A study in Physical Review Research has identified when amplification in certain quantum-like systems can be predicted from only the starting point and ending point of a slow change. The work, led by Tomoki Ozawa at the Advanced Institute for Materials Research at Tohoku University and Henning Schomerus of Lancaster University, reveals a geometric rule inside non-Hermitian systems.
The finding focuses on systems that exchange energy with their surroundings. These include platforms related to optics, classical mechanics and metamaterial design. In such settings, waves can grow or fade as system parameters change. Ozawa and Schomerus show that geometry can govern part of that growth in a surprisingly simple way.
At the heart of the result is a question with practical stakes. If a signal becomes stronger while a system is slowly tuned, do researchers need to know every detail of the route taken through parameter space? In special cases, the answer becomes much simpler. The amplification can be determined by a ratio between geometric quantities at the two endpoints.
Geometry sets the gain
Quantum geometry describes how the state of a system changes as its conditions are adjusted. A famous example is the Berry phase, a geometric effect that appears when a quantum state follows a slow path through a landscape of possible settings. This concept has helped physicists understand phenomena ranging from electrical conductivity to superconductivity.
Ozawa and Schomerus worked in a broader setting called non-Hermitian physics. In this framework, a system can effectively gain or lose energy. That makes it useful for describing real wave systems where light, sound, or mechanical motion may leak away or be amplified.
In non-Hermitian quantum mechanics, the Berry phase can contain an imaginary part. That feature matters because it can change the intensity of a wave. A geometric phase can then affect whether the final signal grows or shrinks during a slow process.
“We wanted to identify geometric phenomena that are truly intrinsic to non-Hermitian quantum mechanics,” Ozawa said. The study answers that goal by connecting geometric amplification to a static property of the system’s eigenstates.
When the path stops mattering
The study examines adiabatic amplification, which occurs when system parameters change slowly enough for the system to follow its evolving state. In ordinary terms, imagine gently tuning a complex wave system while tracking how the signal intensity changes. The question is whether the route through the tuning landscape controls the final strength.
Ozawa and Schomerus found that a geometric amplification factor can become route-independent when the imaginary part of the Berry curvature is zero. Under that condition, the gain depends only on the initial and final points in parameter space.
This kind of path independence is valuable because it turns a potentially complicated history into a boundary calculation. Researchers can focus on the state at the beginning and the state at the end. The entire path connecting those points becomes secondary for this specific geometric contribution.
The result is theoretical and the authors supported it with model calculations. As the paper’s abstract states, “We validate our theory using a couple of concrete examples of physical relevance.” Those examples show how the mathematical rule behaves in systems that resemble experimentally relevant wave platforms.
The Petermann factor connection
A key part of the discovery is the Petermann factor. This quantity measures how much the eigenstates of a non-Hermitian system fail to behave like the clean, perpendicular states familiar from ideal closed quantum systems. In practical wave physics, that lack of orthogonality can strongly influence noise, sensitivity and amplification.
Before this study, the Petermann factor was mainly viewed as a static geometric property. Ozawa and Schomerus connected it to a dynamic effect. They showed that, in some symmetry classes, the geometric contribution to amplification can be written using only the Petermann factors at the initial and final points.
That connection gives the result its striking simplicity. A wave can be amplified along a slow process and the geometric part of that amplification can be tied to a ratio between two endpoint quantities. The study links a property measured at fixed settings with behavior that unfolds during a slow change.
For general readers, the message is that geometry can act like hidden accounting. The system keeps track of how its states are shaped and that shape can determine the signal’s final intensity. The Petermann factor becomes one of the numbers that records this hidden structure.
Symmetry provides the shortcut
The route-independent behavior appears when suitable symmetries constrain the non-Hermitian Hamiltonian. A Hamiltonian is the mathematical object physicists use to describe a system’s energy structure and time evolution. In the non-Hermitian case, it can also encode gain and loss.
One important symmetry is reciprocity. In a reciprocal system, signals propagate symmetrically in opposite directions. Ozawa explained that when such symmetries are present, “the amplification becomes path-independent and depends solely on the ratio of the Petermann factors at the start and end points.”
This provides a shortcut through a complicated calculation. Instead of integrating the geometric contribution along every step of a slow path, researchers can evaluate endpoint properties. The shortcut works only in the symmetry classes identified by the theory.
The team confirmed the prediction through numerical simulations of two physically realistic models. Those simulations matter because they show how the formal result behaves in concrete examples. They also help connect the abstract geometry to systems that may resemble laboratory platforms.
A possible way to measure a hidden quantity
The Petermann factor is important, yet it can be difficult to measure directly. The new work suggests a different route. If the geometric amplification is governed by endpoint Petermann factors, then observing how the wave norm changes during a slow process may reveal the quantity indirectly.
This idea could be especially useful in optical, mechanical, or metamaterial systems where non-Hermitian effects are engineered. Researchers often have strong control over parameters in these platforms. They may be able to design slow changes and measure the resulting signal intensity with precision.
The proposed measurement route remains grounded in theory. The study lays out the framework and tests it in simulations. Future experiments would need to implement the right symmetry conditions and separate the geometric contribution from other sources of amplification or decay.
Even with that caution, the implication is clear. A dynamical measurement could expose a geometric property that usually hides inside the structure of the system’s eigenstates. That makes amplification a possible diagnostic tool, rather than only an effect to manage.
What comes next for non-Hermitian physics
The collaboration behind the work came together quickly. Schomerus arrived at AIMR through the GI3 program and the project developed through frequent discussions with Ozawa. “This project was unique in how it came together,” Ozawa said.
According to Ozawa, the main results emerged within about two months. Schomerus arrived in July 2024, left in August and the paper was submitted in September 2024. The pace underscores how close collaboration can sharpen a theoretical question.
The next steps include extending the framework to more complex parameter spaces. The researchers also aim to explore non-adiabatic processes, where systems change too quickly for the slow-following assumption to fully apply. That direction may connect geometric amplification to non-Hermitian topological phase transitions.
Such phase transitions are of growing interest because non-Hermitian systems can behave in ways that closed quantum systems rarely do. Gain, loss and state geometry can combine to produce unusual sensitivity and wave behavior. The new study gives physicists another tool for sorting out which effects are controlled by geometry.
For now, the result offers a clean principle. In the right non-Hermitian systems, the geometric part of amplification can be read from two points. That turns a winding path through parameter space into a simpler comparison between where the system begins and where it ends.
