🔢 mathematics

The role of density in phase transitions and new perspectives on theta functions

This paper establishes a complete phase diagram for the minimization of the functional $\theta(\alpha;z)-b\theta(2\alpha;z)$ , revealing a unique "hexagonal–square–rectangular" phase transition governed by density and competition parameters that resolves Bétermin's conjecture by proving the absence of rhombic lattice minimizers and correcting the predicted transition type.

Original authors: Kaixin Deng, Senping Luo

Published 2026-10-27

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Kaixin Deng, Senping Luo

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the microscopic world of atoms and electrons, particles often arrange themselves into repeating patterns called lattices. Imagine a vast floor tiled perfectly with identical shapes; the way these shapes fit together determines the material's energy and stability. For decades, scientists have known that when particles interact through a single, simple force, the most efficient, lowest-energy arrangement is a honeycomb pattern, where every point has six neighbors in a perfect triangle. This hexagonal structure is so robust that it appears to be the universal winner, no matter how densely the particles are packed. However, nature is rarely so simple. When different forces compete—some pulling particles together, others pushing them apart—the optimal shape can change. The question then becomes: as the density of the particles shifts, does the lattice simply stay hexagonal, or does it transform into a different shape to find a new balance?

A team of mathematicians has now mapped out exactly how this transformation happens when two competing forces are at play. They studied a specific mathematical model where one force tries to organize particles into a hexagonal pattern, while a second, opposing force tries to disrupt that order. By rigorously calculating the energy of every possible lattice shape across all densities, they discovered that the path to the lowest energy is not always a straight line. Instead, the system undergoes a series of distinct phase transitions, shifting from a hexagonal arrangement to a square one, and then to a rectangular one. Most surprisingly, they found that for certain conditions, the lattice passes through a temporary, diamond-like shape known as a rhombus before settling into its final form. For other conditions, this intermediate step vanishes entirely, and the system jumps directly from hexagonal to square.

The researchers focused on a specific scenario that had puzzled the scientific community. A previous study had suggested that for a particular balance of forces, the lattice would follow a complex four-stage journey: starting as a hexagon, morphing into a rhombus, then a square, and finally a rectangle as the density changed. This idea was based on numerical observations that hinted at the existence of that elusive rhombic phase. The new work, however, proves that this specific prediction was incorrect. Through a complete and rigorous analysis, the authors demonstrate that for this specific balance of forces, the rhombic phase never actually appears as the most stable shape. Instead, the system skips that middle step entirely, moving directly from the hexagonal structure to the square one. This finding corrects a long-standing conjecture and reveals that the "rhombic" phase is a fragile state that only exists under a very narrow set of conditions, vanishing completely once the competition between forces reaches a certain threshold.

To reach this conclusion, the authors had to navigate a vast mathematical landscape. They treated the lattice shape as a variable that could be stretched, squashed, and rotated, calculating the energy for every possible configuration. They identified precise tipping points where the most stable shape suddenly changes. One key discovery was the existence of a critical threshold value that acts as a switch. Below this value, the system exhibits the full four-stage transition, including the rhombic phase. Above this value, the rhombic phase is impossible, and the system follows a simpler three-stage path. The researchers calculated this threshold with extreme precision, determining it to be approximately 0.0767767081. This number is the dividing line between two fundamentally different behaviors of the material.

The study also clarified the sequence of shapes the lattice takes as it moves from low density to high density. At low densities, the particles prefer a rectangular arrangement, which is stretched out. As the density increases, this rectangle tightens into a perfect square. At a specific point, the square abruptly snaps into a hexagonal honeycomb pattern, which remains the most stable shape at very high densities. The transition from square to hexagonal is sharp and sudden, occurring at a precise density value. The researchers also identified the exact mathematical conditions under which the system might briefly pause in a rhombic state, but they confirmed that for the specific case of the corrected conjecture, this pause does not happen. The lattice flows smoothly from square to hexagon without getting stuck in the middle.

This work is significant because it moves beyond numerical guesses to provide a complete, proven map of how these competing forces dictate the structure of matter. It shows that the "winner" in the battle for the lowest energy is not always the same shape, and the path it takes to get there depends delicately on the strength of the competing forces. The researchers did not just simulate these changes; they proved them mathematically, ruling out the possibility of any other hidden shapes or unexpected transitions. Their results settle a specific debate about the behavior of double-Gaussian lattice energies, showing that the complex four-stage transition predicted by earlier numerical work is actually a misinterpretation of a simpler, three-stage reality for that specific case.

The implications of this work extend beyond abstract mathematics. The models used here appear in various physical systems, from the arrangement of atoms in crystals to the behavior of vortices in superconductors and the organization of particles in soft matter. Understanding exactly how and when these structures change shape helps scientists predict the properties of new materials. By knowing that a rhombic phase is absent under certain conditions, researchers can avoid looking for a structure that simply does not exist in that regime. The paper provides a definitive guide to the "phase diagram" of these systems, showing exactly where the boundaries lie between different geometric worlds.

In the end, the study reveals a world where the most efficient shape is not a static ideal but a dynamic response to changing conditions. The hexagonal lattice, long thought to be the undisputed champion, shares the stage with squares and rectangles, and occasionally, fleeting rhombuses. The transition between these shapes is governed by precise mathematical laws that the authors have now fully charted. Their work corrects a previous misunderstanding, proving that for a key case, the system does not take a detour through a rhombic phase but instead follows a more direct route. This clarity allows for a deeper understanding of how competition between forces shapes the fundamental architecture of the physical world.

Technical Summary: Minimization of the Difference of Theta Functions and Phase Transitions in Lattice Energies

Problem Statement
This paper investigates the minimization of the functional $\theta(\alpha; z) - b\theta(2\alpha; z)$ over the upper half-plane $\mathbb{H}$ , where $\theta(\alpha; z)$ denotes the theta function associated with a two-dimensional Bravais lattice of unit covolume parametrized by $z = x + iy$. The parameter $\alpha > 0$ acts as an inverse density (or Gaussian scale), while $b > 0$ represents a competition coefficient between two Gaussian interaction terms.

The study addresses a specific gap in the literature regarding lattice energy minimization. While Montgomery's theorem establishes that the hexagonal lattice minimizes a single theta function $\theta(\alpha; z)$ at all scales, and recent work by B'etermin, Faulhuber, and Steinerberger confirms hexagonal optimality in min-max structures for translated sums, the behavior of the difference of two Gaussian sums remains less understood. Specifically, the authors aim to determine how the optimal lattice shape evolves as density varies when a second Gaussian term with a negative coefficient is introduced. This problem is motivated by B'etermin's numerical observation of a "triangular–rhombic–square–rectangular" phase transition for the specific case $b = 1/10$ , which the authors rigorously analyze and correct.

Methodology
The authors employ a combination of modular symmetry, calculus of variations, and rigorous analytic estimates to reduce the infinite-dimensional minimization problem to a finite set of boundary cases.

Symmetry Reduction: Utilizing the modular group $SL_2(\mathbb{Z})$ and the extended symmetry group $G$ , the authors restrict the search for minimizers to the fundamental domain $D_G = \{z \in \mathbb{H} : |z| > 1, 0 < \text{Re}(z) < 1/2\}$ . The problem is further reduced to analyzing the function on the boundary components $\Gamma_1 = \{iy : y \ge 1\}$ (rectangular lattices) and $\Gamma_2 = \{e^{i\theta} : \theta \in [\pi/3, \pi/2]\}$ (rhombic lattices).
Boundary Reduction via Monotonicity: The core technical strategy involves proving that the global minimizer must lie on the boundary $\Gamma_1 \cup \Gamma_2$ $Γ_{1} \cup Γ_{2}$ . This is achieved by establishing horizontal and vertical monotonicity properties of the functional $E(\alpha; z) = \theta(\alpha; z) - \frac{1}{20}\theta(\alpha/2; z)$ $E (α; z) = θ (α; z) - \frac{1}{20} θ (α /2; z)$ (equivalent to the original problem via Poisson summation).
- For large $\alpha$ (low density), the authors prove horizontal monotonicity ( $\partial_x E \ge 0$ ) by analyzing quotients of theta derivatives and bounding error terms in series expansions.
- For intermediate $\alpha$ , they apply a minimum principle based on weighted convexity inequalities involving the operator $(\partial_y^2 + \frac{2}{y}\partial_y)$ and mixed derivative positivity ( $\partial_{xy} E \ge 0$ ).
Critical Point Analysis: Once reduced to the boundaries, the authors analyze the critical points of the restricted functions. They utilize:
- Derivative Quotients: Analyzing the monotonicity of ratios like $\frac{\theta_y(\alpha; iy)}{\theta_y(\alpha/2; iy)}$ to locate unique critical points on $\Gamma_1$ .
- Shape of Quotients on $\Gamma_2$ : Proving that the ratio of derivatives on the arc $\Gamma_2$ is "U-shaped" (decreasing then increasing), which limits the number of interior critical points.
- Energy Comparisons: Rigorously comparing energy values at critical points (square, hexagonal, and interior rhombic) to determine global minimizers.
Numerical and Asymptotic Verification: The proofs rely on precise numerical bounds for theta function derivatives and series tails, alongside asymptotic expansions for small $\alpha$ derived from the Jacobi theta function $\vartheta_3$ .

Key Contributions and Results

Resolution of B'etermin's Conjecture: The paper rigorously resolves the conjecture regarding the functional with $b = 1/10$ . Contrary to the numerical prediction of a "triangular–rhombic–square–rectangular" sequence, the authors prove that the rhombic phase is absent. The system undergoes a Type II phase transition:
- Rectangular ( $y > 1$ ) for low densities ( $\alpha < \alpha_a$ ).
- Square ( $z=i$ ) for intermediate densities ( $\alpha_a \le \alpha < \alpha_b$ ).
- Hexagonal ( $z=e^{i\pi/3}$ ) for high densities ( $\alpha > \alpha_b$ ).
  The transition from rectangular to square is continuous, while at the critical threshold $\alpha_b$ , the minimizer jumps from the square to the hexagonal lattice, with both configurations being minimizers at the transition point.
Complete Phase Diagram for $0 < b < \sqrt{2}$ : The authors establish a comprehensive classification of phase transitions dependent on the competition parameter $b$ . They identify a critical threshold $b_1 \approx 0.0767767$ :
- Type I ( $0 < b < b_1$ ): The system exhibits a "hexagonal–rhombic–square–rectangular" transition. As density decreases, the minimizer passes through an intermediate rhombic phase where the angle $\theta$ varies continuously.
- Type II ( $b_1 \le b < \sqrt{2}$ ): The rhombic phase vanishes. The system transitions directly from hexagonal to square to rectangular, mirroring the behavior found for $b=1/10$ .
Threshold Characterization: The paper provides explicit definitions and numerical values for the critical thresholds ( $\alpha_a, \alpha_b, b_1$ , etc.) that separate these phases. These thresholds are defined as unique solutions to equations involving theta functions and their derivatives (e.g., $\theta_{yy}(\alpha; i) = b\theta_{yy}(2\alpha; i)$ ).
Asymptotic Analysis: The authors derive precise asymptotic expansions for the minimizer $y_\alpha$ as $\alpha \to 0^+$ , characterizing the approach to the rectangular limit.

Significance
The paper claims significance in providing the first rigorous classification of global minimizers for the difference of two Gaussian theta functions across the full range of densities and competition parameters. It corrects a previously held numerical conjecture regarding the existence of a rhombic phase for $b=1/10$ , demonstrating that the rhombic phase only emerges for sufficiently small competition coefficients ( $b < b_1$ ).

By distinguishing between Type I and Type II transitions, the work elucidates how the competition parameter $b$ controls the emergence of intermediate lattice symmetries. This contrasts with previous results where the hexagonal lattice was universally optimal or where transitions were studied in different contexts (e.g., min-max structures or single potentials). The results offer a systematic framework for understanding density-dependent lattice shape selection in systems with competing Gaussian interactions, relevant to models in soft matter and periodic defect structures.

More like this