Moving mechanochemistry forward: tribochemistry

When one thinks about friction, the chemical reactions occurring at the interfaces of sliding contacts are not usually the first thing that comes to mind. Most considerations of friction focus on values such as the friction coefficient, a unitless parameter that we are taught in our careers, that relates an applied load to the resistance to motion (friction). This simple relationship, known as Amonton’s law, however, fails to reveal the true nature of friction, where compressive and shear forces can conspire to alter friction, sometimes by changing the local structure to lead to wear of surfaces and interfaces or, in many cases, inducing chemical transformations. These combined processes can have deleterious effects on machined interfaces and, in the UK in 1966, Jost published a seminal report¹ illustrating the dramatic economic impact that friction and wear can have as a direct result of energy loss (due to friction) and the resulting breakdown of machines that result from wear, with the associated economic losses estimated to be on the order of 1–2% of GDP.

More recently, Holmberg and Erdemir published a study that the true impact was that ca. 20% of all of the world’s energy is wasted combating friction,² with a follow-on paper highlighting the corresponding impact of friction on sustainability.³ Moreover, while much research effort has gone into new approaches to generate energy, efforts to reduce energy losses have received much less attention; this remains a daunting challenge to develop true energy independence and a sustainable energy economy. Consequently, a fundamental understanding of tribochemistry, that is, reactions occurring under the influence of normal and shear stresses, may currently be one of the most commercially significant areas of mechanochemistry. However, studies of the mechanically driven chemical changes that occur in tribological contacts, has only truly been a topic of intense research for ca. 35 years and has remained largely in the purview of mechanical engineers. Importantly, insights from the tribology/tribochemistry community have much to bring to advancing the fundamental understanding of mechanochemical processes as they share much in common, but with greater focus on the role of the mechanics of the interfaces, which is often ignored in many mechanochemistry studies. This offers the means to advance our overall understanding of the role force plays in driving chemical reactions, a central element that needs greater attention in the community to truly advance the application of mechanochemistry for sustainable chemical manufacturing. As such, in this editorial, we look to highlight these connections and the role that mainstream mechanochemistry can bring to solving challenges in tribology and vice versa.

Many of the first experiments on how normal and shear stresses influence reaction rates on surfaces were carried out on tribological systems to understand the way in which additives to lubricants could react with machine elements to reduce the friction between the surfaces and to mitigate the wear rate.^4–23 The strategy in the early days was to post analyze surfaces that had been rubbed in the presence of some model lubricant – often containing chemically complex additives such as zinc dialkyl dithiophosphate (ZDDP) – and to try to understand what had happened on the surface. Similar to the challenge of developing a clear understanding of mechanochemical systems, understanding such tribochemical processes was hampered by the difficulty of analyzing moving solid–solid interfaces, where the most relevant materials, steels, ceramics, etc., were not optically transparent. The main difficulty, in these early stages, was that the composition and topography of tribologically formed films were completely different from those obtained by heating the sample in the same fluid: there appeared to be something fundamentally different between thermal and tribological formed films. These experimental challenges were so great, that at the first World Tribology Congress, held in London in 1997, only one invited paper and a handful of presentations and posters appeared that touched on tribochemistry.²⁴

Several main contenders were suggested for these unusual phenomena. The first was that the local “flash” temperatures (aka hot spots) at the contacting protuberances or asperities on the rough contacting surfaces were much higher than the average surface temperature, and perhaps was no longer even in thermal equilibrium with a Boltzmann energy distribution, so perhaps was not even a temperature in the conventional thermodynamic sense. The second was the idea that interfacial sliding could induce the emission of electrons, and tribo-plasmas were detected, in particular for insulating materials under high-speed sliding.²⁵ This resulted in the Negative-Ion-Radical Action Mechanism (NIRAM) proposal,²⁶ that tribochemical reaction pathways could be predicted based on mass spectrometer fragmentation patterns. A third idea, based on what was being gleaned from surface scientists about how low-coordination sites on surfaces could influence chemical reactivity, was that more reactive sites were created by the surfaces sliding past each other. Perhaps not surprisingly, the parallels between early tribological theories and those suggested to control mechanochemistry now are similar.

Because tribochemistry – and also arguably mechanochemistry – are inherently surface problems, the techniques of surface analysis were being brought to bear on tribochemistry; synchrotron radiation,⁵ for example, was being used to analyze tribological interfaces, optical techniques, such as Raman^27,28 and infrared spectroscopies,^29,30 and X-ray photoelectron spectroscopy^31–33 were used to do in situ analysis, although the complexity of the interfaces often defeated these attempts.

An important advance in analyzing tribological interfaces came with the development of the atomic force microscope (AFM).^34,35 Akin to an old-fashioned profilometer, AFM opened the door to exploring the molecular forces acting between surfaces. Because it had been proposed that the interactions between real surfaces in sliding contact arose from the high compressive and shear forces at nanoscopic asperities at surfaces, the fact that the AFM tip had a radius of curvature on the order of 10 nm, made it possible to consider the AFM probe as a single sliding asperity, leading the advancement of the field of nanotribology.

At the same time, surface analytical methods in ultrahigh vacuum were being used to measure the high-temperature reaction rates of so-called extreme pressure (EP) additives such as halides and sulfides that were used under the harshest conditions such as machining and grinding.^35–38 The argument was that any sporadic flash temperatures would coalesce into an average temperature close to the thermodynamic value to allow EP lubrication chemistry to be theoretically modelled, as a foundation for understanding other regimes. This proved to be a successful approach and the tribofilm growth rate could be modelled by surface-electric-field driven processes proposed by Mott and Cabrera.³⁹ It was found that the seizure loads due to the failure of the lubricating film occurred when the interfacial temperature reached the melting point of the tribofilm material. This allowed the nature of the lubricating film to be identified.^40,41

In parallel with these experimental approaches was the development of simulation methods, particularly reactive molecular dynamics (MD) simulations, which solve Newton’s equations of motion of a system controlled by some model intermolecular potential. Simulations provide atomic-level insights into how the interface developed under (often very fast) sliding, as well as with the application of load. Arguably, the major impact was conceptual because tribologists could, for the first time, visualize what might be occurring on the atomic scale when two surfaces were rubbed together. The ability of MD simulations to provide insights into chemical reactions driven by shear forces was first demonstrated for diamond surfaces with attached alkyl groups.⁴² In that work, interfacial chemical reactions were initiated by the shear-induced removal of hydrogen atoms from the tails of the alkyl groups, producing radicals that then bonded to others on the opposite surface. Since that work, advancements in computing power, the development and improvement of potential energy functions^43,44 including through machine learning,⁴⁵ and significant progress with ab initio MD (using quantum instead of classical mechanics)^46,47 has opened up a large number of systems to examination using MD simulations, as discussed in several recent reviews.^48–50

Despite these advancements, MD simulations still suffer from some limitations. While more accurate ab initio MD is based on quantum mechanics, the required computational power limits simulation sizes to scales well below those in experiments. For example, simulation box lengths on the order of 1–2 nm can be achieved in ab initio MD,^46,48 while 10s to 100s of nm are achievable using empirical potentials. While advanced reactive classical MD can match experimental scales (such as the size of the stressed zone in a realistic AFM tip–sample contact),⁵¹ the motion of atoms is governed by forces derived from the potential-energy function. Potentials vary in complexity and accuracy (see Harrison et al.⁴³ for a discussion of the differences in potentials) due to approximations made while attempting to reproduce quantum mechanical reality. For instance, some potential-energy functions are unable to reproduce activation energies of chemical reactions with sufficient accuracy. While bond dissociation energies for covalent materials are in the range of 100s of kJ mol⁻¹, errors of a few kiloJoules per mole in the reaction coordinate can drastically change the rate of a reaction. This limitation typically stems from the fitting of potential parameters to equilibrium properties or empirical data without attention to transition-state properties, or from other features of the potential such as short-range distance cut-offs required for practical computational efficiency. It is possible to develop empirical potentials that more accurately reflect processes that involve the breaking of bonds, such as crack propagation⁵² and wear⁵³ in diamond. Machine-learning models, if trained properly, are also making strides in capturing the essential chemistry and physics of these processes.⁴⁸ In addition, because chemical reactions are rare occurrences on the timescale of the motion of individual atoms and molecules, this makes reaction statistics difficult to obtain. On the other hand, when specific rare events of interest are known, one can leverage that by using accelerated MD techniques.⁵⁴

Importantly, a major advantage of MD simulations is that they are able to model asperity-level tribochemical processes in a reliable way, permitting direct comparison with AFM measurements. Some of these joint studies have examined the wear mechanisms of diamond-like carbon tips in contact with ultrananocrystalline diamond (UNCD),⁵¹ DLC in contact with diamond,⁵⁵ a Si tip and diamond surfaces,⁵⁶ between silica surfaces,⁵⁷ and a Si tip sliding on the basal plane of graphite.^58,59 In all these works, MD simulations have provided key insights into wear mechanisms and other factors that influence the AFM measurements. For example, atomic-scale mechanisms that led to a sliding-induced increase in adhesion were identified. Namely, sliding caused shear stresses that induced removal of passivating surface species (–H and –OH groups on silicon), followed by rapid covalent bond formation across the interface between the exposed radicals. This required much larger forces to separate than in the case of passivated surfaces whose adhesion was largely due to weaker van der Waals bonds.⁵⁶

As noted above, the use of AFM tips, where the contacts were, in principle, much better defined than rough engineering surfaces because they represented a single asperity, allowed much more precise measurements of reaction rates to be made. In addition, the small tip contact size meant that contact stresses in the GPa range could be attained relatively straightforwardly. Importantly, this approach could be implemented in various ways. First, it could be carried out in solution using conventional lubricant additives such as ZDDP, where the reaction rate could be measured from the thickness of the film that was formed.⁶⁰ Including wear models revealed that these “single asperity” contacts yielded growth kinetics that corresponded well to real lubricants.⁶¹ However, these molecules are sufficiently complex that their surface reaction pathways are not well understood, thus limiting their use for developing molecular-scale models.

Analogous experiments were carried out on simple overlayer systems, such as functionalized graphene, where the AFM-tip-induced reactions could be tracked and the energetics of the reaction barriers with respect to applied forces examined.^62,63 Model surface lubricant films, such alkyl thiolates on copper offered the possibility of exploiting their relative simplicity to use first principles quantum calculations to obtain reaction energetics.^64–68 In fact the literature is now replete with examples of where AFM has been used to probe the energetics of tribochemical wear,⁶⁹ and of tribofilm formation.⁷⁰ This allowed analytical models of the tribo/mechanochemical rates to be proposed and tested based on classical kinetic models such as transition-state theory. They predict that the reaction rate should vary exponentially with the applied stress.^71–73

Results such as these have represented a very significant advance in our understanding of tribochemical reactions over the decades since the World Tribology Congress, that envisages mechanisms that vary from those in which the process is dominated by a direct coupling of the stress into the reaction under mild conditions, to those that are dominated by the creation of high interfacial temperatures under the most severe rubbing conditions. It is clear that the fates and development of the fields of tribochemistry and mechanochemistry are intimately intertwined, and that advances in one area will benefit the other. Hopefully, what has been learned during the development of tribochemistry from the surface-science perspective will help inform the ideas of how mechanochemical reactions proceed. For example, one of the greatest challenges for mechanochemistry is to understand the nature of the contacting interfaces. This will be undoubtedly underpinned by the extensive work that has already been carried out calculating how real rough surfaces contact each other, and the effect of their elastic plastic deformation. But similarly, tribologists can hopefully also see ways in which new mechanically driven organic reactions can be adapted to facilitate the design of better friction modifiers in a predictable way.

An emerging area in which synergistic development might be expected is in the field of artificial intelligence (AI) and machine learning (ML) models. Both fields suffer from similar obstacles of being able to define the associated metadata involved with any process that involves a buried interface. However, the potential benefit for both fields are very significant, with the possibility of the bespoke prediction of new lubricant formulations and coatings in tribochemistry, to expanding databases of mechanochemical reactions by using AI to suggest new reaction pathways and to facilitate and advance more sustainable approaches for chemical synthesis that extend beyond traditional trial-and-error approaches.

References

1. H. P. Jost, Wear, 1990, 136, 1–17.
2. K. Holmberg, P. Andersson and A. Erdemir, Tribol. Int., 2012, 47, 221–234.
3. K. Holmberg and A. Erdemir, Tribol. Int., 2019, 135, 389–396.
4. Z. Yin, M. Kasrai, M. Fuller, G. M. Bancroft, K. Fyfe and K. H. Tan, Wear, 1997, 202, 172–191.
5. G. W. Canning, M. L. Suominen Fuller, G. M. Bancroft, M. Kasrai, J. N. Cutler, G. De Stasio and B. Gilbert, Tribol. Lett., 1999, 6, 159–169.
6. M. N. Najman, M. Kasrai, G. M. Bancroft and A. Miller, Tribol. Lett., 2002, 13, 209–218.
7. Z. Zhang, M. Kasrai, G. M. Bancroft and E. S. Yamaguchi, Tribol. Lett., 2003, 15, 377–384.
8. E. S. Yamaguchi, Z. Zhang, M. Kasrai and G. M. Bancroft, Tribol. Lett., 2003, 15, 385–394.
9. Z. Zhang, E. S. Yamaguchi, M. Kasrai and G. M. Bancroft, Tribol. Trans., 2004, 47, 527–536.
10. M. N. Najman, M. Kasrai and G. M. Bancroft, Tribol. Lett., 2004, 17, 217–229.
11. Z. Zhang, E. S. Yamaguchi, M. Kasrai, G. M. Bancroft, X. Liu and M. E. Fleet, Tribol. Lett., 2005, 19, 221–229.
12. N. J. Mosey, T. K. Woo, M. Kasrai, P. R. Norton, G. M. Bancroft and M. H. Müser, Tribol. Lett., 2006, 24, 105–114.
13. Z. Zhang, E. S. Yamaguchi, L. Yu, M. Kasrai and G. M. Bancroft, Tribol. Trans., 2007, 50, 58–67.
14. J. M. Martin, M. Belin, J. L. Mansot, H. Dexpert and P. Lagarde, ASLE Trans., 1986, 29, 523–531.
15. M. I. de Barros’Bouchet, J. M. Martin, T. Le-Mogne and B. Vacher, Tribol. Int., 2005, 38, 257–264.
16. D. Philippon, M. I. de Barros-Bouchet, T. L. Mogne, E. Gresser and J. M. Martin, Tribol.-Mater., Surf. Interfaces, 2007, 1, 113–123.
17. D. Philippon, M. I. De Barros-Bouchet, T. Le Mogne, O. Lerasle, A. Bouffet and J. M. Martin, Tribol. Int., 2011, 44, 684–691.
18. M. I. De Barros Bouchet, J. M. Martin, J. Avila, M. Kano, K. Yoshida, T. Tsuruda, S. Bai, Y. Higuchi, N. Ozawa, M. Kubo and M. C. Asensio, Sci. Rep., 2017, 7, 46394.
19. P. M. Cann and H. A. Spikes, Tribol. Trans., 1991, 34, 248–256.
20. H. Spikes, Tribol. Lett., 2004, 17, 469–489.
21. J. Zhang and H. Spikes, Tribol. Lett., 2016, 63, 24.
22. J. Zhang, J. P. Ewen, M. Ueda, J. S. S. Wong and H. A. Spikes, ACS Appl. Mater. Interfaces, 2020, 12, 6662–6676.
23. J. Zhang, J. P. Ewen and H. A. Spikes, Mol. Syst. Des. Eng., 2022, 7, 1045–1055.
24. I. M. Hutchings, Abstracts of Papers from World Tribology Congress, 8-12 September 1997, Mechanical Engineering Publications Ltd, 1997.
25. K. Nakayama, in Encyclopedia of Tribology, ed. Q. J. Wang and Y.-W. Chung, Springer US, Boston, MA, 2013, pp. 3750–3760.
26. C. Kajdas and K. Hiratsuka, Proc. Inst. Mech. Eng., Part J, 2009, 223, 827–848.
27. I. L. Singer, Langmuir, 1996, 12, 4486–4491.
28. T. W. Scharf and I. L. Singer, Tribol. Lett., 2003, 14, 3–8.
29. P. M. Cann and H. A. Spikes, Tribol. Lett., 2005, 19, 289–297.
30. F. M. Piras, A. Rossi and N. D. Spencer, Langmuir, 2002, 18, 6606–6613.
31. A. Rossi, F. M. Piras, D. Kim, A. J. Gellman and N. D. Spencer, Tribol. Lett., 2006, 23, 197–208.
32. M. A. Scorciapino, G. Navarra, B. Elsener and A. Rossi, J. Phys. Chem. C, 2009, 113, 21328–21337.
33. F. Mangolini, A. Rossi and N. D. Spencer, J. Phys. Chem. C, 2010, 115, 1339–1354.
34. C. M. Mate and R. W. Carpick, in Tribology on the Small Scale: A Modern Textbook on Friction, Lubrication, and Wear, ed. C. M. Mate and R. W. Carpick, Oxford University Press, 2019.
35. P. V. Kotvis and W. T. Tysoe, Tribol. Trans., 1998, 41, 117–123.
36. T. J. Blunt, P. V. Kotvis and W. T. Tysoe, Tribol. Trans., 1998, 41, 129–139.
37. J. Lara, K. K. Surerus, P. V. Kotvis, M. E. Contreras, J. L. Rico and W. T. Tysoe, Wear, 2000, 239, 77–82.
38. M. Kaltchev, P. V. Kotvis, T. J. Blunt, J. Lara and W. T. Tysoe, Tribol. Lett., 2001, 10, 45–50.
39. N. Cabrera and N. F. Mott, Rep. Prog. Phys., 1949, 12, 163.
40. F. Gao, O. Furlong, P. V. Kotvis and W. T. Tysoe, Tribol. Lett., 2005, 20, 171–176.
41. O. Furlong, F. Gao, P. Kotvis and W. T. Tysoe, Tribol. Int., 2007, 40, 699–708.
42. J. A. Harrison and D. W. Brenner, J. Am. Chem. Soc., 1994, 116, 10399–10402.
43. J. A. Harrison, J. D. Schall, S. Maskey, P. T. Mikulski, M. T. Knippenberg and B. H. Morrow, Appl. Phys. Rev., 2018, 5, 031104.
44. T. Liang, Y. K. Shin, Y.-T. Cheng, D. E. Yilmaz, K. G. Vishnu, O. Verners, C. Zou, S. R. Phillpot, S. B. Sinnott and A. C. T. van Duin, Annu. Rev. Mater. Res., 2013, 43, 109–129.
45. V. L. Deringer, M. A. Caro and G. Csányi, Adv. Mater., 2019, 31, 1902765.
46. R. Iftimie, P. Minary and M. E. Tuckerman, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 6654–6659.
47. H. Ustunel and D. Toffoli, Electron. Struct., 2022, 4, 023002.
48. L. Pastewka, A. I. Vakis, S. J. Eder, R. Aghababaei, A. Almqvist, G. Carbone, M. Chandross, D. Dini, H. J. Ehrich, J. P. Ewen, N. Menga, J.-F. Molinari, G. Moras, L. Nicola, M. Paggi, C. Putignano, M. Scaraggi, V. A. Yastrebov and M. H. Müser, Tribol. Int., 2026, 218, 111326.
49. A. I. Vakis, V. A. Yastrebov, J. Scheibert, L. Nicola, D. Dini, C. Minfray, A. Almqvist, M. Paggi, S. Lee, G. Limbert, J. F. Molinari, G. Anciaux, R. Aghababaei, S. Echeverri Restrepo, A. Papangelo, A. Cammarata, P. Nicolini, C. Putignano, G. Carbone, S. Stupkiewicz, J. Lengiewicz, G. Costagliola, F. Bosia, R. Guarino, N. M. Pugno, M. H. Müser and M. Ciavarella, Tribol. Int., 2018, 125, 169–199.
50. A. Martini, S. J. Eder and N. Dörr, Lubricants, 2020, 8, 44.
51. V. Vahdat, K. E. Ryan, P. L. Keating, Y. Jiang, S. P. Adiga, J. D. Schall, K. T. Turner, J. A. Harrison and R. W. Carpick, ACS Nano, 2014, 8, 7027–7040.
52. L. Pastewka, P. Pou, R. Pérez, P. Gumbsch and M. Moseler, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 78, 161402.
53. L. Pastewka, S. Moser, P. Gumbsch and M. Moseler, Nat. Mater., 2011, 10, 34–38.
54. D. Perez and T. Lelièvre, in Comprehensive Computational Chemistry, ed. M. Yáñez and R. J. Boyd, Elsevier, Oxford, 1st edn, 2024, pp. 360–383.
55. R. A. Bernal, P. Chen, J. D. Schall, J. A. Harrison, Y.-R. Jeng and R. W. Carpick, Carbon, 2018, 128, 267–276.
56. Z. B. Milne, J. D. Schall, T. D. B. Jacobs, J. A. Harrison and R. W. Carpick, ACS Appl. Mater. Interfaces, 2019, 11, 40734–40748.
57. C. Luo, Y. Jiang, Y. Liu, Y. Wang, J. Sun, L. Qian and L. Chen, Front. Chem., 2022, 10, 852371.
58. D.-C. Yue, T.-B. Ma, Y.-Z. Hu, J. Yeon, A. C. T. van Duin, H. Wang and J. Luo, Langmuir, 2015, 31, 1429–1436.
59. Z. Chen, A. Khajeh, A. Martini and S. H. Kim, ACS Appl. Mater. Interfaces, 2021, 13, 1895–1902.
60. N. N. Gosvami, J. A. Bares, F. Mangolini, A. R. Konicek, D. G. Yablon and R. W. Carpick, Science, 2015, 348, 102–106.
61. A. Akchurin and R. Bosman, Tribol. Lett., 2017, 65, 59.
62. J. R. Felts, A. J. Oyer, S. C. Hernández, K. E. Whitener Jr, J. T. Robinson, S. G. Walton and P. E. Sheehan, Nat. Commun., 2015, 6, 6467.
63. S. Raghuraman, M. B. Elinski, J. D. Batteas and J. R. Felts, Nano Lett., 2017, 17, 2111–2117.
64. N. Hopper, F. Sidoroff, R. Rana, R. Bavisotto, J. Cayer-Barrioz, D. Mazuyer and W. T. Tysoe, Phys. Chem. Chem. Phys., 2023, 25, 15855–15861.
65. N. Hopper, F. Sidoroff, J. Cayer-Barrioz, D. Mazuyer and W. T. Tysoe, Tribol. Lett., 2023, 71, 121.
66. R. Rana, G. Djuidje Kenmoe, F. Sidoroff, R. Bavisotto, N. Hopper and W. T. Tysoe, J. Phys. Chem. C, 2022, 126, 11585–11593.
67. R. Rana, N. Hopper, F. Sidoroff and W. T. Tysoe, Chem. Sci., 2022, 13, 12651–12658.
68. R. Rana, N. Hopper, F. Sidoroff, J. Cayer-Barrioz, D. Mazuyer and W. T. Tysoe, Tribol. Lett., 2024, 72, 76.
69. S. Kopta and M. Salmeron, J. Chem. Phys., 2000, 113, 8249–8252.
70. S. Kumar, S. H. Kim and A. Martini, Chem. Rev., 2026, 126, 3883–3906.
71. J. Yeon, X. He, A. Martini and S. H. Kim, ACS Appl. Mater. Interfaces, 2017, 9, 3142–3148.
72. X. He and S. H. Kim, Langmuir, 2018, 34, 2432–2440.
73. W. Tysoe, Tribol. Lett., 2017, 65, 48.

---