Source-linked AI summary
Quantum optical circulator controlled by a single chirally coupled atom
Michael Scheucher, Adèle Hilico, Elisa Will, Jürgen Volz, Arno Rauschenbeutel
TL;DR
The paper asks how a quantum optical circulator’s routing direction can be controlled by a single atom’s internal state. The circulator’s routing direction is defined by the atom’s internal state. The circulator exhibits photon-number-dependent routing, with clear photon bunching for simultaneously arriving photons.
Problem
The paper asks how a quantum optical circulator’s routing direction can be controlled by a single atom’s internal state.
Source: p. 4Method
The circulator’s routing direction is defined by the atom’s internal state.
Source: p. 4Results
The circulator exhibits photon-number-dependent routing, with clear photon bunching for simultaneously arriving photons.
Source: p. 6Takeaways & Limitations
The atom-controlled system also supports operation in coherent superpositions of routing states.
Source: p. 6Optical pumping effects affect circulator performance for the 2→3 and 4→1 directions.
Source: Experimental procedure · p. 15
Abstract
from arXiv · showhide
We demonstrate a fiber-integrated quantum optical circulator that is operated by a single atom and that relies on the chiral interaction between emitters and transversally confined light. Like its counterparts in classical optics, our circulator exhibits an inherent asymmetry between light propagation in the forward and the backward direction. However, rather than a magnetic field or a temporal modulation, it is the internal quantum state of the atom that controls the operation direction of the circulator. This working principle is compatible with preparing the circulator in a coherent superposition of its operational states. Such a quantum circulator may thus become a key element for routing and processing quantum information in scalable integrated optical circuits. Moreover, it features a strongly nonlinear response at the single-photon level, thereby enabling, e.g., photon number-dependent routing and novel quantum simulation protocols.
Correspondence
A single atom controls the circulator’s routing direction through its internal Zeeman state and chirally unequal coupling to oppositely propagating light. The study characterizes port transmissions, operation fidelity, photon survival, and photon correlations.
Operation principle
- For an atom in F = 3, mF = +3, σ+ transition strength is 28 times larger than σ−, routing inputs 1→2, 2→3, 3→4, and 4→1.Preparing the atom in mF = −3 reverses which polarization couples more strongly.Source: Figure 1 · p. 11
Circulator performance
- The study plots port-to-port transmissions against κtot/2κ0 and calculates operation fidelity F and photon survival probability η from those data, alongside model predictions.The model predictions use gccw = 2π × 12 MHz; the intrinsic field decay rate is κ0 = 2π × 5 MHz.Source: Figure 2 · p. 12
Transmission matrices
- Transmission matrices show that preparing the atom in mF = −3 reverses the routing direction, while measurements without the atom show reciprocal operation.The fiber–resonator coupling was set to κtot/2κ0 = 2.2 for the measured matrices.Source: Figure 3 · p. 13
Quantum nonlinearity
- Photon correlations show antibunching for 1→2 and 3→4, and bunching for 1→4 and 3→2, consistent with the expected intra-resonator photon statistics.Antibunching is more pronounced for 1→2 because κa > κb. The nonzero g(2)(0) for 1→2 and 3→4 is theoretically expected from a coherent superposition of resonator light and probe-laser light at the outputs.Source: Figure 4 · p. 14
Materials and Methods
The methods describe single-atom detection and probe sequences, calibrated transmission and second-order-correlation measurements, and analytical atom–resonator models. They also specify the resonator-coupling assumption and account for direction-dependent coupling, imperfect circular polarization, and photon survival with respect to polarization overlap.
Experimental procedure
- Transmission through fiber a increases by two orders of magnitude; within approximately 150 ns, detection light is switched off and probe light applied, while optical pumping affects 2→3 and 4→1 circulator performance.A 1.5 G bias magnetic field is applied along the resonator axis for both cases.Source: Experimental procedure · pp. 15–16
Simplified model
- The simplified model uses orthogonally polarized counter-propagating modes to reduce the atom to independent two-level systems with direction-dependent coupling, then calculates low-saturation transmissions.The atom–resonator field is obtained by solving the master equation; coupling asymmetry reflects the transition-strength difference between the two systems.Source: Simplified model · pp. 18–19
Coupling to both resonator modes
- To account for imperfect circular polarization, a second model treats the atom as a circular-dipole two-level system and neglects coupling to the weaker atomic transition.The authors state that this approximation accurately describes most physical situations because weak-transition coupling is typically smaller than residual coupling to the strong transition; it yields theoretical transmission curves for Fig. 2.Source: Coupling to both resonator modes · p. 19
- With equally coupled fibers (κa = κb = κ), photon survival probability is independent of polarization overlap α.The process fidelity is mainly limited by the non-unit overlap of the evanescent field with σ+ polarization (α² < 1).Source: Coupling to both resonator modes · p. 20
Second order correlation functions
- Second-order correlations were measured for relevant output-port pairs; other pairs lacked enough transmitted photons for meaningful estimates.Figure 5 presents the full set of second-order correlation functions for the relevant port pairs.Source: Second order correlation functions · p. 20, Figure 5 · p. 22
Transmission matrices
- Transmission matrices were measured at κtot/2κ0 = 2.2 for atoms in F = 3, mF = +3 or −3 states and with no atom.Rows denote input ports, columns output ports, and error bars indicate ±1σ statistical error.Source: Transmission matrices · pp. 20–21