A research team led by Wenbo Bai and Qudong Wang of Shanghai Jiao Tong University, working with colleagues at Inner Mongolia University of Technology and the University of Maragheh in Iran, fabricated copper/aluminum/copper trilayered plates using a technique called high-temperature oxygen-free rolling. The process is notable for what it avoids: by rolling the metals together at elevated temperature in an oxygen-free environment, it prevents the stubborn oxide films that normally sabotage bonding between aluminum and other metals. The result was a genuinely metallurgically bonded interface, meaning the two metals are joined by atomic-scale bonding and interdiffusion rather than mere mechanical interlocking, which is the gold standard for load transfer across a laminated joint.

That bonding, however, comes with chemical consequences. When copper and aluminum meet at high temperature, their atoms diffuse into each other and form intermetallic compounds, brittle phases with their own crystal structures that are neither copper nor aluminum. In the rolled plates, the team found a thin intermetallic region measuring just 4.8 micrometers across, yet containing not one but three distinct compounds stacked in layers: Al2Cu, AlCu and Al4Cu9. The fact that the region remains so thin is a virtue, since thick intermetallic layers are notorious for cracking, but even a thin multilayer of brittle phases concentrated at the interface is a potential weak link when the material is stressed.

Using electron backscatter diffraction, a technique that maps the orientation and size of individual grains, the researchers uncovered a striking gradient in the microstructure around the interface. Grains near the boundary were heavily refined, broken down by the intense plastic deformation of rolling, and then gradually coarsened back toward the normal grain size of the bulk matrix. This interface-induced gradient is not merely decorative; gradient structures are known to contribute extra strengthening in heterogeneous metals because the mismatch between fine and coarse grains forces dislocations to pile up and interact. The aluminum layer also showed more dynamic recrystallization than the copper layer, meaning its deformed grains had more thoroughly re-nucleated into new, strain-free grains during hot rolling, which weakened the crystallographic rolling textures in the aluminum compared with the copper.

The dislocation densities told a similar story of asymmetry. By estimating geometrically necessary dislocation densities from the diffraction data, the team found roughly 18.57 x 10^14 dislocations per square meter in the copper layer against only 6.57 x 10^14 in the aluminum, nearly a threefold difference. Copper, with its higher melting point and greater resistance to thermal recovery, retains far more of the deformation damage imparted by rolling, while aluminum, which recovers and recrystallizes more readily, sheds much of it. This built-in heterogeneity between the two layers sets the stage for how the composite behaves when it is later pulled apart at temperature.

To probe that behavior, the researchers performed tensile tests at elevated temperatures and tracked three key metrics: yield strength, ultimate tensile strength and elongation. All three declined as the test temperature rose, but for different reasons. The loss of strength was traced mainly to thermal softening, the straightforward weakening of both metals as heat gives dislocations the mobility to move and annihilate more easily. The loss of ductility was more subtle and more worrying. The team attributed the falling elongation to the mismatch in thermal expansion between copper and aluminum, which generates internal stresses as the laminate heats and cools, and to progressive damage accumulating at the interface itself.

To understand how the material lost its ability to stretch, the researchers analyzed strain-hardening, the phenomenon by which metals actually become stronger as they deform, which is what allows them to neck slowly rather than snap suddenly. Using strain-hardening-rate curves and Kocks-Mecking plots, a classical framework that separates dislocation storage from dislocation recovery, they showed that the hardening capability dropped in a temperature-dependent way. Three culprits emerged: thermally activated recovery, which erases stored dislocations before they can strengthen the material; heterogeneous strain partitioning, in which the softer aluminum layer stretches more than the stiffer copper, overloading the interface; and interfacial damage, which progressively removes the very bonding that holds the sandwich together.

The most dramatic evidence came from the fracture surfaces. When specimens were pulled apart at 250 to 300 degrees Celsius, the fractures showed pronounced separation along the copper-aluminum interface, littered with debris of the intermetallic compounds. In other words, at these temperatures the composite no longer fails like a single material but peels apart at its seam, with the brittle intermetallic phases acting as crack nucleation sites and the fragments scattering across the failure surface. This intermetallic-assisted interfacial damage, the authors conclude, increasingly dominates premature fracture as temperature rises into this range.

The findings carry practical weight for anyone designing laminated metal components for warm service environments. They suggest that the operating window for copper/aluminum/copper composites should be defined not just by the softening of the constituent metals but by the integrity of the micrometers-thick intermetallic seam, and that processing routes which keep that layer thin and well-bonded, as oxygen-free rolling does, buy real performance. The work also feeds a broader research wave on heterogeneous and laminated metals, where controlled gradients and layer interfaces are deliberately engineered to deliver strength-ductility combinations that homogeneous alloys cannot match. Knowing precisely how those interfaces degrade under heat, the study argues, is essential before such composites can be trusted in demanding, temperature-cycling applications.

For now, the study stands as a detailed microstructural autopsy of a failing laminate, connecting the atomic-scale chemistry of intermetallic growth, the grain-scale gradients left by hot rolling, and the macroscopic loss of strength and ductility at temperature. It is precisely this kind of multiscale linkage, from dislocation densities to fracture surfaces, that turns a promising materials concept into an engineering-grade design rule, and it offers a clear target for the next generation of copper-aluminum composites: interfaces that stay intact when the mercury rises.

Subject of Research: Interfacial microstructure and elevated-temperature tensile behavior of Cu/Al/Cu composite plates made by high-temperature oxygen-free rolling

Article Title: Interfacial microstructural heterogeneity and elevated-temperature tensile behavior of Cu/Al/Cu composite plates fabricated by high-temperature oxygen-free rolling

Article References: Bai, W., Cai, H., Ebrahimi, M., Zhao, Y., & Wang, Q. (2026). Interfacial microstructural heterogeneity and elevated-temperature tensile behavior of Cu/Al/Cu composite plates fabricated by high-temperature oxygen-free rolling. Journal of Materials Science. https://doi.org/10.1007/s10853-026-13839-w

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13839-w

Keywords: copper-aluminum composites, laminated metals, intermetallic compounds, oxygen-free rolling, elevated-temperature tensile testing, electron backscatter diffraction, dynamic recrystallization, strain hardening, interfacial fracture, thermal expansion mismatch, dislocation density, Journal of Materials Science