The takeaway from this article is that they came up for a technique to use cheaper materials instead of silicon. This is significant because silicon is very expensive and every increasing demand is also increasing the prices of it at a very fast rate. I took a class on CMOS Digital Design where the professor went over manufacturing overhead for a processor. The cost of the silicon alone was astounding, not even taking into account all the other overhead that comes with the process.
Unfortunately, like a lot of other research in this field, its real world applicability may be relatively limited. One of the reasons for this is that so much time, money, and infrastructure has been put into modern silicon semiconductor manufacturing that no one really wants to touch anything else. It could mean starting from scratch and requiring massive amounts of R&D and process planning to break even.
The real breakthroughs come when someone keeps the existing silicon process in mind and makes discoveries that use the existing infrastructure. That's the kind of research that really "changes" things.
If someone could come up with a manufacturing system that was cheap and easy to swap out or modify the process, they could literally change technology as we know it. If you could have the capability to scale processes easily, a lot of the really cool and cutting-edge research could get implemented on a large scale.
EDIT: I graduated as an electrical engineer and have taken several clean room processing classes, in case you were wondering.
The price of polysilicon, the stuff that a large portion of commercially available solar panels use, is NOT increasing. In fact, just the opposite has happened. If you want to know why Solyndra really failed so dramatically, you need to understand what this chart means to the solar industry:
Prices of polysilicon are forecasted to drop over the long term. That means any technology that wants to compete in solar cells had better be really, really cheap. First Solar can produce modules at around $.70/Watt. Many Si module manufacturers are producing at sub $.90/Watt. Given that Si modules that have been in place for decades and still function, any other materials are a risk for a 25+ year investment.
(I also graduated as an EE, but I also work in PV test and measurement, so I work with this stuff everyday, and my paycheck depends on it.)
From what I've learned, silicon is abundant in the earth crust. The problem is that you need a very very high purity to make useful wafers. So technically, its not a problem of scarcity, just a lack of miners producing it in a very pure form, which could be solved by having more of them.
Are there some other difficulties that I'm overlooking?
It's expensive to manufacture silicon at semiconductor quality, even when talking about polycrystalline wafers. (Monocrystalline wafers, used for "chips", is even more expensive and prone to defect." People are increasing production of polycrystalline silicon wafers but the demand is much greater than the supply, even with this increase. The cycle basically goes like this: polycrystalline gets more expensive because of demand -> supply increases -> prices go down (in theory) -> demand increases -> polycrystalline gets expensive again.
This is an endless loop right now, as the total amount of possible demand is >> than the growth of polycrystalline foundries. Furthermore, LCD displays also use this form of polycrystalline silicon, which doesn't help with the demand problem. Decently-graded silicon is inherently expensive to manufacture because of the process involved.
The takeaway is that if these metal-oxides are cheaper to produce, even if they are more expensive for the raw material, the cost savings would carry over to the products. Equally relevant when talking about solar cells is how much energy is needed to produce the cells themselves. Right now, an enormous amount of energy is required for silicon solar cells. Helping the energy crisis doesn't help if something takes that much energy to produce. (I do not know the ratio of lifetime energy output versus energy to manufacture but I am sure it's not very good.)
Again, you have this 180 degrees backwards. There was a huge rush to build out polysilicon module fabs in China that has caused a huge oversupply. Now module manufacturers are going out of business daily, leaving an oversupply of polysilicon capacity.
Unfortunately, like a lot of other research in this field, its real world applicability may be relatively limited. One of the reasons for this is that so much time, money, and infrastructure has been put into modern silicon semiconductor manufacturing that no one really wants to touch anything else. It could mean starting from scratch and requiring massive amounts of R&D and process planning to break even.
The real breakthroughs come when someone keeps the existing silicon process in mind and makes discoveries that use the existing infrastructure. That's the kind of research that really "changes" things.
If someone could come up with a manufacturing system that was cheap and easy to swap out or modify the process, they could literally change technology as we know it. If you could have the capability to scale processes easily, a lot of the really cool and cutting-edge research could get implemented on a large scale.
EDIT: I graduated as an electrical engineer and have taken several clean room processing classes, in case you were wondering.