Research · White paper · The electron gas

Lower bounds for jellium, and the limit of smearing bounds

We prove, with computer assistance, that classical jellium (point charges in a uniform neutralising background) has energy per particle eJel≥−1.448178733e_{\rm Jel}\ge-1.448178733 in three dimensions (density 1, kernel 1/∣x∣1/|x|). Lieb and Narnhofer proved −35(9π/2)1/3≈−1.4507927586-\tfrac35(9\pi/2)^{1/3}\approx-1.4507927586 in 1975 1. We found no published improvement, but read only the abstract of Caillol's finite-coupling bound, and not Rosenfeld's or Totsuji's, so ours may not be the first. No smearing bound of the Lieb–Narnhofer kind, a radial charge profile with non-negative Fourier transform and non-negative potential deficit, exceeds −1.44800699488-1.44800699488; the body-centred cubic (bcc) crystal has −1.4442307515-1.4442307515. In two dimensions such a bound comes within 6.4×10−296.4\times10^{-29} of the triangular-lattice energy, an upper bound for eJele_{\rm Jel}, after Marzo's improvement of 3.6×10−63.6\times10^{-6} 2 and unrefereed claims of the exact value 3, 4, 5. For the Lieb–Oxford constant, Li reported lower constants first 6.

The question

What is the lowest energy per electron of a classical electron gas at zero temperature, held by a uniform positive background? In three dimensions it equals the energy of the classical uniform electron gas 7, 8, the low-density limit against which the quantum gas is compared 9.

Jellium is NN point charges at x1,…,xNx_1,\dots,x_N and a uniform neutralising background of density 1 on a bounded set Ω\Omega with ∣Ω∣=N|\Omega|=N. With v(x)=1/∣x∣v(x)=1/|x| in R3\mathbb R^3 and v(x)=−log⁡∣x∣v(x)=-\log|x| in R2\mathbb R^2 the energy is

EΩ(x)=∑i<jv(xi−xj)−∑i∫Ωv(xi−y) dy+12∬Ω×Ωv(x−y) dx dy.E_\Omega(x)=\sum_{i<j}v(x_i-x_j)-\sum_i\int_\Omega v(x_i-y)\,dy+\frac12\iint_{\Omega\times\Omega}v(x-y)\,dx\,dy .

The ground-state energy per particle is eJel=lim⁡N→∞min⁡EΩN/Ne_{\rm Jel}=\lim_{N\to\infty}\min E_{\Omega_N}/N, with ΩN=N1/dΩ1\Omega_N=N^{1/d}\Omega_1 and Ω1\Omega_1 convex of volume 1. Dividing a three-dimensional energy at density 1 by (4π/3)1/3=1.6119919540(4\pi/3)^{1/3}=1.6119919540 gives units of e2/ae^2/a, a=(3/(4πρ))1/3a=(3/(4\pi\rho))^{1/3}.

The bcc crystal gives eJel≤ζBCC(1)=−1.44423075152697…e_{\rm Jel}\le\zeta_{\rm BCC}(1)=-1.44423075152697\ldots 7, 10, where ζBCC(s)=∑x≠0∣x∣−s\zeta_{\rm BCC}(s)=\sum_{x\ne0}|x|^{-s} runs over the bcc lattice of density 1, continued to s=1s=1. Equality is conjectured. Muratov and Novaga report an unrefereed computer-assisted proof that bcc is optimal among Bravais lattices 11. Recent work 10, 12, 13 states Lieb and Narnhofer's lower bound for jellium 1, −35(9π/2)1/3≈−1.4508-\tfrac35(9\pi/2)^{1/3}\approx-1.4508 at density 1, that is −0.9 e2/a-0.9\,e^2/a. Their argument smears each charge into a uniform ball of volume 1, the optimal positive smearing 12.

In two dimensions the same argument gives −(3/8+14log⁡π)=−0.6611824715-(3/8+\tfrac14\log\pi)=-0.6611824715 12, 14, 15. Marzo improved it by about 3.6×10−63.6\times10^{-6} (arXiv, 2025) 2, through the constant Clog⁡=2eJel+12log⁡4πC_{\log}=2e_{\rm Jel}+\tfrac12\log4\pi of the logarithmic energy on the sphere 15. The triangular lattice of density 1 has energy per particle etri=−0.6605587142e_{\rm tri}=-0.6605587142.

What we found

Theorem 1 (Smearing bound). Let h1,…,hph_1,\dots,h_p be real radial finite combinations of uniform shell (circle) measures with ∑k(∫dhk)2=1\sum_k(\int dh_k)^2=1. Put u=∑khk∗hku=\sum_kh_k*h_k, f=v∗uf=v*u, w=v−fw=v-f, and assume w≥0w\ge0 on Rd∖{0}\mathbb R^d\setminus\{0\}. Then for all NN, all bounded Ω\Omega with ∣Ω∣=N|\Omega|=N and all x∈(Rd)Nx\in(\mathbb R^d)^N, EΩ(x)≥−N2f(0)−NW+W2∣Ω−r∣E_\Omega(x)\ge-\tfrac N2f(0)-NW+\tfrac W2|\Omega_{-r}|, where W=∫w=cd⟨∣y∣2⟩uW=\int w=c_d\langle|y|^2\rangle_u, c3=2π/3c_3=2\pi/3, c2=π/2c_2=\pi/2, ⟨g⟩u=∫g du\langle g\rangle_u=\int g\,du, rr is the radius of the support of uu and Ω−r={y:B(y,r)⊂Ω}\Omega_{-r}=\{y:B(y,r)\subset\Omega\}. Hence eJel≥−J(u)e_{\rm Jel}\ge-J(u) with J(u)=12⟨v⟩u+cd2⟨∣y∣2⟩uJ(u)=\tfrac12\langle v\rangle_u+\tfrac{c_d}2\langle|y|^2\rangle_u. For d=2d=2 the conclusion eJel≥−J(u)e_{\rm Jel}\ge-J(u) also holds for a radial Schwartz function uu with ∫u=1\int u=1, u^≥0\hat u\ge0 (where u^(k)=∫u(x)e−ik⋅xdx\hat u(k)=\int u(x)e^{-ik\cdot x}dx) and w≥0w\ge0.

Result 2 (Three dimensions). eJel≥−1.448178733e_{\rm Jel}\ge-1.448178733. The certificate is u=∑1≤a,b≤100Qab saδ∗sbδ/∑a,bQabu=\sum_{1\le a,b\le100}Q_{ab}\,s_{a\delta}*s_{b\delta}/\sum_{a,b}Q_{ab} with δ=3/250\delta=3/250, sRs_R the uniform probability measure on the sphere of radius RR, and Q=LLT/q2Q=LL^{T}/q^2 for an integer matrix LL and an integer qq. It satisfies w≥0w\ge0 and has J(u)=1.44817873288J(u)=1.44817873288, rounded outward.

The bound exceeds Lieb–Narnhofer's by 0.00261402560.0026140256, 39.8 %39.8\,\% of the distance 0.00656200710.0065620071 to bcc, and reads eJel≥−0.8983784 e2/ae_{\rm Jel}\ge-0.8983784\,e^2/a.

Result 3 (Two dimensions). Let κ\kappa be a rational within 3×10−313\times10^{-31} of 1/(4π)1/(4\pi), t=r2/(4κ)t=r^2/(4\kappa) and u(r)=14πκ∑p=064cp e−ttp/p!u(r)=\frac1{4\pi\kappa}\sum_{p=0}^{64}c_p\,e^{-t}t^p/p! with rational cpc_p, ∑pcp=1\sum_pc_p=1. For a certified choice of the cpc_p, u^≥0\hat u\ge0, w≥0w\ge0 and J(u)−Jtri<6.4×10−29J(u)-J_{\rm tri}<6.4\times10^{-29}, where Jtri=−etri=0.6605587142140189574949345979…J_{\rm tri}=-e_{\rm tri}=0.6605587142140189574949345979\ldots Hence etri−6.4×10−29≤eJel≤etrie_{\rm tri}-6.4\times10^{-29}\le e_{\rm Jel}\le e_{\rm tri}.

The upper bound is the triangular-lattice competitor 15.

Theorem 4 (Ceiling). Let d=3d=3. Let uu be a radial finite signed measure with ∫u=1\int u=1 and finite second moment, f=v∗uf=v*u with f(0)f(0) finite, u^≥0\hat u\ge0 and w=v−f≥0w=v-f\ge0. Let g=1+γ≥0g=1+\gamma\ge0 with γ\gamma bounded, radial and of compact support, and suppose its structure factor S(k)=1+γ^(k)≥0S(k)=1+\hat\gamma(k)\ge0 for all kk. Then E(g)+J(u)=12∫w g d3x+12(2π)−3∫S(k)f^(k) d3k≥0E(g)+J(u)=\tfrac12\int w\,g\,d^3x+\tfrac12(2\pi)^{-3}\int S(k)\hat f(k)\,d^3k\ge0, with E(g)=12∫γ(y)/∣y∣ d3yE(g)=\tfrac12\int\gamma(y)/|y|\,d^3y. In particular −J(u)≤E(g)-J(u)\le E(g).

Result 5 (Ceiling in three dimensions). There is such a gg, piecewise constant on 770 cells out to radius 64 and equal to 1 beyond, with E(g)=−1.44800699488215…E(g)=-1.44800699488215\ldots. Every bound eJel≥−J(u)e_{\rm Jel}\ge-J(u) of the family of Theorem 4 therefore has −J(u)≤−1.44800699488-J(u)\le-1.44800699488.

With Result 2, the best bound of this family lies in [−1.448178733, −1.44800699488][-1.448178733,\,-1.44800699488]. The bcc value is 3.8×10−33.8\times10^{-3} above its upper end, so the family closes at most 42.5 %42.5\,\% of the Lieb–Narnhofer-to-bcc distance.

The Lieb–Oxford constant

The Lieb–Oxford constant cLOc_{\rm LO} is the best constant in Eind[ρ]≥−c∫ρ4/3E_{\rm ind}[\rho]\ge-c\int\rho^{4/3} for all NN and all ρ\rho with ∫ρ=N\int\rho=N. Here Eind[ρ]E_{\rm ind}[\rho] is the least expected Coulomb repulsion of NN-particle states of density ρ\rho, minus D(ρ,ρ)=12∬ρ(x)ρ(y)/∣x−y∣D(\rho,\rho)=\tfrac12\iint\rho(x)\rho(y)/|x-y|. Lewin, Lieb and Seiringer report cLO≤1.5765c_{\rm LO}\le1.5765 from a floating-point evaluation of their inequality 16. Li's repository reports exact-rational certificates, committed on 2026-10-18 (UTC), from 1.5447013101.544701310 with one covariance channel to 1.5341733601.534173360 with five 6. Li has priority; our certificate, from his one-channel family with a signed pair, is higher.

The method of 16 gives cLO≤32(2ID2)1/3c_{\rm LO}\le\tfrac32(2ID^2)^{1/3} for a probability measure μ\mu, a signed measure ν\nu of mass 1, both finite sums of uniform shells of radii rir_i with weight vectors p\mathbf p and ν\boldsymbol\nu (radius 0 allowed), and any ff with Ψ(a,b)≤f(a)+f(b)\Psi(a,b)\le f(a)+f(b) for all a,b>0a,b>0. Here D=D(μ,μ)D=D(\mu,\mu), I=4π∫0∞f(r)r−5drI=4\pi\int_0^\infty f(r)r^{-5}dr, Ψ(a,b)=a3b3∑ijWij ω(a/ri,b/rj)\Psi(a,b)=a^3b^3\sum_{ij}W_{ij}\,\omega(a/r_i,b/r_j), W=pνT+νpT−ννTW=\mathbf p\boldsymbol\nu^T+\boldsymbol\nu\mathbf p^T-\boldsymbol\nu\boldsymbol\nu^T, and ω(a/r,b/s)\omega(a/r,b/s) is 1 minus the mean of 1/∣x−y∣1/|x-y| over uniform spheres of radii r/ar/a and s/bs/b with centres a unit distance apart.

Result 6 (Certificate). Let μ\mu be the radial probability measure on 12 shells with radii in [0.121,0.443][0.121,0.443], and ν\nu the radial signed measure of total mass 1 with a Dirac mass 0.3940.394 at the origin and nine shells of radii up to 1.01.0, with weights from −0.7229-0.7229 to 1.82121.8212, four of them negative (radii in units of ρ−1/3\rho^{-1/3}). Then D(μ,μ)=1.4407261477D(\mu,\mu)=1.4407261477, I=0.2693539685I=0.2693539685 and cLO≤32(2ID2)1/3=1.5569094067c_{\rm LO}\le\tfrac32(2ID^2)^{1/3}=1.5569094067.

Proposition 7 (Floor). In this setting let ζ1,…,ζm\zeta_1,\dots,\zeta_m be signed shell measures of total mass 0 with weight vectors ζℓ\boldsymbol\zeta_\ell, put W=pνT+νpT−ννT−∑ℓζℓζℓTW=\mathbf p\boldsymbol\nu^T+\boldsymbol\nu\mathbf p^T-\boldsymbol\nu\boldsymbol\nu^T-\sum_\ell\boldsymbol\zeta_\ell\boldsymbol\zeta_\ell^T, and let ff be a majorant of the resulting Ψ\Psi. Li counts m+1m+1 channels (m=0m=0 is Result 6). Then 32(2ID2)1/3≥35(9π/2)1/3=1.4507927586\tfrac32(2ID^2)^{1/3}\ge\tfrac35(9\pi/2)^{1/3}=1.4507927586.

Why it is true

Write θ=∑iδxi−1Ω\theta=\sum_i\delta_{x_i}-\mathbf 1_\Omega, a neutral measure. For radial real hkh_k, hk∗θh_k*\theta is neutral with finite energy, and the Coulomb (logarithmic) energy of a neutral measure is non-negative because the Fourier transform of the kernel is. Summing over kk, the energy with vv replaced by ff is at least −N2f(0)-\tfrac N2f(0). By Newton's theorem v∗u=vv*u=v outside the support of uu, so ww has compact support.

Result 3 has infinitely supported profiles, covered by the last sentence of Theorem 1. There −log⁡∣z∣=12∫0∞(e−t∣z∣2−e−t) dt/t-\log|z|=\tfrac12\int_0^\infty(e^{-t|z|^2}-e^{-t})\,dt/t writes ff as a mixture of Gaussians convolved with uu, whose Fourier transforms are non-negative, and a perimeter estimate for convex ΩN\Omega_N makes the boundary term o(N)o(N).

The bound is a linear programme of Cohn–Elkies and Cohn–Kumar type 17, 18 in sum-of-squares form, and Theorem 4 is its weak duality, the jellium analogue of Proposition 9.3 of 18. For a lattice Λ\Lambda of covolume 1 with dual lattice Λ∗\Lambda^*, Poisson summation gives its energy per particle in its background, eΛ=−J(u)+12∑ℓ∈Λ∖{0}w(ℓ)+12∑G∈Λ∗∖{0}f^(G)e_\Lambda=-J(u)+\tfrac12\sum_{\ell\in\Lambda\setminus\{0\}}w(\ell)+\tfrac12\sum_{G\in\Lambda^*\setminus\{0\}}\hat f(G), with both sums non-negative. Sharpness at Λ\Lambda needs double zeros of ww and u^\hat u at the norms of Λ\Lambda and Λ∗\Lambda^*, which the 2D certificate has to within 10−4010^{-40} at the first 16 norms.

The Lieb–Oxford method applies Onsager's inequality D(F−η,F−η)≥0D(F-\eta,F-\eta)\ge0 to the smeared charge FF; it holds for every finite-energy η\eta (inequality (16) of 16), so ν\nu may be signed. For Proposition 7, I=∫R3f(∣z∣)∣z∣−7d3zI=\int_{\mathbb R^3}f(|z|)|z|^{-7}d^3z, and the diagonal a=ba=b of Ψ≤f(a)+f(b)\Psi\le f(a)+f(b) gives f(a)≥12Ψ(a,a)f(a)\ge\tfrac12\Psi(a,a). Half the integral of Ψ(∣z∣,∣z∣)∣z∣−7d3z\Psi(|z|,|z|)|z|^{-7}d^3z is 2π3∫r2dμ\tfrac{2\pi}3\int r^2d\mu whatever the neutral channels, so I≥2π3∫r2dμI\ge\tfrac{2\pi}3\int r^2d\mu. As 32(2ID2)1/3=min⁡t>0[tD+t−2I]\tfrac32(2ID^2)^{1/3}=\min_{t>0}[tD+t^{-2}I], the constant is at least min⁡[D(μ′,μ′)+2π3∫r2dμ′]\min[D(\mu',\mu')+\tfrac{2\pi}3\int r^2d\mu'] over all radial probability measures μ′\mu', and the uniform ball attains 1.45079275861.4507927586.

How it was checked

Results 2, 3, 5 and 6 are computer-assisted proofs; Theorems 1 and 4 and Proposition 7 are proved, and their proofs were re-derived step by step in a separate check, with the constants recomputed in separately written code. The 3D certificate is checked by an exact rational test that r w(r)r\,w(r) is non-negative on 200 intervals where it is quadratic, with a ball enclosure of π\pi; a separately written checker reproduced J=1.4481787328783099509J=1.4481787328783099509. The 2D and dual certificates are checked in ball arithmetic, and separately written checkers accept all 15 2D certificates and all eight dual certificates (truncation radii 6 to 64, Result 5 the last) with the same values. Controls: 400 random Bravais lattices, minimised cells and clusters of up to 150 particles stay above Result 2, and discs of up to 31453 particles confirm the normalisation of etrie_{\rm tri}.

For Result 6, a ball-arithmetic branch and bound proves the majorant inequality on a compact region, with analytic bounds near the axes and ∞\infty. A separately written float64 branch and bound, without outward rounding, closed the whole domain too (evidence, not proof); DD, II and the constant were recomputed to 19 digits with separately written code. For Li's certificates we recomputed the constants from his integers, each at or below his value, and ran the float64 code on the whole node domain, not his exact verifier. The proofs have not been peer reviewed, and the existence of eJele_{\rm Jel} is quoted.

What it does not show

  • This is classical jellium at zero temperature. We state no consequence for quantum jellium or a real material. Neither bcc nor the triangular lattice is shown optimal.
  • Result 5 concerns the family of Theorem 4, not every two-point method, and is loose at its upper end: all eight dual certificates respect a bound on the radius-weighted total variation of gg, and without it the certificates of radii 6 and 12 lie 8.4×10−58.4\times10^{-5} and 8.6×10−58.6\times10^{-5} lower. Its cost at radius 64 was not measured.
  • The gg of Result 5 is not a pair correlation. It vanishes below r=1.131r=1.131 yet has 13.3913.39 neighbours on average within r<1.175r<1.175. Neighbours that close and 1.1311.131 apart subtend at least 57.5∘57.5^\circ, which 13 points on a sphere cannot reach (their best separation is 57.14∘57.14^\circ 19), so at most 12 fit. Its hard core is not a proved property of minimisers.
  • Lieb's hard core (3/4π)1/3=0.6204(3/4\pi)^{1/3}=0.6204 and the radial inequalities of isotropic averaging hold for minimisers away from the boundary. The eight certified gg satisfy both, so imposing them can raise the bound of Result 2 by at most 1.72×10−41.72\times10^{-4}. A discretised jellium version of the three-point inequality of 20 returned no gain, as its construction forces at the resolutions we could run, so whether three-point bounds beat two-point ones is open.

Earlier and parallel work

We searched arXiv, Zenodo, GitHub web pages and the web, 2026-11-05 to 2026-11-15.

Three dimensions. We found no published improvement and no earlier ceiling for smearing bounds. We read only the abstract of Caillol's finite-coupling bound for the one-component plasma 21, which places it above Mermin's (1968), Lieb–Narnhofer's and Totsuji's 22, and a search snippet of Rosenfeld's 23. We do not know their large-coupling limits and do not claim that −1.448178733-1.448178733 is the first improvement.

Method. Petrache and Serfaty 24 showed that the Cohn–Kumar conjecture implies minimality of the same lattices for jellium energies. We found no earlier statement of Theorem 1 in sum-of-squares form as a Cohn–Kumar-type linear programme for jellium with its background. Lieb's hard core is an unpublished result of Lieb, proved in 12 (Lemma 25); isotropic averaging is due to Leblé and Thoma 25, 26.

Two dimensions. The OpenAI Math release 3 (unrefereed, by an unnamed model) holds manuscripts, dated 2026-09-23 to 2026-09-26, that claim triangular minimality for completely monotone functions of squared distance and, by heat-kernel comparison, for the Coulomb energy and the jellium minimum; sharp Gaussian Fourier minorants; and, by Voronoi cells with a finite computation, the Coulomb energy. By 24 the Gaussian result implies eJel=etrie_{\rm Jel}=e_{\rm tri}. Callens (2026-10-26) re-ran a finite certificate of the Coulomb manuscript in an AI-produced, unaudited record that "does not prove, verify or refute" the claim 27. Cynamon (2026-10-30) builds a sharp Fourier pair for completely monotone potentials of squared distance with finite triangular reference energy 28. Lei 4 (Lean proof unfinished) claims the same value and Kitamura 5 (two README sections labelled AI generated) a sphere conjecture equivalent to it through Clog⁡C_{\log}, both in October 2026. None that we read contains an auxiliary-function certificate for the Coulomb kernel with its background. Result 3 depends on none of them and is not a priority result for the value; its gap to etrie_{\rm tri} is 6.4×10−296.4\times10^{-29}, where Marzo's bound leaves 6.2×10−46.2\times10^{-4}.

Lieb–Oxford. The floor is the Lieb–Narnhofer constant, quoted in 16 after inequality (31) for general states. We found no earlier statement of Proposition 7; Li's manuscript was not available and may contain it.

References

  1. 1.E. H. Lieb and H. Narnhofer, "The thermodynamic limit for jellium", J. Stat. Phys. 12, 291–310 (1975). doi.org/10.1007/BF01012066
  2. 2.J. Marzo, "An improved lower bound for the logarithmic energy on S²", arXiv:2506.01660 (2025; no journal reference on the arXiv page). arxiv.org/abs/2506.01660
  3. 3.OpenAI Math release, github.com/openai/math (first commit 2026-10-06; manuscripts "produced by an unnamed internal OpenAI model", not refereed). github.com/openai/math
  4. 4.Y. Lei, "Lean 4 formalization: the triangular lattice minimizes the 2D Coulomb renormalized energy", GitHub repository yuhaolei-ctrl/Triangular-Lattice-Coulomb-Lean (work in progress, the proof is unfinished). github.com/yuhaolei-ctrl/Triangular-Lattice-Coulomb-Lean
  5. 5.K. Kitamura, GitHub repository KitaKen1/riesz-energy-spheres (Lean sources, a proof note and a README with two sections labelled AI generated, October 2026). github.com/KitaKen1/riesz-energy-spheres
  6. 6.Y. Li, "Data and code for An Improvement on the Lieb-Oxford Bound (2026)", GitHub repository pkupkupku123/improvement-lieb-oxford-bound, two commits of 2026-10-18 (UTC); the manuscript, submitted to Phys. Rev. A, is not in the repository. github.com/pkupkupku123/improvement-lieb-oxford-bound
  7. 7.M. Lewin, E. H. Lieb and R. Seiringer, "Floating Wigner crystal with no boundary charge fluctuations", Phys. Rev. B 100, 035127 (2019). doi.org/10.1103/PhysRevB.100.035127
  8. 8.C. Cotar and M. Petrache, "Equality of the jellium and uniform electron gas next-order asymptotic terms for Coulomb and Riesz potentials", arXiv:1707.07664 (2017). arxiv.org/abs/1707.07664
  9. 9.M. Lewin, E. H. Lieb and R. Seiringer, "Statistical mechanics of the uniform electron gas", J. Éc. polytech. Math. 5, 79–116 (2018). doi.org/10.5802/jep.64
  10. 10.R. L. Frank, M. Lewin and R. Seiringer, "Liquid drop model for nuclear matter in the low density limit", arXiv:2507.14012v2 (2026; to appear in Comm. Pure Appl. Math.). arxiv.org/abs/2507.14012
  11. 11.C. B. Muratov and M. Novaga, "BCC optimality for the three-dimensional Coulomb lattice energy", arXiv:2610.29732 (v1, 2026-10-21; no journal reference on the arXiv page). arxiv.org/abs/2610.29732
  12. 12.M. Lewin, "Coulomb and Riesz gases: the known and the unknown", J. Math. Phys. 63, 061101 (2022). doi.org/10.1063/5.0086835
  13. 13.M. A. Csirik, "Equality of the jellium and the uniform electron gas energies", arXiv:2610.20910 (2026). arxiv.org/abs/2610.20910
  14. 14.R. R. Sari and D. Merlini, "On the ν-dimensional one-component classical plasma: the thermodynamic limit problem revisited", J. Stat. Phys. 14, 91–100 (1976). doi.org/10.1007/BF01011761
  15. 15.A. B. Lauritsen, "Floating Wigner crystal and periodic jellium configurations", J. Math. Phys. 62, 083305 (2021). doi.org/10.1063/5.0053494
  16. 16.M. Lewin, E. H. Lieb and R. Seiringer, "Improved Lieb–Oxford bound on the indirect and exchange energies", Lett. Math. Phys. 112, 92 (2022). doi.org/10.1007/s11005-022-01584-5
  17. 17.H. Cohn and N. Elkies, "New upper bounds on sphere packings I", Ann. of Math. 157, 689–714 (2003). doi.org/10.4007/annals.2003.157.689
  18. 18.H. Cohn and A. Kumar, "Universally optimal distribution of points on spheres", J. Amer. Math. Soc. 20, 99–148 (2007). doi.org/10.1090/S0894-0347-06-00546-7
  19. 19.O. R. Musin and A. S. Tarasov, "The strong thirteen spheres problem", Discrete Comput. Geom. 48, 128–141 (2012). doi.org/10.1007/s00454-011-9392-2
  20. 20.H. Cohn, D. de Laat and A. Salmon, "Three-point bounds for sphere packing", arXiv:2206.15373 (2022). arxiv.org/abs/2206.15373
  21. 21.J. M. Caillol, "A new (and better) lower bound for the excess internal energy of the one-component plasma", J. Chem. Phys. 111, 9695–9699 (1999), with erratum J. Chem. Phys. 112, 6940 (2000). doi.org/10.1063/1.480302
  22. 22.H. Totsuji, "Lower bounds for thermodynamic quantities of two- and three-dimensional one-component plasmas", Phys. Rev. A 19, 1712–1716 (1979). doi.org/10.1103/PhysRevA.19.1712
  23. 23.Y. Rosenfeld, "Mean spherical model for the D-dimensional ν-component classical plasma", Phys. Rev. A 25, 1206–1208 (1982). doi.org/10.1103/PhysRevA.25.1206
  24. 24.M. Petrache and S. Serfaty, "Crystallization for Coulomb and Riesz interactions as a consequence of the Cohn–Kumar conjecture", arXiv:1908.09714 (2019). arxiv.org/abs/1908.09714
  25. 25.E. Thoma, "Overcrowding and separation estimates for the Coulomb gas", arXiv:2210.05902 (2022). arxiv.org/abs/2210.05902
  26. 26.S. Serfaty, "Lectures on Coulomb and Riesz gases", arXiv:2407.21194 (2024). arxiv.org/abs/2407.21194
  27. 27.X. Callens, "The Sandier–Serfaty triangular-lattice conjecture: an independent reproduction of a claimed finite certificate, an adversarial audit of the claimed reduction, and a Lean statement (no proof of the conjecture is claimed)", Zenodo (2026-10-26). doi.org/10.5281/zenodo.23653827
  28. 28.B. Z. Cynamon, "Sharp mixed-potential Fourier pairs and planar universal optimality", Zenodo (2026-10-30; read as version v993 on 2026-11-15). doi.org/10.5281/zenodo.23827153