Metal–organic frameworks (MOFs) have attracted attention as drug carriers in drug-delivery systems (DDSs). MOFs, which are composed of metal ions and organic linkers, exhibit high porosity, tunable pore structures, and large surface areas, making them promising DDS carriers. Molecular simulation methods are powerful tools for investigating drug-loading mechanisms and predicting the drug-loading amount. However, a simulation method for liquid-phase drug adsorption under the appropriate conditions has not yet been developed. In this study, a combined canonical and grand canonical Monte Carlo (CMC-GCMC) approach is developed to simulate drug and solvent adsorption under realistic solution conditions. The fugacities of the drugs and solvents, which correspond to their chemical potentials, are determined using the CMC with the Widom insertion method. The fugacities are then applied to GCMC simulations for drug and solvent adsorption. The proposed approach successfully reproduces experimental drug-loading behaviors for various MOF–drug–solvent systems, such as ZIF-8, UiO-66-H, UiO-66-NH2, and MIL-101(Cr). We also investigate the effect of the solvent on drug loading for ZIF-8 by analyzing the molecular interactions, radial distribution functions, and spatial probability distributions. Hydrogen bond-like interactions and solvent-dependent configurations govern drug stabilization within MOF pores. These results highlight the critical role of solvent properties, such as polarity and molecular size, in determining the drug-loading amount. This work provides a quantitative and molecular-level understanding of MOF–drug–solvent interactions and establishes a versatile computational strategy for designing MOF-based drug carriers with optimized loading and release performances.
Ohshima et al. (2026) studied this question.