{"id":"94cad19a-d258-458a-809f-4c5a5666b866","arxiv_id":"1908.08693","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Photodoping a repulsive Hubbard Mott insulator can stabilize a metastable eta-pairing superconducting phase, detectable through a zero-frequency conductivity peak and negative high-frequency conductivity.","lead":"Shining light to create extra electrons and holes in a strongly interacting insulator could make it superconduct through a hidden eta-pairing state. The paper predicts a clear optical fingerprint, a zero-frequency conductivity spike plus negative conductivity at high frequencies, that experiments could look for.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed spontaneous η-pairing phase rests on finite-seed steady-state responses; no zero-field extrapolation is shown, and the paper's own direct real-time protocol fails its stated seed-independence criterion.","rationale":"The reader's weakest assumption was that the shifted-fermion-bath steady state faithfully represents a generic photodoped state. That is a valid concern about external validity, and the failed direct-excitation protocol in Appendix D.2 supports it. However, I find an even more immediate and decisive internal gap: even if the bath representation is accepted, the evidence for spontaneous symmetry breaking in the steady-state phase diagram is based on finite h_x responses without extrapolation to h_x = 0. The paper's own Appendix D.2 states the correct seed-independence criterion and then reports that the direct real-time protocol fails it. This does not disprove the eta-paired phase, but it means the central claim is not closed by the data presented in this manuscript. The analytic effective model, the D ∼ |η|^2 scaling, and the independent companion-paper evaporative protocol are genuine supporting evidence, which is why I would not move the verdict to REJECT. A conditional acceptance with a request for the zero-field extrapolation is the appropriate disposition, matching the reader's CONDITIONAL verdict. Thus I leave the verdict unchanged while sharpening the specific concern that must be settled.","tokens_in":16198,"tokens_out":4696,"duration_ms":47642,"concrete_test":"Recompute the steady-state η order parameter at the Fig. 1(b) arrow point (d ≈ 0.4, β_eff ≈ 7.8, U = 8, Γ = 0.05) for h_x = 10^-1, 10^-2, 10^-3, 10^-4, and 10^-5 with the same NCA/DMFT setup, and fit ⟨η_x⟩ versus h_x. If the h_x → 0 intercept is zero within numerical error, then the claimed spontaneous η-pairing phase in the steady-state calculation is not supported; if the intercept is nonzero and stable, the finite-seed concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires a spontaneously symmetry-broken η-paired state, not merely an enhanced pairing response. The steady-state phase diagram in Section IV is computed with a fixed pairing seed h_x = 0.0001, and the paper states that Re⟨d↓d↑⟩ ≳ 0.2 'clearly indicates the symmetry breaking.' However, no extrapolation to h_x → 0 is presented, so a large but finite susceptibility (χ ≈ 2000 at that seed) is not excluded. The phase boundary is likewise drawn at a finite susceptibility threshold (χη ∼ 10^3), which is not the same as a divergent susceptibility or a nonzero order parameter in the zero-field limit. The paper itself supplies the strict criterion in Appendix D.2: 'a strict criterion for the spontaneous symmetry breaking can be that the final state becomes independent of the size of the initial pairing field.' For the direct excitation protocol, a test with h_x = 0.001 shows that the long-time state still depends on the initial pairing field, and the authors conclude that the state 'has not yet entered the η-pairing state' despite χη ∼ 15. This is an explicit, in-scope admission that the strongest real-time test performed within this manuscript does not establish the spontaneous phase. The companion evaporative-cooling protocol (Ref. 45) does show a persistent order parameter beyond t ∼ 100, but the data in this manuscript do not independently confirm the zero-field and seed-independent limit for the steady-state phase diagram. Because the claimed hidden phase is defined by spontaneous symmetry breaking, the absence of a zero-field extrapolation is the most load-bearing gap in the argument.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the repulsive Hubbard model under photodoping, modeled by coupling the lattice to shifted fermion reservoirs, using non-equilibrium dynamical mean-field theory. The authors report a large eta-pairing susceptibility and a finite eta-pairing order parameter when a small pairing seed field hx = 0.0001 is applied, and map out a phase diagram in doublon density d and effective inverse temperature beta_eff. They derive an effective two-liquid model with an eta-exchange interaction, compute the phase stiffness D = 4 J_ex |eta|^2, and show that the optical conductivity in the paired state has a delta-function peak at zero frequency and negative conductivity near the Hubbard U. They also present real-time protocols, one of which (evaporative cooling) produces a persistent eta-paired state. The central claim is that photodoping can stabilize a metastable eta-pairing superconducting phase in a Mott insulator.","tokens_in":16388,"tokens_out":2339,"duration_ms":23277,"significance":"If the central claim holds, this is a significant step toward understanding light-induced superconductivity in Mott insulators, providing a concrete microscopic mechanism (doublon-hole exchange) and falsifiable optical fingerprints. The paper has strong analytical components: the effective Hamiltonian derivation in Appendix B and the phase-stiffness expression D = 4 J_ex |eta|^2 are independent analytic results, and the data collapse in Fig. 4 is a nontrivial consistency check across protocols and parameters. The main weakness is that the spontaneous symmetry breaking is not fully demonstrated: the steady-state phase diagram relies on a finite pairing seed, and the paper's own direct real-time protocol fails its stated seed-independence criterion. The distinction between a large but finite pairing susceptibility and a true symmetry-broken phase is load-bearing for the claim.","major_comments":[{"comment":"The steady-state phase diagram is computed at fixed pairing seed hx = 0.0001, and the order parameter Re<d_down d_up> >~ 0.2 is taken as evidence of symmetry breaking. No extrapolation to hx -> 0 is shown, so a large but finite susceptibility (chi ~ 2000 at that seed) is not excluded. A divergence of chi or an order parameter that survives in the hx -> 0 limit is required to establish a spontaneous eta-pairing phase; the present data do not provide it.","section":"Section IV, Fig. 2"},{"comment":"The paper itself states that a strict criterion for spontaneous symmetry breaking is that the final state becomes independent of the initial pairing field size. For the direct excitation protocol, the test with hx = 0.001 shows such dependence, and the authors conclude that the state 'has not yet entered the eta-pairing state' despite chi_eta ~ 15. This admission directly undermines the claim that the hidden phase is a generic consequence of photodoping; at minimum, the steady-state results need to pass the same seed-independence test or be accompanied by a zero-field extrapolation.","section":"Appendix D.2"},{"comment":"The phase boundary is drawn at a finite susceptibility threshold (chi_eta ~ 10^3), which is not equivalent to a divergent susceptibility. The authors should either demonstrate that chi_eta diverges at the boundary (e.g., by scaling with hx or with lattice size) or rephrase the boundary as a crossover line. As written, the phase diagram conflates an enhanced response with a true thermodynamic phase.","section":"Section IV, Fig. 2"}],"minor_comments":[{"comment":"There is a typo in 'orgainzed' in the introduction; the organization paragraph should be corrected.","section":"Section I"},{"comment":"The text 'instrinsic doublon-holon pairing mechanism' contains a spelling error ('instrinsic' should be 'intrinsic').","section":"Section IV.B"},{"comment":"The delta-function peak in Re sigma at omega = 0 is said to be 'not shown'; it would be clearer to state this explicitly in the main text as well, and to explain how the Drude weight D is extracted from the numerical data.","section":"Fig. 3 caption"},{"comment":"The negative-temperature region is described as obtained by reflection, but no data points are shown for that region; a brief statement on the numerical or symmetry-based reasoning would help.","section":"Fig. 2 and Section IV"},{"comment":"Typo: 'narraow' should be 'narrow' in the description of the energy bands.","section":"Appendix D.1"}],"recommendation":"major_revision","confidential_remarks":"The paper makes an interesting and potentially important claim, and the analytic effective model is a genuine contribution. However, the central evidence for a spontaneous eta-pairing phase rests on finite-seed steady-state calculations and on a real-time protocol that is described in a companion paper (Ref. 45). The manuscript's own direct real-time test fails the stated seed-independence criterion. I would need to see either a clear hx -> 0 extrapolation or a successful seed-independence test for the steady-state phase diagram before recommending acceptance. The authors should also clarify whether the phase boundary is thermodynamic or a crossover."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is worth reading if you work on light-induced superconductivity or DMFT. What's genuinely new: it argues that generic photodoping of a repulsive Hubbard Mott insulator can produce a metastable eta-pairing superconductor without relying on SO(4) symmetry. Earlier work needed symmetry protection or specific driving; this one proposes an intrinsic doublon-hole exchange mechanism, derives a two-liquid effective model that gives D ~ |eta|^2, and checks it against steady-state DMFT and two real-time protocols. The optical conductivity prediction (delta peak plus negative conductivity near U) gives experiments a concrete target. The D vs |eta|^2 collapse in Fig. 4 across protocols is a real consistency check, not a fit.\n\nThe soft spot is the symmetry-breaking claim itself. The steady-state phase diagram is computed with a fixed pairing seed h_x = 0.0001, and the phase boundary is drawn at a susceptibility threshold. There is no h_x -> 0 extrapolation, so a large but finite susceptibility is not excluded. The paper's own strict criterion in Appendix D.2 — final state independent of the initial seed — is not met by the direct excitation protocol, and they say so plainly. That is honest and good, but it means the central claim of a robust spontaneously symmetry-broken phase is not fully closed by the data in this manuscript. The evaporative-cooling protocol from Ref. 45 does show persistence, but it is not independently demonstrated here beyond the reference.\n\nMinor points: the bath-coupling model is a reasonable stand-in for photodoping but the universality assumption (d and beta_eff suffice) is not rigorously justified; NCA error bars are absent; the Meissner effect is inferred from the phase stiffness, not computed. None of those are deal-breakers.\n\nThe math in the effective model derivation is careful and the paper is honest about its limits. I'd send it to referees — the within-subfield importance is high and the gaps are technical and addressable rather than fatal. The authors should be asked to do the zero-field extrapolation or explicitly soften the spontaneous-symmetry-breaking language. If they can't close that gap, the paper still has value as a proposed mechanism with a concrete experimental fingerprint. So engage with it, but read the phase diagram with the finite-seed caveat in mind.","headline":"A plausible and partly novel mechanism for eta-pairing in photodoped Mott insulators, with the symmetry-breaking claim not fully closed by the numerics.","tokens_in":17068,"tokens_out":1890,"would_cite":true,"duration_ms":20174,"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":"Photodoping a strongly repulsive Hubbard Mott insulator can stabilize a metastable eta-pairing superconducting phase without relying on the SO(4) symmetry of half filling.","keywords":["eta pairing","photodoped Mott insulator","Hubbard model","light-induced superconductivity","non-equilibrium dynamical mean-field theory","optical conductivity","doublon-hole condensate","metastable phase"],"falsifier":"Measure the optical conductivity of a strongly photodoped Mott insulator at $d \\gtrsim 0.3$ and low effective temperature: finding no zero-frequency delta peak in $\\mathrm{Re}\\,\\sigma(\\omega)$, no $1/\\omega$ divergence in $\\mathrm{Im}\\,\\sigma(\\omega)$, and no negative conductivity near $\\omega \\approx U$ would contradict the eta-pairing phase. Equivalently, a dynamical mean-field simulation of the direct excitation protocol run to much longer times that never develops a spontaneous eta order parameter for $d \\gtrsim 0.3$ and $\\beta_{\\mathrm{eff}} \\gtrsim 6$ would falsify the claimed universality.","tokens_in":15888,"feed_emoji":"⚡","tokens_out":17042,"duration_ms":139344,"temperature":0.7,"pith_summary":"The paper argues that injecting doublon-hole pairs into a repulsive Hubbard Mott insulator, by photodoping or any protocol that creates cold charge carriers, can drive the system into a metastable superconducting phase built from eta pairs. The phase is claimed to be intrinsic to the local electron-electron interaction, to persist over a wide range of doublon densities and effective temperatures, and to require no fine-tuned driving or the SO(4) charge-spin symmetry of the half-filled model. If correct, this gives a generic microscopic route to light-induced superconductivity, with a zero-resistivity delta peak in the optical conductivity and negative conductivity near the Hubbard gap as distinctive pump-probe signatures. The claim is supported by steady-state dynamical mean-field calculations, an effective two-liquid model, and real-time simulations that leave a symmetry-broken eta state after the drive is removed.","feed_headline":"Photodoping can turn a Mott insulator into a superconductor","feed_subtitle":"Photocreated doublon-hole pairs condense into a zero-resistance state with a clear optical fingerprint.","key_machinery":"The central object is the eta-pairing pseudospin $\\eta_i$, defined by $\\eta_i^+ = \\theta_i d_{i\\uparrow}^\\dagger d_{i\\downarrow}^\\dagger$ and $\\eta_i^z = (n_i - 1)/2$, whose transverse expectation value is a staggered doublon-hole condensate. The argument runs through a Schrieffer-Wolff-derived two-liquid effective model in which doublon-hole pairs interact through an exchange term $-J_{\\mathrm{ex}} \\eta_i \\cdot \\eta_j$ while singly occupied sites interact through $+J_{\\mathrm{ex}} S_i \\cdot S_j$, with a common coupling $J_{\\mathrm{ex}} = 2t_0^2/U$; this shared coupling is what lets eta order develop away from half filling. Numerically, the phase diagram and optical response are carried by non-equilibrium dynamical mean-field theory in the Nambu-Keldysh formalism on the Bethe lattice, solved with the non-crossing approximation.","core_discovery":"Using non-equilibrium dynamical mean-field theory on an infinite-coordination Bethe lattice with the lattice coupled to two shifted fermion reservoirs, the authors find a spontaneous staggered eta-pairing order at $U=8$ once the doublon density reaches $d \\gtrsim 0.3$ and the inverse effective temperature reaches $\\beta_{\\mathrm{eff}} \\gtrsim 6$. The order parameter is $\\eta_i^+ = \\theta_i d_{i\\uparrow}^\\dagger d_{i\\downarrow}^\\dagger$, and the instability is traced to an intrinsic doublon-hole exchange interaction $J_{\\mathrm{ex}} = 2t_0^2/U$ that appears in the effective two-liquid Hamiltonian alongside the usual spin exchange. The condensed phase shows ideal-metallic optical response: a zero-frequency delta function in $\\mathrm{Re}\\,\\sigma(\\omega)$, a $1/\\omega$ tail in $\\mathrm{Im}\\,\\sigma(\\omega)$, zero DC resistivity, and a London equation $j = -D A$ with phase stiffness $D \\approx 4 J_{\\mathrm{ex}} |\\eta|^2$. The same physics is found with protocols beyond the bath setup, including an evaporative-cooling real-time protocol that leaves the eta-pairing phase intact after the external coupling is switched off.","pith_inferences":["A testable extension: measure the real part of the optical conductivity of a photodoped Mott insulator across the Hubbard gap; a sign change to negative values near $\\omega \\approx U$ at high doublon density and low effective temperature would be a direct discriminator for eta-pairing.","By the particle-hole duality noted in the paper, the eta-pairing phase is dual to a ferromagnetic state, which suggests that disorder or longer-range hopping should destabilize it less than it destabilizes antiferromagnetism; this could be checked by adding such terms to the steady-state calculation.","If the bath steady state is truly generic, then any scheme that produces cold doublon-hole pairs, including chemical doping or cold-atom experiments, should show eta-pairing correlations rather than only optical pumping; this would widen the search for hidden superconductivity in strongly correlated insulators."],"forward_implications":["A strongly photodoped Mott insulator with $d \\gtrsim 0.3$ and $\\beta_{\\mathrm{eff}} \\gtrsim 6$ should behave as an ideal conductor: a zero-frequency delta peak in $\\mathrm{Re}\\,\\sigma(\\omega)$, a $1/\\omega$ tail in $\\mathrm{Im}\\,\\sigma(\\omega)$, and zero DC resistivity.","The nonzero zero-frequency current-current correlation implies the London equation $j = -D A$, so the eta phase should exhibit the Meissner effect like a true superconductor.","Negative conductivity at frequencies near $U$ in the eta phase gives a concrete optical fingerprint that distinguishes eta-pairing from ordinary metallic photodoping in pump-probe experiments.","Small symmetry-breaking perturbations such as next-nearest-neighbor hopping $t_1 = 0.1t_0$ leave the eta-pairing phase intact, so the effect is not tied to the SO(4) symmetry of the half-filled Hubbard model.","The superfluid stiffness scales as $D \\approx 4 J_{\\mathrm{ex}} |\\eta|^2$, so the phase stiffness is controlled by short-range eta correlations and may already be observable at doublon densities below the symmetry-breaking threshold."],"supporting_citations":[{"why":"Defines the eta-paired states of the Hubbard model and their staggered superconducting order parameter.","marker":"[16]"},{"why":"Supplies the photodoped doublon-hole liquid picture and the extreme d=0.5 limit that motivates treating photodoping as a universal state.","marker":"[26]"},{"why":"Provides the evaporative cooling mechanism used to reach low effective temperatures in the real-time protocol.","marker":"[32]"},{"why":"Gives the direct-excitation baseline that found only enhanced local pairing susceptibility, which the paper's symmetry-broken result extends.","marker":"[34]"},{"why":"Supplies the non-equilibrium steady-state theory of photodoped Mott insulators on which the bath setup is built.","marker":"[37]"},{"why":"Provides the dynamical mean-field mapping of the lattice problem to a self-consistent impurity model.","marker":"[38]"},{"why":"Supplies the non-equilibrium Keldysh formalism and optical-conductivity framework used in the calculation.","marker":"[39]"},{"why":"Gives the Schrieffer-Wolff transformation used to derive the two-liquid effective model with the eta exchange interaction.","marker":"[44]"}],"fun_headline_variants":["Light-induced eta-pairing superconductivity in Mott insulators","Photodoping creates hidden eta-paired superconductor","Mott insulator turned superconductor by photodoping","Eta-pairing phase emerges in photodoped Mott insulators","Photodoping unlocks eta-paired superconducting state"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a steady state obtained by coupling the Hubbard lattice to shifted fermion reservoirs faithfully represents any generic photodoped state, characterized only by doublon density $d$ and effective temperature $T_{\\mathrm{eff}}$; the paper's own direct-excitation protocol did not reach a symmetry-broken state, so if a real pump does not thermalize into this two-parameter description, the computed phase could be a bath artifact.","fun_headline_variants_meta":{"raw":{"variants":["Light-induced eta-pairing superconductivity in Mott insulators","Photodoping creates hidden eta-paired superconductor","Mott insulator turned superconductor by photodoping","Eta-pairing phase emerges in photodoped Mott insulators","Photodoping unlocks eta-paired superconducting state"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000335,"raw_usage":{"total_tokens":1860,"prompt_tokens":948,"completion_tokens":912,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":564,"completion_tokens_details":{"reasoning_tokens":829}},"tokens_in":564,"tokens_out":912,"duration_ms":7607,"temperature":1.0,"reasoning_tokens":829,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:32:39.395875+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the optical conductivity of a strongly photodoped Mott insulator at $d \\gtrsim 0.3$ and low effective temperature: finding no zero-frequency delta peak in $\\mathrm{Re}\\,\\sigma(\\omega)$, no $1/\\omega$ divergence in $\\mathrm{Im}\\,\\sigma(\\omega)$, and no negative conductivity near $\\omega \\approx U$ would contradict the eta-pairing phase. Equivalently, a dynamical mean-field simulation of the direct excitation protocol run to much longer times that never develops a spontaneous eta order parameter for $d \\gtrsim 0.3$ and $\\beta_{\\mathrm{eff}} \\gtrsim 6$ would falsify the claimed universality.","supporting_citations":[{"cited_title":"Rosch , author D","cited_arxiv_id":null,"evidence_quote":"Supplies the photodoped doublon-hole liquid picture and the extreme d=0.5 limit that motivates treating photodoping as a universal state."},{"cited_title":"Cooling by photo-doping $--$ Light-induced symmetry breaking in the Hubbard model","cited_arxiv_id":"1904.00822","evidence_quote":"Provides the evaporative cooling mechanism used to reach low effective temperatures in the real-time protocol."},{"cited_title":"Enhancement of Local Pairing Correlations in Periodically Driven Mott Insulators","cited_arxiv_id":"1904.00857","evidence_quote":"Gives the direct-excitation baseline that found only enhanced local pairing susceptibility, which the paper's symmetry-broken result extends."},{"cited_title":"Nonequilibrium steady-state theory of photodoped Mott insulators","cited_arxiv_id":"2007.12511","evidence_quote":"Supplies the non-equilibrium steady-state theory of photodoped Mott insulators on which the bath setup is built."}],"review_version":1}