Carbon allotropes

Why do diamond and pencil lead behave so differently when both are nothing but carbon?

Diamond scratches every other natural mineral and conducts no electricity. Pencil graphite smears off onto paper under the weight of your hand and conducts perfectly well. A chemical analysis of both returns the same single answer: carbon. If the atoms are identical, what is left to differ — and can arrangement alone really account for a gap that large?

Take the question seriously rather than accepting "structure" as a slogan. What can an atom actually choose? Carbon has four outer electrons, and it can commit them in different geometries. In one arrangement — call it the tetrahedral one — all four go into equivalent bonds pointing to the corners of a tetrahedron, at about 109 degrees to each other. In another, three go into bonds lying flat in a plane at 120 degrees, and the fourth is left in a lobe standing perpendicular to that plane.

Follow the first choice through the whole crystal. Every atom bonded tetrahedrally to four neighbours, each of those to four more: there is no way to stop, so the entire stone is one continuous covalent network in three dimensions. Now ask what it would take to deform such a thing. Every path through the material runs along strong bonds, so there is no cheap direction — you cannot slide one part past another without breaking bonds outright. That is diamond, and its hardness is not a separate fact from its geometry but a restatement of it. All four electrons are locked into localised bonds, none free to wander, so it is an electrical insulator — yet an exceptional conductor of heat, because a stiff, light, near-perfect lattice carries vibrations superbly.

Now follow the second choice. Three in-plane bonds build a flat honeycomb of hexagons, and the sheet is enormously strong within its own plane — the in-plane bond is about 0.142 nanometres, shorter and stronger than diamond's. But the sheet has nothing structural above or below it. The fourth electrons, one per atom, merge into a cloud spread over the whole sheet, which is why graphite conducts electricity along its layers and not nearly as well across them. Successive sheets sit about 0.335 nanometres apart, held only by dispersion forces. Same atoms; a material with two utterly different directions built into it.

Here is where the usual story stops, and it stops slightly too soon. The standard explanation says graphite is slippery because those weak interlayer forces let sheets slide. That is incomplete, and there is a clean piece of evidence against it. In the 1940s, carbon brushes in aircraft generators wore out catastrophically at high altitude — the phenomenon was called dusting. Graphite in dry vacuum is not a lubricant at all; it abrades. Its slipperiness depends on adsorbed species, water vapour chiefly, sitting on the sheet surfaces and passivating the edges. Take the air away and the layers grip. Molybdenum disulfide, whose lubricity does not depend on adsorbates in the same way, is what actually gets used in space mechanisms.

One more question worth asking: if the two structures are so different, why does a diamond not quietly turn into graphite? At ordinary pressure graphite is in fact the thermodynamically stable form — diamond is metastable. It survives because getting from one to the other requires rebuilding every bond in the crystal at once, and the barrier is enormous. Diamonds are not forever on principle; they are forever on kinetics.

"Same atoms, different arrangement" is the right headline but hides how much is doing the work. It is not simply that the atoms are packed differently in space, like two stackings of identical balls. The bonding is different: how many neighbours each atom commits electrons to, at what angles, and how many electrons are left over to roam. Density follows from it too — diamond about 3.51 grams per cubic centimetre against graphite's 2.27 — but density is a consequence, not the cause.

It is also worth saying that these two are not the whole family. Buckminsterfullerene, nanotubes and single-sheet graphene are all carbon under different bonding topologies, and the sooty carbon in a pencil-shaded page or a candle flame is largely disordered rather than either crystal.

And a note on the honest limit of the account above: the mechanism of graphite lubrication is better understood than it was, but it is not a solved textbook item. Adsorbate dependence is well established experimentally; the details — how much comes from passivating dangling bonds at sheet edges, how much from the near-frictionless sliding of mismatched lattices seen in clean nanoscale contacts — are still argued over.

• Why graphene, a single graphite sheet, behaves so differently from the stack it came from.
• What conditions actually convert graphite into diamond, and how synthetic diamond is made.

Every term the collection defines is gathered in the glossary.