HISTORY OF SQUID TECHNOLOGY IN THE UNITED STATES AND JAPAN

SQUID Technology and Markets in the United States

In 1970, not long after the discovery of the Josephson Effect and the subsequent demonstration of the first SQUIDs, the company SHE (Superconducting Helium Electronics) was founded in San Diego by Professor John Wheatley and a number of his colleagues. Its initial products were primarily dilution refrigerators, with superconducting magnets and SQUIDs being of secondary interest. The first commercial SQUID products were sold in about 1974. SHE was renamed Biomagnetic Technologies, Inc. (BTi) in 1985 and was focused on development and manufacture of magnetoencephalography (MEG) systems.

During the early 1980s, a number of SQUID companies were started in San Diego by scientists and engineers, many of whom had previously been employed by SHE. Quantum Design, which is well known for its SQUID susceptometer product, recently split off its research organization, Quantum Magnetics, as a separate company. Tristan Technologies, founded by Dr. D. Crum and his colleagues in 1991, focused on the manufacture of custom SQUID systems, particularly for nondestructive evaluation (NDE). Conductus acquired Tristan in 1993 and renamed it the Conductus Instruments and Systems Division, although in the summer of 1997 Conductus sold most of the SQUID technology and products division back to some of the original Tristan employees, who reformed the company under the same name, Tristan.1

Thus, the United States was alone in the SQUID market for over 20 years. Admittedly, the market was small, initially comprising sensors and electronics for use by physicists for low temperature research. The susceptometer of Quantum Design was the first system-level SQUID product to be sold in some volume, notably in 1987-88 after the discovery of the high Tc superconductors. BTi has sold MEG systems of increasing complexity, 7 channels, followed by 37 channels, and most recently it has developed a "whole head system" (i.e., the head is surrounded by a helmet-like Dewar containing the SQUIDs).

Much of the SQUID sensor technology of the United States, both LTS and HTS, was developed by Prof. J. Clarke of the University of California, Berkeley, and by Drs. Mark Ketchen and Roger Koch, and Prof. Fred Wellstood, who were all his students at Berkeley. At IBM, Koch and Ketchen also contributed numerous advances in their development of submarine and mine detection systems for the U.S. Navy. Prof. J. Wikswo of Vanderbilt University has been at the forefront of NDE and medical studies using commercial (but custom) LTS systems. The first HTSQUID system, which was also the first HTS SCE product of any kind, was "Mr. SQUID," introduced by Conductus in 1992. This early success has not resulted in the introduction of any further system-level HTSQUID products over the past 5 years by any U.S. firm, except for a few custom systems sold by Conductus.

SQUID Technology in Japan/the Superconducting Sensor Laboratory

The history of SQUID technology in Japan is largely the history of the Superconducting Sensor Laboratory, which was a centralized consortium (the types of consortia used for funding SCE in Japan are discussed in detail in Chapter 8) funded by MITI from 1990 to 1996.

In 1984, Dr. Masao Koyanagi and Prof. Kado at MITI's Electrotechnical Lab (ETL) began to make SQUIDs using the Nb trilayer process technology that had been developed as part of the Josephson Computer Project. Using these SQUIDs, from 1984 to 1987 they developed at ETL both a 9-channel MEG system and many of the associated technologies for MEG imaging. NbN SQUIDs operating at 11 K were developed just before the discovery of the HTS materials. In order to make these technologies available to Japanese industry and to further develop the system-level technology that would be necessary in any type of later product, the decision was made by MITI to fund a consortium for a period of six years, with the objective of demonstrating the operation of a number of MEG and other systems, culminating in a 256-channel whole-head MEG system by the end of the consortium activity in March 1996. Considering the state of the technology in Japan in 1990, this was an extremely ambitious and aggressive system-level goal. The SSL consortium used the style that had already been established at ISTEC: a new central facility was built, and the supporting companies sent researchers (almost all of whom had no previous experience with SQUIDs) to work for periods of time in this central facility. The member companies were Daikin, Hitachi, Seiko, Shimadzu, Shimizu, Sumitomo Electric, Takenaka, Toshiba, Ulvac, and Yokogawa. Prof. Kado was appointed as the Research Director of SSL, a position equivalent to Chief Technical Officer of a small company, which had 36 technical staff. Fig. 6.1 shows SSL's organization chart.


Fig. 6.1. Organization of MITI's Superconducting Sensor Laboratory (SSL).

SSL's building, located approximately halfway between Tokyo and Narita Airport, had two wings and a central reception and office area. One wing contained the class 10,000 cleanroom (class 1,000 in critical areas) for the fabrication of the Nb sensors; the other had a large laboratory for the shielded rooms and MEG systems, plus a machine shop. Most of the work within this building was on LTS sensors and systems. Meanwhile, in a part of SSL that rented from Sumitomo Electric within its Itami laboratories, Dr. Itozaki and his colleagues carried out a program to first develop HTSQUID sensors and later, working with Prof. Kado and other members of the SSL, build magnetocardiography (MCG) systems using arrays of such sensors.

Figure 6.2 shows the objectives of SSL, all of which were achieved. In fact, there is a parallel between the extraordinary progress made at SSL and that made by the four small U.S. companies Conductus, ISC, SCT, and STI with their wireless prototypes during the mid-1990s. It is remarkable how rapid progress can be once system-level objectives are defined. The facilities of SSL were not completed until October 1991, yet only 3 years later, a 256-channel system was completed and operating in a heavily shielded room designed and assembled on-site. In the same period from 1991-94, sensor fabrication was brought on-line and control electronics and software were developed; MEG and MCG systems of 16, 32, and 64 channels (completed 12/92) were built before the final 256-channel system; and small systems for earthquake monitoring were also demonstrated. In addition, since many of the scientists and engineers sent by the companies had no prior experience in SQUIDs or cryogenics, another function SSL had to assume was the education of its staff. Prof. Kado has commented that the industry staff brought essential and varied expertise in both analog and digital electronics, cryogenics, signal processing, inverse problem and computer architecture, and so on.


Fig. 6.2. Objectives of the Superconducting Sensor Laboratory, 1989-95.

The (unanticipated) educational function of SSL and other consortia in Japan is perhaps not appreciated in the United States, where collaborations or consortia occur between groups that are already expert in the science and technology. In Japan, numerous staff received education and training in SQUID technology and cryogenics at SSL and returned to their companies with this knowledge. At ISTEC, over 250 scientists from industry have been informally trained in the field of superconducting materials and characterization before returning to their companies. There is no equivalent educational function of participants from industry achieved by any consortium, in any technical field, in the United States, due in part to the typically strong U.S. focus on applications. The HTSQUID activity at SSL, partly carried out within the rented laboratory of Sumitomo Electric, resulted in the development of flux-focusing-type magnetometers. At the system level, a 32-channel MCG system was demonstrated (Fig. 6.3). When operated in a moderately shielded room, cardiograms of patients were measured by the SQUID array, as shown in Fig. 6.4. After SSL closed in March 1996, Sumitomo purchased all the facilities of SSL that were located in Sumitomo during the HTS project and started their own activities on sensors and multichannel HTS systems (see the Sumitomo site visit report in Appendix B).


 
Fig. 6.3. Sumitomo's 32-channel SQUID Fig. 6.4. Magnetocardiograms measured by the array

1Also in 1997, Biomagnetic Technologies, Inc. (BTi), and Quantum Magnetics (QM) formed a new company, Magnesensors, to focus on HTSQUIDs.
Published: July 1998; WTEC Hyper-Librarian