Overview of crystallography in Japan (1913 – 1980) 

Modern crystallography in Japan started right after Laue’s discovery of diffraction of X-rays by crystals (1912), and many valuable achievements were realized during those early years. To our regret, most of them were not recognized outside of Japan. This article will present an outline of crystallography in Japan, especially crystal structure determination, from its advent to approximately 1980. The reviewer apologizes that he cannot fully review research in physics and mineralogy since he is a chemist.

The early years of Crystallography to about 1945

In 1913, T. Terada tried X-ray experiments using a rock salt crystal and a fluorescent screen in a dark room. He observed several diffraction spots and that they moved according to the movement of the crystal. He explained this phenomenon as reflection of the incident X-ray beam from the atomic net planes in the crystal (Nature 91, 135 (1913)). His work paralleled that reported by Bragg (Nature 91, 557 (1913)).
S. Nishikawa, stimulated by Terada’s experiment, began X-ray diffraction studies with Pt white X-rays of (i) fibrous materials such as cotton, silk, and asbestos, (ii) minerals such as mica and talc, and (iii) crystalline powders of fluorite and zinc blende. He found that materials of group (i) were aggregates of micro crystals having a common (fiber) axis, group (ii) were stacked aggregates of thin leafy crystals, and group (iii) were randomly oriented powdered crystals (Proc. Math. Phys. Soc. Tokyo, 7, 131 (1913)). He was a pioneer in the area of X-ray studies of structures of complex compounds. Nishikawa (1915) then studied the structure of spinel, using Laue photographs with space group theory. This was the first case in which space group theory was applied to crystal structure determination. Nishikawa and K. Matsukawa (1928) verified, by experiments on zinc blende, the breakdown of Friedel’s law for crystals containing atoms which anomalously dispersed incident X-rays. This work was reported two years earlier than that of Coster, Knol, and Prins (1930). In 1922, Nishikawa organized a crystallography research group that used both X-ray and electron diffraction in the Institute for Physical and Chemical Research (Tokyo). I. Nitta, a member of the group, studied the crystal structure of pentaerythritol C(CH₂OH)₄ to establish tetrahedral carbon valence bonds. At that time, however, there was a report based on a space group assignment that the central C atom of pentaerythritol in the crystal was (tetragonal) pyramidal. Nitta (1926) found that the reported space group was erroneous and the C atom was actually tetrahedral. This was later verified by a full structure determination, and the tetrahedral carbon valence bonds were established. Nitta served as Vice President of IUCr from 1963 to 1969.
S. Kikuchi, a member of Nishikawa’s group, began research on electron diffraction right after the experiments by Davisson, Germer and Thomson in 1927. Kikuchi then succeeded in doing high-energy electron diffraction (HEED) studies on single crystals of mica (1928). He observed several characteristic features of HEED, now known as the Kikuchi pattern. Several crystallographers in this group, S. Miyake, R. Uyeda, N. Kato and others studied the diffraction physics of both X-rays and electrons.
In 1924 T. Ito, who studied crystallography with P. Niggli and later with W. L. Bragg, organized a research group on mineralogical crystallography at the Univ. of Tokyo. In 1933, Nitta organized a crystallographic research group at Osaka Univ.

(1945 – 1980)

Diffraction physics : N. Kato experimentally and theoretically studied X-ray dynamical diffraction, section topographs and secondary extinction. He was President of IUCr (1978~1981), and a winner of the IUCr Ewald Prize. K. Kohra was the principal crystallographer working on the construction of the Tsukuba Photon Factory, which is still one of the most productive SR sources in the world.

Mineralogical crystallography: from 1945 – 1955 T. Ito and his collaborators, R. Sadanaga, Y. Takeuchi, N. Morimoto and others, determined crystal structures of important minerals. Ito developed a powerful indexing method for X-ray powder diffraction patterns, the Ito method. During this period, Ito established the general idea of ‘cell twin’ or ‘Ito twin’ for ‘repeated twinning of unit cell level’. This twin was quite different from the macroscopic one. These results were published in X-ray Studies on Polymorphism (Maruzen, 1950).

Sadanaga and others systematized Ito’s theory of the twin space group. Takeuchi studied the crystal structures of minerals under high pressure and/or high temperature. Morimoto, M. Tokonami and Y. Koto studied structures, superstructures and phase-transitions of many minerals.

Instruments: One of the first large computers built in Japan, comparable to IBM systems, that had enough capacity for computing in crystal structure determination came into operation in 1967. A program system UNICS (universal crystallographic computing system) consisting of fundamental programs for structure determination was written by T. Sakurai, T. Ashida and others (1967), and used by many crystallographers. The first Japanese computer-controlled four-circle diffractometer was produced in 1968 by Rigaku. Then the precision as well as the speed of X-ray single crystal structure determination in Japan improved dramatically. Tsukuba photon factory 2.5GeV came into operation in 1983. It provided a large step forward for the development of crystallography in Japan which saw an ensuing growth in the number of researchers involved in structure determination.

Anomalous dispersion: Using anomalous dispersion Y. Saito and collaborators (1954) were among the first to determine absolute configuration with their work on the complex ion [Co(en)₃]³⁺. Y. Okaya, Saito and R. Pepinsky proposed the theoretically attractive Ps(u) function. Okaya and Pepinsky formulated the relationship between structure factor phases and anomalous dispersion terms. This relationship is used in the phasing method which is now most widely applied for protein crystallography.

Electron density distribution: Y. Tomiie et al. (1958) studied precise electron density distribution in diformylhydrazine and obtained a reasonable coincidence with the one calculated by the MO method. He also found the electron density of a lone pair of electrons belonging to the O atom. Saito, F. Marumo and others studied precise 3d electron distribution in some transition metal ions and compared their results with theoretical predictions. Saito, H. Iwasaki and others studied electron density distributions in charge transfer complexes.

Important compounds: Crystal structures of many important compounds, such as natural organic compounds, coordination compounds, organometallics (A. Shimada, N. Kasai and N. Yasuoka), charge transfer complexes, synthetic polymers (H. Tadokoro), oligopeptides (Ashida), etc, have been studied. Many crystallographers are engaged in research in this category, but only a few of them could be presented here for want of space.

Crystalline phase reactions : In 1977, Y. Ohashi and Y. Sasada observed that racemization occurred when crystals of an (S-α-methylbenzylamine)cobaloxime complex were exposed to X-rays. This reaction could be traced by observing changes in electron density with exposure time. Similar reactions were found in the crystals of related compounds and were extensively studied. This research exploited a new attractive research field in chemical crystallography.

Protein crystals: M. Kakudo, Ashida, N. Tanaka and others began protein crystallography in 1965, and reported a 2.3 Å resolution structure of bonito ferrocytochrome c at the IUCr meeting in 1972. Most of the MIR programs necessary were prepared by Ashida. This was the first high-resolution analysis of a protein structure in Japan. In an international collaboration involving the staff of Dorothy Hodgkin’s laboratory in England, N. Sakabe and others studied 2Zn-insulin at 3.1 Å resolution in 1972 and 1.1 Å resolution in 1984. Y. Iitaka and Y. Mitsui and others studied 2.6 Å resolution subtilisin inhibitor in 1979. Katsube’s group reported the structure of ferredoxin in 1981. From that point on protein crystallography in Japan developed rapidly in both quality and quantity.

Other biological materials: DNA, RNA and related compounds were studied by Y. Mitsui and Iitaka, Sasada’s group and K. Tomita’s group. Bacterial cilia and flagella, muscle, biomembranes, etc have been studied. Among many biophysicists who engaged in the research of these material were T. Mitsui’s group including T. Ueki, and K. Wakabayashi.

Crystallographic community

Japan joined the IUCr in 1951 with its adhering body being the National Committee for Crystallography (NCC-Japan), which adheres to the Science Council of Japan (SCJ). The Crystallographic Society of Japan (CrSJ) was organized in 1950 as the electorate for the national committee. The number of members is now about 1,300. CrSJ has held annual meetings since 1973, and has published the Journal of the Crystallographic Society of Japan (bimonthly) since 1960.

The Japanese crystallographic community has hosted several successful international meetings including the IXth Congress and General Assembly of the IUCr, Kyoto (1972), the 7^th Sagamore Conference, Nikko (1982) and the International Summer School on Crystallographic Computing, Kyoto (1983)

CrSJ and NCC-Japan jointly celebrated the 50th anniversary of crystallography in Japan in Tokyo on July 1, 2000.

Tamaichi Ashida
(ashida-t@mpd.biglobe.ne.jp)

Activities in big facilities

SPring-8

SPring-8, the largest third-generation synchrotron radiation facility in the world, provides the most powerful synchrotron radiation currently available. The Japanese Atomic Energy Research Institute (JAERI) and RIKEN (The Institute of Physical and Chemical Research) began constructing SPring-8 (Super Photon ring-8GeV) in 1991 with support from Hyogo Prefecture, universities, research institutes and industries, and opened the facility in October, 1997.

Since its completion, all aspects of its operation have been conducted by the Japan Synchrotron Radiation Research Institute (JASRI), which was designated as the sole institute for the management, operation and development of SPring-8. SPring-8 consists of a 1GeV linear accelerator and 8GeV booster synchrotron that generates an incident electron beam of 8GeV and a storage ring (electron energy, 8GeV; current, 100mA; circumference, 1436m) that holds the generated electrons. Figure 3-1-1 shows the Bird’s-eye view of SPring-8.

There are two types of light sources in SPring-8, insertion devices and bending magnets. Insertion devices are either undulators or wigglers. SPring-8 beam ports include insertion device beamlines with a 6m straight section (max. 34, existing 26), long insertion device beamlines with a 30m long straight section (max. 4, existing 1) and bending magnet beamlines (max. 24, existing 21). In-vacuum type undulators developed at SPring-8 have magnet arrays sealed in a vacuum chamber. The arrangement results in a smaller gap between arrays, allowing radiation with shorter wavelength and higher power to be generated.

There is an in-vacuum revolver variable polarization undulator, an in-vacuum figure-8 undulator, a twin helical undulator, a tandem vertical undulator, an elliptical wiggler, and others that generate a variety of polarized radiation. The characteristic photon energy of a bending magnet is 28.9 keV.

62 beamlines are available. 46 are in operation and 2 are under construction. The beamlines are classified as public beamlines, contract beamlines, JAERI/RIKEN beamlines, R&D beamlines, and beamlines for accelerator diagnosis.

Two beamlines, whose lengths are 225m, are already operational at the biomedical imaging center. Two beamlines, longer than 100m, have been developed at the RI laboratory for research on radioactive isotopes and actinide materials. A 1 km-long beamline is used for research on advanced coherent X-ray optics.

The top-up operation was started successfully at SPring-8 in May, 2004. The electron beam is injected at short intervals during user beamtime, so that the current stored in the storage ring is kept constant. The top-up operation has a number of innovative features such as (1) a perturbation-free beam injection scheme, (2) minimum injection beam loss, and (3) a high purity singlebunch beam. This “ideal top-up operation” is used by many other light source facilities around the world. The top-up operation offers the following benefits to users: (1) time averaged brilliance is increased by a factor of two, (2) improvement of thermal stability of the X-ray optics, and (3) a new filling pattern with a high bunch current that was not available in more conventional operations due to the extremely short beam life-time.

SPring-8 invites proprietary and non-proprietary research proposals in June and November and does not charge users for beam time if their research is non-proprietary. The annual operation time of Spring-8 is around 5400 hours, with more than 4200 hours available for experiments. More than 1000 research subjects are accepted annually by SPring-8, and the number of researchers involved exceeds 10000.

Hideo Ohno, SPring-8

Photon factory

The Photon Factory was the first synchrotron radiation research facility for X-rays in Japan. It operates two electron storage rings, shown in Table 1. One is a 2.5 GeV ring that has been operational since 1982. It’s emittance was 400 nm⋅rad in the beginning, but was reduced to 36 nm⋅rad with upgrades in 1986 and 1997. The other is a single bunch operated 6.5 GeV electron storage ring which had been used parasitically since 1986 and then in dedicated mode for synchrotron radiation research since 1997. This ring has a unique feature that 100 ps wide X-ray pulses are always obtained with a repeat period of 1.24 micro seconds. There are currently 7 and 4 insertion device beamlines on the 2.5 GeV and 6.5 GeV rings, respectively. The BL-2 of the 2.5 GeV ring is a soft X-ray undulator beamline which was the first in Japan and the second in the world. There are two in-vacuum undulator beamlines on the 6.5 GeV ring one of which is the first in-vacuum undulator beamline in the world and has been used extensively for the study of nuclear Bragg scattering, which makes the best use of pulsed X-Rays. A new in-vacuum X-ray undulator beamline is under construction on the 6.5 GeV ring for the study of photo-induced phase transitions in highly correlated materials.
There are 32 X-ray experimental stations out of 58 on the 2.5 GeV ring (energy ranging from 2.2 to 100 keV) and 8 X-ray experimental stations out of 9 on the 6.5 GeV rings (energy ranging from 5 to160 keV). Among those 40 X-ray stations, 22 are for non-biological diffraction/scattering experiments, 5 for protein crystal structure analysis, 2 for small angle scattering, 8 for X-ray spectroscopy, 2 for X-ray imaging and 2 for others. There are 650-700 active experimental proposals every year and approximately 2,700 users repeatedly visit the Photon Factory to carry out their experiments. Eighty to eighty-five percent of proposals are for X-ray experiments, the others are for VUV/soft X-ray experiments. Since the inauguration of the facility in 1982, a number of contributions have been made at the Photon Factory. Those are, for example, MAD analysis of protein crystal structures using synchrotron radiation data, extensive use of image plates for X-ray diffraction experiments, development and extensive use of multi-anvil cells for structure studies under high pressure and high temperature, development of an X-ray Fresnel lens, experimental studies of resonant magnetic X-ray scattering, development of a quarter wave X-ray phase plate using a Bragg-transmitted beam, use of anomalous scattering methods for the study of orbital order in crystals, the development of an in-vacuum undulator and its extensive use for the study of nuclear Bragg scattering, and extensive use of a high energy X-ray elliptical wiggler for magnetic Compton scattering.
To meet requests for more experimental opportunities on insertion device beamlines, the lattice of the 2.5 GeV ring will be modified in the summer of 2005 to increase the number of straight sections for insertion devices. The modifications place new quadrupole magnets of shorter length with higher field gradients closer to neighboring bending magnets to lengthen the existing straight sections or to create new straight sections as shown in Fig. 3-2-1. Four short straight sections of 1.5 m will be created. Furthermore, two 5 m long straight sections will lengthened to 9.2 m, and 8 other sections of 3.5 m will become more than 5 m. Except for one straight section used forelectron beam injection, a total of 13 straight sections will be available for insertion devices on the 2.5 GeV ring. Such an improvement will enable us to install longer undulators with planar or helical magnet configurations and minipole undulators with a narrower gap at the short straight sections. By adding 5 straight sections on the 6.5 GeV ring, the Photon Factory will have a total of 18 straight sections for insertion devices. An in-vacuum short period undulator beamline for structural biology is now being constructed and will be commissioned in the fall of 2005. Another in-vacuum undulator Xray beamline for materials science will be constructed and commissioned in 2006.

Tadashi Matsushita
(tadashi.matsushita@kek.jp)

Neutron facilities

Japan started atomic energy related activities well behind Western countries after World War II. In the mid 1950s the first nuclear research reactor JRR-1 (Japan Research Reactor No.1, 50kW, now preserved in a museum in Tokai) built by the Japan Atomic Energy Research Institute (JAERI) was used for neutron beam research. Neutron scattering research requiring an MW-class reactor started in the early 1960s when JAERI built the JRR-2, which could produce a thermal power of 10MW, thermal neutron flux or 2x1014 neutrons/cm2/sec. There were 9 neutron scattering instruments during its most active period. It was the first neutron source available for inelastic experiments and also played a central role in training many researchers. It was permanently shut down in 1996, as scheduled, after about 35-years of service. Side-by-side with JRR-2, the JRR-3 (10MW, 1x1014 n/cm2/sec, 6 instruments) was built with purely domestic technology for the first time. In the mid 1960s Kyoto University built its 5MW reactor named KUR (3x1013n.cm2/sec, 8 instruments; www-j.rri.kyoto-u.ac.jp/) in Kumatori. It is the only MW-class reactor owned by a university. In late 1960s and 1970s, therefore, three medium-size reactors were available for neutron scattering and other beam experiments while several high flux reactors became operational abroad such as the HFBR at Brookhaven National Laboratory (60MW, 8x1014, USA), HFIR at Oak Ridge National Laboratory (85MW, 1x1015, USA), and HFR at the Institute Laue Langevin (58MW, 3x1015, France). Japanese user’s strong demand for higher neutron flux by the early 1980s culminated in the refurbishment of the JRR-3 to double its thermal power (20MW, 3x1014) and install a new cold neutron source accompanied with a guide hall building. This refurbished reactor renamed as JRR-3M became fully operational in 1990. Currently it supports 25 neutron scattering, 2 neutron radiography and 2 prompt γ-ray analysis instruments and a total number of about 12,000 users per day (www.issp.u-tokyo.ac.jp/labs/neutron/index.html and www.jaeri.go.jp/).

Japan is one of pioneers in the field of accelerator-based neutron sources. In the mid 1960s the electron linear accelerator at Tohoku University in Sendai, dedicated to nuclear studies, was tested as an injector to a metal target to produce neutrons. It was then successfully used as a pulse neutron source for neutron diffraction experiments even after a new spallation neutron source was built at KEK (the present High Energy Accelerator Research Organization; http://kek.jp/) in Tsukuba in 1980. This neutron source, named KENS, was the first dedicated pulse neutron facility in the world and it has been operated at a proton beam power of 4kW for a total of 17 instruments and a cold source. Several more powerful sources were built abroad such as IPNS at Argonne National Laboratory (6kW, 1981, USA), LANSCE at Los Alamos National Laboratory (80kW, 1985, USA), and ISIS at Rutherford Appleton Laboratory (160kW, 1985, UK). The Japanese community demanded a much stronger pulse neutron source to follow post-KENS. KEK belonging to the Ministry of Education, Science and Culture (MONBUSHO) planned the Hadron Project consisting of four major facilities based on an intense proton accelerator (0.6MW) to be used for neutron scattering, muon science, nuclear and high energy physics. JAERI belonging to the Science and Technology Agency (STA) also planned a MWclass proton accelerator that would provide intense proton beams to two facilities for neutron scattering and R&D for nuclear transmutation. Both were very big projects costing nearly one billion US dollar each. When MONBUSHO and STA merged, it was recommended that both projects merge and be promoted jointly by JAERI and KEK. In 2001 the Japanese Government funded it for the following 6 years (later extended to a 7-year project in 2003). Fig. 3-3-1 displays a layout of the facility, now called J-PARC (Japan Proton Accelerator Research Complex; http://j-parc.jp/), which consists of 5 major facilities based on an intense proton accelerator (1MW). According to the plan, proton beams are accelerated by Linac up to 400Mev and transported into the 3GeV rapid cycling synchrotron (3GeV PS in figure) further accelerating up to 3GeV. Most of the beams are transported to the Materials and Life Science Facility (3GeV PS Experimental Area) where both neutrons and muons are produced. This neutron beam facility, tentatively called JSNS (Japan Spallation Neutron Source, 1MW), will have a total of 23 beam lines at which instruments will be installed. A small portion of the proton beams from the 3GeV PS is fed into the 50GeV synchrotron (50GeV PS) for nuclear physics and for long base-line neutrino experiments to be combined with the Super KAMIOKANDE. In the near future, a superconducting Linac will be installed at the end of the 400MeV Linac to further accelerate proton beams up to 600MeV for R&D study of nuclear transmutation.
Fig. 3-3-2 is an aerial photograph of the construction site of the J-PARC taken in February 2004, the fourth year of the 7-year project. One can see both the JRR-3M 20MW reactor-based steady source and the JSNS 1MW accelerator-based pulse source located within 800m of each other on the same campus facing the Pacific Ocean. This co-location of both neutron sources will be a great advantage for neutron science and technology and they will be available to users worldwide. The first neutron beams from JSNS are scheduled in November 2007 and a users program will start in April 2008.

The Japanese neutron community has been dominated by solid state physicists at the forefront of solid state physics and material science particularly in the field of low-dimensional magnetic and strongly-correlated electron systems including high-Tc superconductors and CMR materials. After cold neutrons became available from JRR- 3M in 1990, however, soft-matter scientists and biologists started using the domestic neutron facilities and now account for as many as one quarter of the total proposals. Another innovation in neutron crystallography is a neutron-sensitive imaging plate (IP) specially developed by the JAERI group with Fuji Film Co., which has opened up a new era of structural biology. Another important instrumental development is the wide variety of neutron optical devices such as the magnetic lens, compound prism, and supermirrors cooperatively developed by RIKEN, JAERI and KUR Groups (http://nop.riken.go.jp/indexJ.html). These devices combined with a new neutron source will lead to new instruments to reach an unexplored region of momentum-energy-spin space for promoting new fields of crystallography.

Yasuhiko Fujii, JAERI
fujii@neutrons.tokai.jaeri.go.jp

Recent developments in biological crystallography

In 1913, one year after Laue’s discovery, Nishikawa and Ono of the University of Tokyo successfully observed diffraction patterns from plant and animal fibers. Forty-six years later, M. Kakudo of IPR, founded a laboratory for protein crystallography at Osaka University. This group started to crystallize cytochrome c in 1962. After developing automatic four circle diffractometers and computing programs, they determined the structure of bonito ferrocytochrome c at 2.3 Å resolution in 1973. With the support of the Crystallographic Society of Japan, M. Kakudo and N. Yasuoka established the Research Center for Crystallography at IPR in 1978 to promote protein crystallography in Japan. In the 1970s, two additional groups were active in protein crystallography. N. Sakabe (Nagoya University) and Y. Mitsui (University of Tokyo) were working on insulin and subtilisin inhibitor protein, respectively. Another protein structure determined during the 1970s was a bacterial protease inhibitor (1977), whose structure was solved by Mitsui’s group. Many protein crystallographers were trained in these three groups. In the early 1980s two notable structures, [2Fe-2S] ferredoxin at 2.8 Å (1980) and RNase ST at 2.5 Å (1982), were reported by Kakudo’s group and by Mitsui’s group, respectively. The structures of Taka-amylase A at 3.0 Å (1980), rice ferricytochrome c at 2.0 Å (1983), cytochrome c3 at 1.8 Å (1984), a complex of subtilisin and its inhibitor protein at 2.6 Å (1984), Bowman-Birk type protease inhibitor at 3.0 Å (1986), [4Fe-4S] ferredoxin at 2.3 Å (1988), aspartate aminotransferase at 2.8 Å (1988), and omega-amino acid: pyruvate aminotransferase at 2.0 Å (1989) were determined during the 1980s.
In the 1990s, new research groups studying protein crystallography were organized by a younger generation of scientists both inside and outside of universities. The number of protein structures determined at high resolution has gradually increased. Eight membrane protein structures have been determined at high resolution using X-ray methods in Japan. In 1995, following an almost 20 year collaboration, T. Tsukihara’s group at Osaka University and S. Yoshikawa’s group at Himeji Institute of Technology succeeded in the structure determination of a bovine respiratory membrane protein complex, cytochrome c oxidase, consisting of two copies of 13 different subunits (Fig. 2-2-1). This structure, the first membrane protein structure determined from mammalian cells, was a marked breakthrough not only in crystallographic studies of membrane proteins, but also in the field of bioenergetics. Upon examination of the structures of the enzyme complex in different reaction states, these groups proposed a new theory of the proton pumping mechanism (1998, 2003). C. Toyoshima (University of Tokyo) determined the structure of the sarcoplasmic reticulum calcium pump in 2000 (Fig. 2-2-2). He successfully demonstrated the mechanism of calcium pumping by serial structural analyses of reaction intermediates (2002, 2004). T. Okada’s endeavor to crystallize a G proteincoupled receptor resulted in successful structure determination by a collaboration of SPring-8 group lead by M. Miyano and R. E. Stenkamp’s group of University of Washington in 2000. Collaborating with A. Yamaguchi (Osaka University), three young crystallographers, M. Murakami, S. Nakashima and E. Yamashita, determined the structure of the bacterial multidrug efflux transporter AcrB in 2002. The structure of rat monoamine oxidase A (2004) was the first structure of a membrane protein with an isolated single transmembrane helix. The structures of three other membrane proteins, bacteriorhodopsin by Kouyama (Nagoya University, 1999, 2004), a bacterial photoreaction center by K. Miki (Kyoto University, 2001), and photosystem II by N. Kamiya (RIKEN, 2003) have been determined. Y. Fujiyoshi (Kyoto University) has been working on the technique of cryo-electron microscopy since the 1980s. Collaborating with scientists around the world, he has determined the structures of a number of physiologically important membrane proteins and viruses, including influenza A virus (1994), a plant light-harvesting complex (1994), bacteriorhodopsin (1999), nicotinic acetylcholine receptor (1999, 2003), aquaporin 1 (2000), a voltage-sensitive sodium channel (2001), and others.
K. Namba (Osaka University) elucidated the mechanism of bacterial flagellar assembly and function by combining cryo-electron and X-ray structures (1995-2004) (Fig. 2-2-3). A. Nakagawa elucidated the structural organization of a double-shelled RNA virus based on the structure of Rice Dwarf Virus at 3.5 Å resolution (2003). Many other physiologically important protein structures have been determined in Japan since the mid 1990s. Outstanding structural studies on recombination, replication, transcription, and translation have been performed by K. Morikawa (BERI, 1999-2004) and T. Hakoshima (Nara IST, 2000), S. Yokoyama (RIKEN and University of Tokyo, 1995-2004), and I. Tanaka (Hokkaido University, 1997, 1999). After setting up a Structural Biology Center in the Photon Factory, S. Wakatsuki has made significant progress in understanding the structural biology of lysosomal protein transport (2002, 2003).
In 2002, the Japanese government started a large structural genomics program, “Protein 3000.” In addition to the highthroughput structural genomics approach led by S. Yokoyama (RIKEN), eight target oriented structural genomics projects led by I. Tanaka (Hokkaido University), S. Wakatsuki (Photon Factory), M. Tanokura (University of Tokyo), K. Miki (Kyoto University), A. Nakagawa (Osaka University), and others are included in Protein 3000.

N. Sakabe has been developing IP diffractometers at the Photon Factory since the early 1980s (Fig. 2-2-4). Before SPring-8 began operations, all protein crystallographers were dependent on him for the collection of intensity data. Many Japanese and foreign crystallographers determined a number of novel crystal structures using his detectors at the Photon Factory. The newest version, which has a high quality and high speed detector, was installed at BL-6C in the Photon Factory. S. Wakatsuki has built a new user friendly undulator beamline equipped with a CCD detector. The Photon Factory has one undulator and three bending magnet beamlines and SPring-8 has four undulator and six bending magnet beamlines for protein crystallography. Each beamline has specific features, requiring separate network systems for access. In addition to X-ray beamline facilties, the Japanese Atomic Energy Institute has a neutron beamline facility. A neutron diffractometer dedicated to protein crystals has been developed by Niimura’s group (2004).

Tomitake Tsukihara
(tsuki@protein.osaka-u.ac.jp)

CRYSTALLOGRAPHY IN JAPAN

This issue completes the series of articles describing crytallography in Japan that started in Vol 13 No 2. Our thanks to Yuji Ohashi for assembling the information.

Recent developments

Chemical crystallography

In Japan, many chemical crystallographers are doing forefront research in solid-state reaction-mechanisms, supra-molecular chemistry, phase transition, and materials science. Multi-dimensional structural chemistry which depends on time, temperature, and humidity etc. is actively investigated together with the development of equipment with two-dimensional detectors.
At the Tokyo Inst. of Technology, Y. Ohashi, H. Uekusa and their coworkers developed crystalline-state reaction, which means the reaction occurs in a single crystal without degradation of crystallinity. Any stage of the reaction process, therefore, can be observed by X-ray crystal structure analysis and the reaction mechanism can be analyzed using the intermediate structures as eyewitness accounts. They defined the reaction cavity, which explains quantitative reactivity in solid-state reactions. Recently, in collaboration with Rigaku, they have developed a new diffractometer with an imaging-plate detector, R-AXIS RAPID, which enables three-dimensional intensity data collection within 2 or 3 hours. Using this new diffractometer, they analyzed a variety of metastable intermediate structures, such as radical pairs of hexaarylbiimidazolyl derivatives, triplet carbenes (Fig. 1) and triplet nitrenes. Moreover, they analyzed the excited-state structures of a series of Pt complexes and a vanadium complex at the equilibrium state between ground and excited states. In order to analyze more rapid structural changes, they are now developing a new detector, Micro-Strip-Gas-Chamber (MSGC). Using this detector and a rotating anode, preliminary intensity data were collected within 2 seconds and the structure was successfully analyzed. Combining the detector and Synchrotron radiation at SPring-8 (Fig. 2), they found that the cell change of the Pt complex due to the formation of the excited state continued for more than 20 seconds and then the equilibrium state appeared.
T. Ozeki (Tokyo Inst. of Technology) is working on the structural chemistry of polyoxometalates. His research interests include the crystal structure determinations of novel polyoxometalates, hydrogen bonds involving the polyoxometalates, extended structures using polyoxometalates as building blocks, and inclusion compounds using the polyoxomelatate as the host (Fig. 3). He is also in charge of two diffractometers in the synchrotron radiation facilities: an imaging plate Weissenberg camera at the BL04B2 beamline of SPring-8 and a CCD diffractometer at the NW2 beamline of the Advanced Ring (AR), Photon Factory (PF) of the High Energy Accelerator Research Organization (KEK). Using these facilities, he is exploiting the advantage of the high energy (typically λ=0.33Å) and high flux X-rays to do atomic-resolution single crystal structure analyses of polyoxometalates and other chemically synthesized molecules (sometimes the size of which is stepping out of the realm of smallness and thus the common name “small molecules” is not used here).
K. Toriumi and Y. Ozawa (Univ. of Hyogo) and their coworkers have developed a new low-temperature vacuum X-ray camera (LTV X-ray camera) and installed it at the SPring-8 BL02B1 beamline (Fig. 4) for photo-excited crystallography, phase transition studies at low temperature, and micro-crystal structure analyses. By eliminating X-ray scattering from air and windows, high quality diffraction data with high S/N ratio can be automatically obtained down to 20 K. Direct observation of geometrical changes accompanied by photo-excitation of molecules provides essential information for transient species such as metastable states of chemical reactions. They have developed the multiple-exposure IP method for accurate measurement of intensity changes due to photo-illumination. A photo-excited molecular structure has been observed for the luminescent diplatinum(II) complex [Pt₂(pop)₄]_4-.The observed electron density map shows positive and negative peaks with heights of 1–2 e/Å3 near the Pt atoms (Fig. 5). This indicates that small portions of the metal atoms move toward the positive peaks in the light-on crystal.
The main subjects of H. Kobayashi (Inst. of Molecular Science) and A. Kobayashi (The Univ. of Tokyo) are: (1) development and physical properties of molecular metals and superconductors and (2) molecular design, development and physical properties of single-component molecular metals. Until the early 1990s they engaged in the development of various molecular superconductors including a k-type BEDT-TTF superconductor (k-(BEDT-TTF)₂I₃) and a p-acceptor (M(dmit)₂, M=Ni, Pd) superconductor without TTF-like donors. Recently they have reported an organic conductors exhibiting colossal magnetoresistance and field-induced superconductivity, l-(BETS)₂FeCl₄ and the first antiferromagnetic organic superconductor, k-(BETS)₂FeBr₄ (BETS = Bis(ethylenedithio)tetraselenafulvalene). They have also developed a single-component molecular metal with extended-TTF (tetrathiafulvalene) ligands, [Ni(tmdt)₂] (tmdt = trimethylenetetrathia fulvalenedithiolate). The planar molecules are closely packed and three-dimensional S…S interactions were observed in the crystal (Fig. 6). The room temperature conductivity was 400 S cm^–1 and metallic behaviour was observed down to 0.6 K. Experimental evidence for the existence of Fermi surfaces was obtained by detecting the quantum oscillations in magnetization (the de Haas-van Alphen (dHvA) effect). Metallic [Au(tmdt)₂] is iostructural with [Ni(tmdt)₂] and has an unusually high temperature SDW antiferromagnetic transition at around 85 K.
K. Tanaka (Nagoya Inst. of Technology) has reported 4f-electron transfer to B6 with a decrease in temperature in the study of the electron density distribution (EDD) analysis of the Kondo crystal CeB₆ by X-ray atomic orbital methods. This analysis opened the door to many interesting heavy-atoms materials and they have tried to explain precise electron densities for high Tc super conductors, skutteldites and rare-earth garnets. Diffraction measurements under vacuum by the VCIP method is also being developed to get super accurate structure factors for retrieving molecular orbitals from X-ray diffraction.
The research activity of J. Mizuguchi and co-workers (Yokohama National Univ.) is focused on molecular structure, crystal structure and intermolecular interactions with potential electronic applications. Organic pigments are basically organic semiconductors which can be used in electronic materials such as photoconductors, electrophotographic photoreceptors, solar cells, optical recording materials, organic transistors etc.
Y. Sugawara (Kitasato Univ.), investigated humidity and temperature dependent phase transitions of nucleosides and nucleotide hydrates. Change in the numbers of water molecules or in their ordering at low temperature accompanies reconstruction of hydrogen-bonding networks. Structural transformations and dynamics of phase transitions are being examined using X-ray and neutron diffraction methods, Raman spectroscopy, and computer simulation calculations. These phenomena are interesting from the view points of crystal engineering, pseudopolymorphism of pharmaceuticals, and water structure in biological systems.
K. Ogawa and J. Harada (The Univ. of Tokyo) have discovered a pedal motion in crystals (Fig. 7), an essential molecular motion of stilbene-type molecules. They have also been successful in providing new insight into the chromism of organic crystals.
The main research interests of M. Yasui, F. Iwasaki (Univ. of Electro-Communications) and D. Hashizume (Riken) are: (1) weak intra and intermolecular interactions by a topological analysis of experimental charge density distributions, and (2) the mechanism of phase transitions of organic crystals using detailed temperature-resolved diffraction methods. Topological analyses for tetraazathiapentalenes, pentacoordinated hypervalent boron and carbon compounds show small positive Laplacian values at bond critical points which indicate the electrostatic character of these hypervalent bonds (Fig. 8). Analyses of intermolecular Br...Br contacts shorter than the normal van der Waals contacts revealed an electro-static interaction corresponding to the dispersion force. The magnitude of Br...Br interactions can be estimated to be about half of those of NH...O hydrogen bonds. They also studied intermolecular CH...O interactions in TEMPO radical derivatives. Topological properties prove that at least one H atom has a significant interaction with the p* orbital of the radical O atom which can explain intermolecular magnetic interaction. They have developed a ‘detailed temperature resolved diffraction method’ during the phase transition of organic crystals. This method can make the mechanism of the transition clear by looking at a set of stop-motion photographs. The mechanisms of phase transition of acylurea derivatives and thiocatechole crystals were clarified, and energy barriers of these transitions were estimated (Fig. 9).

R. Tamura (Kyoto Univ.) reported the first instance where enantiomeric resolution by simple recrystallization of a racemic crystal was observed. This unusual enantiomeric resolution phenomenon was referred to as preferential enrichment. Mechanistically, it has been proven that preferential enrichment is a secondary, dynamic enantiomeric resolution phenomenon caused by a solvent-assisted solid-to-solid transformation of a metastable polymorphic form into a thermodynamically stable polymorphic form. It occurs during crystallization of certain kinds of racemic mixed crystals composed of two enantiomers in a supersaturated solution. This phenomenon has been detected in (i) the crystal structures of the stable and metastable polymorphic forms (ii) by a direct-space approach employing the Monte Carlo method followed by Rietveld refinement, (iii) in the in situ ATR-FTIR (ReactIR) spectral data during crystallization, and (iv) in the DSC analytical and solid-state ReactIR spectral data of the deposited crystals (Fig. 10).

R. Kuroda’s research group (Univ. of Tokyo) explores and exploits solid-state chiral chemistry. In crystals, interactions between molecules are expected to be orders of magnitude stronger than in solution. Thus, it is safe to assume that chiral discrimination, recognition, generation and transfer occur most strongly in the solid state. They have studied chirality recognition in solvent-free, solid-solid reactions. Upon co-grinding and heating (without melting) of the crystals of a template compound and a substrate compound, the chirality of the substrate compound was inverted to fit with the chirality of the template compound. They have developed two novel instruments, UCS-1 (Universal Chiroptical Spectrophotometer) and UCS-2, for measuring the chirality of solid materials.

Today, many scientists in Japan, whose main research involves organic or coordination chemistry, have their own diffractometers. Apart from the above examples, splendid work is going on related to chemical crystallography, such as, multicolor phototropism of single crystals by M. Irie (Kyushu Univ.), solid-solid synthesis using host-guest interactions by F. Toda (Okayama Univ. of Science) and K. Tanaka (Kansai Univ.), inclusion phenomena of choric acids by M. Miyata (Osaka Univ.), and synthesis of organic semiconductors with metallic luster by K. Ogura (Chiba Univ.).

F. Iwasaki, Univ. Electro-communication

Materials science

Structural science of materials in Japan is studied not only by crystallographers but also by materials scientists, physicists, chemists and many other researchers in science and technology. This report cannot be expected to cover all the aspects of structural science of materials in Japan but a few aspects of the field give some flavor of our country.
In 1991, S. Iijima and coworkers (NEC Co. LTD.) reported a structure of a nanotube. There are many varieties, such as multi-wall, double-wall, and single-wall carbon nanotubes. These structures are mainly determined by TEM. Structural studies of nanotubes by X-ray are also active in Japan. One of the interesting properties of nanotubes is an inner hollow cavity of nanometer size. Kataura and coworkers reported that a one-dimensional crystal of fullerene molecules were found inside the nanotubes. Takenouchi and coworkers suggested that the nanotube absorbed not only fullerene molecules but also exotic molecules like TCNQ and TMTSF.
Another nano-structured carbon material, for which structures are mainly determined in Japan, is metallo-fullerene. M. Sakata, M. Takata and coworkers confirmed the endohedral nature of metallo-fullerene by X-ray powder diffraction at SPring-8. Since then, the structures of endohedral metallofullerenes have shown very unique structures including the disordered states of two La atoms in a C₈₀ carbon cage (Fig. 11).
J. Akimitsu and coworkers (Aoyama Gakuin Univ.) discovered that MgB₂ has one of the highest transition temperature (T_c = 39 K) of the known metallic superconductors. The superconductivity relating the BCS (Bardeen-Cooper-Schrieffer) mechanism has a superconducting energy gap associated with formation of the superconducting pairs. T. Takahashi (Tokyo Univ.) and S. Sasaki (Tokyo Inst. Tech.) suggested the existence of two kinds of superconducting gaps in MgB₂ on the basis of an ARPES (angle-resolved photoemission spectroscopy) study, where the σ and surface bands have large gaps of 6-7 meV and the σ band is dominant in the superconductivity with a stronger coupling to phonons. The electron-phonon coupling in MgB2 causes the high T_c in a particular phonon mode with the E_2g symmetry at Γ. The electron-phonon coupling is only possible for phonons with momenta appropriate to move electrons within the neighborhood of a Fermi surface. The softening and broadening of the E_2g mode were observed inside the σ surfaces, by using a highly efficient X-ray spectrometer with a 4 meV resolution at SPring-8. K. Ohshima and coworkers reported a Kohn anomaly in TaS₂ from the phonon dispersion curves.

Charge ordering and charge fluctuation of mixed-valence compounds are commonly investigated in Japanese universities and governmental insitutes. The valence-difference contrast method with X-ray anomalous scattering and electron-microscopic techniques are used for materials such as RM₂O₅ (R: Y or a rare earth; M = Mn, Fe) by Y. Yamada, and K. Kohn (Waseda Univ.), La_1-xSr_xMnO₃ by H. Sawa (Photon Factory), Fe₃O₄ by S. Sasaki (Tokyo Inst. of Technology), CuIr₂S₄ by H. Ishibashi (Osaka Prefecture Univ.), Pr_1-xCa_xMnO₃ and Nd_1-xSr_xMnO₄ by Y. Matsui (National Inst. for Materials Science(NIMS)), CaFeO₃ by S. Morimoto (Osaka Univ.), NaV₂O₅ by Y. Fujii (Univ. of Tokyo), and La_1-x(Ba,Sr)_xCuO₄ by Y. Noda (Tohoku Univ.). An example of the research by Noda is Eu₃S₄, where Eu²⁺ and Eu³⁺, in a tetragonal cell below T_c = 188.5 K, occupy 4a and 8d sites, having the cation distribution of [Eu³⁺]^4a[Eu²⁺Eu³⁺]^8dS₄.as shown in Fig. 12.

X-ray resonant magnetic scattering was found for Ni single crystals by K. Namikawa (Tokyo Gakugei Univ.) in 1985. Resonant magnetic scattering factors were then obtained from experimental data from Fe₃O₄ by H. Kawata (Photon Factory) and coworkers. Systematic work on nonresonant magnetic scattering has been devoted to metallic and simple oxides by M. Ito (Gunma Univ.). The resonant X-ray scattering technique has been proved to be a powerful probe of orbital states, and is now being applied to a wide range of systems with improved accuracy. One example is the case of La_0.45Sr_0.55MnO₃/La_0.6Sr_0.4MnO₃ multilayers reported by M. Izumi (NIMS).

The horizontal-type high-speed four-circle diffractometer at beam line 14A at the Photon Factory has been used in a number of electron density studies. Recently a high-speed detector called a stacked avalanche photodiode detector (APD) was employed with the collaboration of S. Kishimoto (Photon factory) and N. Ishizawa (Nagoya Inst. Tech). Studies of La(Sr)₂CuO₄ and MnS₂ revealed an enhanced accuracy of the experimentally determined electron densities.

Micro-beams obtained by using synchrotron radiation are powerful tools. Interplanetary dust particles and meteorites are usually examined by optical microscopy, chemical analysis, micro Raman spectroscopy, and so on. The lower limit of the area for these examinations is about a micrometer. Diffraction data from the same samples is indispensable especially in cases of polymorphs and polytypes. Interplanetary dust particles of iron sulfides and micrometer-sized areas of the same kind in a thin section of meteorite are identified, and the structure is refined based on intensities of the Laue spots obtained by using polychromatic synchrotron radiation at the Photon Factory by K. Ohsumi.

High-resolution synchrotron powder diffractometers are available at beam line 4B2 of the Photon Factory (δd/d=0.04%) and at the BL15XU of SPring-8 (δd/d=0.03%). Two angle-dispersive-type neutron powder diffractometers (HRPD) and HERMES are installed at the JRR-3M research reactor in JAERI. There are two time-of-flight (TOF) neutron powder diffractometers (Sirius and VEGA) at KENS of KEK in Tsukuba. The KENS facility will be shut down soon, but a new powerful neutron facility (JSNS at J-PARC) has been built. At the JSNS, a new high-resolution TOF neutron powder diffractometer will be installed. At the Ceramics Research Laboratory, Nagoya Inst. Tech., the profile functions of powder diffraction data are characterized by T. Ida. F. Izumi (NISM) developed computer programs for Rietveld analysis, maximum-entropy methods and visualization of crystal structure and electron/nuclear density distribution. The superspace group approach to structure analysis of composite crystals has been a successful collaboration between NIMS and Tohoku Univ. T. Ikeda (Tohoku Univ.) analyzed the crystal structures and properties of synthesized zeolites through ab-initio computational methods.

M. Yashima (Tokyo Inst. Tech.) and coworkers have been studying high-temperature neutron and synchrotron powder diffraction techniques for precise structure analysis up to 1900 K. By using the furnace diffusion path, the disorder in some ceramic superionic conductors were visualized (Fig. 13).

Coordinated by Kazumasa Ohsumi

(kazumasa.ohsumi@kek.jp) and

Matt Sakata (sakata@cc.nagoya-u.ac.jp) 

High-pressure activity

For high pressure studies, a large volume press and diamond anvil cell (DAC) are used. There are two types of large volume presses; one is a DIA type cubic anvil system and the other is a Kawai type double stage anvil system.
The DIA type apparatus, MAX80 at the Photon Factory and SMAP-180 at SPring-8, can generate a pressure and temperature around 10 GPa and 2000K, and it is mainly used for materials science. Recent work with SMAP-180 involved the observation of congruent melting of GaN under high pressure and temperature ^[1]. Fig. 15 shows the phase boundaries of GaN obtained at high pressure and temperature. This observation enables one to obtain bulk single crystals of GaN by slow cooling from a high pressure congruent melt. GaN, a famous green fluorescent material, had so far been made only in thin films. Another exciting result is finding the pressure induced first order transition in liquid states. Phosphorus was found to show a clear first order transition at around 1 GPa and 1000K (Y. Katayama, SPring-8/JAERI). This was confirmed by measurements of XAFS, density, X-ray diffraction and X-ray transmission imaging at the transition. Density was measured by X-ray absorbance by filling the sample in a sapphire or diamond ring. A sharp liquid-liquid transition in liquid germanate accompanied by a fourfold to six fold coordination change was also found around 3 GPa at 1273 K(3) (O. Ohtaka, Osaka Univ.)

The first Kawai type press for SR study named SPEED1500 can generate pressure and temperature around 30 GPa and 2000K with tungsten carbide anvils, and will be used to study the interior of the earth across the lower mantle region. The first result using SPEED 1500 is the determination of the phase boundary of Mg₂SiO₄ (T. Irifune, Ehime Univ.). Assuming a geotherm of 1650K, dissociation pressure is estimated to be 21GPa, which is apparently lower than the previous result. This was a sensational result to understand the upper to lower mantle structures.

In order to extend the pressure range, a new system named SPEED Mk-II using sintered diamond anvils was developed at SPring-8. At room temperature, pressure generation higher than 60GPa can be routinely achieved. SPEED Mk-II has an oscillation mechanism to avoid the effect of preferred orientation and grain growth. With this oscillation mechanism, the diffraction profiles of NaCl were clearly observed with reasonable integrated intensities very close to the melting temperature, and the phase relations among the B1, B2 and liquid phases are precisely determined (N. Nishiyama, Ehime Univ.).

A unique technique with SR is viscosity measurements using a falling ball method. An image is recorded each 1/125 second and the linear part of the falling distance against time is fit to give an accurate viscosity. Sulfur content dependence in the Fe-FeS system is clearly seen, as well as temperature dependence.

We recently succeeded in measuring ultrasonic velocity at high pressure and temperature. A transducer is set outside the anvil to prevent damage by pressure and temperature. The length of the sample is measured directly by a transmission image of the sample using a CCD camera, and volume and pressure are measured by X-ray diffraction. With this technique, concentration dependence of bulk and shear moduli for ringwoodite are beautifully observed, which agrees very well with finite strain fitting simulations (Y. Higo, Ehime Univ.).

At dedicated DAC stations (BL10XU at SPring-8) one can perform high and low temperature diffraction experiments. Strongly correlated materials, nano-materials, etc. are extensively studied by using the quasi hydrostatic pressure medium of He. Structural behaviors of simple materials such as Sc, BeO, Hg, Cs, H₂, O₂, and LiH are investigated in the ultra high pressure region above 200GPa.

High temperature experiments with DAC are performed by use of a laser heated DAC. The characteristics of the system at SPring-8 are dual lasers (YAG and YLF) and dual detectors (IP for precise measurement and CCD for rapid measurement). The most exciting result with laser heated DAC is the structural phase transition of MgSiO₃ at 125GPa and 2400K^[2]. The structure of the lower pressure phase is a well known perovskite type structure with apical sharing. The structure of the high pressure phase, which is explained by a Cmcm structure type, is unique in that it is a layered type two dimensional one (Fig. 16). This pressure and temperature condition corresponds to the mantle to core boundary of the Earth, and this result is believed to help clarify the structure of earth’s core.

Other experiments using DAC such as nuclear X-ray scattering, inelastic scattering, infrared spectroscopy, etc. are performed at each general beamline.

The high pressure neutron group, headed by H. Kagi (Tokyo Univ.) submitted a proposal for the construction of a high pressure system consisting of a DIA type high pressure cell to J-PARC. The proposal is on the waiting list.

Osamu Shimomura, simomura@spring8.or.jp

References

[1] W. Utsumi, H. Saitoh, H. Kaneko, T. Watanuki, K. Aoki and O. Shimomura. Nature Mater. 2, 735 (2003).
[2] M. Murakami, K. Hirose, K. Kawamura, N. Sata and Y. Ohishi, Science 305, 855 (2004).

Electron diffraction

The electron microscope is unique in that it enables us to simultaneously conduct experiments in diffractometry, microscopy and spectroscopy on a nanometer scale.

Tsuda et al.^[1] developed a method to refine crystal structural parameters and charge density using convergent-beam electron diffraction (CBED). The method is based on a least-squares fit between full dynamical calculations and energy-filtered intensities from two-dimensional higher-order Laue zone (HOLZ) and zeroth-order Laue zone (ZOLZ) CBED patterns. Application of this method to the rhombohedral phase of LaCrO₃ revealed clear anisotropy of thermal vibration of the oxygen atoms and charge transfer from the metal atoms to the oxygen atoms. With the aid of parallel computations, the structure (30 positional parameters and Debye-Waller factors) of the intermediate phase of hexagonal BaTiO₃ was refined. In the orbital ordering phase of LaMnO₃, the anisotropic charge distribution caused by orbital ordering of the 3d-electrons of the Mn atoms was found.

Fujiyoshi et al.^[2] developed a high-resolution electron cryo-microscope equipped with a top-entry specimen stage. Using the microscope, they determined the atomic structures of several membrane proteins with the use of two-dimensional (2D) and tubular crystals. Image shifts due to beam-induced specimen charging were found to be the most severe problem in imaging of biological macromolecular 2D crystals, especially at highly specimen tilt conditions. To reduce beam-induced movement, Gyobu et al. developed a new sample preparation method, the carbon sandwich method, in which the crystals are put between two carbon films. This method has improved the success ratio of obtaining high-resolution images from tilted specimens, namely from 30 to 90% at a tilt angle of 45 degrees. The method enabled them to collect a full data set (87.0% complete) of aquaporin-4 (AQP4), a water channel protein, and to conduct the structural analysis at 3.2 Å resolution with a Friedel factor and merging R-factor of 11.4 % and 22.3 %, respectively.

Nagayama et al.^[3] developed a series of transmission electron microscopes (TEM) capable of retrieving object-retarding phases in electron waves. The Zernike phase contrast (ZPC) method uses a classical π/2 phase plate in the form originally developed by Zernike. The Hilbert differential contrast (HDC) method uses a half-plane π phase plate to generate images similar to the differential-interference-contrast conveniently used in light microscopy. The Foucault differential contrast (FDC) method uses dynamical control for a Foucault knife-edge to perform an accurate differential operation for phases involved in the wave function of electrons. The remarkable high contrast achieved with these phase contrast TEMs without sample staining is now opening a novel field of in vivo electron microscopy in biology.

Matsui et al.^[4] carried out intensive structural studies of various new superconductors by means of high-resolution transmission electron microscopy (HRTEM). They found incommensurate superstructures in Bi-2212 and Bi-2223 superconducting phases, and proposed modulated structure models with strongly distorted lattice planes. They developed a new high-voltage (1300kV) HRTEM with an 0.1nm point-resolution and used it to identify new oxycarbonate superconductors which contain CO₃ groups forming various types of order/disorder structures. Since new phenomena of colossal magnetoresistivity (CMR) were reported, Matsui et al. also studied the magnetic nanostructures of various magnetic materials by cryo-Lorentz electron microscopy.^[5] They examined the formation of ferromagnetic domains in a “double-perovskite” Ba₂MoFeO₆, and proved that the crystallographic anti-phase boundaries tend to pin the ferromagnetic domains.

Saitoh et al.^[6] applied HAADF-STEM and ALCHEMI to quasicrystals and their approximants for the first time. Using the HAADF-STEM method, they first observed an asymmetric atom-cluster and a high degree of quasiperiodic arrangement of clusters in decagonal Al₇₂Ni₂₀Co₈, which led to the quasi-unit-cell model. Using the two-dimensional angular-scanning ALCHEMI, Saitoh et al. also found that, in decagonal Al-Ni-Co and Al-Ni-Fe, two kinds of transition metal elements occupy the same sublattice site (chemical disorder), which is compatible with the fact that the quasicrystals are stabilized by the Hume-Rothery mechanism.

Abe et al.^[7] demonstrated the real-space imaging of a local thermal vibration anomaly in a solid through atomic-resolution annular dark-field scanning transmission electron microscope (ADF-STEM) observations of an Al₇₂Ni₂₀Co₈ quasicrystal. They found significant changes of the ADF intensity at some Al atomic sites, which depend on the observation temperature and the scattering angle range. This anomalous ADF intensity, which is due to an anomaly of the thermal diffuse scattering (TDS) intensity, is explained fairly well by an anomalously large value of the temperature (Debye-Waller) factor of Al at the sites or by a larger mean-square thermal vibration amplitude of the atoms. By introducing angle-resolved and/or in-situ heating/cooling techniques, they have extended the ADF-STEM method as a tool to determine local Debye-Waller factors that crucially affect the physical properties of materials.

Suenaga et al.^[8] demonstrated chemical analysis by means of electron energy-loss spectroscopy (EELS) with a sensitivity of a single atom limit and the high-resolution imaging of individual metallofullerene molecules encapsulated in a single-wall carbon nanotube. They directly observed nanotubes composed of single-graphene layers and their structural defects using phase contrast transmission electron microscopy.

Takayanagi et al.^[9] developed an ultra-high vacuum electron microscope combined with a scanning tunneling microscope to study structure and electronic conductance quantization of gold atomic chains and nanowires. They observed that a gold atomic chain, which was fabricated between the gold STM tip and the gold substrate, has the conductance quantum of 2e2/h, where e is electron charge and h is the Planck constant. They found gold nanowires having single-wall and multi-wall tubular structures (helical multi-shell magic number seven structure).

Using electron holography, Hirayama et al.^[10] observed two-dimensional electric potential distributions in a cross section of a Metal-Oxide-Semiconductor Field-Effect-Transistor (MOSFET) fabricated from a silicon wafer with a boron concentration of 1015cm^-3. This implies that electron holography allows the two-dimensional mapping of dopant distributions as low as 1015cm^-3. This technique is useful for developing new devices for failure analysis in the semiconductor industry.

Kimoto et al.^[11] applied electron energy-loss spectroscopy (EELS) to the analysis of an ultra-thin film of amorphous-Al₂O₃ on Si. EEL spectra were successfully acquired with a small interval of 0.28 nm in depth using their spatially-resolved technique. From the spectra, they found different Al coordinations of an AlO₄ tetrahedron and an AlO₆ octahedron with the aid of first-principle calculations. They revealed the detailed depth dependence of Al coordination in the film with sub-nanometer resolution.

Terauchi et al.^[12] constructed a high energy-resolution EELS microscope equipped with a Wien-filter monochromator and a Wien-filter analyzer. The EELS microscope was used to measure the bandgap energies and the density of states (DOS) of the conduction bands of BN nano-cones and metal-doped boron micro-crystals with energy resolutions of 0.2-0.26eV. Terauchi et al. have developed a high resolution soft-X-ray spectrometer for X-ray emission spectroscopy (XES) for a transmission electron microscope (TEM). This instrument enables us to obtain the DOS of the valence band with an energy resolution better than 1ev from a small specimen area identified by observing an electron microscope image. They demonstrated using h-BN that the total DOS of the valence and conduction bands can be obtained in the electron microscope from XES and EELS spectra.

Michiyoshi Tanaka
(tanakam@tagen.tohoku.ac.jp)

References

[1] K. Tsuda and M. Tanaka, Acta Cryst., A55 (1999) 939-954.
[2] Y. Fujiyoshi, Adv. Biophys. 35 (1998) 25-80.
[3] R. Danev and K. Nagayama, Ultramicroscopy 88 (2001) 243-252.
[4] Y. Matsui et al., Jpn. J. Appl. Phys. 27, L372 (1988).
[5] Y. Anan et al., J. Electron Microsc. 50, 457 (2001).
[6] K. Saitoh, K. Tsuda, M. Tanaka, K. Kaneko, A. P. Tsai, Jpn. J. Appl. Phys. 36 (1997) L1400.
[7] E. Abe, S. J. Pennycook and A. P. Tsai, Nature 421 (2003) 347-350.
[8] K. Suenaga et al., Science, 290 (2000) 2280-2282.
[9] H.Ohnishi et al., Nature 395 (1998) 780-782.
[10] Z. Wang, T. Hirayama, K. Sasaki, H. Saka and N. Kato, Appl. Phys. Lett. 80 (2002) 246-248.
[11] K. Kimoto et al., Appl. Phys. Lett. 83 (2003) 4306-4308.
[12] M.Terauchi, M.Tanaka, K.Tsuno and M.Ishida, J. Microscopy, 194 (1999), 203-209.

XAFS activities

XAFS activities in Japan started in 1982. The EXAFS beamline BL-10B, one of the first beamlines constructed at the Photon Factory (PF), is equipped with a Si(311) channel-cut type monochromator without any focusing optics. A number of papers for structural analysis of materials have been published using data from this beamline. New XAFS beamlines constructed include a sagittal focusing double crystal monochromator (BL-7B), and a Si(111) double crystal monochromator with curved cylindrical mirrors both up- and down-stream of the monochromator. Several upgrades of the PF ring lowered the emittance to 27 nm rad.
SPring-8, a third generation hard X-ray 8 GeV ring has two beamlines dedicated to XAFS experiments; BL-1B at the bending magnet port and BL-10XU in an undulator port. The former has been used for XAFS experiments in the higher energy region extending to 100 keV, and the latter for focused (0.1-1 mm2), high photon flux XAFS with more than 1013 photons/s. The later beamline is especially useful for XAFS under high pressures. Micro XAFS experiments are conducted at BL37XU using Kirpatrick and Baez (K-B) mirror optics. The beam size was 2 (H) × 4 (V) μm2, with a flux of more than 1010 photons/s at 10 keV.
Another 6 GeV storage ring in KEK, which had been used as a booster ring for the TRISTAN main ring (25 GeV), has been renewed as a high current ring with a single bunch mode dedicated for synchrotron radiation use. This is especially useful for time resolved XAFS. In the AR facility, the dispersive XAFS technique was installed, which enables one to measure EXAFS in less than 1 ms. Quick XAFS is also under construction and will be dedicated to structure analysis of catalysts.

One unique feature of XAFS activities is that the laboratory XAFS instruments are prevailing in Japan. TECHNOS I.T Co. Ltd and RIGAKU Co. have independently developed compact XAFS instruments equipped with an X-ray tube. By using a Johansson-type curved crystal, EXAFS measurements can be taken in a reasonable time; typically, 20 min for Cu K-EXAFS of a Cu foil.

There are more than 150 users of XAFS experiments in Japan. The society of XAFS was organized in 2000. Its annual meetings attract more than 100 participants. More than 50% of its members specialize in catalysis and 15% are from industry.

Although XAFS studies are diverse in various fields, two recent highlights:

(1) Self-regeneration of a Pd-perovskite catalyst for automotive emission control. Catalytic conversion of exhaust gases is an essential part of automobiles. Recently, the Japanese motor company, Daihatsu Co. developed an intelligent conversion catalyst, LaFe_0.75Co_0.38Pd_0.05O₃, perovskite-based catalysts which has a much longer lifetime than the conventional catalyst, Pd/Al₂O₃. The structural change of the perovskite catalyst was studied in SPring-8 by using Pd and Co K-edge XAFS as well as X-ray diffraction (Y. Nihshihata et al. Nature 418, 164 (2002)) Fig. 17 shows the Fourier transformed spectra of Pd K-edge EXAFS oscillations of the sample at oxidized, reduced and re-oxidized states together with that of a Pd foil. In the oxidized state, Pd is surrounded by oxygen, while in the reduced state, Pd is surrounded by Co and Pd. Combined with the XRD data, it revealed that Pd is in the B-site (octahedral site) of the perovskite lattice in the oxidized stateand in the reduced state, Pd is segregated with Co to form a PdCo solid solution. This process is reversible and explains the retention of high catalytic activity during long-term use and aging.

(2) Little has been known about the dynamical structural change of active metal sites in supported metal cluster/nanoparticles catalysts. The in-situ time resolved XAFS study of a CO-induced disintegration process of Rh clusters on an Al₂O₃ surface was performed at PF (Suzuki et. al, Ang. Chem. Int. Ed. 42 (2003) 4795). Rh K-edge EXAFS of an Rh/Al₂O₃ catalyst was taken under 26.7 kPa of CO at 298 K every 100 ms with the energy dispersive mode. Fig. 18 shows a series of Fourier transforms calculated during the carbonylation process. The peak at 0.2 nm (Rh-Rh) suddenly reduces and the peak at 0.1 nm (Rh-C) increases rapidly. Analysis of the coordination number and bond distances (Rh-Rh, Rh-C) as a function of CO exposure reveals that there are three elementary steps for the surface dynamic structural rearrangement of Rh clusters involving two intermediate states as depicted in Fig. 19. Before CO exposure, each Rh cluster consists of seven atoms in the first layer and three atoms in the second layer on Al₂O₃. CO exposure causes Rh-CO bond formation for the second layer in 600 ms, and further CO exposure induces Rh-CO bond formation with cluster disintegration and finally each Rh atom adsorbs in the threefold hollow site, forming Rh(CO)₂. These results demonstrate that dispersive XAFS is useful to elucidate the mechanism for dynamic surface processes.

Toshiaki Ohta (ohta@chem.s.u-tokyo.ac.jp)

---

First published for the International Union of Crystallography 1962 by N.V.A. Oosthoek's Uitgeversmaatschappij, Utrecht, The Netherlands
Digitised 1999 for the IUCr XVIII Congress, Glasgow, Scotland
© 1962, 1999 International Union of Crystallography