4 ms·

Breakthrough: IBM Makes a Solar Cell Out of "Earth Abundant" Materials

54 points by MikeCapone 17y ago

go37pi 17y ago
This solar cell has a "conversion efficiency of 9.6 percent, which is 40 percent higher than previous attempts to create a solar cell made of similar materials." Current solar cells have approximately 20 percent efficiency. Unless this solar cell is extremely cheap, I think this is a long way from practicality. http://upload.wikimedia.org/wikipedia/commons/a/a4/PVeff%28rev110707%29d.png http://upload.wikimedia.org/wikipedia/commons/a/a4/PVeff%28r... - graph of best research cell efficiencies EDIT: "The researchers estimate that if the technology can be taken beyond its basic state right now and achieve 12 percent efficiency, it could be an economically viable alternative to current products." - from CNET news http://news.cnet.com/8301-11128_3-10451641-54.html http://news.cnet.com/8301-11128_3-10451641-54.html
- machrider 17y ago
  Is efficiency really the #1 criterion for practicality? I thought cost (per watt) was.
  - go37pi 17y ago
    That's a good point, I was implicitly assuming that the lost power due to efficiency losses would be greater than the gained power due to cheaper costs.
  - anamax 17y ago
    #1 enough. My roof has finite surface area. If the efficiency is too low, even zero cost isn't competitive. That's true even of commercial facilities. Various enviro groups are already trying to kill solar farms. Somewhere in here, wire and other infrastructure (such as the frame) costs matter too.
  - Groxx 17y ago
    It's a sliding scale for priority. More space per watt takes up more space for sun-power farms, and they often cause more heat, making the eco people go nuts, what with migrating geese turning into popcorn chicken above them (</exaggeration>). For many individuals, who just have roof-space (for instance), space is really at a premium, and increased output means more money saved / made per time period, so it can pay for itself and more over time. Power companies could go either way, as they need a LOT of space. Doubling the area they need may not sound too good to them, as land costs money every year in taxes. If land is cheap enough though, cost per watt likely wins out.
- presidentender 17y ago
  The point isn't that it's close to being efficient as our current technology, it's that the materials are cheaper. Indium and gallium are a lot more expensive than tin and zinc.
- akgerber 17y ago
  I think there's a lot more surplus sunshine than surplus Ga or In.
  - camccann 17y ago
    Depends on where you are. Sunshine is a little harder to box up and ship to where it's needed.
    - MikeCapone 17y ago
      Indeed. That's why they put the best solar panels on space probes and satellites, and why they use those with the best performance/price ratio for solar farms in the desert.
- patrickgzill 17y ago
  IF they can make them cheap enough, you can basically afford to put them everywhere. Also from a manufacturing standpoint, it will be easier to get financing to build a factory, as you are not dependent on commodity swings in scarcer metals affecting your ability to make a profit.
- Zarathust 17y ago
  Conversion rates aren't as important as with fuel driven electricity generators since the sun has unlimited fuel. Making hydrogen by burning gas however is much more crucial, why aren't we fueling cars and power plants directly instead? Solar cells have different metrics, like cost/square meter
- MikeCapone 17y ago
  "Current solar cells have approximately 20 percent efficiency." Not the vast majority of commercial ones (closer to 10-15% range, afaik - the ones that get around 20% (for non-concentrating) are in the lab or on satellites), and certainly not the cheaper thin film (GICS) kind with which this will compete.
dkimball 17y ago
This is extremely heartening. I think we always over-estimate the level of threat that we face at a given time, and underestimate the possibility of finding some sort of a solution. But on the other hand, not needing enormous quantities of rare earths means that human expansion into space no longer gets a kickstart from attempts to mine the asteroid belt... :)
anonjon 17y ago
This makes me wonder what the conversion efficiency and price per watt for a tree is. Trees (plants in general) already do photosynthesis really well, so why can't we genetically engineer a tree that produces electricity?
- ars 17y ago
  Why do you need to genetically engineer it? Grow a tree, cut it down, burn it. Done. Or grow switchgrass and burn it instead of coal. Or burn corn. And ignore the stupid ethanol thing.
  - camccann 17y ago
    If you want to take that route, one of the most promising approaches is via "algaculture"--basically cultivating oily pond scum in shallow pools, then skimming off the algae and drying/pressing it to extract useful oil.
- graphene 17y ago
  Trees (plants in general) already do photosynthesis really well As I understand it, that's not exactly true; photosynthetic compounds are very efficient at harvesting light, but the chemical conversion to energy-carrying compunds is rather inefficient, I believe on the order of 1%. That said, those compounds are probably more valuable per joule than electricity since there is no storage problem, so it's hard to compare the two. The reason we can't simply genetically engineer an organism that converts sunlight into electricity (or fuels) is that at the moment, neither the photon harvesting process nor the chemical conversion process are well enough understood to try to replicate.
  - MikeCapone 17y ago
    http://en.wikipedia.org/wiki/Photosynthetic_efficiency http://en.wikipedia.org/wiki/Photosynthetic_efficiency
- Groxx 17y ago
  Put simply, trees make sugar, not electricity. We could burn the sugar... as coal...
CoryMathews 17y ago
IBM also said "But this is just a start. More improvements to power conversion should be possible." So now that the poc is done they can refine it and make it better. Also I agree with @machrider its all about the cost per watt. Thats what is going to make it popular or not.
InclinedPlane 17y ago
Approximately 1/4 of the Earth's crust is made out of Silicon. Tin and Zinc may be more abundant than Gallium, but this is a ridiculous phrasing given that Silicon based PV cells already exist.
MikeCapone 17y ago
"Even with selenium this type of cell has materials cost advantages over existing commercial thin films from First Solar made of cadmium and telluride. Also, this cell has advantages over the CIGS (copper indium gallium selenide) cells of the newer thin film manufacturers since indium and gallium cost more and CIGS also uses selenium." (source: http://www.futurepundit.com/archives/006938.html http://www.futurepundit.com/archives/006938.html)

54 points by MikeCapone 17y ago

This solar cell has a "conversion efficiency of 9.6 percent, which is 40 percent higher than previous attempts to create a solar cell made of similar materials." Current solar cells have approximately 20 percent efficiency. Unless this solar cell is extremely cheap, I think this is a long way from practicality. http://upload.wikimedia.org/wikipedia/commons/a/a4/PVeff%28rev110707%29d.png http://upload.wikimedia.org/wikipedia/commons/a/a4/PVeff%28r... - graph of best research cell efficiencies EDIT: "The researchers estimate that if the technology can be taken beyond its basic state right now and achieve 12 percent efficiency, it could be an economically viable alternative to current products." - from CNET news http://news.cnet.com/8301-11128_3-10451641-54.html http://news.cnet.com/8301-11128_3-10451641-54.html

machrider 17y ago
Is efficiency really the #1 criterion for practicality? I thought cost (per watt) was.
- go37pi 17y ago
  That's a good point, I was implicitly assuming that the lost power due to efficiency losses would be greater than the gained power due to cheaper costs.
- anamax 17y ago
  #1 enough. My roof has finite surface area. If the efficiency is too low, even zero cost isn't competitive. That's true even of commercial facilities. Various enviro groups are already trying to kill solar farms. Somewhere in here, wire and other infrastructure (such as the frame) costs matter too.
- Groxx 17y ago
  It's a sliding scale for priority. More space per watt takes up more space for sun-power farms, and they often cause more heat, making the eco people go nuts, what with migrating geese turning into popcorn chicken above them (</exaggeration>). For many individuals, who just have roof-space (for instance), space is really at a premium, and increased output means more money saved / made per time period, so it can pay for itself and more over time. Power companies could go either way, as they need a LOT of space. Doubling the area they need may not sound too good to them, as land costs money every year in taxes. If land is cheap enough though, cost per watt likely wins out.
presidentender 17y ago
The point isn't that it's close to being efficient as our current technology, it's that the materials are cheaper. Indium and gallium are a lot more expensive than tin and zinc.
akgerber 17y ago
I think there's a lot more surplus sunshine than surplus Ga or In.
- camccann 17y ago
  Depends on where you are. Sunshine is a little harder to box up and ship to where it's needed.
  - MikeCapone 17y ago
    Indeed. That's why they put the best solar panels on space probes and satellites, and why they use those with the best performance/price ratio for solar farms in the desert.
patrickgzill 17y ago
IF they can make them cheap enough, you can basically afford to put them everywhere. Also from a manufacturing standpoint, it will be easier to get financing to build a factory, as you are not dependent on commodity swings in scarcer metals affecting your ability to make a profit.
Zarathust 17y ago
Conversion rates aren't as important as with fuel driven electricity generators since the sun has unlimited fuel. Making hydrogen by burning gas however is much more crucial, why aren't we fueling cars and power plants directly instead? Solar cells have different metrics, like cost/square meter
MikeCapone 17y ago
"Current solar cells have approximately 20 percent efficiency." Not the vast majority of commercial ones (closer to 10-15% range, afaik - the ones that get around 20% (for non-concentrating) are in the lab or on satellites), and certainly not the cheaper thin film (GICS) kind with which this will compete.

dkimball 17y ago

This is extremely heartening. I think we always over-estimate the level of threat that we face at a given time, and underestimate the possibility of finding some sort of a solution. But on the other hand, not needing enormous quantities of rare earths means that human expansion into space no longer gets a kickstart from attempts to mine the asteroid belt... :)

anonjon 17y ago

This makes me wonder what the conversion efficiency and price per watt for a tree is. Trees (plants in general) already do photosynthesis really well, so why can't we genetically engineer a tree that produces electricity?

ars 17y ago
Why do you need to genetically engineer it? Grow a tree, cut it down, burn it. Done. Or grow switchgrass and burn it instead of coal. Or burn corn. And ignore the stupid ethanol thing.
- camccann 17y ago
  If you want to take that route, one of the most promising approaches is via "algaculture"--basically cultivating oily pond scum in shallow pools, then skimming off the algae and drying/pressing it to extract useful oil.
graphene 17y ago
Trees (plants in general) already do photosynthesis really well As I understand it, that's not exactly true; photosynthetic compounds are very efficient at harvesting light, but the chemical conversion to energy-carrying compunds is rather inefficient, I believe on the order of 1%. That said, those compounds are probably more valuable per joule than electricity since there is no storage problem, so it's hard to compare the two. The reason we can't simply genetically engineer an organism that converts sunlight into electricity (or fuels) is that at the moment, neither the photon harvesting process nor the chemical conversion process are well enough understood to try to replicate.
- MikeCapone 17y ago
  http://en.wikipedia.org/wiki/Photosynthetic_efficiency http://en.wikipedia.org/wiki/Photosynthetic_efficiency
Groxx 17y ago
Put simply, trees make sugar, not electricity. We could burn the sugar... as coal...

CoryMathews 17y ago

IBM also said "But this is just a start. More improvements to power conversion should be possible." So now that the poc is done they can refine it and make it better. Also I agree with @machrider its all about the cost per watt. Thats what is going to make it popular or not.

InclinedPlane 17y ago

Approximately 1/4 of the Earth's crust is made out of Silicon. Tin and Zinc may be more abundant than Gallium, but this is a ridiculous phrasing given that Silicon based PV cells already exist.

MikeCapone 17y ago

"Even with selenium this type of cell has materials cost advantages over existing commercial thin films from First Solar made of cadmium and telluride. Also, this cell has advantages over the CIGS (copper indium gallium selenide) cells of the newer thin film manufacturers since indium and gallium cost more and CIGS also uses selenium." (source: http://www.futurepundit.com/archives/006938.html http://www.futurepundit.com/archives/006938.html)