There are multiple ways to approach this. Concentration will typically depend on the matrix - soil, water, tissue, etc. I have found that number of particles per weight or volume to be the most often reported (and is what I use) as it is more descriptive of what is found in the sample.
As for the counting procedure itself, some type of visual sorting is most commonly done. FTIR is used for identification and/or confirmation of polymer type. If you are purchasing MPs to use from a company, they will have concentration listed in the product's specs, but it is good to double-check that with your own counting so that you confirm your starting value for dosing. I have done this by taking multiple subsamples of the product in a known unit (e.g., ul), visually counted them, and averaged. If you are recovering MPs from a sample of something, it gets a little more complicated. You should still count (taking multiple subsamples if applicable), but you should also record various measurements within that including shape (e.g., number of spherical, fibers, etc.), size, and polymer (e.g., number polystyrene, polyethylene, etc.). These types of counts and distributions are far more descriptive of what was recovered from your sample than just simple total counts.
Dear Melissa Chernick, I imply FTIR microscope i.e. following youtube video link describes an automated evaluation routine for particle statistics by LUMOS II FTIR microscope.
Adeel Rafiq If you have access to such a machine then, yes, it will work for counting and identification. However, as the video says, it is to be used for samples that have a low load of particles. If you have many plastics in your samples, this technique would be exhaustive and visual counting better. That said, if you decide on visual counting then this type of machine is still useful to identify polymers and verify counts in subsamples.
There are some techniques such as nile red, to identify and count microplastics, in addition to the other techniques aforementioned by other collegues.