Imaginary gauge potentials in a non-Hermitian spin-orbit coupled quantum gas

This paper experimentally realizes a continuum analog of the Hatano-Nelson model in a spin-orbit coupled Bose-Einstein condensate by introducing tunable spin-dependent loss, demonstrating collective nonreciprocal transport and self-acceleration while revealing how strong interactions suppress topological edge states in favor of localized excited states.

The Gist

The Big Idea: A Quantum World with "One-Way" Wind

Imagine you are in a giant, empty room (a quantum gas) filled with thousands of tiny, invisible balls (atoms). Usually, if you push these balls, they bounce around randomly, and if you push them left, they go left; push them right, they go right. Physics is usually fair and symmetrical.

But in this experiment, the scientists from the Joint Quantum Institute and Harvard did something magical: they created a "wind" that only blows in one direction, but you can't feel it with your hands.

They call this an "Imaginary Gauge Potential." It sounds like a sci-fi term, but think of it like a one-way street for quantum particles. In this street, particles don't just move; they get a mysterious "self-acceleration" that pushes them faster and faster in one direction, even though no one is physically pushing them.

How Did They Do It? (The Recipe)

To create this strange world, they used a Bose-Einstein Condensate (BEC). Think of a BEC not as a gas, but as a super-cool, super-organized dance troupe. All the atoms are holding hands and moving in perfect unison.

1. The Spin-Orbit Coupling (The Dance Move): First, they used lasers to make the atoms "spin" and move in a specific way. It's like teaching the dance troupe a complex routine where their spin determines their speed.
2. The "Leaky" Subspace (The Exit Door): This is the secret sauce. They added a special "loss" mechanism. Imagine that if a dancer spins a certain way (let's call it "Spin Up"), there is a tiny, invisible trapdoor in the floor that lets them fall out of the room.
3. The Result: Because the "Spin Up" dancers are constantly falling out of the room, the remaining dancers (the "Spin Down" ones) start to behave strangely. The system effectively creates a mathematical wind that pushes the remaining dancers to one side.

The Two Main Discoveries

The paper describes two cool things that happened when they turned on this "wind."

1. The "Self-Accelerating" Crowd

Usually, if you have a crowd of people in a box, they just sit there or bounce around. But here, the crowd started to scoot across the room on its own.

• The Analogy: Imagine a crowd of people in a hallway. Suddenly, the people on the left start to vanish (fall through the floor). The people on the right, seeing their neighbors disappear, start to drift left to fill the gap. But because of the quantum rules, this drift turns into a run. The whole crowd starts accelerating to the left, getting faster and faster.
• The Twist: The scientists found that the bigger the crowd (the more atoms), the slower this acceleration was. It's like a heavy truck is harder to push than a bicycle. The atoms were pushing back against each other (interacting), which slowed down the "wind" effect.

2. The Missing "Skin" Effect

In the world of non-interacting particles (like single, lonely dancers), this "wind" usually causes a phenomenon called the Non-Hermitian Skin Effect.

• The Analogy: Imagine a crowd of people in a room with a one-way wind. In a normal scenario, everyone would get blown into a tight pile against the left wall. This is the "skin effect"—the crowd clumps up at the edge.
• What Happened Here: Because the atoms in this experiment were holding hands (interacting strongly), they refused to clump up at the wall. Instead of piling up, they stayed spread out, but they kept trying to accelerate. The "clumping" was suppressed by their social nature (repulsion). Instead of a static pile at the edge, they formed a dynamic, moving state that looked like a shockwave rippling through the gas.

Why Does This Matter?

1. It's Not Just Math, It's Real:
For a long time, "Imaginary Numbers" in physics were just tools for calculation. This experiment proves that you can build a real physical system where these "imaginary" forces act like real wind, pushing atoms around.

2. It Works Forever (Sort of):
Usually, when particles fall out of a system (like our dancers falling through the trapdoor), the math breaks down. But the scientists showed that because the atoms that "fall out" are kicked out of the room so hard they never come back, the system stays stable. It's like a leaky boat that keeps sinking, but the water is removed so fast that the boat never actually fills up. This allows them to study these weird effects for a long time (milliseconds, which is an eternity in quantum physics).

3. Future Tech:
Understanding how to control these "one-way winds" could help us build new types of quantum computers or sensors that are immune to noise, or even simulate how particles move in curved space (like near a black hole) right here in a lab.

Summary

The scientists built a quantum dance floor where they could make the floor "leak" in a specific way. This created an invisible, one-way wind that made the entire crowd of atoms accelerate on its own. While they expected the crowd to pile up against the wall, the atoms' strong friendship (interactions) kept them spread out, creating a new, unique state of matter. It's a step toward mastering the strange, "non-Hermitian" side of the quantum universe.

Technical Summary

1. Problem Statement and Motivation

The paper addresses the experimental realization and characterization of non-Hermitian quantum systems, specifically focusing on the Hatano-Nelson (HN) model.

• Theoretical Context: The HN model describes a lattice with a constant imaginary Peierls phase, leading to the Non-Hermitian Skin Effect (NHSE), where eigenstates exponentially localize at boundaries. While observed in non-interacting systems (photonic, electrical, cold atoms), the behavior of interacting quantum gases under imaginary gauge potentials remains largely unexplored.
• Key Challenge: In open quantum systems, non-Hermitian Hamiltonians are often approximations valid only until a "quantum jump" (spontaneous emission) occurs. A major theoretical hurdle is determining if a non-Hermitian description remains valid over long timescales in the presence of interactions and continuous dissipation.
• Goal: To experimentally realize a continuum analog of the HN model in a Bose-Einstein Condensate (BEC) with tunable spin-dependent loss, investigate the interplay between interactions and non-Hermiticity, and validate the non-Hermitian description against a full master-equation treatment.

2. Methodology

The researchers utilized a homogeneous, spin-orbit coupled (SOC) $^{87}\text{Rb}$ Bose-Einstein Condensate confined in a box-shaped optical trap.

A. System Implementation

• Spin-Orbit Coupling (SOC): Two internal states ($|\downarrow\rangle$ and $|\uparrow\rangle$) were coupled via Raman lasers, inducing a momentum-dependent coupling with recoil momentum $p_0$ and Rabi frequency $\Omega$.
• Imaginary Gauge Potential ($B$): To introduce non-Hermiticity, the $|\uparrow\rangle$ state was coupled via microwaves to an intermediate state, which was then driven by a resonant laser to a lossy excited state. This created a spin-dependent loss rate $\gamma$.
• Effective Hamiltonian: In the adiabatic limit (large $\Omega$), the system is projected onto the lowest SOC band, yielding an effective non-Hermitian Hamiltonian:
  $$\hat{H}_{\text{nh}} = \frac{(\hat{p} - iB)^2}{2m^*} + V(\hat{x}) - i\frac{\hbar\gamma}{2}$$
  where the imaginary gauge potential is $B = -p_0 \frac{\gamma}{\Omega - 4\omega_0}$.

B. Experimental Protocol

1. Preparation: A BEC was prepared in the SOC ground state ($p=0$).
2. Dynamics: The microwave and loss-inducing laser fields were abruptly switched on, allowing the system to evolve for up to 9 ms.
3. Measurement: High-resolution absorption imaging (partial transfer) was used to measure the spatial density distributions of both spin components and the total density.
4. Validation: A simplified configuration using RF coupling was employed to compare the non-Hermitian predictions against a multi-level quantum master equation (QME) to verify the validity of the approximation beyond the quantum jump timescale.

3. Key Contributions and Theoretical Insights

• Heisenberg Equations of Motion (EoM): The authors derived that for non-Hermitian systems with imaginary gauge fields, the EoMs for position and momentum are not closed. They explicitly depend on the system's phase-space distribution (e.g., momentum variance $\langle \delta \hat{p}^2 \rangle$).
  • This leads to "self-acceleration": The center of mass accelerates even in the absence of external forces, driven by the interplay between the imaginary gauge field and the momentum spread.
  • It also leads to "self-displacement": An anomalous term arising from correlations between position and momentum fluctuations.
• Interaction Effects: The study highlights a unique role of interactions (described by the Gross-Pitaevskii Equation, GPE). Interactions significantly enhance the self-acceleration compared to non-interacting predictions but simultaneously suppress the formation of topological edge states (NHSE).

4. Key Results

A. Collective Nonreciprocal Transport

• Observation: The BEC exhibited a clear displacement in real space. The center of mass (CoM) accelerated quadratically with time ($x \propto t^2$).
• Dependence: The acceleration was proportional to the imaginary gauge potential $B$. Reversing the sign of $B$ (by coupling the loss to the $|\downarrow\rangle$ state instead of $|\uparrow\rangle$) reversed the direction of transport.
• Role of Interactions: Non-interacting simulations failed to match the data. Only simulations including mean-field interactions (GPE) reproduced the experimental acceleration. The interactions broaden the momentum-space wavefunction and extend it into regions of rapid potential variation, amplifying the self-acceleration effect.

B. Suppression of the Non-Hermitian Skin Effect (NHSE)

• Expectation vs. Reality: In non-interacting systems, the HN model predicts exponential localization at the boundary (NHSE).
• Result: In the interacting BEC, the strong repulsive interactions prevented the formation of these localized edge states. Instead, the system formed a dynamical state where self-acceleration was counterbalanced by repulsive interactions, resulting in a "static" configuration that is actually a dynamic equilibrium. The CoM displacement was significantly reduced compared to the non-interacting case.

C. Validity of the Non-Hermitian Description

• Quantum Jump Timescale: Spontaneous emission typically limits the validity of non-Hermitian Hamiltonians to the time between quantum jumps (approx. 26 ns for the excited state).
• Experimental Proof: By operating in the Quantum Zeno regime (strong continuous measurement/loss), the researchers ensured that any atom undergoing a quantum jump was immediately ejected from the trap.
• Conclusion: Since no observed atoms remained in the system after a jump, the non-Hermitian description remained quantitatively accurate for the entire millisecond-scale evolution, validated against full master-equation simulations.

5. Significance and Outlook

• New Paradigm for Open Systems: This work demonstrates that non-Hermitian physics is not limited to single-particle or non-interacting regimes. It establishes a framework for studying interacting non-Hermitian many-body systems.
• Interplay of Topology and Interactions: The finding that interactions can suppress the NHSE while enhancing transport dynamics challenges the standard bulk-boundary correspondence in non-Hermitian systems and suggests new phases of matter (e.g., non-Hermitian Mott insulators).
• Experimental Benchmark: The successful comparison with multi-level master equations provides a rigorous benchmark for future theoretical work on non-Hermitian topological invariants in interacting systems.
• Future Directions: The platform opens avenues for exploring non-Hermitian analog gravity, asymmetric fermionization, and the impact of disorder on non-Hermitian topological phases in cold atoms.

In summary, this paper provides the first experimental realization of an imaginary gauge potential in an interacting quantum gas, revealing that interactions fundamentally alter non-Hermitian dynamics by enhancing transport while suppressing topological skin effects, all while validating the non-Hermitian formalism over long timescales.