Polarization Entanglement in Atomic Biphotons via OAM-to-Spin Mapping

We demonstrate polarization-entangled biphotons in a cold-atom double-$Λ$ system, overcoming atomic selection rules that suppress polarization correlations and favor orbital angular momentum (OAM) entanglement. Using spatial light modulators, we coherently map a selected two-dimensional OAM subspace onto the polarization basis and thereby open an otherwise inaccessible polarization channel. Quantum-state tomography confirms that the mapping preserves the biphoton coherence. The four polarization Bell states are generated with fidelities of $92\text{-}94%$ with few-percent statistical uncertainties, and an average Clauser-Horne-Shimony-Holt parameter of $S=2.44$ verifies the survival of nonlocal correlations. To the best of our knowledge, this work presents the first demonstration of OAM-to-polarization entanglement transfer in a cold-atom spontaneous four-wave mixing platform and establishes a practical interface for integrating atomic OAM resources with polarization-based quantum communication networks.

💡 Research Summary

In this work the authors address a long‑standing limitation of cold‑atom spontaneous four‑wave‑mixing (SFWM) sources: in the standard double‑Λ configuration the atomic selection rules suppress polarization correlations, leaving only orbital angular momentum (OAM) entanglement between the Stokes and anti‑Stokes photons. By inserting a purely external mode‑conversion stage they succeed in transferring the OAM entanglement of a selected two‑dimensional OAM subspace (ℓ = ±1) into the two‑dimensional polarization degree of freedom, thereby opening a polarization channel that is otherwise inaccessible.

The experimental platform consists of a magneto‑optical trap of ⁸⁷Rb atoms optically pumped into the |5 S₁/₂,F = 1⟩ ground state. Counter‑propagating σ⁺ driving (Ωd = 2Γ) and σ⁺ coupling (Ωc = 4Γ) beams generate narrowband biphotons via SFWM with an optical depth of ≈48, yielding a pair production rate of 5 × 10⁶ s⁻¹. After polarization filtering, each photon passes through a quarter‑wave plate and a polarizing beam splitter to enforce σ⁺ polarization, then encounters a spatial light modulator (SLM) displaying a fork hologram. The hologram diffracts the photon into ±1 orders, each of which is filtered by a temperature‑stabilized etalon that transmits only the Gaussian LG₀₀ mode. Consequently, only the ℓ = ±1 components survive, defining the effective OAM subspace {|1,1⟩,|−1,−1⟩}.

The −1 order is rotated by a half‑wave plate so that its linear polarization becomes orthogonal to the +1 order. Both paths are recombined on a polarizing beam splitter, which maps the OAM superposition |1,1⟩ + |−1,−1⟩ onto the polarization state (|H,H⟩ + e^{iθ}|V,V⟩)/√2. An electro‑optic phase modulator (EPM) in one arm provides active control of the relative phase θ, enabling deterministic switching between the |Φ⁺⟩ and |Φ⁻⟩ Bell states. By digitally rotating the hologram on one SLM by 180°, the OAM sign is flipped, converting the output to (|H,V⟩ + e^{iθ}|V,H⟩)/√2 and thus generating the |Ψ⁺⟩ and |Ψ⁻⟩ states.

Full quantum‑state tomography is performed for both the native OAM entangled state and the transferred polarization states. Six holograms on each SLM provide the required projective measurements in the OAM basis, while a quarter‑wave plate and linear polarizer implement the six polarization bases (H, V, D, A, R, L). Sixteen joint measurement settings constitute a tomographically complete set. Maximum‑likelihood reconstruction yields the following results: the native OAM Bell state |Φ⁻⟩_L is measured with a fidelity of 96.4 % (statistical uncertainty ±2.2 %). After the OAM‑to‑polarization interface, the four polarization Bell states are produced with fidelities of 93.6 % (|Φ⁺⟩), 92.9 % (|Φ⁻⟩), 92.2 % (|Ψ⁺⟩) and 93.7 % (|Ψ⁻⟩). Corresponding CHSH parameters are S = 2.34 ± 0.12, 2.54 ± 0.13, 2.46 ± 0.12 and 2.42 ± 0.11, all clearly violating the classical bound of 2.

The authors identify the main sources of imperfection as (i) insertion loss and phase noise introduced by the four etalons required for the polarization tomography, (ii) finite diffraction efficiency and alignment errors of the SLM holograms, and (iii) slow thermal drifts that are compensated by re‑optimizing the EPM phase every fifteen minutes. Within the current statistical precision (2–6 % uncertainty) no additional decoherence attributable to the OAM‑to‑polarization mapping itself is observed.

This demonstration is significant for several reasons. First, it preserves the narrowband, memory‑compatible nature of cold‑atom SFWM while providing a deterministic interface to the ubiquitous polarization qubit, facilitating integration of high‑dimensional OAM resources into existing fiber‑based quantum communication infrastructure. Second, the ability to generate all four Bell states on demand using only external optical elements (SLMs, wave plates, EPM) offers a flexible platform for protocols such as quantum key distribution, teleportation, and entanglement swapping without modifying the atomic level structure or pump geometry. Third, the approach is readily transferable to other atomic species or solid‑state ensembles because it relies solely on mode conversion rather than internal atomic engineering.

Future directions suggested include extending the mapping to higher‑dimensional OAM subspaces to create high‑dimensional polarization‑orbital hybrid entanglement, implementing real‑time feedback for phase stabilization, and optimizing the interface for low‑loss transmission through standard single‑mode or few‑mode fibers. Such advances would enable efficient quantum‑memory‑to‑network transduction, a key requirement for scalable quantum repeater architectures and long‑distance quantum networks.