Dual-Rail Photonic Qubit

Figure

Description

The dual-rail photonic qubit encodes one logical qubit into the single-photon subspace of two optical modes:

$|0_L⟩=|1{⟩}_a|0{⟩}_b,|1_L⟩=|0{⟩}_a|1{⟩}_b$

This is the canonical path-encoded qubit for linear-optical quantum computing and for many integrated-photonics architectures. The logical state is carried by which mode contains the photon, not by a microscopic two-level system inside the photon. More generally, arbitrary qubit states live in the one-photon manifold

$|{\psi }_L⟩=\alpha |1,0⟩+\beta |0,1⟩,$

with $|\alpha {|}^2+|\beta {|}^2=1$.

Passive linear optics acts naturally on this subspace. Beam splitters mix the two modes, while phase shifters control their relative phase, so arbitrary single-qubit SU(2) operations are available without requiring optical nonlinearities. A phase shift that is identical on both rails is only a global phase and is therefore irrelevant, but differential path-length or phase noise between the rails directly dephases the qubit.

Entangling gates are the hard part. With linear optics alone there is no deterministic photon-photon interaction, so two-qubit logic is implemented through interference, ancilla photons, measurement, and feed-forward, as in the Knill-Laflamme-Milburn (KLM) scheme and its cluster-state / fusion-based descendants.

The dominant hardware failure mode is photon loss, which usually ejects the state from the computational subspace into vacuum and therefore behaves like a flagged erasure rather than an unknown Pauli error.

Hamiltonian

In the absence of optical nonlinearities, passive two-mode photonic control is generated by a number-conserving bilinear Hamiltonian

$H={\sum }_{i,j\in \{a,b\}}{h}_{ij}{a}_i^{†}{a}_j.$

A beam splitter with a real phase convention is generated by

$H_{\mathrm{BS}}=\hslash \kappa \left(a^{†}b+b^{†}a\right),$

which mixes the two rails and acts as a logical $X$-type rotation within the single-photon dual-rail subspace.

A relative phase shift between the rails can be written as

$U_{\varphi }=e^{i\varphi b^{†}b},$

so that

$|0_L⟩\mapsto |0_L⟩,|1_L⟩\mapsto e^{i\varphi }|1_L⟩,$

which is a logical $Z$ rotation up to an irrelevant global phase convention.

These number-conserving linear-optical generators are sufficient for arbitrary single-qubit control. By contrast, universal two-qubit control requires measurement-induced nonlinearities, ancillary photons, or resource-state constructions rather than a native interacting Hamiltonian between dual-rail qubits.

Motivation

• Dual-rail encoding is the cleanest qubit abstraction for path-encoded linear optics: one photon, two modes, no need for a material qubit memory element.
• Common-mode optical phase is irrelevant, while photon loss is often detectable as a vacuum event, making erasure-aware fault tolerance especially natural.
• The same encoding underlies KLM, photonic cluster-state MBQC, and fusion-based photonic architectures.
• It maps naturally onto bulk optics, fiber interferometers, and integrated silicon / silicon-nitride photonic circuits.

Experimental Status

Early constructive proposal, Adami and Cerf (1998 preprint / 1999 LNCS):

• Showed how quantum circuits can be mapped onto linear-optical networks using multi-rail single-photon encodings.
• Established the dual-rail picture as a practical computational encoding rather than just an interferometric toy model.

Scalable LOQC proposal, KLM (2001):

• Proved that efficient quantum computation is possible using dual-rail photonic qubits, ancilla photons, linear optics, and photon counting with feed-forward.
• Made clear that deterministic matter-mediated optical nonlinearities are not strictly required for scalability in principle.

First canonical two-qubit gate demonstration, O’Brien et al. (2003):

• Demonstrated an all-optical probabilistic CNOT operating on dual-rail qubits.
• Verified entangling behavior by generating all four Bell states for suitable logical inputs.

Recent architecture and state-generation relevance (2023-2024):

• Fusion-based photonic computation (Bartolucci et al. 2023) treats dual-rail or closely related photonic encodings as the natural physical qubit layer for fault-tolerant optics.
• Blau et al. (2024) reported a 4-photon dual-rail cluster state with lower-bound fidelity 0.81, a useful modern benchmark for multi-photon dual-rail state generation.

Key Metrics

Metric	Value	Notes	Fidelity reference
Encoding overhead	2 optical modes / logical qubit	One photon delocalized across two rails	Adami and Cerf 1999
Native 1Q control	Deterministic SU(2)	Beam splitters + phase shifters act within the one-photon subspace	Reck et al. 1994
Basic KLM entangling-gate success	1/16	Unboosted linear-optical entangling gate in the original KLM construction	Knill et al. 2001
Recent dual-rail cluster-state benchmark	4 photons, fidelity lower bound 0.81	Dual-rail photonic cluster state from dual-color photon-pair sequences	Blau et al. 2024
Dominant error channel	Photon loss (erasure)	Vacuum leakage out of the computational subspace dominates over in-subspace decoherence	Kok et al. 2007

References

Foundational proposals and theory

• C. Adami and N. J. Cerf, “Quantum Computation with Linear Optics,” Lecture Notes in Computer Science (1999) — arXiv:quant-ph/9806048
• E. Knill, R. Laflamme, and G. J. Milburn, “A scheme for efficient quantum computation with linear optics,” Nature 409, 46 (2001) — arXiv:quant-ph/0006088
• M. Reck, A. Zeilinger, H. J. Bernstein, and P. Bertani, “Experimental realization of any discrete unitary operator,” Phys. Rev. Lett. 73, 58 (1994)
• P. Kok et al., “Linear optical quantum computing with photonic qubits,” Rev. Mod. Phys. 79, 135 (2007) — arXiv:quant-ph/0512071

Experimental and architectural milestones

• J. L. O’Brien, G. J. Pryde, A. G. White, T. C. Ralph, and D. Branning, “Demonstration of an all-optical quantum controlled-NOT gate,” Nature 426, 264 (2003)
• S. Bartolucci et al., “Fusion-based quantum computation,” Nat. Commun. 14, 912 (2023) — arXiv:2101.09310
• M. Blau, R. Oliver, X. Ji, M. Lipson, and A. L. Gaeta, “Realization of a Dual-Rail Photonic Cluster State,” CLEO 2024, FM1K.5

Linked Papers

• adami-1998-quantum-computation-with-linear
• knill-2001-klm
• reck-1994-experimental-realization-any
• kok-2005-review-article-linear-optical
• obrien-2003-all-optical-cnot
• bartolucci-2023-fbqc
• blau-2024-dual-rail-photonic-cluster-state

Evergreen context

• erasure-error-vs-pauli-error — photon loss naturally appears as a flagged vacuum event, so dual-rail photonics strongly rewards erasure-aware decoding rather than pretending everything is depolarizing noise.
• noise-bias-and-asymmetric-error-channels — scalable photonic schemes only look attractive when loss remains the dominant, interpretable error channel.
• threshold-theorem — KLM, cluster-state optics, and fusion-based approaches are all different ways of turning nondeterministic optical primitives into a fault-tolerant computational model.

Related Entries

• linear-optical-photonic-qubit — broader linear-optical computing architecture built around encodings like dual rail
• time-bin-photonic-qubit — another canonical photonic encoding, especially for fiber links
• photonic-cluster-state-mbqc-qubit — measurement-based photonic architecture that often uses dual-rail-style physical qubits
• fusion-based-photonic-qubit — fault-tolerant photonic architecture built from probabilistic fusion operations
• dual-rail-superconducting-qubit — superconducting analog of the same logical encoding idea
• erasure-qubit — umbrella entry for platforms where loss-like flagged errors become a central design feature