In a cold 87Rb ensemble, backward-wave four-wave mixing assisted by electromagnetically induced transparency (EIT) demonstrated phase-matched difference-frequency conversion at sub-picojoule energies but with efficiency of only about 10−410^{-4}, and left three quantitative questions open: the unexplained growth of the ground-state dephasing rate γ12\gamma_{12} as the coupling Rabi frequency Ωc\Omega_c is reduced, the lack of a quantitative multilevel (Zeeman-sublevel) model reproducing the measured detuning dependences and efficiencies, and the absence of quantitative limits and an optimization route toward high-efficiency conversion at the single-photon level. The paired-photon emission predicted for the backward geometry has since been realized experimentally, and conversion efficiency has been substantially advanced in related EIT mixing schemes, but the microscopic origin of the excess ground-state dephasing and a validated multilevel model of the backward geometry remain unresolved, and with them a quantitative account of the efficiency ceiling of backward EIT frequency conversion in cold atoms.

Background

Electromagnetically induced transparency (EIT) is a quantum-interference effect in which a strong 'coupling' laser renders an otherwise absorbing atomic transition transparent to a weak 'probe' field while greatly enhancing certain nonlinear susceptibilities. In a double-Lambda system — an atom with two ground states coupled to two excited states by four optical fields, so that the level diagram traces the Greek letter Lambda twice — two of the fields can mix through the atoms and generate new fields at Stokes and anti-Stokes frequencies; this resonant four-wave mixing (FWM) is the basis of narrowband frequency conversion and of paired-photon sources compatible with atomic quantum memories. A Rabi frequency (here Ωc\Omega_c, Ωp\Omega_p) measures the strength of each light-atom coupling; Γ3=5.9 MHz\Gamma_3 = 5.9\ \mathrm{MHz} is the natural linewidth of the Rb D1D_1 line involved. If the generated Stokes field counterpropagates with respect to the driving fields, the mixing is automatically phase-matched in a short sample, a 'backward-wave' geometry that is attractive for separating weak generated light from strong pumps. The ground-state dephasing rate γ12\gamma_{12} is the decay rate of the coherence between the two ground states; it sets the width of the transparency window, so excess γ12\gamma_{12} causes residual absorption exactly where conversion should occur. In a cold atomic cloud trapped in a dark magneto-optical trap (a 'dark SPOT', a standard cold-atom trap in which atoms are optically pumped into a dark hyperfine state to raise density), Doppler broadening is small and large optical depth NσL\mathcal{N}\sigma L (atom density times length times resonant cross section) is available, which in principle maximizes the EIT enhancement. Braje, Balić, Goda, Yin, and Harris (2004) reported the first experimental demonstration of EIT-based FWM in cold atoms: backward-wave, phase-matched difference-frequency conversion at optical powers of a few nanowatts and sub-picojoule energies, with a four-level double-Lambda theory that reproduced the observed (NL)2(\mathcal{N}L)^2 efficiency scaling and detuning dependence. Their experiment, however, hit a limit they could not explain: when the coupling Rabi frequency was reduced toward the natural linewidth, the fitted γ12\gamma_{12} grew, transparency was lost, and Stokes generation dropped — and their model was an idealized four-level simplification of the real atom with its many Zeeman sublevels, nonuniform ground-state populations, and multiple coupling channels. They also predicted that the same setup, driven spontaneously without an applied anti-Stokes field, would emit counterpropagating photon pairs with a bandwidth of order 5 MHz5\ \mathrm{MHz} and coherence length of order 60 m60\ \mathrm{m}. The question is what ultimately limits this backward-wave EIT frequency conversion in cold atoms: the physical origin of the excess dephasing, the correct multilevel description, and the quantitative efficiency ceiling on the way to single-photon-level conversion.

Problem Statement

For backward-wave, phase-matched four-wave mixing assisted by electromagnetically induced transparency in a cold 87Rb ensemble (double-Lambda system on the D1D_1 line, optical depth of order 10 or more, coupling Rabi frequency Ωc\Omega_c of a few natural linewidths Γ3=5.9 MHz\Gamma_3 = 5.9\ \mathrm{MHz}), determine the physical limits of the frequency conversion, specifically: (i) identify the microscopic origin — experimental or fundamental — of the excess ground-state dephasing rate γ12\gamma_{12} that grows as the coupling Rabi frequency Ωc\Omega_c is reduced toward the natural linewidth, where it destroys the transparency window and suppresses Stokes generation, and establish how it can be mitigated; (ii) construct a quantitative model of the real multilevel atom (Zeeman sublevels, nonuniform ground-state populations, and mixing between the multiple four-wave-mixing channels) that reproduces the measured detuning dependences and conversion efficiencies of the cited backward-geometry experiment, replacing the idealized four-level treatment; and (iii) determine quantitatively the maximum conversion efficiency achievable in this backward EIT geometry and the route to it — optical-depth scaling, spatial-mode overlap, and the other loss channels including the dephasing of (i) — at input energies down to the single-photon level. The paired-photon emission projected for this geometry (bandwidth ∼5 MHz\sim 5\ \mathrm{MHz}, coherence length ∼60 m\sim 60\ \mathrm{m}) has been realized experimentally and is no longer part of the open question; the three items above are jointly posed and share one objective, the quantitative physical limit of backward EIT frequency conversion in cold atoms, because the dephasing mechanism and multilevel structure are precisely the loss channels that bound the efficiency.

References

  1. D. A. Braje, V. Balić, S. Goda, G. Y. Yin, S. E. Harris, 'Frequency Mixing Using Electromagnetically Induced Transparency in Cold Atoms', Phys. Rev. Lett. 93, 183601 (2004). DOI: 10.1103/PhysRevLett.93.183601, https://doi.org/10.1103/PhysRevLett.93.183601
  2. V. Balić, D. A. Braje, P. Kolchin, G. Y. Yin, S. E. Harris, 'Generation of paired photons with controllable waveforms', Phys. Rev. Lett. 94, 183601 (2005). DOI: 10.1103/PhysRevLett.94.183601, https://doi.org/10.1103/PhysRevLett.94.183601
  3. C. S. Chuu, S. E. Harris, 'Ultrabright backward-wave biphoton source', Phys. Rev. A 83, 061803 (2011). DOI: 10.1103/PhysRevA.83.061803, arXiv:1105.0144, https://doi.org/10.1103/PhysRevA.83.061803
  4. Y. Mei, X. Guo, L. Zhao, S. Du, 'Mirrorless optical parametric oscillation with tunable threshold in cold atoms', Phys. Rev. Lett. 119, 150406 (2017). DOI: 10.1103/PhysRevLett.119.150406, https://doi.org/10.1103/PhysRevLett.119.150406
  5. Y. Jiang, Y. Mei, Y. Zuo, Y.-H. Zhai, J. Li, J.-M. Wen, S. Du, 'Anti-parity-time symmetric optical four-wave mixing in cold atoms', Phys. Rev. Lett. 123, 193604 (2019). DOI: 10.1103/PhysRevLett.123.193604, https://doi.org/10.1103/PhysRevLett.123.193604
  6. Y. Jiang, Y. Mei, S. Du, 'Quantum Langevin theory for two coupled phase-conjugated electromagnetic waves', Phys. Rev. A 107, 053703 (2023). DOI: 10.1103/PhysRevA.107.053703, https://doi.org/10.1103/PhysRevA.107.053703
  7. H.-H. Jen, T. A. B. Kennedy, 'Efficiency of light-frequency conversion in an atomic ensemble', Phys. Rev. A 82, 023815 (2010). DOI: 10.1103/PhysRevA.82.023815, https://doi.org/10.1103/PhysRevA.82.023815
  8. C.-Y. Lee, B.-H. Wu, G. Wang, Y.-F. Chen, Y.-C. Chen, I. A. Yu, 'High conversion efficiency in resonant four-wave mixing processes', Opt. Express 24, 1008 (2016). DOI: 10.1364/OE.24.001008, https://doi.org/10.1364/OE.24.001008
  9. H.-C. Li, G.-Q. Ge, M. S. Zubairy, 'Efficient nonlinear frequency mixing using Autler-Townes splitting', Phys. Rev. A 102, 053701 (2020). DOI: 10.1103/PhysRevA.102.053701, https://doi.org/10.1103/PhysRevA.102.053701
  10. R. Kumar, D. Manchaiah, M. Ahmad, R. Easwaran, 'Effect of relaxation on the transfer of orbital angular momentum via four-wave mixing process in the four-level double lambda atomic system', New J. Phys. 26 (2024). DOI: 10.1088/1367-2630/ad4b59, https://doi.org/10.1088/1367-2630/ad4b59
  11. J.-S. Shiu, C.-W. Lin, Y.-F. Chen, 'Asymmetric biphoton generation under ground-state decoherence and phase mismatch in a cold atomic ensemble' (2025). DOI: 10.1002/qute.202500052, arXiv:2505.06650, https://doi.org/10.1002/qute.202500052