Dialysis filtration - spontaneous process by diffusion molecules through the membrane. The cleaning process of the nanoparticles from impurity molecules is slow (24hours) . Why do you need this treatment? Which solvent? What is the size of nanoparticles? Your question is not clear. So you do not respond.
THe commercial cells I am aware of (eg, Amicon) use a stirred cylindrical chamber that contains the dispersion/suspension or solution you wish to dialyze. the base of the cylinder lies atop a membrane filter with a given molecular weight cutoff (MWCO). Solution passing through this filter can be collected with volume and contents analyzed if you wish; it is useful to track volume, as once you measure three time the volume of your original sample, you have probably washed your sample sufficiently. The top of the chamber is attached to a solvent source (eg, water). As solvent exits the chamber through the membrane, it creates a partial vacuum that draws in more solvent from the source; this allows one to operate at constant volume. Some models have simple swtchs that allow one to switch between diafiltration and ultrafiltration; in the ultrafiltration mode, with auxiliary pressure provided by a tank of N2 gas, one can decrease the volume to a desired level.
It is not clear to me that separating one type of nanoparticle from another can be done by diafiltration. The separation is done by a filter that has a given MWCO. It works for salts and small polymers, but you would have to have a dramatically differently sized pair of nanoparticles to separate by any filtration process. If one type is ten times the diameter of another, it may be feasible.