Is the silica functionalized with amino groups? Are you reducing the bond between aldehyde and amino groups? Are you reducing the free aldehyde groups from glutaraldehyde? The toxicity comes from free aldehyde groups. 

Thanks for your reply. Yes silica is functionalized with amine groups. The 20 mg of amine functionalized silica is treated with 500 microlitre of glutaraldehdye and then sonicate for 20  minutes. Then glutaraldehyde treated amine functionalized silica added to 10 ml of chitosan solution (20 mg in 2% glacial acetic acid) and then stir at room temperature for 24 hrs. Then do centrifugation at 14000 rpm for 10 minutes to get the chitosan coated silica as a pellet. My main doubt is that how much amount of glutaraldehyde can be used to cross link amine functionalised silica with chitosan. How to remove free aldehyde groups?

I am not familiar with you particles but GTA (glutaraldehyde) has to be in excess otherwise you could crosslink your particles (if that is possible). If you know the surface area of your particles you could estimate the number of amino groups and then add an excess of GTA to ensure that each amino group reacts with a single GTA molecule. Run this reaction using phoshate buffer at pH 7 in the presence of sodium cyanoborohydride. Wash your particles to remove excess GTA. Afterwards add your chitosan solution containing sodium cyanoborohydride again (the reaction will be slow and inefficient due to the acidic pH), wash you particles with your acidic solution and add either a blocking solution (containing tris or other molecule with amino groups) or a solution containing sodium borohydride (to convert the remaining aldehyde groups into alcohol groups). 

Thanks for the answer