{"id":"e17cc6e6-10a6-4b83-98dd-4d784a20286d","arxiv_id":"2412.01329","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Intercalated tin forms a √3×√3 Mott lattice under graphene, with Hubbard bands visible in STS and EELS beside graphene's Dirac cone.","lead":"Tin atoms placed under a graphene sheet on silicon carbide form a small triangular pattern that shows signs of strongly correlated 'Mott' electronic bands. The result is a new platform for studying how a correlated electron layer interacts with graphene's fast-moving Dirac electrons.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Whether the √3×√3 phase is a Sn-pz Mott lattice is the load-bearing premise, but the identification rests only on LEED symmetry and bias-dependent STM contrast, and the paper itself says 'nominally Sn-√3'; no element-specific or momentum-resolved probe excludes a different √3 origin.","rationale":"I found no internal inconsistency that would invalidate the argument; the load-bearing concern is the one the reader identified. The experimental chain is internally coherent: √3 LEED reflexes appear with annealing; the STS pair-correlation function shows √3 symmetry at exactly the two bias voltages (-0.5 V, +1.0 V) where the Hubbard bands are claimed; the 1.2 eV EELS loss and the √3 LEED intensity grow together (Figs. 2e, 5b); and the DMFT yields LHB/UHB at roughly ±(0.5-0.7) eV for both T4 and H registries, matching the L1/L2-U1/U2 features. The theory is not the fragile part: the Hubbard bands are robust to local registry and hybridization strength, and the C6v selection rule suppressing hybridization at the H position is a clean symmetry argument. The fragile part is the identification of the √3×√3 lattice with Sn. The paper's own hedges ('nominally Sn-√3 phases', 'contains locally Sn-√3 phases'), the absence of any element-specific or momentum-resolved measurement of this phase, the mixed-phase LEED superposition, and the known ability of intervalley scattering to produce a graphene √3 reconstruction near Sn islands (Ref. 65) together leave a concrete alternative interpretation open: the √3 symmetry, the -0.5/+0.7 V peaks, and the 1.2 eV loss could originate from something other than a half-filled Sn-pz band. The empty NOMAD DOI (Ref. 50) additionally prevents verification of the DFT underpinning. The Q1/Q2 features at ±14/18 meV sit inside the ±60 meV phonon-related gap structure of the same spectra, so the correlated-metal reading of those small peaks is the least supported part of the paper, though it is secondary to the main Mott claim. These are verification gaps, not demonstrated errors, so the appropriate verdict remains CONDITIONAL: the claim is plausible and the evidence is consistent, but an element-specific structural/chemical check is required to confirm the assignment. I therefore leave the reader's verdict unchanged.","tokens_in":15460,"tokens_out":19786,"duration_ms":168704,"concrete_test":"Perform the same de-intercalation in a system with in-situ X-ray photoelectron spectroscopy (Sn 3d5/2, Si 2p) and quantitative LEED-I(V) of the √3 spots, then compare the experimental I(V) curves with DFT-simulated I(V) of the Sn-√3 T4 and H models and check that the Sn 3d coverage matches the √3-domain fraction (near 1/3 ML). If the Sn coverage is inconsistent or the √3 I(V) curves cannot be reproduced by the Sn models, the structural premise, and with it the Mott-Hubbard interpretation, fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing premise is that the √3×√3 structure formed by partial de-intercalation (Figs. 2-4) is a half-filled Sn-pz triangular lattice, so that the ~1.2 eV gap seen in STS and EELS (Figs. 5, 6) constitutes Mott-Hubbard bands. That premise is not element-specifically established. The phase is assigned from the appearance of √3 LEED reflexes after annealing and from √3 distances in the angularly averaged pair-correlation function at -0.5 V and +1.0 V (Fig. 4d); no XPS or AES shows Sn in these domains, and the tin coverage is known only globally from a quartz microbalance (Methods). The paper itself refers to 'nominally Sn-√3 phases' in the EELS section and says the target STM area 'contains locally Sn-√3 phases' (Fig. 3e). Since a graphene √3×√3 reconstruction from intervalley scattering is observed over the Sn-(1×1) phase (Ref. 65) and the de-intercalated surface also contains Sn-(1×1) patches and disordered buffer regions, the √3 LEED pattern is a superposition and could in principle arise from graphene or SiC-related remnants rather than an ordered Sn layer. If the √3 periodicity is not the Sn lattice, the DMFT calculation, which assumes the Sn-√3 geometry, one Sn electron per cell, and U=1.2 eV, does not apply to the measured spectra, and L1/L2-U1/U2 could have another origin (e.g., SiC-derived interface states). The paper's annealing correlation (√3 LEED intensity and the 1.2 eV EELS peak grow together, Figs. 2e, 5b) and the robustness of the DMFT Hubbard bands to T4/H and V0 are genuine points in its favor; the problem is an unverified premise, not an internal contradiction. Additionally, the promised DFT repository (Ref. 50) has an empty DOI, so the supporting data cannot be checked.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a combined SPA-LEED, STM/STS, EELS, DFT, and DMFT study of Sn intercalated under epitaxial graphene on SiC(0001). After full Sn intercalation and partial de-intercalation, the authors observe √3×√3 LEED reflexes and a surface phase with 6×6/6√3 periodicities; STS shows peaks near -0.5 V and +0.7 V with L1/L2 and U1/U2 substructure, and EELS shows a non-dispersing 1.2 eV loss. DFT+DMFT with U=1.2 eV produces lower and upper Hubbard bands from Sn-pz orbitals for both T4 and H registries, with hybridization- and charge-transfer-dependent spectra. The paper interprets these observations as Mott-Hubbard bands of a Sn-√3 triangular lattice coexisting with Dirac electrons, and emphasizes T4/H hybridization selection rules and charge-transfer tunability.","tokens_in":15872,"tokens_out":9475,"duration_ms":86593,"significance":"If the structural identification is correct, the work is significant: it would extend Mott physics into an epitaxial graphene platform, with a correlated triangular Sn layer proximitized to Dirac electrons, and the T4/H selection-rule asymmetry is an elegant and falsifiable prediction. The experimental probes are mutually consistent (STS gap ≈1.2 eV, EELS loss ≈1.2 eV, DMFT Hubbard gap on the same scale), and the DFT data are stated to be deposited on NOMAD. The robustness of the Hubbard bands to T4/H stacking is a strength. However, the significance is conditional: the element-specific identification of the √3 phase as Sn is missing, and the agreement between experiment and theory is partly enforced by the choice U=1.2 eV. These limitations need to be addressed before the central claim can be taken as established.","major_comments":[{"comment":"The load-bearing premise that the √3×√3 domains are an ordered Sn-pz triangular lattice is not established. The paper itself calls the phase 'nominally Sn-√3 phases' (Fig. 5 caption) and describes the STM area as 'contains locally Sn-√3 phases' (Fig. 3e caption); the √3 LEED pattern after de-intercalation is a superposition over a surface that contains Sn-(1×1) patches, disordered buffer regions, and the 6√3 graphene/SiC reconstruction (Fig. 3a). The Sn coverage is known only globally from a quartz microbalance, and the √3 distance in the pair-correlation function at -0.5 V and +1.0 V (Fig. 4d) is a symmetry analysis of tunneling contrast, not a chemical identification. Since a graphene √3 reconstruction from intervalley scattering is already observed on the Sn-(1×1) phase (Ref. 65), and SiC or buffer remnants can also produce √3 periodicities, the assignment of all subsequent 1.2 eV features to Sn-pz Mott bands requires element-specific confirmation (e.g., XPS/AES of the de-intercalated phase or element-specific STM/STS). Without it, the DMFT model in Eqs. (1)–(3), which assumes one Sn pz orbital per √3 cell, is not guaranteed to apply to the measured spectra.","section":"Structure of EG/Sn interface: SPA-LEED and STM (Figs. 2–4)"},{"comment":"The claimed 'excellent agreement' between theory and experiment is partly by construction. U is set to 1.2 eV 'as a representative value considering the previous studies on Sn surface systems' (Methods), and the Sn hopping parameters are taken from Sn/Si(111) (Refs. 33, 54), while V0 is selected so that the noninteracting tight-binding bands fit the DFT bands (Figs. 8c,d). The DMFT Mott gap at half-filling is controlled by U, so the computed ~1.2 eV separation between lower and upper Hubbard bands in Fig. 9 is not an independent prediction of the measured 1.2 eV STS gap and EELS loss; it is largely a restatement of the input U. An independent estimate of U for this specific Sn/graphene/SiC interface (e.g., constrained RPA or GW) and a scan over U would be needed to claim that the energy scale is a falsifiable outcome. This is a major issue because the paper's conclusion that the system is in the Mott regime rests on this scale.","section":"DMFT calculations and Figs. 8–9"},{"comment":"The assignment of the non-dispersing 1.2 eV EELS loss to Hubbard-band excitations is not uniquely supported by the data shown. The loss peak has FWHM=0.4 eV, no comparison with the disordered-buffer or Sn-(1×1) phases in the same loss range is presented, and the correlation with annealing (growth of √3 LEED intensity and loss intensity, Figs. 2e and 5) is also compatible with growth of an ordered non-Sn √3 phase. The paper does not calculate the EELS loss function or the joint density of states from the DMFT spectra, so the identification rests on energy coincidence with the chosen U. Given that the structural premise is also unproven, the 1.2 eV feature could also arise from SiC-related interface states or interband transitions. A momentum-resolved analysis or at least a computed loss function from the model would substantiate the assignment.","section":"Electronic structure: EELS and STS (Figs. 5–6)"}],"minor_comments":[{"comment":"Reference 50 has an empty DOI ('DOI: .'), and the Data Availability statement is therefore incomplete; the NOMAD link/identifier should be supplied.","section":"References"},{"comment":"Typos and broken text should be corrected, including 'caclulations are availble' in Ref. 50 and the stray 'and' at the end of the section before 'Interplay of hybridization' ('agrees well our theoretical findings. and').","section":"General"},{"comment":"The number of spectra and the procedure used to obtain the averaged peak positions and the Mott-gap variation in the inset are not stated; please clarify whether error bars were evaluated.","section":"Fig. 7b"},{"comment":"The caption states that the color bar indicates spectral weight in the lowest-lying non-interacting band, but the figure panels show momentum-integrated A(ω); please reconcile the caption with the panels.","section":"Fig. 9 caption"},{"comment":"The Methods state that EELS-LEED provides a momentum resolution around 0.001 Å^-1, but no details of the primary energy and angle-mapping geometry used for the dispersion in Fig. 5c are given; please specify.","section":"Methods"}],"recommendation":"major_revision","confidential_remarks":"This is an interesting and generally well-written manuscript, but the central claim depends on the chemical identity of the √3 phase as Sn. I recommend requiring element-specific evidence before publication. If such evidence cannot be provided, the manuscript should be reframed as a model-based interpretation with the structural premise explicitly labeled as an assumption. The missing DOI in the data-availability statement is a minor issue but should also be fixed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. The paper is worth reading: it reports a new geometry — Sn-√3 domains under epitaxial graphene by partial de-intercalation — and shows a ~1.2 eV gap in both STS and EELS that DMFT reproduces as Hubbard bands, with T4/H stacking explaining spatial variation. That internal consistency is genuine. But the load-bearing premise, that the √3 structure is a half-filled Sn triangular lattice, is inferred from LEED symmetry and bias-dependent STM contrast, not element-specific probes. The paper itself says \"nominally Sn-√3\" in the EELS section. If the √3 periodicity comes from graphene or SiC remnants, the Mott interpretation loses its basis.\n\nWhat is new: prior work had the Sn Mott state on bare SiC(0001), and Sn intercalation for doping and plasmons, but not the combined proximitized geometry. The symmetry argument — T4 vs H sites modulating hybridization via selection rules — matches the observed variation of Hubbard band positions across the 6√3 unit cell. The annealing correlation between √3 LEED intensity and the 1.2 eV loss peak is good evidence tying the loss to the √3 phase. DMFT robustness to T4/H and V0 also helps.\n\nSoft spots: phase identification is the first. No XPS or AES shows Sn in these domains; tin coverage is only globally known. The √3 LEED could have other sources, and the paper is honest about the uncertainty. Second, the theory is not independent: U is from prior Sn surface work, hoppings from prior studies, and V0 is calibrated to match the paper's own DFT. The Mott gap follows from the choice of U, not from first principles. It's interpretive, not predictive. Third, the DFT repository DOI is empty.\n\nNone of this is fatal. The data are mutually consistent, the paper does not overclaim, and the hypothesis is testable. A serious referee should push for element-specific confirmation (XPS/AES, ARPES of the Sn bands) and a filled repository.\n\nFor surface scientists and 2D correlated-electron researchers this is a valuable paper. I'd accept it for peer review. If the Sn-√3 assignment holds, it's an important platform.","headline":"The Sn-√3 Mott layer under graphene is plausible and the STS/EELS/DMFT consistency is real, but the phase assignment is not element-specifically proven.","tokens_in":16505,"tokens_out":3857,"would_cite":true,"duration_ms":32951,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.27.+a","73.22.Pr","73.20.-r"],"model":"deepseek-v4-flash","headline":"The paper claims that a √3×√3 tin layer intercalated under epitaxial graphene on SiC(0001) is a strongly correlated Mott-Hubbard system whose pz electrons split into lower and upper Hubbard bands while graphene's Dirac cone survives.","keywords":["epitaxial graphene","tin intercalation","Mott-Hubbard bands","√3×√3 Sn superstructure","dynamical mean-field theory","scanning tunneling spectroscopy","electron energy loss spectroscopy","Dirac electrons"],"falsifier":"A spatial map of the √3 domains by element-specific core-level photoemission or Auger spectroscopy showing that tin coverage is far below one atom per √3 cell, or an angle-resolved photoemission spectrum showing no flat Sn-$p_z$ band with Hubbard satellites near -0.5 eV and +0.7 eV, would refute the central claim.","tokens_in":116,"feed_emoji":"🔬","tokens_out":13972,"duration_ms":237496,"temperature":0.7,"pith_summary":"This paper sets out to show that tin atoms intercalated beneath epitaxial graphene on SiC(0001) form a √3×√3 triangular lattice whose out-of-plane orbitals are so strongly correlated that they split into lower and upper Hubbard bands, while the graphene sheet above keeps its relativistic Dirac bands. The evidence is tunneling spectra with paired peaks near -0.5 V and +0.7 V (with L1/L2 and U1/U2 substructure), a non-dispersing 1.2 eV energy-loss peak, and dynamical mean-field calculations that produce the same Hubbard bands from a flat tin-derived band. If the assignment holds, the interface becomes a single epitaxial platform where Mott physics coexists with, and hybridizes to, a Dirac electron gas, with the tin registry (T4 versus H) and charge transfer as control knobs.","feed_headline":"Tin under graphene produces Mott-Hubbard bands beside Dirac electrons","feed_subtitle":"Experiments and DMFT calculations agree: tin's pz electrons split into Hubbard bands while graphene keeps its Dirac cone.","key_machinery":"The engine is an effective Hubbard-model Hamiltonian $H = H_0 + H_{\\mathrm{int}}$ built from graphene and tin $p_z$ orbitals (the electron lobes sticking out of the layer): carbon and tin hopping terms, a single hybridization parameter $V_0$ between Sn and C, a crystal-field offset $\\Delta = \\epsilon_{\\mathrm{Sn}} - \\epsilon_{\\mathrm{C}}$, and an on-site Hubbard interaction $U = 1.2$ eV on tin, solved by dynamical mean-field theory, a method that treats local quantum fluctuations beyond static mean field, using a continuous-time quantum Monte Carlo impurity solver. The flat Sn-$p_z$ band near the Fermi energy is the seed Mott state; the graphene Dirac bands act as the relativistic bath. A symmetry selection rule controls their coupling: at the graphene K point the Dirac states transform as the two-dimensional representations $E_1$ and $E_2$ under the sixfold symmetry of the hexagon, while the tin $p_z$ orbital transforms as the one-dimensional $A_1$ representation, so tin at the hexagon center (H) barely hybridizes with graphene, whereas tin below a carbon atom (T4) hybridizes strongly. The offset $\\Delta$ governs charge transfer, and together $V_0$ and $\\Delta$ determine whether the interface behaves as a Mott insulator or a doped correlated metal.","core_discovery":"The central claim is that a locally ordered Sn-√3×√3 phase at the graphene/SiC interface hosts Mott-Hubbard bands formed by Sn pz orbitals, directly in proximity to the Dirac cone of graphene. In STS, the occupied side shows split peaks L1/L2 near -0.43 and -0.68 eV and the unoccupied side shows U1/U2 near +0.7 eV; their separation matches the 1.2 eV non-dispersing EELS loss and the roughly 1 eV Mott gap obtained from DMFT. The calculations show that the flat Sn-pz band gives rise to robust lower and upper Hubbard bands for both T4 and H registries, while the T4 registry allows hybridization with graphene that broadens and splits the Hubbard bands and the H registry suppresses hybridization by symmetry, leaving sharper, atomic-limit-like bands. When charge transfer is increased, quasiparticle peaks appear near the Fermi level, matching the correlated-metal signatures seen close to zero bias.","pith_inferences":["An immediate test is angle-resolved photoemission of the √3 domains: it should resolve a nearly flat Sn-$p_z$ band split into Hubbard satellites near -0.5 eV and +0.7 eV, cleanly separating tin states from the graphene Dirac cone.","The registry-controlled hybridization suggests that other group-IV intercalants, such as Ge or Pb, might form the same proximitized Mott family on SiC(0001) with different spin-orbit coupling and Hubbard U.","Since the √3 phase is grown by partial de-intercalation of a full Sn monolayer, the domain boundaries are a natural place to look for one-dimensional correlated or metallic states not discussed in the paper.","If the Mott assignment is correct, the system offers a Kondo-like platform in epitaxial graphene: the flat Sn band plays the role of a correlated lattice and the Dirac cone the conduction bath, with the T4/H registry tuning the effective hybridization."],"forward_implications":["The 1.2 eV non-dispersing EELS loss and the roughly 1.2 eV separation between the L1/L2 and U1/U2 STS peaks are the same Mott-Hubbard gap, so the correlated state is visible in both tunnel and loss spectroscopy.","Because the flat Sn-$p_z$ band produces Hubbard bands for both T4 and H registries, the Mott state is robust to local stacking; the registry only controls how much hybridization broadens and splits the bands.","The electron doping found in the graphene sheet (about 7.8×10^12 cm^-2, Fermi level around 330 meV above the Dirac point) provides the charge reservoir that can drive the Sn layer from a Mott insulator toward a doped correlated metal with quasiparticle peaks near the Fermi energy.","In the H registry the doped system resembles a doped Mott insulator beside a charge reservoir, while in the T4 registry the doped Mott state hybridizes with Dirac electrons, a situation the paper compares to magic-angle twisted bilayer graphene.","Molecular doping of the graphene side (for example with F4-TCNQ) should tune the correlated state, since the paper identifies charge transfer, not only U and hybridization, as a control parameter."],"supporting_citations":[{"why":"It supplies the reference √3×√3 Sn Mott state on SiC(0001), including the hopping parameter and Hubbard U used in the model.","marker":"[38]"},{"why":"It provides the two-dimensional Mott-state reference for Sn on Si(111) and the longer-range tin hopping parameters used in the tight-binding model.","marker":"[33]"},{"why":"It establishes the Sn intercalation and de-intercalation preparation and the SPA-LEED calibration used to form the √3 phase.","marker":"[39]"},{"why":"It gives the graphene sheet-plasmon dispersion analysis used to determine the n-type doping and to separate the 1.2 eV loss peak from the plasmon.","marker":"[40]"},{"why":"It supplies the buffer-layer graphene growth method on SiC(0001) that the intercalation experiments start from.","marker":"[21]"},{"why":"It underpins the identification of the (6√3×6√3) buffer-layer reconstruction and its structural relation to the √3 phase.","marker":"[51]"},{"why":"It justifies the simplified √3×√3 supercell approximation used in the DFT calculations of the interface.","marker":"[52]"},{"why":"It provides the comparison system of a doped Mott insulator beside a charge reservoir, used to interpret the H-registry result.","marker":"[13]"},{"why":"It supplies the dynamical mean-field theory framework used to solve the Hubbard Hamiltonian and obtain the Hubbard bands.","marker":"[57]"}],"fun_headline_variants":["Sn interlayer induces Mott-Hubbard bands near Dirac cone","Tin proximity yields Mott-Hubbard bands beside Dirac cone","Hubbard bands from tin proximity in graphene","Mott states from Sn intercalation coexist with graphene's Dirac electrons"],"cache_read_input_tokens":18304,"weakest_assumption_plain":"The load-bearing premise is that the repeated diffraction pattern and the bias-dependent microscope contrast come from a tin triangular lattice beneath the graphene, an assignment made from symmetry and imaging rather than from element-specific or momentum-resolved measurements; if those domains are instead silicon-carbide or buffer-layer reconstructions, or if the local tin coverage differs from one atom per repeat unit, the strong-correlation interpretation has no substrate.","fun_headline_variants_meta":{"raw":{"variants":["Sn interlayer induces Mott-Hubbard bands near Dirac cone","Tin proximity yields Mott-Hubbard bands beside Dirac cone","Hubbard bands from tin proximity in graphene","Mott states from Sn intercalation coexist with graphene's Dirac electrons"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001596,"raw_usage":{"total_tokens":6328,"prompt_tokens":879,"completion_tokens":5449,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":495,"completion_tokens_details":{"reasoning_tokens":5380}},"tokens_in":495,"tokens_out":5449,"duration_ms":34125,"temperature":1.0,"reasoning_tokens":5380,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T04:29:42.869822+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A spatial map of the √3 domains by element-specific core-level photoemission or Auger spectroscopy showing that tin coverage is far below one atom per √3 cell, or an angle-resolved photoemission spectrum showing no flat Sn-$p_z$ band with Hubbard satellites near -0.5 eV and +0.7 eV, would refute the central claim.","supporting_citations":[{"cited_title":"Triangular Spin-Orbit-Coupled Lattice with Strong Coulomb Correlations: Sn Atoms on a SiC(0001) Substrate","cited_arxiv_id":null,"evidence_quote":"It supplies the reference √3×√3 Sn Mott state on SiC(0001), including the hopping parameter and Hubbard U used in the model."},{"cited_title":"ager, M.; Brand, C.; Weber, A. P.; Fanciulli, M.; Dil, J. H.; Pfn\\","cited_arxiv_id":null,"evidence_quote":"It provides the two-dimensional Mott-state reference for Sn on Si(111) and the longer-range tin hopping parameters used in the tight-binding model."},{"cited_title":"Exploring graphene-substrate interactions: plasmonic excitation in Sn-intercalated epitaxial graphene","cited_arxiv_id":null,"evidence_quote":"It gives the graphene sheet-plasmon dispersion analysis used to determine the n-type doping and to separate the 1.2 eV loss peak from the plasmon."},{"cited_title":"V.; Bostwick, A.; Horn, K.; Jobst, J.; Kellogg, G","cited_arxiv_id":null,"evidence_quote":"It supplies the buffer-layer graphene growth method on SiC(0001) that the intercalation experiments start from."},{"cited_title":"Structural and electronic properties of epitaxial graphene on SiC(0001): a review of growth, characterization, transfer doping and hydrogen intercalation","cited_arxiv_id":null,"evidence_quote":"It underpins the identification of the (6√3×6√3) buffer-layer reconstruction and its structural relation to the √3 phase."},{"cited_title":"Ab Initio Study of Graphene on SiC","cited_arxiv_id":null,"evidence_quote":"It justifies the simplified √3×√3 supercell approximation used in the DFT calculations of the interface."},{"cited_title":"I.; Yan, B.; Sangiovanni, G.; Wehling, T.; Valent \\' , R","cited_arxiv_id":null,"evidence_quote":"It provides the comparison system of a doped Mott insulator beside a charge reservoir, used to interpret the H-registry result."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the dynamical mean-field theory framework used to solve the Hubbard Hamiltonian and obtain the Hubbard bands."}],"review_version":1}