Humidity dependence on the friction and wear behavior of diamond-like
carbon film in air and nitrogen environments
Hongxuan Li, Tao Xu, Chengbing Wang, Jianmin Chen , Huidi Zhou, Huiwen Liu
State Key Laboratory of Solid Lubrication, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, Lanzhou 730000, China
Received 6 July 2004; received in revised form 28 December 2005; accepted 28 December 2005
Abstract
Diamond-like carbon (DLC) films were deposited on Si (100) wafers by a plasma enhanced chemical vapor deposition (PECVD) technique
using CH4 plus Ar as the feedstock. The friction and wear behaviors of the resulting film sliding against Si3N4 balls were investigated on a ball-on-
disk test rig in air and nitrogen environments at a relative humidity from 5% to 100%. The worn surface morphologies of the DLC film and the
Si3N4 counterpart were observed on a scanning electron microscope (SEM), while the chemical states of some typical elements thereon were
investigated by means of X-ray photoelectron spectroscopy (XPS). It was found that the DLC film recorded continuously increased friction
coefficient and wear rate with increasing relative humidity in air. It showed linearly increased friction coefficient with increasing relative humidity
in nitrogen, in this case the wear rate sharply decreased and reached the minimum at a relative humidity of 40%, which was followed by an
increase with further increase of the relative humidity. The interruption of the transferred carbon-rich layers on the Si3N4 balls, and the friction-
induced oxidation of the films in higher relative humidity were proposed to be the main reasons for the increases of the friction coefficient and
wear rate. Moreover, the oxidation and hydrolysis of the Si3N4 ball in higher relative humidity, leading to the formation of a tribochemical film
that mainly consists of silica gel on the wearing surface, were also thought to have effects on the friction and wear behaviors of the DLC films.
 2006 Published by Elsevier B.V.
Keywords: PECVD; DLC film; Friction and wear behavior; Relative humidity
1. Introduction
Diamond-like carbon (DLC) films exhibit attractive me-
chanical and tribological properties (e.g. high hardness, low
friction, high wear resistance, etc.), which makes them suitable
for numerous potential tribological applications such as
magnetic hard disks, gears, bearings, cutting tools and other
moving mechanical assemblies [1­3]. However, previous
studies have shown that the tribological properties of the DLC
films are strongly dependent on the nature of the films and on
the deposition methods and conditions, and are very sensitive to
the testing environment, especially to the relative humidity
(RH) [4­14]. Depending on the above-mentioned factors, the
friction coefficients for the DLC films have been reported to
broadly range from 0.003 to more than 0.5 [7,9,13­19].
Various theories have been proposed to explain the
friction and wear behaviors of DLC films. The build-up of
a friction-induced transfer film on the counterpart, followed
by easy shear within the interfacial material is the most
frequently observed friction-controlling mechanism for DLC
films [3,10,20,21]. Liu et al. said that the steady-state low
friction of DLC films was due to the wear-induced
graphitization, i.e., the formation of a low friction graphitized
tribolayer [7,22]. Erdemir et al. investigated the tribological
properties of different DLC films in open air and dry
nitrogen, and argued that the general low friction coefficient
of hydrogenated DLC films was mainly attributed to the high
chemical inertness of the DLC films by passivating the
surface dangling bonds by hydrogen [17,23]. Many other
researchers proposed that the tribochemical reactions in the
tribo-system played an important role in determining the
friction and wear behaviors of DLC films [2,3,13,20].
However, these reported friction and wear mechanisms for
the DLC films are the subjects to controversy and the role of
Diamond & Related Materials xx (2006) xxx­xxx
+ MODEL
DIAMAT-04097; No of Pages 8
www.elsevier.com/locate/diamond
 Corresponding author. Tel.: +86 931 4968018; fax: +86 931 8277491.
E-mail address: chenjm@ns.lzb.ac.cn (J. Chen).
0925-9635/$ - see front matter  2006 Published by Elsevier B.V.
doi:10.1016/j.diamond.2005.12.048
ARTICLE IN PRESS
the environments on the friction and wear behaviors of DLC
films is still not well understood.
With that perspective in mind, the effect of relative humidity
(RH) on the friction and wear behavior of DLC film in air and
nitrogen environments was investigated, in a hope to reveal the
humidity dependence of the friction and wear behavior of the
DLC film in air and nitrogen and to clarify the friction and wear
mechanism of the DLC film as well.
2. Experimental details
2.1. Deposition and characterization of the DLC film
DLC film was deposited on Si (100) wafers making use of
PECVD technique, using CH4 plus Ar as the feedstock. The
details about the deposition equipment and process were
described elsewhere [5,24]. Prior to deposition, the substrates
were cleaned with Ar plasma sputtering at a bias voltage of
-400 V for 15 min so as to remove the native oxide on the Si
surface. The deposition conditions and some properties of the
resulting DLC film are summarized in Table 1.
The deposited DLC films are dark brown in color and are
extremely smooth and featureless when viewed with unaided
eyes. The film thickness measured on a profilometer is about
640 nm using. The hardness and Young's modulus of the film
are determined to be about 15.2 GPa and 120 GPa, respectively,
using a nano-indenter.
2.2. Friction and wear test
The friction and wear behavior of the DLC film sliding
against a Si3N4 ball was evaluated on a ball-on-disk test rig
equipped with an environmental chamber with which the
relative humidity and gaseous environment could be controlled.
The friction and wear tests were performed in air and nitrogen
environments, respectively, with the relative humidity ranging
from 5% to 100%, at a normal load of 2 N, a sliding velocity of
about 125 m/min, to a maximum sliding duration of 60 min.
The wear rate of the DLC film was calculated from the
wear track profiles measured on a surface profilometer. The
worn surfaces of the DLC film and the counterpart Si3N4 ball
were analyzed on a JSM-5600LV scanning electron micro-
scope (SEM) equipped with a KEVEX SIGMA energy
dispersive X-ray spectrometer (EDS). The chemical composi-
tions and chemical states of the worn surfaces were analyzed
on a PHI-5702 X-ray photoelectron spectroscope (XPS)
operating with monochromated Al-K irradiation at a pass
energy of 29.4 eV.
3. Results and discussion
3.1. Friction and wear behaviors
Fig. 1 shows friction behaviors of the DLC film at different
relative humidity in air and nitrogen environments. It is seen
that the friction coefficients continuously increase with
increasing relative humidity in both air (Fig. 1(a)) and nitrogen
(Fig. 1(b)) environments. Moreover, the friction coefficient in
nitrogen is always lower than that in air under the same relative
humidity (Fig. 1(c)). Particularly, in dry nitrogen (RH-5% and
RH-20%), the DLC film fails very quickly, though it records a
much low friction coefficient.
Fig. 2 shows the variation of the wear rates of the DLC film
with the relative humidity in air and nitrogen. It is seen that the
wear rate of the DLC film is closely dependent on the relative
humidity and environment. Namely, the wear rate of the DLC
film continuously increases with increasing relative humidity in
air, while it decreases sharply with increasing relative humidity
in nitrogen and reaches the minimum at a relative humidity of
40%. In the latter case, the friction coefficient assumes an
increase with further increase of the relative humidity above
40%. At a relative humidity below 40%, the DLC film has a
much higher wear rate in nitrogen than that in air, while above a
relative humidity of 40% it shows a smaller wear rate in
nitrogen than in air. In particular, the DLC film failed quickly in
dry nitrogen (relative humidity about 5%) and recorded a
specific wear rate of approximately 6.5×10-7
mm3
/N m, which
was larger than that in dry air (7.1×10-9
mm3
/N m) by nearly
two magnitude orders.
3.2. SEM analyses of the worn surfaces
In order to gain more insights into the friction and wear
mechanisms of the DLC film sliding against the Si3N4 ball
under different relative humidity in air and nitrogen environ-
ments, SEM and EDS analyses of the worn DLC film and the
counterpart Si3N4 surfaces were performed. Figs. 3 and 4 show
the SEM pictures of the worn surfaces of the DLC film and the
counterpart Si3N4 ball at different relative humidity in air and
nitrogen environments. It is seen that the worn surface of the
DLC film in air is very much smooth and shows only signs of
slight scuffing (Fig. 3(a)). The worn surface of the counterpart
Si3N4 ball in this case is covered by transferred materials, with
the transferred layer mainly composed of carbon to be compact
and homogeneous (Fig. 3(b)). Contrary to the above, the DLC
film sliding against the Si3N4 ball in dry nitrogen is
characterized by obvious flake-like peeling off (Fig. 3(c)),
though the counterpart Si3N4 ball is still covered by a
transferred carbon-rich layer in this case (Fig. 3(d)). The
accumulation of a large number of wear debris around the
contact region of the Si3N4 ball indicates that the DLC film
Table 1
Summary of the deposition conditions and some properties of the DLC film
Item Parameters
CH4 gas flow rate 11.6 sccm
Ar gas flow rate 11 sccm
Deposition pressure 5 Pa
RF power 300 W
DC bias 200 V
Deposition time 3 h
Thickness 640 nm
Hardness 15.2 GPa
Young's modulus 120 GPa
2 H. Li et al. / Diamond & Related Materials xx (2006) xxx­xxx
ARTICLE IN PRESS
experiences a mass-transfer to Si3N4 ball during the friction
process.
As the relative humidity in nitrogen increases to 40%, the
worn surface of the DLC film is smooth and shows many
granular wear debris scattered at the edge of the wear track (Fig.
4(a)). The corresponding contact area of the Si3N4 ball is
covered by a compact and dense transferred carbon layer aside
from the surrounding loose wear debris (Fig. 4(b)). However,
when the relative humidity in nitrogen increases to 100%,
obvious differences will be seen on the worn surfaces of the
DLC film and the counterpart Si3N4 as compared with those at a
relative humidity of 40% in nitrogen. As shown in Fig. 4(c), at a
relative humidity as high as 100% in nitrogen, a discontinuous
layer is formed in the central wear track of the DLC film, and
more wear debris is gathered on the wear track edge. The
corresponding worn surface of the Si3N4 ball is larger and much
rougher, indicating a severe wear of the ball (see Fig. 4(d)). The
similar changes of the worn surfaces of the DLC film and the
counterpart Si3N4 ball are also observed with increasing relative
humidity in air.
3.3. XPS analysis of the worn surface
Table 2 summarizes the changes in the concentrations (at.%)
of C, Ar, O, and Si, determined by XPS analysis (the XPS data
are estimated to have an error of 10%), on the original and worn
surfaces of the DLC film at varied testing conditions. The
presence of Ar in the film is attributed to the CH4 plus Ar as the
feedstock. The Ar concentration on the worn surface of the DLC
film decreases owing to the removal of the Ar atoms during the
friction process. The concentration of O on the worn film
surfaces tested at a relative humidity of 5%, 40% and 100%,
respectively, is 24.7 at.%, 27.7 at.%, and 28.3 at.%, which is
much larger than 5.3 at.%, that of the original DLC film.
Moreover, as the relatively humidity exceeds 40% in air, the
concentration of Si increases as well. Similar variation trends
are also observed for the O and Si concentrations on the worn
surfaces of the DLC film with increasing relative humidity in
nitrogen environment. It should be noted that in dry nitrogen
(relative humidity about 5%), the O concentration on the wear
track of the DLC film is 6.4 at.%, which is higher than that of
the original film (5.3 at.%) but much smaller than 24.7 at.%, that
on the wear track of the DLC film tested in dry air, due to the
protection action of nitrogen. Thus, it can be inferred that the
1
0 20 40 60 80 100
10
100
1000
Wearrate(10-9
mm3
N-1
m-1
)
Relative humidity (%)
In air
In nidrogen
Fig. 2. Wear rate of the DLC film as a function of relative humidity in air and
nitrogen environments.
0 10 20 30 40 50 60 70
0.00
0.05
0.10
0.15
0.20
a: RH-5%
b: RH-20%
c: RH-40%
d: RH-60%
e: RH-80%
f: RH-100%
c
d
e
f
b
a
Frictioncoefficient
Sliding time (min)
(a)
0 10 20 30 40 50 60 70
Sliding time (min)
0.00
0.04
0.08
0.12
0.16
0.20
a: RH-5%
b: RH-20%
c: RH-40%
d: RH-60%
e: RH-80%
f: RH-100%
f
e
d
c
ba
Frictioncoefficient()
(b)
0 20 40 60 80 100
0.00
0.02
0.04
0.06
0.08
0.10
0.12
Frictioncoefficient()
Relative humidity (%)
In air
In nitrogen
(c)
Fig. 1. Friction behaviors of the DLC film at different relative humidity: (a)
friction coefficient as a function of sliding time in air; (b) friction coefficient as a
function of sliding time in nitrogen; (c) average friction coefficient as a function
of relative humidity.
3H. Li et al. / Diamond & Related Materials xx (2006) xxx­xxx
ARTICLE IN PRESS
Fig. 4. SEM morphologies of the worn surfaces of the DLC film (a, c) and counterpart Si3N4 ball (b, d) in humid nitrogen: (a) and (b) 40%RH; (c) and (d) 100%RH.
Fig. 3. SEM morphologies of the worn surfaces of the DLC film (a, c) and counterpart Si3N4 ball (b, d) in dry air and nitrogen (relative humidity about 5%): (a) and (b)
in dry air; (c) and (d) in dry nitrogen.
4 H. Li et al. / Diamond & Related Materials xx (2006) xxx­xxx
ARTICLE IN PRESS
DLC film experiences structure change and tribochemical
reactions during the friction process.
Fig. 5 shows the XPS spectra of C1s and O1s on the original
surface of the DLC film. The C1s spectra can be fitted into three
components around 284.8 eV, 286.3 eV, and 288.7 eV,
respectively (Fig. 5(a)). The one around 284.8 eV corresponds
to C­C or C­H bonds, and that around 286.3 eV corresponds to
C­O bonds [25,26]. The third peak of a much smaller intensity
near 288.7 eV is assigned to CfO bonds [25,26]. At the same
time, the symmetrical O1s peak at 532.6 eV is attributed to C­O
bonds caused by the absorbed oxygen on the film surface due to
air exposure (Fig. 5(b)).
Figs. 6 and 7 show the XPS spectra of C1s on the wear tracks
of the DLC films tested at different relative humidity in air and
278 280 282 284 286 288 290 292 294
0.0
0.2
0.4
0.6
0.8 C1s
C=O
C-O
C-C
C-H
Inetnsity(a.u.)
Binding energy (eV)
Binding energy (eV)
(a)
524 526 528 530 532 534 536 538 540 542
0.0
2
4
6
8
0
2
4
C-O
O1s
Intensity(a.u.)
(b)
Fig. 5. XPS spectra of C1s (a) and O1s (b) on the surface of the as­deposited
DLC film.
0.0
0.2
0.4
0.6
0.8 C1s
CO2
C=O
C-O
C-C
C-H
(c)
Inetnsity(a.u.)
278 280 282 284 286 288 290 292 294
Binding energy (eV)
Inetnsity(a.u.)
(b)
0.2
0.4
0.6
0.8 C1s
CO2
C=O
C-O
C-C
C-H
278 280 282 284 286 288 290 292 294
Binding energy (eV)
C1s
C=O
C-O
C-C
C-H
Inetnsity(a.u.)
(a)
278 280 282 284 286 288 290 292 294
Binding energy (eV)
CO2
Fig. 6. XPS spectra of C1s on the wear tracks of the DLC film tested under
different relative humidity in air: (a) 5%RH; (b) 40%RH; (c) 100%RH.
Table 2
Changes of concentrations (at.%) of C, Ar, O, and Si on the original and worn
surfaces of the DLC film
Element As-
deposited
film
Worn surfaces of the DLC films tested in different
environments
In air (RH%) In nitrogen (RH%)
5% 40% 100% 5% 40% 100%
C 89.8 73.2 69.5 62.6 92.3 82.0 68.1
Ar 4.9 2.1 2.6 2.3 1.3 1.2 1.9
O 5.3 24.7 27.7 28.3 6.4 16.3 22.5
Si 0.0 0.0 0.2 6.8 0.0 0.5 7.5
5H. Li et al. / Diamond & Related Materials xx (2006) xxx­xxx
ARTICLE IN PRESS
nitrogen, respectively. It is seen that the C1s spectra of the wear
tracks tested at different relative humidity in air (Fig. 6) are
much different from that of the original film (Fig. 5). With the
increase of the relative humidity in air, the C1s core level
spectrum shifts to a higher binding energy, the intensity of C­C
(or C­H) peak decreases, while the intensities of C­O and CfO
peaks significantly increase. A new C1s peak near 290.1 eVon
the wear track is ascribed to the adsorbed CO2 groups [27].
Similar variation trends are also observed for the C1s states on
the wear tracks of the DLC films with increasing relative
humidity in nitrogen (see Fig. 7). It is interesting to note that the
XPS C1s states on the wear track of the DLC film tested in dry
nitrogen (relative humidity about 5%) are similar to that on the
original surface of the film. Moreover, the C­C (or C­H) bond
intensity on the wear tracks of the DLC films tested in nitrogen
is higher than that in air under the same relative humidity. This
could also be attributed to the protection action of nitrogen.
The XPS Si2p spectra on the wear tracks of the DLC film and
the counterpart Si3N4 ball tested at a relative humidity of 100%
in nitrogen are shown in Fig. 8. It is seen that the Si2p spectrum
on the wear track of the DLC film is centered at 103.4 eV (see
Fig. 8(a)), which is assigned to gel-type amorphous SiO2 [28].
The asymmetrically broad Si2p spectrum on the worn surface of
the Si3N4 ball can be fitted into two components around 101.8
eV and 103.4 eV, respectively (see Fig. 8(b)). The one around
101.8 eV corresponds to the Si­N bonds of the Si3N4 ball, and
that around 103.4 eV to the Si­O bonds of gel-type amorphous
SiO2 [28].
C1s
C-C
C-H
C-O
C=O
C1s
CO2
C=O
C-O
C-C
C-H
C1s
CO2
C=O
C-O
C-C
C-H
Inetnsity(a.u.)Inetnsity(a.u.)
(a)
(b)
0.
0.
(c)
278 280 282 284 286 288 290 292 294
Binding energy (eV)
Inetnsity(a.u.)
278 280 282 284 286 288 290 292 294
Binding energy (eV)
278 280 282 284 286 288 290 292 294
Binding energy (eV)
Fig. 7. XPS spectra of C1s on the wear tracks of the DLC film tested under
different relative humidity in nitrogen: (a) 5%RH; (b) 40%RH; (c) 100%RH.
0.0
0.2
0.4
0.6 SiO2(gel)Si2p
Intensity(a.u.)
94 96 98 100 102 104 106 108 110
-0.2
0.0
0.2
0.4
0.6
0.8
SiO2 (gel)
Si3
N4
Intensity(a.u.)
Binding energy (eV)
96 98 100 102 104 106 108 110
Binding energy (eV)
(a)
(b)
Fig. 8. XPS spectra of Si2p on the worn surfaces of the DLC film (a) and
counterpart Si3N4 ball (b) tested at a relative humidity of 100% in nitrogen.
6 H. Li et al. / Diamond & Related Materials xx (2006) xxx­xxx
ARTICLE IN PRESS
From the above-mentioned XPS analyses of the worn
surfaces of the DLC film and the counterpart ceramic ball, it
is deduced that two kinds of tribochemical reactions are
involved during the friction testing under a higher relative
humidity. One is the oxidation of the DLC film, and the other is
the oxidation and hydrolysis of the Si3N4 ball. As reported by
Kim et al. [20], the oxidation of the DLC film was expected to
be similar to that of hydrocarbon polymers described by Scott
[29]. Namely, the oxidation of the DLC film during the friction
process was divided into three steps. First, the C­H bonds on
the DLC film surface were mechanically broken by friction
force and the water molecules (and O2 molecules in air) were
chemisorbed to form the C­OH or C­O groups. Secondly, the
chemisorbed peroxides were transformed to COOH or remained
as active radicals such as COO­. Thirdly, further shearing of the
C­H bonds in the neighbors donated electrons to form CfO.
Therefore, the DLC film experienced a change in the surface
chemical states from C­H bonds to oxygen-containing groups
during the friction testing. On the basis of a simple model for
asperity temperature rise due to friction [30], Liu et al. had
proposed that the temperature on the local asperity contact could
reach about 1200 °C for the ceramic ball sliding against DLC
film [22]. The temperature was so high that the contact points of
the Si3N4 ball could react with water to form SiO2 by way of the
reaction below:
Si3N4  6H2O--3SiO2  4NH3:
The resultant SiO2 would further react with water to form
silicic acid [28]:
SiO2  2H2O­SiOH4:
The silicic acid would accumulate on the friction surface and
form a thin hydrated film, therefore to reduce friction and wear.
3.4. General discussion
Erdemir et al. said that the low friction coefficient (0.025­
0.093 in our work) of DLC films was mainly attributed to the
high chemical inertness of the films [17,23]. Namely, in a highly
hydrogenated plasma, constant hydrogen bombardment on the
DLC films during growth process increases the hydrogen
content of these films, and most of the hydrogen atoms are
expected to be paired with the covalent or free -bonds, leading
to the elimination of dangling bonds on the sliding DLC
surfaces. Furthermore, the intense hydrogen ion bombardment
also prevents cross-linking or CfC double bonding in growing
DLC films and etches out the graphitic phases. Thus, the
residual ­* interactions that can result from graphitic phases
are also minimized when a highly hydrogenated plasma is used.
Finally, some of the carbon atoms (at least those on the surface)
can be di-hydrated (i.e. two hydrogen atoms bonded to one
carbon atom). The presence of di-hydrated carbon atoms on
DLC surfaces is expected to provide better shielding or a higher
degree of chemical passivation. Subsequently, the dramatically
lowered friction coefficients are recorded for the DLC films,
owing to the elimination of the possibility of strong covalent
and ­* interactions at sliding DLC interfaces, plus better
shielding and higher chemical passivation by di-hydration [23].
However, the significant effect of the relative humidity on
the friction and wear behavior of the DLC film could never be
underestimated (see Figs. 1 and 2). In other words, water had a
specific effect on the formation of a DLC transferred layer on
the counterpart surface and on the tribochemical reactions of the
film itself and the counterpart ceramic as well (see Figs. 3­8).
As shown in Figs. 3 and 4, the formation of the transferred
layer on the Si3N4 ball is strongly influenced by H2O and O2. In
dry air, the worn surface of the Si3N4 ball is covered by a thick
and compact transferred carbon-rich layer (see Fig. 3(b)). This
transferred carbon-rich layer helps to prevent the direct contact
between the ceramic ball and the DLC film and acts as a
"lubricant" of a low-shear-strength at the interface to result in a
much small friction coefficient (0.03) [2,20,21]. With the
increase of the relative humidity, water vapor can be condensed
on the contact zone to prevent the formation of the transferred
carbon-rich layer on the Si3N4 ball [7] (see Fig. 4(d)) and hence
to result in a continuously direct contact between the DLC film
and the ceramic ball, thus higher friction coefficient and wear
rate of the DLC film are observed.
The oxidation of the DLC film in the presence of H2O and O2
during friction testing also has effect on the friction and wear
behavior of the DLC film. With the increase of the relative
humidity in air, the oxidation of the DLC film became
increasingly severe and the DLC film experienced a surface-
chemical-state change from C­H bonds to oxygen-containing
groups (see Figs. 6 and 7). Subsequently, during the friction
process, the bonding strength increased from about 0.08 eV per
bond for the Van der Waals bonding of hydrocarbons to about
0.21 eV per bond for the hydrogen bonding at C­O and CfO
sites in the presence of H2O molecules, which resulted in a
continuous increase in the friction coefficient and wear rate of
the DLC film [3,18].
The thin tribochemical film mainly composed of silica gel
and formed by way of the oxidation and hydrolysis of the Si3N4
ball due to friction-induced temperature-rise on the sliding
surface (see Fig. 8) has also an effect on the friction and wear
behavior of DLC film. On one hand, the tribochemical reactions
of the Si3N4 ball acted to prevent the formation of the
transferred carbon-rich layer thereon, which led to increased
friction coefficient of the DLC film and wear of the Si3N4 ball.
On the other hand, a thin hydrated film of silicic acid formed on
the sliding surfaces of both DLC film and ceramic ball could act
as a "lubricant" of a low shear strength and hence resulted in a
friction coefficient 0.090 at a higher relative humidity, which
was much smaller than 0.2, that reported elsewhere under a high
relative humidity [19,31].
In nitrogen environment, the friction coefficient is always
smaller than that in air under the same relative humidity. The
nitrogen molecules are expected to hinder the access of active
species like water and oxygen molecules to the counterface and
prevent or decelerate the friction-induced tribochemical oxida-
tion of the DLC films, thus to decrease the friction coefficient
[32,33]. The DLC film experienced mass-transfer of carbon to
the Si3N4 ball during the sliding in nitrogen (see Fig. 3(d)),
7H. Li et al. / Diamond & Related Materials xx (2006) xxx­xxx
ARTICLE IN PRESS
which caused a much larger wear rate and hence the early failure
of the DLC film. Yet it has not been known why the DLC film is
liable to mass-transfer carbon to the Si3N4 ball in a dry nitrogen
environment. As the relative humidity increases to an
intermediate value (about 40%) in nitrogen, water molecules
may passivate the freshly formed dangling bonds on the sliding
surface [14] and prevent the mass-transfer of carbon from the
DLC film to the Si3N4 ball. Therefore, the wear rate of the DLC
film decreases sharply to a minimum value and the wear track of
the film retains smooth after tested at a lower relative humidity
(RH40%) in nitrogen.
Andersson et al. [34] investigated the friction behavior of a
DLC film under varying water vapor pressure and pointed out
that water molecules adsorbed to the friction surface could
cause the dipole-like interaction, increasing the adhesive forces
and hence friction coefficient. In addition, more water would
initiate a greater viscous drag and capillary forces. This could
account for the linear increase of the friction coefficient with
increasing relative humidity in our work in some sense.
4. Conclusions
The friction and wear behavior of the DLC film deposited on
a Si (100) wafer, making use of PECVD technique with CH4
plus Ar as the feedstock, is highly dependent on the relative
humidity. The DLC film sliding against Si3N4 in a ball-on-disk
configuration registers continuously increased friction coeffi-
cient and wear rate with increasing relative humidity in air. It
records linearly increased friction coefficient with increasing
relative humidity in nitrogen, while its wear rate sharply
decreases and reaches the minimum at a relative humidity of
40%, followed by an increase with further increase of the
relative humidity above 40% in this case. A transferred carbon-
rich layer is formed on the counterpart Si3N4 ball surface, along
with the tribochemical reactions involving the frictional pair
materials and the environmental H2O and O2. The oxidation and
hydrolysis of the Si3N4 ball in higher relative humidity leads to
the formation of a tribochemical film mainly composed of silica
gel on the sliding surface, which, together with the transferred
carbon-rich layer and the friction-induced oxidation of the DLC
film at an increased relative humidity, accounts for the
environment dependence of the tribological behavior of the
DLC film and its variation therewith.
Acknowledgments
The authors are grateful to the National Natural Science
Foundation of China (grant nos. 50575217 and 50323007),
the "863" Program of China (no. 2003AA305670) and
the Innovative Group Foundation from NSFC (grant no.
50421502) for financial support.
References
[1] J. Robertson, Mater. Sci. Eng., R Rep. 37 (2002) 129.
[2] A. Grill, Diamond Relat. Mater. 8 (1999) 428.
[3] C. Donnet, Surf. Coat. Technol. 100­101 (1998) 180.
[4] M. Suzuki, T. Watanabe, A. Tanaka, Y. Koga, Diamond Relat. Mater. 12
(2003) 2061.
[5] H.X. Li, T. Xu, J.M. Chen, H.D. Zhou, H.W. Liu, Appl. Surf. Sci. 227
(2004) 364.
[6] H. Ronkainen, S. Varjus, K. Holmberg, Wear 222 (1998) 120.
[7] Y. Liu, A. Erdemir, E.I. Meletis, Surf. Coat. Technol. 94­95 (1997) 463.
[8] J. Andersson, R.A. Erck, A. Erdemir, Wear 254 (2003) 1070.
[9] W. Zhang, A. Tanaka, K. Wazumi, Y. Koga, Thin Solid Films 413 (2002)
104.
[10] E.S. Yoon, H. Kong, K.R. Lee, Wear 217 (1998) 262.
[11] T. Krumpiegl, H. Meerkemm, W. Fruth, C. Schaufler, G. Erkens, H.
Bohner, Surf. Coat. Technol. 120­121 (1999) 555.
[12] K. Enke, H. Dimigen, H. Hubsch, Appl. Phys. Lett. 36 (1980) 291.
[13] Y. Kokaku, M. Kitoh, J. Vac. Sci. Technol., A, Vac. Surf. Films 7 (1989)
2311.
[14] J. Jiang, S. Zhang, R.D. Arnell, Surf. Coat. Technol. 167 (2003) 221.
[15] C. Donnet, Surf. Coat. Technol. 80 (1996) 151.
[16] J.A. Heimberg, K.J. Wahl, I.L. Singer, A. Erdemir, Appl. Phys. Lett. 78
(2001) 2449.
[17] A. Erdemir, O.L. Eryilmaz, I.B. Nilufer, G.R. Febske, Diamond Relat.
Mater. 9 (2000) 632.
[18] C. Donnet, A. Grill, Surf. Coat. Technol. 94­95 (1997) 456.
[19] S.J. Park, J.K. Kim, K.R. Lee, D.H. Ko, Diamond Relat. Mater. 12 (2003)
1517.
[20] D.S. Kim, T.E. Fischer, B. Gallois, Surf. Coat. Technol. 49 (1991) 537.
[21] K. Jia, Y.Q. Li, T.E. Fischer, B. Gallois, J. Mater Res. 10 (1995) 1403.
[22] Y. Liu, A. Erdemir, E.I. Meletis, Surf. Coat. Technol. 82 (1996) 48.
[23] A. Erdemir, Surf. Coat. Technol. 146­147 (2001) 292.
[24] H.X. Li, J.M. T.Xu, H.D. Chen, H.W. Zhou, J. Phys., D, Appl. Phys. 36
(2003) 3183.
[25] T. Xu, S.R. Yang, J.J. Lu, Q.J. Xue, J.Q. Li, W.T. Guo, Y.N. Sun, Diamond
Relat. Mater. 10 (2001) 1441.
[26] F. Fusallba, N. EI Mehdi, L. Breau, D. Belanger, Chem. Mater. 11 (1999)
2743.
[27] Z.R. Yue, W. Jiang, L. Wang, S.D. Gardner, C.U. Pittman Jr., Carbon 37
(1999) 1785.
[28] Y.M. Gao, L. Fang, J.Y. Su, X.W. Xu, Tribol. Int. 30 (1997) 693.
[29] G. Scott, Polym. Eng. Sci. 24 (1984) 1007.
[30] E. Rabinowicz, Friction and Wear of Materials, Wiley, New York, 1965.
[31] A. Tanaka, T. Nishibori, M. Suzuki, K. Maekawa, Diamond Relat. Mater.
12 (2003) 2066.
[32] J.C. Sánchez-López, M. Belin, C. Donnet, C. Quirós, E. Elizalde, Surf.
Coat. Technol. 160 (2002) 138.
[33] C. Quirós, J. Gómez-García, F.J. Palomares, L. Soriano, E. Elizalde, J.M.
Sanz, Appl. Phys. Lett. 77 (2000) 803.
[34] J. Andersson, R.A. Erck, A. Erdemir, Surf. Coat. Technol. 163­164 (2003)
535.
8 H. Li et al. / Diamond & Related Materials xx (2006) xxx­xxx
ARTICLE IN PRESS