Fluctuation theorems for a non-Gaussian system

This study numerically verifies that the Jarzynski equality and Crook fluctuation theorem hold for a non-Gaussian Brownian particle in a heterogeneous thermal bath described by a diffusing-diffusivity model, demonstrating that these fundamental relations remain valid even when the work distribution is non-Gaussian.

The Gist

Here is an explanation of the paper, translated into everyday language with some creative analogies.

The Big Picture: Testing the Rules of the Game

Imagine you are playing a board game where the rules are supposed to be fair and predictable. In the world of physics, there are two famous "rules of the game" called the Jarzynski Equality and the Crooks Fluctuation Theorem.

These rules tell us how much "work" (energy effort) is needed to change a system from one state to another. For a long time, scientists believed these rules only worked perfectly for "normal" systems—systems where things move in a predictable, bell-curve pattern (like a calm crowd walking in a straight line).

But what happens if the system is "weird"? What if the crowd is chaotic, or the ground they are walking on is uneven? Do the rules still hold?

This paper asks: Do these fundamental laws of physics survive when the system is chaotic and "non-Gaussian" (meaning it doesn't follow the standard bell curve)?

The Setup: A Hiker on a Shifting Trail

To test this, the authors created a digital simulation of a Brownian particle. Think of this particle as a tiny hiker.

1. The Environment (The Thermal Bath): Usually, we imagine the hiker walking on a flat, smooth floor. But in this experiment, the floor is a heterogeneous thermal bath. Imagine the hiker is walking on a trail where the mud depth changes randomly. Sometimes the mud is thin (easy to walk), and sometimes it's deep (hard to walk).
2. The "Diffusing-Diffusivity" Concept: The key twist is that the hiker's ability to move (their "mobility") isn't fixed. It fluctuates. One moment they are sprinting, the next they are stuck in deep mud. This is called the diffusing-diffusivity model. It's like the hiker's shoes are changing size and grip every second.
3. The Task: The hiker is trapped in a "breathing parabola." Imagine a valley that gets wider and narrower over time (like a giant, invisible hand squeezing and releasing a spring). The scientists change the shape of this valley to force the hiker to do work.

The Experiment: Pushing the Hiker

The researchers ran a massive computer simulation with one million different hikers (trajectories). They pushed these hikers through the shifting mud and changing valleys.

They wanted to see two things:

1. The Jarzynski Equality: Does the average "effort" of all these hikers still match the theoretical energy difference, even if some hikers got stuck in deep mud and others sprinted?
2. The Crooks Theorem: If you run the movie backward (un-squeezing the valley), does the probability of the hikers' movements match the forward movie in a specific mathematical way?

The Results: The Rules Still Hold!

Here is the surprising discovery: Yes, the rules still work.

Even though the hikers were moving through chaotic, shifting mud (a non-Gaussian system), the Jarzynski Equality and the Crooks Fluctuation Theorem remained perfectly valid.

• The Analogy: Imagine you are betting on a horse race. Usually, you expect the horses to run at a steady pace. But in this race, the track is made of jelly that changes shape every second. Some horses slip, some fly. You might think the old betting rules are broken. But the authors found that if you look at the average of all the crazy outcomes, the old betting rules still predict the outcome perfectly.

The Twist: The "Weirdness" Lasts Longer

While the big rules held up, the details were different.

In a normal system (smooth floor), if you watch the hikers for a long time, their movements eventually settle into a predictable, smooth pattern (a Gaussian distribution).

But in this "shifting mud" system:

• The hikers' movements remained chaotic and unpredictable for a very long time.
• Even after a long process, the distribution of their positions didn't smooth out into a nice bell curve. It stayed "spiky" and weird.
• The Work Distribution: The amount of "work" done by the hikers also stayed weird. In normal systems, the work eventually becomes predictable. Here, the work kept fluctuating wildly, showing that the "shifting mud" creates a lasting memory of chaos.

Why Does This Matter?

This is a big deal for physics and engineering.

1. Robustness: It proves that the fundamental laws of thermodynamics are incredibly strong. They don't break just because the environment is messy or the particles are behaving strangely. They work even in the "jelly mud."
2. Real-World Applications: Many real-world systems (like proteins folding inside a cell, or nanoparticles moving through complex fluids) are exactly like this "shifting mud" system. They are non-Gaussian. This paper tells us we can use these powerful thermodynamic tools to understand and design those complex systems with confidence.

Summary

The authors took a system that was supposed to be "too messy" for standard physics rules, simulated it with a million virtual particles, and found that the universe's fundamental rules are tougher than we thought. Even in a chaotic, shifting environment, the math of energy and work still holds true, even if the path to get there is much wilder than expected.

Technical Summary

Here is a detailed technical summary of the paper "Fluctuation theorems for a non-Gaussian system" by Saravanan and Iyyappan.

1. Problem Statement

The paper addresses a fundamental question in non-equilibrium statistical mechanics: Do the Jarzynski equality and the Crooks fluctuation theorem remain valid for systems exhibiting non-Gaussian position distributions that arise from thermal bath heterogeneity?

• Context: Classical fluctuation theorems (Jarzynski equality and Crooks theorem) relate non-equilibrium work to equilibrium free energy differences. While extensively verified for Gaussian systems (where fluctuations are driven by standard thermal noise), their validity in non-Gaussian regimes is debated.
• Specific Gap: Previous studies confirmed these theorems for non-Gaussian systems where the non-Gaussianity stems from non-harmonic potentials. However, this study investigates a different scenario: a particle confined by a harmonic potential (Gaussian potential) where the non-Gaussianity arises solely from the heterogeneity of the thermal bath.
• System Characteristics: The system exhibits "Brownian (Fickian) yet non-Gaussian diffusion," meaning the mean-squared displacement grows linearly with time (normal diffusion), but the position probability distribution is non-Gaussian (characterized by non-zero excess kurtosis).

2. Methodology

The authors employed numerical simulations based on stochastic thermodynamics to model a Brownian particle in a heterogeneous thermal bath.

• Physical Model (Diffusing-Diffusivity):
  • The system is modeled using the diffusing-diffusivity concept, where the particle's mobility ($\mu$) is a fluctuating quantity rather than a constant.
  • Langevin Equations:
    • Particle position: $\frac{dx}{dt} = -\mu(t)\frac{\partial V}{\partial x} + \sqrt{2\mu(t)k_B T}\xi(t)$
    • Fluctuating mobility: $\mu(t) = Y^2(t)$, where $Y(t)$ follows an Ornstein-Uhlenbeck process: $\frac{dY}{dt} = -\frac{Y}{s} + \sigma\eta(t)$.
  • Here, $\xi(t)$ and $\eta(t)$ are independent Gaussian white noises. The parameters $s$ (correlation time) and $\sigma$ (fluctuation strength) control the bath heterogeneity.
• Thermodynamic Process:
  • The particle is confined by a time-dependent harmonic potential: $V(x, \lambda) = \frac{1}{2}\lambda(t)x^2$.
  • Protocol: The stiffness coefficient $\lambda(t)$ is varied sinusoidally over a period $\tau$: $\lambda(t) = A - B \sin(2\pi t/\tau)$.
  • Isothermal Cycle: Since $\lambda(0) = \lambda(\tau)$, the initial and final free energies are identical ($\Delta F = 0$). This simplifies the Jarzynski equality to $\langle e^{-\beta w} \rangle = 1$.
• Simulation Parameters:
  • Integration: Euler-Maruyama method was used to solve the coupled Langevin equations.
  • Ensemble: $10^6$ independent trajectories were simulated.
  • Comparison: Results were compared against a control case with constant mobility (standard Gaussian system) where $\mu = 1$.
  • Metrics: The study analyzed work distributions ($P(w)$), the Jarzynski equality, the Crooks relation, and higher-order moments (variance and kurtosis) of both position and work.

3. Key Contributions

1. Validation of Theorems in Heterogeneous Baths: The study provides numerical evidence that the Jarzynski equality and Crooks fluctuation theorem hold true even when the non-Gaussian nature of the system originates from bath heterogeneity (fluctuating diffusivity) rather than potential shape.
2. Characterization of Work Fluctuations: The authors demonstrate that while the theorems hold, the statistics of the work differ significantly from Gaussian systems. Specifically, the work distribution remains non-Gaussian for significantly longer times in diffusing-diffusivity systems compared to constant-diffusivity systems.
3. Non-Monotonic Dynamics: The paper identifies unique non-monotonic behaviors in the average work and work variance for non-Gaussian systems, contrasting with the monotonic relaxation observed in standard Gaussian systems.

4. Results

• Jarzynski Equality:

  • Numerical results confirmed that $\langle e^{-\beta w} \rangle \approx 1$ for both Gaussian and non-Gaussian (diffusing-diffusivity) systems across various process times ($\tau$).
  • Small fluctuations were observed in the non-Gaussian case, but the equality was satisfied within statistical error, confirming its robustness.

• Crooks Fluctuation Theorem:

  • The relation $\ln[P_F(w)/P_R(-w)] = \beta(w - \Delta F)$ was verified.
  • Plots of $\ln[P_F(w)/P_R(-w)]$ vs. $w$ showed a linear dependence with a slope of $\beta$ for both system types.
  • Intersection Point: For Gaussian systems, the forward and reverse work distributions intersect exactly at $\langle w \rangle = 0$. For non-Gaussian systems, the intersection point was slightly shifted, suggesting that larger statistical sampling is required to resolve the crossing point accurately in heterogeneous baths.

• Non-Gaussianity of Position:

  • The kurtosis of the position distribution ($K_x$) remained significantly higher than 3 (the Gaussian value) throughout the process, even for long times ($\tau=10$), confirming the persistent non-Gaussian nature of the position distribution.

• Work Distribution Statistics:

  • Average Work: In Gaussian systems, mean work decreases monotonically to zero as $\tau$ increases. In non-Gaussian systems, mean work initially increases, reaches a maximum at an intermediate $\tau$, and then decreases.
  • Work Variance ($\sigma_w^2$): Gaussian systems show monotonic decrease in variance with time. Non-Gaussian systems exhibit non-monotonic behavior, with enhanced fluctuations induced by the bath heterogeneity.
  • Work Kurtosis ($K_w$):
    • For Gaussian systems, the work distribution converges to a Gaussian form relatively quickly (around $\tau \approx 100$).
    • For non-Gaussian systems, the convergence to Gaussian behavior is delayed. The kurtosis increases initially, reaches a maximum, and only decays to 3 at very long times. This delay is attributed to the large correlation time of the fluctuating mobility, which sustains non-Gaussian characteristics in the work distribution.

5. Significance

• Theoretical Robustness: The findings reinforce the universality of the Jarzynski equality and Crooks theorem, showing they are not limited to Gaussian fluctuations or specific potential shapes. They remain valid even in complex environments with time-correlated, fluctuating diffusivity.
• Implications for Microscopic Engines: The results highlight that non-Gaussianity in thermal baths can qualitatively alter the performance and fluctuation statistics of microscopic heat engines (e.g., average work and variance), even if the fundamental thermodynamic bounds (free energy differences) remain unchanged.
• Search and Transport: The paper connects to broader applications where non-Gaussianity plays a dual role: it can be advantageous for search processes with few agents but disadvantageous when many agents are required. Understanding the relaxation times of work distributions in such systems is crucial for optimizing such processes.
• Experimental Relevance: The model (diffusing diffusivity) is highly relevant to biological systems and active matter where the environment is inherently heterogeneous, suggesting that experimental verification of fluctuation theorems in these contexts should account for delayed relaxation to Gaussian statistics.