The fin-based LED chip architecture could be a game-changer in terms of device efficiency.
A novel LED chip architecture developed jointly by the National Institute of Standards (NIST), University of Maryland, Rensselaer Polytechnic Institute, and the IBM Thomas J. Watson Research Center has the potential to be a real game-changer in terms of device efficiency.
The team has successfully demonstrated its concept in the lab and recently published a paper discussing its work in the Science Advances peer-reviewed journal. The new design promises the potential for an orders-of-magnitude increase in brightness, anywhere from 100 to 1,000 times that of current submicron-sized LEDs, making this design an attractive alternative for emerging technologies in a variety of applications.
By now we are all familiar with the degree to which LEDs have taken over the lighting industry, in large part due to their vast improvement in operating efficiency over earlier technologies. But a roadblock still remains that imposes a limit on just how efficient LEDs can be; a phenomenon known as “efficiency droop” that has long frustrated designers of products that use LEDs. Efficiency droop is caused by a decrease in external quantum efficiency (EQE), a measure of how efficiently the LED converts electrons to photons and how easily those photons are able to escape the LED materials. This decrease in EQE occurs with increasing operating current and is attributed to a rise of non-radiative recombination and an elevated p-n junction temperature which, in turn, impacts the recombination process in the diode’s active region.
The research team that developed the new architecture was working on physical designs different from the usual flat LED “chips” to provide a solution for applications like NIST’s NOAC (NIST on a Chip) technology. The end result was an LED source made up of zinc oxide “fins,” each approximately 5 microns long, combined into an array resembling a comb. The structure and materials used (ZnO-GaN) produce light in a band at the boundary between violet and UV. Via three-dimensional finite-difference time-domain (FDTD) modeling, the team was able to determine that the fins emit light from facets open to the air with an extraction efficiency of about 15%. In addition, spectral radiant flux measurements confirmed that the output power of this fin-based design increased linearly with increasing drive current, indicating that the factors that contribute to efficiency droop (electron leakage, Auger recombination, defect-related recombination, and temperature effects) were insignificant.
This diagram shows the structure of a “fin-based” LED.
The team attributes the elimination of efficiency droop to the physical geometry of the fin design. It is important to note that in an LED, the current in the n-type material must be equal to the current in the p-type material. Therefore, for planar LED designs in which the n-type and p-type materials are the same size, the dimensions of the recombination region are essentially constant regardless of injected current. However, in order for the fin LED to provide sufficient numbers of electrons to combine with the holes generated over a larger physical area, the recombination region expands further into the fin as higher current is applied, resulting in an elimination of efficiency droop.
Even more surprising, the team found that at drive currents above about 50 mA, the emission spectrum of this new design gradually changed from a broader output spectrum centered around 385 nm to two narrow lines at 403 nm and 417 nm. Based on their calculations, the team believes this transition takes place when the recombination region reaches the top of the fin and can, therefore, no longer expand with increasing current. At this point, the fin behaves like a Fabry-Perot cavity, enabling lasing.
LEDs have been incorporated into myriad applications beyond lighting, including displays, biomedicine, disinfection, and sensors and security systems. The main challenge up to now has been that efficiency droop limits output power to nanowatts, which, in turn, limits the performance of these devices. A fin-based architecture appears to be a viable solution to this problem, allowing for radiant power output in the microwatt range. And because of their high radiant power output, submicron LED and laser devices like this new fin design have the potential to facilitate yet another major shift in how LEDs are deployed in the products and systems we use every day.
This article was originally published on EDN.
—Yoelit Hiebert has worked in the field of LED lighting for over 10 years and has experience in both the manufacturing and end-user sides of the industry.