Symposium CP05—Materials Evolution in Dry Friction---Microstructural, Chemical and Environmental Effects

An important, and also sometimes controversial issue in nanotribology is how does a microstructure evolve? Is it purely stochastic, or are there some systematics underlying the science which occur naturally and/or can be exploited to reduce friction and wear. Over the last decade there has been a slow increase in the use of in-situ sliding within transmission electron microscopes. While these instruments normally are less sophisticated than dedicated scanning probe instruments or aberration-corrected microscopes, by combing the two one can obtain information about what matters, what does not which is often obscured by the buried interface in less direct experiments. This talk will focus on some older and some new results from in-situ observations inside electron microscopes. We find that the evolution is not purely stochastic, but instead involves fracture of weak links and the reduction of high friction local structures.

In probe-based microscopy, nanomanufacturing, and small-scale devices, the performance and reliability often depend on the size and nature of contacts between nanoscale bodies. However, conventional models of contact based on continuum mechanics rely on assumptions that may not hold at the nanometer length scale. Here we used a combination of in situ experiments and atomistic simulations to quantitatively investigate nanoscale asperities during formation, loading, and separation of contact. Experimentally, contact tests were performed inside of a transmission electron microscope, with in situ measurements of load and contact properties taken simultaneously with high-resolution characterization of the materials structure and geometry. In simulations, molecular dynamics modeling was performed on identical nanocontacts that were precisely matched in terms of materials, geometry, and loading conditions. The results demonstrate that important contact properties such as adhesion, contact size, and electrical current flow deviate significantly from the predictions of conventional models. The physical origin of the deviations is shown to be chemical and mechanical interactions at the atomic scale.

In absence of physically-based constitutive relations describing mechanisms of abrasive wear, design of wear-resistant materials has been driven either by (i) trial-and-error based alloy screening approaches, or (ii) following the Archard equation that linearly relates wear resistance to the hardness of the annealed material. There are, however, observations in literature that demonstrate that hardness is not the only factor controlling abrasive wear response. In an effort to unravel the key constitutive relations, a systematic study on the role of crystal structure and orientation has been performed. Grains with specific orientations of IF steel and pure nickel, heat treated to obtain similar micro-hardness values, have been selected using EBSD to perform in-situ SEM scratch tests with the same applied force. Focused Ion Beam has been used in order to confirm that the dry friction tests were conducted in ploughing mode, and Atomic Force Microscopy as a way to quantify wear from the surface profile. Comparing the results for the two different crystal structures and for grains with different crystal orientations enabled the identification of guidelines for design of advanced wear-resistant materials.

Atomic friction is imperative to the performances and reliability of micro-/nano-electromechanical systems and magnetic storage devices. Nevertheless, the atomic mechanisms underlying friction are still elusive, primarily due to a lack of direct observation on the contacting interface during the process. Here, by combining in situ atomic-scale transmission electron microscopy (TEM) and atomic force microscopy (AFM), we discovered diffusion-mediated formation of an interfacial layer between two metallic asperities under trivial normal forces, which was responsible for ultra-low friction. In addition, the atomic friction mechanism was found to depend on the mutual orientation of the contacted crystals. The work demonstrates fundamental insights and a new avenue for atomic friction.

New insights into friction and lubrication from atomic force microscopy (AFM) are presented. First, nanocontacts with 2-dimensional materials including graphene and MoS2 are discussed. Friction between nanoscale tips and 2D materials has been repeatedly shown to depend on the number of layers. An initial model attributing this to puckering [1] is now enhanced by molecular dynamics (MD) simulations showing a strong role of energy barriers due to interfacial pinning and commensurability [2]. New results where 2D materials are used on both the tip and the sample will also be presented. Then, nanoscale asperity-on-asperity sliding experiments conducted using a nanoindentation apparatus inside a transmission electron microscope will be discussed. This approach is used to obtain atomic-scale resolution of contact formation, sliding, and adhesive separation of two silicon nanoasperities. Forming and separating the contacts without sliding revealed small adhesion forces; separating during sliding resulted in a nearly 20 times increase. These effects were repeatable multiple times, and increased with both the sliding speed and the applied load used. We attribute this surprising sliding-dependent adhesion to the removal of passivating terminal species (likely hydrogen atoms) from the surfaces, followed by re-adsorption of these species after separating the surfaces. This helps explain mechanisms of nanoscale dry adhesion, and demonstrates that adhesion can be reversibly varied, which offers potential applications for nanoscale manipulation.

[1] C. Lee et al. Frictional Characteristics of Atomically-Thin Sheets. Science, 328, 76 (2010).
[2] S. Li et al. The Evolving Quality of Frictional Contact with Graphene. Nature 539, 541 (2016).

We present a model that predicts the macroscale temperature-dependent interfacial shear strength of 2D materials like MoS₂ based on atomistic mechanisms and energetic barriers to sliding. Atomistic simulations were used to systematically determine the lamellar size-dependent rotation and translation energy barriers, that were used to accurately predict a broad range of experimental data. This framework provides insight about the origins of characteristic shear strengths of 2D materials.

There has been a sustained interest in 2D and layered materials for use as solid lubricants due to their intrinsically low friction, and in advanced electronics applications including flexible electronics due to their intrinsic bendability and novel electronic properties. It is known that both temperature and sliding speed affect friction, but there is little understanding about the relative importance of these effects. We use matched ultrahigh vacuum atomic force microscopy (AFM) experiments and molecular dynamics (MD) simulations to study the atomic stick slip friction for nanoscale Si₃N₄ and Si tips sliding on monolayer and multilayer MoS₂ from cryogenic to elevated temperatures, and over several decades of sliding speed. We observe that friction increases as temperature decreases and as speed increase. In agreement with the qualitative trends predicted by the thermal Prandtl-Tomlinson model, friction shows an approximately logarithmic increase with increasing sliding speed. In some cases, we observe dramatically decreases in friction as temperature increases, suggesting the effective role of thermolubricity. Similar trends are seen for monolayer and multilayer MoS₂. However, the temperature dependence differs between different tip materials, potentially due to the difference in thermal conductivity of the tip and across the tip-sample interface. These new findings have important implications for macroscale and nanoscale solid lubrication, and for applications using layered and 2D materials including the assembly and functionality of flexible electronics.

Environmental contaminations are known to affect the use of nanoscale materials in devices. In this contribution, we show that friction in graphene is significantly altered by the presence of contamination that comes from air exposure. The influence of contamination in tribological properties was studied by friction force microscopy. It was observed that friction in graphene increases due to surface contamination, decreasing the difference in friction between mono and bilayer graphene. The physisorption of contaminants was observed to begin at the edges or mechanical defects in graphene monolayers. Time series of friction force images enable us to observe the kinetics of this contamination spreading over the surface.

A major focus of tribology research has been to understand the origins of practical-scale multi-asperity friction by dissecting it down to nano-scale single-asperity contacts. The Prandtl-Tomlinson model with thermal activation (PTT) has laid the theoretical foundations for understanding friction of single-asperity contacts. Experimentally, both molecular dynamics and friction force microscopy have agreed with PTT model’s prediction that friction increases logarithmically with velocity. Multi-asperity contacts’ velocity dependence, on the other hand, is much more challenging to describe theoretically and verify experimentally. In this presentation, we show an experimental setup that consists of a microtribometer that simultaneously uses two lateral motion devices with vastly different characteristic velocities: a piezo stage (<0.1 mm/s) and a quartz crystal microbalance (QCM) (up to 1 m/s). We also have two independent opportunities to measure the frictional responses of the contact: using either the capacitance sensor-based microtribometer load cell or the QCM network analyzer that tracks the change in resonant frequency and quality factor of the crystal’s oscillations. When run independently, the two techniques produce similar results: μ_gold = 0.2 and μ_MoS2 = 0.05. The integration of the two techniques, however, produced an unexpected set of results indicating that the QCM’s four orders of magnitude faster motion had a significant effect on the friction measured by the microtribometer. When the quartz crystal was placed between the tribometer’s lateral piezo stage and the contact, significant decrease in the microtribometer’s kinetic friction signal was observed: an order of magnitude for gold and a factor of two for single-crystal MoS₂. At the same time, the microtribometer continued measuring high static friction coefficient and strong adhesion. These microtribometry results are complemented by information obtained from QCM data analysis such as contact area, shear strength and nanoscale static friction. Pairing two tools with such vastly different characteristic time scales enables our results to be sensitive to inherently atomic-scale phenomena, such as adhesion and contact aging, in a macro-scale contact.

Dry friction is commonly desired in extreme environments where traditional lubrication solutions such as oils and greases are precluded. Polytetrafluoroethylene is an excellent candidate material for dry sliding applications due to its low friction coefficient, high resistance to chemical attack, vacuum compatibility, and wide range of operating temperatures. Although it suffers from a notoriously high wear rate, k~7 x 10^-4 mm^3/(Nm), filler materials from microparticles to fibers provide modest to moderate improvements in wear resistance, typically by one to two orders of magnitude. The remarkable tribological behavior of PTFE filled with a few volume percent of a particular alpha-phase alumina filler (k ~ 1 x 10^-7 mm^3/(Nm)) has attracted significant scientific interest over the past decade, yet the mechanisms responsible for the over 5,000x improvement in wear rate compared to unfilled PTFE remain elusive. The formation of a stable transfer film is largely believed to be a critical step towards attaining ultralow wear in sliding against metal countersurfaces; however, the specific tribochemical mechanisms of transfer film formation have only very recently been investigated. In this study, a “stripe” experiment permitted spectroscopic analyses of PTFE/alumina transfer films in various stages of development. This study led to a proposed mechanism that details the formation of chemically-distinct films on both surfaces of the tribological pair. PTFE chains break due to the mechanical stresses generated during sliding and, in the presence of oxygen and water in the ambient environment, produce carboxylic acid end groups, which chelate to both the surface of the alumina filler and the exposed metal countersurface and nucleate the formation of the transfer film. These tribochemical reactions form a thin and robust polymer-on-polymer tribological system that after the initial run-in period can endure hundreds of thousands of reciprocating sliding cycles with virtually no polymer composite wear. The Hamaker model for van der Waals interactions between polymer fibrils and the countersurface provides mathematical support for the mechanical scission of carbon–carbon bonds in the backbone of PTFE under sliding contact conditions. Model experiments utilizing small molecules have confirmed the proposed reactions.

INTRODUCTION: Though it has similar mechanical, thermal, and chemical properties as polytetrafluoroethylene (PTFE), perfluoroalkoxy polymer (PFA) has rarely been evaluated in tribological (friction and wear) applications. In addition to its unique physical, thermal, and chemical properties, PFA can be melt-processed instead of the sintering-based methods of PTFE. This allows filler particles or fibers to be melt mixed with PFA and this mixture can be used to produce complex geometries via screw-injection molding.
The wear rate of PFA was shown to improve nearly four orders of magnitude with the addition of 5-10 wt. % of α-alumina filler particles, highlighting the need for a better understanding of the fundamental aspects of how shear stress may affect the microstructure and chemistry of PFA. The objective of this work is to link the chemical and microstructural observations of unfilled PFA the observations seen in PTFE-alumina and PFA-alumina composite systems.
METHODS: Differential scanning calorimetry (DSC), x-ray diffraction (XRD), transmission infrared spectroscopy, and attenuated total reflectance fourier transform infrared spectroscopy (ATR-FTIR) were used on bulk and worn samples of unfilled PFA and PFA-alumina composites were slid against stainless steel counterfaces at 6.25 MPa of applied pressure and a sliding velocity of 50 mm/s.
RESULTS & DISCUSSION: Tribological experiments revealed a 10,000x improvement in wear of PFA-alumina composites (K~4.0x10^-8mm³/Nm) compared to unfilled PFA (K~4.0x10^-4 mm³/Nm). Additionally, PFA-alumina composites were found to have similar friction (µ~0.25) which is surprising considering unfilled PTFE (µ~0.12 has a friction coefficient nearly 2x lower than that of unfilled PFA (µ~0.24). Wear properties of PTFE-alumina composites (K~4.4x10^-8 mm³/Nm) matched those observed in filled PTFE composites. The reduction in wear rate for PTFE-alumina and PFA-alumina composites has been attributed to tribochemical reactions of broken fluoropolymer chains, environmental constituents, and the metallic countersurface/aluminum oxide filler material. These reactions promote the development of running films and transfer films that protect the surface of the polymer composite and steel counterface.
To better understand this tribochemical mechanism maybe different in PFA compared to PTFE, the effects of sliding on the microstructure and chemical makeup of PFA were investigated. From observations of DSC endotherms, PFA wear debris had higher crystallinity compared to bulk PFA. Additionally, XRD of unworn and worn PFA showed an increase in crystallite size for the wear debris. Both these observations point to microstructural rearrangement of the PFA backbone due to stress applied during sliding, which have also been reported in the wear of unfilled PTFE and drawing of PFA films. Higher concentrations of PAVE comonomer were seen in ATR-IR spectra of PFA wear debris, suggesting preferential crack propagation through the amorphous region of the microstructure of PFA during sliding. Similar repeated loading during MIT flex life experiments on PFA showed that PFA samples with lower crystallinity of PFA (higher % amorphous region) exhibited improved fatigue performance, suggesting that cracks preferentially travel through the amorphous region of PFA. Transmission IR spectra of PFA wear debris revealed the presence of free carboxylic acid endgroups that were not present in the bulk PFA surface. The presence of free carboxylic acids is due to the breaking of C-C bonds in the PFA backbone reacting with environment constituents. The formation of free-carboxylic acids endgroups is a critical intermediate step of the tribochemical reaction necessary to form robust transfer films and running films proposed for PTFE-alumina composites.

This presentation focuses on recently proposed fundamental, predictive correlations between microstructural evolution, materials properties, and the shear strength – and thus friction coefficient – demonstrated for FCC metals. We discuss recent efforts to expand these ideas to BCC metals and alloys, and provide highlights of how this work has impacted the development of innovative materials solutions for a wide range of commercial and industrial applications. Examples include the development of higher efficiency renewable energy power generation systems (wind turbines), longer life electrical contacts for consumer electronics, higher reliability slip-rings for satellites and other aerospace systems, and new prospects for achieving practical superlubricity via in situ tribochemistry.

This work was funded by the Laboratory Directed Research and Development program at Sandia National Laboratories, a multi-mission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

During sliding of surfaces the near surfaces undergo significant changes in terms of topography, composition and microstructure and a so-called “third body” or “tribomaterial” forms which differs strongly from the bulk materials in terms of topography, composition and microstructure. This has been studied in a series of experiments using different metals sliding against metal, ceramics or coatings. In this work we use multilayer model alloys of an Au/Ni layer system to study effects of grain size on steady-stady friction by varying the layer spacing and the number of layers. Using microtribometer under ultra-high-vacuum conditions, we performed friction experiments sliding with a ruby sphere against Au/Ni multilayers. We find that the friction strongly depends on the layer spacing, and we are able to link the friction to the tribologically induced tribomaterial as found by FIB and TEM/EDS.

Shear force applied to metallic materials can induce deformation and microstructural changes which can be used to process materials in solid state, helping avoid melt processing and in achieving highly refiled microstructures not achievable by conventional processing. In order to develop such solid phase processing methods for metallic alloys, we aim to better understand the fundamental atomic scale mechanisms of mass and energy transfer in materials under shear deformation. Formation of metastable interfacial states in metallic multilayer thin film systems with different miscibility between individual layers under shear deformation, unique deformation mechanisms and accelerated diffusion observed only under high shear deformation etc are crucial factors to be understood. We employ a pin-on-disk tribometer deformation of various thin film stacks based on Cu and Al systems followed by detailed microstructural characterization before and after shear deformation by transmission electron microscopy, atomic force microscopy and atom probe tomography to understand the structural and compositional changes near atomic scale resolution. Our results on structural and chemical modifications at interfaces and individual layers of thin films after shear deformation versus thermal annealing provide new insights on the unique role of shear deformation on modifying the phase transformation pathways of these alloy systems. These new insights will be presented and compared with what is currently understood about role of shear deformation on generating super saturated solid solution states or self-organized nanolayers in immiscible metallic alloys.

Molecular dynamics simulations of diamond-cubic silicon and carbon under combined shear and compression show the formation of an amorphous solid with liquidlike structure at room temperature. Consistent with the opposite density changes of the two crystals upon melting, the amorphous material is denser than the crystal in silicon and less dense than the crystal in carbon. As a result, its rate of formation is enhanced by pressure in silicon but suppressed in carbon. These results are particularly unexpected for silicon, whose amorphous structure is supposed to be liquidlike (high-density amorphous Si, HDA-Si) only when hydrostatically compressed above the polyamorphic transition pressure (∼14 GPa). Below this pressure, amorphous silicon is expected to have a low-density structure (LDA-Si) with density close to that of the diamond-cubic crystal. Our simulations show that this polyamorphic transition disappears under shear and high-density, liquidlike amorphous silicon with metallic ductility forms even at low pressure. These results are potentially transferable to other diamond-cubic crystals, like germanium and ice I_h, and provide insights into nonequilibrium materials transformations that govern friction and wear in tribological systems.

Wear of metals in dry sliding contacts is a dynamic process that leads to adhesive transfer, material mixing and oxidation. All of these phenomena, along with the mechanical deformation of the near-surface, lead to the formation of tribofilms, which are also called mechanically mixed layers. Microstructural and chemical modification of tribofilms lead to significantly different properties from the bulk material that is subjected to dry sliding wear. In some instances, tribofilms can be stable and resist wear and in other cases, there can be persistent wear flow related to tribofilm instability.

Recent advances in nanomechanical testing has made it possible to study mechanical properties of small volumes of materials. Tribofilms present an interesting and challenging vista for nanomechanical testing. In recent years, researchers have used nanoindentation and pillar compression testing to determine the properties of tribofilms. While a limited number of studies of this type exist, and measurements are difficult to quantify, there is clearly a trend with tribofilm properties and the wear resistance of a metals and metal-matrix composites (MMCs). In this study, nanoindentation testing of tribofilms revealed that for a wide range of metals (Al, Cu, Ti, Ni) and their MMCs (Al-Al₂O₃, Ti-TiC, Cu-MoS₂ and Ni-WC) that there is a trend of decreasing wear rate with tribofilm hardness. Metals and MMCs are tested in sliding wear and sometimes fretting wear test conditions. For both cases, post-characterization of cross-sectioned wear scars reveals microstructural evolution near surface leading to formation of tribofilms that provide wear resistance and friction control. Structure and properties of the tribofilms are determined with SEM, TEM, EDS, Raman spectroscopy and nanoindentation. Generally, tribofilms are found to be mixtures of the two components in the MMC, but with finer microstructure and some level of oxidation that leads to higher hardness. Using a wider range of test conditions (e.g. sliding speed, normal load), one can establish additional relationships between tribofilm properties and wear resistance. Furthermore, using recently developed techniques for toughness measurements using nanomechanical methods, a better picture of the overall mechanical properties required for tribofilm stability and wear resistance can be established.

One of the critical challenges in designing materials with superior tribological properties is that at present there is limited understanding of how the contact interface and the microstructures of contacting materials evolve during sliding. This evolution may include chemical reactions at the buried interface, grain growth and refinement, evolution of dislocation networks, interaction of dislocations with interfaces, chemical mixing etc. All these phenomena can contribute to energy dissipation (friction) and they can affect the dominant type and amount of wear. Changes taking place in the near-interfacial regions of the contacting materials are difficult to monitor experimentally in operando because imagining techniques for buried interfaces under mechanical loads are still in their infancy. Computer simulations provide a powerful approach to investigating interfacial evolution and to discovering phenomena that control friction and wear. Here, I will discuss our developments of a multiscale simulation methodology to determine how materials evolve in sliding contacts.

In the first part of the talk I will focus on chemical evolution of frictional interfaces, a phenomenon referred to as chemical aging. Specifically, I will discuss the combined effects of chemical evolution of interfaces and of surface roughness on the origin of static friction, which is a highly debated topic in the field of tribology. Our approach combines density functional theory calculations of chemical reactions at interfaces, kinetic Monte Carlo simulations that extend simulation time scales to those that can be accessed in experiments, and boundary element method to extend simulation length scale and investigate multi-asperity contacts. For most solid/solid contacts, static friction increases logarithmically with time, a phenomenon known as aging. We discovered molecular mechanisms that are responsible for aging of silica/silica contacts in the presence of water and that are based purely on interfacial chemistry. By combing simulations with atomic force microscopy (AFM) experiments we identified the effects of contact pressure on both friction and aging, showing that aging is accelerated when the contact pressure is increased. Our model shows a quantitative agreement with AFM experiments on load-dependence of aging in single-asperity contacts. In addition, our model predicts that aging is a non-monotonic function of temperature and therefore the temperature can be used to control this phenomenon. Finally, we make predictions of how chemical aging will scale to large rough surfaces and how it will depend on the nature of the contact roughness.

In the second part of the talk I will discuss our developments of multiscale models of microstructural evolution of polycrystalline metals subject to mechanical loads. We combine molecular dynamics simulations with crystal plasticity and phase field models to determine microstructural evolution on experimentally relevant time scales. I will also discuss specific predictions from our models regarding the effects of grain size on mechanical response of metal alloys. For instance, we have discovered an intermetallic alloy that can accommodate large plastic strain without help of dislocations due to formation of thin amorphous shear bands during deformation. This finding has important consequences for mechanical response because such intermetallics do not exhibit the classical Hall-Petch grain-size strengthening.

This paper reports our attempt to prepare icosahedral quasicrystalline nanoparticles (IQNP). It is well-known that, at the nanometer order, the properties of materials depend on their size. For example, the emission spectrum of quantum dots can be simply tuned by adjusting their size. Although nanoparticles of crystalline and amorphous structures have been reported, nanoparticles of icosahedral quasicrystalline structure, which combines a novel structure with nanometer size, have not.
It is believed that a mechanical approach is the easiest way to prepare IQNP. In our study, we proposed to grind alloy mechanically in to nanoparticles followed by size sorting using filtering. An alloy of Al₆₀Cu₂₅Fe₁₅ composition was used to experimentally establish this procedure. Alloy of this composition is closely related to the formation of pure icosahedral phase, although the alloy used in the present study is not of pure icosahedral phase. The alloy was grinded for 40 hours and the existence of nanoparticles, particles smaller than 100 nanometers, was confirmed using Scanning Electron Microscopy (SEM). Such particles were suspended in methanol, forming a solution of 1 milligram per milliliter (1 mg/mL). This solution was then passed through a syringe filter. The passed solution was expected to contain particles with sizes smaller than the filter pore size and was subsequently passed through another filter with a smaller pore size. This filtering process began with filters of 5µm pore size, and progressed through 1 µm, and down to 0.45 µm. The volume of the raw solution started with 1 mL and was maintained to this volume by adding extra methanol at different stages. SEM specimens were prepared by dispersing approximately 10 µL of the through-solution onto a silicon wafer and allowing to dry.
We were successful in filtering out nanoparticles, indicating that our procedure has been established, and is ready to be applied to an alloy of Al-Pd-Mn pure icosahedral quasicrystalline phase. During this process, we encountered the following problems and were able to establish effective solutions to overcome them.
We first began with hydrophobic polypropylene syringe filters. Upon examination of the specimen prepared with the flow through, it was frequently found that the desired particles were buried within unknown objects. The situation was worse with filters of a smaller pore size. The morphology of the objects resembled that of the filter fibers. This speculation was confirmed by Energy Dispersive X-ray Spectroscopy (EDS) analysis. The flow through of filter fibers is believed to be related to the presence of alloy particles in our solution, since repeated experiments with pure methanol did not cause this effect. With this knowledge, we switched to a stronger filter made of glass fiber. We found that the glass filter had little to no filter fiber flow through.
The second problem we encountered is that the alloy particles obtained through filtering were mixed with salt (NaCl), which affected the examination of the nanoparticles by SEM. The effect became worse with filters of smaller pore size. We speculate that the salt is from the filter since salt was not observed on SEM specimens prepared from the raw solution prior to filtration. Several efforts were made to eliminate the salt. First, the filters were rinsed before use, although this turned out to also affect the integrity of the filters. Second, we scratched the silicon wafers with diamond paper and placed the particles on this relatively rough surface. Then we rinsed the sample with water with the aim of dissociating the salt, leaving the nanoparticles in the grooves. This was successful in removing almost all of the salt and allowing us to get a clear view of the desired particles.

Nanomaterials have great potential as lubricating materials and for the development of advanced lubrication technologies. Moreover, applied electric field is considered to be of great interest and importance for controlling friction in systems with nanoparticle additives. Copper sulfide nanoparticles can be synthesized both as single nanoplates and nanoplate arrays single by using single source, single surfactant technique and their surfactant coating can be re-organized in different structures by post synthesis thermal treatments.¹ Surfactant coated CuS nanoparticle additives dispersed in diethyl succinate lubricant were studied in friction experiments under applied electric potential from -30 V to +30 V. A remarkable decrease in the coefficient of friction was observed upon application of an electric field perpendicular to the shearing interface. Friction tests indicate that the nanoparticles can reduce the friction coefficient of diethyl succinate lubricant by ~25%, from 0.24 to 0.18. When a negative potential lower than -14 V was applied, the friction coefficient markedly decreased to 0.05. Finally, different models were used for adsorption-settlement kinetics of nanoparticles implied that both the adsorption and settlement of CuS nanoparticles play an important role in the mass changes on the solid surface in the suspension of the nanoparticles. According to this research, shape controlled nanoscale lubricant additives are likely to have profound technological importance in reducing and controlling friction in engineering systems.

1. Rabkin A., Friedman O. & Golan Y. Surface plasmon resonance in surfactant coated copper sulfide nanoparticles: Role of the structure of the capping agent. J. Colloid. Interf. Sci. 457, 43-51 (2015).

The energy dissipated in a metallic frictional contact is largely converted into plastic deformation. The subsurface material is thereby exposed to extreme amounts of shear deformation and often forms layered subsurface microstructures with reduced grain size. We analyze the early stages of these processes with the aim of elucidating the elementary mechanisms for the formation of the subsurface microstructures by systematic model experiments and Discrete Dislocation Dynamics simulations. The simulations show how preexisting dislocations transform into dislocation structures within the stress field underneath a moving spherical contact. The crystallographic orientation is decisive for the formation of these structures and for dislocation transport with the moving contact[1]. Experimentally, we find a localized dislocation structure at a depth of ~100-150 nm already after the first loading pass[2]. This dislocation structure is shown to be connected to the inhomogeneous stress field under the moving contact. Its generation and evolution under different conditions of load and loading rate are investigated. All subsequent microstructural transformations and the mechanical properties of the surface layer to some extent depend on this structure.

[1] J. Gagel, et al., Acta Mater. 156 (2018) 215
[2] C. Greiner, et al. Scripta Mater. 153 (2018) 63

Frictional sliding induces a distinct discontinuity in the microstructure between a subsurface layer and the underlying bulk material, which strongly influences the tribological performance. Here, the strain rate and temperature dependence of such tribologically induced microstructure evolution was systematically investigated during reciprocating sliding of copper. It was found that an increase in strain rate and a decrease in temperature each result in a transition in the dominating deformation mechanism from dislocation slip to twinning-mediated plasticity at the very beginning of sliding. A sequence of deformation mechanisms was revealed under high rate and/or cryogenic sliding (strain rate ∼ 10⁴ s⁻¹; liquid nitrogen temperature): First, nanoscale dislocation trace lines form beneath the surface during the first forward pass; Second, partial dislocation nucleation from the sliding surface accompanied by nano-twinning and abundant stacking faults in the backward pass; Third, formation of a nanocrystalline layer upon further sliding. Sliding induced surface roughening is found to assist partial dislocation nucleation from the surface during high rate and cryogenic sliding. Our results suggest that the sliding surface can act as an effective source of dislocations to initiate and accommodate associated plastic deformation, which may be explored to model the microstructure evolution during sliding.

Microstructural evolution induced by plastic deformation during dry sliding and frictional contact is quantitatively examined via electron microscopy over several length scales. Individual dislocations, medium to high angle and dislocation boundaries, crystal orientation/texture, as well as local changes in chemical composition were measured as a function of depth from the sliding surface. These quantified structures are analyzed within a universal framework developed for metals and alloys undergoing different modes of plastic deformation. New and prior statistical scaling laws are used to evaluate structural evolution and the gradient in microstructural size scale that develops with increasing depth from the surface. In the near surface regions crystallite sizes with an average of 5 nm and a minimum of 2nm are observed. This average size scale coarsens with increasing depth and is measured as a function of depth. A new scaling law connects the microstructural evolution of two measured structural parameters. In turn those structural parameters are related to strength parameters that estimate the flow stress utilizing a linear addition of the classical Taylor and Hall-Petch formulations. The gradient in flow stress with depth is used to evaluate the overlapping stress fields created by sliding deformation.

Mechanisms that promote this refinement and counteract structural coarsening are examined. Solutes and dislocations, in addition to the applied stress field and deformation temperature, are all important elements in structural refinement. Solutes may be present both in the original alloy or introduced via mechanical mixing during deformation. Their presence is important to suppress dynamic recrystallization. In the absence of dynamic recrystallization, coarsening occurs by the migration of sharp triple junctions between deformation-induced boundaries that enclose lamellar regions. Attached dislocations and solutes can suppress junction migration via pinning mechanisms. Solutes hinder migration via both their presence within high angle boundaries and within the matrix as they stabilize dislocations and low angle dislocation boundaries. The observed and continuous presence of dislocations is a key point. Dislocation mechanisms remain the dominant deformation mechanism even below 5 nm.

D. A. Hughes, N. Hansen, “The microstructural origin of work hardening stages,” Acta Materialia, 148, (2018), 374-383.
D. A. Hughes, N. Hansen, “Exploring the limit of dislocation based plasticity in nanostructured metals,” Physical review letters, 112, (2014), 135504.

Supercomputer K was constructed in 2011 in Kobe, Japan. Supercomputer K contains over 80,000 nodes and 700,000 cores and its calculation speed is 10¹⁶ flops. Then, it won a championship in the supercomputer world ranking in June, 2011. However, in the supercomputer world ranking in June, 2018, supercomputer K ranks sixteenth. The ministry of education, culture, sports, science, and technology in Japan plans to construct new supercomputer post-K in 2021. The above ministry also started the “Post-K” projects for the development of Japanese-made simulation software on the supercomputer Post-K. Then, our proposed project “Challenge of Basic Science - Exploring Extremes through Multi-Physics and Multi-Scale Simulations” was adopted in June, 2016 and started in August, 2016. In this symposium, we introduce our supercomputer Post-K project and the recent outcomes of the tribology simulations by the supercomputer. The previous small-scale molecular dynamics simulations can clarify the super-low friction mechanism induced by the chemical reactions. However, the tribo-wear cannot be elucidated by the previous small-scale molecular dynamics simulations, because the tribo-wear simulations require huge amount of atoms. Super-large scale calculations by supercomputer and the molecular dynamics simulator optimized for supercomputer enable to simulate the tribo-wear dynamics induced by the chemical reactions. This is the paradigm-shift of the tribology simulation research. In the present presentation, we also discuss how the super-large scale simulations by the supercomputer “Post-K” will give the impacts on the tribology research.

We apply state-of-the-art methods used in materials design and computational physics/chemistry, such as multiscale quntum mechanics – molecular mechanics (QM/MM) dynamics simulations and high throughput calculations, for the first time to tribology. These innovative approaches allowed us to unravel fundamental mechanisms of friction and describe tribochemistry mechanisms in important solid lubricants. In particular:
i) We developed a computational protocol to calculate from first principles the work of adhesion, and the ideal hear strength of solid interfaces in a high throughput way.[1] By screening a wide series of interactions including metallic, covalent, and physical bonds, we derive a connection between the intrinsic tribological properties and the electronic properties solid interfaces. In particular, we show that the adhesion and frictional forces are dictated by the electronic charge redistribution occurring due to the relative displacements of the two surfaces in contact. This suggests unconventional ways of measuring friction by recording the evolution of the interfacial electronic charge during sliding. Finally, we explain that the key mechanism to reduce adhesive friction is to inhibit the charge flow at the interface and provide examples of this mechanism in common lubricant additives. [2]
ii) We adopted a quantum Mechanics/Molecular Mechanics (QM/MM) approach to simulate a tribological interface. This method makes computationally feasible to effectively simulate systems with the size required to properly model chemical reactions occurring in tribological conditions. The comparison with ab initio (Car-Parrinello) and classical (using the ReaxFF force field) molecular dynamics calculations highlights the advantages of this novel approach. Using QM/MM, we accurately describe the tribochemistry processes reproducing the dissociation paths and rates of water molecules interacting with graphene ribbons within sliding graphene layers, a system which is relevant for technological applications.[3] Our simulations provide useful insights to understand the effects of humidity in graphitic systems, which I will discuss in comparison with MoS₂, another important solid lubricant, where humidity has opposite effects.[4]
[1] P. Restuccia, G. Levita, M. Wolloch, G. Losi, G. Fatti, M. Ferrario and M. C. Righi Ideal adhesive and shear strengths of solid interfaces: A high troughput ab initio approach In printing, Computational Materials Science.
[2] M. Wolloch, G. Levita, P. Restuccia, and M. C. Righi, Ideal adhesive and shear strengths of solid interfaces: A high throughput ab initio approach Phys. Rev. Lett. 121, 026804 (2018).
[3] P. Restuccia, M. Ferrario and M. C. Righi Quantum Mechanics/Molecular Mechanics (QM/MM) applied to tribology: real-time monitoring of tribochemical reactions of water at graphene edges, Submitted.

[4] G. Levita, and M. C. Righi, Effects of Water Intercalation and Tribochemistry on MoS2 Lubricity: An Ab Initio Molecular Dynamics Investigation, ChemPhysChem 18, 1475 (2017).

The discipline of tribology investigates the effects of mechanical forces during sliding, including energy dissipation, wear, and chemical reactions. Various models have been developed to describe these phenomena, but these apparently disparate models are based on the same fundamental concept; that an external force modifies the potential energy surface to decrease the energy barrier for the transition from one state of the system to another. In chemical systems, this causes an increase in the reaction rate. However, obtaining a detailed molecular understanding of the way in which an external force modifies reaction rates requires knowing the reaction pathway.
Strategies for measuring reaction pathways and their kinetics are illustrated using a model tribochemical reaction consisting of the gas-phase lubrication of copper by dimethyl disulfide. Two stress-activated elementary-step reactions are identified. The first is the tribochemical decomposition of adsorbed methyl thiolate species, that form rapidly by dimethyl disulfide reacting with copper, to form gas-phase hydrocarbons and chemisorbed sulfur. In a second stress-induced reaction, surface sulfur is transported into the subsurface region of the copper. This process causes an increase in oxidation state of copper, and the creation of vacant surface sites that allows the tribochemical reaction cycle to continue.
The rate constants for the elementary reaction steps can be measured and incorporated into a kinetic model for the gas-phase lubrication of copper by dimethyl disulfide that successfully describes the variation in friction coefficient as a function of the number of rubbing cycles on the surface. The model also reproduces measurements of the total amount of subsurface sulfur as a function of the number of cycles, as well as its depth distribution.
This fundamental understanding of the nature of the elementary-step processes enables their molecular origins to be investigated, both experimentally and theoretically using first-principles quantum calculations.

Tribo-oxidation is an often observed but far from fully understood phenomenon during friction and wear, taking place by tribochemical reactions of the tribo-partners or the surrounding medium. These reactions in most cases lead to a change in the surface properties such as the chemical composition and the mechanical properties and consequently result in a different friction and wear behavior. The aim of our research is to elucidate the elementary mechanisms of tribologically-induced oxidation by paring polycrystalline high-purity copper plates with sapphire spheres. The experiments are performed at room temperature in a strictly controlled atmosphere with reciprocating linear sliding under mild tribological loading.
Our results show the formation of amorphous oxygen-rich cuprous-oxide (Cu₂O) islands which are formed below the surface and grow to hemispherical amorphous/nanocrystalline cuprous-oxide clusters. We relate the growth of the clusters to the diffusion of oxygen on and in copper but the exact pathways of the oxygen are still elusive. After the islands have grown, they coalescence and form a continuous oxide layer on the surface, determining from then on the tribological properties of the contact. This oxidation process takes place at rates order of magnitudes faster than the native oxidation of copper in contact with the same environmental conditions. This works aims to understand the influence of the sliding velocity on the formation of these oxide clusters. We systematically vary the sliding speed from 0.1 to 5.0 mm/s and investigate the resulting microstructure. Scanning electron microscopy techniques are used in order to reveal the fundamental mechanisms of tribologically-induced oxidation. Once understood, this will help for tailoring the materials properties in order to achieve superior tribological performance.

This research project acknowledges funding by the European Research Council (ERC) under Grant No. 771237. Furthermore, it was possible through funding by the German Research Foundation under Project GR 4174/1.

Nickel-based alloys exhibit superior thermo-mechanical properties and are currently being considered for structural components of very-high-temperature gas-cooled reactors (VHTRs). During the operation of a VHTR, the structural components are exposed to extreme conditions including high temperatures and stresses; they are also susceptible to corrosion from moisture and impurity gases such carbon dioxide and methane. Fretting wear occurs in several VHTR components such as valve stems and seats, control rod drive mechanisms, and fuel handling mechanisms.
Recently, a set of in-situ fretting wear measurements have been carried out on a nickel-based superalloy – Inconel 617 [Ahmadi, A., F. Sadeghi and S. Shaffer (2018)] in helium and air environment at elevated temperatures. In this work, we perform microstructural analysis of the samples from these tests using confocal laser microscopy and scanning/transmission electron microscopy techniques. The surface roughness measurements show that small amount of oxides, which are more prominent at higher temperatures, inhibit wear damage. The elemental mapping using energy-dispersive x-ray spectroscopy (EDS) also reveal enhanced oxygen presence on the wear scars. We will finally describe the possible formation of complex carbides that can have a deleterious effect on the long-time integrity of the structural components.

References

Ahmadi, A., F. Sadeghi and S. Shaffer (2018). "In-situ friction and fretting wear measurements of Inconel 617 at elevated temperatures." Wear 410-411: 110-118.

Extraordinary tribological properties of diamond and tetrahedral amorphous carbon (ta-C) make them promising candidates for protective coatings in a variety of industrial products such as bearings and seals. Their ultra- (0.01 < μ < 0.1) and super-lubricity (μ < 0.01) have been observed under boundary lubrication with water and oxygen-containing organic friction modifiers. State-of-the-art spectroscopic measurements provided evidence for chemical and structural modifications of ultrathin carbon layers that play a role in establishing near-frictionless states. However, the tribochemical mechanisms behind the formation of superlubricious tribolayers have not been fully understood yet. Here, we will present recent results of atomistic tribology simulations of diamond and ta-C surfaces in contact with various lubricant molecules.

Large-scale quantum molecular dynamics simulations reveal five different friction regimes for water-lubricated diamond (111) surfaces [1]. While water starvation gives rise to cold-welding and amorphization of the tribological interface, traces of water are sufficient to preserve crystallinity. This can result either in high friction due to cold-welding via ether groups or in ultralow friction due to aromatic surface passivation triggered by shear-induced Pandey reconstruction. As increasing amounts of water molecules, dense H/OH surface passivation also provides another ultralow friction state. For other low-index diamond surfaces [i.e. (110) and (100)], three additional friction regimes involving ether and keto oxygen were found [2]. The surface-orientation-dependent friction regimes are rationalized by the structural and energetic peculiarities of the different low-index surfaces.

Experimentally observed superlubricity of ta-C under boundary lubrication with oleic acid and glycerol is explained by our atomistic model [1]. The presence of two reactive centers in the lubricants (e.g. carboxylic polar head and C=C double bond in oleic acid) allows them to concurrently chemisorb on both ta-C surfaces and bridge the tribogap. Sliding-induced mechanical strain triggers a cascade of molecular fragmentation reactions releasing passivating hydroxyl, keto, epoxy, hydrogen and olefinic groups. Similarly, glycerol’s three hydroxyl groups react simultaneously with both ta-C surfaces, causing the molecule’s complete mechano-chemical fragmentation and formation of aromatic passivation layers with superlow friction. Interestingly, our finding is not limited to oleic acid and glycerol, but applicable to other common friction modifiers.

[1] T. Kuwahara, G. Moras, M. Moseler, Phys. Rev. Lett. 119, 096101 (2017).
[2] T. Kuwahara, G. Moras, M. Moseler, Phys. Rev. Mater. 2, 073606 (2018).
[3] T. Kuwahara, P. A. Romero, S. Makowski, V. Weihnacht, G. Moras, M. Moseler, under review.

The triboelectric effect, which is defined as the charge transfer when two surfaces come into contact and slide against each other, is a well-established phenomenon underlying tribological systems ranging from a balloon rubbing on hair to triboelectric nanogenerators. While the phenomenon is well-established, the driving force for why there is charge transfer is less well-defined. A potential difference must exist between the sliding surfaces which will then drive the charge transfer as a reduction of free energy. The deformation induced by the microscopic asperities in these systems yields a strain gradient, which will produce a flexoelectric polarization. Using Hertzian contact mechanics and a single asperity model, flexoelectric polarization was determined as function of the asperity radius of curvature. The flexoelectric polarization of asperities ranging from 0.1 to 1 μm agree with typical magnitudes of triboelectric surface charge.

One of the most common examples of dry friction occurs when two materials in intimate contact with each other are peeled off. This frequently occurs in polymer materials used for replica molding or soft-lithography in which one material serves as a mold with a predefined pattern, and another is the target of the pattern. Replica-molding is extensively employed to inexpensively create micro- and nano-scale patterns. When two surfaces slide against each other tribocharges are generated. To date, most of tribocharges are studied on flat surfaces. We show that frictional contact between patterned surfaces during replica molding generates nanoscale patterned tribocharges.
We study replica molding with a polycarbonate mold with a nanoscale pattern of periodic nanocones. We replicated the inverse of this pattern with the common elastomer poly(dimethylsiloxane) (PDMS), generating nanocup arrays in PDMS. The surface topography was characterized by atomic force microscopy and the surface electric potential measured by scanning Kelvin probe microscopy (SKPM). We found that the PDMS had lower electric potential inside the nanocups, with a small maxima of the electric potential at the bottom of the nanocup. This measurement indicated an intricate ensemble of charged nano-domains. The measurements could only be accounted for by negative tribocharges highly localized inside the nanocups in an unusual ring-like distribution below the rim of the nanocups. We found the nanoscale variations of the frictional stress, induced by the demolding action, as the main factor governing the tribocharge’s nanoscale distribution pattern. Surface potential measurements and numerical modeling confirms that the frictional stress is maximum just below the rim of the nanocups, where the ring-like tribocharges are formed. The nanopatterned charges can be utilized to generate novel nanostructure in polymer materials. We will discuss how these nanoscale charges can be employed to fabricate unusual nano-volcano arrays [2]. We will also discuss the origin of charging at the interface of tribo-materials.
[1] Q. Li, A. Peer, I. Cho, R. Biswas, J. Kim, Nature Communications 9, 974 (2018).
[2] Physics Today, 71 (5), 72 (2018). (Backscatter; highlight)
Partially supported by NSF grant CMMI- 1760348