Unheralded-error and large-N scalability limits of error-detected repeat-until-success N-photon cluster state generation with a quantum dot in an optical microcavity

Liang, Zheng, Zhang and Zhang (Front. Phys. 15, 21601, 2020) proposed generating an N-photon linear cluster state by sending photons one by one through an error-detected controlled-phase gate mediated by an electron spin of a charged quantum dot (QD) in a single-sided optical microcavity: any imbalance of the hot- and cold-cavity reflection amplitudes is converted into a heralded failure (a click on a photon detector), so that, conditioned on no clicks, the cluster-state fidelity is unity regardless of side leakage, detuning, cooperativity, and photon number, and failed attempts are simply retried without re-initializing the spin. This unity-fidelity claim, however, holds only within an error model in which every imperfection is heralded. The open problem is to determine the realistic performance and scaling limits of this repeat-until-success protocol when the non-heralded experimental nonidealities are included — electron-spin decoherence accumulating over the stochastic retry sequence, photon loss into undetected modes, detector inefficiency and dark counts (spurious detector clicks), finite photon bandwidth and mode mismatch, and imperfect final spin readout — and to establish the maximum cluster size achievable as a function of the device parameters.

Background

A cluster state is a multi-qubit entangled state on a graph, introduced by Briegel and Raussendorf, that serves as the universal resource for measurement-based (one-way) quantum computation: the computation is driven by sequential single-qubit measurements rather than by a circuit of gates. Photonic cluster states are especially attractive because photons decohere weakly, but they are hard to produce deterministically, which motivates using a stationary spin as an entangling ancilla that stitches an incoming photon string into a cluster.

The platform of the source paper is a singly charged self-assembled quantum dot embedded in a single-sided micropillar optical cavity (one perfectly reflecting mirror, one partially transmitting mirror). By spin-selection rules, a circularly polarized photon couples to the optical trion transition (a charged exciton, $X^-$ , consisting of two electrons and a hole) only for one orientation of the excess electron spin, so the photon reflection coefficient depends on the spin state. Input-output theory of cavity QED gives a 'hot-cavity' (coupled) reflection coefficient $r_1$ and a 'cold-cavity' (uncoupled, $g=0$ ) coefficient $r_0$ , which differ whenever the cavity is imperfect: the side leakage rate $\kappa_s$ (field decay into unwanted modes, relative to the decay $\kappa$ through the accessible mirror), the exciton dipole decay rate $\gamma$ , the coupling $g$ , and detunings of the photon from the cavity and trion frequencies all enter $r_1$ and $r_0$ . The cooperativity $C = g^2/(\gamma\kappa_T)$ with $\kappa_T = \kappa_s + \kappa$ measures coupling against dissipation; $C = 1$ is a typical value in the Purcell regime (cavity-enhanced spontaneous emission) of present devices.

Because $r_1 \neq r_0$ in any real device, spin-photon gates based on this interface suffer infidelity. Li and Deng (Phys. Rev. A 94, 062310, 2016) introduced the error-rejecting paradigm for such gates: with suitable linear optics and a passive attenuator that equalizes the successful amplitudes, the unwanted $r_1+r_0$ component is routed to a photon detector, so every imperfection of the input-output process becomes a heralded failure rather than an infidelity, and the gate can simply be repeated until the detector does not click. Liang et al. (Front. Phys. 15, 21601, 2020) built on this paradigm an N-photon linear cluster-state source: each photon passes through an error-detected hybrid controlled-phase (C-phase) gate block, a microwave Hadamard pulse (a spin rotation that interchanges the computational basis with the superposition basis) is applied to the electron spin between photons, and at the end the spin is measured to disentangle it from the photon string (a byproduct correction on the last photon — a Pauli flip conditioned on the random spin-measurement outcome — fixes the outcome branch). Conditioned on no detector clicks ever occurring, the output is exactly an N-photon cluster state with fidelity $F = |\langle\psi_r|\psi_i\rangle|^2 = 1$ , independent of $\kappa_s/\kappa$ , of the detuning $(\omega_c-\omega)/\kappa$ , of $C$ , and of $N$ ; the success probability of one full pass is $\eta_s = ((r_1-r_0)/2)^{2N}$ , decaying exponentially in $N$ , and a recycling analysis (retrying only the failed photon, which leaves the spin state intact) gives $\eta = \eta_s/(1-\eta_e)$ with a computable heralded-error probability $\eta_e$ . The paper plots these quantities for $N = 2, 3, 4$ at representative parameters ( $C=1$ , $\gamma/\kappa = 0.1$ ), e.g. a 3-photon success probability of about 0.24–0.27.

Two assumptions delimit the published analysis. First, every error source in the model is heralded: the derivation conditions on the photon emerging and the detector behaving perfectly, so any physical process that produces no click — decoherence of the electron spin during the (randomly long) retry sequence and the interleaved microwave pulses, loss of a photon into undetected modes (side leakage en route, optical elements, or the attenuator arm), a photon in the error arm failing to register on an inefficient detector, or dark counts falsely aborting or mis-heralding an attempt — silently degrades the fidelity that the scheme claims to protect. Second, timing: the paper assumes the photon spacing $\Delta t$ stays within the electron-spin coherence time (it cites 0.44 µs in GaAs quantum dots) and asserts the scheme 'can work many times before the decoherence time'; but in repeat-until-success the expected number of attempts per accepted photon grows as the per-photon success probability falls, so the total generation time is a random variable that grows with $N$ and worsens in weak-coupling, high-leakage devices — precisely the regime where the error detection is most needed. The paper analyzes neither this time budget nor any concrete large- $N$ scaling statement.

A parallel, deterministic family of proposals generates photonic cluster strings from a dot spin by repeated excitation and spin rotation (Lindner and Rudolph, Phys. Rev. Lett. 103, 113602, 2009, and descendants), without heralding. That family has by now received exactly the analysis whose absence defines the present problem: complete treatments including spin decoherence, phonon baths, finite lifetimes, and cavity-environment imperfections with explicit regime maps (Tiurev et al. 2022; Prasad et al. 2025; Köcher and Schumacher 2026) and experimental demonstrations (e.g. Peniakov et al. 2026, telecom C-band, cluster-process fidelity (fidelity of the full state-generating process) 0.71). The heralded, repeat-until-success family of Liang et al. — whose distinctive selling point, unity conditional fidelity, is precisely the property that unheralded errors break — has not been given any comparable treatment.

Problem Statement

For the error-detected repeat-until-success N-photon cluster-state protocol of Liang et al. (2020) — sequential photons interacting with an electron spin in a single-sided QD microcavity through an error-detected C-phase gate block, with detector clicks heralding failure and triggering a retry of the failed photon without spin re-initialization — determine the realistic performance and scalability limits when the analysis is extended beyond the all-errors-heralded model. Specifically: (i) build an end-to-end theoretical description that includes the physically distinct non-heralded channels: electron-spin decoherence during the randomly long retry sequence and the interleaved microwave Hadamard pulses, photon loss into undetected modes (side leakage en route, optical elements, attenuator arm, and error-arm photons that fail to trigger the detector), detector inefficiency and dark counts, finite photon bandwidth and cavity-mode mismatch beyond the monochromatic reflection coefficients $r_0$ and $r_1$ , and imperfect final electron-spin readout — with each channel either modeled or shown negligible in the considered regime; (ii) within this model, derive or compute the achievable cluster-state fidelity and the total expected generation time as functions of $N$ and the device parameters ( $g$ , $\kappa$ , $\kappa_s$ , $\gamma$ , detunings, spin coherence time, detector efficiency); and (iii) determine the maximum cluster size attainable at a fiducial acceptance level of $F \ge 0.9$ as an explicit function of the device parameters on which it depends — the cooperativity $C$ , the decay ratios $\gamma/\kappa$ and $\kappa_s/\kappa$ , the detuning $(\omega_c-\omega)/\kappa$ , the absolute cavity decay rate $\kappa$ (or the coupling $g$ ) that sets the per-attempt timescale, the detector dark-count rate, the final spin-readout infidelity, the electron-spin coherence time, and the detector efficiency — anchored at the source paper's representative values of $C = 1$ and $\gamma/\kappa = 0.1$ , replacing the paper's qualitative assertion that the scheme 'can work many times before the decoherence time' with a quantitative scaling statement. The analysis must reproduce the source paper's idealized limits — unity conditional fidelity and the closed-form success probabilities of Eqs. (19)–(21) — when the added non-heralded channels are switched off.

References

Lei-Xia Liang, Yan-Yan Zheng, Yuan-Xia Zhang, Mei Zhang. Error-detected N-photon cluster state generation based on the controlled-phase gate using a quantum dot in an optical microcavity. Frontiers of Physics 15(2), 21601 (2020). DOI: 10.1007/s11467-019-0931-2. https://doi.org/10.1007/s11467-019-0931-2 (source paper)
Netanel H. Lindner, Terry Rudolph. Proposal for Pulsed On-Demand Sources of Photonic Cluster State Strings. Physical Review Letters 103, 113602 (2009). DOI: 10.1103/PhysRevLett.103.113602. https://doi.org/10.1103/PhysRevLett.103.113602
Tao Li, Fu-Guo Deng. Error-rejecting quantum computing with solid-state spins assisted by low-Q optical microcavities. Physical Review A 94, 062310 (2016). DOI: 10.1103/PhysRevA.94.062310. https://doi.org/10.1103/PhysRevA.94.062310
Konstantin Tiurev, Martin Hayhurst Appel, Pol Llopart Mirambell, Mikkel Bloch Lauritzen, Alexey Tiranov, Peter Lodahl, Anders Søndberg Sørensen. High-fidelity multi-photon-entangled cluster state with solid-state quantum emitters in photonic nanostructures. Physical Review A 105, L030601 (2022). DOI: 10.1103/PhysRevA.105.L030601. https://doi.org/10.1103/PhysRevA.105.L030601 (preprint arXiv:2007.09295, 2020: https://arxiv.org/abs/2007.09295)
Rohit Prasad, Simon Reiß, Giora Peniakov, Yorick Reum, Peter van Loock, Sven Höfling, Tobias Huber-Loyola, Andreas Theo Pfenning. Analytical Fidelity Calculations for Photonic Linear Cluster State Generation (2025). arXiv:2505.11078. https://arxiv.org/abs/2505.11078
Giora Peniakov, Reza Hekmati, Johannes Michl, Mohamed Helal, Moritz Meinecke, Jochen Kaupp, Yorick Reum, Martin Kamp, Monika Emmerling, Andreas Theo Pfenning, Sven Höfling, Tobias Huber-Loyola. Photonic Cluster State Generation from a Quantum Dot Emitting in the Telecom C-band (2026). arXiv:2607.04896. https://arxiv.org/abs/2607.04896
Nikolas Köcher, Stefan Schumacher. Deterministic Generation of Linear Photonic Cluster States with Semiconductor Quantum Dots: A Detailed Comparison of Different Schemes (2026). arXiv:2607.09373. https://arxiv.org/abs/2607.09373
Yi-Ming Wu, Gang Fan, Fang-Fang Du. Error-detected three-photon hyperparallel Toffoli gate with state-selective reflection. Frontiers of Physics 17, 125062 (2022). DOI: 10.1007/s11467-022-1172-3. https://doi.org/10.1007/s11467-022-1172-3
Jiu-Ming Li, Jun-Yan Liu, Yuan-Yuan Liu, Xiao-Ming Xiu, Shao-Ming Fei. General quantum computation on photons assisted with double single-sided cavity system. Physics Letters A 590, 131854 (2026). DOI: 10.1016/j.physleta.2026.131854. https://doi.org/10.1016/j.physleta.2026.131854 (preprint arXiv:2507.22773, 2025: https://arxiv.org/abs/2507.22773)