a high-entropy ceramic
Most ceramics are built from one or two kinds of metal cation sitting on their lattice sites — magnesium in MgO, or barium and titanium in BaTiO3. A high-entropy ceramic throws that habit out. It blends five or more different metal cations in roughly equal shares and dares them all to share the same site of a single crystal structure at random, the way five colours of sprinkles mix evenly on one cake. The idea is borrowed straight from the high-entropy alloys that shook up metallurgy: the sheer disorder of the mixing — its high configurational entropy — helps hold the single phase together instead of letting it separate into simpler compounds.
The engine is a term from thermodynamics. The configurational entropy of random mixing is ΔS_config = -R times the sum of x_i times ln(x_i), where x_i is the fraction of each cation and R is the gas constant; it is largest when all fractions are equal, reaching R times ln(5), about 1.61 R, for five equal cations. The first clean demonstration was an entropy-stabilised oxide, (Mg,Co,Ni,Cu,Zn)O, which takes the rock-salt structure even though some of those oxides would rather adopt other structures on their own (Rost and colleagues, 2015). The family now spans high-entropy oxides, borides and carbides (some of them ultra-high-temperature ceramics), fluorites and perovskites. Cramming five different-sized cations onto one site distorts the lattice and slows diffusion, which tends to lower thermal conductivity and can raise hardness.
High-entropy ceramics matter because they open an almost unlimited compositional playground and, with it, tunable properties — low thermal conductivity for thermal-barrier coatings, promising catalysts, and new battery electrodes — while offering the UHTC community fresh borides and carbides. Be honest about two things. First, 'entropy-stabilised' is contested: many of these phases are actually enthalpy-favourable too, so entropy is not always doing the work the name implies, and some will decompose into separate phases if cooled slowly. Second, the composition space is so vast (billions of possible five-cation recipes) that you cannot explore it by trial and error — which is exactly why this frontier leans so heavily on computation and machine learning.
The entropy-stabilised oxide (Mg,Co,Ni,Cu,Zn)O forms one uniform rock-salt crystal at high temperature, but if you quench and then anneal it too low, it can unmix into separate copper-rich and other phases — proof that the single phase is held together by heat-driven entropy, not by chemistry alone.
Disorder itself can be the glue — but only while it stays hot enough for entropy to win.
The name oversells the mechanism: not every 'high-entropy' ceramic is truly stabilised by entropy, and single-phase formation is never guaranteed. Always check whether the phase is stable at the temperature you will actually use it.