Recycling Lithium Ion Batteries

One of the biggest contributors to greenhouse warming is transport, and the solution that seems to be advocated is to switch to electric vehicles as they do not release CO2, and the usual option is to use the lithium ion battery A problem that I highlighted in a previous blog is we don’t have enough cobalt, and we run out of a lot of other things if we do not recycle. A recent review in Nature (https://doi.org/10.1038/s41586-019-1682-5)   covered recycling and the following depends on that review. The number of vehicles in the world is estimated to reach 2 billion by 2035 and if all are powered by lithium ion batteries the total pack wastes would be 500 million tonnes, and occupy a billion cubic meters. Since the batteries last about nine years, we eventually get drowned in dead batteries, unless we recycle. Also, dead lithium ion batteries are a fire hazard. 

There are two initial approaches, assuming we get the batteries cleanly out of the vehicle. One is to crush the whole and burn off the graphite, plastics, electrolyte, etc, which gives an alloy of Co, Cu, Fe and Ni, together with a slag that contains aluminium and manganese oxides, and some lithium carbonate. This loses over half the mass of the batteries and contributes to more greenhouse warming, which was what we were trying to avoid. Much of the lithium is often lost this way to, and finally, we generate a certain amount of hydrogen fluoride, a very toxic gas. The problem then is to find a use for an alloy of unknown composition. Alternatively, the alloy can be treated with chlorine, or acid, to dissolve it and get the salts of the elements.

The alternative is to disassemble the batteries, and some remaining electricity can be salvaged. It is imperative to avoid short-circuiting the pack, to prevent thermal runaway, which produces hydrofluoric acid and carcinogenic materials, while fire is a continual hazard. A further complication is that total discharge is not desirable because copper can dissolve into the electrolyte, contaminating the materials that could be recycled. There is a further problem that bedevils recycling and arises from free market economics: different manufacturers offer different batteries with different physical configurations, cell types and even different chemistries. Some cells have planar electrodes, others are tightly coiled and there are about five basic types of chemistries used. All have lithium, but additionally: cobalt oxide, iron phosphorus oxide, manganese oxide, nickel/cobalt.aluminium oxide, then there are a variety of cell manufacturers that use oxides of lithium/manganese/cobalt in various mixes. 

Disassembling starts with removing and the wiring, bus bars, and miscellaneous external electronics without short-circuiting the battery, and this gets you to the modules. These may have sealants that are difficult to remove, and then you may find the cells inside stuck together with adhesive, the components may be soldered, and we cannot guarantee zero charge. Then if you get to the cell, clean separation of the cathode, anode, and electrolyte may be difficult, we might encounter nanoparticles which provide a real health risk, the electrolyte may generate hydrogen fluoride and the actual chemistry of the cell may be unclear. The metals in principle available for recycling are cobalt, nickel, lithium, manganese and aluminium, and there is also graphite.

Suppose we try to automate? Automation requires a precisely structured environment, in which the robot makes a pre-programmed repetitive action. In principle, machine sorting would be possible if the batteries had some sort of label that would specify precisely what it was. Reading and directing to a suitable processing stream would be simple, but as yet there are no such labels, which, perforce, must be readable at end of life. It would help recycling if there were some standardised designs, but good luck trying to get that in a market economy. If you opt for manual disssembling, this is very laboour intensive and not a particularly healthy occupation.

If the various parts were separated, metal recovery can be carried out chemically, usually by treating the parts with sulphuric acid and hydrogen peroxide. The next part is to try to separate them, and how you go about that depends on what you think the mixture is. Essentially, you wish to precipitate one material and leave the others, or maybe precipitate two. Perhaps easier is to try to reform the most complex cathode by taking a mix of Ni, Mn, and Co that has been recovered as hydroxides, analysing it and making up what is deficient with new material, then heat treating to make the desired cathode material. This assumes you have physically separated the anodes and cathodes previously.

If the cathodes and anodes have been recovered, in principle they can be directly recycled to make new anodes and cathodes, however the old chemistry is retained. Cathode strips are soaked in N-methylpyrrolidine (NMP) then ultrasonicated to make the powder to be used to reformulate a cathode. Here, it is important that only one type is used, and it means new improved versions are not made. This works best when the state of the battery before recycling was good. Direct recycling is less likely to work for batteries that are old and of unknown provenance. NMP is a rather expensive solvent and somewhat toxic. Direct recycling is the most complicated process.

The real problem is costs. As we reduce the cobalt content, we reduce the value of the metals. Direct recycling may seem good, but if it results in an inferior product, who will buy it? Every step in a process incurs costs, and also produces is own waste stream, including a high level of greenhouse gases. If we accept the Nature review, 2% of the world’s cars would eventually represent a stream of waste that would encircle the planet so we have to do something, but the value of the metals in a lithium ion battery is less than 10% of the cost of the battery, and with all the toxic components, the environmental cost of such electric vehicles is far greater than people think. All the steps generate their own waste streams that have to be dealt with, and most steps would generate their own greenhouse gases. The problem with recycling is that since it usually makes products of inferior quality because of the cost of separating out all the “foreign” material, economics means that in a market economy, only a modest fraction actually gets recycled.