Technical Insight
Multiple factors are blamed for droop in blue LEDs
Theorists point the finger at carrier leakage, compositional fluctuations and threading dislocations
Calculations from a US-Italian partnership offer a new view on the role of Auger processes in droop, the decline in LED efficiency as its current is cranked up. The researchers from Boston University and Politecnico di Torino, Italy, claim that indirect Auger processes are only a contributing factor to droop in blue LEDs, and play a bigger part in green and yellow variants.
The conclusions of this team’s work differ from the findings of those of Chris Van de Walle’s team from the University of California, Santa Barbara (UCSB) – this is the only other group to have performed and reported the results of a full-band calculation of Auger rates in InGaN. Both teams agree that direct Auger recombination is very weak in bulk InGaN and indirect Auger processes are far stronger. How much stronger is the crux of the debate.
The west coast team claims that these Auger-based processes are the primary cause of droop for all LEDs emitting from the ultraviolet to the yellow, while the US-Italian collaboration argues that they are only one of several important factors. Team member Francesco Bertazzi from Boston University and Politecnico di Torino says that the strength of their approach is the combination of the incorporation of realistic electronic structures and ab-initio phonon dispersion relations. “The effects of phonons are formally included to the infinite order by means of a spectral density function, thus avoiding the divergence problem inherent in second-order perturbation theory.”
When the team performs its perturbation theory calculations, they find that phonon-related processes involving two electrons and one hole have a broadly similar strength to those involving two holes and one electron. Those involved in the debate over the origin of droop number far more than just the US-Italian partnership and the UCSB theorists. Many other theorists and experimentalists have worked on this topic, and many of their claims over the origin of droop can be divided into two camps: Auger and electron leakage.
However, arguing that it can only be one of these two causes is fundamentally flawed, according to Michele Goano from Politecnico di Torino, for the same reasons it was wrong to argue more than 20 year’s ago that just one process could explain the behaviour in InGaAsP alloys. Back then Witold Bardyszewski – who is now at the University of Warsaw, Poland, but was then at the University of Lund, Sweden – co-wrote a paper explaining that Auger recombination can promote carriers to higher energy states, where they can undergo carrier leakage. In other words, it is impossible to treat leakage and Auger recombination as two distinct, unrelated processes.
“In our opinion, carrier leakage is one of the main causes of droop in blue LEDs, but several other mechanisms come into play, from compositional fluctuations to threading dislocations and so on. Each may be important, depending on the LED structures,” explains Goano.
He and his co-workers are now trying to refine their calculations so that they can determine Auger coefficients in quantum wells, rather than bulk material. Results could be different, due to the breaking of the conservation of momentum along the confined direction.
F. Bertazzi et al. Appl. Phys. Lett. 101 011111 (2012)
