Analyticity, asymptotics and natural boundary for a one-point function of the finite-volume critical Ising chain

This paper demonstrates that the finite-volume critical Ising chain's one-point function, when analytically continued via Borel resummation, possesses a natural boundary along the negative real axis governed by odd-divisor-squared sums, which also dictate the strengths of its factorially divergent large-$N$ asymptotics.

The Gist

Imagine you are trying to predict the weather in a very specific, tiny room. You have a formula that works perfectly when the room is huge (like a stadium). But what happens when you shrink the room down to the size of a shoebox?

This paper is about a physicist, Yizhuang Liu, who decided to look at a famous model of magnetism (the Ising chain) inside a tiny, finite-sized room. He wanted to see how the "spin" of the atoms (think of them as tiny compass needles) behaves when the room gets smaller and smaller.

Here is the story of what he found, explained without the heavy math:

1. The Smooth Road and the Sudden Wall

Usually, in physics, if you have a formula that works for big numbers, you can imagine "sliding" it to work for negative numbers or complex numbers, just like driving a car from a highway onto a side street. You expect the road to stay smooth.

Liu took his formula for the tiny room and tried to drive it into the "negative" territory (mathematically speaking).

• The Expectation: He thought the road would stay smooth, maybe with a few potholes.
• The Reality: As he drove closer to a specific line (the negative real axis), the road didn't just get bumpy; it turned into a cliff.

In math, this cliff is called a Natural Boundary. It's like hitting a wall of fog so thick that you can't see through it, and you can't drive any further. No matter how hard you try to extend the formula, it simply breaks down. This is surprising because usually, these "walls" only appear in very abstract math, not in the physical properties of a simple magnet.

2. The "Divisor" Mystery

Why did the wall appear? Liu discovered that the behavior of the magnet is secretly controlled by a very old, very picky number game called Divisor Sums.

Imagine you have a pile of bricks. You want to stack them in rows.

• If you have 6 bricks, you can stack them in rows of 1, 2, 3, or 6.
• If you have 7 bricks, you can only stack them in rows of 1 or 7.

The formula for the magnet's behavior depends on counting these stacking possibilities in a very specific, weird way.

• The Analogy: Think of the magnet's behavior as a song. The "notes" of the song are determined by how many ways you can stack bricks.
• The Problem: When the room size (the number of bricks) changes, the "notes" jump around wildly. Sometimes the song is smooth; other times, it becomes a chaotic, screaming mess.

This chaos is what creates the "Natural Boundary." The math is so sensitive to the exact number of atoms in the room (and whether that number is odd, even, or a specific type of fraction) that the formula refuses to work smoothly across the line.

3. The "Lambert" Series (The Infinite Echo)

The paper mentions something called a Lambert Series. Imagine you are in a canyon and you shout.

• In a normal canyon, the echo fades away.
• In this math canyon, the echo bounces back, but every time it returns, it changes slightly based on the "divisor" rules.

Liu found that the chaotic behavior near the cliff is exactly the same as this infinite, echoing song. The "strength" of the echo (how loud the chaos is) is determined by those same brick-stacking numbers.

4. Why This Matters

You might ask, "Who cares about a tiny magnet in a shoebox?"

• The Bridge: This discovery connects two worlds that usually don't talk to each other: Physics (how magnets work) and Pure Number Theory (the study of integers and their secrets).
• The Surprise: It shows that even in a simple, perfect system (like the Ising model), if you look closely enough at the "finite" details (the size of the room), the universe behaves like a chaotic number puzzle.
• The "Leg" Function: The paper also looks at how particles "walk" (form factors) in this tiny room. It turns out their walking pattern is also governed by these same weird number rules, linking the magnet to advanced theories in string theory and quantum geometry.

The Big Takeaway

Think of the universe as a giant, smooth tapestry. Usually, we can zoom out and see a nice, continuous picture. But Liu found a spot where, if you zoom in too close to the "size" of the system, the tapestry unravels into a chaotic knot of numbers.

He proved that nature has a "no-entry" zone for this specific calculation. You can approach the edge, but you cannot cross it, because the underlying math is too jagged and full of number-theoretic traps to allow a smooth path. It's a beautiful reminder that even in the simplest models of reality, the deep structure of mathematics is always hiding in the shadows, ready to surprise us.

Technical Summary

1. Problem Statement

The paper investigates the analytic properties of the finite-volume expectation value of the spin operator (the one-point function) in the critical periodic Ising chain. Specifically, it focuses on the overlap between the $Z_2$-even and $Z_2$-odd ground states, denoted as $|\langle \Omega_R | \hat{\sigma} | \Omega_{NS} \rangle|^2$.

While Conformal Field Theory (CFT) predicts simple power-law scaling for this quantity in the thermodynamic limit, finite-size corrections introduce a function $v(1/N)$ (where $N$ is the system size). The central problem is to understand the analytic continuation of this function $v(1/N)$ from the positive real axis (physical system sizes) into the complex plane. The author demonstrates that, contrary to standard expectations for integrable lattice models, this function possesses a natural boundary of analyticity along the negative real axis, preventing further analytic continuation.

2. Methodology

The author employs a combination of complex analysis, number theory, and resurgent asymptotics:

• Integral Representations: The study begins with an integral representation of $v(1/N)$ derived from the lattice model.
• Mellin-Barnes Representation: The integral is converted into a Mellin-Barnes integral involving Riemann zeta functions ($\zeta(u-1)\zeta(-u)$). This allows for the systematic derivation of the large-$N$ asymptotic expansion by shifting the integration contour.
• Reflection Formulas: To analyze the behavior near the negative real axis, the author derives reflection formulas relating $v(1/N)$ to $v(-1/N)$. This transforms the problem into analyzing an infinite sum involving logarithmic terms and trigonometric functions.
• Lambert Series and Divisor Sums: The singular part of the reflection formula is identified as a Lambert series. The coefficients of this series are linked to a specific divisor sum, $\sigma_{-2}^o(n)$, which sums over odd divisors.
• Borel Resummation: The paper utilizes the Nevanlinna-Sokal theorem to prove that the factorially divergent large-$N$ expansion is Borel summable to the original function $v(1/N)$.
• Connection to Chern-Simons Theory: The author connects the finite-volume one-point function to the partition function of $U(N)$ Chern-Simons theory on $S^3$, providing an alternative derivation of the asymptotics.

3. Key Contributions and Results

A. Existence of a Natural Boundary

The primary result is the proof that $v(1/N)$ has a natural boundary along the negative real axis ($N \in \mathbb{R}_{<0}$).

• Mechanism: The singularity arises from the behavior of the function as the system size $N$ approaches negative values. The analytic continuation is blocked by a dense set of singularities.
• Number-Theoretic Origin: The nature of the singularity depends on the arithmetic properties of the system size (treated as a complex variable).
  • For dyadic rationals (numbers of the form $x = \frac{2q+1}{2^p}$), the singularity is logarithmic: $\sim \ln|y|$.
  • For rationals with odd denominators (no factors of 2), the singularity is quadratic: $\sim y^{-2}$.
  • For irrational numbers, the singularity is bounded by $y^{-2}$ but exhibits complex fluctuations.
• Since rational numbers are dense, the accumulation of these singularities creates a natural boundary, making analytic continuation beyond the negative real axis impossible.

B. Role of Divisor Sums and Lambert Series

The singular behavior is governed by a Lambert-type series $S(z)$:
$$S(z) = -4 \sum_{k=0}^{\infty} \frac{1}{2k+1} \frac{z^{2k+1}}{(z^{2k+1}+1)^2}$$
where $z = e^{-i\pi x - \pi y}$.

• The coefficients of the Taylor expansion of this series are determined by the odd-divisor-squared sum:
  $$\sigma_{-2}^o(n) = \sum_{d|n, d \text{ odd}} \frac{1}{d^2}$$
• The irregularities in the growth of $\sigma_{-2}^o(n)$ are directly responsible for the non-modular nature of the series and the formation of the natural boundary. This contrasts with standard CFT thermal averages, which often involve modular forms.

C. Borel Singularities and Resurgence

The large-$N$ expansion of $v(1/N)$ is factorially divergent but Borel summable.

• The Borel transform $B[v](t)$ possesses infinitely many double poles along the imaginary axis at $t = 2\pi i (2k+1)n$.
• The strength of the discontinuities (Stokes phenomena) across the Stokes line is controlled by the same divisor sum $\sigma_{-2}^o(n)$.
• This establishes a direct link between the non-perturbative ambiguities (Borel singularities) and the natural boundary in the complex plane.

D. Connection to Chern-Simons Theory

The paper reveals that the finite-volume one-point function can be expressed as a ratio of Chern-Simons partition functions on $S^3$:
$$P(q) = \frac{Z^2(2N, 2N)}{Z^8(N, N)}$$
This connection explains the appearance of specific combinations of Bernoulli numbers in the asymptotic expansion and suggests that the natural boundary is a feature shared with the perturbative expansion of Chern-Simons theory at specific coupling constants.

4. Significance

• Rare Example in Physics: Natural boundaries are common in pure mathematics (e.g., lacunary series) but rare in the ground-state properties of integrable lattice models, especially at infinite volume limits. This paper provides a concrete physical example where the lattice structure imposes a natural boundary on a ground-state observable.
• Lattice vs. Continuum: The result highlights that certain analytic properties (like the natural boundary) are intrinsic to the lattice level and are lost in the continuum CFT limit. The boundary arises from the discrete nature of the system size and its number-theoretic properties.
• Resurgent Structure: The work deepens the understanding of resurgence in quantum field theory by linking the coefficients of the asymptotic expansion (divisor sums) directly to the geometry of the natural boundary and the Stokes discontinuities.
• Cross-Disciplinary Links: It bridges statistical mechanics (Ising model), number theory (divisor sums, Lambert series), and topological field theory (Chern-Simons theory), suggesting that similar "non-modular" behaviors may appear in other contexts involving finite-volume effects or specific matrix models.

In summary, the paper demonstrates that the finite-size corrections to the critical Ising chain's one-point function are not merely smooth analytic functions but possess a complex singular structure dictated by number theory, fundamentally limiting their analytic continuation.