3 月 13 日（木）16：00－18

物理工学域セミナーのお知らせ
日時：3 月 13 日（木）16：00－18：00
場所：3F800
発表者：
北京大学・物理学部・State Key Laboratory of Artificial Microstructure and
Mesoscopic Physics
B. Shen 教授
X. Q. Wang 教授
平成 25 年度文部科学省研究大学強化促進事業研究者招へいプログラムにより，
北京大学の Shen 教授と Wang 教授を筑波大学へ招聘し，北京大学のワイドギャ
ップ半導体，パワーエレクトロニクスの研究の状況についてご講演を頂くこと
になりました．お話の内容は，北京大学のご紹介，GaN バルク基板，MOCVD およ
び MBE による III-V 族半導体薄膜（InN，GaN，AlN およびその合金）の研究につ
いてです．年度末のお忙しい中，たいへん恐縮ですが，ご出席頂ければ幸いで
す．
現在，北京大学と筑波大学，NIMS，AIST により，パワーエレクトロニクス分
野で研究を開始しようとしており，共同研究等についてもご提案頂ければ幸い
です．
世話人：上殿明良
電子メール：[email protected]
Spin transport study in GaN-based heterostructures
Bo SHEN
State Key Laboratory of Artificial Microstructure and Mesoscopic Physics, School of
Physics, Peking University, Beijing 100871, China
In this talk, the research topics and achievements on III-nitride semiconductor
materials, physics, and devices at the Research Center of Wide Bandgap
Semiconductors in Peking University will be introduced firstly and briefly, including (1)
MOCVD and MBE growth of III-nitride thin films, heterostructures and quantum wells
of high quality, (2) the defect control, transport properties, and spin properties of
GaN-based low dimensional quantum structures, (3) HVPE fabrication of stand-free
GaN substrates.
Then, it will be focused on the spin transport properties of two-dimensional
electron gas (2DEG) in AlxGa1-xN/GaN heterostructures, which have been investigated
by means of helicity-dependent photocurrent measurements at room temperature.
Spin-related photocurrent is detected under normal incidence of a circularly polarized
laser with a Gaussian distribution. On one hand, spin-polarized electrons excited by the
laser generate a diffusive spin polarization current, which leads to a vortex charge
current as a result of reciprocal spin Hall effect. On the other hand, photo-induced
spin-polarized electrons driven by a longitudinal electric field give rise to a transverse
current via anomalous Hall Effect. Both of these effects originated from the Rashba
spin-orbit coupling. By analyzing spin-related photocurrent varied with laser position,
the contributions of the two effects are differentiated and the ratio of the spin diffusion
coefficient (D) to photo-induced anomalous spin Hall mobility (µ) Ds /µ s = 0.08 V is
extracted at room temperature.
Molecular beam epitaxy of InN and InGaN
X.Q. Wang
State Key Laboratory of Artificial Microstructure and Mesoscopic Physics, School of
Physics, Peking University, Beijing 100871, China
InN has the narrowest direct bandgap of ~0.64 eV at room temperature among
III-nitrides, and the ternary alloys of InGaN cover a very wide wavelength region and
thus provide the possibility for fabricating high efficiency solar cells. All of these
devices require high quality materials, in particular InGaN alloys. However, the epitaxy
of InN and InGaN is very difficult due to the low maximum epitaxial temperature
(500-600°C) and the large lattice mismatch of InN with common buffer layers and/or
substrates (for example 11% with GaN and 25% with sapphire). More efforts are
necessary to improve the quality of InGaN in order to satisfy the requirement of devices.
In this talk, we would like to report our recent progress on growth of high quality InN
layer, InGaN layers with In composition in a range of 0 to 100% and InN/GaN short
period superlattices by molecular beam epitaxy.
We have set up a method called boundary-temperature-controlled epitaxy, where
the growth temperature of InN is controlled at its maximum one, to get high electron
mobility InN layers. The Hall-effect measurement shows a recorded directly-probed
mobility of 3010 cm2/Vs and a residual electron concentration of 1.77 cm-3 at room
temperature, which corresponds to a mobility of 3280 cm2/Vs and a residual electron
concentration of 1.47×1017 cm-3 in InN bulk layer. The enhanced electron mobility and
the reduced residual electron concentration are mainly due to the reduction of threading
dislocation density.
Then, InxGa1-xN alloys (0≤x≤1) have been grown on GaN/sapphire templates.
Growth temperature controlled epitaxy (GTCE) was proposed to modulate In
composition. The GTCE method shows two advantages over the common growth
method: (1) The growth can be kept at the highest temperature for each for InxGa1-xN
alloy with specific x. This leads to better crystalline quality and surface due to enhanced
migration of adatoms, in particular for samples with relatively large In composition. (2)
The growth is always kept under In-rich condition. This is helpful to obtain smooth
surface since the strong enhancement of adatom diffusion length can be realized
underneath a thin In layer on the growing surface. As a result, InxGa1-xN alloys (0≤x≤1)
have been achieved. The quality is relatively good and surface morphology is also fine.
Moreover, the bandgap energies of the InxGa1-xN alloys have been evaluated by optical
transmission spectroscopy, where effect of residual strain and electron concentration on
bandgap energy shift has been considered. Finally, a bowing parameter of ~1.9±0.1 eV
has been obtained by the well fitting for the In-composition dependent bandgap energy.
At last, we will report growth of InN/GaN short period superlattices, which is
proposed to work as digital alloys to further improve quality of InGaN. The designed
structure are (InN)1/(GaN)10, (InN)1/(GaN)4, (InN)1/(GaN)3, (InN)1/(GaN)2 grown on
GaN template from 5-100 periods. TEM measurements imply that the growth is
successful, whereas the growth is not so uniform due to the monolayer epitaxy.
Unfortunately, it is found that the effective bandgap of digital alloys is much larger than
the normal ternary alloys. This is basically due to the coherent growth of InN on GaN.
Optical properties of the digital alloys will be reported as well.