🔬 mesoscale physics

Preparation-protocol-dependent quantum Mpemba dynamics in a magnetically tunable graphene nanotorus qubit

This study demonstrates that the strength of the quantum Mpemba effect in a magnetically tunable graphene nanotorus qubit is critically dependent on the initial state preparation protocol, with bare Gibbs preparation enabling strong anomalous relaxation while driven steady-state preparation suppresses it due to differences in Liouvillian modal weights.

Original authors: J. Furtado

Published 2026-09-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: J. Furtado

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 quiet world of quantum physics, where particles behave more like waves than solid objects, scientists have long been fascinated by a peculiar paradox known as the Mpemba effect. Named after a Tanzanian student who noticed that hot water sometimes freezes faster than cold water, this counterintuitive phenomenon suggests that a system starting further away from a stable state can sometimes reach that state more quickly than one starting closer. In the quantum realm, this idea translates to a system that is "hotter" or more excited potentially relaxing to a calm, steady condition faster than a "cooler" system. For years, researchers have searched for the conditions that allow this to happen, focusing on the specific rules that govern how energy dissipates and how a system's initial setup influences its journey toward equilibrium. The question has shifted from simply observing the effect to understanding how to control it, asking whether the way we prepare a system before it begins its relaxation can determine the speed of its recovery.

A recent study by physicist J. Furtado explores this question using a unique and highly controllable quantum system: a tiny ring made of graphene, known as a nanotorus, which acts as a qubit, the fundamental unit of quantum information. Imagine this ring as a microscopic track where a single electron is trapped. By applying a magnetic field, researchers can shape the electron's path and define two specific energy states that serve as the "zero" and "one" of a quantum bit. To manipulate this bit, they use an oscillating electric field, much like a radio wave, to nudge the electron between these states. The researchers set up a scenario where they prepare two versions of this system: one that is initially "hot" and one that is "cold." They then suddenly change the environment, cooling both systems down to a much lower temperature while simultaneously turning on the electric drive, and watched to see which one reached the new stable state first.

The study compared two distinct ways of preparing these initial hot and cold states. In the first method, called bare Gibbs preparation, the system is allowed to settle into a thermal state based on its natural energy levels without any external electric drive. Only after the system is ready does the researchers switch on the electric field and cool the environment simultaneously. In the second method, known as driven steady-state preparation, the electric field is already active while the system is being heated or cooled. The system settles into a state that is already balanced with the drive present before the final cooling step begins. The researchers then subjected both setups to the exact same final conditions: the same magnetic field, the same electric drive, and the same cooling bath.

The results revealed a dramatic difference depending entirely on which preparation method was used. When the researchers used the bare Gibbs method, the hotter system consistently raced past the cooler one, reaching the final stable state much faster. In many of the simulations, the effect was so strong that the integrated measure of this speed advantage approached its maximum possible value, indicating a robust and powerful quantum Mpemba effect. The crossing point, where the hot system overtook the cold one, happened very early in the process, often within a fraction of a microsecond. However, when they switched to the driven steady-state preparation, the story changed completely. Under these conditions, the Mpemba effect almost vanished. The two systems relaxed at nearly the same pace, and the hot system rarely overtook the cold one. When it did, the event was delayed and weak, occurring much later in the timeline.

To understand why this happened, the researchers looked inside the mathematical machinery driving the system's behavior. They found that the final rules governing how the system relaxes were identical in both cases; the difference lay entirely in the starting point. The way the system was prepared determined how much of its initial energy was stored in the "slow" modes of relaxation versus the "fast" ones. In the successful bare Gibbs scenario, the hot system happened to have very little energy trapped in the slowest, most sluggish part of the system, allowing it to shed its excess energy quickly. In contrast, the driven steady-state preparation loaded the hot system with a heavy weight in that slow sector, effectively bogging it down and preventing it from gaining the speed advantage. The study demonstrates that the quantum Mpemba effect is not just a fixed property of the materials or the environment, but a dynamic outcome that can be turned on or off simply by changing how the system is prepared before the experiment begins.

This finding is significant because it identifies the preparation protocol itself as a control knob for quantum relaxation. It suggests that in future quantum technologies, where managing heat and decoherence is critical, the way a qubit is initialized could be just as important as the cooling mechanism used to stabilize it. While the current results are based on detailed computer simulations using a model of the graphene ring, they provide a clear roadmap for what experimentalists should look for. If a real-world graphene nanotorus behaves as predicted, engineers could potentially design systems that cool down or reset their quantum states with unprecedented speed simply by choosing the right initial setup. The work underscores a broader lesson in quantum physics: the history of how a system gets to a starting point matters, and that history can dictate the speed of its future.

=== SUMMARY ===

Technical Summary: Preparation-protocol-dependent quantum Mpemba dynamics in a magnetically tunable graphene nanotorus qubit

Problem Statement
The quantum Mpemba effect (QMpE) describes a counterintuitive phenomenon where a system initially farther from equilibrium relaxes to a stationary state faster than a system starting closer to equilibrium. While the existence and strength of the QMpE depend on both the decay spectrum of the open-system Liouvillian and the projection of the initial state onto decay modes, previous studies have largely focused on optimizing arbitrary initial states or specific many-body models. This paper addresses a distinct question within the context of a realistic physical platform: how strongly does the QMpE depend on the specific preparation protocol used to generate the "hot" and "cold" initial states, assuming the final dynamical evolution (Hamiltonian and bath) is identical?

Methodology
The author investigates this problem using a graphene nanotorus qubit, a platform where an electron is confined to the inner region of a nanotorus, forming curvature-induced bound states. The qubit is defined by two low-lying states ( $|l=0, m=0\rangle$ and $|l=1, m=0\rangle$ ) tunable via a static magnetic field and controlled by a local oscillating electric field.

The study employs a thermal Lindblad master equation to model the open-system dynamics, incorporating thermal bosonic rates for relaxation and dephasing. The core of the methodology involves comparing two distinct preparation protocols that share the same final Hamiltonian, thermal bath, and Markovian Liouvillian:

Bare Gibbs Preparation: The hot and cold states are prepared as Gibbs states of the undriven qubit Hamiltonian. At $t=0$ , the system undergoes a thermal quench to a target temperature $T_0$ , and the coherent electric drive is simultaneously switched on.
Driven Steady-State Preparation: The coherent drive is present during the preparation phase. The hot and cold states are the stationary states of the driven Liouvillian at their respective initial temperatures. At $t=0$ , only the bath temperature is quenched to $T_0$ , while the drive remains active.

The author quantifies the QMpE using the Hilbert-Schmidt distance ( $B(t)$ ) between the evolving state and the final stationary state. They define an integrated Mpemba parameter, $M_B$ , which measures the weight of the time interval where the initially hotter state is closer to the target than the initially colder state. A comprehensive numerical analysis is performed across various parameter spaces (magnetic field $B$ , electric field amplitude $E_0$ , radii $R, r$ , and temperatures $T_h, T_c, T_0$ ).

Key Contributions and Results
The primary finding is a stark, protocol-dependent dichotomy in the relaxation dynamics:

Bare Gibbs Protocol: This protocol yields broad regions of strong anomalous relaxation. In the reference parameter set, the integrated parameter reaches $M_B \approx 0.661$ , with the hot state crossing the cold trajectory at $\gamma_0 t \approx 1.13$ . In extensive two-dimensional scans (e.g., varying $B$ and $E_0$ ), $M_B$ frequently approaches unity, indicating a very strong Mpemba effect.
Driven Steady-State Protocol: Under the exact same final Liouvillian, this protocol almost completely suppresses the effect. The integrated parameter remains close to zero ( $M_B \approx 2.5 \times 10^{-3}$ ) over most of the parameter space, with any crossings occurring significantly later ( $\gamma_0 t \approx 11.1$ ) and being weak.

Mechanism Analysis
To explain this contrast, the author performs a Liouvillian-mode decomposition. Since the final spectrum is identical for both protocols at corresponding parameter points, the difference must arise from the initial state projections onto the decay modes.

The author defines a ratio $R_s$ , representing the relative weight of the hot versus cold initial states in the "slow active sector" of the Liouvillian.
Bare Gibbs: Typically yields $R_s < 1$ , meaning the initially hot state has a smaller projection onto the slowest-decaying modes than the cold state. This suppression of the slow component in the hot state allows it to relax faster.
Driven Steady-State: Predominantly yields $R_s > 1$ , where the hot state carries a larger weight in the slow sector, hindering rapid relaxation.
The study confirms that the spectral spread (the hierarchy of decay rates) remains largely unchanged between protocols; thus, the control mechanism is the preparation-dependent population of decay modes, not a restructuring of the Liouvillian eigenvalues.

Significance and Claims
The paper claims that state preparation acts as a critical control parameter for anomalous quantum relaxation in curved graphene qubits. The results demonstrate that the QMpE is not merely a property of the system's Hamiltonian or bath but is highly sensitive to the history of how the initial state was generated.

The author emphasizes that the "strong bare-Gibbs response" arises because the quench projects the system into a newly driven Hamiltonian, coupling population and coherence sectors in a way that creates favorable modal weights ( $R_s < 1$ ). Conversely, the driven steady-state protocol pre-optimizes the state for the driven dynamics, inadvertently populating the slow modes.

The paper concludes modestly, noting that while the results identify a robust mechanism for controlling the QMpE, the quantitative predictions rely on phenomenological thermal rates. The author states that the decisive next step is a microscopic treatment of electron-phonon coupling specific to the nanotorus geometry to determine if these strong effects survive when the bath itself acquires geometry-dependent characteristics. They do not claim immediate experimental realization but rather establish a theoretical framework for understanding preparation-dependent relaxation in this specific qubit architecture.

Technical Summary: Preparation-protocol-dependent quantum Mpemba dynamics in a magnetically tunable graphene nanotorus qubit

More like this