A Real Lava Lamp

Thanks Mark Rober

In this weekend’s edition of ‘Bram gets nerd sniped by something ridiculous so makes a blog post about it to make it somebody else’s problem’ Mark Rober said something about ‘A Lava Lamp made out of real lava’. Unfortunately he just poured lava on a regular lava lamp to destroy it but this does raise the question of whether you could have a real lava lamp which uses a molten salt instead of water.

First the requirements. The lamp is made of three substances: the lamp itself, the ‘liquid’ inside, and the ‘solid’ inside. The lamp must be transparent and remain solid across the range of temperatures used. The ‘liquid’ must be solid at room temperature, become liquid at a high but not too high temperature, and be transparent in its liquid phase. It should also be opaque in its solid phase to give a cool reveal of what the thing does as it heats up but but that’s hard to avoid. The ‘solid’ should have a melting point higher than the ‘liquid’ but not so high that it softens the lamp and be opaque. The density of the ‘solid’ should be just barely below that of the ‘liquid’ in its melted form and just barely above in its solid form to give it that distinctive lava lamp buoyancy effect. The ‘solid’ and ‘liquid’ should not react with each other or stick to the lamp or decompose over time.

That was a lot of requirements, but it does seem to be possible to meet them. The choice for the lamp is obvious: Borosilicate glass. That’s physically strong, transparent, can withstand big temperature changes (due to low thermal expansion) and is chemically inert. All the same reasons why it’s ideal for cookware. It doesn’t get soft until over 800C, so the melting points of the other materials should be well below that.

For the ‘liquid’ there also turns out to only be one real option: Zinc Chloride. That’s transparent and has a melting point of 290C and a density of 2.9 (it’s also opaque at room temperature). The other transparent salts aren’t dense enough.

For the ‘solid’ there once again only seems to be one option: Boron Trioxide. That has a melting point of 450C and a density of 2.46. Every other oxide has a density which is way too high, but this one overshoots it a bit. It’s much easier to get get the densities closer together by making mixing the Boron Trioxide with something heavy than the Zinc Chloride with something light, so some Lead(II) oxide can be mixed in. That has a density of 9.53 so not much of it is needed and a melting point of 888C so the combined melting point will still be completely reasonable. (Due to eutectic-type effects it might be barely higher at all.) It should also add some color, possibly multiple ones because the colors formed depend on how it cools. Bismuth(III) oxide should also work and may be a bit more colorful.

I’m crossing my fingers a bit on these things not reacting but given that they’re glasses and salts it seems reasonable. The glasses may have a bit of a tendency to stick to each other. Hopefully not so much because one is a solid at these temperatures and the other is a liquid, but it’s probably a good idea to coat the top and bottom of the insides of the lamp with Silicon and to use an overall shape where the pieces inside never come close to the walls, in particular having an inverted cone shape at the bottom and a similar tapering at the top. The whole lamp should also be sealed because oxygen and water might react at the high temperatures reached, and there should be an argon bubble at the top because there is some expansion and contraction going on. Those same concerns apply to regular lava lamps which explains a lot about how they’re shaped.

Anyone who wants to feel free to try this build. You don’t need any more permission from me. I’d like to see it happen and don’t have the time to spend on building it myself.