Chemosphere, Vol. 36. No. 9. pp. 2007-2017, 1998 0 1998 Elsevier Science Ltd Pergamon

PII: SO0456535(97)10085-6 All rights reserved. Printed in Great Britain

0045-6535/98 $19.00+0.00

INDUSTRIAL POLLUTANTS IN GROUND WATERS FROM NORTHERN MILAN

Elena Fattore’t, Laura Mtlllerl, Enrico Davolit, Donata CastellP and Emilio Benfenatil

1Laboratorio di Farmacologia e Tossicologia Ambientali, Istituto di Ricerche Farmacologiche “Mario Negri”,. Via Eritrea 62,20157 Milano.

2Dipartimento di Prevenzione, Azienda USSL Ambito Territoriale No 32.

(Received in Germany 22 July 1997; accepted 10 October 1997)

Abstract Ground water samples from an industrialised area near Milan were analysed by gas chromatography-mass spectrometry (CC-MS) to identify the main pollutants and to quantify two classes of chemicals: polychloro-

1,3-butadienes (PCBD) and some aromatic amines. The water contained several halogenated aromatic and aliphatic compounds and heavy contamination due to PCBD, probably arising from contaminated land where a disused chemical plant is located. All the samples contained low levels of aromatic amines indicating a diffuse contamination probably arising from different sources. 01998 Elsevier Science Ltd

Introduction The north-west area of Milan is one of the most populated and industrialised regions in Italy and ground water contamination has been already described [ 11. This area houses an important dye manufacturing plant, disused since the ‘70. Chemical wastes, many of them not yet identified, have been widely dispersed into the soil or stored in basins and tanks, some of them without any waterproofing. Qualitative and quantitative analysis were done to investigate the type and the degree of contamination in ground water samples from Limbiate, Senago and Bollate, three municipalities lying south of this ex-

industrial area (Figure 1). On the grounds of qualitative analysis and of chemical production in the area, levels

of polychloro- 1,3-butadienes (PCBD) and aromatic amines were investigated. Very little can be found on PCBD in the literature. They are of industrial origin since they are by-products of the production of chlorinated solvents or intermediates of rubber compounds manufacture [ 1,2], and they can persist in the environment [3]. Hexachloro-1,3-butadiene (HCBD), the most widely studied, is included in the EPA list of priority pollutants. It was used as a pesticide in the former Soviet Union [4] and 1,1,3,4- tetrachloro- 1,9butadiene has been detected in the environment in Eastern Europe [5]. Aromatic amines such as aniline and its substituted derivatives are much better known. Like PCBD, they may be released into the environment from industrial sources since they are residues and intermediates of several chemical manufacturing processes such as dyestuffs, cosmetics, medicines, rubber compounds, etc. [6-91. They are generally dangerous on account of their toxicity and carcinogenicity [8-91 and they can be converted to dangerous N-nitroso compounds in the environment [5].

2007

2008

Figure 1. Map of the area showing the wells sampled and the ground water levels.

Experimental part

Sampling

Ground water samples were collected at Limbiate, Senago and Bollate (Figure 1) before and after active

carbon purifying treatment during four different sampling campaigns (Nov 95, Jan 96, Ott 96 and Nov 96).

2009

The bottles were filled completely so as to avoid loss of volatile compounds. The wells analysed, and their

depths, are listed in Table I.

Table I. Wells sampled.

Town Well Depth (m)

Limbiate 1L 105.7 2L 105 3L 114.5 4L 130 5L 110 6L 108 1OL 102 12L 102 2(1L 80

Senago 4s 70 6s 103.2

Bollate 3B 60.5 1OB 64.1

Chemicals

1,1,4,4-Tetrachloro-1,3-butadiene (TCBD), purity 91%, 1,1,2,4,4-pentachloro-1,3-butadiene (1,1,2,4,4-

PCBD), purity 97%, cis-1,1,2,3,4-pentachloro-1,3-butadiene (cis-1,1,2,3,4,-PCBD), purity 95% were synthesised in the laboratories of the Institute of Physical Organic Chemistry, Minsk, Belarus. HCBD was

purchased from Sigma-Aldrich (St. Louis, MO, USA), and 1,3,5-trichloro-2,4,6trifIuorobenzene, used as internal standard, from Fluorochem (Derby, UK). The following aromatic amines were tested for in the ground water (Figure 2): aniline (Ciba Geigy, Bologna, Italy), N,N dimethylaniline, 2,6 dimethylaniline (Fluka, Buchs, Switzerland), 2,6-diethylaniline, ortho-toluidine, para-chloroaniline, 2,4-dichloroaniline, 2.5 dichloroaniline, 3,4-dichloroaniline and 3,5dichloroaniline (Aldrich, Steinheim, Germany), 2,4- dimethylaniline, diphenylamine (Merck, Darmstadt, Germany), para-nitroaniline (Carlo Erba, Milan, Italy).

Extraction

Solid phase extraction (SPE) was used for qualitative analysis of 1 L ground water samples from well 4 at Limbiate. To identify the largest possible number of compounds in the samples, different extractive materials

were used: Carbopack B (Supelco, Bellefonte, PA, USA), Cl8 (International Sorbent Technology, UK), and Lichrolut EN (Merck). The SPE extraction procedures are shown in Table II. Solid phase microextraction (SPME) [lo- 121 was also employed for qualitative and quantitative analysis of PCBD and aromatic amines.

The detailed analytical methods have been described [13,14]. Qualitative analysis by SPME was done with all the coating fibers commercially available (polydimethylsiloxane 100 pm, polyacrylate, polydimethylsiloxane/divinylbenzene, Carbowax/divinylbenzene) and a 80 pm carbon fiber. The SPME

holders and the different coating fibers were supplied by Supelco. All the analysis were carried out within 72 hours of sampling. For SPME analysis, 10 mL ground water samples were transferred to 10 mL vials and extracted by direct immersion of the fiber for 30 min under magnetic stirring at room temperature.

2010

Table II. Extraction procedures for qualitative analysis.

Carbopack Cl8 Lichrolut EN

Washing 10 mL ethyl acetate 10 mL methylene chloride/methanol 8/2 5 mL acetonitie

Activation 10 mL methanol a. 5 mL methanol 5 mL distilled water b. 20 mL sol. 10 mg/ml ascorbic acid HCI 0.01 M solution

Extraction 1 L of sample 1 L of sample 1 L of sample

Eluition 10 mL ethyl acetate a 10 mL methylene chloride/methanol 8t2 6 mL acetonitrile b.10 mL chloride/methanol 8/2 and hifluoroacetic acid 0.2% v

Final volume lOOa lOOIlL 1OOUL

Instrumental Analysis

GC-MS analysis was carried out on a HP 5980

- MSD 5971. The capillary column used for qualitative analysis was a CP Sil 8, 25 m x 0.25 mm, 0.25 pm of stationary phase,

purchased from Chrompack (Middelburg, The Netherlands). The temperature programme was 70°C x 2 min, 8Wmin until 3OO’C x 4 min. The injector temperature was 240°C. For

PCBD quantitative analysis a Chrompack CP

Sil 8, 25 m x 0.25 mm, film thickness 1.2 pm

was used; temperature programme 60°C x 5

min, 12Wmin until 2OO’C; injector

temperature 2OO’C [ 131. The gas

chromatography conditions for the aromatic

amines were: capillary column from Supelco

PTA 5 30 m x 0.25 mm, film thickness 0.5

pm, base-deactivated; temperature programme

60°C x 5 min, 10”Clmin until 250°C x 3 min;

injector temperature 26O’C [14]. The

desorption times for SPME fiber in the GC

injector for PCBD and aromatic amines were 5

and 10 min respectively.

93 12

107 15

12, 18

121 18

127 57

149 23

162 03

162.03

162.03

13s 12

169 22

Figure 2. Aromatic amines sought in ground water samples.

Results Quulitative analysis

Table III lists the compounds, extracted by SPE, identified in well 4L at Limbiate, one of the most

contaminated. A large number of compounds were also identified by the various extractive phases of the

2011

SPME. An investigation concerning on the sensitivity and selectivity of the SPME fibers toward the pollutants

will be presented elsewhere.

Table III. Compounds identified by GC-MS in untreated ground water sample from well 4L at Limbiate extracted by C 18 cartridge, with their relative GC peak abundance.

Ret. Time Compound Abundance

13.16

13.61

13.95

14.21

14.86

15.02

15.08

15.33

15.45

16.04

16.14

16.87

17.21

17.35

17.69

18.88

18.99

19.16

19.54

20.36

21.65

22.08

23.14

24.26

29.68

30.63

31.64

33.97

1,1,2,2-Tetrachloroethane

Erhene, l,Zdichlorc-

Benzene, l.3dichloro-

2,T-Azobis (2-methyl-propanenitrile)

Tetrachlom-1,3-bute

TetrachIom-1,3-butiene

Tetmchloro-1,3-butadiene

Ethane, hexachloro

1 -Propene, bromochlorodifluoro

Tetrachloro-1,3-buticne

Tefmchlom-1,3-butiiene

Penrachloro-1,3-butadiene

Pentachlor~1,3-bute

1,2,3,3-Tetrachloro-I-propene

Hexachlorc+l,3-butadiene

I-Dodecanol, 3,7,1 I-@imethyl-*

Cyclohexasiloxane, dodecamethyl-

Dodecane, 2,6,10+imethyl-*

2,3-Dichloroaniline

1,2,3,3-Tetrachloro 1-propcne*

2-Isopropyl-5-oxohexanal*

Butylated hydroxytoluene

Propanoic acid, 2-methyl-, I-(l,ldiiethylethyl)-2-methyl-1,3-propanedyl ester

Benzamide, 2,6dichloro-*

Pentadecane*

Hea*

Heptadecane, 2-methyl*

Nona&cane*

18OCQ

25000

12ooo

98ooo

1OCQO

3cGW

172OCO

5oooo

13ooo

3OOOO

5OOcQ

16000

44ooo

15OCKl

2oooo

26OC0

8000

8OcG

loo00

9ooo

25OiM

6ooO

13000

13OiM

8000

2OOiM

34000

45OW

50000

* Uncertain identification

The largest number of compounds were detected by SPE with Cl8 cartridges. Except for some PCBD that were subsequently identified by reference standards, identification was obtained by comparison with mass spectra in the NBS75K library of the analytical instrument. Several peaks in the chromatogram (about 30%) could not be identified from the library. A tetrachloro-1,3-butadiene was the most abundant compound detected in this analysis. Most of the substances were aromatic or aliphatic halogenated (mainly chlorinated)

2012

compounds. Several pollutants were the same as those detected by Botta and co-workers [I] in a bottom

fraction from a rectifying column of tetrachloroethene. At the end of the chromatogram various long-chain

alkanes were recognized even though the lengh of the chain was not identified with certainty. A chlorinated

aniline was also identified at the retention time of 19.54 min. These results showed heavy chemical pollution

of this well, probably related to the chlorinated solvent production.

well 1L

2 -: .z

well 1OL

Figure 3. F’CBD concentrations (ng/mL) in raw ground water samples at Limbiatc.

2013

Figure 4. PCBD concentrations (ng/mL) in raw ground water samples at Senago and Bollate.

PCBD quantitative analysis

Figure 3 shows the average concentrations of PCBD in wells lL-6L at Limbiate, arising from two different

sampling campaigns, and the concentrations in private wells 1OL and 12L that were sampled in a different campaign. Except for 1,1,2,4,4-PCBD and HCBD in well 6L, that were near the limit of detection of the analytical method (0.025 ng/mL and 0.05 ng/mL respectively), these compounds were always found. These and the following results refer to the raw well waters; in the corresponding samples after the purifying treatment done before distribution of drinking water, the PCBD content was below the limit of detection.

The highest levels were detected in wells 4L and 2L where the average for TCBD was 28.3 ng/mL and 8.54 ng/mL respectively, while the lowest levels were detected in well 6L where TCBD was 0.24 ng/mL. Wells

1OL and 12L subsequently sampled showed concentrations of TCBD of 2.24 ng/mL and 17.34 ng/mL respectively. The TCBD concentrations, as well as the concentrations of the other PCBD, generally decrease going south, excepting wells lOL, 6L and 5L. Nevertheless if we assume that the source of the pollution lies to the north of Limbiate this could be explained by the direction of the underground flow: the ground water levels shown in Figure 1 indicate a prevailing path of the waters in the area of Limbiate. Thus the wells with the highest

concentrations should be located along this main flow, while the wells with lower concentrations should be

further out, and would only be affected by pollutants arising from lateral diffusive processes. The relative amounts of these compounds were generally similar for all the Limbiate samples: TCBD was always the most abundant, representing about 80% of the total PCBD, while 1,1,2,4,4-PCBD, cis-1,1,2,3,4- PCBD and HCBD were about 3, 10 and 7%. This suggests that the source of these pollutants is the same for all these wells.

2014

Figure 4 shows the results of wells 4s and 6s at Senago and wells 3B and 10B at Bollate. The highest levels

were in well 4s in Senago where TCBD and HCBD values were 1.9 and 0.9 ng/mL, respectively. In these

wells the relative amount of HCBD was higher than in Limbiate, representing from 17 to 49% of total PCBD.

This suggests another source for this compound since several industrial activities are present in that area. In a

recent investigation HCBD was quantified in the some wells from Senago and Bollate. The values were a little

lower but comparable to this study.

Aromatic Amines

Several aniline derivatives for which a SPME analytical method was optimised (Figure 2), were not found in

the ground water samples. The levels of the aromatic amines detected are presented in Figure 5a and 5b. The

concentrations were substantially lower than those of PCBD. Except for 2,4-2,5dichloroaniline in wells 4L

and 12L at Limbiate the concentrations do not reach the guideline value of 0.1 ng/mL set by Italian regulations

for pesticides in general (no specific law limits aromatic amines). PCBD and 2,4-25 dichloroaniline showed a

similar pattern, with a correlation between TCBD and 2,4-2.5 dichloroaniline (r*=0.96, P< 0.001) (Figure 6).

That suggests that PCBD and 2,4-2,5-dichloroaniline arise from the same place; nevertheless, since only a

few wells were analysed, further studies are needed to investigate this. A meaning correlation can also be seen

for 3.4. and 3,5-dichloroaniline (r*=0.78, P< 0.01) (Figure 7). 3,4-dichloroaniline is also a product of

degradation of various herbicides as diuron, propanil, linuron etc., nevertheless there is not evidence that such

pesticides were applied in the area. No correlation were found for other amines. Thus contamination due to

these compounds may arise from different sources

0.12 , 1 0.1 I ::::::::::::: well 12L ::::::::::::: well 20L ~~~1.~~~ :Ij ~ ~

Figure 5a. Aromatic amines (ng/mL) in raw ground water samples at Limbiate.

Legend: p-CA = pm-chloroaniline; 2,4-25DCA = 2,4-dichloroaniline and 2,Michloroaniline; 3,4-DCA = 3,4-dlcbloroaniline; 3,SDCA = 35dichloroaniline.

2015

0.075

0.05

0.025

0

0.014

0.007

0

0.045 I

0.03

0.015

0

well IOB

Figure 5b. Aromatic amines (ng/mL) in raw ground water samples at Senago and Bollate.

Figure 6. Correlation curve between PCBD and 2,4-2,6-DCA.

2016

0.025

3.5.DCA

Figure 7. Correlation curve between 3,4 DCA and 2,5-DCA

Acknowledgements We thank the Fondazione Lombardia per l’Ambiente, Milan, Italy and the EC prqject ENV4-CT95-020 for

financial support. We also thank Dipartimento di Prevenzione Azienda USSL Ambito Territoriale N 32 for

sampling. L. M. is recipient of a grant from Banca Commerciale Italiana.

References

1.

2.

3.

4.

5.

6.

D. Bona, E. Dancelli and E. Mantica, A case of history of contamination by polychloro-1,3-butadiene

congeners, Environ. Sci. Technol. 30, 453-462 (1996).

Ingannamorte B., Inquinanti organici non convenzionali nelle acque di falda a nord di Milano: il case dei

policlorobutadieni, Graduation Thesis, Universita degli Studi di Milano, Milan, Italy (1995-1996).

T. N. P. Bosma, F. H. M. Cottaar, M. A. Posthumus, C. J. Teunis, A. van Veldhulzen, G. Schraa and

A. J. B. Zehnder, Comparison of reductive dechlorination of hexachloro-1,3-butadiene in Rhine

sediment and model systems with hydroxocobalamin, Environ. Sci. Technol. 28, 1 I24- 1128 (1994).

WHO (World Health Organization), Hexachloro-1,3-butadiene, Environmental Health Criteria 156

(1994).

I. Liska, Long Term Pesticide Screening in the Nitra River Basin, presented at the 5th Symposium on

Chemistry and Fate of Modem Pesticides, Paris 6-8 September, 1995, abstract L3 1.

H. Kataoka, Derivatization reactions for the determination of amines by gas chromatography and their

applications in environmental analysis, J. Chromatogr. A. 733, 19-34 (1996).

2017

7. L. M. Games and R. A. Hites, Composition, treatment efficiency, and environmental significance of dye

manufacturing plant effluents, Anal. Chem. 49, 1433-1440 (1977).

8. R. D. Voyksner, R. Straub, J. T. Keever, H. S. Freeman and W-N. Hsu, Determination of aromatic

arnines originating from azo dyes from chemical reduction combined with liquid chromatography/mass

spectrometry, Environ. Sci. Technol. 27, 16651672 (1993). 9. S. Laha and R. G. Luthy, Oxidation of aniline and other primary aromatic amines by manganese dioxide,

Environ. Sci. Technol. 24, 363-373 (1990).

10. C-S. Lu and S-D. Huang, Trace determination of aromatic amines or phenolic compounds in dyestuffs

by high-performance liquid chromatography with on-line preconcentration J. Chromatogr. A 696,201-

208 (1995).

11. Z. Zhang, M. J. Yang and J. Pawliszyn, Solid Phase Microextraction, Anal. Chem. 66, 844A-853A (1994).

12. R. Eisert and K. Levsen, Solid-phase microextraction coupled to gas chromatography: a new method for the analysis of organic in water, J. Chromatography A 733, 143-157 (1996).

13. E. Fattore, E. Benfenati and R. Fanelli, Analysis of chlorinated 1,3-butadienes by solid-phase

microextraction and gas chromatography-mass spectrometry, Journal of Chromatography A 737, 85-9 1 (1996).

14. L. Mtiller, E. Fattore and E. Benfenati, Analysis of Aromatic Amines by SPME and GC-MS in Water Samples, J. Chromatography A, submitted.