{"id":"4f4fa020-4e18-4a98-a826-7beb1c7f56fb","arxiv_id":"2412.03648","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"The easy-axis triangular-lattice J1-J2 spin model hosts a substantial spin-liquid phase between the supersolid and stripe phases over a wide anisotropy range.","lead":"This paper maps the phase diagram of a frustrated quantum magnet on a triangular lattice with easy-axis anisotropy using large-scale numerical simulations. It finds that a spin-liquid phase survives even when the anisotropy is strong, and it refines the properties of the magnetic supersolid phase.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"At Δ=1.9 the claimed SL window is only ΔJ2≈0.017 and is set by a 36×6 scan with a ⟨S⟩<0.05 cutoff; at Δ=1.3, where a 1/Ly extrapolation was possible, it shifts the Y-SL boundary by ~0.014, so the large-Δ portion of the SL region is not yet established.","rationale":"The reader's weakest assumption was the linear 1/Ly extrapolation from 6×3 and 12×6 clusters, and the negative extrapolated m∞_U(1) values support that concern. However, the more damaging gap for the paper's central claim is that this extrapolation was only applied at Δ≤1.3; for Δ=1.6 and 1.9, the entire Y-SL and SL-stripe boundary set comes from a single 36×6 scan with an ad hoc ⟨S⟩<0.05 cutoff. The one place where the scan and the extrapolation can be compared, Δ=1.3, shows a shift of 0.014 in J2, which is nearly the full width of the Δ=1.9 SL window. Thus the 'surprising resilience ... to rather large Δ values' is not supported by the same level of evidence as the SL at Δ≤1.3. This is not a disagreement with consensus; it is an internal robustness problem in the boundary determination. The paper does provide multiple independent probes at Δ≤1.3 (S(q), VBS and chirality checks, correlation-length analysis), and those give credible evidence for an SL there, so the overall conditional verdict remains appropriate. The recommended action is to keep the CONDITIONAL verdict, with the condition being that the large-Δ SL window be validated by wider-cylinder non-scans and error-aware extrapolation.","tokens_in":24592,"tokens_out":5621,"duration_ms":58399,"concrete_test":"Run Sz-conserved DMRG non-scans on 20×6 and 20×9 YC cylinders (m up to 8000) at Δ=1.6, J2=0.09 and Δ=1.9, J2=0.09 and 0.095, with no pinning fields, and compute S(q) plus the exponential correlation length of the ordered moment induced by a pinned classical Y edge. If S(q) develops sharp K- or M-point peaks, or if the induced-moment correlation length exceeds ~3a, the putative SL window at Δ≥1.6 is a finite-size artifact. Separately, extend the fixed-aspect-ratio 1/Ly extrapolation of ⟨S⟩ to Ly=3,6,9 cylinders at Δ=1.6 and 1.9; if the extrapolated ⟨S⟩∞ remains above 0.05 across the nominal window, the Y-SL boundary shifts toward larger J2 and the large-Δ SL region shrinks or closes.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the spin-liquid region is 'surprising[ly] resilient' and extends to rather large Δ values rests on the location of the Y-to-SL boundary at Δ≥1.6, which is determined only by J2-scans on 36×6 YC cylinders using the finite-size criterion ⟨S⟩<0.05 (Fig. 2(b) and Table S.I). At Δ=1.9 the entire SL window spans J2∈[0.082,0.099], a width of only 0.017. Where an independent determination exists, at Δ=1.3, the non-scan 1/Ly extrapolation moves the Y-SL boundary from J2≈0.062 (scan) to J2≈0.076 (diamond in Fig. 8(b)), a shift of 0.014 that is comparable to the whole Δ=1.9 window. No S(q), correlation-length, or 1/Ly checks are reported for Δ=1.6 or 1.9. The paper itself acknowledges that the linear 1/Ly extrapolation becomes unreliable near the Ising limit and that 'possible non-linear effects in the finite-size extrapolations for the already small values of mU(1)' prevent a definite conclusion (SM Figs. S8–S9); the negative extrapolated m∞_U(1) values in those figures confirm the failure of the linear assumption in exactly the regime used to set boundaries. Because the headline result is the extent of the SL phase, and the largest-Δ portion of that phase is supported only by an unvalidated finite-size scan, the quantitative claim of resilience to large easy-axis anisotropy is the weakest load-bearing element.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies the S=1/2 easy-axis triangular-lattice J1-J2 XXZ model by DMRG on YC cylinders, supplemented by a minimally augmented spin-wave-theory (MAGSWT) analysis. It maps the easy-axis (Delta>1) phase diagram and claims a substantial spin-liquid region between the Y (supersolid) and collinear stripe-z phases, with the SL surviving to rather large easy-axis anisotropy. It also analyzes supersolid order parameters (m_uud, m_U(1), m_F), proposes an explanation for the vanishing ferromagnetic moment, and compares the Y-to-stripe-z boundary with MAGSWT.","tokens_in":24978,"tokens_out":7308,"duration_ms":73283,"significance":"If the phase diagram is correct, the paper fills a largely unexplored easy-axis region of a canonical frustrated quantum magnet and provides concrete guidance for rare-earth and transition-metal triangular-lattice compounds. The strengths of the paper are real: multiple independent probes support the SL at moderate anisotropy (exponential decay of induced order, broadened S(q), absence of VBS and chiral order, consistency on Ly=6 and Ly=9 cylinders), the transition-point tables are explicit, and the MAGSWT calculation is presented with enough detail to be checked. The main quantitative claim about the large-Delta extent of the SL, however, is not yet supported at the same level as the moderate-Delta part of the phase diagram.","major_comments":[{"comment":"The claim that the SL region is 'surprisingly resilient' and extends to rather large Delta values rests on the Y-to-SL boundary points at Delta=1.6 and 1.9, which are determined solely by 36x6 J2 scans using the <S><0.05 cutoff. At Delta=1.9 the entire claimed SL window is J2 in [0.082, 0.099], a width of 0.017. Where an independent determination exists, at Delta=1.3, the 1/Ly non-scan extrapolation moves the Y-SL boundary from J2 approximately 0.062 (scan) to J2 approximately 0.076 (diamond in Fig. 8(b)), a shift of 0.014 that is comparable to the whole Delta=1.9 window. No S(q), correlation-length, or 1/Ly checks are reported for Delta=1.6 or 1.9, so the large-Delta portion of the phase diagram is not quantitatively established by the presented data.","section":"Phase diagram and DMRG; Fig. 2(b); Table S.I; Fig. 8(b)"},{"comment":"The non-scan Y-SL boundary itself relies on a linear 1/Ly extrapolation from only two fixed-aspect-ratio clusters (6x3 and 12x6). With two points the linearity of the scaling cannot be tested, and the authors' own SM shows that the linear extrapolation yields negative m_infinity_U(1) values near the Y-SL boundary (SM Fig. S9), attributed to 'possible non-linear effects in the finite-size extrapolations for the already small values of mU(1)'. The blue 'conservative' boundary in Fig. 8(a) therefore carries an unquantified systematic error of the same order as the observed Delta=1.3 boundary shift (about 0.01 in J2).","section":"End Matter, 'Conservative phase diagram'; Fig. 8(b); SM S9"},{"comment":"The conclusion that m_U(1) vanishes continuously at the Ising limit, stated as 'contrary to previous works', is based on the same two-point linear extrapolation of m_U(1). In SM Fig. S8 the extrapolated values become small or negative near 1/Delta to 0, and the authors acknowledge that nonlinear effects make the extrapolation problematic in exactly this regime. The continuous-vanishing claim is therefore not established at the level of evidence used for the other order parameters; it should either be supported by additional data (for example, a Ly=9 check) or be presented as a tentative extrapolation with an explicit uncertainty estimate.","section":"Supersolid order parameters; Fig. 6; SM S8"}],"minor_comments":[{"comment":"The text states that the error bars for the transitions are the J2 steps in the scans, but it does not report the J2 step size used at each Delta; reporting these step sizes alongside Table S.I would allow readers to assess the boundary uncertainty directly.","section":"Phase diagram and DMRG"},{"comment":"The caption says 'Other symbols are as described in the text', but the stars, triangles, squares, and diamonds are introduced in several different paragraphs; a compact legend in the caption itself would improve readability.","section":"Fig. 8(a) caption"},{"comment":"In the SM figure panels labeled with the vertical axis 'mu', the plotted quantity appears to be the tilt angle theta; please relabel the axis as theta or define mu in the caption to avoid confusion with the MAGSWT chemical potential mu.","section":"SM Figs. S8 and S9"},{"comment":"The sentence stating that the level-crossing transition in the non-scans coincides with the inflection-point transition 'within the error bars' would be more useful with the numerical values or a pointer to the relevant SM table, since the main text alone does not quantify the comparison.","section":"Supersolid order parameters"}],"recommendation":"major_revision","confidential_remarks":"The moderate-Delta part of the spin-liquid region (Delta between 1.0 and 1.3) is supported by several independent probes, and the MAGSWT and supersolid-order-parameter analyses are solid contributions. The main issue is that the headline quantitative claim about the large-Delta extent of the SL is supported only by a single finite-size scan with a cutoff criterion, and the one available cross-check at Delta=1.3 shifts the boundary by an amount comparable to the claimed Delta=1.9 SL window. This is fixable in revision by adding non-scan or S(q) checks at Delta=1.6 and 1.9, or by softening the abstract and phase-diagram claims to reflect the current evidence. No concerns about novelty or citation practice."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take on arXiv:2412.03648. The easy-axis J1–J2 triangular phase diagram was genuinely unmapped, and this is the first serious DMRG attack on it. The paper does the standard things well: S(q) maps, correlation length checks, VBS and chiral probes, wider Ly=9 cylinders, and the MAGSWT boundary is an independent analytical input that agrees with DMRG near the Ising limit. The order parameter analysis of the Y phase is careful, and the suggestion that the ferromagnetic moment cancels via the S^z_triangle = 0 sector is a nice heuristic that deserves attention.\n\nThe soft spot is the quantitative extent of the spin liquid at large Δ. At Δ=1.9 the entire SL window is J2∈[0.082,0.099], a width of 0.017, and that rests on a 36×6 scan with a ⟨S⟩<0.05 cutoff. The one place where a non-scan extrapolation was possible, Δ=1.3, it shifts the Y-SL boundary by ~0.014—comparable to the whole Δ=1.9 window. Add that the 1/Ly extrapolations use only two cylinder sizes and produce negative extrapolated m_U(1) near the transitions, and the authors' own admission that nonlinearities are likely, and it becomes clear that the 'surprising resilience' of the SL to large easy-axis anisotropy is plausible but not yet established. The claim that m_U(1) vanishes continuously at the Ising limit is also under-supported for the same reason.\n\nThe Heisenberg-limit calibration of the 0.05 cutoff is sensible, and the existence of an SL for Δ up to at least 1.3 is well-supported by multiple probes. So the broad picture is probably right. What is load-bearing and weak is the large-Δ part of the phase diagram and the continuous-vanishing claim.\n\nThis paper deserves a serious referee. It will be cited as the reference for the easy-axis J1–J2 phase diagram, so the boundary uncertainties should be addressed before publication. I'd ask the authors for either wider cylinders (Ly=9) along the Y-SL boundary at Δ=1.6 and 1.9, or a more careful treatment of the extrapolation error, and a softened statement about the large-Δ SL extent.","headline":"A solid first map of the easy-axis J1–J2 triangular phase diagram whose large-Δ spin-liquid boundary needs more than one 36×6 scan to believe.","tokens_in":25515,"tokens_out":2943,"would_cite":true,"duration_ms":28063,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82B20","82B26","82B27"],"pacs":["75.10.Jm","75.40.Mg","75.50.Ee"],"model":"deepseek-v4-flash","headline":"This paper maps the easy-axis $S=1/2$ triangular-lattice $J_1$--$J_2$ phase diagram and finds that the spin liquid survives to large anisotropy, sitting between a supersolid and stripe order.","keywords":["spin liquid","triangular lattice","J1-J2 model","XXZ model","easy-axis anisotropy","supersolid","density-matrix renormalization group","frustrated magnetism"],"falsifier":"Compute the same order parameters on at least three fixed-aspect-ratio cylinders (for example $18\\times9$ and $24\\times12$ alongside the existing $6\\times3$ and $12\\times6$) and test whether the extrapolated $m_{U(1)}$ stays non-negative at the claimed Y-to-SL boundary. If the $1/L$ extrapolation still produces negative values, or if the spin-liquid region shrinks below the reported $J_2$ range, the central phase diagram would need revision.","tokens_in":24367,"feed_emoji":"🧲","tokens_out":12943,"duration_ms":119866,"temperature":0.7,"pith_summary":"The paper maps the ground-state phase diagram of the spin-1/2 $J_1$--$J_2$ XXZ model on the triangular lattice for the entire easy-axis regime, from the Heisenberg point to the Ising limit. Its central finding is that the spin liquid known at the Heisenberg point does not die out as easy-axis anisotropy grows: at $\\Delta=1.3$ it occupies a large interval of $J_2$ between the three-sublattice supersolid Y phase and the collinear stripe-z phase, and the liquid region persists to substantially larger $\\Delta$. The paper also quantifies the Y phase's supersolid order parameters and argues that its in-plane superfluid component is much smaller than previously reported, vanishing continuously toward the Ising limit, while the ferromagnetic moment is essentially zero. This matters because it tells experimentalists searching rare-earth and Ising-like triangular magnets what phases to expect and which order parameters to measure.","feed_headline":"Spin liquid survives strong easy-axis anisotropy","feed_subtitle":"DMRG maps the S=1/2 triangular-lattice J1-J2 model and finds a large liquid region before stripe order sets in.","key_machinery":"The argument is carried by three tools. DMRG scans and non-scans on open Y-cylinders directly measure local ordered moments and the structure factor $S(q)$, distinguishing the Y, spin-liquid, and stripe-z phases by the location and sharpness of peaks and by the decay of edge-induced order. Fixed-aspect-ratio $1/L$ extrapolations of the ordered moment and of $m_{uud}$, $m_{U(1)}$, and $m_F$, using $6\\times3$ and $12\\times6$ clusters with edges pinned to the classical Y state, convert finite-cylinder data into estimates of thermodynamic-limit order parameters. Minimally augmented spin-wave theory (MAGSWT), which stabilizes the otherwise unstable spin-wave spectrum of the Y state by adding a chemical potential, gives the quantum energy comparison between the Y, Y$'$, and stripe-z states and produces the linear Y-to-stripe boundary near the Ising limit, matching DMRG. The identity that explains the missing ferromagnetic moment is the rewriting of the $J_1$ Hamiltonian as a sum over triangles of $(S^\\perp_\\triangle)^2$ and $(S^z_\\triangle)^2$, where $S^\\alpha_\\triangle = S^\\alpha_A+S^\\alpha_B+S^\\alpha_C$: for $S=1/2$ the single-site term is constant, so a state with zero total spin on every triangle simultaneously minimizes every component.","core_discovery":"Within the easy-axis regime $\\Delta>1$, the spin-liquid state previously identified in the isotropic $J_1$--$J_2$ triangular model survives the addition of easy-axis anisotropy: for $\\Delta=1.3$, cylinders with $J_2$ between roughly 0.05 and 0.15 show a liquid with no magnetic Bragg peak, exponentially decaying edge-induced order with correlation length below $2a$, and no valence-bond or scalar-chiral order. The liquid's structure factor has broadened maxima at the $K$ points, making it a \"molten $120^\\circ$\" state, and its spin-spin correlations likely retain $SU(2)$ symmetry even though the Hamiltonian is easy-axis. On the ordered side, the Y phase has a finite solid (up-up-down) order parameter $m_{uud}$ but a small in-plane superfluid order parameter $m_{U(1)}$ that extrapolates to zero before the transition, and the ferromagnetic moment $m_F$ is essentially zero throughout. The paper explains the vanishing $m_F$ by rewriting the nearest-neighbor exchange as a sum over triangles of the squared total spin components, $\\frac{1}{2}\\sum_\\triangle[(S^\\perp_\\triangle)^2+\\Delta(S^z_\\triangle)^2]-(3/2)(\\Delta-1)\\sum_i(S^z_i)^2$, in which the last term is a constant for $S=1/2$ and the triangle sums $S^\\alpha_\\triangle=S^\\alpha_A+S^\\alpha_B+S^\\alpha_C$ vanish in the natural ground-state sector, killing the net ferromagnetic moment.","pith_inferences":["Inference: If the spin liquid extends to large $\\Delta$, then compounds such as $K_2Co(SeO_3)_2$ and $Na_2BaCo(PO_4)_2$ may have parameter regions currently labeled supersolid that are actually liquid; a test would be to search for broad excitation continua and the absence of magnetic Bragg peaks in those regions.","Inference: The triangle-local zero-total-spin mechanism for the missing ferromagnetic moment may be general: rewriting exchange as sums over simplex squared components should suppress local net moments in other $S=1/2$ frustrated lattices whenever the single-site term becomes a constant, which could predict which supersolid analogues are moment-free.","Inference: The apparently continuous vanishing of $m_{U(1)}$ near the Ising limit leaves open a thin pure up-up-down sliver between the supersolid and the stripe/SL phases; the paper itself says this layer cannot be ruled out, and targeted calculations at $J_2$ just below the transition for $\\Delta\\gtrsim5$ could settle it.","Inference: The abrupt appearance of the M-point peak in $S(q)$ at $\\Delta=1.3$ suggests the SL-to-stripe transition is first order; if so, finite-size energy level crossings on wider cylinders should show an avoided crossing or a sharp kink in the ground-state energy, a signature a future DMRG study could look for."],"forward_implications":["Materials described by the easy-axis $J_1$--$J_2$ triangular model can host a spin liquid even far from the Heisenberg point, so spin-liquid searches in rare-earth and transition-metal compounds should not be limited to isotropic or easy-plane parameters.","In the Y supersolid phase, the in-plane superfluid order $m_{U(1)}$ is much smaller than earlier numerical studies suggested and appears to vanish continuously as the Ising limit is approached, while the solid up-up-down order stays finite; experiments should see a nearly pure solid order near the Ising side.","The ferromagnetic moment of the $S=1/2$ Y phase should be zero, in sharp contrast to the classical value reaching $1/3$ in the Ising limit, so net magnetization measurements cannot be used to detect the Y phase.","The Y-to-stripe-z phase boundary is linear in $1/\\Delta$ near the Ising limit, not quadratic as classical minimization gives, because quantum fluctuations restore the linear energy term; this changes which materials are predicted to order into stripes.","The spin-liquid correlations retain approximate $SU(2)$ symmetry even for easy-axis anisotropy, meaning the liquid's short-distance spin correlations look isotropic despite the Hamiltonian's explicit anisotropy."],"supporting_citations":[{"why":"It establishes the spin-liquid phase of the isotropic $J_1$--$J_2$ model on the triangular lattice, the starting point whose resilience the paper tests under easy-axis anisotropy.","marker":"[37]"},{"why":"It supplies the easy-plane phase diagram and the numerical protocols, including the $\\langle S\\rangle<0.05$ spin-liquid criterion and the $1/L$ scaling, that the paper adapts to the easy-axis case.","marker":"[48]"},{"why":"It introduces the \"molten $120^\\circ$\" characterization of the triangular-lattice spin liquid and shows its resilience to anisotropy, the template for the liquid analysis here.","marker":"[53]"},{"why":"It demonstrates that quantum fluctuations in the easy-axis triangular model suppress the ferromagnetic moment and select the Y state, grounding the quasiclassical part of the argument.","marker":"[28]"},{"why":"It provides previous DMRG values for the supersolid order parameters in the $J_2=0$ limit, the comparison set that the paper's smaller $m_{U(1)}$ and $m_F$ results update.","marker":"[35]"},{"why":"It is the previous variational Monte Carlo study claiming finite $m_{U(1)}$ near the Ising limit, the result the paper's extrapolation contradicts.","marker":"[36]"},{"why":"It establishes the fixed-aspect-ratio $1/L$ extrapolation of ordered moments that the paper uses to locate the Y-to-spin-liquid boundary.","marker":"[60]"},{"why":"It supplies the chemical-potential stabilization of spin-wave spectra used by MAGSWT to compare the Y and stripe states.","marker":"[64]"}],"fun_headline_variants":["Spin liquid persists in easy-axis J1-J2 triangular model","Easy-axis anisotropy fails to kill spin liquid in J1-J2","Molten 120° liquid survives easy-axis anisotropy","DMRG: spin liquid lives in easy-axis J1-J2 despite anisotropy","Zero ferromagnet explained in easy-axis triangular model"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the linear extrapolation in $1/L_y$ from only the $6\\times3$ and $12\\times6$ fixed-aspect-ratio cylinders gives the true thermodynamic-limit values; near the transition the extrapolated in-plane superfluid order parameter turns negative, showing the premise is already under strain there.","fun_headline_variants_meta":{"raw":{"variants":["Spin liquid persists in easy-axis J1-J2 triangular model","Easy-axis anisotropy fails to kill spin liquid in J1-J2","Molten 120° liquid survives easy-axis anisotropy","DMRG: spin liquid lives in easy-axis J1-J2 despite anisotropy","Zero ferromagnet explained in easy-axis triangular model"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000605,"raw_usage":{"total_tokens":2834,"prompt_tokens":971,"completion_tokens":1863,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":587,"completion_tokens_details":{"reasoning_tokens":1776}},"tokens_in":587,"tokens_out":1863,"duration_ms":15255,"temperature":1.0,"reasoning_tokens":1776,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T22:13:59.245793+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the same order parameters on at least three fixed-aspect-ratio cylinders (for example $18\\times9$ and $24\\times12$ alongside the existing $6\\times3$ and $12\\times6$) and test whether the extrapolated $m_{U(1)}$ stays non-negative at the claimed Y-to-SL boundary. If the $1/L$ extrapolation still produces negative values, or if the spin-liquid region shrinks below the reported $J_2$ range, the central phase diagram would need revision.","supporting_citations":[{"cited_title":"Kleine, E","cited_arxiv_id":null,"evidence_quote":"It demonstrates that quantum fluctuations in the easy-axis triangular model suppress the ferromagnetic moment and select the Y state, grounding the quasiclassical part of the argument."},{"cited_title":"Heidarian and A","cited_arxiv_id":null,"evidence_quote":"It is the previous variational Monte Carlo study claiming finite $m_{U(1)}$ near the Ising limit, the result the paper's extrapolation contradicts."},{"cited_title":"Wenzel, T","cited_arxiv_id":null,"evidence_quote":"It supplies the chemical-potential stabilization of spin-wave spectra used by MAGSWT to compare the Y and stripe states."}],"review_version":1}