Materials Transactions, Vol. 46, No. 3 (2005) pp. 716 to 719
#2005 The Japan Institute of Metals
EXPRESS REGULAR ARTICLE
Local Electronic Structure of Adatom Vacancies on Si(111)-7 7 Surface
Takashi Yamauchi* , Yoshihiro Takahara and Nobutaka Narita
Department of Applied Science for Integrated System Engineering, Graduate School of Engineering,
Kyushu Institute of Technology, Kitakyushu 804-8550, Japan
Local electronic structure around the adatom vacancies on Si(111)-7 7 surface has been investigated using the methods of STM
(scanning tunneling microscopy), STS (scanning tunneling spectroscopy) and the molecular orbital calculation for the cluster of local structure
around each adatom vacancy. In view of the difference of surrounding local structure, the adatoms are classified into four types, i.e., the cornerand center-adatoms in a faulted half (F) cell, and the corner- and center-adatoms in an unfaulted half (UF) cell.
The brightness for the nearest neighbor adatom of each adatom vacancy differs from that of each adatom in the STM images. In the STS
spectra for each type of adatom vacancies, characteristic state is revealed. The new state forms at about 0.15 eV above the HOMO (highest
occupied molecular orbital) level for the adatom. The change of the local electronic structure by the presence of a vacancy is shown by the MO
calculation using the cluster models with and without a vacancy. The new state also is observed in the calculated LDOS (local density of states).
(Received November 18, 2004; Accepted January 19, 2005)
Keywords: adatom, scanning tunneling microscopy, scanning tunneling spectroscopy, local density of states, molecular orbital calculation
1.
Introduction
It has long been known that defects in bulk semiconductors
affect powerful effects on the electronic properties of these
materials and therefore on the performance of devices
fabricated from them. As microelectronic device dimensions
continue to scale downward and surface to volume ratios
increase, there is increasing recognition that the defect exerts
correspondingly powerful effects on semiconductor surfaces.
Recently, the surface analysis with atomic resolution has
become possible by the development of scanning probe
microscopes (SPM)1–3) which usually provide the STS
(scanning tunneling spectroscopy) measurement and the
STM (scanning tunneling microscopy) observation. Defects
on semiconductor surfaces have been studied by the STM
measurement. The 7 7 reconstruction of the Si(111)
surface is one of the most attractive structures. The presence
of adatoms is a unique feature of the 7 7 reconstruction as
compared with other reconstructed structures on the semiconductor surfaces. Recent experiments report that the
adatoms are more weakly bound than the other surface
atoms and can be removed by field evaporation using
STM.4,5) The adatom vacancy forms easier than the vacancy
of the other surface atoms. It is consider that the presence of
the adatom vacancy influences the electronic structure and
the bonding states of the adatom around the vacancy. The
adatom and the adatom vacancy are considered to play an
important role in the dynamical behavior on Si(111) surface.
Thus, understanding the structure and properties of the
adatom vacancy is of great importance. There have been
many reports on the experimental and theoretical analysis of
the reconstructed Si(111)-7 7 surface in recent years.6–9)
However, little attempt has been made to analyze the
electronic states on the particular adatom vacancies.
In the present study, the influence of the adatom vacancy
upon the local electronic structure of its neighbor adatom is
examined by means of the STM and the atom-resolved STS.
To clarify the cause of the configurational effect, molecular
*Graduate
Student, Kyushu Institute of Technology
orbital (MO) calculation is carried out on the cluster of local
structure around each adatom vacancy. The change in the
LDOS (local density of states) by the difference of local
structure around adatom vacancies is discussed on the basis
of experimental and calculated results.
2.
Experiments
The heat treatments and STM measurements were performed with an Omicron SPM system under the ultrahigh
vacuum (UHV) of the pressure <1 10 8 Pa. The STM tips
were prepared from tungsten wire by electrochemical etching
in a 1N-NaOH solution. The tips were annealed to degas
before measurements.
The Si samples used (13:0 3:0 0:5 mm3 ) were cut
from phosphorus-doped (0.375–0.625 cm) Si(111) wafers.
The samples were flashed up to 1450 K to get the clean
surface by direct current heating. In order to make the 7 7
reconstruction of the Si(111) surface, the sample temperature
were decreased to 1200 K and then slowly cooled down to
room temperature. All heat treatments were made in the UHV
chamber. The STM measurements were performed at room
temperature in the constant-height mode with tunneling
current of between 0.3 and 0.4 nA, and sample voltage of
between 1:2 and 1.2 V. The STS measurements were done
simultaneously at each adatom vacancy and each neighbor
adatom to the vacancy on the 7 7 surface. The LDOS was
obtained from (dI/dV)/(I/V)-V curves which were automatically computed from measured I-V relations.
On the other hand, the LDOS on each adatom and each
adatom vacancy was also calculated using the discrete
variational (DV)-X molecular orbital method.10) The 7 7
unit cell model was constructed using the atomic coordinates
reported by Qian and Chadi.11) The clusters used in this
calculation were cut out from the 7 7 unit cell. Except for
the surface atoms, the ends of the cluster were terminated
with hydrogen atoms to prevent surface effect. We obtained
the LDOS for the particular atoms from the calculated results.
Local Electronic Structure of Adatom Vacancies on Si(111)-7 7 Surface
717
V
V
(a) UF-C
V
(c) F-E
(b) UF-E
V
(d) F-C
Fig. 1 STM images of various adatom vacancies (denoted by V) on the Si(111)-7 7 reconstruction obtained under the condition of
occupied states with tunneling current 0.4 nA and tip bias of 1.2 V at room temperature. The upper panels show the image of UF-C adatom
vacancy (a) and UF-E adatom vacancy (b) while the lower panels show the image of F-E adatom vacancy (c) and F-C adatom vacancy (d).
3.
Results and Discussions
There are 12 adatoms in each Si(111)-(7 7) reconstructed unit cell evenly distributed in the unfaulted (UF) and
faulted (F) halves of unit cell, respectively. Among the 12
adatoms, there are only four types of non-equivalent adatoms
which are classified as corner-adatom (C) and center-adatom
(E) on the unfaulted half of surface (UF-C and UF-E), and
corner-adatom and center-adatom on the faulted half of
surface (F-C and F-E), respectively. There is difference
among the atomic arrangements around each adatom, which
result in the difference of the local electronic structures. The
presence of the adatom vacancy is considered to affect the
local electronic structures.
Figures 1(a)–(d) show STM images of various adatom
vacancies on Si(111)-7 7 reconstruction surface obtained
under the condition of occupied states with tip bias 1.2 V at
room temperature. Figures 1(a) and (b) are STM images of
corner-adatom vacancy and center-adatom vacancy in an
unfaulted half cell, respectively. Figures 1(c) and (d) are
STM images of center-adatom vacancy and corner-adatom
vacancy in a faulted half cell, respectively.
As is seen from Fig. 1(a), the nearest neighbor F-C
adatoms of an UF-C adatom vacancy are brighter than those
of the UF-C adatom. The neighbor UF-E adatoms of an UF-C
adatom vacancy are slightly darker than those of the UF-C
adatom. By comparison the neighbor F-E adatom of an UF-C
adatom vacancy with that of the UF-C adatom, it is clear that
the brightness of each adatom differs.
As shown in Fig. 1(b), the nearest neighbor F-E adatoms
of an UF-E adatom vacancy are clearly brighter than those of
the UF-E adatom. However, there is no difference in the
brightness between the neighbor UF-C and UF-E adatoms of
an UF-E adatom vacancy and that of the UF-E adatom. The
same features are observed in STM images (Figs. 1(c) and
(d)) of adatom vacancies in the F half cells. The brightness of
T. Yamauchi, Y. Takahara and N. Narita
the STM image is associated with the tunneling current. The
difference in the tunneling current is closely related to that in
the local electronic structure. The results of the STM
observations show the same features around each adatom
vacancy with and without the stacking-fault. Thus, it is
considered that the difference of the local electronic structure
is produced by correlation between each adatom above the
atomic layer with the stacking-fault.
We think about the local structure around each adatom
vacancy based on the STM observation. The distance from
the F-C adatom vacancy to its neighbor UF-C adatom is
shorter than that from the F-C adatom
about 0.1 nm (1 A)
vacancy to its neighbor F-E adatom. Furthermore, it can be
seen that the location of the nearest neighbor UF-C adatom
for a F-C adatom vacancy shifts to the vacancy side about
0.04 nm. The location of the other neighbor adatoms does not
change. In previous paper,12) we report that adatoms and restatoms have the positive and the negative net charges,
respectively, owing to the charge transfer from adatoms to
rest-atoms. Dev and Seebauer report that the adatom vacancy
with negative charges is more stable than that with positive
charges.13) Formation of the adatom vacancy causes the
change in the interaction between the vacancy with the
negative charge and the nearest neighbor adatom with the
positive charge. The nearest neighbor adatom slightly
approaches to the vacancy, which changes the local electronic states of the adatom vacancy and its nearest neighbor
adatom. As a result, it is considered that the nearest neighbor
adatom of an adatom vacancy shows the bright image in STM
observation.
The neighbor adatoms of an adatom vacancy in same
domain, as the near F-E adatoms of a F-C adatom vacancy,
are slightly darker than those of an adatom. As already
mentioned, the distance between the adatom vacancy and its
nearest neighbor adatom in the different type of domain is
about 0.1 nm shorter than that between an adatom vacancy
and its neighbor adatom in the same domain. The rest-atom is
between the adatom vacancy and its neighbor adatom.
However the rest-atom is not between the adatom vacancy
and its nearest neighbor adatom. The rest-atom has the
negative charges.12) The adatom vacancy with the negative
charges is more stable.13) Therefore, the direct interaction
between the adatom vacancy and its neighbor adatom in the
same domain is very small, because the rest-atom is
interposed between the adatom vacancy and its neighbor
adatom in the same domain. It is considered that the restatom play the role of the intermediary to change the local
electronic structures as the neighbor adatoms of the adatom
vacancy is slightly dark in the STM images. However, in case
the adatom became a vacancy, the interaction between the
neighbor adatom of the vacancy and the rest-atom changes.
As a result, the local electronic structure of the neighbor restatom of the adatom vacancy varies and the interaction
between the rest-atom and its neighbor adatom changes. The
different type of neighbor adatoms of an adatom vacancy in
same domain is slightly darker than those of an adatom.
The STS spectra taken for each adatom and adatom
vacancy site are shown in Fig. 2. Figure 2(a) shows the STS
spectra for the UF-C adatom and the UF-C adatom vacancy.
The HOMO (highest occupied molecular orbital) level for the
On adatom
On vacancy
(a) UF-C
(dI/dV) / (I/V)
718
(b) UF-E
(c) F-E
(d) F-C
-2.0
-1.0
0.0
1.0
Energy, E / eV
Fig. 2 STS spectra of the adatoms and the adatom vacancies in the energy
range from 2:0 eV to 1:5 eV. STS spectra on adatom and on vacancy are
plotted using solid and dashed lines, respectively.
adatom appears at about 0.7 eV below the Fermi level. The
state for the same level exists in the LDOS for the adatom
vacancy. However, the new state forms at about 0.15 eV
above the HOMO level for the UF-C adatom. Figure 2(b)
shows the STS spectra for the UF-E adatom and the UF-E
adatom vacancy. In the spectrum for the UF-E adatom
vacancy, the new state also appears at about 0.15 eV above
the HOMO level for the UF-E adatom. The STS spectra for
the F-E adatom and the F-E adatom vacancy and for the F-C
adatom and the F-C adatom vacancy are shown in Figs. 2(c)
and (d), respectively. From these spectra, the new state for
the adatom vacancy similarly appears at about 0.15 eV above
the HOMO level for the adatom. The new state which is
observed in the STS spectra taken at each adatom vacancy
site is not detected in the STS spectra taken at each adatom
site. Thus, it is considered that the new state is related to the
adatom vacancy.
Figure 3 shows the calculated LDOS using the MO
method. The calculated LDOSs for the UF-C adatom and
the UF-C adatom vacancy are shown in Fig. 3(a). The
HOMO level for the UF-C adatom appears at about 1.0 eV
below the Fermi level. That for the UF-C adatom vacancy
also appears at about 1.0 eV below the Fermi level. However,
the new state for the UF-C adatom vacancy forms at about
0.2 eV above the HOMO level. The calculated adatom
vacancy cluster model only removed an adatom from the
adatom cluster model. Thus, it is considered that the new
Local Electronic Structure of Adatom Vacancies on Si(111)-7 7 Surface
On adatom
On vacancy
719
about 0.2 eV above the HOMO level. The results are
consistent with the LDOSs of the STS measurements.
4.
Conclusion
Local Density of States
(a)UF-C
Interrelation between LDOS and local structure around
adatoms and adatom vacancies on the reconstructed Si(111)7 7 surface has been investigated using STM, STS and MO
calculation. In the STM images, the nearest neighbor
adatoms of the adatom vacancy are brighter than those of
the same kind of the adatom. However, there is no difference
in the brightness between the neighbor adatom except for the
nearest neighbor adatom of the vacancy and the neighbor
adatom of the adatom. This result indicates that the formation
of adatom vacancy changes the electronic structure for the
nearest neighbor adatom of the vacancy. In the STS spectra
for the adatom vacancy, the new state forms at about 0.15 eV
above the HOMO level for the adatom. The calculated LDOS
also shows that the new state associated with the vacancy
forms at about 0.2 eV above the HOMO level for the adatom.
The results are consistent with the STS measurements.
(b)UF-E
(c)F-E
(d)F-C
REFERENCES
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
Energy, E / eV
Fig. 3 Calculated LDOSs of the adatom clusters and the adatom vacancy
clusters in the energy range from 3:0 eV to 3.0 eV. LDOS on adatom
cluster and on vacancy cluster are plotted using solid and dashed lines,
respectively.
state for the cluster model with the adatom vacancy is
associated with the localized vacancy level. Figure 3(b)
shows the calculated LDOSs for the cluster models with and
without the UF-E adatom. Also in these calculated LDOSs,
the new state for the cluster model with vacancy has appeared
at about 0.2 eV higher than that without vacancy. The
calculated LDOSs for the cluster model with and without the
F-E and F-C adatom (Figs. 3(c) and (d)) shows the same
results. Therefore, when the adatom becomes the vacancy, it
turns out that the new state related to the vacancy appears at
1) G. Binning and H. Rohrer: Rev. Mod. Phys. 59 (1987) 615–625.
2) P. K. Hansma and J. Tersoff: J. Appl. Phys. 61 (1987) R1–R24.
3) M. Tsukada, K. Kobayashi, N. Isshiki and H. Kageshima: Surf. Sci.
Rep. 13 (1991) 1–72.
4) H. Uchida, D. Huang, F. Grey and M. Aono: Phys. Rev. Lett. 70 (1993)
2040–2043.
5) B. C. Stipe, M. A. Rezaei and W. Ho: Phys. Rev. Lett. 79 (1997) 4397–
4400.
6) A. Ichimiya, H. Nakahara and Y. Tanaka: Thin Solid Films 281–282
(1996) 1–4.
7) K. Miyake, M. Ishida, M. Uchikawa, K. Hata, H. Shigekawa, Y.
Nannichi and R. Yoshizaki: Surf. Sci. 357–358 (1996) 464–467.
8) T. Hoshino, K. Kumamoto, K. Kokubun, T. Ishimaru and I. Ohdomari:
Phys. Rev. B 51 (1995) 14594–14597.
9) T. Hoshino, K. Kokubun, H. Fujiwara, K. Kumamoto, T. Ishimaru and
I. Ohdomari: Phys. Rev. Lett. 75 (1995) 2372–2375.
10) H. Adachi, M. Tsukada and C. Satoko: J. Phys. Soc. Jpn. 45 (1978)
875–883.
11) G. X. Qian and D. J. Chadi: Phys. Rev. B 35 (1987) 1288–1293.
12) T. Yamauchi, Y. Takahara and N. Narita: Mater. Trans. 42 (2001)
1843–1845.
13) K. Dev and E. G. Seebauer: Surf. Sci. 538 (2003) L495–L499.