Nanoparticles in agglomerated state lose their outstanding properties; hence, it is essential to break them up prior to use and prevent their re-agglomeration. Even though there are several dry techniques to disperse nanopowders, none of them have been able to produce truly nanoscale aerosols so far. Here, we study de-agglomeration of dry silica nanopowder via a jet impactor-assisted fluidized bed (JIAFB). The particle size distribution of fragmented powders was characterized by in-line scanning mobility particle spectrometry (SMPS) and offline transmission electron microscopy (TEM). In order to ascertain the jet length and that the kinetic energy of particles is sufficient for de agglomeration, a CFD simulation was carried out. Both SMPS and TEM measurements imply that at a certain fluidization velocity, increasing the jet velocity shifts the particle size distribution towards smaller diameters, and at higher velocities the mode value reduced from 113 130 to 55-60 nm. However, the geometric standard deviation or degree of polydispersity rises from 1.5 to 2.0 by increasing the jet velocity up to 197 m/s, as it will increase the total superficial velocity and consequently entrainment of larger particles from the bed. In addition, the TEM results indicate that the range of individual particle sizes in the supplied nanopowder is wide; hence, increasing the geometric standard deviation can be an indicator of a higher level of agglomerate dispersion. (C) 2017 Elsevier B.V. All rights reserved.