Effect of microstructure and thickness on the friction and wear

behavior of CrN coatings

C. Lorenzo-Martin
a,n
, O. Ajayi
a
, A. Erdemir
a
, G.R. Fenske
a
, R. Wei
b
a
Argonne National Laboratory, Tribology Section, Lemont, IL 60439, USA
b
Southwest Research Institute, San Antonio, TX 78228, USA
a r t i c l e i n f o
Article history:
Received 31 August 2012
Received in revised form
4 February 2013
Accepted 7 February 2013
Keywords:
Hard coatings
Thickness
Microstructure
Friction
Wear
a b s t r a c t
One of the most commonly used tribological thin-lm coatings is chromium nitride (CrN), typically
deposited by physical vapor deposition. Examples of current applications of this coating include cutting
and forming tools, dies and automotive components, such as injection valves and piston rings for diesel
engines. In selecting coatings for different tribological applications, one of the critical parameters is the
coating thickness. In the present work, the effects of microstructure and coating thickness on the
friction and wear behavior of CrN coatings were determined under unidirectional sliding conditions.
Tests were conducted with a dry ball-on-at contact conguration using 1-, 5-, and 10-mm thick
coatings deposited on a hardened H-13 steel substrate by plasma enhanced magnetron sputtering. The
ball specimen was made of WC. The friction behavior was observed to be strongly dependent on coating
thickness and microstructure, especially at relatively low loads (5 N). At higher loads, however, the
thinnest coating (1 mm) was quickly worn through, while the thicker ones (5 and 10 mm) remained
intact. Wear in both the counterface WC ball material and the coatings also depended on coating
thickness and microstructure. In all coatings, there was localized damage but minimal wear. Additional
tests were done with Si
3
N
4
and 52100 steel balls, and the results indicated different wear and friction
behavior from that for WC balls. The observed effect of coating thickness on tribological behavior is
attributed to differences in the microstructure and mechanical behaviors of CrN coatings as a function
of thickness.
& 2013 Published by Elsevier B.V.
1. Introduction
Thin lm coatings are being increasingly used for tribological
applications. Surface engineering is a fast growing area of
research because of the high industrial demands for friction
control and wear resistance, coupled with enabling technology
that produces new coatings with desirable tribological perfor-
mance as well as mechanical properties. Currently, many com-
mercial coatings are available for tribological applications.
Because of the variety of available coatings in terms of thickness,
hardness, and fracture toughness, among other properties, there
is usually a suitable coating for a given engineering application.
Thin lm coatings of different composition are among the most
common ones used for tribological purposes, because they require
minimal to no post-deposition processing. Some examples of
materials used for thin lm coatings are TiN, CrN, and diamond-
like carbon. Of these coatings, CrN is increasingly being used for
tribological applications, ranging from automotive components
[1,2] to forming dies [3,4]. In general, CrN coatings exhibit
superior ductility, fracture toughness, and corrosion and
oxidation resistance compared to the widely used TiN counter-
part. Also the lower coefcients of friction and higher wear
resistance under dry sliding conditions make CrN coatings excel-
lent candidates for a variety of applications, such as metal-
forming and die-casting tools [5].
CrN coatings are produced mainly by physical vapor deposi-
tion (PVD) techniques, including magnetron sputtering and arc
evaporation. The average CrN coating thickness is 15 mm, and
this material exhibits hardness values of HV 14.523.5 GPa and a
surface morphology consisting of dense, ne columnar structures.
One challenge for PVD coatings (including CrN) is related to the
thickness limitation. The thickness affects the residual stress state
in PVD coatings and also determines the stress eld distribution
under contact conditions, both of which are relevant to failure
pathways and mechanisms. Usually, the thicker the coating,
the larger the residual tensile stress accumulation and the worse
the adhesion to the substrate. There is often an optimal coating
thickness in terms of failure and performance. It is also known
that the PVD coating microstructure and morphology are depen-
dent on thickness [6]. In addition to deposition parameters, the
properties of coatings, including their tribological properties, will
be inuenced by the microstructure. Recently, a technique cap-
able of producing relatively thick CrN lms without the stress
problem has been developed. Plasma enhanced magnetron
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/wear
Wear
0043-1648/$ - see front matter & 2013 Published by Elsevier B.V.
http://dx.doi.org/10.1016/j.wear.2013.02.005
n
Corresponding author. Tel.: 1 630 252 8577; fax: 1 630 252 5568.
E-mail address: lorenzo-martin@anl.gov (C. Lorenzo-Martin).
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i
Wear ] (]]]]) ]]]]]]
sputter (PEMS) deposition is an improved version of conventional
magnetron sputtering. The PEMS technology can produce CrN
coatings as thick as 20 mm [7]. With the advent of PEMS, it is thus
possible to evaluate the impact of coating thickness on tribologi-
cal performance over a broad range of coating thicknesses. This
type of study was not possible prior to PEMS because the largest
thickness achievable, without stress problems, was typically less
than 3 mm.
In this study, we used a tungsten lament and a discharge power
supply to generate a global plasma in the entire vacuum system and
fabricated CrN coatings of different thickness (110 mm). The goal
was to assess the effect of CrN coating thickness and microstructure
on tribological performance.
2. Experimental details
2.1. Coating material
CrN coatings of different thicknesses were deposited on
polished H-13 steel at specimens with dimensions of
21.50.25 in.
3
and hardness of 58.5 Rc by the PEMS process.
The H-13 material was hardened by heat treatment through
quenching and tempering at 550 1C. Fig. 1 is a schematic of the
PEMS system developed at Southwest Research Institute [6]. Prior
to coating deposition, substrates were sputtered with Ar for
6090 min to remove residual contaminant on the surface.
A solid target of 170-mm diameter was employed as a source
of Chromium (Cr). The power was set at 1.5 kW. During the
deposition process, a mixture of argon (Ar) and nitrogen (N
2
)
gases was introduced, and the chamber pressure was maintained
between 0.33 and 0.47 Pa (2.53.510
3
Torr). The surface
temperature of the substrate was maintained at about
400 1C [8]. Because this temperature is lower than the tempering
temperature, no change in microstructure and properties is
expected for the substrate during the coating process. Coatings
were deposited to approximately 1-, 5-, and 10-mm thicknesses
with a corresponding surface nish of 0.011, 0.024, and 0.025 mm
Ra, respectively. The roughness of the original polished steel at
surface was about 0.005 mm. Coatings surface morphologies, as
well as cross sections, were examined by using optical and
scanning electron microscopy (SEM) (Quanta 400F ESEM) operat-
ing at 10 kV.
2.2. Hardness measurement
Coating surface hardness (H) and reduced Youngs modulus
(E
r
) were measured by nanoindentation method using a triboin-
denter TI-950 (Hysitron Inc.). A Berkovich diamond tip with an
approximate radius of curvature of 50 nm was used for the
hardness measurements. A xed displacement of 100 nm was
selected, and indents were performed at a loading/unloading rate
of 20 nm/s. The reduced Youngs modulus was obtained from the
slope of the unloading part of the load-displacement curve using
the Oliver and Pharr method, as described in Ref. [9]. Because the
roughness of the coating seemed to be too large to obtain
consistent measurements, indentations were placed on a polished
annular ring of the coating produced by a micro-abrasion ball
cratering method to measure the hardness and elastic modulus
accurately close to the coating surface (less than 0.25 mm from
the top surface). A minimum of six indents were considered for an
average of hardness and reduced modulus calculation.
The crystalline structure of the coatings was assessed by using
a Philips X Pert X-ray diffractometer. A Cu K
a
radiation source
(l0.1542 nm) was used, operating at 40 kV and 40 mA. X-ray
scans covered from 201 to 1001 at a step size of 0.011 and scan
speed of 0.1251/s. The total number of steps was 8000. Peaks were
identied by using the X Pert data viewer software.
2.3. Friction and wear testing
Friction and wear tests were conducted in unidirectional dry
sliding contact using a ball-on-at conguration (Fig. 2). The balls
had 1/2 in. diameter, were made of WC, and had a hardness value
of 16 GPa and a precision surface nishing grade of 25 (38 nm
measured surface roughness). According to manufacturer speci-
cations, the ball material consisted of 9094% tungsten and 610%
cobalt as a binder material. The steel substrate and CrN coatings
with three thicknesses were tested against the WC balls. For
comparison, 52100 steel and Si
3
N
4
balls were also tested.
Tests were conducted at normal loads of 5, 10, and 20 N. The
maximum Hertzian contact pressures calculated for steel at tested
against WC ball were (0.91, 1.15, and 1.44 GPa, respectively). For
Si
3
N
4
ball tested against steel at, the maximum Hertzian contact
pressures were ( 0.77, 0.96, and 1.22 GPa, respectively), while for the
52100 steel ball on steel at the maximum Hertzian contact
pressures were (0.68, 0.86, and 1.09 GPa, respectively). The effect of
coating on contact pressure will be dependent on coatings properties
(see Table 1). Compared to steel substrate, coatings with lower
elastic modulus than steel (such as CrN-3) will produce lower
contact pressures, while coatings with higher elastic modules than
steel (such as CrN-1 and CrN-2) will produce higher contact
pressures than steel substrate. The contribution of the coating to
contact pressure will be also dependent on load and coating
thickness. The higher the load and the thinner the coating, the more
important contribution of substrate to contact pressure. All tests
were run for 60 min and at a linear speed of 1 cm/s in ambient room
air (relative humidity of 65%). The friction coefcient was continu-
ously measured during the sliding. Wear on ats and counterface ball
materials was evaluated at the conclusion of the test by using optical
prolometry. We obtained 3D maps of the ball scars and measured
volume wear for each track. To calculate the wear volume of the
coatings, we measured the wear track proles across the sliding
direction of the track at four locations, also using the prolometry
Cr Target
S
a
m
p
l
e
s
To Pump
Filament
Global
Plasma
Ar, N2
Worktable
Magnetron
Power
Supply
-
AC
Discharge
Power
supply
+
Bias
Power
Supply
-
+
Magnetron-
generated
Plasma
Magnetron
Fig. 1. Schematic of plasma enhanced magnetron sputtering (PEMS).
C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 2
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i
technique. Wear rates were obtained by normalizing wear volume
by the total sliding distance. A repeat test was conducted for each
condition. Worn surfaces were also characterized by optical and
scanning electron microscopy equipped with energy dispersive X-ray
spectroscopy (EDS) analysis.
3. Results and discussion
3.1. Surface morphology
Fig. 3 shows the surface morphology for the three CrN coat-
ings. Average grains size (d), as determined by surface image
analysis, is dependent on coating thickness: d0.1 mm for CrN-1,
d0.40.5 mm for CrN-2, and d0.81.2 mm for CrN-3. Although
grain size distribution is mostly homogeneous for all the coatings,
the surface morphology (specically the aspect ratio) is depen-
dent on coating thickness.
For the thinnest coating, the grains are uniaxial, while the
thickest coatings possess a clear columnar structure with the
grains preferentially elongated perpendicular to the surface.
Cross-sectional micrographs of the coatings are shown in Fig. 4.
For CrN-1, the coating is mostly uniaxial. Over a Cr bond layer
of 400 nm, a columnar CrN structure that is tens of nanometers
thick and few hundreds of nanometers high is starting to grow.
For the thicker coatings, CrN-2 and CrN-3, the structure is clearly
columnar with cauliower-like texture on the surface (Fig. 5b).
The columns grow wider with thickness of the coatings, forming a
fan shape. The columns are slightly tilted 351 with respect to the
perpendicular direction of the substrate. In the thickest coating,
CrN-3, the columns grow from a width of 50 nm (close to the
substrate) to 500750 nm at the thickest area, giving a coarser
structural morphology. Also, surface localized defects (known as
macro particles or macro droplets) were observed for all coatings,
especially for thicker ones.
Fig. 5 shows such a typical nodular-like defect. The particle
defect diameter is in the 210 mm range. These types of defects
are produced during the deposition process, and because the
bonding of the macro particle with the coating is relatively poor,
they can be easily removed during the wear process, leaving
craters on the coating surface. The presence of macro particles
may be responsible for the accelerated initial wear.
3.2. Crystal structure of the coatings
X-ray diffraction (XRD) spectra for the coatings are presented
in Fig. 6. These spectra show how the relative peaks for Fe, Cr, and
CrN change with coating thickness. The uncoated steel (baseline)
shows a clear iron peak at 44.451. For the thinnest coating, CrN-1,
the most intense peak is located at 44.521. Two possible peaks
overlap at this position. One peak is Fe-a from the substrate;
considering the relative thin coating of less than a micron, the
effect of the substrate can be important.
The second possible overlapping peak is Cr (110), most probably
from an initial Cr bond layer during deposition. Another peak almost
equally intense for the CrN-1 coating is CrN (111). Several other
weaker peaks, corresponding to Cr and CrN, are also observed. As
the coating thickness is increased to 5 mm (CrN-2), the CrN (111)
peak dominates, followed by CrN (220) and several other weaker
peaks for CrN and Cr. With coating thickness of 10 mm (CrN-3), the
strongest peak is CrN (220), followed by CrN (200). The CrN (111)
peak is still present but is less intense than for the thinner coatings,
and a new peak corresponding to CrN (311) appears. In sum, the
XRD analysis showed that the three coatings consist of CrN crystals
primarily, regardless of coating thickness.
3.3. Nano-mechanical properties of the coatings
Table 1 summarizes the measured hardness and reduced
Youngs modulus obtained for the three CrN coatings, determined
using the nano-indentation method described earlier in experi-
mental details section. Other relevant properties of the coating
and substrate material, such as thickness and roughness, are also
included.
Note that CrN-1 and CrN-2 have very similar hardness
(21 GPa) while CrN-3 exhibits a considerably lower hardness
(14 GPa). The reason for differences in the measured hardness
values is unclear, perhaps a reection of differences in structural
morphologies and/or differences in residual stresses in the coatings.
3.4. Friction and wear results
The friction behavior of the three CrN coatings and uncoated
steel substrate tested under unidirectional dry sliding with the
WC ball is summarized in Fig. 7. For the uncoated H13 steel at,
the friction behavior is nearly identical at 5 and 10 N loads. In
both cases, the friction coefcient increased rapidly at the start of
the test to a maximum value of about 0.6, followed by a gradual
Fig. 2. Ball-on-at test rig for unidirectional sliding testing.
Table 1
Properties of coatings used in this work.
Coatings Deposition
method
Thickness
(mm)
Reduced elastic
modulus (GPa)
Hardness
(GPa)
Roughness
(nm)
CrN-1 PEMS 1 256 21 11
CrN-2 PEMS 5 269 21 25
CrN-3 PEMS 10 190 14 24
Steel none 0 212 8.4 5
C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 3
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i
decrease to a minimum value of about 0.4, after which a
continuous gradual increase occurred to the end of test. For the
rst 1000 s of sliding at 20 N (Fig. 7c), a similar trend was
observed: rapid increase, followed by both gradual sequential
decrease and increase. However, this trend was followed by a
sudden rapid decrease to a near constant value of 0.35 for the
duration of the test.
This friction behavior was observed in repeat tests for the WC and
H13 steel sliding pair. It is attributed to the extensive formation of a
transfer lm on the uncoated steel at, as shown in Fig. 8. Higher
loads produced more transfer lm. In the 20-N load test, the rapid
decrease in friction coefcient after 1000 s of testing is attributed to
large coverage of the contact area by the WC transfer lm, such that
in the later stage of testing, the contact interface consisted of the
sliding of the WC transfer lm on the WC, which is known to have
friction coefcient of about 0.350.45 [10]. There is also the possibility
of oxidation at the tribo contact interface resulting in the formation of
iron and tungsten oxides. Oxidation of wear debris trapped at the
contact interface can also occur. Indeed, tribo layers are well known
to consist of a complex mixture of both material pair in contact, as
well as species from the environment.
The friction coefcients in the tests with CrN-coated ats were
almost identical at 5 and 10 N loads. For the three coatings at
these loads, the friction coefcient showed a general gradual
increase with time. For both loads, the friction coefcient (m) in
CrN-1 was consistently higher than for the thicker coatings, with
a value of m0.7 at the test conclusion. The CrN-2 and CrN-3
coatings showed identical friction coefcients for the rst few
hundred seconds of testing, and by the end of test, the friction
coefcient with CrN-2 was about the same or slightly lower than
that for CrN-3 (m0.5 vs. 0.55).
At 20 N, the friction behavior in tests with CrN-2 and CrN-3
remained unchanged, but a sudden rapid increase in noise was
evident in the test with CrN-1 after 1500 s. This transition in CrN-
1 coincided with the coating being worn through, as shown in
Fig. 9a. It is possible that the sudden increase in friction was due
to the ball sliding against the Cr interlayer, although unlikely.
Rather, the friction increase is more likely due to the abrasive
Fig. 4. Cross-section SEM micrographs of CrN coating: (a) 1 mm, (b) 5 mm, and (c) 10 mm thickness.
Fig. 5. Characteristic surface defect (macro particles) for CrN coatings.
Fig. 3. Surface morphology of CrN coating of different thicknesses: (a) 1 mm, (b) 5 mm, and (c) 10 mm.
C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 4
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i
plowing action of the hard coating debris generated by the
coating failure.
Subsequent WC transfer occurred from the ball to the partially
exposed steel substrate (Fig. 9b), as indicated by EDS analysis
(Fig. 9c). This WC transfer may account for the decrease in friction
coefcient from 0.7 to 0.5 at the very end of the test, considering
that the new contact interface consists of WC ball material against a
at surface with CrN-steel-WC areas. For the thicker coatings (CrN-2
and CrN-3), the friction behavior and magnitude of coefcients are
similar to those in the lower load tests. Although no steady friction
coefcient value was achieved by the conclusion of testing for most
of the materials and loads tested, an average friction coefcient was
calculated for the duration of the test, excluding the rst few
minutes of testing. Average friction coefcient values are summar-
ized in Fig. 10.
The highest average friction coefcients are observed for the
thinnest coating (CrN-1), for which friction decreases with load,
exhibiting values ranging from 0.64 (for 5 N) to 0.58 (for 20 N). For
the thickest coatings (CrN-2 and -3), the average friction coefcients
values are similar to those of the uncoated steel baseline. While for
the uncoated steel at average friction values decrease with load too,
from 0.51 (for 5 N load) to 0.37 (for 20 N load), they do not for the
CrN-2 and CrN-3 coatings. Of the coatings CrN-2 exhibits the lowest
average values: 0.38, 0.48, and 0.44 for 5 N, 10 N, and 20 N,
respectively. For CrN-3 the values are slightly higher: 0.44, 0.51,
and 0.49 for 5 N, 10 N, and 20 N respectively. While the friction
coefcients for repeat tests were quiet close (less than 4% variability),
more variations were observed in the wear values (up to 10%).
Fig. 11 shows the wear results, as measured by optical prolo-
metry, in the ball-and-at specimens after testing. For the test with
CrN-1 at highest load (20 N), in which the WC ball counterface had
worn through the coating, the highest wear rates were measured in
the WC ball and at. With that exception, wear in both WC balls and
ats appears to increase nearly linearly with load for the other test
pairs. Wear in the uncoated steel at was, on average, at least an
order of magnitude higher than that for the coating material. Coating
thickness seems to have a minor effect on coating wear, with the
two thickest coatings wearing slightly more than the thinnest one at
low loads (Fig. 11a). The same tendency was observed for ball wear
(Fig. 11b). While coating thickness showed minimal effect on WC
ball wear, the material tested against it had a major effect on wear.
The WC ball wear against the uncoated steel at was, on average, at
least four times higher than ball wear against CrN coatings. Although
steel hardness is considerably lower than CrN coating hardness (see
Table 1), this higher wear is attributed to the presence of relatively
hard second-phase carbide particles, notably VC, MoC, and CrC,
sliding against the H-13 steel sample microstructure and surface.
Fig. 7. Friction evolution with time for unidirectional dry sliding at different loads: (a) 5 N, (b) 10 N, and (c) 20 N.
Fig. 8. Steel at tested against WC ball in dry unidirectional sliding: (a) overall track, (b) transfer layer from ball, and (c) EDS analysis of transfer layer.
20 30 40 50 60 70 80 90 100
-50
0
50
100
150
200
250
300
350
2 Theta
I
n
t
e
n
s
i
t
y

(
c
o
u
n
t
s
)
H-13
CrN-1
CrN-2
CrN-3
Fig. 6. XRD peaks of CrN coatings deposited on H-13 steel substrate.
C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 5
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i
Wear on the uncoated steel at was predominantly abrasive wear.
The transfer layer of WC from the ball onto the surface protected it
partially against further wear (as shown in Fig. 8). In terms of wear
mechanisms in the CrN coatings, polishing wear and some transfer
from the ball material are the dominant mechanisms. The exception
is CrN-1 tested at high load (20 N). As the coating was worn through,
severe abrasive wear became the dominant mechanism (as seen in
Fig. 9b). By contrast, CrN-2 showed mostly polishing wear with
minimal transfer lm formation regardless of the test load. Some
local damage, consisting of cracking and delamination of the transfer
surface layer, was observed (Fig. 12).
For CrN-3, load seemed to affect the degree of damage on the
wear track. For low load (5 N), there is minimal damage to the
original surface morphology. Only mild polishing and some smooth-
ing of the wear track by minimal transfer are apparent. At higher load
(20 N), the amount of transfer is increased considerably, and exten-
sive patches of WC transfer are observed (Fig. 13b). This transfer
layer, which has a smoothing effect on the track, eventually grows to
a critical thickness and fails by cracking and chipping (Fig. 13c).
3.5. Effect of counterface material
Additional friction and wear tests were conducted with the CrN-2
coated at sliding against polished, commercially available bearing
balls composed of either hardened 52100 steel (Ra36 nm) or Si
3
N
4
(Ra10 nm). The hardness of the 52100 balls is about 7.2 GPa
(62 Rc), and that of the Si
3
N
4
balls is about 14.5 GPa. Tests with these
balls were conducted by the same procedure as used with WC balls.
Fig. 14 shows the friction variation with time during the test with
the three balls (WC, 52100 steel, and Si
3
N
4
) when sliding against
CrN-2 at 10 and 20 N loads. For WC, the test started at relatively low
friction coefcient of 0.2, but increased gradually for the 1-hour
duration of the test, ending with a nal value of about 0.5. The rate of
friction increase decreased with sliding distance or time. For this
material, the friction coefcient appears to be independent of load, as
the tests at 10 and 20 Nshowed nearly identical friction behavior. The
presence of the cobalt binder phase in the ball may have contributed
to the friction coefcient and noise reduction. For the 52100 steel
balls, sliding started with a friction coefcient of about 0.8 at both 10
and 20 N load. This was followed by a slight gradual decrease to a
steady value of about 0.70 for 20 N and about 0.75 for 10 N. The
friction coefcient data showed considerably more noise in the tests
with the steel ball compared to the WC ball. The high friction and
noise in the test data for the steel ball are attributed to extensive
metal transfer into the coated at surface, as illustrated in Fig. 15.
The tests with Si
3
N
4
balls started with a friction coefcient of
about 0.25 at both 10 and 20 N. In both cases, the friction coefcient
increased rapidly during the rst 5 min (300 s) of sliding, reaching a
nearly constant value of 0.7 for the remainder of the one-hour test.
Furthermore, the friction coefcient data increased in noise through-
out the test. This friction behavior is the result of both transfer of
material from the Si
3
N
4
ball into the coating and the wear and
damage of the coating, as shown in Fig. 16.
Fig. 17 compares the wear for the three different balls with CrN-2
coated ats. Wear was the greatest for the steel ball and the least for
the WC ball (Fig. 17a). Although more wear occurred in the Si
3
N
4
ball compared to the WC ball, it was substantially lower than the
wear in the steel ball. This behavior is perhaps a reection of
differences in ball hardness. The steel ball produced no wear in
terms of material removal from the coated at, while the Si
3
N
4
ball
produced the most wear on the CrN coated at (Fig. 17b). The wear
produced by the WC ball is signicantly less than that for the Si
3
N
4
ball due to the more extensive formation of a transfer layer, even
though the WC ball is harder than the Si
3
N
4
ball. Note that the
friction coefcients in the test with the WC ball are also much lower
than those in the test with the Si
3
N
4
ball. This condition will reduce
the shear stresses imposed on the coating and, hence, the amount
of wear.
4. Summary
Although CrN coatings with different thicknesses can be
produced by PEMS, they have signicant differences in terms of
microstructure and surface morphology:
For the 1-mm thick coating the grain structure is uniaxial. The
crystalline structure is mostly CrN with possible small
amounts of Cr. In terms of morphology, this coating surface
is relatively smooth.
Fig. 9. CrN-1 coating tested against WC ball in unidirectional sliding: (a) overall track, (b) transfer layer from ball and coating damage, and (c) EDS analysis of coating
damage and transfer layer.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
H-13 CrN-1 CrN-2 CrN-3
5N
10N
20N
A
v
e
r
a
g
e

F
r
i
c
t
i
o
n

C
o
e
f
f
i
c
i
e
n
t
Fig. 10. Average friction coefcient for CrN coatings and uncoated substrate
sliding against WC ball.
C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 6
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i
For the 5-mm thick coating, the grain shape distribution is
bimodal. There are uniaxial grains close to the substrate and
columnar-shaped grains in the upper region of the coating.
The crystal structure is mainly CrN. Surface morphology is
coarser with small cauliower texture.
The 10-mm thick coating is mostly columnar with fan shape
(wider columns as the coating thickens). Also, the crystal
structure is mainly CrN, and its surface morphology is the
coarsest with cauliower texture.
These differences have a signicant impact on the coating
properties and tribological behavior. For example, high hardness in
the CrN-1 and CrN-2 coatings is most probably due to the grain
morphology. The CrN-3 has lower hardness due to grain morphology
and/or residual stresses. In unidirectional sliding against the WC ball,
friction in the 1-mmcoating is higher, while it is similar for the 5- and
10-mm coatings. Since CrN-1 is the only coating with signicantly
different microstructure, its higher friction coefcient may be due to
its different microstructure, including higher level of Cr content.
Usually, differences in surface morphologies result in differences in
Fig. 12. CrN-2 surface after testing in dry sliding against WC ball: (a) overall track, (b) mild polishing, and (c) localized damage.
Fig. 13. CrN-3 surface after testing in dry sliding against WC ball: (a) mild polishing at low load, (b) extensive transfer at high load, and (c) localized damage in
transfer layer.
1.3
0.27
0.26
0.51
2
0.32 0.45
0.81
4.4
6.5
0.51
1.1
0
1
2
3
4
5
6
7
8
Steel CrN-1 CrN-2 CrN-3
5N
10N
20N
3.3
0.9
0.2
1.7
15.4
1.5 1.0 3.5
69.9
134.0
0.8
4.3
0
20
40
60
80
100
120
140
Steel CrN-1 CrN-2 CrN-3
5N
10N
20N
Flat Wear x10
3
m
3
/m Ball Wear x10
5
m
3
Fig. 11. Flat-and-ball wear for CrN coatings and uncoated substrate at different loads.
Fig. 14. Friction behavior of CrN-2 coating when sliding against different ball
materials at 10 and 20 N.
C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 7
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i
interactions between coating and ball surface asperities. This effect
may result in differences in the amount of material transfer and,
hence, in friction and wear behavior. The CrN-2 and CrN-3 samples
(thicker coatings) exhibited more local surface defects (macro-parti-
cles/macro-droplets). This condition resulted in local surface damage
in these coatings during sliding contact. There are other possible
differences in the structure and microstructure of the three coatings
in this study that may account for some of our observations. More
comprehensive comparative structural analysis by multiple techni-
ques and tools are needed to further elucidate the observations
reported in the present paper.
In general, thicker coatings showed better friction and wear
behavior, while the thinnest coating was easily worn through at high
load, resulting in a substantial increase in friction and wear. Com-
pared with the uncoated steel surface, the coatings have minimal
effect on friction but signicant effect on wear reduction. Indeed,
lower friction was observed in tests with uncoated steel under some
conditions after a good transfer layer had formed. This transfer layer
on steel at consists of a signicant amount of tungsten. Thus, it is
possible that lower friction in uncoated steel is due to the formation
of the W-rich transfer layer [11].
When sliding against different counterface materials, the
CrN-2 coating showed signicant differences in friction and wear
behavior. Although more studies are needed to further elucidate
the role of counterface materials, this preliminary study indicates
that CrN coatings should be paired carefully with counterface
materials. The current paper suggests that thicker coatings can be
deposited for tribological applications, but more work is needed
to optimize coating thickness for different applications.
Acknowledgments
This work was supported by U.S. Department of Energy,
Energy Efciency and Renewable Energy, Ofce of Vehicle Tech-
nologies, under contract DE-AC0206CH11357. The electron
microscopy was accomplished at the EMC at Argonne National
Laboratory, a U.S. Department of Energy Ofce of Science
Laboratory operated under Contract No. DE-AC0206CH11357
by UChicago Argonne, LLC.
Fig. 15. (a) Transfer from steel ball to CrN-2 coating during dry sliding; (b) EDS analysis of transfer layer.
Fig. 16. (a) Transfer from Si3N4 ball to CrN-2 coating during dry sliding; (b) EDS analysis of transfer layer.
0.0
1.0
2.0
3.0
4.0
5.0
WC Si3N4 Steel
W
e
a
r

V
o
l
u
m
e

(
1
0
5

m
3
)
W
e
a
r

V
o
l
u
m
e

(
1
0
5

m
3
)
10N
20N
356.6
0
100
200
300
400
W C Si3N4 Steel
10N
20N
Fig. 17. Comparison of ball and at wear for tests at 10 and 20 N: (a) CrN-2 coating wear against different balls and (b) wear of different balls against CrN-2 at.
C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 8
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i
References
[1] H. Scheerer, H. Hoche, E. Broszeit, C. Berger, Surface and Coatings Technology
142144 (2001) 1017.
[2] S.C. Tung, H. Gao, Wear 255 (712) (2003) 1276.
[3] J. Vetter, R. Knaup, H. Dweletzki, E. Schneider, S. Vogler, Surface and Coatings
Technology 8687 (1996) 739.
[4] B. Navinsek, P. Panjan, Surface and Coatings Technology 7475 (1995) 919.
[5] P. Panjan, M. Cekada, R. Kirn, M. Sokovic, Surface and Coatings Technology
180 (2004) 561.
[6] J.A. Thorton, Annual Review of Materials Science 7 (1997) 239.
[7] Ronghua Wei, Edward Langa, Christopher Rincon, James H. Arps, Surface and
Coatings Technology 201 (2006) 44534459.
[8] Feng Cai, Xiao Huang, Qi Yang, Ronghua Wei, Doug Nagy, Surface and
Coatings Technology 205 (2010) 182188.
[9] W.C. Oliver, G.M. Pharr, J. Mater. Res. 7 (1992) 1564.
[10] K. Bonny, P. De Baets, J. Vleugels, S. Huang, B. Lauwers, Tribology Transac-
tions 52 (2009) 481491.
[11] H. Engqvist, H. Hogberg, G.A. Botton, S. Ederyd, N. Axen, Wear 239 (2000)
219228.
C. Lorenzo-Martin et al. / Wear ] (]]]]) ]]]]]] 9
Please cite this article as: C. Lorenzo-Martin, et al., Effect of microstructure and thickness on the friction and wear behavior of
CrN coatings, Wear (2013), http://dx.doi.org/10.1016/j.wear.2013.02.005i