Hi, Maximilian:
That does look like you have some insoluble crystalline material in your sensitizer. I guess the filtering results in loss of expensive material as well - given how absorbent those Whatman filters can be. Not something one would want to have to do routinely.
Once this filtered solution that presumably has no crystals is coated, somehow they are formed again on the paper. Actually that is itself is not unusual - as all sensitizers when they are dried form particles, but they are usually small enough to escape notice. If I remember correctly my crystallization kinetics from way back, slower drying should result in bigger crystals than faster drying. Is there something that can be done in how you are drying the paper after coating that might affect the results?
I wonder why Malde/Ware process calls for such a high solids content in the sensitizer - seems 60% FAO is close to its saturation point- may be to maximize the Dmax, don't know. But that is also asking for trouble as there would be greater propensity for precipitates with small variations in temperature and weights/volumes. In any case, it seems you are making a combined sensitizer by mixing all of the the FAO and palladium together and store. Perhaps a better way might be to not do that and only mix drops of the two when you are ready to coat. One, you are not messing up the whole quantity of palladium in case something goes wrong. Two, by coating the mixture right away, perhaps there is not enough time to form crystals in the solution itself. I don't know, but that might be the safer thing to do before the root cause of the problem can be figured out.
Just as a philosophical aside: why would one prefer this process over the conventional tried-and-true developed-out process that is far more straight forward. This seems to me a bit too complicated with all its preparations and various conditions that need to be controlled just to eliminate the necessity of a developer.
In any case, good luck....
:Niranjan.