Records — quantum-networking distance and time
Quantum-networking progress is most legible in the distance and time numbers: how far a key was distributed, how far two matter qubits were entangled, how long a qubit kept its phase, how long a quantum state stayed retrievable from a memory. The four tables below collect the best-published numbers across the field, each row grounded in a primary peer-reviewed source.
Click any column header to sort. The criteria for what counts as a record are at the bottom of the page.
QKD distance
QKD records split by protocol family. Decoy-state BB84 is the workhorse for terrestrial fibre; twin-field QKD (TF-QKD) breaks the repeaterless rate–loss bound by interfering single photons at a central station; satellite links use Micius-class downlinks to skip the fibre-loss budget entirely Azuma et al. 2023 .
| Protocol | Year | Distance (km) | Medium | Key rate | Group / lab | Citation |
|---|---|---|---|---|---|---|
| Decoy-state BB84 | 2018 | 421 | Fibre (ultra-low-loss) | 0.25 bit/s @ 405 km | Geneva (Zbinden) | Boaron PRL 121.190502 Boaron et al. 2018 |
| TF-QKD (proposal) | 2018 | 550 | Fibre (modelled) | Beats PLOB bound | Toshiba CRL (Lucamarini) | Lucamarini Nature 557.400 Lucamarini et al. 2018 |
| TF-QKD | 2019 | 90.8 | Fibre (lab) | First sub-PLOB demo | Toshiba CRL (Minder) | Minder NatPhoton 13.334 Minder et al. 2019 |
| TF-QKD (dual-band) | 2021 | 605 | Fibre | 0.0034 bit/s @ 605 km | Toshiba Europe (Pittaluga) | Pittaluga NatPhoton 15.530 Pittaluga et al. 2021 |
| TF-QKD (sending-or-not) | 2023 | 1002 | Fibre (ultra-low-loss) | 0.0034 bit/s @ 952 km | USTC (Pan) | Liu PRL 130.210801 Liu et al. 2023 |
| Decoy-state BB84 (satellite-to-ground) | 2017 | 1200 | Satellite downlink (Micius) | ~kbit/s | USTC / Micius (Liao, Pan) | Liao Nature 549.43 Liao et al. 2017 |
| Entanglement-based (E91/BBM92, satellite) | 2017 | 1203 | Satellite double downlink (Micius) | ~1 Hz pair rate | USTC / Micius (Yin) | Yin Science 356.1140 Yin et al. 2017 |
| Entanglement-based QKD (satellite) | 2020 | 1120 | Satellite double downlink (Micius) | ~0.12 bit/s | USTC / Micius (Yin) | Yin Nature 582.501 Yin et al. 2020 |
| Decoy-state BB84 (microsatellite, intercontinental) | 2025 | 12900 | Microsatellite downlink (Jinan-1, ~23 kg, 500 km LEO) | Real-time keying, one-time-pad image transfer | USTC + Stellenbosch (Li, Pan) | Li Nature (2025) Li et al. 2025 |
Entanglement distribution between matter qubits, by modality
The relevant number here is the separation between two stationary qubits that have been heralded into a Bell pair — not the distance a photon travelled. Quantum frequency conversion (QFC) into the telecom band is the enabling technique on every fibre-based row.
| Modality | Year | Distance (km) | Wavelength path | Technique | Group | Citation |
|---|---|---|---|---|---|---|
| SiV centre (diamond) | 2024 | 35 | 737 nm ↔ 1350 nm telecom O | QFC + nuclear-spin memory | Harvard (Lukin) | Knaut Nature 629.573 Knaut et al. 2024 |
| Neutral atom (Rb) | 2022 | 33 | 780 nm ↔ 1517 nm telecom L | QFC, single atoms in optical traps | LMU Munich (Weinfurter) | van Leent Nature 607.69 Leent et al. 2022 |
| NV centre (diamond) | 2021 | ~0.03 | 637 nm native (no QFC) | Three-node repeater, swapping | Delft (Hanson) | Pompili Science 372.259 Pompili et al. 2021 |
| NV centre (deterministic teleport) | 2022 | ~0.03 | 637 nm native | Memory-assisted teleportation | Delft (Hanson) | Hermans Nature 605.663 Hermans et al. 2022 |
| Trapped ion (Ba⁺/Yb⁺, long-lived) | 2026 | 10 | Visible ↔ telecom (QFC) | Long-lived memory; rate exceeds loss | USTC (Liu, Bao, Pan) | Liu Nature 652.51 Liu et al. 2026 |
| Trapped ion (Ca⁺) | 2023 | 0.23 | 854 nm ↔ 1550 nm telecom C | QFC, two-photon BSM | Innsbruck (Northup) | Krutyanskiy PRL 130.050803 Krutyanskiy et al. 2023 |
| Trapped ion (Sr⁺) | 2020 | 0.002 | 422 nm native | High-rate two-photon BSM | Oxford (Lucas) | Stephenson PRL 124.110501 Stephenson et al. 2020 |
| Atomic ensemble (DLCZ) | — | — | Survey reports demonstrations to ~50 km | QFC of DLCZ heralding photons | various; see Meddeb 2025 §3.7 | [primary cite pending] Meddeb 2025 |
| Superconducting transmon | — | 0.030 | Microwave, cryogenic waveguide | Direct microwave link (no transduction) | survey (Meddeb 2025 §3.7) | [primary cite pending] Meddeb 2025 |
| Quantum dot (single emitter) | — | — | — | Spin–photon entanglement reported; remote-spin entanglement record at network range [unverified] | — | [unverified] |
"Distance" is the fibre or free-space path length between the matter qubits, not photon path through repeaters. The trapped-ion and transmon rows reflect that those platforms are early in the remote-entanglement-over-fibre curve — neither has a published demo past a few hundred metres.
Coherence times by qubit modality
T2 is the timescale over which a qubit retains phase information. T2* is the inhomogeneous-dephasing time without active correction. Long T2 is achieved with dynamical decoupling (XY8, CPMG) and isotopic purification; numbers below are the best published values for each modality and reflect the measurement protocol in the source.
| Modality | T₂ | Conditions | Year | Group | Citation |
|---|---|---|---|---|---|
| Atomic ensemble (¹⁵¹Eu³⁺:Y₂SiO₅ nuclear spin) | 6 h | 2 K, dynamical decoupling at zero-first-order-Zeeman field | 2015 | ANU (Sellars) | Zhong Nature 517.177 Zhong et al. 2015 |
| Trapped ion (¹⁷¹Yb⁺ ground state) | 92 min | Magnetic-field-insensitive clock state, dynamical decoupling | 2021 | Tsinghua (Kim) | Wang NatComms 12.233 Wang et al. 2021 |
| NV centre (¹³C nuclear spin) | 75 s | 1.5 K, isotopically purified diamond, XY8 decoupling | 2019 | Delft (Taminiau) | Bradley PRX 9.031045 Bradley et al. 2019 |
| Neutral atom (Rb ground state) | ~10 s | Magic-wavelength optical-lattice trapping | — | various | see Meddeb 2025 §2.1 Meddeb 2025 |
| SiV centre (²⁹Si nuclear spin) | ~1 s | Mixing-chamber temperatures, XY8 decoupling | 2024 | Harvard (Lukin) | Knaut Nature 629.573 Knaut et al. 2024 |
| Superconducting transmon (2D) | 1.68 ms (T₁); T₂ₑ ≳ T₁ | ~10 mK, tantalum-on-silicon, refined junction deposition | 2025 | Princeton (Bland, de Leon, Houck) | Bland Nature 647.343 Bland et al. 2025 |
| NV centre (electron spin) | 1.2 ms | Room temperature, ¹²C purified, dynamical decoupling | — | Stuttgart / Delft | see Azuma RMP 95.045006 Azuma et al. 2023 |
| Quantum dot (electron spin) | ~300 ns | InGaAs, optical dynamical decoupling | — | various | see Meddeb 2025 Meddeb 2025 |
| Photonic (flying qubit) | — | Loss-limited rather than dephasing-limited; coherence not directly comparable | — | — | — |
Rows where the year or citation reads as "—" are well-established field benchmarks reported by the survey literature (Meddeb 2025 §2.1; Azuma RMP 2023) rather than a single primary paper.
Memory storage time
Storage time is the demonstrated retrieval of a quantum state from a memory after a controllable delay, not the bare coherence time of the underlying qubit. Numbers below pair the storage duration with the fidelity at that time, where reported.
| Platform | Storage time | Fidelity at that time | Year | Group | Citation |
|---|---|---|---|---|---|
| Eu:YSO atomic-frequency-comb memory (3-level AFC + spin control) | 1 h | ~0.96 retrieval fidelity (classical light) | 2021 | USTC (Li, Zhou) | Ma NatComms 12.2381 Ma et al. 2021 ; reviewed in Tittel et al. 2025 |
| NV ¹³C nuclear-spin register (10 qubits) | 75 s | Single-qubit fidelity preserved at T₂ scale | 2019 | Delft (Taminiau) | Bradley PRX 9.031045 Bradley et al. 2019 |
| Cold-atom ensemble (Rb spin wave) | 16 s | ~0.5 retrieval efficiency at >1 min | 2013 | Georgia Tech (Kuzmich) | Dudin PRA 87.031801 Dudin et al. 2013 |
| Trapped-ion remote-entanglement memory | ~550 ms | Heralded ion-ion entanglement lifetime over 10 km fibre | 2026 | USTC (Liu, Bao, Pan) | Liu Nature 652.51 Liu et al. 2026 |
| SiV centre ²⁹Si nuclear memory (network-grade) | 200 ms | 0.69(7) Bell-state fidelity over 35 km fibre | 2024 | Harvard (Lukin) | Knaut Nature 629.573 Knaut et al. 2024 |
| Superconducting cavity (3D) | ~1 ms | Logical-qubit gate teleportation between cavities | 2018 | Yale (Schoelkopf) | Chou Nature 561.368 Chou et al. 2018 |
| Photonic delay-line (fibre loop) | ~1 ms | Loss-limited; ~0.2 dB/km accumulation | — | various | see Heshami JMO 63.2005 Heshami et al. 2016 |
| Rare-earth-doped waveguide (Tm:LiNbO₃) | ~7 ns | Heralded entangled-photon retrieval; AFC delay | 2011 | Calgary (Tittel) | Saglamyurek Nature 469.512 Saglamyurek et al. 2011 |
What counts as a record
Inclusion criteria for the tables above:
- Primary peer-reviewed source. Nature, Science, PRL, PRA, Nature Photonics, Nature Communications, and equivalents. Preprints accepted only when the published version is not yet available. Press releases and conference talks are not enough on their own.
- Demonstrated, not projected. Modelled extrapolations and proposal-only numbers are excluded except where flagged as such (e.g. Lucamarini 2018's TF-QKD proposal).
- Modality means the qubit that stored or carried the state — for entanglement-distribution rows that is the matter qubit being entangled, not the photonic carrier.
- Coherence vs storage are different. A long bare T2 does not imply a long demonstrated storage; the storage table records actual write-store-retrieve experiments, not Ramsey or spin-echo measurements alone.
- Records age. Rows marked "[unverified]" or "—" are placeholders for entries we have not yet tied to a primary source in the indexed library. They will be replaced as the catalogue is updated.
The field's compilation papers — Meddeb 2025 Table 5 for memory TRL, Azuma 2023 §III for repeater milestones, Wehner 2018 for stage gates — are the survey-level companion reading Meddeb 2025 Azuma et al. 2023 .