Molten salt is regarded as one of most promising candidates for solar energy storage due to possessing stable properties and large energy storage densities. However, the intrinsically low thermal conductivity of molten salt has become a bottleneck for rapid heat storage and transport. The addition of nanoparticles is generally considered to be a most effective way to improve the thermal conductivity of molten salt phase change materials (PCMs), while the phase change enthalpies of the nanocomposite phase change materials usually show two opposite trends of enhancement or decrement. Furthermore, the reason for the abnormal change of phase change enthalpy has not been clear in the literature so far, so the mechanism of change needs to be further explored. In this work, graphene nanosheets (GNS)@NaNO ₃(sodium nitrate) nanocomposite phase change materials are prepared by the hydration ultrasonic method. The materials are characterized by scanning electron microscope, and the phase change characteristics are measured using differential scanning calorimeter. Molecular dynamics simulation is carried out to explain the mechanism for the formation of the NaNO ₃dense layer and the non-collateral decrease of the enthalpy from the microscopic level. With the increase of GNS mass fraction, the melting point of the GNS@NaNO ₃composite phase change material decreases slightly while the phase change enthalpy decreases significantly with a non-colligative trend. A 13.81% decrease of the theoretical phase change enthalpy is observed with a GNS doping ratio of 1.5%. The NaNO ₃clusters observed on the surface of GNS are considered to have not melted, thereby resulting in a reduction in the phase change enthalpy. The mechanism is further investigated by molecular dynamics simulation, showing that the strong van der Waals attraction between GNS and NaNO ₃leads the 2–4 Å-thick NaNO ₃dense layer to form in the vicinity of GNS. With the increase of GNS mass fraction, the centroid equivalent distance between the dense layer and GNS gradually increases, which leads their mutual attraction to first increase and then weaken. When GNS mass fraction is 1.5%, the centroid equivalent distance reaches the position closest to the potential well, leading to a strongest mutual attraction. In other words, the phase change enthalpy decreases most obviously at this mass fraction. Thus, some conclusions can be drawn as follows. The type of interaction between molten salt and nano-enhancers and the position of the potential well are the fundamental reasons for the thickness of molten salt dense layer and the reduction of phase change enthalpy. The calculation of the interaction energy can be used to guide the selection of the mass fraction of the nano-enhancers, so as to avoid the loss of core material cluster and phase change enthalpy caused by the introduction of the nano-enhancers to a greatest extent. The preparation cost of the composite phase change material can also be reduced to a certain extent.