Signs of high order derivatives for the theta and Epstein zeta functions and application

Deng Kaixin; Luo Senping

REVIEW 3 major objections 4 minor 30 references

Signs of high order derivatives for the theta and Epstein zeta functions and application

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read For theta and Epstein zeta functions, the paper proves strict sign control on second and third mixed derivatives over the relevant fundamental region, with a corollary locating derivative minima on a boundary arc.

desk verdict New sign results for higher-order theta/zeta derivatives, but the third-derivative proof rests on a false algebraic identity; the result is likely true and worthy of review, with a repairable gap. read the letter →

arxiv 2501.01265 v1 pith:7JVFO65C submitted 2025-01-02 math.AP math-phmath.MP

Deng Kaixin , Luo Senping This is my paper · ORCID

classification math.APmath-phmath.MP MSC 11F2711H0611M41

keywords thetafunctionsEpsteinzetafunctionlatticeenergymixedpartialderivativessignoffundamentaldomainminimizationhexagonal

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper aims to establish strict sign control on higher-order mixed derivatives of the two central lattice-energy functionals, the $\theta$ function and the Epstein zeta function, over the region where the hexagonal lattice is the conjectured minimizer. It proves that the second mixed derivative $\frac{\partial^2}{\partial x \partial y}$ is strictly positive and the third mixed derivative $\frac{\partial^3}{\partial x \partial y^2}$ is strictly negative in the relevant domain, for every Gaussian parameter $\alpha>0$ and every exponent $s>1$. These are higher-order analogues of the classical first-derivative sign inequalities. A corollary locates the minima of certain first and second derivatives on a one-parameter boundary arc, sharpening the search in lattice-minimization problems. If the theorem holds, it gives new rigidity constraints on the shape of these energy functionals near the hexagonal minimizer.

What carries the argument

The engine of the proof is the one-dimensional $\theta$ function $\vartheta(X;Y)=\sum_{n\in\mathbb{Z}} e^{-\pi n^2X}e^{2\pi i nY}$, together with sharp bounds on the quotients $\vartheta_Y(X;kY)/\vartheta_Y(X;Y)$, $\vartheta_{XY}/\vartheta_Y$, and $\vartheta_{XXY}/\vartheta_Y$ obtained by Poisson summation and explicit series estimates. These bounds feed into explicit expressions for the mixed derivatives of $\theta(\alpha;z)$, where the parameter $y/\alpha$ determines which family of estimates applies. The sign of the whole expression is then reduced to checking finitely many numerical constants at the endpoints $x=1/2$, $y=3/5$, and $y=\sqrt{3}/2$. A second load-bearing object is the integral identity linking $\zeta$ to $\theta$, which converts each $\theta$ sign into the corresponding zeta sign without new estimates.

What would settle it

Compute the explicit series for $\mu,\nu,\omega$ and their quotients on $x\in[1/2,\infty)$ and test whether each monotonicity assertion holds; if any fails, the replacement of quotients by their $x=1/2$ values is invalid. Independently, evaluate $\frac{\partial^2}{\partial x\partial y}\theta(\alpha;x+iy)$ at $x=0.49$, $y=0.6$, $\alpha=2$: the theorem predicts a strictly positive value, so a nonpositive result would refute Theorem 1.1(1).

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is Theorem 1.1: for $\alpha>0$ and $s>1$, the mixed derivative $\frac{\partial^2}{\partial x \partial y}$ of both $\theta(\alpha;z)$ and $\zeta(s;z)$ is strictly positive whenever $0<x<1/2$ and $y\ge 3/5$, whereas $\frac{\partial^3}{\partial x \partial y^2}$ is strictly negative for $z$ in the fundamental domain $D_G=\{z:|z|>1,\ 0<x<1/2\}$. The signs are proved first for the $\theta$ function, using an exponential expansion into one-dimensional $\theta$ terms, and then transferred to the zeta function by the integral representation $\zeta(s;z)=\frac{\pi^s}{\Gamma(s)}\int_0^\infty (\theta(\alpha;z)-1)\alpha^{s-1}\,d\alpha$, which preserves signs because the prefactor is positive.

Load-bearing premise

The load-bearing premise is a set of monotonicity assertions about the series $\mu,\nu,\omega$ and the quotients $(1+\mu)/(1-\mu)$, $(1+\nu)/(1-\mu)$, $(1+\omega)/(1-\mu)$, $(1+\hat\nu)/(1+\hat\mu)$, and $(1+\hat\omega)/(1+\hat\mu)$ on $x\ge 1/2$; the paper states these without proof, and the endpoint bounds in (3.32) and (3.40) require them.

Editorial extensions

If this is right

For $\alpha>0$ and $s>1$, strict positivity of $\frac{\partial^2}{\partial x\partial y}$ holds on the whole strip $0<x<1/2$, $y\ge 3/5$, not just on the fundamental domain.
The minima of $\frac{\partial}{\partial x}$ and $\frac{\partial^2}{\partial x\partial y}$ over $D_G$ are attained on the boundary arc $\Gamma=\{z=e^{i\theta}: \theta\in[\pi/3,\pi/2]\}$, reducing a two-dimensional minimization to a one-dimensional one.
The same signs transfer from theta to zeta for every $s>1$, so zeta-function energy landscapes inherit the second- and third-derivative information.
Because $\frac{\partial^3}{\partial x\partial y^2}<0$ on $D_G$, the second mixed derivative is strictly decreasing in $y$ across the fundamental domain, a new rigidity constraint in the neighborhood of the hexagonal point.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The same two-case comparison of $y/\alpha$ may yield sign theorems for higher mixed derivatives, provided the analogous one-dimensional quotient bounds can be established; the paper itself does not claim such an extension.
The monotonicity assumptions on the series $\mu,\nu,\omega$ and their quotients are checkable numerically to arbitrary precision because the series are explicit and convergent; a failure would locate a counterexample before any full evaluation is needed.
If the third-derivative sign persists under smooth perturbations of the interaction potential, it could constrain the local energy landscape near hexagonal lattices in non-monotone or multi-species models; this goes beyond the paper's stated scope.
A natural next test is whether analogous mixed-derivative sign theorems hold for sums or differences of theta functions, for which similar expansions are already available.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The thing to know about arXiv:2501.01265: it proves a genuinely new sign result for mixed second and third derivatives of theta and Epstein zeta functions, but the third-derivative part has a concrete algebraic error in Lemma 3.5, Case B. The error is localized and likely repairable, but as written the proof of Theorem 1.1(2) doesn't go through.

What's new: the first-order derivative signs go back to Rankin, Cassels, Ennola, Diananda, Montgomery. The mixed second and third derivative signs are not in the literature, as far as I can tell. The estimates on ϑ_XXY/ϑ_Y in Lemma 2.2 also look new. If the signs hold, the corollary reducing minimization over D_G to a boundary arc is a useful monotonicity tool for the crystallization work. The derivation of the zeta-function case from the theta case via the Mellin integral is clean.

The proof of the second mixed derivative, Lemma 3.4, seems plausible. It splits into y/α ≥ 1/2 and ≤ 1/2, and the bounds are concrete. The weaker spot is Lemma 3.5. In Case B, equation (3.46) asserts an identity that is false. With t=α/y≥2, the bracketed expression E is order t^{-2}, while the two right-hand sides are order t^{-1} and t. The subsequent substitution of the upper bound (3/2+16πe^{-2π})/(1-8πe^{-2π}) * α/y is not justified. This is not a corner case; α ≥ 2y is an open region for every z in the fundamental domain. The error looks repairable because a correct crude bound on E is much smaller, but the manuscript doesn't contain it.

There are also unproved monotonicity claims: in both Lemmas 3.4 and 3.5 the bounding at x=1/2 relies on µ, ν, ω and certain quotients being monotone for x ≥ 1/2. These are asserted without proof. That's less damaging—they're likely true and checkable—but they're load-bearing. The numerical values in (3.33), (3.43) are given as 'by computation' with no code or reproducible detail; minor in this genre.

Net: the first part of Theorem 1.1 looks solid; the second part is conditional on a corrected Case B bound. This deserves a serious referee, not a desk reject. I'd send it out with instructions to fix (3.46), justify the monotonicity, and make the numerics reproducible. If the fix works, it's a publishable contribution to the crystallization literature.

Referee Report

3 major / 4 minor

Summary. The paper proves two sign statements: for α>0 and s>1, in the strip 0<x<1/2, y≥3/5 the mixed derivative ∂²/∂x∂y of the theta and Epstein zeta functions is positive, and in the fundamental domain D_G the third derivative ∂³/∂x∂y² is negative. The proof expands the theta function as a sum of one-dimensional theta functions (Lemmas 3.1–3.3), applies estimates for quotients of theta derivatives (Lemmas 2.1–2.5), and transfers the theta signs to the zeta function via the integral representation (3.47). The paper also derives a corollary locating minima of ∂_x and ∂²_{xy} on the arc Γ.

Significance. If the main theorem is correct, it extends the classical first-order sign results of Rankin, Cassels, Ennola, Diananda, and Montgomery to second and third mixed derivatives, giving new monotonicity information for lattice-energy functionals relevant to the crystallization literature. The paper's strategy is natural: explicit derivative expansions, a reduction to one-dimensional theta estimates, and a theta-to-zeta transfer. The stated bounds are quantitative and falsifiable, and the derivative expressions in Lemmas 3.2 and 3.3 are explicit and useful. However, as written the proof contains a false algebraic identity in a central step and several unproved monotonicity and numerical assertions, so the significance is contingent on a corrected derivation.

major comments (3)

[Lemma 3.5, Eq. (3.46)] Equation (3.46) is algebraically false. With t=α/y≥2, the subtracted quantity E=(3/2)(y/α)^2+4π^2e^{-πα/y}(y/α)^3−4π(y/α)^2e^{-πα/y} equals t^{-2}(3/2+4π^2e^{-πt}/t−4πe^{-πt}), which is O(t^{-2}); it is not equal to ((3/2)/t+4πt^3e^{-πt})/(1−4πte^{-πt}) nor to ((3/2+4πt^2e^{-πt})/(1−4πte^{-πt}))t, which are O(t^{-1}) and O(t), respectively. Therefore the subsequent bound using (3/2+16πe^{-2π})/(1−8πe^{-2π}) is unjustified, and the displayed proof of the Case B lower bound 1/50 does not hold. Because y/α≤1/2 is an open region of D_G, this issue is load-bearing for Theorem 1.1(2). The error appears localized and likely repairable with a correct O(t^{-2}) estimate, but it must be fixed.
[Lemmas 3.4 and 3.5, around (3.31)–(3.32) and (3.39)–(3.43)] The proof relies on unproved monotonicity of the functions μ, ν, ω and of the quotients (1+μ)/(1−μ), (1+ν)/(1−μ), (1+ω)/(1−μ) (decreasing) and (1+ν̂)/(1+μ̂), (1+ω̂)/(1+μ̂) (increasing) for x≥1/2. These statements are asserted immediately after (3.31) and again before (3.41), and they are exactly what justifies replacing those quotients by their endpoint values at x=1/2. No proof or reference is given. Since the positivity of the final bounds (3.34), (3.37), and (3.43) depends on these assertions, the authors need to include a verification (termwise differentiation should suffice) or cite a source that contains it.
[Eqs. (3.33), (3.35), (3.41), (3.43), and ε2≤10^{-4}] Several numerical constants are used as rigorous inequalities but are only stated after 'by computation' or 'it can be bounded': the values in (3.33), the bound 0.039 in (3.35), the values 1.1042..., 0.8884..., 0.4435... in (3.41), the bound ≤2/50 in (3.43), and ε2≤10^{-4} after (3.44). These numbers enter directly into the final inequalities, so the manuscript should provide either explicit analytic estimates with error control or a reproducible computation (for example interval arithmetic) for each of them. In particular the ε2 bound is essential in Case B of Lemma 3.5.

minor comments (4)

[Proposition 1.1] In parts (3) and (4) the function θ(s,z) is written where θ(α,z) is meant; the same notation appears in the abstract.
[Section 3, Eq. (3.48)] The step from (3.47) to (3.48) differentiates under the integral and then concludes that ζ_xy and ζ_xyy have the same sign as the corresponding theta derivatives; the manuscript should spell out the standard dominated-convergence justification, especially for ζ_xyy, since the integrand is not uniformly bounded near α=0 without using the Fourier identity.
[Lemma 2.5, Eq. (2.26)] In the display after (2.25), the argument 'α' is used in f'_n(α;Y) and in the denominator, but the variable is a; this should be corrected to avoid confusion.
[References] References [24] and [28] are identical; this duplicate should be removed or replaced with distinct entries if the authors intended two different works.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the higher-order sign theorems for theta and zeta are derived from independent one-dimensional theta estimates and the classical Mellin representation, not from their own conclusions.

full rationale

The proof chain begins with the Poisson-summation formula for the one-dimensional theta function and with external bound lemmas, then manipulates explicit derivative formulas (Lemmas 3.2 and 3.3). The self-citations to Luo-Wei and Deng-Luo, such as Lemma 2.1 and Lemma 2.3(3)-(4), are parameter-free technical estimates whose stated hypotheses do not contain Theorem 1.1, so they are independent support rather than assumed conclusions. The reduction to α ≥ 1 uses the exact identity θ(1/α; z) = α θ(α; z), and the zeta-function statement follows from the theta statement by the Mellin representation in (3.47)-(3.48), which preserves signs. No fitted parameter is renamed as a prediction, no ansatz is introduced through self-citation, and the target sign inequalities never appear among the inputs. The false algebraic identity in (3.46) identified by the skeptic is a correctness defect in the bounding argument, not a circular dependency, because it is asserted within the proof rather than assumed as an equivalent input. Therefore no circular step is present.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new entities; it relies on standard functions and a set of unproved-by-this-paper estimates from the authors' and collaborators' prior work, plus several assertions that are effectively computational claims.

assumptions (5)

domain assumption Known quotient estimates for the 1-d theta function (Lemma 2.1) from Luo-Wei and Deng-Luo.
Invoked without proof in Lemmas 3.4 and 3.5 to bound terms in θ_xy and θ_xyy.
domain assumption Known sign and bound properties of the Gaussian sums in Lemma 2.3 from Luo-Wei.
Used to derive Lemma 2.2 and Lemma 2.4.
domain assumption Exponential expansion of the theta function in Lemma 3.1 from Montgomery and Luo-Wei.
Provides the starting point for all derivative computations in Section 3.
ad hoc to paper Monotonicity of the quotient functions μ, ν, ω and their ratios for x ≥ 1/2.
Stated without proof or reference in Lemmas 3.4 and 3.5; used to replace quotients by their endpoint values at x=1/2.
ad hoc to paper Numerical bounds in (3.33), (3.35), (3.43), and ε2 ≤ 10^-4.
Asserted 'by computation' without reproducible derivation; these bounds are needed to show the lower bounds are positive.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Signs of high order derivatives for the theta and Epstein zeta functions and application." pith.science (2026). https://pith.science/paper/7JVFO65C

@misc{pith2026250101265,
  author       = {Pith},
  title        = {Pith review of: Signs of high order derivatives for the theta and Epstein zeta functions and application},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7JVFO65C}},
  note         = {Machine review of arXiv:2501.01265}
}

read the original abstract

In the 1950s, 1960s and 1988, number theorists Rankin \cite{Ran1953}, Cassels \cite{Cas1959}, Ennola \cite{Enn1964a}, Diananda \cite{Dia1964}, and Montgomery \cite{Mon1988} derived the signs of first order derivatives of Epstein zeta and theta functions, respectively. In this note, we shall derive the signs of higher order derivatives of such functions. Application to lattice minimization problems will be given.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 29 canonical work pages

[1]

T. M. Apostol, Modular functions and Dirichlet series in number theory, Springer-Verlag, Berlin Heidelberg, 1976

work page 1976
[2]

B´ etermin and P

L. B´ etermin and P. Zhang, Minimization of energy per par ticle among Bravais lattices in R2: Lennard-Jones and Thomas-Fermi cases, Commun. Contemp. Math. , 17(6) (2015), 1450049

work page 2015
[3]

B´ etermin, Two-dimensional theta functions and crystallization among Bravais lattices, SIAM J

L. B´ etermin, Two-dimensional theta functions and crystallization among Bravais lattices, SIAM J. Math. Anal. , 48(5) (2016), 3236-3269

work page 2016
[4]

B´ etermin, Local variational study of 2d lattice ener gies and application to Lennard-Jones type interactions, Nonlinearity, 31(9) (2018), 3973-4005

L. B´ etermin, Local variational study of 2d lattice ener gies and application to Lennard-Jones type interactions, Nonlinearity, 31(9) (2018), 3973-4005

work page 2018
[5]

B´ etermin, Minimizing lattice structures for Morse p otential energy in two and three dimensions, J

L. B´ etermin, Minimizing lattice structures for Morse p otential energy in two and three dimensions, J. Math. Phys., 60(10) (2019), 102901

work page 2019
[6]

B´ etermin and M

L. B´ etermin and M. Petrache, Dimension reduction techn iques for the minimization of theta functions on lattices, J. Math. Phys. , 58(7) (2017), 071902

work page 2017
[7]

B´ etermin, M

L. B´ etermin, M. Faulhuber and H. Kn¨ upfer, On the optimality of the rock-salt structure among lat tices with charge distributions, Math. Models Methods Appl. Sci. , 31(2): 293-325, 2021

work page 2021
[8]

B´ etermin and M

L. B´ etermin and M. Petrache, Optimal and non-optimal la ttices for non-completely monotone interaction potentials, Anal. Math. Phys. , 9(4): 2033-2073, 2019

work page 2019

Show all 30 references

[9]

B´ etermin, On energy ground states among crystal latt ice structures with prescribed bonds, J

L. B´ etermin, On energy ground states among crystal latt ice structures with prescribed bonds, J. Phys. A , 54(24): 245202, 2021

2021
[10]

B´ etermin, Eﬀect of periodic arrays of defects on lat tice energy minimizers

L. B´ etermin, Eﬀect of periodic arrays of defects on lat tice energy minimizers. Ann. Henri Poincar´ e22 (2021), no. 9, 2995-3023

2021
[11]

B´ etermin, L

L. B´ etermin, L. De Luca and M. Petrache, Crystallizati on to the square lattice for a two-body potential, Arch. Ration. Mech. Anal. , 240, 987-1053 (2021)

2021
[12]

B´ etermin, M

L. B´ etermin, M. Friedrich and U. Stefanelli, Lattice g round states for embedded-atom models in 2D and 3D, Lett. Math. Phys. , 111, 107 (2021)

2021
[13]

Blanc and M

X. Blanc and M. Lewin, The crystallization conjecture: a review, EMS Surv. Math. Sci . 2 (2015), no. 2, pp. 255-306

2015
[14]

J. W. S. Cassels, On a problem of Rankin about the Epstein zeta function, Proc. Glasgow Math. Assoc. , 4 (1959), 73-80. (Corrigendum, ibid. 6 (1963), 116.) SIGNS OF HIGH ORDER DERIV ATIVES 11

1959
[15]

P. H. Diananda, Notes on two lemmas concerning the Epste in zeta-function, Proc. Glasgow Math. Assoc. 6 (1964), 202-204

1964
[16]

Deng and S

K. Deng and S. Luo, Minimizing lattice energy and hexago nal crystallization, arXiv:2411.17199v1

arXiv
[17]

Ennola, A lemma about the Epstein zeta function, Proc

V. Ennola, A lemma about the Epstein zeta function, Proc. Glasgow Math. Assoc. 6 (1964), 198-201

1964
[18]

Ennola, On a problem about the Epstein zeta-function , Proc

V. Ennola, On a problem about the Epstein zeta-function , Proc. Cambridge Philos, Soc. 60(1964), 855-875

1964
[19]

Evans, A fundamental region for Hecke modular group, J

R. Evans, A fundamental region for Hecke modular group, J. Number Theory , 5(2) (1973), 108-115

1973
[20]

S. Luo, X. Ren and J. W ei, Non-hexagonal lattices from a t wo species interacting system, SIAM J. Math. Anal. , 52(2) (2020), 1903-1942

2020
[21]

Luo and J

S. Luo and J. W ei, On minima of sum of theta functions and a pplication to Mueller-Ho conjecture, Arch. Ration. Mech. Anal. , 243 (2022), no. 1, 139-199

2022
[22]

Luo and J

S. Luo and J. W ei, On minima of diﬀerence of theta functio ns and application to hexagonal crystallization, Math. Ann. , 387 (2023), no. 1-2, 499-539

2023
[23]

Luo and J

S. Luo and J. W ei, On lattice hexagonal crystallization for non-monotone potentials, J. Math. Phys. , 65 (2024), no. 7, Paper No. 071901, 28 pp

2024
[25]

Luo and J

S. Luo and J. W ei, On a variational model for the continuo us mechanics exhibiting hexagonal to square phase transitions, arxiv:2312.02497

arXiv
[26]

Luo and J

S. Luo and J. W ei, On minima of diﬀerence of Epstein zeta f unctions and exact solutions to Lennard-Jones lattice energy, arxiv:2212.10727

arXiv
[27]

Luo and J

S. Luo and J. W ei, On extremum of quotient of Theta and Zet a functions and their applications, preprint
[28]

S. Luo, J. W ei and W. Zou, On universally optimal lattice phase transitions and energy minimizers of completely monotone potentials, arxiv:2110.08728

arXiv
[29]

Montgomery, Minimal theta functions, Glasgow Math

H. Montgomery, Minimal theta functions, Glasgow Math. J. , 30 (1988), 75-85

1988
[30]

Radin, Low temperature and the origin of crystalline symmetry, Int

C. Radin, Low temperature and the origin of crystalline symmetry, Int. J. Mod. Phys. B , Vol. 01, No. 05n06, pp. 1157-1191 (1987)

1987
[31]

R. A. Rankin, A minimum problem for the Epstein zeta-fun ction, Proc. Glasgow Math. Assoc. , 1 (1953), 149-158. (K. Deng) School of Mathematics and statistics, Jiangxi Normal Univers ity, Nanchang, 330022, China (S. Luo) School of Mathematics and statistics, Jiangxi Normal Uni...

1953

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

T. M. Apostol, Modular functions and Dirichlet series in number theory, Springer-Verlag, Berlin Heidelberg, 1976

work page 1976

[2] [2]

B´ etermin and P

L. B´ etermin and P. Zhang, Minimization of energy per par ticle among Bravais lattices in R2: Lennard-Jones and Thomas-Fermi cases, Commun. Contemp. Math. , 17(6) (2015), 1450049

work page 2015

[3] [3]

B´ etermin, Two-dimensional theta functions and crystallization among Bravais lattices, SIAM J

L. B´ etermin, Two-dimensional theta functions and crystallization among Bravais lattices, SIAM J. Math. Anal. , 48(5) (2016), 3236-3269

work page 2016

[4] [4]

B´ etermin, Local variational study of 2d lattice ener gies and application to Lennard-Jones type interactions, Nonlinearity, 31(9) (2018), 3973-4005

L. B´ etermin, Local variational study of 2d lattice ener gies and application to Lennard-Jones type interactions, Nonlinearity, 31(9) (2018), 3973-4005

work page 2018

[5] [5]

B´ etermin, Minimizing lattice structures for Morse p otential energy in two and three dimensions, J

L. B´ etermin, Minimizing lattice structures for Morse p otential energy in two and three dimensions, J. Math. Phys., 60(10) (2019), 102901

work page 2019

[6] [6]

B´ etermin and M

L. B´ etermin and M. Petrache, Dimension reduction techn iques for the minimization of theta functions on lattices, J. Math. Phys. , 58(7) (2017), 071902

work page 2017

[7] [7]

B´ etermin, M

L. B´ etermin, M. Faulhuber and H. Kn¨ upfer, On the optimality of the rock-salt structure among lat tices with charge distributions, Math. Models Methods Appl. Sci. , 31(2): 293-325, 2021

work page 2021

[8] [8]

B´ etermin and M

L. B´ etermin and M. Petrache, Optimal and non-optimal la ttices for non-completely monotone interaction potentials, Anal. Math. Phys. , 9(4): 2033-2073, 2019

work page 2019