Site: Nagoya Institute of Technology

Department of Electrical and Computer Engineering
Research Center for Microstructure Devices
Gokiso-cho, Showa-ku
Nagoya 466-8555, Japan
Tel.: (81) 527 35 5544
Fax: (81) 527 35 5546

Date Visited:12 June 1998

TTEC Attendees:
P. M. Stipan (report author)
T. P. Chow
S. DenBaars
C. Uyehara

Dr. Masayoshi Umeno, Professor,

    Director, Dept. of Electrical and Computer Engineering,
    Research Center for Microstructure Devices, NIT

Dr. Takashi Egawa, Associate Professor
Dr. Hiroyasu Ishikawa, Assistant
Dr. Kalaga Murali Krishna
Dr. Takashi Jimbo, Professor,
    Dept. of Environmental Technology & Urban Planning,
    Graduate School of Engineering, NIT

Dr. Tetsuo Soga, Associate Professor,
    Dept. of Environmental Technology & Urban Planning,
    Graduate School of Engineering, NIT


The Nagoya Institute of Technology (NIT) had the first MOCVD reactor in Japan. At the time of the TTEC panel's visit, it had six and claimed that Japan had about 300. In 1984, NIT was first to have GaAs on silicon. The Research Center for Microstructure Devices was founded in 1993 with its new building set up in 1997 (530 m2). Its main aim is to produce III-V compound semiconductors on silicon. It is doing basic device science, continuing to research crystal growth of GaN (grown by MOCVD), and analyzing the electrical and optical characteristics of III-V compounds. It plans to combine the advantages of silicon and gallium nitride. Researchers are confident about producing GaN on Si since they have been successful with many evaluations of GaAs on Si.

As a note, Prof. Umeno was Vice President of NIT last year, with Prof. Jimbo serving as the director of this research center. Now, Prof. Umeno has come back to head this research center.

Status of Activities

Professor Umeno wants the speed of GaN combined with the benefits of silicon. His goal is to have the material characteristics of stable oxide-high density, fast speed, a robust and lightweight large diameter wafer-at low cost. He feels this can be accomplished by 3-D structures. There are problems though with high density dislocations (>10e6 cm-2). Large residual tensile stress (2 x 10e9 dyn/cm2) leads to rapid degradation of LED or laser diodes.

But as NIT pointed out, in 1984 it was successful with GaAs grown on silicon. Oki, IBM, MIT, and the University of Illinois (to name a few) have used this technique. NIT's GaAs on a Si solar cell was 22.1% efficient. Researchers are working towards 40% efficiency with a monolithic tandem solar cell.

Dr. Ishikawa described an oil research company that is interested in a 200° C high pressure light emitting and detecting device. NIT is working with GaN material, but the main problems are the buffer layer and the soft oxide. The bow of the wafer needs to be reduced by selective epitaxial growth.

Dr. Egawa discussed GaN on silicon. He is evaluating GaN on sapphire. But Si is less costly so it is more desirable, especially since NIT's work with GaAs is on Si. NIT has InGaN/AlGaN laser diodes. Dr. Egawa described the sources for various elements: Ga, Al, In, N, Si, Mg, and Zn. A graph showed GaN versus Mg flow rate. Maximum hole concentration is 1.5 x 10e 18 /cm3 with hole mobility of 6 cm2/Vs.

He then described an aging test under DC current. It is used as a laser life test.

A GaN MESFET has been developed on sapphire. It has mesa etching (Reactive Ion Etching used - BCl3). The Schottky contact was Pt/Ti/Au (10/40/100 nm). S=1.77, h =1.04 and F b (Schottky barrier height)=0.89eV.

AlGaN layer = 0.2m m

GaN layer = 2.0 m m

Sapphire layer

When the Al concentration is increased, researchers observe cracks. But they do not see any polarization dependency on the laser diode.

In the very near future, NIT will try SiC-or diamond. Its visiting scientist from India, Dr. Krishna, presented the following information on amorphous carbon for electrical applications. He has started carbon on a silicon heterojunction and achieved 2.1% efficiency. At first, he experienced a problem with thick carbon peeling off. He moved from sputter to pulsed with good results and will present his findings at a Vienna conference. He is now researching diamond on silicon.

Table Nagoya.1

Research Field and Some Equipment From Lab Tour

Field of Research

Main Equipment

Atomic layer epitaxy

Atomic layer epitaxial equipment

Physics of microstructure materials

Atomic force microscope

Optoelectronic integrated circuits

Fabrication system of nanostructure electron device

3-D integrated circuits

Deep level transient spectroscopy excimer laser


Reactive ion etcher

Precision machining for crystalline materials and ceramics

Film-thickness measuring system

Planarization polishing for wafers

Chemical-mechanical polishing machine


Jeol transmission electron microscope


Hitachi scanning electron microscope


Seiko focused ion beam

NIT researchers believe that the big problem with GaN on Si is material epitaxy. They are focusing on MOCVD processing. First, they need to grow high quality material; then they want to build devices. They are planning a blue laser (AlGaN using LPCVD is thought to be the best).

MESFET N-type ohmic contact only is no problem. The high temperature behavior of the Schottky contact is difficult. NIT has observed hysteresis curve MOS on GaN. It used SiN and SiO2 by plasma and e-beam. The plasma is the best method because the e-beam-deposited SiO2 was very porous. NIT has not observed any e-beam damage though. Its experiences show that defects cause GaAs on Si to degrade rapidly. InGaN works longer than GaAs on Si even when defect densities are the same.

Professor Jimbo presented information on blue light. Originally, researchers thought that only the II-VI compounds could emit blue light, but this is not the case.


Nagoya Institute of Technology. 1997. Overview of Institute. Brochure.

NIT. Research Center for Micro-Structure Devices. (in Japanese) On processes and equipment. Pamphlet.

Technical Report at Research Center for Micro-Structure Devices. 1998. Volume 5, March. Booklet of activities from last year.