⚛️ quantum physics

A Refined Biorthogonal Framework for Non-Hermitian Quantum Theory and Its Application in Dynamical Phase Transition

This paper proposes a refined biorthogonal framework for non-Hermitian quantum theory, grounded in the premise that both left- and right-vectors satisfy the Schrödinger equation, which resolves existing theoretical controversies and successfully generalizes the conditions for dynamical phase transitions in the Su-Schrieffer-Heeger model to the non-Hermitian regime.

Original authors: Fei Wang, Guoying Liang, Zecheng Zhao, Bao-Ming Xu

Published 2026-03-24

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Fei Wang, Guoying Liang, Zecheng Zhao, Bao-Ming Xu

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

Imagine you are trying to describe the motion of a ball. In the standard world of physics (Hermitian quantum mechanics), the ball is like a perfect marble rolling on a smooth, frictionless table. It bounces, it rolls, but it never loses energy, and its path is perfectly predictable. The rules are simple: what goes in must come out, and everything balances perfectly.

But in the real world, things are messier. Think of a ball rolling through a swamp. It loses energy to the mud (loss), or maybe it's being pushed by a hidden fan (gain). This is the world of Non-Hermitian Quantum Systems. Here, energy isn't conserved in the usual way; things can appear out of nowhere or vanish into thin air.

For a long time, physicists have been arguing about how to write the "rulebook" for these swampy, messy systems. The paper you shared, by Fei Wang and his team, proposes a new, cleaner rulebook that finally makes sense of the chaos.

Here is the breakdown of their discovery using simple analogies:

1. The Problem: The "Shadow" Mismatch

In standard physics, to describe a state, you just need one vector (let's call it the Right-Hander). It's like describing a person by looking at their face.

In non-Hermitian physics, things are trickier. You need two things to describe a state:

The Right-Hander: The state moving forward in time.
The Left-Hander: A "shadow" or "mirror image" state that helps you calculate probabilities.

The Old Rulebook's Flaw:
Previously, scientists treated the Right-Hander as the "real" actor following the rules of the game (the Schrödinger equation). They treated the Left-Hander as a passive spectator, just a mathematical tool to help with calculations.

The Analogy: Imagine a dance duo. The old rulebook said, "The male dancer (Right) follows the choreography perfectly. The female dancer (Left) just stands there and holds a sign that says 'I'm here,' but she doesn't actually dance to the music."
The Issue: This created a weird disconnect. If the music changes, the male dancer reacts, but the female dancer's "shadow" didn't react the same way. The theory was inconsistent.

2. The Solution: The "Perfect Dance Duo"

The authors propose a Refined Biorthogonal Framework. Their big idea is simple but profound: Both dancers must follow the music.

The New Rule: Both the Right-Hander and the Left-Hander must obey the Schrödinger equation. They are a unified team. You can't have one dancing to the beat while the other stands still.
The Result: This creates a consistent, self-contained theory. It fixes the "asymmetry" of the old method. When you turn off the "swamp" (remove the gain and loss), this new rulebook naturally turns back into the standard, perfect marble physics we already know. It's a universal translator that works for both simple and messy worlds.

3. The Application: The "Tipping Point" (Phase Transitions)

To prove their new rulebook works, the authors tested it on a specific model called the SSH Model (think of it as a chain of atoms connected by springs, like a necklace). They simulated a "quantum quench"—suddenly snapping the springs from one tension to another.

In physics, when you snap a system suddenly, it sometimes undergoes a Dynamical Phase Transition (DQPT). This is like a sudden "snap" in the system's behavior, similar to water suddenly turning to ice, but happening in time rather than temperature.

What they found:

The Old Way: In simple systems, you could predict this "snap" by checking if two arrows (vectors) were perpendicular (at a 90-degree angle).
The New Way: In the messy, non-Hermitian world, the old "90-degree" rule breaks down. The authors found a new, more complex rule: The "Real Part" of the angle between the arrows must be zero.
- Analogy: Imagine trying to balance a broom on your finger. In a calm room, you just need to keep it straight up. In a windy room (non-Hermitian), you have to adjust for the wind. The new rule tells you exactly how to adjust so the broom doesn't fall, even when the wind is blowing in weird directions.

4. The Surprise: New Types of "Snaps"

The most exciting part of their discovery is that they found entirely new types of phase transitions that the old rules couldn't see.

The Winding Number: In the past, physicists used a "winding number" (like counting how many times a rubber band twists around a finger) to predict these transitions.
The Discovery: The authors found transitions where the rubber band doesn't twist in a way that can be counted, yet the system still undergoes a dramatic change. It's like finding a new type of knot that doesn't look like a knot at all, but still holds the rope together (or breaks it).

Summary

This paper is like fixing the instruction manual for a complex video game.

The Bug: The old manual told you how to move the main character but ignored the shadow, leading to glitches.
The Patch: The new manual says, "The character and the shadow must move together."
The Feature: With this new manual, they discovered hidden levels and secret endings (new phase transitions) that players didn't even know existed.

By making the theory consistent, the authors have given scientists a reliable tool to explore the strange, gain-and-loss-filled world of non-Hermitian physics, which is crucial for developing future technologies like better lasers, sensors, and quantum computers.

1. Problem Statement

Non-Hermitian quantum mechanics (NHQM) is essential for describing open systems with gain and loss, yet the theoretical framework for describing their states and dynamics remains a subject of debate.

The Core Issue: The standard biorthogonal framework (introduced by Brody et al.) relies on both left ( $\langle \tilde{\psi}|$ ) and right ( $|\psi\rangle$ ) eigenvectors. However, a fundamental inconsistency exists: while the right-vectors evolve according to the Schrödinger equation ( $\partial_t |\psi\rangle = -iH|\psi\rangle$ ), the associated left-vectors (defined via complex conjugation of coefficients) do not satisfy the corresponding Schrödinger equation ( $\partial_t \langle \tilde{\psi}| \neq \langle \tilde{\psi}|iH$ ).
Consequence: This asymmetry implies that the standard biorthogonal formulation is incomplete at the dynamical level. It treats the left-vector merely as an auxiliary mathematical tool for spectral decomposition rather than a physically evolving entity, leading to conceptual ambiguities in defining states, expectation values, and time evolution.

2. Methodology

The authors propose a Refined Biorthogonal Framework based on a foundational physical premise: Both the left- and right-vectors of a non-Hermitian system must independently satisfy the Schrödinger equation.

Reformulation of States:
- Instead of defining a state solely by a right-vector $|\psi\rangle$ with an associated conjugate left-vector, the authors define a pure state as a density matrix $\rho = |\psi\rangle\langle\tilde{\psi}|$ .
- Crucially, $|\psi\rangle$ and $\langle\tilde{\psi}|$ are treated as an inseparable entity. They cannot be chosen independently; they must be determined simultaneously (e.g., by diagonalizing a physical operator or applying a transformation to an eigenstate).
- The coefficients in the expansion of the left-vector ( $\tilde{c}_n$ ) are distinct from the complex conjugates of the right-vector coefficients ( $c_n^*$ ), i.e., $\tilde{c}_n \neq c_n^*$ .
Dynamics:
- The time evolution is governed by:
  $\frac{\partial}{\partial t}|\psi(t)\rangle = -iH|\psi(t)\rangle$
  $\frac{\partial}{\partial t}\langle\tilde{\psi}(t)| = \langle\tilde{\psi}(t)|iH$
- This leads to the density matrix evolution: $\rho(t) = e^{-iHt}\rho(0)e^{iHt}$ .
- This formulation allows for complex probabilities (diagonal elements of $\rho$ can be negative or complex), a feature intrinsic to NHQM, while maintaining the normalization of the trace ( $\text{Tr}[\rho]=1$ ).
Measurement and Geometry:
- Expectation values are defined as $\langle A \rangle = \text{Tr}[A\rho]$ .
- The Loschmidt echo (return probability) is redefined as $L(t) = \langle\tilde{\psi}(0)|\psi(t)\rangle\langle\tilde{\psi}(t)|\psi(0)\rangle$ .
- Geometric phases (Pancharatnam) and dynamical phases are derived consistently using the biorthogonal inner product.

3. Key Contributions

Resolution of Dynamical Asymmetry: The paper resolves the long-standing inconsistency in the biorthogonal framework by enforcing the Schrödinger equation on both left and right vectors, establishing a self-consistent dynamical theory.
Generalization of DQPT Conditions: The authors derive a unified condition for Dynamical Quantum Phase Transitions (DQPT) in non-Hermitian systems.
- For Hermitian two-band systems, the condition is $d_i^k \cdot d_f^k = 0$ (orthogonality of initial and final Hamiltonian vectors).
- The refined framework generalizes this to the non-Hermitian case as:
  $\text{Re}\left[ \frac{d_i^k}{|d_i^k|} \cdot \frac{d_f^k}{|d_f^k|} \right] = 0$
  where $d^k$ represents the vector of coefficients in the Pauli matrix expansion of the Hamiltonian.
Discovery of New DQPTs: The framework reveals DQPTs that cannot be characterized by the winding number, a topological invariant traditionally used to detect such transitions.

4. Results: Application to the SSH Model

The theory was applied to a one-dimensional Su-Schrieffer-Heeger (SSH) model with gain and loss ($PT$-symmetric).

Phase Diagram: The model exhibits three phases: $PT$-symmetric (real spectrum), $PT$-broken Phase I (purely imaginary spectrum), and $PT$-broken Phase II (complex spectrum).
Quench Dynamics: The authors simulated quantum quenches (sudden parameter changes) and analyzed the Loschmidt echo rate function $r(t)$ $r (t)$ and Fisher zeros.
- Case A ($PT$-symmetric to $PT$-symmetric): DQPTs occur when quenching between specific sub-regions of the symmetric phase. These transitions are accompanied by integer jumps in the winding number, consistent with Hermitian intuition.
- Case B (Boundary/Exceptional Points): Quenches from critical points (e.g., $q=1$ ) to the symmetric phase result in DQPTs characterized by half-integer jumps in the winding number.
- Case C ($PT$-broken to $PT$-symmetric):
  - When quenching from the $PT$-antisymmetric phase, all modes satisfy the DQPT condition. The critical times are non-periodic, and the winding number becomes ill-defined (staying zero), failing to detect the transition.
  - When quenching from the $PT$-broken phase, modes with purely imaginary energies trigger non-periodic DQPTs (winding number fails), while modes with real energies trigger periodic DQPTs (winding number jumps).
Key Finding: The refined framework successfully identifies DQPTs in regimes where the winding number is zero or undefined, proving that the topological order parameter is insufficient for characterizing all non-Hermitian dynamical transitions.

5. Significance

Theoretical Unification: This work provides a rigorous, consistent foundation for non-Hermitian quantum dynamics, bridging the gap between spectral properties and time evolution. It clarifies how to properly define states and observables in open systems.
New Physical Phenomena: By identifying DQPTs that evade detection by standard topological invariants (winding numbers), the paper suggests that non-Hermitian systems possess a richer landscape of dynamical phenomena than previously understood.
Experimental Relevance: The framework offers a consistent theoretical tool for interpreting experiments in photonic, atomic, and electronic systems where gain and loss are engineered, particularly in the study of exceptional points and non-Hermitian skin effects.
Broader Impact: The ability to handle complex probabilities and non-unitary evolution consistently opens new avenues for studying non-equilibrium dynamics in open quantum systems.