The Lee-Yang property of isotropic vector ferromagnets and lattice fields

This paper proves that isotropic spin and field models on the one-dimensional integer lattice $\mathds{Z}$ satisfy the generalized Lee-Yang property for all even dimensions $D$, extending a result previously known only for two-dimensional isotropic models.

The Gist

Imagine you are trying to predict the weather in a giant, complex city. You have millions of sensors (spins) scattered everywhere, each pointing in a different direction. Sometimes, they all want to point the same way because they are "friends" (ferromagnetic interaction), and sometimes an outside force (like a magnetic field) tries to push them in a specific direction.

In physics, there's a famous rule called the Lee-Yang Theorem. It's like a crystal ball that tells us something very specific about this system: if you look at the mathematical formula that describes the whole city's behavior (called the partition function), all the "dangerous spots" where the math breaks down (the zeros) are hidden in a very specific, safe place—they are purely imaginary numbers.

For decades, physicists knew this rule worked perfectly for simple systems where the sensors only pointed Left or Right (1-dimensional). Later, they figured out it also worked for systems where sensors pointed in a flat plane (2-dimensional, like a compass needle on a map).

But what if the sensors could point in 3D, 4D, or even higher dimensions? Like a spinning top that can wobble in every direction at once? For a long time, nobody knew if the "crystal ball" still worked for these complex, multi-dimensional systems, especially when the dimensions were 4 or higher.

The Big Breakthrough

This paper by Yuri Kozitsky is like finding the missing piece of a massive jigsaw puzzle. The author proves that for a specific type of "perfectly round" (isotropic) spinning systems, the Lee-Yang rule does hold true, but with a catch: it works for all even dimensions (2, 4, 6, 8, etc.).

Here is how the author solved it, using some creative analogies:

1. The "Magic Mirror" (Isotropy)

Imagine a spinning top that looks exactly the same no matter how you rotate it. In math, this is called isotropy. The author focuses on these perfectly symmetrical tops. Because they are so symmetrical, the math describing them becomes much simpler, almost like looking at a reflection in a magic mirror.

2. The "Dimensional Elevator"

The author discovered a clever trick to move between dimensions. Think of dimensions as floors in a building.

• We already knew the rule worked on the 2nd floor (2D).
• The author built an "elevator" that allows you to take a solution from the 2nd floor and instantly upgrade it to the 4th, 6th, or 8th floor without breaking the math.
• This elevator works because of a special property of the "single-spin measure" (the rules governing how one individual sensor behaves). If the single sensor follows a specific "safe" pattern (mathematically called a Laguerre entire function), the whole building stays safe.

3. The "Domino Effect" (Induction)

The proof uses a method called induction, which is like knocking over a line of dominoes.

• Step 1: Prove the rule works for a tiny system (just 2 sensors).
• Step 2: Show that if the rule works for a system of $N$ sensors, it must also work for a system of $N+1$ sensors.
• Step 3: Because the "elevator" connects the 2D world to the 4D, 6D, etc., worlds, proving it for the 2D case automatically proves it for all the even-numbered higher dimensions.

Why Does This Matter?

You might ask, "Who cares about imaginary zeros in a math formula?"

In the real world, these zeros tell us about phase transitions. This is the moment when a material changes its state, like ice melting into water or a magnet losing its magnetism when it gets too hot.

• If the zeros are in the "safe" imaginary place, the system behaves predictably.
• If the zeros wander off into dangerous territory, the system might behave chaotically or unpredictably.

By proving that these zeros stay in the safe zone for even-dimensional "rotors" (spinning tops), the author gives physicists and mathematicians a powerful tool. It means they can use advanced, high-speed mathematical techniques to study complex quantum materials and fields, knowing that the underlying math is stable and reliable.

The Bottom Line

Think of this paper as a master key. For a long time, we had a key that opened doors for 1D and 2D systems. This paper forges a new key that opens the doors for all even-dimensional systems. It confirms that nature's symmetry in these complex, multi-directional spinning systems is robust, keeping the mathematical "danger zones" neatly tucked away where they belong.

Technical Summary

1. Problem Statement and Context

The Core Problem:
The paper addresses a long-standing open problem in mathematical physics regarding the Lee-Yang property for isotropic vector spin models (rotors) in dimensions $D \geq 4$.

• The Lee-Yang Property: Originally proven by Lee and Yang (1952) for the Ising model ($D=1$), this property states that the zeros of the partition function $Z(h)$, viewed as a function of the external magnetic field $h$, lie purely on the imaginary axis.
• Generalization: For vector models where spins are $D$-dimensional vectors ($D \geq 2$), the property implies that the partition function can be represented as an infinite product involving only $z^2$ (where $z$ is the complex magnetic field vector) with positive coefficients. Specifically, $Z_N(z) = Z_N(0) \prod (1 + \gamma_j z^2)$.
• The Gap: While the property was established for $D=2$ (by Lieb and Sokal), it remained unproven for $D \geq 4$. Barry Simon (2019) explicitly listed proving a Lee-Yang theorem for isotropic classical $D$-rotors for $D \geq 4$ as a major open conjecture.

The Specific Model:
The author studies a model on the one-dimensional lattice $\mathbb{Z}$ (a chain) with nearest-neighbor ferromagnetic interactions ($J > 0$). The spins $\sigma_l$ are $D$-dimensional vectors, and the single-spin measure $\chi$ is isotropic (invariant under orthogonal transformations $O(D)$).

2. Methodology

The proof relies on a sophisticated combination of statistical mechanics, complex analysis, and the theory of entire functions.

A. Strongly Isotropic Measures and Laplace Transforms

• The author introduces the concept of strongly isotropic measures. A measure $\chi$ is strongly isotropic if it can be expressed as $\chi(d\sigma) = \tau(\sigma^2)d\sigma$, where $\tau$ is a distribution on $\mathbb{R}_+$.
• The core assumption is that the Laplace transform of the single-spin measure, $\hat{\chi}(z)$, belongs to the class $\mathcal{L}$ of Laguerre entire functions of the first kind. These are functions of the form $C \prod (1 + \gamma_j z^2)$ with $\gamma_j > 0$.
• The author utilizes the property that differentiation preserves the class $\mathcal{L}$ (i.e., if $f \in \mathcal{L}$, then $f' \in \mathcal{L}$).

B. Reduction from Dimension $D$ to $D=2$

• A critical technical innovation is the reduction of the problem from arbitrary even dimensions $D$ to the two-dimensional case ($D=2$).
• The author defines differential operators $\Delta_{k,D}$ that relate the partition functions in dimension $D$ to those in dimension $D-2$.
• By using the identity $\partial_1 \Delta_{k,D} = \Delta_{k,D} \partial_1$, the author shows that if the Lee-Yang property holds for $D=2$, it can be extended to $D = 2 + 2m$ via repeated differentiation. This bypasses the need to prove the property directly for high dimensions.

C. Inductive Proof via Lieb-Sokal Theorem

• The proof proceeds by induction on the number of spins $N$.
• It relies on a theorem by Lieb and Sokal (1979), which establishes that if a partition function with zero interaction ($J=0$) has no zeros in a specific domain $L \subset \mathbb{C}^2$, then the partition function with ferromagnetic interactions ($J>0$) also has no zeros in that domain.
• The author defines a mapping $\ell_{2,2}$ from the complex plane $\mathbb{C}^2$ to $\mathbb{C}^3$ (representing squared norms and dot products). They prove that the preimage of the domain where the function is non-zero (the cut plane $\mathbb{C}_-$) under this mapping corresponds exactly to the domain $L$ required by Lieb and Sokal.

D. Gram Map and Symmetry

• Due to the isotropic nature of the spins, the partition function $F_N(z_1, z_2)$ depends only on the scalar products $z_i \cdot z_j$.
• The author uses Gram maps to reduce the variables from vectors in $\mathbb{C}^D$ to scalar invariants. For $D \geq 2$, the map is surjective onto $\mathbb{C}^3$, allowing the problem to be treated in terms of scalar variables $\zeta_1, \zeta_2, \zeta_{12}$.

3. Key Contributions

1. Resolution of Simon's Conjecture: The paper proves the Lee-Yang theorem for isotropic classical $D$-rotors and Euclidean quantum lattice fields for all even dimensions $D$. This resolves Conjecture 9.2.27 from Barry Simon's list of open problems.
2. Generalization of Lieb-Sokal: The work extends the Lieb-Sokal theorem (originally for $D=2$) to all even $D$ by constructing a rigorous reduction mechanism using differential operators and the properties of Laguerre entire functions.
3. Introduction of Strong Isotropy: The paper formalizes the class of "strongly isotropic" measures, providing a unified framework to handle measures across different dimensions via a single generating distribution $\tau$.
4. Handling of Quantum Fields: The results apply not just to classical rotors but also to specific models of Euclidean quantum lattice fields (e.g., $\phi^4$ theory and higher-order polynomial interactions), provided the single-spin measure satisfies the Laguerre condition.

4. Main Results

Theorem 2.4 (Main Result):
Let $\chi$ be a strongly isotropic measure on $\mathbb{R}^D$ that possesses the Lee-Yang property (i.e., its Laplace transform is in class $\mathcal{L}$). Then, for every ferromagnetic coupling $J > 0$ and every even integer $D$, the partition function of the model on the lattice $\mathbb{Z}$:
$$Z_N(z) = \int \exp\left( \sum_{l=1}^N z \cdot \sigma_l + J \sum_{l=1}^{N-1} \sigma_l \cdot \sigma_{l+1} \right) \prod_{l=1}^N \chi(d\sigma_l)$$
can be represented as:
$$Z_N(z) = Z_N(0) \prod_{j=1}^\infty (1 + \gamma_{j,N} z^2)$$
where $\gamma_{j,N} > 0$ and $\sum \gamma_{j,N} < \infty$.

Implications:

• The zeros of the partition function are purely imaginary (since $z^2$ must be negative real for the term to vanish).
• The result holds for measures such as the uniform distribution on a sphere (classical rotors) and measures with potentials like $e^{-a\sigma^2 - b(\sigma^2)^2}$ (quantum fields).

5. Significance

• Mathematical Physics: The Lee-Yang property is a cornerstone for understanding phase transitions, stability, and the analyticity of thermodynamic functions. Proving this for $D \geq 4$ removes a significant barrier in the rigorous analysis of high-dimensional vector models.
• Quantum Field Theory: The results provide rigorous foundations for the study of Euclidean quantum lattice fields in higher dimensions, ensuring that certain pathological behaviors (like zeros in the physical region) do not occur for these specific isotropic models.
• Analytic Methods: The paper demonstrates the power of combining Lieb-Sokal's zero-free domain techniques with the theory of Laguerre-Pólya class functions. The reduction technique (using derivatives to move between dimensions) offers a new tool for tackling similar problems in statistical mechanics where dimensionality is a hurdle.

Limitations:
The proof currently relies on the lattice being one-dimensional ($\mathbb{Z}$) to manage the recurrence relations and the rank constraints of the Gram matrices. Extending this to higher-dimensional lattices (e.g., $\mathbb{Z}^d$) remains an open challenge, as noted in the concluding remarks. Additionally, the result is restricted to even dimensions $D$; the case for odd $D$ is not covered by this specific reduction method.