Kinetic, mechanistic, and modeling study of the OH-radical-initiated oxidation of di-n-butoxymethane (DNBM)

Article abstract of DOI:10.1021/jp001901x

Di-n-butoxymethane (DNBM) is a potential diesel fuel component to improve the atmospheric properties of the fuel exhaust. The widespread use of organic compounds such as DNMB will result in their release to the atmosphere in significant quantities, thus the potential arises for possible adverse effects on tropospheric chemistry. Product formation in the OH-radical-initiated oxidation of DNBM in the presence of NO_x was studied in an indoor photoreactor and in the outdoor simulation chamber EUPHORE. The major reaction products obtained in the air were n-butoxymethyl formate (NBMF) and propionaldehyde with a molar yield of ～ 80 mole % for each compound. Smaller amounts of di-n-butyl carbonate (DNBC), n-butyl formate, and n-butoxymethyl butyrate were also formed. The short tropospheric lifetimes of NBMF and DNBC indicated that reaction with OH is the dominant sink for both compounds and that transport to remote areas plays a minor role in their regional and global atmospheric chemistry. A photochemical mechanism was developed to describe the OH-initiated degradation of DNBM in the presence of NO_x. Experimentally obtained and modeled concentration - time profiles for selected reactants showed good agreement.

Full text of DOI:10.1021/jp001901x

J. Phys. Chem. A 2000, 104, 11087-11094
11087
Kinetic, Mechanistic, and Modeling Study of the OH-Radical-Initiated Oxidation of
Di-n-butoxymethane (DNBM)
T. Maurer, H. Geiger, I. Barnes,* and K. H. Becker
FB 9-Physikalische Chemie, Bergische UniVersita¨t GH Wuppertal, Gaussstrasse 20,
D-42097 Wuppertal, Germany
L. P. Thu1ner†
Centro de Estudios Ambientales del Mediterraneo, Parque Tecnolo´gico, Calle Charles R. Darwin 14,
46980 Paterna, Valencia, Spain
ReceiVed: May 23, 2000; In Final Form: September 14, 2000
FTIR product studies of the OH-radical-initiated oxidation of di-n-butoxymethane (DNBM) in the presence
of NOx were performed in an indoor photoreactor and in the outdoor simulation chamber EUPHORE in
Valencia, Spain. The reaction products observed in both reaction chambers were n-butoxymethyl formate
(NBMF), propionaldeyhde, and di-n-butyl carbonate (DNBC). In the indoor reactor yields for NBMF and
propionaldeyhde of 77 ( 15 and 78 ( 16 mol % were obtained in the system DNBM/MeONO/NOx/air/hν
and 88 ( 18 and 69 ( 14 mol % in the system DNBM/H2O2/NOx/air/hν, respectively. In the outdoor chamber,
yields of 80 ( 8 and 44 ( 11 mol % were obtained after sunlight irradiation of a DNBM/NOx/HONO/air
mixture. For di-n-butyl carbonate (DNBC), an upper limit of e10 mol % was estimated for both reaction
chambers. In the indoor photoreactor small amounts of n-butyl formate (NBF) and n-butoxymethyl butyrate
(NBMB) were also detected with upper limits of 3 mol % for each compound. Bimolecular rate coefficients
for the reactions of NBMF and DNBC with OH radicals were determined in the indoor photoreactor using
the relative rate technique. Values of kOH+NBMF ) (8.00 ( 0.91) × 10-12 cm3 s-1 and kOH+DNBC ) (7.07 (
1.64) × 10-12 cm3 s-1 were obtained. NBMF was synthesized and authentic samples were used for calibration.
A photochemical mechanism was developed to describe the OH-initiated degradation of di-n-butoxymethane
(DNBM) in the presence of NOx. The reaction scheme was tested by comparison of computer box model
calculations and experimental data. Experimentally obtained and modeled concentration-time profiles for
selected reactants are in excellent agreement.
1. Introduction
effects on tropospheric chemistry. It is well-known, for example,
that the reformulation of fuel by oxygenated VOCs (volatile
Ethers are an important class of automotive fuel additives.1
These compounds enhance the octane level, increase the
efficiency of combustion, and reduce the emission of atmo-
spheric pollutants, e.g., hydrocarbons, CO, and particles.2-5 The
possible industrial application of a large number of new
oxygenated compounds such as acetals or esters is currently
under discussion.6 For example, both classes of compound are
being considered as alternative solvents.
The compound under investigation here, di-n-butoxymethane
(DNBM), is currently under consideration as a potential diesel
fuel component to improve the atmospheric properties of the
fuel exhaust. It has been shown that the addition of formalde-
hyde acetals with the general structure ROCH2OR to diesel fuel
significantly reduces the emission of particulate matter from
diesel engines,5 with DNBM being particularly effective in this
respect.
organic compounds) increases the exhaust concentration of
aldehydes7 and the discovery of methyl tert-butyl ether (MTBE)
in groundwater and reservoirs used for drinking water8 has
resulted in the prohibition of MTBE as a fuel additive in
California from the beginning of the year 2003.9 These examples
show the necessity to improve the knowledge about the kinetic,
mechanistic, and environmental behavior of oxygenated VOCs
and their reaction products in the troposphere prior to their
widespread application.
In continuation of previous work on the OH radical kinetics
of a series of acetals,10 the present study is focused on the
product formation in the OH-initiated oxidation of di-n-
butoxymethane (DNBM) in the presence of NOx and also the
further reactions of the products. Time-resolved product con-
centrations were obtained from experiments in both an indoor
photoreactor and an outdoor simulation chamber.
However, the widespread use of these organic compounds
will result in their release to the atmosphere in significant
quantities and thus the potential arises for possible adverse
Mechanistic studies on tropospheric VOC oxidation are often
limited to the detection of the primary reaction products.
Reaction mechanisms, which are derived from these experi-
mental data, have not usually been subject to validation by
comparison of the measured product concentrations against
model calculations. In other instances, where no suitable
experimental results are available, reaction mechanisms are
* To whom correspondence should be addressed. E-mail: barnes@
physchem.uni-wuppertal.de. Phone: +49-202-439-2510. Fax: +49-202-
439-2505.
† Present address: Bergische Universita¨t GH Wuppertal.
10.1021/jp001901x CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/04/2000

11088 J. Phys. Chem. A, Vol. 104, No. 47, 2000
Maurer et al.
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 using the results obtained from the photoreactor
experiments. The reaction scheme was developed by comparison
of simulated and experimental concentration-time profiles.
Altogether, the present work provides a detailed kinetic,
product, and modeling study on the OH-initiated oxidation of
di-n-butoxymethane in the presence of NOx.
of selected reactants and products were recorded using in situ
FTIR-spectroscopy with a Nicolet Magna 550 spectrometer
using a resolution of 1 cm-1 and a path length of 553.5 m. Each
spectrum consisted of 550 interferograms co-added over a time
period of 10 min. Typical initial concentrations were 350 ppb
DNBM, 60 ppb HONO, 100 ppb NO, and 70 ppb NO2. The
mixtures were irradiated for approximately 3 h.
2.3. Relative Rate Technique. The rate coefficients for the
reaction of OH with n-butoxymethyl formate (NBMF) and di-
n-butyl carbonate (DNBC) were determined in the 1080 L
photoreactor using the relative rate technique.13 Since this
method has already been described numerous times in the
literature, it is only briefly outlined here. Provided that the
reactant and the reference compound are removed solely by the
reaction of interest and are not reformed in any process, which
was the case for NBMF
2. Experimental and Chemical Modeling System
The different experimental systems used for the present
investigations are described only briefly here. Details can be
found in the literature.11,12
2.1. Indoor Photoreactor. The majority of the experiments
were carried out in a 1080 L quartz-glass indoor photoreactor
equipped with a built-in White mirror system. OH radicals were
generated either by the UV photolysis of hydrogen peroxide
(H2O2) using 32 low-pressure mercury lamps (Philips TUV 40,
λmax ) 254 nm) or by irradiation of methyl nitrite (CH3ONO)/
NO mixtures using 16-32 fluorescent lamps (Philips Tl 40W/
05, 320 < λ < 450 nm). Concentration-time profiles of
reactants and products were recorded by long-path in situ FTIR
spectroscopy. The FTIR spectrometer (Bruker IFS-88) was
operated using a resolution of 1 cm-1 and a path length of 484.7
m. Most experiments were carried out at 1000 ( 20 mbar total
pressure in synthetic air at 298 ( 2 K. However, in order to
obtain mechanistic information some experiments were also
performed in 1000 ( 20 mbar of N2. During the irradiations
10 infrared spectra (128 interferograms co-added over ca. 2 min)
were collected. Typical initial concentrations employed for the
product studies were 0.7-1.0 ppm DNBM, 2-4 ppm CH3ONO,
10-20 ppm H2O2, and 0.5-7 ppm NO (1 ppm ) 2.46 × 1013
cm-3 at 1000 mbar and 298 K).
OH + reactant f products
OH + reference f products
(R1)
(R2)
it can be shown that
[reactant]t)0
ln
k
[reactant]t)0
[reactant]t
)
1 ln
k2
(1)
[reactant]t
where [reactant]t)0 and [reactant]t, and [reference]t)0 and
[reference]t are the reactant and reference concentrations at times
t ) 0 and t, respectively, and k1 and k2 are the rate coefficients
for reactions R1 and R2.
A plot of ln([reactant]t)0/[reactant]t) versus ln([reference]t)0
/
[reference]t) yields a slope of k1/k2 with zero intercept. When
there is an additional wall loss process for the reactant, which
was the case for DNBC
reactant wall8 products
(R3)
Since the infrared spectra of DNBM and its oxidation
products, n-butoxymethyl formate (NBMF) and di-n-butyl
carbonate (DNBC), are not available in the literature, calibrated
IR spectra for these compounds have been measured and
integrated band intensities (IBI) determined for the main
absorption bands.
the data can be analyzed using the following expression
[reactant]
[reactant]t
k
[reactant]t)0
[reactant]t
ln
t)0 - k3t ) 1 ln
(2)
k2
The IBIs for DNBM, NBMF, and DNBC were measured in
the 1080 l quartz-glass room temperature photoreactor. In order
to allow reproducible and quantitative transfer of the compounds
into the gas phase, weighed amounts of the compound were
dissolved in HPLC grade CH3Cl and known volumes of the
solution were subsequently injected into the reactor. The use
of a solution in contrast to the pure compounds minimized
interference by wall loss processes due to the low vapor pressure
of this compound. The measurements were performed in 1000
mbar of N2 at 298 K with spectrometer conditions as given
above.
2.2. Outdoor Smog Chamber EUPHORE. Additional
experiments on the OH-initiated oxidation of DNBM were
carried out in the outdoor simulation chamber EUPHORE in
Valencia, Spain. The chamber consists of a half-spherical 200
m3 Teflon bag with a transmission of more than 80% in the
range 280-640 nm. Two high-capacity ventilators ensure
homogeneous mixing in the chamber. Cooling of the chamber
floor compensates for the heating of the chamber by solar
radiation. The photooxidation was initiated in the chamber by
sunlight irradiation of a mixture of DNBM, nitrous acid (HONO)
and NOx in 1000 mbar of dry air. Concentration-time profiles
where k3 is the rate coefficient for reaction R3.
The kinetic experiments were performed in 1000 mbar air at
298 K using the photolysis of H2O2 as the OH radical precursor
and 2-propanol (kOH+isopropanol ) 5.7 × 10-12 cm3 molecule-1
s-1 14
s-1 14
)
)
and n-butane (kOH+n-butane ) 2.17 × 10-12 cm3 molecule-1
as reference compounds. Typical initial concentrations
for the kinetic experiments were 0.3-0.4 ppm DNBC, 0.3-
0.4 ppm NBMF, 4 ppm n-butane, 1.0-1.5 ppm 2-propanol, and
25 ppm H2O2.
2.4. Chemicals and Gases. Di-n-butoxymethane (DNBM)
(Lambiotte & Cie S.A, >99%), di-n-butyl carbonate (DNBC)
(Lancaster, >98%), 2-propanol (Aldrich, >99.5%), and H2O2
(Peroxid Chemie, 85%) were used without further purification.
All gases were supplied by Messer Griesheim and had the
following stated purities: synthetic air (99.995%), N2 (99.999%),
NO (99.8%), NO2 (98%), and n-butane (99.5%). The experi-
ments in the EUPHORE reactor were performed in dried and
purified air.12
Methyl nitrite was synthesized as described in the literature15
by adding sulfuric acid dropwise to an ice-cooled mixture of
methanol and sodium nitrite and collection in a cold trap at
-78 °C.

OH-Radical-Initiated Oxidation of DNBM
J. Phys. Chem. A, Vol. 104, No. 47, 2000 11089
HONO was synthesized by dropwise addition of 15 mL of a
1% NaNO2 solution to 30 mL sulfuric acid (30%) at about 20
°C. The HONO formed in this reaction was flushed in an air
stream into the dark EUPHORE chamber.
n-Butoxymethyl formate (NBMF) was synthesized in analogy
to the synthesis of methoxymethyl formate described by Pihlaja
et al.16 and Weeks et al.17 Briefly, 2 mol of finely ground sodium
formate was placed in a flask, heated to 30 °C, and mixed with
0.5 mol n-butoxymethyl chloride. Mixing was stopped when
the temperature in the system started to decrease. The product
was extracted with ether. After drying by manganese sulfate,
the solution was distilled under reduced pressure yielding pure
NBMF. The purity of NBMF was determined by GC and NMR
to be >98%.
2.5. Computer Simulation System. All computer simulations
were carried out using the box model SBOX by Seefeld and
Stockwell.18,19 This FORTRAN program incorporating the Gear
algorithm20 was operated on an SGI Origin 200 workstation
running under IRIX 6.5. The program uses the public domain
library VODE21 to integrate the ordinary differential equations.
3. Results and Discussion
3.1. Rate Coefficients for the Reaction of n-Butoxymethyl
Formate (NBMF) and Di-n-butyl Carbonate (DNBC) with
OH Radicals. Over the time period of all the experiments on
NBMF, both adsorption on the wall and photolysis were
negligible, whereas for DNBC an additional loss reaction due
to wall adsorption was observed. This loss was well character-
ized by kloss ) 1.5 × 104 s-1. Figure 1 shows examples of the
data from the kinetic experiments plotted according to eq 1 and
eq 2 for the reaction of OH with NBMF and DNBC, respec-
tively. The rate coefficient ratios obtained from the slopes of
such plots are given in Table 1 for the experiments on NBMF
and DNBC with the two reference compounds 2-propanol and
n-butane. Also given are the rate coefficients the reaction of
OH with NBMF and DNBC derived from these ratios using
the literature values for the analogue reactions with 2-propanol
and n-butane. We choose to quote values for the rate coef-
ficients, which are an average of all the experiments; hence
Figure 1. Plots of the kinetic data for the reaction of OH with
n-butoxymethyl formate (NBMF) versus n-butane (O) and 2-propanol
(b) (top) and di-n-butyl carbonate (DNBC) versus n-butane (0) and
2-propanol (9) (bottom). The solid lines are the results of linear least-
squares fits to the corresponding data.
kOH + NBMF ) (8.00 ( 0.91) × 10-12 cm3 s-1
kOH + DNBC ) (7.07 ( 1.64) × 10-12 cm3 s-1
TABLE 1: Kinetic Data Obtained from the Investigations
on the Reactions of OH with n-Butoxymethyl Formate
(NBMF) and Di-n-butyl Carbonate (DNBC) at 298 ( 2 K
The quoted errors reflect the systematic errors due to the
uncertainties in the rate coefficients of the reference compounds.
They were derived from the standard deviations from averaging
the values determined using the different reference compounds.
Both the measured rate coefficients are about a factor of 4 lower
than that for the reaction of OH with their atmospheric precursor
kreactant × 1012
reactant
NBMF
reference
kreactant/kreference
cm3 molecule-1 s-1
isopropanol
butane
isopropanol
butane
1.35 ( 0.06
3.84 ( 0.49
1.14 ( 0.05
3.53 ( 0.42
7.68 ( 0.34
8.32 ( 0.49
6.49 ( 0.30
7.65 ( 0.93
DNBC
DNBM. Taking an average OH concentration of 5 × 106 cm-3
,
typical for a summer day in a polluted region, the tropospheric
lifetimes for NBMF and DNBC with respect to reaction with
OH are 7.9 and 6.9 h, respectively. These fairly short lifetimes
indicate that NBMF and DNBC will mainly be removed from
the troposphere by reaction with OH. Transport to remote areas
will be of minor importance. However, both compounds are
highly water-soluble. Therefore, it can be assumed that, under
appropriate conditions, wet deposition might also be an impor-
tant tropospheric sink for these compounds.
spectra which include that of DNBM (panel A) and NBMF
(panel C) recorded with a resolution of 1 cm-1 in the region
from 700 to 2000 cm-1. The spectrum of DNBM contains a
main absorption in the region of the C-O stretching vibration
between 1000 and 1200 cm-1. The spectrum of NBMF contains
two main absorption bands, one in the region of the CdO
stretching vibration between 1700 and 1790 cm-1 and one in
the region of the C-O stretching vibration between 1000 and
1200 cm-1. For the main bands of the DNBM and NBMF
spectra and the two bands of the DNBC spectrum (not shown
in Figure 2) the following integrated band intensities (IBI) have
been calculated between the stated limits (the errors reflect the
3.2. Infrared Integrated Band Intensities (IBI) of Di-n-
butoxymethane (DNBM), n-Butoxymethyl Formate (NBMF),
and Di-n-butyl Carbonate (DNBC). Figure 2 shows infrared

11090 J. Phys. Chem. A, Vol. 104, No. 47, 2000
Maurer et al.
Figure 2. Infrared spectra recorded for a MeONO/NOx/DNBM/air
photolysis system in the wavenumber region 2000-700 cm-1. Spectrum
A was recorded before irradiation, B is a difference spectrum derived
from a spectrum recorded after 10 min of irradiation minus spectrum
A, C is a reference spectrum of n-butoxymethyl formate (NBMF), and
D is the residual spectrum obtained after 10 min irradiation and
subtraction of DNBM and all identified products multiplied by a factor
of 5.
Figure 3. Formation of NBMF (b), propionaldehyde (9), and DNBC
(2) versus loss of DNBM for an example experiment. Initial condi-
tions: 0.68 ppm DNBM, 15 ppm H2O2, and 6.0 ppm NO in air. The
solid lines are the results of linear least-squares fits to the corresponding
data.
statistical 2σ precision only):
DNBM: IBI (1100-1170 cm-1) ) (1.35 ( 0.04) × 10-17
TABLE 2: Molar Yields of the Major Products Obtained
from Experiments on the OH-Radical-Initiated Oxidation of
cm molecule-1
a
Di-n-butoxymethane (DNBM) in the Presence of NOx
NBMF: IBI (1711-1786 cm-1) ) (1.49 ( 0.24) × 10-17
yield (mol %)
cm molecule-1
experimental system
NBMF
propionaldehyde
HCHO
indoor photoreactor
air/H2O2/NOx/hν
air/CH3ONO/NOx/hν 77 ( 15
N2/H2O2/NOx/hν
outdoor smog chamber
air/HONO/hν
IBI (1060-1168 cm-1) ) (2.10 ( 0.23) × 10-17
88 ( 18
69 ( 14
78 ( 16
27 ( 5
25 ( 5
n. p.b
50 ( 10
cm molecule-1
76 ( 15
DNBC: IBI (1200-1354 cm-1) ) (4.35 ( 0.47) × 10-17
80 ( 8
44 ( 11
6 ( 3
cm molecule-1
a Minor products observed: di-n-butyl carbonate (DNBC, e10 mol
%), n-butyl formate (NBF, e3 mol %), and n-butoxymethyl butyrate
(NBMB, e3 mol %). b Determination not possible because of form-
aldehyde formation from CH3ONO.
IBI (11725-1800 cm-1) ) (1.24 ( 0.16) × 10-17
cm molecule-1
shown in Figure 3. The molar yields of all the products identi-
fied in the different experimental systems are summarized in
Table 2. An upper limit of e10% for the yield of DNBC was
estimated from the experimental data. For the yields of NBF
and NBMB, an upper limit of e3% has been obtained for each
compound.
The yields of n-butoxymethyl formate, propionaldehyde, and
formaldehyde were corrected for secondary reactions with OH
radicals using an expression given by Atkinson et al.22 This
equation defines a correction factor F to be applied to the
measured concentration in order to obtain the “true” concentra-
tion:
3.3. Products of the OH-Initiated Oxidation of Di-n-
butoxymethane (DNBM) in the Presence of NOx. Figure 2
gives an example of the spectral information obtained from an
experiment on a CH3ONO/NOx/DNBM system carried out in
the indoor photoreactor. Spectrum A was recorded before
irradiation. Spectrum B was obtained by subtraction of spectrum
A from a spectrum recorded 10 min after irradiation. Spectrum
C is a reference spectrum of n-butoxymethyl formate. Spectrum
D is the residual product spectrum obtained after subtraction
of all the identified compounds. This spectrum has been
multiplied by a factor of 5 in order to highlight the residual
absorptions.
In all systems investigated, n-butoxymethyl formate and
propionaldeyhde were identified as the major products of the
OH-initiated oxidation of DNBM in the presence of NOx. In
the indoor photoreactor runs using H2O2 as the OH radical
precursor, formation of formaldehyde in high yield was
observed. Di-n-butyl carbonate was identified as a minor product
in all of the reaction systems. Additionally, formation of n-butyl
formate and n-butoxymethyl butyrate in low yield was observed
in the indoor photoreactor runs. The molar yields for the
products in the different experimental systems have been
obtained from the slopes of plots of the product concentration
versus the amount of DNBM consumed. Examples of such plots
for an experiment carried out in the indoor photoreactor are
F )
[DNBM]t
1 -
kDNBM - kproduct
[DNBM]t)0
kproduct/kDNBM
‚
kDNBM
[DNBM]t
[DNBM]t
-
(
)
[DNBM]t)0
[DNBM]t)0
(3)
where kDNBM and kproduct are the rate coefficients for the reactions
of DNBM and the products (n-butoxymethyl formate, formal-
dehyde, or propionaldeyhde) with OH. [DNBM]t)0 and
[DNBM]t are the concentrations of DNBM at t ) 0 and t,

OH-Radical-Initiated Oxidation of DNBM
J. Phys. Chem. A, Vol. 104, No. 47, 2000 11091
Figure 4. Tropospheric degradation scheme for the OH-initiated oxidation of DNBM in the presence of NOx: OH attack at the R-CH2 group of
the butoxy group. The branching ratios given were determined by chemical modeling.
respectively. For the correction of the yields of n-butoxymethyl
For DNBM the OH radical attack will mainly occur at the
-CH2- entities of the butyl side groups adjacent to the O atoms
and also the central -CH2- group.
formate, propionaldeyhde, and formaldehyde, the following rate
coefficients were used: kOH+DNBM ) 3.21 × 10-11 cm3 s-1 23
,
kOH+NBMF ) 8.00 × 10-12 cm3 s-1, kOH+propionaldeyhde ) 1.96 ×
Figure 4 illustrates the reaction chain initiated by OH attack
at the -CH2- entity of the butyl group, leading to the major
products, n-butoxymethyl formate and propionaldehyde. H atom
abstraction from the -CH2- entity forms the corresponding
substituted alkyl radical. Addition of O2 followed by reaction
of the peroxy radical (I) with NO forms the alkoxy radical (II).
Further reaction by cleavage of the C-C bond forms n-
butoxymethyl formate and an n-propyl radical (III), which
further reacts to form propionaldehyde. C-O bond cleavage
forms butanal and an n-butoxymethoxy radical (VI). Re-
action of radical VI with O2 yields n-butyl formate. Because of
the low upper limit of 3 mol % estimated for n-butyl formate
and lack of experimental evidence for the formation of bu-
tanal, this reaction channel would appear to be of minor
importance.
10-11 cm3 s-1 14
,
and kOH+HCHO ) 9.6 × 10-12 cm3 s-1 14
.
The reaction of OH with all other products is slow. Accord-
ingly, its influence on the product yields obtained over the time
scale of the experiments is negligibly small and the correction
is not necessary.
In the residual product spectrum (see Figure 2, spectrum D),
distinctive spectral features remain with bands at 1650, 1280,
and 850 cm-1. These absorptions are characteristic for organic
nitrates, and lead to the conclusion that the OH-initiated
oxidation of DNBM results in the formation of organic nitrates.
Using the absorption cross section for octyl nitrate24 of 1.25 ×
1018 cm3 molecule-1 at 1228 cm-1, the overall yield for the
sum of all nitrates formed in the OH-initiated oxidation of
DNBM has been estimated to be e8-10%.
In all the systems investigated, n-butoxymethyl formate and
propionaldehyde have been identified as the major products of
the OH-initiated oxidation of DNBM in the presence of NOx.
In every experiment performed, the measured molar yield of
n-butoxymethyl formate is, within the experimental errors, very
similar. In the indoor photoreactor experiments, the yield of
propionaldehyde measured in the absence of molecular oxygen
is significantly lower than that observed in air. In contrast, the
yield of formaldehyde in air is a factor of 2 lower than in the
absence of O2. This behavior can be rationalized from a
consideration of the degradation mechanism for DNBM.
The OH-radical-initiated oxidation of DNBM proceeds via
H atom abstraction. It is well-known that the preferred positions
for OH attack on acetals of the general structure ROCH2OR
are the -CH2- groups in R-position to the oxygen atoms.25
OH radical attack at the central -CH2- group of DNBM
leads to di-n-butyl carbonate as illustrated in Figure 5. The
initially formed substituted alkyl radical adds molecular oxygen
yielding the peroxy radical (VII). Conversion of NO to NO2
leads to the alkoxy radical (VIII), which further reacts with O2
forming the stable carbonate.
As noted above, there is strong evidence for the formation
of organic nitrates in the oxidation of DNBM. Accordingly, the
reactions of the alkyl radicals I, IV, and VII (see Figures 4 and
5) with NO forming the corresponding nitrates seems to be
reasonable. Isomerization reactions of the oxy radicals II and
VIII appear to be negligible; this is supported by the nearly
100% carbon balance. The observation of only very weak
absorptions in the residual product spectra also supports that
isomerization products are probably not important.

11092 J. Phys. Chem. A, Vol. 104, No. 47, 2000
Maurer et al.
Figure 5. Tropospheric degradation scheme for the OH-initiated oxidation of DNBM in the presence of NOx: OH attack at the central CH2 group.
The branching ratios given were determined by chemical modeling.
In order to ensure that the formation of n-butoxymethyl
formate and propionaldehyde is initiated by H atom abstraction
from the -CH2- entity of the butyl group adjacent to the O
atom, experiments were performed in N2. In N2 the yield of
propionaldehyde decreased significantly compared to air, whereas
the yield of n-butoxymethyl formate remained constant and the
yield of formaldehyde increased. This behavior is consistent
with the degradation mechanism proposed in Figure 4. OH
attack at the -CH2- entity of the butyl group adjacent to the
O atom of DNBM is expected to lead to the formation of
n-butoxymethyl formate and propionaldehyde in equivalent
amounts when the oxygen partial pressure is sufficiently high.
If the oxygen concentration is significantly reduced, the lifetime
of radical V increases and C-C bond cleavage forming HCHO
and an ethyl radical becomes possible and consequently the yield
of propionaldehyde will decrease. In contrast, OH attack to the
â-CH2 position of the butoxy group of DNBM would directly
form propionaldehyde by C-C bond scission. In this case the
yield in propionaldehyde would be independent of oxygen partial
pressure.
mechanism was enhanced by the reactions describing the
photolysis of CH3ONO and the degradation of DNBM. The
chemistry of CH3ONO has been described in detail elsewhere.27
The NO2 photolysis rate (JNO ) 3.76 × 10-3 s-1 for 32
2
fluorescent lamps) was taken from previous measurements in
the photoreactor.28 The CH3ONO photolysis rate cannot be
obtained directly from the observed CH3ONO decay since this
compound is re-formed by recombination of CH3O radicals with
NO. However, photolysis of CH3ONO is the most predominant
OH source in the experiments and it was possible to obtain a
value of JCH ONO ) 8 × 10-4 s-1 from a fit of the simulated
3
VOC profiles to the experimental data. All other photolysis
frequencies were evaluated relative to JNO using the algorithm
2
of Madronich.29
The degradation scheme developed for DNBM is listed in
Table 3 and illustrated in the Figures 4 and 5. The rate
coefficient for the reaction of OH radicals with n-butoxymethyl
butyrate (NBMB) has not been measured and was estimated.
NBMB has a molecular structure similar to that of n-butoxy-
methyl formate (NBMF). It seems reasonable to assume that
both compounds will react with OH radicals with very similar
rate coefficients. Accordingly, for the reaction OH + NBMB,
a value of k ) 8.0 × 10-12 cm3s-1 was used in the present
model.
This observed behavior of the propionaldehyde yield in the
experiments with and without O2 thus confirms that n-butoxy-
methyl formate and propionaldehyde are formed via H atom
abstraction from the -CH2- entity of the butyl group adjacent
to the O atom.
In the modeling exercise the branching ratio of the initial
reaction of OH with DNBM kR4/kR16 and the nitrate from the
reactions of peroxy radicals (I and VII) (R7 and R19) with NO
were varied to provide the best fit of the experimental data.
The product profiles were sensitive to the choice of kR4/kR16
and the NO and NO2 profiles were sensitive to the nitrate yield.
Nitrate yields in reactions of peroxy radicals with NO are
generally determined by the size of the peroxy radical. It was
found that the best fits were achieved using kR4:kR16 ) 88:12
and a nitrate formation yield of 7.5% for reaction of NO with
each of the peroxy radicals I and VII (R7 and R19). The total
nitrate yield in the simulations is about 9.6 mol %, which agrees
well with the upper limit of 8-10 mol % estimated from the
experiments.
The yield of propionaldehyde obtained in the outdoor smog
chamber EUPHORE is significantly lower than in the indoor
photoreactor. This is caused by removal of propionaldehyde
from the system by photolysis. Under sunlight conditions, the
rate of photolysis is notably higher than in the indoor photo-
reactor experiments, since the radiation intensity of the lamps
used is much weaker. Accordingly, photolysis of propionalde-
hyde in the indoor chamber is negligible.
3.4. Chemical Modeling. An explicit reaction mech-
anism for the OH-initiated degradation of di-n-butoxy-
methane (DNBM) in the presence of NOx was developed using
a simple box model.19 The reaction scheme was developed by
fitting the model to the experimental concentration-time profiles
for DNBM, the primary reaction products NBMF, pro-
pionaldehyde, DNBC, NO, NO2, and methyl nitrite. The
simulations were focused on the indoor photoreactor experi-
ments using methyl nitrite photolysis as the precursor for OH
radicals.
For DNBM the major fraction of the OH radical attack occurs
at the -CH2- entity of the butyl group adjacent to the O atom
(R4). Addition of O2 to the alkyl radical formed and further
conversion of NO to NO2 by the corresponding peroxy radical
leads to the oxy radical (II). The branching ratio for the reactions
of II (R8-R10) was adjusted by fitting the model to the
experimental data. The major reaction channel is the thermal
decay forming NBMF and an n-propyl radical (R8). Further
The reactions, describing the inorganic chemistry of the
system, as well as the photolysis reactions, were taken from
the RACM mechanism of Stockwell et al.,26 which is widely
used to model the chemistry occurring in urban air. The present

OH-Radical-Initiated Oxidation of DNBM
J. Phys. Chem. A, Vol. 104, No. 47, 2000 11093
TABLE 3: Chemical Mechanism for the OH-Initiated Oxidation of Di-n-butoxymethane (DNBM) in the Presence of NOx Used
for the Computer Simulations in the Present Work
no.
reaction
k(298 K)a)
ref
OH attack to DNBM
in position 4
R4
R5
R6
R7
R8
R9
R10
R11
R12
R13
R14
OH + DNBM
f
f
f
f
f
f
f
f
f
f
f
f
R + H2O
2.83 × 10-11
1 × 10-11
1 × 10-11
7.5 × 10-13
6.7 × 105
8 × 10-15
8.0 × 10-12
4.8 × 10-12
2.1 × 10-13
8 × 10-15
8 × 10-15
23b
c
c
d
d
R + O2 (+ M)
RO2 (+ M)
RO2 + NO
RO + NO2
RONO2 (+ M)
NBMF + CH3CH2CH2(•) (+ M)
n-C4H9OCH2O(•) + n-C4H10 (+ M)
NBMB + HO2
n-C3H7O2 (+ M)
n-C3H7O + NO2
n-C3H7ONO2 (+ M)
C2H5CHO + HO2
NBF + HO2
RO2 + NO (+ M)
RO (+ M)
RO (+ M)
RO + O2
4.0 × 104
d
c
CH3CH2CH2(•) + O2 + (M)
n-C3H7O2 + NO
n-C3H7O2 + NO (+ M)
n-C3H7O + O2
30
30
30
c
R15
n-C4H9OCH2O(•) + O2
OH attack to DNBM
in position 6 (center)
R16
R17
R18
R19
R20
reactions of
OH + DNBM
R′ + O2 (+ M)
R′O2 + NO
R′O2 + NO (+ M)
R′O + O2
f
f
f
f
f
R′ + H2O
3.85 × 10-12
1 × 10-11
2.5 × 10-13
8 × 10-15
23b
c
c
d
c
R′O2 (+ M)
R′O + NO2
R′ONO2 (+ M)
DNBC + HO2
1 × 10-11
primary products
R21
R22
R23
R24
R25
R26
OH + NBMF
OH + DNBC
OH + NBMB
OH + NBF
f
f
products
products
products
products
products
CH3O + CO + HO2
8 × 10-12
e
e
f
7.1 × 10-12
8 × 10-12
f
f
f
3.54 × 10-12
2 × 10-11
OH + C2H5CHO
C2H5CHO + hν
1.5 × 10-06
26
a In cm3 s-1 (second order) and s-1 (first order). b Derived from total rate coefficient,23 ratio adjusted in this work to fit the chamber data.
c Estimated from similar reactions.14,30 d This work (estimated). e This work (measured). f Estimated from reaction R21.
Figure 7. Oxidation of DNBM: comparison of experimental and
simulated concentration-time profiles for methyl nitrite (9), NO ([),
and NO2 (b). The solid lines are the results of computer simulations.
For the initial conditions see Figure 6.
Figure 6. Oxidation of DNBM: comparison of experimental and
simulated concentration-time profiles for DNBM (2), NBMF ([),
propionaldehyde (b), and DNBC (9). The solid lines are the results
of computer simulations. Initial conditions: 0.85 ppm DNBM, 6.0 ppm
NO, 0.41 ppm NO2, 1.28 ppm CH3ONO, 760 Torr of air.
at the -OCH2O- group of DNBM (R16) which leads to
formation of di-n-butyl carbonate (DNBC) and small amounts
of the corresponding nitrate (see Figure 5).
In total, the chemical model used in the present work
consisted of 76 elementary reactions. Figures 6 and 7 demon-
strate the quality of the fits achieved. For all the reactants
illustrated, experimental and modeled concentration-time pro-
files are in excellent agreement.
reaction of this n-propyl radical yields propionaldehyde, the
second major product of the DNBM oxidation. Since upper
limits for the formation of n-butyl formate and n-butoxymethyl
butyrate (NBMB) were determined in the Experimental Section
(see above), their formation reactions (R9 and R10) from radical
II were also considered in the simulations. A small but
nevertheless significant fraction of the OH radical attack occurs

11094 J. Phys. Chem. A, Vol. 104, No. 47, 2000
Maurer et al.
(5) Dodge, L.; Naegeli, D. W. Final Report No. NREL/-TP-425-6345,
National Renewable Energy Laboratory, Golden, CO, 1994.
(6) Wallington, T. J.; Hurley, M. D.; Ball, J. C.; Straccia, A. M.; Platz,
J.; Christensen, L. K.; Sehested, J.; Nielsen, O. J.; J. Phys. Chem. A 1997,
101, 5302.
(7) Reuter, R. M.; Benson, J. D.; Burnes, V. R.; Gorse, R. A.;
Hochhauser, A. M.; Koehl, W. J.; Painter, L. J.; Rippon, B. H.; Rutherford,
J. A. SAE Technical Paper Series, No. 920326, Society of Automotive
Engineers, Warrendale, PA, 1992.
(8) Squillace, P. J.; Zogorski, J. S.; Wilber , W. G.; Price, C. V.
EnViron. Sci. Technol. 1996, 30, 1721.
(9) Morse, P. M. Chem. Eng. News 1999, 77, 26.
(10) Thu¨ner, L. P. Barnes, I., Maurer, T. Sauer, C. G.; Becker, K. H.
Int. J. Chem. Kinet. 1999, 31, 797.
(11) Barnes, I.; Becker, K. H.; Mihalopoulos, N. J. Atmos. Chem. 1994,
18, 267.
(12) Becker, K. H., Ed. The European Photoreactor EUPHORE, Final
Report, EV5V-CT92-0059 1996.
(13) Bierbach, A.; Barnes, I.; Becker, K. H.; Wiesen, E. EnViron. Sci.
Technol. 1994, 28, 715.
(14) Atkinson, R. J. Phys. Chem. Ref. Data 1994, Monograph No. 2.
(15) Taylor, W. D.; Allston, T. D.; Moscato, M. J.; Fazekas, G. B.;
Kozlowski, R.; Takacs, G. A. Int. J. Chem. Kinet. 1980, 12, 231.
(16) Pihlaja, K.; Lampi, A. Acta Chim. Scand. B 1986, 40, 196.
(17) Weeks, D. P.; Field, F. H. J. Am. Chem. Soc. 1970, 92, 1600.
(18) Seefeld, S. Ph.D. Thesis, Swiss Federal Institute of Technology
Zurich (ETH), Switzerland, 1997.
(19) Seefeld, S.; Stockwell, W. R. Atmos. EnViron. 1999, 33, 2941.
(20) Gear, C. W. Prentice Hall Series in Automatic Computation;
Prentice-Hall: Englewood Cliffs, NJ, 1971; Vol. 17.
(21) Brown, P. N.; Byrne, G. D.; Hindmarsh, A. C. J. Sci. Stat. Comput.
1989, 10, 1038.
4. Summary and Conclusions
The atmospheric oxidation of di-n-butoxymethane (DNBM)
was studied in the presence of NOx. The major reaction products
in air are n-butoxymethyl formate (NBMF) and propionaldeyhde
with a molar yield of ∼80 mol % for each compound. Further
products formed in smaller amounts are di-n-butyl carbonate
(DNBC), n-butyl formate (NBF) and n-butoxymethyl butyrate
(NBMB).
For the reactions of OH radicals with NBMF and DNBC,
bimolecular rate coefficients of kOH+NBMF ) (8.00 ( 0.91) ×
10-12 cm3 s-1 and kOH+DNBC ) (7.07 ( 1.64) × 10-12 cm3 s-1
,
respectively, were determined for the first time. From these data,
tropospheric lifetimes of several hours for each of both
compounds were calculated. These relatively short lifetimes
indicate that reaction with OH will be the dominant sink for
both NBMF and DNBC and that transport to remote areas will
play a minor role in their regional and global atmospheric
chemistry. Due to their water solubility, wet deposition of these
compounds will probably also be an additional removal path.
For the OH-initiated degradation of DNBM in the presence
of NOx, a chemical mechanism has been constructed. Experi-
mental and modeling results are in excellent agreement. After
further validation tests under realistic atmospheric conditions
in the presence of other VOCs, the developed reaction scheme
will be able to be used in applications such as field modeling.
(22) Atkinson, R.; Aschmann, S. A.; Carter, W. P. L.; Winer, A. M.;
Pitts, J. N., Jr. J. Phys. Chem. 1982, 86, 4563.
(23) Becker, K. H.; Dinis, C.; Geiger, H.; Wiesen, P. Chem. Phys. Lett.
1999, 300, 460.
(24) Wirtz, K. Ph.D. Thesis, University of Wuppertal, FRG, 1991.
(25) Becker, K. H.; Freitas Dinis, C. M.; Geiger, H.; Wiesen, P. Phys.
Chem. Chem. Phys. 1999, 1, 4721, and references therein.
(26) Stockwell, W. R.; Kirchner, F.; Kuhn, M.; Seefeld, S. J. Geophys.
Res. 1997, 102, 25847.
(27) Geiger, H.; Becker, K. H. Atmos. EnViron. 1999, 33, 2883.
(28) Sauer, C. G.; Barnes, I.; Becker, K. H.; Geiger, H.; Wallington, T.
J.; Christensen, L. K.; Platz, J.; Nielsen, O. J. J. Phys. Chem. A 1999, 103,
5959.
Acknowledgment. Financial support of the work by the
European Commission and the Bundesministerium fu¨r Bildung
und Forschung (BMBF) is gratefully acknowledged.The sample
of di-n-butoxymethane was supplied by Lambiotte & Cie S.A.,
Brussels. The authors thank W. R. Stockwell, Reno, NV, for
the supply of software and very helpful discussions. L.T.
acknowledges the cooperation of the CEAM employees at
EUPHORE and financial support by the Generalidad Valenci-
ana, Fundacio´n BANCAIXA and the Deutscher Akademischer
Austauschdienst (DAAD).
(29) Madronich, S. J. Geophys. Res. 1987, 92, 9740.
(30) DeMore, W. B.; Sander, S.; Howard, C. J.; Ravishankara, A. R.;
Golden, D. M.; Kolb, C. E.; Hanson, R. F.; Kurylo, M. J.; Molina, M. J.
Chemical Kinetics and Photochemical Data for Use in Stratospheric
Modeling, Evaluation No. 12, JPL Publication 97-4, Pasadena, CA, 1997.
(31) Le Calve´, S.; Le Bras, G.; Mellouki, A. J. Phys. Chem. A 1997,
101, 5489.
References and Notes
(1) Ha¨lsig, C. P.; Gregory, R.; Peacock, T. Spec. Publ.sR. Soc. Chem.
Chem. Oil Ind.: DeV. Appl. 1991, 97, 311.
(2) Dolislager, L. J. J. Air Waste Manage. Assoc. 1997, 47, 775.
(3) Guttmann H.; Schug, K. P. SAE Technical Paper Series, No.
900274, Society of Automotive Engineers, Warrendale, PA, 1990.
(4) CONCAWE Automotive Emissions Management Group, Report
No. 2/95, Brussels, Belgium, 1995.
(32) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, Jr., R. F.; Kerr,
J. A.; Rossi, M. J.; Troe, J. J. Phys. Chem. Ref. Data 1997, 26, 521.