Modeling and optimization of wurtzite ferroelectrics in memory applications
Restricted (Penn State Only)
- Author:
- Baksa, Steven Michael
- Graduate Program:
- Materials Science and Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 12, 2024
- Committee Members:
- John Mauro, Program Head/Chair
Susan Trolier-McKinstry, Major Field Member
Jon-Paul Maria, Major Field Member
Susan Sinnott, Major Field Member
Ismaila Dabo, Chair & Dissertation Advisor
Adri van Duin, Outside Unit, Field & Minor Member - Keywords:
- Ferroelectrics
heterostructures
device scaling
energy efficiency
Ferroelectrics
Heterostructures
Device scaling
Energy efficiency
Ceramics
Semiconductors
Thin films
First principles - Abstract:
- Addressing climate change necessitates a holistic approach from developing carbon-neutral renewable energy resources to improving the energy efficiency of existing technologies, and especially of microelectronic devices. Although the von Neumann computer architecture has been the standard for microelectronic computing since the 1950s, circumventing inherent energy losses due to the "von Neumann bottleneck'' requires a paradigm shift in materials science and manufacturing for integrated circuits. Due to their reorientable, crystallographically-defined polarization, ferroelectrics are of practical interest for non-volatile data storage. Yet efforts to integrate conventional ferroelectrics into ultrathin memories have been frustrated by film-thickness limitations, which impede polarization reversal under low voltage. One appealing alternative is the wurtzite class of materials, including magnesium-substituted zinc oxide (Zn,Mg)O and boron-substituted aluminum nitride (Al,B)N, which have been shown to exhibit scalable ferroelectricity as thin films; however, the atomistic mechanisms for ferroelectric polarization reversal remain elusive. With the need to develop fundamental knowledge of ferroelectric reversal in wurtzite crystal structures, the goals of this dissertation are to determine the atomistic mechanisms for ferroelectric reversal in substituted wurtzite oxides and nitrides and to propose general design criteria to discover new and optimize existing ferroelectrics. First, the theoretical background of density functional theory is presented with a discussion of the Wang-Landau Monte Carlo technique, the nudged-elastic-band algorithm, and the ``modern theory of polarization''. Second, the computational framework for determining the thermodynamic stability of multi-component systems which includes first-principles calculations, fitting of the cluster expansion model, and performing Monte Carlo simulations at finite temperature. Third, atomistic mechanisms of ferroelectric switching in wurtzite oxides are presented with a focus on Mg-substituted ZnO. Strain fluctuations from cation environments reach up to 1.5% in ε⊥ and –2.5% in ε‖, driving ferroelectric polarization reversal through sequential switching and domain wall formation and motion. Co-validation from computation and experiment offers conclusive evidence that strain fluctuations can effectively promote ferroelectricity in wurtzites. Fourth, atomistic mechanisms of ferroelectric wurtzite nitrides are presented with a focus on AlN-based systems, including (Al,B)N, (Al,Sc)N, and (Al,Sc,B)N. Computation and experiment demonstrate that AlN engages in uniform switching while (Al,B)N engages in sequential switching with improvement in (Al,Sc,B)N for ferroelectric random-access memory applications. Finally, future directions on the dissertation research are presented.