ArticlePDF Available

Abstract and Figures

Time-averaged and time-resolved ion fluxes during reactive HiPIMS deposition of Ti1-xAlxN thin films are thoroughly investigated for the usage of Ti1-xAlx composite targets – Al/(Ti+Al) ratio x = 0.4 and 0.6. Ion mass spectroscopy analysis revealed, that increasing x in the target material or reducing the N2 flow-rate ratio leads to a proportional increase of the Al⁺-ion count fraction, whereas that of Tiⁿ⁺-ions (n = 1, 2) remains unaffected despite of comparable primary ionisation energies between Al and Ti. In fact, energetic Ti²⁺-ions account for the lowest flux fraction incident on the substrate surface, allowing for a high Al-solubility limit in cubic-structured Ti1-xAlxN thin films (xmax ~ 0.63) at low residual stresses. In addition, time-resolved plasma analysis highlights the simultaneous arrival of metal- and process-gas-ions throughout the entire HiPIMS pulse duration. These ion-bombardment conditions, which were dominated by gas-ion irradiation with a significant contribution of Al⁺-ions (up to ~ 20 %) and negligible energetic Ti²⁺-ions, allowed for the growth of cubic Ti0.37Al0.63N coatings exhibiting high indentation hardness of up to ~36 GPa at a low compressive stress level (σ = -1.3 GPa).
Content may be subject to copyright.
Surface & Coatings Technology 424 (2021) 127638
Available online 23 August 2021
0257-8972/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Time-averaged and time-resolved ion uxes related to reactive HiPIMS
deposition of Ti-Al-N lms
L. Zauner
a
,
*
, A. Bahr
a
, T. Koz´
ak
b
, J. ˇ
Capek
b
, T. Wojcik
a
,
c
, O. Hunold
d
, S. Kolozsv´
ari
e
,
P. Zeman
b
, P.H. Mayrhofer
c
, H. Riedl
a
,
c
a
Christian Doppler Laboratory for Surface Engineering of high-performance Components, TU Wien, Austria
b
Department of Physics and NTIS European Center of Excellence, University of West Bohemia, Czech Republic
c
Institute of Materials Science and Technology, TU Wien, Austria
d
Oerlikon Balzers, Oerlikon Surface Solutions AG, Liechtenstein
e
Plansee Composite Materials GmbH, Germany
ARTICLE INFO
Keywords:
Ti-Al-N
HiPIMS
Thin lm
Mass spectroscopy
Time-resolved
ABSTRACT
Time-averaged and time-resolved ion uxes during reactive HiPIMS deposition of Ti
1-x
Al
x
N thin lms are
thoroughly investigated for the usage of Ti
1-x
Al
x
composite targets Al/(Ti +Al) ratio x =0.4 and 0.6. Ion mass
spectroscopy analysis revealed, that increasing x in the target material or reducing the N
2
ow-rate ratio leads to
a proportional increase of the Al
+
-ion count fraction, whereas that of Ti
n+
-ions (n =1, 2) remains unaffected
despite of comparable primary ionisation energies between Al and Ti. In fact, energetic Ti
2+
-ions account for the
lowest ux fraction incident on the substrate surface, allowing for a high Al-solubility limit in cubic-structured
Ti
1-x
Al
x
N thin lms (x
max
~ 0.63) at low residual stresses. In addition, time-resolved plasma analysis highlights
the simultaneous arrival of metal- and process-gas-ions throughout the entire HiPIMS pulse duration. These ion-
bombardment conditions, which were dominated by gas-ion irradiation with a signicant contribution of Al
+
-
ions (up to ~20 %) and negligible energetic Ti
2+
-ions, allowed for the growth of cubic Ti
0.37
Al
0.63
N coatings
exhibiting high indentation hardness of up to ~36 GPa at a low compressive stress level (
σ
= − 1.3 GPa).
1. Introduction
The past decades in the development of plasma-based physical vapor
deposition (PVD) techniques have meticulously focused on accessing
innovative routes of controlled ion-bombardment during thin lm
growth [14]. Within this context, high-power impulse magnetron
sputtering (HiPIMS) a technological advancement to drastically
enhance the ux of ionized target species involved in conventional DC
magnetron sputtering (DCMS) is rendered particularly attractive for
introducing novel means of inuencing working gas- and metal-ions
incident on substrate and growing lm [59]. HiPIMS utilises short
(10500
μ
s), yet highly energetic plasma discharges (up to several
kWcm
2
) at relatively low repetition rates (tens of Hertz to kilo-Hertz)
to generate signicantly increased plasma densities on the target sur-
face, thus leading to an enhanced fraction of ions present in the lm-
forming vapor [6,7]. The increased contribution of ions to thin lm
growth allows for an enhanced control over the energy delivered to the
growing lm, revealing new kinetic pathways of controlling the
structure-property relationship [1014]. Moreover, ion-driven phe-
nomena such as process gas rarefaction, self-sputtering, or working gas
recycling are characteristic for HiPIMS discharges, hence creating
distinctly different process conditions when compared to DCMS, espe-
cially under reactive deposition environments [1519]. Nevertheless,
numerous studies already highlighted the benecial impact of HiPIMS
on thin lm quality, producing densied microstructures, improved
mechanical properties, reduced surface roughness, or enhanced lm
uniformity on complex shaped substrates [2024].
Many of the advantageous aspects of HiPIMS are driven by or related
to the dynamics of the ion distribution bombarding the growing lm
surface. Consequently, complementary detailed plasma analysis in the
time- and energy-domain has evolved into a vital extension for such ion-
based PVD techniques through unfolding the correlation between
deposition parameters and ion distributions (i.e., composition, energy,
or temporal sequence) arriving on the substrate surface [11,2533].
Utilising ion mass spectroscopy allows for methodically altering the
incident ion composition in conjunction with given deposition
* Corresponding author.
E-mail address: lukas.zauner@tuwien.ac.at (L. Zauner).
Contents lists available at ScienceDirect
Surface & Coatings Technology
journal homepage: www.elsevier.com/locate/surfcoat
https://doi.org/10.1016/j.surfcoat.2021.127638
Received 2 August 2021; Received in revised form 17 August 2021; Accepted 18 August 2021
Surface & Coatings Technology 424 (2021) 127638
2
parameters and to identify the distinct inuence of individual ionic
species on the resulting thin lm quality [34]. Moreover, combining a
known temporal sequence of process gas- and metal-ions with a specially
tailored substrate bias potential synchronised to a precise time-domain
within the HiPIMS pulse, provides advanced pathways of controlling
coating stoichiometry, lm adhesion, and even phase formation during
ion-assisted thin lm growth [3538].
Taking up the concept from Ref. [39], Nedfors et al. [40] have shown
that using time- and energy-resolved ion mass spectroscopy in
conjunction with a substrate bias synchronised to different time-frames
of the HiPIMS discharge can be utilised to inuence the average energy
per deposited species E
D
even when using a multi-element target.
Moreover, specically attracting lm-forming constituents (there B
+
-
ions during the deposition of TiB
2
) was demonstrated to mitigate Ar
+
incorporation, and thus also residual compressive stresses. A comple-
mentary work by Bakhit et al. [41] also highlighted an alternative
approach of using time-dependent ion mass spectroscopy combined with
decreasing HiPIMS pulse lengths to effectively control the B/Ti-ratio in a
wide range for this thin lms class.
Detailed in-situ plasma analysis during HiPIMS deposition of Al
alloyed cubic TM-N-based coatings has further contributed to access
additional pathways of maintaining, or even exceeding previously
established solubility limits for these typically supersaturated cubic
structures e.g. the well-studied Ti-Al-N system [42,43]. This bench-
mark system exemplies the capabilities of kinetically-limited lm
growth during PVD, providing excellent high hardness and wear pro-
tection when synthesised in the metastable face-centred cubic structure
(c, B1, NaCl-prototype). Since the high-temperature oxidation resistance
of c-Ti
1-x
Al
x
N scales with the AlN mole fraction (x), a commonly adopted
strategy is to maximise the Al/(Ti +Al)-ratio of the metal sublattice,
while aiming to maintain the cubic structure type. Exceeding the solu-
bility limit (x
max,
~ 0.67 for DCMS or arc evaporation), results in the
precipitation of the thermodynamically stable hexagonal (w, B4, ZnS-
wurtzite prototype) AlN phase, thus deteriorating both the thermal
stability and mechanical properties [4448]. In fact, most studies uti-
lising reactive HiPIMS (R-HiPIMS) for the synthesis of Ti
1-x
Al
x
N thin
lms reported on drastically reduced solubility limits compared to
conventional plasma-based techniques [4951].
Utilising a hybridAl-DCMS/Ti-HiPIMS (and Al-HiPIMS/Ti-DCMS)
approach, Greczynski et al. [34] identied the detrimental effect of en-
ergetic, doubly-charged Ti-ions on the structure of reactively deposited
Ti
1-x
Al
x
N thin lms. It was demonstrated that lattice defects induced
from intense Ti
2+
-ion bombardment act as preferred nucleation sites for
the wurtzite phase, causing a drastic reduction of x
max
. Building on this
concept, a novel synthesis route was developed where high temporal
uxes of Al
+
-ions are specically employed during a HiPIMS pulse to
extend the metastable Al-solubility limit of c-TM-Al-N based coatings
beyond values observed for conventional PVD techniques. By applying a
high substrate bias synchronised to the Al-rich portion of the discharge,
energetic Al
+
-ions are incorporated into the c-TM-N host structure
referred to as subplantation” – thereby effectively suppressing any
diffusion driven precipitation of a second phase. This method was
proofed successful especially for the V-Al-N system, where an increase of
x
max
from x =0.52 to 0.75 could be achieved [36,52,53].
The common concept of these advances in the application of (R)-
HiPIMS technology is built around a precise intersection between a
detailed knowledge of the time- and energy-resolved ion distribution
originating in the discharge and the structure-property relation of the
deposited lm. In this work, we present detailed results on the effect of
altered deposition parameters on the ion ux distribution arriving at the
substrate during R-HiPIMS deposition of Ti
1-x
Al
x
N thin lms by means of
time- and energy-resolved ion mass spectroscopy. Different discharge
conditions involving an increasing Al content in the powder-
metallurgically prepared Ti
1-x
Al
x
target (x =0.4 and 0.6) as well as
changing nitrogen ow rate ratios are employed. Variations observed in
the average ion count fractions as well as the temporal sequence of the
arriving ions are related to structural changes observed within the
synthesised thin lms. Furthermore, the thin lms are analysed in more
detail with respect to changes in morphology and local phase formation.
Finally, the obtained results are discussed in relation to the mechanical
properties of all Ti
1-x
Al
x
N thin lms investigated.
2. Experimental
Mass spectroscopy measurements were performed to investigate the
composition, the energy, as well as the time-resolved sequence of ions
incident on the substrate plane during HiPIMS sputtering of Ti
1-x
Al
x
composite targets in various nitrogen containing atmospheres. Two
different target compositions (Ti
0.6
Al
0.4
and Ti
0.4
Al
0.6
; powder-
metallurgically prepared, Plansee Composite Materials GmbH) were
operated in varying N
2
/Ar-mixtures, with the nitrogen ow-rate ratio
fnorm
[N2]=f
[N2]
/(f
[N2]
+f
[Ar]
) increased from 0, to 0.23, to 0.3, and 1. Prior
to all measurements, the cylindrical vacuum system was pumped to a
base pressure below 310
4
Pa using a diffusion pump, while the total
process gas pressure (p) was xed to p =0.4 Pa during magnetron
operation. The system was equipped with a 4-in. cathode (Vtech, Gencoa
Ltd), allowing for an in-situ control of the magnetic eld conguration
on the Ti
1-x
Al
x
targets. Since the mass spectroscopy measurements could
not be performed in the same chamber used for the subsequent depo-
sition of Ti
1-x
Al
x
N thin lms, the magnetic eld strength had to be
adjusted to reproduce the waveforms of the target voltage U
T
(t) and the
discharge current I
D
(t) observed during the deposition.
The magnetron was powered by a 5 kW plasma generator (ADL
GmbH), with the HiPIMS signal modulated using a SIPP2000USB pulse
power controller (Melec GmbH). For all measurements, the repetition
frequency (f) and pulse duration (t
on
) were maintained at 500 Hz and 75
μ
s, respectively, with the corresponding duty cycle t
on
/T =3.75 %,
where the pulse period T equals 1/f. The waveforms of the magnetron
voltage and the discharge current were recorded on a digital oscillo-
scope (PicoScope 6403C, Pico Technology) utilising a voltage-(Testec
TT-HV 150) and current-(Tektronix TCP303) probe. The average target
power density in a pulse period was evaluated as
P=1
ATTT
0
UT(t)ID(t)dt
where A
T
represents the total target area (~ 78.5 cm
2
). The discharge
peak power density within a period was calculated as
Ppk =max(UT(t)ID(t)
AT)
with U
T,pk
and I
D,pk
denoting the target peak voltage and discharge peak
current, respectively. In the presented experiments, the average power
density was maintained constant at P ~ 10.25 W/cm
2
, resulting in peak
power densities ranging from P
pk
~ 0.5 to 1 kW/cm
2
, depending on
f
[N2]ow
used. Moreover, both the employed process gas as well as the
HiPIMS discharge parameters were based on our recently published
work [20].
Time-averaged ion energy distributions of positive ions in the
discharge plasma were measured using an energy-resolved mass spec-
trometer (EQP 300, Hiden Analytical) with the sampling orice placed
parallel to the target surface at a xed distance of d =110 mm. A shutter
construction was further placed in front of the mass spectrometer orice,
to allow for stable discharge conditions prior to all measurements per-
formed. The mass spectrometer was tuned to
40
Ar
+
ions and the ob-
tained setting was kept xed for all measurements. Standard time-
averaged acquisitions of ion spectra were measured for
40
Ar
+
,
27
Al
+
,
27
Al
2+
,
14
N
+
,
28
N
2+
,
48
Ti
+
and
48
Ti
2+
ions. The extractor voltage was set
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
3
to 10 V with respect to ground potential and the electrode controlling
the ion energy was scanned from 5 to 80 V at a step size of 0.1 V
ensuring that practically all ions (with different energies) were recorded
during the measurement. The dwell time for ion detection at each point
was 10 ms. The ion energy distribution for each species was recorded
during ve consecutive scans to increase the signal-to-noise ratio and to
identify any potential long-term changes in the discharge conditions
during data acquisition. Assuming consistent sensitivity of the instru-
ment to ions at various energies, the total ux of all species was calcu-
lated by integrating the corresponding time-averaged energy
distribution over the full energy range. The recorded counts for the
48
Ti
n+
species were additionally corrected for natural isotope abundance
(
48
Ti ~ 73.72 % of all stable isotopes [54]). Moreover,
40
Ar
+
and
48
Ti
n+
species were corrected for system transmission using the manufacturers
calibration data (transmittance of
40
Ar
+
~ 86.17 % and
48
Ti
+
~ 76.02
%, respectively [55]). Also, no indications for the molecular species
62
TiN
+
and
41
AlN
+
were found, hence their uxes into the mass spec-
trometer are likely several orders of magnitude lower compared to the
uxes of metal ions. With the assumption of an identical transmission
characteristic of the instrument for all species when implementing the
corrections, the composition of the total ux of ions onto the substrate
surface was calculated on the basis of these integral uxes determined.
In addition, to obtain the temporal sequence of ions arriving at the
substrate, time-resolved ion counts were recorded for specic HiPIMS
discharge conditions. In that case, the mass spectrometer was set to
continuously scan for one ion species in a narrow energy range (typically
20 eV) corresponding to the signal peak in the ion energy distribution
function with a dwell time of 100 ms. The raw pulse-stream output of the
ion detector was fed to a multichannel scaler (SR430, Stanford Research
Systems). This device registered the incoming pulses and accumulated
them successively in evenly spaced time intervals once a trigger signal
for the negative voltage pulse of the HiPIMS power supply was regis-
tered. Thus, the time of ion arrival onto the detector with respect to the
beginning of the negative pulse was discriminated with a time resolution
of 1.28
μ
s (width of the accumulation time interval). A correction for the
time-of-ight of ions in the mass spectrometer was applied during post-
processing of the data [55,56]. Several tens of thousands of pulse periods
were accumulated to increase the signal-to-noise ratio.
To analyse the inuence of the recorded ion spectra on the thin lm
properties, Ti
1-x
Al
x
N coatings were synthesised using a separate, in-
house developed magnetron sputtering system holding two 6-in. cath-
odes equipped with identical Ti
1-x
Al
x
composite targets as were used
during the mass spectroscopy studies. The magnetrons (large circular,
Gencoa Ltd.) were individually powered by a 5 kW HiPIMS plasma
generator (HIP3, Solvix). The system uses a confocal, bottom-up
conguration for the two cathodes resulting in an included angle of
α
=20between the target and substrate normal. All depositions were
carried out with a xed distance of h =110 mm between the rotating
substrate holder (0.25 Hz) and the target surface in alignment with the
mass spectroscopy analysis. Further details on the deposition system can
be extracted from [20].
Before each deposition, a base pressure below 310
4
Pa was estab-
lished. The Ti
1-x
Al
x
N thin lms were deposited on Si platelets ((100)-
oriented, 20 ×7 ×0.38 mm
3
), monocrystalline Al
2
O
3
platelets
((1102)-oriented, 10 ×10 ×0.53 mm
3
), as well as polished austenite
platelets (DIN EN 1.4571, 20 ×7 ×0.8 mm
3
). Prior to all depositions,
the substrate materials were pre-cleaned in an ultrasonic bath using
acetone and isopropyl alcohol, consecutively. A 30 min heating
sequence to a substrate temperature of T
S
=300 C (measured directly
on the substrate holder, corresponds to a heater temperature of T
H
=
500 C) was followed by a 10 min Ar-ion etching step conducted at a
process pressure of p
etch
=3 Pa and a substrate potential of U
S
= − 1000
V. The discharge parameters and overall process conditions (i.e., pulse
frequency and duration, average power density, deposition pressure,
nitrogen ow-rate ratios) for all Ti
1-x
Al
x
N depositions were selected
according to the settings used for the mass spectroscopy measurements.
Details of the deposition conditions and coating properties are sum-
marised in Table A1 of Appendix A.
Investigations on the coating structure were performed using X-ray
diffraction (XRD) analysis on a PANalytical XPert Pro MPD system
equipped with a Cu-K
α
radiation source (wave length λ =1.54 Å) in
Bragg-Brentano geometry. The chemical composition of all Ti
1-x
Al
x
N
thin lms was characterised utilising energy dispersive X-ray spectros-
copy (EDS) in top-view conguration (EDAX EDS detector, 10 kV ac-
celeration voltage). Furthermore, scanning electron microscopy (SEM,
FEI Quanta 200, operated at 10 kV) was employed to investigate the
coating morphology based on fracture cross-sections of single-side
coated Si substrates. Additional analysis on local phase formation as
well as growth characteristics of selected Ti
1-x
Al
x
N thin lms was ob-
tained by transmission electron microscopy (TEM, FEI TECNAI F20)
complemented by selected area electron diffraction (SAED). The eval-
uation of SAED diffractograms was conducted using the CrysTBox soft-
ware package [57].
The mechanical properties including both the indentation hardness
(H) and modulus (E) were characterised using an ultra-micro indenta-
tion system (UMIS) equipped with a Berkovich diamond indenter. For
each coating, 30 load-displacement curves were recorded with the
indentation load varied between 6 and 45 mN. The evaluation of the
collected data was conducted according to the Oliver and Pharr method
[58]. Residual macro-stresses within the Ti
1-x
Al
x
N coatings were further
calculated from the modied Stoney-equation combined with curvature
measurements of single-side coated Si substrates obtained from optical
prolometry (PS50, Nanovea) [59].
0
25
50
75
100
125
150
Discharge current, I
D
(A)
Ti0.4Al0.6
f
norm
[N
2
]
=0
f
norm
[N
2
]
=0.3
f
norm
[N
2
]
=0.23
f
norm
[N
2
]
=1
(a)
(b)
0 255075100125
-0.8
-0.6
-0.4
-0.2
0.0
Magnetron voltage, U
T
(kV)
Time (µs)
75 80 85
0.0
0.2
0.4
0.6
0.8
(
c
)
Fig. 1. Time-evolution of (a) the discharge current I
D
(t) and (b) the magnetron
voltage on a Ti
0.4
Al
0.6
target as functions of the applied nitrogen ow rate ratio
fnorm
[N2]for a pre-set pulse duration of 75
μ
s. Insert (c) shows details of the cor-
responding positive voltage overshoot recorded at the end of each pulse.
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
4
3. Results & discussion
3.1. Discharge characteristics
Fig. 1 shows the time-evolution of the discharge current I
D
(t) and the
magnetron voltage U
T
(t) recorded during HiPIMS sputtering of a
Ti
0.4
Al
0.6
target in mixed N
2
/Ar-atmospheres as functions of time.
During all measurements, the discharge parameters were kept constant
at f =500 Hz, t
on
=75
μ
s, and P ~ 10.25 W/cm
2
, respectively, while
fnorm
[N2]was increased from 0 to 1. Irrespective of the nitrogen content in
the atmosphere, the discharge current increases over time, reaching a
respective maximum at the end of the pulse. When increasing the ni-
trogen ow-rate ratio from 0 to 0.23, the discharge current shifts from a
logistic to an exponential growth until voltage shutdown at t =75
μ
s,
reaching values of I
D,pk
=67 and 103 A, respectively (see Fig. 1a). Upon
further increasing fnorm
[N2]to 0.3 and 1, the exponential characteristic is
even more pronounced, with I
D,pk
increasing up to 135 A for the highest
nitrogen content (corresponds to a peak current density of 1.72 A/cm
2
).
Concurrently, by increasing the nitrogen ow-rate ratio, the onset of the
current rise becomes retarded and the initial slope of the waveforms is
observed to be less steep. Moreover, the uniform evolution of the current
waveforms suggests that no distinct contribution through working gas
recycling and/or self-sputter recycling indicated by a second
maximum in I
D
(t) occurs at higher peak current densities [7]. Since the
average energy imparted into each pulse (Ep=P/f) was constant for all
experiments, the increasing peak current densities can be clearly
attributed to the altered nitrogen ow-rate ratio, consequently shifting
the Ti
1-x
Al
x
composite target from a metallic to a more poisoned sput-
tering mode. Considering that the emission of secondary electrons is
signicantly amplied for AlN, and only marginally lowered for TiN,
when compared to their pure metallic states, the increasing peak current
densities can be explained by the progressive formation of surface ni-
trides upon increasing fnorm
[N2][60]. Moreover, self-sputtering of the
nitrided target surface by energetic N
+
-ions at high values of fnorm
[N2]
additionally supports the emission of secondary electrons [61].
Following typical oscillations after the discharge ignition, the
magnetron voltage is almost constant throughout the pulse for all ni-
trogen variations conducted, resembling an almost ideal rectangular
signal with U
T
~ 600 V (see Fig. 1b). The magnetron shutdown at t =
75
μ
s is accompanied by a steep decrease in the applied potential, fol-
lowed by a 35
μ
s long positive overshootof the magnetron voltage of
up to +800 V for fnorm
[N2]=1 (see insert c). This application of a positive
voltage after the HiPIMS pulse derives from the inherent inductance of
the power supply used, where the voltage amplitude can be estimated by
the following expression:
V+LdID,pk
dt
Through this almost instantaneous inversion in magnetron polarity
and the resulting difference in plasma and substrate potential, ions
located in the vicinity of the target get accelerated towards the substrate
plane at energies corresponding to the positive pulse voltage [12].
The measurements were repeated for a Ti
0.6
Al
0.4
target (see Fig. A1
in Appendix A) with an identical stepwise increase of the nitrogen ow-
rate ratio from 0 to 1, revealing analogous results for both the evolution
of the discharge current and magnetron voltage. For a pure Ar-
atmosphere, a similar peak current of ~70 A is recorded at the end of
the pulse. However, upon increasing fnorm
[N2]to 0.23 and further, no in-
termediate target poisoning state can be observed, instead the discharge
current immediately saturates at peak values around 135 A. Hence, it
can be concluded that upon increasing the Ti/Al-ratio within the target
material, less nitrogen will be required to transfer the target to a fully
poisoned state under the same HiPIMS conditions, being in perfect
agreement with previous observations for DCMS. Moreover, both the
magnetron voltage as well as the positive overshoot after the pulse
were recorded at similar values with U
T
~ 600 V and V
+max
~ 800 V,
when compared to the Ti
0.4
Al
0.6
target and the corresponding nitrogen
ow-rate ratios.
3.2. Time-averaged composition of the total ion ux
In Fig. 2, the individual fractions of the ionic species as well as the
overall composition of the total ion ux measured at a distance of d =
110 mm from the magnetron source is presented for all nitrogen ow-
rate ratios applied for both Ti
1-x
Al
x
target compositions. The obtained
results provide qualitative information on the fractional changes in the
total ion ux arriving at the substrate surface with respect to the ni-
trogen ow-rate variations. Here it should be noted, that the contribu-
tions from Al
2+
- and N
+
-ions cannot be separated due to an intensity
overlap in the recorded spectra (m/e =13.5 and 14, respectively).
However, experiments in pure Ar atmosphere showed that Al
2+
-signals
are insignicant compared to those from Al
+
-ions, hence they were not
considered in the evaluation even at higher peak current densities with
increasing fnorm
[N2]. This is in agreement with previous observations for
HiPIMS powered Al targets and mainly related to the considerably
0
10
20
30
40
50
60
70
80
90
100
Ion count fraction (%)
N
+
N
+
2
Ar
+
Al
+
Ti
+
Ti
2+
(a) Ti
0.4
Al
0.6
J
Me
+
J
g
+
0.0 0.2 0.4 0.6 0.8 1.0
0
10
20
30
40
50
60
70
80
90
100
N
2
flow-rate ratio, f
norm
[N
2
]
(-)
Ion count fraction (%)
Ar
+
Al
+
Ti
+
Ti
2+
N
+
N
+
2
(b) Ti
0.6
Al
0.4
J
Me
+
J
g
+
Fig. 2. Integral ion count fractions of the ionic species arriving at the substrate
plane as function of the nitrogen ow rate ratio f
norm
[N2], presented for (a) a
Ti
0.4
Al
0.6
and (b) a Ti
0.6
Al
0.4
target, respectively. The data was acquired at a
distance of d =110 mm from the target surface. Singly charged species are
presented by lled symbols and solid lines, whereas doubly charged ions are
presented by open symbols and dashed lines. In addition, (a) and (b) contain the
total ux of process gas ions J
g
+
and metal ions J
Me
+
, indicated by half-lled
circles and dot dashed lines.
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
5
higher secondary ionisation energy of aluminium (IP
2Al
=18.83 eV)
compared to the primary ionisation energy of argon (IP
1Ar
=15.76 eV)
[34,54]. In addition, Fig. 3 presents a direct comparison of the ux
compositions recorded at fnorm
[N2]=0.3, for both the Ti
0.4
Al
0.6
(blue, lled
bars) and Ti
0.6
Al
0.4
(red, dashed bars) target.
Over the entire range of fnorm
[N2]considered, the ux of process gas ions
J
g
+
(i.e., the combined ux of Ar
+
, N
2+
, and N
+
ions) dominates the
composition of the total ion ux emitted from the plasma discharge. In
pure argon-atmosphere, the integral ux of metal ions J
Me
+
(i.e., the
combined ux of Al
+
, Ti
+
, and Ti
2+
ions) constitutes 36.3 and 29.6 %
within the total ionic species ejected from the Ti
0.4
Al
0.6
and Ti
0.6
Al
0.4
target, respectively. Upon stepwise increase to pure nitrogen-
atmosphere at fnorm
[N2]=1, the corresponding values for J
Me
+
decrease
almost linearly to 15.4 and 15.6 %, respectively. This reduction in the
metal-to-gas ion ux ratio J
Me
+
/J
g
+
from 0.57 to 0.18 for the Ti
0.4
Al
0.6
target and from 0.42 to 0.18 for the Ti
0.6
Al
0.4
target with increasing
nitrogen ow results from a progressive nitride formation on the target
surface, and hence a decreasing sputter yield of the metal constituents.
In addition, the higher overall sputter yield of the Ti
0.4
Al
0.6
target
compared to the Ti
0.6
Al
0.4
target the sputter yield of Al is nearly twice
that of Ti [62] results in higher values of J
Me
+
throughout all nitrogen
ow rates.
The poisoning behaviour of the target material is of course strongly
related to the Ti/Al-ratio and driven by the difference in the formation
energy of the respective nitrides (305.6 kJ/mol for TiN vs. -241.6 kJ/
mol for AlN) [46,63]. This implies that the higher Al-containing
Ti
0.4
Al
0.6
target can sustain the metallic sputtering behaviour up to
higher values of fnorm
[N2]than the Ti
0.6
Al
0.4
target, as seen in the current
waveforms presented in Figs. 1 and A1. Accordingly, for fnorm
[N2]=0.23 and
0.3 the data also show an increased population of N
2+
and N
+
ions for
the Ti
0.6
Al
0.4
target due to less absorption of reactive gas (i.e., a more
poisoned target for the same N
2
ow rate), leading to a higher partial
pressure and thus ionisation probability for the nitrogen species (see
Figs. 2 and 3). In addition, for both target-types a predominance of N
+
-
ions (dark-yellow, lled triangles) over N
2+
-ions (light-yellow, open
triangles) is observed in all ux compositions. This highlights the
preferred ionisation of atomic N sputtered from the nitrided target
surface as well as the promoted dissociation of N
2
-molecules within the
dense plasma created for high values of P
pk
close to 1 kW/cm
2
(see
Fig. 2).
Regarding the individual contribution of metal ions generated in the
discharge, the data presented in Fig. 2a and b shows that Al
+
-ions (red
diamonds) account for a major fraction within J
Me
+
, irrespective of the
Al/Ti-ratio in the target material or fnorm
[N2]. Especially for the Ti
0.4
Al
0.6
target, high values of J
Al
+
up to 23 % of the total ion ux are observed in
pure Ar-atmosphere (see Fig. 2a). Increasing the nitrogen ow to fnorm
[N2]=
0.23 and 0.3 being typical values for the deposition of Ti-Al-N thin
lms the fraction of Al
+
-ions in the sputtered ux decreases slightly to
J
Al
+
~ 18 %. Following a near linear trend, this value further decreases
down to J
Al
+
~ 8.1 % for fnorm
[N2]=1. Contrary, uxes for singly (dark-blue
squares) as well as doubly (light-blue open squares) charged titanium
ions remain almost unaffected by the N
2
ow-rate ratio, with J
Ti
+
and
J
Ti
2+
accounting for values below 14 % and 3 % within the total ionic ux
from both targets, respectively.
Upon further comparison of J
Al
+
to the combined ux of process gas
ions J
g
+
for both target chemistries, the increased ionisation of Al-atoms
within the Ti
0.4
Al
0.6
discharge is predominantly realised at the expense
of a less efcient ionisation of the process gas species (see Figs. 2 and 3).
This effect could be explained by an enhanced gas rarefaction occurring
in front of the higher Al-containing target due to an increased density of
metal species entering the plasma mind the difference in sputter yield
Al vs. Ti thus resulting in a higher probability for atomic collisions and
gas heating [16]. In addition, the signicantly lower ionisation energy of
aluminium atoms (IP
1Al
=5.99 eV [54]) over the gaseous species (IP
1Ar
=15.76 eV, IP
1N
=14.53 eV, and IP
1N2
=15.58 eV [54]) results in
quenching of the electron energy distribution, thereby further contrib-
uting to a less efcient ionisation of process gas atoms [29,35]. This
explanation is further underlined when reconsidering the evolution of
both J
Ti
+
and J
Ti
2+
between the two target compositions, revealing
negligible differences despite the signicant changes in J
Me
+
/J
g
+
over the
entire N
2
ow range considered.
Consequently, it can be concluded that increasing the Al content in a
Ti
1-x
Al
x
composite target can be utilised for amplifying the fraction of
Al
+
ions present in the sputtered ux, while reducing the overall fraction
of process gas ions. Although a high Al/Ti ratio in a composite target also
has to be considered with respect to phase formation and general solu-
bility limits for the deposition of cubic structured Ti
1-x
Al
x
N thin lms,
this provides a means of tuning J
Al
+
while maintaining J
Ti
+
, and espe-
cially the detrimental J
Ti
2+
, low.
3.3. Time-resolved composition of the total ion ux
In order to study the temporal sequence of ions impinging on the
substrate plane especially from the aspect of time domains with high
values of J
Me
+
/J
g
+
time-resolved ion mass spectroscopy was performed
Fig. 3. Integral ion count fractions of the total ionic ux emitted from a HiPIMS
discharge for a nitrogen ow rate ratio of f
norm
[N2]=0.3, presented for both a
Ti
0.4
Al
0.6
(blue, lled bars), as well as a Ti
0.6
Al
0.4
(red, dashed bars) target.
Fig. 4. Normalised time-evolution of the energy-integrated total ionic ux
emitted during HiPIMS sputtering of a Ti
0.4
Al
0.6
target, measured at fnorm
[N2]=0.3
and a distance of d =110 mm from the target. Singly charged species are
presented by solid lines, whereas doubly charged ions are presented by
dashed lines.
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
6
for a specic discharge condition. In Fig. 4, all major ion uxes (Ar
+
,
Al
+
, N
+
, N
2+
, Ti
+
and Ti
2+
) emitted from the Ti
0.4
Al
0.6
target were
recorded at the substrate position (d =110 mm) for a nitrogen ow-rate
ratio of fnorm
[N2]=0.3. The total number of ion-counts normalised to the
number of recorded pulse cycles is shown as a function of the time delay
with respect to the voltage application to the magnetron.
Regarding the ux chemistry prior to the pulse ignition (t <0
μ
s)
comprised of the residual ionized species from the previous pulse a
predominance of Ar
+
ions over the ux of N
2+
, N
+
, Al
+
, Ti
+
and Ti
2+
ions is observed, with the latter two species accounting for the lowest
fractions. Upon pulse ignition (t =0
μ
s), and thus a rapid increase in the
electron temperature T
e
, all ux signals especially the process gas ions
immediately gain intensity until reaching a rst maximum at t ~ 25
μ
s.
Since the ionic species generated in the target vicinity will require at
least several tens of
μ
s before arriving at the analyser, this initial inux
of ions corresponds to ions created close to the mass spectrometer orice
by the rst wave of energetic electrons [64]. In the following time in-
terval until the cathode shutdown at t =75
μ
s, the ion intensities in-
crease to their respective maximum due to the arrival of ions generated
within the main magnetron discharge to approximately one order of
magnitude higher than the corresponding initial values at t =0
μ
s
coinciding well with I
D,pk
(see Fig. 1). Interestingly, the uxes of metal
ions J
Me
+
and N
+
-ions show a concomitant increase to their respective
maximum at t ~ 60
μ
s, indicating that the latter species originates
strongly from sputtering of target surface nitrides. Still, it should be
noted that the highly dense plasma additionally contributes to the
dissociation of N
2
-molecules, thus explaining the extended N
+
peak.
Contrary, uxes for Ar
+
- and N
2+
-ions show a slightly delayed arrival
with an intensity maximum at t =75
μ
s, suggesting for the ionisation to
occur preferably in the bulk plasma.
After reaching the peak ion ux at the end of the 75
μ
s pulse, an
abrupt decay of all ion signals can be observed, suggesting for a decrease
in the number of arriving ions. However, this loss of ion intensity is due
to a temporal broadening of the ion energy distribution function (IEDF)
due to the positive voltage overshoot occurring on the magnetron at t ~
78
μ
s (see insert c in Fig. 1, U
T
~ 500 V). The application of this positive
pulse accelerates ions in the target proximity towards the analyser and
the bulk plasma potential increases signicantly, as it is always more
positively charged than the most positive electrode within the plasma
represented by the magnetron during this time frame. Consequently, all
ions obtain kinetic energies beyond the respective scanning range by this
instantaneous difference between plasma and mass spectrometer po-
tential, causing the apparent dip in the ion count evolution. This effect of
a positive pulse on the IEDF is also described in more detail in recent
studies by Kozak et al. [11] and Santiago et al. [12]. So in fact, the
presented ion counts can be rather understood as obtaining a smooth
decay between the shown maxima at t =75 and 110
μ
s for all species.
Finally, in the time domain of t >80
μ
s (U
T
~ 0 V), the IEDFs gradually
collapse, resulting in the recorded ion counts to slightly regain intensity
up to t ~ 110
μ
s, to then decay at an equal rate in the following 150
μ
s
until the initial state between the HiPIMS pulses is attained.
A further interesting observation when considering the collective
evolution of all ion signals is that the relative ux composition is pre-
served throughout the entire pulse duration, with J
Me
+
/J
g
+
remaining
fairly unchanged except for a slight dip at t ~ 35
μ
s due to the large
inux of initial Ar
+
ions. Hence, no time domain obtaining a preferred
arrival of either metal or process gas ions can be identied resulting
from well-known effects such as process gas rarefaction or quenching of
T
e
[16,28,35]. This continuous arrival of J
Me
+
and J
g
+
has direct impli-
cations on the deposition process, thereby minimising the controllability
of lm growth kinetics through synchronised metal-ion irradiation when
using a Ti
1-x
Al
x
composite target in a reactive HiPIMS process. More-
over, analogous to the results obtained for the integral ion ux (see
Figs. 2 and 3), both J
Ti
+
and J
Ti
2+
constitute the lowest fractions during
the entire pulse period with respect to the total ionic ux, and in
particular J
Al
+
. This is especially interesting, when taking the moderate
rst ionisation energy of titanium (IP
Ti
=6.83 eV) which is close to that
of Al, yet signicantly lower when compared to the present process gas
species as well as the almost equiatomic target composition into ac-
count [54]. Nevertheless, the presented results clearly underline our
previous observations outlined in Ref. [20], where synchronising the
bias potential to the HiPIMS pulse showed only minor effects on the
phase formation of cubic structured Ti-Al-N.
3.4. Phase formation and chemical composition
Fig. 5a and b containing the XRD diffractograms of lms deposited
from Ti
0.6
Al
0.4
or Ti
0.4
Al
0.6
targets, respectively show the inuence of
different target chemistries and the nitrogen ow-rate ratio fnorm
[N2]on the
cubic phase stability of Ti
1-x
Al
x
N. In addition, Fig. 5c and d contain the
corresponding chemical composition to the coatings presented in Fig. 5a
and b, with square data points indicating the Al/(Ti +Al) ratio on the
metal sublattice (lower axis) and circular data points denoting the ni-
trogen content (upper axis).
For a Ti
0.6
Al
0.4
target chemistry (Fig. 5a and c), altering fnorm
[N2]from
0.23 to 1.0 results in a slight decrease in the Al-content from x =0.46 to
0.42 of the purely fcc-structured Ti
1-x
Al
x
N thin lms. This decrease can
be understood based on the previously discussed difference in the heat of
Fig. 5. (a) and (c) show XRD diffractograms of Ti
1-x
Al
x
N thin lms deposited
from two 6-in. Ti
0.6
Al
0.4
and Ti
0.4
Al
0.6
cathodes, respectively, arranged with
increasing fnorm
[N2]from bottom (0.23) to top (1.0). The 2θ peak positions for cubic
structured TiN (open square, [69]) and AlN (lled square, [70]), wurtzite AlN
(open hexagon, [65]) and the substrate material (cubic Si, half-lled triangle,
[71]) are indicated in (a) and (c). Furthermore, (b) and (d) show the chemical
composition of all Ti
1-x
Al
x
N thin lms presented in (a) and (c), respectively
bottom axis Al/Me-ratio (red squares) and top axis N-content (blue circles)
with the data points aligned horizontally with the corresponding XRD
diffractogram.
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
7
formation for TiN and AlN (see Section 2), the Al target particles
commonly show a retarded poisoning state over the Ti particles, thus
leading to a preferential Al sputter erosion at lower nitrogen ow rates
[46,63]. Also, the N-content on the non-metal sublattice increases
accordingly from 52 to 55 at.%, due to the higher availability of N
+
- and
N
2+
-ions in the plasma (see Fig. 2b) as well as the concomitant decrease
in deposition rate at higher values of fnorm
[N2]the latter decreasing from
22.1 to 17.4 to 10.8 nm/min. Moreover, with increasing reactive gas
ow fnorm
[N2]from 0.23 to 1, the XRD diffractograms present a transition
from a preferred (111)-oriented growth, to a randomly oriented growth
with diminishing peak intensities (i.e., smaller coherently diffracting
domain sizes), towards a slightly preferred (200)-oriented growth.
In the case of the higher Al containing target Ti
0.4
Al
0.6
(see Fig. 5b),
the Al metal-fraction of the Ti
1-x
Al
x
N thin lms is x =0.59 for a low and
high fnorm
[N2]of 0.23 and 1, while it increases to x =0.63 at fnorm
[N2]=0.3 (see
Fig. 5d). Also, this behaviour is the result of selective poisoning pro-
cesses of the Ti-Al target. With increasing fnorm
[N2]from 0.23 to 0.30, the
sputter yield of Ti decreases faster than that of Al (as Ti is more afne to
N than Al and thus poisons earlier). However, if fnorm
[N2]is sufciently high
to also poison Al, the reduction in sputter yield is more severe for Al due
to the electrically isolating nature of AlN. The XRD diffractogram of the
highest Al-containing coating (x =0.63), prepared at fnorm
[N2]=0.3, depicts
a predominantly face-centered cubic structure having a preferred (111)-
oriented growth, with slight indications for an additional wurtzite AlN-
based phase (see the XRD peaks at 2θ ~ 32.2and 35). The offset of
these peaks from the ideal w-AlN reference pattern [65] hints towards
the formation of a w-Al
x
Ti
1-x
N solid solution [44,47]. Reducing fnorm
[N2]
from 0.3 to 0.23 promotes the growth of a highly (111)-oriented cubic
phase (even leading to the formation of the K
β1
diffraction spot at 2θ ~
34). This c-Ti
0.41
Al
0.59
N thin lm is understoichiometric with a nitro-
gen content close to 40 at.%. The diffractogram further shows a clear
shift towards the c-AlN reference line despite an actual reduction in the
aluminium content indicative of a reduced compressive stress state and
nitrogen deciency. This drastic transition in lm growth characteristics
is in excellent agreement with previous studies [20,46] and follows the
suggestions by Gall et al. [66], that a reduced availability of excess ni-
trogen (see Fig. 2a) yields a preferred growth in the (111)-direction. An
increase of fnorm
[N2]to 1 results in a multi-phased coating of (200)-oriented
c-Ti
1-x
Al
x
N and an almost X-ray amorphous w-Al
x
Ti
1-x
N. Analogous to
the coatings prepared from the lower Al containing target, increasing
fnorm
[N2]drastically reduces the deposition rate from 66.2 to 50.6 to 10.9
nm/min. This allows for a less disturbed surface diffusion and provides
more time for the growing lm to approach the thermodynamically
favoured crystal structures (cubic for Ti-rich and wurtzite for Al-rich).
Moreover, the increased bombardment with N
+
/N
2+
-ions (see Fig. 2a)
additionally contributes to the wurtzite phase growth, hence the
reduction in cubic phase fraction when increasing fnorm
[N2]from 0.3 to 1.0
despite the concomitant reduction of the Al/(Ti +Al) ratio [46]. The
competitive growth between c-TiN and the preferred w-AlN phase then
results in a nano-composite structure [20]. Overall, the coatings pre-
pared with fnorm
[N2]=0.23 and 0.3 (Fig. 5b) provide signicantly higher
diffracted peak intensities when compared with the corresponding lms
prepared from the Ti-richer Ti
0.6
Al
0.4
target (Fig. 5a), indicating altered
lm growth pathways when increasing the Al content in the target.
Taking the ion count fractions presented in Fig. 3 into account, the
increased bombardment with medium energetic Al
+
-ions in favour of
less Ti
n+
- and N
+
/N
2+
- species in the lm forming ux contributes to
enhanced surface diffusion and thus larger coherently diffracting
domain sizes, which is also in good agreement with previous observa-
tions [20,49].
3.5. Growth morphology and phase formation
The Ti
1-x
Al
x
N thin lms were investigated in more detail with respect
to growth morphology and phase formation by means of cross-sectional
SEM and TEM, see Fig. 6. The fracture cross-sectional SEM image of
Ti
0.54
Al
0.46
N (Fig. 6a) shows a dense and highly columnar growth
morphology, which is also underlined by the bright- (BF) and dark-eld
(DF) TEM investigations (Fig. 6b). Furthermore, the corresponding
SAED studies taken in the interface near region (Fig. 6c1) as well as close
to the lm surface (Fig. 6c2) conrm the polycrystalline face-centred
cubic crystal structure observed by XRD (Fig. 5a, pattern in the mid-
dle) i.e., with no preferred orientation by ring-type diffraction pat-
terns throughout the coating thickness. Investigations of the Ti
1-x
Al
x
N
coatings synthesised from the Al-richer Ti
0.4
Al
0.6
target (Fig. 6d till l)
nicely depict the inuence of the increasing fnorm
[N2]on the growth char-
acteristics of these higher Al containing thin lms. Both, the SEM frac-
ture cross-sections as well as the corresponding BF- and DF-TEM images
indicate the transition from a highly oriented Ti
0.41
Al
0.59
N coating (N
content of ~40 at.%) with large columnar crystals at fnorm
[N2]=0.23 (see
Fig. 6d, e, and f), to a slightly less preferred-oriented Ti
0.37
Al
0.63
N
coating at fnorm
[N2]=0.3 with shorter columnar crystals (see Fig. 6g, h and
i). A further increase in fnorm
[N2]to 1.0 leads to the formation of an even
ner microstructure, exhibiting a featureless morphology during SEM
(Fig. 6j) with small-sized slightly elongated grains (Fig. 6k). The selected
area electron diffraction patterns in Fig. 6f1 and f2 recorded for the
Ti
0.41
Al
0.59
N coating at fnorm
[N2]=0.23 conrm the cubic crystal structure
suggested by XRD (despite the N sub-stoichiometry) and show the highly
crystalline nature of this coating through distinct point type diffractions,
especially during later growth stages (see inset f2). Interestingly, the
SAED investigations of the Ti
0.37
Al
0.63
N coating prepared with fnorm
[N2]=
0.3 show that the wurtzite Al
x
Ti
1-x
N phase as also observed during
XRD, Fig. 5c merely prevails in the interface near region (Fig. 6i1),
whereas the main part of the coating is essentially cubic structured
(Fig. 6i2). Since the deposition temperature was chosen rather low at T
S
~ 300 C, this effect could be seen in progressive substrate heating due
to the interaction with the dense HiPIMS plasma combined with
enhanced ion bombardment. Certainly, these results show that the cubic
Ti
1-x
Al
x
N structure is accessible from Ti
1-x
Al
x
compound targets even up
to x ~ 0.63 (i.e., close to the theoretical solubility limit) and despite the
presence of doubly charged Ti-ions in the lm forming ux. Again, in
accordance with the XRD investigations (Fig. 5c), detailed SAED ana-
lyses of the coating synthesised with fnorm
[N2]=1.0 underline the formation
of cubic and wurtzite phases (see Fig. 6 l1-l2).
3.6. Mechanical properties
Finally, in addition to the structural and morphological changes
discussed from the aspect of changing deposition and HiPIMS discharge
conditions, the consequential inuence on the mechanical properties of
all Ti
1-x
Al
x
N thin lms was studied. Fig. 7a-c summarize the biaxial re-
sidual stresses, hardness, and indentation modulus for the coatings
prepared from the two different targets as function of fnorm
[N2], respectively.
Single-phase cubic Ti
1-x
Al
x
N coatings on Si substrates sputtered using
Ti
0.6
Al
0.4
targets (red, lled squares) exhibit compressive stresses be-
tween 1.2 and 2.5 GPa, with the lowest value observed for the fully
polycrystalline Ti
0.53
Al
0.46
N when using fnorm
[N2]=0.3 (see Figs. 6ac and
7a). The corresponding nanoindentation measurements give high
hardness values, decreasing from H =36.4 ±1.5 to 32.6 ±2.5 GPa with
increasing fnorm
[N2], whereas the indentation modulus remains at E ~ 420
GPa almost unaffected by the nitrogen ow rate ratio variation, Fig. 7b
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
8
and c, respectively. Analogous to the ndings in our previous work, the
continuous reduction in hardness with increasing fnorm
[N2]can be attributed
to the increased fraction of nitrogen ions bombarding the lm surface
(see Fig. 2b) thus leading to a weaker grain boundary structure
[20,67]. Upon increasing the Al content in the target material to
Ti
0.4
Al
0.6
(see Fig. 7, blue, open squares) the obtained coatings exhibit
biaxial compressive stresses which increase from 0.4 to 2.5 GPa with
increasing fnorm
[N2]from 0.23 to 1.0, respectively. A maximum hardness of
36.1 ±3.5 GPa is obtained for the predominantly cubic structured
Ti
0.37
Al
0.63
N thin lm prepared at fnorm
[N2]=0.3. The minor wurtzite phase
content of this coating is essentially present only at the near substrate
region (Fig. 7i), thus its contribution to the hardness measurement can
be neglected. Whereas preparing the lm with fnorm
[N2]=0.23 results in an
understoichiometric phase, too much N
2
(fnorm
[N2]=1.0) favours the for-
mation of a signicant amount of the wurtzite phase. Both scenarios lead
to lower hardness (~ 29.5 GPa) and indentation modulus (380 ±10 and
300 ±9 GPa), Fig. 7b and c, respectively. Interestingly, when compared
with conventional DCMS or cathodic arc evaporation (CAE) techniques,
all Ti
1-x
Al
x
N coatings obtained here have relatively low compressive
stresses especially at fnorm
[N2]=0.3 which is likely linked to the large
fraction of low energetic lm forming ions bombarding the growing
coating surface, whilst the overall contribution from energetic Ti
2+
-ions
remains low (see Fig. 2). Consequently, this avoids the trapping of excess
Ar interstitial atoms and reduces the point defect density [68]. In gen-
eral, the presented evolution of the mechanical properties follows all
previous interpretations on these coatings, yet again highlighting the
excellent potential of R-HiPIMS deposited Ti
1-x
Al
x
N thin lms syn-
thesised from composite targets.
4. Conclusion
In this work, the correlation between time-averaged and time-
resolved ion uxes to the N
2
/Ar-ow-rate ratio utilised during reac-
tive HiPIMS deposition of Ti
1-x
Al
x
N thin lms from Ti
0.4
Al
0.6
or Ti
0.6
Al
0.4
composite targets is studied. Detailed ion mass spectroscopy in the time-
and energy-domain shows that for given discharge conditions the frac-
tion of Al
+
-ions is signicantly affected by the Al/(Ti +Al)-ratio of the
target material, whereas the amount of Ti
n+
-ions (n =1, 2) only shows a
minor correlation to the target composition despite the similar primary
ionisation energies for Ti and Al. Moreover, resulting from progressing
nitride formation, the data show a concomitant linear decrease for the
total metal-to-gas-ion ux ratio J
Me
+
/J
g
+
with increasing N
2
/Ar-ow-rate
ratio, reducing J
Me
+
from 36.3 and 29.6 % (fnorm
[N2]=0) down to 15.4 and
15.6 % (fnorm
[N2]=1) for the Ti
0.4
Al
0.6
and Ti
0.6
Al
0.4
target, respectively.
Interestingly, Ti
n+
-ion fractions remain almost unaffected by the applied
N
2
-ow-rate ratio, with the ux fractions of Ti
+
- and Ti
2+
-ions corre-
sponding to less than 14 and 3 % of the total ionic ux from both targets,
Fig. 6. (a), (d), (g), and (j) show SEM fracture cross-sections including the lm chemistry and f
norm
[N2]of selected Ti
1-x
Al
x
N thin lms from Fig. 5. (b), (e), (h), and (k)
present bright- and dark-eld TEM micrographs of the corresponding lm cross-sections depicted in (a), (d), (g), and (j). The respective SAED diffractograms taken in
the interface near region are shown in (c1), (f1), (i1), and (l1), whereas (c2), (f2), (i2), and (l2) contain SAED diffractograms recorded closer to the lm surface, as
indicated by the circles in (a), (d), (g), and (j). In addition, all SAED diffractograms contain an azimuthal integration of the pattern.
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
9
respectively. Additional time-resolved ion distributions recorded for
specic discharge conditions on the Ti
0.4
Al
0.6
target conrm the fairly
continuous arrival of both metal- and process-gas-ions throughout the
entire HiPIMS pulse cycle, thus hinting at the reduced effectiveness of
using synchronised bias potentials in tuning the lm growth kinetics.
Based on the recorded ion uxes, predominantly cubic-structured Ti
1-
x
Al
x
N thin lms with x
max
as high as 0.63 could be synthesised thor-
oughly proven by detailed structural analysis using XRD and TEM
exhibiting excellent mechanical properties (H up to ~36 GPa) at mod-
erate stress states. This is realised by an increasing fraction of low-
energy lm forming ions present in the sputterred ux, while the frac-
tion of energetic Ti
2+
ions remains low. Thereby, the probability for
trapped Ar interstitials and density of structural defects can be kept low.
Overall, the presented results show a clear pathway of controlling the
ion ux composition and thin lm structure during reactive HiPIMS
deposition of Ti
1-x
Al
x
N thin lms from composite targets through the use
of target-driven ion mass spectrosctopy.
CRediT authorship contribution statement
L. Zauner: Conceptualization, Investigation, Visualization, Writing
original draft. A. Bahr: Investigation, Writing review & editing. T.
Koz´
ak: Investigation, Writing review & editing. J. ˇ
Capek: Investiga-
tion, Writing review & editing. T. Wojcik: Investigation, Writing
review & editing. O. Hunold: Writing review & editing. S. Kolozsv´
ari:
Writing review & editing. P. Zeman: Writing review & editing. P.H.
Mayrhofer: Writing review & editing. H. Riedl: Supervision,
Conceptualization, Writing review & editing, Project administration.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgements
The nancial support by the Austrian Federal Ministry for Digital and
Economic Affairs, the National Foundation for Research, Technology
and Development and the Christian Doppler Research Association is
gratefully acknowledged (Christian Doppler Laboratory Surface Engi-
neering of high-performance Components). The ion mass spectroscopy
measurements done at University of West Bohemia were supported in
part by the project LO 1506 of the Czech Ministry of Education, Youth
and Sports under the program NPU I. We also thank for the nancial
support of Plansee SE, Plansee Composite Materials GmbH, and Oerlikon
Balzers, Oerlikon Surface Solutions AG. In addition, we want to thank
the X-ray center (XRC) of TU Wien for beam time as well as the electron
microscopy center - USTEM TU Wien - for providing the SEM and TEM
facilities. The authors acknowledge TU Wien Bibliothek for nancial
support through its Open Access Funding Programme.
Fig. 7. (a) Biaxial residual stress, (b) nanoindentation hardness, as well as (c) nanoindentation modulus of all Ti
1-x
Al
x
N thin lms deposited onto (100)-oriented
silicon substrates presented as function of fnorm
[N2]. Coatings deposited from Ti
0.6
Al
0.4
cathodes are denoted by red, lled squares, whereas those from Ti
0.4
Al
0.6
targets
are indicated by blue, open squares.
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
10
Appendix A
Table A1
Detailed overview of the deposition parameters, chemical composition and mechanical properties of the synthesised coatings.
Coating Target fnorm
[N2]p
dep
T
S
f t
on
U
S
Al/Me N H E
σ
dep. Rate
[] [Pa] [C] [Hz] [
μ
s] [V] [] [at.%] [GPa] [GPa] [GPa] [nm/min]
Ti
0.54
Al
0.46
N Ti
0.6
Al
0.4
0.23 0.4 300 500 75 50 0.46 52.5 36.4 ±1.6 435 ±10 2.4 ±0.019 22.1
Ti
0.54
Al
0.46
N 0.3 0.46 53 35.1 ±1.5 406 ±10 1.2 ±0.004 17.4
Ti
0.58
Al
0.42
N 1 0.42 55 32.6 ±2.6 417 ±16 2.0 ±0.001 10.8
Ti
0.41
Al
0.59
N Ti
0.4
Al
0.6
0.23 0.59 38.5 29.4 ±0.7 380 ±10 0.4 ±0.003 66.2
Ti
0.37
Al
0.63
N 0.3 0.63 51 36.1 ±3.5 541 ±35 1.3 ±0.011 50.6
Ti
0.41
Al
0.59
N 1 0.59 53 29.5 ±1.9 300 ±9 2.5 ±0.012 10.9
0
25
50
75
100
125
150
Discharge current, I
D
(A)
Ti0.6Al0.4
f
norm
[N
2
]
=0
f
norm
[N
2
]
=0.3
f
norm
[N
2
]
=0.23
f
norm
[N
2
]
=1
(a)
(b)
0 25 50 75 100 125
-0.8
-0.6
-0.4
-0.2
0.0
Magnetron voltage, U
T
(kV)
Time (µs)
75 80 85
0.0
0.2
0.4
0.6
0.8
(
c
)
Fig. A1. Time-evolution of (a) the discharge current I
D
(t) and (b) the magnetron voltage on a Ti
0.6
Al
0.4
target as functions of the applied nitrogen ow rate ratio fnorm
[N2]
for a pre-set pulse duration of 75
μ
s. Insert (c) shows details of the corresponding positive voltage overshoot recorded at the end of each pulse.
References
[1] D. Gall, C.S. Shin, T. Spila, M. Od´
en, M.J.H. Senna, J.E. Greene, I. Petrov, Growth
of single-crystal CrN on MgO(001): effects of low-energy ion-irradiation on surface
morphological evolution and physical properties, J. Appl. Phys. 91 (2002)
35893597, https://doi.org/10.1063/1.1446239.
[2] M. Lattemann, U. Helmersson, J.E. Greene, Fully dense, non-faceted 111-textured
high power impulse magnetron sputtering TiN lms grown in the absence of
substrate heating and bias, Thin Solid Films 518 (2010) 59785980, https://doi.
org/10.1016/j.tsf.2010.05.064.
[3] J.E. Greene, J.-E. Sundgren, L. Hultman, I. Petrov, D.B. Bergstrom, Development of
preferred orientation in polycrystalline TiN layers grown by ultrahigh vacuum
reactive magnetron sputtering ARTICLES YOU MAY BE INTERESTED IN, Appl.
Phys. Lett. 67 (1995) 2928, https://doi.org/10.1063/1.114845.
[4] I. Petrov, P.B. Barna, L. Hultman, J.E. Greene, Microstructural evolution during
lm growth, J. Vac. Sci. Technol. A Vacuum, Surfaces, Film. 21 (2003) S117S128,
https://doi.org/10.1116/1.1601610.
[5] V. Kouznetsov, K. Mac´
ak, J.M. Schneider, U. Helmersson, I. Petrov, A novel pulsed
magnetron sputter technique utilizing very high target power densities, Surf. Coat.
Technol. 122 (1999) 290293, https://doi.org/10.1016/S0257-8972(99)00292-3.
[6] J.T. Gudmundsson, N. Brenning, D. Lundin, U. Helmersson, High power impulse
magnetron sputtering discharge, J. Vac. Sci. Technol. A 30 (2012), https://doi.org/
10.1116/1.3691832, 030801.
[7] D. Lundin, J.T. Gudmundsson, T. Minea, High Power Impulse Magnetron
Sputtering, Elsevier, 2020, https://doi.org/10.1016/C2016-0-02463-4.
[8] A.P. Ehiasarian, High-power impulse magnetron sputtering and its applications,
Pure Appl. Chem. 82 (2010) 12471258, https://doi.org/10.1351/PAC-CON-09-
10-43.
[9] K. Sarakinos, J. Alami, S. Konstantinidis, High power pulsed magnetron sputtering:
a review on scientic and engineering state of the art, Surf. Coat. Technol. 204
(2010) 16611684, https://doi.org/10.1016/j.surfcoat.2009.11.013.
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
11
[10] A.P. Ehiasarian, A. Vetushka, Y.A. Gonzalvo, G. S´
afr´
an, L. Sz´
ekely, P.B. Barna,
Inuence of high power impulse magnetron sputtering plasma ionization on the
microstructure of TiN thin lms, J. Appl. Phys. 109 (2011), 104314, https://doi.
org/10.1063/1.3579443.
[11] T. Koz´
ak, A.D. Pajdarov´
a, M. ˇ
Cada, Z. Hubiˇ
cka, P. Mareˇ
s, J. ˇ
Capek, Ion energy
distributions at substrate in bipolar HiPIMS: effect of positive pulse delay, length
and amplitude, Plasma Sources Sci. Technol. 29 (2020), 065003, https://doi.org/
10.1088/1361-6595/ab8fbb.
[12] J.A. Santiago, I. Fern´
andez-Martínez, T. Koz´
ak, J. Capek, A. Wennberg, J.
M. Molina-Aldareguia, V. Bellido-Gonz´
alez, R. Gonz´
alez-Arrabal, M.A. Monclús,
The inuence of positive pulses on HiPIMS deposition of hard DLC coatings, Surf.
Coat. Technol. 358 (2019) 4349, https://doi.org/10.1016/j.
surfcoat.2018.11.001.
[13] Batkov´
a, J. ˇ
Capek, J. Rezek, R. ˇ
Cerstvý, P. Zeman, Effect of positive pulse voltage
in bipolar reactive HiPIMS on crystal structure, microstructure and mechanical
properties of CrN lms Surf. Coat. Technol. 393 (2020) 125773. doi:https://doi.
org/10.1016/j.surfcoat.2020.125773.
[14] G. Greczynski, J. Jensen, J. B¨
ohlmark, L. Hultman, Microstructure control of CrNx
lms during high power impulse magnetron sputtering, Surf. Coat. Technol. 205
(2010) 118130, https://doi.org/10.1016/j.surfcoat.2010.06.016.
[15] A. Anders, Tutorial: reactive high power impulse magnetron sputtering (R-
HiPIMS), J. Appl. Phys. 121 (2017), 171101, https://doi.org/10.1063/1.4978350.
[16] S.M. Rossnagel, Gas density reduction effects in magnetrons, J. Vac. Sci. Technol. A
Vacuum, Surfaces, Film. 6 (1988) 1924, https://doi.org/10.1116/1.574988.
[17] A. Anders, J. ˇ
Capek, M. H´
ala, L. Martinu, The ‘recycling trap: a generalized
explanation of discharge runaway in high-power impulse magnetron sputtering,
J. Phys. D. Appl. Phys. 45 (2012), 012003, https://doi.org/10.1088/0022-3727/
45/1/012003.
[18] J. ˇ
Capek, S. Kadlec, Return of target material ions leads to a reduced hysteresis in
reactive high power impulse magnetron sputtering: experiment, J. Appl. Phys. 121
(2017), 171911, https://doi.org/10.1063/1.4977816.
[19] A. Anders, Discharge physics of high power impulse magnetron sputtering, Surf.
Coat. Technol. 205 (2011) S1S9, https://doi.org/10.1016/J.
SURFCOAT.2011.03.081.
[20] L. Zauner, P. Ertelthaler, T. Wojcik, H. Bolvardi, S. Kolozsv´
ari, P.H. Mayrhofer,
H. Riedl, Reactive HiPIMS deposition of Ti-Al-N: inuence of the deposition
parameters on the cubic to hexagonal phase transition, Surf. Coat. Technol. 382
(2020), 125007, https://doi.org/10.1016/j.surfcoat.2019.125007.
[21] J. Paulitsch, P.H. Mayrhofer, W.D. Münz, M. Schenkel, Structure and mechanical
properties of CrN/TiN multilayer coatings prepared by a combined HIPIMS/UBMS
deposition technique, Thin Solid Films 517 (2008) 12391244, https://doi.org/
10.1016/j.tsf.2008.06.080.
[22] M. Samuelsson, D. Lundin, J. Jensen, M.A. Raadu, J.T. Gudmundsson,
U. Helmersson, On the lm density using high power impulse magnetron
sputtering, Surf. Coat. Technol. 205 (2010) 591596, https://doi.org/10.1016/j.
surfcoat.2010.07.041.
[23] T. Shimizu, H. Komiya, Y. Teranishi, K. Morikawa, H. Nagasaka, M. Yang, Pressure
dependence of (Ti, Al)N lm growth on inner walls of small holes in high-power
impulse magnetron sputtering, Thin Solid Films 624 (2017) 189196, https://doi.
org/10.1016/j.tsf.2016.09.041.
[24] M. Balzer, M. Fenker, Three-dimensional thickness and property distribution of TiC
lms deposited by DC magnetron sputtering and HIPIMS, Surf. Coat. Technol. 250
(2014) 3743, https://doi.org/10.1016/j.surfcoat.2014.02.011.
[25] G. Greczynski, I. Zhirkov, I. Petrov, J.E. Greene, J. Rosen, Control of the metal/gas
ion ratio incident at the substrate plane during high-power impulse magnetron
sputtering of transition metals in Ar, Thin Solid Films 642 (2017) 3640, https://
doi.org/10.1016/j.tsf.2017.09.027.
[26] G. Greczynski, I. Petrov, J.E. Greene, L. Hultman, Strategy for tuning the average
charge state of metal ions incident at the growing lm during HIPIMS deposition,
Vacuum. 116 (2015) 3641, https://doi.org/10.1016/j.vacuum.2015.02.027.
[27] D. Lundin, M. ˇ
Cada, Z. Hubiˇ
cka, Ionization of sputtered Ti, Al, and C coupled with
plasma characterization in HiPIMS, Plasma Sources Sci. Technol. 24 (2015),
035018, https://doi.org/10.1088/0963-0252/24/3/035018.
[28] N. Britun, S. Konstantinidis, R. Snyders, An overview on time-resolved optical
analysis of HiPIMS discharge, Plasma Process. Polym. 12 (2015) 10101027,
https://doi.org/10.1002/ppap.201500051.
[29] G. Greczynski, L. Hultman, Time and energy resolved ion mass spectroscopy
studies of the ion ux during high power pulsed magnetron sputtering of Cr in Ar
and Ar/N2 atmospheres, Vacuum. 84 (2010) 11591170, https://doi.org/
10.1016/j.vacuum.2010.01.055.
[30] M. Palmucci, N. Britun, T. Silva, R. Snyders, S. Konstantinidis, Mass spectrometry
diagnostics of short-pulsed HiPIMS discharges, J. Phys. D. Appl. Phys. 46 (2013),
215201, https://doi.org/10.1088/0022-3727/46/21/215201.
[31] A. Ferrec, J. Keraudy, S. Jacq, F. Schuster, P.Y. Jouan, M.A. Djouadi, Correlation
between mass-spectrometer measurements and thin lm characteristics using
dcMS and HiPIMS discharges, Surf. Coat. Technol. 250 (2014) 5256, https://doi.
org/10.1016/j.surfcoat.2014.02.030.
[32] J. Bohlmark, M. Lattemann, J.T. Gudmundsson, A.P. Ehiasarian, Y.
Aranda Gonzalvo, N. Brenning, U. Helmersson, The ion energy distributions and
ion ux composition from a high power impulse magnetron sputtering discharge,
Thin Solid Films 515 (2006) 15221526, https://doi.org/10.1016/j.
tsf.2006.04.051.
[33] G. Greczynski, I. Zhirkov, I. Petrov, J.E. Greene, J. Rosen, Time evolution of ion
uxes incident at the substrate plane during reactive high-power impulse
magnetron sputtering of groups IVb and VIb transition metals in Ar/N 2, J. Vac.
Sci. Technol. A 36 (2018), 020602, https://doi.org/10.1116/1.5016241.
[34] G. Greczynski, J. Lu, M.P. Johansson, J. Jensen, I. Petrov, J.E. Greene, L. Hultman,
Role of tin+and Aln+ion irradiation (n=1, 2) during Ti1-xAlxN alloy lm growth
in a hybrid HIPIMS/magnetron mode, Surf. Coat. Technol. 206 (2012) 42024211,
https://doi.org/10.1016/j.surfcoat.2012.04.024.
[35] G. Greczynski, I. Petrov, J.E. Greene, L. Hultman, Paradigm shift in thin-lm
growth by magnetron sputtering: from gas-ion to metal-ion irradiation of the
growing lm, J. Vac. Sci. Technol. A 37 (2019), 060801, https://doi.org/10.1116/
1.5121226.
[36] G. Greczynski, S. Mr´
az, M. Hans, J. Lu, L. Hultman, J. Schneider, Control over the
phase formation in metastable transition metal nitride thin lms by tuning the Al+
subplantation depth, Coatings. 9 (2018) 17, https://doi.org/10.3390/
coatings9010017.
[37] M. Lattemann, A.P. Ehiasarian, J. Bohlmark, P.Å.O. Persson, U. Helmersson,
Investigation of high power impulse magnetron sputtering pretreated interfaces for
adhesion enhancement of hard coatings on steel, Surf. Coat. Technol. 200 (2006)
64956499, https://doi.org/10.1016/j.surfcoat.2005.11.082.
[38] A.P. Ehiasarian, J.G. Wen, I. Petrov, Interface microstructure engineering by high
power impulse magnetron sputtering for the enhancement of adhesion, J. Appl.
Phys. 101 (2007), https://doi.org/10.1063/1.2697052.
[39] G. Greczynski, J. Lu, J. Jensen, I. Petrov, J.E. Greene, S. Bolz, W. K¨
olker,
C. Schiffers, O. Lemmer, L. Hultman, Metal versus rare-gas ion irradiation during
Ti1-xAlxN lm growth by hybrid high power pulsed magnetron/dc magnetron co-
sputtering using synchronized pulsed substrate bias, J. Vac. Sci. Technol. A 30
(2012), 061504, https://doi.org/10.1116/1.4750485.
[40] N. Nedfors, O. Vozniy, J. Rosen, Effect of synchronized bias in the deposition of TiB
2 thin lms using high power impulse magnetron sputtering, J. Vac. Sci. Technol. A
Vacuum, Surfaces, Film. 36 (2018), 031510, https://doi.org/10.1116/1.5003194.
[41] B. Bakhit, I. Petrov, J.E. Greene, L. Hultman, J. Ros´
en, G. Greczynski, Controlling
the B/Ti ratio of TiB x thin lms grown by high-power impulse magnetron
sputtering, J. Vac. Sci. Technol. A Vacuum, Surfaces, Film. 36 (2018), 030604,
https://doi.org/10.1116/1.5026445.
[42] W. Münz, Titanium aluminum nitride lms: a new alternative to TiN coatings,
J. Vac. Sci. Technol. A 4 (1986) 27172725, https://doi.org/10.1116/1.573713.
[43] O. Knotek, T. Leyendecker, On the structure of (Ti, Al)N-PVD coatings, J. Solid
State Chem. 70 (1987) 318322, https://doi.org/10.1016/0022-4596(87)90071-5.
[44] P.H. Mayrhofer, D. Music, J.M. Schneider, Inuence of the Al distribution on the
structure, elastic properties, and phase stability of supersaturated Ti1xAlxN,
J. Appl. Phys. 100 (2006), 094906, https://doi.org/10.1063/1.2360778.
[45] L. Chen, Y. Du, P.H. Mayrhofer, S.Q. Wang, J. Li, The inuence of age-hardening on
turning and milling performance of Ti-Al-N coated inserts, Surf. Coat. Technol. 202
(2008) 51585161, https://doi.org/10.1016/j.surfcoat.2008.05.036.
[46] L. Chen, M. Moser, Y. Du, P.H. Mayrhofer, Compositional and structural evolution
of sputtered Ti-Al-N, Thin Solid Films 517 (2009) 66356641, https://doi.org/
10.1016/j.tsf.2009.04.056.
[47] K. Kutschej, P.H. Mayrhofer, M. Kathrein, P. Polcik, R. Tessadri, C. Mitterer,
Structure, mechanical and tribological properties of sputtered Ti1-xAlxN coatings
with 0.5x0.75, Surf. Coat. Technol. 200 (2005) 23582365, https://doi.org/
10.1016/j.surfcoat.2004.12.008.
[48] L. Chen, J. Paulitsch, Y. Du, P.H. Mayrhofer, Thermal stability and oxidation
resistance of TiAlN coatings, Surf. Coat. Technol. 206 (2012) 29542960,
https://doi.org/10.1016/j.surfcoat.2011.12.028.
[49] G. Greczynski, J. Lu, J. Jensen, S. Bolz, W. K¨
olker, C. Schiffers, O. Lemmer, J.
E. Greene, L. Hultman, A review of metal-ion-ux-controlled growth of metastable
TiAlN by HIPIMS/DCMS co-sputtering, Surf. Coat. Technol. 257 (2014) 1525,
https://doi.org/10.1016/j.surfcoat.2014.01.055.
[50] C.L. Chang, F.C. Yang, Effect of target composition on the microstructural,
mechanical, and corrosion properties of TiAlN thin lms deposited by high-power
impulse magnetron sputtering, Surf. Coat. Technol. 352 (2018) 330337, https://
doi.org/10.1016/j.surfcoat.2018.08.023.
[51] S. Severin, M. Naveed, S. Weiß, Effect of HPPMS pulse-frequency on plasma
discharge and deposited AlTiN coating properties, Adv. Mater. Sci. Eng. (2017) 18,
https://doi.org/10.1155/2017/4850908.
[52] G. Greczynski, S. Mr´
az, M. Hans, D. Primetzhofer, J. Lu, L. Hultman, J.
M. Schneider, Unprecedented Al supersaturation in single-phase rock salt structure
VAlN lms by Al+subplantation, J. Appl. Phys. 121 (2017), 171907, https://doi.
org/10.1063/1.4977813.
[53] G. Greczynski, S. Mr´
az, L. Hultman, J.M. Schneider, Selectable phase formation in
VAlN thin lms by controlling Al+subplantation depth, Sci. Rep. 7 (2017) 17544,
https://doi.org/10.1038/s41598-017-17846-5.
[54] D.R. Lide, CRC Handbook of Chemistry and Physics, 84th ed., CRC Press, 2003.
[55] Hiden Analytical Ltd. EQP and EQS Analysers, (n.d.). https://www.hiden.de/wp-c
ontent/uploads/pdf/EQP_and_EQS_-_Hiden_Analytical_Techinical_Inofrmation.pdf
(accessed November 26, 2020).
[56] Private communications between Hiden Analytical Ltd. and T. Koz´
ak, (2019).
[57] M. Klinger, A. J¨
ager, Crystallographic tool box (CrysTBox): automated tools for
transmission electron microscopists and crystallographers, J. Appl. Crystallogr. 48
(2015) 20122018, https://doi.org/10.1107/S1600576715017252.
[58] G.M. Pharr, An improved technique for determining hardness and elastic modulus
using load and displacement sensing indentation experiments, J. Mater. Res. 7
(1992) 15641583, https://doi.org/10.1557/JMR.1992.1564.
[59] G.C.A.M. Janssen, M.M. Abdalla, F. van Keulen, B.R. Pujada, B. van Venrooy,
Celebrating the 100th anniversary of the Stoney equation for lm stress:
developments from polycrystalline steel strips to single crystal silicon wafers, Thin
Solid Films 517 (2009) 18581867, https://doi.org/10.1016/j.tsf.2008.07.014.
L. Zauner et al.
Surface & Coatings Technology 424 (2021) 127638
12
[60] D. Depla, X.Y. Li, S. Mahieu, R. De Gryse, Determination of the effective electron
emission yields of compound materials, J. Phys. D. Appl. Phys. 41 (2008), https://
doi.org/10.1088/0022-3727/41/20/202003.
[61] F. Magnus, O.B. Sveinsson, S. Olafsson, J.T. Gudmundsson, Current-voltage-time
characteristics of the reactive Ar/N 2 high power impulse magnetron sputtering
discharge, J. Appl. Phys. 110 (2011), 083306, https://doi.org/10.1063/
1.3653233.
[62] J.Y. Rauch, C. Rousselot, N. Martin, Structure and composition of TixAll-xN thin
lms sputter deposited using a composite metallic target, Surf. Coat. Technol. 157
(2002) 138143, https://doi.org/10.1016/S0257-8972(02)00146-9.
[63] S. Inoue, H. Uchida, A. Hioki, K. Koterazawa, R.P. Howson, Structure and
composition of (Ti, Al)N lms prepared by r.f. planar magnetron sputtering using a
composite target, Thin Solid Films 271 (1995) 1518, https://doi.org/10.1016/
0040-6090(95)06817-1.
[64] A.D. Pajdarov´
a, J. Vlˇ
cek, P. Kudl´
aˇ
cek, J. Luk´
aˇ
s, Electron energy distributions and
plasma parameters in high-power pulsed magnetron sputtering discharges, Plasma
Sources Sci. Technol. 18 (2009), https://doi.org/10.1088/0963-0252/18/2/
025008, 025008.
[65] ICDD, Powder Diffraction File - wurtzite AlN - 04-016-3965, (2013).
[66] D. Gall, S. Kodambaka, M.A. Wall, I. Petrov, J.E. Greene, Pathways of atomistic
processes on TiN(001) and (111) surfaces during lm growth: an ab initio study,
J. Appl. Phys. 93 (2003) 90869094, https://doi.org/10.1063/1.1567797.
[67] L. Hultman, J.E. Sundgren, J.E. Greene, Formation of polyhedral N 2 bubbles
during reactive sputter deposition of epitaxial TiN(100) lms, J. Appl. Phys. 66
(1989) 536544, https://doi.org/10.1063/1.343570.
[68] G. Greczynski, J. Lu, J. Jensen, I. Petrov, J.E. Greene, S. Bolz, W. K¨
olker,
C. Schiffers, O. Lemmer, L. Hultman, Strain-free, single-phase metastable
Ti0.38Al0.62N alloys with high hardness: metal-ion energy vs. momentum effects
during lm growth by hybrid high-power pulsed/dc magnetron cosputtering, Thin
Solid Films 556 (2014) 8798, https://doi.org/10.1016/j.tsf.2014.01.017.
[69] ICDD, Powder diffraction le - cubic TiN - 00-038-1420, (1970).
[70] ICDD, Powder diffraction le - cubic AlN - 00-025-1495, (1970).
[71] ICDD, Powder diffraction le - cubic Si - 00-027-1402, (2017) 78.
L. Zauner et al.
Article
Metal-ion etching is widely used in substrate surface modification before coating by physical vapor deposition. In this work, TiAlN coatings were synthesized onto cemented carbides with Ti-, Cr-, and Zr-ion pre-etching by cathodic arc evaporation and investigated concerning surface topography and mechanical failure. The pre-etching of energetic metal ions introduces droplet defects and increases surface roughness through uneven bombardment. The high interface roughness and deposition of nanoscale metal interlayers contributed to the upgraded adhesion between the coating and the substrate. As a result, the adhesion strength of TiAlN coatings advances to more than 95 N by applying Cr- and Zr-ion pre-etching. The rough substrate surface also gives rise to a rough coating surface associated with the pre-etching time. Under normal loading, the microscopic pit-like surface morphology, stemming from ion etching, interrupts the transverse propagation of the circumferential cracks but does not affect the longitudinal crack propagation inside the coating. In addition, the higher roughness of coatings with ion pre-etching causes a shorter run-in stage during ball-on-disc tests due to the higher contact stress between the friction pair.
Article
Full-text available
Ion irradiation is a key tool for controlling the nanostructure, phase content, and physical properties of refractory ceramic thin films grown at low temperatures by magnetron sputtering. However, in contrast to gas-ion bombardment, the effects of metal-ion irradiation on properties of refractory ceramic thin films have not been extensively studied due to (i) low metal-ion concentrations (a few percents) during standard direct-current magnetron sputtering (DCMS) and (ii) difficulties in separating metal-ion from gas-ion fluxes. Recently, the situation has changed dramatically, thanks to the development of high-power impulse magnetron sputtering (HiPIMS), which provides highly-ionized metal-ion plasmas. In addition, careful choice of sputtering conditions allows exploitation of gas-rarefaction effects such that the charge state, energy, and momentum of metal ions incident at the growing film surface can be tuned. This is possible via the use of pulsed substrate bias, synchronized to the metal-ion-rich portion of each HiPIMS pulse. In this review, the authors begin by summarizing the results of time-resolved mass spectrometry analyses performed at the substrate position during HiPIMS and HiPIMS/DCMS cosputtering of transition-metal (TM) targets in Ar and Ar/N2 atmospheres. Knowledge of the temporal evolution of metal- and gas-ion fluxes is essential for precise control of the incident metal-ion energy and for minimizing the role of gas-ion irradiation. Next, the authors review results on the growth of binary, pseudobinary, and pseudoternary TM nitride alloys by metal-ion-synchronized HiPIMS. In contrast to gas ions, a fraction of which are trapped at interstitial sites, metal ions are primarily incorporated at lattice sites resulting in much lower compressive stresses. In addition, the closer mass match with the film-forming species results in more efficient momentum transfer and provides the recoil density and energy necessary to eliminate film porosity at low deposition temperatures. Several novel film-growth pathways have been demonstrated: (i) nanostructured N-doped bcc-CrN0.05 films combining properties typically associated with both metals and ceramics, (ii) fully-dense, hard, and stress-free Ti0.39Al0.61N, (iii) single-phase cubic Ti1−xSixN with the highest reported SiN concentrations, (iv) unprecedented AlN supersaturation in single-phase NaCl-structure V1−xAlxN, and (v) a dramatic increase in the hardness, due to selective heavy-metal ion bombardment during growth, of dense Ti0.92Ta0.08N films deposited with no external heating.
Article
Full-text available
The performance of transition metal nitride based coatings deposited by magnetron sputtering, in a broad range of applications including wear-protective coatings on cutting tools and components in automotive engines, is determined by their phase content. The classical example is the precipitation of thermodynamically-favored wurtzite-AlN while alloying TiN with Al to obtain ternary single phase NaCl-structure films with improved high-temperature oxidation resistance. Here, we report on reactive high-power impulse and direct current magnetron co-sputtering (HiPIMS/DCMS) growth of Ti0.31Al0.69N and Zr0.48Al0.52N thin films. The Al concentrations are intentionally chosen to be higher than theoretically predicted solubility limits for the rock salt structure. The goal is to investigate the effect of the incident Al+ energy EAl+, controlled by varying the amplitude of the substrate bias applied synchronously with the Al+-rich portion of the ion flux from the Al-HiPIMS source, on the crystalline phase formation. For EAl+ ≤ 60 eV, films contain predominantly the wurtzite phase. With increasing EAl+, and thus, the Al subplantation depth, the relative fraction of the NaCl structure increases and eventually for EAl+ > 250 eV, Ti0.31Al0.69N and Zr0.48Al0.52N layers contain more than 95% of the rock salt phase. Thus, the separation of the film forming species in time and energy domains determines the phase formation of Ti0.31Al0.69N and Zr0.48Al0.52N layers and enables the growth of the cubic phase outside of the predicted Al concentration range. The new film growth concept can be applied to the entire family of multinary transition metal aluminum nitrides, where one of the metallic film constituents is available in the ionized form while the other arrives as neutral.
Article
Full-text available
Reactive transition-metal (TM) nitride film growth employing bias-synchronized high power impulse magnetron sputtering (HiPIMS) requires a detailed knowledge of the time evolution of metal- and gas-ion fluxes incident at the substrate plane in order to precisely tune momentum transfer and, hence, provide the recoil density and energy necessary to eliminate film porosity at low deposition temperatures without introducing significant film stress. Here, the authors use energy- and time-dependent mass spectrometry to analyze the evolution of metal- and gas-ion fluxes at the substrate plane during reactive HiPIMS sputtering of groups IVb and VIb TM targets in Ar/N2 atmospheres. The time- and energy-integrated metal/gas ion ratio NMe+/Ng+ incident at the substrate is significantly lower for group IVb TMs (ranging from 0.2 for Ti to 0.9 for Hf), due to high N2 reactivity which results in severely reduced target sputtering rates and, hence, decreased rarefaction. In contrast, for less reactive group VIb metals, sputtering rates are similar to those in pure Ar as a result of significant gas heating and high NMe+/Ng+ ratios, ranging from 2.3 for Cr to 98.1 for W. In both sets of experiments, the peak target current density is maintained constant at 1 A/cm². Within each TM group, NMe+/Ng+ scales with increasing metal-ion mass. For the group-VIb elements, sputtered-atom Sigmund-Thompson energy distributions are preserved long after the HiPIMS pulse, in contradistinction to group-IVb TMs for which the energy distributions collapse into narrow thermalized peaks. For all TMs, the N⁺ flux dominates that of N2⁺ ions, as the molecular ions are collisionally dissociated at the target, and N⁺ exhibits ion energy distribution functions resembling those of metal ions. The latter result implies that both N⁺ and Me⁺ species originate from the target. High-energy Ar⁺ tails, assigned to ionized reflected-Ar neutrals, are observed with heavier TM targets.
Article
CrN films were prepared using three different configurations of the HiPIMS discharge mode and the substrate holder potential. We investigate the effect of a positive pulse voltage (30–400 V) in bipolar HiPIMS on the crystal structure, microstructure and resulting mechanical properties of the films, and compare it to the effect of a standard DC bias voltage applied to the substrate holder in unipolar HiPIMS. We find that when the substrate holder is at a floating potential, its charging causes the loss of the plasma-substrate potential difference, necessary for ion acceleration, and no obvious evolution is thus observed with increasing positive pulse voltage. However, when the substrate holder is grounded, the effect of the positive pulse voltage is apparent and different from the effect of the DC bias substrate voltage. That is mainly due to differences in energies delivered into the growing film by bombarding ions. Films prepared using bipolar HiPIMS at a positive pulse voltage of 90 and 120 V exhibit the most interesting properties, namely high hardness (23.5 and 23.1 GPa, respectively) at a relatively low residual compressive stress (1.7 and 1.5 GPa, respectively). The results indicate that as long as the growing film is conductively connected with the ground, bipolar HiPIMS is a suitable method to tailor and improve the film properties.
Article
Reactive high-power impulse magnetron sputtering (R-HiPIMS) is seen as a key-technology for the deposition of future hard and multifunctional coatings. Increased ionisation rates allow for additional possibilities in tuning specific coating characteristics based on growth mechanisms varied by surface-diffusion. Especially within the well-established Ti-Al-N system, the Al solubility limit (xmax) of metastable face-centred-cubic (fcc) Ti1−xAlxN is an everlasting scientific topic. Here, we investigate in detail the dependence of xmax on various deposition parameters (i.e. pulse frequency and duration, N2-to-Ar flow rate ratio, and substrate bias potential) during R-HiPIMS of Ti-Al-N coatings using Ti0.6Al0.4, Ti0.5Al0.5 and Ti0.4Al0.6 composite targets. The systematic studies showed that the highest solubility limit of xmax ∼0.55 could be obtained for duty cycles around 3.75% (peak power densities of ∼1.0 kW/cm2) and a N2-to-Ar flow rate ratio of 0.3. A further decisive fact for the deposition of high Al containing fcc-structured Ti1−xAlxN coatings is surface diffusion controlled by bias potentials (DC as well as modulated pulses) ensuring sufficient intermixing of the arriving film species. Despite the presence of very small amounts of wurtzite-typed phases, excellent hardness values of ∼36 GPa for Ti0.40Al0.60N – which further increased to ∼ 40 GPa upon annealing for 1 h at 700 °C – could be achieved for a DC bias potential of -50 V, irrespective of all variations conducted. Based on our results we can further conclude, that the ratio and energy of Tin+- and Aln+-ions, simultaneously arriving at the substrate surface, are decisive for stabilising the highly preferred cubic modification. A distinct promotion of specific discharge regimes – selected by synchronised bias pulses – can thus positively influence the cubic phase formation through altered gas-to-metal ion ratios arriving at the film surface.
Article
Diamond-like Carbon (DLC) coatings were deposited by a novel HiPIMS method that incorporates positive voltage pulses at the end of the conventional HiPIMS discharge. Different positive voltage amplitudes (100, 200, 300, 400 and 500 V) were used to evaluate the effect of this operation mode on the discharge process and the mechanical properties of the deposited DLC coatings. The application of positive pulses was observed to enhance the ionization of both the sputtered carbon and argon species. Mass spectroscopy measurements showed that a larger amount of high-energy C⁺ ions are generated, with ion energies proportional to the amplitude of the overshoot voltage. The ion bombardment induced by the positive pulses led to higher compressive residual stresses and densification of deposited DLC coatings. Moreover, their Raman spectra exhibited lower D-band and G-band intensity ratios (ID/IG) as the pulses voltage was increased which is indicative of higher sp³ content. Mechanical properties were evaluated by nanoindentation testing and the hardness of the deposited DLC films was observed to increase from 9.6 GPa (for no voltage pulse applied) to 22.5 GPa (for an applied positive pulse voltage of 500 V).
Article
Ti1−xAlxN thin films were deposited using reactive high-power impulse magnetron sputtering on silicon Si (100), tungsten carbide (WC-Co), and SUS304 substrates using three metallurgical TixAly alloy targets with Al/(Al + Ti) atomic ratios of 0.19, 0.36, and 0.57. A fixed duty cycle of 1% and a N2/Ar ratio of 15.8% was used. The current–voltage characteristics of the TixAly alloy targets were first investigated. Then, the microstructural evolution and the mechanical and corrosion properties of the as-deposited Ti1−xAlxN thin films were analysed by scanning electron microscopy, grazing incident X-ray diffractometry, X-ray photoelectron spectroscopy, transmission electron microscopy, nanoindentation, and anodic potentiodynamic polarisation tests. The power density of 1.86 kW/cm² achieved with the square-like power output wave obtained in the constant current mode was higher than the 1.37 kW/cm² power density obtained with the triangular-like power output wave. The deposition rates of the Ti1−xAlxN thin films increased from 13.4 to 28.6 nm/min with an increasing alloy target Al/(Al + Ti) atomic ratio due to the higher sputter yield of Al than Ti. The Al/(Al + Ti) atomic ratio of the alloy target determines the structure, which changed from a highly preferred TiN (111) orientation (Al/(Al + Ti) ratio = 0.19) to a nanocomposite structure (Al/(Al + Ti) ratio = 0.57). The nanoindentation results indicated that the highest hardness and elastic modulus values of the Ti1−xAlxN thin films were 33.3 GPa and 405.9 GPa, respectively, obtained using the Al/(Al + Ti) = 0.19 alloy target. In addition, the friction coefficients of the Ti1−xAlxN thin films were between 0.68 and 0.77. However, the potentiostatic polarisation measurements showed that the amorphous TiAlN films exhibited the highest corrosion resistance in 3.5% NaCl solution, as the corrosion resistance of the Ti1−xAlxN coatings increased as the Al/(Al + Ti) atomic ratio increased.
Article
TiBx thin films grown from compound TiB2 targets by magnetron sputter deposition are typically highly over-stoichiometric, with x ranging from 3.5 to 2.4, due to differences in Ti and B preferential-ejection angles and gas-phase scattering during transport from the target to the substrate. Here, the authors demonstrate that stoichiometric TiB2 films can be obtained using high-power impulse magnetron sputtering (HiPIMS) operated in power-controlled mode. The B/Ti ratio x of films sputter-deposited in Ar is controllably varied from 2.08 to 1.83 by adjusting the length of HiPIMS pulses ton between 100 and 30 μs, while maintaining average power and pulse frequency constant. This results in peak current densities JT,peak ranging from 0.27 to 0.88 A/cm2. Energy- and time-resolved mass spectrometry analyses of the ion fluxes incident at the substrate position show that the density of metal ions increases with decreasing ton due to a dramatic increase in JT,peak resulting in the strong gas rarefaction. With ton < 60 μs (JT,peak > 0.4 A/cm2), film growth is increasingly controlled by ions incident at the substrate, rather than neutrals, as a result of the higher plasma dencity and, hence, electron-impact ionization probablity. Thus, since sputter-ejected Ti atoms have a higher probability of being ionized than B atoms, due to their lower first-ionization potential and larger ionization cross-section, the Ti concentration in as-deposited films increases with decreasing ton (increasing JT,peak) as ionized sputtered species are steered to the substrate by the plasma in order to maintain charge neutrality.
Article
Titanium diboride thin films have been deposited from a compound TiB2 target on Si(001) substrates at a temperature of 500 °C using high power impulse magnetron sputtering (HiPIMS) at a frequency of 1000 Hz and pulse lengths of 20 and 40 μs. A −60 V bias pulse of different pulse length was applied at different time delay relative to the HiPIMS pulse. The average energy per deposited species, ⟨ED⟩ = Ei(Ji/Jt), where Ei is the average ion energy and Ji/Jt is the ratio of the ion bombarding flux to the total flux of deposited species, is strongly dependent on bias mode. A change in preferred orientation from (101) to (001) is observed when ⟨ED⟩ increase above 50 eV. The limited adatom mobility at ⟨ED⟩ below 50 eV promote growth of fast growing planes resulting in a (101) texture, while ⟨ED⟩ above 50 eV supply sufficient energy for development of the thermodynamically more favorable (001) texture. A linear increase in compressive residual stress with the increase in ⟨ED⟩ is also found, due to more intensive Ar⁺ ion bombardment. Analysis of charge-state-resolved plasma chemistry and ion energy shows that the total flux of bombarding ions contains a higher fraction of B⁺ when the bias is applied in synchronous with the HiPIMS pulse instead of after, resulting in a lower residual stress at similar values of ⟨ED⟩ (cf. −2.0 ± 0.2 and −2.6 ± 0.1 GPa). This study shows that use of a bias synchronized in different modes relative to the HiPIMS pulse, can be used as a tool to control ⟨ED⟩ and to some extent the type of bombarding species, and hence the microstructure of TiB2 thin films.