Abstract
This paper explores the fundamental connection between two phenomena in nanoscale friction: thermally assisted lubricity (TAL) and negative dissipated work (NDW) events. Using the one-dimensional Prandtl-Tomlinson model, we identify a general crossover when the total frictional work per cycle crosses the thermal energy threshold, $k_B T$. Below this crossover, the system exhibits linear viscous friction (TAL regime), and the Jarzynski equality (JE) is numerically verifiable. Above it, the system enters a hard stick-slip regime where JE verification requires exponentially large sampling. We also derive a parameter-free formula for the linear friction coefficient, significantly improving upon previous empirical models.
Introduction: The Nanoscale Friction Problem
At the macroscopic scale, friction is typically described by simple empirical laws. However, at the nanoscale, the behavior of a slider—such as an atomic force microscope (AFM) tip—is dominated by discrete "stick-slip" events as it moves over an atomic potential. Under specific conditions of high temperature or low velocity, this jerky motion transitions into a smooth, viscous-like regime known as thermally assisted lubricity (TAL).
Simultaneously, the Jarzynski Equality (JE) provides a bridge between non-equilibrium work and equilibrium free energy. Its validity hinges on the existence of rare thermal fluctuations that result in negative work—essentially "free lunches" where the environment gives energy back to the slider. This paper seeks to determine if a rigorous physical link exists between the onset of smooth sliding and the observability of these thermodynamic fluctuations.
Methods: Modeling the Slider
The researchers utilized the Prandtl-Tomlinson (PT) model, which represents a point mass slider dragged by a spring over a sinusoidal potential. To analyze the system, they applied adiabatic perturbation theory to the Fokker-Planck equation, allowing them to track the evolution of the slider's position probability distribution. These theoretical findings were validated through extensive overdamped Langevin dynamics simulations, spanning a wide range of temperatures ($k_B T = 0.01$ to $0.05$) and velocities.
Key Insights: The $k_B T$ Crossover
The central finding is that the transition between friction regimes is governed by the thermal energy, $k_B T$. When the average work per cycle $\langle W \rangle$ is less than $k_B T$, the system resides in the thermolubric regime. In this state, thermal fluctuations are strong enough to frequently push the slider over potential barriers, leading to smooth motion and a linear relationship between force and velocity.
This regime is also where the Jarzynski Equality is most accessible. Because the system frequently samples negative work events (the "tails" of the work distribution), the JE can be verified with a reasonable number of experimental trials. As $\langle W \rangle$ exceeds $k_B T$, the probability of these negative work events drops exponentially, making the JE practically impossible to verify without an astronomical number of samples.
Key Results and a New Friction Formula
The study yielded several significant results:
- Universal Scaling: When normalized by temperature ($\beta \langle W \rangle$), data from various temperatures collapse onto a single curve when plotted against rescaled velocity, confirming the universal nature of the $k_B T$ crossover.
- JE Verification: The authors quantified the "bias error" in the Jarzynski estimator, showing that the required sample size $N$ scales as $e^{\beta \langle W \rangle}$.
- Parameter-Free Prediction: A major theoretical contribution is the derivation of a new formula for the linear friction coefficient ($\gamma_{eq}$). Unlike previous models that relied on empirical fitting, this formula uses only the potential geometry and system damping, matching simulation data within ~15% across three orders of magnitude.
Limitations and Future Directions
While the theory holds for single-degree-of-freedom systems, the connection weakens in massive multi-contact interfaces. In large systems, local "free lunches" tend to cancel out, making global negative work unobservable even during smooth sliding.
The authors suggest that these findings could be tested in experimental setups involving trapped cold ions, where thermolubricity is known but work distributions are unexplored, or in forced protein unfolding, where the Jarzynski Equality is a standard tool. Future work may also extend these analytical derivations to underdamped systems where inertial effects become significant.
Summary updated