Surface & Coatings Technology 260 (2014) 290302

Contents lists available at ScienceDirect

Surface & Coatings Technology

journal homepage: www.elsevier.com/locate/surfcoat

Tribological behavior and machining performance of non-hydrogenated

diamond-like carbon coating tested against Ti6Al4V: Effect of surface
passivation by ethanol
S. Bhowmick, A. Banerji, A.T. Alpas
Mechanical Automotive and Materials Engineering Department, University of Windsor, Windsor, ON N9B3P4, Canada

a r t i c l e

i n f o

Available online 15 July 2014

Keywords:
Diamond-like carbon
Running-in friction
Surface passivation mechanism
Ethanol blended lubricant
Machining performance
Ti alloys

a b s t r a c t
The role of an ethanol environment on the tribological behavior of non-hydrogenated DLC (NH-DLC) coatings sliding against Ti6Al4V was investigated. Pin-on-disk tests were performed under different test atmospheres consisting of dry argon (0% RH) and ambient air (40% RH) and also while the samples were
submerged in water (H2O) and ethanol (C2H5OH). The running-in friction of the NH-DLC coatings under
each condition was studied and the results were used to rationalize the drilling behavior of Ti6Al4V
by NH-DLC coated tools. The highest running-in coefcient of friction (R = 0.55) was observed under a
dry argon atmosphere. In ambient air, the R was 0.42, while a reduction in R to 0.21 was recorded
when the tests were carried out in H2O. The observed reductions in R were attributed to OH passivation
of the NH-DLC surfaces. Ethanol was effective in providing OH passivation, as evident from the lowest R
of 0.15. Consistent with the friction measurements, a 45% reduction in drilling torque was obtained
when 10% ethanol was blended with the metal removal uid (MRF) consisting of wateroil (1:3) mixture
compared to an ethanol free MRF.
2014 Published by Elsevier B.V.

1. Introduction
Diamond-like carbon (DLC) coatings have consistently displayed
low coefcient of friction, COF or , values when tested against technologically signicant lightweight aluminum, magnesium and titanium alloys. Several studies that used counterfaces coated with DLC have
shown that DLC coatings minimized the adhesion and restricted material transfer from the surfaces of these alloys. The DLC coatings provide
signicant opportunities for expanding the applications of lightweight
materials by improving their performance in components operating
under sliding contact conditions (e.g. powertrain elements in combustion engines). The use of DLC coated tools and dies was reported to improve the surface quality of aluminum and magnesium sheets or
castings during machining and shaping operations [18]. A shortcoming
of DLC coatings is the fact that their tribological properties are highly
sensitive to the environmental conditions as their friction, wear, adhesion behavior may change drastically with small changes in the composition of the gas or liquid atmospheres present in the tribosystem. The

Corresponding author. Tel.: +1 519 253 3000x2602; fax: +1 519 973 7085.
E-mail address: aalpas@uwindsor.ca (A.T. Alpas).

http://dx.doi.org/10.1016/j.surfcoat.2014.06.066
0257-8972/ 2014 Published by Elsevier B.V.

friction behavior of the hydrogenated (a-C:H or H-DLC) and nonhydrogenated DLC (a-C or NH-DLC) coatings was found to exhibit almost opposite trends under vacuum and inert (i.e. N2 or Ar) atmospheres. Among the studies conducted on ferrous materials, Erdemir
[9] showed that H-DLC coatings tested in dry N2 atmosphere against
H-DLC coated H13 grade steel disks generated a coefcient of friction
(COF), of 0.005, while the NH-DLC produced a of 0.75. The very
low value in the case of H-DLC was attributed to the passivation of
-carbon bonds at the contact surface by the hydrogen atoms. For aluminum alloys, Konca et al. [10] reported a high of 0.52 for NH-DLC
against an Al-6.5% Si alloy, under vacuum (6.65 104 Pa). The same
NH-DLC when tested in an ambient atmosphere (47% RH) produced a
low of 0.16. The of NH-DLC tested against Al alloys was shown to decrease with increasing the moisture in the surrounding environment [5,
13]. The decrease in of the NH-DLC coating in moist air was attributed
to the passivation of the dangling carbon bonds by the OH groups dissociated from the water vapor in the atmosphere during the tribological
contact as described in [10]. First principle atomistic study [11] provided
quantitative support for this mechanism, by showing that the dissociation of water to OH molecules at the interface is possible and the static
work of separation between the OH\C (111)/OH\C(111) atoms is
considerably lower than Al/non-passivated coating surface thus
resulting in lower magnitudes of in the case of OH passivated surfaces.
An ab initio study [12] based on density-functional theory (DFT) and

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

generalized gradient approximations (GGA) of two (2 1)-C(001)

surfaces terminated by hydroxyl groups showed that repulsive interactions between them were greater than termination due to H
groups and thus could account for the low friction experimentally
observed in [10].
Another requirement for the attainment of low observed in the
case of NH-DLC is the formation of a carbonaceous transfer layer on
the counterface [13]. Surface analysis studies [7,14] suggested that the
carbonaceous transfer layers were possibly hydrated, and hence contributed to low friction. These transfer layers form early in the sliding
process during the running in period where the peak running-in coefcient of friction (R) is high followed by a low steady state coefcient of
friction (S). Notable reductions in COF have been observed in case of
NH-DLC (0.140.18) sliding against Mg and Ti alloys, in addition to Al
alloys, for tests conducted in air with 2985% RH [5,1517] again with
a high S and a low and stable R.
Parallel to the tribological studies summarized above, Bhowmick
and Alpas [18] examined cutting performances of NH-DLC coated HSS
drills using an aqueous minimum quantity of lubrication method
(H2O-MQL). During H2O-MQL drilling of Al-6.5% Si with NH-DLC coated
drills, using water sprayed at a rate of 30 ml/h, an average drilling
torque of 1.65 N-m was measured, which represented a signicant reduction compared to dry drilling that generated a torque of 4.11 N-m.
Machining of magnesium alloy AZ91 using NH-DLC coated drills under
similar (H2O-MQL) conditions was also examined [19]. The application
of H2O-MQL reduced the drilling torque to 1.46 N-m compared to the
dry drilling using NH-DLC (12.73 N-m). The OH passivation of NH-DLC
was cited as the possible reason for the reduced torque values. It was
found that the range of average drilling torque values generated during
H2O-MQL drilling using NH-DLC coated tools (0.981.05 N-m) was lower
than that observed in the case of ooded drilling (1.932.42 N-m) of Mg
using conventional tools.
In summary, the above review indicates that surface passivation of
NH-DLC by OH has a positive effect on the friction performance of
these coatings. There is also some evidence that OH enrichment of the
lubricants used in tribological tests may lead to the friction reduction.
Kano et al. [20] reported very low of 0.03 for NH-DLC sliding against
itself for tests lubricated with ester based lubricant glycerol monooleate (GMO). Using time-of-ight secondary ion mass spectroscopy
(ToF-SIMS) formation of an OH-terminated carbon surface was detected. The low was discussed in terms of Van der Waals interactions between the two carbon surfaces passivated by OH groups originating
from the GMO lubricant. Matta et al. [21] performed similar sliding experiments in which NH-DLC sliding against itself was lubricated with
glycerol in the presence of hydrogen peroxide resulting in a value of
0.03. X-ray photoelectron spectroscopy (XPS) and ToF-SIMS observations supported the earlier view that carbon surfaces terminated by
OH would lead to low values of friction.
The objective of this paper is to investigate the effect of surface passivation mechanism on the tribological behavior of NH-DLC coatings.
Sliding experiments are designed to measure friction and wear under
controlled atmospheres that provided potential different degrees of
passivation. Ti6Al4V pins were used as counterfaces against NH-DLC
coated M2 steel disks in such a way that the Ti6Al4V/NH-DLC contact
interface was submerged either in water or in ethanol Tests were also
performed in dry argon (0% RH) and ambient air (40% RH) atmospheres,
i.e., under the conditions that are known to create high values. Surface
characterization methods were used to investigate the compositions of
the contact surfaces and transfer layers with particular attention to the
detection of OH passivation effect. The surface tribochemistry was studied using Fourier transform infrared spectroscopy (FTIR) and microRaman spectroscopy. Low friction results obtained from tribological
tests during NH-DLC sliding against Ti6Al4V alloy in ethanol suggested that the use of the NH-DLC in combination with metal removal
uid (MRF) blended with ethanol would facilitate machining of this
alloy.

291

2. Experimental details
2.1. Material properties
NH-DLC coatings were deposited on AISI M2 grade tool steel disks
and WCCo drill using an unbalanced magnetron sputtering system
equipped with one chromium and two graphite targets. A 0.10 m Cr interlayer was deposited on 25.10 mm diameter M2 disks to promote
coating adhesion followed by the deposition of 1.50 m thick NH-DLC
coating on top. The hydrogen content of the NH-DLC coating measured
using elastic recoil detection (ERD) was b 2.00 at.%. The hardness and
the elastic modulus of the coatings were calculated from the loading
unloading curves obtained by a Berkovich type nano-indenter that penetrated to a maximum depth of 200 nm below the surface. Average
hardness and elastic modulus values were 13.00 1.10 GPa and
158.55 6.82 GPa. The NH-DLC coated coupons were tested against
Ti6Al4V pins with a rounded tip of 4.01 mm in diameter. Ti6Al4V
had the following composition (in wt.%): 5.506.75% Al, 0.08%
C, 0.4% Fe, 0.03% N, 0.2% O, 0.02% H, 3.504.50% V and the
balance Ti. The microhardness of the Ti6Al4V was 380.00 3.01 HV.
2.2. Pin-on-disk tests
Pin-on-disk type wear tests were performed on NH-DLC coated coupons using a CSM tribometer. Tests were done in four different environments consisting of (i) dry Ar atmosphere with 0% RH, (ii) ambient air
with 40% RH (iii) distilled water and (iv) anhydrous (99.9%) ethanol.
During the last two tests the NH-DLC coated M2 steel disks were
submerged in reservoirs lled with uid. Sliding tests were performed
at a linear sliding speed of 0.12 m/s and under the application of 5.00
N normal load (Hertzian contact pressure = 1095.5 MPa) for 3.00
103 cycles where the peak running in coefcient of friction (R), number
of revolutions of the running-in (tR) and the steady state coefcient of
friction (S) were recorded. Each test was repeated at least 3 times.
Following the sliding tests, the Ti6Al4V pin contact surfaces were
examined by a scanning electron microscope (SEM) and their compositions were analyzed using energy dispersive spectroscopy (EDS), Fourier transform infrared spectroscopy (FTIR) and micro-Raman analyzer.
The energy dispersive X-ray spectroscopy (EDS) spectra were recorded
using an FEI Quanta 200 FEG SEM equipped with an EDAX SiLi detector
spectrometer. FTIR analyses were conducted using a Thermoelectron
Nicolet 760 spectroscope in reectance mode at two different spots
using a 100 m 100 m aperture. The Raman spectra of the transfer
layers were obtained using a 50 mW Nd-YAG solid state laser
(532.00 nm excitation line) through the 50 objective lens of a Horiba
Raman micro-spectrometer.
2.3. Drilling tests
Drilling tests were conducted on the Ti6Al4V alloy using NH-DLC
coated and uncoated WCCo drills with 6.35 0.01 mm diameter. The
drilling tests were performed at a cutting speed of 2500 rpm (50 m/
min) using a feed rate of 0.25 mm/rev. The time interval between consecutive holes was 5 s. The ow rate used for supplying MRF was
30,000 ml/h and the temperature of the MRF at the inlet was 18 C. A
total of 100 holes were drilled with a horizontal center to center spacing
of 10 mm between them. Each hole was 19 mm deep. The temperature
measurements presented in the paper are taken from the last hole. The
tests were conducted under the following conditions: (i) dry: drilling
without using MRF; (ii) ooded: drilling using a water based MRF,
consisting of a mixture of water and oil blended at a ratio of 1 to 3 and
(iii) ooded with 10 vol.% ethanol: drilling using the same water/oil
(1:3) based MRF as in (ii) but with an addition of 10 vol.% ethanol. Exploratory experiments that used different volume percentages of ethanol mixtures showed that MRF blended with 10% or greater
percentage of ethanol by volume showed no difference in the torque

292

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

response whereas the average torque increased by 20% for ethanol contents lower than 10%. Consequently, a 10% ethanol blend was used in
the drilling experiments with MRF.
The cutting torque was measured individually for each drilled hole
using a magneto-static sensor as described in [22]. In order to monitor
the drill tip temperature variations, a contact thermocouple was
inserted inside to the mid-point of the 100th hole drilled. The details
of temperature measurements are given in [8].

3. Results
3.1. Changes in tribological behavior of NH-DLC tested at different environmental conditions
The variations in values of the NH-DLC coatings tested against Ti
6Al4V, under dry argon (Ar) atmosphere with 0% RH, and ambient
atmosphere with 40% RH (ambient air), are shown in Fig. 1 as a function
of number of revolutions. Fig. 1 also shows the plots obtained while
the NH-DLC sample and the counterface were immersed in water and
ethanol. According to the friction plots in Fig. 1, testing the NH-DLC
using a dry Ar atmosphere produced a high R of 0.55. Following this
maximum initial R, the friction plot showed a decrease with an average
value of 0.40 0.06 for the next 4.00 102 cycles and decreased again
to an average of 0.37 0.02 occurred until 19.00 102 cycles and a
steeper decrease to 0.27 0.05 with large uctuations until the test
was stopped at 30.00 102 cycles. The uctuations in the friction plot
were always high and no S was observed in this and the two other
tests that were done under the same conditions. Under the ambient
atmosphere (with 40% RH) a maximum R of 0.42 was recorded during
the running-in period which lasted for 1.50 102 revolutions followed
by a decrease to a S of 0.11 0.01. The decrease in R and S is in agreement with the previously suggested surface passivation mechanism of
carbon bonds by H and OH molecules dissociated from the moisture
during sliding contact [10,23]. Both R and S further decreased for
tests conducted while the NH-DLC sample was immersed in water;
resulting in a low R peak of 0.22 followed by a low S of 0.08 0.02.
A signicant difference between these tests that showed a distinct running-in period and those conducted while NH-DLC was submerged in
ethanol was that in ethanol the running-in period was almost nonexistent. It should be noted that wear tests in ethanol and water were
conducted under the boundary lubrication regime conditions as determined from the ratio () of minimum lubricant thickness (hmin) and
the root mean square (r.m.s.) roughness (r*) of the surfaces in contact

Maximum running-in COF= 0.55 (dry Argon)

Maximum running-in COF= 0.42 (ambient air)

Dry Argon

Maximum running-in COF= 0.22 (water)

Maximum running-in COF= 0.14 (ethanol)

Ambient
Ethanol

[24]. The r* value in the case of Ti-6Al-4V and NH-DLC at the start of
the wear tests was 0.575 m, which increased to 0.720 m at the end
of tests in ethanol and to 1.170 m at the end of tests in water. These
changes in the r.m.s. roughness of the surfaces in contact, as a result of
sliding wear, led to lower magnitudes of (=hmin/r*) as compared to
the beginning of the tests. The value of at 25 C was 0.0030 in
the case of tests in ethanol and 0.0027 for tests in water at the beginning of the tests. They decreased slightly to = 0.0023 in ethanol
and = 0.0013 in water at the end of the tests. The parameters
used to calculate for tests in water and ethanol are summarized
in Table 1. As the b 1, boundary lubrication conditions were satised throughout the tests. The running-in COF of NH-DLC in ethanol
(R = 0.14) was the lowest value of R for the all test conditions considered. The average S was low at 0.13 but higher than other testing
environments that produced a steady state friction regime. The salient observations, namely the reduction in both the running-in
and steady state COF in water and the elimination of R in ethanol,
are discussed in terms of passivation of surface carbon atoms and
formation of transfer layers on the counterface in Section 4.

3.2. Detailed examination of the running-in COF

According to Fig. 1, tR and R varied with the test environment. This
section focuses on the detailed characterization of both of these parameters because the contact conditions of the running-in period simulate
the contact conditions during machining, more appropriately than the
steady state conditions (as fresh Ti6Al4V surfaces come in contact
with the coated tool surface continuously). Results of three sets of pinon-disk tests, conducted in order to quantify the running in period parameters under each testing environment, are shown in Fig. 2 and are
also summarized in Table 2.
The COF reached a maximum value and then decreased to a lower
steady state value within the rst 50 cycles for ambient air atmosphere,
30 cycles for tests in water and 10 cycles for tests in ethanol. In the case
of tests with dry Ar atmosphere, a decrease in COF was observed after
1.80 102 cycles but no steady state regime was found as explained
in Section 3.1. Presented in Fig. 2 are the results of pin-on-disk
tests conducted on fresh Ti6Al4V pins with the purpose of determining the average tR and R values. The standard deviation values
presented here reect the variations in average R values obtained
from the three tests performed in each test environment. The R, typically, for dry Ar was 0.55 0.02. The R decreased to 0.42 0.01
when the Ti6Al4V was tested against NH-DLC in an ambient atmosphere within tR = 50.0 revolutions. The R decreased by 50% to 0.21
0.01 for tests in water compared to the tests conducted in ambient condition, within tR = 30-40 revolutions. The lowest R (0.15 0.01) was
for tests with NH-DLC immersed in ethanol and assumed a steady
friction in tR = 10 revolutions.
The wear tracks formed on the NH-DLC surfaces were examined
using SEM. A typical secondary electron image (SEI) shown in
Fig. 3(a), after tR = 185 cycles, indicated that considerable amount of

Table 1
Parameters used to calculate the lubrication conditions ( values) for tests in water and
ethanol.
R (m)

U (m/s)

W (N)

E* (GPa)

0 (Pas)

r* (m)

2.29 108

0.57

0.01
(Ethanol)
Water

Fig. 1. Variations of COF with the number of revolutions for NH-DLC coating tested against
Ti6Al4V under different environmental conditions.

0.002

0.12

5.0

74.1
0.00089
(water)

R represents the radius of the Ti6Al4V pin in m, U represents the sliding velocity in m/s,
W represents the normal load in N, E* represents the composite elastic modulus in GPa, 0
and are the viscosity constants of water and ethanol, and r* is the r.m.s. surface roughness of the contacting surfaces before (0.57 m) and after the test (0.72 m).

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

293

(a)

(b)

(c)

(d)

Fig. 2. Running-in COF ( R) and duration of the running in (tR) values of the NH-DLC coating sliding against Ti -6Al -4V in (a) dry Ar, (b) ambient air, (c) water and (d) ethanol. The tR and R
calculations have been shown in the rst column. Three friction curves are shown for each test environment each curve represents a different test on a new sample.

Table 2
Running-in coefcient of friction, duration and wear rate values and steady state COF of
NH-DLC coatings tested against Ti6Al4V at different environmental conditions.
Test
conditions

Test Duration of
no. running-in
(number of
revolutions, tR)

Highest
running-in
COF (R)

Steady state
COF (S)

Running-in
wear rate
(mm3/Nm)

Dry Ar

1
2
3
1
2
3
1
2
3
1
2
3

0.55
0.57
0.54
0.42
0.42
0.43
0.22
0.21
0.19
0.15
0.14
0.16

0.39
0.40
0.37
0.14
0.12
0.11
0.07
0.06
0.07
0.12
0.11
0.12

5.73
5.74
5.71
2.85
2.86
2.88
1.59
1.62
1.64
0.75
0.75
0.76

Ambient
(40% RH)
Water

Ethanol

185
180
180
50
50
50
30
40
30
10
10
10

0.023
0.031
0.042
0.002
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.002

105
105
105
105
105
105
105
105
105
105
105
105

material transfer occurred from Ti6Al4V surface to the NH-DLC

surface when tested in dry Ar atmosphere. In fact the section of
wear track shown in Fig. 3(a) was entirely covered by the adhered
material transferred from Ti6Al4V surface whereas other sections
of the wear track exhibited lesser amounts of adhesion. Quantitative
metallography was used to determine the amount of material transfer to the NH-DLC coating's wear track (by examining three different
wear tracks under each condition). The details of the image analysis
and measurement techniques are given in [7]. It was found that 56%
of the total contact area was covered with the transferred material.
On the other hand, the area of the coating contact surface covered
by the transferred material was small (5%) for tests conducted
under the ambient atmospheric condition (Fig. 3(b)). Consistent
with the decrease in R , a notable reduction in the amount of
material transfer (b 0.5%) occurred for tests conducted with NH-DLC
immersed in water (Fig. 3(c)). For the tests performed with NH-DLC
immersed in ethanol material transfer to the NH-DLC that could
be detected was also very small and estimated to be about 0.1%
(Fig. 3(d)).

294

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

Dry Ar

(a)

Wear Track

(b)

Ambient

Wear Track

Adhered Ti

Adhered Ti

Water

100 m

(c)

100 m
Ethanol

No adhesion

(d)

Adhered Ti

Wear Track
Wear Track

100 m

100 m

Fig. 3. Typical secondary electron images of wear tracks formed on the NH-DLC coating surfaces when tested against Ti6Al4V after the running-in period for test in
(a) dry Ar (t R = 185 revolutions), (b) ambient air (t R = 50 revolutions), (c) water (t R = 30 revolutions) and (d) ethanol (t R = 10 revolutions).

The wear rates of the NH-DLC for the running in period, WR, were
measured using a non-contact surface prolometry technique and
have been summarized in Table 2. In addition to Ti6Al4V adhesion,
damage to NH-DLC surface occurred by surface scratching, namely by
formation of longitudinal scratches along the wear track. Volume of
coating material removal was measured by taking an average of 24
readings from six different regions of the wear track. The details of the
volumetric wear loss measurements can be found in [25,26]. After the
running in period the wear rates decreased to very low values at the
steady state wear regime. Thus the cumulative wear, almost equal to
WR, was determined at the end of the running in period under each condition. In dry Ar atmosphere, the wear loss was 5.73 105 mm3/Nm
and decreased to 2.86 105 mm3/Nm in ambient atmosphere. The
measured wear rates were 1.62 10 5mm3/Nm for tests in water
and 0.75 10 5 mm3/Nm for tests in ethanol resulting in 45% and
75% reduction in wear rates respectively compared to the tests in
ambient air with 40% RH.
The Ti6Al4V pin contact surfaces were examined for wear damage
and also for possible occurrence of transfer layers. Tests in dry Ar atmosphere produced a wear track of diameter of 220 2 m and showed
scratch marks along the sliding direction. There is also evidence for material transfer from the coating surface with a dark gray appearance and
debris accumulation along the periphery of the worn area as shown in
Fig. 4(a). The EDS maps in Fig. 4 indicate that the darker gray area that
covered a part of worn Ti-6Al-4V contact surface was rich in C
(Fig. 4(b, c)) that should have been transferred from the DLC while it
was in sliding contact with the pin. Oxygen map is given as a reference
but no oxygen was detected as expected as the sliding tests were conducted in an inert atmosphere (Fig. 4(d)). In ambient atmosphere, the
dark gray transfer layers covered greater area percentage of the contact

surface of Ti6Al4V, compared to tests in dry Ar, and are shown in

Fig. 4(e). The contact surface area was smaller and consisted of the
alloying elements (Ti shown hereFig. 4(f)) in addition to the transfer
layer comprising of C and O according to elemental EDS maps shown
in Fig. 4(gh). It is important to note that the elemental C map in
Fig. 4(g) suggests that almost all the contact surface was covered with
a carbonaceous layer. The results of surface analyses conducted on samples tested in liquid environments namely water (Fig. 5(aj)) and especially ethanol (Fig. 5(eh)) suggest that much smaller wear damage was
inicted to the surface contact surface of Ti6Al4V compared to those
in Fig. 4. Fig. 5(a) shows the wear marks on the Ti6Al4V surface and a
portion near the middle that was covered with the transfer layer
(darker). The EDS map shows the presence of C and O (Fig. 5(c, d)) in
the transfer layersimilar to the transfer layer formed in the case of
tests in ambient air. The diameter of wear scar formed on the Ti6Al
4V pin was very small, 27 5 m, in ethanol environment (Fig. 5(e)).
There was no noticeable carbon transfer layer formation on the pin contact surface Fig. 5(f, g) although some C was detected outside the wear
scar. On the pin contact surface an increase in concentration of O was
detected (Fig. 5(h)).
3.3. Characterization of transfer layers by FTIR and micro-Raman analyses
FTIR analyses conducted on the transferred material generated on
the Ti6Al4V tested against NH-DLC showed that the layers had different chemical compositions depending on the test atmosphere
(Fig. 6(ac)). In dry Ar atmosphere, the transfer layer exhibited a peak
at 848 cm1 corresponding to the C\H group that may be attributed
to the carbonaceous material transferred from the NH-DLC coating. According to the FTIR spectra shown in Fig. 6(b), the material transferred

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

(a)

Dry Ar

295

Ti

(b)

(c)

(d)

Ti

(f)

(g)

(h)

Transfer layer

100 m

Ambient

(e)

Transfer layer

100 m

Fig. 4. (a) Secondary electron image of the Ti6Al4V pin surface taken after sliding against NH-DLC coating in dry Ar atmosphere. The elemental EDS maps taken from the whole area
shown in (a) are for (b) Ti, (c) C and (d) O. (e) Secondary electron image of Ti6Al4V pin surface after the sliding test against NH-DLC coating in ambient air. The elemental EDS
maps taken from the whole area shown in (e) are for (f) Ti, (g) C and (h) O.

to Ti6Al4V during tests conducted in water consisted of hydrocarbon

compounds; where bands attributed to C\O groups can be observed
between 900 and 1000 cm 1 while C\H groups were identied at
1454 cm 1 and 2925 cm 1. A broad band observed at 3458 cm 1
could be attributed to hydroxyl (OH) groups [2729].
The material transferred to Ti6Al4V during the tests conducted
under an ethanol environment showed a doublet that appeared between 900 and 1000 cm1 attributed to C\O groups that are expected
to form as a result of OH adsorption (Fig. 6(c)). The intensity of these
peaks was higher than those observed in water (Fig. 6(b)). C\H groups
were identied at 1610 cm1 and 2925 cm1 [2729]. Similar to the
transfer layer formed in water environment, a broad band was observed
at 3458 cm1 and attributed to the hydroxyl (OH) groups. Consequently, the FTIR analyses suggested that water and ethanol would promote
passivation of carbon atoms at the surface leading to a reduced R and
also a reduced tR as shown in Fig. 2. Ethanol, however, was more effective passivating agent, as indicated by the prominent C\O and OH
peaks (Fig. 6(c)), derived from the OH adsorption, as compared to
tests in water.

The micro-Raman spectra that were obtained from both the unworn
NH-DLC coating surface and the transfer layers that were formed on the
top of pin surfaces under different testing conditions are presented in
Fig. 7. The Raman spectra of the unworn NH-DLC coating showed a
broad peak between 1250 and 1650 cm1which is a common feature
of amorphous NH-DLC coatings [30]. On the other hand, the Raman
spectra of the transfer layers that were generated on the Ti6Al4V
pin surface showed a noticeable shift of the peak values in Fig. 7.
Deconvolution of the Raman spectra into two Lorentzian peaks indicated that the peaks centered at 1388 cm1 and 1583 cm 1 can be
assigned to the D and G bands of the carbon, respectively. The distinct
D and G bands observed for the transfer layers formed on the Ti6Al
4V pins under different test conditions as opposed to the single broad
peak observed in the case of the NH-DLC coating indicated an increase
in sp2 content. This may suggest sliding induced graphitization during
sliding as previously discussed in the literature [3133]. In addition to
the D and G bands, distinct peaks were also observed at 480, 1090,
1200, 1460 cm1 and 2950 cm 1 for transfer layers formed during
the tests in ethanol, which may be attributed to the adsorbed ethanol

296

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

Fig. 5. (a) Secondary electron image of the Ti6Al4V pin surface taken after sliding against NH-DLC coating in water. The elemental EDS maps taken from the whole area shown in (a) are
for (b) Ti, (c) C and (d) O. (e) Secondary electron image of Ti6Al4V pin surface after the sliding test against NH-DLC coating in ethanol. The elemental EDS maps taken from the whole
area shown in (e) are for (f) Ti, (g) C and (h) O.

molecules [34], responsible for passivating the transfer layers formed on

the Ti6Al4V surface reiterating the observations previously made
from FTIR observations (Fig. 6(c)).
In summary, the results presented in Section 3.3 show that ethanol can act as an effective passivating agent for the coating as well as
the transfer layer, to reduce the magnitude and duration of runningin COF. The results of these tribological experiments underline the
important role that ethanol plays in surface passivation when NHDLC is placed in sliding contact against Ti6Al4V. Hence the use of
ethanol as a lubricant in machining of the same alloy may present
an opportunity to reduce the cutting forces. The drilling experiments
described in the next section were performed to test this hypothesis.
Results of drilling experiments performed on Ti6Al4V, under
ooded conditions using an oilwater based conventional MRF and
ethanol blended with the oilwater based MRF, are presented in
Section 3.4.

3.4. Results of drilling tests

Fig. 8 shows the torque proles generated when Ti6Al4V workpiece was drilled with and without using NH-DLC coated tools as a function of time. Torques generated under dry (nominally ambient air with
40% RH), ooded using an MRF consisting of water/oil (1:3), and
ooded using ethanol containing ethanol blended with water/oil (1:3)
based MRF, were measured and plotted as shown in Fig. 8(a). When
ooded MRF drilling was used, occasional spikes were observed
reaching as high as 5.00 N-m compared to 13.00 N-m in the case of
dry drilling. The occurrence of spikes is related to the material transfer
to the tool surface and under dry drilling environment, signicant adhesion occurred immediately after the start of the test leading to short tool
life. When drilling was conducted under ethanol blended MRF environment, the average torque was low (1.60 0.02 N-m) and relatively stable and spikes were sparsely observed, indicating that this was the most

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

(a)

297

Transfer layer formed in dry Ar

Transfer layer formed in ambient

Transfer layer formed in water

G
D

(b)

D
C-C-O

Transfer layer formed in ethanol

CH3

C-H
CH3

NH-DLC

Fig. 7. Micro-Raman spectra of the transfer layer formed on Ti6Al4V pins tested against
NH-DLC coating in dry Ar, ambient air, water, ethanol and as-deposited NH-DLC coating.

(c)

Fig. 6. FTIR spectra of the material transfer red on the Ti6Al4V pin wear track tested
against NH-DLC in (a) dry Ar, (b) water, and (c) ethanol. The CO2 gas peak at 2350 cm1
observed in this and other FTIR spectra was due to contamination from the surrounding
atmosphere [14,25].

suitable drilling environment among the ones considered. In the case of

drilling with uncoated WCCo all three environments resulted in generation of a large number of spikes superimposed to high average torque
as shown in Fig. 8(b). Dry drilling resulted in torque spikes at almost
every hole. Note that the tool life however is longer in this case compared to NH-DLC coated drill, a point that will be explored further
later. Under MRF ooded drilling environment, large number of spikes
were also observed, although with a lower magnitude. Frequent spikes
were generated again for tests conducted using ethanol blended MRF.
Thus the use of NH-DLC coating clearly provided an advantage over

the uncoated WCCo drills. The addition of ethanol to the MRF resulted
in a favorable drilling condition, an observation that mimics the results
of the tribological tests.
Fig. 9(a) shows the average torque values calculated from Fig. 8(a,
b). The NH-DLC coated drills showed a low and stable torque response
when the drilling was conducted in ooded conditions using water/oil
(1:3) based MRF. When this MRF was mixed with 10% ethanol, a 50% reduction in the average torque was observed. The uncoated WCCo drill
showed a prolonged tool life compared to the NH-DLC coated drill
under dry conditions; as shown in Fig. 9(b) the uncoated drill did
not fail during the course of tests that involved drilling 100 holes
(Fig. 9(a)). The average torque however was high. The addition of 10%
ethanol to the MRF initially decreased the drilling torque but no difference between the two ooded conditions (with or without ethanol)
was apparent after drilling the 45th hole (Fig. 9(b)). This result indicates
that ethanol addition to the MRF was not an advantageous for ooded
drilling using uncoated WCCo tools contrary to the use of NH-DLC
coated drills under the same ooded condition (Fig. 9(a)).
A sharp increase in the torque value, that manifested itself as a spike
in drilling torque vs. time curve, is usually accompanied by the rise in
local contact temperature causing high friction and adhesion [8].
Fig. 10(a) shows the temperature variations in the hole depth during
drilling of the Ti6Al4V using NH-DLC coated drills under each
machining environment. As shown in Fig. 10(a), during dry drilling of
Ti6Al4V using the NH-DLC coated drills the temperature increased
to 566 C. The drills, as reported earlier, failed at the rst hole. High temperatures generated during dry drilling cause softening of Ti6Al4V
promoting adhesion to the tool and consequently result in high drilling
torque. The NH-DLC coating was removed during frictional contact and
it was the WCCo substrate that came in contact with Ti6Al4V that
resulted in high temperatures and high adhesion (see end of

298

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

Dry

NH-DLC

WC-Co

Dry
Spike
Spike

Spike

Flooded

Spike

Flooded

Spike
Spike

Flooded with Ethanol

Flooded with Ethanol

Spike

Spike

(a)

(b)

Fig. 8. Torque proles generated during drilling in Ti6Al4V in dry, ooded and ooded with ethanol conditions using (a) NH-DLC and (b) uncoated WCCo.

(a)
NH-DLC

Dry

Flooded

Flooded + ethanol

(b)
Uncoated WC-Co
Dry

Flooded + ethanol
Flooded

Fig. 9. Variations of average torque with the number of holes using (a) NH-DLC and
(b) uncoated WCCo coated drills against Ti6Al4V under different drilling conditions.

Section 3.4). During the ooded drilling tests using MRF coolant, the
drilling-induced temperature did not exceed 280 C. When the
ooded drilling tests were conducted using MRF blended with
ethanol (10%) the highest temperature recorded was 102 C. Consequently lower friction and adhesion would be expected compared to
other conditions.
The temperature variations during drilling of Ti6Al4V using
uncoated WCCo drill are shown in Fig. 10(b). The temperature increase during dry drilling reached about 553 C whereas the maximum temperatures were 437 C and 388 C during drilling in ooded
conditions without and with ethanol addition. These are higher compared to those when NH-DLC coated tools were used under the same
conditions.
The results presented in Figs. 810 indicate that an MRF blended
with ethanol decreased the temperature and the average torque. It is
thus expected that there will be less adhesion during drilling of Ti
6Al4V using NH-DLC coated drills using the ethanol (10 vol.%)
blended with the MRF. These trends are reected in the SEM micrographs of the cutting edges shown in Fig. 11.
Adhesion of Ti6Al4V was observed on the NH-DLC coated WCCo
drills used in dry condition (Fig. 11(b)). The adhered titanium was as
thick as 410500 m at the cutting edge of NH-DLC coated drills
forming a built-up edge (BUE) generating high drilling torques
(Fig. 9(b)). The use of a water/oil (1:3) based MRF reduced the extent of Ti6Al4V softening by lowering the drilling induced temperature and consequently less material transfer to the cutting edge
occurred (Fig. 11(c)). The 50% reduction in the torque value and
near-complete mitigation of adhesion due to 10% ethanol blended
with water/oil (1:3) based MRF (Fig. 11(d)) may be explained on the
basis of effective passivating ability of the ethanol compared to water,
as will be explained in Section 4.
The BUE was as thick as 350460 m at the cutting edge of the uncoated WCCo drills (Fig. 11(e)). In the case of ooded and ooded
with ethanol conditions, using the uncoated WCCo drill led to low

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

(a)

299

4. Discussion: surface passivation mechanisms in water and

ethanol environments

NH-DLC
Dry

Flooded

Flooded+ ethanol

(b)
Uncoated WC-Co
Dry
Flooded

Flooded+ ethanol

Fig. 10. Variations of drilling temperature with the depth of holes using (a) NH-DLC coated
drills and (b) uncoated WCCo against Ti6Al4V under different drilling conditions.

drilling temperature and resulted in a thinner BUE (240 m for

ooded and 220 m for ooded with ethanol) compared to dry drilling as evident from Fig. 11(f, g). The use of ooded and ooded
with ethanol could not eliminate the transfer of Ti6Al4V to the
uncoated tool drilling surface unlike the NH-DLC/ethanol MRF
condition.
The key features observed during the drilling experiments on Ti
6Al4V using uncoated WCCo drills, are high average torques (Fig. 9)
and high drilling temperatures (Fig. 10) causing considerable adhesion
(Fig. 11) of the alloy to the uncoated drill surface. To conrm this
trend and for better understanding of friction reduction mechanisms, tribological tests were performed on uncoated WCCo against
Ti6Al4V alloy in an ambient air, MRF and ethanol environments.
The tests in MRF showed the lowest R (0.22 0.005) and tR = 33
revolutions compared to the ambient air (R = 0.34 0.060, tR =
90 revolutions) and ethanol ( R = 0.28 0.005, tR = 50 revolutions). It can be recalled that the NH-DLC coating and ethanol combination performed better ( R =0.14 0.002, tR = 10 revolutions)
compared to these conditions.
In summary the MRF blended with ethanol decreased the temperature, average torque and adhesion signicantly while using
NH-DLC coated drills whereas no such signicant improvements
could be observed for tests with uncoated WCCo drills. The passivation of NH-DLC surfaces by OH groups, that was thought to be responsible for the observed low magnitudes of torque and
temperatures, consistent with low COF values in tribological tests
will be discussed in Section 4.

The tribological behavior of the NH-DLC coatings in vacuum and

inert gas environments is determined by the interactions of the dangling bonds of the carbon atoms on the NH-DLC coating surface with
the atoms on the counterface material [3537]. As shown in Fig. 2, a
high running-in COF of 0.55 was observed under a dry Ar environment.
The degree of the interactions between the coating and the counterface
material can be expected to increase on increasing the chemical afnity
of the counterface material towards C atoms in the NH-DLC coating. Ti
has high driving force for compound formation with C. The standard
Gibbs free energy of formation (Gf) for TiC is 180.0 kJ/mol at 300
K [38]. It can be suggested that the high chemical afnity of Ti towards
C would promote adhesion of Ti to the NH-DLC surface resulting in
high COF values and high wear rates in Ar and other inert atmospheres.
Adsorption and dissociation processes of H2O on NH-DLC surfaces and
formation of C\OH and C\H bonds have been previously described
[35,39,40]. Here the decrease in COF in ambient air with 40% RH and
in water can be explained by OH passivation of surface C atoms. The
mechanism of OH passivation is supported by the rst principles calculations of energies required for water dissociation and passivation at a
diamond surface [11]. The passivation led to signicant reduction in
the adhesive interactions between NH-DLC coating surfaces and the
Ti6Al4V. Thus, the running-in COF dropped from 0.55, observed in
dry Ar atmosphere, to values ranging between 0.22 and 0.42 in water
and ambient air with 40% RH. It should be noted that a carbonaceous
layer was formed on Ti6Al4V alloy's surface during the running in
stage of sliding. As such C atoms on both sides of the interface became
passivatedan effect that appears to be intensied with the increase
in OH supply to sliding contact area. The presence of OH groups on
the contact surfaces was conrmed by the FTIR and Raman analyses.
FTIR spectra shown in Fig. 6(b) and (c) indicated that in the presence
of water and ethanol, prominent OH and C\H peaks were generated.
Although the OH passivation mechanism appears to be similar in
water and in ethanol, a signicant difference is the reduction in the R
0.15 0.01) and tR (10 cycles) when the tests were run in ethanol.
This implies that the ethanol molecules may provide more effective passivation of the coating surface compared to water. The appearance of
prominent micro-Raman peaks (Fig. 7) corresponding to CH3 stretching
and bending observed for the tests conducted in an ethanol environment may provide support to this suggestion. In addition, the FTIR spectra of the transfer layers generated from tests in ethanol indicated a
peak at 3458 cm 1 (Fig. 6(c)), corresponding to OH molecules, that
was of a higher intensity than the OH peak observed in Fig. 6(b) for
tests performed in water.
The dissociation of a methanol molecule and the subsequent generation of OH groups have been previously studied in [41] where the authors performed tight-binding quantum chemical molecular dynamics
(TB-QCMD) studies on a model that considered a methanol molecule
sandwiched between two H-DLC surfaces. Results showed that the
methanol molecule dissociated and produced OH groups, which terminated the surface carbon atoms. Accordingly, an ethanol molecule may
be assumed to dissociate into ethyl and OH groups. A schematic representation of the dissociation and adsorption mechanisms of water
(H2O) molecules and ethanol (C2H5OH) molecules to generate OH molecules is given in Fig. 12. According to the bond dissociation energy calculations [4244] the energy required to break the C\O bond is 381
8 kJ/mol at 298 K, whereas a higher energy of 485.8 19 kJ/mol is
needed to break an O\H bond. Thus, from thermodynamics point of
view generation of an OH as a result of dissociation of a C2H5OH (by
C\O bond cleavage) would be more favorable compared to H2O dissociation (via O\H bond cleavage). Consequently the OH molecules
would be more readily produced at the sliding interface in the presence
of an ethanol atmosphere compared to water. This semi-qualitative argument needs to be further developed but higher surface passivation

300

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

(a)

(b)

(e)

(c)

(f)

(d)

(g)

Fig. 11. (a) Schematic representation of the location of SEM observations. Typical SEM secondary electron images of the drill cutting edges for NH-DLC under (b) dry, (c) ooded and
(d) ooded with ethanol conditions and for uncoated WCCo drills under (e) dry, (f) ooded and (g) ooded with ethanol conditions.

efciency in the presence of C2H5OH environment may account for the

observed reduction in the running-in COF during the sliding tests performed while the NH-DLC was immersed in ethanol.
It is also pertinent to comment on the role of transfer layers. In
the case of tests performed in water, formation of a carbonaceous transfer layer was observed (Fig. 5(a)) on the Ti6Al4V surface. The formation of this transfer layer and its subsequent passivation led to low and
stable S as observed in Fig. 1. On the other hand according to Fig. 5(e),
no distinct carbonaceous transfer layer formation was observed in the
case of tests with ethanolleading to slightly higher S (Fig. 1). It can
be suggested that the, OH molecules, (Fig. 12), passivated the coating
surface leading to a low R. During the drilling process, fresh Ti6Al

4V surfaces came in contact with the NH-DLC coated drillnot allowing

the formation of a stable transfer layer. Ethanol blended MRF provided
the most effective passivation of the NH-DLC coated drill surface considerably reducing the average torque and adhesion of Ti6Al4V to the
drill surface.

5. Conclusions
Pin-on-disk tests were performed on NH-DLC coatings sliding
against Ti6Al4V pins, where the Ti6Al4V/NH-DLC contact surfaces
were submerged in water and ethanol liquid media. Sliding wear tests

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

301

Fig. 12. Schematic representation of dissociation of water molecule and ethanol molecule leading to formation of OH molecules that are adsorbed by the NH-DLC coating surface. The bond
dissociation energy values represent the ease of OH bond cleavage from the parent molecule.

were also performed in dry argon (0% RH) and ambient air (40% RH)
atmospheres, to study the surface passivation mechanisms.
The main conclusions that arise from this work are as follows:
1. NH-DLC tested in dry Ar resulted in the highest running-in COF
(R) of 0.55 0.02. The R values reduced for tests in ambient
air to R = 0.42 0.01 and R = 0.21 0.01 for tests in water. Experiments performed in ethanol showed the lowest R = 0.15
0.01 lasting for the shortest duration of 10 sliding cycles (tR).
2. The low R in ethanol was attributed to passivation of surface carbon
atoms by OH molecules. The lower bond dissociation energy of the
C\O bond in ethanol as compared to the O\H bond in water was
suggested as a possible reason for effective passivation.
3. The passivation effect of ethanol was useful for machining of the Ti
6Al4V. Using NH-DLC coated drills in the presence of a lubricant
blend containing 10% ethanol in a conventional MRF (wateroil mixture), a 45% reduction in drilling torque occurred and eliminated
almost all material transfer from Ti6Al4V to the drill's cutting edge.
Conict of interest
There is no conict of interest.
Acknowledgments
Financial support for this work was provided by Natural Sciences
and Engineering Research Council of Canada (NSERC) through a Strategic
Network (CANRIMT) Grant.
References
[1]
[2]
[3]
[4]

J. Robertson, Mater. Sci. Eng. R. Rep. 37 (4) (2002) 129281.

F.G. Sen, Y. Qi, A.T. Alpas, Acta Mater. 59 (2011) 26012614.
F.G. Sen, Y. Qi, A.T. Alpas, J. Mater. Res. 24 (2009) 24612470.
T. Yokota, T. Sawa, M. Yokouchi, K. Tozawa, M. Anzai, T. Aizawa, Precis. Eng. 38
(2014) 365370.

[5] E. Konca, Y.T. Cheng, A.T. Alpas, Surf. Coat. Technol. 201 (7) (2006) 43524356.
[6] S. Bhowmick, M.J. Lukitsch, A.T. Alpas, J. Mater. Process. Technol. 210 (2010)
21422153.
[7] A. Banerji, S. Bhowmick, A.T. Alpas, Surf. Coat. Technol. 241 (2014) 93104.
[8] S. Bhowmick, A.T. Alpas, J. Manuf. Sci. Eng. Trans. ASME 135 (2013) 0610191-15.
[9] A. Erdemir, Surf. Coat. Technol. 146147 (2001) 292.
[10] E. Konca, Y.T. Cheng, J.M. Dasch, A.T. Alpas, Wear 259 (2005) 795799.
[11] Y. Qi, E. Konca, A.T. Alpas, Surf. Sci. 600 (2006) 29552965.
[12] G. Zilibotti, M.C. Righi, M. Ferrario, Phys. Rev. B 79 (075420) (2009) 110.
[13] X. Meng-Burany, A.T. Alpas, Thin Solid Films 516 (2) (2007) 325335.
[14] A.A. Gharam, M.J. Lukitsch, Y. Qi, A.T. Alpas, Wear 271 (2011) 21572163.
[15] E. Konca, Y.T. Cheng, A.M. Weiner, J.M. Dasch, A.T. Alpas, Surf. Coat. Technol. 200
(2005) 17831791.
[16] E. Konca, Y.T. Cheng, A.T. Alpas, Diam. Relat. Mater. 15 (2006) 939943.
[17] E. Konca, Y.-T. Cheng, A.M. Weiner, J.M. Dasch, A.T. Alpas, Tribol. Trans. 50 (2)
(2007) 178186.
[18] S. Bhowmick, A.T. Alpas, Int. J. Mach. Tools Manuf. 48 (2008) 14291443.
[19] S. Bhowmick, A.T. Alpas, Surf. Coat. Technol. 205 (2011) 53025311.
[20] M. Kano, Y. Yasuda, Y. Okamoto, Y. Mabuchi, T. Hamada, T. Ueno, J. Ye, S. Konishi, S.
Takeshima, J.M. Martin, M.I. De Barros Bouchet, T. Le Mognee, Tribol. Lett. 18 (2)
(2005) 245251.
[21] C. Matta, M.I. De Barros Bouchet, T. Le Mognee, B. Vachet, J.M. Martin, T. Sagawa,
Lubr. Sci. 20 (2) (2008) 137149.
[22] S. Bhowmick, A.T. Alpas, Int. J. Mach. Tools Manuf. 48 (2008) 802814.
[23] H. Ronkainen, S. Varjus, J. Koskinen, K. Holmberg, Wear 249 (2001) 260266.
[24] A. Banerji, A. Edrisy, V. Francis, A.T. Alpas, Wear 311 (2014) 113.
[25] A. Banerji, H. Hu, A.T. Alpas, Wear 301 (2013) 626635.
[26] A. Banerji, H. Hu, A.T. Alpas, Adv. Mater. Res. 445 (2012) 503508.
[27] A.A. Gharam, M.J. Lukitsch, M.P. Balogh, A.T. Alpas, Thin Solid Films 519, 2010.
16111617.
[28] C.E. Bell Jr., D.F. Taber, A.K. Clark, Infrared Absorption Spectroscopy, Organic Chemistry Laboratory: With Qualitative Analysis: Standard and Microscale Experiments,
3rd ed., Brooks/Cole Thomson Learning, Pacic Grove, CA, 2001. 95.
[29] V.P. Tolstoy, I.V. Chernyshova, V.A. Skryshevsky, Infrared Spectroscopy of Thin
Layers in Silicon Microelectronics, Handbook of Infrared Spectroscopy of Ultrathin
Films, John Wiley & Sons, Hoboken, NJ, 2003. 448.
[30] E. Konca, Y.T. Cheng, A.M. Weiner, J.M. Dasch, A.T. Alpas, Surf. Coat. Technol. 200
(2006) 39964005.
[31] F.G. Sen, X. Meng-Burany, M.J. Lukitsch, Y. Qi, A.T. Alpas, Surf. Coat. Technol. 215
(2013) 340349.
[32] T.W. Scharf, J.A. Ohlhausen, D.R. Tallant, S.V. Prasad, J. Appl. Phys. 101 (2007) 06352
1-11.
[33] A.C. Ferrari, J. Robertson, Phys. Rev. B Condens. Matter Mater. Phys. 61 (2000)
1409514107.
[34] J.F. Mammone, S.K. Sharma, M. Nicol, J. Phys. Chem. 84 (23) (1980) 31303134.

302
[35]
[36]
[37]
[38]

S. Bhowmick et al. / Surface & Coatings Technology 260 (2014) 290302

J. Andersson, R.A. Erck, A. Erdemir, Surf. Coat. Technol. 163164 (2003) 535540.
J. Andersson, R.A. Erck, A. Erdemir, Wear 254 (2003) 10701075.
C. Donnet, A. Grill, Surf. Coat. Technol. 9495 (1997) 456462.
J.A. Dean (Ed.), Lange's Handbook of Chemistry, 15th ed., McGraw Hill, New York,
1999, p. 8.81.
[39] Y. Kokaku, M. Kitoh, J. Vac. Sci. Technol. A 7 (3) (1989) 23112314.
[40] D.S. Kim, T.E. Fischer, B. Gallois, Surf. Coat. Technol. 49 (1991) 537.

[41] K. Hayashi, S. Sato, S. Bai, Y. Higuchi, N. Ozawa, T. Shimazaki, K. Adachi, J.M. Martin,
M. Kubo, Faraday Discuss. 156 (2012) 137146.
[42] Y.R. Luo, Comprehensive Handbook of Chemical Bond Energies, CRC Press, 2010.
255262.
[43] S.J. Blanksby, G.B. Ellison, Acc. Chem. Res. 36 (4) (2003) 255263.
[44] J.A. Dean (Ed.), Lange's Handbook of Chemistry, 15th ed., McGraw Hill, New York,
1999, pp. 4.414.53.