Are Buckminsterfullerenes Molecular Ball Bearings? - The Journal of

Subscriber access provided by UniSA Library

B: Fluid Interfaces, Colloids, Polymers, Soft Matter, Surfactants, and Glassy Materials

Are Buckminsterfullerenes “Molecular Ball Bearings”? Romain Lhermerout, Christophe Diederichs, Sapna Sinha, Kyriakos Porfyrakis, and Susan Perkin J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.8b10472 • Publication Date (Web): 10 Dec 2018 Downloaded from http://pubs.acs.org on December 12, 2018

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 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 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.

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 13 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

The Journal of Physical Chemistry

Are Buckminsterfullerenes Molecular Ball Bearings? †

Romain Lhermerout,

†

Christophe Diederichs,

‡

Sapna Sinha,

Kyriakos Porfyrakis,

‡

∗,†

and Susan Perkin

†Department

of Chemistry, Physical and Theoretical Chemistry Laboratory, University of Oxford, Oxford OX1 3QZ, UK

‡Department

of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, UK E-mail: [email protected]

Abstract

29

prehistorical ages to the harvesting of energy

30

with wind turbines in present times. The ques-

2

Buckminsterfullerenes (C60 ) are near-spherical 31

tion of whether logs were used as rollers to

3

molecules which freely rotate at room temper- 32

move megaliths has been debated,

4

ature in the solid state and when dissolved in 33

it is clear that the idea of using wheels to re-

5

solution.

An intriguing question arises as to 34

place sliding by rolling was early found to be

6

whether C60 molecules can act as molecular 35

an ecient way to reduce energy dissipation

7

ball bearings, i.e.

during motion, i.e.

8

between two solid surfaces whilst simultane- 37

ing the rst fundamental laws of solid friction,

9

ously dissipating shear stress through fast ro- 38

Da Vinci conceived of many ingenious machines

10

tation. To explore this, we performed measure- 39

involving rotary parts where the friction is lim-

11

ments of friction across a solution of C60 in the 40

ited to the axis.

12

boundary lubrication regime.

balls running along a groove (e.g.

13

shear and normal force measurements between 42

assembly), are designed to reduce the friction

14

mica sheets separated by the C60 solution were 43

further by transforming sliding into rolling. It

15

made using a Surface Force Balance to provide a 44

is also known, since ancient times,

16

single-asperity contact and sub-nanometer res- 45

bricating the contact, i.e.

17

olution in lm thickness.

We nd that, even 46

between the moving solids, is an ecient way

18

at small volume fraction, C60 forms a solid-like 47

to reduce wear and produce a moderate, or

19

amorphous boundary lm sustaining high nor- 48

at least stable friction. However, nding good

20

mal load, suggesting that this system undergoes 49

lubricants is complex because the appropriate

21

a glass transition under connement. The C60 50

mixture has to remain eective in harsh condi-

22

lm gives rise to a low friction coecient up 51

tions (high loads/shear stresses/temperatures,

23

to moderate applied loads, and we discuss the 52

humidity etc.), and the eld of tribology has

24

possible relevance of the ball bearing eect at 53

been active over the past century and until to-

25

the molecular scale.

54

day.

Introduction

55

friction are often complicated by processes act-

56

ing over multiple length-scales and timescales

57

simultaneously.

58

bricate motion as a thick lm (hydrodynamic

59

regime) or in molecular connement between

60

close asperities (boundary regime), either at dif-

1

26

preventing direct contact 36

High resolution 41

27

Human activity has always required the mo-

28

tion of objects, from the building of edices in

46

nonetheless

While formulat-

Ball bearings, consisting of in an axle

3

that lu-

inserting a liquid

Detailed, mechanistic interpretations of

ACS Paragon Plus Environment 1

2

friction.

1

For example a uid may lu-

The Journal of Physical Chemistry 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 13

61

ferent regions of a rough contact or under dif- 110

such nanocars was found to be indeed due to

62

ferent shearing conditions.

the rotation of the C60 on the Au-(111) sur◦ 15 face at ∼ 200 C. Experimental and numer-

63

111

Buckminsterfullerene (C60 ) is a molecule of 112

7

64

almost ideal spherical shape

(structure in Fig- 113

ical studies investigated the frictional behavior

65

ure 1(a)). In its pure form under ambient con- 114

on a C60 single crystal around an orientational

66

ditions C60 forms a solid in which the molecules 115

order-disorder phase transition at

67

are able to rotate freely; inspection of the tem- 116

no signicant change of friction coecient was

68

perature dependence of the heat capacity shows 117

found.

69

that the energy associated with rotation of C60 118

fullerene-like molecules have been used as an

70

in the crystal is smaller than the ambient ther- 119

additive in a liquid, they formed a protective

71

mal energy.

boundary lm that prevents wear and induces

72

bined with relative chemical stability, led to 121

73

the early proposal that C60 could act as an 122

Campbell et al.

74

eective boundary lubricant:

it might be ex- 123

in toluene between atomically-smooth mica sur-

75

pected that even when conned between two 124

faces in a Surface Force Apparatus, and studied

76

surfaces it could rotate to dissipate stress, the 125

the hydrodynamic lubrication regime by look-

77

molecules eectively performing as molecular 126

ing at the viscous response to a normal oscilla-

78

ball bearings. Direct observations have shown 127

tion. They showed that this system exhibits a

79

that spherical nanoparticles can rotate between 128

full slip boundary condition, suggesting a par-

80

sheared surfaces, in a manner which is remi- 129

ticular uidity of the C60 that are adsorbed

81

niscent of the rolling without slipping motion 130

on the surfaces.

82

of macroscopic ball bearings.

By analogy to 131

port the rst measurements of friction across a

83

the macroscopic mechanism of ball bearings, a 132

dispersion of C60 in the boundary lubrication

84

molecular ball bearing system should sustain 133

regime.

85

normal load (i.e.

86

surfaces apart) yet without hindering molec-

87

ular rotations (which serve to dissipate shear 134

88

stress and so reduce friction).

89

provides a useful motivating concept to con-

90

sider the eect of rotations in modifying fric-

91

tion at the nanoscale, we note that the macro-

92

scopic mechanism of ball bearings does not map

93

perfectly onto molecular systems under ambient

94

conditions: the molecular rotation rate will typ-

95

ically be much higher than the imposed shear

96

under ambient conditions. To dierentiate be-

97

tween these two situations, we call the molec-

98

ular mechanism the molecular ball bearing ef-

99

fect.

100

8

This interesting property, com- 120

9

keeping the shearing solid

Although this

In this work we measure and appraise

whether C60 might satisfy these criteria.

101

Previous studies have investigated lubrica-

102

tion by C60 or fullerene-like molecules in dif-

103

ferent ways.

104

used as a solid lubricant in dry conditions

105

(from

106

sublimated thin lm),

107

vide particularly exceptional frictional proper-

108

ties.

109

molecule nanocars, and the displacement of

simply

1014

On one hand, they have been dispersed

powder

to

carefully

but they didn't pro-

C60 have been used as wheels for single-

16,17

∼ 260 K, but

On the other hand, when C60 or

a stable frictional response.

1820

In particular,

conned a dispersion of C60

18

In the present paper, we re-

Methods

135

Materials.

136

discharge method according to the procedure

137

rst

138

was further isolated by high performance liq-

139

uid chromatography (HPLC) to a purity of

140

99.5%. Tetralin, 1,2,3,4-tetrahydronaphthalene

C60 was synthesised via the arc

published

by

Krätschmer

21

and

141

(Sigma-Aldrich,

as solvent (chemical structure in Figure 1(a)).

143

Tetralin was chosen as a good solvent for C60

144

147

(solubility of C60 in tetralin is 16 mg/mL 25 ◦ C 22 ) with low volatility (vapor pres◦ 23 sure of 0.05 kPa at 25 C ) and a mod◦ 24 erate viscosity (2.015 mPa.s at 25 C ).

148

The tetralin was dried with molecular sieves

149

(0.4

150

for a week,

146

99%),

al.

142

145

anhydrous,

et

was used

at

nm

pore

size,

from

Fisher

Chemical)

and ltered before use (Ultra-

0.22 µm).

151

Cruz Syringe Filter, PTFE,

152

molecules were dispersed at a concentration of

153

5.60 ± 0.01 mg/mL, corresponding to a molar fraction of 0.1065 ± 0.0002% (given the ◦ 23 tetralin density of 0.9645 g/mL at 25 C ), or

154 155


C60

Page 3 of 13

The Journal of Physical Chemistry 6

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

3

0

-3 0

Figure

1:

(a)

Chemical

structures

and

sizes

of

2

4

buckminsterfullerene

6

(C60 )

8

and

10

tetralin.

(b) Schematic of the SFB experiment, that allows to determine the interaction and friction forces between two mica surfaces separated by a liquid lm of controlled thickness. (c) Normal force

R of vN ∼ 1 nm/s, for

D

FN

rescaled by the radius of curvature

the surfaces as a function of the separation

obtained

at approach velocities

tetralin (approach in brown, pull-o force measured on

retraction indicated by the black arrow) and the solution of C60 in tetralin (approach in purple, retraction in black).

156

a mean distance between the C60 molecules of 185

and shear force can be measured in parallel (e.g.

157

5.979 ± 0.004 nm ∼ 5 × (C60 diameter).

186

during approach of the surface from large dis-

The measurements 187

tances to contact). The details of the procedure

158

Force measurements.

2528

159

were performed with a Surface Force Balance 188

have explained elsewhere;

160

(SFB), which is a method ideal for the study 189

we note the quantities and details particular to

161

of normal and lateral forces transmitted across 190

the present experiments.

162

uid and soft lms with high resolution.

A 191

Muscovite mica is cleaved, backsilvered and

163

schematic diagram of the key aspects is in Fig- 192

glued on glass cylindrical lenses using dextrose,

164

ure 1(b).

D-(+)-glucose (Sigma-Aldrich, 99.5%, chosen

165

optical lenses, hemi-cylindrical in shape (ra- 194

166

dius of curvature

167

crossed-cylinder conguration.

The liquid lm is held between two 193

R ∼ 1 cm)

in the following

for its insolubility in tetralin).

Two surfaces

and arranged in 195

are mounted in a crossed-cylinder geometry

This geometry 196

to make a single contact between atomically

168

provides a point of closest approach between 197

smooth surfaces, and the liquid is injected in

169

the two surfaces, which is model experiment for 198

between to form a capillary bridge. The cham-

170

study of a single-asperity contact. The optical 199

ber is dried with P2 O5 , phosphorus pentoxide

171

lenses are coated with single crystal sheets of 200

172

mica, so that the roughness is sub-molecular 201

(Sigma-Aldrich, 99%) and the room is regulated ◦ to 25 C. FECO are analyzed to measure the

173

(roughness arising only from the atomic cor- 202

radius of curvature

174

rugation of the crystalline surfaces). The pre- 203

liquid thickness,

175

cise geometry and liquid thickness are measured 204

176

directly in-situ using white light interferome- 205

R of the surfaces and the D. D is measured with a precision of 0.02 nm (RMS noise) and accuracy of 1 nm and D = 0 is dened as the mica-mica

177

try; so-called Fringes of Equal Chromatic Or- 206

contact position measured in dry air before liq-

178

der (FECO). The surfaces (lenses) can be trans- 207

179

lated in both normal and lateral directions rel- 208

uid injection. The refractive indeces of 1.5413 ◦ 23 ◦ at 20 C for tetralin and 1.5417 at 22 C for

180

ative to one another, and the resulting forces 209

the mixture were measured with a Bellingham+

181

between them are detected via the deection 210

Stanley Abbe 60 ED refractometer.

182

of normal and lateral springs. These measure- 211

per motor is used to approach or retract the

183

ments can be performed simultaneously, so that 212

top surface at a normal velocity

184

information about lm thickness, normal force, 213

tored piezo-electric tube allows application of


vN

29

A step-

and a sec-


vL .

214

a shearing motion at lateral velocity

215

mal force

216

sured using springs, with respective stiness of 262

is reached at

217

133.8 ± 3.0 N/m

263

nm with increased load (appearing even more

264

clearly in Figure 3(b)).

265

sion corresponds to the thickness of approxi-

266

mately 5 C60 molecules, and the soft wall holds

267

for loads up to

268

sure of

218

FN

and lateral force and

FL

Nor- 260

Page 4 of 13

are then mea- 261

441 ± 4 N/m.

Results

219

Normal force.

220

tained are shown in Figure 1(c).

221

tetralin,

222

Waals) attractive interaction on approach of

223

the surfaces to

224

to contact.

225

the jump-in is close to the direct mica-mica

226

contact value (within the systematic experi-

227

mental error).

228

wall is due to the compression of the mica,

229

the single material remaining in the optical

230

interferometer.

231

faces jump-out when reaching a pull-o force

232

of

233

the JKR theory corresponds to an adhesion en-

234

ergy of

235

This value is comparable with the adhesion en-

236

ergy between

237

measured previously for mica-mica contact in

238

dry nitrogen.

239

force is not observed in this case, as might be

240

expected by comparison with similar measure-

241

ments with apolar liquids like cyclohexane, ben-

242

zene or toluene which do each give rise oscilla-

243

tory structural surface forces .

244

a similarly strong adhesion minimum has also

245

been observed for these liquids, and attributed

246

to the presence of traces of water that wets the

247

mica surfaces. Particular care was taken to use

248

tetralin in dry conditions (storage in molecular

249

sieves, measurements with P2 O5 in the cham-

250

ber), however it is still possible that traces of

251

water may remain giving rise to an adsorbed

252

layer or part-layer on the mica.

253

sequent experiments we take this control mea-

254

surement with pure tetralin as the reference sit-

255

uation in order to investigate the eect of the

256

addition of small quantities of C60 under the

257

same conditions.

The normal force proles obFor pure

the surfaces experience a (van der

∼ 10 nm

causing a jump-in

The surface separation just after

The nite gradient of this soft

On retraction the two sur-

Fadh /R = −435 mN/m,

which according to

W = 2Fadh /(3πR) = −92 mN/m. 25 −160 mN/m

30

and

−108 mN/m

It is not clear why a structural

26,3135

However,

In the sub-

258

When 0.1 mol % C60 is added to the tetralin

259

the normal force between the surfaces is mod-

ied signicantly, as clear in Figure 1(c).

On

approach of the surfaces a repulsive soft wall

∼ 6 nm,

compressing by several The onset of repul-

270

80 µN, corresponding to a pres∼ 6 MPa (given the contact radius of ∼ 2 µm), and on retraction an adhesive minimum of Fadh /R = −2.7 mN/m is obtained.

271

This value is comparable with what Campbell

272

et al.

273

mica.

269

274

18

measured for C60 in toluene between

In summary of the normal force proles, we

275

nd that very small volume fractions of C60

276

lead to substantial modication to the inter-

277

action force between conning surfaces, giving

278

rise to a monotonic repulsive force extending to

279

∼ 5

280

us subsequently to interpret the direct friction

281

measurements, as follows.

282

molecular diameters.

Lateral force.

This insight allows

Having characterised the nor-

283

mal interaction between the surfaces, we next

284

applied a lateral (shearing) motion of the top

285

surface relative to the bottom surface and de-

286

tected the resulting lateral force transmitted

287

across the liquid. In the SFB this can be per-

288

formed at the same time as approaching the sur-

289

faces and measuring the lm thickness and nor-

290

mal force; in this section we present results of

291

the measured friction as a function of separation

292

and load. We compare the case of pure tetralin,

293

as control, to the tetralin with C60 .

294

ure 2 we show the result of measurements with

295

pure tetralin (part (a)) and with C60 in tetralin

296

(part(b)). For the control experiment with pure

297

tetralin, we show the temporal evolution of the

298

liquid thickness

299

the top surface is moved downward and then

300

upward at

In Fig-

D and the lateral force FL when

vN = 0.92±0.05 nm/s and simultaneoscillated laterally at vL = 287 ± 1 nm/s.

301

ously

302

When the surfaces are separated by a nite

303

lm of tetralin, before reaching surface con-

304

tact, no measurable lateral force is detected

305

(smaller than the sensitivity of

306

ing there is no mechanical coupling between

307

the surfaces.

308

tact the lateral force instantaneously increases


∼ 1 µN), mean-

At the point of jump-in to con-

Page 5 of 13 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


30 20 10

0 150 100 50 0

-50

-100 -150

0

5

150

10

565

15

570

575

580

100

40

50 0

30

-50

-100

20

-150 410

10 0

100

415

200

420

425

300

Figure 2: (a) Temporal evolutions of liquid thickness

650

400

655

500

660

665

600

700

D (red traces) and lateral force FL (blue traces)

when approaching and retracting mica surfaces separated by tetralin with simultaneous constantvelocity shearing of one surface. The lateral force is below the experimental resolution until the surfaces reach direct contact, at which point they are rigidly coupled. (b) Temporal evolution of liquid thickness

D

when approaching the surfaces and simultaneously applying constant-velocity

lateral motion to one surface across the solution of force

FL

C60

at two time intervals (or equivalently two loads

in tetralin.

FN

The insets show the lateral

as indicated), showing how the lateral

force during shearing cycles evolves as the applied load is increased. At higher loads a clear yield spike followed by smooth sliding is observed; from this the kinetic friction


FL,k

is deduced.


Page 6 of 13

309

in magnitude and varies directly with the ap- 358

which point the layer is already substantially

310

plied shearing amplitude; the saw-tooth shape 359

compressed and the load is high, qualitatively

311

is the signature that the surfaces are rigidly cou- 360

similar to what has been observed for lubrica-

312

pled, i.e with no relative motion throughout the 361

tion by polymer brushes.

313

cycle. Because of the strong adhesion between 362

ure 3(a), the relationship between the kinetic

314

the surfaces, friction is controlled by adhesion

363

friction force and the load is not linear, instead

315

and is so high that the yield point of the con- 364

it has a strongly convex shape. At small loads,

316

tact is not reached within the range of lateral 365

kinetic friction is proportional to the load, with

317

force explored. Thus we nd that the yield force 366

no signicant contribution from adhesion (zero

318

of the contact must be higher than

319

this case.

320

ing amplitudes, we found that the yield force 369

321

was in fact higher than

322

to a lower limit for the contact shear stress of 371

with the load, and reaches

323

σL ∼ 26 MPa.

During retraction, a damped os- 372

maximum imposed load. Thus the friction coef-

324

cillation at the resonance frequency of

∼ 25 Hz 373

cient evolves from a very low value, indicating

325

is obtained when the surfaces jump-out, and 374

ecient lubrication at moderate load, up to a

326

then no lateral force is detected, consistently 375

high value under strong compression. To com-

327

with the behavior on approach. In sum, we nd 376

ment on these values, it is useful to compare

328

that tetralin alone cannot support any applied 377

them to the friction coecient of dierent liq-

329

load, and so squeezes out of a contact when - 378

uids between mica surfaces, all measured with

330

nite load is applied, giving rise to high friction 379

a Surface Force Balance. Simple apolar liquids

331

in accordance with direct mica-mica contact.

380

are generally characterized by a single friction

332

Lateral forces measured across the solution of 381

coecient: 1.1 for octamethylcyclotetrasiloxan

333

C60 in tetralin as a function of load - examples 382

(OMCTS), 2.2 for cyclohexane.

334

of which are in Figure 2(b) - show an entirely 383

exhibit quantized friction,

335

dierent behavior. When the surfaces approach 384

ecient indexed by the number of ordered lay-

336

to the distance corresponding to the repulsive 385

ers of ions in the lm, typically varying from

337

wall the lateral force is still below the sensi- 386

0.007 to 0.5 for 1-decyl-1-methylpyrrolidinium

338

tivity limit.

bis[(triuoromethane)sulfonyl]imide, [C10 C1 Pyrr]

339

load before any detectable lateral force could 388

[NTf2 ].

340

be recorded between the surfaces; the left-hand 389

cient, the C60 solution is thus between apo-

341

inset show the emergence of tiny lateral forces 390

lar liquids and ionic liquids, and is comparable with the 0.12 obtained for 2,6,10,15,19,23-

110 µN

36

in 367

In experiments with higher shear- 368

330 µN;

corresponding 370

It was necessary to increase the 387

37

As shown in Fig-

friction at zero imposed load) and a coecient of proportionality

µ = 0.072 ± 0.002

(with

the coecient of friction dened as the local slope

µ = dFL,k /dFN ).

40

This quantity increases

µ = 4.1 ± 1.6

39

38

i.e.

at the

Ionic liquids a friction co-

Regarding the low-load friction coef-

342

when the surfaces are being pushed together 391

343

with a force of

The amplitude of the 392

hexamethyltetracosane (squalane), a branched

344

lateral force then increases with the load, and 393

hydrocarbon liquid that has been reported for

345

the signal exhibit a strong stiction spike fol- 394

exhibiting glassy behavior in certain conne-

346

lowed by a plateau (right-hand side inset). The 395

ment conditions.

347

clear yield point followed by sliding behavior is

348

typical of a solid-like response to lateral applied

349

stress. We systematically extracted the ampli- 396

350

tude of the plateau, which we identied as the

351

kinetic friction force

352

compare the normal force and kinetic friction

353

force proles as a function of surface separation,

354

D.

355

the range of the friction force is much smaller

356

than the range of the normal force, becoming

357

measurable only at separations of

28 µN.

FL,k .

In Figure 3(b) we

In this representation, we clearly see that

∼ 4 nm

at

41

Discussion

397

A solid lm.

398

itative picture of what is happening at the

399

molecular scale to interpret the observed behav-

400

ior. The molecular forces governing the inter-

401

action between C60 , tetralin and mica include

402

van des Waals and steric forces.

403

lar, and maybe covered with a (sub-)molecular


We now propose a simple qual-

Mica is po-

Page 7 of 13 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


80 60 80 40 20 60 0 3

4

5

6

40

20

0

20

Figure 3: (a) Kinetic friction

40

FL,k

60

80

as a function of load

FN .

Friction in pure tetralin was above

the range limit of the experiment, as indicated by the shaded bar, whereas the solution of C60 in tetralin gave rise to measurable friction (in purple). The two stars correspond to the lateral force traces shown in insets of Figure 2(b), and the red lines are the linear ts used to deduce the local friction coecient

µ.

(b) Normal (red) and kinetic friction (blue) force proles, for the solution of

C60 in tetralin. (c) Schematic representation of the system in connement: the C60 molecules are randomly packed, steric interactions lead to a signicant repulsion between the solid surfaces, but fast rearrangements induce a relatively small shearing resistance.



Tetralin is slightly polar and 453

Page 8 of 13

404

lm of water.

405

served a structural force over the explored spots

406

polarizable (calculated static polarizability of 454 1.8 × 10−39 C · m2 /V), whereas C60 is more 455

407

polarizable because of the highly delocalized 456

tween

408

likely to be due to the dierent solvent and

409

π electrons (measured static polarizability of 457 8.6 × 10−39 C · m2 /V 42 ). When the mica sur- 458

410

faces are far apart, the C60 molecules in the 459

411

bulk are separated by an average distance of 460

412

about 5 times their diameter (deduced from the 461

413

chosen concentration) and are attracted by a 462

414

dispersion (London) interaction but the liquid 463

415

dispersion is thermodynamically stable because 464

and having no time to order.

We never ob-

on the surfaces, experiments and velocities be-

∼ 1 nm/s

and

∼ 12 nm/s.

concentration chosen:

This is most

Campbell et al.

used

416

the concentration is (just) below the saturation 465

0.095 mg/mL corresponding to a molar fraction of ∼ 0.001% or ∼ 3% of the saturation limit in toluene (2.8 mg/mL ◦ 22 at 25 C ), whereas here we use a concentration of 5.60 mg/mL corresponding to a molar fraction of ∼ 0.1% or ∼ 35% of the saturation ◦ 22 limit in tetralin (16 mg/mL at 25 C ). In our

417

limit. A monolayer of C60 is probably initially 466

case, the C60 molecules in the bulk are much

418

adsorbed on each mica surface due to the induc- 467

more concentrated and close to the agglomera-

419

tion (Debye) interaction and the preference of 468

tion limit, and the order/disorder kinetic tran-

420

mica for the more polar C60 rather than tetralin. 469

sition may have been shifted to approach ve-

421

When conning the liquid, tetralin tends to be 470

locities much smaller than

422

squeezed-out, as shown by the reference mea- 471

the observations of no structural force and of

423

surement.

a solid-like friction response, we conclude that

424

the gap therefore increases as the surfaces ap- 473

under the experimental conditions used the C60

425

proach and the C60 molecules - which are al- 474

must form a disordered/amorphous solid, i.e. a

426

ready close to the aggregation limit in the bulk 475

glass, where the molecules are randomly packed

427

solution - are expected to agglomerate, forming 476

in the gap (as illustrated in Figure 3(c)). The

428

a strongly bound solid lm (as shown by the 477

situation is thus related to the dicult problem of glass transition in connement, already

The local concentration of C60 in 472

a concentration of

1 nm/s.

Combining

429

soft wall in the normal force prole). The fact 478

430

that this lm holds for pressures up to

431

is remarkable, given that a yield stress of about 480

liquids,

432

∼ 1 MPa

loids.

433

single crystal along the

This 482

sured kinetic friction-load curve as a kind of

∼ 6 MPa 479

is measured when compressing a C60 481

h1 1 0i direction. 43

investigated for various systems

48

45

polymers,

46

44

like simple

liquid crystals

47

or col-

It is then tempting to interpret our mea-

434

is probably due to the fact that we are dealing 483

Angell plot, that conventionally represents how

435

with a nanometric lm, the C60 having attrac- 484

the logarithm of the viscosity increases with the

436

tive interactions with the conning surfaces and 485

inverse temperature (for molecular liquids) or

437

being packed in a disordered arrangement with 486

with the packing fraction (for colloidal liquids)

438

no cleaving plane.

when approaching the glass transition.

A disordered lm.

487

49,50

Fol-

The absence of order 488

lowing this scenario, it would be the increase

440

in the lm is revealed by the normal force pro- 489

of the C60 packing fraction with the load (as

441

le showing a soft wall at distance of approx- 490

shown by the compression of the lm) that leads

442

imately 5 C60 diameters.

This behavior was 491

to a dramatic slowing down of the dynamics,

443

unexpected, because a structural force is usu- 492

i.e. of the rearrangement timescale of the glassy

444

ally observed for apolar liquids between mica. 493

lm in the gap, as measured by the increase of

445

When Campbell et al.

the friction force.

446

C60 in toluene between mica, they observed 495

447

a soft wall at

whether the rotation of C60 contributes to the

448

ties higher than

frictional behavior. As shown by a 2D molecu-

449

of period of

439

studied a solution of 494

∼ 3.4 nm at approach veloci- 496 5 nm/s, or a structural force 497 1.1 nm equal to the diameter of 498 18

Role of rotation.

One can nally ask

lar dynamics simulation of lubrication by cir-

450

the C60 molecules at smaller velocities.

They 499

cular molecules, a high concentration in the

451

interpret this transition as C60 molecules be- 500

gap can lead to jamming, which hinders the

452

ing kinetically trapped between the surfaces 501

molecular rotation because the two sides of two


Page 9 of 13 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


502

molecules in contact would have to rotate in 548

molecular ball bearing eect may be impor-

503

opposite directions.

tant for reducing shear stress.

504

plays a role in our system, it is likely to happen 550

Future investigations should involve a sys-

505

at small loads, for which we indeed measured a 551

tematic study of the eect of the concentra-

506

small friction coecient.

tion on kinetic trapping and on friction re-

507

tively strong adhesion of the C60 on mica (De- 553

sponse.

508

bye interaction), the lateral motion is probably 554

tion, comparative Surface Force Balance mea-

509

distributed in the middle of the gap and not 555

surements with C60 chemically grafted to the

510

at the lm/mica interfaces.

surfaces and thus unable to rotate

511

tance to ow in the gap indicates that energy 557

performed. It would also be of interest to study

512

is dissipated eciently in the lm; the mech- 558

513

anism for this could originate from (at least) 559

dierent surface materials such as graphene; 2 with only sp -hybridized carbon materials ex-

514

three modes. First, there is only small adhesion 560

ceptional properties can emerge.

515

between C60 molecules (London interaction).

516

Second, there is room and time for rearrange- 562

R.L. and C.D. performed the experiments. S.P.

517

ments, since C60 molecules are randomly packed 563

and R.L. conceived of the project, interpreted

518

with vacancies, at a packing fraction far enough 564

the data and wrote the paper.

519

from the glass transition. Third, the C60 may 565

synthesised the fullerenes and contributed to

520

rotate freely in the thin disordered lm, like 566

the experimental design.

521

they do in the bulk crystal, and this free ro-

522

tation could contribute to a dissipation mecha-

523

nism leading to low friction. In other words, a

524

molecular ball bearing eect may be occurring

525

and be partly responsible for the low friction

526

regime.

527

51

So if the rotation of C60 549

Because of the rela- 552

The small resis- 556

18

561

To clarify the role of molecular rota-

52,53

could be

54

5557

Author Contributions

Acknowledgement ported

The

K.P. and S.S.

S.P. and R.L. are sup-

Leverhulme

Trust

(RPG-2015-

328) and the ERC (under Starting Grant No. 676861, LIQUISWITCH). R.L. is supported by the EPA Cephalosporin Junior Research Fellowship and Linacre College (University of Oxford). S.P. is grateful for research leave enabled

Conclusions

by the Philip Leverhulme Prize.

References

528

To summarize our ndings, C60 performs as

529

an excellent boundary lubricant additive in our

530

system;

531

strong boundary lm that reduces wear by pre-

532

venting direct contact between the sliding sur-

533

faces and gives rise to a small friction coecient

534

at moderate loads. A key ingredient leading to 571

535

this useful outcome is that C60 has a nite sol- 572

Studies of Friction.

536

ubility and tends to solidify when the concen- 573

51 66.

537

tration exceeds the saturation point locally in

538

conned contacts thus preventing squeeze-out 574

539

of the lm.

The fact that the C60 appears to 575

Lenoir, N.; Shahidzadeh, N.; Fiscina, J.;

540

form a disordered solid, evidenced by no regu- 576

Wagner, C.; Bonn, D. Sliding Friction on

541

lar oscillatory structural force despite solid-like 577

Wet and Dry Sand.

542

mechanical response, indicates that the convex 578

543

friction-load relationship may be the manifes-

544

tation of a glass transition under connement.

545

The friction coecient at moderate loads is re-

546

markably small, and so it is hypothesised that

547

the molecular rotations intrinsic to C60 the

∼ 0.1%

of C60 is sucient to form a

567 568

(1) Richards, J.; Whitby, M. In

Science and

569

Stonehenge ;

570

Eds.; British Academy, 1997; pp 231256.

579

(2) Hutchings,

(3) Fall,

A.;

Cunlie, B.,

I.

M.

Weber,

Renfrew, C.,

Leonardo

Wear

B.;

da

2016,

Vinci's

360-361,

Pakpour,

M.;

Phys. Rev. Lett. 2014,

112, 175502. (4) Bowden, F. P.; Tabor, D.

The Friction and

580

Lubrication of Solids ; Oxford University

581

Press, 1950.



582

(5) Urbakh, M.; Klafter, J.; Gourdon, D.; Is- 625

583

raelachvili, J. The Nonlinear Nature of 626

584

Friction.

Nature

2004, 430, 525.

627

Page 10 of 13

Friction of Smooth C60 Surfaces.

Mater. (15) Shirai,

1997, 9, 11661171. Y.;

Osgood,

A.

J.;

Chem.

Zhao,

Y.;

585

(6) Lhermerout, R.; Diederichs, C.; Perkin, S. 628

Kelly, K. F.; Tour, J. M. Directional Con-

586

Are Ionic Liquids Good Boundary Lubri- 629

trol in Thermally Driven Single-Molecule

587

cants?

588

cants

A Molecular Perspective.

2018, 6 .

Lubri- 630

Xiao, X. Eect of C60 Molecular Rotation

(7) Kroto, H. W.; Heath, J. R.; O'Brien, S. C.; 632 Curl, R. F.; Smalley, R. E. C60 : Buckmin- 633

591

sterfullerene.

592

(8) Bagatskii,

M.

I.;

Dolbin,

V.

V.; 635

A.

V.; 636

Barabashko,

594

Sundqvist, B. The Low-Temperature Heat 637

595

Capacity

596

Phys.

597

S.;

634

593

of

M.

Sumarokov,

Fullerite

C60 .

2015, 41, 630636.

(9) Lahouij,

I.;

Dassenoy,

J.-M.

Real

Low Temp. 638 639

F.;

Time

Vacher, TEM

2005, 5, 23302334.

(16) Liang, Q.; Tsui, O. K. C.; Xu, Y.; Li, H.;

589

1985, 318, 162.

Nano Lett.

631

590

Nature

Nanocars.

B.; 640

on Nanotribology.

2003,

Phys. Rev. Lett.

90, 146102. (17) Benassi, A.; Vanossi, A.; Pignedoli, C. A.; Passerone, D.;

Tosatti, E. Does Rota-

tional Melting Make Molecular Crystal Surfaces More Slippery?

2014,

Nanoscale

6, 1316313168. (18) Campbell,

S.

E.;

Luengo,

G.;

Sr-

598

Martin,

Imag- 641

danov, V. I.; Wudl, F.; Israelachvili, J. N.

599

ing of Compression and Shear of Single 642

Very Low Viscosity at the Solid-Liquid In-

600

Fullerene-Like MoS2 Nanoparticle.

Tribol. 643

terface Induced by Adsorbed C60 Mono-

601

Lett.

602

2012, 45, 131141.

644

layers.

(10) Blau, P. J.; Haberlin, C. E. An Investiga- 645

(19) Golan,

603

tion of the Microfrictional Behavior of C60 646

604

Particle Layers on Aluminum.

605

Films

1992, 219, 129 134.

R.;

Fullerene-Like

Mater.

609

and Mica.

610

891893.

1993,

63, 652

611

(12) Bhushan, B.; Gupta, B. K. Friction and

612

Wear of Ion-Implanted Diamondlike Car-

613

bon and Fullerene Films for Thin-Film

614

Rigid Disks.

615

61566158. E.;

Tenne,

650

Investigation of C60 Adsorbed on Silicon 651

Meyer,

Y.;

Homy-

649

Compton, R. N. Atomic Force Microscope

R.;

Feldman,

C.;

rect Force Measurement of WS2 Nested

608

(13) Lüthi,

M.;

Drummond,

648

607

616

Y.;

Israelachvili, J. Microtribology and Di-

(11) Thundat, T.; Warmack, R. J.; Ding, D.;

J. Appl. Phys.

1996, 382, 520.

Thin Solid 647

606

Appl. Phys. Lett.

onfer,

Nature

1994,

Haefke,

Nanostructures.

1999, 11, 934937.

(20) Drummond, raelachvili,

C.; J.;

Alcantar,

Tenne, and

R.;

653

Microtribology

654

Material Transfer in WS2

655

Additives.

656

348354.

Adv.

N.;

Is-

Golan,

Y.

Friction-Induced Nanoparticle

Adv. Funct. Mater.

2001,

11,

75, 657

(21) Krätschmer, W.; Lamb, L. D.; Fostiropou-

658

los, K.; Human, D. R. Solid C60 : A New

659

Form of Carbon.

H.;

Nature

1990, 347, 354.

617

Howald, L.; Gutmannsbauer, W.; Guggis- 660

(22) Ruo, R. S.; Tse, D. S.; Malhotra, R.;

618

berg, M.; Bammerlin, M.; Güntherodt, H.- 661

Lorents, D. C. Solubility of C60 in a Vari-

619

J. Nanotribology: an UHV-SFM Study on 662

ety of Solvents.

620

Thin Films of C60 and AgBr.

621 622

1995, 338, 247 260. (14) Luengo,

G.;

Campbell,

Surf. Sci. 663 664

S.

E.;

Sr- 665

623

danov, V. I.; Wudl, F.; Israelachvili, J. N. 666

624

Direct Measurement of the Adhesion and

1993, 97,

33793383. (23) Lide, D. R.

CRC Handbook of Chemistry

and Physics, 90th Edition (CD-ROM Version 2010) ; CRC Press, 2010.


J. Phys. Chem.

Page 11 of 13 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

667


(24) Gonçalves,

668

gers,

669

Tetralin and

670

mophys.

671

J.

F.

V.

A.;

Hamano,

Density

and

K.;

Sen- 710

Viscosity

of 711

turing in Liquid Alkanes Between Solid Surfaces: Force Measurements and Mean-

714

Field Theory.

J.

N.

Intermolecular and 715

Surface Forces (Third Edition) ; Academic

673

Press, 2011. J.;

Kumacheva,

E.

Simple

Liq-

675

uids Conned to Molecularly Thin Lay-

676

ers.

677

Solid Phase Transitions.

678

I.

Connement-Induced

1998, 108, 69967009.

Liquid-to-

J. Chem. Phys.

680

1987,

J. Chem. Phys.

87,

18341841.

716

(35) Christenson, H. K.; Blom, C. E. Solvation

717

Forces and Phase Separation of Water in

718

a Thin Film of Nonpolar Liquid Between

719

Mica Surfaces.

720

419424.

721

(36) Homola,

722 679

Israelachvili, J. N. Struc-

713

(25) Israelachvili,

(26) Klein,

Horn, R. G.;

Gruen, D. W. R.;

Trans -Decalin. Int. J. Ther- 712

1989, 10, 845856.

672

674

(34) Christenson, H. K.;

(27) Perkin, S.; Chai, L.; Kampf, N.; Raviv, U.; 723 Briscoe, W.; Dunlop, I.; Titmuss, S.; 724

A.

J. Chem. Phys.

M.;

1987,

Israelachvili,

J.

86,

N.;

McGuiggan, P. M.; Gee, M. L. Fundamental

Experimental

Studies

in

Tribology:

The Transition from Interfacial Friction

Seo, M.; Kumacheva, E.; Klein, J. Forces 725 Between Mica Surfaces, Prepared in Dif- 726

of Undamaged Molecularly Smooth Sur-

Wear

684

ferent Ways, Across Aqueous and Non- 727 aqueous Liquids Conned to Molecularly

685

Thin Films.

681 682 683

686 687 688 689 690 691

2006,

Langmuir

22, 6142 728

6152.

2018, 3, 014201.

to Normal

Friction

with Wear.

1990, 136, 65 83.

(37) Klein, J.;

Kumacheva, E.;

Mahalu, D.;

729

Perahia, D.; Fetters, L. J. Reduction of

730

Frictional Forces Between Solid Surfaces

(28) Lhermerout, R.; Perkin, S. Nanoconned 731 Ionic Liquids: Disentangling Electrostatic 732 and Viscous Forces.

faces

Phys. Rev. Fluids

Nature

Bearing Polymer Brushes.

1994,

370, 634636.

733

(38) Kumacheva, E.; Klein, J. Simple Liquids

734

Conned to Molecularly Thin Layers. II.

(29) Israelachvili, J. N. Thin Film Studies Us- 735

Shear and Frictional Behavior of Solidied

692

ing Multiple-Beam Interferometry.

693

loid Interface Sci.

J. Col- 736

1973, 44, 259 272.

737

Films.

J. Chem. Phys.

1998,

108, 7010

7022.

694

(30) Horn, R. G.; Israelachvili, J. N.; Pribac, F. 738

(39) Smith, A. M.; Lovelock, K. R. J.; Gos-

695

Measurement of the Deformation and Ad- 739

vami, N. N.; Welton, T.; Perkin, S. Quan-

696

hesion of Solids in Contact.

J. Colloid In- 740

tized Friction Across Ionic Liquid Thin

697

terface Sci.

698

1987, 115, 480 492.

(31) Horn, R. G.; Israelachvili, J. N. Direct Forces

742

15, 1531715320.

Be- 743

2013,

Measurement

700

tween Two Surfaces in a Nonpolar Liquid. 744

Molecular

701

J. Chem. Phys.

745

Nanolms of a Bilayer-Forming Ionic Liq-

746

uid.

747

4037.

(32) Christenson,

Structural

Films.

699

702

of

Phys. Chem. Chem. Phys.

741

1981, 75, 14001411.

H.

K.;

Horn,

R.

G.;

Is-

703

raelachvili, J. N. Measurement of Forces

704

Due to Structure in Hydrocarbon Liquids. 748

705

J. Colloid Interface Sci.

1982, 88, 79 88. 749

706

(33) Christenson, H. K. Experimental Mea-

707

surements of Solvation Forces in Nonpolar

708

Liquids.

709

6913.

J. Chem. Phys.

1983,

(40) Smith, A. M.; Parkes, M. A.; Perkin, S. Friction

Mechanisms

J. Phys. Chem. Lett.

(41) Gourdon, D.; sitions

2014, 5, 4032

Israelachvili, J. N. Tran-

Between

Smooth

and

750

Stick-Slip Sliding of Surfaces.

751

E

78, 6906 752 753

2003, 68, 021602.

(42) Antoine, R.;

Dugourd, P.;

Complex

Phys. Rev.

Rayane, D.;

Benichou, E.; Broyer, M.; Chandezon, F.;


Across


754

Guet,

C.

755

Electric

Direct

Measurement

Polarizability

of

of

the 797

Films: C60 Anchored to a Self-Assembled

Isolated

C60 798

Monolayer.

1999,

J. Chem. Phys.

756

Molecules.

757

97719772.

110, 799 800

758

(43) Fomenko, L.;

Lubenets, S.;

R.;

Sidorov,

N.

tion

Properties of C60 Single Crystals.

761

Sci. Eng. A.

762

(44) Alba-Simionesco,

Mater. 803

2005, 400-401, 320 324. Coasne,

Shon,

S.

and

S.

S.,

Y-S;

Structural

Frictional

Lee,

Properties

Terminated

Self-Assembled

804

on Au(111).

Thin Solid Films

805

152 158.

Gub- 806

T.

R.;

Characterizaof

C60 -

Monolayers

2000, 358,

Dosseh,

764

bins, K. E.; Radhakrishnan, R.; Sliwinska- 807

Coles,

765

Bartkowiak, M. Eects of Connement on 808

Babenko, V.; Murdock, A. T.; Koós, A.;

766

Freezing and Melting.

767

Matter

2006, 18, R15.

G.;

B.;

12, 3905

763

768

Dudziak,

S.;

Mechanical 802

Nikolaev,

G.;

(53) Lee,

Parry,

759

1996,

Langmuir

3911.

Izotov, A.; 801

760

C.;

Page 12 of 13

J. Phys. Condens. 809

(45) Rosenhek-Goldian, I.; Kampf, N.; Yere-

Friction.

Proc. Natl. Acad. Sci. U.S.A. 814

(46) Ellison, C. J.; Torkelson, J. M. The Distri-

774

bution of Glass-Transition Temperatures

775

in Nanoscopically Conned Glass Form-

776

ers.

777

Nat. Mater.

The

Connement-Induced

779

Lamellar

Phase

Transition.

1999, 15, 77747788.

Sponge

to

Langmuir

781

(48) Leocmach, M.; Tanaka, H. Roles of Icosa-

782

hedral and Crystal-Like Order in the Hard

783 784 785

Spheres Glass Transition.

2012, 3, 974.

786

Liquids and Biopolymers.

267, 19241935.

imberg,

(50) Royall, C. P.; Williams, S. R. The Role

789

of Local Structure in Dynamical Arrest.

790

Phys. Rep.

Lett.

817 818

820

Molecular

(57) Sasaki,

Phys. Rev. Lett.

95, 126104.

794

(52) Tsukruk, V. V.;

795

der, L. M.;

796

logical Properties of Composite Molecular

Everson, M. P.;

D.;

2014, S.;

Frenken, J. W. M.;

He-

of

Zandbergen, Graphite.

Kamiya, S.; Bearings.

Itamura,

H.

W.

Phys. Rev.

Sasaki, N. C60

Phys.

N.;

Rev.

Lett.

Miura,

K.

821

Atomic-Scale Ultralow Friction - Simula-

822

tion of Superlubricity of C60 Molecular

823

Bearing.

824

012001.

2005,

793

N.;

C.

Graphene

Verhoeven,

2003, 90, 055509.

(51) Braun, O. M. Simple Model of Microscopic Rolling Friction.

A

A.;

Langmuir

2004, 92, 126101.

(56) Miura, K.;

2015, 560, 1 75.

792

N.

E.

G.

A.;

816

1995,

788

J.

Superlubricity

Nat. Commun.

Science

M.;

Pradeep, N.;

(49) Angell, C. A. Formation of Glasses from

787

791

(55) Dienwiebel,

N.

Engers,

Grobert,

815

819

(47) Antelmi, D. A.; Kékiche, P.; Richetti, P.

778

780

2003, 2, 695.

S.;

van

30, 1148511492.

771

773

Perkin,

W.;

811

dor, A.;

2015, 112, 71177122.

S.

Cousens,

Surface Force Balance.

770

772

J.;

810

Klein, J. On the Question of 812 Whether Lubricants Fluidize in Stick-Slip 813

769

(54) Britton,

Lan-

Brittain, W. J. Nanotribo-


J. Phys. Conf. Ser.

2007,

89,

Page 13 of 13 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

825


Graphical TOC Entry

826


Are Buckminsterfullerenes Molecular Ball Bearings? - The Journal of

Recommend Documents