Rationally designing plasmonic devices using imaging and simulation

Hank Hogan, hank.hogan@photonics.com

Researchers intent on building plasmonic sensors to a set of specifications now have a new tool to help in that quest, according to a group of investigators. The team has demonstrated that electron tomography and electrodynamic simulations can create an accurate optical model of highly irregular gold nanoparticles, the heart of many plasmonic sensors.

The method could prove a powerful aid in designing devices to detect the faintest of substances, said researcher Eduardo A. Coronado. “The ultimate goal is to have a very precise optical characterization of real nanoparticles to address their capabilities to enhance the spectroscopic signal of nearby molecules, which is useful in ultrasensitive detection.”

Coronado, a chemical sciences professor at the National University of Córdoba, headed the part of the group in Argentina. Other members were from the University of Cambridge, where an electron microscopy research team was led by Paul A. Midgley.

By taking 2-D electron microscope images (left) at different angles and combining them, researchers create a 3-D electron tomogram that reveals the real shape of a gold nanoparticle (right). This provides accurate information for simulation of its optical properties by the discrete dipole approximation (DDA) method without any assumption about its shape or morphology. Courtesy of Eduardo A. Coronado, National University of Córdoba.

Gold and silver nanoparticles enhance faint Raman spectroscopy and fluorescence microscopy signals. In the case of the first, this can be by many orders of magnitude, sometimes a hundred trillionfold. The effect is not nearly as great for fluorescence but is still significant. These are just some examples of the ability of nanoparticles to potentially control, manipulate and amplify light, making them of great interest to researchers and possibly of considerable commercial importance.

Achieving the right response from nanoparticles requires understanding the relationship between their structure and optical properties, both in the far- and near-field regions. The latter is particularly important when it comes to determining the precise enhancement of a signal and what is needed to invoke it.

There are some analytical methods that yield an exact answer to the question about the relationship of a nanoparticle’s size, shape and composition to its optical behavior. But, Coronado noted, they apply only to nanoparticles of a regular shape – such as a sphere – that are tens of nanometers in diameter. They don’t work for irregular shapes and smaller nanoparticles, two categories that could be the majority in a given sample.

The solution demonstrated by the researchers in a Nano Letters paper published earlier this year involves the combination of two technologies. The first is electron tomography, a form of electron microscopy. In it, two-dimensional images sensitive to nanoparticle composition are captured at different tilt angles.

From this data, the investigators reconstructed a highly accurate three-dimensional picture of the gold nanoparticles; e.g., they achieved nanometer resolution of the features of triangular particles measuring 100 nm or more on a side.

The second technology is electrodynamic simulations. The investigators modeled the irregularly shaped gold nanoparticles, approximating them as an array of discrete dipoles for the purpose of predicting their optical response.

Coronado noted that this modeling approach has been shown to be accurate to within 5 or 10 percent when used on nanoparticles that are simple spheroids. Applying it, however, requires accurate 3-D information about the object being modeled.

Being able to predict a nanoparticle’s optical characteristics could prove useful. Not only could the right shape, size and composition be designed for a given situation, but the right illumination and working distance also could be determined.

One application might be to mount a nanoparticle on the probe tip of an atomic force microscope. That would then be used to detect the sequence of bases in a strand of DNA molecule by molecule.

Such a capability could be a boon, Coronado said. “This kind of technology would, for example, speed up considerably the process of DNA sequencing, which now takes a lot of time and effort using PCR [polymerase chain reaction] chemical procedures.”