{"id":"eff0ee1f-a87f-49c7-9368-92a27090323b","arxiv_id":"2501.18168","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In a MoSe2/hBN/WSe2 electron-hole bilayer, an excitonic insulator shows magnetic-field-driven quantum oscillations and multiple transitions into quantum Hall insulators.","lead":"This experiment puts a bilayer of two different atomically thin semiconductors in a magnetic field and watches its electrons and holes pair up, oscillate, and switch between insulating states. It reports the first clear sign that excitonic insulators can show quantum oscillations, a behavior normally reserved for metals.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Zero drag and quantized R_h at integer fillings may stem from edge-channel shunting rather than bulk exciton dissociation; a Corbino geometry test is needed to confirm the claimed EI-QHI phase transitions.","rationale":"The reader identified essentially the same load-bearing concern: the closed-circuit drag measurement along the CNL is treated as a faithful bulk probe, while edge-channel shunting is only included qualitatively. My analysis sharpens this by noting that the drive layer's own edge channels can carry the excitation current at integer fillings, making the vanishing drag ratio independent of the bulk exciton state. Since the central claim includes multiple EI-QHI phase transitions at integer fillings, this ambiguity directly affects the phase diagram and the novelty of the result. However, the observation of quantum oscillations in both drag and resistance at low to medium fields, and the qualitative agreement with theory, are not invalidated by this concern. The paper's internal consistency, reproducible device fabrication described in Methods, and the use of a second device (Extended Data Fig. 1) provide supporting evidence. The correct stance remains conditional: the manuscript should be accepted with the requirement that the edge-shunting alternative be addressed, either by the proposed Corbino experiment or by a quantitative model showing that edge channels cannot account for the observed zero-drag regions. My proposed test is decisive and feasible with existing fabrication techniques, so the verdict remains CONDITIONAL, unchanged from the reader's assessment.","tokens_in":12234,"tokens_out":5343,"duration_ms":63915,"concrete_test":"Fabricate a Corbino-geometry device (annular MoSe2/hBN/WSe2 stack with independent inner and outer Ohmic contacts to both layers) and repeat the CNL linecut of Fig. 2e at B = 12 T. In a Corbino geometry, edge channels are geometrically eliminated, so the two-terminal resistance and closed-circuit drag current measure bulk transport only. If the drag ratio remains near unity and R_h stays large at integer fillings, the zero-drag dips in Fig. 2e are edge-shunting artifacts and the claimed EI-QHI transitions are not bulk transitions. If drag still vanishes and R_h still quantizes at h/νe², the bulk phase-transition interpretation is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Fig. 2e, the identification of EI-QHI phase transitions rests on the interpretation that a simultaneous zero drag ratio and quantized R_h = h/νe² at integer fillings proves complete exciton dissociation. However, the authors' own edge-channel model (paragraph beginning 'A qualitative explanation...') points out that a portion of the induced hole current can return through the hole layer's edge channels instead of the measurement circuit. At an integer filling, the electron layer also develops a dissipationless edge channel, so a drive current can flow predominantly through the electron edge, bypassing the bulk excitons and yielding a vanishing drag signal even if the bulk remains an excitonic insulator. Simultaneously, the hole layer's edge channels, which may be populated by unpaired holes at the sample boundary or in the heavily doped reservoir regions, can produce a quantized two-terminal R_h. Thus the coexistence of zero drag and quantized R_h is not unique evidence for bulk EI destruction; it is also consistent with an EI bulk coexisting with QH-like edge transport in both layers. The temperature-induced recovery of R_h quantization (Fig. 4e-f) could likewise be dominated by edge channels rather than a bona fide bulk phase transition. The phase boundaries in Figs. 2e and 4 are therefore underdetermined unless the bulk contribution is isolated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports transport experiments on MoSe2/hBN/WSe2 electron-hole bilayers under a perpendicular magnetic field. At zero field the device exhibits an excitonic insulator with perfect Coulomb drag. With field, the authors observe oscillations in the closed-circuit drag ratio and the two-terminal hole resistance along the charge-neutrality line, and at high field they identify multiple transitions between the excitonic insulator and bilayer quantum Hall insulator phases at integer Landau-level fillings. A fan diagram and finite-temperature measurements are presented as evidence for these transitions. The data are supported by a second device for resistance measurements only.","tokens_in":12469,"tokens_out":7390,"duration_ms":71655,"significance":"If the interpretation is correct, this is the first experimental observation of quantum oscillations in an excitonic insulator and of magnetic-field-driven dissociation of an excitonic insulator into two independent quantum Hall insulators, directly testing the predictions of Refs. [5,6]. The authors take care to rule out graphite-gate-induced oscillations via a capacitor model and to exclude interlayer tunnelling by measuring a lower-bound resistance of 40 GΩ. The central observation of oscillations in both drag and resistance is compelling. However, the uniqueness of the phase-transition assignment is compromised by edge-channel transport, which the authors themselves discuss qualitatively. For this reason, the significance of the paper is high if the edge-channel alternative can be eliminated, but the current evidence is insufficient to establish the phase transitions unambiguously.","major_comments":[{"comment":"The identification of EI-QHI phase transitions at integer filling factors rests on the simultaneous vanishing of the drag ratio and the quantization of R_h to h/νe^2. The authors' own edge-channel model, introduced in the paragraph beginning 'A qualitative explanation can be built considering the quantized edge conducting channels...', shows that at integer filling the electron-layer edge channel can carry the drive current without coupling to bulk excitons, and the hole-layer edge channels can return a significant portion of the induced current to the other terminal without passing through the measurement circuit. Consequently, a zero drag ratio and quantized R_h are also consistent with a bulk that remains an excitonic insulator while edge channels dominate the transport. Because the central new claim is the existence of multiple EI-QHI phase transitions, a Corbino-geometry measurement, a four-terminal bulk-conductivity measurement, or an explicit quantitative treatment of the edge channels is needed to exclude this alternative. As presented, the phase boundaries in Figs. 2e, 3, and 4 are underdetermined.","section":"Electron-hole fluids in a strong magnetic field (Fig. 2e)"},{"comment":"The temperature-induced EI-QHI transitions are inferred from the rapid decrease of the drag ratio and the recovery of R_h quantization at temperatures of roughly 5 K (ν=3) and 2 K (ν=4). The same edge-channel alternative applies: the observed temperature dependence may reflect thermally activated hopping in edge channels or changes in contact resistance rather than a bulk phase transition. Without a bulk-sensitive measurement, the assignment of these features as thermally driven EI-QHI phase transitions is not unique. I therefore consider this a second instance of the same load-bearing issue.","section":"Finite-temperature phase diagram (Fig. 4e-f)"}],"minor_comments":[{"comment":"The definition of the drag ratio and the method for measuring I_drive should be stated explicitly; currently the reader must infer the drive current from the circuit diagram, and it is unclear whether I_drive is measured directly or computed from the excitation voltage and series resistance.","section":"Methods, Coulomb drag measurements"},{"comment":"The contact resistance is quoted as 2.4 kΩ in the text but as 2.2 kΩ in the Extended Data Fig. 3 caption; please reconcile these values.","section":"Electron-hole fluids in a strong magnetic field"},{"comment":"The claim that the oscillations are periodic in 1/B is presented visually; a quantitative analysis (e.g., FFT or zero-crossing positions) would substantiate this periodicity and allow a comparison with the theoretical period.","section":"Fig. 3c"},{"comment":"The drag ratio color scale appears saturated at unity; error bars or a noise floor estimate would help readers assess the precision of the perfect-drag claim.","section":"Fig. 1h"},{"comment":"The statement that similar results are reproduced in device D2 is based only on R_h measurements; the text should clarify that drag measurements were not possible in D2.","section":"Extended Data Fig. 1"},{"comment":"The term 'excitonic quantum oscillations' in the abstract might be more precise as 'quantum oscillations in an excitonic insulator'.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"This is a strong experimental paper with a clear central observation. The main barrier to acceptance is the edge-channel alternative for the claimed EI-QHI phase transitions; the authors themselves introduce the edge-channel mechanism to explain the gradual decrease of the drag ratio at half-integer fillings, which makes the concern particularly relevant. Given the overlapping authorship with one of the theory papers (ref. 6), the qualitative agreement should be discussed conservatively. I recommend major revision, and I would want to see either a direct bulk measurement (e.g., Corbino) or a quantitative edge-channel model that can account for the observed drag ratios and R_h at integer fillings."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this one: it reports the first experimental observation of quantum oscillations inside an excitonic insulator and of magnetic-field-driven transitions between an excitonic insulator and bilayer quantum Hall insulators, using a MoSe2/hBN/WSe2 stack. The zero-field EI baseline in these devices is already well established by the same group, so the starting point is solid.\n\nThe genuinely new content is the finite-field transport: oscillatory drag and R_h along the charge-neutral line, Landau-fan-like structure in both quantities, and a sequence of dips at integer fillings where the drag ratio goes to zero and R_h quantizes. There is also a temperature-driven recovery of R_h quantization at small integer fillings. Two devices are used, one with full drag capability, and the authors do careful things: they subtract the Pt-WSe2 contact resistance, check that graphite-gate densities are too high to explain the oscillations, and show that interlayer tunneling is negligible. That is a credible experimental package.\n\nThe main soft spot is the interpretation of zero drag plus quantized R_h as proof of complete bulk exciton dissociation. The authors themselves note that part of the induced hole current can return through hole-layer edge channels, and at integer filling the electron layer also has dissipationless edge states. So a vanishing drag ratio is not unique evidence for a bulk EI-to-QHI transition; it is also consistent with edge-dominated transport coexisting with an EI bulk. The phase boundaries in Fig. 2e and the temperature-driven transitions in Fig. 4e-f inherit that ambiguity. The authors do not isolate the bulk response, and a Corbino geometry would settle it. That said, the quantum oscillations at partial fillings, with drag peaking at half-integer fillings, are much harder to explain by edge effects alone and give real support to the central claim.\n\nOther limitations: no error bars or statistics, no deposited data, and only one device for the drag measurements. The comparison to theory is qualitative, but that is appropriate here.\n\nOverall, the central observation holds up; the phase-boundary identification is underdetermined but not obviously wrong. I would send this to a serious referee. The paper deserves a careful review that pushes on the edge-channel versus bulk distinction and asks for raw data. With that revision, it would be a strong contribution.","headline":"First experimental report of magnetic-field-driven excitonic quantum oscillations and EI-QHI transitions in a TMD bilayer; the data are credible, the integer-filling phase boundary interpretation has a real edge-channel caveat.","tokens_in":13017,"tokens_out":1710,"would_cite":true,"duration_ms":19293,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper reports the observation of excitonic quantum oscillations in Coulomb drag and resistance of MoSe2/hBN/WSe2 electron-hole bilayers, and quantum phase transitions between the excitonic insulator and bilayer quantum Hall insulator…","keywords":["excitonic insulator","Coulomb drag","quantum oscillations","quantum Hall insulator","electron-hole bilayer","MoSe2/hBN/WSe2 heterostructure","Landau quantization","charge-neutral line"],"falsifier":"Measure the same charge-neutral line in a device with a Corbino or four-terminal geometry that removes the hole layer's quantum Hall edge channels, and check whether the drag ratio still drops to zero at filling factor five and whether the hole resistance remains quantized. If the dip disappears when edge shunting is suppressed, the claimed complete exciton dissociation at integer fillings is not established; if it survives, the EI-QHI transition is a bulk effect. A second check is to vary the hBN spacer thickness: the gap-oscillation mechanism predicts that the magnetic field at which the first drag dip appears should move with the interlayer exciton binding energy.","tokens_in":12053,"feed_emoji":"🧲","tokens_out":11459,"duration_ms":108972,"temperature":0.7,"pith_summary":"This paper claims that a magnetic field can tune the ground state of an excitonic insulator, a phase in which electrons and holes in separate layers bind into charge-neutral excitons. In MoSe2/hBN/WSe2 bilayers, the authors observe quantum oscillations in both the Coulomb drag current and the hole-layer resistance at fields above roughly five tesla, with periodicity in $1/B$. At stronger fields and higher exciton densities, the oscillations deepen until the excitonic insulator gives way to two independent quantum Hall insulators at integer Landau-level fillings. If correct, the result shows that insulating exciton fluids can show metallic-style quantum oscillations and that Landau quantization can reversibly dissociate interlayer excitons, making the bilayer a switchable platform for composite bosonic phases.","feed_headline":"12 tesla switches an excitonic insulator into two quantum Hall states","feed_subtitle":"At lower fields, drag and resistance oscillate with $1/B$ as Landau levels modulate the gap.","key_machinery":"The carrying object is the effective charge gap of the excitonic insulator, defined as the energy difference between the highest occupied and lowest unoccupied dressed Landau levels. An out-of-plane magnetic field quantizes the two-dimensional electron and hole gases; when the exciton filling factor approaches an integer, Landau quantization lowers the gap, which shows up as a $1/B$-periodic oscillation in both the drag ratio and the resistance. In the strong-field, high-density regime, the same oscillation can drive the gap to zero, turning the bound-exciton insulator into two decoupled quantum Hall insulators whose signatures are quantized hole resistance and a vanishing drag signal.","core_discovery":"The paper reports the first experimental observation that an excitonic insulator in a MoSe2/hBN/WSe2 electron-hole bilayer responds to a perpendicular magnetic field by developing quantum oscillations. Along the net charge-neutral line, both the Coulomb drag ratio and the hole-layer resistance oscillate with period $1/B$ at fields above about five tesla; at higher fields and higher pair densities the oscillations deepen until, at integer Landau-level fillings, the drag ratio falls to zero and the hole resistance becomes quantized at $h/\\nu e^2$, marks of two independent integer quantum Hall insulators. The authors interpret these features as oscillations of the effective charge gap, which Landau quantization lowers whenever the exciton filling factor approaches an integer, and at high pair density can close completely and dissociate the excitons. They take the coexistence of large drag and a small deviation from perfect quantization at low integer fillings as evidence that Coulomb interactions mix Landau levels and stabilize excitonic pairing even inside the quantum Hall regime.","pith_inferences":["Editorial extension: if the gap-oscillation mechanism is generic, other strongly coupled dipolar exciton systems should show the same $1/B$ drag oscillations once many-body screening reduces their binding energy; the authors observe this only in the TMD platform, not in other materials.","Implicit in their edge-channel picture: a zero drag ratio at integer fillings could also be produced if the induced hole current returns through the hole layer's quantum Hall edge channels instead of the measurement circuit, so four-terminal or Corbino drag measurements would decide whether the EI-QHI boundary is sharp or is partly an edge-shunting artifact.","Going beyond the paper, the depth of each drag dip as a function of field supplies a quantitative measure of exciton binding as a function of Landau index, which could be compared directly with exact-diagonalization calculations of Landau-level-mixed exciton energies.","The paper does not discuss it, but the same devices with a small electron-hole imbalance should test the predicted excitonic insulator states at filling-factor differences of one; the authors note that these states are absent, which is an immediate experimental target for theory."],"forward_implications":["Quantum oscillations can appear in a correlated insulator: inside the excitonic insulator, both the Coulomb drag ratio and the hole resistance are periodic in $1/B$ even though the charge response is gapped.","At integer Landau fillings the magnetic field can fully destroy exciton binding, producing independent bilayer quantum Hall insulators with quantized resistance and zero drag.","Temperature acts like a knob on the same competition: heating melts the excitonic domes at low integer fillings and restores quantized quantum Hall plateaus.","Because the oscillation frequency tracks the pair density along the charge-neutral line, the fan diagram maps where cyclotron energy and exciton binding energy become comparable.","TMD-based electron-hole bilayers provide a tunable platform in which an insulating exciton fluid can be switched, by field, density, or temperature, between a bound-exciton phase and two independent Landau-quantized phases."],"supporting_citations":[{"why":"Establishes the charge-neutral phase diagram and electrostatic doping control of the MoSe2/hBN/WSe2 bilayer used here, including the optical method for extracting electron and hole densities.","marker":"[1]"},{"why":"Provides the reference strongly correlated excitonic insulator in atomic double layers with a large charge gap, the zero-field state this paper subjects to a magnetic field.","marker":"[2]"},{"why":"Reports perfect Coulomb drag in a dipolar excitonic insulator, supplying the drag-ratio-unity signature that defines the excitonic insulator phase in this work.","marker":"[3]"},{"why":"Reports perfect Coulomb drag and exciton transport in an excitonic insulator in the same TMD platform, establishing the closed-circuit drag measurement used along the charge-neutral line.","marker":"[4]"},{"why":"Predicts $1/B$-periodic quantum oscillations in an excitonic insulating electron-hole bilayer, the effect this paper claims to observe.","marker":"[5]"},{"why":"Predicts electrically controlled transitions between electron-hole fluid states in the quantum Hall regime, including exciton dissociation at high field.","marker":"[6]"},{"why":"Explains why two-terminal resistance of a quantum Hall device is quantized at $h/\\nu e^2$ when contact resistance is small, the criterion used to identify the bilayer quantum Hall insulator phase.","marker":"[40]"},{"why":"Identifies graphite-gate-induced potential oscillations as a spurious cause of quantum oscillations in two-dimensional insulators, the alternative the authors rule out with their capacitor model.","marker":"[41]"},{"why":"Provides the early theory that a magnetic field can increase exciton binding by reducing screening, which the authors invoke to explain the excitonic insulator region surviving to higher pair density at higher field.","marker":"[42]"}],"fun_headline_variants":["Field turns excitonic insulator into two quantum Hall states","Excitonic insulator drag oscillates, then quantizes into Hall states","Magnetic field tunes excitonic insulator to Hall insulator","Drag oscillations signal excitonic insulator's Hall transition"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on treating the measured drag current and the two-terminal resistance along the charge-neutral line as faithful measures of the bulk exciton fluid: if the hole layer's quantized edge channels carry away a large fraction of the induced current when the Landau level is full, a zero drag ratio would not by itself prove that all excitons have dissociated.","fun_headline_variants_meta":{"raw":{"variants":["Field turns excitonic insulator into two quantum Hall states","Excitonic insulator drag oscillates, then quantizes into Hall states","Magnetic field tunes excitonic insulator to Hall insulator","Drag oscillations signal excitonic insulator's Hall transition"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001362,"raw_usage":{"total_tokens":5504,"prompt_tokens":902,"completion_tokens":4602,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":518,"completion_tokens_details":{"reasoning_tokens":4536}},"tokens_in":518,"tokens_out":4602,"duration_ms":28201,"temperature":1.0,"reasoning_tokens":4536,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T00:25:53.992906+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the same charge-neutral line in a device with a Corbino or four-terminal geometry that removes the hole layer's quantum Hall edge channels, and check whether the drag ratio still drops to zero at filling factor five and whether the hole resistance remains quantized. If the dip disappears when edge shunting is suppressed, the claimed complete exciton dissociation at integer fillings is not established; if it survives, the EI-QHI transition is a bulk effect. A second check is to vary the hBN spacer thickness: the gap-oscillation mechanism predicts that the magnetic field at which the first drag dip appears should move with the interlayer exciton binding energy.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the charge-neutral phase diagram and electrostatic doping control of the MoSe2/hBN/WSe2 bilayer used here, including the optical method for extracting electron and hole densities."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the reference strongly correlated excitonic insulator in atomic double layers with a large charge gap, the zero-field state this paper subjects to a magnetic field."},{"cited_title":"Perfect Coulomb drag and exciton transport in an excitonic insulator","cited_arxiv_id":"2309.15357","evidence_quote":"Reports perfect Coulomb drag and exciton transport in an excitonic insulator in the same TMD platform, establishing the closed-circuit drag measurement used along the charge-neutral line."},{"cited_title":"& Dai, X","cited_arxiv_id":null,"evidence_quote":"Predicts $1/B$-periodic quantum oscillations in an excitonic insulating electron-hole bilayer, the effect this paper claims to observe."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Predicts electrically controlled transitions between electron-hole fluid states in the quantum Hall regime, including exciton dissociation at high field."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Explains why two-terminal resistance of a quantum Hall device is quantized at $h/\\nu e^2$ when contact resistance is small, the criterion used to identify the bilayer quantum Hall insulator phase."},{"cited_title":"F., Shan, J","cited_arxiv_id":null,"evidence_quote":"Identifies graphite-gate-induced potential oscillations as a spurious cause of quantum oscillations in two-dimensional insulators, the alternative the authors rule out with their capacitor model."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the early theory that a magnetic field can increase exciton binding by reducing screening, which the authors invoke to explain the excitonic insulator region surviving to higher pair density at higher field."}],"review_version":1}