Some compounds decompose before they melt. Distillation can also be more efficient at low pressure, where the liquid doesn't exist. So here's a sublimation distillator - on the paper at least.

 

A rotating cylinder with compartments lets the powder fall through the vapour, aided by paddles or buckets, as preferred to a shaker or a gas jet. Heating at one end and cooling at the other establishes a temperature gradient. The rotation axis is adjustably tilted over the horizontal, so the powder overflows through holes from a compartment to the next warmer one. The holes are small enough that the vapour flows gently towards the colder compartments, and the shape of their edge, aided by the shape of the paddles or buckets, can prevent the falling powder from reaching unduly the neighbour compartments.

Condensation seeds, big enough and nonvolatile, can regulate the powder size.

Marc Schaefer, aka Enthalpy

I could dig out a book I have called something like "coal tar and related industry" which includes a section on the purification of ammonium chloride by sublimation. The book is about a hundred years old.

The least volatile compound may deposit as a tight layer on the powder grains, preventing the evaporation of the more volatile compound. A parry is to heat the cylinder's end enough to evaporate both compounds.

I could dig out a book I have called something like "coal tar and related industry" which includes a section on the purification of ammonium chloride by sublimation. The book is about a hundred years old.

 

Thanks JC!