An overview of state of the art strained interface gate stacks is given and their mathematical description with the aid of the k.p method is presented. The description of the influence of strain on the semiconductor is extended to ultra-thin body FETs with the help of a confinement potential and a two-band k.p model. The effects on the dispersion relation, energy splitting, and effective masses for ultra-thin body FETs along for primed and unprimed subbands with respect to the point at different strain levels and body thicknesses are studied. The utilized two-band k.p Hamiltonian accurately describes the band structure up to energies of and incorporates a shear strain component, neglected in the parabolic approximation. Due to the modification of the effective mass by shear strain it is an important source of mobility enhancement one has to account for in ultra-thin silicon films.
It is shown that the effective masses along and become different for decreasing film thickness. In ultra-thin films the large separation in energy between the primed and the unprimed subbands even without stress leads to a de-population of the primed subbands. Tensile stress in direction generates a shear component which changes the transport effective masses of the unprimed subbands and shifts the primed subbands with unfavorable effective masses up in energy. With decreasing film thickness the decrease of the effective mass along direction induced by shear strain becomes more pronounced enabling mobility enhancement even in ultra-thin films. Unfortunately, the density of states effective mass in unprimed subbands increases with shear strain and thus results in higher scattering rates, which reduces the mobility gained due to the thickness-enhanced transport mass decrease at high stress values. However, the mobility enhancement remains significant.
Thereafter, a detailed introduction into the modeling of electrolytic interfaces is presented and its peculiarities like, the double layer, the Stern layer, the site-binding model, and the different available models for the electrolyte are elaborated. Several examples illustrate the exploitation of electrolytic gate stacks for pH, DNA, and biotin-streptavidin sensitive devices. The presented pH sensor (ISFET) was the first device utilizing an electrolytic gate stack. The simulations performed show good agreement with the values reported in literature. The examples for detecting DNA predict an angular dependence of the DNA molecule orientation with respect to the surface and show that for low salt concentrations the commonly employed Poisson-Boltzmann model can not reproduce the screening of the DNA in the electrolyte correctly, while the Debye-Hückel model is able to fit the experimental data. This can be explained by the in this work introduced extended Poisson-Boltzmann model, which is able to adjust the screening behavior with respect to the average closest possible distance between two ions. Finally, a BioFET for detecting biotin-streptavidin is analyzed, studying the influence of several dielectric materials, molecule surface density, and molecule orientation with respect to the surface.
Even though the end of scaling is beginning to rise on the horizon, there are plenty of white spots on the map of microelectronics and there is more than enough room for further improvements and fascinating new applications.