Synthesis of imidazolium-mediated Poly(benzoxazole) Ionene and composites with ionic liquids as advanced gas separation membranes

Article abstract of DOI:10.1016/j.polymer.2020.123239

Thermally rearranged (TR) polymers and ionic polymers are two material classes which have been employed in leading gas separation membranes. This work introduces a novel approach of combining the benzoxazole functionality associated with TR polymers with

Full text of DOI:10.1016/j.polymer.2020.123239

Surface & Coatings Technology 235 (2013) 475–483
Contents lists available at ScienceDirect
Surface & Coatings Technology
journal homepage: www.elsevier.com/locate/surfcoat
Improved properties of TiAlN coatings through the multilayer structure
a
a
b
a
a
a
a
A. Rizzo a, , L. Mirenghi , M. Massaro , U. Galietti , L. Capodieci , R. Terzi , L. Tapfer , D. Valerini
⁎
a
ENEA — Italian National Agency for New Technologies, Energy and the Sustainable Economic Development, Technical Unit for Materials Technologies, Brindisi Research Center,
S.S. 7 Appia km. 706, 72100 Brindisi, Italy
b
Politecnico di Bari, DIMEG — Dipartimento di Ingegneria Meccanica e Gestionale, Viale Japigia 182, I-70126 Bari, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 25 March 2013
Accepted in revised form 3 August 2013
Available online 13 August 2013
TiAlN/AlN multilayers are attracting great interest for the possibility to modulate their mechanical and tri-
bological properties through the variation of multilayer design. In this work TiAlN single layer, TiAlN/AlN
intermixed-multilayer and nano-multilayer were prepared using a reactive magnetron sputtering system
starting from targets of TiAl and Al. The aim is to analyze how the multilayer design affects the thermal
and tribological properties of the coatings. The microstructure of as-deposited and annealed films has
been studied using X-ray diffraction. The chemical composition has been deduced by XPS analyses. Thermal
behavior was assessed by means of differential thermal analysis (DTA) and thermogravimetric analysis
(TGA), while mechanical properties have been investigated by wear tests.
Keywords:
TiAlN
Multilayers
Reactive magnetron sputtering deposition,
Thermal stability
© 2013 The Authors. Published by Elsevier B.V.
Open access under CC BY-NC-ND license
.
1. Introduction
that the properties of the resulting coating are optimized for the specific
application required [10–12]. Then, once selected the proper materials,
an accurate engineering of the coating structure is necessary to identify
the powerful system with the best functional properties. In this
work TiAlN and AlN were chosen as basic materials and combined
under different architectures, namely, nano-multilayer, intermixed-
multilayer and single layer structures. All coatings were deposited by
magnetron sputtering. Afterwards, microstructure, chemical composi-
tion, thermal stability and mechanical properties of the single layer
and the multilayers were investigated by using a combination of X-ray
diffraction (XRD), X-ray photoelectron spectroscopy (XPS), thermal
analysis TGA (thermogravimetric analysis) and DTA (differential ther-
mal analysis) and tests of wear and friction. Finally, the worn surfaces
of these coatings were analyzed by means of scanning electron micros-
copy (SEM) and energy dispersive spectroscopy (EDS).
Surface coating on tools can confer them a high wear resistance, an
increased surface hardness at high temperatures and create a chemical
barrier for diffusion or reaction between the tool and the working-
piece, thus reducing tool wear [1,2]. Therefore, the demand to develop
new wear resistant hard coatings with good thermal stability has be-
come crucial to enhance tool life as machining speed increases [3]. The
first coating successfully used in steel machining industry and still
employed for its attractive bright gold color, is titanium nitride (TiN).
However, its use at elevated temperature is restricted due to its poor
chemical stability and it has been replaced by TiAlN in several applica-
tions [4]. It has been shown that the mechanical properties of TiAlN
material are strongly dependent on its chemical composition [5,6].
A coating with a composition of Ti0.35Al0.65N shows a slightly higher
oxidation resistance than Ti0.62Al0.38N. On the other hand, an increase
of Al amount in the TiAlN coating, such as in Ti0.19Al0.81N, revealed a
worse oxidation resistance, similar to that found for AlN [7]. Then,
improvements of TiAlN coating properties can be obtained optimizing
the Al content, the microstructure and the crystal orientation. Multi-
layer coatings and superlattices are the most widely employed to
provide a better alternative to single layer structures [8,9]. These sys-
tems allow to suitably combine different materials, selected to ensure
2. Experimental details
TiAlN and TiAlN/AlN multilayered coatings were deposited on sil-
icon and WC–6%Co using a multi-cathode reactive RF magnetron
sputtering system.
TiAl (99.95%) and Al (99.9%) targets were sputtered in high purity
plasma of Ar (99.999%) and N2 (99.999%). The composition of the TiAl
target was approximately 50:50 in atomic percentage. For all the exper-
iments, the power densities were approximately 3.88 and 1.94 W/cm2
for the TiAl and Al targets, respectively.
The coatings were deposited under a base pressure of 2.0 × 10−5 Pa
and a working total pressure of Ar + N2 gas equal to 4.0 Pa. The flow
rates of N2 and Ar were separately controlled by two mass flow control-
lers. The flow rate ratio of N2 and Ar + N2 was fixed at 5%.
⁎
Corresponding author. Tel.: +39 0831201428.
E-mail address: antonella.rizzo@enea.it (A. Rizzo).
0257-8972 © 2013 The Authors. Published by Elsevier B.V.
http://dx.doi.org/10.1016/j.surfcoat.2013.08.006
Open access under CC BY-NC-ND license
.

476
A. Rizzo et al. / Surface & Coatings Technology 235 (2013) 475–483
The substrates were cleaned in acetone and isopropyl alcohol by
sputtering the samples at 65° respect to the surface normal. The
use of this organic source ensures the absence of alteration of the
chemical composition and bonds along the specimen thickness, un-
like conventional Ar+ sources. In the depth profiles, a normalization
respect to the C1s signal has been operated, because the C1s signal
detected inside the films was due to the implanted carbon from
coronene. During spectrum acquisition the charge neutralizer was al-
ways on. “CasaXPS” software [14] was used both for quantitative peak
fitting on narrow spectra of Ti 2p, Al 2p, N1s and O 1s regions and on
survey spectra to built up the depth profiling.
Differential thermal analysis (DTA) with thermogravimetry (TGA)
was performed in a calibrated Netzsch-STA 409C from room tempera-
ture (RT) up to 1300 °C with a heating rate of 20 °C/min in flowing Ar
(99.999% purity, 20 sccm flow rate) and synthetic air (79% N2, 21% O2,
20 sccm flow rate) to mimic application conditions. Every DTA curve
was corrected using as baseline the second scan performed on the
same material. The annealing process of the coatings was performed
in vacuum (base pressure below 5 mPa) with heating and cooling
rates of 20 °C/min at temperatures of 700, 800, 900, and 1000 °C on
silicon substrates. The same coatings have been analyzed by XRD in
order to support the DTA results.
A pin-on-disk wear method was used to investigate the wear
resistance of the coatings. A bearing steel ball, 5.5 mm in diameter,
was adopted as the stationary pin. A normal load of 1 N was applied.
The sliding speed was 0.2 m/s on the circular track, for 500 rotations
of the sample. The test temperatures were 30 °C and 400 °C and the rel-
ative humidity was kept at 60%. The wear time was 40 min for each test.
Three friction tests were conducted for each specimen and for every
wear scan five measurements of the total wear volume were performed
at different points of the wear path and their average values are re-
ported. The wearing morphology of the films was investigated by
scanning electron microscope (SEM) XL40 Philips in plan view and
significant areas inside and outside the wear tracks were examined
by energy dispersive spectroscopy analyzer (EDS).
ultrasonic cleaning. The substrates were subsequently etched in situ
by Ar+ ion bombardment for 45 min, obtained by applying a RF bias
of −20 V to the substrate at an argon pressure of 6.0 × 10−1 Pa. The
coatings were deposited at a substrate temperature of 150 °C.
A Ti interlayer was deposited on the WC–6%Co substrates to improve
the adhesion of the coatings. Coatings with controlled layer thicknesses
were deposited using a stepper motor connected to the substrate holder
through a rotary feed. The substrate was moved in circular motion be-
tween the two sputtering targets using the stepper motor. The substrate
was kept underneath each target for a well predetermined time to
achieve the required multilayer thickness.
The multilayers were realized with two different methods of alter-
nation of TiAlN and AlN layers, i.e. continuously deposited with both
targets (TiAl and Al) turned on and the substrate rotating and passing
under each of the target (TiAlN/AlN-i, where “i” stands for “intermixed”)
or, alternatively, deposited with only one target turned on and with the
power switching between the two targets at each cycle (TiAlN/AlN-n,
where “n” stands for “nano-multilayer”). Look at Fig. 1 which shows
schematically in the deposition system the relative position of the
sputtering sources and the substrate table. In this way, it is easier to un-
derstand how the deposition was carried out for samples TiAlN/AlN-i
and TiAlN/AlN-n.
The multilayer period is about 4.5 nm. Coatings deposited on silicon
substrates with thickness of about 40 nm (corresponding to 9 periods in
multilayered films) were used for structural, chemical and thermal sta-
bility studies, while coatings deposited on WC–6%Co substrates with
thickness of approximately 2.2 μm (corresponding to 500 periods in
multilayered films) were used for wear tests.
The X-ray diffraction and reflectivity experiments were carried out
by using an X-ray diffractometer in parallel beam geometry (Philips
MPD PW1880, 3 kW generator) optimized for small-angle scattering
measurements. For all the measurements CuKα-radiation (λCuKα
=
0.154186 nm) was used. The X-ray diffraction measurements (XRD)
were performed at glancing incidence angle (with fixed incidence
angle ωi = 1.0°) in a range between 5° and 85°. Specular (ω, 2θ)
scans (XSR) were acquired at the grazing angle of incidence of the
X-rays equal to the exit angle 2θ measured in the range between 0°
and 9° with a step size of 0.01°.
3. Results
3.1. Structure of coatings
For X-ray photoelectron spectroscopy (XPS) studies, samples were
analyzed by AXIS Ultra DLD KRATOS using an Al Kα monochromatic
X-ray source at 600 W. Survey scans were acquired at a pass energy of
the analyzer equal to 160 eV, while narrow scans were acquired with
a pass energy of 20 eV. All spectra were recorded from an analysis
area of 700 × 300 μm2, while depth profiling was performed all along
the films' thickness using a coronene (C24H+12) organic source as
sputtering source [13]. The energy of the C24H+12 gun was 8 keV,
Fig. 2 presents the X-ray diffractograms of TiAlN, TiAlN/AlN-n and
TiAlN/AlN-i coatings. The TiAlN/AlN-n coating (see Fig. 2 in the middle)
shows the characteristic peak of the (101) direction of h-AlN s.g. P63mc
(JCPDS card #87-1054) at 2θ = 38.29°. Two broad peaks, centered at
~35° and ~40.2°, not expected, can be attributed to the hexagonal
phase of Ti2AlN (JCPDS card #18-0070). In addition there are
two peaks related to the (111) and (200) directions of the cubic
phase of TiAlN. Experimental investigations show that Ti1 − xAlxN
Fig. 1. Diagram of the deposition system and the relative positions of the target and the substrate. In this way it is easier to understand how the deposition was carried out for samples
TiAlN/AlN-i and TiAlN/AlN-n.

A. Rizzo et al. / Surface & Coatings Technology 235 (2013) 475–483
477
Fig. 2. The X-ray diffractograms of TiAlN, TiAlN/AlN-n and TiAlN/AlN-i coatings.
The symbols indicated in the legend of the figure correspond to the different phases
present in the analyzed films.
crystallizes in the cubic NaCl modification (space group Fm3m) for
AlN mole fractions x b 0.7 [15]. These latter peaks are present also
in the diffractogram of the TiAlN single layer and the TiAlN/AlN-i
multilayer (top and bottom of Fig. 2) where, however, the peaks corre-
sponding to aluminum nitride phases are missing. The other symbols
indicated in the legend correspond to the literature peak positions for
cubic TiN, cubic AlN, hexagonal AlN and hexagonal Ti2AlN [16]. The
peak position of c-TiAlN (111) XRD reflection is equal to 37.46° for the
single layer and becomes 37.71° for sample TiAlN/AlN-i while comes
back to 37.63° for sample TiAlN/AlN-n. In the literature it is reported
that the (111) diffraction peak shifts towards higher 2θ values at in-
creasing aluminum content and then the two multilayers have a higher
aluminum content with respect to the single layer, as expected by the
growth deposition. Kutschej et al. in Ref. [17] assess that the incorpo-
ration of Al in the fcc Ti1 − xAlxN solid solution leads to a shift of the
lattice parameter to lower values, caused by the substitution of Ti
atoms by smaller Al atoms, and therefore to a significant shift of
the fcc Ti1 − xAlxN peak to higher 2θ values compared to the standard
position of TiN. According to the (111) c-TiAlN peak position of sam-
ple TiAlN/AlN-i, with respect to film TiAlN/AlN-n and to the single
layer, we can say that sample -i has the higher Al content. Taking
into account the dynamics of the deposition process we can conclude
that when both targets are contemporarily swithced on, the forma-
tion of a TiAlN phase rich in aluminum takes place.
Fig. 3. XSR diffractograms for (a) the TiAlN/AlN-i coating and (b) the TiAlN/AlN-n coating.
The dashed arrows indicate the frequency due to the whole film while the solid arrows
show the period of the alternation of the two layers.
This suggests that the deposition approach used for the TiAlN/AlN-i
sample gives rise to an intermixing of the different materials in each
rotation cycle thus preventing the formation of nano-multilayer and
also negatively affecting the crystalline quality of the film.
3.2. XPS results
Instead, when the two targets are alternately shuttered, we assist to
the formation of three phases: a cubic TiAlN phase with less aluminum
content with respect to the other multilayer TiAlN/AlN-n but more alu-
minum content compared to the single layer TiAlN; a well separated
AlN phase; and a second ternary phase like h-Ti2AlN richer in titanium.
This can explain the XRD diffraction behavior of film TiAl/AlN-i.
XSR diffractogram [18] for the TiAlN/AlN-i coating shows only the
interference due to the two film-substrate and film–air interfaces
(Fig. 3a). In fact, a period equal to the thickness of the overall film
is obtained (40 nm). A different spectrum is evidenced for sample
TiAlN/AlN-n (Fig. 3b), whose reflectivity shows interfering peaks due
to two different periodicities: the asterisks show the period of the alter-
nation of the two layers (Λ = 4.28 nm) while the other peaks are due
to the whole film thickness (D = 37.7 nm). These values agree with
those obtained by the profilometer measurements of the total thickness
and the number of rotations.
The analyses of the surface region for the analyzed samples strongly
reveal the presence of nitride species on both spectral regions of Ti 2p
and Al 2p. After a peak fitting of the N 1s, Al 2p, Ti 2p and O 1s regions
of a representative coating (see Fig. 4), the binding energy (BE) posi-
tions of the nitrides (on Ti 2p and Al 2p experimental peaks) result at
lower values with respect to those predicted by the XPS NIST database
[19] for the TiN and AlN single phases and those collected on reference
samples [20].
In particular the N1s region (Fig. 4c) presents three different
components: the first one (1) at 397.4
(2) at 398.5 0.2 eV and the third (3) at 400.7
0.2 eV, the second one
0.2 eV. They cor-
respond to TiAlN and AlN compounds (1), TiON (2) and N2 surface
adsorbates (3), respectively. It can exclude the presence of TiN be-
cause no peak arises at 397 eV [21], as also confirmed by the absence
of TiN peaks in the XRD analysis. A deconvolution of Al 2p region
(Fig. 4b), with two unresolved 2p components, points out two
sub-peaks: the first one at lower BE centered at 73.5 eV and another
one at higher BE at 74.6 eV. Reminding that Al 2p line of AlN can be
found in the range (74.3–74.7 eV) [19], the second component is un-
doubtedly attributed to AlN (together with a certain content of alumi-
num oxide as shown later in the test), while the first one could be
Therefore, the two deposition routines give rise to two structures
with different stacking: TiAlN/AlN-n is a multilayered structure with
an alternation of AlN and TiAlN films, on the other hand, TiAlN/AlN-i
appears like a single film with only a cubic TiAlN phase with a level of
order lower than the TiAlN reference sample.

478
A. Rizzo et al. / Surface & Coatings Technology 235 (2013) 475–483
Fig. 4. XPS peak fittings of (a) Ti 2p, (b) Al 2p, (c) N 1s and (d) O 1s regions for a TiAlN/AlN-n representative coating.
attributed to AlNx unstoichiometric compound, as Chen has interpreted
[22], as well as to the TiAlN ternary compound. In our case the peak
at 73.5 eV has been attributed to the ternary compound, because of
the match with the peak position of the reference TiAlN single layer
(see Fig. 5) and also because of the binding energy position read on Ti
2p region, recently attributed to the TiAlN compound [20].
As shown in Fig. 4a, three different components are revealed after
peak fitting of the Ti 2p3/2 line: the first one at 455.0 eV is assigned to
TiAlN and it is shifted of 0.3 eV towards low binding energy respect
to the Ti 2p3/2 of the TiN compound acquired in our laboratories on
samples deposited with the same technique [23]; the second one at
456.4 eV is attributed to TiOxNy compound; the third at 458.2 eV is
identified as TiO2.
(Fig. 4b) because its energetic range (74.1–74. 6) eV falls in the same
predicted for AlN. Therefore we do not exclude the presence of Al2O3
under the peak at 74.6 eV, that in literature has been uniquely attribut-
ed to Al2O3 [20]. The results of the XPS surface analyses are summarized
in Table 1. Here there are the identification and the quantitative
percentage for the different components of Ti2p3/2, Al 2p, N 1s and O
1s for the main analyzed typologies of samples: TiAlN single layer,
TiAlN/AlN-i, and TiAlN/AlN-n.
As it can be noticed the reference single layer sample is a typical
TiAlN coating with titanium native oxides on the surface region. In par-
ticular the Al 2p signal shows the unique presence of a TiAlN chemical
environment (see Fig. 5). In the complex samples (TiAlN/AlN-i and
TiAlN/AlN-n) the Al signal has been equally divided into two com-
pounds, TiAlN and AlN, confirming what is expected by the growing
procedure. Concerning the Ti 2p signal, it is evident that there is a
very different quantitative distribution of the three compounds
evidenced after the fit. The TiAlN/AlN-i sample is much more oxidized
with respect to the other two samples. Results for the oxynitride com-
ponents in table 1 for sample TiAlN/AlN-i were overestimated due to a
resonance effect between the peaks associated to two different chemi-
cal titanium environments, as it has been presented by our structural
analysis and supposed by Kutschej et al. in Ref. [17].
O 1s after peak fitting (Fig. 4d) with the gaussian–lorentzian peak
shape shows two different components: the first one at 530.4 eV asso-
ciated to TiO2 and the second one at 532.15 eV associated to aluminum
oxide. The Al2O3 contribution is not well separated on Al 2p peak
In order to get qualitative information inside the films, XPS measure-
ments were taken at different depths through the coronene sputtering
source, allowing depth profiling of the relative concentration of the
four main elements in the films (Ti, Al, N, and O). The depth resolution
at optimized operative parameters was about 10 nm, that is higher
than the bilayer period, so it was not possible to calculate a reliable stoi-
chiometry for each layer. As a consequence, the depth profiling allowed
us to obtain an average concentration of the overall Ti, Al, N and O con-
tents in the films. The average relative atomic percentage inside the sin-
gle TiAlN film (spectrum not reported) was Ti 27%, Al 23%, N 48% and O
2%. In Fig. 6a and b the depth profiles for TiAlN/AlN-i and TiAlN/AlN-n
samples are reported, indicating an overall composition of Ti 19%, Al
38%, N 39% and O 4% for sample TiAlN/AlN-i and Ti 23%, Al 40%, N 36%
Fig. 5. XPS Al 2p signal obtained for typical single layer TiAlN coating.

A. Rizzo et al. / Surface & Coatings Technology 235 (2013) 475–483
479
Table 1
The identification and the quantitative percentage for the different components of Ti 2p3/2, Al 2p, N 1s and O 1s for the main analyzed typologies of samples: TiAlN single layer, TiAlN/AlN-i,
and TiAlN/AlN-n.
Ti 2p3/2
Al 2p
N 1 s
O 1 s
Ti–Al–N at.%
Ti–O–N at.%
Ti–O at.%
Al–Ti–N at.%
Al–N at.%
N–Ti–Al at.%
N–Ti–O at.%
N2 at.%
O–Ti at.%
O–Ti–N at.%
TiAlN
TiAlN/AlN-i
TiAlN/AlN-n
35
5
64
38
60
19
27
35
17
100
51
46
0
49
54
84
45
41
16
35
31
0
20
29
45
42
33
55
58
67
and O 1% for sample TiAlN/AlN-n. The nitrogen amount is lower than
the overall metallic species contribution and, correspondingly, the
amount of aluminum is always higher than titanium atomic percentage.
It can be noticed that the overall Ti and Al percentages are about the
same for both TiAlN/AlN samples, which is reasonable because both
coatings are realized with the same deposition parameters (power,
voltage and pressure). At the same time, the Ti and Al contents in
these two films are respectively lower and higher than those in the
single layer, as expected since in these two samples the Ti-containing
layers (i.e. TiAlN layers) are alternated with AlN layers, thus giving a
lower overall Ti percentage and higher overall Al percentage with
respect to the single TiAlN layer.
where three distinct phases, Ti2AlN, TiAlN aluminum-rich and AlN are
identified, and XPS provides a chemical composition ratio Ti/Al b1.
3.3. Thermal behavior of TiAlN/AlN multilayer coatings
Although the decomposition of metastable TiAlN with NaCl-type
structure (c-TiAlN) upon annealing has been the subject of considerable
research effort in the last decade [24–26], the structural evolution of
films with mixed hexagonal Ti2AlN and cubic TiAlN phases under
annealing process is still little studied [27]. Then, according to the re-
sults of XRD investigation, the sample TiAlN/AlN-n appears more inter-
esting than the other samples that showed peculiarities already known.
The DTA curve for TiAlN/AlN-n coating, shown in Fig. 7a, evidences dif-
ferent peaks (labeled with numbers from 1 to 4) which indicate that
several exothermic reactions occurred during annealing in Ar atmo-
sphere from RT up to 1300 °C. XRD spectra taken after annealing at
different temperatures can give information about physical transfor-
mations happening in the film. From the bottom to the top, Fig. 7b
shows the GIXRD patterns for the film as-deposited and annealed
at 700, 800, 900 and 1000 °C.
The reported XPS average quantifications give place to a Ti/Al ratio
around 0.5, depending on the architecture of the multilayers. This is in
agreement with the XRD results, also in the case of sample TiAlN/AlN-n
The diffractogram of the as-deposited sample clearly exhibits the
presence of the cubic c-TiAlN and hexagonal h-Ti2AlN ternary phases.
Reflections in the vicinity of the theoretical positions of the h-AlN
phase are also observed.
The two GIXRD patterns taken after annealing at 700 and 800 °C are
very similar, as expected due to the absence of any significant reaction
between these two temperatures in the DTA curve. These spectra
show that: peaks at 35°, 40.2° and 44° associated to h-Ti2AlN disappear
as well as those associated to h-AlN; the c-TiAlN becomes sharper; three
additional peaks at 39.2°, 42.3° and 43.4° appear but they do not match
any previous reflection. The position and intensity of these peaks do not
change at increasing the annealing temperature, indicating that this
phase is thermally stable up to 1000 °C. Persson et al. [27] found a
broad peak at 2θ = 38.5° indicative of the presence of intermetallic
phases with residual Ti2AlN content attributed to Ti3Al and TiAl phases,
in TiAlN coating realized at a substrate temperature of 1050 °C and
lower nitrogen flux. Zhang et al. [28] realized films composed of
Ti2AlN and tetragonal Ti2N phases (JCPDS Card No. 77-1893), finding
that their phase structure did not change during annealing process,
indicating phase stability of Ti2AlN–Ti2N during vacuum annealing at
800 °C. From the phase diagram, we get that for the above identified
phases, TiAlN has tie lines with the intermetallics and Ti2AlN, so it is nat-
ural for them to stay together in this material [29]. In our case the cubic
phase c-TiAlN still remains after annealing at 700 °C, but the hexagonal
phases h-AlN and h-Ti2AlN have together given rise to an unique terna-
ry phase rich in titanium and aluminum probably related to h-Ti3Al2N2
phase. Reasonably the Ti in excess forms the tetragonal Ti2N already
observed at 700 °C [30]. The intermetallic phases are excluded in our
identification, because the starting ternary compounds are not decom-
posable and the chemical bond with nitrogen is not broken at 700 °C.
(111) c-TiN reflection comes out at 900 °C, with its intensity in-
creasing at increasing the annealing temperature. So after annealing at
1000 °C, TiN, Ti2N, and Ti3Al2N2 phases are copresent, as confirmed by
the phase diagram [16].
Fig. 6. XPS depth profiles for TiAlN/AlN-i and TiAlN/AlN-n films obtained using C24H1+2
(coronene) as sputtering source.
From these observations, the peaks in Fig. 7a can be identified as
follows: the first two peaks (1 and 2) are probably associated to an

480
A. Rizzo et al. / Surface & Coatings Technology 235 (2013) 475–483
3.4. Friction and wear
Single layer, intermixed-multilayer and nanolayered coatings de-
posited on WC–6%Co inserts have undergone wear tests at 30 °C and
400 °C.
The wear volume was assessed through the evaluation of the re-
moved material section area, by a high precision profilometer, multi-
plied by the wear length. The average of 5 measurements for each
track is here reported. Uncoated WC–6%Co insert has been used as
reference sample for the tests. In Table 2 tribological test results are
reported at two different temperatures: 30 °C and 400 °C. The values re-
ported are the average of the experimental results. At 30 °C the friction
coefficient μ is equal to 0.64 for the substrate inserts, while it decreases
going down to a value 0.56 for the single layers, 0.54 for TiAlN/AlN-i and
to 0.42 for the TiAlN/AlN-n coatings. The decreasing of the friction coef-
ficient is already a very important result, but the tests performed on
complex samples at high temperature (400 °C) evidenced even more
interesting features. Fig. 8 represents the average friction coefficient of
coatings as a function of the wear distance against steel balls at the tem-
perature of 400 °C for all the representative samples. The friction coeffi-
cient for all samples has a very low value; a possible reason is the low
normal load (1 N), which was used for the measurements and the low
average value of surface roughness. The average RMS surface roughness
is about 3.5 nm. For TiAlN single layer it appears clear that after 25 m a
delamination occurred with sudden removal of the coating. This may
suggest a weakness in term of adhesion between coating and substrate
for this kind of coating. Friction coefficient of the coating is assessed
where a clear contact between steel ball and coating occurred. Still it
is very interesting the reduction of friction before delamination. The be-
havior for coating TiAlN/AlN-i appears more regular with a lower fric-
tion coefficient and a contact morphology less sensitive to loads, with
an evident better adhesion between coating and substrate respect
to the single layer. A small variation in the friction coefficient
shows the presence of interaction between steel ball and coating.
The TiAlN/AlN-n coating friction coefficient is very low indicating a
smooth surface and suggesting a gradual exfoliation typical of the
laminated structure.
Summarizing the results, the friction coefficient increases to a value
of 0.53 at 400 °C both for the single layer and the TiAl/AlN-i, while it
decreases to 0.33 in the case of sample TiAl/AlN-n. A lower friction
value indicates less adhesive areas between the ball and the surface of
the tests, highlighting a different surface behavior of the two complex
samples and giving the opportunity to better identify the kind of the
wear mechanism. At the same time, the wear rate value is the highest
in the bare substrate, while it is lower with the single layer coating
and it further decreases to a limit value around 1.4 × 10−5 mm3/N/m
for the two complex sample typologies, indicating a better wear resis-
tance in the case of the complex samples.
The SEM observations after wear tests at 400 °C and in the dam-
aged regions are useful to better understand the mechanisms hap-
pening during the wear tests at the microscopic level. Fig. 9 shows
the back-scattered electrons (BSE) image for a wear track in the
Fig. 7. a) DTA signal for sample TiAlN/AlN-n in flowing Ar. (b) GIXRD spectra of the sample
as-deposited and after annealing at different temperatures. (c) DTA and TGA in flowing
synthetic air.
initial recovery process of deposition-induced lattice point defects,
followed by the disappearance of the AlN phase and the formation of
the new Ti3Al2N2 ternary phase; the third one (3) is related to the
appearance of Ti2N domains; finally, the last peak (4) shows the appear-
ance of cubic TiN.
In Fig. 7c the DTA curve is pictured for the representative TiAlN/AlN-n
coating after annealing in synthetic air up to 1300 °C. DTA signal is not
corrected by baseline because its signal is very low and it is comparable
to the error introduced by its correction. It exhibits an onset of the pro-
nounced exothermic peak due to oxidation at temperature higher than
900 °C, hence, at least 400 °C above the oxidation temperature of TiN
coatings. This information is more evident in the TGA signal compared
to the DTA curve. The excellent oxidation resistance is confirmed by
the preliminary results obtained by Wang et al. [31] on Ti2AlN coatings.
Table 2
Wear rate and friction coefficient μ, together with their corresponding standard deviation
σ, at temperature of 30 °C and 400 °C.
Sample
T °C
Wear rate
σWR
μ
σμ
(mm3/N/m) × 10−5
(mm3/N/m) x 10−6
TiAlN
30
400
30
400
30
400
30
400
5.11
8.89
4.41
1.35
4.19
1.42
15.8
19
3
2
5
5
4
5
3
5
0.56
0.54
0.42
0.33
0.54
0.53
0.64
0.66
0.026
0.024
0.023
0.020
0.018
0.025
0.026
0.025
TiAlN/AlN-n
TiAlN/AlN-i
WC–Co

A. Rizzo et al. / Surface & Coatings Technology 235 (2013) 475–483
481
it can be concluded that the lighter zones (like x) are likely the substrate
(flakings), the gray zone (y) is the remaining coating, and the darker
zone (z) is that part of the coating containing the ferrous debris of the
ball. Flakings are actually areas uncovered by the coating, especially in
the aftermath of the wear test. It means that the initial impact created
these delaminated areas, generally due to a surface fatigue phenome-
non, with consequent release of small debris of the coating material
which takes part of the subsequent erosion both of the surface and the
ball. In this case, abrasive wear of three bodies interacting can be spoken
of, where the debris formed in the initial phase of the test then contrib-
utes to abrasion. This situation is found every time there is a great differ-
ence between the hardness of the two bodies interacting. Although Fe
was detected on the coating, there was no Al or Ti element on the abra-
sion mark of the steel ball. The debris layer and the obvious debris ad-
herence represented the main characteristics of the wear track.
In Fig. 11 there is a comparison between the SEM image with
SE (a) and BSE (b), for the same groove considered for sample
TiAlN/AlN-n. It can be observed that there are many areas within
the groove where the film has not been completely removed, even
after 500 cycles. It drives towards the abrasive wear where, at the
same working conditions, the film can be removed with more diffi-
culty. In Fig. 11c a detail of the same groove is considered. The groove
generated by the wear test has a well defined and delimited contour
(right side of the image) and also the wall section of the coating is
well visible and presents a certain number of overlayered slides
(Fig. 11d), indicating the multilayered structure of the coating. Pre-
cisely this laminar structure is probably linked to the lower value
of the coefficient of friction.
Fig. 8. The average friction coefficient of coatings as a function of the wear distance against
steel balls at the temperature of 400 °C for all the representative samples.
single TiAlN layer, where different shades of gray show a different
chemical composition: the lightest areas (with greater average atomic
number) represent the substrate while the darkest correspond to the
film. EDS analysis performed inside the groove (labeled with a black
circle in the left side of Fig. 9) reveals a composition corresponding to
the WC–6%Co substrate, therefore indicating a complete removal of
the TiAlN coating (white circle). The inset at the top right of Fig. 9
shows a secondary electron (SE) image which highlights a borderline
region with not well defined contours, from where a certain number
of cracks start to propagate onto the coating surface. Although the
load was very weak the material is very fragile and not tough, so that
it can be spoken of as having “brittle failure wear”.
4. Summary and conclusions
In Fig. 10 the SEM image with SE (a) and with BSE (b) shows the
surface region of a wear groove in the multilayer TiAlN/AlN-i. It is well
evidenced that the groove region is not homogeneous and presents
localized pits of about 50 μm. EDS analyses were performed at three
different zones, labeled with x, y and z. Looking to the EDS peaks,
Single layer, intermixed-multilayer, and nanolayered coatings of
TiAlN/AlN were fabricated using reactive magnetron sputtering in
Ar and N2 gas mixture. The experimental results can be summarized
as follow:
o
o
Fig. 9. Worn surface BSE micrograph of single layer TiAlN deposited on WC–6%Co after wear test at 400 °C. At the left side, EDS analyses performed inside the groove (labeled with
a black circle) and on the TiAlN (labeled with a white circle). At the top right there is a particular zone at higher magnification taken with SE, that shows a certain number of
cracks onto the coating surface.

482
A. Rizzo et al. / Surface & Coatings Technology 235 (2013) 475–483
a
x
y
z
b
x
z
y
Fig. 10. SEM image with SE (a) and with BSE (b) relative to a surface region of a wear groove for the multilayer TiAlN/AlN-i. At the right side EDS analyses performed inside the groove in
three different points labeled with x, y and z.
Fig. 11. SEM image with SE (a) and BSE (b) of the same groove for sample TiAlN/AlN-n. (c) and (d) Details of the same groove.

A. Rizzo et al. / Surface & Coatings Technology 235 (2013) 475–483
483
1. XRD examinations revealed that TiAlN single layer had a cubic struc-
ture of TiAlN with (111) preferential orientation. The Al 2p XPS sur-
face spectrum consisted of only one peak at 73.5 eV attributed to the
ternary nitride titanium aluminum compound. The surface Ti 2p3/2
peak presents, apart the oxide and oxynitride components, a clear
component attributed to the nitride titanium aluminum compound.
The relative atomic percentage inside the single TiAlN film was: Ti
27%, Al 23%, N 48% and O 2%. TiAlN single layer showed a brittle fail-
ure wear in fact, at 400 °C and a weak load of 1 N applied, wear test
grooves showed regions where the coating was completely removed.
2. TiAlN/AlN-i has been realized by continuous deposition with both
targets (TiAl and Al) turned on and the substrate that rotates and
passes under each of the targets. TiAlN/AlN-i appears like a single
film with the two materials intermixed, forming only a cubic TiAlN
phase with a level of order less than the TiAlN reference sample but
more rich in aluminum. The XPS Al 2p signal consisted of two
peaks: the first one at 73.5 eV attributed to the ternary nitride titani-
um aluminum compound and the other one at higher BE at 74.6 eV
attributed to AlN. The depth profile indicates the overall composition
for the main elements inside the coatings: Ti 19%, Al 38%, N 39%
and O 4%.
3. The multilayer TiAlN/AlN-n was realized with only one target alter-
natively shuttered. The XPS Al signal has been equally divided into
two compounds, TiAlN and AlN, confirming what was expected by
the growing procedure. The depth profile indicates the overall com-
position for the main elements of the coatings, that is: Ti 23% Al 40% N
36% and O 1%. TiAlN/AlN-n is a multilayered structure with an alter-
nation of AlN and TiAlN films where there is evidence of hexagonal
AlN phase besides the ternary phases of hexagonal Ti2AlN and
cubic aluminum rich-TiAlN.
to the crack propagation, a better wear resistance was registered for
smooth surface and hard multilayered coatings.
Acknowledgments
The authors are thankful to KRATOS Ltd-UK for its technical support
and in particular to Mr. Simon Hutton, Mr. Adam Roberts and
Mr. Jonathan Counsell for their precious contribution to XPS results.
References
[1] W. Grzesik, P. Nieslony, Wear 256 (2004) 108.
[2] G.M. Robinson, M.J. Jackson, M.D. Whitfield, J. Mater. Sci. 42 (2007) 2002.
[3] M. Kathrein, C. Michotte, M. Penoy, P. polcik, C. Mitter, Surf. Coat. Technol. 200
(2005) 1867.
[4] S.Y. Yoon, J.K. Kim, K.H. Kim, Surf. Coat. Technol. 161 (2002) 237.
[5] D. Mclntyre, J.E. Greene, G. Hakansson, J. Sundgren, W.D. Eand Munz, J. Appl. Phys.
67 (1990) 1542.
[6] J. Musil, H. Hrubý, Thin Solid Films 365 (2000) 104.
[7] J.-L. Huang, B.-Y. Shew, J. Am. Ceram. Soc. 82 (1999) 696.
[8] Z. Zhou, W.M. Rainforth, Q. Luo, P.Eh. Hovsepian, J.J. Ojeda, M.E. Romero-Gonzalez,
Acta Mater. 58 (2010) 2912.
[9] E. Voglia, W. Tillmannb, U. Selvadurai-Lasslb, G. Fischerb, J. Herper, Appl. Surf. Sci.
257 (2011) 8550.
[10] L. Chen, Y. Du, X. Xiong, K.K. Chang, M.J. Wu, Int. J. Refract. Met. Hard Mater 29
(2011) 68.
[11] L. Chen, Y. Du, P.H. Mayrhofer, S.Q. Wang, J. Li, Surf. Coat. Technol. 202 (2008)
5158.
[12] L. Chen, M. Moser, Y. Du, P.H. Mayrhofer, Thin Solid Films 517 (2009) 6635.
[13] A. Rafat, M.C. Davies, A.G. Shard, S. Hutton, G. Mishra, M.R. Alexande, J. Controlled
Release 139 (2009) 40.
[14] www.casaxps.com.
[15] T. Ikeda, H. Sato, Thin Solid Films 195 (1991) 99.
[16] Y.S. Han, K.B. Kalmykov, S.F. Dunaev, A.I. Zaitsev, J. Phase Equilib. Diffus. 25 (2004)
427.
[17] K. Kutschej, P.H. Mayrhofer, M. Kathrein, P. Polcik, R. Tessadri, C. Mitterer, Surf. Coat.
Technol. 200 (2005) 2358.
[18] E.E. Fullerton, I.K. Schuller, H. Vanderstraeten, Y. Bruynseraede, Phys. Rev. B 45
(1992) 9292.
DTA results indicate that several exothermic reactions occur during
annealing in Ar atmosphere from RT up to 1300 °C. XRD after annealing
at temperatures corresponding to exothermic peaks' position has
allowed the identification of peaks: the first two peaks are probably
associated to an initial recovery process of deposition-induced lat-
tice point defects, followed by the disappearance of the AlN phase
and the formation of the new Ti3Al2N2 ternary phase; the third one
is related to the appearance of Ti2N domains; the last peak shows
the appearance of cubic TiN. The films exhibited excellent oxidation
resistance and thermal stability at 900 °C in air. Compared with
other coatings, the multilayered coated samples presented better
wear resistance properties, due to structure stability and coating de-
sign features. In the case of sample TiAl/AlN-n the friction coefficient
has a lower value of 0.33 at 400° with respect to the value 0.53 both
for the single layer and the TiAl/AlN-i. A lower friction value indi-
cates less adhesive areas between the ball and the surface of the coat-
ings, highlighting a different surface behavior of the two complex
samples. Moreover TiAlN/AlN-n presented a better wear resistance
than the other typologies of films. This behavior can be associated to
the presence of interfaces: since each interface becomes an obstacle
[19] www.nistgov.com.
[20] M. Čekada, Surface properties and engineering of complex intermetallics, Cap 1,
pag.1 © World Scientific Publishing Co. Pte. Ltd., http://www.worldscibooks.com/
physics/7733.html.
[21] J.F. Marco, J.R. Gancedo, M.A. Auger, O. Sanchez, J.M. Albella, Surf. Interface Anal. 37
(2005) 1082.
[22] J.T. Chen, J. Wang, F. Zhang, G.A. Zhang, X.Y. Fan, Z.G. Wu, P.X. Yan, J. Alloys Compd.
472 (2009) 91.
[23] A. Rizzo, M.A. Signore, L. Mirenghi, T. Di Luccio, Thin Solid Films 517 (2009)
5956.
[24] A. Knutsson, M.P. Johansson, P.O.Å. Persson, L. Hultman, M. Odén, Appl. Phys. Lett. 93
(2008) 143110.
[25] P.H. Mayrhofer, A. Hörling, L. Karlsson, J. Sjölen, T. Larsson, C. Mitterer, L. Hultman,
Appl. Phys. Lett. 83 (2003) 2049.
[26] P.H. Mayrhofer, D. Music, J.M. Schneider, Appl. Phys. Lett. 88 (2006) 071922.
[27] P.O.Å. Persson, S. Kodambak, I. Petrov, L. Hultman, Acta Mater. 55 (2007) 4401.
[28] T.F. Zhang, Q.M. Wang, J.L. Peiling Ke, R. Nowak, K. Ho Kim, Surf. Coat. Technol. 212
(2012) 199.
[29] T. Joelsson, A. Flink, J. Birch, L. Hultman, J. Appl. Phys. 102 (2007) 074918.
[30] T. Ikeda, H. Sato, Thin Solid Films 195 (1991) 99.
[31] Q.M. Wang, W. Garkas, A. Flores Renteria, C. Leyens, K.H. Kim, J. Nanosci.
Nanotechnol. 11 (2011) 8959.