Solid-State Reactions and Reaction Kinetics

From Moving Atoms to Making New Compounds

The last four guides taught you the machinery of motion: Fick's laws for the flux and the profile, the coupled march of cations and anions, vacancy hopping, and the Arrhenius climb of the diffusion coefficient with temperature. All of that was preparation. Now we put diffusion to work, because moving atoms are how a ceramist actually makes things. A solid-state reaction takes two solid powders, presses them together, heats them below their melting points, and lets the atoms crawl across the contact until a genuinely new compound grows in between — no liquid, no vapour, just diffusion doing chemistry.

This is the mixed-oxide route, the workhorse behind most ceramic powders on Earth — the barium titanate in a capacitor, the ferrite in a magnet, the YSZ in a fuel cell all begin this way. The recipe sounds almost too simple: weigh out the oxides (or carbonates) in the right ratio, mix them, and fire. Picture two stacked slabs of coloured clay left in a kiln: at room temperature nothing happens for a million years, but heat them until the atoms unfreeze and a fuzzy new band slowly grows along the seam where they touch. The whole art of this guide is understanding how fast that band grows, and why.

Here is the honest catch that shapes everything below. Unlike stirring two solutions, where dissolved atoms meet freely and react in seconds, in a solid every atom is locked in its cage. Reaction happens only where two particles actually touch, and only as fast as diffusion can carry ions across the growing product. That makes solid-state reactions slow, thirsty for high temperature and long time, and — a point beginners always underestimate — almost never truly complete in a single firing. Understanding that limitation is what separates guessing from engineering.

The Classic Case: Spinel Grows Between MgO and Al2O3

The textbook example, studied for a century, is spinel: MgO + Al2O3 -> MgAl2O4, the same spinel structure you met in the structures rung. Press a grain of magnesia against a grain of alumina, hold them at 1400 to 1600 degrees C, and a layer of spinel appears at the contact. But the moment that layer exists, it walls the two reactants apart. For the reaction to go any further, ions can no longer simply meet at the interface — they must diffuse all the way through the spinel that already formed. The product becomes its own barrier.

The mechanism is beautiful, and it leans directly on guide 4. The oxygen ions are big and close-packed; they form a rigid scaffold that barely moves. It is the small cations that shuttle. To keep the crystal charge-neutral, Mg2+ and Al3+ must counter-diffuse: for every 3 Mg2+ that migrate one way, 2 Al3+ migrate the other, because 3 times (+2) exactly balances 2 times (+3). This is ambipolar diffusion wearing a new hat — two cations coupled by charge neutrality, so the slower one is the rate-controlling species that sets the whole pace.

MIXED-OXIDE ROUTE: spinel grows BETWEEN the two reactants
(the oxygen framework stays put -- only the cations shuttle through it)

   MgO   |          MgAl2O4  (spinel)          |   Al2O3
         | <----------  2 Al3+  ---------------|
         |------------  3 Mg2+  ----------->   |
   ------+-------------------------------------+------
     left interface                       right interface
     2 Al3+ + 4 MgO ->                     3 Mg2+ + 4 Al2O3 ->
          MgAl2O4 + 3 Mg2+                      3 MgAl2O4 + 2 Al3+

   net:   MgO + Al2O3 -> MgAl2O4
   charge stays balanced:  3 Mg2+ (=6+)  counter  2 Al3+ (=6+)
   3 units of spinel form on the Al2O3 side, 1 on the MgO side
     -> the layer thickens ~3x faster into the alumina
   every ion must cross the WHOLE product layer -> path grows -> slows

Trace the sketch and the whole reaction snaps into place. At the left interface, arriving Al3+ react with MgO to lay down fresh spinel and release Mg2+; at the right interface, arriving Mg2+ react with Al2O3 to lay down spinel and release Al3+. Add the two and the leftover ions cancel, leaving the clean net reaction. Because three formula units of spinel are built on the alumina side for every one on the magnesia side, the layer visibly grows faster into the Al2O3 — a real, measurable asymmetry that first told scientists the cations were doing all the travelling.

Why the Layer Thickens as the Square Root of Time

Now the kinetics, and it follows from Fick's first law with almost no algebra. The flux across the product is proportional to the concentration gradient, and the gradient is just the fixed difference in ion concentration across the layer divided by its thickness x. So as the layer grows, the driving gradient falls as 1/x: the thicker the wall, the gentler the slope, the slower the crossing. Since the growth rate dx/dt is proportional to that flux, dx/dt is proportional to 1/x. Rearrange to x dx = (constant) dt and integrate, and you land on the famous parabolic rate law: x^2 = k times t.

That square is the entire story, so make it concrete. Suppose a mixed powder grows a 5 micron spinel layer after 1 hour at 1500 degrees C. Then k = x^2/t = 25 micron^2 per hour. To reach 10 micron you need t = 10^2/25 = 4 hours; to reach 20 micron, t = 20^2/25 = 16 hours. Every doubling of thickness costs four times the hold. The reaction throttles itself: the layer it builds is the wall it must then diffuse through. And k itself is Arrhenius, k = k0 times exp(-Q/RT), so it climbs steeply with temperature — the same exponential you met in guide 3 — which is why a hundred degrees hotter can shrink a day-long firing to an hour. (The numbers here are illustrative, but the sqrt-time shape is real.)

One honest refinement: the parabolic law rules only once the layer is thick enough that diffusion is the bottleneck. At the very start, when the product is just a few atoms thick, ions cross it instantly and the slow step is instead the chemistry of attaching them at the interface. That regime is phase-boundary-controlled, and it grows linearly, x proportional to t, not as a square root. Real reactions often begin linear and then bend over into parabolic as the layer thickens — so if your early data look like a straight line, you have not broken physics, you have just caught the reaction before diffusion took command.

Calcination: Reactions that Breathe Out a Gas

Not every solid-state reaction happens between two oxides. Many ceramic recipes start from a carbonate or a hydroxide, and the first firing simply drives off a gas. This is calcination, the most common flavour of thermal decomposition: one solid becomes another solid plus a gas that escapes. Limestone is the everyday case: CaCO3 -> CaO + CO2, which runs near 900 degrees C and is how quicklime and cement clinker begin. Magnesium hydroxide behaves the same way, Mg(OH)2 -> MgO + H2O. The kinetics differ from the spinel case because there is no thickening product layer to cross — instead the gas must find its way out, and the fresh oxide nucleates as it goes.

Why bother starting from a carbonate at all, instead of buying the oxide? Because decomposition is a gift: as the CO2 tears out of the crystal it leaves behind a very fine, highly porous oxide riddled with fresh surface, far more reactive for the next step than a coarse, dead-burned powder. That reactive powder then feeds straight into a mixed-oxide reaction. Be honest about the flip side, though: fresh calcined oxides like CaO and MgO are so eager that they grab CO2 and water back out of the air, so they must be kept dry and used promptly. Reactive is a double-edged word.

Turning the Knobs: Making the Reaction Go

Everything above hands you four levers, and each traces straight back to an earlier guide. Raise the temperature and both the diffusion coefficient and the rate constant k climb exponentially (Arrhenius, guide 3). Shrink the powder and you win three ways at once: shorter distances for ions to cross, far more contact area between particles, and heavier reliance on the fast grain-boundary and surface short-circuits that dominate fine-grained bodies (guide 4). Mix more intimately so every magnesia grain actually touches an alumina grain. And, more subtly, dope to raise the very defect population that carries the flux — more vacancies mean more hops per second.

1. Weigh the oxides or carbonates to the target stoichiometry — the ratio you want in the final compound, since solid-state reactions do not fix a bad recipe.
2. Wet-mill the mixture: milling both blends the powders intimately and grinds them finer, cutting the diffusion distance and multiplying contact area.
3. Dry, then calcine at moderate temperature to decompose any carbonates and let the first reaction begin in a fresh, reactive powder.
4. Grind the calcined cake to shatter product shells and re-expose unreacted cores — resetting the parabolic clock — then fire again; repeat until conversion is high enough.
5. Check phase purity, typically by X-ray diffraction, to see whether any starting oxide or a metastable intermediate is still hiding in the product.