Redneck Engineering News: A Molten Lithium Battery Renaissance, Or How We Took “Rechargeable” Rather Too Literally

Introduction

One does read, with the sort of mild interest one reserves for particularly ambitious gardening columns, that the scientific press has been positively giddy about the latest developments in molten lithium batteries. These are not, one hastens to add, your common-or-garden 18650s that occasionally grace the bottom of a kitchen drawer. No, these are altogether more ambitious affairs: grid-scale wonders wherein lithium, or some equally characterful metal alloy, exists in a state of cheerful liquidity at several hundred degrees Celsius, sloshing about like particularly excitable soup in a ceramic cauldron. The promise, apparently, is prodigious current density, immortality of cycle life, and the sort of robust indifference to abuse that would make a Victorian battleship blush.

Naturally, we felt it only proper to have a crack ourselves. One mustn’t let the boffins have all the fun.

The Preliminary Investigation

It seemed only reasonable, before committing to the full grandeur of a molten lithium system, to acquaint ourselves with the subject through what one might call a graduated programme of optimism. We would begin small. We would be methodical. We would, in the fullness of time, work up to the proper thing.

We did not, in the fullness of time, work up to the proper thing. But the journey was educational.

So, how to get hold of a molten lithium battery? Of course, by melting a lithium battery! Our first subjects were the humble 18650 — that stalwart cylinder of modern civilisation, powering everything from electric velocipedes to the sort of torch one buys with great confidence and loses within a fortnight. We had several dozen of these, sourced from a battery pack of uncertain provenance and, as it transpired, equally uncertain state of charge. This latter detail would prove significant.

The literature — and here we use the term "literature" generously, to include certain forum posts of the more adventurous variety — suggested that lithium melts at a perfectly manageable 180 degrees Celsius, which is barely more than one's oven at a vigorous roast. What the literature was somewhat coy about was the steel casing surrounding the lithium, the separator, the electrolyte, the cathode material, and approximately seventeen other constituents that have no intention whatsoever of participating in an orderly phase transition. The 18650, it turns out, does not so much melt as object, loudly and at some length, to the entire proposition. The first cell expressed this objection at approximately 140 degrees. The second held its counsel until 160, then expressed it more emphatically. The workbench acquired a patina that we chose to regard as characterful.

We regrouped.

Our second line of enquiry involved a collection of CR2032s — those flat, silvery discs that lurk in the backs of remote controls, bathroom scales, and the sort of novelty birthday card that plays "Happy Birthday" until the batteries die, which in our experience takes approximately eleven years longer than anyone would wish. These had been accumulated over a period best described as "geological" and stored in a drawer alongside expired coupons and a mysterious allen key of unknown provenance. Their state of charge was, charitably, a matter of philosophical inquiry rather than engineering certainty.

The CR2032, one must acknowledge, has a certain dignity. It is small. It is round. It contains lithium, technically speaking, in a solid rather than metallic form, which means that melting it produces not a gleaming puddle of reactive metal but rather a sort of acrid soup that smells of regret and chemistry gone wrong. Several cells, when introduced to moderate heat, simply declined to participate at all, offering instead a gentle hiss and a wisp of something that our health and safety intuitions suggested we should not be breathing. Others were more theatrical. One, whose state of charge turned out to be rather higher than its companions, contributed a small but vigorous fountain of sparks that would not have disgraced a modest fireworks display. Mark observed that this was "not ideal." It was also the first (though not last) occasion on which Przemek remarked, “o kurwa“. We were developing a sense of what ideal was not.

The CR2032s were retired from active service. Their drawer awaits them, alongside the allen key.

We turned, with the sort of determined optimism that has historically preceded both great discoveries and significant property damage, to the pouch cells. These had come from a laptop of distinguished vintage, and bore the characteristic slight convexity of cells that have spent some years in quiet contemplation of their own internal chemistry. "Slightly swollen," we had noted in our preliminary inventory. In retrospect, "slightly swollen" was doing considerable heavy lifting as a descriptor, in much the same way that one might describe the North Sea in February as "a bit fresh."

The pouch cell is, by design, a thing of elegant minimalism: a laminated foil envelope containing layers of electrode material separated by a polymer membrane, the whole assembly resembling nothing so much as an extremely dangerous sandwich. When heated, the pouch cell progresses through several distinct stages, which we were able to observe with the sort of scientific detachment that becomes easier with practice and a longer pair of tongs. First: softening of the foil. Second: venting, with an assertiveness that suggested the cell had been waiting some time for this opportunity. Third: what we have since taken to calling the enthusiasm phase, in which the cell's opinions about the experimental protocol became briefly but vividly visible across an area somewhat larger than we had cordoned off.

The electrolyte, for those keeping score at home, has a flash point that proved to be rather lower than the temperature required to melt anything useful.

We noted all of this carefully. We made several amendments to our methodology. We extended the cordon.

It was at this point, with the accumulated wisdom of three failed specimen types and a workbench that had begun to acquire strong opinions about retirement, that we procured the graphite crucibles.

The Methodology

The first, but by no means least, discovery, made rather earlier than anticipated, was that the principal constituent of any lithium battery — molten or otherwise — is, in point of fact, fire. Quite a lot of it. We had scarcely begun proceedings when several cells expressed their enthusiasm for the project by liberating generous quantities of this fiery essence across the workbench. I've never heard Mark saying “well, that's not quite ideal“ so quickly. A minor conflagration ensued, the sort one might charitably describe as “lively.” After a brief interlude involving a garden hose and some colourful language, we regrouped with the calm determination of Englishmen who have queued for hours only to discover the café has run out of scones.

Undeterred, we procured a pair of proper graphite crucibles (the sort normally reserved for chaps melting down wedding rings after unfortunate divorces) and a propane burner that had previously seen service rendering particularly stubborn paint from a 1978 Land Rover. With the air of gentlemen conducting a perfectly reasonable experiment in the potting shed, we applied heat. Slowly at first, then with the sort of increasing confidence that tends to precede insurance claims.

Reader, they melted. Not gracefully, you understand. More in the manner of a reluctant dinner guest finally agreeing to leave at half past three in the morning. The lithium salts and assorted electrode innards surrendered to the flames with a series of pops, fizzes, and the occasional small jet of something that looked worryingly like it wanted to join the local amateur dramatics society as a flamethrower. The graphite held, which was sporting of it. The workbench, less so.

Emboldened by this modest triumph, we turned our attention to the next logical step: charging the molten battery. After all, what is a battery if not a device one may recharge? Here we encountered what the more euphemistic academic papers might term “a materials compatibility challenge.”

The copper lead wires, those faithful servants of low-temperature electrical work, took one look at the incandescent soup and decided they would rather liquefy and alloy themselves enthusiastically with the contents. The result was less a functioning electrochemical cell and more a sort of metallic stew with coppery undertones — the battery equivalent of adding too much Bovril to an already questionable casserole. Voltage readings became erratic. Sparks were thrown. A small but determined plume of smoke suggested that the experiment had achieved thermal runaway with all the understated drama of a minor volcanic event.

We did, eventually, achieve something that might, in the right light and with a following wind, be called molten. Whether it could ever be coaxed into storing and releasing energy in an orderly fashion remains, at time of writing, an open question best pondered over a stiff drink and several yards of safety tape.

Still, one must look on the bright side. The neighbours now regard us with the sort of respectful wariness usually reserved for men who keep tigers as pets. And we have produced, if nothing else, the world’s most committed road flare from what began life as a perfectly innocent power tool battery pack. We have not yet reached the stage of our shed having become technically exterior.

One rather suspects the chaps in the proper laboratories, with their fancy solid electrolytes and sealed high-temperature casings, might raise an eyebrow at our methodology. But then, they probably haven’t tried it with a propane torch and a positive attitude.

All in all, a most educational afternoon. We shan’t be repeating it immediately. Probably.