Hacker News Reader

Comment by syntaxing | Hacker News Reader

To put things in perspective, the current theoretical maximum efficiency for solar panels is about 34% [1]. There are a couple assumptions that are made to derive this number so there are ways to exceed this maximum efficiency. Though the assumptions made in the equations are fairly application to most solar panels product (not the ones made in labs).

[1] https://en.wikipedia.org/wiki/Shockley%E2%80%93Queisser_limi...

View on HN

There are already 46% efficient solar cells, the 33.7% limit is for a specific type of solar cell (single p-n junction) which are cheaper to manufacture.

The maximum theoretical limit is ~86.8% for solar panels.

Is the maximum theoretical efficiency I often see stated just for type of technology we use today (pn junctions)? Or is it a hard physical limit for any light -> electricity system? Would it be possible that some new technology for solar cells could be more efficient?

I think the highest physical limit would be Carnot efficiency for a heat engine driven by the temperature difference between the sun and the earth. (6000K - 300 K) / 6000 K = 95%. There could be a different lower limit we hit before that though.

That does make sense for an absolute upper limit, thanks.

Yeah, the limit only applies to single junction solar panels which I believe most consumer level solar panels and solar farms are this type [something I heard during graduate school, I have no source for this on top of my head]. I'm pretty sure the solar panel in the article is a single junction solar panel since they use PECVD (which is just CVD with a plasma generated above the wafer to improve uniformity) which allows them to get better results.

Most double junctions require either MOCVD or even ALD but both technologies require longer fab time, have a higher reject rate, and the machines are almost 2X as expensive. Even then, the highest I've seen for solar panel efficiency product that you can purchase is around 32% and that is "state of the art" for today.

Single junction Silicon* cells.

Reply on news.ycombinator.com