Phase transfer catalysts generally have in common their amphoteric properties (soluble in both phases) and this soapy property also ensures they are super difficult to remove by liquid-liquid extraction, at least. NMP can be regarded in the case of a DCM based synthesis as something similar even though it isn't actual PTC, I figured out it's best to use about a third of the DCM added and the required volume of the aqueous amine can be decreased by about 15%. I had also used GVL instead of NMP but only about 1/6 of the DCM and it went super ok even being able to recover most of the product that didn't precipitate but it had the GVL smell.