R. Chaler et al., Sensitivity effects in U-37(k') paleotemperature estimation by chemical ionization mass spectrometry, ANALYT CHEM, 72(24), 2000, pp. 5892-5897
Analysis of C-37 alkenone mixtures by gas chromatography (GC) with flame io
nization detection (FID) and GC coupled to mass spectrometry (GC/MS) in the
chemical ionization mode (CI) shows that the later is useful for paleotemp
erature estimation when ammonia is used as reagent gas. Conversely, the use
of isobutane gives rise to U-37(k') readings that are dependent on the amo
unt of C37 alkenones introduced in the system, being unreliable for paleocl
imatic studies. However, ammonia CI GC/MS may produce U-37(K)' measurements
that deviate from those obtained by GC-FID, the method calibrated for temp
erature estimation from algal cultures and marine sedimentary data. The dif
ferences result from changes in relative sensitivity between the di- and tr
iunsaturated alkenones and depend on the instrument used and operational co
nditions. This problem is solved in the present study by determination of t
he response factor linear equations for each alkenone and their average rel
ative sensitivity (R) using mixtures of known composition. These parameters
allow the transformation of the GC/MS readings into the GC-FID equivalents
using the following equation: UG = R x UM/(1 - UM(1 - R)). Examples of the
suitability of this approach are given.