Jul 24, 2008 - CrAlN films were prepared by a pulsed DC magnetron sputtering in .... 4. A pronounced fcc-CrN phase was obtained in the films deposited at ...
Materials Transactions, Vol. 49, No. 9 (2008) pp. 2082 to 2090 #2008 The Japan Institute of Metals
Effect of Deposition Conditions on the Structure and Properties of CrAlN Films Prepared by Pulsed DC Reactive Sputtering in FTS Mode at High Al Content Sara Khamseh1; * , Masateru Nose2 , Tokimasa Kawabata1 , Atsushi Saiki1 , Kenji Matsuda1 , Kiyoshi Terayama1 and Susumu Ikeno1 1 2
Graduate School of Science and Engineering, University of Toyama, Toyama 933-0855, Japan Faculty of Art and Design, University of Toyama, Takaoka 933-8588, Japan
CrAlN films were prepared by a pulsed DC magnetron sputtering in FTS mode with Cr-Al alloy targets (Cr/Al = 30 at%/70 at%) in a mixed atmosphere of Ar and N2 . Effects of different deposition conditions (frequency, duty cycle, nitrogen flow rate and pulsed DC power) on the films structure and phase formation have been investigated. XRD analyses were carried out to determine the phases of the films. The surface morphology was observed using FE-SEM. Transmission electron microscopy studies were carried out for selected films. In order to investigate the relationship between the mechanical properties and microstructure of the films, the hardness was measured by a nanoindentation system. All films exhibited mixture of fcc-CrN and hcp-AlN structures with changing the sputtering conditions. With increasing pulse frequency, the relative percent of fcc-CrN phase decreased from 100% to 20%. In the current study it was found that there is an optimal duty cycle (around 82%) where the fcc-CrN relative% reaches to its highest value (about 98%). On the other hand, grain size of the films kept in the range of 20–25 nm below 82% of duty cycle, then increased up to 50 nm with increasing duty cycle. Changing the nitrogen flow rate affected the fcc-CrN relative % and films morphology. Under optimal nitrogen flow rate, fcc-CrN relative % reached to about 100% in addition to the formation of micro columnar morphology, which resulted in the highest plastic hardness of the films. It seems that the mechanical properties of the CrAlN films are directly influenced not only by fcc/hcp phase ratio, but also by morphology. These results suggest that the dominant fcc-CrN phase and high hardness of CrAlN films containing high Al content (Cr : Al ¼ 3 : 7) can be obtained under the restricted condition by the pulsed DC magnetron sputtering in FTS mode. [doi:10.2320/matertrans.MRA2008604] (Received February 5, 2008; Accepted June 9, 2008; Published July 24, 2008) Keywords: CrAlN, nanoindentation, transmission electron microscopy (TEM), facing target-type sputtering (FTS), pulsed magnetron sputtering
1.
Introduction
Transition metal nitride films such as TiN, CrN, TiAlN and CrAlN have been prepared intensively by magnetron sputtering system. Pulsed DC magnetron sputtering was developed initially as a means of suppressing the arcing at the target during reactive sputtering and offers an excellent opportunity to deposit non-conductive, high-quality coatings with outstanding deposition rates.1,2) Recently, pulsed DC unbalanced magnetron sputtering is shown to bring higher ion energy and ion flux to the growing film at the substrate, with modifying the intrinsic plasma parameters, which, in turn may have a strong influence on film structure and properties.1) Many papers have reported a positive effect of increased plasma density and higher ionization. The high mobility of ions and, in particular, electrons results from a high ionization of the pulsed plasma.3,4) This high mobility can then lead to higher adatom mobility, resulting in modified coating structures.2) There are some reports on the effect of pulsing parameters on the properties and structure of hard nitride coatings in Unbalanced Magnetron Sputtering system, but there are a few reports on these effects in Balanced Magnetron Sputtering system. Facing target type sputtering is one of the balanced magnetron sputtering system, which technique is frequently used for the deposition of magnetic materials5,6) hard coatings7) and transparent conducting coatings.8) We need to investigate the effect of pulsing parameters on the mechanical properties and morphology of nitride films in Balanced Magnetron Spattering system. *Graduate
Student, University of Toyama
In the last decades, research on hard coatings has been carried out into the development of new ternary nitrides consisting of transition elements and other metals.9–12) CrAlN coating is one of the most attractive nitride, because CrAlN coatings have been reported to exhibit higher oxidation resistance than TiAlN.13,14) According to previous reports, aluminum content strongly affects not only the structure of CrAlN coatings (fcc-CrN or hcp-AlN)15–18) but also the oxidation resistance.19,20) Although the theoretical solubility of Al in fcc-CrN is larger than in fcc-TiN,21) excess Al content over the Al solubility in fcc-CrAlN films causes the segregation of hcp-AlN, which alters the mechanical properties of the coatings.22–24) The maximum hardness of fcc-Cr1x Alx N coatings attributed to solid solution hardening was reported for x between 0.5 and 0.7.16,18) The hardness and maximum residual compressive stress was also reported to decrease with the appearance of the hcp-AlN phase.24) In order to address the problems for CrAlN coatings, the present study seeks to obtain experimental data for the effect of sputtering conditions, in the pulsed DC system in FTS mode, on the morphology and crystal structure of CrAlN coatings with high Al content. 2.
Experimental Procedure
CrAlN films were prepared in a pulsed DC sputtering apparatus which has a pair of targets facing each other (referred to as the facing target-type sputtering, Osaka Vacuum Co., Ltd., FTS-2R) as shown in Fig. 1, where a magnetic field of about 0.02 T is applied perpendicular to
Effect of Deposition Conditions on the Structure and Properties of CrAlN Films
Magnet
films. TEM samples were prepared and observed parallel (plane view) to the film surface. The plane-view sample was thinned by mechanical grinding, followed by ion milling with Arþ in a BAL-TEC, RES010 system operated at 5 kV with an incident angle ranging from 10 to 25 . TEM studies were carried out with a (TOPCON, EM-002B type) TEM operated at 120 kV. The crystal size in the films was estimated from the width of XRD peaks using Scherer’s equation.
Target
3.
Exhaust
Ar N2 Substrates Magnet
NS
2083
NS
Target
Results and Discussion
Substrates
fcc-CrN% ¼ ½I(CrN) =I(AlN) þ I(CrN) 100
(1122)
(1013)
(220)
(1120)
f=190kHz
(1012)
Where I(CrN) is the peak intensity of the fcc-CrN phase and I(AlN) is the peak intensity of the hcp-AlN phase from the XRD patterns in -2 mode. The peak intensity of (111) plane of the fcc-CrN phase and (101 0) plane of hcp-AlN phase were used for this calculation. When the sputtered film does not show a strong preferred orientation, phase ratio in the mixed phase film can be estimated roughly from the peak intensity of XRD pattern. Figure 3 depicts the variation of phase ratio of fcc-CrN with the pulsing frequency. The fccCrN% decrease from 100% to 20% with increasing frequency. This fact means that the film structure changed from dominant fcc-CrN at 120 kHz to the dominant hcp-AlN at higher frequencies.
(200)
the target planes to confine the plasma between the two targets and thereby maintain discharge even under a lower gas pressure than the limit of planar magnetron sputtering. Two rectangular plates (100 mm 160 mm 10 mm thickness) of alloy targets (Cr/Al = 30 at%/70 at%) (99.9%), were sputtered in a mixture of argon and nitrogen, both of 99.9999% purity. The system was evacuated to a vacuum better than 5 105 Pa (¼ 3:8 107 Torr) prior to deposition. The flow rate of each gas (Ar/N2 ) was independently controlled using a mass-flow controller. In order to avoid target poisoning and control the nitride formation precisely, the argon gas flow rate was fixed at 15 sccm and employed a nitrogen gas flow rate brought back from a high nitrogen flow rate to a lower nitrogen flow rate along the hysteresis loop of reactive sputtering.25,26) A mirror-polished silicon wafer 25 mm square was used as the substrate. All the substrates were cleaned ultrasonically with acetone, ethanol, and 2propanol, in this sequence, before sputtering deposition. The input power of pulsed DC, which was applied to both targets synchronously, was controlled in the range from 1.5 to 2 kw. The target-to-substrate distance was fixed at 115 mm. The substrate temperature was increased to about 150 C during deposition due to particle bombardment of the substrate even without bias application and substrate heating. The film thickness was constrained between 1.8 and 2 mm by controlling the sputtering time. The hardness was measured by a nanoindentation system (Fischer scope, H100C) at room temperature. The indentation was performed using a triangular Berkovitch diamond pyramid. The load was selected to keep an impression depth not more than 10% of the film thickness, so that the influence from the substrate can be neglected. The crystal structure of the films was identified by X-ray diffractometery using Cu K radiation with either a thin film or -2 goniometer (Philips X’pert system). When using the thin film goniometer, scans were made in the grazing angle mode (Seeman–Bohlin mode) with an incident beam angle of 2 . A (JEOL, JSM-6700F) field emission scanning electron microscopy (FE-SEM) was used to provide a high resolution scan on the plane view and fractured cross-section of the
(1010)
Schematic drawing of a facing target type sputtering apparatus.
(0002) (111)
Fig. 1
Intensity, I/(arbit, unit)
Pulsed-DC
3.1 Influence of frequency It is well known that pulsing condition such as frequency or duty cycle has a significant effect on the properties of CrAlN films.1,27) A series of CrAlN films were deposited at N2 /Ar flow rate of 30/15 sccm and pulsing power of 1.5 kW with a different frequencies varying from 120 to 190 kHz, when the duty cycle was adjusted between 6080% with respect to the stability of plasma at each frequency. The XRD patterns of the films prepared under different frequencies in the grazing angle mode are shown in Fig. 2. Analysis of the XRD results showed that with increasing frequency, the structure changed from the dominant fcc-CrN structure at 120 kHz to a mixed structure of fcc-CrN and hcp-AlN. The relative percentage of the fcc-CrN phase was calculated according to the following equation:
f=170kHz
f=150kHz
f=120kHz
20
30
40
50
60
70
80
2θ Fig. 2 X-ray diffraction patterns of CrAlN films at different frequencies measured by thin-film method.
S. Khamseh et al.
20
(1122)
(220)
Duty cycle=78% Duty cycle=68%
0 110
Duty Cycle=82%
(1120)
40
(111)
60
(0002)
CrN(%)
80
(1010)
Intensity, I/(arbit, unit)
100
(200)
2084
20
120
130
140
150
160
170
180
190
30
40
50
200
60
70
80
2θ
Frequancy,f/kHz
Lin et al. revealed that the pulsing conditions have a significant effect on the Al and Cr concentrations of the films. In an unbalanced magnetron sputtering system; with an increase in the pulse frequency and reverse time the Cr content of the films decreases, which is correlated to a simultaneous increase of the Al content of the films.27) Given that sputtering phenomena in our system have a similar tendency as those in Unbalanced Magnetron Sputtering, where Al content in the film seems to increase with increasing the pulse frequency.
Fig. 4 X-ray diffraction patterns of CrAlN films at different duty cycles measured by thin-film method.
grain size,
CrN%
100
55 90
50
80
45
70
40
60
35
50
30
40
25
30 20
20
3.2 Influence of duty cycle The films were prepared at different duty cycles with a fixed N2 /Ar flow rate of 30/15 sccm, the pulsing power of 1.5 kW and the frequency of 120 kHz. XRD patterns of the films in the grazing angle mode are shown in Fig. 4. A pronounced fcc-CrN phase was obtained in the films deposited at 82% duty cycle. The films that prepared at lower duty cycles showed mixed structure of fcc-CrN and hcp-AlN phases. The dependence of the grain size and fccCrN% on the duty cycle is shown in Fig. 5. The fcc-CrN% in the films increase with increasing duty cycle, reaching to almost 100% of fcc-CrN at 82% of duty cycle, then decrease to about 50%, whereas the grain size of the films keeps a constant value in the range of 20–25 nm below 82% of duty cycle, then increasing abruptly up to 50 nm at 88%. Lin et al. demonstrated that in the P-CFUBMS system, the calculated ion flux decreases with an increase in the duty cycle. On their view, with a decrease of the duty cycle (an increase of positive pulse time), the cathode is switched proportionally to a positive voltage for a longer period of time, which in turn provides more time for positive ions to stream away from the target. This increased escape time results in a higher number of ions gaining the extra kinetic energy made available from the positive potential switch, thereby increasing the flux of higher energy ions at the substrate.27) In the present study, similar phenomena have possibly occurred. Ion energy and flux seem to decrease in the plasma, with increasing the duty cycle. The low ion energy and ion flux leads to a low nucleation density due to the low mobility and diffusivity of the adatoms on the
CrN%
Influence of frequency on fcc-CrN% of CrAlN films.
Grain size, D/(nm)
Fig. 3
10
15 0
70
72
74
76
78
80
82
84
86
88
Duty Cycle(%) Fig. 5 Influence of duty cycle on the fcc-CrN% and grain size of CrAlN films.
substrate, which consequently increases the grain size of the films. It was reported that in unbalanced magnetron sputtering system, increasing the duty cycle leads to the decrease of the Al content and an increasing of the Cr content in the films.1,27) However, it seems that there is an optimal duty cycle (82%) where the fcc-CrN% reaches to its highest value (98%) for the FTS system. 3.3 Influence of nitrogen flow rate and pulsed DC power Table 1 summarizes the details of the film deposition, magnetron pulsing parameters and film properties. Nitrogen flow rate was changed in the range of 10 to 35 sccm while pulsed DC power was chosen 1.5 or 2 kW under the fixed frequency of 120 kHz and the duty cycle of 82%. XRD patterns in the grazing angle mode for the films prepared at different nitrogen flow rates and pulsed DC powers are shown in Fig. 6. The XRD results show that an amorphous like structure with broadened peaks is formed at a nitrogen flow rate of 10 sccm (S-1). In contrast, fcc-CrN and hcp-AlN were obtained in the film deposited at 20 sccm of nitrogen flow rate (S-2). Some unknown peaks are also found
Effect of Deposition Conditions on the Structure and Properties of CrAlN Films
N2 (sccm)
Power (W)
Hpl (GPa)
EPMA Cr/[Cr+Al]
Crat%
Alat%
Nat%
S-1
10
1.5
17.9
—
—
—
—
S-2
20
1.5
6.8
0.32
15.8
33.1
50.2
S-3
25
2.0
8.3
0.31
15.2
32.9
51.6
S-4
30
1.5
27.2
0.34
17.1
31.91
51.1
S-5
30
2.0
35.7
0.35
16.9
31.1
52.1
S-6
35
2.0
22.4
—
—
—
—
CrN%
35
100
30 80
25 20
60
15
40
10 20
5 0
(311)
(2021)
(220) (1013)
(1120) unknown
2kW 1.5kW 2kW 1.5kW
N2=30sccm N2=25sccm N2=20sccm N2=10sccm
1.5kW 40°
28
30
32
34
0 36
Fig. 7 Influence of nitrogen flow rate on the fcc-CrN% and Hpl of CrAlN films that prepared under pulsed DC power of 2 kW.
N2=30sccm
30°
26
N2 flow rate/sccm
(200)
(111) (1010) unknown (0002)
Intensity, I/(arbit unit)
24
20°
120
CrN(%)
Sample No
Hpl,
Grain size,
40
Grain size,D/nm/Hpl/Gpa
Table 1 Details of deposition parameters of the films that prepared under different deposition conditions (f = 120 kHz, Ar = 15 sccm, Duty cycle = 80%).
2085
50°
60°
70°
80°
2θ Fig. 6 X-ray diffraction patterns of CrAlN films at different nitrogen flow rates and pulse powers measured by thin-film method.
in the film deposited at 25 sccm of nitrogen flow rate (S-3). With further increase of the nitrogen flow rate to 30 sccm (S-4, 5), the fcc-CrN phase becomes predominate. Only the weak (101 0) peak of the hcp-AlN phase was observed in S-4, indicating low content of hcp-AlN phase in the coating. This result was examined by TEM observation, which will be discussed later. Peak positions of fcc-CrN phase shifted to higher 2 values in comparison with the standard JCPDS card values for fccCrN phase. This shift in 2 can be attributed to a decrease in lattice parameter of CrAlN (approximately 1%) due to the substitution of some of the Cr atoms by Al atoms in the fccCrN lattice. In the hcp-AlN phase, the peak of (101 0) shifted to lower diffraction angles while those of (0002) shifted to higher angles in comparison with the standard JCPDS card values for hcp-AlN phase. Change of lattice parameter was evaluated by the equation, a(AlN) ¼ ½a(XRD) a(AlN) =a(AlN) . The lattice parameters changed between 1%2% in a-axis, and between 0:06% 4% in c-axis. Therefore, the lattice parameter anisotropically changed with respect to the (101 0) and (0002) planes. Kimura et al.18) showed that the crystal structures of Crð1xÞ Alx N film changed from the NaCl-type to wurtzitetype at Al contents x ¼ 0:6{0:7. For Al content X > 0:7, the lattice spacing of hcp-AlN expanded in the a-direction and shrank in the c-direction with addition of Cr atom into AlN
lattice. In the present study, in spite of coexistence of two fcc-CrN and hcp-AlN phases in the films, the anisotropic lattice change seems to happen. One can see that plastic hardness, Hpl , of the films ranges between 6.8 to 35.7 GPa. Figure 7 illustrates the grain size, D, and fcc-CrN% and the plastic hardness, Hpl , as a function of nitrogen flow rate for the films prepared at 2 kW. Grain size, D, which was calculated from the peak width of (111) plane, is almost constant around 25 nm versus N2 flow rate. On the other hand, change of fcc-CrN% and Hpl show same tendency and reach to a maximum at nitrogen flow rate of 30 sccm. In a same way, for the film prepared at 1.5 kW (S-4 film) dominant fcc-CrN phase of about 98% was obtained at N2 flow rate of 30 sccm. The Gibb’s free energy of formation of hcp-AlN phase is lower than that of fcc-CrN phase.28) Additionally, in the unbalanced magnetron sputtering system, the amount of aluminum contained in the (Cr,Al)N coatings was minimal at high nitrogen pressures.10) Mientus et al. reported that the minimum nitrogen pressure for the formation of hcp-AlN phase was about 0.08 Pa, while the pressure for fcc-CrN phase was about 0.2 Pa in the DC magnetron sputtering system of elemental targets.29) This may mean that with increasing the nitrogen flow rate, the Al content decreases in the coatings, which consequently increases the Cr content in the films, leading to the formation of fcc-CrN phase at higher nitrogen flow rates. Film composition and the content ratio of metal atoms (Cr/[Cr+Al]) for S-2, 3, 4 and 5 are shown in Table 1. Although there was a small difference less than 2% in composition between these films, content ratio of Cr and Al was affected more significantly by nitrogen flow rates. As for the film S-2, 3 (N2 flow rates of 20, 25 sccm) Cr content ratios (Cr/[Cr+Al]) are 0:31{0:32, which are similar to the ratio for the target (0.3). On the other hand, the ratios for the films S-4 and 5 are 0:34{0:35, that means, the Cr content ratio increased a little for the films that were prepared under the nitrogen flow rate of 30 sccm in both powers compared to S-2 and 3. However, the maximum fcc-CrN% at 30 sccm of N2 flow rate in this study cannot be explained perfectly by this speculation. It might be attributed to the difference of deposition system.
2086
S. Khamseh et al.
(a)
300nm
(e)
300nm
(b)
1 µm
(f)
1 µm
(c)
(d)
300nm
1 µm
(g)
(h)
300nm
1 µm
Fig. 8 Plane-view and cross-sectional FE-SEM images of CrAlN films prepared under different deposition conditions (a), (b) N2 = 10 sccm, pulse power = 1.5 kW (c), (d) N2 = 20 sccm, pulse power = 1.5 kW (e), (f) N2 = 25 sccm, pulse power = 2 kW (g), (h) N2 = 30 sccm, pulse power = 1.5 kW.
3.4 Microstructure evaluation FE-SEM observations were performed to determine the morphology of the coatings. Figure 8 provides the development of the morphology upon the nitrogen flow rate and pulsed DC power. In the films prepared under a low nitrogen flow rate (S-1), fine grained structure was observed to develop Figs. 8(a) and 8(b). As shown in Figs. 8(a) and 8(b) the FE-SEM imaging of the film prepared under the nitrogen flow rate of 10 sccm (S-1) was very smooth with a grain size that was out of the resolution of FE-SEM. This film seems to consist of very fine grains as can be seen in an amorphous or nanocrystalline films. In the film prepared under nitrogen flow rate of 20 sccm (S-2) the nodular like outgrowth was observed on the matrix that having fine columnar structure as shown in Fig. 8(c). This kind of morphological feature is called node or nodular defect.30) Xing-Zhao Ding et al. prepared CrAlN films having high aluminum content with an unbalanced magnetron sputtering system and reported that arcing on the target surface may have caused this kind of inhemogenity in the film.31) For the film prepared at a pulsed DC power of 2 kW under nitrogen flow rate of 25 sccm (S-3), a pebble-like microstructure without any node was observed (Fig. 8(e)). However, a significant increase of the column width was observed with an increase in nitrogen flow rate and pulsed DC power. The cross-sectional micrograph (Fig. 8(f)) of S-3 film indicates typical Zone-I structure,32) showing a wide columnar structure with large column width between 660 to 1200 nm. The films of this type are characterized by a spiky topography due to tightly packed columns with hemispherical tops.30) In contrast, S-4 film showed a pyramid-like surface feature. This condition results in a finer columnar size, denser structure and smoother surface morphology, without any nodes as shown in Fig. 8(g) and (h). The column width ranged between 160 to 200 nm, which is slightly larger than that of matrix for S-2. However, the structure of S-4 films possess a micro-columnar structure,
comparable to the Zone-I structure of Thornton’s zone models. TEM observations were performed in order to determine the differences in the microstructure and phase formation of the films. Figure 9(a) is the bright field TEM micrograph of the film prepared under the nitrogen flow rate of 20 sccm (S-2). In comparison with FE-SEM image (Fig. 8(c)) bright field TEM image of the film exhibits a polycrystalline structure with some nodes which seem thicker than matrix as can be seen in the dark contrast of the image (Fig. 9(a)). From dark field images of the film, average grain sizes of the matrix and node ranged between 60–65 nm and 270–300 nm, respectively. The corresponding electron diffraction patterns of white contrast part (matrix) and dark contrast part (node) of Fig. 9(a) are shown in Figs. 9(b) and 9(c). Analysis of these diffraction patterns for the matrix and the node reveals the crystal structure of fcc-CrN phase and hcp-AlN phase respectively. These observations confirmed the results of XRD shown in Fig. 6. Plane view TEM observation was carried out for S-3 film. Figure 10(a) exhibits a bright field TEM image of the film corresponding to the polycrystalline structure with pebblelike morphology as shown in Fig. 8(e). The related diffraction pattern shown in Fig. 10(b) indicates sharp diffraction spots arranged on a circle, which suggest polycrystalline material consisting of relatively large grains. SAED analysis reveals the co-existence of fcc-CrN, hcp-AlN and some unknown phase in the film. Figure 10(c) is a dark field image for Spot-1 illustrated in Fig. 10(b). The grain size, obtained from the dark field image of the film, was in the range of 200300 nm. Analysis of SAED revealed the existence of hcp-AlN phase in this area as shown in Fig. 10(d). Dark field image and related diffraction pattern for spot-2 marked on the main diffraction pattern are shown in Fig. 10(e). The grain size of the film obtained from this dark field image is between 80300 nm. Analyses of related diffraction pattern
Effect of Deposition Conditions on the Structure and Properties of CrAlN Films
2087
(a)
500nm (c)
(b) 200
220
0002 1011 1010
Fig. 9 Plane-view TEM micrograph and electron diffraction patterns of the film with nitrogen flow rate of 20 sccm (power = 1.5 kW, Hpl ¼ 6:8 GPa, S-2) (a) bright field image (b) SAED of main matrix with fcc-CrN structure ([001] oriented grain) (c) SAED of staked area with hcp-AlN structure ([211 0] oriented grain).
(c)
(a)
200nm
(b)
(e)
100nm Spot-3
Spot-1
Spot-2
300nm
(f)
(d)
200
0002 1011
220 200
1010
(g)
(h)
61°
28°
300nm Fig. 10 Plane-view TEM images and electron diffraction patterns of the film with nitrogen flow rate of 25 sccm (power = 2.0 kW, Hpl ¼ 8:3 GPa, S-3) (a) bright field image (b) SAED with mixed structure of fcc-CrN and hcp-AlN and unknown phase (c) dark field image of Spot-1 (d) SAED pattern of bright grain of image (c) with hcp-AlN structure ([211 0] oriented grain) (e) dark filed image of Spot2 (f) SAED of bright grain in (e) with fcc-CrN structure ([001] oriented grain) (g), dark field image of Spot-3 (h), SAED of bright grain in (g) with unknown structure.
2088
S. Khamseh et al.
(a)
(b)
500nm
500nm
(c) 222 311 220 200 111
Fig. 11 Plane-view TEM micrograph and corresponding electron diffraction pattern of the film with nitrogen flow rate of 30 sccm with dominant fcc-CrN structure (a) bright field image (b) dark field image of first ring with grain size between 60–100 nm (c) SAED with fcc-CrN structure (Hpl ¼ 27:2 GPa, S-4).
(Fig. 10(f)) revealed the growth of fcc-CrN phase in this grain. Dark field image and the related diffraction pattern for spot-3 are illustrated in Fig. 10(g). The related SAED pattern (Fig. 10(h)) has proved to be similar to the hcp-AlN phase since the angles are same as that for the hcp phase with foil plane of (211 0), but the lattice spacing are different from those of hcp-AlN phase. Lattice spacing d of the (0002) plane of this unknown hcp phase was shifted about 2.4% from d of hcp-AlN (0002) plane. However, d of (101 0) and d of (101 1) planes of this unknown hcp phase was shifted about 5:9% and 6:5% from the lattice spacing of the corresponding hcpAlN planes, respectively (Fig. 10(d)). In any case, microstructure evaluation using FE-SEM and TEM revealed that the S-3 film exhibits typical Zone-I structure with large columnar width of about 660 to 1200 nm, consisting of fccCrN and hcp-AlN phases mainly. The low plastic hardness of 8.3 GPa for the S-3 film is probably attributed to high amount of hcp-AlN phase in the film, which hardness is about half of fcc-CrN phase (with hardness about 19 GPa), and the rough morphology.33) Since a weak (101 0) peak (2 ¼ 33:2 ) of hcp-AlN phase were observed in the XRD pattern of the S-4 film consisting of the dominant presence (100%) of fcc-CrN, TEM observation was carried out for S-4 film. Figure 11(a) shows a bright field image of this film. Sharp tips of the pyramid like structure observed in bright field image of the film are in conformity with the FE-SEM image (Fig. 8(g)). Analysis of the related diffraction pattern revealed a dominant crystal
structure for fcc-CrN phase. Dark field image of this film indicates that the average grain size is between 6787 nm. In order to examine the existence of hcp-AlN phase in the film TEM observation was done at higher magnifications. Figure 12(a) shows the bright field image of the same film. Analysis of corresponding electron diffraction pattern of the film reveals the existence of a mixed structure of fcc-CrN and hcp-AlN phases in the film (Fig. 12(b)). Dark field image corresponding to (101 0) diffraction ring is shown in Fig. 12(c). Analysis of the related diffraction pattern of the bright grain indicates the existence of hcp-AlN phase with small grain size (26 nm). As shown in Fig. 12(c), a low distribution of this hcp-AlN phase was also observed in the other areas of the film. From precise analysis of the weak XRD peaks and SAED patterns, hcp-AlN phase is determined as an exact hcp-AlN phase. Therefore, we can conclude that the pure hcp-AlN has been formed in S-4 film and that no Cr atom is substituted to Al atom in the hcp-AlN crystals. It should be noticed that even in the S-5 film exhibiting no hcp-AlN phase in XRD chart, traces of the hcp-AlN phase were detected by TEM diffraction (is not shown here). According to the previous reports as for unbalanced magnetron sputtering, CrAlN films maintained fcc-CrN structure even for the Al content more than 71%.1) On the other hand, it has been reported that the structure of CrAlN film prepared by AIP method changes from the fcc-CrN to hcp-AlN phase at the Al/(Al+Cr) ratios of X0:6{0:7.18)
Effect of Deposition Conditions on the Structure and Properties of CrAlN Films
2089
(c)
(a)
100nm
100nm
(d)
(b)
0110
1120
60°
1010 222 200
1010
Fig. 12 Plane-view TEM micrograph and corresponding electron diffraction pattern of the film with nitrogen flow rate of 30 sccm with dominant fcc-CrN structure (Hpl ¼ 27:2 GPa, S-4) (a) bright field image (b) SAED pattern with mixed structure of fcc-CrN and hcp-AlN structures (c) dark field image of third ring with hcp-AlN structure[112 1] oriented grain.
Although the details are not known at this moment, we can conclude that it is very difficult to prepare CrAlN films consisting of 100% fcc-CrN phase without any content of hcp-AlN phase using Cr70%-Al30% alloy target with a facing target type sputtering system. On the other hand, it is worth noting that the films prepared at the narrow range of sputtering condition exhibited higher Hpl value ranging from 28 to 35 GPa (S-4 and S-5 films). The increased hardness of S-4 and S-5 films can be explained by two facts: one is improved density of microstructure showing a fine columnar morphology with relatively small grain size (60 nm), and the second is the contribution of the dominant fcc-CrN phase with very little amount of hcp-AlN phase. 4.
Conclusion
CrAlN films has been prepared by a pulsed DC reactive sputtering in FTS system with (Cr/Al = 30 at%/70 at%) alloy targets. Effects of different deposition conditions (nitrogen flow rate, frequency and duty cycle) on the film structure and phase formation have been investigated using XRD, FE-SEM, and TEM. It was found that hcp-AlN/fcc-CrN phase ratio and grain size of the films changed with varying the frequency and duty cycle. Under the optimum condition of the frequency and duty cycle, nitrogen flow rate did not affect the grain size significantly but the morphology and hcp-AlN/fcc-CrN phase ratio. The comparative study by nanoindentation measurement and microstructure evaluation demonstrated that the plastic hardness of these films were in the range of 6.8 GPa to 35.7 GPa, which were affected strongly by fcc-
CrN% and the morphology. We can conclude that in addition to the sputtering conditions, the pulsing parameters should carefully controlled in order to obtain fcc-CrN phase with proper morphology and grain size, consequently leading to good mechanical properties in the films. However, the precise observation of microstructure of the films using TEM revealed that even the hardest films contained hcp-AlN phase. Accordingly, it seems to be difficult to obtain CrAlN films consisting of 100% fcc-CrN phase using Cr30-Al70 alloy target in the FTS system. Acknowledgements The authors wish to express their thanks for the financial support by a grant for scientific research from Japan society for the promotion of science. Grateful thanks are specially dedicated to Dr. Y. Sakamoto for help in executing the TEM sample preparation and to Mr. H. Ura and Mrs. Juli Ma for their help in sample preparation. REFERENCES 1) J. Lin, J. J. Moore, B. Mishra, W. D. Sproul and J. A. Rees: Surf. Coat. Technol. 201 (2007) 4640–4652. 2) K. Bobzin, E. Lugscheider and M. Maes: Surf. Coat. Technol. 200 (2005) 1560–1565. 3) J. C. Sellers: Surf. Coat. Technol. 94–95 (1997) 184–188. 4) G. Erkens, R. Cremer, T. Hamoudi, K. D. Bouzakis, I. Mirisidis, S. Hadjiyiannis, G. Skordaris, A. Asimakopoulos, S. Kombogiannis, J. Anastopoulos and K. Efstathiou: Surf. Coat. Technol. 177–178 (2004) 727. 5) H. Ito, M. Yamaguchi and M. Naoe: J. Appl. Phys. 67 (1990) 5307.
2090
S. Khamseh et al.
6) E. Y. Jiang and Z. J. Wang: J. Mag. Magn. Mater. 81 (1989) 25. 7) W. K. Lee, T. Machino and T. Sugihara: Thin Solid Films 224 (1993) 105. 8) D. C. Sun, E. Y. Jiang, H. Liu Ming and C. Lin: Z. Phys. D 28 (1995) 4. 9) H. Willmann: Scripta Materialia. 54 (2006) 1847–1851. 10) R. Wuhrer: Scripta Materialia. 50 (2004) 1461–1466. 11) H. Hasegawa: Surf. Coat. Technol. 188–189 (2004) 234. 12) M. Fenker, M. Balzer and H. Kappl: Thin Solid Films 515 (2006) 27–32. 13) O. Banakh, P. E. Schmid, R. Sanjines and F. Levy: Surf. Coat. Technol. 163 (2003) 57. 14) M. Kawate, A. K. Hashimoto and T. Suzuki: Surf. Coat. Technol. 165 (2003) 163. 15) Y. Makino: Surf. Coat. Technol. 193 (2005) 185–191. 16) H. C. Barshilia, N. Selvakumar, B. Deepthi and K. S. Rajam: Surf. Coat. Technol. 201 (2006) 2193–2201. 17) A. Sugishima, H. Kajioka and Y. Makino: Surf. Coat. Technol. 97 (1997) 590–594. 18) A. Kimura, M. Kawate, H. Hasegawa and T. Suzuki: Surf. Coat. Technol. 169–170 (2003) 367. 19) M. Kawate, A. K. Hashimoto and T. Suzuki: Surf. Coat. Technol. 165 (2003) 163.
20) A. E. Reiter, V. H. Derflinger and B. Hanselmann: Surf. Coat. Technol. 200 (2005) 2114–2122. 21) Y. Makino: ISIJ Int. 38 (1998) 925. 22) S. PalDey and S. C. Deevi: Mat. Sci. Eng. A 342 (2003) 58. 23) A. Horling, L. Hultman and M. Oden: J. Vac. Sci. Technol. A 20 (2002) 1815. 24) T. Suzuki, Y. Makina and M. Samandi: J. Mat. Sci. 35 (2000) 4193. 25) A. Mumtaz and W. H. Class: J. Vac. Sci. Technol. 20 (1982) 345. 26) R. G. Jahanson and W. G. Carruthers: J. Vac. Sci. Technol. A 5 (1987) 2246. 27) J. Lin, J. J. Moore, B. Mishra, M. Pinkas, W. D. Sproul and J. A. Rees: Surf. Coat. Technol. 202 (2008) 1418–1436. 28) R. C. Weast and M. J. Astle (Eds.): CRC Hand Book of Chemistry and Physics, (CRC Press Inc., Florida, 1982). 29) R. Mientus and K. Ellmer: Surf. Coat. Technol. 116–119 (1999) 1093–1101. ˆ ny and G. RadnoA ˆ czi: Thin Solid Films 347 (1999) 133– 30) Z. CzigaA 145. 31) X.-Z. Ding and X. T. Zeng: Surf. Coat. Technol. 200 (2005) 1372– 1376. 32) J. A. Thornton: Ann. Rev. Mater. Sci. 7 (1977) 239–266. 33) P. Hones and R. Sanjines: Surf. Coat. Technol. 94–95 (1997) 398.
