betaFeedback

WorksMathematics

A Gap Theorem for the Möbius Cross Energy Near the Hopf Link

Abstract

We prove that the critical value 2π22π^2 is isolated for the Möbius cross energy of two-component links in the round three dimensional sphere S3\mathbb{S}^{3}. Specifically, there exists ε0>0\varepsilon_0>0 such that any non-split, regular H2H^2 pair of curves with disjoint images, having vanishing first variation and energy at most 2π2+ε02π^2+\varepsilon_0, must lie in the Möbius-reparametrization orbit of the standard Hopf link.

Introduction and main theorem

Set S1=R/(2πZ)\mathbb{S}^{1} = \mathbb{R}/(2\pi\mathbb{Z}), and equip S3={z∈R4:∣z∣=1}\mathbb{S}^{3} = \{z \in\mathbb{R}^{4} : |z| = 1\} with the round metric g0g_{0}. For a pair Γ=(γ1,γ2)\Gamma= (\gamma_{1}, \gamma_{2}) of rectifiable closed curves with disjoint component images, the Möbius cross energy is

E(Γ)=∫γ1∫γ2dℓ1(x) dℓ2(y)∣x−y∣2,(1)E(\Gamma) = \int_{\gamma_{1}} \int_{\gamma_{2}} \frac{\mathrm{d}\ell_{1}(x)\,\mathrm{d}\ell_{2}(y)}{|x-y|^{2}}, \tag*{(1)}

where ∣x−y∣|x-y| is the Euclidean chordal distance in R4\mathbb{R}^{4}. Integration along each parametrized curve counts multiplicity. O’Hara [6] introduced the related knot energy. Freedman, He, and Wang [2] studied the cross interaction (1), which is invariant under Möbius transformations.

A pair of disjoint compact connected sets is called non-split if no smooth embedded two dimensional sphere disjoint from them separates the two sets. Hopf links are non-split, and the standard Hopf link is

H=(x,y),x(s)=(cos⁡s,sin⁡s,0,0),y(t)=(0,0,cos⁡t,sin⁡t).(2)H = (x,y), \qquad x(s) = (\cos s,\sin s,0,0), \qquad y(t) = (0,0,\cos t,\sin t). \tag*{(2)}

Its components are unit-speed great circles in orthogonal 2-planes. Since ∣x(s)−y(t)∣2=2|x(s)-y(t)|^{2} = 2,

E(H)=2π2.E(H) = 2\pi^{2}.

Freedman, He, and Wang [2] conjectured that this is the minimum cross energy among all non-split two-component links. An important step was taken by He [4], who proved that a global minimizer exists and that every such minimizer is ambiently isotopic to the Hopf link. Agol, Marques, and Neves subsequently proved the sharp lower bound and classified all equality configurations [1].

Theorem 1.1 (Agol–Marques–Neves). Every non-split two-component rectifiable link in S3\mathbb{S}^{3} satisfies

E(Γ)≥2π2.E(\Gamma) \ge2\pi^{2}.

Equality holds precisely when a conformal transformation carries the component images to the standard Hopf circles, each counted with multiplicity one.

2020 Mathematics Subject Classification. Primary 53A31, 58E05; Secondary 57K10, 49Q10,

Key words and phrases. Möbius cross energy, Hopf link, critical-value gap. For a pair of curves Γ^=(γ^1,γ^2)\widehat{\Gamma}=(\widehat{\gamma}_1,\widehat{\gamma}_2) in R3\mathbb{R}^3, the Gauss linking number is defined by

lk⁡(Γ^)=14π∫S1∫S1det⁡(γ^1′(s),γ^2′(t),γ^1(s)−γ^2(t))∣γ^1(s)−γ^2(t)∣3 ds dt.\operatorname{lk}(\widehat{\Gamma})=\frac{1}{4\pi}\int_{\mathbb{S}^1}\int_{\mathbb{S}^1}\frac{\det(\widehat{\gamma}_1'(s),\widehat{\gamma}_2'(t),\widehat{\gamma}_1(s)-\widehat{\gamma}_2(t))}{\lvert\widehat{\gamma}_1(s)-\widehat{\gamma}_2(t)\rvert^3}\,ds\,dt.

It is conformal invariant, therefore the linking number of a pair of curves Γ\Gamma in S3\mathbb{S}^3 is defined to be that of its stereographic projection in R3\mathbb{R}^3. A nonzero linking number implies non-splitness, but the converse fails, as illustrated by the Whitehead link. A direct calculus implies

E(Γ)≥4π∣lk⁡(Γ)∣.E(\Gamma)\geq4\pi\lvert\operatorname{lk}(\Gamma)\rvert.

Allowing a choice of component orientations, He [4] shows that a pair of non-split curves is link-homotopic to the Hopf link provided E(Γ)<8πE(\Gamma)<8\pi.

As a continuation of the study of He, Agol, Marques, and Neves [4, 1], and motivated by Mondino and Nguyen [7], we study critical points rather than minimizers of the cross energy. The equality case in Theorem 1.1 characterizes the minimizers, but does not rule out non-minimizing critical pairs with energy arbitrarily close to 2π22\pi^2. Our main result shows that such critical pairs do not exist.

To state the result precisely, we fix the following notation.

Definition 1.2. A closed curve γ\gamma is regular if ∣γ′∣>0\lvert\gamma'\rvert>0 everywhere, we fix the extrinsic Sobolev space

H2(S1,S3)={γ∈H2(S1,R4):∣γ∣=1 a.e.}.H^2(\mathbb{S}^1,\mathbb{S}^3)=\{\gamma\in H^2(\mathbb{S}^1,\mathbb{R}^4):\lvert\gamma\rvert=1\ \text{a.e.}\}.

(i) The space of pairs we study is

A={(γ1,γ2): γ1,γ2∈H2(S1,S3) are regular curves s.t. γ1(S1)∩γ2(S1)=∅}.\mathcal{A}=\{(\gamma_1,\gamma_2):\ \gamma_1,\gamma_2\in H^2(\mathbb{S}^1,\mathbb{S}^3)\ \text{are regular curves s.t.}\ \gamma_1(\mathbb{S}^1)\cap\gamma_2(\mathbb{S}^1)=\varnothing\}.

(ii) A pair is weakly critical if the first variation δEΓ=0\delta E_\Gamma=0 on every H2H^2 tangent variation.

(iii) Let Diff⁡2(S1)\operatorname{Diff}^2(\mathbb{S}^1) denote the H2H^2 circle maps that are C1C^1 diffeomorphisms, allowing both orientations, and set

G=Mo¨b⁡(S3)×Diff⁡2(S1)×Diff⁡2(S1).(3)G=\operatorname{Möb}(\mathbb{S}^3)\times\operatorname{Diff}^2(\mathbb{S}^1)\times\operatorname{Diff}^2(\mathbb{S}^1). \tag*{(3)}

We write G⋅HG\cdot H for the set of pairs obtained from HH by a common orientation-preserving Möbius transformation and separate reparametrizations.

We can now state our main gap theorem.

Theorem 1.3. There exists ε0>0\varepsilon_0>0 such that every non-split, weakly critical pair Γ∈A\Gamma\in\mathcal{A} satisfying

E(Γ)≤2π2+ε0E(\Gamma)\leq2\pi^2+\varepsilon_0

belongs to G⋅HG\cdot H. In particular, E(Γ)=2π2E(\Gamma)=2\pi^2.

Remark 1.4. For fixed γ2\gamma_2, set

ρ2(x)=∫γ2dℓ2(y)∣x−y∣2.\rho_2(x)=\int_{\gamma_2}\frac{d\ell_2(y)}{\lvert x-y\rvert^2}.

Then E(γ1,γ2)E(\gamma_1,\gamma_2) is the length of γ1\gamma_1 in the conformal metric ρ22g0\rho_2^2g_0 on S3∖γ2\mathbb{S}^3\setminus\gamma_2; see [2]. Thus a critical pair consists of unparametrized closed geodesics in the metrics generated by the opposite components. We derive the corresponding equation in § 2.

Exact criticality is essential here, since non-Möbius perturbations of the Hopf link can have energies arbitrarily close to 2π22\pi^2. Although we work with regular H2H^2 pairs, the critical equation implies smoothness after constant-speed reparametrization.

Outline of the paper. § 2 derives the first variation and the Euler–Lagrange equation for weakly critical pairs. § 3 establishes a Möbius normalization, proves regularity of weakly critical pairs, and develops the compactness theory near the minimum energy. In § 4, we compute the second variation at the Hopf link, identify its kernel, and establish transverse coercivity. § 5 develops the nonlinear estimates and proves the renormalization lemma. Finally, in § 6, we combine the transverse analysis of § 4 with the renormalization argument of § 5 to prove the local gap theorem, and then use the compactness result of § 3 to complete the proof of the main theorem.

Acknowledgments. The authors grateful to Professor Jian Ge for useful discussions. They also thank Xingwang Xu, Xingdong Tang and Jie Zhou for their interest in this problem. The first author is supported by the China Postdoctoral Science Foundation (Grant No. 2026M793350). The second author is supported by the Open Project Program of the Key Laboratory of Mathematics and Complex System at Beijing Normal University (Grant No. 202501). The principal arguments and the core of the paper were developed by the second author while visiting SIMIS.

AI assistance declaration: The main ideas and most proofs are solely the work of the authors. ChatGPT (version 5.6sol) was used only to assist in proving Proposition 3.1 and to aid in manuscript preparation. The authors have verified all contributions and assume complete accountability for the paper.

First variation and the Euler–Lagrange equation

In this section, we establish the basic preliminaries for the Möbius cross energy and derive its first variation and Euler–Lagrange equation.

We denote DD the Euclidean connection on R4\mathbb{R}^{4} and ∇\nabla the Levi-Civita connection of g0g_{0} on S3\mathbb{S}^{3}. Let X,YX,Y be tangent vector fields on S3\mathbb{S}^{3}, then for each x∈S3x \in\mathbb{S}^{3}, the Gauss formula gives

DXY=∇XY+II(X,Y),II(X,Y)=−⟨X,Y⟩g0x,(4)D_XY=\nabla_XY+II(X,Y),\qquad II(X,Y)=-\langle X,Y\rangle_{g_0}x, \tag*{(4)}

where IIII is the second fundamental form of S3⊂R4\mathbb{S}^{3}\subset\mathbb{R}^{4}.

Thus, for a unit-speed curve, γ′′=κ−γ\gamma''=\kappa-\gamma, where κ=∇γ′γ′\kappa=\nabla_{\gamma'}\gamma' is its curvature vector in S3\mathbb{S}^{3}.

For a variation Γτ\Gamma_\tau with initial velocity ξ\xi, we write δEΓ[ξ]=ddτE(Γτ)∣τ=0\delta E_{\Gamma}[\xi]=\left.\frac{d}{d\tau}E(\Gamma_\tau)\right|_{\tau=0} and δ2\delta^{2} for the second differential in a fixed chart. Ordinary coordinate derivatives are denoted by ∂\partial or dd.

The parameter torus is T2=Ss1×St1\mathbb{T}^{2}=\mathbb{S}^{1}_{s}\times\mathbb{S}^{1}_{t}. Throughout the paper,

∬F(s,t) ds dt:=∫02π∫02πF(s,t) ds dt\iint F(s,t)\,ds\,dt:=\int_{0}^{2\pi}\int_{0}^{2\pi}F(s,t)\,ds\,dt

always refers to this fixed torus with the product parameter measure.

Lemma 2.1. Every regular closed H2H^{2} curve admits a constant-speed parametrization obtained by an orientation-preserving H2H^{2} circle diffeomorphism. Separate such reparametrizations preserve the cross energy, non-splittness, and weak criticality.

Proof. For a lift of the parameter to [0,2π][0,2\pi], set

θ(s)=2πL(γ)∫0s∣γ′(r)∣ dr,s∈[0,2π].\theta(s)=\frac{2\pi}{L(\gamma)}\int_{0}^{s}|\gamma'(r)|\,dr,\qquad s\in[0,2\pi].

Since γ′∈H1↪C0\gamma'\in H^{1}\hookrightarrow C^{0} and γ\gamma is a regular closed curve, composition with the Euclidean norm gives ∣γ′∣∈H1|\gamma'|\in H^{1}. Consequently θ∈H2\theta\in H^{2}, and θ′=2π∣γ′∣/L(γ)\theta'=2\pi|\gamma'|/L(\gamma) is continuous and bounded below by a positive constant. The identity θ(s+2π)=θ(s)+2π\theta(s+2\pi)=\theta(s)+2\pi makes θ\theta a degree-one C1C^1 circle diffeomorphism. Denote its inverse by ψ=θ−1\psi=\theta^{-1}, change of variables gives ψ∈H2\psi\in H^2. Thus we can define γˉ:=γ∘ψ\bar{\gamma}:=\gamma\circ\psi satisfies γˉ∈H2\bar{\gamma}\in H^2 and ∣γˉ′∣=L(γ)/(2π)|\bar{\gamma}'|=L(\gamma)/(2\pi).

Changing variables separately in each of the two integrals proves the energy invariance, since the curve images remain unchanged, the non-split property is also preserved. Composition by each fixed H2H^2 diffeomorphism preserves H2H^2 tangent fields. Pulling back a variation and differentiating energy invariance proves preservation of weak criticality in both directions. A fixed smooth Möbius transformation preserves weak criticality by the same argument. □

Next, we derive the first variation in the spherical model. Let Γ=(γ1,γ2)∈A\Gamma=(\gamma_1,\gamma_2)\in\mathcal{A} be a regular H2H^2 pair with disjoint component images in S3S^3. Write Ti=∂ℓiγiT_i=\partial_{\ell_i}\gamma_i and κi=∇TiTi\kappa_i=\nabla_{T_i}T_i. The projection onto the curve-normal plane inside TγiS3T_{\gamma_i}S^3 is

PNi=Id⁡−γi⊗γi−Ti⊗Ti.(5)P_{N_i}=\operatorname{Id}-\gamma_i\otimes\gamma_i-T_i\otimes T_i. \tag*{(5)}

Define the positive interaction potentials by

ρ1(y)=∫γ1dℓ1(x)∣x−y∣2,ρ2(x)=∫γ2dℓ2(y)∣x−y∣2.(6)\rho_1(y)=\int_{\gamma_1}\frac{d\ell_1(x)}{|x-y|^2},\qquad\rho_2(x)=\int_{\gamma_2}\frac{d\ell_2(y)}{|x-y|^2}. \tag*{(6)}

For r=∣x−y∣r=|x-y| and K(x,y)=r−2K(x,y)=r^{-2}, spherical differentiation yields

grad⁡xK=2(y−⟨x,y⟩x)r4,grad⁡yK=2(x−⟨x,y⟩y)r4.(7)\operatorname{grad}_x K=\frac{2(y-\langle x,y\rangle x)}{r^4},\qquad\operatorname{grad}_y K=\frac{2(x-\langle x,y\rangle y)}{r^4}. \tag*{(7)}

In particular, PN1grad⁡xK=2PN1y/r4P_{N_1}\operatorname{grad}_x K=2P_{N_1}y/r^4, and similarly for the second component.

Proposition 2.2. For every Γ∈A\Gamma\in\mathcal{A}, define the L2L^2 normal fields

F1=PN1grad⁡ρ2−ρ2κ1,F2=PN2grad⁡ρ1−ρ1κ2.F_1=P_{N_1}\operatorname{grad}\rho_2-\rho_2\kappa_1,\qquad F_2=P_{N_2}\operatorname{grad}\rho_1-\rho_1\kappa_2.

For every H1H^1 sphere-tangent field ξ=(ξ1,ξ2)\xi=(\xi_1,\xi_2), the extended first variation is

δEΓ[ξ]=∫γ1⟨F1,ξ1⟩ dℓ1+∫γ2⟨F2,ξ2⟩ dℓ2.(8)\delta E_\Gamma[\xi]=\int_{\gamma_1}\langle F_1,\xi_1\rangle\,d\ell_1+\int_{\gamma_2}\langle F_2,\xi_2\rangle\,d\ell_2. \tag*{(8)}

Consequently, weak criticality is equivalent to the Euler–Lagrange equations

κ1=PN1grad⁡log⁡ρ2,κ2=PN2grad⁡log⁡ρ1(9)\kappa_1=P_{N_1}\operatorname{grad}\log\rho_2,\qquad\kappa_2=P_{N_2}\operatorname{grad}\log\rho_1 \tag*{(9)}

almost everywhere.

Proof. We write

E(Γ)=∫γiρj dℓi,i,j=1,2,i≠j.E(\Gamma)=\int_{\gamma_i}\rho_j\,d\ell_i,\qquad i,j=1,2,\quad i\ne j.

For an H2H^2 sphere-tangent variation field ξi\xi_i, differentiation and integration by parts along the closed curve give

δEΓ[ξi]=∫γi(⟨grad⁡ρj,ξi⟩+ρj⟨Ti,∇Tiξi⟩) dℓi=∫γi(⟨grad⁡ρj−∇Ti(ρjTi),ξi⟩) dℓi=∫γi⟨Fi,ξi⟩ dℓi,\begin{aligned} \delta E_\Gamma[\xi_i] &=\int_{\gamma_i}\left(\langle\operatorname{grad}\rho_j,\xi_i\rangle+\rho_j\langle T_i,\nabla_{T_i}\xi_i\rangle\right)\,d\ell_i\\ &=\int_{\gamma_i}\left(\left\langle\operatorname{grad}\rho_j-\nabla_{T_i}(\rho_jT_i),\xi_i\right\rangle\right)\,d\ell_i\\ &=\int_{\gamma_i}\langle F_i,\xi_i\rangle\,d\ell_i, \end{aligned}

@end_mid=1

where the last equality uses

∇Ti(ρjTi)=⟨grad⁡ρj,Ti⟩Ti+ρjκi.\nabla_{T_i}(\rho_j T_i)=\langle\operatorname{grad}\rho_j,T_i\rangle T_i+\rho_j\kappa_i.

In arclength coordinates, Ti∈H1T_i\in H^1 and κi∈L2\kappa_i\in L^2, so the integration by parts is valid in the Sobolev sense. Since Fi∈L2F_i\in L^2, the right-hand side extends continuously to H1H^1 sphere-tangent test fields. Summing the two partial variations proves (8).

Given an arbitrary smooth periodic R4\mathbb{R}^4-valued field φ\varphi, take an admissible sphere-tangent field

ξi=(Id⁡−γi⊗γi)φ∈H2.\xi_i=(\operatorname{Id}-\gamma_i\otimes\gamma_i)\varphi\in H^2.

Since Fi⊥γiF_i\perp\gamma_i, weak criticality gives ∫⟨Fi,φ⟩ dℓi=0\int\langle F_i,\varphi\rangle\,d\ell_i=0. The speed is bounded away from zero implies Fi=0F_i=0 almost everywhere. Conversely, Fi=0F_i=0 makes every first variation vanish. Dividing by the positive potentials proves (9).

Remark 2.3. He [4 Lemma 3.1, equations (3-2)–(3-3)] derived the first variation for smooth links and smooth variation fields. We repeat the computation here in order to justify the formula for regular H2H^2 links and H1H^1 variation fields, which obtain the weak Euler–Lagrange equation used below.

Global normalization and compactness of critical pairs

In this section, we establish a global Möbius gauge for non-split pairs in A\mathcal{A}. Combined with the critical equation, this yields regularity and uniform derivative estimates for weakly critical pairs. We then prove a compactness theorem for critical sequences approaching the Hopf energy, which will be used in the proof of the gap theorem.

We first give a global Möbius gauge. Let N=(0,0,0,1)N=(0,0,0,1) and let σ:S3∖{N}→R3\sigma:\mathbb{S}^3\setminus\{N\}\to\mathbb{R}^3 be stereographic projection. Its inverse satisfies

σ−1(z)=(2z1+∣z∣2,∣z∣2−11+∣z∣2),∣σ−1(z)−σ−1(w)∣=2∣z−w∣(1+∣z∣2)(1+∣w∣2).(10)\sigma^{-1}(z)=\left(\frac{2z}{1+|z|^2},\frac{|z|^2-1}{1+|z|^2}\right),\qquad \left|\sigma^{-1}(z)-\sigma^{-1}(w)\right|=\frac{2|z-w|}{\sqrt{(1+|z|^2)(1+|w|^2)}}. \tag*{(10)}

Proposition 3.1. Fix 0<Λ<∞0<\Lambda<\infty. Every non-split pair Γ=(γ1,γ2)∈A\Gamma=(\gamma_1,\gamma_2)\in\mathcal{A} satisfying E(Γ)≤ΛE(\Gamma)\le\Lambda has an orientation-preserving Möbius image for which

2π3≤L(γi)≤6Λπ,i=1,2,(11)\frac{2\pi}{3}\le L(\gamma_i)\le\frac{6\Lambda}{\pi},\qquad i=1,2, \tag*{(11)}

and

dist⁡R4(γ1,γ2)≥dΛ>0.(12)\operatorname{dist}_{\mathbb{R}^4}(\gamma_1,\gamma_2)\ge d_\Lambda>0. \tag*{(12)}

The constant dΛd_\Lambda depends only on Λ\Lambda.

Proof. Send a point of γ1\gamma_1 to the north pole NN, and then apply a translation and dilation in the stereographic chart σ\sigma so that

N∈γ1,0∈σ(γ2),diam⁡R3σ(γ2)=1.N\in\gamma_1,\qquad0\in\sigma(\gamma_2),\qquad\operatorname{diam}_{\mathbb{R}^3}\sigma(\gamma_2)=1.

Thus σ(γ2)⊂B‾1(0)\sigma(\gamma_2)\subset\overline{B}_1(0). Non-splitness forces γ1\gamma_1 to meet σ−1(B‾1(0))\sigma^{-1}(\overline{B}_1(0)); otherwise the inverse image of a slightly larger round sphere would separate the components. We also use the elementary estimate

L(η)≥2diam⁡S3η(S1)(13)L(\eta)\ge2\operatorname{diam}_{\mathbb{S}^3}\eta(\mathbb{S}^1) \tag*{(13)}

for every closed rectifiable curve η:S1→S3\eta:\mathbb{S}^{1}\to\mathbb{S}^{3} by considering the two arcs joining a diameter-realizing pair. The first component contains NN and a point q∈γ1(S1)q\in\gamma_{1}(\mathbb{S}^{1}) with ∣σ(q)∣≤1|\sigma(q)|\le1, so its spherical diameter is at least π/2\pi/2. Applying (13) gives

L(γ1)≥π.(14)L(\gamma_{1})\ge\pi. \tag*{(14)}

The compact set σ(γ2(S1))\sigma(\gamma_{2}(\mathbb{S}^{1})) realizes its diameter, so there exist z,wz,w in this set such that

∣z−w∣=1.|z-w|=1.

Note that ∣z∣,∣w∣≤1|z|,|w|\le1. If dd denotes their spherical distance, then formula (10) gives

2sin⁡d2=∣σ−1(z)−σ−1(w)∣=2∣z−w∣1+∣z∣21+∣w∣2≥1,2\sin\frac{d}{2}=|\sigma^{-1}(z)-\sigma^{-1}(w)|=\frac{2|z-w|}{\sqrt{1+|z|^{2}}\sqrt{1+|w|^{2}}}\ge1,

and hence d≥π/3d\ge\pi/3. Another application of (13) yields

L(γ2)≥2π3.(15)L(\gamma_{2})\ge\frac{2\pi}{3}. \tag*{(15)}

Since ∣x−y∣2≤4|x-y|^{2}\le4 for x,y∈S3x,y\in\mathbb{S}^{3},

E(Γ)=∫γ1∫γ2dℓ1dℓ2∣x−y∣2≥14L(γ1)L(γ2).(16)E(\Gamma)=\int_{\gamma_{1}}\int_{\gamma_{2}}\frac{\mathrm{d}\ell_{1}\mathrm{d}\ell_{2}}{|x-y|^{2}}\ge\frac{1}{4}L(\gamma_{1})L(\gamma_{2}). \tag*{(16)}

Combining E(Γ)≤ΛE(\Gamma)\le\Lambda with (14)–(15) gives the stronger individual estimates

L(γ1)≤4ΛL(γ2)≤6Λπ,L(γ2)≤4ΛL(γ1)≤4Λπ.L(\gamma_{1})\le\frac{4\Lambda}{L(\gamma_{2})}\le\frac{6\Lambda}{\pi},\qquad L(\gamma_{2})\le\frac{4\Lambda}{L(\gamma_{1})}\le\frac{4\Lambda}{\pi}.

Together with (14)–(15), these imply the symmetric bounds (11).

To obtain a uniform lower bound for the distance between the two components, we set

d:=dist⁡R4(γ1,γ2)=min⁡x∈γ1(S1), y∈γ2(S1)∣x−y∣.d:=\operatorname{dist}_{\mathbb{R}^{4}}(\gamma_{1},\gamma_{2})=\min_{x\in\gamma_{1}(\mathbb{S}^{1}),\,y\in\gamma_{2}(\mathbb{S}^{1})}|x-y|.

Using unit-speed arclength coordinates on R/L(γi)Z\mathbb{R}/L(\gamma_{i})\mathbb{Z} with the two origins chosen at a closest pair, so that

∣γ1(0)−γ2(0)∣=d.|\gamma_{1}(0)-\gamma_{2}(0)|=d.

Put ℓ:=2π3\ell:=\frac{2\pi}{3}. The lower length bounds ensure that both curves contain arclength intervals [0,ℓ][0,\ell]. For 0≤s,t≤ℓ0\le s,t\le\ell, the triangle inequality and the fact that chordal distance is bounded above by arclength give

∣γ1(s)−γ2(t)∣≤∣γ1(s)−γ1(0)∣+∣γ1(0)−γ2(0)∣+∣γ2(0)−γ2(t)∣≤s+d+t.|\gamma_{1}(s)-\gamma_{2}(t)|\le|\gamma_{1}(s)-\gamma_{1}(0)|+|\gamma_{1}(0)-\gamma_{2}(0)|+|\gamma_{2}(0)-\gamma_{2}(t)| \le s+d+t.

Restricting the energy integral to this parameter rectangle, we obtain

Λ≥E(Γ)≥∫0ℓ∫0ℓds dt(d+s+t)2=log⁡(d+ℓ)2d(d+2ℓ).\Lambda\ge E(\Gamma)\ge\int_{0}^{\ell}\int_{0}^{\ell}\frac{\mathrm{d}s\,\mathrm{d}t}{(d+s+t)^{2}}=\log\frac{(d+\ell)^{2}}{d(d+2\ell)}.

Solving this inequality gives an explicit choice:

d≥dΛ:=2π3((1−e−Λ)−1/2−1)>0.d\ge d_{\Lambda}:=\frac{2\pi}{3}\left((1-e^{-\Lambda})^{-1/2}-1\right)>0.

Thus we complete the proof. □

Remark 3.2. Ge [3] proved that any two disjoint linked Jordan curves γ1,γ2⊂S3\gamma_{1},\gamma_{2}\subset\mathbb{S}^{3} satisfy

dist⁡S3(γ1,γ2)≤π2,\operatorname{dist}_{\mathbb{S}^{3}}(\gamma_{1},\gamma_{2})\le\frac{\pi}{2},

with equality precisely for two great circles in orthogonal 2-planes.

From the Euler–Lagrange equation, we obtain regularity and uniform derivative estimates for weakly critical pairs.

Lemma 3.3. Assume dist⁡(γ1,γ2)≥d>0\operatorname{dist}(\gamma_1,\gamma_2) \ge d > 0. Then, for every integer m≥0m \ge0, i,j={1,2}i,j = \{1,2\}, i≠ji \ne j,

∣∇mρj∣L∞(γi)≤CmL(γj)d−m−2,ρj≥L(γj)4.\lvert\nabla^{m}\rho_j\rvert_{L^\infty(\gamma_i)} \le C_m L(\gamma_j)d^{-m-2}, \qquad\rho_j \ge\frac{L(\gamma_j)}{4}.

For m≥1m \ge1,

∣∇mlog⁡ρj∣L∞(γi)≤Cm(d)(17)\lvert\nabla^{m}\log\rho_j\rvert_{L^\infty(\gamma_i)} \le C_m(d) \tag*{(17)}

with Cm(d)C_m(d) independent of L(γj)L(\gamma_j).

Proof. We only prove the case i=1i=1, j=2j=2. Repeated Euclidean differentiation of K(x,y)=∣x−y∣−2K(x,y)=\lvert x-y\rvert^{-2} yields a finite sum of tensors

DxmK(x,y)=∣x−y∣−m−2Pm(x−y∣x−y∣),D_x^m K(x,y)=\lvert x-y\rvert^{-m-2}P_m\left(\frac{x-y}{\lvert x-y\rvert}\right),

where PmP_m is a universal tensor-valued polynomial. The Gauss formula (4) gives

∣∇xmK(x,y)∣≤Cm∣x−y∣−m−2≤Cmd−m−2.\lvert\nabla_x^m K(x,y)\rvert\le C_m\lvert x-y\rvert^{-m-2} \le C_m d^{-m-2}.

Arclength parametrization and dominated convergence justify differentiation under the integral:

∇mρ2(x)=∫γ2∇xmK(x,y) dℓ2(y).\nabla^m \rho_2(x)=\int_{\gamma_2}\nabla_x^m K(x,y)\,\mathrm{d}\ell_2(y).

This proves the first estimate. Since ∣x−y∣2≤4\lvert x-y\rvert^2 \le4, the kernel is at least 1/41/4, giving the lower bound.

For the logarithmic estimates, write

ρ2=L(γ2)ρˉ2,ρˉ2(x)=1L(γ2)∫γ2K(x,y) dℓ2(y).\rho_2=L(\gamma_2)\bar{\rho}_2,\qquad\bar{\rho}_2(x)=\frac{1}{L(\gamma_2)}\int_{\gamma_2}K(x,y)\,\mathrm{d}\ell_2(y).

The function ρˉ2\bar{\rho}_2 is bounded below by 1/41/4, and all of its ambient derivatives are bounded by constants depending only on mm and dd. Since log⁡ρ2=log⁡L(γ2)+log⁡ρˉ2\log\rho_2=\log L(\gamma_2)+\log\bar{\rho}_2, every derivative of positive order is independent of the constant factor L(γ2)L(\gamma_2). The ordinary chain rule for log⁡\log proves (17). □

Proposition 3.4. Every constant-speed weakly critical pair in A\mathcal{A} is smooth. Moreover, let Γn=(γ1,n,γ2,n)\Gamma_n=(\gamma_{1,n},\gamma_{2,n}) be sequences of constant-speed weakly critical pairs, under the bounds

0<L−≤L(γi,n)≤L+<∞,dist⁡(γ1,n,γ2,n)≥d0>0.0<L_- \le L(\gamma_{i,n}) \le L_+<\infty,\qquad\operatorname{dist}(\gamma_{1,n},\gamma_{2,n})\ge d_0>0.

one has ∣γi,n∣Cm≤Cm(L−,L+,d)\lvert\gamma_{i,n}\rvert_{C^m}\le C_m(L_-,L_+,d) for every $m\ge0.

Proof. By Lemma 2.1, take

∣γi′∣≡ai=Li2π>0,Ti=γi′ai.\lvert\gamma_i'\rvert\equiv a_i=\frac{L_i}{2\pi}>0,\qquad T_i=\frac{\gamma_i'}{a_i}.

Thus the curvature and γi′′\gamma_i'' satisfy

γi′′=ai2(κi−γi).\gamma_i''=a_i^2(\kappa_i-\gamma_i).

Combining this identity with (9) gives the following Euler–Lagrange equation:

γi′′=ai2(PNiGj(γi)−γi),Gj=grad⁡log⁡ρj,j≠i.(18)\gamma_i''=a_i^2\left(P_{N_i}G_j(\gamma_i)-\gamma_i\right),\qquad G_j=\operatorname{grad}\log\rho_j,\qquad j\ne i. \tag*{(18)}

Here PNiP_{N_i} is the projection onto the normal space. By Lemma 3.3, GjG_j extends smoothly near γi\gamma_i. Initially H2↪C1,1/2H^2 \hookrightarrow C^{1,1/2}, so γi∈C1,1/2\gamma_i \in C^{1,1/2}, which implies γi∈C2,1/2\gamma_i \in C^{2,1/2}. Repeating this argument yields γi∈C∞\gamma_i \in C^\infty.

Under the stated bounds, ∣γi,n∣=1|\gamma_{i,n}| = 1 and ∣γi,n′∣=ai,n|\gamma'_{i,n}| = a_{i,n} give a uniform C1C^1 bound. Since Lemma 3.3 gives common bounds for every ambient derivative of Gj,nG_{j,n} in a fixed neighborhood of the opposite component, (18) gives the C2C^2 bound. The higher-order estimates follow inductively. Indeed, if

∣γi,n∣Cm≤Cm|\gamma_{i,n}|_{C^m} \le C_m

for some m≥2m \ge2, then differentiating (18) m−1m-1 times expresses γi,n(m+1)\gamma_{i,n}^{(m+1)} as a finite sum involving only derivatives of γi,n\gamma_{i,n} of order at most mm and ambient derivatives of Gj,nG_{j,n} of order at most m−1m-1. All these terms are uniformly bounded by the induction hypothesis and (17). Hence

∣γi,n∣Cm+1≤Cm+1(L−,L+,d).|\gamma_{i,n}|_{C^{m+1}} \le C_{m+1}(L_-,L_+,d).

Induction proves the result. □

Lemma 3.5. Suppose that non-split curve pairs Γn\Gamma_n converge uniformly to a curve pair Γ∞\Gamma_\infty whose component images are disjoint. Then Γ∞\Gamma_\infty is non-split.

Proof. If a smooth embedded sphere SS separated the two limiting images, compactness would give a positive distance from SS to each image. Uniform convergence then places each approximating image, for large nn, on the same side of SS as its limit. The same fixed sphere would split Γn\Gamma_n, a contradiction. □

Combining the preceding normalization, regularity, and uniform estimates, we obtain the following compactness theorem for sequences of weakly critical pairs.

Theorem 3.6. Let Γn∈A\Gamma_n \in\mathcal{A} be non-split and weakly critical, with E(Γn)→2π2E(\Gamma_n) \to2\pi^2 as n→∞n \to\infty. Up to Möbius transformations and separate reparametrizations of the components, a subsequence converges to HH in CmC^m for every mm.

Proof. Discarding finitely many terms, fix Λ\Lambda with E(Γn)≤ΛE(\Gamma_n) \le\Lambda. Apply Proposition 3.1 and Lemma 2.1 to obtain two-sided length bounds, a uniform positive chordal separation, and constant-speed parametrizations. The curves are smooth and have uniform CmC^m bounds for every mm by Proposition 3.4. From Arzelà–Ascoli theorem and a diagonal subsequence, we get a limit Γ∞\Gamma_\infty in every CmC^m norm.

The lower speed bound persists under C1C^1 convergence, and the separation bound persists under uniform convergence. Thus Γ∞∈A\Gamma_\infty\in\mathcal{A}. It is non-split by Lemma 3.5. On the fixed parameter torus the energy densities converge uniformly, since their denominators stay away from zero. Hence

E(Γ∞)=lim⁡n→∞E(Γn)=2π2.E(\Gamma_\infty) = \lim_{n\to\infty} E(\Gamma_n) = 2\pi^2.

The equality case of Theorem 1.1 implies that after a conformal map Φ∈Mob⁡+(S3)\Phi\in\operatorname{Mob}^+(\mathbb{S}^3) (labeling and orientation fixed as explained after (3)), each Φ∘γi,∞\Phi\circ\gamma_{i,\infty} is a regular covering of a standard Hopf circle of some degree mi≠0m_i \ne0, and

2π2=E(Γ∞)=E(ΦΓ∞)=2π2∣m1m2∣.2\pi^2 = E(\Gamma_\infty) = E(\Phi\Gamma_\infty) = 2\pi^2 |m_1m_2|.

So equality forces ∣m1∣=∣m2∣=1|m_1| = |m_2| = 1, both are circle diffeomorphisms. Thus, g∞⋅Γ∞=Hg_\infty\cdot\Gamma_\infty= H for some g∞∈Gg_\infty\in G. These preserve convergence in every CmC^m norm and give the stated conclusion. □

The second variation and infinitesimal Möbius action at the Hopf Link

In this section, we firstly compute the normal directional second variation (the Hessian) of the Möbius cross energy at the Hopf Link, and we obtain its kernel. Then we identify the kernel with the infinitesimal Möbius action at the Hopf Link in the normal direction.

At first, we calculate the second variation of the Möbius cross energy at Hopf Link. Set

P1=span⁡{e1,e2},P2=span⁡{e3,e4}.P_1 = \operatorname{span}\{e_1,e_2\}, \qquad P_2 = \operatorname{span}\{e_3,e_4\}.

For the Hopf pair (2), the curve-normal bundles are the constant bundles Nxx=P2N_xx = P_2 and Nyy=P1N_yy = P_1. Thus normal fields have the form

U:S1→P2,V:S1→P1.U:\mathbb{S}^1 \to P_2, \qquad V:\mathbb{S}^1 \to P_1.

For k=0,1,2k=0,1,2, we set Hk=Hk(S1,P2)⊕Hk(S1,P1)\mathcal{H}^k=H^k(\mathbb{S}^1,P_2)\oplus H^k(\mathbb{S}^1,P_1) and H−1=(H1)∗\mathcal{H}^{-1}=(\mathcal{H}^1)^*. In particular,

∥W∥H12=∫02π(∣U∣2+∣U′∣2) ds+∫02π(∣V∣2+∣V′∣2) dt,W=(U,V).(19)\lVert W\rVert_{H^1}^2=\int_0^{2\pi}(|U|^2+|U'|^2)\,\mathrm{d}s+\int_0^{2\pi}(|V|^2+|V'|^2)\,\mathrm{d}t,\qquad W=(U,V). \tag*{(19)}

Proposition 4.1. The Hopf link is critical. For W=(U,V)∈H2W=(U,V)\in\mathcal{H}^2 its normal Hessian is given by

δ2EH(W)=π∫02π∣U′∣2 ds+π∫02π∣V′∣2 dt+2∫02π∫02π⟨y(t),U(s)⟩⟨x(s),V(t)⟩ ds dt.(20)\delta^2 E_H(W)=\pi\int_0^{2\pi}|U'|^2\,\mathrm{d}s+\pi\int_0^{2\pi}|V'|^2\,\mathrm{d}t+2\int_0^{2\pi}\int_0^{2\pi}\langle y(t),U(s)\rangle\langle x(s),V(t)\rangle\,\mathrm{d}s\,\mathrm{d}t. \tag*{(20)}

Proof. It suffices first to take smooth fields and use the pointwise spherical exponential variations xτ(s)=exp⁡x(s)(τU(s))x_\tau(s)=\exp_{x(s)}(\tau U(s)) and yτ(t)=exp⁡y(t)(τV(t))y_\tau(t)=\exp_{y(t)}(\tau V(t)). Orthogonality of the two planes gives

xτ=x+τU−τ22∣U∣2x+O(τ3),yτ=y+τV−τ22∣V∣2y+O(τ3).x_\tau=x+\tau U-\frac{\tau^2}{2}|U|^2x+O(\tau^3),\qquad y_\tau=y+\tau V-\frac{\tau^2}{2}|V|^2y+O(\tau^3).

Using x,x′∈P1x,x'\in P_1 and U,U′∈P2U,U'\in P_2, and symmetrically for y,Vy,V, gives

∣xτ′∣=1+τ22(∣U′∣2−∣U∣2)+O(τ3),∣yτ′∣=1+τ22(∣V′∣2−∣V∣2)+O(τ3).|x_\tau'|=1+\frac{\tau^2}{2}(|U'|^2-|U|^2)+O(\tau^3),\qquad|y_\tau'|=1+\frac{\tau^2}{2}(|V'|^2-|V|^2)+O(\tau^3).

Let a(s,t)=⟨U(s),y(t)⟩+⟨x(s),V(t)⟩a(s,t)=\langle U(s),y(t)\rangle+\langle x(s),V(t)\rangle. For cτ=⟨xτ,yτ⟩c_\tau=\langle x_\tau,y_\tau\rangle, orthogonality of P1P_1 and P2P_2 gives c0=0c_0=0, c˙0=a\dot{c}_0=a, and c¨0=2⟨U,V⟩=0\ddot{c}_0=2\langle U,V\rangle=0. Therefore

1∣xτ−yτ∣2=12(1−cτ)=12+τ2a+τ22a2+O(τ3).\frac{1}{|x_\tau-y_\tau|^2}=\frac{1}{2(1-c_\tau)}=\frac{1}{2}+\frac{\tau}{2}a+\frac{\tau^2}{2}a^2+O(\tau^3).

The first-order term is linear in x(s)x(s) and y(t)y(t). Since both Hopf circles have vanishing mean, ∫S1x(s) ds=∫S1y(t) dt=0\int_{\mathbb{S}^1}x(s)\,\mathrm{d}s=\int_{\mathbb{S}^1}y(t)\,\mathrm{d}t=0, every such term integrates to zero on the parameter torus, hence δEH=0\delta E_H=0 and HH is critical. Twice the coefficient of τ2\tau^2 after integration is

∬a2 ds dt+π∫02π(∣U′∣2−∣U∣2) ds+π∫02π(∣V′∣2−∣V∣2) dt.\iint a^2\,\mathrm{d}s\,\mathrm{d}t+\pi\int_0^{2\pi}(|U'|^2-|U|^2)\,\mathrm{d}s+\pi\int_0^{2\pi}(|V'|^2-|V|^2)\,\mathrm{d}t.

Finally,

∫02πx⊗x ds=πId⁡P1,∫02πy⊗y dt=πId⁡P2.(21)\int_0^{2\pi}x\otimes x\,\mathrm{d}s=\pi\operatorname{Id}_{P_1},\qquad\int_0^{2\pi}y\otimes y\,\mathrm{d}t=\pi\operatorname{Id}_{P_2}. \tag*{(21)}

Hence the pure squares in ∬a2 ds dt\iint a^2\,\mathrm{d}s\,\mathrm{d}t equal π∫∣U∣2\pi\int|U|^2 and π∫∣V∣2\pi\int|V|^2 and cancel the negative zeroth-order terms. The remaining cross term is exactly (20). □

Remark 4.2. In fact, the first variation extends continuously to H1\mathcal{H}^{1}, while the second variation extends to a bounded bilinear form on H1×H1\mathcal{H}^{1} \times\mathcal{H}^{1}; see Lemma 5.1 for details. These extensions will be used in the local analysis.

Decompose the normal fields into their constant, first-harmonic, and higher-frequency parts:

U(s)=U0+Ax(s)+U≥2(s),V(t)=V0+By(t)+V≥2(t)U(s) = U_0 + Ax(s) + U_{\geq2}(s), \qquad V(t) = V_0 + By(t) + V_{\geq2}(t)

where

U0=12π∫02πU(s) ds,A=1π∫02πU⊗x ds,U_0 = \frac{1}{2\pi}\int_0^{2\pi} U(s)\,\mathrm{d}s, \qquad A = \frac{1}{\pi}\int_0^{2\pi} U \otimes x\,\mathrm{d}s,

and V0V_0, BB are defined analogously. Thus A∈Hom⁡(P1,P2)A \in\operatorname{Hom}(P_1,P_2), B∈Hom⁡(P2,P1)B \in\operatorname{Hom}(P_2,P_1), the Fourier expansion of U≥2U_{\geq2} contains only frequencies ∣n∣≥2|n| \geq2.

If U∈H1U \in H^1, the same orthogonal decomposition holds in H1H^1, because Fourier projection commutes with differentiation.

Proposition 4.3. For every W∈H1W \in\mathcal{H}^1 one has the exact identity

δ2EH(W)=π2∣B+AT∣F2+π∫∣U≥2′∣2+π∫∣V≥2′∣2.(22)\delta^2 E_{\mathrm{H}}(W) = \pi^2 \lvert B + A^T\rvert_F^2 + \pi\int\lvert U'_{\geq2}\rvert^2 + \pi\int\lvert V'_{\geq2}\rvert^2. \tag*{(22)}

where ∣A∣F2=tr⁡(ATA)\lvert A\rvert_F^2 = \operatorname{tr}(A^T A). Consequently, the nullspace KH\mathcal{K}_{\mathrm{H}} of the bilinear form is

KH={(U0,V0):U0∈P2, V0∈P1}⊕{(Ax,−ATy):A∈Hom⁡(P1,P2)},(23)\mathcal{K}_{\mathrm{H}} = \{(U_0,V_0): U_0 \in P_2,\ V_0 \in P_1\} \oplus\{(Ax,-A^T y): A \in\operatorname{Hom}(P_1,P_2)\}, \tag*{(23)}

and dim⁡KH=8\dim\mathcal{K}_{\mathrm{H}} = 8. Its L2L^2-orthogonal complement is

KH⊥={(Ax+U≥2,ATy+V≥2):A∈Hom⁡(P1,P2)}.(24)\mathcal{K}_{\mathrm{H}}^\perp= \{(Ax + U_{\geq2}, A^T y + V_{\geq2}): A \in\operatorname{Hom}(P_1,P_2)\}. \tag*{(24)}

For every W∈H1∩KH⊥W \in\mathcal{H}^1 \cap\mathcal{K}_{\mathrm{H}}^\perp,

δ2EH(W)≥4π5∥W∥H12.(25)\delta^2 E_{\mathrm{H}}(W) \geq\frac{4\pi}{5}\lVert W\rVert_{H^1}^2. \tag*{(25)}

Proof. Only the first Fourier modes contribute to the nonlocal term in (20). Indeed, put

I(U,V)=∫02π∫02π⟨y(t),U(s)⟩⟨x(s),V(t)⟩ ds dt.I(U,V) = \int_0^{2\pi}\int_0^{2\pi} \langle y(t),U(s)\rangle\langle x(s),V(t)\rangle\,\mathrm{d}s\,\mathrm{d}t.

If U=U0U=U_0 is constant, integration in ss and ∫x(s) ds=0\int x(s)\,\mathrm{d}s=0 give I(U0,V)=0I(U_0,V)=0; similarly I(U,V0)=0I(U,V_0)=0. If U=U≥2U=U_{\geq2}, then for each fixed tt the function s↦⟨y(t),U≥2(s)⟩s\mapsto\langle y(t),U_{\geq2}(s)\rangle contains only frequencies ∣n∣≥2|n|\geq2, whereas s↦⟨x(s),V(t)⟩s\mapsto\langle x(s),V(t)\rangle has frequency ∣n∣=1|n|=1. Their integral therefore vanishes. Interchanging ss and tt shows that every term containing V≥2V_{\geq2} also vanishes. Hence

I(U,V)=I(Ax,By).I(U,V)=I(Ax,By).

Now we compute this remaining term. Since ⟨y,Ax⟩=⟨ATy,x⟩\langle y,Ax\rangle=\langle A^T y,x\rangle, the moment identity (21) gives, after first integrating in ss,

∫02π⟨y,Ax⟩⟨x,By⟩ ds=⟨ATy,(∫02πx⊗x ds)By⟩=π⟨ATy,By⟩=π⟨y,ABy⟩.\int_0^{2\pi}\langle y,Ax\rangle\langle x,By\rangle\,\mathrm{d}s = \left\langle A^T y,\left(\int_0^{2\pi}x\otimes x\,\mathrm{d}s\right)By\right\rangle = \pi\langle A^T y,By\rangle = \pi\langle y,ABy\rangle.

A second use of (21) yields

I(Ax,By)=πtr⁡(AB∫02πy⊗y dt)=π2tr⁡(AB).I(Ax,By)=\pi\operatorname{tr}\left(AB\int_0^{2\pi}y\otimes y\,\mathrm{d}t\right)=\pi^2\operatorname{tr}(AB).

If ∣A∣F:=(tr⁡(ATA))1/2|A|_F := \left(\operatorname{tr}(A^T A)\right)^{1/2} denotes the Frobenius norm, then

∫02π∣(Ax)′∣2 ds=∫02π⟨Ax′,Ax′⟩ ds=tr⁡(ATA∫02πx′⊗x′ ds)=π∣A∣F2.\int_0^{2\pi} |(Ax)'|^2\,\mathrm{d}s = \int_0^{2\pi} \langle Ax',Ax'\rangle\,\mathrm{d}s = \operatorname{tr}\left(A^T A\int_0^{2\pi} x'\otimes x'\,\mathrm{d}s\right) = \pi|A|_F^2.

because ∫x′⊗x′ ds=πId⁡P1\int x'\otimes x'\,\mathrm{d}s = \pi\operatorname{Id}_{P_1}. The second component gives ∫∣(By)′∣2 dt=π∣B∣F2\int|(By)'|^2\,\mathrm{d}t = \pi|B|_F^2 in the same way. Thus the first-mode contribution is

π2(∣A∣F2+∣B∣F2+2tr⁡(AB))=π2∣B+AT∣F2.\pi^2\left(|A|_F^2 + |B|_F^2 + 2\operatorname{tr}(AB)\right) = \pi^2|B+A^T|_F^2.

so (22) and (23) are proved. Orthogonality to the constant kernel removes U0,V0U_0,V_0, while orthogonality to every (Cx,−CTy)(Cx,-C^T y) gives B=ATB=A^T, this implies (24).

On the transverse first modes,

δ2EH((Ax,ATy))=4π2∣A∣F2and∣(Ax,ATy)∣H12=4π∣A∣F2.\delta^2 E_H((Ax,A^T y)) = 4\pi^2|A|_F^2 \qquad\text{and}\qquad|(Ax,A^T y)|_{H^1}^2 = 4\pi|A|_F^2.

On frequencies ∣n∣≥2|n|\ge2, Poincaré’s inequality gives ∫∣f′∣2≥4∫∣f∣2\int|f'|^2 \ge4\int|f|^2, and hence

π∫∣f′∣2≥(4π/5)∣f∣H12.\pi\int|f'|^2 \ge(4\pi/5)|f|_{H^1}^2.

The Fourier pieces are orthogonal, which proves (25). □

The kernel in (23) consists of constant normal fields and mixed rotational fields. We will construct an explicit family of Möbius transformations realizing these velocities.

For A∈Hom⁡(P1,P2)A\in\operatorname{Hom}(P_1,P_2), the skew-symmetric matrix

(0−ATA0)\begin{pmatrix} 0 & -A^T\\ A & 0 \end{pmatrix}

generates a rotation mixing the two orthogonal planes. Its normal trace along the Hopf pair is (Ax,−ATy)(Ax,-A^T y). The constant normal fields are generated by gradients of height functions. For a=a1+a2∈P1⊕P2a=a_1+a_2\in P_1\oplus P_2, set fa(z)=⟨a,z⟩f_a(z)=\langle a,z\rangle. On the unit sphere,

∇fa(z)=a−⟨a,z⟩z,∇2fa=−fag0.\nabla f_a(z)=a-\langle a,z\rangle z,\qquad\nabla^2 f_a=-f_a g_0.

Thus ∇fa\nabla f_a is a conformal vector field. Its normal trace along the Hopf pair is (a2,a1)(a_2,a_1), since the curve-normal planes along xx and yy are P2P_2 and P1P_1, respectively.

Accordingly, for the uniquely decomposed kernel field

h=(U0+Ax,V0−ATy)∈KH,h=(U_0+Ax,V_0-A^T y)\in\mathcal{K}_H,

define

a(h)=U0+V0,Ω(h)=(0−ATA0),(26)a(h)=U_0+V_0,\qquad\Omega(h)= \begin{pmatrix} 0 & -A^T\\ A & 0 \end{pmatrix}, \tag*{(26)}

and

Xh(z)=Ω(h)z+∇fa(h)(z)=Ω(h)z+a(h)−⟨a(h),z⟩z.(27)X_h(z)=\Omega(h)z+\nabla f_{a(h)}(z)=\Omega(h)z+a(h)-\langle a(h),z\rangle z. \tag*{(27)}

This assignment is linear in hh, and (PP2Xh(x),PP1Xh(y))=h(P_{P_2}X_h(x),P_{P_1}X_h(y))=h. In particular, a(h)a(h) is chosen to realize the constant part of hh, while Ω(h)\Omega(h) realizes its first-harmonic part.

Every Möbius transformation of S3\mathbb{S}^3 is a composition of a rotation in SO⁡(4)\operatorname{SO}(4) and a standard conformal diffeomorphism Tb(z)T_b(z) of S3\mathbb{S}^3 as in the following:

Tb(z)=(1−∣b∣2)z−b∣z−b∣2−b,z∈S3,(28)T_b(z)=(1-|b|^2)\frac{z-b}{|z-b|^2}-b,\qquad z\in\mathbb{S}^3, \tag*{(28)}

where ∣b∣<1|b|<1.

We see T0=Id⁡T_0=\operatorname{Id} and TτbT_{\tau b} is a path of sphere diffeomorphisms for 0≤τ≤10\leq\tau\leq1.

ddτ∣τ=0Tτb(z)=−2(b−⟨b,z⟩z)=−2∇fb(z).(29)\left.\frac{\mathrm{d}}{\mathrm{d}\tau}\right|_{\tau=0}T_{\tau b}(z)=-2\left(b-\langle b,z\rangle z\right)=-2\nabla f_b(z). \tag*{(29)}

Consequently, T−τa/2T_{-\tau a/2} has initial velocity ∇fa\nabla f_a. This fixes the normalization of the conformal parameter in the following construction.

Proposition 4.4. For h∈KHh\in\mathcal{K}_{H} sufficiently close to zero, define

Φh=exp⁡(Ω(h))∘T−a(h)/2.(30)\Phi_h=\exp\left(\Omega(h)\right)\circ T_{-a(h)/2}. \tag*{(30)}

Then Φh∈Mob⁡+(S3)\Phi_h\in\operatorname{Mob}^{+}(\mathbb{S}^{3}), the map (h,z)↦Φh(z)(h,z)\mapsto\Phi_h(z) is smooth, and Φ0=Id⁡\Phi_0=\operatorname{Id}. Moreover,

ddτ∣τ=0Φτh(z)=Xh(z),(31)\left.\frac{\mathrm{d}}{\mathrm{d}\tau}\right|_{\tau=0}\Phi_{\tau h}(z)=X_h(z), \tag*{(31)}

so the normal differential of this family along HH is the identity on KH\mathcal{K}_{H}. Thus KH\mathcal{K}_{H} is exactly the space of normal infinitesimal Möbius motions of the Hopf pair.

Proof. The exponential in (30) is the ordinary matrix exponential and belongs to SO⁡(4)\operatorname{SO}(4). Linearity of a(h)a(h) and Ω(h)\Omega(h), together with (29), gives

Φτh(z)=z+τ(Ω(h)z+a(h)−⟨a(h),z⟩z)+O(τ2).(32)\Phi_{\tau h}(z)=z+\tau\left(\Omega(h)z+a(h)-\langle a(h),z\rangle z\right)+O(\tau^{2}). \tag*{(32)}

By combining (27), we have (31), i.e. every kernel field is realized as a normal Möbius velocity.

Conversely, let us eliminate the tangential velocity of a Möbius motion by reparametrization, then the energy invariance implies zero second variation in its remaining normal velocity, which therefore belongs to KH\mathcal{K}_{H}.

We now derive the linearized operator from the quadratic form. For W=(U,V)W=(U,V) and W~=(U~,V~)\widetilde{W}=(\widetilde{U},\widetilde{V}) define the polarization

BH(W,W~):=12(δ2EH(W+W~)−δ2EH(W)−δ2EH(W~)).B_H(W,\widetilde{W}):=\frac{1}{2}\left(\delta^{2}E_H(W+\widetilde{W})-\delta^{2}E_H(W)-\delta^{2}E_H(\widetilde{W})\right).

Substitution of (20) and integrating by parts gives

BH(W,W~)=∫⟨(LHW)1,U~⟩ ds+∫⟨(LHW)2,V~⟩ dt,B_H(W,\widetilde{W})=\int\left\langle(\mathcal{L}_H W)_1,\widetilde{U}\right\rangle\,\mathrm{d}s+\int\left\langle(\mathcal{L}_H W)_2,\widetilde{V}\right\rangle\,\mathrm{d}t,

where the self-adjoint operator is

(LHW)1(s)=−πU′′(s)+∫02πy(t)⟨x(s),V(t)⟩ dt,(LHW)2(t)=−πV′′(t)+∫02πx(s)⟨y(t),U(s)⟩ ds.(33)\begin{aligned} (\mathcal{L}_H W)_1(s)&=-\pi U''(s)+\int_{0}^{2\pi}y(t)\langle x(s),V(t)\rangle\,\mathrm{d}t,\\ (\mathcal{L}_H W)_2(t)&=-\pi V''(t)+\int_{0}^{2\pi}x(s)\langle y(t),U(s)\rangle\,\mathrm{d}s. \tag*{(33)} \end{aligned}

On H0\mathcal{H}^{0} with domain H2\mathcal{H}^{2}, LH\mathcal{L}_{H} is a self-adjoint operator. It also admits the weak realization

LH:H1⟶H−1\mathcal{L}_{H}:\mathcal{H}^{1}\longrightarrow\mathcal{H}^{-1}

where the second derivatives in (33) are understood distributionally. Moreover,

⟨LHW,W~⟩=BH(W,W~).\left\langle\mathcal{L}_{H}W,\widetilde{W}\right\rangle=B_H(W,\widetilde{W}).

Its kernel is KH\mathcal{K}_{H}. Since KH\mathcal{K}_{H} is finite-dimensional and consists of smooth fields, the L2L^{2}-orthogonal projections

PKH,P⊥=I−PKH(34)P_{\mathcal{K}_{H}},\qquad P_{\perp}=I-P_{\mathcal{K}_{H}} \tag*{(34)}

extend boundedly to H1\mathcal{H}^{1} and H−1\mathcal{H}^{-1}.

Nonlinear estimate and renormalization lemma

In this section, we derive a nonlinear estimate for the first variation near the Hopf link. Motivated by the renormalization lemma of Mondino and Nguyen [7], we then construct a local Möbius gauge in which any sufficiently nearby link is represented as a normal graph over HH with graph field orthogonal to KH\mathcal{K}_{H}.

Fix 0<r0<π/40<r_{0}<\pi/4. The normal exponential maps

Ψ1(s,ξ)=exp⁡x(s)ξ,(s,ξ)∈S1×Br0(0;P2),(35)\Psi_{1}(s,\xi)=\exp_{x(s)}\xi,\qquad(s,\xi)\in S^{1}\times B_{r_{0}}(0;P_{2}), \tag*{(35)}
Ψ2(t,η)=exp⁡y(t)η,(t,η)∈S1×Br0(0;P1),(36)\Psi_{2}(t,\eta)=\exp_{y(t)}\eta,\qquad(t,\eta)\in S^{1}\times B_{r_{0}}(0;P_{1}), \tag*{(36)}

are diffeomorphisms onto disjoint tubular neighborhoods of the Hopf circles. For a small field v=(U,V)v=(U,V), set

Graph⁡H(v)=(XU,YV),XU(s)=Ψ1(s,U(s)),YV(t)=Ψ2(t,V(t)),\operatorname{Graph}_{H}(v)=(X_{U},Y_{V}),\qquad X_{U}(s)=\Psi_{1}(s,U(s)),\qquad Y_{V}(t)=\Psi_{2}(t,V(t)),

and define

E(v)=E(Graph⁡H(v)),⟨G(v),ζ⟩=δE(v)[ζ].\mathcal{E}(v)=E(\operatorname{Graph}_{H}(v)),\qquad\langle\mathcal{G}(v),\zeta\rangle=\delta\mathcal{E}(v)[\zeta].

Here v∈H2v\in\mathcal{H}^{2} has small C1C^{1} norm, and ζ∈H2\zeta\in\mathcal{H}^{2}. The notation δG(v)[ζ]\delta\mathcal{G}(v)[\zeta] denotes its linearization in the direction ζ\zeta and satisfies

⟨δG(v)[ζ],η⟩=δ2E(v)[ζ,η].\langle\delta\mathcal{G}(v)[\zeta],\eta\rangle=\delta^{2}\mathcal{E}(v)[\zeta,\eta].

The graph variation in the first component is

∂ξΨ1(s,U(s))[ζ1(s)].\partial_{\xi}\Psi_{1}(s,U(s))[\zeta_{1}(s)].

Its projection to the normal plane of the current curve is an isomorphism at v=0v=0, and remains so for small ∣v∣C1|v|_{C^{1}}. The second component has the same property. For ∣v∣C1|v|_{C^{1}} small, the normal projection of the graph variations onto the normal bundles of the current curves is pointwise invertible. Hence, Proposition 2.2 gives

G(v)=0⟺Graph⁡H(v) is weakly critical.(37)\mathcal{G}(v)=0\quad\Longleftrightarrow\quad\operatorname{Graph}_{H}(v)\text{ is weakly critical}. \tag*{(37)}

Lemma 5.1. There are C,r1>0C,r_{1}>0 such that, for v,w∈H2v,w\in\mathcal{H}^{2} with ∣v∣C1,∣w∣C1<r1|v|_{C^{1}},|w|_{C^{1}}<r_{1}, the linearization δG(v)\delta\mathcal{G}(v) extends to a uniformly bounded map H1→H−1\mathcal{H}^{1}\to\mathcal{H}^{-1} and

∣δG(v)−δG(w)∣Lin⁡(H1,H−1)≤C∣v−w∣C1.(38)|\delta\mathcal{G}(v)-\delta\mathcal{G}(w)|_{\operatorname{Lin}(\mathcal{H}^{1},\mathcal{H}^{-1})}\leq C|v-w|_{C^{1}}. \tag*{(38)}

Here Lin⁡(X,Y)\operatorname{Lin}(X,Y) is the space of bounded linear maps with its operator norm. Moreover,

G(0)=0,δG(0)=LH,\mathcal{G}(0)=0,\qquad\delta\mathcal{G}(0)=\mathcal{L}_{H},

and

∣G(v)−LHv∣H−1≤C∣v∣C1∣v∣H1.(39)|\mathcal{G}(v)-\mathcal{L}_{H}v|_{H^{-1}}\leq C|v|_{C^{1}}|v|_{H^{1}}. \tag*{(39)}

Proof. The chain rule gives

XU′=∂sΨ1(s,U)+∂UΨ1(s,U)[U′],YV′=∂tΨ2(t,V)+∂VΨ2(t,V)[V′].X_{U}'=\partial_{s}\Psi_{1}(s,U)+\partial_{U}\Psi_{1}(s,U)[U'],\qquad Y_{V}'=\partial_{t}\Psi_{2}(t,V)+\partial_{V}\Psi_{2}(t,V)[V'].

Hence

E(v)=∬f(s,t,U(s),U′(s),V(t),V′(t)) ds dt,(40)\mathcal{E}(v)=\iint f(s,t,U(s),U'(s),V(t),V'(t))\,\mathrm{d}s\,\mathrm{d}t, \tag*{(40)}

where

f(s,t,U,U′,V,V′)=∣∂sΨ1(s,U)+∂UΨ1(s,U)[U′]∣∣∂tΨ2(t,V)+∂VΨ2(t,V)[V′]∣∣Ψ1(s,U)−Ψ2(t,V)∣2.f(s,t,U,U',V,V')= \frac{ \left|\partial_{s}\Psi_{1}(s,U)+\partial_{U}\Psi_{1}(s,U)[U']\right| \left|\partial_{t}\Psi_{2}(t,V)+\partial_{V}\Psi_{2}(t,V)[V']\right| }{ \left|\Psi_{1}(s,U)-\Psi_{2}(t,V)\right|^{2} }.

Let q=(U,U′,V,V′)q=(U,U',V,V'). For ∣v∣C1<r1|v|_{C^{1}}<r_{1}, the speeds and the separation of the two components are uniformly bounded away from zero. Hence DqkfD_{q}^{k}f is uniformly bounded for $k\leq 3.

To display the differentiation, abbreviate the tuple of values and first derivatives by

zv(s,t)=(U(s),U′(s),V(t),V′(t)).z_v(s,t) = (U(s), U'(s), V(t), V'(t)).

We define zζ,zηz_\zeta, z_\eta analogously for ζ,η∈H2\zeta,\eta\in H^2. Twice differentiating the finite-dimensional integrand gives

⟨δG(v)[ζ],η⟩=∬∂q2f(s,t,zv)[zζ,zη] ds dt.(41)\langle\delta\mathcal{G}(v)[\zeta],\eta\rangle= \iint\partial_q^2 f(s,t,z_v)[z_\zeta,z_\eta]\,ds\,dt. \tag*{(41)}

Since ∂q2f\partial_q^2 f is uniformly bounded, each term in Dq2f(s,t,zv)[zζ,zη]D_q^2 f(s,t,z_v)[z_\zeta,z_\eta] is bounded by a constant times a product acac, where aa is one of

∣ζ1(s)∣,∣ζ1′(s)∣,∣ζ2(t)∣,∣ζ2′(t)∣,|\zeta_1(s)|,\quad|\zeta_1'(s)|,\quad|\zeta_2(t)|,\quad|\zeta_2'(t)|,

and cc is the corresponding quantity associated with η\eta. Hence, by Cauchy-Schwarz inequality

∣D2E(v)[ζ,η]∣≤C∬T2∣zζ(s,t)∣ ∣zη(s,t)∣ ds dt≤C∣zζ∣L2(T2)∣zη∣L2(T2).|D^2\mathcal{E}(v)[\zeta,\eta]| \leq C\iint_{\mathbb{T}^2}|z_\zeta(s,t)|\,|z_\eta(s,t)|\,ds\,dt \leq C|z_\zeta|_{L^2(\mathbb{T}^2)}|z_\eta|_{L^2(\mathbb{T}^2)}.

Since

∣zζ∣L2(T2)=2π∣ζ∣H1,∣zη∣L2(T2)=2π∣η∣H1,|z_\zeta|_{L^2(\mathbb{T}^2)} = 2\pi|\zeta|_{H^1},\qquad|z_\eta|_{L^2(\mathbb{T}^2)} = 2\pi|\eta|_{H^1},

we obtain ∣D2E(v)[ζ,η]∣≤C∣ζ∣H1∣η∣H1|D^2\mathcal{E}(v)[\zeta,\eta]| \leq C|\zeta|_{H^1}|\eta|_{H^1}. Thus D2E(v)D^2\mathcal{E}(v) extends uniquely to a bounded bilinear form on H1×H1H^1 \times H^1.

Finally, the bound for the third derivatives of ff and the mean-value theorem give

sup⁡s,t∣∂q2f(s,t,zv)−∂q2f(s,t,zw)∣≤C∣v−w∣C1.\sup_{s,t}|\partial_q^2 f(s,t,z_v)-\partial_q^2 f(s,t,z_w)| \leq C|v-w|_{C^1}.

Applying the same product estimates to this difference proves (38). At v=0v=0 the chart Hessian is the normal Hopf Hessian, so δG(0)=LH\delta\mathcal{G}(0)=\mathcal{L}_H. Integration along τv\tau v now gives

G(v)−LHv=∫01(δG(τv)−δG(0))[v] dτ,\mathcal{G}(v)-\mathcal{L}_H v = \int_0^1(\delta\mathcal{G}(\tau v)-\delta\mathcal{G}(0))[v]\,d\tau,

and (38) proves (39). □

Write Ψi−1=(πi,νi)\Psi_i^{-1}=(\pi_i,\nu_i). For Z=(Z1,Z2)Z=(Z_1,Z_2) close to HH in C1C^1, let

θi=πi∘Zi,ηi=νi∘Zi.\theta_i=\pi_i\circ Z_i,\qquad\eta_i=\nu_i\circ Z_i.

Then θi\theta_i is a degree-one circle diffeomorphism close to the identity. The field of its normal-graph representative is

ng⁡H(Z):=(η1∘θ1−1,η2∘θ2−1).(42)\operatorname{ng}_H(Z):=(\eta_1\circ\theta_1^{-1},\eta_2\circ\theta_2^{-1}). \tag*{(42)}

Lemma 5.2. On a sufficiently small C1C^1 neighborhood of HH, the functions θi,ηi\theta_i,\eta_i satisfy θi′>1/2\theta_i'>1/2, after taking lifts close to the identity. They define degree-one circle diffeomorphisms and

ui=ηi∘θi−1,Zi(r)=Ψi(θi(r),ui(θi(r))).u_i=\eta_i\circ\theta_i^{-1},\qquad Z_i(r)=\Psi_i(\theta_i(r),u_i(\theta_i(r))).

The assignment Z↦ng⁡H(Z)=(u1,u2)Z\mapsto\operatorname{ng}_H(Z)=(u_1,u_2) is continuous in C1C^1, with ng⁡H(H)=0\operatorname{ng}_H(H)=0. If Z∈H2Z\in H^2, then θi,θi−1\theta_i,\theta_i^{-1}, and uiu_i belong to H2H^2.

Proof. At Z=HZ=H, θi(r)=r\theta_i(r)=r and ηi=0\eta_i=0. Smoothness of the tubular coordinates gives the lower derivative bound and the identity θi(r+2π)=θi(r)+2π\theta_i(r+2\pi)=\theta_i(r)+2\pi on a small C1C^1 neighborhood. Thus each θi\theta_i is a degree-one diffeomorphism.

For completeness, suppose Z(n)→ZZ^{(n)} \to Z in C1C^1. Smooth pointwise composition gives θi(n)→θi\theta_i^{(n)} \to\theta_i and ηi(n)→ηi\eta_i^{(n)} \to\eta_i in C1C^1. The inverse lifts are uniformly Lipschitz, with constant at most 2, so their uniform convergence follows from that of the original lifts. The formula

((θi(n))−1)′=1(θi(n))′∘(θi(n))−1((\theta_i^{(n)})^{-1})' = \frac{1}{(\theta_i^{(n)})' \circ(\theta_i^{(n)})^{-1}}

and uniform continuity of the limiting derivative give convergence in C1C^1. Applying the chain rule to ui(n)u_i^{(n)} proves ui(n)→uiu_i^{(n)} \to u_i in C1C^1.

If Z∈H2Z \in H^2, the ordinary Sobolev chain rule gives θi,ηi∈H2\theta_i,\eta_i \in H^2. Write ψi=θi−1\psi_i = \theta_i^{-1}. Almost everywhere,

ψi′′=−θi′′∘ψi(θi′∘ψi)3,ui′′=(ηi′′∘ψi)(ψi′)2+(ηi′∘ψi)ψi′′.\psi_i'' = -\frac{\theta_i'' \circ\psi_i}{(\theta_i' \circ\psi_i)^3}, \qquad u_i'' = (\eta_i'' \circ\psi_i)(\psi_i')^2 + (\eta_i' \circ\psi_i)\psi_i''.

The derivative bounds for θi\theta_i and change of variables imply ψi′′∈L2\psi_i'' \in L^2. The first term in ui′′u_i'' lies in L2L^2 by the same change of variables, and the second does because ηi′\eta_i' and ψi′\psi_i' are bounded. This proves the H2H^2 assertions. □

Let PKHP_{\mathcal{K}_H} denote the L2L^2-orthogonal projection onto the kernel of the Hopf Hessian. We define M(Z)M(Z) by

M(Z)≔PKHng⁡H(Z).M(Z) \coloneqq P_{\mathcal{K}_H}\operatorname{ng}_H(Z).

For h∈KHh \in\mathcal{K}_H sufficiently close to zero, let Φh\Phi_h be the Möbius transformation defined in (30). Set

X1=C1(S1,P2)⊕C1(S1,P1),\mathcal{X}^1 = C^1(\mathbb{S}^1,P_2) \oplus C^1(\mathbb{S}^1,P_1),

and define, on sufficiently small zero neighborhoods in KH×X1\mathcal{K}_H \times\mathcal{X}^1,

F(h,w)=M(Φh(Graph⁡H(w))).\mathcal{F}(h,w) = M(\Phi_h(\operatorname{Graph}_H(w))).

Based on the above argument, the following lemma, motivated by the renormalization lemma in [7], provides a local Möbius gauge near HH: after a small Möbius transformation and reparametrization, a nearby link can be written as a normal graph over HH, with its graph field orthogonal to KH\mathcal{K}_H.

Lemma 5.3. For every sufficiently small w∈X1w \in\mathcal{X}^1, there is a unique sufficiently small h(w)∈KHh(w) \in\mathcal{K}_H such that

v=ng⁡H(Φh(w)(Graph⁡H(w)))satisfiesPKHv=0.(43)v = \operatorname{ng}_H(\Phi_{h(w)}(\operatorname{Graph}_H(w))) \quad\text{satisfies} \quad P_{\mathcal{K}_H}v = 0. \tag*{(43)}

Moreover, h(w)h(w) depends smoothly on ww, v→0v \to0 in C1C^1 as w→0w \to0, and H2H^2-regularity is preserved.

Proof. Let k1,…,k8k_1,\ldots,k_8 be a smooth L2L^2-orthonormal basis of KH\mathcal{K}_H with ka=(ka,1,ka,2)k_a=(k_{a,1},k_{a,2}). Define

ma(Z)=∑i=12∫02π⟨ηi(r),ka,i(θi(r))⟩θi′(r) dr.(44)m_a(Z) = \sum_{i=1}^{2} \int_0^{2\pi} \langle\eta_i(r), k_{a,i}(\theta_i(r))\rangle\theta_i'(r)\,\mathrm{d}r. \tag*{(44)}

From the definition of Ψ−1\Psi^{-1}, changing variables q=θi(r)q=\theta_i(r) in the formula for mam_a yields

ma(Z)=∑i=12∫02π⟨ηi(θi−1(q)),ka,i(q)⟩ dq=⟨ng⁡H(Z),ka⟩L2.m_a(Z) = \sum_{i=1}^{2} \int_0^{2\pi} \langle\eta_i(\theta_i^{-1}(q)), k_{a,i}(q)\rangle\,\mathrm{d}q = \langle\operatorname{ng}_H(Z), k_a\rangle_{L^2}.

Thus, we have

M(Z)=∑a=18ma(Z)ka.M(Z)=\sum_{a=1}^{8}m_a(Z)k_a.

Indeed, we can define

fa,i(z,p):=⟨νi(z),ka,i(πi(z))⟩ dπi∣z[p],f_{a,i}(z,p):=\langle\nu_i(z),k_{a,i}(\pi_i(z))\rangle\,d\pi_i|_z[p],

which is smooth in (z,p)(z,p). Thus

ma(Z)=∑i=12∫02πfa,i(Zi(r),Zi′(r)) drm_a(Z)=\sum_{i=1}^{2}\int_{0}^{2\pi}f_{a,i}(Z_i(r),Z_i'(r))\,dr

is smooth in the C1C^1 topology, which means MM is smooth in the C1C^1 topology.

For a C1C^1 sphere-tangent field ζ\zeta along HH, we now compute dM(H)[ζ]dM(H)[\zeta]. At Z=HZ=H,

ηi=0,θi(r)=r,θi′=1.\eta_i=0,\qquad\theta_i(r)=r,\qquad\theta_i'=1.

In differentiating (44), the terms differentiating ka,i(θi)k_{a,i}(\theta_i) and θi′\theta_i' therefore vanish. The remaining term is

dma(H)[ζ]=∑i=12∫02π⟨dνi∣Hi(r)[ζi(r)],ka,i(r)⟩ dr.dm_a(H)[\zeta]=\sum_{i=1}^{2}\int_{0}^{2\pi}\left\langle d\nu_i|_{H_i(r)}[\zeta_i(r)],k_{a,i}(r)\right\rangle\,dr.

The differentials of the inverse tubular coordinates give

dM(H)[ζ]=PKH(PP2ζ1,PP1ζ2).dM(H)[\zeta]=P_{\mathcal K_H}\left(P_{P_2}\zeta_1,P_{P_1}\zeta_2\right).

The graph map and the finite-dimensional Möbius action are smooth in the C1C^1 topology. Hence F\mathcal F is smooth near (0,0)(0,0), and F(0,0)=0\mathcal F(0,0)=0. By Proposition 4.4, for every k∈KHk\in\mathcal K_H,

dhF(0,0)[k]=dMH[(Xk∘x,Xk∘y)]=PKHk=k.d_h\mathcal F(0,0)[k]=dM_H[(X_k\circ x,X_k\circ y)]=P_{\mathcal K_H}k=k.

Thus dhF(0,0)d_h\mathcal F(0,0) is the identity on KH\mathcal K_H. The implicit-function theorem gives a unique smooth small parameter h=h(w)h=h(w), with h(0)=0h(0)=0 and F(h(w),w)=0\mathcal F(h(w),w)=0. This proves (43) and the asserted local uniqueness of the parameter. The remaining continuity and regularity statements follow from Lemma 5.2. □

Proof of the critical-value gap

In this section, we establish a local gap theorem near the Hopf orbit. Combining this local rigidity result with the compactness theorem in Theorem 3.6, we then prove the main theorem.

Theorem 6.1. There is a C1C^1 neighborhood of the Hopf orbit such that every weakly critical pair in this neighborhood belongs to $\mathbf G\cdot\mathbf H.

Proof. By Lemma 5.2, a pair Γ∈A\Gamma\in\mathcal A sufficiently close to H\mathbf H in C1C^1 can be separately reparametrized as Graph⁡H(w)\operatorname{Graph}_{\mathbf H}(w), with w∈H2w\in\mathcal H^2 small in C1C^1.

Apply Lemma 5.3 to obtain a Möbius-equivalent graph Graph⁡H(v)\operatorname{Graph}_{\mathbf H}(v) with v⊥L2KHv\perp_{L^2}\mathcal K_H. The continuity statements in the two lemmas allow us to shrink the original neighborhood so that ∣v∣C1<r∗|v|_{C^1}<r_*, where r∗<r1r_*<r_1 and Cr∗<4π/5Cr_*<4\pi/5 for the constants in Lemma 5.1.

Weak criticality is preserved by both operations from Lemma 2.1. We write

R(v)=G(v)−LHv.\mathcal R(v)=\mathcal G(v)-\mathcal L_Hv.

then Lemma 5.1 implies LHv=−R(v)\mathcal{L}_{H}v=-\mathcal{R}(v). Pairing with vv gives

4π5∣v∣H12≤δ2EH(v)=−⟨R(v),v⟩≤Cr∗∣v∣H12.(45)\frac{4\pi}{5}|v|_{H^1}^{2}\leq\delta^{2}E_{H}(v)=-\langle\mathcal{R}(v),v\rangle\leq Cr_{*}|v|_{H^1}^{2}. \tag*{(45)}

Since Cr∗<4π/5Cr_{*}<4\pi/5, one has v=0v=0. Undoing the transformations proves Γ∈G⋅H\Gamma\in G\cdot H. Finally, each fixed element of GG acts by a C1C^{1} homeomorphism and preserves weak criticality, which proves the statement on the union of translates. □

Proof of Theorem 1.3. Suppose that no positive gap exists. Then, there is a sequence of non-split, weakly critical pairs Γn∈A∖(G⋅H)\Gamma_{n}\in\mathcal{A}\setminus(G\cdot H) such that

2π2≤E(Γn)≤2π2+1n.2\pi^{2}\leq E(\Gamma_{n})\leq2\pi^{2}+\frac{1}{n}.

By Theorem 3.6, a subsequence converges smoothly to HH after common Möbius transformations and separate reparametrizations. For large nn the resulting pairs lie in the neighborhood of Theorem 6.1, and hence belong to G⋅HG\cdot H. Invariance of this orbit gives a contradiction. □

References

  1. [1]Ian Agol; Fernando C. Marques; André Neves. Min-max theory and the energy of links. J. Amer. Math. Soc. 29 (2016), no. 2, 561–578.DOI
  2. [2]Michael H. Freedman; Zheng-Xu He; Zhenghan Wang. Möbius energy of knots and unknots. Ann. of Math. (2) 139 (1994), no. 1, 1–50.DOI
  3. [3]Jian Ge. Gehring link problem, focal radius and over-torical width. Amer. J. Math. 146 (2024), no. 5, 1295–1304.arxiv.org/abs/2102.05901
  4. [4]Zheng-Xu He. On the minimizers of the Möbius cross energy of links. Experiment. Math.11 (2002), no. 2, 244–248.DOI
  5. [6]O’Hara Jun. Energy of a knot. Topology. 30 (1991), no.2, 241-247DOI
  6. [7]Andrea Mondino; Huy The Nguyen. A gap theorem for Willmore tori and an application to the Willmore flow. Nonlinear Anal. 102 (2014), 220–225.DOI

Paper details

vlnc:W_mvehjs7t
Version
v2 · latest
Published
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

GPT-5.6 Sol (bare)

View statement
Looking for related papers…
Contents