- Email: [email protected]
Behaviour of solid adsorbents for the sampling of atmospheric organochlorine compounds
Talanra,Vol. 40, No. 11,pp. 1769-1774, 1993 Rintai in Great Britain. All right.areserved
0039-9140/93 $6.00+ 0.00
copyright 0 1993PcrgamonPfcw Ltd
BEHAVIOUR OF SOLID ADSORBENTS FOR THE SAMPLING OF ATMOSPHERIC ORGANOCHLORINE COMPOUNDS C.NERIN,M. MARTINEZ, B. PONS and J. CACHO Dept.QuimicaAnalitica, Centro Politknico Superior de Ingenieros, Universidad de Zaragoza, Zaragoxa, Spain F-y-The behaviour of Tenax@ GC, Polyurethane foam (density 0.022 g/cm)), Amber&@ XAD-2 and Amberlite@ XAD-4 alone or in combination has been studied. Standard atmospheres containing different concentration levels of hexachlorocyclohexane and chlorobenxene isomers were generated and trapped in adsorbent cartridges. The generation of the atmosphere, the adsorption by the cartridges. the extraction of the compounds, the evaporation of the final solution and the analysis of GC/ECD have been studied. The most efficient system for trapping the test gases is the use of two cartridges conne-cted in series, one containing Polyurethane foam and the second one containing Tenax GC. Recovery values ranging from 72% for 1,3_dichlorobenzene to 98% for gamma-hexachlorocyclohexane are obtained, most of them >90%. The SD values for all the compounds under study are around 4% for a total sampled amount of 0.5 peg of each compound.
The environmental distribution of organochlorine compounds in water, sediments, fish, etc, has been studied.‘-‘3 There are many papers about their distribution in these different media. However, very few papers address the atmospheric distribution of these compounds even though they are used as pesticides and appear in the atmosphere around industrial waste dumps, e.g., from Lindane factories. These emissions contain hexachlorocyclohexane and chlorobenzene isomers. Nevertheless, the atmospheric distribution of these compounds and the typical concentration level is not well known. Due to their relatively higher vapour pressure, the presence of these species in the atmosphere may be important.lW3 The classical methods of sampling organochlorine compounds in the atmosphere consist of trapping them in an organic solvent, usually iso-octane. Due to the evaporation of the solvent during sampling and the consequent difficulty of long term sampling, this is not very practical. Also, the use of an organic solvent in the impinger demands attention. To avoid these disadvantages, relatively few studies have been conducted to determine trapping efficiencies of adsorbents to airborne pesticides.‘“16 Polyurethane foams, Porapak@ N and R, Tenax@ GC and Chromosorb@ 102 have been used for trapping several pesticides or related com-
pounds; only hexachlorocyclohexanes, chlordane and heptachlor have been studied.” Their ability to trap more volatile compounds such as chlorobenzenes has not previously been checked.16 This paper describes the results of studies carried out with Polyurethane foam, Tenax@ Amberlite@ XAD-4 and Amberlite@ XAD-2 in order to establish the optimum procedure for sampling organochlorine compounds, such as hexachlorocyclohexanes (HCHs) and chlorobenzenes, in the atmosphere.
Apparatus A Varian (Palo Alto, CA) Aerograph Gas Chromatograph and a silanized hollow glass column were used to generate the standard atmosphere. A Hewlett-Packard (Palo Alto, CA) Gas Chromatograph 5890 equipped with an Electronic Capture Detector (ECD) and automatic injector was used to analyse the samples. A capillary column DB-1701, 60 m x0.25 mm x 0.25 pm fihn thickness, was used under the following conditions: inj. temperature: 250”; detector temperature: 350”. Split ratio 1: 10. Initial temperature 50” for 5 min, 7”/min up to 250” and then held at 250” for 9.5
1769
1770
C. Nm
Column bead pressure: 95 kPa; carrier Hz; make up gas N2.
min.
et al.
amount of the internal analysed by GC/ECD.
Reagents Alpha-, beta-, gammaand delta1,2 - dichlorobenzene, hexachlorocyclohexane, 1,3 - dichlorobenzene, 1,4-dichlorobenzene, 1,2,3- trichlorobenzene, 1,2,4-trichloroben, 1,3,5-trichloroben1,2,5 - trichlorobenzene , zene , 1,2,3,~te~~hloro~~e and 1,2,4,5tetrachloro~~ene were obtained from Chem Service (West Chester, PA). Nitrobenzene standard was from Fluka (Buchs, Switzerland) and methyl chlorpyriphos was from Riedel-de Ha&i (Seelze, Germany). Polyurethane foam (PUF) of 0.022 g/cm’ density was specially produced by Pikolin (Zaragoza, Spain); Tenax GC, Amberlite XAD2 and Amberhte XAD-4 were supplied by Supelco (Bellefonte, PA) and Florisil was from Fluka. All glassware was cleaned using an ultrasonic bath with distilled water and neutral soap for 6 hr, after which it was washed sequentially with distilled water, acetone and hexane. PUF cartridges of 100 mm length and 20 mm diameter were cleaned before using a hexane diethylether mixture (19 : 1, v/v) for 12 hr in a Soxhlet extractor. After this treatment the cartridges were dried under a nitrogen current and stored in a clean glass container in the dark in a nitrogen atmosphere. Tenax was heated at 250” under Nz current for 2 hr, in order to clean the adsorbent before using it. Amberlite XAD-2 and XAD-4 were extracted with a hexanediethyl ether (19: 1, v/v) mixture before use. Figure 1 shows the chromatogram of a blank of each adsorbent used after the mentioned cleanup.
standard,
and then
RESULTS AND DWCUSSION
Generation of standard atmosphere Figure 2 depicts the scheme of the system employed. A GC with a hollow sifanized glass column was used to evaporate a solution of known concentration of the compounds to be (a)
” ISTD 14
~1
_ U
10
20
30
40
10
20
30
40
Procedure A hexane solution (0.5 ml) containing about 1 PgJg of the compounds mentioned was injected in an empty silanized glass column (y x 2 m, i.d.) placed in a GC at 60”, using synthetic air as carrier. The carrier flow is set at 0.8 l./min. In the GC outlet a glass cartridge containing a solid adsorbent is placed. After 1 hr the solid adsorbent sampling cartridge was introduced in a Soxhlet and a 12 hr extraction with the hexane-diethyl ether (19: 1, v/v) mixture carried out. The extract obtained was con~ntrat~ under Nr current up to 2 ml. This extract was transferred to a volumetric flask with a suitable
@)
A
Fig. 1. (a) Chromatogram of a blank of PUF. (b) Chromatogram of a blank of Tenax GC. (c) Cbromatogram of a blank of Amberlite XAD-2.
Solid adsorbents in the sampling of organochlorines
1771
Bxt?8ction
Carrier gas (rynthetic ait -
soxhlet extractor
0.8 l/mill)
adsorbent
Bv8poration
-
70’C 12 hours
I’
To - 2 mi N2 cuzrent 2s*c ISTD -
Hollow glass column
I h]SCtiOtl
Fig. 2. Scheme of the system used to generate the standard atmosphere and to study the behaviour of solid adsorbents.
studied. To stimulate the real sampling conditions, the injection port was heated at 90”, whereas the empty column was kept at 60”, under the boiling point of the solvent used in the injected solution. In the first part of the glass column a small amount of glass wool was placed to give more homogeneity to the evaporation process. In these conditions the compounds were evaporated slowly and because of the low oven temperature, the adsorption occurred essentially at room temperature. One or more glass cartridges containing the solid adsorbents 13
4
l-
10
were placed at the column outlet to trap the compounds. Fifty minutes after injecting the solution in the glass column, the temperature of the oven was increased to 100” and all the exhaust from the column was trapped in the same cartridge. Synthetic air (0.8 l./min) was used as carrier gas for the evaporation of the compounds. After each injection, 0.5 ml of hexane was injected into the system and the exhaust was trapped in another cartridge; this was used as a blank. No analyte compounds were observed in this blank. Figure 3 shows a 6 7
89
10
11 12 13 1415 16
i, i
20
30
40
Fig. 3. Chromatogram of a standard solution. 1: 1,3-dichlorobemme; 2: 1,4-dichlorobemene; 3 1,2_dichlorobenzene; 4: 1,3,5-trichlorobenzene; 5: nitrobenzene (1st ISTD); 6: 1,2&trichlorobenzen~ 7: 1,2,3-t~~~oro~~ne; 8: 1,2,4,5-tet~c~o~~ne; 9: 1,2,3,~~~oro~~; ltk pentachlorobenz.ene; 11: hexa~~oro~~~e; 12 alpha-HCB, 13: gamma-HCH, 11: methyl c~o~~ph~ (2nd ISTD); 1% beta-HCH, 16: delta-HCH.
C. NBRINet al.
1772
Table 1. Percentage recoveries of hexachlorocyclohexanes and chlorobenxuws from spiked polyurethane foam samples after extraction and evaporation steps Compound
Mean
88.04 1,3-Dichlorobenzene 68.06 1,4-Dichlorobenzene 77.45 1,2Dichlorobenzene 1,3,5-Trichlorobenzene 88.87 94.63 1,2,4_Trichlorobenzene 97.70 1,2,3-Trichlorobenzene 1,2,4,5-Trichlorobenzene 100.70 1,2,3,4_Trichlorobenzene 98.56 95.57 Pentachlorobenzene 94.17 Hexachlorobenzene Alpha-HCH 99.63 102.19 Gamma-HCH Beta-HCH 88.89 104.05 Delta-HCH
SD
n
5.07 5.02 6.82 5.26 6.93 5.93 6.75 4.80 4.17 7.74 8.19 3.19 13.10 7.12
5 5 6 6 5 5 5 5 5 4 4 5 4 4
temperature was selected as the most appropriate. The extracts obtained from the spiked samples were evaporated and analysed. The results are shown in Table 1. No losses are apparent from the extraction and evaporation steps. Behaviour of solid adrorbents
chromatogram of the compounds studied under the conditions mentioned above. Extraction study Once the compounds were trapped on the solid sorbents it was necessary to extract them. A study was carried out to optimize the extraction procedure. A 50 ml Soxhlet extractor was used with a hexanediethyl ether (19 : 1, v/v) mixture during 12 hr. Several spiked samples were prepared by adding 0.5 ml of a solution containing 2 pg/ml to a cartridge of PUF. After drying the PUF well, it was extracted in the Soxhlet extractor under the conditions mentioned above. According to previous studies, one of the major sources of error in the analysis of volatile compounds is due to the loss of compounds during the concentration/evaporation step. Among evaporation methods reported in the literature,18 the use of N, current at constant
Some organic compounds cannot be sampled on charcoal owing to decomposition reactions, irreversible adsorption, or due to the adsorption of other compounds which can interfere in the analysis. In such cases adsorbents of the porous polymer type may be useful alternatives. Previous worki on PUF showed that this can be a good adsorbent for HCHs, but its applicability to the sampling of chlorobenzenes was unknown. Therefore, the first adsorbent checked was PUF. Tenax, Amberlite XAD-4 and Amberlite XAD-2 are mentioned in the literature as very efficient adsorbents of volatile compounds.‘6*‘~22A second cartridge containing one of these adsorbents was connected serially after the cartridge containing PUF. Several standard atmospheres as calibrants were generated and these were sampled in the cartridges. In order to compare the trapping efficiency, when more than one cartridge was sequentially employed, the solid adsorbents from each cartridge were extracted together in the same Soxhlet. Table 2 shows the results obtained as an average of five independent determinations. These values include the recoveries obtained in the overall process, which includes the steps of generation, adsorption, extraction, concentration and analysis. Two different aspects can be emphasized. The first is the different trapping efficiency of PUF
Table 2. Trapping e5ciency of different solid adsorbents and combinations expressed as percentage recoveries PUF alone Compound 1,3-Dichlorobenzene 1,CDichlorobenzene 1,2-Dichlorobenzene 1,3,5-Trichlorobenxene 1,2,4-Trichlorob 1,2,3-Trichlorobcnxcne 1,2,4,5-Trichlorob 1,2,3,4-Trichlorobenne Pentachlorobenzene Hexachlorobenzene Alpha-HCH Gamma-HCH Beta-HCH Delta-HCH
PUF-TENAX
PUF-XAD2
PUF-XAD4
Mean
SD
n
Mean
SD
n
Mean
SD
R
Mean
SD
n
63.84 94.31 88.23 83.69 79.42 77.00 78.38 60.65 101.41 72.13 81.04 59.51 61.77 48.27
17.59 5.29 0.05 6.04 4.72 5.21 3.65 24.64 5.46 3.78 6.02 5.40 1.41 1.35
4 4 4 4 4 4 4 4 4 4 4 4 4 4
72.41 96.81 81.70 81.59 78.27 75.22 79.96 89.35 66.45 79.45 86.27 63.85 66.12 53.43
6.09 8.47 1.98 5.24 7.47 7.00 7.44 4.82 4.97 4.83 5.72 4.19 6.65 6.98
5 5 5 5 5 5 5 5 5 5 5 5 5 5
69.89 94.20 87.78 73.24 88.08 79.18 84.46 87.03 115.53 81.20 95.46 74.56 84.13 71.77
4.43 4.44 5.36 3.67 3.81 3.43 3.34 4.08 4.71 0.16 0.87 0.59 1.05 4.41
5 5 5 5 5 5 5 5 5 5 5 5 5 5
86.28 106.05 103.29 81.73 109.17 83.18 141.87 105.70 106.98 76.42 144.20 72.06 64.03 56.16
4.37 5.42 7.61 2.24 3.45 4.61 6.00 4.31 2.95 1.39 18.23 7.23 2.41 4.12
5 5 5 5 5 5 5 5 5 5 5 5 5 5
Solid adsorbents in the sampling of organochlorines
for chlorobenzene isomers compared to that for HCHs. The recovery of chlorobenzene is quite low. According to the literature,” the recoveries of volatile compounds from PUF are very low. As chlorobenzenes are quite volatile, especially dichlorobenzenes and trichlorobenzenes, PUF does not seem to be appropriate to trap them in the atmosphere. On the other hand, volumes of air around 1 m3 have to be sampled in order to measure the low concentration of such compounds commonly found in the atmosphere. Obviously, in these conditions the breakthrough volumes for these volatile compounds on PUF are surpassed, which means that this adsorbent is not efficient enough to sample these compounds. It can be further observed that the higher the boiling point of the compound, the higher is the recovery. The second aspect is that when two cartridges are combined in series, one of them serves mainly to trap the more volatile compounds and the other to capture the HCHs. No differences were found in the behaviour of Tenax and the Amberlite adsorbents, except that the standard deviation was slightly higher for the PUF + Amberlite combination. Taking into account these considerations, the PUF + Tenax combination in series was selected for trapping atmospheric HCHs and chlorobenzenes. Clew-up
of the PUF cartridges
The typical atmospheric concentration of the compounds of interest is usually very low. Consequently, it is necessary to improve the sensitivity of the procedure. This can be achieved through the concentration step. However, in this way, other compounds present in the PUF which can also be extracted are concentrated as well and some interferences appear. On the other hand, it was observed experimentally that when PUF was recently prepared no interferences appeared from other compounds present in the PUF, even after this concentration step. This fact can be seen in Fig. la, in which a chromatogram of a new PUF cleaned with hexanediethyl ether (19 : 1, v/v) as mentioned above is shown. Nevertheless, as soon as the PUF became older, which happens after approximately 2 months of being produced, new interferences from the PUF appeared in the concentration step, as shown in Fig. 3a. Figure 4a shows a chromatogram of an extract of an old PUF before the clean-up step. It is obvious that a clean-up step is necessary in such a case.
1773
04
61
1 I,.,.i....i....i.,..:....:.
0
5
10
15
20
25
Fig. 4. (a) Chromatogram of an old PUF before clean-up. (b) Chromatogram of a spiked old PUF after clean-up with 10% Florisil. The peak identifications are the same as in Fig. 3.
It is well known that both FlorisilQy and silica are good adsorbents for the clean-up of different matrices.2~*5 Although in this particular case the interferences to be removed are not major components, several experiments were carried out using 10% deactivated Florisil. The columns used were of 20 mm internal diameter, filled with about 7 g of Florisil and topped with 1 g of anhydrous sodium sulphate. The Soxhlet extract, about 70 ml, was added to the column and the compounds were eluted with 15 ml of hexane. The final solution obtained was evaporated to 2 ml and analysed. Figure 4b shows the chromatogram of a spiked PUF obtained after the clean-up step. The results obtained show that the use of Florisil in the conditions mentioned is adequate to remove the interferences from the old PUF. Furthermore, no losses of the OCPs occur in this clean-up step. This is detailed in the data presented in Table 3.
C. NERINet al.
1774
Table 3. Percentage recoveries of spiked polyurethane Florisil Compound 1,3Dichlorobenzene 1,CDichlorobenzene 1,2-Dichlorobenzene 1,3,5-Trichlorobenzene 1,2,4-Trichlorobenzene 1,2,3-Trichlorobenzene 1,2,4,5-Trichlorobenzene 1.2,3,4-Trichlorobe.nzene Pentachlorobenzene Hexachlorobenzene Alpha-HCH Gamma-HCH Beta-HCH Delta-HCH
foams after clean-up on
Amount added @g)
Percent recovered
CV
n
1.14 0.75 1.06 0.37 0.39 0.32 0.32 0.49 0.37 0.34 z
74.1 79.9 79.3 83.2 89.5 91.1 88.3 86.4 85.9 78.1 87.2 84.5 83.3 54.0
5.06 5.02 6.82 5.27 6.93 5.94 6.75 4.80 4.17 8.19 7.74 8.11 3.19 7.12
4 4 4 4 4 4 4 4 4 4 4 4 4 4
0:34 0.39
CONCLUSIONS The system used here to generate standard calibrant gases is efficient and lets the user obtain a great variety of contaminated atmospheres at different concentration levels. Among the solid adsorbents checked, the use of two cartridges connected in series, one containing polyurethane foam (PUF) and the other Tenax CC, is the most efficient trap for HCHs and chlorobenzenes in the atmosphere. This system has the advantage over other combinations that PUF is quite rigid and porous and the structure facilitates the handling of the cartridge. On the other hand, as two cartridges are independent, each cartridge can be recycled after use. Furthermore, in a real atmosphere, the PUF is not clogged by humidity from the air, and for this reason it is more convenient to place PUF in the first cartridge. As was shown in previous work,‘( humidity does not influence the efficiency of PUF to trap HCHs; this is a significant advantage over particulate polymeric adsorbents.
Acknowledgements-This paper was ilnancially supported by the Project “Bstudio de Accidentes Mayores en la Industria Quiica y sus consecuencias Medioambientales” de la Diputacibn General de Aragon. REFERENCES
1. S. Tanabe, R. Tatsukawa, M. Kawano and H. Hidaka, J. Ocean. Sot. Japan, 1982, 38, 137. 2. R. W&linger and K. Balls&miter, Chemosphere, 1987, 16, 2497. 3. K. Balls&miter and R. Wittlinger, Environ. Sci. Techno/., 1991, 25, 1103.
C. Nerin, M. Martinez and J. Cache, Toxicol. Environ. Ckem., 1992,35, 125. R. C. Fischer, W. Kramer and K. Balls&miter, Chemosphere, 1991, 23, 889. C. Ang, K. Meleady and L. Wallace, Bull. Environ. Con&m. Toxicol., 1989, 42, 595. D. M. Lockerbie and T. A. Clair, Bull. Environ. Contam. Toxicof., 1988, 41, 625. 8. C. Nerin, I. Echarri and J. Cache, 2. Anal. Chem., 1991, 339, 684. 9. K. W. McDougall, G. Singh, C. R. Harris and F. R. Higginson, Bull. Environ. Contam. Toxicol., 1987, 39, 286. 10. T. Samuel, H. C. Agarwal and M. K. K. Pillai, Pestic. Sci., 1988, 22, 1. 11. S. Panda, M. Sharmila, K. Ramanand, D. Panda and N. Sethunathan, Pestic. Sci., 1988, 23, 199. 12. L. Granier, M. Chevreuil, A. M. Cart-u and R. I_&tolle, Chemosphere, 1990, 21, 1101. 13. L. M. Hemandez, M. A. Femandez and M. J. Gonzalez, Bull. Environ. Contam. Toxicol., 1991, 46, 9. 14. R. G. Lewis and K. E. MacLeod, Anal. Chem., 1982.54, 310. 15. R. G. Lewis and M. D. Jackson, Anal. Chem., 1982,54, 592. 16. R. B. Leidy and C. G. Wright, J. Environ. Sci. Health, 1991, 4, 367. 17. C. Nerin, S. Ballestar, J. Cache and V. Ferreira, Man and his Ecosystem (Proc. 8th World Clean Air Congress), 1989, 3, 701. 18. V. Ferreira, Ph.D., University of Zaragoza, 1992. 19. H. Rothweiler, R. T. Wager and C. S&latter, Arm. Environ., 1991, 25, 231. 20. Y. Yokouchi and M. Sano, J. Chromarogr., 1991,556, 297. 21. J. 0. Levin and L. C. Carleborg, Ann. Occup., 1987, Hyg., 31, 31. 22. P. Ciccoli, E. Brancaleoni and A. Ceniuato, J. Ckromatogr., 1986, 35, 433. 23. I. Levi and T. W. Nowicki, J. AOAC, 1974, 57, 924. 24. M. J. Chessells, D. W. Hawker, D. W. Connell and I. A. Papajcsik, Chemosphere, 1988, 17, 1741. 25. T. S. S. Dikshith, S. N. Kumar and G. S. Tandon, Bull. Environ. Contam. Toxicol., 1989, 42, 50.
0039-9140/93 $6.00+ 0.00
copyright 0 1993PcrgamonPfcw Ltd
BEHAVIOUR OF SOLID ADSORBENTS FOR THE SAMPLING OF ATMOSPHERIC ORGANOCHLORINE COMPOUNDS C.NERIN,M. MARTINEZ, B. PONS and J. CACHO Dept.QuimicaAnalitica, Centro Politknico Superior de Ingenieros, Universidad de Zaragoza, Zaragoxa, Spain F-y-The behaviour of Tenax@ GC, Polyurethane foam (density 0.022 g/cm)), Amber&@ XAD-2 and Amberlite@ XAD-4 alone or in combination has been studied. Standard atmospheres containing different concentration levels of hexachlorocyclohexane and chlorobenxene isomers were generated and trapped in adsorbent cartridges. The generation of the atmosphere, the adsorption by the cartridges. the extraction of the compounds, the evaporation of the final solution and the analysis of GC/ECD have been studied. The most efficient system for trapping the test gases is the use of two cartridges conne-cted in series, one containing Polyurethane foam and the second one containing Tenax GC. Recovery values ranging from 72% for 1,3_dichlorobenzene to 98% for gamma-hexachlorocyclohexane are obtained, most of them >90%. The SD values for all the compounds under study are around 4% for a total sampled amount of 0.5 peg of each compound.
The environmental distribution of organochlorine compounds in water, sediments, fish, etc, has been studied.‘-‘3 There are many papers about their distribution in these different media. However, very few papers address the atmospheric distribution of these compounds even though they are used as pesticides and appear in the atmosphere around industrial waste dumps, e.g., from Lindane factories. These emissions contain hexachlorocyclohexane and chlorobenzene isomers. Nevertheless, the atmospheric distribution of these compounds and the typical concentration level is not well known. Due to their relatively higher vapour pressure, the presence of these species in the atmosphere may be important.lW3 The classical methods of sampling organochlorine compounds in the atmosphere consist of trapping them in an organic solvent, usually iso-octane. Due to the evaporation of the solvent during sampling and the consequent difficulty of long term sampling, this is not very practical. Also, the use of an organic solvent in the impinger demands attention. To avoid these disadvantages, relatively few studies have been conducted to determine trapping efficiencies of adsorbents to airborne pesticides.‘“16 Polyurethane foams, Porapak@ N and R, Tenax@ GC and Chromosorb@ 102 have been used for trapping several pesticides or related com-
pounds; only hexachlorocyclohexanes, chlordane and heptachlor have been studied.” Their ability to trap more volatile compounds such as chlorobenzenes has not previously been checked.16 This paper describes the results of studies carried out with Polyurethane foam, Tenax@ Amberlite@ XAD-4 and Amberlite@ XAD-2 in order to establish the optimum procedure for sampling organochlorine compounds, such as hexachlorocyclohexanes (HCHs) and chlorobenzenes, in the atmosphere.
Apparatus A Varian (Palo Alto, CA) Aerograph Gas Chromatograph and a silanized hollow glass column were used to generate the standard atmosphere. A Hewlett-Packard (Palo Alto, CA) Gas Chromatograph 5890 equipped with an Electronic Capture Detector (ECD) and automatic injector was used to analyse the samples. A capillary column DB-1701, 60 m x0.25 mm x 0.25 pm fihn thickness, was used under the following conditions: inj. temperature: 250”; detector temperature: 350”. Split ratio 1: 10. Initial temperature 50” for 5 min, 7”/min up to 250” and then held at 250” for 9.5
1769
1770
C. Nm
Column bead pressure: 95 kPa; carrier Hz; make up gas N2.
min.
et al.
amount of the internal analysed by GC/ECD.
Reagents Alpha-, beta-, gammaand delta1,2 - dichlorobenzene, hexachlorocyclohexane, 1,3 - dichlorobenzene, 1,4-dichlorobenzene, 1,2,3- trichlorobenzene, 1,2,4-trichloroben, 1,3,5-trichloroben1,2,5 - trichlorobenzene , zene , 1,2,3,~te~~hloro~~e and 1,2,4,5tetrachloro~~ene were obtained from Chem Service (West Chester, PA). Nitrobenzene standard was from Fluka (Buchs, Switzerland) and methyl chlorpyriphos was from Riedel-de Ha&i (Seelze, Germany). Polyurethane foam (PUF) of 0.022 g/cm’ density was specially produced by Pikolin (Zaragoza, Spain); Tenax GC, Amberlite XAD2 and Amberhte XAD-4 were supplied by Supelco (Bellefonte, PA) and Florisil was from Fluka. All glassware was cleaned using an ultrasonic bath with distilled water and neutral soap for 6 hr, after which it was washed sequentially with distilled water, acetone and hexane. PUF cartridges of 100 mm length and 20 mm diameter were cleaned before using a hexane diethylether mixture (19 : 1, v/v) for 12 hr in a Soxhlet extractor. After this treatment the cartridges were dried under a nitrogen current and stored in a clean glass container in the dark in a nitrogen atmosphere. Tenax was heated at 250” under Nz current for 2 hr, in order to clean the adsorbent before using it. Amberlite XAD-2 and XAD-4 were extracted with a hexanediethyl ether (19: 1, v/v) mixture before use. Figure 1 shows the chromatogram of a blank of each adsorbent used after the mentioned cleanup.
standard,
and then
RESULTS AND DWCUSSION
Generation of standard atmosphere Figure 2 depicts the scheme of the system employed. A GC with a hollow sifanized glass column was used to evaporate a solution of known concentration of the compounds to be (a)
” ISTD 14
~1
_ U
10
20
30
40
10
20
30
40
Procedure A hexane solution (0.5 ml) containing about 1 PgJg of the compounds mentioned was injected in an empty silanized glass column (y x 2 m, i.d.) placed in a GC at 60”, using synthetic air as carrier. The carrier flow is set at 0.8 l./min. In the GC outlet a glass cartridge containing a solid adsorbent is placed. After 1 hr the solid adsorbent sampling cartridge was introduced in a Soxhlet and a 12 hr extraction with the hexane-diethyl ether (19: 1, v/v) mixture carried out. The extract obtained was con~ntrat~ under Nr current up to 2 ml. This extract was transferred to a volumetric flask with a suitable
@)
A
Fig. 1. (a) Chromatogram of a blank of PUF. (b) Chromatogram of a blank of Tenax GC. (c) Cbromatogram of a blank of Amberlite XAD-2.
Solid adsorbents in the sampling of organochlorines
1771
Bxt?8ction
Carrier gas (rynthetic ait -
soxhlet extractor
0.8 l/mill)
adsorbent
Bv8poration
-
70’C 12 hours
I’
To - 2 mi N2 cuzrent 2s*c ISTD -
Hollow glass column
I h]SCtiOtl
Fig. 2. Scheme of the system used to generate the standard atmosphere and to study the behaviour of solid adsorbents.
studied. To stimulate the real sampling conditions, the injection port was heated at 90”, whereas the empty column was kept at 60”, under the boiling point of the solvent used in the injected solution. In the first part of the glass column a small amount of glass wool was placed to give more homogeneity to the evaporation process. In these conditions the compounds were evaporated slowly and because of the low oven temperature, the adsorption occurred essentially at room temperature. One or more glass cartridges containing the solid adsorbents 13
4
l-
10
were placed at the column outlet to trap the compounds. Fifty minutes after injecting the solution in the glass column, the temperature of the oven was increased to 100” and all the exhaust from the column was trapped in the same cartridge. Synthetic air (0.8 l./min) was used as carrier gas for the evaporation of the compounds. After each injection, 0.5 ml of hexane was injected into the system and the exhaust was trapped in another cartridge; this was used as a blank. No analyte compounds were observed in this blank. Figure 3 shows a 6 7
89
10
11 12 13 1415 16
i, i
20
30
40
Fig. 3. Chromatogram of a standard solution. 1: 1,3-dichlorobemme; 2: 1,4-dichlorobemene; 3 1,2_dichlorobenzene; 4: 1,3,5-trichlorobenzene; 5: nitrobenzene (1st ISTD); 6: 1,2&trichlorobenzen~ 7: 1,2,3-t~~~oro~~ne; 8: 1,2,4,5-tet~c~o~~ne; 9: 1,2,3,~~~oro~~; ltk pentachlorobenz.ene; 11: hexa~~oro~~~e; 12 alpha-HCB, 13: gamma-HCH, 11: methyl c~o~~ph~ (2nd ISTD); 1% beta-HCH, 16: delta-HCH.
C. NBRINet al.
1772
Table 1. Percentage recoveries of hexachlorocyclohexanes and chlorobenxuws from spiked polyurethane foam samples after extraction and evaporation steps Compound
Mean
88.04 1,3-Dichlorobenzene 68.06 1,4-Dichlorobenzene 77.45 1,2Dichlorobenzene 1,3,5-Trichlorobenzene 88.87 94.63 1,2,4_Trichlorobenzene 97.70 1,2,3-Trichlorobenzene 1,2,4,5-Trichlorobenzene 100.70 1,2,3,4_Trichlorobenzene 98.56 95.57 Pentachlorobenzene 94.17 Hexachlorobenzene Alpha-HCH 99.63 102.19 Gamma-HCH Beta-HCH 88.89 104.05 Delta-HCH
SD
n
5.07 5.02 6.82 5.26 6.93 5.93 6.75 4.80 4.17 7.74 8.19 3.19 13.10 7.12
5 5 6 6 5 5 5 5 5 4 4 5 4 4
temperature was selected as the most appropriate. The extracts obtained from the spiked samples were evaporated and analysed. The results are shown in Table 1. No losses are apparent from the extraction and evaporation steps. Behaviour of solid adrorbents
chromatogram of the compounds studied under the conditions mentioned above. Extraction study Once the compounds were trapped on the solid sorbents it was necessary to extract them. A study was carried out to optimize the extraction procedure. A 50 ml Soxhlet extractor was used with a hexanediethyl ether (19 : 1, v/v) mixture during 12 hr. Several spiked samples were prepared by adding 0.5 ml of a solution containing 2 pg/ml to a cartridge of PUF. After drying the PUF well, it was extracted in the Soxhlet extractor under the conditions mentioned above. According to previous studies, one of the major sources of error in the analysis of volatile compounds is due to the loss of compounds during the concentration/evaporation step. Among evaporation methods reported in the literature,18 the use of N, current at constant
Some organic compounds cannot be sampled on charcoal owing to decomposition reactions, irreversible adsorption, or due to the adsorption of other compounds which can interfere in the analysis. In such cases adsorbents of the porous polymer type may be useful alternatives. Previous worki on PUF showed that this can be a good adsorbent for HCHs, but its applicability to the sampling of chlorobenzenes was unknown. Therefore, the first adsorbent checked was PUF. Tenax, Amberlite XAD-4 and Amberlite XAD-2 are mentioned in the literature as very efficient adsorbents of volatile compounds.‘6*‘~22A second cartridge containing one of these adsorbents was connected serially after the cartridge containing PUF. Several standard atmospheres as calibrants were generated and these were sampled in the cartridges. In order to compare the trapping efficiency, when more than one cartridge was sequentially employed, the solid adsorbents from each cartridge were extracted together in the same Soxhlet. Table 2 shows the results obtained as an average of five independent determinations. These values include the recoveries obtained in the overall process, which includes the steps of generation, adsorption, extraction, concentration and analysis. Two different aspects can be emphasized. The first is the different trapping efficiency of PUF
Table 2. Trapping e5ciency of different solid adsorbents and combinations expressed as percentage recoveries PUF alone Compound 1,3-Dichlorobenzene 1,CDichlorobenzene 1,2-Dichlorobenzene 1,3,5-Trichlorobenxene 1,2,4-Trichlorob 1,2,3-Trichlorobcnxcne 1,2,4,5-Trichlorob 1,2,3,4-Trichlorobenne Pentachlorobenzene Hexachlorobenzene Alpha-HCH Gamma-HCH Beta-HCH Delta-HCH
PUF-TENAX
PUF-XAD2
PUF-XAD4
Mean
SD
n
Mean
SD
n
Mean
SD
R
Mean
SD
n
63.84 94.31 88.23 83.69 79.42 77.00 78.38 60.65 101.41 72.13 81.04 59.51 61.77 48.27
17.59 5.29 0.05 6.04 4.72 5.21 3.65 24.64 5.46 3.78 6.02 5.40 1.41 1.35
4 4 4 4 4 4 4 4 4 4 4 4 4 4
72.41 96.81 81.70 81.59 78.27 75.22 79.96 89.35 66.45 79.45 86.27 63.85 66.12 53.43
6.09 8.47 1.98 5.24 7.47 7.00 7.44 4.82 4.97 4.83 5.72 4.19 6.65 6.98
5 5 5 5 5 5 5 5 5 5 5 5 5 5
69.89 94.20 87.78 73.24 88.08 79.18 84.46 87.03 115.53 81.20 95.46 74.56 84.13 71.77
4.43 4.44 5.36 3.67 3.81 3.43 3.34 4.08 4.71 0.16 0.87 0.59 1.05 4.41
5 5 5 5 5 5 5 5 5 5 5 5 5 5
86.28 106.05 103.29 81.73 109.17 83.18 141.87 105.70 106.98 76.42 144.20 72.06 64.03 56.16
4.37 5.42 7.61 2.24 3.45 4.61 6.00 4.31 2.95 1.39 18.23 7.23 2.41 4.12
5 5 5 5 5 5 5 5 5 5 5 5 5 5
Solid adsorbents in the sampling of organochlorines
for chlorobenzene isomers compared to that for HCHs. The recovery of chlorobenzene is quite low. According to the literature,” the recoveries of volatile compounds from PUF are very low. As chlorobenzenes are quite volatile, especially dichlorobenzenes and trichlorobenzenes, PUF does not seem to be appropriate to trap them in the atmosphere. On the other hand, volumes of air around 1 m3 have to be sampled in order to measure the low concentration of such compounds commonly found in the atmosphere. Obviously, in these conditions the breakthrough volumes for these volatile compounds on PUF are surpassed, which means that this adsorbent is not efficient enough to sample these compounds. It can be further observed that the higher the boiling point of the compound, the higher is the recovery. The second aspect is that when two cartridges are combined in series, one of them serves mainly to trap the more volatile compounds and the other to capture the HCHs. No differences were found in the behaviour of Tenax and the Amberlite adsorbents, except that the standard deviation was slightly higher for the PUF + Amberlite combination. Taking into account these considerations, the PUF + Tenax combination in series was selected for trapping atmospheric HCHs and chlorobenzenes. Clew-up
of the PUF cartridges
The typical atmospheric concentration of the compounds of interest is usually very low. Consequently, it is necessary to improve the sensitivity of the procedure. This can be achieved through the concentration step. However, in this way, other compounds present in the PUF which can also be extracted are concentrated as well and some interferences appear. On the other hand, it was observed experimentally that when PUF was recently prepared no interferences appeared from other compounds present in the PUF, even after this concentration step. This fact can be seen in Fig. la, in which a chromatogram of a new PUF cleaned with hexanediethyl ether (19 : 1, v/v) as mentioned above is shown. Nevertheless, as soon as the PUF became older, which happens after approximately 2 months of being produced, new interferences from the PUF appeared in the concentration step, as shown in Fig. 3a. Figure 4a shows a chromatogram of an extract of an old PUF before the clean-up step. It is obvious that a clean-up step is necessary in such a case.
1773
04
61
1 I,.,.i....i....i.,..:....:.
0
5
10
15
20
25
Fig. 4. (a) Chromatogram of an old PUF before clean-up. (b) Chromatogram of a spiked old PUF after clean-up with 10% Florisil. The peak identifications are the same as in Fig. 3.
It is well known that both FlorisilQy and silica are good adsorbents for the clean-up of different matrices.2~*5 Although in this particular case the interferences to be removed are not major components, several experiments were carried out using 10% deactivated Florisil. The columns used were of 20 mm internal diameter, filled with about 7 g of Florisil and topped with 1 g of anhydrous sodium sulphate. The Soxhlet extract, about 70 ml, was added to the column and the compounds were eluted with 15 ml of hexane. The final solution obtained was evaporated to 2 ml and analysed. Figure 4b shows the chromatogram of a spiked PUF obtained after the clean-up step. The results obtained show that the use of Florisil in the conditions mentioned is adequate to remove the interferences from the old PUF. Furthermore, no losses of the OCPs occur in this clean-up step. This is detailed in the data presented in Table 3.
C. NERINet al.
1774
Table 3. Percentage recoveries of spiked polyurethane Florisil Compound 1,3Dichlorobenzene 1,CDichlorobenzene 1,2-Dichlorobenzene 1,3,5-Trichlorobenzene 1,2,4-Trichlorobenzene 1,2,3-Trichlorobenzene 1,2,4,5-Trichlorobenzene 1.2,3,4-Trichlorobe.nzene Pentachlorobenzene Hexachlorobenzene Alpha-HCH Gamma-HCH Beta-HCH Delta-HCH
foams after clean-up on
Amount added @g)
Percent recovered
CV
n
1.14 0.75 1.06 0.37 0.39 0.32 0.32 0.49 0.37 0.34 z
74.1 79.9 79.3 83.2 89.5 91.1 88.3 86.4 85.9 78.1 87.2 84.5 83.3 54.0
5.06 5.02 6.82 5.27 6.93 5.94 6.75 4.80 4.17 8.19 7.74 8.11 3.19 7.12
4 4 4 4 4 4 4 4 4 4 4 4 4 4
0:34 0.39
CONCLUSIONS The system used here to generate standard calibrant gases is efficient and lets the user obtain a great variety of contaminated atmospheres at different concentration levels. Among the solid adsorbents checked, the use of two cartridges connected in series, one containing polyurethane foam (PUF) and the other Tenax CC, is the most efficient trap for HCHs and chlorobenzenes in the atmosphere. This system has the advantage over other combinations that PUF is quite rigid and porous and the structure facilitates the handling of the cartridge. On the other hand, as two cartridges are independent, each cartridge can be recycled after use. Furthermore, in a real atmosphere, the PUF is not clogged by humidity from the air, and for this reason it is more convenient to place PUF in the first cartridge. As was shown in previous work,‘( humidity does not influence the efficiency of PUF to trap HCHs; this is a significant advantage over particulate polymeric adsorbents.
Acknowledgements-This paper was ilnancially supported by the Project “Bstudio de Accidentes Mayores en la Industria Quiica y sus consecuencias Medioambientales” de la Diputacibn General de Aragon. REFERENCES
1. S. Tanabe, R. Tatsukawa, M. Kawano and H. Hidaka, J. Ocean. Sot. Japan, 1982, 38, 137. 2. R. W&linger and K. Balls&miter, Chemosphere, 1987, 16, 2497. 3. K. Balls&miter and R. Wittlinger, Environ. Sci. Techno/., 1991, 25, 1103.
C. Nerin, M. Martinez and J. Cache, Toxicol. Environ. Ckem., 1992,35, 125. R. C. Fischer, W. Kramer and K. Balls&miter, Chemosphere, 1991, 23, 889. C. Ang, K. Meleady and L. Wallace, Bull. Environ. Con&m. Toxicol., 1989, 42, 595. D. M. Lockerbie and T. A. Clair, Bull. Environ. Contam. Toxicof., 1988, 41, 625. 8. C. Nerin, I. Echarri and J. Cache, 2. Anal. Chem., 1991, 339, 684. 9. K. W. McDougall, G. Singh, C. R. Harris and F. R. Higginson, Bull. Environ. Contam. Toxicol., 1987, 39, 286. 10. T. Samuel, H. C. Agarwal and M. K. K. Pillai, Pestic. Sci., 1988, 22, 1. 11. S. Panda, M. Sharmila, K. Ramanand, D. Panda and N. Sethunathan, Pestic. Sci., 1988, 23, 199. 12. L. Granier, M. Chevreuil, A. M. Cart-u and R. I_&tolle, Chemosphere, 1990, 21, 1101. 13. L. M. Hemandez, M. A. Femandez and M. J. Gonzalez, Bull. Environ. Contam. Toxicol., 1991, 46, 9. 14. R. G. Lewis and K. E. MacLeod, Anal. Chem., 1982.54, 310. 15. R. G. Lewis and M. D. Jackson, Anal. Chem., 1982,54, 592. 16. R. B. Leidy and C. G. Wright, J. Environ. Sci. Health, 1991, 4, 367. 17. C. Nerin, S. Ballestar, J. Cache and V. Ferreira, Man and his Ecosystem (Proc. 8th World Clean Air Congress), 1989, 3, 701. 18. V. Ferreira, Ph.D., University of Zaragoza, 1992. 19. H. Rothweiler, R. T. Wager and C. S&latter, Arm. Environ., 1991, 25, 231. 20. Y. Yokouchi and M. Sano, J. Chromarogr., 1991,556, 297. 21. J. 0. Levin and L. C. Carleborg, Ann. Occup., 1987, Hyg., 31, 31. 22. P. Ciccoli, E. Brancaleoni and A. Ceniuato, J. Ckromatogr., 1986, 35, 433. 23. I. Levi and T. W. Nowicki, J. AOAC, 1974, 57, 924. 24. M. J. Chessells, D. W. Hawker, D. W. Connell and I. A. Papajcsik, Chemosphere, 1988, 17, 1741. 25. T. S. S. Dikshith, S. N. Kumar and G. S. Tandon, Bull. Environ. Contam. Toxicol., 1989, 42, 50.