Note 20: Using Direct Thermal Desorption to Assess the Potential Pool of Styrene and 4-Phenylcyclohexene In Latex-Backed Carpets

National Center For Toxioogical Research, FDA, Jefferson, AR 72079

INTRODUCTION

Indoor air pollution, including the emission of volatile organic compounds (VOC's) from carpets, has been associated with Sick building syndrome and Building-related illness. The styrene-butadiene-rubber (SBR) latex glue used in the manufacturing of carpets gives rise to styrene and 4-phenylcyclohexene (4-PC), two VOC's associated with these phenomena. Styrene is a residue from the polymerization reaction, and 4-PC results from a thermally controlled (and thermally allowed) [4+2] cycloaddition reaction between styrene and butadiene during the polymerization reaction (3). Several researchers have identified 4-PC in synthetic carpets using headspace analysis (4) and gas phase stripping (5).

Short path direct thermal desorption (DTD) is a technique used to desorb VOC's directly from solid samples.(6) The advantages of short path DTD are (a) no sample preparation is required, (b) Memory effects and analyte degradation are minimized due to short transfer lines, and (c) increased sample delivery leads to modest sample sizes and improved sensitivity. VOC's in plastic wrap, a high molecular weight polymer, have been identified via DTD (7). The classes of compounds found included aliphatic and olefinic hydrocarbons, alcohols, aromatics and phthalates. Using DTD, we describe the semi-quantification of styrene and 4-PC in another polymer, SBR latex-backed carpets.

Experimental

Styrene (Aldrich Chemical Co., Milwaukee, WI), styrene-d8 (Aldrich Chemical Co.) and absolute ethanol (Midwest Grain Products, Perkin, IL) were purchased from sources indicated. A sample of 4-PC was provided by Dr. John W. Waechter, Jr., Dow Chemical Co., Midland, MI.

Carpet samples were obtained directly from the manufacturers by representatives of the U.S. Consumer Product Safety Commission. The samples (nine samples of the same carpet type per bag, each approximately 30 cm X 30 cm) were heat-sealed in Tedlar bags and shipped to the National Center for Toxicological Research for analysis. The samples remained in the sealed bags until the analyses were conducted.

The linear regression analyses were performed on a personal computer using the statistics package SPSS for Windows, Release 5.0 by SPSS Inc. (Chicago, IL).

Direct Thermal Desorption

Thermal Desorption Tubes:

Glass wool was silylated with dimethyldichlorosilane. The glass wool (10g) was immersed in dimethyldichloro-silane/methylene chloride (10 mL/100 mL), and sonicated for 10 min. It was then rinsed two times for 30 min with methanol/methylene chloride (1:1, 100 mL) using the ultrasonic bath, and dried in an oven at 100 degrees C.

Stainless steel glass-lined thermal desorption tubes were packed with 50 mg of Tenax® TA (35/60 mesh, Alltech Associates, Deerfield, IL) or a mixed bed of 50 mg each Tenax TA and Carbotrap (20/40 mesh, Supelco, Bellefonte, PA). The tubes were prepared by inserting, in order, a plug of silylated glass wool, Tenax TA or Tenax TA and Carbotrap, and another plug of silylated glass wool. The packed tubes were conditioned in a Scientific Instrument Services Short Path Thermal Desorption Conditioning System (Ringoes, NJ) at 320 degrees C for four hours with a flow of helium at 40 mL/min. The tubes were removed from the heater block, cooled for 10 min under a stream of helium, and capped as soon as possible. The conditioned tubes were stored in capped test tubes until needed.

The standard response curves were generated using the mixed bed desorption tubes. The DTD of the carpet samples was carried out in the desorption tubes packed with Tenax TA.

Thermal Desorption

Samples were thermally desorbed using a Scientific Instrument Services Short Path Thermal Desorption System, model TD-1. The desorption time was 10 min at 250 degrees C, with a helium flow of 20 mL/min.

Samples were desorbed directly into the GC injection port of a Hewlett Packard GC-MS system equipped with a 30 m DB-1 column, with the head being held at -40 degrees C. The sample was cryofocused at the head of the GC column, the column temperature was raised to 20 degrees C at 30 degrees C/min, held for three min, then increased to 230 degrees C at 10 degrees C/min, where it was held for an additional 10 minutes. Limited scan data from m/z 90 to 190 were collected 17.5 min after the start of the chromatographic run.

Standard Solutions

A stock solution of styrene and 4-PC was prepared by dissolving 1 µL (910 µg) of styrene and 1 µL (1020 µg) of 4-PC in 10.0 mL of absolute ethanol. The resulting solution concentration was 91 µg/mL and 102 µg/mL of styrene and 4-PC, respectively. The stock solution was diluted serially with ethanol to yield a solution of 18.2 µg/mL and 20.4 µg/mL of styrene and 4-PC, respectively.

A stock solution of styrene-d8 was prepared by dissolving 7.1 mg of styrene-d8 in 10.0 mL of absolute ethanol, resulting in a concentration of 710 µg/mL. The stock solution was further diluted with ethanol to yield a solution of concentration 14.2 µg/mL.

Direct Solution Loading

The desorption tubes were fitted with caps containing a 1.0 mm diameter hole and high temperature green septa. A 10 µL syringe was used to load 1-5 µL of standard solution through the septum and into the glass wool plug in the tube. The ethanol was purged with helium at a flow of 60 mL/min for five min through the septum. The standards were loaded from the Tenax TA side of the mixed beds. The internal standard, styrene-d8, was added prior to either the styrene or 4-PC.

Direct Thermal Desorption Of Carpet Samples

Samples (approximately 1 cm X 3 cm) of carpet were cut from the larger piece of carpet. The carpet pile was removed, leaving the carpet backing. A subsample of the carpet backing (8 -112 mg) was placed in a Tenax TA thermal desorption tube and analyzed as previously described.

Results And Discussion

Carpet Sample Description

The industrial designations of the collected commercial carpet samples were nylon with SBR latex backing (CNL1, CNL2 and CNL3).

The industrial designations of the collected residential carpet samples were as follows: nylon with latex backing (RNL4, RNL5 and RNL6), wool with latex backing (RWL7 and RWL8), polyester with latex backing (REL9 and REL10) and polypropylene with latex backing (RPL11).

Standard Response Curve Generation

Typically, standard response curves have been generated via gas phase loading of desorption tubes (8) or via direct injection (9). We developed a direct loading method, using solutions of analyte in ethanol. The samples were loaded in Tenax TA/Carbotrap thermal desorption tubes and analyzed via GC-MS as delineated in the experimental section.

The suitability of styrene-d8 as an internal standard was verified by generating a standard response curve of detector response (determined via the area of 112 m/z) versus weight of styrene-d8 The equation of the resulting line was y = 0.926 x - 3.69, with a correlation coefficient of 0.985. The m/z 112 response versus the amount of styrene-d8 was demonstrated to be linear between 21.3 and 106.5 ng. The amount of styrene-d8 added to each sample was within the linear region of the line.

The styrene and 4-PC standard response curves were generated by loading styrene-d8 and a styrene/4-PC solution in a Tenax TA/Carbotrap desorption tube. The samples were analyzed by GC-MS, and the ratio of the area of styrene m/z 104 to the styrene-d8 m/z 112 versus the amount of styrene was analyzed by linear regression. The equation of the resulting line was y = 0.1965 x + 0.0908, with a correlation coefficient of 0.997 (Figure 1). In analogous fashion, the 4-PC standard response curve was generated. The equation of the line was y = 0.0542 x + 1.35, and the correlation coefficient was 0.993 (Figure 1).

Direct solution loading versus gas phase loading (8) of standards eliminated the surface sink of the gas-expansion chamber, the syringe and the packed column injection port (used to introduce the gaseous VOC's onto the Tenax TA/Carbotrap desorption tubes), as well as any Memory Effects associated with gas phase loading, and the limitations caused by the volatilization of the semi-volatile 4-PC.

Figure 1 - Standard Response Curve: Ratio of Styrene To Styrene-d8 vs ng of Styrene and Ratio of 4-PC To Styrene-d8 vs ng of 4-PC.

Direct Thermal Desorption Analysis

Styrene and 4-PC released from a carpet sample were quantified by placing a known amount of carpet backing in a Tenax TA desorption tube that had previously been loaded with styrene-d8. The samples were analyzed by GC-MS.

Initially, DTD of approximately 50 mg of carpet backing with styrene-d8 was used to determine the sample size necessary for the analytes to be within the range of the standard response curve. This allowed the observation of styrene at 22.1 min and 4-PC at 30 min. The weight of the carpet backing was adjusted accordingly, and the amount of styrene and 4-PC was determined for the latex backed carpets. The results are summarized in Table I.

Table I: Amount of Styrene and 4-PC in Latex Backing Carpet                                   Stryene/Carpet        4-PC/Carpet

Carpet                   Weight         (ng/mg)             (ng/mg)

Sample                    mg)           m/z 104          m/z 104, 158

----------         --------------     ------------      --------------

CNL1                      40              2.41               **

                           7                *               45.0

CNL2                      39              1.30               **

                          11                *               37.6

CNL3                     111              2.03               **

                           8                *               17.6

RNL4                      10              2.39              23.2

RNL5                      34             0.536              11.7

RNL6                      25              3.46              16.8

RWL7                     105             0.231               **

                          15                *               21.0

RWL8                      41                *               11.8

REL9                      12              5.49              3.49

REL10                     10              9.35              10.2

RPL11                     18              9.34              8.44

   * Not detected, measurement below detection limit (10ng)

   ** Measurements above the highest calibration point (500ng)

Styrene and 4-PC were present in both residential and commercial carpets having latex glue backing: styrene, because it was present in excess from the polymerization of styrene and butadiene to form SBR latex, and 4-PC due to a side reaction in the preparation of SBR latex.

Conclusion

These studies should not preclude the possibility of other VOC's in carpets. They should, however, illustrate the use of solution loading for standard response curve generation, and the subsequent ease of quantifying VOC's in solid samples via DTD.

DTD provided a measure of the total emission pool that may be expected for 4-PC and styrene in latex-back carpet.

Acknowledgments

Partial support for this work was provided by the Consumer Product Safety Commission under Interagency Agreement number CPSC-IAG-90-1212.

One of us (N.E.T.) was supported in part by an appointment to the ORISE Postgraduate Research Program at the National Center for Toxicological Research, which is Administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U.S. Department of Energy and the U.S. Food and Drug Administration.

References

1. a. Reviewed in Sage, G.; Ward, J.; Neal, M. W.; Howard, P. H. Ò4-Phenylcyclohexene ScreeningÓ, Syracuse Research Corp.: New York, 1982.

b. Steyer, R. Across the Board, 1986, 34.

c.Skov, R., Valbjorn, O. DISG Environment International, 1987, 13, 339.

2. a.Walsh, D. W., Masters Thesis, University of Arizona, 1986.

b.Vulkelich, D. ÒThomas Admits EPA Needs Some Fresh AirÓ in The Washington Times, Washington D. C., June 17, 1988.

3. Woodward, R. B., Hoffmann, R. Angew. Chem. Inter. Ed. 1969, 81, 797.

4. a.Miksch, R. R.; Hollowell, C. D.; Schmidt, H. E. Environment International 1982, 8, 129.

b.Crabbe, C. L. Masters Thesis, University of Arizona, 1984.

c.Roderick, G. S., Anderson, C. P. ÒVolatile Organic Compounds from some Carpet MaterialsÓ presented at The Pittsburg Conference, Chicago, IL, 1991.

d.Miller, D. W., Heinze, T., Beland, F. A. ÒQualitative Evaluation of Volatile Compounds from Commercial and Residential CarpetsÓ to Consumer Product Safety Commission under Interagency Agreement number CPSC-IAG-90-1212, Washington D.C., 1992.

5. Schroder, E. Textilveredlung, 1989, 21, 254.

6. a.Manura, J. J., Overton, S. V., Baker, C.W., Manos, J. N. The Mass Spec Source, 1990, XIII, 22.

b.Manura, J. J. The Mass Spec Source, 1991, XIV, 22.

7. a.Hartman, T. G., Lech, J., Rosen, R.T. The Mass Spec Source, 1990, XIII, 30.

b.Manura, J. J., Overton, S. V., Baker, C. W. The Mass Spec Source, 1990, XIV, 14.

8. a.Berkley, R. E. ÒStandard Operating Procedure for the Preparation and Use of Standard Organic Gas Mixtures in a Static Dilution Bottle,Ó U.S. Environmental Protection Agency, Method EMSL/RTP-SOP-EMD-001, July 1982.

b.Hodgson, A. T., Girman, J. R. ÒApplication of a Multisorbent Sampling Technique for Investigations of Volatile Organic Compounds in BuildingsÓ in Design and Protocol for Monitoring Indoor Air Quality, ASTM STP 1002, Nagda, N. L., Harper, J. P., Eds., Philadelphia: 1989, 244-256.

9. Tichenor, G. A. Indoor Air Ô87, 1989, 1,8.

aArticle contributed by Michelle E. Lee, Natalie E. Takenaka, Frederick A. Beland, Dwight W. Miller

National Center for Toxicological Research

Food and Drug Administration

Jefferson, AR 72079