-
A
Nr1Httiatriw©
K
1
mbcibpapis
s
0d
Materials Science and Engineering A 445–446 (2007) 223–236
Properties of titanium nitride films prepared by directcurrent
magnetron sputtering
Y.L. Jeyachandran a, Sa.K. Narayandass a,∗, D. Mangalaraj a,Sami
Areva b, J.A. Mielczarski c
a Department of Physics, Bharathiar University, Coimbatore 641
046, Tamil Nadu, Indiab Department of Physical Chemistry, Åbo
Akademi University, FIN-20500 Turku, Finland
c Laboratoire ‘Environnement et Mineralurgie,’ UMR 7569 CNRS,
INPL-ENSG, B.P. 40, 54501 Vandoeuvre-les-Nancy, France
Received 19 July 2006; received in revised form 28 August 2006;
accepted 11 September 2006
bstract
Titanium nitride (TiN) thin films of different thickness were
deposited by direct current (dc) magnetron sputtering under
conditions of various2 concentrations (0.5–34%). The electrical,
optical, structural, compositional and morphological properties of
the films were studied and the
esults were discussed with respect to N2 concentration and
thickness of the films. At low N2 concentration of 0.5% (of the
total sputtering pressure.1 Pa), golden coloured stoichiometric TiN
films were obtained and with increase in the N2 concentration
non-stoichiometric TiNx phases resulted.owever, irrespective of the
N2 concentration, the TiN stoichiometry in the films increased with
increase in the film thickness. In the surface of
he films the presence of nitride (TiN), oxynitride (TiOxNy) and
oxide (TiO2) phases were observed and the quantity of these phases
varied withhe N2 concentration and thickness. The films of lower
thickness were found to be amorphous and the crystallinity was
observed in the films withncrease in the thickness. The crystalline
films showed reflections corresponding to the (1 1 1), (2 0 0) and
(2 2 0) orientation of the cubic TiN andlso features associated
with TiNx phases. The transmission spectra of the films revealed
the typical characteristics of the TiN films i.e. a narrow
ransmission band, however, the width varied with thickness, in
the wavelength range of 300–600 nm and exhibited low transmission
in the infraredegion. The TiN films deposited at low N2
concentration of 0.5% showed smooth and uniform morphology with
densely packed crystallites. Withncrease in N2 concentration
various characteristics such as needle type crystallization, bubble
precipitates and after bubble burst morphologiesere observed in the
films. However, at higher N2 concentration conditions, uniformity
developed in the films with increase in thickness.2006 Elsevier
B.V. All rights reserved.
n con
id
[sftb
eywords: Titanium nitride films; dc magnetron sputtering;
Properties; Nitroge
. Introduction
The nitride compounds of titanium (TiNx) are the uniqueaterials
exhibiting both metallic (Ti–Ti) and covalent (Ti–N)
onding characteristics. The metallic properties are
electricalonductivity and metallic reflectance; and the covalent
bond-ng properties are high melting point, extreme hardness
andrittleness, and excellent thermal and chemical stability.
These
roperties of Ti and TiN have been frequently exploited
forpplications in microelectronic devices [1], solar cells [2] and
asrotective and decorative coatings [3]. Additionally, due to
thentrinsic biocompatibility and hemocompatibility, TiN has
beenuccessfully used as surface layers and electrical
interconnects
∗ Corresponding author. Tel.: +91 422 2425458; fax: +91 422
2422387.E-mail addresses: [email protected] (Y.L.
Jeyachandran),
[email protected] (Sa.K. Narayandass).
twu
bepce
921-5093/$ – see front matter © 2006 Elsevier B.V. All rights
reserved.oi:10.1016/j.msea.2006.09.021
centration; Thickness
n orthopaedic prostheses, cardiac valves and other
biomedicalevices [4,5].
Various methods have been employed for TiN deposition6–10].
Among them sputtering techniques (dc and rf) are con-idered as most
suitable methods and are being extensively usedor TiN deposition.
The importance of sputtering methods is thathey involve a number of
parameters such as nitrogen pressure,ase pressure, sputtering
pressure, cathode power and substrate-arget separation in addition
to substrate bias and temperaturehereby a number of combination of
these parameters may besed to obtain high quality films with
required properties.
However, a non-linear relationship was found to existetween the
reactive gas pressure and other processing param-
ters exhibiting hysteresis effects thereby restraining the
finalroperties of the films. Under low N2 concentrations
goldenoloured stoichiometric TiN films were obtained [11–13],
how-ver, controlling low N2 concentration was found to be
difficult.
mailto:[email protected]:[email protected]/10.1016/j.msea.2006.09.021
-
2 e and Engineering A 445–446 (2007) 223–236
AtptHi[
raiosudgt
sooNAtrswcto
2
2
smKcmfaeptbabatppp(da
Table 1Conditions employed for the preparation of TiN films
Parameters Values
(A) Deposition period from 10 to 15 minCathode power (CPw) 100
WSputtering pressure (SPr) 1.1 PaBase pressure (BPr) 4 × 10−4
PaSubstrate–target distance 100 mm
N2 concentration
0.5%3%6%11%17%22%27%34%
Thickness (nm)
0.5% 27% 34%
(B) Deposition period up to 30 min23 88 9542 142 13482 160
112
outtn1afi3
nidawf
2
aoirmts
24 Y.L. Jeyachandran et al. / Materials Scienc
t higher N2 concentrations, nitrogen oversaturation occurred
inhe films sometimes leading to unusual effects such as
nitrogenrecipitation at grain boundaries or as gas bubbles,
superstruc-ure formation and partial crystallisation were observed
[14–16].ence, due to the criticalness of nitrogen partial pressure,
much
nterested has been paid to investigate its effect on TiN
films17,18].
Furthermore, by sputtering methods, obtaining stiochiomet-ic TiN
films were found possibly only under substrate biasingnd/or at
higher temperature conditions [13]. However, the goaln
microelectronics and industrial applications determines tworiginal
constraints: (i) the substrate could not be heated to
avoidtructural or compositional changes; (ii) the substrates must
bender low or unbiased to avoid contact damages in electronicevices
and also for coil-coating process. Therefore, there is areat
interest to attain stoichiometric TiN films at low tempera-ures and
without substrate biasing.
In the present work, by direct current (dc) planar
magnetronputtering method nearly stoichiometric TiN films were
obtainedn unbiased substrates at room temperature under conditionsf
low N2 concentration. Together the effect of thickness and2
concentration on the properties of the films was studied.nalytical
techniques such as resistance measurements, optical
ransmission spectroscopy, spectroscopic ellipsometry (SE), X-ay
photoelectron spectroscopy (XPS), X-ray diffraction (XRD),canning
electron microscopy (SEM) and optical microscopyere used to
characterise the prepared films. The electrical, opti-
al, structural, compositional and morphological properties ofhe
films were analysed with a particular emphasize on the effectf N2
concentration and film thickness.
. Experimental details
.1. Preparation of TiN films
TiN films were sputter deposited onto cleaned p-type (1 0
0)ilicon and glass substrates at room temperature from a
titaniumetal target (75 mm diameter × 5 mm thick, 99.995% pure,
PI-EM, England) mechanically clamped to the dc magnetron
athode of a conventional sputtering system (Vacuum Instru-ents
Company, India). The silicon substrates were cleaned by
ollowing RCA (Radio Corporation of America) procedure [19]nd the
glass substrates were cleaned by the method describedlsewhere [20].
Commercial argon (Ar2) and N2 gases (99%ure) were used as the
sputtering and reactive gases respec-ively. The gas flow into the
preparation chamber was controlledy two stage needle valves. Prior
to Ti film deposition vacuumnd target conditioning were performed.
The deposition cham-er was pumped down to the ultimate vacuum (4 ×
10−4 Pa)nd repeatedly charged with Ar2 and pumped down in ordero
minimize the residual gas components. The Ti target wasre-sputtered
at a sputtering pressure (SPr) of 2 Pa and cathodeower (CPw) of 125
W to sputter out the surface oxide layer. The
re-sputtering was done until the Ti characteristic plasma
glowdark blue colour) appeared. The TiN films were deposited
underifferent N2 concentrations (0.5–34%) of the total SPr (1.1
Pa)nd of different thicknesses by fixing all the other parameters
at
aocP
123153
ptimum values (details specified in Table 1). The optimum val-es
of the fixed parameters were derived from our experience onitanium
film deposition [20]. The effect of N2 concentration onhe
properties of the films was studied by keeping the film thick-ess
constant (∼60 nm) via adjusting the deposition period from0 to 15
min (section A in Table 1). Additionally, at low (0.5%)nd higher
(27 and 34%) N2 concentrations, higher thicknesslms were prepared
by extending the deposition period up to0 min (section B in Table
1).
The thicknesses of the films were measured using MBI tech-ique
[21] and cross checked by SE measurements. The variationn the
thickness values of the films as evaluated using the twoifferent
techniques was found to be ±8 nm. In Table 1, the aver-ge
thicknesses of films were presented. All the deposited filmsere
found to be uniform over an area of 36 mm2 as observed
rom the sheet resistance measurements.
.2. Characterisation experiment details
All the characterisation experiments were made ex situ andt room
temperature. For SE measurements the films depositednto silicon
substrates were used and for all the other character-sations the
films prepared on glass substrates were used. Sheetesistance and
resistance versus temperature (R–T) measure-ents for the TiN films
were made using a four point probe sys-
em (Scientific Equipment, Roorkee, India). A constant
currentource and a digital microvoltmeter were used to apply
current
nd measure the voltage respectively and a PID controlledven was
used as the temperature source. The crystallographicharacteristics
of the films were studied by XRD (PHILIPSW 3040 X-ray
diffractometer) using Cu K�1 (λ = 0.1542 nm)
-
e and
rfoU
wcTu(rsTtt(ofi
fetwwmwo
lwi
bcttr
3
3
ocRaftvoi
Fc
Y.L. Jeyachandran et al. / Materials Scienc
adiation operated at 30 kV and 30 mA. The scan was per-ormed in
continuous mode for a 2θ range of 10–80◦. Theptical transmittance
spectra of the films were recorded from aV–vis–NIR
spectrophotometer (VARIAN, Cary 500).For the TiN films of different
thicknesses (23–153 nm) that
ere deposited at an N2 concentration of 0.5%, the chemi-al
composition was obtained by XPS (KRATOS, XSAM800).he measurements
were performed at a vacuum of 10−6 Pasing a Mg K� X-ray source (λ =
1253.6 eV). The UNIFITTUVersion 2.1) [22,23] software was used for
peak fitting the high-esolution scans and for quantitative chemical
analysis, applyingensitivity factors given by the manufacturer of
the instrument.he high-resolution spectra were charge compensated
by set-
ing the binding energy (BE) of the C (1s) contamination peako
284.4 eV. The morphology of the films was studied by SEMSiemens,
UK). The images were taken at an accelerating voltagef 5 kV. The
optical constants spectra of the films were obtainedrom the SOPRA
GESP5 spectroscopic ellipsometer operatingn a spectral range
250–850 nm with variable angle of incidence.
For those films deposited at the other N2 concentrations,
theollowing characterisation tools were used. The XPS systemmployed
was Perkin-Elmer PHI 5400 ESCA System Spec-rometer using Mg X-ray
source. The SEM instrument usedas Philips XL20 electron microscope.
The SEM micrographs
ere taken at acceleration voltages of 10 and 15 kV. The
opticalicroscopic images of the films were obtained from Leica
(Q-in) optical microscope. The SE measurements of the films
werebtained from a VASE spectroscopic ellipsometer (J. A. Wool-
twi
ig. 1. Sheet resistance, resistivity and temperature coefficient
of resistance values aoncentration of 0.5%.
Engineering A 445–446 (2007) 223–236 225
am Inc.). The angle of incidence and the polarization azimuthere
set at 70◦ and 135◦, respectively. The data were measured
n the wavelength range of 300–1100 nm.The use of different
instruments (XPS, SEM and SE) was
ecause the characterisation of TiN films prepared at 0.5%
N2oncentration and those prepared at all the other N2
concen-rations was performed at two different laboratories.
However,he basic accuracy of the instruments was comparable and
theesults obtained were within the instrumental error.
. Results
.1. Resistance studies
Fig. 1 shows the room temperature sheet resistance (R) valuesf
the TiN film of different thicknesses deposited at the N2
con-entration of 0.5%. Also shown in the figure are the behaviour
ofof the films with temperature (303–373 K), and the
resistivity
nd TCR values of the films. The R value of the films
decreasedrom 85 to 15 �/sq with the increase in thickness. Films of
allhicknesses followed metal type R–T behaviour, that is, the
Ralues increased with increase in temperature. The resistivityf the
films was within ±20 of 210 �� cm and the TCR valuencreased from
0.002 to 0.05%/K with thickness.
The R values of the films deposited at higher N2 concen-rations
(3–34%), in general, increased from 7.2 to 87.5 k�/sqith increase
in N2 concentration and decreased with increase
n thickness. A low R value of 133 �/sq was obtained for the
nd resistance–temperature plots of films of different thickness
prepared at N2
-
226 Y.L. Jeyachandran et al. / Materials Science and Engineering
A 445–446 (2007) 223–236
Fig. 2. Resistance–temperature plot of the film of thickness 160
nm prepared atavr
fia1r(Rpt
3
Ne(rs
Fot
N2 concentration of 34%. Along the y-axis are the normalised
sheet resistancealues. R is the sheet resistance value at any
temperature (K) and R0 is the sheetesistance value at room
temperature (303 K).
lm of thickness 160 nm deposited at N2 concentration of 34%nd a
value of 260 �/sq was observed in the films of thickness42 and 134
nm prepared at N2 concentration of 27 and 34%espectively. All the
films prepared at higher N2 concentration3–34%) showed
semiconductor type R–T behaviour, that is, the
value decreased with increase in temperature. A typical R–Tlot
of the film of thickness 160 nm deposited at N2 concentra-ion of
34% is shown in Fig. 2.
.2. Optical transmission studies
The optical transmission spectra of the films prepared at2
concentration of 0.5% are shown in Fig. 3. The films
xhibited a transmission band in the visible wavelength
region
300–1000 nm) and a low transmission in the higher
wavelengthegion. The transmission percentage and width of the
transmis-ion band increased, and the transmittance peak shifted
towards
ig. 3. Optical transmission spectra of the films prepared at N2
concentrationf 0.5%. The inset shows the shift in the peak
transmittance wavelength withhickness.
Fco
ho
hfivitn
ig. 4. Optical transmission spectra of the films (60 nm)
prepared at N2 con-entration of 3–34% and higher thickness films
prepared at N2 concentrationsf 27 and 34%.
igher wavelength (376–475 nm) with decrease in the thicknessf
the films.
The optical transmission spectra of the films deposited atigher
N2 concentrations are shown in Fig. 4. The spectra of thelms of
thickness 60 nm exhibited a transmission band in the
isible wavelength region (300–700 nm) and the
transmissionncreased in the higher wavelength region. With an
increase inhe N2 concentration, the transmittance of the films and
broad-ess of the visible transmission band increased. At a high
N2
-
Y.L. Jeyachandran et al. / Materials Science and Engineering A
445–446 (2007) 223–236 227
F us thp
cltfitbw
3
(bpnomo3ttw5oifit[
3
The XPS survey spectra of normal unetched and 15 nm
etchedsurfaces of the films of thickness 42–153 nm prepared at the
N2concentration of 0.5% exhibited the characteristic Ti 2p, O
1s
Table 2Refractive index and extinction coefficient values of the
TiN films at 632.8 nmas derived from and spectroscopic ellipsometry
and optical transmissionmeasurements
N2 concentration (%) Thickness (nm) Optical constants 632.8
nm
n k
0.5
82 1.33 2.24112 1.34 2.07123 1.32 2.26153 1.42 2.16
3
60
1.83 1.176 1.92 1.2611 1.82 1.0717 1.77 1.0722 1.81 1.1427 1.92
1.0234 1.83 0.92
27a88 1.62 1.08
ig. 5. Refractive index and extinction coefficient spectra of
the films of variorepared at N2 concentrations of 3–34%.
oncentration of 34% the transmission band was completelyost and
the films recorded high transmittance with an absorp-ion band edge
at 350 nm. With increase in the thickness of thelms prepared at N2
concentrations of 27 and 34%, the transmit-
ance decreased and the narrowness of the visible transmissionand
increased with relatively low transmission in the higheravelength
region.
.3. Spectroscopic ellipsometry studies
The spectra of the optical constants such as refractive indexn)
and extinction coefficient (k) of the TiN films as obtainedy SE are
shown in Fig. 5. The films of different thicknessesrepared at the
N2 concentration of 0.5% exhibited similar
and k spectral behaviour. A maximum in the n patternsf the films
appeared at wavelength ∼300 nm and a mini-um at the wavelength
range of 500–600 nm. The k plots
f the films showed a minimum in the wavelength range of00–400 nm
with a steep increase with wavelength. Likewise,he films of
thickness 60 nm prepared at various N2 concen-rations (3–34%)
showed similar n and k spectral behaviourith a minimum in the n and
k plots at wavelength ranges of00–530 nm and 372–450 nm,
respectively. The n and k values
f the films at 632.8 nm as obtained from the Fig. 5 are
presentedn Table 2. Also given in the table are the n and k values
of thelms of higher thicknesses deposited at higher N2
concentra-
ions 27–34% as evaluated from the optical transmittance
data24].
3
icknesses prepared at N2 concentration of 0.5% and films of
thickness 60 nm
.4. X-ray photoelectron spectroscopy studies
142 1.46 1.2
4a95 1.78 0.89
134 1.69 0.89160 1.43 0.67
a Values derived from optical transmission data.
-
228 Y.L. Jeyachandran et al. / Materials Science and Engineering
A 445–446 (2007) 223–236
F face of the film of thickness 82 nm prepared under 0.5% N2
concentration condition.
a4wTpc1geptieTts
fpfc4p3r
Table 3Assignments to the component peaks of the high resolution
XPS scans of thefilms prepared at 0.5% N2 concentration
condition
Peak Component Binding energy (eV) Assignment
Ti 2p3/2 I 458.1 TiO2II 456.1 TiOxNyIII 455.6 TiN
O 1s IV 529.8 TiO2V 531.6 O–H [Ti(OH)2 or H2O]
N 1s VI 396.1 TiNVII 398.1 O–N [TiO N ]
TT
icpoawo
ig. 6. XPS survey spectra of the (a) unetched and (b) 15 nm
sputter etched sur
nd N 1s peaks at the corresponding binding energies 528.2,56.5
and 396.2 eV, respectively [25,26]. The surface of the filmsas
etched in the XPS chamber by sputtering using argon gas.he typical
survey spectra of the film of thickness 82 nm areresented in Fig.
6. The C 1s peaks in the spectra may be theontributions from
organic carbon. The observed sodium (Nas) in the spectrum of the
unetched surface may be from thelass substrate whereby Na atoms are
highly mobile and so canasily diffuse to the surface of the films.
Additionally, the Ar 2peak identified in the spectra of the etched
surface may be fromhe adsorbed argon during etching or argon
species incorporatednto the films during growth [27]. The elemental
Ti/N ratios asvaluated from the spectra of the films are plotted in
Fig. 7. Thei/N ratio decreased on surface etching. The Ti/N ratio
was in
he range of 1.12–1.25 in the normal surface and 1.1–1.22 in
theputter etched surface of the films.
From high-resolution XPS measurements of the normal sur-ace of
the films, the spin orbit doublet Ti 2p1/2 and Ti 2p3/2eaks at
binding energies 462.5 and 458.1 eV, respectively wasound in the Ti
2p spectra. The Ti 2p3/2 peaks included threeomponents whose peaks
centered at 458.1 (I), 456.1 (II) and55.6 eV (III). Both the O 1s
and N 1s peaks showed two com-
onents resolution centered at 529.8 eV (IV), 531. 6 eV (V)
and96.1 eV (VI), 398.1 eV (VII), respectively. The typical
high-esolution spectra of the film of thickness 82 nm are shown
Fig. 7. Ti/N ratio in the films prepared at N2 concentration of
0.5%.
gas
naeiT
TReN
F(
11
x y
iO2: titanium dioxide; TiOxNy: titanium oxynitride; TiN:
titanium nitride;i(OH)2: titanium hydroxide.
n Fig. 8. The most probable assignments to the origin of
theomponents are presented in Table 3 [28–31] and the
relativeercentage of the components as evaluated from the high
res-lution spectra are presented in Table 4. The components IIInd
VI of the Ti 2p3/2 and N 1s peaks respectively associatedith the
TiN phases increased with the increase in thicknessf the films. The
TiOxNy component (II) of the Ti 2p3/2 peakenerally, however a
higher percentage in 82 nm film, remainedlmost constant and the OH−
component (V) of the O 1s peakhowed a decreasing trend.
The survey spectra of the normal surface of the films of
thick-ess 60 nm deposited under different N2 concentrations
(3–34%)
nd films of higher thickness prepared at 22 and 34%
conditionsxhibited the characteristic Ti 2p, O 1s and N 1s peaks
sim-lar to that of the films of 0.5% N2 concentration condition.he
high resolution Ti 2p peaks of the films displayed the spin
able 4elative percentage of the components of Ti 2p3/2, O 1s and
N 1s peaks asvaluated from the high resolution XPS spectra of the
films prepared at 0.5%
2 concentration condition
ilm thicknessnm)
Relative percentage of the components
Ti 2p3/2 peak O 1s peak N 1s peak
I II III IV V VI VII
42 67.8 20.6 11.6 51.2 48.8 52.6 47.482 57.1 28.7 14.2 56.5 43.5
65.6 34.423 63.9 21.7 14.4 58.2 41.8 69.4 30.653 61.0 22.3 16.7
67.5 32.5 81.8 18.2
-
Y.L. Jeyachandran et al. / Materials Science and Engineering A
445–446 (2007) 223–236 229
Fig. 8. XPS high resolution spectra of the film of thickness 82
nm prepared at a N2 concentration of 0.5%.
F N2 co3
otTOfirasn
tba
surement was carried out for the 523 K annealed film. The
XPSsurvey and high resolution scan spectra of the 523 K
annealedfilm are shown in Fig. 10. The characteristic Ti 2p, O 1s
andN 1s peaks in the survey spectra and the doublet peaks in
the
Table 5Assignments to the component peaks of the N 1s high
resolution spectra of thefilms deposited under N2 concentrations of
3–34%
Peak Component Binding energy (eV) Assignment
VI 395.6 TiN
ig. 9. High resolution XPS scans of the films of thickness 60 nm
prepared at4% condition.
rbit splitting characteristics and the Ti 2p3/2 peaks
exhibitedhe components resolution corresponding to TiO2, TiOxNy
andiN species. The O 1s peaks exhibited the typical TiO2 andH−
components peak; and the N 1s peaks displayed three to
our (only for 3% condition) component resolutions at the bind-ng
energies 395.6, 396.8, 399.1 and 400.2 eV. The typical
highesolution scans of the films of thickness 60 nm prepared at 3nd
17% conditions, and 95 nm thick film of 34% condition arehown in
Fig. 9. The most probable assignments to the compo-ents of the N 1s
peaks are presented in the Table 5 [32–34].
The film prepared under 22% N2 concentration conditionshat
showed bubble precipitate morphology (results presentedelow in
Section 3.6) was subjected to annealing in argontmosphere (10 Pa)
for 60 min at 523 and 723 K. XPS mea-
N
T
ncentrations of 3 and 17% and that of the film of thickness 95
nm prepared at
1sVII 396.8 O–N [TiOxNy]VIII 399.1 N–HIX 400.2 Adsorbed N2
iOxNy: titanium oxynitride; TiN: titanium nitride.
-
230 Y.L. Jeyachandran et al. / Materials Science and Engineering
A 445–446 (2007) 223–236
cknes
hetasnmN
ta
poptaTN
TRs
N
112223
Fig. 10. XPS survey spectra and high resolution scans of a film
of thi
igh resolution Ti 2p peaks were observed. The Ti 2p3/2
peakxhibited showed only one resolved component correspondingo TiO2
phase and the O 1s peak showed two componentsssigned to TiO2 and
OH− species. The N 1s peaks did nothowed a definite peak, however,
two crests observed in theoisy background at the binding energies
399.4 and 402.2 eVay be assigned to the contributions from N–H and
adsorbed
2 species, respectively [32–34].The relative percentage of the
components as evaluated from
he high resolution spectra of the films of 3–34% conditionsre
given in Table 6. Also given in the table are the relative
cctc
able 6elative percentage of the components and elements on the
surface of the films prepa
pectra
2 concentration (%) Relative % of the components
Ti 2p3/2 peak O 1s peak
I II III IV V
3a 88.4 11.6 – 77.8 22.26 72.0 17.9 10.1 67.2 32.81 69.2 18.5
12.3 52.9 47.17 70.1 21.5 8.4 60.1 39.92 72.2 18.3 9.5 56.2 43.87
76.5 17.8 5.7 52.8 47.22b – – – 62.2 37.84c 57.2 23.1 19.7 51.8
48.2
a For this film no TiN component in the Ti 2p3/2 peak was
observed and in N 1s peb Data for the film annealed at 523 K.c Data
for the film of thickness 95 nm while the other data are for the
films of thick
s 60 nm prepared at N2 concentration of 22% and annealed at 523
K.
ercentage of the elements Ti, O and N present in the surfacef
the films. In general, the TiN component (III) of the Ti 2p3/2eak
decreased with increase in N2 concentration, however,he TiN
component increased with increase in the thicknesss for the film
prepared at a N2 concentration of 34%. TheiOxNy component (II) was
almost constant with change in2 concentration. In the O 1s peak the
percentage of TiO2
omponent (V) decreased and in the N 1s peak the O–Nomponent
(VII) increased and the component (VIII) assignedo the adsorbed N2
species decreased with increase in N2oncentration. The elemental Ti
and N elemental percentage
red at N2 concentrations of 3–34% as evaluated from the high
resolution XPS
at.% of elements
N 1s peak Ti O N
VI VII VIII
25.8 24.2 35.0 22.2 70 7.846.6 28.3 25.1 24.8 57.9 17.344.0 32.0
24.0 22.8 58.5 18.741.1 36.1 22.8 21.5 61.9 16.646.0 36.5 17.5 21.6
63.3 15.143.2 34.8 22.0 22.3 61.6 16.1
– – – 20.3 76.6 3.143.5 34.5 22.0 22.2 63.8 14.0
ak four components were observed.
ness 60 nm.
-
Y.L. Jeyachandran et al. / Materials Science and Engineering A
445–446 (2007) 223–236 231
F % an
vg
3
catt
2o∼tot
TDm
P
0
11222
2
3
ig. 11. X-ray diffractograms of the films prepared at a N2
concentration of 0.5
aried from 21.5 to 24.5 and 7.8 to 18.5, respectively
withenerally a decreasing trend with increase in N2
concentration.
.5. X-ray diffraction studies
The XRD patterns of the TiN films prepared at 0.5% N2
oncentration and the films of higher thickness prepared at 27nd
34% conditions are shown in Fig. 11. For 0.5% condi-ion, an
amorphous like XRD pattern with a broad peak inhe 2θ range 15–39◦
was obtained for the film of thickness
[tos
able 7imensions of the morphological features of the TiN films
of different thicknessicrographs
reparation conditiona Thickness (nm) Feature dimension
Particle size (nm)
.5%
23 1642 23
123 38153 34
1% –7% 602%2% (523 K) –2% (723 K) 60
7%88 32
142 85
4%95 25
160 60
a N2 concentration conditions.
d the films of higher thickness prepared at N2 concentrations of
27 and 34%.
3 nm. With increase in thickness of the films the intensityf the
broad peak decreased and diffraction peaks at 2θ ∼ 22◦,27.3◦ and
∼36.7◦ resolved in the amorphous background pat-
ern. The peaks observed at 2θ ∼ 22◦ and ∼27.3◦ in the patternsf
the films of thickness 42–112 nm may be due to the contribu-ions
from the crystallites with substoichiometric phases (TiNx)
35,36]. The peak at 2θ ∼ 36.7◦ observed for the films of
higherhickness (123 and 153 nm) could be associated with the (1 1
1)rientation of stoichiometric TiN crystal particles with
cubictructure [37].
prepared at various N2 concentration conditions as obtained from
the SEM
Needle crystals Bubble diameter
Length (nm) Width (nm)
– – –
120–800 25–62 –100–560 20–70 780–1000 nm
– – 300–2200 nm– – 69–550 �m
9–125 �m
– – –
– – –
-
2 e and
6pataeeodNspmtt[
3
dTatw
(tmt
32 Y.L. Jeyachandran et al. / Materials Scienc
For higher N2 concentrations (3–34%), the films of thickness0 nm
exhibited an amorphous diffraction pattern with a broadeak in the
2θ range of 15–39◦. In the Fig. 11, the broad peak alsoppeared in
the films of higher thickness prepared at N2 concen-rations of 27
and 34%. Additionally, the higher thickness (88nd 142 nm) films
prepared at 27% N2 concentration conditionxhibited diffraction
peaks at 2θ ∼ 22◦, ∼24◦ and ∼43◦. How-ver, at N2 concentration of
34%, an amorphous pattern wasbserved for the films of thickness up
to 134 nm and the 43◦iffraction peak was observed for the film of
thickness 160 nm.evertheless, for the films of thickness 95 and 134
nm small
houlders at 2θ ∼ 43◦ and 62◦ were observed. The lower angleeaks
may be associated with the particles of the substoichio-
etric phases [35,36]. The 43◦ diffraction peak was assigned
o the contribution from cubic TiN particles with (2 0 0)
orienta-ion and the 62◦ resolution was assigned to the (2 2 0)
orientation37].
cgin
Fig. 12. SEM micrographs of the film prepared
Engineering A 445–446 (2007) 223–236
.6. Microscopic studies
From the SEM micrographs, the films prepared at 0.5% con-ition
were found to possess uniform and void free morphology.he
micrographs showed better particle resolution. The aver-ge particle
size evaluated from the micrographs of the films ofhickness 42, 82,
123 and 153 nm as presented in the Table 7as found to be in the
range of 18–42 nm.Films of thickness 60 nm prepared at higher N2
concentration
3–34%) showed varying characteristics such as smooth, needleype
crystallization, bubble precipitation and bubble burst-like
orphologies. The low (3%) and high (34%) N2 concentra-ion
conditions produced films with smooth morphology. At 6%
ondition cluster like features was observed in the smooth
back-round morphous. The needle type crystallites were observedn
the films for 11 and 17% conditions and the number of theeedle
crystallites decreased with the increase in the N2 con-
at different N2 concentration conditions.
-
Y.L. Jeyachandran et al. / Materials Science and
Fp
cfwcmnbdmfibot9i1d
plw
4
f
fic
tnwvci
3rtblr(atctTaalfse
d[fsfibrw
otthkpbt(olfi
ti
ig. 13. Optical microscope images of the annealed films of
thickness 60 nmrepared at a N2 concentration of 22%.
entration. Large bubble precipitates in the films were
observedor 17 and 22% conditions and the number of bubbles
increasedith increase in the N2 concentration. The individual
needle
rystallites and the bubbles were spaced by relatively
smoothorphology. The length (L) and maximum width (Wmax) of the
eedle crystallites and the horizontal diameter (Dh) of the
gasubbles observed in the films are presented in Table 7. At a
con-ition of higher N2 concentration of 27% rough and
non-uniformorphology resulted in the films. However, the higher
thicknesslms prepared at N2 concentrations 27 and 34% were found
toe relatively smooth and uniform. The average particle size
asbtained from the high magnification (100× k) of the films
ofhickness 88 and 142 nm prepared at 27% N2 concentration and5 and
160 nm prepared at 34% N2 concentration are presentedn Table 7. The
typical SEM images of the film of thickness53 nm of 0.5%
conditions, 60 nm of 11, 17, 22 and 27% con-itions and 95 nm of 34%
condition are shown in Fig. 12.
The optical micrographs obtained for the 523 and 723 K
filmsrepared at 22% N2 concentration are shown in Fig. 13. Veryarge
bubbles were observed and the bubble diameter decreasedith increase
in annealing temperature.
. Discussion
In this section, for the purpose of clarity, the results
obtainedor the films prepared at 0.5% N2 concentration are
discussed
thaT
Engineering A 445–446 (2007) 223–236 233
rst and then the results of the films deposited at higher
N2oncentrations are dealt with.
From Fig. 1, the decrease in R value with thickness may behe
thickness effect; however, a slight deviation in the chemicalature
of the films, still retaining the metallic characteristics,as also
evident from the other results. The increase in TCRalue with
thickness reveals the existence of the chemical naturehanges in the
films and improvement in metallic character withncrease in
thickness of the films.
A narrow transmission window in the wavelength region00–600 nm
and low transmission towards higher wavelengthegion are the typical
characteristics of TiN films [38]. Fromhe transmission spectra of
the films (Fig. 3) the transmissionand for the film of thickness
153 nm was observed in the wave-ength range of 300–500 nm and the
band width wavelengthange increased together with a red shift in
the peak transmissionTmax) wavelength with decrease in thickness.
These featureslso show the existence of change in chemical
composition ofhe films with thickness with a possibility of
stoichiometric TiNomposition at higher thicknesses (>82 nm)
[2,38]. Generally,he optical features of TiN arise from the free
carriers in Tid band.herefore, an increase in the number of Ti
vacancies results indecrease in the density of free carriers and
shifts in the Tmax,
ssociated with the screened plasma frequency, to higher
wave-ength thereby creating the red-shifted spectra [38,39].
Alsorom the spectra, the observed increase in transmittance
inten-ity with decrease in thickness of the films may be the
thicknessffect [38].
The optical quality of the films of 0.5% N2 concentration
con-ition was found superior when compared to the earlier
report2,38]. For example, the peak transmission percentage
obtainedor a film of thickness 82 nm was 12.7% (Fig. 3), however,
Duru-oy et al. [38] obtained a transmission percentage of 6.5% for
thelms of thickness 90 nm prepared under a condition of
substrateias 160 V and temperature 573 K. The superiority of the
presentesults is in the fact that the films were prepared at
conditionsithout substrate bias and temperature.In the optical
constants spectra (Fig. 5) of the films, the
bserved maximum and minimum in the n and k plots, respec-ively
at the wavelength range of 300–400 nm may be relatedo E2-type
transitions associated with the transitions betweenybridized Np–Tid
bands [40–42]. The gradual increase in thevalues towards higher
wavelength and the minimum in the natterns may be associated with
the Drude–Lorentz-type (inter-and) transitions [40]. The interband
transitions may correspondo those between the Γ v15 valence and
Γ
c12 conduction band
Γ v15 − Γ c12, E1 gap) [39]. The values of n and k in the rangef
1.32–1.42 and 2.07–2.26, respectively as obtained at a wave-ength
of 632.8 nm (Table 2) compare well with values for TiNlms reported
elsewhere [43].
From Fig. 7, the Ti/N ratio decreased with increase inhickness
of the films. The decrease in Ti/N ratio shows themprovement in TiN
stoichiometry of the films as evidenced in
he above studies. In coherence the TiN component in both theigh
resolution Ti 2p3/2 peak and N 1s peak increased (Table 4)nd the
TiO2 component (I) in the Ti 2p3/2 peak decreased.he observed spin
orbit doublet in the Ti 2p peak (Ti 2p3/2 and
-
2 e and
Trc(wc
paipescf
mTdtst
(fico2ti(wmis
bhitiettsdmfi(ifitt[tth
ut
(ttaamacitcifsc
switiocatitCNoT
owooisc(tmb2stsNps
34 Y.L. Jeyachandran et al. / Materials Scienc
i 2p1/2) suggests that oxidation was the predominant
surfaceeaction in all the films [26]. Correspondingly a strong
TiO2omponent at 529.8 eV was observed in O 1s peak of the filmsFig.
8). The II (456.1 eV) component in the Ti 2p3/2 peakas assigned to
TiOxNy due to the observation of the O–N
omponent at 398.1 eV in the N 1s peak of these films [26].In the
XRD plots of the films (Fig. 11), the observed broad
eak features in the 2θ range of 15–39◦ and the diffraction
peakst the 2θ ∼ 22◦ and ∼27.3◦ may be attributed to the oxidationn
the films and subsequent formation of substoichiometric TiNhases.
The decrease in the intensity of the broad peak and thevolution of
a peak at 2θ ∼ 36.7◦ corresponding to TiN phasehow the improvement
in the crystallinity and also chemical stoi-hiometry in the films
with increase in thickness as substantiatedrom the above
discussion.
The morphological studies showed the presence of
nanoscaleicrostructural features such as particle size (16–38 nm,
seeable 7) in the films. The increase in particle size reveals
theevelopment of better crystallinity in the films with increase
inhickness. The particles would have been formed by clusters oftill
smaller sized crystallites, however, the present magnifica-ion of
the images revealed little about these features.
In the films that were prepared at higher N2
concentrations3–34%), the observed high R values (7.2–87.5 k�/sq)
in thelms of thickness 60 nm may be due to a large deviation in
theirhemical composition from TiN stoichiometry. The decreasef R
value with increase in thickness of the films prepared at7 and 34%
N2 concentration conditions may be attributed tohe thickness
effect; however, significant improvement in sto-chiometry with
thickness was evidenced from other studiesdiscussed below). The R–T
plots of the films of all thicknessesere of semiconductor type, as
observed in Fig. 2, in contrast toetal type conductivity of the TiN
films. This shows that, even
f the higher thickness films are supposed to have improved
TiNtoichiometry they still hold semiconducting nature.
From the optical transmission results (Fig. 4) the observedroad
transmission band in the visible wavelength region andigher
transmission towards longer wavelengths together withncrease of
these features with N2 concentration reveal the exis-ence of
non-stoichiometry (TiNx) in the films and also decreasen nitride
stoichiometry with increase in N2 concentration. Thevolution of
narrowness of transmission band and decrease inransmission towards
higher wavelengths in the films of higherhicknesses prepared at 27
and 34% N2 concentration conditionshows the increase in nitride
stoichiometry with thickness. Theecrease of transmittance intensity
with increase in thicknessay be the thickness effect [2,38]. The n
and k spectra of thelms of thickness 60 nm prepared at different N2
concentrationFig. 5) show the presence of E1 and E2 type transition
character-stics. The increased n and decreased k values obtained
for theselms at a wavelength of 632.8 nm (Table 2) when compared
with
hat values obtained for the films prepared at 0.5% N2
concen-ration condition may be attributed to the compositional
changes
43]. The composition of the film inherently affects the
dielec-ric constant of the film thereby causing an associated
change inhe optical constants [44]. The values presented for the
films ofigher thickness prepared at 27 and 34% N2 concentrations
are
eice
Engineering A 445–446 (2007) 223–236
ncertain because they were derived from the transmission datahat
do not account the reflectance and absorption losses.
The XPS survey spectra and high resolution Ti 2p and O 1sFig. 9)
peaks of most of the films exhibited similar characteris-ics as
observed for the other films (TiN films, 0.5% condition)hat have
been already discussed. However, in the N 1s peaks andditional
component (VIII) at 399.1 eV was observed that wasssigned to the
contribution from N–H species. The N–H bondsight have been formed
through the activity of the surface N
toms with the adsorbed water molecules (H2O). In support,
aonsiderable amount of O–H component (V) could be observedn the O
1s peaks of the films (Table 6). The H2O adsorption in thehese
films may be high, than in films prepared at 0.5% N2 con-entration,
due to the existence of chemical non-stoichiometryn the films. The
non-stoichiometric compositions would haveavoured surface charges
(Ti3+ defect states) thereby mediatingignificant H2O adsorption and
causing the evolution of N–Homponent [45,46].
The film of thickness 60 nm prepared at 3% N2 concentrationhowed
distinct XPS spectral characteristics (Fig. 9) comparedith the
films prepared at other N2 concentrations. No signif-
cant TiN component in the Ti 2p3/2 peak was observed andhe N 1s
peak showed four components. The low nitridationn the films may be
due to high surface oxidation. The sourcef oxidation may be the
nitrogen gas. As in the present caseommercial nitrogen gas was used
and it may have a consider-ble percentage of oxygen impurities
[47]. Since Ti have lowhreshold towards oxidation the available
percentage of oxygenn the nitrogen gas together with the background
residual quan-ity would have mediated the high surface oxidation of
the films.orrespondingly the component IX at 400.2 eV evolved in
the1s peak that was assigned, in literature, to adsorbed
nitrogen
riginating from the release of nitrogen during oxidation of
theiN component in the film [32].
At N2 concentrations higher than 3%, nitridation wasbserved in
the films, however, the nitride percentage decreasedith increase in
N2 concentration as partially evidenced fromptical transmission
studies. This could be due to the effect ofxidation, as just
mentioned, that may have sourced from thencreasing impurity
component from the reactive gas. Corre-pondingly the increase in
TiO2 component and decrease of TiNomponent of the Ti 2p3/2 with
increase of N2 concentration>3%) could be observed from Table 6.
At these N2 concentra-ion conditions (6–27%) the unreacted N2
present in the films
ay get buried in the films and sometime be precipitated
asubbles. The bubble precipitates in the films prepared at 17 and2%
N2 concentrations could be observed from the SEM imageshown in Fig.
12. As the case, the XPS spectra (Fig. 10) ofhe annealed film
prepared at 22% N2 concentration conditionhowed no indication for
the nitride component in both Ti 2p and
1s peaks. At the same time, the N 1s spectra revealed a
com-onent at 402.2 eV that may be due to poorly screened
nitrogentates [32]. At higher temperature the small nitride
component
xisted in the film may have also been oxidized. However,
withncrease in thickness of the films prepared at higher N2
con-entration (34%), as observed from Fig. 9 and Table 6 and
alsovidenced from the optical transmission studies, the nitride
com-
-
e and
pm
ficstindiawc(
(taasptT2aactimdmrrrl
5
nttisScqdpinmhcc
pi
wdprmaotta
A
fsd
R
[
[
[
[
[
[[
[[[
[
Y.L. Jeyachandran et al. / Materials Scienc
onent increased. The increased nitridation at higher
thicknessesay be due to the increased dissolution of nitrogen in
the films.In the XRD studies, the amorphous nature observed in
the
lms of thickness 60 nm prepared at different N2
concentrationonditions may be due to oxidation. Additionally, the
glass sub-trates (amorphous character) and the room temperature
prepara-ion condition may also be the reason for the amorphous
naturen the films. The XRD patterns of the films of higher
thick-esses prepared at 27 and 34% conditions (Fig. 11) showed
theevelopment of nitride phases in the films with thickness, whichs
in consistent with the optical and XPS results as discussedbove.
However, a significant contribution from TiNx phasesas also
observed, which may be responsible for the semi-
onducting nature in the films as obtained in the R–T studiesFig.
2).
From the SEM images of the films of thickness 60 nmFig. 12), the
mechanism of formation and composition of needleype of crystallites
in the films prepared at 11 and 17% conditionsre not clear at the
present stage. Different microstructures suchs selective
crystallization in an amorphous matrix and super-tructure formation
in TiNx films were reported [15,16]. Theresent result of needle
type crystallization is one of such dis-inct microstructure of TiNx
films and effect of N2 concentration.he bubble structures observed
in the films prepared at 17 and2% conditions may be due to the
precipitation of unreacted N2nd also Ar incorporated during growth
[14]. The non-uniformnd rough morphology observed in the films
deposited at N2oncentration of 27% resemble something like bubble
burst fea-ures. On annealing the film of 22% condition at 523 K,
thencrease of bubble dimension (Fig. 13) may be due to the ther-al
expansion of the gas precipitates and the decrease of bubble
imension on annealing at further higher temperature (723 K)ay be
due to the burst of the large bubbles to smaller ones as a
esult of over expansion. Bubble precipitates in TiN films
wereeported in the literature; however, the uniqueness of the
presentesult is that the dimension of the observed bubbles was
veryarge when compared to the reported values of 5–10 nm [14].
. Conclusions
TiN films of different thickness were prepared by dc mag-etron
sputtering method under various N2 concentration condi-ions. The
effect of thickness and N2 concentration (0.5–34%) onhe electrical,
optical, compositional, structural and morpholog-cal properties of
the films were studied by using resistance mea-urements, optical
transmission spectroscopy, SE, XPS, XRD,EM and optical microscope
techniques. The metal type electri-al properties, characteristic
TiN optical transmission, structuraluality and chemical
stoichiometry of the films improved withecrease in N2 concentration
and increase in thickness. The filmsrepared at low (0.5–3%) and
high (34%) N2 concentrationsndependent of the thickness and those
films of higher thick-esses independent of the N2 concentration
exhibited uniform
orphology with better microstructural properties. On the
other
and, the films of thickness 60 nm that were prepared at N2
con-entrations of 6–27% exhibited various morphologies such
asluster formation, needle type selective crystallization,
bubble
[
[
Engineering A 445–446 (2007) 223–236 235
recipitation and non-uniform morphologies respectively
withncrease in N2 concentration.
As a concluding remark, in the present work TiN films with aide
range of chemical stoichiometry starting from highly oxi-ized
composition through TiNx phase to stoichiometric TiNhase and with
different morphologies were prepared and theesults were discussed.
The possibility of obtaining stoichio-etric TiN films with good
optical quality at room temperature
nd substrate unbiased condition has been demonstrated. Basedn
this result together with the results of the effect of N2
concen-ration and thickness further investigations could be carried
ono obtain good quality TiN films both optically and structurallyt
conditions of low temperature and bias.
cknowledgement
We sincerely thank Dr. Tudor Jenkins, University of Wales,or his
kind help in performing spectroscopic ellipsometry mea-urements for
the samples of 0.5% condition and for the valuableiscussion on the
results.
eferences
[1] M. Tao, D. Udeshi, S. Agarwal, E. Maldonado, W.P. Kirk,
Solid StateElectron. 48 (2004) 335.
[2] G.B. Smith, A. Ben-David, P.D. Swift, Renew. Energ. 22
(2001) 79.[3] F. Vaz, P. Cerqueira, L. Rebouta, S.M.C. Nascimento,
E. Alves, Ph.
Goudeau, J.P. Riviere, K. Pischow, J. de Rijk, Thin Solid Films
447–448(2004) 449.
[4] K.H. Chung, G.T. Liu, J.G. Duh, J.H. Wang, Surf. Coat.
Technol. 188–189(2004) 745.
[5] W. Franks, I. Schenker, P. Schmutz, A. Hierlemann, IEEE
Trans. Biomed.Eng. 52 (2005) 1295.
[6] N. Mustapha, R.P. Howson, Surf. Coat. Technol. 92 (1997)
29.[7] K. Lal, A.K. Meikap, S.K. Chattopadhyay, S.K. Chatterjee, M.
Ghosh, K.
Baba, R. Hatada, Physica B 307 (2001) 150.[8] L. Fouilland, L.
Imhoff, A. Bouteville, S. Benayoun, J.C. Remy, J. Perrière,
M. Morcrette, Surf. Coat. Technol. 100–101 (1998) 146.[9] V.
Craciun, D. Craciun, C. Ghica, L. Trupina, C. Flueraru, N.
Nastase,
Appl. Surf. Sci. 138–139 (1999) 593.10] F. Richter, H. Kupfer,
H. Giegengack, G. Schaarschmidt, F. Schoize, F.
Elstner, G. Hecht, Surf. Coat. Technol. 54–55 (1992) 338.11]
H.A. Jehn, in: Y. Pauleau (Ed.), Materials and Processes for
Surface and
Interface Engineering, Kluwer Academic, Dordrecht, 1995, p.
359.12] G. Reiners, H. Hantsche, H.A. Jehn, U. Kopacz, A. Rack,
Surf. Coat. Tech-
nol. 54–55 (1992) 273.13] Ph. Roquiny, F. Bodart, G. Terwagne,
Surf. Coat. Technol. 116–119 (1999)
278.14] L. Hultman, J.-E. Sundgren, L.C. Markert, J.E. Greene,
J. Vac. Sci. Technol.
A 7 (1989) 1187.15] A. Tarniowy, R. Mania, M. Rekas, Thin Solid
Films 311 (1997) 93.16] S. Zerkout, S. Achour, A. Mosser, N. Tabet,
Thin Solid Films 441 (2003)
135.17] R. Banerjee, R. Chandra, P. Ayyub, Thin Solid Films 405
(2002) 64.18] L.J. Meng, M.P. dos Santos, Surf. Coat. Technol. 90
(1997) 64.19] W. Kern, Handbook of Semiconductor Wafer Cleaning
Technology, Noyes
Publications, New Jersey, 1993, 20.20] Y.L. Jeyachandran, B.
Karunagaran, Sa.K. Narayandass, D. Mangalaraj,
T.E. Jenkins, P.J. Martin, Mater. Sci. Eng. A 431 (2006) 277.21]
S. Tolansky, Multiple Beam Interferometry of Surfaces and Films,
Oxford
University Press, London, 1948, 147.22] R. Hesse, T. Chasse, P.
Streubel, R. Szargan, Surf. Interface Anal. 36 (2004)
1373.
-
2 e and
[[
[[[
[[[[
[
[
[[
[
[
[[[[[
[
[
[
36 Y.L. Jeyachandran et al. / Materials Scienc
23] R. Hesse, T. Chasse, R. Szargan, Anal. Bioanal. Chem. 375
(2003) 856.24] J.C. Manifacier, J. Gasiot, J.P. Fillard, J. Phys. E
Sci. Instrum. 9 (1976)
1002.25] F.-H. Lu, H.-Y. Chen, Thin Solid Films 355–356 (1999)
374.26] J. Zhao, E.G. Garza, K. Lam, C.M. Jones, Appl. Surf. Sci.
158 (2000) 246.27] A. Pruymboom, P. Berghuis, P.H. Kes, H.W.
Zandbergen, Appl. Phys. Lett.
50 (1987) 1645.28] N.C. Saha, H.G. Tompkins, J. Appl. Phys. 72
(1992) 3072.29] P. Prieto, R.E. Kirby, J. Vac. Sci. Technol. A 13
(1995) 2819.30] K.S. Robinson, P.M.A. Sherwood, Surf. Interface
Anal. 6 (1984) 261.31] I. Bertoti, M. Mohai, J.L. Sullivan, S.O.
Saied, Appl. Surf. Sci. 84 (1995)
357.32] H.Z. Wu, T.C. Chou, A. Mishra, D.R. Anderson, J.K.
Lampert, S.C.
Gujrathi, Thin Solid Films 191 (1990) 55.33] F. Esaka, K.
Furuya, H. Shimada, M. Matsubayashi, H. Sato, A. Nishijima,
A. Kawana, H. Ichimura, T. Kikuchi, J. Vac. Sci. Technol. A 15
(1997)2521.
34] N.D. Shinn, K.L. Tsang, J. Vac. Sci. Technol. A 9 (1991)
1558.35] B. Granier, R. Renard, J. Coutures, Rev. Int. Hautes Temp.
Refract. 17
(1980) 235, JCPDS file no. 44-0951.
[
[
Engineering A 445–446 (2007) 223–236
36] Holmberg, Acta. Chem. Scand. 16 (1962) 1255; (JCPDS file no.
17-0386);JCPDS file no. 23-1455.
37] W. Wong-Ng, H. McMurdie, B. Paretzkin, C. Hubbard, A.
Dragoo, NBS(USA), ICDD Grant-Aid, 1987 (JCPDS file no.
38-1420).
38] H.Z. Durusoy, O. Duyar, A. Aydinli, F. Ay, Vacuum 70 (2003)
21.39] V. Ern, A.C. Switendick, Phys. Rev. 137 (1965) A927.40] S.
Adachi, M. Takahashi, J. Appl. Phys. 87 (2000) 1264.41] B.
Karlsson, Thin Solid Films 87 (1982) 181.42] B. Karlsson, R.P.
Shimshock, B.O. Seraphin, J.C. Haygarth, Sol. Energy
Mater. 7 (1983) 401.43] H.G. Tompkins, R. Gregory, B. Boeck,
Surf. Interface Anal. 17 (1991)
22.44] B. Karunagaran, S.J. Chung, E.-K. Suh, D. Mangalaraj,
Physica B 369
(2005) 129.45] F. Guillemot, M.C. Porte, C. Labrugere, Ch.
Baquey, J. Colloid Interface
Sci. 255 (2002) 75.46] R. Schaub, P. Thostrup, N. Lopez, E.
Lægsgaard, I. Stensgaard, J.K.
Nørskov, F. Besenbacher, Phys. Rev. Lett. 87 (2001)
266104–266111.47] K. Deenamma Vargheese, G. Mohan Rao, T.V.
Balasubramanian, Sanjiv
Kumar, Mater. Sci. Eng. B 83 (2001) 242.
Properties of titanium nitride films prepared by direct current
magnetron sputteringIntroductionExperimental detailsPreparation of
TiN filmsCharacterisation experiment details
ResultsResistance studiesOptical transmission
studiesSpectroscopic ellipsometry studiesX-ray photoelectron
spectroscopy studiesX-ray diffraction studiesMicroscopic
studies
DiscussionConclusionsAcknowledgementReferences