Metals go through extreme processing: heating, rolling, bending, and reworking. The prevailing belief has been that such treatment randomizes atomic arrangements fully. But a new study from MIT shows that chemical order can survive, deeply influencing material properties.
The team, led by Assistant Professor Rodrigo Freitas, used large-scale molecular dynamics simulations and machine learning to track millions of atoms in alloys as they underwent cycles of deformation and heat. They found that atoms did not converge to a fully random distribution, instead, residual patterns of short-range order remained.
These patterns, described as nonequilibrium short-range order, arise because dislocations (defects in the crystal structure) guide atomic swaps preferentially. In other words, these defects don’t just disrupt; they also bias which bonds break and reform during processing. The team even built a simplified analytical model that captures how this process plays out: atomic reordering competes between disordering forces and selective reorganization around defects.
What makes this finding significant is that these hidden patterns affect physical properties, i.e., strength, durability, thermal capacity, and radiation tolerance, even though engineers have largely ignored them in standard alloy design. Until now, such ordering was only seen in controlled lab samples, not conventional industrial metals.
The research further maps how different processing paths (temperatures, deformation histories) lead to distinct chemical order regimes. Freitas and colleagues envision that engineers could someday “dial in” atomic patterns to fine-tune performance.
This work uncovers a layer of physics once believed negligible. It suggests that even in heavily worked metals, atoms don’t forget their past and that by understanding those memories, we might design alloys that perform beyond what we thought possible.