IR characterization of a-C:H:N films sputtered in Ar/CH4/N2 plasma
-
Upload
gabriel-lazar -
Category
Documents
-
view
212 -
download
0
Transcript of IR characterization of a-C:H:N films sputtered in Ar/CH4/N2 plasma
Journal of Non-Crystalline Solids 331 (2003) 70–78
www.elsevier.com/locate/jnoncrysol
IR characterization of a-C:H:N films sputteredin Ar/CH4/N2 plasma
Gabriel Lazar *, Iuliana Lazar
Bacau University, Calea Marasesti 157, Bacau, Romania
Received 27 September 2002; received in revised form 6 June 2003
Abstract
Amorphous nitrogenated carbon films (a-C:H:N) were deposited by rf magnetron sputtering of a graphite target in
an Ar/CH4/N2 plasma. The films were characterized by infrared spectroscopy. An increase in the intensity of absorption
maximum at 3300–3400 and 1620 cm�1 of the IR spectra was observed. For high nitrogen concentrations, a new
maximum appears at 2150 cm�1 indicating that the C and N atoms are chemically bonded in the film. The analysis of
the broad absorption band observed between 1700 and 1000 cm�1 was made. Deconvolution of this band showed a six
band system, indicating a lack of aromatic structures. The content of nitrogen and sp2 bonded carbon in the films was
seen to increase with a growth of the N2 partial pressure in the deposition gas.
� 2003 Elsevier B.V. All rights reserved.
1. Introduction
Much effort has been devoted to the study of
amorphous carbon nitride (a-C:N:H) films because
of their remarkable mechanical and tribological
properties and more recently, for their potentialapplications as electronic materials in cold cathode
displays and in electrochemical studies of water
treatments [1]. The crystalline phase of carbon
nitride, especially b-C3N4, is expected to be a very
hard material, as shown theoretically by Liu and
Cohen [2].
One of the key factors in determining the
structure and properties of hard amorphous car-bon films is the hybridization state of the atoms
* Corresponding author.
E-mail address: [email protected] (G. Lazar).
0022-3093/$ - see front matter � 2003 Elsevier B.V. All rights reserv
doi:10.1016/j.jnoncrysol.2003.09.004
that form the amorphous network [3]. Various
kinds of preparation methods have been applied to
fabricate carbon nitride films, but samples with
sufficient amount of crystallized C3N4 phase or
with mechanical properties comparable to the
predicted values have not been reported. The in-ternal stress considerably decreases with the addi-
tion of nitrogen, without sacrificing hardness, up
to an N incorporation of 2%, by the decrease of
the amount of unbound hydrogen in the film [4,5].
At higher concentrations, hardness is reduced due
to the development of graphitic domain. The in-
corporation of nitrogen could also affect the
amount of bonded hydrogen in the films [6].Franceschini et al. [3] reported that nitrogen in-
corporation strongly affects the hybridization state
of carbon atoms by increasing the sp2 fraction for
both nitrogen precursor gas N2 and NH3. At the
same time the nitrogen can be used to dope the
ed.
G. Lazar, I. Lazar / Journal of Non-Crystalline Solids 331 (2003) 70–78 71
material; however, the doping process remainedinefficient because the diversity of bonding con-
figuration allowed nitrogen to be incorporated in
different environments [7].
Infrared spectroscopy (IR) is a non-destructive
and widely available technique to probe the
bonding in carbon based films. IR absorption is
widely used in a-C:H films to determine the CH
bonding configuration [8]. The interpretation ofthe absorption IR spectra for nitrogen containing
carbon films is not very clear yet. The problem is
the broad band between 1700 and 1000 cm�1 that
represents a superposition of several bond contri-
butions.
In literature, three major assignments can be
found for this band: (i) the Raman �G� and �D�bands rendered infrared active by the incorpora-tion of nitrogen into the graphitic rings [9,10], (ii)
various bending modes of CC, NC, CH and NH
bonds [11,12] and (iii) a combination of the first
two [13–15].
This paper is an attempt to make this problem
clearer, starting from interpreting IR absorption
spectra for hydrogen and sp3 carbon rich, poly-
mer-like nitrogenated amorphous carbon films.
5000 4000 3000 2000 10000
20
40
60
80
100
120
140
160
180
200
1375
14501610218029103400
12% N2 8% N2 5% N2 2.5% N2 1% N2 0% N2
IR T
rans
mita
nce
(a.u
.)
wavelength (cm -1)
Fig. 1. The transmission spectra in the range 4600–650 cm�1,
for six samples deposited in argon/methane/nitrogen atmo-
sphere, with different nitrogen proportions and for a sample
deposited without nitrogen.
2. Experiment
Thin films were produced by magnetron sput-
tering in a home-built system from pure graphite
target. In brief, the system consists of a cylindrical
stainless-steel chamber with one magnetron(cathode) and two anodes [16]. In this work, a
13.56 MHz rf generator supplies power to the
system of the cathode and circular anode, and the
planar anode is biased by a dc voltage. Films, 0.2–
2.2 lm thick, were deposited on boro-silicate
(microscope slides) glass and silicon wiskers at a
target-to-substrate distance of 3 cm. The propor-
tion between Ar and methane was maintainedconstant at 1/1 and the concentration of the ni-
trogen in the gas mixture was varied from 0% to
12%. The unusual use of methane in the gas mix-
ture in a sputtering process implies a combination
of sputtering and CVD deposition. The discharge
chamber was first pumped down to 10�5 Torr and
then filled with gas. During sputtering, the gas
pressure was kept at 10�2 Torr. The substratetemperature remains below 100 �C. The appliedsubstrate bias voltage was )200 V for all samples.
IR transmission of films deposited on Si substrates
was measured with a IR double beam spectro-
photometer. Spectra were recorded in the wave-
number range from 650 to 4600 cm�1. Parts of the
spectra were fitted using Gaussian curves, centered
in the positions determined from the literature.Optical absorption of the films was measured in
the wavelength range 700–400 nm using an UV–
VIS spectrophotometer. The absorption coefficient
a was calculated using Lambert�s Law and was
explored as a function of photon energy hm. Theoptical band gap is estimated from a ðahmÞ1=2 ¼f ðhmÞ Tauc plot.
3. Experimental results
The results for nitrogen-free a-C:H films are
presented in an earlier paper [16]. Basically, the
films deposited by magnetron sputtering from
graphite target in Ar/CH4 atmosphere presents a
sp3 carbon proportion of 85–90% and an atomicconcentration of carbon-bonded hydrogen of 45%.
The great proportion of the sp3 fraction in the
films and the high amount of the hydrogen are
features of the soft polymeric films. As shown in
Fig. 1, the transmission spectrum of the nitrogen-
3100 3000 2900 2800 27000
500
1000
1500
5
4
321
α (c
m-1
)
wavenumber (cm-1)
Fig. 2. The infrared absorption spectrum for the C–H
stretching mode region for a sample obtained in following
conditions: working gas Ar/CH4/N2¼ 46%/46%/8%, gas pres-sure 10�2 Torr and substrate negative bias voltage 200 V.
Table 1
C–H stretch absorption bands deconvoluted from the Fig. 2
absorption peak
Curve
number
Observed vibration
frequency (cm�1)
Configuration
1 3027 sp2 CH2 (olefinic)
2 2998 sp2 CH (olefinic)
3 2956 sp3 CH3 (asymmetrical)
4 2921 sp3 CH2 (asymmetrical)
sp3 CH
5 2858 sp3 CH2 (symmetrical)
72 G. Lazar, I. Lazar / Journal of Non-Crystalline Solids 331 (2003) 70–78
free film presents an absorption peak at 2900 cm�1,common for all hydrogenated carbon films and
other tree peaks: at 1370, 1450 cm�1 (both as-
signed to CH bending modes [17]) and 1600 cm�1
(related to olefinic C@C bonds [18]). These three
peaks are very important for nitrogenated films,
because of their position, inside of the main ab-
sorption band (1700–1000 cm�1).
Fig. 1 shows the transmission spectra in therange 4600–650 cm�1, for six samples deposited in
argon/methane/nitrogen atmosphere, with differ-
ent nitrogen proportions and for a sample depos-
ited without nitrogen. All the films were deposited
in the following conditions: gas pressure 10�2 Torr
and substrate negative bias voltage 200 V.
The spectra show major modifications when
nitrogen concentration increases. All the nitro-genated films present a broad feature between 3700
and 3000 cm�1, associated with N–H bonds [11].
The amplitude of this absorption band increases
with the nitrogen proportion in the deposition gas
mixture indicating an increasing of the nitrogen
proportion in films.
The 2900 cm�1 absorption peak, which shows
the C–H bonds type and concentration, initiallyincreases with the growth of the nitrogen propor-
tion, then it decreases, but the position of the peak
remains the same. This behavior shows that, first,
the concentration of the carbon bonded hydrogen
is increasing and then it is decreasing, because of
the growth of the nitrogen proportion in films.
These would mean that carbon atoms were re-
placed by nitrogen atoms with an increase of theN2 partial pressure. The shape and position of the
peak are similar for all the samples, indicating a
low variation of the proportion of sp3 bonds. The
role of hydrogen in passivating the dangling bonds
or defect states is very important since it controls
the properties of film [5]. Addition of nitrogen in
the precursor may affect these defect states and
reduce the unbound hydrogen, which will effec-tively improve the adhesion of the film on the
substrate. The evidence of the very high value of
the sp3 fraction is shown in Fig. 2 where it is
presented the infrared absorption spectrum for the
C–H stretching mode region for a sample obtained
in the following conditions: working gas Ar/CH4/
N2¼ 46%/46%/8%, gas pressure 10�2 Torr and
substrate negative bias voltage 200 V. The spec-
trum can clearly be decomposed to five Gaussian
components, as shown in Fig. 2. The positions ofthese components are in very good agreement with
other results [1,16,18] and are summarized in
Table 1. From the great difference between sp3
peaks (3–5 in Fig. 2) and sp2 peaks (1 and 2 in
Fig. 2) area we may conclude that the type bond-
ing of C adjacent to H is predominantly sp3 [16].
When the proportion of the nitrogen increases,
a new absorption maximum appears at 2150 cm�1,related to CBN (nitrile) group [12,14,15] and as-
sociated to a high ion energy during the film de-
position process [6]. This peak is often observed in
the IR spectra of the samples with high nitrogen
content, suggesting that the C and N atoms are
chemically bonded in the film.
The broad band observed between 1700 and
1000 cm�1 for all nitrogenated samples evidently
Table 2
Major peaks positions in the 1700–1000 cm�1 absorption band
of hydrogenated carbon nitride films
Wavenumber
(cm�1)
Assignment Reference
1620–1650 C@C, C@N, NHx [15,17,19–22]
�1600 C@N, C@C [17,23,24]
1550–1570 Raman G, C@N [9,10,13–15,25]
1500–1510 C–N, C@N, C@C [12,14,19,26]
1450 sp3 CHx [17]
1360–1380 Raman D, CH3, CC,
C@N[9,10,13–15,17,19]
1300–1350 sp2 carbon, C–N,
C@N[12,14,23]
1220–1265 C–N (in C3N4), CC,
C@N[19,20,25,27,28]
1020–1150 C–N (aliphatic), N–H [19,29]
G. Lazar, I. Lazar / Journal of Non-Crystalline Solids 331 (2003) 70–78 73
represents a superposition of several contributionsand the paper is focused on this feature.
At a first glance, this broad band is composed
of at least five components. For a possible as-
signment of the individual absorption bands to the
corresponding bond configuration an analysis of
the published data is necessary. Despite a large
variation in the peaks positions and assignments,
nine main bands can be identified (Table 2).Not all these maximums can be found in the
absorption spectrum for a certain film. Each peak
can be correlated with other peaks in the absorp-
tion spectrum or with the results of other experi-
ments. An attempt to deconvolute the absorption
band of the films using nine gaussians failed.
4. Discussion
The analysis of the whole absorption spectrum
and the known IR absorption properties of the
nitrogen-free films impose some conditions.
The peaks at 1600, 1450 and 1370 cm�1 must
appear in the deconvolution. The 2900 cm�1 peak
position and amplitude suggest large quantities ofsp3 bonded CH2 and CH3 groups, that implies the
presence of the 1450 and 1370 cm�1 peaks. The
presence of the C@C peak at 1600 cm�1 in nitro-
gen-free films suggests a similar position of the
peak in the nitrogenated samples spectra.
The carbon based polymer-like materials are
mainly composed of a complicated and entangled
network of cross-linked alkane-like chains withsome double or triple bonds [17]. In the absorption
spectra, the 1600 cm�1 C@C band is not accom-
panied by sharp bands in the 950–650 cm�1 region
or at 3050 cm�1, and this fact offers evidence that
sixhold aromatic structures are not major com-
ponents in these polymeric films [30]. In this
conditions, the influence of the Raman D (at 1360–
1380 cm�1) and Raman G (at 1550–1570 cm�1)peaks can be neglected but the 1370 cm�1 feature
still remains in the deconvolution. The 1500–1510
cm�1 peak is associated mainly with nitrogen
bonded in pyridine-like structures [31] and can be
neglected as well. The major contribution of the
1220–1265 cm�1 is from C–N bonds in C3N4
structures [20,31]. The great disorder of the poly-
meric films implies a small contribution of thisbond to the absorption band and we can search
the C–N bonds at 1300–1350 and 1020–1150 cm�1.
The contribution of NH bonds must also appear at
1620–1650 and 1020–1100 cm�1 [21,29]. In con-
clusion, for the investigated films, a six peaks de-
convolution must be tried.
Fig. 3 shows the 1000–1800 cm�1 absorption
band of the films presented in Fig. 1. The ab-sorption intensity clearly increases in this domain
with the nitrogen proportion in the deposition gas.
The results of the peaks deconvolution for nitro-
gen containing films are presented in Table 3.
The verification of the proposed band system
can be made by analysing the peaks position for all
samples. If the system is good, the peaks centre
must be in the proposed region and small differ-ences can appear between the peaks position for
different films. As shown in Fig. 4, the position of
the all six deconvoluted peaks has a very low
variation for all the nitrogenated films. The posi-
tion of the three peaks common for all the films
(nitrogenated or not) has a remarkably low vari-
ation around proposed values: 1375, 1450 and
1600 cm�1. This fact confirms the absence of theother absorption bands in 1450–1600 cm�1 region.
The other peaks present law amplitude variations
around a value from the proposed domain.
For an assignation of the absorption bands, the
area and proportion in the total band as a function
of the nitrogen proportion in the deposition gas
are presented for all deconvoluted peaks in Fig. 5.
0
500
1000
1500
2000
2500
3000
6
5
4
3
2
1
12 % N2
abso
rban
ce (c
m-1)
0
500
1000
1500
2000
2500
3000
8 % N2
0
500
1000
1500
2000
2500
3000
5 % N2
abso
rban
ce (c
m-1)
0
500
1000
1500
2000
2500
3000
2.5 % N2
800 1000 1200 1400 1600 18000
500
1000
1500
2000
2500
3000
1 % N2
wavenumber (cm-1)
abso
rban
ce (c
m-1)
800 1000 1200 1400 1600 18000
500
1000
1500
2000
2500
3000
0 % N2
wavenumber (cm-1)
Fig. 3. The deconvolution of the 1000–1800 cm�1 absorption band of the six films presented in Fig. 1.
74 G. Lazar, I. Lazar / Journal of Non-Crystalline Solids 331 (2003) 70–78
The peak proportion is calculated making the ratio
between the peak area and the total band area.
The first peak, centred around 1100 cm�1, in-
creases rapidly for small proportion of the nitro-
gen in working gas, and decreases for great
nitrogen proportions. This absorption cannot be
considered to be related to the N–H bond, because
of the constant increases of the absorption in the
3300 cm�1 region with an increase of the nitrogen
partial pressure. The band would mostly be due to
the C–N bond. The decrease of the band intensity
must be correlated with an increase of the sp2
carbon proportion, as a result of the increases of
the nitrogen content in the films [29].
The intensity of the second peak increases al-
most linearly with an increase of the nitrogen
Table 3
The results of the peaks deconvolution in the 1000–1800 cm�1 domain for different proportion of nitrogen in the deposition gas
Nitrogen proportion Peak position (cm�1) Peak area (cm�2) Peak proportion in the band (%) Total band area (cm�2)
1% 1077.1 36 188 14.4 251 932
1341.3 93 357 37.1
1374.1 4731 1.9
1450.3 26 121 10.4
1599.1 36 540 14.5
1668.4 54 995 21.8
2.5% 1134.9 79 511 13.7 578 919
1332.5 159 772 27.6
1374.6 12 355 2.1
1449.5 68 001 11.7
1599.7 178 086 30.7
1635.2 81 194 14.0
5% 1120.3 61 300 10.8 565 276
1329.7 181 940 32.2
1374.4 11 941 2.1
1450.7 68 268 12.1
1608.1 17 7588 31.4
1637.7 64 239 11.4
8% 1112.5 61 491 8.1 758 607
1337.7 244 814 32.3
1375.9 11 335 1.5
1448.8 69 592 9.2
1594.6 301 118 39.7
1642.1 70 256 9.3
12% 1105.9 36 572 4.7 782 492
1357.8 322 295 41.2
1377.4 7725 1.0
1448.4 47 993 6.1
1598.6 273 677 35.0
1639.4 94 229 12.0
0 2 4 6 8 10 121000
1100
1200
1300
1400
1500
1600
1700
1800
Peak 1 Peak 2 Peak 3 Peak 4 Peak 5 Peak 6
Wav
elen
gth
(cm
-1)
Nitrogen proportion (%)
Fig. 4. The peak position for the investigated bands as a
function of nitrogen proportion in the working gas.
G. Lazar, I. Lazar / Journal of Non-Crystalline Solids 331 (2003) 70–78 75
proportion and must be related to the C@N andsp2 carbon bonds [12,23].
There is a great similarity between the data
variation for the next two peaks. The ratio be-
tween the peaks area, shown in Fig. 6, presents a
small variation. The initial increase of the ab-
sorption, followed by a decrease can be correlated
to the same variation of the 2900 cm�1 peak. These
facts offer evidence that the two peaks are relatedto the sp3 CHx bonds.
The last two peaks must be related to the C@Cand C@N bonds. The increase of the absorption in
this region indicates that the sp2 carbon propor-
tion in the film increases with the nitrogen content.
The evidence of this fact can be the decrease of the
optical band gap. Fig. 7 shows the variation of the
0 2 4 6 8 10 121x10 4
2x10 4
3x10 4
4x10 4
5x10 4
6x10 4
7x10 4
8x10 4
Peak 1 area
Peak
are
a (c
m-2)
Nitrogen proportion (%)0 2 4 6 8 10 12
0.0
5.0x104
1.0x105
1.5x105
2.0x105
2.5x105
3.0x105
3.5x105
Nitrogen proportion (%)
Peak
are
a (c
m-2)
Peak 2 area
0 2 4 6 8 10 124.0x103
5.0x103
6.0x103
7.0x103
8.0x103
9.0x103
1.0x104
1.1x104
1.2x104
1.3x104
Nitrogen proportion (%)
Peak
are
a (c
m-2)
Peak 3 area
0 2 4 6 8 10 122x104
3x104
4x104
5x104
6x104
7x104
Nitrogen proportion (%)
Peak
are
a (c
m-2)
Peak 4 area
0 2 4 6 8 10 120.0
5.0x104
1.0x105
1.5x105
2.0x105
2.5x105
3.0x105
3.5x105
4.0x105
Nitrogen proportion (%)
Peak
are
a (c
m-2)
Peak 5 area
0 2 4 6 8 10 120.0
2.0x104
4.0x104
6.0x104
8.0x104
1.0x105
Nitrogen proportion (%)
Peak
are
a (c
m-1)
Peak 6 area
4
8
12
16
Peak proportion (%)
Peak 1 proportion
20
25
30
35
40
45
50
55
60
Peak proportion (%)
Peak 2 proportion
0
1
2
3
4
5 Peak proportion (%
)
Peak 3 proportion
4
8
12
16
20
24
Peak proportion (%)
Peak 4 proportion
0
10
20
30
40
50
Peak proportion (%)
Peak 5 proportion
10
15
20
25
30
Peak proportion (%)
Peak 6 proportion
Fig. 5. The peak area and proportion for the investigated bands as a function of nitrogen proportion in the working gas. The peak
proportion is calculated making the ratio between the peak area and the total band area.
76 G. Lazar, I. Lazar / Journal of Non-Crystalline Solids 331 (2003) 70–78
0 2 4 6 8 10 120.10
0.15
0.20
0.25
0.30
Peak
are
a 3/
Peak
are
a 4
Nitrogen proportion (%)
Fig. 6. The ratio between the area of the peaks related to the
sp3 CHx bonds as a function of nitrogen proportion in the
working gas.
0 2 4 6 8 10 12
1.9
2.0
2.1
2.2
2.3
2.4
2.5
E g (eV
)
nitrogen (%)
Fig. 7. The variation of the Tauc gap as a function of the ni-
trogen proportion in the feed gas.
G. Lazar, I. Lazar / Journal of Non-Crystalline Solids 331 (2003) 70–78 77
Tauc gap as a function of the nitrogen proportionin the feed gas. The optical band gap decreases
from 2.5 to 1.9 eV with the growth of nitrogen flow
rate within the investigated range. Other groups
[10,11] have also reported a similar decrease in the
optical band gap of nitrogenated carbon.
5. Conclusions
Amorphous nitrogen-containing hydrogenated
carbon films are deposited from rf magnetron
sputtering of graphite target in Ar/CH4/N2 atmo-
sphere. In the IR absorption spectra, with the in-creasing of N2 proportion, there is an enlarging in
the intensity of absorption maxima at 3300–3400
and 1620 cm�1. For high nitrogen concentrations, a
newmaximum appears at 2150 cm�1 indicating that
theCandNatoms are chemically bonded in the film.
The IR spectroscopy is an easy and widely
available technique, but, because of the lack of
standards and reliable assignments, the interpre-tation of the absorption IR spectra for nitrogen
containing carbon films is not very clear yet. As one
small step in the attempt to render IR spectroscopy
appropriate for routine characterization of carbon
nitride films, the analysis of the broad absorption
band observed between 1700 and 1000 cm�1 has
been made. For polymer-like films investigated in
this paper, the deconvolution shows a six bandssystem, indicating the lack of the aromatic struc-
tures and an increase of the nitrogen and sp2 car-
bon proportion in the films with an increase of the
N2 partial pressure in the deposition gas.
Acknowledgements
The authors are grateful to Dr I. Vascan and to
Dr M. Caraman, Bacau University, for expert
technical assistance and for stimulating discussions
and comments.
References
[1] M. Lacerda, M. Lejeune, B.J. Jones, R.C. Barklie, R.
Bouzerar, K. Zellama, N.M.J. Conway, C. Godet, J. Non-
Cryst. Solids 299–302 (2002) 907.
[2] A.Y. Liu, M.L. Cohen, Science 245 (1989) 841.
[3] D.F. Franceschini, F.L. Freire Jr., S.R.P. Silva, Appl.
Phys. Lett. 68 (1996) 2645.
[4] K.J. Clay, S.P. Speakman, G.A.J. Amaratunga, S.R.P.
Silva, J. Appl. Phys. 79 (1996) 7227.
[5] K. Chakrabarti, M. Basu, S. Chaudhuri, A.K. Pal, H.
Hanzawa, Vacuum 53 (1999) 405.
[6] S.F. Yoon, Rusli, J. Ahn, Q. Zhang, C.Y. Yang, H. Yang,
F. Watt, Thin Solid Films 340 (1999) 62.
[7] N.P. Barradas, R.U.A. Khan, J.V. Anguita, S.R.P. Silva,
U. Kreissig, R. Grotzschel, W. Moller, Nucl. Instr. Meth.
Phys. Res. B 161–163 (2000) 969.
[8] T. Heitz, B. Drevillon, C. Godet, J.E. Bouree, Phys. Rev. B
58 (1998) 13957.
[9] J.K. Kaufman, S. Metin, D.D. Saperstein, Phys. Rev. B 39
(1989) 13053.
78 G. Lazar, I. Lazar / Journal of Non-Crystalline Solids 331 (2003) 70–78
[10] Y. Hayashi, G. Yu, M.M. Rahman, K.M. Krishna, T.
Soga, T. Jimbo, M. Umeno, J. Appl. Phys. 89 (2001) 7924.
[11] S.E. Rodil, A.C. Ferrari, J. Robertson, W.I. Milne,
J. Appl. Phys. 89 (2001) 5425.
[12] W. Xu, T. Fujimoto, B. Li, I. Kojima, Appl. Surf. Sci.
175&176 (2001) 456.
[13] M. Lejeune, O. Durand-Drouhin, K. Zellama, M. Benlah-
sen, Solid State Commun. 120 (2001) 337.
[14] Y.-S. Jin, T. Shibata, Y. Matsuda, H. Fujiyama, Thin Solid
Films 345 (1999) 18.
[15] A.G. Fitzgerald, L. Jiang, M.J. Rose, T.J. Dines, Appl.
Surf. Sci. 175&176 (2001) 525.
[16] G. Lazar, J. Phys.: Condens. Matter 13 (2001) 3011.
[17] J.V. Anguita, S.R.P. Silva, A.P. Burden, B.J. Sealy, S. Haq,
M. Hebbron, I. Sturland, A. Pritchard, J. Appl. Phys. 86
(1999) 6276.
[18] A.J.M. Buuron, M.C.M. van de Sanden, W.J. van Ooij,
R.M.A. Driessens, D.C. Schram, J. Appl. Phys. 78 (1996)
528.
[19] T. Szorenyi, C. Fuchs, E. Fogarassy, J. Hommet, F. Le
Normand, Surf. Coat. Technol. 125 (2000) 308.
[20] Y.H. Cheng, Z.H. Sun, B.K. Tay, S.P. Lau, X.L. Qiao,
J.G. Chen, Y.P. Wu, C.S. Xie, Y.Q. Wang, D.S. Xu, S.B.
Mo, Y.B. Sun, Appl. Surf. Sci. 182 (2001) 32.
[21] T. Ujvari, A. Toth, M. Mohai, J. Szepvolgyi, I. Bertoti,
Solid State Ionics 141&142 (2001) 63.
[22] M. Zhang, Y. Nakayama, S. Harada, J. Appl. Phys. 86
(1999) 4971.
[23] M. Ricci, M. Trinquecoste, F. Auguste, R. Canet,
P. Delhaes, C. Guimon, G. Pfister-Guillouzo, B. Nysten,
J.P. Issi, J. Mater. Res. 8 (1993) 480.
[24] X.-M. He, L. Shu, W.-Z. Li, H.-D. Li, J. Mater. Res. 12
(1997) 1595.
[25] A.K.M.S. Chowdhury, M. Monclus, D.C. Cameron,
J. Gilvarry, M.J. Murphy, N.P. Barradas, M.S.J. Hashmi,
Thin Solid Films 308&309 (1997) 130.
[26] M. Tabbal, P. Merel, S. Moisa, M. Chaker, E. Gat, A.
Ricard, M. Moisan, S. Gujrathi, Surf. Coat. Technol. 98
(1998) 1092.
[27] J.P. Zhao, Z.Y. Chen, T. Yano, T. Ooie, M. Yoneda,
J. Sakakibara, J. Appl. Phys. 89 (2001) 1634.
[28] O. Durand-Drouhin, M. Lejeune, M. Clin, D. Ballutaud,
M. Benlahsen, Solid State Commun. 118 (2001) 179.
[29] A.K.M.S. Chowdhury, D.C. Cameron, M.S.J. Hashmi,
Thin Solid Films 332 (1998) 62.
[30] N. Mutsukura, S. Inoue, Y. Machi, J. Appl. Phys. 72
(1992) 43.
[31] Y. Liu, C. Jiaa, H. Do, Surf. Coat. Technol. 115 (1999) 95.