If the filter has pores not larger than 220 nm, particles with a hydrodynamic diameter higher than this value should be retained by the filter. In addition, once a 'cake' (a layer of blocked particles) has been formed, also particles with a smaller dimension will be blocked, because the cake acts as a filter with lower porosity. As a result, you should find increasing difficulties in filtering your suspension.

Concerning the aggregation, it depends not only on the concentration of the suspended particles, but also on the ionic strength of the solution, and on the possible presence of surfactants. In order to limit the aggregation rate, you may consider to decrease the ionic strength (e.g. by diluting the solution) or to add suitable stabilizers. However, the latter approach will modify the surface properties of your nanoparticles.

If the filter has pores not larger than 220 nm, particles with a hydrodynamic diameter higher than this value should be retained by the filter. In addition, once a 'cake' (a layer of blocked particles) has been formed, also particles with a smaller dimension will be blocked, because the cake acts as a filter with lower porosity. As a result, you should find increasing difficulties in filtering your suspension.

Concerning the aggregation, it depends not only on the concentration of the suspended particles, but also on the ionic strength of the solution, and on the possible presence of surfactants. In order to limit the aggregation rate, you may consider to decrease the ionic strength (e.g. by diluting the solution) or to add suitable stabilizers. However, the latter approach will modify the surface properties of your nanoparticles.

Thanks, Sergio. But will the surface property of my nanoparticle affect the filtration efficiency? 

What do you mean with "surface properties"? I wrote that the recourse to suitable stabilizers may modify the surface properties of your nanoparticles: in fact, the stabilization is obtained through an interaction between the stabilizer and the surface of your nanoparticles, with the result that the nanoparticles will be covered by a monolayer (more or less continuous, it depends on concentration) of the added chemical specie. Then, if your nanoparticles are used for a given reason, you have to check whether the monolayer disturbs or not!

Thanks again. Actually, I mean that if I use the stabilizer to stabilize my nanoparticles, and then I filtered these particles with 0.22um filter, will the filtration change the properties of my NPs ( including the possible monolayer that covering the NPs) that can pass through the filter?

And, actually I think I have this question because I don't know the following one:

When we measure the hydrodynamic diameters of NPs in suspension with addition of stabilizer such as citrate, is the monolayer contribute to the particle diameter? 

I think that the monolayer practically does not contribute to the particle diameter: we are speaking about fractions of microns, while the "thickness" of a relatively big molecule like citric acid cannot be higher than 1-2 nm!

Regarding the filtering operation, theoretically it may allow the particles to enter in contact, thus facilitating the aggregation... but I think it is not easy to make predictions

Hello,

I just want to add that for soft particles the situation can get more complicated:

G. R. Hendrickson, L. A. Lyon, Angew. Chem., Int. Ed. 2010, 49, 2193-2197.

Here, particles could pass pores 10 times smaller then their original size...

Regards

if the particles do not change shape beside the nominal cutoff even smaller particles will be retained to some degree. Further one has to consider also interactions between particles and filter material. Ideally each, filter material and particles should love solvent more than each other. The smaller the particles the higher the effect of layer associated with particles.

Thanks Felix and Titus, I think the property of particles and interaction of particles and filter will have influence on the filtration process.  This is pretty interesting. 

Hi Xiaoyu Gao, you asked several questions and I thought I'd add my experience to this as well.:

1. Yes, there is a correlation between nominal pore size and hydrodynamic diameter. Ideally the two should be the same, in reality they are not always exactly the same but instead close to each other, and that is due to both filter quality and particle properties (for example, flexible particles can squeeze through smaller pores).

2. This depends on the filter quality and the particle properties. Even though the filter says 220nm there is a distribution of different pore sizes in the filter membrane. Most likely this will mean that some particles slightly larger than 220nm may still get through. Along the same line, non-spherical particles of for example oblate ellipsoid shape 400nm x 100nm x 100nm - which will have a hydrodynamic size larger than 100nm - will still "fit" through the filter.

3. Yes, the surface chemistry of the particles will certainly affect the filtration efficiency. In an extreme case, if the particle coating really likes the filter membrane material, not much will pass the filter, even though the particles are smaller than the nominal pore size.

4. If the aggregation was related to the production process, then "cleaning" the sample by filtration may be a permanent solution. If however there is an equilibrium between the aggregated and non-aggregated state, then removing the large aggregates will only be a temporary solution: the larger aggregates will reform. The same is true if the overall sample is aggregating over time: just filtering out the large components will most likely be a temporary relief only.

It may be possible to check a filtration membrane before and after to see if the surface was altered. Best regards, Ulf

http://www.materials-talks.com/blog/2015/01/06/surface-zeta-potential-what-it-is-and-how-to-measure-it/