A reliable compound-specific nitrogen isotope analysis of amino acids by GC-C-IRMS following derivatisation into N-pivaloyl-iso-propyl (NPIP)esters for high-resolution food webs estimation

- Amino acids were baseline separated in GC-C-IRMS analysis.
- High correlations were observed between the δ15N values obtained by GC-C-IRMS and EA-IRMS.
- Cation-exchange chromatography was effective to purify the sample.
- The actual food web structure in natural environments can be accurately estimated with the presented approach.

The signatures of natural stable nitrogen isotopic composition (δ15N) of individual amino acid (AA) have been confirmed to be a potentially effective tool for elucidating nitrogen cycling and trophic position of various organisms in food webs. In the present study, a two-stage derivatisation approach of esterification followed by acylation was evaluated. The biological samples underwent acid hydrolysis and the released individual AA was derivatived into corresponding N-pivaloyl-isopropyl (NPIP) esters for nitrogen isotopic analysis in gas chromatography-combustion-isotope ratio mass spectrometry (GC-C-IRMS). Usually, 13 individual AA derivatives were separated with fine baseline resolution based on a nonpolar gas chromatography column (DB-5 ms). The minimum sample amount required under the presented conditions is larger than 20 ngN on column in order to accurately determine the δ15N values. The δ15N values determined by GC-C-IRMS with a precision of better than 1‰, were within 1‰ after empirical correction compared to the corresponding measured by element analysis (EA)-IRMS. Bland-Altman plot showed highly consistency of the δ15N values determined by the two measurement techniques. Cation-exchange chromatography was applied to remove interfering fraction from the extracts of plant and animal samples and without nitrogen isotope fractionation during the treatment procedure. Moreover, this approach was carried out to estimate the trophic level of various natural organisms in a natural lake environment. Results highly proved that the trophic level estimated via the presented AA method well reflected the actual food web structure in natural environments.