Science · 2010–2016 · Hsinchu & Berkeley

Watching materials grow

My research focused on how materials form and how their structure affects their properties.

I earned my PhD in Materials Science and Engineering at National Chiao Tung University, now part of National Yang Ming Chiao Tung University, in 2016. My advisor was Prof. Ying-Hao Chu, and my co-advisor was Dr. Haimei Zheng.

From 2012 to 2016, I was a visiting scholar in Haimei Zheng’s group at Lawrence Berkeley National Laboratory. There I used liquid-cell electron microscopy, which seals a thin layer of liquid inside the microscope so crystals can be filmed as they grow.

Iron oxide dendrite at 0.1 seconds — Fig. 01 — An iron oxide dendrite growing in liquid, at 0.1, 29.8, and 75.7 seconds. From my PhD thesis.

Iron oxide dendrite at 29.8 seconds — Fig. 01 — An iron oxide dendrite growing in liquid, at 0.1, 29.8, and 75.7 seconds. From my PhD thesis.

When tips split

Some crystals grow into branches, called dendrites. They also form on battery electrodes, where they damage the cell.

I filmed iron oxide and nickel ferrite dendrites from the first nanoparticles to fully branched structures. A tip widens before it splits, and faster-growing tips split wider.

Related paper: Nano Letters, 2018

A core, then a shell

The first question was which metals end up in the crystal. Nickel and cobalt grew into the crystals with iron, while manganese and zinc stayed in the solution. The outcome followed how easily each metal is reduced and how readily its starting compound breaks down.

I then mixed platinum with iron. With four times more iron than platinum, an iron–platinum core formed first. When the platinum ran out, an iron oxide shell grew around it. We recorded the process at atomic resolution.

Core–shell nanoparticles at 22.0 seconds — Fig. 02 — Core–shell nanoparticles forming in liquid, at 22.0, 49.6, and 125.3 seconds. Liang et al., *JACS*, 2015.

Core–shell nanoparticles at 49.6 seconds — Fig. 02 — Core–shell nanoparticles forming in liquid, at 22.0, 49.6, and 125.3 seconds. Liang et al., *JACS*, 2015.

Building the holder

To run electrochemistry experiments inside the microscope, I built a new sample holder. The group’s earlier holder had two electrodes and needed wire bonding for each experiment, which often broke the liquid cell.

With Haimei Zheng’s guidance, I designed a three-electrode holder for FEI microscopes, drew the 3D model, and worked with a machine shop to CNC-machine the parts. In testing, it gave a much steadier current signal than the earlier holder.

Three-electrode TEM holder on its stand, wired to an electrochemical workstation — Fig. 03 — The three-electrode holder connected to an electrochemical workstation. From my PhD thesis.

In the cleanroom

I made the liquid cells at the Marvell Nanofabrication Laboratory at UC Berkeley, working through lithography, etching, and thin-film deposition on silicon wafers. Some cells carried electrodes, so we could watch electrochemical reactions, including battery charge and discharge.

As the only PhD student in the group at the time, I contributed hands-on work across several projects.

UV exposure tool for photolithography under yellow cleanroom light — Fig. 04 — The UV exposure tool used for photolithography at the Marvell Nanofabrication Laboratory. From my PhD thesis.

Thin films

Before Berkeley, I spent about three years growing oxide thin films in Hsinchu, mainly BiFeO₃ and self-assembled oxide nanocomposites, using pulsed laser deposition. I also grew samples by molecular beam epitaxy for collaborators.

BiFeO₃ can form a nanoscale boundary between two crystal phases, where its electrical and magnetic response is enhanced. That boundary had been made on oxide substrates. We made it on silicon, a step toward using it in semiconductor devices, and I mapped how the phases develop as the film gets thicker.

Cross-section of BiFeO₃ on silicon and the phase boundary at atomic resolution — Fig. 05 — BiFeO₃ on silicon in cross-section, and the phase boundary at atomic resolution. Liang et al., *Nanoscale*, 2016.

Publications

The research resulted in 29 journal papers, including four first-author papers.

I also co-authored three papers in Nature Communications and three in Nano Letters.

Google Scholar, October 2026: 2,044 citations · h-index 22 · i10-index 25

Selected first-author papers

In situ study of Fe₃Pt–Fe₂O₃ core–shell nanoparticle formation
Journal of the American Chemical Society, 2015
In situ study of spinel ferrite nanocrystal growth using liquid cell TEM
Chemistry of Materials, 2015
Epitaxial integration of a nanoscale BiFeO₃ phase boundary with silicon
Nanoscale, 2016
Design of magnetoelectric coupling in a self-assembled epitaxial nanocomposite via chemical interaction
Journal of Materials Chemistry C, 2014

PhD thesis

The Growth Kinetics Study of Functional Nanostructures using In-situ Transmission Electron Microscopy in Solution
National Chiao Tung University

Full publication list on Google Scholar ↗

From research to a hardware startup

In 2014, while I was still at Berkeley Lab, I began helping a hardware startup in the Bay Area. That work led into the next part of my career.

The Elos story →