Lithium morphology discovery could lead to safer lithium-metal batteries | Hacker News Reader

Lithium morphology discovery could lead to safer lithium-metal batteries

phys.org

190 pointswglb37 comments

Seems the artwork i did, for my friends research, that got the cover on "Advanced Energy Materials" wasn't quite right?

https://onlinelibrary.wiley.com/toc/16146840/2012/2/3

Crazy that it's over 10 years ago he made a material with so huge gains, and nothing has come of it yet. Scaling really is hard!

I’ve read this comment three times and cannot understand it

> I’ve read this comment three times and cannot understand it

1) He did a nice paint for a journal cover showing the Lithium morphology ten years ago.

2) Good research is slow. Requiring ten years from initial crawls and mistakes to reaching the market is not uncommon.

Thank you. Yeah it was the first sentence in particular I was struggling with

> But the UCLA investigators developed a technique that prevents that corrosion and showed that, in its absence, lithium atoms assemble into a surprising shape—the rhombic dodecahedron, a 12-sided figure similar to the dice used in role-playing games like Dungeons and Dragons.

I've never seen a rhombic dodecahedron d12 before. They do exist after a quick search, but the much more common d12 is just a dodecahedron, with pentagonal faces.

They're not common, but you can get them https://mathsgear.co.uk/products/rhombic-d12-dice

these look like someones trying to play a perspective trick on me

how are they not the common/normal d12s

Count the edges of each flat face - five on a normal d12, four on this one

ah i meant "how are they not" incredulously, as in these should be the ones endorsed by the creators

i've always had a hard time differentiating between a d12 and its neighbors at a glance, but this would make it clearly stand out from the other shapes

"It was surprising for us to discover that when we prevented surface corrosion, instead of these ill-defined shapes, we saw a singular polyhedron that matches theoretical predictions based on the metal's crystal structure."

Why is it surprising that it matches theoretical predictions? So people knew the true shape of lithium in the first place.

In theory, theory and practice are the same. In practice they are (usually) not.

And there are lots of theories, often incompatible.

It’s nice when one of them actually pans out and has positive benefits, as it helps show a potentially useful direction to continue in.

The researchers are looking into "leveraging Li rhombic dodecahedra as nucleation seeds, enabling the subsequent growth of dense Li that improves battery performance compared with a baseline" [1]. Does that means being able to run a more-conventional manufacturing process once you have the seeds?

[1] https://www.nature.com/articles/s41586-023-06235-w

Possibly, I was looking at it thinking they were going for seeding the re-crystallization process during recharging in order to avoid dendrites and maintain cell energy density. I find it is often hard to deduce exactly where they are thinking of applying this due to the incredible competition around patenting and rechargeable battery ideas. (electrolytes, electrode configurations, etc). I've gotten to the point where I search patent applications with the lead authors names in them to get better clarity on why the author thinks they have made a breakthrough :-)

I enjoy the fact that the study’s author is named Yuzhang Li.

Why's that?

Li(thium).

https://en.wikipedia.org/wiki/Nominative_determinism

> "Now that we know the shape of lithium, the question is, Can we tune it so that it forms cubes, which can be packed in densely to increase both the safety and performance of batteries?"

Is the idea that they would stack the dodecahedrons into an approximate cube shape, or can they tweak the synthesis so that lithium cubes are deposited directly?

The latter. Though you can pack rhombic dodecahedrons with 100% efficiency I doubt you could get them to grow that way at an industrial scale.

https://en.wikipedia.org/wiki/Rhombic_dodecahedral_honeycomb

Depends on the economic value - we grow monocrystaline silicon at large scale now, for instance.

Silicon has an inherently cubic crystalline structure and is way less reactive compared to lithium both of which help a lot. Maybe we could figure out monocrystalline lithium but the structure is a lot less forgiving.

We've never been so close to dilithium crystals. `2(5)6 dilithium 2(:)l diallosilicate 1:9:1 heptoferranide` for the uninitiated.

Matt Parker has made a few videos on the rhombic dodecahedron. This one is the best:

https://www.youtube.com/watch?v=oJ7uOj2LRso

A while ago, I saw that someone developed a really safe lithium battery that can be cut and pierced and yet not explode. Whatever happened to that???

lithium iron phosphate perhaps? it's in widespread use...

Is it the shape of lithium's electron shell?

It is the crystalline shape of pure lithium metal which is shown.

aren't shells just these vague clouds where an electron "might" be?

No.

Shells are rather precise representations of the wavelike properties of electrons as they exist bound in atoms.

Electrons are neither exactly waves or particles, they have properties of both.

If you try to interact with an electrons’ particle like properties you tend to get particles, when you try to inject with electrons’ wavelike properties you tend to get waves.

The dual nature can be hard to say just right and communicate.

The math of what’s going on describes it extremely well… the “this is what’s actually going on” descriptions are more like metaphorical stories because on a human scale quantum objects just don’t exist.

On this dual nature, can it be that the electron cloud is the electron itself? i.e. it's not a point-like particle that jumps around, but it's the probability function itself, that upon interaction with another cloud-particle may shrink into a tiny ball-like cloud, but never a point. In atoms, electrons interact in such a way that they assume curious drum-like shapes, but when they are set free somewhere in the interstellar space, the same electron will be a gigantic planet-size, albeit very thin, cloud. Edit: continuing this speculation, in the double-slit experiment an electron passes thru both slits, then shrinks into an atom-size cloud upon contact with the screen, and a "weak measurement" would be a way to slightly disturb the shape of the electron, to correspondingly disturb the shape of the detector.

The “reality” of any QM interpretation should be taken with a grain of salt and not too seriously. There is some debate, but yet what you’re proposing goes considerably too far. The dual nature is more or less real, you can’t distill it into one exact thing.

QED describes the behavior of electrons very well. Interpretations are projected from equations but what a thing “actually” is only comes from the models.

It can be hard to let go of wanting to come up with an intuitively satisfying explanation but there is not one. Your brain evolved in s world where experiences are classical not quantum. Unless you spend a ton of time interacting with quantum systems (and still maybe not) you’re not going to come up with an intuitively satisfying explanation of what an electron is.

The short answer is no.

There are experiments in which the electron does behave as a point-like particle (definitely smaller than anything that is currently measurable, if it has any size at all). In particular, any interaction always consumes or creates exactly one electron, there is never a half electron being affected (of course, it could be 2 or 3 etc electrons - but always a whole number, never a fraction). So, it doesn't behave like a cloud of something.

Secondly, there are experiments where it behaves like a wave in a field, showing self-interaction and peaks and troughs etc. This continues to happen even if you space it very far apart from other electrons. Experiments have been done using an electron cannon that can only fire one electron every hour or so at a double slit, and still you get the same diffusion pattern on the screen. So, the electron wave is still there and shows self-interference, even if you send it slowly. Clouds again don't have this type of behavior.

Not to mention, if the charge of an electron were distributed in a large cloud, that would have very measurable consequences for the electic and magnetic fields of that electron.

As others are pointing out, there is no way to think classically about quantum phenomena and still match all experiments.

What's the magnetic field of an electron in the double slit experiment before the electron has decided where it wants to be? The electron travels slowly, the detector screen may be very far, while the magnetic field travels fast.

In a sense you're right - it's not an electron that jumps around, the electron is its quantum amplitude wavefunction. Until you measure it, that is what you need to calculate the details of and time evolution of and interactions with.

And there can be electron wavefunctions that are macroscopically distributed for sure.

But as the electron's eventual interaction shape is a point, and never more than one point (even if the wavefunction was a large drumbell shape before!) we can't say that the electron "grows thinner and larger" itself, we have to conclude that there is a thing called a wavefunction (or quantum amplitude) that evolves that is not a point, that can be used to predict where/when the single-point "full electron" interaction with something else might take place.

A quantum weak measurement is a measurement where the object you want to study is allowed to interact with a part of the detector that doesn't immediately reduce everything to a single point interaction. As you correctly note, it does change the shape of the wavefunction, but what you also need to know is that the wavefunction change is not of the electron and the detector-part by themselves - there is a combined wavefunction for the electron + detector-part pair now, with a distribution of amplitudes for their possible states.

It quickly gets intuitively messy but this is the core...

No. The 'faceting' of a crystal is determined by a minimization of the 'Gibbs free energy' [a thermo quantity determined mostly by the relative energies of different faces of the crystal, at low temp] over all the faces. The relative energies of the facets are in turn determined by the crystal structure, and the favored crystal structure is determined by the 'electronic structure' of the material itself. Li is an alkali atom, so presumably in the metal donates its outermost 's' electron to make a relatively simple metal. Almost never does crystal shape in any way reflect directly the atomic structure

Next: The Shape of Water

I laughed

Reply on news.ycombinator.com