Reaction between OH and CH₃CHO

Part 2. Temperature dependent rate coefficients (201–348 K)

Abstract

Absolute rate coefficients for the reaction OH + CH₃CHO → products (1) were measured over the temperature range 201–348 K using the technique of pulsed laser photolytic generation of OH coupled to detection by pulsed laser induced fluorescence. The experiments were carried out at total pressures of 50 or 100 Torr (He or Ar bath gas). The accuracy of the rate constants obtained was enhanced by on-line optical absorption measurements of the acetaldehyde concentration. The temperature dependence of the rate coefficient, expressed in Arrhenius form, is given by k₁(201–348 K) = 4.38 × 10⁻¹² exp{(366 ± 40)/T} cm³ s⁻¹ with the rate coefficient at room temperature of 1.53 × 10⁻¹¹ cm³ s⁻¹. The estimated total error (95% confidence) associated with the rate coefficient derived from this expression is estimated as 5%, and is independent of temperature. The present results, which extend the database on this reaction to cover temperature relevant for the upper troposphere, are compared to previously published measurements, and values of k₁ for atmospheric modelling are recommended.

---

1 Introduction

Recent field measurements have revealed higher than expected concentrations of partially oxidised hydrocarbons (POH) in large areas of the free and upper troposphere (UT).^1,2 The impact of POH in the UT has been investigated in modelling studies that identified a strong influence on the formation rate of odd hydrogen radicals (HOx).^3–7 The resultant enhanced HOx concentrations increase the rates of oxidation processes in the tropopause region and the net photochemical production of O₃, for given levels of NOx, and increase the sensitivity of O₃ production to increasing emissions of NOx from e.g. aircraft and the polluted boundary layer.^8–10 The formation of PAN, an important temporary reservoir for both NOx and HOx, and mediator for NOx transport from more to less polluted regions, was also predicted to be strongly enhanced.³

Much attention has focussed on acetone (CH₃C(O)CH₃) in the UT, with new data on its photolysis and OH sinks at low temperatures becoming available over the last few years.^11,12 Similar to acetone, acetaldehyde is formed in the atmosphere in secondary processes following the OH initiated degradation of non methane hydrocarbons, and is also directly emitted in the boundary layer as a result of biogenic processes.¹³ Acetone has been measured at mixing ratios that frequently approach 2000 ppt¹⁴ whilst those of the shorter lived acetaldehyde are typically 50–100 ppt.²

Whereas acetone is degraded by photolysis and reaction with OH, the fate of acetaldehyde is expected to be predominantly reaction (1), which proceeds mainly by abstraction of the aldehydic H-atom to form the acetyl radical, CH₃CO.¹⁵

OH + CH₃CHO → CH₃CO + H₂O (1)

OH + CH₃C(O)CH₃ → products (2)

CH₃C(O)CH₃ + hν → CH₃ + CH₃CO (3)

CH₃CO + O₂ + M → CH₃C(O)O₂ + M (4)

CH₃C(O)O₂ + NO₂ + M → CH₃C(O)O₂NO₂ + M (5)

Under atmospheric conditions, the acetyl radical formed in reaction (1) or (3) reacts with O₂ [reaction (4)] to form the acetylperoxy radical, CH₃C(O)O₂, which can react with NO₂ [reaction (5)] to form peroxyacetylnitrate, PAN, CH₃C(O)O₂NO₂.¹⁶

The presence of 2000 ppt acetone is accompanied by enhanced OH levels of [OH] ≈ 5 × 10⁵ cm⁻³ in the upper troposphere. Under these conditions, and assuming 50 ppt acetaldehyde, the rates of acetyl formation (R_acetyl) from acetone photolysis and OH reaction with acetaldehyde are, to a first approximation given by R_acetyl = J_acetone[CH₃C(O)CH₃] ≈ 8 × 10⁻⁴ ppt s⁻¹ for acetone photolysis and R_acetyl = k₁[OH][CH₃CHO] ≈ 6 × 10⁻⁴ ppt s⁻¹ from CH₃CHO. In these calculations, J_acetone is the diurnally averaged photolysis rate of acetone in the UT, and is approximately 4 × 10⁻⁷ s⁻¹,^12,17 and k₁ is the rate coefficient for reaction between OH + CH₃CHO, for which a value of ≈2.5 × 10⁻¹¹ cm³ s⁻¹ at 220 K is presently recommended.¹⁶ Clearly, PAN formation from acelaldehyde and acetone in the UT can be comparable. Despite the potential importance of acetaldehyde in the UT, an examination of the kinetic database for the OH reaction shows that relatively little data has been obtained in the relevant temperature regime (200–250 K), as most studies have been carried out at room temperature and above and have focussed on its role in combustion processes. The available literature will be discussed in the context of the present study in a later section.

The goal of the present study was to obtain accurate data on the kinetics of the overall reaction of OH with acetaldehyde [reaction (1)] at low temperatures. An associated study of the product channels in this reaction has recently been published.¹⁵

2 Experimental

The experimental investigation of the title reaction was carried out using the pulsed laser photolysis, pulsed laser induced fluorescence (PLP-PLIF) technique in which OH radicals are generated by pulsed photolysis of suitable precursors on time scales that are short compared to the subsequent decay, and are detected by pulsed laser induced fluorescence. The experimental set-up for PLP-PLIF has been described previously¹¹ and is described only briefly here.

2.1 Pulsed laser photolysis-pulsed laser induced fluorescence

The experiments were carried out using a double jacketed reactor of volume ≈500 cm³. Gas mixtures entered the reactor ≈15 cm upstream of the photolysis region to ensure thermal equilibrium with the walls of the vessel, and were pumped out through ports in the baffle arms, and in front of the fluorescence collection lens. The cell was thermostatted to the desired temperature by circulating a cryogenic fluid through the outer jacket, and the temperature was measured by using a J-Type thermocouple that could be inserted into the reaction volume where the focus of the telescopic lens system and the laser beams intersect. The pressure in the cell, monitored with 10, 100 and 1000 Torr capacitance manometers, was held constant at 50 or 100 Torr, using either Ar or He as bath gases. Typical flow rates, regulated using calibrated mass flow controllers, were between 250 and 500 sccm, resulting in linear gas velocities in the reaction cell of ≈40 cm s⁻¹, ensuring that a fresh gas sample was photolysed at each laser pulse, and prevented build up of products.

OH radicals were generated from suitable precursors (see below) by the ≈20 ns pulse of an excimer laser at 193 or 248 nm. Fluorescence from OH was detected by a photomultiplier tube (PMT) filtered by 309 nm interference filter and a BG 26 glass cut off filter. Excitation of the A ²Σ (v = 1) ← X ²Π (v = 0), Q₁₁ (1) transition at 281.997 nm was achieved using the frequency doubled emission from a Nd-Yag pumped dye laser (Rhodamine 6G). The PMT signals were recorded by a gated box-car integrator. The sensitivity of this set up to OH was estimated by photolysis of a known concentration of OH precursor (e.g. H₂O₂) with a known laser fluence. The detection limit in the absence of CH₃CHO was found to be ≈10⁸ cm⁻³ for a S/N = 1 (20 scans).

2.2 Generation of OH radicals

Two different precursors were used to generate OH radicals in the present experiments. The photolysis of H₂O₂ at 248 nm (KrF excimer laser) was used to generate OH radicals for experiments between 273 and 348 K. Typically, about 10¹⁴ H₂O₂ cm⁻³ were photolysed at a laser fluence of 5–10 mJ cm⁻² to generate ≈(5–10) × 10¹⁰ cm⁻³ of OH. H₂O₂ was transported to the photolysis cell by bubbling a 5 sccm flow of Ar through a concentrated H₂O₂ solution.

H₂O₂ + hν (248 nm) → 2 OH (X ²Π, v = 0) (6)

Below ≈270 K, this source of OH could not be used due to the condensation of H₂O₂ on the reactor wall. For temperatures between 295 and 202 K, OH generation was initiated by the following scheme.

N₂O + hν (193 nm) → O(¹D) + N₂ (7)

O(¹D) + CH₄ → OH(X ²Π, v > 0) + CH₃ (8)

OH(X ²Π, v > 0) + CH₄ → OH(X ²Π, v = 0) + CH₄ (9)

About 1 × 10¹⁴ cm⁻³ of N₂O was photolysed at 193 nm with a fluence of 5–10 mJ cm⁻² per pulse [reaction (7)] to generate O(¹D). This was rapidly converted to OH by reaction (8) in an excess of CH₄. The vibrationally hot¹⁸ OH formed in reaction (8) is quenched to the ground state by CH₄ [reaction (9)] which is a reasonably efficient process for v = 2 (k₉ = 2.3 × 10⁻¹² cm³ s⁻¹).¹⁹ The CH₄ concentration was varied between ≈5 × 10¹⁶ and 5 × 10¹⁷ cm⁻³ in order to find the best experimental conditions that represent a balance between sufficiently rapid formation and quenching of the vibrationally hot OH on a time scale that was short compared to its reaction with CH₃CHO, and loss of the thermalised OH due to reaction with CH₄. At a concentration of 1 × 10¹⁷ CH₄ cm⁻³, vibrationally excited OH will be quenched to OH(X ²Π, v = 0) in a few microseconds.

Evidence for the formation of vibrationally hot OH in the present experiments was provided by the observation of non-exponential OH decays in the absence of CH₄, which is caused by the quenching of higher vibrational states to the X ²Π (v = 0) state on the time-scale of OH decay due to reaction with CH₃CHO. In these experiments the reaction of O(¹D) with CH₃CHO provides the excited OH. Experiments carried out with H₂ used to scavenge the O(¹D) always resulted in non-exponential OH decays as He and H₂ are inefficient quenchers of OH(X ²Π, v > 0).¹⁹

As the rate coefficient of OH with CH₄ is strongly temperature dependent, decreasing by a factor of ≈40 over the temperature range of the present study,¹⁶ the loss of OH by reaction with CH₄ is only significant at the higher temperatures and higher CH₄ concentrations. For example, at 202 K and [CH₄] = 1 × 10¹⁷ cm⁻³ the first order decay coefficient for OH is ≈40 s⁻¹ which is small compared to first order decay coefficients of up to 3 × 10⁴ s⁻¹ for loss of OH by reaction with CH₃CHO. For this reason, the photolysis of N₂O/CH₄/He mixtures as the OH source is particularly well suited for work at lower temperatures. In previous work¹¹ we used the 351 nm photolysis of dry, flowing HONO as OH source at low temperatures. This source proved to be inappropriate for the present experiments as the simultaneous formation of O(³P) (from the photolysis of NO₂ impurity) resulted in secondary OH formation via its reaction with CH₃CHO.

2.3 Optical absorption measurements of CH₃CHO

The concentration of the excess reactant (CH₃CHO) was determined by on-line optical absorption, whereby the absorption by CH₃CHO of UV light from a D₂ lamp was determined in a separate absorption cell, which was equipped with external multi-pass optics to enhance the effective absorption pathlength to 892 cm (8 passes). Tests with static gas mixtures showed that the effective optical absorption pathlength was within a few percent of the geometric one and showed no systematic bias. The entire absorption region of CH₃CHO between 220–340 nm was recorded by diode array spectrograph with a resolution of ∼0.4 nm, and the concentrations were determined by least-squares fitting of the measured data to a reference UV absorption spectrum measured in this laboratory.²⁰ This reference spectrum is in excellent agreement with published^21,22 measurements of the CH₃CHO absorption spectrum, which we consider to be known to an estimated accuracy of ≈5% at wavelengths close to λ_max.

The multi-pass optical absorption cell was connected serially in flow prior to the reactor cell and maintained at room temperature. Corrections for molecular density changes due to the difference in pressure (usually less than 1%) and temperature between the cells were made to calculate the concentration of CH₃CHO in the photolysis cell.

For determining the concentration of CH₃CHO for use in the analysis of the kinetic data, wavelength dependent absorption traces as shown in Fig. 1 were obtained simultaneously with the measurement of an OH decay profile. Fig. 1 shows the measured optical density, scaled reference spectrum and fit residuals. The quality of the least squares fit of the optical density measurements to the UV spectrum of CH₃CHO was such that errors in [CH₃CHO] are estimated at ⩽5%, mainly from errors in the CH₃CHO cross-sections. The fine structure in the residuals at wavelengths longer than 290 nm is due to rotational fine structure in the reference spectrum, which was not observed in the present experimental set-up as it employed a lower spectral resolution. As the fine structure is of negligible amplitude compared to the overall absorption continuum, this has no impact on the overall precision of the fit, or calculation of the concentration.

Fig. 1 Measured optical density over a pathlength of 892 cm due to dilute flowing mixture of CH₃CHO (solid line) along with the least-squares fit using a reference spectrum (dashed line) and the residuals (dotted line). In this particular case the concentration of CH₃CHO was determined as 1.03 × 10¹⁵ cm⁻³, with statistical errors of less than 1%.

This procedure worked well for measurements of CH₃CHO that were taken close in time to measurements of the un-attenuated analysis light intensity, I₀ (empty cell) at the beginning of the experiment. As the experiment progressed, and different CH₃CHO concentrations were used, the quality of the fit was reduced due to changes in the relative intensity of the D₂ lamp. For this reason series of three absorption spectra and OH decays over a range of CH₃CHO were obtained in relatively rapid succession after measuring I₀. The linearity of the optical CH₃CHO measurement with the flow controller reading was then established to derive a conversion factor, which was used to convert subsequent, intermediate mass flow controller readings to CH₃CHO concentrations. As documented previously,²³ we found no evidence for the decomposition of CH₃CHO on passage through the flow controller.

2.4 Chemicals

CH₃CHO (Aldrich, 99.9%), was degassed several times before being transferred and diluted in darkened glass storage bulbs. H₂O₂ (Peroxid-Chemie GmbH, ≈80 wt.%) was concentrated to >90 wt.% by vacuum distillation. N₂O (Aldrich, 99.999%), Ar (Linde, 99.999%), He (Linde, 99.999%), CH₄ (Linde, 99.99%) were used without further purification.

3 Results

The PLP-PLIF studies were carried out under pseudo-first order condition, [CH₃CHO] ≫ [OH], and the decay of OH is described by:

[OH]_t = [OH]₀ exp−{(k₁[CH₃CHO] + d)t} (i)

where [OH]_t is the OH concentration at time = t after the laser pulse, k₁ is the bimolecular rate coefficient for the reaction with CH₃CHO and d accounts for diffusion of OH out of the reaction zone, and reaction of OH with either CH₄ or H₂O₂ depending on the OH source chemistry. The use of low OH concentration and hence low conversion of CH₃CHO ensures that secondary loss of OH, e.g. with CH₃CO or CH₃ can be disregarded fully. This could be confirmed by showing that the variation of the photolysis laser fluence, and thus radical concentrations, by a factor of three had no measurable influence on the OH decay.

Fig. 2 shows two data sets exemplifying the OH decays measured in different excess concentration of CH₃CHO at 296 K using reaction (6) as OH source (A) and at 223 K using reaction sequence (7)–(9) as OH source (B). In both cases the OH decays were exponential over at least 2 half-lives, though the data obtained using reaction sequence (7)–(9) has slightly inferior signal-to-noise characteristics, which is a result of quenching of the fluorescent A ²Σ state of OH by CH₄. The first-order decay rate coefficient, k′, was obtained by non-linear least squares fitting of the OH decays and is related to the desired rate coefficient, k₁ by:

k′ = k₁[CH₃CHO] + d (ii)

Fig. 2 OH decays in the presence of various excess concentrations of CH₃CHO. A: 296 K, H₂O₂ photolysis at 248 nm used as OH source. The CH₃CHO concentrations (in units of cm⁻³) were 1.65 × 10¹⁴ (slowest decay), 2.47 × 10¹⁴, 4.11 × 10¹⁴, 6.16 × 10¹⁴ and 1.03 × 10¹⁵ (fastest decay). B: 223 K, 193 nm photolysis of N₂O/CH₄ used as OH source. The CH₃CHO concentrations (in units of cm⁻³) were 0 (slowest decay), 7.27 × 10¹³, 1.93 × 10¹⁴, 3.93 × 10¹⁴ and 9.93 × 10¹⁴ (fastest decay).

Fig. 3 shows a plot of k′ versus [CH₃CHO] at selected temperatures, and using both OH source reactions. The low scatter in these plots is the result of accurate online measurement of the CH₃CHO concentration and small statistical errors in the pseudo-first-order OH decay coefficients (k′). The measured rate coefficients were found to be independent of the variation of the bath gas pressure (50–100 Torr) and also the bath gas (Ar or He), and the OH generation scheme.

Fig. 3 Plot of k′ versus [CH₃CHO] at different temperatures according to eqn. (ii). The data at 261, 296 and 342 K were obtained using the 248 nm photolysis of H₂O₂ as OH source, the data at 223 K were obtained using the193 nm photolysis of N₂O/CH₄ as OH source.

The complete data set obtained between 201–348 K is listed in Table 1, and shown in Arrhenius format in Fig. 4. Over the range covered, the temperature dependence of the rate coefficient can be accurately reproduced by an equation of the form:

k₁(T) = Aexp(−E/RT) (iii)

to give k₁ (201–348 K) = 4.38 × 10⁻¹² exp {(366 ± 40)/T)} cm³ s⁻¹, where the quoted errors in E/R encompass all data points and their estimated 5% total error (see below). The rate coefficient shows a distinct negative temperature dependence (i.e. the rate coefficient increases as the temperature decreases), which has previously been observed for this reaction, at least in the same temperature range (see section 3.1). As discussed in an earlier study that focussed on the product channels in the title reaction, the negative temperature dependence most likely reflects the formation and stabilisation of an association complex, that dissociates to form CH₃CO and H₂O.¹⁵

Fig. 4 Arrhenius plot of all data from the present study. The open circles represent data obtained using the 248 nm photolysis of H₂O₂ as OH source, the data represented by solid triangles were obtained using the 193 nm photolysis of N₂O/CH₄ as OH source.

Table 1 Summary of present measurements of k₁

Temperature/K	Bath gas,^a OH source	Range of CH3CHO/10^14 cm^−3	10^11k1^bcm^3 s^−1
a All experiments carried out at 100 Torr total pressure, except for those where the pressure is indicated. b Error limits are 2σ precision only. The total estimated error is 5% at each temperature.
202	He, N2O/CH4	1.4–14.4	2.58 ± 0.03
202	He, N2O/CH4	1.4–15.4	2.62 ± 0.04
202	He, N2O/CH4	1.6–14.0	2.87 ± 0.06
211	He, N2O/CH4	1.0–9.9	2.51 ± 0.04
212	He, N2O/CH4	1.4–14.2	2.47 ± 0.04
224	He, N2O/CH4	1.1–13.5	2.27 ± 0.02
234	He, N2O/CH4	1.1–13.0	2.14 ± 0.02
244	He, N2O/CH4	1.0–12.8	1.90 ± 0.02
253	He, N2O/CH4	0.9–11.7	1.82 ± 0.02
263	He, N2O/CH4	0.9–11.5	1.73 ± 0.03
273	He, N2O/CH4	0.8–10.7	1.73 ± 0.02
283	He, N2O/CH4	0.8–10.7	1.59 ± 0.03
297	He, N2O/CH4	0.8–10.3	1.56 ± 0.02
295 (50 Torr)	He, N2O/CH4	1.2–16.1	1.52 ± 0.02
294	He, N2O/CH4	0.7–09.9	1.58 ± 0.02
273	Ar, H2O2	0.9–11.2	1.65 ± 0.03
273	Ar, H2O2	0.9–11.8	1.68 ± 0.02
278	Ar, H2O2	0.9–11.6	1.66 ± 0.02
284	Ar, H2O2	0.8–10.8	1.60 ± 0.03
284	Ar, H2O2	0.9–09.3	1.61 ± 0.02
289	Ar, H2O2	0.9–11.3	1.53 ± 0.03
292	Ar, H2O2	0.8–10.5	1.51 ± 0.02
296	Ar, H2O2	1.0–13.5	1.51 ± 0.01
296 (50 Torr)	Ar, H2O2	0.8–10.2	1.53 ± 0.02
304	Ar, H2O2	0.8–10.6	1.44 ± 0.02
314	Ar, H2O2	0.8–10.2	1.41 ± 0.02
319	Ar, H2O2	0.7–09.5	1.38 ± 0.03
324	Ar, H2O2	0.8–07.9	1.36 ± 0.03
333	Ar, H2O2	0.7–09.5	1.30 ± 0.02
342	Ar, H2O2	0.7–07.3	1.25 ± 0.02
348	Ar, H2O2	0.7–09.0	1.26 ± 0.02

---

The total error in k₁ at any particular temperature covered by the present work is ≈5% and is derived mainly from estimated errors in the measurement of the CH₃CHO concentration. The 5% error is represented by the vertical error bars at each temperature in Fig. 4 and the 95% confidence limits are represented by the outlying upper and lower bounds to the central fit line. Note that where there is overlap (273–300 K), excellent agreement exists between data obtained using the two different OH sources.

Discussion

In the present work, accurate kinetic data on the reaction of OH with CH₃CHO have been obtained at temperatures down to almost 200 K for the first time. We now compare our results with previous measurements of the rate coefficient, that have been carried out over a period spanning about 30 years, and by a number of different methods, which are listed in Table 2. Excluding the present study, there are a total of 12 determinations of the rate coefficient at room temperature, including both relative-rate, and absolute determinations. The data are in good agreement, with most results consistent with k₁ (298 ± 2 K) = (1.5 ± 0.1) × 10⁻¹¹ cm³ s⁻¹ within combined error limits. As in the present study, the independence of k₁ on pressure and bath gas has been documented several times.

Table 2 Summary of literature determinations of k₁^a

Ref.	Method	OH source	T/K	P/Torr	10^11k1/cm^3 s^−1 (room temp.)	10^12A/cm^3 s^−1	(E/R)/K
a Overall rate coefficient given by k1(T) = A exp (−E/RT). PLP = pulsed laser photolysis; PLIF = pulsed laser induced fluorescence; DF = discharge flow; MS = mass spectrometry; RR = relative-rate; FTIR = Fourier transform infrared spectroscopy; FP = flash photolysis; RF = resonance fluorescence; GC = gas chromatography. b Estimated total uncertainty of 25%. c Derived using presently recommended^28 value of k(OH + C2H4) and associated uncertainties. d Includes estimation of overall uncertainty. e Errors are 2σ, precision only. f Total errors of 8–27% given for all data at all temperatures. g Derived using presently recommended^28 value of k(OH + 1-butene) and associated uncertainties.
This work	PLP-PLIF	H2O2 (λ = 248 nm)	201–348	50, 100 (He, Ar)	1.53 ± 0.08	4.38	−366 ± 40
		N2O/CH4 (λ = 193 nm)
Morris et al.^29	DF-MS	H + NO2	300	1 (He)	1.55 ± 0.39^b
Niki et al.^30	RR-FTIR (relative to C2H4)	HONO	298	700 (air)	1.62 ± 0.18^c
Atkinson and Pitts^31	FP-RF	H2O (λ > 105 nm)	299–426	50 (Ar)	1.60 ± 0.16^d	6.87	−257 ± 151
Kerr and Sheppard^32	RR-GC (relative to C2H4)	HONO	298	760	1.28 ± 0.43^c
Michael et al.^26	DF-RF	H + NO2	248–528	1.1–3.4 (He)	1.47 ± 0.28^e	5.52	−307 ± 52
Semmes et al.^24	FP-RF	H2O (λ > 165 nm)	250–425	100 (Ar)	1.22 ± 0.27^d	7.1	−165 ± 91
Dóbé et al.^33	DF-RF/PLIF	H + NO2 F + H2O	297–517		1.69 ± 0.33^d	8.6	−200 ± 60
Balestra et al.^34	PLP-RF	HNO3 (λ = 248 nm)	298	27	1.7 ± 0.3^d
Scollard et al.^35	RR-GC (relative to C2H4)	CH3ONO2 (λ = 350–450 nm)	298	760	1.62 ± 0.1^c
Tyndall et al.^23	DF-PLIF	H + NO2 F + H2O	298	2–3 (He)	1.45 ± 0.25^d
Taylor et al.^25	PLP-PLIF	N2O/H2O (λ = 193 nm)	295–550	100–740 (He)	1.23^f	4.31	−309 ± 19
Taylor et al.^25	PLP-PLIF	N2O/H2O (λ = 193 nm)	600–900	100–740 (He)	1.23^f	1.89	597 ± 108
D'Anna et al.^36	RR-FTIR (relative to 1-butene)	RONO	298	760 (air)	1.44 ± 0.08^g

---

A comparison between the present result and literature data is presented in graphical form in Fig. 5. The rate coefficients of Semmes et al.,²⁴ and Taylor et al.²⁵ are slightly lower than most of the others, which may be due to problems associated with CH₃CHO decomposition in storage bulbs, as these authors have already discussed. Still, these results agree, within combined error limits with most of the other studies at room temperature, both relative and absolute. The average value of the existing room temperature data is 1.49 × 10⁻¹¹ cm³ s⁻¹, which is in excellent agreement with the present determination of (1.53 ± 0.08) × 10⁻¹¹ cm³ s⁻¹ where the error limits define the total estimated error as 5%.

Fig. 5 Compilation of literature data. The solid line is the least squares fit to the present data set (k₁ = 4.38 × 10⁻¹² exp{366/T} cm³ s⁻¹). The data of Taylor 96 were not tabulated in their publication and were therefore taken from their Arrhenius plot. The references refer to Morris et al.,²⁹ Niki et al.,³⁰ Atkinson and Pitts,³¹ Kerr and Sheppard,³² Michael et al.,²⁶ Semmes et al.,²⁴ Dóbé et al.,³³ Balestra et al.,³⁴ Tyndall et al.,²³ Scollard et al.,³⁵ Taylor et al.,²⁵ and D'Anna et al.³⁶

Apart from the present study, there are five absolute measurements of the temperature dependence of k₁. Much of the previous work has focussed on higher temperatures, and no data had previously been obtained at temperatures below 244 K. An examination of the data in Fig. 5 shows considerable scatter, though all previous studies of the rate coefficient at temperatures less than ≈400 K determined a negative temperature dependence that varied between E/R = −165 and −307. A common feature of a number of the previous data sets is also some curvature in the Arrhenius plot at the higher range of temperatures, though the onset of the curvature is quite variable and partially disguised by scatter in the data. The data set of Taylor et al.²⁵ most clearly reveals strong curvature with an onset at ≈550–600 K, with the rate coefficient displaying a reversed temperature dependence (i.e. positive) at temperatures between 600 and 900 K. The present data set is represented by the straight line between 200 and 350 K which was calculated according to eqn. (iii) using A = 4.38 × 10⁻¹² cm³ s⁻¹ and E/R = −366. The slope of this plot agrees well with those of Semmes et al.²⁴ (below 360 K), Michael et al.²⁶ and Taylor et al.²⁵ (below 500 K), though the absolute values of the rate coefficient are somewhat higher (see discussion above concerning the room temperature data). The present data, which extends over the entire temperature range expected to be encountered by acetaldehyde in the atmosphere, accurately defines both the absolute values of the rate coefficient and their temperature dependence. This significantly reduces the uncertainty associated with calculation of the atmospheric lifetime of acetaldehyde with respect to its OH sink reaction, and formation rates of PAN. For very low temperatures in the upper troposphere, e.g. 210 K, the present parametrisation of the rate coefficient results in a value of 2.5 × 10⁻¹¹ cm³ s⁻¹, which is entirely consistent with recent IUPAC recommendations¹⁶ but which is significantly higher than the value of 2.0 × 10⁻¹¹ cm³ s⁻¹ that is presently recommended by the NASA data evaluation panel.²⁷

Conclusions

The rate constant for the reaction of OH with CH₃CHO was determined using pulsed laser photolysis-pulsed laser induced fluorescence. The kinetic data obtained shows a negative temperature dependence over the range covered, and the overall rate coefficient is given by k₁(201–348 K) = 4.38 × 10⁻¹² exp{(366 ± 40)/T} cm³ s⁻¹. Use of this expression should reduce uncertainties in modelling of the lifetime and role of CH₃CHO in the atmosphere.

Acknowledgements

The authors thank the European Union for financial support within the environment programme (EVK2-CT2001-00099, UTOPIHAN-ACT).

References

1. H. Singh, Y. Chen, A. Tabazadeh, Y. Fukui, I. Bey, R. Yantosca, D. Jacob, F. Arnold, K. Wohlfrom, E. Atlas, F. Flocke, D. Blake, N. Blake, B. Heikes, J. Snow, R. Talbot, G. Gregory, G. Sachse, S. Vay and Y. Kondo, J. Geophys. Res., 2000, 105, 3795–3805.
2. H. Singh, Y. Chen, A. Staudt, D. Jacob, D. Blake, B. Heikes and J. Snow, Nature, 2001, 410, 1078–1081.
3. H. B. Singh, M. Kanakidou, P. J. Crutzen and D. J. Jacob, Nature, 1995, 378, 50–54.
4. I. Folkins, P. O. Wennberg, T. F. Hanisco, J. G. Anderson and R. J. Salawitch, Geophys. Res. Lett., 1997, 24, 3185–3188.
5. I. Folkins, R. Chatfield, H. Singh, Y. Chen and B. Heikes, Geophys. Res. Lett., 1998, 25, 1305–1308.
6. L. Jaeglé, D. J. Jacob, W. H. Brune, D. Tan, I. C. Faloona, A. J. Weinheimer, B. A. Ridley, T. L. Campos and G. W. Sachse, Geophys. Res. Lett., 1998, 25, 1709–1712.
7. L. Jaeglé, D. J. Jacob, P. O. Wennberg, C. M. Spivakovsky, T. F. Hanisco, E. J. Lanzendorf, E. J. Hintsa, D. W. Fahey, E. R. Keim, M. H. Proffitt, E. L. Atlas, F. Flocke, S. Schauffler, C. T. McElroy, C. Midwinter, L. Pfister and J. C. Wilson, Geophys. Res. Lett., 1997, 24, 3181–3184.
8. G. P. Brasseur, J.-F. Müller and C. Granier, J. Geophys. Res., 1996, 101, 1423–1428.
9. J. F. Müller and G. Brasseur, J. Geophys. Res., 1999, 104, 1705–1715.
10. S. K. Meilinger, B. Kärcher, R. v. Kuhlmann and T. Peter, Geophys. Res. Lett., 2001, 28, 515–518.
11. M. Wollenhaupt, S. A. Carl, A. Horowitz and J. N. Crowley, J. Phys. Chem. A, 2000, 104, 2695–2705.
12. T. Gierczak, J. B. Burkholder, S. Bauerle and A. R. Ravishankara, Chem. Phys., 1998, 231, 229–244.
13. P. Carlier, H. Hannachi and G. Mouvier, Atmos. Environ., 1986, 20, 2079–2099.
14. F. Arnold, V. Bürger, B. Droste-Fanke, F. Grimm, A. Krieger, J. Schneider and T. Stilp, Geophys. Res. Lett., 1997, 24, 3017–3020.
15. M. Cameron, V. Sivakumaran, T. J. Dillon and J. N. Crowley, Phys. Chem. Chem. Phys., 2002, 4, 3628–3638.
16. R. Atkinson, D. L. Baulch, R. A. Cox, R. F. J. Hampson, J. A. Kerr, M. J. Rossi and J. Troe, J. Phys. Chem. Ref. Data, 1999, 28, 191–393.
17. S. A. McKeen, T. Gierzak, J. B. Burkholder, P. O. Wennberg, T. F. Hanisco, E. R. Keim, R.-S. Gao, S. C. Liu, A. R. Ravishankara and D. W. Fahey, Geophys. Res. Lett., 1997, 24, 3177–3180.
18. P. M. Aker, J. J. A. O'Brien and J. J. Sloan, J. Chem. Phys., 1986, 84, 745–749.
19. K. J. Rensberger, J. B. Jeffries and D. R. Crosley, J. Chem. Phys., 1989, 90, 2174–2181.
20. W. Schneider and G. K. Moortgat, unpublished data.
21. R. D. Martinez, A. A. Buitrago, N. W. Howell, C. H. Hearn and J. A. Joens, Atmos. Environ., 1992, 26, 785–792.
22. H. G. Libuda, Spektroskopische und kinetische Untersuchungen an halogenierten Carbonylverbindungen von atmosphärischem Interesse, PhD Thesis, University of Wuppertal, 1992.
23. G. S. Tyndall, T. A. Staffelbach, J. J. Orlando and J. G. Calvert, Int. J. Chem. Kinet., 1995, 27, 1009–1020.
24. D. H. Semmes, A. R. Ravishankara, C. A. Gump-Perkins and P. H. Wine, Int. J. Chem. Kinet., 1985, 17, 303–313.
25. P. H. Taylor, M. S. Rahman, B. Dellinger and P. Marshall, Twenty-Sixth Symposium (International) on Combustion, The Combustion Institute, 1996.
26. J. V. Michael, D. G. Keil and R. B. Klemm, J. Chem. Phys., 1985, 83, 1630–1636.
27. W. B. DeMore, S. P. Sander, D. M. Golden, R. F. Hampson, M. J. Kurylo, C. J. Howard, A. R. Ravishankara, C. E. Kolb and M. J. Molina, Chemical Kinetics and Photochemical Data for Use in Stratospheric Modelling, No 12, Jet Propulsion Laboratory, 1997.
28. R. Atkinson, J. Phys. Chem. Ref. Data, 1997, 26, 215–290.
29. E. D. Morris, Jr., D. H. Stedman and H. Niki, J. Am. Chem. Soc., 1971, 93, 3570–3572.
30. H. Niki, P. D. Maker, C. M. Savage and L. P. Breitenbach, J. Phys. Chem., 1978, 82, 132–134.
31. R. Atkinson and J. N. Pitts, J. Chem. Phys., 1978, 68, 3581–3584.
32. J. A. Kerr and D. W. Sheppard, Environ. Sci. Technol., 1981, 15, 960–963.
33. S. Dóbé, L. A. Khachatryan and T. Bérces, Ber. Bunsen-Ges. Phys. Chem., 1989, 93, 847–852.
34. C. Balestra-Garcia, G. Le Bras and H. Mac Leod, J. Phys. Chem., 1992, 96, 3312–3316.
35. D. J. Scollard, J. J. Treacy, H. W. Sidebottom, C. Balestra-Garcia, G. Laverdet, G. Lebras, H. Macleod and S. Teton, J. Phys. Chem., 1993, 97, 4683–4688.
36. B. D'Anna, W. Andresen, Z. Gefen and C. J. Nielsen, Phys. Chem. Chem. Phys., 2001, 3, 3057–3063.

---

Footnote

† For Part I see ref. 15.

---