Implementation of nonlocal multi-photon interference by mode swapping

AI summary

Overview

Research area: Quantum optics and optical quantum information — specifically multi-photon interference, entanglement generation, and nonlocal correlations between spatially separated optical systems.

Technical level: Intermediate. The algebra is elementary (two-photon Fock states in a four-mode basis), but comfort with second quantization, creation operators, and post-selection is assumed.

Scope (one sentence): The paper shows theoretically that exchanging one optical mode between two spatially separated two-mode systems — a "mode swap" — implements a nonlocal version of the Hong-Ou-Mandel effect, producing a maximally entangled state between two local qubits when one photon is detected at each location.

What This Paper Is About

Multi-photon interference, such as the Hong-Ou-Mandel effect, is normally observed when photons meet at a beam splitter and interfere between transmission and reflection. The authors ask whether an equivalent interference can be produced when the interfering alternatives are not "transmit or reflect" but "swap a mode between two locations or leave it alone" — with the photons ending up detected at two spatially separated sites.

The goal is to give a minimal optical circuit that generates entanglement between two local multi-mode systems and to explain what the resulting nonlocal interference physically means, in terms of modes that are shared between the two locations rather than confined to one.

Key Contributions

1. A basic mode-swap circuit for nonlocal interference. The paper introduces a circuit in which one mode of system A is exchanged with one mode of system B inside a pair of local Mach-Zehnder interferometers, with no physical interaction between the two sides.

2. Explicit state derivation. It writes the pre-swap state as an equal-weight superposition of four two-photon Fock states (Eq. 1), the post-swap state (Eq. 2), and shows that projecting onto the subspace with one photon in A and one in B leaves a maximally entangled two-qubit state, up to normalization by the post-selection probability (Eq. 3).

3. A mode-based explanation of the effect. By expressing the input mode creation operators as superpositions of output mode operators (Eq. 4), the authors identify two "nonlocal modes," each an equal superposition of a mode in A with a mode in B, and show that both photons leave the circuit in the same nonlocal mode (Eqs. 5–6). This is presented as nonlocal photon bunching.

4. A framing of multi-photon interference as mode coherence. The paper argues that multi-photon interference is an effect caused by nonlocal phase coherence between spatially separated modes, and proposes nonlocal modes as a practical bookkeeping tool for larger multimode, multiphoton systems.

Main Findings

• The swap itself creates entanglement. Before the swap, the state in Eq. 1 is a product of local two-mode states; after the swap, the state in Eq. 2 is entangled across A and B, supported on the four-dimensional subspace {|0,0⟩, |0,1⟩, |1,0⟩, |1,1⟩} of each side.

• Post-selection gives a maximally entangled qubit state. Because |0,0⟩ and |1,1⟩ lie outside the qubit subspace {|0,1⟩, |1,0⟩}, the authors post-select on one photon per side. The normalized result is (|0,1;1,0⟩ − |1,0;0,1⟩)/√2, a maximally entangled state for which any parallel linear-optics transformation applied to A and B gives perfectly anti-correlated outcomes.

• The interference is between "swapped" and "not swapped" photon-number distributions. The entangled state is a superposition of a Fock state with no photons in the swapped modes and one with both photons in the swapped modes; indistinguishability means the photons' swap history cannot be known, so these distributions interfere.

• The two input modes reduce to two nonlocal output modes. Both input creation operators can be written as orthogonal superpositions of only two modes, each of which is an equal superposition of one mode from A and one from B.

• Nonlocal bunching correlates separated detectors. When one photon is detected in each location, a detection in â₁ always coincides with a detection in b̂₂, and a detection in â₂ with b̂₁. The paper states that the probabilities of detecting photons in â₁ and b̂₁, or in â₂ and b̂₂, are "close to zero." In the post-selected state of Eq. 6 those terms do not appear at all.

• Photons exit in the same nonlocal mode despite spatial separation. The authors describe this as nonlocal photon bunching, where separately detected photons can be traced back to the same original mode.

• No experimental data are reported. The paper contains no measured counts, fidelities, visibilities, or post-selection probability values; the post-selection probability appears only as the symbol P_Π.

Methodology in Plain English

The work is entirely theoretical and algebraic. The authors define two qubits, each encoded in a single photon occupying one of two spatial modes at a given location, then model a mode swap as literally exchanging one mode label between the two locations. They write the joint state of the two photons in a four-mode photon-number basis, apply the swap by relabeling modes, and then restrict to the cases where one photon remains at each location.

To explain why the result is an interference effect rather than a trivial relabeling, they work backwards: they take the creation operators of the two input modes and rewrite them as combinations of the output-mode creation operators. Applying both input operators to the vacuum then shows which output mode pairs can be populated and which cancel. This operator bookkeeping makes the interference visible as a pairing of one mode in A with one mode in B.

Why This Matters

The result gives a route to entanglement distributed between two distant parties using a mode exchange rather than photons meeting at a shared beam splitter, which fits protocols where quantum information must be delivered to separate locations. It also offers a conceptual claim: multi-photon interference is driven by coherence between nonlocal modes, not only by photons coinciding in space.

Potential applications the paper's framing points toward:

• Distributed quantum information protocols that require entanglement between spatially separated parties, which the authors cite as a particular suitability of this implementation.
• Entanglement generation in post-selected optical circuits, the context in which the authors situate the work (heralded and output-post-selected schemes).
• Large-scale quantum optical networks, which the authors cite as a motivation for an efficient construction using simple mode swapping.
• Efficient description of multimode, multiphoton optical circuits, using nonlocal modes to track nonclassical correlations when information is encoded in photon-number distributions over a local mode set.

Industry relevance: The work is foundational rather than device-oriented. Its relevance is to developers of photonic quantum computing and quantum communication hardware who need compact, scalable ways to entangle separate modules; the paper itself reports no hardware, implementation, or performance data.

Future Directions

• Extending beyond two photons and two modes per side. The authors explicitly propose applying the multi-mode analysis to interferences of more than two photons, and to generating and distributing entanglement between separate multi-mode systems.
• Using nonlocal modes as a general design tool. The paper presents the analysis as "a promising starting point for a systematic exploration" of interacting multi-mode multi-photon systems; developing that into a general method is left open.
• Experimental realization. No experiment, implementation, or error analysis is reported, so demonstrating the mode swap and the predicted correlation patterns in the laboratory remains an open step.
• Quantifying the post-selection cost. The post-selection probability appears only symbolically as P_Π; how it behaves for larger photon numbers and mode counts is not reported and would matter for scaling.

Target Audience

Quantum optics researchers and graduate students working on multi-photon interference and entanglement generation; photonic quantum information theorists interested in distributed entanglement and multimode encodings; and readers who want a clear, algebraically explicit account of what "nonlocal" means for multi-photon interference. Experimentalists will find the circuit concept useful but will not find implementation details or data here.

Authors’ abstract

Multi-photon interference can be observed using independently generated photons as input. In the most simple case, these photons meet up at a beam splitter, resulting in quantum interference between transmission and reflection of the photons. Here, we show that non-local multi-photon interference can be implemented by using a mode swap operation to generate entanglement between the photons detected in the outputs of two spatially separated interferometers. The spatial separation of the output photons makes this implementation of multi-photon interference particularly suitable for quantum protocols that distribute quantum information to different parties.