ANALYSIS OF PCDDs AND PCDFs IN CONTAMINATED SOIL AND SEDIMENT USING SUPERCRITICAL FLUID EXTRACTION

Project leader: Terttu Vartiainen1,2, 1National Public Health Institute, Division of Environmental Health, P.O. Box 95, FIN-70701 Kuopio, Finland, 2 Department of Environmental Sciences, University of Kuopio, P.O. Box  1627 FIN-70211 Kuopio, Finland, tel. +358-17-201346, e-mail: Terttu.Vartiainen@ktl.fi
 
 

PUBLICATIONS

TIIVISTELMÄ SUOMEKSI

Researchers:
Mari Mannila, National Public Health Institute, tel. +358-17-201161, e-mail: Mari.Mannila@ktl.fi
Jaana Koistinen, National Public Health Institute,  tel. +358-17-201350, e-mail: Jaana.Koistinen@ktl.fi

Consortium: Environmental health risk of dioxins
Financing SYTTY organisation: The Ministry of Environment
Funding from SYTTY / Total funding of project (€): 45410 /45410
Person-months of work funded by SYTTY / Total person-months of work: 20 /25

KEY WORDS: SFE, PCDD/PCDF, soil, sediment
 

EXTENDED ABSTRACT

1 Introduction

Polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/PCDFs) are persistent and toxic compounds, which are formed as by-products in several different industrial and combustion processes. Thus they are distributed all over the world and some certain areas are highly contaminated with PCDD/PCDFs. For example, in Finland over 250 sawmills are contaminated with PCDD/PCDFs due to the previous use of a wood preservative, Ky-5, which is a chlorophenol formulation containing PCDD/PCDFs as impurities [3].

Traditional analysis of PCDD/PCDFs is very expensive and time-consuming process due to many isolation and clean-up steps, which consume lots of hazardous solvents. To overcome these disadvantages, more economical and faster extraction methods, such as supercritical fluid extraction (SFE), have been developed. SFE has successfully been used in analyses of environmental pollutants such as polychlorinated biphenyl (PCB), polysyclic aromatic hydrocarbon (PAHs) [4,5,6], but detailed studies of SFE in analysis of PCDD/PCDFs in soil and sediment have not been performed, so far. However, when remediation of sites contaminated with PCDD/PCDFs demand large scale analyses, an economic and fast method for analysis of PCDD/PCDFs is needed.

Supercritical fluid extraction (SFE) was optimised to replace the traditional liquid based methods (Soxhlet and ultrasonic) in routine analysis of polychlorinated dibenzo-p-dioxins (PCDDs) and dibenzofurans (PCDFs) in soil and sediment. For collection system in SFE was developed an activated carbon trap that could permit minimal solvent consumption and ensure that the samples were clean enough for direct detection after concentration. The SFE conditions were optimised by modifying dynamic time, temperature, pressure, and carbon dioxide flow rate. To verify the efficiency of SFE method the results were compared to those obtained in international intercalibration studies [1,2] and to results obtained by Soxhlet and Ultra-sonic performed in our laboratory.

2 Experimental section

All extractions were performed by Suprex Autoprep 44TM instrument with SFE grade carbon dioxide (5.2 AGA Gas or 5.5 Messer) as an extraction fluid. Following conditions were used in all extractions: static extraction time 10 min, trap temperature 40 oC and restrictor temperature 45 oC. An extraction cell (10 ml steal vessel) was packed with a layer of activated Na2SO4 (5 g; Merck p.a.), 1 g soil/sediment (or native standard solution in tests of trap), basic Al2O3 (2 g; Merck 90 standardised) and finally a layer of activated Na2SO4 (2 g).

Standards. The standard solution 12C-PCDD/PCDFs, that was used in trap tests, contained 17 toxic PCDD/PCDF congeners (4 ng/congener/extraction) in toluene. The solution of 12C-PCDEs contained alltogether 45 congeners (tetra-deca chlorinated, 2-4 ng/congener/extraction, University of Jyväskylä, Finland), and that of 12C-PCBs 28 congeners (tri-deca chlorinated, 2-4 ng/congener/extraction, AccuStandard).

A mixture of internal standard solution used for the quantitation of PCDD/PCDFs in soil and sediment contained altogether sixteen 13C-labelled PCDD/PCDFs (ED-998 tetra-octa chlorodioxin standard solution and EF-999 tetra-octa chlorofuran standard solution, Cambro Scientific, Netherlands). 100 ul of this standard (115 pg/congener/extraction) was added before extraction to the samples that were concentrated as a whole for the further analysis. A recovery standard, a mixture of 13C-labelled 1234-TCDD and 123789-HxCDD (40 pg/congener, Cambro Scientific, Netherlands), was added to each sample before the analysis.

Development of the trap [13]. The solid phase adsorption material optimised in the trap was a mixture of activated carbon (Carbopack C, 60/80 mesh, Supelco, Bellefonte, USA) and Celite 545 (0,01-0,04 mm, E. Merck, Darmstad, Germany). Optimisation of the adsorption material in the trap was performed by extracting native standard solutions of PCDD/PCDFs, PCBs and PCDEs as samples (composition of standard solutions described earlier). Three different configurations of carbon content were tested by changing the ratio of carbon/Celite: 1:25 (w/w), 1:10 (w/w) and 1:5 (w/w) in traps A, B and C, respectively. Hexane was tested for flushing of co-extracted impurities from the trap, and toluene and xylene for collecting PCDD/PCDFs. The solid phase trap was filled with 0.38 g of adsorbent material. The adsorption capacity of trap C was investigated also by extracting PCDD/PCDF standard solution with CO2 modified with methanol (5%). The modifier was added to the fluid by a separate pump. The possibility to use the same adsorption material for several sample extractions was also studied, because with SFE all of the samples go through the same adsorbent and this can result in the contamination of the instrument, the trap and the next sample.

Optimisation of SFE conditions [14,15]. Three sediment samples and seven soil samples were used in optimisation of SFE conditions, which included tuning of temperature (50-150°C), pressure (300-400 atm) and fluid flow rate (1-3 ml/min) [14]. Sediments were prepared by International Sediment Exchange for Tests on Organic Contaminants (SETOC) and soils originated from Finnish sawmills that were contaminated with a chlorophenolformulation, KY-5. Dynamic extraction time was optimised with three soil samples (with 3 ml/min fluid flow) by four sequential 20 min extractions (total of 80 min). Also the fluid flow rates of 1 ml/min and 3 ml/min were studied with three soil samples using 60 min dynamic extraction [15].

Sediments were recently analysed in an international intercalibration study [1], so the average values of the best performing laboratories in that study and the Soxhlet results obtained in our laboratory were used as the reference values for SFE. All soil samples were extracted with multiple extractions (2-5) at the best obtained SFE conditions and the results were compared to those obtained with Soxhlet. Ultrasonic extraction was also performed with three soil samples. Furthermore, the efficiency of the developed SFE method was confirmed by extracting five additional sediment samples [2] as triplicates with SFE .

Soxhlet and ultrasonic extractions. Soxhlet and ultrasonic methods used in this study have been accredited in our laboratory by Finnish certification system (SFS-EN 45001, ISO/IEC Guide 25). Soxhlet extraction was carried out with 300 ml of toluene for at least 18 hours. Ultrasonic extraction was performed by extracting each sample three times with 10 ml toluene for 30 min at a room temperature. Following ultrasonic extractions, the solvent was decanted and the extracts were combined. Both Soxhlet and ultrasonic extracts were cleaned-up by column chromatography using silica gel, basic alumina and activated carbon columns before analyses. The clean-up procedure has been presented in details previously [7].

Analysis. The samples were analysed with a high-resolution gas chromatograph (Hewlet Packard 5890, column DB-Dioxin: 60 m, 0.25 mm, 0.15 µm) which was coupled to a high-resolution mass spectrometer (magnetic sector instrument VG-70-250SE, Manchester, England). The samples were splitlessly injected at 270 oC, and helium (purity 4.6, AGA Gas, Hamburg, Germany) at a flow rate of 1 ml/min was used as a carrier gas. The oven temperature was held at 140 oC for 4 min, then increased at a rate of 20 oC/min to 180 oC, and finally increased at a rate of 2 oC/min to 270 oC, where it was retained for 41 min. Analyses were performed in selected ion monitoring (SIM) mode using 10,000 resolution. The concentrations of each toxic PCDD/PCDFs congeners were analysed, and the toxic load (I-TEQ) was calculated using international TCDD equivalent factors (I-TEFs) [8].

3 Results and discussion

Solid phase trap. In the optimization of the activated charbon content in the trap, the best results were obtained with trap C that contained the highest amount of carbon. In traps A and B the carbon content was too low to keep the lower chlorinated PCDD/PCDFs absorbed during elution with hexane. With trap C the co-extractive compounds were efficiently collected with hexane (4 ml) after which PCDD/PCDFs were quantitatively collected with toluene (10 ml). To clean and recondition the system for the next sample, the trap and lines were flushed with additional fractions of xylene (5 ml) and hexane (5 ml). The adsorption capacity of the trap was maintained good over a long time period and did not alter in use: extraction of about 100 samples could be performed with the same adsorbent.

The addition of a modifier (5% methanol) to the extraction fluid (CO2) was not suitable with this trap, since it resulted in partial elution of PCDD/PCDFs to the hexane fraction weakening the fractionation.

Temperature and Pressure. The lower chlorinated PCDD/PCDFs were quantitatively extracted from sediment even with the mildest studied conditions (300 atm and 50 °C), but the extraction of more chlorinated PCDD/PCDFs (congeners containing over six chlorine atoms) were more challenging. The increase of temperature from 50 to 100 °C at 300 atm improved significantly (20-50%) the extraction of octa-chlorinated PCDD/PCDFs. Extraction efficiency was further improved by raising the pressure from 300 to 400 atm at 100 °C.  However, the concentrations of more than six chlorine atoms containing congeners were still lower than the average concentrations of those obtained in the intercalibration study. Because the toxicity of OCDD and OCDF are low compared to other 2378-substituted PCDD/PCDFs, the low extraction of these compounds did not have influence to the total PCDD/PCDF toxic load (TEQ) of the sample. The TEQs were similar with all studied SFE conditions and corresponded well to the results obtained by Soxhlet and intercalibration study [

Dynamic extraction time. In analysis of PCDD/PCDFs in soil contaminated by chlorophenols, SFE with 60 min dynamic extraction time (at 400 atm and 100 °C) showed results comparable to the traditional liquid based extraction methods (Soxhlet and ultrasonic) [14]. With these conditions comparable concentrations and toxic loads of PCDD/PCDFs were obtained with SFE and Soxhlet for all studied soil samples. Also the reproducibility of SFE was comparable to that of Soxhlet.

Fluid flow rate. The results of fluid flow rate test (at 400 atm and 100 °C) showed that the flow rate of CO2 did not affect to the extraction efficiency of PCDD/PCDFs in sediment [15] or soil [14]. However, slightly more PCDD/PCDF compounds were retained on the lines of the SFE instrument after extraction with the lowest fluid flow rate (1 ml/min). Because that was only 1% of the concentration found in soil, this does not affect to the results of consecutive samples when the PCDD/PCDFs concentrations are at the same level. However, when the levels of consecutive samples differ by orders of magnitudes, the contamination risk exist. This can be minimized by extracting the cleaner samples at the beginning of the series, if the levels of the samples are known, and/or by extractions of several blanks in sample series.

4 Conclusions

At the best obtained conditions (400 atm, 100 °C, 60 min dynamic time) SFE method showed to be effective in extraction of lower chlorinated PCDD/PCDFs and in analysis of toxic load of the sample.

Compared to the traditional extraction methods and clean-up procedures, the SFE method developed here is significantly faster and more economic. Furthermore, the exposure of laboratory personnel to hazardous chemicals and reagents is clearly decreased. The consumption of solvents was over 30 times less with SFE than with Soxhlet and ultrasonic methods. Because the same carbon was used for all samples with SFE, the relative costs of reagents for one sample were significantly decreased. The more samples are extracted with SFE the more economic method SFE is compared to Soxhlet and ultrasonic methods.

In addition to lowered consumption of solvents and reagents, also the working time used for analysis was remarkably decreased with SFE. Laboratory personnel’s working time used for pre-treatment of eight samples was about 40 hours with Soxhlet method, whereas with SFE that was only 6 hours. The total time consumed for pre-treatment of eight samples (including extraction) was six days with Soxhlet and two days with SFE.

Due to the benefits of SFE compared to traditional methods, SFE is a preferable method for analysis of levels of PCDD/PCDFs in contaminated soil and sediment. The developed SFE method has been applied, with minor modifications, also for extraction of PCBs in contaminated soil and sediment. Both SFE methods (PCDD/PCDFs and PCBs) have now been accredited by Finnish certification system in our laboratory.

5 References

[1] B. Bavel, 5th Round of international Intercalibration Study, Uumeå, Sweden 2000.Wageningen evaluation programmes for analytical laboratories, WEPAL, International Sediment Exchange for Tests on Organic Contaminants, SETOC, Amsterdam, Netherlands).
[2] Ontarion interkalibrointi
[3] T. Vartiainen, K. Lampi, K. Tolonen, J. Tuomisto, Chemosphere, 30 (1995) 1439.
[4] M. M Schantz, S. Bowadt, A. Bruce, B. A. Benner, et al., 1998, J. Chromatog. A.
816, 213-220.
[5] Berg, B. E.; Lund, H. S.; Kringstad, A.; Kvernheim, A. L., 1999, Chemosphere 38, 587-599.
[6] Lee, H-B.; Peart, T. E., 1994 J. Chromatog. A. 663, 87-95.
[7] P.Lampi, K. Tolonen, T. Vartiainen, J. Tuomisto, 1992. Chlorophenols in Lake     Bottom Sediments: a Retrospective Study of Drinking Water Contamination. Chemosphere 24,1805-1824.
[8] Van den Berg M., Birnhaum L., Bosveld A.T.C., et al., 1998, Toxic Equivalency Factors (TEFs) for PCBs, PCDDs, PCDFs for Humans and Wildlife. Environ Health Perspect. 106, 775-792.

6 Publications

Extended abstracts

[9] Mannila M., Koistinen J. and Vartiainen T., Development of a Solid Phase  Carbon Trap in Supercritical Fluid Extraction for Determination of PCDDs and PCDFs in Soil Samples, Dioxin `98, Stockholm, August 17-21, 1998, Organohalogen Compounds, 1998, 35,137-140.

[10] Koistinen J., Mannila M., and Vartiainen T., Estimation of the Level of PCDD/PCDFs in Soil Contaminated with a Chlorophenol Formulation using Supercritical Fluid Extraction, Dioxin `98, Stockholm, August 17-21, 1998, Organohalogen Compounds, 1998, 35,123-126.

[11]  Mannila M., Koistinen J. and Vartiainen T., Comparison of SFE with Soxhlet and Ultrasonic Extractions for the Determination of PCDD/PCDF in Soil Samples, Dioxin ’99, Venice, September 12-17, 1999, Organohalogen Compounds, 1999, 40, 197-200.

[12] Mannila M., Koistinen J., Vartiainen T., Optimization of Supercritical Fluid Extraction for Determination of PCDD/PCDF in Soil, 5th International Symposium of Supercritical Fluis, Atlanta, April , 2000., poster presentation.

Submitted articles

[13] Mannila M., Koistinen J., Vartiainen T., Solid Phase Adsorption Trap in Supercritical Fluid Extraction for Determination of Dioxins in Soil, J Chromatogr. A, Submitted for publication

[14] Mannila M., Koistinen J., Vartiainen T., Optimisation of Supercritical Fluid Extraction and Comparison to Traditional Extraction Methods in the Estimation of Toxic Load of Dioxins in Soil, J Chromatogr. A, Submitted for publication.

[15] Mannila M., Koistinen J., Vartiainen T., Effects of Temperature, Pressure and Fluid Flow Rate on Supercritical Fluid Extraction of PCDD/PCDFs in Sediment, Anal. Chem., Submitted for publication.
 

[ Projects | Main Page ]