Reevaluating the crystallographic textures in Pb(Zr, Ti)O3 piezoelectric ceramics
In a recent publication, Li et al. claimed the achievement of superior (001)-oriented textures in pseudo-cubic Pb(Zr,Ti)O3 ceramics with Zr/Ti ratios ≥55/45 by utilizing Ba(Zr0.1Ti0.9)O3 microplatelet templates. The texture fractions were reported to be ≥94% based on X-ray pole figures and electron backscattered diffraction (EBSD)-derived pole figures, even reaching as high as an incredible 99.7% from mere X-ray diffraction (XRD) θ-2θ scans. However, upon reevaluating the presented results, it appears that the Pb(Zr,Ti)O3 ceramics likely contain only about 40.6%–44.5% of (001)-oriented grains, with the remaining grains displaying near-random orientation distributions.
The quantification of the (001)-textured component requires determining the volume fraction fv(001,ωmax) for crystallites with (001) orientation within a specified solid angle spread ΔΩ. Operationally, this volume fraction is calculated through numerical integration of the expression
where P002(ω,φ) denotes the pole density distribution of the {002} pole figure, and ωmax symbolizes the predefined maximum permissible tilt angle (typically ±5°). This angular tolerance parameter can be optimized based on experimental requirements. Given its significant impact on texture fraction calculations, precise specification of the employed ωmax value is necessary when reporting any fv(001,ωmax) quantification. Unfortunately, Li’s article exhibits two critical methodological shortcomings concerning texture quantification: absence of proper numerical integration and failure to specify the angular tolerance parameter.
X-ray pole figure analysis remains the most established method for evaluating macrotextures using XRD with a texture goniometer. The macrotexture obtained needs to be validated through microtextural characterizations. The ideal approach for such validation would involve thorough EBSD investigations and/or systematic selected area electron diffraction (SAED) analyses combined with transmission electron microscopy (TEM) imaging. Regrettably, in Li’s article,1 the fv(001,ωmax) value is quantified from a mere XRD θ-2θ scan, which is inadequate for semiquantitative texture analysis. Additionally, SAED patterns are missing, and both the X-ray pole figure and EBSD are incorrectly applied in quantitative texture assessments. I am addressing the methodological flaws and presenting my reassessment of the (001) textures as follows.
