Jihan Kim

Timeline

1. 2001–2009

   Degrees at Berkeley and Illinois

   He earns a B.S. in electrical engineering and computer sciences at the University of California, Berkeley, in 2001. He then completes an M.S. in 2004 and a Ph.D. in 2009, both in electrical and computer engineering, at the University of Illinois at Urbana-Champaign.

2. 2009–2013

   Postdoc at Berkeley Lab

   He works as a postdoctoral researcher at Lawrence Berkeley National Laboratory, in the Computational Sciences Division from 2009 to 2011 and in the Materials Sciences Division from 2011 to 2013.

3. 2019

   Algorithm-designed MOF composites

   In 2019, he coauthors a Nature Communications paper whose joint computational and experimental workflow screens thousands of metal-organic frameworks for pairs that connect seamlessly. Six predicted pairs, including HKUST-1@MOF-5, are synthesized as single crystalline MOF@MOF.

4. 2020

   A GAN for zeolite design

   In January 2020, KAIST announces that his research team has used AI to inverse design porous materials with desired properties. The team trains a generative adversarial network on three-dimensional grid data from molecular simulations and generates zeolites with structures the AI had not learned, aiming at energy storage materials such as gas adsorption materials.

5. 2024

   ChatMOF in Nature Communications

   In 2024, he and Yeonghun Kang publish ChatMOF in Nature Communications, an AI system built on GPT-4 and GPT-3.5-turbo that predicts and generates metal-organic frameworks. With GPT-4 it reaches accuracy rates of 96.9% for searching, 95.7% for predicting, and 87.5% for generating tasks.

6. June 2025

   Screening MOFs for air capture

   In June 2025, a team he leads, working with Imperial College London, reports in Matter a machine learning force field for the interactions of CO2 and water with MOFs. The method screens over 8,000 experimentally synthesized MOF structures and identifies more than 100 promising candidates for CO2 capture.

7. September 2025

   Quantum computing for porous materials

   In September 2025, a KAIST team he leads reports using quantum computers to design multivariate porous materials, with the findings published in ACS Central Science. The team encodes the framework structures as qubits and synthesizes four predicted structures, and the experimental results match the simulations.

8. As of October 2026

   Professor at KAIST

   He is a professor in the Department of Chemical and Biomolecular Engineering at KAIST, with research interests in computational chemistry, nanomaterials simulations, machine learning and AI, and materials design. He is also an adjunct professor in KAIST's AX faculty, with generative AI and inverse design among his core AI competencies.