Next Generation of Sensors and Low-Powered Memory Storage: Hafnia
Motivation: "If cloud and electronic devices operated on non-volatile memory the U.S. could trim $6 billion in energy costs annually."
Not all crystals are created equal. Some of them have special properties as a result of their non-centromeric geometry. Atoms in the crystal are oriented in a way that creates a dipole, or a net charge. When a collection of these crystals orients in the same direction they create a domain wall. This is a key characteristic of a ferroelectric material.
We can use ferroelectric materials to encode and store information in the form of non-volatile memory. Think: the tiny chips in your cell phone that allow you to store pictures and apps. Non-volatile memory, in comparison to random access memory (RAM), allows information to remain stored even when the power is turned off.
I study a recently discovered ferroelectric material, hafnia, that has shown promise to be an excellent candidate for non-volatile memory storage. Hafnia’s compatibility with Si-based devices, ease of fabrication, and large processing window makes it an especially exciting topic to researchers.
University of New South Wales (UNSW) – Sydney, Australia
Ferroelectrics are also piezoelectric, meaning that they generate an electric charge when mechanically stressed. In summer 2017 I traveled to UNSW in Sydney, Australia to study the piezoelectric response using a technique called Piezoresponse Force Microscopy (PFM) under the supervision of Dr. Jan Seidel. We found that the piezoelectric response increases at least 30% when hafnia is deposited on a flexible surface. This finding is significant because it opens the door to many novel applications for hafnia such as wearable sensors, self-charging electronics, and actuators.
I successfully first-authored my first publication on this work, “Enhanced piezoelectricity of thin film hafnia-zirconia (HZO) by inorganic flexible substrates” in the July 2018 issue of Applied Physics Journal. I have also had the opportunity to present this work in Hiroshima, Japan, at the Joint-ISAF conference and in Barcelona, Spain, at the Symposium on Ferroic Domains. In November 2018 my poster won 1st place at the Carolina Science Symposium in Raleigh, NC.
NC State University – Raleigh, NC
Now, I'm continuing my work from UNSW at NC State under the supervision of Dr. Jacob L. Jones in the Materials Science and Engineering Department. I work at the Nanofabrication Facility (NNF) where I grow ferroelectric crystals into thin films using a technique called atomic layer deposition. Growing my own thin films allows our lab to study the effects of important processing parameters on the film's electrical and ferroelectric properties. By understanding the kinetics of crystal growth, meta-stable phase formation, and the mechanical effects of strain engineering, we can build faster memory devices that require no power to store information.
As a solid-state chemist and process engineer, I use tools to make and detect things far beyond human limits.
Film Growth & Device Fabrication
I grow films via atomic layer deposition with various dopants such as Zr, Si, and La. I can change dopant concentration by manipulating the ratio of ALD cycles Hf:X. I then deposit electrodes for electrical characterization using lithography and sputtering. After my device architecture is complete, I perform rapid thermal annealing to crystalize the films into their final state.
Since my films are polycrystalline, I also use X-Ray Diffraction, Scanning Electron Microscopy, Spectroscopy, and Transmission Electron Microscopy to study their stoichiometry and structure. I can also use these tools to study defects and phase equilibria.
I study the structure of domain walls using piezoresponse force microscopy (PFM) which is a subset of atomic force microscopy (AFM). Using nanoscale resolution, I can look at polarization orientation, domain wall kinetics, and piezoelectric coefficient of films.
I also look at the electrically "useful" properties of my films like remnant polarization, coercive field, and leakage current. These allows me to determine if my films are truly ferroelectric and how effective they would be in a real electronic device.