FTIR kinetic, product, and modeling study of the OH-initiated oxidation of 1-butanol in air

Article abstract of DOI:10.1021/es010220s

A kinetic and product study was conducted on the reaction of OH radicals with 1-butanol in a 480 L indoor photoreactor and in the EUPHORE outdoor smog chamber in Valencia, Spain. Long path in situ FTIR spectroscopy and GC with photoionization detection were used to analyze reactants and products. The reaction products observed were butanal, propanal, ethanal, and formaldehyde. The results supported the probable formation of 4-hydroxy-2-butanone. Propanal, ethanal, and formaldehyde could also be formed in secondary reactions of some of the primary aldehydic products. A detailed atmospheric degradation mechanism was constructed and evaluated against experimental data by chemical box model calculations based on product studies. The difference between modeled and measured formaldehyde concentration was always < 20%. The OH-initiated degradation scheme of 1-butanol in the presence of NO_x developed in this study showed the experimental data generated in a large outdoor smog chamber well and could be used for the purposes of modeling oxidant formation in urban air masses, either in an explicit form or after suitable lumping, in chemical modules applied in chemistry transport models.

Full text of DOI:10.1021/es010220s

Environ. Sci. Technol. 2002, 36, 1263-1270
Alcohols are a class of highly volatile oxygenates, which
FTIR Kinetic, Product, and Modeling
Study of the OH-Initiated Oxidation
of 1-Butanol in Air
are currently finding increasing use as solvents because they
typically combine satisfactory technical performances in
chemical industry applications with low costs and low health
risks.
The compound investigated here, 1-butanol (propyl
carbinol, n-butyl alcohol), currently finds a variety of
applications in chemical industry (4). It is principally used
in the field of surface coating either directly as a solvent or
converted into derivatives, which then serve as solvents or
monomer components. 1-Butanol is an excellent thinner and
useful for regulating the viscosity and improving the flow
properties of varnishes; it is also applied in plastic and fiber
industries as well as in the production of butylamines. Besides
its anthropogenic sources, the compound is directly emitted
by plants (5).
F A B R I Z I A C A V A L L I , H A R A L D G E I G E R , *
I A N B A R N E S , A N D K A R L H E I N Z B E C K E R
Bergische Universit a¨ t-GH Wuppertal, Physikalische Chemie/
FB 9, D-42097 Wuppertal, Germany
A kinetic and product study was performed on the
reaction of OH radicals with 1-butanol in a 480 L indoor
photoreactor and also in the EUPHORE outdoor smog chamber
in Valencia, Spain. Long path in situ FTIR spectroscopy
and gas chromatography with photoionization detection
were used to analyze reactants and products. Using a kinetic
relative rate technique, a rate coefficient of k(OH +
While rate coefficients for the reactions of OH radicals
with many alcohols are well-established, the atmospheric
oxidation pathways of alcohols have received relatively little
attention, and their representation in atmospheric models
is, therefore, associated with substantial uncertainties.
-
12
3
-1
1
-butanol) ) (8.28 ( 0.85) × 10 cm s was measured
To improve the knowledge of the atmospheric chemistry
of alcohols and to facilitate their accurate description in
atmospheric models, a kinetic and product investigation of
the room-temperature gas-phase reaction of OH radicals with
1-butanol has been conducted. The kinetic studies were
performed in a large-volume indoor photoreactor, while the
product investigations were carried out in the outdoor
EUPHORE (6-8) smog chamber facility in Valencia, Spain.
To our knowledge, a detailed reaction mechanism for
in 740 Torr synthetic air at 298 ( 2 K. The reaction
products observed and their fractional molar yields were
(
in percent) butanal (51.8 ( 7.1), propanal (23.4 ( 3.5), ethanal
(12.7 ( 2.2), and formaldehyde (43.4 ( 2.4). In addition,
the results support the probable formation of 4-hydroxy-2-
butanone. Propanal, ethanal, and formaldehyde could
also be formed in secondary reactions of some of the primary
aldehydic products. However, under the conditions
employed in the experiments, the contribution from
secondary reactions is very minor. On the basis of the
product studies, a detailed atmospheric degradation
mechanism was constructed and tested against experimental
data by chemical box model calculations. Measured and
simulated concentration-time profiles for selected reactants
were in excellent agreement.
1
-butanol based on experimental measurements, such as
those developed in this paper, has not yet been reported in
the literature. Such a mechanism may now be applied in
chemistry transport models on local or regional scales using
scenarios suitable for estimating the influence of 1-butanol
on tropospheric photosmog formation if the industrial
emissions of this compound are significantly increased.
2. Experimental Section
The two experimental systems used for the present study
have been described in detail elsewhere (6-9) and are only
briefly outlined here.
1
. Introduction
Oxygenated organic compounds are currently under con-
sideration as potentially attractive replacements for tradi-
tional chlorinated and aromatic solvents. The drive behind
the search for replacements is the need for more environ-
mentally acceptable solvents with respect to ambient air
quality (i.e., compounds which will help to reduce the
2
.1. Kinetic Study of the Reaction of OH Radicals with
1
-Butanol. Kinetic experiments on the reaction of OH radicals
with 1-butanol were carried out at 298 ( 2 K and 740 Torr
total pressure of synthetic air in a 480 L cylindrical Duran
glass chamber. The chamber is surrounded by 24 fluorescent
lamps (Philips T 40W/ 05, 320 nm < λ < 450 nm) (9) and is
equipped with a built-in White mirror system for long path
3
tropospheric O and photo-oxidant burden). However, the
use of oxygenated replacements will also result in the release
of these compounds into the atmosphere, where they will
undergo photo-oxidative degradation processes, the most
important and often dominant being reaction with the OH
radical (1, 2). The photo-oxidation of oxygenates might
contribute to some extend to the formation of photochemical
air pollution in urban and regional areas (3). Therefore, before
large-scale usage of these compounds, a quantification of
their potential environmental impacts is desirable and, in
many countries, mandatory. Accurate kinetic data and
mechanistic information on the atmospheric fate of alterna-
tive oxygenated solvents are essential components in any
attempts to reliably assess the ozone forming potential of
these substances.
(
51.6 m) in situ monitoring of reactants.
The rate coefficient for the gas-phase reaction of OH
radicals with 1-butanol was determined using a relative rate
method, in which the decays of the reactant (1-butanol) and
of a suitable reference compound, whose rate coefficient
with OH is reliably known (10), are monitored in parallel in
the presence of OH radicals.
k1
OH + 1-butanol
OH + reference
9
8 products
(R1)
(R2)
k2
98 products
Provided that 1-butanol and the reference are removed solely
by reaction with OH radicals, then
*
Corresponding author phone: +49-202-439-3832; fax:+49-202-
4
39-2757; e-mail: geiger@uni-wuppertal.de.
1
0.1021/es010220s CCC: $22.00

2002 Am erican Chem ical Society
VOL. 36, NO. 6, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 6 3
Published on Web 02/08/2002

[
1-butanol]0
k1
k2
[reference]0
[reference]t
An FTIR spectrometer (NICOLET Magna 550, MCT
ln
)
ln
(1)
-1
detector; 1 cm resolution) coupled to a White mirror system
base length, 8.17 m; total optical path length, 554.5 m) was
used to monitor HCHO, HONO, and SF . The absorption
{
}
{
}
[
1-butanol]t
(
6
where [1-butanol]
0
and [reference]
0
are the concentrations
cross-sections for the compounds used for the evaluation of
the FTIR spectra were determined in separate calibration
studies.
of 1-butanol and the reference compound, respectively, at
time 0; [1-butanol] and [reference] are the corresponding
concentrations at time t; k and k are the rate coefficients
for reactions 1 and 2, respectively.
Hydroxyl radicals were generated by the photolysis of
methyl nitrite (CH ONO) in synthetic air using the fluorescent
lamps. NO was included in the reactant mixtures to suppress
the formation of O and, hence, of NO radicals. The kinetic
measurements were performed relative to cyclohexane. The
t
t
1
2
Because the FTIR technique was found to be unsuitable
for the quantitative detection of 1-butanol, butanal, propanal,
and ethanal, these compounds were analyzed by gas chro-
matography combined with photoionization detection (GC-
PID). Gas samples were automatically collected from the
chamber every 10 min and introduced via a heated sampling
loop of 2 mL capacity onto a DB624 column at 80 °C (J & W
Scientific; 30 m) mounted in a Fisons GC 8000 instrument.
The GC-PID response factors for the alcohol and the
products were determined by introducing known concen-
trations of these compounds into the chamber. This tech-
nique allowed for the determination of the higher aldehydes,
with an error of (15-20%.
3
3
3
-
3
approximate initial reactant concentrations (molecule cm
were as follows: CH
)
1
4
14
3
4
ONO, 4.1 × 10 ; NO, 4.1 × 10 ;
1
13
1
-butanol, 3.4 × 10 ; and cyclohexane, 9.3 × 10 . The
concentration-time behavior of 1-butanol and cyclohexane
were monitored over a 10-20 min irradiation period by in
situ FTIR long path spectroscopy. The FTIR spectrometer
(
Nicolet Magna 520) was operated with a resolution of 1 cm-1
.
In addition, carbonyl compounds were sampled from the
chamber using solid-phase 2,4-dinitrophenyl hydrazone
(DNPH)-silica cartridges. Hydrazones, formed by derivati-
zation, were separated and quantitatively measured by HPLC
(Hewlett-Packard model 1050; UV detector).
A few experiments using 1-butanol/ cyclohexane/ air mix-
tures (i.e., without OH precursor) were also carried out (i) in
the dark and (ii) at maximum light intensity over a period
of 60 min to check for complications associated with
adsorption on the wall or photolysis of these compounds; no
such complications were observed.
All chemicals used (1-butanol, cyclohexane, and NO) had
a stated manufacturer purity of >99%. Methyl nitrite was
prepared and stored as described in the literature (11).
2 x
NO and NO were measured by two NO analyzers
equipped with a molybdenum converter (Monitor Labs. Inc.,
ML 9841 A) and a photolytical converter (Eco Physics, CLD
770 AL with PLC 760), respectively.
To determine the dilution rate of the chamber, due to
1
1
-3
2
.2. Products of the Reaction of OH Radicals with
-Butanol. All product investigations were performed in the
small leaks in the FEP foil, 3.7 × 10 molecule cm of SF
6
1
were added, as an inert tracer, to the reaction mixture. In the
experiments presented here, the chamber leak rate was about
European Photoreactor (EUPHORE), a large-volume outdoor
smog chamber which is part of the Centro de Estudios
Ambientales del Mediterraneo (CEAM) in Valencia, Spain.
The facility consists of two identical half-spherical fluo-
rine-ethene-propene (FEP) foil chambers mounted on
aluminum floor panels covered with FEP foil. The FEP foil
transmits 85-90% of the light with wavelengths from >500
to 320 nm, dropping to 75% transmission at the atmospheric
threshold of 290 nm. To avoid unwanted heating of the
chambers, the floor panels are connected to a cooling system
allowing for realistic atmospheric temperature conditions
to be maintained, even during extended irradiation periods.
The chamber used for the experiments described here
had a volume of about 195 m and two mixing fans to ensure
homogeneity of the reaction mixtures. An air-drying and
purification system supplied the chamber with oil vapor,
hydrocarbon and NO
was measured in the chamber with a spectral radiantmeter
scanning double monochromator Bentham TM300). Fil-
-
1
3% h
.
2.3. Com puter Sim ulation System . All computer simula-
tions were carried out using the box model SBOX by Stockwell
et al. (12). This FORTRAN program incorporating the Gear
algorithm (13) was operated on an SGI Octane running under
IRIX 6.5. The program used the public domain library VODE
(14) to integrate the ordinary differential equations.
3. Results and Discussion
3.1. Rate Coefficient for the Reaction of OH with 1-Butanol.
Test experiments on 1-butanol/ cyclohexane/ air mixtures in
the 480 L indoor reactor showed that losses of the compounds
to the wall or via photolysis were <3% for both processes
over a 60 min period. Compared to the measured decays for
the compounds during the kinetic experiments, contributions
from wall loss and photolysis were negligible.
3
y
-free dry air. The solar light intensity
The kinetic investigations were performed by irradiation
(
of CH ONO/ NO/ 1-butanol/ cyclohexane/ air mixtures. Figure
3
terband radiometers (Meteorologie Consult, Glash u¨ tten) were
used to measure the photolysis frequencies of J(O D) and
1 illustrates the kinetic data obtained from three experiments
plotted according to eq 1. An excellent straight line plot is
obtained. Least-squares analysis of these data led to the rate
1
J(NO
2
) during the days of the experiments.
1
-Butanol was added to the chamber by gently heating
coefficient ratio k
put on an absolute basis using a value of k
1
/ k
2
) 1.10 ( 0.02 ((2σ), which was then
(OH + cyclo-
the required amount of the substance in a stream of air
entering the chamber. OH radicals were produced by the
photolysis of nitrous acid (HONO), also introduced into the
chamber in an air stream. HONO was prepared by dropwise
2
-
12
3
-1
hexane) ) 7.49 × 10 cm s at 298 K (2). This resulted in
a rate coefficient for the reaction of OH with 1-butanol of
-
12
3 -1
k
1
)(8.24 (0.84) ×10 cm s . The indicated error contains
addition of 15 mL of an aqueous 1% NaNO
2
solution to 30
an additional 10% uncertainty due to the potential systematic
errors associated with the value of the reference rate
coefficient.
mL of a 30% sulfuric acid solution at room temperature.
After allowing for the mixing of the reactants, photo-oxidation
was initiated by opening the light-tight protective housing
of the chamber and exposing the photoreactor contents to
sunlight. The initial concentrations of 1-butanol and HONO
The value of the rate coefficient measured in the present
study can be compared with previous room temperature
determinations reported in the literature (15-17). Using an
absolute technique, Wallington and Kurylo (15) obtained
1
3
12
-3
were approximately 1.5 × 10 and 1.8 × 10 molecule cm
respectively. The initial concentrations of NO and NO were
both approximately 1.8-2.3 × 10 molecule cm and their
final concentrations were 2.6 × 10 and 3.5 × 10 molecule
cm , respectively.
,
-
12
3
-1
2
k(OH + 1-butanol) ) (8.31 ( 0.63) × 10
cm s . Nelson
1
2
-3
12
et al. (16) have reported values of k(OH + 1-butanol) )
1
1
-12
-12
3
-1
(7.80 ( 0.20) × 10 and (8.56 ( 0.70) × 10 cm s , using
-
3
absolute and relative techniques, respectively. A unit-
1
2 6 4
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 6, 2002

FIGURE 1. Plot of eq 1 for the gas-phase reaction of OH radicals
with 1-butanol, using cyclohexane as reference compound.
weighted least-squares average of these three rate coefficients
leads to the recommendation of k(OH + 1-butanol) )
-
12
3
-1
8
.57 × 10
cm
s
given by Atkinson (2). Very recently,
Yujing and Mellouki (17) determined a value of (8.47 ( 0.34)
-
12
3 -1
×
10 cm s from an absolute rate technique experiment.
The value obtained here is in excellent agreement with all
previous determinations.
The Structure-Activity Relationship (SAR) method of
Kwok and Atkinson (18) predicts k(OH + 1-butanol) )
-
12
3
-1
6
.9 × 10 cm s at 298 K, which is in fair agreement with
the experimental findings. The discrepancy between calcu-
lated and measured rate coefficient presumably indicates
that the OH substituent group has a more significant effect
on the -CH
2
- groups than that assumed in the estimation
FIGURE 2. Plot of the aldehyde yields (O) uncorrected and (b)
corrected for the photolysis and reaction with OH radicals (see
text) versus the amount of 1-butanol consumed for (a) butanal, (b)
propanal, and (c) ethanal.
method. This has been confirmed in a very recent SAR study
by Bethel et al. (19), who calculated new substitution factors
for hydroxy compounds such as 1-butanol. The updated
-
12
cm3
s ,
-1
approach yields k(OH + 1-butanol) ) 7.8 × 10
which is fairly close to the experimental value. The rate
coefficient k can be combined with the estimated ambient
2
an NO/ NO ratio of about 1. Because the NO concentration
is rapidly decreasing after opening the chamber, the main
fate of the alkyl peroxy radicals, formed in the degradation
reactions of the higher aldehydes, will be formation of the
corresponding stable acyl peroxynitrate and not a lower
aldehyde.
To derive the formation yields of the aldehydic products,
their measured concentrations had to be corrected for further
reaction with OH radicals, photolysis, and dilution losses.
Corrections have been performed using the mathematical
procedure described by Tuazon et al. (21) and described in
a previous study from this laboratory (22).
1
6
tropospheric OH concentration (20) of 1.0 × 10 molecule
-
3
cm (24 h average) to obtain a tropospheric lifetime of 1.4
days for 1-butanol due to gas-phase reaction with OH radicals.
3
.2. Products of the OH Radical Reaction with 1-Butanol.
GC-PID and HPLC analysis of the products formed in the
irradiation of HONO/ 1-butanol/ air mixtures in the EUPHORE
photoreactor facility showed the formation of butanal,
propanal, ethanal, and formaldehyde among the products.
These products have been quantified by GC-PID and HPLC
response factors (except formaldehyde, which was measured
by FTIR) determined from calibrations with authentic
standards.
The OH radical concentration was calculated from the
decay of the alcohol using the rate coefficient for k
1
(OH +
1
-butanol) determined in this study, and the dilution rate
In addition, a product eluted at or very close to the
retention time of 4-hydroxy-2-butanone. Because the HPLC
retention time of this product closely matches that of an
authentic standard of 4-hydroxy-2-butanone, it is believed
that this product is 4-hydroxy-2-butanone. The HPLC
response factor, HPLC peak area/ concentration, used to
quantify this product was obtained using an authentic
standard.
Propanal, ethanal, and HCHO might also be formed in
secondary reactions of some of the primary aldehydic
products. However, under the conditions employed in the
experiments, the contribution from secondary reactions is
coefficients were determined for each experiment as previ-
ously described. Rate coefficients for reaction of OH radicals
with the aldehydic products were taken from the literature
(23, 24). The following values of kOH (in units of 10-11 cm
3
-1
s )
were used: butanal, 2.47 ( 0.15 (23); propanal, 2.02 ( 0.14
(23); and ethanal, 1.22 ( 0.27 (24).
Photolysis frequencies for the photolysis of the aldehydes
khν were taken from the literature (25). The mean of J(NO ),
2
measured during the experiments, was used as an indicator
of the light intensity in order to select the frequencies. The
-
5
-1
following values of khν (in units of 10
s ) were calculated
relative to J(NO ) and employed in the present study: butanal,
2
very minor. The initial NO
x
concentration is fairly high, with
1.26; propanal, 1.12; and ethanal, 0.23.
VOL. 36, NO. 6, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 6 5

The hydroxy alkyl radicals formed after H-atom abstrac-
tion from the -CH - group at the â and γ positions will add
to form hydroxy alkyl peroxy radicals, which can react
with NO or NO . Addition of NO will form thermally labile
hydroxy alkyl peroxy nitrates. In the presence of NO, the
hydroxy alkyl peroxy radicals form hydroxy alkyl nitrate and
TABLE 1. Product Formation Yields from the Gas-Phase
2
Reaction of the OH Radical with 1-Butanol in the Presence of
O
2
NO
x
2
2
product
formation yielda
butanal
propanal
ethanal
form aldehyde
0.518 ( 0.071
0.234 ( 0.035
0.127 ( 0.022
0.434 ( 0.024
0.05 ( 0.01
e0.04
2
hydroxy alkoxy radicals plus NO .
The contribution of the channel forming organic nitrates
•
b
in these RO
2
+ NO reactions is considered to be small (e4%).
b
This value was derived from the yield of butyl nitrates,
0.077 ( 0.009 (26), obtained in the irradiation of CH ONO/
4
-hydroxy-2-butanone
organic nitrates
carbon yield
3
e0.92 ( 0.16
NO/ n-butane/ air mixtures. The results of previous studies
(27-29) suggest that hydroxy butyl nitrate formation yields
from the reaction of hydroxy butyl peroxy radicals with NO
is a factor of ∼2 lower than the butyl nitrate formation yields
from the OH reaction with n-butane in the presence of NO.
The further expected reactions of the alkoxy radicals are
outlined in Scheme 1. In this reaction scheme, the alkyl and
alkyl peroxy radicals have been omitted for clarity, and only
the reactions of the alkoxy and acyl radicals involved are
shown (the given branching ratios refer to the box model
calculations described in the following paragraphs).
a
Corrected for photolysis and reaction with OH radicals as described
in the text. Indicated errors are two least-squares standard deviations
of the data shown in Figure 2 com bined with overall uncertainties in
the alcohol and carbonyl GC-PID calibration factors of (10% and (5%,
respectively. Additional uncertainties in the form ation yields due to
uncertainties in the m ultiplicative correction factor F (see text) are <(8%
for the form ation of butanal, <(10 for the form ation of propanal, and
<
(13% for the form ation of ethanal. b Uncorrected form ation yields,
derived in the early stage of the reaction; indicated error lim its are two
standard deviations.
The correction factors, [RHO]
t
corr/ [RHO]
t
were e1.5 for
The â-hydroxy alkoxy radical (I) formed after the H-atom
butanal, e1.3 for propanal, and e1.1 for ethanal. The
estimated overall uncertainties in the OH reaction rate
coefficient of 1-butanol and in the rate coefficients for
reaction with OH and photolysis of the products lead to
maximum uncertainties in the correction factors of (8% for
the formation of butanal, (10% for the formation of propanal,
and (13% for the formation of ethanal.
abstraction from the -CH
react with O to form 4-hydroxy-3-butanone (II) or decom-
pose to propanal and HCHO (via reaction of CH OH (III)
with O ). Isomerization by a 1,5-H shift via a six-membered
2
- group at the â position can
2
2
2
transition state cannot occur for this â-hydroxy alkoxy radical.
Furthermore, isomerizations by a 1,4-H shift via a five-
membered transition state are expected to be 3-4 orders of
magnitude slower at 296 K than 1,5-H shift isomerizations
30) because of the higher ring strain involved and are,
therefore, considered to be of negligible importance.
The corrected aldehydes concentrations are plotted
against the amount of 1-butanol consumed in Figure 2.
(
The corrected values for the formation yields are given in
Table 1. The molar formation yield of 0.434 ( 0.024 given for
HCHO was derived from the uncorrected concentrations
plotted against the consumption of 1-butanol in the early
stage of the reaction, when secondary processes can largely
be neglected. Similarly, the value of 0.05 ( 0.01 indicated for
Literature estimates (31) of the enthalpy for the decom-
-1
position (∆H
d
) 3.3 kJ mol ) and reaction with O
2
(∆H 2 )
O
-1
-
1423.7 kJ mol ) of this â-hydroxy alkoxy radical (I) can be
used to predict the relative importance of the two reaction
pathways using the empirical estimation method of Atkinson
and Carter (32). This approach assumes that a correlation
exists between the thermochemical behavior of these reac-
tions and the rate constants for alkoxy radical decomposition
4
-hydroxy-2-butanone represents its semiquantitative molar
formation yield obtained by plotting the uncorrected HPLC
concentrations of 4-hydroxy-2-butanone against the amounts
of 1-butanol consumed in the early stages of the reaction.
and reaction with O
whenever ∆(∆H) ) (∆H
reaction normally dominates for ∆(∆H) > 184 kJ mol
Thus, on the basis of the previously reported heats of reaction
for decomposition and O reaction and using the Atkinson
2
; in particular, decomposition dominates
-1
d
- ∆H 2) < 163 kJ mol , while the
O
-
1
In summary, the carbonyl and the hydroxycarbonyl
products observed and quantified in this work account for
O
2
.
9
2 ( 16% C of the overall OH radical reaction with 1-butanol.
.3. Mechanistic Considerations Based on SAR and the
Product Yield Data. The reaction of OH with 1-butanol
proceeds by H-atom abstraction from the various -CH
CH , and -OH groups. Using the SAR technique (18), it
can be estimated that the relative importance of OH attack
at the various -CH , -CH -, and -OH groups of 1-butanol
is the following: from the -OH group, 2.4%; from the -CH
group at the â position, 58.3%; from the -CH - group at the
â position, 20.4%; from the -CH - group at the γ position,
6.5%; and from the -CH group, 2.4%. Very recently, Bethel
2
estimation method (33), the decomposition of I is calculated
to dominate over the reaction with oxygen with a decom-
3
5
-1
position rate k
reaction rate k 2
d
of 1.1 × 10 s at 298 K compared to an O
2
3
,
4
-1
O
[O
2
] of 4.1 × 10 s
.
-
2
The identification among the degradation products of
propanal in a molar formation yield of 0.234 ( 0.035 clearly
supports that the decomposition channel is the main fate of
the â-hydroxy butoxy radical (I), which is calculated to
account for ∼20% of the overall reaction of 1-butanol with
OH radicals. The present results are in good agreement with
the experimental data of previous studies on the degradation
3
2
2
-
2
2
1
3
et al. (19) updated the SAR parameters. The new substitution
factors reported by these authors lead to probabilities for
the OH attack at the OH group and the R, â, γ, and δ positions
of 2.2%, 42.7%, 38.3%, 14.7%, and 2.2%, respectively. In
particular, the values for the â and γ attack are significantly
different to those predicted by the former set of parameters
developed by Kwok and Atkinson (18).
mechanism of 1-butene (28, 34) (note that CH
3
CH
2
CH-
•
(
1
O )CH
2
OH is also formed by OH radical addition to
-butene), which showed that the sole fate of I was
decomposition. The present results are also consistent with
the literature estimations (32) of the decomposition and O
reaction rate constants for I.
2
The possible reaction sequences for the γ-hydroxy alkoxy
radical (IV) formed after the H-atom abstraction from the
CH - group at the γ position are also shown in Scheme 1.
2
The R-hydroxy alkyl radical, formed by H-atom abstraction
from the -CH
2
- group at the 1-position, reacts with O
2
to
-
form butanal in unit yield (2)
The present GC-PID and HPLC product data show that
ethanal is formed with a yield of 12.7 ( 2.2%. Ethanal is
•
CH CH CH C( )HOH + O f CH CH CH CHO + HO
expected to be formed from the decomposition of (IV). The
3
2
2
2
3
2
2
2
•
(
R3)
2 2
CH CH OH radical (V), formed as “coproduct”, undergoes
1
2 6 6
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 6, 2002

x
SCHEME 1. Reaction Scheme for the OH-Initiated Oxidation of 1-Butanol in the Presence of NO Developed on the Basis of the
a
Present Smog Chamber Studies
a
The stated branching ratios were obtained by fitting the chem ical m odel to experim ental concentration-tim e profiles (see text).
addition of O
sponding alkoxy radical, which subsequently decomposes
82%) to form HCHO and reacts with O to form glycolal-
dehyde (VI) (18%) (35). While HCHO was identified by GC-
PID, HPLC and FTIR, glycoladehyde was not observed with
any of these analytical techniques, indicating that its yield
must be very low, as expected.
2
and reaction with NO to form the corre-
tration-time profiles for 1-butanol, butanal, propanal,
ethanal, HCHO, NO, and ozone.
(
2
The reactions describing the inorganic chemistry of the
system as well as the photolysis reactions for NO
2 3
, O , HONO,
HNO , HNO , NO , H , and formaldehyde were taken from
3
4
3
2 2
O
the Regional Atmospheric Chemical mechanism (RACM) of
Stockwell et al. (12), which is widely used to model the
The other decomposition channel involving C
sion would produce 3-hydroxy propanal (VII) and a CH
γ
-C
δ
scis-
chemistry occurring in urban air. The NO
2
photolysis
•
3
frequency was taken from the filterband radiometer mea-
surements in EUPHORE. All other photolysis frequencies,
except those for ethanal, propanal, and butanal, were
evaluated relative to JNO2 using the algorithm of Madronich
radical. However, this channel is considered to be of minor
importance because no clear evidence for 3-hydroxy propanal
formation could be obtained from the product analysis.
•
Furthermore, because of the greater stability of CH
2
CH
2
OH
(
38). For the evaluation, the radiation data were used as
•
(
V) compared to CH
3
(36, 37), the channel leading to ethanal
described by Stockwell et al. (12) for the RACM mechanism,
except those for the higher aldehydes which were taken from
an EUPHORE study performed by Moortgat (25).
•
and CH
2
CH
2
OH is predicted to be dominant.
On the basis of the experimental ratio k
as derived from the molar formation yields of the products
arising from the two reaction channels, the reaction of the
d
/ k 2
O
[O
2
] ≈ 2.6
The reactions used for the description of the OH-initiated
degradation of 1-butanol are summarized in Table 2.
Unknown rate coefficients in the mechanism were set to
values for similar reactions taken from established literature
data (e.g., the evaluations of DeMore et al. (39) and Atkinson
et al. (40)). The rate coefficients for the reactions of the four
γ-hydroxy butoxy radical (IV) with O
2
to form 4-hydroxy-
2
-butanone would appear to be much less important than
decomposition. A similar observation was made in a previous
study in this laboratory (22) on the γ-hydroxy pentoxy radical.
•
The δ-hydroxy alkoxy radical, OCH
2
CH
2
CH
2
CH
2
OH (VIII),
group,
-11
3
-1
different alkyl radicals with O
2
were set to 1 × 10 cm s
,
formed after the H-atom abstraction from the -CH
3
while for the alkylperoxy + NO reactions forming the
corresponding alkoxy radicals I, IV, and VIII, a k value of
was used. Here, an additional nitrate
formation fraction of 2% was considered for each reaction
is only of marginal importance in the atmospheric OH
initiated degradation mechanism of 1-butanol, accounting
for only an estimated 2.2% (SAR estimation (19)) of the overall
reaction. The possible products of the alkoxy radical reactions
could not be identified in the present study, presumably
because of their very low concentrations.
-12
cm3
-1
7
.5 × 10
s
(
see Section 3.3.). The rate coefficients for the alkoxy + O
2
-15
cm3
-1
reactions were all set to 8 × 10
s
. The first-order
decay constants for the other alkoxy reaction pathways were
calculated with respect to this value considering the branch-
ing ratios given in Scheme 1. For the further degradation of
3
.4. Chem ical Modeling. To confirm the experimental
findings on the OH-initiated degradation of 1-butanol,
computer simulations using a simple box model were
performed on the basis of the oxidation scheme described
previously and illustrated in Scheme 1. The branching ratios
for the OH attack to 1-butanol as well as the reactions of the
alkoxy radicals I and IV given in Scheme 1 and Table 2 were
obtained by fitting the model to the experimental concen-
3 2 2 2
CH , CH OH (III) and CH CH OH (V) radicals, well-
established literature data (39, 40) were incorporated into
the model. The OH-initiated oxidation of butanal, propanal,
and ethanal was also explicitly treated in the simulations.
Besides the rate coefficients for the OH reactions discussed
previously, mechanistic and rate data recommended by
VOL. 36, NO. 6, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 6 7

TABLE 2. Chemical Mechanism for the OH-Initiated Oxidation of 1-Butanol in the Presence of NO
x
Used for the Computer
a
Simulations in the Present Work
no.
reaction
k(298 K)b
ref
-
12
R4
R5
OH + 1-butanol
f
f
0.55 R-alkyl + 0.25 â-alkyl + 0.17 γ-alkyl + 0.03 δ-alkyl + H2O 8.28 × 10
c
d
-
11
R-alkyl + O2
â-alkyl + O2
1-butanal + HO2
â-Alkyl Chemistry
1 × 10
1 × 10
-
11
R6
R7
R8
R9
f
f
f
f
f
CH3CH2C(OO•)CH2OH
I + NO2
d
d
c
d
-
12
CH3CH2C(OO•)CH2OH + NO
I
I + O2
7.5 × 10
7.4 × 10
5
propanal + III
II + HO2
-
15
8 × 10
-
12
12
R10 III + O2
HCHO + HO2
9.1 × 10
39
γ-Alkyl Chemistry
CH3C(OO•)CH2CH2OH
IV + NO2
4-hydroxy-2-butanone + HO2
ethanal + V
-
11
R11 γ-alkyl + O2
R12 CH3C(OO•)CH2CH2OH + NO
R13 IV + O2
R14 IV
R15 V + O2
R16 CH2(OO•)CH2OH + NO
R17 CH2(O•)CH2OH + O2
R18 CH2(O•)CH2OH
R19 IV
R20 •CH3 + O2
R21 CH3 OO• + NO
R22 CH3O• + O2
f
f
f
f
f
f
f
f
f
f
f
f
1 × 10
d
d
d
c
d
42
d
35
c
39
39
39
-
7.5 × 10
-
15
5
8 × 10
3.3 × 10
-
-
-
11
CH2(OO•)CH2OH
CH2(O•)CH2OH + NO2
HOCH2CHO + HO2
HCHO + III
1 × 10
9 × 10
8 × 10
12
15
5
1.8 × 10
1.9 × 10
4
VII + •CH3
-
16
CH3 OO•
3 × 10
-
12
15
CH3O• + NO2
7.7 × 10
1.9 × 10
-
HCHO + HO2
δ-Alkyl Chemistry
CH2(OO•)CH2CH2CH2OH
VIII + NO2
-
11
R23 δ-alkyl + O2
f
f
f
1 × 10
d
d
d
-
12
11
R24 CH2(OO•)CH2CH2CH2OH + NO
R25 VIII + O2
7.5 × 10
-
15
HO(CH2)3CHO + HO2
8 × 10
OH + Aldehyde Chemistry
CO + HO2 + H2O
CH3C(OO•)O + H2O
•CH3 + NO2 + CO2
peroxy acetyl nitrate (PAN)
CH3CH2C(OO•)O + H2O
CH3CH2• + NO2 + CO2
CH3CH2OO•
CH3CH2O• + NO2
ethanal + HO2
-
R26 OH + HCHO (+ O2)
R27 OH + ethanal (+ O2)
R28 CH3C(OO•)O + NO
R29 CH3C(OO•)O + NO2
R30 OH + propanal (+O2)
R31 CH3CH2C(OO•)O + NO
R32 CH3CH2• + O2
R33 CH3CH2OO• + NO
R34 CH3CH2O• + O2
R35 CH3CH2C(OO•)O + NO2
R36 OH + 1-butanal (+ O2)
R37 CH3CH2CH2C(OO•)O + NO
R38 CH3CH2CH2O• + O2
R39 CH3CH2CH2C(OO•)O + NO2
f
f
f
f
f
f
f
f
f
f
f
f
f
f
1.0 × 10
39
39
40
44
40
40
39
39
39
44
23
40
d
-
11
1.21 × 10
-
11
11
11
11
12
2.0 × 10
-
1.0 × 10
-
2.0 × 10
-
2.0 × 10
-
8.0 × 10
-
-
12
8.85 × 10
9.95 × 10
15
-
12
peroxy propionyl nitrate (PPN)
CH3CH2CH2C(OO•)O + H2O
CH3CH2CH2O• + NO2 + CO2
propanal + HO2
6.4 × 10
-
11
2.47 × 10
-
11
2.0 × 10
-
15
8 × 10
-
12
peroxy butyryl nitrate (PBN)
6.4 × 10
44
a
The structures of I-VIII are illustrated in Schem e 1. b In cm 3 s-1 (second-order) and s-1 (first-order). c This work. d Estim ated from sim ilar
reactions (39, 40).
DeMore et al. (39) and Atkinson et al. (40) were used for this
purpose. The photolytic degradation schemes of these
compounds were also taken from the literature (25). In total,
the reaction scheme used for the simulations consisted of 88
elementary reactions.
It was decided to start the computer simulations 5 min
after opening the chamber, because the slow opening
procedure (over 3 min) was difficult to model and would
have led to large uncertainties of the simulated concentra-
tion-time profiles. Accordingly, the measured concentrations
for 1-butanol, HCHO, ethanal, propanal, butanal, NO
x
, and
O
3
at t )5 min served as initial parameters for the calculations.
Figures 3 and 4 illustrate the comparison of experimental
and modeled concentration-time profiles for selected
reactants for a typical run. Nearly all data are in excellent
agreement, except those for NO whose experimentally
obtained concentrations are slightly overestimated by the
model. The reason for this disparity remains unclear. At longer
reaction times, the calculated formaldehyde concentration
is also slightly larger than measured in the experiment. This
might be due to an unknown wall loss of HCHO or
uncertainties of its photolysis frequencies used in the
simulations. In particular, formaldehyde is often overesti-
mated in the modeling of smog chamber studies (41), a known
FIGURE 3. OH-Initiated oxidation of 1-butanol; comparison of
experimental and simulated concentration-time profiles for 1-bu-
3
tanol (b), NO ((), O (9), and HCHO (O). The solid lines are the
results of computer simulations (initial conditions (ppbV): 1-butanol,
591.4; NO, 56.0; NO , 87.0; O , 35.2; butanal, 6.1; propanal, 2.1; ethanal,
2
3
1.6; formaldehyde, 3.4; 760 Torr air; 294 K).
1
2 6 8
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 6, 2002

(
(
8) The European Photoreactor EUPHOREs2nd Report 1998-1999;
Barnes, I., Brockmann, K. J., Eds.; Wuppertal, Germany,
2
000 (http:/ / www.physchem.uni-wuppertal.de/ PC-WWW_Site/
Publications/ Publications.html).
9) Barnes, I.; Becker, K. H.; Zhu, T. J. Atmos. Chem. 1993, 17, 353-
3
73.
(
(
10) Bierbach, A.; Barnes, I.; Becker, K. H.; Wiesen, E. Environ. Sci.
Technol. 1994, 28, 715-729.
11) Taylor, W. D.; Allston, T. D.; Moscato, M. J.; Fezekas, G. B.;
Kozlowski, R.; Takacs, G. A. Int. J. Chem. Kinet. 1980, 12, 231-
2
40.
(
(
12) Stockwell, W. R.; Kirchner, F.; Kuhn, M.; Seefeld, S. J. Geophys.
Res. 1997, 102, 25847-26879.
13) Gear, C. W. Numerical Initial Value Problems in Ordinary
Differential Equations; Prentice Hall Series in Automatic Com-
putation; Prentice-Hall: New York, 1971; Vol. 17.
(
(
(
14) Brown, P. N.; Byrne, G. D.; Hindmarsh, A. C. J. Sci. Stat. Comput.
1
989, 10, 1038-1051.
15) Wallington, T. J.; Kurylo, M. J. Int. J. Chem. Kinet. 1987, 19,
015-1023.
1
16) Nelson, L.; Rattingan, O.; Neavyn, R.; Sidebottom, H. Int. J. Chem.
Kinet. 1990, 22, 1111-1126.
(
(
17) Yujing, M.; Mellouki, A. Chem. Phys. Lett. 2001, 333, 63-68.
18) Kwok, E. S. C.; Atkinson, R. Atmos. Environ. 1995, 29, 1685-
FIGURE 4. OH-Initiated oxidation of 1-butanol; comparison of
1
695 and references therein.
19) Bethel, H. L.; Atkinson, R.; Arey, J. Int. J. Chem. Kinet. 2001, 33,
10-316.
experimental and simulated concentration-time profiles for butanal
(
(
(9), propanal ((), and ethanal (b). The solid lines are the results
3
of computer simulations (initial conditions: see Figure 3).
20) Prinn, R. G.; Weiss, R. F.; Miller, B. R.; Huang, J.; Alyea, F. N.;
Cunnold, D. M.; Fraser, P. J.; Hartley, D. E.; Simmonds, P. G.
Science 1995, 269, 187-192.
21) Tuazon, E. C.; MacLeod, H.; Atkinson, R.; Carter, W. P. L. Environ.
Sci. Technol. 1986, 20, 383-387.
22) Cavalli, F.; Barnes, I.; Becker, K. H. Environ. Sci. Technol. 2000,
problem which could also not be solved in the present study.
However, in this work the disparity between modeled and
measured HCHO concentration was always less than 20%.
(
(
(
NO
the NO
2
has not been plotted, because there was interference of
measurements at longer reaction times by other
3
4, 4111-4116.
2
23) Papagni, C.; Arey, J.; Atkinson, R. Int. J. Chem. Kinet. 2000, 32,
nitrogen oxides. Accordingly, these experimental data could
not be used.
79-84.
(24) Semmes, D. H.; Ravishankara, A. R.; Gump-Perkins, C. A.; Wine,
P. H. Int. J. Chem. Kinet. 1985, 17, 303-313.
In summary, the OH-initiated degradation scheme of
(
25) Evaluation of Radical Sources in Atmospheric Chemistry through
Chamber and Laboratory Studies (RADICAL); Final report of the
EC Project, Contract no. ENV4-CT97-0419; Moortgat, G. K., Ed.;
Mainz, Germany, 2000.
(26) Atkinson, R.; Aschmann, S. M.; Carter, W. P. L.; Winer, A. M.;
Pitts, J. N., Jr. J. Phys. Chem. 1982, 86, 4563-4569.
1
x
-butanol in the presence of NO developed in the present
study describes the experimental data generated in a large
outdoor smog chamber well and can be used for the purposes
of modeling oxidant formation in urban air masses, either
in an explicit form or after suitable lumping, in chemical
modules applied in chemistry transport models.
(27) Atkinson, R. J. Phys. Chem. Ref. Data 1997, 26, 215-290.
28) Atkinson, R.; Tuazon, E. C.; Carter, W. P. L. Int. J. Chem. Kinet.
(
(
(
1
985, 17, 725-734.
Acknowledgments
29) Atkinson, R.; Tuazon, E. C.; Aschmann, S. M. Environ. Sci.
Technol. 1995, 29, 1674-1680.
30) Baldwin, A. C.; Barker, J. R.; Golden, D. M.; Hendry, D. G. J.
The authors thank the technical staff of CEAM for their
assistance with the measurements in the EUPHORE chamber
and the financial support by the “Generalitat Valenciana”.
F.C. gratefully acknowledges the European Commission for
funding her research at the Bergische Universit a¨ t Wuppertal
within the framework of the Environmental and Climate
Research and the Specific RTD Program. Support of the work
within the 4th Framework Program (EUROSOLV) is also
gratefully acknowledged.
Phys. Chem. 1977, 81, 2489-2492.
(
(
(
(
31) Atkinson, R. Atmos. Environ. 1990, 24, 1-41.
32) Atkinson, R.; Carter, W. P. L. J. Atmos. Chem. 1991, 13, 195-210.
33) Atkinson, R. Int. J. Chem. Kinet. 1997, 29, 99-111.
34) Kwok, E. S. C.; Atkinson, R.; Arey, J. Environ. Sci. Technol. 1996,
30, 1048-1052.
(35) Orlando, J. J.; Tyndall, G. S.; Blide, M.; Ferronato, C.; Wallington,
T. J.; Vereecken, L.; Peeters, J. J. Phys. Chem. A 1998, 102, 8116-
8
123.
(
(
36) McMillen, D. F.; Golden, D. M. Annu. Rev. Phys. Chem. 1982,
3, 493-532.
37) Baghal-Vayjooee, M. H.; Colussi, A. J.; Benson, S. W. Int. J. Chem.
Kinet. 1979, 11, 147-154.
3
Literature Cited
(
(
1) Atkinson, R. Atmos. Environ. 1990, 24, 1-41.
2) Atkinson, R. J. Phys. Chem. Ref. Data (Monograph No. 2) 1994,
(38) Madronich, S. J. Geophys. Res. 1987, 92, 9740-9752.
(39) 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-
1
-216.
3) Lamanna, M. S.; Goldstein, J. J. Geophys. Res. 1999, 104, 21247-
1262.
(
(
2
4) Ullman’s Encyclopedia of Industrial Chemistry; 5th Completely
Revised edition; VCH: Weinheim, Germany, 1993; A19, pp 49-
4
, Pasadena, CA, 1997.
(
40) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr,
6
0.
J. A.; Rossi, M. J.; Troe, J. J. Phys. Chem. Ref. Data 1997, 26,
(
(
5) Kesselmeier, J.; Staudt, M. J. Atmos. Chem. 1991, 13, 23-88.
5
21-1011.
6) The European Photoreactor EUPHORE; Final report of the
EC project, Contract no. EV5V-CT92-0059; Becker, K. H.,
Ed.; Wuppertal, Germany, 1996 (http:/ / www.physchem.
uni-wuppertal.de/ PC-WWW_Site/ Publications/ Publications.
html).
(41) Sampling Device for the Measurement of Peroxy Radicals in
Atmospheric Systems (SAMPLER); Final report of the EC project,
Contract no. ENV4-CT97-0389; Geiger, H., Ed.; Wuppertal,
Germany, 2000 (http:/ / www.physchem.uni-wuppertal.de/ PC-
WWW_Site/ pub/ Projektberichte/ SAMPLER. PDF).
(42) Becker, K. H.; Geiger, H.; Wiesen, P. Chem. Phys. Lett. 1991, 184,
256-261.
(
7) The European Photoreactor EUPHOREsReport 1997; Barnes, I.,
Wenger, J., Eds.; Wuppertal, Germany, 1998.
VOL. 36, NO. 6, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 6 9

(
(
43) Taylor, P. H.; Rahman, M. S.; Arif, M.; Dellinger, B.; Marshall,
Received for review September 3, 2001. Revised manuscript
received December 10, 2001. Accepted December 17, 2001.
P. Symp. (Int.) Combust., [Proc.] 1996, 26, 497-504.
44) Sehested, J.; Christensen, L. K.; Nielsen, O. J.; Bilde, M.;
Wallington, T. J.; Schneider, W. F.; Orlando, J. J.; Tyndall, G. S.
Int. J. Chem. Kinet. 1998, 30, 475-489.
ES010220S
1
2 7 0
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 6, 2002