Skip to main content

Chemical News

Burning In Reverse

This paper has picked up some coverage and headlines, not all of which are accurate. It’s from a group at Oak Ridge that’s looking at methods to use carbon dioxide as a chemical feedstock, and that’s a worthy goal. If we could usefully turn CO2 back into methanol, methane, or some other such material we would effectively be running combustion in reverse, with great benefits for our own atmosphere and probably the world’s economy as well.

This paper is unusual in that it has carbon dioxide being turned into ethanol, so there’s not only a reduction reaction going on, but a carbon-carbon coupling. It’s an electrochemical reaction, done with an exotic new material, copper nanoparticles deposited on the surface of a “carbon nanospike” electrode. This seems to have unusual properties, because what look like broadly similar electrode materials don’t accomplish the same result at all. The whole thing is a good illustration of what large changes can occur at the nanoscale – you’d think that copper is copper, and carbon is carbon, but that’s not true. Carbon can be diamond, graphite, or a zillion things in between, and copper (or any other metal) is a very different beast when it’s partitioned down into tiny particles. What size and what shape those particles can also make a vast difference. Nanotechnology is all about exploiting these differences, but the tricky parts are (1) getting the same materials reproducibly, (2) exploring their gigantic landscape of possibilities in some sort of productive way, and (3) trying to understand what’s going on and why once you find something useful. None of those steps are straightforward.

The headline writers should have read the conclusions section of the paper, however, where it says that “The overpotential. . .probably precludes economic viability for this catalyst“. Basically, you have to use more electric power to get ethanol this way than the resulting ethanol can possibly be worth. The authors suggest some ways that this might be overcome, but those will be matters for a lot of further experimentation. It’s worth noting (and the paper has many references to this effect) that if you just want to chew up carbon dioxide electrochemically, you can already do that with existing technology. The big issues are the cost of the electricity you need to run such a process, and where that electricity comes from. If the amount of carbon dioxide emitted to generate all that electricity is more than the amount you’re removing and turning back into reduced carbon feedstocks, the whole thing is as useful as a vacuum cleaner that sprays extra dirt out the back.

Since combustion itself gives off energy, you’re always going to be using energy to run it in reverse – you can’t escape thermodynamics. But that still gives you a lot of room, with one sort of catalyst or another, to do it in an efficient enough way that makes economic sense. After all, plants are quietly taking up carbon dioxide and converting it to reduced carbon compounds all around us, albeit at only a handful of carbon dioxide molecules per second per enzyme molecule, using sunlight as their energy source in the water-oxidation side of the process. It could well be that some sort of new electrolysis setup, as in this latest work, could do the trick. But as the Oak Ridge team and the others working in this area well know, scale is going to be a big problem here.

Carbon dioxide is found at its largest concentrations where it’s being emitted, but there’s often a fair amount of other gunk coming along with it. You not only have to be able to do your conversion on a gigantic scale, but your catalytic system also has to be able to operate for extended periods without fouling or being consumed by side reactions. (A number of such installations are already running to see how they perform under real-world conditions). If on the other hand you’re going to just quietly pull CO2 out of the ambient atmosphere, you’re going to be working at much lower concentrations of your starting material, so being able to run on a useful scale becomes difficult for that reason as well. The size, surface area, and flow characteristics of such an installation will be the subject of a lot of engineering expertise, and I think it’s fair to say that we’re not yet close to the efficiency needed to make something like this viable.

This latest work is certainly interesting and worth following up on – once you get up to the high electric potentials needed, it works at pretty good efficiency from an electron-use standpoint, and the production of ethanol (instead of methanol) could be a good feature, too. The amount we don’t know about nanoengineered electrode surfaces is vast, and there are surely a lot of useful things to be found in the field. But this isn’t the magic carbon-source solution that some of the stories are claiming. Not yet.

48 comments on “Burning In Reverse”

  1. bearing says:

    Not to mention that if you actually want to burn all that ethanol, you have to spend the energy to distill it out of the water it is dilutely dissolved in.

    1. SP123 says:

      I’m sure they can come up with other uses for ethanol diluted in water.

      1. Isidore says:

        As long as it’s not diluted to less than 80 proof 😉

        1. WilG says:

          Dissolve some of the CO2, dilute to 4%, voila, LITE BEER!

          Sometimes known as sex-in-a-canoe beer.

    2. Samira Peri says:

      Not really, if you can get to whatever the max is allowing for hygroscopic effects. Ethanol only needs to be ~40% to burn. Blended to gasoline you could probably just ignore the 4% water and dump the stuff right in.

  2. Pierre R Bedard says:

    Some of your readers may have forgotten about “Overpotential”.
    Overpotential is the thermodynamic cost of going from the bulk solution (a 3 dimension world) to an electrode surface (a 2 D world). That reduction in entropy (visualize a change in the range of molecular motions) is huge. That alone makes industrial scale commodity production with electrodes impractical.
    By the way hydrogen, on industrial scale, is not produced electrochemically, it is produced from the decomposition of water at high temperature from the burning of natural gas and catalysis.

    1. John Wayne says:

      I thought most hydrogen was produced by cracking hydrocarbons.

      1. Pierre R Bedard says:

        Actually hydrogen is consumed by cracking hydrocarbons.
        Molecules are broken during cracking: carbon-carbon bonds are replaced with carbon-hydrogen bonds. Also, unsaturated bonds are replaced by saturated bonds, a process that also require hydrogen

        1. John Wayne says:

          Thanks for commenting, I just read up on it. For some reason I thought that a lot of the hydrogen came from an ethane to ethylene conversion. As you said, breaking all the carbon-carbon bonds in the feed is the dominant activity in the reactors that actually make hydrogen. Interesting stuff.

    2. Chemperor says:

      Pierre is correct, here. Most H2 is derived from hydrocarbon feedstocks, as electrochemical conversion is still not economically viable or scalable.

      By the way, thanks to Pierre for the wonderful description of overpotential. I don’t believe that I’ve ever considered it in those terms before!

      1. Paul D. says:

        With PV solar electricity becoming very cheap in some places ($0.0242/kWh in a recent utility scale bid in the middle east) the electrolytic route to hydrogen is not as far out of the running as you might think.

    3. c says:

      “By the way hydrogen, on industrial scale, is not produced electrochemically, it is produced from the decomposition of water at high temperature from the burning of natural gas and catalysis.”

      Are you referring to steam reforming? I’ve never heard it described quite that way. It’s not really the “burning” of NG, is it? Oxygen isn’t involved (for good reason) in steam reforming, as far as I am aware.

      1. yuri says:

        I think he was referring to the source of heat for the reforming as in burning of natural gas.

        Btw, great explanation of overpotential and kudo’s to Derek for such a great summary and critique of this paper and what it really means in terms of practicality.

    4. RTW says:

      “That alone makes industrial scale commodity production with electrodes impractical”

      That’s just not accurate. There are many electrochemical processes used on industrial scale, with the most prominent examples being the chloralkalai process and the Hall–Heroult process.

    5. tangent says:

      So why do you not recoup that thermodynamic cost when the molecule goes back into the bulk?

      1. 2nd law and disorder says:

        Overpotential is like irreversible work. An analogy might be this. If your tire pressure is low, you might stop at a gas station and pay for “air” to fill the tires. Of course, you’re really paying for electricity to do the work to compress the air to a suitable pressure to fill the tires. A non -negligible amount of the energy is wasted (heat) to do this. If you then run over a nail, the (useless) PV work that is done by the air leaving your tire is less than the work required to fill the tire. The work that was wasted in compression is not recoverable.

  3. MSP says:

    Why is this not economically feasible? Don’t we make ethanol from corn, likely using more energy from planting seed to the anhydrous solvent than is obtained from using the ethanol as a fuel in our autos? Oh right all we need is a government subsidy to have it all make sense.

    1. Scott says:

      If you’re trying this as a carbon-capture/sequestration method, it doesn’t do any good to be using any fossil-fuel generated electricity to make the ethanol. You need to be using Nuclear or some renewable source (including hydropower).

      But in terms of making auto fuel and telling the Middle East to go pound sand, this may be a useful idea.

      1. loupgarous says:

        We can tell the Middle East to go pound sand right now. Almost all of our crude oil storage capacity is full. US and Canadian oil reserves reachable by fracking or processing of oil shales are capable of supporting current demand for at least a century.

        But it seems to me that the best use of this way of making ethanol from atmospheric CO2 is to use off-peak electrical power. That power’s largely wasted for want of a way to store it (organic bulk batteries may change that), so it largely doesn’t matter if it’s used inefficiently to make ethanol – the power will be lost in any event if no one uses it.

        Optimally, you’d co-locate a CO2 –> ethanol plant near an electrical power station, wind farm or solar farm to reduce transmission losses from Joule heating of wires, and you’d have ethanol made from power which is now wasted.

        1. Scott says:

          Ah, yeah, didn’t think about using Ethanol for off-peak energy storage, which is a real issue for Solar and Wind power generation. Could definitely see co-locating a CO2-Ethanol plant with a wind farm or solar farm, and then either shipping said ethanol to another powerplant or having an actual ethanol powerplant there with the CO2-Ethanol plant.

          Not being a chemist (I’m honestly here for the “Things I won’t work with” and “How not to do it” entries), how well does this process scale up to industrial/continuous production?

          1. Barry says:

            for off-peak energy storage, this is vastly less efficient than existing:
            -pumped water
            -compressed air
            -hydrogen generation/fuel cells

            spending energy to wring CO2 out of the atmosphere and to then reduce it to ethanol and then spend more energy to separate that ethanol from the product stream just to burn it later to get your energy back is ridiculous. The authors eplicitly conclude that the energy required is more than the ethanol produced is worth

  4. Barry says:

    what’s under-played is the selectivity that makes one carbon-carbon bond but doesn’t proceed to oligomeric gunk.

  5. Pierre R Bedard says:

    So Yes
    Two reservoirs inside one of the other: 1- The high temperature is achieved with burning natural gas with oxygen AND 2- Natural gas is reformed in absence of oxygen to produce hydrogen and other building blocks over catalysis.
    So yes my shortcut was too short for chemists, I should have known better on a chemical blog!

  6. Anon says:

    “you have to use more electric power to get ethanol this way than the resulting ethanol can possibly be worth.”

    Reminds me of synthetic hamburger meat – academically interesting but commercially completely non-viable.

    1. Hap says:

      But synthetic hamburger doesn’t have to generate more hamburger than you put in – if you want a niche market (people who don’t want hamburger from animals) then it just has to cost little enough to be affordable to that market. If you were getting it from animals, then it has to kill fewer animals than needed for hamburger itself, but I assume any market for artificial hamburger would be either not wanting any animals to be killed for it or be looking for really cheap, which using more animals to make would probably not be,) Ethanol being used to replace hydrocarbon fuels, though, has to cost less fuel than it can replace – otherwise you’d be better just burning the hydrocarbons than using the ethanol. (If you get electricity to make the ethanol from non-hydrocarbon sources – wind, solar, nuclear- then you could get ethanol without hydrocarbon fuels, but in that case, you’re probably costing a whole lot more than anyone wants to pay, which would be closer to the issue with hamburger.)

  7. Cu NanoTubeS says:

    I think this could be a good application for Copper NanoTubeS

    1. Samira Peri says:

      I see what you did there. 😛

  8. Curt F. says:

    I collaborated on a project that involved electrochemical reduction of CO2. (My job was to make bacteria that ate the products of the reduced CO2; I’m not an electrochemist.)

    I agree that the noteworthy aspects of this paper are the catalyst material and the product. Ethanol is a very unusual and cool product. Most CO2 electroreduction papers make either formate or CO, with formate being more common in aqueous media and CO in nonaqueous media.

    The Faradaic efficiency in this paper is actually worse than seen in many formate or CO-focused studies. But it’s not faradaic efficiency (molar yield per mole of electrons) that matters, it’s energy efficiency. Overpotential measures how much energy that mole of electrons needs to have relative to the thermodynamic minimum in order for you to obtain product. I don’t think Pierre R. Bedard’s description of overpotential is correct. Overpotential varies dramatically by electrode material and by the reaction particulars happening in the electrolyte; it’s not as simple as 2D vs. 3D. Every electrochemical reaction goes from 2D to 3D but overpotentials vary dramatically.

    Hydrogen is not produced at “industrial” scales by electrolysis, but commercial hydrogen generators have long been available for low-volume users of hydrogen. The best overpotentials for the proton reduction reaction to form H2 are only ~100 mV, the best I’ve seen for aqueous CO2 reduction was at least 10 times higher, probably closer to 20 times higher. Since the standard reduction potentials for formate and H2 are similar, those differences in overpotentials translate directly to energy efficiencies.

    Another critical parameter is the attainable current densities. Practical electrolysis schemes get 10s to 100s of milliamps per square cm of electrode surface. The largest-scale electrochemical process I’m aware of is the chlor-alkali process for production of elemental chlorine. Current densities of ~500 mA per square centimeter can be obtained at commerical scale. The paper in the OP didn’t really try to optimize current density, so the fact that they only got up to 2 mA / cm^2 isn’t really a fault, in fact to me it seems pretty good for not having optimized anything.

    1. Pierre R Bedard says:

      Again I had no intention of producing here a complete thesis on overpotential or industrial H2 production.
      In thermodynamics there are oceans of “if buts and when”.
      I just proposed a broad visualization.
      Some found it helpful.

  9. Benny says:

    The real question remains, can these researchers at ORNL modify their setup so that the water-ethanol output drains into, say, large oak casks, and, if, as Oak Ridge is in Eastern Tennessee, can this then be labelled as nanotech whiskey?

    I think that could help with the economic viability of it just a bit.

    1. wlm says:

      You can make “nanotech whiskey” anywhere, but I imagine that if you want to make nanotech Tennessee whiskey, you at least have to filter it through charcoal:

  10. Thomas says:

    The energy use of the whole thing makes it useless from the start, in my opinion.
    To get CO2 from the atmosphere? Plants seem to be the ticket. They even self-reproduce.
    To store electricity? Ethanol is a useful and safe energy carrier. But one needs fairly pure CO2 to feed this process and your car isn’t going to store it while driving. Burning coal to generate electricity and then using solar power to make ethanol out of the CO2 seems to be ‘meh’ at best for a ‘green’ solution.

  11. Mark Murcko says:

    There have been a number of groups doing pretty cool work on using carbonic anhydrase (one of my favorite enzymes) to convert CO2 to baking soda. Has that gone anywhere? Is there any consensus on whether such methods could be economically viable?

  12. zero says:

    There is an entire field devoted to using CO2 on, say, Mars and producing methane or ethanol. Being able to go from CO2 directly to ethanol in one ‘box’ would be tremendously useful, as you can go from ethanol to polyethylene in a second one-step ‘box’.
    As long as it takes less energy than electrolysis + Fischer-Tropsch there is room for it. Even if it’s not competitive on an energy basis, simplicity of the equipment and the lack of unwanted byproducts would be significant advantages.
    This specific process may or may not pan out on Earth, but this kind of nano-scale research has deeper potential.

  13. LiqC says:

    Got into a discussion on Linkedin with an exec from BASF who proudly announced their newest amine-based CO2 scrubbing technology. I asked a few questions and did not receive good answers. Perhaps someone can answer this:

    1) what are they planning to do with the CO2 they recovered from, for example, coal-fired plants?
    2) what is the energy cost of producing this material?
    3) how is it going to make a dent in the amount of atmospheric CO2? There are a trillion tons of CO2 up there, and about a billion is added every year. As a comparison, the largest commodity chemical by volume is sulfuric acid, produced on ~300 MT/yr scale.

    1. Paul D. says:

      The most likely use for scrubbed CO2 is for enhanced oil recovery, at least until they run out of old oil fields that could benefit from EOR.

  14. Anon says:

    Once we approach 100% renewable/no-carbon energy, we can entertain carbon capture methods even if they’re energy inefficient. We would be able to use other energy-inefficient technologies for lots of other things at that point too.

  15. Barry says:

    Day by day, year by year, termites produce more CO2 than all human activity. Fungal celllulases in the soil produce, too. (Both termites and fungi also produce methane, which is a more potent greenhouse gas) For the foreseeable future, if we are to scrub Carbon out of the atmosphere, the best we can do is to pull “waste cellulose” (newsprint, straw, bagasse, sawdust…) out of the cycle. “Biochar” or “anthropogenic peat” are technologies already in hand. And low-valent carbon (peat, charcoal, coal) is far easier to store stably for geologically-relevant periods than is CO2.

    1. Paul D. says:

      Barry: termites do not produce more CO2 than human activities. Where did you obtain that misinformation?

      1. David Gerard says:

        It’s a standard climate change denial claim, e.g. here. That cites “the journal Science (Nov. 5, 1982)”, which is of course just enough cite to look sciencey without actually being checkable. Looking at the table of contents, the relevant paper appears to be Termites: A Potentially Large Source of Atmospheric Methane, Carbon Dioxide, and Molecular Hydrogen, the abstract for which says “Termites may emit large quantities of methane, carbon dioxide, and molecular hydrogen into the atmosphere. Global annual emissions calculated from laboratory measurements could reach 1.5 x 10^14 grams of methane and 5 x 10^16 grams of carbon dioxide.” Last year’s CO2 output is estimated at 38.2 billion tons, which (assuming US short tons) is 3.47 x 10^16 grams. Someone with access to the paper itself should check how strong the claim actually is.

        1. Paul D. says:

          It is my understanding that result was disputed soon after.

          Of course, one has to wonder why termites, which have been around for 100 million years, would have suddenly started causing CO2 to increase in just the past century or two.

          1. Barry says:

            Termites’* production of CO2 and methane has played a large role in out atmosphere’s heat balance for not less than 100million years. The anthropogenic contributions are on top of that. If we could reduce termites’ activity by 20% worldwide, we might offset all current anthropogenic CO2 (or the accumulated cellulose might just feed more/bigger wildfires with the same CO2 released)

            *and their gut commensals, of course

          2. Thomas says:

            The termite thing might actually be true. But termites probably ’emit’ this as it was recently captured by a plant or tree. And actually if we can keep all plant and trees growing forever and never burning down or being eaten or otherwise consumed this would constitute a huge CO2 sink (and trees would grow into heaven no less).
            But this cycle is hard to influence. Fewer trees doesn’t help as trees sequester more CO2 than is subsequently emitted during digesting those trees.
            But out of context, yes, it is a nice ‘denial’ claim.

          3. Barry says:

            in case I wasn’t clear; the last two hundred years of global warming are anthropogenic. The atmospheric heat balance had been managed by termites (and their gut commensals, and soil fungi, and commensals in ungulates’ guts) for millennia before that. The anthropogenic contribution is on top of what had been a steady state.

          4. King Rocker says:

            @Barry: it’s a mix – termite activity has changed due to humans clearing forests and giving them lots of new stuff to feed on! So in a way it’s them, but then it’s also us – just not by burning fuels!

  16. slashdotty says:

    Saw this on /.

    “As I read it, TFA _does_ give a clue as to efficiency.
    60% of the electrons are used for producing ethanol.
    Equilibrium potential for the ethanol reaction is 84 mV.
    The total voltage that is used is 1.2V, which is 14 times as high.
    That means that only 7% of the voltage is used effectively.
    This gives a total energy of a little over 4%.
    In the conclusion, this is mentioned as “The overpotential (which might be lowered with the proper electrolyte, and by separating the hydrogen production to another catalyst) probably precludes economic viability for this catalyst”

    So, they don’t (dare to) mention efficiency directly, but data is presented by which it can be calculated.”

    Does this analysis look correct?

    1. anon electrochemist says:

      No. Faradaic efficiency only describes the selectivity of the catalyst, and is not directly related to the energetic efficiency of the cell.

      That equilibrium potential is at standard state where all species are at 1M concentration, not their open circuit potential with no products and 100% reactants. The competing side reactions here are making hydrogen, methane, and CO, all of which are useful and easy to separate, so you won’t just be venting those into the air. The overpotentials for each of these will depend on their equilibrium potential and their Faradaic efficiencies at a particular current density. Since we’re talking gaseous reactants at mass-transport limits, they will be extremely sensitive to reactor configuration, temperature, and fluid dynamics at the electrode. Even given that, without specifying a conversion factor (ie final EtOH concentration), the energetic efficiency would be meaningless. That’s why they don’t really bother, there’s just too many parameters here.

      Their proposed mechanism explaining why their catalyst is more selective than straight copper is hand-wavy graphene magic. The DFT calculations for graphene electrocatalysis are absolute garbage. We don’t actually know what the mechanism is, so we can’t model the intermediates, which control the selectivity of the whole process. My money is on high index Cu facets having weirdly strong C2 adsorption, something they hint at. This should be fairly easily measured using old-school UHV surface science: mis-cut single xtals, LEED, TPD, etc

  17. Rich Lambert says:

    Why would we want to deprive plants of carbon dioxide?

Comments are closed.