Hardness is an essential property for a wide range of applications. However, hardness alone, typically accompanied by brittleness, is not sufficient to prevent failure in ceramic films exposed to high stresses. Using VN as a model system, we demonstrate with experiment and density functional theory (DFT) that refractory VMoN alloys exhibit not only enhanced hardness, but dramatically increased ductility. V0.5Mo0.5N hardness is 25% higher than that of VN. In addition, while nanoindented VN, as well as TiN reference samples, suffer from severe cracking typical of brittle ceramics, V0.5Mo0.5N films do not crack. Instead, they exhibit material pile-up around nanoindents, characteristic of plastic flow in ductile materials. Moreover, the wear resistance of V0.5Mo0.5N is considerably higher than that of VN. DFT results show that tuning the occupancy of d–t2g metallic bonding states in VMoN facilitates dislocation glide, and hence enhances toughness, via the formation of stronger metal/metal bonds along the slip direction and weaker metal/N bonds across the slip plane.

Hardness is an essential property for a wide range of thin film applications, including wear- and abrasion-resistant coatings. Over the past two decades, a major quest in materials science has been the development of artificial materials with increasing hardness. Successful examples include phase-stability tuning,1 superhardening of refractory nitrides via the development of artificial superlattices,2,3 nano-scale composites,4,5 vacancy-induced hardening,6–9 and polytype-mixtures in transition-metal carbides and nitrides.10 However, each of these hardening mechanisms extracts a steep price in terms of increased brittleness.

In order to avoid brittle failure due to cracking, thin films must be both hard and ductile. This combination of properties is referred to as toughness, a measure of the resistance of a material to crack formation. Several thin-film toughening approaches have been developed, including the incorporation of ductile phases,11–13 multilayer structures,14,15 compressive stresses,16,17 phase transformations,18,19 and carbon-nanotubes.20–22 However, these approaches, which are based on methods applied to bulk materials, raise significant problems in coatings.23 The slow progress in obtaining hard, yet tough, ceramic coatings is due to the relatively primitive state of the present understanding of thin film toughness, with only a small number of studies dedicated to the electronic origin of brittleness vs. ductility.24,25

Recently, Sangiovanni et al.,26,27 using ab initio density functional theory (DFT), predicted enhanced toughness in pseudobinary B1-NaCl structure TM nitride alloys of TiN or VN and MoN or WN with Cu-Pt ordering on the cation sublattice. Toughness enhancement in such alloys stems from the high valence electron concentration (VEC) which leads, during shear, to an electronic structure consisting of alternating layers of high and low charge density orthogonal to the applied stress. This, in turn, allows a selective response to tetragonal and trigonal deformations: if compressive/tensile stresses are applied, the structure responds by resisting deformation. Upon application of shear stress, however, the layered electronic arrangement allows the sample to respond in a ductile/tough manner as dislocation glide along the {110}/⟨|$1\bar 10$|11¯0⟩ slip system becomes energetically favored.

Here, we experimentally demonstrate that the combination of increased hardness with enhanced ductility is inherent to B1 NaCl-structure V0.5Mo0.5N/MgO(001), compared to the binary parent compound VN/MgO(001), and TiN/MgO(001) reference samples, independent of atomic ordering. In addition, the wear resistance of V0.5Mo0.5N is considerably higher than that of VN. New DFT calculations show that in disordered V0.5Mo0.5N alloys, the filling of d–t2g states, due to the high VEC, also leads to enhanced dislocation mobility. In this case, however, toughness arises not from a layered charge density structure, as for Cu-Pt ordered alloys, but from the formation of stronger metal/metal bonds along the slip direction and weaker metal/N and metal/metal bonds across the slip plane as shown by crystal orbital overlap population (COOP)28 analyses.

Epitaxial V0.5Mo0.5N/MgO(001) alloys are grown by dual-target reactive magnetron sputtering in a stainless-steel UHV system with a base pressure ∼2 × 10−9 Torr. Depositions are carried out at Ts = 700 °C in mixed Ar/N2 atmospheres at a total pressure of 5 mTorr, with a N2 partial pressure of 3.2 mTorr. The targets are V (99.95% purity) and Mo (99.95% purity) discs, with diameters of 76 mm, tilted 30° with respect to the substrate normal and separated from the substrate by 15 cm. Alloy film growth rates are 70 Å/min. VN/MgO(001) and TiN/MgO(001) reference samples are grown at Ts = 700 °C with a pressure of 10 mTorr in pure N2 (VN), and 800 °C and 5 mTorr in mixed Ar/N2 with PN2 = 1 mTorr (TiN). A bias of −30 V is applied to the substrate during deposition of all layers.

Film composition is determined by a combination of energy-dispersive electron spectroscopy (EDS), Rutherford backscattering spectrometry (RBS), and elastic recoil detection analysis (ERDA). Two separate sets of V0.5Mo0.5N samples, with thicknesses of ∼1000 and 3000 Å, are grown. The thinner samples are used to determine film composition by RBS, while EDS measurements are carried out on both thick and thin samples. The thicker sample set is used for the remaining analyses. RBS measurements are performed using a 2 MeV 4He+ beam. Results are analyzed based upon SIMNRA 6.06 software.29 ERDA compositional analyses are carried out with a 36 MeV 127I8+ probe beam. The data are quantified using CONTES software.30 

Structure and phase composition of as-deposited alloys are analyzed by x-ray diffraction (XRD), cross-sectional transmission electron microscopy (XTEM), and selected-area electron diffraction (SAED). A Bragg-Brentano diffractometer with Cu Kα radiation is used to acquire ω-2θ XRD scans. Relaxed alloy lattice parameters ao are obtained from reciprocal space maps (RSM) around the asymmetric 113 reflection using Cu Kα1 radiation. XTEM and SAED analyses are carried out in an FEI Tecnai G2 TF20UT microscope operated at 200 kV.

Film hardness and elastic moduli are determined, based on the Oliver and Pharr method,31 from nanoindentation measurements in a Hysitron TI950 Triboindenter using a Berkovich diamond probe (radius ∼1500 Å), and calibrated using a fused silica reference sample. The maximum indentation depth is ≤10% of the film thickness, ∼3000 Å in all cases. Film toughness is assessed based upon analyses of sample cracks, or material pile-up adjacent to indent edges, via scanning electron and scanning probe microscopies (SEM and SPM), following nanoindentation experiments at constant depth, with a sharp cube-corner tip. Wear tests are performed over 2 × 2 μm2 areas using a cube-corner tip with a constant load of 100 μN at a scan rate of 8 μm/s for 5 cycles. Each scan consists of 256 lines with a lateral displacement of 80 Å per line.

DFT calculations are accomplished using the Vienna ab initio simulation package (VASP),32 implemented with the generalized gradient approximation (GGA),33 and the projector augmented wave (PAW) method.34 The rectangular simulation supercell consists of four (001) planes with 24 atoms (6 × 4) each. The non-ordered alloy configuration is generated by employing a modified special quasirandom structure (SQS) method35,36 which provides negligible short-range order.37 Randomization is carried out on individual (001) planes to achieve low metal/metal correlation on circular neighboring shells around each metal atom. In addition, since the deposited films do not exhibit Cu-Pt ordering, simulation supercell (001) layers are shuffled until nearly equal concentrations of Mo and V atoms are obtained on each (111) plane. Equilibrium structures are determined by minimizing the total energy to within an accuracy of 10‑5 eV/atom upon relaxing the atomic positions, cell shape, and volume. Elastic constants and moduli are calculated, as described in Refs. 26 and 27, and 38, using a plane-wave energy cutoff of 500 eV and 6 × 6 × 6 k-point grids. For electron density and COOP calculations, we use 8 × 8 × 8 k-point grids. Charge transfer maps are obtained by subtracting gas-phase atom electron densities from the alloy charge density. COOP corresponds to the electron density of states (DOS) resolved into bonding and antibonding states and is used to quantify the strength of the covalent component of chemical bonds. We calculate COOP values following the procedures used in Refs. 27 and 39, and 40. The integration of COOP functions over all occupied states (ICOOP) is a measure of bond strengths. The calculated alloy mixing enthalpy is negative (−0.121 eV/atom); i.e., the alloy is stable with respect to mixing of the parent cubic binary compounds.

The combination of XTEM and XRD analyses reveals that all films – VN/MgO(001), V0.5Mo0.5N/MgO(001), and TiN/MgO(001) – are epitaxial with the B1-NaCl structure and exhibit a cube-on-cube orientation relationship to the substrate: e.g., (001)VMoN∥(001)MgO and [100]VMoN∥[100]MgO. The only Kα1 V0.5Mo0.5N/MgO(001) ω-2θ XRD peaks observed over the 2θ range 10°–95° are the 002 and 004 MgO reflections at 42.95° and 93.96° and the 002 V0.5Mo0.5N film peak at 43.32°. XRD reciprocal space maps (Fig. 1) reveal that V0.5Mo0.5N(001) alloy films are fully relaxed with ao = 4.175 ± 0.005 Å. The N/(V+Mo) ratio obtained from RBS measurements is 0.94 ± 0.05. VN(001) and TiN(001) reference layers have out-of-plane lattice parameters of 4.130 ± 0.005 and 4.245 ± 0.005 Å, in agreement with previous reports.6 N/V = 0.89 ± 0.05 and N/Ti = 0.96 ± 0.05.

FIG. 1.

Typical XRD reciprocal space map around the 113 reflection from an epitaxial V0.5Mo0.5N/MgO(001) layer grown at 700 °C.

FIG. 1.

Typical XRD reciprocal space map around the 113 reflection from an epitaxial V0.5Mo0.5N/MgO(001) layer grown at 700 °C.

Close modal

Fig. 2 shows typical high-resolution XTEM (HR-XTEM) images, with corresponding SAED patterns as insets, of epitaxial V0.5Mo0.5N/MgO(001) layers viewed along [100] and [110] zone axes. The [110] HR-XTEM image exhibits uniform lattice-fringe contrast characteristic of a randomly distributed metal sublattice. The SAED patterns are composed of symmetric single-crystal reflections. There is no evidence in the 011 pattern of ½{111} superstructure reflections characteristic of Cu-Pt ordering on the metal sublattice which, in this case, would correspond to alternating V- and Mo-rich (111) planes, as observed for Ti0.5W0.5N/MgO(001) alloys.41 

FIG. 2.

Typical HR-XTEM images, with SAED patterns as inserts, of V0.5Mo0.5N/MgO(001) viewed along (a) the [100] and (b) [110] zone axes.

FIG. 2.

Typical HR-XTEM images, with SAED patterns as inserts, of V0.5Mo0.5N/MgO(001) viewed along (a) the [100] and (b) [110] zone axes.

Close modal

The elastic moduli E obtained from nanoindentation results using calculated Poisson ratios νTiN = 0.22,27 νVN = 0.25,27 and νVMoN = 0.319 are EVMoN = 376 ± 30, ETiN = 421 ± 40, and EVN = 386 ± 20 GPa. The nanoindentation modulus EVMoN is in reasonable agreement with the modulus calculated for the disordered alloy, 328 GPa. The hardness of VMoN, HVMoN = 20 ± 1 GPa, is higher than that of VN, HVN = 16 ± 1 GPa, and comparable with TiN, HTiN = 23 ± 2 GPa. Thus, alloying VN with MoN enhances hardness by 25%, while the elastic modulus remains essentially constant. The combined results are one indication of enhanced toughness.

The ductility of epitaxial V0.5Mo0.5N(001) layers, compared to epitaxial VN(001) and TiN(001) reference layers, is also evaluated via nanoindentation experiments, but in this case with a sharp cube-corner diamond tip. Ten or more indents to a constant depth of 4000 Å are produced in each 3000-Å-thick film, thus extending into the MgO substrates by 1000 Å. Typical SEM and SPM images of the results are presented in Fig. 3. VN and TiN reference samples (Figs. 3(c) and 3(d)) exhibit severe cracking, characteristic of brittle ceramic films, along ⟨110⟩ directions around the indents. In distinct contrast, cracks are never observed in V0.5Mo0.5N films (see Fig. 3(a)). Instead, indented alloy layers exhibit material pile-up along indent edges (Figs. 3(a) and 3(b)), characteristic of plastic flow in ductile materials. From analyses of SPM results, the volume of the pile-up is approximately 0.018 μm3. These results, combined with the hardness data, are a second, and more direct, demonstration of the enhanced toughness of V0.5Mo0.5N alloys. In addition, the wear rate, defined as the difference in absolute values between the mean height outside and inside the scanned area, decreases by a factor of five, from 5 nm/cycle for VN to 1 nm/cycle for V0.5Mo0.5N.

FIG. 3.

Typical (a) SEM and (b) SPM images of nanoindentations in V0.5Mo0.5N/MgO(001). (c) and (d) are SEM images of nanoindentations on VN/MgO(001) and TiN/MgO(001) reference samples, respectively. All films have the same thickness, 3000 Å, and indentation depths are 4000 Å.

FIG. 3.

Typical (a) SEM and (b) SPM images of nanoindentations in V0.5Mo0.5N/MgO(001). (c) and (d) are SEM images of nanoindentations on VN/MgO(001) and TiN/MgO(001) reference samples, respectively. All films have the same thickness, 3000 Å, and indentation depths are 4000 Å.

Close modal

In parallel with our experiments, we perform calculations on non-ordered B1-NaCl-structure VMoN alloys. The calculated Cauchy pressure (C12-C44) and shear/bulk modulus ratio (G/B), often used as ductility criteria for cubic systems,42,43 indicate that irrespective of the degree of ordering, V0.5Mo0.5N [(C12-C44) = 95 GPa; G/B = 0.411] alloys are significantly more ductile than both TiN [(C12-C44) = −44 GPa; G/B = 0.690] and VN [(C12-C44) = 1 GPa; G/B = 0.597]. However, since these measures only provide qualitative indications of toughness, we also probe the V0.5Mo0.5N electronic structure.

To provide atomistic insights into the electronic origin of the enhanced toughness of V0.5Mo0.5N, we show charge transfer plots for relaxed and sheared disordered alloys in Figs. 4(a) and 4(b). Stress is applied along ⟨110⟩ directions to obtain 10% shear strain, in order to simulate massive deformation under penetration by the cube corner tip. The plots are oriented to depict the most common slip system for the NaCl-structure, {110}/⟨|$1\bar 10$|11¯0⟩, with the horizontal axis along the direction of the Burgers vector and the glide plane viewed edge-on. Corresponding charge transfer maps for VN and TiN (not shown) consist of spherical charge distributions around both cation and anion sites, due to pronounced ionic bonding. It is evident that alloying with MoN significantly alters bonding configurations.

FIG. 4.

Charge transfer maps for (a) relaxed and (b) 10% sheared disordered V0.5Mo0.5N. The color scale is expressed in electrons/Å3.

FIG. 4.

Charge transfer maps for (a) relaxed and (b) 10% sheared disordered V0.5Mo0.5N. The color scale is expressed in electrons/Å3.

Close modal

Mo forms weak pd-eg bonds (Fig. 4(a)) with both the under- and overlying N planes as evidenced by gold-colored lobes (see color code) between Mo and N atoms; one set of lobes is indicated by arrows. The filling of the dt2g bonding states, due to the high VEC, leads to delocalization of charge between metal atoms along ⟨110⟩ directions. Upon shearing (Fig. 4(b)), electron charge transfer occurs from metal/N and metal/metal bonds orthogonal to the glide plane to metal/metal bonds oriented along the slip direction, thus promoting dislocation mobility.

Upon 10% shearing V0.5Mo0.5N, ICOOP values orthogonal to the slip plane decrease by 9% for metal/N bonds [V/N (−13%) and Mo/N (−7%)] and 53% for metal/metal bonds [V/V (−55%), V/Mo (−49%), and Mo/Mo (−55%)]. However, along the glide plane, ICOOP for metal/metal bonds increases by 72% [V/V (+6%), V/Mo (+59%), and Mo/Mo (+86%)]. Changes in ICOOP due to shearing of VN are analogous, but smaller; ICOOP decreases 7% for V/N bonds and 59% for V/V bonds across the slip plane, while increasing 20% for V/V bonds within the glide plane. Overall, upon shearing, V0.5Mo0.5N forms considerably stronger metal/metal bonds within the glide plane, metal/metal ICOOP = 229.0, compared to VN for which V/V ICOOP = 65.8, with much weaker metal/N bonds across the slip plane for which V0.5Mo0.5N metal/N ICOOP = 900.8 and VN V/N ICOOP = 1176.4.44 Moreover, the pronounced ionic character of V/N bonds in VN (see charge transfer maps in Fig. 10 of Ref. 27) further hinders dislocation glide.

Overall, the addition of MoN to VN provides a VEC which optimizes the occupancy of metallic d–t2g bonding states in V0.5Mo0.5N, without filling antibonding states, leading to the formation of stronger metal/metal bonds within {110} planes and weaker metal/N bonds parallel to the applied strain. This, in turn, allows easier slip of {110} planes along ⟨|$1\bar 10$|11¯0⟩ directions and explains why V0.5Mo0.5N is inherently more ductile than VN and TiN reference samples. The alloy reacts in a hard manner upon compression and in a ductile manner upon shearing.

In summary, we show that alloying the Group-V TM nitride VN with Group-VI MoN not only increases film hardness, but also enhances ductility. That is, the alloy has higher toughness than the parent VN compound, consistent with our DFT results showing that the optimized occupancy of d–t2g metallic states in V0.5Mo0.5N facilitates dislocation glide via the formation of stronger metal/metal bonds along the slip direction and weaker metal/N bonds across the slip plane. The combination of XRD, HR-XTEM, and SAED reveals that the metal species in our tough V0.5Mo0.5N alloys are disordered on the cation sublattice, which is also consistent with DFT simulations indicating that an ordered structure is not necessary to access the combination of high hardness and ductility.

The present results provide new insights into the electronic origin of toughness and demonstrate the possibility of tuning the hardness/ductility ratio of pseudobinary transition-metal nitride alloys via manipulation of electronic d–t2g state occupancy.

The authors acknowledge the financial support of the Knut and Alice Wallenberg Foundation, the Swedish Research Council, and the Swedish Government Strategic Research Area Grant in Materials Science (AFM-SFO Mat-LiU). We are also grateful to E. Broitman for fruitful discussions. Calculations were performed, using resources provided by Swedish National Infrastructure for Computing, on the Neolith and Triolith Clusters located at the National Supercomputer Center in Linköping.

1.
H. W.
Hugosson
,
U.
Jansson
,
B.
Johansson
, and
O.
Eriksson
,
Science
293
,
2434
(
2001
).
2.
U.
Helmersson
,
S.
Todorova
,
S.
Barnett
,
J. E.
Sundgren
,
L. C.
Markert
, and
J.
Greene
,
J. Appl. Phys.
62
,
481
(
1987
).
3.
P.
Mirkarimi
,
L.
Hultman
, and
S.
Barnett
,
Appl. Phys. Lett.
57
,
2654
(
1990
).
4.
S.
Veprek
,
M. G. J.
Veprek-Heijman
,
P.
Karvankova
, and
J.
Prochazka
,
Thin Solid Films
476
,
1
(
2005
).
5.
L.
Hultman
,
J.
Bareno
,
A.
Flink
,
H.
Soderberg
,
K.
Larsson
,
V.
Petrova
,
M.
Oden
,
J. E.
Greene
, and
I.
Petrov
,
Phys. Rev. B
75
,
155437
(
2007
).
6.
C. S.
Shin
,
D.
Gall
,
N.
Hellgren
,
J.
Patscheider
,
I.
Petrov
, and
J. E.
Greene
,
J. Appl. Phys.
93
,
6025
(
2003
).
7.
S. H.
Jhi
,
S. G.
Louie
,
M. L.
Cohen
, and
J.
Ihm
,
Phys. Rev. Lett.
86
,
3348
(
2001
).
8.
C. S.
Shin
,
S.
Rudenja
,
D.
Gall
,
N.
Hellgren
,
T. Y.
Lee
,
I.
Petrov
, and
J. E.
Greene
,
J. Appl. Phys.
95
,
356
(
2004
).
9.
T.
Lee
,
K.
Ohmori
,
C. S.
Shin
,
D. G.
Cahill
,
I.
Petrov
, and
J. E.
Greene
,
Phys. Rev. B
71
,
144106
(
2005
).
10.
T.
Joelsson
,
L.
Hultman
,
H. W.
Hugosson
, and
J. M.
Molina-Aldareguia
,
Appl. Phys. Lett.
86
,
131922
(
2005
).
11.
L.
Hultman
, “
Synthesis, structure, and properties of super-hard supperlattices coatings
,”
Nanostructured Coatings
(
Springer
,
New York
,
2006
), Chap. 13.
12.
S.
Zhang
,
X. L.
Bui
, and
Y. Q.
Fu
,
Thin Solid Films
467
,
261
(
2004
).
13.
S.
Zhang
,
X. L.
Bui
,
X. T.
Zeng
, and
X. M.
Li
,
Thin Solid Films
482
,
138
(
2005
).
14.
S.
Hogmark
,
S.
Jacobson
, and
M.
Larsson
,
Wear
246
,
20
(
2000
).
15.
Y. H.
Cheng
,
T.
Browne
,
B.
Heckerman
,
C.
Bowman
,
V.
Gorokhovsky
, and
E. I.
Meletis
,
Surf. Coat. Technol.
205
,
146
(
2010
).
16.
O.
Abe
and
Y.
Ohwa
,
Solid State Ionics
172
,
553
(
2004
).
17.
B. D.
Beake
,
V. M.
Vishnyakov
, and
A. J.
Harris
,
Tribol. Int.
44
,
468
(
2011
).
18.
Z.
Ji
,
J. A.
Haynes
,
M. K.
Ferber
, and
J. M.
Rigsbee
,
Surf. Coat. Technol.
135
,
109
(
2001
).
19.
B.
Basu
,
T.
Venkateswaran
, and
D. Y.
Kim
,
J. Am. Ceram. Soc.
89
,
2405
(
2006
).
20.
G. D.
Zhan
,
J. D.
Kuntz
,
J. L.
Wan
, and
A. K.
Mukherjee
,
Nature Mater.
2
,
38
(
2003
).
21.
Z.
Xia
,
L.
Riester
,
W. A.
Curtin
,
H.
Li
,
B. W.
Sheldon
,
J.
Liang
,
B.
Chang
, and
J. M.
Xu
,
Acta Mater.
52
,
931
(
2004
).
22.
H. J.
Choi
and
D. H.
Bae
,
Mater. Sci. Eng., A
528
,
2412
(
2011
).
23.
S.
Zhang
,
H. L.
Wang
,
S. E.
Ong
,
D.
Sun
, and
X. L.
Bui
,
Plasma Processes Polym.
4
,
219
(
2007
).
24.
S.
Kamran
,
K. Y.
Chen
, and
L.
Chen
,
Phys. Rev. B
79
,
024106
(
2009
).
25.
T.
Dasgupta
,
U. V.
Waghmare
, and
A. M.
Umarji
,
Phys. Rev. B
76
,
174110
(
2007
).
26.
D. G.
Sangiovanni
,
V.
Chirita
, and
L.
Hultman
,
Phys. Rev. B
81
,
104107
(
2010
).
27.
D. G.
Sangiovanni
,
L.
Hultman
, and
V.
Chirita
,
Acta Mater.
59
,
2121
(
2011
).
28.
R.
Hoffmann
,
Solids and Surfaces: A Chemist's View of Bonding in Extended Structures
(
Wiley-VCH
,
1988
).
29.
M.
Mayer
,
AIP Conf. Proc.
475
,
541
(
1999
).
30.
M. S.
Janson
,
Internal Report
,
Uppsala University
,
2004
.
31.
W. C.
Oliver
and
G. M.
Pharr
,
J. Mater. Res.
7
,
1564
(
1992
).
32.
G.
Kresse
and
J.
Furthmuller
,
Phys. Rev. B
54
,
11169
(
1996
).
33.
J. P.
Perdew
and
Y.
Wang
,
Phys. Rev. B
45
,
13244
(
1992
).
34.
P. E.
Blochl
,
Phys. Rev. B
50
,
17953
(
1994
).
35.
A.
Zunger
,
S. H.
Wei
,
L. G.
Ferreira
, and
J. E.
Bernard
,
Phys. Rev. Lett.
65
,
353
(
1990
).
36.
B.
Alling
,
C.
Hoglund
,
R.
Hall-Wilton
, and
L.
Hultman
,
Appl. Phys. Lett.
98
,
241911
(
2011
).
37.
A. V.
Ruban
and
I. A.
Abrikosov
,
Rep. Prog. Phys.
71
,
046501
(
2008
).
38.
F.
Tasnadi
,
M.
Oden
, and
I. A.
Abrikosov
,
Phys. Rev. B
85
,
144112
(
2012
).
39.
D. G.
Sangiovanni
,
V.
Chirita
, and
L.
Hultman
,
Thin Solid Films
520
,
4080
(
2012
).
40.
A.
Grechnev
,
R.
Ahuja
, and
O.
Eriksson
,
J. Phys.: Condens. Matter
15
,
7751
(
2003
).
41.
F.
Tian
,
J.
D’Arcy-Gall
,
T. Y.
Lee
,
M.
Sardela
,
D.
Gall
,
I.
Petrov
, and
J. E.
Greene
,
J. Vac. Sci. Technol. A
21
,
140
(
2003
).
42.
D. G.
Pettifor
,
Mat. Sci. Technol.
8
,
345
(
1992
).
43.
44.
See supplementary material at http://dx.doi.org/10.1063/1.4822440 for ICOOP values of bonds in V0.5Mo0.5N and VN.
All article content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 Unported License.

Supplementary Material