Imaginary-time Mpemba effect in quantum many-body systems

This paper reports the discovery of the imaginary-time Mpemba effect (ITME) in quantum many-body systems, where higher-energy initial states relax faster than lower-energy ones, a phenomenon demonstrated via quantum Monte Carlo simulations that offers a potential pathway to accelerate ground-state computations, particularly for models with sign problems.

The Gist

Imagine you have two cups of coffee: one is scalding hot, and the other is just warm. Common sense tells you that the warm cup will cool down to room temperature faster than the hot one, right?

Well, in the strange world of quantum physics, sometimes the hotter cup actually cools down faster. This counter-intuitive phenomenon is called the Mpemba effect. It's a bit like a sprinter who starts the race running backward at full speed, somehow crossing the finish line before the runner who started slowly.

For decades, scientists have studied this in the real world (like water freezing) and in real-time quantum systems. But in this new paper, the authors discovered a brand new version of this effect that happens in a "time machine" used by computers to solve complex physics problems. They call it the Imaginary-Time Mpemba Effect (ITME).

Here is a simple breakdown of what they found, using everyday analogies:

1. The "Time Machine" (Imaginary Time)

To understand this, you first need to understand what "Imaginary Time" is. It's not time travel in the Back to the Future sense. Instead, think of it as a mathematical filter or a sieve.

• Real Time: If you watch a quantum system evolve in real time, it's chaotic. It's like a crowded dance floor where everyone is spinning and bumping into each other.
• Imaginary Time: This is a special mathematical trick used by supercomputers. When you run a simulation in "imaginary time," it acts like a slow-motion filter. It gradually washes away the "noise" (the high-energy, chaotic parts of the system) and leaves only the most stable, calm state (the "ground state").

Scientists use this to find the "perfect" lowest-energy state of a material, which is crucial for designing new batteries, superconductors, or understanding how matter behaves.

2. The Race to the Bottom

Usually, when you use this "filter," you start with a guess.

• The Standard Guess: You pick a starting state that is already pretty close to the answer (low energy). You expect it to reach the bottom quickly because it's already close.
• The Surprising Discovery: The authors found that sometimes, if you start with a wild, messy, high-energy guess (the "hot" state), it actually reaches the bottom faster than the "calm, low-energy" guess.

The Analogy:
Imagine you are trying to find the lowest point in a foggy valley.

• Scenario A (Low Energy Start): You start on a gentle hill. You think, "I'm close to the bottom, I'll just walk down." But the path is full of small bumps and detours (low-energy excitations) that slow you down.
• Scenario B (High Energy Start): You start on a steep, rocky cliff (high energy). You slide down so fast that you shoot right past all the little bumps and land directly in the deepest part of the valley.

In the quantum world, the "hot" state sometimes has a secret shortcut that the "cold" state doesn't have.

3. Why Does This Happen? (The "Ghost" Particles)

The paper explains that this happens because of low-energy excitations. Think of these as "ghosts" or "echoes" of the system's structure.

• In some materials, there are these "ghosts" (like ripples in a pond) that are very easy to create.
• If your starting state is "cold" (low energy), it gets stuck interacting with these ghosts, which slows it down.
• If your starting state is "hot" (high energy), it might be structured in a way that ignores these ghosts entirely, allowing it to slide straight to the solution.

The authors found this happens in many different types of quantum materials, especially near Quantum Critical Points—these are like the "tipping points" where a material is about to change its nature (like water turning to ice, but for quantum magnets).

4. Why Should We Care? (The Supercomputer Boost)

This isn't just a cool physics trick; it's a game-changer for computing.

Currently, simulating quantum materials on computers is incredibly hard and slow. It's like trying to solve a Rubik's cube while wearing oven mitts. The "Imaginary Time" method is the best tool we have, but it's often slow because the computer has to wait for the "cold" starting guesses to settle down.

The New Superpower:
Now that we know the "hot" starting guesses can sometimes win the race, scientists can optimize their simulations. Instead of picking the "safest" low-energy guess, they can pick a "risky" high-energy guess that might zoom to the answer much faster.

• The Benefit: This could make quantum simulations exponentially faster.
• The Sign Problem: There's a notorious bug in quantum computing called the "Sign Problem" that makes some calculations impossible. The authors suggest that by choosing the right "hot" starting state, we might be able to bypass this bug and solve problems that were previously unsolvable.

Summary

The paper reveals a new rule of the quantum universe: Sometimes, starting with a messier, higher-energy state gets you to the solution faster than starting with a clean, low-energy one.

It's like realizing that to get to the bottom of a maze, sometimes it's better to run wildly through the walls (high energy) than to carefully follow the winding path (low energy). This discovery gives computer scientists a new, powerful strategy to simulate the quantum world, potentially speeding up the discovery of new materials and technologies.

Technical Summary

Here is a detailed technical summary of the paper "Imaginary-time Mpemba effect in quantum many-body systems."

1. Problem Statement

The Mpemba effect is a counter-intuitive non-equilibrium phenomenon where a hotter system freezes (relaxes to equilibrium) faster than a colder one. While extensively studied in classical thermodynamics and recently in real-time quantum dynamics, its existence in imaginary-time dynamics remained unexplored.

Imaginary-time evolution ($\tau = it$) is a cornerstone tool in quantum many-body physics, primarily used to numerically access ground-state properties (e.g., via Quantum Monte Carlo, QMC). The standard assumption is that an initial state with lower energy (closer to the ground state) will converge to the ground state faster. This paper investigates whether the Mpemba effect exists in this context: Can an initial state with higher energy relax to the ground state faster than a lower-energy initial state under imaginary-time evolution?

2. Methodology

The authors employed numerically exact Projective Quantum Monte Carlo (PQMC) simulations to investigate the imaginary-time relaxation dynamics of various interacting fermionic and bosonic models.

• Theoretical Framework: The system evolves according to the imaginary-time Schrödinger equation:
  $$-\frac{\partial}{\partial \tau} |\psi(\tau)\rangle = (\hat{H} - E_\tau) |\psi(\tau)\rangle$$
  The energy at time $\tau$ is given by $E_\tau = \langle \psi(\tau) | \hat{H} | \psi(\tau) \rangle$. The relaxation speed depends on the overlap coefficients ($a_n$) of the initial state $|\psi_I\rangle$ with the eigenstates of the Hamiltonian.
• Models Investigated:
  1. Interacting Dirac-Fermion Systems:
     • $\pi$-flux square Hubbard model (Spinful).
     • Honeycomb Hubbard model (Spinful).
     • Spinless $t$-$V$ honeycomb model (Density interaction).
  2. Fermi Surface Nesting (FSN) Systems:
     • Square-lattice Hubbard model (Zero flux).
  3. Bosonic Systems (Supplementary):
     • 1D XXZ spin chain.
• Initial States: The study utilized Mean-Field (MF) wave functions with varying order-parameter fields (e.g., Antiferromagnetic $M_N$ or Charge Density Wave $M_C$ fields) as initial states. These states allow for the tuning of initial energy and symmetry breaking.
• Simulation Details: The authors used Trotter decomposition and Hubbard-Stratonovich transformations to decouple interactions, employing Metropolis sampling for auxiliary fields. They focused on early-time dynamics to observe the crossing of energy curves.

3. Key Contributions

• Discovery of ITME: The paper identifies and names the Imaginary-time Mpemba Effect (ITME), demonstrating that higher-energy initial states can converge to the ground state faster than lower-energy states in imaginary-time evolution.
• Universality: ITME is shown to be a generic phenomenon occurring across different classes of interacting models (Hubbard, $t$-$V$, spin chains) and different interaction types (spin-spin, density-density).
• Mechanism Identification: The authors establish that the emergence of ITME is intimately linked to the presence of low-energy excitations (gapless modes) in the system, such as those found near Quantum Critical Points (QCPs) or due to Goldstone modes in ordered phases.
• Algorithmic Implication: The work challenges the conventional wisdom in numerical simulations that the lowest-energy MF state is the optimal starting point. It suggests that selecting an initial state with faster convergence (even if higher in energy) can significantly reduce computational cost.

4. Key Results

• Energy Curve Crossing: In all investigated models, the authors observed that energy curves $E_\tau$ vs. $\tau$ for different initial states cross. Specifically, a state with a higher initial energy (e.g., $M_N = 0.9$ in the $\pi$-flux Hubbard model at $U=4.0$) decayed faster and converged to the ground state at a smaller imaginary time $\tau$ compared to a state with lower initial energy (e.g., $M_N = 0.0$).
• Role of Quantum Criticality:
  • Near QCP: Pronounced ITME was observed in the vicinity of Quantum Critical Points (e.g., $U \approx 5.5$ for the $\pi$-flux Hubbard model, $V \approx 1.35$ for the $t$-$V$ model) where the system is gapless.
  • Deep in Ordered Phases:
    • In the AFM phase of the Hubbard model (large $U$), ITME persists due to gapless Goldstone modes arising from spontaneous symmetry breaking.
    • In the CDW phase of the spinless $t$-$V$ model (large $V$), where all excitations are fully gapped, ITME vanishes.
  • Non-interacting Limit: In the non-interacting limit ($U=0$) with generic MF initial states, ITME was absent. However, the authors showed that ITME can be artificially induced in non-interacting systems by constructing specific initial states with tailored overlaps with low-energy excited states.
• Ground State Properties: A surprising correlation was found: the initial state exhibiting the fastest convergence (the "hotter" state in the Mpemba sense) often yields physical observables (like structure factors $S_{AFM}$ or $S_{CDW}$) that align more closely with the true ground-state results than the initial state with the lowest energy.

5. Significance and Impact

• Fundamental Physics: ITME enriches the understanding of non-equilibrium quantum dynamics, bridging the gap between thermal relaxation, real-time quantum dynamics, and imaginary-time projection. It highlights the critical role of low-energy spectral structures in relaxation pathways.
• Computational Efficiency: This discovery has immediate practical implications for Quantum Monte Carlo (QMC) and other imaginary-time evolution algorithms:
  • Optimizing Initial States: The standard practice of minimizing the initial energy to select the starting wavefunction may be suboptimal. Selecting an initial state that exhibits ITME could drastically reduce the imaginary time required to reach convergence.
  • Mitigating the Sign Problem: In QMC simulations suffering from the sign problem, computational cost scales exponentially with imaginary time. By identifying initial states that converge faster (via ITME), the required simulation time can be shortened, potentially alleviating the severity of the sign problem.
• Future Directions: The paper opens avenues for exploring ITME in bosonic systems, experimental demonstrations on quantum computers (where imaginary-time evolution is increasingly implemented), and the development of a unified theoretical framework for the effect.

In summary, the paper uncovers a novel non-equilibrium phenomenon in quantum many-body systems that defies intuitive energy-relaxation hierarchies, offering both a deeper theoretical understanding of quantum relaxation and a concrete strategy for accelerating numerical simulations.