{"id":"ff751416-d41b-4d31-bc61-7dc298607c1b","arxiv_id":"1908.04453","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":3,"one_line_summary":"The claimed result is a one-particle Green's function G_k(ω) = δ/(ω + μ_eff - ε^eff_k), in which quasiparticle weight and bandwidth both scale with hole doping δ.","lead":"The paper derives an analytic Green's function for the doped Hubbard model and concludes that spin and charge separate, with the low-energy spectral weight proportional to hole doping. It is worth reading because it connects a formal equation-of-motion method to a long-standing picture of high-temperature superconductors, but the central formula largely reproduces slave-boson mean-field theory.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The closure that produces Eq. (10) is uncontrolled and self-inconsistent: n2=s2=1-δ forces zero double occupancy while Eq. (12) predicts nonzero D, and Eq. (11) violates the exact spectral sum rule in the atomic limit.","rationale":"The reader's verdict correctly identifies the SCA truncation as the load-bearing assumption. My independent check of the derivation reinforces this: Eq. (10) is obtained only after replacing (n_i)^2 and (s_i)^2 by c-numbers and discarding L≥3 correlation functions and static terms, with no small parameter controlling the discarded terms. The internal contradiction with Eq. (12) is a concrete symptom: the calculation uses D=0 as input and returns D≠0 as output, so the approximation is not self-consistent even at the level of its own equations. Additionally, the interpretation of δ as the lower-Hubbard-band spectral weight conflicts with the exact sum rule and the U→∞ atomic limit; if the claim is restricted to the coherent quasiparticle peak, it is standard Gutzwiller/slave-boson phenomenology and the paper's novelty must be scaled back accordingly. The paper remains a plausible but unvalidated analytic derivation; it should be conditioned on a numerical check and on a clarified statement of what 'spectral weight' means. Therefore the reader's CONDITIONAL verdict is appropriate and no change is needed.","tokens_in":15688,"tokens_out":12249,"duration_ms":127344,"concrete_test":"Perform exact diagonalization on a 4x4 or 8-site Hubbard cluster at U=8t0 and hole doping δ=1/8 (and δ=0), computing the spin-averaged spectral function A(k,ω). Extract the coherent quasiparticle weight Z_k from the pole residue of the low-energy peak and the peak dispersion; compare with Eq. (10). Independently measure D=<n↑n↓>, n2=<n^2>, and s2=<s^2>. If D differs from zero by more than O(t0/U), or if n2≠s2 at that order, the SCA input n2=s2=1-δ is falsified and the δ-proportional result of Eq. (10) is not established. Also check the integrated weight of the lower Hubbard band against the exact sum rule: it should be ≈(1+δ)/2 in the atomic limit, not δ.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The decisive step is the soft cut-off approximation (SCA) in Sec. III: in Eqs. (7)-(8), the operators (n_i)^2 and (s_i)^2 are replaced by constants n2 and s2, and then n2=s2=1-δ is imposed because double occupancy is 'completely depleted'. This closure is what turns the hierarchy into the linear equation (9), whose spin-averaged solution (10) has numerator δ and effective dispersion ε_k^eff. The concern is that this closure is both uncontrolled and internally inconsistent. (i) n2=s2=1-δ is equivalent to D=<n↑n↓>=0, but Eq. (12) of the same paper derives D ∝ δ(1-δ) t0/U. Since exactly n2=1-δ+2D and s2=1-δ-2D, the SCA relation is wrong at O(t0/U), changing the coefficients in Eqs. (38)-(39), and hence changing ε_k^eff and the alleged spinon term at precisely the order the paper claims to control. No error bound is given for dropping all L=3 correlation functions and the static correlators hidden in P^(n), P^(s) (Eqs. (43)-(44)). (ii) As written, Eq. (11) states that the lower Hubbard band spectral weight is δ and vanishes at half-filling. But the exact spin-resolved sum rule is ∫dω A_{kσ}(ω)=1; in the atomic limit U≫t at δ=0 the lower Hubbard band has weight 1/2. Thus the δ in Eq. (10) can only be the coherent quasiparticle weight, not the claimed lower-Hubbard-band spectral weight, and the approximation contains no controlled separation of coherent and incoherent parts.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies the single-band Hubbard model in the large-U limit using a hierarchical equation-of-motion approach. It closes the hierarchy with a 'soft cut-off approximation' (SCA) in which (n_i)^2 and (s_i)^2 are replaced by their expectation values n2 and s2, with n2=s2=1-δ, and all L=3 correlation functions and static correlators are discarded. The resulting spin-averaged Green's function is claimed to be G_k(ω)=δ/(ω+μ_eff-ε^eff_k), giving a lower-Hubbard-band spectral weight proportional to hole doping δ, a δ-renormalized hopping, and a 'spinon' contribution; the paper interprets these as spin-charge separation in a doped Mott insulator and compares the results with cuprate spectroscopy and numerical calculations.","tokens_in":16099,"tokens_out":16444,"duration_ms":176264,"significance":"If correct, the result would provide a simple analytic account of the doping dependence of the coherent spectral weight and of spin-charge separation in the Hubbard model. The starting equations of motion (1)-(6) are rigorous, and the algebraic steps leading to Eqs. (38)-(45) are explicit and checkable. However, the paper supplies no error estimate or numerical benchmark for the truncations, and the central prediction conflicts with exact spectral sum rules; as it stands the significance is therefore limited to a possibly instructive but uncontrolled approximation.","major_comments":[{"comment":"The claimed lower-Hubbard-band spectral function A^L_k(ω)=2πδ δ(ω+μ_eff-ε^eff_k) violates the exact one-particle spectral sum rule. In the atomic limit U→∞ at half-filling, the removal spectral weight per spin is <n_σ>=1/2; more generally, the occupied part of A_{kσ} integrates to the momentum distribution n_{kσ}, whose average is (1-δ)/2 per spin. A delta-function spectrum of weight δ cannot represent the lower Hubbard band and in particular vanishes at δ=0, where the lower Hubbard band is known to have weight 1/2 per spin. The factor δ in Eq. (10) is at most a coherent quasiparticle weight, but the paper provides no coherent/incoherent decomposition, so the abstract's central claim that the spectral weight of the one-particle Green's function is proportional to δ is not supported.","section":"Sec. III, Eq. (11)"},{"comment":"The SCA closure is not self-consistent. Since n2=<n_i>+2D and s2=<n_i>-2D, with D=<n↑n↓>, the assumption n2=s2=1-δ is equivalent to D=0. Equation (12) of the same paper predicts D∝δ(1-δ)t0/U, so the input and the output disagree at first order in t0/U. No small parameter is given for neglecting the L=3 correlation functions or the static terms P^(n) and P^(s) in Eqs. (43)-(44). The δ numerator of Eq. (10) simply reflects the exact filling factor 1-<n_i>=δ, not a dynamical prediction of the hierarchy; the approximate content of the result is therefore confined to ε^eff_k, whose coefficients are uncontrolled at the order where D contributes.","section":"Sec. III and Appendix, Eqs. (7)-(8), (12), (38)-(46)"},{"comment":"The derivation is explicitly restricted to |ω|≪t0 (see the text before Eq. (9) and after Eq. (10)), but the pole of Eq. (10) lies at ω=-μ_eff+ε^eff_k≈-U/2 for the value μ≈U/2 used in the Appendix. For U∼W=8t0 this is an order of magnitude outside the stated validity window, so the Green's function is evaluated in the very regime in which the low-frequency limits (45)-(46) were taken. The self-consistent location of the pole invalidates the use of those limits.","section":"Sec. III and Appendix C, Eqs. (9)-(10) and (45)-(46)"}],"minor_comments":[{"comment":"There are several typographical errors, including 'equation s of motion' and 'correlation function s'; the manuscript would benefit from a careful proofreading pass.","section":"Abstract"},{"comment":"The statement that 'most of spectral weight is transferred to the upper Hubbard band in the underdoping region' appears opposite to the atomic-limit expectation, where doping transfers weight from the upper to the lower Hubbard band; this should be clarified.","section":"Sec. III, after Eq. (11)"},{"comment":"The claimed consistency with spectroscopy and numerical calculations is qualitative only; a quantitative comparison table or benchmark against determinant quantum Monte Carlo or cluster dynamical mean-field theory would substantially strengthen the paper.","section":"Conclusion and Discussion"}],"recommendation":"reject","confidential_remarks":"For the editor: in my view the sum-rule violation alone is decisive, and the uncontrolled truncation makes the quantitative claims unreliable. A revision would require a new controlled closure or at least a coherent/incoherent decomposition with explicit error bounds, which is beyond a textual revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper's main claim—that the one-particle spectral weight in a doped Mott insulator is proportional to δ—is not new, and the derivation that leads to it has a load-bearing uncontrolled approximation. The final Green's function, Eq. (10), is, by the authors' own admission, essentially slave-boson mean-field theory with condensed holons. The truly new piece, a small spinon-like term, is a subleading correction that depends on the questionable soft cut-off approximation (SCA).\n\nWhat the paper does well: the hierarchical equation-of-motion setup is rigorous up to the point of truncation, and the identity (n_i)^2 + (s_i)^2 = 2n_i is used creatively. The equations are laid out carefully, and the authors are honest about the relation to existing slave-boson results. If you work on EOM truncation schemes, there is something to learn here about how such hierarchies can be organized.\n\nThe soft spots are real and substantial. First, the SCA sets n2 = s2 = 1 - δ, which is equivalent to double occupancy D = 0, but later the paper derives D ∝ δ(1-δ) t0/U in Eq. (12). That is an internal contradiction at order t0/U. You cannot assume D = 0 for the closure and then compute a nonzero D from the closed equations. The approximation therefore has no consistent small-parameter control. Second, Eq. (11) presents the spectral function in the lower Hubbard band as 2πδ δ(ω + ...), vanishing at half-filling. That violates the exact atomic-limit sum rule: for U ≫ t at δ = 0, the lower Hubbard band carries weight 1/2 per spin, not zero. The δ in Eq. (10) should be interpreted as the coherent quasiparticle weight, not the full lower-Hubbard-band spectral weight. The paper does not make that distinction and consequently overstates what it has proven. There is also no benchmark against exact diagonalization or DMFT, and no error estimate for neglecting all L=3 correlators and static quantities.\n\nWho is this for? Readers interested in analytic EOM approaches to strong correlations might find the machinery worth discussing, but the phenomenology is already known from slave-boson theory. The paper deserves a serious referee because the calculation is substantive and the EOM path is nonstandard, but as written it would require major revision to separate the coherent weight from the incoherent background and to reconcile the D = 0 closure with Eq. (12). I would not cite it in its current form.","headline":"A serious EOM calculation undermined by an uncontrolled closure and a spectral-weight sum-rule violation; the main result is slave-boson mean-field in new clothing.","tokens_in":16626,"tokens_out":4564,"would_cite":false,"duration_ms":49109,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In a doped Mott insulator described by the Hubbard model, the electron spectral weight is proportional to the hole doping $\\delta$, and spin-charge separation emerges at large on-site repulsion.","keywords":["Hubbard model","Mott insulator","spin-charge separation","hierarchical Green's function","soft cut-off approximation","hole doping","spectral weight","cuprate superconductors"],"falsifier":"Numerically solve the two-dimensional Hubbard model at $U\\approx 8t_0$ for several small dopings and integrate the lower-Hubbard-band spectral function; if the integrated weight does not track $\\delta$ in the low-energy window above the pseudogap, the central claim fails. A photoemission measurement of an underdoped cuprate comparing spectral weight with hole concentration would test the same scaling.","tokens_in":15435,"feed_emoji":"⚛️","tokens_out":10189,"duration_ms":94588,"temperature":0.7,"pith_summary":"This paper derives an analytic form for the one-particle Green's function of a doped Mott insulator described by the Hubbard model at large on-site repulsion $U$. The central result is that the lower-Hubbard-band spectral weight equals $2\\pi\\delta$ times a delta function, so the electron's spectral weight is proportional to the hole doping concentration $\\delta$ and vanishes at half filling. The excitation spectrum splits into a term proportional to $\\delta$, attributed to spin fluctuations, and a term proportional to $1-\\delta$, attributed to coupling between charge and spin fluctuations; the paper reads this as spin-charge separation intrinsic to the doped Mott insulator. If correct, the result explains why only doped holes contribute to low-energy transport and why the Drude weight scales with $\\delta$ in cuprate normal states.","feed_headline":"Hole doping sets the electron's spectral weight in a Mott insulator","feed_subtitle":"An analytic Green's function ties electron spectral weight to hole doping, vanishing exactly at half filling.","key_machinery":"The central machinery is the hierarchical Green's function approach, closed by the soft cut-off approximation. The hierarchy of equations of motion for the one-particle Green's function and its related multipoint correlation functions is truncated using the identity $(\\hat{n}_i)^2+(\\hat{s}_i)^2=2\\hat{n}_i$ and the relation $n_2+s_2=2(1-\\delta)$, with $n_2=s_2=1-\\delta$ at large $U$ because double occupancy is completely suppressed. Dropping all third-level correlation functions and static quantities closes the equations, yielding analytic expressions for the charge- and spin-fluctuation correlation functions $F^{(n)}_{ijq}(\\omega)$ and $F^{(s)}_{ijq}(\\omega)$, which then produce the one-particle Green's function in Eq. (10).","core_discovery":"For the Hubbard model with on-site repulsion $U$ of order the bandwidth, in the low-energy window above the pseudogap and below the hopping scale, the paper claims the one-particle Green's function takes the form $G_k(\\omega)=\\delta/(\\omega+\\mu_{\\mathrm{eff}}-\\varepsilon^{\\mathrm{eff}}_k)$, with $\\varepsilon^{\\mathrm{eff}}_k=(\\delta+n_2 J_U/(2U))\\varepsilon^0_k$ and $J_U=4t_0^2/U$. The numerator $\\delta$ means the lower-Hubbard-band spectral function is $A^L_k(\\omega)=2\\pi\\delta\\,\\delta(\\omega+\\mu_{\\mathrm{eff}}-\\varepsilon^{\\mathrm{eff}}_k)$, so the spectral weight of electrons is proportional to hole doping and disappears at half filling. The effective hopping is renormalized by $\\delta$, the spin-fluctuation contribution to the excitation spectrum is proportional to $\\delta$, and the charge-spin coupling term is proportional to $1-\\delta$, taking its maximum at undoping. The paper also derives that double occupancy is proportional to $\\delta(1-\\delta)t_0/U$, vanishing at both $\\delta=0$ and $\\delta=1$.","pith_inferences":["The derivation suggests a sharp experimental signature the paper does not spell out: integrated spectral weight in the lower Hubbard band, measured above the pseudogap across different cuprate families, should fall on a single straight line in $\\delta$ if the theory is right.","The closure at third level implies the first corrections to the $\\delta$ factor are of order $t_0/U$; measuring deviations from linear-in-$\\delta$ scaling at larger doping could locate the validity boundary of the approximation.","Spin-charge separation here is driven by large $U$ rather than by one-dimensional kinematics, so the theory predicts spinon-like continua in two dimensions near momenta such as $(\\pi,0)$ and $(\\pi/2,\\pi/2)$, which could be tested by neutron scattering.","The same closure could be extended to include pseudogap order parameters; if spin and charge separate above the pseudogap, introducing those orders into the hierarchy may modify the $1-\\delta$ coupling term."],"forward_implications":["At half filling $\\delta=0$ the lower Hubbard band contributes no spectral weight, so the system becomes a Mott insulator while the $1-\\delta$ charge-spin coupling term reaches its maximum.","In an underdoped cuprate above the pseudogap, the low-energy spectral weight and the Drude weight should scale linearly with hole doping, meaning only doped holes participate in low-energy transport.","The effective hopping is multiplied by $\\delta$, so spin fluctuations strongly suppress hole itinerancy and tend to localize electrons in the underdoped regime.","The same Green's-function form holds on any lattice, including a chain or a square lattice, and next-nearest-neighbor hopping $t'$ is renormalized to $\\delta t'$.","Double occupancy is proportional to $\\delta(1-\\delta)t_0/U$, vanishing both at $\\delta=0$ and at $\\delta=1$."],"supporting_citations":[{"why":"Supplies the hierarchical equations of motion for one-particle and multipoint Green's functions on which the derivation rests.","marker":"[31]"},{"why":"Provides the spinon-like excitation spectrum against which the $1-\\delta$ coupling term is identified.","marker":"[32]"},{"why":"Links the one-particle Green's function to measured quantities such as Drude weight, Hall coefficient, and spectral function.","marker":"[33]"},{"why":"Gives the Drude-weight result that the $\\delta$-proportional transport prediction extends.","marker":"[34]"},{"why":"Reports optical spectroscopy of cuprates used as the consistency check for the derived spectral weight.","marker":"[35]"},{"why":"Reports optical and spectral measurements of cuprates used as the consistency check for the derived spectral weight.","marker":"[36]"},{"why":"Reports infrared spectroscopy of underdoped cuprates used as the consistency check for the derived spectral weight.","marker":"[37]"},{"why":"Provides numerical spectral calculations for the doped Mott insulator with which the analytic Green's function is compared.","marker":"[38]"},{"why":"Provides normal-state numerical results above the pseudogap used as the consistency check.","marker":"[39]"}],"fun_headline_variants":["Spectral weight of Mott insulator electrons scales with hole doping","Doping controls electron weight and double occupancy in Mott insulator","Spin-charge separation found analytically in doped Mott insulator","Hole doping sets electron spectral weight, vanishing at half filling"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire $\\delta$ factor rests on treating $(\\hat{n}_i)^2$ and $(\\hat{s}_i)^2$ as their expectation values $n_2=s_2=1-\\delta$ and dropping all correlations involving three electron operators; if double occupancy is not fully suppressed or those correlations matter, the proportionality to $\\delta$ changes.","fun_headline_variants_meta":{"raw":{"variants":["Spectral weight of Mott insulator electrons scales with hole doping","Doping controls electron weight and double occupancy in Mott insulator","Spin-charge separation found analytically in doped Mott insulator","Hole doping sets electron spectral weight, vanishing at half filling"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000847,"raw_usage":{"total_tokens":3702,"prompt_tokens":975,"completion_tokens":2727,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":591,"completion_tokens_details":{"reasoning_tokens":2660}},"tokens_in":591,"tokens_out":2727,"duration_ms":17760,"temperature":1.0,"reasoning_tokens":2660,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:42:51.304759+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Numerically solve the two-dimensional Hubbard model at $U\\approx 8t_0$ for several small dopings and integrate the lower-Hubbard-band spectral function; if the integrated weight does not track $\\delta$ in the low-energy window above the pseudogap, the central claim fails. A photoemission measurement of an underdoped cuprate comparing spectral weight with hole concentration would test the same scaling.","supporting_citations":[{"cited_title":"Mancini and A","cited_arxiv_id":null,"evidence_quote":"Supplies the hierarchical equations of motion for one-particle and multipoint Green's functions on which the derivation rests."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the spinon-like excitation spectrum against which the $1-\\delta$ coupling term is identified."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Links the one-particle Green's function to measured quantities such as Drude weight, Hall coefficient, and spectral function."},{"cited_title":"Timusk, and B","cited_arxiv_id":null,"evidence_quote":"Gives the Drude-weight result that the $\\delta$-proportional transport prediction extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports optical spectroscopy of cuprates used as the consistency check for the derived spectral weight."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports optical and spectral measurements of cuprates used as the consistency check for the derived spectral weight."},{"cited_title":"Uchida, T","cited_arxiv_id":null,"evidence_quote":"Reports infrared spectroscopy of underdoped cuprates used as the consistency check for the derived spectral weight."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides numerical spectral calculations for the doped Mott insulator with which the analytic Green's function is compared."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides normal-state numerical results above the pseudogap used as the consistency check."}],"review_version":1}