PMA additives were never originally meant to fight friction. They entered lubricants as viscosity index improvers, long-chain polymers added to keep oil from thinning too much at high temperatures. But researchers later discovered that certain PMA molecules carry polar groups that latch onto metal surfaces, forming lubricating films roughly 10 to 100 nanometers thick that reduce friction and wear even when the bulk oil can no longer separate the surfaces. Subsequent work showed that these adsorbed layers, sometimes called polymer brushes, push the opposing surfaces apart through strong intermolecular repulsion described by the Flory exclusion volume effect. If that repulsion is the source of friction reduction, then the molecular architecture of the polymer—its polar groups, side chains, and molecular weight—should determine how well it performs. What remained unclear was how these adsorbed films behave dynamically: how they change during sliding, and how those changes feed back into friction.

To answer that question, the Kyoto team built a friction tester unlike anything in the standard tribology toolkit. Instead of point contacts between a ball and a disk, the researchers fabricated film-shaped test specimens on silicon-on-insulator wafers using microelectromechanical systems, or MEMS, processes. Each specimen was a 5-micrometer-thick silicon film with a sliding surface so smooth that its arithmetic mean roughness measured just 0.34 nanometers. The sliding region measured 2.0 by 0.6 millimeters, giving a wide-area contact of roughly 1 square millimeter—far larger than the point contacts of conventional testers, and therefore involving far more molecules in the friction phenomenon. Because the tester applies load electrostatically, by generating an attractive force between the film and a silicon-nitride-coated grip, it can hold surfaces in parallel contact with a nanoscale oil gap while a piezoelectric stage drives unidirectional sliding at speeds as low as 0.05 micrometers per second.

The same electrostatic architecture doubles as a measurement instrument. By superimposing a 1-volt AC signal at 10 kilohertz on the DC bias that generates the load, the researchers measured the electrical impedance of the sandwiched oil film and calculated its capacitance, from which the film thickness followed directly. The frequency response analyzer offered gain accuracy of ±0.01 dB and phase accuracy of ±0.06 degrees, and the team confirmed the circuit was capacitive at the measurement frequency. The researchers also took care to rule out confounding electrostatic effects: prior work has shown that friction-induced static voltages under oil lubrication reach only about 1 volt even over 100 meters of sliding, negligible compared with the 100 to 200 volts applied here. Each test oil was adjusted so its kinematic viscosity at 30 degrees Celsius was approximately 30 square millimeters per second, ensuring that viscosity differences could not masquerade as additive effects, and all oils shared a dielectric constant of 2.2.

The team compared four oils: a base oil of API Group III mineral oil, an oil containing an olefin copolymer, or OCP, as a non-adsorbing control, and three oils containing PMA additives with different molecular weights and polar group contents. LowPMA1 and LowPMA2 both had a molecular weight of 30,000, but LowPMA1 contained many polar groups while LowPMA2 contained almost none. HighPMA, with a molecular weight of 430,000, carried many polar groups. When the researchers swept shear velocities from 0.05 to 0.80 micrometers per second, a clear divide emerged. The base oil and the OCP-containing oil showed textbook behavior: a static coefficient of friction immediately after shearing began, followed by a constant dynamic coefficient. All three PMA-containing oils, by contrast, exhibited pronounced stick-slip motion at low shear velocities—and, crucially, the friction coefficient remained low throughout the stick-slip region, with the PMA oils showing lower friction than the base oil and OCP oil at every applied load.

Stick-slip motion, in which surfaces alternately seize and release as they slide, is usually a nuisance associated with elevated friction. Here it marked a low-friction regime instead. The researchers calculated average friction coefficients for the base oil and OCP oil from the dynamic friction interval, and for the PMA oils from the onset of stick-slip motion. At each of the three applied voltages—100, 150, and 200 volts, corresponding to loads of roughly 1 to 3 newtons and a surface pressure near 0.83 megapascals—the PMA oils maintained lower friction. Notably, no clear difference in friction coefficient appeared among the three PMA additives, despite their very different molecular structures. The friction story, it turned out, was only half the picture; the film thickness measurements told the rest.

Statistical analysis of 30 measurements per oil, reported with standard errors and 95 percent confidence intervals based on the Student’s t-distribution, revealed that the oil film thicknesses of all three PMA oils exceeded those of the base oil and OCP oil. But the additives diverged under load. LowPMA1, rich in polar groups, held its film at approximately 30 nanometers regardless of the applied voltage, forming what the researchers interpret as a stiff, strongly adsorbed interfacial film. LowPMA2, nearly devoid of polar groups, saw its film thickness decrease as the load increased, thinning to about 23 nanometers at 200 volts—still thicker than the base oil or OCP film under the same conditions, but clearly more compressible. The team attributes this to weaker repulsive forces between sparsely adsorbed molecules: neutron reflectometry measurements cited in the study showed that LowPMA2 adsorbs in smaller amounts onto silicon substrates, producing a sparser film that yields under compression.

HighPMA told the most surprising story. Before any friction was applied, its film was the thickest of all the test oils, averaging roughly 36 nanometers at 200 volts, and it showed no tendency to thin as the load increased. After the friction load, however, the film thinned to approximately 27 nanometers—comparable to LowPMA1—even though its friction coefficient remained essentially unchanged. The researchers propose that within the molecular weight distribution of HighPMA, the longest polymer chains adsorb preferentially onto the surface, in line with adsorption theory: a polymer with more segments adsorbs when the combined adsorption energy of its monomers overcomes the entropic cost, and higher-degree-of-polymerization polymers adsorb at lower concentrations. Shorter chains then find fewer available adsorption sites and settle weakly on top of the strongly adsorbed layer. During sliding, shear forces desorb or rearrange this weakly bound topmost layer, leaving only the robustly anchored film beneath—yet the presence of that sacrificial layer before sliding may have created a low-shear zone that kept friction low throughout the test.

This layered picture echoes earlier work on hydration lubrication, in which free polymer chains adsorbing onto immobilized polymer brushes formed composite films that sheared more easily than the brushes alone. The Kyoto team’s data suggest a similar mechanism operates in HighPMA oil: a composite adsorbed film, built from chains of different molecular weights, delivers low friction while its weakly bound component is shed under shear. The researchers also found that the PMA oils showed weak load dependence of friction force—unlike the base oil, whose friction force and load correlated strongly with a coefficient of 0.916—indicating that PMA additives not only reduce friction in the boundary regime but can extend the low-friction region to higher loads. Given that comparable macroscopic tests typically use specimens with root-mean-square roughness of 10 to 20 nanometers, the team expects the same adsorption mechanism to operate at least at that roughness level on practical engineering surfaces.

The findings translate molecular structure directly into film behavior: the number of polar groups governs the stiffness of the adsorbed film in the load direction, while molecular weight governs film thickness and the formation of the shear-sensitive topmost layer. For lubricant designers, the message is that PMA’s molecular weight and polar group content must be controlled together to build a strongly adsorbed layer that covers the surface at the lowest possible concentration, with a weakly adsorbed complement that expands the low-friction zone. The researchers note that polymer desorption is expected to vary with temperature, and they identify temperature-dependent desorption behavior as the key next question. If future work confirms how these nanoscale films restructure under heat as well as shear, the path from polymer architecture to cooler, more efficient engines becomes considerably shorter.

Subject of Research: A Kyoto University team used a MEMS friction tester to show how polar groups and molecular weight of polymethyl methacrylate additives control the stiffness, thickness, and shear-induced restructuring of nanoscale adsorbed films that reduce friction under boundary lubrication.

Article Title: Friction properties of boundary film formed by polymer additives in wide-area contacts with nanoscale gap

Article References: Yagi, W., Tsuchiya, T., Yamashita, N., & Hirayama, T. (2026). Friction properties of boundary film formed by polymer additives in wide-area contacts with nanoscale gap. Frontiers in Mechanical Engineering, 12, Article 1989288. https://doi.org/10.3389/fmech.2026.1989288

Image Credits: AI Generated

DOI: 10.3389/fmech.2026.1989288

Keywords: boundary lubrication, polymethyl methacrylate, friction modifiers, adsorbed polymer film, MEMS, tribology, oil film thickness, stick-slip, viscosity index improver, lubricant additives, Friction, properties