Bright idea for nanosensors

The transmission of light can be affected by the suspension of metal particles in a clear medium, an effect that has been used for centuries in making red stained glass using gold dust. Researchers are now exploiting this property to construct nanosensors that could be used to detect explosives or toxins, or identify infections.

Lead researcher on the project, Stefan Maier, professor of nanophotonics at

, said: 'Small metallic structures are well known to work efficiently as light absorbers, scatterers and concentrators in certain frequencies. The red glass that you see in churches is normal glass doped with small gold particles. Because gold particles absorb and scatter preferentially in the green, the green light is taken out of the white light coming through so that it looks red.'

Scientists have for some time used techniques such as electron beam lithography to create nanoparticles of a very specific shape that could be tuned to the resonant frequency where there is preferential scattering. The problem is that all these resonances are usually quite broad, and they get broader the more the resonance is pushed into the near infrared. This is because scattering increases as the particle size increases. For sensing applications, the resonances need to be less broad so that small shifts can be more easily detected.

To overcome this, the researchers are using two similar metallic structures, a flat disk and a ring, next to each other so they can interact. 'The disk by itself shows limited transmission as it has a broad resonance, absorption and scattering,' explained Maier. 'The ring has a very narrow, higher-order resonance that cannot be excited if it is just by itself. But when you put the disk next to it, it can be excited, it destructively interferes with the broad transmission of the disk and gives a narrow frequency band in which the total transmission is increased.' The frequency where the interaction occurs is also highly sensitive to the environment around the particle. A bare gold surface transmits in green frequencies, but if molecules bind to the gold surface or if the gold surface is covered with glass, for example, that resonance shifts to the red by a small amount.

'This way, you can detect if something binds to your particle without having to rely on a fluorescent process or another active detection method,' said Maier. 'You coat a metallic structure with a trap — an antibody that combines with an antigen or something that combines with the molecule you want to sense. When that binds, the bioelectric environment changes, the light refractive index close to the surface of the metal gets a little bit higher and the frequency of transmitted light shifts.'

So far the researchers have demonstrated experimentally that the concept of this quenching of scattering works using silver ring and disk nanostructures interacting at a distance of around 10-15 nanometres. A computer simulation has shown that the overlap of sharper resonance and broad resonance where the two interact is very sensitive to molecules that bind to it.

'The next step will be binding different molecules to the surface of our nanoparticles,' added Maier. 'So far we have done the fabrication and the spectroscopy, and studied the fundamental physics of the structures by themselves to show this very pronounced decrease in scattering as you would expect from electromagnetically induced transparency.'

In use, the nanostructures would be suspended in a fluid channel with a light source on one side and a simple transmission frequency measuring device on the other. 'Due to the interaction, the transmission in a very narrow window would increase simply because the scattering is reduced, so you would get more light through,' Maier explained.

Detection depends only on the refractive index contrast of what binds, so the system could be used to detect almost any molecule so long as the surface of the particles can be functionalised to bind to the substance being detected. Silver nanostructures have been used to date, but the researchers plan to use gold for sensing as the surface chemistry of gold is well known.

Collaborators in the project are European nanotechnology research centre

, Leuven University in Belgium and

in the US. IMEC and Leuven did the fabrication, Rice carried out the theory and Imperial performed the experiments. The project is at too early a stage to have involved any industrial partners.