Figure 1.

Issues with direct determination of ferroelectric parameters. (a) A typical ferroelectric hysteresis loop ($Q_{\mathit{FE}}$-$E_{\mathit{FE}}$), where $E_{c,ext}$ directly determined at $Q_{\mathit{FE}}$ = 0 μC/cm² is smaller than the actually $E_c$ from the $P_{\mathit{FE}}$-$E_{\mathit{FE}}$ loops. In addition, $P_s$ is not obvious from the $Q_{\mathit{FE}}$-$E_{\mathit{FE}}$ curve. (b) A hysteresis loop with asymmetrical $E_{c\pm }$ and non-zero $P_{\mathit{offset}}$, where 2$P_{r,ext}$ is smaller than 2$P_r$, leading to underestimation. The green square indicates the endpoint of the 2$P_r$ when the start point (top yellow square) of the 2$P_r$ aligned with that (top yellow circle) of the 2$P_{r,ext}$. (c) A minor hysteresis loop (blue) and the corresponding saturated hysteresis loop (dashed red), where 2$P_{r,minor}$ is smaller than 2$P_r$ and $E_{c,minor}$ is less than $E_{c,ext}$. Extracting ferroelectric parameters from a minor loop leads to underestimation.