- The current state of the art of capturing CO2 from the atmosphere uses about 240 to 800 kilojoules of energy per mole of CO2.

This newly discovered electrolysis system uses 50 to 120 kilojoules of energy per mole.

That feels big to me.

But then again, the scientific literature is littered with the corpses of discoveries that never quite lived up to their promise or just never went anywhere at all.

And that brings up an existential question for me, which is, how good theoretically could carbon capture get?

And can we figure that out by looking at the beginning and end states and calculating their energies?

Doing some thermodynamics?

Venturing into chemistry's most annoying subfield?

Yes, we can.

(water burbling) The stuff in the flask is calcium hydroxide dissolved in water.

Now, calcium hydroxide is a base.

My exhaled breath is about 4% CO2.

That CO2 reacts with water to form carbonic acid, H2CO3, which then reacts with the base calcium hydroxide to form calcium carbonate, which is not very soluble in water at all.

So, it crashes out of solution as this thing that you're seeing.

This white precipitate.

Now, it looks milky because it's suspended in the water that I'm agitating, but if I were to let this sit for a few hours, it would all settle out to the bottom.

Now, I'm gonna bubble plain old air instead of my breath.

Air is roughly 0.04% CO2, but a hundred times less concentrated than in my exhaled breath, but the underlying chemistry is exactly the same.

So, this too will eventually turn milky.

What I'm doing in both of these cases is removing CO2 and trapping it.

This is analogous to what you might do at a natural gas power plant where the CO2 is concentrated.

This is analogous to direct air capture, directly removing CO2 from the atmosphere.

Either way, the chemistry is pretty magical.

You are selectively removing one invisible gas from a mixture of many invisible gases, even though they're all completely mixed together.

Unfortunately though, I cannot remove all of the CO2 that humanity has ever produced using this one Erlenmeyer flask because the calcium hydroxide solution will just get saturated.

But you can regenerate the calcium hydroxide solution by collecting this precipitate and then heating it.

This is the basic cycle behind carbon capture.

CO2 in the atmosphere dissolves in water to give you carbonic acid.

You react that carbon acid with a base, in this case calcium hydroxide.

That traps your CO2 inside of calcium carbonate.

You then heat the calcium carbonate, which liberates pure carbon dioxide gas and calcium oxide.

You then take the calcium oxide and add water to regenerate your original base, which reacts with more carbonic acid from the CO2 in the atmosphere to give you... (chipmunk squeaking) This underlying chemistry, this exact reaction actually was discovered in the mid 1700s.

And as far back as 1977, an Italian scientist named Cesare Marchetti proposed using the technology to combat climate change.

Now, his idea was to capture CO2 from power plants using potassium hydroxide, and then dump it in the Mediterranean Ocean.

And the problem with this entire idea, some might even call it a fatal flaw, is this.

This step takes a lot of heat and if you generate that heat using fossil fuels, well, that could defeat the whole purpose because you might be releasing more CO2 here than you absorbed here.

So, a lot of research in the carbon capture field has focused on reducing how much heat you need.

For example, this paper describes an extremely high tech powder that can absorb CO2 all on its own without needing to be dissolved in anything.

And you only need to heat it to a gentle 60 degrees C to release that CO2.

Now this stuff is just yellow number 5, but the real stuff is yellow 2.

And 225 grams of it can absorb as much CO2 as one tree does every single year.

But the regeneration step still takes heat.

Not this much heat but still heat.

And then, I discovered this paper which describes a completely different way to capture CO2.

It starts out actually kind of similar.

Using a base dissolved in water, sodium hydroxide, which reacts with CO2 to form sodium carbonate and sodium bicarbonate and solution.

But to regenerate the base and release pure CO2, the authors here do not use heat, they use electrochemistry.

Animation.

Here at the anode, you pull electrons out of hydrogen gas to form hydrogen ions.

The electrons go to the cathode and the hydrogen ions permeate all this stuff to get to the center part where the sodium carbonate and sodium bicarbonate are.

The hydrogen ions react with the carbonate and bicarbonate to form carbonic acid which breaks down into water, CO2, and sodium ions.

That's your pure CO2 right there.

The water and the sodium go to the cathode where they meet the electrons that came off the anode earlier.

The electrons combine with the water to form hydrogen gas and hydroxide ions.

Now, the hydrogen gas gets funneled back to the anode and the hydroxide ions reunite with the sodium ions which forms sodium hydroxide.

So there, you've regenerated the base and that can now go back and absorb more CO2 from the atmosphere.

That whole thing uses no heat at all, none.

Now it does use electricity 'cause it is electrolysis.

And the authors measure an energy consumption of 50 to 120 kilojoules per mole of CO2 captured.

Compare that to the industry standard of 240 to 800.

That's a big improvement.

But how much can this technology keep improving?

Is there a limit here?

Chemistry can actually answer that question.

(exhales deeply) Thermodynamics.

And yes, these are my actual notes from the first and last thermo class that I will ever take.

And one of the most important concepts from this incredibly annoying but very important subfield is something called a state quantity or a state function.

And that is a quantity whose value does not depend on the path it took to get there.

Think of it like this.

Let's say you hike up a hill, and then your buddy skydives and joins you on the same hill.

You now both have the same elevation.

It doesn't matter that you took two very different paths to get there.

And the same thing is true of some quantities in chemistry like internal energy.

And what that means is that if we wanna approximate the energy required to separate CO2 from air, we just have look at the beginning and end states.

The beginning state being CO2 thoroughly mixed with air and the end state being the two things totally separate.

Everything I'm about to do is adapted from Rhett Allain's blog over at Wired.

There's no way I'd be able to do this on my own 'cause thermo is my kryptonite.

Okay, this is one meter.

One meter high by one meter wide by one meter deep gives us a one cubic meter box of air.

And let's simplify things a little and just assume that the air is nitrogen and carbon dioxide.

Capturing the carbon dioxide just means separating it from the nitrogen.

If you do a bunch of math, you'll find that the difference in.. between plain old air and air minus CO2 is roughly 313 joules.

So, you would need to put in 313 joules of energy to pull out all the CO2 from this box of air.

And that translates to 29 kilojoules per mole of CO2 re.. Now again, thermo is my kryptonite.

So, I wanted to confirm this with published scientific papers.

And I found two whose authors seem to disagree about basically everything except this number.

They calculated it more sophisticatedly than I did and they got an even smaller number, 20 kilojoules per mole of CO2 removed.

So, right now we're here.

This new paper could get us down to here, which is a big jump, but there's still a long way to go before we hit the theoretical minimum.

Now, assuming we could ever get down here, how would this handle today's emissions?

42 billion metric tons of CO2 were released in 2024, which is about 950 trillion moles of CO2.

If we wanted to soak all that up, it would take 20 kilojoules per mole times 950 trillion moles, which is 1.9 times 10 to the 16 kilojoules or 19 billion gigajoules of energy.

Great Scott!

19 billion gigajoules!

This is a reference to the movie "Hot Tub Time Machine."

The world produced about 108 billion gigajoules of electricity in 2022.

So, to pull out all of the CO2 we emitted in one year, it would take about 18% of that electricity.

That's not terrible.

Unfortunately, this 20 kilojoule per mole number is the absolute rock bottom minimum the universe would allow if we could somehow build a 100% efficient process.

We cannot build a 100% efficient process because the second law of thermodynamics says that you cannot break even in a cycle, and there is no potential breakthrough that will get around that.

So, what is a realistic efficiency number here?

Let's think about that in two ways.

First, let's assume that the science in this paper gets scaled up while holding on to its 50 to 120 kilojoules per mole number.

That would make it 17 to 40% efficient.

At 40% efficiency, let's be optimistic, it would take 48 billion gigajoules of energy to pull one year's worth of CO2 out of the atmosphere.

That is almost half of global electricity generation.

Second, I called up Climeworks, which is one of the few companies pulling CO2 out of the atmosphere at scale today.

And I asked them, "Hey, how much energy do you think you will be using when you reach gigaton scale?"

And they said about 1500 terawatt hours per gigaton of CO2 removed, which is about 238 kilojoules per mole, which translates to an efficiency of about 8%.

And that number is more realistic because it includes the energy required to run all the fans, pressurize the liberated CO2, and inject it deep underground.

The earlier number we were talking about, that 20 kilojoules per mole theoretical minimum?

That includes none of those steps.

So, at a total process efficiency of around 8%, we would need more than double all electricity generation in a year to pull one year's worth of CO2 out of the atmosphere.

And what that means is that direct air capture is not efficient enough, nor will it ever be efficient enough that we can just keep emitting CO2 like we have been for the past 50 years.

But if we can reduce our emissions by 80 to 90%, then direct air capture could conceivably take care of the rest and get us into that super important carbon negative zone.

So, we should probably get on that like now.