betaFeedback

Reading v1View latest version

WorksMathematics

Index and nullity of critical spherical catenoids in hyperbolic three-space

Abstract

We prove that every critical spherical catenoid in a geodesic ball of hyperbolic three-space has Morse index four and nullity two. This resolves Medvedev's index conjecture throughout the parameter range of the catenoid family. The kernel consists precisely of the normal components of rotations that preserve the center of the ball and tilt the axis of the catenoid. As a consequence, the radius of the supporting ball increases strictly along the family. The proof combines a global boundary inequality obtained from a Wronskian, a positive Dirichlet eigenfunction supplied by the time component of the unit normal, and a Riccati comparison for the angular mode k=2k=2. A ground-state identity then yields positivity of every angular mode k≥2k\ge2.

Introduction

The critical catenoid is a basic example in the study of the Morse index of free boundary minimal surfaces. In the Euclidean unit ball in R3\mathbb{R}^{3}, its index is four, as proved by Devyver [3], Smith and Zhou [11], and Tran [12]; another proof was given by Medvedev [6]. Devyver also proved that every compact orientable free boundary minimal surface in this ball, other than a flat disk, has index at least four [3]. Thus the critical catenoid attains this lower bound.

The analogous index problem in curved space forms concerns critical spherical catenoids. In H3\mathbb{H}^{3}, these are rotational minimal annuli obtained by free boundary truncation of the spherical catenoids constructed by Mori [8]. The geometry of rotational hypersurfaces in spaces of constant curvature was developed further by do Carmo and Dajczer [1], and the free boundary truncation was considered by Li and Xiong [4]. Lima [5] proved that an embedded free boundary minimal annulus in a geodesic ball of S+3\mathbb{S}^{3}_{+} or H3\mathbb{H}^{3} that is invariant under the geodesic symmetry about the center of the ball is congruent to a critical rotational annulus; in H3\mathbb{H}^{3}, these are the critical spherical catenoids. Medvedev [7] proved that the critical spherical catenoid in a geodesic ball of S+3\mathbb{S}^{3}_{+} has index four, and that its hyperbolic counterpart has index at least four. He conjectured equality in the hyperbolic case. We prove this equality throughout the parameter range of the critical spherical catenoid family in H3\mathbb{H}^{3} and determine the nullity.

The analysis in [10] determines the index and nullity contributions of the angular mode ∣k∣=1|k|=1 for every parameter. The local result in [9] establishes index four and nullity two when the parameter is sufficiently close to its lower endpoint. The theorem below establishes both equalities for every a>12a>\frac{1}{2}.

Theorem 1. For each a>12a>\frac{1}{2}, let Σa\Sigma_a be the critical spherical catenoid with angular period 2π2\pi, parametrized in § 2, and let B(r(a))⊂H3\mathbb{B}(r(a))\subset\mathbb{H}^{3} be its supporting geodesic ball. Then

ind⁡(Σa)=4,nul⁡(Σa)=2.\operatorname{ind}(\Sigma_a)=4,\qquad\operatorname{nul}(\Sigma_a)=2.

The Robin kernel is generated by the normal components of the two rotations that tilt the axis of the catenoid while fixing the center of B(r(a))\mathbb{B}(r(a)).

2020 Mathematics Subject Classification. Primary 53A10; Secondary 53C42, 58J50, 34B24.

Key words and phrases. Free boundary minimal surfaces, hyperbolic space, critical catenoid, Morse index, nullity, Robin eigenvalue problem.

Here the index and nullity refer to the second variation of area for normal variations whose boundary remains on the supporting sphere. In particular, the ambient ball is fixed when taking variations.

The theorem supplies the global hyperbolic counterpart of the Euclidean index-four results and identifies all infinitesimal degeneracies with rotations of the ambient ball. At fixed ambient curvature −1-1, both the geometry of the annulus and its supporting radius vary with aa. The associated Jacobi operator contains the curvature term +2+2, and the Robin coefficient is coth⁡r(a)\coth r(a); these terms are recorded in Section 3. The proof controls this family of boundary value problems over the entire interval a>12a > \frac{1}{2}.

The proof has three main ingredients. A Wronskian identity gives a strict boundary inequality valid for every a>12a > \frac{1}{2}. The time component of the unit normal is a positive Dirichlet eigenfunction with eigenvalue two; a codimension-two min–max argument therefore determines the radial mode k=0k=0. Finally, a comparison of logarithmic derivatives, together with an explicit integral inequality, gives a boundary inequality for the mode k=2k=2; a ground-state identity turns it into positivity of that mode, and monotonicity in k2k^2 completes the spectral count. The geometric and spectral identities used below are included in the proof.

Two consequences are recorded in Section 8. The radius of the supporting ball is a strictly increasing function of aa (Corollary 9). The restriction to angular period 2π2\pi is essential: for m≥2m \ge2, the mm-fold cover of Σa\Sigma_a has index at least 4m4m, and its nullity exceeds two exactly when mk∗(a)mk_*(a) is an integer, for a threshold frequency k∗(a)∈(1,2)k_*(a) \in(1,2) (Remark 10). This agrees with the non-rotational free boundary minimal annuli of Cerezo [2], which bifurcate from multiple covers of critical spherical catenoids.

Geometry of the critical catenoids

We use the hyperboloid model

H3={X∈R3,1:⟨X,X⟩L=−1, X0>0},⟨X,Y⟩L=−X0Y0+∑j=13XjYj.\mathbb{H}^{3} = \{X \in\mathbb{R}^{3,1} : \langle X,X\rangle_{L} = -1,\ X^{0} > 0\}, \qquad\langle X,Y\rangle_{L} = -X^{0}Y^{0} + \sum_{j=1}^{3} X^{j}Y^{j}.

The center of the balls under consideration is e0=(1,0,0,0)e_{0}=(1,0,0,0). Fix a>12a > \frac{1}{2} and define

K=a2−14,A(s)2=acosh⁡(2s)+12,B(s)2=acosh⁡(2s)−12.(1)K = \sqrt{a^{2}-\frac{1}{4}}, \qquad A(s)^{2}=a\cosh(2s)+\frac{1}{2}, \qquad B(s)^{2}=a\cosh(2s)-\frac{1}{2}. \tag*{(1)}
φ(s)=∫0sKA(σ)2B(σ)2 dσ.(2)\varphi(s)=\int_{0}^{s}\frac{K}{A(\sigma)^{2}B(\sigma)^{2}}\,\mathrm{d}\sigma. \tag*{(2)}

The positive square roots are understood. The rotational parametrization is

X(s,θ)=(Acosh⁡φ, Asinh⁡φ, Bcos⁡θ, Bsin⁡θ),(s,θ)∈R×(R/2πZ).(3)X(s,\theta)=\left(A\cosh\varphi,\ A\sinh\varphi,\ B\cos\theta,\ B\sin\theta\right), \qquad(s,\theta)\in\mathbb{R}\times(\mathbb{R}/2\pi\mathbb{Z}). \tag*{(3)}

A prime denotes differentiation with respect to ss throughout the paper. Set v=B′v=B'. The identities

A2=1+B2,A′=BvA,v=asinh⁡(2s)B,(4)A^{2}=1+B^{2}, \qquad A'=\frac{Bv}{A}, \qquad v=\frac{a\sinh(2s)}{B}, \tag*{(4)}
v2=1+B2−K2B2,v′=B+K2B3>0(5)v^{2}=1+B^{2}-\frac{K^{2}}{B^{2}}, \qquad v'=B+\frac{K^{2}}{B^{3}}>0 \tag*{(5)}

follow directly from (2.1). In particular, AA and BB are even, whereas vv and φ\varphi are odd. For completeness, introduce the Lorentz-orthonormal frame

E0=(cosh⁡φ,sinh⁡φ,0,0),E1=(sinh⁡φ,cosh⁡φ,0,0),Er=(0,0,cos⁡θ,sin⁡θ),E_{0}=(\cosh\varphi,\sinh\varphi,0,0), \qquad E_{1}=(\sinh\varphi,\cosh\varphi,0,0), \qquad E_{r}=(0,0,\cos\theta,\sin\theta),

together with Eθ=(0,0,−sin⁡θ,cos⁡θ)E_{\theta}=(0,0,-\sin\theta,\cos\theta). Then X=AE0+BErX=AE_{0}+BE_{r} and

Xs=A′E0+Aφ′E1+vEr,Xθ=BEθ.X_{s}=A'E_{0}+A\varphi'E_{1}+vE_{r}, \qquad X_{\theta}=BE_{\theta}.

Using (2.2)–(2.5), we obtain

⟨Xs,Xs⟩L=v2+K2/B2A2=1,⟨Xs,Xθ⟩L=0.\langle X_s,X_s\rangle_L=\frac{v^2+K^2/B^2}{A^2}=1,\qquad\langle X_s,X_\theta\rangle_L=0.

Consequently,

g=ds2+B2dθ2.(6)g=ds^2+B^2d\theta^2. \tag*{(6)}

A unit normal is

ν=KAE0−vAE1+KBEr.(7)\nu=\frac{K}{A}E_0-\frac{v}{A}E_1+\frac{K}{B}E_r. \tag*{(7)}

Indeed, (2.4) and (2.5) imply that ν\nu is orthogonal to X,Xs,XθX,X_s,X_\theta and has Lorentzian norm one. A direct computation using (2.2)–(2.5) gives

νs=−KB2Xs,νθ=KB2Xθ.\nu_s=-\frac{K}{B^2}X_s,\qquad\nu_\theta=\frac{K}{B^2}X_\theta.

Thus, with shape operator −dν-d\nu, the principal curvatures are K/B2K/B^2 and −K/B2-K/B^2. The surface is minimal and

∣II∣2=2K2B4.(8)\lvert II\rvert^2=\frac{2K^2}{B^4}. \tag*{(8)}

The parametrization is embedded: the quotient X1/X0=tanh⁡φ(s)X^1/X^0=\tanh\varphi(s) is strictly increasing in ss, and, for each fixed ss, the pair (X2,X3)=B(s)(cos⁡θ,sin⁡θ)(X^2,X^3)=B(s)(\cos\theta,\sin\theta) determines θ\theta uniquely modulo 2π2\pi because B(s)>0B(s)>0.

Free boundary truncation

Write

g0(s)=A(s)cosh⁡φ(s)=X0,g1(s)=A(s)sinh⁡φ(s)=X1.g_0(s)=A(s)\cosh\varphi(s)=X^0,\qquad g_1(s)=A(s)\sinh\varphi(s)=X^1.

For the distance ρ\rho from e0e_0, one has X0=cosh⁡ρX^0=\cosh\rho and

∇H3X0=X0X−e0,⟨∇H3X0,ν⟩L=ν0.\nabla^{\mathbb{H}^3}X^0=X^0X-e_0,\qquad\langle\nabla^{\mathbb{H}^3}X^0,\nu\rangle_L=\nu^0.

It follows that the surface meets a sphere centered at e0e_0 orthogonally precisely when ν0=0\nu^0=0. At a point s=S>0s=S>0, this condition reads

tanh⁡φ(S)=Kv(S).(9)\tanh\varphi(S)=\frac{K}{v(S)}. \tag*{(9)}

As functions of S>0S>0, the left-hand side increases strictly from zero, whereas the right-hand side decreases strictly from +∞+\infty to zero. Hence (9) has exactly one solution S>0S>0.

The critical catenoid is

Σa=X([−S,S]×(R/2πZ)).\Sigma_a=X\bigl([-S,S]\times(\mathbb{R}/2\pi\mathbb{Z})\bigr).

It is symmetric with respect to the totally geodesic plane {X1=0}\{X^1=0\}, which contains the neck circle s=0s=0, and its supporting ball is centered at the point e0e_0 where this plane meets the axis {X2=X3=0}\{X^2=X^3=0\}. Set

B0=B(S),v0=v(S),r=arcosh⁡g0(S),c=coth⁡r.(10)B_0=B(S),\qquad v_0=v(S),\qquad r=\operatorname{arcosh}g_0(S),\qquad c=\coth r. \tag*{(10)}

Since g0′(s)>0g_0'(s)>0 for s>0s>0, every interior point of Σa\Sigma_a lies strictly inside B(r)\mathbb{B}(r). At the positive boundary the outward unit conormal is XsX_s. Orthogonality to the sphere gives

Xs(S,θ)=cosh⁡rX(S,θ)−e0sinh⁡r.X_s(S,\theta)=\frac{\cosh r X(S,\theta)-e_0}{\sinh r}.

Taking components yields

g0(S)=cosh⁡r,g0′(S)=sinh⁡r,g1′(S)=cg1(S).(11)g_0(S)=\cosh r,\qquad g_0'(S)=\sinh r,\qquad g_1'(S)=c g_1(S). \tag*{(11)}
v0=cB0,g0′(S)−cg0(S)=−1sinh⁡r.(12)v_0=cB_0,\qquad g_0'(S)-c g_0(S)=-\frac{1}{\sinh r}. \tag*{(12)}

The quadratic form and separation of variables

We use the convention Δ=div⁡∇\Delta= \operatorname{div}\nabla. For the ambient Ricci tensor, Ric⁡H3(ν,ν)=−2\operatorname{Ric}_{\mathbb{H}^3}(\nu,\nu)=-2. If NN is the outward unit normal to the supporting sphere, then ⟨∇VH3N,V⟩L=coth⁡r\langle\nabla^{\mathbb{H}^3}_V N,V\rangle_L=\coth r for every unit vector VV tangent to that sphere. Along the free boundary, ν\nu is such a tangent vector. The general free boundary second variation formula [12 Section 2] therefore gives

Q[u]=∫Σa(∣∇u∣2+(2−∣II∣2)u2) dA−c∫∂Σau2 dℓ,u∈H1(Σa).(13)Q[u]=\int_{\Sigma_a}\left(|\nabla u|^2+(2-|\mathrm{II}|^2)u^2\right)\,\mathrm{d}A-c\int_{\partial\Sigma_a}u^2\,\mathrm{d}\ell,\qquad u\in H^1(\Sigma_a). \tag*{(13)}

Its associated self-adjoint operator and boundary condition are

J=−Δ−∣II∣2+2,∂ηu=cu,J=-\Delta-|\mathrm{II}|^2+2,\qquad\partial_\eta u=cu,

where η\eta is the outward unit conormal along ∂Σa\partial\Sigma_a. The index is the number of negative eigenvalues, counted with multiplicity; the nullity is the dimension of the kernel. The form domain is H1(Σa)H^1(\Sigma_a) and imposes no Robin condition on test functions.

For angular frequency k≥0k\geq0, the radial operator is

Jku=−1B(Bu′)′+(2+k2B2−2K2B4)u(14)J_k u=-\frac{1}{B}(Bu')'+\left(2+\frac{k^2}{B^2}-\frac{2K^2}{B^4}\right)u \tag*{(14)}

in L2((−S,S),B ds)L^2((-S,S),B\,\mathrm{d}s), with boundary conditions

u′(S)=cu(S),−u′(−S)=cu(−S).(15)u'(S)=cu(S),\qquad-u'(-S)=cu(-S). \tag*{(15)}

After removing the positive angular integration factor, the radial form is

qk[u]=∫−SS(Bu′2+(2B+k2B−2K2B3)u2) ds−v0(u(S)2+u(−S)2),u∈H1(−S,S).(16)q_k[u]=\int_{-S}^{S}\left(Bu'^2+\left(2B+\frac{k^2}{B}-\frac{2K^2}{B^3}\right)u^2\right)\,\mathrm{d}s-v_0\left(u(S)^2+u(-S)^2\right),\qquad u\in H^1(-S,S). \tag*{(16)}

Each k≥1k\geq1 has the two real angular factors cos⁡(kθ)\cos(k\theta) and sin⁡(kθ)\sin(k\theta); the mode k=0k=0 has one. The coefficients are even in ss, so each radial problem splits into even and odd sectors.

For each fixed a>12a>\frac{1}{2}, one has B(s)≥a−12>0B(s)\geq\sqrt{a-\frac{1}{2}}>0 on the compact interval [−S,S][-S,S]. All radial problems are therefore regular Sturm–Liouville problems with positive smooth weight. We use the min–max principle, simplicity of radial eigenvalues, and the oscillation theorem: the jjth eigenfunction, with eigenvalues numbered starting at j=1j=1, has exactly j−1j-1 interior zeros. These standard facts are discussed in [13].

A global boundary inequality

For the flat connection DD of R3,1\mathbb{R}^{3,1}, the Gauss formula reads

DU(dX(V))=dX(∇UΣV)+II(U,V)ν+g(U,V)X.D_U(\mathrm{d}X(V))=\mathrm{d}X(\nabla^\Sigma_U V)+\mathrm{II}(U,V)\nu+g(U,V)X.

Tracing and using minimality gives ΔΣX=2X\Delta_\Sigma X=2X. The coordinate functions consequently satisfy ΔΣXj=2Xj\Delta_\Sigma X^j=2X^j, so

gi′′+B′Bgi′−2gi=0,J0gi=−∣II∣2gi,i=0,1.(17)g_i''+\frac{B'}{B}g_i'-2g_i=0,\qquad J_0g_i=-|\mathrm{II}|^2g_i,\qquad i=0,1. \tag*{(17)}

Lemma 2. For every a>12a>\frac{1}{2}, the boundary radius in the rotational coordinates satisfies

B02>2K2.B_0^2>2K^2.

Proof. By (5),

v′′=v−3K2vB4,v''=v-\frac{3K^2v}{B^4},

and substitution in (14) gives J0v=0J_0v=0. Moreover, v(0)=0v(0)=0 and v>0v>0 on (0,S](0,S]. Define

W=B(vg1′−v′g1).W=B(vg_1'-v'g_1).

The equations for vv and g1g_1 imply

W(0)=0,W′=B∣II∣2vg1>0on (0,S].W(0)=0,\qquad W'=B\lvert\mathrm{II}\rvert^2vg_1>0\quad\text{on }(0,S].

Using (11), we find

0<W(S)=−B0g1(S)(v′(S)−cv0).0<W(S)=-B_0g_1(S)\bigl(v'(S)-cv_0\bigr).

On the other hand, (5) and c=v0/B0c=v_0/B_0 give

v′(S)−cv0=B0+K2B03−v02B0=2K2−B02B03.v'(S)-cv_0=B_0+\frac{K^2}{B_0^3}-\frac{v_0^2}{B_0}=\frac{2K^2-B_0^2}{B_0^3}.

Since B0g1(S)>0B_0g_1(S)>0, the asserted inequality follows. □

The modes zero and one

A Dirichlet eigenfunction

Lemma 3. The first Dirichlet eigenvalue of J0J_0 on (−S,S)(-S,S) is 22. A corresponding positive eigenfunction is

n(s)=ν0(s)=Kcosh⁡φ(s)−v(s)sinh⁡φ(s)A(s).n(s)=\nu^0(s)=\frac{K\cosh\varphi(s)-v(s)\sinh\varphi(s)}{A(s)}.

Proof. The Gauss–Weingarten equations in R3,1\mathbb{R}^{3,1} imply

ΔΣν=−∣II∣2ν(18)\Delta_\Sigma\nu=-\lvert\mathrm{II}\rvert^2\nu \tag*{(18)}

for a minimal surface in H3\mathbb{H}^3. To see the signs explicitly, let AA denote the shape operator. In a geodesic orthonormal frame on the surface, tracing the derivative of Dν=−dX∘AD\nu=-dX\circ A gives

ΔΣν=−dX(div⁡A)−∣II∣2ν−(tr⁡A)X.\Delta_\Sigma\nu=-dX(\operatorname{div}A)-\lvert\mathrm{II}\rvert^2\nu-(\operatorname{tr}A)X.

The Codazzi equation in constant curvature gives div⁡A=∇(tr⁡A)\operatorname{div}A=\nabla(\operatorname{tr}A), and both trace terms vanish by minimality. This proves (18). Taking its time component gives J0n=2nJ_0n=2n.

The function nn is even and n(0)=K/A(0)>0n(0)=K/A(0)>0. For 0<s<S0<s<S, the uniqueness of the crossing in (9) gives tanh⁡φ(s)<K/v(s)\tanh\varphi(s)<K/v(s), so n(s)>0n(s)>0. At s=±Ss=\pm S, the free boundary condition gives n=0n=0. Thus nn is a Dirichlet eigenfunction with no interior zero. The oscillation theorem identifies its eigenvalue as the first one. □

:::

Proposition 4. The radial Robin problem for J0J_0 has exactly two negative eigenvalues and no zero eigenvalue. One negative eigenfunction is even and the other is odd.

Proof. Integrating (17) by parts in (16), and using (11)–(12), gives

q0[g0]=−∫−SSB∣II∣2g02 ds−2B0coth⁡r<0,(19)q_0[g_0]=-\int_{-S}^{S}B\lvert\mathrm{II}\rvert^2g_0^2\,\mathrm{d}s-2B_0\coth r<0, \tag*{(19)}
q0[g1]=−∫−SSB∣II∣2g12 ds<0.(20)q_0[g_1]=-\int_{-S}^{S}B\lvert\mathrm{II}\rvert^2g_1^2\,\mathrm{d}s<0. \tag*{(20)}

Since g0g_0 is even and g1g_1 is odd, their mixed term in the polarized quadratic form is zero. Hence their span is a two-dimensional negative subspace.

Let λ1<λ2<λ3<⋯\lambda_1<\lambda_2<\lambda_3<\cdots be the radial Robin eigenvalues. The preceding subspace gives λ2<0\lambda_2<0. The Dirichlet form domain H01(−S,S)H_0^1(-S,S) has codimension two in H1(−S,S)H^1(-S,S), since its defining conditions are the two endpoint traces. Therefore the span of the first three Robin eigenfunctions contains a nonzero z∈H01(−S,S)z\in H_0^1(-S,S). Lemma 3 and the min–max principle yield

2∫−SSBz2 ds≤q0[z]≤λ3∫−SSBz2 ds.2\int_{-S}^{S}Bz^2\,\mathrm{d}s\le q_0[z]\le\lambda_3\int_{-S}^{S}Bz^2\,\mathrm{d}s.

Thus λ3≥2\lambda_3\ge2. This proves the exact count and excludes a kernel. The negative tests (5.2) and (5.3) supply one negative direction of each parity, so each parity contributes exactly one negative eigenvalue. □

The first angular mode

Proposition 5. The radial Robin problem for J1J_1 has one negative eigenvalue and a one-dimensional kernel generated by g0′g'_0. The negative eigenfunction is even and the kernel function is odd.

Proof. The identity

(B′B)′=2K2B4−1B2\left(\frac{B'}{B}\right)'=\frac{2K^2}{B^4}-\frac{1}{B^2}

shows, upon differentiating (17), that

J1g0′=0,J1g1′=0.(21)J_1g'_0=0,\qquad J_1g'_1=0. \tag*{(21)}

Set f=g0′f=g'_0. The function ff is odd, positive on (0,S](0,S], and satisfies f′(0)=2g0(0)>0f'(0)=2g_0(0)>0. At s=Ss=S,

f′(S)=2g0(S)−cg0′(S)=2cosh⁡r−csinh⁡r=cosh⁡r=cf(S).f'(S)=2g_0(S)-cg'_0(S)=2\cosh r-c\sinh r=\cosh r=cf(S).

Parity gives the other boundary condition in (15). Thus ff is a Robin eigenfunction with eigenvalue zero and exactly one interior zero. By the oscillation theorem, zero is the second radial eigenvalue, and it is simple. There is exactly one negative eigenvalue.

A first eigenfunction can be chosen strictly positive. Reflection s↦−ss\mapsto-s preserves the problem, so simplicity makes that eigenfunction even. The kernel generator ff is odd. □

Positivity of the higher angular modes

Define the conformal coordinate

t=t(s)=∫0sdσB(σ),T=t(S).t=t(s)=\int_0^s\frac{d\sigma}{B(\sigma)},\qquad T=t(S).

A dot denotes differentiation with respect to tt; primes continue to denote differentiation with respect to ss. We use the same letters for functions composed with s(t)s(t). Equation (5) gives

B˙=Bv,v˙=D(v):=(v2−1)2+4K2,v(0)=0,D(0)=2a.(22)\dot B=Bv,\qquad\dot v=D(v):=\sqrt{(v^2-1)^2+4K^2},\qquad v(0)=0,\qquad D(0)=2a. \tag*{(22)}

Indeed, v˙=Bv′=B2+K2/B2>0\dot v=Bv'=B^2+K^2/B^2>0, and the square of this expression is (v2−1)2+4K2(v^2-1)^2+4K^2.

In this coordinate the zero-eigenvalue equation is

u¨=(2v2+k2−2)u.(23)\ddot u=(2v^2+k^2-2)u. \tag*{(23)}

and the full radial form is

qk[u]=∫−TT(u˙2+(2v2+k2−2)u2) dt−v0(u(T)2+u(−T)2).(24)q_k[u]=\int_{-T}^{T}\left(\dot u^2+(2v^2+k^2-2)u^2\right)\,dt-v_0\left(u(T)^2+u(-T)^2\right). \tag*{(24)}

The norm in the spectral problem is

∫−SSBu2 ds=∫−TTB2u2 dt.(25)\int_{-S}^{S}Bu^2\,ds=\int_{-T}^{T}B^2u^2\,dt. \tag*{(25)}

In particular, the conformal change preserves the form and the zero equation; the transformed eigenvalue problem has weight B2B^2.

A boundary logarithmic derivative

Let u2u_2 solve

u¨2=(2v2+2)u2,u2(0)=1,u˙2(0)=0.(26)\ddot u_2=(2v^2+2)u_2,\qquad u_2(0)=1,\qquad\dot u_2(0)=0. \tag*{(26)}

The coefficient is strictly positive. While u2u_2 is positive, u¨2>0\ddot u_2>0, so u2u_2 increases from 1 and cannot have a first zero. Hence u2>0u_2>0 on [0,T][0,T] and u˙2>0\dot u_2>0 on (0,T](0,T]. Define

p2=u˙2u2,δ2=p2−v.p_2=\frac{\dot u_2}{u_2},\qquad\delta_2=p_2-v.

Lemma 6. For every a>12a>\frac{1}{2}, one has p2(T)>v0p_2(T)>v_0.

Proof. The Riccati equation for p2p_2 and (22) give

δ˙2+(p2+v)δ2=F2(v),δ2(0)=0,F2(v)=v2+2−D(v).(27)\dot{\delta}_2+(p_2+v)\delta_2=F_2(v),\qquad\delta_2(0)=0,\qquad F_2(v)=v^2+2-\mathcal{D}(v). \tag*{(27)}

Since (v2+2)2−D(v)2=6v2+4−4a2(v^2+2)^2-\mathcal{D}(v)^2=6v^2+4-4a^2, we have the exact identity

F2(v)=6v2+4−4a2v2+2+D(v).(28)F_2(v)=\frac{6v^2+4-4a^2}{v^2+2+\mathcal{D}(v)}. \tag*{(28)}

If a≤1a\leq1, then F2(v(t))>0F_2(v(t))>0 for every t>0t>0. An integrating factor in (27) yields δ2(T)>0\delta_2(T)>0.

Suppose now that a>1a>1. By (28), F2(v)F_2(v) is negative for v2<2(a2−1)/3v^2<2(a^2-1)/3 and positive for v2>2(a2−1)/3v^2>2(a^2-1)/3. Since vv is strictly increasing, the function

M(t)δ2(t)=∫0tM(τ)F2(v(τ)) dτ,M(t)=exp⁡(∫0t(p2+v) dτ)>0,M(t)\delta_2(t)=\int_0^t M(\tau)F_2(v(\tau))\,\mathrm{d}\tau,\qquad M(t)=\exp\left(\int_0^t(p_2+v)\,\mathrm{d}\tau\right)>0,

first decreases and then increases, if the sign change is reached. It follows that

δ2(T)≤0⟹δ2(t)<0(0<t<T).(29)\delta_2(T)\leq0\quad\Longrightarrow\quad\delta_2(t)<0\quad(0<t<T). \tag*{(29)}

If the sign change is not reached before TT, the same implication follows from strict decrease throughout the interval.

For a comparison solution, set w=g1′w=g_1'. From the definitions,

w=A′sinh⁡ϕ+Aϕ′cosh⁡ϕ>0(s≥0).w=A'\sinh\phi+A\phi'\cosh\phi>0\qquad(s\geq0).

The function ww is even, and J1w=0J_1w=0 by (21). Consequently, for

p1=w˙w,δ1=p1−v,p_1=\frac{\dot{w}}{w},\qquad\delta_1=p_1-v,

we have

p˙1=2v2−1−p12,p1(0)=0,\dot{p}_1=2v^2-1-p_1^2,\qquad p_1(0)=0,

and

δ˙1+(p1+v)δ1=v2−1−D(v)<0,δ1(0)=0.\dot{\delta}_1+(p_1+v)\delta_1=v^2-1-\mathcal{D}(v)<0,\qquad\delta_1(0)=0.

The strict sign holds because K>0K>0, so D(v)>∣v2−1∣\mathcal{D}(v)>|v^2-1|. Thus p1<vp_1<v for t>0t>0.

Assume, for contradiction, that δ2(T)≤0\delta_2(T)\leq0. By (29), p1+p2<2vp_1+p_2<2v on (0,T)(0,T). The difference ω=p2−p1\omega=p_2-p_1 satisfies

ω˙+(p1+p2)ω=3,ω(0)=0.\dot{\omega}+(p_1+p_2)\omega=3,\qquad\omega(0)=0.

Therefore

ω(T)=3∫0Texp⁡(−∫τT(p1+p2) dt) dτ>3B02∫0TB(t)2 dt=3B02∫0SB(s) ds.(30)\begin{aligned} \omega(T)=3\int_0^T\exp\left(-\int_\tau^T(p_1+p_2)\,\mathrm{d}t\right)\,\mathrm{d}\tau\\ &> \frac{3}{B_0^2}\int_0^T B(t)^2\,\mathrm{d}t=\frac{3}{B_0^2}\int_0^S B(s)\,\mathrm{d}s. \tag*{(30)} \end{aligned}

Here we used B˙/B=v\dot{B}/B=v; the strict inequality holds on every integration interval with τ<T\tau<T.

At the boundary, (17) and g1′(S)=(v0/B0)g1(S)g_1'(S)=(v_0/B_0)g_1(S) give

p1(T)=B0g1′′(S)g1′(S)=2B02v0−v0.p_1(T)=\frac{B_0g_1''(S)}{g_1'(S)}=\frac{2B_0^2}{v_0}-v_0.

Since p2(T)≤v0p_2(T)\leq v_0, equation (5) yields

ω(T)≤v0−p1(T)=2(B02−K2)B02v0.(31)\omega(T)\leq v_0-p_1(T)=\frac{2(B_0^2-K^2)}{B_0^2v_0}. \tag*{(31)}

We now prove that (30) and (31) are incompatible. For s>0s>0, define

I(s)=3∫0sB(σ) dσ−2(B(s)2−K2)v(s).I(s)=3\int_0^s B(\sigma)\,\mathrm{d}\sigma-\frac{2(B(s)^2-K^2)}{v(s)}.

Let x=B2x=B^2 and κ=K2\kappa=K^2. Direct differentiation gives

I′=−B+2(B2−κ)(B+κ/B3)v2=B(−1+2(x−κ)(x2+κ)x(x2+x−κ))=B(x−2κ)(x2−x+κ)x(x2+x−κ).(32)\begin{aligned} \mathcal{I}'&=-B+\frac{2(B^2-\kappa)(B+\kappa/B^3)}{v^2}\\ &=B\left(-1+\frac{2(x-\kappa)(x^2+\kappa)}{x(x^2+x-\kappa)}\right)\\ &=B\frac{(x-2\kappa)(x^2-x+\kappa)}{x(x^2+x-\kappa)}. \tag*{(32)} \end{aligned}

For the last step, the numerator expands as

−x(x2+x−κ)+2(x−κ)(x2+κ)=x3−(1+2κ)x2+3κx−2κ2=(x−2κ)(x2−x+κ).\begin{aligned} -x(x^2+x-\kappa)+2(x-\kappa)(x^2+\kappa) &=x^3-(1+2\kappa)x^2+3\kappa x-2\kappa^2\\ &=(x-2\kappa)(x^2-x+\kappa). \end{aligned}

The denominator in (32) is positive for s>0s>0, because x2+x−κ=xv2>0x^2+x-\kappa=xv^2>0. Since a>1a>1, we have κ>34\kappa>\frac{3}{4}, and

x2−x+κ=(x−12)2+κ−14>0.x^2-x+\kappa=\left(x-\frac{1}{2}\right)^2+\kappa-\frac{1}{4}>0.

Set

s∗=12arcosh⁡(2a),B(s∗)=2K,v(s∗)=2a.(33)s_*=\frac{1}{2}\operatorname{arcosh}(2a),\qquad B(s_*)=\sqrt{2}K,\qquad v(s_*)=\sqrt{2}a. \tag*{(33)}

Lemma 2 implies S>s∗S>s_*. Equation (32) shows that I\mathcal{I} is strictly increasing on (s∗,S](s_*,S].

The identity

(B′B)2=1+1x−K2x2=a2K2−(x−2K2)24K2x2≤a2K2(34)\left(\frac{B'}{B}\right)^2=1+\frac{1}{x}-\frac{K^2}{x^2}=\frac{a^2}{K^2}-\frac{(x-2K^2)^2}{4K^2x^2}\leq\frac{a^2}{K^2} \tag*{(34)}

follows by expanding the square and using a2=K2+14a^2=K^2+\frac{1}{4}. As B′≥0B'\geq0 for s≥0s\geq0, integration of B′≤(a/K)BB'\leq(a/K)B gives

∫0s∗B ds≥Ka(2K−a−12).\int_0^{s_*}B\,ds\geq\frac{K}{a}\left(\sqrt{2}K-\sqrt{a-\frac{1}{2}}\right).

Using (33), we conclude that

I(s∗)≥Ka(22K−3a−12)>0.(35)\mathcal{I}(s_*)\geq\frac{K}{a}\left(2\sqrt{2}K-3\sqrt{a-\frac{1}{2}}\right)>0. \tag*{(35)}

The final strict inequality follows from

8K2−9(a−12)=(a−12)(8a−5)>0.8K^2-9\left(a-\frac{1}{2}\right)=\left(a-\frac{1}{2}\right)(8a-5)>0.

Thus I(S)>0\mathcal{I}(S)>0. In contrast, (30)–(31) imply

3∫0SB ds<2(B02−K2)v0,3\int_0^S B\,ds<\frac{2(B_0^2-K^2)}{v_0},

which says I(S)<0\mathcal{I}(S)<0. This contradiction proves δ2(T)>0\delta_2(T)>0 also when a>1a>1.

A positive quadratic form

Proposition 7. Every radial Robin eigenvalue in angular mode k≥2k\geq2 is strictly positive.

Proof. Extend u2u_2 evenly to [−T,T][-T,T]. The coefficient in (26) is even, so the extension is smooth, positive, and solves the same equation. Its logarithmic derivative p2p_2 is odd. For any z∈H1(−T,T)z\in H^1(-T,T), writing z=u2hz=u_2h and integrating

z˙2+(2v2+2)z2=u22h˙2+ddt(p2z2)(36)\dot{z}^2+(2v^2+2)z^2=u_2^2\dot{h}^2+\frac{d}{dt}(p_2z^2) \tag*{(36)}

gives

q2[z]=∫−TTu22(dzdu2)2 dt+(p2(T)−v0)(z(T)2+z(−T)2).(37)q_2[z] = \int_{-T}^{T} u_2^2\left(\frac{\mathrm{d}z}{\mathrm{d}u_2}\right)^2 \,\mathrm{d}t + (p_2(T)-v_0)\left(z(T)^2+z(-T)^2\right). \tag*{(37)}

The identity first follows for smooth functions and extends to H1H^1 by density and continuity of endpoint traces. Lemma 6 makes the boundary coefficient strictly positive. If the right-hand side vanishes, then z/u2z/u_2 is constant and z(T)=z(−T)=0z(T)=z(-T)=0, which forces z=0z=0. Therefore q2[z]>0q_2[z]>0 for every nonzero z∈H1(−T,T)z\in H^1(-T,T).

The minimum of the Rayleigh quotient with the positive weight B2 dtB^2\,\mathrm{d}t in (25) is attained for this regular problem. Hence its first eigenvalue is strictly positive. Finally,

qk[z]=q2[z]+(k2−4)∫−TTz2 dt(k≥2)q_k[z] = q_2[z] + (k^2-4)\int_{-T}^{T} z^2\,\mathrm{d}t \qquad(k\ge2)

proves the assertion for all higher modes, on the full form domain and therefore in both parity sectors. □

Proof of the main theorem

Proof of Theorem 1. Propositions 4, 5, and 7 give the following contributions. Angular multiplicities are included in the table.

SectorIndex contributionNullity contribution
k=0k=0, even10
k=0k=0, odd10
k=1k=1, even20
k=1k=1, odd02
k≥2k\ge2, both parities00

Table 1.

The angular Fourier decomposition is orthogonal for both the L2L^2 inner product and the quadratic form. Its partial sums converge in H1(Σa)H^1(\Sigma_a), and hence their boundary traces converge in L2(∂Σa)L^2(\partial\Sigma_a). Thus the decomposition also holds on the full form domain. The span of the negative eigenfunctions in the table has dimension four, and its L2L^2-orthogonal complement is nonnegative for QQ. A function in the Robin kernel has each Fourier coefficient in the corresponding radial kernel. The mode counts therefore give

ind⁡(Σa)=4,ker⁡J=span⁡{g0′(s)cos⁡θ, g0′(s)sin⁡θ}.\operatorname{ind}(\Sigma_a)=4,\qquad\ker J=\operatorname{span}\{g'_0(s)\cos\theta,\ g'_0(s)\sin\theta\}.

In particular, nul⁡(Σa)=2\operatorname{nul}(\Sigma_a)=2.

To identify the kernel geometrically, consider the spatial rotation fields

Z2(X)=(0,−X2,X1,0),Z3(X)=(0,−X3,0,X1).Z_2(X)=(0,-X^2,X^1,0),\qquad Z_3(X)=(0,-X^3,0,X^1).

They preserve the center e0e_0 and the geodesic ball. Equation (7) gives

⟨Z2,ν⟩L=g0′cos⁡θ,⟨Z3,ν⟩L=g0′sin⁡θ.\langle Z_2,\nu\rangle_L=g'_0\cos\theta,\qquad\langle Z_3,\nu\rangle_L=g'_0\sin\theta.

Indeed, ⟨Z2,ν⟩L=(−Bν1+(K/B)g1)cos⁡θ\langle Z_2,\nu\rangle_L=(-B\nu^1+(K/B)g_1)\cos\theta and ⟨Z3,ν⟩L=(−Bν1+(K/B)g1)sin⁡θ\langle Z_3,\nu\rangle_L=(-B\nu^1+(K/B)g_1)\sin\theta, while

−Bν1+KBg1=BνAcosh⁡φ+KABsinh⁡φ=g0′.-B\nu^1+\frac{K}{B}g_1=\frac{B\nu}{A}\cosh\varphi+\frac{K}{AB}\sinh\varphi=g'_0.

These rotations tilt the axis of the catenoid. Rotation about the axis itself has zero normal component. The two displayed fields span the entire Robin kernel, as asserted. □

Remark 8. For each fixed a>12a>\frac{1}{2}, the theorem gives nondegeneracy modulo the rotations preserving the supporting ball. All spectral comparisons above are made for that fixed parameter; no uniform spectral gap as the parameter approaches an endpoint is required.

Two consequences

Corollary 9. The supporting radius r(a)r(a) depends smoothly on aa and satisfies

drda>0for every a>12.\frac{\mathrm{d}r}{\mathrm{d}a}>0 \qquad\text{for every }a>\frac{1}{2}.

In particular, distinct members of the family {Σa}a>1/2\{\Sigma_a\}_{a>1/2} are free boundary in distinct geodesic balls centered at e0e_0.

Proof. For s>0s>0, the function tanh⁡φ−K/v\tanh\varphi-K/v has ss-derivative φ′sech⁡2φ+Kv′/v2>0\varphi'\operatorname{sech}^2\varphi+Kv'/v^2>0. By the implicit function theorem, the solution SS of (9) depends smoothly on aa, and so does r=arcosh⁡g0(S)r=\operatorname{arcosh}g_0(S), since g0(S)>1g_0(S)>1.

All quantities of § 2 depend smoothly on (s,a)(s,a). Let Y=∂aXY=\partial_aX, taken at fixed (s,θ)(s,\theta), and set ψ=⟨Y,ν⟩L\psi=\langle Y,\nu\rangle_L. This is the parametric Jacobi field of [9 Definition 8.1]; its evenness, the value ψ(0)=1/(2K)\psi(0)=1/(2K) and the boundary identity (8.1) below are established in [9 Lemmas 8.2 and 8.3], and we include the short derivations for completeness. Each X(⋅,⋅ ;a)X(\cdot,\cdot\,;a) is a minimal immersion of R×(R/2πZ)\mathbb{R}\times(\mathbb{R}/2\pi\mathbb{Z}), so the normal component of the variation field is a Jacobi field. Since ψ\psi does not depend on θ\theta, this gives J0ψ=0J_0\psi=0. The Lorentz reflection R=diag⁡(1,−1,1,1)R=\operatorname{diag}(1,-1,1,1) satisfies X(−s,θ)=RX(s,θ)X(-s,\theta)=RX(s,\theta) and ν(−s,θ)=Rν(s,θ)\nu(-s,\theta)=R\nu(s,\theta) for every aa; hence ψ\psi is even.

At s=0s=0 one has φ=∂aφ=0\varphi=\partial_a\varphi=0 and v=0v=0. Since ∂aA2=∂aB2=cosh⁡(2s)\partial_aA^2=\partial_aB^2=\cosh(2s), the vectors YY and ν\nu at s=0s=0 are

(12A,0,cos⁡θ2B,sin⁡θ2B)and(KA,0,KBcos⁡θ,KBsin⁡θ),\left(\frac{1}{2A},0,\frac{\cos\theta}{2B},\frac{\sin\theta}{2B}\right) \quad\text{and}\quad \left(\frac{K}{A},0,\frac{K}{B}\cos\theta,\frac{K}{B}\sin\theta\right),

with A=A(0)A=A(0) and B=B(0)B=B(0). Using A(0)2−B(0)2=1A(0)^2-B(0)^2=1 and A(0)2B(0)2=K2A(0)^2B(0)^2=K^2, we obtain

ψ(0)=K2(1B(0)2−1A(0)2)=12K≠0.\psi(0)=\frac{K}{2}\left(\frac{1}{B(0)^2}-\frac{1}{A(0)^2}\right)=\frac{1}{2K}\ne0.

Now let s=Ss=S. The boundary relation of § 2 gives e0=cosh⁡r X−sinh⁡r Xse_0=\cosh r\,X-\sinh r\,X_s; in particular ⟨Xs,e0⟩L=−sinh⁡r\langle X_s,e_0\rangle_L=-\sinh r. Differentiating ⟨X,X⟩L=−1\langle X,X\rangle_L=-1, ⟨ν,X⟩L=0\langle\nu,X\rangle_L=0 and ⟨ν,Xs⟩L=0\langle\nu,X_s\rangle_L=0 in aa, and using νs=−(K/B2)Xs\nu_s=-(K/B^2)X_s, we find

⟨Y,X⟩L=0,⟨∂aν,X⟩L=−ψ,⟨∂aν,Xs⟩L=−ψ′−KB2⟨Y,Xs⟩L.\langle Y,X\rangle_L=0,\qquad \langle\partial_a\nu,X\rangle_L=-\psi,\qquad \langle\partial_a\nu,X_s\rangle_L=-\psi'-\frac{K}{B^2}\langle Y,X_s\rangle_L.

Since ∂ag0=−⟨Y,e0⟩L=sinh⁡r⟨Y,Xs⟩L\partial_ag_0=-\langle Y,e_0\rangle_L=\sinh r\langle Y,X_s\rangle_L at s=Ss=S and g0′(S)=sinh⁡rg_0'(S)=\sinh r, differentiating g0(S(a))=cosh⁡r(a)g_0(S(a))=\cosh r(a) and dividing by sinh⁡r\sinh r gives

dSda+⟨Y,Xs⟩L=drda.\frac{\mathrm{d}S}{\mathrm{d}a}+\langle Y,X_s\rangle_L=\frac{\mathrm{d}r}{\mathrm{d}a}.

Since ⟨νs,e0⟩L=(K/B02)sinh⁡r\langle\nu_s,e_0\rangle_L=(K/B_0^2)\sinh r at s=Ss=S, differentiating the free boundary condition ⟨ν,e0⟩L∣s=S(a)=−ν0(S(a))=0\left.\langle\nu,e_0\rangle_L\right|_{s=S(a)}=-\nu^0(S(a))=0, and using the decomposition of e0e_0 together with the identities above, gives

KB02sinh⁡rdSda−cosh⁡r ψ(S)+sinh⁡r(ψ′(S)+KB02⟨Y,Xs⟩L)=0.\frac{K}{B_0^2}\sinh r\frac{\mathrm{d}S}{\mathrm{d}a}-\cosh r\,\psi(S)+\sinh r\left(\psi'(S)+\frac{K}{B_0^2}\langle Y,X_s\rangle_L\right)=0.

Dividing by sinh⁡r\sinh r and combining the last two identities, we obtain

ψ′(S)−cψ(S)=−KB02drda.(38)\psi'(S)-c\psi(S)=-\frac{K}{B_0^2}\frac{\mathrm{d}r}{\mathrm{d}a}. \tag*{(38)}

Suppose that dr/da=0\mathrm{d}r/\mathrm{d}a=0 for some aa. Then (38) and the evenness of ψ\psi show that ψ\psi satisfies both conditions in (15). Since J0ψ=0J_0\psi=0 and ψ(0)≠0\psi(0)\ne0, zero would be an eigenvalue of the radial Robin problem for J0J_0, contrary to Proposition 4. Hence dr/da\mathrm{d}r/\mathrm{d}a never vanishes and has constant sign. By Lemma 2,

cosh⁡r=A(S)cosh⁡φ(S)≥1+B02>1+2K2,\cosh r=A(S)\cosh\varphi(S)\ge\sqrt{1+B_0^2}>\sqrt{1+2K^2},

and the right-hand side tends to infinity as a→∞a \to\infty. Thus rr is unbounded as a→∞a \to\infty, which excludes the negative sign. □

Combined with the limit r(a)→0r(a) \to0 as a→12+a \to\frac{1}{2}^{+}, established in [10], Corollary 9 shows that a↦r(a)a \mapsto r(a) is an increasing diffeomorphism of (12,∞)(\frac{1}{2}, \infty) onto (0,∞)(0, \infty). Thus every geodesic ball of H3\mathbb{H}^{3} is the supporting ball of an isometric image of exactly one Σa\Sigma_a. Corollary 9 answers the question of the strict monotonicity of r(a)r(a); see [9 Remark 1.3].

Remark 10. The restriction to angular period 2π2\pi in Theorem 1 cannot be dropped. The radial form (16) is defined for every real k≥0k \ge0, and

qk[u]−qℓ[u]=(k2−ℓ2)∫−SSu2B ds.q_k[u] - q_\ell[u] = (k^2-\ell^2)\int_{-S}^{S}\frac{u^2}{B}\,ds.

Since B(0)≤B≤B0B(0) \le B \le B_0 on [−S,S][-S,S], the min–max principle gives, for the radial Robin eigenvalues λj(k)\lambda_j(k),

k2−ℓ2B02≤λj(k)−λj(ℓ)≤k2−ℓ2B(0)2(0≤ℓ<k).\frac{k^2-\ell^2}{B_0^2} \le\lambda_j(k)-\lambda_j(\ell) \le\frac{k^2-\ell^2}{B(0)^2}\qquad(0\le\ell<k).

Hence each λj(k)\lambda_j(k) is a continuous, strictly increasing function of k≥0k \ge0. Propositions 5 and 7 give λ1(1)<0<λ1(2)\lambda_1(1)<0<\lambda_1(2), so there is a unique k∗(a)∈(1,2)k_*(a)\in(1,2) with λ1(k∗(a))=0\lambda_1(k_*(a))=0. For 0<k<10<k<1, Propositions 4 and 5 give λ1(k)<λ2(k)<λ2(1)=0<λ3(0)<λ3(k)\lambda_1(k)<\lambda_2(k)<\lambda_2(1)=0<\lambda_3(0)<\lambda_3(k); for 1<k<21<k<2, one has λ2(k)>λ2(1)=0\lambda_2(k)>\lambda_2(1)=0. On the mm-fold cover of Σa\Sigma_a, of angular period 2πm2\pi m, the angular frequencies are j/mj/m with integer j≥0j\ge0. Each frequency j/m∈(0,1)j/m\in(0,1) therefore contributes four negative directions, so for m≥2m\ge2 the index of the cover is at least 4m4m. Moreover, the nullity of the cover exceeds two exactly when mk∗(a)mk_*(a) is an integer, which again requires m≥2m\ge2. This agrees with the construction of Cerezo [2], in which non-rotational free boundary minimal annuli in geodesic balls of H3\mathbb{H}^3 bifurcate from mm-fold covers of critical spherical catenoids (called free boundary hyperbolic catenoids in [2]). The bifurcating annuli are invariant under nn reflections in planes through the axis, where mm and nn are coprime and n/mn/m lies in an interval contained in (2,3)⊂(1,2)(\sqrt{2},\sqrt{3})\subset(1,2).

Remark 11 (Scope of the argument). Lemmas 2 and 6 rest on identities specific to the rotational family of Section 2, and the separation of variables uses its rotational symmetry. Three ingredients nevertheless have a wider scope.

(i) The identity (18) uses neither the explicit parametrization nor the rotational symmetry: for every two-sided minimal hypersurface Σn⊂Hn+1⊂Rn+1,1\Sigma^n\subset\mathbb{H}^{n+1}\subset\mathbb{R}^{n+1,1}, the Gauss–Weingarten and Codazzi equations give ΔΣν=−∣II∣2ν\Delta_\Sigma\nu=-|\mathrm{II}|^2\nu, hence Jν0=nν0J\nu^0=n\nu^0 for J=−ΔΣ−∣II∣2+nJ=-\Delta_\Sigma-|\mathrm{II}|^2+n. If Σ\Sigma is free boundary in a geodesic ball centred at e0e_0, then ν0=0\nu^0=0 on ∂Σ\partial\Sigma. What is specific to the present family is the sign property of Lemma 3: ν0\nu^0 is positive in the interior, hence it is the first Dirichlet eigenfunction, with eigenvalue 22; this is the input of Proposition 4.

(ii) The last step of Proposition 7 is an instance of the ground-state representation for regular Sturm–Liouville problems. Write the Robin conditions as ∂ηz=σz\partial_\eta z=\sigma z at the endpoints. If a positive solution uu of the zero-energy equation satisfies ∂ηu/u>σ\partial_\eta u/u>\sigma at both endpoints, the substitution z=uhz=uh writes the quadratic form as a nonnegative interior square plus boundary terms whose coefficients are positive multiples of ∂ηu/u−σ\partial_\eta u/u-\sigma; the form is therefore positive definite. By symmetry the two conditions reduce here to p2(T)>v0p_2(T)>v_0, which is Lemma 6.

(iii) The proof of Corollary 9 reflects a general transversality principle. Let XaX_a be a smooth family of free boundary minimal hypersurfaces in the geodesic balls of radius r(a)r(a) centred at e0e_0, parametrized over a fixed manifold with boundary, and let ψ=⟨∂aXa,ν⟩\psi=\langle\partial_aX_a,\nu\rangle. Then ψ\psi is a Jacobi field, and differentiating the two free boundary conditions gives

∂ηψ−coth⁡r ψ=−II(η,η)drdaon ∂Σ,\partial_\eta\psi-\coth r\,\psi=-\mathrm{II}(\eta,\eta)\frac{dr}{da}\quad\text{on }\partial\Sigma,

where η\eta is the outward conormal; here one uses that η\eta is a principal direction along ∂Σ\partial\Sigma, the supporting sphere being umbilic. For Σa\Sigma_a one has II⁡(η,η)=K/B2\operatorname{II}(\eta,\eta) = K/B^2, and this is (8.1). Hence, if dr/da\mathrm{d}r/\mathrm{d}a vanished, ψ\psi would lie in the Robin kernel. For Σa\Sigma_a, ψ\psi is nonzero since ψ(0)=1/(2K)\psi(0) = 1/(2K), and Proposition 4 excludes a radial Robin kernel, so dr/da≠0\mathrm{d}r/\mathrm{d}a \ne0; the sign is then fixed by the unboundedness of rr.

References

References

  1. [1]M. do Carmo and M. Dajczer, Rotation hypersurfaces in spaces of constant curvature, Trans. Amer. Math. Soc. 277 (1983), no. 2, 685–709. https://doi.org/10.1090/S0002-9947-1983-0694383-X.
  2. [2]A. Cerezo, Free boundary minimal annuli in geodesic balls of H³, arXiv:2502.20303 (2025). https://arxiv.org/abs/2502.20303.
  3. [3]B. Deyver, Index of the critical catenoid, Geom. Dedicata 199 (2019), 355–371. https://doi.org/10.1007/s10711-018-0353-2.
  4. [4]H. Li and C. Xiong, A gap theorem for free boundary minimal surfaces in geodesic balls of hyperbolic space and hemisphere, J. Geom. Anal. 28 (2018), no. 4, 3171–3182. https://doi.org/10.1007/s12220-017-9953-6.
  5. [5]C. Lima, On uniqueness of free boundary minimal annuli in geodesic balls of S³₊ and H³, arXiv:2503.16763 (2025). https://arxiv.org/abs/2503.16763.
  6. [6]V. Medvedev, On the index of the critical Möbius band in ℝ⁴, J. Geom. Anal. 33 (2023), no. 3, article no. 93. https://doi.org/10.1007/s12220-022-01069-w.
  7. [7]V. Medvedev, On free boundary minimal submanifolds in geodesic balls in Hⁿ and Sⁿ₊, Math. Z. 310 (2025), no. 4, article no. 115. https://doi.org/10.1007/s00209-025-03707-9.
  8. [8]H. Mori, Minimal surfaces of revolution in H³ and their global stability, Indiana Univ. Math. J. 30 (1981), no. 5, 787–794. https://doi.org/10.1512/iumj.1981.30.30057.
  9. [9]A. Pigazzini, Analytic local resolution of Medvedev’s Morse index conjecture for the critical spherical catenoid in H³, arXiv:2605.13562 (2026). https://arxiv.org/abs/2605.13562.
  10. [10]A. Pigazzini, Robin nullity in mode |k| = 1 and asymptotic radius of the critical spherical catenoid, arXiv:2605.11244 (2026). https://arxiv.org/abs/2605.11244.
  11. [11]G. Smith and D. Zhou, The Morse index of the critical catenoid, Geom. Dedicata 201 (2019), 13–19. https://doi.org/10.1007/s10711-018-0412-8.
  12. [12]H. Tran, Index characterization for free boundary minimal surfaces, Commun. Anal. Geom. 28 (2020), no. 1, 189–222. https://doi.org/10.4310/CAG.2020.v28.n1.a6. See also https://arxiv.org/abs/1609.01651v4.
  13. [13]A. Zettl, Sturm–Liouville Theory, Mathematical Surveys and Monographs, vol. 121, American Mathematical Society, Providence, RI, 2005. https://doi.org/10.1090/surv/121.

Paper details

vlnc:W_zwqgjdz5
Published
Fresh — added to the Hub within the last 30 days
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

Not recorded for this paper

View statement
Looking for related papers…
Contents