Date of Graduation


Document Type


Degree Name

Doctor of Philosophy in Physics (PhD)

Degree Level





Huaxiang Fu

Committee Member

Laurent Bellaiche

Second Committee Member

Hameed Naseem


Defect, Ferroelectricity, Hyperferroelectricity, OCBC, Vacancy


This dissertation presents three projects that investigate the complex phenomena of ferroelectricity under different conditions in BaTiO3, LiNbO3, and LiZnSb using first-principles density functional calculations. Extended defects in ferroelectric solids play a crucial role in reducing the lifetime and performance of ferroelectric devices by causing fatigue, domain pinning, and aging. Thus, understanding their impact is of critical importance for the development of reliable and high-performance ferroelectric devices. In addition, hyperferroelectricity is an intriguing phenomenon that has attracted much attention in recent years. Despite the existence of depolarization field, spontaneous polarization persists under an open-circuit boundary condition (OCBC), making hyperferroelectric materials promising candidates for various electronic applications. In the first project, we investigate the energetic, structural, and electric properties caused by extended planar oxygen vacancies in ferroelectric BaTiO3. Oxygen vacancies of different charge states, namely V_O^(2+), V_O^(1+), and V_O^0, are studied. We find that, while planar V_O^q vacancies of all three charge states may take place under the oxygen-poor condition, planar V_O^(2+) vacancies can still occur even under the oxygen-rich condition. Furthermore, our calculations show that BaTiO3 with planar V_O^0 or V_O^(1+) vacancies is metallic, while BaTiO3 with planar V_O^(2+) vacancies is an insulator. This finding reveals an important conclusion that, for BaTiO3 with planar V_O^(2+) vacancies, no mobile charges are available to screen the polarization, electric polarization is thus well defined and can be rigorously computed. Moreover, we discover that extended planar V_O^(2+) vacancies produce a devastating effect on ferroelectricity by drastically reducing the electric polarization, which is markedly different from isolated V_O^(2+) vacancies (which are benign to ferroelectricity). The polarization dead layer caused by planar oxygen vacancies is shown to be around 72Å. In the second project, we investigate the hyperferroelectric properties of LiNbO3. We find that (i) the longitudinal-optic A2u(LO1) phonon is soft with imaginary frequency 96i cm−1 when centrosymmetric LiNbO3 is under OCBC. (ii) The A2u(LO1) phonon can yield, under OCBC, a free-energy minimum with well depth of -9 meV at a nonzero polarization of 0.023 C/m2, thereby capable of producing hyperferroelectricity. (iii) The origin that A2u(LO1) can induce HyFE stems from the extraordinarily small mode effective charge of this phonon. (iv) Despite that A2u(LO1) can induce HyFE, we find the ground state of LiNbO3 under OCBC is not polar, revealing that the existence of a soft LO phonon does not guarantee HyFE. (v) We further show that LiNbO3 under OCBC may exhibit unusual tri-stable polarization states, with two potential wells of depth -23 meV and -1.9 meV. However, small polarization and shallow potential well depth are found so far in most hyperferroelectric materials, including LiNbO3, which severely limits their applications. In the third project, we report the discovery of a giant hyperferroelectricity in LiZbSb. The HyFE polarization is shown to be remarkably large with P=0.282 C/m2 under OCBC, which is one order of magnitude greater than the HyFE polarization in LiNbO3. Furthermore, HyFE in LiZnSb is found to be exceptionally stable with a well depth of electric free energy ∆F=-332 meV, which makes LiZnSb a possible hyperferroelectric solid at room temperature. The origin of the giant hyperferroelectricity in LiZnSb is attributed to the large mode effective charge of the soft longitudinal-optic phonon and the large high-frequency dielectric constant. The HyFE strain dependence in LiZnSb is also examined.