Optical properties of silicon oxynitride thin films
determined by vacuum ultraviolet
spectroscopic ellipsometry
Hyun long Kirn,1 Yong Jai Cho?2 Hyun Mo Cho52* Sang Youl Kirn,1
Changsun Moon,3 Gyungsu Cho,3 and Youngmin Kwon3
2
Department of Molecular Science and Technology, Ajou University, Suwon, Kyunggi-do 442-749, South Korea
Division of Optical Metrology, Korea Research Institute of Standards and Science, Daejeon 305-600, South Korea
3
Anam Semiconductor, Inc. Buchon, Kyunggi-do 420-130, South Korea
Abstract We determined the optical constants of silicon oxynitride (SiOxNy) thin films using a vacuum ultraviolet
spectroscopic ellipsometer. The SiOxNy layers with a nominal thickness of 25 nm - 35 nm were grown on silicon
substrates by using plasma enhanced chemical vapor deposition (PECVD), The precursor gases were nitrous oxide
(N2O) and silane (SiH4). The ratio r of N2O flow rate to SiH one in the deposition process was controlled from 1,16 to
3,05. The ellipsometric measurements were performed at the angle of incidence 75° for the spectral range from 0.75 eV
to 8,75 eV, The complex refractive indices, optical band gaps, and thicknesses of the SiOxNy layers were determined by
using Tauc-Lorentz dispersion model.
phase difference of ARC layer. Silicon oxynitride
(SiOxNy) is one of promising inorganic antireflective
candidates since it has a low deposition temperature.
SiOxNy films have been utilized in the production of
electronics and opto-electronic devices since they have
many advantages like high dielectric constant, low
leakage current, high breakdown voltage, low
dielectric/semiconductor interface state density, low
impurity diffusion, high radiation hardness, and good
mechanical, thermal and chemical stability.
INTRODUCTION
The development of ultra large scale integrated
(ULSI) semiconductor device has led the processing
technology to a smaller linewidth than 0.1 um. It is
impossible to use the previous process technology of
the semiconductor device. It has demanded the shorter
wavelength of lithography light for reducing transistor
dimension. Because it happens an undesirable
reflection on the substrate, an antireflection coating
(ARC) layer is considered for a solution to overcome
these problems. ARC layer is expected to reduce the
standing wave effect of the laser beam that causes the
interference of reflected light, as well as to decrease
the reflection from topography of fabricated circuit.
So far, several workers have studied SiOxNy thin
films to use for ARC layer in photolithography process.
Most of their studies were performed on samples of
SiOxNy thin films grown by plasma enhanced chemical
vapor deposition (PECVD). In the deposition process,
the precursor gases were nitrous oxide (N2O) and
silane (SiH4). These films were thicker than 100 nm.
They reported the results on the optical, structural,
mechanical, physical and chemical characteristics by
using various methods for the controlled flow rate of
N2O and SiH4 gases.[1"8] Especially, the refractive
indices were determined by ellipsometry at laser
The ARC layer is deposited on Si substrate to avoid
the reflection of light on the interface surface between
photoresist and Si substrate. An ideal ARC layer have
the same amplitude and the phase difference of 180°
between the reflected light on the substrate/ARC layer
interface and that of the ARC layer/photoresist
interface. The refractive index and the thickness of
ARC layer is important for the amplitude and the
* Correspond author: hmcho@kriss,re.kr
CP683, Characterization and Metrology for ULSI Technology: 2003 International Conference,
edited by D. G. Seiler, A. C. Diebold, T. J. Shaffner, R. McDonald, S. Zollner, R. P. Khosla, and E. M. Secula
© 2003 American Institute of Physics 0-7354-0152-7/03/$20.00
171
wavelength/1"51 visible range16"71 and deep ultraviolet
range(- 6.5 eV),[8]
The Eqs. (2) and (3) are uniquely defined by the
following fitting parameters: A is proportional to the
transition probability matrix element, C the broadening
term, EQ the peak transition energy, Eg the optical band
gap, and S^ an integration constant,
In this work, we prepared a set of SiOxNy films
deposited by PECVD with different ratios of N2O flow
rate to that of SiH4 . The ratio r of N2O flow rate to
SiH4 one was controlled from 1.16 to 3,05. Their
nominal thicknesses are 25 nm - 35 nm. The refractive
indices and thicknesses of the SiOxNy films have been
determined by using a vacuum ultraviolet
spectroscopic ellipsometer (VUV SE) for the spectral
range from 0.75 eV to 8,75 eV (142 nm - 1653 nm).
The variations of the complex refractive indices and
optical band gaps of the SiOxNy films were
demonstrated by changing the gas flow ratio.
TABLE 1. Deposition conditions for silicon oxynitride
thin films examined in this study.
m
ID
A-l
A-2
A-3
A-4
A-5
A-6
A-7
THEORY
Dep. Time
(sec)
7.2
9,6
9.6
9.6
9.6
9.6
7.2
SiH4
(seem)
41
46
49
52
55
58
75
N20
(seem)
125
112
112
112
112
112
87
Ratio r
(N2O/ SiH4)
3.05
2.43
2,29
2.15
2.04
1.93
1.16
EXPERIMENTS
The theory of ellipsometric analysis is based on the
Fresnel reflection equations for polarized light that
comes from Maxwell equations. The ellipsometry is
expressed with terms of the ellipsometric parameters,
Psi OP) and Delta (A):[9]
The SiOxNy thin films were deposited on p-type,
(100) surface, 200 mm diameter, and crystalline
silicon substrate using the PECVD method. All SiOxNy
films were grown at the temperature of 350 °C and the
RF power of 120 W. The base pressure of the chamber
was kept at 5.5 Torr during film growth. The flow rate
of He gas was 2000 seem. The ratio r of N2O flow rate
to SiH4 one was adjusted from 1.16 to 3.05. The
detailed deposition conditions are shown in Table 1.
(1)
The ellipsometric measurements of SiOxNy thin
films were performed by using a vacuum ultraviolet
spectroscopic ellipsometer (VUV SE: J.A.Woollam
Co.), which adopts the nitrogen purging system to
avoid the absorption by oxygen and moisture. The
angle of incidence was 75° and the spectral range of
photon energy was 0,75 eV ~ 8.75 eV with a step of
0.05 eV,
where rp and rs are the complex Fresnel reflection
coefficients of the light whose electric fields are
parallel and perpendicular to the plane of incidence,
respectively.
We adopt the Tauc-Lorentz (TL) dispersion
function*lo] to characterize the complex dielectric
functions (G = B\ + m2) of the SiOxNy films. The TL
model is defined by the imaginary part s2 of the
complex dielectric function;
RESULTS
2
2 2
2 2
(E -~E } +C E
E '
We have used a simple three-phase optical model
(air/SiOxNy film/c-Si) to analyze the measured VUV
SE data. The regression analysis was carried out using
the TL dispersion model to determine the optical
constant of SiOxNy film. The best-fitted parameters of
the TL model are compiled in Table 2. Fig. 1 shows
the spectra of the refractive indices and extinction
coefficients of SiOxNy films. It is clear that a crossover region of the refractive index exists at ~ 6.3 eV,
As the ratio r increases, the refractive index decreases
for lower photon energies than ~ 6,3 eV, but it
(2)
= 0, (E<E\
The real part 8] of the dielectric function is obtained by
the Kramers-Kronig dispersion relation:
(3)
172
TABLE 2. The best fitted parameters of silicon oxynitride thin films determined by using the Tauc-Lorentz
dispersion model. A, C, E0, and Eg have units of eV, S^ is dimensionless. The fitting residual 0 is the goodness-offit parameter.
Thickness
ID
A
C
a
Eo
Eg
^oo
(nm)
A-l
0.77
53.1
10.5
11,9
2.81
0.0083
26,58
33.92
11.1
0.0071
0.58
84.3
18.9
2.69
A-2
0.0060
A-3
33.90
94.5
20.8
10.5
2,65
0.56
A-4
20,6
9.38
2.60
0.0055
33,70
0.70
92.6
8.41
0.0052
33.59
0.82
92.2
20.6
A-5
2.58
0.0052
34,60
20.6
7.81
2.57
A-6
0.87
93,8
128.4
10
0.0090
24.03
1.18
4.66
2,37
A-7
2.6
1.2
2.4
1.0
2.2
0.8
2.0
0.6
1.8
0.4
1.6
0.2
1.4
0.0
2
4
6
10
8
2
4
6
10
Photon Energy (eV)
Photon Energy (eV)
FIGURE 1. The refractive indices (a) and extinction coefficients (b) of silicon oxynitride films determined by the Tauc-Lorentz
model.
2.4
Boseetal.
Alay o et al.
Uenoetal.
Present Work
2.0
1.4
X=633 nm
0
1
2
3
4
5
8
7
3
Ratio of Gas Flow rate (N2O/SiH4)
FIGURE 2. The refractive indices of the silicon oxynitride
films at 633 nm.
173
I
o
o-
i.
o*
S11
If
g l g c S s
3
^ ^
O
3* 3 3
t£> P (TQ
<1 t—t
<r+
a. Q
Jf2"
3
CD
ca
ll
I-K- X x
T3 3
ft PT
11 ffi i. 5- B- »
ofSi|ti
" <" o
3 5- g. a
gJ a> -a o
t-K
gji £•„ a* «> ot g
«
CD
sT SL £r fl> o CD o 2
££ 2 :s a- o a
11 § i i a H<g 8
3 S 3- g "3 S- «>
•
4^ O*
00 P
°
3 ^«
X
fi3 ft5
r* p
fO P
r
3
^|£
i G? .
21" i
I.
CD
3
O
O-
p
ja ^
CO
g-s
I?
I a,
f
e. a
3 I
£<cr *
ro
I
s,1
S
o
03 in
I?
1 S
|
o
I
o
o
o
o
-A
-*
Extinction Coefficient, k
o
REFERENCE
this work are different from the result of Gaillard et at.
They are explained the abnormal decreasing of the
extinction coefficients from FTIR and AES results, as r
approach near zero, no pure silicon is deposited: the
layer is a mixture of nitrogen and silicon. It is because
nitrogen is used as a carrier gas, and the rf power used
in the reactor is sufficient to decompose a small
amount of nitrogen.
In Fig, 4, appears the complex refractive index
variation of SiOxNy films for the representative laser
wavelengths such as 633 nm, 248 nm, 193 nm, and 157
nm. The refractive index variation shows different
slopes for each wavelength with r increase. At shorter
wavelength (157 nm) it is positive and at longer
wavelength (248 nm and 633 nm) it is negative. At 193
nm, it changes its sign from negative to positive. It is
related with the intersectional region shown in Fig. 1,
From the parameters in Table 2, we draw the relation
between the optical band gap and the ratio of gas flow
rate r as shown in Fig. 5. The optical band gap
increases quite linearly when r increases.
From these results, we can determine the optimal
condition of ARC layer at shorter wavelength for the
same amplitude and the phase difference of 180°. For
example, if the refractive index of ARC layer is 1.9 at
ArF lithography process (193 nm), the thickness of
ARC layer is near 25 nm.
1.
K. Ueno, T. Kikkawa and K, Tokashiki, J, Vac. Sci.
Thechnoi B 13,1447 (1995),
2.
E. A. Joseph, C. Gross. H. -Y, Liu, R. T. Laaksonen and
F. G. Celii, J. Vac. Sci. Thechnoi. B 19,2483 (2001).
3.
M. Bose, D. N. Bose and D. K. Basa, Materials Letters
52,417(2002).
4.
M. I. Alayo, I. Pereyra, W. L. Scopel and M. C. A.
Fantini, Thin Solid Films 402, 154 (2002).
5.
F. Gaillard, P. Schiavone and P. Brault, J. Vac. Sci.
Thechnoi. B 15, 2777 (1997).
6.
M. Modreanu, N. Tomozeiu, P. Cosmin, Mariuca
Gartner, Thin Solid Films 337, 82 (1999).
7.
H. G. Tompkins, R. B. Gregory, P. W. Deal, S. M Smith,
J. Vac. Sci. Thechnoi. A 17, 391 (1999).
8.
C. Defranoux, J. P. Piel, and J. L. Stehle, Thin Solid
Films 313-314, 742 (1998).
9.
R. M. A. Azzam and N. M. Bashara, Ellipsometry and
Polarized Light (North-Holland, Amsterdam, 1987).
10. G, E, Jellison Jr. and F. A. Modine, Appl. Phys. Lett. 69,
371 (1996); 69, 2137 (1996).
CONCLUSION
The effect of the source gas flow rate variation of
SiOxNy films deposited by PECVD was investigated
by a vacuum ultraviolet spectroscopic ellipsometer in
the spectral range from 0.75 eV to 8.75 eV (142 nm 1653 nm). The flow rate ratio (r=N2O/SiH4) was varied
for SiOxNy film depositions and films with different
optical constants were obtained. The optical constants
of SiOxNy films were determined from the measured
ellipsometric data by using the Tauc-Lorentz
dispersion function. The change of r in the deposition
processes caused significant changes in the optical
properties of SiOxNy films and provided it a promising
role as ARC layers in semiconductor devices. Using
vacuum ultraviolet spectroscopic ellipsometer, the
unknown optical constants of material can determine at
vacuum ultraviolet range.
175