I'm interested in minerals and rocks. They are super neat. A mineral is a chemical in one of seven crystal forms or crystal systems. For example, carbon in the hexagonal crystal system is called graphite, and carbon in the cubic crystal system is called diamond. Want to hear all seven crystal systems? Of course you do. There's even a song: ♫ Triclinic, monoclinic, orthorhombic, tetragonal, | trigonal, hexagonal, isometric/cubic! ♫ Pretty catchy, I'd say.
I want to make my own crystals. I want to make lots of different kinds. I want to be able to make any kind of crystal. What can I make?
Kids make crystals out of table salt, or alum, or borax, or epsom salts, by saturating hot water and letting it cool. These are fine. Kind of fragile. Table salt in the cubic system is called halite. I think it's always and only cubic. Kids make gypsum too. That's hydrated anhydrite. Basically the same as plaster of Paris. And um... you can grow aragonite (calcium carbonate in the orthorhombic system) from dolomite in vinegar, I think, which would be more impressive if dolomite were not just calcium carbonate with some extra magnesium. I haven't heard of kids doing it as a science fair project, but potassium chloride is one of the main water-soluble salts in wood ash (behind potassium carbonate and potassium sulfate, both of which can be converted into potassium chloride and some other stuff by treatment with hydrochloric acid) and crystals of it can be grown in the same way by cooling a supersaturated aqueous solution of potassium chloride. The mineral made that way is called sylvite, and it forms in the isometric crystal system, and like almost everything else in this paragraph, it's an evaporite, which means it can form naturally by the evaporation of water. Several evaporites are unfortunately deliquescent, which means they can absorb atmospheric moisture and then dissolve in it, so they have to be sealed in airtight containers. Or coated with something? It would be nice if you could just cover deliquescent evaporites in nail polish or polyurethane or something. Two other evaporites I kind of want to make someday are bischofite (a hydrate of magnesium chloride, which can be produced by combining magnesium hydroxide (milk of magnesia) with hydrochloric acid) and carnallite, which is like a cross between bischofite and sylvite (between hydrated magnesium chloride and potassium chloride). You can also grow deep, deep blue crystals from an aqueous solution of copper sulfate and it's only slightly poisonous. The powder is sold in gardening centers as a fungicide, and the hydrated mineral form is called Chalcanthite.
I also want to make more advanced crystals that don't spontaneously dissolve in air. How about gemstones? Can I make them in a melting furnace? Or even just non-precious crystals that are a little more solid?
Well, you can turn aragonite into calcite (you can polymorph it from the orthorhombic to the trigonal system) with heat, and that's on my to-do list. I like calcite a lot. Sometimes it's fluorescent. Sometimes you can use it to locate the sun in the sky when the sky is cloudy (Iceland spars have refraction that depends on the polarization of incoming light). I already mentioned growing aragonite from dolomite. Google Scholar says there are lots of ways to make whiskers of aragonite by simple chemistry in atmospheric conditions. So there are ways to make feedstock for synthetic calcite, if other sources of calcium carbonate like crushed eggshells or crushed limestone don't work.
What kind of heat do you need to polymorph aragonite into calcite? A kiln will do it after a couple of hours. You just need to hold the crystal at a couple hundred degrees celsius. I think I'm going to try a furnace though. A furnace is a thermally insulated box with a torch pointed at it. If you use it for pouring molten metal into a cast, then you can call it a melting furnace. If you use it to heat up metal and then whack it, you can call it a forge. But either way, it's just a box. The thermally insulating material the box is made from is called refractory. I've seen forges made of dirt. You can use fire bricks if you're fancy. You can make your own fire brick out of kaolin clay or refractory cement. Some people make their own refractory out of like a mixture of sand and perlite and plaster of Paris. Perlite is that foamy stuff in potting soil, which is actually a hydrated version of obsidian glass that puffed up when heated. Pretty cool. It's not a crystal; it's a glass, but still cool. Some clays puff up the same way when you get them really hot. That would be fun. When I've made a furnace, I'll puff things up too.
Let's talk a little bit about the history of making crystals with torches. The first torches used a mix of hydrogen gas and oxygen gas: oxy-hydrogen. It's a nice gas mixture. It burns very hot and it just makes water as a byproduct and you can get it by splitting water. Easy, abundant, clean, effective. Awesome.
Paracelsus (b 1493) was the first person to publish about producing hydrogen. He did it by mixing iron with sulfuric acid. Sendivogius (b 1566) was the first person to publish about producing oxygen. He did it by heating potassium nitrate to the thermal decomposition (around 600°C). Lavoisier was the first person to publish about using an oxy-hydrogen torch (in 1782), which he used to melt platinum. Volta made his voltaic pile in 1800, and people immediately began using it to make oxy-hydrogen at scale by water electrolysis. In 1837, Gaudin was the first person to use an oxy-hydrogen torch to synthesize gemstones - he made rubies by fusing alum and chromium. Apparently, he was trying to make glass and didn't realize they were crystals. Frémy was making small rubies at commercial quality and scale by 1877, and Verneuil worked with him and after him to develop the Verneuil furnace for crystal growth, published in 1902, which still just ran on oxy-hydrogen. The Verneuil process is still basically what we use to grow synthetic corundum crystals, including rubies, sapphires, and emery. They're all aluminum oxide (alumina) with impurities. The same process can be used for growing crystals of rutile, spinel, strontium titanate, and possibly chrysoberyl (also in the spinel group). (Not emerald. I accidentally said emerald instead of emery on twitter. Emerald is a color of beryl, like aquamarine is a color of beryl, and beryl is a special silicate mineral (a cyclosilicate) that has beryllium and aluminum. More on beryls later.)
All of this is to say that I should be able to grow cool crystals with a torch and some refractory cement and a ceramic or graphite crucible to hold the melt. Some minerals are industrially grown in different ways than this, like cooking under pressure in an autoclave for months, but for others, you basically just torch the right ingredients with easily obtained gasses and then drop or pull the melt. And if I can make cool crystals, then of course I will. And I can! So I will.
The Verneuil furnace involves dropping bits of melted mineral into a pile in order to grow crystals. It might work better if you drop them onto a slowly rotating pile. You can also pull the melt, instead of dropping it. Pulling the melt is a different method of crystal synthesis: the Czochralski process. "He made this discovery by accident: instead of dipping his pen into his inkwell, he dipped it in molten tin, and drew a tin filament, which later proved to be a single crystal." Start with a bath of your molten mineral, get a tiny crystal for a seed, drop it in and pull it out slowly, possibly rotating the seed while you pull. Verneuil and Czochralski: sometimes you twist the thing.
In the standard practice of the Czochralski method, some minerals are pulled and twisted very, very slowly - too slowly to do by hand. Not a big impediment to my growing crystals: I think I'd just need a leadscrew linear actuator. In addition to tin crystals, the Czochralski process can be used to make crystals of corundum, spinel, chrysoberyl, garnet, lithium niobate, scheelite, fluorite, silicon, germanium, and apatite I think. Ooh, and bismuth telluride! So there are a few more options than Verneuil. Silicon and germanium crystals are important for electronic semiconductor manufacturing. Fluorite is worth making not least because it's fluorescent; it glows under UV radiation. Wonderful! I like fluorite a lot. Artificial garnets are useful for making lasers, in addition to being gemstones. Lithium niobate isn't even made by nature, but it's a cool mineral with weird properties that engineers (especially communications engineers) have put into lots of things. And bismuth telluride is a sort-of alloy of the metal bismuth and the metalloid tellurium. Bi2Te3 crystals are used as semiconductors in those ceramic plate thermoelectric modules that you ... probably haven't played with unless you build your own computers or you're an electrical engineer, but trust me, they're cool. The Czochralski process: it gets shit done.
Pure elemental metals are technically crystals. For example, native copper forms in either the isometric/cubic system or the octahedral system. But they're mostly not geometric and colorful and sometmes transparent, and those are the crystals I'm after. Not sure why that's less common for plain metal than for metal+sulfur or metal+carbonate, but that's the case. Oh well. There are also amorphous metal alloys or glassy metals. Kind of cool, but they're not for me. There's at least one metal that doesn't suck for making crystals: bismuth. Bismuth crystals are beautiful: geometric and colorful. And bismuth crystals are super easy to make. Take an ingot of bismuth, melt it in a pot on the stove, and slowly lift up the top layer when it starts to cool and solidify. Don't like the colors? Remelt and do it again. The color of a region on the crystal depends on the temperature of the melt as the region enters the air. Bismuth is in Pepto-Bismol, and you could do a chemical extraction, but the metal is pretty cheap compared to how awesome it is and if you want to play with bismuth crystals, it's probably a better use of your resources to just buy an ingot. For sulfides of metals, pyrite is pretty. It's iron sulfide in the cubic system. Iron sulfide also exists in the orthorhombic system, where it's called marcasite. So maybe I'll sulfidate iron some day. And there are other nice sulfides. Sphalerite is okay: ZnS. The metalloid arsenic has very pretty sulfides that are transparent and faceted and colorful: realgar and orpiment. But I don't have arsenic. Antimony sulfide is cool, but I don't have antimony. I'll just focus on iron sulfides first.
Now, these are good projects and I intend to do them. But let's take stock for a second. None of these are silicates! Not one of the minerals I've mentioned has silica (SiO2) groups or orthosilicate (SiO4) grroups or any of the many other silicon oxide groups. We have pure elemental silicon crystals from Czochralksi, but no silicon oxides! This is a huge glaring hole. Let me explain.
Edit: Wait! Am I completely wrong? You can make garnets with Czochralski! Aren't garnets silicates?! Let's take a moment to look at the chemical formulas for the minerals most famously synthesized by the Czochralski method. Corundum is an aluminum oxide, [Al2 O3], better known as alumina in chemistry and metallurgy when we're not considering the crystal system. Spinel and chrysoberyl are also oxides with aluminum: they're [Mg Al2 O4] and [Be Al2 O4] respectively. Garnets are, depending on usage, either a simple group of minerals that are all silicates or huge a super-group of minerals containing normal silicate garnets but also some non-silicates and some non-sense. They all have a common structure in their ratios of metal ions, basically. Here's the structure: minerals in the simple garnet group look like this: [X3 Y2 (SiO4)3]. They have three metal ions with valence +2 (divalent cations), and they have two metal ions with valence +3 (trivalent cations), and then they have three silicate anions. Almandine is a famous garnet, and it's [Fe3 Al2 (SiO4)3], with iron in the +2 oxidation state. In the garnet super-group, the Xs might not all be the same metal, the Ys might not all be the same metal, and the silicon could be anything else that will bind to four oxygens, and again, the silicon stand-ins don't all have to be the same element. *Throws hands up in frustration*. The Czochralski method is famous for making yttrium aluminum garnets, [Y3 Al2 (AlO4)3], and gadolinium gallium garnets, [Gd3 Ga2 (GaO4)3] as laser media. Obviously they're not silicates. I'm still not sure if Czochralski can grow normal silicate garnets, but I've heard that it can grow silicates of some kind, which is why I'm adding formulas into the post in case I missed something. Let's cover the rest of them just for safety: Lithium niobate is [Li Nb O3]. Scheelite is a tungstate, [Ca W O4]. Fluorite is a halide mineral, [Ca F2] (whose name comes from halogen, whose name comes from the pseudo-greek for "salt-forming"). Bismuth telluride is still [Bi2 Te3]. Apatite is a phosphate mineral group with just three members, and every member of the group has calcium phosphate with a little ion hanging off the end, either a fluorine, a chlorine, or a hydroxide (OH). So I haven't lied yet! But more research is necessary.
Post-edit segue: Why is it a glaring hole that I can't make silicates?
Silicon and oxygen are the most abundant elements in the earth's crust, and they're also the main ingredients in the most abundant minerals in the earth's crust: quartz and feldspars. Quartz is silica in either a trigonal or a hexagonal crystal system. Feldspars are alumino-silicates with small bits of other stuff, arranged in the monoclinic crystal system or in the triclinic crystal system. If I can't make those in my back yard, I can't make much. Can I though? And also can I make clays? Clay minerals are super common, and they're also silicates: clays are hydrated aluminosilicate minerals that arrange in microscopic sheets. I'm not sure I want to make clays; they're not exactly gemstones, but I would be sad if I couldn't.
Let's talk about how quartz is made industrially: the hydrothermal process. Quartz crystals are made under moderately high heat (at one end of the container, so that the crystal is under a thermal gradient) and high pressure in autoclaves by the hydrothermal process. It takes months. It's slow, expensive, kind of dangerous, and kind of ridiculous when you consider that natural quartz is absolutely everywhere. But sometimes you just need a really pure single quartz crystal the size of a spoiled cat. I'm not here to judge. I'm here to make crystals. These are my people. I don't have an autoclave. I have a pressure cooker, but I like to not experiment with it. I'm not scared of high pressures, and I'm not scared of high temperatures, but I've never dealt with both at the same time, you see. Maybe in a decade I'll have forty acres where I can blow stuff up in the pursuit of worthless quartz crystals, but not today. Let's keep learning about how other gems are synthesized and we'll come back to quartz and feldspar later.
Beryls are also grown in autoclaves by the hydrothermal process. Emeralds and aquamarine are colors of beryl, and they're all aluminosilicates, just like feldspars and clays. Also beryls can be grown by the flux-growth method. I don't know enough about that, but I've read that it's also expensive and it also takes months or years to make small crystals. Sad. I've read that General Electric once made jadeite-jade under pressure, presumably by the hydrothermal method. Jadite is an aluminosilicate with a little sodium. GE doesn't make jadeite crystals anymore, but it's nice to know that it can be done. I don't know of any other minerals that people have synthesized which also need high pressures, but there are probably others. Other silicates presumably. I really like olivine, which is a silicate with some variable amount of magnesium and iron. Maybe people grow it in autoclaves. Who knows. Someone knows. Not me.
Instead of looking at synthesis methods, I've also just looked up individual gemstones to see if and how they can be synthesized, with mostly negative results. A french guy named Gilson claims to have made synthetic turquoise and lapis lazuli (and opals, which are amorphous/glass/mineraloids rather than minerals), but almost everyone except Gilson agrees that his synthetics are actually imitations, and contain at least some of the wrong stuff. It doesn't really matter either way to me, because the process isn't something I can replicate: it's a secret, bought and owned by some Japanese manufacturer. Another negative result: people on the internet seem very adamant that humans have never produced synthetic tourmaline. Okay. I believe you. Calm down. Tourmaline is a group of silicates with lots of boron, and I have never seen two people give the same formula that generalizes the group, because like garnets, there's a core group that's simple and then a wider group of rare minerals that don't matter at all in practice, and you'll need a different formula to cover your particular choice of included rare minerals. But it's a silicate, and it hasn't been grown. Sad.
I mentioned that apatite can be made by the Czochralski method. It's a calcium phosphate mineral, and it's the main mineral in bones and tooth enamel. Natural apatite crystals are very pretty to me, and they're sometimes striped with colors like a candy cane. We know that the body can make tiny apatite crystals without heating up to furnace or kiln temperatures, and I find that encouraging. Maybe there's an even easier way for me to make it than Czochralski. What other minerals can bodies make? Similar to apatite, there's another calcium phosphate mineral called brushite and it's part of kidney stones. You're probably more familiar with calcium oxalate as a component of kindey stones, which goes by the name Whewellite as a mineral. If bodies can make these things, that seems like a strong hint that I can make them too. So I will! One day! Somehow! With my hands and my ingenuity I mean, not with my kidneys. The general term for these processes is bio-mineralization and I don't know much about it. Wikipedia says some algae and diatoms make silicates! Relevant to my interests! Also, goethite is listed on the page. That's another fun ore, an iron hydroxide in the orthorhombic system. One more to do!
Okay, back to quartz. A guy claims to have made it at standard atmospheric conditions back in the 1970s. Is he a dirty liar? I don't know. Let's talk about it.
The processes that turn sediment into sedimentary rocks are called diagenesis (= across-generations). They're the opposite of weathering. Diagenesis includes stuff that follows deposition/sedimentation, like compaction, deformation, dissolution, hydration, kerogen cracking, post-depositional mobilization (stuff moving after it stopped moving), chemical replacements and alterations including dolomitization (magnesium ions replace calcium ions), cementation, and lithification. The usual way to synthesize quartz is not diagenetic - it doesn't happen at atmospheric conditions. Rather, it's hydrothermal or magmatic.
Werner Flehmig said he made feldspar and quartz diagenetically back in the 70s (here and here) and Harding said he made clay minerals using basically the same diagenetic process (here and here). The method, I think just from looking at the abstracts, is to take a hydroxide of the right metal, like brucite [Mg(OH)2] (milk of magnesia) or gibbsite [Al(OH)3], make an aqueous solution, add in some amorphous silica, but very little or it doesn't work, and then you let it sit and age and nucleate and adsorb for three weeks. I think that's it. Also, they mention iron(III) hydroxide several times. Iron(III) hydroxide is basically limonite, a common ore used as a pigment since antiquity, also called yellow ochre.
I don't get it. These crystals must be microscopic, if no one figured out by accident before the 1970s that any of these hydroxides can be used to make quartz diagenetically. And I'm not sure I care about microscopic crystals. They don't make for a good display case. Or maybe the crystals made by this method are not microscopic. After all, no one figured out until 1915 that you can make large single tin crystals by dipping a pen nib into molten tin. Paracelsus didn't know about electrolysis. Sometimes people just don't know easy ways of making cool things. And if algae and diatoms can make silicate crystals, then what's so strange about Flehmig and Harding doing it? I'll keep reading. Maybe it'll be another case like Gilson. Maybe I'll actually learn to make felsic crystals. Maybe I can just be happy sitting on a mountain of rubies and garnets and halite and kidney stones and rainbow bismuth. Maybe I'll get my forty acres and set up a lab. Maybe I'll figure out a way to devitrify glass intro cristobalite silica and then polymorph it into quartz at low pressures.
Happy synthesizing!
