2026 Internship Opportunities

Carbon dioxide mineralization in seawater produces calcium carbonate (CaCO₃), a stable and abundant sink for atmospheric CO₂. However, magnesium carbonate (MgCO₃) also plays an important role in oceanic carbon cycling, particularly in regions with high Mg²⁺/Ca²⁺ ratios. Gaining molecular-level insight into the Mg²⁺-CO₃²⁻ ion pairing process is foundational for understanding both CO₂ mineralization in seawater and the kinetics of dolomite formation, a long-standing puzzle in geochemistry.

This project will employ machine-learning potential molecular dynamics (MLP-MD) to investigate Mg²⁺-CO₃²⁻ ion pairing in aqueous solution. Machine-learning potentials provide near-quantum mechanical accuracy at a fraction of the computational cost, enabling simulations of long-timescale molecular processes inaccessible to conventional ab initio quantum mechanics methods. The MLPs will be trained on density functional theory (DFT) data and, if time permits, further refined with embedded correlated wavefunction (ECW) theory for higher-level accuracy. The student will analyze free-energy surfaces and the structural characteristics to determine how hydration governs magnesium–carbonate interactions.

Building on our established workflow for Ca²⁺-CO₃²⁻ ion pairing using MLP-MD and ECW theory and combining with the Ca²⁺-CO₃²⁻ results, this study will provide a complementary perspective on how cation identity and solvation dynamics control carbonate ion pairing and crystallization. Together, these insights will advance the understanding of CO₂ mineralization in seawater and the fundamental origins of Mg-rich carbonate formation.

The student will participate in all stages of the computational study of Mg²⁺-CO₃²⁻ ion pairing in water. Under guidance, the student will:

• Set up and run first-principles DFT simulations on magnesium-carbonate aqueous systems to generate training data.
• Train a machine learning potential (MLP) using the DFT dataset and apply it to perform MLP-MD simulations of Mg²⁺-CO₃²⁻ ion pairing in solution.
• Analyze simulation trajectories and construct free-energy surfaces. If time permits, the student will explore ECW refinement and contribute to drafting a manuscript.

This internship aims to help the student build a strong foundation in computational chemistry. Specifically, the student will

• Learn the physical principles governing chemical reactions and molecular interactions in complex condensed-phase systems.
• Gain hands-on experience with modern computational methods, e.g., electronic-structure theory, machine-learning model development, and molecular dynamics simulations.
• Develop general scientific skills, such as data analysis, scientific communication, and presentation.

Previous experience with quantum mechanics/chemistry/physics is preferred.

Basic programming, scripting and data-handling skills (preferably in Python).

• We will provide training for running quantum chemistry calculations, machine learning potential training and molecular dynamics simulations on remote computational cluster(s). Students new to computational modeling are welcome. 
• Students are responsible for securing accommodations for the duration of their internship. On-campus summer housing is typically available the first week of June until the first week of August (details to become available in Spring 2026). If the internship falls outside of on-campus housing availability, students will need to secure alternate housing.