Analytical model for site-specific isotope fractionation in 13 C during sorption: Determination by isotopic 13 C NMR spectrometry with vanillin as model compound Patrick Höhener a,⇑ , Virginie Silvestre b , Anaïs Lefrançois b , Denis Loquet b , Eliot P. Botosoa b,1 , Richard J. Robins b , Gérald S. Remaud b a Aix-Marseille Université – CNRS, Laboratoire Chimie Provence, UMR 6264, Case 29, 3, Place Victor Hugo, F-13331 Marseille Cedex 3, France b Elucidation of Biosynthesis by Isotopic Spectrometry Group, Unit for Interdisciplinary Chemistry: Synthesis, Analysis, Modelling (CEISAM), UMR 6230, University of Nantes – CNRS, 2 rue de la Houssinière, BP 92208, 44322 Nantes Cedex 3, France article info Article history: Received 8 June 2011 Received in revised form 28 November 2011 Accepted 10 December 2011 Available online 9 January 2012 Keywords: Stable isotopes Modeling Sorption Transport NMR abstract The aim of this study was to conceive a reactive transport model capable of providing quantitative site- specific enrichment factors for fractionation in 13 C isotopic content during sorption. As test compound the model treats vanillin, for which the 13 C isotopic content at natural abundance at each of the 8 carbon positions can be measured by quantitative 13 C nuclear magnetic resonance spectrometry. This technique determines the isotope ratios with a resolution better than ±1‰ (0.1%) at each carbon position. Site- specific isotope fractionations were recorded in chromatography column experiments with silica RP-18 as stationary phase. The one dimensional reactive transport model accounted for the sorption/ desorption behavior of 8 individual 13 C-isotopomers and one 12 C-isotopomer of vanillin and reproduced satisfactorily the bulk (average over the whole compound) fractionation observed during elution. After model calibration, the enrichment factors were fitted for each carbon site where a significant fraction- ation was recorded. To show the interest of such a transport model for environmental studies, the model, extended to three dimensions, was exploited to simulate reactive transport in an aquifer. These results show that significant 13 C isotope fractionation is expected for 4 out of 8 13 C-isotopomers in vanillin, and illustrate that bulk isotope ratios measured by conventional compound specific isotope analysis and mass spectrometry would hardly document significant isotope fractionations in vanillin. It is concluded that modeling of site-specific isotope ratios in molecules is a priori feasible and may help to quantify unknown processes in the environment. Ó 2011 Elsevier Ltd. All rights reserved. 1. Introduction The monitoring of stable isotope ratios is increasingly used in different branches of science in order to find the origin of target molecules and/or processes that acted on them. In environmental sciences, stable isotope ratios are either used to identify different sources of pollutants by highlighting subtle but significant differ- ences in isotope abundance, or used to identify processes that gov- ern the fate of these pollutants after release from a source (Meckenstock et al., 2004; Schmidt et al., 2004). Among the pro- cesses that can be assessed by stable isotope analysis in environ- mental sciences are biotic and abiotic degradation, volatilization, gas-phase diffusion, and sorption (Elsner et al., 2005; Aelion et al., 2009). In the agriculture and food industries, stable isotope ratios are used to verify the origin of food ingredients in order to detect fraud. Well-known applications are the detection of forbid- den sugar additions to honey and wine (Martin et al., 2006b), and the identification of synthetic instead of natural vanillin in ice cream (Remaud et al., 1997). In environmental sciences, the most frequently used analytical instrument to measure stable carbon isotope ratios of pollutants at natural abundance is an isotope ratio mass spectrometer cou- pled via a combustion interface to a gas chromatograph (GC–C– IRMS) (Aelion et al., 2009). GC–C–IRMS attains a very low detection limit, since 1 nmol carbon in a target compound is sufficient to re- cord the isotope ratio (Schmidt et al., 2004) and to compare it to an international standard via the d 13 C notation (Eq. S1, Supplemen- tary material). However, the ratio measured is a bulk value aver- aged over the whole compound (designated d 13 C bulk ), since combustion transforms the target compound to CO 2 . This is a prob- lem when dealing with molecules having a large number of carbon atoms, particularly because isotope effects are generally restricted to specific atoms that are at key positions in a molecule. For exam- ple, it is well known that for the aerobic biodegradation of linear alkanes, the initial enzymatic attack occurs at a terminal carbon, 0045-6535/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2011.12.023 ⇑ Corresponding author. Tel.: +33 4 13 55 10 34; fax: +33 4 13 55 10 60. E-mail address: patrick.hohener@univ-provence.fr (P. Höhener). 1 Present address: Laboratoire CSPBAT (UMR 7244), équipe SBMB, Univ. Paris 13, France. Chemosphere 87 (2012) 445–452 Contents lists available at SciVerse ScienceDirect Chemosphere journal homepage: www.elsevier.com/locate/chemosphere