Granite sliding on granite: friction, wear rates, surface topography, and the scale-dependence of rate–state effects

Deep within the Earth’s crust, massive tectonic plates shift. But what happens at the microscopic level? Physicists from Forschungszentrum Jülich and Saarland University have proposed a new explanation: The rock grains do not simply interlock – they bond together at their contact points.

When tectonic plates move, they rarely do so smoothly. Sometimes they slide almost imperceptibly; at other times, stress is suddenly released – resulting in an earthquake. What exactly governs this behavior remains one of the key open questions in earthquake research.

From a physical perspective, the problem comes down to friction. To better understand the underlying processes, the researchers studied friction between granite surfaces – an established model system for tectonic faults. Combining experiments, simulations, and theory, they arrive at a clear conclusion: friction in rocks arises differently than previously thought.

Bonding instead of scratching
Conventional models attribute friction mainly to mechanical effects: rough surfaces interlock, wear down, and sharp asperities scratch grooves into the material. Most earthquake models are based on these assumptions.

However, the new findings challenge this view. “Wear is not the dominant factor. Instead, we were able to show that another mechanism governs friction,” explains Dr. Bo Persson from the Peter Grünberg Institute (PGI-1) in Jülich.

At microscopic contact points, rock surfaces form chemical bonds – a process similar to cold welding. “As rocks slide past each other, these bonds continuously form and break. This requires energy – and that is what generates friction,” says Bo Persson.

The researchers also showed an important size-effect. “For small systems the bond breaking occurs uniformly while for large systems the bond-breaking occurs non-uniformly – some regions start to slip before other regions – which reduces the so called breakloose friction force,” Persson adds.

Insights from simulations
The results are supported by computer simulations carried out at Saarland University. Prof. Martin Müser and his team have been studying friction between solids under extreme conditions for many years.

“In our simulations, we identified the breaking of bonds as the main source of friction,” explains Martin Müser. “In addition, deformation and local melting processes in the material affect the friction – effects that also play a role in existing models.”

A new perspective on earthquakes
The findings suggest that tectonic plates behave differently at the microscopic level than previously assumed. Traditional models assume that stress builds up over time and is then suddenly released in a rupture.

The new model paints a different picture: motion begins much earlier. The plates are never completely at rest but move continuously – albeit extremely slowly, often at rates of just fractions of a nanometer per second. This corresponds to only a few millimeters per year and is known as “creep.”

From creep to slip
At the microscopic level, chemical bonds are constantly breaking and reforming. As velocity increases, friction initially rises because more bonds must be broken per unit time. However, once a critical threshold is reached, the system changes: bonds can no longer reform quickly enough, and local heating effects occur. As a result, friction suddenly drops.

“The system transitions from slow creep to rapid sliding – and that could be a key trigger of earthquakes,” says Persson.

The findings could help refine existing models. “We need a better understanding of how friction depends on motion,” Persson adds. “That could be crucial for describing earthquake processes more realistically.”