sign in sign up
<<
Home , Carbonate

Environ. Sci. Technol. 2007, 41, 7802-7806

Because of their possible emission into the atmosphere, Suboxide, a Highly Reactive the natural formation of volatile compounds has been the Intermediate from the Abiotic subject of numerous publications over the past years (6-9). Biotic processes through mammals, insects, fungi, plants, Degradation of Aromatic Compounds and bacteria leading to methane, chloromethane, and are well-known. However, abiotic processes have in been attracting more and more attention. Many investigations were performed discussing the degradation of soil organic STEFAN G. HUBER,* matter (SOM) through different redox processes. As the GERHARD KILIAN, AND structure of SOM is very complex, it is common to use model HEINZ F. SCHO¨ LER compounds, such as substituted phenols or catechols, which represent fragments of the humic substances in soil (7). Institute of Environmental , University of Heidelberg, Im Neuenheimer Feld 236, 69120 Heidelberg, Hydroxyl radicals play an important part in these redox Germany processes because they readily react with organic molecules initializing subsequent oxidation steps. Their release from the reaction of redox-active metals such as or manganese with members of the family (hydrogen peroxide, The formation of volatile compounds during abiotic superoxide, molecular oxygen, etc.) is explained by the Fenton degradation processes of aromatic compounds in soil has (10) and Fenton-like or Haber-Weiss (11) reactions. Small been the subject of many experimental studies but molecules such as chloroethyne, chloroethene, and carbon should be examined further. In this context, the present dioxide are formed during the degradation of catechol in the presence of Fe(III) and chloride (12-15). For their formation, work investigates the natural formation of cleavage of the aromatic ring is an important step that is using the model compounds catechol and 3,5-dichloro- followed by further redox reactions. catechol and also a soil sample from a peat bog. The Chloroethyne and merely comprise small measurements were performed with a purge and trap GC/ fragments of the aromatic ring; thus, we expected further MS system following various optimization steps. Under volatile compounds that are generated in the course of SOM certain conditions, we obtained 16.7 ng of carbon suboxide decomposition. and ethene are some from a 250 mg soil sample. We also found that the plausible candidates. On the basis of mechanistic models, formation of carbon suboxide requires a definite activation we postulated that carbon suboxide is a conceivable inter- energy and that it is rather short-lived in the natural mediate among the C3-fragment molecules during the environment. A subsequent reaction to malonic is cleavage of the aromatic ring. This finally led to the assumption that the natural formation of C3O2 in soil should expected in the presence of . It is shown that iron- be possible. (III), hydrogen peroxide, and chloride are prerequisites for its To verify this hypothesis, we selected catechol as the formation. Experimental parameters for the highest yield starting model compound because it is a constituent of of carbon suboxide depend on the precise molecular structure natural SOM. Iron as a redox-sensitive element, promoting of the model compound or on the individual soil sample, the necessary electron transfer for the degradation reaction, respectively. The presented results point to a new degradation and hydrogen peroxide as a strong oxidant were chosen. process for aromatic compounds in soil. Additionally, chloride was used to enable halogenation processes during the degradation pathway, which may affect the formation of carbon suboxide. Preliminary results from Introduction the catechol reaction showed an elevated yield, if either Publication Date (Web): October 12, 2007 | doi: 10.1021/es071530z With four successive double bonds, carbon suboxide (Od chloride or chlorine covalently bound to the aromatic CdCdCdO) is a very unstable compound. This linear and structure are present. A chlorinated catechol was therefore

Downloaded by UNIV HEIDELBERG on September 15, 2009 | http://pubs.acs.org symmetric molecule, known as doubly dehydrated malonic applied in a series of optimization steps, wherein the influence acid, is a toxic gas with an unpleasant odor. Its escharotic of different parameters on the production of C3O2 was property is principally due to a remarkable charge distribu- tested: the time dependence, the concentration of hydrogen tion. The central carbon atom is the most nucleophilic center peroxide, the concentration of chlorinated catechol, the pH (1). C3O2 is used in organic chemistry in various addition value, and the temperature. To confirm a possible natural reactions with nucleophilic substances, such as alcohols or formation of carbon suboxide, studies with soil samples were phenols, producing the corresponding esters (2). In addition, carried out, which are discussed in the last section of this carbon suboxide can polymerize to the chemically and paper. thermodynamically stable (C3O2)n (3), which has been considered a constituent of the atmosphere of Venus, and Materials and Methods it also has been proposed to be responsible for the red color Chemicals. The following chemicals were used as received: of the surface of Mars (1, 4). Therefore, a natural occurrence catechol (99%; Lancaster Synth), 4-chlorocatechol (97%; of C3O2 is conceivable. Aldrich), 3,5-dichlorocatechol (97%; Aldrich), 4,5-dichloro- In recent research (5), the use of polymerized carbon catechol (97%; Aldrich), isoprene (99%; Aldrich), hydrogen suboxide with an average oxidation state of 1.33 is discussed peroxide (30%; Merck), iron(III) sulfate (Fe 21-23%; Riedel- as an alternative end product of fossil fuel burning instead de-Hae¨n), potassium chloride (99%+, Merck), and of volatile carbon dioxide. The aim is to obtain a still profitable hydroxide (99%, Aldrich). Doubly distilled deionized water energy yield and yet reduce the release of the greenhouse (18 MΩ cm), from an ELGASTAT UHQ PS water treatment gas, CO2. system, was employed in all experiments. All pH measure- * Corresponding author phone: ++49 6221 544818; fax: ++496221 ments were made using a Mettler Toledo 320 pH meter, 545228; e-mail: [email protected]. calibrated on the free hydrogen scale with 4.0 and 7.0 buffers.

7802 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007 10.1021/es071530z CCC: $37.00  2007 American Chemical Society Published on Web 10/12/2007 Instrumentation. GC/MS analyses were performed on a Varian STAR 3400 Cx gas chromatographic system connected TABLE 1. Properties of Rotwasser Soil to a Saturn 2000 trap mass spectrometer. Sampling, soil type peatland preconcentration, and injection of the volatile analytes were controlled by a Tekmar purge and trap system. The GC was geographic position 49°,36′,39′′ N, 8°,53′,11′′ E a equipped with a BP624 (25 m; 0.53 mm i.d.; 3.0 µm film organic carbon (Corg) 1.64% inorganic carbonb <0.05% thickness) and a BPX5 (60 m; 0.32 mm i.d.; 1.0 µm film Cl- concentrationc 4.5 ppm thickness) capillary column, both connected in series. Fe concentrationd 1.8% Experimental Procedure. The experiments were carried pH value 4.0 out in airtight closed 20 mL headspace vials, filled with a b c 10 mL of aqueous solution. They were shaken for a CS measurement. Via a carbonate bomb (16). IC measurement. d µ-XRF measurement. determined time on a rotary board at 500 rpm at a precise temperature. After introduction of an inlet and an outlet stainless steel needle through the septum, the volatile compounds were purged for 4 min with a stream of helium at 40 mL/min and concentrated on a cooled trap (below -60 °C), filled with Tenax. Moisture was removed on a magnesium perchlorate water trap from the sample stream. The loaded preconcentration trap was switched into the GC carrier gas line (He, 2.0 mL/min) and heated to 180 °C. The desorbed compounds were refocused at the beginning of the capillary column, which was cooled with liquid to -90 °C. After 4 min, the GC oven temperature program started automatically: -90 to 0 °Cat40°C/min, hold 1 min, 0 to 150 °Cat5°C/min, hold 5 min, 150 to 210 °Cat30°C/min, hold 15 min. Preliminary Tests. The formation of carbon suboxide was examined using catechol as well as chlorinated catechols. In these assays, solutions of 10 mL of water with initial FIGURE 1. Chromatogram section, showing the formation of C3O2 concentrations of 0.50 mM , 2.50 mM Fe2- in preliminary tests with catechol, 4-chlorocatechol, 3,5-dichlo- (SO4)3, 5.00 mM H2O2, and 5.00 mM KCl were prepared and rocatechol, 4,5-dichlorocatechol, and tetrachlorocatechol using Fe - ° 2 shaken for 20 s at 40 C. (SO4)3,H2O2, and KCl. Process Optimization. On the basis of the preliminary results, 3,5-dichlorocatechol was chosen for the next series of experiments. The yield of carbon suboxide was to be this, solutions of 10 mL of water with initial concentrations maximized by systematic sequential variation of the following of 0.50 mM catechol, 2.50 mM Fe2(SO4)3, 5.00 mM H2O2, and reaction parameters. 5.00 mM KCl were prepared and shaken for 1 min at 40 °C. Time Dependence. The first optimization step was to Soil Sampling. The next logical step was to apply the elaborate the time dependence of C3O2 formation and knowledge gained with the model compounds to natural degradation. Solutions of 10 mL of water with initial samples. The employed peatland soil was sampled in March concentrations of 0.50 mM 3,5-dichlorocatechol, 2.50 mM 1996 from the natural reserve Rotwasser, Odenwald/Germany Fe2(SO4)3, and 5.00 mM H2O2 were prepared and shaken at (49°,36′,39′′ N, 8°,53′,11′′ E). After the collection, it was freeze- 40 °C for different durations. dried and milled to a fine powder. A total of 250 mg of this Concentration of Hydrogen Peroxide. The second series of powdered soil was used for each experiment. The charac- measurements concerned the influence of hydrogen peroxide terization of the soil is shown in Table 1. on the yield of C3O2. Therefore, solutions of 10 mL of water Identification and Quantification of Carbon Suboxide.

Publication Date (Web): October 12, 2007 | doi: 10.1021/es071530z with initial concentrations of 0.50 mM 3,5-dichlorocatechol For a qualitative identification, C3O2 was produced by the and 2.50 mM Fe2(SO4)3 were prepared with varying con- thermolysis of malonic acid bis(trimethylsilyl) ester (17) and centrations of H2O2 and shaken for 1 min at 40 °C.

Downloaded by UNIV HEIDELBERG on September 15, 2009 | http://pubs.acs.org introduced into the purge and trap GC/MS system. The C3O2 Concentration of 3,5-Dichlorocatechol. The third param- peak appeared after a retention time of 13.40 min showing eter to be checked was the amount of 3,5-dichlorocatechol m/z 68 (M+) and 53 (M+ - O + H) as the most prominent in relation to the oxidants. For this purpose, solutions of 10 fragments of its mass spectrum. Nevertheless, quantification mL of water with initial concentrations of 2.50 mM Fe2(SO4)3 of this organic compound is not simple, due to its escharotic and 0.71 mM H2O2 were prepared together with different behavior and to the lack of standard reference material. concentrations of 3,5-dichlorocatechol and shaken for 1 min Therefore, to estimate the response factor of carbon suboxide, at 40 °C. according to previous arguments in a related case (14), we pH Value. To test for a possible pH dependence of the used a roughly similar molecule: isoprene. Both compounds carbon suboxide production, solutions of 10 mL of water were assumed to have an identical recovery during the with initial concentrations of 0.50 mM 3,5-dichlorocatechol, different steps of the purge and trap GC/MS method. 2.50 mM Fe2(SO4)3, and 0.71 mM H2O2 were prepared. The Furthermore, both analytes have comparable ionization pH was adjusted by adding different quantities of NaOH. energies: 10.6 eV for carbon suboxide and 8.86 eV for isoprene The solutions were shaken for 1 min at 40 °C. (18). The mass spectrometric response of a compound is Temperature. Finally, the applied reaction temperature determined by two factors: the ionization cross-section for was varied. Solutions of 10 mL of water with initial con- 70 eV electrons, which is largely dependent on the ionization centrations of 0.50 mM 3,5-dichlorocatechol, 2.50 mM Fe2- energy of the molecule, and the relative abundance of the (SO4)3, 0.71 mM H2O2, and 12.5 mM NaOH were prepared in the spectra selected for quantification. The difference and were shaken for 1 min at different temperatures. in the ionization potentials of both compounds should be Catechol. Using the results of the optimization procedure of minor importance, as electrons with an energy of 70 eV on 3,5-dichlorocatechol, a further enhancement of the are used in the MS. The selected masses for chromatographic observed C3O2 production from catechol was assayed. For quantitation were m/z 68 in the case of carbon suboxide and

VOL. 41, NO. 22, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7803 Publication Date (Web): October 12, 2007 | doi: 10.1021/es071530z

FIGURE 2. Effect of different parameters on carbon suboxide formation. (A) Time dependence: reaction of 3,5-dichlorocatechol, Fe2(SO4)3, and H O at 40 °C; (B) variation of hydrogen peroxide concentration from 0 to 100 µmol: reaction of 3,5-dichlorocatechol, Fe (SO ) , and

Downloaded by UNIV HEIDELBERG on September 15, 2009 | http://pubs.acs.org 2 2 2 4 3 H2O2 at 40 °C after 1 min; (C) variation of the 3,5-dichlorocatechol concentration from 2.50 to 25.0 µmol: reaction of 3,5-dichlorocatechol, Fe2(SO4)3, and H2O2 at 40 °C after 1 min; (D) pH value: reaction of 3,5-dichlorocatechol, Fe2(SO4)3,H2O2, and NaOH at 40 °C after 1 min; and (E) agitation temperature: reaction of 3,5-dichlorocatechol, Fe2(SO4)3,H2O2, and NaOH after 1 min. Symbols are average data from experiments, and error bars indicate standard deviations from triplicate measurements.

m/z 53 + 67 for isoprene. Therefore, the response factor of Therefore, the abundance of m/z 68 was calculated for each C3O2 can be estimated according to eq 1 measurement individually, assuming that the error becomes more important at higher concentrations. abundance (m/z 68) rf (C O ) ) rf (isoprene) (1) 3 2 abundance (m/z 53 + 67) Results and Discussion Preliminary Tests. The results of the preliminary assays on The relative abundances were calculated from reference carbon suboxide formation from catechol and its chlorinated spectra with a mass scan range from m/z 10 to 249, recorded analogues are shown in Figure 1. Because 3,5-dichlorocat- at different compound concentrations. For isoprene, the echol as well as 4,5-dichlorocatechol were the most effective relative weight of the quantitation ions m/z 53 + 67 remained starting compounds for C3O2 production, the former was approximately constant at 33.7%, whereas in the case of selected for the optimization test series. carbon suboxide, the computation turned out to be more Optimization. Time Dependence. The measurements complicated. At higher concentrations, we observed the show that the optimum time for highest carbon suboxide decrease of m/z 68 and the formation of an ion at m/z 80 yield is very short, approximately 30 s to 1 min. An exponential that was generated by collision processes in the ion trap and decrease is observed, and C3O2 is no longer detectable after that has already been observed by Bortolini et al. (1). 30 min (Figure 2A). This agrees with its known property of

7804 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007 SCHEME 1. Presumed Degradation Pathway of Catechol to suboxide increases by a factor of 10 with respect to the case Carbon Suboxide where no sodium hydroxide is used (pH 2.2). Moreover, the measurements reveal a very sensitive system under these conditions, especially regarding iron speciation. By adding sodium hydroxide to the solution, a part of the Fe(III) precipitates as hydroxide, and the amount of Fe(III) available for the reaction is reduced. Particularly small differences in the preparation, the handling, the shaking, or the transferring of the samples lead to fluctuating yields of C3O2. Therefore, reproducible results are significantly more difficult to achieve being rather unstable and readily reacting with nucleophilic than at lower pH values, explaining the larger error bars in compounds such as H2O to form malonic acid. To reduce the figures. Hence, the employed NaOH concentration for these losses, yet affording a practicable experimental protocol, the last optimization step was chosen to be 12.5 mM. an agitating time of 1 min was chosen for the next optimiza- Temperature. The experiments performed at room tem- tion steps. perature up to 70 °C show approximately constant amounts Concentration of Hydrogen Peroxide. When the concen- of carbon suboxide above 30 °C, with a distinct decrease to tration of hydrogen peroxide is varied, an optimal production lower temperatures (Figure 2E). This finding could be of carbon suboxide is achieved at a molar ratio of 1:7 as explained by an appreciable activation energy necessary for compared to the concentration of iron(III) (Figure 2B). No the production of carbon suboxide, which competes with C3O2 is detected in the absence of H2O2, which indicates C3O2-destroying processes at higher temperatures. that, for the production of carbon suboxide, the presence of Finally, starting from 5 µmol of 3,5-dichlorocatechol in OH radicals is necessary. They are generated via Fenton the preliminary tests, the different optimization steps of the reaction from H2O2 and Fe(II), the latter through the carbon suboxide production permitted us to accomplish a preceding reduction of Fe(III) by the organic substrate. On yield of 7.9 µgofC3O2, which is 1000 times higher than before the other hand, H2O2 concentrations exceeding a 1:7 molar the optimization. Next to the time dependence of the reaction, ratio to iron are not favorable for an optimal yield. Presum- the influence of the pH value proves to be the most important. ably, side reactions, leading to higher oxidized product Nevertheless, it is also obvious that in this sensitive system, compounds (e.g., CO and CO2 (15)) become more efficient further maxima of C3O2 can be expected from small simul- under these conditions. Therefore, a ratio of 1:7 (H2O2/Fe) taneous parameter changes. was used to evaluate the other parameters. Catechol. The experiments with catechol demonstrate Concentration of 3,5-Dichlorocatechol. In the third op- that a production of carbon suboxide is possible. However, timization step, the yield of C3O2 has to be set in relation to the conditions are somewhat different as compared to those the concentration of the organic matter employed because employed with 3,5-dichlorocatechol. An important factor is higher concentrations deliver more material for the degra- potassium chloride. In the absence of KCl, no carbon suboxide dation. This ratio (amount of carbon suboxide)/(employed is detected. However, when larger amounts of KCl are added, molar quantity of 3,5-dichlorocatechol) shows a distinct the C3O2 yield increases significantly. In comparison, the maximum at a molar concentration of 3,5-dichlorocatechol, experiments with 3,5-dichlorocatechol do not show any effect which is very close to 1/10 of the Fe(III) molarity (Figure 2C). of potassium chloride addition. This indicates that the Small deviations of the 3,5-dichlorocatechol concentration presence of chlorine in the system, either as substituent or from this value result in lower amounts of carbon suboxide. as anion, is necessary for the production of C3O2. However, Hence, a definite stoichiometric ratio between iron and 3,5- it remains difficult to decide whether covalently bound Cl dichlorocatechol seems to be as important as an appropriate atoms or anionic Cl- are more efficient for the formation of hydrogen peroxide concentration. Consequently, an initial carbon suboxide. Indeed, the higher yield obtained from 3,5- concentration of 0.25 mM 3,5-dichlorocatechol was used for dichlorocatechol (approximately 109 ng) as compared to that the next optimization steps. obtained with catechol (approximately 8.8 ng) under similar pH Value. Thus far, the pH value of the solution was mainly conditions might favor the covalently bound chlorine.

Publication Date (Web): October 12, 2007 | doi: 10.1021/es071530z determined by the concentration of iron(III) sulfate. The The necessity of chloride for the carbon suboxide forma- addition of sodium hydroxide shows that the optimized C3O2 tion from catechol points to a chlorinated intermediate (e.g., yield from 3,5-dichlorocatechol is more favorable at pH values chlorocatechol) that is subsequently hydroxylated to form Downloaded by UNIV HEIDELBERG on September 15, 2009 | http://pubs.acs.org ranging from 2.7 to 3.5 (Figure 2D). The production of carbon highly unstable tetrahydroxybenzene leading to two C3O2

FIGURE 3. GC chromatogram of a natural soil sample and mass spectrum of carbon suboxide. Initial concentrations: 250 mg of Rotwasser soil sample, 12.5 mM Fe2(SO4)3, 25.0 mM H2O2, and 25.0 mM KCl in 10 mL of H2O. The solution was shaken for 1 min at 40 °C.

VOL. 41, NO. 22, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 7805 molecules after ring fission (Scheme 1). But so far, experi- (3) Ballauff, M.; Rosenfeldt, S.; Dingenouts, N.; Beck, J.; Krieger- mental data do not convincingly imply that chlorinated Beck, P. Analysis of poly(carbon suboxide) by small-angle X-ray - catechols or quinones are always required as intermediates. scattering. Angew. Chem., Int. Ed. 2004, 43, 5843 5846. (4) Chen, H.; Holmes, J. L. The generation of OC O•+ and OC O and Soil Samples. In contrast to 3,5-dichlorocatechol, catechol 2 2 a study of ionized OC3O and C2O by tandem mass spectrometry. can be found as a natural constituent of SOM. Therefore, Int. J. Mass Spectrom. Ion Processes 1994, 133, 111-119. this compound is a more adequate representative for soil, (5) Fridman, A.; Gutsol, A. CO2-free energy and hydrogen production and it is plausible to apply reaction conditions that have from hydrocarbons. Energy Fuels 2006, 20, 1242-1249. (6) Harper, D. B.; Kalin, R. M.; Hamilton, J. T. G.; Lamb, C. Carbon been optimized for a maximum C3O2 yield from catechol rather than from 3,5-dichlorocatechol. The corresponding isotope ratios for chloromethane of biological origin: Potential tool in determining biological emissions. Environ. Sci. Technol. GC chromatogram and mass spectrum are shown in Figure 2001, 35, 3616-3619. 3 A,B. (7) Scho¨ler, H. F.; Keppler, F. Abiotic formation of organohalogens As with catechol, the addition of potassium chloride is during early diagenetic processes. In The Handbook of Envi- crucial. The small amounts of carbon suboxide obtained ronmental Chemistry. The Natural Production of Organohalogen without KCl could arise either from chloride (4.5 ppm) or Compounds; Hutzinger, O., Gribble, G. W., Eds.; Springer: Berlin, 2003; Vol. 3, pp 63-84. from covalently bound chlorine already present in the soil (8) Yokouchi, Y.; Noijiri, Y.; Barrie, L. A.; Toom-Sauntry, D.; Machida, samples. T.; Inuzuka, Y.; Akimoto, H.; Li, H.-J.; Fujinuma, Y.; Aoki, S. A In summary, in the catechol experiments, we obtained strong source of methyl chloride to the atmosphere from tropical - 8.8 ng of C3O2 from 0.36 mg of carbon, whereas 16.7 ng of coastal land. Nature (London, U.K.) 2000, 403, 295 298. s C3O2 resulted from 4.1 mg of carbon in the soil sample (Table (9) Harper, D. B. Halomethane from halide ion A highly efficient 1). In addition, the concentrations of Fe(III), H O , and Cl-, fungal conversion of environmental significance. Nature (Lon- 2 2 don, U.K.) 1985, 315,55-57. required to obtain the elevated carbon suboxide yields from (10) Duesterberg, C. K.; Waite, T. D. Process optimization of Fenton the natural soil samples, were slightly higher than those from oxidation using kinetic modeling. Environ. Sci. Technol. 2006, the model compound catechol. The following hypotheses 40, 4189-4195. are conclusive, due to the complex structure of SOM: the (11) Petigara, B. R.; Blough, N. V.; Mignerey, A. C. Mechanisms of organic carbon of the soil sample basically consists of other hydrogen peroxide decomposition in . Environ. Sci. Technol. - non-degradable organic compounds, or the formed carbon 2002, 36, 639 645. (12) Keppler, F.; Eiden, R.; Niedan, V.; Pracht, J.; Scho¨ler, H. F. suboxide is intercepted before it is liberated into the gas Halocarbons produced by natural oxidation processes during phase, or the oxidants are consumed by other constituents degradation of organic matter. Nature (London, U.K.) 2000, 403, in the soil. Therefore, the study of these processes has to be 298-301. individually optimized for different soils. (13) Keppler, F.; Borchers, R.; Pracht, J.; Rheinberger, S.; Scho¨ler, H. This work suggests that the formation of C O points to F. Natural formation of vinyl chloride in the terrestrial environ- 3 2 - a new degradation pathway for aromatic structures that also ment. Environ. Sci. Technol. 2002, 36, 2479 2483. - (14) Keppler, F.; Borchers, R.; Hamilton, J. T. G.; Kilian, G.; Pracht, takes place in soil. Up to now, cleavage of the C C bond J.; Scho¨ler, H. F. De novo formation of chloroethyne in soil. between the hydroxyl groups of catechols, leading to dicar- Environ. Sci. Technol. 2006, 40, 130-134. bonic such as muconic acid (7, 15, 19-21), or the (15) Pracht, J.; Boenigk, J.; Isenbeck-Schro¨ter, M.; Keppler, F.; Scho¨ler, complete destruction of the aromatic ring into small com- H. F. Abiotic Fe(III) induced mineralization of phenolic sub- stances. Chemosphere 2001, 44, 613-619. pounds such as CO2, vinyl chloride, and chloroethyne (7, 12-15) has been known. From the present study, it seems (16) Gastner, M.; Mu¨ller, G. The “Karbonat-Bombe”, a simple device for the determination of the carbonate content in sediments, conceivable that the electron-rich phenolic C6 ring could soils, and other materials. Neues Jahrb. ., Monatsh. 1971, also be split symmetrically, resulting in two C3O2 molecules 10, 466-469. without a loss of carbon atoms. Further investigations should (17) Birkhofer, L.; Sommer, P. Eine neue synthese von kohlensuboxid focus on polar intermediates to elucidate the precise und seine reaktionen mit silylierten aminen und silanolen. - degradation pathway. Chem. Ber. 1976, 109, 1701 1707. (18) Rosenstock, H. M.; Draxl, K.; Steiner, B. W.; Herron, J. T.; Lias, S. G. Ion Energetics Data; NIST: Gaithersburg, MD, 2003. Acknowledgments (19) Li, X.; Cubbage, J. W.; Tetzlaff, T. A.; Jenks, W. S. Photocatalytic We thank B. Breuninger and D. Pieroth for execution and degradation of 4-chlorophenol. 1. The hydroquinone pathway. - Publication Date (Web): October 12, 2007 | doi: 10.1021/es071530z assistance in a portion of the experimental research during J. Org. Chem. 1999, 64, 8509 8524. this work as well as A. Cheburkin and S. Rheinberger for their (20) Li, X.; Cubbage, J. W.; Jenks, W. S. Photocatalytic degradation of 4-chlorophenol. 2. The 4-chlorocatechol pathway. J. Org.

Downloaded by UNIV HEIDELBERG on September 15, 2009 | http://pubs.acs.org technical support. We also thank J. Pracht and F. Keppler for Chem. 1999, 64, 8525-8536. collecting the Rotwasser soil sample. This work was supported (21) Gupta, S. S.; Stadler, M.; Noser, C. A.; Ghosh, A.; Steinhoff, B.; by the German Research Foundation (DFG). Lenoir, D.; Horwitz, C. P.; Schramm, K.-W.; Collins, T. J. Rapid total destruction of chlorophenols by activated hydrogen peroxide. Science (Washington, DC, U.S.) 2002, 296, 326- Literature Cited 328. (1) Bortolini, O.; Pandolfo, L.; Tomaselli, C.; Traldi, P. Ion-molecule chemistry of carbon suboxide in an ion-trap mass spectrometer. Received for review June 22, 2007. Revised manuscript re- Int. J. Mass Spectrom. 1999, 190-191, 171-179. ceived August 30, 2007. Accepted September 7, 2007. (2) Kappe, T.; Ziegler, E. Carbon suboxide in preparative organic chemistry. Angew. Chem., Int. Ed. Engl. 1974, 13, 491-504. ES071530Z

7806 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 22, 2007