Fate of aromatic sulfonates in fluvial environment

Pergamon Chemosphere, Vol. 35, No. IO, pp. 2295-2305, 1997 Q 1997 Elsevier Science Ltd

All rights reserved. Printed in Great Britain 0045-6535/97 $17.00+0.00

PII: SO0456535(97)00308-l

FATE OF AROMATIC SULFONATES IN FLUVIAL ENVIRONMENT

Orfeo Zerbinati’, Marco Vincenti, Sara Pittavino and Maria Carla Gennaro

Dipartimento di Chimica Analitica, Universita di Torino, v. Giuria 5, I-10125 Torino, Italy

(Received in &many 14 January 1997; accepted 30 May 1997)

ABSTRACT

Water samples of the Italian river Bormida, which is polluted by the wastes of the production

of azo-dyestuff intermediates, were analysed by FAB/MS and the resulting data were compared to

those obtained by HPLC. The presence of sulfonated derivatives of naphthalene was confirmed but,

in addition, several sulfonated compounds having molecular mass other than those of

naphthalenesulfonates (NS) were also found. Those compounds were suspected of originating from

oxidative degradation of NS’s. Laboratory tests showed that some NS’s indeed can undergo

oxidative degradation under physico-chemical conditions similar to those occurring in a river. In

particular, the dark-coloured degradation products of 1-hydroxy-2-naphthalenesulfonic acid

appeared similar to an unknown compound found in the river water. @I 997 Elsevier Science Ltd

INTRODUCTION

The presence of sulfonated derivatives of aromatic hydrocarbons in river and sea water has

been investigated [l-13]. These compounds are utilised in the manufacturing of azodyestuffs,

optical brighteners, ion-exchange resins, concrete plasticisers and pharmaceuticals. Owing to their

good water solubility and to their xenobiotic properties, they are often transported by the rivers quite

far from the point where they have been emitted.

Among aromatic sulfonates, only those classified as surfactants were studied from the point of

view of their biogenic transformation and toxicity in the environment [14], while studies regarding

ecotoxicity [15] and environmental fate of the sulfonic acids generated as substrates, by products

and degradation intermediates have appeared only in the last few years. A few paper dealt with

biogenic transformations of naphthalenesulfonic acids [17-241. All these studies were performed

2295

2296

under laboratory conditions or in water purification plants, but none of them regarded the

transformations occurring in a river.

The water of the Italian river Bormida becomes visibly darker as it flows far from the point

where a chemical industry emits its wastes, containing aromatic sulfonates. As no further emission of

pollutants is present, water darkening was attributed to the formation of coloured by-products from

the chemical transformation of colourless aromatic sulfonates. In order to support this hypothesis,

mass spectrometric analysis on samples of river water was performed, and the fate of a number of

commercially available aromatic sulfonates was evaluated, under physico-chemical conditions

simulating the ones occurring in a river. Fast atom bombardment (FAB) was utilised for ionising the

anaiytes, since FAB proved effective in the analysis of mixtures of ionic substances, as it produces

singly charged molecular ions with negligible fragmentation. Then, the application of tandem mass

spectrometry (MS/MS) to the molecular ions generated by FAB provided fragmentation and,

consequently, structural information. As a matter of fact, FAB-MS/MS has already been applied to

the analysis of sulfonated dyes in water [25].

To reduce the interference of the sample matrix and to increase the sensitivity, FAB-MS

analysis was combined with a solid phase extraction (SPE) sample pre-treatment, aimed to the

isolation and concentration of the anionic analytes from water samples, while liquid

chromatographic techniques and IR spectroscopy were utilised for product quantitation or

identification.

EXPERIMENTAL

Standards - Pure aromatic sulfonic acids standards were obtained from either Merck, Aldrich

or Kodak.

SPE - SPE was employed for isolating analytes from laboratory test solutions and from real

world samples. 100 mL of aqueous sample were extracted with a Merck C18, 400 mg SPE cartridge,

loaded with cetyltrimethylammonium bromide (CTAB). The substances retained in the cartridge were

eluted by 2 mL of methanol, then reduced to 0,5 mL by a gentle stream of nitrogen.

HPLC - HPLC analyses were performed on a C8, $I, 250 x 4.6 mm HPLC column with

acetonitrilelwater 56146 eluent, adjusted at pH 7.0, containing 7 g/L CTAB and 1 g/L citric acid. UV

detection at 254 was used. Complete description of SPE and HPLC procedure has been reported

elsewhere [4].

FAB-tandem MS - A Finnigan MAT 90Q tandem mass spectrometer, equipped with a FAB

interface was employed for the analysis of both river water extracts and laboratory test solutions. 1%

thioglycerol was added to the samples and 1 ul of the resulting solutions were placed into the FAB

vessel. Two experimental techniques were adopted for MS/MS studies of the ions generated by the

FAB source: (i) fragmentation of selected ions, for MS/MS studies and (ii) dual-scan MS/MS

analysis. For MS/MS studies, collision-induced fragmentation of the sample’s negative ions was

produced with 40 eV energy and 1.5 uTorr Argon gas into the second ionisation chamber. For dual

scan MS/MS analysis, preliminary experiments were conducted by fragmenting the molecular ion of standard sulfonic acids, that showed the (m - SO*)- and SOs- fragments. Thus, for obtaining

2297

maximum selectivity when complex mixtures had to be analysed, the first and the second MS

analyser were operated in the dual scan mode, with a 64 m/z gap among them. Therefore, only

those negative ions undertaking a neutral loss of 64 (SO2) were selected, and disturbance due to

interfering ions was largely reduced.

TLC - TLC separation of the degradation products of 1-hydroxy-2-naphthalenesulfonic acid

was effected on Merck Silica Gel 60 W, 5 x 10 cm plates, which were oven dried at 1 IO “C for two

hours before use. A methanol/IO% aqueous ammonia, 90110 eluent was used. Two brown spots

were obtained (Rf 0.77 and 0.87). These two spots were scraped off and then extracted with

methanol. Extracts were evaporated on KBr pellets, for FTIR analysis.

Degradation tests - Separate solutions of 10 mg/L of each compound listed in Table I were

prepared in 1 mM sodium hydrogen carbonate/O,5 mM calcium chloride solution. The resulting

solutions were then filtered at 0.22 urn to avoid microbial contamination and then placed into

transparent glass vessels, previously sterilised, equipped with rubber septa for collecting samples by

means of sterile syringes. The solutions were aerated through 0.22 urn filters (Elga UHQ-4 vent

filter); the vessels were exposed to normal daylight and samples of each solution were collected at

regular time intervals for HPLC analysis.

FTIR - FTIR spectra were collected by a Perkin Elmer 1710 FTIR spectrometer.

RESULTS AND DISCUSSION

The HPLC analysis with combined UV and fluorescence detection of two river Bormida water

samples evidenced the following compounds: 2-naphthalenesulfonic acid, 0.32 mgl-1; 2,7-

naphthalenedisulfonic acid, 3.46 mgl-l ; 2-hydroxy-6,8-naphthalenedisulfonic acid, 0.65 mgl-1; I,5

anthraquinonedisulfonic acid, 1 .I mgl-1 The identity of the fluorescent naphthalenesulfonic acids

was confirmed by means of their fluorescence spectra, while dual-wavelength absorbance ratio

supported the identification of 1 $anthraquinonedisulfonic acid. Additional peaks appeared in the

HPLC chromatograms, but they were not identified because they did not correspond to anyone of the

available standards [5].

To achieve further information, the same extracts were analysed by FABlMS according to the

procedure described previously. Fig. 1 shows the FABldual-scan MS spectra of the SPE extracts of

river water samples. As it was expected, peaks were found at m/z 383, 382, 367, 319, 318, 317, 303,

302, 287, 239, 238, 237, 223, 222 and 207, corresponding to the masses of the position isomers of

naphthalene or anthraquinone substituted by various combinations of amino, hydroxy and sulfonate

groups, As an example, a naphthalene derivative containing one amino and two sulfonic acid

substituents, having thus a molecular mass of 303, will originate a mass peak at 302 m/z, since FAB

extracts from the sample matrix only singly-deprotonated anions. However, the peaks at 352, 309,

271, 194 and 161 can not belong to amino or hydroxy substituted naphthalene or anthraquinone

sulfonates. Further, the intensity of the signal at m/z 207 is unexpectedly high, compared to the

intensities of the peaks corresponding to remaining compounds, which were already identified by

HPLC, also considering their relative concentrations. In fact, the concentration of 2,7-

naphthalenedisulfonic acid (m.w. 288) in the sample is approximately ten times larger than the one

2298

of 2-naphthalenesulfonic acid (m.w. 208) while the peak at m/z 207 is about two times higher than

the peak at m/z 287. Similar observations can be done if 2-hydroxyb,8-naphthalenedisulfonic acid

(m/z 303) or 1 JCanthraquinonedisulfonic acid are considered. Therefore, assuming similar ion-

generating efficiencies for the various naphthalenesulfonates, the relative intensity of the 207 peak

seems much higher than the one expected on the basis of the concentrations of this compound in

the sample.

Further MS/MS experiments were performed on the 207 and 209 ions obtained from the river

water extracts, by fragmenting them in the second collision chamber of the MS spectrometer and

then examining the resulting ions. In Fig. 2 the daughters of the 207 ions found in one of the extracts

and those resulting from 2-naphthalenesulfonic acid standard are compared.

loo-

SO-

60-

40-

20- ZOL

171

IS 1, 194

. .!, 150 200 250

100

80

60

40

20

2

271 223

L&l

230 254

223

250 300 350 400 450

a

302

' 317 3,52 382 L 1 I. 1 42 t _

I * 9 " I '. . n I 300 350 400 450

b

392

I 3,SZ

367

1

-ig 1. FAWdual-scan tandem MS spectra of two river water samples.

2299

Although the masses of their daughter ions are the same, their relative abundance is significantly

different; it should be outlined that the two spectra were collected with minimum delay, therefore no

variation in the response of the mass spectrometer can explain such difference. Similar

considerations can be done when the daughters of 209 ions (Fig. 3) obtained from the river water

extract and standard 2-naphthalenesulfonic acid are examined. In this case the differences are more

evident, as ions having different masses are present; while the 209 fragment resulting from standard

2-naphthalenesulfonic acid is clearly due to “S isotope, the one found in the river water is partly or

totally due to a definitely different chemical compound. Therefore, it was concluded that the peaks at

m/z 207 and 209 are likely to be originated by a mixture of 2-naphthalenesutfonate ion and one or

more different singly charged sutfonated anions.

loo- 143

a 60-

40-

80 20- I

y4

i 50 100 150 200

loo-

b 60-

40-

20-

115 1'rz"P9 92 12 (1,%9 _ ),&6 lf U4 v7 177 v9

I * * * .I.. .., . . . .

50 ,

100 150 200

:ig. 2. Daughters of the 207 ions; a: in the extract of river water; b: 2-naphthalenesulfonic acid

2300

a

100' 209

60-

40-

*? 143 zo-

27 43 II

60 i' 11: L k' 9; 133 K,,

'i' Y" Pl LPI 1.. . * I. a. ' I * 3 s. I * ' 50 100 150 200

loo- 209

60-

40-

b

1?5 117

zo-

30, 4i 'I' i"' "Ip 'I" Lpi I L

\49 165 18; '9" I L

pl

I *. * .I ’ n.. 1. * '.I. * 50 100 150 200

:ig. 3. Daughters of the 209 ions; a: in the extract of river water; b: 2-naphthalenesulfonic acid

In order to ascertain if aromatic sulfonic acids could be oxidised under physico-chemical conditions

similar to those occurring in a river, a number of aromatic sulfonic acids were submitted to the

degradation test described in the Experimental section. The following twenty-two acids were tested;

those marked with an asterisk were found to undergo extensive chemical transformation within one

month: 1 naphthalenesulfonic”, 2-naphthalenesulfonic’, 1 ,!Snaphthalenedisulfonic, 2,6-naphthalene- disulfonic, 2,7-naphthalenedisulfonic. 2-amino-l -benzenesulfonic, 3-amino-l benzenesulfonic, 4- amino-l benzenesulfonic’, 2amino-1 -naphthalenesulfonic, 4-amine-1 -naphthalenesulfonic, 5-

amino-2-naphthalenesulfonic, 8-amino-2-naphthalenesulfonic 3-amino-2,7naphthalenedisutfonic,

2301

4-amino-1,3-naphthalenedisulfonic, 1 -hydroxy-2-naphthalenesulfonic*, 1 -hydrox+naphthalene-

sulfonic, l-hydroxy-3,6-naphthalenedisulfonic, 2-hydroxy3,6-naphthalenedisulfonic, 2-hydroxy-6,8-

naphthalenedisulfonic, 4-amino-3-hydroxy-1-naphthalenesulfonic’, 4amino-5-t iydroxy-l-

naphthalenesulfonic’, 8-amino-l -hydroxy-3,6+-raphthalenedisutfonic’

I I I I I I I

1.0 I 0 0 0

0.5 ’ 0.1

I 1 , 0 I 1 I 1 10 100 0 20 40 60 60

time (hours) time (min)

?g. 4; a: variation with time of the height of the chromatographic peak of 4-amine-3-hydroxy-1. iaphthalenesulfonic acid (0) and 8-amino-l -hydroxy-3,6-naphthalenedisulfonic acid (0) in the aboratory test solutions exposed to air and daylight; b: variation with time of the height of the zhromatographic peak of 4-amino-3-hydroxy-1 -naphthalenesulfonic acid under the followinS sxperimental conditions: (A): solution with 4.0 mgll initial 02 concentration, exposed to daylight; (C): solution with 4.0 mgA initial O2 concentration, maintained in the dark ; (0): solution deoxygenatec Nith nitrogen, maintained in the dark.

To give an idea of the order of magnitude of the time involved in these transformations, the

variation of the chromatographic peak height vs. time for two of the examined compounds are

reported in Fig. 4. As it can be seen, hours or days may be necessary to observe complete

degradation of the two compounds. Fig. 4b shows the influence of light and dissolved oxygen on the

disappearance of one of the investigated substances. The decrease of the HPLC peak was found to

be faster when solution contained oxygen and it was exposed to light, although a slower decrease

was observed even in the absence of light. The peak did not decrease if the solution contained no

oxygen and it was maintained in the dark. Owing to the experimental conditions adopted, microbial

degradation of the substances was excluded, and chemical oxidation, accelerated by light, seemed

the most probable cause of the disappearance of the substances.

The aqueous solutions of all the experimented substances were colourless immediately after

their preparation, but visibly some of them assumed progressively a dark colour. They were

extracted by means of the SPE procedure in order to perform further experiments. The UViVis

spectra of the solutions of the degraded compounds, coincided with those of SPE extract, while the

extracted solutions were colourless, thus indicating the completeness of the extraction.

2302

The SPE extracts of the degraded solutions were analysed by FAB/tandem MS with the dual-

scan mode. Two spectra, obtained by monitoring the neutral loss 64, are reported in Fig. 5. In the

spectrum of the degraded 8-amino-l -hydroxy-3,6naphthalenedisulfonic acid, its molecular peak at

318 is almost indistinguishable from the background, while several peaks at lower masses are much

more intense. As FAB produces no relevant fragmentation, it can be concluded that almost all of the

original substance has undergone chemical modification and that its degradation has involved

several reactions. Also the degradation of 4-amino-3hydroxy-1 -naphthalenesulfonic acid produced

100

80

60

40

20

164 I

100 -

80'

60'

40

20

1’7’ 150

a

! 2 t

269 I

b

236 I

266 3f4 2?3

276 I

‘ig. 5. Dual scan FABlMS spectra of the extracts of the degraded solutions of 8-amino-l-hydroxy-

,,6_naphtha\enedisulfonic

2303

several peaks, some of which were found at masses larger than 239, which -is the mass of the

original substance, thus indicating that polymerisation has occurred, to some extent, for this

compound.

Among the remaining compounds which underwent degradation, 1 hydroxy-2naphthalene

sulfonic acid showed a particularly interesting behaviour. In fact, two of the degradation products of

this acid appeared to be very similar to some of the compounds found in the river water sample

previously analysed. The dual-scan MS spectrum of the oxidation products, reported in Fig. 6, shows

two major peaks at 207 and 209, which, after further collision-induced fragmentation were found to

produce daughter ions very similar to those found in the river water extracts. HPLC analysis

confirmed the absence of both 1- and 2-naphthalenesulfonic acids among the degradation products.

TLC separation of 1 hydroxy-2-naphthalenesulfonic acid degradation products was effected,

and two brown spots were obtained. Their FTIR spectra are reported in Fig. 7, together with those of

their parent compound and of 2-naphthalenesulfonic acid. It can be noted that the bands of the

aromatic C-H stretching, in the spectra of the two naphthalenesutfonic acids (Fig. 7a and 7b), lie at

wavenumbers larger than 3ooO cm-f. On the contrary, both spectra of the brown spots (Fig. 7c and

7d) show C-H stretching at wavenumbers lower than 3000 cm-l .This could be due to disappearance

from the original molecule of the aromatic moiety, which could have transformed into an aliphatic or

an olefinic one. In addition, both spectra show intense and broad bands at 3406 cm-l, which indicate

the presence of O-H bonds. No band is observed at 1700 cm-f, thus the presence of carbonyl

moieties into these molecules can be excluded.

60

223

237

#A ~3 ~1 ;:... 39' 319 3tg 1””

i...., ‘1, ‘f9 ‘7

- . . . . . . . . . 250 300 350 400 4s

:ig. 6. Dual scan FABIMS spectra of the extract of the degraded solution of 1 -hydroxy-2-

raohthalenesulfonic acid

2304

b

:ig. 7. FTIR spectra. a: 1 -hydroxy-2-naphthalenesulfonic acid; b: sodium 2-naphthalenesulfonate; c:

legradation compound, TLC spot at Rf = 0.77; d: degradation compound, TLC spot at Rf = 0.87. The

lands at 2350 cm-f are due to atmospheric carbon dioxide.

CONCLUSIONS

The results of previous HPLC analyses, which indicated the presence of aromatic sulfonic

acids in the water of the river Bormida, were confirmed by means of FABIMS. In addition, FABlMS

evidenced that several sulfonated compounds other than amino and/or hydroxy substituted benzene,

naphthalene or anthraquinone sulfonates are present in the river. In particular, two sulfonated compounds of 207 and 209 molecular mass, different from 1- or 2-naphthalenesulfonic acids were

found. Some aromatic sulfonic acids were observed to undergo oxidative degradation under

laboratory conditions simulating those occurring in a river, thus originating multiple degradation

products. In particular, the FABiMS analysis of the degradation products of I-hydroxy-2-

naphthalenesulfonic acid showed two anions at 207 and 209 m/z, having characteristics similar to

those encountered in real environmental samples. These darkcoloured products are non-aromatic, hydroxy-substituted sulfonated hydrocarbons. If they are originated into the river, they could contribute to the progressive browning of the water which has been observed.

2305

REFERENCES

2

3

8

9

IO 11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

F.T. Lange, H. J. Brauch; Symposium on Polar Organic Pollutants in the Environment, Basef,

1993

E.R. Brouwer, J. Slobodnik, H. Lingeman, U.A.Th. Brinkman; Anafusis, 20:3 (1992) 121

B. Altenbach, W. Giger; Symposium on Polar Organic Pollutants in the Environment, Base/,

1993

0. Zerbinati, G. Ostacoli, D. Gastaldi, V. Zelano; J. Chromatogr., 640 (1993) 231

0. Zerbinati, G. Ostacoli, J. Chromatogr A, 671 (1994) 217-23

E.R. Brower, TM. Tot, H. Lingeman, U.A. Th. Brinkman; Qufm. Analftica, 12 (1993) 88-95

H. Miyoshi, T. Nagai, M. Ishikawa, Bull. Shizuoka Pref Inst. Pub/. Ht. and Environ. sd.; 27, 45

(1984) S. Schullerer, H.-J. Brauch, F.H. Frimmel; Vom Wasser 75 (1990) 83

Association of River Rhine Waterworks inc. (ARNW); 50. Report, ISSN 0343-0391 (1993) p.

69

S. Schullerer, G. Koschenz, H.-J. Brauch, F.H. Frimmel; Vom Wasser 78 (1992) 229

S. Schullerer, FD.H. Frimmel; Anal. Chim. Acta, 283 (1993) 251

F. Th. Lange, M. Wenz, H.-J. Brauch; J. High Resol. Chromatogr. 18 (1995) 243

0. Zerbinati, S. Salomone, G. Ostacoli; Chemosphere, 29 (1994) 2639-43

M.A. Hashim, J. Kulandai, R.S. Hassan; J. Chem.Tech. Biotechnol., 1992,54,207-14

Greim H, J. Ahlers, R. Bias, B. Broecker, H. Hollander, H.-P Gelbke, H.-J. Klimsch, I.

Mangelsdorf, A. Paetz, N. Sccin, G. Stropp, R. Vogel, C. Weber, K. Ziegler-Skylakakis, E.

Bayer; Chemospere 28, 1994,2203-36

H. Grossenbacher, T. Thurnheer, D. Zurrer, A.D. Cook, J. Chromatogr. 523 ( 1990) 247

Diekmann R., Hempel D. C.; Appl. Microbial. Biotechnol. 1989, 32, 113-l 7

Diekmann R., Naujoks M., Gerdes-Kuehn M., Hempel D.C.; Bioprocess. Eng. 1990,5, 13-17

Kerimov T.M., Ismailov, N.M., Izv. Akad. Nauk AZ. SSR, Ser. Biol. Nauk 1988, (2) 99-102

Ohe Tatsuhiko, Watanabe Yasuto, Agric. Biol. Chem. 1988, 52, 2409-l 4

Diekmann R., Noertemann B., Hempel D.C., Knackmuss H.J.; Appf. Microbial. Biotecmtol.

1988, 29(l), 85-8

Luther M.; Ber. Kernforshungsanlage Juelich 1988, Juel-2236, 211 pp.

Wittich R.M., Rast H.G., Knackmuss H.J.; Appl. Environ. Microbial. 1988, 54,1842-7

Menscke J., Bennemann H., Herbst H., Doermier S., Hempel. D.C.; Bioprocess Eng. 1988,

3(4), 151-7

F. Ventura, A. Figueras, J. Caixach, D. Fraisse, J. Rivera; Fresenius’Z. Anal. Chem., 1989

335(3), 272-8