Anal. Chem. 2001, 73, 971-977
Analysis of a GC/MS Thermal Desorption System with Simultaneous Sniffing for Determination of Off-Odor Compounds and VOCs in Fumes Formed during Extrusion Coating of Low-Density Polyethylene Kirsi Villberg*,† and Anja Veijanen‡
Chemical Technology, Processes and Environment, Technical Research Centre of Finland, P.O. Box 1403, FIN-02044 VTT Espoo, Finland, and Department of Bological and Environmental Science, University of Jyva¨skyla¨, P.O. Box 35, FIN-40351 Jyva¨skyla¨, Finland
A thermal desorption equipment introducing volatile organic compounds (VOCs) into the gas chromatographic/ mass spectrometric system (GC/MS) with simultaneous sniffing (SNIFF) is a suitable method for identifying the volatile organic off-odor compounds formed during the extrusion coating process of low-density polyethylene. Fumes emitted during the extrusion coating process of three different plastic materials were collected at two different temperatures (285 and 315 °C) from an outgoing pipe and near an extruder. The VOCs of fumes were analyzed by drawing a known volume of air through the adsorbent tube filled with a solid adsorbent (Tenax GR). The air samples were analyzed by using a special thermal desorption device and GC/MS determination. The simultaneous sniffing was carried out to detect off-odors and to assist in the identification of those compounds that contribute to tainting and smelling. The amounts of offodor carbonyl compounds and the total content of the volatile organic compounds were determined. The most odorous compounds were identified as carboxylic acids while the majority of the volatile compounds were hydrocarbons. The detection and quantification of carboxylic acids were based on the characteristic ions of their mass spectra. The higher the extrusion temperature the more odors were detected. An important observation was that the total concentration of volatiles was dependent not only on the extrusion temperature but also on the plastic material. Manufacturers are usually interested in identifying the off-flavor compounds in their plastics. Being widely used by consumers, the quality aspect of polyolefins is very important, especially for the materials used in food packaging. Even though most identified compounds from polyethylene (PE) samples are hydrocarbons, they are rarely the reasons for off-odors.1 †
Technical Research Centre of Finland. University of Jyva¨skyla¨. (1) Villberg, K.; Veijanen, A.; Gustafsson, I.; Wickstro ¨m, K. J. Chromatogr. 1997, 791, 213-219. ‡
10.1021/ac001114w CCC: $20.00 Published on Web 01/30/2001
© 2001 American Chemical Society
Carbonyl compounds such as aldehydes, ketones, and carboxylic acids are known to be the off-flavor compounds in PE.2-6 Linssen et al.3 noted the taint in water packed in pouches made of the extrusion coating of low-density polyethylene (LDPE) and noticed that the number of the carbonyl compounds was responsible for off-taste. The dominant compounds, hydrocarbons, were not found in leaching water because of their low solubilities. Offflavor problems seemed to be connected with the extrusion coating and the volatile compounds were found to be the reaction products from the extrusion process.4,5 Odors are known4,6-7 to be formed during extrusion coating at high temperatures. Oxidation of low molecular weight hydrocarbons leads to a complex mixture of the compounds. High concentrations of alkoxy radicals in the gas phase during the extrusion process of PE and their possible reaction path have been reported by Hoff et al.6 The mechanism of the oxidation process in the extrusion coating was successfully presented by Warnar.8 The amount and nature of the volatile organic products evolved during the processing depend on temperature and the amount of oxygen access during heating.4 Bravo and Hotchkiss9 gave an excellent list of compounds formed during the thermal oxidation of PE. Despite the fact that oxidation is the most presumptive reason for off-flavor formation, oxidizing is necessary to achieve good adhesion properties.10-13 PE is normally nonpolar and (2) Koszinowski, J.; Piringer, O. J. Chromatogr. 1986, 2, 40-50. (3) Linssen, J. P. H.; Janssens, J. L. G. M.; Roozen, J. P. J. Plast. Film Sheet 1991, 7, 294-305. (4) Brodie, V., III. TAPPI Polymers, Laminations and Coating Conference, 1988; pp 511-515. (5) Hagman, A. Dynamic Headspace, A Versatile Method for Analyzing Volatile Compounds in Polymers. Academic Dissertation, Department of Analytical Chemistry, University of Stockholm, Stockholm, Sweden, 1988. (6) Hoff, A.; Jacobsson, S. Scand. J. Work. Environ. Health 1982, 2, 12-27. (7) Wyatt, D. M. TAPPI Polymers, Lamination & Coating Conference, 1985; pp 325-328. (8) Warnar, T. P. J. TAPPI European Polymer, Film and Coating Conference, Du ¨ sseldorf, Germany, 1997. (9) Bravo, A.; Hotchkiss, J. H. J. Appl. Polym. Sci. 1992, 47, 1741-1748. (10) Briggs, D.; Kendall, C. R. Polymers 1979, 20, 1053-1054. (11) Gregory, B. H.; McIntyre, W. D.; Michiels, D.TAPPI Paper Synthesis Conference, 1982; pp 167-172. (12) Potts, M. W.; Hansen, M. H.; Kuettel, B. T.; Goins, J. D. TAPPI Polymers, Laminations & Coatings Conference, 1993; pp 443-445.
Analytical Chemistry, Vol. 73, No. 5, March 1, 2001 971
oxidizing the PE surface polarizes it for good adhesion. The importance of that phenomenon is apparent in many applications that require good adhesion.14 Laiho and Yla¨nen13 presented the most common variables affecting adhesion. Many reports9,15-18 have shown the GC/MS technique to be suitable for identifying the unknown compounds in the fumes. Tikuisis et al.15 investigated the emission products generated by the processing of PE. Wyatt7,16 showed the effect of the extrusion temperature on volatiles formed by analyzing the volatile organic compounds (VOCs) of PE by GC/MS with headspace sampling. Le Lacheur et al.17 developed a method for identification of the unknown carbonyl compounds using derivatization followed by GC/MS. Rogge et al.18 studied the fumes formed in roofing asphalt and identified the volatile compounds using GC/MS. The concern with reference to airborne emission formed during the extrusion processing is essential. In atmospheric samples, the adsorption/thermal desorption/cold trap system is a suitable way to analyze the volatile organic compounds. The main categories of adsorbents used for VOCs are listed by Tirkkonen et al.19 The widely used adsorbents in thermal desorption are various classes (TA, GC, GR) of the Tenax matrix. The capacity of trapping volatiles of Tenax is high, the thermal stability is good, and the affinity for water is poor.19,20 There are many components with no significant odor in the raw materials used in the extrusion coating of LDPE. Nevertheless, during heating the carbonyl compounds are formed from hydrocarbons and they are supposed to be the most probable reasons for organoleptic problems. Some papers17,21,22 have explained the investigations of aroma adsorption and reported odor-active aliphatic compounds as the byproducts formed by combustion. The most common carbonyl compounds have their own characteristic odor. Their concentrations are very low in the samples studies but still exceeded their threshold odor concentration (TOC) and in some cases it is difficult to identify the carbonyl compounds among the hydrocarbons. Simultaneous sniffing during the MS run is a useful way to recognize the carboxylic acids with low molecular weight (C2C5) responsible for bad smells.23 Such identification was based on odors known to be specific and thus helped in recognition of the acids in question. The two-column system with simultaneous sniffing of the effluent of the GC/MS system was used to determine the components responsible for off-odors.1,24-28 The approach in identifying the odor-causative carboxylic acids under (13) Laiho, E.; Yla¨nen, T. TAPPI Polymers, Laminations & Coatings Conference Toronto, Canada, 1997; pp 93-109. (14) Brewis, D. M.; Briggs, D. Polymers 1981, 22, 7-16. (15) Tikuisis, T.; Phibbs, M. R.; Sonneberg, K. L. Am. Ind. Hyg. Assoc. J. 1995, 56, 809-814. (16) Wyatt, D. M. J. Plast. Film Sheet 1986, 2, 144-152. (17) Le Lacheur, R. M.; Sonnenberg, L. B.; Singer, P. C.; Christman, R. F.; Charles, M. J. Environ. Sci. Technol. 1993, 27, 2745-2753. (18) Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R. Environ. Sci. Technol. 1997, 31, 2726-2730. (19) Tirkkonen, T.; Mroueh, U.-M.; Orko, I. In NKB Committee and Work Reports 06 E, 1995; pp 7-8. (20) Bertsch, W.; Chang, R. C.; Zlatkis, A. J. Chromatogr. Sci. 1974, 12, 175182. (21) Stenstro ¨m, K.; Erlandsson, B.; Hellborg, R.; Wiebert, A. Nucl. Instrum. Methods Phys. Res. 1994, B 89, 256-258. (22) Anker, L. S.; Jurs, P. C.; Edwards, P. A. Anal. Chem. 1990, 62, 2676-2684. (23) Villberg, K. TAPPI European polymer, Film and Coating Conference, Du ¨ sseldorf, Germany, 1997.
972
Analytical Chemistry, Vol. 73, No. 5, March 1, 2001
the large hydrocarbon peaks based on a partial mass fragmentation using the common peak m/z 60 is known to be essential in linear carboxylic acids. The mass spectra of aliphatic acids having γ-hydrogen m/z 60 is usually very intense and it is a base peak.29,30 The objective of this study was to recognize the VOCs in the fumes formed during the extrusion coating of LDPE. The major part of these compounds was recognized as hydrocarbons which were not detected as off-flavors in our previous studies.1,23,28 To test volatile organic compounds and off-odor-causing compounds the fumes were collected into the adsorbent. EXPERIMENTAL SECTION Gas Chromatography. A gas chromatograph (HP 5890 series II) was connected to a mass spectrometer (VG AutoSpec). The GC was equipped with two identical capillary columns (HP-5, 0.25 mm × 30 m with a 1.0-µm phase; Hewlett-Packard); one column led to the mass spectrometer and the other one led to the sniffing funnel. Helium (99.995% purity) was used as the carrier gas at 1.5 mL/min. The GC oven temperature was held at 40 °C for 5 min and programmed at 5 °C/min to 150 °C and then 20 °C/min to 250 °C and held for 5 min. The temperature of injector A was installed at 200 °C and injector B was held at 180 °C. Mass Spectrometer. A VG AutoSpec (Manchester, U.K.) high-resolution mass spectrometer equipped with a VAX data system and Opus 3.3 software was used. The mass spectrometer was scanned from m/z 39 to 350 at a cycle of 1 s. The ion source temperature was 250 °C, and the electron ionization potential (EI) was 70 eV. Thermal Desorption. A special thermal desorption/cold trap device was used to introduce the samples into the GC/MS system.24 The diagram of the system used is shown in Figure 1. Sniffing Funnel. The sampling device was constructed with a funnel of which the eluted compounds from one column were sniffed. In the sniff port, the column is lead through the open wall inside a copper tube to the sniffing funnel. The device is shown in Figure 1. Preparation of Sampling Tubes. The sampling tubes (length 115.0 mm × 6.0 mm (o.d.) × 3.0 mm (i.d.)) was filled by Tenax GR (Chrompack, mesh 60/80), which is poly(p-2,6-diphenylphenylene oxide) with 23% graphitized carbon. The amount of the adsorbent in the glass tube was 225 mg. The filled tubes were thermally cleaned at 320 °C for 8 h with a purified nitrogen flow (20 mL/min) going through the tubes. Sampling Procedure. The volatile organic compounds were collected into the Tenax tubes from inside of the outgoing pipe and from the hot melt near the extruder at two temperatures: 285 and 315 °C. Known volumes (3.1 L from the outgoing pipe and (24) Veijanen, A. An Integrated Sensory and Analytical Method for Identification of Off-flavour Compounds. Academic Dissertation, University of Jyva¨skyla¨, Finland, 1990. (25) Veijanen, A.; Kolehmainen, E.; Kauppinen, R.; Lahtipera¨, M.; Paasivirta, J. Water Sci. Technol. 1992, 20, 2, 165-170. (26) Flφgstad, H. World Water 1984, (Dec), 27. (27) Sa¨venhed, R.; Boren, H.; Grimvall, A. J. Chromatogr. 1985, 328, 219-231. (28) Villberg, K.; Veijanen, A.; Gustafsson, I. Polym. Eng. Sci. 1998, 38, 8, 922925. (29) McLafferty, F. W.; Turezek, F. Interpretation of Mass Spectra, 4th ed.; University Science Books: Mill Valley. CA, 1993; Chapter 8. (30) Davis, R.; Frearson, M. Mass Spectrometry; John Wiley & Sons: London, 1987; Chapter 9.
Figure 1. Diagram of the thermal desorption system. Key: 1, heating regulation; 2, thermocouple measurement; 3, Swagelock fitting; 4, gas valve. Table 1. The Samples
to the areas of these fragment ions in the samples and the concentrations were calculated.
operating conditions sampling amount, L plastic
extrusion temp, °C
near ext
outgoing
1 2a 2b 3
285 285 315 315
2.1 2.6 1.6 2.1
1.5 1.5 1.5 1.5
1.6-2.6 L near the extruder) of air were drawn through the Tenax tubes using a calibrated air pump SKS model 222-3. The flow rate was 100 mL/min. Two sampling tubes were used in series. Quantitative Measurements. Quantification was performed by using an external standard. Two special standard solutions were made, and the contents of the standards are shown in Table 1. Aliquots of 1 µL of both solutions were injected into the Tenax GR filled tube and analyzed under the same conditions as the samples. After the MS run, the detected peak areas of all compounds were observed and compared with the suitable peaks of standard solution 2. For determining the amount of the carboxylic acids, the ion m/z 60 was selected (Figure 2) and the areas of this ion were calculated. The areas of the peak m/z 60 in standard solution 1 were compared to the areas of the peaks m/z 60 in the samples, and the concentrations were calculated. The areas of these fragment ions in standard solution 1 were compared
RESULTS AND DISCUSSION GC/MS/SNIFF System with a Thermal Desorption. The investigated samples were fumes collected from air near the extruder (hot melt) and from the outgoing pipe. There were three different plastic materials, and the extrusion temperatures used were 285 and 315 °C. The samples are encoded in Table 2. The total ion chromatograms of the fumes analyzed by GC/MS/SNIFF with thermal desorption are illustrated in Figures 3 with the odors detected. The intensity of odor was marked with “x’s” as follows: x ) weak odor, xx ) moderate odor, xxx ) strong odor, and xxxx ) very strong odor. Carbonyl Compounds in the Fumes Collected from an Outgoing Pipe. When the extrusion temperature was 285 °C, the number of odors in the fumes samples was less than in the fume samples collected at the extrusion temperature of 315 °C. Aldehyde contents were higher at extrusion temperature 315 °C than at 285 °C, and the total concentration of carbonyl compounds was the highest in the fumes of plastic B2 (4,2 mg/m3). The calculated amounts of all carbonyl compounds are shown in Table 3. In several cases, the same compound caused a different odor in different fume samples. The odor description was changed by the concentration. For example, 3-hexanone caused odor in all fume samples collected during extrusion of three different plastic Analytical Chemistry, Vol. 73, No. 5, March 1, 2001
973
Figure 2. Total ion chromatogram (TIC) with m/z 60 selected for the fume sample B2 collected from an outgoing pipe during the extrusion coating. Table 2. The Standard Solutions compound acetic acid propanoic acid butanoic acid pentanoic acid
dilution Standard: Solution 1 25 µL/25 mL (v/v) in methanol
Standard: Solution 2 acetophenone 50 µL/10 mL (v/v) in H2O benzaldehyde 2,3-butanedione butyl acetate decane 1-phenylcyclohexene hexanal 2-hexanone heptanal 3-hexen-1-one o-xylene cumene limonene methylcyclohexane 2,6-nonadienal 1-octen-3-ol R-pinene styrene 1,3,5-trimethylbenzene
materials (content in parentheses): plastic C, gluelike (0.07 mg/ m3); plastic B2, sweet, plasticlike (0.05 mg/m3); plastic B1 bad, acrid (0.004 mg/m3); plastic A, sweet (
