New perspectives on the very small
Nanotubes, ranging from 2 to 50 nanometers wide and 50 nanometers to several microns long, are consistently regarded as small. Thinner than the wavelength of light, several hundred make up the width of a hair. They're small, except when they aren't. It's a matter of perspective.
Nanotubes are like planks, both are carbon based, give you nasty splinters and only really shine when you control their assembly. Nail planks together at random and, occasionally, something resembling a shed may emerge. But to build something specific, rather than get lucky, you need to cut a joint here, mount a hinge there, and fix a myriad of fasteners along fairly precise lines.
Is this possible with nanotubes and chemistry? Open a chemist's toolbox and you find molecules and bonds rather than screws and nails. Unhindered carbon-carbon bonds can be useful hinges. An azide functional group can provide half of the glue holding things together.
Having the parts is half the battle, placing them exactly is another matter. To understand this adjust your perspective to where small is very big indeed. If molecules were human, the nanotube would be as wide as the Empire State Building. The height would exceed a stack of 200 Empire State Buildings. Standing on such a stack the surroundings are near indistinguishable from space.
Placing our molecule is like ordering it to choose a building, and wave from one of the windows. Each building has 6,500 windows, making in excess of 1 million places for our molecule. Once established, verifying its location means spotting which window, of millions, has a waving molecule. This molecule won't be alone so you also need to check the colour of its shirt. If the job of checking hundreds of sky scrapers and millions of windows, for a handful of waving molecules in snazzy shirts is daunting consider another sense of scale. If you’re trying to analyse these nanotubes from the palm of your hand then it is like trying to spot waving and snazzy shirts from orbit.
This may be difficult but science has a way of making difficult into run of the mill. Conceptually the systems required to achieve this are all in place. The reference data has yet to be compiled, and that means large quantities of tedious double checking and cross-correlation. However with enough thought and with the design of elegant experiments it should be enough to tease out proof that it can be done. What I am trying to do is just this: attach my molecules to my nanotube in just the place I want, figuring out if it worked, and then using it to join nanotubes together intelligently. Whether this can be achieved in a practical manner, out in the real world and inside the time-frame of a PhD is yet to be seen. I may not be able to place molecules on a nanotube as accurately as I'd like, but that doesn't mean I shouldn't spend years trying.