arXiv:2611.02861
Quantum-Nondemolition Probing of Multiphoton Transitions between Dressed States in a Strongly Driven Jaynes--Cummings System
Han-Yu Ren, Zi-Yi Liu, Wen-Qi Fu, Zhuo Chen, Wen-Hua Qin, Tao Rong, Bo-Yang An, ZhiAng Qin, Li-Wei Liu, Rami Mrad, Ming Gong, Yingqiu Mao, Yuan Zhang, Johannes Majer
Abstract
The response of the Jaynes--Cummings (JC) model to strong driving has been examined previously, but the origin of intriguing features near emitter--cavity resonance remains poorly understood because of the large background in conventional transmission spectroscopy. Here we introduce a spectroscopic approach for circuit quantum electrodynamics that combines strong cavity driving with quantum-nondemolition qubit readout. Separating the drive and readout substantially enhances the measurement sensitivity as compared with conventional cavity-transmission spectroscopy. With this technique, we identify the near-resonant features as multiphoton transitions between emitter--photon dressed states of the same parity, and resolve transitions between dressed states of opposite parity. Away from resonance, we also observe selective enhancement and suppression of multiphoton transitions, together with transitions involving the second excited state of the anharmonic qubit. Our approach opens a route to resolving the rich transition structure of the JC ladder and probing strongly driven light--matter interactions.
AI Summary
Strongly driven Jaynes–Cummings spectroscopy contains near-resonant features that conventional cavity transmission does not readily resolve. This paper investigates their origin using a separately driven cavity–transmon system and an independent, dispersively coupled resonator for continuous-wave qubit readout. The device has cavity frequency $3.171\,\mathrm{GHz}$, coupling $g/2\pi=31\,\mathrm{MHz}$, transmon anharmonicity $192\,\mathrm{MHz}$, and cavity linewidth approximately $67\,\mathrm{kHz}$. Separating excitation from detection resolves ground-state multiphoton transitions to excitation manifolds up to $n=20$ on the lower branch and $n=14$ on the upper branch. Additional lines are assigned to thermally assisted transitions from $|1,\pm\rangle$ to higher dressed states, including processes involving more than seven photons. Weaker transitions connecting opposite JC branches, $|1,\pm\rangle\to|2,\mp\rangle$, are also resolved. Flux-dependent measurements track these transitions and avoided crossings involving the transmon’s second excited state, including a virtual two-photon exchange. Assignments use multilevel Hamiltonian diagonalization and steady-state Lindblad simulations, alongside direct-transmission measurements. A global fit to 28 manually selected transition frequencies gives an rms residual of $0.16\,\mathrm{MHz}$. Spectral line-cut fits require an effective cavity-bath occupation $n_{\rm th}=0.010$, corresponding to $33\,\mathrm{mK}$; this is a model parameter, not an independently measured equilibrium temperature. The readout is treated as a phenomenological proxy for transmon occupation rather than absolutely calibrated transmission. Fit uncertainties are not reported. Finite Hilbert-space truncation does not reproduce the high-occupation central response associated with photon-blockade breakdown. Time-resolved dressed-state jumps and transition rates remain future measurements.
Error found: The paper incorrectly calls the ± Jaynes–Cummings branches different parities: both branches at excitation number n have parity (−1)^n, and the weak branch-changing transitions already have nonzero cavity-drive matrix elements in the ideal model.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 6 pages, 4 figures