j m a t e r r e s t e c h n o l .

2 0 1 9;8(2):1653–1662

Available online at www.sciencedirect.com

www.jmrt.com.br

Original Article

Characterization of DLC coatings over nitrided

stainless steel with and without nitriding
pre-treatment using annealing cycles

Eugenia. L. Dalibón a,∗, Thierry Czerwiec b, Vladimir J. Trava-Airoldi c, Naureen Ghafoor d,

Lina Rogström d , Magnus Odén d , Sonia P. Brühl a
a Surface Engineering Group, Universidad Tecnológica Nacional (UTN-FRCU), Ing. Pereira 676, E3264BTD Concepción del Uruguay,
Argentina
b Institut Jean Lamour UMR 7198, CNRS, Université de Lorraine, Dept. CPSS, 54011 Nancy, France
c Instituto Nacional de Pesquisas Espaciáis (INPE), Av. dos Astronautas 1758, 12.227-010 São José dos Campos, SP, Brazil
d Nanostructured Materials, Department of Physics, Chemistry and Biology (IFM), Linköping University, SE-581 83 Linköping, Sweden

a r t i c l e i n f o a b s t r a c t

Article history: Amorphous hydrogenated diamond-like carbon (DLC) coatings were deposited using plasma
Received 24 June 2018 assisted chemical vapour deposition (PACVD) on precipitation hardening (PH) stainless steel.
Accepted 1 December 2018 Plasma nitriding has been used as pre-treatment to enhance adhesion and mechanical prop-
Available online 12 January 2019 erties. Chemical and mechanical properties of DLC coatings are dependent on the hydrogen
content and so on the relation between sp3 /sp2 bondings. The bondings and the structure
Keywords: of the DLC film change with temperature. In this work, a study of the thermal degradation
DLC coatings and the evolution of the mechanical properties of DLC coatings over PH stainless steel have
Thermal stability been carried out, including the effect of an additional nitrided layer.
Duplex treatment Nitrided and non-nitrided steel samples were subjected to the same coated in the same
Nitriding conditions, and they were submitted to the same thermal cycles, heating from room tem-
perature to 600 ◦ C in several steps.
After each cycle, Raman spectra and surface topography measurements were performed
and analyzed. Nanohardness measurements and tribological tests, using a pin-on-disc
machine, were carried out to analyze variations in the friction coefficient and the wear
resistance.
The duplex sample, with nitriding as pre-treatment showed a better thermal stability.
For duplex sample, the coating properties, such as adhesion, and friction coefficient were
sustained after annealing at higher temperatures; whereas it was not the case for only coated
sample.
© 2018 Brazilian Metallurgical, Materials and Mining Association. Published by Elsevier
Editora Ltda. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).

∗
Corresponding author..
E-mails: [email protected] (E.L. Dalibón), [email protected] (T. Czerwiec), [email protected] (V.J. Trava-
Airoldi), [email protected] (N. Ghafoor), [email protected] (L. Rogström), [email protected] (M. Odén), [email protected] (S.P. Brühl).
https://doi.org/10.1016/j.jmrt.2018.12.002
2238-7854/© 2018 Brazilian Metallurgical, Materials and Mining Association. Published by Elsevier Editora Ltda. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1654 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662

performed, and a complete study of the film properties and

1. Introduction degradation after the different annealing cycles was carried
out. Topography measurements were performed to improve
Despite their amorphous characteristics and the presence of
the understanding of the coating degradation and friction
hydrogen carbon bondings, diamond-like carbon (DLC) coat-
coefficients measurements are presented. The influence of the
ings have unique properties similar to diamond such as very
nitrided layer on the thermal behaviour of the DLC coating is
high hardness, chemical inertia, and a very low friction coef-
analyzed.
ficient compared to other hard coatings.
DLC coatings can be deposited via plasma-assisted tech-
niques, from a solid carbon target, such as plasma vapour 2. Experimental
deposition (PVD) or via hydrocarbon precursors by plasma
®
assisted chemical vapour deposition (PACVD). In the latter PH martensitic stainless steels (Corrax from Uddeholm) were
case, the presence of hydrogen in the film cannot be avoided, used as base material. The chemical composition in weight
®
and the chemical bondings are not only C C but also C H, percentage of Corrax is 0.03% C, 12% Cr, 1.4% Mo, 0.3% Mn,
with different proportions of sp3 and sp2 bondings, depending 0.3% Si, 9.2% Ni, 1.6% Al and Fe as balance.
on plasma deposition parameters such as time, temperature Disc-shaped samples were cut from a 25 mm in diameter
and gas flow [1–3]. In a wide range of structures, these films bar and they were hardened by means of an ageing process for
have properties similar to diamond such as high hardness, a two hours at 530 ◦ C according to supplier recommendations.
very smooth surface with a low coefficient of friction (below One group of samples was nitrided in an industrial reactor
0.1) and very good wear and corrosion protection properties. using a DC pulse discharge at 390 ◦ C for 10 h in a 25% N2 + 75%
However, these properties are not only affected by film struc- H2 gas mixture.
ture but also by the tribosystem properties: hardness profile, DLC coatings were deposited on all samples by PACVD
adhesion, mechanical support and stresses. For these reasons using methane as gas precursor. The process total gas pres-
nitriding has been used as pre-treatment, especially over soft sure was of 10 Pa at 10 ml/min gas flow during 2 h. Structural
substrates such as stainless steel [4–6]. properties of the nitrided layer and the DLC film without
Duplex treatments combining nitriding and DLC coatings annealing, so as the duplex system and corrosion properties
can be used to increase the mechanical properties and tribo- have been published before [10,11,21,22]. The group, which
logical performance by increasing the load bearing capacity of was not nitrided, but with a deposited coating was named
the so-formed composite. Such duplex treatments with nitrid- “coated samples” and the one, which was previously nitrided,
ing are often applied on medium alloy steels as substrates [7,8]. was named “duplex samples”. The coating thickness was
It was reported that a good wear resistance was obtained with about 3.0 ± 0.3 ␮m with a silicon interlayer of 0.3 ␮m and the
the DLC deposited over a hard and homogeneous compound nitrided layer thickness was of 14 ± 1 ␮m approximately.
layer [8,9]. Although other authors found that the best load- The heat treatment post DLC deposition consisted of
bearing capacity was achieved for a nitrided layer primarily annealing steps at 200 ◦ C, 300 ◦ C, 400 ◦ C, 500 ◦ C and 600 ◦ C for
constituted by a diffusion layer [7]. 1 h, counting from the time the temperature step was reached.
In the case of stainless steels, where only a nitrogen- The heating rate was 20 ◦ C/min and both coated and duplex
expanded phase arises after low temperature nitriding, some samples were heat treated in a vacuum chamber, following
papers were previously published [10–12]. a procedure reported previously in the literature [22]. In each
DLC coatings are frequently used in applications for which step, one group of samples was taken out. So at the end, five
there may be high work temperature or localized heating groups of samples annealed at different temperatures were
caused by friction. Hence detailed studies of the evolution of studied, each group with equal quantity of coated and duplex
coating properties with thermal solicitations are necessary, samples.
for example, to use it in engine components [13,14]. Several The DLC coatings were characterized by Raman spec-
studies about the thermal stability of DLC were found in the troscopy. The Raman spectra were fitted using two Gaussian
literature, reporting detrimental structural changes caused by lines. The intensity for each band was determined and the
the change in the carbon bondings, from sp3 to sp2 [15,16]. ID /IG ratio was calculated. The hydrogen content was esti-
Also, some papers were published about improving thermal mated using the slope of the photoluminescence background
stability using dopant elements or nanoparticles incorporated in the Raman spectra taking into account the model proposed
to the film [17–19]. by Casiraghi et al. [23].
A preliminary work was published on the thermal The surface was observed by electronic microscopy and
behaviour of duplex coatings [20]. In this work a different white light interferometry (WLI) using 20× objective. Both sur-
thermal degradation between duplex and coated samples was face map and profile were obtained by Metropro© software.
detected. The nitriding treatment improved not only the adhe- Moreover, focused ion beam (FIB) cuttings were performed in
sion of the coating but it also affected the surface properties. It different regions of the surface samples in order to observe
was decided to continue with the research in this topic, ana- the cross section and to measure the DLC thickness after the
lyzing new aspects. The main goal of the present work is to annealing. Mechanical properties were assessed using instru-
study not only changes in structural properties but also the mented nanoindentation with 15 mN load.
tribological behaviour of DLC coatings deposited on nitrided Sliding wear tests were carried out using 2 N load, 1800
and non-nitrided precipitation hardening (PH) stainless steels cycles, 1 mm of amplitude and alumina balls ( = 6 mm) as
going further than in the previous work. Raman spectra were counterpart to evaluate the tribological behaviour of the
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662 1655

This fact can be explained in two ways. The coating was

18 Duplex sample
Coated sample
detached partially or the film suffered a structure or com-
position change. The first one could be easily observed in
17
some regions of the surface and the second one was ana-
Nanohardness (GPa)

16 lyzed with Raman spectroscopy and the results are presented

below. When graphitization occurs in DLC films, the hardness
15 decreases [18].
The Raman spectra allowed the possibility to evaluate the
14
film structure. It can be seen in Fig. 2 that both groups of sam-
13
ples presented similar features after annealing at 200 ◦ C and
300 ◦ C. The DLC without thermal post treatment was analyzed
100 200 300 400 500 600 as well and it was depicted for comparison in black. In this
Temperature (°C) case, the Raman spectrum presented the classical two over-
lapping bands D and G, which, as it was already published,
Fig. 1 – Variation of nanohardness in DLC coatings after the
were well positioned, in 1379 cm−1 and 1554 cm−1 [15,16].
annealing cycles at different temperatures for duplex and
However, in the coated sample at 400 ◦ C, it could be noticed
coated samples.
that the film transformation started because the characteris-
tic G and D bands of DLC coatings could not be distinguished
(Fig. 2). On the other hand, in the duplex sample, this did not
coatings. Three tests were performed for each sample. The occur until reaching a temperature value of 500 ◦ C (Fig. 2).
wear tracks were observed by SEM and WLI. After annealing, After annealing at 200 ◦ C and 300 ◦ C, a slight shift in the
the degradation of the surface could be evaluated from the D and G peak positions was detected, as it is shown in Fig. 2
values of the coefficient of friction and the observation of and Table 1. For these temperatures, in both samples, it was
the surface damage along and at the sides of the wear tracks. possible to measure the D and G bands position, the intensity
The changes in film properties, surface topography and ratio between the D and G bands and the D and G bands width,
tribological behaviour were compared after each thermal indicating only a very small graphite domain for both samples
cycle to describe the coating degradation with temperature. (see Table 1).
For higher annealing temperatures, it was not possible to
perform an accurate analysis of the DLC coatings. A very
3. Results strong fluorescence appeared at 400 ◦ C, revealing degrada-
tion of the structural properties of the film in the case of
3.1. Changes in film properties with thermal cycles the coated sample (Fig. 2). On the contrary, in the duplex
sample, the analysis of the Raman spectrum corresponding
The hardness for the DLC films and for the nitrided layer were to 400 ◦ C annealing could still be carried out. The degrada-
16 ± 1 GPa and 11 ± 1 GPa respectively as it was reported in a tion starts at 500 ◦ C, as it can be observed in the Raman
previous publication [11]. The mean values for nanohardness Spectrum of Fig. 2, where the G and D bands cannot be dif-
measured after each annealing cycle are presented in Fig. 1, ferentiated. Finally, at 600 ◦ C, there was no characteristic DLC
comparing coated and duplex samples. Twenty measure- coating structure in the Raman spectrum for both groups of
ments in different places were carried out and the magnitude samples.
of the error was the standard deviation of the measured val-
ues. As the indenter penetration is less than 10% of the film 3.2. Changes in surface topography and film integrity
thickness, this hardness value can be considered as a film
property without the influence of the substrate. Surface topography did not change after the annealing at
In this test, some differences became visible. For the duplex 200 ◦ C and 300 ◦ C in both coated and duplex samples. More-
sample, the nanohardness values did not have an apprecia- over, this latter did not present changes in the surface
ble variation, up to 500 ◦ C. Only at 600 ◦ C, some changes were topography after the annealing at 400 ◦ C (Fig. 3a). However,
observed, which were probably caused by structural modifica- different features indicating a film transformation appeared
tion. when the surface of the coated samples was observed after
On the other hand, a hardness decrease was noticed in the the 400 ◦ C cycle. In fact, part of the film was delaminated;
coated sample at 300 ◦ C. Probably, a degradation that cannot bubbles and holes were formed, as it is shown in Fig. 3b.
be observed occurred at the interface between the DLC coating Moreover, the existence of holes was confirmed by the differ-
and the substrate. However, no changes were observed neither ent heights detected on WLI Images (Fig. 4a and b). It should
in surface images nor in the Raman spectrum, as it will be be mentioned that the presence of the coating, even thin-
mentioned below. ner, was confirmed by EDS analysis in all regions of interest
At higher temperatures, the dispersion increased in the (not shown).
coated sample as it can be observed in Fig. 1. It is pointless In the duplex samples, some changes could be observed
to indicate a mean nanohardness value when the annealing on the coating surface at the end of the annealing process at
was performed at 600 ◦ C because the dispersion is very high. 500 ◦ C, as it is shown in Fig. 5a. This sample presented some
The nanohardness changes from one point to another on the cracks on the surface (Fig. 5a) but there was not detachment
surface sample were between 4 GPa and 14 GPa. of the coating in any region. After annealing at 600 ◦ C, the
1656 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662

Without annealing
60000
200 °C
50000 300 °C
Duplex sample

Intensity (a.u.)
40000 400 °C
500 °C
30000
600 °C
20000

10000

0
800 1000 1200 1400 1600 1800

Without annealing
60000
200 °C
300 °C
Intensity (a.u.)

50000
400 °C Coated sample
40000
500 °C
30000 600 °C
20000

10000

0
800 1000 1200 1400 1600 1800
Raman shift (cm-1)

Fig. 2 – Raman spectrum of the duplex and coated samples after annealing at different temperatures.

Table 1 – Analysis of Raman spectra.

Samples D position G position ID /IG D width G width

Duplex sample without annealing 1379 1554 0.66 283 132

Duplex sample 200 ◦ C 1390 1570 0.88 289 121
Duplex sample 300 ◦ C 1395 1573 0.83 285 104
Duplex sample 400 ◦ C 1388 1570 0.76 276 109
Coated sample without annealing 1383 1557 0.71 283 127
Coated sample 200 ◦ C 1388 1561 0.74 284 121
Coated sample 300 ◦ C 1394 1572 0.83 284 103

Fig. 3 – SEM image of the samples surfaces after annealing at 400 ◦ C: (a) duplex sample and (b) coated sample.

a +2.02723 +0.60000 b
µm +0.40000
Height (mm)

-1.15330
0.212 +0.20000

mm +0.00000

-0.20000
0.000
0.000 mm 0.282 0.000 0.050 0.100 0.150 0.200 0.250
Distance (mm)

Fig. 4 – (a) WLI image of the coated sample surface after annealing at 400 ◦ C and (b) WLI surface profile.
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662 1657

Fig. 5 – SEM image of the duplex sample surfaces after annealing at (a) 500 ◦ C and (b) 600 ◦ C.

Fig. 6 – SEM image of the coated sample surfaces after annealing at (a) 500 ◦ C and (b) 600 ◦ C.

sample had more cracks as it can be seen in Fig. 5b, indicating 3.3. Tribological behaviour
a major degradation of the coating. However, the degradation
of the duplex coatings is much less than that of the only The evolution of the friction coefficient () is shown in Fig. 8
coated samples. as a function of time for two annealing temperatures. Other
Indeed, in the coated samples some cracks are observed in tribological results (final  values and wear damage) are
regions where the coating was partially detached at the end presented in Table 2 for duplex and coated samples. After
of the annealing cycle at 500 ◦ C (Fig. 6a). The degradation of annealing at 200 ◦ C and 300 ◦ C, the steady value of  for both
the coating was greater after the annealing at 600 ◦ C (Fig. 6b). samples was low and similar, about 0.2. This value is in the
However, the coating was not completely peeled off but some range of those that have been reported for this kind of coat-
surface layers were gone. ings in the literature [24]. The presence of the coating strongly
In the coated sample surface, three FIB cuts were per- reduced the friction coefficient in comparison with steel, with
formed and the location is indicated in Fig. 7. The first one or without the nitrided layer, which is about 0.7. This is due
covered a region of a hole defect from the outside to the inside to the formation of a graphite transfer layer which has a self-
(region named 1), the second one, on the edge of a protuber- lubricating effect [24,25]. At 200 ◦ C and 300 ◦ C the track was
ance with a delaminated part (named 2), and the third one, almost undetectable, indicating that the film was intact. After
over a protuberance (named 3). In region 1, Fig. 7c, a step can higher temperature cycles, the wear tracks were visible but
be observed. The right part of the coating has a thickness very irregular in width and depth. For this reason, a wear vol-
similar to the original coating, showing that the film has not ume loss could not be calculated. In addition, the coefficient of
been damaged there. In the left part of the film, thickness is friction becomes higher and irregular due to the surface inho-
only 15% less than the right part, showing that the film was mogeneities and areas with different coating thickness and
not totally removed. In regions 2 and 3, Fig. 7b and d, some flaking.
delaminated layers of the coating can be observed in the cross The friction coefficient increased for the coated sample
section. It can be seen that the minimum film thickness was after the annealing cycle at 400 ◦ C. It reached values that cor-
1.5 ␮m, indicating that part of the film was still there but thin- respond to the untreated steel friction coefficient (0.8–0.9 in
ner than the not annealed one (2.7 ␮m). It can be concluded the same test conditions). On the other hand, for the duplex
that several layers of the coating were degraded and detached, samples, the friction coefficient did not raise that much after
i.e. the coating delamination occurred by sheets, revealing a annealing at the same temperature; it begins at 0.2 and it
cohesive failure. reached a final value of 0.47, as it can be observed in Fig. 8.
1658 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662

a b
Platinum layer

1.53 µm (cs)
2
3
1

50 µm 2 µm

c d

343.7 nm (cs)
Platinum layer
Platinum layer

1.56 µm (cs)
2.37 µm (cs)

2.72 µm (cs)
2.42 µm (cs)

2µm 2µm

Fig. 7 – (a) SEM image of a surface region, where the film was degraded in the coated sample after annealing at 400 ◦ C,
showing the three regions were FIB cuts were performed. (b) Cross section in region 2, a protuberance that was in part
peeled off, can be observed. (c) The step in the cross section of region 1. (d) The subsurface structure of a protuberance,
corresponding to region 3.

Table 2 – Summary of the tribological results for duplex and coated samples.
Annealing (◦ C) Duplex sample Coated sample

 (steady or final) Wear damage  (steady or final) Wear damage

200 0.20 ± 0.01 Undetectable track 0.20 ± 0.01 Undetectable track

300 0.20 ± 0.01 Undetectable track 0.20 ± 0.01 Undetectable track
400 0.47 ± 0.08 Detectable track 0.86 ± 0.08 Coating detachment
500 0.77 ± 0.07 Coating detachment 0.90 ± 0.09 Coating detachment.
Greater damage than in
the duplex sample.
600 0.90 ± 0.09 Coating detachment 0.90 ± 0.09 Coating detachment

After annealing at 600 ◦ C, the friction coefficient rapidly Section 3.2, the coating is thinner and did not detach com-
reached the value of the base material for both samples pletely, some parts of it remained on the surface.
(Table 2), indicating that the coating has detached during the Fig. 9 presents SEM pictures of parts of the wear tracks at
test and the wear damage occurred on the substrate. In fact, the end of the tribological test for annealed samples at 400 ◦ C.
the wear track depth was greater than the coating thickness A high magnification was selected, with the intention to show
for both samples. topography features within and around the track. In Fig. 9a,
As a summary, the difference in tribological properties which corresponds to the wear track on the coated sample,
between both groups of samples occurred after annealing at the coating detachment is clearly visible; in fact, chemical ele-
400 ◦ C. The  values remained rather low after the anneal- ments corresponding to the steel substrate and aluminium of
ing process only in the duplex samples. It could be confirmed the counterpart could be detected by the EDX analysis. More-
not only that the coating was still present, but also that it over, some dark particles were also observed in the track. Since
has maintained its properties. As it was already mentioned in the EDX analysis shows that they are containing oxygen, it
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662 1659

of the wear scar to observe the damage features. Outside the

1,2
track, the film surface can always be observed and it is clear, by
Coated sample after annealing at 500°C
1,0 comparing both figures, that the damage was already present
before the test. It can also be observed that the degradation
Friction coefficient µ

0,8
was more extended in the coated sample than in the duplex
Coated sample after annealing at 400°C
0,6 one.
Duplex sample after annealing at 400°C Looking inside the wear track in the coated sample
0,4 (Fig. 10a), adhesive wear clearly occurred, and large plate-like
wear debris could be found as other authors also reported
0,2
Duplex sample after annealing at 500°C
[26]. It is well known that bonding occurs between surface
0,0 asperities, and when these junctions are broken, new ones
0 200 400 600 800 1000 1200 1400 1600 1800 2000
are formed as large lumps and transfer of materials comes as a
Time (s)
result. A large amount of plastic deformation is usually associ-
Fig. 8 – Friction coefficient versus time in the wear test for ated with this mechanism [24,26]. The wear was severe in this
duplex and coated samples annealed at 400 ◦ C and 500 ◦ C. case; the elements from the substrate and also Al from the
counterpart were detected by EDX analysis. On the contrary,
in the duplex sample, the damage could be described as mild
can be deduced that they are oxidized debris. The mean track abrasive wear. Soft grooves with some dark regions, which
width was 380 ␮m in this case. correspond to oxidized particles, can be observed (Fig. 10b)
In Fig. 9b, which corresponds to the wear track on the [27].
duplex sample, only some grooves in the direction of move-
ment and some small damaged regions can be observed. The
wear mechanism could be classified as mild abrasive. The 4. Discussion
coating was still or always present in the wear scar and around
it, even in the damaged areas, as it could be confirmed by the An explanation about the different behaviour of DLC coated
EDX spectrum, in which Si and C were detected. With less mag- PH samples with and without nitriding pre-treatment can
nification, the whole scar could be observed, and the mean be proposed and, related to this point, some hypotheses are
width was about 150 ␮m. described below.
At the end of the wear tests on samples after annealing at The first one is the chemical modification of the DLC film
500 ◦ C, some flaking could be detected in the wear tracks as with an increase in temperature. It is known that DLC films
it can be seen in Fig. 10a and b. In these figures, only parts of transform to an aromatic ring structure and then to disor-
the tracks are presented, because focus was made on the edge dered graphitic ring structure during annealing at temperature

a
Feka O ka

Fela
3
EDS zone 4
EDS zone 3
Particles
Damaged regions

crKa Fela
AlKa Feka
AuMa Auma
Cka Sika Nika Detachment coating Alka crKa
C ka Sika
Nika
2 4 6 8 keV 30 µm 2 4 6 8 keV

Sika b Sika
Grooves
1
Particles

EDS zone 1 Small damaged regions EDS zone 2

Mild abrasive wear

2
O ka O ka
C ka CrkaFeka C ka

2 4 6 8 keV 30 µm 2 4 6 8 keV

Fig. 9 – SEM images of the wear track and EDX analysis at the end of the tribological test for annealing at 400 ◦ C: (a) duplex
sample and (b) coated sample.
1660 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662

a Fela FeKa
Adhesive wear

Plate - like wear debris

EDS zone 2
Flaking

O ka AuMa CrKa
Alka

C ka Sika Nika

20 µm 2 4 6 8 keV

b
Sika

Oxidation

EDS zone 1

Flaking

O ka

C ka Feka
Crka

20 µm 1 2 3 4 5 6 keV

Fig. 10 – SEM images of the wear track and EDX analysis at the end of the tribological test after annealing at 500 ◦ C: (a)
duplex sample and (b) coated sample.

higher than 400 ◦ C [28]. With increasing annealing tempera- The difference between both groups of samples is the
ture, the structural disorder decreases and the size of graphitic nitriding pre-treatment. The influence that it has on the film
domains increase [28,29]. adhesion, which was better in the duplex sample than in the
Such phenomenon is associated with plastic deformation, coated sample, was reported in a previous paper [11]. The
which reduces the internal stress and in turn the hardness nitrided layer raises the load bearing capacity of the system.
of the DLC coating. According to the results presented above, This layer generates a graded interface between the coating
from properties such as hardness and the observation of sur- and substrate, increases the hardness of the substrate, reduces
face morphology, it can be stated that the transformation of the stresses and improves the adhesion as it was published
the DLC coating in graphite was produced at a higher tem- also by other authors [11,30,31]. Moreover, in both samples, a
perature in the duplex sample than in the coated sample. silicon interlayer was deposited previous to the coating. As a
This could be confirmed by the Raman analysis, where the result, there is a chemical affinity between this Si layer and
modification and the shift of the classical DLC peaks is a the nitrogen of the nitrided layer. So, in the duplex sample,
sign of coating degradation. Such degradation was detected in Si reacts with N forming silicon nitride bonds, as pointed out
both groups of samples at 600 ◦ C. However, this phenomenon by some of the authors, when XPS analysis were performed
occurred at a higher temperature in the case of the duplex in a thin Si layer deposited onto nitrided steel [32,33]. Since
samples as compared to the only coated samples. It is highly hardness and adhesion are higher for the duplex coating, con-
probable that in the duplex sample, the rate of graphitization sequently it is more stable at higher temperatures.
in the coating was lower than in the coated samples. There- Regarding thermal conductivity, the silicon adhesion layer
fore, the film transformation did not have so much influence has a thermal coefficient (3.2 × 10−6 K−1 ) comparable to
on the coating properties and it was also less noticeable. This those from DLC (2.3 × 10−6 K−1 ) and from the nitrided layer
change in graphitization is probably related to a change in the (7.6 × 10−6 K−1 ) but lower when compared to the one of bare
thermal conductivity as will be discussed later. stainless steel (14.5 × 10−6 K−1 ). As a consequence, thermal
 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662 1661

stress is lower in the duplex sample than in the coated sam- of samples. In the duplex, cracks appear on the film surface
ple [34,35]. On the other hand, the thermal conductivity of DLC and no delamination was observed, meanwhile in the coated
is very low; it can be hundred times lower than the thermal sample, the film has cohesive failures and it detached layer by
conductivity of steel. Therefore, if the adhesion is higher for layer. Changes in structure properties appear at 400 ◦ C in the
the duplex coating, the heat dissipation is higher as well and coated sample and at 500 ◦ C in the duplex one.
the film resists more to the thermal effect. A thin film with- It was shown that the low friction coefficient and good tri-
out good adhesion due to the low thermal conductivity heats bological properties of the DLC coating remained unchanged
rapidly modifying its structure, i.e. going faster to a structure in the duplex samples after an annealing process at 400 ◦ C.
close to the graphite or to a graphite-like film. As the dura- It was also observed that the nanohardness values did not
tion of the thermal cycle is only one hour, a delay can exist decrease after thermal treatments, even at high temperatures.
before the establishment of a stabilized temperature between It was proved that a nitriding treatment previous to the
nitrided and non-nitrided samples. Thus, some bubbles could coating deposition improves the DLC thermal stability through
be formed in the film with worse adhesion caused by the the thermal annealing cycles, if compared to the coating over
release of bonded and unbonded hydrogen during the anneal- the same stainless steel without any nitriding pre-treatment.
ing, as it was reported by other authors [34,36]. However, the It is a fact that a better adhesion is obtained in the duplex
bubble formation was not observed in well adherent films, e.g. samples, and there is also a better matching of the thermal
the duplex sample. This could be due to a more efficient heat coefficients and the mechanical properties of the nitrided
dissipation. layer with the DLC coating, allowing in some way that the
A relation between nanohardness measures and adhesion nitrided layer delayed the thermal degradation of the coating
can thus be established. That is, whenever a thin film that is deposited over it.
well adhered, the hardness measurement is more trustable In this paper, it is shown that the adhesion and the hard-
because the well-adhering film has the substrate as a more ness are related and that can be evaluated from a heat
efficient support. On the other hand, the non-adherent film treatment. This is a new result, which the authors can
is “loose” on the substrate, not allowing the hardness mea- speculate as being of much importance to any measure of
surement to be real, as well as being closer to the graphite nanohardness in thin films.
structure. In this work, it was quite clear that the disper- Further studies are necessary to relate the coatings struc-
sion of the nanohardness measurements in the coated sample ture to the properties of the interface with the nitrided
observed in Fig. 1 was directly related to the dispersion of the substrate and the early stages of growing, and their behaviour
adhesion through the interface, caused by the film degrada- at different temperatures.
tion.
The friction coefficient remained similar to the DLC coating
without thermal treatment after the annealing at 200 ◦ C and Conflicts of interest
300 ◦ C for both samples. After the annealing at 400 ◦ C, the fric-
tion coefficient was low (0.2) at the beginning. After 800 s the The authors declare no conflicts of interest.
value increased for both samples, but at the end it was higher
for the coated sample than the duplex sample. Probably as Acknowledgements
the hardness was lower for the coated sample compared to the
duplex sample and the original coating, the penetration of the The authors would like to thank students of the Surface
counterpart was deeper, so the real contact area increased and Engineering Group (UTN) for their collaboration in the experi-
consequently the friction coefficient too [26,37]. The friction mental tasks. Also to Dipl. Eng. Federico Lasserre at that time
coefficient increased when the annealing temperature was at Saarland University (Germany), for the help in SEM and
higher. This could be due to the fact that the annealing at high WLI experiments.PID UTI 4716 TC UTN, 2018-2021. SUMA2
temperatures produces the effusion of hydrogen and deteri- Network, EU FP7-PEOPLE-2012-IRSES, Project Nr. 318903
orate the coatings tribological properties, as it was reported
in the literature [38]. Moreover, the coating properties (hard-
ness, topography and thickness) effectively changed after the
references
annealing at this temperature, as it was presented in the previ-
ous sections. Consequently, the mechanical resistance of the
system decreased and the wear mechanisms changed. It is [1] Choy K. Chemical vapour deposition of coatings. Prog Mater
Sci 2003;48:57–170.
clear that the wear damage became greater with the increas-
[2] Robertson J. Diamond-like amorphous carbon. Mater Sci Eng
ing annealing temperature but, in all cases, the damage was R Rep 2002;37:129–281.
worse in the coated sample than in the duplex one. [3] Trava-Airoldi VJ, Bonetti LF, Capote G, Santos LV, Corat EJ. A
comparison of DLC film properties obtained by r.f. PACVD,
IBAD, and enhanced pulsed-DC PACVD. Surf Coat Technol
5. Summary and conclusions 2007;202:549–54.
[4] Chicot D, Puchi-Cabrera ES, Decoopman X, Roudet F, Lesage J,
Staia MH. Diamond-like carbon film deposited on nitrided
As a summary of results, it can be stated that at the same tem-
316L stainless steel substrate: a hardness depth-profile
perature, the coating degradation is always lower in the duplex modeling. Diam Relat Mater 2011;20:1344–52.
sample than in the only coated one. Moreover, the degradation [5] Snyders R, Bousser E, Amireault P, Klemberg-Sapieha JE, Park
or the morphology changes are different in the two groups E, Taylor K, et al. Tribo-mechanical properties of DLC
1662 j m a t e r r e s t e c h n o l . 2 0 1 9;8(2):1653–1662

coatings deposited on nitrided biomedical stainless steel. et al., editors. Proc. 3rd Pan Am. Mater. Congress. The
Plasma Process Polym 2007;4:S640–6. minerals, metals & materials series. Springer; 2017.
[6] Azzi M, Paquette M, Szpunar JA, Klemberg-sapieha JE, [21] Guitar A, Trava-airoldi VJ, Mu F, Brühl SP. Plasma nitriding
Martinu L. Tribocorrosion behaviour of DLC-coated 316L and DLC coatings for corrosion protection of precipitation
stainless steel. Wear 2009;267:860–6. hardening stainless steel. Adv Eng Mater 2015;18:826–32.
[7] Shioga JPHT, Binder C, Hammes G, Klein AN, de Mello DB. [22] Zhang S, Lam X, Li X. Thermal stability and oxidation
Effects of different plasma nitrided layers on the tribological properties of magnetron sputtered diamond-like carbon and
performance of DLC coatings. Mater Res 2016;19:1180–8. its nanocomposite coatings. Diam Relat Mater 2006;15:972–6.
[8] Kovaci H, Baran Ö, Yetim AF, Bozkurt YB, Kara L, Çelik A. The [23] Casiraghi C, Ferrari AC, Robertson J. Raman spectroscopy of
friction and wear performance of DLC coatings deposited on hydrogenated amorphous carbons. Phys Rev B 2005;72:1–13.
plasma nitrided AISI 4140 steel by magnetron sputtering [24] Donnet C, Erdemir A. Tribology of diamond-like carbon films.
under air and vacuum conditions. Surf Coat Technol fundamentals and applications. USA: Springer; 2008.
2018;349:969–79. [25] Holmberg K, Matthews A. Coatings tribology properties,
[9] Podgornik B, Vižintin J, Wänstrand O, Larsson M, Hogmark S, mechanisms, techniques and applications in surface
Ronkainen H, et al. Tribological properties of plasma nitrided engineering. UK: Elsevier; 2009.
and hard coated AISI 4140 steel. Wear 2001;249:254–9. [26] Wang L, Nie X, Hu X. Effect of thermal annealing on
[10] Dalibon EL, Charadia R, Cabo A, Trava-Airoldi VJ, Brühl SP. tribological and corrosion properties of DLC coatings. J Mater
Evaluation of the mechanical behaviour of a DLC film on Eng Perform 2013;22:3093–100.
plasma nitrided AISI 420 with different surface finishing. [27] Bayer RG. Wear analysis for engineers. New York: HNB
Surf Coat Technol 2013;235:735–40. Publishing; 2002.
[11] Dalibon EL, Trava-Airoldi VJ, Pereira LA, Cabo A, Brühl SP. [28] Peter S, Günther M, Gordan O, Berg S, Zahn DRT, Seyller T.
Wear resistance of nitrided and DLC coated PH stainless Experimental analysis of the thermal annealing of hard
steel. Surf Coat Technol 2014;255:22–7. a-C:H films. Diam Relat Mater 2014;45:43–57.
[12] Forsich C, Heim D, Mueller T. Influence of the deposition [29] Sheng R, Li L, Su D, Peng J, Gao J, Zhao K, et al. Effect of
temperature on mechanical and tribological properties of unbonded hydrogen on amorphous carbon film deposited by
a-C:H:Si coatings on nitrided and postoxidized steel PECVD with annealing treatment. Diam Relat Mater
deposited by DC-PACVD. Surf Coat Technol 2008;203: 2018;81:146–53.
521–5. [30] Podgornik B, Vižintin J, Leskovšek V. Tribological properties
[13] Kovaleva MG, Kolpakov AJ, Poplavsky AI, Galkina ME. Effect of plasma and pulse plasma nitrided AISI 4140 steel. Surf
of vacuum annealing on tribological behavior of nanosized Coat Technol 1998;108–109:454–60.
diamond like carbon coatings produced by pulse vacuum arc [31] Borges CFM, Pfender E, Heberlein J. Influence of nitrided and
method. J Frict Wear 2013;34:481–6. carbonitrided interlayers on enhanced nucleation of
[14] Ito H, Yamamoto K, Masuko M. Thermal stability of UBM diamond on stainless steel 304. Diam Relat Mater
sputtered DLC coatings with various hydrogen contents. 2001;10:1983–90.
Thin Solid Films 2008;517:1115–9. [32] Li D, Guruvenket S, Azzi M, Szpunar JA, Klemberg-Sapieha JE,
[15] Leng YX, Chen JY, Yang P, Sun H, Wan GJ, Huang N. Martinu L. Corrosion and tribo-corrosion behavior of
Mechanical properties and thermomechanical stability of a-SiCx:H, a-SiNx:H and a-SiCxNy:H coatings on SS301
diamond-like carbon films synthesized by pulsed vacuum substrate. Surf Coat Technol 2010;204:1616–22.
arc plasma deposition. Surf Coat Technol 2003;173:67–73. [33] Chen CW, Huang CC, Lin YY, Chen LC, Chen KH. The affinity
[16] Choi J, Miyagawa S, Nakao S, Ikeyama M, Miyagawa Y. of Si N and Si C bonding in amorphous silicon carbon
Thermal stability of diamond-like carbon films deposited by nitride (a-SiCN) thin film. Diam Relat Mater 2005;14:1126–30.
plasma based ion implantation technique with bipolar [34] Wei Ch, Yen JY. Effect of film thickness and interlayer on the
pulses. Diam Relat Mater 2006;15:948–51. adhesion strength of diamond like carbon films on different
[17] Moolsradoo N, Abe S, Watanabe S. Thermal stability and substrate. Diam Relat Mater 2007;16:1325–30.
tribological performance of DLC-Si-O films. Adv Mater Sci [35] Li XY, Sun Y, Bell T. Stability of the nitrogen S-phase in
Eng 2011. austenitic stainless steel. Int J Mater Res 1999;90:901–7.
[18] Hatada R, Baba K, Flege S, Ensinger W. Long-term thermal [36] Maheswaran R, Thiruvadigal DJ, Gopalakrishnan C. Thermal
stability of Si-containing diamond-like carbon fi lms instability of DLC film surface morphology – an AFM study.
prepared by plasma source ion implantation. Surf Coat AIP Conf Proc 2012;1447:789–90.
Technol 2016;305:93–8. [37] Cloutier M, Harnagea C, Hale P, Seddiki O, Rosei F, Mantovani
[19] Dai W, Gao X, Liu J, Wang Q. Microstructure, mechanical D. Long-term stability of hydrogenated DLC coatings: effects
property and thermal stability of diamond-like carbon of aging on the structural, chemical and mechanical
coatings with Al, Cr and Si multi-doping. Diam Relat Mater properties. Diam Relat Mater 2014;48:65–72.
2016;70:98–104. [38] Li H, Xu T, Wang C, Chen J, Zhou H, Liu H. Annealing effect
[20] Brühl SP, Dalibon EL, Ghafoor N, Rogström L, Trava-Airoldi VJ, on the structure, mechanical and tribological properties of
Odén M. Carbon based coatings deposited on nitrided hydrogenated diamond-like carbon films. Thin Solid Films
stainless steel: study of thermal degradation. In: Meyers M, 2006;515:2153–60.