Jul 20 2017

Bismuth and Solar Cells

bismuth ocyodideThere is something appealing about the fact that while there was so much controversy and public debate about global warming and energy production, solar cell technology incrementally improved in the background, largely unnoticed, until it became an actual cost-effective option for energy production. There were no breakthroughs or big announcements, just a slow increase in efficiency and decrease in cost.

Slow incremental changes add up, like a conservative but reliable investment. In 1941 the first silicon-based solar cells were developed with <1% energy conversion efficiency. In 2009 we broke 25% efficiency. The current record efficiency for mass-produced solar cells is 26.6%.

While more efficiency is always welcome, this is now more than enough for most practical uses of solar cells. The solar panels on the roof of my house (which doesn’t even cover my entire available roof space) produces 100% of the energy (averaged over the year) that my house consumes. Obviously, your mileage may vary, depending on roof orientation, shading, and geographic location. I live in CT, which is hardly a sunny state, so anywhere in the US should be viable for solar energy. If you life somewhere like Arizona, it’s a no-brainer.

With many technologies there is one feature that is considered a limiting factor, or at least the most important factor, at least by the public, and tends to dominate discussion. For computers it used to be processor speed, for cameras it was megapixels. Once, however, we get to the point where these parameters are generous enough to no longer be the dominant limiting factor, we start to pay attention to other factors. Make no mistake, these other factors were important all along, they were just often neglected by the public who focused on one number.

I have been reading about solar panels for decades, and definitely popular discussions were dominated by reference to the energy conversion efficiency. Now that we are north of 20%, however, it is clear that other factors and now more critical. The parameter that I think now dominates discussion is the overall cost of mass produced solar panels. While the price has also been dropping every year, the cost is still high – tens of thousands of dollars for a single home, depending on how much of your roof needs to be covered.

The primary factor that is keeping costs relatively high, apparently, is the fact that silicon-based solar cells have a very low tolerance for defects. This means the silicon purity needs to be very high, which in turn drives high manufacturing costs.

Solar researchers have therefore been looking for alternatives to silicon that have a much higher tolerance for defects. A great deal of recent attention has been focused on perovskite. If you have not heard about perovskite yet, you probably will (or, as we will see, perhaps not). A recent report noted:

In the past decade, research into perovskite solar cells has boomed. At least in the lab, the efficiency of perovskite devices is now more than a match for those based on silicon. In 2006, the first perovskite photovoltaic converted 2.2% of photons into electrons1; by 2016, that figure was 22.1%. Silicon rooftop panels have an efficiency of 16–20%; perovskite cells could in theory could reach 31%. And even higher efficiencies might be achieved by combining silicon and perovskite devices.

So again we are well into the acceptable efficiency range, but the huge advantage of perovskite is that they have a higher defect tolerance than silicon and are therefore much cheaper to mass produce. This technology could break out of the small incremental advances and put us on a different path to cheap and abundant solar panels.

But there is one issue – current perovskite solar cells are alloyed with lead. It is not clear that this is a problem, but it is a concern. Lead is a highly toxic heavy metal, and its use in mass producing anything is a concern. It would also be a concern for disposal. I don’t think this is a deal-killer, but it is not optimal.

As with other similar technologies (like batteries) the optimal solar cell has multiple simultaneous properties: high energy conversion efficiency, long lifespan, resistant to damage, cheap to produce, easy to install, and the manufacture, use and disposal (or recycling) has low environmental impact (meaning not a lot of toxic substances).

Perovskite is efficient and cheap, but contains a toxic heavy metal. Researchers, therefore, are continuing to search for still other solar panel materials that would be even better than perovskite.

A recent study explores the potential of bismuth oxyiodide. Bismuth is similar to lead (in that it is a dense metal), but is extremely non-toxic. The researchers found that bismuth oxyiodide is tolerant to defects, therefore inexpensive to manufacture, has potential efficiencies rivaling perovskite, is stable, and non-toxic. Further, it can be mass produced with existing techniques.

It is too early to tell what the next generation solar panels will be. There is also a class of solar panels that are thin, flexible and cheap. They have less efficiency than current silicon, however, and make up a minority of the market.

What is encouraging is that the basic science of solar voltaics has advanced to the point that applications are now coming fairly quickly. We can’t predict what will dominate the market in a decade, but there are now several viable options all of which promise to be significant improvements over current silicon-based solar panels (which are already cost effective in many contexts).

This can only improve our energy infrastructure, with the potential of greatly expanding our green energy production. However, the grid will need to be significantly upgraded to handle distributed production of energy. Right now the grid can only handle about a quarter of houses having solar panels. As more homes try to add solar, they (or the electric company) will have to pay to update the grid in their area to handle the production.

Battery technology (or some mass storage technology) also needs to improve. If every home could also have a power wall type storage that was cheaper with a greater capacity and life cycles, that could have a huge impact.

Further, while solar energy will help reduce our greenhouse gas production, we cannot assume it will be enough fast enough. But I am excited to see this technology really come into its own.

17 responses so far

17 thoughts on “Bismuth and Solar Cells”

  1. SteveA says:

    “Bismuth is similar to lead, but is extremely non-toxic”

    Slightly eccentric phrasing.

    Surgeon: “Your mother has gone to a better place.”
    Relative: “You mean…she’s dead?”
    Surgeon: “I mean, she’s extremely non-alive.”

    Had a quick read-up on Bismuth. China (typically) has most of it.

    My other half often suggests we go solar. It’s a while since I last looked into it. Perhaps time to revisit the idea.

  2. mumadadd says:

    “Slightly eccentric phrasing.”

    I had exactly the same reaction to that — quite non-elegently phrased.

  3. I added a clarification. It is a dense metal like lead, and apparently has similar properties for use in solar panels, but it is very non-toxic.

  4. jester700 says:

    I thought “non-toxic” meant “not at all toxic”, and as an absolute couldn’t be used with a modifier. Like “unique”. Is that incorrect? Are there degrees of non-toxicity?

  5. there is no such thing as 0% toxic. All toxicity is a matter of degrees, as it is a matter of dose. We are always talking about relative toxicity. This can be further clarified depending on context, such as no toxicity at the doses to which people will be exposed.

  6. jester700 says:

    I understand your points, and the meaning was clear. This is an issue of semantics; I didn’t understand the best way to say “of extremely low toxicity”, and understand the prefix “non-” to mean “not”, so it just looked wrong. But if that usage is indeed accepted and the best way to say it, cool.

    I didn’t mean to be a grammar wonk – especially not an incorrect one. It really just looked goofy to me, like “very unique”.

  7. Charon says:

    I feel like I’m off topic because I’m not posting about grammar 🙂

    Although distributed (rooftop) solar isn’t yet cost competitive, utility-scale solar is already likely cheaper than fossil fuel power plants or nuclear (e.g., https://medium.com/americas-power-plan/wind-and-solar-are-our-cheapest-electricity-sources-now-what-do-we-do-b323082239de ).

    There are challenges, including the grid one mentioned, and also load balancing. But even with tricky things like load balancing, there are good ways forward, like demand management.

  8. DisplayGeek says:

    In speaking of silicon vs. perovskite I believe a significant issue is NOT the cost of manufacture of the cell alone, but of the rest of the physical packaging that is needed. Silicon is very robust. Perovskite is NOT. It has a very short lifetime as moisture and UV degrade the material rather rapidly. The cost of packaging perovskite cells will be higher than the cost saved by replacing the silicon.

    The issue of lead will likely be a deal killer due to fear. Consider that lead in electronics, even though very stable and not likely to cause environmental issues, has been effectively banned world wide.

    Replacing the lead with bismuth will help this… but fails to address the issue of packaging cost.

    Interestingly, this issue of moisture and UV degradation is also present in Organic Light Emitting Diode (OLED) displays… of which I am VERY familiar in my work in the field. My experience in the field suggests that even with cell costs verging on zero, the cost of packaging still outweighs the savings of eliminating silicon.

    Oh… and the cost of silicon is NOT dominated by defect issues… but of the very high capital expense of the factory.

  9. mmcfall88 says:

    I just want to point out that while there is a mineral called perovskite (CaTiO3), it generally refers to a group of materials with the same crystal structure and a chemical composition of the form ABO3. There are many perovskites that are being studied for a variety of interesting electrical properties.

  10. Alex Simmons says:

    The limiting factor for rapid take up isn’t so much technological but rather a lack of suitable financial solutions. Deals like what Steve has, where up front costs and maintenance are not his concern, are not common. Nothing exists like that here in Australia. And of course local policies about feed in tariffs etc vary considerably.

    New tech will continue to reduce the cost incrementally but even half of $10k is still too much of an up front expense for many. Here payback time for solar is still in the 7+ year range. That’s OK but still represents a hurdle, especially if loan finance is the only way to afford the up front cost as that pushes out the payback time further.

    As an aside, while there are early adopters, batteries still don’t make financial sense for a household. Cost per kWh over life of battery is still several multiples of grid power cost. That tech needs to be 1/2 to 1/3 the current cost to begin making financial sense.

  11. MusingJess says:

    I find it so fascinating when all these different areas intersect. Journal articles, patents, affordability, cost of packaging. I had no idea these things could be this complicated. All that time I saw solar cell headlines, I just assumed, oh okay a new advance in solar cells, they will just add it to the existing stuff. But then I’m struck with feeling like oh yeah, everything is really complicated when you get into the details.

  12. Robert Christ says:

    Where I live in Cutter NM it’s not to the utility providers interest to buy back electricity. I am so rural if everyone had solar panels they would be supplying the infrastructure for free. The way it is now where almost nobody has solar they are still practically supplying the infrastructure for free. When I inquired they said they would not buy back power. Some other areas in NM they are required to. Not in Cutter. For some reason they don’t have to where I am. Although it would be a lot of fun I can’t see buying batteries and voltage regulators so that I can take full advantage of the power solar panels produce. That and I don’t live there year round. That and their is always the issue of theft and vandalism. So for now I am sitting on the sidelines.

    I also would, because I have the space, want trackers which could increase output by 10 to 25% but don’t want to spend a fortune on them. I would think bringing down the cost of trackers would certainly be a low tech way to increase efficiency. And would be very popular if you could mount them on roofs and get them to work that way.

    Also at this point I have to consider if I am going to live long enough to offset the initial cost.

  13. SteveA says:

    DisplayGeek: “Silicon is very robust. Perovskite is NOT. It has a very short lifetime as moisture and UV degrade the material rather rapidly.”

    Sounds like a deal-killer for a roof-top installation.

    DisplayGeek: “The issue of lead will likely be a deal killer due to fear. Consider that lead in electronics, even though very stable and not likely to cause environmental issues, has been effectively banned world wide.”

    It’s a shame. Lead only has the reputation it has because, historically, we’ve worked so hard to make it bioavailable (putting it in fuel, house-paint, food etc.). Used responsibly I would consider it a relatively low risk material.

    People are generally over sensitised to the, often, presumed dangers of chemicals, radioactivity, EMR. People who complain bitterly about the heath-risks of living too close to a phone mast, power-lines or factory are oblivious to the stockpile of highly toxic, potentially explosive, flammable fluid they have at the end of their street. When you tell them about it, and they ask what you’re talking about, and you say ‘Gas station’, then the penny drops and they sometimes start thinking about these risks in a different way.

    By any reasonable definition, gasoline (which the civilised call ‘petrol’) is a horribly dangerous substance, but because it is familiar and has obvious utility, most people don’t think twice about it. And why should they? On a day-to-day level it is stored and used responsibly and safely.

  14. Alex – it would be nice if you could finance solar panels like a car, especially if the monthly payments was at or less than your electric bill savings.

    Regarding batteries, I agree, we are not there yet. I suspect we will be perhaps in about a decade.

    However, in some states power companies are successfully pushing for peak billing laws. This is a way for them to recoup lost income from solar panels. Essentially they charge you a penalty for your peak usage, with the justification that they have to maintain extra capacity to meet it.

    In this case, a home battery could be used for peak shaving, and could therefore avoid these substantial costs. So they may ironically be making home batteries cost effective.

  15. ScubaSharky says:

    “it would be nice if you could finance solar panels like a car, especially if the monthly payments was at or less than your electric bill savings.”

    That’s exactly what I’ve done with my home here in Central Florida. I joined a solar co-op to get really low pricing and financed with a HELOC. My monthly loan payments are comparable to what my electric bills were before. There’s been no change to my monthly cash flow but I’m building equity with every payment. The system is churning out enough energy not only to power my home, but also to charge my Chevy Volt. Factoring in that $80-$100 I’m saving in gas, I’ve actually increased my cash flow by going solar and driving an EV. ☀️

  16. Ivan Grozny says:

    “Bismuth is very non-toxic” – sounds like “CNN is very fake news”. )

  17. Gingerbaker says:

    Small differences in cost between two competing PV panels are not crucial. What seems more important to me is longevity of the panel. Contemporary silicon PV looks to have at least a 125 year working lifespan, so the amortized cost is nominal.

    But longevity, so far, is the Achilles heel of perovskites. I believe they currently look to have a very small fraction of the lifespan of silicon. Perhaps, someday, they can be printed onto rolls of linoleum for next to nothing and lifespan won’t matter so much.

    But there is something very satisfying about measuring usable lifespans of silicon in units of centuries, don’t you think?

Leave a Reply