ABSTRACT Transient negative capacitance observed during ferroelectric polarization switching is often interpreted as an intrinsic material property, yet our study shows it can arise as a circuit artifact. We systematically varied the configuration of the measurement circuit used in fast voltage‐pulse experiments to characterise the ferroelectric polarization switching of epitaxial high‐quality thin‐film capacitors. We conclude that even modest resistances, such as the 50 Ω source impedance of a standard pulse generator, serially connected with the ferroelectric capacitor, significantly produced the characteristic voltage “dip” associated with negative capacitance, reducing the effective electric field across the ferroelectric and altering the polarization switching dynamics. Because the applied field is no longer constant, most of the theoretical models, such as conventional Kolmogorov–Avrami–Ishibashi or Landau–Khalatnikov, used to describe the ferroelectric switching, become invalid in the interpretation of the polarization switching in classical PUND (Positive‐Up‐Negative‐Down) measurements. To reveal intrinsic switching dynamics, we implemented a low‐impedance buffer amplifier. This buffered circuit configuration maintained a nearly constant field, eliminating the transient negative‐capacitance artifact, and revealed faster, higher‐current switching. We establish quantitative criteria for circuit design and device geometry to ensure accurate analysis of ferroelectric switching dynamics.
Marin Alexe (2026) studied this question.