Current crowding effects and determination of specific contact resistivity from contact end resistance (CER) measurements

Current crowding effects on Contact End Resistance (CER) test structures due to the finite diffusion overlap of the contact window are studied by experiment and numerical simulation. This finite overlap adds a parasitic resistance component not accounted for by the standard one-dimensional theory, and if uncorrected, this parasitic resistance may lead to gross overestimation of the true specific contact resistivity ρ<inf>c</inf>. The overestimate increases with increasing diffusion sheet resistance and large contact size. Excellent agreement between experiment and simulations has allowed this effect to be modeled. Accurate values of ρ<inf>c</inf>in the range of 5 × 10^-8to 2 × 10^-5Ω cm²are extracted using CER structures.