It will better to centrifuge but after centrifugation it washed it with absolute ethanol then you will get powder form of nanoparticles

It will better to centrifuge but after centrifugation it washed it with absolute ethanol then you will get powder form of nanoparticles

I am not interested in powder nanoparticle...i need to have nanopartcles of size 6 nm or below in colloidal form...and for that i just need any experience to get shared with me if anybody has.

A centrifuge should work assuming that your particles of your required size are sufficiently less dense than the larger component, however as particle size decreases, Brownian and buoyant forces become more closely matched. 

In order to remove any large component of your colloidal dispersion, you could also filter your sample. Syringe filters of 20nm pore size are readily available commercially. 

Alexander Malm!! thank you sir.... can you please explain that if i want to have only nanoparticles of size say less than 6 nm  stabilized in the colloidal remain after centrifuge?? i mean if one can do it using centrifuge, is it possible??

(Although i agree with other suggestion to have syringe filters)

Alexander Malm, When we filter the dispersion with syringe filter does it not effect the morphology of nanoparticles?

You should use Stokes law to calculate how long to centrifuge to remove all particles of a given density larger than x nm. You should be aware that you always remove particles of the desired size. This will also partly the case for filtration.

Alternatively, you should be able to ease the task by selective flocculation. Larger particles are destabilised early than the smaller wanted fraction.