Decoupling Mesoscale Functional Response in PLZT across the Ferroelectric–Relaxor Phase Transition with Contact Kelvin Probe Force Microscopy and Machine Learning

Abstract

Relaxor ferroelectrics exhibit a range of interesting material behavior including high electromechanical response, polarization rotations as well as temperature and electric fielddriven phase transitions. The origin of this unusual functional behavior remains elusive due to limited knowledge on polarization dynamics at the nanoscale. Piezoresponse force microscopy and associated switching spectroscopy provide access to local electromechanical properties on the micro-and nanoscale, which can help to address some of these gaps in our knowledge.However, these techniques are inherently prone to artefacts caused by signal contributions emanating from electrostatic interactions between tip and sample. Understanding functional

To switch or not to switch – a machine learning approach for ferroelectricity

“…10 To further corroborate the applicability of our analysis approach we process cKPFM response measured across the ferroelectricrelaxor phase transition on multiple grains of lanthanum zirconate titanate (PLZT), which even in the ferroelectric state exhibits peculiarities in hysteresis loop, as commonly observed for relaxors. 31 Fig. 1(a) schematically depicts the DC voltage waveform used in cKPFM, which consists of triangular write pulses V write to initiate ferroelectric switching and a probe voltage V read that is applied between the write pulses and stepwise changed with every write cycle.…”

“…4(c). Apart from HAC and DBSC, other algorithms like k-means (e.g., Li et al, 32 Neumayer et al 31 ) can be used to group response to nd trends, e.g., dependent on material, location, temperature, etc. Successful differentiation of switching properties via machine learning of cKPFM maps motivated us to apply nonlinear clustering methods, such as neural networks, that could potentially reveal even more details of dielectric behavior.…”

“…In order to analyze and extract meaning from these data sets, machine learning has become increasingly important. [27][28][29][30][31][32][33][34] Dimensionality reduction without loss of important information, de-noising, clustering and identifying characteristic features in data sets have been achieved using supervised and unsupervised machine-learning algorithms.…”

Machine learning for composite materials

“…Recently, ML has turned our daily life to be more convenient in numerous ways by influencing image recognition, autonomous driving, e-mail spam detection, among others. [31][32][33][34][35][36][37][38] In the field of materials science, most ML applications are concentrated on discovering new chemical compounds or molecules with desired properties. Those studies can be categorized into a subfield of ML for materials chemistry.…”

Generalizable Framework for Algorithmic Interpretation of Thin Film Morphologies in Scanning Probe Images

“…There has been compelling research from various groups on utilizing computer vision and machine learning approaches to automatically identify features in micrographs and SPM measurements. ,− These approaches often use supervised models, which inherently restrict their application to materials or morphologies that are similar to the samples used in training the model. Moreover, the performance of supervised models is connected to the number of samples in the training data set.…”