betaFeedback

WorksMathematics

Uniqueness of the hexagonal lattice's universal optimality

Abstract

We build on a recent proof of the universal optimality of the planar hexagonal lattice. We extend this result to demonstrate that, up to Euclidean motion, the hexagonal lattice is the only periodic configuration with this property.

Energy minimization and universal optimality

We study the problem of arranging points in periodic configurations to minimize certain energies. Towards this goal, we call a function p:R>0→Rp : \mathbb{R}_{>0} \to\mathbb{R} a potential. For a discrete set X⊂RnX \subset\mathbb{R}^{n}, n∈Nn \in\mathbb{N}, we denote the number of points in XX by ∣X∣|X|. For a point configuration C⊂RnC \subset\mathbb{R}^{n}, we define its pp-energy as

Ep(C)=lim⁡R→∞1∣C∩BR∣∑γ,γ′∈C∩BRγ≠γ′p(∥γ−γ′∥).E_p(C) = \lim_{R \to\infty} \frac{1}{|C \cap B_R|} \sum_{\substack{\gamma,\gamma' \in C \cap B_R \\ \gamma\ne\gamma'}} p(\|\gamma-\gamma'\|).

where BR={x∈Rn∣∥x∥<R}B_R = \{x \in\mathbb{R}^{n} \mid\|x\| < R\} denotes the open ball of radius R>0R > 0 centered at the origin. The limit may not exist, in which case we define the energy to be infinite. The quantity

ρ(C)=lim⁡R→∞∣C∩BR∣vol⁡(BR)\rho(C) = \lim_{R \to\infty} \frac{|C \cap B_R|}{\operatorname{vol}(B_R)}

is called the density of the point configuration CC.

The energy minimization problem is as follows. For a fixed potential and a fixed density, we seek to find the minimizer of EpE_p among all point configurations. First, we formulate this problem for Gaussian energies φα:r↦e−αr2\varphi_\alpha: r \mapsto e^{-\alpha r^2}, α>0\alpha> 0. Later, in Subsection 1.1 we describe how the problem can naturally be extended to so-called completely monotone functions of squared distance.

Definition 1.1. A point configuration C0C_0 with ρ(C0)=1\rho(C_0) = 1 is called universally optimal if we have

Eφα(C0)≤Eφα(C),for all C with ρ(C)=1 and all α>0.E_{\varphi_\alpha}(C_0) \le E_{\varphi_\alpha}(C), \qquad\text{for all } C \text{ with } \rho(C) = 1 \text{ and all } \alpha> 0.

If C0C_0 is universally optimal, then for every ρ>0\rho> 0, the rescaled configuration C~0=ρ−1/dC0\widetilde{C}_0 = \rho^{-1/d}C_0 minimizes every Gaussian energy among configurations of density ρ\rho, i.e.,

Eφα(C~0)≤Eφα(C),for all C with ρ(C)=ρ and α>0.E_{\varphi_\alpha}(\widetilde{C}_0) \le E_{\varphi_\alpha}(C), \qquad\text{for all } C \text{ with } \rho(C) = \rho\text{ and } \alpha> 0.

In dimension two, a celebrated result of Montgomery [22] establishes that the hexagonal (or triangular) lattice minimizes every Gaussian energy among lattices of fixed density. The hexagonal lattice of density one is defined as

Λ2=23(112032)Z2.(1)\Lambda_2 = \sqrt{\frac{2}{\sqrt{3}}} \begin{pmatrix} 1 & \frac{1}{2} \\ 0 & \frac{\sqrt{3}}{2} \end{pmatrix} \mathbb{Z}^{2}. \tag*{(1)}

Cohn and Kumar conjectured that the hexagonal lattice is universally optimal among periodic point configurations of fixed density in R2\mathbb{R}^{2} [7 Conjecture 9.4]. This universal optimality was recently proved by OpenAI [23].

Faculty of Mathematics, University of Vienna, Oskar-Morgenstern-Platz 1, 1090 Vienna, Austria

Faculty of Mathematics, University of Vienna, Kolingasse 14-16, 1090 Vienna, Austria

E-mail addresses: [email protected], [email protected].

2020 Mathematics Subject Classification. 52C15.

This research was supported in whole or in part by the Austrian Science Fund (FWF) 10.55776/PAT51022224. For open access purposes, the authors have applied a CC BY public copyright license to any author accepted manuscript version arising from this submission.

A natural associated question is whether Λ2\Lambda_2 is, up to Euclidean motion, the only point configuration that has this property. To make this question precise, we note that uniqueness among arbitrary configurations does not hold. This is because deleting finitely many points from Λ2\Lambda_2 preserves its density and Gaussian energies, but produces a configuration that is not congruent to Λ2\Lambda_2. We therefore restrict our uniqueness question to periodic point configurations.

It is explicitly mentioned in [23] that only minimality of Λ2\Lambda_2 for all Gaussian energies is shown, but not the uniqueness of Λ2\Lambda_2 for the considered problem.

Our contribution is to close this gap by proving the following result.

Theorem 1.2. For every periodic point configuration C⊂R2C \subset\mathbb{R}^2 with ρ(C)=1\rho(C)=1 we have

Eφα(Λ2)≤Eφα(C),for all α>0.(2)E_{\varphi_\alpha}(\Lambda_2) \le E_{\varphi_\alpha}(C), \qquad\text{for all } \alpha> 0. \tag*{(2)}

Moreover, the following statements apply.

(A) If equality in (2) holds for every α>0\alpha> 0, then CC is the hexagonal lattice Λ2\Lambda_2 up to Euclidean motion.

(B) If equality holds in (2) for at least one α>0\alpha> 0, then CC is the hexagonal lattice Λ2\Lambda_2 up to Euclidean motion.

Statement (2) was shown in [23]. Clearly, statement (B) implies (A). We nevertheless give a separate, simpler proof of (A), for which we only need the asymptotic decay of Gaussian energy and uniqueness of the densest periodic sphere packing in the plane. No further results from [23] are used. Establishing (B) additionally uses properties of the auxiliary functions constructed in [23], which we identify below.

General potentials

Using standard techniques, many of the arguments around universal optimality can naturally be extended from Gaussian potentials to completely monotone functions of squared distance.

Definition 1.3. A function f:R>0→Rf:\mathbb{R}_{>0}\to\mathbb{R} is completely monotone if f∈C∞(R>0)f\in C^\infty(\mathbb{R}_{>0}) and

(−1)kf(k)(s)≥0,for all s∈R>0, k=0,1,2,….(-1)^k f^{(k)}(s) \ge0, \qquad\text{for all } s \in\mathbb{R}_{>0},\ k=0,1,2,\ldots.

A potential g:R>0→Rg:\mathbb{R}_{>0}\to\mathbb{R} is completely monotone of squared distance if g(r)=f(r2)g(r)=f(r^2) for some completely monotone function ff.

Observe that completely monotone functions are in particular nonnegative, nonincreasing, and convex. The Bernstein–Widder theorem characterizes completely monotone functions via the Laplace transform of nonnegative Borel measures [1, 31].

Theorem 1.4 (Bernstein-Widder). Let f:R>0→R≥0f:\mathbb{R}_{>0}\to\mathbb{R}_{\ge0} be completely monotone. Then there exists a non-negative Borel measure μf\mu_f such that

f(s)=∫0∞e−αs dμf(α).f(s)=\int_0^\infty e^{-\alpha s}\,\mathrm{d}\mu_f(\alpha).

Consequently, every completely monotone potential of squared distance has the representation

g(r)=f(r2)=∫0∞e−αr2 dμf(α).g(r)=f(r^2)=\int_0^\infty e^{-\alpha r^2}\,\mathrm{d}\mu_f(\alpha).

Thus, completely monotone functions of squared distance are nonnegative, continuous superpositions of Gaussians φα\varphi_\alpha and the constant function. Therefore, inequality (2) remains valid with φα\varphi_\alpha replaced by any completely monotone potential of squared distance.

Moreover, the following uniqueness result for the hexagonal lattice, similar to Theorem 1.2, holds in this case.

Corollary 1.5. Let g≠0g\ne0 be a completely monotone potential of squared distance, and suppose Eg(Λ2)<∞E_g(\Lambda_2)<\infty. Then

Eg(C)≥Eg(Λ2)E_g(C) \ge E_g(\Lambda_2)

for every periodic configuration C⊂R2C\subset\mathbb{R}^2, ρ(C)=1\rho(C)=1, with equality if and only if CC is congruent to Λ2\Lambda_2.

Related work

Universal optimality was shown to hold for the integers Z\mathbb{Z} in dimension 1 [7] and, in seminal work by Cohn, Kumar, Miller, Radchenko, and Viazovska [9], in dimensions 8 and 24, where the minimizers are the E8E_8-lattice and the Leech lattice Λ24\Lambda_{24}, respectively. In all cases, the minimizer is also, up to Euclidean motion, the unique periodic configuration of density 1 that yields the densest sphere packing in that dimension. While sphere packing in one dimension is trivial, the seminal breakthrough by Viazovska [30] solving the sphere packing problem in dimension 8, soon followed by a proof for dimension 24 [8], is based on the linear programming bounds established by Cohn and Elkies [6]. In all proved cases, including OpenAI’s result in dimension 2 [23], universal optimality could be established by constructing a magic function saturating the linear programming bounds for energy minimization [7]. Yet, it should be noted that an alternative, earlier proof in dimension 1 was given by Ventevogel and Nijboer [29].

The problem of identifying optimal configurations for energy minimization in the plane has a rich history deeply intertwined with the Epstein zeta function. For a two-dimensional lattice Λ⊂R2\Lambda\subset\mathbb{R}^2, the function is defined for Re⁡(s)>1\operatorname{Re}(s)>1 by

ζΛ(s)=∑λ∈Λ∖{0}1∥λ∥2s\zeta_{\Lambda}(s)=\sum_{\lambda\in\Lambda\setminus\{0\}}\frac{1}{\lVert\lambda\rVert^{2s}}

and extended via analytic continuation to C∖{1}\mathbb{C}\setminus\{1\}. For real s>1s>1, this sum equals the normalized energy Eps(Λ)E_{p_s}(\Lambda) for the Riesz potential ps(r)=r−2sp_s(r)=r^{-2s}. For 0<s≤10<s\leq1, the positive interaction sum diverges, so the analytically continued zeta values must be distinguished from this energy.

Rankin [26] proved that the hexagonal lattice minimizes the Epstein zeta function among lattices of fixed density for a restricted range of real ss. Cassels [5] proposed an extension, but identified a gap in the proof in his corrigendum. Ennola [12, 13] supplied a repair, and Diananda [11] simplified the argument. Together, these works establish hexagonal-lattice minimality for the analytically continued zeta functions for real s>0s>0, s≠1s\ne1. At s=1s=1, the comparison concerns the finite difference between zeta functions of lattices of the same density, after cancellation of their common pole.

Montgomery [22] studied the lattice theta function

θΛ(α)=∑λ∈Λe−πα∥λ∥2for α>0.\theta_{\Lambda}(\alpha)=\sum_{\lambda\in\Lambda}e^{-\pi\alpha\lVert\lambda\rVert^2}\qquad\text{for }\alpha>0.

For Re⁡s>1\operatorname{Re}s>1, we have the Mellin transform identity, see, e.g., [20]

ζΛ(s)=πsΓ(s)∫0∞(θΛ(α)−1)αs−1 dα,\zeta_{\Lambda}(s)=\frac{\pi^s}{\Gamma(s)}\int_0^\infty\left(\theta_{\Lambda}(\alpha)-1\right)\alpha^{s-1}\,d\alpha,

which relates the zeta function to the theta function. Montgomery proved that, among lattices of fixed density, the hexagonal lattice uniquely minimizes θΛ(α)\theta_{\Lambda}(\alpha) for every α>0\alpha>0, up to orthogonal transformations. The Mellin identity therefore yields its optimality for the convergent Riesz energies. Integration against positive measures similarly gives minimality for completely monotone potentials of squared distance.

Beyond constrained optimization over rigid lattice structures, these geometric energy minimizers are deeply linked to the fundamental crystallization conjecture in mathematical physics. This conjecture asserts that under appropriate physical conditions, a system of NN interacting particles will spontaneously break translation invariance and arrange itself into a periodic configuration as N→∞N\to\infty. As detailed in the comprehensive review by Blanc and Lewin [3], proving this phenomenon from first principles remains one of the most challenging open problems in statistical mechanics. Substantial progress in understanding the foundational limits of such systems has been achieved through the work of Serfaty [27] and Petrache and Serfaty [24], who developed a rigorous variational framework for the next-order energy of Riesz gases, connecting macroscopic distribution profiles to microscopic point configurations. Bétermin [2] established hexagonal-lattice optimality among planar Bravais lattices for completely monotone potentials of squared distance with finite energy at fixed density, and for certain Lennard–Jones-type potentials. Hardin and Saff [18] studied energy asymptotics and limiting distributions of minimizing Riesz configurations on rectifiable manifolds.

Before the unconstrained problem was solved by OpenAI [23], optimality of the hexagonal lattice within certain classes of periodic configurations was obtained by Faulhuber, Shafkulovska, and Zlotnikov [14] and by Hardin and Tenpas [19]. Moreover, recent work by Leblé [21] established that the hexagonal lattice is a local energy minimizer within periodic configurations.

Disclosure on the usage of AI tools

The authors used OpenAI’s ChatGPT-6 Astra model and Microsoft’s Copilot, based on the GPT-5.6 Sol (OpenAI) reasoning model, as interactive mathematical and computational assistants during the development of this work. The models were used to identify the necessary auxiliary results in [23] needed to prove Theorem 1.2 (B). The formulation of the research questions, the assessment of the mathematical arguments, and the selection of results for inclusion remained with the authors. All claims in the final manuscript were reviewed, rewritten, and verified by the authors, who take full responsibility for the mathematical content and its presentation.

Notation and further results

Throughout, we denote the minimal distance dd between distinct points in Λ2\Lambda_2, defined by (1), by

d=23.d=\sqrt{\frac{2}{\sqrt{3}}}.

Definition 2.1. For n∈Nn\in\mathbb{N}, a periodic point configuration C⊂RnC\subset\mathbb{R}^n is of the form

C=⋃j=1N(vj+Λ),ρ(C)=Nvol⁡(Rn/Λ),(3)C=\bigcup_{j=1}^{N}(v_j+\Lambda),\qquad\rho(C)=\frac{N}{\operatorname{vol}(\mathbb{R}^n/\Lambda)}, \tag*{(3)}

where Λ⊂Rn\Lambda\subset\mathbb{R}^n is a full-rank lattice and the vjv_j are distinct modulo Λ\Lambda, i.e., vi−vj∉Λv_i-v_j\notin\Lambda for i≠ji\ne j.

As vol⁡(Rn/Λ)\operatorname{vol}(\mathbb{R}^n/\Lambda) is the volume of a fundamental cell of Λ\Lambda, the periodic configuration CC in (3) has density 11 exactly when vol⁡(Rn/Λ)=N\operatorname{vol}(\mathbb{R}^n/\Lambda)=N.

For a nonnegative potential p:R>0→R≥0p:\mathbb{R}_{>0}\to\mathbb{R}_{\geq0} and a periodic configuration CC as in (3), the pp-energy of CC is given by

Ep(C)=1N∑i,j=1N∑λ∈Λ∖{vi−vj}p(∥vi−vj+λ∥),(4)E_p(C)=\frac{1}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda\setminus\{v_i-v_j\}}p(\lVert v_i-v_j+\lambda\rVert), \tag*{(4)}

where the self-interaction terms with i=ji=j and λ=0\lambda=0 are excluded. For t>0t>0, put

Et(C)=Er↦e−πtr2(C).E_t(C)=E_{r\mapsto e^{-\pi t r^2}}(C).

We call Et(C)E_t(C) the Gaussian energy of CC. In the notation of the introduction, Et(C)=Eφπt(C)E_t(C)=E_{\varphi_{\pi t}}(C). Thus (2) applies with α=πt\alpha=\pi t.

We will prove Theorem 1.2 (A) in § 3 and Theorem 1.2 (B) in § 4. The arguments also yield the following additional results of independent interest. These rely on the connections between energy minimization and optimal sphere packings, which have been stated more informally, e.g., in [7 Section 9], [9 Section 1.1], [25 Section 1], and hold in arbitrary dimension.

Proposition 2.2. For n∈Nn\in\mathbb{N}, let C⊂RnC\subset\mathbb{R}^n be a periodic configuration that minimizes EtE_t among periodic configurations of the same density for every t>0t>0. Then CC maximizes minimum separation among periodic configurations of that density. Equivalently, its centers give a densest periodic sphere packing. In particular, this conclusion holds for every universally optimal periodic configuration with density 11.

Theorem 2.3. For n∈Nn\in\mathbb{N}, let C⊂RnC\subset\mathbb{R}^n be a periodic configuration that minimizes EtE_t among periodic configurations of the same density for every t>0t>0. Suppose that, up to Euclidean motion, CC is the unique periodic configuration of that density maximizing minimum separation. Then, up to Euclidean motion, CC is the unique periodic configuration of that density minimizing every Gaussian energy simultaneously.

Part I: the minimum separation route

The aim of this section is to prove Theorem 1.2 (A). We show that if a periodic configuration of density 11 has the same energy as Λ2\Lambda_2 for every Gaussian potential, then it is congruent to Λ2\Lambda_2. The proof establishes the minimum separation of points in the configuration and uses the uniqueness of the densest periodic circle packing. Together with (2), this gives uniqueness of the simultaneous minimizer of all Gaussian energies.

We begin by showing that for periodic configurations, the decay rate of Gaussian energies for t→∞t \to\infty corresponds directly to the minimum distance between points. For a periodic configuration CC, the minimum distance between distinct points exists and is strictly positive. Indeed, a periodic configuration has only finitely many points modulo its period lattice. Its pair-displacement set is a finite union of translates of a discrete lattice and therefore contains only finitely many vectors in every bounded region. We denote this minimum distance by sep⁡(C)\operatorname{sep}(C).

Lemma 3.1. Let CC be a periodic configuration in Rn\mathbb{R}^{n}. Then

lim⁡t→∞(−1πtlog⁡Et(C))=sep⁡(C)2.(5)\lim_{t \to\infty}\left(-\frac{1}{\pi t}\log E_t(C)\right)=\operatorname{sep}(C)^2. \tag*{(5)}

Proof. Write r0=sep⁡(C)r_0=\operatorname{sep}(C). Let M>0M>0 be the average number of points at distance r0r_0 from a point of CC, where the average is taken over the NN representatives v1,…,vNv_1,\ldots,v_N in one period cell, i.e.,

M=1N∑i=1N∣{x∈C:∥x−vi∥=r0}∣.M=\frac{1}{N}\sum_{i=1}^{N}\lvert\{x\in C:\lVert x-v_i\rVert=r_0\}\rvert.

Then, by (4)

Et(C)≥Me−πtr02.(6)E_t(C)\geq M e^{-\pi t r_0^2}. \tag*{(6)}

Fix t0>0t_0>0. For t≥t0t\geq t_0, we use (4) and the fact that ∥vi−vj+λ∥≥r0\lVert v_i-v_j+\lambda\rVert\geq r_0 in every summand to obtain

Et(C)=1N∑i,j=1N∑λ∈Λ∖{vi−vj}e−πt∥vi−vj+λ∥2=e−πtr02N∑i,j=1N∑λ∈Λ∖{vi−vj}e−πt(∥vi−vj+λ∥2−r02)≤e−πtr02N∑i,j=1N∑λ∈Λ∖{vi−vj}e−πt0(∥vi−vj+λ∥2−r02)=e−πtr02eπt0r02Et0(C).(7)\begin{aligned} E_t(C)&=\frac{1}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda\setminus\{v_i-v_j\}}e^{-\pi t\lVert v_i-v_j+\lambda\rVert^2}\\ &=\frac{e^{-\pi t r_0^2}}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda\setminus\{v_i-v_j\}}e^{-\pi t(\lVert v_i-v_j+\lambda\rVert^2-r_0^2)}\\ &\leq\frac{e^{-\pi t r_0^2}}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda\setminus\{v_i-v_j\}}e^{-\pi t_0(\lVert v_i-v_j+\lambda\rVert^2-r_0^2)}\\ &=e^{-\pi t r_0^2}e^{\pi t_0r_0^2}E_{t_0}(C). \tag*{(7)} \end{aligned}

It is clear from (4) that the quantity Et0(C)E_{t_0}(C) is finite for a periodic configuration. Taking logarithms in (6) and (7), dividing by −πt-\pi t, and letting t→∞t\to\infty proves (5). □

Proof of Proposition 2.2. Let CC satisfy the stated minimality assumption, and let DD be a periodic configuration of the same density. Hence,

Et(C)≤Et(D)for all t>0.E_t(C)\leq E_t(D)\qquad\text{for all }t>0.

After taking logarithms, multiplying by −1/(πt)-1/(\pi t), and passing to the limit in Lemma 3.1, we obtain

sep⁡(C)2≥sep⁡(D)2.\operatorname{sep}(C)^2\geq\operatorname{sep}(D)^2.

Thus CC maximizes minimum separation. Balls of radius sep⁡(C)/2\operatorname{sep}(C)/2 centered at the elements of CC therefore form a densest periodic sphere packing. A universally optimal periodic configuration of density 1 satisfies the minimality assumption by definition. □

For periodic configurations of density 1, Proposition 2.2 gives the implication

C is not an optimal sphere packing⟹C is not universally optimal.C\text{ is not an optimal sphere packing}\quad\Longrightarrow\quad C\text{ is not universally optimal}.

Proof of Theorem 2.3. Let DD be a periodic configuration of the same density as CC, minimizing every Gaussian energy among periodic configurations of density ρ(C)\rho(C). Since CC and DD both minimize EtE_t for every t>0t>0,

Et(D)=Et(C)for all t>0.E_t(D)=E_t(C)\qquad\text{for all }t>0.

Lemma 3.1 gives sep⁡(D)=sep⁡(C)\operatorname{sep}(D)=\operatorname{sep}(C). Thus, DD also maximizes minimum separation, and the assumed uniqueness of CC implies that DD is congruent to CC. □

Proof of Theorem 1.2(A). Let CC be a periodic configuration with ρ(C)=1\rho(C)=1 for which equality holds in (2) for every α>0\alpha>0. Since Et(C)=Eφπt(C)E_t(C)=E_{\varphi_{\pi t}}(C), this assumption gives

Et(C)=Et(Λ2),for all t>0.E_t(C)=E_t(\Lambda_2),\qquad\text{for all }t>0.

Lemma 3.1 yields

sep⁡(C)=sep⁡(Λ2)=d.\operatorname{sep}(C)=\operatorname{sep}(\Lambda_2)=d.

Hence disks of radius d/2d/2 centered at the points of CC form a periodic packing. Since ρ(C)=1\rho(C)=1, its packing density equals

π(d2)2=π23=π12.(8)\pi\left(\frac{d}{2}\right)^2=\frac{\pi}{2\sqrt{3}}=\frac{\pi}{\sqrt{12}}. \tag*{(8)}

This is the maximal planar packing density. Since the hexagonal packing is the unique densest periodic circle packing in the plane up to Euclidean motion [15, 28] see also [4, 10, 16], it follows that CC is congruent to Λ2\Lambda_2. □

Two useful variants follow from the same argument.

Corollary 3.2. Let C⊂R2C\subset\mathbb{R}^2 be a periodic point configuration of density 11, and let Λ2\Lambda_2 be the hexagonal lattice of density 11 defined in (1). Let (tk)k∈N(t_k)_{k\in\mathbb{N}} be a sequence of positive real numbers with tk→∞t_k\to\infty. If

Etk(C)≤Etk(Λ2)E_{t_k}(C)\le E_{t_k}(\Lambda_2)

for every k∈Nk\in\mathbb{N}, then CC is congruent to Λ2\Lambda_2.

Proof. Lemma 3.1 gives sep⁡(C)≥d\operatorname{sep}(C)\ge d. Equation (8) and the uniqueness of the densest periodic circle packing in the plane completes the proof. □

Corollary 3.3. Let g≢0g\not\equiv0 be a completely monotone potential of squared distance with representation

g(r)=∫(0,∞)e−πtr2 dμ(t)g(r)=\int_{(0,\infty)}e^{-\pi t r^2}\,d\mu(t)

where μ\mu is a positive Borel measure on (0,∞)(0,\infty). Suppose Eg(Λ2)<∞E_g(\Lambda_2)<\infty and supp⁡(μ)\operatorname{supp}(\mu) is unbounded. Then Λ2\Lambda_2 is the unique density-one periodic minimizer of EgE_g.

Proof. Let C⊂R2C\subset\mathbb{R}^2 be a periodic configuration of density 11. By Tonelli’s theorem and (2), applied with α=πt\alpha=\pi t,

Eg(C)−Eg(Λ2)=∫(0,∞)(Et(C)−Et(Λ2)) dμ(t)≥0.E_g(C)-E_g(\Lambda_2)=\int_{(0,\infty)}\left(E_t(C)-E_t(\Lambda_2)\right)\,d\mu(t)\ge0.

If equality holds, then Et(C)=Et(Λ2)E_t(C)=E_t(\Lambda_2) for μ\mu-almost every tt. From the unbounded support assumption, we conclude that there is a diverging sequence (tk)k∈N(t_k)_{k\in\mathbb{N}} with Etk(C)=Etk(Λ2)E_{t_k}(C)=E_{t_k}(\Lambda_2), for all k∈Nk\in\mathbb{N}. An application of Corollary 3.2 completes the proof. □

Part II: the auxiliary-function route

We now use specific information from the proof in [23]. This route yields the strong uniqueness statement of Theorem 1.2 (B). We use the Fourier transform

f^(ξ)=∫R2f(x)e−2πix⋅ξ dx,for ξ∈R2.\widehat{f}(\xi)=\int_{\mathbb{R}^2}f(x)e^{-2\pi i x\cdot\xi}\,dx,\qquad\text{for }\xi\in\mathbb{R}^2.

It is well-known (see, e.g., [17 Appendix A]) that the Fourier transform of a Gaussian GtG_t is given by

Gt^(x)=1tG1/t(x),Gt(x)=e−πt∥x∥2,for all x∈R2, t>0.\widehat{G_t}(x)=\frac{1}{t}G_{1/t}(x),\qquad G_t(x)=e^{-\pi t\lVert x\rVert^2},\qquad\text{for all }x\in\mathbb{R}^2,\ t>0.

We recall the following auxiliary-function construction from [23, Lemma 6.2 and equation (6.5)], expressed in our notation.

Proposition 4.1. For every t>0t>0 there is a real radial Schwartz function ftf_t such that

ft≤Gt,ft^≥0,(9)f_t\le G_t,\qquad\widehat{f_t}\ge0, \tag*{(9)}

and

ft^(0)−ft(0)=Et(Λ2).(10)\widehat{f_t}(0)-f_t(0)=E_t(\Lambda_2). \tag*{(10)}

The functions ftf_t and ft^\widehat{f_t} satisfy the contact conditions from [23 Lemma 6.2]:

ft(x)=Gt(x)for all x∈Λ2∖{0},ft^(ξ)=0for all ξ∈Λ2∗∖{0},f_t(x)=G_t(x)\quad\text{for all }x\in\Lambda_2\setminus\{0\},\qquad \widehat{f_t}(\xi)=0\quad\text{for all }\xi\in\Lambda_2^*\setminus\{0\},

where Λ2∗\Lambda_2^* denotes the dual lattice of Λ2\Lambda_2.

Recall that the dual lattice to Λ\Lambda is

Λ∗={ξ∈R2:ξ⋅λ∈Z for every λ∈Λ}.\Lambda^*=\{\xi\in\mathbb{R}^2:\xi\cdot\lambda\in\mathbb{Z}\text{ for every }\lambda\in\Lambda\}.

To prove Theorem 1.2 (B), we need Gt(x)>ft(x)G_t(x)>f_t(x) whenever 0<∥x∥<d0<\lVert x\rVert<d. The following lemma establishes this property using the quantitative estimates in [23].

Lemma 4.2. For every t>0t>0, choose ftf_t as constructed in the proof of [23 Lemma 6.2], and put

Qt=Gt−ftQ_t=G_t-f_t

Then

Qt(x)>0andft^(x)>0,for all 0<∥x∥<d.(11)Q_t(x)>0\qquad\text{and}\qquad\widehat{f_t}(x)>0,\qquad\text{for all }0<\lVert x\rVert<d. \tag*{(11)}

Proof. We first identify the functions in the construction of [23], then explain how its quantitative estimates imply strict positivity. We use the particular auxiliary functions specified in [23] as the properties listed in Proposition 4.1 alone do not give the quantitative estimates below.

Throughout, we put

h=25,k=π(d2t−h),s=∥x∥2d2.h=\frac{2}{5},\qquad k=\pi\left(d^2t-h\right),\qquad s=\frac{\lVert x\rVert^2}{d^2}.

Step 1: identify the functions for $t\ge 1. As is common when working with Gaussian energies, we first treat the case t≥1t\ge1 and then employ the Fourier transform to cover the range $0<t<1$. The Gaussian parameter denoted by α\alpha in [23 Lemma 6.2] is our tt. We have

k≥π(23−25)>2.36,k\ge\pi\left(\frac{2}{\sqrt{3}}-\frac{2}{5}\right)>2.36,

which is requested in the first sentence of [23 Section 5].

Also, we note the following equivalence

0<∥x∥<d⟺0<s<1.0<\lVert x\rVert<d\Longleftrightarrow0<s<1.

By [23 Theorem 2.1] it holds that there are entire functions H1,H2H_1,H_2, real on the real axis, such that for all x∈R2x\in\mathbb{R}^2

f(x)=e−πh∥x∥2d2H1(∥x∥2d2),f^(x)=e−πh∥x∥2d2H2(∥x∥2d2).f(x)=e^{-\pi h\frac{\lVert x\rVert^2}{d^2}}H_1\left(\frac{\lVert x\rVert^2}{d^2}\right),\qquad \widehat{f}(x)=e^{-\pi h\frac{\lVert x\rVert^2}{d^2}}H_2\left(\frac{\lVert x\rVert^2}{d^2}\right).

As in [23 Theorem 2.1], we set

Tk(s)=e−k(s−1)k.T_k(s)=\frac{e^{-k(s-1)}}{k}.

Next, we note that the proof of [23 Lemma 6.2] defines ft=ke−kff_t=ke^{-k}f in the third sentence. The relation πh+k=πtd2\pi h+k=\pi td^2 gives

ke−ke−πhsTk(s)=e−ke−πhse−k(s−1)=e−(πh+k)s=e−πt∥x∥2=Gt(x).ke^{-k}e^{-\pi hs}T_k(s)=e^{-k}e^{-\pi hs}e^{-k(s-1)} =e^{-(\pi h+k)s}=e^{-\pi t\lVert x\rVert^2}=G_t(x).

The second sentence of Section 5 of [23] defines

F1=Tk−H1,F2=H2.F_1=T_k-H_1,\qquad F_2=H_2.

These notations imply that

Qt(x)=Gt(x)−ft(x)=ke−ke−πhs(Tk(s)−H1(s))=ke−ke−πhsF1(s).Q_t(x) = G_t(x) - f_t(x) = k e^{-k} e^{-\pi h s}\left(T_k(s) - H_1(s)\right) = k e^{-k} e^{-\pi h s}F_1(s).

Linearity of the Fourier transform yields

ft^(x)=ke−kf^(x)=ke−ke−πhsF2(s).\widehat{f_t}(x) = k e^{-k}\widehat{f}(x) = k e^{-k}e^{-\pi h s}F_2(s).

The common factor ke−ke−πhsk e^{-k}e^{-\pi h s} is strictly positive. The assertion therefore follows if we prove F1(s),F2(s)>0F_1(s),F_2(s)>0 for 0<s<10<s<1.

Step 2: identify the two estimates needed. We use the intervals [0,1/2][0,1/2] and [1/2,1][1/2,1]. Equation (5.1) of [23], applied at 11, gives

Fi(1)=Fi′(1)=0,i=1,2.F_i(1)=F_i'(1)=0,\qquad i=1,2.

The quotient defined in [23 Section 5.1] therefore simplifies to

D1,2Fi(s)=Fi(s)−Fi(1)−(s−1)Fi′(1)(s−1)2=Fi(s)(s−1)2for s≠1.D_{1,2}F_i(s)=\frac{F_i(s)-F_i(1)-(s-1)F_i'(1)}{(s-1)^2} =\frac{F_i(s)}{(s-1)^2}\qquad\text{for }s\ne1.

Therefore, to show positivity of FiF_i for 0<s<10<s<1, it suffices to prove positivity of FiF_i on [0,1/2][0,1/2] and of D1,2FiD_{1,2}F_i on [1/2,1)[1/2,1).

To match the notation of [23], the two intervals are parametrized by

s=m+yu,0≤u≤1,(m,y,ν)=(0,12,0) or (1,−12,2).(12)s=m+yu,\qquad0\le u\le1,\qquad(m,y,\nu)=\left(0,\frac{1}{2},0\right)\ \text{or}\ \left(1,-\frac{1}{2},2\right). \tag*{(12)}

The choice ν=0\nu=0 corresponds to estimating the function itself, while ν=2\nu=2 corresponds to estimating D1,2FiD_{1,2}F_i. The second parametrization runs from s=1s=1 to s=1/2s=1/2.

Step 3: apply the estimate from the proof of Lemma 5.2. For

ε=0.001−0.00011−0.000001−0.000001=0.000888>0,\varepsilon=0.001-0.00011-0.000001-0.000001=0.000888>0,

the displayed estimate near the end of the proof of [23 Lemma 5.2], applied to the two choices of (m,y,ν)(m,y,\nu) in (12), yields

Fi(s)>εfor 0≤s≤12,F_i(s)>\varepsilon\qquad\text{for }0\le s\le\frac{1}{2},
D1,2Fi(s)>εfor 12≤s<1,i=1,2.D_{1,2}F_i(s)>\varepsilon\qquad\text{for }\frac{1}{2}\le s<1,\qquad i=1,2.

We use this estimate from the proof because the statement of Lemma 5.2 asserts only nonnegativity. With Step 2 and Step 1, this yields (11) for t≥1t\ge1.

Step 4: treat 0<t<10<t<1. Put β=1/t>1\beta=1/t>1. The functions QβQ_\beta and fβ^\widehat{f_\beta} are strictly positive for 0<∥x∥<d0<\lVert x\rVert<d by the case already proved. Equation (6.4) of [23] defines the auxiliary function for tt by

ft=βQβ^.f_t=\beta\widehat{Q_\beta}.

With our Fourier convention, the Gaussian transform in dimension two is

Gβ^(x)=1βe−π∥x∥2/β=1βGt(x).\widehat{G_\beta}(x)=\frac{1}{\beta}e^{-\pi\lVert x\rVert^2/\beta}=\frac{1}{\beta}G_t(x).

Since Qβ=Gβ−fβQ_\beta=G_\beta-f_\beta, it follows that

ft=β(Gβ^−fβ^)=Gt−βfβ^,Qt=Gt−ft=βfβ^.f_t=\beta\left(\widehat{G_\beta}-\widehat{f_\beta}\right)=G_t-\beta\widehat{f_\beta},\qquad Q_t=G_t-f_t=\beta\widehat{f_\beta}.

Moreover, applying the Fourier transform twice reflects a Schwartz function. The function QβQ_\beta is radial and hence even. Therefore,

ft^(x)=βQβ^^(x)=βQβ(−x)=βQβ(x).\widehat{f_t}(x)=\beta\widehat{\widehat{Q_\beta}}(x)=\beta Q_\beta(-x)=\beta Q_\beta(x).

Both Qt=βfβ^Q_t=\beta\widehat{f_\beta} and ft^=βQβ\widehat{f_t}=\beta Q_\beta are strictly positive on 0<∥x∥<d0<\lVert x\rVert<d, because β>1>0\beta>1>0. Together with the case t≥1t\ge1, this proves the lemma. □

Remark 4.3. Lemma 4.2 makes no new computer-assisted estimate. It records the implication of the positive certified margin for the location of the first possible zero. The certificate, its error bounds, and Fourier complementation are all provided by [23].

Proof of Theorem 1.2 (B). Let C⊂R2C \subset\mathbb{R}^2 be a periodic configuration with ρ(C)=1\rho(C)=1 for which equality holds in (2) at some α0>0\alpha_0>0. Set t=α0/πt=\alpha_0/\pi, so that Et(C)=Et(Λ2)E_t(C)=E_t(\Lambda_2). Choose the auxiliary function ftf_t from Proposition 4.1, with the strictness property established in Lemma 4.2. Write CC as in (3);

C=⋃j=1N(vj+Λ),Λ a full-rank lattice in R2.C=\bigcup_{j=1}^{N}(v_j+\Lambda),\qquad\Lambda\text{ a full-rank lattice in }\mathbb{R}^2.

Then ρ(C)=1\rho(C)=1 implies vol⁡(R2/Λ)=N\operatorname{vol}(\mathbb{R}^2/\Lambda)=N. To keep the proof self-contained, we repeat the arguments leading to the linear programming bounds for energy minimization [7], which, just as the bounds for sphere packings [6], follow from the Poisson summation formula. Define

S(ξ)=∑j=1Ne2πiξ⋅vj.S(\xi)=\sum_{j=1}^{N}e^{2\pi i\xi\cdot v_j}.

Since ftf_t is a Schwartz function, the Poisson summation formula applies with absolute convergence of both sums. With our Fourier-transform convention and vol⁡(R2/Λ)=N\operatorname{vol}(\mathbb{R}^2/\Lambda)=N, it yields

∑λ∈Λft(vi−vj+λ)=1N∑ξ∈Λ∗ft^(ξ)e2πiξ⋅(vi−vj).\sum_{\lambda\in\Lambda}f_t(v_i-v_j+\lambda)=\frac{1}{N}\sum_{\xi\in\Lambda^*}\widehat{f_t}(\xi)e^{2\pi i\xi\cdot(v_i-v_j)}.

Summing over i,ji,j and using

∑i,j=1Ne2πiξ⋅(vi−vj)=∣S(ξ)∣2\sum_{i,j=1}^{N}e^{2\pi i\xi\cdot(v_i-v_j)}=\lvert S(\xi)\rvert^2

yields

1N∑i,j=1N∑λ∈Λft(vi−vj+λ)=1N2∑ξ∈Λ∗ft^(ξ)∣S(ξ)∣2.\frac{1}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda}f_t(v_i-v_j+\lambda)=\frac{1}{N^2}\sum_{\xi\in\Lambda^*}\widehat{f_t}(\xi)\lvert S(\xi)\rvert^2.

The energy in (4) omits the NN diagonal terms with i=ji=j and λ=0\lambda=0, which contribute ft(0)f_t(0) after normalization. Also note that S(0)=NS(0)=N. Hence, by (10),

1N∑i,j=1N∑λ∈Λ∖{vi−vj}ft(vi−vj+λ)=1N2∑ξ∈Λ∗∖{0}ft^(ξ)∣S(ξ)∣2+ft^(0)−ft(0)=1N2∑ξ∈Λ∗∖{0}ft^(ξ)∣S(ξ)∣2+Et(Λ2).(13)\begin{aligned} \frac{1}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda\setminus\{v_i-v_j\}}f_t(v_i-v_j+\lambda)=\frac{1}{N^2}\sum_{\xi\in\Lambda^*\setminus\{0\}}\widehat{f_t}(\xi)\lvert S(\xi)\rvert^2+\widehat{f_t}(0)-f_t(0) \\ &=\frac{1}{N^2}\sum_{\xi\in\Lambda^*\setminus\{0\}}\widehat{f_t}(\xi)\lvert S(\xi)\rvert^2+E_t(\Lambda_2). \tag*{(13)} \end{aligned}

Since Gt=ft+QtG_t=f_t+Q_t we conclude from (13)

Et(C)−Et(Λ2)=1N∑i,j=1N∑λ∈Λ∖{vi−vj}Qt(vi−vj+λ)+1N∑i,j=1N∑λ∈Λ∖{vi−vj}ft(vi−vj+λ)−Et(Λ2)=1N∑i,j=1N∑λ∈Λ∖{vi−vj}Qt(vi−vj+λ)+1N2∑ξ∈Λ∗∖{0}ft^(ξ)∣S(ξ)∣2.(14)\begin{aligned} E_t(C)-E_t(\Lambda_2)=\frac{1}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda\setminus\{v_i-v_j\}}Q_t(v_i-v_j+\lambda)+\frac{1}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda\setminus\{v_i-v_j\}}f_t(v_i-v_j+\lambda)-E_t(\Lambda_2) \\ &=\frac{1}{N}\sum_{i,j=1}^{N}\sum_{\lambda\in\Lambda\setminus\{v_i-v_j\}}Q_t(v_i-v_j+\lambda)+\frac{1}{N^2}\sum_{\xi\in\Lambda^*\setminus\{0\}}\widehat{f_t}(\xi)\lvert S(\xi)\rvert^2. \tag*{(14)} \end{aligned}

Every summand in (14) is nonnegative by (9). If Et(C)=Et(Λ2)E_t(C)=E_t(\Lambda_2), both sums on the right-hand side of (14) are zero. In particular, each nonnegative term in the first sum must vanish:

Qt(vi−vj+λ)=0Q_t(v_i-v_j+\lambda)=0

for every i,j∈{1,…,N}i,j\in\{1,\ldots,N\} and λ∈Λ\lambda\in\Lambda with i≠ji\ne j or λ≠0\lambda\ne0. By (11), every such nonzero vector vi−vj+λv_i-v_j+\lambda has length at least dd. Thus sep⁡(C)≥d\operatorname{sep}(C)\ge d, and the packing argument used in the proof of Theorem 1.2 (A) gives C≅Λ2C\cong\Lambda_2. □

Proof of Corollary 1.5. By definition, we have g(r)=f(r2)g(r)=f(r^2) for some nonzero completely monotone function ff. By Theorem 1.4,

f(s)=∫[0,∞)e−πts dμ(t)f(s)=\int_{[0,\infty)} e^{-\pi ts}\,d\mu(t)

for a nonzero positive measure μ\mu. If μ({0})>0\mu(\{0\})>0, then g(r)=f(r2)≥μ({0})>0g(r)=f(r^2)\geq\mu(\{0\})>0 for every r>0r>0, contradicting the finite-energy assumption. Thus μ({0})=0\mu(\{0\})=0, and

g(r)=f(r2)=∫(0,∞)e−πtr2 dμ(t).g(r)=f(r^2)=\int_{(0,\infty)} e^{-\pi tr^2}\,d\mu(t).

Let C⊂R2C\subset\mathbb{R}^2 be periodic of density 1. Tonelli’s theorem and (2), applied with α=πt\alpha=\pi t, give

Eg(C)−Eg(Λ2)=∫(0,∞)(Et(C)−Et(Λ2)) dμ(t)≥0.E_g(C)-E_g(\Lambda_2)=\int_{(0,\infty)}\left(E_t(C)-E_t(\Lambda_2)\right)\,d\mu(t)\geq0.

If Eg(C)=Eg(Λ2)E_g(C)=E_g(\Lambda_2), then Et(C)=Et(Λ2)E_t(C)=E_t(\Lambda_2) for μ\mu-almost every tt. Since μ≠0\mu\ne0, equality holds for at least one t>0t>0, and Theorem 1.2(B), applied with α=πt\alpha=\pi t, gives C≅Λ2C\cong\Lambda_2. □

References

References

  1. [1]S. Bernstein, Sur les fonctions absolument monotones, Acta Math. 52:1–66, 1929. doi:10.1007/BF02592679.DOI
  2. [2]L. Bétermin, Two-dimensional theta functions and crystallization among Bravais lattices, SIAM J. Math. Anal. 48(5):3236–3269, 2016. doi:10.1137/15M101614X.
  3. [3]X. Blanc and M. Lewin, The crystallization conjecture: a review, EMS Surv. Math. Sci. 2(2):255–306, 2015. doi:10.4171/EMSS/13.DOI
  4. [4]K. Böröczky Jr., Finite Packing and Covering, Cambridge Tracts in Mathematics, vol. 154, Cambridge University Press, Cambridge, 2004. doi:10.1017/CBO9780511546587.DOI
  5. [5]J. W. S. Cassels, On a problem about the Epstein zeta-function, Glasg. Math. J. 4(2):73–80, 1959. doi:10.1017/S2040618500033906. (Corrigendum, ibid. 6:116, 1963. doi:10.1017/S2040618500034833.)DOI
  6. [6]H. Cohn and N. Elkies, New upper bounds on sphere packings I, Ann. of Math. (2) 157 (2003), no. 2, 689–714. doi:10.4007/annals.2003.157.689.DOI
  7. [7]H. Cohn and A. Kumar, Universally optimal distribution of points on spheres, J. Amer. Math. Soc. 20(1):99–148, 2007. doi:10.1090/S0894-0347-06-00546-7.
  8. [8]H. Cohn, A. Kumar, S. D. Miller, D. Radchenko, and M. Viazovska, The sphere packing problem in dimension 24, Ann. of Math. (2) 185(3):1017–1033, 2017. doi:10.4007/annals.2017.185.3.8.
  9. [9]H. Cohn, A. Kumar, S. D. Miller, D. Radchenko, and M. Viazovska, Universal optimality of the E₈ and Leech lattices and interpolation formulas, Ann. of Math. (2) 196(3):983–1082, 2022. doi:10.4007/annals.2022.196.3.3.
  10. [10]J. H. Conway and N. J. A. Sloane, Sphere Packings, Lattices and Groups, 3rd ed., Grundlehren der Mathematischen Wissenschaften, vol. 290, Springer-Verlag, New York, 1999. doi:10.1007/978-1-4757-6568-7.DOI
  11. [11]P. H. Diananda, Notes on two lemmas concerning the Epstein zeta-function, Glasg. Math. J. 6(4):202–204, 1964. doi:10.1017/S2040618500035036.
  12. [12]V. Ennola, A lemma about the Epstein zeta-function, Glasg. Math. J. 6(4):198–201, 1964. doi:10.1017/S2040618500035024.DOI
  13. [13]V. Ennola, On a problem about the Epstein zeta-function, Math. Proc. Cambridge Philos. Soc. 60(4):855–875, 1964. doi:10.1017/S0305004100038330.DOI
  14. [14]M. Faulhuber, I. Shafkulovska, and I. Zlotnikov, A note on energy minimization in dimension 2, Proc. Amer. Math. Soc. Ser. B 11:664–679, 2024. doi:10.1090/bproc/247.
  15. [15]L. Fejes Tóth, Über die dichteste Kugellagerung, Math. Z., 48:676–684, 1942. doi:10.1007/BF01180035.
  16. [16]L. Fejes Tóth, Lagerungen in der Ebene, auf der Kugel und im Raum, Die Grundlehren der Mathematischen Wissenschaften, vol. 65, Springer-Verlag, Berlin, 1953. doi:10.1007/978-3-662-01206-2.
  17. [17]G. B. Folland, Harmonic analysis in phase space, Annals of Mathematics Studies, number 122, Princeton University Press, 1989. doi:10.1515/9781400882427.
  18. [18]D. P. Hardin and E. B. Saff, Minimal Riesz energy point configurations for rectifiable d-dimensional manifolds, Adv. Math. 193(1):174–204, 2005. doi:10.1016/j.aim.2004.05.006.
  19. [19]D. P. Hardin and N. Tenpas, Universally Optimal Periodic Configurations in the Plane, Discrete Anal. 26, 2025. doi:10.48550/arXiv.2307.15822.
  20. [20]J. Jorgenson and S. Lang, The Ubiquitous Heat Kernel, in: B. Engquist und W. Schmid (Hrsg.), Mathematics Unlimited — 2001 and Beyond, Springer, Berlin, Heidelberg, 2001, S. 655–683. doi:10.1007/978-3-642-56478-9_34.DOI
  21. [21]T. Leblé, The hexagonal lattice is universally locally optimal, arXiv preprint arXiv:2511.03353, 2025. doi:10.48550/arXiv.2511.03353.
  22. [22]H. L. Montgomery, Minimal theta functions, Glasgow Math. J. 30(1):75–85, 1988. doi:10.1017/S0017089500007047.DOI
  23. [23]OpenAI, Universal optimality of the triangular lattice, preprint, 2026. Available at: https://github.com/openai/math/blob/main/preprints/Universal-optimality-of-the-triangular-lattice-September-23-2026
  24. [24]M. Petrache and S. Serfaty, Next order asymptotics and renormalized energy for Riesz interactions, J. Inst. Math. Jussieu 16(3):501–569, 2017. doi:10.1017/S1474748015000201.
  25. [25]D. Radchenko, Fourier uniqueness pairs and interpolation formulas, in: G. P. Curbera, M. L. Puertas, A. Suárez-Llorens (Eds.), European Congress of Mathematics (Sevilla, July 15–19, 2024), EMS Press, pp. 975–990, 2026. doi:10.4171/9ECM/40.DOI
  26. [26]R. A. Rankin, A minimum problem for the Epstein zeta-function, Glasg. Math. J. 1(4):149–158, 1953. doi:10.1017/S2040618500035668.DOI
  27. [27]S. Serfaty, Smooth Ginzburg-Landau Vortices, Zurich Lectures in Advanced Mathematics, European Mathematical Society (EMS), Zürich, 2015. doi:10.4171/152.
  28. [28]A. Thue, Über die dichteste Zusammenstellung von kongruenten Kreisen in einer Ebene. Norske Vid. Selsk. Skr. 1:1–9, 1910. Available at: https://www.biodiversitylibrary.org/page/15174094
  29. [29]W. J. Ventevogel and B. R. A. Nijboer, On the configuration of systems of interacting particle with minimum potential energy per particle, Physica A 98(1):274–288, 1979. doi:10.1016/0378-4371(79)90178-X.DOI
  30. [30]M. S. Viazovska, The sphere packing problem in dimension 8, Ann. of Math. (2) 185(3):991–1015, 2017. doi:10.4007/annals.2017.185.3.7.DOI
  31. [31]D. V. Widder, The Laplace Transform, Princeton Mathematical Series, vol. 6, Princeton University Press, Princeton, 1941. doi:10.1515/9781400876457.

Paper details

vlnc:W_burudc2u
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 7

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