Failure to replicate long-range tunable attractions in colloidal system

Recently, a temperature-sensitive long-range attraction colloidal system exhibiting melting and pre-melting of colloidal crystals has been reported by Li et al.[1]. If it is true, it would be an ideal system to simulate a board range of phase transitions happening in atomic systems. We have reanalyzed their data in [1] and have discovered great inconsistency in their radial distribution functions g(r) and pair potentials u(r). We have tried to reproduce their experimental results but failed to find the temperature-sensitive long rang attraction as reported by Li et al. Based on our experimental observations, We conclude that the melting and premelting reported in their paper are non-equilibrium phenomena caused by non-uniform temperature in the samples of Li et al..