Artificial Leaf Could Provide Cheap Energy

sciencehabit was one of several readers to tip news of a sunlight-harvesting artificial leaf, writing: "Nearly all the energy we use on this planet starts out as sunlight that plants use to knit chemical bonds. Now, for the first time, researchers at the Massachusetts Institute of Technology have created a potentially cheap, practical artificial leaf that does much the same thing—providing a vast source of energy that's easy to tap. The new device is a silicon wafer about the shape and size of a playing card coated on either side with two different catalysts. The silicon absorbs sunlight and passes that energy to the catalysts to split water into molecules of hydrogen and oxygen. Hydrogen is a fuel that can be either burned or used in a fuel cell to create electricity, reforming water in either case. This means that in theory, anyone with access to water can use it to create a cheap, clean, and available source of fuel."

Comments

So it's a solar cell....

[artor3]

There's only so much insolation to harvest. If this is cheaper and higher efficiency than existing solar cells, then great. Based on the article, it's only 5.5% efficient, so meh. But even if it were 100% efficient, it's not some magical free energy machine, and never can be. While it's true that "nearly all the energy we use on this planet starts out as sunlight", a lot of that energy arrived at earth several millenia ago. In the long run, we're going to need to either use less energy (preferably by making things more efficient, not making do with fewer things) and/or get some near-unlimited fuel source, like fusion.

Re:So it's a solar cell....

[timeOday]

No, actually there is plenty of sunlight to power all current needs and more, if we could capture it efficiently.

Yup, I linked to a page claiming to "debunk" this "myth" on the basis it would take a solar panel the size of Georgia to power the whole earth. Big deal! Vastly more land is consumed by agriculture. Just reclaiming all the space on rooftops, roadways, and parking lots for solar would account for a lot of that, puttng power generation right where it's needed.

And then there's there's the 2/3 of the earth covered by water nobody is making much use of. If cheap solar devices can produce hydrogen, it can be shipped long distances efficiently.

Re:So it's a solar cell....

[Anonymous Coward]

People keep claiming that something the size of Georgia is a sane amount of the earth's surface to cover. You know that agricultural land isn't completely covered with manufactured material, right? Even if it were, how is doubling the human footprint on the planet reasonable? That article really is a successful debunking.

You're talking about mining enough resources to cover Georgia in something manufactured every N years, perhaps N = 30. Think about how bad that is. It's not road bed, either, it's something expensive and chemical-intensive to produce.

I'm completely for solar energy research, and it can totally make sense for some applications, and why not cover an otherwise unused roof when solar cells are cheap enough. But you just have to look at the energy densities of various energy sources to understand that nukes are it. Nukes are much more sustainable than anything that requires to cover 0.01% or more of the earth's surface with anything manufactured. Nukes aren't perfect, but they're by far the most sustainable, safest, cheapest source of energy, if we could ever act rationally as a species.

I mean, energy density (http://en.wikipedia.org/wiki/Energy_density) isn't the whole story, but even the most pedestrian nuke fuel is four orders of magnitude denser than the most exotic chemical fuel. Five would be a realistic comparison, and then another order of magnitude for more refined fuel, and then another three orders of magnitude when fast breeders become really practical. We're talking about eight orders of magnitude denser energy!

Uranium isn't cheap to mine, but it almost doesn't matter, and silicon is not so cheap to refine, either.

So, I'm no expert, and I'm happy to hear some detailed arguments refuting my back-of-the-envelope ones. But even the most cursory glance at the issue from an engineering perspective says nukes are quite safe enough to be the most sustainable form of energy, even if you use TMI technology and just blow up a nuke plant every 10 years. The energy density, and therefor logistical considerations, are just ridiculously lopsided.

Again, I'm all for developing everything, and I'm working hard on residential-scale solar myself as a hobby. But basic physics says there will not be a fairytale ending for solar, wind, biofuels, or fossil fuels. Certainly, there will be local niches where each of these will be the right tool for the job, but at the planetary scale, we can do better than evolution. We have harnessed the power of the atom. WTF is wrong with us.

This is how it goes

[alex_guy_CA]

See cool science article.

Get excited.

Read comments.

Excitement crushed.

Re:Isn't it better with traditional electrolysis?

[loshwomp]

So, unless there's a pretty substantial price benefit to the cell, where's the benefit?

As you have discovered, the economics are precisely the key to solar energy. The power density (Watts/m^2) is unimportant, except for installations with unique constraints (e.g. spacecraft). For terrestrial applications, Watts/$ is the most interesting term.

Similarly, for economic reasons, I don't think electrolysis (or H2) is likely to succeed on a wide scale. The dirty secret of the H2 "economy" is that the hydrogen fuel cell cycle has a round trip efficiency of about 25%. A fuel cell is effectively a battery, and we already have substantially better batteries at a tiny fraction of the cost.

Re:no free energy

[shermo]

The second law of Slashdot:

"No matter how obvious you make the joke, someone will feel the need to correct it for you."