Impact of saponification and silver-nitrate purification on lacustrine alkenone distributions and alkenone-based indices

A growing interest in lacustrine alkenones as a proxy for continental paleotemperature reconstructions accompanied important methodological improvements over the past two decades. New gas chromatography (GC) columns were used for alkenone analysis, that drastically improved alkenone separation, especially for freshwater lakes. However, these recent advances are sometimes not sufficient in separating compounds that interfere with alkenones in the resulting chromatograms and concurrently, new chemical procedures were implemented to further clean up the samples. Here we investigate the impact of two clean-up procedures, saponification and silver-nitrate purification, on alkenone distribution, alkenone quantification, and C₃₇ alkenone-based indices, including the U₃₇^K index. The silver-nitrate purification modified the C₃₇ alkenone distribution and thus the C₃₇ alkenone-based indices, especially the U₃₇^K index, in 6 out of the 9 studied samples by further retaining alkenones with more double bonds. These changes would result on an average error of 3 °C in reconstructed temperatures. Saponification also influenced the C₃₇ alkenone distribution mainly by removing co-eluting compounds, thereby improving the quality of the results. Both saponification and purification resulted in the reduction of the C₃₇ alkenone concentration by almost half. Clean-up steps should thus be used carefully, paying particular attention to any change in alkenone distribution and concentration. Limiting the use of additional clean-up steps reduces the risk of modifying the alkenone distribution.