Introduction
What did you do before the discovery of nanotubes?
Sumio Iijima in his office at Meijo University.
I was a student at the University of Electro-Communications in Tokyo, where I graduated in engineering, but I found that boring, so I moved to Tohoku University in Sendai where I studied pure physics, which I found more interesting, and I went on to receive a Masters degree and then a PhD in 1968. After two more years at Tohoku I went to Arizona State University in the US where I worked in electron microscopy for 12 years. Then, in 1982, I returned to Japan as group leader of an ERATO project to develop a high-resolution electron microscope to study clusters, before joining NEC in 1987. Now I am based at Meijo University in Nagoya, but am also director of the AIST Research Center for Advanced Carbon Materials and a senior research fellow at NEC in Tsukuba1.
How do your experiences of working in the US and Japan, and in universities and industry, compare?
During my research career, including the 12 years I spent in the US, I met many teachers. Many Japanese scientists only ever meet one such teacher, but I met at least four in my career. Moreover, every time I have moved, I have changed subjects. John Cowley2 was my boss at Arizona but he did not interfere with my research. In fact we only published three joint papers in 12 years — it was an excellent environment. As an electron microscopist my top priority is to have the best instrument and I do not care where I work as long as I have the best equipment. When I came back to Japan in 1982 to join the ERATO project, which was funded by the government, I was based at Meijo University. After finishing the project, I thought about moving to a university in Japan on a permanent basis, but they could not provide me with the best instruments so I approached several companies and NEC agreed to take me on and provided the environment where I could conduct world-class research. Another reason for working in industry is that industrial research labs are likely to have interesting materials to study, particularly in electronics.
Was the discovery of carbon nanotubes accidental?
The discovery was serendipitous but not entirely accidental because I had accumulated a lot of experience in looking at short-range order in carbon specimens such as amorphous carbon and very thin graphite sheets, and of course glassy carbon was used as a test specimen for tuning high-resolution electron microscopes, so I was very familiar with carbon. The discovery of C60 molecules by Harry Kroto, Richard Smalley and co-workers in 1985 also motivated me. Indeed, Kroto visited me in Tsukuba around 1989 and told me that he and Smalley were having a few problems at the time because the carbon and cluster communities did not accept their papers on C60 because the results were based solely on mass spectrometry and they did not have any data on the crystalline structure. However, I had electron micrographs of 'onion' structures in graphite3, which I sent to Kroto, saying "this is what you have". So prior to 1990, my pictures were the only evidence for the structure of C60. Kroto suggested that I become part of the C60 community because people were already aware of my pictures. This was the actual trigger that led to the discovery of the carbon nanotubes.
How did you go from C60 to nanotubes?
High-temperature superconductors (HTS) had been discovered in 1986 and I was part of the 'fever' at NEC, working to determine the structure of these materials because I had experience of working on oxides as well. However, in 1990 I realized that electron microscopy, and therefore I myself, could not make any significant contributions to research on HTS, and I was very disappointed. So when the C60 fever started I was careful not to get caught up in it because I am a physicist and C60 belongs to the chemists. My only interest was how the C60 molecules are formed from the chaos in the carbon, so I became interested in studying the growth of C60 and all kinds of carbonaceous materials with electron microscopy. That was how I came to see carbon nanotubes for the first time in 1991. However, other NEC researchers had just reported a HTS with a then record transition temperature of 32 K, so no one was interested when I showed them the structures that are now known as multiwalled carbon nanotubes. Again I was disappointed.
Were you really the first?
There may have been someone else who saw the multiwalled nanotube prior to my 1991 paper but this paper included critical electron diffraction data that was important to define the crystalline structure, as well as the dimensions of carbon nanotubes. We have to be careful about the definition of discovery. In the case of carbon nanotubes, publishing one picture is not enough and not science.
What happened next?
I submitted my manuscript entitled "Helical microtubules of graphitic carbon" to Nature in August 1991, and it was published4 in November. That same month I also reported the discovery at a conference on clusters in the US. Initially, apart from Smalley, most of the interest came from theorists, and it took time before more experimentalists became involved. Single-walled carbon nanotubes were reported for the first time in back-to-back papers in the 17 June 1993 issue of Nature by Toshinari Ichihashi and myself and, independently, Donald Bethune and co-workers at IBM, although our paper was submitted about a month before the IBM paper5, 6.
What will be the main applications of nanotubes?
I have never exaggerated the potential of nanotubes because they were discovered 16 years ago but no one is making money yet, although more than 1,000 PhDs have been awarded for research on nanotubes. As regards to applications, the prospects for fuel cells are not very encouraging because nanotubes are only one part of the overall system and other aspects require extensive development to take full advantage of the properties of nanotubes. However, nanotube-based supercapacitors are more promising and we have recently carried out 'supergrowth' of nanotubes using inexpensive A4-sized steel foil and nitrogen gas (p593). We are aiming to lower the production cost to about $200–500 per kilogram, and Nippon Chemi-Con Corporation plans to use this material to make supercapacitors. In electronics, field-effect transistors that use single nanotubes still have a long way to go, and the integration of the nanotubes with peripheral devices is the major issue (see page 605). I am not very familiar with graphene but it may be easier to produce field-effect transistors with this material. In medicine, it is too early to predict if nanotubes will have an impact on drug delivery but carbon nanohorns7 are promising. Separating semiconducting and metallic tubes from a random mixture of the two also remains a challenge but a group at AIST led by Hiromichi Kataura and Kazuhiro Yanagi has just submitted a paper reporting a major breakthrough in this area.
How safe are nanotubes?
The Japanese government has started a big-budget five-year project to look into environmental issues such as the toxicity of nanomaterials. However, I am a little concerned about crazy reports published in American journals about the toxicity of nanotubes based on animal experiments. I am not convinced about the reliability of such experiments. Were the nanotubes the cause of death, or were impurities — such as the metal catalysts used during synthesis — responsible? This is an important issue. We need materials to be standardized before we start testing, so the various standards and protocols being developed by the ISO will be very important in this area.
Why do you give relatively few talks about nanotubes and nanotechnology?
I believe as a scientist I should not exaggerate the potential of my field. We should be modest and not become political about the technology and research at hand. This is one of the reasons I do not give many talks about nanotechnology.
Interview by Adarsh Sandhu
