Nagoya University (Prof. Hayakawa)|
Department of Quantum Engineering
Nagoya 464-01, Japan
Tel: 81-52-789-3158; Fax: 3160
|Date Visited:||January 29, 1997|
|WTEC Attendees:||M. Nisenoff (report author), M. Beasley, G. Gamota, H. Morishita, F. Patten, R. Ralston, J. Rowell|
|Hosts:||Prof. H. Hayakawa|
The head of the Superconducting Device Laboratory in the Department of Quantum Engineering at Nagoya University is Prof. H. Hayakawa, who, prior to coming to the university in 1988, was the head of the Superconductivity Electronics Section at MITI's Electrotechnical Laboratory in Tsukuba. In that position, he was one of the technical leaders in the Superconducting Digital (Josephson Computer) Project, which was funded by MITI from 1981 to 1990 to develop low temperature superconducting LTS digital logic and memory technologies. Since arriving at Nagoya, Prof. Hayakawa has established one of the largest university research groups in superconducting electronics in Japan, one that is currently carrying out research in both low temperature and high temperature electronics.
The Nagoya group has four staff members including Prof. Hayakawa, an associate professor (Dr. A. Fujimaki), a lecturer (Dr. M. Inoue), an assistant professor (Dr. H. Akaike), and an engineer (K. Sawaki). There are about 20 students in the group, including 16 MS students and four PhD students. This distribution of MS and PhD students is fairly common in many Japanese universities, where most students leave after their MS degrees and join some industrial organization. The PhD degree is usually obtained after the student has had some industrial experience. Two of the four PhD students in Prof. Hayakawa's group are from Sanyo Electric and are working only part-time on their degrees.
In addition to the facilities in the Department of Quantum Engineering, Prof. Hayakawa's group has access to the Center for Cooperative Research in Science and Technology (CCRAST), which is also headed by Prof. Hayakawa and is located in a nearby building. In the Superconducting Device Laboratory, there is a small (100 square meters) cleanroom, optical lithography equipment, a high vacuum system with load lock for fabrication of niobium devices and circuits, several smaller vacuum systems, laser ablation equipment, reactive ion etching and ion mill etching equipment, atomic force microscopy (AFM), wafer dicing equipment, and selected surface analysis equipment. Because of the access to CCRAST, the group also has access to electron beam lithography, focused ion beam (FIB), ion implantation, three target sputtering systems and other equipment. Hence, the group has access to the wide variety of facilities required for state-of-the art superconducting junction fabrication and testing.
Prof. Hayakawa has established a research program exploring both low temperature and high temperature superconducting materials, devices, and circuits. The motivation for these programs was stated to be for use in high performance fiberoptic communications systems, especially for operation at frequencies greater than 10 GHz, for example, for high speed multiplexers (mux) and demultiplexers (demux). The primary advantage of superconductive electronic technology is the very low power dissipation it offers.
In the area of LTS device studies, submicrometer dimension niobium junctions with aluminum oxide barriers have been fabricated with good sub-gap leakage (Rsg /Rn ~20) and with good critical current spreads (about 14 % for 0.7 square junctions). However, as the barrier thickness is reduced to obtain higher critical current densities (a current density of about 20,000 amps per square centimeter is needed for 100 GHz operation), the sub-gap leakage current increases to a level where latching logic is not feasible. If non-latching single flux quantum (SFQ) logic is to be used for these high frequency circuits, latching (that is, hysteretic) junctions are not essential and, therefore, Prof. Hayakawa's group is exploring how to fabricate non-hysteretic niobium nitride (NbN) junctions. The use of NbN with a Tc of 16 K for the junction electrodes permits circuit operation at temperatures up to 10 K. As an alternative for the barrier material, the Nagoya group has investigated the use of plasma nitridation of the surface of the NbN base electrode, that is, plasma treating of the surface in a nitrogen gas. The electrical characteristics of the resulting tunnel junction depend on the surface treatment before the nitridation and the duration of exposure to the plasma and, to lesser extents, the nitrogen pressure during nitridation and the voltage used to sustain the plasma. The surface treatment prior to the plasma treatment determines whether the resulting device characteristics are underdamped (non-hysteretic) or overdamped (hysteretic). The conduction mechanisms in these barriers is thought to be dominated by variable range hopping mechanisms. Critical current densities near 20,000 amps per square centimeter with values of IcRn of about 1 mV (at 4 K) can be obtained for nitrogen pressures up to about 140 millitorr. Good magnetic field modulation of both types of junctions was obtained. It was pointed out that this plasma nitridation technique doesn't work for niobium technology.
The Nagoya group is also investigating new types of logic circuits. Prof. Hayakawa reiterated that the requirements for superconducting circuits for practical applications include compatibility with semiconductor logic circuits, high reliability (error-free operation), dc power supply (to minimize punch-through effects and provide simple powering system) and ultrahigh speed. The Nagoya group has developed a new latching circuit 3/4 a four junction coupled SQUID circuit that has large bias margins (( ( 50%) and fast switching speeds (about 13 psec for critical current densities of 1,000 amps per square centimeter). The circuit has been fabricated using niobium-aluminum oxide-niobium Josephson device technology and molybdenum resistors. The functionality of the circuit has been simulated, and preliminary experiential results show good behavior with a bistable output, that is, low output signal for a range of input current values and a high output signal for the remaining range of input currents, with a fairly abrupt transition between the two output states. Further work on this and other new circuit topologies is to be undertaken. Since the university does not have a highly disciplined fabrication facility, Prof. Hayakawa indicated that his group is exploring the possibility of having LTS chips fabricated for his group by NEC, which has a foundry operation with 2 micron design rules. If follow-on funding for LTS digital circuits becomes available, this joint activity with NEC may be explored.
A number of HTS materials and device projects are underway in the Superconductivity Device Laboratory at Nagoya University. In one project, researchers are fabricating bismuth films with the goal of studying "intrinsic" junctions. In this deposition system, there are four targets, one of bismuth strontium copper oxide, another of calcium copper oxide, another one of bismuth oxide, and finally, one of platinum, each with a dedicated shutter. The sample is rotated over these targets and the shutter opened for varying times to control the amount of material incident on the substrate, thus controlling the composition of the resulting film. A variety of bismuth films, 2201, 2212, 2223, and 2234, have been fabricated, but "intrinsic" Josephson junction behavior has not yet been observed.
Ramp edge junctions with yttrium-barium-copper-oxide (YBCO) electrodes have been fabricated using doped praseodymium-barium-copper-oxide (PBCO) barriers with either calcium or gallium doping. Varying the dopant species and its concentration yielded junctions with a range of electrical characteristics. The barrier electrical characteristics have been explained using variable range hopping mechanisms. Vertical junctions with calcium-doped YBCO and gallium-doped PBCO have been fabricated with resistively shunted Josephson (RSJ) characteristics at high temperatures and flux flow characteristics at low temperatures. The IcRn values for these devices were quite low, about 8 microvolts. The temperature dependence of the critical current for the Ca:PBCO barriers could be explained using the proximity effects mechanism.
The Nagoya group is also investigating a Josephson vortex flow transistor (JVFT), which has been fabricated with multiple current lines, sometimes planar in layout and also in a "stacked" configuration. Current gains of near 2 have been obtained. Work continues to improve current gain and to investigate the maximum operating frequency.
Following the formal presentation, there was an open discussion. Prof. Hayakawa indicated that the only university activities in low temperature superconducting electronics in Japan are at Nagoya and at Tokyo University (Prof. Okabe's group). In order to expand his activities and to accelerate progress, Nagoya, with the support of NEC, has applied to the Science and Technology Agency (STA) for funding in niobium-based circuits technology. Prof. Hayakawa stated that he believed continuing work on LTS circuits technology at this time is important, as HTS device technology will take many years until it matures to the point where modestly complex circuits can be fabricated. There is a strong belief that MITI will fund a new five-year program to develop HTS circuits, although the performers are not known. Furthermore, Prof. Hayakawa pointed out that there is still very little work on digital memory, with the exception of the work at NEC to reduce cell size of its existing 4 kbit design. There is no work to explore improved memory concepts.
When asked why there was so little work on microwave superconductivity in Japanese universities, Prof. Hayakawa said that many electrical engineering professors switched from microwave to fiberoptic research a number of years ago when there was a large emphasis on fiberoptic technology in industry. However, Prof. Hayakawa indicated that he would be interested in starting such work as now there is increasing interest in industry in such work.
Under Prof. Hayakawa's direction, the Superconducting Device Laboratory of the Department of Quantum Engineering at Nagoya University is investigating a very interesting combination of materials and device physics in both low temperature and high temperature superconductivity. In LTS it is investigating novel and innovative approaches to fabricating niobium and niobium nitride Josephson structures and also exploring new concepts in Josephson junction latching gates. In HTS, it is looking at HTS film growth and a variety of techniques for fabricating Superconductor-Normal Metal-Superconductor (SNS) device structures and the underlying conduction mechanisms for these structures.
Overall, the Superconducting Device Laboratory at Nagoya University is a very well rounded group, with excellent thin film fabrication and analysis equipment, a very strong senior staff, and a large group of graduate students. The research topics in both LTS and HTS material and device technologies are well selected and address state-of-the-art issues at the forefront of superconducting digital electronics technology. This group is clearly one of the strongest university groups in superconducting electronics science and technology worldwide.