Postdocs cracking open a black box of catalysis

Inside a reactor at SLAC National Accelerator Laboratory, something unexpected was happening. 

Materials chemist Kang Rui Garrick Lim had loaded a stainless steel tube of powdered catalyst into a reactor, sealed it, and filled it with carbon dioxide and hydrogen at more than thirty times atmospheric pressure and over 500 degrees Fahrenheit. Inside the reactor, which was built from scratch for catalytic reactions involving CO₂, his catalyst was turning the CO₂ and hydrogen into methanol.

"I made a lot of methanol, and I was very happy," says Garrick, who joined Stanford as a Stanford Energy Postdoctoral Fellow in 2025. "But I had no clue how or why."

Garrick’s catalyst is a combination of two metals. When he tested each metal individually, neither made methanol, and instead produced less useful chemicals (methane and carbon monoxide). But together, they generated methanol at rates competitive with those of the industry-standard catalyst. Methanol is one of the world’s largest commodity chemicals, Garrick explains, a building block for countless consumer and industrial products (like paint and glue), and increasingly added to gasoline and other fuels. Turning CO₂ into methanol, he says, recycles a greenhouse gas that would otherwise harm our planet into a valuable product. 

Understanding why Garrick’s two-metal catalyst makes methanol was the next step. "I can see the materials I make before. I can see the materials after," Garrick says. "But I cannot see what happens inside the sealed reactor. And during is where the action is."

Enter Zisheng "ZZ" Zhang, a theoretical chemist who joined the same fellowship in 2024. ZZ, who earned his PhD at UCLA, builds computational models that simulate what happens inside reactors at the atomic level. When Garrick presented his methanol puzzle at one of the fellowship's monthly “lab crawls,” routine gatherings that bring fellows into each other’s research spaces, ZZ saw an opportunity. The two began collaborating to crack open one of catalysis's biggest questions.

"The fellowship is built on the premise that breakthroughs often happen at the intersection of disciplines. Bringing an experimental and a computational chemist into the same room creates the opportunities that turned Garrick and ZZ's separate projects into a shared discovery," said Prof. Yi Cui, who founded the fellowship at the Stanford Doerr School of Sustainability’s Precourt Institute for Energy in 2022 and remains its faculty director. 

Why catalysts matter

A catalyst is a material that speeds up a chemical reaction by lowering the amount of energy necessary to initiate the reaction. Catalysts are not consumed in the reactions and can, in principle, be used over and over again. They help make more than 90% of industrial chemicals produced today, including methanol. 

Nearly all methanol produced globally starts with natural gas. Through steam reforming, the natural gas is converted into a mixture of CO₂, hydrogen, and carbon monoxide, which then react over industrial catalysts comprising three or more metals to produce methanol. This process releases roughly 165 million metric tons of CO₂ annually, equivalent to emissions from 36 million cars driven for a year.

A new catalyst that requires less energy and uses CO₂ directly for methanol production could make methanol production faster, cheaper, and less carbon-intensive. Understanding why Garrick’s two-metal catalyst works could point the way to improved methanol production and better catalysts for CO₂ reactions across the chemical industry. 

To understand a catalyst, researchers zoom in to its surface. “How do the metal atoms arrange and rearrange themselves under the harsh reaction conditions? How do CO₂ and hydrogen molecules bind to them? Which metal surface structures are key to driving the reaction forward with the lowest possible energy penalty? Those are the questions we ask,” says ZZ.

To answer those questions only with lab experiments, Garrick would need to test different metals, ratios, and conditions in many thousands of combinations.

"If I had a million copies of myself, maybe I could test everything experimentally," Garrick says. "But I don't, and I won’t because there must be a more efficient way."

ZZ offered an only-one-Garrick path. 

Living inside the oven

ZZ uses machine learning to speed up simulations of what happens inside reactors under real operating conditions, capturing the irregular, complex, and shifting reality of catalysts at the atomic level. 

"For the past 20 years, most of the field has been simulating catalyst surfaces as perfect, orderly lattices," ZZ says. "In reaction conditions, the lattice messes up. You have different molecules binding to the surface in different amounts." 

Garrick likens their collaboration to baking. "I prepare the ingredients. I see the finished cake. But if I don't know what happens in the oven, I can't make an intelligent guess on how to make the next, better cake," he says. "ZZ lives inside the oven. He simulates what goes on, but for thousands of cakes at once."

After every experiment, Garrick tells ZZ what went into the reactor and what came out. ZZ's simulations map the range of structures the catalyst likely takes on under those conditions. Together, they piece together how the catalyst behaved during the reaction.

"I'm very annoying. I ask a lot of questions,” ZZ says, laughing. "When Garrick sends me new results, we don't wait. We look at it together and decide what to do next to debunk or corroborate our hypothesis."

The iterative loop runs fast because they work so closely together. Garrick and ZZ often begin conversations on the bus between campus – where Garrick synthesizes catalysts in Prof. Matteo Cargnello's lab – and SLAC National Accelerator Laboratory, where they both test and model reactions at the SUNCAT Center for Interface Science and Catalysis, a joint Stanford-SLAC effort. Prof. Thomas Jaramillo, who directs SUNCAT, advises them both. ZZ will call Garrick to his desk to look at data. In turn, Garrick brings ZZ to the reactor to view the experimental results in real time and adjust reaction conditions to test their renewed hypotheses. 

"First, you find the observation," Garrick says. "Then you understand why. Then you make the next iteration."

Built for collaboration

"Garrick and I wound up working together on efficiently producing methanol from CO₂ and hydrogen, but we also need people figuring out how to economically produce green hydrogen and capture CO₂, the latter of which may be solved by policy instead of technology," says ZZ.  "This underscores the urgent need to integrate energy researchers across campus and beyond. That's what the energy fellowship seeks to achieve – bringing together the brightest, early-career energy researchers across disciplines to achieve outcomes beyond the sum of our individual parts."

Both researchers are about to embark on independent faculty research careers. ZZ has joined Stanford's Department of Chemical Engineering as an assistant professor this fall and is building up his research group. Garrick will join Nanyang Technological University in Singapore as an assistant professor next year, but their collaboration will continue.

"We're friends, then colleagues, then collaborators," Garrick says. "We'll get our students to work together. And it all started here in the Stanford Energy Postdoctoral Fellowship program."

For now, they are inching closer to understanding their new catalyst. They have written up and submitted for publication – hopefully early next year – what they have learned about the multi-step chemical reaction of producing methanol from CO₂ and hydrogen. With each iteration, they narrow existing possibilities while finding new unexpected ones. And then the loop goes on.

Stanford Energy Postdoctoral Fellowships are awarded for up to three years to allow recently minted PhDs to explore new avenues of research, which is difficult to do in the more common two-year postdoctoral tenure. This timeframe benefited Garrick, who earned his doctorate at Harvard.

“I never worked on methanol or CO₂ before. In fact, I never worked with such a big reactor before! This is all new and has inspired me to do similar work in my upcoming lab in Singapore,” said Garrick.