Role of Si segregation in the structural, mechanical, and compositional evolution of high-temperature oxidation resistant Cr-Si-B2±z thin films

Abstract

This work investigates the influence of Si-alloying up to 17 at.% on the structural, mechanical, and oxidation properties of magnetron sputtered CrB2±z-based thin films. Density-functional theory calculations combined with atom probe tomography reveal the preferred Si occupation of Cr-lattice sites and an effective solubility limit between 3 to 4 at.% in AlB2-structured solid solutions. The addition of Si results in refinement of the columnar morphology, accompanied by enhanced segregation of excess Si along grain boundaries. The microstructural separation leads to a decrease in both film hardness and Young’s modulus from H ~ 24 to 17 GPa and E ~ 300 to 240 GPa, respectively, dominated by the inferior mechanical properties of the intergranular Si-rich regions. Dynamic thermogravimetry up to 1400 °C reveals a significant increase in oxidation onset temperature from 600 to 1100 °C above a Si content of 8 at.%. In-situ X-ray diffraction correlates the protective mechanism with thermally activated precipitation of Si from the Cr-Si-B2±z solid solution at 600 °C, enabling the formation of a stable, nanometer-sized SiO2¬¬-based scale. Moreover, high-resolution TEM analysis exposes the scale architecture after dynamic oxidation to 1200 °C (10 K/min heating rate) – consisting only of ~20 nm amorphous SiO2 beneath ~200 nm of nanocrystalline Cr2O3. In summary, the study provides detailed guidelines connecting the chemical composition with the respective thin film properties of high-temperature oxidation resistant Cr-Si-B2±z coatings.

Full text

Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jalcom
Research article
Role of Si segregation in the structural, mechanical, and compositional
evolution of high-temperature oxidation resistant Cr-Si-B
2 ± z
thin films
L. Zauner
a,⁎,1
, A. Steiner
a
, T. Glechner
a
, A. Bahr
a
, B. Ott
b
, R. Hahn
a
, T. Wojcik
a,c
, O. Hunold
d
,
J. Ramm
d
, S. Kolozsvári
e
, P. Polcik
e
, P. Felfer
b
, H. Riedl
a,c
a
Christian Doppler Laboratory for Surface Engineering of high-performance Components, TU Wien, Austria
b
Department of Materials Science, Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany
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
Article history:
Received 13 December 2022
Received in revised form 2 February 2023
Accepted 6 February 2023
Available online 7 February 2023
Keywords:
Borides
Thin films
Si alloying
Oxidation resistance
Phase stability
Mechanical properties
Segregation
abstract
This work investigates the influence of Si-alloying up to 17 at.% on the structural, mechanical, and oxidation
properties of magnetron sputtered CrB
2 ± z
-based thin films. Density-functional theory calculations com-
bined with atom probe tomography reveal the preferred Si occupation of Cr-lattice sites and an effective
solubility limit between 3 to 4 at.% in AlB
2
-structured solid solutions. The addition of Si results in refine-
ment of the columnar morphology, accompanied by enhanced segregation of excess Si along grain
boundaries. The microstructural separation leads to a decrease in both film hardness and Young’s modulus
from H ∼ 24 to 17 GPa and E ∼ 300 to 240 GPa, respectively, dominated by the inferior mechanical prop-
erties of the intergranular Si-rich regions. Dynamic thermogravimetry up to 1400 °C reveals a significant
increase in oxidation onset temperature from 600 to 1100 °C above a Si content of 8 at.%. In-situ X-ray
diffraction correlates the protective mechanism with thermally activated precipitation of Si from the Cr-Si-
B
2 ± z
solid solution at 600 °C, enabling the formation of a stable, nanometer-sized SiO
2
-based scale.
Moreover, high-resolution TEM analysis exposes the scale architecture after dynamic oxidation to 1200 °C
(10 K/min heating rate) – consisting only of ∼20 nm amorphous SiO
2
beneath ∼200 nm of nanocrystalline
Cr
2
O
3
. In summary, the study provides detailed guidelines connecting the chemical composition with the
respective thin film properties of high-temperature oxidation resistant Cr-Si-B
2 ± z
coatings.
© 2023 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/).
1. Introduction
Transition metal diboride (TMB
2
) thin films are promising can-
didates to replace state-of-the-art functional and protective coating
materials in a wide range of applications [1–8]. TMB
2
s typically
feature a high melting temperature, excellent thermal stability, as
well as high hardness and strength, thus providing a strong incentive
for ultra-high temperature applications [9–13]. However, this out-
standing property spectrum is usually confined to inert atmospheres
due to the consecutive/competitive formation of both TM- and B-
based oxide scales, both usually incapable of forming a fully pro-
tective layer at temperatures beyond 600–700 °C [14–18]. More
drastically, above ∼1100 °C linear mass gain kinetics are regularly
observed, which coincides with the evaporation of the glassy-like
boria (B
2
O
3
) embedded within the non-protective, porous metal
oxide [18,19].
Different alloying concepts have been studied and implemented
successfully to address the poor oxidation resistance of TMB
2
bulk
and thin film materials. Adding Si-containing compounds such as
SiC, MoSi
2
, Si
3
N
4
, or Ta
5
Si
3
is the most commonly used method for
bulk diboride materials and improves the oxidation resistance by
forming a stable, amorphous (boro-)silicate surface layer [19–21].
For instance, Fahrenholtz et al. demonstrated that adding SiC to ZrB
2
and HfB
2
permits drastically decreased oxidation rates up to
1600 °C [19].
Regarding TMB
2
-based thin film materials, several ternary al-
loying routes, e.g., the addition of Al(B
2
), TaB
2
, or CrB
2
, have been
explored to improve the oxidation resistance [14–16,22,23]. Bakhit
et al. [15] demonstrated that Al alloying into TiB
2
-based thin films
Journal of Alloys and Compounds 944 (2023) 169203
https://doi.org/10.1016/j.jallcom.2023.169203
0925-8388/© 2023 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/).
]]]]
]]]]]]
⁎
Correspondence to: Christian Doppler Laboratory for Surface Engineering of high-
performance Components, TU Wien, Getreidemarkt 9, 1060 Wien, Austria.
E-mail address: lukas.zauner@tuwien.ac.at (L. Zauner).
1
ORCID-ID: 0000–0002-8373–6552

significantly retards the oxide scale growth at temperatures up to
800 °C based on the formation of a dense Al-oxide surface layer.
Moreover, Kashani et al. [17] showed that the Ti-Al-B
2 ± z
system even
outperforms the corresponding nitride system at 700 °C, especially
for stoichiometric compositions close to B/TM ratios of 2. Indeed,
this necessity for tailoring the B/TM-ratio within TiB
2 ± z
thin films to
optimize oxidation properties is highlighted in several works and
rationalized by the fast-track oxidation pathway created through
excess B-rich phases preferentially located at column and grain
boundaries [24–26]. However, this effect appears specific to the
material system and/or annealing treatment conditions, since a
stable, protective boria surface layer was observed for HfB
2.3
thin
films up to 900 °C [27].
A seemingly universal alloying route for improved oxidation re-
sistance was recently published by Glechner et al. [28,29], showing
that co-sputtering of pure Si to various TMB
2
(TM = Ti, Cr, Hf, Ta, W)
drastically improves the oxidation resistance in all materials, with
the onset of oxidation elevated to 1200 °C specifically for Cr-Si-B
2
and Hf-Si-B
2
. Thereby, the protective mechanism relies on the for-
mation of a stable Si-rich oxide scale above the formed TM-Si-B
2
solid solution.
Inevitably, the ensemble of available tools to improve the oxi-
dation resistance within TMB
2
thin films influences the property
spectrum of the initial binary alloy. While strategies involving
ternary TM
1
-TM
2
-B
2 ± z
thin films proved successful only below
800 °C, their mechanical properties, including hardness and fracture
toughness, can often be preserved or even improved over the cor-
responding binary constituents [13–15]. Contrary, Si-based protec-
tive mechanisms can result in superior high-temperature
performance of TM-Si-B
2
thin films, however, typically at the ex-
pense of reduced mechanical properties at higher Si contents
[28,30]. Consequently, finding the optimum alloying content to
achieve the desired oxidation resistance while maintaining good
mechanical properties is a vital prerequisite for an industrial appli-
cation of the entire TM-Si-B
2
material family.
Therefore, within this work we systematically study the influ-
ence of Si-alloying on the structural evolution, phase stability, as
well as the mechanical and oxidation properties of magnetron
sputtered Cr-Si-B
2 ± z
thin films. This novel high-temperature ceramic
is modelled by density functional theory calculations to reveal the
energetically preferred lattice occupation of the alloying atom
within various AlB
2
-structured compositions. Furthermore, limita-
tions in the accessible alloying range to yield solid solutions are
discussed in conjunction with detailed atom probe tomography,
thereby spanning a clear connection to the observed thin film
growth and mechanical properties. The mechanism leading to the
drastically increased oxidation resistance is revisited through dy-
namic oxidation, in-situ X-ray diffraction, and transmission electron
microscopy, thus providing an in-depth correlation to the thin film
properties towards finding an optimum Si alloying composition.
2. Experimental
Cr-Si-B
2 ± z
thin films were synthesized from a 3-inch CrB
2
target
(Plansee Composite Materials GmbH, 99.3% purity) in a pure Ar at-
mosphere (99.999% purity) using direct current magnetron sputtering
in an in-house developed deposition system (base pressure below
1.0 × 10
−4
Pa). The Si content was adjusted by placing 0, 2, 4, 6, 8, 12, or
16 Si platelets (3.5 × 3.5 × 0.38 mm) on the target racetrack. The ro-
tating substrate holder (0.25 Hz) was positioned at a target-to-sub-
strate distance of 90 mm. All thin films were grown on Si ((100)-
oriented, 20 × 7 × 0.38 mm), single crystalline Al
2
O
3
((1−102)-oriented,
10 × 10 × 0.53 mm), and poly-crystalline Al
2
O
3
(20 × 7 × 0.38 mm)
substrates, which were ultrasonically pre-cleaned in acetone and iso-
propanol, respectively. Following a heating sequence to a substrate
temperature of 550 °C, an Ar-ion etching step was performed at a total
pressure of 5 Pa and an applied substrate bias potential of −800 V for
10 min. The target and Si alloying platelets were sputter-cleaned for
3 min prior to all depositions to reduce oxygen contamination. The thin
films were then grown at a total Ar pressure of 0.7 Pa, a target current
of 0.4 A (corresponds to a power density of ∼5 W/cm
2
), and a bias
potential of −40 V. The synthesis conditions resulted in deposition rates
of 16.1 and 18.1 nm/min for 0 and 16 Si platelets on the target surface.
Aiming for a consistent coating thickness of ∼3 µm, deposition times of
190 and 170 min were selected, respectively. Deposition times for in-
termediate Si compositions were calculated from linear interpolation,
resulting in a total thickness variation of ± 0.1 µm between all samples.
The overall chemistry of the Cr-Si-B
2 ± z
coatings was obtained
using liquid inductively coupled plasma-optical emission spectro-
scopy (ICP-OES). A detailed explanation of this methodology is given
in Ref. [28]. Structural analysis was performed by X-ray diffraction
on a PANalytical XPert Pro MPD equipped with a Cu-K
α
radiation
source (wave-length λ = 1.54 Å, operated at 45 kV and 40 mA) in
Bragg-Brentano geometry. The cross-sectional growth morphology
was further investigated by scanning-electron microscopy (ZEISS
Sigma 500VP, operated at 5 kV) on coated Si substrates.
The hardness and elastic modulus of all coatings was studied
using instrumented nanoindentation (ultra-micro indentation
system, UMIS) with a minimum of 30 load-displacement curves
evaluated according to Ref. [31] for each coating. Increasing in-
dentation loads ranging from 5 to 22 mN (steps of 0.5 mN), with
additional measurements up to 45 mN to probe for any substrate
influence, were applied. Moreover, the obtained E values were fitted
over the indentation depth using a power law function and extra-
polated to the sample surface to receive the film-only modulus [32].
Macro-stresses in the coatings were additionally analyzed through
curvature measurements using optical profilometry (PS50, Nanovea)
and the modified Stoney equation [33,34]. All mechanical properties
were determined on coated sapphire substrates.
The oxidation behavior of the Cr-Si-B
2 ± z
thin films was de-
termined from thermogravimetric analysis (TGA, Netzsch STA 449
F1, equipped with a Rhodium furnace) performed on coated poly-
crystalline Al
2
O
3
substrates. The substrates were weighed before and
after deposition to determine the coating-only mass. This value
serves as a reference during dynamic oxidation experiments up to
1400 °C (heating rate of 10 °C/min) in a synthetic air environment
(50 ml/min flow rate). Any oxidation-related mass change was re-
corded at a resolution of 0.1 µg. Pre-tests on uncoated Al
2
O
3
sub-
strates additionally proved their inertness during the oxidation
treatment [28].
Additional oxidation experiments combined with in-situ X-ray
diffraction analysis were carried out on a PANalytical XPert Pro MPD
(Cu-K
α
radiation source, wave-length λ = 1.54 Å, operated at 45 kV
and 40 mA) in Bragg-Brentano geometry using an Anton Paar high-
temperature furnace chamber (HTK 1200 N). Measurements were
taken in a lab-air environment (0.3 l/min flow rate) at room-tem-
perature and from 400 to 1200 °C in 50 °C steps. The sample was
heated at a rate of 50 °C/min between the individual temperature
steps, with each diffraction measurement taking ∼21 min
Furthermore, detailed microstructural and chemical analysis on
selected oxidized samples was performed using transmission elec-
tron microscopy (TEM, FEI TECNAI F20, operated at 200 kV). Bright-
field (BF) and high-angle annular dark field (HAADF) imaging are
utilized to gain information on the microstructure and oxide scale
growth. In addition, energy dispersive X-ray spectroscopy (EDX)
performed in scanning TEM (STEM) revealed the chemical compo-
sition of the entire coating cross-section as well as the oxide layer.
Density functional theory (DFT) coded VASP [35,36] calculations
(projector augmented waves method within the generalized gra-
dient approximation [37]) were performed to study the preferred
atomic configuration for silicon alloying atoms within various AlB
2
-
structured Cr-Si-B
2
compositions. Moreover, the influence of
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
2

increasing Si content as well as the impact of various vacancy con-
figurations on the phase stability were investigated. The influence of
vacancies was studied up to a content of 2 Si atoms within the su-
percell (corresponds to ∼4 at.%), with 2 vacancies introduced either
on the Cr-sublattice, the B-sublattice, or as a Schottky defect. All
2 × 2 × 4 supercell structures (16 metal and 32 boron atoms) were
generated using the special quasi-random structure (SQS) approach
[38]. Values for the formation energy were only extracted from fully
converged supercells. A plane wave cut-off energy of 600 eV and an
automated k-point mesh (length = 60) were chosen to provide a total
energy accuracy of about 10
−3
eV/at. All calculations were conducted
without considering the paramagnetic states of Cr.
Finally, atom probe tomography (APT) analysis was performed on
an exemplary coating in the as-deposited state to reveal the initial
elemental distribution. Sample preparation involved milling of an
initial coating pillar and sharpening to a tip using a focused ion beam
microscope FEI Scios 2 DualBeam operated at 30 kV and stepwise
decreasing milling currents. Final tip sharpening was performed at
50 pA, with a subsequent clean-up step at 5 kV and 28 pA to
minimize possible Ga
+
ion-induced damage. Subsequent APT ana-
lysis was carried out on a CAMECA LEAP 4000X HR in pulsed laser
mode with a set pulse energy of 50 pJ. The system uses a 355 nm UV
laser equipped with a reflection lens, resulting in a detection effi-
ciency of ∼37%. The sample was cooled to a constant temperature of
44 K. Experiments were performed with a target evaporation rate of
1% and pulse repetition rate of 200 kHz. Data analysis was conducted
using an open-source Matlab Toolbox for APT data evaluation [39].
3. Results & discussion
3.1. Computational phase formation & stability boundaries
A regular requirement for alloying strategies to successfully im-
prove the oxidation resistance of a coating material involves un-
altered phase stability for the host structure to maintain a distinct
property profile. This necessity is demonstrated by the well-studied
Ti
1−x
Al
x
N system, where the oxidation resistance of rock-salt struc-
tured TiN scales with the AlN alloying fraction [40]. However, upon
exceeding the Al solubility threshold on the metal sublattice (x ∼
67% for DCMS deposited thin films), precipitation of the thermo-
dynamically favored wurtzite-structured Al
1−x
Ti
x
N phase occurs,
thus deteriorating both the thermal stability and mechanical prop-
erties. Analogously, to probe the effect of an increasing Si alloying
content on the phase stability of prototypical Cr-Si-B
2
thin films from
a theoretical point of view, Fig. 1a presents ab initio calculated for-
mation energies for various AlB
2
structured compositions. This
evaluation allows to assess the influence of Si-addition on the phase
stability but also provides information on the energetically preferred
lattice occupation within the unit cell and hence a guideline for the
Si solubility limit.
The model assumes that up to 8 Si atoms are either placed in-
terstitially or substitutionally within the AlB
2
-structured Cr-Si-B
2
.
Over the entire compositional range, DFT calculations associate the
formation of all possible alloying configurations with an increasing
E
f
compared to the binary CrB
2
composition. This implies an overall
reduced stability of the hexagonal structure with increasing Si
content. In more detail, all structures where Si is placed interstitially
either within B- or Cr-planes show the most substantial increase in
E
f
, already leading to positive formation energies upon alloying 2 Si
atoms (equals ∼4 at.%). Hence, these structures are energetically
unstable, and all substitutional configurations are significantly more
favored.
The structures where Si is placed in substitution for B/ Cr/ or
equally on both sublattices exhibit very similar formation energies in
the low alloying regime and thus can co-exist without any preferred
atomic position of Si. However, upon increasing the alloying content
beyond 3 atoms (equals values ≥6 at.%), structures with Si sub-
stituting solely B atoms and later also upon replacing both B and Cr
equally become unstable and undergo a phase transformation (i.e.,
converge into a different structure type) – see red and grey crosses in
Fig. 1a. Only calculations where Si replaces Cr exclusively within the
supercell yield negative formation energies up to 4 Si atoms (equals
∼8 at.%), while maintaining the hexagonal configuration. In fact,
compositions with up to 8 Si alloying atoms on the Cr sublattice
relax in the AlB
2
-type structure, although thermodynamically un-
stable due to positive E
f
. Consequently, these predictions also in-
dicate a theoretical Si solubility limit within AlB
2
-structured CrB
2
above 4 Si atoms, which equals an overall alloying content of about
8 at.%.
Another factor to consider is the presence of point defects, as
they are typically related to PVD synthesized films, which can
strongly influence the phase stability criteria compared to the
thermodynamic equilibrium [41]. Therefore, three different vacancy
configurations – either two Cr vacancies, two B vacancies, or one
Schottky defect – were analyzed for Cr-Si-B
2
structures with up to 2
alloyed Si atoms. The differences in energy of formation between the
defected and the corresponding prototypical structure
(
=E E E
ff
def
f
prot
) are presented in Fig. 1b. In perfect agreement
with the findings of Moraes et al. [42], the binary CrB
2
compound
favors the formation of boron and boron-containing vacancies over
Fig. 1. (a) DFT-calculated E
f
per atom for prototypical Cr-Si-B
2
structures (AlB
2
-type)
with various Si contents. The lower x-axis gives the number of Si atoms within the
employed supercell, whereas the upper x-axis gives the corresponding atomic con-
centration. The Si atoms are either positioned interstitially on Cr- or B-planes, or in
substitution for Cr, B, or an equal fraction of both Cr and B atoms. Crossed data points
indicate alloying-induced deviations from the prototype structure during supercell
relaxation. (b) DFT-evaluated formation energy differences per atom between the
prototypical Cr-Si-B
2
structures in (a), with up to two Si atoms replacing either Cr
(plane bars) or B (striped bars) atoms in defected structures that hold either two Cr
vacancies (blue bars)/ two B vacancies (red bars)/ or one Schottky defect (grey bars),
respectively.
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
3

Cr point defects. The calculations further confirm this trend for both
Si alloyed compositions, again indicating the preferred incorporation
of B over Cr vacancies, except for the case of two B atoms exchanging
Si. There, both vacancy types contribute to increased stability by
slightly lowering E
f
. Overall, the DFT calculations suggest the pre-
ferred incorporation of synthesis induces point defects on the non-
metal sublattice for all compositions, with only minor negative in-
fluence from transition-metal vacancies.
3.2. Structural & morphological properties
The chemical composition of all Cr-Si-B
2 ± z
thin films is presented
within a ternary phase diagram in Fig. 2. The diagram is extended
with guidelines connecting stoichiometric CrB
2
with single-phased
Si and CrSi
2
(endpoints not visible due to reduced axis ranges),
corresponding to narrow two-phase fields according to the equili-
brium phase diagram [43]. The chemical analysis revealed an in-
creasing silicon content of 0, 1, 3, 8, 11, and 17 at.% for the alloyed thin
films with an increasing number of Si platelets placed on the target
racetrack, respectively. The synthesis approach allowed for a pre-
dictable and linear adjustment of the Si content within the resulting
thin film compositions. The coating prepared with two Si platelets
on the target surface obtains an effective Si content below the de-
tection limit of the employed ICP-OES method, thus the coating is
referenced with a content of 0 at.% Si (Cr
0.32
Si
0.00
B
0.68
). Nevertheless,
compared to the unalloyed coating, a minute fraction of Si is still
expected within this thin film.
During the PVD deposition of (ternary) compound materials,
coatings usually become enriched or depleted in specific con-
stituents due to their preferred sputtering or scattering behavior
within the plasma [26]. Interestingly for the Cr-Si-B
2 ± z
coatings,
increasing the Si content in the thin films leads to a stoichiometric
(B:Cr = 2:1) replacement of the CrB
2
mole fraction. This is also in-
dicated by the direct overlap of all data points with the connecting
line between stoichiometric CrB
2
and pure Si in Fig. 2. Moreover, the
Si-free coating Cr
0.34
B
0.66
obtains an almost nominal stoichiometry
with a B:Cr-ratio close to 2:1, which is consequently preserved for all
further Si-containing depositions. Nevertheless, with increasing Si
content in the Cr-Si-B
2 ± z
thin films, the overall B content decreases
from 66 at.% down to 55 at.% for the coating with the highest Si-
content.
The X-ray diffractograms depicted in Fig. 3 demonstrate that all
Cr-Si-B
2 ± z
thin films, regardless of their chemical composition,
adopt the hexagonal AlB
2
-type structure (space group 191). More-
over, within the accuracy of the employed method, no additional
phases could be determined for any coating. All thin films obtain a
polycrystalline growth, with slightly preferred orientations notice-
able for coatings with a Si content below 8 at.%. Within these sam-
ples, the preferred orientation shifts from (101) for the Si-free
Cr
0.34
B
0.66
coating, towards (001) for Cr
0.32
Si
0.00
B
0.68
, to (100)-or-
iented for both the Cr
0.32
Si
0.01
B
0.67
and Cr
0.32
Si
0.03
B
0.65
thin films,
respectively. Further increasing the Si content results in equally or-
iented grains and causes a reduction of the diffracted intensities,
hinting towards a concomitant decrease in the coherently diffracting
domains (i.e., a reducing grain size). This structural evolution cor-
relates well with the DFT calculated Si solubility threshold close to
8 at.% (see Fig. 1), thus suggesting that the excess alloying fraction
preferably occupies grain boundary sites while also rationalizing
their suggested increase in volume fraction due to smaller grains.
The data further reveals that incorporating Si into the CrB
2
host
structure leads to a slight decrease of the lattice parameter c in
(001)-direction from 3.00 to 2.97 Å, meaning that the bond distance
between adjacent B and Cr lattice planes is reduced. On the other
hand, the lattice parameter a remains relatively unchanged at 2.97 Å,
hence lateral bond distances between similar atoms are maintained.
Further correlating the phase formation with the chemical
composition of all coatings shows that the decrease in B content
with increasing Si fraction takes no influence on the stability of the
hexagonal CrB
2
structure in the as deposited state. Considering that
only minor quantities of Si are chemically stable when located on the
B-sublattice, the missing B-fraction is likely accommodated by in-
troducing vacancies on the non-metal sublattice during the de-
position process (compare with Cr-rich/B-deficient planar defects
previously observed in CrB
1.90
[46]). This is also in excellent agree-
ment with the above DFT calculations, where enhanced thermo-
dynamic stability is indicated for all structures containing B-
Fig. 2. Ternary phase diagram showing the chemical composition of all synthesized
Cr-Si-B
2 ± z
coatings. The dashed lines connecting CrB
2
with Si and CrSi
2
indicate
narrow two-phase fields according to the equilibrium phase diagram [43]. The axis
ranges are reduced for improved separation of the data points.
Fig. 3. X-ray diffractograms of all Cr-Si-B
2 ± z
coatings, arranged with increasing Si
content from bottom to top. The diffractographs are correlated with standardized
reference patterns for hexagonal CrB
2
(space group 191, AlB
2
prototype, [44]) and
cubic silicon (substrate material, [45]). The corresponding coating chemistry for each
diffractograph is included on the left side.
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
4

vacancies – especially pronounced when Si is introduced on the Cr-
sublattice.
The influence of Si on the growth mode of Cr-Si-B
2 ± z
thin films,
specifically the decreasing columnar crystallite size, is further illu-
strated in Fig. 4 through selected SEM fracture cross-sectional stu-
dies. The non- and low-alloyed Cr
0.34
B
0.66
and Cr
0.32
Si
0.03
B
0.65
coatings show a pronounced columnar structure with grains ex-
tending throughout the entire cross-section. The coating with
∼11 at.% Si shows denser and increasingly more fibrous crystal col-
umns – see Fig. 4c. With the highest alloying content of 17 at.%, the
growth morphology appears featureless, with limited indications for
individual columnar structures remaining.
3.3. Atom probe tomography
To obtain an improved view of the distribution of Si atoms within
the Cr-Si-B
2 ± z
thin films – especially at concentrations close to the
proposed solubility limit – an additional coating with a composition
of Cr
0.27
Si
0.09
B
0.64
was analogously prepared and investigated using
detailed atom probe tomography. Fig. 5 shows reconstructions of the
atomic positions for Cr, Si, and B recorded within the tip volume. A
random distribution is observed for both Cr and B atoms in the
entire volume, although slight clustering of Cr atoms is noticeable in
certain regions. In contrast, local chemical analysis of the Si dis-
tribution reveals the formation of Si-enriched regions at defect sites,
identified as grain boundaries and triple junctions within the as-
deposited thin film. A concentration profile in the grain interior (see
Fig. 5i and Supplementary) shows an entirely homogeneous dis-
tribution of the constitutional elements within the undisturbed,
crystalline region. The calculated average chemistry reveals a com-
position close to Cr
0.37
Si
0.04
B
0.60
, thus indicating that the actual Si
solubility within the CrB
2
-structure could be even lower than the
DFT calculated limitation. The corresponding concentration profile
taken at a grain boundary location (see Fig. 5ii and Supplementary)
depicts an increased fluctuation of all elements and a drastically
increased Si content of up to 30 at.% in specific locations. The latter
findings clearly underline the preferred segregation of surplus Si
during the deposition process.
Overall, these findings experimentally underpin the DFT calcu-
lated solubility threshold above 8 at.% Si for the analyzed sample
composition. Evidently, the conducted calculations neglect the
possible impact of temperature and can only incorporate the che-
mical as well as kinetic limitations during PVD synthesis to a limited
extent, thus rationalizing the deviation from the experimentally
observed solubility threshold of ∼4 at.%. Nevertheless, within the
accuracy of the conducted analysis, the agreement between the
chemical composition and DFT calculations is clearly given.
Moreover, the proposed influence of Si segregation to promote grain
refinement, as evidenced in XRD analysis above an alloying content
of 3 at.%, is additionally confirmed.
3.4. Mechanical properties
Fig. 6 presents the mechanical properties of all Cr-Si-B
2 ± z
thin
films deposited. The residual stress state, film hardness, and Young’s
modulus are plotted as function of the Si alloying content. The Si-
free Cr
0.34
B
0.66
coating obtains a compressive residual stress state
with σ ∼ − 0.5 GPa and corresponding hardness and Young’s modulus
values of H = 23.5 ± 2.7 GPa and E = 295 ± 18 GPa, respectively.
When compared to other TMB
2
materials, such as TiB
2
[47,48] or
WB
2
[49], CrB
2
typically features a reduced hardness and a relatively
low elastic modulus [50]. Nevertheless, related works have also re-
ported vastly higher hardness values for this material systems using
similar deposition techniques, yet the origin of the observed varia-
tion remains unresolved [51–53]. Alloying a minute fraction of Si
into CrB
2
shows a reversed residual stress state, with Cr
0.32
Si
0.00
B
0.68
revealing a tensile stress of σ ∼ 0.9 GPa. Interestingly, despite the
adverse effect of tensile stresses on the measurable hardness, this
coating shows identical nanoindentation results with H = 23.9 ± 1.1
and E = 291 ± 6 GPa. The maintained properties are related to the
preferred orientation rather than the influence of Si alloying per se.
In line with a work by Fuger et al. [54], the preserved hardness is
explained by the pronounced orientation towards the (001)-direc-
tion (see Fig. 3), which was demonstrated to yield the highest
hardness for TMB
2
thin films in general. Thus, the anisotropy effect
balances the negative impact of the tensile stress state. Conse-
quently, even higher hardness values could be expected for this
material system by tailoring the residual stress state towards the
compressive regime. The actual shift in the residual stress state
between the Cr
0.34
B
0.66
and Cr
0.32
Si
0.00
B
0.68
coating may also be re-
lated to the preferred orientation. CrB
2
obtains a significant aniso-
tropy in the thermal expansion (
a
CrB2
= 10.8
×
10
−6
K
−1
,
c
CrB2
= 6.3
×
10
−6
K
−1
, [55]), therefore higher in-plane tensile stresses are to be
expected for (001)-textured thin film (a-direction parallel to coating-
substrate interface) when grown on sapphire substrate (
a
Al O
2 3
=
4.5–5.3
×
10
−6
K
−1
, [56]). With a further increase in Si, the residual
tensile stress is gradually reduced from σ ∼ 0.9 GPa for
Cr
0.32
Si
0.01
B
0.67
down to σ ∼ 0.3 GPa for Cr
0.29
Si
0.11
B
0.60
, before again
increasing in the compressive regime to σ ∼ − 0.5 for Cr
0.28
Si
0.17
B
0.55
.
Concomitantly, with the above observed increase in grain boundary
volume – i.e., an increase in regions that are less strongly bound than
the surrounding crystal – a linear decrease in the elastic modulus
down to E = 238 ± 7 GPa for the highest Si content of 17 at.% was
recorded. Interestingly, Si alloying did not result in any solid solution
hardening effect during nanoindentation. Upon introducing more
than 1 at.% Si into CrB
2
, the hardness gradually decreases from H
= 21.6 ± 1.1 GPa for Cr
0.32
Si
0.01
B
0.67
, down to a constant value of H ∼
17 GPa for all coatings having a Si content ≥ 8 at.%. In relation to the
Si segregations observed on grain boundaries for higher alloying
compositions (see Fig. 5), the measured hardness of these thin films
is likely dominated by the inferior mechanical properties of the Si-
rich regions.
3.5. Dynamic oxidation
Dynamic oxidation experiments were conducted in a TGA system
to revisit [28] the impact of Si alloying on the high-temperature
oxidation resistance of CrB
2
-based thin films, and to determine the
minimum alloying quantity necessary to yield enhanced protective
properties. Fig. 7 presents the mass change of the coating material,
deposited onto pre-weighed polycrystalline Al
2
O
3
substrate (inert in
the temperature rage up to 1400 °C, [28]), with respect to the an-
nealing temperature T. Up to a range of T ∼ 600 °C a constant mass
Fig. 4. SEM images depicting the growth morphology of selected Cr-Si-B
2 ± z
coatings
deposited on Si substrate including the corresponding chemical composition.
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
5

signal is recorded for all Cr-Si-B
2 ± z
thin films, indicating no pro-
gressive oxide scale formation. Only the Cr
0.32
Si
0.01
B
0.67
thin film
shows an earlier onset temperature at T ∼ 500 °C (blue dotted line),
visible by the already occurring mass gain due to oxide scale growth.
Upon reaching the oxidation onset temperature, all Cr-Si-B
2 ± z
coatings up to a Si content of 3 at.% show a stepwise increase in the
mass signal until reaching a maximum value at T ∼ 1200 °C, in-
dicating the fully oxidized state. Beyond this temperature, a de-
creasing sample mass is recorded, which is correlated with the
evaporation of B
2
O
3
-based oxides. The mass signal evolution shows
intermediate plateaus for all these coatings between
600 < T < 1200 °C, hinting towards a competitive formation of B
2
O
3
-
and Cr
2
O
3
-based scales and thus limited protection against con-
tinued oxidation. Moreover, with already small alloying fractions
(e.g., Cr
0.32
Si
0.03
B
0.65
), both the slope of the increase and the overall
Fig. 5. Atom probe tomography determined chemical composition of a Cr
0.27
Si
0.09
B
0.64
thin film. Reconstructed positions of Cr, Si, and B atoms are presented. Insets (i) and (ii)
depict concentration profiles of the grain interior and grain boundary, respectively. Both profiles were collected in the corresponding regions of interest marked in the Si atom
distribution. The supplementary material contains animations of the Si distribution to provide an improved view on both regions of interest.
Fig. 6. (a) Residual stress state of all Cr-Si-B
2 ± z
thin films versus the Si alloying
content. (b) Corresponding hardness and Young’s modulus data. All mechanical
properties were determined on coated sapphire substrates.
Fig. 7. Mass change of all Cr-Si-B
2 ± z
coatings as function of the annealing tempera-
ture, recorded during dynamic oxidation in a TGA system (10 K/min heating rate) in
synthetic air environment. The coatings were deposited on pre-weighed Al
2
O
3
sub-
strates, which are inert over the entire temperature range.
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
6

mass gain up to T ∼ 1100 °C are significantly reduced, already
pointing towards the effectiveness of the employed alloying routine.
Increasing the Si content within the Cr-Si-B
2 ± z
thin films beyond
8 at.% leads to a fully preserved coating mass up to T ∼ 1100 °C due to
the formation of a stable, protective oxide scale preventing any
oxidative attack of the underlying coating material. Only at
T > 1100 °C, these higher alloyed thin films show a slight increase in
the overall mass signal up to 1400 °C. Unlike the low-alloyed coat-
ings, the protective mechanism relies on the formation of a con-
tinuous, dense SiO
2
-based scale enabled by a sufficient Si diffusion
provided through the Cr-Si-B
2 ± z
thin film. Moreover, as known for a
Cr
0.26
Si
0.16
B
0.58
coating [28], the formed oxide scale after annealing
at T = 1400 °C should in fact be comprised of a layered amorphous
SiO
2
-based phase with a crystalline Cr
2
O
3
scale on top (discussed in
more detail in Section 3.7). Overall, these results highlight that
achieving high-temperature oxidation resistance for Cr-Si-B
2 ± z
thin
films involves a minimum alloying content close to 8 at.% Si to ac-
tivate the protective mechanism.
3.6. In-situ X-ray diffraction
Comparative in-situ X-ray diffraction studies were performed
during the oxidation of Cr
0.32
Si
0.03
B
0.65
and Cr
0.29
Si
0.11
B
0.60
in lab-air
environment, to reveal the underlying mechanism causing the
drastically improved oxidation resistance above a distinct Si content.
Fig. 8 depicts the diffractographs taken at room temperature (RT) as
well as from 400 to 1200 °C in steps of 50 °C, with the corresponding
annealing temperature included on the right axis. The data for an
“insufficiently” alloyed Cr
0.32
Si
0.03
B
0.65
coating (see Fig. 8a) reveal an
unaltered crystal structure up to a temperature of T = 550 °C, de-
picting a preferred (100)-orientation as shown in Fig. 3. With the
oxidation onset at T = 600 °C (see also Fig. 7), initial indications to-
wards a boron depleted CrB phase are formed (e.g., 2θ ∼ 32.2°, 38.5°,
44.9°, etc.), increasing in intensity up to T = 1100 °C. This suggests the
partial decomposition of the Cr-Si-B
2
structure to form an un-
protective B
2
O
3
scale, which is in line with the mass gain to an initial
plateau observed during the TGA measurements. In the temperature
regime beyond T = 750–800 °C, additional recrystallization of the
remaining Cr-Si-B
2 ± z
solid solution is observed, as indicated by the
decreasing peak width and increase in diffracted intensities (e.g., 2θ
∼ 45.8°). Similar behavior was previously reported for amorphous
Cr-Al-Si-B-(N) coatings, experiencing crystallization of the CrB
2
phase at T = 800 °C [57]. At T = 1050 °C the diffraction signals for the
CrB
2
structure diminish, pointing towards a full decomposition of
the diboride phase. Finally, at T = 900 °C, an additional Cr
2
O
3
scale is
formed – compare with the second mass gain plateau observed in
Fig. 7 – which subsequently consumes the entire coating at
T = 1200 °C. Over the entire temperature range, no Si-based phase is
observed.
A direct comparison to results obtained for a “sufficiently” al-
loyed Cr
0.29
Si
0.11
B
0.60
(Si content ≥ ∼8 at.%) coating is illustrated in
Fig. 8b. In the as-deposited state at RT, the data shows an analogous
diffraction result as depicted in Fig. 3, revealing a hexagonal struc-
tured CrB
2
-based coating with polycrystalline grain distribution.
Moreover, no additional Cr-Si- or Si-B-based phase is detected,
suggesting that Si is dissolved up to the solubility limit of ∼3–4 at.%
Fig. 8. In-situ X-ray diffractographs recorded during subsequent annealing treatments in lab-air environment of (a) Cr
0.32
Si
0.03
B
0.65
and (b) Cr
0.29
Si
0.11
B
0.60
thin films deposited on
polycrystalline Al
2
O
3
. Standardized reference patterns for hexagonal CrB
2
(blue hexagon, [44]), cubic Si (light blue square, [45]), hexagonal CrSi
2
(dark yellow triangle, [58]),
rhombohedral Cr
2
O
3
(green diamond, [59]), orthorhombic CrB (dark red circle, [60]) and rhombohedral Al
2
O
3
(grey star, [61]) are included. The sample temperature corresponding
to each diffraction experiment is added on the right axis.
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
7

within the CrB
2
structure – note that excess Si is located at grain
boundaries as shown in Fig. 5. The structure is fully preserved up to a
temperature of T = 600 °C. A further increase to T = 650 °C leads to
first indications for crystalline Si precipitates (e.g., 2θ ∼ 28.7°, 47.5°,
56.3°). Analogously to Cr
0.32
Si
0.03
B
0.65
, recrystallization of the CrB
2
structure is observed beyond T = 750–800 °C (e.g., 2θ ∼ 29.0°, 34.7°,
45.8°, etc.), being more pronounced due to the absence of a CrB
phase. Furthermore, in accordance with the three-phase field of
CrB
2
-CrSi
2
-Si in Fig. 2 (area between the dashed lines), the inter-
mediate formation of a CrSi
2
phase is suggested between T = 750 and
1050 °C by a set of low intensity reflexes (e.g., 2θ ∼ 26.8°, 42.3°,
49.6°). However, it should be noted that the deviation from the in-
dexed peak positions is significant, so that additional high-resolution
analysis would be required for confirmation. Recrystallization of the
Cr-Si-B
2 ± z
matrix and Si precipitation continue up to T = 1200 °C,
resulting in sharp peaks for both phases. In addition, several in-
dications for a Cr
2
O
3
oxide layer emerge after the annealing ex-
periments above T ≥ 1100 °C (e.g., 2θ ∼ 33.5°, 50.0°, 54.5°), yet no
diffraction peaks pointing towards a B
2
O
3
or the more important
SiO
2
-based structure occur. Thus, in line with previous findings,
especially the latter phase is expected to be in an amorphous state.
Overall, pronounced diffraction peaks indicate that the original Cr-
Si-B
2 ± z
structure is still intact at T = 1200 °C, highlighting the ex-
cellent oxidation resistance of this coating and confirming the pre-
sence of a stable oxide scale protecting the underlying coating
material.
When discussed in relation to the dynamic oxidation experi-
ments (see Fig. 7), the precipitation of Si at T > 600 °C in this “suf-
ficiently” alloyed coating correlates well with the oxidation onset
temperature of the “insufficiently” alloyed samples. Consequently,
this provides a strong indication that Si precipitates within the
Cr
0.29
Si
0.11
B
0.60
coating – either derived from excess Si on grain
boundaries or the surrounding Cr-Si-B
2 ± z
solid solution (discussed
in more detail in Section 3.7) – are the primary source for the in-
creased oxidation resistance. In addition, the formation of crystalline
Cr
2
O
3
above T = 1100 °C is in excellent agreement with the previous
TGA analysis, rationalizing the mass gain for all coatings with Si
content above 8 at.% in the same temperature regime.
Regarding the thermally activated precipitation of Si from the
Cr
0.29
Si
0.11
B
0.60
coating, a possible explanation is seen in the con-
tinuous increase in the DFT calculated E
f
for the AlB
2
structured Cr-
Si-B
2
compositions over the binary CrB
2
with increasing Si content.
Theoretically comparing the difference in energy of formation be-
tween a Cr
0.27
Si
0.06
B
0.67
structure (see Fig. 1) with its corresponding
decomposition products of CrB
2+z
and Si according to:
= + +
=
E x E x E x E E
x
[(1 ) 2 ] ,
with 0. 06
ff
CrB
f
Si f
Bf
Cr Si B
20.27 0. 06 0.67
(1)
a significant energetic benefit of
Ef
= −1.34 eV/at. towards the de-
composed constituents is attained. The DFT calculated formation
energies of elemental Si (cubic, −5.41 eV/at.) and B (rhombohedral,
−6.67 eV/at.) correspond to their stable configuration at room-tem-
perature and ambient pressure.
Analogous results are obtained for all other theoretically and
experimentally studied compositions of Cr-Si-B
2 ± z
, thus underlining
that the precipitation follows the thermodynamically prescribed
equilibrium condition. Finally, the precipitation is believed to be
further supported during recrystallization of the Cr-Si-B
2 ± z
phase
above T = 750–800 °C, allowing for even enhanced Si diffusion.
Considering a melting temperature of T
M
= 2200 °C [62] for pure
CrB
2
, the recrystallization process occurs at a typical homologous
temperature of T
H
∼ 0.4.
Fig. 9. (a) bright-field and (b) high-angle annular dark field TEM micrographs of a Cr
0.29
Si
0.11
B
0.60
coating on polycrystalline Al
2
O
3
substrate, gradually oxidized up to 1200 °C (see
Fig. 8). (c) elemental EDX mapping of the entire sample cross-section according to the insert in (a). (d) detailed scanning-TEM image of the oxide scale indicated in (a) including
the elemental EDX mapping.
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
8

3.7. Structural & chemical analysis post annealing
In order to complete the established viewpoint on the mor-
phological evolution and especially the oxide scale growth during
high-temperature oxidation of Cr-Si-B
2 ± z
thin films, com-
plementary detailed TEM analysis (see Fig. 9) is performed on the
Cr
0.29
Si
0.11
B
0.60
thin film used during the in-situ X-ray studies (see
Fig. 8b). Fig. 9a and b depict bright-field and high-angle annular
dark field micrographs of the entire sample cross-section, including
the interface to the polycrystalline Al
2
O
3
substrate and the formed
oxide scale, respectively. Both images immediately visualize the
pronounced recrystallization of the Cr-Si-B
2 ± z
thin film, revealing
large globular grains throughout the cross-section. Given the
atomic number contrast in the HAADF image, regions of different
elemental compositions – indicated by lighter and darker grey
areas – can be identified next to several black appearing voids.
When combined with the elemental mapping in Fig. 9c, bright
areas can be correlated with a Cr- and B-rich phase (i.e., CrB
2 ± z
),
whereas darker regions correspond exclusively to pure Si. Note the
superposition of Cr and O signals during EDX analysis, thus creating
a slight, artificial O signal (see Fig. 9c-iv) overlapping with all
CrB
2 ± z
regions. Furthermore, also Si and W overlap in the EDX
spectrum, resulting in a misinterpretation of the W protection layer
with an artificial Si region in the top area of Fig. 9c-ii. Several
conclusions can be drawn from these results: (i) Silicon precipita-
tion is not restricted to grain boundary sites already holding excess
Si in the as-deposited state. (ii) Upon thermal activation, the
Cr-Si-B
2 ± z
solid solution fully decomposes into large globular
phase regions containing solely CrB
2 ± z
or Si. (iii) Globular Si pre-
cipitates are formed throughout the coating cross-section in addi-
tion to a continuous surface layer. As a result, several voids are
formed in the thin film volume to compensate for the Si surface
diffusion (note, certain voids may also originate from focus-ion
beam milling preparation of the TEM lamella, resulting from the
weak connection between the individual recrystallized grains). Fi-
nally, the elemental distributions of Al and O show that no inter-
action between the coating and substrate material occurred during
the entire oxidation treatment.
Fig. 9a and b also clearly show a thin, dense oxide scale formed
on the sample surface. Using detailed STEM imaging combined with
EDX analysis (see Fig. 9d), the oxide reveals a defined, layered ar-
chitecture composed of a thin amorphous SiO
2
layer on the coating-
oxide interface and a nanocrystalline Cr
2
O
3
top layer. Similar to the
elemental distribution in Fig. 9c, no intermixing of Cr- and B-rich
sites with Si can be observed in the oxide layer and the unaffected
material below. However, it has to be considered that the employed
chemical analysis is not suitable for tracing minimum quantities of
light elements such as B within, e.g., the SiO
2
layer. Interestingly,
despite the extended annealing time at temperatures above
T > 1000 °C, the SiO
2
layer features a thickness in the range of only
20–40 nm, whereas the Cr
2
O
3
top layer extends over 200–250 nm.
Regarding the temporal sequence of forming the highly pro-
tective oxide scale, the primary mechanism is seen in the pre-
cipitation of Si – especially towards the coating surface – allowing
for the initial growth of a stable SiO
2
layer in the temperature range
from T = 650–1100 °C. Due to the minimal thickness of this layer,
even at T = 1200 °C, no mass gain is visible in the TGA signal.
Furthermore, in line with the TGA and in-situ X-ray diffraction
analysis, the additional Cr
2
O
3
surface layer is subsequently formed
in the temperature regime above T = 1100 °C. Unlike many TM-
oxides, Cr-cations primarily diffuse outwards on grain boundaries
within the Cr
2
O
3
oxide layer, thus allowing for a scale growth on
top of the oxide surface rather than the oxide-coating interface
[63]. Also, resulting from the vastly increased layer thickness, the
formation is clearly relatable to the mass gain signal shown
in Fig. 7.
4. Conclusion
Si alloying was proven a successful concept to significantly en-
hance the oxidation resistance of transition-metal diboride-based
thin films. In this work, DC magnetron sputtered Cr-Si-B
2 ± z
coatings
with Si content up to 17 at.% were analyzed to reveal the impact of
the alloying element on the structural and mechanical properties of
the AlB
2
-type thin films. In addition, the mechanisms leading to the
enhanced oxidation resistance were investigated to deepen the
knowledge on optimized chemical compositions.
DFT calculations performed on various stoichiometric and de-
fected AlB
2
-structured Cr-Si-B
2
compositions indicate the en-
ergetically favored incorporation of Si on the Cr-sublattice over a
wide alloying range. Contrary, already limited occupation of the B-
sublattice destabilizes the hexagonal cell. Detailed APT analysis of a
Cr
0.27
Si
0.09
B
0.64
thin film revealed Si segregation towards grain
boundaries in the as-deposited state, while the grain interior holds
up to 4 at.% Si, being in line with a DFT-calculated solubility limit.
Despite a concomitant increase in B under-stoichiometry with
increasing Si content, all synthesized Cr-Si-B
2 ± z
coatings obtain the
hexagonal AlB
2
structure, irrespective of the chemical composition.
Moreover, increasing the Si content is accompanied by a variation of
the preferred growth orientation and a gradual reduction in the
average columnar grain size. Nanoindentation measurements
showed a direct correlation between the morphological features and
the mechanical properties. The highest film hardness was recorded
for low Si alloyed, (001)-oriented coatings at H ∼ 24 GPa, whereas an
increased alloying content of Si ≥ 8 at.% resulted in H ∼ 17 GPa due to
mechanically weak Si grain boundary segregates.
Thermogravimetric analysis proofed the excellent oxidation re-
sistance of Cr-Si-B
2 ± z
thin films with Si content ≥ 8 at.% up to T ∼
1400 °C, whereas lower alloyed coatings suffer from stepwise oxi-
dation above T ∼ 600 °C related to a non-protecting Cr- and B-based
oxide scale. The enhanced oxidation resistance could be linked to
thermally activated precipitation of Si and the subsequent re-
crystallization of the Cr-Si-B
2 ± z
solid solution, thereby creating a
continuous Si-based surface layer. This layer allows for a dense,
amorphous SiO
2
-based scale (∼20 nm at 1200 °C) in the temperature
range between T = 650–1100 °C, beyond which an additional nano-
crystalline Cr
2
O
3
top layer (∼200 nm at 1200 °C) is formed due to
increased Cr-outward diffusion.
In summary, the results underpin the promising capabilities of
Cr-Si-B
2 ± z
coatings applied in high-temperature oxidative environ-
ments and provide detailed guidelines to connect the chemical
composition with resulting thin film properties.
CRediT authorship contribution statement
L. Zauner: Conceptualization, Investigation, Visualization,
Writing – original draft. A. Steiner: Investigation, Writing – review &
editing. T. Glechner: Investigation, Writing – review & editing. A.
Bahr: Investigation, Writing – review & editing. B. Ott: Investigation,
Writing – review & editing. R. Hahn: Investigation, Writing – review
& editing. T. Wojcik: Investigation, Writing – review & editing. O.
Hunold: Project administration, Writing – review & editing. J.
Ramm: Conceptualization, Project administration, Writing – review
& editing. S. Kolozsvári: Project administration, Writing – review &
editing. P. Polcik: Conceptualization, Project administration, Writing
– review & editing. P. Felfer: Investigation, Writing – review &
editing. H. Riedl: Supervision, Conceptualization, Project adminis-
tration, Writing – review & editing.
Data Availability
Data will be made available on request.
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
9

Declaration of Competing Interest
The authors declare that they have no known competing fi-
nancial interests or personal relationships that could have appeared
to influence the work reported in this paper.
Acknowledgments
The financial 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 Engineering of high-performance
Components"). We also thank for the financial 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.
We also thank Dr. M. Weiss and Prof. A. Limbeck from the Institute of
Chemical Technologies and Analytics, TU Wien, for their support
with chemical analysis of our samples. The authors acknowledge TU
Wien Bibliothek for financial support through its Open Access
Funding Programme.
Appendix A. Supporting information
Supplementary data associated with this article can be found in
the online version at doi:10.1016/j.jallcom.2023.169203.
References
[1] C..Subramanian, A.K..Suri, T.S.R..Ch.Murthy, Development of Boron-based mate-
rials for nuclear applications, BARC Newsl. (2010) 6–9. https://inis.iaea.org/
search/search.aspx?orig_q=RN:45027141 (Accessed 18 August 2022).
[2] D. Sciti, L. Silvestroni, L. Mercatelli, J.L. Sans, E. Sani, Suitability of ultra-refractory
diboride ceramics as absorbers for solar energy applications, Sol. Energy Mater.
Sol. Cells 109 (2013) 8–16, https://doi.org/10.1016/j.solmat.2012.10.004
[3] D.C. Larbalestier, L.D. Cooley, M.O. Rikel, A.A. Polyanskii, J. Jiang, S. Patnaik,
X.Y. Cai, D.M. Feldmann, A. Gurevich, A.A. Squitieri, M.T. Naus, C.B. Eom,
E.E. Hellstrom, R.J. Cava, K.A. Regan, N. Rogado, M.A. Hayward, T. He, J.S. Slusky,
P. Khalifah, K. Inumaru, M. Haas, Strongly linked current flow in polycrystalline
forms of the superconductor MgB2, Nature 410 (2001) 186–189, https://doi.org/
10.1038/35065559
[4] F. Monteverde, L. Scatteia, Resistance to thermal shock and to oxidation of metal
diborides-SiC ceramics for aerospace application, J. Am. Ceram. Soc. 90 (2007)
1130–1138, https://doi.org/10.1111/j.1551-2916.2007.01589.x
[5] M. Magnuson, L. Hultman, H. Högberg, Review of transition-metal diboride thin
films, Vacuum 196 (2022) 110567, https://doi.org/10.1016/j.vacuum.2021.110567
[6] A. Bellosi, F. Monteverde, Fabrication and properties of zirconium diboride-
based ceramics for UHT applications, in: Eur. Sp. Agency, (Special Publ. ESA SP,
2003: pp. 65–72. https://www.researchgate.net/publication/234354208
(Accessed 18 August 2022).
[7] W. Tan, C.A. Petorak, R.W. Trice, Rare-earth modified zirconium diboride high
emissivity coatings for hypersonic applications, J. Eur. Ceram. Soc. 34 (2014)
1–11, https://doi.org/10.1016/j.jeurceramsoc.2013.07.016
[8] C. Rhodes, J. Stuart, R. Lopez, X. Li, M. Waje, M. Mullings, J. Lau, S. Licht,
Evaluation of properties and performance of nanoscopic materials in vanadium
diboride/air batteries, J. Power Sources 239 (2013) 244–252, https://doi.org/10.
1016/j.jpowsour.2013.03.071
[9] W.G. Fahrenholtz, G.E. Hilmas, Ultra-high temperature ceramics: materials for
extreme environments, Scr. Mater. 129 (2017) 94–99, https://doi.org/10.1016/j.
scriptamat.2016.10.018
[10] C. Mitterer, Borides in thin film technology, J. Solid State Chem. 133 (1997)
279–291, https://doi.org/10.1006/jssc.1997.7456
[11] J.W. Zimmermann, G.E. Hilmas, W.G. Fahrenholtz, R.B. Dinwiddie, W.D. Porter,
H. Wang, Thermophysical properties of ZrB2 and ZrB2-SiC ceramics, J. Am.
Ceram. Soc. 91 (2008) 1405–1411, https://doi.org/10.1111/j.1551-2916.2008.
02268.x
[12] V. Moraes, L. Zauner, T. Wojcik, M. Arndt, P. Polcik, H. Riedl, P.H. Mayrhofer,
Thermally stable superhard diborides: An ab initio guided case study for V-W-
diboride thin films, Acta Mater. 186 (2020) 487–493, https://doi.org/10.1016/j.
actamat.2020.01.014
[13] C. Fuger, V. Moraes, R. Hahn, H. Bolvardi, P. Polcik, H. Riedl, P.H. Mayrhofer,
Influence of Tantalum on phase stability and mechanical properties of WB 2,
MRS Commun. 9 (2019) 375–380, https://doi.org/10.1557/mrc.2019.5
[14] C. Fuger, B. Schwartz, T. Wojcik, V. Moraes, M. Weiss, A. Limbeck, C.A. Macauley,
O. Hunold, P. Polcik, D. Primetzhofer, P. Felfer, P.H. Mayrhofer, H. Riedl, Influence
of Ta on the oxidation resistance of WB2−z coatings, J. Alloy. Compd. 864 (2021)
158121, https://doi.org/10.1016/j.jallcom.2020.158121
[15] B. Bakhit, J. Palisaitis, J. Thörnberg, J. Rosen, P.O.Å. Persson, L. Hultman, I. Petrov,
J.E. Greene, G. Greczynski, Improving the high-temperature oxidation resistance
of TiB2 thin films by alloying with Al, Acta Mater. 196 (2020) 677–689, https://
doi.org/10.1016/j.actamat.2020.07.025
[16] J. Thörnberg, S. Mráz, J. Palisaitis, F.F. Klimashin, P. Ondracka, B. Bakhit, P. Polcik,
S. Kolozsvari, L. Hultman, I. Petrov, P.O.Å. Persson, J.M. Schneider, J. Rosen,
Oxidation resistance and mechanical properties of sputter-deposited
Ti0.9Al0.1B2-y thin films, Surf. Coat. Technol. 442 (2022) 128187, https://doi.
org/10.1016/j.surfcoat.2022.128187
[17] A.H. Navidi Kashani, S. Mráz, D.M. Holzapfel, M. Hans, L. Löfler, P. Ondračka,
D. Primetzhofer, J.M. Schneider, Synthesis and oxidation behavior of
Ti0.35Al0.65By (y = 1.7–2.4) coatings, Surf. Coat. Technol. 442 (2022) 128190,
https://doi.org/10.1016/j.surfcoat.2022.128190
[18] T.A. Parthasarathy, R.A. Rapp, M. Opeka, R.J. Kerans, A model for the oxidation of
ZrB2, HfB2 and TiB2, Acta Mater. 55 (2007) 5999–6010, https://doi.org/10.1016/j.
actamat.2007.07.027
[19] W.G. Fahrenholtz, G.E. Hilmas, Oxidation of ultra-high temperature transition
metal diboride ceramics, Int. Mater. Rev. 57 (2012) 61–72, https://doi.org/10.
1179/1743280411Y.0000000012
[20] L. Silvestroni, K. Stricker, D. Sciti, H.J. Kleebe, Understanding the oxidation be-
havior of a ZrB2–MoSi2 composite at ultra-high temperatures, Acta Mater. 151
(2018) 216–228, https://doi.org/10.1016/j.actamat.2018.03.042
[21] I.G. Talmy, J.A. Zaykoski, M.M. Opeka, High-temperature chemistry and oxidation
of ZrB 2 ceramics containing SiC, Si 3N 4, Ta 5Si 3, and TaSi 2, J. Am. Ceram. Soc.
91 (2008) 2250–2257, https://doi.org/10.1111/j.1551-2916.2008.02420.x
[22] T.S.R.C. Murthy, J.K. Sonber, C. Subramanian, R.K. Fotedar, M.R. Gonal, A.K. Suri,
Effect of CrB2 addition on densification, properties and oxidation resistance of
TiB2, Int. J. Refract. Met. Hard Mater. 27 (2009) 976–984, https://doi.org/10.1016/
j.ijrmhm.2009.06.004
[23] B. Bakhit, S. Dorri, A. Kooijman, Z. Wu, J. Lu, J. Rosen, J.M.C. Mol, L. Hultman,
I. Petrov, J.E. Greene, G. Greczynski, Multifunctional ZrB2-rich Zr1-xCrxBy thin
films with enhanced mechanical, oxidation, and corrosion properties, Vacuum
185 (2021), https://doi.org/10.1016/j.vacuum.2020.109990
[24] S. Dorri, J. Palisaitis, G. Greczynski, I. Petrov, J. Birch, L. Hultman, B. Bakhit,
Oxidation kinetics of overstoichiometric TiB2 thin films grown by DC magnetron
sputtering, Corros. Sci. 206 (2022) 110493, https://doi.org/10.1016/J.CORSCI.
2022.110493
[25] J. Thörnberg, B. Bakhit, J. Palisaitis, N. Hellgren, L. Hultman, G. Greczynski,
P.O.Å. Persson, I. Petrov, J. Rosen, Improved oxidation properties from a reduced
B content in sputter-deposited TiBx thin films, Surf. Coat. Technol. 420 (2021)
127353, https://doi.org/10.1016/j.surfcoat.2021.127353
[26] C. Fuger, R. Hahn, A. Hirle, P. Kutrowatz, M. Weiss, A. Limbeck, O. Hunold,
P. Polcik, H. Riedl, Revisiting the origins of super-hardness in TiB2+z thin films –
impact of growth conditions and anisotropy, Surf. Coat. Technol. 446 (2022)
128806, https://doi.org/10.1016/j.surfcoat.2022.128806
[27] T. Glechner, O.E. Hudak, T. Wojcik, L. Haager, F. Bohrn, H. Hutter, O. Hunold,
J. Ramm, S. Kolozsvári, E. Pitthan, D. Primetzhofer, H. Riedl, Influence of the non-
metal species on the oxidation kinetics of Hf, HfN, HfC, and HfB2 coatings, Mater.
Des. 211 (2021) 110136, https://doi.org/10.1016/j.matdes.2021.110136
[28] T. Glechner, H.G. Oemer, T. Wojcik, M. Weiss, A. Limbeck, J. Ramm, P. Polcik,
H. Riedl, Influence of Si on the oxidation behavior of TM-Si-B2 ± z coatings (TM =
Ti, Cr, Hf, Ta, W), Surf. Coat. Technol. 434 (2022) 128178, https://doi.org/10.1016/
j.surfcoat.2022.128178
[29] T. Glechner, A. Bahr, R. Hahn, T. Wojcik, M. Heller, A. Kirnbauer, J. Ramm,
S. Kolozsvari, P. Felfer, H. Riedl, High temperature oxidation resistance of phy-
sical vapor deposited Hf-Si-B2 ± z thin films, Corros. Sci. 205 (2022) 110413,
https://doi.org/10.1016/j.corsci.2022.110413
[30] B. Grančič, M. Mikula, T. Roch, P. Zeman, L. Satrapinskyy, M. Gregor, T. Plecenik,
E. Dobročka, Z. Hájovská, M. Mičušík, A. Šatka, M. Zahoran, A. Plecenik, P. Kúš,
Effect of Si addition on mechanical properties and high temperature oxidation
resistance of Ti-B-Si hard coatings, Surf. Coat. Technol. 240 (2014) 48–54,
https://doi.org/10.1016/j.surfcoat.2013.12.011
[31] W.C. Oliver, G.M. Pharr, Measurement of hardness and elastic modulus by in-
strumented indentation: Advances in understanding and refinements to meth-
odology, J. Mater. Res. 19 (2004) 3–20, https://doi.org/10.1557/jmr.2004.19.1.3
[32] A.C. Fischer-Cripps, Critical review of analysis and interpretation of na-
noindentation test data, Surf. Coat. Technol. 200 (2006) 4153–4165, https://doi.
org/10.1016/j.surfcoat.2005.03.018
[33] 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 film stress:
Developments from polycrystalline steel strips to single crystal silicon wafers,
Thin Solid Films 517 (2009) 1858–1867, https://doi.org/10.1016/j.tsf.2008.07.014
[34] G.G. Stoney, The Tension of Metallic Films Deposited by Electrolysis, Proc. R. Soc.
A Math. Phys. Eng. Sci. 82 (1909) 172–175. doi:10.1098/rspa.1909.0021.
[35] G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy
calculations using a plane-wave basis set, Phys. Rev. B - Condens. Matter Mater.
Phys. 54 (1996) 11169–11186, https://doi.org/10.1103/PhysRevB.54.11169
[36] G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector aug-
mented-wave method, Phys. Rev. B. 59 (1999) 1758–1775, https://doi.org/10.
1103/PhysRevB.59.1758
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
10

[37] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made
simple, Phys. Rev. Lett. 77 (1996) 3865–3868, https://doi.org/10.1103/
PhysRevLett.77.3865
[38] A. Van de Walle, M. Asta, G. Ceder, The alloy theoretic automated toolkit: A user
guide, Calphad Comput. Coupling Phase Diagr. Thermochem 26 (2002) 539–553,
https://doi.org/10.1016/S0364-5916(02)80006-2
[39] P. Felfer, GitHub - Atom-Probe-Toolbox: Matlab toolbox for APT analysis, (2022).
https://github.com/peterfelfer/Atom-Probe-Toolbox (Accessed 19 August 2022).
[40] L. Chen, J. Paulitsch, Y. Du, P.H. Mayrhofer, Thermal stability and oxidation re-
sistance of Ti–Al–N coatings, Surf. Coat. Technol. 206 (2012) 2954–2960, https://
doi.org/10.1016/j.surfcoat.2011.12.028
[41] H. Euchner, P.H. Mayrhofer, Vacancy-dependent stability of cubic and wurtzite
Ti1−xAlxN, Surf. Coat. Technol. 275 (2015) 214–218, https://doi.org/10.1016/j.
surfcoat.2015.05.017
[42] V. Moraes, H. Riedl, C. Fuger, P. Polcik, H. Bolvardi, D. Holec, P.H. Mayrhofer, Ab
initio inspired design of ternary boride thin films, Sci. Rep. 8 (2018) 9288,
https://doi.org/10.1038/s41598-018-27426-w
[43] P. Villars, H. Okamoto, B-Cr-Si Isothermal Section of Ternary Phase Diagram,
(2012). https://materials.springer.com/isp/phase-diagram/docs/c_0977925.
[44] ICDD, Powder Diffraction File - hexagonal CrB2 - 04–004-1734, Powder Diffr. File
- Hexag. CrB2 - 04–004-1734. (2011).
[45] ICDD, Powder Diffraction File - cubic Si - 00–027-1402, Powder Diffr. File - Cubic
Si - 00–027-1402. (2017).
[46] M.M. Dorri, J. Thörnberg, N. Hellgren, J. Palisaitis, A. Petruhins, F.F. Klimashin,
L. Hultman, I. Petrov, P.O.Å. Persson, J. Rosen, Synthesis and characterization of
CrB2 thin films grown by DC magnetron sputtering, Scr. Mater. 200 (2021)
113915, https://doi.org/10.1016/j.scriptamat.2021.113915
[47] M. Stüber, H. Riedl, T. Wojcik, S. Ulrich, H. Leiste, P.H. Mayrhofer, Microstructure
of Al-containing magnetron sputtered TiB2 thin films, Thin Solid Films 688
(2019) 137361, https://doi.org/10.1016/j.tsf.2019.06.011
[48] P.H. Mayrhofer, C. Mitterer, J.G. Wen, J.E. Greene, I. Petrov, Self-organized na-
nocolumnar structure in superhard TiB2 thin films, Appl. Phys. Lett. 86 (20 05)
1–3, https://doi.org/10.1063/1.1887824
[49] R. Hahn, V. Moraes, A. Limbeck, P. Polcik, P.H. Mayrhofer, H. Euchner, Electron-
configuration stabilized (W,Al)B2 solid solutions, Acta Mater. 174 (2019)
398–405, https://doi.org/10.1016/j.actamat.2019.05.056
[50] L. Zauner, R. Hahn, E. Aschauer, T. Wojcik, A. Davydok, O. Hunold, P. Polcik,
H. Riedl, Assessing the fracture and fatigue resistance of nanostructured thin
films, Acta Mater. 239 (2022) 118260, https://doi.org/10.1016/j.actamat.2022.
118260
[51] M. Audronis, A. Leyland, P.J. Kelly, A. Matthews, The effect of pulsed magnetron
sputtering on the structure and mechanical properties of CrB2 coatings, Surf.
Coat. Technol. 201 (2006) 3970–3976, https://doi.org/10.1016/j.surfcoat.2006.08.
006
[52] K.L. Dahm, L.R. Jordan, J. Haase, P.A. Dearnley, Magnetron sputter deposition of
chromium diboride coatings, Surf. Coat. Technol. 108–109 (1998) 413–418,
https://doi.org/10.1016/S0257-8972(98)00568-4
[53] S. Zhang, Z. Wang, P. Guo, P. Ke, M. Odén, A. Wang, Temperature induced su-
perhard CrB2 coatings with preferred (001) orientation deposited by DC mag-
netron sputtering technique, Surf. Coat. Technol. 322 (2017) 134–140, https://
doi.org/10.1016/j.surfcoat.2017.05.030
[54] C. Fuger, R. Hahn, L. Zauner, T. Wojcik, M. Weiss, A. Limbeck, O. Hunold, P. Polcik,
H. Riedl, Anisotropic super-hardness of hexagonal WB2 ± z thin films, Mater. Res.
Lett. 10 (2022) 70–77, https://doi.org/10.1080/21663831.2021.2021308
[55] C.L. Jiang, Z.L. Pei, Y.M. Liu, H. Lei, J. Gong, C. Sun, Determination of the thermal
properties of AlB 2 -type WB 2, Appl. Surf. Sci. 288 (2014) 324–330, https://doi.
org/10.1016/j.apsusc.2013.10.027
[56] CrysTec GmbH, Data sheet: Sapphire for Research and Development, (2022).
http://www.crystec.de/daten/al2o3.pdf (Accessed 19 August 2022).
[57] P.V. Kiryukhantsev-Korneev, J.F. Pierson, K.A. Kuptsov, D.V. Shtansky, Hard Cr-Al-
Si-B-(N) coatings deposited by reactive and non-reactive magnetron sputtering
of CrAlSiB target, Appl. Surf. Sci. 314 (2014) 104–111, https://doi.org/10.1016/j.
apsusc.2014.06.160
[58] ICDD, Powder Diffraction File - hexagonal CrSi2 - 04–001-7333, Powder Diffr.
File - Hexag. CrSi2 - 04–001-7333. (2011).
[59] ICDD, Powder Diffraction File - rhombohedral Cr2O3 - 00–0 06-0504, Powder
Diffr. File - Rhombohedral Cr2O3 - 00–006-0504. (1970).
[60] ICDD, Powder Diffraction File - orthorhombic CrB - 04–003-6125, Powder Diffr.
File - Orthorhombic CrB - 04–003-6125. (2011).
[61] ICDD, Powder Diffraction File - rhombohedral Al2O3 - 00–046-1212, Powder
Diffr. File - Rhombohedral Al2O3 - 00–046-1212. (1970).
[62] P.K. Liao, K.E. Spear, The B-Cr (Boron-Chromium) system, Bull. Alloy Phase Diagr.
7 (1986) 232–237, https://doi.org/10.1007/BF02868996
[63] N. Birks, G.H. Meier, F.S. Pettit, Introduction to the high temperature oxidation of
metals, Second edition., Cambridge University Press, 2006, https://doi.org/10.
1017/CBO9781139163903
L. Zauner, A. Steiner, T. Glechner et al. Journal of Alloys and Compounds 944 (2023) 169203
11