Analysis of Volatile Fragrance and Flavor Compounds by Headspace

Determination of Xylitol in Sugar-Free Gum by GC–MS with Direct Aqueous Injection: A Laboratory Experiment for Chemistry Students. Journal of Chemic...
0 downloads 0 Views 86KB Size
In the Laboratory

Analysis of Volatile Fragrance and Flavor Compounds by Headspace Solid Phase Microextraction and GC–MS An Undergraduate Instrumental Analysis Experiment Randolph C. Galipo, Alfredo J. Canhoto, Michael D. Walla, and Stephen L. Morgan* Department of Chemistry and Biochemistry, The University of South Carolina, Columbia, SC 29208

This experiment is intended to familiarize senior-level undergraduate students with analytical separations by way of a hands-on instrumental application. Laboratory experiments that employ high-resolution capillary gas chromatography coupled with mass spectrometry (GC–MS) make excellent “real-world” learning experiences (1–3). GC–MS can be used to separate volatile, thermally stable compounds and is particularly advantageous for the analysis of complex samples. GC–MS is widely used in industrial and research laboratories, and student exposure to these modern hyphenated techniques is essential for a well-rounded laboratory program. The main learning goals of undergraduate experiments involving gas chromatography usually include GC separation fundamentals and instrumentation, and qualitative compound identification via mass spectrometry. These goals can be addressed by having students separate a complex mixture by GC and then decide what components are present using MS. Such an exercise is more interesting to the students if they get to choose the sample. However, the samples chosen should be amenable to analysis during a 3-hour laboratory period and must not adversely affect the instrumentation. Students should not choose samples without guidance from the instructor because they may pick samples that are too difficult to analyze. Samples may be considered difficult because of involved sample preparation. One class of samples that comes to mind is shampoos. Students are disappointed to find out that shampoos cannot just be injected directly into a gas chromatograph. Before a sample is suitable for GC–MS analysis, cleanup steps involving extractions may be necessary to separate volatile analytes from background matrices. A relatively new extraction technique known as solid phase microextraction (SPME) minimizes sample preparation and concentrates volatile analytes in a solvent-free manner. Background SPME was developed by Pawliszyn’s research group at the University of Waterloo in the late 1980s (4). SPME is a sensitive, reproducible, cost-efficient, solventless technique that incorporates extraction, concentration, and sample introduction into a single step (5). A syringe-like device (Fig. 1) with an outer septum-piercing needle and a plunger houses a fused silica fiber coated with a stationary phase. The fiber can be inserted into the sample matrix (aqueous samples) or the gaseous phase above the sample (headspace). Liquids can be sampled by inserting the fiber directly into the solution. Volatile analytes from solids can be sampled by inserting the fiber into the headspace region above the sample. Analytes are partitioned *Corresponding author. Email: [email protected].

between the stationary-phase coating and the gas phase when equilibrium is established. After concentration of analytes on the fiber, the syringe assembly is inserted into the injection port of a gas chromatograph, where the analytes are thermally desorbed from the fiber and cold-trapped on the head of the capillary column. If an unknown sample has volatile components that can be detected by the human nose, SPME coupled to GC or GC–MS might be employed to identify and quantitate those compounds. Thus, a common industrial application of SPME might involve analysis of odors from consumer products. Applications of SPME have included extraction of environmental contaminants from aqueous matrices (4, 6 ), headspace extraction of flavor and fragrance compounds (7), and forensic investigations of drugs of abuse in biological fluids (8–10). This paper presents a SPME experiment we designed for undergraduate classes at The University of South Carolina. SPME sampling facilitates extraction of volatile flavor and fragrance compounds from a broad range of samples selected by students. For example, students in our classes have analyzed commercial products such as candies, gums, perfumes, and shampoos. Most of these samples were household items easily obtained by students. Criteria for sample selection include strong odor; volatile, noncorrosive components; and availability of a set of similar samples to compare different sample types. Extracts were analyzed by GC–MS to identify major constituents.

plunger

plunger retaining screw

septum piercing needle

adjustable needle guide/depth gauge

septum fiber attachment needle fiber

liquid/solid sample

Figure 1. Schematic of the headspace SPME apparatus.

JChemEd.chem.wisc.edu • Vol. 76 No. 2 February 1999 • Journal of Chemical Education

245

In the Laboratory

Experimental Procedure Results from the analysis of chewing gums and shampoos are presented here. Chewing gums were cut into small squares and approximately half of a gram was transferred to GC vials and capped. Approximately 200 µ L of shampoos and perfumes was transferred with disposable Pasteur pipets to GC vials and capped. Vials of different sizes are appropriate as long as they can be tightly sealed. The SPME apparatus and fibers (7 µ m bonded polydimethylsiloxane) were obtained from Supelco (Bellefonte, PA). An SPME kit, which includes the sample holder and a package of fiber assemblies (one fiber each of 85 µm polyacrylate coating, and 100- and 7-µ m bonded polydimethylsiloxane coating), can be purchased from Supelco for about $330. Sampling was performed by inserting the syringe needle of the SPME assembly through the septum cap into the headspace above the sample. Volatiles were sorbed by extending the fiber into the headspace. After adequate sorption time, the fiber was withdrawn into the outer septum-piercing needle, removed from the vial, and subsequently desorbed in the heated injection port of the GC. Chewing gums R and S were adsorbed for 5 and 10 min, respectively. Both chewing gums were desorbed for 2 min at 250 °C. Shampoos X, Y, and Z were adsorbed for 5 min. All samples were desorbed for 2 min at 250 °C, except shampoo X, which was desorbed for 1 min at 250 °C. Fibers can be cleaned to avoid analyte carry-over by placing them in the injection port of the gas chromatograph for a few minutes while maintaining the oven temperature at 280 °C. Any remaining analytes are in this way removed from the stationary phase of the fiber. For comparison purposes, conventional headspace sampling was performed by inserting an airtight syringe needle through the septum cap into the headspace above the sample and withdrawing 25 µl of the headspace gas. The syringe was then inserted into the injection port of a GC–MS instrument. All extracts were analyzed with a Hewlett-Packard (HP, Palo Alto, CA) 5890 gas chromatograph interfaced to an HP 5970 series mass selective detector (MSD). A Merlin (Merlin, Half Moon Bay, CA) microseal was used in place of a standard septum. A 30-m × 0.25-mm i.d. × 0.25-µ m Rtx®-5 (95% dimethyl–5% diphenyl polysiloxane) capillary column (Restek, Bellefonte, PA) was used for all separations. SPME fibers were desorbed under splitless conditions. For all samples, the gas chromatograph oven temperature was held at 40 °C during desorption, increased to 250 °C at 10 °C/min, and then held for 2 min (run time varied depending on desorption time). Injection port and transfer line temperatures were set at 250 and 270 °C, respectively. The GC carrier gas was helium at a column head pressure of 5 psi. The mass spectrometer was operated in electron impact mode and tuned to perfluorotributylamine (PFTBA). The mass spectrometer was scanned from m/z 50 to 350. After a short yet simple demonstration of the MS ChemStation software by the course instructor, students should be able to select peaks for identification. In our experience, students are increasingly computer-literate and have no difficulty with operation of the Microsoft Windows-based HP ChemStation software. The software allows ready identification of mass spectra by reference comparison to spectra in a standard library database system (NIST). 246

Figure 2. Total ion chromatograms from GC–MS analyses of shampoo X: (A) headspace SPME; and (B) conventional headspace. Peak identification: (1) limonene, (2) 3,7-dimethyl-1,6-octadien-3-ol, (3) unknown, (4) dodecane, (5) 1,6-octadien-3-ol, 3,7-dimethyl acetate (linalyl acetate), (6) unknown.

Results and Discussion Chewing gums and shampoos were among the many interesting samples analyzed. Perfumes were another popular choice of the students. These samples were easily extracted by SPME fibers owing to their volatility. A comparison of conventional headspace analysis with that of headspace SPME of shampoo X is shown in Figure 2. Although the same compounds were extracted by both techniques, a noticeable difference in total signal is evident. The total ion abundance of the headspace SPME GC–MS results was about 100 times greater than the result from conventional headspace sampling (based on analyses of shampoo X performed in our laboratory). Examples of headspace SPME/GC–MS results for shampoos and chewing gums are shown in Figures 3 and 4, respectively. A major volatile component of shampoos Y and Z (Fig. 3) was determined to be 2-(1-dimethylethyl)-cyclohexanol. The abundance of limonene (found in oils of lemon and orange) was greater in the total ion chromatogram (TIC) of shampoo Z than in that of shampoo Y. The number of volatile components extracted from shampoo X was also greater than that extracted from shampoo Y. Limonene was identified in many of the samples analyzed (perfumes, chewing gums, shampoos). The HP ChemStation software permits students to select a peak from the total ion chromatogram and immediately view its mass spectrum. Figure 5 presents mass spectra for limonene (1-methyl-4-(1-methylethenyl)cyclohexene, molecular weight

Journal of Chemical Education • Vol. 76 No. 2 February 1999 • JChemEd.chem.wisc.edu

In the Laboratory

Figure 3. Total ion chromatograms from headspace SPME GC–MS analyses of: (A) shampoo Y; and (B) shampoo Z. Peak identification: (1) limonene, (2) allyl heptanoate, (3) dodecane, (4) 2-(1dimethylethyl)-cyclohexanol, (5) phenoxy ethyl isobutyrate.

Figure 4. Total ion chromatograms from headspace SPME GC–MS analyses of (A) chewing gum R; (B) chewing gum S. Peak identification: (1) ethyl butyrate, (2) isobutyl propionate, (3) limonene, (4) eucalyptol, (5) L-menthone, (6) menthol, (7) methyl salicylate, (8) menthyl acetate, (9) caryophyllene, (10) BHT.

136.23), and menthol (5-methyl-2-(1-methylethyl)-cyclohexanol, molecular weight 156.26). Exploring results interactively provides an opening for teaching the fundamentals of mass spectrometry. The molecular ion peak in Figure 5A, located at m/z 136, represents the molecular weight of limonene. Depending on the time allotted, other peaks might also be interpreted to identify the structure: in Figure 5A, the peaks at m/z 121 and 107 represents loss of methyl (– CH3) and ethyl (–C2H5) groups, respectively. In Figure 5B, menthol fragments with loss of water to produce the peak at m/z 138; loss of methyl and ethyl fragments are seen at m/z 123 and 109, respectively. The software provides a matchquality index assessing the percent similarity between mass spectra of peaks in the TIC and reference spectra from the NIST library. The mass spectrum in Figure 5A produces a match quality of 94% against the limonene library spectrum; the mass spectrum in Figure 5B matches that of menthol with 91% match quality. Chewing gums vary from flavor to flavor, as shown by the differences in the TICs (Fig. 4). The TIC of chewing gum R (a mint-flavored gum) consisted mostly of early-eluting compounds, with the most abundant compound identified as limonene. Menthol, L-menthone, and menthyl acetate (obtained from peppermint or mint oils) were just a few of the volatile flavor components identified in the TIC of chewing gum S (a fruit-flavored gum). These components were also identified in candies (data not shown). It is interesting to note that butylated hydroxytoluene (BHT), an antioxidant, was also identified in the TICs of both chewing gums.

Comparing and contrasting volatile sample components was found to be of particular interest to the students. This experiment was purely an educational experience for them, and samples were chosen based on convenience and simplicity, not novelty. Similar analyses, involving the extraction of peppermint oil from cookie bars, spearmint flavor from chewing gum, and fragrance compounds from perfume, have been published in technical literature by Supelco (11). There are a few noteworthy points from our experiences with SPME. First of all, the fiber assembly is fragile, making it possible to bend the outer syringe needle or damage the fiber. Although the fiber assembly is easy to use, great care must be taken to avoid damaging the fiber. Second, we liked the 7-µm bonded polydimethylsiloxane fiber because of its robustness. Bonded-phase fibers can be submersed directly into a liquid matrix during sampling and can be rinsed with organic solvents. Another point of concern involves injection seals. We chose to use a Merlin microseal instead of a septum for several reasons. The diameter of the outer syringe needle cores standard septa easily, requiring them to be replaced frequently. Predrilled septa work better and last longer, but the septum retainer nut has to be tightened beyond normal tightening to keep them from leaking. Owing to the nature of the septum retainer nut, the fiber assembly will not rest firmly unless it is held in place. We found that a ring stand with a clamp works well for holding the fiber assembly in place during desorption. A Merlin microseal not only lasts longer than standard septa, but also allows the base of the fiber assembly to rest freely on the microseal’s housing.

JChemEd.chem.wisc.edu • Vol. 76 No. 2 February 1999 • Journal of Chemical Education

247

In the Laboratory

Conclusion This laboratory experiment presents an opportunity for students to use a relatively new, simple, yet rapid, solventless technique to extract volatile fragrance and flavor components from the headspace region of samples. A typical experiment involving SPME requires minimal sample preparation and analysis times are under 30 min. Multiple analyses can be performed within the time span allotted for a typical laboratory class period. Students can be creative in choosing samples of personal interest. SPME offers a relatively inexpensive way to analyze fairly complex samples such as shampoos and chewing gums that might otherwise be difficult to analyze by conventional headspace analysis. Literature Cited

Figure 5. Mass spectra: (A) limonene (shampoo X); (B) menthol (chewing gum S).

Although the samples reported in this paper were adsorbed at room temperature, samples can be heated to increase the concentration of volatiles in the headspace region. We desorbed volatiles from the sampling fiber with the oven temperature set at 40 °C. Cold-trapping analytes on the column by cryogenic cooling might also be used to sharpen peaks. This experiment can also be conducted by gas chromatography with a flame ionization detector (FID) if a GC–MS instrument is not available. Quantitative SPME experiments (12–14 ) can also be performed. For example, Yang et al. (15) described determination of caffeine in coffee, tea, and soft drinks by SPME/GC–MS using trimethyl 13C-labeled caffeine as an internal standard.

248

1. Kostecka, K. S.; Lerman, Z. M.; Angelos, S. A. J. Chem. Educ. 1996, 73, 565–567. 2. Elderd, D. M.; Kildahl, N. K.; Berka, L. H. J. Chem. Educ. 1996, 73, 675–676. 3. McGoran, E. C.; Melton, C.; Taitch, D. J. Chem. Educ. 1996, 73, 88–92. 4. Arthur, C. L.; Pawliszyn, J. Anal. Chem. 1990, 62, 2145–2148. 5. Zhang, Z.; Yang, M. J.; Pawliszyn, J. Anal. Chem. 1994, 66, A844–A852. 6. Potter, D. W.; Pawliszyn, J. Environ. Sci. Technol. 1994, 28, 298– 305. 7. Yang, X.; Peppard, T. J. Agric. Food Chem. 1994, 42, 1925–1930. 8. Nagasawa, N.; Yashiki, M.; Iwasaki, Y.; Hara, K.; Kojima, T. Forensic Sci. Int. 1996, 78, 95–102. 9. Yashiki, M.; Kojima, T.; Miyazaki, T.; Nagasawa, N.; Iwasaki, Y.; Hara, K. Forensic Sci. Int. 1995, 76, 169–177. 10. Brewer, W. E.; Galipo, R. C.; Morgan, S. L.; Habben, K. H. J. Anal. Toxicol. 1997, 21, 286–290. 11. Technical Bulletin #869; Supelco Inc.: Bellefonte, PA, 1995. 12. Hawthorne, S. B.; Miller, D. J.; Pawliszyn, J.; Arthur, C. L. J. Chromatogr. 1992, 603, 185–191. 13. Fromberg, A.; Nillson, T.; Larsen, B. R.; Montanarella, L.; Facchetti, S.; Madsen, J. O. J. Chromatogr. A 1996, 746, 71–81. 14. Clark, T. J.; Bunch, J. E. J. Chromatogr. Sci. 1996, 34, 272–275. 15. Yang, M. J.; Orton, M. L.; Pawliszyn, J. J. Chem. Educ. 1997, 74, 1130–1132.

Journal of Chemical Education • Vol. 76 No. 2 February 1999 • JChemEd.chem.wisc.edu