Dimethyl Ether Oxidation: Kinetics and Mechanism of the CH3OCH2 + O2 Reaction at 296 K and 0.38-940 Torr Total Pressure

Article abstract of DOI:10.1021/jp961821m

The title reaction was studied at 296 K and 0.38-940 Torr total pressure using a FTIR smog chamber technique.The overall rate constant for reaction of CH3OCH2 radicals with O2 may be written, k₁ = k_RO2 + k_prod, where k_RO2 is the rate constant for peroxy radical production and k_prod is the rate constant for the production of other species from reaction 1. k₁ was measured relative to the pressure independent reaction of CH3OCHJ2 radicals with Cl2 (k₄).There was no discernible effect of pressure on k₁ in the range 200 - 940 Torr.Between 200 and 2 Torr total pressure k₁ decreased by approximately a factor of 2.For pressures below 2 Torr k₁ was again independent of pressure.The reaction proceeds via the formation of an activated complex, CH3OCH2O2(excit.), that is either collisionally stabilized to form the peroxy radical, CH3OCH2O2, or undergoes intramolecular H-atom abstraction followed by decomposition to give two formaldehyde molecules and an OH radical: CH3OCH2 + O2 ->/<- CH3OCH2O2(excit.), CH3OCH2O2(excit.) + M -> CH3OCH2O2 + M, CH3OCH2O2(excit.) -> CH2OCH2O2H(excit.) -> 2HCHO + OH.The products from this reaction were studied as a function of total pressure.The molar yield of formaldehyde increased from <2percent at 700 Torr total pressure to ca. 200percent at 0.38 Torr total pressure, while the combined yield of methyl formate and methoxy methylhydroperoxide decreased from ca. 100percent to 4percent over the same pressure range.Fitting the product yields and relative rate data using a modified Lindemann expression gave the following rate constants: k_{RO2,0^{/k4 = (1.97 +/- 0.28)E-19 cm3 molecule-1, k}RO2,infinite^/k4 ^{= 0.108 +/- 0.004, and k}prod.0^/k4 ^{= (6.3 +/- 0.5)E-2 where k}RO2.0} and k_RO2.infinite are the overall termolecular and bimolecular rate constants for formation of the CH3OCH2O2 radical and k_prod.0 represents the bimolecular rate constant for the reaction of CH3OCH2 radicals with O2 to yield formaldehyde in the limit of low pressure.These data and absolute rate data from the literature were used to derive a rate constant for the reaction of CH3OCH2 radicals with Cl2 of (1.00 +/- 0.16)E-10 cm³ molecule^-1 s^-1.The results are discussed in the context of the use of dimethyl ether as an alternative diesel fuel.

Full text of DOI:10.1021/jp961821m

17218
J. Phys. Chem. 1996, 100, 17218-17225
Dimethyl Ether Oxidation: Kinetics and Mechanism of the CH3OCH2 + O2 Reaction at
296 K and 0.38-940 Torr Total Pressure
J. Sehested* and T. Møgelberg
Section for Chemical ReactiVity, EnVironmental Science and Technology Department,
Risø National Laboratory, DK-4000 Roskilde, Denmark
T. J. Wallington* and E. W. Kaiser
Ford Motor Company, Mail Drop, SRL-3083, P.O. Box 2053, Dearborn, Michigan 48121-2053
O. J. Nielsen*
Ford Motor Company, Ford Forschungszentrum Aachen, Dennewartsstrasse 25, D-52068 Aachen, Germany
ReceiVed: June 19, 1996; In Final Form: August 19, 1996X
The title reaction was studied at 296 K and 0.38-940 Torr total pressure using a FTIR smog chamber technique.
The overall rate constant for reaction of CH3OCH2 radicals with O2 may be written, k1 ) kRO2 + kprod, where
kRO2 is the rate constant for peroxy radical production and kprod is the rate constant for the production of other
species from reaction 1. k1 was measured relative to the pressure independent reaction of CH3OCH2 radicals
with Cl2 (k4). There was no discernible effect of pressure on k1 in the range 200-940 Torr. Between 200
and 2 Torr total pressure k1 decreased by approximately a factor of 2. For pressures below 2 Torr k1 was
#
again independent of pressure. The reaction proceeds via the formation of an activated complex, CH3OCH2O2 ,
that is either collisionally stabilized to form the peroxy radical, CH3OCH2O2, or undergoes intramolecular
H-atom abstraction followed by decomposition to give two formaldehyde molecules and an OH radical: CH3-
#
OCH2 + O2 h CH3OCH2O2 , CH3OCH2O2# + M f CH3OCH2O2 + M, CH3OCH2O2# f CH2OCH2O2H# f
2HCHO + OH. The products from this reaction were studied as a function of total pressure. The molar
yield of formaldehyde increased from <2% at 700 Torr total pressure to ∼200% at 0.38 Torr total pressure,
while the combined yield of methyl formate and methoxy methylhydroperoxide decreased from ∼100% to
4% over the same pressure range. Fitting the product yields and relative rate data using a modified Lindemann
expression gave the following rate constants: kRO2,0/k4 ) (1.97 ( 0.28) × 10-19 cm3 molecule-1, kRO2,∞/k4 )
0.108 ( 0.004, and kprod,0/k4 ) (6.3 ( 0.5) × 10-2 where kRO2,0 and kRO2,∞ are the overall termolecular and
bimolecular rate constants for formation of the CH3OCH2O2 radical and kprod,0 represents the bimolecular rate
constant for the reaction of CH3OCH2 radicals with O2 to yield formaldehyde in the limit of low pressure.
These data and absolute rate data from the literature were used to derive a rate constant for the reaction of
CH3OCH2 radicals with Cl2 of (1.00 ( 0.16) × 10-10 cm3 molecule-1 s-1. The results are discussed in the
context of the use of dimethyl ether as an alternative diesel fuel.
1. Introduction
octane (i.e., low cetane) numbers and are used to boost the
octane number of gasoline. It is surprising that DME, which
is structurally similar to MTBE, ETBE, and TAME, has a high
cetane number and hence is a good diesel fuel, while the other
ethers have low cetane numbers and would make poor diesel
fuels. A possible explanation for this difference is discussed
below.
At this point it is germane to note that unlike conventional
diesel fuel, DME is a gas at ambient temperature. Hence, any
widespread use of DME would require significant investment
in new infracture associated with fuel distribution and delivery.
The atmospheric chemistry of DME has been studied by Japar
et al.,3 Jenkin et al.,4 Wallington et al.,5 and Langer et al.6
Surprisingly high formaldehyde yields following Cl atom
initiated oxidation of dimethyl ether in N2/O2 diluent at total
pressures below 300 Torr were reported by Jenkin et al.4 It
was suggested that formaldehyde was formed in the reaction
of CH3OCH2 radicals with O2,
Dimethyl ether (DME) has been proposed as an alternative
diesel fuel by Topsøe and AMOCO.1 Dimethyl ether has the
following desirable properties that make it an attractive diesel
fuel. First, it has a high cetane (i.e., low octane) number.
Second, when used in diesel engines, it reduces combustion
noise and emissions of particles. Third, the trade-off between
NOx emissions and soot using conventional diesel fuel is
eliminated in DME-fueled engines because the combustion is
soot-free. Hence, a DME-fueled engine can be tuned to give
reduced NOx emissions.1 Fourth, DME can be produced from
hydrocarbons via a relatively cheap one-step synthesis. Engine
tests have shown that DME-fueled diesel engines have emission
levels that surpass the California ultralow emissions vehicle
(ULEV) regulation for medium duty vehicles.2 Interestingly,
ethers like MTBE (methyl tert-butyl ether), ETBE (ethyl tert-
butyl ether), and TAME (tert-amyl methyl ether) have high
* To whom correspondence may be addressed.
CH3OCH2 + O2 (+ M) f products
(1)
X Abstract published in AdVance ACS Abstracts, October 1, 1996.
S0022-3654(96)01821-7 CCC: $12.00 © 1996 American Chemical Society

Dimethyl Ether Oxidation
J. Phys. Chem., Vol. 100, No. 43, 1996 17219
via the following mechanism:4
formaldehyde, 1700-1800; formic acid, 1050-1150; CO,
2050-2250. All compounds were obtained from commercial
vendors and used without further purification.
#
CH3OCH2 + O2 h CH3OCH2O2
CH3OCH2O2# f CH2OCH2O2H# f 2HCHO + OH (1b)
CH3OCH2O2# + M f CH3OCH2O2 + M
(1a,-1a)
Four sets of experiments were performed. First, the kinetics
of the CH3OCH2 + O2 reaction were studied relative to the
CH3OCH2 + Cl2 reaction as a function of total pressure.
Second, the product distribution following the reaction of CH3-
OCH2 radicals with O2 was studied over the pressure range
0.38-700 Torr. Third, the products of DME oxidation at 2, 5,
and 700 Torr total pressures in the presence of a large excess
of 13CH3OH, and hence HO2 radicals, were studied. Finally,
the apparent photolysis rate of Cl2 (calculated by the loss of
DME) was measured at 2 and 700 Torr total pressure.
(1c)
where formaldehyde is formed by intramolecular hydrogen
transfer in the energetically excited CH3OCH2O2 radical
followed by decomposition, (reaction 1b). At sufficiently low
pressures CH3OCH2O2# radicals are not stabilized and reaction
1 proceeds essentially exclusively via reaction 1b. At higher
#
#
pressures the highly energetic CH3OCH2O2 complex is stabi-
3. Results
lized by collisional quenching to give thermalized CH3OCH2O2
radicals. CH3OCH2O2 radicals react to give CH3OCHO and
CH3OCH2OOH under atmospheric conditions.4
3.1. Relative Rate Study of CH3OCH2 + O2 vs CH3OCH2
+ Cl2. To study the rate constant ratio, k1/k4, mixtures of 10-
31 mTorr of DME, 28-234 mTorr Cl2, and 0.34-2.5 Torr O2
with N2 added to a total pressure of 0.38-940 Torr were subject
to UV irradiation. The following reactions are important in the
mixture:
Alkylperoxy radicals are thermally unstable at >1000 K and
consequently are not important in high-temperature combustion
chemistry. However, at lower temperature (500-950 K)
alkylperoxy radicals have sufficient stability to assume an
important role in autoignition phenomena, which cause un-
wanted audible mechanical vibrations in gasoline-fueled engines
known as “knock”.7,8 Isomerization reactions similar to reaction
1b are important sources of reactive free radicals that promote
autoignition.9 The importance of reaction 1b in the high-
temperature oxidation of DME has been considered by Dagaut
et al.10 using a detailed kinetic modeling approach at temper-
atures of 850-1300 K and pressures of 1-10 bar. It was found
that reaction 1b was an important loss of CH3OCH2 radicals in
their model at high pressures and low temperatures.10 In terms
of the use of DME as a diesel fuel, reaction 1b is desirable,
since it propagates the reaction chain producing reactive OH
radicals and formaldehyde. Combustion of formaldehyde gives
little or no particulate matter.11
Cl2 + hν f 2Cl
(2)
(3)
(1)
(4)
Cl + CH3OCH3 f CH3OCH2 + HCl
CH3OCH2 + O2 (+ M) f products
CH3OCH2 + Cl2 f CH3OCH2Cl + Cl
Reaction 4 is expected to be independent of total pressure and
provides a convenient reference for determining the pressure
dependence of k1. The fraction of the CH3OCH2 radicals that
react with Cl2, YDMECl, was determined by measuring the
formation of CH3OCH2Cl (∆DMECl) relative to the loss of
DME (∆DME). Hence, YDMECl ) ∆DMECl/∆DME ) k4[Cl2]/
(k1[O2] + k4[Cl2]). The measured yield of CH3OCH2Cl needs
a small correction to account for loss of CH3OCH2Cl via
reaction 5:
The objective of this work is to investigate the kinetics and
mechanism of reaction 1. The results are discussed in the
context of DME as an alternative diesel fuel.
Cl + CH3OCH2Cl f products
(5)
2. Experimental Section
The rate constants for reactions 5 and 3 are k5 ) 2.9 × 10-11
The apparatus and experimental technique used in this work
have been described in detail previously12 and are only briefly
discussed here. Experiments were performed in a 140 L Pyrex
reactor surrounded by 22 fluorescent blacklamps (GE F15T8-
BL) that emit in the region 300-450 nm. Methoxy methyl
radicals were generated by the photolysis of molecular chlorine
in the presence of dimethyl ether with and without NO in 0.34-
700 Torr of O2. N2 was added to reach the desired total pressure
of 0.38-940 Torr at 296 ( 2 K.
and k3 ) 1.9 × 10-10 cm3 molecule-1 s-1 4
. Conversion of DME
was in the range 0-59%, and the corrections applied to YDMECl
were <7%.
k1/k4 can then be calculated from eq I:
k1/k4 ) (1 - YDMECl)[Cl2]/(YDMECl[O2])
(I)
The results are shown in Figure 1. As expected, k1/k4 ap-
proaches a high-pressure limit at pressures above 100 Torr and
falls off below 100 Torr total pressure. It is interesting to note
that k1/k4 does not approach zero at low pressures. This behavior
can be explained if reaction 1 proceeds via a complex that
decomposes via two pathways, either to reform the reactants
or to give other products. We show later that these “other
products” are formaldehyde and OH radicals. The reactions
involved in the formation and decomposition of the complex,
Cl2 + hν f 2Cl
(2)
(3)
Cl + CH3OCH3 f CH3OCH2 + HCl
Loss of dimethyl ether and formation of products were
monitored by Fourier transform infrared spectroscopy using an
infrared path length of 27 m and a resolution of 0.25 cm-1
.
Infrared spectra were derived from 32 coadded interferograms.
Reference spectra were obtained by expanding known volumes
of reference materials into the long path length cell at appropriate
total pressures. Products were identified and quantified by
fitting reference spectra of the pure compounds obtained at the
appropriate total pressures to the observed product spectra using
integrated absorption features over the following wavelength
ranges (in cm-1): dimethyl ether, 900-1200; methyl formate,
1000-1100; methoxy methylhydroperoxide, 775-850 and
1700-1750; chlorodimethyl ether, 650-850 and 1200-1300;
#
CH3OCH2O2 , are
#
CH3OCH2 + O2 h CH3OCH2O2
(1a,-1a)
CH3OCH2O2# f CH2OCH2O2H# f 2HCHO + OH (1b)
CH3OCH2O2# + M f CH3OCH2O2 + M
(1c)
At high pressures the rate constant for the reaction of CH3-

17220 J. Phys. Chem., Vol. 100, No. 43, 1996
Sehested et al.
limit, kprod,0 ) k1ak1b/(k-1a + k1b), it follows that
kprod
)
{k1ak1b/(k-1a + k1b)}/{1 + k1ak1c[M]/(k1a(k-1a + k1b))}
) kprod,0/(1 + kRO2,0[M]/kRO2,∞
Equation II,
)
k1/k4 ) kRO2/k4 + kprod/k4
(II)
was fitted to the data in Figure 1 with kprod,0/k4, kRO2,0/k4, and
kRO2,∞/k4 as parameters. The fit is shown as the smooth line
through the data in Figure 1. As seen from Figure 1, the
experimental data are well fitted by eq II. The best fit was
obtained using kprod,0/k4 ) 0.059 ( 0.005, kRO2,0/k4 ) 0.0069
( 0.0032 Torr-1 ) (2.1 ( 1.0) × 10-19 cm3 molecule-1, and
kRO2,∞/k4 ) 0.111 ( 0.004.
Figure 1. k1/k4 vs total pressure. The smooth line is a fit to the
data using a modified Lindemann mechanism (eq II). See text for
details.
Two absolute measurements of the rate constant for reaction
1 have been reported. Masaki et al.13 measured k1 at 0.8-8
Torr total pressure and 298 K using a laser flash photolysis-
mass spectrometry absolute technique. Masaki et al.13 reported
k1 ) (6.5 ( 0.7) × 10-12 cm3 molecule-1 s-1 independent of
total pressure over the pressure range studied. We also find
that k1 is independent of total pressure below 6 Torr. Sehested
et al.14 measured k1 at 296 K at 1 bar total pressure using pulse
radiolysis transient UV absorption spectroscopy and report k1
) (1.09 ( 0.05) × 10-11 cm3 molecule-1 s-1. The ratio of the
absolute rate constants determined by Masaki et al.13 at low
pressure and by Sehested et al.14 at high pressure is 0.59 (
0.07. This is in good agreement with the ratio of the average
of the relative rate data below 8 Torr total pressure, 0.064 (
0.05, and the high-pressure limit determined in this study, 0.111
( 0.004, k1(0-8 Torr)/kRO2,∞ ) 0.58 ( 0.05. Using the values
for k1 of Masaki et al.13 and Sehested et al.,14 k1(0-8 Torr) )
0.064 ( 0.005, and kRO2,∞/k4 ) 0.111 ( 0.004, we calculate k4
) (1.02 ( 0.14) × 10-10 and k4 ) (0.98 ( 0.06) × 10-10 cm3
molecule-1 s-1, respectively. We choose k4 ) (1.00 ( 0.16)
× 10-10 cm3 molecule-1 s-1, which is the average of two k4
values with an uncertainty that encompasses the extremes of
the error limits of the individual determinations.
#
OCH2 radicals with O2 equals k1a, since all CH3OCH2O2
radicals are quenched by the third body “M”, mainly O2 and
N2 (reaction 1c), and the peroxy radical, CH3OCH2O2, is formed.
#
As the total pressure is decreased, not all of the CH3OCH2O2
radicals are quenched and the rate constant ratio k1/k4 decreases.
#
At the low-pressure limit none of the CH3OCH2O2 radicals
are quenched; however, because of intramolecular hydrogen
#
transfer in the excited CH3OCH2O2 radical, the overall rate
constant k1 does not drop to zero at zero total pressure.
#
Hydrogen transfer in CH3OCH2O2 is expected to proceed via
a six-membered ring transition state.
Reactions 1a-1c can be viewed as a modified Lindemann
mechanism. The overall reaction of CH3OCH2 radicals with
O2 proceeds via two pathways to give either the peroxy radical
CH3OCH2O2 (RO2) or other products (prod). Hence,
-d[CH3OCH2]/dt ) d[RO2]/dt + d[prod]/dt )
k1[CH3OCH2][O2]
It is interesting to compare our result for k4 with the large
kinetic data base for reactions of alkyl radicals with Cl2. Rate
constants at 298 K for such reactions are typically in the range
#
d[RO2]/dt ) kRO2[CH3OCH2][O2] ) k1c[M][RO2 ]
(1-50) × 10-12 cm3 molecule-1 s-1 15
.
For reasons that are
#
d[prod]/dt ) kprod[CH3OCH2][O2] ) k1b[RO2 ]
unclear, the CH3OCH2 radical displays an unusually high
reactivity toward Cl2.
#
Applying a steady state analysis for RO2 gives
3.2. Product Study of CH3OCH2 + O2. Six experiments
were carried out to study the products of dimethyl ether
oxidation as a function of total pressure. The initial conditions
and the results are given in Table 1. In each experiment, a
CH3OCH3/Cl2/O2/N2 mixture was subject to two to six succes-
sive periods of irradiation, each of 4-30 s duration. After each
irradiation the reaction mixture was analyzed using FTIR
spectroscopy.
Parts A and B of Figure 2 show IR spectra acquired before
and after irradiation of a mixture of 23 mTorr of dimethyl ether,
23 mTorr of Cl2, and 9.9 Torr of O2 in 690 Torr of N2
(experiment 6). Parts C and D of Figure 2 show IR spectra
acquired before and after irradiation of a mixture of 23.1 mTorr
of dimethyl ether and 23 mTorr of Cl2 in 10.4 Torr of O2
(experiment 5). The loss of dimethyl ether was 2.89 mTorr
from spectrum A to B of Figure 2 and 2.77 mTorr from C to
D. Reference spectra of methyl formate and formaldehyde are
shown in the bottom panels of Figure 2. It is easily seen from
the product spectra in parts B and D of Figure 2 that different
products are observed at different total pressures. The product
kRO2 ) k1ak1c[M]/(k-1a + k1b + k1c[M])
kprod ) k1ak1b/(k-1a + k1b + k1c[M])
Defining kRO2,0 ) k1ak1c/(k-1a + k1b) and kRO2,∞ ) k1a, we
can express the effective second-order rate constant for the
formation of RO2 radicals at any given pressure, kRO2, in terms
of the limiting low- and high-pressure rate constants, kRO2,0 and
kRO2,∞
:
kRO2
)
{k1ak1c[M]/(k-1a + k1b)}/{1 + k1ak1c[M]/(k1a(k-1a + k1b))}
) kRO2,0[M]/(1 + kRO2,0[M]/kRO2,∞
)
Similarly, by definition of the rate constant for formation of
“products” (formaldehyde and OH radicals) in the low-pressure

Dimethyl Ether Oxidation
J. Phys. Chem., Vol. 100, No. 43, 1996 17221
TABLE 1: Product Yieldsa Following the Irradiation of CH3OCH3/Cl2/O2/N2 at 0.38-700 Torr Total Pressure
Y(CH3OCHO)
Y(CH3OCH2OOH)
Y(HCHO)
Y(HCOOH)
Y(CO)
%
[DME]b
[O2]c
[N2]c
[Cl2]o
Ptot
%
%
%
%
b
e
set
1
2
3
4
10.3
10.4
10.4
21.8
23.1
23.0
29.6
33.8
0.36
1.06
1.64
5.7
10.4
9.9
0
0
0
0
0
690
699
580
10.3
10.4
10.4
15.4
23.0
23.0
15.8
213.0
0.38
1.08
1.66
5.7
10.4
700
4 ( 2
6 ( 2
181 ( 15
185 ( 20
173 ( 32
120 ( 8
89 ( 10
22 ( 4
14 ( 9
13 ( 5
20 ( 8
6 ( 3
11 ( 2
32 ( 7
42 ( 4
67 ( 2
88 ( 6
96 ( 20
7 ( 2
14 ( 3
30 ( 5
5
2 ( 2
6
7d
8d
1.2
120
700
700
<1.9
<2
a Molar yields relative to CH3OCH3 loss corrected for formation of CH3OCH2Cl. b Concentrations in units of mTorr. c Concentrations in units
of Torr. d NO present in 15 mTorr concentration. e Total pressure in units of Torr.
Figure 2. IR spectra before (A) and after (B) irradiation of a mixture
of 23 mTorr of DME, 23 mTorr of Cl2, and 9.9 Torr of O2 with N2
added to 700 Torr total pressure. (C) and (D) show IR spectra before
and after irradiation of a mixture of 23.1 mTorr of DME, 23 mTorr of
Cl2, and 10.4 Torr of O2. Reference spectra of methyl formate and
Figure 3. Observed formation of methyl formate (3), methoxy
methylhydroperoxide (b), formaldehyde (1), CO (9), and formic acid
(2) in experiment 6 (A) and experiment 5 (B).
formaldehyde are given in the bottom panels. See text for details.
formaldehyde, CO, methyl formate, methoxy methylhydroper-
observed in Figure 2B is methyl formate. Methyl formate and
methoxy methylhydroperoxide were the only products observed
in experiment 6. The yields of methyl formate and methoxy
methylhydroperoxide were 1.95 and 0.86 mTorr, respectively.
The product spectrum in Figure 2D shows that methyl formate
and formaldehyde are formed at 10.4 Torr total pressure
(experiment 5). The yield of methyl formate was 1.17 mTorr
while that of HCHO was 2.46 mTorr. Examination of other
regions of the product spectrum showed that formic acid,
methoxy methylhydroperoxide, and CO were also formed in
the low-pressure experiments in yields of 0.07, 0.39, and 0.20
mTorr, respectively.
Parts A and B of Figure 3 show the observed yields of
formaldehyde, CO, methyl formate, methoxy methylhydroper-
oxide, and formic acid plotted vs the loss of CH3OCH3 in
experiments 6 (700 Torr total pressure) and 5 (10.4 Torr total
pressure), respectively. Linear least-squares analysis of the data
in Figure 3A gives molar yields for methyl formate and methoxy
methylhydroperoxide of 67 ( 2% and 30 ( 5%, respectively.
Analysis of the data in Figure 3B gives molar yields for
oxide, and formic acid of 89 ( 10%, 6 ( 3%, 42 ( 4%, 14 (
3%, and 2 ( 2%, respectively. These values are also reported
in Table 1. With one exception (the HCHO yield plotted in
Figure 4) all product yields in this paper are expressed as moles
of product formed per mole of DME consumed. To calculate
the carbon yield balance, allowance must be made for the fact
that HCHO, CO, and HCOOH contain one carbon atom while
methyl formate and methoxy methylhydroperoxide contain two.
Quoted errors are 2 standard deviations and do not include
potential systematic errors associated with uncertainties in the
calibrations of the reference spectra. We estimate that potential
systematic errors could contribute an additional 10% range.
Figure 4 shows the yields of formaldehyde, methyl formate,
and methoxy methylhydroperoxide as a function of the total
pressure. The yields and experimental conditions are also shown
in Table 1. The product yields are corrected for CH3OCH2Cl
formed via reaction 4. The corrections were small (<10%).
The product data determined by Jenkin et al.4 are also shown
in Figure 4. Clearly, there is good agreement between the data
obtained here and those reported by Jenkin et al.4 The total

17222 J. Phys. Chem., Vol. 100, No. 43, 1996
Sehested et al.
(11)
HCHO + Cl f HCO + HO2
HCO + O2 f CO + HO2
HCHO + HO2 h HOCH2O2
(12)
(-13,13)
HOCH2O2 + HOCH2O2 f 2HOCH2O + O2 (14)
HOCH2O + O2 f HCOOH + HO2
(15)
(16)
HOCH2O2 + HO2 f HCOOH + H2O + O2
As seen from Figure 4, the yield of CH3OCH2OOH drops
with decreasing pressure. There was no detectable CH3OCH2-
OOH product (<5% yield) for experiments conducted below 2
Torr. There are two reasons for the declining yield of CH3-
OCH2OOH. First, there is the decreased yield of CH3OCH2O2
radicals at low pressure. Second, with low total pressures of
O2 diluent, H atom elimination rather than reaction with O2
becomes an important fate of CH3OCH2O radicals.4 At low
total pressures H atoms react with Cl2, thereby avoiding HO2
radical formation. Without HO2 radicals there can be no CH3-
OCH2OOH formation.
There are four possible sources of the large yield of
formaldehyde observed at low pressures. The formaldehyde is
formed from the thermal decomposition of CH3OCH2 radicals
(reactions 17-20), from the decomposition of the alkoxy radical,
CH3OCH2O (reaction 21), from a bimolecular channel of
reaction 1 (to give ethylene oxide (reaction 1d) followed by
conversion of the ethylene oxide into formaldehyde by some
(unknown) mechanism (reaction 22), or from the reaction of
CH3OCH2 with O2 via reaction 1b:
Figure 4. Molar yields (relative to the dimethyl ether loss) of methyl
formate (3), methoxy methylhydroperoxide (b), and the molar yield
of formaldehyde (1) divided by 2 vs total pressure. The total carbon
yield is shown with hollow squares. Results reported previously by
Jenkin et al.4 are given by the symbols enclosed by large circles. The
smooth lines are fits to polynomial expressions to aid visual inspection
of the data.
product yields are also shown in Figure 4 and are, within the
experimental uncertainties (∼10%), indistinguishable from
100%.
At 700 Torr total pressure the C2 compounds, methyl formate
and methoxy methylhydroperoxide, are the only observable
products of dimethyl ether oxidation. This can be explained
by the following reactions:
CH3OCH2 + M f CH3 + HCHO + M
CH3 + O2 + M f CH3O2 + M
(17)
(18)
(19)
(20)
(21)
Cl2 + hν f 2Cl
(2)
(3)
(1)
CH3O2 + RO2 f RO + CH3O + O2
CH3O + O2 f HCHO + HO2
Cl + CH3OCH3 f CH3OCH2 + HCl
CH3OCH2 + O2 + M f CH3OCH2O2 + M
CH3OCH2O + M f CH3O + CH2O + M
CH3OCH2O2 + CH3OCH2O2 f
CH3OCH2O + CH3OCH2O + O2 (6a)
CH3OCH2 + O2 f
CH2OCH2 (ethylene oxide) + HO2 (1d)
CH3OCH2O2 + CH3OCH2O2 f
CH3OCH2OH + CH3OCHO + O2 (6b)
CH2OCH2 (ethylene oxide) f f HCHO + HCHO (22)
CH3OCH2O + M f CH3OCHO + H + M
CH3OCH2O + O2 f CH3OCHO + HO2
H + O2 + M f HO2 + M
(7)
(8)
(9)
CH3OCH2 + O2 f 2HCHO + OH
(1b)
Two pieces of information show that reaction 17 is not important
in the present reaction system. First, the thermal decomposi-
tion of CH3OCH2 radicals has been studied by Louck and
Laidler18 and Sehested et al.14 The following expression was
derived for the high-pressure limit of the thermal decomposi-
tion of CH3OCH2 radicals: k14 ) 1.6 × 1013 exp(-12832/T)
CH3OCH2O2 + HO2 f CH3OCHO + O2 + H2O (10a)
CH3OCH2O2 + HO2 f CH3OCH2OOH + O2 (10b)
s-1 14,18
From this expression it can be calculated that the
.
lifetime of CH3OCH2 radicals at 296 K is 4.9 days and that
reaction 17 is not important in the present chemical system.
Second, no CH3Cl was observed in the relative rate experiments
described in section 3.1. If CH3 radicals were produced in
reaction 17, then CH3Cl formed by reaction 23 would be a major
product in the reaction system where [O2]/[Cl2] ≈ 10.
Reaction 6 proceeds via two channels with k6a/(k6a + k6b) )
0.7 ( 0.1.4 The yield of CH3OCH2OH in this system is <15%.4
At lower pressures the C1 products, formaldehyde, CO, and
HCOOH, are the dominant products of the oxidation of dimethyl
ether. At 0.38 Torr total pressure the yield of methyl formate
is only 4% while formaldehyde is observed in a yield that
accounts for greater than 90% of the loss of DME. CO and
HCOOH are seen in small yields at low total pressure. Their
CH3 + O2 + M f CH3O2 + M
(18)
CH3 + Cl2 f CH3Cl + Cl
(23)
formation can be rationalized by the following reactions16,17
:

Dimethyl Ether Oxidation
J. Phys. Chem., Vol. 100, No. 43, 1996 17223
At 5 Torr total pressure the rate constant ratio k18/k23 ) 0.05 19
and k18/k23 will decrease further at lower pressures. In a system
where [O2]/[Cl2] ≈ 10 the main fate of methyl radicals formed
in reaction 17 is formation of CH3Cl via reaction 23. No CH3-
Cl was observed in any of the experiments described in section
3.1. Hence, we conclude that reaction 17 is not important.
Decomposition of CH3OCH2O radicals via C-O bond
cleavage is another possible source of formaldehyde. The
decomposition rate of CH3OCH2O radicals is expected to be
greatest at high pressures. To provide insight into the impor-
tance of the thermal decomposition of CH3OCH2O radicals, two
experiments were performed where mixtures of DME/Cl2/NO
with either 1.2 or 140 Torr of oxygen with N2 added to 700
Torr total pressure were irradiated. Under such experimental
conditions essentially all CH3OCH2 radicals will be converted
into CH3OCH2O2 radicals and thence to CH3OCH2O radicals.
The experimental details are given in Table 1 (experiments 7
and 8). Methyl formate was the only observable product in
experiments 7 and 8. An upper limit of 2% was established
for the HCHO yield. This result shows that the fate of CH3-
OCH2O radicals is elimination of a H atom via reaction 7, or
reaction with O2, or both.
Figure 5. Plot of the combined yield of the C2 products, methyl
formate, and methoxy methylhydroperoxide × k1/k4 (kRO2/k4 (b)) and
the C1 products, formaldehyde, HCOOH, and CO × k1/k4 (kprod/k4 (3))
vs the total pressure. The smooth, dashed, and dotted lines are fits using
Fc ) 1 (Lindemann expression), 0.8, and 0.6, respectively. See text
for details.
CH3OCH2O + M f CH3OCHO + H + M
CH3OCH2O + O2 f CH3OCHO + HO2
(7)
(8)
where kRO2 and k1 are the rate constants for the formation of
CH3OCH2O2 radicals and the overall rate constant for reaction
1, respectively. Similarly, the fraction of reaction 1 that gives
formaldehyde and OH radical is given by
Since thermal decomposition of CH3OCH2O radicals via C-O
bond cleavage is not important at 700 Torr total pressure and
[O2] ) 1.2 Torr, it cannot be important at the lower total
pressures used for experiments 1-5 (see Table 1).
A direct hydrogen abstraction mechanism for reaction 1
followed by reaction 22 can be proposed as a source of
formaldehyde at low pressures:
â ) kprod/k1
(IV)
where kprod is the rate constant for the formation of formaldehyde
and OH radicals. kRO2/k4 and kprod/k4 can be calculated directly
from the experimental data by multiplying the fraction of the
products that are C1 and C2 products by the values of k1/k4
derived in section 3.1. Figure 5 shows the resulting values of
kRO2/k4 and kprod/k4 calculated by multiplying the product data
given in Figure 4 by appropriate values of k1/k4 derived from
the fit through the data in section 3.1. These two data sets can
be fitted by the following expressions derived in section 3.1:
CH3OCH2 + O2 f
CH2OCH2 (ethylene oxide) + HO2 (1d)
CH2OCH2 (ethylene oxide) f f HCHO + HCHO (22)
However, if the formation of formaldehyde and the nonzero
intercept in Figure 1 were due to such a pressure independent,
direct bimolecular channel of reaction 1, then inspection of
Figure 1 reveals that reaction 1d would account for ap-
proximately half the overall rate constant at 700 Torr. At
pressures below 1 Torr the reaction of CH3OCH2 radicals with
O2 results in the conversion of essentially all CH3OCH2 radicals
into formaldehyde. If reaction 1d was the explanation of the
formaldehyde formation and the nonzero intercept in Figure 1
then we would expect that at 700 Torr total pressure ap-
proximately 50% of CH3OCH2 radicals would be converted into
formaldehyde. In contrast, the formaldehyde yield from dim-
ethyl ether oxidation at 700 Torr total pressure is <2%. Hence,
we conclude that reaction 1d is unimportant.
Finally, we need to consider the possible formation of HCHO
directly in the reaction of CH3OCH2 radicals with O2. As
discussed in the previous section, the mechanism proposed for
reaction 1, which includes reactions 1a, 1b, and 1c, explains
the pressure dependence of k1 very well. We will show below
that the reaction mechanism also explains the pressure depen-
dence of the product distribution shown in Figure 4. Assuming
that all thermalized peroxy radicals react to give C2 products,
methyl formate or methoxy methylhydroperoxide, the fraction
of the CH3OCH2 radicals reacting with O2 that give the peroxy
radical, CH3OCH2O2, is given by
kRO2/k4 ) kRO2,0[M]/{k4(1 + kRO2,0[M]/kRO2,∞)} (V)
kprod/k4 ) kprod,0/{k4(1 + kRO2,0[M]/kRO2,∞)}
(VI)
using kRO2,0/k4 and kRO2,∞/k4 (eq V) and kprod,0/k4 and kRO2,0
kRO2,∞ (eq VI) as parameters. The fits are given as solid lines
in Figure 5. The parameters obtained from the fits were kRO2,0
/
/
k4 ) (6.45 ( 0.92) × 10-3 Torr-1 ) (1.97 ( 0.28) × 10-19
cm3 molecule-1 and kRO2,∞/k4 ) 0.104 ( 0.007 (kRO2 fit) and
kprod,0/k4 ) (6.6 ( 0.9) × 10-2 and kRO2,0/kRO2,∞ ) (8.4 ( 3.2)
× 10-2 Torr-1 ) (2.6 ( 1.0) × 10-18 cm3 molecule-1 (kprod
fit). These values are consistent with the rate constant ratios
obtained in section 3.1. We choose to report the average of
the values of kRO2,∞/k4 and kprod,0/k4 determined in section 3.1
and in this section. Hence, kRO2,∞/k4 ) 0.108 ( 0.004 and kprod,0
/
k4 ) (6.3 ( 0.5) × 10-2; uncertainties were calculated by normal
error propagation methods. For kRO2,0/k4 we choose to report
the more precise value determined in this section, kRO2,0/k4 )
(1.97 ( 0.28) × 10-19 cm3 molecule-1
.
As a check of the validity of the use of a modified Lindemann
mechanism to analyze the data obtained in this work, the kRO2
/
k4 data in Figure 5 were also analyzed by using a Troe type
expression:
R ) kRO2/k1
(III)

17224 J. Phys. Chem., Vol. 100, No. 43, 1996
kRO2/k4 ) kRO2,0[M]/(k4(1 +
Sehested et al.
thermal decomposition of CH3OCH2 radicals is too slow to be
of any importance under the present experimental conditions.
Therefore, the experiments conducted in the presence of a large
excess of methanol suggest that reaction 1 is indeed the source
of formaldehyde.
3.4. Apparent Photolysis Rate of Cl2 at 2, 5, and 700 Torr
Total Pressure. The apparent photolysis rate of Cl2, as
measured from the rate of dimethyl ether loss, was determined
for three sets of initial experimental conditions. During these
experiments only 4 out of the 22 blacklamps were used to
photolyze molecular chlorine. The initial concentrations were
[DME] ) 10 mTorr, [Cl2] ) 33 mTorr, and [O2] ) 2
(experiment 1), 2 (experiment 2), or 140 (experiment 3) Torr
with N2 added to 700 Torr total pressure in experiments 2 and
3. The photolysis rates were calculated by eq VIII:
RO2,0[M]/kRO2,•)]2)-1
kRO2,0[M]/kRO2,∞))Fc(1+[log(k
(VII)
where kRO2,0 ) k1ak1c/(k-1a + k1b), kRO2,∞ ) k1a, and kRO2 is the
overall rate constant for RO2 formation. kRO2,0, kRO2,∞, and kRO2
are also used in eq II and discussed in section 3.1. The Troe
type expression (eq VII) describes the pressure dependence of
a reaction system consisting of reactions 1a-c. Introducing
the additional reaction 1b in the present chemical system should
not alter the pressure dependence of kRO2. The fits to the data
using eq VII and Fc ) 0.6 (dotted lines) and Fc ) 0.8 (dashed
lines) are shown in Figure 5. The Troe expression (eq VII) is
equal to the Lindemann expression (eq II) when Fc ) 1. As
seen from Figure 5, the fits using Fc ) 1 give the best
representation of the kRO2/k4 data.
3.3. Product Study of CH3OCH2 + O2 in the Presence of
13CH3OH. As a part of this work, a product study of the
oxidation of dimethyl ether in the presence of a large excess of
13CH3OH was performed at 2, 5, and 700 Torr total pressure.
The initial conditions and the results are given in Table 2.
Chlorine-initiated oxidation of methanol gives HO2 radicals:
JCl2
)
(∆[CH3OCH3] - ∆[CH3OCH2Cl])/(2[Cl2]∆t) (VIII)
where ∆[CH3OCH3] is the observed loss of dimethyl ether,
∆[CH3OCH2Cl] is the observed formation of CH3OCH2Cl, ∆t
is the irradiation time, and [Cl2] the initial chlorine concentration.
The photolysis rates were determined to be (1.05 ( 0.10) ×
10-3 s-1 (experiment 1), (3.8 ( 0.2) × 10-4 s-1 (experiment
2), and (3.8 ( 0.2) × 10-4 s-1 (experiment 3). As expected,
the apparent photolysis rate of Cl2 in 700 Torr of air (experiment
3) is the same as that measured in 2 Torr of oxygen and 698
Torr of N2 (experiment 2). However, the apparent photolysis
rate in 2 Torr total pressure of O2 (experiment 1) is 2.8 ( 0.3
times higher than those observed at 700 Torr total pressure
(experiments 2 and 3). This observation suggests that a chain
reaction is operating at low pressures (i.e., more than one
molecule of CH3OCH3 is consumed for each Cl atom formed
in the system). In light of the reaction mechanism proposed
for the reaction of CH3OCH2 radicals with O2 in the previous
sections, it seems reasonable to conclude that the apparent
enhancement of the Cl2 photolysis rate can be ascribed to
additional loss of CH3OCH3 via reaction with OH radicals
produced in reaction 1b.
As an additional check on this chemical system, the concen-
tration of HCl after 25 s irradiation of reaction mixture 1 was
compared with that in reaction mixture 3 after 55 s irradiation.
The loss of dimethyl ether in these two experiments were
comparable; 1.55 mTorr in experiment 1 and 1.34 mTorr in
experiment 3. The yield of HCl in reaction mixture 1 was found
to be 0.55 ( 0.20 times that in reaction mixture 3. Hence, the
HCl yield per dimethyl ether molecule consumed in experiment
1 is 0.48 ( 0.17 of that in experiment 3. This result is entirely
consistent with the magnitude of the enhancement of the
apparent Cl2 photolysis rate.
CH3OH + Cl f CH2OH + HCl
CH2OH + O2 f HCHO + HO2
(24)
(25)
HO2 radicals react with CH3OCH2O2 radicals via two reaction
pathways:
CH3OCH2O2 + HO2 f CH3OCHO + O2 + H2O (10a)
CH3OCH2O2 + HO2 f CH3OCH2OOH + O2 (10b)
All experiments were conducted using an initial concentration
ratio of [13CH3OH]/[DME] ) 10. This ratio was selected
because previous experiments5 have shown that under such
conditions the sole fate of CH3OCH2O2 radicals is reaction with
HO2 radicals. 13CH3OH was used as opposed to CH3OH in
this study to avoid formation of H12CHO from methanol in the
system. By use of 13CH3OH, dimethyl ether oxidation is the
only source of H12CHO.
As seen from Tables 1 and 2, the yields of formaldehyde
and C2 products (methyl formate and methoxy methylhydrop-
eroxide) obtained with 13CH3OH present are indistinguishable
from those obtained without methanol present in the chamber
apart from the yields of C2 products reported at 2.2 Torr total
pressure. The yield of C2 products with 13CH3OH present is
29 ( 6% (2.2 Torr total pressure). The yields of C2 products
reported in Table 1 are 11 ( 2% at 1.66 Torr and 32 ( 7% at
5.7 Torr total pressure. This apparent discrepancy may be due
to a very high third-body efficiency of methanol, which would
increase the yield of C2 products at the expense of formaldehyde.
We therefore believe that the yields of C1 and C2 products from
dimethyl ether oxidation with and without 13CH3OH are
indistinguishable.
4. Discussion
The experimental observations presented in sections 3.1-
3.4 constitute a large body of self-consistent kinetic and product
data that show that the reaction of CH3OCH2 radicals with O2
proceeds via the formation of an excited complex (CH3-
The relative yields of methyl formate and the methoxy
methylhydroperoxide with methanol present in the system are
consistent with those reported by Wallington et al.5 of 40 (
4% methyl formate and 53 ( 8% methoxy methylhydroperox-
ide.
#
OCH2O2 ). This complex has one of three fates: decomposition
back into reactants, collisional stabilization to give thermalized
CH3OCH2O2 radicals, or isomerization via intramolecular
H-atom abstraction followed by decomposition to give two
formaldehyde molecules and an OH radical:
It is interesting to note that the yield of formaldehyde from
13
dimethyl ether oxidation is not affected by the presence of
-
#
CH3OH. CH3OCH2O2 radicals are known to be rapidly
scavenged by HO2 radicals in the system to give CH3OCHO
and CH3OCH2OOH.5 The insensitivity of the HCHO yield to
the presence of CH3OH suggests that HCHO is formed before
CH3OCH2O2 radicals are produced. As discussed above, the
CH3OCH2 + O2 h CH3OCH2O2
(1a,-1a)
CH3OCH2O2# f CH2OCH2O2H# f 2HCHO + OH (1b)
CH3OCH2O2# + M f CH3OCH2O2 + M
(1c)

Dimethyl Ether Oxidation
J. Phys. Chem., Vol. 100, No. 43, 1996 17225
TABLE 2: Product Yieldsa Following the Irradiation of CH3OCH3/13CH3OH/Cl2/O2 Mixtures
Y(CH3OCHO)
Y(CH3OCH2OOH)
Y(HCHO)
Y(CO)
%
[DME]/[13CH3OH]b
[O2]c
[N2]c
[Cl2]o
%
%
%
b
set
1
2
3
10.6/117
10.4/104
10.4/104
2.2
10.5
140
0
0
560
33
33
33
14 ( 1
29 ( 3
46 ( 6
15 ( 5
42 ( 13
56 ( 8
130 ( 14
78 ( 12
9 ( 4
a Molar yields relative to CH3OCH3 loss corrected for formation of CH3OCH2Cl. b Concentrations in units of mTorr. c Concentrations in units
of Torr.
In section 3.1 it was shown that the rate constant for reaction 1
possible. It is well-known that intramolecular hydrogen transfer
is important in excited alkyl peroxy radicals at high temperatures
(500-1000 °C).8 Reaction 1b may be the major pathway for
the reaction of CH3OCH2 radicals with O2 under compression-
ignition conditions. If so, this might explain the high cetane
number of dimethyl ether. In addition, the formation of
formaldehyde during dimethyl ether combustion would explain
the low particle emissions from diesel engines powered by
DME. A study of the mechanism for reaction 1 at higher
temperatures is necessary to determine the importance of
reaction 1b in the ignition and combustion of dimethyl ether.
is independent of total pressure below 2 Torr. In section 3.2
the products of reaction 1 were studied and it was shown that
formaldehyde is the dominant product below 10 Torr. In section
3.3 the products of dimethyl oxidation were studied in the
presence of a source of a large excess of HO2 radicals such
that any CH3OCH2O2 radical would be scavenged by reaction
with HO2 radicals. The formaldehyde yield was insensitive to
the presence of excess HO2 radicals, suggesting that formalde-
hyde formation occurs before the formation of CH3OCH2O2
radicals. This experiment provides evidence that formaldehyde
is formed via reaction 1b. Finally, the apparent chlorine
photolysis rate and the HCl yield were studied at 2 and 700
Torr total pressure. The apparent photolysis rate was enhanced
by a factor of 2.8 ( 0.3, while the HCl yield was suppressed
by a factor of 0.48 ( 0.17 at 2 Torr total pressure compared to
700 Torr. These observations are consistent with the formation
of OH radicals in reaction 1b.
Acknowledgment. We thank Steve Japar for commenting
on the manuscript and Geoff Tyndall for helpful insight into
the mechanism for reaction 1.
References and Notes
(1) Roubi, M. A. Chem. Eng. 1995, 44, 37.
(2) Fleisch, T.; McCarthy, C.; Basu, A.; Udovich, C.; Charnonneau,
P.; Slodowske, W.; Mikkelsen, S.; McCandless, J. Soc. Automot. Eng. 1995,
950061, 1.
5. Implications for Atmospheric Chemistry
(3) Japar, S. M.; Wallington, T. J.; Richert, J. F. O.; Ball; J. C. Int. J.
Chem. Kinet. 1990, 22, 1257.
(4) Jenkin, M. E.; Hayman, G. D.; Wallington, T. J.; Hurley, M. D.;
Ball, J. C.; Nielsen, O. J.; Ellermann, T. J. J. Phys. Chem. 1993, 97, 11723.
(5) Wallington, T. J.; Hurley, M. D.; Ball, J. C.; Jenkin, M. E. Chem.
Phys. Lett. 1993, 211, 41.
(6) Langer, S.; Ljungstro¨m, E.; Ellermann, T.; Nielsen, O. J.; Sehested,
J. Chem. Phys. Lett. 1995, 240, 53.
(7) Cox, R. A. In Modern Gas Kinetics; Pilling, M. J., Smith, I. W.
M., Eds.; Blackwell: Oxford, 1987; p 262.
(8) Hughes, K. J.; Lightfoot, P. D.; Pilling, M. J. Chem. Phys. Lett.
1992, 191, 581.
(9) Westbrook C. K.; Pitz, W. J.; Leppard, W. R. Soc. Automot. Eng.
1991, 912314, 605.
(10) Dagaut, P.; Boettner, J. C.; Cathonnet, M. 26th International
Symposium on Combustion, Napoli, Italy, July 28-August 2, 1996.
(11) Calcote, H. F. Combust. Flame 1981, 42, 215.
(12) Wallington, T. J.; Gierczak, C. A.; Ball, J. C.; Japar, S. M. Int. J.
Chem. Kinet. 1989, 21, 1077.
(13) Masaki, A.; Tsunashima, S.; Washida, N. J. Phys. Chem. 1995,
99, 13126.
The atmospheric oxidation of dimethyl ether is initiated by
reaction with OH radicals. The reaction of OH radicals with
dimethyl ether gives CH3OCH2 radicals whose sole fate is
reaction with O2. The rate constant at 298 K for the reaction
of OH radicals with dimethyl ether is 260 times (2.5 × 10-12 15
/
9.6 × 10-15 20) that of reaction with CH3CCl3, which has an
atmospheric lifetime of 5.7 years. Hence, the atmospheric
lifetime of DME is approximately 8 days. We have shown
herein that at 296 K the reaction of CH3OCH2 radicals with O2
proceeds via two channels to give either CH3OCH2O2 radicals
or HCHO and OH radicals. For pressures above ∼10 Torr the
formation of CH3OCH2O2 radicals dominates, while for pres-
sures less than ∼10 Torr the formation of HCHO and OH
radicals is dominant. Since the vast majority of the atmospheric
oxidation of dimethyl ether occurs at pressures substantially
above 10 Torr and with the reasonable assumption that the
importance of reaction 1c does not decrease dramatically with
temperature over the atmospherically relevant range 220-296
K, it follows that the vast majority of the dimethyl ether released
to the atmosphere is oxidized via reaction 1c.
(14) Sehested, J.; Sehested, K.; Platz, J.; Egsgaard, H.; Nielsen, O. J.
Int. J. Chem. Kinet., submitted for publication.
(15) Mallard, W. G.; Westley, F.; Herron, J. T.; Hampson, R. F. NIST
Chemical Kinetics Database, Version 4.0; NIST Standard Reference Data;
NIST: Gaithersburg, MD, 1994.
(16) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F.; Kerr, J.
R.; Troe, J. J. Phys. Chem. Ref. Data 1992, 21, 1125.
(17) Burrows, J. P.; Moortgat, G. K.; Tyndall, G. S.; Cox, R. A.; Jenkin,
M. E.; Hayman, G. D.; Veyret, B. J. Phys. Chem. 1989, 93, 2374.
(18) Loucks, L. F.; Laidler, K. Can. J. Chem. 1967, 45, 2768.
(19) Kaiser, E. W. J. Phys. Chem. 1993, 97, 11681.
(20) Ravishankara, A. R.; Lovejoy, E. R. J. Chem. Soc., Faraday Trans.
1994, 90, 2159.
6. Implications for Combustion Chemistry
The mechanism for reaction 1 at 296 K found in this work
may explain why DME is a good diesel fuel. By analogy with
the formation of ethene from the reaction of ethyl radicals with
O2,21 it is expected that the relative importance of reaction 1b
will increase with temperature because reaction 1c becomes
reversible and multiple passes through reaction 1a become
(21) Kaiser, E. W. J. Phys. Chem. 1995, 99, 707.
JP961821M