Environ. Sci. Technol. 2010, 44, 3819–3824
Benzene Photolysis on Ice: Implications for the Fate of Organic Contaminants in the Winter T A R A F . K A H A N †,§ A N D D . J . D O N A L D S O N * ,†,‡ Department of Chemistry, University of Toronto, 80 Saint George Street, Toronto, Ontario, Canada M5S 3H6, and Department of Physical and Environmental Sciences, University of Toronto, 1265 Military Trail, Toronto, Ontario, Canada M1C 1A4
Received February 9, 2010. Revised manuscript received April 7, 2010. Accepted April 12, 2010.
The members of the important class of organic pollutants known as BTEX (benzene, toluene, ethylbenzene, and xylenes) do not undergo direct photolysis in natural waters, because their absorption spectra do not overlap that of the solar radiation which reaches the Earth’s surface. Recent work has shown that aromatic compounds undergo significant red-shifts in their absorption spectra when they are present at air-ice interfaces, suggesting that BTEX components could undergo direct photolysis at ice surfaces. Using glancing-angle laser-induced fluorescence (LIF) as a probe, we measured benzene photodegradation at λ > 295 nm having a rate constant of (3 ( 1 × 10-4 s-1) under our experimental conditions. We predict that the photolysis rate at environmental ice surfaces will be similar, based on the photon flux dependence we measured. This study presents the first report of direct benzene photolysis under environmentally relevant conditions. The results suggest that direct photolysis could be an important removal pathway for organic pollutants such as BTEX in snow-covered regions, for example, in polar or urban areas contaminated by oil spills or leaks.
Introduction Organic contaminants such as BTEX (benzene, toluene, ethylbenzene, and xylenes) are considered to be important pollutants due to their health effects. They can have high environmental concentrations due to oil leaks and spills, as well as long-range transport. For example, benzene concentrations in excess of 100 µg L-1 (1.4 × 10-6 mol L-1) have been reported in contaminated surface waters and groundwater (1-4). In rural and semirural snowpacks, benzene concentrations on the order of 30 µg L-1 have been reported (5), with total BTEX concentrations ranging from 30 to 300 µg L-1 (5, 6). The photolysis of several aromatic pollutants by actinic radiation has been observed in ice (7-14). As expected, these compounds all absorb in the actinic region and, so, also undergo photolysis in aqueous solution. However, as shown in Figure 1, the aqueous absorption spectra of individual * Corresponding author e-mail:
[email protected]; phone: (416) 978-3603; fax: (416) 978-8775. † Department of Chemistry. ‡ Department of Physical and Environmental Sciences. § Present address: Department of Chemistry, University of CaliforniasIrvine, Irvine, CA 92697. 10.1021/es100448h
2010 American Chemical Society
Published on Web 04/27/2010
BTEX components do not overlap the expected solar intensity in Toronto, Canada, in June (15). It is clear from the figure that BTEX compounds should not absorb solar radiation at Earth’s surface; therefore, direct photolysis has not been considered to be a viable removal pathway for these contaminants in the environment. Reaction with hydroxyl radicals is often the dominant removal pathway for organic pollutants in natural waters (16). However, when aromatics are present at the surface of frozen water (ice), this reaction pathway may not occur: Recent work in our laboratory (8, 17) has shown that hydroxyl radicals do not react to an appreciable extent with aromatic compounds, including benzene, at air-ice interfaces, whereas in aqueous solution these reactions occur at near diffusion-limited rates. On the basis of these results, one would predict that BTEX components should not undergo chemical transformation (at least, via reaction with OH) on snow and ice surfaces. In the following, we present evidence to suggest that direct photolysis of these compounds could be important there. This hypothesis is based on our previous observation (17) that aromatic compounds such as benzene and phenol undergo significant red-shifts in their absorption spectra when they are present at ice surfaces. Figure 2 shows excitation spectra of neat benzene, dilute aqueous benzene, and benzene present at air-ice interfaces after being frozen out of solution. Both neat benzene and benzene at ice surfaces absorb longer wavelength radiation than does benzene in aqueous solution. Importantly, their absorbance overlaps with the solar output present at Earth’s surface in the winter, also shown in the figure. This suggests that benzene and other BTEX components could undergo direct photolysis at ice surfaces. In this work, we measured benzene photolysis kinetics at air-ice interfaces to determine whether direct photolysis of BTEX compounds could occur on ice under atmospherically relevant conditions.
Experimental Section Aqueous solutions containing benzene (ACP, g99%) in 18 MΩ · cm deionized water were prepared daily; concentrations ranged from 10-6 to 10-3 mol L-1. Ice samples were prepared by placing approximately 0.8 mL of the aqueous benzene solution on a small stainless steel plate inside a Teflon reaction chamber (8). The copper floor of the chamber was maintained at approximately -15 °C by copper cooling tubes running underneath the floor. After freezing, the samples were inverted to present a flat surface for spectroscopic interrogation. Photolysis of aqueous samples was performed in 10 mm path length quartz cuvettes, and ice samples were photolyzed in the reaction chamber. Samples were irradiated with the output of a 100 W xenon arc lamp, which passed through a 295 nm long-pass cutoff filter. The lamp’s output was directed onto the sample by a first-surface mirror with maximum reflectance at 355 nm. For ice samples, the lamp’s output entered the reaction chamber through a quartz window in the chamber roof. The distance between the lamp and the sample was ∼90 cm for irradiation of both aqueous and ice samples. The lamp power was measured by a power meter using an energy-absorbing head optimized between 250 and 400 nm. For most experiments, the measured power was ∼0.34 W. For some experiments, the power was decreased by placing wire mesh between the lamp and the sample to determine the photon flux dependence for the reaction (9). Measured lamp powers for these experiments ranged from 0.023 to 0.33 W. VOL. 44, NO. 10, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3819
FIGURE 1. Absorbance spectra of aqueous solutions: (left-hand y-axis) 1.8 × 10-2 mol L-1 benzene (solid red trace), 1.2 × 10-3 mol L-1 toluene (long-dashed green trace), 1.6 × 10-3 mol L-1 ethylbenzene (short-dashed blue trace), and 1.6 × 10-3 mol L-1 p-xylene (dotted gray trace); black dash-dotted trace is the average solar irradiance at Earth’s surface for Toronto, Canada, in midsummer at noon (right-hand y-axis).
FIGURE 2. Excitation spectra (left-hand y-axis) of 10-3 mol L-1 benzene in aqueous solution (solid blue trace), the surface of a frozen 10-5 mol L-1 benzene solution (red dashed trace), and neat benzene (dotted green trace). The black dash-dotted trace is the average solar irradiance at Earth’s surface for Toronto, Canada, in midwinter at noon (right-hand y-axis). Excitation spectra of benzene were acquired using laserinduced fluorescence (LIF) for aqueous samples and glancing-angle LIF (8) for ice samples, using a Nd:YAG-pumped optical parametric oscillator (OPO) as the excitation source. The laser output entered the reaction chamber through a quartz window in the front of the chamber at a shallow angle (>85° from the surface normal) and impinged upon the sample surface. A liquid light guide suspended approximately 10 mm above the sample surface collected fluorescence, which was passed through a monochromator and detected by a photomultiplier tube. The resulting signal was imaged on a digital oscilloscope, and a 200 ns section of the fluorescence decay was captured and stored for analysis. The excitation wavelength was scanned in 2 nm increments from 245 to 271 nm for aqueous samples and to 320 nm for ice samples and for neat benzene. Emission was monitored at 318 nm for aqueous samples and 344 nm for ice samples. Benzene photolysis kinetics were determined by monitoring benzene emission intensity during irradiation with excitation 3820
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 10, 2010
at 261 nm and emission at 318 or 344 nm for aqueous and ice samples, respectively. The wavelengths monitored correspond to strong emission bands from benzene in aqueous solution and on ice, respectively, as discussed in ref 18. The laser output was blocked between data acquisitions to avoid inducing photolysis from the 261 nm output (17). Absorption spectra of 1.8 × 10-2 mol L-1 benzene, 1.9 × -3 10 mol L-1 toluene (EMD USA, 99.98%), 1.6 × 10-3 mol L-1 ethylbenzene (Fisher, reagent grade), and 1.6 × 10-3 mol L-1 p-xylene (Caledon, 98%) in 18 MΩ · cm deionized water were acquired on a commercial UV-vis spectrometer.
Results and Discussion Spectra. As discussed in the Introduction and shown in Figure 2, the absorption spectrum of benzene at air-ice interfaces is similar to that of neat benzene, but is distinctly red-shifted compared to benzene in aqueous solution. The shape of the excitation spectra at air-ice interfaces was independent of the benzene concentration in solution prior to freezing, in
FIGURE 3. Log-normal benzene emission intensity vs. irradiation time in aqueous solution (solid blue circles) and at an air-ice interface (open red squares) during irradiation with the 295 nm filtered output of a 100 W xenon arc lamp. the range from 10-6 to 10-3 mol L-1. This suggests that even at the lowest initial concentrations, benzene undergoes significantly enhanced self-association (causing its spectrum to appear similar to that of neat benzene) at air-ice interfaces compared to in aqueous solution, as we have previously reported for polycyclic aromatic hydrocarbons (PAHs) (8, 19). This enhanced absorption in the actinic region of the spectrum could give rise to photochemical processes on ice surfaces. Photolysis Kinetics in Aqueous Solution and at Air-Ice Interfaces. Figure 3 shows benzene emission intensity as a function of irradiation time in aqueous solution and at an air-ice interface. No decrease in intensity was observed for the aqueous sample, but at air-ice interfaces benzene intensity decreased during photolysis. The data were well-fit by a single exponential decay, indicating that the reaction was first order with respect to benzene concentration. Varying the benzene concentration in the original solution between 9.0 × 10-6 and 2.7 × 10-4 mol L-1 did not affect the kinetics. No benzene loss was observed in the dark, and the reaction chamber was maintained at a constant temperature during irradiation, so the observed intensity loss during irradiation is not likely due to sample heating. The lack of reactivity in the dark also suggests that the laser was not responsible for benzene loss, through either heating or chemistry. This result supports the suggestion that the red-shift in benzene’s absorption spectrum on ice allows it to undergo direct photolysis at wavelengths relevant to Earth’s surface. Neat benzene, which also absorbs at wavelengths above 290 nm, has also been reported to undergo direct photolysis under solar radiation (20). Phenol is reported to be the major product of the direct photolysis of neat benzene by sunlight (20). We have previously observed phenol formation at air-ice interfaces from the direct photolysis of benzene at wavelengths below 295 nm (17). In those studies, phenol growth was indicated by an increase in emission intensity in the region between 300 and 330 nm, as illustrated in Figure 4a. Figure 4b shows emission spectra of a frozen benzene solution at an air-ice interface before and after irradiation by the filtered (λ g 295 nm) output of the xenon arc lamp. There is a clear decrease in emission intensity at higher wavelengths, consistent with the loss of benzene illustrated in Figure 3, but no growth in intensity at wavelengths below 330 nm, which would indicate phenol formation. It appears that although the absorbance
spectra of neat benzene and frozen aqueous solutions containing benzene are similar, the products formed from their photolysis at environmentally relevant wavelengths are different. Indeed, HPLC analysis of melted samples after photolysis shows evidence for the formation of product(s) at longer elution times than benzene, perhaps from photopolymerization of the parent compound. No further attempt to identify products was undertaken in this work, however. Photon Flux Dependence. The photon flux of the lamp over the wavelength range 295-410 nm used in this work is approximately 3 orders of magnitude lower than that of solar radiation at Earth’s surface in Toronto, Canada, at noon in the winter (9). One might expect, therefore, that the photolysis rate in the environment would be 1000 times more rapid than the rate measured in our laboratory at full lamp power [(3 ( 1) × 10-4 s-1]. However, we have recently reported (9) that PAH photolysis depends nonlinearly on photon flux both in aqueous solution and at air-ice interfaces. Figure 5 shows measured benzene photolysis rate constants at air-ice interfaces as a function of lamp power. The vertical arrows indicate total photon fluxes between 295 and 410 nm, as determined by nitrite actinometry (17). Above ∼0.1 W, the observed rate constant becomes independent of lamp power. This supports our previous suggestion (9) that at high photon fluxes, the absorption of photons by strong absorbers such as aromatic species may cease to be the rate-limiting step of photolysis, resulting in an observed saturation in photolysis kinetics at high photon fluxes. It also indicates that extrapolating laboratory-measured kinetics to those expected in the environment by scaling for the differences in photon fluxes may result in underpredictions of benzene lifetimes in the environment. Atmospheric Implications. The benzene concentrations used in this study are representative of contaminated environmental sites (1-6). Therefore, photolysis could be an important transformation route for benzene and other BTEX components at the surface of contaminated snow and ice. On the basis of our measured kinetic data, we calculated expected lifetimes toward photolysis for benzene at the surface of snow and ice in Toronto in midwinter and in the Arctic in midsummer. Photolysis rate constants are given as J)
∫ σ(λ)φ(λ)F λ
λ0
λ
dλ
VOL. 44, NO. 10, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
(1)
9
3821
where σ gives the absorption cross section of the molecule, φ is the photolysis quantum yield, and Fλ represents the photon flux. The first two variables should be the same in the laboratory and in the environment, but Fλ can differ by orders of magnitude. Therefore, predicting environmental photolysis lifetimes is usually accomplished by multiplying the rate constant measured in the laboratory by an appropriate factor to account for differences in the photon flux from the lamp used in the laboratory experiments and the photon flux of the sun:
( )
Jenv ) Jlab
Fenv Flab
(2)
The subscripts “env” and “lab” denote “environment” and “laboratory”, respectively. From Figure 5, we see that a linear extrapolation of photon flux from our lamp to that of the sun’s output would not be appropriate in the case of benzene. We therefore predict the lifetime of benzene using the fit to the data in Figure 5
J)
a × Fc b + Fc
(3)
where a and b are constants obtained from the fit to the data and c is a conversion factor relating lamp power to photon flux in our experiments. The values for these constants are given in Table 1. The photon flux at noon in the Arctic in late June, calculated using the TUV model (15), is ∼3 × 1016 photon cm-2 s-1 over the wavelength range of 290-400 nm. This gives us a lifetime toward photolysis of ∼1 h for benzene at the surface of snow and ice. As expected, this lifetime is the same as that observed in our laboratory, despite a 1000-fold increase in the photon flux. Some caution is required in predicting environmental lifetimes of benzene and other BTEX compounds in snow and ice. Although we measure rapid degradation kinetics at air-ice interfaces, it is likely that photolysis within ice, that is, benzene that is trapped in liquid pockets or veins within the ice bulk, will show a similar lack of reactivity as observed for benzene in aqueous solution. We have recently dem-
FIGURE 4. Emission spectra of the surface of a frozen 10-5 mol L-1 benzene solution (solid blue trace) and of the same sample after irradiation (dashed red trace) by (a) the 261 nm output of the OPO and (b) the 295 nm filtered output of a 100 W xenon arc lamp. The sharp peaks around ∼285 nm are due to Raman scattering from the ice. Figure 4a: reprinted with permission from ref 17. Copyright 2009 American Chemical Society. 3822
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 10, 2010
FIGURE 5. Observed benzene photolysis rate constants at air-ice interfaces as a function of lamp power. Error bars represent the standard deviation about the mean for at least three trials. The trace shows a hyperbolic fit to the data and is included to guide the reader’s eye. The arrows indicate total photon fluxes over the wavelength range 295-410 nm at various lamp powers, based on nitrite actinometry (17); the fluxes are in units of photon cm-2 s-1.
TABLE 1. Values of the Constants in Equation 3 constant a b c
important transformation process for BTEX components that are associated with a range of environmental surfaces.
value -4
2.7 × 10 s-1 2.6 × 10-2 photon cm-2 s-1 2.4 × 10-13
onstrated (9, 17) that aromatic species react at different rates in bulk ice than at air-ice interfaces. The direct photolysis of anthracene (9) and the reaction between benzene and hydroxyl radicals (17) occur at similar rates in aqueous solution and in bulk ice, but the rates of both reactions are very different at air-ice interfaces: They are enhanced in the former reaction and suppressed in the latter. It is therefore likely that benzene’s lifetime toward photolysis in ice will depend strongly on its distribution between the bulk and the surface. The type of reactions that benzene undergoes will also depend on its distribution between the bulk and the surface. For example, in bulk ice, reactions with OH may be the most important transformation pathway for benzene, but at air-ice interfaces, where OH is unreactive toward benzene (17), other reactions, including direct photolysis, will likely dominate. It is worth reiterating that the products formed from these reactions will be different: benzene in bulk ice reacts with OH to form phenol, but its direct photolysis on ice forms other, as yet unknown, products. The presence of contaminants other than benzene, either inorganic or organic, could affect the degree to which benzene self-associates at the air-ice interface, and therefore its photolability in the environment might be different from that measured in our experiments. This could be tested by measuring benzene photolysis kinetics on more chemically complex ice samples or even on actual environmental ice samples. Finally, the photolysis of benzene and other aromatic contaminants could also occur at surfaces other than ice; aromatic compounds such as PAHs have been reported to self-associate at surface coverages of well under a monolayer on surfaces such as silica and alumina (21, 22). If this also occurs for BTEX components, then the resulting red-shift in their absorption spectra would likely make them photoactive. Therefore, we predict that direct photolysis could be an
Acknowledgments We thank NSERC and CFCAS for funding this research. T.F.K. thanks NSERC for a Canadian Graduate Doctoral Scholarship. We thank the anonymous referees for helpful comments on the manuscript.
Literature Cited (1) Landmeyer, J. E.; Chapelle, F. H.; Petkewich, M. D.; Bradley, P. M. Assessment of natural attenuation of aromatic hydrocarbons in groundwater near a former manufactured-gas plant, South Carolina, USA. Environ. Geol. 1998, 34, 279–292. (2) Memon, B. A.; Burston, M. R. Benzene contamination at a metal plating facility. Environ. Geol. 2005, 48, 430–436. (3) Westbrook, S. J.; Rayner, J. L.; Davis, G. B.; Clement, T. P.; Bjerg, P. L.; Fisher, S. T. Interaction between shallow groundwater, saline surface water and contaminant discharge at a seasonally and tidally forced estuarine boundary. J. Hydrol. 2005, 302, 255– 269. (4) Zemo, D. A. Sampling in the smear zone: evaluation of nondissolved bias and associated BTEX, MTBE, and TPH concentrations in ground water samples. Ground Water Monit. Rem. 2006, 26, 125–133. (5) Kos, G.; Ariya, P. A. Determination of a wide range of volatile and semivolatile organic compounds in snow by use of solidphase micro-extraction (SPME). Anal. Bioanal. Chem. 2006, 385, 57–66. (6) Kos, G.; Ariya, P. A. Volatile organic compounds in snow in the Quebec-Windsor corridor. J. Geophys. Res. 2010, 115, D01302. (7) Grannas, A. M.; Bausch, A. R.; Mahanna, K. M. Enhanced aqueous photochemical reaction rates after freezing. J. Phys. Chem. A 2007, 111, 11043–11049. (8) Kahan, T. F.; Donaldson, D. J. Photolysis of polycyclic aromatic hydrocarbons on water and ice surfaces. J. Phys. Chem. A 2007, 111, 1277–1285. (9) Kahan, T. F.; Zhao, R.; Jumaa, K. B.; Donaldson, D. J. Anthracene photolysis in aqueous solution and ice: photon flux dependence and comparison of kinetics in bulk ice and at the air-ice interface. Environ. Sci. Technol. 2010, 44, 1302-1306. (10) Abbatt, J. P. D. Interactions of atmospheric trace gases with ice surfaces: adsorption and reaction. Chem. Rev. 2003, 103, 4783– 4800. (11) Klan, P.; Holoubek, I. Ice (photo)chemistry. Ice as a medium for long-term (photo)chemical transformations-environmental implications. Chemosphere 2002, 46, 1201–1210. VOL. 44, NO. 10, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3823
(12) Klan, P.; Klanova, J.; Holoubek, I.; Cupr, P. Photochemical activity of organic compounds in ice induced by sunlight irradiation: the Svalbard project. Geophys. Res. Lett. 2003, 30, 46-41–46-44. (13) Klanova, J.; Klan, P.; Nosek, J.; Holoubek, I. Environmental ice photochemistry: monochlorophenols. Environ. Sci. Technol. 2003, 37, 1568–1574. (14) Ram, K.; Anastasio, C. Photochemistry of phenanthrene, pyrene, and fluoranthene in ice and snow. Atmos. Environ. 2009, 43, 2252–2259. (15) Madronich, S.; Flocke, S. J. The role of solar radiation in atmospheric chemistry. In Handbook of Environmental Chemistry; Boule, P., Ed.; Springer-Verlag: Heidelberg, Germany, 1998; pp 1-26. (16) Viskari, E. L.; Rekila, R.; Roy, S.; Lehto, O.; Ruuskanen, J.; Karenlampi, L. Airborne pollutants along a roadside: assessment using snow analyses and moss bags. Environ. Pollut. 199797, 153–160. (17) Ardura, D.; Kahan, T. F.; Donaldson, D. J. Self-association of naphthalene at the air-ice interface. J. Phys. Chem. A 2009, 113, 7353–7359.
3824
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 10, 2010
(18) Kahan, T. F.; Zhao, R.; Donaldson, D. J. Reactivity of hydroxyl radicals at air-ice interfaces. Atmos. Chem. Phys. 2010, 10, 843– 854. (19) Kahan, T. F.; Donaldson, D. J. Heterogeneous ozonation kinetics of phenanthrene at the air-ice interface. Environ. Res. Lett. 2008, 3, 045006. (20) Stella, L.; Seraglia, R.; Sturaro, A.; Rella, R.; Traldi, P. On the photochemical oxidation of benzene and its relevance at environmental level. Rapid Commun. Mass Spectrom. 2008, 22, 257–260. (21) Dabestani, R.; Ellis, K. J.; Sigman, M. E. Photodecomposition of anthracene on dry surfaces: products and mechanism. J. Photochem. Photobiol., A 1995, 86, 231–239. (22) Wilkinson, F.; Worrall, D. R.; Williams, S. L. Primary photochemical processes of anthracene adsorbed on silica gel. J. Phys. Chem. 1995, 99, 6689–6696.
ES100448H
