Short Path Thermal Desorption is a versatile technique for the analyses of solid, liquid and gaseous samples. Previous articles in this newsletter have described the theory and operation of short path thermal desorption (1) and have demonstrated its versatility in the analysis of volatile organic chemicals (VOC's) and semi-volatile organic chemicals in solid matrices (2 & 3). The detection and analysis of VOC's in water is routinely performed utilizing Purge and Trap techniques with standard EPA methodology (4). The detection of VOC's in other water-based matrices presents its own unique problems. The purpose of this article is to demonstrate the versatility of the short path thermal desorption technique to the analyses of VOC's in commercial water based products including liquid formulations, colloidal suspensions, and liquid pastes. Using a newly designed liquid purging system for the collection of the VOC's followed by trapping on an adsorbent trap and subsequent analysis via the Short Path Thermal Desorption technique, it is possible to detect and identify the various flavors, fragrances, off-flavors, off-odors, and manufacturing by-products in this wide diversity of liquid samples. This technique can easily be incorporated into a troubleshooting technique to detect problems in a wide variety of liquid products, to compare various competing manufacturers products, as well as a quality control program.
Volatile organic compounds are present in commercial liquid formulations, a diverse range of colloidal suspensions, and liquid pastes. Flavors and fragrances are added for their aesthetic values and to increase consumer appeal. The flavor/fragrance qualities of liquid commercial products is greatly dependent on the plethora of volatile and semivolatile organic compounds contained both in the liquid matrix and the headspace aroma (5). Volatile compounds are also used in the manufacturing process to obtain the desired physical properties. Trace residues of these manufacturing by-products are often present in the final products. Analytical techniques are needed to profile and identify flavors, fragrances, off-flavors, off-odors and potential contaminants that may be present as flavor and fragrance additives, residual solvents from the manufacturing process or as impurities in raw materials. Results of such analyses may be used for developmental research as well as production quality control.
The detection of VOC's from liquids can be accomplished by using a headspace sampling technique combined with gas chromatography (GC) and/or gas chromatography/mass spectrometry (GC/MS). The most sensitive of the techniques is the dynamic headspace technique. Dynamic headspace analysis depends on the vapor phase extraction of the sample liquid achieved by purging it with an inert gas. VOC's present will diffuse into the vapor phase and are collected on a solid sorbent material, as Tenax®. Older desorption type purge and trap methodology and apparatus, such as utilized by the EPA, is only applicable to very volatile components, as solvents.
However, the Short Path Thermal Desorption system permits the analyses of liquid samples by desorbing the samples previously collected on adsorbent resins directly into the GC injection port for subsequent analyses by conventional GC detectors or via mass spectrometers. Due to its Short path of sample flow, this new system overcomes the shortcomings of previous desorption systems by eliminating transfer lines which are easily contaminated by samples and by providing for the optimum delivery and therefore maximum sensitivity of samples to the GC injector via the shortest path possible (1). By making the transfer path as short as possible, the maximum sample size is delivered to the GC, samples are not lost or destroyed in hot transfer lines, and no memory effects occur due to contamination of transfer lines from previous samples. The high recovery rates, time savings for sample preparation and the high sensitivity of short path thermal desorption make this not only an attractive technique to the analyst but an essential detection apparatus because of toxicology and quality control, as well as a means for determining human health impacts.
The objective of this investigation is to demonstrate the feasibility of combining a headspace sampling technique with the technique of Short Path Thermal desorption to analyze VOC's that can be purged from diverse liquids, colloidal suspensions, or water-based pastes. Commercial beverages including fruit drinks, carbonated colas, and wine coolers as well as olive oil, shampoo, latex paint, toothpaste and a spiked water sample were analyzed by thermal desorption-gas chromatography-mass spectrometry (TD-GC-MS) using a Scientific Instrument Services (S.I.S.) model TD-1 Short Path Thermal Desorber accessory.
A. Liquid Purging System
Figure 1a. Schematic of Enhanced Dry Purge Headspace Sampling Apparatus With 5 ml Test Tube
Figure 1b. Liquid Purging System
Samples were collected using a S.I.S. enhanced dry purge headspace sampling apparatus. This apparatus (Fig. 1a) consists of a sparge gas inlet connected to stainless steel purging needle that is inserted through an adaptor fitting into a 50 ml round bottom flask or into a 5 ml test tube (Figs. 1a & b). The dry purge gas inlet is located at a right angle to the sparge gas inlet at the top of the apparatus (Figs. 1a & b). This can be left in the closed or open position. The purpose of the dry purge is to reduce the water vapor condensation on the adsorbent trap. This problem can be especially troublesome when isolating volatiles from aqueous solutions at high temperatures. Although the adsorbent traps packed with Tenax have a low affinity for water, it is inevitable that some condensation will occur in the trap due to the high relative humidity of the sparge gas as it exits the apparatus. When moisture condenses on the adsorbent, it can block the pores of the resin matrix and thereby drastically reduce the diffusion of volatile organics into the trapping agents. This will result in reduced trapping efficiency. Opposite the dry purge inlet is the connector for the glass-lined stainless steel (GLT) desorption tube (Fig 1b). Many different sizes of glassware can be used as the collection vessel for the sample depending on the size of the sample to be analyzed.
The liquid purging system contains two ball rotameters with adjustable needle valves mounted on a stationary base and permit the visual indication and independent adjustment of the carrier gas flow to each of the gas inlets (Fig. 1b). The 150 mm long flow tube contains a sapphire ball for flow ranges of 0 to 72 ml/min of helium or air.
B. GC, GC/MS
Initial screening of samples was performed on a Varian 3400 GC with a flame ionization detector in order to determine the sample size and thermal desorption temperature and time. Split ratios of 1 to 20 were used for overloaded samples. Otherwise, direct splitless analysis was used. The GC column was a 30 meter x .32 mm i.d. DB-5 capillary column containing a 1.0 um film thickness with a flow rate of 2.0 ml per min (He). The column was temperature programmed from -40 degrees C (hold for 10 minutes to cryofocus during thermal desorption interval) to 280 degrees C at a rate of 10 degrees C per minute.
Sample identification was performed on an HP 5890 GC interfaced to an HP 5971 MSD. The GC injector was maintained at 260 degrees C. The GC column was a 25 meter x .25 mm i.d. DB-5 capillary column containing a 0.25 um film thickness with a flow rate of 1.0 ml per minute (He). The column was temperature programmed from -40 degrees C (hold for 10 minutes to cryofocus during thermal desorption interval) to 280 degrees C at a rate of 10 degrees C per minute.
C. Short Path Thermal Desorption System
All experiments were conducted using a S.I.S. model TD-1 Short Path Thermal Desorber accessory connected either to a Varian 3400 gas chromatograph or an HP 5890 GC interfaced to an HP 5971 MSD. The theory and operation of the Short Path Thermal Desorption System was previously described in this newsletter (1). GLT desorption tubes were purged for 10 minutes at 40 ml/min to remove any excess water and then thermally desorbed for 10 minutes at a temperature of 150 degrees C to 200 degrees C.
Samples to be analyzed are collected on silanized glass-lined stainless steel desorption tubes packed with 100 mg. of Tenax TA 60/80 mesh. The ends of the tubes are plugged with silanized glass wool approximately one centimeter on each end. Using an S.I.S. Desorption Tube Conditioning System, traps are then temperature programmed from ambient temperature to 320 degrees C at a rate of 4 degrees C per minute while purging with helium at a flow rate not less than 20 ml per minute. The traps are held at the upper temperature limit for not less than four hours. After conditioning, the tubes are promptly sealed on both ends using column end caps fitted with PTFE ferrules. Traps prepared in this manner exhibit excellent adsorptive capacity and contain no organic background (bleed or artificial peaks) when analyzed by GC-MS.
Sample sizes of 1 to 25 mls of various commercial products were measured in a graduated cylinder and transferred to a 5 ml test tube or a 50 ml round bottom flask. Samples were sparged with helium at 15 to 30 ml/min with an additional 15 to 30 ml/min dry purge for 10 to 15 minutes using a S.I.S. enhanced dry purge headspace sampling apparatus. All samples were collected at room temperature. Volatile analytes were gas extracted and carried to a preconditioned 4.0 mm i.d. glass-lined stainless steel desorption tube packed with 100 mg of Tenax TA. The tubes were then fitted with S.I.S. syringe adaptors and connected to the TD-1 Thermal Desorber.
Results and Discussion
The following liquid based products depict the wide diversity of commercial liquid products that can be analyzed via this technique. All the samples were analyzed by purging as described above at room temperature. Additional compounds of less volatility could have been detected if the samples were heated, and this may be desirable in many instances. Liquid samples with low concentrations of foreign materials such as the spiked water sample are the easiest type of sample to analyze. No foaming occurred and the volatiles are easily purged from these samples. The commercial beverages contain many more compounds and begin to foam. In many instances, it is necessary to use larger vessels, foam breakers, or other techniques to minimize these foam forming problems. Colloidal solutions, as the latex paint and shampoo, present additional problems due to their thick nature. It may be preferable to dilute these samples with additional water to make them less viscous and enhance the purging of the VOC's. Liquid pastes, as toothpaste, are the final extreme of sample density and need to be diluted in order to purge the volatiles. In addition, foaming is a major problem and large vessels (relative to the sample size) are required to prevent the foam from entering the desorption trap.
A calibration mixture (Polyscience) containing 2 mg/ml in methanol of benzene, toluene, 2-bromo-1-chloropropane (int. std.), m-xylene, p-isopropyltoluene and naphthalene was diluted in glass distilled water with a final concentration of 20 ppm. Figure 2 shows a chromatogram of the components as they are separated by gas chromatography. A desorption temperature of 200 degrees C proved to be the optimum temperature for the detection of the less volatile compound naphthalene.
Figure 2 - Gas Chromatograph of Calibration Mixture, 50 ml. Collected For 15 min at 20 ml/m with 20 ml/m Dry Purge Thermally Desorbed at 200 Degrees C for 10 min.
Several brands of non-alcoholic fruit drinks were studied to identify the flavor volatiles present. The fruit drinks were found to contain numerous mono-and sesquiterpenoid compounds and flavors such as ethyl acetate, ethyl butyrate and benzaldehyde (Figs. 3, 4 & 5). The negative control (empty desorption tube) was found to be free of interferences with the exception of two minor peaks which were organosilicone compounds from the injection port system bleed. Monoterpenes were identified as limonene in orange, cranberry and raspberry drinks (Figs. 3, 4 & 5) and terpinene and myrcene in orange (Fig. 3). Sesquiterpenes included cadinene in orange (Fig. 3) and caryophyllene in raspberry (Fig. 5). The Merck Index lists these compounds as constituents of several plant derived essential oils which are used as fragrance materials.
Figure 3 - Orange drink, 25 ml. Collected For 10 min at 30 ml/m With 30 ml/m Dry Purge Thermally Desorbed At 150 Degrees C For 10 min.
The flavor ethyl acetate was detected in all drinks with ethyl butyrate found in orange and cranapple (Figs. 3 & 4). High concentrations of the flavors furfural and 1-butanol, -3-methyl-, acetate were found in cranberry (Fig. 4) and raspberry (Fig. 5), respectively. A trace amount of chloroform was identified in orange (Fig. 3) as well as trace amounts of methyl chloroform in cranapple and raspberry (Figs. 4 & 5). The aromatic compounds toluene and xylene were also detected in raspberry (Fig. 5). Trace amounts of these compounds may be present as residual solvents from the manufacturing process or occur naturally. Styrene which may have leached from the plastic cap liner was found in cranapple (Fig. 4).
Figure 4 - Cranapple drink, 25 ml. Collected For 10 min at 30 ml/m With 30 ml/m Dry Purge Thermally Desorbed At 150 Degrees C For 10 min.
Figure 5 - Orange drink, 25 ml. Collected For 10 min at 30 ml/m With 30 ml/m Dry Purge Thermally Desorbed At 150 Degrees C For 10 min.
Several brands of carbonated colas were analyzed to determine the feasibility of this technique to distinguish the different manufacturers brands. Monoterpenes such as camphene, cymene, limonene and terpinene were identified in both of the carbonated colas, in addition to the alcohols endo-fenchol and endo-borneol (Figs. 6 & 7). While high concentrations of cymene and limonene were detected in Cola A (Fig. 6), Cola B was found to contain high concentrations of isocineole and 1,8 cineole (Fig. 7). Trace amounts of chloroform, most likely from water treatment, were found to be present in each of the colas. Cola A was also found to contain the aromatic compound toluene (Fig. 6) which may be present as a natural product from essential oils, synthetic or as a residue from the manufacturing process.
Figure 6 - Cola A, 25 ml. Collected for 10 min At 15 ml/m With 10 ml/m Dry Purge Thermally Desorbed At 150 Degrees C For 10 min.
Figure 7 - Cola B, 25 ml. Collected For 10 min At 15 ml/m With 10 ml/m Dry Purge Thermally Desorbed At 150 Degrees C For 10 min.
Alcohol containing wine coolers were analyzed to demonstrate the variation of the flavor compounds in these drinks. The dry purge step in the sample set-up purges the ethanol from the sample in addition to water. Even though each wine cooler had its own distinct chromatograph, they were found to contain common compounds such as the flavors ethyl acetate, ethyl butyrate, ethyl caproate and propyl acetate, as well as several monoterpenes with limonene the most common (Figs. 8 & 9). Additional compounds which were found in the wine coolers included aldehyde and alcohol derivatives.
Figure 8 - Wine Cooler A, 25 ml. Collected For 10 min at 30 ml/m With 30 ml/m Dry Purge 20:1 split Thermally Desorbed At 150 Degrees C For 10 min.
Figure 9 - Wine Cooler B, 25 ml. Collected For 10 min At 30 ml/m With 30 ml/m Dry Purge 20:1 Split Thermally Desorbed At 150 Degrees C For 10 min.
The analysis of olive oil demonstrates the detection of volatiles from other non-water based liquids. A series of straight chain hydrocarbons and linear alkenes such as 1- and 2-octene from lipid oxidation decomposition products were identified in the olive oil sample (Fig. 10). Minor constituents, as ketones, aldehydes and alcohols, were identified, as well as trace amounts of the aromatic compounds benzene and toluene (Fig. 10). The presence of these compounds may be the result of the manufacturing process or occur naturally from essential essences. Olive oil was also found to contain the compound styrene.
Figure 10 - Olive Oil, 5 ml. Collected For 10 min at 20 ml/m With 20 ml/m Dry Purge 20:1 Split Thermally Desorbed At 150 Degrees C For 10 min.
A popular hair shampoo was analyzed to determine the volatiles which could be purged from a thick water based suspension. Shampoo was found to contain numerous monoterpenoid compounds such as alpha and beta pinene, camphene, terpinene, linalool and a high concentration of limonene (Fig. 11). Several ketones, in addition to a trace amount of the solvent 1,4-dioxane for removing oils from hair, were identified. A high concentration of the fragrance compound benzyl acetate which occurs in a number of plants, particularly jasmine, and linalyl acetate, a valuable constituent of bergamot and lavender oils, were also detected (Fig. 11). Additional compounds included a series of sesquiterpenoid compounds such as alpha-copaene and caryophyllene.
Figure 11 - Shampoo, 5 ml. Collected For 10 min at 20 ml/m With 20 ml/m Dry Purge 20:1 Split Thermally Desorbed At 150 Degrees C For 10 min.
The identification and quantitation of VOC's in latex paints is of importance to paint manufacturers. A latex enamel paint was found to contain over 100 volatile organic compounds (Fig. 12). Numerous straight and branched chain hydrocarbons and alcohols were identified. Butyl esters of propanoic and butanoic acids were also found in latex enamel paint (Fig. 13). In addition to a number of ketone compounds, trace amounts of the aromatic compounds benzene, toluene and other benzene derivatives were present (Fig. 13).
Figure 12 - Gas Chromatograph of Latex Enamel Paint, 1 ml Collected For 10 min at 15 ml/m With 15 ml/m Dry Purge Thermally Desorbed At 150 Degrees C For 10 min
Figure 13 - Latex Enamel Paint, 1 ml Collected For 10 min at 15 ml/m With 15 ml/m Dry Purge Thermally Desorbed At 150 Degrees C For 10 min
A method is needed to detect the volatile manufacturing by products in toothpaste. Liquid paste, as toothpaste, was diluted 1:5 in glass distilled water to enhance the purging of VOC'S. The collection time was reduced to 6 minutes due to sample foaming; however, the liquid paste was found to contain over 50 VOC'S (Fig. 14). The sample foams precautions must be taken to prevent the foam from reaching and contaminating the adsorbent trap. This can be accomplished using a foam breaker such as a distillation trap which is usually used on rotary evaporators. It is fitted directly below the adsorbent trap. As an alternative, a larger collection vessel or a tube packed with a small amount of glass wool often suffices to break the foam and allow the liquid to drop back down into the vessel. A heated dynamic headspace purging technique may also be used where helium or another suitable gas is purged over the surface of a heated sample.
Figure 14 - Gas Chromatography of Toothpaste Diluted In Glass-Distilled Water (1.5), 5 ml Collected For 6 min at 20 ml/m With 20 ml/m Dry Purge Thermally Desorbed At 150 Degrees C For 10 min
The S.I.S. model TD-1 Short Path Thermal Desorber accessory, used in combination with GC-MS is an ideal instrument for the detection of volatile organics from a liquid matrix. It can quickly and easily analyze a small amount of sample to yield both qualitative and quantitative data on volatile organic composition. In this study, the apparatus was successfully employed for the determination of volatile components in a variety of liquid matrices. Combined with the dynamic headspace sampling technique, the technique of Short Path Thermal Desorption can be used during the production of commercial products, for quality control of flavor/fragrance additives as well as for the detection of residual solvents.
Food Science Applications. Food Technology. Vol. 45 (7): 104-105.
EPA Methods for the Determination of Organic Compounds in Drinking Water. 1988. Environmental Monitoring Systems Laboratory-Cincinnati. EPA-600/4-88/039. 378 pp.
Hartman, T. G., J. Lech and R. T. Rosen. 1990. Determination of Off-Odors and Other Volatile Organics in Food Packaging Films by Direct Thermal Analysis-GC-MS. The Mass Spec Source. Vol. XIII (4): 30-33.
Hartman, T. G., S. V. Overton, J. J. Manura, C. W. Baker and J. N. Manos. 1991. Short Path Thermal Desorption:
Manura, J. J. 1991. Direct Thermal Analysis Using the Short Path Thermal Desorption System. The Mass Spec Source. Vol. XIV (1): 22-27.
Manura, J. J., S. V. Overton, C. W. Baker and J. N. Manos. 1990. Short Path Thermal Desorption-Design and Theory. The Mass SpecSource. Vol. XIII (4): 22-28.