One picture, a whole zoo of crystals
Four guides have quietly built one picture. In a ceramic the big anions pack together like stacked marbles, and the small cations drop into the gaps between them — so once you fix the anion stacking (cubic FCC or hexagonal HCP) and say which holes the cations take, you have named the crystal. Sorted by how many anions ride each cation — the stoichiometry — the canonical structures form a short list: the AX rock salts and their cousins, the AX2 fluorites and rutiles, the sesquioxide corundum of Al2O3, and the mixed-cation perovskite and spinel.
What every one of them shares is that it is nothing more than a specific filling of the octahedral and tetrahedral holes in that anion array — MgO fills every octahedral hole, zinc-blende fills half the tetrahedral ones, spinel splits its cations between both. This final guide adds no new structure. It cashes those pictures in for two things: a number you can put on a balance — the theoretical density — and a warning that the same atoms can re-stack into a completely different crystal, which we call polymorphism.
Weighing a unit cell
A crystal is just its unit cell copied endlessly in every direction, so the density of the whole is exactly the density of one cell: the mass packed inside it divided by its volume. Both pieces are things you already know how to get. The mass is the mass of the atoms the cell contains; the volume is the cell's edge cubed, a^3, for a cubic structure. Nothing here needs a laboratory — only the structure you have learned to read, plus a periodic table.
Written out, the theoretical density is rho = n x M / (N_A x V). Here n is the number of formula units in one cell — read straight off the structure, so rock salt has n = 4; M is the formula weight in grams per mole, from the periodic table; N_A is Avogadro's number, 6.022 x 10^23 per mole; and V is the cell volume. Every symbol is either counted from the crystal or looked up, which is why this single line is the first honest bridge from an invisible atomic arrangement to a property you can actually measure.
A worked example: the density of MgO
- Read the structure. MgO is rock salt: an FCC array of O2- with an Mg2+ in every octahedral hole. Counting the cell contents gives 4 O2- and 4 Mg2+, that is 4 formula units, so n = 4.
- Get the formula weight. M(MgO) = 24.31 (Mg) + 16.00 (O) = 40.31 g/mol.
- Take the cell edge. Measured (or from the structure) a = 0.4212 nm = 4.212 x 10^-8 cm, so V = a^3 = 7.47 x 10^-23 cm^3.
- Assemble. rho = n x M / (N_A x V) = (4 x 40.31) / (6.022 x 10^23 x 7.47 x 10^-23) = 161.2 / 45.0 = 3.58 g/cm^3.
- Sanity-check. A handbook lists MgO at 3.58 g/cm^3 — dead on. The structure, read correctly, predicted a property you can put on a balance.
The same recipe runs on every structure in the rung — only n and the cell geometry change. Swap in the fluorite cell of ZrO2, or the rutile cell of TiO2, and you predict their densities too. This is not a party trick: it is exactly how you tell one polymorph from another. Because different structures pack the same atoms with different tightness, each one carries its own theoretical density, so a density measurement becomes a fingerprint for which crystal you actually made — and that is the door into the second half of this guide.
Same atoms, a different crystal
Here is the twist that unsettles beginners: a composition does not own a single structure. Many ceramics are polymorphic — the identical formula adopts different crystal structures in different windows of temperature and pressure. Polymorphism is why carbon can be both soft graphite and hard diamond, and why the three headline ceramics of this rung — silica, zirconia, and titania — each come in several distinct crystals rather than one.
The logic behind the map is simple. A denser, higher-coordination packing is favoured by high pressure, which squeezes atoms together; a more open, higher-entropy arrangement is favoured by high temperature, which rewards floppiness. So each polymorph owns a home on a pressure-temperature map — a one-component phase diagram — and crossing a boundary tips the atoms from one stacking into another.
SILICA (SiO2): one composition, many crystals
--------------------------------------------------------
raise TEMPERATURE (1 atm) --> more open, LESS dense
Si stays 4-coordinate
alpha-quartz ~2.65 g/cm^3 (alpha->beta at 573 C)
tridymite ~2.26 g/cm^3 (above ~867 C)
cristobalite ~2.32 g/cm^3 (above ~1470 C)
liquid -> glass ~2.20 g/cm^3 (melt ~1713 C; frozen on cooling)
raise PRESSURE --> tighter packing, MORE dense
coesite ~2.91 g/cm^3 (Si still 4-coordinate)
stishovite ~4.29 g/cm^3 (Si now 6-coordinate: rutile type!)Read the ladder and both halves of this guide lock together. As silica heats at ordinary pressure it steps from quartz to tridymite to cristobalite, each a more open framework of the same corner-sharing SiO4 tetrahedra, so the density falls as the temperature climbs. Squeeze it instead and you reach coesite, then stishovite — and in stishovite silicon finally abandons its lifelong tetrahedral habit and sits in a six-fold octahedron, the rutile structure, which is why stishovite is nearly twice as dense. Same formula throughout; the density is simply announcing which stacking you have got.
Zirconia: the change that breaks and saves
No polymorphism matters more in engineering than zirconia's. Pure ZrO2 steps monoclinic at room temperature, to tetragonal near 1170 degrees C, to cubic near 2370 degrees C, and melts near 2715 degrees C. The catch lives on the way back down: the tetragonal-to-monoclinic change on cooling comes with a roughly 3 to 5% volume expansion, and that sudden swelling shatters a dense pure-zirconia part into rubble. Left to itself, pure zirconia is useless as a fired ceramic — the zirconia system simply will not hold together through cooling.
Two moves turn the curse into a gift. Add a stabiliser such as yttria and the cubic or tetragonal form survives, metastable, all the way to room temperature — yttria-stabilized zirconia. Better still, keep the tetragonal grains only just metastable, poised on the edge: a passing crack's stress tips them into the monoclinic form, and the expansion that accompanies the flip clamps down on the crack and squeezes it shut. That is transformation toughening — an airbag that deploys in the path of a crack — and it is what lifts zirconia from brittle curiosity to one of the toughest ceramics we have.
The through-line to phase change
Notice how the two halves are really one idea. Because each polymorph packs its atoms differently, each carries its own density — so density fingerprints the structure, telling loose anatase from dense rutile in titania, or metastable tetragonal from stable monoclinic in zirconia. And because switching polymorph changes the density, it changes the volume, and a volume change forced through a rigid solid means stress, warping, and cracking. That single fact — a polymorphic transformation is a volume change — is the engine behind quartz inversion cracking pottery at 573 degrees C, behind pure zirconia's self-destruction, and behind the phase transformations the next rungs are built on.
Carry three honest caveats up the ladder. Theoretical density is a pore-free ideal, so real bodies always fall short by their porosity — sometimes porosity is even wanted, in filters and bone scaffolds. The radius-ratio scheme that named these structures is a guide, not a law, so strong directional bonds bend it. And equilibrium diagrams describe stability, not what firing freezes in — window glass, anatase, and stabilized tetragonal zirconia all persist metastably for the same reason. Underneath it all runs one liberating idea: composition is not destiny. The same atoms, re-stacked, become a new crystal with a new density and new properties — and that lever is exactly what the rest of this ladder learns to pull.