174, Fujitsuka-cho, Hodogaya-ku
Yokohama-shi 240-0031, Japan
Tel.: (81) 453 53 6866
Fax: (81) 453 53 6905
Date Visited:11 June 1998
TTEC Attendees:
J. H. Maurice (report author)
T. P. Chow
S. DenBaars
P. M. Stipan
C. Uyehara
Hosts:
Dr. Hiroji Kawai, Chief Research Scientist, N Project
Dr. Masco Ikeda, Project Leader, N Project
Mr. Hiroshi Ooki, General Manager, Materials Research Laboratory
INTRODUCTION
Our hosts first provided a briefing on the Sony Research Center's business structure. The center has electronic components and devices groups. Structurally, there are two companies: a semiconductor company and a components and computer peripherals company. Mr. Ooki briefed the panel on the center's structure and general activities (details under that heading, below). Within the bigger picture, the focus then moved to work within three entities:
Regarding high-temperature electronics and wide bandgap work, the center is primarily a GaN facility. Discussion of its GaN work included a comparison with GaAs and a tour of the GaN (or "N") project area.
Status of Activities and Comments
Growth is the most important issue to the N Project people, indeed the main focus of work at the facilities toured. Process is next in importance. For that, the researchers particularly follow the lead of Dr. DenBaars' work. In fact, an engineer at Sony (Dr. Uchida) is currently working in Dr. DenBaars' University of California at Santa Barbara group.
The core of Sony's technical presentation was based on a GaAs and GaN comparison of key material parameters for the two semiconductor materials:
The advantages a GaN-based FET offer are as follows:
GaN offers these characteristics in a stable material with large-area substrate (sapphire) available. For transport devices, 2DEG in GaN is highly desirable since with reduced dimensionality, electron mobility becomes high. This cannot be done in SiC.
Cornell University researchers in this area (under Prof. Lester Eastman) have calculated a transit-time comparison of the two materials (GaAs vs. GaN). Dr. Kawai mentioned that Sony's results of 230 mS/mm of transconductance for GaN FETs is too low and that it is smaller than calculated. The drift velocity must be much higher, and he attributes the problem to surface roughness and dislocations, as well as poor contacts. That is, growth issues remain a central concern.
Some of the problem areas (bad features) are:
All in all, the implications of the comparison for FETs is that the piezoelectric effects in the AlGaN/GaN system's heterostructure allow for a built-in 2DEG effect, thus making a high-temperature, high-power GaN FET device both attractive and possible. Since GaN could then replace GaAs, certainly it can compare with GaAs in devices.
Among the challenges to meet for device fabrication are the following:
Is high electron mobility transfer (HEMT) the best structure for a transport device? Well, given the wide bandgap, the built-in 2DEG concentration, and the high Schottky barrier height, if no deep levels are in the AlGaN layer, then a hetero MIS-HFET is possible, and Sony proposes such. (The MISFET is not created by an inversion structure, though an inversion mode is very possible.)
Evident from plots of the conduction-band lineup in the Sony metal/AlN/n+GaN/AlGaN structure, the channel carrier decreases as the AlN thickness decreases. Using a very thin A1N layer, Sony researchers thus have succeeded in fabricating the first GaN IG-HFET. They expect AlN/GaN HFETs to be the most promising structures for HTE and high frequency devices.
Regarding blue laser work, Sony's continuous (CW) blue laser had not been made public at the time of the TTEC panel's visit; that is, no details had yet been published, and information was shared by verbal mention only. Sony projects that an optical storage device using blue lasers will be available in 2000 or 2001.
Many device structure improvements have been made in Sony's blue laser diodes. Among these are the following:
This is for a blue-green ZnMg /ZaSSe /ZnCdSe /ZnSSe /ZnMgSSe structure. In the future of next-generation optical disk systems, they must have a low power consumption, a lifetime greater than 10,000 hours, and an optical output power of approximately 30 mW. The systems must also feature a high quality laser beam.
General Activities
The Materials Research Center has 4 divisions. These are the 1) Materials Research Lab (with a staff of about 80 researchers), 2) Magnetic Recording Research Lab, 3) Frontier Science Research Lab (which includes advanced devices), and 4) Display Technology Lab. Activities within the Materials Research Center include:
At the Kuboto Laboratory, there are optoelectronic (OE), nonlinear optic (NLO), and solid-state laser (materials processing) labs. Projects at the Kuboto Lab include the aforementioned GaN work and also a ferroelectrics lab. Kuboto also includes the Center for Environmental Technology (staff of about 50), the Center for Materials Analysis (which is primarily involved with characterization), and the Center for Technical Information. As for activities within the Center for Environmental Technology, they include reclamation technology (based on polystyrene with the citrus-based oil, limonene, and flocculants from polystyrene), ecological materials (such as lead-free solder, biodegradable plastics, and cellulose-based molds, and life-cycle assessment (of energy-related materials such as solar cells). Researchers there have developed a new recycling system for the foam packaging material EPS (or expanded polystyrene) and foam molds using limonene. EPS is mostly air (with a 1/20 volume ratio). Recovered waste EPS is dissolved by limonene at the recycling plant. The limonene and high-quality polystyrene are recovered in the process.
LAB TOUR
Sony has many MOCVD reactors, all of which are constructed and maintained there. The researchers mentioned that horizontal 2-flow, in their opinion, is the superior growth method. A photoluminescence test station is used in situ, basically as a monitoring tool between growth steps of GaN epilayers. The station measures conductivity via electron mobility. An X-ray "double crystal X-ray" diffractometer measures lattice spacing and mismatch via strain.
Main Points of Discussion
Q.What is the future market for microwave power devices?
A.The most promising application Sony researchers see is industry use in base-station-to-base-station communications. This is highly desirable as digital systems are spreading and fewer stations will be needed. The high frequency range for GaN (from 2.5 GHz to 30 GHz) means that GaN-based microwave power devices could also replace vacuum tube technology. GaN has bigger consumer applications than GaAs. SiC does not have as many consumer applications. SiC may become a replacement for Si devices only. In the 20-30 GHz range, many and varied laser and LED consumer applications exist.
In lighting devices, a microwave lamp is available (from Fusion, a U.S. company, stateside). In this lamp, microwaves stimulate sulfur, and light is emitted very efficiently. Maybe microwave devices can be used for this. Currently microwaves are produced in a magnetron tube. A much smaller generating device would be a very good thing. Another consumer application lies in cellular car-to-car communications, since high levels of microwave radiation are not good for anyone. Another application is a personal uplink to space satellites, which requires high frequency and high power.
The amount of value added to a device is important in assessing consumer applications.
Q.Between the fabrication process and material growth, which is the main issue at present for HTE?
A.Both are experiencing problems.
REFERENCES
Imanaga, S., and H. Kawai. 1997. Novel AlN/GaN insulated gate heterostructure field effect transistor with modulation doping and one-dimensional simulation of charge control. Journal of Applied Physics. 82 (II):5843-5858. Reprint.
Kawai, H., M. Hara, F. Nakamura, and S. Imanaga. 1998. AlN/GaN insulated gate heterostructure FET with regrown n* GaN ohmic contact. Electronics Letters. 34 (6, 19 March):592-393. Reprint.