Data-mining from reliable extant data on basis of material informatics can lead to unveil those hidden qualitative and quantitative rules and make predictions of new materials faster and cheaper, which would reduce human synthesis effort than traditional trial-and-error way. In contrast to traditional scenario of material research initiated by scientists’ intuition and in a chain of synthesis-observation-analyses of mechanism, modern material research will change into productions and applications of data provoked by data science paradigm, while the present bottleneck of material informatics is still limited by the number of data. In the field of functional materials of ferroelectric piezoceramics, there are two ongoing trends of developing lead-free piezoceramics with thermal aging stability and high temperature piezoceramics with higher response for applications like in Internet-of-Things (IoT) and smart systems. Recent years, data-mining on oxide perovskites and designing new perovskite-type high temperature piezoceramics have been carried out and progresses were firstly reviewed in a short. Based on those quantitative relationships and rules data-mined out previously, various kinds of substitution elements for BiFeO3-BaTiO3(BF-xBT) solid solution ferroelectrics were proposed on basis of the mass and cation size features of atom and their effects on ferroelectric phase transition temperature, dielectric and piezoelectric properties observed experimentally. The compositions were tentatively chosen near the rhombohedral-pseudocubic structural phase boundary in order to check predominant contribution from substitution element. Among those elements of Al, Sc, Zn1/2Ti1/2 substitution for Fe and Bi1/2Nb1/2, Al1/2Nb1/2, Zn1/3Nb2/3, Zr, Sn for Ti, it was demonstrated experimentally that Bi(Zn1/2Ti1/2)O3 (BZT) exhibits a very promising effect in the BF-xBT system to obtain piezoceramics prepared using conventional solid state electroceramic processing with robust insulating and low dielectric loss properties fulfilled for the requirements of engineering applications. For those low dielectric loss BF-xBT-yBZT ternary solid solution ceramics, their ferroelectric Curie temperature (TC) follows a quadratic polynomial relation of TC (μ)= a+bμ+ cμ² and reduced mass of unit cell (μ ) is a good descriptor to represent BF-xBT and BF-xBT-yBZT solid solution perovskite numerically. Full poling for piezoelectricity was obtained in Mn-doped BF-xBT-yBZT (x≤0.32, y<0.05) ceramics. So far, two ferroelectric piezoceramic materials were selected out: one is Mn-doped 0.74BF-0.22BT-0.04BZT exhibiting piezoelectric constant d33~75 pC/N, dielectric constant ε33T/ε0~260, dielectric loss tanδ ~0.01, thickness electromechanical coupling factor kt~0.42, radial electromechanical coupling factor kp~0.29 and TC~630°C, four times piezoresponse of commercial K-15 Aurivillius-structured bismuth titanate ceramics; another is Mn-doped 0.69BF-0.27BT-0.04BZT exhibiting d33~145 pC/N, ε33T/ε0~570, tanδ ~0.03, kt~0.41, kp~0.33, TC~510°C, 50% beyond K-81 tungsten-bronze-structured lead metaniobate ceramics. Similar to PbZr1−xTixO3(PZT) system, BF-BT-BZT system exhibits a structural phase boundary and thus their piezoelectric performance and thermal stability could be modified in a wide range through adjusting composition and doping. At this point, it is time performing systematical characterizations on BF-BZT-BT perovskite-type piezoceramics for high temperature engineering application testing. Based on the correlation between d33 and ε33 for piezoceramics data-mined out previously, how to reduce synthesis effort for designing new materials was also argued using “bad” dielectric data of those high dielectric loss samples. Our essay demonstrates data-mining driven designing based on material informatics sure able to reduce time-to-insight and human effort on synthesis, accelerating new
materials discovery and deployment.
perovskite-type oxides, data-mining, piezoceramics, ferroelectric phase transition, piezoresponse
