Cetrifuge force depends on the radius of your centrifuge, thus rpm is never enough information. Anyway, increase centrifuge velocity and centrifugation time. If that does not work add some NaCl to reduce zeta potential. 

Adding NaCl as suggested by Omar will probably destroy the double layer and lead to irreversible segregation - but they'll certainly settle and be removable by centrifuge!  CuO has a high density, so your particles are small if centrifugation won't separate them.  If they're really small then 25000 rpm will probably not be sufficient depending on your make of centrifuge...

I know that aggregation could occur by adding nacl Alan. That is something to be determined (nacl concentration). That is how we purify hollow gold nanoshells that are tremendously stable. Obviously adding a high nacl concentration will aggregate the particles in no time. 

Cleaning via dialysis might be necessary if the sedimentation is not possible or if a aggregation have to be prevented. Okay it's more time consuming but works quite good for small to moderate sample amounts. 

Article Functionalized polymer colloids bearing primary amino groups

Article Tuning the relaxation behaviour by changing the content of c...

Thanks Omar for your response, I'll try adding some NaCl of several amounts to check the better concentration that I should add

Thanks Alan for your response, our instrument of centrifugation 6000 rpm max....I'll try resolving the problem by adding some NaCl...using different concentration 

Thanks Michael for your response, yes dialysis should be necessary, thanks for your orientation