What is the best way to assay silicone dioxide?

Is this a 3 day late a
April Fool posting to the LC site?

No really I am not taking the mickey! We have had real problems trying to determine the assay for this material. Currently we make a complex using ammonium molybdate then measure the blue colour using UV absorption. The derivatisation reaction is horrendous and involves burning the sample in a Pt crucible overnight at 900C, digging the sample out, washing everything to ensure no losses then reducing the complex before analysis. It is fraught with potential errors! I realise now that this is not a chromatography question but if anybody knows of any useful links or could point me in the right direction I would be really grateful.

I don't know another method, but i vaguely remember running a method like this many years ago. I think one of the problems is that you need to cover the crucible because some of the SiO2 floats out on thermal drafts if you don't.

Question: Is this SiO2 in a product, or are you trying to do it as a "pure" material. I know that silicone oil can be done by IR. If I remember SiO2 also has specific bands.

Good luck.

Chris

Many thanks for your help Chris, we have noticed that you need to cover the crucible too, and also remember to wash the lid. Just to clarify we are trying to assay this as a bulk material. We use it as a raw material test to blend with another pharmaceutical ingredient. Relatively large doses of the SiO2 will be taken by the patient hence the necessity for the analysis. I will try the IR approach, thanks again.

I am still not clear on this need. Do you want to determine the purity of a sample of SiO2 where the SiO2 is essentially pure or are you determining Si in a silicone, or other? If the need is to determine the purity of silica, SiO2, the colorimetric method would not be the method of choice, and I don't see that IR would be appropriate. If you will define the problem more specifically, including how many(or often) of these you will need to do, I can help with a method. Do you have ICP capability? It is not trivial to determine silica purity and if you don't have many to do I would suggest getting a contract lab that is experienced to do them, or get your supplier to provide a certificate of analysis for the product. I don't have experience with pharma stuff so I don't know what is possible, but in an industrial environment I would dump the responsibility for determining silica purity on the supplier who is no doubt set up to do this analysis right. If you have to do it, this will not be one of your favorite activities.

Bill thanks for your comments too, I really appreciate this. The purpose of this analysis is to determine the purity of SiO2 that is essentially pure. The SiO2 is then blended with the pharmaceutical product roughly in a 50:50 ratio. We use the suppliers assay results but have committed to testing one batch per year to ensure that we get the same result. Since the test is difficult we give ourselves no end of problems, apparently we are using the suppliers method when we do perform the test. In the quest for alternatives we have ICP as an option. Would you recommend pursuing this avenue? Thanks again.

There is also a method for silica assay of raw materials which involves ashing, adding methanol, sulfuric acid, and HF then evaporating to dryness, adding and evaporating HF, then ashing at 1000 C. The silica assay is then determined from the difference of the two ashes. It is obviously time consuming, there are problems with using HF, and the method is probably not as accurate as ICP.

Margaret,
First a disclaimer that I know nothing about pharma stuff and what is required to satisfy the legal stuff.

A place to begin is to consider what risk you are trying to minimize by analyzing silica. So, if the risk is extractable metals that might be toxic, then one might extract the sample with .1-1M HCl and determine extracted metals by ICP. The HCl would simulate digestive conditions and HCl is a recognized procedure for determining extractable metals in silica.

If the risk is organic stuff that might come from somewhere,then there is a sensitave procedure for total carbon in silica that involves "burning" the silica in special sample boat that provides a temperature boost. See you local elemental analysis company for details, eg Perkin Elmer, or a contract lab. FDA accepted this approach for a food contact silica product I once worked on.

If you want to determine all the inorganic stuff that might be there, including Si, then dissolve the sample in a Li Borate flux, dissolve flux in nitric and determine elements of interest by ICP. Any lab doing minerals will be familiar with this procedure, or get it from ICP company.

The best reference way to determine how much silica is in a sample is the classical approach to mineral analysis. Consult a classic text, eg "Quantitative Chemical Analysis, by Kohltoff et al. for details. A summary: Dissolve the sample in a flux, eg sodium carbonate, precipitate silica and extract soluble metals with an acid. Dry and weigh. Add HF and volatilize silica, leaving behind nonvolatile stuff, eg alumina. Silica is determined by difference. In good hands this procedure has exceptional precision and accuracy. Good hands implies one does a lot of them and gets good at it.

If you only intend to do a few of these, an experinece contract lab would be most cost effective. Pick one that knows about mineral analysis.

I am going to be away for 10 days, "sunny CA", which seems to be freezing, my bad luck.

Bill many thanks for all your help - very much appreciated. Enjoy your trip!