(2gq) Machine Learning-Assisted Multiscale Modeling for Materials Design

Lightweight and sustainable high-performance electronic devices serve as key drivers of innovation and efficiency across different fields such as automotive, telecommute healthcare, and beyond. In my lab, we will focus on the design of new advanced materials for the electronics industry applications using integrated computational materials engineering and artificial intelligence through understanding the fundamental science in nanoscale. My research experiences have equipped me with a deep understanding of different materials and computational techniques. I see three main research areas where I can apply my expertise.

Design new MOF-polymer materials for gas sensors material.

Hazard gas sensing involves the detection and quantification of specific hazard gases in the environment, providing essential information for safety, and environmental monitoring in the chemical industry. To improve gas sensitivity and selectivity, MOFs are a class of porous materials consisting of varieties of metal ions and organic ligands, which exhibit extraordinary potential for gas adsorption/selection due to their porosity and structural tunability. However, MOFs have naturally low electrical conductivity, limiting their application as electrical gas sensors, resulting in low response time. To overcome this inherent limitation, one possible approach is to modify the MOF/MOF-derivatives with conductive polymers (CP) to form MOF-polymer composites. My research group will perform high-throughput screening of molecular dynamics simulations to create a dataset for the adsorption of the target gas using different MOF-polymer composites. With the limited experimental data on MOF@CP, ML models could be used to design possible candidate composites with enhanced electrochemical properties. I hypothesized this study would leverage the superior attributes of both MOFs and conductive polymers, leading to new materials with enhanced sensing capabilities with gas selectivity and response time in different operating conditions.

Studying the effect of long-period stacking order (LPSO) on deformation behavior in Mg-Y-Zn alloy.

The Mg-Y-Zn alloys with LPSO structures have shown superior strength and lightweight characteristics because of which they find applications in the automotive, aerospace, and electronics industry. However, a fundamental understanding of the effect of LPSO on deformation behavior such as dislocation, twinning, and their interactions on improving the alloys' mechanical in the nanoscale is still lacking. Though in-situ TEM has been used to study the interaction between basal dislocation with LPSO, the energetics related to this interaction, as well as interaction with non-basal dislocation are difficult to study through experiments. However, in the atomistic simulation aspect, the face-centered cubic (FCC) LPSO easily transforms back to hexagonal close-packed (HCP) structures during relaxation with the currently available interatomic potentials. I will develop a new machine learning potential to effectively describe the LPSO. Upon the successful validation of the new potential, it will be applied to study the deformation behavior of Mg-Y-Zn alloys and reveal how the LPSO contributes to the improvement of mechanical properties at the atomistic level, which might make significant strides towards lightweight alloy desire for electronics applications.

Design new high entropy oxides with enhanced magnetic and electronic properties.

High entropy oxides (HEOs) are multi-component systems to exhibit tunable electronic and magnetic features due to their intrinsic disorder and randomness. The application of these materials linked to these properties could be magnetic storage and quantum computing. By changing the selection and ratio of the metal elements as well as the structure of the oxides, the target design properties could be achieved. However, the large design space of HEOs makes it difficult to design new HEOs through trial and error. Therefore, my group will perform the first principle calculation to create the HEO dataset. Then apply machine learning with optimization methods to accelerate the new HEOs with optimal magnetic and electronic properties.

Research Experience

My academic journey commenced with a bachelor’s degree in polymer engineering. Subsequently, during my master’s program, I conducted laboratory experiments in heat treatment for steel. My Ph.D. work focused on designing new lightweight magnesium alloys with enhanced properties through understanding the deformation behavior using multiscale modeling. Currently, in my postdoc work, I have been working on the design of functionalized metal-organic frameworks for gas adsorption/separation using high-throughput screening. Additionally, I am also working on developing the hybrid modeling approach for the design of high entropy alloys and high entropy oxides with enhanced properties by integrated atomistic simulation, optimization algorithms, and machine learning, due to the tremendous design space.

Teaching Interests

Drawing from my personal learning and teaching experiences, I believe in the philosophy that “Passion drives learning”. During my two-year teaching as a chemistry instructor at an International Baccalaureate World School reaffirmed this belief, to spark curiosity and ignite passion, I consistently engaged students with interactive chemistry experiment demonstrations rather than dedicating prolonged periods to equation derivations or theory explanations. I found the truth in the adage, “The proof of the pudding is in the tasting”. The hands-on experiences, such as conducting experiments or coding tutorials, resonate more deeply with students.

This teaching journey afforded me invaluable insights into the concerns, academic challenges, and cognitive processes of high school students, especially those transitioning into university life. I've carried these insights forward into my mentoring roles during my Ph.D. and Postdoctoral periods, where I guided both undergraduate and graduate students in conducting computational modeling to understand the structure-property relationships of materials. I've realized that a combination of academic professionalism, patience, effective communication, and mutual respect cultivates an atmosphere of productive mentorship and efficient learning. I'll keep this philosophy in the classroom, ensuring an optimal learning environment that encourages collaboration and mutual growth.

As I look ahead, I am eager to impart my experiences and expertise to those young talents in the field of chemical engineering or related majors. I am open to teaching any course that the Department needs me to teach. However, given my background in multi-scale modeling and machine learning, I would be comfortable teaching core courses on “Thermodynamics”, “Transport Phenomena”, etc. I will not just ensure students understand the concepts but also promote a culture of critical examination and idea connection. Besides, I would like to develop specialized courses such as “Computational Modeling for Materials Design” or “Applications of Machine Learning in Materials” to provide students with a firm understanding of cutting-edge AI algorithms and enhance their competitiveness in this “AI era”. Helping students transition smoothly from university to academia or industry and succeed in their chosen path is a responsibility I deeply value as an educator. With this in mind, I will continually improve my teaching methods to address student concerns and create a conducive learning environment that encourages curiosity, deep understanding, and critical thinking.

I commit to creating a nurturing academic environment, in which Diversity, Equity, and inclusion (DEI) are of great importance. To foster diversity, I plan to actively participate in the university’s recruitment efforts, collaborating on workshops and open days, to attract and engage underrepresented groups, women, and differently-abled individuals in the field of chemical engineering. And I will try my best to provide every student with resources and opportunities for their success, including one-to-one conversations to help sessions. I will actively seek feedback from my students and adapt my teaching methods to accommodate different learning styles. For me, DEI is not a static goal but a long-term dynamic process. Therefore, I will also engage in discussions with colleagues and be open to new advice and approaches to advocate DEI for all young talents.