🔬 mesoscale physics

Characterizing Liouvillian Exceptional Points Through Newton Polygons and Tropical Geometry

This paper demonstrates that Newton polygons and tropical geometry provide an effective analytical framework for identifying and characterizing the order and anisotropy of Liouvillian exceptional points in open quantum systems, as validated through dissipative spin and superconducting qubit models.

Original authors: Sayooj P, Awadhesh Narayan

Published 2026-08-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Sayooj P, Awadhesh Narayan

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 quantum world, the rules of stability are often written in the language of energy levels. For decades, physicists have studied how these levels behave in systems that are perfectly isolated, where energy is conserved and the mathematics describing them is straightforward. But the real universe is rarely isolated. Open quantum systems constantly exchange energy and information with their surroundings, a process known as dissipation. When scientists describe these open systems, they use a specific mathematical framework called the Lindblad master equation. Unlike the simpler models for isolated systems, this framework allows for a peculiar phenomenon called an exceptional point. At these points, the usual distinction between different states of a system collapses; the energy levels and the shapes of the states themselves merge into a single, degenerate entity. These points are not just mathematical curiosities; they are known to create strange effects, such as extreme sensitivity to tiny changes or unusual shifts in the system's behavior, making them a hot topic for researchers trying to build better sensors or quantum computers.

A team of researchers at the Indian Institute of Science has now developed a new way to find and understand these exceptional points in open quantum systems. Instead of relying solely on heavy numerical calculations, they applied two powerful mathematical tools from the fields of algebraic geometry and polyhedral geometry. One tool is the Newton polygon, a method originally devised centuries ago to solve complex polynomial equations, and the other is tropical geometry, a modern branch of mathematics that simplifies the study of solutions by focusing on their most dominant features. By combining these approaches, the researchers created a method that can directly identify the order and nature of exceptional points in open systems. They tested their method on two distinct physical models: a simple spinning particle that loses energy to its environment, and a more complex superconducting qubit, a type of artificial atom used in quantum computing.

The researchers found that their geometric approach could map out the entire landscape of these exceptional points with surprising clarity. In the case of the spinning particle, the method successfully located all the points where the system's states merge. More importantly, the shape of the geometric maps they constructed revealed exactly how the system would react if it were nudged. They could see that the energy levels would split apart in a specific way, following a square-root pattern as the system moved away from the exceptional point. This was confirmed by running direct computer simulations, which showed that when the researchers circled around these points, the energy levels swapped places in a predictable cycle, a hallmark of these special degeneracies.

The study took a deeper turn when they examined the superconducting qubit model, where they discovered a crucial detail about how these points behave. They found that the "order" of an exceptional point—essentially, how many states are merging together—is not a fixed property of the system alone. Instead, it depends entirely on how the system is disturbed. In one scenario, when they perturbed the system by changing the rate of energy loss, four states merged into a single point, but they split apart into two distinct groups when the disturbance was applied. One group of states separated linearly, while the other three separated according to a cube-root pattern. However, when they changed the type of disturbance to a different parameter, the behavior shifted dramatically. In this second case, only two of the states actually split apart in a square-root fashion, while the other two remained completely unchanged, as if the disturbance had no effect on them at all.

This discovery highlights a profound anisotropy, or directional dependence, in how these quantum systems respond to the outside world. The researchers demonstrated that by choosing a specific type of perturbation, one could effectively lower the order of an exceptional point, turning a complex four-way merger into a simpler two-way split. Their geometric tools, the Newton polygons and the associated shapes called amoebas, acted like a blueprint, allowing them to predict these outcomes before running any simulations. The "amoeba" in this context is a specific geometric shape derived from the system's equations that visually displays the scaling behavior of the energy levels. The researchers showed that the slopes of the lines forming this shape directly correspond to the mathematical rules governing how the states split.

The work suggests that these geometric methods are not just a way to analyze existing systems but could be a guide for designing them. If scientists want to create a quantum system with a specific type of exceptional point for a particular application, they can use these tools to determine exactly which parameters to tune and which disturbances to apply. The paper concludes that this analytical approach offers a powerful, direct route to understanding the intricate topology of open quantum systems. By translating the complex algebra of quantum dissipation into clear geometric shapes, the researchers have provided a new lens through which to view the delicate balance of energy and information in the quantum realm, potentially aiding in the design of more robust and sensitive quantum devices.

Technical Summary: Characterizing Liouvillian Exceptional Points Through Newton Polygons and Tropical Geometry

Problem Statement
Open quantum systems described by the Lindblad master equation are governed by non-Hermitian operators, which can host non-Hermitian degeneracies known as Liouvillian exceptional points (EPs). At these points, both eigenvalues and eigenvectors coalesce, leading to unique phenomena such as geometric phases, enhanced sensitivity, and specific phase transitions. While EPs in non-Hermitian Hamiltonians have been extensively studied, Liouvillian EPs present distinct challenges due to the effects of quantum jumps and the structure of the Liouvillian superoperator. A critical aspect of understanding these systems is determining the order of an EP and its anisotropy—how the eigenvalue scaling depends on the specific form of the perturbation applied. Existing methods often require direct numerical diagonalization or rely on specific perturbative expansions that may not capture the full structural dependence on the perturbation matrix.

Methodology
The authors propose a unified analytical framework utilizing Newton polygons and Tropical geometry to characterize Liouvillian EPs. The method relies on the following mathematical foundations:

Newton-Puiseux Theorem: This theorem establishes that solutions to algebraic equations (specifically the characteristic equation of the Liouvillian) can be expressed as Puiseux series in a perturbation parameter $\epsilon$ . The order of the EP corresponds to the valuation (lowest power) of this series.
Newton Polygons: By plotting the valuations of the coefficients of the characteristic polynomial against their powers, a convex hull (the Newton polygon) is formed. The slopes of the polygon's segments directly reveal the valuations of the solutions. Specifically, the negative of a segment's slope corresponds to the scaling exponent of the eigenvalues ( $\epsilon^{1/m}$ ), and the horizontal projection of the segment indicates the number of eigenvalues sharing that scaling.
Tropical Geometry and Amoebas: The authors employ the concept of the "amoeba" of an algebraic variety. The amoeba is the image of the solution set under a logarithmic map. The asymptotic "tentacles" of the amoeba correspond to the normal vectors of the Newton polygon's edges. The slopes of these tentacles provide a geometric visualization of the eigenvalue scaling and can identify invariant solutions (where eigenvalues do not split under perturbation).
Perturbation Analysis: The method distinguishes between "generic" perturbations (random matrices where no fine-tuned relations exist) and "non-generic" perturbations. The authors demonstrate that the structure of the Newton polygon changes if specific coefficients in the characteristic polynomial vanish, which occurs for non-generic perturbations.

Key Contributions and Results
The paper validates this methodology using two specific models:

Dissipative Spin-1/2 System:
- The authors apply the method to a spin-1/2 particle with three decay channels ( $\gamma_x, \gamma_y, \gamma_-$ ).
- By analyzing the characteristic polynomial, they derive the exhaustive set of Liouvillian EPs, identifying three distinct parameter regimes.
- The Newton polygon analysis reveals a slope of $-1/2$ , and the corresponding amoeba shows a tentacle with a slope of 2. This confirms the presence of second-order EPs, where eigenvalues scale as $\epsilon^{1/2}$ .
- Numerical diagonalization confirms the square-root scaling of eigenvalues and the characteristic cyclic exchange of eigenvalues when encircling the EP.
Dissipative Superconducting Qubit System:
- The authors study a model based on recent experiments involving a superconducting qutrit (states $|g\rangle, |e\rangle, |f\rangle$ ) with decay channels and a microwave drive.
- Anisotropy of EPs: The study highlights that the order of a Liouvillian EP is not an intrinsic fixed property but depends on the perturbation direction.
  - Case A (Perturbation in decay rate $\gamma_f$ ): The Newton polygon exhibits two segments with slopes $-1$ and $-1/3$ . The amoeba shows tentacles with slopes 1 and 3. This indicates that under this perturbation, the four coalescing eigenvalues split into two groups: one linear ( $\epsilon^1$ ) and three following a cube-root scaling ( $\epsilon^{1/3}$ ). This corresponds to a Jordan block structure of sizes $1 \times 1$ and $3 \times 3$ .
  - Case B (Perturbation in drive amplitude $J$ ): The Newton polygon shows a segment with slope $-1/2$ and a vertical segment. The amoeba shows a tentacle with slope 2 and a horizontal tentacle. This indicates that only two eigenvalues split with a square-root scaling ( $\epsilon^{1/2}$ ), while the other two remain invariant (valuation $\infty$ ).
- Numerical calculations of eigenvalue scaling and cyclic permutations verify these analytical predictions, demonstrating that the same Liouvillian point can exhibit different effective orders depending on the perturbation matrix.

Significance and Claims
The paper claims that the combination of Newton polygons and tropical geometry provides a direct, analytical tool for:

Identification and Characterization: Locating Liouvillian EPs and determining their order without solely relying on numerical brute-force methods.
Revealing Anisotropy: Explicitly capturing how the order and scaling behavior of an EP depend on the specific form of the perturbation (generic vs. non-generic).
Design Utility: The authors suggest that this approach can be used to understand and potentially design Liouvillian EPs of a desired order by selecting appropriate perturbations that alter the structure of the Newton polygon.

The work is presented as an analytical supplement to numerical studies, offering a geometric interpretation of the algebraic properties of Liouvillian operators. The authors maintain a modest scope, focusing on the characterization of existing models and suggesting that future work could extend these methods to non-linear and non-Markovian variants of EPs.

More like this