Hydrodefluorination and Hydrogenation of Fluorobenzene under Mild

Aug 7, 2012 - Al2O3 and H2) for hydrodefluorination and hydrogenation of fluorobenzene under mild aqueous conditions (1 atm of H2, ambient temperature...
3 downloads 0 Views 1MB Size
Article pubs.acs.org/est

Hydrodefluorination and Hydrogenation of Fluorobenzene under Mild Aqueous Conditions Rebekka Baumgartner and Kristopher McNeill* Institute of Biogeochemistry and Pollutant Dynamics, ETH Zurich, 8092 Zurich, Switzerland S Supporting Information *

ABSTRACT: Fluorinated organic compounds are increasingly used in many applications, and their release to the environment is expected. It is therefore important to find suitable methods for degradation of fluorinated compounds under environmentally relevant conditions. In this study, a simple heterogeneous rhodium-based catalytic system (Rh/ Al2O3 and H2) for hydrodefluorination and hydrogenation of fluorobenzene under mild aqueous conditions (1 atm of H2, ambient temperature) was developed and the underlying reaction mechanism was investigated. Fluorobenzene degraded rapidly (t1/2 ≈ 0.2 h) to form cyclohexane and fluoride (F−) as the stable end products, with benzene and cyclohexene observed as intermediates. Cyclohexadiene intermediates were not observed but were expected to form during the hydrogenation of benzene. Three postulated but unobserved fluorinated intermediates were subjected to the catalytic reaction conditions, and it was concluded that they most likely do not form during the fluorobenzene degradation reaction. Isotope labeling experiments showed that the unsaturated intermediates undergo rapid and reversible hydrogenation/dehydrogenation under the reaction conditions and also that fully saturated compounds are unreactive in the catalytic system. Both molecular hydrogen and water were sources of hydrogen in the final cyclohexane product. Kinetic fitting indicated that sorption/desorption of fluorobenzene onto the catalyst surface plays an important role in the mechanism.



concentrations of fluorobenzene in groundwater up to 700 μg L−1 have been found close to an abandoned industrial complex.13 The catalytic degradation of fluorobenzene described herein is conducted in water under mild reaction conditions, features that are important for remediation of fluorocarbon-contaminated natural waters. In the past decades, research on the occurrence and degradation of fluorochemicals has mainly been focused on perfluorinated compounds, which are highly persistent in the environment and for which degradation methods are hard to find.14 For fluorinated aromatics, several microbial (e.g., ref 15), chemical (e.g., refs 16 and 17), photochemical (e.g., ref 18), and photocatalytic (e.g., ref 19) methods have been explored. Fluorobenzene specifically has been shown to degrade microbially,20−22 with microorganisms isolated from contaminated sites, and chemically using various catalysts.23−29 Rh-based catalysts have found successful use in several reported fluorochemical degradation studies. In fact, Rh complexes are generally considered among the best for C−F activation and have been employed as homogeneous C−F activation catalysts in many studies.30,31 For fluorobenzene, Blum et al.23 found that the active catalyst formed from RhCl3·3H2O, Aliquat 336, catalyzes defluorination and hydro-

INTRODUCTION Fluorinated organic chemicals are used widely in numerous applications, ranging from agrochemicals and pharmaceuticals to specialty solvents, surfactants, and materials. One of the selling points of organic fluorochemicals is their high stability, resulting from both fluoride’s poor leaving group ability and the strong carbon−fluorine bond.1−3 The increased industrial use of fluorochemicals has raised concerns about their fate in the environment.1,4 Furthermore, the chemical stability of fluorocarbons has fueled the search for mild and selective fluorochemical degradation methods.5−7 Fluoroaromatics are an important subclass of fluorinated organics. In 1994 their estimated world market was at 10 000 tons per year,1 and the global production capacity of the most important fluoroaromatics was approximately 35 000 tons per year in 2000.8 The increase of fluorinated pharmaceuticals on the market9 and recent advancements in fluoroorganic synthesis10 both indicate that the production of fluorinated compounds has further increased since 2000. The present work is focused on fluorobenzene, which is the archetypal and simplest fluorinated aromatic compound. Fluorobenzene is listed as a high production volume (HPV) chemical by the U.S. EPA,11 with an estimated yearly global production capacity of 5000 tons in 2000.8 It is considered highly persistent in air, and also resistant to hydrolysis, and not readily biodegradable.12 Fluorobenzene is mainly used as an intermediate in the fluorochemical industry,8 so fluorobenzene groundwater contamination would be expected near industrial sites. Indeed, © 2012 American Chemical Society

Received: Revised: Accepted: Published: 10199

June 1, 2012 August 6, 2012 August 7, 2012 August 7, 2012 dx.doi.org/10.1021/es302188f | Environ. Sci. Technol. 2012, 46, 10199−10205

Environmental Science & Technology

Article

genation to give cyclohexane under 16 atm of H2 at 80 °C. In other studies, it was shown that rhodium pyridylphospine bipyridine complexes lead to defluorination of fluorobenzene under 4 atm of H2 at 70 °C in aqueous−organic biphasic systems.27,29 Buil et al.24 developed Rh-based nanoparticles that degraded fluorobenzene to benzene under 1 atm of H2 at 60 °C, and Ukisu and Miyadera28 showed hydrodehalogenation of fluorobenzene with Rh/C below 60 °C with 2-propanol as the solvent and H-donor. In addition to the Rh-based catalysts, there are examples of the degradation of fluorobenzene with Ni-based25 and Al-based26 catalysts. While each of these studies is interesting, none of them were performed under environmentally relevant conditions. Many employ high temperatures that are unfeasible given the energy required to heat large volumes of water, and many also employ organic solvents or biphasic organic/aqueous mixtures, which would most likely contaminate the water with solvent, even as fluorobenzene is removed. In addition, the use of homogeneous catalysts and high H2 pressure is less practical than that of a heterogeneous catalyst and low H2 pressure. As some of the Rh-based defluorination catalysts are compatible with water,27,29 they served as precedents for our development of an aqueous catalytic C−F activation system. Reinhard and co-workers developed and implemented an important system for groundwater remediation involving halocarbon degradation under environmentally relevant conditions. They employed heterogeneous Pd-based catalysts and used hydrogen as the reducing agent to degrade a wide range of chlorinated contaminants, including chlorinated ethylenes, chloroform, chlorinated alkanes, and chlorinated aromatics.32,33 The process has been tested with groundwater constituents,34,35 and the fouling of the catalysts and their regeneration were investigated.36,37 Successful remediation of contaminated sites with Pd/Al2O3 or Pd/zeolite was achieved in different field studies.38−40 While these heterogeneous Pd catalysts have proven quite versatile, they and other Pd-based catalysts generally display low activity toward fluorobenzene.41−43 Inspired by these practice-oriented Pd-based studies, we sought to take advantage of the notable defluorination capabilities of Rh-based catalysts to develop a degradation method for fluorobenzene that could also function under environmentally relevant conditions. Here we describe the complete degradation of fluorobenzene to cyclohexane with the heterogeneous catalyst Rh/Al2 O3 under mild, aqueous conditions (ambient temperature, 1 atm of H2) and present the results of experiments aimed at uncovering the underlying mechanistic details involved.

LA-950 laser scattering particle size distribution analyzer). No special precautions were taken to avoid exposure to air prior to the batch experiments. Batch Experiments. Batch experiments with fluorobenzene and reaction intermediates were performed in glass bottles (246 mL) sealed with crimp caps lined with rubber septa in buffered Milli-Q water (164 mL, pH 7, 0.01 M phosphate buffer) at room temperature (21 ± 2 °C). The solution was sparged with H2 (or D2) for 15 min prior to the reaction and kept under constant H2 pressure during the course of the reaction (1 atm). Starting materials (fluorobenzene and intermediates) were individually added by syringe through the septum to provide an initial compound concentration of 100 μM. The system was equilibrated for 15 min under constant stirring. Rh/Al2O3 (6−8 mg) powder was added using a glass syringe by rinsing it with the reaction solution. Headspace samples (100 μL) were taken at defined time points using a gastight syringe (500 μL) and directly injected into the gas chromatograph. Analytical Methods. The headspace aliquots were analyzed by gas chromatography coupled to mass spectrometry (GC−MS) using a Thermo Scientific Trace GC Ultra and a Thermo Scientific DSQ quadrupole MS system. The column used was a Restek Rtx-1 Crossbond 100% dimethylpolysiloxane (30 m × 0.32 mm × 4 m film thickness). The oven temperature profile was 70 °C (2 min) and ramping at 40 °C min−1 to 140 °C (2.5 min). Standard calibration curves were derived from 20 mL crimp cap headspace vials with the same water-toheadspace ratio as the bottles used for the batch experiments and were based on the total amount added to the flask. Limit of detection (LOD) and limit of quantification (LOQ) values for the substances with available analytical standards are reported in the Supporting Information. Evaluation of Reaction Kinetics. The degradation of fluorobenzene and intermediates in the batch experiments can be described by a pseudo-first-order kinetic rate law. Hydrogen pressure was assumed to be constant and available in excess. Pseudo-first-order rate constants (kobs) were estimated using AQUASIM44 by minimizing weighted least-squares estimates.



RESULTS AND DISCUSSION Fluorobenzene Degradation. Fluorobenzene underwent rapid degradation in water under mild conditions (ambient temperature, 1 atm of H2) in the presence of Rh/Al2O3 catalyst. With a catalyst loading corresponding to 1.89 mgRh L−1 (18.3 mol % Rh relative to fluorobenzene), degradation of fluorobenzene was complete within 1 h (t1/2 ≈ 0.2 h). Derived rate constants (kobs) of four replicate batch experiments deviated up to 20% from each other, which we ascribe to the heterogeneity of the catalyst. The initial rate corresponded to a turnover frequency (TOF) of about 80 fluorobenzene molecules per Rh atom per minute. No attempt was made to explore the maximum possible TOF of the reaction. Both the Rh/Al2O3 and the H2 were essential for the reaction to proceed. Benzene and cyclohexene were observed as the major and minor reaction intermediates, respectively, and the final stable organic product of the catalytic reaction was cyclohexane (see Figure 1). Stoichiometric release of fluoride (F−) was confirmed by ion chromatography. The carbon mass balance dropped to 80−90% (four replicates) of the initial fluorobenzene after approximately 15 min but recovered to 94−113% at the end of the reaction (60 min). We speculate that the missing components of the mass balance after 15 min correspond to surface-bound species. Thermodynamically, the



MATERIALS AND METHODS Reagents. Hydrogen (99.999%), N2 (99.999%), and He (99.999%) were purchased from Carbagas (Switzerland). Deuterium (D2; 99.8% D), cyclohexane, cyclohexene, benzene, fluorobenzene, and 1,3- and 1,4-cyclohexadiene were purchased from Sigma-Aldrich (Switzerland). Fluorocyclohexane was purchased from TCI Europe (Belgium). Milli-Q water (18.2 MΩ cm−1), generated from a NANOpure Diamond purifying system (Barnstead), was used for all experiments. Catalyst. The catalyst material, 5% Rh by weight on γ-Al2O3 (Rh/Al2O3), was purchased from Sigma-Aldrich. The manufacturer-specified metal weight was assumed to be accurate. The Brunauer−Emmett−Teller (BET) surface area was 150.7 m2 g−1, determined by N2 adsorption (Quantachrome Autosorb). The mean particle diameter of the catalyst was 55 μm (Horiba 10200

dx.doi.org/10.1021/es302188f | Environ. Sci. Technol. 2012, 46, 10199−10205

Environmental Science & Technology

Article

occur in either order. If fluorobenzene were hydrogenated prior to defluorination, the formation of fluorocyclohexadienes (five isomers), fluorocyclohexenes (three isomers), and fluorocyclohexane would be expected (Figure 2). These intermediates are Figure 1. Fluorobenzene is degraded to cyclohexane and fluoride (100% conversion) via benzene (observed, major intermediate), cyclohexadienes (not observed), and cyclohexene (observed, minor intermediate) using Rh/Al2O3 and H2 in water. Details of the assignments as being reversible or irreversible are presented in the text. For a full depiction of the proposed mechanism, including information from the kinetic fitting, see Figure 5.

overall reaction from fluorobenzene to cyclohexane in water is exothermic (ΔHrxn = −357 kJ mol−1; see the Supporting Information). Proposed Mechanistic Pathway. On the basis of the observed intermediates, reaction kinetics, and mechanistic studies using deuterium-labeled reaction components (D2, D2O, and fluorobenzene-d5), a reaction mechanism for the Rhcatalyzed fluorobenzene degradation process can be proposed. The favored mechanistic model is presented in Figure 1 and is comprised of fluorobenzene hydrodefluorination to give benzene as a reaction intermediate, followed by hydrogenation to cyclohexane. The involvement of specific intermediates was examined and is described in the following sections. Observed Reaction Intermediates. The observation of benzene and cyclohexene as reaction intermediates is consistent with the proposed pathway in which rapid hydrodefluorination to benzene is followed by sequential hydrogenation reactions. In control experiments, both benzene and cyclohexene were hydrogenated to form cyclohexane when they were used as starting material for the Rh/Al2O3 batch experiment. The end product, cyclohexane, was stable under the conditions of the batch experiments. Rh-based catalysts are well-known to hydrogenate arenes,45,46 and several studies have reported hydrogenation of benzene derivatives under conditions as mild as those reported in the current study.47 We note that aqueous solvent appears to be the key common factor for achieving hydrogenation under such mild conditions.48−51 Unobserved Reaction Intermediates. To further understand the reaction mechanisms operating in this system, specifically the intermediacy of molecules that could be formed by hydrogenation or hydrodefluorination, a series of reactions were conducted with postulated, but unobserved intermediates. In these experiments, postulated intermediates were subjected to the batch reaction conditions to determine if they were labile and formed the observed end products, or if they were stable under the reaction conditions. In the batch experiments, benzene is hydrogenated to cyclohexane presumably through the intermediacy of 1,3- and 1,4-cyclohexadiene; however, neither diene is observed in these reactions. Using either 1,3-cyclohexadiene or 1,4-cyclohexadiene as starting material in the batch reactions resulted in rapid reaction to form cyclohexene and ultimately cyclohexane. We conclude that 1,3- or 1,4-cyclohexadiene or both are formed during the hydrogenation of benzene and are rapidly further hydrogenated under the batch reaction conditions to give cyclohexene and eventually cyclohexane. Although the presence of benzene, cyclohexene, and the end product cyclohexane suggests that hydrodefluorination occurs before hydrogenation, the two processes could theoretically

Figure 2. Possible routes from fluorobenzene to the observed intermediates involving cyclohexadienyl, cyclohexenyl, and cyclohexyl fluoride intermediates in the presence of Rh/Al2O3 and H2. Experimentally tested intermediates are depicted in black, and untested intermediates are shown in gray. Experimentally confirmed reactions are depicted in green, while disproven processes are indicated in red and crossed. Processes that could not be tested are shown in gray.

not observed, but this does not preclude their formation on the catalyst surface or at concentrations or for lifetimes that are below our detection limits. To test this possibility, we examined the stability of representative, putative fluorinated intermediates under the reaction conditions. The fluorinated cyclohexadiene, cyclohexene, and cyclohexane compounds can be subdivided into three functional categories: alkyl fluorides (fluorocyclohexane and 4-fluorocyclohexene), alkenyl fluorides (1-fluorocyclohexene, 1-fluorocyclohexa-1,3-diene, 2-fluorocyclohexa-1,3-diene, and 1-fluorocyclohexa-1,4-diene), and allylic fluorides (3-fluorocyclohexene, 5-fluoro-1,3-cyclohexadiene, and 3-fluoro-1,4-cyclohexadiene) (Figure 2). We examined the reactivity of one representative intermediate from each of these three categories. Exposure of fluorocyclohexane (an alkyl fluoride) to the batch reaction conditions resulted in no reaction. This observed stability of fluorocyclohexane under the reaction conditions excludes it as an intermediate in the degradation of fluorobenzene (Figure 2). Similarly, 1-fluorocyclohexene (an alkenyl fluoride) and 3-fluorocyclohexene (an allylic fluoride) were synthesized (see the Supporting Information) and exposed to the Rh/Al2O3 and H2 batch reaction conditions. 1-Fluorocyclohexene degraded first to cyclohexene and then formed cyclohexane. These observed products are consistent with the degradation pattern of fluorobenzene; therefore, 1fluorocyclohexene could be an intermediate along the reaction pathway. However, isotope labeling studies (discussed below) argue against the intermediacy of partially hydrogenated fluorinated compounds. 3-Fluorocyclohexene degraded quickly in the batch system even in the absence of Rh/Al2O3. The main product of this noncatalytic degradation was 3-cyclohexen-1-ol (detected by 10201

dx.doi.org/10.1021/es302188f | Environ. Sci. Technol. 2012, 46, 10199−10205

Environmental Science & Technology

Article

HPLC−UV/vis), suggesting nucleophilic substitution of fluoride by water. Treatment of 3-fluorocyclohexene with the batch reaction conditions (Rh/Al2O3 and H2) gave cyclohexanol (eq 1); however, cyclohexanol was not detected in the fluorobenzene batch experiment. As 3-fluorocyclohexene is unstable under the applied conditions and does not yield the observed products, its intermediacy is ruled out.

Taken together, the degradation behavior of these three potential intermediates indicates that if hydrogenation of fluorobenzene is occurring prior to defluorination, complete hydrogenation to fluorocyclohexane cannot occur. Additionally, the degradation of fluorobenzene likely does not proceed via 3fluorocyclohexene. These restrictions narrow the possibilities, but still leave room for a pathway involving some fluorinated intermediates. A summary of all observed and postulated intermediates, information about whether they are observed, and their observed or postulated reactivity are given in Table SI-6 (Supporting Information). Isotope Labeling Studies. To explore the mechanism further, experiments were conducted with deuterium-labeled reaction components, D2, D2O, and fluorobenzene-d5, individually and in combination. The results of these experiments led to three major conclusions. First, both molecular hydrogen and water are sources of hydrogen atoms in the final cyclohexane product. Second, some of the intermediates undergo rapid reversible hydrogenation/dehydrogenation under the reaction conditions. Third, fully saturated cyclohexane and fluorocyclohexane are unreactive in this system. In addition to these conclusions, the results of the isotope labeling experiments also suggest that defluorination occurs prior to hydrogenation. Each of these results is discussed in greater detail below. For the studies with labeled materials, no isotope effect was observed, but considering the uncertainty related to the degradation rate constants of the experiments with unlabeled material, we note that an isotope effect of 50 h) relative to the time scale of product formation, which was completed within less than 2 h. We therefore favor mechanisms in which H atoms come directly from each source (water and molecular hydrogen). Hydrogenation Is Reversible. When cyclohexene was added to the D2/H2O batch experiment, the end product was a mixture of deuterated cyclohexane products. The number of D atoms incorporated ranged from zero to a maximum of eight (Figure 3a). The percentages of the sum of the converted molecular ion intensities fit well to a Gaussian distribution with the parameters mean, amplitude, and variance (σ2). These fits were performed with the Levenberg−Marquardt algorithm (pro Fit, version 6.2.3, Quantum Soft). The fitted mean showed incorporation of 1.0 D atom into the cyclohexane products from cyclohexene, with a variance of 1.8 D atoms. The Gaussian-shaped distribution and the high maximum number of D atoms incorporated indicate that the hydrogenation process is reversible. If hydrogenation were irreversible, a maximum number of two D atoms could be incorporated into cyclohexane from cyclohexene. The catalyst Rh/Al2O3 thus catalyzes dehydrogenation as well as hydrogenation (Figures 1 and 3). Also, reaction at both faces of the ring must occur to have more than six deuterium atoms incorporated into a cyclohexane molecule. This is in contrast to systems which deliver six D atoms with an all-cis relationship and have been interpreted as binding to one face of the arene and not releasing it until all six D atoms have been 10202

dx.doi.org/10.1021/es302188f | Environ. Sci. Technol. 2012, 46, 10199−10205

Environmental Science & Technology

Article

transferred.46,47,52 Furthermore, as the distribution showed no preference for even numbers of D atoms, it suggests that the D atoms are not incorporated in an exclusively pairwise fashion. A distribution of deuterated cyclohexane products was also observed for the degradation of benzene and fluorobenzene with Rh/Al2O3 and D2 in H2O, confirming the reversibility of hydrogenation. The maximum number of deuterium atoms incorporated was 9 for benzene and fluorobenzene, and the fitted means were 1.5 and 1.9 deuterium atoms for benzene and fluorobenzene, respectively (Figure 3b,c). Saturated Substances Are Unreactive. Fluorocyclohexane did not react with Rh/Al2O3 and D2/H2O; this result is consistent with the results of the experiments conducted using unlabeled materials. More importantly, no D atom incorporation was observed. Similarly, no D incorporation was observed when cyclohexane was treated with Rh/Al2O3 and D2/H2O. This confirmed that the two substances are unreactive under the reaction conditions and do not undergo dehydrogenation/hydrogenation. The last hydrogenation step leading from cyclohexene to cyclohexane must be irreversible (see also Figure 1). Defluorination Precedes Hydrogenation. Two observations from the isotope labeling experiments indicate that defluorination of fluorobenzene occurs prior to hydgrogenation. In the treatment of fluorobenzene with Rh/Al2O3 and D2/H2O, the deuteration pattern of the observed benzene intermediate corresponded to a very narrow Gaussian distribution (σ2 = 0.6) with a mean of zero D atoms. About 93% of the benzene contained zero or one deuterium atom. This narrow distribution indicates that the majority of the observed benzene has been produced by the first defluorination step and has not undergone hydrogenation/dehydrogenation yet. Furthermore, in the batch experiment of fluorobenzene with Rh/Al2O3 and D2/H2O, no incorporation of deuterium into fluorobenzene was observed during the course of the reaction. By contrast, small amounts of deuterated benzenes were detected when benzene was used as the starting material in the catalytic system. This suggests that benzene can undergo reversible hydrogenation/dehydrogenation, whereas fluorobenzene must be defluorinated before hydrogenation/dehydrogenation can be observed. These observations are consistent with a model in which hydrodefluorination occurs before hydrogenation and in which the defluorination step is not reversible. Tests for Reversibility of Defluorination. To examine the reversibility of the defluorination step from fluorobenzene to benzene, two additional tests were conducted. First, 0.1 M sodium fluoride (NaF) was added to the catalytic reaction of benzene with Rh/Al2O3. If the defluorination reaction were reversible, fluorobenzene could be formed from benzene and fluoride (F−). Second, 4-fluorotoluene (100 μM), which is rapidly degraded to methylcyclohexane under the batch reaction conditions, was added to the catalytic reaction of benzene to see whether fluorobenzene could be formed from benzene and the fluoride ion freshly abstracted from fluorotoluene. The production of fluorobenzene was not observed in either of the two tests. These results further indicate that the hydrodefluorination step is not reversible. Thermodynamic calculations also support an irreversible defluorination step, as the reaction step from fluorobenzene to benzene is highly exothermic (ΔH° = −148 kJ mol−1), whereas the subsequent reversible hydrogenation reaction steps exhibit smaller reaction enthalpies (see the Supporting Information).

Kinetic Fitting. Additional information about the degradation pathway of fluorobenzene was gained from fitting the timecourse data of the degradation of fluorobenzene and the formation of the major intermediate benzene, as well as the end product cyclohexane, to a kinetic model. Cyclohexene was not included because it was observed in amounts lower than the limit of quantification. Several models were tested for the fitting. A model with two consecutive steps (i.e., conversion of fluorobenzene to benzene and conversion of benzene to cyclohexane) did not represent the data very well. A model allowing for the two-step consecutive pathway and an additional direct pathway from fluorobenzene to cyclohexane also did not provide a satisfactory fit. The best fit model incorporated an unknown intermediate (I) from which both the two-step pathway (through benzene) and a direct pathway led to cyclohexane (Figure 4a). The conversion of fluoroben-

Figure 4. (a) Kinetic model for the degradation of fluorobenzene to cyclohexane via an unknown intermediate (I) that best represents the batch experimental data. Arrows correspond to pseudo-first-order degradation. (b) Degradation of fluorobenzene (○), formation and degradation of the intermediate benzene (◆), and formation of cyclohexane (■) in a batch experiment with a fit to the kinetic model (dotted lines).

zene (C6H5F) to cyclohexane (C6H12) was assumed to be 100%, with a fraction α of I taking the consecutive pathway through benzene (C6H6) and the other fraction (1 − α) taking the direct pathway. The best fit α value averaged from four batch experiments was 0.78 ± 0.10. All parameters of the differential equation system representing the model (eqs 2−5) were simultaneously fitted using AQUASIM.44 d[C6H5F] = −kobs[C6H5F] dt

(2)

d[I ] = kobs[C6H5F] − k1[I ] dt

(3)

d[C6H6] = αk1[I] − k 2[C6H6] dt

(4)

d[C6H12] = (1 − α)k1[I] + k 2[C6H6] dt

(5)

The good fit of the experimental data to this model (Figure 4b) and the missing part in the carbon mass balance during the 10203

dx.doi.org/10.1021/es302188f | Environ. Sci. Technol. 2012, 46, 10199−10205

Environmental Science & Technology



first 20 min of the batch experiments both support a surfacemediated reaction. I must represent all surface-bound species during the reaction of fluorobenzene to cyclohexane. As the fluorinated intermediates have been excluded from the mechanism, the surface-bound species are, namely, fluorobenzene, benzene, the cyclohexadienes, and cyclohexene. A portion (∼22%) of the reaction species stay bound to the surface and react to cyclohexane (direct pathway), while part of the benzene and cyclohexene desorb to the bulk phase before they resorb and react to form cyclohexane (consecutive pathway). Current Mechanistic Hypothesis. The current mechanistic hypothesis for the hydrodefluorination/hydrogenation reaction of fluorobenzene with Rh/Al2O3 and H2 in water is depicted in Figure 5. Fluorobenzene first sorbs to the Rh

Article

ASSOCIATED CONTENT

S Supporting Information *

Thermodynamic calculations, LOD and LOQ values, matrix approach description, a summary of the fluorinated intermediates, and synthetic pathways. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank Dr. Michael Plötze (ETH Zurich) for conducting the BET and particle size measurements, Dr. Sarah Kliegman (ETH Zurich) for experimental support and helpful comments, and Prof. Peter J. Alaimo (Seattle University) for helpful comments on the manuscript.



REFERENCES

(1) Key, B. D.; Howell, R. D.; Criddle, C. S. Fluorinated organics in the biosphere. Environ. Sci. Technol. 1997, 31, 2445. (2) Banks, R. E. Fluorine chemistry at the millenium. Fascinated by Fluorine; Elsevier: Amsterdam, The Netherlands, 2000. (3) Hiyama, T.; Kusumoto, T.; Morizawa, Y.; Shimizu, M. OrganoFluorine Compounds: Chemistry and Applications; Springer: Berlin, Germany, 2000. (4) Adams, D. E. C.; Halden, R. U. Fluorinated chemicals and the impacts of anthropogenic use. ACS Symp. Ser. 2010, 1048, 539. (5) Aizenberg, M.; Milstein, D. Catalytic activation of carbon-fluorine bonds by a soluble transition metal complex. Science 1994, 265, 359. (6) Amii, H.; Uneyama, K. C-F bond activation in organic synthesis. Chem. Rev. 2009, 109, 2119. (7) Douvris, C.; Ozerov, O. V. Hydrodefluorination of perfluoroalkyl groups using silylium-carborane catalysts. Science 2008, 321, 1188. (8) Bryant, R. Pharmaceutical Fine Chemicals. Global Perspectives 2000; Informa Chemicals Industry Report; Informa Publishing Group: Kent, U.K., 2000. (9) O’Hagan, D. Fluorine in health care: Organofluorine containing blockbuster drugs. J. Fluorine Chem. 2010, 131, 1071. (10) Soloshonok, V. A.; Mikami, K.; Yamazaki, T.; Welch, J. T.; Hoenk, J. F. Current Fluoroorganic Chemistry: New Synthetic Directions, Technologies, Materials, and Biological Applications (ACS Symposium); American Chemical Society: Washington, DC, 2006. (11) High Production Volume (HPV) ChallengeRobust Summaries and Test Plans (accessed on July 31, 2012). http://www.epa.gov/ chemrtk/pubs/summaries/viewsrch.htm. (12) Robust Summaries and Test Plans: Fluorobenzene (accessed on July 31, 2012). http://www.epa.gov/chemrtk/pubs/summaries/ flurbenz/c14602tc.htm. (13) Boyd, E. M.; Killham, K.; Wright, J.; Rumford, S.; Hetheridge, M.; Cumming, R.; Meharg, A. A. Toxicity assessment of xenobiotic contaminated groundwater using lux modified Pseudomonas fluorescens. Chemosphere 1997, 35, 1967. (14) Rayne, S.; Forest, K. Perfluoroalkyl sulfonic and carboxylic acids: A critical review of physicochemical properties, levels and patterns in waters and wastewaters, and treatment methods. J. Environ. Sci. Health, Part A 2009, 44, 1145. (15) Travkin, V.; Solyanikova, I.; Rietjens, I.; Vervoort, J.; Van Berkel, W.; Golovleva, L. Degradation of 3,4-dichloro- and 3,4-difluoroaniline by Pseudomonas fluorescens 26-K. J. Environ. Sci. Health, Part B 2003, 38, 121. (16) Reade, S. P.; Mahon, M. F.; Whittlesey, M. K. Catalytic hydrodefluorination of aromatic fluorocarbons by ruthenium Nheterocyclic carbene complexes. J. Am. Chem. Soc. 2009, 131, 1847.

Figure 5. Current mechanistic hypothesis of fluorobenzene degradation, including the sorption/desorption of reactant, intermediates, and end product to the catalyst surface.

surface. On the surface, an irreversible hydrodefluorination step to benzene is followed by reversible hydrogenation steps and a last irreversible hydrogenation step to cyclohexane. When cyclohexane is formed on the surface, it is released to the bulk phase, where it is stable. During the hydrogenation reaction, the intermediates benzene and cyclohexene exchange between the surface and the bulk phase. Although the pathway involving fluorinated intermediates could not be completely ruled out on the basis of degradation experiments with the postulated intermediates, their formation was finally excluded from the current mechanistic hypothesis on the basis of isotope labeling experiments that strongly indicated that defluorination occurs prior to hydrogenation. Environmental Significance. In this work, the degradation of fluorobenzene under mild aqueous conditions has been demonstrated. The ambient reaction temperature and low pressure of H2 make this method potentially viable for the remediation of fluorobenzene-contaminated groundwater, analogous to Pd-catalyzed remediation of chlorocarbonpolluted groundwater.38−40 Because defluorination is followed by hydrogenation to cyclohexane, benzene, a compound of serious toxicological concern, would not contaminate remediated natural waters. Further work to be performed includes establishing the range of fluorinated substrates that can be degraded by Rh/Al2O3 with H2 in water, as well as the reaction’s sensitivity to common constituents in natural waters that could act as inhibitors. 10204

dx.doi.org/10.1021/es302188f | Environ. Sci. Technol. 2012, 46, 10199−10205

Environmental Science & Technology

Article

(17) Young, R. J.; Grushin, V. V. Catalytic C-F bond activation of nonactivated monofluoroarenes. Organometallics 1999, 18, 294. (18) Ge, L.; Chen, J.; Wei, X.; Zhang, S.; Qiao, X.; Cai, X.; Xie, Q. Aquatic photochemistry of fluoroquinolone antibiotics: Kinetics, pathways, and multivariate effects of main water constituents. Environ. Sci. Technol. 2010, 44, 2400. (19) Ravichandran, L.; Selvam, K.; Swaminathan, M. Highly efficient activated carbon loaded TiO2 for photo defluoridation of pentafluorobenzoic acid. J. Mol. Catal. A 2010, 317, 89. (20) Carvalho, M. F.; Alves, C. C. T.; Ferreira, M. I. M.; De Marco, P.; Castro, P. M. L. Isolation and initial characterization of a bacterial consortium able to mineralize fluorobenzene. Appl. Environ. Microbiol. 2002, 68, 102. (21) Carvalho, M. F.; Ferreira, M. I. M.; Moreira, I. S.; Castro, P. M. L.; Janssen, D. B. Degradation of fluorobenzene by Rhizobiales strain F11 via ortho cleavage of 4-fluorocatechol and catechol. Appl. Environ. Microbiol. 2006, 72, 7413. (22) Li, H.; Liu, Y. H.; Luo, N.; Zhang, X. Y.; Luan, T. G.; Hu, J. M.; Wang, Z. Y.; Wu, P. C.; Chen, M. J.; Lu, J. Q. Biodegradation of benzene and its derivatives by a psychrotolerant and moderately haloalkaliphilic Planococcus sp. strain ZD22. Res. Microbiol. 2006, 157, 629. (23) Blum, J.; Rosenfeld, A.; Gelman, F.; Schumann, H.; Avnir, D. Hydrogenation and dehalogenation of aryl chlorides and fluorides by the sol-gel entrapped RhCl3-Aliquat 336 ion pair catalyst. J. Mol. Catal. A 1999, 146, 117. (24) Buil, M. L.; Esteruelas, M. A.; Niembro, S.; Oliván, M.; Orzechowski, L.; Pelayo, C.; Vallribera, A. Dehalogenation and hydrogenation of aromatic compounds catalyzed by nanoparticles generated from rhodium bis(imino)pyridine complexes. Organometallics 2010, 29, 4375. (25) Desmarets, C.; Kuhl, S.; Schneider, R.; Fort, Y. Nickel(0)/ imidazolium chloride catalyzed reduction of aryl halides. Organometallics 2002, 21, 1554. (26) Klahn, M.; Fischer, C.; Spannenberg, A.; Rosenthal, U.; Krossing, I. Hydrodefluorination of non-activated C-F bonds by diisobutyl-aluminiumhydride via the aluminium cation [i-Bu2Al]. Tetrahedron Lett. 2007, 48, 8900. (27) Stanger, K. J.; Angelici, R. J. Hydrodefluorination of fluorobenzene catalyzed by rhodium metal prepared from [Rh(COD)2]+BF4− and supported on SiO2 and Pd-SiO2. J. Mol. Catal. A 2004, 207, 59. (28) Ukisu, Y.; Miyadera, T. Hydrogen-transfer hydrodehalogenation of aromatic halides with alcohols in the presence of noble metal catalysts. J. Mol. Catal. A 1997, 125, 135. (29) Yang, H.; Gao, H. R.; Angelici, R. J. Hydrodefluorination of fluorobenzene and 1,2-difluorobenzene under mild conditions over rhodium pyridylphosphine and bipyridyl complexes tethered on a silica-supported palladium catalyst. Organometallics 1999, 18, 2285. (30) Braun, T.; Wehmeier, F. C-F bond activation of highly fluorinated molecules at rhodium: From model reactions to catalysis. Eur. J. Inorg. Chem. 2011, 5, 613. (31) Perutz, R. N.; Braun, T. In Comprehensive Organometallic Chemistry III; Crabtree, R. H., Mingos, D. M. P., Eds.; Elsevier Ltd.: Amsteram, The Netherlands, 2007; Vol. 1, p 725. (32) Lowry, G. V.; Reinhard, M. Hydrodehalogenation of 1- to 3carbon halogenated organic compounds in water using a palladium catalyst and hydrogen gas. Environ. Sci. Technol. 1999, 33, 1905. (33) Lowry, G. V.; Reinhard, M. Pd-catalyzed TCE dechlorination in groundwater: Solute effects, biological control, and oxidative catalyst regeneration. Environ. Sci. Technol. 2000, 34, 3217. (34) Chaplin, B. P.; Shapley, J. R.; Werth, C. J. Regeneration of sulfur-fouled bimetallic Pd-based catalysts. Environ. Sci. Technol. 2007, 41, 5491. (35) Hildebrand, H.; Mackenzie, K.; Kopinke, F.-D. Pd/Fe3O4 nanocatalysts for selective dehalogenation in wastewater treatment processesInfluence of water constituents. Appl. Catal., B 2009, 91, 389.

(36) Angeles-Wedler, D.; Mackenzie, K.; Kopinke, F.-D. Sulphideinduced deactivation of Pd/Al2O3 as hydrodechlorination catalyst and its oxidative regeneration with permanganate. Appl. Catal., B 2009, 90, 613. (37) Heck, K. N.; Nutt, M. O.; Alvarez, P.; Wong, M. S. Deactivation resistance of Pd/Au nanoparticle catalysts for water-phase hydrodechlorination. J. Catal. 2009, 267, 97. (38) Davie, M. G.; Cheng, H.; Hopkins, G. D.; Lebron, C. A.; Reinhard, M. Implementing heterogeneous catalytic dechlorination technology for remediating TCE-contaminated groundwater. Environ. Sci. Technol. 2008, 42, 8908. (39) Schuth, C.; Kummer, N. A.; Weidenthaler, C.; Schad, H. Field application of a tailored catalyst for hydrodechlorinating chlorinated hydrocarbon contaminants in groundwater. Appl. Catal., B 2004, 52, 197. (40) McNab, W. W.; Ruiz, R.; Reinhard, M. In-situ destruction of chlorinated hydrocarbons in groundwater using catalytic reductive dehalogenation in a reactive well: Testing and operational experiences. Environ. Sci. Technol. 2000, 34, 149. (41) Aramendia, M. A.; Borau, V.; Carcia, I. M.; Jimenez, C.; Marinas, A.; Marinas, J. M.; Urbano, F. J. Hydrogenolysis of aryl halides by hydrogen gas and hydrogen transfer over palladium-supported catalysts. C. R. Acad. Sci., IIC 2000, 3, 465. (42) Cellier, P. P.; Spindler, J. F.; Taillefer, M.; Cristau, H. J. Pd/Ccatalyzed room-temperature hydrodehalogenation of aryl halides with hydrazine hydrochloride. Tetrahedron Lett. 2003, 44, 7191. (43) Mackenzie, K.; Frenzel, H.; Kopinke, F. D. Hydrodehalogenation of halogenated hydrocarbons in water with Pd catalysts: Reaction rates and surface competition. Appl. Catal., B 2006, 63, 161. (44) Reichert, P. Concepts Underlying a Computer Program for the Identification and Simulation of Aquatic Systems; Swiss Federal Institute for Environmental Science and Technology: Dübendorf, Switzerland, 1994. (45) Landis, C. R.; Halpern, J. Homogeneous catalysis of arene hydrogenation by cationic rhodium arene complexes. Organometallics 1983, 2, 840. (46) Russell, M.; White, C.; Maitlis, P. Stereoselective homogeneous hydrogenation of arenes to cyclohexanes catalysed by [Rh(η5C5Me5)Cl2]2. J. Chem. Soc., Chem. Commun. 1977, 427. (47) Widegren, J. A.; Finke, R. G. A review of soluble transition-metal nanoclusters as arene hydrogenation catalysts. J. Mol. Catal. A 2003, 191, 187. (48) Blum, J.; Amer, I.; Zoran, A.; Sasson, Y. Catalytic hydrogenation of olefins, acetylenes and arenes by rhodium trichloride and Aliquat336 under phase-transfer conditions. Tetrahedron Lett. 1983, 24, 4139. (49) Fache, F.; Lehuede, S.; Lemaire, M. A catalytic stereoselective and chemoselective method for the reduction of substituted aromatics. Tetrahedron Lett. 1995, 36, 885. (50) Januszkiewicz, K. R.; Alper, H. Exeedingly mild, selective and stereospecific phase-transfer-catalyzed hydrogenation of arenes. Organometallics 1983, 2, 1055. (51) Landre, P. D.; Richard, D.; Draye, M.; Gallezot, P.; Lemaire, M. Colloidal rhodiumA new catalytic system for the reduction of dibenzo-18-crown-6 ether. J. Catal. 1994, 147, 214. (52) Muetterties, E. L.; Rakowski, M. C.; Hirsekorn, F. J.; Larson, W. D.; Basus, V. J.; Anet, F. A. L. Hydrogenation of arenes with discrete coordination catalysts. III. Synthesis and nuclear magnetic resonance spectrum of all-cis-cyclohexane-d6. J. Am. Chem. Soc. 1975, 97, 1266.

10205

dx.doi.org/10.1021/es302188f | Environ. Sci. Technol. 2012, 46, 10199−10205