About this Abstract |
| Meeting |
2027 TMS Annual Meeting & Exhibition
|
| Symposium
|
2D Materials – Preparation, Properties, Modeling & Applications
|
| Presentation Title |
Deep Neural Network Molecular Dynamics Modeling of Oxidized Graphene Aerogels |
| Author(s) |
John Crosby, Theodore Maranets, Haoran Cui, Yan Wang |
| On-Site Speaker (Planned) |
John Crosby |
| Abstract Scope |
Graphene aerogels offer tunable thermal transport, but the atomistic mechanisms governing heat conduction across their porous, chemically heterogeneous networks remain unclear. Here, deep neural network molecular dynamics is used to investigate pristine, edge-functionalized, and edge-and-surface-functionalized graphene aerogels. Equilibrium simulations show that functionalization increases bulk thermal conductivity by approximately two- to sixfold, with both conductive and convective contributions enhanced. Transient simulations further reveal that functionalization increases interfacial thermal conductance for representative edge-on-surface and cross-plane junctions, while per-atom heat-current maps identify functional groups as localized pathways for intersheet energy exchange. Green–Kubo modal analysis shows that functionalized aerogels exhibit globally elevated frequency-resolved thermal-conductivity spectra, particularly at low frequencies, together with substantial additional high-frequency contributions associated with low-mass hydrogen atoms. These results demonstrate that chemical functionalization enhances aerogel thermal transport by simultaneously strengthening interflake coupling and expanding the spectrum of heat-carrying vibrational modes, providing design guidance for thermally engineered lightweight porous graphene materials. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Computational Materials Science & Engineering, Machine Learning, Modeling and Simulation |