Mix, Calcine, Mill: The Oldest Recipe
Guide 1 made the case that in ceramics the powder decides everything — a part is only ever as good as the powder it is built from. So how do you actually make one? The oldest answer, and still the one behind most of the ceramics in the world, is the mixed-oxide route: take the plain oxides or carbonates off the shelf, mix them, fire them until they react into the compound you want, and grind the result back to a powder. Three verbs sum the whole thing up — mix, calcine, mill — and it reads more like a cook's recipe than a chemist's synthesis.
Why does this plodding recipe still dominate industry? Because it is cheap, forgiving, and scales from a lab crucible to tonnes without drama. The barium titanate in your phone's capacitors, the ferrites in its antennas, the bulk of the alumina and zirconia parts made each year — most of it starts here. The catch, which the last three sections make painfully concrete, is that the very simplicity that makes the route robust also leaves its powder coarse, uneven, and prone to hard lumps. It is the workhorse precisely because it is good enough, not because it is good.
- Weigh each oxide or carbonate to the target stoichiometry, correcting for assay and any absorbed moisture.
- Wet-mill the batch in water or alcohol with grinding media — this both mixes the components intimately and breaks up loose clumps.
- Dry the slurry back into a powder cake.
- Calcine: fire below the final sintering temperature to decompose the carbonates and react the oxides into the phase you want.
- Mill the calcined cake back into a fine, free-flowing powder — and watch for contamination creeping in from the media.
- Characterize the powder — particle size, the full distribution, surface area, morphology — before you dare to form it.
Weigh and Mix: Getting the Atoms Close
Batching is arithmetic. Say you want barium titanate, BaTiO3, the backbone of the capacitor world. The reaction is BaCO3 + TiO2 -> BaTiO3 + CO2, so you weigh one mole of barium carbonate (molar mass about 197.3 g) against one mole of titania (about 79.9 g) — a 197.3-to-79.9 ratio by mass. Get it wrong by even a percent and you pay: a little excess TiO2 leaves a second phase at the grain boundaries, a little excess Ba leaves unreacted carbonate. Because the carbonate holds moisture and is never quite 100 percent pure, real batching corrects for assay — which is why the recipe starts by trusting a good balance over a good intention.
Then comes mixing, and here the last rung pays a dividend. Calcination is a solid-state reaction, and you already know from the diffusion rung that atoms crawl only a few microns through a hot crystal — the depth goes as sqrt(D times t). So the reaction can only happen where a barium grain actually touches a titania grain; anywhere the mixing left a lonely clump, the phase never forms. That is why the batch is not merely stirred but wet-milled in a liquid with media: the slurry carries particles past one another far more thoroughly than dry tumbling, coating every titania grain in barium neighbours so the diffusion distances stay short.
Calcine: Firing the Powder into the Right Phase
Calcination is the firing that does two jobs at once, both below the temperature you will later sinter at. First it drives off gas: heating a carbonate or hydroxide until it decomposes — a thermal decomposition — so BaCO3 sheds carbon dioxide and becomes BaO on the spot. For the BaTiO3 batch this is a hefty loss: 44 grams of CO2 leave for every mole made, about 16 percent of the starting mass gone up the flue. Second, the freed oxide reacts with its neighbour by solid-state diffusion, BaO plus TiO2 knitting into BaTiO3. Decompose, then react — that is the whole of calcination.
CALCINATION: a Goldilocks window (BaCO3 + TiO2 -> BaTiO3 + CO2) stage what happens the point ----- ------------ --------- ramp up ~5 deg C/min toward the hold near 600 C BaCO3 decomposes, CO2 leaves (~16%) thermal decomposition HOLD ~1150 C BaO + TiO2 react into BaTiO3 finish the phase (about 2 h) (kept BELOW the sintering temperature) cool take the still-loose powder out too COOL / too SHORT -> reaction unfinished (leftover BaCO3, TiO2) too HOT / too LONG -> powder sinters to itself -> HARD aggregates (ruined)
That schedule hides the route's central trap. Calcine too cool or too briefly and the reaction stalls: you pull out a cake still speckled with leftover BaCO3, TiO2, and half-baked intermediates such as Ba2TiO4. Push hotter or longer to finish the job and you cross an invisible line — the fresh powder starts to sinter to itself, particles welding at their contact necks into coarse, rock-hard lumps. Those are hard aggregates, and they are exactly what you do not want, because no amount of later milling will break them. Calcination is therefore a balancing act: hot enough to complete the phase, cool enough to leave the powder loose.
Mill: Break It Down, and Pay the Price
The cake that comes out of the calciner is coarse and clumped, so the last step is to mill it back down. Milling tumbles the powder with hard media — balls of alumina or zirconia — that fracture the particles on impact. A simple ball mill, a jar rolling for hours or days, grinds most oxides down toward a micron; an attrition mill, which stirs the media hard with a paddle, pours in far more energy and reaches well below a micron much faster. Milling does three good things at once: it shrinks the particle size, it narrows and homogenizes the size distribution, and it tears apart the loose clumps.
But milling always sends you a bill: grinding-media contamination. Every impact that chips the powder also wears a sliver off the media, so the batch slowly eats its own grinding balls. Zirconia media wear least, alumina media dose the batch with a little Al2O3, and old-fashioned steel adds iron oxide — typically anywhere from a few tenths of a weight percent up to a few percent over a long grind. For an advanced ceramic meant to be high-purity, that stray oxide is an unplanned dopant that can shift colour, conductivity, or sintering. There is also a floor: below roughly a tenth of a micron, particles stop breaking and simply deform and re-weld, so mechanical milling cannot cleanly reach the nanoscale — a real grinding limit.
What the Solid-State Route Can and Cannot Give
Lay the ledger out honestly. In the plus column, the mixed-oxide route is cheap, robust, endlessly scalable, and works for almost any oxide — reasons enough to keep it the industrial default. In the minus column, its powder comes out coarse (microns, not nanometres), with a broad size distribution, chemically uneven at the finest scale because mixing is only ever as good as the powders you started with, and forever threatened by hard aggregates from the calciner. Recall the maxim of this rung — a good powder is fine, pure, uniform, and free of hard aggregates — and you can see the solid-state route strains against all four. Coarse, clumped powder is hard to densify, so the part you fire from it fights you too.
When those limits start to hurt — when you need a truly fine, pure, uniform powder — the fix is to stop mixing solids and start mixing in solution, so the components blend not grain-against-grain but ion-against-ion. That is the family of chemical, or wet, routes: the sol-gel process, coprecipitation, hydrothermal synthesis, and the more exotic spray pyrolysis and combustion synthesis. Because the atoms are already intimate, the phase forms at a far lower temperature and the powder emerges fine and often nano-scale — the whole subject of guide 3. They cost more and can be finicky, which is exactly why the plodding solid-state route still refuses to die.
Whichever route you pick, you cannot manage what you do not measure — so before forming you weigh the powder up with numbers, the business of guide 4. One shortcut ties it together nicely: for equal-sized spheres the BET surface area and the particle size are two views of the same thing, d = 6/(rho times S), with d in microns when rho is in g per cm^3 and S in m^2 per g. A coarse solid-state alumina (rho about 3.98) with S near 1 m^2/g implies particles about 1.5 microns across; a good sub-micron powder with S near 10 m^2/g implies about 0.15 micron. One number, measured in an afternoon, tells you at a glance which kind of powder your route just made — and whether it stands any chance of sintering dense.