2010-11-30 23:17:59Series on solar cell physics?
Mark Richardson


I'm not sure whether this is appropriate for SkS but John sounded interested, so I wanted to know what all y'all think. I spent a year in a solar lab doing my masters. I plan several articles.


The current script I have for 1)  is this. I would re-write it and add pictures to break it up but I'll only put the effort in if people think it's something SkS could cover.

I'm also not sure about the level; I can confidently go up to 'advanced' level, but I think condensed matter physics is a pretty exclusive club so I've gone for simpler language. First draft with no pictures follows:

 (800+ words atm, I want to simplify to 600 if possible, feedback would be great!)


The Physics of Solar Cells


As SkepticalScience has moved into solutions it seemed a great opportunity to share some of the physics behind solar cells.

To start, we should know that a solar cell works by sticking together two slightly different semiconductors that form a voltage which will push around any electrons that are knocked off their atoms (e.g. by light), and therefore generate an electric current.

In quantum physics some setups are ‘allowed’ and some aren’t: electrons in sodium are only ‘allowed’ to be in certain places, which are different distances from the nucleus. You can move electrons between two states that happen to be the right distance apart to give off light with a certain energy, and in sodium this is the orange you see in streetlights.

Semiconductor crystals are similar – some energy levels are ‘allowed’ and some are ‘forbidden’. The highest filled energy level is where electrons are stuck to their atoms. Then there is a ‘bandgap’ where electrons can’t go, and above that a ‘conduction band’ where electrons can move about after they’ve been knocked off their atoms.

Once an electron is away from its atom, it moves around the crystal until it finds a ‘hole’ around another atom that it can drop into. However, if you apply a voltage, you can whisk it away before it finds a hole and use it to do work.

Crystals have a property that individual sodium atoms don’t: because there are so many atoms all interacting with each other you end up with very regions that electrons are allowed to be in. Whilst atoms like sodium and hydrogen can only absorb specific colours, crystals can absorb a broad range of energies.

Once an electron has been knocked off its atom by a high energy photon it bumps into other electrons in the crystal and these collisions can make the crystal wobble, sucking energy away from the electron. If it can’t find a hole to drop into it won’t fall out of the conduction band, but it loses energy until it’s at the border of the forbidden region and can’t lose any more.

Solar cells can absorb light that has more energy than its bandgap, but the electrons will bounce around and lose this energy until the only electrical energy you can get out again is the bandgap energy. It might absorb blue light, but only ever give out the equivalent energy of red, for example!

This means the bandgap is very important; any less energetic light is too weak to knock electrons off their atoms. Any light that is far more energetic than the bandgap might knock an electron off, but the electron will then lose a lot of this energy through collisions before it can be used for electricity.

The cheapest and most abundant semiconductor was silicon: and it has a bandgap equivalent to infrared light. It can absorb higher energy visible light like nice white sunlight, but it loses a lot of that energy before it can be turned into electricity so the theoretical maximum efficiency is about 29%.

There are other materials like cadmium telluride which have a bigger energy bandgap and they happily absorb visible light and waste less of it as heat before you can extract the electricity. Unfortunately, the bandgap is so big that infrared light doesn’t have enough energy to knock electrons off the atoms so it can’t be used at all. However, the balance works out better for CdTe: it can turn up to 33% of the sunlight into electricity.

Indeed, the cheapest solar cells in the world are cadmium telluride ones produced by FirstSolar, at $0.76 per Watt of peak power, compared with the cheapest silicon solar cells from SunPower at $1.08 for the same output. However, SunPower’s modules are actually higher efficiency than FirstSolar’s. What’s going on? Please come back for the next article where we’ll try to make sense of this.

2010-12-01 08:44:08


MarkR, I hate to discourage you. However, I have to wonder who the intended audience is? And how does learning about the physics of solar cells shed any light on what's happening with the climate?

 If you want to focus on a part of the solution, doesn't this relate more to the possibilities and limitations of practical application, rather than solid-state physics? Isn't there the danger of fuzzing up the image of SkS, so that it begins to be perceived as a generic science site?

Just asking...

2010-12-01 17:42:54Off-topic post
John Cook


I confess, I do find it interesting, not least because I just put solar panels on last Wednesday and find the technology quite fascinating. I'm thinking I do an intro to this first post, just saying that yes, this is a change of pace for SkS but having just installed solar panels on my house and Mark suggesting this post around the same time, was interested in learning more of the technology.

Mark, what is the subject of the other posts? Are there any particular "myths" about solar power that you conceivably tackle in any of them?

Who are FirstSolar and SunPower? Are they solar power companies in the states?

2010-12-01 20:31:14

This post is somewhere in between the problem (climate) and the solution (energy production). It's true that it's a more generic scientific issue but once in a while it won't hurt.

I think your description of how a semiconductor works is a bit confusing. I know, it's not easy to explain solid state physics in a blog post but that's why you need to go step by step from the atomic energy levels (which I guess/hope/assume readers know) to the energy bands in solids, to semiconductors, to conduction in semiconductors and charge separation under illumination. These concepts are necessary if you want people understand how a solar cell works.
It's a series of post, afterall, the first should be more focused on semiconductor physics. As John asked, it would be helpfull to have the outline of the series.
2010-12-01 20:33:14
Mark Richardson

I'm thinking another blog might be the best place to put the introductory ones on physics...



There are several silly little myths I've heard about solar panels that I wanted to deal with later: they take more energy to make than they deliver; they are made with materials that are too rare to ever be of any use; total output can't be enough; they cause more global warming by cutting albedo; etc. Many of these will be put in my summary article so I'm thinking perhaps I can see if someone else wants to put up the series once it's finished and cross post the myths one to SkS.

In which case any feedback here would be nice!



And John: they are both yank solar companies, I don't know too much about sunpower (we worked on the First Solar tech) but FS produce modules for the commercial market and don't sell to residential afaik, so they aren't that well known. In terms of output they were the second biggest producers in the world. In 5 years it has gone from ~50 MW annual production to 1400 MW and plans to double production by 2012.

 Also, as well as producing cells much more cheaply, they have much bigger profit margins than SunPower. $600m profit from $2b sales versus $60m from $1.5b. I suppose that's what you get for only having one major company producing CdTe...

2010-12-02 11:09:37Solar myths
John Cook

Mark, just throwing this out there - would it be feasible to reframe the content of your article/s so they address solar myths? For example, the above post about semi-conductors could be integrated into a rebuttal of the "materials are too rare" myth. That way, you get to communicate the science but also address a myth in the one go. I know it's not exactly what you had in mind but just a thought - you get to tick a few different boxes plus your content will go straight into the iphone/android apps as they'll be part of rebuttals.
2010-12-05 05:02:22agree with John
Dana Nuccitelli

I'd be interested to see some solar cells articles, but for this site they would be most appropriate if integrated into rebutting various solar-related myths.  But if you're going to do several of these, I could also see doing an introductory 'guest post' sort of thing to introduce the basics and set up the series of rebuttals.  That would just be a blog post, while the myth debunkings could be made into rebuttals.

If you want to pursue the introductory post idea, I would add to the introduction by saying that there are a number of solar-related myths, maybe mention a few of them.  I'd also mention in the intro that you're not just talking about the physics, but also how it relates to the ultimate efficiency and cost of the solar cells, because some people might be turned off from reading further if they think the post is just about physics.  And at the end I would add more detail about why you chose SunPower and FirstSolar specifically as examples, otherwise it sort of comes out of the blue.  Is SunPower a representative company, for example?

2010-12-05 08:37:08
Rob Painting
I agree, a series of rebuttals on solar myths would be good. A lot of bogus comments have cropped up on Dana's renewable energy post.
2010-12-05 18:53:51Solar myths
John Cook


Would be great if the main solar myths could be collected in one place - perhaps use Dana's baseload discussion thread as a good place to harvest them.

Even better, to find webpages featuring examples of these myths. Every rebuttal needs to have a link to a skeptic webpage giving the argument, ideally.

2010-12-09 16:24:13Another solar myth to counter
John Cook


Someone emailed me this solar myth they just encountered:

...there was one person who was openly in denial. I was unable to counter one argument about solar cells, though. He claimed that they'd never produce the energy it took to produce them during their whole life time (of course they do within 4-6 months...).

2010-12-11 03:39:28
Mark Richardson

I'm going to have a go at solar myths, but right now I'm very busy with my PhD meetings and a pantomime I'm being in next Thursday...


I was going to do 'solar cells/renewable energy don't earn back their energy' and 'solar cells cause global warming because they're dark'. Unfortunately, the second one has few/no literature calculations as far as I can tell, but I can do them pretty easily myself.