desired for modern applications which depend upon an understanding of coupled-cavity phenomena at a high level. In this talk, I will present my efforts to remedy these deficiencies via a first-principles description of strongly coupled photonic molecules, providing predictive analytic power which outpaces that of current theoretical and numerical methods. I will show that the resulting model not only suggests that standard coupled-mode theories are inaccurate at close cavity-cavity separations, but the very intuition they rely on – that coupled cavity modes are akin to coordinate-coupled oscillators – is flawed. Finally, I will demonstrate that this generalized model suggests the possibility for exotic phenomena typically associated with ultrastrong coupling, predicted here for comparably modest parameter regimes. I will then conclude with a brief description of ongoing projects related to quantum simulation.