## Graduation Year

2022

## Document Type

Dissertation

## Degree

Ph.D.

## Degree Name

Doctor of Philosophy (Ph.D.)

## Degree Granting Department

Physics

## Major Professor

Inna Ponomareva, Ph.D.

## Committee Member

Robert Hoy, Ph.D.

## Committee Member

Manh-Huong Phan, Ph.D.

## Committee Member

Razvan Teodorescu, Ph.D.

## Committee Member

Sarath Witanachchi, Ph.D.

## Keywords

Electrocaloric, Hybrid Perovskite, Perovskite, Physics, Simulations

## Abstract

Ferroelectric materials are a type of multifunctional material that exhibit spontaneous polarization reversable by the application of an electric field. They are used in many technologies such as ferroelectric RAM (FeRAM), piezoelectric devices, RFID chips, and capacitors. However, the most commonly used ferroelectrics are often made with rare, expensive elements and are not environmentally friendly. For example, many prototypical ferroelectrics contain elements such as lead, zirconium, and titanium. As technology grows more advanced, there is a need to discover or manufacture cheaper ferroelectrics and to make them less impactful on human health and the earth as a whole. We set out to determine some potential alternatives to inorganic ferroelectric perovskites and to predict their structural, electronic, electric, and energy converting properties. Addtionally, we address some issues with computational methodologies that are used to study inorganic ferroelectrics before they can be applied to hybrid organic-inorganic ferroelectric perovskites.

First, we investigate the energy converting property of the antiferroelectric PbZrO3. Specifically, we investigate its electrocaloric effect, which is the the reversible change in temperature under adiabatic application of an electric field. We begin by applying the direct and indirect methods to predict the electrocaloric effect for lead zirconate for a range of temperatures. The direct method is an NVE simulation which allows the temperature to change under the application of an electric field, and therefore, the electrocaloric effect can be directly measured by this change temperature. We use the direct method to predict the electrocaloric effect in antiferroelectrics with the phase competition. We predict a large electrocaloric effect in the region of the ferroelectric-antiferroelectric phase transition, as well as the coexistence of positive and negative electrocaloric effect that is tunable by the application of an electric field. Additionally, we predict a large electrocaloric effect in the vicinity of lossless polar-antipolar phase transition. We also find that phase switching with hysteretic losses causes irreversible heating and is explained via thermodynamics. The indirect method is used in experiment and numerically integrates the Maxwell equation to calculate the electrocaloric effect. To simulate these conditions, we use adiabatic and isothermal Monte Carlo computations, which use a weighted random walk method within the Monte Carlo technique and statistical mechanics to solve for the equilibrium properties of a system. However, there are various controversies associated with it such as the inability to identify the true value of the electrocaloric effect which is different for the polar and antipolar states which we aim to address. Finally, we determine the potential of antiferroelectric lead zirconate for its ability to be used successfully in solid-state refrigeration cycles. We test a refrigeration cycle using direct simulations that tames the irreversible heating and is shown to outperform convention cycles that are based on only fully reversible regions.

We notice however that in these simulations the coercive field is overestimated and we had to rescale our results by a factor of 4.2 in order to bring our results in line with experiment. An effective Hamiltonian is a model of the energy of a ferroic and which reproduces the static and dynamical properties of a perovskite ferroelectric, including accurately capturing the phase transitions in a material. However, computations often overestimate the coercive electric field for inorganic ferroelectrics. This is primarily due to the fact that defects are neglected in computations but are inherent within experiment. We investigate various potential causes of this overestimation using the lead titanate and determine what may be added to computations that would bring our predictions in line with experimental findings without greatly increasing computation time and cost. In fact, we find that the depolarizing field and the development of polar domains are the main reason for the mismatch.

Another issue with the effective Hamiltonian is that the Curie temperature is often underestimated. It is believed that the issue originates in density functional theory computations which are used to parameterize the effective Hamiltonian. Density functional theory solves the time-independent Scho ̈dinger equation using the Kohn-Sham approximation and the Hohenberg-Kohn theorems. The Kohn-Sham equation includes the kinetic energy, potential energy from the electron-nuclei interactions, Hartree potential which is the energy from the Coulombic interactions between electrons, and the exchange-correlation potential whose exact solution is unknown. We know that effective Hamiltonians that are parameterized using different exchange-correlation potentials can yield different predictions of the Curie temperature. From this knowledge, we screen ten relatively new functionals from the Minnesota suite that, for the most part, have not been used previously to predict the structural properties and polarization of inorganic ferroelectric perovskites; specifically, we test them on lead and barium titanate. We had hoped to find one functional that would predict the energy well and polarization similar to the values obtained in experiment. However, we find that the results of the functionals are material dependent but we find a few that predict these better than the popular functionals.

The first half of this work is dedicated to the study of conventional ferroelectrics, including their energy converting and electrical properties, and using them with the aim to find solutions to some failings of our computational methodologies. However, the previous studies all include the material PbTiO3 which, due to it inclusion of lead, is toxic and is relatively expensive due to its inclusion of titanium. Thus, we investigate emerging alternative to these conventional inorganic ferroelectrics. Recently, there have been theoretical predictions showing that inverse-hybrid perovskites (IHP) may be a promising new ferroelectric. These novel materials may hold the key to cheap, environmentally friendly ferroelectric technologies. Since IHPs have just been predicted, their properties are mostly unknown. Computations provide a fast, reliable, and inexpensive way to study their properties. However, there needs to be a methodological framework developed to investigate them computationally in order to address their unique problems. We investigate one IHP, specifically (CH3NH3)3OI, using density functional theory and was predicted to have a large spontaneous polarization from recent first-principles computations. We propose a route to predict the structural and electrical properties, including the polarization, polarization reversibility, and the associated coercive field for hybrid organic-inorganic perovskites. Within this route, a polarization reversal path is constructed which models experimental approaches. Along this polarization reversal path, we identify competing ground state structures. We also predict the piezoelectric stress and strain constants and elastic moduli, as well as the polarization response to epitaxial strain.

In the last study, we investigate another emerging alternative for conventional ferroelectrics, hybrid inorganic-organic perovskites, which have the same chemical formula ABX3 but contain one or more organic molecules which can be placed on any of the sites. Unlike IHPs, hybrid perovskites are found in nature and have been synthenized succesfully for over a decade. They are of interest because of the large variety of organic molecules that can be placed in the perovskite framework. The formate family of hybrids, where a formate molecule is placed on the X-site, need further investigation. However, there are a large number of formate hybrids that need further investigation. We choose formate perovskites that have been grown experimentally to investigate further using density functional theory. We predict the structural parameters and polarization of nearly 20 hybrid perovskites. We also compute the piezoelectric coefficients for a few of them. To our knowledge, we do not know of any predictions of the piezoelectric constants for the selected hybrids, allowing our study to fill in this gap. This work aims to develop and employ computational techniques in order to study conventional and emerging ferroelectrics. We aim to achieve a fundamental atomistic understanding of conventional and emerging ferroelectrics, as well as to predict their structural, electrical, electronic, and energy converting properties. Finally, we use computations to explore and propose novel functionalities of conventional and emergent ferroelectrics.

## Scholar Commons Citation

Kingsland, Maggie, "First-principles-based Modeling of Energy Converting Properties of Conventional and Emerging Ferroelectrics" (2022). *USF Tampa Graduate Theses and Dissertations.*

https://digitalcommons.usf.edu/etd/9389