ABSTRACT
Insulators with a large static dielectric constant (usually referred to as ‘high-κ insulators’) are currently being considered as possible replacements for SiO2, because of the necessity of increasing the gate capacitance of Si metal–oxide–semiconductor field-effect transistors (MOSFETs), while avoiding the problems that arise when the SiO2 thickness is reduced below the 1.5–1.0 nm range, as demanded by device scaling [1, 2]. At least at present, these efforts are still mainly aimed at improving the chemical and physical properties of the insulating materials. Yet, in this paper, we point out an intrinsic, possibly unavoidable, and unwanted property of these materials, namely, the fact that their high dielectric constant may necessarily cause a reduction of the electron mobility in the inversion layer of Si MOSFETs. The dielectric constant of a (non-metallic) solid results from the contribution of the ionic and the electronic polarization. The latter is roughly inversely proportional to the square of the direct band gap of the solid, averaged over the Brillouin zone. Insulators, by definition, have large band gaps, so that there is little one can do to increase the electronic polarization and a larger (static) dielectric constant can only stem from a larger ionic polarization, often due to highly polarizable (‘soft’) metal–oxygen bonds. Associated with soft bonds are low-energy optical phonons. By contrast, the ‘hard’ Si–O 398bonds in SiO2 yield a reduced ionic polarization. Associated with ‘hard’ bonds are ‘stiff optical phonons.
