Atmospheric Chemistry of Acetylacetone - American Chemical Society

Sep 24, 2008 - Chemistry, University of Iasi, Carol I Boulevard 11,. 700 506 Iasi, Romania. Received April 14, 2008. Revised manuscript received. Augu...
0 downloads 0 Views 197KB Size
Download PDF
Environ. Sci. Technol. 2008, 42, 7905–7910

Atmospheric Chemistry of Acetylacetone S H O U M I N G Z H O U , †,‡ I A N B A R N E S , * ,‡ T O N G Z H U , † I U S T I N I A N B E J A N , ‡,§ M I H A E L A A L B U , ‡,§ A N D THORSTEN BENTER‡ State Key Joint Laboratory for Environmental Simulation and Pollution Control, College of Environmental Science, Peking University, 100871 Beijing, China, Physikalische Chemie/FBC, Bergische Universitaet Wuppertal, Gauss Strasse 20, D-42097 Wuppertal, Germany, and “Al.I. Cuza” Faculty of Chemistry, Department of Inorganic and Analytical Chemistry, University of Iasi, Carol I Boulevard 11, 700 506 Iasi, Romania

chelate compounds with a wide range of transition metals. It is also a building block for the synthesis of heterocyclic compounds and is the raw material for sulfonamide drugs. The current global capacity for AcAc is estimated at 20 000 t a-1 with Japan, the United States, and Europe being the biggest users and suppliers; http://findarticles.com/p/ articles/mi_hb048/is_200005/ai_hibm1G162871527. Its presence in the atmosphere will result mainly from industrial releases and not from in situ atmospheric formation. Due to its wide application in industry, there is a need to assess its atmospheric fate and any potential detrimental environmental impacts. AcAc is a diketone, however, it belongs to the category of R, β-enones due to keto-enolic tautomerization:

Received April 14, 2008. Revised manuscript received August 11, 2008. Accepted August 11, 2008.

A kinetic study on the reactions of the OH radical and ozone with acetylacetone (AcAc) has been performed in a 1080 L quartz glass reaction chamber using in situ FTIR spectroscopy analysis. Temperature dependent rate coefficients for the reaction of AcAc with the OH radical were determined over the temperature range 285-310 K using the relative kinetic method. The following Arrhenius expression was derived: k ) 3.35 × 10-12 exp((983 ( 130)/T) cm3 molecule-1 s-1, where the indicated error is the two least-squares deviation. A rate coefficient (in units of cm3 molecule-1 s-1) of (1.03 ( 0.31) × 10-18 has been obtained at (298 ( 3) K for the reaction of ozone with AcAc. A product investigation on the gas-phase reaction of OH radical with AcAc was conducted in a 405 L borosilicate glass chamber using in situ FTIR spectroscopy to monitor reactants and products. Methylglyoxal, acetic acid, peroxy acetic nitrate (PAN) were positively identified as products with molar yields of (20.8 ( 4.5)%, (16.9 ( 3.4)%, and (2.0 ( 0.5)%, respectively. From the residual infrared spectrum the main products are attributed to 2,3,4-pentantrione (CH3sCOsCOsCOsCH3) and its hydrated analogue pentan-2,3dione-4-diol (CH3sCOsCOsC(OH)2sCH3). Based on the observed products, a simplified mechanism for the reaction of the OH radical with AcAc is proposed.

Introduction Ketones are widely used by industry and also produced in the atmospheric oxidation of volatile organic compounds (VOCs). From the available kinetic and mechanistic studies it is well-known that the atmospheric degradation of saturated carbonyl compounds is mainly initiated by reaction with the OH radical and photolysis, whereas losses of the carbonyls through reaction with O3 and NO3 are of minor importance (1-3). Acetylacetone (AcAc) is used extensively in industry as an additive and is an important reagent for the preparation of * Corresponding author phone: +49 202 439 2510; fax: +49 202 439 2505; e-mail: [email protected]. † College of Environmental Science. ‡ Bergische Universitaet Wuppertal. § University of Iasi. 10.1021/es8010282 CCC: $40.75

Published on Web 09/24/2008

 2008 American Chemical Society

This equilibrium exists in both the solution and gas phases, and numerous experimental and theoretical studies on the equilibrium and spectroscopic and photochemical properties of AcAc have been reported in the literature (4-21). Experimental and theoretical studies have concluded that more than 90% of gas-phase AcAc exists as the enolic isomer, which has intramolecular hydrogen bonding and a conjugated π-electron system (12). Using the pulsed laser photolysis-laser induced fluorescence (PLP-LIF) technique, rapid production of OH radicals has been observed after photoexcitation of the enolic form of AcAc to its π-π* state (17-20). Studies on the atmospheric chemistry of AcAc are very limited (21-25). Using the photolysis of AcAc as the source of OH radicals Holloway et al. (23) and Bell et al. (25) have reported very fast rate coefficients for the reaction of OH with AcAc at 298 K of 8.8 and 7.9 × 10-11 cm3 molecule-1 s-1, respectively. These high values are in sharp contrast to an earlier study performed by Dagaut et al. (22) who reported the much lower rate coefficient of 0.12 × 10-11 cm3 molecule-1 s-1. Holloway et al. (23) have attributed this large discrepancy to additional OH radicals generated from the photolysis of AcAc by the flash photolysis technique employed by Dagaut et al. (22), which would cause problems in the analysis of the OH decays. Holloway et al. (24) have also determined a rate coefficient for the reaction of ozone with AcAc. Interestingly, Bell et al. (25) found that the reaction of OH with AcAc has a strong negative temperature dependence, i.e., the reaction becomes significantly faster with decreasing temperature. They report an activation parameter (Ea/R) of 1260 K. To the best of our knowledge, no other experimental information is currently available in the peer-reviewed literature on the gasphase atmospheric chemistry of AcAc. In order to improve the atmospheric chemistry database for AcAc, we report here kinetic and product studies on the OH radical and ozone initiated oxidation of AcAc.

Experimental Section Kinetic Studies. All the kinetic studies were performed in a 1080 L quartz glass reactor (26). The chamber is equipped with 32 super actinic fluorescent lamps (Philips TL 05/40 W: 320 e λ e 480 nm, λmax ) 360 nm) and 32 low-pressure mercury lamps (Philips TUV40W; λmax ) 254 nm). A Whitetype mirror system (total optical absorption path of 484.7 m) coupled to an FTIR spectrometer enables in situ infrared monitoring of reactants in the ppmV range. VOL. 42, NO. 21, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7905

The relative kinetic technique was used to investigate the kinetics of the reactions of OH radicals and ozone with AcAc. The relative disappearance of AcAc and reference compounds, whose rate coefficients with the reactive species are reliably known, were monitored in the present of OH or O3, which react with AcAc and the reference compound: AcAc + OH f products, k1

(R1)

reference + OH f products, k2

(R2)

AcAc + O3 f products, k3

(R3)

reference + O3 f products, k4

(R4)

Additionally, AcAc could be loss to the reactor walls or photolyze in the OH experiments (R5). (R5)

AcAc(wall loss, hν) f products, k5

In the ozonolysis experiments only wall loss needs to be considered (R6). (R6)

AcAc(wall loss) f products, k6

Test experiments showed that wall loss and photolysis was negligible for all the reference compounds employed. The following equation has been used to evaluate the kinetic data: ln

{

}

{

}

k1 [reference]t0 [AcAc]t0 - k5 × (t - t0) ) × ln or [AcAc]t k2 [reference]t k3 [AcAc]t0 [reference]t0 ln - k6 × (t - t0) ) × ln [AcAc]t k4 [reference]t

{

}

{

}

(I)

where [AcAc]t0 and [reference]t0 are the concentrations of AcAc and reference compounds, at time t0, respectively; [AcAc]t and [reference]t are the corresponding concentrations at time t; k1 and k2 are the rate coefficients for the reaction of AcAc and references with OH, and k3 and k4 are the corresponding reactions with O3; k5 is the combined first order loss rate of AcAc due to deposition at the walls and photolysis in the OH kinetic investigations and k6 is the wall loss rate in the ozonolysis experiments. Plots of ln([AcAc]t0/ [AcAc]t) - k5(t - t0) or ln([AcAc]t0/[AcAc]t) - k6(t - t0) versus ln([reference]t0/[reference]t) should give straight lines with slopes of k1/k2, or k3/k4, and zero intercepts for the reactions of AcAc with OH or O3, respectively. The rate coefficients k1 and k3 can be derived from the known rate coefficients k2 and k4, respectively. The corrections for the first order loss processes in the OH and O3 kinetic studies were always between 5 and 10% in both instances and were mainly due to wall loss. Isobutene and isoprene were used as reference compounds to study the reaction of OH with AcAc over the temperature range 285 to 310 K at a total pressure of 760 Torr of synthetic air. A total of 3-4 individual experiments were performed to obtain k at each temperature. The photolysis of methyl nitrite (CH3ONO) with the fluorescent lamps was used for the production of OH radicals, CH3ONO + hν(λmax ) 360 nm) f CH3O + NO

(R7)

CH3O + O2 f CH2O + HO2

(R8)

HO2 + NO f NO2 + OH

(R9)

The initial concentrations (in ppmV, 1ppmV ) 2.46 × 1013 molecule cm-3 at 298 K and 1 atm) of AcAc, the reference compound(s), CH3ONO and NO were approximately 2.0, 1.5-2.5, 0.7-2.0, and 5-10, respectively. All the experiments on the reaction of O3 with AcAc were carried out at (760 ( 10) Torr total pressure and (298 ( 2) K in synthetic air, which contained an excess of cyclohexane 7906

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 21, 2008

FIGURE 1. Plot of the kinetic data according to eq I for the gas-phase reaction of OH with AcAc at (a) 285 K using isobutene as reference and (b) 304 K using isoprene as reference. The plot of (a) is offset by 0.1 for clarity. to scavenge more than 95% of the OH radicals produced by the ozonolysis reactions. Ozone was generated as a mixture in O2 by passing O2 through an ozone generator. The initial concentrations (in ppmV) used in the ozonolysis experiments were: AcAc 1.5-2.5; ethene (used as reference) 1.0-2.7; O3 0.5-1.0; cyclohexane 150. Product Study. The product studies on the reaction of AcAc with the OH radical were carried out in a 405 L borosilicate glass chamber (27, 28) at (760 ( 10) Torr total pressure and (298 ( 3)K in synthetic air. Two types of photolysis sources are available in the chamber: 18 fluorescent lamps (Philips TLA 40 W/05; λmax ) 360 nm) and three low-pressure mercury vapor lamps (Philips TUV40W; λmax ) 254 nm). The photolysis of CH3ONO with the fluorescent lights in the presence of NO was employed as the OH radical source. The initial concentrations of the AcAc, CH3ONO and NO were approximately 2.0, 0.7-2.0, and 5-10 ppmV, respectively. Chemicals. Chemicals, purities and suppliers were as follows: O2 (99.995%), ethane (99.95%), NO (99.5%), and isobutene (99%) were obtained from Messer-Griesheim GmbH; acetylacetone (99%), isoprene (98%), cyclohexane (99.9%) were obtained from Sigma-Aldrich Company, methanol (99.8%) and NaNO2 (99.5%) were purchased from Fluka Company.

Results and Discussion Kinetic Studies. OH Radical Reaction. Examples of the kinetic data, plotted according to eq I, for AcAc at 285 and 304 K are shown in Figure 1. Good linear relationships were generally found for the plots of the data. The rate coefficient ratios k1/k2, obtained from the experiments at different temperatures are listed in Table 1. The rate coefficients, k2, (in cm3 molecule-1 s-1 units) for the reaction of OH with the reference compounds, i.e., isobutene and isoprene, were calculated using the recommended respective Arrhenius expressions k2(T ) 297-425 K) ) 9.47 × 10-12exp(504/T) and k2(T ) 249-425 K) ) 2.54 × 10-11 exp(410/T) cm3 molecule-1 s-1 (29). The quoted errors for k1 are a combination of the leastsquares standard deviations 2σ from the plots plus an additional 20% to account for errors in the rate coefficients of OH with the reference compounds. Figure 2 shows the Arrhenius plot for the reaction of OH with AcAc using the rate coefficients given in Table 1. From the plot of the data the following Arrhenius expression for the reaction of OH with AcAc, valid over the temperature range 285-310 K, has been obtained: k1 ) 3.35 × 10-12exp((983 ( 130)/T) cm3 molecule-1 s-1, where the error is the 2σ standard deviation obtained from the plot.

TABLE 1. Measured Rate Coefficient Ratios, k1/k2, and Values of the Temperature Dependent Rate Coefficients k1 and k2 (in cm3 Molecule-1 s-1) for the Reactions of OH Radical with AcAc and Reference(s) in 760 Torr of Synthetic Air temperature (K)

reference

k1/k2

k2 × 1011

k1 × 1011

285 290 298 304 310

isobutene isoprene isobutene isoprene isobutene

1.92 ( 0.02 0.93 ( 0.02 1.76 ( 0.03 0.88 ( 0.02 1.66 ( 0.03

5.55a 10.47b 5.14a 9.78b 4.81a

10.7 ( 2.2 9.77 ( 1.99 9.05 ( 1.81 8.59 ( 1.77 7.98 ( 1.62

a The rate coefficients were calculated from the recommended Arrhenius expression k2(T ) 297-425 K) ) 9.47 × 10-12 exp(504/T) cm3 molecule-1 s-1 for the reaction of OH with isobutene (29). b The rate coefficients were calculated from the recommended Arrhenius expression k2(T ) 249-425 K) ) 2.54 × 10-11 exp(410/T) cm3 molecule-1 s-1 for the reaction of OH with isoprene (29).

FIGURE 2. Arrhenius plot for the reaction of the OH radical with AcAc over the temperature range 285-310 K. As stated in the Introduction AcAc exists in the gas-phase as an equilibrated mixture of keto and enol tautomers, however, according to the equilibrium data of Nakanishi et al. (12) the fraction of AcAc existing in the enol form will always be 95% or more over the 285-310 K temperature range of the present kinetic investigations. Since there is conjugation between the CdC and CdO entities in the enolic form of AcAc it is expected that the OH radical and ozone initiated oxidation of AcAc will both mainly proceed by the addition of the reactive species to the CdC double bond. For the OH radical kinetic studies H-atom abstraction, mainly from the central sCH2s entity of the diketone form of AcAc but also from the terminal methyl groups, will also contribute to the overall measured rate coefficients for OH reaction with AcAc. Based on structure-activity relationships (SAR) (30) the estimated rate coefficient for H-atom abstraction from the central sCH2s group in the diketone will be 1.42 × 10-11 cm3 molecule-1 s-1 at room temperature, i.e., of similar magnitude to that for the addition reaction of OH to the enol tautomer. H-atom abstraction from the end methyl groups will occur for both the keto and enol forms of AcAc, however, this contribution to the overall rate coefficient will be relatively minor and is expected to be similar for both tautomers. Although H-atom abstraction reactions have an opposite temperature dependence compared to OH radical addition reactions, because of the e5% contribution of the diketone form of AcAc to the overall AcAc in the gas-phase, it is not expected that the H-atom abstraction from the diketone form will have any experimentally discernible influence on the measured rate coefficients within the

experimental precision of the present work. This is expected to be true independent of whether the tautomer re-equilibrium during the course of the experiments is fast or slow. This reasoning is supported by the good linearity of the kinetic data at all temperatures, even at large conversions of AcAc. The rate coefficient (in 10-11 cm3 molecule-1 s-1 units) of (9.05 ( 1.81) measured in the present work for the reaction of the OH radical with AcAc at 298 K agrees well with the value of (8.78 ( 0.58) reported by Holloway et al. (23). It is somewhat higher than the nonpeer reviewed value of 7.89 reported by Bell et al. (25) but the values agree within the combined experimental errors of the determinations. All these values are higher than a value reported by Dagaut et al. (22) by almost 2 orders of magnitude. As stated in the Introduction this discrepancy has been attributed by Holloway et al. (23) to problems arising in the analysis of the OH radical decays in the flash photolysis system employed by Dagaut et al.. The good agreement between the rate coefficient obtained at 298 K in this work with those determined by Holloway et al. (23) using (i) a relative method with different reference compounds to those employed in this study and (ii) an absolute method (PLP-LIF) support that the present determination is free of interferences from possible secondary chemistry. The rate coefficients for the reaction of the OH with AcAc determined in this work exhibit a strong negative temperature dependence, i.e., the rate coefficient increases substantially with decreasing temperature. This observation agrees with the results obtained recently by Bell et al. (25), who determined the rate coefficient for the reaction of OH with AcAc as a function of temperature (250-350 K) and pressure (20-200 Torr) in argon bath gas. The activation parameter (Ea/R) determined in the present work of 983 K is somewhat lower than that reported by Bell et al. (25), who determined a best-fit Arrhenius expression of k ) 1.15 × 10-12 exp(1260/ T) cm3 molecule-1 s-1 from their kinetic investigations. The pre-exponential factor determined in the present work is three times higher than that reported by Bell et al. (25). It can be seen from Table 2, where the rate coefficients for the reaction of the OH radical with AcAc at different temperatures obtained in the present work are compared with those derived from the Arrhenius expression reported by Bell et al. (25), that the rate coefficients of Bell et al. are all slightly lower than those of the present work, however, they agree within the reported combined error limits. In Supporting Information (SI) Figure S1, the absorptions around 1625 cm-1 and 1725 cm-1 show the presence of both keto and enol forms of AcAc in the gas-phase. The strong band around 1625 cm-1 is due to the CdO stretching vibration in the enolic tautomeric form where it is modified by hydrogen bonding and resonance in the ring, while the absorption at 1725 cm-1 is characteristic of the CdO stretching vibration in unchelated ketones (31, 32). The band intensity at 1725 cm-1 increases slightly as the pressure decreases, indicating that the equilibrium moves to the keto tautomer at lower pressure. Such a change in equilibrium could result in a pressure dependence for the reaction rate coefficient. However, both Holloway et al. (23) and Bell et al. (25) in their studies varied the total pressure from 100 to 300 and 20 to 200 Torr, respectively, and observed no pressure dependence for the rate coefficient of OH with AcAc. The good agreement between the rate coefficient for the reaction of OH with AcAc obtained at atmospheric pressure in this work with those obtained by absolute rate methods at pressures between 20 and 300 Torr would seem to indicate that any pressure dependence for the reaction is very small, at least over the pressure range covered by the experiments. O3 Reaction. SI Figure S2 shows the kinetic data for the reaction of ozone with AcAc plotted according to eq I. From a minimum of three experiments a rate coefficient ratio, k3/ VOL. 42, NO. 21, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7907

TABLE 2. Rate Coefficients (in 1011 cm3 Molecule-1 s-1) for the Reaction of AcAc with the OH Radical over the Temperature Range 285-310 K temperature (K)

285

290

298

304

310

present work Bell et al.

10.7 ( 2.2 9.57

9.77 ( 1.99 8.86

9.05 ( 1.81 7.89

8.59 ( 1.77 7.26

7.98 ( 1.62 6.70

k4, of 0.65 was obtained. The rate coefficient, k3, determined for the reaction of ozone with AcAc has been put on an absolute basis using a value of k4(ethene) ) 1.59 × 10-18 cm3 molecule-1 s-1 (29), which resulted in a value of k3 ) (1.03 ( 0.31) × 10-18 cm3 molecule-1 s-1. The error for k3 was calculated in a manner similar to that described for the OH radical rate coefficients. The rate coefficient of (1.03 ( 0.31) × 10-18 cm3 molecule-1 -1 s determined in the present work for the reaction of ozone with AcAc is a factor of 2 higher than the value of (5.23 ( 0.18) × 10-19 cm3 molecule-1 s-1 reported by Holloway et al. (24). The reason for this discrepancy is presently not known. It is presently not possible to discuss reasons for the discrepancy since experimental details were not given in the Holloway et al. (24) symposium proceedings contribution. Compared to its high reactivity toward the OH radical, it is surprising that AcAc is so unreactive toward ozone. The rate coefficients for the reactions of unsaturated alcohols such as methyl butenol, hexenol and pentenol isomers with O3, for example, are of the order 10-16 to 10-17 cm3 molecule-1 s-1 (33-36), i.e., 1 to 2 orders of magnitude faster. Assuming average tropospheric concentrations (in molecules cm-3) of OH radicals and ozone of ca. 1.6 × 106 (12 h daytime average) (37) and 7 × 1011 (24 h average concentration) (38), respectively, results in atmospheric lifetimes of AcAc with respect to reaction with OH and ozone of 1.9 h and 16.1 days, respectively. Reaction with NO3 radicals, photolysis, and reactive uptake are other possible atmospheric loss processes for AcAc. A rate coefficient for the reaction of NO3 with AcAc is not available; however, Orlando et al. (34) have estimated a lifetime of around 20 h for the reaction of the pentenol isomers with NO3 and it is quite probable that the lifetime for the reaction of NO3 with AcAc is longer. AcAc only absorbs very weakly in the actinic region of the atmosphere (12) and photolysis is therefore probably only a minor loss process. This is supported by its low photolysis loss in the experiments; however, its photolysis frequency under atmospheric sunlight conditions needs to be determined to confirm this. Nozie`re and Riemer (39) have estimated a loss rate for AcAc on acidic aerosol of 2 × 10-6 s-1 corresponding to a lifetime of around 139 h for this process. Based on the present rate coefficient for the reaction of the OH radical with AcAc and the above considerations reaction with OH will be the dominant atmospheric gasphase degradation pathway for AcAc. It will therefore be quickly degraded on a local scale by this reaction. Product Study on the OH Radical Reaction. Acetic acid and methylglyoxal have been identified as major products of the reaction of OH radicals with AcAc and formation of peroxy acetic nitrate (PAN) has been observed in low yields. Figure 3 shows typical product spectra recorded during the OH radical initiated oxidation of AcAc. Trace (a) is a product spectrum after 25 min reaction and subtraction of some organic and inorganic products, e.g. HCHO, CO, NO, NO2, and HNO3. Figure 3, traces (b), (c), and (d) show reference spectra of acetic acid, methylglyoxal and PAN, respectively. SI Figure S3 shows a plot of the concentrations of methylglyoxal and acetic acid versus the amount of reacted AcAc from which the yields of these products have been determined. Quantification of acetic acid and methylglyoxal were made using FTIR cross sections recorded in this laboratory; 4.76 × 10-19 cm2 molecule-1 (base 10) for acetic 7908

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 21, 2008

acid at 1796 cm-1 and 1.07 × 10-19 cm2 molecule-1 (base 10) for methylglyoxal at 2835 cm-1. From three experiments, averaged molar formation yields of (20.8 ( 4.5%) and (16.9 ( 3.4%) were obtained for methylglyoxal and acetic acid, respectively. Using the absorption cross section reported by Allen et al. (40) a formation yield of (2.0 ( 0.5%) has been obtained for PAN. The errors are the least-squares standard deviations 2σ from the plots shown in SI Figure S3. After subtraction of these identified products from trace (a) in Figure 3 strong distinct absorption bands remained in the residual spectrum in the OsH stretching region (3484 cm-1), the carbonyl-stretching region (1727.7 cm-1) and also in the fingerprint region at 1367, 1192.9, and 1120.6 cm-1 (see trace (a) in SI Figure S4). Trace (b) in SI Figure S4 shows an infrared spectrum of a sample of 2,3,4-pentanetrione synthesized using the method described by Calvin et al. (41). It can be seen that the infrared spectrum of the synthesized triketone sample exhibits an O-H stretching absorption in the 3497 cm-1 region suggesting that the sample is partly hydrolyzed; if hydrolyzed the compound is probably pentan-2,4-dione-3-diol (CH3s COsC(OH)2sCOsCH3) since hydration at the central carbonyl is expected to dominate (42). Based on mechanistic considerations (see later) and also on the fact that the absorption bands around 3484 (OsH stretching) and 1727.7 cm-1 (carbonyl stretching) in SI Figure S4 trace (a) are slightly shifted in comparison with those in SI Figure S4 trace (b), it is considered that the bands in SI Figure S4 trace (a) probably belong to a mixture of the vicinal triketone 2,3,4pentantrione (CH3sCOsCOsCOsCH3) and its hydrated analogue pentan-2,3-dione-4-diol (CH3sCOsCOsC(OH)2s CH3). Since AcAc exists predominantly in the enolic form in the gas phase at room temperature, as discussed in the kinetic section, the OH radical will be expected to mainly add to the double bond of enolic AcAc and form β-hydroxy intermediates:

FIGURE 3. Product spectra recorded during the reaction of OH radicals with AcAc; (a) product spectrum after subtraction of HCHO, CO, NO, NO2, and HNO3; (b) scaled reference spectrum of acetic acid; (c) scaled reference spectrum of methylglyoxal; (d) scaled reference spectrum of peroxy acetyl nitrate (PAN).

•

CH3C(O)CH ) C(OH)CH3 + OH f CH3C(O)CHC(OH)2CH3 + •

CH3C(O)CH(OH)C(OH)CH3 (R10)

The sequential reactions of the β-hydroxy intermediates with O2 and NO will result in the formation of the corresponding alkoxy radicals: •

CH3C(O)CHC(OH)2CH3+O2 + NO f CH3C(O)CH(O)C(OH)2CH3+NO2 (R11) •

•

CH3C(O)CH(OH)C(OH)CH3+O2 + NO f CH3C(O)CH(OH)C(O)(OH)CH3+NO2 (R12) •

available, quantification of the product yields for these compounds is currently not possible. To the best of our knowledge no studies on the atmospheric chemistry of vicinal tricarbonyls have been published in the literature. In past work in this laboratory the UV absorption spectrum of 2,3,4-pentanetrione has been recorded and its Cl atom initiated oxidation has been investigated; however, these results have never been published (43). The major atmospheric fate of 2,3,4-pentanetrione under tropospheric conditions is likely to be similar to that of 2,3butanedione, i.e., photolysis (44). In addition to photolysis, wet deposition will probably be of importance for the hydrated analogues of 2,3,4-pentanetrione.

Acknowledgments

The alkoxy radical from R11 will either react with O2 to form pentan-2,3-dione-4-diol (R13), which could eliminate a molecule of H2O to form 2,3,4-pentanetrione (R14),

Financial support of this work by the European Commission within the project MOST (contract EVK2-CT-2001-00114) and the EU project EUROCHAMP is gratefully acknowledged.

CH3C(O)CH(O)C(OH)2CH3+O2 f HO2+

Supporting Information Available

•

CH3C(O)C(O)C(OH)2CH3 (R13) CH3C(O)C(O)C(OH)2CH3 + M f CH3C(O)C(O)C(O)CH3+ H2O + M (R14) or it could decompose. The decomposition of the alkoxy radical from R11 could result in methylglyoxal and acetic acid (R15 and R16), CH3C(O)CH(O)C(OH)2CH3 + M f CH3C(O)CHO +

Gas-phase IR spectra of AcAc at 297 K for different pressures of air (Figure S1); a plot of the kinetic data according to eq I for the gas-phase reaction of O3 with AcAc at room temperature (Figure S2); plots of the measured product concentrations as function of the amount of AcAc reacted with OH radicals (Figure S3); and a product spectrum for the reaction of OH radicals with AcAc (Figure S4).This material is available free of charge via the Internet at http:// pubs.acs.org.

•

•

C(OH)2CH3 + M (R15) •

C(OH)2CH3+O2 f CH3COOH + HO2

(R16)

or an acetyl radical and propanal-2,2-diol (R17). The diol could give rise to the formation of methylglyoxal by elimination of a molecule of H2O (R18). •

CH3C(O)CH(O)C(OH)2CH3 f CH3C(O) + CH(O)C(OH)2CH3 (R17) CH(O)C(OH)2CH3 + M f CH3C(O)CHO + H2O + M (R18) PAN can be formed by the consecutive reactions of the acetyl radical with O2 and NO2. •

CH3C(O) + O2 + NO2 f CH3C(O)OONO2

(R19)

Similarly the alkoxy radical from R12 can decompose in two ways: (i) to form acetic acid and methylglyoxal (R20 and R21), CH3C(O)CH(OH)C(O)(OH)CH3 f HOC(O)CH3+ •

CH3C(O)CH(OH) (R20) •

CH3C(O)CH(OH) + O2 f CH3C(O)CH(O) + HO2 (R21) and (ii) to give 2-hydroxy-3-keto-butanoic acid and the methyl radical (R22). CH3C(O)CH(OH)C(O)(OH)CH3 f •

•

CH3C(O)CH(OH)C(O)OH + CH3 (R22) From the above-proposed reaction mechanism the sum of the yields of acetic acid and PAN should equal that of methylglyoxal. Within the experimental error limits this is what is observed in the experiments. Since calibrated spectra of 2,3,4-pentanetrione and its hydrates are currently not

Literature Cited (1) Atkinson, R. Gas-phase Tropospheric chemistry of organic compounds. J. Phys. Chem. Ref. Data, 1994, Monograph, 2. (2) Mellouki, A.; Le Bras, G.; Sidebottom, H. Kinetics and mechanisms of the oxidation of oxygenated organic compounds in the gas-phase. Chem. Rev. 2003, 103, 5077–5096. (3) Wayne, R. P.; Barnes, I.; Biggs, P.; Burrows, J. P.; Canosa.Mas, C. E.; Hjorth, J.; Le.Bras, G.; Moortgat, G. K.; Perner, D.; Poulet, G.; Restelli, G.; Sidebottom, H. The nitrate radical: physics, chemistry and the atmosphere. Atmos. Environ. 1991, 25A, 1– 203. (4) Iglesias, E. Determination of the keto-enol equilibrium constants and the kinetic study of the nitrosation reaction of β-dicarbonyl compounds. J. Chem. Soc., Perkin Trans. 2 1997, 431–439. (5) Ishada, T.; Hirata, F.; Kato, S. Thermodynamic analysis of the solvent effect on tautomerization of acetylacetone: an ab initio approach. J. Chem. Phys. 1999, 110, 3938–3945. (6) Folkend, M. M.; Weiss-Lopez, B. E.; Chauvel, J. P.; True, N. S. Gas-phase 1H NMR studies of keto-enol tautomerization of acetylacetone, methyl acetylacetone, and ethyl acetylacetone. J. Phys. Chem. 1985, 89, 3347–3352. (7) Bauer, S. H.; Wilcox, C. F. On malonaldehyde and acetylacetone: are theory and experiment compatible. Chem. Phys. Lett. 1997, 279, 122–127. (8) Nagashima, N.; Kudoh, S.; Takayanagi, M.; Nakata, M. UVinduced photoisomerization of acetylacetone and identification of less-stable isomers by low-temperature matrix-isolation infrared spectroscopy and density functional theory calculation. J. Phys. Chem. A 2001, 105, 10832–10838. (9) Choudhury, T. K.; Lin, M. C. Homogeneous pyrolysis of acetylacetone at high temperatures in shock waves. Int. J. Chem. Kinet. 1990, 20, 491–504. (10) Al-Awadi, N. A.; El-Nagdi, M. H.; Mathew, T. Gas-phase kinetics for elimination reactions of pentane-2,4-dione derivatives. Int. J. Chem. Kinet. 1995, 27, 517–523. (11) Lazaar, K. I.; Bauer, S. H. Intramolecular unsymmetric OHO bonds. Thermochemistry. J. Phys. Chem. 1983, 87, 2411–2416. (12) Nakanishi, H.; Morita, H.; Nagakura, S. Electronic structures and spectra of the keto and enol forms of acetylacetone. Bull. Chem. Soc. Jpn. 1977, 50, 2255–2261. (13) Tayyari, S. F.; Milani-nejad, F. Vibrational assignment of acetylacetone. Spectrochim. Acta, Part A 2000, 56, 2679–2691. (14) Trivella, A.; Roubin, P.; Theule´.Rajzmann, M.; Coussan, S.; Manca, C. UV and ir photoisomerization of acetylacetone trapped in a nitrogen matrix. J. Phys. Chem. A 2007, 111, 3074– 3081. VOL. 42, NO. 21, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7909

(15) Mohacek-Grosev, V.; Furic, K.; Ivankovic, H. Luminescence and Raman spectra of acetylacetone at low temperatures. J. Phys. Chem. A 2007, 111, 5820–5827. (16) Broadbent, S. A.; Burns, L. A.; Chatterjee, C.; Vaccaro, P. H. Investigation of electronic structure and proton transfer in ground state acetylacetone. Chem. Phys. Lett. 2007, 434, 31–37. (17) Yoon, M.-C.; Chio, Y. S.; Kim, S. K. The OH production from the π-π* transition of acetylacetone. Chem. Phys. Lett. 1999, 300, 207–212. (18) Yoon, M.-C.; Chio, Y. S.; Kim, S. K. Photodissociation dynamics of acetylacetone: the OH product state distribution. J. Chem. Phys. 1999, 110, 11850–11855. (19) Yoon, M.-C.; Chio, Y. S.; Kim, S. K. The OH product state distribution from the photodissociation of hexafluoroacetylacetone. J. Phys. Chem. A 2000, 104, 4352–4355. (20) Upadhyaya, H. P.; Kumar, A.; Naik, P. D. Photodissociation dynamics of enolic- acetylacetone at 266, 248, and 193 nm: mechanism and nascent state product distribution of OH. J. Chem. Phys. 2003, 118, 2590–2598. (21) Chen, X.-B.; Fang, W.-H.; Phillips, D. L. Theoretical studies of the photochemical dynamics of acetylacetone: Isomerzation, dissociation, and dehydration reactions. J. Phys. Chem. A 2006, 110, 4434–4441. (22) Dagaut, P.; Wallington, T. J.; Liu, R.; Kurylo, M. J. A kinetics investigation of the gas-phase reactions of OH radicals with cyclic ketones and diones: mechanistic insights. J. Phys. Chem. 1988, 92, 4375–4377. (23) Holloway, A.-L.; Treacy, J.; Sidebottom, H.; Mellouki, A.; Daele, V.; Le.Bras, G.; Barnes, I. Rate coefficients for the reactions of OH radicals with the keto/enol tautomers of 2,4-pentanedione and 3-methyl-2,4-pentanedione, allyl alcohol and methyl vinyl ketone using the enols and methyl nitrite as photolytic sources of OH. J. Photochem. Photobiol. A 2005, 176, 183–190. (24) Holloway, A.-L.; Sidebottom, H.; Mellouki, A.; Le Bras, G.; Wirtz, K. A Kinetic and Mechanistic Study of Atmospheric Oxidation of 1,3-Diketone. Proceedings of 19th International Symposium on Gas Kinetics. Orleans, France, July 22-27, 2006; pp 61-62. (25) Bell, P.; Nicovich, J. M.; Wine, P. H. Acetylacetone photolysis at 248 nm: hydroxyl radical yield and temperature-dependent rate coefficients for the OH + acetylacetone reaction. Proceedings of 19th International Symposium on Gas Kinetics, Orleans, France, July 22-27, 2006; pp 79-80. (26) Barnes, I.; Becker, K. H.; Mihalopoulos, N. An FTIR product study of the photooxidation of dimethyl disulfide. J. Atmos. Chem. 1994, 18, 267. (27) Barnes, I.; Bastian, V.; Becker, K. H.; Fink, E. H.; Zabel, F. Reactivity studies of organic substances towards hydroxyl radicals under atmospheric conditions. Atmos. Environ. 1982, 16, 545–550. (28) Barnes, I.; Becker, K. H.; Fink, E. H.; Reimer, A.; Zabel, F.; Niki, H. Rate constant and products of the reaction CS2 + OH in the presence of O2. Int. J. Chem. Kinet. 1983, 15, 631–645.

7910

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 21, 2008

(29) Calvert, J. G.; Atkinson, R.; Kerr J. A.; Madronich, S., Moortgat G. K.; Wallington T. J.; Yarwood G. The Mechanisms of Atmospheric Oxidation of the Alkenes; Oxford University Press: Oxford, 2000. (30) Kwok, E. S. C.; Akinson, R. Estimation of hydroxy radical reaction rate constants for gas-phase organic compounds using a structure-reactivity relationship: An update. Atmos. Environ. 1995, 29, 168–1695. (31) Gordy, W. Spectroscopic evidence for hydrogen bonds: effects of chelation on the carbonyl frequency. J. Chem. Phys. 1940, 8, 516–519. (32) Pauling, J.; Bernstein, H. J. The effect of solvents on tautomeric equilibria. J. Am. Chem. Soc. 1951, 73, 4353–4356. (33) Fantechi, G.; Jensen, N. R.; Hjorth, J.; Peeters, J. Mechanistic studies of the atmospheric oxidation of methyl butenol by OH radicals, ozone and NO3 radicals. Atmos. Environ. 1998, 32, 3547– 3556. (34) Orlando, J. J.; Tyndall, G. S.; Ceazan, N. Rate coefficients and product yields from reaction of OH and 1-penten-3-ol, (Z)-2Penten-1-ol, and ally alcohol (2-propen-1-ol). J. Phys. Chem. A 2001, 105, 3564–3569. (35) Atkinson, R.; Arey, J.; Aschmann, S. M.; Corchnoy, S. B.; Shu, Y. Rate constants for the gas-phase reactions of cis-3-hexen1-ol, cis-3-hexenylcaetate, trans-2-hexenal, and linalool with OH and NO3 radicals and O3 at 296 ( 2 K and OH radical yields from the O3 reactions. Int J. Chem. Kinet 1995, 27, 941–955. (36) Barnes, I.; Zhou, S.; Klotz, B. Final Report of the EU project MOST; contract EVK2-CT-2001-00114; European Union: Brussels, August, 2005. (37) Prinn, R. G.; Weiss, R. F.; Miller, B. R.; Huang, J.; Alyea, F. N.; Cunnold, D. M.; Fraser, P. J.; Hartley, D. E.; Simmonds, P. G. Atmospheric trends and lifetime of CH3CCl3 and global OH concerns. Science 1995, 269, 187–192. (38) Logan, J. A. Tropospheric ozone: Seasonal behavior, trends, and anthropogenic influence. J. Geophys. Res. 1985, 90, 10463–10482. (39) Nozie`re, B.; Riemer, D. D. The chemical processing of gas-phase carbonyl compounds by sulfuric acid aerosols: 2,4-pentanedione. Atmos. Environ. 2003, 37, 841–851. (40) Allen, G.; Remedios, J. J.; Newnham, D. A.; Smith, K. M.; Monks, P. S. Improved mid-infrared cross-sections for the peroxyacetyl nitrate (PAN) vapour. Atmos. Chem. Phys. 2005, 5, 47–56. (41) Calvin, M.; Wood, C. L. Conjugation of carbonyl group and the absorption spectrum of triketopentane. J. Am. Chem. Soc. 1940, 62, 3152–3155. ¨ ber das Triketopentan. I. Ber. Chem. (42) Sachs, F.; Barschall, H. U Ges. 1901, 34, 3047–3054. (43) Barnes, I.; Thomas, W. personal communication. (44) Betterton, E. A.; Hoffmann, M. R. Henry′s law constants of some environmentally important aldehydes. Environ. Sci. Technol. 1988, 22, 1415–1418.

ES8010282