It’s looking like modern electrochemistry might be a big part of the chemical industry in the years to come, and part of the shifts to carbon-neutral production as well. Electricity, after all, is what green/renewable energy production is directly producing (solar panels, wind turbines, hydroelectric power) without the intermediacy of burning carbon-based fuels. Nuclear fission (and eventually nuclear fusion) are in the electricity-without-burning-fossil-fuel category as well, of course, with one intermediate step: the generator turbines being run by internal steam systems off the nuclear reaction heat.
So it would be useful to take the increasing amount of non-carbon-dioxide-emission electricity and use it to improve some of the nastier fuel-consuming parts of the chemical economy, and indeed, ideally to use it to make carbon dioxide itself into a chemical feedstock. Plants have the prior art on that with photosynthesis, of course, and there are all kinds of efforts to mimic that chemically. But using CO2 as a starting material is a matter of chemical reduction – putting electrons into the system – and electrochemistry is a direct way of doing that. I wrote about one of these in 2016, a catalyst system that (unusually) produced ethanol rather than carbon monoxide or formic acid, which are the products you get from an initial two-electron reduction.
The tricky part with doing electrochemistry on large scale is that it’s a surface phenomenon. That leads us to the idea of “overpotential”. For these reduction (or oxidation) steps, there are of course thermodynamic potentials for each step. Those are the energy tolls that you cannot avoid paying – energy is liberated when you burn carbon-based compounds and generate carbon dioxide, and energy needs to be put back in to run combustion in reverse and generate any such compounds from carbon dioxide in turn. If an electrochemical transformation worked perfectly, the voltage (potential difference) where the reaction took place would be ths same as that thermodynamic potential, but that is never the case. It’s going to depend on the reaction, the medium, the electrode, the cell design and more factors besides, but there are many factors which cause the reaction to have to be pushed harder than the ideal prediction. Diffusion of electrolytes, formation of layers of ions at the electrodes, formation of gaseous bubbles at the surface, surface polarization – these phenomena and many others, some not well understood, will mess with your process. And if the overpotential is too high, it can make the whole idea nonviable: spending one hundred dollars of electricity to make seventy-five dollars worth of product is not an attractive proposition.
Now, if the electricity gets cheaper, or if there are externalities that make the product more valuable than it seems, you can still be in business. This especially applies to carbon capture. Non-carbon-dioxide electricity has indeed been getting cheaper, and the benefits of large scale carbon dioxide scrubbing (if it can be achieved, and with such power sources) are indeed high – higher than the book monetary value of the ethanol or what have you.
Here’s a new paper that goes all the way from carbon dioxide to butanol. And while smelly, butanol has definite appeal in this area. It has comparable energetics to gasoline, and can be burned in internal combustion engines in the same way. One of the reasons it’s so difficult to shake liquid fossil fuels is their volumetric energy density – there’s a lot of oomph per gallon, a property definitely not shared by hydrogen, to pick another “green fuel” candidate. Gasoline comes in at about 35 MJ/liter, and butanol at about 30 (ethanol is about 25. Hydrogen at one atmosphere is just barely above zero on this scale, and even at rather impractical pressures of hundreds of atmospheres barely hits 5 MJ/liter.
The work uses a chromium-gallium oxide catalyst doped with small amounts of (crucial) nickel, and the overpotential starts at about 320 mV, which is lower than many such systems, although the highest efficiencies take place at higher overpotentials. At lower voltages the system produces 3-hydroxybutanal, and without nickel in the system you get neither of those products. It looks like acetaldehyde is the key intermediate, with the nickel sequentially coupling that to give you the higher-carbon products. This system is not going to start pumping out the butanol on an industrial scale for us, but it’s a useful step along the way.
A key step in using carbon dioxide in this way would be a way to remove it from the environment and concentrate it without spending too much energy along the way. That’s a big topic all its own, but sticking with the electrochemistry theme, it’s worth noting this recent paper, proposing a silver/bismuth-driven electrochemical route to removing CO2 from seawater. The team claims that their preliminary analysis indicates that this process could be economically feasible, which is a good start, because many such ideas don’t even make it past that hurdle as things stand now.
And while I’m on the electrochemistry, this new work gets at one of the other uses mentioned above, the possibility of replacing energy-intensive chemical processes with better ones. In this case, it’s the production of elemental phosphorus, which is pretty painful. Although there’s a huge amount of mineral phosphate in the world, the whole phosphoric acid industrial chemical production world is based on reducing those minerals (calcium phosphate specifically) at high temperature with silica and a carbon source, to produce plenty of carbon monoxide, plenty of metasilicate slag, and some P4 white phosphorus. High-quality phosphoric acid and its derivatives, along with all the phosphite and phosphonates, etc., comes from that reduction step to the pure element, followed by various sorts of re-oxidation.
There have been many attempts to get something more efficient working, and the paper linked above shows a new electrochemical route. It’s still done at high temperatures (800C in a glassy carbon crucible), but it’s more efficient than the existing route and could be run by non-carbon-emitting electrical sources. In this case, you add phosphoryl anhydride species to the mix as oxide acceptors (instead of the silica in the all-thermal process) – a similar electrochemical route has also recently been reported for the production of elemental silicon, another element that is only really found in its (very stable) oxidized state and has to be beaten back down at high temperatures if you want to get more uses out of it. The classic example in this area is of course aluminum, where electrochemistry is the only way.
So bring on the electrons! We’re only starting to find out what’s possible. . .
