Use of compound specific δ¹³C stable isotope analysis for intact triacylglycerides using high temperature gas chromatography up to 430°C

Preliminary results from compound specific 13C isotope ratio mass spectrometry measurements of intact triacylglycerides in olive oil achieved by coupling high temperature gas chromatography with oven cycling up to 430C with IRMS


Introduction
Esters of glycerol and three fatty acids (triacylglycerides or TAGs; allows the molecular characterisation of the separated FAMEs using their 'fingerprint' mass spectra and relative retention times (Farkas et al., 2008). Liquid chromatographic (LC) techniques with ultraviolet detection and MS have also been used to identify and quantify FAMEs (e.g. Wei and Zeng 2011). Similarly, LC coupled with refractive index detection or MS (electrospray ionisation, atmospheric pressure chemical ionisation, atmospheric pressure photoionisation or matrix assisted laser desorption ionisation; ESI, APCI, APPI or MALDI) have been used for intact TAGS, as recently reviewed by Tena et al 2019. In addition to quantitative analysis and molecular identification, measurement of the stable isotopes of constituent carbon, hydrogen and oxygen atoms can provide a profile of the biological, chemical and physical processes that each molecule has undergone (Lichtfouse 2000). Stable isotopic values of individual FAMEs (compound specific) are usually obtained when GC is coupled with a stable isotope ratio mass spectrometer (GC-IRMS;Camin et al., 2010;Paolini et al., 2017). Methyl esterification has been reported to have a modest effect on the  13 C and  2 H values of FAMEs (Chivall et al., 2012;Paolini et al. (2017). Although LC has been coupled with IRMS it is less well established due to difficulties removing organic solvent necessary to effect some chromatographic separations. When GC is coupled to the IRMS via a catalytic combustion reactor operated at around 1000°C (GC-C-IRMS) the FAMEs are converted to CO2 and the   C of the CO2 measured relative to a reference gas.
Similarly,   H (including H3 + correction) and   O can be measured from H2 and CO, respectively, when the GC is interfaced with the IRMS via a pyrolysis reactor (GC-Py-IRMS; preconditioned with solvent) and operated at around 1420°C. Only recently has the GC-Py-IRMS   H measurement of FAMEs been reported (Paolini et al., 2017). Carbon fractional isotopic abundance ( 13 F; Eq 1), carbon isotope ratio ( 13 R; Eq 2) and the representation of isotope values using the delta notation ( 13 C) can be calculated from isotope measurements (Eq 3). An adjustment is made for the isotopic contribution of the Here we present preliminary data comparing  13 C isotope data obtained using bulk and compound specific isotope ratio mass spectrometry with the latter utilising HTGC operated up to 430°C.

Methyl trans-esterification of tripalmitolein
An aliquot of tripalmitolein was partially methyl trans-esterified by heating in a closed vial with 10% HCl/MeOH (v/v; 2 mL; 70°C; 2 h). The extract was recovered, following the addition of water (2 mL), in cyclohexane (3 x 2 mL) after mixing and centrifugation.
Solvent was removed (N2/65°C) and the extract re-dissolved in cyclohexane. The incomplete methyl trans-esterification of tripalmitolein allowed the CSIRMS carbon isotope values of the intact TAG and FAME derivative to be measured in the same analysis. process data.
2.5 13 C High temperature gas chromatography compound specific combustion isotope ratio mass spectrometry (HTGC-C-

CSIRMS)
Compound specific carbon isotope ratio mass spectrometry analysis was undertaken using a Thermo Scientific™ IRMS instrument: a TRACE™ 1310 GC (schematic Figure 2)  Isodat™ software suite (Thermo Scientific™) was used for data acquisition and evaluation.
Carbon isotope ratios of olive oils are influenced by a complex of factors such as latitude, water availability, coastal proximity, temperature and relative humidity (Camin et al., 2010;Chiocchini et al., 2016). We were interested to see whether this difference was evident in dominant constituent TAGs.
The present study used a Vf-5ht column to separate the main TAGs prior to combustion for isotope analysis which may have led to some overlap with minor components (e.g. Figure S4 where OIL2 and Oil 8 were separated on an MXT-biodiesel TG column). One of the replicate 50 ng on column injections produced peaks that were around four times higher than the other two replicates,

Nevertheless, peaks
gave lighter values for all three peaks and was most variable for peak A, which may relate to linearity. OIL2 was analysed twice with injections of 50 and 100 ng on column. Whilst there was an order of magnitude difference in peak amplitude the  13 C values for peaks B and C were reasonably consistent whilst the smallest peak A varied by around 3.6 ‰ (Figure 3).

Conclusions
This preliminary work has confirmed our hypothesis and demonstrated for the first time that high temperature gas chromatography using GC oven temperatures up to 430°C can be coupled with compound specific isotope ratio mass spectrometry   Peak A ( Figure S5 and S6) mass spectrum ( Figure S7), molecular ion m/z 833 (C53H100O6) and a two fragment ions at m/z 577 (C37H69O4) and m/z 551 (C39H71O4) in a 2:1 ratio indicating that one of the constituent fatty acids has molecular weight: mass 833 -551 = 282 (normally there would be an allowance for the additional hydrogen but it is not needed in this case because of rounding up the monoisotopic mass) which equates to oleic acid (C18H34O2) and two fatty acids have the same molecular weight: mass 833 -577 = 256 which equates to palmitic acid (C16H32O2).
Peak B ( Figure S5 and S6) mass spectrum ( Figure S8) molecular ion m/z 859 (C55H102O6) and a two fragment ions at m/z 577 (C37H69O4) and m/z 603 (C39H71O4) in a 2:1 ratio indicating that one of the constituent fatty acids has molecular weight: mass 859 -603 = 256 (normally there would be an allowance for the additional hydrogen but it is not needed in this case because of rounding up the monoisotopic mass) which equates to hexadacanoic acid (C16H32O2) and two fatty acids have the same molecular weight: mass 859 -577 = 282 which equates to oleic acid (C18H34O2).
Peak C ( Figure S5 and S6) mass spectrum ( Figure S9) molecular ion m/z 885 (C57H104O6) and a single fragment ion at m/z 603 (C39H71O4) indicating that constituent fatty acids have the same structure and molecular weight: mass 885 -603 = 282 (normally there would be an allowance for the additional hydrogen but it is not needed in this case because of rounding up the monoisotopic mass) which equates to oleic acid (C18H34O2).