Endothermic versus exothermic transport

Going deeper

The reaction runs both ways on purpose

Chemical vapour transport works by turning an involatile solid into a gas at one temperature and turning it back into a solid at another. A transport agent – iodine, chlorine, a metal halide – reacts with the charge to form volatile species; those species diffuse or convect down the tube; where conditions favour the reverse reaction they decompose, and the solid reappears as a crystal.

What makes this a growth method rather than a curiosity is that the same reaction runs in both directions at different temperatures. The equilibrium constant of the transporting reaction depends on temperature through its enthalpy, so the gas that is in equilibrium with the solid at one end is supersaturated at the other. The material moves down that gradient in vapour phase, which is why crystals grown this way can be large and well-formed without ever melting anything.

Which way the material moves

If the transporting reaction is endothermic – it takes heat in to make the gas – then raising the temperature makes more gas. The hot end therefore dissolves the charge into the vapour, and the cool end, where the equilibrium favours the solid, is where crystals grow. If the reaction is exothermic, everything reverses: gas forms preferentially at the cool end and deposits at the hot end.

This is not a detail to be discovered experimentally. It decides where the charge is loaded, which end of the furnace runs hotter, and where the seed or the nucleation zone should be. A tube loaded on the wrong assumption comes out after two weeks looking exactly as it went in. The enthalpy of the transport reaction can be estimated from tabulated thermodynamic data before anything is sealed, and for the classical systems it has been tabulated for decades – the method has been studied since mineral-forming reactions were recognised in the mid-nineteenth century, and was made quantitative by Schäfer’s work.

Choosing conditions in practice

The magnitude of the enthalpy matters as much as its sign. A reaction with a very small enthalpy barely responds to the gradient, so almost nothing is transported. One with a very large enthalpy drives everything to one end immediately and nucleates a mass of small crystals instead of a few large ones. The useful window is intermediate, and it is tuned with the agent, its quantity, the mean temperature and the size of the gradient.

For layered materials this is the standard route to research-grade single crystals – the ones that get exfoliated afterwards – and the choices leave fingerprints. Iodine, chlorine and tellurium chlorides transport different systems with different efficiency, and residual agent is incorporated as an unintentional dopant, which is one reason crystals of nominally the same compound from different groups have different carrier concentrations. Reporting the agent, its concentration, the temperatures at both ends and the duration is the minimum for a result anyone can reproduce.