"Effect of interfacial bridging atoms on the strength of Al/CNT composites: machine-learning-based prediction and experimental validation"
Abstract : Weak interfacial adhesion is one of key obstacles to develop aluminum matrix composites containing carbon nanotubes (CNTs). This study suggests the concept of bridging atoms to enhance the interfacial wetting between aluminum and CNTs. Machine learning and
sensitivity analyses were employed to determine the most favorable element as a bridging atom. Copper was identified as the most effective bridging atom, and its bridging efficiency (enhancement of strengthening efficiency of CNTs) was experimentally validated by
comparison with those in the monolithic Al and AleSi matrix. As a result, the strengthening efficiencies of the CNTs were measured to be ~43, 27, and 73 MPa/vol% for the Al, AleSi, and AleCu matrices, respectively, which is comparable with the prediction by the machine learning model

"Synthesis of high-entropy alloy thin films via grain boundary diffusion–assisted solid-state alloying"
Abstract
The interface reaction and its effect on the mechanical properties have been experimentally studied for aluminum-based composites reinforced with titanium dioxide nanoparticles (TiO2). Aluminum-based composites containing TiO2 nanoparticles (20 nm in size) are developed by hot-rolling the ball-milled powder. High chemical potential energy of the nanoparticles induces the fast formation of the interface layer during the heat treatment process, and high yield stress of 514 MPa in a composite containing 5 vol.% TiO2 nanoparticles can be achieved. Furthermore, dislocations are emitted at the nanoparticle/matrix interface during deformation due to the high stress concentration. This study can provide useful insights for the design of metal-matrix composites.

"Mechanism for self-formation of Al matrix composites using nitridation-induced manufacturing processes"
Abstract
The present study investigates the effect of multi-walled carbon nanotubes (MWCNTs) on atomic diffusion processes in the metal matrix, by comparing growth kinetics of interfacial phases in Al–Cu and Al/MWCNT–Cu diffusion couples. Multiple intermetallic layers such as Al2Cu, AlCu, Al3Cu4, Al2Cu3, and Al4Cu9 are formed at the interface between Al and Cu during heat treatment of an Al/Cu diffusion couple at 530C. For the diffusion couple of ultrafine-grained Al and cast Cu, the growth rate of intermetallic layers is comparable with theoretical expectations based on Fick’s second law. On the other hand, MWCNTs significantly restrict the diffusion of Al atoms in the composite because the atoms tend to detour around the long tube for diffusion, particularly when MWCNTs are oriented perpendicular to the atomic diffusion path. This accelerates the growth of voids at the contact interface of the diffusion couple during heat treatment.

"Effect of interfacial bridging atoms on the strength of Al/CNT composites: machine-learning-based prediction and experimental validation"
Abstract : Weak interfacial adhesion is one of key obstacles to develop aluminum matrix composites containing carbon nanotubes (CNTs). This study suggests the concept of bridging atoms to enhance the interfacial wetting between aluminum and CNTs. Machine learning and
sensitivity analyses were employed to determine the most favorable element as a bridging atom. Copper was identified as the most effective bridging atom, and its bridging efficiency (enhancement of strengthening efficiency of CNTs) was experimentally validated by
comparison with those in the monolithic Al and AleSi matrix. As a result, the strengthening efficiencies of the CNTs were measured to be ~43, 27, and 73 MPa/vol% for the Al, AleSi, and AleCu matrices, respectively, which is comparable with the prediction by the machine learning model