The non-uniform electron gas

Mihály A. Csirik ∗,1 and Andre Laestadius 1,2 · 2026

Read The non-uniform electron gas by Mihály A. Csirik ∗,1 and Andre Laestadius 1,2 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-21200.txt.

Introduction

Csirik and Laestadius give a firm math footing to the non-uniform electron gas. This gas was an early tool in density functional theory. Workers used it to build gradient approximations for exchange and correlation. Their quantum form rests on the grand-canonical Levy-Lieb functional. Their classical form rests on the strictly correlated electrons functional. They set both up as true thermodynamic limits and work out their basic traits. The gas is non-uniform because a lattice-periodic background density is imposed on it.

Catalog record

CREATORMihály A. Csirik ∗,1 and Andre Laestadius 1,2
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE204095
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.21200

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The non-uniform electron gas Mihály A. Csirik ∗,1 and Andre Laestadius 1,2 Abstract . The non-uniform (or inhomogeneous) electron gas has received much attention in many-body quantum mechanics and quantum chemistry in the early days of density functional theory, mainly as a theoretical device to construct gradient approximations via linear response theory. In this article, motivated by the recent works of Lewin, Lieb and Seiringer, we propose a definition of the quantum (resp. classical) non-uniform electron gas through the use of the grand-canonical Levy–Lieb functional (resp. the grand-canonical strictly correlated electrons functional), establish these systems as rigorous thermodynamic limits and analyze their basic properties. The non-uniformity of the gas comes from an arbitrary lattice-periodic background density. 1. Introduction Before the widespread acceptance of the name “density functional theory” early researchers of the field referred to what we call DFT today as the theory of the “inhomogeneous electron gas” [ 16 , 26 ] . As far as we can tell, the (weakly) inhomogeneous electron gas is regarded a “fictitious system”, or a stepping stone, and not a “real” physical system like the uniform electron gas (UEG). Furthermore, it is always understood in a perturbative sense, i.e. through its linear response [ 11 , 9 ] . These density perturbations are measured relative to the constant density of the homogeneous gas. In [ 9 ] “both the case of a localized perturbation as well as some periodic structure” are claimed to be covered. The treatment of a localized perturbation in the thermodynamic limit remains somewhat unclear. Hence, we will consider the case of a lattice-periodic inhomogeneity in this work. Rather than relying on linear response theory, we treat the inhomogeneous electron gas as a genuinely nonlinear (i.e. non‑perturbative) system. The weakly inhomogeneous electron gas may be considered within this nonlinear framework as a perturbation of the homogeneous gas by a slowly varying inhomogeneity. Henceforth, we shall adopt the term non-uniform electron gas (NUEG) in order to better align with the existing mathematical physics literature pertaining to the UEG [ 18 ] . The main reason for this is that we will also employ density functionals to define the NUEG energy. The alternative path would be to consider the thermodynamic limit of an electronic ground-state problem like in the case of the Jellium [ 23 ] . In the Jellium model the external potential is obviously constant. However, it is unclear what external potential (or whether such potential exists) would generate a prescribed lattice-periodic inhomogeneity as a ground-state density. This issue is related to the so-called v -representability problem, which is poorly understood at the present—there is no analytic description of the v -representability set except in 1D on a finite interval with Neumann boundary conditions [ 5 , 4 ] . The approach of [ 18 ] elegantly avoids these complications by defining the UEG energy through the use of density functionals and demanding the density to be constant. The price to pay is that we lose the connection with the ground-state problem and have to deal with the constrained optimization problem, which may not even have the usual Schrödinger equation as its Euler–Lagrange equation. The said connection is still unresolved for the quantum UEG, but is settled for the classical case [ 6 , 19 ] . Our definition of the NUEG in the 3D Coulomb case involves a fixed lattice-periodic inhomogeneity \zeta:\mathbb{R}^{3}\to\mathbb{R}_{+} having natural integrability and regularity properties. This \zeta replaces the constant density \rho_{0} in the definitions of [ 18 ] based on the strictly-correlated electrons (SCE) functional (classical case) and the grand-canonical Levy–Lieb functional [ 22 ] (quantum case). In order to obtain a well-behaved energy per volume, we average the energy over the translations and rotations of \zeta . More precisely, the (3D Coulomb) indirect energy per volume of the classical NUEG with inhomogeneity \zeta in a bounded domain \Omega\subset\mathbb{R}^{3} is taken to be e_{\Omega}^{\mathrm{cl}}(\zeta)=\frac{1}{|\Omega|}\int_{\mathrm{SO}(3)}\mathrm{d}R\mathchoice{{\vbox{\hbox{$\textstyle-$ }}\kern-7.83337pt}}{{\vbox{\hbox{$\scriptstyle-$ }}\kern-6.11674pt}}{{\vbox{\hbox{$\scriptscriptstyle-$ }}\kern-5.48965pt}}{{\vbox{\hbox{$\scriptscriptstyle-$ }}\kern-5.31259pt}}\!\int_{\mathbb{R}^{3}}\mathrm{d}a\,\Bigg[\inf_{\begin{subarray}{c}\mathbb{P}\,\text{GC prob.}\\ \rho_{\mathbb{P}}=\mathbb{1}_{\Omega}\zeta(R(\cdot-a))\end{subarray}}\sum_{n\geqslant 2}\sum_{1\leqslant j<k\leqslant n}\int_{\mathbb{R}^{3n}}\frac{\mathbb{P}_{n}(\mathrm{d}x_{1},\ldots,\mathrm{d}x_{n})}{|x_{j}-x_{k}|}\\ -\frac{1}{2}\iint_{\mathbb{R}^{3}\times\mathbb{R}^{3}}\frac{\mathbb{1}_{\Omega}(x)\zeta(R(x-a))\mathbb{1}_{\Omega}(y)\zeta(R(y-a))}{|x-y|}\,\mathrm{d}x\mathrm{d}y\Bigg] in the classical case and analogously in the quantum case, which we denote by e_{\Omega,\eta_{\delta}}^{\hbar}(\zeta) . Here, a regularization function \eta_{\delta} must be used, because quantum densities cannot be cut off sharply. Noting that the a -integrand above is lattice-periodic and locally integrable, the mean value over \mathbb{R}^{3} exists and equals the mean value over a unit cell of the lattice. This is a “floating crystal-like” construction, where the lattice is allowed to float in the container \Omega . Under the sole assumption that \zeta^{4/3} is locally integrable (and that \sqrt{\zeta} is in H_{\mathrm{loc}}^{1} in the quantum case) we can prove that the thermodynamic limit of e_{\Omega}^{\mathrm{cl}}(\zeta) (resp. e_{\Omega,\eta_{\delta}}^{\hbar}(\zeta) ) exists for rather general domain sequences. We denote the limiting functional by e_{\mathrm{NUEG}}^{\mathrm{cl}}(\zeta) (resp. e_{\mathrm{NUEG}}^{\hbar}(\zeta) ). In the quantum case, we also offer a slightly different definition based on a transition region near the boundary where the density is allowed to “relax” in an arbitrary manner analogously to [ 18 ] . We prove the equivalence of the two definitions. Clearly, by putting \zeta\equiv\rho_{0} we recover the UEG energy exactly, even at finite volume. We consider the slowly varying limits e_{\mathrm{NUEG}}^{\mathrm{cl}/\hbar}(\zeta(\lambda\cdot)) as \lambda\to 0 and establish the rate of its convergence to the UEG energy dictated by the local density approximation (LDA) both in the classical-, and the quantum case. These follow from an adaptation of the method of [ 20 ] . In addition to establishing the NUEG as a physically reasonable system with expected properties, our work provides a complementary perspective on DFT—in fact, instead of integrable densities , the functionals e_{\mathrm{NUEG}}^{\mathrm{cl}/\hbar}(\zeta) introduced involve locally integrable inhomogeneities \zeta that are not globally integrable. It would be interesting to extend the allowed set of inhomogeneities to a broader class, such as to some space of almost periodic functions. On the technical side, we improve a very useful spatial decoupling estimate in Theorem 5.9 that appeared first in [ 20 ] , which is a crucial tool in the quantum case. Organization of the paper. In Section 2 , we recall the necessary definitions and facts for our discussion. In Section 3 , we present our main results in detail. Section 3.1 contains the definition of the classical non-uniform electron gas through its thermodynamic limit ( Theorem 3.1 ). We also establish the local density approximation for slowly varying inhomogeneities ( Theorem 3.2 ). The proofs of these results are provided in Section 4 . The quantum case is covered in Section 3.2 , our main results being Theorem 3.3 and Theorem 3.4 , which are analogs of the classical case. The proofs for the quantum gas may be found in Section 5 . Finally, the Appendices contain the proofs of some auxiliary results. Acknowledgments. The authors are grateful to Erik I. Tellgren for many helpful discussions. A.L. and M.A.Cs. have received funding from the ERC-2021-STG under grant agreement No. 101041487 REGAL. A.L. was also funded by the Research Council of Norway through CoE Hylleraas Centre for Quantum Molecular Sciences Grant No. 262695. 2. Preliminary notions 2.1. Notations Throughout the paper, we set C_{1}=(-\frac{1}{2},\frac{1}{2})^{d} for the origin-centered d -dimensional open unit cube. Its scalings will be denoted as C_{\ell}=\ell C_{1} for any side length \ell>0 . We will make use of

Read the complete text