Analysis of Fatty Acid Methyl Esters (FAMES), and Examination of Biodiesel Samples for these Components, by GCxGC-FID

Transcription

1 Analysis of Fatty Acid Methyl Esters (FAMES), and Examination of Biodiesel Samples for these Components, by GCxGC-FID Introduction P Gorst-Allman (LECO Africa Pty. Ltd) and B-J de Vos (NMISA). The analysis of fatty acid methyl esters is important both in the food industry, and in the biofuels area where these compounds form the major components of e.g. biodiesel. In the food industry it can be important to determine the absolute amount of both double bond isomers in unsaturated fatty acids, as the trans isomers have significant negative health effects. This consideration is, however, not important in the analysis of fatty acid methyl esters (FAMES) in the biofuels market as naturally occurring FAMES possess almost exclusively cis double bonds, and both isomers are equivalent as fuels. With these considerations in mind the analysis of FAMES in biodiesel produced from plant material can be successfully undertaken using GCxGC-FID. This technique allows successful determination of the fatty acids differing by carbon number, and unsaturated FAME isomers differing by the number and position of double bonds in the compounds. The application of this technique in the determination of FAMES in biodiesel sample is described in this note. Samples Several samples were used during the course of the study. Those described below are: (i) An NLEA FAME Mix, Cat No 35078, sold by Restek. This contains the following components: 4:0, 6:0,, 10:0, 11:0, 12:0, 13:0,, 14:1(cis-9), 15:0,, 16:1(cis-9), 17:0,, (cis-9), (trans- 9) (cis-9,12), (trans-9,12), 18:3 (cis-9,12,15), 20:0, 20:1 (cis- 11), 22:0, 22:1 (cis-13), 23:0, 24:0, 20:5 (cis-5,8,11,14,17), 24:1 (cis- 15), and 22:6 (cis-4,7,10,13,16,19). (ii) A biodiesel sample produced from palm oil provided by LECO Thailand. (iii) A biodiesel sample produced from Jatropha Curcas Linn provided by LECO Thailand. Analysis Conditions GC Parameters: Injection parameters: Primary column: Secondary column: Carrier gas: Agilent 6890N GC modified for GCxGC by LECO Corporation 0.05 µl, splitless at 250ºC. DB-23 (20 m x 0.18 mm id x 0.20 µm film) Rtx-5 (1.1 m x 0.18 mm id x 0.20 µm film) Helium, 1 ml/min, corrected constant flow

2 Oven programme: 80ºC for 2 min, 3ºC/min to 250ºC, hold 2 min Secondary Oven: 100ºC for 2 min, 3ºC/min to 270ºC, hold 2 min Modulator offset: 30ºC Modulation time: 8 seconds Analysis time: min FID Conditions: Temperature: 250ºC Hydrogen flow: 50 ml/min Air flow: 450 ml/min Make Up gas: Nitrogen Make Up flow: 40 ml/min Data collection rate: 50 Hz Acquisition delay: 120 sec Results and Discussion The column set described in the conditions above was chosen for the analysis for the following reasons: 1) The primary column (DB-23) is a highly polar column with a high cyano phase which has been shown to be capable of separating cis and trans isomers. However, for this separation either a 60 m or a 100 m column is normally used. 2) As we were not interested in this separation, the column selected was 20m. The second column was a standard Rtx-5 used for a boiling point separation. The column set worked particularly well as can be seen from the results described below. Initially the NLEA FAME Standard mix was run under the conditions above to determine the positions of the standard components in the chromatogram. The GCxGC chromatogram for this mix is shown in Fig 1. Component identification is indicated on the main chromatogram (Fig 1), and also on an expanded chromatogram (Fig 2) showing the higher boiling point components. Wrap around is visible on the 10:0 FAME. All compounds between 11:0 and 20:1 are wrapped around once, and 20:0 is wrapped around twice, as are all the components eluting after 20:1. Separation of all the compounds is good and this chromatogram could be easily used for quantitative work. The components are not separated for and, but this is not to be expected on a 20 m column. A longer column would permit this to be realized. An unknown component appears in the main chromatogram between the and compounds.

3 Figure 1: GCxGC-FID Chromatogram of the NLEA FAME Standard Mix 10:0 17:0 20:1 cis & trans 15:0 6:0 13:0 16:1 cis & trans 11:0 12:0 14:1 18:3 20:0 Figure 2: Expanded GCxGC-FID Chromatogram of the NLEA FAME Standard Mix 20:1 24:0 22:0 23:0 20:5 22:6 24:1 20:0 22:1

4 A biodiesel sample produced from palm oil, supplied by LECO Thailand, was analysed next. The same conditions were used, except that the sample was diluted before analysis, and the results of this analysis are shown in Fig 3 below. The FAME components are well separated and can be easily identified by comparison with the standard sample. A number of additional compounds are present in this sample which were not seen in the standard. Figure 3: GCxGC-FID Chromatogram of the Palm Oil Biodiesel Sample 12:0 Finally a biodiesel sample produced from Jatropha Curcas Linn, and also supplied by LECO Thailand was analysed. The same conditions were used and the sample was diluted before analysis. The results for this analysis are shown in Fig 4. In this sample many extra components are seen, which are not present in the NLEA FAME Standard mixture. The identity of these compounds cannot be determined from the FID run, and it would be necessary to repeat the analysis using GCxGC-TOFMS to unambiguously identify these compounds. As can be seen the major biodiesel FAME components,, 16:1,, and are easily located and could be quantified routinely. Again use of a longer column would facilitate the separation of isomers for and should this be deemed necessary.

5 Figure 4: GCxGC-FID Chromatogram of the Jatropha Curcas Linn Biodiesel Sample 16:1 Conclusions An NLEA FAME standard sample was analysed by GCxGC-FID using a polar : non-polar column combination. All components of the sample were separated easily, although no separation of isomers was deemed necessary for biofuel samples, and this was not attempted. Using the results obtained from the FAME standard, two biodiesel samples were analysed. These samples were produced from Palm Oil, and from Jatropha Curcas Linn. The major FAME components found to be present in these samples were,, and. This was determined by comparison of the results with those obtained from the standard. Acknowledgements The samples were run on the GCxGC-FID instrument belonging to the National Metrology Institute of South Africa (NMISA). Ms Betty-Jayne de Vos provided the NLEA FAME Standard sample, and assisted with the analyses.