Environ Sci Pollut Res DOI 10.1007/s11356-014-2913-9
ATMOSPHERIC POLLUTANTS IN A CHANGING ENVIRONMENT
Reactions of Cl atoms with alkyl esters: kinetic, mechanism and atmospheric implications Stefanie Ifang & Thorsten Benter & Ian Barnes
Received: 3 February 2014 / Accepted: 14 April 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Rate coefficients have been measured for the reaction of Cl atoms with a series of alkyl esters at 298±2 K and atmospheric pressure in a large volume photoreactor using the relative kinetic technique. The kinetic data have been used in conjunction with other literature studies on the reactions of Cl atoms with esters to revise the existing values for ester substituent factors in a structure activity relationship (SAR) for Cl reactions. Product studies are reported for the reactions of Cl atoms with isopropyl ethanoate and methyl-2-methylpropanoate under NOx-free conditions. These studies highlight the types of products that can be expected when oxidation occurs at R groups on the acyl (-C(O)OR) and oxy (RC(O)O-) sides of the ester functionality where R is a straight or branched chain alkyl entity. Possible atmospheric repercussions of the atmospheric chemistry of esters are considered. Keywords Alkyl esters . Chlorine atoms . Oxidation mechanism . Relative kinetic . Structure activity relationship
Introduction Alkyl esters (RC(O)OR, R=alkyl group) are an important class of oxygenated volatile organic compounds (OVOCs) which are emitted directly to the atmosphere from both natural and anthropogenic sources. They also are produced in situ in the atmosphere via the oxidation of ethers. Natural sources of Responsible editor: Gerhard Lammel Electronic supplementary material The online version of this article (doi:10.1007/s11356-014-2913-9) contains supplementary material, which is available to authorized users. S. Ifang : T. Benter : I. Barnes (*) FB C - Department of Physical Chemistry, University of Wuppertal, Gauss Strasse 20, 42119 Wuppertal, Germany e-mail: [email protected]
esters include emissions from vegetation and fruit. Esters are used to varying degrees as solvents in the manufacture of fats, cellulose, varnishes, paints and insecticides and as reagents during the manufacture of food flavourings (Calvert et al. 2011; Graedel et al. 1986; Koppmann 2007). They are often added to perfumes since many have pleasant fruity odours (Paillard 1990; Rowan et al. 1996). Some alkyl and cyclic esters (lactones) are under consideration as candidates for biofuels (Alonso et al. 2013; Chan-Thaw et al. 2013; Hayes and Burgess 2009; Huynh and Violi 2008). Anthropogenic emissions of the unsaturated esters to the atmosphere can occur during their production, processing, storage and disposal. In the atmosphere, saturated esters will be subject to photooxidation initiated mainly by OH radicals during the daytime and by NO3 radicals during the night. In certain environments, Cl atom-initiated oxidation will also be important. Conventionally, the major sources of Cl atoms in the troposphere have been thought to be (i) the photolysis of molecular chlorine for which a number of possible heterogeneous production sources have been proposed (Finlayson-Pitts et al. 1989; Finley and Saltzman 2006, 2008; Vogt et al. 1996) and (ii) the reaction of OH radicals with hydrogen chloride. On a global scale, the atmospheric Cl atom concentration is low, ~103 molecules cm−3 (Rudolph et al. 1996; Singh et al. 1996), and atmospheric levels sufficiently high enough to be of significance for VOC decay have been thought to have been confined largely to coastal areas and possibly some polluted industrial regions with molecular chlorine emission sources (Finley and Saltzman 2006, 2008; Spicer et al. 1998; Wingenter et al. 1999, 2005). However, recent field work has presented evidence for Cl chemistry in continental regions remote from coastal regions involving the photolysis of nitryl chloride (ClNO2) as the Cl atom source suggesting that Clinitiated photooxidation chemistry may play a more important role than previously thought in polluted continental regions
Environ Sci Pollut Res
(Mielke et al. 2011; Philips et al. 2012; Riedel et al. 2012, 2013; Sarwar and Bhave 2007; Thornton et al. 2010). Presented here is a relative kinetic and limited product study on the reaction of Cl atoms with a series of alkyl esters. The aims of the study are to extend the kinetic database for esters, improve structure reactivity relationships (SARs) for Cl atom reactions and provide photooxidation mechanisms for esters for inclusion in chemistry transport (CT) models. Of the 18 alkyl esters investigated in this work, room temperature rate coefficients have been reported previously in the literature for nine of the esters. Rate coefficients have been reported for the reactions of Cl with ethyl, propyl, butyl, isopropyl and tert-butyl ethanoate using both absolute (Notario et al. 1998; Cuevas et al. 2005; Xing et al. 2009) and relative kinetic (Xing et al. 2009) techniques. Rate coefficient determinations using a relative kinetic technique have also been reported for the reactions of Cl with ethyl propanoate (Andersen et al. 2012; Cometto et al. 2009), propyl propanoate, methyl-2-methyl and ethyl butanoate (Cometto et al. 2009). For the remainder of the ester compounds, this study represents the first-time determination of their room temperature rate coefficient with Cl atoms.
Experimental The experiments have been performed in two large volume photoreactors with volumes of 1,080 and 480 L at 298±2 K and 760±10 Torr total pressure of synthetic air. The reactors are equipped with fluorescent lamps (Philips TLA 40W/05, 300≤λ≤450 nm, λmax =360 nm) which are used for the photolytic production of radicals. White-type mirror systems are mounted internally within each of the chambers and coupled to FTIR spectrometers (Fourier transform-infrared spectrometer; Nicolet Nexus and Nicolet Magna 550) each equipped with a globar as IR source and a liquid nitrogen cooled MCT detector (mercury-cadmium-telluride). The setups enable in situ infrared monitoring of both reactants and products. The White systems in the reactors were operated with total optical absorption paths of 487 m (1,080 L) and 50.4 m (480 L), and infrared spectra were recorded with a spectral resolution of 1 cm−1. Further details on the reactors can be found in Barnes et al. (1993, 1994). In some cases GCFID was also used to monitor the reactants. Samples, taken from the reactor in gas-tight syringes, were injected into the inlet port of a Hewlett Packard HP5890A Series II gas chromatograph containing a 30m-long, 0.53 mm ID capillary RTX 225 (fused silica) column from Restek operated at 60 °C.
Chlorine atoms were produced in both reactors by the photolysis of molecular chlorine. Cl2 þ hν→Cl þ Cl Experiments were performed on mixtures of Cl2/ester/ethene/air. In the case of the measurements using FTIR, the mixtures were irradiated for periods of 15–20 min during the course of which infrared spectra were recorded at 1 cm−1 resolution with the FTIR spectrometer. Typically, 64 interferograms were co-added per spectrum over a period of approximately 1 min. For the experiments using GC-FID for the analysis, the mixtures were irradiated for 40 min and samples taken from gas-tight syringes as indicated above. Product analyses were performed on the reactions of Cl atoms with isopropyl ethanoate and methyl-2-methylpropanoate using FTIR for in situ monitoring of the reactants and products. The experimental conditions used were similar to those given above for the kinetic experiments, the only difference being the omission of the reference hydrocarbon from the reaction mixture. The initial concentrations used in the experiments in ppmV (1 ppmV=2.46×1013 molecule cm−3 at 298 K and 760 Torr of total pressure) were typically 6–10 for the esters and the reference compound and between 10 to 15 ppm for Cl2. The chemicals used in the experiments had the following purities as given by the manufacturer and were used as supplied: synthetic air 99.999 % (Air Liquide), esters ≥98 % (Aldrich and Alfa Aesar), ethane 99.5 % (Aldrich) and chlorine 99.8 % (Messer Griesheim).
Data evaluation and results Kinetic studies In the presence of the oxidant, in this study, the Cl atom, the ester and reference compound decay through the following reactions: ester þ Cl→products; k ester
ð1Þ
reference þ Cl→products; k ref
ð2Þ
Provided that the ester under investigation and the reference compound are lost only by reactions (1) and (2), then it can be shown that: ½Estert0 k ester ½Referencet0 ln ln ¼ ð3Þ ½Estert k ref ½Referencet where [Ester]t0, [Reference]t0, [Ester]t and [Reference]t are the concentrations of the ester and reference compound at times
Environ Sci Pollut Res
t=0 and t, respectively, and kester and kref are the rate coefficients of reactions (1) and (2), respectively. The relative rate technique relies on the assumption that both the ester and reference compound are removed solely by reaction with the Cl atoms. To verify this assumption, various tests were performed to assess possible contributions to the decays of the organics by reaction with the radical precursor Cl2, photolysis and loss to the reactor surface. Mixtures containing an ester, a reference compound and the radical precursor Cl2 were prepared and allowed to stand in the dark for 1 h. In all cases, loss of both the ester and reference compound in the presence of the radical precursor and in the absence of UV light was negligible showing that both wall deposition and reaction of the compounds with Cl2 are negligible loss processes. Additionally, to test for possible photolysis of the esters, mixtures of the esters alone in air were irradiated for 30 min. No measurable loss of the esters within this time period was observed showing that photolysis is also a negligible loss process for the esters. In the experiments using FTIR where ethene was used as the reference compound, the products formed from its reaction with Cl atoms, mainly HCHO and HCClO, did not interfere with the spectral analysis. The kinetic data obtained from the experiments on the reactions of Cl with the 18 alkyl esters using FTIR have been plotted according to Eq. (3). Figure 1 shows example plots for four of the esters. Plots for the remainder are shown in Figs. SI-1 to SI-4 in the Supporting Information (SI). In all cases, reasonably good linear relationships with zero intercept were obtained. Similarly, good linear relationships were obtained for the alkyl esters investigated using GC-FID for the reactant analysis. For all of the alkyl ester and reference compound combinations, a minimum of three experiments was performed. Table 1 lists the alkyl esters investigated, the method applied for the analysis, the reference compound used and the average rate coefficient ratio kester/kref obtained for each ester and reference combination investigated. Ethene, with a rate coefficient of kCl+ethene = 1.1×10−10 cm3 molecule−1 s−1 (Atkinson et al. 2006), was used as a reference compound for the experiments employing FTIR, and ethyl ethanoate (EE) and propyl ethanoate (PE), with rate coefficients of kCl + EE =1.76× 10−11 and kCl+PE =7.19×10−11 cm3 molecule−1 s−1 (Xing et al. 2009), respectively, were used for the experiments employing GC-FID for the analysis. The absolute values of the rate coefficients, kester, obtained for the reaction of Cl with the alkyl esters from the rate coefficient ratios kester/kref using the appropriate value of the rate coefficient for the reference compound are listed in the right-hand column in Table 1. The quoted errors for the rate coefficients are the 2σ errors from the linear regression analyses of the data plots plus an additional 10 % to account for the uncertainty in the reference compound rate coefficient.
Fig. 1 Examples of plots of the kinetic data for the reactions of Cl atoms with ethyl-2-methyl-pentanoate, ethyl-2-methyl-butanoate, ethyl-2methylpropanoate and methyl-2-methyl-propanoate relative to the reaction of Cl with ethene obtained at 298±2 K and atmospheric pressure using in situ FTIR for the analysis
In cases where both analytical methods were used to determine the rate coefficient, good agreement was found in nearly all cases. The only exception was for methyl-3-methylbutanoate where the difference was ~25 %.
Product studies Cl+isopropyl ethanoate Figure 2(a) shows a spectrum obtained after irradiation of an isopropyl ethanoate/Cl2/air mixture and subtraction of the absorptions due to isopropyl ethanoate. Panel (b) shows a reference spectrum of acetic acid anhydride (AAA: CH3C(O)OC(O)CH3). Visual inspection of the spectra shows quite clearly that AAA is a major reaction product. Other products detected in the product spectrum included CO, formaldehyde (HCHO), formic acid (HC(O)OH), acetic acid (CH3C(O)OH) and minor amounts of methanol (CH3OH). Concentration-time profiles of isopropyl ethanoate and the detected products, with the exception of methanol, are shown in Fig. 3. The upward curvature of both the AAA and HCHO concentration-time profiles shows that they are primary products. The flattening and downward trend of the HCHO concentration-time profile at later stages in the reaction shows that it is being consumed in secondary reactions with Cl atoms. The slow initial formation of CO and acetic acid and the continually increasing concentrations with reaction time are consistent with products formed in secondary reactions. Figure 4 shows a plot of the amount of AAA formed as a function of the amount of isopropyl ethanoate reacted. The slope of the plot gives a value of 84±9 % for the molar yield of AAA in the reaction of Cl atoms with isopropyl ethanoate.
Environ Sci Pollut Res Table 1 Rate coefficient ratio kester/kref and rate coefficients for the reaction of Cl atoms with a series of esters determined at 298 ±2 K in 1 atm of synthetic air
Ester
Method
Reference
kester/kref
kester (cm3 molecule−1 s−1)
Ethyl ethanoate Propyl ethanoate
FTIR FTIR
Ethene Ethene
0.155±0.018 0.700±0.075
(1.71±0.20)×10−11 (7.70±0.83)×10−11
Butyl ethanoate
GC-FID FTIR
Ethyl ethanoate Ethene
4.18±0.37 1.09±0.15
(7.35±0.65)×10−11 (12.00±1.62)×10−11
Pentyl ethanoate Isopropyl ethanoate tert-Butyl ethanoate
GC-FID FTIR FTIR FTIR
Propyl ethanoate Ethene Ethene Ethene
1.84±0.21 1.63±0.14 0.258±0.032 0.185±0.0
(13.23±1.54)×10−11 (17.90±1.53)×10−11 (2.84±0.35)×10−11 (2.04±0.23)×10−11
Ethyl propanoate Propyl propanoate Methyl-2-methyl-propanoate
GC-FID FTIR FTIR FTIR
Propyl ethanoate Ethene Ethene Ethene
0.243±0.0 0.381±0.044 0.890±0.107 0.338±0.039
(1.75±0.19)×10−11 (4.19±0.48)×10−11 (9.84±1.18)×10−11 (3.72±0.43)×10−11
Ethyl-2-methyl-propanoate
GC-FID FTIR
Propyl ethanoate Ethene
0.575±0.053 0.553±0.061
(4.13±0.38)×10−11 (6.08±0.67)×10−11
GC-FID FTIR
Propyl ethanoate Ethene
0.912±0.083 0.342±0.030
(6.56±0.60)×10−11 (3.76±0.33)×10−11
Ethyl-2,2-dimethyl-propanoate Ethyl butanoate Methyl-3-methyl-butanoate
GC-FID FTIR FTIR FTIR
Propyl ethanoate Ethene Ethene Ethene
0.528±0.056 0.486±0.051 0.950±0.076 0.734±0.101
(3.79±0.40)×10−11 (5.35±0.56)×10−11 (10.45±0.84)×10−11 (8.07±1.11)×10−11
Ethyl-2-methyl-butanoate
GC-FID FTIR
Propyl ethanoate Ethene
1.48±0.15 1.28±0.11
(10.64±1.09)×10−11 (14.10±1.19)×10−11
Ethyl-3-methyl-butanoate
GC-FID FTIR
Propyl ethanoate Ethene
1.96±0.13 1.14±0.13
(14.09±0.94)×10−11 (12.50±1.41)×10−11
Methyl-2-methyl-pentanoate
GC-FID FTIR
Propyl ethanoate Ethene
1.70±0.09 1.27±0.15
(12.24±0.65)×10−11 (13.97±1.59)×10−11
Ethyl-2-methyl-pentanoate
FTIR
Ethene
1.57±0.18
(17.20±1.95)×10−11
Methyl-2,2-dimethylpropanoate
Figure 5(a) shows a spectrum obtained after irradiation of a methyl-2-methyl-propanoate/Cl2/air mixture after subtraction
of the absorptions due to methyl-2-methyl-propanoate. Panel (b) shows a reference spectrum of methyl-2-oxopropanoate (MOP: CH3C(O)C(O)OCH3). Visual inspection of the spectra shows quite clearly that MOP is a major reaction product.
Fig. 2 a A spectrum of isopropyl ethanoate (IPE) in a IPE/Cl2/air mixture before irradiation. b The product spectrum obtained after irradiation and subtraction of the spectral features due to IPE. c A reference spectrum of acetic acid anhydride in air
Fig. 3 Concentration-time plot for the Cl atom-induced decay of isopropyl ethanoate and the products CO, HCHO, acetic acid and acetic acid anhydride
Cl+methyl-2-methyl-propanoate
Environ Sci Pollut Res
Fig. 4 Yield plot for the formation of acetic acid anhydride from the Cl atom-initiated photooxidation of isopropyl ethanoate
Other products detected in the product spectrum included CO, HCHO, acetone and formic acid (HC(O)OH). The concentration-time profiles of methyl-2-methyl-propanoate and the reaction products are shown in Fig. 6. The plot contours of HCHO, acetone and methy-oxopropanoate are consistent with primary product formation, whereby the contour of HCHO also shows that it is being consumed in secondary reactions with Cl atoms. The concentration-time contours of CO and HC(O)OH are in line with secondary production through reaction of Cl atoms with products. Figure 7 shows plots of the concentrations of methyl-oxopropanoate and acetone against the amounts of methyl-2-methylpropanoate consumed. The slopes of the plots give yields of methyl-oxopropanoate and acetone of 62±6 and 30±5 %, respectively. CO is a product of the reaction of Cl with HCHO and HC(O)OH a product of the reaction of HCHO with HO2 radicals (Niki et al. 1980). A plot of the combined yields of HCHO, CO and HC(O)OH against the amount of methyl-2methyl-propanoate reacted was linear and gave a combined yield for the three compounds over the course of the reaction of 80±5 %.
Fig. 5 a A spectrum of methyl-2-methyl-propanoate (M2MP) in a M2MP/Cl2/air mixture before irradiation. b The product spectrum obtained after irradiation and subtraction of the spectral features due to M2MP. c A reference spectrum of methyl-2-oxopropanoate in air
consideration the fairly large errors associated with some of the measurements. There are a number of potential difficulties when measuring Cl rate coefficients with both absolute and relative kinetic methods, which are discussed by Xing et al. (2009), and these can account for the sometimes significant differences between some of the rate coefficients, which have been measured using both absolute and relative kinetic methods. Shown in the right-hand column in Table 2 are the values of the rate coefficients for Cl with the esters estimated using the SAR and currently available substituent factors as outlined below. For the calculation of the SAR values of the rate coefficients, the SAR method of Aschmann and Atkinson (1995) developed for the reaction of Cl atoms with alkanes has been adopted with amendments to account for the ester, C(O)O-, functionality. In the SAR method, an overall rate coefficient is calculated from the sum of the estimated group rate coefficients for H atom abstraction from all the -CH3, CH2- and>CH- moieties in the compound. The group rate coefficients depend on the identity of the substituents around the -CH3, -CH2- and>CH- moieties and are defined as
Discussion Kinetics As can be seen from Table 1, in cases where both analytical methods were used to determine the rate coefficient good agreement, within the experimental errors of the measurements, was found in nearly all cases with the only exception being methyl-3-methyl-butanoate where the difference was ~25 %. In Table 2, the esters are listed with their chemical structures and a comparison is made between the rate coefficients measured in this study and those reported in the literature. Where literature rate coefficients are available, the agreement with the rate coefficients measured in this study is acceptable in the majority of cases when one takes into
Fig. 6 Concentration-time plot for the Cl atom-induced decay of methyl2-methyl-propanoate and the products CO, HCHO, acetone and methyl2-oxopropanoate
Environ Sci Pollut Res
Fig. 7 Yield plot for the formation of acetone and methyl-2oxopropanoate from the Cl atom-initiated photooxidation of methyl-2methyl-propanoate
k ðCH3 ‐X Þ ¼ k prim F ðX Þ ; k ðX ‐CH2 ‐Y Þ ¼ k sec F ðX Þ F ðY Þ ; k ðX ‐CH‐Y ðZ ÞÞ ¼ k tert F ðX Þ F ðY ÞF ðZ Þ
where kprim, ksec and ktert are the rate coefficients per -CH3, -CH2-, and >CH- moiety for a standard substituent; X, Y and Z are the substituent groups; and F(X), F(Y) and F(Z) are the substituent factors. In this method, the standard substituent group is CH3with F(CH3-)=1. For the SAR estimations, we have used kprim = 3.32, ksec = 8.34 and ktert = 6.09 (all in units of 10−11 cm3 molecule−1 s−1) and F(-CH2-)=F(>CH-)=0.79 as reported by Aschmann and Atkinson (1995). Other values for kprim, ksec and ktert are available in the literature (Seakan and Quam 1992; Tyndall et al. 1997), whereby those of Tyndall et al. differ only marginally from those of Aschmann and Atkinson and those of Senkan and Quam for ksec and ktert differ quite significantly. Xing et al. (2009) in their study on the reactions of Cl atoms with a series of acetates have derived several substituent factors for the ester functionality in acetates, namely F(CH3C(O)O-)=0.066, F(CH3C(O)OCH2-)=0.36 and F(CH3C(O)OCH
ATMOSPHERIC POLLUTANTS IN A CHANGING ENVIRONMENT
Reactions of Cl atoms with alkyl esters: kinetic, mechanism and atmospheric implications Stefanie Ifang & Thorsten Benter & Ian Barnes
Received: 3 February 2014 / Accepted: 14 April 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Rate coefficients have been measured for the reaction of Cl atoms with a series of alkyl esters at 298±2 K and atmospheric pressure in a large volume photoreactor using the relative kinetic technique. The kinetic data have been used in conjunction with other literature studies on the reactions of Cl atoms with esters to revise the existing values for ester substituent factors in a structure activity relationship (SAR) for Cl reactions. Product studies are reported for the reactions of Cl atoms with isopropyl ethanoate and methyl-2-methylpropanoate under NOx-free conditions. These studies highlight the types of products that can be expected when oxidation occurs at R groups on the acyl (-C(O)OR) and oxy (RC(O)O-) sides of the ester functionality where R is a straight or branched chain alkyl entity. Possible atmospheric repercussions of the atmospheric chemistry of esters are considered. Keywords Alkyl esters . Chlorine atoms . Oxidation mechanism . Relative kinetic . Structure activity relationship
Introduction Alkyl esters (RC(O)OR, R=alkyl group) are an important class of oxygenated volatile organic compounds (OVOCs) which are emitted directly to the atmosphere from both natural and anthropogenic sources. They also are produced in situ in the atmosphere via the oxidation of ethers. Natural sources of Responsible editor: Gerhard Lammel Electronic supplementary material The online version of this article (doi:10.1007/s11356-014-2913-9) contains supplementary material, which is available to authorized users. S. Ifang : T. Benter : I. Barnes (*) FB C - Department of Physical Chemistry, University of Wuppertal, Gauss Strasse 20, 42119 Wuppertal, Germany e-mail: [email protected]
esters include emissions from vegetation and fruit. Esters are used to varying degrees as solvents in the manufacture of fats, cellulose, varnishes, paints and insecticides and as reagents during the manufacture of food flavourings (Calvert et al. 2011; Graedel et al. 1986; Koppmann 2007). They are often added to perfumes since many have pleasant fruity odours (Paillard 1990; Rowan et al. 1996). Some alkyl and cyclic esters (lactones) are under consideration as candidates for biofuels (Alonso et al. 2013; Chan-Thaw et al. 2013; Hayes and Burgess 2009; Huynh and Violi 2008). Anthropogenic emissions of the unsaturated esters to the atmosphere can occur during their production, processing, storage and disposal. In the atmosphere, saturated esters will be subject to photooxidation initiated mainly by OH radicals during the daytime and by NO3 radicals during the night. In certain environments, Cl atom-initiated oxidation will also be important. Conventionally, the major sources of Cl atoms in the troposphere have been thought to be (i) the photolysis of molecular chlorine for which a number of possible heterogeneous production sources have been proposed (Finlayson-Pitts et al. 1989; Finley and Saltzman 2006, 2008; Vogt et al. 1996) and (ii) the reaction of OH radicals with hydrogen chloride. On a global scale, the atmospheric Cl atom concentration is low, ~103 molecules cm−3 (Rudolph et al. 1996; Singh et al. 1996), and atmospheric levels sufficiently high enough to be of significance for VOC decay have been thought to have been confined largely to coastal areas and possibly some polluted industrial regions with molecular chlorine emission sources (Finley and Saltzman 2006, 2008; Spicer et al. 1998; Wingenter et al. 1999, 2005). However, recent field work has presented evidence for Cl chemistry in continental regions remote from coastal regions involving the photolysis of nitryl chloride (ClNO2) as the Cl atom source suggesting that Clinitiated photooxidation chemistry may play a more important role than previously thought in polluted continental regions
Environ Sci Pollut Res
(Mielke et al. 2011; Philips et al. 2012; Riedel et al. 2012, 2013; Sarwar and Bhave 2007; Thornton et al. 2010). Presented here is a relative kinetic and limited product study on the reaction of Cl atoms with a series of alkyl esters. The aims of the study are to extend the kinetic database for esters, improve structure reactivity relationships (SARs) for Cl atom reactions and provide photooxidation mechanisms for esters for inclusion in chemistry transport (CT) models. Of the 18 alkyl esters investigated in this work, room temperature rate coefficients have been reported previously in the literature for nine of the esters. Rate coefficients have been reported for the reactions of Cl with ethyl, propyl, butyl, isopropyl and tert-butyl ethanoate using both absolute (Notario et al. 1998; Cuevas et al. 2005; Xing et al. 2009) and relative kinetic (Xing et al. 2009) techniques. Rate coefficient determinations using a relative kinetic technique have also been reported for the reactions of Cl with ethyl propanoate (Andersen et al. 2012; Cometto et al. 2009), propyl propanoate, methyl-2-methyl and ethyl butanoate (Cometto et al. 2009). For the remainder of the ester compounds, this study represents the first-time determination of their room temperature rate coefficient with Cl atoms.
Experimental The experiments have been performed in two large volume photoreactors with volumes of 1,080 and 480 L at 298±2 K and 760±10 Torr total pressure of synthetic air. The reactors are equipped with fluorescent lamps (Philips TLA 40W/05, 300≤λ≤450 nm, λmax =360 nm) which are used for the photolytic production of radicals. White-type mirror systems are mounted internally within each of the chambers and coupled to FTIR spectrometers (Fourier transform-infrared spectrometer; Nicolet Nexus and Nicolet Magna 550) each equipped with a globar as IR source and a liquid nitrogen cooled MCT detector (mercury-cadmium-telluride). The setups enable in situ infrared monitoring of both reactants and products. The White systems in the reactors were operated with total optical absorption paths of 487 m (1,080 L) and 50.4 m (480 L), and infrared spectra were recorded with a spectral resolution of 1 cm−1. Further details on the reactors can be found in Barnes et al. (1993, 1994). In some cases GCFID was also used to monitor the reactants. Samples, taken from the reactor in gas-tight syringes, were injected into the inlet port of a Hewlett Packard HP5890A Series II gas chromatograph containing a 30m-long, 0.53 mm ID capillary RTX 225 (fused silica) column from Restek operated at 60 °C.
Chlorine atoms were produced in both reactors by the photolysis of molecular chlorine. Cl2 þ hν→Cl þ Cl Experiments were performed on mixtures of Cl2/ester/ethene/air. In the case of the measurements using FTIR, the mixtures were irradiated for periods of 15–20 min during the course of which infrared spectra were recorded at 1 cm−1 resolution with the FTIR spectrometer. Typically, 64 interferograms were co-added per spectrum over a period of approximately 1 min. For the experiments using GC-FID for the analysis, the mixtures were irradiated for 40 min and samples taken from gas-tight syringes as indicated above. Product analyses were performed on the reactions of Cl atoms with isopropyl ethanoate and methyl-2-methylpropanoate using FTIR for in situ monitoring of the reactants and products. The experimental conditions used were similar to those given above for the kinetic experiments, the only difference being the omission of the reference hydrocarbon from the reaction mixture. The initial concentrations used in the experiments in ppmV (1 ppmV=2.46×1013 molecule cm−3 at 298 K and 760 Torr of total pressure) were typically 6–10 for the esters and the reference compound and between 10 to 15 ppm for Cl2. The chemicals used in the experiments had the following purities as given by the manufacturer and were used as supplied: synthetic air 99.999 % (Air Liquide), esters ≥98 % (Aldrich and Alfa Aesar), ethane 99.5 % (Aldrich) and chlorine 99.8 % (Messer Griesheim).
Data evaluation and results Kinetic studies In the presence of the oxidant, in this study, the Cl atom, the ester and reference compound decay through the following reactions: ester þ Cl→products; k ester
ð1Þ
reference þ Cl→products; k ref
ð2Þ
Provided that the ester under investigation and the reference compound are lost only by reactions (1) and (2), then it can be shown that: ½Estert0 k ester ½Referencet0 ln ln ¼ ð3Þ ½Estert k ref ½Referencet where [Ester]t0, [Reference]t0, [Ester]t and [Reference]t are the concentrations of the ester and reference compound at times
Environ Sci Pollut Res
t=0 and t, respectively, and kester and kref are the rate coefficients of reactions (1) and (2), respectively. The relative rate technique relies on the assumption that both the ester and reference compound are removed solely by reaction with the Cl atoms. To verify this assumption, various tests were performed to assess possible contributions to the decays of the organics by reaction with the radical precursor Cl2, photolysis and loss to the reactor surface. Mixtures containing an ester, a reference compound and the radical precursor Cl2 were prepared and allowed to stand in the dark for 1 h. In all cases, loss of both the ester and reference compound in the presence of the radical precursor and in the absence of UV light was negligible showing that both wall deposition and reaction of the compounds with Cl2 are negligible loss processes. Additionally, to test for possible photolysis of the esters, mixtures of the esters alone in air were irradiated for 30 min. No measurable loss of the esters within this time period was observed showing that photolysis is also a negligible loss process for the esters. In the experiments using FTIR where ethene was used as the reference compound, the products formed from its reaction with Cl atoms, mainly HCHO and HCClO, did not interfere with the spectral analysis. The kinetic data obtained from the experiments on the reactions of Cl with the 18 alkyl esters using FTIR have been plotted according to Eq. (3). Figure 1 shows example plots for four of the esters. Plots for the remainder are shown in Figs. SI-1 to SI-4 in the Supporting Information (SI). In all cases, reasonably good linear relationships with zero intercept were obtained. Similarly, good linear relationships were obtained for the alkyl esters investigated using GC-FID for the reactant analysis. For all of the alkyl ester and reference compound combinations, a minimum of three experiments was performed. Table 1 lists the alkyl esters investigated, the method applied for the analysis, the reference compound used and the average rate coefficient ratio kester/kref obtained for each ester and reference combination investigated. Ethene, with a rate coefficient of kCl+ethene = 1.1×10−10 cm3 molecule−1 s−1 (Atkinson et al. 2006), was used as a reference compound for the experiments employing FTIR, and ethyl ethanoate (EE) and propyl ethanoate (PE), with rate coefficients of kCl + EE =1.76× 10−11 and kCl+PE =7.19×10−11 cm3 molecule−1 s−1 (Xing et al. 2009), respectively, were used for the experiments employing GC-FID for the analysis. The absolute values of the rate coefficients, kester, obtained for the reaction of Cl with the alkyl esters from the rate coefficient ratios kester/kref using the appropriate value of the rate coefficient for the reference compound are listed in the right-hand column in Table 1. The quoted errors for the rate coefficients are the 2σ errors from the linear regression analyses of the data plots plus an additional 10 % to account for the uncertainty in the reference compound rate coefficient.
Fig. 1 Examples of plots of the kinetic data for the reactions of Cl atoms with ethyl-2-methyl-pentanoate, ethyl-2-methyl-butanoate, ethyl-2methylpropanoate and methyl-2-methyl-propanoate relative to the reaction of Cl with ethene obtained at 298±2 K and atmospheric pressure using in situ FTIR for the analysis
In cases where both analytical methods were used to determine the rate coefficient, good agreement was found in nearly all cases. The only exception was for methyl-3-methylbutanoate where the difference was ~25 %.
Product studies Cl+isopropyl ethanoate Figure 2(a) shows a spectrum obtained after irradiation of an isopropyl ethanoate/Cl2/air mixture and subtraction of the absorptions due to isopropyl ethanoate. Panel (b) shows a reference spectrum of acetic acid anhydride (AAA: CH3C(O)OC(O)CH3). Visual inspection of the spectra shows quite clearly that AAA is a major reaction product. Other products detected in the product spectrum included CO, formaldehyde (HCHO), formic acid (HC(O)OH), acetic acid (CH3C(O)OH) and minor amounts of methanol (CH3OH). Concentration-time profiles of isopropyl ethanoate and the detected products, with the exception of methanol, are shown in Fig. 3. The upward curvature of both the AAA and HCHO concentration-time profiles shows that they are primary products. The flattening and downward trend of the HCHO concentration-time profile at later stages in the reaction shows that it is being consumed in secondary reactions with Cl atoms. The slow initial formation of CO and acetic acid and the continually increasing concentrations with reaction time are consistent with products formed in secondary reactions. Figure 4 shows a plot of the amount of AAA formed as a function of the amount of isopropyl ethanoate reacted. The slope of the plot gives a value of 84±9 % for the molar yield of AAA in the reaction of Cl atoms with isopropyl ethanoate.
Environ Sci Pollut Res Table 1 Rate coefficient ratio kester/kref and rate coefficients for the reaction of Cl atoms with a series of esters determined at 298 ±2 K in 1 atm of synthetic air
Ester
Method
Reference
kester/kref
kester (cm3 molecule−1 s−1)
Ethyl ethanoate Propyl ethanoate
FTIR FTIR
Ethene Ethene
0.155±0.018 0.700±0.075
(1.71±0.20)×10−11 (7.70±0.83)×10−11
Butyl ethanoate
GC-FID FTIR
Ethyl ethanoate Ethene
4.18±0.37 1.09±0.15
(7.35±0.65)×10−11 (12.00±1.62)×10−11
Pentyl ethanoate Isopropyl ethanoate tert-Butyl ethanoate
GC-FID FTIR FTIR FTIR
Propyl ethanoate Ethene Ethene Ethene
1.84±0.21 1.63±0.14 0.258±0.032 0.185±0.0
(13.23±1.54)×10−11 (17.90±1.53)×10−11 (2.84±0.35)×10−11 (2.04±0.23)×10−11
Ethyl propanoate Propyl propanoate Methyl-2-methyl-propanoate
GC-FID FTIR FTIR FTIR
Propyl ethanoate Ethene Ethene Ethene
0.243±0.0 0.381±0.044 0.890±0.107 0.338±0.039
(1.75±0.19)×10−11 (4.19±0.48)×10−11 (9.84±1.18)×10−11 (3.72±0.43)×10−11
Ethyl-2-methyl-propanoate
GC-FID FTIR
Propyl ethanoate Ethene
0.575±0.053 0.553±0.061
(4.13±0.38)×10−11 (6.08±0.67)×10−11
GC-FID FTIR
Propyl ethanoate Ethene
0.912±0.083 0.342±0.030
(6.56±0.60)×10−11 (3.76±0.33)×10−11
Ethyl-2,2-dimethyl-propanoate Ethyl butanoate Methyl-3-methyl-butanoate
GC-FID FTIR FTIR FTIR
Propyl ethanoate Ethene Ethene Ethene
0.528±0.056 0.486±0.051 0.950±0.076 0.734±0.101
(3.79±0.40)×10−11 (5.35±0.56)×10−11 (10.45±0.84)×10−11 (8.07±1.11)×10−11
Ethyl-2-methyl-butanoate
GC-FID FTIR
Propyl ethanoate Ethene
1.48±0.15 1.28±0.11
(10.64±1.09)×10−11 (14.10±1.19)×10−11
Ethyl-3-methyl-butanoate
GC-FID FTIR
Propyl ethanoate Ethene
1.96±0.13 1.14±0.13
(14.09±0.94)×10−11 (12.50±1.41)×10−11
Methyl-2-methyl-pentanoate
GC-FID FTIR
Propyl ethanoate Ethene
1.70±0.09 1.27±0.15
(12.24±0.65)×10−11 (13.97±1.59)×10−11
Ethyl-2-methyl-pentanoate
FTIR
Ethene
1.57±0.18
(17.20±1.95)×10−11
Methyl-2,2-dimethylpropanoate
Figure 5(a) shows a spectrum obtained after irradiation of a methyl-2-methyl-propanoate/Cl2/air mixture after subtraction
of the absorptions due to methyl-2-methyl-propanoate. Panel (b) shows a reference spectrum of methyl-2-oxopropanoate (MOP: CH3C(O)C(O)OCH3). Visual inspection of the spectra shows quite clearly that MOP is a major reaction product.
Fig. 2 a A spectrum of isopropyl ethanoate (IPE) in a IPE/Cl2/air mixture before irradiation. b The product spectrum obtained after irradiation and subtraction of the spectral features due to IPE. c A reference spectrum of acetic acid anhydride in air
Fig. 3 Concentration-time plot for the Cl atom-induced decay of isopropyl ethanoate and the products CO, HCHO, acetic acid and acetic acid anhydride
Cl+methyl-2-methyl-propanoate
Environ Sci Pollut Res
Fig. 4 Yield plot for the formation of acetic acid anhydride from the Cl atom-initiated photooxidation of isopropyl ethanoate
Other products detected in the product spectrum included CO, HCHO, acetone and formic acid (HC(O)OH). The concentration-time profiles of methyl-2-methyl-propanoate and the reaction products are shown in Fig. 6. The plot contours of HCHO, acetone and methy-oxopropanoate are consistent with primary product formation, whereby the contour of HCHO also shows that it is being consumed in secondary reactions with Cl atoms. The concentration-time contours of CO and HC(O)OH are in line with secondary production through reaction of Cl atoms with products. Figure 7 shows plots of the concentrations of methyl-oxopropanoate and acetone against the amounts of methyl-2-methylpropanoate consumed. The slopes of the plots give yields of methyl-oxopropanoate and acetone of 62±6 and 30±5 %, respectively. CO is a product of the reaction of Cl with HCHO and HC(O)OH a product of the reaction of HCHO with HO2 radicals (Niki et al. 1980). A plot of the combined yields of HCHO, CO and HC(O)OH against the amount of methyl-2methyl-propanoate reacted was linear and gave a combined yield for the three compounds over the course of the reaction of 80±5 %.
Fig. 5 a A spectrum of methyl-2-methyl-propanoate (M2MP) in a M2MP/Cl2/air mixture before irradiation. b The product spectrum obtained after irradiation and subtraction of the spectral features due to M2MP. c A reference spectrum of methyl-2-oxopropanoate in air
consideration the fairly large errors associated with some of the measurements. There are a number of potential difficulties when measuring Cl rate coefficients with both absolute and relative kinetic methods, which are discussed by Xing et al. (2009), and these can account for the sometimes significant differences between some of the rate coefficients, which have been measured using both absolute and relative kinetic methods. Shown in the right-hand column in Table 2 are the values of the rate coefficients for Cl with the esters estimated using the SAR and currently available substituent factors as outlined below. For the calculation of the SAR values of the rate coefficients, the SAR method of Aschmann and Atkinson (1995) developed for the reaction of Cl atoms with alkanes has been adopted with amendments to account for the ester, C(O)O-, functionality. In the SAR method, an overall rate coefficient is calculated from the sum of the estimated group rate coefficients for H atom abstraction from all the -CH3, CH2- and>CH- moieties in the compound. The group rate coefficients depend on the identity of the substituents around the -CH3, -CH2- and>CH- moieties and are defined as
Discussion Kinetics As can be seen from Table 1, in cases where both analytical methods were used to determine the rate coefficient good agreement, within the experimental errors of the measurements, was found in nearly all cases with the only exception being methyl-3-methyl-butanoate where the difference was ~25 %. In Table 2, the esters are listed with their chemical structures and a comparison is made between the rate coefficients measured in this study and those reported in the literature. Where literature rate coefficients are available, the agreement with the rate coefficients measured in this study is acceptable in the majority of cases when one takes into
Fig. 6 Concentration-time plot for the Cl atom-induced decay of methyl2-methyl-propanoate and the products CO, HCHO, acetone and methyl2-oxopropanoate
Environ Sci Pollut Res
Fig. 7 Yield plot for the formation of acetone and methyl-2oxopropanoate from the Cl atom-initiated photooxidation of methyl-2methyl-propanoate
k ðCH3 ‐X Þ ¼ k prim F ðX Þ ; k ðX ‐CH2 ‐Y Þ ¼ k sec F ðX Þ F ðY Þ ; k ðX ‐CH‐Y ðZ ÞÞ ¼ k tert F ðX Þ F ðY ÞF ðZ Þ
where kprim, ksec and ktert are the rate coefficients per -CH3, -CH2-, and >CH- moiety for a standard substituent; X, Y and Z are the substituent groups; and F(X), F(Y) and F(Z) are the substituent factors. In this method, the standard substituent group is CH3with F(CH3-)=1. For the SAR estimations, we have used kprim = 3.32, ksec = 8.34 and ktert = 6.09 (all in units of 10−11 cm3 molecule−1 s−1) and F(-CH2-)=F(>CH-)=0.79 as reported by Aschmann and Atkinson (1995). Other values for kprim, ksec and ktert are available in the literature (Seakan and Quam 1992; Tyndall et al. 1997), whereby those of Tyndall et al. differ only marginally from those of Aschmann and Atkinson and those of Senkan and Quam for ksec and ktert differ quite significantly. Xing et al. (2009) in their study on the reactions of Cl atoms with a series of acetates have derived several substituent factors for the ester functionality in acetates, namely F(CH3C(O)O-)=0.066, F(CH3C(O)OCH2-)=0.36 and F(CH3C(O)OCH
