Bhushan B. Nanotribology and Nanomechanics: An Introduction

Подождите немного. Документ загружается.

1224 Bharat Bhushan

logical layers, biocompatibility, and biofouling for BioMEMS/BioNEMS are im-

portant. Most mechanical properties are known to be scale-dependent [133]. The

properties of nanoscale structures need to be measured [134]. Tribology is an

important factor aﬀecting the performance and reliability of MEMS/NEMS and

BioMEMS/BioNEMS [13,46,47,80].There is a need for the development of a fun-

damental understanding of adhesion, friction/stiction, wear, and the role of sur-

face contamination, and the environment [13]. MEMS/NEMS materials need to

exhibit good mechanical and tribological properties on the micro/nanoscale. There

is a need to develop lubricants and identify lubrication methods that are suitable

for MEMS/NEMS. Methods need to be developed to enhance adhesion between

biomolecules and the device substrate, referred to as bioadhesion. Component-level

studies are required to provide a better understanding of the tribological phenom-

ena occurringin MEMS/NEMS. The emergenceof micro/nanotribologyand AFM-

based techniques has provided researchers with a viable approach to address these

problems [46,47,135].This chapter presents an overview of micro/nanoscaletribo-

logical studies of materials and lubrication studies for MEMS/NEMS, bioadhesion,

friction and wear of BioMEMS/BioNEMS, and component-levelstudies of stiction

phenomena in MEMS/NEMS devices.

22.2 Tribological Studies of Silicon and Related Materials

The materials of most interest for planar fabrication processes using silicon as the

structural material are undoped and boron-doped (p

-type) single-crystal

silicon for bulk micromachining, and phosphorus (n

-type) doped and undoped

low-pressure chemical vapor deposition (LPCVD) polysilicon ﬁlms for surface mi-

cromachining. Silicon-based devices lack high-temperature capabilities with re-

spect to both mechanical and electrical properties. SiC has been developed as

a structural material for high-temperature microsensor and microactuator applica-

tions [138, 139]. SiC can also be desirable for high-frequency micromechanical

resonators, in the GHz range, because of its high ratio of modulus of elasticity to

density and consequently high resonance frequency. Table 22.2 compares selected

bulk properties of SiC and Si(100). Researchers have found low-cost techniques

of producing single-crystalline 3C–SiC (cubic or β-SiC) ﬁlms via epitaxial growth

on large-area silicon substrates for bulk micromachining [140] and polycrystalline

3C–SiC ﬁlms on polysilicon and silicon dioxide layers for surface micromachin-

ing of SiC [141]. Single-crystalline 3C–SiC piezoresistive pressure sensors have

been fabricated using bulk micromachining for high-temperature gas turbine appli-

cations [142]. Surface-micromachined polycrystalline SiC micromotors have been

fabricated and satisfactory operation at high temperatures has been reported [143].

As will be shown, bare silicon exhibits inadequate tribological performance and

needs to be coated with a solid and/or liquid overcoat or be surface treated (by,

e.g., oxidation and ion implantation, commonly used in semiconductor manufac-

turing), which exhibits lower friction and wear. SiC ﬁlms exhibit good tribological

22 Characterization of MEMS/NEMS and BioMEMS/BioNEMS 1225

Table 22.2. Selected bulk properties of 3C (β- or cubic) SiC and Si(100)

Sample Density

(kg/m

)

Hardness

(GPa)

Elastic modulus

(GPa)

Fracture toughness

(MPam

1/2

)

β-SiC 3210 23.5–26.5 440 4.6

Si(100) 2330 9–10 130 0.95

Sample Thermal

conductivity

(W/mK)

Coeﬀ.ofther-

mal expansion

(×10

−6

◦

Melting point

(

◦

Band gap

(eV )

β-SiC 85–260 4.5–6 2830 2.3

Si(100) 155 2–4.5 1410 1.1

Data from Bhushan and Gupta [136].

Data from Shackelford et al. [137].

performance. Both macroscale and microscale tribological properties of virgin and

treated/coated silicon, polysilicon ﬁlms and SiC are presented next.

22.2.1 Virgin and Treated/Coated Silicon Samples

Tribological Properties of Silicon and the Eﬀect of Ion Implantation

Friction and wear of single-crystallineand polycrystallinesilicon samples havebeen

studied and the eﬀect of ion implantation with variousdoses of C

and Ar

ion species at an energy of 200keV to improve their friction and wear properties

has been studied [144–146]. The coeﬃcient of macroscale friction and wear factor

of virgin single-crystal silicon and C

-implanted silicon samples as a function of

ion dose are presented in Fig. 22.14 [144]. The macroscale friction and wear tests

were conducted using a ball-on-ﬂattribometer. Each data bar represents the average

value of four to six measurements. The coeﬃcient of friction and wear factor for

bare silicon are very high and decrease drastically with ion dose. Silicon samples

bombarded above an ion dose of 10

−2

exhibit extremely low values of co-

eﬃcients of friction (typically 0.03–0.06 in air) and the wear factor (reduced by as

much as fourorders of magnitude). Gupta et al. [144] reported that a decrease in the

coeﬃcient of friction and the wear factor of silicon as a result of C

ion bombard-

ment occurred because of the formation of silicon carbiderather than amorphization

of silicon. Gupta et al. [145] also reported an improvementin friction and wear with

ion implantation.

Microscale friction measurements were performed using an atomic force/fric-

tion force microscope (AFM/FFM) [46, 47]. Table 22.3 shows values of surface

roughness and coeﬃcients of macroscale and microscale friction for virgin and

doped silicon. There is a decrease in the coeﬃcients of microscale and macroscale

friction values as a result of ion implantation. When measured for the small con-

tact areas and very low loads used in microscale studies, indentation hardness and

1226 Bharat Bhushan

Coefficient of friction

1.0

0.8

0.6

0.4

0.2

0.0

Ion dose (ions/cm

)

Wear factor (mm

/Nm)

–

Ion dose (ions/cm

)

S,P

Fig. 22.14. Inﬂuence of ion

doses on the coeﬃcient of

friction and wear factor on

-ion-bombarded single-

crystal and polycrystalline

silicon slid against alumina

ball. V corresponds to virgin

single-crystal silicon, while

S and P denote tests that

correspond to doped single-

crystal and polycrystalline

silicon, respectively [144]

Table 22.3. Surface roughness and micro- and macroscale coeﬃcients of friction of selected

samples

Material RMS roughness

(nm)

Coeﬃcient

of microscale

friction

Coeﬃcient

of macroscale

friction

Si(111) 0.11 0.03 0.33

-implanted

Si(111)

0.33 0.02 0.18

Versus S i

tip; tip radius of 50 nm in the load range 10–150 nN (2.5–6.1GPa)

at a scanning speed of 5 µm/s over a scan area of 1 µm×1 µminanAFM

Versus S i

ball, ball radius of 3 mm at a normal load of 0.1N(0.3GPa)atan

average sliding speed of 0.8mm/s using a tribometer

22 Characterization of MEMS/NEMS and BioMEMS/BioNEMS 1227

elastic modulus are higher than those at the macroscale. This, added to the eﬀect

of the small apparent area of contact reducing the number of trapped particles on

the interface, results in less plowing contribution and lower friction in the case of

microscale friction measurements. Results of microscale wear resistance studies of

ion-implanted silicon samples studied using a diamond tip in an AFM [147] are

shown in Figs. 22.15a and 22.15b. For tests conducted at various loads on Si(111)

Wear depth (nm)

400

300

200

100

Normal load (

μN)

0 40 80 120 160

Wear depth (nm)

400

300

200

100

Number of cycles

020406080

100

Nanohardness (GPa)

Normal load (

μN)

Depth (nm)

200

150

100

10 20 30

1 cycle

Si(111)

-implanted Si(111)

Si(111)

-implanted Si(111)

40 μN normal load

Si(111)

-implanted Si(111)

Fig. 22.15. Wear depth as

a function of (a)load(after

one cycle), and (b)cycles

(normal load = 40 mN) for

Si(111) and C

-implanted

Si(111). (c) Nanohardness

and normal load as func-

tion of indentation depth

for virgin and C

-implanted

Si(111) [147]

1228 Bharat Bhushan

and C

-implanted Si(111), it is noted that wear resistance of implanted sample is

slightly poorer than that of virgin silicon up to about 80µN. Above 80 µN, the wear

resistance of implanted Si improves. As one continues to run tests at 40 µNfor

a larger number of cycles, the implanted sample, which forms hard and tough sili-

con carbide, exhibits higher wear resistance than the unimplanted sample. Damage

from the implantationin the top layer results in poorer wear resistance, however,the

implanted zone at the subsurface is more wear-resistant than the virgin silicon.

Hardness values of virgin and C

-implanted Si(111) at various indentation

depths (normal loads) are presented in Fig. 22.15c [147]. The hardness at a small

indentationdepthof 2.5nmis16.6GPa and it dropsto a valueof 11.7GPa at a depth

of 7nm and a normal load of 100µN. Higher hardness values obtained in low-load

indentationmay arise fromthe observedpressure-inducedphase transformationdur-

ing thenanoindentation[148,149].Additionalincrease inthe hardnessat an theeven

lower indentation depth of 2.5nm reported here may arise from the contribution by

complex chemical ﬁlms (not from native oxide ﬁlms) present on the silicon sur-

face. At small volumes there is a lower probability of encountering material defects

(dislocations, etc.). Furthermore, according to the strain-gradient plasticity theory

advanced by Fleck et al. [150], large strain gradients inherent in small indentations

lead to accumulation of geometrically necessary dislocations that cause enhanced

hardening. These are some of the plausible explanations for an increase in hardness

at smaller volumes. If the silicon material were to be used at very light loads such as

in microsystems, the high hardness of surface ﬁlms would protect the surface until

it is worn.

From Fig. 22.15c, hardness values of C

-implanted Si(111) at a normal load of

50µNis20.0GPa with an indentation depth of about 2nm, which is comparable

to the hardness value of 19.5GPaat70µN, whereas measured hardness value for

virgin silicon at an indentation depth of about 7 nm (normal load of 100µN) is only

about 11.7GPa. Thus, ion implantation with C

results in an increase in hardness in

silicon. Note that the surface layer of the implanted zone is much harder compared

with the subsurface, and may be brittle leading to higher wear on the surface. The

subsurface of the implanted zone (SiC) is harder than the virgin silicon, resulting in

higher wear resistance, which is also observed in the results of the macroscale tests

conducted at high loads.

The Eﬀect of Oxide Films on Tribological Properties of Silicon

Macroscale friction and wear experiments have been performed using a magnetic

disk drive with bare, oxidized, and implanted pins sliding against amorphous-

carbon-coated magnetic disks lubricated with a thin layer of perﬂuoropolyether lu-

bricant [151–154]. Representative proﬁles of the variation of the coeﬃcient of fric-

tion with number of sliding cycles for Al

–TiC slider and bare and dry-oxidized

silicon pins are shown in Fig. 22.16. For bare Si(111), after initial increase in the

coeﬃcient of friction, it drops to a steady-state value of 0.1 following the increase,

as seen in Fig. 22.16. The rise in the coeﬃcient of friction for the Si(111) pin is as-

sociated with the transfer of amorphous carbon from the disk to the pin, oxidation-

22 Characterization of MEMS/NEMS and BioMEMS/BioNEMS 1229

1.00

0.80

0.60

0.40

0.20

0.00

Coefficient of friction

0 1200 2400 3600 4800 6000

Revolutions

1.00

0.80

0.60

0.40

0.20

0.00

Coefficient of friction

0 10000

Revolutions (×10

)

20000 30 000 40000 50000

Ambient air

Si(111) pin

Dry nitrogen

Ambient air

Dry-oxidized silicon pin

Fig. 22.16. Coeﬃcient of fric-

tion as a function of number

of sliding revolutions in am-

bient air for a Si(111) pin in

ambient air and dry nitrogen

and a dry-oxidized silicon pin

in ambient air [151]

enhanced fracture of pin material followed by tribochemical oxidation of the trans-

fer ﬁlm, while the drop is associated with the formation of a transfer coating on

the pin. Dry-oxidized Si(111) exhibits excellent characteristics and no signiﬁcant

increase was observed over 50,000 cycles (Fig. 22.16). This behavior has been at-

tributed to the chemical passivity of the oxide and lack of transfer of diamond-like

carbon (DLC) from the disk to the pin. The behavior of PECVD oxide (data not

presented here) was comparable to that of dry oxide but for the wet oxide there was

some variation in the coeﬃcient of friction (0.26–0.4). The diﬀerence between dry

and wet oxide was attributed to increased porosity of the wet oxide [151]. Since

tribochemical oxidation was determined to be a signiﬁcant factor, experimentswere

conducted in dry nitrogen [152,153]. The variation of the coeﬃcient of friction for

a silicon pin sliding against a thin-ﬁlm disk in dry nitrogen is shown in Fig. 22.16. It

is seen that, in a dry nitrogen environment, the coeﬃcient of friction of Si(111)slid-

ing against a disk decreased from an initial value of about 0.05–0.2 with continued

sliding. Based on SEM and chemical analysis, this behavior has been attributed to

the formation of a smooth amorphous-carbon/lubricanttransfer patch and suppres-

1230 Bharat Bhushan

sion of oxidation in a dry nitrogen environment. Based on macroscale tests using

disk drives, it is found that the friction and wear performance of bare silicon is not

adequate. With dry-oxidized or PECVD SiO

-coated silicon, no signiﬁcant friction

increase or interfacial degradation was observed in ambient air.

Table 22.4 and Fig. 22.17 show surface roughness, microscale friction and

scratch data and nanoindentation hardness for the various silicon samples [147].

Scratch experiments were performed using a diamond tip in an AFM. Results on

polysilicon samples are also shown for comparison. Coeﬃcients of microscale fric-

tion values for all the samples are about the same. These samples could be scratched

at 10 µN load. Scratch depth increased with normal load. Crystalline orientation of

silicon has little inﬂuence on scratch resistance because natural oxidation of silicon

in ambient masks the expected eﬀect of crystallographic orientation. PECVD-oxide

samplesshowed thebest scratchresistance, followedby dry-oxidized,wet-oxidized,

and ion-implanted samples. Ion implantation with C

does not appear to improve

scratch resistance.

Wear data onthe silicon samplesare alsopresented in Table22.4 [147].PECVD-

oxide samples showed superior wear resistance followed by the dry-oxidized, wet-

oxidized, and ion-implanted samples. This agrees with the trends seen in scratch

resistance. In PECVD, ion bombardment during the deposition improves the coat-

ing properties such as suppression of columnar growth, freedom from pinhole, de-

crease in crystalline size, and increase in density, hardness and substrate-coating

adhesion. These eﬀects may help in improving mechanical integrity of the sample

surface. Coatings and treatments improved nanohardness of silicon. Note that dry-

oxidized and PECVD ﬁlms are harder than wet-oxidized ﬁlms as these ﬁlms may

be porous. The high hardness of oxidized ﬁlms may be responsible for the high

measured scratch/wear resistance.

Undoped Si(111)

PECVD-oxide Si(111)

Dry-oxidized Si(111)

Wet-oxidized Si(111)

-implanted Si(111)

Scratch depth (nm)

Normal load (

μN)

250

200

150

100

0 20 40 60 80 100

Fig. 22.17. Scratch depth as

a function of normal load af-

ter 10 cycles for various sili-

con samples: virgin, treated,

and coated [147]

22 Characterization of MEMS/NEMS and BioMEMS/BioNEMS 1231

Table 22.4. RMS, microfriction, microscratching/microwear and nanoindentation hardness

data for various virgin, coated and treated silicon samples

Material Rms

roughness

(nm)

Coeﬃcient

of microscale

friction

Scratch

depth

at 40µN

(nm)

Wear depth

at 40µN

(nm)

Nanohard-

ness

at 100µN

(GPa)

Si(111) 0.11 0.03 20 27 11.7

Si(110) 0.09 0.04 20

Si(100) 0.12 0.03 25

Polysilicon 1.07 0.04 18

Polysilicon

(lapped)

0.16 0.05 18 25 12.5

PECVD-oxide

coated Si(111)

1.50 0.01 8 5 18.0

Dry-oxidized

Si(111)

0.11 0.04 16 14 17.0

Wet-oxidized

Si(111)

0.25 0.04 17 18 14.4

-implanted

Si(111)

0.33 0.02 20 23 18.6

Scan size of 500 nm×500 nm using AFM

Versus S i

tip in AFM/FFM, radius 50 nm; at 1 µm×1 µm scan size

Measured using an AFM with a diamond tip of radius 100 nm

22.2.2 Tribological Properties of Polysilicon Films and SiC Film

Studies have also been conducted on undoped polysilicon ﬁlm, heavily doped (n

type) polysilicon ﬁlm, heavily doped (p

-type) single-crystal Si(100) and 3C–SiC

(cubic or β-SiC) ﬁlm [155–157]. The polysilicon ﬁlms studied here are diﬀerent

from those discussed previously.

Table 22.5 presents a summary of the tribologicalstudies conducted on polysili-

con and SiC ﬁlms. Values for single-crystal silicon are also shown for comparison.

Polishing of the as-deposited polysilicon and SiC ﬁlms drastically aﬀect the rough-

ness as the values reduce by two orders of magnitude. Si(100) appears to be the

smoothest, followed by polished undoped polysilicon and SiC ﬁlms, which have

comparable roughness. The doped polysilicon ﬁlm shows higher roughness than

the undoped sample, which is attributed to the doping process. Polished SiC ﬁlm

shows the lowest friction followed by polished and undoped polysilicon ﬁlm, which

strongly supports the candidacy of SiC ﬁlms for use in MEMS/NEMS devices.

Macroscale friction measurements indicate that SiC ﬁlm exhibits one of the lowest

friction values as compared to the other samples. Doped polysilicon sample shows

1232 Bharat Bhushan

Table 22.5. Summary of micro/nanotribological properties of the sample materials

Sample Rms

roughness

P-V

distance

Coeﬃcient

of friction

Scratch

depth

(nm)

(nm) (nm) Micro

Macro

Undoped Si(100) 0.09 0.90.06 0.33 89

Undoped polysilicon

ﬁlm (as deposited)

46 340 0.05

Undoped polysilicon

ﬁlm (polished)

0.86 6 0.04 0.46 99

-type polysilicon

ﬁlm (as deposited)

12 91 0.07

-type polysilicon

ﬁlm (polished)

1.07 0.02 0.23 61

SiC ﬁlm

(as deposited)

25 150 0.03

SiC ﬁlm (polished) 0.89 6 0.02 0.20 6

Sample Wear depth

(nm)

Nano-

hardness

(GPa)

Young’s

modulus

(GPa)

Fracture tough-

ness

(MPa m

1/2

)

Undoped Si(100) 84 12 168 0.75

Undoped polysilicon

ﬁlm (as deposited)

Undoped polysilicon

ﬁlm (polished)

140 12 175 1.11

-type polysilicon

ﬁlm (as deposited)

-type polysilicon

ﬁlm (polished)

51 9 95 0.89

SiC ﬁlm

(as deposited)

SiC ﬁlm (polished) 16 25 395 0.78

Measured using AFM over a scan size of 10 µm×10 µm

Measured using AFM/FFM over a scan size of 10 µm×10 µm

Obtained using a 3-mm-diameter sapphire ball in a reciprocating mode at a normal load

of 10 mN and average sliding speed of 1 mm/s after 4 m sliding distance

Measured using AFM at a normal load of 40 µN for 10 cycles; scan length of 5 µm

Measured using AFM at normal load of 40 µN for 1 cycle, wear area of 2 µm×2 µm

Measured using nanoindenter at a peak indentation depth of 20 nm

Measured using microindenter with Vickers indenter at a normal load of 0.5N

22 Characterization of MEMS/NEMS and BioMEMS/BioNEMS 1233

low friction on the macroscale as compared to the undoped polysilicon sample, pos-

sibly due to the doping eﬀect.

Figure 22.18a shows a plot of scratch depth versus normal load for various sam-

ples [155, 156]. Scratch depth increases with increasing normal load. Fig. 22.19

shows AFM three-dimensional (3D) maps and averaged two-dimensional (2D) pro-

ﬁles of the scratch marks on the various samples. It is observed that scratch depth

increases almost linearly with the normal load. Si(100) and the doped and undoped

polysilicon ﬁlm show similar scratch resistance. From the data, it is clear that the

SiC ﬁlmis much more scratch-resistantthan the other samples. Figure22.18bshows

Search depth (nm)

300

200

100

Normal load (

μN)

0 20 40 60 80 100

Wear depth (nm)

300

200

100

0 20 40 60 80 100

Normal load (

μN)

600

450

300

150

0 5 10 15 20 25

Number of cycles

Wear depth (nm)

1 cycle

20 μN normal load

Undoped Si(100)

Undoped

polysilicon film

-polysilicon film

SiC film

Fig. 22.18. (a)Scratch depths

for 10 cycles as a function

of normal load and (b) wear

depths asa function of normal

load and of number of cycles

for various samples [155]