Molecular-scale Wear and Tear

Drivers, start your engines! To motorsport fans, that famous call brings to mind the roar of powerful automobiles and the start of high-speed competition. To some engineers and surface scientists, however, the words mark the onset of tribochemical reactions.

As drivers rev their engines, machine components such as valves and gear wheels undergo furious revolutions and other types of motions that cause engine parts to slide past one another at high speed. Even in well-lubricated engines, the relative motions of the machine parts cause friction and wear and tear and stimulate chemical changes at the interface between the sliding surfaces and a lubricant film that's less than a micrometer thick.

Tribology-the study of friction, wear, and lubrication-has clear applications to race car engines and other topics related to classical mechanical engineering. But while today's tribologists continue to study traditional engineering subjects, some of the field's practitioners focus on applications in developing areas. Among other specialties, the list includes tribochemical reaction mechanisms, tribological properties of nanoscale materials, and tribology of micromachines and biological systems.

More than 1,000 scientists from roughly 50 countries gathered in Washington, D.C., last month to discuss those topics and others at the 3rd World Tribology Congress. Sponsored by the American Society of Mechanical Engineers and the Society of Tribologists & Lubrication Engineers, the conference, which is held every four years, gave attendees an opportunity to catch up on the latest research in a wide variety of areas.

In tribology, expertise runs from A to Z, remarked Stephen M. Hsu, nanotribology group leader at the National Institute of Standards & Technology (NIST), Gaithersburg, Md. Hsu, who was one of the meeting's general organizers, explained that the field is made up of physicists, engineers, chemists, and materials scientists. It also includes people with expertise in manufacturing bearings and engines and in formulating lubricants.

The groups are rather distinct, and they don't usually interact very much, Hsu acknowledged. But interdisciplinary collaborations are essential, and they're on the rise. One of the principal aims of the congress, according to Hsu, was to foster cross-pollination-to encourage interaction among tribologists with distinct backgrounds.

In one of the symposia devoted to tribochemistry, Hsu presented an overview of the connections between tribochemistry and lubrication that included some of his group's results and those of other researchers. In an opening comment, he noted that, at the interface between surfaces in sliding contact, tribological processes initiate chemical reactions that alter the applied lubricant, transforming it into a protective film. To be an effective lubricant, the reaction product must be a strong and durable film that minimizes friction and wear between the surfaces.

We want the protective film to form as rapidly as possible and then be used as a sacrificial lamb that's worn away during contact, Hsu said. The question is, what stimulates the onset of tribochemical reactions?

The NIST researcher offered several possibilities. For example, depending on surface roughness, forces associated with rubbing motions can induce spikes in the temperature and pressure in the vicinity of microscopic bumps (asperities), which lead to chemical reactions. Or rubbing metal surfaces can scrape surface films, thereby exposing fresh metal sites that may react with a lubricant directly or catalyze lubricant reactions.

Another possibility is that mechanical forces disrupt surface bonds, leading to charged-particle emission and formation of dangling bonds (unsatisfied valencies), which in turn may stimulate chemical reactions.

Sorting out the roles played by various tribological processes in lubricant chemistry and the exact nature of the reaction products has been challenging. But some pieces of the puzzle have fallen into place. More than 20 years ago, for example, Hsu and coworkers examined reaction products from paraffin lubricants used in wear tests on steel parts and identified oil-soluble organo-iron compounds. Later, through the use of gel permeation chromatography and atomic absorption spectroscopy methods, Hsu, Richard S. Gates, and coworkers at NIST determined that the wear processes form organometallic polymers with molecular weights ranging from 1,000 to 100,000 atomic mass units.

In follow-up studies, scientists at Pennsylvania State University, in collaboration with the NIST team, proposed that the high-molecular-weight species-so-called friction polymers-are formed via a multistep mechanism that begins with oxidation of the hydrocarbon lubricant in the contact area to form carboxylic acids. On steel surfaces, the organic acids react with iron oxides, which function as hydrogen-abstraction catalysts, leading to formation of conjugated molecules.

According to Hsu, the conjugated species can grow via polymerization reactions until they reach a molecular weight of about 100,000 amu. Molecules that large precipitate from the oil solution in the form of a brown sludge that is often observed near wear spots. Interestingly, Hsu pointed out that in wear tests in which no high-molecular-weight products were detected, lubrication was generally found to be ineffective.

To gauge the importance of thermal chemistry on the lubrication reactions, the team oxidized thin films of lubricants under static conditions (no rubbing) at 225300 C-the range associated with temperature spikes during sliding contact. They found similar products under static and dynamic conditions and concluded that the tribochemical reactions that formed organometallic friction polymers were initiated by heat generated at sliding contacts. Additional evidence for the dominance of thermal effects came from wear tests in which the researchers exposed fresh metal sites in a controlled manner. In those tests, the product yield increased, but by no more than 15%, Hsu reported.

More recently, the NIST scientists designed an experiment to evaluate the significance of bond disruption and charged-particle emission in tribochemistry. In that study, a glass tip was coated with diamond particles of known size and was used to scratch specimens in a reproducible manner with minuscule force to avoid heating the sample surface. Moving the tip across the surface dislodged molecules in shallow swaths of nanometer width and depth and pushed them aside. Hsu explained that dislodging surface molecules liberates electrons that can stimulate subsequent reactions. The samples were two model systems that the group had studied in detail previously: stearic acid films deposited on copper and on steel.

Using a surface-sensitive infrared spectroscopy technique, the team observed that contact between the tip and sample led to formation of copper and iron stearate species, but the effect was modest, according to Hsu (Tribol. Lett. 2002, 13, 131). The same types of products were detected in rubbing experiments and in static heating measurements. But when the sample was heated, even gently, the yield was much greater than the yields observed from the other methods, leading Hsu to conclude that tribochemistry in metal systems is dominated by thermal effects.

Another piece of the tribochemical puzzle that has been examined by the NIST team and other research groups is the role of antiwear additives. Oil formulators typically add compounds such as zinc dialkyldithiophosphate (ZDDP) to protect engine parts. In light of recent experiments, Hsu explained that under typical wear conditions, ZDDP decomposes at the contact site to form a hard, glassy zinc phosphate film. Although the film may be hard, he said, the protection it provides would be limited if it functioned alone, because the material can be cracked easily. But through a synergistic effect, in which a soft, shearable friction polymer coats the glassy phosphate, the two materials protect and lubricate the contact area (Tribol. Int. 2005, 38, 305).

Nanostructured materials are another type of lubricant additive under investigation in several laboratories. At Ecole Centrale de Lyon, in France, for example, assistant professor Fabrice Dassenoy studies tribological properties of inorganic fullerene-like (IF) nanoparticles and carbon nanotubes. Dassenoy, who also is affiliated with the French National Research Center (CNRS), pointed out that MoS₂, a common solid-phase lubricant with a layered structure, undergoes oxidizing reactions that generate polluting sulfur compounds. In contrast, the spherical IF form of MoS₂ and related metal disulfides are more stable-making those compounds (and sulfur-free materials) candidates for environmentally friendly lubricant substitutes.

In one study, conducted in collaboration with materials science professor Reshef Tenne and coworkers at Weizmann Institute of Science, Rehovot, Israel, and others, the team prepared a series of nanoparticle-poly--olefin oil mixtures of various concentrations. In friction tests, in which a steel pin was rubbed against a steel plate in ambient conditions under carefully applied loads and speeds (the instrument is known as a pin-on-flat tribometer), the group found that adding just 0.1 weight % IF-WS₂ reduces the coefficient of friction nearly 50% and leads to very little wear compared with the pure oil (Tribol. Lett. 2005, 18, 477). Increasing the concentration up to 1 wt % reduces the friction even further, Dassenoy said. He noted that similar results were obtained with IF-MoS₂ and IF-NbS₂.

In contrast with pure oil, which gave high initial friction values that decreased with continued rubbing, the nanoparticle formulations gave an immediately observable friction benefit, the Lyon scientist pointed out. That property may reduce friction during cold engine start-up, he commented. The research team also observed-based on microscopy and spectroscopy measurements-that the structures of the fullerene-like compounds evolve during the wear test into the sheet forms, which may be one of the sources of the good tribological properties.

Using the same types of methods in a related study, the Lyon team and their coworkers at Shizuoka University and Kobe University, both in Japan, found that 1 wt % of single-walled carbon nanotubes reduces the friction coefficient some 70% compared with pure oil. According to Dassenoy, the wear process converts the nanotubes into a hard carbon-based film that protects the surfaces from abrasion (Tribol. Int. 2004, 37, 1013).

Animal joints are a little off the beaten path for most tribologists, but for decades, Van C. Mow has been examining the topic from an engineering perspective, using computational techniques and the methods of applied mechanics. Mow, a professor and chairman of the biomedical engineering department at Columbia University, described the structures of various animal joints and nature's methods for lubricating them in one of the conference's plenary lectures.

Ordinary activities subject diarthrodial joints, such as shoulders, knees, and elbows to great compressive stresses and other types of forces, Mow pointed out. A deep knee bend, for example, can lead to pressures in excess of 10 megapascal (about 1,500 psi) in an adult's hips and knees. Healthy joints can support the forces and slow grinding motions without pain or bone damage for several decades. But according to Mow, medical surveys show that arthritis and rheumatism are leading causes of disability among adults in the U.S. For that reason, Mow and coworkers have probed the details of joint lubrication and the causes of breakdown and failure.

In a human knee, the ends of the femur and tibia (the leg bones) are coated with a thin layer of cartilage (less than 7 mm thick) and separated from one another by a thin film of synovial fluid (less than 50 m thick). The fluid is produced by a metabolically active tissue known as the synovium, which encloses the joint and, together with the bones and cartilage, provides a smooth bearing system, Mow explained.

In knees and other healthy animal joints, coefficients of friction are remarkably low compared with common engineering materials, Mow pointed out. To study animal joints, researchers use customized tribometers to quantify frictional forces between sliding surfaces. One type of device is based on a pendulum, in which the joint being studied serves as the fulcrum. In that setup, friction is measured by monitoring the decay of the pendulum amplitude.

According to Mow, cartilage in joints (articular cartilage) is composed of a collagen-proteoglycan matrix. The porous and permeable material retains water and dissolved electrolytes in interstitial sites of the matrix. Synovial fluid contains hyaluronic acid, which is an unbranched macromolecule composed of glucuronic acid linked with N-acetylglucosamine. In addition to other components, synovial fluid contains glycoproteins, which aid in lubrication, and phospholipids, which provide wear protection.

As loads are applied to a joint, while running or jumping, for example, the cartilage minimizes friction and wear of the bearing surfaces by spreading the load over a large area. Cartilage begins to fail when the body's natural mechanisms for repairing microscale damage can no longer keep pace with the cartilage wear rate, the Columbia researcher explained. Using photos and micrographs to compare a healthy dissected knee joint with a diseased one, Mow pointed out fissures, gaps, and overall thinning of the cartilage in a patient who had suffered from advanced arthritis.

Assessing the detailed condition of cartilage in an intact knee (in a live patient) is a challenging problem. But computational techniques developed by Mow and others make it possible to prepare engineering-type models from X-ray images of joints. The models enable exact shapes of joint surfaces and the thickness of cartilage to be determined precisely. You can't get that level of detail from X-ray or MRI [magnetic resonance imaging] images, Mow declared. The information can be used in several ways-for example, to advise patients in a rehabilitation program about the most effective angles and loads that should be used in knee-flexing exercises.

A key outcome of Mow's work-based on models that address the pressurization and flow of fluids in the cartilage matrix-is that in healthy joints, 95% of the load is supported by fluid at the cartilage tissue surface. Elaborating on that finding, Mow explained that, because the proteoglycan portion of cartilage is highly negatively charged, the tissue is hydrophilic and attracts a high concentration of freely mobile sodium ions. The ions, in turn, generate an osmotic pressure that draws water into the cartilage and prevents it from being totally squeezed out while being compressed by an applied load.

It's a wear-protection mechanism that enables the tissue to retain water while supporting 95% of the load, Mow said. It's no wonder the coefficient of friction is so small, he added with a chuckle. In effect, we're all walking on water.

Another area of interest for tribologists is microelectromechanical systems (MEMS), tiny machines with micrometer-sized gears, actuators, and other moving parts. Commercial examples of MEMS devices include automobile airbag accelerometers and video-display semiconductor chips that contain roughly 1 million fast-moving micromirrors. To keep lab-on-a-chip and other future MEMS devices working properly, or working at all, for that matter, the microscopic parts will need to move smoothly. Enter tribology.

Tribology plays a big role in preparing MEMS devices and using them, according to Kathryn J. Wahl, a senior scientist at the Naval Research Laboratory, Washington, D.C. If micrometer-sized machine components stick together after fabrication or rub against nearby surfaces in an unintended way when the device is switched on, it will have a short lifetime. Wahl and Nicholas D. Spencer, a professor of surface science and technology at the Swiss Federal Institute of Technology (ETH), Zurich, organized a symposium that focused on tribological issues associated with MEMS devices.

Michael T. Dugger, a researcher at Sandia National Laboratories, reported on surface treatments for controlling friction, adhesion, and other device properties. Dugger and coworkers conduct fundamental studies that may lead to various applications, such as microelectromechanical locks-miniature safety switches that can be used to prevent weapons systems from being armed accidentally or without authorization. These devices may sit in storage for years, Dugger said. We need to make sure they're chemically stable and work reliably when they're needed.

Using micromachine tribometers as test devices, the Sandia team compared various coating procedures to evaluate the level of protection they provide. The group found that a commonly used solution-phase method for depositing hydrophobic trichlorosilane monolayers left critical but hard-to-reach areas unprotected. A vapor-phase method for depositing an aminosilane compound was more effective, Dugger reported. They also observed that as little as 500 ppm of water in a hermetically sealed package is sufficient to hydrolyze the protective films and cause an increase in friction.

Switching to solids, Dugger noted that a thin film of metallic tungsten grown from WF₆ decomposition provides wear protection due to the film's hardness but lowers the strength of the polycrystalline silicon device parts. Atomic-layer deposition of WS₂, however, is one of the most promising methods, Dugger remarked, because the technique effectively delivers the lubricant to complex, buried, sliding surfaces.

Understanding the relation between the molecular architecture of protective monolayers and the films' tribological properties is one of Robert W. Carpick's research goals. Carpick, an associate professor of engineering physics at the University of Wisconsin, Madison, teamed up with Sandia staff scientist Maarten P. de Boer to compare friction measurements made with a MEMS-type tribometer (a nanotractor) with results from an atomic force microscope (AFM).

The nanotractor tests show that the friction coefficient measured between silicon surfaces coated with a fluorinated aminosilane monolayer is three times greater than the value obtained when using a hydrogenated trichlorosilane film. Despite a difference in scale of several orders of magnitude, the AFM study provided similar results, Carpick said. Now the team is working on experiments to uncover the molecular basis of the friction differences.

In collaboration with Wisconsin physics professor Gelsomina DeStasio, the group has recently developed a photoemission electron microscopy method for mapping the chemical composition of wear tracks (worn areas) with nanometer resolution. Initial results from a silicon sample coated with a fluorinated film show a decrease in fluorine and an increase in oxygen in the wear track compared with the surrounding area.

From powerful automobile engines to nimble micromachines and animal joints, tribological properties control the mobility of a vast number of moving parts. By deepening our understanding of the molecular basis of friction, lubrication, and wear, tribologists ensure those parts continue to move smoothly and avoid rubbing each other the wrong way.

The next meeting of the World Tribology Congress is scheduled to be held in Kyoto in 2009.