[Rnonymous]

Well first of all this claim is quite a bit exagerated, as a single QD could produce twice more electrons in some situations, but only in some. A process called MEG - Multiple exciton generation.

Let me explain a bit simplified. Normally in semiconductors, when a high energy photon is absorbed, excess energy is converted into heat, and you end up with an electron with energy equal to the bandgap of the semiconductor.

With QDs you could have two QDs in close proximity, and when one of them absorbs a photon with energy more than twice the bandgap, there is a possibility that instead of only one of them getting excited with excess energy, two excitons will form of half the photon energy in each of the two QDs. Resulting in double the amount of excited electrons (A doubling of the electrons!). If this was really efficient it would mean that now you need to optimize your PV material bandgap differently. And instead of a theoretical maximum efficiency (Shockley-Queisser) for single bandgap solar cells of ~32%, you could theoretically get 45% with a single bandgap.[1]

[1] https://www.sciencedirect.com/science/article/pii/S000926140...

[hilbert42]

Thanks for the info. My abrupt comment was because—as with so many articles like this—bold claims are made without explanation or reference and it is particularly common in the semiconductor world whether it be LEDs, solar cells, camera sensors or even semiconductor materials themselves.

I've worked on the periphery of this stuff for years, by that I mostly mean the applied side, and it's so hard to cut to the chase and get overall input/output efficiencies, etc. One has to do searches and wade through research papers just to get some basic info and it's very tedious and ought to be unnecessary.

For example, say I want a spreadsheet-like table that compares overall efficiency of a range of white LEDs using practical or in-use materials versus the theoretical maximum efficiency for same LEDs (materials under ideal conditions) versus theoretical efficiencies for materials that have theoretical† potential but have yet to be fabricated—and then those efficiencies expressed in terms of power in (electron Watts) versus photon Watts out.

A similar scenario is trying to find the efficiencies of image sensors to do comparisons between them, for instance, photons in (captured) versus 'modulated' (signal) electrons out(put) versus S/N—noise of various kinds, shot noise etc.

The same goes for solar cells. Even though the SQ limit is reasonably well known, it's hard to do comparisons between single junction devices let alone those with multiple (stacked) junctions. What's the inter-junction (material and coupling) losses etc? How do losses add up in multiple junctions for practical devices versus their theoretical counterparts? And that's long before we start worrying about trying to develop and or optimize any MEG-like action.

Unless one is on the cutting edge of research or closely involved in developing the technology then it's nigh on impossible to get info like this without months of research. Frankly, it's damn annoying. Moreover, much of the info is either unavailable for various reasons, or hidden behind publishing firewalls and or locked up in proprietary environments—companies such as Sony which make image sensors, etc.

Some would ask why would one want such info, well it could be any number of reasons, trying to determine the state of the art versus the theoretical and trying to figure whether a particular tech is worth pursuing or whether another approach is needed, and so on.

It used to be much easier. For instance the RCA Electro-Optics Handbook was a wealth of information but there's nothing like that available now to my knowledge.

I've been around long enough, that's when one could go to say the RCA Electro-Optics ref and just look up the quantum efficiencies of, say, trialkali photocathodes as in photomultipliers, orthicons, and isocons etc. This stuff is still very important, one needs to know the probability of a photon being captured by a photomultiplier in say a neutrino detector—and it's good to know that this info is easily available.

Come semiconductors however and it's just not. The collated info either doesn't exist or it's in a damn mess.

____

† I've an old book on semiconductors but I can't immediately put my hands on it so I can't give you an exact reference at present but it has a comprehensive table comparing electron mobility in different semiconductors. It compares practical materials with those that haven't yet been able to be fabricated against theoretical ones that are unlikely to be practical.

This was extremely informative, before reading the book I used to think gallium arsenide had one of the fastest mobilities but many others beat it hands down, as for silicon it's positively glacial.

Until one sees tables such as this it's almost impossible to grasp the whole picture. This is precisely what I am complaining about.