Metamaterial Discovery the Combinatorial Way
British engineers perfect a way to characterise the properties of new photonic metamaterials using the brute force methods of combinatorial chemistry
kfc 09/07/2010
- 1 Comment
The ability to make and test thousands of different but structurally related molecules has had a huge impact on the pharmaceutical industry. These days, it's not uncommon for a company to make (or simulate) tens of thousands of similar molecules in an attempt to identify one in which the required biochemical activity is optimised. This brute force method, called combinatorial chemistry, is also used to identify new catalysts, light emitting materials and electronic devices.
So it's not really a stretch to imagine the same technique being used to discover new metamaterials with unusual electromagnetic properties and indeed that is exactly what Eric Plum and pals at the University of Southampton today suggest.
Their idea is to create an array of metamaterial samples in which various design parameters vary in a regular way. They then test each new material with a view to finding one in which the properties are optimised for the kind of photons the team want to manipulate.
They test the idea by creating an array cells each containing a new metamaterial made of gold square split ring resonators and in which the size of the squares varies from cell to cell in steps of 25 nm.
They then measured the reflection and transmission properties of each cell and compared the results to computer simulations of their properties.
What these guys are looking for are resonance effects that occur at very specific frequencies and which can be fine-tuned by varying factors such as the size of the split rings and the number and thickness of extra layers. These so-called Fano resonances are important because they act like optical switches that are flicked on by light of a specific frequency.
There's one problem though: small but inescapable variations in the manufacturing process inevitably change the resonances. But by comparing the simulated and experimental performance, Plum and co were able to characterise the changes caused by manufacturing variations.
That's a potentially lucrative ability. Engineers expect to use Fano resonances in future generations of electronic devices as everything from ultracompact nanoantennas to optical memories. And that means there's a potentially huge payoff at stake for anyone who can master the technology.
Ref: arxiv.org/abs/1009.0391: A Combinatorial Approach To Metamaterials Discovery
On Twitter
Become a Fan on Facebook
Subscribe to the Feed
flared0ne
395 Comments
One common criteria applied while reading through "here"
tends to be "how can this be added to the 'toolbox' of functional technologies we can use to DO things?" As well as "hmmm -- how might we use this BESIDES the 'obvious' applications?"
One analogy that comes to mind for the "process variation" issue is the post-production laser trimming of R2R networks to "normalize" and tweak variations in ADCs and DACs... The problem would seem to be that "post-facto" tweaks probably sacrifice already-'iffy' "Q" (discriminator width) characteristics.
The mental image I get is of peripheral "buds" off the ring resonators which are "tweaked", perhaps via embedded optical pathways -- or maybe something similar to a DRAM "refresh" strobe happening at a rate fast enough to sustain the "adjustment" with a minimum "droop" in tuned characteristic.
NOT clear on the available "tweak" mechanisms, but wonder if extremely localized changes (in thermal, dielectric, magnetic (?), or impedance characteristics) are the only options?
Again, the mental image is of a ring where one quadrant varies smoothly from "normal" width to "thicker", with the "adjustment" taking place under that extended "shelf" area. Perhaps as an adjacency overlap of a spiraling ring conductor, with an adjustable-gate "shorting mechanism", for impedance changes, or an underlying dielectric and opposing-plate... One risk: introducing an eccentricity to the material behavior in the axis of that quadrant.
Nice -- don't HAVE to "know" all the details in order to visualize a few of the "perhaps" mechanisms available to "tweak" things. Really only works in a forum like this, though. With a low background noise level... (So far, anyway)
Reply