Atmospheric Chemistry of 1,3-Dioxolane: Kinetic, Mechanistic, and Modeling Study of OH Radical Initiated Oxidation

Article abstract of DOI:10.1021/jp991025r

An absolute rate pulse radiolysis technique was used to measure k(OH + 1,3-dioxolane) = (8.8 ± 0.9) × 10^-12 cm³ molecule^-1 s^-1 at 295 K in 1000 mbar of Ar. Relative rate techniques were used to study the reactions of OH radicals and Cl atoms with 1,3-dioxolane and Cl atoms with ethylene carbonate and methylene glycol diformate at 300 K in 1 bar of synthetic air. Rate coefficients were k(OH + 1,3-dioxolane) = (1.04 ± 0.16) × 10^-11, k(Cl + 1,3-dioxolane) = (1.6 ± 0.3) × 10^-10, k(Cl + ethylene carbonate) = (7.1 ± 1.7) × 10^-12, and k(Cl + methylene glycol diformate) = (5.6 ± 0.7) × 10^-13 cm³ molecule^-1 s^-1. OH radical and chlorine atom initiated oxidation of 1,3-dioxolane in 1 bar of N₂/O₂ mixtures at 298 K in the presence of NO_x gives ethylene carbonate and methylene glycol diformate. Molar yields of ethylene carbonate and methylene glycol diformate were 0.48 ± 0.07 and 0.50 ± 0.14 for OH radical initiation and 0.43 ± 0.07 and 0.53 ± 0.07 for Cl atom initiation. Product yields were independent of O₂ partial pressure over the range studied (60-800 mbar). A photochemical mechanism was developed to describe the OH-initiated degradation of 1,3-dioxolane in the presence of NO_x. The results are discussed with respect to the available literature data concerning the atmospheric chemistry of ethers.

Full text of DOI:10.1021/jp991025r

J. Phys. Chem. A 1999, 103, 5959-5966
5959
Atmospheric Chemistry of 1,3-Dioxolane: Kinetic, Mechanistic, and Modeling Study of OH
Radical Initiated Oxidation
C. G. Sauer, I. Barnes,* K. H. Becker, and H. Geiger
Bergische UniVersita¨t GH Wuppertal/FB 9sPhysikalische Chemie, Gaussstrasse 20,
D-42119 Wuppertal, Germany
T. J. Wallington* and L. K. Christensen
Ford Motor Company, 20000 Rotunda DriVe, Mail Drop SRL-3083, Dearborn, Michigan 48121-2053
J. Platz and O. J. Nielsen
Atmospheric Chemistry, Plant Biology and Biogeochemistry Department, Risø National Laboratory,
DK-4000, Roskilde, Denmark
ReceiVed: March 24, 1999; In Final Form: June 1, 1999
An absolute rate pulse radiolysis technique was used to measure k(OH + 1,3-dioxolane) ) (8.8 ( 0.9) ×
10-12 cm3 molecule-1 s-1 at 295 K in 1000 mbar of Ar. Relative rate techniques were used to study the
reactions of OH radicals and Cl atoms with 1,3-dioxolane and Cl atoms with ethylene carbonate and methylene
glycol diformate at 300 K in 1 bar of synthetic air. Rate coefficients were k(OH + 1,3-dioxolane) ) (1.04 (
0.16) × 10-11, k(Cl + 1,3-dioxolane) ) (1.6 ( 0.3) × 10-10, k(Cl + ethylene carbonate) ) (7.1 ( 1.7) ×
10-12, and k(Cl + methylene glycol diformate) ) (5.6 ( 0.7) × 10-13 cm3 molecule-1 s-1. OH radical and
chlorine atom initiated oxidation of 1,3-dioxolane in 1 bar of N2/O2 mixtures at 298 K in the presence of NOx
gives ethylene carbonate and methylene glycol diformate. Molar yields of ethylene carbonate and methylene
glycol diformate were 0.48 ( 0.07 and 0.50 ( 0.14 for OH radical initiation and 0.43 ( 0.07 and 0.53 (
0.07 for Cl atom initiation. Product yields were independent of O2 partial pressure over the range studied
(60-800 mbar). A photochemical mechanism was developed to describe the OH-initiated degradation of
1,3-dioxolane in the presence of NOx. The results are discussed with respect to the available literature data
concerning the atmospheric chemistry of ethers.
1. Introduction
validation by comparison of the measured product concentrations
against model simulations. In other instances, where no suitable
experimental results are available, reaction mechanisms are
postulated without any further verification. Both methods result
in the transfer of considerable uncertainty to the results of
models when applied, e.g., to the modeling of field data. As a
consequence, this work also includes the development of an
explicit photochemical degradation mechanism, which was
constructed by using the results obtained from the photoreactor
experiments. The reaction scheme was validated by comparison
of simulated and experimental concentration/time profiles.
Altogether, the present work provides an intensive kinetic,
product, and modeling study on the OH-initiated oxidation of
1,3-dioxolane.
Ethers are an important class of automotive fuel additives.
For example, methyl tert-butyl ether (MTBE), ethyl tert-butyl
ether (ETBE), tert-amyl methyl ether (TAME), and diisopropyl
ether (DIPE) are widely used octane-enhancing additives in
gasoline. In addition, ethers such as dimethyl ether1 and
dimethoxy methane are under consideration as alternative diesel
fuels. To predict the environmental impact of ether fuels and
fuel additives, it is desirable to understand how the structure of
such molecules impacts their atmospheric reactivity and deg-
radation pathways. While 1,3-dioxolane is not used currently
as an automotive fuel additive, it is used as an industrial solvent.
1,3-Dioxolane is a convenient model compound whose chem-
istry provides insight into the behavior of alkyl peroxy and
alkoxy radicals connected to oxygen ether linkages.
2. Experimental Section
As for most chemicals released into the troposphere, the major
atmospheric fate of 1,3-dioxolane will be reaction with OH
radicals. We report here the results of a collaborative experi-
mental study of the kinetics and mechanism of the OH radical
initiated oxidation of 1,3-dioxolane.
Four experimental systems were used.2-5 They are described
briefly here.
2.1. Pulse Radiolysis System at Risø National Laboratory.
To study the reaction between OH radicals and 1,3-dioxolane,
Mechanistic studies are often limited to the detection of the
primary reaction products. Reaction mechanisms, which are
derived from these data, have not usually been subject to
OH + 1,3-dioxolane f products
(1)
OH radicals were generated from H2O by pulsed radiolysis of
Ar/H2O gas mixtures in a 1 L stainless steel reaction cell with
a 30 ns pulse of 2 MeV electrons from a Febetron 705 field
* To whom correspondence should be adressed. E-mail: barnes@
physchem.uni-wuppertal.de. E-mail: twalling@ford.com.
10.1021/jp991025r CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/14/1999

5960 J. Phys. Chem. A, Vol. 103, No. 30, 1999
Sauer et al.
emission accelerator2. Ar diluent was always in great excess
and was used to generate Ar* atoms as follows:
[reactant]t
[reactant]t
[reference]t
0
k
ln
0 - k8t ) 6ln
k7
[reference]t
Ar + 2MeV e- f Ar*
(2)
(3)
where k8 is the rate coefficient for reaction 8.
Ar* + H2O f OH + H + Ar
Reagent compounds were injected into the reactor in a stream
of synthetic air using microliter syringes or gastight syringes.
Ethylene carbonate is a solid and was introduced into the
chamber by heating and evaporating a known mass of the
compound into a stream of synthetic air. Initial concentrations
were in the range 0.1-20 ppm (1 ppm ) 2.46 × 1013 molecules
cm-3). In the 20 L photoreactor, initial concentrations were
about 1 order of magnitude higher. Samples were taken from
the small reactor with a gastight syringe and analyzed by GC/
FID (gas chromatograph HP 5890, RTx1 capillary column 30
m × 0.53 mm × 1 µm or m-AT 1701 capillary column 30 m
× 0.53 mm × 1.2 µm, electronic integrator HP 3396).
The radiolysis dose was varied by insertion of stainless steel
attenuators between the accelerator and the reaction cell. OH
radicals were monitored using their characteristic absorption at
309.25 nm. The initial OH concentration was approximately
1013 cm-3. The analysis light was provided by a pulsed 150 W
Xenon arc lamp and was multipassed through the reaction cell
using internal White cell optics to give a total optical path length
of 120 cm. After leaving the cell, the light was guided through
a monochromator (operated with a spectral resolution of 1 nm)
and detected by a photomultiplier. All absorption transients were
derived from single pulse experiments.
Reagent purities and concentrations were as follows: Ar
(ultrahigh purity) 980-990 mbar; H2O (triple distilled deionized
water) 5-15 mbar; 1,3-dioxolane (>99%) 0.8-2.4 mbar. All
experiments were performed at 295 ( 2 K and 1000 mbar total
pressure. 1,3-Dioxolane was degassed by freeze-pump-thaw
cycling before use, all other reagents were used as received.
2.2. FTIR-Smog Chamber System at Wuppertal. The
majority of experiments were performed in a 1080 L quartz
photoreactor surrounded by 32 fluorescent lamps (Philips TL05/
40 W) and 32 mercury lamps (Philips TUV/40 W)3. A
multireflection White mirror system (484.7 m optical path
length) mounted in the reactor and coupled to a Fourier
transform infrared (FTIR) spectrometer Bruker IFS-88 allowed
in situ monitoring of both reactants and products. Additional
experiments were carried out in a 20 L glass photoreactor
surrounded by four fluorescent lamps.4
Product studies were performed in the 1080 L quartz
photoreactor at 298 ( 4 K and 1000 ( 20 mbar total pressure
of synthetic air or O2/N2 mixtures with different O2 partial
pressures in the range 60-800 mbar. Photolysis of methyl nitrite
in the presence of NO was used as an OH source.
CH3ONO + hν f CH3O + NO
CH3O + O2 f HO2 + HCHO
HO2 + NO f OH + NO2
(9)
(10)
(11)
The photolysis of chlorine, reaction 5, served as a source for
Cl atoms. Initial concentrations were 2.8-4.9 ppm of 1,3-
dioxolane, 2-20 ppm of methyl nitrite, 1-5 ppm of NO, and
1-5 ppm of Cl2. Individual experiments lasted 5-23 min in
which time 10 FTIR spectra (32 or 64 scans, 1 cm-1 spectral
resolution) were collected. After irradiation, another 10 spectra
were collected to determine the wall loss rate of ethylene
carbonate.
Relative rate kinetic experiments were performed in synthetic
air or nitrogen at total pressures of 1000 ( 40 mbar at 300 (
4 K. OH radicals were generated by the photolysis of hydrogen
peroxide using mercury lamps; Cl atoms were generated by the
photolysis of Cl2 using fluorescent lamps.
1,3-Dioxolane (Aldrich, 99.9%), ethylene carbonate (Aldrich,
98%), cyclohexane (Aldrich, 99.9%), dichloromethane (Aldrich,
HPLC grade), and hydrogen peroxide (Peroxid Chemie, 85%)
were used without further purification. Methyl nitrite was
synthesized by the dropwise addition of concentrated sulfuric
acid to a saturated solution of sodium nitrite in methanol. The
gaseous methyl nitrite was passed over solid calcium chloride,
collected in an ethanol/dry ice cold trap, and stored at -78 °C
in the dark. FT-IR analysis showed neither methanol nor water.
Methylene glycol diformate was prepared from ethylene carbon-
ate in a five-step synthesis7-10 and was handled as a solution
in dichloromethane or deuterotrichloromethane. The following
spectroscopic data were obtained. 1H NMR (400 MHz,
CDCl3): δ(ppm) ) 5.91 (s, 2H), 8.09 (s, 2H). IR, gas (cm-1):
2949, 1769, 1168, 1096, 957. No impurities were observed. All
gases were supplied by Messer Griesheim and had the following
stated purities in vol %: synthetic air (99.995), N2 (99.999),
O2 (99.995), NO (99.8), NO2 (98), Cl2 (99.8), methane (99.995),
ethane (99.95), propane (99.95), n-butane (99.5), chloroethane
(99.0).
H2O2 + hν f 2OH
Cl2 + hν f 2Cl
(4)
(5)
Provided that the reactant and the reference compound are
removed solely by the reaction of interest and are not reformed
in any process,
OH/Cl + reactant f products
OH/Cl + reference f products
(6)
(7)
it can be shown that
[reactant]t
[reference]t
k
0
0
ln
)
6 ln
k7
[reactant]t
[reference]t
where [reactant]t and [reactant]t and [reference]t and [reference]t
0
0
are the reactant and reference concentrations at times t0 and t,
respectively, and k6 and k7 are the rate coefficients for reactions
2.3. FTIR-Smog Chamber System at Ford Motor Com-
pany. Experiments were performed in a 140 L Pyrex reactor
interfaced to a Mattson Sirus 100 FTIR spectrometer.5 The
reactor was surrounded by 22 fluorescent blacklamps (GE
F15T8-BL). Loss of 1,3-dioxolane and formation of products
were monitored by FTIR spectroscopy using an infrared path
length of 28 m and a resolution of 0.25 cm-1. Infrared spectra
were derived from 32 co-added interferograms.
6 and 7. A plot of ln([reactant]t /[reactant]t) versus ln([referen-
0
ce]t /[reference]t) yields a slope of k6/k7. When there is an
0
additional loss process for the reactant,
reactant f products
(8)
the data can be analyzed using the following expression:6

Atmospheric Chemistry of 1,3-Dioxolane
J. Phys. Chem. A, Vol. 103, No. 30, 1999 5961
The products of the reaction of 1,3-dioxolane and Cl atoms,
Cl + 1,3-dioxolane f products
(12)
were investigated by irradiating 1,3-dioxolane/Cl2/NO/O2/N2
mixtures at 296 ( 2 K. Initial concentrations of the gas mixtures
were 17-37 mTorr of 1,3-dioxolane, 96-148 mTorr of Cl2,
16-43 mTorr of NO, and 77-600 Torr of O2 in 700 Torr (760
Torr ) 1013 mbar) total pressure of N2 diluent. A reference
spectrum of ethylene carbonate was acquired by introducing
vapor from a commercial sample into the chamber. Ethylene
carbonate has a vapor pressure at room temperature of ap-
proximately 25 mTorr. It is difficult to measure such low
pressures with great accuracy, and absolute calibration of the
spectrum was achieved using the infrared cross sections
measured at Wuppertal. A reference spectrum of methylene
glycol diformate was generated from irradiation of 1,3,5-
trioxane/Cl2/O2/N2 mixtures11 and calibrated using the infrared
cross section measured at Wuppertal.
In smog chamber experiments, unwanted loss of reactants
and products via photolysis, dark chemistry, and wall reactions
have to be considered. Control experiments were performed to
check for such unwanted losses in the chamber. No significant
loss of 1,3-dioxolane (<1%) due to photolysis was observed.
No significant loss of 1,3-dioxolane (<2%) or methylene glycol
diformate (<2%) was observed when mixtures of these com-
pounds in air were left in the dark for 7-45 min, showing that
heterogeneous loss of 1,3-dioxolane and methylene glycol
diformate in the chamber are not important. When gas mixtures
containing ethylene carbonate were allowed to stand in the dark,
a slow loss of this compound was observed (with first-order
rate constants of 0.01-0.03 min-1) presumably due to deposition
on the reactor walls.
Figure 1. Pseudo-first-order loss of OH radicals versus 1,3-dioxolane
concentration.
3. Results
3.1. Pulse Radiolysis Study of Reaction of OH Radicals
with 1,3-Dioxolane at Risø. A rapid increase in absorption was
observed 0-1 µs after the radiolysis pulse which was attributed
to the formation of OH radicals, which absorb strongly at 309.25
nm. The subsequent decay of absorption followed first-order
kinetics and is ascribed to loss of OH radicals via reaction with
1,3-dioxolane. All transients were well described using the
following expression:
Figure 2. Decay of 1,3-dioxolane versus n-C4H10 and c-C6H12 when
exposed to OH radicals in 1000 mbar total pressure of air at 300 K.
The kinetics of reactions 13 and 14 are well established. In 1
atm of air diluent at 296 K, k13 ) 2.44 × 10-12 and k14 ) 7.49
× 10-12 cm3 molecule-1 s-1 13
. Figure 2 shows plots of the loss
of 1,3-dioxolane versus those of n-C4H10 and cyclohexane.
Linear least-squares analysis gives k1/k13 ) 3.96 ( 0.32 and
k1/k14 ) 1.50 ( 0.08. Hence, values of k1 ) (9.66 ( 0.78) ×
10-12 and (1.12 ( 0.06) × 10-11 cm3 molecule-1 s-1 are
derived. Consistent results were obtained using the two different
reference compounds. We choose to quote the average of the
results above with error limits which encompass the extremes
of the ranges, hence k1 ) (1.04 ( 0.16) × 10-11 cm3 molecule-1
s-1. This result is indistinguishable, within the experimental
uncertainties, from that of k1 ) (8.8 ( 0.9) × 10-12 cm3
molecule-1 s-1 measured using the pulse radiolysis technique
described in the previous section.
[OH]t ) [OH]0 exp(-k1stt)
where [OH]t and [OH]0 are the OH radical concentrations at
time t and time 0 and k1st is the first-order rate constant of
reaction 1. Figure 1 shows a plot of k1st versus the 1,3-dioxolane
concentration. Linear least-squares regression of the data in
Figure 1 gives k1 ) (8.8 ( 0.8) × 10-12 cm3 molecule-1 s-1
.
To account for additional uncertainties associated with the
nonlinear relationship between [OH] and the UV absorption at
309.25 nm,12 we choose to propagate an additional 5%
uncertainty range and obtain a final value of k1 ) (8.8 ( 0.9)
3.3. Cl Atom Relative Rate Studies at Wuppertal and
Ford. Relative rate experiments were performed to investigate
the kinetics of the reactions of Cl atoms with 1,3-dioxolane,
ethylene carbonate, and methylene glycol diformate.
× 10-12 cm3 molecule-1 s-1
.
3.2. Relative Rate Study of Reaction of OH Radicals with
1,3-Dioxolane at Wuppertal. The kinetics of reaction 1 were
studied using n-butane and cyclohexane as reference compounds.
Cl + 1,3-dioxolane f products
(12)
(15)
OH + 1,3-dioxolane f products
OH + n-C4H10 f products
(1)
Cl + ethylene carbonate f products
(13)
Cl + methylene glycol diformate f products (16)
OH + c-C6H12 f products
(14)
The kinetics of reaction 12 were measured relative to reactions

5962 J. Phys. Chem. A, Vol. 103, No. 30, 1999
Sauer et al.
TABLE 1: Measured Cl Atom Rate Constant Ratios
reactant
ethylene
carbonate
methylene glycol
diformate
reference
1,3-dioxolane
CH4
C2H6
C3H8
C2H4
5.53 ( 0.35
2.67 ( 0.15
1.21 ( 0.06
1.78 ( 0.08
CH3OCHO
CH2Cl2
C2H5Cl
4.71 ( 0.70
0.96 ( 0.10
1.70 ( 0.10
TABLE 2: Measured Cl Atom Rate Constants
(cm3 molecule-1 s-1
)
reactant
1,3-dioxolane
ethylene
methylene glycol
reference
(×10-10
)
carbonate (×10-12
)
diformate (×10-13
)
CH4
C2H6
C3H8
C2H4
5.53 ( 0.35
1.52 ( 0.09
1.69 ( 0.09
1.67 ( 0.08
CH3OCHO
CH2Cl2
C2H5Cl
6.59 ( 0.98
7.68 ( 0.77
5.61 ( 0.33
) (1.63 ( 0.20) × 10-10, k15 ) (7.1 ( 1.5) × 10-12, and k16
) (5.6 ( 0.4) × 10-13 cm3 molecule-1 s-1. We estimate that
potential systematic errors associated with uncertainties in the
reference rate constants could add an additional 10% to the
uncertainty range. Propagating this additional 10% uncertainty
gives k12 ) (1.6 ( 0.3) × 10-10, k15 ) (7.1 ( 1.7) × 10-12
,
and k16 ) (5.6 ( 0.7) × 10-13 cm3 molecule-1 s-1. Our result
for k16 is in good agreement with the value of k16 ) (5.1 ( 1.0)
× 10-13 cm3 molecule-1 s-1reported previously.11 There are
no literature data for k12 or k15 to compare with our results.
3.4. Mechanistic Study of 1,3-Dioxolane Oxidation at
Wuppertal and Ford. The atmospheric oxidation of 1,3-
dioxolane proceeds via OH radical attack which gives two
different alkyl radicals.
Figure 3. Decay of 1,3-dioxolane versus C2H4 and C2H6 (top) and
ethylene carbonate (EC, corrected for wall loss) versus C2H5Cl and
CH3OCHO (bottom) when exposed to Cl atoms at 300 K.
18-20, reaction 15 was studied relative to (21) and (23), and
reaction 16 was compared to (17) and (22).
Cl + CH4 f products
Cl + C2H6 f products
Cl + C3H8 f products
Cl + C2H4 f products
Cl + CH3OCHO f products
Cl + CH2Cl2 f products
Cl + C2H5Cl f products
(17)
(18)
(19)
(20)
(21)
(22)
(23)
In 1 atm of air, these alkyl radicals add O2 to give peroxy
radicals which react with NO to produce alkoxy radicals.
Representative results obtained at Wuppertal and Ford using
1,3-dioxolane and ethylene carbonate as reactants and C2H6,
C2H4, CH3OCHO, and C2H5Cl as references are shown in Figure
3. The ethylene carbonate decay was corrected for wall loss as
described in section 2.2. Linear least-squares analyses of the
data shown in Figure 3 and analogous plots give the rate constant
ratios shown in Table 1. Indicated errors are two standard
deviations of the linear regression.
The rate constant ratios in Table 1 can be placed upon an
14
absolute basis using k17 ) 1.0 × 10-13 14
,
,
k18 ) 5.7 × 10-11
,
,
k19 ) 1.4 × 10-10 14
,
k20 ) 9.4 × 10-11 14
k21 ) 1.4 × 10-12 15
The alkoxy radicals will either react with O2 to give cyclic
carbonyl compounds or undergo ring-opening bond scission
leading to the formation of linear multifunctional compounds.
To provide an accurate description of the atmospheric chemistry
of 1,3-dioxolane, we need information concerning the relative
importance of attack by OH radicals on the two different
k22 ) 3.3 × 10-13 14
,
and k23 ) 8.0 × 10-12
,
16 to give the data
in Table 2. Consistent results were obtained using the different
reference compounds. We choose to quote values for k12, k15,
and k16 which are averages of the results in Table 2 with error
limits which encompass the extremes of the ranges, hence k12

Atmospheric Chemistry of 1,3-Dioxolane
J. Phys. Chem. A, Vol. 103, No. 30, 1999 5963
Figure 5. Formation of methylene glycol diformate MGDF (2) and
ethylene carbonate (EC), measured (b) and corrected for wall loss (O),
versus loss of 1,3-dioxolane for an experiment using 1.34 ppm of 1,3-
dioxolane, 4.6 ppm of methyl nitrite, 1.9 ppm of NO, and 206 mbar of
O2 in 1000 mbar total pressure of N2 diluent at 297 K. The methylene
glycol diformate data are shifted vertically by 75 ppb for clarity.
reactions, the concentration of methylene glycol diformate
increased linearly with the loss of 1,3-dioxolane over the range
of 1,3-dioxolane consumptions used (2-52%). The correspond-
ing plot for ethylene carbonate shows curvature (Figure 5) due
to deposition to the reactor walls. The curvature is minor for
short irradiation times and was not observed in the chlorine-
initiated experiments which employed shorter irradiation times.
Correction17 of the observed ethylene carbonate concentrations
for wall loss results in a linear increase of the ethylene carbonate
production with the loss of 1,3-dioxolane. The yields of ethylene
carbonate and methylene glycol diformate were independent of
O2 partial pressure over the range studied (60-800 mbar).
The observed molar yields of ethylene carbonate and meth-
ylene glycol diformate were 0.48 ( 0.07 and 0.50 ( 0.14 for
OH radical initiation and 0.43 ( 0.07 and 0.53 ( 0.07 for Cl
atom initiation. It should be noted at this point that Cl atom
initiated oxidation of 1,3-dioxolane produces HO2 radicals which
react with NO to form OH radicals. Thus, in the Cl-initiated
experiments there may be a contribution from OH radicals.
Experiments involving 1,3-dioxolane have been performed
previously in the EUPHORE outdoor smog chamber facility.18
Mixtures of 2 ppm of 1,3-dioxolane and 100-200 ppb of NOx
in air were prepared in the 168 m3 Teflon EUPHORE chamber
and exposed to sunlight. Molar yields of (0.42 ( 0.10) for both
ethylene carbonate and methylene glycol diformate were
obtained from the EUPHORE experiments.18 The results from
the present study are consistent with the outdoor smog chamber
studies.
Figure 4. IR spectra acquired before (A) and after (B) a 20 s irradiation
of a mixture of 18.7 mTorr of 1,3-dioxolane, 16.3 mTorr of NO, 96
mTorr of Cl2, and 249 Torr of O2 in 700 Torr total pressure of N2
diluent. During the irradiation, 23% of the 1,3-dioxolane was consumed.
Subtraction of features attributable to 1,3-dioxolane, NO, NO2, HNO3,
and HONO from panel B gives panel C. Panels D and E show reference
spectra of methylene glycol diformate and ethylene carbonate, respec-
tively. Subtraction of the infrared features attributable to methylene
glycol diformate and ethylene carbonate from panel C gives panel F.
-CH2- groups and the fates of the corresponding alkoxy
radicals. To provide such data, experiments were performed
using the FTIR-smog chamber systems at Wuppertal and Ford.
Experiments employed the UV irradiation of 1,3-dioxolane/
CH3ONO/NO and 1,3-dioxolane/Cl2/NO mixtures in 1 atm of
O2/N2 diluent at 298 ( 4 K. Figure 4 shows typical IR spectra
acquired before (A) and after (B) a 20 s irradiation of a mixture
of 18.7 mTorr of 1,3-dioxolane, 16.3 mTorr of NO, 96 mTorr
of Cl2, and 249 Torr of O2 in 700 Torr total pressure of N2
diluent. During the irradiation, 23% of the 1,3-dioxolane was
consumed. Subtraction of features attributable to 1,3-dioxolane,
NO, NO2, HNO3, and HONO from panel B gives panel C.
Panels D and E show reference spectra of methylene glycol
diformate and ethylene carbonate, respectively, which are
subtracted from the product spectrum in panel C. As shown in
panel F, there are no products other than methylene glycol
diformate and ethylene carbonate formed.
The observation of ethylene carbonate and methylene glycol
diformate in yields which are independent of [O2] over the range
60-800 mbar and account for 100% of the loss of 1,3-dioxolane
serves to define the atmospheric oxidation mechanism. The
absence of any effect of [O2] shows that there is only one fate
of each of the alkoxy radicals. Alkoxy radicals produced in
reaction 25 react with O2, while those formed in reaction 27
decompose.
For the Cl-initiated reactions, the concentration of methylene
glycol diformate and ethylene carbonate increased linearly with
the loss of 1,3-dioxolane over the range of 1,3-dioxolane
consumptions used (6-70%). After each experiment, the
mixture was left to stand in the dark for about 10 min to measure
the loss rate of ethylene carbonate in the dark and hence
calculate appropriate corrections.17 Such corrections were in the
range 2-5%. No changes in the ethylene carbonate and
methylene glycol diformate yields were observed when the
initial NO and Cl2 concentrations were varied independently
by factors of 5 and 2, respectively. For the OH-initiated
It is interesting to compare the behavior of the alkoxy radical
derived from 1,3,5-trioxane with the analogous alkoxy radical

5964 J. Phys. Chem. A, Vol. 103, No. 30, 1999
Sauer et al.
formed from 1,3-dioxolane. As shown below, the structures of
the two alkoxy radicals are similar.
There was no discernible loss (<2%) of the peroxynitrates on
standing in the dark in the chamber for 10 min. However,
addition of NO resulted in a complete (>95%) loss of the
peroxynitrates within the 2 min taken for the addition and the
formation of approximately equal amounts of ethylene carbonate
and methylene glycol diformate. The peroxy nitrates derived
from 1,3-dioxolane are thermally unstable and decompose to
give the peroxy radicals and NO2. In the presence of NO2,
peroxy nitrate is reformed via reactions 33 and 34. Addition of
NO serves to scavenge the peroxy radicals and prevents the
reformation of the peroxy nitrate. The thermal stability of the
peroxy nitrates derived from 1,3-dioxolane is low and, like other
alkyl peroxy nitrates,6 will not play any role in atmospheric
chemistry.
3.6. Chemical Modeling at Wuppertal. An explicit reaction
mechanism for the OH-initiated degradation of 1,3-dioxolane
in the presence of NOx was developed using a simple box
model.24 The reaction scheme was checked by fitting the model
to the experimental concentration-time profiles for 1,3-diox-
olane, methylene glycol diformate, ethylene carbonate, NO,
NO2, and methyl nitrite, simultaneously using 10 sets of
experimental data with [O2] ) 60-800 mbar. The mechanism
was based upon the RACM mechanism,25 which is used widely
to model the chemistry occurring in urban air. The RACM
mechanism was modified by inclusion of the reactions describ-
ing CH3ONO photolysis and 1,3-dioxolane degradation given
However, the atmospheric fates of these radicals are very
different. There are two possible fates of these radicals:
decomposition or reaction with O2. Under atmospheric condi-
tions, the alkoxy radical derived from 1,3,5-trioxane undergoes
essentially 100% ring opening decomposition via C-O bond
scission. In contrast, the corresponding alkoxy radicals derived
from 1,3-dioxolane react with O2 to form ethylene carbonate.
It is interesting that the relative importance of decomposition
and reaction with O2 changes so dramatically for these species.
The cause of this change in behavior probably lies in the
energetics associated with the decomposition reaction. A
computational study of the energetics of these systems would
be interesting but is beyond the scope of the present study.
The behavior of the other alkoxy radical derived from 1,3-
dioxolane can be compared to the corresponding alkoxy radical
formed in the oxidation of 1,4-dioxane.
The structures of these two alkoxy radicals are similar, and their
atmospheric fate is the same, decomposition via C-C bond
scission.19,20 The preference of these two radicals for decom-
position via C-C bond scission rather than C-O bond scission
or reaction with O2 can be rationalized by their relatively low
C-C bond strengths (decomposition results in the loss of a C-C
bond and the formation of a C-O bond). This behavior is
entirely consistent with the available database concerning the
behavior of similar alkoxy radicals. Thus, decomposition via
C-C bond scission has been shown to be the major fate of
alkoxy radicals R to oxygen atoms in the ether linkages in
diethyl ether,21 2-ethoxy ethanol,22 and 2-butoxy ethanol.23 The
in Table 3. The NO2 photolysis rate (JNO ) 3.76 × 10-3 s-1
)
2
was measured in the Wuppertal chamber. The CH3ONO
photolysis rate cannot be obtained directly from the observed
CH3ONO decay since this compound is reformed by recombi-
nation of CH3O radicals with NO. However, photolysis of CH3-
ONO was the only OH source in the experiments and it was
possible to obtain a value of JCH ONO ) (0.95 ( 0.17) × 10-3
3
s-1 from a fit of the simulated VOC profiles to the experimental
data. All other photolysis frequencies were evaluated relative
•
alkyl radical formed in reaction 29, HC(O)OCH2OCH2 , reacts
to JNO using the algorithm of Madronich.26
further to give methylene glycol diformate, HC(O)OCH2OCHO.
2
The degradation scheme developed for 1,3-dioxolane is
illustrated in Figure 6. The rate constants for the reactions of
OH radicals with methylene glycol diformate and ethylene
carbonate have not been measured and were estimated. Ethylene
glycol diformate and dimethyl carbonate have molecular
structures similar to those of methylene glycol diformate and
ethylene carbonate and react with OH radicals with rate
constants of 4.7 × 10-13 20 and 3.2 × 10-13 cm3 molecule-1
The observation of stoichiometric conversion of HC(O)OCH2-
OCH2 radicals into methylene glycol diformate in the present
s-1 27
, respectively. On the basis of these data, it seems
•
work is consistent with the recent work in our laboratories
showing that the atmospheric oxidation of 1,3,5-trioxane
produces essentially 100% yield of methylene glycol difor-
mate.11
reasonable to assume that the reactions of OH radicals with
methylene glycol diformate and ethylene carbonate will proceed
with rate constants of the order of 5 × 10-13 cm3 molecule-1
s-1. While such slow reactions will not play an important role
in the present system, they were included in the mechanism for
completeness. The observed wall loss for ethylene carbonate
was also included in the model.
In the modeling exercise, the branching ratio of the initial
reaction of OH with 1,3-dioxolane k1a/k1b and the nitrate from
the reactions of peroxy radicals (RO2, R′O2, and HC(O)-
(OCH2)2O2; see Table 3) with NO were varied to provide the
best fit of the experimental data. The methylene glycol diformate
and ethylene carbonate profiles were sensitive to the choice of
k1a/k1b, while the NO and NO2 profiles were sensitive to the
3.5. Cl Atom Initiated Oxidation of 1,3-Dioxolane in the
Presence of NO2 at Wuppertal. The UV irradiation of 1,3-
dioxolane/Cl2/NO/O2/N2 mixtures resulted in consumption of
NO and formation of ethylene carbonate, methylene glycol
diformate, and NO2. When all of the NO had been consumed,
additional product features at 792, 1300, 1726, and 1755 cm-1
were observed. The same IR product features were observed
following irradiation of 1,3-dioxolane/Cl2/NO2/O2 mixtures.
These features increased linearly with the consumption of 1,3-
dioxolane and are ascribed to the formation of peroxynitrates.

Atmospheric Chemistry of 1,3-Dioxolane
J. Phys. Chem. A, Vol. 103, No. 30, 1999 5965
Figure 6. Degradation scheme for 1,3-dioxolane in the presence of NOx.
TABLE 3: Chemical Mechanism Used in Simulations
reaction
k (298 K)a)
ref
(a) Methyl Nitrite Chemistry
CH3ONO + hν f CH3O + NO
CH3O + O2 f HCHO + HO2
HO2 + NO f OH + NO2
OH + CH3ONO f products
OH + HCHO f HCO + H2O
HCO + O2 f HO2 + CO
CH3O + NO f CH3ONO
CH3O + NO f HCHO + HNO
CH3O + NO2 f CH3ONO2
9.5 × 10-4
1.9 × 10-15
8.1 × 10-12
2.6 × 10-13
1.0 × 10-11
5.5 × 10-12
3.6 × 10-11
4.0 × 10-12
2.0 × 10-11
this work
14
14
31
14
14
14
14
14
(b) Degradation Scheme for 1,3-Dioxolaneb
OH + 1,3-dioxolane f R + H2O
R + O2 f RO2
5.4 × 10-12
5.0 × 10-12
1.0 × 10-11
2.5 × 10-13
4.2 × 105
this work
estimated
estimated
this work
estimated
estimated
estimated
this work
estimated
this work
estimated
estimated
this work
estimated
this work
estimated
estimated
RO2 + NO f RO + NO2
RO2 + NO f RONO2
RO f HC(O)(OCH2)2
HC(O)(OCH2)2 + O2 f HC(O)(OCH2)2O2
HC(O)(OCH2)2O2 + NO f HC(O)(OCH2)2O + NO2
HC(O)(OCH2)2O2 + NO f HC(O)(OCH2)2ONO2
HC(O)(OCH2)2O + O2 f MGDF + HO2
OH + 1,3-dioxolane f R′ + H2O
R′ + O2 f R′O2
5.0 × 10-12
1.0 × 10-11
2.5 × 10-13
8.0 × 10-15
4.9 × 10-12
5.0 × 10-12
1.0 × 10-11
2.5 × 10-13
8.0 × 10-15
1.9 × 10-4
5.0 × 10-13
5.0 × 10-13
R′O2 + NO f R′O + NO2
R′O2 + NO f R′ONO2
R′O + O2 f EC + HO2
EC f wall loss
OH + EC f products
OH + MGDF f products
a In units of cm3 molecule-1 s-1 (second order) or s-1 (first order). b MGDF ) methylene glycol diformate; EC ) ethylene carbonate.
nitrate yield. Nitrate yields in reactions of peroxy radicals with
NO are generally determined by the size of the peroxy radical.
Since the peroxy radicals formed during the oxidation of 1,3-
dioxolane are all of a similar size, their nitrate yields were
constrained to be equal. It was found that the best fits were
achieved using k1a/k1b ) 47.5:52.5 and with a nitrate formation
yield of 2.5% for reaction of NO with each of the peroxy radicals
(RO2, R′O2, and HC(O)(OCH2)2O2; see Table 3). Figure 7 shows
an example of the quality of the fits achieved. The ethylene
carbonate concentration is underestimated slightly by the model
for reaction times less than 14 min. This may be caused by the
fact that the rate constant of the observed wall loss for ethylene

5966 J. Phys. Chem. A, Vol. 103, No. 30, 1999
Sauer et al.
radicals will react with NO within a few minutes leading to the
formation of methylene glycol diformate and ethylene carbonate.
Both species are relatively unreactive toward OH radicals and
will not participate in further gas-phase reactions in urban
airsheds.
Acknowledgment. We thank Bill Stockwell (Desert Re-
search Institute) for sharing his computer code and for helpful
discussions and Tobias Maurer (Bergische Universita¨t GH
Wuppertal) for a critical reading of the manuscript. Work at
Wuppertal was funded by the EU and the German Bundesmin-
isterium fu¨r Bildung, Wissenschaft, Forschung und Technologie
(BMBF), Project “Fo¨rderschwerpunkt Tropospha¨renforschung
(TFS)”, Contract 07TFS30. T. J. Wallington acknowledges
financial support by the Alexander von Humboldt Stiftung.
References and Notes
(1) Roubi, M. A. Chem. Eng. 1995, 44, 37.
(2) Hansen, K. B.; Wilbrandt, R.; Pagsberg, P. ReV. Sci. Instrum. 1979,
50, 1532.
(3) Barnes, I.; Becker, K. H.; Mihalopoulos, N. J. Atmos. Chem. 1994,
18, 267.
(4) Bierbach, A. Ph.D. Thesis, University of Wuppertal, Wuppertal,
F.R.G., 1994.
(5) Wallington, T. J.; Japar, S. M. J. Atmos. Chem. 1989, 9, 399.
(6) Atkinson, R. Chem. ReV. 1986, 86, 69.
(7) Horspool, W. M., Song, P-S., Eds. CRC Handbook of Organic
Photochemistry and Photobiology; CRC Press: New York, 1995; p 337.
(8) Field, N. D. J. Am Chem. Soc. 1961, 83, 3504.
(9) Posner, G. H.; Nelson, T. D. Tetrahedron 1990, 46, 4573.
(10) Shaap, A. P. Tetrahedron Lett. 1971, 21, 1757.
(11) Platz, J.; Christensen, L. K.; Sehested, J.; Nielsen, O. J.; Wallington,
T. J.; Sauer, C.; Barnes, I.; Becker, K. H.; Vogt, R. J. Phys. Chem. A 1998,
102, 4829.
(12) Nielsen, O. J.; Sidebottom, H. W.; Nelson, L.; Treacy, J. J.;
O’Farrel, D. J. Int. J. Chem. Kinet. 1989, 21, 1101.
(13) Atkinson, R. J. Phys. Chem. Ref. Data 1989, Monograph 1.
Figure 7. Comparison of experimental (symbols) and simulated
(curves) concentration-time profiles. Top panel: 1,3-dioxolane ([),
CH3ONO (2), NO (9), and NO2 (b). Bottom panel: ethylene carbonate
(9) and methylene glycol diformate (b).
(14) DeMore, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina,
M. J. Jet Propulsion Laboratory Publication 97-4; Pasadena, CA, 1997.
(15) Wallington, T. J.; Hurley, M. D.; Ball, J. C.; Jenkin, M. E. Chem.
Phys. Lett. 1993, 211, 41.
(16) Wine, P. H.; Semmes, D. H. J. Phys. Chem. 1983, 87, 3572.
(17) Atkinson, A.; Aschmann, S. M.; Carter, W. P. L.; Winer, A. M.;
Pitts, J. N., Jr. J. Phys. Chem. 1982, 86, 4563.
(18) Sauer, C. G.; Barnes, I.; Donner, B. In The European Photoreactor,
EUPHORE Report 1997; Barnes, I., Wenger, J., Eds.; University of
Wuppertal: Wuppertal, F.R.G., 1998; p 32.
(19) Platz, J.; Sehested, J.; Møgelberg, T. E.; Nielsen, O. J.; Wallington,
T. J. J. Chem. Soc., Faraday Trans. 1997, 93, 2855.
(20) Maurer, T.; Hass, H.; Barnes, I.; Becker, K. H. J. Phys. Chem. A,
submitted.
(21) Wallington, T. J.; Japar, S. M. EnViron. Sci. Technol. 1991, 25,
410.
(22) Stemmler, K.; Mengon, W.; Kerr, J. A. EnViron. Sci. Technol. 1996,
30, 3385.
(23) Stemmler, K.; Mengon, W.; Kinnison, D. J.; Kerr, J. A. EnViron.
Sci. Technol. 1997, 31, 1496.
carbonate is weakly dependent on the reaction time. For the
present computer simulations, an averaged value was used. For
reaction times longer than 12 min, the modeled concentration
of methylene glycol diformate is slightly higher than the
experimental values. Nevertheless, the fact that within the
experimental uncertainties the mechanism described above is
able to account for the temporal changes of NO, NO2, CH3-
ONO, 1,3-dioxolane, ethylene carbonate, and methylene glycol
diformate in the chamber demonstrates that we understand the
chemical degradation mechanism of 1,3-dioxolane.
4. Conclusions
A substantial body of kinetic and mechanistic data pertaining
to the atmospheric chemistry of 1,3-dioxolane is presented. It
is expected that the atmospheric lifetime of 1,3-dioxolane is
determined by reaction with OH radicals. While the OH radical
concentration in the atmosphere varies with location, time of
day, season, and meteorological conditions, a reasonable 24 h
(24) Seefeld, S.; Stockwell, W. R. Atmos. EnViron., in press.
(25) Stockwell, W. R.; Kirchner, F.; Kuhn, M.; Seefeld, S. J. Geophys.
Res. 1997, 102, 25847.
(26) Madronich, S. J. Geophys. Res. 1987, 92, 9740.
(27) Bilde, M.; Møgelberg, T. E.; Sehested, J.; Nielsen, O. J.; Wallington,
T. J.; Hurley, M. D.; Japar, S. M.; Dill, M.; Orkin, V. L.; Buckley, T. J.;
Huie, R. E.; Kurylo, M. J. J. Phys. Chem. A 1997, 101, 3514.
(28) Ravishankara, A. R.; Lovejoy, E. R. J. Chem. Soc., Faraday Trans.
1994, 90, 2159.
(29) Dorn, H.-P.; Brandenburger, U.; Brauers, T.; Ehhalt, D. H. Geophys.
Res. Lett. 1996, 23, 2537.
(30) Hofzumahaus, A.; Aschmutat, U.; Hessling, M.; Holland, F.; Ehhalt,
global average is (0.5-1.0) × 106 molecules cm-3 28-30
.
At 295
K, the rate constant for reaction of OH radicals with 1,3-
dioxolane is 8.8 × 10-12 cm3 molecule-1 s-1; hence, the
atmospheric lifetime of 1,3-dioxolane will be 30-60 h. Reaction
with OH gives two different alkyl radicals in approximately
equal yield, which then add O2 rapidly (within 1 µs) to give
the peroxy radicals. In polluted urban air masses, the peroxy
D. H. Geophys. Res. Lett. 1996, 23, 2541.
(31) Nielsen, O. J.; Sidebottom, H. W.; Donlon, M.; Treacy, J. Int. J.
Chem. Kinet. 1991, 23, 1095.