Analyzing aerosol particles collected on polycarbonate filters might be a rather difficult thing to do at least from my experience of analyzing asbestos particles collected on similar filter materials. One problem is very low electric conductivity of filter materials and consequent electrical charging of the samples which prevents you from viewing your samples under significantly high magnification to actually observe the aerosol particles and with high enough voltage to do the EDX analysis.

To avoid the electrical charging problem, the best solution would be to transfer the captured particles onto a conductive surface, like carbon tape commonly used in electron microscopy, and to make them more electrically conductive you can use sputtering method to cover your samples with thin layer of gold, gold-palladium, platinum or some other highly conductive material.

The other problem is doing EDX analysis on a single particle, because you have to consider the depth of the high-energy electron beam penetration into the sample, which is usually around one micrometer or more and consequent depth from which the characteristic X-ray emission lines used in sample analysis are coming from. This penetration depth of the high energy electron beam depends on the accelerating voltage, material density and the atomic number. It can be calculated using R. Castaing's expression. So if your aerosol particles are smaller than the penetration depth you have to consider that your EDS analysis results will also contain significant contribution of the background holder material on which your aerosol particles have been laid upon and prepared for electron microscopy.

I hope this information will be somewhat helpful.

Thanks Marijan, We do follow quite similar procedure for PM10 analysis using our SEM-EDX system.. We do well on single particle analysis with manual mode, but I really interested in automated and computer aided analysis for multiple particles e.g. hundreds of single particles on one filter automatically. As you see it is very difficult and time consuming procedure if you analyze one particle each time manually. I have decided to use a software recently, now working on receipt for PM10 identification, yet I am not happy since the software is quite expensive :( it would be better to fine a cheaper preferable a free one..

Dear Dr. Ferhat,

As far as I know, the data acquisition process can be automatic using CCSEM (Computer controlled SEM)...where a brightness threshold is set so that SEI images and EDX spectra are acquired from the relatively bright regions on the image field (on the substrate) taken as the aerosol particles.

You mentioned an expensive software. Which one is it?

Dr. Alexander Laskin of Pacific Northwest Lab uses a CCSEM and has many related publications on Aerosol characterization.