4.3 Comparison of the Two-Dimensional Model to Measurements

A key question for the simulation model is whether it is justified to ignore the existence of domains and to treat the material as a continuum. Indeed, experimental data for PZT directly deposited on Ti/Pt electrodes show a dependence of ferroelectric properties on the thickness of the film, thus questioning this approach. These difficulties were solved by the assumption of a non-ferroelectric blocking layer (Fig. 4.12) at the interface of the contact electrodes.

Figure 4.12: Non-ferroelectric blocking layer

Thus good correspondence of the measured data to the continuum approach was obtained [LDTvV94]. These calculations were carried out using a one-dimensional continuum model. The decisive parameter for the blocking layer is the ratio between its relative dielectric constant $\epsilon_\mathrm{bl}$ and its thickness $d_\mathrm{bl}$. The reported, specimen dependent, values of $\epsilon_\mathrm{bl}/d_\mathrm{bl}$ are 20-30 $\mathrm{nm}^{-1}$.

Later experiments with $\mathrm{La_{0.5}Sr_{0.5}CoO_3}$ electrodes immediately matched the expected predictions of linear scaling of field properties [CPW97], thus giving further evidence that the model is appropriate.