Surface & Coatings Technology 201 (2006) 164 – 167

www.elsevier.com/locate/surfcoat

Synthesis of carbon films with ultra-low friction in dry and humid air
Christina A. Freyman, Yanfeng Chen, Yip-Wah Chung ⁎
Department of Materials Science and Engineering, Northwestern University, 2220 N Campus Dr., Evanston, IL 60208, USA
Received 8 July 2005; accepted in revised form 7 November 2005
Available online 13 December 2005

Abstract

Using pulsed dc magnetron sputtering, we synthesized hydrogenated amorphous carbon films with and without sulfur doping. These films
were smooth and amorphous, with low compressive stress (b1 GPa). Auger electron spectroscopy confirmed the incorporation of sulfur, and
X-ray photoelectron spectroscopy showed that sulfur atoms are chemically bound. Most significant, ball-on-flat tribo-testing showed that
hydrogenated carbon films doped with 5 at.% sulfur have ultra-low friction coefficients (b0.01) in air with relative humidity from 0% to
50%.
© 2005 Elsevier B.V. All rights reserved.

Keywords: Hydrogenated carbon films; Sulfur doping; Friction; Magnetron sputtering

1. Introduction transfer layer [12,13]. The strong dependence on humidity has

Hydrogenated amorphous carbon films (CHx) have been
extensively investigated due to their wide range of properties.
These films can be synthesized through a variety of physical
and chemical techniques. Tribological properties of these films
depend on testing conditions (environment, load, speed and
geometry), film structure and composition, and surface rough-
ness [1,2]. Heavily hydrogenated carbon films deposited by
chemical vapor deposition have been shown to provide ultra-
low friction coefficients (b0.01) in vacuum and dry nitrogen
environments. This ultra-low friction between two coated sur-
faces has been attributed to hydrogen-terminated surfaces with
weak van der Waals interaction, resulting in weak adhesion and
hence low shear strength against sliding [3–7]. Their friction
coefficients increase rapidly in the presence of oxygen and
humidity. Oxygen can interact with both the carbon and hydro-
gen atoms [8,9]. Tribochemical reactions resulting in oxidation
of the film have also been observed to increase friction. Atomic
oxygen increases the friction coefficient much more than mole-
cular oxygen [10,11]. Tribochemical oxidation from oxygen or
water has also been observed to affect the transfer layer, even at
low partial pressures, increasing friction coefficients. This is
attributed to the oxygen removal of sp2 carbon atoms from the
⁎ Corresponding author. Tel.: +1 847 491 3112; fax: +1 847 491 7820.
E-mail address: ywchung@northwestern.edu (Y.-W. Chung).
0257-8972/$ - see front matter © 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.surfcoat.2005.11.075
been attributed to viscous and capillary forces induced by
adsorbed water [14].
There have been efforts to mitigate the humidity problem.
Silicon, fluorine, boron, gold, and titanium doping have been
attempted, but they have only marginal effects on the moisture
sensitivity of CHx films. Friction coefficients of 0.04–0.1 were
measured in ambient air, at least 10 times higher than those mea-
sured in dry air by Erdemir et al. [1,15–18]. Silicon and boron
dopants are thought to form a silicon oxide or boric acid lubricous
layer on the surface in humid air, decreasing the ambient friction
coefficient [1,19,20]. Gold acts as a solid lubricant because of its
low shear strength [21]. Reduction in the surface energy using
fluorine-doped films is believed to be the primary reason for the
reduced sensitivity of friction towards moisture [22].
Tagawa et al. observed a direct link between water coverage
on a hydrogenated carbon surface and friction coefficient. The
friction coefficient increased markedly when the surface water
coverage exceeded about one monolayer [23]. Shulka et al.
measured the adsorption of water to be ∼1.5 monolayers on
hydrogenated carbon films at 60% relative humidity [24]. There-
fore, reduction in the surface water coverage under humid con-
ditions may allow these films to retain their low friction
properties. The ability of certain surface dopants to suppress
gas adsorption is well-documented in the surface science litera-
ture, due to physical site-blocking or modification of surface
electronic properties [25]. In principle, one can minimize water
 C.A. Freyman et al. / Surface & Coatings Technology 201 (2006) 164–167 165

adsorption by proper surface texturing [26,27], but this may lead

to high friction due to mechanical interlocking or plastic defor-
mation of asperities.
Here, we report the synthesis and characterization of sulfur-
doped hydrogenated carbon films. These films exhibit ultra-low
friction behavior in ambient humid air. We believe that this is
due to suppressed water adsorption on these sulfur-doped
hydrogenated carbon film surfaces.

2. Experimental

Hydrogenated carbon films were deposited in a single-cath-

ode unbalanced magnetron sputter-deposition system with a
base pressure of better than 6.6 × 10− 6 Pa (5 × 10− 8 Torr). A
5-cm-diameter pure graphite target was used. The target power
was set at 100 W pulsed at 150 kHz (80% duty cycle), with
the positive voltage set at 10% of the negative voltage. The
sputter-gas was an argon–hydrogen–hydrogen sulfide mix-
ture. The argon–hydrogen ratio remained constant at 15%
hydrogen, balance argon at a sputtering pressure of 6 mTorr
(0.8 Pa). For results reported in this paper, the hydrogen
sulfide partial pressure was set at 0.02 mTorr, resulting in
5.0 at.% sulfur concentration in these carbon films. Low-
resistivity silicon (100) wafers and 6.35-mm stainless steel
ball bearings were used as substrates. Prior to carbon film
deposition, the ball bearings were coated with 75 nm of
silicon by sputtering to improve film adhesion. All substrates
were ultrasonically cleaned in acetone and methanol before
being introduced into the deposition chamber and sputter-
cleaned in argon before deposition. The substrate was pulse-
biased at − 25 V, 20 kHz to improve the efficiency of ion
bombardment during film growth. The film thickness was set
at 1.0 μm for all depositions.
Auger spectra were taken using a double-pass cylindrical
mirror analyzer (Staib Instruments). X-ray photoelectron spec-
troscopy was performed on an ESCA Probe (Omicron). Fric-
tion properties were measured using a CETR Micro-Tribometer
(UMT) in a standard ball-on-flat configuration at room tem-
Fig. 2. Sulfur 2p core level spectrum from a sulfur-doped hydrogenated carbon
film.
perature. The sliding speed was 0.01 m/s. Both the ball and flat
were coated. The tribometer was enclosed in a glove box to
enable controlled atmosphere testing. Humidity was reduced by
the introduction of dry air. Surface roughness was measured
with a Nanoscope IIIa atomic force microscope (Digital Instru-
ments) in contact mode. Hardness and elastic modulus were
measured using a nanoindentor (Hysitron) with the supplied
software, which analyzed the load–displacement curves with
the method of Pharr and Oliver [28].
3. Results and discussion
Sulfur incorporation was verified using Auger electron
spectroscopy. As illustrated in Fig. 1, the surface composition
is stable with respect to annealing at 300 °C. Auger sputter-
profiling showed that sulfur is uniformly distributed within
these films. Separate Rutherford backscattering measurements
showed that these films contained ∼ 30 at.% hydrogen. X-ray

Fig. 1. Auger analysis of 5 at.% sulfur-doped hydrogenated carbon films. The

bottom spectrum was taken after sputter-cleaning at room temperature, followed
by a series of Auger spectra taken after the indicated time of annealing at 300 Fig. 3. Transmission electron micrograph from a sulfur-doped hydrogenated
°C. There is no change in the sulfur-to-carbon ratio as a result of annealing. carbon film.
166 C.A. Freyman et al. / Surface & Coatings Technology 201 (2006) 164–167
Fig. 4. (a) A 10-ml drop of water on an undoped hydrogenated amorphous carbon film. (b) A 10-ml drop of water on 5 at.% sulfur-doped hydrogenated amorphous
carbon film.
photoelectron spectroscopy showed the sulfur 2p-core level at
163.3 eV for the sulfur-doped carbon film. Compared with
the 2p binding energy [29] of elemental sulfur (164.05 eV),
one may infer that the incorporated sulfur atoms are in a
chemically bound state as shown in Fig. 2. Film hardness
decreases from 13 GPa to 8 ± 1 GPa with sulfur doping of 5
atomic percent (at.%). This is most likely due to sulfur
interruption of the 3D-bonding network of carbon atoms.
As shown in Fig. 3, plan-view transmission electron micro-
scopy showed that these films are amorphous. The internal
stress in these sulfur-doped films was typically 1 GPa or less,
as determined by the wafer curvature method [29]. The RMS
surface roughness, when measured at a scan size of 20 μm,
was 0.33 ± 0.10 nm for undoped films and 0.52 ± 0.15 nm for
the doped films.
Fig. 4 shows the marked increase of water contact angle
from 64° to 86° after the addition of 5 at.% sulfur, an indication
of increased hydrophobicity with sulfur doping. Fig. 5(a) shows
Fig. 5. Friction-versus-time plots for undoped and doped films in different
humidity environments. (a) Friction coefficient versus time for an undoped
hydrogenated carbon film in dry air (relative humidity b 1%). (b) Friction
coefficient versus time for undoped and 5 at.% sulfur-doped hydrogenated
carbon films at a relative humidity of 50%.
a typical friction-versus-time trace with undoped hydrogenated
carbon films in dry air in a ball-on-flat configuration. The initial
transient is typical of what has been reported in the literature
and this was attributed to the removal of adsorbed contami-
nants. The steady-state friction coefficient is 0.003 ± 0.002,
comparable to that reported by Erdemir and coworkers. Level-
ing of the testing stage was critical when measuring friction
coefficients in this range.
Fig. 5(b) shows that when tested under exactly the same
conditions but in air with 50% relative humidity, the friction
coefficient of undoped films increased to 0.020 ± 0.005. On the
other hand, doping these carbon films with sulfur reduces the
humidity sensitivity of friction. These doped carbon films have
friction coefficients in the range of 0.003 ± 0.002 in dry air and
0.004 ± 0.002 in laboratory air with 50% relative humidity. The
dependence of friction coefficient on relative humidity for
undoped and doped carbon films is summarized in Fig. 6.
As suggested by the XPS data, sulfur doping may result in
the formation of thiol-like (–C–S–H) groups on carbon film
surfaces. The binding of water molecules on hydrogenated
carbon film surfaces is primarily due to dipole–dipole forces.
The dipole moment of C–H species, the major polar species
on the hydrogenated carbon surface, is 1.46 D (1 D = 3.34 ×
10− 30 C m), whereas the dipole moment of S–H species is 0.76
D. The smaller dipole moment of S–H species results in weaker
binding between water and the surface, and hence lower equili-
brium water coverage. Together with the result of Tagawa et al.,
this may explain the improved friction performance of these
sulfur-doped hydrogenated carbon films in humid air.
Fig. 6. Friction coefficient versus humidity for undoped and 5 at.% sulfur-doped
films.
 C.A. Freyman et al. / Surface & Coatings Technology 201 (2006) 164–167
4. Conclusions
In summary, we have synthesized smooth hydrogenated
carbon films by magnetron sputtering, with and without sulfur
doping. Ball-on-flat tribo-testing at various humidity levels at
room temperature showed that sulfur doping retains the ultra-
low friction performance of undoped films, with much reduced
sensitivity towards humidity. Reduction of adsorbed water
molecules by sulfur doping may be the primary cause of the
improved friction performance observed in our experiments.
Acknowledgments
We gratefully acknowledge many fruitful discussions with
Kathy Wahl, Nick Wu, Ray Twesten, and especially Ali Erde-
mir for his critique of the manuscript. This work was supported
by NSF IGERT Grant #DGE-0114429/007. TEM studies were
carried out in the Center for Microanalysis of Materials, Uni-
versity of Illinois, which is partially supported by the U.S.
Department of Energy under grant DEFG02-91-ER45439.
XPS work was carried out at the Keck-II Facility of North-
western University.
References
[1] A. Erdemir, Diamond-like Carbon Films, ASME Press, New York, 2004,
p. 139.
[2] J.A. Heimberg, K.J. Wahl, I.L. Singer, A. Erdemir, Appl. Phys. Lett. 78
(2001) 449.
[3] A. Erdemir, O.L. Eryilmaz, G. Fenske, J. Vac. Sci. Technol., A, Vac. Surf.
Films 18 (2000) 1987.
167
[4] A. Erdemir, O.L. Eryilmaz, I.B. Nilufer, G.R. Fenske, Surf. Coat. Technol.
133 (2000) 448.
[5] J. Fontaine, C. Donnet, A. Grill, T. LeMogne, Surf. Coat. Technol. 146
(2001) 286.
[6] J. Andersson, et al., Surf. Coat. Technol. 163–164 (2003) 535.
[7] J. Fontaine, C. Donnet, A. Grill, T. Le Mogne, Surf. Coat. Technol. 146–
147 (2001) 286.
[8] A. Erdemir, Trib. Int. 37 (2004) 577.
[9] B. Marchon, M.R. Khan, N. Heiman, P. Pereira, A. Lautie, IEEE Trans.
Magn. 26 (1990) 2670.
[10] H. Li, et al., Appl. Surf. Sci. 249 (2005) 257.
[11] D. Paulmier, H. Zaidi, H. Nery, T.L. Huu, T. Mathia, Surf. Coat. Technol.
62 (1993) 570.
[12] J. Fontaine, T. Le Mogne, J.L. Loubet, M. Belin, Thin Solid Films 482
(2005) 99.
[13] W. Zhang, A. Tanaka, K. Wazumi, Y. Koga, Tribol. Lett. 14 (2003) 123.
[14] J. Andersson, R.A. Erck, A. Erdemir, Surf. Coat. Technol. 163 (2003) 535.
[15] A.K. Gangopadhyay, P.A. Willermet, M.A. Tamor, W.C. Vassell, Trib. Int.
30 (1997) 9.
[16] R. Gilmore, R. Hauert, Surf. Coat. Technol. 133–134 (2000) 437.
[17] R. Gilmore, R. Hauert, Thin Solid Films 398–399 (2001) 199.
[18] B. Racine, et al., J. Appl. Phys. 90 (2001) 5002.
[19] S.H. Yang, H. Kong, K.R. Lee, S. Park, D.E. Kim, Wear 252 (2002) 70.
[20] K. Oguri, T. Arai, J. Mater. Res. 7 (1992) 1313.
[21] X.B. Yan, et al., Appl. Phys., A Mater. Sci. Process. 81 (2005) 197.
[22] R. Gilmore, R. Hauert, Thin Solid Films 398 (2001) 199.
[23] M. Tagawa, M. Ikemura, Y. Nakayama, N. Ohmae, Tribol. Lett. 17 (2004)
575.
[24] N. Shukla, et al., Tribol. Lett. 15 (2003) 9.
[25] D.W. Goodman, M. Kiskinova, Surf. Sci. 105 (1981) L265.
[26] A. Marmur, Langmuir 19 (2003) 8343.
[27] A. Lafuma, D. Quere, Nat. Mater. 2 (2003) 457.
[28] G. Pharr, W. Oliver, MRS Bull. 17 (1992) 28.
[29] M. Ohring, The Materials Science of Thin Films, Academic Press, San
Diego, 2001.