Determination of electron escape depth in ultrathin silicon oxide

Article abstract of DOI:10.1063/1.1868066

Using the high-brilliance synchrotron radiation at SPring-8, we determined the electron escape depths in approximately 1-nm -thick low-temperature oxide layers, which were formed on Si(100) at 300 °C using three kinds of atomic oxygen and that in approxim

Full text of DOI:10.1063/1.1868066

Determination of electron escape depth in ultrathin silicon oxide
H. Nohira, H. Okamoto, K. Azuma, Y. Nakata, E. Ikenaga, K. Kobayashi, Y. Takata, S. Shin, and T. Hattori
Citation: Applied Physics Letters 86, 081911 (2005); doi: 10.1063/1.1868066
View online: http://dx.doi.org/10.1063/1.1868066
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/86/8?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Band offsets of nitrided ultrathin hafnium silicate films
Appl. Phys. Lett. 88, 162906 (2006); 10.1063/1.2196235
Roads to ultrathin silicon oxides
J. Vac. Sci. Technol. A 23, 201 (2005); 10.1116/1.1842113
Suppression of subcutaneous oxidation during the deposition of amorphous lanthanum aluminate on silicon
Appl. Phys. Lett. 84, 4629 (2004); 10.1063/1.1759065
Hydrogen redistribution in thin silicon dioxide films under electron injection in high fields
J. Appl. Phys. 82, 5573 (1997); 10.1063/1.366417
The role of extra-atomic relaxation in determining Si 2p binding energy shifts at silicon/silicon oxide interfaces
J. Appl. Phys. 82, 2298 (1997); 10.1063/1.366037
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 131.155.81.2
On: Sun, 23 Nov 2014 02:57:49

APPLIED PHYSICS LETTERS 86, 081911 ͑2005͒
Determination of electron escape depth in ultrathin silicon oxide
H. Nohiraa͒ and H. Okamoto
Department of Electrical & Electronic Engineering, Musashi Institute of Technology, 1-28-1 Tamazutsumi,
Setagaya-ku, Tokyo 158-8557, Japan
K. Azuma and Y. Nakata
Advanced LCD Technologies Development Center Co. Ltd., 292 Yoshida-cho, Totsuka-ku,
Yokohama, 244-0817, Japan
E. Ikenaga and K. Kobayashi
JASRI/SPring-8, Mikaduki-cho, Sayo-gun, Hyogo 679-5198, Japan
Y. Takata and S. Shin
RIKEN/SPring-8, Mikaduki-cho, Sayo-gun, Hyogo 679-5148, Japan
T. Hattori
Research Center for Silicon Nano-Science, Advanced Research Laboratories,
Musashi Institute of Technology, 8-15-1 Todoroki, Setagaya-ku, Tokyo 158-0082, Japan
͑Received 23 September 2004; accepted 28 December 2004; published online 16 February 2005͒
Using the high-brilliance synchrotron radiation at SPring-8, we determined the electron escape
depths in approximately 1-nm-thick low-temperature oxide layers, which were formed on Si͑100͒ at
300 °C using three kinds of atomic oxygen and that in approximately 1-nm-thick thermally grown
oxide layer formed in 1 Torr dry oxygen at 900 °C by measuring angle-resolved Si 2p
photoelectron spectra at the photon energy of 1050 eV. The results indicated that the electron escape
depths in the three kinds of low-temperature oxide layers were 18%–24% smaller than that in the
thermally grown oxide layer. Furthermore, the electron escape depth in the thermally grown oxide
layer, whose thickness was close to that of the structural transition layer, was 7% smaller than that
in bulk SiO2. © 2005 American Institute of Physics. ͓DOI: 10.1063/1.1868066͔
The electron escape depth in silicon oxide layer and that
in silicon used for the determination of silicon oxide thick-
ness by x-ray photoelectron spectroscopy have been deter-
mined by measuring the number of x-ray excited photoelec-
trons arising from silicon oxide layer and that arising from
silicon as a function of thickness of silicon oxide layer.1,2
However, such method cannot be applied to the oxide layer
with the thickness on the order of 1 nm, which mostly deter-
mines the quality of a SiO2/Si interfacial transition layer.
One of our goals is to establish a low-temperature oxidation
process for the gate insulator of thin-film transistors on glass
or plastic substrates. Radical oxidation is one of the most
effective ways to conduct an oxidation reaction at low tem-
perature. We evaluated three approaches for effectively gen-
erating oxygen radicals. Photo-oxidation uses a xenon exci-
mer lamp producing 172 nm wavelength light, which has a
photon energy sufficient to selectively obtain O͑1D͒ radicals
and lower than the SiO2 band gap in order to suppress defect
production at the SiO2/Si interface. Plasma-enhanced oxida-
tion, which is further enhanced by adding a rare gas such as
Kr, is also effective for obtaining high-density oxygen radi-
cals. In the present letter the method of determining the elec-
tron escape depths in ultrathin silicon oxide layers by mea-
suring angle-resolved Si 2p photoelectron spectra is
proposed and the electron escape depths in low-temperature
oxides formed at 300 °C using three kinds of atomic oxygen
are found to be clearly different from that in thermally grown
oxide layer formed at 900 °C using molecular oxygen.
On a vicinal Si͑100͒ 0.01° surface, a 0.72-nm-thick ox-
ide layer was formed at 300 °C by using atomic oxygen
produced in a krypton-mixed oxygen ͑Kr:O2=97:3͒ plasma
excited by microwave radiation ͑2.45 GHz͒; this layer is re-
ferred to as Kr/O2 plasma oxide hereafter.3 On the same kind
of surface, a 0.86-nm-thick oxide layer was formed at
300 °C by using atomic oxygen produced in an oxygen
plasma excited by microwave radiation ͑2.45 GHz͒; this
layer is referred to as O2 plasma oxide hereafter.3 Finally, on
a
third sample with the same kind of surface, a
0.85-nm-thick oxide layer, referred to as photo-oxide hereaf-
ter, was formed at 300 °C by using atomic oxygen produced
by exciting molecular oxygen with a xenon excimer lamp.
As a reference, a high-quality 1.25-nm-thick silicon oxide
layer, referred to as thermal oxide hereafter, was formed in
dry oxygen at 900 °C on another vicinal Si͑100͒ 0.01° sur-
face covered with a 0.3-nm-thick pre-oxide layer formed in
dry oxygen at 300 °C.4 The thicknesses of SiO2 layers were
evaluated from the angle-resolved Si 2p photoelectron spec-
tra discussed and the thicknesses of compositional transition
layers were evaluated from the measurement of Si 2p photo-
electron spectra at photoelectron take-off angle of 55°, where
the effect of elastic scattering can be effectively neglected.5
Using electron energy analyzer ESCA-200 angle-resolved
1050 eV photons’ excited Si 2p photoelectron spectra were
measured with energy resolution of 100 meV at soft-x ray
undulator beam line ͑BL27SU͒ of the Super Photon Ring
8 GeV ͑SPring-8͒.
Figure 1 shows 1050 eV photons’ excited Si 2p3/2 pho-
toelectron spectra arising from Kr/O2 plasma oxide, O2
plasma oxide, photo-oxide and thermal oxide with photo-
a͒
Electronic mail: nohira@ee.musashi-tech.ac.jp
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 131.155.81.2
0003-6951/2005/86͑8͒/081911/3/$22.50 86, 081911-1 © 2005 American Institute of Physics
On: Sun, 23 Nov 2014 02:57:49

081911-2
Nohira et al.
Appl. Phys. Lett. 86, 081911 ͑2005͒
FIG. 1. 1050 eV photons’ excited Si
2p3/2 photoelectron spectra with pho-
toelectron take-off angle ␪ as a param-
eter measured for ͑a͒ the Kr/O2
plasma oxide, ͑b͒ the O2 plasma oxide,
͑c͒ the photo-oxide, and ͑d͒ the ther-
mal oxide, which were formed on
Si͑100͒ surfaces.
electron take-off angle ␪ as a parameter. Here, the refraction
of the electron wave at vacuum/SiO2 interface was taken
into account.6 With decreasing ␪, the Si 2p photoelectron
spectral intensities arising from the Si substrate decreased.
This confirmed that the Si substrates were, in fact, covered
with silicon oxide layers. Furthermore, the Si 2p3/2 spectral
intensity, NI, arising from the compositional transition layer
consisting of intermediate oxidation states of Si measured at
photoelectron take-off angle of 20° is indicated by the
shaded region in each graph. The amounts of NI are 0.759
ML for Kr/O2 plasma oxide, 0.865 ML for O2 plasma oxide,
0.881 ML for photo-oxide, and 0.835 ML for thermal oxide,
respectively. Here, 1 ML expresses 6.80ϫ1014 cm−2.
function of the electron escape depth ⌳O in the SiO2 layer by
the following equation. Here, it should be noted that the
effect of photoelectron diffraction on the dependence of
NI/NO on ␪ does not have to be taken into account, because
the oxide layers are in amorphous states,
t
n ⌳ sin ␪ 1 − exp −
ͩ ͪ
t
t
ͭ ͮ
NI
⌳t sin ␪
=
.
͑1͒
NO
d
n ⌳ sin ␪ exp
− 1
ͩ ͪ
o
o
ͭ ͮ
⌳o sin ␪
Here, nO and nt, d and t, and ⌳t denote the densities of
silicon atoms in the SiO2 layer and the compositional transi-
tion layer, the thicknesses of the SiO2 layer and the compo-
sitional transition layer, and the electron escape depth in the
compositional transition layer, respectively. The following
relation between ⌳O and d should be satisfied.
The dependencies of NI/NO on ␪ measured for three
kinds of low-temperature oxide layers formed on Si͑100͒ are
compared with that measured for thermal oxide in Fig. 2.
Here, NO represents the Si 2p spectral intensity arising from
SiO2. The dependence of NI/NO on ␪ can be expressed as a
d
no⌳o exp
− 1
ͩ ͪ
ͭ ͮ
NO
NS
⌳o sin ␪
=
.
͑2͒
t
n ⌳ exp −
ͩ ͪ
s
s
⌳t sin ␪
Here, ⌳s represents the electron escape depth in the silicon
substrate. The value of ⌳s ͑1.59 nm͒ was calculated from the
escape depth ͑2.11 nm͒ of Si 2p photoelectrons in silicon
excited by Al K␣ radiation,2 by considering the dependence
of the electron escape depth on the kinetic energy of the
electrons.7 The following relation between ⌳t and t should
also be satisfied,
t
nt⌳t exp
− 1
ͩ ͪ
ͭ ͮ
NI
⌳t sin ␪
FIG. 2. Dependence of NI/NO on ␪ for ͑a͒ Kr/O2 plasma oxide, ͑b͒ O2
plasma oxide, and ͑c͒ photo-oxide. As a reference, the dependence of
=
.
͑3͒
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 131.155.81.2
NI/NO on ␪ for the thermal oxide is also shown.
NS
ns⌳s
On: Sun, 23 Nov 2014 02:57:49

081911-3
Nohira et al.
Appl. Phys. Lett. 86, 081911 ͑2005͒
TABLE I. Electron escape depths in low-temperature oxides.
Kr/O2 plasma oxide and in the O2 plasma oxide was 6%
smaller than that in the photo-oxide.
Kr/O2 plasma
oxide
O2 plasma
oxide
In conclusion, the electron escape depths in three kinds
of ultrathin low-temperature-oxide and that in ultrathin ther-
mal oxide were determined from angle-resolved Si 2p pho-
toelectron spectra. We found that the electron escape depths
in the low-temperature oxide layers were 18%–24% smaller
than that in the thermal oxide. The results also indicated that
the electron escape depth in the thermal oxide, whose thick-
ness was close to that of the structural transition layer, was
7% smaller than that in bulk SiO2.
Photo-oxide
2.19
Thermal oxide
2.67
⌳
͑nm͒
2.04
2.06
O
The values of ⌳0 for the four kinds of oxides were de-
termined by least-squares fitting, shown by the solid and
dashed curves in Fig. 2, which were calculated using Eq. ͑1͒,
with the experimental data. Here, we assumed that nO was
the value for bulk SiO2, and that nt=͑ns+nO͒/2 and ⌳t
=͑⌳s+⌳O͒/2. We used values of 5ϫ1028 cm−2 for ns and
2.28ϫ1028 cm−2 for nO.8 The resulting values of ⌳0 are
listed in Table I. As shown in Table I, the electron escape
depth ͑2.67 nm͒ in thermal oxide, whose thickness
͑1.25 nm͒ was comparable to the thickness of the structural
transition layer consisting of SiO2,9 was smaller than
2.86 nm, whose value calculated from the escape depth
͑3.80 nm͒10 of Si 2p photoelectrons excited by Al K␣ radia-
tion in thermally grown bulk SiO2 by considering the depen-
dence of the electron escape depth on the kinetic energy of
the electrons.7 This must be correlated with the observation
that the density of the structural transition layer is approxi-
mately 5% larger than that of bulk SiO2.11 The electron es-
cape depths in three kinds of low temperature oxides listed in
Table I are 18%–24% smaller than that in thermal oxide.
Because the difference between the densities of low-
temperature oxide and thermal oxide is nearly 5%, the dif-
ference between the electron escape depths in each type of
oxide cannot be ascribed to the density difference; rather, it
may be ascribed to the difference in the electron scattering
mechanisms. Specifically, because the fraction of oxygen at-
oms that are not incorporated in the oxide network may be
larger in low-temperature oxide than in thermal oxide, the
scattering frequency of electrons may be larger in the low-
temperature oxide, resulting in a smaller electron escape
depth as compared to that in the thermal oxide. Furthermore,
it should be noted that the electron escape depth in the
The synchrotron radiation experiments were performed
at SPring-8 with the approval of Japan Synchrotron Radia-
tion Research Institute as a Nanotechnology Support Project
of The Ministry of Education, Culture, Sports, Science and
Technology. This work was partially supported by the Min-
istry of Education, Science, Sports and Culture through a
Grant-in-Aid for Scientific Research ͑A͒ ͑No. 32678͒, and
partially by the Ministry of Economy, Trade and Industry and
the New Energy and Industrial Technology Development.
1R. Flitsch and S. I. Raider, J. Vac. Sci. Technol. 12, 305 ͑1975͒.
2Z. H. Lu, J. P. McCaffrey, B. Brar, G. D. Wilk, R. M. Wallance, L. C.
Feldman and S. P. Tay, Appl. Phys. Lett. 71, 2764 ͑1997͒.
3M. Goto, K. Azuma, T. Okamoto, and Y. Nakata, Jpn. J. Appl. Phys., Part
1 42, 7033 ͑2003͒.
4T. Hattori, K. Takahashi, M. B. Seman, H. Nohira, K. Hirose, N. Ka-
makura, Y. Takata, S. Shin, and K. Kobayashi, Appl. Surf. Sci. 212–213,
547 ͑2003͒.
5T. Hattori, K. Hirose, H. Nohira, K. Takahashi, and T. Yagi, Appl. Surf.
Sci. 144–145, 297 ͑1999͒.
6D.-A. Luh, T. Miller, and T.-C. Chiang, Phys. Rev. Lett. 79, 3014 ͑1997͒.
7M. P. Seah and W. A. Dench, Surf. Interface Anal. 1, 2 ͑1979͒.
8M. F. Hochella Jr. and A. H. Carim, Surf. Sci. 197, L260 ͑1988͒.
9K. Hirose, H. Nohira, T. Koike, K. Sakano, and T. Hattori, Phys. Rev. B
59, 5617 ͑1999͒.
10K. Takahashi, K. Hirose, H. Nohira, and T. Hattori, Appl. Phys. Lett. 83,
3422 ͑2003͒.
11Y. Sugita, S. Watanabe, N. Awaji, and S. Komiya, Appl. Surf. Sci. 100/
101, 268 ͑1996͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 131.155.81.2
On: Sun, 23 Nov 2014 02:57:49