Growth of ordered titanium oxide films on Ag(100)

Article abstract of DOI:10.1016/j.ssc.2007.01.037

Titanium oxide films grown on Ag(100) have been investigated with low-energy electron diffraction, X-ray photoelectron spectroscopy and ultraviolet photoelectron spectroscopy. As Ti is deposited on Ag(100) at room temperature in an O₂ atmosphere at 10^-5-10^-7?Torr, a disordered TiO₂-like film is formed. As the surface is subsequently heated at 600 ^{{ring operator}}C, though the composition of the film is nearly unchanged, the film changes into an ordered film with (5×1) periodicity [(5×1) TiO₂ film]. On the other hand, as Ti is deposited in O₂ at 5.0×10^-9-1.0×10^-8?Torr, a disordered TiO-like film is formed. As the surface is subsequently heated at 600 ^{{ring operator}}C, the composition of the film is nearly unchanged and the film changes into an ordered film with (1×1) periodicity [(1×1) TiO film]. Ultraviolet photoelectron spectroscopy measurements show that the (1×1) TiO film has an electronic state around the Fermi level, and the state is ascribed to a partially filled Ti 3d state which has been theoretically expected to exist in a TiO crystal.

Full text of DOI:10.1016/j.ssc.2007.01.037

Solid State Communications 142 (2007) 32–35
www.elsevier.com/locate/ssc
Growth of ordered titanium oxide films on Ag(100)
H. Kanekoa, M. Onoa, K. Ozawab, K. Edamotoa,∗
a
Department of Chemistry, Rikkyo University, Nishi-Ikebukuro, Toshima-ku, Tokyo 171-8501, Japan
b
Department of Chemistry and Materials Science, Tokyo Institute of Technology, Ookayama, Meguro-ku, Tokyo 152-0033, Japan
Received 24 November 2006; accepted 28 January 2007 by S. Ushioda
Available online 2 February 2007
Abstract
Titanium oxide films grown on Ag(100) have been investigated with low-energy electron diffraction, X-ray photoelectron spectroscopy and
−5
−7
ultraviolet photoelectron spectroscopy. As Ti is deposited on Ag(100) at room temperature in an O atmosphere at 10 –10 Torr, a disordered
2
◦
TiO -like film is formed. As the surface is subsequently heated at 600 C, though the composition of the film is nearly unchanged, the film changes
2
−9
−8
into an ordered film with (5 × 1) periodicity [(5 × 1) TiO film]. On the other hand, as Ti is deposited in O at 5.0 × 10 –1.0 × 10 Torr,
2
2
◦
a disordered TiO-like film is formed. As the surface is subsequently heated at 600 C, the composition of the film is nearly unchanged and the
film changes into an ordered film with (1 × 1) periodicity [(1 × 1) TiO film]. Ultraviolet photoelectron spectroscopy measurements show that the
(1 × 1) TiO film has an electronic state around the Fermi level, and the state is ascribed to a partially filled Ti 3d state which has been theoretically
expected to exist in a TiO crystal.
c
ꢀ 2007 Elsevier Ltd. All rights reserved.
PACS: 61.14.Hg; 68.55.-a; 79.60.Dp
Keywords: A. Thin films; C. Surface electron diffraction (LEED); E. Photoelectron spectroscopies
The Ti oxides have attracted much attention in recent years
because of their technological importance. In particular, the
physical and chemical properties of TiO2 have been extensively
studied because of its use as a catalytic support, photocatalyst,
gas sensor and so on [1]. Ti monoxide (TiO) is also of
interest because it has an interesting combination of useful
properties such as extreme hardness, a high melting point
and metallic conductivity [2]. In practical applications, TiO is
expected to be useful as a diffusion barrier [3], anticorrosive
coating material [3], gas sensor [4] and so on. To explore
further the functionality of TiO, it is important to elucidate
the electronic structure in the valence band region. Some
theoretical studies have been reported on the band structure
of TiO [2,5,6]; however, very few experimental investigations
have been performed on the electronic structure of TiO. This
is primarily due to the fact that TiO is rather unstable and it is
difficult to synthesize the single crystal. Recently, Suzuki et al.
found that a TiO single crystal film can be epitaxially grown
on MgO(100) [7,8]. In order to explore the band structure,
however, it is desirable to form the single crystal film on
a conductive substrate. Recently, we have found that a TiO
single crystal film can be formed on TiC(100), which is a
conductive material, by heating the oxygen-adsorbed TiC(100)
surface at ∼1000 ◦C [9]. The two-dimensional band structure
of the film has been measured by angle-resolved photoemission
spectroscopy (ARPES) [10]. However, the formed TiO layer
passivates TiC(100) from further oxidation, and thus a very thin
˚
TiO film whose thickness is estimated to be 1.3–2.0 A is formed
on the TiC(100) surface [9]. Therefore, it is unclear whether
the measured band structure is that inherent to TiO itself or
that modified by the interaction with the TiC(100) substrate.
In the present work, we tried to form a TiO epitaxial film on
Ag(100) by Ti deposition in an O2 atmosphere. The Ag(100)
surface was chosen as a substrate because the Ag has a fcc-type
˚
crystal structure whose lattice constant is 4.09 A [11], which is
˚
similar to that of the TiO crystal (4.18 A [2]). In this work, it
was found that the composition of the Ti oxide film on Ag(100)
can be controlled by the O2 pressure during Ti deposition, and
∗
Corresponding author. Tel.: +81 3 3985 2365; fax: +81 3 3985 2365.
E-mail address: edamoto@rikkyo.ac.jp (K. Edamoto).
c
0038-1098/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ssc.2007.01.037

H. Kaneko et al. / Solid State Communications 142 (2007) 32–35
33
Fig. 1. Ti 2p spectra of the Ag(100) surface after Ti deposition in O at
2
−9
−8
−6
5.0 × 10 , 1.0 × 10 and 1.0 × 10 Torr (gray lines), and of the surfaces
◦
subsequently heated at 600 C (solid lines). The energies of the Ti 2p
levels
3/2
2+
3+
4+
of Ti , Ti and Ti states [12] are shown as vertical bars.
that a TiO film is formed on Ag(100) by Ti deposition in O2 at
1.0 × 10−8–5.0 × 10−9 Torr.
The experiments were performed in an ultra-high vacuum
chamber equipped with a dual anode X-ray tube, a He I
discharge lamp, a 180◦ spherical-sector type electron energy
analyzer (VG 100AX), and low-energy electron diffraction
(LEED) optics. A Mg Kα radiation source was used for all
X-ray photoelectron spectroscopy (XPS) measurements. The
binding energy of the XPS spectra was calibrated by the energy
position of the Ag 3d5/2 peak which is observed in the spectra
at 368.2 eV. For ultraviolet photoelectron spectroscopy (UPS),
the binding energy was calibrated by the Fermi cut-off in the
spectra of the Ta sample holder. The base pressure in the
vacuum system was 1.2 × 10−10 Torr.
Fig. 2. LEED patterns of (a) the Ag(100) clean surface (E = 55 eV), (b)
p
−8
the surface after Ti deposition in O at 1.0 × 10
Torr and subsequent
2
◦
heating at 600 C (E = 60 eV), (c) the surface after Ti deposition in O
p
2
−6
◦
at 1.0 × 10 Torr and subsequent heating at 600 C (E = 58 eV).
p
The Ag(100) surface was cleaned by several cycles of Ar+
ion sputtering and annealing at 600 ◦C. The cleanliness of the
surface was checked by XPS. The prepared Ag(100) clean
surface gives a sharp (1 × 1) LEED pattern (Fig. 2(a)). The
Ti oxide films were grown on the Ag(100) surface at room
temperature by Ti deposition in an O2 atmosphere (5.0 ×
10−9–1.0 × 10−5 Torr). Ti was deposited onto the surface
from an electron-beam evaporator (Omicron EFM3). During
the deposition, the flux current of the Ti cations at the exit of
the evaporation source was kept at ∼10 nA, and the deposition
time was 9 min for all experiments. During the deposition,
the Ag(100) sample was kept at room temperature for all
experiments.
1.0 × 10−6 Torr, the Ti 2p peaks (2p3/2, 2p1/2) are observed
at (458.9, 464.5 eV). Nearly the same spectrum is observed
independent of the O2 pressure during 10−5–10−7 Torr. As
Ti is deposited in O2 at 1.0 × 10−8 and at 5.0 × 10−9 Torr,
the Ti 2p peaks are observed at (454.8, 460.0 eV) and (454.6,
459.5 eV), respectively. The Ti 2p peaks of Ti atoms in the
oxidation state of Ti4+, Ti3+ and Ti2+ have been observed at
(458.8, 464.5 eV), (457.1, 462.8 eV) and (455.4, 461.1 eV),
respectively [12]. Therefore, the Ti oxide films formed on the
Ag(100) surface in O2 at 10−5–10−7 Torr are ascribed to a
TiO2-like film, and those formed in O2 at 1.0 × 10−8 and at
5.0×10−9 Torr are ascribed to a TiO-like film. The ratios of the
integrated intensities of the O 1s and Ti 2p peaks (IO1s/ITi2p) are
estimated to be 0.74, 0.40 and 0.26 for Ti oxide films formed in
O2 at 1.0×10−6, 1.0×10−8 and 5.0×10−9 Torr, respectively.
Fig. 1 shows Ti 2p spectra of the Ag(100) surface after Ti
deposition in an O2 atmosphere of 1.0 × 10−6, 1.0 × 10−8 and
5.0 × 10−9 Torr (gray lines). When Ti is deposited in O2 at

34
H. Kaneko et al. / Solid State Communications 142 (2007) 32–35
By the correction using the photoionization cross sections of
O 1s and Ti 2p levels [13], the compositions of the Ti oxides
can be estimated from the ratio of the O 1s and Ti 2p peak
intensities (IO1s/ITi2p). The results show that the compositions
of the Ti oxide films formed in O2 at 1.0 × 10−6, 1.0 × 10−8
and 5.0×10−9 Torr are TiO2.1, TiO1.1 and TiO0.73, respectively,
also suggesting that TiO2-like film is formed in O2 at 1.0×10−6
Torr while the TiO-like film is formed in O2 at 1.0 × 10−8 and
5.0 × 10−9 Torr.
In Fig. 1, the Ti 2p spectra of the Ti oxide films after heating
at 600 ◦C are also shown as solid lines. Fig. 1 shows that
the spectral intensities are decreased while the peak energies
are nearly unchanged upon heating at 600 ◦C for all measured
films, indicating that the oxidation state of Ti in the oxide
films is nearly unchanged by heating at 5600 ◦C. The surface
covered with the Ti oxide films formed at room temperature
gives diffuse LEED patterns with broadened (1 × 1) spots
independent of the O2 pressure during Ti deposition, and
therefore the TiO2-like film and TiO-like film formed at room
temperature are thought to have structures with disorder over
a long-range scale. However, when the surface covered with
the TiO2-like film is heated at 600 ◦C, the disordered pattern
changes into a (5 × 1) pattern. The LEED pattern of the surface
covered with the TiO2-like layer formed by Ti deposition in O2
at 1 × 10−6 Torr and subsequently heated at 600 ◦C is shown
in Fig. 2(b). The surface gives a (5 × 1) pattern, and this film
will be called a (5 × 1) TiO2 film hereafter. On the other hand,
when the surface covered with the TiO-like film is heated at
600 ◦C, the disordered pattern changes into a sharp (1 × 1)
pattern. Fig. 2(c) shows the pattern of the surface on which Ti
is deposited in O2 at 1.0 × 10−8 Torr and subsequently heated
at 600 ◦C. The surface gives a clear (1 × 1) pattern, indicating
that the periodicity of the Ti oxide film after heating at 600 ◦C
is the same as that of the Ag(100) surface. This film will be
called a (1 × 1) TiO film hereafter. The thickness (d) of the
ordered Ti oxide film is estimated from the measurements of
Fig. 3. He I UPS spectra of (a) the Ag(100) clean surface, (b) the surface after
−6
◦
Ti deposition in O at 1.0×10 Torr and subsequent heating at 600 C and (c)
2
−8
the surface after Ti deposition in O at 1.0 × 10 Torr and subsequent heating
2
◦
at 600 C.
film has a distorted rutile TiO2 structure whose lattice constants
5
˚
˚
are a = rAg = 4.82 A and c = rAg = 2.89 A with the a and
3
¯
c axes parallel to the [110] and [110] directions of Ag(100),
respectively. In this case, the (5×1) LEED pattern (Fig. 2(b)) is
interpreted as originating from the two types of (5×1) domains
which are rotationally misaligned by 90◦.
TiO2 is a typical semiconductor with a band gap of
∼3 eV [11], while TiO is expected to have metallic conductivity
due to the existence of a partially filled Ti 3d band around
the Fermi level (EF) [2,5,6]. Therefore, the electronic states
of the Ti oxide films, especially in the vicinity of EF, are of
interest. Fig. 3 shows UPS spectra of the Ag(100) clean surface
and of the surfaces covered with Ti oxide films. For the clean
surface (Fig. 3(a)), the Ag 4d band is observed at 4–7 eV
and the spectrum is typical of the UPS spectra of Ag. As the
Ag(100) surface is covered by the (5 × 1) TiO2 film, the Ag
4d band emissions are suppressed and a broad band appears
at 4–8 eV (Fig. 3(b)). The band at 4–8 eV has been typically
observed in UPS spectra of TiO2, and has been ascribed to
the O 2p induced valence band [15,16]. Thus the UPS result is
compatible with the model deduced from the XPS results that
the Ti oxide film with (5 × 1) periodicity has a composition
of TiO2. Fig. 3(b) shows that there is no state in the vicinity
of EF in the spectrum of the (5 × 1) TiO2 film, suggesting
that the film corresponds to nearly stoichiometric TiO2 with a
negligible number of O vacancies. On the other hand, as the
surface is covered by the (1 × 1) TiO film, the peaks appear
at just below EF, 2.8 eV, and ∼6 eV (Fig. 3(c)). The O 2p
band of a TiO film was observed at 6.5 eV and 6.4 eV in
the spectra of TiO/MgO(100) [8] and of TiO/TiC(100) [10],
respectively, and thus the band at ∼6 eV is ascribed to the O
2p band of TiO. In contrast to the case of the (5 × 1) TiO2
the attenuation of Ag 3d peaks, using the equation I/I0
=
exp(−d/λ), where I0 and I are the Ag 3d peak intensities in
XPS spectra of the clean surface and of the surface covered
with the oxide film, respectively, and λ is the inelastic mean free
˚
path of Ag 3d photoelectrons (λ is estimated to be 19.3 A in
˚
TiO2 and 17.8 A in TiO [14]). The estimated thicknesses are
12.7 A for the (5 × 1) TiO2 film and 11.5 A for the (1 × 1) TiO
˚
˚
film.
TiO is known to have a NaCl-type crystal structure whose
˚
lattice constant (4.18 A) is slightly larger than that of Ag
˚
(4.09 A). Therefore, it is deduced from the XPS and LEED
results that a slightly compressed TiO(100) single-crystal film
is formed on the Ag(100) surface as the (1 × 1) TiO film when
the surface covered by the TiO-like film is heated at 600 ◦C. It
is more difficult to deduce the structure of the (5×1) TiO2 film.
If TiO2 crystallizes into a rutile structure, the lattice constants
˚
˚
are a = 4.58 A and c = 2.95 A [1], and thus the lattice is not
commensurate with that of Ag(100). However, since the lattice
constants a and c are close to rAg and rAg, respectively, where
5
3
˚
rAg = 2.89 A is the nearest neighbor distance in Ag crystal, one
of the possible models of the (5 × 1) TiO2 structure is that the

H. Kaneko et al. / Solid State Communications 142 (2007) 32–35
35
film, a peak is observed at just below EF in the spectrum of the
Acknowledgements
(1 × 1) TiO film. The peak is assigned as the partially filled
Ti 3d state, which has been expected to exist in TiO from the
theoretical calculations [2,5,6]. The peak just below EF has also
been observed in UPS spectra of the TiO/MgO(100) [8] and
TiO/TiC(100) [10] systems, and thus the state is characteristic
of the electronic structure of a TiO film. We think that the state
is responsible for the metallic nature of the TiO crystal. The
origin of the peak at 2.8 eV is unknown at present. In the
case of TiO/MgO(100), a peak was observed at 3.5 eV and
the peak was tentatively ascribed to the adsorbed oxygen on
the TiO surface [8]. The existence of the ∼3 eV peak in the
spectra of TiO/TiC(100) is not clear due to the existence of
large emissions from the underlying TiC(100) substrate [10].
As discussed above, the spectrum of the (1 × 1) TiO film is
essentially compatible with those of TiO/MgO(100) [8] and
TiO/TiC(100) [10].
In conclusion, it was found that two types of ordered Ti
oxide films can be formed on the Ag(100) surface; one is
the (1 × 1) TiO film which is formed by Ti deposition in
an O2 atmosphere at 1.0 × 10−8–5.0 × 10−9 Torr at room
temperature and subsequent heating at 600 ◦C, and the other
is the (5 × 1) TiO2 film which is formed by Ti deposition in
an O2 atmosphere at 10−5–10−7 Torr at room temperature and
subsequent heating at 600 ◦C. Since the lattice constant of TiO
is slightly larger than that of Ag, a slightly compressed TiO
single-crystal film is thought to be formed as the former film.
The UPS measurements showed that the electronic structure of
the (1 × 1) TiO film is characterized by the existence of the
state around EF. The state is ascribed to the partially filled Ti
3d band which has been theoretically expected to be formed in
a TiO crystal [2,5,6].
The work was financially supported by a Grants-in-Aid
for Scientific Research (No. 17550139) from the Ministry of
Education, Culture, Sports, Science, and Technology of Japan,
and by Nippon Sheet Glass Foundation for Materials Science
and Engineering.
References
[1] U. Diebold, Surf. Sci. Rep. 48 (2003) 53.
[2] A. Neckel, P. Rastl, R. Eibler, P. Weinberger, K. Schwarz, J. Phys. C: Solid
State Phys. 9 (1976) 579.
[3] R.A. Za´rate, V.M. Fuenzalida, Vacuum 76 (2004) 13.
[4] E. Comini, G. Sberveglieri, M. Ferroni, V. Guidi, G. Martinelli, Sensor.
Actuator. B 68 (2000) 175.
[5] R. Ahuja, O. Eriksson, J.M. Wills, B. Johansson, Phys. Rev. B 53 (1996)
3072.
[6] A. Yamasaki, T. Fujiwara, Phys. Rev. B 66 (2002) 245108.
[7] T. Suzuki, R. Souda, Surf. Sci. 445 (2000) 506.
[8] T. Suzuki, R. Souda, W. Maus-Friedrichs, V. Kempter, Phys. Rev. B 62
(2000) 8306.
[9] Y. Shirotori, K. Sawada, K. Ozawa, K. Edamoto, S. Otani, Jpn. J. Appl.
Phys. 42 (2003) 1725.
[10] Y. Shirotori, K. Sawada, K. Ozawa, K. Edamoto, S. Otani, Surf. Sci. 584
(2005) 237.
[11] C. Kittel, Introduction to Solid State Physics, John Wiley & Sons, New
York, 2004.
[12] S. Pe´tigny, H. Moste´fa-Sba, B. Domenichini, E. Lesniewska,
A. Steinbrunn, S. Brourgeois, Surf. Sci. 410 (1998) 250.
[13] J.J. Yeh, I. Lindau, At. Data Nucl. Data Tables 32 (1985) 1.
[14] S. Tanuma, C.J. Powell, D.R. Penn, Surf. Interface Anal. 35 (2003) 268.
[15] J. Nerlov, Q. Ge, P.J. Møller, Surf. Sci. 348 (1996) 28.
[16] A.G. Thomas, W.R. Flavell, A.R. Kumarasinghe, A.K. Mallick,
D. Tsoutsou, G.C. Smith, R. Stockbauer, S. Patel, M. Gra¨tzel, R.
Hengerer, Phys. Rev. B 67 (2003) 035110.