Analytical and Bioanalytical Chemistry

, Volume 399, Issue 10, pp 3579–3588

Transformations of polycyclic musks AHTN and HHCB upon disinfection with hypochlorite: two new chlorinated disinfection by-products (CDBP) of AHTN and a possible source for HHCB-lactone

Authors

    • BAM Federal Institute for Materials Research and Testing
  • Robert Göstl
    • Department of ChemistryHumboldt-University Berlin
  • Philip Teichert
    • BAM Federal Institute for Materials Research and Testing
  • Christian Piechotta
    • BAM Federal Institute for Materials Research and Testing
  • Irene Nehls
    • BAM Federal Institute for Materials Research and Testing
Original Paper

DOI: 10.1007/s00216-011-4674-3

Cite this article as:
Kuhlich, P., Göstl, R., Teichert, P. et al. Anal Bioanal Chem (2011) 399: 3579. doi:10.1007/s00216-011-4674-3

Abstract

In this work, the behavior of the polycyclic musks 6-acetyl-1,1,2,4,4,7-hexamethyltetraline (AHTN) and 1,3,4,6,7,8-hexahydro-4,6,6,7,8,8-hexamethylcyclopenta-γ-2-benzopyran (HHCB) was investigated upon disinfection by using sodium hypochlorite as disinfectant in a model disinfection basin in order to find new disinfection by-products (DBP). In the case of AHTN, the carboxylic acid 3,5,5,6,8,8-hexamethyl-5,6,7,8-tetrahydronaphthalene-2-carboxylic acid (AHTN-COOH) was generated by a haloform reaction, being the origin for two new chlorinated DBPs. In the case of HHCB, disinfection via hypochlorite led to the HHCB-lactone. All reaction products and intermediates were synthesized and isolated. The relevant degradation mechanisms are discussed in detail.

Keywords

MuskAHTNHHCBHHCB-lactoneChlorinationDisinfection

Introduction

Nitro musks and polycyclic musks are synthesized in order to emulate the scent of natural musk, obtained from male musk deer in the past. However, these animals are protected from extinction since 1979 [1]. Today, the polycyclic musks 6-acetyl-1,1,2,4,4,7-hexamethyltetraline (AHTN) and 1,3,4,6,7,8-hexahydro-4,6,6,7,8,8-hexamethylcyclopenta-γ-2-benzopyran (HHCB) are widely used fragrances in cosmetics, products of daily care, and cleaning products for households and industry. Since 1990, both musks were increasingly used to replace nitro musks, which are costly [2], may trigger photoallergic reactions, and transform into ecotoxic aniline derivatives [3].

In the last decade, both musks are produced in the ton-scale, e.g., in 2000, approximately 1,427 metric tons (t) and 343 t of HHCB and AHTN, respectively [4]. They are introduced into the environment mainly by sewage treatment plants supplied by municipal wastewater. HHCB and AHTN were shown to occur in sewage sludge at low milligrams per kilogram concentration (dry weight), in surface waters at low micrograms per liter concentrations, in swimming pool water at mid-nanograms per liter concentrations, and could also be detected in fatty tissues of fish and human fat samples [510].

Recent examinations towards the estrogenic potential testified low estrogenic activities for both musks. However, it must be stated that different assays used led to ambiguous data, e.g., Bitch et al. using a human MCF-7 assay stated that AHTN has a low estrogenic activity, whereas HHCB was shown to be estrogenically inactive in vitro [11]. On the other hand, Yamauchi et al. showed the existence of vitellogenin (VTG) in the liver of male medaka in vivo upon exposure of HHCB proofing its estrogenic activity [12]. These facts might induce health concerns and require further investigations.

The guideline for drinking-water quality from the World Health Organization (WHO) mentions chlorination as a first and cheap step to disinfect water from microorganisms in order to extract water of high quality [13]. Therefore, chlorine is added to the water as gas, hypochlorite solution, or in the form of hypochlorite-generating precursor. Also, water for swimming pools is disinfected by the use of hypochlorite tablets. Besides the desired disinfection, the high reactivity of chlorine can lead to the formation of disinfection by-products (DBP) from organic compounds [14]. First identified DBPs in drinking water were trihalomethanes (THM) [15]. For the UV agents benzophenone and oxybenzone, as well as for the monomer bisphenol A (BPA) from plastics, the formation of chlorinated disinfection by-products (CDBP) could be observed in model experiments [1618]. However, to our knowledge, the behavior of HHCB and AHTN under disinfection conditions has not been investigated yet, and little is known about their possible transformation products. In order to achieve a comprehensive risk assessment for a given environmental pollutant, the identification of possible degradation pathways and products is highly desirable, as the latter might show toxicological effects different from the parent compound.

Hence, we applied a model simulating disinfection to HHCB and AHTN aiming at the identification of novel CDBP and/or transformation products.

Materials and methods

Chemicals

All chemicals were purchased from Sigma-Aldrich (Taufkirchen, Germany). All solvents were of HPLC grade; all standard chemicals were of analytical-reagent grade. The sodium hypochlorite solution contained 10–15% available chlorine (as labeled). AHTN was purchased as racemate (containing two isomers). HHCB was purchased as technical product (containing four isomers) stabilized with 50% diethyl phthalate (DEP). Before usage, HHCB was purified and separated from DEP by column chromatography (silica gel; cyclohexane/ethyl acetate, 4:1 (v)).

Instruments

High-performance liquid chromatography diode array detector mass spectrometry (HPLC/DAD/MS) measurements were carried out on an Agilent 1100 HPLC series coupled to an extended capacity trap (XCT) mass spectrometer equipped with an electrospray interface (ESI) ion source (Agilent Technologies, Waldbronn, Germany). Separation was done using a Wicom Prontosil ACE-EPS column (300 × 4.6 mm, 3-μm particle size; Heppenheim, Germany).

Gas chromatography mass spectrometry (GC/MS) measurements were performed using a Varian CP3800 gas chromatograph coupled to a Varian 1200L quadrupole mass spectrometer (Agilent Technologies, Waldbronn, Germany) with a CB-PAH column from J & K Scientific (12 m, 0.25 mm ID, 0.15-μm film).

Semi-preparative HPLC was done on an Agilent 1100 HPLC series coupled to the fraction collector Foxy R1 (Teledyne Isco, Lincoln, NE, USA). The preparative column was purchased from SepServ (Berlin, Germany): 250 × 20 mm, 10-μm, RP18.

The single crystal X-ray data collection was carried out on a Bruker AXS SMART diffractometer.

1H and 13C NMR measurements were carried out on a Bruker AVANCE III 500 (Bruker Daltonik, Bremen, Germany).

Methods

GC/MS and HPLC/DAD/MS setup

All parameters are given in Tables 1 and 2.
Table 1

Instrument parameters for GC/MS

Parameter

Setting or value

Solvent

Hexane

Oven program

Initial, 60 °C; held for 3 min

25.5 min, 285 °C; held for 10 min

Injection volume

1 μL

Injector temperature

250 °C

Split

Initial, on; ratio 10

0.01 min, off

1 min, on; ratio 50

Carrier gas and flow rate

Helium; 1 mL/min (constant)

Ionization

Electron impact ionization (EI); 70 eV

Source temperature

230 °C

Filament current

50 μA

Transfer line temperature

230 °C

Table 2

Instrument parameters for HPLC/DAD/MS

Parameter

Setting or value

Injection solvent

Identical to initial LC eluent composition

Solvents

A, water

B, methanol both modified with 0.1% (v) formic acid

Solvent gradient

Initial, A/B 15:85 (v/v)

20 min, A/B 0:100 (v/v)

30 min, A/B 0:100 (v/v)

Oven temperature

30 °C

Flow rate

0.7 mL/min

Injection volume

10 μL; if not mentioned differently

DAD

210 nm (bandwidth, 4 nm), 254 nm (bandwidth, 16 nm), 310 nm (bandwidth, 8 nm); reference wavelengths were 360 nm with a bandwidth of 100 nm for all signals

Nebulizer

60 psi

Dry gas (N2)

11 L/min

Drying temperature

350 °C

Mode

Ultra scan

Maximum accumulation time

200 ms

Averaging

5 scans

Trap drive

100%

Compound stability

100%

Scan range

250 to 350 amu

ICC

200,000

Ion source

ESI ion source; operating in positive and negative mode (alternating)

Semi-preparative HPLC

The separation was done isocratic with a flow of 4 mL/min at room temperature. Please see “HPLC/DAD/MS setup” for eluents. The respective eluent composition is given in the synthetic part below.

X-ray crystallography

Measurements were carried out at room temperature using Mo Kα radiation (λ = 0.71073 Å), monochromatized by a graphite crystal. Intensities were measured with an exposure time of 60 s per frame. The data reduction was performed by using the Bruker AXS SAINT and SADABS packages. The structures were solved by direct methods and refined by full-matrix least squares calculation using SHELX. Anisotropic thermal parameters were employed for non-hydrogen atoms. The hydrogen atoms were treated isotropically with Uiso = 1.2 times the Ueq value of the parent atom. In case of methylene groups, Uiso = 1.5 times the Ueq value was chosen. See synthesis part for respective CCDC identifiers, containing the supplementary crystallographic data for this paper. These data can be obtained free of charge from the Cambridge Crystallographic Data Centre via http://www.ccdc.cam.ac.uk/data_request/cif.

NMR

The solvents and internal references are given for each spectrum. The signals are characterized as follows: s for singlet, d for doublet, dd for double doublet, t for triplet, and m for multiplet. The chemical shifts are given in parts per million, relative to tetramethylsilane (TMS). The coupling constants (J) are given in Hertz. Data handling was carried out with TOPSPIN 2.1.

Experiments and syntheses

An overview of all compounds with their numbering, molecular structures, molecular weights, and abbreviation used is given in Table 3.
Table 3

Overview of all compounds occurring in this work

https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Tab3_HTML.gif

Pre-experiments on the disinfection of AHTN 1 and HHCB 9 in a model disinfection basin

In two 1.5-mL HPLC vials, 300 μL of a solution of 1 and 9, respectively, in methanol (c = 1 mg/mL) were mixed with 250 μL water and 250 μL of NaOCl solution. The mixtures were shaken for 10 min and left to stand for 72 h. Afterwards, 800 μL of ethyl acetate was added, and the mixtures were shaken again for 10 min. After removing the aqueous layers, the organic solvents were dried over anhydrous sodium sulfate and evaporated to dryness. The residues were then reconstituted in the solvent required for the respective measurement.

Synthesis of 2-chloro-1-(3,5,5,6,8,8-hexamethyl-5,6,7,8-tetrahydronaphthalen-2-yl)ethan-1-one 2 (AHTN-Halo-I) and 2,2-dichloro-1-(3,5,5,6,8,8-hexamethyl-5,6,7,8-tetrahydronaphthalen-2-yl)ethan-1-one 3 (AHTN-Halo-II)

A suspension of 1 (5.17 g, 20 mmol) in NaOCl solution (300 mL) was acidified to pH 1 with 6 M hydrochloric acid and stirred for 3 h. Afterwards, sodium sulfite (20 g) was added to quench free chlorine. The mixture was extracted with dichloromethane (3 × 50 mL). The extracts were combined, washed with brine, dried over anhydrous sodium sulfate, and filtered. After evaporation of the solvent under vacuum, the residue was cleaned by column chromatography (silica gel; dichloromethane) resulting in two fractions: the one eluating first corresponded to the product AHTN-Halo-II 3 (yellowish oil; yield, 500 mg, 6%), the second to the product AHTN-Halo-I 2 (yellowish oil; yield, 800 mg, 12%). Both oils gave yellowish crystals after several hours at 4 °C.

For crystal structure of AHTN-Halo-II 3, see Fig. 8 or for further information, see CCDC-784582, respectively.

1H-NMR of AHTN-Halo-I 2 (500 MHz, CD3OD, TMS)

δ/ppm = 7.71 (1H, s), 7.28 (1H, s), 4.80 (2H, s), 2.43 (3H, s), 1.88 (1H, m), 1.64 (1H, dd, 2J = 13.6 Hz, 3J = 13.2 Hz), 1.42 (1H, dd, 2J = 13.6 Hz, 3J = 2.7 Hz), 1.32 (3H, s), 1.32 (3H, s), 1.27 (3H, s), 1.07 (3H, s), and 1.00 (3H, d, J = 6.9 Hz).

13C-NMR of AHTN-Halo-I 2 (125 MHz, CD3OD, TMS)

δ/ppm = 196.5, 152.3, 144.0, 137.0, 133.4, 132.0, 132.0, 45.7, 44.7, 39.1, 35.8, 35.1, 32.6, 32.3, 28.8, 25.0, 21.4, and 17.3.

1H-NMR of AHTN-Halo-II 3 (500 MHz, CD3OD, TMS)

δ/ppm = 7.81 (1H, s), 7.35 (1H, s), 7.33 (1H, s), 2.44 (3H, s), 1.89 (1H, m), 1.65 (1H, dd, 2J = 13.6 Hz, 3J = 13.2 Hz), 1.42 (1H, dd, 2J = 13.6 Hz, 3J = 2.6 Hz), 1.33 (3H, s), 1.33 (3H, s), 1.27 (3H, s), 1.08 (3H, s), and 1.01 (3H, d, J = 6.9 Hz).

13C-NMR of AHTN-Halo-II 3 (125 MHz, CD3OD, TMS)

δ/ppm = 189.7, 153.1, 144.1, 138.5, 132.3, 131.1, 129.2, 71.1, 44.7, 39.1, 35.8, 35.3, 32.6, 32.3, 28.8, 25, 21.3, and 17.2.

Synthesis of 3,5,5,6,8,8-hexamethyl-5,6,7,8-tetrahydronaphthalene-2-carboxylic acid 5 (AHTN-COOH)

Hypochlorite solution (150 mL) was added to a solution of 1 (5.17 g, 20 mmol) in acetonitrile (150 mL). The mixture was stirred vigorously for 72 h at room temperature. Afterwards, water was added to dissolve precipitated salt, sodium sulfite (10 g) to quench free chlorine, and finally, 6 M hydrochloric acid to adjust the pH to 1. The organic compound was extracted with diethyl ether (3 × 100 mL). The extracts were combined, dried over anhydrous sodium sulfate, and filtered. Evaporation of the solvent under vacuum gave a white crystalline residue that was washed with cyclohexane. Recrystallization from diethyl ether resulted in colorless crystals (yield, 2.73 g; 52%). The crystal structure has been reported previously by our group [19].

1H-NMR (500 MHz, CD3OD, TMS)

δ/ppm = 7.88 (1H, s), 7.24 (1H, s), 2.52 (3H, s), 1.89 (1H, m), 1.64 (1H, dd, 2J = 13.5 Hz, 3J = 13.3 Hz), 1.41 (1H, dd, 2J = 13.5 Hz, 3J = 2.6 Hz), 1.33 (3H, s), 1.30 (3H, s), 1.25 (3H, s), 1.07 (3H, s), and 1.01 (3H, d, J = 6.9 Hz).

13C-NMR (125 MHz, CD3OD, TMS)

δ/ppm = 171.5, 151.6, 143.4, 137.8, 131.3, 130.5, 128.5, 44.7, 38.9, 35.8, 35.0, 32.8, 32.4, 28.9, 25.1, 21.9, and 17.3.

Synthesis of 4-chloro-3,5,5,6,8,8-hexamethyl-5,6,7,8-tetrahydronaphthalene-2-carboxylic acid 6 (AHTN-Cl-COOH)

This compound was prepared by a modification of a literature procedure [20]. Sodium periodate (102 mg, 0.5 mmol) was slowly added to a mixture of 5 (500 mg, 1.92 mmol) in acetonitrile/water (6 mL, 2:1 (v/v)), sodium chloride (135 mg, 2.3 mmol), and sulfuric acid (0.1 mL, 30%). The mixture was then heated to 80 °C for 24 h. After cooling to room temperature, a saturated solution of sodium sulfite (50 mL) was added, and the pH was adjusted to 1 with 6 M hydrochloric acid. The mixture was extracted with diethyl ether (3 × 20 mL). The extracts were combined, dried over anhydrous sodium sulfate, and filtered. Afterwards, the solvent was evaporated to dryness, and the residue was reconstituted in methanol. The product was purified by semi-preparative column chromatography (eluent: 5:95 A/B (v/v)). The methanol of the resulting fraction was removed in vacuo, and the residue was extracted with dichloromethane (3 × 50 mL). Finally, the extracts were combined, dried over anhydrous sodium sulfate, and filtered. Evaporation of the solvent under vacuum yielded a white solid.

For crystal structure of AHTN-Cl-COOH 6, see Fig. 9 or for further information, see CCDC-784581, respectively.

1H-NMR (500 MHz, CD3OD, TMS)

δ/ppm = 7.73 (1H, s), 2.54 (3H, s), 1.88 (1H, m), 1.64 (1H, dd, 2J = 13.5 Hz, 3J = 13.4 Hz), 1.59 (3H, s), 1.37 (3H, s), 1.37 (1H, dd, 2J = 13.5 Hz, 3J = 2.2 Hz), 1.33 (3H, s), 1.25 (3H, s), and 1.04 (3H, d, J = 6.8 Hz).

13C-NMR (125 MHz, CD3OD, TMS)

δ/ppm = 171.4, 148.0, 146.7, 137.2, 136.4, 132.2, 128.7, 43.7, 40.7, 38.4, 36.2, 33.4, 32.2, 27.2, 18.8, 18.6, and 17.4.

Synthesis of 6-chloro-1,1,2,4,4,7-hexamethyl-1,2,34-tetrahydronaphthalene 7 (AHTN-Cl-Kochi)

Synthesis is identical to the one of 5, with the pH adjustment step being replaced by washing with brine and subsequent isolation of the acetonitrile layer. The acetonitrile layer was then dried over anhydrous sodium sulfate, filtered, and evaporated to dryness. The residue was taken up in methanol, and the product was purified by semi-preparative column chromatography (eluent: B, 100%). The methanol of the relevant product fraction was removed in vacuo, the residue taken up in methanol again, dried over anhydrous sodium sulfate, and filtered. Evaporation of the solvent under vacuum gave a yellow solid.

For crystal structure of AHTN-Cl-Kochi 7, see Fig. 10 or for further information, see CCDC-784580, respectively.

1H-NMR (500 MHz, CD3OD, TMS)

δ/ppm = 7.24 (1H, s), 7.21 (1H, s), 2.29 (3H, s), 1.85 (1H, m), 1.62 (1H, dd, 2J = 13.5 Hz, 3J = 13.3 Hz), 1.37 (1H, dd, 2J = 13.5 Hz, 3J = 2.6 Hz), 1.30 (3H, s), 1.25 (3H, s), 1.22 (3H, s), 1.04 (3H, s), and 0.99 (3H, d, J = 6.8 Hz).

13C-NMR (125 MHz, CD3OD, TMS)

δ/ppm = 145.9, 145.5, 133.8, 132.4, 130.7, 127.6, 44.6, 38.5, 35.8, 35.0, 32.7, 32.4, 29.1, 25.2, 19.8, and 17.2.

One-step synthesis of methyl 3,5,5,6,8,8-hexamethyl-5,6,7,8-tetrahydronaphthalene-2-carboxylate 8 (AHTN-COOMe)

Hypochlorite solution (150 mL) was added slowly to a cooled solution of 1 (5.17 g, 20 mmol) in methanol (150 mL). The mixture was stirred vigorously for 4 h at room temperature. Afterwards, water was added to dissolve precipitated salt. The mixture was then extracted with ethyl acetate (3 × 200 mL). The extracts were combined, dried over anhydrous sodium sulfate, and filtered. Evaporation of the solvent under vacuum gave a white solid residue that was suspended in ethyl acetate (50 mL) and filtered again. Upon evaporation to dryness in vacuo, colorless crystals were obtained (yield, 3.43 g; 66%).

1H-NMR (500 MHz, CD3OD, TMS)

δ/ppm = 7.83 (1H, s), 7.23 (1H, s), 3.83 (3H, s), 2.48 (3H, s), 1.85 (1H, m), 1.61 (1H, CH2, dd, 2J = 13.5 Hz, 3J = 13.6 Hz), 1.38 (1H, CH2, dd, 2J = 13.5 Hz, 3J = 2.6 Hz), 1.30 (3H, s), 1.26 (3H, s), 1.22 (3H, s), 1.04 (3H, s), and 0.98 (3H, d, J = 6.8 Hz).

13C-NMR (125 MHz, CD3OD, TMS)

δ/ppm = 169.7, 151.8, 143.5, 137.8, 131.8, 130.2, 127.8, 52.1, 44.6, 38.9, 35.7, 34.9, 32.6, 32.4, 28.9, 25.1, 21.7, and 17.1.

Synthesis of 4,6,6,7,8,8-hexamethyl-1H,3H,4H,6H,7H,8H-indeno[5,6-c]pyran-1-one 10 (HHCB-lactone)

Sodium hypochlorite solution (50 mL) was added to a solution of HHCB 9 (1.04 g, 4 mmol) in acetonitrile (100 mL). The mixture was stirred vigorously for 5 h at room temperature. Afterwards, water was added to dissolve precipitated salt. The mixture was then extracted with ethyl acetate (3 × 50 mL). The extracts were combined, dried over anhydrous sodium sulfate, and filtered. After evaporation of the solvent under vacuum, the residue was cleaned by column chromatography (silica gel; cyclohexane/ethyl acetate, 4:1 (v)), and colorless crystals were obtained upon evaporation to dryness (yield, 0.49 g; 45%).

1H-NMR (500 MHz, CD3OD, TMS)

δ/ppm = 7.81 (1H, s), 7.21 (1H, s), 4.48 (1H, m, CH2-a), 4.25 (1H, m, CH2-b), 3.15 (1H, m), 1.87 (1H, m), 1.34 (3H, d, J = 13.6 Hz,), 1.32 (3H, d, J = 5.3 Hz), 1.31 (3H, d, J = 5.0 Hz), 1.11 (3H, d, J = 8.3 Hz), 1.091 (3H, d, J = 6.7 Hz), and 1.04 (3H, d, J = 7.4 Hz).

13C-NMR (125 MHz, CD3OD, TMS)

δ/ppm = 168.1, 160.0, 152.2, 145.7, 125.6, 123.9, 121.6, 74.0, 55.6, 46.4, 45.6, 33.3, 29.6, 29.1, 26.3, 26.0, 17.6, and 8.89.

Results and discussion

Identification of novel transformation products of AHTN obtained in a model disinfection basin

The treatment of a solution of AHTN in methanol with NaOCl led to four identifiable transformation products (58), all of which were observable by HPLC/DAD/MS-Trap and GC/MS (see Fig. 1, “AHTN NaOCl Mix” for HPLC/DAD chromatogram).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig1_HTML.gif
Fig. 1

LC/DAD chromatograms of AHTN NaOCl mix, AHTN 1 itself, and all synthesized compounds (DAD wavelength, 210 nm). The mix is joined out of two chromatograms with varying injection volumes for the sake of clarity

Initial hints on the identity of the transformation products could be obtained from the GC/MS spectra (see Table 4). The mass shift of +2 amu of compound 5 is consistent with a haloform reaction occurring at the acetyl side chain of AHTN [21]. Compounds 6 and 7 feature the characteristic chlorine isotope pattern in their mass spectra (35Cl:37Cl, ratio 3:1, Fig. 2). The mass shift of compound 6 is consistent with the chlorination of the haloform product 5 (mass shift of +34 amu). The mass of the second chlorinated compound 7 is 44 amu lower compared to 6, indicating decarboxylation. The remaining compound 8 shows a mass shift of +14 amu compared to 5, indicating methylation.
Table 4

Attributions of observed mass shifts in GC/MS

Compound

Mass/Da (intensity)

Mass shift/Da compared to 1

Attribution of mass shifts

1

257 [M − H]

5

259 [M − H]

+2

1 − CH3 + OH

6

293 [M − H]; 295 [M − H + 2]; (3:1)

+36

1 − CH3 + OH − H + 35Cl

7

250 [M]; 252 [M + 2]; (3:1)

−8

1 − acetyl + 35Cl

8

273 [M − H]

+16

1 + [O]

https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig2_HTML.gif
Fig. 2

GC/MS mass spectra of compounds 6 ([M − H] = 293.2 Da) and 7 ([M] = 250.2 Da)

Based on the considerations above, degradation mechanisms based on Haloform and Kochi reactions were derived (Figs. 3 and 4) [21, 22]. For structural confirmation, the presumed transformation products were synthesized and characterized by NMR and X-ray crystallography in case of 3, 6, and 7, thus also confirming the regiochemistry of chlorination of 6 and 7.
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig3_HTML.gif
Fig. 3

Haloform reaction of AHTN 1 to AHTN-COOH 5 and subsequent chlorination [21]

https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig4_HTML.gif
Fig. 4

Kochi-like reaction of AHTN-COOH 5 to AHTN-Cl-Kochi 7; Kochi reaction is a variation of Hunsdiecker reaction on which the mechanism shown is based [22, 33]

Comparison of the HPLC/DAD and GC/MS retention times (see Fig. 1 for all HPLC/DAD chromatograms) as well as the associated MS fragmentation patterns and NMR data confirmed the mechanisms shown.

Products 5 and 8 were previously known [23]; however, compounds 6 and 7 were not yet isolated. Lately, our group reported the crystal structure of 5 [19].

Identification of transformation product of HHCB obtained in a model disinfection basin

The treatment of a solution of HHCB in methanol with NaOCl led to only one major transformation product 10, observable by HPLC/DAD/MS-Trap and GC/MS (see Fig. 5, “HHCB NaOCl Mix” for HPLC/DAD chromatogram).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig5_HTML.gif
Fig. 5

LC/DAD chromatograms of HHCB NaOCl mix, HHCB 9 itself, and the synthesized lactone 10 (DAD wavelength, 254 nm)

The mass shift of +14 amu in comparison to 9 indicated the formation of a lactone 10 (Fig. 6). For confirmation, the lactone was synthesized and identified by NMR. Comparison of the HPLC retention times (see Fig. 5 for all chromatograms) as well as the associated GC/MS mass-spectra and NMR data confirmed the formation of the lactone.
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig6_HTML.gif
Fig. 6

GC/MS mass spectra of compounds 9 ([M − 15] = 243.1 Da) and 10 ([M − 15] = 257.1 Da)

Interpretation of the formation mechanisms of AHTN transformation products

In aqueous solution, NaOCl is hydrolyzed to free chlorine and sodium hydroxide. These reagents enable the Haloform reaction, which results in the carboxylic acid 5 of AHTN 1 (Fig. 3) [21].

The two intermediates of the haloform reaction, AHTN-Halo-I 2 and AHTN-Halo-II 3, could not be isolated directly from the model disinfection basin, but were obtained through different routes, i.e., chlorination of AHTN with sodium periodate and sodium chloride or chlorination with N-chlorosuccinimide (NCS) and p-toluenesulphonic acid [20, 24]. These approaches yielded compound 2 directly, while compound 3 was obtained by acidifying the model disinfection basin to pH 1.

The absent direct chlorination of AHTN 1 at the aromatic ring may be explained by the deactivating effect of the acetyl side chain. In the haloform product 5, the acetyl side chain is replaced by a carboxylic acid moiety, which activates the only available meta-position of the benzene ring with respect to electrophilic chlorination. Thus, chlorination can only occur subsequently to the haloform reaction resulting in the chlorinated carboxylic acid AHTN-Cl-COOH 6 (Fig. 3).

Product 7 is formed by a Kochi-like reaction [22], in which the function of the typical Kochi oxidant lead(IV) acetate [Pb(OAc)4] is taken by NaOCl. Judging from the position of the chlorine atom in 7, a decarboxylation of 6 to form 7 can be ruled out. The expected mechanism is shown in Fig. 4.

The third compound, AHTN-COOMe 8, is the esterification product of the acid 5. This product is accounted to taking up AHTN 1 in methanol in the model disinfection basin. Thus, CDBPs 6 and 7 can also be obtained, if AHTN 1 is taken up in acetonitrile.

Summarized, the carboxylic acid 5 is the precursor of all other transformation products of AHTN 1.

Interpretation of the formation of HHCB-lactone

NaOCl as a strong oxidant is capable to oxidize HHCB to its lactone. The HHCB-lactone was first identified by Franke et al. in surface water [25]. The group synthesized HHCB-lactone as standard compound themselves by using potassium permanganate as oxidizing agent. By searching the literature, lactones, originating from cyclic ethers, are nearly always obtained by using oxidizing agents such as sodium bromate, lithium hypochlorite, or even sodium hypochlorite itself [2628]. Thus, we showed the formation of HHCB-lactone via sodium hypochlorite in our disinfection model basin. A possible reaction mechanism for the oxidation of the cyclic ether tetrahydrofurane to its lactone is given by Metsger et al. [29], using bromate as oxidizing agent. A biological pathway for the formation of HHCB-lactone is given by Martin et al. [30], who investigated the influence of fungi towards HHCB and AHTN. Recent studies reported that the concentration of HHCB-lactone even rises within wastewater plants [9, 31, 32], indicating a biological degradation. However, anthropogenic causes like the disinfection with sodium hypochlorite have not been taken into account so far and might be a further source for the transformation of HHCB to HHCB-lactone (Fig. 7).
https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig7_HTML.gif
Fig. 7

Transformation of HHCB 9 by NaOCl; monoisotopic masses (Mmi) are given

https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig8_HTML.jpg
Fig. 8

Crystal structure of AHTN-Halo-II 3; CCDC 784582

https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig9_HTML.jpg
Fig. 9

Crystal structure of AHTN-Cl-COOH 6; CCDC 784581

https://static-content.springer.com/image/art%3A10.1007%2Fs00216-011-4674-3/MediaObjects/216_2011_4674_Fig10_HTML.jpg
Fig. 10

Crystal structure of AHTN-Kochi 7; CCDC 784580

Conclusions

In this article, we have shown two new CDBPs of AHTN: a mono-chlorinated carboxylic acid AHTN-Cl-COOH 6 and a decarboxylated Kochi-like compound AHTN-Cl-Kochi 7. Furthermore, we showed that the disinfection of HHCB via hypochlorite generating precursors led to the transformation product HHCB-lactone 10. Methods for the convenient syntheses of 5, 6, 7, 8, and 10 and the reaction intermediates 2 and 3 were given.

We are aware of the fact that the concentrations of free chlorine and musk used in this pilot study are much higher than expected in swimming pools or plants extracting drinking water. However, at this stage, we are not yet capable of measuring such low concentrations of musk in swimming pool water reported by Regueiro et al. [10] without establishing a solid-phase extraction method. Therefore, to examine and confirm the environmental occurrence of the novel CDBPs and also the known transformation product 10, we are currently developing analytical methods capable of detecting 6, 7, and 10 in water treated with hypochlorite generating chemicals, e.g., swimming pool water. Furthermore, we want to state that this disinfection might be another source of HHCB-lactone in the environment besides biological metabolization.

Acknowledgements

The authors wish to thank Prof. Dr. Mügge and Ms. Thiesies (Humboldt University, Berlin, Department of Chemistry) for measuring NMR samples. David Siegel, Stefan Merkel, Sebastian Schmidt, and Robert Rothe (Federal Institute for Materials Research and Testing, Berlin, Germany) are thanked for valuable discussion and technical support. Dr. Franziska Emmerling and Werner Kraus (Federal Institute for Materials Research and Testing, Berlin, Germany) are acknowledged for measuring and obtaining X-ray crystal structures.

Copyright information

© Springer-Verlag 2011