Structural contribution to the ferroelectric fatigue in lead zirconate titanate ceramics

Many ferroelectric devices are based on doped lead zirconate titanate (PZT) ceramics with compositions near the morphotropic phase boundary (MPB), at which the relevant material's properties approach their maximum. Based on a synchrotron x-ray diffraction study of MPB PZT, bulk fatigue is unambiguously found to arise from a less effective field induced tetragonal-to-monoclinic transformation, at which the degradation of the polarization flipping is detected by a less intense and more diffuse anomaly in the atomic displacement parameter of lead. The time dependence of the ferroelectric response on a structural level down to 250 μs confirms this interpretation in the time scale of the piezolectric strain response.

Figure 1

Difference in the physical and structural behavior for an unfatigued and a fatigued PZT. (a) Cycling dependence of the spontaneous polarization and of (b) field induced strain; hysteresis cycle (HC) electric field values range (1–4). Influence of the applied electric field on the ${002}_{\mathrm{c}}$ reflection for (c) unfatigued and (d) fatigued sample.

Figure 2

Rietveld refinements for an unfatigued and a fatigued PZT. (a) Unfatigued PZT at 1.635 kV/mm and (b) PZT at 1.618 kV/mm previously fatigued at 50 Hz kV/mm. The difference plot (blue) has the same scale. The vertical green tickmarks indicate the calculated positions of the $P4mm$ phase (top) and Cm phase reflections (bottom). The inset corresponds to an enlargement of the ${002}_{\mathrm{c}}$ reflection and the high-2θ range.

Figure 3

Structural dependence of the PZT sample as a function of the applied electric field obtained by Rietveld refinements. Monoclinic-tetragonal phase ratio (percent) for (a) the unfatigued and (b) the fatigued sample. Isotropic mean square displacement of the lead atom in the monoclinic phase $B$_iso(Pb)${}_{\mathrm{m}}$ for (c) the unfatigued and (d) the fatigued sample. Hysteresis cycles (HC) electric field values range (1–4) are indicated.

Figure 4

Rectangular on-off stroboscopic pump-probe measurements of PZT. (a) Unfatigued precycled sample; the sample was precycled with ${10}^5$ cycles to assure a reproducible reaction throughout the whole experiment. (b) PZT fatigued with 50 Hz. In this setup the maximum field of 2.0 kV/mm is applied immediately and complete diffraction patterns are measured stroboscopically. The modulation of the field follows a rectangular on-off function with 50 Hz.