Abstract
Efforts for realizing microscopic and nanoscopic ferroelectric devices have contributed to uncover new prospects of the ferroelectricity. Examples of such efforts are memories based on the ferroelectric field effect. In this article, we discuss the origin of the ferroelectricity in metal oxides in view of these applied researches. In particular, the long-range electrostatic interactions are presumably not the primary origin of ferroelectricity, because it is incompatible with the experimental existence of free electrons/holes at polarization discontinuities. On the other hand, local covalent interactions are compatible with free electrons/holes at polarization discontinuities can explain ferroelectricity as well as the existence of multiferroics.
Acknowledgments
The author acknowledges enlightening comments by Prof. T. Janssen and Prof. Y. Ishibashi and Prof. J. Scott for very interesting discussions during the visit in Colorado Spring in 1992.