This article describes some recent work on ‘direct air capture’ of carbon dioxide—essentially, sucking it out of the air:

• Jon Gerntner, The tiny Swiss company that thinks it can help stop climate change, New York Times Magazine, 12 February 2019.

There’s a Swiss company called Climeworks that’s built machines that do this—shown in the picture above. So far they are using these machines for purposes other than reducing atmospheric CO2 concentrations: namely, making carbonated water for soft drinks, and getting greenhouses to have lots of carbon dioxide in the air, for tastier vegetables. And they’re just experimental, not economically viable yet:

The company is not turning a profit. To build and install the 18 units at Hinwil, hand-assembled in a second-floor workshop in Zurich, cost between $3 million and $4 million, which is the primary reason it costs the firm between $500 and $600 to remove a metric ton of CO₂ from the air. Even as the company has attracted about $50 million in private investments and grants, it faces the same daunting task that confronted Carl Bosch a century ago: How much can it bring costs down? And how fast can it scale up?

If they ever make it in these markets, greenhouses and carbonation might want 6 megatonnes of CO₂ annually. This is nothing compared to the 37 gigatonnes of CO₂ that we put into the atmosphere in 2018. In principle the technology Climeworks is using could be massively scaled up. After all, Napoleon used aluminum silverware, back when aluminum was more precious than gold… and only later did the technology for making aluminum improve to the point where the metal gained a mass market.

But can Climeworks’ technology actually be scaled up? Some are dubious:

M.I.T.’s Howard Herzog, for instance, an engineer who has spent years looking at the potential for these machines, told me that he thinks the costs will remain between $600 and $1,000 per metric ton. Some of Herzog’s reasons for skepticism are highly technical and relate to the physics of separating gases. Some are more easily grasped. He points out that because direct-air-capture machines have to move tremendous amounts of air through a filter or solution to glean a ton of CO₂ — the gas, for all its global impact, makes up only about 0.04 percent of our atmosphere — the process necessitates large expenditures for energy and big equipment. What he has likewise observed, in analyzing similar industries that separate gases, suggests that translating spreadsheet projections for capturing CO₂ into real-world applications will reveal hidden costs. “I think there has been a lot of hype about this, and it’s not going to revolutionize anything,” he told me, adding that he thinks other negative-emissions technologies will prove cheaper. “At best it’s going to be a bit player.”

What actually is the technology Climeworks is using? And what other technologies are available for sucking carbon dioxide out of the air—or out of the exhaust from fossil-fuel-burning power plants, or out of water?

I’ll have a lot more to say about the latter question in future articles. As for Climeworks, they describe their technology rather briefly here:

• Climeworks, Our technology.

They write:

Our plants capture atmospheric carbon with a filter. Air is drawn into the plant and the CO2 within the air is chemically bound to the filter.

Once the filter is saturated with CO2 it is heated (using mainly low-grade heat as an energy source) to around 100 °C (212 °F). The CO2 is then released from the filter and collected as concentrated CO2 gas to supply to customers or for negative emissions technologies.

CO2-free air is released back into the atmosphere. This continuous cycle is then ready to start again. The filter is reused many times and lasts for several thousand cycles.

What is the filter material?

The filter material is made of porous granulates modified with amines, which bind the CO2 in conjunction with the moisture in the air. This bond is dissolved at temperatures of 100 °C.

So, it seems their technology is an example of ‘amine gas treating’:

• Wikipedia, Amine gas treating.

In future posts I’ll talk a bit more about amine gas treating, but also other methods for absorbing carbon dioxide from air or from solution in water. Maybe you can help me figure out what’s the best method!

28 Responses to Climeworks

  1. Paul Whaley says:

    Another “negative emissions” direct air capture company is Carbon Engineering in Squamish, BC. (

    • John Baez says:

      Thanks, I’ll check out Carbon Engineering. I’m interested in the various technologies being used. The company’s website doesn’t make that information easy to find! Wikipedia says this:

      Carbon Engineering’s DAC system integrates two main cycles. The first cycle is the absorption of CO2 from the atmosphere in a device called an “air contactor” using an alkaline hydroxide solution.The second cycle regenerates the capture liquid used in the air contactor, and delivers pure CO2 as an end product. These cycles operate in tandem continuously, producing a concentrated stream of CO2 gas as an output, and requiring only energy, water, and small material make up streams as inputs. Energy is used in such a way that no new CO2 emissions are incurred, and thus do not counteract what was captured from the air.

      Not sure how that last part is supposed to work; all I can imagine is that they get energy from carbon-free sources (wind, solar, etc.). More interesting is the alkaline hydroxide approach: I don’t know how that works. I see that it would react with CO2, but not how you get the CO2 back out.

      • Steve Wenner says:

        I commented about Carbon Engineering in your later post before seeing this. Sorry about the duplication. Still wondering about the viability of these technologies for renewable energy storage and net-zero carbon fuels.

  2. davidwlocke says:

    From waste to gold.

    It took Superfund rules to build a network for consolidating small quantities of toxics, so they were economic rather than waste.

    Everything has a beginning. Think who else moves air? Every skyscraper moves air so these machines could be everywhere. We are constantly moving air.

  3. tomate says:

    I have the eerie feeling that this technology might violate the second law of thermodynamics…

    • John Baez says:

      Right! That’s a very important point: if they’re burning carbon to produce the electricity to remove CO2 from the atmosphere, they’ll be putting more CO2 into the atmosphere than they’re taking out. But if they use power generated by renewables, they could actually reduce atmospheric CO2 concentrations.

      As I move further with these articles I’ll talk about sequestering CO2 from coal-fired power plants. Here one gets less power from the power plant but manages to sequester most of the CO2. So you’re not reducing atmospheric carbon dioxide concentrations, you’re just increasing them less. This might be worthwhile—if the legal regime allows coal-fired power plants but demands that they cut their carbon emissions.

      • davidwlocke says:

        We end up powering these units via wind, solar, and nuclear. I don’t see nuclear as a long-term solution since the nuclear waste problem persists and has resisted solutions. Burying it is a known bad.

        • ecoquant says:

          Burying it is a known bad.

          Well, it’s complicated, at least in the United States. There are problems with nuclear power but technical solutions to waste disposal in my opinion are the least.

          The first problem is affordability. Nuclear power, as practiced, has a negative learning curve. Until recently (2015) why this is the case was not entirely understood, but now we know it’s because rather than designing and building relatively small, even possibly portable nuclear reactors that could be lashed together to achieve scale, the industry chose to be mega-reactors, and keeps escalating their size. With modularity and replication comes turning nuclear power into commodity power, and it automatically achieves reliability. Now, if a reactor goes offline, it takes its hundreds of megawatts with it.

          The second problem is enormous amount of concrete these require. Making concrete is a huge emitter of CO2.

          The third problem is time-to-construct. Right now, while modular reactors could have been designed and built many years ago, waiting for the industry to switch will take unacceptably long.

          The fourth is waste, and this is in part, I believe, due to the close relationship between the civilian commercial nuclear power industry and the military-supporting one. The military very much wanted to piggyback their needs on top of a commercial one so they didn’t have to build the entire infrastructure. And, naturally, it extends to waste disposal, as proposed at Yucca Mountain. Fact is the French manner of disposal is pretty safe, even long term: Vitrification. You combine wastes with molten glass, let it cool, then submerge the resulting globules in water in long, deep columns drilled in relatively stable rock. The U.S. doesn’t want to do that, namely, the U.S. nuclear military, because some of the wastes are nuclear weapon cores, retired by treaties. They want to be able to reconstitute these if needed. Once vitrified, you can’t. And they don’t want to pay for their own separate waste disposal system.

          Of course, there’s transport problems, and so on.

          Nuclear would have been nice to have once upon a time. But, projections of need for reactors convinced commercial builders of reactors that there was no future in innovation and, at least in the United States, all serious innovation on commercial scale reactors stopped in the 1960s. I worked as a test engineer on contract for what was once Westinghouse Nuclear Automation and they were using 20+ year old designs and safety calculations.

  4. domenico says:

    I am not an expert, but in the industrial plants to produce liquid nitrogen and liquid oxygen, a little fraction of carbon dioxide 0.038% could be produced; if there was a political demand for the obligation to extract and sell carbon dioxide for industrial plants (to reduce and mitigate global warming) then any diffusion of little plants would not be able to reach the global industrial capacity.
    I found only this site:
    for the global market in dollars, not in volumes.
    The problem is the industrial use of carbon dioxide: if the production of industrial plants, or diffuse plants, exceeds the global market request, then there is not economic use.
    I think that in the industrial greenhouse production, an increase of the carbon dioxide can accelerate the growth, and increase the capture of photons (as in the greenhouse effect), but I understand that there are cheaper methods to produce the carbon dioxide (burn something for free), so that only the politics can direct the economy giving tax advantages to those who use, produce and distribute carbon dioxide from air separation.

    • John Baez says:

      Greenhouses and carbonation might want 6 megatonnes of CO₂ annually, which is tiny compared to the 37 gigatonnes of CO₂ that we put into the atmosphere in 2018.

      Plastic manufacture is a bit better, but still not enough. In 2016, worldwide production of plastic was 335 megatonnes, while 9.90 gigatonnes of carbon was emitted by production of energy and cement manufacture. So, even if it we made all our plastics from carbon removed from the air, we’d only be taking out about 1/30th as much carbon from the air as we’re putting in!

      My conclusion: commercial uses of carbon removed from the air can only serve as a stepping-stone toward carbon sequestration at the scales that would significantly slow global warming. But this could in theory be a useful stepping-stone. To bring down the price of carbon sequestration, we’d need to start doing it. The commercial uses might be a good excuse to start doing it.

  5. ecoquant says:

    Thoughts on overall negative emissions economics: It’s difficult to keep up with emissions at present rate. If emissions, could try for drawdown if we, say, peaked over 700 ppm CO2. Still would need to capture or offset 7-11 GtCO2 per annum emissions from just pure agriculture, although that might be done by afforestation. My net was that once ocean carbonic acid equilibration is considered, really need a beginning-to- end cost for the engineering of less than US$100/metric tonne CO2, zero Carbon energy sources for it, and a lot of patience: Drawdown of 200 ppm could take a couple of centuries, depending upon scale of build out.

  6. Wolfgang says:

    As a chemist I think this approach is not scaling up well. Chemical industry removes large amounts of nitrogen from the air to produce liquid nitrogen and rare gases and a lot of nitrogen is also really fixed by turning it into ammonia and its resulting products by the Haber Bosch process. But it is tremendously expensive to do so and needs a lot of valuable energy to run the compressors or maintain the high-temperature, high-pressure reaction conditions and thereby produces a lot of wast energy, i.e. heat (one can visit such production plants and convince oneself easily about the negative balance of energy). The net effect is positive since the products are valuable goods. But there is no huge market for CO2. And it would not makes sense either, because the goal of CO2 fixation would be to remove it permanently. So, the two main obstacles of such processes are: 1) How to not use more energy and thereby produce more CO2 in the beginning, than to remove from air (negative net CO2 balance), 2) How to store large amounts of solid CO2 for indefinite times (CO2 sequestration). It is kind of ironic, that nature has invented the solution for both processes already: plants. And indeed plant growth has increased worldwide in response to higher CO2 levels in the atmosphere. So this could give one a clue, that chemistry is needed (or maybe biology, i.e. genetics) and not physics, to solve this problem effectively. Anyone who tells me, he will solve the problem by physical means alone, seems to me like someone who cannot make up the energy balance. Compare it with another example: While it would be technically possible to extract gold from sea water (the Germans researched this possibility at some time in history to pay off their WW I debts), no one does it, because it is a huge waste of money, more money is spent on the process, than regained by the product. And this is even so, while the normal process of gaining gold is already very costly.

    • John Baez says:

      I agree with all this. The approach being used by Climeworks does use chemistry: amine gas treating. But we may need more clever chemistry, more like trees. Or maybe we just need more trees—or genetically modified trees that sequester carbon in forms that don’t rot and go back to the atmosphere.

      Two vaguely relevant facts:

      One of the larges forms of material processing we engage in is making cement. We’re making about 4 gigatonnes a year, mostly in China (where the statistics are so bad I could easily be 1 gigatonne off). Ordinary cement emits carbon dioxide as it hardens, but it’s possible to make cement that absorbs it:

      • Ben Block, Capturing carbon emissions… in cement?, WorldWatch Institute.

      Each tonne of cement absorbs 0.4 tonnes of carbon dioxide! So I’m curious how cheap can such stuff become, and how sturdy?

      As for trees, a new paper says the world is getting greener. Apparently most of the change comes from new agriculture in India (basically bad for carbon emissions and other forms of pollution) and new trees planted in China (basically good):

      • Dan Charles, You may be surprised to learn which 2 countries are making the globe a lot greener, Morning Edition, NPR, 14 February 2019.

      So, it’s possible to plant significant numbers of trees. At present, forests only remove significant amounts of CO2 from the atmosphere when they’re first growing; then they come into approximate equilibrium.

      • Wolfgang says:

        Yes, I read about their amine process. But this, unfortunately, is part of another problem for me. This amine has to be synthesized (and possibly a lot of energy is needed for that step, essentially amine synthesis traces back to the energy expensive Haber Bosch process before, that has to go into the balancing), and recycled (again, energy consuming). There will be opportunistic costs with transport of the amine etc. If the process works in principle, to solve the problem, the amine synthesis has to be scaled up considerably. I wonder, if this is chemically possible. If so, a bunch of another costs come into play, build the production plants, make some steel for them and the amine carrying train wagons, trucks, and ships (energy, energy, energy). Then, no recycling process is ever perfect. There will be amine leakage into the environment (the amines in question are somehow toxic), and degradation processes rendering them unusable (What are the chemical products of this? Could it be, that some following products are harmful? Could it be, that side products of the synthesis are harmful? Chemists have some bad experiences with scaling up products, once considered beneficial, like tetraethyllead , halocarbons, asbestos, now plastic) I do not even consider the costs of scaling up, in principle, because if this would be the solution, then we should spent the money without thinking about it. I agree, of course, that someone has to come up with ideas, that we have to test them, that it is always good to have a lot of alternatives, but in particular this approach seems not to be as thoroughly researched by now, to stir up some hope that it will be “the thing” to do. Ironically, if done, it would rather fix a lot of nitrogen instead of CO2.

        I think, a solution could be the genetically modification of microorganisms that could fix CO2 somehow by using it for production of energetic molecules (carbohydrates) or structure materials. I mean, microorganisms DID prove their potential to change the atmosphere on a global scale when they started to produce oxygen some million years ago by exactly this process. Dead microorganisms would fix some amounts of CO2 again, like they did before in producing oil. We just have to speed up this process extremely, so that it does not take geological timescale but could keep up with our release of CO2 from fossil sources. At best, this should work without need of human interference anymore. You could think of coupling the task we would like to be performed to the well being of another organism, letting evolution doing the scaling process for you at no costs at all. If we tell these microorganisms to use nitrate and phosporus from their surroundings, or plastic, as a source of chemical energy, this would even solve some other pollution problems as well. Unfortunately, large-scale genetic modification of our environment is not a thing appreciated by “green” people.

        I wonder, too, if nuclear fusion, if ever it could be make work to produce energy on an industrial scale, could help in solving these problems. We would need an abundant, “clean” source of energy to use it for caring for these problems. Until then, we should save energy, where ever we can, for sure.

      • John Baez says:

        Wolfgang wrote:

        I wonder, too, if nuclear fusion, if ever it could be make work to produce energy on an industrial scale, could help in solving these problems.

        My latest hope is while that our civilization will take some serious punishment from global warming during this century, it will survive, and technology will keep improving, along with our wisdom, so we can spend the next two centuries cleaning up the mess we made. Fusion-powered carbon sequestration on a large scale might be part of the answer—or biological carbon sequestration with the help of clever genetic engineering.

        A bunch of species may go extinct; we can try to revive some… but it seems nobody remembered to get the DNA of the bramble Cay melomys: “not charismatic enough”.

  7. domenico says:

    I read on a different blog of an article on the tobacco plant optimization of the Calvin-Benson cycle
    there is an increase in the biomass (>25%), and an increase of the efficiency of the energy of 17%, blocking an archaic pathway that absorb oxygen,
    I am thinking that if the spirulina could be modified genetically with the same efficiency, and if the spirulina could be used for animal feed (fish and cows), and human food source, then there could be a emission reduction and capture of carbon dioxide: a biological mean to obtain the same results like climeworks.
    I am thinking that if the optimization of the Calvin cycle was applied to the natural flora through the seeds, then the global absorption variation would be appreciable.

  8. Guillaume ANDRIEU says:

    And then, once CO2 has been captured from the air, what do we do with it? Where do we store gigatons of CO2? Oceans are already acidified from too much CO2. So… pumping it back in the oil fields? We need to account for transitory storage, energy for transport and reinjection. If used for gas/oil extraction, 50% of the CO2 gets back up (and it defeats the point of targeting negative emissions).

    “However, when the CO2 is injected into the oil field typically up to 50% of the injected volume remains in the reservoir either in the immobile oil or dissolved in the formation water. The remaining CO 2 is extracted from the produced oil and recycled for re-injection.”

  9. Rogier Brussee says:

    Nature stores CO2 in plants on the timescale of centuries. But on the timescale of millions of years, CO2 is sequestered by weathering rocks. We can speed up weathering a million times by increasing the surface area. In particular there is the idea of “green beaches”, crushing the green mineral olivine (Mg2+, Fe2+)2SiO4 Crushed Olvine in the surf of tropical beaches where it is constantly milled and scratched reacts with CO2 according to

    Mg2SiO4 + 4CO2 + 4H2O → 2Mg2+ + 4HCO3− + H4SiO4

    This means in particular that the oceans become less acid. Also note that this reaction is exothermal: we use chemical energy to extract CO2 which is provided by the Olivine itself.

    Olivine is a reasonably common mineral: In fact the earth mantle consists largely of Olivine and in can be found in some volcanic deposits.

    • ecoquant says:

      @Rogier Brussee,

      I think those Urey reactions are slow but can be sped up with catalysts.

      • Rogier Brussee says:

        To make a real dent, it would be necessary to mine and crush cubic kilometres worth of olivine. That is a lot of material to react with catalysers. Fortunately, the rate of weathering of olivine (or serpentine) in a natural environment especially one where it is kept in constant motion so that weathering products (dunerite) are scratched off is much higher than it is the laboratory:

        Crushed to fine sand, it should weather on the timescale of years.

        To be sure, more research is needed, and geo engineering does have environmental consequences too.

        Some numbers:

        1 km^3 olivine = 10^15 cm^3 = 3.310^15 g = 3.3 Gton olivine
        = 3.3
        10^15 / ((224 + 28 + 416) Mole olivine

        Which according to the above reaction scrubs (ideally)

        3.3 * (412 + 816)/(224 + 28 + 416) = 4.1 Gton CO2

        Note that the world CO2 production is ~40Gton/y

        • John Baez says:

          You might like this article I wrote:

          • Azimuth Wiki, Enhanced weathering.

          One calculation I did back then:

          A kilogram of serpentine can dispose about two-thirds of a kilogram of CO2. According to Wikipedia, in 2008 about 31.8 gigatonnes of CO2 were emitted from burning fossil fuels (and more from land use change). So, to counter that we’d need to grind up about 48 gigatonnes of serpentine a year. For comparison, total world cement production in 2009 was about 2.8 gigatonnes. The total amount of material handled by US mines in 2008 was about 5.6 gigatonnes.

          So, we’d need serious pressure from the government to crush enough rock to absorb the CO2 we’re putting out now. By 2018 we reached 37.1 gigatonnes of CO2 emissions, which would require 55.6 gigatonnes of ground serpentine.

        • ecoquant says:

          @John Baez,

          The late Wally Broecker in his talk at BU reported that he and a colleague estimated that to stop increases in forcing from CO2, a system would need to be established that consumed the world’s entire annual supply of Sulfur and converted it into Sulfuric Acid droplets to achieve the albedo hack of solar geoengineering.

          Don’t know if David Keith at Harvard has improved on that at all.

        • ecoquant says:

          @Rogier Brussee,

          I see such capture as less as response to direct emissions as a mechanism which could be used after cessation to help bring down climate (atmosphere + ocean) CO2 concentrations. Going after direct emissions is a fool’s errand, as they’ll increase as pricier sources of fossil fuels are found and exploited.

          Also keeping up with emissions is tough … You calculated quantity, but what’s capacity per unit time?

  10. @alex_sehdeva Twitter says:

    Did biochemists ever find a more efficient system for reducing carbon than photosynthesis? Whether we invent it or not there will be a point when photosynthesis goes ambient to match the atmospheric distribution of co2. Can a photosynthesis apparatus fit in a virus?

  11. Hi John
    Sounds interesting to me
    Room temperature CO2 reduction to solid carbon species on liquid metals featuring atomically thin ceria interfaces

  12. If we’re exploring various methods to achieve negative carbon emissions, a key aspect is figuring out economically viable pathways to scale up those methods. They’ll start small and they’ll inevitably be expensive at first. The ones that get big will get cheaper—per tonne of CO2 removed—as they grow.

    This has various implications. For example, suppose someone builds a machine that sucks CO2 from the air and uses it to make carbonated soft drinks and to make plants grow better in greenhouses. As I mentioned, Climeworks is actually doing this!

    In one sense, this is utterly pointless for fighting climate change, because these markets only use 6 megatonnes of CO2 annually—less than 0.02% of how much CO₂ we’re dumping into the atmosphere!

    But on the other hand, if this method of CO2 scrubbing can be scaled up and become cheaper and cheaper, it’s useful to start exploring the technology now. It could be the first step along some economically viable pathway.

  13. This is from a website that reprinted a blog post without crediting the author (John Baez).

You can use Markdown or HTML in your comments. You can also use LaTeX, like this: $latex E = m c^2 $. The word 'latex' comes right after the first dollar sign, with a space after it.

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.