From the TLC analysis of the fermentations carried out with the three yeast strains 240, 1026, and 1087 we can see that there is a preferential use of the available carbohydrate. However, such a simple analysis does not provide any information as to how the sugars are processed during the fermentation, the amount of alcohol produced or indeed how such changes affect the taste or aroma of a finished brew.
Flavour and fragrance are in part due to the presence of volatile organic compounds, many of which are produced by yeasts (others come from the malt and hops). A quick and simple method for the analysis of volatile organics is Gas chromatography-mass spectrometry (GC-MS).
GC-MS (Gas Chromatography-Mass spectrometry) consists of two analytical techniques working in tandem. The first technique, GC, separates mixtures of compounds into their individual component parts. The Mass Spectrometer detects and then characterises each of the individual components based on their chemical structure and, usually, molecular weight.
As with all chromatography, separation of a mixture is obtained due to the differential partitioning of individual chemicals between a mobile phase (in this case helium gas) and a stationary phase (the column packing). Those compounds that have a greater affinity for the mobile (gas) phase will elute (leave) from the column first, thus having a short column retention time, whilst those that have a greater affinity for the stationary phase will be retained on the column for longer and hence have a longer retention time. The resolution of mixtures can be improved by alteration of the nature of the stationary phase, the temperature of the column (and injector) and the flow rate of the eluent gas. This separation can be visualised on the GC chromatogram.
Each individual compound will have a specific retention time on a specific column under set conditions. Comparison of retention times with those of known standards can enable tentative identification of the individual components. However, some compounds may have identical retention times, and so cannot be fully identified by this method. Similarly, if no known standards exist which match the RT of a fraction then identification will not be possible. This is where the Mass Spectrometer becomes vital.
Each resolved fraction is analysed by the Mass Spectrometer. Once separated, the individual fractions elute from the GC column and enter the mass spectrometer ion source. Here they are bombarded with a stream of electrons which causes them to ionise to give positive ions, and break into small fragments. Essentially it is the pattern of this fragmentation that allows identification. Each compound has a distinct and individual fragmentation pattern which is independent of the type of column used for separation in the GC phase. So whereas two compounds may have the same retention time on a given GC column they are extremely unlikely to have same fragmentation pattern – thus allowing for their identification.
After fragmentation the ionized fragments enter the mass analyzer where they are separated on the basis of their mass to charge ratio (m/z). In our case the mass spectrometer used separates the fragment ions by passing through a quadropole ion trap. The ions are separated and ejected from the trap in order of increasing m/z. The separation is achieved by varying the voltage across the quadropoles. The ions with the lowest m/z emerge first and are picked up by the detector.
The detector then produces a mass spectrum which shows the intensity or relative abundance of ions for each m/z ratio with respect to size as shown below.
Follow here for an explanation on deciphering mass spectra
The first stage of any analytical process is to establish a method with which you are able to detect a range of known standards. Standards are chosen that will provide a representative sample set of the analytes that are being looked for in any given sub-set of unknowns (real experimental products!). Using a standard mix we were able to develop a GC-MS method which permitted the separation and identification of 12 potential fermentation products.
Time in mins 1 = 1- propanol RT = 1.9 2 = diacetyl RT = 2.1 3 = ethyl acetate RT = 2.3 4 = isobutanol RT = 2.6 5 = 1-butanol RT = 3.1 6 = isoamyl alcohol 4.3 7 = active isoamyl alcohol RT = 4.4 8 = isobutyl acetate RT = 4.8 9 = isoamyl acetate RT = 6.3 10 = ethyl caproate RT = 7.575 11 = phenethanol RT = 8.7 12 = ethylcaprilate RT = 9.3
The standard mix contained ethanol at a final concentration of 40,000ppm, diacetyl at 200ppm, and all the other standards at 20ppm in water. From the chromatogram above it can be seen that peak height is not directly related to concentration. If it were, all the peak heights with the exception of diacetyl, would be identical. This is not the case. This non-correlation is due to the differences in the volatility of the analytes, and the differences in their partitioning between the liquid and vapour phases in the vial. An analyte's hydrophobicity is also important. Those that are most volatile will evaporate more completely into the headspace, and therefore be at a higher gaseous concentration than those of a lower volatility. By running a range of standard concentrations (10, 15, 20ppm for minor components, 100,150 and 200ppm for diacetyl) a standard curve can be generated. By reading from the standard curve we can accurately measure the concentration of each component in any given sample set. For example from the standard curve below we can see that if the GCMS output is 17.5 (Ay) then by reading across we find that the concentration is 17.5ppm (Ax). A standard curve needs to be generated for each individual component.
By running our fermentation samples under identical conditions to those used for detection of the standards we were able to identify the range of compounds produced by each strain, and their concentrations.
TIC Chromatogram of Headspace analysis of Strain 240 after 6 days Fermentation.
The chromatogram above shows the resolution of the individual components of the fermentation broth at the end of a 4-day fermentation. Some of the peaks are well defined and can be easily identified by comparison of their retention times and mass spectra with those of our known standards.
Experimental mass spectrum of 1-propanol (RT = 1.92mins)
Spectrum of 1-propanol from a MS library
Spectrum from NIST Chemistry WebBook This shows a very good match between the spectrum generated from our sample and the library spectrum and so this provides unequivocal evidence that the peak assignment is correct. Comparison of mass spectra really becomes important when dealing with structural isomers of the same compound for instance 1- and 2-propanol.
Very small peaks can be visualised by zooming in on the spectrum. Enhancing the spectrum can give a much clearer resolution as shown below:
Spectrum enhanced between 4 mins and 6.5 mins
Mass Spectrum of Isobutyl acetate (RT = 4.9mins)
Library Spectrum of isobutyl acetate
By comparison of the retention time and mass spectrum of this peak with those of the standards run previously we can assign this peak to the presence of isobutylacetate. Hence, the ability to enhance spectra is particularly useful when dealing with low concentrations of analyte or analytes that are of relatively low volatility.
We employed this technique to allow each fermentation to be followed precisely over a number of days. We were able to see which compounds were produced first, and the extent to which they accumulated. By overlaying spectra it is possible to see which components are produced/depleted over time.
However, the total ion chromatogram becomes exceedingly ‘busy’ when multiple analyses are aligned over each other, but by enhancing a particular area, the results are much clearer:
From this it can easily be seen that isoamyl acetate gradually increases over the course of the fermentation.
A detailed analysis of the fermentation process can be seen in the fermentation results section and analysis of the volatile components produced during the fermentations can be seen at the GC-MS results section.