Site: Tohoku University
Institute for Materials Research (IMR)
Katahira 2-1-1, Aoba-ku
Sendai 980-77, Japan
Tel: (81) 022-215 2000; Fax: (81) 022-215 2002
Date Visited: 21 July 1997
WTEC: E. Hu, C. Koch, and D. Cox (report authors), L. Jelinski, M. Roco, R.W. Siegel, D. Shaw, C. Uyehara
Most of the WTEC team's visit to Tohoku University was focused on the Institute for Materials Research (IMR) which is directed by Professor Kenji Suzuki. IMR's historical roots go back to 1916; at first it was devoted to research on iron and steel. However, under the leadership of Professor Matsumoto it became a leading research laboratory in the 1970s and 1980s in the area of nonequilibrium processing techniques for producing metastable materials, in particular metallic glasses, an area in which it was a world leader. At present IMR is a large modern materials research center containing 26 research laboratories in which approximately 160 scientists, 120 technicians, 190 graduate students, and 70 visiting scientists carry out a variety of research projects. The major part of the financial support for the IMR is provided by the Ministry of Education, Science, Sports, and Culture. The specific groups in the IMR concerned with nanoparticles/nanostructured materials that the panel visited are those of Professors K. Suzuki and K. Sumiyama (metallic nanocluster assemblies), Professor Kasuya (semiconductor nanocluster assemblies), Professor H. Fujimori (magnetic nanostructured materials), Professor A. Inoue (nanostructured bulk materials), and Professor T. Yao (semiconductor nanodevices). Brief descriptions of these research efforts are given below. In addition to visiting IMR, the WTEC team visited the Tohoku University laboratory of Professor Esashi in the Department of Machine Intelligence and Systems Engineering. Professor Esashi's research on micro/nano machines is also described.
Microsystems by Silicon Machining: This program at Tohoku University, under the direction of Prof. M. Esashi, is located on the "mountain" campus of Tohoku, about a 15-minute drive from the main campus. The program has five components and is staffed by 35 professionals. The components are:
The professionals comprised postdoctoral fellows from a number of countries, resident staff, and employees of a number of companies, including Samsung, Ford, Hitachi, Asahi Optical, and Honda.
Prof. Esashi has been working in the area of microelectromechanical systems (MEMS) devices for the past 25 years and over that time built up a fairly sophisticated, albeit homemade, facility for producing MEMS devices and another facility for characterizing them. A number of biomedical devices have been produced over the years, including blood gas sensors, a 1 mm-diameter, navigable catheter based on shape-memory alloy material, produced with an STM tip, using a heated silicon substrate.
In the past several years there have been major capital investments into the program, resulting in the building and equipping of a microfabrication facility and a separate nanofabrication facility.
The microfabrication facility, also called a Venture Business Laboratory, was constructed and outfitted with government money. Companies are encouraged to make use of the facilities. The names of 27 companies, most of them large and well-known, were displayed in the entryway.
The microfabrication facility consisted of a 0.5 micron CMOS line for 2-inch wafers, and housed an impressive array of modern equipment, including He-Cd and Ar lasers for laser fabrication, and laser-induced CVD. This facility was very heavily invested in etch equipment, which included a molecular beam etch station for fairly sophisticated controlled etching. There was room after room of characterization equipment, including field emission SEM, SIMS, an SEM for characterizing biological systems under water, and an in situ monitor. There was also highly specialized equipment, including an entire room devoted to sensor characterization, which housed an accelerometer tester that worked in vacuum and a position-sensitive device that also worked in vacuum.
The nanofabrication facility, a Class-1000 cleanroom, was about a year old. Like the microfabrication facility, it was equipped with an impressive array of new and sophisticated equipment, including an e-beam direct exposure system and a stepper/pattern generator. It is conservatively estimated that the facility cost $10 million.
There was also a nanodevice characterization room, including ultrahigh vacuum AFM, STM, STEM, and ESCA. An in situ infrared ellipsometer was available for characterization of epitaxial layers.
While the size scale of the micro- and nanomachines we learned about were larger than that covered in this report, two projects bear description:
The focus of Professor Fujimori's group is on new developments in the field of magnetic materials. A major part of this effort is directed at nanocrystalline (nc) Fe-based soft magnetic materials. A major method of processing utilizes preparation of amorphous precursors by melt spinning or sputtering followed by partial crystallization of nc bcc Fe particles. The nc Fe particles provide high magnetic moments and, if < about 20 nm in diameter, exhibit small crystalline anisotropy and superior soft ferromagnetic properties. An example given is an Fe-base amorphous alloy produced by sputtering that contains Hf, Ta, C, which on crystallization gives 10 nm Fe particles whose grain boundaries are pinned by about 10 nm diameter carbide (e.g., Fe80Hf8C12) particles, which stabilize the nm grain size up to elevated temperatures (~ 700°C). This material exhibits almost constant permeability µ', with frequency up to about 10 MHz while the µ" (imaginary, lossy) component rises at values between 1 and 10 MHz.
Another interesting topic involves preparation and property measurements of metal/nonmetal granular nc composites. Sputtering was used to form systems such as Fe/SiO2 , Co/Al2O3. As the oxygen concentration in the films increases, the electrical resistivity increases from metallic (101-102 µW cm), to a r vs. T behavior with a r minimum with r = 102-103 µW cm, to high resistivity (negative coefficient of r ) with r = 104-1010 µW cm. The low oxygen content composites show ferromagnetic behavior with high permeabilities - about double that of ferrites but with higher losses. A potential application is use as a high frequency inductor.
The oxygen-rich composites exhibit giant magnetoresistance (GMR). For example, the Co, Fe in Al2O3 provides an assembly of tiny tunneling junctions with D r /r o about 7.8%.
Fujimori's magnetic materials program is expected to receive an additional $3.8 million in government funding for the next five years in a new program entitled, "Nanostructurally controlled spin dependent quantum effects and new electronics and magnetics." A goal is a one-electron spin memory.
Professor Inoue's large group (41 persons) studies amorphous, quasi-crystalline, and nc materials. The nc materials are typically prepared by crystallization of amorphous precursors, which can be bulk amorphous alloys, ribbons or wires produced by melt spinning, or rapidly quenched powders produced by powder atomization methods. Mechanical tests on nc/amorphous alloys from bulk amorphous samples exhibit increased yield strengths but little elongation in tension - that is, behavior similar to amorphous alloys even though 60-70 vol.% of the material is crystalline. Mg-base alloys have been successfully studied in this regard.
The equipment for processing, material characterization, and property measurements in Inoue's laboratory is very impressive indeed, with most nonequilibrium processing tools, sophisticated characterization, and property measurement facilities available.
The innovative work on metastable materials Professor Inoue has carried on over the years cannot be praised too highly. He and his coworkers are responsible for a new class of soft ferromagnetic materials based on Fe-Zr-B compositions which consist of nanocrystalline bcc Fe particles in an amorphous matrix. These materials exhibit the lowest core losses at power frequencies and at frequencies up to 400 Hz of any known material. They have also developed alloys based on Al or Mg, which are two-phase nanocrystalline and amorphous and which exhibit high strength, e.g., the Al alloys have more than double the strength of the strongest existing Al alloys, along with some ductility.
A major factor in his group is its close association with Japanese industry. This is illustrated by the fact that 20 of his 41-member group are employees of many companies, specially assigned to carry out research in his group.
Prof. Kenji Sumiyama and Dr. Changwu Hu gave highlights of some of the work being done in the areas of metallic nanocluster assemblies and work on fullerenes and carbon nanotubes, respectively. In the metallic nanocluster area the main focus is on developing different techniques for production of metallic clusters.
Prof. Sumiyama described five different techniques for production of metallic clusters that have been utilized by his group:
For characterization their main tools are electron microscopes, e.g., SEM, FE-TEM, and STM examinations of clusters deposited on substrates. The primary use appears to be for evaluation of the cluster size and cluster size distribution from the various cluster production techniques. Every researcher appears to be well equipped, with each possessing electron microscopes for individual use. They reported plans to add electron holography capability to the field-emission TEM next year. In addition to use as cluster size and size distribution measurements, recent STM studies of selenium clusters deposited on HOPG were interpreted as the first evidence for six or eight metal atom rings covering an HOPG surface.
The ionized cluster beam apparatus is the most developed and was an early workhorse, with several publications reporting work using this source. More recently, a new plasma sputtering apparatus has been constructed over the last four years and has become the primary cluster synthesis apparatus at IMR. Drs. Suzuki and Sumiyama reported being able to produce clusters of smaller sizes and with narrower size distributions than was possible with the ionized cluster beam apparatus and are optimistic that this source can be scaled up for much larger production than currently feasible. For example, with the plasma sputtering apparatus, they reported production of chromium clusters with sizes ranging between 3-4 nm with about 10% variation in size. Previously, the average cluster size was about 8-9 nm, but with a larger variation in size. Optimization of the helium and argon gas mixtures used in the plasma sputtering have resulted in better control of the cluster size and cluster size distribution, according to Prof. Sumiyama. The laser ablation cluster source is now being developed for production of transition metal clusters. The other two techniques, field emission and liquid metal ion source, are not as versatile, being limited to materials that have relatively low melting points.
Dr. Hu described the efforts at Tohoku in the area of fullerene and carbon nanotube production and characterization. Dr. Hu described three main items:
In (1), the chemical reaction of C60 on Si(111), C60 is deposited on Silicon(111) and then heated to 800°C. At 800°C STM shows that a monolayer of C60 in registry with the Si(111) surface is covering the surface. Upon further heating to 850°C, the C60 layer becomes disordered. Heating even further to 1100°C results in formation of SiC in the form of SiC clusters of about 50 nm diameter and 2-4 nm height. This technique is reported to be a novel low temperature route to SiC.
In (2) the fullerenes are observed by STM to polymerize upon irradiation by an Ar ion laser and form large (150 nm diameter) clusters. Additional studies showed that the STM pattern changed with changing bias voltage, suggesting some polymerization is induced by the STM electric fields.
Lastly, Dr. Hu reported recent results on characterization of nanotubes produced in the laboratory. The researchers report single-walled nanotube yields of 20-30%, and that the diameter of the single-walled nanotube depends on the metal used in their (metal catalyzed) production, e.g., nanotube radii of 0.5 nm, 0.65 nm, and 1.0 nm for Fe/Ni, Co, and La, respectively. Raman spectroscopy and STM are used to characterize the nanotube deposits. Attempts to interrogate the electronic structure have been unsuccessful thus far.
The overall picture is that this group at Tohoku is well equipped as far as the electron microscopy techniques are concerned and is primarily interested in developing larger scale production techniques for both metallic clusters and for carbon nanotubules. It was not clear what the ties to future technology development may be.
Dr. Takafumi Yao's general research goal is "to exploit new optoelectronic materials for the 21st century." The WTEC team's visit to Professor Yao's laboratories was hosted by two postdocs working with him: Dr. Elisabeth Kurtz and Dr. Darren Bagnall. Drs. Kurtz and Bagnall are two of three foreign (JSPS) postdocs working within Dr. Yao's lab. (This laboratory is apparently one of three under Professor Yao's supervision; another one is located in Tsukuba). They described projects in the nanostructured growth of wide bandgap materials: ZnO and CdSe.
Dr. Bagnall spoke of Plasma-Assisted MBE of ZnO for Blue-UV Emitters. A plasma source of oxygen was used to assist the epitaxial growth of ZnO on sapphire substrates. Free exciton emission dominates at room temperature, and pulsed, optically-excited lasing was observed up to 500 K, with fairly large thresholds: 2.5 MW/cm2. With the large degree of strain between ZnO and sapphire, it would not be surprising to observe nanostructure growth (e.g., islands). In fact, AFM traces showed evidence of pyramidal structure: this was not thought to be the basis of the excitonic emission. Bagnall hopes to grow quantum wells in this material structure.
Dr. Kurtz described the growth of self-organized quantum dots in the CdSe material system: (111)A ZnSe was grown on a GaAs substrate, and CdSe dot structures were subsequently nucleated on this surface. The (111)A surface was utilized because it provided a smooth, featureless surface. The dots had typical diameters of 47 nm, with a height/diameter ratio of 19%. There are some interesting differences in these dots compared to self-organized dots in the III-V materials, such as InAs/GaAs:
Both projects used cathodoluminescence, photoluminescence, and AFM measurements in their analysis. There was a lack of modeling effort complementing the experimental work of the group. There seemed to be access to a broad range of characterization tools within or external to the Yao group ¾ such as near field scanning optical microscopy. The WTEC team was taken for tours of some of the Yao labs, which include five MBE stations, CL, PL, X-ray analysis, and a UHV STM.
Ono, T., H. Saitoh, and M. Esashi. 1997. Si nanowire. In growth with ultrahigh vacuum scanning tunneling microscopy. Applied Physics Letters 70(14)(7 April).
Hamanaka, H., T. Ono, and M. Esashi. 1997. Fabrication of self-supported Si nano-structure with STM. In Proc. of IEEE, MEMS '97 (January), pp. 153-158.