Spectroscopic And Microscopic Investigation On First-Order Phase Transition And Polarization Properties In Ferroelectric Lattice

The studies of lattice structure and phase transition of matter and phonon-electron coupling property are of fundamental interests in condensed matter physics and material science.The structure characteristics could determine the variety of functionalities and some potential applications of material.In recent years,with the rapid development of information technology,microelectronic material and technology,the limitations in Moore's law have led to a series of bottlenecks in semiconductor and information materials research.It requires the related novel materials and semiconductor studies to meet some new criteria,such as low dimension,high efficiency of photoelectric/piezoelectric conversion,good mechanical property,and stable chemical property.However,companied with the investigation of novel material systems with the brilliant physical characteristics,there are some ambiguous basic physics in plenty of traditional functional material system of microelectronics.In order to solve these problems,except for the controllable synthesis of these conventional material,it also depends on futher understanding the structural and other physical properties on both micro and macroscopic dimensions by improving the analyzing tools.In this dissertation,we consider K0.5Na0.5NbO3(KNN)based ferroelectrics as the main samples to carry out a series of studies,where KNN based ferroelectric crystal possesses Pb-free environment-friendly composition and high electromechanical response.The study aims at explaining the lattice structure details and ferroelectric polarization characteristics by in-situ spectroscopic and scanning probe microscopic(SPM)techniques,proposing a new method for analyzing spectroscopic data,and designing a set of novel technique used to imaging electromechanical property and structure characteristics.(1)Designing the in-situ probing technique based on Raman spectroscopy and piezoresponse force microscopy(PFM)to investigating the physical correlations among ferroelectric spontaneous polarization,internal energy evolution in lattice,crystalline symmetry,and first-order phase transition:In terms of phase transition in ferroelectrics with complicated structures,such as the typical ABO3 perovskite and LiNbO3 ferroelectric classes,spontaneous polarization is regarded as order parameter for studying ferroelectric lattice dynamics under the doping and multi-field dependence.Many mature characterization techniques have been used to probe the lattice structure and phase transition so far,such as X-ray diffraction(XRD),temperature dependence of dielectric constant,infrared spectroscopy and Raman scattering,etc.As one of SPM modes,PFM technique is widely used to image ferroelectric surface polarization properties and indicate the polarization switching,domain wall,and polarization direction.However,the specific evolution of spontaneous polarization as the order parameter companied with phase transition is still unclear,especially for KNN perovskite system.Therefore,this study focuses on investigate the physical relationship among temperature dependent evolutions of spontaneous polarization,internal energy and strain,crystalline symmetry and phase transition by in-situ imaging technique with the resolution of 0.1-1 ?m.Most of domain-wall directions(180°,90°-a,90°-a-c,and 60°)in orthorhombic(O)phase and their evolution under first-order phase transition of KNN crystal have been systematically observed.During the O to tetragonal(T)phase transition,the thermal dynamics in various types of spontaneous polarization is closely related to the bonding strength and internal lattice energy in molecule.According to the energy minimization of system,we discuss the underlying physics,for example,90° domain wall density could determine the change of bonding strength and lattice energy.(2)Exploring the fundamental mechanism of Raman scattering on analyzing ferroelectric first-order phase transition,structural evolution and phonon characteristics,as well as presenting three scattering criteria for studying first-order phasr transition and symmetry evolution:In combination of the experimental discussion based on unpolarized and polarized Raman scattering with the theoretical analysis of light scattering mechanism,this issue deeply discusses the correlations between the scattering criteria and lattice structure details during the first-order phase transition of KNN based system.We clearly point out that three criteria(phonon frequency,polarized scattering intensity and depolarization ratio)could be used to study the molecular structure,symmetry and phase transition.Frequency shift of phonon during the temperature changing maybe ascribed to factors,including the lattice distortion,change of bond length,phonon occupation,and phase transition.Thus,we indicate the intrinsic relationship between the frequency shift and these factors.It is also proven in theory that the polarized scattering intensity is proportional to the molecular polarizability of lattice.And the related explaination on the correlation between the scattering intensity and the electron dynamics in the atomic space in molecule has been revealed.In addition,we have also studied the temperature dependent symmetry of different lattice vibrational modes accompanied with first-order phase transition by calculating the values of depolarization ratio.Except for the KNN based crystals,these Raman scattering criteria used to indicate the structure and symmetry are also suitable to more the developing fields of material and condensed matter.(3)Proposing an intelligent and novel method to investigate the structure of matter by spectroscopic exploration based on machine learning and big data techniques:Material informatics,as one of blooming subjects,has recently attracted broad attentions.Machine learning and big data analysis methods play important roles in assisting and accelerating the process of experimental data from material characterization and computational modelling.The spectroscopic and microscopic imaging data from experiments in this work is multi-dimensional and abundant.And these multi-dimensional datasets are suitable to be analyzed by statistical algorithm.Hence,we first give an attempt on establishing a spectroscopic dataset from Raman scattering on KNN system,and further choosing,training,optimizing,and using a classifier algorithm to predict lattice structure and construct a phase diagram.The radial basis function(RBF)kernel support-vector-machine(SVM)algorithm and principal component analysis algorithm are chosen to design a phase classifier,and realized the reliable structure prediction on a set of ‘new' scattering data.This work has built a unifying framework in the application of machine learning in material's structural identification,enabling the synergy of material synthesis,experimental characterization and computational modelling.The integral workflow for a ferroelectric phase classifier from material synthesis to structural prediction has been demonstrated,and could be popularized to the machine-dominant studies in more extensive materials,biological and chemical fields.(4)Imaging the electromechanical coupling property in conductive liquid environments on the nanoscale by PFM technique:PFM plays a crucial role in investigating ferroelectric surface polarization property,and precisely recognizing the domain wall information and polarization direction.The polarization behavior in ferroelectrics is one of the classic electromechanical coupling characteristics.As the property describing the interaction between electric and mechanical energy,electromechanical coupling widely exists in many systems,such as biological cell and tissue,energy storage and conversion,as well as all piezoelectric materials.However,it is still a technical challenge for in vivo or in operando detection of electromechanical coupling or electrochemical reaction in electrolytes or other conductive liquids.Therefore,we aim at realizing electrical microscopic imaging in conductive liquid environments by PFM and other derived SPM modes.In NaCl electrolytes of less than 0.5 M,ferroelectric domain visualization by PFM has been successfully achieved by the steps of choosing probe,experimental designation,and numerical modelling.Simultaneously,we simulate electric field screening effect in tip-surface junction by finite element analysis,when tip is drived by a modulation ac bias.It clearly explains the difficulty in PFM imaging in highly conductive liquids.The technique of imaging electromechanical coupling by PFM will pave the way and promote the development of the investigations in the fields of under-water electronics,in vivo bio-piezoelectrics,and electrochemical systems.