{"id":"98e17189-dfe7-4a88-a882-97cf079e3625","arxiv_id":"2508.18800","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For anisotropic elasticity -3/2<L<0, the Landau-de Gennes flow converges rigorously to a sharp interface moving by mean curvature with the director satisfying harmonic-map-type dynamics and strong anchoring.","lead":"A team of mathematicians proved that a standard model of liquid crystal phase transitions, with slightly anisotropic elasticity, converges to a sharp interface that moves by mean curvature with molecules anchored along the interface normal. The result rigorously confirms a 1971 prediction by de Gennes about the surface tension of these interfaces.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1 is not established as written: the existence of each order of the approximate solution requires solving the coupled parabolic system (4.10) for (d_{k+1}, Q^J_{k+1}), but Sections 4.2-4.3 assert solvability without a fixed-point argument, so the Q^K used in Theorem 1.3 is not rigorously…","rationale":"The paper develops a long and intricate spectral analysis, and much of the energy estimate appears internally coherent. However, the proof of the key existence theorem for approximate solutions is not complete: the solvability of the coupled free-boundary parabolic system (4.10) is asserted rather than demonstrated, and the same issue appears already for k=1 in Section 4.2.1 Step 4. The reader's stated weakest assumption was the external smoothness of the limit solution (Γ,n), which is a scope condition rather than an internal correctness risk. The reader did also note in the rationale that the d_1 existence step is sketched, and that is exactly the concern identified here. Because Q^K is an essential ingredient of Theorem 1.3, the main convergence theorem cannot be considered proven as written until the fixed-point construction is supplied. This is an addressable gap, so I would not say the result is false, but the manuscript as it stands leaves the central claim unverified. The recommended verdict is therefore UNVERDICTED rather than ACCEPT or the previous CONDITIONAL, which typically presumes the missing details are routine. The proposed concrete test isolates the minimal missing argument: a rigorous fixed-point proof for the k=1 system and the corresponding trace and regularity estimates. If the authors can provide that argument, the rest of the proof may well be sound, but that step is load-bearing and currently absent.","tokens_in":61401,"tokens_out":20709,"duration_ms":197288,"concrete_test":"Provide a complete fixed-point proof for the k=1 system. Define a bounded ball in X_T = L^∞(0,T;H²(Γ)) ∩ H^1(0,T;L²(Γ)) with d_1(0)=0. Show that solving (4.1)-(4.3) with boundary data determined by d_1 defines a map d_1 ↦ d_1^*, where d_1^* solves (4.5) with coefficients depending on the resulting q_{1,i}; prove the map sends the ball to itself and is a contraction or compact in L²(0,T;H¹(Γ)) for small T. In particular, verify the trace estimate ‖q_{1,i}|_Γ‖_{L²(0,T;H^{1/2}(Γ))} ≤ C(T,‖d_1‖) and the needed regularity of the nonlinear terms in (4.1). If this cannot be done with the estimates stated in Lemmas 3.7-3.9, the existence of Q^K is not established.","verdict_should_be":"UNVERDICTED","load_bearing_attack":"The central convergence result rests on Theorem 1.1, which supplies the approximate solution Q^K with residual O(ε^{K-1}). The construction in Section 4 reduces the problem at each order to a coupled system: a quasilinear parabolic system for the director components Q^J_k in Ω_+, with Dirichlet boundary data on the moving interface given by (3.65)/(3.81), coupled to a parabolic equation for d_k on Γ whose coefficients depend on Q^J_k. For k=1, Section 4.2.1 Step 4 states that the map P:d1↦(q_{1,1},q_{1,2}) is bounded from L²(0,T;H¹(S²)) to L²(0,T;H^{1/2}(S²)) and 'thus' a solution exists; no fixed-point theorem, compactness argument, or a priori estimates are given. For k≥2, Section 4.3.1 simply asserts 'We can obtain Q^J_{k+1} in Ω_+ and d_{k+1} on Γ by the system (4.10)', again without proof. Since Theorem 1.3 needs K=10 and Theorem 1.1 promises arbitrary K, this is load-bearing: if (4.10) is not solvable, Q^K does not exist and the nonlinear stability estimate has no base profile. This is an internal proof gap, not a dispute about mathematical consensus.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the sharp-interface limit, as ε→0, of the Landau–de Gennes gradient flow (1.3) for liquid crystals with anisotropic elasticity parameter L∈(-3/2,0). The claimed limit is the two-phase system (1.6): the interface moves by mean curvature, Q=0 in the isotropic phase, Q=s_+(nn-I/3) in the nematic phase, and the director n satisfies (2s_+^2∂_t n + h)×n=0 with the strong anchoring condition n=∇d on the interface. The authors state three main results: existence of approximate solutions to arbitrary order (Theorem 1.1), a uniform spectral lower bound for the linearized operator around the approximate solution (Theorem 1.2), and a nonlinear stability estimate with error O(ε^{2k}) for k=9 (Theorem 1.3). The proof strategy combines matched outer/inner asymptotic expansions with a div-curl decomposition and a reduction of the spectral estimate to scalar one-dimensional operators. The paper is technically ambitious and contains many detailed estimates, but the current version has a load-bearing gap in the construction of the approximate solutions, because the solvability of the coupled parabolic systems at each order is asserted rather than proved.","tokens_in":61793,"tokens_out":10076,"duration_ms":93671,"significance":"If the theorems are correct, the paper would be a substantial contribution: it would rigorously justify an anisotropic sharp-interface limit with strong anchoring, extending the isotropic results of Fei–Wang–Zhang–Zhang [21] and the matrix-valued Allen–Cahn analysis of Fei–Lin–Wang–Zhang [19], and it would verify a dynamical version of de Gennes's claim on isotropic–nematic interfacial tension. The technical machinery is impressive: matched asymptotic expansions to arbitrary order, spectral gap estimates for the operators G0 and G1, endpoint L∞ estimates, and a weighted nonlinear energy estimate. However, the existence of the approximate solution QK is not fully established because the coupled parabolic system (4.10) is not solved in Sections 4.2–4.3. Since Theorem 1.3 uses QK with K=10, this gap affects the central convergence claim. With a complete fixed-point argument, the paper would be a valuable and publishable contribution.","major_comments":[{"comment":"The construction of V1 is incomplete. The functions q1,1 and q1,2 are defined by solving the nonlinear parabolic system (4.1) with Dirichlet boundary conditions (4.2)–(4.3) that depend on d1, while d1 is then supposed to solve (4.5), whose right-hand side depends on q1,1 and q1,2. Step 4 only states that the map P:d1↦(q1,1,q1,2) is bounded from L2(0,T;H1(S2)) to L2(0,T;H^{1/2}(S2)) and then concludes 'Thus, there exists a solution d1'. Boundedness of P does not imply existence of a fixed point; the argument needs continuity, compactness or a priori estimates, and a fixed-point theorem. Without d1 and q1,i, the inner coefficients s1,i and hence Q1 are not constructed, so the base profile QK in Theorem 1.1 is not defined.","section":"Section 4.2.1, Step 4; Eqs. (4.1)–(4.5)"},{"comment":"For k≥1 the induction step is asserted rather than proved. After writing the coupled system (4.10), the text says 'We can obtain Q^J_{k+1} in Ω+ and d_{k+1} on Γ by the system (4.10)'. This system is genuinely coupled: d_{k+1} appears in the Dirichlet boundary condition for q_{k+1,1} and q_{k+1,2}, while the equation for d_{k+1} has a right-hand side depending on ∇d_{k+1} and on the q's. No fixed-point argument, function-space setting, or regularity theory is provided. Since Theorem 1.1 promises arbitrary K and Theorem 1.3 uses K=10, this missing induction step is load-bearing and prevents the current manuscript from establishing Theorem 1.1 as written.","section":"Section 4.3.1, Eq. (4.10)"},{"comment":"The spectral lower bound is a central ingredient, but its proof relies on Lemmas 6.6, 6.7, and 6.8 as black boxes imported from [19] and [21]. In particular, Lemma 6.7 contains an unspecified positive weight ω that 'decays exponentially to zero at +∞'; in the proof of Lemma 6.18 the weight is later chosen as ω = |d/ds κ(ξ_{1,ε}) s_{1,1}|^2 |ξ_{0,ε}|, but the paper does not verify that this choice is positive, bounded, and exponentially decaying on the interval I, nor that the orthogonality condition in Lemma 6.7 is satisfied. The authors should state the precise hypotheses of the imported lemmas and verify them explicitly for the weight actually used.","section":"Theorem 1.2 and Lemmas 6.6–6.8"}],"minor_comments":[{"comment":"The phrase 'there exits an approximate solution' should read 'there exists an approximate solution'; also the quantifier 'for any K ∈ Z+' should be 'for any K ∈ N' if Z+ denotes the positive integers.","section":"Section 1.3, Theorem 1.1"},{"comment":"Reference [10] contains the typo 'Allen-Chan', which should be 'Allen-Cahn', and reference [49] contains 'Sch :ordinger', which should be 'Schrödinger'.","section":"References [10] and [49]"},{"comment":"The symbol η is used for two different cut-off functions: η(z) in (3.34) and η(d0/δ) in (5.7). This is confusing and should be clarified, for example by using different symbols.","section":"Sections 3.3 and 5"},{"comment":"The statement that g1,0 is independent of d1 uses the identity ∫ s^2 s1 dz = 0; this identity should be displayed explicitly, since it is used again in Remark 3.4 and in the proof of Lemma 3.9.","section":"Remark 3.3"},{"comment":"The flowchart in Figure 1 is difficult to read; the authors should enlarge the text and include explicit equation numbers in each box so that the induction procedure can be followed.","section":"Figure 1"},{"comment":"The hypothesis that (Γ,n) is a smooth solution of the limit system (1.6) on [0,T] should be emphasized as a conditional assumption; the result does not address finite-time singularities of the limit flow, which can form under mean curvature motion.","section":"Theorem 1.1 and Section 1.3"},{"comment":"The gluing formula (5.6) uses the cut-off η(d0/δ), while the inner expansion QK_in is defined with dK; the justification that this mismatch yields the claimed O(ε^{K-1}) residual after (5.4)–(5.5) would be clearer if written out explicitly.","section":"Section 5, Eq. (5.6)"}],"recommendation":"major_revision","confidential_remarks":"I recommend major revision rather than rejection. The overall strategy is plausible and the paper contains substantial new technical work, but the missing solvability proof for the coupled parabolic systems in Section 4 is a genuine gap in the proof of Theorem 1.1, and Theorem 1.3 depends on that theorem. The gap may be fixable by adding a fixed-point argument with appropriate function spaces, so the work is within the scope of the journal. The authors should also clarify the import of Lemmas 6.6–6.8 and verify the hypotheses of the weight in Lemma 6.18."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the quick take: this is a serious step forward for the sharp interface program in liquid crystals, extending the rigorous L=0 result to anisotropic -3/2<L<0 with strong anchoring n=∇d on the interface. The main theorems are the right ones: arbitrary-order approximate solutions, a spectral lower bound, and a nonlinear stability estimate. The div-curl decomposition and reduction to scalar ODE operators are genuinely new and well motivated. If the proof holds, this closes the gap left by Fei-Wang-Zhang-Zhang's formal derivation, which matters.\n\nBut the stress-test note lands. Theorem 1.1 is the load-bearing piece, and the existence of the higher-order corrections is not actually proved. Section 4.2.1 Step 4 says the map P from d1 to (q_{1,1},q_{1,2}) is bounded and 'thus' a solution exists; that is not a fixed-point argument. Section 4.3.1 simply asserts that (4.10) can be solved. This is the coupled parabolic system for (d_{k+1}, Q^J_{k+1}) with Dirichlet boundary data tied to the inner expansion—exactly the kind of quasilinear system that needs a Schauder or Leray-Schauder degree argument, or at least a stated iteration scheme with a priori estimates. Since Theorem 1.3 needs K=10, the gap is not cosmetic. I suspect it's fixable, because the structure is a standard parabolic system with tame nonlinearities and the estimates are already there. But as written, Theorem 1.1 is not established.\n\nTwo smaller points. The abstract says the paper rigorously verifies de Gennes's claim about surface tension strength. I don't see σ computed from the heteroclinic; the paper derives V=σκ but never connects to de Gennes's explicit sqrt(1+2L/3) formula. That's an overstatement, easily fixed. Also the paper leans heavily on [19] and [21] for spectral lemmas from overlapping authors; that's normal in this line and the results are published, so I don't treat it as a flaw.\n\nBottom line: this deserves a serious referee, not a desk reject. The referee should push on Section 4 and require a complete existence proof for the coupled system (4.10) before acceptance. I'd want to see that fixed before citing it myself.","headline":"Rigorous-looking anisotropic sharp interface limit, but the approximate-solution construction has a real existence gap; deserves refereeing.","tokens_in":62264,"tokens_out":2602,"would_cite":false,"duration_ms":25633,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35B25","35K55","35Q35","82D30"],"pacs":[],"model":"deepseek-v4-flash","headline":"For anisotropic elasticity $L\\in(-3/2,0)$, the Landau-de Gennes flow has a sharp-interface limit: mean-curvature motion with strong anchoring of the director field.","keywords":["Landau-de Gennes","isotropic-nematic phase transition","sharp interface limit","matched asymptotic expansions","anisotropic elasticity","mean curvature flow","strong anchoring","spectral gap estimate"],"falsifier":"Compute numerically the lowest eigenvalues of the scalar operators $G_0$ and $G_1$ defined around (6.48) for a fixed $L\\in(-3/2,0)$ on a sequence of shrinking intervals $\\varepsilon\\to0$; if the second eigenvalue of either operator is not bounded below by a positive constant independent of $\\varepsilon$, the spectral lower bound (1.9) fails and the error estimate collapses.","tokens_in":61141,"feed_emoji":"","tokens_out":7480,"duration_ms":76566,"temperature":0.7,"pith_summary":"The paper proves that the diffuse Landau-de Gennes description of an isotropic-nematic interface has a sharp-interface limit when the elastic anisotropy parameter $L$ lies in $(-3/2,0)$. In the limit, the interface moves by mean curvature, the order parameter is zero in the isotropic phase and uniaxial in the nematic phase, and the director field satisfies the Oseen-Frank gradient-flow equation with strong anchoring at the interface. The proof constructs approximate solutions to arbitrarily high order and proves a uniform spectral lower bound that turns the formal expansion into a quantitative convergence statement: an initial error of order $\\varepsilon^{18}$ (in the paper's weighted energy) remains of that order up to time $T$. This establishes, in a dynamical setting, a surface-tension claim about isotropic-nematic interfaces that goes back to 1971.","feed_headline":"Sharp-interface limit proved for anisotropic nematic fronts","feed_subtitle":"Landau-de Gennes flow converges to mean-curvature motion with strong director anchoring.","key_machinery":"The argument is carried by a matched asymptotic expansion whose inner profile is the heteroclinic solution $Q_0(z)=s(z)(nn-\\frac13 I)$ with $s(z)=\\frac12(1+\\tanh(\\gamma z/2))$ and $\\gamma=(1+2L/3)^{-1/2}$. Around this profile, the proof establishes a uniform spectral lower bound for the linearized operator $H_{Q^K}$: for every traceless symmetric $Q\\in H^1$, $\\int |\\nabla Q|^2 - L\\int |\\nabla\\cdot Q|^2 + \\varepsilon^{-2}\\int H_{Q^K}Q:Q \\le C\\int |Q|^2$, with $C$ independent of $\\varepsilon$. The decisive step is a div-curl decomposition $|\\nabla Q|^2 = \\frac32|\\nabla\\cdot Q|^2 + \\frac14|T(Q)|^2 + \\cdots$, followed by a basis decomposition and a change of coordinates that reduce the tensor spectral problem to two scalar one-dimensional operators $G_0,G_1$ plus singular product terms; coercivity and spectral gap estimates for those scalar operators close the energy estimate.","core_discovery":"The central claim is Theorem 1.3: for any smooth solution $(\\Gamma,n)$ of the limit system (1.6) on $[0,T]$, a solution $Q^\\varepsilon$ of the Landau-de Gennes flow that starts within $\\varepsilon^{18}$ of the matched asymptotic solution $Q^K$ stays within that distance for all later times. Consequently, as $\\varepsilon\\to0$, the tensor field converges to the sharp-interface system: $Q=0$ in the isotropic region, $Q=s_+(nn-\\frac13 I)$ with the director $n$ obeying $(2s_+^2\\partial_t n+h)\\times n=0$, and $n=\\nabla d$ on the interface, with $d$ the signed distance function and the interface evolving by $V=\\sigma\\kappa$. The result upgrades a formal derivation for anisotropic elasticity into a rigorous stability theorem and replaces the Neumann-type boundary condition of the isotropic case by strong anchoring.","pith_inferences":["The method should carry over to other elliptic differential operators with a scalar heteroclinic profile whenever the reduced one-dimensional operators have a uniform spectral gap; the proof's real content is that gap, not the specific liquid-crystal structure.","One can test the sharpness of the $k=9$ power by tracking the constants: the argument likely yields a similar statement for any $k\\ge9$ order term, so the limitation is technical rather than structural.","Near $L\\to -3/2$, the profile width $\\gamma^{-1}$ diverges, so the interface layer is no longer thin; a separate scaling would be needed, suggesting that convergence may fail or require a rescaled limit at that endpoint."],"forward_implications":["For $L\\in(-3/2,0)$ and well-prepared data, solutions of the Landau-de Gennes flow (1.3) converge to the sharp-interface system (1.6) on the whole smooth-existence interval $[0,T]$.","The convergence is quantitative: the weighted energy $E(Q^\\varepsilon-Q^K)$ stays at order $\\varepsilon^{18}$, so the interface profile is captured at that precision.","The interface speed is $V=\\sigma\\kappa$ with $\\sigma$ determined by the elastic constants, so the surface tension predicted for anisotropic elasticity is realized dynamically, not only statically.","In the nematic bulk the limit director obeys $(2s_+^2\\partial_t n+h)\\times n=0$, the Oseen-Frank analogue of harmonic map heat flow, with the strong anchoring boundary condition $n=\\nabla d$ replacing Neumann conditions."],"supporting_citations":[{"why":"Supplies the quasi-minimal connecting orbit construction that produces approximate solutions to arbitrary order.","marker":"[19]"},{"why":"Gives the sharp-interface limit for the isotropic case $L=0$, the baseline result the present paper extends to anisotropic elasticity.","marker":"[21]"},{"why":"Provides the formal matched-asymptotic derivation of the strong-anchoring sharp interface model for $L<0$.","marker":"[20]"},{"why":"Establishes stability of uniaxial solutions for $L<0$, justifying the profile ansatz used in the expansion.","marker":"[43]"},{"why":"States the surface-tension conjecture for isotropic-nematic interfaces that the dynamical convergence result verifies.","marker":"[11]"}],"fun_headline_variants":["Rigorous proof of de Gennes surface tension claim","Anisotropic nematic interface limit: exact asymptotics","Landau-de Gennes flow converges to sharp interface","Director anchoring validated in nematic phase transition","Rigorous asymptotics for anisotropic nematic fronts"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The theorem assumes a smooth solution of the limit system exists on the whole time interval $[0,T]$; if the interface develops a curvature singularity before $T$, the expansion and the convergence proof stop at the first singularity.","fun_headline_variants_meta":{"raw":{"variants":["Rigorous proof of de Gennes surface tension claim","Anisotropic nematic interface limit: exact asymptotics","Landau-de Gennes flow converges to sharp interface","Director anchoring validated in nematic phase transition","Rigorous asymptotics for anisotropic nematic fronts"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000809,"raw_usage":{"total_tokens":3632,"prompt_tokens":1107,"completion_tokens":2525,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":723,"completion_tokens_details":{"reasoning_tokens":2449}},"tokens_in":723,"tokens_out":2525,"duration_ms":18978,"temperature":1.0,"reasoning_tokens":2449,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:54:45.566866+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute numerically the lowest eigenvalues of the scalar operators $G_0$ and $G_1$ defined around (6.48) for a fixed $L\\in(-3/2,0)$ on a sequence of shrinking intervals $\\varepsilon\\to0$; if the second eigenvalue of either operator is not bounded below by a positive constant independent of $\\varepsilon$, the spectral lower bound (1.9) fails and the error estimate collapses.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes stability of uniaxial solutions for $L<0$, justifying the profile ansatz used in the expansion."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"States the surface-tension conjecture for isotropic-nematic interfaces that the dynamical convergence result verifies."}],"review_version":2}