Provided you are using a sealed ampoule (quartz glass) combustion technique and
do the combustion at a high enough temperature (1000 C +) there are no problems
with running low TOC samples that have been fully decarbonated. We routinely
analyse C samples at the ppm level in various matrices including clay minerals
and have not found matrix effects that affect the yield, provided the
temperature is high enough. The matrix does influence the combustion
temperature of some forms of carbon and in this respect we have had problems
with organic carbon in cherts where the carbon is occluded.
Hope this is of help.
Earth Sciences, Open University