Theory of Two-Photon Absorption with Broadband Squeezed Vacuum

TL;DR

This paper develops an analytical quantum model for two-photon absorption using broadband squeezed vacuum, analyzing the effects of molecular linewidth and optical bandwidth on TPA efficiency in different gain regimes.

Contribution

It provides a comprehensive theoretical framework for understanding TPA with broadband squeezed vacuum, including the transition between low-gain and high-gain regimes and the impact of dispersion compensation.

Findings

Bright squeezed vacuum is as effective as classical pulses in high-gain, narrow linewidth cases.

TPA rate depends on g^(2)(0) when molecular linewidth exceeds optical bandwidth.

Dispersion compensation is necessary for optimal g^(2)(0) in BSV.

Abstract

We present an analytical quantum theoretic model for non-resonant molecular two-photon absorption (TPA) of broadband, spectrally multi-mode squeezed vacuum, including low-gain (isolated entangled photon pairs or EPP) and high-gain (bright squeezed vacuum or BSV) regimes. The results are relevant to the potential use of entangled-light TPA as a spectroscopic and imaging method. We treat the scenario that the exciting light is spatially single-mode and is non-resonant with all intermediate molecular states. In the case of high gain, we find that in the case that the linewidth of the final molecular state is much narrower than the bandwidth of the exciting light, bright squeezed vacuum is found to be equally (but no more) effective in driving TPA as is a quasi-monochromatic coherent-state (classical) pulse of the same temporal shape, duration and mean photon number. Therefore, in this case the sought-for advantage of observing TPA at extremely low optical flux is not provided by broadband bright squeezed vacuum. In the opposite case that the final-state linewidth is much broader than the bandwidth of the BSV exciting light, we show that the TPA rate is proportional to the second-order intensity autocorrelation function at zero time delay g^(2)(0), as expected. We derive and evaluate formulas describing the transition between these two limiting cases, that is, including the regime where the molecular linewidth and optical bandwidth are comparable, as is often the case in experimental studies. We also show that for g^(2)(0) to reach the idealized form g^(2)(0) = 3 + 1/n, with n being the mean number of photons per temporal mode, it is required to compensate the dispersion inherent in the nonlinear-optical crystal used to generate the BSV.