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

The central problem the KAIST group, led by corresponding author Ki-Uk Kyung, attacked is one that anyone who has dragged a fingernail across a chalkboard will recognize intuitively. When two surfaces in contact begin to slide, friction does not behave politely. Static friction, the force that must be overcome to start motion, is typically higher than kinetic friction, the force that resists motion once sliding is underway. In an electrostatic clutch, where voltage presses two films together and controls how much shear force the interface can transmit, this disparity produces the notorious stick–slip instability: the interface grabs, builds up elastic strain, breaks loose, snaps forward, and grabs again, cycling hundreds of times per second. The result is a vibration that ruins precision and, worse, makes the transmitted force effectively binary. Most electrostatic clutches reported to date can only be commanded fully on or fully off, which is a poor match for robots that need to modulate stiffness continuously, the way a human hand modulates its grip.

The KAIST team’s answer lies in the choice of dielectric. Conventional electrostatic clutches sandwich ordinary insulating films between electrodes, relying on simple capacitive attraction to press the surfaces together. The new device instead uses a plasticized polyvinyl chloride gel, a soft polymer material with an unusual electrostatic property: when a voltage is applied, charges accumulate at the interface between the gel and the electrode rather than merely sitting on capacitor plates. This interfacial charge accumulation generates electrostatic pressures far beyond what a conventional dielectric of the same thickness can produce. The numbers reported in the paper are striking. The device achieves a shear stress capacity of approximately 29 newtons per square centimeter at just 100 volts. For comparison, typical electrostatic clutches in the literature require driving voltages in the kilovolt range to reach comparable force densities, which means heavy insulation, bulky high-voltage supplies, and serious safety considerations for any wearable application.

Force alone, however, would not have been enough, and this is where the materials science becomes genuinely clever. The researchers systematically tuned the amount of plasticizer blended into the PVC gel, discovering that the formulation affects not just the magnitude of the electrostatic force but the entire dynamic character of the friction. The optimized composition does two things simultaneously. First, it minimizes the gap between static and kinetic friction, so the interface no longer has a strong incentive to grab and release cyclically. Second, it induces a velocity-strengthening friction response, meaning that as the sliding speed increases, the friction force rises rather than falls. That behavior is dynamically stabilizing: any tendency of the interface to accelerate is met with increasing resistance, which damps the oscillation that would otherwise grow into stick–slip. The combination ensures continuous, judder-free sliding even while the interface is transmitting substantial shear loads.

The practical payoff is quantified in the paper’s headline demonstration: precise control of kinetic friction while the device carries a shear stress of 22.5 newtons per square centimeter. In other words, the clutch can hold a heavy load, then release it into smooth, metered sliding, all under low-voltage electronic command. Because the friction can be set continuously anywhere between locked and free rather than snapped between two states, the device functions as a true variable clutch. A robotic arm equipped with such an interface could stiffen its joints to manipulate a heavy object, soften them the instant it detects an unexpected collision, and modulate the transition smoothly enough that neither the robot nor whatever it touches experiences a jolt. Programmable mechanical compliance of this kind is exactly what safety standards for collaborative robots have been pushing the industry toward.

To show the mechanism is more than a laboratory curiosity, the team validated it across three distinct robotic control scenarios. The first is programmable impact damping: by electronically setting the friction level, the device can absorb and dissipate impact energy on demand, acting like a shock absorber whose damping curve is written in software rather than cast into a viscous fluid. The second is the variable clutch itself, where the modulator couples and decouples mechanical power transmission with continuously adjustable strength. The third, and perhaps the most immediately tangible for consumers, is a wearable haptic device. Because the thin film is lightweight and operates at 100 volts rather than several kilovolts, it can plausibly be integrated into a glove or sleeve, rendering realistic tactile sensations by varying the friction a user’s fingertip experiences as it strokes a surface, from slick ice to gritty sandpaper.

The low-voltage operation deserves particular emphasis because it changes the engineering envelope of the whole field. High-voltage electrostatic clutches have historically been confined to tethered laboratory setups, with isolation circuitry that makes them impractical for mobile robots or battery-powered wearables. Dropping the drive requirement by more than an order of magnitude means the clutch electronics can shrink toward the scale of the display drivers and motor controllers already embedded in commercial robots. The thin-film form factor compounds the advantage: the modulator adds negligible mass and can conform to curved joint surfaces, in contrast to electromagnetic clutches whose copper windings and steel cores concentrate weight exactly where a robot can least afford it, at the distal end of its limbs.

There is also a broader materials-science lesson embedded in the result. PVC gels have been studied for years as electroactive polymers, best known for creeping and deforming under high voltage in actuator demonstrations. This work reframes the material as a charge accumulator whose interfacial electrostatics can be engineered through formulation chemistry. The finding that plasticizer content governs the velocity dependence of friction, and through it the stability of sliding, connects polymer physics directly to a control problem that has limited electrostatic clutching for a decade. It suggests a design space in which tribological behavior, not just force output, becomes a tunable property of the material itself, opening the door to gels optimized for specific load ranges, sliding speeds, or environmental conditions.

The research, led by Jihyeong Ma, Jongseok Nam, and Nakhyeong Lee together with Kyung in KAIST’s Department of Mechanical Engineering, was supported by the National Research Council of Science and Technology, the Institute of Information and Communications Technology Planning and Evaluation, and the National Research Foundation of Korea. The authors report no competing interests, and the paper is open access, with supplementary videos documenting the damping, clutching, and haptic demonstrations. If the stick–slip-free performance holds up under the wear, temperature swings, and millions of duty cycles that real robots inflict on their components, the PVC-gel friction modulator could become a standard building block of a new generation of machines that are simultaneously strong, safe, and exquisitely adjustable, gripping the world with a firmness chosen moment by moment in software.

Subject of Research: PVC-gel electrostatic friction modulation for variable robotic clutches

Article Title: Thin-film dynamic friction modulator interfaced with PVC-gel charge accumulator for variable clutching of robotic systems

Article References: Ma, J., Nam, J., Lee, N., & Kyung, K.-U. (2026). Thin-film dynamic friction modulator interfaced with PVC-gel charge accumulator for variable clutching of robotic systems. Advanced Composites and Hybrid Materials. https://doi.org/10.1007/s42114-026-02056-3

Image Credits: AI Generated

DOI: 10.1007/s42114-026-02056-3

Keywords: PVC gel, electrostatic clutch, friction modulation, stick-slip, soft robotics, haptics, variable clutch, mechanical compliance, interfacial charge accumulation, wearable devices, KAIST, electroactive polymers