{"id":"1ce01c26-bb2a-44f7-8f1b-04a3b45123ff","arxiv_id":"2501.15912","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A flux-tunable superconducting loop provides phase control over the coupling between two quantum dots, enabling continuous tuning to poor man's Majorana sweet spots and a direct probe of their wavefunction.","lead":"Researchers used a superconducting loop in a semiconductor-superconductor device to control the coupling between two distant quantum dots by adjusting magnetic flux. This adds a new tuning knob for engineering poor man's Majorana bound states, a step toward building topological qubits.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Middle-probe zero-bias peak may reflect a zero-energy ABS rather than the PMM wavefunction, leaving the spatial-distribution claim uncontrolled.","rationale":"The reader's weakest assumption identifies the same load-bearing concern: the middle-probe zero-bias peak is interpreted as PMM wavefunction overlap without a control baseline. The paper's other central claims—phase-controlled ECT/CAR tuning and a continuous sweet-spot line—are supported by the 2π-periodic modulation of avoided crossings, repeated cool-downs in the SI, and spectroscopic verification of PMM behavior at the sweet spot. The continuous-line claim is based on the standard CSD straight-crossing criterion for Γe = Γo, which is a reasonable proxy once the mapping to PMM behavior is calibrated. The terminology issue (MBS vs PMM) and missing error bars are real but do not change the verdict. The middle-probe concern is the most consequential because it affects a distinct advertised result, and the alternative—that the zero-bias conductance arises from a zero-energy ABS level visible in Fig.2c—is physically plausible and not excluded by the reported data. Therefore the paper should remain CONDITIONAL: the phase-control result can stand, but the spatial wavefunction claim needs an explicit control experiment or a softened conclusion.","tokens_in":11376,"tokens_out":9958,"duration_ms":109436,"concrete_test":"Re-analyze the existing detuning spectroscopy data (main-text Fig.3c and/or SI Fig.S8b): extract GMM at zero bias as a function of the two QD detunings ε_L and ε_R around the sweet spot. The spatial claim is supported only if the GMM zero-bias peak persists when one QD is detuned and splits with the same gap as GLL and GRR when both QDs are detuned. Additionally, measure GMM at the same Δφ and VABS with both QDs pinched off (no dots formed); if a zero-bias peak remains, it is a bare ABS/YSR signal, not the PMM wavefunction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's third headline claim—that the middle probe reveals the spatial distribution of the Majorana wavefunction—rests solely on the zero-bias peak in GMM shown in Fig.3c. The alternative is concrete: Fig.2c shows that the ABS spectrum itself crosses zero energy as a function of both Δφ and VABS, and the PMM sweet spot likely sits near such a crossing. A zero-bias peak in GMM could then appear even if the delocalized zero-energy mode were a bare ABS or YSR state not belonging to the PMM subspace, or if the middle lead couples to an ABS component that is not part of the QD–QD PMM state. The main text does not show that the GMM peak tracks the PMM: it does not report GMM away from the sweet spot or GMM under one-dot/both-dot detuning. The SI caption for Fig.S8 only says all three probes were used to verify PMM conditions, without stating that the middle-probe peak splits or vanishes together with the L/R PMM peaks. Without a control establishing that the GMM zero-bias peak is destroyed or split with the same energy scale as the PMM when the PMM is detuned, the spatial wavefunction interpretation is not established. This concern does not threaten the phase-control and sweet-spot-line claims, but it does undermine one of the three headline results.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports measurements of a two-site Kitaev chain formed by two quantum dots coupled through an extended Andreev bound state (ABS) in an InSbAs 2DEG, with the ABS embedded in a superconducting loop. The authors show that the superconducting phase difference Δφ, controlled by an out-of-plane magnetic field, changes the relative strength of the spin-conserving and spin-non-conserving couplings Γo and Γe inferred from charge stability diagrams, and that a PMM sweet spot (Γe = Γo) can be found for a range of ABS gate voltages VABS, tracing a path in the (VABS, Δφ) plane. They also report a zero-bias conductance peak in a middle tunnel probe attached to the ABS region, which they interpret as evidence that the PMM wavefunctions extend into the ABS. The data come from several cooldowns, and raw data and plotting scripts are deposited on Zenodo.","tokens_in":11578,"tokens_out":6957,"duration_ms":68694,"significance":"If the phase-control result holds, it adds a genuinely useful control knob: the sweet-spot manifold is extended from a discrete set of ABS gate values to a continuous path in a two-dimensional parameter space, and the demonstration of 10–30 μeV couplings over ~1 μm distances is valuable for longer Kitaev chains. The 2π periodicity of the effect and the matching of the 28 μT field period to the loop area provide independent experimental checks, and the absence of fitted parameters is a clear strength. The middle-probe result, if properly controlled, would be a rare direct probe of PMM wavefunction delocalization. However, the skeptical concern about the middle-probe identification is justified and is the main reason I cannot accept the paper in its present form. The circularity concern raised in the reader's report does not land: the PMM criteria are taken from prior work, and the phase-periodicity and loop-area checks are independent of the PMM interpretation.","major_comments":[{"comment":"The inference that the middle-probe zero-bias peak in GMM demonstrates that γ1 and γ2 reside partially in the ABS region is not controlled. Figure 2c shows that the ABS spectrum itself crosses zero energy as a function of both Δφ and VABS, so a zero-energy ABS or Yu-Shiba-Rusinov state independent of the two-QD PMM subspace could also produce a peak in GMM. The manuscript does not report GMM spectra away from the PMM sweet spot, nor does it show that the GMM peak splits or vanishes with the same energy scale as the left/right PMM peaks when one or both dots are detuned. Please add such a control measurement, or explicitly weaken the abstract and Section III claim from 'spatial distribution of the Majorana wave function' to a zero-bias signal whose assignment to the PMM subspace remains to be established. The phase-control and sweet-spot results do not depend on this identification.","section":"Section III, final paragraph and Fig. 3c"},{"comment":"The statement that sweet spots 'span a continuous line' in the (VABS, Δφ) plane is stronger than the evidence shown. Fig. 4b displays a discrete set of extracted Δφ* values with no error bars and no quantitative criterion for the 'straight transition line' diagnostic, and the text does not state how densely VABS was swept. Since the continuous-range claim is one of the paper's headline results, please either show a dense sweep with error analysis or rephrase to 'a sweet spot was obtained for every VABS setting explored, with a monotonic shift of Δφ*'.","section":"Section IV, Fig. 4"}],"minor_comments":[{"comment":"Please specify the detuning configuration for each trace (sweet spot, one dot detuned, both dots detuned) and mark the middle-probe zero-bias peak more explicitly; the main text currently leaves this to the reader.","section":"Fig. 3c caption"},{"comment":"The attribution of the absence of the predicted excitation-gap variation to 'multiple states' and 'spin-splitting' is plausible but not demonstrated; please present it as an open question or support it with data.","section":"Section IV, final paragraph"},{"comment":"In the preprint rendering, some figure labels (e.g., 'φ/two.denominator' in Fig. 1) appear garbled; please check the final compiled figure files.","section":"Fig. 1"}],"recommendation":"major_revision","confidential_remarks":"The phase-control and sweet-spot-line results are likely sound and will be a useful contribution. The required revision is the middle-probe control; if the missing control cannot be provided, the spatial-distribution claim should be explicitly downgraded. The sparse sampling in Fig. 4b is not fatal if the wording is adjusted. The data availability statement and the use of multiple cooldowns are positive. No concerns about citation practice or novelty disclosure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a serious look. The core result—using the superconducting phase difference to tune the ECT/CAR balance and move PMM sweet spots along a continuous line in (VABS, Δφ)—is well supported and genuinely new. The 2π-periodic response, the loop-area check, and reproducibility across cool-downs make the phase-control claim credible. The long-range coupling between QDs separated by ~1 µm with interaction strengths around 18–30 µeV is also a useful design advance for PMM devices. The paper is clearly written, the device characterization is careful, and raw data are on Zenodo.\n\nThe soft spot is the middle-probe zero-bias peak in GMM. The paper interprets that peak as evidence that the PMM wavefunctions γ1 and γ2 extend into the ABS region, but it does not show a control measurement away from the sweet spot or with one or both QDs detuned. Given that Fig. 2c shows the ABS spectrum itself crosses zero energy as a function of both Δφ and VABS, a bare ABS or YSR state could produce the same middle-probe peak. Without a control showing the GMM peak splits or vanishes together with the L/R PMM peaks, the spatial-distribution claim is not established. This does not threaten the phase-control or sweet-spot-line results, but it is one of the three headline claims and needs the missing baseline.\n\nMinor issues: the abstract says MBSs while the text correctly clarifies these are non-topological PMMs, and the quoted coupling amplitudes have no error bars. The paper honestly reports that no systematic variation of the excitation gap was observed along the sweet-spot line, contrary to theory prediction, and attributes this to multiple states and spin-splitting; that is a reasonable caveat, not a flaw.\n\nWho this is for: experimental groups working on poor man's Majoranas, Kitaev chains in quantum dots, or superconducting coupling between spin qubits. They will find the flux knob and the large-separation coupling genuinely useful. The middle-probe method could be valuable once controlled properly.\n\nRecommendation: send it to peer review. It deserves referee time, with the request that the middle-probe control data be provided and the MBS/PMM terminology fixed.","headline":"Flux phase control of ABS-mediated QD coupling is a real new knob and the continuous sweet-spot line is solid; the middle-probe spatial wavefunction claim is under-supported.","tokens_in":12206,"tokens_out":1664,"would_cite":true,"duration_ms":16883,"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 paper shows that magnetic flux through a superconducting loop gives continuous control over the coupling between two quantum dots, turning an extended Andreev bound state into a tunable mediator for poor man's Majorana states.","keywords":["poor man's Majorana","Andreev bound state","crossed Andreev reflection","elastic cotunneling","superconducting phase control","quantum dots","two-site Kitaev chain","2DEG hybrid device"],"falsifier":"A control measurement would settle it: record the middle-probe conductance with both quantum dots detuned away from the sweet spot (for instance, in Coulomb blockade) and look for the zero-bias peak; if the peak persists or does not split in the way the PMM model predicts when one dot is detuned, the spatial-distribution claim would be falsified.","tokens_in":11170,"feed_emoji":"🧲","tokens_out":7949,"duration_ms":64605,"temperature":0.7,"pith_summary":"This paper reports a new way to control the couplings that form a minimal Kitaev chain from two quantum dots. The key idea is that an extended Andreev bound state in a flux-tunable Josephson junction mediates both elastic cotunneling and crossed Andreev reflection between the dots, and the superconducting phase difference sets the balance between the two. The experiment shows that this phase control makes the poor-man's-Majorana sweet spot reachable for a continuous range of ABS chemical potentials, rather than only at isolated gate values. A middle spectroscopic probe reveals a zero-bias peak, which the authors take as evidence that the Majorana wavefunction extends into the hybrid region. These results add a practical tuning knob and relax design constraints for Majorana-based devices.","feed_headline":"Superconducting phase becomes a tuning knob for Majorana states","feed_subtitle":"A magnetic-flux knob adds a continuous gate–phase tuning axis for poor man's Majoranas, easing device design constraints.","key_machinery":"The central object is the extended Andreev bound state formed in a proximitized semiconductor segment that connects two superconducting electrodes in a loop. Its energy $E_{\\mathrm{ABS}}$ and its coherence factors $u$ and $v$ determine the amplitudes of elastic cotunneling (ECT) and crossed Andreev reflection (CAR) between the two adjacent quantum dots. The superconducting phase difference $\\Delta\\phi$, controlled by a magnetic field through the loop, changes $u$ and $v$ and therefore the ECT/CAR ratio; the gate voltage $V_{\\mathrm{ABS}}$ tunes the ABS chemical potential. The poor-man's-Majorana sweet spot is defined as the point where the effective spin-conserving coupling $\\Gamma_o$ equals the spin-non-conserving coupling $\\Gamma_e$, and the paper maps how this point moves as $V_{\\mathrm{ABS}}$ and $\\Delta\\phi$ are varied.","core_discovery":"The central claim is that the superconducting phase difference, set by an out-of-plane magnetic field threading a loop, changes the effective coupling between two spin-polarized quantum dots separated by about one micrometre. The phase difference modifies the coherence factors of the mediating Andreev bound state, thereby changing the relative strengths of elastic cotunneling and crossed Andreev reflection. As a consequence, the condition for a poor-man's-Majorana sweet spot, where the spin-conserving and spin-non-conserving couplings are equal, can be satisfied for every value of the ABS chemical potential within a range, with the sweet spot tracing a continuous line in the gate–phase plane. The paper also claims that a zero-bias conductance peak measured from a probe attached to the middle of the hybrid segment shows that both Majorana wavefunctions reside partially in the ABS region, a spatial distribution that had been predicted but not directly probed previously.","pith_inferences":["A natural next step would be to use a fast on-chip flux line instead of a global magnetic field to switch between coupling regimes; because phase changes are local and fast, this could enable time-resolved tuning of the ECT/CAR ratio without disturbing the dot electrostatics.","The observed continuous sweet-spot line in the $(V_{\\mathrm{ABS}},\\Delta\\phi)$ plane could be used as a sensitive probe of the ABS spectrum; the systematic deviations from theory that the authors note may reflect spin-orbit and Zeeman effects that a simple two-level ABS model omits.","If the middle-probe peak truly tracks the PMM wavefunction, then the same probe could test the predicted exchange braiding of poor-man's Majoranas by watching how the zero-bias peak evolves during a manipulation sequence."],"forward_implications":["PMM sweet spots become reachable for a continuous range of ABS chemical potentials by choosing the appropriate phase difference, not only at discrete gate voltages.","Quantum dots separated by about one micrometre can still be coupled strongly enough (10–30 µeV) to form a two-site Kitaev chain, relaxing geometric constraints on device layouts.","A zero-bias peak at a probe in the middle of the hybrid segment indicates that the Majorana wavefunctions extend into the ABS region, providing a direct way to probe their spatial distribution.","Phase control of the ABS can serve as an additional tuning mechanism for the interaction between quantum dots, complementing electrostatic gate control."],"supporting_citations":[{"why":"Supplies the theory that ABS energy and coherence factors determine ECT and CAR amplitudes between quantum dots.","marker":"[14]"},{"why":"Demonstrated tunable crossed Andreev reflection and elastic cotunneling in hybrid nanowires, establishing the baseline this work extends.","marker":"[15]"},{"why":"Reported the previous realization of a minimal Kitaev chain in coupled quantum dots, the experimental foundation for the present device.","marker":"[18]"},{"why":"Presented a two-site Kitaev chain in a two-dimensional electron gas, supplying the $\\Gamma_o/\\Gamma_e$ framework and sweet-spot condition used here.","marker":"[19]"},{"why":"Predicted that flux (phase) can tune the Kitaev chain and controls the ECT/CAR ratio, motivating this experiment.","marker":"[20]"},{"why":"Predicted that PMM wavefunctions overlap inside the ABS segment in the strong-coupling regime, which the middle-probe measurement tests.","marker":"[21]"}],"fun_headline_variants":["Flux knob tunes quantum dots into Majorana sweet spots","Magnetic flux guides Majorana states in two-site chain","Phase control unlocks Majorana parameter space"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the zero-bias peak seen with the middle probe comes from the poor-man's-Majorana wavefunctions, not from a zero-energy Andreev or Yu-Shiba-Rusinov state that exists independently of the two-dot sweet-spot configuration.","fun_headline_variants_meta":{"raw":{"variants":["Flux knob tunes quantum dots into Majorana sweet spots","Magnetic flux guides Majorana states in two-site chain","Phase control unlocks Majorana parameter space"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000151,"raw_usage":{"total_tokens":1166,"prompt_tokens":878,"completion_tokens":288,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":494,"completion_tokens_details":{"reasoning_tokens":239}},"tokens_in":494,"tokens_out":288,"duration_ms":3271,"temperature":1.0,"reasoning_tokens":239,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T13:50:01.227911+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A control measurement would settle it: record the middle-probe conductance with both quantum dots detuned away from the sweet spot (for instance, in Coulomb blockade) and look for the zero-bias peak; if the peak persists or does not split in the way the PMM model predicts when one dot is detuned, the spatial-distribution claim would be falsified.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the theory that ABS energy and coherence factors determine ECT and CAR amplitudes between quantum dots."},{"cited_title":"Bordin, G","cited_arxiv_id":null,"evidence_quote":"Demonstrated tunable crossed Andreev reflection and elastic cotunneling in hybrid nanowires, establishing the baseline this work extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reported the previous realization of a minimal Kitaev chain in coupled quantum dots, the experimental foundation for the present device."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Presented a two-site Kitaev chain in a two-dimensional electron gas, supplying the $\\Gamma_o/\\Gamma_e$ framework and sweet-spot condition used here."},{"cited_title":"Flux-tunable Kitaev chain in a quantum dot array","cited_arxiv_id":"2402.07575","evidence_quote":"Predicted that flux (phase) can tune the Kitaev chain and controls the ECT/CAR ratio, motivating this experiment."},{"cited_title":"Enhancing the excitation gap of a quantum-dot-based Kitaev chain","cited_arxiv_id":"2310.09106","evidence_quote":"Predicted that PMM wavefunctions overlap inside the ABS segment in the strong-coupling regime, which the middle-probe measurement tests."}],"review_version":1}