Battery research is a fraught area to report on. In the lab, researchers manage to show something spectacular, like high current density or excellent recharge characteristics. But the part where the battery catches fire and destroys the lab is left out when the story makes the popular press. I am guilty of this myself, and I’m about to do it again. (Not really.)
Lithium-oxygen batteries are very promising, but most current iterations manage to destroy themselves after a few charge/discharge cycles. A recent publication in the Journal of the American Chemical Society shows some progress in overcoming the problems associated with lithium-oxygen batteries.
We all love our lithium-ion batteries. Even though they are the best that we have, they still suck pretty hard. To put it in perspective, lithium-ion batteries top out at about 200mAh/g, so you need a very heavy battery to get much energy. Lithium-oxygen, on the other hand, promises 1675mAh/g, a very respectable energy density.
Unfortunately, batteries are not simple devices. It’s not enough to have a positive ion (lithium) give an electron to oxygen via some pathway that allows us to extract work. To make the battery function, we must have electrodes to collect the electrons, and we need a material that allows the ions to move around the battery (called an electrolyte, which is, basically, a salt). The presence of all these other materials means that there are a multitude of other undesirable reactions that might take place.
This is where lithium oxygen batteries are currently stuck: the side reactions destroy the battery much too quickly for it to be useful. To put this simple sentence in perspective, here is an example of what can happen. The lithium ions proceed to the carbon electrode, where they donate electrons as planned. In doing so, some of the lithium ions react to form a different lithium oxide than the desired form.
This lithium oxide can’t dissolve in the electrolyte, so a non-conducting oxide starts to form on the electrode. At the same time, other lithium ions react with both the electrode and the oxygen to form an insoluble non-conducting carbonate that also coats the electrode. Meanwhile, the electrolyte decides that the situation is all too much for it and tears itself apart, creating yet more useless byproducts that slow the transfer of ions.
After a few cycles, the battery coughs up some electrolyte and drowns in its own fluids.
Some of these problems can be overcome by accepting a lower charging efficiency. For example, by applying a much higher voltage (compared to the operating voltage of the battery) on recharge, the oxide can be broken up. A high enough voltage will even get rid of the carbonate, but this treatment will speed the destruction of the electrolyte. Either way, the battery dies.
Alternatively, if the oxide and the carbonate didn’t form a layer on the electrode, the loss in capacity would be much less dramatic—we would lose ions from cycle to cycle, but the electrodes would still function properly.
Taming of the chemistry
The idea behind the new research is charmingly simple. The electrolyte is key: a liquid electrolyte will allow ions and reaction products to diffuse to and from the electrode faster. But most electrolytes are a gel-like organic material because most inorganic salts (like table salt) are not liquid at room temperature.
If you heat the battery up, however, even an inorganic salt will melt. You then have a liquid electrolyte that is very difficult to destroy. At high temperature, more of the reaction byproducts can be dissolved into the electrolyte. This should prevent the destruction of the electrolyte and transport more of the undesirables away from the electrode.
To test this idea, the researchers constructed a sealed cell that contained the inorganic salt, a lithium salt, and oxygen gas. They ran the battery through several discharge/charge cycles at an elevated temperature. During these tests, they monitored not just the current and voltage profile of the battery but also the oxygen pressure. They could see from the changes in oxygen pressure that the indestructible oxide and carbonate products were still being formed, though at a lower rate.
More importantly, however, was that the discharge/recharge profile was much more stable (in comparison with a battery that had an organic electrolyte), indicating that the byproducts were not coating the electrode as fast.
After cycling a few times, the researchers broke open the battery and examined the electrode and the content of the electrolyte. They found that the electrodes were mainly coated with the carbonate, while the indestructible oxide had diffused away into the electrolyte to form small, free-floating crystals. This meant that over the time period they observed, the majority of reduction in capacity was due to carbonate formation on the electrode.
Are hot batteries useful?
Although this is a nice demonstration, it doesn’t come close to being a complete solution for lithium air batteries. Although it might be possible to put up with the oxide forming in the electrolyte, the growth of a carbonate layer on the electrodes must be prevented. The good news is that there is a solution for carbonate formation as well. It’s just a question of whether that solution is compatible with molten salt.
The operating temperature (around 100 degrees Celsius, depending on the salt) of the battery also means that these batteries will never be used in electronic devices. The researchers are aiming for electric vehicles, aircraft, and solar storage. It also puts users in the interesting situation of needing a battery to provide the heat to get your main battery started.
This result represents an unavoidable efficiency loss, but how bad is it? That’s hard to say because you only need a small amount of electrolytes between the two electrodes. However, salts have quite a high heat capacity, which means that they require quite a lot of energy to melt. But even with that additional cost, this looks like a promising bit of progress.
Journal of the American Chemical Society, 2016, DOI: 10.1021/jacs.5b11744