Copper is most widely used interconnect material in present silicon microelectronic technologies. As such, multiple interfaces formed by a thin Cu layer and other materials must be engineered to achieve the desired chemical, mechanical, and electrical properties. Adhesion between Cu and the barrier layer, as well as between Cu and the dielectric, is of particular interest, due to its role in controlling interfacial stability and Cu electromigration behavior . This work focuses on understanding how the interface chemistry affects adhesion. Firstprinciples density functional theory (DFT) calculations were used to determine chemical adhesion energies of interfaces formed by Cu and various metals considered as a diffusion barrier, including Ta, TiN, and W. Calculations predicted increasing adhesion strength in the order of TiN < Si-doped TiN < TaN < Ta, consistent with wetting experiments done using 100Å thick Cu layer samples. The effects of doping at the interface using light elements (C, N, O) were also determined. Calculations were also done for interfaces of Cu with two different classes of amorphous dielectric materials, i.e. silicon nitride and silicon carbide, for which detailed material characterizations are often difficult and time consuming. Calculations predicted Cu/Si-nitride and Cu/Si-carbide adhesion strengths consistent with 4-pt-bend experiments, including the improvement in adhesion energies when silicon was used to dope the interface. In addition, weaker interfaces provide low-resistance diffusion paths for Cu atoms during electromigration. The first-principles based modeling, validated by select adhesion measurements, provides a predictive approach to effectively determine adhesion strengths and predict electromigration reliability in interconnects.