Go for high energy ultrasonic treatment to deagglomerate nanoparticles and measure Zeta potential 

As-made silica nanoparticles often indicate a high dispersibility in water or ethanol solvent. But when these nanoparticles were calcined at high temperature, the dispersibility greatly decreased and more serious problem is that these calcined silica nanoparticles can not disperse in water or other organic solvent in some cases. This result means high-temperature calcination markedly influences the dispersibility of silica NPs due to the aggregation of silica nanoparticles. Hence I suggest if as-made silica NPs contain surfactant you can use solvent-extraction method to remove surfactant, and make these NPs redisperse in solvent for storage and further use. Of course you have to determine concentration of NPs in a solvent.

Thank you very much for your answers! Let me ask some clarifications.

@Susi Kumar: Do you have any suggestions for the sonication parameters (e.g. intensity)? About the Zeta potential, is it relevant for silica NPs?

@Yucang Liang: I thought that calcination was the major issue. At the moment I am not using a surfactant, but a silane that bears a long organic chain. I can switch to the common surfactants (e.g. CTAB), but in this case: is solvent-extraction enough to remove the surfactant? I read that even in this case calcination is the most common technique.

Thank you again and sorry for bothering.

Sure. Calcination is quite often used to remove organic composite. But it is truth to make particles aggregate together and decrease the dispersibility of particles during calcination process. If you use silane containing long carbon chain as silica source, as-made silica NPs have higher dispersibility due to existence of long-carbon chain. I think if you obtain silica particles containing long-carbon chain and this long organic chain do not influence your next step reaction, you can directly use as-made silica particles. If you have to remove long organic chain from particles, it is good way to decrease calcined temperature to reduce the chance of aggregation. Of course as Susi Kumar's statement, you can use sonication method to redisperse your particle in an expected solvent and increase sonication time. In this case you have to make very fine powder.

Hi Alessio

I had some personal experience with the Stober method quite recently. Do you have an SEM/TEM image of your sample? The condensation may be occuring on inter-particle hydroxyl groups which means particles are covalently bonded and will sediment. From my experience centrifugation did not work and i obtained mine from dialysis and filtration (to remove non-coating silica). About calcination, i am not sure whether that temperature is enough to condense surface hydroxyl groups. If this is the case, it is no wonder why they are not dispersible anymore, you are forming big chunks of silica.

Good luck

Hello Georgios,

Thanks for your answer. Do you have a good preparative to suggest? What I would like to obtain is mesoporous silica NPs with diameters between 100 and 170 nm.

I do have some SEM images, however they were registered from samples of the suspension, where the particles size is appropriate. The precipitate instead is formed by a multitude of differently sized particles.

First of all i dont understand how you form straight away mesoporous silica shell by adding TEOS half-reaction. my guess is that you need to form solid particles first and then etch those to form pores.

How do you precipitate the reaction mixture? are you sure what you get is the particles and not random condensed silica? 

What i was doing is quite different from yours. You are making silica particles, i was coating stuff with silica, similar process, different reagents. There are tons of literature on silica. 

I didnt get what you did with the SEM, nevertheless a TEM will show whether particles are linked together and an SEM showed me non-coating silica (i couldnt observe these with either instruments alone). 

Hello again,

You are right: I think I got a bit muddler.

This is how I carry out the reaction:

A solution of EtOH, H2O and ammonia is prepared, then TEOS is added under vigorous stirring. After 1h a mixture of TEOS and hexadecyltrimethoxysilane is added dropwise. The stirring is continued for other 2h, then the mixture is centrifuged and repeatedly washed with ethanol (dispersion and centrifugation included).

The so obtained particles are firstly dried in air (heated at 50 °C on a hot plate), then calcined at 550 °C (6h) to remove the hexadecyl- chains.

This is what I do.

I am attaching a pic of a final product (here there was a double distribution).

What I was trying to say is that this SEM is only representative of the part that dispersed in ethanol. I also took a sample from the precipitate, and you can see in the second pic what I got (it is just an optical microscope, but you can still observe the uneven size distribution).

I wasn't sure whether it was possible to redisperse the particles or not (that's why I asked the question), but form yours and the previous answers I understand that they probably became covalently bonded.

Anyway, do you have any experience in coating polymer particles with silica? I found many opposite examples in the literature, but not this. In my case I would be interested in making a silica shell onto PMMA particles.

if you can see anything under an optical microscope then most probably the reaction is not as specific as they say. 

I was trying to coat polymeric particles with silica as well. I found that the alcohol used and water amount all affect the resulting particles. I had huge troubles with non-coating silica so i droped that pathway to my target. The first image doesnt say much, other than your yield is not great (nor was mine). There seems to be bonding between the particles from the TEM/SEM whatever that is. 

I will attach a couple of images showing 2 silication reactions. On the first you can see the sites where bonding occured and you can see the difference from the second where a sheet of silica coats many. I speculated then that the first one where particles are attached at some sites is caused post synthesis while the continuous sheet must have formed during the reaction. The method i followed is dialysis and lyophilisation (no calcination). Since you try to coat polymeric particles in silica i see no reason why you should use the hexadecyl whatever silane, i just used TEOS itself since you already have a template. 

My polymer had available amines though which could attack TEOS and have a site for growth (or not). You have to consider how your polymer behaves and readjust the reaction to your needs.

Sorry i have to change the image type to upload. I will do so asap