On this page
Some discoveries need a new idea. Others need someone who knows exactly how to look. Sumio Iijima spent two decades learning to see individual atoms with an electron microscope, so when he examined a smear of carbon soot in 1991 he recognised something others had passed over. This post covers his path, what his famous paper reported, a worked example on the size of a nanotube, and what students can take from his way of working.
Who is Sumio Iijima?
The Kavli Prize biography gives the outline. Iijima studied at the University of Electro-Communications in Tokyo and completed a PhD in physics at Tohoku University in Sendai. From 1970 to 1982 he worked at Arizona State University on high-resolution electron microscopy, and in 1977 he was able to observe individual tungsten atoms. In 1979 he studied graphite with the electron microscope as a visiting scholar at Cambridge.
He returned to Japan in 1982, became a Research Fellow at NEC Corporation in 1987, and has been a professor at Meijo University in Nagoya since 1999.
What did he see in the soot?
The Kavli Prize page for 2008 says that after a conference where he discussed the new forms of carbon with pioneers of the field (see how Kroto, Curl and Smalley found C60), he returned to Japan to build on his earlier work. His breakthrough came on 23 June 1991, while he was examining soot made by passing an electric current between carbon rods.
His paper, Helical microtubules of graphitic carbon, appeared in Nature in November 1991. Its abstract reports needle-like tubes made by an arc-discharge method similar to the one used to make fullerenes, growing at the negative electrode. Under the electron microscope each needle turned out to be a set of coaxial tubes of graphite-like sheets, from 2 up to about 50 in number. On each tube the hexagons of carbon atoms wind in a helix around the axis, and the needles range from a few to a few tens of nanometres in diameter.
Was he the first to see a nanotube?
Not quite, and the prize citations say so. The Kavli biography notes that carbon nanotubes had been observed before, but that his paper generated unprecedented interest and a great deal of further research. That is an honest way to describe many famous discoveries: the credit often goes to the person who characterised a thing clearly and showed why it mattered.
The next step came in 1993. Iijima and Toshinari Ichihashi reported single-shell carbon nanotubes of about 1 nm diameter, growing in the gas phase rather than on the electrode. In the same issue of Nature, a team led by Donald Bethune at IBM described cobalt-catalysed nanotubes with single-atomic-layer walls. Two groups reached the same kind of structure independently.
In 2008 Iijima shared the Kavli Prize in Nanoscience with Louis Brus, who created the field of colloidal semiconductor nanocrystals. The prize page also notes that he received the Balzan Prize for nanoscience in 2007.
Worked example: how wide is a nanotube with many walls?
You need one number: the spacing between neighbouring graphite layers. The Nobel Prize popular information on graphene says one millimetre of graphite holds about three million layers, so the spacing is 1 mm ÷ 3,000,000 ≈ 0.33 nm. The same page notes that a rolled-up layer of graphene forms a carbon nanotube. Assume nested tubes sit about one layer spacing apart.
Problem. A needle has an innermost tube of diameter 2 nm (a sample value) and 10 tubes in total. Estimate its outer diameter.
Each extra tube adds one spacing on each side, so the diameter grows by 2 × 0.33 = 0.66 nm per tube.
From the first tube to the tenth there are 9 steps.
Outer diameter ≈ 2 + 9 × 0.66 = 2 + 5.94 ≈ 7.9 nm.
Check against the paper. With 50 tubes and the same 2 nm core, the outer diameter is 2 + 49 × 0.66 ≈ 34 nm, which fits "a few tens of nanometres".
One more. A single-shell tube 1 nm across has a circumference of π × 1 ≈ 3.1 nm. That is the width of the strip of carbon that was rolled to make it.
Common mistakes:
Adding the spacing once. A new shell surrounds the old one on both sides, so it adds two spacings to the diameter.
Mixing units. Crystal data are often in ångströms (1 Å = 0.1 nm). Convert before you add.
Treating nanotubes as fullerenes. Both are curved carbon, but a fullerene is a closed cage while a nanotube is a long tube.
What can a student take from how he worked?
Iijima's discovery rested on years of practice with one instrument. He knew what graphite looked like at atomic scale, so an unusual pattern stood out. He also looked at what others treated as waste: the deposit on the electrode, not the product people wanted. And the citations show that clear description and careful measurement can matter as much as being first.
Shapes like nested tubes are hard to picture from text alone. The Fermi Simulation Lab has interactive simulations for ideas that are easier to see than read, which can help when a structure will not settle in your head.
Try it now: roll a sheet of hexagon-printed paper into a tube and count how many hexagons fit around it.
