Chemical disorder and structural order are assumed to trade off. More disorder in composition means worse crystallinity. Alloys are messier than pure metals. High-entropy materials — five or more elements mixed in roughly equal proportions — should be the messiest of all.
Almishal and colleagues (arXiv:2603.00814) grew high-entropy oxide thin films that combine extreme chemical disorder with exceptional crystalline quality. Lattice strains exceeding five percent are stabilized through compositional design and epitaxial constraints. The atoms are randomly arranged by species, but the lattice they sit on is nearly perfect. The composition is chaos. The structure is order.
The payoff appears at the interface. When two entropy-stabilized oxide layers meet at a “pseudomorphic valence interface,” cobalt atoms shift between oxidation states. This produces doubled exchange bias compared to single layers — a magnetic effect that emerges from the boundary between two kinds of designed disorder. Neither layer alone produces it. The interface, where one random composition meets another on a shared crystalline scaffold, is where the new physics lives.
The assumption being overturned is that order and disorder are opposite ends of a single axis. They are not. They are orthogonal. You can independently tune how disordered the chemistry is and how ordered the structure is. The high-entropy composition is not fighting the crystallinity — it is occupying a different dimension entirely. Maximizing along one axis does not require minimizing along the other.
This decoupling changes what is designable. If compositional disorder and structural order were antiparallel, then any improvement in one would cost the other. Because they are orthogonal, the design space is a plane, not a line. The corner where both are extreme — maximum chemical entropy, maximum crystalline perfection — is not forbidden. It is merely unexplored.