From literature, people always filter the PEDOT: PSS solution before spin-coating. I think it will lose part of big-molecule materials, which may influence its thermoelectric performance.
The filters used for PEDOT:PS have quite large porous sizes such as 1 micrometer to remove big particles of its solutions. The molecular weight of PEDOT:PSS is normally in certain range for different applications.
If you are worried about the use of filters for PEDOT:PSS solutions, you could try to compare with the performances between the solutions with using filters and without using filters. You can optimise the PEDOT:PSS solutions for your thermoelectric applications.
PEDOT:PSS aqueous "solution" isn't actually a solution, but a dispersion with PEDOT:PSS particles. And a lot of people, including me, don't filter it as filtering a dispersion can lead to a loss of a part of the material.
Hi! In our labortory we don´t filter PEDOT for application in solar cells. This bacause we did some experiments with and without, and performance was better without filtering. The recommendation, try the two cases and select the best experimental result. Rgds