Ultralow Friction of Steel Surfaces Using a 1,3 ... - ACS Publications

Subscriber access provided by - Access paid by the | UCSB Libraries

Article

Ultralow Friction of Steel Surfaces Using a 1,3Diketone Lubricant in the Thin Film Lubrication Regime Ke Li, Tobias Amann, Mathias List, Michael Walter, Michael Moseler, Andreas Kailer, and Jürgen Rühe Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.5b02315 • Publication Date (Web): 12 Aug 2015 Downloaded from http://pubs.acs.org on August 25, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Ultralow Friction of Steel Surfaces Using a 1,3-Diketone Lubricant in the Thin Film Lubrication Regime Ke Li a,d,e, Tobias Amannb, Mathias Listc, Michael Walterbc, Michael Moselerb,c, Andreas Kailer b and Jürgen Rühea a University of Freiburg, IMTEK - Department of Microsystems Engineering, Georges-KoehlerAllee 103, Freiburg, Germany. b Fraunhofer Institute for Mechanics of Materials IWM, Woehlerstraße 11, Freiburg, Germany. c University of Freiburg, FMF - Freiburger Materialforschungszentrum, Stefan-Meier-Straße 21, Freiburg, Germany. d National Engineering Research Center for Water Transport Safety, Wuhan, 430063, China. e Intelligent Transport Systems Research Center, Wuhan University of Technology, Wuhan, 430063, China.

1 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

Abstract: Ultralow friction (coefficient of friction μ≈0.005) is observed when two steel surfaces are brought into sliding contact in the presence of a particular 1,3-diketone lubricant (1-(4-ethyl phenyl) nonane-1,3-dione). We investigate the friction process of such a system both experimentally and theoretically and show that the superlubricity is caused by a novel, unique mechanism: The formation of iron-1,3-diketonato complexes during frictional contact leads to a self-limiting, tribochemical polishing process while at the same time a self-assembled monolayer of the diketone is formed on the employed steel surfaces. This polishing process reduces the contact pressure and at the same time leads to formation of a boundary lubricant layer. During sliding the system transits from the original boundary lubrication regime towards hydrodynamic lubrication. Conductivity measurements across the friction gap during sliding show that the lubricant layer present in the gap between the two shearing surfaces is a only few ten nanometers thick, so that the molecules experience under typical sliding conditions shear rates of a few 106 s1

. Simulations show that under such strong shear the molecules become strongly oriented in the

friction gap and the effective viscosity in sliding direction is significantly reduced so that the system is in the thin film lubrication regime and superlubricity is observed. The results of the experiments suggest that such diketones are promising lubricants for the decrease of energy loss and frictional damage in steel based mechanical devices.


Page 3 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

1 Introduction Although in many cases friction is a desired or even essential physical phenomenon for moving objects (for example during walking or for braking a vehicle), the most common consequences of friction are a nuisance, since they lead to energy loss and wear. For example, friction consumes 10% of a car’s or truck’s fuel. For example in the U.S. alone almost 31 billion US dollar worth of petrol are lost on friction in automobile engines every year1. Additionally, about 80% of machinery component failure is caused by wear2. To improve energy efficiency and material durability, friction reduction is mandatory - especially for micro- and nanomechanical systems as such systems are to a large extent dominated by surface interactions3. To minimize the effect of friction many different lubricant systems have been developed and are today widely used in standard technology. To reduce friction a broad spectrum of lubricants are employed, including oils4, greases5, or polymers such as polydimethylsiloxanes6 or perfluoropolyethers7 . In the past years a number of systems have been developed, which show, compared to standard lubricants, much lower, frequently called ultralow, friction coefficients such as molybdenum disulfide8, graphite9, nanocrystalline diamond-like carbon films10 and polymer brushes11 to name just a few examples. For such systems sometimes the term “superlubricity” is employed, which was firstly proposed in 1990 by M. Hirano12, who described a theoretical sliding regime in which the resistance to sliding completely vanishes. In practical systems the friction never completely disappears and frequently tribological contacts with coefficients of friction (COF) µ on the order of µ ≈ 10-3 are considered as superlubricious13. In superlubricity, the ultralow friction is induced by a layered structure of the lubricant molecules in which the interaction between two neighboring layers is very small. The superlubricity of polyelectrolyte brushes for example, is caused by the strong chain stretching and the accompanying changes in configurational


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

entropy and the high osmotic pressure generated from the counter ions11. For silicon nitride/aluminum oxide surfaces in contact with water and other aqueous solutions like phosphoric acid14, the ultralow friction is based on a tribochemical reaction between water and sliding surfaces. However, as far as practical applications are concerned, a severe limitation of most superlubricious system is that the ultralow friction is commonly observed only on a very small scale and can hardly be realized on the scale of macroscopic objects. Additionally, the effect is usually restricted to very specific environmental conditions. Recently, Zhang et al.15 reported superlubricity in centimeterlong double-walled carbon nanotubes under ambient conditions. However, in order to achieve ultralow friction it is required that no defect exist in such long nanotubes, so the fabrication poses a severe technical challenge. The lubrication properties of oils, which are the most popular lubricants in mechanical systems such as bearings or gears, are usually limited by the viscosity, so that the minimum coefficients of friction of traditional oil based lubricants are around 0.01~0.05. In molecular dynamics simulations, a liquid-to-solid transition has been reported when some oils are confined into very thin films16. To understand the nature of this thin confined film, Luo et al.17 proposed an ordered liquid film model, which consists of a solid-like monolayer on the surface (adsorbed layer), an aligned layer in a short distance from the surface (ordered layer), and an isotropic fluid layer in the bulk (fluid layer). When the film thickness is reduced to molecular dimensions, the tribological properties are dominated by the ordered layer and a so-called “thin film lubrication” regime is proposed between boundary lubrication and fluid lubrication. Based on this theory Zhang et al.18 studied a dynamic rheological model for thin film lubrication. They claimed that when a lubricant exhibits both strong interfacial interactions with the substrate and strong intermolecular interactions, thin film lubrication with a very low coefficient of friction is promoted and maintained over a rather wide


Page 5 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

range of velocities. Zhu et al.19,20 found that confined and ordered molecular layers of squalane show an oscillatory force-distance profile, which causes low energy dissipation during sliding. Jabbarzadeh et al.21 also reported similar aligned multilayers of confined dodecane, which induce an ultralow effective viscosity under shear and an ultralow COF. However, most descriptions of ultralow friction in confined thin films are based on simulations, whereas the superlubricity of oilbased lubricants has been rarely reported in experiments22. In prior work23 a 1,3-diketone lubricant (1-(4-ethyl phenyl) nonane-1,3-dione (Fig. 1), abbreviated in the following as EPND-02/06, has been investigated, which shows much lower COF than a variety of standard oil-based lubricants on steel surfaces after a short running-in period. A tribochemical polishing of the steel surfaces in the presence of the 1,3-diketone lubricants has been found. This polishing process causes self-limiting wear and lead to two strongly conformal surfaces23. As the complex formation is strongly temperature dependent, the wear process does not continue at the same rate throughout the friction process, but stops once a certain reduction of the contact pressure and a concurrent elimination of the solid-solid contacts have been reached24-25. This is evidenced by the fact that the formation of iron-diketonate complexes levels off after the running in time has been completed.25 It has been demonstrated that when a continuous thin lubricant film is formed the lubrication regime transits from boundary lubrication to fluid lubrication and the COF decreases dramatically. However, the observed low value of the COF (≈ 0.005) is unusual even for fluid lubrication. Here, we provide deeper insights into the lubrication mechanism of EPND-02/06 by a combination of experiments with multiscale simulations of the 1,3-diketone lubricant between two sliding interfaces.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

Fig 1: Tribochemical polishing processes of EPND-02/06 on steel surfaces (load 50 N, 90 ℃, 1 mm sliding path, 50 Hz; adapted from Ref. 25).

2 Methods 2.1 Experimental methods EPND-02/06 (1-(4-ethyl phenyl) nonane-1,3-dione, Fig. 1) and PCND-05/06 (1-(4-pentylcyclohexyl)-nonane-1,3-dione, Nematel GmbH, Fig. 6) are used as lubricants in this study. They have similar dynamic viscosities at 90 °C at a shear rate of 1000 s-1 (2.6mPa·s and 3.5 mPa·s respectively). Tribological experiments were performed using an oscillating cylinder (diameter = 15.0 mm, length = 22.0 mm) - on- disc (diameter = 24.0 mm, height = 7.9 mm) sliding geometry (SRV-III, Optimol Instruments). For the tribological tests standardized 100Cr6 specimens (steel 1.3505) were used with a hardness of 62 HRD (Rockwell C hardness). The tests parameters were 50 N normal load, 50 Hz frequency, 90 °C temperature, 1 mm sliding path and 16 h testing time. The initial cylinder-on-disc contact pressure was calculated using the geometrical parameters and employing the Hertz contact theory23. During the test the worn area increased due to the tribochemical polishing processes. After renewed determination of the contact area the contact


Page 7 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

pressure was recalculated. One drop of the lubricant fluid (approximately 20 µl) was placed on the cylinder before the test. For reference purposes a pair of glass specimen with the same size as the steel ones was used. To analyze the surfaces, X-ray photoelectron spectroscopy was carried out with a Physical Electronics 5600 ci spectrometer equipped with a concentric hemispherical analyzer and an Al Kα X-ray source (15 keV, filament current 20 mA; 10-9~10-8 mbar). All spectra were measured at a 45 ° angle with respect to the sample surface. Electrical contact resistance measurements were performed with a standard commercial voltmeter to study the intermittent contact of the surfaces and the thickness of the lubricant film.

2.2 Computational methods Molecular dynamics simulations were performed in order to obtain a deeper understanding of the friction in EPND-02/06 lubricated sliding contacts. To represent the dispersive interaction between EPND-02/06 molecules a Gay-Berne potential (GB) was used. In the GB coarse graining approach molecules are represented by rigid ellipsoids that interact via an anisotropic LennardJones (LJ) potential26. The pairwise LJ interaction is calculated with respect to the relative orientation of two ellipsoids and their minimal distance. Details of the GB potential

were described

by Berardi et al.27

(1)

(2)


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

The distance dependence of the interaction is given by potential.

The

maximum

depth

of

the

GB

which has the form of a shifted LJ potential

is

given

min |

is the distance between the centers of mass of two particles and is the minimal distance between two ellipsoids. The 3

3 rotational matrices

,

by

and

|

represent

the orientation of the GB ellipsoids. The directional vector between two particles is given by ̂ /|

|. Expressions for the functions

and

can be found in

28,29

. The potential can be

adjusted to the special properties of a molecule by tuning the empirical parameters

, , ,

and

. These parameters were obtained from a fit to intermolecular potentials described by the universal force field as detailed in the supporting information (SI). The model for the tribo system consisted of 897 GB particles as displayed in Fig. 2. The GB fluid (white particles in Fig. 2) was confined between two solid surfaces. The surfaces were represented by two fixed layers of GB particles (depicted in violet color in Fig. 2). Two neighboring layers of GB particles were used for thermalization via velocity scaling (particles depicted in green in Fig. 2).The interaction strength in the green subsystems was increased by a factor of 5 to make them solid-like. A normal pressure of 1 GPa was introduced by applying a corresponding force to the top rigid layer of GB particles. Subsequently, the system was sheared by moving the upper rigid layer with a constant velocity of 100 m/s relative to the lower rigid layer. In a second simulation, the role of additional Fe(III) complexes in the EPND-02/06 liquid was investigated. The complexes where represented by spherical particles that interact with the EPND02/06 GB particles via LJ potentials (see SI for a detailed description of the parameterization of GB/LJ potential that describes the resulting mixture of ellipsoids and spheres). The initial configuration of this system is depicted in Fig. 2c.


Page 9 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Fig 2: Model structure for the thin film sliding simulations. The EPND-02/06 is represented by Gay-Berne particles (in white). Two rigid walls are shown by violet GB particles, while thermalization of the system is established with velocity scaling of the green GB particles. (a) System before sliding; sliding is initiated by moving the top violet layer with constant velocity (indicated by the red arrow). (b) Configuration of particles at the end of the shear simulation. Note, that all particles with bold solid green edges were initially on a straight vertical line. These particles serve as markers to visualize the strain accommodation mode. (c) Simulation setup that includes Fe(III) 1,3 diketonato complexes as spherical LJ particles (light green). (d) Final configuration after sliding.

3 Results and discussion 3.1 Monolayer formation A key requirement of any thin film lubricant system is to have a thin layer of a firmly anchored boundary lubricant at the sliding surfaces. Figure 3a shows the XPS analysis of an iron sheet (0.5 mm thick, 99.99% pure) before and after the reaction with EPND-02/06 at 90 °C for 16 h. To remove any loosely attached organic liquid, the iron sheet was thoroughly cleaned by acetone in both cases. Nearly no carbon signal was observed on the unmodified iron surface, which was also measured as a reference. However, a strong carbon signal is observed after the reaction of this


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

surface with EPND-02/06, which indicates that an organic layer has been chemically bound to the iron surface. The high-resolution C1s spectrum of the iron sheet after reaction with EPND-02/06 indicates that carbon atoms in two different chemical environments (C=O and C-C) are present at the surface giving rise to signals at 285.3 eV and 287.0 eV (Figure 3b). The positions of those signals are in good agreement with the literature on similar compounds30. The area fractions of the two signals are 13% and 87% respectively, which is consistent with the molecular structure of EPND-02/06 (C=O 11.7%, C-C 88.3%). This result shows that the surface adsorbed layer is caused by the coordination of EPND-02/06 with the iron substrate.

Fig 3: (a) XPS analysis of an iron sheet before and after reaction with EPND-02/06 at 90 °C for 16h. Before the measurement, iron sheet was cleaned by acetone and dried in both cases. (b) The resolved high-resolution XPS C 1s spectrum from an iron sheet after reaction with EPND-02/06, showing peaks arising from the carbons of C=O and C-C bonds.

To investigate the importance of this adsorbed layer a comparison of the frictional behavior was made between two surfaces, where one has no surface-attached layer, while the other surface is covered with such a monolayer. As the monolayer formation is based on a specific complexation of the 1,3-diketone with the iron surface, glass surfaces cannot react in a similar way. So as a control experiment a glass cylinder and a glass disc with the same size as the steel specimens were


Page 11 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

used, while all test parameters in the friction experiments were kept the same (50 N, 1 mm, 50 Hz, 90 °C). At first, a dry friction test without any lubricant was carried out for 1 min. This was carried out to remove some of the asperities and this way to reduce the contact pressure (Figure 4a). The resultant worn area of the glass cylinder was 34.1 mm2 and the calculated contact pressure for a 50 N load was 1.5 MPa. This contact pressure was even lower than the one (6 MPa) for the friction test of EPND-02/06 on steel surfaces after ultralow friction was achieved (Figure 1). Afterwards, one drop of EPND-02/06 lubricant was placed on the worn area and the friction test was restarted (Figure 4b). Compared with dry friction on the same glass surface, the COF was reduced from 0.8 to 0.2 upon addition of the EPND-02/06 lubricant. However, this COF value on the lubricant/glass system is still almost three orders of magnitude far away from the ultralow friction observed on a steel surface of the same geometry (μ≈0.005). It can be concluded that the adsorbed layer plays a very important role in the lubrication of EPND-02/06; without it, ultralow friction cannot be achieved even under very low contact pressure.

Fig 4: Reciprocating friction tests (load 50 N, 90 °C, 1 mm sliding path, 50 Hz) on glass surfaces. (a) Dry friction for 1 min. (b) Friction test with EPND-02/06 lubricant on the worn area of (a).

3.2 Molecular Shape of Lubricant


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

Fig 5: (a) Schematic molecular structure of a rod like mesogenic compound. Two rigid rings with flexible tails are linked by a connecting group. (b) Molecular structure of EPND-02/06.

Although the EPND-02/06 is not a liquid-crystalline substance itself as evidenced by DSC and polarization microscopy, the EPND-02/06 lubricant seems to have a rod-like structure, not unlike many liquid-crystalline substances (Fig 5). To experimentally investigate the influence of the molecular shape onto the lubricating properties, friction experiments were performed on a structurally similar compound, a 1,3-diketone substituted with a cyclohexane ring (PCND-05/06) using the same steel specimen and same test parameters as those in sliding experiments with EPND-02/06 (50 N, 1 mm, 50 Hz, 90 °C). (Fig. 6). PCND-05/06 has, similar to EPND-02/06, a 1,3-diketone group, which can react with the steel surface during the friction test. Due to the tribochemical polishing effect, also in the PCND-05/06 experiment the contact pressure was reduced during the running-in period and the coefficient of friction decreased with test time. However, even though the final contact pressure was lower than that in the EPND-02/06 experiment, the COF of KI-911 was still around µ=0.1, which is much higher than the one of EPND-02/06 (μ≈0.005). Similar results have been reported by Mori et al.31, who compared the tribological behavior of two liquid crystalline lubricants: cyanobiphenyl (CB) and cyanophenylcyclohexyl (CPC). When a phenyl group was substituted with a cyclohexyl group, the coefficient of friction increased from 0.035 (CB) to 0.045 (CPC).


Page 13 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Fig 6: Reciprocating friction test of EPND-02/06 and PCND-05/06 (load 50 N, 90 °C, 1 mm sliding path, 50 Hz, 16 h) on steel surfaces and the worn area of cylinder after the test.

This result can be explained by differences of the intermolecular forces between the two molecular structures. For EPND-02/06, the central 1,3-diketone group connected with aromatic ring forms a rigid flat structure, and the entire molecule assumes a more or less rod-like structure. When a benzene ring, however, is substituted by a cyclohexane ring, the planar structure is destroyed due to the chair/envelope structure of the cyclohexane ring which decreases packing and thus reduces molecular interactions and renders ordering in the layer much more difficult32

3.3 Confinement An important parameter in the lubricating process is the value of the shear rates, which in turn depends on the extent of confinement of the lubricant under sliding conditions. However, it is


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

difficult to measure the thickness under confinement by optical methods, especially when opaque surfaces as the steel surfaces are employed, however, the distance between the sliding surfaces can be easily elucidate from the electrical resistance between the two sliding steel surfaces33. Firstly, the electrical resistance of a lubricant film between the two steel surfaces was measured, while the surfaces were separated by a distance of 1 mm, as this allows to estimate the resistivity of the lubricant (Table1). Then the resistivity was measured during sliding. In the beginning of the friction experiments, a very low electrical resistance was observed. Apparently, the surfaces are in direct contact with each other, most likely due to asperity contacts between the two steel surfaces. However, during the friction test the asperities are sheared off and a continuous lubricant film is formed which separated the solid surfaces. Due to the presence of this liquid lubricant layer, an electric resistance of 1.2 KΩ was observed. When it is assumed that the specific conductivity and the contact area remain the same, the separation between the two surfaces can be calculated according to: R1/R2 = ρL1A/ρL2A = L1/L2, showing that under the applied conditions the distance is about 40 nm (Table 1). Even when this value is only a rough estimate and the precision should not be overestimated, it shows that the lubricant film is confined in the nanometer range. If one now considers that the sliding speed is 100 mm/s, it can be concluded that the film is sheared with shear rates on the order of γ. = v/L ≈ 2.5x106 s-1. Such shear rates are very high and are much higher than those commonly used for aligned of liquid crystals.

Table 1: Electrical resistance between steel surfaces during reciprocating friction test (load 50 N, 90 °C, 1 mm sliding path, 50 Hz, 16 h).

Lubricant film

Electric resistance R

Calculated distance L

1 mm confinement (no load)

30.0 MΩ

1 mm*


Page 15 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

start of friction test (50 N)

0.0 Ω

asperity contact

end of friction test (50 N)

1.2 KΩ

40 nm

*optical measurement

3.4 Molecular dynamics simulations The shear induced structural evolution in the EPND-02/06 film was further investigated using molecular dynamics simulations with the GB system displayed in Fig. 2. Figure 7 shows the distribution of angles

between the direction of the long molecular axis (director) and the

sliding direction for the sheared and the not sheared system. Initially (after thermalization – before sliding)

is approximately randomly distributed (see blue curve in Fig. 7). Note, that the small

dip of

can be explained by a small ordering effect that is caused by the confinement and

pressurization in the direction perpendicular to the sliding direction. After sliding the system for 4.1 ns at v=100 m/s (corresponding to a shear rate of

9x109 s-1) the director of the molecules

partly aligns into the direction of sliding (see high density values of red curves in Fig. 7a at cos 1 . . α=0°).

Fig 7: Distribution of the angle between the long GB axis and the shear direction. (a) pure GB liquid (ellipsoids only) and (b) mixture between GB ellipsoids and spherical LJ particles. Blue


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

curves: distribution in initial configuration. Red curves: distribution in final configuration after sliding.

This alignment of the GB particles is connected with a pronounced shear thinning. Figure 8 displays the dependence of the kinematic viscosity The viscosity

in the GB film as a function of shear rate .

was obtained from the force on the upper rigid wall. It clearly shows a non-

Newtonian power law34,35 decrease with corresponds to ~

with the power

(note the double logarithmic plot). The slope 0.94 and

0.92 for pure GB particles and the

mixture respectively. Thus there is very strong shear thinning in both cases, even larger than what is found in thin films of long chain organic molecules35, most likely due to the rather rigid structure of the diketone.

. Fig 8: Dependence of the kinematic viscosity

on the shear rate

in the GB simulations. Red

squares: pure GB liquid (ellipsoids only), blue circles: mixture between ellipsoids and spherical LJ particles.


Page 17 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

The EPND-02/06 is known to form FeIII complexes during wear of the steel surfaces resulting in deep red color of the lubricant. 24,25 Their presence might influence the self-organization of the remaining EPND-02/06 molecules as these complexes consist of three diketonate groups around the iron central atom in an octagonal complex and thus resemble a more sphere. They are therefore modeled by Lennard-Jones spheres. The starting point of the simulations depicted in figure 2c is very similar to that of figure 2a, only that we have now used 924 GB particles (representing EPND02/06) mixed with 49 (5%) spherical particles (representing the FeIII complexes). Depending on the amount of lubricant used and the duration of the friction test, the amount was determined by FTIR and UV-spectroscopy to be between 0 (start of the friction experiment) and 23% (16 h).24,25 Interestingly, a sliding simulation including the FeIII complexes revealed no qualitative influence on the shear thinning behavior of the EPND-02/06 liquid (see blue symbols in Fig. 8). This is in agreement with the director distribution

in Fig. 7b that indicates also in this case an alignment

of the EPND-02/06 in the mixture that is of comparable magnitude to the alignment in the pure phase (Fig. 7a). This alignment can already be detected by a visual inspection of the final configuration after sliding (see Fig. 2d).

3.5 Confinement and Friction To elucidate the influence of confinement on the friction behavior, friction tests under varying normal loads were performed (Figure 9). To allow for reproducible conditions, the sliding (load 50 N, 90 °C, 1 mm sliding path, 50 Hz) was first performed for more than 20 hours where an ultralow coefficient of friction was observed. Then the normal load was reduced from 50 N to 25 N. In contrast to standard systems where the friction decreases with decreasing load, in the diketo-


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 25

system the COF increased. By increasing the load back to 50 N, the coefficient of friction decreased to an ultralow level again. This seems to indicate that the with decreasing normal load and thus shear force the ordering of the lubricant molecules decreases, which would lead to appropriate changes in the (effective) viscosity of the system. In other terms the strongly confined and sheared lubricant has a lower effective viscosity than the unconfined liquid.

Fig 9: Reciprocating friction test of EPND-02/06 (90 °C, 1 mm sliding path, 50 Hz) under varying normal loads after 20 hours.

4 Discussion Two steel surfaces in sliding contact in the presence of an alkyl-aryl-substituted 1,3-diketone constitute a very unique tribological system. Together with the results of previous work24 the


Page 19 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

obtained results suggest a novel, complex mechanism for lubrication in such systems. From wear analysis and electrical conductivity measurements across the lubricant filled gap it can be concluded that upon frictional contact a self-limiting, tribochemical polishing of the surfaces takes place. This eliminates all asperity contacts, so that the surfaces become extremely conformal, which in turn leads to a strong reduction of the contact pressure. All sheared off iron species become completely solubilized in the lubricant oil through the formation of 1,3-diketonato iron complexes. This way that no abrasion particles are formed, which would be very detrimental for the frictional properties as the presence of hard particles in sheared systems usually leads to excessive wear. This surface-attached monolayer controls the interactions of the lubricants with the surfaces and might act as a template for a possible orientation of molecules composing the fluid. At the same time a self-assembled monolayer of the diketone is formed at the steel surface as evidenced by XPS analyses, which can act as a boundary lubricant. The tribochemical polishing process induces the formation of a full fluid film and the system transitions from a regime dominated by boundary lubrication towards a fluid lubrication regime. From the conductivity measurements it can be concluded that under the shearing conditions applied here, the liquid layer present in the gap between the two shearing surfaces a few ten nanometers thick, so that the molecules experience rather strong shear rates which are higher than γ.= 106 s-1 . Simulations via Gay-Berne particles shows that the high amount of energy imparted this way into the molecules could lead to an orientation of the molecules with an accompanying reduction of effective viscosity in shear direction. This effect is occurs even in the presence of additional Fe(III) complexes formed during the initial abrasion. The shear thinning depends strongly on the shear rate and thus on the size of the gap between the two sliding surfaces, which of course is strongly influenced by the load. Thus a reduction of the load should strongly reduce the orientation and through this increase the


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 25

local viscosity in shearing direction, which would consequently increase the coefficient of friction in a hydrodynamically lubricated system. Such an unusual behavior, with an increase of the friction up decreasing the load and vice versa is indeed observed (Fig. 9) The strong molecular orientation and a consequent reduction of the effective viscosity therefore explains why the friction coefficient in such a thin film lubrication system is even lower than that which would be expected for a regular hydrodynamic lubrication. The overall lubrication mechanism, which leads eventually to superlubricity, is illustrated in Figure 10.

Fig 10: Artists impression of the lubrication mechanism of 1,3-diketones on steel surfaces. No implication is made about the extent of orientation and the tilt angle in respect to the shear direction.

5 Conclusions The thin film lubrication behavior of a novel di-ketone based lubricant on steel surfaces has been investigated. The system exhibits excellent lubrication properties through a rather complex


Page 21 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

mechanism including tribochemical polishing and shear induced alignment in the nanoscale friction gap. While some of the individual features of the mechanism alone would be sufficient to give a system with more or less low coefficients of friction, all these factors need come together to give a thin film lubrication system which exhibits what might be called superlubricity. The proposed mechanism suggests, that the lubricant is only effective for surfaces of metals which can form diketonato complexes, However, as iron is the most common metal in mechanically strongly challenged system, 1,3 diketones could be an interesting lubricant for such materials..

Corresponding Author: Jürgen Rühe. University of Freiburg, IMTEK - Department of Microsystems Engineering, Georges-Koehler-Allee 103, Freiburg, Germany.

Acknowledgments: We gratefully acknowledge the DFG (Deutsche Forschungsgemeinschaft RU 489/22-1 and Mo 879/8-1) for funding this project. In addition we also thank Dr. Rudolf Eidenschink (Nematel GmbH, Mainz) for the supply of the 1,3-diketones. Computational resources from FZ Jülich are gratefully acknowledged.

Supporting information: Details on the parameters used for the molecular dynamics simulations are provided in the supporting information. This information is available free of charge via the Internet at http://pubs.acs.org/.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

11. 12. 13. 14. 15.

Holmberg, K.; Andersson, P.; Erdemir, A., Global energy consumption due to friction in passenger cars. Tribology International 2012, 47 (0), 221-234. Voith, M., Fighting Friction. Chemical & Engineering News 2010, 88 (41), 14-17. Bhushan, B.; Israelachvili, J. N.; Landman, U., Nanotribology - Friction, Wear and Lubrication at the Atomic-Scale. Nature 1995, 374 (6523), 607-616. Fox, N. J.; Stachowiak, G. W., Vegetable oil-based lubricants - A review of oxidation. Tribology International 2007, 40 (7), 1035-1046. Lugt, P. M., A Review on Grease Lubrication in Rolling Bearings. Tribology Transactions 2009, 52 (4), 470-480. McGuiggan, P. M.; Zhang, J.; Hsu, S. M., Comparison of friction measurements using the atomic force microscope and the surface forces apparatus: the issue of scale. Tribology Letters 2001, 10 (4), 217-223. Kasai, P. H.; Raman, V., Perfluoropolyethers with dialkylamine end groups: ultrastable lubricant for magnetic disk application. Tribology Letters 2002, 12 (2), 117-122. Chhowalla, M.; Amaratunga, G. A. J., Thin films of fullerene-like MoS2 nanoparticles with ultra-low friction and wear. Nature 2000, 407 (6801), 164-167. Dienwiebel, M.; Verhoeven, G. S.; Pradeep, N.; Frenken, J. W. M.; Heimberg, J. A.; Zandbergen, H. W., Superlubricity of graphite. Physical Review Letters 2004, 92 (12). De Barros Bouchet, M. I.; Matta, C.; Vacher, B.; Le-Mogne, T.; Martin, J. M.; von Lautz, J.; Ma, T.; Pastewka, L.; Otschik, J.; Gumbsch, P.; Moseler, M., Energy filtering transmission electron microscopy and atomistic simulations of tribo-induced hybridization change of nanocrystalline diamond coating. Carbon 2015, 87 (0), 317-329. Chen, M.; Briscoe, W. H.; Armes, S. P.; Klein, J., Lubrication at Physiological Pressures by Polyzwitterionic Brushes. Science 2009, 323 (5922), 1698-1701. Hirano, M.; Shinjo, K., Atomistic locking and friction. Physical Review B 1990, 41 (17), 11837-11851. Erdemir, A.; Mantin, J.-M., Superlubricity. Elsevier Academic Press: New York, 2007. Li, J.; Zhang, C.; Luo, J., Superlubricity Behavior with Phosphoric Acid-Water Network Induced by Rubbing. Langmuir 2011, 27 (15), 9413-9417. Zhang, R.; Ning, Z.; Zhang, Y.; Zheng, Q.; Chen, Q.; Xie, H.; Zhang, Q.; Qian, W.; Wei, F., Superlubricity in centimetres-long double-walled carbon nanotubes under ambient conditions. Nature Nanotechnology 2013, 8 (12), 912-916.


Page 23 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

16. 17. 18. 19. 20. 21.

Langmuir

Hu, Y. Z.; Wang, H.; Guo, Y.; Zheng, L. Q., Molecular dynamics simulation of ultra-thin lubricating films. Proceedings of the Institution of Mechanical Engineers Part J-Journal of Engineering Tribology 1998, 212 (J3), 165-170. Luo, J. B.; Wen, S. Z.; Huang, P., Thin film lubrication .1. Study on the transition between EHL and thin film lubrication using a relative optical interference intensity technique. Wear 1996, 194 (1-2), 107-115. Zhang, X. J.; Huang, Y.; Guo, Y. B.; Tian, Y.; Meng, Y. G., A dynamic rheological model for thin-film lubrication. Chin. Phys. B 2013, 22 (1), 8. Zhu, Y. X.; Granick, S., Superlubricity: A paradox about confined fluids resolved. Physical Review Letters 2004, 93 (9). Zhu, Y.; Granick, S., Reassessment of solidification in fluids confined between mica sheets. Langmuir 2003, 19 (20), 8148-8151. Jabbarzadeh, A. Harrowell, P, Tanner, I., The structural origin of the complex rheology in thin dodecane films: Three routes to low friction; Tribology International 2007, 40 (10-12) 1574-1586.

22.

Kano, M.; Yasuda, Y.; Okamoto, Y.; Mabuchi, Y.; Hamada, T.; Ueno, T.; Ye, J.; Konishi, S.; Takeshima, S.; Martin, J. M.; Bouchet, M. I. D.; Le Mogne, T., Ultralow friction of DLC in presence of glycerol mono-oleate (GMO). Tribology Letters 2005, 18 (2), 245-251.

23.

Amann, T.; Kailer, A., Ultralow Friction of Mesogenic Fluid Mixtures in Tribological Reciprocating Systems. Tribology Letters 2010, 37 (2), 343-352. Walter, M.; Amann, T.; Li, K.; Kailer, A.; Ruehe, J.; Moseler, M., 1,3-Diketone Fluids and Their Complexes with Iron. Journal of Physical Chemistry A 2013, 117 (16), 3369-3376. Li, K.; Amann, T.; Walter, M.; Moseler, M.; Kailer, A.; Ruehe, J., Ultralow Friction Induced by Tribochemical Reactions: A Novel Mechanism of Lubrication on Steel Surfaces. Langmuir 2013, 29 (17), 5207-5213. Gay, J. G.; Berne, B. J., Modification of the overlap potential to mimic a linear site–site potential. The Journal of Chemical Physics 1981, 74 (6), 3316-3319. Berardi, R.; Fava, C.; Zannoni, C., A Gay–Berne potential for dissimilar biaxial particles. Chemical Physics Letters 1998, 297 (1–2), 8-14. R. Berardi, C. Fava, and C. Zannoni. A gay-berne potential for dissimilar biaxial particles. Chem. Phys. Lett., 297(1–2), 1998. W. Michael Brown, Matt K. Petersen, Steven J. Plimpton, and Gary S. Grest, Liquid crystal nanodroplets in solution. J. Chem Phys., 130(4):044901, 2009. Lopez, G. P.; Castner, D. G.; Ratner, B. D., Xps O 1s Binding-Energies for Polymers Containing Hydroxyl, Ether, Ketone and Ester Groups. Surface and Interface Analysis 1991, 17 (5), 267-272. Mori, S.; Iwata, H., Relationship between tribological performance of liquid crystals and their molecular structure. Tribology International 1996, 29 (1), 35-39. Jaworski, A., Chemistry: the central science. 12th ed.; Prentice Hall: United States of America, 2011; p 429-430. Lord, J.; Larsson, R., Film-forming capability in rough surface EHL investigated using contact resistance. Tribology International 2008, 41 (9-10), 831-838. Hu, H.-W.; Carson, G. A.; Granick, S., Relaxation time of confined liquids under shear. Physical Review Letters 1991, 66 (21), 2758-2761. Barnes, H. A.; Walters, K., The yield stress myth? Rheol Acta 1985, 24 (4), 323-326.

24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.


Langmuir

Page 24 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 25 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Table of content graphics


Ultralow Friction of Steel Surfaces Using a 1,3 ... - ACS Publications

Recommend Documents