arXiv:2610.40216
Detuning-assisted cavity enhancement in spectroscopy
Subhasish Guha, Sourav Das, Francesco Albarelli, Aiman Khan, Elnaz Darsheshdar, Adam D. Wright, Animesh Datta
Abstract
Cavity-enhanced spectroscopy is a widely used experimental technique. We show that detuning the cavity frequency from the transition of interest can enhance the precision for dipole moment estimation by up to two orders of magnitude beyond the improvement when the cavity is resonant with the transition. We show this for both classical continuous-wave (CW) and pulsed single-photon probes, even in the presence of losses. We trace this enhancement to a detuning-controlled redistribution of information between the light- and matter-like polaritons, with losses determining the optimal branch. We show that this enhancement remains accessible using conventional photodetection and time-correlated single-photon counting.
AI Summary
Can detuning a spectroscopy cavity from a matter transition improve dipole-moment estimation relative to resonant operation? This paper studies a single two-level atom coupled to a cavity, estimating $\mathcal G=g^2$, proportional to the squared transition dipole moment. It treats weak continuous-wave coherent illumination using transfer-matrix transmission and photon-counting Fisher information (FI), and pulsed single-photon illumination using single-excitation input–output dynamics and quantum Fisher information (QFI). The calculations predict detuning-assisted enhancement, with precision defined as inverse estimator variance and reported gains reaching over two orders of magnitude relative to resonance. Detuning redistributes information between the polariton branches: the atom-like branch is preferable at low atomic loss, while the light-like branch can retain more information when atomic excitation is strongly lossy. Numerically optimized FI and QFI show enhancement with nonzero losses. In the lossless model, optimization over single-photon input states gives approximate dimensionless bounds $\mathcal G^2\mathcal Q_{\mathcal G}\lesssim16(\Delta_{\rm AC}^2+4\mathcal G)/\mathcal K^2$ in the strong-coupling limit and $\mathcal G^2\mathcal Q_{\mathcal G}\lesssim4$ in the bad-cavity limit. Fixed-duration pulse families instead saturate at finite QFI as detuning grows. Calculated time-correlated single-photon-counting FI also exhibits enhancement. The analysis assumes local parameter estimation, weak unsaturated coherent drive or single-photon excitation, and a single cavity mode. Ideal optimized pulses involve monochromatic frequency superpositions and may be inaccessible experimentally. Lossless continuous-wave divergence requires infinite spectral resolution; finite laser linewidth regularizes it. With losses, enhancement need not persist at arbitrarily large detuning or loss rates.
Central claim may not hold: The lossy Fisher information holds atomic decay Γ fixed while varying G=g², so it does not establish the advertised physical dipole-moment or linewidth precision when Γ shares the dipole dependence G=λΓ stated in the appendix. The strong-coupling expression G²Q≤16Ω_R²/K² is not a rigorous upper bound at finite K, because evaluating the true phase-sensitivity extrema yields a strictly larger optimal QFI.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 33 pages, 11 figures (appendix starts from page 8)