With these very sensitive methods, it may be that your sample is not
completely homogeneous, especially if you've added a totally enriched
tracer for the additions. For 13C analysis, and also 15N, we make up the
solution, dry it down, redissolve it again, dry it, grind it up, and then
sample it from there. That works pretty well.
I have been involved in experiments that resulted in huge variations,
because of heterogeneity rather than faulty mixing.
Dr. Marilyn L. Fogel, Senior Scientist
Carnegie Institution of Washington
5251 Broad Branch Rd., NW
Washington, DC 20015
202-686-2410 ext 2480