The quantitative semi-classical limit of a large Fermi system at zero temperature

Esteban Cárdenas · 2025

Read The quantitative semi-classical limit of a large Fermi system at zero temperature by Esteban Cárdenas (2025) 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-2505-24706.txt.

Introduction

Esteban Cárdenas, writing alone, measures how fast a cold gas of fermions becomes classical. This 2025 paper studies a large three-dimensional Fermi system at zero temperature, in the joint limit where interactions average out and quantum ripples shrink. The ground state's Wigner function, a phase-space portrait of the quantum state, should approach the classical Thomas-Fermi profile, and the question is the rate.

The main contribution is the rate itself, a quantitative bound in the semiclassical parameter rather than a bare statement of convergence. A key input is a recent family of commutator estimates from Hartree theory, which control how badly position and momentum fail to commute in the mean-field setting. Singular interactions are allowed, up to the strength of the Coulomb force between charges. The paper sharpens a foundational approximation of quantum statistical mechanics into one with an error bar.

Catalog record

CREATOREsteban Cárdenas
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE82 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-24706

Browse the full library · About the Tower of Babel

Full text

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

The quantitative semi-classical limit of a large Fermi system at zero temperature Esteban Cárdenas Abstract . In this article we consider a large system of fermions in a combined mean-field and semi-classical limit, in three dimensions. We investigate the convergence of the Wigner function of the ground state, towards the classical Thomas-Fermi theory. The main novelty of the present article is quantifying the convergence rate with respect to the semi-classical parameter. One of the main ingredients is a recent result on the validity of semi-classical commutator estimates satisfied by the Hartree theory. Singular potentials, up to the Coulomb interaction, are included. 1. Introduction 1.1. Introduction In this article, we consider N (spinless) fermions moving in \mathbb{R}^{d} . The Hilbert space of the system is the antisymmetric tensor product. \mathscr{H}_{N}\equiv\bigwedge^{N}L^{2}(\mathbb{R}^{d})\ . (1.1) We consider the N -body Hamiltonian with a combined mean-field and semi-classical scaling {\rm H}_{N}\equiv\sum_{i=1}^{N}\Big(-\hbar^{2}\Delta_{x_{i}}+U(x_{i})\Big)+\frac{1}{N}\sum_{i<j}V(x_{i}-x_{j})\ ,\qquad\hbar\equiv N^{-1/d}\,. (1.2) Here U:\mathbb{R}^{d}\rightarrow\mathbb{R} and V:\mathbb{R}^{d}\rightarrow\mathbb{R} account for an external trapping potential, and a pair potential. They are assumed to be sufficiently regular, so that {\rm H}_{N} is self-adjoint in its natural domain. The parameter \hbar\in(0,1) can be understood as an effective Planck constant, and the scaling is chosen so that the kinetic energy per particle remains bounded. In particular, the present N -body problem is naturally linked with a semi-classical limit \hbar\rightarrow 0. Let \Psi\in\mathscr{H}_{N} be a normalized state. We define the one-particle reduced density matrix as the operator \gamma_{\Psi}:L^{2}(\mathbb{R}^{d})\rightarrow L^{2}(\mathbb{R}^{d}) with kernel \displaystyle\gamma_{\Psi}(x,y)\equiv N\int_{\mathbb{R}^{d(N-1)}}\Psi(x,x_{2},\cdots x_{N})\overline{\Psi}(y,x_{2},\cdots,x_{N})dx_{2}\cdots dx_{N}\ , (1.3) here (x,y)\in\mathbb{R}^{d}\times\mathbb{R}^{d} . In particular, \gamma_{\Psi} is self-adjoint, trace-class, and verifies 0\leqslant\gamma_{\Psi}\leqslant\bm{1}\ . (1.4) For short, we refer to an operator \gamma which verifies these properties a density matrix . Here, we have the normalization \textnormal{Tr}\gamma_{\Psi}=N and we call \varrho_{\Psi}(x)=N^{-1}\gamma_{\Psi}(x,x) the position density. We are interested in the limit of \gamma_{\Psi_{N}} as N\rightarrow\infty when \Psi_{N} is an approximate ground state of {\rm H}_{N} \left\langle\Psi_{N},{\rm H}_{N}\Psi_{N}\right\rangle=\inf\sigma_{\mathscr{H}_{N}}({\rm H}_{N})+N\varepsilon_{N}\ , (1.5) for a sequence \varepsilon_{N}\rightarrow 0 at a certain rate. Due to the semi-classical nature of the scaling under consideration, the limit of \gamma_{\Psi_{N}} is better understood in terms of its Wigner function. This is the real-valued function on phase space \mathbb{R}^{d}\times\mathbb{R}^{d} defined for a general density matrix as f_{\gamma}(x,p)\equiv\int_{\mathbb{R}^{d}}\gamma\big(x+\tfrac{1}{2}y\,,\,x-\tfrac{1}{2}y\big)e^{-i\hbar^{-1}y\cdot p}dy\ , (1.6) with normalization (2\pi)^{-d}\int_{\mathbb{R}^{2d}}f_{\gamma}(x,p)dxdp=1. We also introduce the Weyl quantization map, which assigns to a real-valued function f on \mathbb{R}^{d}\times\mathbb{R}^{d} the self-adjoint operator \gamma_{f} on L^{2}(\mathbb{R}^{d}) with kernel \gamma_{f}(x,y)=\frac{1}{(2\pi\hbar)^{d}}\int_{\mathbb{R}^{d}}f\left(\frac{x+y}{2},p\right)e^{-i\hbar^{-1}p\cdot(x-y)}dp\ . (1.7) It is well-known [ 11 ] that under very general conditions, the large N limit of f_{\gamma_{\Psi_{N}}} can be described by Thomas-Fermi theory. More precisely, the limit function on phase space corresponds to the generalized Fermi ball f_{\rm TF}(x,p)\equiv\mathds{1}(|p|^{2}\leqslant{\rm C}_{\rm TF}\,\varrho_{\rm TF}(x)^{2/d}) (1.8) where {\rm C}_{\rm TF}\equiv 4\pi^{2}|B_{\mathbb{R}^{d}}(0,1)|^{-2/d} is a constant, and \varrho_{\rm TF}(x) is a measure on \mathbb{R}^{d} which verifies the Thomas-Fermi equation \varrho_{\rm TF}(x)={\rm C}_{\rm TF}^{-d/2}\big(U(x)+V*\varrho_{\rm TF}(x)-\mu\big)_{-}^{d/2}\ . (1.9) Here, \mu\in\mathbb{R} is a Lagrange multiplier (chemical potential) which guarantees the normalization \int_{\mathbb{R}^{d}}\varrho_{\rm TF}(x)dx=1 . In particular, \varrho_{\rm TF} is the minimizer of the Thomas-Fermi functional \displaystyle\mathcal{E}_{\rm TF}(\varrho) (1.10) \displaystyle\equiv\frac{d\,{\rm C}_{\rm TF}}{d+2}\int_{\mathbb{R}^{d}}\varrho(x)^{1+\frac{2}{d}}dx+\int_{\mathbb{R}^{d}}U(x)\varrho(x)dx+\frac{1}{2}\int_{\mathbb{R}^{2d}}\varrho(x)V(x-y)\varrho(y)dxdy which is minimized over all 0\leqslant\varrho\in L^{1}\cap L^{\frac{d+2}{2}}(\mathbb{R}^{d}) of unit mass. The validity of the limit f_{\gamma_{\Psi_{N}}}\rightarrow f_{\rm TF} is often refered to as convergence of states , and was proven in great generality by Lewin, Fournais and Solovej [ 11 ] , for a wide variety of potentials ( U,V ), in any dimension d\geqslant 1 , with magnetic fields, and also for higher-order marginals. This is contrast to the convergence of the position densities \lim_{N\rightarrow\infty}\varrho_{\Psi_{N}}=\varrho_{\rm TF}, whose analysis goes back to the pioneering works of Lieb and Simon [ 18 , 19 ] . In particular, the convergence of states addressed in [ 11 ] was proven in the following topology (assuming here uniqueness of minimizers) \lim_{N\rightarrow\infty}\langle\varphi,f_{\gamma_{\Psi_{N}}}\rangle=\left\langle\varphi,f_{{\rm TF}}\right\rangle\ ,\qquad\forall\varphi\in W^{1,\infty}(\mathbb{R}^{d}\times\mathbb{R}^{d})\ . (1.11) The mode of convergence was then improved by the author [ 7 ] , and it was shown that ( 1.11 ) holds in the strong sense, with respect to various norms of interest. For instance, for all s<0 \lim_{N\rightarrow\infty}\|f_{\gamma_{\Psi_{N}}}-f_{{\rm TF}}\|_{H^{s}(\mathbb{R}^{2d})}=0\ , (1.12) although with slightly stronger conditions on the pair (U,V) . We would also like to mention that the convergence of states obtained in [ 11 ] has been extended in various other directions: [ 17 ] considered a system of fermions at positive temperature; [ 12 ] studied various scalings for the magnetic field; and [ 13 ] analyzed the case of anyons with almost fermionic statistics. We would also like to mention [ 14 ] which considers the convergence of the position densities in the sense of \Gamma convergence. The main goal of this article is to investigate the quantitative convergence of states . As a corollary of our main result in Theorem 1 we obtain the L^{2} convergence \|f_{\gamma_{\Psi_{N}}}-f_{\rm TF}\|_{L^{2}(\mathbb{R}^{2d})}\leqslant C\hbar^{\delta/2}|\!\ln\hbar|^{1/4} (1.13) in terms of an explicit power \delta>0 , where up to Coulomb interactions V(x)=|x|^{-1} are included. To the authors best knowledge, there are no previous results in the literature which are quantitative in the semi-classical parameter. In particular, previous work on the convergence of states [ 11 , 12 , 17 , 13 , 7 ] relies on compactness arguments and, therefore, are not suitable for proving quantitative estimates. The only exception corresponds to systems on the torus \mathbb{T}^{d} with no external trap, for which the problem becomes translation invariant \rho_{\rm TF}=1 . See e.g. [ 5 , 10 ] for Fermi systems in the random phase approximation. Additionally, in the context of large atomic/molecular Coulomb systems, related estimates have been employed with the purpose of proving relevant energy asymptotics. See for instance [ 15 , Theorem 1.6] . Here, however, it is sufficient to analyze the difference between the position densities of a mean-field density matrix (trial state) and the Thomas-Fermi theory; no information about the N -body is deduced. 1.2. Semi-classical commutator estimates and Hartree theory One of the key challenges regarding the problem under consideration is the lack of high regularity of the limit ( 2.4 ), given by a characteristic function. In addition, in order to quantify the convergence as N\rightarrow\infty it becomes crucial to understand the uniform-in- N regularity at the quantum level. In this regard, there is an extremely helpful notion that we may use to understand such properties. Namely, we say (\gamma_{N})_{N\geqslant 1} satisfies the semi-classical commutator estimates if there is C>0 such that for all N\geqslant 1 \|[x,\gamma_{N}]\|_{\textnormal{Tr}}\leqslant CN\hbar\qquad\textnormal{and}\qquad\|[-i\hbar\nabla,\gamma_{N}]\|_{\textnormal{Tr}}\ \leqslant CN\hbar\ , (1.14) where \|A\|_{\textnormal{Tr}}=\textnormal{Tr}|A| is the trace norm for operators on L^{2}(\mathbb{R}^{d}) . The bounds ( 1.14 ) where introduced as key requirements in the initial data by Benedikter, Porta and Schlein [ 2 ] , in the context of the derivation of Hartree-Fock dynamics for large fermionic systems. Physically, these bounds encode the semi-classical structure of the states, meaning they can only vary over macroscopic scales. Mathematically, they contain information about the uniform regularity of the density matrix \gamma_{N} . For instance, for orthogonal projections \gamma_{N}=\gamma_{N}^{2} they imply the optimal fractional regularity of the Wigner functions in the Sobolev scale \sup_{N\geqslant 1}\|f_{\gamma_{N}}\|_{H^{s}(\mathbb{R}^{d}\times\mathbb{R}^{d})}<\infty\ ,\qquad\forall s<\tfrac{1}{2}\ . (1.15) We refer the reader to the article [ 16 ] by Lafleche for more details on the regularity of orthogonal projections verifying the semi-classical commutator estimates. In this article, our strategy is to

Read the complete text