Machine-learning simulations reveal true hydrophobic nature of graphene–water interface
A research team at Korea's Institute for Basic Science (IBS) recently resolved a long-standing question in graphene science - whether the material intrinsically attracts or repels water - by combining machine-learning-enhanced molecular dynamics simulations with vibrational spectroscopy modeling.
The interaction between graphene and water underpins applications ranging from filtration membranes to nanoelectronics, yet its intrinsic wettability has remained controversial. Experimental observations have been inconsistent: in some cases, water droplets bead up (hydrophobic behavior), while in others they spread (hydrophilic behavior). This led to the widely discussed concept of “wetting transparency,” in which atomically thin graphene was assumed to transmit the wettability of its underlying substrate.



