ARTICLE IN PRESS Atmospheric Environment 39 (2005) 1935–1944 www.elsevier.com/locate/atmosenv Study of reaction processes of furan and some furan derivatives initiated by Cl atoms B. Cabañas, F. Villanueva, P. Martı́n, M.T. Baeza, S. Salgado, E. Jiménez Departamento de Quı´mica-Fı´sica. Facultad de Ciencias Quı´micas, Universidad de Castilla-La Mancha. Avda. Camilo José Cela, 10, 13071, Ciudad-Real, Spain Received 5 August 2004; received in revised form 22 November 2004; accepted 9 December 2004 Abstract The reactions of chlorine (Cl) atoms with volatile organic compounds (VOCs) emitted into the atmosphere are of interest to understand the role of Cl in the marine and coastal chemistry. The rate coefficients for Cl-atom reactions with organic compounds are typically about one or two order of magnitude larger than the ones corresponding to OH reaction. We report here the first kinetic measurements of the reactions of atomic Cl with some VOCs: furan, 2methylfuran, 3-methylfuran, 2-ethylfuran, and 2,5-dimethylfuran. The reactions of atomic Cl with furan and its derivatives were studied at 29872 K and 1 atm pressure using a relative rate technique with GC-FID/MS detection of the organic compounds. Thionyl chloride and trichloroacetyl chloride were used as Cl-atom precursors since molecular Cl reacted in the dark with the studied compounds. The ratios of rate coefficients for Cl-atom reactions with furan and its derivatives relative to n-nonane were as follows: furan (0.4170.05); 2-methylfuran (0.8570.03); 3-methylfuran (0.8870.04); 2-ethylfuran (0.9570.05); and 2,5dimethylfuran (1.1770.06). Taking k(Cl+n-nonane) ¼ (4.8270.14)  1010 cm3 molecule1 s1 the absolute rate coefficients obtained (in units of 1010 cm3 molecule1 s1) were: furan (2.070.2); 2-methylfuran (4.170.2); 3methylfuran (4.270.3); 2-ethylfuran (4.670.3); and 2,5-dimethylfuran (5.770.3). All errors are 72s. The influence of the structure on the reactivity of these compounds and the atmospheric implications are discussed. r 2005 Elsevier Ltd. All rights reserved. Keywords: Furan; Chlorine-atoms; Rate coefficient; Relative method; Tropospheric chemistry 1. Introduction There has been long-standing interest in the possible role of halogen atoms as tropospheric oxidants (Singh and Kasting, 1988; Chatfield and Crutzen, 1990). The best evidence so far comes from measurements of alkanes and acetylene in Arctic surface air (Jobson et al., 1994) which indicate a sink in April (polar sunrise) consistent with oxidation by chlorine (Cl) atoms present Corresponding author. Fax: +34 926 295 318. E-mail address: Beatriz.Cabanas@uclm.es (B. Cabañas). at a concentration 1  104 atoms cm3. The source of the halogen oxidants is not well established but likely involves chemical production from sea salt accumulated on the ice over the polar night (Impey et al., 1999). Generation of halogen oxidants from sea salt would be of little interest for global tropospheric chemistry if they were confined to Arctic sunrise. However, measurements of hydrocarbons and non-radical Cl species in the marine boundary layer (MBL) at midlatitudes and in the tropics suggest that Cl atoms may be present at least occasionally at concentrations in the range of 104–105 atoms cm3 (Keene et al., 1990; Pszenny et al., 1352-2310/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2004.12.013 ARTICLE IN PRESS 1936 B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 1993; Singh et al., 1996; Wingenter et al., 1996; Spicer et al., 1998; Wingenter et al., 1999). Therefore, the sources of Cl atoms are still subject of investigation. Although the concentration of atomic Cl can be lower than the OH concentration, Cl reacts with some hydrocarbons faster than OH does, implying that Cl can play an important role in the atmospheric chemistry. Extensive kinetic and mechanistic studies on the oxidation reactions involving Cl atom are therefore required to properly describe the fate of VOCs in coastal regions as well as in the Arctic. The interaction of Cl atoms with VOCs can occur as sea breezes carry marine air masses inland. Recently, environmental studies in continental areas show that F, Cl, and S are the origin of pollutants. In this sense, an anthropogenic source of precursors of tropospheric Cl atoms results from atmospheric emissions by brick factories, local concentrations of HCl and chlorides as high as 40–200 ppm in the emission plume have been reported (Galán et al., 2002). Among volatile organic compounds that can react with Cl atoms, furan and its derivatives have to be considered. Furans are a kind of heterocyclic aromatics emitted into the atmosphere as primary anthropogenic pollutants from the combustion of fossil fuels, refuse, plants and, in particular, from biomass burning (Isidorov et al., 1985; Graedel et al., 1986; Knudsen et al., 1993; Andreae and Merlet, 2001). These compounds are also known to be the products of the photooxidation of hydrocarbons with the structure CH2QCH–CR1QCHR2, such as 1,3-butadiene, isoprene and 1,3-pentadiene (Ohta, 1984; Gu et al., 1985; Tuazon and Atkinson, 1990; Ruppert et al., 1992, Rasmussen and Khalil, 1988; Ruppert and Becker, 2000) and aromatic compounds such as toluene and o-xylene (Shepson et al., 1984). Most of these compounds are mainly biogenic in origin emitted by the ocean such as isoprene (Bonsang et al., 1992; Ratte et al., 1998). Heretofore, rate coefficients of furan and some derivatives have been reported in the literature for the reaction of OH radical (Atkinson et al., 1983; Bierbach et al., 1992; Grosjean and Williams, 1992), NO3 (Atkinson and Aschmann, 1985; Atkinson, 1989; Kind et al., 1996), Br and O3 (Bierbach et al., 1996, 1999), but there is no data on the reactivity of furan and alkylfurans with Cl atoms. In this context, the study of Cl-atoms reaction with a series of VOCs like furans, whose reactivity is not established, is proposed. So, in this work we report the first kinetic study of the reaction of Cl atoms with a series of furans under atmospheric conditions. The rate coefficients were used to establish the structure–reactivity relationship for Cl–furans reactions. In addition, the reactivity of furans with Cl atoms is compared with other important atmospheric oxidants and tropospheric fates of these compounds are discussed. 2. Experimental A relative kinetic technique was used to determine the rate coefficients for Cl-atom reaction with furan, 2methylfuran, 3-methylfuran, 2-ethylfuran, and 2,5-dimethylfuran. The experiments were carried out at 29872 K in a 200 L FEP Teflon reaction bag at 1 atm of total pressure of N2 or synthetic air. That bag was placed inside a rectangular cage with 4 fluorescent lamps (Philips TUV G13 36W) mounted on the walls. The Cl atom precursor was added together with the organic compound to test for a possible interference from dark reactions. In preliminary experiments, it was found that Cl2 could not be used as a Cl atom source because it reacts in the dark with the furans (furan and furan derivatives). As a result, photolysis at 254 nm of thionyl chloride (SOCl2) or trichloroacetyl chloride (CCl3COCl) was used to generate atomic Cl (Finlayson-Pitts et al., 1999). Photolysis was typically carried out in steps of 11 s using SOCl2 and 2 min using CCl3COCl, followed by turning off the lamps and sampling the reaction mixture. Total photolysis times were ranged from 1 to 22 min. The decay of the reactant and reference compound concentrations was followed using gas chromatography (GC) with mass spectrometry, MS, (Shimadzu GC-17A, MS-QP5000) or flame ionization detection, FID (Shimadzu GC-14A). The GC column (size: 30 m  0.32 mm  0.1 mm Teknokroma TRB-1701) was run either at 45 1C or temperature programmed from 45 1C (9 min) to 100 1C (3 min) at a rate of 10 1C per min. In the experiments with GC-MS the kinetics were followed using the single ion-monitoring (SIM) mode. However previously, a run was performed in the scanning mode to determine the retention time of each reactant. The m/z values to be monitored by the mass spectrometer were selected taking into account the abundance in the mass spectra of the reference or reactant compounds. According to the column temperature, the measured retention time varied from 1.8 min for propane to 13.8 min for n-nonane. A separate set of runs were also carried out with the individual reference compound or the furans to ensure that the reactions did not produce products with retention times which could interfere with the reactant peaks used in the kinetic analysis. Additionally, all organics were subjected to the 254 nm radiation alone to study the possible photolysis and to ensure that there were no unrecognized reactions occurring in the absence of the Cl atom source. Finally, for each mixture of organics, a number of injections (typically 12 or more) of the unreacted mixture were carried out to obtain an estimate of the precision associated with the measurements in order to use in the error analysis (Brauers and Finlayson-Pitts, 1997). The standard deviations (2s) of these replicate injections were typically in the range 1–4% for the furans and 3–6% for the reference compounds. These ARTICLE IN PRESS B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 measurements also included the losses with the walls of the reaction bag. The reproducibility of the results show that the wall losses of the reactants were not significant. The furans and the reference compound were introduced into the reaction bag by injecting a measured amount of these compounds into a stream of the carrier gas (either air or N2). Concentrations of furans and reference organic ranged from 3 to 13 ppm and from 5 to 18 ppm, respectively. Concentration ranges of 5–16 and 32–80 ppm were used for the internal standard and for Cl atom sources, respectively. Sampling after the measurement of photolysis times was carried out using a VICI VALCO gas-sampling valve. The reaction mixture was slowly pumped through the sampling loop (100 mL) to be sure that it was thoroughly flushed and then allowed to come to equilibrium with the reaction chamber pressure of 1 atm prior to injection. The reactions of interest are the simultaneous reaction of the furans (substrate) and the reference compound with Cl atoms, Cl þ substrate ðSÞ ! products (1) ðkS Þ Cl þ reference compound ðRÞ ! products ðkR Þ (2) As described in detail elsewhere (Atkinson and Aschmann, 1985; Finlayson-Pitts and Pitts, 2000), the decay of the substrate from [S]0 at time t ¼ 0 to [S]t at time t, and the simultaneous loss of the reference compound from an initial concentration of [R]0 to [R]t at time t is given by     ½S 0 ½R 0 kS ln ¼ (I) ln ½S t ½R t kR Thus, a plot of {ln [S]0/[S]t} versus {ln [R]0/[R]t} should be a straight line passing through the origin and whose slope yields the ratio of rate coefficients kS/kR. Benzene was used as an internal standard, as its reaction rate with Cl is neglected due to its low rate coefficient (k298 ¼ (1.370.3)  1015 cm3 molecule1 s1) (Shi and Bernhard, 1997). At each sampling, the ratio of concentration of the reference or substrate against the concentration of internal standard was used in rate coefficients calculations. This method allowed us to minimize the sampling errors. Previously to the determination of the rate kS, the possible photolysis of these compounds was studied. The results showed a significant decrease in the concentration of 2-methylfuran, 3-methylfuran, 2-ethylfuran, and 2,5dimethylfuran after 10 min of UV irradiation at 254 nm. Therefore, when CCl3COCl is used (total time of photolysis of 14–22 min), it is necessary to correct the observed second-order rate coefficient in the kinetic experiments in order to obtain the real second-order rate coefficient (kS). So in the case of the compounds that are subjected to photolysis under the conditions of our experiments 1937 (alkylfurans ¼ 2-methylfuran, 3-methylfuran, 2-ethylfuran, and 2,5-dimethylfuran), the decay of these compounds and the reference compound are governed by the following rate laws (Bierbach et al., 1992, Olariu et al., 2000): d½alkylfuran ¼ kS ½Cl ½alkylfuran þ kph ½alkylfuran dt (II) d½reference (III) ¼ kR ½Cl ½reference dt Integration and rearrangement of Eqs. (II) and (III) leads to the following expression: ln ½alkylfuran 0 kS ½reference 0  kph t ¼ , ln ½alkylfuran t kR ½reference t (IV) where kS and kR are the rate coefficients for reaction of the alkylfuran and reference compound (n-nonane) with Cl atoms, respectively; kph is the photolysis rate constant of alkylfurans and t is the reaction time. Plotting ln([alkylfuran]0/[alkylfuran]t)-kpht versus ln([n-nonane]0/[n-nonane]t) gives kS/kR as the slope of the straight lines according to Eq. (6). When the Cl atom source used is SOCl2, photolysis is carried out in steps of 10 s, the total time of photolysis being 70–110 s. During this time the photolysis undergone by the furans is totally negligible; thus, it was not necessary to make any correction. The chemicals were as follows: propane (Praxair, 99%), 1-butene (Aldrich, 99%), butane (Aldrich, 99%), n-nonane (Aldrich, 99.5%), benzene (Panreac, 99%), thionyl chloride (Aldrich, 99%) and trichloroacetyl chloride (Aldrich, 99%) used as received. a-pinene (Aldrich, 98%), furan (Aldrich, 99%), 2-methylfuran (Aldrich, 99%), 3-methylfuran (TCI America 98%), 2,5dimethylfuran (Aldrich, 99%) and 2-ethylfuran (Aldrich, 97%) were purified by several trap-to-trap distillations. N2 and air (Praxair, ultrahigh purity, 99.999%) were purified by means of an Oxisorb trap and a molecular-sieve trap. 3. Results and discussion 3.1. Control experiments Two Cl reactions were studied in order to validate the GC-MS as a detection technique in the kinetic studies. First, the rate coefficients for Cl+a-pinene and nnonane obtained in our system by using GC-MS detection were compared with the ones obtained by Finlayson-Pitts et al. (1999) and Atkinson et al. (1994) using GC-FID (see Table 1). Secondly, the reaction between Cl atoms and n-nonane was studied using CCl3COCl as an alternate to Cl2, comparing with ARTICLE IN PRESS 1938 B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 secondary reactions or heterogeneous processes are insignificant. The second-order rate coefficient for n-nonane, kR, is accurately known allowing kS to be determined. Fig. 2 shows a plot of Eq. (6) for the reaction of Cl with 2methyl, 3-methyl, 2-ethyl and 2,5-dimethylfuran taking into account kph previously determined. These plots present a very good linearity and have intercepts close to zero, within the experimental error. Table 2 summarizes the rate coefficients for the reactions of Cl atom and furans using SOCl2 and CCl3COCl sources in either N2 or air. The error limit for the ratio of rate coefficients kS/kR and kS includes only the precision of the fit to our experimental data ( 2s). The data show that within the error bars there is no significant difference between runs in N2 and in air, indicating that interference from OH reaction is negligible. As it can be seen in Table 2, the large values of the rate coefficients (order of 1010 cm3 molecule1 s1) show a great reactivity of Cl towards furan and alkylfurans. The larger inductive effect of methyl group increases the reactivity of the cycle in Cl reaction, although the rate coefficient is hardly affected by the length or the position of the alkyl group (compare rate constants of 2/3methylfuran with 2-ethylfuran). Since Cl atom could compete with OH radical in the oxidation of some organics in the marine boundary layer, coastal regions (Finlayson-Pitts et al., 1999) and continental areas, it seems reasonable to compare the reactivity of these oxidants with the same group of compounds. So Bierbach et al. (1992, 1995) proposed as the main reaction channel for OH-furans reactions the electrophilic addition to the p system; the mechanism also proposed here for Cl-furans due to the electrophilic nature of Cl atom. In the case of furan, the attack of the radical for both Cl and OH reactions could happen at C2 (or C-5) and C-3 (or C-4). Both positions undergo radicals stabilized by resonance. Nevertheless, the attack at ortho position (C-2 or C-5) leads to a radical intermediate with an additional resonant structure, Atkinson’s data (Atkinson, 1994) where Cl2 is used as a precursor of Cl atoms. In Table 1, the obtained results for the studied reactions are shown. In light of the agreement between our rate coefficients and those from bibliography, it could be concluded that CCl3COCl is a good alternate chlorinate source for these kind of studies. The photolysis rate constants (kph) were measured in separate experiments under typical experimental kinetic conditions obtaining a range of kph of (4–6)  105 s1 for 2-methylfuran, 3-methylfuran and 2-ethylfuran and of (1.970.8)  104 s1 for 2,5-dimethylfuran. Photolysis represents a contribution less than 20% of the total decay of the alkylfuran concentrations. For furan and the reference compounds photolysis were not observed under our experimental conditions. 3.2. Rate coefficients of Cl+furan and alkylfurans reactions In this work, all reactants were studied relative to nnonane. In addition for furan, the rate coefficient was also measured relative to propane, 1-butene, trans-2butene in order to ensure that the rate coefficient was not dependent on the reference compound used. This fact is showed by the data obtained for kS that are equal within the experimental errors (see Table 2). So, the rate coefficient for furan was obtained from the weighted average of the several rate coefficients. Also, separate sets of experiments were carried out for all organic pairs using air or N2 as a carrier gas to test for potential systematic errors due to OH reaction with the organics (Kaiser and Wallington, 1996). Fig. 1 shows some typical data plotted in the form of Eq. (I) for the Cl+furan reaction. In agreement with this equation, the data yield straight lines that passes through the origin with a slope of kS/kR indicating that Table 1 Rate coefficients for the reactions of n-nonane and a-pinene with Cl atoms at atmospheric pressure and room temperature. Comparison between our results and the bibliographic ones Reaction Chlorine source k298 K/1010 (cm3 molecule1 s1) k/kreference a c Technique n-Nonane+Cl Cl2 CCl3COCl 2.2170.06 2.3770.29a 4.870.1 5.270.8d GC-FID GC-MS a-Pinene+Cl CCl3COCl CCl3COCl CCl3COCl 0.9570.24b 1.1070.02b 0.9970.06b 4.671.2e 5.370.1f 4.870.3d GC-FID GC-FID GC-MS a With a reference rate coefficient of butane of (2.1870.22)  1010 cm3 molecule1 s1 (Atkinson et al., 1995). With a reference rate coefficient of n-nonane of (4.8270.14)  1010 cm3 molecule1 s1 (Atkinson, 1997). c Atkinson et al. (1995). d This work. e Finlayson-Pitts et al. (1999). f Timerghazin and Parisa, (2001). b ARTICLE IN PRESS 1939 B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 Table 2 Summary of relative rate coefficients and the absolute values derived from them for the reactions of furan and furan derivatives with Cl atoms at atmospheric pressure and 29872 K. The absolute rate coefficients are in units of 1010 cm3 molecule1 s1 Organic Reference Chlorine atom source Carrier gas Detection system ks/kr ( 2s) Average ks/ kr ( 2s)e k298 k ( 2s) Average k̄298 k Propanea CCl3COCl N2 GC-MS 1.470.1 1.970.4 1-Buteneb n-Onanec CCl3COCl CCl3COCl N2 N2 GC-MS GC-MS 0.6870.02 0.4170.05 2.070.3 2.070.2 E-2-butened CCl3COCl Air N2 Air N2 GC-MS GC-MS GC-MS GC-MS 0.6370.05 0.5870.04 0.670.1 0.5770.1 CCl3COCl Air GC-MS 0.9470.06 SOCl2 N2 Air N2 GC-MS GC-MS GC-MS 0.8070.03 0.9970.08 0.9170.05 CCl3COCl Air GC-FID 1.070.2 SOCl2 N2 Air N2 GC-FID GC-MS GC-MS 0.8370.05 0.9970.10 0.9770.10 CCl3COCl Air GC-FID 1.170.2 SOCl2 N2 Air N2 GC-FID GC-MS GC-MS 0.9370.05 1.170.2 0.970.3 CCl3COCl Air GC-MS 1.370.2 SOCl2 N2 Air N2 GC-MS GC-MS GC-MS 1.1670.06 1.270.2 1.1670.1 2.070.2e Furan O SOCl2 2-Methylfuran O n-Nonane 0.5970.03 2.470.3 0.8570.03 4.170.2 CH3 3-Methylfuran n-Nonane 4.270.3 CH3 O 2-Ethylfuran O n-Nonane CH2CH3 2,5-Dimethylfuran CH3 0.8870.04 O n-Nonane 0.9570.05 4.670.3 1.1770.06 5.770.3 CH3 a Rate constant of the reaction Cl+propane ¼ (1.470.27)  1010 cm3 molecule1 s1 (Atkinson, 1997). Rate constant of the reaction Cl+1-butene ¼ (3.070.4)  1010 cm3 molecule1 s1 (Orlando et al., 2003). c Rate constant of the reaction Cl+trans-2-butene ¼ (4.070.5)  1010 cm3 molecule1 s1 (Orlando et al., 2003). d Rate constant of the reaction Cl+n-nonane ¼ (4.8270.14)  1010 cm3 molecule1 s1 (Atkinson and Aschmann, 1995) which has been adjusted to the recommended value of k (Cl+n-butane) ¼ 2.18  1010 cm3 molecule1 s1 (Atkinson, 1997). e Weighted average according to the precision of the measurement (w ¼ 1=s2 ). b reason why this position would be more favoured than another one (Bierbach et al., 1992). In reference to alkylfurans, there are several sites of attack for Cl or OH that leads to different radical intermediates (see Scheme 1). For 2-methylfuran only the attack at C-5 leads to an intermediate stabilized by resonance with three resonant structures versus 3-methylfuran that presents two sites of attack (C-2 and C-5) that leads to ARTICLE IN PRESS 1940 B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 3 2.5 ln ([Furan]o/[Furan] t) Propane 2 1-Butene 1.5 1 Trans-2-butene 0.5 n-nonane 0 0 1 2 ln ([R]o/[R]t) 3 4 Fig. 1. ln([furan]0[furan]t) versus ln([R]0/[R]t) for the reactions of Cl atoms and furan using different reference compounds. 3 ln [Alkylfuran]0/[Alkylfuran]t-kpht 2.5 2 1.5 1 0.5 0 0 0.5 1 1.5 2 2.5 ln ([n-nonane]0/[n-nonane]t) 3 Fig. 2. Relative rate plots according to Eq. (6) for alkylfurans with Cl atoms using CCl3COCl in N2, (’) 2,5-dimethylfuran, (K) 2-ethylfuran, (J) 3-methylfuran, and (m) 2-methylfuran. For all, the reference compound is n-nonane. There are no error bars since they are negligible (estimated at 1%). an intermediate stabilized by resonance with three resonant structures in each position. By comparison between the rate coefficients measured in this work (kS) and those found in the literature for the corresponding OH reaction, it can be seen that there is a clear positive reactivity dependence on the number of –CH3 groups for OH and Cl as it is expected due to the electrodonating nature of the methyl substituent. In contrast, the effect of the position of such a group on the rate coefficient is different for Cl and OH reactions. While the value of the rate coefficients of OH-reaction with furans drops with the stability of the radicals formed (for example, kS for 3-methylfuran is rather larger than that for 2-methylfuran, because more stable radicals are produced in the first case, as explained above), the Cl reactions are independent of methyl position, as it can be seen from the data of Table 3. This could be explained by the fact that Cl atoms are less selective although it is more reactive than hydroxyl radical. Its selectivity can be due to its high reactivity, with rate coefficients approaching collision control (collision limits estimated using collision theory, 1.6  1010 for furan to 2.5  1010 cm3 molecule1 s1 for 2,5 dimethylfuran and 2-ethylfuran) and therefore there is no difference of reactivity in function of the position or the length of alkyl chain. The effect of a second methyl group on the reactivity of furan is again larger in the case of OH radical than Cl atom due to the same reason (see Table 3). Although the rate coefficients obtained in this work indicate that the first step is an electrophilic addition of Cl to one of the CQC of furans, the determination of products are necessary to establish the reaction mechanism. There are no data about product determination in Cl reactions. Only Bierbach et al. (1995) reported the reaction products of OH with furan, 2-methylfuran and furanaldehydes. That study demonstrated that unsaturated dicarbonyls were formed in high yields in the OH-initiated oxidation of heterocyclic aromatics rings. It suggested a primary ring opening product of the OH-initiated degradation of furan and its alkylated derivatives. Other reaction partners for furan and its derivatives in the atmosphere apart from OH and Cl include NO3 and O3. Rate coefficients reported in the literature for the reactions of O3 and NO3 with the compounds studied in this work are also shown in Table 3. These oxidants show a similar trend in the reactivity to that of OH. As it can be seen in this table, the rate coefficients for NO3 (1012–1011 cm3 molecule1 s1) and O3 (1017 cm3 molecule1 s1) are significantly lower than the ones for OH and Cl reactions. According to the rate coefficients reported in Table 3, we can compare the reactivity of the different tropospheric oxidants. So, the rate coefficients, in decreasing reactivity order, can be written as kCl 4kOH 4kNO3 bkO3 : ARTICLE IN PRESS B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 1941 Sterically hindered C-2 X O O 2-methylfuran X C-5 O O X X O X O C-4 O X X C-3 +X X O X O X O Radical more stable by resonance O Radical more stable by resonance C-2 X O C-3 +X X O O X X X Sterically hindered O 3-methylfuran O X O X C-4 C-5 O O X O X O X O Radical more stable by resonance Scheme 1. Possible attack positions of X (Cl or OH radical) to methylfurans. 4. Conclusion and atmospheric implications To date, the measurements reported here constitute the first determination of the rate coefficients for the reaction of Cl atoms with furan, 2-methylfuran, 3methylfuran, 2-ethylfuran and 2,5-dimethylfuran. Our measurements confirm that these reactions are very fast, and therefore there is no difference of reactivity in the function of the position or the length of alkyl chain. Although the reactivity of furan and its derivatives towards the tropospheric oxidants is Cl4OH4NO3 O3 ; the role of these oxidants in the degradation tropospheric process depends on their atmospheric concentrations. Using the kinetic data obtained in this work in combination with an average tropospheric concentrations or a peak value of the different oxidants, upper limits for atmospheric lifetime (t) of furan and its alkyl derivatives have been calculated according to the relationship t ¼ 1=k½X ; X ¼ OH, Cl, O3 or NO3. The estimated lifetimes are summarized in Table 3. Assuming an average global concentration of 104 molecule cm3 for Cl atoms, the tropospheric lifetimes vary from 136 h for furan to 58 h for 2,5dimethylfuran. Nevertheless, the contribution of Cl atoms may be significant in those areas with higher concentration. In fact, peak concentrations as high as 1  105 atoms cm3 (Spicer et al., 1998) are expected in the marine boundary layer at dawn and much earlier than OH where [OH] is 5  105 radicals cm3 (Brauers et al., 1996). Under such eventual conditions the tropospheric lifetimes for Cl vary from ca. 13.6 h for furan to ca. 5.8 h for 2,5-dimethylfuran versus tropospheric OH lifetimes of 34 h for furan to 10 h for 2,5-dimethylfuran. ARTICLE IN PRESS 1942 B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 Table 3 Summary of the rate coefficients for the reactions of Cl, OH, NO3 and O3 with the compounds studied in this work and typical tropospheric lifetimes (k in units of cm3 molecule1 s1) a VOCs kCl/1010 kOH/1011 Furan 2-Methylfuran 3-Methylfuran 2.0 4.1 4.2 4.05b 6.19c 9.35d 2-Ethylfuran 2,5-Dimethylfuran 4.6 5.7 10.77c 13.21c kNO3 =1011 kO3 =1017 tCl/h 0.14e 2.57f 2.86f 1.31g — 5.78f 0.24h — 2.05g — — — 136i 73i 70i — 63i 58i i tOH/h 6.8 4.5 3 — 2.5 2.0 j tNO3 = min k 23.8 1.3 1.2 1.9 — 0.6 tO3 =h l 165 — 19 — — — a This work. Atkinson (1994). c Bierbach et al. (1992). d Atkinson, 1989. e Atkinson, 1991. f Kind et al. (1996). g Alvarado et al. (1996). h Atkinson et al. (1983). i Assuming an average global concentration of [Cl] ¼ 1  104 molecule cm3 (Wingenter et al., 1996). j Assuming a 12-h concentration of [OH] ¼ 1  106 radicals cm3 as a 12-h daytime average (Spivakovsky et al., 2000). k Assuming a 12-h concentration of [NO3] ¼ 5  108 radicals cm3 (Shu and Atkinson, 1995). l Assuming a 24-h concentration of [O3] ¼ 7  1011 molecule cm3 (Logan, 1985). b These values show a loss of the furans by reaction with Cl atoms that is comparable to that by reaction with OH. Hence, Cl atoms clearly can contribute significantly to the loss of this kind of compounds in competition with OH radical playing a significant role in the degradation of studied furans at dawn in the marine boundary layer and the coastal areas. That may also be the case of some urban contaminated areas where high levels of Cl may be originated from industrial emission (Galán et al., 2002). The tropospheric removal of these compounds in the presence of NO3 radical, assuming a 12-h night time average of [NO3] ¼ 5  108 radicals cm3 (Shu and Atkinson, 1995) also constitutes a very important sink (lifetime about few minutes) that is going to dominate the nocturnal chemistry of furans. Regarding to the reaction with ozone this oxidant hardly contributes to the degradation of these kinds of compounds since the lifetimes are too large in comparison to the other tropospheric oxidants. All these numbers should be treated with caution because the Cl, OH, NO3 and O3 concentrations vary substantially depending on the environment, location and season. The identification of reaction products for these reactions is of great importance because if these compounds are unsaturated carbonyls they are secondary pollutants that are involved in processes such as the photochemical smog (Comittee on Aldehyde, 1981) or peroxyacyl nitrates (PANs) formation (Wayne, 2000). This fact justified the need of the characterization of reaction products for furan and its derivatives with Cl and shows the atmospheric implications of these reactions. Acknowledgment Florentina Villanueva Garcı́a thanks ‘‘Junta de Comunidades de Castilla La Mancha’’ for a personal grant. References Alvarado, A., Atkinson, R., Arey, J., 1996. Kinetics of the gasphase reactions of NO3 radicals and O3 with 3-methylfuran and the OH radical yield from the O3 reaction. International Journal of Chemical Kinetics 28, 905–909. Andreae, M.O., Merlet, P., 2001. Emission of trace gases and aerosols from biomass burning. Global Biogeochemical Cycles 15 (4), 955–966. Atkinson, R., 1989. Kinetics and mechanisms of the gas-phase reactions of the hydroxyl-radical with organic compounds. Journal of Physical and Chemical Reference Data Monograph No. 1. Atkinson, R., 1991. Gas-phase tropospheric chemistry of organic compounds. Journal of Physical and Chemical Reference Data Monograph No. 2. Atkinson, R., 1994. Gas-phase tropospheric chemistry of organic compounds. Journal of Physical Chemical Reference Data Monograph 2. Atkinson, R., 1997. Gas-phase tropospheric chemistry of volatile organic compounds: 1. Alkanes and alkenes. Journal of Physical and Chemical Reference Data 26. Atkinson, R., Aschmann, S.M., 1985. Kinetics of the gas phase reaction of Cl atoms with a series of organics at 29672 K and atmospheric pressure. International Journal of Chemical Kinetics 17, 33–41. ARTICLE IN PRESS B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 Atkinson, R., Aschmann, S.M., 1995. Rate constants for the gas-phase reactions of alkanes with Cl atoms at 29672 K. International Journal of Chemical Kinetics 27, 613–622. Atkinson, R., Aschmann, S.M., Carter, W.P.L., 1983. Kinetics of the reactions of O3 and OH radicals with furan and thiophene at 29872 K. International Journal of Chemical Kinetics 15, 51–61. Bierbach, A., Barnes, I., Becker, K.H., 1992. Rate coefficients for the gas-phase reactions of hydroxyl radicals with furan, 2-methylfuran, 2-ethylfuran and 2,5-dimethylfuran at 30072 K. Atmospheric Environment 26A (5), 813–817. Bierbach, A., Barnes, I., Becker, K.H., 1995. Product and kinetic study of the OH-Initiated Gas-Phase oxidation of furan, 2-methylfuran and furanaldehydes at E300 K. Atmospheric Environment 29 (19), 2651–2660. Bierbach, A., Barnes, I., Becker, K.H., 1996. Rate coefficients for the gas-phase reactions of bromine radicals with a series of alkenes, dienes, and aromatic hydrocarbons at 29872 K. International Journal of Chemical Kinetics 28, 565–577. Bierbach, A., Barnes, I., Becker, K.H., 1999. Rate constants of the Br-initiated gas-phase oxidation of a series of alcohols, furans and benzenes at 30072 K. Atmospheric Environment 33, 2981–2992. Bonsang, B., Polle, C., Lambert, G., 1992. Evidence for marine production of isoprene. Geophysical Research Letters 19, 1129–1132. Brauers, T., Finlayson-Pitts, B.J., 1997. Analysis of relative rate measurement. International Journal of Chemical Kinetics 29 (9), 665–672. Brauers, T., Aschmutat, U., Brandenburger, U., Dorn, H.P., Hausmann, M., Hessling, M., Hofzumahaus, A., Holland, F., Plas-Dulmer, C., Ehhalt, D.H., 1996. Intercomparison of tropospheric OH radical measurements by multiple folded long-path laser absorption and laser-induced fluorescence. Geophysical Research Letters 23, 2545–2548. Chatfield, R.B., Crutzen, P.J., 1990. Are there interactions of iodine and sulfur species in marine air photochemistry? Journal of Geophysical Research 95, 22,319–22,341. Comittee on Aldehyde, 1981. National Research Council; Formaldehyde and Other Aldehydes. National Academy Press, Washington, DC, pp. 1–340. Finlayson-Pitts, B.J., Pitts Jr., J.N., 2000. Chemistry of the Upper and Lower Atmosphere-Theory, Experiments, and Applications. Academic Press, San Diego. Finlayson-Pitts, B.J., Keoshian, C.J., Buehler, B., Ezell, A.A., 1999. Kinetics of reaction of chlorine atoms with some biogenic organics. International Journal of Chemical Kinetics 31, 491–499. Galán, E., González, I., Fabbri, B., 2002. Estimation of fluorine and chlorine emissions from Spanish structural ceramic industries. The case study of the Bailén area, Southern Spain. Atmospheric Environment 36, 5289–5298. Graedel, T.E., Hawkings, D.T., Claxton, L.D., 1986. Atmospheric Chemical Compounds: Sources, Occurrence, and Bioassay. Academic press, New York. Grosjean, D., Williams II, E.L., 1992. Environmental persistence of organic compounds estimated from structurereactivity and linear free-energy relationships, unsaturated aliphatics. Atmospheric Environment Part A 26, 1395–1405. 1943 Gu, C.E., Rynard, C.M., Hendry, D.G., Mill, T., 1985. Hydroxyl radical oxidation of isoprene. Environmantal Science and Technology 19, 151–155. Impey, G.A., Mihele, C.M., Anlauf, K.G., Barrie, L.A., Hastie, D.R., Shepson, P.B., 1999. Measurements of photolyzable halogen compounds and bromine radicals during the Polar Sunrise Experiment 1997. Journal of Atmospheric Chemistry 34, 21–37. Isidorov, V.A., Zenkevich, I.G., Ioffe, B.V., 1985. Volatile organic compounds in the atmosphere of forests. Atmospheric Environment 19, 1–8. Jobson, B.T., Niki, H., Yokouchi, Y., Bottenheim, J., Hopper, F., Leaitch, R., 1994. Measurements of C2–C6 hydrocarbons during the Polar Sunrise 1992 Experiment: evidence for Cl atom and Br atom chemistry. Journal of Geophysical Research 99 (25), 355–368. Kaiser, E.W., Wallington, T.J., 1996. Kinetics of the reactions of chlorine atoms with C2H4 (k1) and C2H2 (k2): a determination of DH f;2988 for C2H3. Journal of Physical Chemistry 100, 4111. Keene, W.C., Pszenny, A.A.P., Jacob, D.J., Duce, R.A., Galloway, J.N., Schultz-Tokosm, J.J., Sievering, H., Boatman, J.F., 1990. The geochemical cycling of reactive chlorine through the marine troposphere. Global Biogeochemical Cycles 4, 407–429. Kind, I., Berndt, T., Böge, O., Rolle, W., 1996. Gas-phase rate constants for the reaction of NO3 radicals with furan and methyl-substituted furans. Chemical Physics Letters 256, 679–683. Knudsen, J.T., Tollstein, L., Bergström, L.G., 1993. Floral scents- a checklist of volatile compounds isolated by headspace techniques. Phytochemistry 33, 253–284. Logan, J.A., 1985. Tropospheric ozone-seasonal behaviour, trends, and anthropogenic influence. Journal of Geophysical Research 90, 463–482. Ohta, T., 1984. Furan ring formation on the OH-initiated photooxidation of 1,3-butadiene and 1,3-pentadiene. Bulletin of the Chemical Society of Japan 57, 960–966. Olariu, R.I., Barnes, I., Becker, K.H., Klotz, B., 2000. Rate coefficients for the gas-phase reaction of OH Radicals with selected dihydroxybenzenes and benzoquinones. International Journal of Chemical Kinetics 32, 696–702. Orlando, J.J., Tyndall, G.S., Apel, E.C., Riemer, D.D., Paulson, S.E., 2003. Rate coefficients and mechanisms of the reaction of Cl-atoms with a series of unsaturated hydrocarbons under atmospheric conditions. International Journal of Chemical Kinetics 35, 334–353. Pszenny, A.A.P., Keene, W.C., Jacob, D.J., Fan, S., Maben, J.R., Zetwo, M.P., Springer-Young, M., Galloway, J.N., 1993. Evidence of inorganic chlorine gases other than hydrogen chloride in marine surface air. Geophysical Research Letters 20, 699–702. Rasmussen, R.A., Khalil, M.A.K., 1988. Isoprene over the Amazon basin. Journal of Geophysical Research 93 (D2), 1417–1421. Ratte, M., Bujok, O., Spitzy, A., Rudolph, J., 1998. Photochemical alkene formation in seawater from dissolved organic carbon. Results from laboratory experiments. Journal of Geophysical Research 103, 5707–5717. Ruppert, L., Becker, K.H., 2000. A product study of the OH radical-initiated oxidation of isoprene: formation of ARTICLE IN PRESS 1944 B. Cabañas et al. / Atmospheric Environment 39 (2005) 1935–1944 C5-unsaturated diols. Atmospheric Environment 34 (10), 1529–1542. Ruppert, L., Barnes, I., Becker, K.H., 1992. A product study of the OH radical-initiated oxidation of isoprene in the absence of NOx. In: Peeters, J. (Ed.), CEC Air Pollution Research Report 45, pp. 193–198. Shepson, P.B., Edney, E.O., Corse, E.W., 1984. Ring fragmentation reactions on the photooxidations of toluene and o-xylene. Journal of Physical Chemistry 88, 4122–4126. Shi, J., Bernhard, J., 1997. Kinetic studies of Cl-atom reactions with selected aromatic compounds using the photochemical reactor-FTIR spectroscopy technique. International Journal of Chemical Kinetics. 29, 349–358. Shu, Y., Atkinson, R., 1995. Atmospheric lifetimes and fates of a series of sesquiterpenes. Journal of Geophysical Research 100, 7275–7282. Singh, H.B., Kasting, J.F., 1988. Chlorine-hydrocarbon photochemistry in the marine troposphere and lower stratosphere. Journal of Atmospheric Chemistry 7, 261–285. Singh, H.B., Thakur, A.N., Chen, Y.E., Kanakidou, M., 1996. Tetrachloroethylene as an indicator of low Cl atom concentrations in the troposphere. Geophysical Research Letters 23, 1529–1532. Spicer, C.W., Chapman, E.G., Finlayson-Pitts, B.J., Plastidge, R.A., Hubbe, J.M., Fast, J.D., Berkowitz, C.M., 1998. Observations of molecular chlorine in coastal air. Nature 394 (6691), 353–356. Spivakovsky, C.M., Logan, J.A., Montzka, S.A., Balkanski, Y.J., Foreman-Fowler, M., Jones, D.B.A., Horowitz, L.W., Fusco, A.C., Benninkmeijer, C.A.M., Prather, M.J., Wofsy, S.C., McElroy, M.B., 2000. Three-dimensional climatological distribution of tropospheric OH: update and evaluation. Journal of Geophysical Research 105, 8931–8980. Timerghazin, Q.K., Parisa, A.A., 2001. Kinetics of the gasphase reaction of atomic chlorine with selected monoterpenes. Physical Chemistry Chemical Physics 3, 3981–3986. Tuazon, E.C., Atkinson, R., 1990. A product study of the gasphase reaction of methacrolein with the OH radical in the presence of NOx. International Journal of Chemical Kinetics. 22, 591–602. Wayne, R.P., 2000. Chemistry of Atmospheres. Oxford University Press, New York. Wingenter, O.W., Kubo, M.K., Blake, N.J., Smith, T.W., Blake, D.R., Rowland, F.S., 1996. Hydrocarbon and halocarbon measurements as photochemical and dynamical indicators of atmospheric hydroxyl, atomic chlorine, and vertical mixing obtained during Lagrangian fights. Journal of Geophysical Research 101, 4331–4340. Wingenter, O.W., Sive, B.C., Blake, D.R. Rowland, F.S., 1999. Atomic chlorine concentrations determined from ethane and hydroxyl measurements made over the central Pacific Ocean, in American Geophysical Union, San Francisco.