■THE DISCOVERY of CARBON NANOTUBES

                              Figure1
                                                  in NEC Co.

 In June, 1991, I found an extremely thin needle-like material when examining carbon materials under an electron microscope. Soon thereafter the material was proved to have a graphite structure basically, and its details were disclosed. I named these materials "carbon nanotubes" since they have a tubular structure of carbon atom sheets, with a thickness scaled in less than a few nanometers. The name has been widely accepted now. Carbon nanotubes have attracted a lot of researchers in a wide range of fields from academia to industry, not only because of their uniqueness when compared with conventional materials, but also because they are very promising materials in nanotechnology in future technology.


 The simplest carbon nanotube is composed of a single sheet of a honeycomb network of carbon atoms. Called graphene, it is rolled up seamlessly into a tubular form. The carbon nanotubes reported in the first report were composed of multi-tubes (multi-wall tubes) nesting in a concentric fashion [Iijima, Nature, 354, 56 (1991)]. Later, a single-wall carbon nanotube was discovered [Iijima et al., Nature, 363, 603 (1993)]. The tubular structures are rarely found in nature, except for some silicate minerals such as chrysotile asbestos. However, tubular carbon structures have essentially never been found in nature, and are thus a new, artificially created type of carbon solid. They may be categorized in the fourth allotrop of carbon, following diamond, graphite and fullerenes. Another more general difference of carbon nanotubes with respect to conventional crystalline materials is their sizes, that is to say, their diameters are in a range of a few nanometers and their lengths are typically a few micrometers. The diameter of a nanotube is almost comparable to the size of an individual molecule, and therefore the carbon nanotube could be considered as a single molecule. Because of this size, carbon nanotubes behave as both molecule and solid, or their hybrid, and exhibit unique physical and chemical properties. Single-wall nanotubes, in particular, show more dominant properties originating from this one-dimensionality, which becomes more dominant as its diameter becomes smaller. Another important factor controlling unique properties comes from a variation of tubule structures that are caused by the rolling up of a honeycomb sheet of carbon atoms. There are many possible ways to do this, depending upon the direction of rolling. This could result in many chiral arrangements (spiral arrangement of carbon atoms) of the nanotube structures as well as a variety of diameters.

from 1991 Nature  *Sumio Iijima, Nature, 354, 56 (1991).


 After the first report of the discovery of carbon nanotubes, a few theoretical condensed matter physicists got interested in this new material and predicted electronic band structures for single-wall carbon nanotubes that depend on diameter and chirality [Hamada et al., Phys. Rev. Lett., 68, 1579 (1992)]. A carbon nanotube is a simple system composed of a reasonable number of atoms, which enable us to calculate theoretical electronic structures in detail through computer simulations. As a result, single-wall carbon nanotubes were found to be electrically semiconducting or metallic depending upon their diameters and chirality. Such important physical properties were later proved experimentally in various electrical and optical measurements. Therefore, a single piece of single-wall carbon nanotube can be a transistor. Indeed, such a transistor was built some years later. This is one of the reasons why for industry became interested in carbon nanotubes. Interest in carbn nanotubes is not limited to transistors, and many other industrial applications utilizing the unique properties of carbon nanotubes include the electron emitter source with high current density, highly conductive electrical wire, the high thermal conductor for the heat radiator, the probe needle for scanning probe microscopes, molecular-sieves, gas adsorbers, carriers for drug delivery systems in nano-bio medicine, etc. These possible applications play an important role in nanotechnology, and are currently being investigated the wourld over.


 A frequently asked question after the discovery of carbon nanotubes is "How did you find the carbon nanotube" and my first answer is a "Serendipity", but this is not really the way things went. My real answer is "Logic or Rationale". Here I would like to explain the reason for my answer. The origin of the discovery could be traced back in the beginning of my research carrier, which started in 1970, when I was a post-doctoral research fellow in Prof. J. M. Cowley's group at Arizona State University. We developled a high resolution electron microscope and its use, and successfully recorded the first electron micrographs showing individual metal atoms in some oxide crystals [Iijima et al., J. Appl. Phys., 42, 5891 (1971)].

TEM operation *with Prof. J. M. Cowley.

The work involved understanding electron diffraction in the crystal and electron optics of the magnetic lenses of the electron microscope. An essential part of the physics of high resolution electron microscopy was almost finished by the end of the 1970s. During this period I had many occasions to examine carbon materials among many other materials under a high resolution electron microscope at the level of atomic resolution. Works related to the materials are: the imaging of grassy carbon, atomic structure analysis of amorphous carbon materials [Iijima, J. Microscopy, 119, 99 (1980)], imaging atomic steps on thin graphite films [Iijima, Optik, 48, 193 (1977)], imaging single tungsten atoms distributed on a thin graphite film [Iijima, Optik, 48, 193 (1977)] and the imaging of small graphite particles [Iijima, J. Cryst. Growth, 50, 675 (1980), Iijima, J. Phys. Chem., 91, 3466 (1987)]. This last work is concerned with the imaging of multi-wall carbon nanotubes and even fullerene molecules of C60 that were recorded five years before the discovery of fullerene molecules [Kroto et al., Nature, 318, 162 (1985)]. All of these accumulated experiences and knowledge obtained from various types of carbon materials allowed me to immediately get to the problem and solve it, when I accidentally came across carbon nanotubes later. The high resolution electron microscopy is my life-long research theme. In this context, I should say that the carbon nanotube is only one of my many research subjects.


 Besides my experience with carbon materials, work with other materials led me to the discovery of carbon nanotubes. One of them is thin filaments of silver crystal, called 'whiskers' in the 1960s. I discovered these filaments growing on AgBr crystal particles separated from photo-graphic emulsion, and studied the crystallography of this filament crystal by analyzing its electron diffraction patterns. I found that twin-defects were responsible for such an anisotropic growth of the silver filament [Iijima, Jpn. J. Appl. Phys., 8, 1377 (1969)]. Its morphology is quite similar to that of the carbon nanotube as is the way to examine crystal structures by electron diffraction patterns of carbon nanotubes. These studies were a part of my PhD thesis work, which was conducted under the supervision of Prof. T. Hibi at Tohoku University some 25 years before the discovery of carbon nanotubes. Another important experience was the crystal structure analysis of chrysotile asbestos with a tubular structure similar to carbon nanotubes. This work, however, was not my own - a senior research in Prof. Hibi's laboratory was studying this mineral, so I was aware of this tubular mineral structure [Yada, Acta Cryst. A23, 704 (1971)]. When I saw carbon nanotubes under the electron microscope, my old experiences came immediately to mind and helped me to figure out what they were, as I mentioned previously.


 After going back to Japan in 1982, I joined the government research project ERATO, on "Ultra-Fine Particles" (presently nano-particles or nano-crystals are commonly used), and worked there for five years. We produced a large quantity of ultra-fine particles of approximately less than 10 nm in diameter. We followed the physical evaporation method for particle generation, that is to say, materials were evaporated in an inert gas atmosphere using arc-discharge [Iijima, Jpn. J. Appl. Phys., 23, L347 (1984)]. The materials we dealt with were Si, Sic, Al2O3, SiO2 and other metal oxides. It is interesting to note that the arc-discharge evaporation set-up was exactly the same as equipment used to make carbon nanotubes and fullerene molecules in the present day. My major task in this project was to develop a new high resolution electron microscope accessible for characterizing ultra-fine crystals without exposure to air. One successful result was the discovery of the structural instability of small gold crystals [Iijima et al., Phys. Rev. Lett., 56, 616 (1986]. I found a gold cluster specifically composed of less than 1000 gold atoms changing continuously its external form, as well as internal atom arrangements. The phenomenon was thought to occur under electron beam irradiation during microscope observation, and is one of the interesting subjects in nano-science. The size of gold crystals in this case was less than 2-3 nm, and it resembled the metal catalyst particles used for growing carbon nanotubes. Once again I would like to emphasize the technique for high resolution electron microscopy and nanometer-sized crystals that are common important issues for nanotube research.
*There is movie in "Research" page.


 In 1990, there was a great deal of excitement in the world communities of condensed matter physics, chemistry and materials science caused by two discoveries. One was the discovery of high-Tc oxides in superconductivitiy and the other was the discovery of the mass production of fullerene molecules and their crystallization, which showed superconductivity shortly thereafter. These great movements attracted many researchers to the MRS meeting held in Bostonin early December of that year. I also attended the meeting -not the fullerene session, but a session on electron microscopy. After finishing my session, I moved to the fullerene session, which lasted until midnight. In the meeting hall I met Prof. H. Kroto whom, by that time, I was acquainted with since he had visited me at the NEC lab. to see my electron micrographs of graphite balls. He liked one of those micrographs showing a fullerene-like feature and we discussed it as good evidence of a succor ball model for the C60 molecule that Kroto and Smalley had postulated at that time [Kroto et al., Nature, 318, 162 (1985)]. In any event, he strongly urged me to work on graphite balls because of my previous experience with carbon materials. After returning from Boston, I started to search for carbon materials of various forms in order to elucidate the growth mechanism of fullerene molecules in terms of high resolution electron microscopy. I thought the clue was hidden in the growth of graphite balls, which were much larger in size than fullerene molecules although their formation mechanism should be the same.


 For this purpose, I collected and examined carbon materials from various sources, such as dehydrated polymer resign, vacuum evaporated amorphous carbon films, activated carbon, grassy carbon, carbon black, soot produced for fullerene production, and the anodes and cathodes of graphite carbon used for fullerene production and so on. In June 1991, I saw some peculiar elongated objects lying on the electron microscope grid though the binoculars of an electron microscope: they were needle-like features coexisting with graphite balls all over the grid. My primay search for the graphite balls was accomplished, but my eye was caught by those elongated filaments. They reminded me of my old experiments with silver filaments and chrysotile asbestos. It was certainly an exciting moment in my research carrier. These objects were multi-wall carbon nanotubes, and the results were published in "Nature" magazine in November 1991 [Iijima, Nature, 354, 56 (1991)]. It was very beginning of prosperous year of carbon nanotube research, which took place at my laboratory in the NEC Tsukuba Research Laboratory, where I started work in 1987. The second discovery followed, with the finding of single-wall carbon nanotubes in 1993 [Iijima et al., Nature, 363, 603 (1993)] at the same laboratory. This encouraged even more researchers coming into the carbon nanotube community.


 17 years have passed since the discovery of the carbon nanotube. More recently, it has stimulated another carbon research area where "graphene" has reappeared as a new research subject in condensed matter physics. Industrial applications of carbon nanotubes seems to be in progress, and in the near future we will see some commercial products.

Sumio Iijima [Premi Balzan 2007]