supplemental

Lieb-Schultz-Mattis theorems and generalizations in long-range interacting systems

Abstract

In a unified fashion, we establish Lieb-Schultz-Mattis (LSM) theorems and their generalizations in systems with long-range interactions. We show that, for a quantum spin chain, if the interactions decay fast enough as their ranges increase and the Hamiltonian has an anomalous symmetry, the Hamiltonian cannot have a unique gapped symmetric ground state. If the Hamiltonian contains only 2-spin interactions, these theorems hold when the interactions decay faster than $1/r^{2}$, with $r$ the distance between the two interacting spins. Moreover, any pure state with an anomalous symmetry, which may not be a ground state of any natural Hamiltonian, must be long-range entangled. The symmetries we consider include on-site internal symmetries combined with lattice translation symmetries, and they can also extend to purely internal but non-on-site symmetries. Moreover, these internal symmetries can be discrete or continuous. We explore the applications of the theorems through various examples.

Introduction – Understanding and realizing interesting quantum phases of matter is a central goal of condensed matter physics. In this regard, Lieb-Schultz-Mattis-type (LSM) constraints are extremely powerful, which, as initially stated, rule out a unique gapped symmetric ground state based on some basic symmetry-related properties of the system’s Hamiltonian, without referring to any other detail of the Hamiltonian Lieb et al. (1961); Oshikawa (2000); Hastings (2004). Recently, LSM constraints have been interpreted from various perspectives and generalized to different contexts Cheng et al. (2016); Po et al. (2017); Jian et al. (2018); Cho et al. (2017); Metlitski and Thorngren (2018); Cheng (2019); Kobayashi et al. (2019); Else and Thorngren (2020); Jiang et al. (2021); Ogata et al. (2021); Ye et al. (2022); Ma and Wang (2023); Cheng and Seiberg (2023); Kawabata et al. (2024); Seifnashri (2024); Zhou et al. (2023); Kapustin and Sopenko (2024); Garre Rubio et al. (2024). Furthermore, these constraints are identified as a key ingredient to study the classification of quantum phases of matter in a lattice system Zou et al. (2021); Ye et al. (2022); Ye and Zou (2023).

Previous studies of LSM constraints often focus on systems with local interactions. However, many systems feature long-range interactions, which usually take the form of a 2-body interaction that decays as $1/r^{\mathfrak{a}}$, with $r$ the distance between the two interacting objects and $\mathfrak{a}$ an exponent. As examples, electronic systems have Coulomb interaction with $\mathfrak{a}=1$, Rydberg atoms have dipolar or van der Waals interactions with $\mathfrak{a}=3$ or $\mathfrak{a}=6$, and for trapped ions $\mathfrak{a}$ can be tuned between 0 and 3 Defenu et al. (2023). So an important question is: Are LSM constraints applicable to long-range interacting systems?

In this paper, we prove generalized LSM theorems in quantum spin chains with long-range interactions, detailed in theorems 1 and 3 below. In essence, we show that if 1) the long-range interactions decay fast enough as their ranges increase and 2) the system has an anomalous symmetry Kapustin and Sopenko (2024), then the system cannot have a unique gapped symmetric ground state, and all symmetric pure states must be long-range entangled, regardless whether the states are ground states or not. We remark that the Hamiltonians we consider can contain generic $k$-body interactions with $k>2$. For 2-body interactions decaying as $1/r^{\mathfrak{a}}$, our theorems hold when $\mathfrak{a}>2$ (for $k$-body interactions with $k>2$, the condition under which our theorems hold is stated in Eq. (4)). The type of symmetries under consideration is also very broad, including an on-site symmetry combined with the lattice translation symmetry, as featured in the original LSM theorems. Additionally, the symmetry can be purely internal but non-on-site. Furthermore, the internal symmetries can be either discrete or continuous. Besides incorporating long-range interactions, our theorems generalize the original LSM theorems in two ways. First, our theorems apply to a more general class of symmetries. Second, the original LSM theorems often concern about the Hamiltonians’ spectra, but our results also govern the entanglement properties of general states. Our results have wide applicability, and we will discuss some examples below.

Operator algebra formalism – To have a clean notion of locality, we wish to work with systems of infinite size. The operator algebra formalism deals with both finite and infinite systems conveniently. Below we first apply this formalism to infinite systems, which can be viewed as the thermodynamic limits where a sequence of finite systems converge to. From these results, we will extract important implications on finite systems.

For finite systems, the operator algebra formalism is just the usual quantum mechanics in the Heisenberg picture. Here we briefly review this formalism in the context of infinite systems before applying it. We start with the notions of operator algebras and states of infinite size. Then we discuss the symmetry actions and the associated anomaly index developed in Ref. Kapustin and Sopenko (2024), which characterizes the interplay between locality and symmetry.

Given an infinite lattice $\Lambda$ and its finite subset $\Gamma\subset\Lambda$, operators acting trivially outside $\Gamma$, including $c$-numbers, form a local operator algebra, denoted by $\mathcal{A}_{\Gamma}^{\ell}$. The algebra of all local operators is defined as $\mathcal{A}^{\ell}:=\bigcup_{\Gamma\subset\Lambda,~{}|\Gamma|<\infty}\mathcal{A}^{\ell}_{\Gamma}$, with $|\Gamma|$ the cardinality of $\Gamma$.

The Hilbert space $\mathcal{H}_{\Gamma}$ associated with the finite subset $\Gamma$ is the tensor product of the finite-dimensional on-site Hilbert space $\mathcal{H}_{k}$ for each site $k$ in $\Gamma$, i.e., $\mathcal{H}_{\Gamma}=\bigotimes_{k\in\Gamma}\mathcal{H}_{k}$. However, in contrast to finite systems, the total Hilbert space for infinite systems is not well-defined. So how should we represent a quantum state? Recall that a quantum state in finite systems can be specified by the expectation values of all operators with respect to it, so we can define states in infinite systems analogously. Concretely, a state $\psi$ is a linear functional $\psi:\mathcal{A}^{\ell}\to\mathbb{C}$ that satisfies positivity (i.e., $\psi(A^{\dagger}A)\geqslant 0$ for any local operator $A$) and normalization (i.e., $\psi(I)=1$ with $I$ the identity operator).

We remark that, under this definition, states in different superselection sectors of an infinite system (i.e., states that cannot be related by local operators) can only form classical mixtures, but not quantum superpositions. For example, although the $N$-qubit GHZ state $\frac{1}{\sqrt{2}}\left(|0\cdots 0\rangle+|1\cdots 1\rangle\right)$ is pure, its infinite-system version is a mixed state, because no local operator can couple the states $|0\cdots 0\rangle$ and $|1\cdots 1\rangle$ (see supplemental material (SM) for more discussion sup ).

To discuss the entanglement structure of quantum states in infinite systems, it is useful to introduce the split property Matsui (2013); Ogata (2019); Naaijkens and Ogata (2022). When we cut the chain at any point, say, the origin, a pure state $\psi$ of the whole chain may not be factorized as $\psi\simeq\psi_{<0}\otimes\psi_{\geqslant 0}$ for some pure states $\psi_{<0}$ on the left-half chain and $\psi_{\geqslant 0}$ on the right-half chain. If $\psi$ factorizes in this way indeed, then we say $\psi$ splits or it satisfies the split property.

The above discussion on operators and states pertains to general infinite chains. However, we are specifically interested in quantum spin chains with symmetries. Symmetries in this formalism are described by automorphisms associated with the algebra $\mathcal{A}^{\ell}$. An automorphism is an invertible linear map $\varphi:\mathcal{A}^{\ell}\to\mathcal{A}^{\ell}$ satisfying $\varphi(AB)=\varphi(A)\varphi(B)$ and $\varphi(A^{\dagger})=\varphi(A)^{\dagger}$ for any $A,B\in\mathcal{A}^{\ell}$. Automorphisms of $\mathcal{A}^{\ell}$ form a group under finite compositions, denoted by $\mathrm{Aut}(\mathcal{A}^{\ell})$, and symmetry operations can be modelled by automorphisms. There is a special subgroup of $\mathrm{Aut}(\mathcal{A}^{\ell})$ called quantum cellular automata (QCA), denoted by $\mathcal{G}^{\mathrm{QCA}}$, which preserves the locality of operators. More precisely, an automorphism $\varphi$ is a QCA if $\varphi(A)\in\mathcal{A}^{\ell}_{B(\Gamma,r_{\varphi})}$ for each $A\in\mathcal{A}^{\ell}_{\Gamma}$, where $B(\Gamma,r_{\varphi}):=\{x\in\Lambda\,|\,d(x,\Gamma)\leqslant r_{\varphi}\}$, with $r_{\varphi}>0$ depending only on $\varphi$ and $d(x,\Gamma)$ the distance between $x$ and $\Gamma$. The structure of $\mathcal{G}^{\mathrm{QCA}}$ is well-understood in 1D Gross et al. (2012). In essence, 1D QCA are combinations of finite-depth quantum circuits and translations (see Refs. Arrighi (2019); Farrelly (2020) for review).

In the main text, we will focus on unitary symmetries implemented by QCA, as they preserve locality in the most strict sense (in SM, our considerations are extended to a more general class of symmetry actions, i.e., locality preserving automorphisms, and our main theorems still hold). Concretely, given a symmetry group $G$, the symmetry action can be represented by a group homomorphism $\alpha:G\to\mathcal{G}^{\mathrm{QCA}}$. This symmetry may contain internal and/or translation symmetry, and the internal symmetry may be discrete or continuous, on-site or non-on-site.

Figure 1: An illustration on how to obtain anomaly index $\omega\in\mathrm{H}^{3}(G;\mathrm{U}(1))$ from the symmetry action $\alpha$.

Given such a symmetry action $\alpha:G\to\mathcal{G}^{\mathrm{QCA}}$, an important concept is the anomaly index, which takes values in $\mathrm{H}^{3}(G,\mathrm{U}(1))$ Kapustin and Sopenko (2024). The construction of this anomaly index is similar to the previous work Else and Nayak (2014), and the innovation of this new anomaly index is that it applies to translation symmetries and continuous internal symmetries. Below we sketch the definition of the anomaly index, and more details can be found in Ref. Kapustin and Sopenko (2024) and SM sup .

First, suppose $\alpha$ is an internal symmetry action (i.e., it contains no translation) and choose an arbitrary site, say, the origin, then it can be shown that $\alpha$ can be decomposed as

$$\alpha=\alpha_{<0}\,\alpha_{0}\,\alpha_{\geqslant 0}\;,$$ (1)

where $\alpha_{\geqslant 0}$ (resp. $\alpha_{<0}$) is an automorphism of $\mathcal{A}^{\ell}_{\geqslant 0}$ ($\mathcal{A}^{\ell}_{<0}$), and $\alpha_{0}$ is a local unitary (see Fig. 1). Although $\alpha$ is a group homomorphism, in general $\alpha_{\geqslant 0}$ is not. In fact, for any $g,h\in G$,

$$\alpha_{\geqslant 0}(g)\,\alpha_{\geqslant 0}(h)=\mathrm{Ad}_{V(g,h)}\,\alpha_{\geqslant 0}(gh)\;,$$ (2)

where $V:G\times G\to\mathcal{U}^{\ell}$ with $\mathcal{U}^{\ell}$ the group of local unitaries is not necessarily a homomorphism, and $\mathrm{Ad}_{V}(A):=VAV^{\dagger}$ for any $A\in\mathcal{A}^{\ell}$. The associativity of $\alpha_{\geqslant 0}$, i.e., $\left(\alpha_{\geqslant 0}(g)\,\alpha_{\geqslant 0}(h)\right)\,\alpha_{\geqslant 0}(k)=\alpha_{\geqslant 0}(g)\,\left(\alpha_{\geqslant 0}(h)\,\alpha_{\geqslant 0}(k)\right)$, puts further constraints on $V$: $\mathrm{Ad}_{\omega(g,h,k)}=1$, where

$$\begin{split}&\omega(g,h,k)\\ &=V(g,h)V(gh,k)V(g,hk)^{-1}(\alpha_{\geqslant 0}(g)(V(h,k)))^{-1}\;.\end{split}$$ (3)

This means the above $\omega$ is actually a phase since it commutes with all local operators. It can be checked that $\omega$ satisfies the 3-cocycle condition, and multiplying $V(g,h)$ by a phase $\rho(g,h)\in\mathrm{U}(1)$ shifts $\omega$ by a 3-coboundary. Therefore, $\omega$ specifies an element in $\mathrm{H}^{3}(G,\mathrm{U}(1))$, and this element is defined as the anomaly index associated with the symmetry action $\alpha$.

If $\alpha$ contains translation, one can stack the system with another copy on which the translation acts oppositely. The symmetry action on this composite system (denoted by $\alpha_{\otimes}$) contains no translation, and the anomaly index of $\alpha$ is defined to be the index of $\alpha_{\otimes}$.

In SM sup , we prove that this anomaly index is independent of the choice of the site to decompose $\alpha$ in Eq. (1), which was not explicitly proved in Refs. Else and Nayak (2014); Kapustin and Sopenko (2024).

With the above definition of anomaly index, we say that the $G$-symmetry is anomalous if $\omega\not=1\in\mathrm{H}^{3}(G;\mathrm{U}(1))$. Otherwise, we say it is anomaly-free or non-anomalous.

To connect the above discussion with the more familiar notions, let us discuss an example. Consider a quantum spin chain with a symmetry $G=\mathbb{Z}\times G_{\text{int}}$, where $\mathbb{Z}$ represents translation and $G_{\text{int}}$ is an internal symmetry (taken as either a discrete group or a finite dimensional Lie group). Then $\mathrm{H}^{3}(G,\mathrm{U}(1))\simeq\mathrm{H}^{2}(G_{\text{int}},\mathrm{U}(1))\oplus\mathrm{H}^{3}(G_{\text{int}},\mathrm{U}(1))$ Cheng et al. (2016). The part $\mathrm{H}^{2}(G_{\text{int}},\mathrm{U}(1))$ means if the degrees of freedom in a unit cell form a projective representation under $G_{\text{int}}$, which is precisely the condition of the original LSM theorems, the $G$-symmetry is anomalous. The part $\mathrm{H}^{3}(G_{\text{int}},\mathrm{U}(1))$ means that even for a purely internal symmetry $G_{\text{int}}$, the $G$-symmetry can be anomalous if its anomaly index corresponds to a nontrivial element in $\mathrm{H}^{3}(G_{\text{int}},\mathrm{U}(1))$. We will present an example of such internal symmetries below.

LSM theorems and generalizations in long-range interacting systems – Now we proceed to our main theorems, which accommodate long-range and many-body interactions. Consider a 1D Hamiltonian with at most $k$-body interactions, $H=\sum_{|Z|\leqslant k}h_{Z}$, that satisfies Kuwahara and Saito (2020)

$$\begin{split}\max_{i\in\mathbb{Z}}\left(\sum_{Z:Z\owns i,~{}\mathrm{diam}(Z)=d}||h_{Z}||\right)&=O(r^{-\mathfrak{a}}),~{}\mathfrak{a}>2\;,\\ \text{and}~{}\max_{i\in\mathbb{Z}}\|h_{i}\|&\leqslant B\;,\end{split}$$ (4)

where $\mathrm{diam}(Z)=\sup_{x,y\in Z}|x-y|$, and $h_{i}$ is the on-site potential at site $i$. If $H$ satisfies Eq. (4), it is deemed as admissible. Specifically, if the Hamiltonian includes at most 2-body long-range interactions, Eq. (4) indicates that the interactions decay faster than $r^{-2}$, with $r$ the distance between the two interacting spins. Eq. (4) ensures that for any disjoint intervals $X,Y$ (separated by $d$) as in Fig. 2, their interaction $V_{I_{1},I_{2}}=\sum_{Z:Z\cap I_{1},Z\cap I_{2}\not=\emptyset}h_{Z}$ goes to 0 as $d\to\infty$.

Figure 2: The interaction of two disjoint intervals $X,Y$ seprated by distance $r$. The range of interaction is denoted by $Z$ with $\mathrm{diam}(Z)=d$.

To derive our theorem 1, we present several lemmas about the properties of the ground states of admissible Hamiltonians.

Lemma 1

A gapped ground state of an admissible Hamiltonian in 1D must split.

This lemma is deduced by combining Sec. II of Ref. Kuwahara and Saito (2020) and theorem 1.5 of Ref. Matsui (2013).

Lemma 2 (Theorem I.5 in SM sup )

A locally-unique gapped ground state of an admissible Hamiltonian $H$ is pure.

Lemma 3 (Remark 4.1 of Ref. Kapustin and Sopenko (2024))

Given a symmetry action $\alpha:G\to\mathcal{G}^{\mathrm{QCA}}$ on a quantum spin chain, if there exists a pure state $\psi$ which splits and is $G$-symmetric, then the associated anomaly index $\omega=1$.

Our first main theorem can be obtained by considering lemmas 1, 2 and 3.

Theorem 1

If $\alpha:G\to\mathcal{G}^{\mathrm{QCA}}$ is a symmetry action on a quantum spin chain with an anomaly index $\omega\not=1$, then there cannot be a locally-unique $G$-symmetric gapped ground state for a $G$-symmetric admissible $H$.

To prove this theorem, we employ an argument by contradiction. Assume that $\psi$ is a locally-unique gapped ground state of a $G$-symmetric admissible $H$. According to lemmas 1 and 2, $\psi$ must be pure and split. Additionally, it is $G$-symmetric by assumption. However, lemma 3 states that the anomaly index $\omega=1$, which contradicts our initial assumption.

Theorem 1 concerns about infinite systems, but real systems are of finite size. To extract useful implications on finite systems, we utilize another theorem.

Theorem 2 (Theorem V.1 in SM sup )

Suppose a sequence of $G$-symmetric admissible Hamiltonians, $\{H_{L}\}$, converges to an admissible Hamiltonian $H$ as $L\rightarrow\infty$. If each $H_{L}$ has a unique $G$-symmetric gapped ground state, then this sequence of ground states converges to a $G$-symmetric locally-unique gapped ground state of $H$ as $L\rightarrow\infty$.

Combining theorems 1 and 2, we deduce that if (a set of) large but finite systems described by admissible Hamiltonians with an anomalous symmetry have a well-defined thermodynamic limit, then they cannot have a unique $G$-symmetric gapped ground state.

Next, we move to theorem 3, which concerns about the entanglement property of a pure state with an anomalous symmetry in a finite system, and this state may not be a ground state of any natural Hamiltonian. Recall that a pure state in a finite but large system is short-range entangled (SRE) if it can be deformed into a product state by a finite time evolution under a local Hamiltonian, otherwise it is long-range entangled. In addition, consider the following lemma.

Lemma 4 (Propositions V.1 and V.2 in SM sup )

In the thermodynamic limit, an SRE state must be pure and splits.

Recall that pure states in finite systems can become mixed in the thermodynamic limit (e.g., the GHZ state). Lemma 4 ensures that this does not occur for SRE states. It also clarifies the relation between being SRE and the split property, i.e., the former implies the latter, but the converse may not be true in general.

By using Lemmas 3 and 4, we deduce

Theorem 3

There cannot be a $G$-symmetric short-range entangled pure state for sufficiently long spin chains if the anomaly index $\omega\not=1\in\mathrm{H}^{3}(G;\mathrm{U}(1))$.

Special versions of theorem 3 were proved before Else and Nayak (2014); Gioia and Wang (2022). For example, in a spin-1/2 chain no SRE state is compatible with a $G=\mathbb{Z}\times SO(3)$ symmetry Gioia and Wang (2022). A widely studied $G$-symmetric state in this case is the ground state of the nearest-neighbor anti-ferromagnetic Heisenberg model, which indeed realizes a conformal field theory with long-range entanglement. Our theorem applies to general symmetries described by QCA, which is extended to symmetries described by locality preserving automorphisms in SM sup .

Examples and applications – The first example is the spin-$1/2$ XXZ chain with long-range interactions Geier et al. (2021); Scholl et al. (2022). The Hamiltonian is

$$H=\sum_{i>j}\frac{1}{|i-j|^{\mathfrak{a}}}\left(J^{z}_{ij}\,S^{z}_{i}S^{z}_{j}-S^{x}_{i}S^{x}_{j}-S^{y}_{i}S^{y}_{j}\right)\;,$$ (5)

where $J^{z}_{ij}$ can be positive or negative. Notice that $H$ satisfies the admissible condition in Eq. (4) when $\mathfrak{a}>2$. This model has a $O(2)\times\mathbb{Z}$ symmetry, where $\mathbb{Z}$ is lattice translation and $O(2)$ is generated by rotation around the $z$-axis and reflection about the $xy$ plane. We have $\mathrm{H}^{3}(O(2)\times\mathbb{Z};\mathrm{U}(1))\simeq\mathbb{Z}_{2}$, which measures the on-site spin quantum number $S$ sup . If $S\in\mathbb{Z}+\frac{1}{2}$, the anomaly index is nontrivial and our theorems apply, which means that this model cannot have a unique symmetric gapped ground state.

From Ref. Maghrebi et al. (2017), for $\mathfrak{a}>2$ the phase diagram of Eq. (5) contains a ferromagnetic phase, an anti-ferromagnetic phase and a continuous symmetry breaking phase, which all spontaneously break some symmetries, and an XY phase, which is symmetric but gapless. Indeed, none of these phases has a unique symmetric gapped ground state, agreeing with our theorems.

In our second example, the only relevant symmetry is an anomalous $\mathbb{Z}_{2}$ internal symmetry. For each lattice site, we place a qubit. This $\mathbb{Z}_{2}$ symmetry acts as Levin and Gu (2012)

$$\begin{split}\alpha(Z_{i})&=-Z_{i}\;,\\ \alpha(X_{i})&=Z_{i-1}X_{i}Z_{i+1}\;,\end{split}$$ (6)

where $X_{j},Z_{j}$ are usual Pauli matrices, we also denote $Y_{j}=iX_{j}Z_{j}$. Formally, this symmetry is generated by conjugation with following infinite product

$$\prod_{j\in\mathbb{Z}}e^{\frac{i\pi}{4}Z_{j}Z_{j+1}}\prod_{k\in\mathbb{Z}}X_{k}\;.$$ (7)

This choice of symmetry action corresponds the nontrivial anomaly class in $\mathrm{H}^{3}(\mathbb{Z}_{2};\mathrm{U}(1))$ Kapustin and Sopenko (2024); sup . We consider the following Hamiltonian, which may be realizable in experimental setups similar to those in Refs. Geier et al. (2021); Scholl et al. (2022):

$$H=-\sum_{i,j}J_{ij}Z_{i}Z_{j}-\sum_{i}g_{i}(X_{i}+Z_{i-1}X_{i}Z_{i+1})-\sum_{j}h_{j}Y_{j}(1-Z_{j}Z_{j+1})\;,$$ (8)

where $J_{ij}=O(|i-j|^{-\mathfrak{a}})$ when $|i-j|\to\infty$ with $\mathfrak{a}>2$. For example, $J_{ij}$ can be chosen as

$$J_{ij}=\frac{C_{ij}}{|i-j|^{\mathfrak{a}}}\;,$$ (9)

where $\mathfrak{a}>2$ and $C_{ij}$ depend on $i,j$ and are bounded by constant $D$ for all $i,j$. This choice breaks the translation symmetry explicitly. Our theorems imply that this system is either gapless or breaks $\mathbb{Z}_{2}$ symmetry spontaneously.

In the special case where $h_{i}=0$, $g_{i}=0$ and $\mu_{i}=\eta_{j}=1$, this system is the classical long-range Ising model. The ground state is gapped and breaks the $\mathbb{Z}_{2}$ symmetry spontaneously. In the regime $g>0$ and $J_{ij}=h_{i}=0$, this model realizes a gapless Luttinger liquid Levin and Gu (2012). Both cases agree with our results. For more general couplings such as Eq. (9), other phases are also possible, which are left to future works.

Discussion.– In this work, we have proved generalized Lieb-Schultz-Mattis theorems in quantum spin chains with long-range interactions and anomalous symmetries. Our results apply to both discrete and continuous internal symmetries, as well as lattice translation symmetries. Currently it is unclear whether theorem 1 can be extended to the case where the 2-body interactions decay as $1/r^{\mathfrak{a}}$ with $\mathfrak{a}\leqslant 2$, and it is interesting to better understand this. Nevertheless, it should be obvious that theorem 1 cannot be extended to systems with extremely non-local interactions, such as a system whose Hamiltonian is simply a projector into a symmetric long-range entangled state. Also, it is useful to generalize our results to systems with point-group symmetry, fermionic systems and higher dimensional systems.

Note added – While completing this work, we became aware of an independent work by Ruochen Ma on related subjects, whose paper will appear in the same arXiv post.

Acknowledgement. We thank Theo Johnson-Freyd for helpful discussion. LZ thanks the Kavli Institute for Theoretical Physics for their hospitality during the course of this research. Research at Perimeter Institute is supported in part by the Government of Canada through the Department of Innovation, Science and Industry Canada and by the Province of Ontario through the Ministry of Colleges and Universities. This research was supported in part by grant NSF PHY-2309135 to the Kavli Institute for Theoretical Physics (KITP).