betaFeedback

WorksPhysics and Astronomy

A Finite-Range, Translation-Invariant Counterexample to Weak Gibbsianity

Abstract

We construct a finite-range, translation-invariant counterexample to the Weak Gibbsianity property for quantum spin systems on Z2\mathbb{Z}^2. The proof relies on showing how a local potential can induce macroscopically large deviations from an equilibrium configuration after a perturbation is introduced at the boundary of the lattice.

Notation

We introduce now the basic notation that we will employ in the following.

Notation 0.1 (Lattice structures).

  • We denote by Λ⊆Z2\Lambda\subseteq\mathbb{Z}^{2} finite sublattices of Z2\mathbb{Z}^{2}.

  • We denote by XcX^{c} the complements of XX in Z2\mathbb{Z}^{2}: Xc:=Z2\XX^{c} := \mathbb{Z}^{2}\backslash X.

  • For any finite set X⊆Z2X \subseteq\mathbb{Z}^{2} we denote its cardinality by ∣X∣∈Z+\lvert X\rvert\in\mathbb{Z}_{+}.

Notation 0.2 (Quasi-local algebraic setting). We denote a net of local algebras indexed by finite subsets of Z2\mathbb{Z}^{2} by (AX)X⊆Z2(\mathcal{A}_{X})_{X\subseteq\mathbb{Z}^{2}}. In this work, for every i∈Z2i\in\mathbb{Z}^{2} we have A{i}=M2(C)\mathcal{A}_{\{i\}}=M_{2}(\mathbb{C}) and

AX=⨂i∈XA{i}.\mathcal{A}_{X}=\bigotimes_{i\in X}\mathcal{A}_{\{i\}}.

The quasi local algebra A\mathcal{A} is constructed by standard inductive limit argument from the local algebras AΛ\mathcal{A}_{\Lambda}; see [BR87, Section 2.6] and [6 Section 6.2] for additional details.

Notation 0.3 (Spin matrices and computational basis). We denote by SiαS_{i}^{\alpha}, i∈Z2i\in\mathbb{Z}^{2}, α={x,y,z}\alpha=\{x,y,z\} the spin-12\frac{1}{2} angular momentum operators at site ii, satisfying the SU(2)SU(2) commutation relations

[Siα,Sjβ]=iδij∑γεαβγSiγ.[S_{i}^{\alpha},S_{j}^{\beta}]=\mathrm{i}\delta_{ij}\sum_{\gamma}\varepsilon^{\alpha\beta\gamma}S_{i}^{\gamma}.

We denote by σiα=2Siα\sigma_{i}^{\alpha}=2S_{i}^{\alpha} the corresponding Pauli matrices.

We define the computational basis in a finite region Λ\Lambda as the tensor product over all points of Λ\Lambda of the eigenstates of σiz\sigma_{i}^{z}. These basis elements will be labeled as σΛ\sigma_{\Lambda}, where σΛ∈{+,−}Λ\sigma_{\Lambda}\in\{+,-\}^{\Lambda}. Alternatively, when we need to specify the configuration we will adopt Dirac’s bra-ket notation

σ{i}=∣±⟩ii∈Λ.\sigma_{\{i\}}=\lvert\pm\rangle_{i}\qquad i\in\Lambda.

Introduction

The concept of a Weak Gibbs state has its roots in the thermodynamic formalism of classical statistical mechanics [15]. This property was subsequently formulated explicitly and explored in the theory of dynamical systems [19, 20], as well as in connection with equilibrium measures and large deviations [13, 14].

Even though they were not precisely formalized before recent times, quantum Weak Gibbs state already appeared in the quantum spin system literature [2, 3], and more recently found application in the study of large deviation principles for quantum lattice spin systems [12].

More recently, the study of Weak Gibbs states has received renewed attention in the question of approach to equilibrium for quantum statistical mechanics through the works of Jakšić, Pillet and Tauber [11, 10], who gave the first formalized notion for quantum spin systems, and Jakšić, Pillet and Szczepanek [9, 8, 7]. A standard formulation of Weak Gibbsianity for quantum spin systems is the following; see [11 Definition 2.9].

Definition 1.1 (Weak Gibbs state). Let Φ=(ΦX)X⊆Zd\Phi= (\Phi_X)_{X \subseteq\mathbb{Z}^d} be a self-adjoint interaction for a quantum spin system with quasi-local algebra A\mathcal{A} and local algebras (AX)X⊆Zd(\mathcal{A}_X)_{X \subseteq\mathbb{Z}^d}, so that ΦX=ΦX∗∈AX\Phi_X = \Phi_X^* \in\mathcal{A}_X.

Fix an inverse temperature β∈R\beta\in\mathbb{R}. For any finite region Λ⊂Zd\Lambda\subset\mathbb{Z}^d, let HΛ:=∑X⊆ΛΦXH_\Lambda:= \sum_{X \subseteq\Lambda} \Phi_X and define the finite-volume Gibbs state by

ωΛ(AΛ):=tr⁡[eβHΛAΛ]tr⁡[eβHΛ],AΛ∈AΛ.\omega_\Lambda(A_\Lambda) := \frac{\operatorname{tr}[e^{\beta H_\Lambda} A_\Lambda]}{\operatorname{tr}[e^{\beta H_\Lambda}]}, \qquad A_\Lambda\in\mathcal{A}_\Lambda.

A state ω∈S(A)\omega\in\mathcal{S}(\mathcal{A}) is said to be a Weak Gibbs state with respect to Φ\Phi at inverse temperature β\beta if there exists a family of constants (cΛ)Λ⊆Zd(c_\Lambda)_{\Lambda\subseteq\mathbb{Z}^d}, with cΛ≥0c_\Lambda\ge0, such that

lim⁡Λ↑ZdcΛ∣Λ∣=0,\lim_{\Lambda\uparrow\mathbb{Z}^d} \frac{c_\Lambda}{|\Lambda|} = 0,

where the limit is taken along increasing cubes, and

e−cΛωΛ≤ω∣AΛ≤ecΛωΛ(1)e^{-c_\Lambda}\omega_\Lambda\le\omega|_{\mathcal{A}_\Lambda} \le e^{c_\Lambda}\omega_\Lambda \tag*{(1)}

for every finite region Λ⊂Zd\Lambda\subset\mathbb{Z}^d. The inequalities are understood in the order on positive linear functionals, that is, after evaluation on every positive element of $\mathcal{A}_\Lambda}.

Usually, this definition is given for restricted classes of interactions Φ\Phi, such as those that are translation-invariant and exponentially decaying.

It was widely expected, and explicitly conjectured in [7], that Weak Gibbsianity should hold for thermal equilibrium states of any quantum spin system on the lattice Zd\mathbb{Z}^d, d≥1d \ge1, under suitable regularity assumptions on the interaction, such as finite range or exponential decay. Previous results in this direction date back to Araki [1] for one-dimensional lattice systems with finite-range interactions, and to Araki and Ion [3] for arbitrary d≥1d \ge1 at sufficiently high temperatures. The Weak Gibbsianity property was explicitly established in these settings in [11 Theorem 2.11].

In this paper, we show that the Weak Gibbsianity conjecture is false for d≥2d \ge2, even for finite-range, translation-invariant interactions.

The counterexample is constructed starting from the following nearest-neighbors potential

Φ(X):=J{1i−σiz2σjx+σix1j−σjz2if X={i,j} and ∣i−j∣=1,0otherwise.(2)\Phi(X) := J \begin{cases} \frac{\mathbb{1}_i-\sigma_i^z}{2}\sigma_j^x+\sigma_i^x\frac{\mathbb{1}_j-\sigma_j^z}{2} & \text{if } X=\{i,j\}\text{ and }|i-j|=1,\\ 0 & \text{otherwise}. \end{cases} \tag*{(2)}

for all X⊆Z2X \subseteq\mathbb{Z}^2. The potential in (2) does not belong to the usual class of XYZ interactions. For every pair of neighboring sites (i,j)(i,j), it is given by a coupling J>0J>0 times the sum of two terms: the tensor product of the projection onto the ∣−⟩i|-\rangle_i spin state at site ii,

Piz:=1i−σiz2,P_i^z := \frac{\mathbb{1}_i-\sigma_i^z}{2},

and the Pauli matrix σjx\sigma_j^x at site jj, and the corresponding term with ii and jj interchanged. Therefore, the action of the interaction term PizσjxP_i^z\sigma_j^x on the computational basis is as follows

Pizσjx∣σi,σj⟩=(δσi,−)∣σi,−σj⟩.(3)P_i^z\sigma_j^x|\sigma_i,\sigma_j\rangle=(\delta_{\sigma_i,-})|\sigma_i,-\sigma_j\rangle. \tag*{(3)}

We see that every interaction potential has positive matrix coefficients in the computational basis.

For every finite subset Λ⊂Z2\Lambda\subset\mathbb{Z}^2, we denote the Hamiltonian of the model as

HΛ=∑X⊆ΛΦ(X)=J∑i,j∈Λ∣i−j∣=1PizσjxH_\Lambda=\sum_{X\subseteq\Lambda}\Phi(X)=J\sum_{\substack{i,j\in\Lambda\\ |i-j|=1}}P_i^z\sigma_j^x

By standard results [6 Section 6.2], the net of Gibbs states (ωΛ)Λ⊂Z2(\omega_\Lambda)_{\Lambda\subset\mathbb{Z}^2} for the interaction in (2) admits convergent subsequences to β\beta-KMS states on A\mathcal{A} for the infinite volume dynamics generated by Φ\Phi.

The central result of the paper is formulated in the following theorem.

Theorem 1.2 (Counterexample to the Weak Gibbsianity property). Let ω\omega be any limit point of the net of finite volume Gibbs states (ωΛ)Λ⊂Z2(\omega_\Lambda)_{\Lambda\subset\mathbb{Z}^2} for the interaction potential in (2). Suppose that the inverse temperature β\beta satisfies

Jβ>2112614.J\beta>2^{\frac{11}{2}}6^{\frac{1}{4}}.

Then, there exists a positive constant CJβ>0C_{J\beta}>0 and an increasing sequence of squares (Xk)k∈N(X_k)_{k\in\mathbb{N}} converging to Z2\mathbb{Z}^2 such that

ω(1{12}(mXkz))≥eCJβ∣Xk∣ωXk(1{12}(mXkz)),(4)\omega\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k}^z\right)\right)\geq e^{C_{J\beta}|X_k|}\omega_{X_k}\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k}^z\right)\right), \tag*{(4)}

where

mXkz=1∣Xk∣∑i∈XkSizm_{X_k}^z=\frac{1}{|X_k|}\sum_{i\in X_k}S_i^z

is the mean magnetization along the zz-axis in the region XkX_k.

We see that Equation (4) provides an explicit counterexample to the upper bound in the Weak Gibbsianity condition (1) for any limit point of the net of Gibbs states (ωΛ)Λ⊂Z2(\omega_\Lambda)_{\Lambda\subset\mathbb{Z}^2}.

The proof of the counterexample in Theorem 1.2 relies on techniques similar to those used to establish the blow-up of complex-time evolution in lattice spin systems in dimensions d≥2d\geq2. These techniques were originally developed by Bouch [4] for a specific quantum spin model. Tasaki [18] subsequently provided a clearer exposition of Bouch’s construction and, in a later revision, incorporated an LLM-generated proof extending the combinatorial result from a particular sequence of tree sizes to every positive integer size. It is possible that the “Extraction lemma” [18 Lemma 2.2] could be used to improve our proof, but we did not check this. Shiraishi and Tasaki [16 Appendix A.4, Theorem A.6] also employed Bouch’s construction to extend the complex-time blow-up result to a general class of XY and XYZ models.

In § 3, we prove Theorem 1.2 and discuss some of its immediate consequences.

Proof of the counterexample

In this section, we will present a proof of Theorem 1.2, showing that there exist finite-range, translation-invariant models for which the Weak Gibbsianity property fails on local observables. We will work in a two dimensional lattice Z2\mathbb{Z}^2 by first restricting to a finite subregion of the lattice Λ⊂Z2\Lambda\subset\mathbb{Z}^2. It is fairly easy to extend all the results to d≥2d\geq2 dimensional lattices.

We need a few preliminary observations. The first one concerns the mean magnetization along the third direction, and its relation with the interaction potential defined in Equation (2).

Remark 2.1. We can immediately read from (3) that the “all spin up” state

ΩΛ+:=⨂i∈Λ∣+⟩i\Omega_\Lambda^+:=\bigotimes_{i\in\Lambda}|+\rangle_i

is in the Kernel of HΛH_\Lambda

HΛΩΛ+=0.H_\Lambda\Omega_\Lambda^+ = 0.

The projection onto ΩΛ+\Omega_\Lambda^+ is also the projection corresponding to the value 12\frac{1}{2} of the mean magnetization along the third direction

mΛz=1∣Λ∣∑i∈ΛSiz,∣ΩΛ+⟩⟨ΩΛ+∣=1{12}(mΛz).m_\Lambda^z = \frac{1}{|\Lambda|}\sum_{i\in\Lambda} S_i^z,\qquad|\Omega_\Lambda^+\rangle\langle\Omega_\Lambda^+| = \mathbb{1}_{\{\frac{1}{2}\}}(m_\Lambda^z).

Next, we fix a positive inverse temperature β>0\beta> 0 and we consider the finite volume Gibbs state

ωΛ(AΛ):=tr⁡[eβHΛAΛ]tr⁡[eβHΛ]\omega_\Lambda(A_\Lambda) := \frac{\operatorname{tr}[e^{\beta H_\Lambda}A_\Lambda]}{\operatorname{tr}[e^{\beta H_\Lambda}]}

for any region Λ⊆Z2\Lambda\subseteq\mathbb{Z}^2.

Remark 2.2. For any X⊆Z2X \subseteq\mathbb{Z}^2 we have

ωX(1{12}(mXz))=1tr⁡[eβHX]=ZX−1,\omega_X\left(\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)\right) = \frac{1}{\operatorname{tr}[e^{\beta H_X}]} = Z_X^{-1},

where ZXZ_X is the partition function for the model in (2), localized in X⊆Z2X \subseteq\mathbb{Z}^2.

We consider the boundary interaction for any finite subregion X⊂ΛX \subset\Lambda

WXΛ:=∑Y⊆Λ,Y∩X≠∅Y∩Xc≠∅Φ(Y)=J∑i∈Xc,j∈X∣i−j∣=1σizσjx+σixσjz.W_X^\Lambda:= \sum_{\substack{Y\subseteq\Lambda,Y\cap X\ne\varnothing\\Y\cap X^c\ne\varnothing}}\Phi(Y) = J\sum_{\substack{i\in X^c,j\in X\\|i-j|=1}} \sigma_i^z\sigma_j^x+\sigma_i^x\sigma_j^z.

We can bound the norm of WXΛW_X^\Lambda as

∥WXΛ∥≤2J∣∂X∣,\lVert W_X^\Lambda\rVert\le2J|\partial X|,

where

∂X:={{i,j}⊂Z2∣∣i−j∣=1, i∈X, j∈Xc}\partial X := \{\{i,j\}\subset\mathbb{Z}^2 \mid|i-j|=1,\ i\in X,\ j\in X^c\}

is the boundary of XX.

Remark 2.3. Since the interaction potential Φ\Phi has positive matrix elements with respect to the computational basis, it easily follows that

tr⁡[eβHΛ1{12}(mXz)]≥tr⁡[eβHΛγY]tr⁡[eβHY1{12}(mXz)]\operatorname{tr}[e^{\beta H_\Lambda}\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)] \ge\operatorname{tr}[e^{\beta H_\Lambda} \gamma_Y]\operatorname{tr}[e^{\beta H_Y}\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)]

for any X⊆Y⊆ΛX\subseteq Y\subseteq\Lambda. Moreover, by a standard Golden-Thompson inequality, the partition function satisfies

ZΛ≤e2βJ∣∂Y∣ZYZΛ∖Y.Z_\Lambda\le e^{2\beta J|\partial Y|}Z_Y Z_{\Lambda\setminus Y}.

Hence, we see that

ωΛ(1{12}(mXz))≥e−2Jβ∣∂Y∣ωY(1{12}(mXz)).\omega_\Lambda\left(\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)\right)\ge e^{-2J\beta|\partial Y|}\omega_Y\left(\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)\right).

We can now outline the main steps in the construction of our counterexample. Let us define for any finite set X⊂ΛX\subset\Lambda

X(i):=X∪{i},i∈Λ,X(i):=X\cup\{i\},\qquad i\in\Lambda,

where i∉Xi\notin X and we suppose that there exists a unique j∈Xj\in X such that ∣i−j∣=1|i-j|=1. Combining Observations 2.2 and 2.3 we arrive at the inequality

ωΛ(1{12}(mXz))≥e−2βJ∣∂X(i)∣tr⁡[eβHX(i)1{12}(mXz)]ZX(i)−1≥12e−2βJ∣∂X(i)∣−2Jβtr⁡[eβHX(i)1{12}(mXz)]ωX(1{12}(mXz)).\begin{aligned} \omega_\Lambda\left(\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)\right)\ge e^{-2\beta J|\partial X(i)|}\operatorname{tr}[e^{\beta H_{X(i)}}\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)]Z_{X(i)}^{-1} \\ &\ge\frac{1}{2}e^{-2\beta J|\partial X(i)|-2J\beta}\operatorname{tr}[e^{\beta H_{X(i)}}\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)]\omega_X\left(\mathbb{1}_{\{\frac{1}{2}\}}(m_X^z)\right). \end{aligned}

To further estimate the last line, let us choose an element of the computational basis σX(i)\sigma_{X(i)} in the range of the projector 1{12}(mXz)\mathbb{1}_{\{\frac{1}{2}\}}(m^{z}_{X}), i.e. 1{12}(mXz)σX(i)=σX(i)\mathbb{1}_{\{\frac{1}{2}\}}(m^{z}_{X})\sigma_{X(i)}=\sigma_{X(i)}. This essentially means that σX(i)\sigma_{X(i)} coincide with ΩX+\Omega^{+}_{X} on XX, while it can be chosen arbitrarily on X(i)\XX(i)\backslash X. Then, for any other computational basis element ξX(i)\xi_{X(i)} we have

tr⁡[eβHX(i)1{12}]≥⟨σX(i)∣eβHX(i)σX(i)⟩≥∣⟨σX(i)∣eβ2HX(i)∣ξX(i)⟩∣2.\begin{aligned} \operatorname{tr}\left[e^{\beta H_{X(i)}}\mathbb{1}_{\{\frac{1}{2}\}}\right]\geq\langle\sigma_{X(i)}\mid e^{\beta H_{X(i)}}\sigma_{X(i)}\rangle\\ &\geq\left|\left\langle\sigma_{X(i)}\left|e^{\frac{\beta}{2}H_{X(i)}}\right|\xi_{X(i)}\right\rangle\right|^{2}. \end{aligned}

Using again that the matrix elements of HΛH_{\Lambda} are positive in the computational basis we can write

∣⟨σX(i)∣eβ2HX(i)∣ξX(i)⟩∣=⟨σX(i)∣eβ2HX(i)∣ξX(i)⟩≥1n!⟨σX(i)∣(β2HX(i))n∣ξX(i)⟩,\begin{aligned} \left|\left\langle\sigma_{X(i)}\left|e^{\frac{\beta}{2}H_{X(i)}}\right|\xi_{X(i)}\right\rangle\right| =\left\langle\sigma_{X(i)}\left|e^{\frac{\beta}{2}H_{X(i)}}\right|\xi_{X(i)}\right\rangle\\ &\geq\frac{1}{n!}\left\langle\sigma_{X(i)}\left|\left(\frac{\beta}{2}H_{X(i)}\right)^{n}\right|\xi_{X(i)}\right\rangle, \end{aligned}

where, in the last line, we have estimated the full Taylor series of the exponential using its term of order nn, for arbitrary n≥0n\geq0. The last bracket can be written as a sum of terms of the form

1n!(βJ2)n⟨σX(i)∣Plnzσknx⋯Pl1zσk1x∣ξX(i)⟩,∣lm−km∣=1, m=1,…,n.\frac{1}{n!}\left(\frac{\beta J}{2}\right)^{n}\left\langle\sigma_{X(i)}\left|P^{z}_{l_{n}}\sigma^{x}_{k_{n}}\cdots P^{z}_{l_{1}}\sigma^{x}_{k_{1}}\right|\xi_{X(i)}\right\rangle,\qquad|l_{m}-k_{m}|=1,\ m=1,\ldots,n.

We can interpret the above as a path (k1,l1…kn,ln)(k_{1},l_{1}\ldots k_{n},l_{n}) on the set X(i)X(i), where at every step m=1,…,nm=1,\ldots,n the spin at a site kmk_{m} is reversed if the spin at site lml_{m} is ∣−⟩lm|-\rangle_{l_{m}}. The only non zero contribution will be those arising from paths changing the configuration ξX(i)\xi_{X(i)} into the configuration σX(i)\sigma_{X(i)}. For all such paths, the bracket will be equal to one. Therefore, it remains to estimate the number of these admissible paths. If this number, denoted by N(ξX(i),σX(i),n)N(\xi_{X(i)},\sigma_{X(i)},n), were to satisfy

N(ξX(i),σX(i),∣X∣)≥∣X∣!C∣X∣,N(\xi_{X(i)},\sigma_{X(i)},|X|)\geq\frac{|X|!}{C^{|X|}},

for some constant C>0C>0, then we would obtain the estimate

tr⁡[eβHX(i)1{12}(mXz)]≥(βJ2)2∣X∣(N(ξX(i),σX(i),∣X∣)∣X∣!)2≥(βJ2C)2∣X∣.\operatorname{tr}\left[e^{\beta H_{X(i)}}\mathbb{1}_{\{\frac{1}{2}\}}(m^{z}_{X})\right]\geq\left(\frac{\beta J}{2}\right)^{2|X|}\left(\frac{N(\xi_{X(i)},\sigma_{X(i)},|X|)}{|X|!}\right)^{2}\geq\left(\frac{\beta J}{2C}\right)^{2|X|}.

Putting all together, this would imply the inequality

ωΛ(1{12}(mXz))≥e2∣X∣ln⁡(βJ2C)−2Jβ∣∂X(i)∣−2Jβ−ln⁡(2)ωX(1{12}(mXz)).(5)\omega_{\Lambda}\left(\mathbb{1}_{\{\frac{1}{2}\}}(m^{z}_{X})\right)\geq e^{2|X|\ln\left(\frac{\beta J}{2C}\right)-2J\beta|\partial X(i)|-2J\beta-\ln(2)}\omega_{X}\left(\mathbb{1}_{\{\frac{1}{2}\}}(m^{z}_{X})\right). \tag*{(5)}

We will now describe in detail the estimate for the number N(ξX(i),σX(i),∣X∣)N(\xi_{X(i)},\sigma_{X(i)},|X|). The core idea of our construction is the following. To allow for many different paths connecting the configurations σX(i)\sigma_{X(i)} and ξX(i)\xi_{X(i)}, these vectors must differ at a macroscopic number of points in the lattice. The second requirement is that there should be many distinct ways to change from one configuration to the other by repeatedly applying the potential Plnzσknx⋯Pl1zσk1xP^{z}_{l_{n}}\sigma^{x}_{k_{n}}\cdots P^{z}_{l_{1}}\sigma^{x}_{k_{1}}. As we will see, these conditions are satisfied by our choice of the potential (2). Indeed, the path Plnzσknx⋯Pl1zσk1xξX(i)P^{z}_{l_{n}}\sigma^{x}_{k_{n}}\cdots P^{z}_{l_{1}}\sigma^{x}_{k_{1}}\xi_{X(i)} can have many different branches, giving rise to a fractal tree structure. This procedure follows the idea in Bouch’s analysis of complex time singularity [4], although the details of the computations and the choice of the tree differ from those of Bouch.

Construction of the spanning tree

We consider a sequence of lattice squares XkX_{k} constructed as follows. Fix Lk=2kL_{k}=2^{k} for k∈Nk\in\mathbb{N} and take Xk={0,…,Lk−1}2X_{k}=\{0,\ldots,L_{k}-1\}^{2}. We enlarge it by adding the point (−1,0)(-1,0), so that its enlarged version is given by Xk(−1,0)={0,…,Lk−1}2∪{(−1,0)}X_{k}(-1,0)=\{0,\ldots,L_{k}-1\}^{2}\cup\{(-1,0)\}. The box’s sides length satisfies 2Lk−1=Lk2L_{k-1}=L_{k}. Our choice for the vectors ξXk(−1,0)\xi_{X_{k}(-1,0)} and σXk(−1,0)\sigma_{X_{k}(-1,0)} is the following

σXk(−1,0)=∣−⟩(−1,0)⨂i∈Xk∣+⟩i,\sigma_{X_{k}(-1,0)}=|-\rangle_{(-1,0)}\bigotimes_{i\in X_{k}}|+\rangle_{i},
ξXk(−1,0)=⨂i∈Xk(−1,0)∣−⟩i.\xi_{X_{k}(-1,0)}=\bigotimes_{i\in X_{k}(-1,0)}|-\rangle_{i}.

We see that the two configurations overlap only at the vertex (−1,0)(-1,0). This is the starting point of our construction. Iteratively applying the operators P(−1,0)zσ(0,0)x→P(0,0)zσ(1,0)x→…P^{z}_{(-1,0)}\sigma^{x}_{(0,0)} \to P^{z}_{(0,0)}\sigma^{x}_{(1,0)} \to\ldots we will construct a path of operators covering the whole square XkX_k and flipping the spin at every point by repeated actions

P(−1,0)zσ(0,0)x∣−⟩(−1,0)⨂i∈Xk∣+⟩i=∣−⟩(−1,0)⊗∣−⟩(0,0)⨂i∈Xk∖{(0,0)}∣+⟩iP^{z}_{(-1,0)}\sigma^{x}_{(0,0)}\lvert-\rangle_{(-1,0)} \bigotimes_{i\in X_k}\lvert+\rangle_i = \lvert-\rangle_{(-1,0)}\otimes\lvert-\rangle_{(0,0)} \bigotimes_{i\in X_k\setminus\{(0,0)\}}\lvert+\rangle_i

and so on. This path must contain many branches, otherwise the number of possible combinations that generate the admissible paths would note be sufficient to violate the Weak Gibbsianity property. We will rely on a fractal tree construction, inspired by the results presented in [17].4 The resulting path is constructed by inductively building a spanning tree Tk=(Vk,Ek)T_k=(V_k,E_k). here Vk=XkV_k=X_k are the vertices of the tree, whereas EkE_k are the edges.

Let us describe the first iterative step of the construction. We take T1T_1 to be the spanning tree for X1X_1 with edges

{e(0,0)−(1,0),e(1,0)−(1,1),e(0,1)−(1,1)}=E1.\{e_{(0,0)-(1,0)}, e_{(1,0)-(1,1)}, e_{(0,1)-(1,1)}\}=E_1.

To construct T2T_2, we attach four identical copies of T1T_1 at the four corners of X2X_2, and we connect these by another copy of T1T_1 in the middle square. We repeat this procedure iteratively. To construct the spanning tree TkT_k in XkX_k, we take four translated copies of Tk−1T_{k-1} and we attach these to the four corners of XkX_k. We connect these four trees thorough another copy of T1T_1 in the middle of the square XkX_k. The first tree steps of the construction are illustrated in Figure 1. Note that the resulting graph is always connected and contains all vertices of XkX_k.

Fractal spanning tree construction

Figure 1. Fractal spanning tree construction. This figure is a partial reproduction of [17 Figure 1].

Let us prove a first geometric lemma that will be needed for future estimates

Lemma 2.4. Let TkT_k be the spanning tree defined above. Call DkD_k the diameter of TkT_k defined as the longest non-intersecting path along the edges of TkT_k. Then,

Dk≤3(2k−1).D_k \le3(2^k-1).

Proof. It is clear that D1=3D_1=3. For any two points x,y∈Xkx,y\in X_k, these satisfy one of the following:

  1. Both xx and yy belong to the central square T1T_1;

  2. One of the two points belongs to the central square say x∈T1x\in T_1, while the other belongs to one of the four copies of Tk−1T_{k-1};

  3. They both belong to the same copy of Tk−1T_{k-1};

  4. They belong to different copies of Tk−1T_{k-1}.

In the first case, their distance along the edges of TkT_k is at most 3. As for the second case, their distance is less than the number of edges that one of the two points has to travel to reach the central square, plus the diameter of the square. This is less than Dk−1+3D_{k-1}+3. In the third case, their distance is always less than Dk−1D_{k-1}. In the fourth case, their distance is always less than the number of edges it takes for both of them to travel to the cross, plus the length of the cross. This is less than 2Dk−1+32D_{k-1}+3. Therefore, we see that

Dk≤2Dk−1+3.D_k \le2D_{k-1}+3.

Iterating this inequality k−1k-1 times we get

Dk≤3∑n=0k−12n=3(2k−1).D_k \le3\sum_{n=0}^{k-1}2^n=3(2^k-1).

This concludes the proof of the Lemma. □\square

Next, we count the number of ways to construct the tree TkT_k as a nested tree, starting from a given root r∈Xkr\in X_k. In [4 Lemma 6.2], Bouch proved that the number of different constructions leading to the same rooted tree with root rr and edges e1r,…,enre_1^r,\ldots,e_n^r is given by

n!∏i=1nw(eir),\frac{n!}{\prod_{i=1}^{n}w(e_i^r)},

where w(eir)w(e_i^r) is the weight of the edge eire_i^r, defined as one plus the number of its descendants. These are the edges that follow eire_i^r as we move away from the root rr. Therefore, after relabeling the edges of TkT_k as e1r,…,e∣Xk∣−1re_1^r,\ldots,e_{|X_k|-1}^r, we obtain a lower bound on the number of possible constructions by finding an upper bound on

∏i=1∣Xk∣−1w(eir).\prod_{i=1}^{|X_k|-1}w(e_i^r).

Let us define

Wk:=sup⁡r∈Tk∑i=1∣Xk∣−1ln⁡(w(eir)).W_k:=\sup_{r\in T_k}\sum_{i=1}^{|X_k|-1}\ln(w(e_i^r)).

We will estimate WkW_k iteratively. In particular, we shall prove

Lemma 2.5. Take WkW_k as above. Then,

Wk≤(92ln⁡(2)+14ln⁡(6))∣Xk∣W_k\le\left(\frac{9}{2}\ln(2)+\frac{1}{4}\ln(6)\right)|X_k|

Proof. We distinguish two cases. i)i) We place the root of TkT_k on the central square T1T_1. Then, every edges on the central tree will have necessarily less than ∣Xk∣|X_k| descendant (this is the trivial bound). The product of the weights over all the edges of the four copies Tk−1T_{k-1} is bounded by exp⁡Wk−1\exp W_{k-1} by definition. Therefore, summing all the contributions we have

∑i=1∣Xk∣−1ln⁡(w(eir))≤4Wk−1+3ln⁡(∣Xk∣)\sum_{i=1}^{|X_k|-1}\ln(w(e_i^r))\le4W_{k-1}+3\ln(|X_k|)

ii)ii) Suppose we put the root on one of the four squares Tk−1T_{k-1}. The unique path connecting the root with the central copy of T1T_1 contains less than Dk−1D_{k-1} edges. Each of these will have less than ∣Xk∣|X_k| descendants. All the other branches starting from this path give a total contribution less than exp⁡Wk−1\exp W_{k-1}. We can repeat the same argument as in i)i) on the edges of the central square T1T_1 and of the other three copies of Tk−1T_{k-1}. Therefore,

∑i=1∣Xk∣−1ln⁡(w(eir))≤4Wk−1+(3+Dk−1)ln⁡(∣Xk∣)\sum_{i=1}^{|X_k|-1}\ln(w(e_i^r))\le4W_{k-1}+(3+D_{k-1})\ln(|X_k|)
≤4Wk−1+3ln⁡(2)2kk\le4W_{k-1}+3\ln(2)2^k k

where we have used Lemma 2.4 to estimate Dk−1D_{k-1}. Since W1=ln⁡(6)W_1=\ln(6), iterating the above inequality we get

Wk≤3ln⁡(2)22k∑n=2kn2n+22k−2ln⁡(6)W_k \le3\ln(2)2^{2k}\sum_{n=2}^{k}\frac{n}{2^n}+2^{2k-2}\ln(6)
≤3ln⁡(2)22k∑n=2+∞n2n+22k−2ln⁡(6)=(92ln⁡(2)+14ln⁡(6))∣Xk∣.\le3\ln(2)2^{2k}\sum_{n=2}^{+\infty}\frac{n}{2^n}+2^{2k-2}\ln(6)=\left(\frac{9}{2}\ln(2)+\frac{1}{4}\ln(6)\right)|X_k|.

This last estimate concludes the proof.

Upper bound on the Gibbs state

As we already argued in the previous subsection, if we start with a configuration σXk(−1,0)\sigma_{X_k(-1,0)} with all spin up, except for the spin at (−1,0)(-1,0), the contribution to

⟨ξXk(−1,0)|HXk(−1,0)∣Xk∣|σXk(−1,0)⟩\left\langle\xi_{X_k(-1,0)}\middle|H_{X_k(-1,0)}^{|X_k|}\middle|\sigma_{X_k(-1,0)}\right\rangle

consists in all the possible paths of operators PizσjxP_i^{z} \sigma_j^{x}, ∣i−j∣=1|i-j|=1 flipping all the spin in XkX_k. Some of the admissible choices are given by the all the possible paths generating the rooted tree TkT_k with root in (−1,0)(-1,0); see Figure 2.

Path construction

Figure 2. Path construction.

Combining Lemma 2.5 with Bouch’s argument, we see that the number of way to construct TkT_k satisfies

N(ξXk(−1,0),σXk(−1,0),∣Xk∣)≥∣Xk∣!∣Xk∣−1e−92ln⁡(2)∣Xk∣−14ln⁡(6)∣Xk∣.N(\xi_{X_k(-1,0)},\sigma_{X_k(-1,0)},|X_k|)\ge|X_k|!|X_k|^{-1}e^{-\frac{9}{2}\ln(2)|X_k|-\frac{1}{4}\ln(6)|X_k|}.

Finally, inserting the latter estimate in (5) we conclude

ωΛ(1{12}(mXkz))≥e2∣Xk∣ln⁡(βJ2264)−2Jβ∣∂Xk(−1,0)∣−2ln⁡(∣Xk∣)−ln⁡(2)−2JβωXk(1{12}(mXkz)).\omega_{\Lambda}\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k}^{z}\right)\right)\ge e^{2|X_k|\ln\left(\frac{\beta J}{2^{2}6^{4}}\right)-2J\beta|\partial X_k(-1,0)|-2\ln(|X_k|)-\ln(2)-2J\beta}\omega_{X_k}\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k}^{z}\right)\right).

Since this inequality holds for any Λ⊇Xk(−1,0)\Lambda\supseteq X_k(-1,0), we can take a converging subsequence for ωΛ\omega_{\Lambda} as Λ↑Z2\Lambda\uparrow\mathbb{Z}^2. To recover the exact statement of Theorem 1.2 it sufficient to repeat the construction using translated boxes XkX_k that eventually cover the whole lattice Z2\mathbb{Z}^2. Thus, we have proved Theorem 1.2.

Remark 2.6 (Invariance under rotations). Since the magnetization commutes with any rotation around the zz-axis, the same result holds for all the rotated Hamiltonian

∏i∈ΛUθ,iHΛ∏i∈ΛUθ,i∗,Uθ,i=e−iθiSiz.\prod_{i\in\Lambda} U_{\theta,i}H_\Lambda\prod_{i\in\Lambda} U_{\theta,i}^*, \qquad U_{\theta,i}=e^{-i\theta_i S_i^z}.

In particular, by choosing θi=π\theta_i=\pi for every i∈Λi\in\Lambda we can change the sign of the coupling JJ (or equivalently, of the inverse temperature β\beta). In a similar fashion, the same result holds for the model where we substitute all the projections PizP_i^z onto ∣−⟩i\lvert-\rangle_i with the projections Qiz=1i+σiz2Q_i^z=\frac{1_i+\sigma_i^z}{2} onto the spin-up configurations ∣+⟩i\lvert+\rangle_i.

Remark 2.7 (Translation invariance). The result of Theorem 1.2 readily generalizes to non-translation invariant models with potentials of the form

Φ({i,j})=Ji,jPizσjx+Ki,jσixPjz,∣i−j∣=1, i,j∈Z2,\Phi(\{i,j\})=J_{i,j}P_i^z\sigma_j^x+K_{i,j}\sigma_i^xP_j^z,\qquad|i-j|=1,\ i,j\in\mathbb{Z}^2,

under suitable hypothesis on the couplings, e.g. Ji,j,Ki,j≥J>0J_{i,j},K_{i,j}\ge J>0 and

sup⁡i,j∈Z2{Ji,j,Ki,j}<+∞\sup_{i,j\in\mathbb{Z}^2}\{J_{i,j},K_{i,j}\}<+\infty

would be sufficient.

Remark 2.8 (Magnetization bounds). The bound presented in Theorem 1.2 can be equivalently written as an inequality between different macroscopic sizes for the magnetization(2) observable. Indeed, we can write

ωXk(−1,0)(1{12}(mXk(−1,0)z))=ZXk(−1,0)−1≤ZXk−1e2Jβ\omega_{X_k(-1,0)}\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k(-1,0)}^z\right)\right)=Z_{X_k(-1,0)}^{-1}\le Z_{X_k}^{-1}e^{2J\beta}
=ωXk(1{12}(mXkz))e2Jβ=\omega_{X_k}\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k}^z\right)\right)e^{2J\beta}
≤ωXk(−1,0)(1{12}(mXkz))e2Jβ−CJβ∣Xk∣.\le\omega_{X_k(-1,0)}\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k}^z\right)\right)e^{2J\beta-C_J\beta|X_k|}.

Therefore, we see that changing the magnetization observable only at the boundary produces a macroscopically relevant factor.

ωXk(−1,0)(1{12}(mXkz))≥e−2Jβ+CJβ∣Xk∣ωXk(−1,0)(1{12}(mXk(−1,0)z)).\omega_{X_k(-1,0)}\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k}^z\right)\right)\ge e^{-2J\beta+C_J\beta|X_k|}\omega_{X_k(-1,0)}\left(\mathbf{1}_{\{\frac{1}{2}\}}\left(m_{X_k(-1,0)}^z\right)\right).

Remark 2.9 (Limits of the current approach). It is not clear at present to what extent one can generalize the above argument to other projections of the magnetization. If we were to consider the projection onto states with one or more spins down, it would be impossible to reproduce the argument used in Theorem 1.2. Indeed, we would have one “activation site” inside the region XX. This alone would produce a macroscopic contribution to the expectation value in the Gibbs state.

AI disclosure. Following the release of the first version of this manuscript, Hal Tasaki drew my attention to technical errors in the proofs that had been identified using large language models (LLMs). I subsequently corrected these errors. LLMs were not used in any other part of the manuscript.

Acknowledgements. I am grateful to Yoshiko Ogata for many illuminating and stimulating discussions on this topic. Without her support, I could not have developed the arguments presented in this manuscript. I also thank Vojkan Jakšić and Anna Szczepanek for drawing my attention to the relevance of Weak Gibbsianity and for their many useful comments. I thank Hal Tasaki for pointing out errors in the first version of this manuscript.

I acknowledge the support of the National Group of Mathematical Physics (GNFM–INdAM).

Data availability statement. Data sharing is not applicable to this article as no new data were created or analysed in this study.

Conflict of interest statement. The authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter discussed in this manuscript.

References

  1. [1]H. Araki. “Gibbs states of a one dimensional quantum lattice”. In: Communications in Mathematical Physics 14.2 (1969), pp. 120–157. DOI: 10.1007/BF01645134. URL: https://doi.org/10.1007/BF01645134.
  2. [2]H. Araki. “Positive cone, Radon–Nikodým theorems, relative Hamiltonian and the Gibbs condition in statistical mechanics. An application of the Tomita–Takesaki theory”. In: C*-Algebras and Their Applications to Statistical Mechanics and Quantum Field Theory. Ed. by D. Kastler. Vol. 60. Proceedings of the International School of Physics “Enrico Fermi”. Course LX, Varenna, 1973. Amsterdam: North-Holland, 1976, pp. 64–100.
  3. [3]H. Araki and P. D. F. Ion. “On the equivalence of KMS and Gibbs conditions for states of quantum lattice systems”. In: Communications in Mathematical Physics 35.1 (1974), pp. 1–12. DOI: 10.1007/BF01646450. URL: https://doi.org/10.1007/BF01646450.
  4. [4]G. Bouch. “Complex-time singularity and locality estimates for quantum lattice systems”. In: Journal of Mathematical Physics 56.12 (2015), p. 123303. DOI: 10.1063/1.4936209.DOI
  5. [6]O. Bratteli and D. W. Robinson. Operator Algebras and Quantum Statistical Mechanics 2: Equilibrium States. Models in Quantum Statistical Mechanics. 2nd ed. Berlin, Heidelberg: Springer, 1997. ISBN: 978-3-540-61443-2. DOI: 10.1007/978-3-662-03444-6. URL: https://doi.org/10.1007/978-3-662-03444-6.
  6. [7]V. Jaks̆ić, C.-A. Pillet, and A. Szczepanek. Approach to equilibrium in translation-invariant quantum systems: Structural results II. arXiv:2610.21475. 2026. arXiv: 2610.21475 [math-ph]. URL: https://arxiv.org/abs/2610.21475.
  7. [8]V. Jaks̆ić, C.-A. Pillet, and A. Szczepanek. Dynamical conservation laws in quantum spin systems. arXiv:2610.21489. 2026. arXiv: 2610. 21489 [math-ph]. URL: https://arxiv.org/abs/2610.21489.
  8. [9]V. Jaks̆ić, C.-A. Pillet, and A. Szczepanek. “Entropy maximization and weak Gibbsianity of quasi-free fermionic states”. In: Letters in Mathematical Physics 116.4 (2026), p. 103. DOI: 10.1007/s11005-026-02100-9. arXiv: 2603.14020. URL: https://doi.org/10.1007/s11005-026-02100-9.
  9. [10]V. Jaks̆ić, C.-A. Pillet, and C. Tauber. “A note on adiabatic time evolution and quasi-static processes in translation-invariant quantum systems”. In: Annales Henri Poincaré 25.1 (2024), pp. 751–771. DOI: 10.1007/s00023-023-01282-5. arXiv: 2204.02177. URL: https://doi.org/10.1007/s00023-023-01282-5.
  10. [11]V. Jaks̆ić, C.-A. Pillet, and C. Tauber. “Approach to equilibrium in translation-invariant quantum systems: Some structural results”. In: Annales Henri Poincaré 25.1 (2024), pp. 715–749. DOI: 10.1007/s00023-023-01281-6. arXiv: 2204.00440. URL: https://doi.org/10.1007/s00023-023-01281-6.
  11. [12]Y. Ogata and L. Rey-Bellet. “Ruelle–Lanford functions and large deviations for asymptotically decoupled quantum systems”. In: Reviews in Mathematical Physics 23.2 (2011), pp. 211–232. DOI: 10.1142/S0129055X11004291.DOI
  12. [13]C.-E. Pfister and W. G. Sullivan. “Weak Gibbs measures and large deviations”. In: Nonlinearity 31.1 (2018), pp. 49–53. DOI: 10.1088/1361-6544/aa99a3. arXiv: 1705.06939. URL: https://doi.org/10.1088/1361-6544/aa99a3.
  13. [14]C.-E. Pfister and W. G. Sullivan. “Asymptotic decoupling and weak Gibbs measures for finite alpha-bet shift spaces”. In: Nonlinearity 33.9 (2020), pp. 4799–4817. DOI: 10.1088/1361-6544/ab8fb7. URL: https://doi.org/10.1088/1361-6544/ab8fb7.
  14. [15]D. Ruelle. Statistical Mechanics: Rigorous Results. Mathematical Physics Monograph Series. New York: W. A. Benjamin, 1969. ISBN: 0805383603.
  15. [16]N. Shiraishi and H. Tasaki. “The S = ½ XY and XYZ Models on the Two- or Higher-Dimensional Hypercubic Lattice Do Not Possess Nontrivial Local Conserved Quantities”. In: Annales Henri Poincaré (2026). DOI: 10.1007/s00023-026-01699-8. arXiv: 2412.18504 [cond-mat.stat-mech].
  16. [17]K. Tapp. “On the minimum spanning tree distribution in grids”. In: European Journal of Combinatorics 133 (2026), p. 104325. ISSN: 0195-6698. DOI: https://doi.org/10.1016/j.ejc.2025.104325. URL: https://www.sciencedirect.com/science/article/pii/S0195669825002185.
  17. [18]H. Tasaki. Trees that can be grown in “too many” ways: Bouch’s construction and a bound for every size. Version 4. Originally posted in December 2024. 2026. DOI: 10.48550/arXiv.2412.16912. arXiv: 2412.16912 [math-ph]. URL: https://arxiv.org/abs/2412.16912v4.
  18. [19]M. Yuri. “Weak Gibbs measures for certain non-hyperbolic systems”. In: Ergodic Theory and Dynamical Systems 20.5 (2000), pp. 1495–1518. DOI: 10.1017/S014338570000081X. URL: https://doi.org/10.1017/S014338570000081X.
  19. [20]M. Yuri. “Weak Gibbs measures and the local product structure”. In: Ergodic Theory and Dynamical Systems 22.6 (2002), pp. 1933–1955. DOI: 10.1017/S0143385702000780. URL: https://doi.org/10.1017/S0143385702000780.

Paper details

vlnc:W_wx2dnuen
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 6

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents