Henini M. Handbook of Self Assembled Semiconductor Nanostructures for Novel devices in Photonics and Electronics
Подождите немного. Документ загружается.
Optical and Vibrational Properties of Self-assembled GaN Quantum Dots 269
79. N. Garro , A. Cros , J.A. Budagosky , A. Cantarero , A. Vinattieri , M. Gurioli , S. Founta , H. Mariette ,
and B. Daudin , Appl. Phys. Lett. 87 , 011101 ( 2005 ) .
80. M. Cardona , in: M. Cardona and G. Günterodt, eds., Light Scattering in Solids II , ( Springer , Berlin ,
1982 ) .
81. J. Gleize , F. Demangeot , J. Frandon , M.A. Renucci , M. Kuball , B. Damilano , N. Grandjean , and
J. Massies , Appl. Phys. Lett. 79 , 686 ( 2001 ) .
82. M. Kuball , J. Gleize , S. Tanaka , and Y. Aoyagi , Appl. Phys. Lett. 78 , 987 ( 2001 ) .
83. A. Cros, J.A. Budagosky, N. Garro, A. Cantarero, J. Coraux, H. Renevier, M.G. Proietti, V. Favre-
Nicolin, and B. Daudin, Phys. Status Solidi (c), 4, 2379 (2007).
84. M. Miyamura , K. Tachibana , and Y. Arakawa , Appl. Phys. Lett. 80 , 3937 ( 2002 ) .
85. S. Tanaka , H. Hirayama , Y. Aoyagi , Y. Narukawa , Y. Kawakami , S. Fujita , and S. Fujita , Appl.
Phys. Lett. 71 , 1299 ( 1997 ) .
86. F. Widmann , J. Simon , N.T. Pelekanos , B. Daudin , G. Feuillet , J.L. Rouvière , and G. Fishman ,
Microelect. J. 30 , 353 ( 1999 ) .
87. A. Niwa , T. Ohtoshi , and T. Kuroda , Jpn. J. Appl. Phys. 35 , L599 ( 1996 ) .
88. R. Cingolani , A. Botchkarev , H. Tang , H. Morkoc , G. Traetta , G. Coli , M. Lomascolo , A. Di Carlo ,
F. Della Sala , and P. Lugli , Phys. Rev. B . 61 , 2711 ( 2000 ) .
89. V. Ranjan , G. Allan , C. Priester , and C. Delerue , Phys. Rev. B . 68 , 115305 ( 2003 ) .
90. V.R. Velasco , J. Tutor , and H. Rodriguez-Coppola , Surf. Sci. 565 , 259 ( 2004 ) .
91. I. Lo , W.T. Wang , M.H. Gau , S.F. Tsay , and J.C. Chiang , Phys. Rev. B . 72 , 245329 ( 2005 ) .
92. S. Schulz , S. Schumacher , and G. Czycholl , Phys. Rev. B . 73 , 245327 ( 2006 ) .
93. A. Lakdja , B. Bouhafs , and P. Ruterana , Phys. Status Solidi (a) . 203 , 2247 ( 2006 ) .
94. S.L. Chuang and C.S. Chang , Phys. Rev. B . 54 , 2491 ( 1996 ) .
95. D. Fritsch , H. Schmidt , and M. Grundmann , Phys. Rev. B . 67 , 235205 ( 2003 ) .
96. Y.M. Sirenko , J.B. Jeon , K.W. Kim et al., Phys. Rev. B . 53 , 1997 ( 1996 ) .
97. S.L. Chuang and C.S. Chang , Sem. Sci. Tech. 12 , 252 ( 1997 ) .
98. F. Mireles and S.E. Ulloa , Phys. Rev. B . 60 , 13659 ( 1999 ) .
99. B. Gil , in: J.I. Pankove and T.D. Moustakas, eds., Gallium Nitride II ( Academic Press , San Diego ,
1999 ) .
100. T. Taliercio , P. Lefebvre , M. Gallart , and A. Morel , J. Phys.: Condens. Matter . 13 , 7027 ( 2001 ) .
101. V.A. Fonoberov and A.A. Balandin , J. Appl. Phys. 94 , 7178 ( 2003 ) .
102. N. Vukmirovic , Z. Ikonic , D. Indjin , and P. Harrison , J. Phys.: Condens. Matter . 18 , 6249
( 2006 ) .
103. J.F. Nye, Physical Properties of Crystals ( Oxford University Press, 1985 ) .
104. M. Grundmann , O. Stier , and D. Bimberg , Phys. Rev. B . 52 , 11969 ( 1995 ) .
105. F. Bernardini , V. Fiorentini , and D. Vanderbilt , Phys. Rev. B . 56 , R10024 ( 1997 ) .
106. P. Waltereit , O. Brandt , A. Trampert , H.T. Grahn , J. Menniger , M. Ramsteiner , M. Reiche , and
K.H. Ploog , Nature 406 , 865 ( 2000 ) .
107. S.F. Chichibu , A.C. Abare , M.S. Minsky , S. Keller , S.B. Fleischer , J.E. Bowers , E. Hu , U.K. Mishra ,
L.A. Coldren , S.P. DenBaars , and T. Sota , Appl. Phys. Lett. 73 , 2006 ( 1998 ) .
108. J.S. Im , H. Kollmer , J. Off , A. Sohmer , F. Scholz , and A. Hangleiter , Phys. Rev. B . 57 , R9435
( 1998 ) .
109. O. Ambacher , J. Smart , J.R. Shealy , N.G. Weimann , K. Chu , M. Murphy , W.J. Schaff , L.F. Eastman ,
R. Dimitrov , L. Wittmer , M. Stutzmann , W. Rieger , and J. Hilsenbeck , J. Appl. Phys. 8 5 , 3222
( 1999 ) .
110. F. Widmann , J. Simon , B. Daudin , G. Feuillet , J.L. Rouvière , and N.T. Pelekanos , Phys. Rev. B . 58 ,
R15989 ( 1998 ) .
111. R.A. Hogg , C.E. Norman , A.J. Shields , M. Pepper , and M. Iizuka , Appl. Phys. Lett. 7 6 , 1428
( 2000 ) .
112. J. Simon , N.T. Pelekanos , C. Adelmann , E. Martínez-Guerrero , R. André , B. Daudin , L.S. Dang ,
and H. Mariette , Phys. Rev. B . 68 , 035312 ( 2003 ) .
113. T. Bretagnon , P. Lefebvre , P. Valvin , R. Bardoux , T. Guillet , T. Taliercio , B. Gil , N. Grandjean ,
F. Semond , B. Damilano , A. Dussaigne , and J. Massies , Phys. Rev. B . 73 , 113304 ( 2006 ) .
114. T. Nakaoka , S. Kako , and Y. Arakawa ,
Phys. Rev. B . 73 , 121305 ( 2006 ) .
115. T. Andreev , Y. Hori , X. Biquard , E. Monroy , D. Jalabert , A. Farchi , M. Tanaka , O. Oda , L.S. Dang ,
and B. Daudin , Phys. Rev. B . 71 , 115310 ( 2005 ) .
CH007-I046325.indd 269CH007-I046325.indd 269 6/27/2008 5:16:20 PM6/27/2008 5:16:20 PM
270 Handbook of Self-assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
116. T. Andreev , N.Q. Liem , Y. Hori , M. Tanaka , O. Oda , L.S. Dang , and B. Daudin , Phys. Rev. B . 73 ,
195203 ( 2006 ) .
117. G. Salviati , F. Rossi , N. Armani , V. Grillo , O. Martinez , A. Vinattieri , B. Damilano , A. Matsuse , and
N. Grandjean , J. Phys.: Condens. Matter . 16 , S115 ( 2004 ) .
118. S. De Rinaldis , I. D’Amico , E. Biolatti , R. Rinaldi , R. Cingolani , and F. Rossi , Phys. Rev. B . 65 ,
081309 ( 2002 ) .
119. P. Riblet , S. Tanaka , P. Ramvall , S. Nomura , and Y. Aoyagi , Solid State Comm . 109 , 377 ( 1999 ) .
120. S. Kalliakos , T. Bretagnon , P. Lefebvre , T. Taliercio , B. Gil , N. Grandjean , B. Damilano ,
A. Dussaigne , and J. Massies , J. Appl. Phys. 96 , 180 ( 2004 ) .
121. Y.-H. Cho , B.J. Kwon , J. Barjon , J. Brault , B. Daudin , H. Mariette , and L.S. Dang , Appl. Phys. Lett.
81 , 4934 ( 2002 ) .
122. S. Kako , M. Miyamura , K. Tachibana , K. Hoshino , and Y. Arakawa , Appl. Phys. Lett. 8 3 , 984
( 2003 ) .
123. T. Bretagnon , S. Kalliakos , P. Lefebvre , P. Valvin , B. Gil , N. Grandjean , A. Dussaigne , B. Damilano ,
and J. Massies , Phys. Rev. B . 68 , 205301 ( 2003 ) .
124. C. Santori , S. Götzinger , Y. Yamamoto , S. Kako , K. Hoshino , and Y. A r a k a w a , Appl. Phys. Lett. 8 7 ,
051916 ( 2005 ) .
125. A. Neogi , H. Morkoç , T. Kuroda , A. Tackechi , T. Kawazoe , and M. Ohtsu , Nano Lett . 5 , 213
( 2005 ) .
126. A. Petersson , A. Gustafsson , L. Samuelson , S. Tanaka , and Y. Aoyagi , Appl. Phys. Lett. 7 4 , 3513
( 1999 ) .
127. Y. Yao , T. Sekiguchi , Y. Sakuma , M. Miyamura , and Y. Arakawa , Scripta Materialia . 55 , 679 – 682
( 2006 ) .
128. D.P. Williams , A.D. Andreev , D.A. Faux , and E.P. O’Reilly , Physica E . 21 , 358 ( 2004 ) .
129. T. Nakaoka , S. Kako , and Y. Arakawa , Physica E . 32 , 148 ( 2006 ) .
130. F. Rol , B. Gayral , S. Founta , B. Daudin , J. Eymery , J.-M. Gérard , H. Mariette , L.S. Dang , and
D . P e y r a d e , Phys. Status Solidi . 243 , 1652 ( 2006 ) .
131. F. Rol , B. Gayral , S. Founta , H. Mariette , B. Daudin , L.S. Dang
, J. Bleuse , D. Peyrade , J.M. Gérard ,
and B. Gayral , Phys. Rev. B . 75 , 125306 ( 2007 ) .
132. N. Iizuka , K. Kaneko , N. Suzuki , T. Asano , S. Noda , and O. Wada , Appl. Phys. Lett. 7 7 , 648
( 2000 ) .
133. C. Gmachl , S.V. Frolov , H.M. Ng , S.-N.G. Chu , and A.Y. Cho , Electron. Lett. 37 , 378 ( 2001 ) .
134. Kh. Moumanis , A. Helman , F. Fossard , M. Tchernycheva , A. Lusson , F.H. Julien , B. Damilano ,
N. Grandjean , and J. Massies , Appl. Phys. Lett. 82 , 868 ( 2003 ) .
135. M. Tchernycheva , L. Nevou , L. Doyennette , A. Helman , R. Colombelli , F.H. Julián , F. Guillot ,
E. Monroy , T. Shibata , and M. Tanaka , Appl. Phys. Lett. 87 , 101912 ( 2005 ) .
136. F. Guillot , E. Bellet-Amalric , E. Monroy , M. Tchernycheva , L. Nevou , L. Doyennette , F.H. Julien ,
L.S. Dang , T. Remmele , M. Albrecht , T. Shibata , and M. Tanaka , J. Appl. Phys. 100 , 044326
( 2006 ) .
137. A. Vardi , N. Akopian , G. Bahir , L. Doyennette , M. Tchernycheva , L. Nevou , F.H. Julien , F. Guillot ,
and E. Monroy , Appl. Phys. Lett. 88 , 143101 ( 2006 ) .
CH007-I046325.indd 270CH007-I046325.indd 270 6/27/2008 5:16:21 PM6/27/2008 5:16:21 PM
GaSb/GaAs Quantum Nanostructures by Molecular Beam Epitaxy
Koichi Yamaguchi
,
1
Shiro Tsukamoto
,
2
and Kazunari Matsuda
3
1
Department of Electronic Engineering, The University of Electro-Communications, 1-5-1
Chofugaoka, Chofu, Tokyo 182-8585, Japan
2
Center for Collaborative Research, Anan National College of Technology, 265 Aoki,
Minobayashi, Anan, Tokushima 774-0017, Japan
3
International Research Center for Elements Science, Institute for Chemical Research,
Kyoto University, Gokasho, Uji, Kyoto 611-0011, Japan
8.1 Introduction
GaSb-based quantum nanostructures in a GaAs matrix have unique band alignment of a stag-
gered type-II [1, 2] large bowing of the band gap [2] and strong hole localization [3] . Based on
these properties, GaSb/GaAs nanostructures including quantum wells (QWs) and quantum dots
(QDs) are of considerable interest for use in various applications, such as infrared emitting and
detecting devices, and memory and spin-controlled devices. In the GaSb/GaAs heteroepitaxial
growth, an exchange of group V elements and a large lattice mismatch of 7.8% between GaSb
and GaAs should be noted to control the heterointerface. Chidley et al. fi rst reported the growth
of GaSb QWs on GaAs by metalorganic chemical vapour deposition (MOCVD) [4] . In their report,
a transition from the pseudomorphic thin fi lm growth to the island growth was observed at a
thickness of above 1.5 nm. The growth of GaSb QDs on GaAs was fi rst reported by Hatami et al.
[1] . The GaSb/GaAs QDs were spontaneously formed by molecular beam epitaxy (MBE) using a
Stranski–Krastanov (SK) growth technique, which has been developed in InAs/GaAs and Ge/Si
heterostructures [5–7]. Currently, the SK growth technique of the self-assembled QDs is the most
promising way for producing coherent QDs. Recently, the self-assembled GaSb QDs on GaAs have
been attempted by many groups.
This chapter focuses on the GaSb/GaAs quantum nanostructures by MBE. Section 8.2 intro-
duces the surface reconstruction of GaSb on GaAs and the heterointerface structure of GaSb/
GaAs. Section 8.3 introduces the self-assembled GaSb/GaAs QDs and the optical properties of
GaSb/GaAs QDs.
8.2 Surface and interface structures of GaSb/GaAs
In order to fabricate GaSb/GaAs quantum nanostructures, we have to control surface and inter-
face structures. In this section, we will discuss the surface and interface structures for GaSb
growth on GaAs and for GaAs growth on GaSb. First, we describe surface reconstructions of the
GaSb grown on the GaAs(001) in Section 8.2.1. Then, we analyse a broadening of GaAs/GaSb
heterointerface by using a kinetic model of surface exchange reaction between Sb and As atoms
in Section 8.2.2.
CHAPTER 8
CH008-I046325.indd 271CH008-I046325.indd 271 6/25/2008 2:30:02 PM6/25/2008 2:30:02 PM
272 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
8.2.1 Surface reconstruction of GaSb on GaAs and in situ STM observation
The fi ne deposition control of various semiconductor materials by MBE has enabled the fab-
rication of quantum-size effect devices. III–V semiconductor materials such as GaAs and InP
have been used as diverse devices, in particular, optoelectronic device applications. Recently,
the novel fabrication method that irradiates antimony (Sb) on a GaAs(001) surface and forms
an Sb-terminated layer by As–Sb exchange reaction has been devised [8] . Sb irradiation brings
those surface reconstruction changes from c (4 4) to ( n 3 ) ( n 1, 2, 4). In the new appli-
cation of Sb/GaAs(001) it has been proposed that InAs quantum dots grown on Sb-irradiated
GaAs(001) by MBE obtained high density and high uniformity [9, 10] . However, a detailed struc-
ture of this surface has not been investigated. Although surface reconstruction of Sb-terminated
GaAs(001) has been reported to exist ((2 4) [11, 12] and (2 8)) [12] , ( n 3) surface recon-
struction has not been reported yet. The 3 ordered reconstructions have also been studied for
Sb-passivated GaAs (001) at room temperature [13–15] . Although scanning tunneling micros-
copy (STM) observation of Sb-passivated surface-(1 3) [15] and GaSb on GaAs(001)- c (4 4 )
[16] have been reported, any reconstruction model is not discussed for these 3 ordered recon-
structions, including (2 3) and (4 3). Recently, using in situ STM and fi rst-principles calcula-
tion, surface reconstructions were investigated on Sb-irradiated GaAs(001) formed by MBE and
its structure models were proposed.
It was used a non-doped GaAs(001) just-oriented substrate. After oxide layer removal, a GaAs
buffer layer of 200 nm was grown at 570°C under an As
4
pressure of 5 1 0
5
Pa and Ga pres-
sure of 3.3 1 0
6
Pa. Next, under As
4
fl ux, the substrate was set up to an irradiation tempera-
ture of 435°C. The c (4 4) reconstruction of the As-rich surface was confi rmed by refl ection
high energy electron diffraction (RHEED). When the Sb shutter was opened, simultaneously clos-
ing the As shutter, an Sb
4
fl ux of 1.0 1 0
6
Pa was irradiated on the As-terminated GaAs so
that an As–Sb exchange reaction was carried out [8] . The irradiation amount depends on the
RHEED pattern changing from c (4 4 ) t o ( 2 3) or (4 3). The substrate was annealed for
over 10 minutes at the same temperature after the irradiation. The (4 3 ) o r ( 2 3) reconstruc-
tions were unchanged during substrate quenching to 200°C. The sample was observed at the
same temperature on the same substrate heater by the in situ STM located inside the MBE growth
chamber [17] STM experiments used a tungsten tip and were performed under the background
pressure of 8 1 0
8
Pa. All STM images were acquired by a constant current of 0.1 nA, a scan
speed of 250 nm/s and a tip bias of 3 V. Then, fi rst-principles density-functional total-energy
calculations were performed based on the density-functional theory using the Vienna ab initio
simulation package [18] . The local-density approximation was used for the exchange correlation
and the projector augmented wave [19, 20] pseudopotentials. The Ga 3d electrons were treated
as part of the valence band. The cutoff energy for the plane-wave basis was 400 eV. We employed
supercells containing six atomic layers of GaAs and two layers of GaSb where the bottom side of
the slab was terminated with fractionally charged hydrogen in order to maintain bulk properties.
Figure 8.1 shows RHEED patterns in the [110] and
[]110
azimuth and a fi lled-state STM image
on an Sb-irradiated surface. The RHEED patterns indicate three-fold order distinctly in the
[]110
azimuth ( Fig. 8.1 a) and weak 2 and very weak 4 in the [110]
azimuth ( Fig. 8.1 b) so that the
[110]
[110]
-
(a) (b)
[1-10]
[110]
10 nm
(c)
Figure 8.1 RHEED patterns and STM image of Sb-irradiated GaAs(001) [21]. (a) Slightly obscure two-fold in
[110] azimuth and (b) three-fold in [11
–
0] azimuth. (c) STM image at 200°C, tunnelling current of 0.1 nA, tip bias
of 3 V, scan size of 50 50 nm, scan speed of 250 nm/s and background pressure of 8 1 0
8
Pa. The height
range of the surface is approximately 1.0 nm. The periodicity perpendicular to the [11
–
0] direction is 1.2 nm.
CH008-I046325.indd 272CH008-I046325.indd 272 6/25/2008 2:30:02 PM6/25/2008 2:30:02 PM
GaSb/GaAs Quantum Nanostructures by Molecular Beam Epitaxy 273
surface is dimerized (2 3) (or (4 3)). STM image ( Fig. 8.1 c) shows that anisotropic islands
are formed with multilayer behaviour and kinks that result from steps. The step density of this
surface, defi ned here as the total length of step edges per unit area, is 183 μ m
1
, whereas that
of GaAs(001) c (4 4 ) i s 4 0 μ m
1
[22] . This Sb-irradiated surface was thus found to be rough,
which agreed with the ribbon-like structures reported by Bennett et al. [16]. They suggested that
the additional growth of GaSb of 2.5 ML on GaSb of 1 ML (3.5 ML total)/GaAs(001) surface
enhances the anisotropy of the surface attributed to the direction-dependent strain associated
with the dimer-based surface reconstructions of the wetting layers. The result reinforces their
suggestion because of the anisotropy along the
[]110
direction. The Sb irradiation causes an
increase of surface step density. The same phenomenon is reported in the co-deposition of Sb and
Si on Si(001) [23] and AlAs on GaAs(001) [24] . The high QD density can be achieved by using
these high-step density surfaces since dangling bonds on the step edges plays a role of nucleation
sites [10, 24] . However, the incorporation mechanism of InAs deposition into Sb/GaAs(001) and
the segregation mechanism of Sb remain unclear. Therefore, it is important to know its initial
surface which means the (2 3) (or (4 3)) reconstructed Sb-terminated GaAs(001).
Figure 8.2 shows a high-resolution empty-state image of Sb-irradiated GaAs(001). A 3
structure of approximately 1.2 nm periodicity was confi rmed perpendicular to the
[]110
direc-
tion and many swinging dimer rows were observed along the []110
direction, similar to the dis-
ordered (1 3) observed by Moriarty et al. [15]. On the other hand, the RHEED pattern showed
slightly obscure two-fold order in the [110] azimuth. Therefore, these surface reconstructions
can be classed as disordered (2 3) structures with swinging dimer rows along the
[]110
direc-
tion. In order to confi rm the interpretation of these structures, the models proposed by Righi
et al. [25] are useful. They studied the respective stability for (1 3) [26] (1 3 ) / c (2 6) [27],
as well as for (4 3) reconstructions for the Sb-rich surface on GaSb(001). Since the model of
h0(4 3) proposed by Righi et al. [25] has the lowest surface energy, we consider three (2 3 ) -
reconstructed structures for Sb/GaAs(001), as shown in Fig. 8.3 , based on this h0(4 3 ) s t r u c -
ture. It was considered that fi rst- and second-layer atoms consisted of Sb only since a 3 RHEED
pattern in the
[]110
azimuth can be achieved by about 2 ML Sb-irradiation on GaAs(001). Note
that Sb–Sb bonds are stronger than Ga–Sb bonds [28, 29] . It is assumed that a little re-evaporation
of Sb atoms from the top layer might occur during annealing at 435
o
C, but it does not affect
reconstruction formations.
3 nm(a) 1 nm(b)
[110]
[1-10]
Figure 8.2 The high-resolution empty-state STM image of Sb/GaAs(001) [21] , tip bias of 3 V, tunnelling
current of 0.1 nA, scan speed of 250 nm/s and background pressure of 8 1 0
8
Pa. The height range of the surface
is approximately 1.0 nm.(b) is a magnifi ed image as indicated by the white open square in (a).
In order to study the imbalance of charge, two bias-alternative STM images were observed
in the identical region. Following a fi lled state ( Fig. 8.3 a), an empty-state image ( Fig. 8.3 b) was
scanned. In empty state many contrasts are shown over the surface (C
1
shown in Fig. 8.3 b,
for example). It is confi rmed that the identical regions in fi lled state of contrasts C
1
do not change
CH008-I046325.indd 273CH008-I046325.indd 273 6/25/2008 2:30:03 PM6/25/2008 2:30:03 PM
274 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
3 periodicity along []110
but have not detected a clear change perpendicular to the
[]110
direc-
tion. Since C
1
is sensitive in empty state, we might assume that C
1
originates not from contami-
nation but from a segregated Ga or Sb cluster. If a III-rich or V-rich condition is matched locally,
some reconstructed changes should arise from those concentrations [30]. On the other hand, C
2
and C
3
shown in Fig. 8.3a and 8.3b are probably contamination and lattice defect, respectively.
The presence of such contrasts may relate strongly to the charge excess on this surface, but for
detailed analysis it is assumed that fi rst-principles calculation considering defect structures is
needed. Moreover, there are still many possibilities of the structure model of this surface. We show
here only some of them. For example, a recent report describes the possibility of Sb multilayers
[31] . Finally, Sb irradiation on GaAs(001) causes high step density and has specifi c structures
such as the swinging dimers and the contrasts C
1
. InAs deposition on this surface causes drastic
changes such as wetting layer and high density of dots. However, since surplus Sb supply gives
rise to the deterioration of crystal quality [10] , close attention must be paid to the Sb amount.
In order to analyse the details of InAs deposition on this surface, in situ STM observations during
growth [17] are suitable and powerful. Therefore, its details are described in next paragraphs.
Scanning tunnelling microscopy (STM) has been applied to GaAs and related materials since
the 1980s. Somewhat more recently it has been applied to GaAs grown in vacuo by molecular
beam epitaxy (MBE). The ability to image surface reconstructions and morphology as well as
island structures and dopants has added substantially to the understanding of fundamental MBE
processes. In a typical STM–MBE experiment, the ultra-high vacuum (UHV) chambers hous-
ing the MBE system and STM system are separate. After halting the MBE growth, the sample is
moved through UHV to the STM for imaging. An early system of this type is described by Orr
et al. [32] while a more recent design is detailed by Geng et al. [33] . These instruments enable one
to interrupt growth and obtain “ snapshots ” of the growing surface. The length scales of interest
range from 10
1 0
m (surface reconstruction, dopant incorporation, alloying) to 10
7
m (islands,
domains, step-terrace structure). Of course, MBE is a dynamic process covering time scales from
seconds or tens of seconds, which are typical monolayer (ML) completion times, to the vibra-
tional periods of adatoms (10
1 3
s) which govern attempt frequencies for migration.
Two conditions are placed on the system if one wishes to faithfully image the state of the sur-
face during growth. First, the surface should not change during imaging. Adsorption of contam-
inants and MBE materials is easily limited by the UHV conditions and the isolation of the MBE
chamber from the STM system, respectively. However, the possibility of rearrangement of surface
structures during imaging remains, so that the imaged surface may not represent the as-grown
surface. MBE growth of III–V materials normally takes place at elevated surface temperatures
( 500°C for GaAs, 400°C for InAs). In most GaAs STM–MBE systems, STM imaging occurs at
room temperature, and this is suffi cient to suppress adatom migration. Furthermore, the stability
of surface structures is easily checked by repeated imaging.
Tip bias 3 V (empty state) Tip bias 3 V (filled state)
(b)(a)
C
2
C
1
C
2
C
3
C
3
Figure 8.3 Bias-alternative STM images of Sb-irradiated GaAs(001) surface at the identical position [21] .
The scan parameters are scan size of 50 nm 50 nm and tunnelling current of 0.1 nA, respectively. The tip bias
conditions are (a) a tip bias of 3 V (fi lled state) and (b) 3 V (empty state), respectively. The grey-scale height range
is approximately 1.2 nm.
CH008-I046325.indd 274CH008-I046325.indd 274 6/25/2008 2:30:04 PM6/25/2008 2:30:04 PM
GaSb/GaAs Quantum Nanostructures by Molecular Beam Epitaxy 275
The second condition relates to the process of quenching the sample from the MBE chamber.
In the case of GaAs MBE, there is normally an excess fl ux of As
2
or As
4
molecules which stabi-
lizes the surface at high temperatures. In some STM–MBE experiments, the substrate tempera-
ture is lowered at a fi xed rate while the As fl ux is also lowered. The two rates must be balanced
to avoid the surface becoming As defi cient or As rich. Surface reconstructions have been imaged
successfully at atomic resolution in this type of experiment: see, for example, Avery et al. [34]
and LaBella et al. [35] . These surfaces are fairly uniform and can be stabilized without disrupt-
ing the surface reconstruction, which can be monitored throughout using refl ection high energy
electron diffraction (RHEED). Moreover, discrete structures such as islands can be affected by the
continued annealing inherent in such a cooling protocol. A detailed cooling/growth scheme for
the STM study of GaAs islands on GaAs(001) has been presented by Yang et al. [36] .
An alternative to controlled cooling is to quench the surfaces to clean UHV as rapidly as pos-
sible. This technique has been applied to image discrete growth structures such as coexisting
reconstruction domains [37], GaAs islands [38] , and InAs quantum dots (QDs) [39] . In these
studies, the small size of the samples required for STM combined with the relatively large thermal
mass of the transfer mechanism produced an estimated cooling rate of around 50°C s
1
. The
samples could be quickly rotated away from the impinging As fl ux and the total time to remove
samples from the MBE chamber was around 5 s by rapid-quench methods [37–39] . Comparing
the sample with rapid-quenched immediately to the one allowed to anneal for 10 s at the growth
temperature of 550°C under the As
2
fl ux prior to rapid quenching, in the latter case, the number
density has dropped by nearly a factor of three with a corresponding increase in the mean island
area. This is due to adatom migration during the short anneal and demonstrates that metrics
such as the island size distribution can evolve markedly during the quench process. Note that, in
both cases, approximately 0.1 ML GaAs was deposited on to GaAs(001)(2 4) to produce an
array of ML-height islands and both the islands and substrate are (2 4) reconstructed.
It is not possible to perform true dynamic imaging by using in vacuo remote STM–MBE. The
development of a particular surface feature during growth cannot be followed since the tip can-
not normally be returned to the same location after additional MBE growth. Of course, observa-
tions of “ typical ” features and metrics such as island size distributions are still extremely useful.
In order to overcome these limitations and any uncertainties about the effects of sample quench-
ing, it is desirable to place the STM inside the MBE chamber (STMBE) and perform true in situ
imaging during MBE growth [17] . This technique presents signifi cant experimental challenges
and shows the possibility of STM observations on GaAs surfaces even under As
4
irradiation at
440°C [40]; it also shows key results for InAs–GaAs(001) heteroepitaxial growth [41] and the
fundamental processes governing QD formation [42–44] .
8.2.2 Heterointerface structure of GaSb/GaAs
In heteroepitaxy of GaSb/GaAs nanostructures, fabrication of abrupt heterointerface is
requested for many device applications. However, a broadening of the GaAs/GaSb heterointer-
face is frequently induced by surface exchange reaction between Sb and As atoms during a GaAs
growth on the GaSb layer. In particular, a large lattice mismatch of about 7% between GaSb and
GaAs enhances Sb surface segregation due to the surface exchange reaction. The segregated Sb
atoms are incorporated into the GaAs overgrown layer, and then GaAsSb alloy layer is formed at
the heterointerface. As a result, the quantum confi nement effect changes in the modifi ed GaSb/
GaAs quantum nanostructure. Here, in order to control the GaAs/GaSb heterointerface and to
design quantum energy levels, we consider a kinetic model of the Sb–As surface exchange reac-
tion [45], which is based on a kinetic model of indium surface segregation in heteroepitaxy of
InGaAs/GaAs [46] . Figure 8.4 shows a schematic diagram of surface exchange reaction between
Sb and As atoms.
The surface exchange reaction of Sb and As atoms between surface and underlying bulk layers
is described as:
[As] [Sb] [As] [Sb] (s: surface, b: bulk).
Sb bs
⇔
(8.1)
CH008-I046325.indd 275CH008-I046325.indd 275 6/25/2008 2:30:05 PM6/25/2008 2:30:05 PM
276 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
[As] and [Sb] are concentrations of As and Sb atoms. Reaction rates of both forward and reverse
exchange reactions can be respectively given by
Rv EkT
gSb Sb
exp( / )
(8.2)
and
Rv EkT
gAs As
exp( / ).
(8.3)
Here, E
Sb
and E
As
are activation energies for the surface exchange reaction. T
g
and ν are the
growth temperature of the GaAs capping layer and the vibration frequency (10
1 3
s), respec-
tively. During GaAs capping growth on the GaSb layer, a variation of Sb surface concentration in
a small time is given by the following continuous equation:
dtdtR t tR t t[Sb] ( )/ [Sb] ( )[As] ( ) [Sb] ( )[As] ( ).
s
Sb
bs
As
sb
(8.4)
Therefore, the Sb composition profi le in the GaAs/GaSb heterostructure can be calculated using
the above equations and mass conservation law of Sb and As. In this model, the growth parame-
ters of growth rate and temperature are included. However, the impinging As fl ux was neglected
because the sticking coeffi cient of As is very low on the V-group surface.
Figure 8.5 shows calculated Sb composition profi les as a function of the growth temperature
(a) and the growth rate (b) of the GaAs capping layer on a 1 monolayer (ML) thick GaSb layer.
In this calculation, E
Sb
1.68 eV and E
As
1.82 eV were used as an example. At a high growth
temperature, the Sb surface exchange reaction becomes active, and the heterointerface broadens
as shown in Fig. 8.5a . In addition, broadening of the heterointerface is similarly enhanced with
a decreasing growth rate of the GaAs capping layer. Therefore, the low temperature and high
growth rate conditions in the GaAs capping growth are requested to suppress broadening of the
GaAs/GaSb heterointerface. However, such growth conditions are not suitable for keeping high
crystal quality of the GaAs.
Next, we will consider prediction of transition energy in the broadened Ga(As)Sb/GaAs QWs,
calculated by using this kinetic model. For GaAs
1
x
Sb
x
alloys, the bowing effect of the band gap
energy is used as:
Ex x x x x
g
() () ()at K.1521 081 12 1 12...
(8.5)
R
Sb
E
Sb
E
As
Energy
Sb As
(surface) (bulk)
R
As
As Sb
(surface) (bulk)
Configuration coordinate
Figure 8.4 Schematic diagram of a surface exchange reaction model of Sb and As atoms.
CH008-I046325.indd 276CH008-I046325.indd 276 6/25/2008 2:30:05 PM6/25/2008 2:30:05 PM
GaSb/GaAs Quantum Nanostructures by Molecular Beam Epitaxy 277
Here, the band gap bowing effect in the valence band edge is represented as:
Ex x x bx x
vv
() () ()0481 116 1..
(8.6)
where b
v
is the added bowing parameter [47] . The conduction band edge is given by:
Ex x Ex
cvg
() E() ().
(8.7)
Both E
v
( x ) and E
c
( x ) in GaAs
1
x
Sb
x
are raised by compressive strain due to lattice mismatch [47] .
For example, the energy band diagrams of the broadened Ga(As)Sb/GaAs QW with and without
strain are shown in Fig. 8.6 , where b
v
of 0.67 was used tentatively. Parameters used for the cal-
culation of the band line-up of GaAs
1
x
Sb
x
are summarized in Table 8.1 . The optical transition
energy in the calculated QW structure was obtained from the difference in energy between the
conduction band edge of the GaAs and the confi ned heavy-hole level in the strained GaAs
1
x
Sb
x
QW. The bowing parameter b
v
and activation energies E
Sb
and E
As
can be estimated from a fi tting
between experimental results (photoluminescence (PL)) and calculated ones, as follows. In this
Sb composition
510
(a) Thickness (ML)
0.5
15
0
0
G.R. = 0.14 nm/s
Ts
1.0
450 °C
350 °C
400 °C
(No segregation)
Sb composition
10 nm/s
1nm/s
0.1nm/s
Ts = 470°C
G.R.
(b)
(No segregation)
5
10
Thickness (ML)
15
0.5
0
1.0
0
Figure 8.5 Sb composition profi les of GaSb(1 ML)/GaAs QWs, calculated using a kinetic model, as functions of
the growth temperature (a) and the growth rate (b) of GaAs capping layer. Activation energies of the Sb–As surface
exchange reaction, E
Sb
1.68 eV and E
As
1.82 eV, were used.
E
c
2.0
Energy (eV)
Thickness (ML)
0
4
8
12
1.0
0.5
with strain
E
v
E
HH
Without
Figure 8.6 Energy band diagrams of conduction band edge ( E
c
) and valence band edge ( E
v
) for broadened
Ga(As)Sb/GaAs QWs with strain and without strain. Bowing parameter b
v
0.67 was used in this calculation.
CH008-I046325.indd 277CH008-I046325.indd 277 6/25/2008 2:30:06 PM6/25/2008 2:30:06 PM
278 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
analysis, it was assumed that activation energy of the surface exchange reaction is kept constant
during GaAs capping growth, and a slight expansion of the QW thickness due to the strain effect
was neglected.
Figure 8.7 a shows PL experimental and calculated results of optical transition energy
obtained from 1 ML thick GaSb/GaAs QWs as a function of the growth temperature of the
GaAs capping layer. The samples were grown by MBE. For the ideal thickness of the GaSb QW
of 1 ML, the calculated optical transition energy is 1.364 eV at 12 K. When the growth tempera-
ture decreases from 470°C to 350°C, the PL peak energy shifts toward the low-energy side in
this experiment. Here, it should be noted that the PL peak energy decreases to below 1.364 eV
(ideal 1 ML thick QW) at lower than 390°C. Furthermore, when the growth temperature is
lower than 350°C, the peak energy oppositely shifts toward the high-energy side. That is, the
minimum PL peak energy of 1.33 eV is obtained at 350°C. The broadening of the GaSb/GaAs
heterointerface induces an increase in the QW width and an increase in the band gap of the
GaAsSb, due to lowering the Sb composition. The former mainly induces a red shift of the tran-
sition energy and the latter mainly induces a blue shift. Thus both effects determine the tran-
sition energy of the broadened QW. Therefore, it is predicted that the dominant effect changes
from the former to the latter at about 350°C. Consequently, in order to fi t the PL experimental
results to the calculated ones, we determine the bowing parameter ( b
v
) for the valence band
edge in GaAsSb and the activation energies ( E
Sb
and E
As
) of the surface exchange reaction in
this kinetic model. Here, when the bowing effect for the valence band becomes strong (i.e. b
v
increases), the minimum transition energy decreases. In contrast, for large E
Sb
, the minimum
transition energy appears at a higher growth temperature because of the suppression of the Sb–
surface segregation effect. In addition, the optical transition energy shifts toward the low energy
side in the higher temperature region. For large E
As
, although the growth temperature indicat-
ing the minimum transition energy is almost maintained, the optical transition energy at the
high-temperature side mainly increases. Based on these behaviours of the three parameters, we
can fi t the calculated results to the PL experimental results by changing the three parameters, as
shown in Fig. 8.7 a. In particular, the calculated minimum transition energy of 1.337 eV appears
at 350°C, similar to the experimental result. From the above fi tting, we could empirically obtain
b
v
0.67 0.06, E
Sb
1.68 0.01 eV and E
As
1.80 0.01 eV. In Fig. 8.7 a, the calculated
results, shown by the solid line, were obtained using b
v
0.67, E
Sb
1.68 eV and E
As
1.80 eV.
Table 8.1 Parameters used for calculation of band line-up of GaAs
1
x
Sb
x
by an
empirical method [47] and the model solid theory [48] .
Parameter Symbol (unit) GaAs GaSb
Lattice constant a (Å) 5.6533 6.0959
Energy gap E
g
(eV) 1.522 0.812
Elastic stiffness constant C
11
1.223 0.908
C
12
0.571 0.413
Hydrostatic deformation potential
for conduction band a
c
(eV) 7.17 6.85
for valence band a
v
(eV) 1.16 0.79
Shear deformation potential
for valence band b (eV) 1.70 2.00
Model solid theory
average valence band position E
v
, a
v
(eV) 6.92 6.25
spin-orbit split-off energy Δ (eV) 0.34 0.82
Empirical method
conduction band position E
E
c0
(eV) 1.91 1.89
valence band position E
E
v0
(eV) 0.48 1.16
Electron effective mass m
e
/ m
v
0.067 0.045
Heavy-hole effective mass m
HH
/ m
o
0.50 0.40
CH008-I046325.indd 278CH008-I046325.indd 278 6/25/2008 2:30:06 PM6/25/2008 2:30:06 PM