PhD candidate in Engineering Thermodynamics (Process & Energy, Faculty of Mechanical Engineering) at TU Delft.

I work on the thermophysical properties and phase behavior of CO₂ mixtures with impurities — the kind of question that decides how you actually design a CO₂ transport pipeline or a sequestration well. My toolkit spans molecular simulation, classical density functional theory, equations of state, and machine learning — most recently, ML surrogate models that predict interfacial tension of impure CO₂ roughly 1000× faster than cDFT.

• Molecular simulation (Monte Carlo & molecular dynamics) of CO₂-rich mixtures
• Vapour–liquid equilibria and interfacial tension of multi-component mixtures
• Equations of state: PC-SAFT, SAFT-VR Mie, cubic EoS
• Classical density functional theory (cDFT) for interfacial properties
• Entropy scaling and finite-size effects in transport/thermodynamic properties
• Machine learning surrogates for phase equilibria and interfacial properties (TabPFN, Gaussian processes, symbolic regression, active learning)
• Carbon capture, transport and storage (CCS)

• Machine Learning-Based Prediction of Phase Equilibria and Interfacial Properties of Multicomponent CO₂ Mixtures with Impurities: Application to CO₂ Transportation — Ind. Eng. Chem. Res. (2026, ASAP) · doi:10.1021/acs.iecr.6c03388
• Vapor–Liquid Interfacial Properties of CO₂ Mixtures for Sequestration Applications: Molecular Simulations, Classical Density Functional Theory, and Equations of State — Ind. Eng. Chem. Res. (2026) · doi:10.1021/acs.iecr.5c04932
• Finite-size effects of the excess entropy computed from integrating the radial distribution function — Molecular Physics (2025) · doi:10.1080/00268976.2025.2456115
• Thermophysical Properties and Phase Behavior of CO₂ with Impurities: Insight from Molecular Simulations — J. Chem. Eng. Data (2024) · doi:10.1021/acs.jced.4c00268

Full list on Google Scholar.

• co2-impurities-ml-surrogates — code behind the ML surrogate paper: PCP-SAFT + cDFT workflows, kij fitting, and the TabPFN / Gaussian-process / symbolic-regression surrogates with active learning
• CO2-Impurities-ML-Supporting-Data — PC-SAFT/cDFT datasets and ML supporting data for the ML surrogate paper on phase equilibria and interfacial properties of multi-component CO₂ mixtures with impurities
• entropy-scaling — excess entropy from integrating the radial distribution function, with the Wang–Frenkel potential
• thermoift — PC-SAFT thermodynamics and interfacial tension for N-component mixtures
• parachorpy — interfacial tension from empirical correlations, the Parachor model and Winterfeld–Scriven–Davis

TU Delft profile · Google Scholar

📫 d.raju@tudelft.nl · rajudarshan1997@gmail.com