Professor, Department of Applied Physics, Graduate School of Engineering, Tohoku University
Development of MRAM - next-generation memory device
—Realization of tunnel magneto-resistance effect at room temperature—
Magnetic random access memories, or MRAMs, are attracting attention from researchers as nonvolatile, high-speed, low-power consumption, large-capacity, next-generation memory devices. Prof. Miyazaki was the first to realize the large tunnel magneto-resistance effect (TMR) at room temperature, paving the way for the development of MRAMs.
Application of voltage on an insulating material does not generate electric current inside the material. Applying voltage on a very thin insulator sandwiched between conductive materials, however, generates tunnel current due to the quantum effect. The electric resistance of film composed of an insulator layer sandwiched between two layers of ferromagnetic material becomes lower when the magnetization of the two ferromagnetic layers is parallel. When their magnetization is anti-parallel, the film’s resistance rises. This is the tunnel magneto-resistance effect. The principle of MRAMs is the application of this effect as a memory device by designating the higher electric resistance of such film as [1] and the lower one as [0].
The discovery of TMR was first unveiled in 1975. In experiments with Fe/Ge/Co junction film, a magneto-resistance ratio of 14% was reported. The ratio is a rate of change in electric resistance by a magnetic field. The result, however, did not draw much attention from researchers as a phenomenon with a large application potential because it was obtained at an extremely low temperature of 4.2K with a large magnetic field.
In 1988, researchers paid attention to the giant magneto-resistance effect (GMR), which was discovered in Fe/Cr artificial lattices. Prof. Miyazaki, who had been studying the anisotropic magneto-resistance effect of Permalloy (a Ni-Fe alloy) film since the mid-1980s, decided to realize TMR at room temperature, feeling sure that what would come after GMR was TMR. Most researchers around the world had already forgotten about TMR.
In June 1994, a student conducting experiments under Prof. Miyazaki’s instruction reported that the pen of a recorder had swung so widely that it had jumped off the paper. The magneto-resistance ratio of the material that the student was experimenting with suddenly jumped from a range of 2-3% up to 18%. This was the first observation of large TMR in Fe/Al2O3/Fe thin film at room temperature. The thinner the insulating Al2O3 layer, the larger the ratio, because of a rise in the tunneling probability of electrons. However, if the insulator is too thin, it forms holes that may cause short-circuits between the Fe layers.
Prof. Miyazaki now makes such film using the sputtering method. He formerly used an old vacuum deposition system thrown away by a company but had difficulty using it to make a thin uniform alumina layer. Even in the experiments in which his team observed the magneto-resistance ratio of 18%, the sample film was not of a uniform thickness due to poor performance of the system. The team aimed for producing an alumina layer with a uniform thickness of 2nm. He says, “I now presume that the Fe layers did not short-circuit incidentally and that the thickness of the alumina layer also happened to become an ideal thinness for the circulation of the tunnel current.”
His landmark achievement initially went unnoticed among most Japanese researchers. However, at the end of 1994, his oral presentation at a symposium attracted attention and, in February 1995, he was invited to a symposium in the US. “I was asked about the details of my experiments at the symposium and fully explained all of them. Now, in retrospect, I believe US researchers may have been planning to begin their own TMR projects,” he says. In March, he accepted an offer to conduct joint research with Motorola. The US company offered to shoulder all research costs and he accepted it immediately. He shared all of his research results with Motorola. Soon after the joint research, the US started a national MRAM project under the lead of IBM and Motorola. Motorola has developed a 4-megabit MRAM prototype and is leading its rivals in commercializing such chips.
Japan is now behind the US in application research of TMR although Prof. Miyazaki’s discoveries constitute the foundation in this field. “No Japanese researchers, including me, envisioned its application for developing memory devices, ” he says. Seven years later, an industry / government / academia collaboration project of the Ministry of Economy, Trade and Industry (METI) embarked on a project to commercialize MRAMs. Prof. Miyazaki has been taking the lead in the project. The Ministry of Education, Culture, Sports and Science and Technology (MEXT) in 2002 also kicked off a research project, “Universal Low Power Spin Memory Project,”in which Prof. Miyazaki has also been playing a central role in researching MRAMs.
He stresses the importance of patents in conducting research and carefully watching trends in the related fields. He says, “Finding an interesting phenomenon is not enough. Considering how it will be practically applied is crucial.” His remarks are based on his knowledge that the US has been leading Japan in the field in which he has been achieving landmark results.
Prof. Miyazaki advises young researchers to recognize the importance of concentration. As he puts it, “There is no substitute for hard work. People’s abilities do not differ greatly. I think researchers’ ability to achieve good work depends on how much they like and focus on their research.”
