Bhushan B. Nanotribology and Nanomechanics: An Introduction

Подождите немного. Документ загружается.

920 Bharat Bhushan

Xiao et al. [71] and Lio et al. [72] also studied the eﬀect of the length of the alkyl

chain on the frictional properties of n-alkanethiolate ﬁlms on gold and n-alkylsilane

ﬁlms on mica. Friction was foundto be particularlyhigh for short chainsof less than

eight carbon atoms. Thiols and silanes exhibit similar friction force for the same n

when n> 11; for n < 11, silanes exhibit higher friction, about three times larger than

that for thiols for n = 6. The increase in friction was attributed to the large number

of dissipative modes in the less ordered chains that occur when going from a thiol

to a silane anchor or when decreasing n. Longer chains (n > 11), stabilized by van

der Waals attractions, form more compact and rigid layers and are better lubricants.

Schonherr and Vancso [73] also correlated the magnitude of the friction with the

order among the alkane chains. The disorder in short-chain hydrocarbon disulﬁde

SAMs was found to result in a signiﬁcant increase in the magnitude of the friction.

Tsukruk and Bliznyuk [74] studied the adhesion and friction between a Si sam-

ple and a Si

tip, in which both surfaces were modiﬁed by

−

-and

−

H-terminated silane-based SAMs. Various polymer molecules were used for

the backbone. They reported a very broad maximum adhesive force at pH 4–8, with

minimumadhesion at pH> 9 and pH< 3, for all ofthe studiedmating surfaces. This

observation can be understood by considering a balance of electrostatic and van der

Waals interactions between composite surfaces with multiple isoelectric points. The

friction coeﬃcients of NH

/NH

-andSO

H/SO

H-mating SAMs are very high in

aqueous solution. Capping NH

-modiﬁed surfaces (3-aminopropyltriethoxysilane)

with rigid and soft polymer layers resulted in a signiﬁcant reduction in adhesion,

to a level lower than that of the untreated surface [75]. Fujihira et al. [76] studied

the inﬂuence of surface terminal groups of SAMs and functional tips on adhesive

force. It was found that the adhesive force measured in air increases in the order

/CH

,CH

/COOH, COOH/COOH.

Bhushan and Liu [39], Liu et al. [40], and Liu and Bhushan [41, 42] studied

adhesion, friction and wear properties of alkylthiol and biphenylthiol SAMs. They

explained the friction mechanisms using a molecular spring model in which local

stiﬀness and intermolecularforces governfriction properties.They studiedthe inﬂu-

ence of relative humidity, temperature and velocity on adhesion and friction. They

also investigated the wear mechanisms of SAMs using a continuous microscratch

AFM technique.

Fluorinated carbon (ﬂuorocarbon) molecules are known to have low surface en-

ergy and are commonly used for lubrication [2,9,10]. Bhushan et al. [43,46], Kasai

et al. [44] and Lee et al. [45] studied the friction and wear properties of methyl- and

perﬂuoro-terminated alkylsilanes on silicon. Bhushan et al. [5] and Kasai et al. [44]

reported that perﬂuoroalkylsilaneSAMs exhibit lower surface energies, higher con-

tact angles, lower adhesive forces, and lower wear than those of alkylsilanes. Kasai

et al. [44] also reported the inﬂuence of relative humidity, temperature and veloc-

ity on adhesion and friction. Tambe and Bhushan [47] studied the nanotribological

properties of methyl-terminated alkylphosphonate on aluminium, which is of in-

dustrial interest. They found that these SAMs perform as well on aluminium as on

17 Self-Assembled Monolayers (SAMs) for Controlling Adhesion 921

silicon. Tao and Bhushan [48] studied degradation mechanisms of SAMs. They re-

ported that oxygen from the air causes thermal oxidation of SAMs.

To date, the nanotribological properties of alkanethiol, biphenylthiol, alkylsi-

lane and perﬂuoroalkylsilane SAMs have been widely studied. In this chapter, we

review, in some detail, the nanotribological properties of various SAMs that have

alkyl and biphenyl spacer chains with diﬀerent surface terminal groups (

−

) and head groups (

−

OH, and P

−

), which were inves-

tigated by AFM at various operating conditions, Fig. 17.8a,b [5,39–42,44,47,48].

Hexadecanethiol (HDT), 1,1



-biphenyl-4-thiol(BPT) andcrosslinked BPT (BPTC)

were deposited on Au(111) substrates by immersing the substrate in a solution

containing a precursor (ligand) that reacts with the substrate surface. Crosslinked

BPTC wasproduced by irradiatingBPT monolayerswith low-energyelectrons.Per-

ﬂuorodecyltricholorosilane (PFTS), CF

−

(CF

)

−

(CH

)

−

SiCl

, n-octyldimethyl

(dimethylamino) silane (ODMS), CH

−

(CH

)

−

Si(CH

)

−

N(CH

)

(n = 7), and

n-octadecylmethyl(dimethylamino)silane (n = 17) (ODDMS) were deposited on Si

by exposing thesubstrate to the vaporof the reactive chemicalprecursor. Octylphos-

phonate (OP)

−(CH

)

−

−OH(n = 7)

Hexadecane thiol

(HDT)

1,1'–biphenyl–4–thiol

(BPT)

Cross-linked 1,1'–biphenyl–4–thiol

(BPTC)

SiCH

(CH

)

(CF

)

(CH

)

(CH

)

(CH

)

SiCH

(CH

)

n-Dimethyl(dimethylamino)silane Perfluorodecyltrichlorosilane n-Phosphonate

Octodecyl (ODDMS) Deca (PFTS) Octadecyl (ODP)Octa (ODMS) Octa (OP)

Alkyl

–(CH

)

–

Au(111)

Biphenyl

–(C

)

–

Au(111)

SSSS

Fig. 17.8. Schematics of the structures of (a) hexadecane and biphenyl thiol SAMs on

Au(111) substrates, and (b) perﬂuoroalkylsilane and alkylsilane SAMs on Si with native ox-

ide substrates, as well as alkylphosphonate SAMs on Al with native oxide

922 Bharat Bhushan

and octadecylphosponate(n = 17)(ODP) were depositedon Al by liquiddeposition.

Thermally evaporated Au(111)ﬁlms on Si(111) substrate were selected because the

epitaxial ﬁlm is smooth and defect-free, which is desirable for SAM applications.

Bulk Si(100) and Al with natural oxide layers were selected because they are used

in the construction of MEMS/NEMS.

17.4.1 Measurement Techniques

Static Contact Angle Measurement Using DI Water

The static contact angle, a measure of how water repellent a material is, was meas-

ured using a Rame–Hart model 100 contact angle goniometer (Mountain Lakes,

NJ, USA) [77,78]. Ten microliter droplets of DI water were typcially used for the

contact angle measurements. At least two measurements of the contact angle were

taken. The contact angles were reproducible within ±2

◦

AFM Adhesion and Friction Measurements

Adhesion and friction tests were conducted using a commercial AFM system

(Dimension 3000, Nanoscope IIIa controller, DI, Santa Barbara, USA). Square-

pyramidal Si

tips with a 30–50nm tip radius were used on a gold back-coated

triangularSi

cantileverwith a typicalspring constant of 0.58 N/m. The adhesion

can be calculated using either force calibration plots or from the negative intercepts

on friction force versus normal load plots. Both methods generally yield similar re-

sults. The force calibration plot technique was used in this study. The coeﬃcient of

friction was obtained from the slope of a plot of the friction force versus the normal

load. Normal loads typically ranged from 5 to 100nN. Friction force measurements

were performed at a scan rate of 1 Hz along the fast scan axis and over a scan size of

2 µm. The fast scan axis was perpendicular to the longitudinal direction of the can-

tilever. The friction force was calibrated by the method described in Bhushan [4].

Eﬀects of Relative Humidity, Temperature and Sliding Velocity

The inﬂuence of the relative humidity on the adhesive force, the friction force and

the wear was studied in an environmentally controlled chamber. Relative humidity

was controlled by introducing a mixture of dry and moist air into the chamber. The

temperature was maintained at 22±1

◦

C. The sample was kept in the environmental

chamber at the desired humidity for at least 2 h prior to the tests so that the system

could reach equilibrium.

In order to study the eﬀect of temperature on adhesion and friction force, the

samples were placed on a thermal stage during the measurements. A glass plate

was placed under the thermal stage to preventthe heat from being transported away.

The temperature range studied was from 20 to 110

◦

C. The relative humidity was

maintained at 50±5% during the measurements.

The eﬀect of sliding velocity on friction force was monitored in ambient con-

ditions using a high-velocity piezo stage designed to achieve high relative sliding

17 Self-Assembled Monolayers (SAMs) for Controlling Adhesion 923

velocities on a commercial AFM set-up [79]. The traveling distance of the sample

(the scan size)was set at 25µm, while the scan frequencywas varied between0.1Hz

(5µm/s) and 100Hz (5000µm/s).

AFM Wear Measurements

Wear tests were conducted using a diamond tip with a nominal radius of 50nm

and a cantilever with a nominal stiﬀness of 10N/m. Wear tests were performed on

a scan area of 1 µm×1 µm at the desired normal load and at a scan rate of 1 Hz.

After each wear test, an area of 3 µm×3µm was imaged and the average wear depth

was calculated.

Degradation and Environmental Studies

The lubricant degradation experiments were carried out in a high-vacuum tribotest

apparatus [82, 83]. The system was equipped with a mass spectrometer so that

gaseous emissions from the interface could be monitored in situ during the slid-

ing in high vacuum. The normal loads and friction forces at the contacting interface

were measured using resistive-type strain-gauge transducers. Sliding tests were con-

ducted by rubbing the sample against a ﬂat sample of Si(100) at a vacuum pressure

of 2×10

−7

Torr at a sliding speed of 0.3m/s. The environmental eﬀects were inves-

tigated in high vacuum (2×10

−7

Torr), argon, dry air (less than 2% RH), ambient

air (30% RH) and in high humidity air (70% RH).

17.4.2 Hexadecane Thiol and Biphenyl Thiol SAMs on Au(111)

Hexadecane thiol on Au(111) was selected as it is a widely studied ﬁlm. Biphenyl

thiol was selected to study the eﬀect of rigidity on nanotribological performance.

The biphenyl thiol ﬁlm was crosslinked to further increase its stiﬀness.

Surface Roughness, Adhesion and Friction

Surfaceheight and friction force images of SAMs were recorded simultaneously on

an area of 1 µm×1 µm by an AFM, and adhesive forces were measured using an

AFM in force calibration mode [39].

Detailed analysis is presented later in this chapter, but the measured rough-

nesses, thickness, tilt angles and spacer chain lengths of Si(111), Au(111) and var-

ious SAMs are listed in Table 17.10 [39]. The roughness of BPT is very close to

that of Au(111), but the roughness of BPTC is lower than that of Au(111) and BPT;

this is caused by electron irradiation. Table 17.10 indicates that the roughness of

HDT is much higher than the substrate roughness of Au(111). This is caused by

local aggregation of organic compounds on the substrates during SAM deposition.

Table 17.5 also indicates that the thicknesses of biphenyl thiol SAMs are generally

smaller than those of the alkylthiol, because of the shorter spacer chain in biphenyl

thiol.

924 Bharat Bhushan

Table 17.10. R

roughnesses, thicknesses, tilt angles and spacer chain lengths of SAMs

Samples R

roughness

Thickness

Tilt angle

Spacer length

(nm) (nm) (degrees) (nm)

Si(111) 0.07

Au(111) 0.37

HDT 0.92 1.89 30 1.91

BPT 0.36 1.25 15 0.89

BPTC 0.14 1.14 25 0.89

Measured by an AFM with a 1 µm×1 µm scan size, using a Si

tip under 3.3 nN normal

load

The thickness and tilt angles of BPT and BPTC are reported by Geyer et al. [67]. The thick-

ness and tilt angles of HDT are reported by Ulman [25]

The spacer chain lengths of alkylthiols were calculated by the method reported by Miura

et al. [80]. The spacer chain lengths of biphenyl thiols were calculated from the data reported

by Ratajczak-Sitarz et al. [81]

Average values and standard deviations for the adhesive force and coeﬃcient of

friction are presented in Fig. 17.9 [39]. Based on the data, the adhesive forces and

friction coeﬃcients of SAMs are less than those of their corresponding substrates.

Among the various ﬁlms, HDT exhibits the lowest values. The adhesive force F

is ranked as follows: F

a–Au

> F

a–BPT

> F

a–BPTC

> F

a–HDT

. The rankings for the

friction coeﬃcients μ are: μ

>μ

BPTC

>μ

BPT

>μ

HDT

. Note that many SAMs have

similar rankings for both adhesive force and coeﬃcient of friction. This suggests

that alkylthiol and biphenyl SAMs would both make eﬀective molecular lubricants

for micro/nanodevices.

Liquid capillary condensation is one source of adhesion and friction in micro/

nanoscale contact. For a sphere in contact with a ﬂat surface, the attractive Laplace

force caused by a water capillary is

= 2πRγ

(cosθ

+ cosθ

) , (17.1)

where R is the radius of the sphere, γ

is the surface tension of the liquid against

air, and θ

and θ

are the contact angles between the liquid and ﬂat and spherical

surfaces, respectively [9,10]. In an AFM adhesive study, the tip–ﬂat sample contact

is just like a sphere in contact with a ﬂat surface, and the liquid is water. Since

a single tip is used in the adhesionmeasurements,cosθ

can be treated as a constant.

Therefore,

= 2πRγ

(1+ cosθ

)−2πRγ

(1−cosθ

)

= 2πRγ

(1+ cosθ

)−C , (17.2)

where C is a constant.

Based on the following Young–Dupre equation, the work of adhesion W

(the

work required to pull apart a unit area of the solid–liquid interface) can be written

17 Self-Assembled Monolayers (SAMs) for Controlling Adhesion 925

Adhesive force (nN)

HDT BPT BPTC

Coefficient of friction

0.08

0.06

0.04

0.02

Materials

HDT BPT BPTC

Fig. 17.9. Adhesive forces

and coeﬃcients of friction for

Au(111) and various SAMs

as [84]

= γ

(1+ cosθ

) . (17.3)

This indicates that W

is determined by the SAM contact angle; in other words it is

inﬂuenced by the surface chemistry properties (polarization and hydrophobicity)of

the SAM. By substituting (17.3) into (17.2), F

can be expressed as

= 2πRW

−C (17.4)

When the inﬂuence of other factors, such as van der Waals force, on the adhesive

force is very small, then the adhesive force F

≈ F

. Thus the adhesive force F

should be proportional to the work of adhesion W

The contact angle is a measure of the wettability of a solid by a liquid, and it

determines the W

value [77,78]. The contact angles for distilled water on Au(111)

and SAMs have been measured, and are summarized in Fig. 17.10a [39]. For water,

= 72.6mJ/m

at 22

◦

C.Therefore, usingthisvalue and(17.3),it is possibleto ob-

tain W

data, and these are presented in Fig. 17.10b. The W

values can be ranked in

the following order: W

a–Au

(97.1) > W

a–BPT

(86.8) > W

a–BPTC

(82.1) > W

a–HDT

(61.4).

Except for W

a–Au

, this order mimics the order of adhesion force in Fig. 17.9. The re-

lationship between F

and W

is summarized in Fig. 17.11 [39]. It indicates that the

adhesive force F

(nN) increases with the work of adhesion W

(mJ/m

) as follows:

= 0.57W

−22 . (17.5)

926 Bharat Bhushan

100

120

100

BPTCBPTHDTAu

Materials

Contact angle (deg)

(mJ/m

)

Fig. 17.10. (a) The static

advancing contact angle, and

(b) the work of adhesion for

Au(111) and various SAMs.

Each point in this ﬁgure

represents the mean value

of six measurements. The

uncertainty associated with

the average contact angle is

±2

◦

(mJ/m

)

Adhesive force (nN)

70 80 90 100

Au(111)

HDT

BPT

BPTC

Fig. 17.11. Relationship

between the adhesive force

and the work of adhesion for

diﬀerent specimens

These experimental results agree well with the modeling prediction presented

earlier in (17.4). It proves that, on the nanoscale, and under ambient conditions, the

adhesiveforces of SAMs are mainly inﬂuencedby the water capillary force. Though

neither HDT nor BPT has polar surface groups, the surface terminal group of HDT

has a symmetrical structure, which causes a smaller electrostatic attractive force and

17 Self-Assembled Monolayers (SAMs) for Controlling Adhesion 927

yields a smaller adhesive force than BPT. It is believed that the easy attachment of

Au to the tip is one of the reasons for the unexpectedly large adhesive force, which

means that it does not ﬁt the linear relationship described by (17.5).

Stiﬀness, Molecular Spring Model and Micropatterned SAMs

The friction mechanisms of SAMs were also examined. Monte Carlo simulation

of the mechanical relaxation of a CH

(CH

)

SH self-assembled monolayer,as per-

formedby Siepmanand McDonald [85],indicatedthat SAMs compressand respond

nearly elastically to microindentation by an AFM tip when the load is below a criti-

cal normal load. Compression can lead to major changes in the mean molecular

tilt (the orientation), but the original structure is recovered as the normal load is

removed.

Stiﬀness properties were measured by an AFM in force modulation mode [4,

5,41]. They reported that BPT was stiﬀer than HDT. Since BPT has rigid benzene

structure,it is more diﬃcultto compressthanHDT. Figure17.12 showsthe variation

in the displacement with normal load during indentation mode. It clearly indicates

that SAMs can be compressed. At a given normal load, SAMs with long carbon

chain structures such as HDT are easy to compress compared to SAMs with rigid

benzene ring structures, such as BPT. Garcia-Parajo et al. [86] have also reported

on the compression and relaxation of octadecyltrichlorosilane (OTS) ﬁlm obtained

from loading and unloading tests.

A molecular spring model is presented in Fig. 17.13 to explain the diﬀerence

in friction between the SAMs observed in the friction and stiﬀness measurements

by AFM and the Monte Carlo simulation. It is believed that the self-assembled

molecules on a substrate act just like assembled molecular springs anchored to the

substrate [39]. A Si

tip sliding on the surface of a SAM is like a tip sliding on

the top of molecular springs or a brush. The molecular spring assembly has com-

pliant features and can experience compression and orientation under normal load.

Normal load (nN)

Displacement (nm)

0.0

2.0

30.0

20.0

10.0

0.0

0.5 1.0 1.5

HDT

BPT

BPTC

Fig. 17.12. Normal load

versus displacement curves

for Au(111) and various

SAMs

928 Bharat Bhushan

Substrate

Fig. 17.13. Molecular spring model for SAMs. In this ﬁgure, α

<α

, which is caused by re-

orientation under the normal load applied by the AFM tip. The reorientation of the molecular

springs reduces the shearing force at the interface, which in turn reduces the friction force.

The molecular spring constant and the intermolecular forces determine the magnitude of the

coeﬃcient of friction. In this ﬁgure, the size of the tip and the molecular springs are not drawn

tothesamescale[39]

The orientation of the molecular springs or brush reduces the shearing force at the

interface, which in turn reduces the friction force. The possibility of orientation is

determined by the spring constant of a single molecule (local stiﬀness), as well as

the interactions between neighboring molecules, which is reﬂected in the packing

density or packing energy. It should be noted that the orientation can lead to con-

formational defects along the molecular chains, which lead to energy dissipation. In

the study of BPT by AFM, it was found that the friction force is signiﬁcantly re-

duced after the ﬁrst several scans, but the surface height does not show any apparent

change. This suggests that molecular orientation can be facilitated by initial sliding

and is reversible [42].

Basedonthestiﬀness measurement results presented in Fig. 17.12 and the view

of molecular structures given in Fig. 17.13, biphenyl is a more rigid structure due

to the contribution of the two rigid benzene rings. Therefore, the spring constant

of BPT is larger than that of HDT. The hydrogen (H

) in the biphenyl chain has

an electrostatic attraction to the π electrons in the neighboring benzene ring. Thus

the intermolecular forces between biphenyl chains are stronger than those between

alkyl chains. The larger spring constant of BPT and stronger intermolecular forces

mean that it requiresa larger external force to allow it to orient, thus causing a higher

coeﬃcient of friction. The crosslinking of BPT leads to a larger packing energy for

BPTC. Therefore BPTC orientation requires a larger external force: the coeﬃcient

of BPTC is higher than BPT.

An elegant way to demonstrate the inﬂuence of molecular stiﬀness on friction is

to investigate SAMs with diﬀerent structures on the same wafer. A micropatterned

SAM was prepared for this purpose. First biphenyldimethylchlorosilane (BDCS)

was deposited on the silicon using a typical self-assembly method [41]. Then the

ﬁlm was partially crosslinked using a mask technique by low-energy electron irra-

diation. Finally, micropatterned BDCS ﬁlms that had the diﬀerent coating regions

on the same wafer were realized. The local stiﬀness properties of these micropat-

terned samples were investigated by a force modulation AFM technique [87]. The

17 Self-Assembled Monolayers (SAMs) for Controlling Adhesion 929

Surface height Stiffness

6μm

Surface height Friction force

0 10 nm

010010μm

stiff soft

0 10 nm

4.5 nN

Fig. 17.14. (a)AFM

grayscale surface height

and stiﬀness images, and

(b) AFM grayscale surface

height and friction force

images of micropatterned

BDCS [41]

variationin the deﬂection amplitude provides a measure of the relativelocal stiﬀness

of the surface. Surface height, stiﬀness and friction images of the micropatterened

biphenyldimethylchlorosilane (BDCS) specimen were obtained and are presented

in Fig. 17.14 [41]. The circular areas correspond to the as-deposited ﬁlm, and the

remaining area to the crosslinked ﬁlm. Figure 17.14a indicates that crosslinking

caused by the low-energy electron irradiation leads to a decrease of about 0.5nm

in the surface height of the BDCS ﬁlm. The corresponding stiﬀness images indi-

cate that the crosslinked area has a higher stiﬀness than the as-deposited area. Fig-

ure 17.14b indicates that the as-deposited area has lower friction force. Obviously,

these data for the micropatterned sample prove that the local stiﬀness of the SAM

inﬂuences its friction performance. Higher stiﬀness leads to larger friction force.

These results correlate well with the suggested molecular spring model.

In summary, it was found that SAMs exhibit compliance and can experience

compression and orientation under normal load. SAM orientation reduces the shear

stress at the interface, so SAMs provide good lubricants. The molecular spring con-

stant (local stiﬀness) and intermolecular forces can both inﬂuence the friction coef-

ﬁcient of a SAM.