Tribological properties of polished diamond filmsBharat Bhushan, Vish V. Subramaniam, Ajay Malshe, B. K. Gupta, and Juai Ruan Citation: Journal of Applied Physics 74, 4174 (1993); doi: 10.1063/1.354421 View online: http://dx.doi.org/10.1063/1.354421 View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/74/6?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Tribological properties of nanocrystalline diamond films deposited by hot filament chemical vapordeposition AIP Advances 2, 032164 (2012); 10.1063/1.4751272 Diamondlike carbon film synthesized by ion beam assisted deposition and its tribological properties J. Vac. Sci. Technol. A 14, 2039 (1996); 10.1116/1.580079 Ultrasound effects on the tribological properties of synthesized diamond films J. Vac. Sci. Technol. B 13, 2124 (1995); 10.1116/1.588087 Tribological improvements of polished chemically vapor deposited diamond films by fluorination Appl. Phys. Lett. 65, 1109 (1994); 10.1063/1.112113 Tribological properties of diamond films grown by plasmaenhanced chemical vapor deposition Appl. Phys. Lett. 54, 2006 (1989); 10.1063/1.101197

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

130.113.86.233 On: Fri, 19 Dec 2014 13:54:10

Tribological properties of polished diamond films Bharat Bhushan, Vish V. Subramaniam, Ajay Malshe, B. K. Gupta, and Juai Ruan Department of Mechanical Engineering, The Ohio State University, Columbus, Ohio 43210-I IO7

(Received 14 December 1992; accepted for publication 21 May 1993)

Despite the rapid progress being made in the synthesis of diamond films and recent interest in polishing of diamond films, no systematic measurements of friction and wear on polished diamond films have been reported. In the present study, chemomechanical and laser polishing techniques are used, and friction and wear data on the chemomechanically polished diamond films are presented. With the chemomechanical polishing technique used in this study, the rms roughness of hot Filament chemical vapor deposited diamond films can be reduced from about 657 to about 170 nm with rounding off of sharp asperities with no change in the diamond structure. The polished films exhibit coefficient of friction ( -0.1) and wear rates much lower than that of unpolished films. Friction and wear properties of the polished films are comparable to that of single-crystal natural diamond. Based on this study, it is concluded that polished films are potential candidates for tribological applications.

1. INTRODUCTION

A natural diamond is well known for its low friction and wear properties. l-5 Natural diamonds are very expen- sive and cannot be formed as thin films. The physical and chemical properties of chemical vapor deposition (CVD) diamond films have been found to be comparable to those of bulk diamond.6 Sawabe and Inuzuka7 reported that electron-assisted CVD films had electrical resistivities greater than 1013 J-I cm, microhardness of about 10 000 HV, and a thermal conductivity of about 1100 W m-* K-’ close to that of bulk natural diamond (lo’- 1020 ti cm, 8000-10 400 HV, and 900-2100 W m-’ K-l). Kuo et aL8 reported that microwave- assisted CVD diamond tirn had a thermal diffusivity of about 200-300 mm2/s, a value close to that of bulk dia- mond (490-1150 mm2/s). Based on thermal gravimetric analysis,’ oxidation rates of PECVD diamond films in the range of 500 to 750 “C have been found to be lower than those of natural diamond. Zhu et al. lo have reported that the starting temperature of oxidation for microwave- assisted CVD diamond film is about 800 “C!, as evidenced by weight loss, while the surface morphology shows visible oxidation etching pits at temperatures as low as 600 “C!.

A major roadblock to the wide spread use of diamond &lms in tribological, optical, and thermal management ap- plications, is the surface roughness. Growth of the dia- mond phase on a nondiamond substrate is initiated by nu- cleation either at randomly seeded sites or at thermally favored sites due to statistical thermal Auctuation at the substrate surface. Based on growth temperature and pres- sure conditions, favored crystal orientations dominate the competitive growth process. As a result, the grown films are polycrystalline in nature with relatively large grain size ( > 1 pm), and terminating in very rough surfaces with roughnesses ranging from about few tenth of a micron to tens of microns. These films have been reported to have high friction and wear and are not suitable for tribological applications.‘1-‘5 A crucial need exists to either develop

methods to deposit smooth diamond films or to post- process these films to reduce their roughness. Fine-grained diamond films have been deposited with low-surface roughness by several investigators’e--‘9 which exhibit low friction. However, these films are believed not to have true diamond structure. Therefore polishing of the true dia- mond film is required. Thermochemical, chemomechani- Cal, and plasma/ion beam/laser techniques have been ex- plored to polish diamond films. However, no friction and wear data exist on polished diamond films. In this paper, we describe chemomechanical and laser polishing tech- niques used in the present study and present friction and wear data on chemomechanically polished CVD diamond films in sliding contact with smooth alumina balls.

II. POLISHING TECHNIQUES FOR DIAMOND FILMS

During polishing, material removal takes place initially on the top of asperities resulting in a smoother surface. Since 1988, various physical and chemical means have been explored to polish diamond films. Traditional me- chanical polishing methods using diamond or other hard abrasive particles” are unsuitable because of extremely low polishing rates (and hence long polishing times) and pref- erential polishing along specific crystal orientations leaving protruding crystallites on the surface.20-22 The existing and potential methods can be broadly classified into three cat- egories: ( 1) thermochemical,23-27 (2) chemomechanical,28 and (3) plasma/ion beam/laser polishing.‘y-3y A summary of aforementioned polishing techniques is given in Table I. Thermochemical and chemomechanical techniques are contact techniques so these can be used for surfaces that are planar. Between these two techniques, chemomechan- ical is preferred as it does not require high sample temper- atures. Plasma, ion beam, and laser techniques do not re- quire bulk sample heating and are noncontact techniques which can be used on nonplanar surfaces, though these are line-of-sight processes. The material removal rates using these techniques have thus far been small. Consequently,

4174 J. Appl. Phys. 74 (6), 15 September 1993 0021-8979/93/74(6)/4174/7/$6.00 @I 1993 American Institute of Physics 4174

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

130.113.86.233 On: Fri, 19 Dec 2014 13:54:10

TABLE I. Summary of various polishing techniques.

Processing temperature Mechanism of material removal

Shape limitation

Size limitation

Special requirements Equipment cost Processing time

Reported roughness achieved

Potential for commercialization

graphitization and diffusion into metal plate only planar surfaces none

need controlled environment low several hours

100 nm

oxidation

only planar surfaces none

low medium few hours several hours

165 nm (present study)

Ambient

sputtering, etching

complex shapes possible

large-limited to reactor size

need high vacuum

50 mn (combined with mechanical polishing) high (for in situ polishing)

Ambient

sputtering, etching

complex shapes possible

large-limited to beam size (for large areas scan beam or sample) need high vacuum

high time dependent on the area of polishing 500 nm

Ambient

etching, graphitization and sublimation complex shapes possible

large-limited to beam size (for large areas scan beam or sample) need controlled environment high time depends on the area of polishing 97 nm (present study)

high (for in situ and ex situ polishing)

these techniques have primarily been used for final polish- ing (in order to achieve low surface roughness of about 5 to 10 nm rms) after coarse polishing using other tech- niques such as thermochemical or chemomechanical. High-energy ion beams and laser sources can be developed for polishing. High-energy ion beams have the potential for high material rates. However, heat generated during sput- tering results in graphitization of the diamond film, so that the laser technique emerges as the most desirable among the noncontact techniques. Ideally, development of in situ polishing techniques are needed for commercialization of polished diamond films. Reactive ion etching (RIE) can be readily used for in situ polishing in most existing plasma deposition methods, but laser technique would exhibit fast polishing times both in sim and ex situ, however at greater capital cost. Achievement of surface smoothness without large material removal using noncontact techniques re- quires surface planarizing before polishing. Planarizing can be achieved by coating the film surface with a photoresist or gold as in the case of RIE and ion bombardment in order to match material removal rates.

In this paper, we have used the chemomechanical and laser techniques to polish HFCVD diamond films, and fric- tion and wear data for the films polished using chemome- chanical technique are reported.

Ill. PREPARATION OF POLISHED DIAMOND FILMS AND EXPERIMENTAL PROCEDURE

Diamond films were deposited on ( 100) single-crystal silicon wafers using a previously described HFCVD technique.40 Before growth, the silicon wafer was abraded with 4-8 pm-sized diamond grit and then cleaned in an

4175 J. Appl. Phys., Vol. 74, No. 6, 15 September 1993

ultrasonic bath using acetone. The deposition conditions used were: initial substrate to filament distance= 10 mm, substrate temperature = 800 “C!, hydrogen flow rate = 99 seem, methane flow rate= 1 seem, chamber pressure==20 Torr, deposition time=24 h, average film thickness=20 pm, and deposition area= 1 cm2.

In the chemomechanical polishing technique used in this study, a fused oxidizer, potassium nitrate ( KN03) was utilized. The choice of the oxidizer was made due to its well-known strong reactive etching affinity towards diamond.4 The etching rate in the presence of a shearing force field is expected to be higher and more uniform. Therefore, diamond films were polished by mechanical lap- ping against polycrystalline alumina for coarse polishing and then against another coarsely polished diamond film with the fused oxidizer in the gap. The polishing apparatus shown in Fig. 1 consisted basically of a hot plate and a rotating stage. The hot plate was used to mount either the mating alumina or one of the diamond films. The rotating stage was used to mount another diamond film and oper- ated at 7 rpm. During polishing, the hot plate was main- tained at 325 “C! in order to maintain the potassium nitrate in the fused form. During the polishing process, the dia- mond films were coarsely polished against the alumina sur- face for 3 h followed by fine polishing against another coarsely polished diamond film for 2 h.

In the laser polishing technique, diamond films were polished using an 193 nm wavelength ArF excimer laser pulsed at 100 mJ at a repetition rate of 20 Hz, focused at normal incidence on the diamond surface and in the pres- ence of an oxygen ambient. We were able to achieve pol- ishing down to 97 nm rms over an area of approximately

Bhushan et a/. 4175 [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

130.113.86.233 On: Fri, 19 Dec 2014 13:54:10

Unidirectional sliding 7 rev. per minute

Silicon substrate lOxlOx0.3mm

Oxidizer Rough diamond film

Hot plate (Temp. 325’C)

FIG. 1. Schematic of the polishing setup using chemomechanical tech- nique.

50 mm2 in 5 min. We clearly note that laser polishing has significant potential for polishing diamond films in a mat- ter of a few minutes.

Unpolished and polished HFCVD diamond films were characterized using Raman spectroscopy, scanning elec- tron microscopy (SEM), atomic force microscopy ( AFM) , and a stylus profiler. Statistical roughness param- eters are dependent on the lateral resolution of measuring instruments, therefore, roughness values obtained using the AFM and the stylus profiler were different. For conve- nience, the data obtained by stylus profiler are only re- ported in the text. Polished single-crystal natural diamond (IIa) (from Double Dee Harris Co.) was also used for reference. In Raman measurements, 5 14.5 nm radiation at a power of 150 mW with a spot size of about 1 mm was used.

Accelerated friction and wear tests were conducted on polished diamond films. Tests were carried out in recipro- cating mode on test coupons using a ball-on-flat tribome- ter. Alumina balls of 5 mm diameter and with a surface finish of about 3 nm rms were fixed in a stationary holder and the diamond-coated silicon coupons were mounted on a reciprocating table. The friction force was measured us- ing a strain gage ring. Wear of the diamond film and the mating balls was estimated using microscopy and surface profiling before and after the wear tests. Additionally, Ra- man spectroscopy and SEM imaging were performed be- fore and after the wear test. The tests were carried at dif- ferent loads, various speeds, in air at ambient temperatures (22 “C) and under dry conditions. Limited tests were con- ducted under lubricated conditions. Multiple tests were run at different conditions and the reproducibi&v was more than 95%. The typical test conditions were as follows: re- ciprocating amplitude 0.8 mm, normal load 0.1 to 10 N, frequency 0.1 to 10 Hz, average linear speed 0.08 to 8 mm/s, test temperature 20 “C, ambient humidity (35% RH) and the ambient air. For lubricated tests, about 10 cc of lubricant was placed in a reservoir surrounding the ball- film interface. The interface was flooded with a petroleum- based oil (68% paraffinics, 28% naphthenes, 4% aromat- ics) lubricant without any additive.

I 1332 cm-l (cl

I , t ! 1 I

1CiM 1400 1800

RAMAN SHIFT, cm-’

FIG. 2. Raman spectra of (a) unpolished diamond film, (b) chemome- chanically polished diamond films, and (c) worn chemomechanically pol- ished diamond films. Higher background in the Raman spectra of pol- ished and worn films is due to more fluorescence because of their damage.

IV. EXPERIMENTAL RESULTS AND DlSCUSSlON

A. Surface characterization

Figure 2 shows the Raman spectra of unpolished film, polished film before the wear test, and polished film after the wear test. The sharp line at or near 1332 cm-’ is attributed to diamond and can easily be distinguished from other allotropes such as graphite (D peak at 1343 cm-’ and G peak at 1600 cm-‘) and sp’-bonded amorphous carbon (broad-band centered at 1550 cm-‘) also known as diamond-like carbon. The Raman spectrum of polished di- amond film after wear test shows that there is no phase change, however, the reduction in intensity of the diamond Raman peak may have been from the presence of alumina wear debris. The Raman data show that there is no signif- icant increase in graphitization or amorphization as a re- sult of polishing. Figures 3-5 show SEM micrographs, AFM images, and stylus profiler scans of unpolished and polished surfaces. We note that the film surface has been planarized with asperity peaks having been knocked off or rounded. We also note that the rms roughness for polished surfaces has decreased from about 657 nm to about 170 and 97 nm for chemomechanically and laser-polished sam- ples, respectively. Although the rms roughness of the ch- emomechanically polished film is slightly higher than that of the laser-polished film, the peak-to-valley (P-V) dis- tance is lower and as a result of run in during mechanical lapping, the asperities appear to be more rounded which may be desirable for low friction and wear.

B. Friction and wear data

The friction data of typical as-deposited diamond films, chemomechanically polished diamond films, and single-crystal natural diamond (Ha) are shown in Fig.

4176 J. Appl. Phys., Vol. 74, No. 6, 15 September 1993 Bhushan et a/. 4176 [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

130.113.86.233 On: Fri, 19 Dec 2014 13:54:10

FIG. 3. SEM micrographs of (a) unpolished, (b) chemomechanically polished, (c) laser polished diamond films. Note (111) orientation of the diamond films.

6 (a). The polished diamond films exhibit low coefficients of friction (about 0.11) when slid against the alumina ball, much lower than that for unpolished films and very close to the values exhibited by natural diamond. We note that the coefficient of friction of polished film and natural dia- mond remained constant throughout the sliding test. How- ever, friction decreased during sliding in the case of unp- olished film. This observation suggests that the chemomechanically polished films do not exhibit any run-in effect. Figure 7 shows SEM micrographs for pol- ished and unpolished films and mating ball surfaces after wear. The wear of the mating alumina ball slid against an unpolished film indicated by the scratches on the ball sur- face and wear debris generated during the test and depos- ited on the unpolished diamond film, is significant. The rough as-deposited diamond film cuts through the alumina ball easily like “cheese” and a large concentration of alu- mina shavings on the film surface can be clearly seen. The wear of the alumina ball was negligible in the case of the

chemomechanically polished diamond film, comparable to wear for the natural diamond sample. AFM imaging of the polished diamond film and of natural diamond after wear showed negligible change as a result of the sliding process.

Coefficient of friction was found to be strongly depen- dent on the direction of the polishing tracks. We found that the coefficient of friction was high (0.25) when the alu- mina ball was slid perpendicular to the polishing tracks (i.e., radial direction of the sample during polishing) and low (0.12) when slid parallel to or along the polishing tracks. This difference in friction is attributed to the dif- ference in the surface roughness along (- 125 nm rms roughness) the polished track versus across ( -200 nm rms roughness) the polished tracks.

The role of roughness of polished diamond film on friction is illustrated in Fig. 6(b). The coetlicient of friction decreased from 0.4 to 0.09 as the rms roughness of the polished films was reduced from 500 to 30 nm. There are three mechanisms of friction-adhesion, ploughing, and ratchet that may be operative either individually or in com- bination for diamond films sliding against a softer material. If two rough surfaces are placed in contact, in addition to adhesion at the asperity contacts, asperities of the mating surfaces may interlock. The interlocked asperities either ride over each other (ratchet mechanism) or push past each other or fracture (ploughing) during relative motion. The adhesion mechanism does not appear to account for the observed reduction in friction as the real area of con- tact or friction is expected to increase with a decrease in roughness. With rough diamond films sliding against a softer material, ploughing and ratchet may be operative. Since the slope of the mating ball asperities are lower than that of polished diamond films, ball asperities can not ride over asperities of the diamond film. Therefore, ratchet mechanism is not relevant here and ploughing is expected to be the major contributor to the etTect of roughness on the friction. The rounding off of asperities from polishing facilitate an easy slip which minimize interface damage.

Blau et al. I1 reported that the coefficient of friction of as-deposited HFCVD diamond films were up to ten times the value for smooth bulk diamond surfaces sliding against various solids in air. They also reported very high wear of a mating sapphire ball surface. Gardos and Soriano12 and Kohzaki et al. l3 also reported rather high values of coeffi- cient of friction and wear rates of PACVD diamond films slid against Sic pins. Jahanmir et a1.16 and Hayward et al. ” reported lower values of coefficient of friction ( - 0.1) of various diamond films sliding against a diamond or sapphire slider. The surface roughnesses of their films were as low as 7 nm rms which suggests that these films were fine grained and probably with significant nondia- mond components. (For friction and wear properties of as-deposited diamond films, also see Refs. 14 and 15.) Plano et al. ‘* and Wu et al. tg have reported the deposition of c‘fine-grained” (20-100 nm) diamond films having a relatively smooth surface ( 15 nm rms) . They produced fine-grained films using a low-pressure and low-power microwave-assisted CVD technique on (100) Si surface having scratched with diamond paste prior to growth.

4177 J. Appl. Phys., Vol. 74, No. 6, 15 September 1993 Bhushan et al. 4177 [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

130.113.86.233 On: Fri, 19 Dec 2014 13:54:10

(b) “”

FLG. 4. AFM images of (a) unpolished (rms=304 nm, PV=1610 nm), (b) chemomechanically polished film (rms=150 nm, PV=995 nm), (c) laser polished film (rms=46 run, PV=320 nm), and (d) natural diamond surface (for reference) (rms=2.4 nm, PV=28 nm).

Wu et ai. l9 reported coefficient of friction of 0.03 for their “fine-grained” diamond film sliding against natural dia- mond in air. Although no data were presented in this work regarding friction of CVD diamond sliding against alu-

‘” pJy~/-JlPf~~

1000 nm

"

400 nm

0

40 nm

0 I I 0 250 so0 Pm

FICr. 5. Surface roughness profiles using a stylus profiler of (a) unpol- ished film (rms=657 mn), (b) chemomechanically polished film (rms =170 nm), (c) laser polished film (rms=97 nm), and (d) natural dia- mond surface (rms= 1.7 nm).

4178 J. Appl. Phys., Vol. 74, No. 6, 15 September 1993 Bhushan et a/. 4178

mina, it is likely that the coefficient of friction for such a case would be larger than 0.03. Further, the significant amorphous carbon peak evident from the Raman spectra of these fine grained diamond films shows that these may not be true diamond, but composite films of diamond and nondiamond carbon. Therefore if the coarse-grained (true) diamond films are required for tribological applications with low friction and wear, they need to be polished.

Coefficient of friction of polished diamond films at dif- ferent loads and speeds is shown in Fig. 8. Friction data for natural diamond at different loads and speeds are also shown for comparison. The coefficient of friction of the polished diamond tim did not increase significantly as the normal load16 increased from 0.1 to 10 N and the speed increased from 0.08 to 8 mm/s. However, the lower coef- ficient of friction (0.05) at higher load (10 N) was ob- served when alumina ball slid over the natural diamond.

Some roughness on the surface of the diamond film may actually be useful in lubricated sliding because the valleys could act as reservoirs for a liquid lubricant. Fric- tion and wear tests on polished films were also carried out under lubricated conditions. The coefficient of friction val- ues and optical micrographs of worn ball surfaces (alu- mina and 52100 steel balls) slid against polished diamond film under dry and lubricated conditions are shown in Fig. 9. As expected, the coefficient of friction and wear scar diameters are significantly lower for the case of lubricated sliding. It appears that lower friction and wear of soft mat- ing surfaces can be achieved in lubricated sliding.

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

130.113.86.233 On: Fri, 19 Dec 2014 13:54:10

t2 3 Unpolished film

3 0.4 j-p++--+ ---- +-f--j

%

5 10 Sliding distance, m

(4

3 3 -g 0.4

‘tj

3 :$ 0.2

x v

0 0 100 200 300 400 500

Rms roughness, nm

(b)

FIG. 6. Variation of the coefficient of friction (a) as a function of sliding distance for unpolished (rms =657 nm) , chemomechanically polished (rms=170 nm), and natural diamond (rms=1.7 nm) films (b) as a function of roughness of a chemomechanically polished film. These tests were carried out at 1 N normal load, 0.8 mm/s sliding speed and at 22 “C temperature and 30% relative humidity.

V. CONCLUSIONS

We have shown that HFCVD “true” diamond films can be polished without any graphitization or significant amorphization during polishing. The polished diamond films and natural single-crystal diamond (IIa) slid against alumina balls (of about 3 nm rms roughness) have com- parable coefficients of friction and observed wear of the mating alumina balls. The coefficient of friction of polished diamond films (with average roughness on the order of 170 nm) exhibit reproducible decrease in the coefficient of fric- tion from about 0.4 down to about 0.09 when compared with their unpolished counterparts. The wear debris gen- erated from the alumina ball during the test with the,un- polished films was significant but was negligible with pol- ished diamond films. The friction and wear data of the polished films are almost identical to those obtained for single-crystal natural diamond samples with average roughness of 1.7 nm.

The operative mechanism for the role of roughness on friction appears to be ploughing. In sliding with polished diamond films, the dominant effect on the coefficient of friction and wear of the mating ball surface is the rounding off of sharp asperities along with the reduction in the sur- face roughness (reduction in the asperity slope) while slid- ing against a smoother mating surface. We find that, it is

FIG. 7. SEM micrographs of worn diamond surfaces and alumina balls, (a) unpolished diamond film, (b) chemomechanically polished diamond ftlm, and (c) natural diamond film.

_

O3 :

Polished film

1.0

Normal load, N (4

o.3 I

Polished film

1

01 I 0.1 1.0 10.0

Sliding speed, mm/s (b)

FIG. 8. Variation of the coefficient of friction as a function of sliding distance for chemomechanically polished diamond film slid against alu- mina ball (a) at different loads and at 0.8 mm/s sliding speed, and (b) at different speeds and at 1 N normal load. Data for natural diamond at different loads and speeds are also included for comparisons. These tests were conducted at 22 “C! temperature and 30% relative humidity.

4179 J. Appl. Phys., Vol. 74, No. 6, 15 September 1993 Bhushan et al. 4179 [This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

130.113.86.233 On: Fri, 19 Dec 2014 13:54:10

Steel, dry, P = 0.25

Steel, lubricated, p = 0.10

FIG. 9. Optical micrographs of alumina and 52100 steel balls slid against polished diamond film (170 nm rms roughness) under dry and lubricated conditions. These tests were carried out at 1 N normal load, 0.8 mm/s sliding speed, 0.4 m m track length, and at 22 ‘C temperature and 30% relative humidity. Each test was conducted for a sliding distance of 8 m.

not necessary to achieve roughnesses on the order of a few nm for low friction and wear as long as the mating surface has a lower surface roughness compared to that of dia- mond film. Some roughness on the surface of the diamond film may actually be useful in lubricated sliding because the valleys could act as reservoirs for a liquid lubricant. From this study, it is concluded that polished diamond films can be used for tribological applications.

ACKNOWLEDGMENTS

Financial support for this research was provided in parts by Center for Automotive Research, industrial mem- bership of the Computer Microtribology and Contamina- tion Laboratory, and MSS 9157303 Grant from the Na- tional Science Foundation.

‘M. Seal, Ind. Diarn. Rev. 25, 111 (1965). zM. Seal, Philos. Mag. A 43, 587 (1979). ‘D. Tabor, in The Properties of Diamond, edited by J. E. Field (Aca- demic, New York, 1979), pp 325-350.

4J. Wilks and E. M. Wilks, in Ref. 3, pp. 351-382. “R. J. Wedlake, in Ref. 4, pp. 501-535. ‘B. Bhushan and B. K. Gupta, Handbook of Tribologv: Materials, Coat-

ings, and Surface Treatments (McGraw-Hill, New York, 1991). i A. Sawabe, and T. Inuzuka, Thin Solid Fi lms 137, 89 (1986). “P. K. Kuo, L. Wei, R. L. Thomas, and R. W . Pryor, ‘in Photoacoustic

and Photothermal Phenomena fl, edited by J. C. Murphy, J. W . M. Spicer, L. C Aamodt, and B. S. H. Royce, Springer Series in Optical Sciences Vol. 62 (Springer, Heidelberg, 1990), pp. 124-126.

‘C. E. Johnson, A. S. Hastings, and W . A. Weimer, .I. Mater. Res. 5, 2320 (1990).

4180 J. Appl. Phys., Vol. 74, No. 8, 15 September 1993

lo W . Zhu, X. H. Wang, A. R Badzian, and R. Messier, in New Diamond Science and Technology (MRS Int. Conf. Proc., 1991), pp. 821-826.

“P. J. Blau, C. S. Yust, L. J. Heatherly, and R. E. Clausing, in Mechanics of Coatings, edited by D. Dowson, C. M. Taylor, and M. Godet, Leed- Lyon Tribology Series 17 (Elsevier Science, Amsterdam, 1990), pp. 399407.

“M. N. Gardos and B. L. Soriano, J. Mater. Res. 5, 2599 (1990). t3M. Kohzaki, K. Higuchi, S. Noda, and K. Uchida, J. Mater. Res. 7,

1769 (1992). 14K. Okada and Y. Narnba, J. Vat. Sci. Technol. A 7, 132 (1989). “C. T. Kuo, T. Y. Yen, T. H. Huang, and S. E. Hsu, J. Mater. Res. 5,

2515 (1990). 16S. Jahanmir, D. E. De&man, L. K. Ives, A. Feldman, and E. Fara-

baugh, Wear 133, 73 (1989). “I. P. Hayward, I. L. Singer, and L. E. Seitzman, Wear 157,215 (1992). ‘sL. S. Plano, I. Hayward, and J. Wegand, Report No. AD-A 220 265,

GNR, Code 1114, Arlington, Virginia, March 1990. 19R L. C. Wu, A. K. Rai, A. Garscadden, P. Kee, H. D. Desai, and K.

Miyoshi, J. Appl. Phys. 72, 110 (1992). “M. N. Gardos, B. W . Buller, and J. M. Pinneo, in Applications of

Diamond Films and Related Materials, edited by Y. Tzeng, M. Yoshikawa, M. Murakawa, and A. Feldman (Elsevier Science, Amster- dam, 1991), pp. 129-135.

2’C. F. Hickey, T. P. Thorpe, A. A. Morrish, J. E. Butler, C. Void, and K. A. Snail, SPIE Diamond Optics IV 1534, 67 (1991).

‘sM. B. Moran, K. A. Klemm, and L. F. Johnson, in Ref. 20, pp. 233- 240.

z’C. F. Yang, H. Tokura, and M. Yoshikawa, Tokyo Institute of Tech- nology, Ookayama, Tokyo 152, Japan.

“T. P. Thorpe, A. A. Morrish, and L. M. Hanssen, SPIE Diamond Optics III 1325, 230 (1990).

“M. Yoshikawa, SPIE Diamond Optics III 1325, 210 (1990). 26Applications of Diamond Eilms and Related Materials, edited by Y.

Izeng, M. Yoshikawa, M. Murakawa, and A. Feldman (Elsevier Sci- ence, Amsterdam, 1991).

*‘H Tokura, C. F. Yang, and M. Yoshikawa, Thin Solid films 212, 49 ( 1992).

“A. G. Thronton and J. W&s, Suppl. Ind. Diamond Rev. 39 (1974). “M. Seal, Phys. Status Solidi 3, 658 (1963). “T Evans and P. F. James, Proc. R. Sot. London Ser. A 277, 260

(1964). 3’ 4 B. Harker, J. Flintoff, and J. F. De Natale, SPIE Diamond Optics III

Ii,, 222 ( 1990). “A. Hirata, H. Tokura, and M. Yoshikawa, in Ref. 20, pp. 227-232. 33A. Hirata, H. Tokura, and M. Yoshikawa, Thin Solid Fi lms 212, 43

(1992). “T. Zhao, D. F. Grogan, B. G. Bovard, and H. A. Macleod, SPIE

Diamond Optics III 1325, 142 (1990). 35B. G. Bovard, T. Zhao, and H. A. Macleod, Appl. Opt. 31, 2366

(1992). 36D F Grogan, T. Zhao, B. G. Bovard, and H. A. Macleod, Appl. Opt.

3;, 1483 (1992). “V P Ageev, L. L. Bouilov, V. I. Konov, A. V. Kuzmichev, S. M. . .

Pimenov, A. M. Prokhorov, V. G. Ralchenko, B. V. Spitsyn, and N. I. Chapliev, Sov. Phys.-Dokl. 33, 840 ( 1988).

‘*L. L. Bouilov, N. I. Chapliev, V. I. Konov, S. M. Pimenov, A. A. Smolin, and I. V. Spitsyn, J. Electrochem. Sot. 91, 357 (1991).

39U. Bogli, A. Blatter, S. M. Piienov, A. A. Smolin, and V. I. Konov, Diamond Relat. Mater. 1, 782 ( 1992).

4oJ. H. D. Rebello, D. L. Straub, V. V. Subramaniam, E. K. Tan, S. A. Dregia, B. L. Preppernau, and T. A. Miller, Mater. Manufactur. Pro- cesses 6, 501 (1992).

Bhushan et al. 4180

[This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to ] IP:

130.113.86.233 On: Fri, 19 Dec 2014 13:54:10