a Scharrer and Mosandl (2002).
b Gatfield et al. (2007).
c Bensaid et al. (2002).
d Schmidt et al. (2007).
e IOFI (1989).
f Kempe and Kohnen (1999).
g Koziet (1997).
h Hoffman and Salb (1979).
i Culp and Noakes (1992).
j Hoffman (1997).
k Brenninkmeijer and Mook (1982).
l Werner (1998).
m Note d’information (2003).
n Krammer et al. (2000).
Apart from the extraction of vanilla beans, vanillin can also be produced by chemical synthesis or biotechnological pathways from different sources (lignin, guaiacol, eugenol, curcumin, and ferulic acid).
The δ13C values of vanillin derived from other sources than vanilla beans or its extracts are in the range from −24.9 to −37.0‰. Therefore, vanillin in vanilla extracts or derived from vanilla beans can be analytically differentiated from other sources by its δ13C value. It can be used to indicate the authenticity of vanilla products based on vanilla or its extracts.
According to a recommendation of IOFI, the authenticity of vanilla flavors is estimated by the determination of the δ13C value for vanillin. δ13C values in the region of –21.0‰ ± 0.5‰ are considered to indicate the plant origin of the analyzed vanillin ex-vanilla beans (IOFI Information Letter, 1989). The variances resulting from the applied isolation techniques are currently under investigation through ring tests performed by a dedicated analytical working group of the German Chemical Society (GDCh).
As long as the vanillin is derived from biotechnological processes such as by fermentation starting from natural ferulic acid, it can be considered as natural and subsequently applied in natural flavors. Owing to the fact, that the δ13C values of vanillin derived from natural ferulic acid by fermentation are very low, this quality can usually be differentiated from synthetic vanillin qualities that do not meet the naturalness requirements.
In order to detect an adulteration of vanilla bean extract with vanillin from other sources, it is possible to apply multielement GC-IRMS; this means, to also determine the ratios 2H/1H and 18O/16O and to evaluate this tridimensional dataset. The standard for these two ratios is Vienna-Standard Mean Ocean Water (V-SMOW). In a recent paper, Bensaid et al. (2002) show that the sp2 oxygen atom in the aldehydic position can chemically exchange with water in industrial or laboratory procedures. Therefore, the authors used the δ18O value of guaiacol, which is formed by the degradation of the vanillin molecule. However, they state that in practice, oxygen isotope ratios do not play an important role in authentication procedures.
The degradation of the vanillin molecule was already used by Krueger and Krueger (1983), when the instrumentation for the multielement IRMS was not yet well developed. Fraudulent adulterators had learned to imitate the δ13C value of natural vanillin by blending it with “nature-identical” vanillin, with products artificially enriched in 13C concentration (Krueger and Krueger, 1983, 1985).
IRMS measures the average δ13C value of the whole vanillin molecule. Since it is relatively simple to enrich its methyl or carbonyl group in 13C, synthetic vanillin blended with a 13C-enriched fraction may be mistakenly taken as natural.
To check for [methyl-13C]-vanillin, the methyl carbon is removed as CH3I in refluxing HI. Then, IRMS is performed on the CH3I (Krueger and Krueger, 1983).
For analyzing [carbonyl-13C]-vanillin, the molecule has to be oxidized to vanillic acid. After decarboxylation, the CO2 formed is analyzed by IRMS (Krueger and Krueger, 1985).
NMR
Presently, the checking for enriched portions is more easily analyzed by quantitative NMR measurements of the abundance of 2H or 13C in the different positions of the molecule. This method was named site-specific natural isotope fractionation nuclear magnetic resonance spectroscopy (SNIF-NMR®) and it is based on the fact that the distribution of 2H or 13C at the different sites of the molecule is not statistical and depends on the origin of the particular compound.
SNIF-2H-NMR spectroscopy is a powerful tool and allows a deeper insight into the biochemical mechanisms by the determination of isotope contents at specific molecule sites. In this respect IRMS, which usually needs to burn the sample before analysis, is not suited to the measurement of an “intramolecular” isotope distribution. Basic research has shown that deuterium is far from being randomly distributed in organic molecules (Martin and Martin, 1981). Therefore, the D/H-ratios of the different positions of the molecule have to be determined.
The relative sensitivity of NMR for 13C is theoretically about 100 times higher than that for deuterium. But due to the fact that the kinetic isotope effects for carbon are much smaller than those for deuterium and due to shorter relaxation time for deuterium, quantitative 2H-NMR became the best tool for the authenticity control in the flavor field.
The first application of this method to the vanillin molecule appeared in 1983 (Toulemonde et al., 1983). Now SNIF-2H-NMR® is accepted as the official method by AOAC (No. 2006.05) for the authentication of vanillin. SNIF-2H-NMR combined with 13C-IRMS also allows the characterization of biotechnological vanillin derived from ferulic acid, eugenol/isoeugenol, or curcumin.
The vanillin molecule has six monodeuterated isotopomers. Five signals can be seen in the 2H-NMR-spectra of the three different vanillin qualities in Figure 15.2, the two deuteriums in the ortho-position fall into one signal. Table 15.4 shows that vanil-lin synthesized from guaiacol (2-methoxyphenol), by introducing an aldehyde group, is clearly separated from the other variants by a high degree of deuterium enrich ment on the carbonyl function. Vanillin from lignin exhibits deuterium depletion at all positions compared with natural vanillin derived from the vanilla bean. Experimental values for the hydroxyl group are not included in Table 15.4, since they can easily be falsified by H/D exchange processes.
FIGURE 15.2 2H-NMR spectra of vanillin ex-vanilla beans, vanillin from synthetic origin (ex-lignin, ex-guaiacol), and vanillin from biotechnological origin ex-ferulic acid.
| Sample | Position in Molecule | |||
|---|---|---|---|---|
| D−C = O | Ortho | Meta | OCH2D | |
| Ex-beansa | 130.8 ± 3.1 | 157.3 ± 3.0 | 196.4 ± 2.5 | 126.6 ± 1.7 |
| Ex-beansb | 128.3 | 156.3 | 180.4 | 126.4 |
| Ex-lignina | 119.9 ± 6.4 | 132.1 ± 2.6 | 168.8 ± 5.9 | 105.9 ± 1.4 |
| Ex-guaiacola | 315.2 ± 56.9 | 138.8 ± 6.7 | 143.8 ± 5.3 | 139.1 ± 8.4 |