Unveiling Carbon\nCluster Coating in Graphene CVD on\nMgO: Combining Machine Learning Force field and DFT Modeling

In this study, we investigate the behavior of carbon\nclusters (C<sub><i>n</i></sub>, where <i>n</i> ranges from 16\nto 26) supported on the surface of MgO. We consider the impact of\ndoping with common impurities (such as Si, Mn, Ca, Fe, and Al) that\nare typically found in ores. Our approach combines density functional\ntheory calculations with machine learning force field molecular dynamics\nsimulations. It is found that the C<sub>21</sub> cluster, featuring\na core–shell structure composed of three pentagons isolated\nby three hexagons, demonstrates exceptional stability on the MgO surface\nand behaves as an “enhanced binding agent” on MgO-doped\nsurfaces. The molecular dynamics trajectories reveal that the stable\nC<sub>21</sub> coating on the MgO surface exhibits less mobility compared\n to other sizes C<i><sub>n</sub></i> clusters and the flexible\ngraphene layer on MgO. Furthermore, this stability persists even at\ntemperatures up to 1100K. The analysis of the electron localization\nfunction and potential function of C<sub><i>n</i></sub> on\nMgO reveals the high localization electron density between the central\ncarbon of the C<sub>21</sub> ring and the MgO surface. This work proposes\nthat the C<sub>21</sub> island serves as a superstable and less mobile\nprecursor coating on MgO surfaces. This explanation sheds light on\nthe experimental defects observed in graphene products, which can\nbe attributed to the reduced mobility of carbon islands on a substrate\nthat remains frozen and unchanged.