PhD candidate doing my research in new materials for photovoltaics here.
I'm sick and tired of all this mis-reporting. These are NOT 86% efficient cells. If they were, (and they were inexpensive) it would be the greatest discovery in 50 years and it would have been all over every newspaper in the world 2 weeks ago when this paper was published.
They simply absorb 86% of light that hits them. When you say a cell is X% efficient without qualifying it, it's taken to mean power conversion efficiency [PCE] (optical power in/ electrical power out) That and dollars per watt are the numbers that really matter. Read the Nature Materials paper that drove this and you'll see that theory says this design could be up to 17% efficient. That compares unfavorably to mid to high-end commercial cells on the market today.
I'm not saying that this research is a worthless endeavor, maybe they can hit the maximum theoretically possible PCE and keep the cost down. That might have real-world impact.
The caltech news brief quotes Atwater (the PI for this research) as saying that the photons are not only absorbed, but they're also convertedto charge carriers (which is a good step). The problem he doesn't mention here is, these charge carriers loose all their energy (voltage) before they exit the cell. Solve that problem and we've got a winner.
The fundamental issue with nano-structured designs like this is the surface area of the P-N junctions in them. Large surface area means high dark current which means low voltage output. Low voltage output means low PCE. Unfortunately, nothing in this research solves that problem.