Many questions concerning ferroelectric non-volatile memory cells are still not clarified, since this is a comparatively young technology. These include key issues like material properties, device design, circuit design and sensing schemes.
This leads to a difficult situation for research engineers. For example device and circuit designers should be able to transform the steady flow of innovation into new, more powerful memory cells, but each innovation might lead to side effects which can hardly be predicted from the present state of the research. This situation is even intensified by the highly nonlinear properties of the ferroelectric materials.
Consequently, the utilization of simulation tools becomes very attractive. In order to match the demands of device research, the following features have to be provided:
Up to date several approaches for the simulation of ferroelectric materials have been tried, but none of them combines all of the characteristics outlined above. They concentrate either on circuit simulation [BEL97][Dun94] or on effects inside the material like domain wall movement [Ish92][OAI91] or material phase transitions [RW95][WR97]. More relevant for development tasks is an approach that adds the ability to simulate ferroelectric materials to a state-of-the art device simulator like MINIMOS-NT [BDG+98], as it is the case for the work presented in this thesis.