Journal of Chromntography, 405 ( 1987) 3 1 l-3 17 Elsevier Science Publishers B.V., Amsterdam ~ Printed in The Netherlands

CHROM. 19 692

DETERMINATION TUTE UGILEC (TE

OF THE POLYCHLORINATED BIPHENYL SUBSTI- TRACHLOROBENZYLTOLUENES) IN FISH

PETER FlfRST*, CHRISTIANE KRUGER, HANS-ALBERT MEEMKEN and WILHELM GROE- BEL

Chemi.rche.s Landesuntersuchungsamt Nordrhein- Westfalen. Sperlichstr. 19, 4400 Miinster IF.R.G.)

(First received February IOth, 1987: revised manuscript received May 4th, 1987)

SUMMARY

A method is presented for the determination of tetrachlorobenzyltoluenes (TCBTs) in fish. TCBTs represent the major constituents of Ugilec, which is com- monly used as a substitute for polychlorinated biphenyls (PCBs) in underground mining. The compounds are extracted from the freeze-dried material together with fat and other lipophilic substances. Gel chromatography on Bio-Beads S-X3 is used for removal of fat followed by adsorption chromatography on silica gel (1.5% water). Combined gas chromatography-mass spectrometry employing an ion trap detector is used for the separation and detection of the compounds. The developed procedure provides identification and determination of TCBTs even in the presence of an excess amount of PCBs. The application of the procedure to the investigation of fish caught in rivers from areas with extensive mining indicates that Ugilec has already found its way into the food chain.

INTRODUCTION

Owing to their excellent physical properties, polychlorinated biphenyls (PCBs)

have enjoyed widespread use as industrial, particularly as dielectric fluids for capac- itors and transformers because they possess good thermal conductivity while their electric conductivity is extremely low. They have also played a significant role as hydraulic fluids in underground mining. However, PCBs represent’ a potential toxi- cological risk as they are accumulated in the food chain and lead to considerable bioconcentrations in animal and human adipose tissue. Their ubiquitous distribution is clearly demonstrated by PCB levels found in human milk.

Although since I972 the use of PCBs in the F.R.G. has been allowed only in closed systems, so far there has been no evidence of a decrease in PCB concentrations in human milk, in contrast to decreases in the concentration of DDT, dieldrin and hexachlorobenzene after they had been banned IJ. Moreover the disposal of PCB-

containing waste represents a severe problem, because incomplete incineration may lead to formation and emission of polychlorinated dibenzodioxins (PCDDs) and dibcnzofurans (PCDFs). Also, significant amounts of PCDDs and PCDFS have been introduced into the environment from fires in capacitors and tranSformers3’4.

0021-9673/87/$03.50 @) 1987 Elsevier Science Publishers B.V.

312 P. FtiRST et al.

In the past few years, the knowledge of these risks has led to efforts to replace PCBs, and suitable substitutes are available for almost all purposes. One group of substances that has replaced PCBs to a large extent in underground mining are tetra- chlorobenzyltoluenes (TCBTs) (Fig. 1). TCBTs are the major components in prod- ucts marketed as Ugilec 141, which is a mixture of pure TCBTs, and Ugilec T, which contains about 40% of trichlorobenzenes in addition.

Because of their good technical properties, which are comparable to those of PCBs, and in accordance with safety requirements, TCBT-containing products are widely used in hydraulic systems for underground mining. Although this use has been denoted a “closed” application, strictly it should be viewed as an “open” system, as the hydraulic fluid is “consumed” and, consequently, set free underground. Finally, TCBTs can enter the environment and contaminate the food chain, especially fish, through pit waters, mine outputs and ventilation systems5.

The simultaneous presence of PCBs makes the determination of TCBTs in fish and other foodstuffs very difficult. Because of their similar polarities, it is almost impossible to separate them selectively. Moreover, both groups have comparable volatilities, which leads to overlapping elution in gas chromatographic (GC) analysis. Hence, it is obvious that a reliable determination in such a complex matrix as foods needs the use of specific detectors. For this purpose a mass spectrometer is prefer- entially employed as a mass-selective detector. The great selectivity of mass-selective detection combined with the excellent resolving power of capillary GC provides an efficient determination of TCBTs, e.g., in fish, even in the presence of a large excess of PCBs.

EXPERIMENTAL

Principle of method After removal of the internal organs, head, tail, bones and scales, about 500

g of the fish samples were cut into small pieces and freeze-dried. From the freeze- dried material the fat was extracted with light petroleum (b.p. 40-60°C) together with TCBTs and other lipophilic substances such as organochlorine pesticides and PCBs. An aliquot of 3 g fat was used for analysis.

Gel chromatography on Bio-Beads S-X3 with cyclohexane-ethyl acetate (1: 1) was chosen for removal of neutral fat, followed by adsorption column chromato- graphy on silica gel (1.5% water) according to the multi-residue procedure of Specht and Tillk&. Elution with hexane provided separation of TCBTs together with PCBs from the main portion of organochlorine pesticides and other more polar com- pounds. Finally, the n-hexane extract was concentrated to 0.5 ml and analysed by capillary gas chromatography-mass spectrometry (GC-MS).

Fig. 1. Structure of tetrachlorobenzyltoluenes.

DETERMINATION OF UGILEC IN FISH 313

GC-MS system

All analytical work was carried out on an ion trap detector (Finnigan) con- nected with a DAN1 6500 gas chromatograph. The gas chromatograph was equipped with a PTV (programmed temperature vaporizer) injector and a 30 m x 0.32 mm I.D. fused-silica capillary column coated with DB-5 (0.1 pm) (J & W). A I-@ volume of each sample extract was injected in the splitless mode at an injector temperature of 50°C and an oven temperature of 80°C. After injection the injector was heated ballistically to 280°C. The oven temperature was held at 80°C for l min, then in- creased at 30”C/min to 140°C and at a rate of 4”C/min from 140 to 260°C. The gas chromatograph was directly connected to the ion trap detector, the fused-silica cap- illary column being pushed through the transfer line into the trap. The temperature of the transfer line was maintained at 280°C.

The ion trap detector was equipped with the new AGC (automatic gain con- trol) software. In comparison with the former version, the linear range could be increased by a factor of 100-1000. Moreover, the new software provided a consider- able improvement in sensitivity. As a result of this new acquisition technique, the sensitivity attainable in the linear scanning mode was comparable to that in selec- ted-ion monitoring (STM) which is normally employed in trace analysis. Hence the full mass spectra1 information important for structure elucidation of unknown com- pounds is combined with the high sensitivity of SIM. In this instance, linear scanning allowed the determination of Ugilec in fish at levels down to 0.02 mg/kg even in the presence of a large excess of PCB. Subsequently the fragmentograms of the com- pounds of interest were extracted from the total ion chromatogram by the choice of suitable masses. Control of the system and processing of the data were effected with an IBM-XT personal computer with a 20 Mbyte Winchester disc.

The quantification of TCBTs was accomplished by an external standard method. Calibration graphs were obtained by analysing different concentrations of Ugilec 141.

RESULTS AND DISCUSSION

Fig. 2 shows the gas chromatogram of Ugilec 141. The mass spectra (Fig. 3)

1m

TM

l!db+J i 1299 1398 1689

1821 29:el ai:u #?l 2#l 26:4l n-l,”

Fig. 2. Gas chromatogram of Ugilec 141, a technical mixture of TCBTs.

314

l&r&

P. FURST et al.

285

328

Fig. 3. Electron-impact mass spectrum of a single TCBT isomer.

of the individual isomers are very similar. The intense molecular peak at m/z 3 18 has an isotope pattern typical of four chlorine atoms. This is important for analysis by mass selective detection with regard to high selectivity and sensitivity.

Fig. 4 compares the total ion chromatogram of a PCB mixture containing Clophen A30 and A60 with that of Ugilec 14 1. It can be clearly seen that the TCBTs elute within the range of retention times of PCBs. Therefore, GC analysis with elec- tron-capture detection (ECD) of TCBTs is affected whenever there are residues of

A

Fig. 4. Total ion chromatograms of a mixture of Clophen A30/A60 (A) and Ugilec 141 (B).

DETERMINATION OF UGILEC IN FISH 315

318

318

1im 1688 28:91 21:41 2Fi 25: 01 26141

min

Fig. 5. Mass fragmentograms (m/z 318-320) reconstructed from the total ion chromatograms (shown in Fig. 4) of the Clophen A30/A60 mixture (A) and Ugilec 141 (B).

PCBs present. As residues of PCBs are probable contaminants when examining fish, a reliable GC-ECD analysis of small amounts of TCBTs is nearly impossible.

Mass-selective detection, in contrast, permits an accurate determination of TCBTs even in the presence of high PCB levels. Fig. 5 shows the fragmentograms of the mass range m/z 3 18-320 extracted from the total ion chromatogram of a Clophen and Ugilec mixture. Although the PCB concentration is about 100 times higher, the mass fragmentogram reveals only a few small peaks that do not interfere with the determination of TCBTs. The only interfering pesticide so far recognized is the DDT metabolite DDE, which is also registrated within this mass range but elutes directly in front of the first TCBT isomer.

The choice of suitable masses consequently provides a selective reconstruction of fragmentograms of special compounds or groups of substances from the total ion chromatogram.

Fig. 6 shows the total ion chromatogram of a pike extract, which is charac- terized by a large number of peaks due mainly to PCBs. The simultaneous presence of TCBTs becomes clear on extracting the fragmentogram for masses m/z 318-320

iis 25:01

min

Fig. 6. Total ion chromatogram of a pike extract.

316 P. FtiRST et ~11.

Fig. 7. Chromatograms of a pike extract. (A) Total ion chromatogram. (B) Reconstructed mass fragmen- togram of mass range miz 318-320.

(Fig. 7). The resulting peak pattern is extremely similar to that of Ugilec 141. The identification of TCBTs was additionally confirmed by full mass spectra. In this sample the concentration calculated as Ugilec 141 was 2.5 mg/kg based on eatable fish tissue.

All fish samples caught in rivers from areas with extensive mining revealed the presence of TCBTs at a mean level of 2.6 mg/kg calculated as Ugilec 141 based on the eatable portion. The levels ranged from 0.1 to 25 mg/kg. On the other hand, no TCBTs could be determined down to a detection limit of 0.02 mg/kg of Ugilec in the eatable portion in fish from rivers into which no effluents from underground mining are introduced.

The identification of TCBTs in fish demonstrates that PCB substitutes are already contaminating the food chain. In fact, normally fish from the examined areas is not put on to the domestic market but is eaten at most by anglers. Nevertheless, it is to be feared that TCBTs will lead to ubiquitous pollution similar to that with PCBs. The results also indicate that a mass-selective detector is necessary for iso- mer-specific PCB determination at least in strongly contaminated samples to exclude the possibility of false positive results due to residues of TCBTS.

ACKNOWLEDGEMENT

The authors gratefully acknowledge the gift of the Ugilec mixture from British Petroleum, F.R.G.

REFERENCES

Riickstiinde und Verunreinigungen in Frauenmilch, Deutsche Forschungsgemeinschaft, Mitteihmg XII der Kommission zur Priifung van RtickstBnden in Lebensmitteln, Verlag Chemie, Weinheim, 1984. H.-A. Meemken and Chr. Kriiger (Editors). Bericht fiber die Untersuchung von Frauenmilch auf Or- ganochlorpestizide und polychlorierte Biphenyle im Jahre 1984, Chemisches Landesuntersuchungsamt NW, Miinster, 1985. A. Schecter, Chemosphere, 12 (1983) 669.

DETERMINATION OF UGILEC IN FISH 317

4 B. Janson and G. Sundstriim, in 0. Hutzinger, R. W. Frei, E. Marian and F. Pocchiari (Editors), Chlorinated Dioxins and Related Compounds. Impact on the Environment (Environmental Sciences Series, Vol. 5), Pergamon Press, Oxford, 1982, pp. 201-205.

5 H. Lorenz and G. Neumeier (Editors), Polychlorierte Biphenyle. Gemeinsamer Bericht des Bundesge- sundsheitsamtes und des Umweltbundesamtes, MMV Medizin Verlag, Munich, 1983.

6 W. Specht and M. Tillkes, Fresenius’ 2. Anal. Chem., 322 (1985) 443.