{"id":"484d4c00-1220-4874-8062-46d7022b3203","arxiv_id":"2602.14426","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The enhanced electron readout contrast in parallel-nuclear exchange-coupled donor pairs is caused by a second spin-up tunneling event to the SET during one readout period.","lead":"This paper explains why reading out an electron spin on a pair of coupled phosphorus atoms in silicon gives a stronger signal when the two atomic nuclei are aligned parallel: an extra electron hops to the detector, producing a second current blip. The work turns a previously unexplained ~40% readout contrast jump into a known mechanism that can guide better spin-qubit readout.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Two-blip evidence does not uniquely prove the exchange-mediated second tunneling of electron 1 unless electron 2's tunnel coupling to the SET is experimentally excluded; the paper only asserts this.","rationale":"Reader's verdict CONDITIONAL is appropriate, but the most load-bearing assumption is not Eq. 5; it is the unverified single-electron tunnel coupling. Eq. 5 is a secondary consistency check; even if the independence assumption fails, the double-blip histograms would still support an additional tunneling event. The tunnel-coupling assumption, however, is what makes the two blips evidence for the specific exchange-mediated re-tunneling of electron 1 rather than a trivial two-electron readout. Because the paper provides no control (e.g., |↓1↑2> anti-parallel readout) or raw-trace audit, the central claim is not uniquely established. This does not contradict the authors; it identifies missing evidence. The concern is concrete and testable from the open data or a straightforward experiment. For these reasons, the verdict remains CONDITIONAL; a clean control would move it toward ACCEPT.","tokens_in":16602,"tokens_out":21687,"duration_ms":230144,"concrete_test":"Perform a control readout in the anti-parallel nuclear configuration with the electrons prepared in |↓1↑2> (electron 1 spin-down, electron 2 spin-up). If only electron 1 is tunnel-coupled, the blip rate should equal the thermal false-positive background (~0.21 from |T−>); a significantly higher rate would indicate electron 2 also tunnels, invalidating the unique assignment of the second blip to electron 1. Alternatively, from the existing Dryad raw traces, measure the SET current between consecutive blips in the parallel |T+> dataset: the exchange mechanism requires a return to the Coulomb-blockade baseline before the second blip, whereas simultaneous/sequential ionization of two tunnel-coupled donors can produce a different inter-blip current/intervals; also check whether the second blip's amplitude and tunnel-out statistics match electron 1.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section IV.A states 'Only electron 1 was tunnel-coupled to the SET and hence could be read out via spin-dependent tunneling,' but no measurement supporting this is shown. This is the linchpin of the central claim. The observed two-blip statistics (Fig. 2c,e: 1.77±0.15 vs 1.02±0.09) and the state-resolved histograms (Fig. 2f-i: 0.21 for |T−>, 1.67 for |T+>, 1.12 for |TX>) are equally consistent with a simpler alternative: if both donor electrons are tunnel-coupled to the SET, then |T−> has zero blips, |T+> has two (one from each spin-up electron), and a 50/25/25 mixture of T0/T−/T+ yields ~1.0 blips on average. This alternative would not require the exchange-mediated S/T0 mechanism of Fig. 2(a) at all. The paper's specific assertion that the same electron (electron 1) tunnels off, is replaced by a spin-down electron, and then tunnels off again is only testable if electron 2 is known not to tunnel. No control experiment in the anti-parallel |↓1↑2> state is presented to demonstrate that electron 2 produces no blips, and the Dryad dataset is not audited in the manuscript. Until this is excluded, the double-blip histograms do not uniquely support the proposed mechanism.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a mechanism for the enhanced electron readout contrast observed in exchange-coupled donor pairs when the nuclear spins are parallel. In the J≫ΔA regime, after the spin-up electron on donor 1 tunnels to the SET and a spin-down electron tunnels back, the readout electron still has significant |↑1> character from the weakly hybridized T0/S manifold and tunnels off a second time, producing two current blips in one readout window. This is supported by blip-counting statistics (1.77±0.15 blips per readout for parallel nuclei vs 1.02±0.09 for anti-parallel), state-resolved histograms for T−, TX, and T+ initial states, and an adiabatic spectroscopy comparison using Eq. (5), P(⇓⇓/⇑⇑)=1−(1−P(⇓⇑/⇑⇓))².","tokens_in":16924,"tokens_out":7537,"duration_ms":79259,"significance":"If established, the result would explain a long-standing, unexplained observation in donor-based qubit readout and would guide improvements in readout fidelity. The paper's strengths are its use of direct experimental statistics (blip counting) rather than only indirect contrast measurements, and the construction of a parameter-free prediction in Eq. (5) from independently measured anti-parallel spin-up proportion. The open data statement is also a positive feature. However, the central mechanism distinguishing sequential tunneling of the same electron from the simpler scenario in which both donor electrons tunnel to the SET is not experimentally established; this is the main gap.","major_comments":[{"comment":"The sentence 'Only electron 1 was tunnel-coupled to the SET and hence could be read out via spin-dependent tunneling' is load-bearing but is not supported by any measurement shown in the manuscript. The two-blip statistics in Fig. 2(c)–(i) and the prediction in Eq. (5) are equally consistent with an alternative in which both donor electrons have tunnel coupling to the SET: in that case |T−> gives no blips, |T+> gives two blips (one from each spin-up electron), the |TX> mixture gives about one blip on average, and Eq. (5) still follows from two independent missed-blip events. A control experiment is needed, e.g., preparing |↓1↑2> in the anti-parallel nuclear configuration and showing that the |↑2> electron produces no blip, or independently calibrating the tunnel coupling of electron 2. Without this, the data do not uniquely support the proposed exchange-mediated sequential tunneling of e","section":"Section IV.A, Fig. 2"},{"comment":"The quantitative validation assumes that missed blips are the dominant readout error and that the two blips in the parallel case are missed independently and with the same per-event probability as the single anti-parallel blip. The manuscript itself lists thermal tunneling and ionization shock as additional error channels (Appendix A). The dashed prediction in Fig. 2(j) is plotted without propagated uncertainty or a goodness-of-fit statistic; 'good correspondence' is not quantified. Please provide uncertainty bands and a statistical comparison, or explicitly present Eq. (5) as a qualitative consistency check rather than a quantitative validation.","section":"Section IV.C, Eq. (5)"},{"comment":"The claim that tunneling into the |T~0> and |S~> states occurs with approximately equal probability is asserted without tunnel-rate measurements or estimates. Since states with different orbital character generally have different tunnel couplings to the SET, this assumption is not obvious and is relevant to the expected two-blip probability. The authors should either provide a measurement or an order-of-magnitude calculation supporting this assumption, or state it as an approximation whose uncertainty does not affect the main conclusion.","section":"Section III, step 2"}],"minor_comments":[{"comment":"The J/|∆| labels for the parallel and anti-parallel cases appear interchanged: the text gives J/|∆|≈0.1 for anti-parallel nuclei and J/|∆|≈133 (≈10²) for parallel nuclei, but the caption and panel (d) state 10⁻¹ for parallel and 10² for anti-parallel.","section":"Figure 1 caption and panel (d)"},{"comment":"Typo: 'the the |T0⟩ or |S~⟩' should read 'either the |T0⟩ or |S~⟩ state'.","section":"Section III, step 2"},{"comment":"Typo: 'donor spinspin qubits' should be 'donor spin qubits'.","section":"Appendix A"},{"comment":"The x-axis label 'Frequency idx' is informal; specify the frequency range and units (or state that it is a linear frequency ramp index).","section":"Fig. 2(j)"},{"comment":"The preparation of the |TX> state is described as a mixture of ≈50% |T~0>, ≈25% |T−>, and ≈25% |T+>, but no derivation or simulation is shown for these fractions. Please clarify how these numbers are obtained.","section":"Section IV.B"},{"comment":"The notation P(⇓⇓/⇑⇑) could be confused with a conditional probability. Define explicitly: the denominator is the prepared nuclear spin configuration and the numerator is the nuclear state after readout.","section":"Eq. (5)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript comes from a group with strong experimental track record, and the direct blip-counting data are compelling evidence for an additional tunneling event. My main concern is that the microscopic identity of that event is underdetermined: the absence of a control for electron-2 tunnel coupling means the data fit a simpler two-electron-tunneling picture. This is fixable by adding a control experiment or by reframing the claim as 'an additional tunneling event' without attributing it to the same electron. If the control cannot be performed, the paper should be revised to clearly separate the established observation (two blips in the parallel regime) from the proposed microscopic mechanism."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know: this paper finally explains the ~40% readout-contrast enhancement seen in exchange-coupled donor qubits when the nuclei are parallel. The proposed mechanism is a second spin-up tunneling event in the same readout window, producing two current blips instead of one. The direct blip-counting data support this: 1.77 ± 0.15 blips per readout for parallel nuclei versus 1.02 ± 0.09 for anti-parallel, with histograms showing a clear two-blip peak. Equation 5, derived from the two-blip picture, matches the measured adiabatic resonance heights. The data are open, which is good practice.\n\nThe paper does well on the central phenomenology. The effect was previously unexplained; this is the first mechanistic attempt, and the experimental statistics are convincing. The practical lever is real: double blips reduce missed-blip errors, so the readout contrast improves. That is a useful, actionable insight for the donor-qubit community.\n\nThe soft spot is the uniqueness of the mechanism. Section IV.A asserts that only electron 1 was tunnel-coupled to the SET, but no measurement establishing this is shown. If electron 2 were also tunnel-coupled, the parallel-regime histograms would look much the same: |T+> would give two blips (one from each electron), |T-> zero, and |TX> about one. That alternative does not require the exchange-mediated re-tunneling of the same electron at all. Equation 5 is unaffected because it only assumes two independent blips, not which electron produced them. So the data support the two-blip picture, but not specifically the same-electron re-tunneling mechanism. This is a moderate gap—probably fixable with a control experiment or existing device characterization—not a fatal flaw.\n\nMinor issues: the Eq. 5 comparison is qualitative, with no propagated uncertainty, and the assumption that missed blips are the dominant error ignores ionization shock, which the authors themselves note as a potential concern. Neither undermines the core observation.\n\nWho this is for: anyone working on donor-based or quantum-dot spin readout, especially readout-fidelity optimization. It deserves a serious referee; a good referee will ask for the tunnel-coupling control for electron 2 and a more quantitative fit. I would send it to peer review, and I would cite the two-blip phenomenology while hedging on the specific mechanism.","headline":"A credible two-blip explanation for the long-standing readout-contrast anomaly in exchange-coupled donor pairs, backed by new blip-counting data, but the claim that the same electron tunnels twice is not fully distinguished from a simpler two-electron-tunneling alternative.","tokens_in":17461,"tokens_out":3668,"would_cite":true,"duration_ms":38006,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The electron readout contrast enhancement in exchange-coupled donor qubits with parallel nuclear spins arises from a second electron tunnelling event to the single-electron transistor within one readout period.","keywords":["exchange-coupled donors","electron spin readout","Elzerman readout","single-electron transistor","readout contrast","double tunnelling","silicon spin qubits","parallel nuclear spins"],"falsifier":"Record SET current traces with a detection bandwidth high enough to resolve single tunnelling events far faster than the current ~50 kHz limit (sub-microsecond rise time). If the double-tunnelling explanation is right, the average blip count for parallel nuclei should approach exactly 2.0 per spin-up readout while anti-parallel stays at 1.0, and the readout-contrast advantage should vanish once no blips are missed. If the contrast gap persists under conditions where only one resolvable blip per readout occurs, the mechanism as proposed is incomplete.","tokens_in":16493,"feed_emoji":"⚛️","tokens_out":6046,"duration_ms":52187,"temperature":0.7,"pith_summary":"This paper claims that the puzzling 40% improvement in electron readout contrast seen in exchange-coupled donor spin qubits when the two donor nuclei are initialized in a parallel spin orientation is not a new physical effect but a straightforward consequence of two-electron quantum mechanics. When the nuclei are parallel, the exchange coupling dominates the energy mismatch between the two electrons, so the two-electron states acquire singlet-triplet character. As a result, after a spin-up electron tunnels off to the single-electron transistor and a spin-down electron returns, the remaining two-electron state still contains spin-up character on the first donor, allowing a second spin-up electron to tunnel off within the same readout window. This produces two current blips instead of one, which halves the probability that a spin-up electron is missed because its blip is too brief for the measurement bandwidth. The paper supports this with direct blip-counting statistics and a quantitative prediction of the parallel-nuclear spin-up fraction.","feed_headline":"Two blips explain donor qubits' 40% readout contrast boost","feed_subtitle":"When donor nuclei align, the electron can tunnel twice per readout, halving the chance of missing the signal.","key_machinery":"The load-bearing object is the two-electron energy-level ladder of an exchange-coupled donor pair in the J >> |ΔA| (parallel-nuclear) regime, specifically the odd-parity states tilde|S> and tilde|T0>. These are linear combinations of |↑↓> and |↓↑> with appreciable single-spin-up probability on the tunnel-coupled donor, which lets a second spin-up electron escape after the first has been replaced by a spin-down electron. The paper uses this level structure to predict a double 'blip' of SET current and to derive the quantitative error-suppression relation for the readout.","core_discovery":"The central claim is that the readout contrast enhancement in the parallel nuclear regime is caused by an additional electron tunnelling event from donor 1 to the SET island within a single readout period. In this regime the exchange interaction J is much larger than the hyperfine difference ΔA between the two donors, so the eigenstates of the two electrons are hybridised versions of the singlet and triplet states. Starting from the |T+> = |↑↑> state, the first spin-up electron tunnels off, a spin-down electron tunnels back in, and because the system now occupies an odd-parity state (tilde|S> or tilde|T0>) with a significant |↑1> component, a second spin-up electron can tunnel off before the","pith_inferences":["If the double-tunnelling mechanism holds, the same physics should also affect the effective measurement back-action on the nuclear spins: each extra electron tunnelling event is an additional opportunity for 'ionization shock', so in devices operated with many repetitive readouts the parallel-nuclear advantage in contrast could come with a slightly higher nuclear-spin flip probability.","A testable engineering extension follows: instead of simply thresholding the SET current once, a readout protocol that counts blips (or triggers on either of the two blips) should recover even more of the parallel-nuclear contrast, and could be applied to anti-parallel readout by transiently driving the system into the high-exchange regime during the readout window.","The independence assumption behind the squared-error formula could be checked by examining the statistics of inter-blip intervals: the model predicts the second blip waiting time follows the same exponential distribution as the first, with the measured ~33 µs tunnel-in time; deviations would indicate correlations or additional tunnelling channels."],"forward_implications":["In the parallel-nuclear configuration, a spin-up electron produces two current blips per readout, so the probability of missing it is the square of the single-blip miss probability, directly improving readout contrast.","The measured average of 1.77 blips per parallel-nuclear readout (vs 1.02 for anti-parallel) quantitatively matches the double-tunnelling picture, with the shortfall from 2.0 attributable to missed blips in the ~50 kHz measurement bandwidth.","Readout fidelity for exchange-coupled donor qubits can be improved either by increasing measurement bandwidth or by deliberately operating in regimes where the extra tunnelling event occurs.","The mechanism explains previously observed, device-to-device variations in readout contrast as arising from differences in exchange coupling and hyperfine detuning, which shift the system between the J>>|ΔA| and J<<|ΔA| regimes."],"fun_headline_variants":["Double tunnelling event explains donor qubit readout contrast boost","Parallel nuclear spins cause extra electron hop in donor qubit readout","Why aligned nuclei double electron tunnelling in spin qubits","Extra tunnelling step when donor nuclei align boosts readout fidelity","Donor qubit readout: additional tunnel event when spins parallel"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The quantitative fit rests on treating missed blips as the only significant readout error and assuming the two parallel-nuclear blips are detected independently with the same per-event miss probability as the single anti-parallel blip; if thermal activation, control errors, or ionization shock contribute substantially, the error-propagation formula would not be a clean test, even though the direct observation of two blips would remain.","fun_headline_variants_meta":{"raw":{"variants":["Double tunnelling event explains donor qubit readout contrast boost","Parallel nuclear spins cause extra electron hop in donor qubit readout","Why aligned nuclei double electron tunnelling in spin qubits","Extra tunnelling step when donor nuclei align boosts readout fidelity","Donor qubit readout: additional tunnel event when spins parallel"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000236,"raw_usage":{"total_tokens":1320,"prompt_tokens":703,"completion_tokens":617,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":447,"completion_tokens_details":{"reasoning_tokens":530}},"tokens_in":447,"tokens_out":617,"duration_ms":6198,"temperature":1.0,"reasoning_tokens":530,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T23:10:51.109484+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Record SET current traces with a detection bandwidth high enough to resolve single tunnelling events far faster than the current ~50 kHz limit (sub-microsecond rise time). If the double-tunnelling explanation is right, the average blip count for parallel nuclei should approach exactly 2.0 per spin-up readout while anti-parallel stays at 1.0, and the readout-contrast advantage should vanish once no blips are missed. If the contrast gap persists under conditions where only one resolvable blip per readout occurs, the mechanism as proposed is incomplete.","supporting_citations":[],"review_version":1}