摘要： Organic–inorganic halide perovskite solar cells have been in the limelight in recent years due to their outstanding power conversion efficiency (PCE) and facile synthesis. Despite of the ever increasing PCE that astonishes the solar cell community, many fundamental structural and theoretical aspects remain elusive, which inhibit our understanding on this fascinating material. Methylammonium iodide (MAI) is one important precursor to fabricate the halide perovskite materials based on the prototypical CH3NH3PbI3. Some MAI might remain in the CH3NH3PbI3 layer, forming a MAI-rich layer that is considered strongly affecting the perovskite solar cell performance. Therefore it is helpful to take one step back to re-evaluate the structures of MAI that is overlooked in many perovskite solar cell studies. In this manuscript, we identify the alternative structures of MAI aggregates via the ab-initio calculations, including the monomeric, dimeric, trimeric, tetrameric, pentameric, hexameric, heptameric, octameric and decameric structures of MAI. These aggregate structures are proposed to act as a simplified, yet efficient model to represent the MAI-rich region in perovskite solar cells that are usually simulated using the computationally expensive slab model. We find that the aggregates of MAI are universally stabilized by the hydrogen bonds. We further evaluate these MAI aggregate structures via studying the interactions between the MAI aggregates and an experimentally-proven additive iodopentafluorobenzene (IPFB) as an example, and show that the iodine atoms in the MAI aggregates acting as potential charge traps could be passivated by the additive molecule. The present study provides an efficient modeling platform for the MAI-rich halide perovskite materials without high computational cost.