Comparison of Compounds in Bourbon Vanilla Extract and Vanilla Flavour

This application note presents a gradient method using a core–shell column for the determination of various substances in Bourbon vanilla extracts as well as in natural and artificial vanilla flavours. The aim is to compare the ingredients to achieve proof of the authenticity of Bourbon vanilla. By applying the Knauer AZURA Analytical system, a very short analysis time and low eluent consumption could be achieved. The high speed and reliability of the method makes it well-suited for routine analyses in food control.

Vanillin is one of the most popular flavouring agents used in various food products and beverages as well as in the pharma and perfume industry. With a high demand for the supply of vanilla pods and the continued increase in price, artificial vanilla flavouring agents of synthetic origin are now available. Artificial vanilla is a phenolic aldehyde, primarily obtained from the extracts of the pods of Vanilla planifolia, a species of vanilla orchids. It is also found in roasted coffee and Chinese red pine. Chemically, vanillin is 4-Hydroxy-3-methyl benzaldehyde (1).

Figure 1: Chemical structures of the analytes.

The high demand for vanillin far exceeds the supply from all sources of vanilla orchids, which is the only vanilla flavour allowed to use the name “Bourbon vanilla”. The high price of natural vanillin, compared with that of synthetic vanillin, and the poor availability are the reasons for the production of vanillin via chemical synthesis since the 1870s. These processes use coniferin, guaiacol, or eugenol as a precurser (2).

Biotechnological processes like fermentation that use ferulic acid and rice bran as precursors of vanillin are relatively new. Biotechnologically produced vanillin is much more cost intensive than chemically synthesized vanillin, but the EU allows this product to use the designation “natural vanilla flavour” while the chemically synthesized flavour has to use the name “vanilla flavour”. Some substances from the chemical or biotechnological processes are undesirable in food products as a result of negative health effects. This makes an analytical control indispensable. These molecules as well as the precursors used in the chemical synthesis are appropriate markers for the differentiation between synthetic vanilla flavour and Bourbon vanilla extract. Analysis of this kind is getting more and more attention caused by food scandals in the last years.

While an exact statement about the origin of vanilla flavour is only possible after complex analytical methods like isotopic analysis, a first statement about the origin of vanilla flavour is already possible by screening for marker substances using relatively easy, cost-effective, and robust HPLC methods. Therefore, in this work ethanolic extracts of Bourbon vanilla pods are compared with synthetic vanilla flavours to find marker substances for the origin of the flavour.

Experimental Sample Preparation
Bourbon vanilla can be bought in supermarkets as a baking supplement. Vanillin can be easily extracted from Bourbon vanilla pods using an ethanol as the solvent. To enhance the extraction process, the mixture is put into an ultrasonic bath and left to stay for extraction overnight. After filtration through a syringe filter and dilution with the mobile phase, the sample is ready for analysis by HPLC.

Experimental Preparation of Standard Solution
Standards were solved and diluted in the mobile phase. 4-hydroxybenzoic acid, vanillic acid, and 4-hydroxybenzaldehyde were analyzed - in addition to the vanillin - as typically occurring substances in Bourbon vanilla extract. In addition, guaiacol, ethylvanillin, coumarin, and eugenol were analyzed as markers for synthetic vanilla flavour or even unwanted additives (3).

Results
Figure 2 shows the overlay of the standard with a real sample of Bourbon vanilla extract. It is obvious that all substances could be separated in less than 2 min and elute in very sharp peaks. This makes the method very attractive for routine analyses because of its speed. High resolution is achieved by the very low dead volume of the AZURA Analytical HPLC system and the high resolution core–shell column.

Figure 2: Chromatograms at 280 nm, blue: standard, green: Bourbon vanilla extract.

Figure 2 also shows that in the Bourbon vanilla extract the substances 4-hydroxybenzoic acid, vanillic acid, 4-hydroxybenzaldehyde, and vanillin were found as expected. At the same time, the unwanted substances guaiacol, ethylvanillin, coumarin, and eugenol were not.

3D data were acquired using the DAD 6.1 L diode array detector. This allows spectra of every single substance to be recorded to get further information about the identity of the occurring peaks. Figure 3 shows the recorded spectra.

Figure 3: Spectra of standards recorded with the DAD 6.1L. [Click Image to Enlarge].

Based on these results, a commercially available “Bourbon vanilla extract” and a synthetic vanilla flavour were analyzed. In the Bourbon vanilla extract, the typical markers for the vanilla pods could be found while the markers for the synthesized flavour could not be detected (see Figure 4).

Figure 4: Chromatograms at 280 nm, blue: standard, green: Bourbon vanilla flavour.

Figure 5: DAD spectra, blue: 4-hydroxybenzaldehyde, green: unknown peak from Bourbon vanilla flavour at ca. 1.3 min.

In addition, the peak eluting at about 1.3 min could be identified as a substance that may be analogous to 4-hydroxybenzaldehyde because the spectra match. For proof of this thesis, further analysis would be needed, such as the coupling of the HPLC to a mass spectrometer.

For the synthetic vanilla flavour, results were completely different. Markers for the natural extract were not detected in significant amounts while the relative amount of vanillin itself was much higher.

Figure 6: Chromatograms at 280 nm, blue: standard, green: artificial vanilla flavour.

The peak eluting at about 1.2 min warrants special attention because it elutes at a time very close to coumarin, which is a particularly unwanted substance in food products because of its adverse health effects. The German Federal Institute for Risk Assessment (Bundesinstitut für Risikobewertung BfR) defines a daily tolerable intake (TDI) of 0.1 mg per kg body weight, which can be consumed over a lifetime without negative health effects (4). Coumarin can be found in significant amounts in tonka beans, which also have a strong vanilla flavour and are used as a different source for vanilla flavour. Once again the DAD spectra of coumarin and the peak from synthetic vanilla flavour were compared. From Figure 7(a) it is obvious that the substance is not coumarin because the spectra differ at several points.

Figure 7: DAD spectra: (a): blue: coumarin, green: peak from artificial vanilla flavour at ca. 1.2 min. (b): blue: guaiacol, green: peak from artificial vanilla flavour at ca. 1.0 min.

Figure 7(b) shows an overlay of the spectra from guaiacol and the sample at 1.0 min. The similarity means it is likely to be the substance guaiacol, which is used as a precursor in the vanillin synthesis process. This is today’s established process for the synthesis of vanillin and therefore the occurrence of this substance in the sample is not surprising. At the same time we can see from these results that guaiacol is a good marker for chemically synthesized vanillin.

Conclusion
The presented HPLC method with diode array detection is well-suited for fast and easy screening of the origin of vanilla flavours in food products. Using comparable easy-to-handle, cost-effective, and robust equipment this difficult problem could be solved. The substances 4-hydroxybenzoic acid, vanillic acid, and 4-hydroxybenzaldehyde as well as their relative concentrations to vanillin could be used as markers for natural Bourbon vanilla flavour while guaiacol, ethylvanillin, coumarin, and eugenol could be used as markers for synthetic vanilla flavour. Guaiacol in particular, the chemical precursor for today’s most widely used vanillin synthesis process, could be detected in artificial vanilla flavour.

References

1. Krishna Veni et al., J Adv Sci Res 4(1), 48–51 (2013).
2. Jagerdeo et al., Journal of AOAC International 83(1), 2000
3. Elke Anklam, Authenticity of vanilla and vanilla extracts, Joint Research Centre European Comission, Environment Institute Food & Drug Unit, EUR 15561 EN (1993).
4. Neue Erkenntnisse zu Cumarin in Zimt, Stellungnahme Nr. 036/2012 des BfR vom 27 September 2012, http://www.bfr.bund.de/cm/343/neue-erkenntnisse-zu-cumarin-inzimt.pdf

Knauer
Wissenschaftliche Geräte GmbH
Hegauer Weg 38, 14163 Berlin, Germany
Tel: +49 30 809727-0  Fax: +49 30 8015010
E-mail: info@knauer.net  Website: www.knauer.net