And it's day-glow orange. Srsly.

July 14, 2008 6:56 AM

All of the light, 40 times the power. Improve the performance of run-of-the-mill solar cells by standing orange pieces of glass on them. MIT's Mark Baldo describes the technology.

The performance claims are outstanding, manufacture is cheap (and defect-tolerant), and the time to market is predicted to be three years. It can also be used to turn building window panes into solar energy collectors. The paper's published in the latest Science Magazine -- but you can only read the abstract unless you're a subscriber.

posted by ardgedee (41 comments total)

Solar concentrators in use today 'track the sun to generate high optical intensities, often by using large mobile mirrors that are expensive to deploy and maintain,'...

I've never understood this objection. Surely tracking the sun is no harder than locating, extracting, processing and delivering oil.

The dyes work together to absorb light across a range of wavelengths, which is then re-emitted at a different wavelength and transported across the pane to waiting solar cells at the edges.

So they are turning blue photons into red ones. Sounds good, but with current PV around 15% efficient, I'm having trouble figuring out how this can be an improvement of 40x.

On viewing the video, I see that they are being a little disingenuous. First they hate on concentrators, but then they've basically made a new concentrator (admittedly, one that doesn't need to track, but still has space and maintenance "problems"). AND they compare the efficiency versus no concentrator at all. What's the actual improvement value compared to existing, comparable technology?
posted by DU at 7:10 AM on July 14

(I am able to read the subscriber link on a word by word basis, but the sentences form no coherent answer to my question that I am able to identify.)
posted by DU at 7:19 AM on July 14

We report single- and tandem-waveguide organic solar concentrators with quantum efficiencies exceeding 50% and projected power conversion efficiencies as high as 6.8%. The exploitation of near-field energy transfer, solid-state solvation, and phosphorescence enables 10-fold increases in the power obtained from photovoltaic cells, without the need for solar tracking.

The Wikipedia Solar cell article discusses the distinction between quantum efficiency and power conversion efficiency.
posted by jedicus at 7:20 AM on July 14

I'm confused too, it seems like a good idea, but in the video, they say:

'double the power of existing solar cells!'

in the article it says:

'We were able to substantially reduce light transport losses, resulting in a tenfold increase in the amount of power converted by the solar cells.'

and again in the article it says:

'... the focused light increases the electrical power obtained from each solar cell 'by a factor of over 40'

Which is it? 2, 10, or 40?
posted by Mach5 at 7:23 AM on July 14

So yeah, cool science, good idea, promising line of research, keep working. But 1/2 < 40, let's don't hype too much.
posted by DU at 7:24 AM on July 14

DU writes "I've never understood this objection. Surely tracking the sun is no harder than locating, extracting, processing and delivering oil."

Because for micro installations (every house in the neighbourhood kind of thing) a tracker is a hassle and requires on going maintence. This tech, in theory anyways, is fire and forget. It just keeps working and could be built into windows.
posted by Mitheral at 7:49 AM on July 14 [1 favorite]

Yeah, but it's ORANGE. Yuck.
posted by tadellin at 8:00 AM on July 14

If you have a tiny solar cell in the sun, it won't produce much energy because a small amount of light is striking its surface. If you have a wave guide like this, you increase its effective area enormously because now its 'surface' is more like the surface of the glass. If you've increased its effective area by 50x, you increase its power output by 40x. The amount of the increase depends on how big you build the waveguide. The new waveguide is 10 times more efficient than older versions of the same thing, which is where that number comes from.

The double comes from someplace else. There are two options in solar cells: cheap large and inefficient, or small expensive and efficient. If you were to compare using a waveguide + efficient cell vs cheap large film on the same area, you'd get 2x the power. That's my guess on it anyway.
posted by a robot made out of meat at 8:04 AM on July 14 [1 favorite]

It's probably not better in pure energy-collecting performance compared to a motorized collection array. On the other hand, it doesn't require any of the energy it collects to track the sun, so that's one form of savings. It should also cost much less to maintain since there are no moving parts.

Above all a high defect tolerance means it's going to be very cheap to make.

The goal is to make stationary collectors useful -- currently they aren't. In most parts of the U.S. they provide their owners little or no economic advantage over staying fully on the grid, and some only make back their costs through tax breaks. So buy-in will be higher than current solar panels -- this is an add-on to current collectors, after all -- but with the potential for considerably greater return than current panels.

The advantage is mass efficiency and economy, not records-setting.
posted by ardgedee at 8:15 AM on July 14 [1 favorite]

D'oh, I fell into the efficiency trap again! When the fuel is free, efficiency (inside of some very broad limits) is irrelevant. The question isn't watts/m², it's watts/$ (and watts/ton_of_CO2). Reducing the area of expensive PV cells by simply increasing the area of cheap colored plastic is a massive gain.

Good job, MIT! And don't keep spreading the efficiency misnomer!
posted by DU at 8:24 AM on July 14

I see nothing in either the video or the abstract which suggests that these can be retrofitted onto existing flat-panel arrays. The whole idea is to use smaller edge collectors rather than larger sheets.
posted by KirkJobSluder at 8:25 AM on July 14

Why did you expect to see something about retrofitting?
posted by DU at 8:34 AM on July 14

For the most important metric (which DU correctly identifies as $/W), the organic concentrators are quite good, though from a pure efficiency point of view, they are quite unimpressive. DU is incorrect; efficiency is important for 2 reasons: (1) there is a limited amount of the elements which make up many solar cells, so low efficiency means $/W goes up quickly as supply of, say, indium goes down, (2) low efficiency can result in thermal effects (beyond the scope of this MeFi comment) and (3) high efficiency can, under the right circumstances, be a big driver for $/W.

The value of expensive tracking concentrators depends strongly on how much concentration is possible, as this determines the cost of the solar cell. Let's say I can make a very small 50% efficient cell with 1000x concentration. That means that my materials costs can be quite small (since I can use a very small cell to collect the sunlight from a very large area). The concentration costs are generally not too bad. The energy/environmental costs of this system can be quite manageable.

Of course, the difficultly is making a cell with such high efficiencies and capable of standing up to such large levels of concentration. There are several approaches with varying degrees of viability to achieve these ends. (I think the one I've recently started developing will work quite well, but there are surely hundreds of scientists who think the same thing about their own ideas, so only time (and funding!) will tell if I am correct.)

The key message I hope people will take away from this is that there is no one right answer for solar. Surely, something like the organic concentrators (which are very clever and reasonably easy to commercialize) has value, but that doesn't negate the value of high efficiency solar. Sadly, the funding for high efficiency solar isn't there as much as one might assume, as the route to market is a bit longer.
posted by JMOZ at 8:38 AM on July 14 [2 favorites]

Oh, and KirkJobSluder is correct: this MIT technology cannot be retrofitted. Given the tiny market cap of solar to date, this is, at worst, a minor weakness.
posted by JMOZ at 8:39 AM on July 14

And, using smaller PV surfaces can help with another resource crunch, the rising cost of semiconducting metals. Although some analysis has suggested that indium isn't really physically scarce. It's currently a byproduct of nickel production and refineries have little incentive to push more ore through the system.

DU: I was responding to argedee's, "this is an add-on to current collectors," and "by standing orange pieces of glass on them."
posted by KirkJobSluder at 8:44 AM on July 14

Sorry for the confusion -- I misunderstood the ability for this to be retrofitted.
posted by ardgedee at 8:53 AM on July 14

Dyson Sphere. Come on people, the solution is waiting for us!
posted by blue_beetle at 8:55 AM on July 14

tadellin : Yeah, but it's ORANGE. Yuck.

I'm sure that's just an effort to live up to the expectations set by the science fiction films of the 60s and 70s. Things that were orange were high tech.

This will be a great technology to power our talking mainframes and lava-lamps.
posted by quin at 8:58 AM on July 14

Surely tracking the sun is no harder than locating, extracting, processing and delivering oil.

Tracking mounts become expensive as the mass you need to track increases. Your choices are to massively overbuild them, like dedicated satellite uplink stations or deal with limited lifespan and high maintenance costs, like portable sat uplinks stations do.

Massively overbuilt means heavy, this makes roof mounting difficult -- esp. if you have to retrofit the roof to support the extra mass. This is why many TV stations, by the way, don't have their main uplink antennas on the roof -- the roof can't support them, and the extra structure costs too much.

Single axis tracking is easier, but still -- moving parts break. Bearings fail. If you can get a useful amount of power without tracking, it's a big, big win.

Win #2. Lets say (and these are made up numbers) it costs $100 to a make a square meter of solar cells. If this new concentrator costs $10/m², and uses one tenth the cells that a full power cells (so, .1m², cost, $10) and results in the same power output, you've just cut the cost of solar cells by a factor of *five*.

Big, Big win. The implication is that they're getting a factor of 40 increase in power density *at the solar cell*. What we don't know is how much will these collector/amplifier cost to make in large quantities -- but since it looks like, well, dyed glass, I suspect they won't cost nearly as much as the solar cells proper.

We're missing lots of real-world data here, but this looks very promising.
posted by eriko at 9:01 AM on July 14 [1 favorite]

I have a question about the claim that these could be built into windows. Given that the entire point of them is to capture and redirect light to the edges, where it's used by solar cells for power, it seems like the resulting windows would either be very inefficient solar collectors (i.e. they let a lot of light through rather than capturing it, which is, after all the point of windows, which can be conceived of as the exact opposite of solar collectors) or they'd be very dark.

Not to criticize the basic premise here -- this looks like a huge win for anyone who would like a cheap home solar PV array, which, I can say from very recent research, is still a complete waste of money. A home rooftop array that consisted of enough solar cells to line the edges of a bunch of sheets of dyed glass, rather than enough cells to cover the whole damn roof, seems like it would have to be many times cheaper, and thus possibly a workable solution.

But the solar-collector windows idea strikes me as dumb.
posted by rusty at 9:09 AM on July 14

Single axis tracking is easier, but still -- moving parts break. Bearings fail. If you can get a useful amount of power without tracking, it's a big, big win.

You don't need to tell me. I've spent a lot of the last 3 months building and rebuilding a homemade solar concentrating system. Even if parts didn't fail, there's still the additional material cost, space requirements, etc.

DU is incorrect; efficiency is important for 2 reasons: (1) there is a limited amount of the elements which make up many solar cells, so low efficiency means $/W goes up quickly as supply of, say, indium goes down, (2) low efficiency can result in thermal effects (beyond the scope of this MeFi comment) and (3) high efficiency can, under the right circumstances, be a big driver for $/W.

These are all true, but kind of beside the point I was making.

With a gas engine, you care about the efficiency of a certain mass of materials making the best possible use of a certain volume of fuel. With a solar engine, you still care about the mass but no longer care about the fuel. If you lump both of those things together under "efficiency" it gets confusing.
posted by DU at 9:13 AM on July 14

> the solar-collector windows idea strikes me as dumb.

A building's wall covered with windows generating a little electricity is better than a wall that generates none, but also a quality-of-life improvement for its occupants when compared to a building faced entirely with solar panels.

This is only practical if the cost difference between a generating and non-generating window is sufficiently compensated with the power generated. I wonder about the lifespan of the solar collectors too -- if replacing old or broken cells requires disassembling the whole window, that's not so good.
posted by ardgedee at 9:29 AM on July 14

it seems like the resulting windows would either be very inefficient solar collectors (i.e. they let a lot of light through rather than capturing it, which is, after all the point of windows, which can be conceived of as the exact opposite of solar collectors) or they'd be very dark.

Well, it's not a binary problem like you frame it here. In theory, you could have both a slightly darkened window (which will still be useful as a window in an office building, not necessarily a house) and a slightly less efficient solar collector (which will still be useful when scaled up to skyscraper size).

Now, consider that skyscraper will stand there, delivering energy to the grid, for 50+ years. And all the while, you're making improvements to the technology, so you can easily swap out window panes...
posted by Cool Papa Bell at 9:29 AM on July 14

But the solar-collector windows idea strikes me as dumb.

It would be a matter of balancing transparency with energy-gathering. Allow enough light through for interior lighting and viewing, capture the rest as electricity. Right now, in a lot of skyscrapers, a lot of money is spent (a) making the windows reflective to reduct heat gain, (b) putting shades and drapes inside the glass to keep the light out. Much of that reflected light off mirrored buildings intensifies heat gain in neighboring buildings without reflective coatings, and at street level. If, instead, you can absorb it into solar cells, it's a win-win. Another approach would be to maximize energy collection on a portion of the surface, say two-thirds, and allow clear glass on the remaining third. Plenty of window area for light and views, but lots of energy captured that would otherwise be reflected.
posted by beagle at 9:33 AM on July 14

I just can't believe it took this long for someone to go from 'Hey, this Space Lego piece is bright along the edge' to building this.
posted by sixswitch at 9:44 AM on July 14

Quick question while we're thinking about this. Can anyone provide me a useful set of links about the material needs and costs for various kinds of solar cells? I was talking to a solar energy entrepreneur type who was curious about where we got tellurium. What are the other rare / uncommon metals and materials? Any pointers would be appreciated -- I can and will do the leg work myself, but getting a push in the right direction would be very nice.
posted by bumpkin at 9:58 AM on July 14

That's all well and good, until the human race scorches the sky.
posted by bwg at 10:02 AM on July 14

That's all well and good, until the human race scorches the sky.

It'll be OK. We have Keanu Reeves and Laurence Fishburne on our side.
posted by Cool Papa Bell at 10:05 AM on July 14

Well, actually lots of buildings have 'darkened' windows already, usually to try to keep the buildings cool. I don't know about orange though, it would be nice if they could figure out some other colors.

Also, even if you don't want a window, these things could still be useful. You put silicon around the edges of this glass, then you just glue the glass to a board or something. You could even put lower-efficiency organic solar panels behind the glass to boost efficiency just a little bit more.

Anyway, I like this advance because I own stock in a bunch of silicon based solar companies, and this would just make their products more efficient, rather then replacing them.
posted by delmoi at 10:33 AM on July 14

What are the other rare / uncommon metals and materials?

Most solar companies just use just use polysilicon. I'm not sure if they dope them with anything rare but silicon is very abundant on earth, obviously. First Solar uses tellurium, which is a trace element and is only produced right now as a byproduct of copper mining.
posted by delmoi at 10:39 AM on July 14

bumpkin: search for prevalence of various elements. Tellurium is/was fairly common, but demand has expanded, and so the (previously VERY low) price has been increasing of late.

Another serious consideration are environmental factors. Lead and cadmium, for example, are things to be avoided whenever possible.

CIGS (copper indium gallium selenide) has been hailed as the magic bullet for cheap solar, but copper and indium are both expensive/becoming expensive and gallium isn't cheap. Silicon isn't so bad (there's PLENTY of it), but the contacts can require indium or other expensive metals. Going to small amounts of materials producing large amounts of power is extremely helpful.

DU: Yes, I understand your point regarding efficiency, but in practice, efficiency can be measured in W/kg, which is largely equivalent of W/m^2. There's an unlimited supply of energy, but a limited supply of the materials for solar cells. I understand there's (virtually) unlimited fuel, but the difference is that here, we care about the cost of the engine rather than the cost of the fuel.

Regarding windows: The efficiency of the device determines (well, approximately) how transparent they are. If this were a highly efficient solar cell, it would make a poor window. Given that the solar cells themselves would be in the window frame (that is, at the edge), the waveguide would make a decent (if orange) window. It would, of course, reduce the amount of light going through the window, though that's usually desirable for thermal management of the indoor environment. (Too much direct sunlight = hot office or lots of A/C).
posted by JMOZ at 10:40 AM on July 14

Solar paint. Give it six years.
posted by autodidact at 10:51 AM on July 14

Solar paint.
posted by autodidact at 10:53 AM on July 14

This technology and sun tracking systems are not mutually exclusive. It would seem that with a large panel of this material as a target a fairly crude (and robust) reflector system could be quite effective.
posted by speug at 11:11 AM on July 14

Everyone (re windows): Yeah, that all makes sense. Especially if orange is not the only color option. I was thinking of these things in terms of like 75-80% light collection or thereabouts. Obviously (in retrospect) that's not a necessary feature. Duh.
posted by rusty at 11:26 AM on July 14

autodidact....you realize you have linked to a description of the same research, right? It's all the same researchers as in the science paper.
posted by HighTechUnderpants at 12:04 PM on July 14

Do I really have to be the first to welcome our new 70s-color-palette overlords? Apparently so. *humph*
posted by Halloween Jack at 1:16 PM on July 14

I was thinking of these things in terms of like 75-80% light collection or thereabouts.

Twenty percent transmittance would make a fine office window. Even 10% might work well. Remember, human eyes don't have linear sensitivity. That's why a room lit with a 100 watt bulb appears only somewhat brighter than a room lit with a 60 watt bulb, rather than twice as bright.
posted by ryanrs at 2:53 PM on July 14

I just can't believe it took this long for someone to go from 'Hey, this Space Lego piece is bright along the edge' to building this.

High-efficiency solar cell research is driven by space applications. Boosting a satellite into orbit is hard, so W/kg is king, not W/$.
posted by ryanrs at 3:02 PM on July 14

I think what's interesting about this paper is that they aren't developing new molecules or even finding new effects - it's stuff that's been around but just wasn't necessarily applied to solar energy (very well). Just reminds me that we (as a society) are barely crawling into the renewable energy age. The low hanging fruit are still very, very low hanging. My only concern would be stability of those organic molecules used in sunlight, but I'm sure they've thought of that already and UV can be filtered.
posted by peppito at 3:38 PM on July 14

Product specs for several types of energy-saving window films. Visible light transmission ranges from 51% to 8%.
posted by ryanrs at 7:29 PM on July 14