Nanoshells are (generally) spherical particles with a core of one material overlaid by a shell of another material, measuring ~100nm. As part of a summer project I worked at the University of Leicester to characterise nanoshells with a silica core and gold shell using a spectrophotometer (in this case a rather antique piece called a Perkin-Elmer Model 330 - it was produced and discontinued before I was born, and communicated with microcomputers through RS232 at a blazing 9600 baud!).
These nanoshells exhibit plasmon resonance (i.e. absorbs the energy of light incident on it, instead of reflecting it or transmitting it) with a frequency that depends on their geometry. Hence we could engineer the resonant frequency by changing the core radius or shell thickness. This behaviour is described classically by Mie theory (basically an extension of Maxwell's equations for layered spheres).
The final aim of the project is to produce nanoshells that have plasmon resonance at ~2000nm in the infra-red range, and to attach them to cancer cells (the gold coating is functionalized [chemist speak: made so we can attach stuff that does like it to it] with thiol [basically sulphur which likes to bond with loose gold atoms] terminated groups whose other end is some biological molecule which attaches to cancerous cells). Then a laser would be shone onto the cancer cells, the nanoshells would resonate absorb the energy, heat up and kill the cells. 2000nm was chosen since water does not absorb with at that wavelength so surrounding (healthy) cells and tissue would not be damaged
That is in the future though, and my project involved conducting readings on the spectrophotometer and comparing it with Mie theory predictions to (gu)estimate the core radius/shell thickness because the synthesis method we used is not very precise and produces nanoshells of varying radius/thicknesses. Once we could be sure of the size of nanoshells we were making then we could refine the techniques to make the sizes we wanted. Of course the ideal wave to find the size of the nanoshells would be to use something like a scanning tunneling microscope - but the Department of Physics and Astronomy doesn't have one of those - instead we had a rather cheap atomic force microscope which produces some pretty pictures but not really and useful data.
The shells were synthesised by chemical reactions which are summarised here, and discussed in detail in my report but the short story is that it involved pouring apparently arbitary amounts of chemicals (some very expensive - tetra-chloro-auric [gold!] acid costs about 100 pounds per gram) into beakers, stirring it (mechanically of course) for hours and hoping it turns some curious shade of blue-green.
The results are:
The error bars indicate the spectrophotometer data and solid lines indicate best fits using Mie theory. The fitting program using a non-linear regression algorithm and the GNU Scientific Library could do with a bit of an improvement, but I think the problem is mostly with the range of nanoshell sizes which tends to blur the resonant peaks. Unfortunately this is due to the chemical synthesis methods - and I'm not a chemist - so I wasn't able to go much further with the project.
The writeup in pdf is here
