The food industry relies heavily on HPLC in many areas, including quality control and authentication. It is generally regarded as a robust technique and has been used to establish many protocols for detecting and measuring components in foods and flavours. One example is the widely used flavour vanilla, which is added to a vast range of foods, confectionery and beverages.

In this case, analysts are usually interested in the four major phenolic compounds which contribute to the flavour: vanillin (4-hydroxy-3-methylbenzaldehyde), vanillic acid (4-hydroxy-3-methylbenzoic acid), 4-hydroxybenzaldehyde and 4-hydroxybenzoic acid. There are several published methods for these components and quantitation by HPLC is recommended by ISO and the AOAC.

However, there are still some research groups who would like to see improvements. For instance, Arun Sinha and colleagues from the Institute of Himalayan Bioresource Technology at Palampur, India, have concerns about the speed. High throughput methods are desirable in the industry when there are many samples to be analysed on a regular basis.

A second group, from Firmenich SA in Geneva, are also troubled by speed, as well as the need for validated methods. "A quantitative method cannot be considered complete without a validation step" declared Alain Chaintreau and Esmeralda Cicchetti. The groups have approached the problems in different ways, both being published in recent issues of the Journal of Separation Science.

The Indian group developed and compared two methods for the quantitation of the four key flavour compounds, which were validated according to ICH guidelines. They were based on HPLC with a monolithic column and ultra-high-pressure LC (UPLC) with a specialised reversed-phase C₁₈ column, in which the column was linked to photodiode array and mass spectrometry detectors. The target analytes were extracted from powdered vanilla pods with ethanol.

The isocratic mobile phase was acetonitrile in aqueous trifluoroacetic acid in both cases and the conditions were optimised for speed and good resolution. The resolution, capacity factors and chromatographic tailing were better for the monolithic column but the UPLC method displayed a higher number of theoretical plates. Retention times on the monolithic column were about 25-30% lower at 0.86-1.45 minutes.

For each analyte, the detection and quantitation limits were lower by UPLC by approximately 50%, covering the range from 0.04-0.38 µg/mL compared with 0.08-0.78 µg/mL for the monolithic column.

After full validation, the team concluded that both methods were suitable for the quantitation of the vanilla flavour compounds, the monolithic column procedure having the slight advantage in terms of cost and availability.

In Switzerland, Chaintreau and Cicchetti compared reversed-phase HPLC with UPLC, both with UV detection for measurement of the same four components in vanilla beans. Once again, they were validated according to ICH requirements. Two types of C₁₈ column were employed and the mobile phase was a gradient of acetonitrile in aqueous phosphoric acid.

Peak resolution using UPLC was about 15-fold better than with HPLC, although the peak capacities were similar. However, the difference in detection limits was remarkable. In HPLC they ranged from 0.039-0.244 µg/mL but these figures plunged to 0.63-0.88 pg/mL in UPLC, which is about 1000-fold lower than those found in conventional HPLC. The change was attributed to better sensitivity due to reduced band broadening.

The levels of the four phenolics determined by the two methods were very similar with a relative difference of less than 6% but several other advantages, apart from the detection limit, pointed towards the UPLC method. The run time was about four times shorter than that in HPLC, allowing higher throughput. The injection volumes could also be reduced by a factor of 10 and solvent consumption reduced by a factor of 20 to values of 0.5 µL and 0.31 ml/min, respectively.

Although all four methods from both research groups are suitable for the vanilla constituents, the UPLC method from the Swiss group would seem to perform the best. Adoption of a method in a particular lab can depend on the equipment already in house, to save on the expense of new kit. Now, the analyst's choice has been extended by these four new validated methods which will help to accelerate sample throughput.

Related links:

• Journal of Separation Science 2009, 32, 3425-3431: "Ultrafast UPLC-ESI-MS and HPLC with monolithic column for determination of principal flavor compounds in vanilla pods"

• Journal of Separation Science 2009, 32, 3043-3052: "Quantitation of the main constituents of vanilla by reverse phase HPLC and ultra-high-pressure-liquid-chromatography with UV detection: Method validation and performance comparison"

Article by Steve Down

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.