I Used Hydrogen to Grow Aluminum Crystals

Added: 2026-05-05

[00:00:00]Just like water can dissolve carbon dioxide, molten aluminum can dissolve hydrogen gas. It's a little weird to think about, we don't normally imagine gases dissolving into metals, but it happens and it's a big problem. But why does this happen? And what's the limit here? We'll figure that out and we'll use this principle to grow some cool metal geodess. Hydrogen dissolution is an issue in any aluminum foundry because it causes porocity. If one of these bubbles ended up in any part that was expected to carry any sort of load, that part is toast. And it's all caused by water. Well, I mean, foundaries aren't pouring molten aluminum into buckets of water. It's instead caused by the water in the air. Humidity. When a water molecule in the air interacts with the surface of molten aluminum, it underos a chemical reaction forming aluminum oxide and hydrogen gas. This aluminum oxide sits on the surface of the molten aluminum while the hydrogen builds up in the surrounding air. The molten aluminum can then act as a catalyst in a source of energy to cleave the hydrogen hydrogen bond in diatomic hydrogen producing monotomic hydrogen. This atom is small enough to squeeze inside the interstitial spaces of molten aluminum and essentially dissolve. But we don't actually see that yet. This is because how much hydrogen can dissolve is governed by Sever's law where one of the factors the dissolved hydrogen is dependent on is the square root of the partial pressure of hydrogen. So we won't see any meaningful dissolution until we build up a higher concentration of hydrogen in the local atmosphere. How do we get more hydrogen? More water. So in very humid environments where water is plentiful, it can react with the aluminum and form more and more diatomic hydrogen molecules which go on again to be cleaved into monotomic hydrogen.

[00:01:40]Eventually once we build up enough hydrogen in the local atmosphere, we start to see a significant amount of atomic hydrogen always being dissolved within the liquid aluminum itself. And after some more time, we reach the saturation point, the maximum amount of hydrogen that can be dissolved. This saturation point is governed by both the hydrogen's partial pressure as well as the temperature of the aluminum. If we begin to cool down the system, we can see that the solubility limit drops and that the aluminum can't hold as much hydrogen and that dissolved hydrogen is simply rejected from the melt, becomes diatomic hydrogen gas, bubbles up through the liquid, and floats away. So, as the melt cools and less and less hydrogen remains soluble, we evolve some gas until we reach the melting point where Severt's law no longer applies.

[00:02:23]Here, the solubility drops dramatically stepwise as the aluminum solidifies. At the solidification front, all of the dissolved hydrogen is rejected as the atoms solidify. But this rejected hydrogen has nowhere to go because it's trapped by solid aluminum all around it.

[00:02:39]This leads to trapped gas bubbles within the solid part. Now, what I want to do is see how large I can make these bubbles and see if I can use that principle to create some metallic geodess using an alloy we developed in a prior video. The way I plan to do this is to maintain a crucible full of molten aluminum right above the melting point, then capture the steam produced by boiling water and deliver it straight into the melon. I'll let the aluminum remain molten for a while as steam bubbles through it, allowing plentiful hydrogen to dissolve. This is kind of ironic as in foundaries, one way they get rid of dissolved hydrogen is by bubbling an inert gas like argon through the melt. These bubbles have zero hydrogen partial pressure. And so the dissolved hydrogen moves from regions of high concentration or the melt to low concentration, those inert bubbles, and is then carried away. Anyways, let's get started. For my steam source, I just used a beaker of boiling water connected to a copper pipe, which was then compression fitted to 10 in of steel pipe at the end. My rubber stopper here is tight enough to force the steam into the melt, but loose enough to prevent steam pressure buildup. As I melted the aluminum, I also melted/ burnt off the zinc coating on the steel to reduce any random alloying elements. And once the aluminum was molten, I stuck the steel pipe in, and we can see some nice bubbling action. I let the steam bubble through the metal for about 30 minutes or so. One thing that I was surprised about was the sheer amount of slag I was producing. If we take a look at it, we can see it's really foliated. And each of these little sheets here represent one bubble passing through the aluminum and forming a single layer of aluminum oxide. I thought that was pretty cool.

[00:04:12]Anyways, once I thought I bubbled for long enough, I poured off the metal. We can see that as the metal cools, it degasses as we lose hydrogen solubility, confirming that we did dissolve hydrogen into the metal. Taking a closer look at the solid ingot, we can see that this thing is full of pores. And some pieces that I allowed to slow cool were even more porous. Now, this is a interesting concept to illustrate, but I mean, it's just it's just a bubbly piece of metal.

[00:04:37]It's not anything special, let's be honest. So, let's try to make it a little more interesting. To do that, let's change the alloy to one that readily grows large inner metallic crystals, an aluminum copper alloy at around about 45 weight% copper, 55 weight% aluminum. In this material, we have two different phases that exist.

[00:04:55]the theta phase and the alpha phase. As we cool from the melt, we first reach what we can consider the solidification temperature of the theta phase and large theta crystals begin to form. These crystals have a distinct composition of Al2CU.

[00:05:10]So, as they solidify, the composition of the melt changes until we reach this point, the utctic point at which the remaining liquid all solidifies as a dualphase utctic structure. Now this alloy is cool but I think if we dissolve hydrogen in it it can be even cooler. As the primary theta crystals grow they will reject hydrogen into the remaining liquid. So they can grow uninterrupted but as the remaining liquid eventually solidifies we should see that same hydrogen void forming. But this time we should have a bunch of crystals that already exist inside of that void giving us a metal geode. So I alloed my aluminum with copper and turned on the steam. 30 minutes later I was ready to pour. and there was a incredible amount of slag. I poured the metal into this sacrificial zirconia crucible and quickly popped it in the furnace. But before I closed the door, I added a piece of wet wood to the surface just to give some extra water and hydrocarbons to ensure that less hydrogen leaves the metal. After that, I closed up the furnace and let it cool slowly for about 4 hours. And the next morning, we had some solid metal. If we free it from the confines and break it open, I saw exactly what I was looking for. a pretty large region of hydrogen pocity, but filled with these metallic crystals. I was super happy to see this, but as awesome as these looked, they could look even cooler. While I was obsessing over hydrogen paracity, Kevin at the backyard scientist was experimenting with cooling this alloy under vacuum. He DM'd me and said that he had used his vacuum chamber to slow cool the alloy, allowing the inner metallic crystals to grow and pour off the liquid prior to complete solidification. This process created a large central void full of inner metallic crystals. Now, I don't have a vacuum chamber, but I imagined I could mimic his cooling process by just using careful temperature control. So, I melted down the alloy and left the crucible to slowly cool within the furnace. Once the surface was almost solidified, I poured out the excess liquid and allowed the solid that remained to cool down. I could then bust this thing open to reveal that there are in fact a bunch of crystals inside.

[00:07:13]While I was obsessing over dissolving hydrogen to produce a void, the answer was much simpler this entire time.

[00:07:19]Literally, all we have to do is slow cool the alloy to allow those inner metallic crystals to form and then just pour it off before we reach the utctic composition. I found I could easily modify the shape of the crystals by changing the cooling time or changing the composition of the alloy. I'll share those more technical details in the comments. Now I will say though after cracking it open and taking a closer look at the crystals they are of worse quality than what is produced via hydrogen pocity. While pouring off the liquid is much much easier than introducing hydrogen pocity. The hydrogen grown crystals basically only ever existed in a reducing atmosphere and they avoided much more intimate contact with the liquid. So they didn't really pick up any surface oxidation or any liquid scum that aderes to the crystals during a manual pour off. But the crystals produced by both of these methods are really reminiscent of the mineral stibite. It's actually really crazy how similar they look. But I really enjoy these and I made a lot of them during this video and I would like to share them with anyone who is interested. So if you want to pick one of these up for yourself, I put them on the Etsy shop for basically the cost of supplies and materials. So like a dollar plus shipping.