I finally decided just to send the samples out. Just too difficult. I tried and and ran into trouble on three fronts.

1. HMDS even with TMCS wasn't strong enough to achieve good derivatization yield I saw multiple peaks for ITSD (DACP) and low yeild for melamine.

2. Methanol extract and blow down left the samples too dirty. I saw a lawn of junk and a lot of methyl esters. If I try again I will have to do a better deproteination (Agilent recommends lead acetate and trichloroacetic acid) followed by SCX SPE cleanup then evaporation and silylation.

3. Programming the backflush in Chemstation MSD is hard in constant flow mode. Agilent has a trick where you use a nonexistent column 2 with an inlet and aux 5 EPC except I have a column 2 on the rear inlet.