Investigation on the influence of doping on ferroelectricity of hafnium oxide thin film grown using Pulsed Laser Deposition (PLD)

Ferroelectric materials have a wide range of applications including memory devices, energy harvesting, negative capacitance systems, etc. A well-known ferroelectric materials are perovskite structured, such as (Pb,Zr)TiO_3, known as PZT. Its performance at ambient temperature is satisfying, yet it has limitation in application to non-volatile memory devices due to scalability, complexity, CMOS compatibility, etc. Therefore, simpler and benign ferroelectric material is needed.

Hafnium oxide (HfO_2), on the other hand, is of great interest due to its CMOS compatibility as it is already in use as gate dielectric, scalability as it could be fabricated as thin film with thickness of a few nm and most importantly, ferroelectricity. It shows ferroelectric behaviour with specific phases, orthorhombic and rhombohedral. They are metastable phases which requires specific conditions for stabilization such as growth condition, post-annealing, induced strain, doping, etc., and it is yet to be confirmed which is dominating. HfO_2 thin film has high coercive field, larger than 1 MV cm^(-1), so that it requires high voltage for polarization switching. Also, after ferroelectric phase is formed, the films experience "wake-up" effect, which is increase in remnant polarization to certain number of field cycles. It is an intrinsic property that all ferroelectric materials experiences including PZT. Wake-up effect is unfavourable as once it is used for memory application, it might lead to misinformation storage. Additionally, it is hard to control the phases formed in polycrystalline films. A single crystalline ferroelectric phase is ideal to minimize contribution of non-ferroelectric phases and achieve high capacitance and low leakage. Epitaxial films grown on PLD system would enable formation of single phase, single crystalline thin film, that enables fundamental understanding of the material's intrinsic property.

In order to overcome aforementioned challenges, I aim to investigate separate effect of dopants and strain. I will learn to grow HfO_2 thin film with optimized ferroelectric behaviour with high saturation polarization, low coercive field, and reduction of wake-up effect.

By far, many dopants have been tested on HfO_2, yet there needs to be a clear understanding of the combination of dopant size, ion size variance, doping fraction and charge mismatch. This will be studied using co-doping. Main interest being Lanthanum and Tantalum dopants on HfO_2 which have 3+ and 5+ valence charge respectively. By co-doping them on HfO_2, the effect of average charge in the system and average cation ion size will be investigated separately.

Also, strain is crucial in controlling the lattice structures. In order to explore strain effect, superlattice structures of a few unit cells will be grown. HfO_2 film will be strained using different oxide lattice structures such as SrTiO_3. Superlattice will allow me to observe interface effect precisely using synchrotron methods. This will enable me to learn about chemical states and electronic states using XPS. We can also explore the influence of thickness and number of interfaces on structure formation and wake up effects.

All the films will be tested on Piezo-response Force Microscopy (PFM) and Polarization-Electric Field measurement to demonstrate ferroelectricity. Using Positive-Up Negative-Down (PUND) technique will eliminate influence of leakage current during ferroelectricity testing, allowing only displacement current to be in consideration.
As a result, the perfect structures with carefully tuned doping, strain and interfaces will enable understanding and control of the scientifically and industrially fascinating ferroelectric system of HfO_2. This will suggest the path to the next generation of nano-scaled electronics promoting CMOS performance and non-volatility.