New silicon etching drops solar power costs by more than 10 percent

Researchers have found a way to reduce production costs of solar cells by more than 10%.

Switching the Silicon Used in Solar Cells for Renewable Energy Drops Costs

Silicon is the standard light-capturing material used in solar photovoltaic (PV) cells. It comes in two main forms: perfect crystals that cost more and produce higher efficiencies and multicrystalline silicon that cost less, but offers lower efficiencies. With common etching to reduce reflected light both types still lose some light, which is what gives most solar panels their blue color.

Researchers already knew that nano-texturing silicon with dry etching makes black silicon (black-Si) that is more efficient at capturing light than standard etching treatments. It has no color because the dry etching process takes a normally flat silicon surface and “etches it into a forest of nanoscale needles,” Pearce says. “Those needles grab the light and don’t let it get away. It’s like looking into the eyes of Darth Vader.”

Normally such a high surface area with many surface defects would hurt electrical performance, but researchers at Aalto University found that when the silicon is also treated with an appropriate atomic layer deposition (ALD) coating, the effects of surface defects are mitigated.

Typical thinking has been that the cost of black-Si cells from dry etching and ALD are too expensive for practical use, especially in an industry where, Pearce says, “margins are extremely tight. Everyone’s trying to push costs as low as possible.”

However, the results of their study shocked even Pearce. While researchers did find that production of individual black-Si passive emitter rear cells (PERC) were between 15.8 and 25.1 percent more expensive than making conventional cells, they also found that the efficiency gains and the ability to go to the less-expensive multicrystalline silicon starting material far outweighed those extra costs: overall the cost per unit power dropped by 10.8 percent.

The Future of Renewables and Solar Energy Production through Material Science

Black is not only better than blue when it comes to solar panels. The improvements could start to beat out renewables’ top energy competitor in the climate change arena.

“For the people that think coal technology is going to be able to compete with solar, they should know solar costs are still coming down. Most coal companies are already, or near, bankrupt now,” Pearce says. “There’s no way coal’s going to be able to compete with solar in the future.”

He adds, “This study points to where the future is going to go in PV manufacturing and what countries might want to do to give themselves a competitive advantage.”

Teaming Up Across the Atlantic for Solar Energy Efficiency

Pearce completed this study while on sabbatical as a Fulbright distinguished chair at Aalto University in Finland. He worked with the Hele Savin’s Electron Physics Group and had access to their data on these processes. Researchers were also able to get information on manufacturing costs from companies, which is not public, but were allowed to use for this study, along with published literature on solar cells.

While the spot price for solar cells may change day by day – or even by hour – the results still hold. “That’s 10 percent decline between cell types from whatever the number is that day,” he says. This is because the comparisons were made on relative costs, not absolute costs. That’s also why arbitrarily fluctuating tariffs were not factored into the calculations.

What’s Next for Solar Energy and Renewables

Pearce says that while the production process can still be optimized to pull out a few more percentage points of efficiency, the next step for this study is to be used by policy makers to accelerate PV manufacturing. For a country like China, which already dominates global PV manufacturing, “to make this relatively small change is pretty trivial.” The European Union, which currently makes a lot of the manufacturing equipment, should also “look carefully at scaling up deep reactive ion etching and ALD tools to meet the needs of the rapidly expanding PV market”. He hopes that countries like the U.S., which used to dominate the solar field, will use this data at a policy level to leap frog international manufacturers, and invest in producing the new machines to manufacture these types of solar cells.

“I don’t know which technology will end up being the one to dominate the solar field,” he said, however “The study shows the clear economic impetus to move in the direction of dry-etched black silicon PERC that wasn’t there before.”

14 thoughts on “New silicon etching drops solar power costs by more than 10 percent”

  1. ‘For a country like China, doing this is trivial!’ Except for a coupla things… Heat buildup is a problem for PV, and I’m not sure your black silicon etched wafers would not be a self-incinerating black body with the wrong IR wavelengths to convert to electricity. And as a related article here indicated, “fixing” China’s dependence on imported foreign chips to offset fuel imports, no amount of energy independence will satisfy Xi, and no amount of export self reliance by China will now encourage the global mob to import Chinese anything — what would China import f4rom the world, besides T-Bills, which Beijing derides as being the reserve currency of the declining US?

    Reply
  2. ‘For a country like China doing this is trivial!’Except for a coupla things…Heat buildup is a problem for PV and I’m not sure your black silicon etched wafers would not be a self-incinerating black body with the wrong IR wavelengths to convert to electricity.And as a related article here indicated fixing”” China’s dependence on imported foreign chips to offset fuel imports”” no amount of energy independence will satisfy Xi and no amount of export self reliance by China will now encourage the global mob to import Chinese anything — what would China import f4rom the world besides T-Bills”” which Beijing derides as being the reserve currency of the declining US?”””

    Reply
  3. ‘For a country like China, doing this is trivial!’ Except for a coupla things… Heat buildup is a problem for PV, and I’m not sure your black silicon etched wafers would not be a self-incinerating black body with the wrong IR wavelengths to convert to electricity. And as a related article here indicated, “fixing” China’s dependence on imported foreign chips to offset fuel imports, no amount of energy independence will satisfy Xi, and no amount of export self reliance by China will now encourage the global mob to import Chinese anything — what would China import f4rom the world, besides T-Bills, which Beijing derides as being the reserve currency of the declining US?

    Reply
  4. ‘For a country like China doing this is trivial!’Except for a coupla things…Heat buildup is a problem for PV and I’m not sure your black silicon etched wafers would not be a self-incinerating black body with the wrong IR wavelengths to convert to electricity.And as a related article here indicated fixing”” China’s dependence on imported foreign chips to offset fuel imports”” no amount of energy independence will satisfy Xi and no amount of export self reliance by China will now encourage the global mob to import Chinese anything — what would China import f4rom the world besides T-Bills”” which Beijing derides as being the reserve currency of the declining US?”””

    Reply
  5. ‘For a country like China, doing this is trivial!’

    Except for a coupla things…

    Heat buildup is a problem for PV, and I’m not sure your black silicon etched wafers would not be a self-incinerating black body with the wrong IR wavelengths to convert to electricity.

    And as a related article here indicated, “fixing” China’s dependence on imported foreign chips to offset fuel imports, no amount of energy independence will satisfy Xi, and no amount of export self reliance by China will now encourage the global mob to import Chinese anything — what would China import f4rom the world, besides T-Bills, which Beijing derides as being the reserve currency of the declining US?

    Reply
  6. So the solar cell industry is piggybacking on the chip making industry. I always took for granted that there was going to be some technology transfer, but I never realized that there was also a transfer of castoff equipment also. You learn something new everyday.

    Reply
  7. So the solar cell industry is piggybacking on the chip making industry. I always took for granted that there was going to be some technology transfer but I never realized that there was also a transfer of castoff equipment also. You learn something new everyday.

    Reply
  8. So the solar cell industry is piggybacking on the chip making industry. I always took for granted that there was going to be some technology transfer, but I never realized that there was also a transfer of castoff equipment also. You learn something new everyday.

    Reply
  9. Hey, higher efficiency is a good thing. Right? So long as it doesn’t negatively impact the overall life of the cells. Somehow, I kind of doubt that an additional step can be added to silicon PV cell fabrication, yet still reduce the overall cost-per-watt by 10%. Seems pretty steep a belief. Moreover, the use of “black silicon” doesn’t really impact the cells all that much. They already (“blue type”) intercept well over 97% of all incident sunlight. Going from 97% to 99.5% is a small change. Perhaps this announcement only impacts one kind of photocell: the polysilicon or polycrystalline type. Raising their efficiency substantially (from relatively pathetic untreated starting point) would be a good thing; it should be noted that polycrystalline silicon is WAY cheaper than the monocrystalline stuff nominally. Monocrystalline Si is made by loading a huge bowl-like graphite container with radically pure polycrystalline chunks, then heating them to just a few degrees above bulk silicon’s melting point. It all liquifies. Carefully lowering a “seed crystal” firmly bonded to a strong steel cable into the melt causes some of the melt to solidify onto it. Rotating this while retracting upwards, allows the growing crystal to both expand sideways, and in length, while maintaining a single crystal’s aspect over both width and length. Some many hours after the pulling started, the BOULE as it is called, is now maybe a meter to 2 meters in length, and either 100, 150, 200 or 300 millimeters in diameter. (For chip-making, these numbers are “magic”. For PV, they’re not, but since the PV industry economizes by deploying well-used pre-owned chip-making boule-pullers to make their raw silicon, there is a strong correlation to the chip-wafer sizes. ) Anyway. Once ‘they’ have 250 kilogram boules, they pull ’em out, let ’em cool off then using diamond-loaded tungsten wire saws, saw them into round wafers. Then polish them with a variety of grits on large laps. Over

    Reply
  10. Hey higher efficiency is a good thing. Right?So long as it doesn’t negatively impact the overall life of the cells. Somehow I kind of doubt that an additional step can be added to silicon PV cell fabrication yet still reduce the overall cost-per-watt by 10{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12}. Seems pretty steep a belief.Moreover the use of black silicon”” doesn’t really impact the cells all that much. They already (“”””blue type””””) intercept well over 97{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} of all incident sunlight. Going from 97{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} to 99.5{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} is a small change. Perhaps this announcement only impacts one kind of photocell: the polysilicon or polycrystalline type. Raising their efficiency substantially (from relatively pathetic untreated starting point) would be a good thing; it should be noted that polycrystalline silicon is WAY cheaper than the monocrystalline stuff nominally. Monocrystalline Si is made by loading a huge bowl-like graphite container with radically pure polycrystalline chunks”””” then heating them to just a few degrees above bulk silicon’s melting point. It all liquifies. Carefully lowering a “”””seed crystal”””” firmly bonded to a strong steel cable into the melt causes some of the melt to solidify onto it. Rotating this while retracting upwards”” allows the growing crystal to both expand sideways and in length while maintaining a single crystal’s aspect over both width and length. Some many hours after the pulling started the BOULE as it is called is now maybe a meter to 2 meters in length and either 100150 200 or 300 millimeters in diameter. (For chip-making”” these numbers are “”””magic””””. For PV”” they’re not but since the PV industry economizes by deploying well-used pre-owned chip-making boule-pullers to make their raw silicon”

    Reply
  11. Hey, higher efficiency is a good thing. Right?
    So long as it doesn’t negatively impact the overall life of the cells.

    Somehow, I kind of doubt that an additional step can be added to silicon PV cell fabrication, yet still reduce the overall cost-per-watt by 10%. Seems pretty steep a belief.

    Moreover, the use of “black silicon” doesn’t really impact the cells all that much. They already (“blue type”) intercept well over 97% of all incident sunlight. Going from 97% to 99.5% is a small change.

    Perhaps this announcement only impacts one kind of photocell: the polysilicon or polycrystalline type. Raising their efficiency substantially (from relatively pathetic untreated starting point) would be a good thing; it should be noted that polycrystalline silicon is WAY cheaper than the monocrystalline stuff nominally.

    Monocrystalline Si is made by loading a huge bowl-like graphite container with radically pure polycrystalline chunks, then heating them to just a few degrees above bulk silicon’s melting point. It all liquifies. Carefully lowering a “seed crystal” firmly bonded to a strong steel cable into the melt causes some of the melt to solidify onto it. Rotating this while retracting upwards, allows the growing crystal to both expand sideways, and in length, while maintaining a single crystal’s aspect over both width and length.

    Some many hours after the pulling started, the BOULE as it is called, is now maybe a meter to 2 meters in length, and either 100, 150, 200 or 300 millimeters in diameter. (For chip-making, these numbers are “magic”. For PV, they’re not, but since the PV industry economizes by deploying well-used pre-owned chip-making boule-pullers to make their raw silicon, there is a strong correlation to the chip-wafer sizes. )

    Anyway.

    Once ‘they’ have 250 kilogram boules, they pull ’em out, let ’em cool off then using diamond-loaded tungsten wire saws, saw them into round wafers. Then polish them with a variety of grits on large laps. Over and over, until nearly shiny. Then chemical etching to get them to the glossy point. Then into ion implanters that accelerate hydrogen, various kinds of ions at the wafers, depositing them at different depths. Wonderful stuff. All automated. Assembly lines.

    Robots.
    Wafers.

    They’re then trimmed, tested, binned, and shipped (or used locally) to fabricate panels. Which typically have glass on upper and lower surfaces and very-low-oxidation-rate polymers between the layers to environmentally seal the cells against the weathering depredations.

    Anyway.
    Fun stuff.

    GoatGuy

    Reply

Leave a Comment