Atmospheric chemistry of C3-Co cycloalkanecarbaldehydes

Article abstract of DOI:10.1021/jp044495g

The rate coefficients for the gas phase reaction of NO₃ and OH radicals with a series of cycloalkanecarbaldehydes have been measured in purified air at 298 ± 2 K and 760 ± 10 Torr by the relative rate method using a static reactor equipped with long-path Fourier transform infrared (FT-IR) detection. The values obtained for the OH radical reactions (in units of 10^-11 cm³ molecule^-1 s ^-1) were the following: cyclopropanecarbaldehyde, 2.13 ± 0. 05; cyclobutanecarbaldehyde, 2.66 ± 0.06; cyclopentanecarbaldehyde, 3.27 ± 0.07; cyclohexanecarbaldehyde, 3.75 ± 0.05. The values obtained for the NO₃ radical reactions (in units of 10^-14 cm ³ molecule^-1 s^-1) were the following: cyclopropanecarbaldehyde, 0.61 ± 0.04; cyclobutanecarbaldehyde, 1.99 ± 0.06; cyclopentanecarbaldehyde, 2.55 ± 0.10; cyclohexanecarbaldehyde, 3.19 ± 0.12. Furthermore, the reaction products with OH radicals have been investigated using long-path FT-IR spectroscopy and proton-transfer-reaction mass spectrometry (PTR-MS). The measured carbon balances were in the range 89-97%, and the identified products cover a wide spectrum of compounds including nitroperoxycarbonyl cycloalkanes, cycloketones, cycloalkyl nitrates, multifunctional compounds containing carbonyl, hydroxy, and nitrooxy functional groups, HCOOH, HCHO, CO, and CO₂.

Full text of DOI:10.1021/jp044495g

5104
J. Phys. Chem. A 2005, 109, 5104-5118
Atmospheric Chemistry of C3-C6 Cycloalkanecarbaldehydes
Barbara D’Anna,*, Armin Wisthaler,* Øyvind Andreasen, Armin Hansel, Jens Hjorth,
†,⊥
,‡,⊥
†
‡
§
§
†
|
§
Niels R. Jensen, Claus J. Nielsen, Yngve Stenstrøm, and Jyrki Viidanoja
Department of Chemistry, UniVersity of Oslo, P.O. Box 1033 Blindern, N-0315 Oslo, Norway, Institut f u¨ r
Ionenphysik, UniVersit a¨ t Innsbruck, Technikerstrasse 25, A-6020 Innsbruck, Austria, European Commission,
DG - Joint Research Centre, Institute for EnVironment and Sustainability, Climate Change Unit, TP 290,
I - 21020 Ispra (VA), Italy, and Department of Chemistry, Biotechnology and Food Science, Chemistry Section,
Agricultural UniVersity of Norway, P.O. Box 5003, N-1432 Aas, Norway
ReceiVed: December 3, 2004; In Final Form: April 14, 2005
3
The rate coefficients for the gas phase reaction of NO and OH radicals with a series of cycloalkanecarbal-
dehydes have been measured in purified air at 298 ( 2 K and 760 ( 10 Torr by the relative rate method
using a static reactor equipped with long-path Fourier transform infrared (FT-IR) detection. The values obtained
-
11
3
-1 -1
for the OH radical reactions (in units of 10
cm molecule s ) were the following: cyclopropanecar-
baldehyde, 2.13 ( 0.05; cyclobutanecarbaldehyde, 2.66 ( 0.06; cyclopentanecarbaldehyde, 3.27 ( 0.07;
-
14
cyclohexanecarbaldehyde, 3.75 ( 0.05. The values obtained for the NO
3
radical reactions (in units of 10
cm molecule s ) were the following: cyclopropanecarbaldehyde, 0.61 ( 0.04; cyclobutanecarbaldehyde,
.99 ( 0.06; cyclopentanecarbaldehyde, 2.55 ( 0.10; cyclohexanecarbaldehyde, 3.19 ( 0.12. Furthermore,
3
-1 -1
1
the reaction products with OH radicals have been investigated using long-path FT-IR spectroscopy and proton-
transfer-reaction mass spectrometry (PTR-MS). The measured carbon balances were in the range 89-97%,
and the identified products cover a wide spectrum of compounds including nitroperoxycarbonyl cycloalkanes,
cycloketones, cycloalkyl nitrates, multifunctional compounds containing carbonyl, hydroxy, and nitrooxy
2
functional groups, HCOOH, HCHO, CO, and CO .
1
. Introduction
mainly via Hald abstraction (herein referred to as route A) and
to a minor extent via Hring abstraction (herein referred to as route
B). In the presence of NOx, aldehydic H abstraction will form
On a global scale, biogenic emissions of volatile organic
compounds (VOCs) are estimated to dominate over those from
anthropogenic sources by an order of magnitude. Monoterpenes
[(nitroperoxy)carbonyl] cycloalkanes (NPCCAs) and cycloalky-
lperoxy radicals which will react to form cycloalkyl peroxyni-
trates, cyloalkyl nitrates, and cycloalkoxy radicals. The chem-
istry of cycloalkoxy radicals has previously been studied for
(C10H16) represent a relevant fraction of the total biogenic
1
nonmethane hydrocarbon emissions. Due to their high reactiv-
ity, monoterpenes play a dominant role in the chemistry of the
lower troposphere and the planetary boundary layer, in particular
1
7-25
cyclopentoxy and cyclohexoxy radicals.
The fate of cy-
cloalkoxy radicals is either H abstraction to form the corre-
sponding cycloketones or decomposition through ring open-
with respect to the formation of tropospheric ozone and
secondary organic aerosol.2-11
17,18,20,25
ing.
Under atmospheric conditions, cyclopentoxy radicals
The oxidation of monoterpene species such as R-pinene,
3
undergo predominately ring opening with only a minor contri-
â-pinene, ∆ -carene, and sabinene forms multifunctional prod-
bution from R-H abstraction by O2, while for cyclohexoxy
ucts with an aldehyde group in the â-position to the ring.
Substituted cyclobutyl aldehydes (norpinonaldehyde and 2,2-
dimethyl-cyclobutyl-1,3-dicarboxylaldehyde) have been identi-
1
8,20,25
radicals both reaction pathways are important.
The main
reaction products upon ring opening for both cyclopentoxy and
cyclohexoxy radicals include a wide spectrum of mono- and
multifunctional species containing carbonyl, hydroxy, and
fied both as primary9 and as secondary products of R-pinene
,11
4
and â-pinene gas phase degradation. The oxidation of primary
products such as hydroxycaronaldehyde, caronaldehyde, and
pinonaldehyde may result in the formation of substituted
cyclopropane- and cyclobutanecarbaldehydes.
1
7,18,20,21,25
nitrooxy functional groups.
Similar species are expected to be formed in the degradation
of smaller cycloalkoxy radicals. Ring opening is expected to
be the major reaction pathway due to the ring strain of cyclic
1
2-15
On the basis of the reactivity of aliphatic aldehydes
and
on structure activity relationships (SARs),16 the OH radical
reaction with cycloalkanecarbaldehydes is expected to proceed
compounds with less than five carbons.
25
In the present study, the rate coefficients of OH and NO3
radical reactions with cyclopropanecarbaldehyde, cyclobutan-
ecarbaldehyde, cyclopentanecarbaldehyde, and cyclohexanecar-
baldehyde have been determined. In addition, the gas phase
reaction products of the cycloalkanecarbaldehydes with OH
radicals in the presence of NOx have been studied. The structure
of the four investigated compounds is shown at the top of the
respective Schemes 1-4.
*
To whom correspondence should be addressed. E-mail:
barbara.danna@kjemi.uio.no (B.D.’A.); armin.wisthaler@uibk.ac.at
A.W.).
(
†
University of Oslo.
Universit a¨ t Innsbruck.
Institute for Environment and Sustainability.
Agricultural University of Norway.
These authors contributed equally to this work.
‡
§
|
⊥
1
0.1021/jp044495g CCC: $30.25 © 2005 American Chemical Society
Published on Web 05/21/2005

Atmospheric Chemistry of Cycloalkanecarbaldehydes
. Experimental Section
.1. Kinetic Studies. The kinetic investigations were per-
J. Phys. Chem. A, Vol. 109, No. 23, 2005 5105
-
1
2
by coadding 20-40 scans recorded at a 1 cm instrumental
-
1
resolution in the range from 600 to 4000 cm . This resulted in
a FT-IR time resolution of approximately 4 min. More details
2
formed at the University of Oslo. The compounds were mixed
in synthetic air at 298 ( 2 K and 760 ( 10 Torr in a 250 L
electropolished stainless steel reactor equipped with a White-
type multiple reflection mirror system of 120 m optical path
length for on-line Fourier transform infrared (FT-IR) detection.
Infrared spectra were recorded with a Bruker IFS 88 instrument
2
7
can be found in Ballesteros et al. FT-IR spectroscopy was
used to quantify CO, CO2, and HCHO through comparison with
reference spectra of pure compounds whose concentrations were
determined using a nonlinear least-squares fitting procedure
2
8
(
NML4) based on absorption coefficients from the HITRAN
2
9
-
1
database. Other products (e.g., individual nitrate compounds)
could not be unambiguously identified by FT-IR or were below
the detection limit of the instrument. CO and CO2 yields were
corrected for artifact formation observed in a “blank” experiment
where only the OH precursor (2-PN-d6) and NO were irradiated.
The accuracy of CO, CO2, and HCHO measurements is
estimated to be (30%. HCHO was above the FT-IR detection
limit only in the cyclopropanecarbaldehyde study; for the higher
cycloalkanecarbaldehydes, HCHO was quantified only by PTR-
MS.
employing a nominal resolution of 0.5 cm and Harp-Genzel
apodization and adding 100 scans; the time of registration was
∼
60 s.
NO3 radicals were generated in situ by thermal decomposition
of N2O5. OH radicals were generated by the photolysis of
organic nitrites (2-propyl-nitrite (2-PN), 2-propyl-nitrite-d6 (2-
PN-d6), and methyl nitrite) using two Philips TLD-08 fluores-
cence lamps (λmax ∼ 370 nm) mounted in a quartz tube and
inserted into the reaction chamber as described previously by
2
6
D’Anna et al. The lamps were turned off during registration
of the spectra. Typical initial volume fractions were 1-3 ppmV
for cycloalkanecarbaldehydes and reference compounds and
PTR-MS is a chemical ionization (CI) technique for on-line
3
0,31
detection of VOCs.
Ionicon GmbH, Innsbruck, Austria) was interfaced to the
reaction chamber through a Silcosteel capillary (inner diameter
.02 mm, length 30 cm, temperature 70 °C, flow rate 200 mL/
A commercial PTR-MS instrument
(
5
-20 ppmV for organic nitrites or N2O5.
NO3 and OH reaction rate coefficients of cycloalkanecarbal-
1
dehydes were determined by the relative rate method. The
method takes into account the simultaneous decay of a cycloal-
kanecarbaldehyde and a reference compound with the respective
rate coefficients kald and kref, where kref is reliably known.
Assuming that there are no other loss processes for the
cycloalkanecarbaldehyde and the reference compound, the
following relation holds:
min). PTR-MS standard operation (E/N ) 120-140 Td; E
electric field strength, N buffer gas number density, 1 Td )
-
17
2
-1
1
0
cm V molecule ) results in extensive fragmentation of
32
cycloalkanecarbaldehydes and their oxidation products. The
PTR-MS instrument was thus operated at E/N ∼ 80 Td, shifting
+
the maximum in the H3O (H2O)n)0,1,2 CI reagent ion distribution
from n ) 0 to n ) 1. The various implications of nonstandard
33
[
ald]0
kald
kref
[ref]0
[ref]t
PTR-MS operation are described in detail by De Gouw et al.
and Hewitt et al.
ln
)
× ln
(1)
31
{
}
{
}
[
ald]t
CI pathways of a series of target compounds including acyclic
and cyclic aldehydes, acyclic and cyclic ketones, one dialdehyde
and one hydroxyaldehyde species, alkyl nitrates, and NPCCAs
were investigated as part of the calibration experiments described
below. With the exception of nitrooxy and nitroperoxy com-
pounds, CI mass spectra of compounds M were dominated by
where [ald]0, [ald]t, [ref]0, and [ref]t denote the concentrations
of the cycloalkanecarbaldehyde and the reference at times zero
and t, respectively. A plot of ln{[ald]0/[ald]t}versus ln{[ref]0/
[ref]t} yields the ratio kald/kref as the slope. The ratio between
the concentrations of the cycloalkanecarbaldehydes and the
reference compounds was found by spectral subtraction using
reference spectra of the pure starting compounds, spectra of
other compounds identified in the reaction mixture, and a
linearly sloping background.
The chemical and photochemical stability of references and
cycloalkanecarbaldehyde was investigated separately; the com-
pounds showed lifetimes in the order of days and wall loss or
direct photolysis could thus be neglected in the data analyses.
+
+
MH ions. Dehydration to form [MH - H2O] ions was
observed for larger cycloalkanecarbaldehydes and for hydroxy-
acetaldehyde. Dehydration is expected to be a dominant CI
3
4
+
pathway for larger hydroxy compounds. [MH + H2O] ions
were typically observed with relative abundances of <20%.
Fragmentation of the carbon skeleton, which complicates the
interpretation of the CI spectra, was observed as a minor route
(
<25%) for cyclopropanecarbaldehyde, n-butyl nitrate, and
pentane-1,5-dial. Alkyl nitrates and NPCCAs gave only weak
2
.2. Product Studies. The OH-cycloalkanecarbaldehyde
+
35
(
<1%) MH signals. CI mass spectra of n-propyl nitrate and
product studies were performed at the Institute for Environment
and Sustainability (IES) in Ispra. Reactants were mixed in
synthetic air at 299 ( 3 K and 740 ( 5 Torr total pressure in
a 480 L cylindrical Teflon-coated evacuable glass reaction
chamber. UV radiation was provided by 18 UV-vis lamps (λ
+
n-butyl nitrate were dominated by [MH - HNO3] ions; minor
+
CI pathways (<30%, in total) included [MH + H2O - HNO3]
+
ion formation, nitrous acid loss to give [MH - HNO2] ions,
fragmentation into NO2 ions, and fragmentation of the carbon
skeleton. Kames et al. observed a similar product ion pattern
for C -C alkyl nitrates using methane CI. CI mass spectra of
NPCCAs were dominated by [MH + H2O - HNO3] ions; for
details, see Hansel and Wisthaler. Fragmentation into the
respective cycloalkyl ions and cycloalkylcarboxy ions was also
observed. Due to the lack of authentic standards, CI pathways
of multifunctional nitrates could not be investigated. Kames et
+
3
6
)
300-500 nm with λmax ∼ 370 nm) surrounding the chamber.
2
6
OH radicals were generated by the photolysis of 2-PN-d6.
1
8
+
Irradiations were carried out for about 30 min. Typical initial
volume fractions for cycloalkanecarbaldehydes and 2-PN-d6
were 1 ppmV and 2-3 ppmV, respectively. An additional 6-8
ppmV of NO was added to the reactant mixtures.
The analytical techniques used to detect the reaction products
were FT-IR spectroscopy and proton-transfer-reaction mass
spectrometry (PTR-MS). On-line FT-IR spectroscopy was
performed using a White-type multiple reflection mirror system
of a total optical path length of 81.24 m coupled to a Bruker
IFS 113 V FT-IR spectrometer. FT-IR spectra were obtained
3
7
3
6
+
+
+
al. found abundant MH , [MH - HNO2] , NO2 , and [MH
+
-
HNO3] ions from methane CI of 2-nitrooxy-ethanol.
The PTR-MS instrument was calibrated for a series of target
compounds by sampling known quantities of these species from
the reaction chamber. C3-C6 cycloalkanecarbaldehydes, C4-

5
106 J. Phys. Chem. A, Vol. 109, No. 23, 2005
D’Anna et al.
C6 cycloketones, HCHO, and butyl nitrate were quantitatively
added to the reaction chamber using a standard manometric
method. Pentane-1,5-dial and hydroxyacetaldehyde, both in
aqueous solution, were first quantified using the FT-IR technique
by comparison with calibrated spectra of aldehydes and then
fed to the PTR-MS instrument for calibration. The IR absorption
coefficient of pentane-1,5-dial was assumed to be twice as large
as that of pentanal.38 Known quantities of NPCCAs were
generated in the reaction chamber from cycloalkanecarbaldehyde
into consideration the fragmentation routes described above),
C abundance, and thermal behavior of the observed signals
1
3
which was studied at the end of each experiment when
photolysis lamps were switched off. Sampling was then
performed through a heated stainless steel inlet (T ) 150 °C).
Peroxyacyl nitrates are known to decompose at 150 °C, and
ion signals associated with peroxyacyl nitrates can thus be
37
selectively identified and quantified. After tentative identifica-
tion, ion count rates were converted into volume mixing ratios
using the calibration factors reported above. Carbon yields of
individual compounds were determined from least-squares
regressions; no corrections for secondary reactions were applied
except where explicitly stated.
2.3. Chemicals. For the kinetic study, N2O5 was synthesized
by mixing gas streams of O3 and NO2 and trapping the products
at -78 °C. N2O5 was purified by vacuum distillation prior to
its use.
+
NO3 in excess of NO2. NPCCA was the only product (g99%)
of this reaction detected by both FT-IR and PTR-MS. NPCCA
calibration was thus performed on the basis of a 1:1 conversion
of known quantities of C3-C6 cycloalkanecarbaldehydes into
the respective NPCAAs. In addition, the PTR-MS instrument
was calibrated using an oxygenate standard (Apel-Riemer
Environmental Inc., Denver, CO) containing 1 ppmV of
methanol, ethanol, acetaldehyde, acetone, and 2-butanone,
respectively.
2PN and 2PN-d6 were prepared from the corresponding
alcohols following the procedure reported for butyl nitrite.40
The sensitivity (or calibration factor) of PTR-MS with respect
to a certain compound is defined as the number of mass
discrimination corrected product ions produced at a trace gas
2
-Propanol-1,1,1,3,3,3-d6 was prepared from acetone-d6 by
reduction with LiAlH4 in dry ether. CH3ONO was synthesized
4
1
+
according to Taylor et al.
Cyclobutanecarbaldehyde and cyclopentanecarbaldehyde were
mixing ratio of 1 ppbV and at a total H3O (H2O)n)0,1,2 ion signal
of 1 million normalized counts per second (ncps). For all
investigated alcohols, ketones, and aldehydes with the exception
of HCHO, the observed sensitivity was in the range 33 ( 5
ncps/ppbV. This calibration factor was assigned to all oxygen-
ates for which no authentic standards were available; the
associated uncertainty of (15% corresponds to the accuracy
of oxygenate measurements. The individual PTR-MS sensitivity
values are given as Supporting Information (Table 7S). The
reduced HCHO sensitivity of 1.6 ncps/ppbV is explained by
4
2
synthesized according to a standard procedure adding rapidly
4
5
.05 g (47.1 mmol) of cyclobutyl- or cyclopentyl-methanol in
0 mL of methylene chloride to a stirred suspension of 10.15 g
(47.1 mmol) of pyridinium chlorochromate in 100 mL of
methylene chloride. The mixture was stirred at room temperature
for 3 h, diluted with 700 mL of anhydrous ether, filtered through
Celite, and carefully evaporated at atmospheric pressure using
a Vigreux column. The residue was dissolved in 50 mL of ether
and dried with CaCl2 to eliminate unreacted starting alcohol.
Evaporation followed by distillation yielded 2.98 g (75%) of
cycloalkanecarbaldehyde. To get an analytical sample, the
the loss of protonated HCHO ions through reaction with H2O,
as discussed in detail by Hansel et al.39 For butyl nitrate and
NPCCAs, sensitivities in the range 25 ( 6 ncps/ppbV were
found. This value was generally used for nitrate species for
which no authentic standards were available; the associated
uncertainty of (25% corresponds to the accuracy of nitrate
measurements. An additional uncertainty arises from the fact
that ion signals may not be unambiguously attributable to simple
oxygenates or to nitrate species. For such signals, the associated
error is on the order of (30%. This is in agreement with the
previously described method-inherent feature of PTR-MS which
4
3
distillation was repeated three times. The purity of cyclobu-
tanecarbaldehyde and cyclopentanecarbaldehyde was determined
by PTR-MS analysis to be >85% and >92%, respectively.
Other chemicals used for the kinetic studies were as fol-
lows: cyclopropanecarbaldehyde (Aldrich, 98%), cyclohexan-
ecarbaldehyde (Fluka, ∼90-95%), but-1-ene (Fluka, 99%),
propene (Fluka, 99%), and synthetic air (AGA plus; CO and
NOx < 100 ppb, CnHm < 1 ppm). The same batches of
cycloalkanecarbaldehydes and 2-PN-d6 were used for the
product studies. Other chemicals used in the product studies
were the following: 1-butyl-nitrate (synthesized by nitration of
constrains the variability of calibration factors for oxygenated
hydrocarbons to (30%.30,31
Product study data were taken from the first 18-22 min of
reaction, during which about 50% of the initially present
cycloalkanecarbaldehyde had reacted and a linear increase of
products versus reactant loss was observed. Mass scans in the
range from m/z 20 to 200 were performed with a dwell time of
1
-butanol), pentane-1,5-dial (Fluka, solution, 70% in H2O),
HCHO (Fluka, 90% CH2O with 10% CH3OH for stabilization),
cyclobutanone (Fluka, >99%), cyclopentanone (Fluka, >99%),
cyclohexanone (Fluka, >99%), hydroxyacetaldehyde (Sigma-
Aldrich, dimer, >98%), cyclohexane (Carlo Erba, 99.5%), NO
1
s per m/z, resulting in a sample time resolution of 3 min. The
(
Alpha Gaz, >99%), NO2 (Messer, >98%), O2 (SIO >99.95),
length of the reaction time was determined by the 3-4 min
time resolution of the two analytical instruments and the
necessity to collect sufficient data points for an accurate linear
regression analysis. PTR-MS ion signals were corrected for mass
discrimination, instrumental background, and artifact formation
observed in blank runs of air containing the OH precursor (2-
PN-d6) and NO. The m/z 31 signal (HCHO) was corrected for
signal loss as discussed above. Only products with a signal-
based yield >1% were further investigated. The number of
carbon atoms was derived from the 1 C-signal abundance or,
in the case of interferences, estimated. An error due to
fragmentation of the carbon skeleton during CI is expected to
be minor but cannot be excluded completely. Compound
assignment was based on molecular weight information (taking
and synthetic air (SIO, >99.95%).
3
. Results and Discussion
3.1. Kinetics. For each compound, data of the independent
experiments were analyzed according to eq 1 using a weighted
least-squares procedure including uncertainties in both reactant
4
4
concentrations and allowing a zero-point offset. The spectral
ranges and the compounds included in the data analysis are
specified in Table 1. But-1-ene was used in all the experiments
as a reference compound. In addition, propene was used as a
reference in the OH-cyclopentanecarbaldehyde studies. Plots
of ln{[ald]0/[ald]t} versus ln{[ref]0/[ref]t} are shown in Figures
1 and 2 for the NO3 and OH experiments, respectively. Table
3

Atmospheric Chemistry of Cycloalkanecarbaldehydes
TABLE 1: Spectral Regions and Compounds Included in the Spectral Subtraction Procedures
compounds included in the spectral analysis
J. Phys. Chem. A, Vol. 109, No. 23, 2005 5107
compound
spectral range (cm-1)
OH
NO
3
cyclopropanecarbaldehyde
cyclobutanecarbaldehyde
cyclopentanecarbaldehyde
cyclohexanecarbaldehyde
but-1-ene
2775-2670
2760-2670
2750-2670
2740-2670
3140-3050
3150-3050
HCHO, propanal, 2-PN, acetone
HCHO, propanal
HCHO, acetaldehyde, propanal
HCHO, propanal,
propanal
acetaldehyde
HCHO, propanal
HCHO, propanal
HCHO, propanal
HCHO, propanal
2 5
N O
, cyclopropanecarbaldehydea
b
propene
a
Used only in the study of cyclopropanecarbaldehyde. b Used only in the study of cyclopentanecarbaldehyde with 2-PN.
Figure 1. Plots of ln{[ald]
with NO radicals. Different symbols indicate independent experiments. The errors quoted are the 3σ statistical errors: (A, cyclopropanecarbaldehyde)
4 data points, three experiments, krel ) 0.452 ( 0.027; (B, cyclobutanecarbaldehyde) 31 data points, three experiments, krel ) 1.477 ( 0.042; (C,
cyclopentanecarbaldehyde) 29 data points, three experiments, krel ) 1.891 ( 0.072; (D, cyclohexanecarbaldehyde) 24 data points, two experiments,
rel ) 2.363 ( 0.084.
0 t 0 t
/[ald] } versus ln{[ref] /[ref] } for the decays of cycloalkanecarbaldehydes and reference compounds during reaction
3
2
k
2
summarizes the measured reaction rate coefficients. The error
of organonitrites. Both O3 and NO3 react relatively fast with
the reference compound 1-butene. However, a kinetic model
of the reaction system indicates that the concentrations of these
oxidants were too low ([NO3] , 1 ppt, [O3] < 5 ppb) to be
relevant. The derived relative rate coefficients are believed to
be accurate within (10%.
limits correspond to 3σ from the statistical analyses, and neither
include uncertainties in the reaction rate coefficients of the
reference compounds nor possible systematic errors in the
experiments. Random “experimental” errors arise from tem-
perature fluctuations and from the quantitative determination
of relative reactant concentrations by spectral subtraction. These
errors are reflected in the statistical error from the least-squares
analysis and amount to less than (10% (3σ) in all cases. In
addition, systematic errors may arise from competing reactions
in the smog chamber. In the NO3 studies, OH radicals may be
produced from HO2 (HO2 + NO3 f OH + NO2 + O2). The
influence of such systematic errors in the derived rate coef-
ficients has been shown to be negligible for the NO3/acetalde-
hyde system.26 This should also be valid for the present study
performed under similar experimental conditions. In the OH
studies, O3 and NO3 radicals are formed during the photolysis
3.2. Tropospheric Lifetimes. The estimated ambient con-
centrations of OH and NO3 radicals combined with the measured
rate coefficients may be used to derive the tropospheric lifetimes
of cycloalkanecarbaldehydes. Using a globally averaged diurnal
5
-3 46
OH radical concentration of 9.4 × 10 radicals cm , the
calculated lifetimes range from 14 h for cyclopropanecarbal-
dehyde to 9 h for cyclohexanecarbaldehyde. Using a 12 h
average nighttime tropospheric NO3 radical concentration of
4
5
8
-3 47,48
∼5.0 × 10 molecule cm ,
the calculated lifetimes range
from 8 days for cyclopropanecarbaldehyde to 1.5 days for
cyclohexanecarbaldehyde.

5108 J. Phys. Chem. A, Vol. 109, No. 23, 2005
D’Anna et al.
Figure 2. Plots of ln{[ald]
0
/[ald]
t
} versus ln{[ref]
0
/[ref]
t
} for the decays of cycloalkanecarbaldehydes and reference compounds during reaction
with OH radicals. Different symbols indicate independent experiments. The errors quoted are the 3σ statistical errors: (A, cyclopropanecarbaldehyde)
8 data points, three experiments, krel ) 0.678 ( 0.006; (B, cyclobutanecarbaldehyde) 18 data points, two experiments, krel ) 0.847 ( 0.007; (C,
cyclopentanecarbaldehyde) 27 data points, two experiments, krel ) 1.043 ( 0.007; (D, cyclohexanecarbaldehyde) 33 data points, two experiments,
rel ) 1.195 ( 0.017.
4
k
TABLE 2: Rate Coefficients (in Units of cm molecule-1 s-1
3
)
fragmentation pattern obtained in the NO3-cyclopropanecar-
baldehyde experiment.
for the Reaction of Cycloalkanecarbaldehydes with NO
3
and
OH Radicals at 298 K
Other major reaction products are CO and CO2 which account
for a carbon yield of 11.5 and 19.4%, respectively. The ion
compound
kNO (×10-14) kOH (×10-11
)
3
a
a
a
a
b
cyclopropanecarbaldehyde 0.61 ( 0.04
2.13 ( 0.05
2.66 ( 0.06
3.27 ( 0.07
3.75 ( 0.05
+
signal at m/z 31 is associated with HCHO (MH ), which was
b
cyclobutancarbaldehyde
1.99 ( 0.06
also detected by FT-IR. The FT-IR derived HCHO carbon yield,
20.7 ( 1.4%, is in good agreement with the PTR-MS derived
yield, 19.6 ( 2.0%. The formation of CO2 may be explained
by decomposition of cyclopropylcarboxy radicals, and the
formation of CO and HCHO, by the unimolecular decomposition
b
b
2.74 ( 0.05c
cyclopentanecarbaldehyde 2.55 ( 0.10
cyclohexanecarbaldehyde 3.19 ( 0.12
a
-14
3
-1 -1 78
Reference but-1-ene, kNO3 ) 1.35 × 10
cm molecule
s .
b
c
-11
3
-1 -1 78
Reference but-1-ene, kOH ) 3.14 × 10
cm molecule
s .
-11
3
-1 -1 78
Reference propene, kOH ) 2.63 × 10 cm molecule
.3. Product Study. Due to the complexity of the reaction
mechanism, the degradation schemes (Schemes 1-4) are
simplified. Reactants such as NO, NO2, and O2 are omitted.
The identified or tentatively identified products generated by
the OH-cycloalkanecarbaldehyde radical reaction are presented
in the degradation schemes with an asterisk and in bold.
s .
•
•
of CHOCH2CH2O and CHOCH2O radicals (see Scheme 1).
Part of the observed CO may also arise from the secondary
reaction of HCHO with OH radicals. OH reaction rate coef-
3
-
11
3
-1 -1
ficients of 2.13 × 10
cm molecule
s
for cyclopropan-
-
12
3
-1
ecarbaldehyde (this work) and 9.37 × 10 cm molecule
for HCHO result in a corrected HCHO carbon yield of
4.4% at 50% cyclopropanecarbaldehyde conversion. After the
-
1
49
s
2
3
.3.1. Cyclopropanecarbaldehyde. Table 3 summarizes the
lights had been turned off, the formation of CO, CO2, and
HCHO was observed from NPCCPr decomposition, confirming
that these species are primary reaction products. The observation
products and the respective molar and carbon yields of the OH-
cyclopropanecarbaldehyde reaction. Scheme 1 shows the pro-
posed degradation paths. [(Nitroperoxy)carbonyl]cyclopropane
NPCCPr) is the major reaction product with a carbon yield of
5.9%. The ion signal at m/z 103 ([MH + H2O - HNO3] )
and cyclopropyl and cyclopropylcarbonyl fragment ions at m/z
1 and m/z 69, respectively, have been fully assigned to
NPCCPr. NPCCPr identification and quantification is based on
results from the thermal decomposition study and the NPCCPr
•
of CHOCH2CH2O unimolecular decomposition contradicts
(
3
+
previous studies on similar systems: calculations on the carbonyl
•
alkoxy radical, CH3COCH2CH2O , predict that unimolecular
decomposition does not occur under atmospheric conditions,50
and studies of the 1-propoxy radical also indicate that unimo-
4
2
-1 51-56
lecular decomposition is quite slow (∼(1-4) × 10 s
).

Atmospheric Chemistry of Cycloalkanecarbaldehydes
J. Phys. Chem. A, Vol. 109, No. 23, 2005 5109
TABLE 3: Ion Signals, Molar Yields, and Carbon Yields of the OH-Cyclopropanecarbaldehyde Reaction Products (the
Indicated Errors Are Standard Errors of the Least-Squares Fits)
compound
signal (m/z)
molar yield (%)
C yield (%)
interferences (m/z)
nitroperoxy carbonyl cyclopropane (NPCCPr)
HCHO
41, 69, 103
31
35.9
78.4
82.7
3.7
2.0
0.9
35.9 ( 1.5
19.6 ( 2.0
20.7 ( 1.4
1.9 ( 0.2
1.5 ( 0.1
0.9 ( 0.1
60.8
11.5 ( 0.6
19.4 ( 0.7
91.8
(FT-IR)
2
-nitrooxy-ethanol
61
73
87
2-hydroxy-ethanal
3-nitrooxy-propanal
propane-1,3-dial
unidentified
total VOCs
CO
(FT-IR)
(FT-IR)
46.1
77.8
CO
total
2
SCHEME 1: Degradation Paths in the OH-Initiated Oxidation of Cyclopropanecarbaldehyde
•
+
Ethanedial, which may also arise from CHOCH2O decom-
propane-1,3-dial (MH ). The signal at m/z 73 may also be partly
+
+
position, has been observed in traces (<1%) at m/z 59 (MH ).
Studies on the acetonoxy radical, CH3COCH2O , and propi-
onoxy radical, CH3CH2COCH2O , show that decomposition
to HCHO and the carbonyl radical is the dominant reaction
pathway. Similarly, it is expected that the carbonyl alkoxy
associated with 3-nitrooxy-propanal ([MH - HNO2] ). Traces
of its molecular ion (MH ) have been detected at m/z 120.
The ion signal at m/z 61 with a carbon yield of 1.9% may be
associated with 2-nitrooxy-ethanol ([MH - HNO2] ). Other ion
signals possibly associated with 2-nitrooxy-ethanol have been
•
57
+
•
50
+
•
+
radical, CHOCH2O , will predominantly decompose to form
observed in traces at m/z 45 ([MH - HNO3] ) and m/z 108
+
HCHO and CO.
(MH ). In CI of 2-nitrooxy-ethanol, however, HNO3 loss (m/z
Some minor products have also been detected. The ion signal
at m/z 73 with a carbon yield of 1.5% has been assigned to
45) apparently dominates over HNO2 loss (m/z 61),36 which is
inconsistent with the signal ratio observed here. The ion signal

5110 J. Phys. Chem. A, Vol. 109, No. 23, 2005
D’Anna et al.
TABLE 4: Ion Signals, Molar Yields, and Carbon Yields of the OH-Cyclobutanecarbaldehyde Reaction Products (the
Indicated Errors Are Standard Errors of the Least-Squares Fits)
compound
signal (m/z)
molar yield (%)
C yield (%)
interferences (m/z)
nitroperoxy carbonyl cyclobutane (NPCCB)
55, 83, 117
57, 75
31
30.3
31.3
52.4
3.7
30.3 ( 1.3
18.8 ( 1.0
10.5 ( 0.8
3.0 ( 0.3
3
-hydroxy-propanal
3-nitrooxy-propanol (75)
HCHO
4
-nitrooxy-butanal
-nitrooxy-ethanol
71
cylobutyl nitrate
cyclobutanone
2-hydroxy-ethanal (61)
4-nitrooxy-butanal
ethanedial (59)
2
45, 61
87
59, 77
7.6
3.7
2.6
3.0 ( 0.2
2.9 ( 0.1
1.5 ( 0.2
butane-1,4-dial
-nitrooxy-propanol
3
unidentified species (77)
propane-1,3-dial
unidentified
cyclobutyl nitrate
unidentified
total VOCs
CO
73
2.4
1.3
<1.3
1.5
1.4 ( 0.1
1.3 ( 0.2
<1.0
0.6 ( 0.2
74.4
5.6 ( 0.5
14.8 ( 1.1
94.8
101
55
29
(FT-IR)
(FT-IR)
27.9
74.1
CO
total
2
at m/z 61 may also partly arise from 2-hydroxy-ethanal (MH+).
Fragmentation of cyclopropanecarbaldehyde produced an intense
m/z 43 ion signal, hindering the detection of the 2-hydroxy-
ethanal fragment ([MH - H2O] ). In the current absence of
other explanations for the m/z 61 ion signal, the formation of
decomposition study and the NPCCB fragmentation pattern
obtained in the NO -cyclobutanecarbaldehyde experiment.
3
The ion signals m/z 57 and m/z 75 have been assigned to
+
+
+
3
-hydroxy-propanal ([MH - H2O] , MH ), the major acyclic
product accounting for 18.8% of the carbon yield. OH reaction
2
-nitrooxy-ethanol and/or 2-hydroxy-ethanal may be explained
-11
3
-1 -1
rate coefficients of 2.66 × 10
cm molecule
s
for
cm
•
via a 1,4-H shift of the CHOCH2CH2O radical. Such a reaction
involves a five-membered transition state with some strain
energy. The activation energy for a 1,4-H shift in the 1-butoxy
radical has been calculated to be in the range from 54 to 75
-11
3
cyclobutanecarbaldehyde (this work) and 1.99 × 10
molecule
-hydroxy-propanal yield of 24.8% at 50% cyclobutanecarbal-
-1 -1
64
s
for 3-hydroxy-propanal result in a corrected
3
dehyde conversion. However, part of the m/z 75 signal may be
2
-1 58
kJ/mol, with the rate coefficient at 1 atm being ∼10 s . In
+
attributable to 3-nitrooxy-propanol ([MH - HNO2] ) with the
addition, there is no experimental evidence of a 1,4-H shift in
+
molecular ion (MH ) at m/z 122 being present in traces.
9,60
3
-hexoxy and 3-pentoxy radicals.5
However, the hydrogen
3
-Hydroxy-propanal may be formed via route A after ring
of an aldehydic group is weakly bound to the substrate and Hald
opening, isomerization through a 1,5-H shift, and further
decomposition (see Scheme 2). The rate coefficient for a 1,5-H
shift from a secondary group (-CH2-) has been evaluated to
abstraction is expected to be easier. The formed â-hydroxy
•
alkoxy radical, HOCH2CH2O , will preferentially decompose
•
5
to HCHO + CH2OH with a rate coefficient of ∼(1-2) × 10
6
-1 65,66
-
1 61,62
be ∼10 s .
The same compound may also be formed via
s ,
while reaction with O2 will produce 2-hydroxy-ethanal
4
-1 61,63
route B after ring opening, isomerization through a 1,6-H shift,
and decomposition of the carbonyl hydroxy alkoxy radical. This
process is less favorable than the 1,5-H shift for entropic
with a rate coefficient of ∼4 × 10 s .
The ion signal at m/z 87 has been corrected for a minor
contribution from NPCCPr and amounts to a carbon yield of
5
8,67
reasons.
However, isomerization via a 1,6-H shift has been
0
.9%. This signal remains unidentified and will be discussed
later.
Using the SAR16 method, a fractional contribution of route
observed when the hydrogen to be removed is bound to an
6
8
3
oxygen-substituted carbon. A rate coefficient of 5.6 × 10
-
1
•
s
was derived for a 1,6-H shift in CH3OCH2OCH2O by
B of ∼1% was calculated. For cyclopropanecarbaldehyde, this
6
9
quantum chemical calculations, but no data are available for
a 1,6-H shift from an aldehydic group. Abstraction via a 1,6-H
shift from an aldehydic group has been proposed in a previous
decomposition route can thus be safely neglected.
It can thus be concluded that the OH-cyclopropanecarbal-
dehyde reaction essentially proceeds via Hald abstraction. To a
first approximation, 36% of the reacted cyclopropanecarbalde-
hyde produces NPCCPr with the remaining 64% forming mainly
CO2 + 2HCHO + CO. The observed CO2, HCHO, and CO
molar yields were 78, 98 (corrected for the secondary OH
reaction), and 46%, respectively. The CO2 yield implies a
cyclopropoxy radical formation percentage higher than that
expected from HCHO and CO yields. It must, however, be
considered that these estimates are affected by a relatively large
uncertainty due to the low measurement accuracy of (30%.
2
0
study as a possible decomposition pathway.
CO and CO2 are also major products accounting for carbon
yields of 5.6 and 14.8%, respectively. CO2 may arise primarily
from decomposition of cyclobutylcarboxy and 4-hydroxy bu-
tyryloxy radicals and from the secondary OH reaction of
3
-hydroxy-propanal. CO may arise from the decomposition of
a carbonyl alkoxy radical via route B, from secondary OH
reactions of C3 and C4 dialdehydes, and also from the reaction
of HCHO with OH radicals.
3
.3.2. Cyclobutanecarbaldehyde. Table 4 summarizes the
HCHO has been quantified by PTR-MS only, as levels just
barely exceeded the detection threshold of the FT-IR instrument.
The HCHO signal at m/z 31 (MH ) accounts for a carbon yield
products and the respective molar and carbon yields for the
reaction of cyclobutanecarbaldehyde with OH radicals. Scheme
+
2
illustrates the proposed degradation pathways.
of 10.5%. No HCHO formation was observed from NPCCB
decomposition after the lights had been turned off, indicating
that HCHO is a secondary reaction product. HCHO may partly
arise from the secondary reaction of 3-hydroxy-propanal with
OH radicals. Secondary OH reaction of highly reactive C3 and
C4 dialdehydes may also generate minor amounts of HCHO.
[(Nitroperoxy)carbonyl]cyclobutane (NPCCB) accounts for
a carbon yield of 30.3% and forms ion signals at m/z 117 ([MH
+
+
H2O - HNO3] ) and cyclobutyl and cyclobutylcarbonyl
fragments at m/z 55 and m/z 83, respectively. NPCCB identi-
fication and quantification is based on results from the thermal

Atmospheric Chemistry of Cycloalkanecarbaldehydes
J. Phys. Chem. A, Vol. 109, No. 23, 2005 5111
SCHEME 2: Degradation Paths in the OH-Initiated Oxidation of Cyclobutanecarbaldehyde
A possible degradation product of 3-hydroxy-propanal is
2-Hydroxy-ethanal may also be formed via route B, but the
contribution from this pathway is probably negligible.
+
2
-nitrooxy-ethanol observed at m/z 45 ([MH - HNO3] ) and
+
m/z 61 (([MH - HNO2] ) with a total carbon yield of 3.0%.
Contrarily to the cylopropanecarbaldehyde experiment, the ion
signal at m/z 45 was found to dominate over the m/z 61 signal
which is consistent with the expected CI pathways for 2-ni-
The ion signal at m/z 87 with a C yield of 2.9% has been
+
assigned to butane-1,4-dial (MH ). Butane-1,4-dial may be
•
formed by H abstraction from the CHO(CH ) CH O radical
2
2
2
•
(route A) or by HCO elimination from the CHOCH(O )CH2-
CH2CHO radical (route B). The signal may also be associated
36
trooxy-ethanol. Traces of the 2-nitrooxy-ethanol molecular ion
+
+
(
(
(
MH ) have been detected at m/z 108. The ion signal at m/z 61
with 4-nitrooxy-butanal ([MH - HNO2] ). Traces of its
+
0.6% C yield) may also be partly assigned to 2-hydroxy-ethanal
molecular ion (MH ) have been detected at m/z 134.
+
MH ) formed from the decomposition of 3-hydroxypropanal.
Secondary OH reactions of butane-1,4-dial may generate
+
As discussed previously, the â-hydroxy alkoxy radical, HOCH2-
propane-1,3-dial (m/z 73, MH ) observed with a carbon yield
•
•
61,62
+
CH2O , will mainly decompose to give HCHO + CH2OH
of 1.4% as well as traces of ethanedial (m/z 59, MH ). A fraction
6
2,63
and only to a minor extent to form 2-hydroxy-ethanal.
of propane-1,3-dial may also arise via route B through decom-

5112 J. Phys. Chem. A, Vol. 109, No. 23, 2005
D’Anna et al.
TABLE 5: Ion Signals, Molar Yields, and Carbon Yields of the OH-Cyclopentanecarbaldehyde Reaction Products (the
Indicated Errors Are Standard Errors of the Least-Squares Fits)
compound
signal (m/z)
molar yield (%)
C yield (%)
interferences (m/z)
nitrooxy carbonyl cyclopentane (NPCCP)
69, 97, 131
71, 89
85
33.4
18.4
10.5
33.4 ( 1.3
12.2 ( 0.9
8.7 ( 0.6
4
-hydroxy-butanal
-nitrooxy-pentanal
5
cyclopentanone
-hydroxybutane-1,4-dial
2
cyclopentyl nitrate
5
-hydroxy-2-nitrooxy-pentanal
101
6.3
5.3 ( 0.3
pentane-1,5-dial
5
-nitrooxy-pentanal
cyclopentyl nitrate
unidentified
69
115
87
75
31
47
43
4.4
1.8
2.4
2.5
6.9
3.7
1.1
3.7 ( 0.3
1.8 ( 0.2
1.6 ( 0.1
1.3 ( 0.1
1.2 ( 0.4
0.6 ( 0.1
0.4 ( 0.1
70.1
2
-hydroxy-4-nitrooxy-butanal
-hydroxy-propanal
butane-1,4-dial
3
HCHO
HCOOH
unidentified
total VOCs
CO
(FT-IR)
(FT-IR)
40.5
75.7
6.7 ( 0.2
12.6 ( 0.3
89.4
CO
total
2
•
position of the CHOCH(O )CH2CHO radical. As discussed
previously, it is expected that the carbonyl alkoxy radical,
It can be concluded that the OH-cyclobutanecarbaldehyde
reaction proceeds essentially via Hald abstraction (g90%) leading
to the formation of 30% NPCCB and ∼60% cyclobutylperoxy
•
CHOCH2O , will decompose to form HCHO and CO. The ion
signals at m/z 59 ([MH - HNO3] ) and m/z 77 ([MH + H2O
HNO3] , with a total carbon yield of 1.5%, are thus primarily
assigned to 3-nitrooxy-propanol. As mentioned above, traces
of the to 3-nitrooxy-propanol molecular ion (MH ) have been
detected at m/z 122. An additional unidentified compound is
likely to contribute to the ion signal at m/z 77.
The ion signal at m/z 71 accounting for a carbon yield of
+
radical + CO . Of the 60% cyclobutylperoxy radical fraction,
2
+
-
∼10% goes into minor products (cyclobutyl nitrate, cyclobu-
tanone, 4-nitrooxy-butanal, butane-1,4-dial, and its oxidation
+
products) and ∼50% goes into CO
2
+ 3-hydroxy-propanal. After
correction for secondary reaction with OH radicals, the observed
molar 3-hydroxy-propanal yield is 41%; the CO2 molar yield
is 74% which is lower than that expected from decomposition
of cyclobutylcarboxy radicals and 4-hydroxy butyryloxy radi-
cals. Again, it must be pointed out that these estimates have a
relatively large uncertainty due to the low measurement ac-
curacy, that is, (30% for CO2 and (25% for NPCCB.
+
3
.0% may be assigned to 4-nitrooxy-butanal ([MH - HNO3] ),
+
+
cyclobutanone (MH ), or cyclobutyl nitrate ([MH - HNO2] ).
The formation of the nitrooxyaldehyde is considered to be most
+
favorable. As stated above, traces of its molecular ion (MH )
have been detected at m/z 134. Previous experimental results
are available for the cyclopentoxy radical (lower strain energy
than cyclobutoxy radical), and ring opening represents the major
decomposition pathway.25 Other studies show that the elimina-
tion of hydrogen through reaction with O2 implies a change
3.3.3. Cyclopentanecarbaldehyde. Table 5 summarizes prod-
ucts and respective molar and carbon yields for the reaction of
cyclopentanecarbaldehyde with OH radicals. Scheme 3 shows
the proposed degradation paths. [(Nitroperoxy)carbonyl]cyclo-
pentane (NPCCP) is the main reaction product with a carbon
yield of 33.4%. The ion signal at m/z 131 ([MH + H2O -
3
2
from sp to sp hybridization of the carbon in the four-membered
ring, increasing the ring strain, while ring opening causes the
+
HNO3] ) and the cyclopentylcarbonyl fragment ion at m/z 97
release of ring strain in the cyclobutoxy radical, reducing the
energy barrier for the reaction.70
have been fully assigned to NPCCP; the cyclopentyl fragment
ion at m/z 69 is partly attributed to NPCCP. Qualitative and
quantitative assignments are based on results from the thermal
decomposition study and the NPCCP fragmentation pattern
obtained in the NO3-cyclopentanecarbaldehyde experiment.
An accurate analysis of the ion signal at m/z 55 associated
+
with cyclobutyl nitrate ([MH - HNO3] ) is not possible due to
+
the interfering H3O (H2O)2 CI reagent ion signal. It is, however,
possible to analyze the 13C isotope on m/z 56, which does not
suffer from this interference. Though affected by a large error,
an upper limit for the m/z 55 carbon yield of 1% may be derived.
The fragment ion signal at m/z 29 cannot be attributed to a
specific compound. The ion signal at m/z 101 remains unidenti-
fied and will be discussed later.
A fraction of the ion signal at m/z 69 accounting for a carbon
yield of 3.7% has been assigned to cyclopentyl nitrate ([MH -
+
HNO3] ). The given value may slightly underestimate the
cyclopentyl nitrate yield because it does not take into account
HNO2 loss as a possible CI pathway. Using the SAR method,16
the decomposition via route A was calculated to amount to 70%
with 33.4% going into NPCCP and the remaining 36.6% going
into cyclopentyl radicals (see Scheme 3). Product yields can
thus be calculated relative to cyclopentyl radical decomposition
and be compared with literature values reported for cyclopentane
oxidation. The cyclopentyl nitrate yield relative to cyclopentyl
radical decomposition is 12.0% which is considerably higher
than the 4.5 ( 1.5% yield reported by Takagi et al.25 Aside
from the inaccuracy of the measurements, this discrepancy may
be explained by the presence of interfering species at m/z 69
An ∼11% fractional contribution of route B was calculated
using the SAR1 method. There is no unambiguous experimental
evidence for product formation via route B. Ion signals which
are expected to be unequivocally attributable to multifunctional
ring opening products from route B (e.g., 2-keto-pentane-1,5-
dial and 2-keto-5-nitrooxy-pentanal, see Scheme VS in the
Supporting Information) have not been observed. Other products
such as 3-hydroxypropanal, dialdehydes, HCHO, and CO are
primarily generated via route A, but a minor contribution from
route B cannot be ruled out.
6
On the basis of a SAR16 evaluation and considering the
current knowledge on alkoxy radical degradation, it can,
however, be assumed that the large majority of the observed
products arises from degradation via route A.5
16
and by the fact that the SAR method underestimates the
18
fractional abundance of route A. Hanson et al. did not observe
cyclopentyl nitrate, but HNO3 loss in their PTR-MS study was
not considered as a possible CI pathway for cyclopentyl nitrate.
5,58,67-69,71

Atmospheric Chemistry of Cycloalkanecarbaldehydes
J. Phys. Chem. A, Vol. 109, No. 23, 2005 5113
SCHEME 3: Degradation Paths in the OH-Initiated Oxidation of Cyclopentanecarbaldehyde
Previous studies suggest that 4-hydroxy-butanal is the major
may also arise from route B through decomposition of the
CHOCOCH(O )(CH2)3OH radical. On the basis of the similarity
of the OH reaction rate coefficients for 3-hydroxy-propanal, 1.99
18,20
•
acyclic product for cyclopentoxy ring opening.
at m/z 71 ([MH + H2O] ) and m/z 89 (MH ) were thus assigned
to 4-hydroxy-butanal, observed with a total carbon yield of 12.2.
Ion signals
+
+
-
11
3
-1 -1 64
-11
× 10
cm molecule s , and propanal, 2.00 × 10
3
-1 -1 72
4
-Hydroxy-butanal may be formed via route A through a 1,6-H
cm molecule s , the OH 4-hydroxy-butanal rate coefficient
-
11
3
shift from a weakly bound H (such as the aldehydic -CHO) or
can be approximated to that of butanal, 2.4 × 10
cm
-
1 -1 72
through a 1,5-H shift followed by elimination of HCO. The
molecule s . Thus, assuming OH rate coefficients of 3.27
5
8,67
-11
3
-1 -1
1
,6-H shift is considered to be less favorable.
However,
× 10 cm molecule
s
for cyclopentanecarbaldehyde (this
2
0
-11
3
-1 -1
Orlando et al. found indirect experimental evidence for the
feasibility of this process. A minor fraction of 4-hydroxy-butanal
work) and 2.4 × 10
cm molecule
s
for 4-hydroxy-
7
2
butanal, a corrected 4-hydroxy-butanal carbon yield of 16.1%

5
114 J. Phys. Chem. A, Vol. 109, No. 23, 2005
D’Anna et al.
is derived at 50% cyclopentanecarbaldehyde conversion. The
corrected 4-hydroxy-butanal yield relative to cyclopentyl radical
The ion signal at m/z 87 may be assigned to 2-hydroxy-4-
+
nitrooxy-butanal ([MH - HNO3] ). Traces of its molecular ion
+
decomposition is 66.1% which is significantly higher than the
have been detected at m/z 150 (MH ). Another candidate is
4% yield estimated by Hanson et al.1 This finding may again
8
butane-1,4-dial (MH ) arising from route B.
+
5
16
be explained by the fact that the SAR method underestimates
the fractional abundance of route A.
As mentioned above, the decomposition via route B was
16
calculated to amount to about 30% using the SAR method.
A
CO and CO2 represent a large portion of the reaction products
with respectively 6.7 and 12.6% of the total carbon yield. CO2
arises primarily from decomposition of cyclopentylcarboxy
radicals and from the secondary OH reaction of 4-hydroxy-
butanal. Part of the CO2 may also be generated through
decomposition of the 5-hydroxy-1-pentanoyl radical, as sug-
gested by Orlando et al.,20 or via route B. CO may arise from
detailed degradation scheme of route B is given in the
Supporting Information (Scheme VIS). As in the case of
cyclobutanecarbaldehyde, no unambiguous experimental evi-
dence for product formation via route B was found. Ion signals
which are expected to be unequivocally attributable to multi-
functional ring opening products from route B (e.g., 2-keto-
hexane-1,6-dial and 2-keto-6-nitrooxy-hexanal) have not been
observed. Other products such as dialdehydes or 4-hydroxy-
butanal are primarily generated via route A, but the contribution
from route B may be significant. As discussed above, a signifi-
cant fraction of the observed CO may arise from route B.
•
decomposition of the CHOC(O )(CH2)3OH radical, as indicated
2
0
by Orlando et al. Minor amounts of CO may also be formed
via route B (see Scheme 3).
Hanson et al.18 observed 2-hydroxypentane-1,5-dial from the
Cl-atom-initiated oxidation of cyclopentane. 2-Hydroxypentane-
It can thus be concluded that the OH-cyclopentanecarbal-
1
9
,5-dial is expected to be predominantly detected at m/z
9 ([MH - H2O] ). In the present study, the reactant cyclo-
dehyde reaction proceeds predominantly via H abstraction.
ald
+
16
The SAR method predicts 70% decomposition via route A,
pentanecarbaldehyde produced an intense ion signal at m/z 99
but the experimental data indicate that this value should be taken
as a lower limit. With an ∼33% NPCCP yield, the formation
of the cyclopentylperoxy radical amounts to g37%. The major
product of cyclopentylperoxy radical decomposition is 4-hy-
droxy-butanal with a corrected molar yield of 24%; minor
products (cyclopentyl nitrate, cyclopentanone, 5-nitrooxy-pen-
tanal, pentane-1,5-dial, and 5-hydroxy-2-nitrooxy-pentanal) ac-
count for a molar yield of ∼21% in total. The CO and CO2
molar yields of 40 and 76%, respectively, are somehow higher
than those expected from their major production channels: 24%
+
(MH ), hindering the detection of 2-hydroxypentane-1,5-dial.
However, cyclopentanecarbaldehyde also formed an ion signal
+
at m/z 81 ([MH - H2O] ) and the m/z 99:m/z 81 signal ratio
did not increase during the reaction with OH radicals. This
finding suggests that 2-hydroxypentane-1,5-dial formation is
only minor.
The ion signal at m/z 85 accounting for a carbon yield of
8
.7% may be assigned to several compounds. The dominant
contribution is most likely due to 5-nitrooxy-pentanal ([MH -
+
•
HNO3] ). Traces of its molecular ion have been detected at m/z
CO yield from CHOC(O )(CH2)3OH radical decomposition and
+
25
1
48 (MH ). Takagi et al. report an overall yield of acyclic
nitrates of 7.5% which are believed to consist mainly of
-nitrooxy-pentanal, 5-hydroxy-2-nitrooxy-pentanal, and 2-hy-
g37% CO2 yield from cyclopentylcarboxy radical decomposi-
tion. As pointed out before, the observed CO and CO2 yields
are affected by a large uncertainty ((30%). Furthermore,
additional production channels for both species have been
indicated above.
5
droxy-4-nitrooxy-butanal (see Scheme 3). For the latter two
nitrates, the carbon yields were e5.3% and e1.6%, respectively;
see below. Hanson et al.18 report a yield of 3% for 5-nitrooxy-
pentanal. A minor fraction of the ion signal at m/z 85 may thus
3.3.4. Cyclohexanecarbaldehyde. Table 6 shows products and
the respective molar and carbon yields for the reaction of
cyclohexanecarbaldehyde with OH radicals. Scheme 4 illustrates
the proposed reaction scheme.
+
arise from cyclopentyl nitrate ([MH - HNO2] ) and cyclopen-
+
tanone (MH ). Previous studies report low cyclopentanone
2
5
20
18
yields: <0.2%, <0.5%, and <5%. 2-Hydroxybutane-1,4-
[(Nitroperoxy)carbonyl]cyclohexane (NPCCH) is the major
+
dial ([MH - H2O] ) may also contribute to the ion signal at
reaction product with a carbon yield of 33.3%. The ion signal
m/z 85.
+
at m/z 145 [MH + H2O - HNO3] and the cyclohexylcarbonyl
The ion signal at m/z 101 accounting for a carbon yield of
fragment ion at m/z 111 have been fully assigned to NPCCH;
the cyclohexyl fragment ion at m/z 83 is partly attributed to
NPCCH. NPCCH identification and quantification is based on
results from the thermal decomposition study and the NPCCH
fragmentation pattern obtained in the cyclohexanecarbaldehyde
5
.3% has been tentatively assigned to 5-hydroxy-2-nitrooxy-
+
pentanal ([MH - HNO3] ). 5-Nitrooxy-pentanal ([MH -
+
+
HNO2] ) and pentane-1,5-dial (MH ) may also be observed at
m/z 101. A low abundance of m/z 57 (a fragment of m/z 101
observed in the pentane-1,5-dial calibration) suggests that the
contribution from pentane-1,5-dial is minor.
+
NO3 experiment.
The second major cyclic compound is cyclohexanone ob-
The ion signal at m/z 75 accounting for a carbon yield of
+
served at m/z 99 (MH ) with a carbon yield of 22.2%. As
discussed below, several species may produce an ion signal at
m/z 99, and thus, the reported carbon yield for cyclohexanone
1
4
.3% is associated with 3-hydroxy-propanal arising from
-hydroxy-butanal decomposition.
The ion signal at m/z 47 accounting for 0.6% carbon yield
16
should be regarded as an upper limit. Using the SAR method,
25
has been assigned to HCOOH. Takagi et al. and Orlando et
the decomposition via route A was calculated to amount to 57%
with 33.3% going into NPCCH and the remaining 23.7% going
into cyclohexyl radicals (see Scheme 4). Product yields can thus
be calculated relative to cyclohexyl radical decomposition and
be compared with literature values reported for cyclohexane
oxidation. The cyclohexanone yield relative to cyclohexyl radical
decomposition is 109.3%, which largely exceeds previously
2
0
al. also observed HCOOH from cyclopentyl radical decom-
position.
+
The ion signal at m/z 31 has been assigned to HCHO (MH ),
1
8
observed with a carbon yield of 1.2%. Hanson et al. and
2
0
Orlando et al. did not observe HCHO in the decomposition
of cyclopentyl radicals, indicating that it may be formed via
route B.
25
18
17
73
20
reported values: 23 ( 13, 28, 32 ( 4, 35 ( 4, 36 ( 6,
2
1
The ion signal at m/z 43 is a fragment ion that cannot be
attributed to a specific compound.
and 39%. This large discrepancy indicates that the SAR
method used underestimates the fractional abundance of route
16

Atmospheric Chemistry of Cycloalkanecarbaldehydes
J. Phys. Chem. A, Vol. 109, No. 23, 2005 5115
TABLE 6: Ion Signals, Molar Yields, and Carbon Yields of the OH-Cyclohexanecarbaldehyde Reaction Products (the
Indicated Errors Are Standard Errors of the Least-Squares Fits)
compound
signal (m/z)
molar yield (%)
C yield (%)
interferences (m/z)
nitroperoxy carbonyl cyclohexane (NPCCH)
cyclohexanone
83, 111, 145
99
33.3
25.9
33.3 ( 1.1
22.2 ( 0.5
6-nitrooxy-hexanal
cyclohexyl nitrate
-hydroxypentane-1,5-dial
3
cyclohexyl nitrate
83, 101
15.4
<8.7
6.9
13.2 ( 1.0
<7.5 ( 0.3
5.9 ( 0.2
3-hydroxy-5-nitrooxy-pentanal (101)
pentane-1,5-dial (101)
4-hydroxy-butanal (71)
3
-hydroxyhexane-1,6-dial
71, 85, 113
97, 115
2
-hydroxybutane-1,4-dial (85)
hexane-1,6-dial
6-hydroxy-3-nitrooxy-hexanal (115)
6
2
-nitrooxy-hexanal (115)
,3-dihydroxypentane-1,5-dial (115)
unidentified
129
127
31
47
87
2.8
1.4
9.1
4.2
1.1
2.8 ( 0.1
1.4 ( 0.1
1.3 ( 0.2
0.6 ( 0.1
0.6 ( 01
88.7
2.5 ( 0.3
5.8 ( 0.2
97.0
2
-keto-7-nitrooxy-heptanal (route B)
nitrooxy-cyclohexanecarbaldehyde
HCHO
HCOOH
2
,4-dihydroxy-butanal
total VOCs
CO
(FT-IR)
(FT-IR)
17.4
40.7
CO
total
2
A. Part of the discrepancy may also be explained by cyclohex-
anone formation via route B (see Scheme 4). HCO elimination
from an acyclic â-carbonyl alkoxy radical has been previously
observed as a dominant reaction pathway.1 The same process
may occur for cyclohexoxy carbonyl radicals. The presence of
interfering species at m/z 99 may also partly explain the
discrepancy in cyclohexanone yields. A fraction of the ion signal
at m/z 99 may derive from 6-nitrooxy-hexanal ([MH -
Some minor acyclic products have also been detected.
However, compound assignment is ambiguous.
Aschmann et al.17 and Hanson et al.18 identified 3-hydroxy-
8
hexane-1,6-dial as a major decomposition product of cyclohex-
oxy radicals applying similar CI methods as those used here.
In both studies, which involved the OH-radical- and Cl-atom-
initiated oxidation of cyclohexane, 3-hydroxyhexane-1,6-dial
+
was predominantly detected at m/z 113 ([MH - H2O] ). In the
+
+
HNO3] ). Traces of the molecular ion (MH ) at m/z 162 have
present study, the reactant cyclohexanecarbaldehyde produced
been observed, suggesting that this species is present. Takagi
+
an intense ion signal at m/z 113 (MH ), hindering the detection
2
5
et al. report an overall acyclic nitrate yield of 11% from
cyclohexane photooxidation. Acyclic nitrates are believed to
consist mainly of 6-nitrooxy-hexanal, 6-hydroxy-3-nitrooxy-
hexanal, and 3-hydroxy-5-nitrooxy-pentanal (see Scheme 4). A
minor (<1%) fraction of the carbon yield at m/z 99 may be
of 3-hydroxyhexane-1,6-dial. However, cyclohexanecarbalde-
+
hyde also formed an ion signal at m/z 95 ([MH - H2O] ) and
the m/z 113:m/z 95 signal ratio remained constant throughout
the reaction with OH radicals. Considering the uncertainty in
this analysis, it can be concluded that no products with a carbon
yield g2% were formed at m/z 113. Additional experiments
were performed under the same experimental conditions using
cyclohexane as a reactant. A reactive intermediate product was
+
attributed to cyclohexyl nitrate ([MH - HNO2] ), if CI
pathways are assumed to be similar to those observed for alkyl
nitrates. 3-Hydroxypentane-1,5-dial may also give rise to a signal
+
at m/z 99 ([MH - H2O] ). This compound may be formed from
observed at m/z 113 with only low yield. In addition to the m/z
the decomposition of 3-hydroxyhexane-1,6-dial (see below).
18
1
13 ion signal, Hanson et al. assigned fragment ion signals at
Another relevant cyclic product is cyclohexyl nitrate observed
m/z 71 and m/z 85 to 3-hydroxyhexane-1,6-dial. In the present
study, the two ion signals account for 2.1 and 3.4% of the carbon
yield, respectively. Increasing m/z 85:m/z 113 and m/z 71:m/z
113 ion signal ratios were observed in the OH-cyclohexane
experiments, suggesting the presence of additional species at
m/z 85 and m/z 71. A potential compound with a m/z 85 signal
+
at m/z 83 ([MH - HNO3] ) and its hydrate at m/z 101 with a
total carbon yield of 13.2%. The NPCCH contribution to the
ion signal at m/z 83 has been subtracted. Again, the nitrate yield
has a higher uncertainty as minor CI pathways have not been
taken into account. A minor fraction of the m/z 101 signal may
be attributable to pentane-1,5-dial and to 3-hydroxy-5-nitrooxy-
+
is 2-hydroxybutane-1,4-dial ([MH - H O] ) which arises from
2
+
+
pentanal ([MH - HNO3] ). Traces of the molecular ion (MH )
at m/z 164 have been observed, suggesting that 3-hydroxy-5-
nitrooxy-pentanal is present. The cyclohexyl nitrate yield relative
to cyclohexyl radical loss is 65%, again significantly higher than
the decomposition of 3-hydroxyhexane-1,6-dial. The ion signal
+
at m/z 71 may derive from 4-hydroxy-butanal ([MH - H2O] ).
The carbon yields of 2.1, 3.4, and e2.0% for m/z 71, m/z 85,
and m/z 113, respectively, result in an upper limit of 7.5% carbon
yield for 3-hydroxyhexane-1,6-dial for the OH-cyclohexan-
ecarbaldehyde reaction. This corresponds to an upper limit of
17
74
21
previously reported values: 16 ( 2, 16 ( 1, 16 ( 4, and
2
0
1
5 ( 4%. This discrepancy may again be explained by the
presence of interfering species at m/z 83 and by the fact that
3
6.7% for the 3-hydroxyhexane-1,6-dial yield relative to cy-
16
the SAR method underestimates the fractional abundance of
route A. Hanson et al.18 report a significantly lower cyclohexyl
nitrate yield of 3.2%, but HNO3 loss in their PTR-MS study
was not considered as a possible CI pathway for cyclohexyl
nitrate.
clohexyl radical decomposition, which is similar to the 39%
18
yield reported by Hanson et al. However, 3-hydroxyhexane-
1
,6-dial loss due to secondary reaction with OH has to be
considered. The OH-cyclohexanecarbaldehyde reaction rate
-11
3
-1 -1
coefficient is 3.75 × 10 cm molecule
s
(this work), and
CO and CO2 account for 2.4 and 5.8% of the carbon yield,
respectively. CO2 arises primarily from the decomposition of
cyclohexylcarboxy radicals. Possible pathways for minor CO
formation via route B are shown in Scheme 4.
the OH-3-hydroxyhexane-1,6-dial rate coefficient is estimated
-
11
3
-1 -1 16,75
to be 8.04 × 10
cm molecule s .
This results in a
corrected 3-hydroxyhexane-1,6-dial yield of 76.7% at 50%
cyclohexanecarbaldehyde conversion. Secondary reactions are

5116 J. Phys. Chem. A, Vol. 109, No. 23, 2005
D’Anna et al.
SCHEME 4: Degradation Paths in the OH-Initiated Oxidation of Cyclohexanecarbaldehyde
1
8
+
insignificant in the study of Hanson et al. Again, the
nitrooxy-hexanal ([MH - HNO3] ), 6-nitrooxy-hexanal ([MH
16
+
discrepancy suggests that the SAR method used underestimates
the fractional abundance of route A.
- HNO2] ), and also 2,3-dihydroxypentane-1,5-dial ([MH -
H2O] ), which is the main decomposition product of 3-hydroxy-
+
+
+
hexane-1,6-dial.
Ion signals at m/z 115 (MH ) and m/z 97 ([MH - H2O] )
with a total carbon yield of 5.9% have been assigned to hexane-
Traces of several other compounds have been detected.
+
1
,6-dial. The given hexane-1,6-dial yield should, however, be
Signals at m/z 31 and m/z 47 are assigned to HCHO (MH )
+
regarded as an upper limit, since other species are expected to
and HCOOH (MH ) with a carbon yield of 1.0 and 0.6%,
contribute significantly to the signal at m/z 115: 6-hydroxy-3-
respectively. The ion signal at m/z 87 accounts for a carbon

Atmospheric Chemistry of Cycloalkanecarbaldehydes
J. Phys. Chem. A, Vol. 109, No. 23, 2005 5117
•
yield of 0.6% and has been tentatively assigned to 2,4-
dihydroxy-butanal ([MH - H2O] ).
The signal at m/z 129 accounting for a carbon yield of 2.8%
corresponds to the unidentified m/z 87, m/z 101, and m/z 115
signals observed for C3-C5 cycloalkanecarbaldehydes and will
be discussed later.
position of CHOCH2CH2O radicals. For larger carbonyl alkoxy
+
radicals, unimolecular decomposition was not observed.
Acyclic and cyclic nitrates have been observed. While it was
difficult to quantify the yield of acyclic nitrates, the yield of
cycloalkyl nitrates was found to increase with the ring size:
cyclopropyl nitrate (<1%), cyclobutyl nitrate (<1%), cyclo-
pentyl nitrate (3.7%), and cyclohexyl nitrate (13.2%). A similar
trend has been observed for n-alkyl nitrates. Cycloketone yields
were also found to increase with ring size, but accurate
quantification was hindered by the presence of interfering
species.
As discussed above, the decomposition via route A was
77
16
calculated to amount to 57% using the SAR method. However,
the observed cyclohexanone and cyclohexyl nitrate yields
relative to cyclohexyl radical decomposition are all well above
1
6
reported literature values, indicating that the SAR method
Ring opening was found to be the only reaction pathway for
the cyclopropoxy and cyclobutoxy radicals formed during OH
degradation of the respective cycloalkanecarbaldehydes. For the
cyclopentoxy radical, ring opening is the dominant decomposi-
underestimates the fractional abundance of route A. Quantitative
estimates of the cyclohexyl radical decomposition products are
thus difficult, also because unambiguous product identification
is not possible. It can just be concluded that a relevant fraction
of the OH-cyclohexanecarbaldehyde reaction proceeds via Hald
abstraction, giving NPCCH as the major reaction product with
a yield of ∼33%. The major products arising from cyclohexy-
lperoxy radicals are cyclohexyl nitrate and cyclohexanone. The
minor products of route A are discussed in detail above.
A detailed degradation scheme of route B is given in the
Supporting Information (Scheme VIIS). Most of the possible
reaction products of route B are identical to those formed along
route A: 2- and 3-hydroxy-pentane-1,5-dial, hexane-1,6-dial,
and pentane-1,5-dial. The ion signal at m/z 127 accounting for
20
tion pathway, as already reported by Orlando et al., Hanson
1
8
25
et al., and Takagi et al. For the cyclohexoxy radical,
competition between ring opening and the formation of cyclo-
1
7,18,20
hexanone was observed.
The reaction of OH radicals with cycloalkanecarb-
aldehydes may occur via H abstraction from the aldehydic
group (route A) or to a lesser extent via H abstraction from the
ring (route B). Using the SAR developed by Kwok and
Atkinson, the fractional contribution of route B was calculated
to be ∼1, 11, 30, and 43% for cyclopropanecarbaldehyde,
cyclobutanecarbaldehyde, cyclopentanecarbaldehyde, and cy-
clohexanecarbaldehyde, respectively. The experimental data did
not allow evaluation of the relative contribution of route B.
However, in the C5 and C6 cycloalkanecarbaldehyde experi-
ments, observed product yields relative to cycloalkyl radical
decomposition were well above reported literature values,
16
1.4% of the carbon yield seems to be the only signal unequivo-
cally attributable to route B. It has been tentatively assigned to
+
2
-keto-7-nitrooxy-heptanal ([MH - HNO3] ) or isomeric spe-
cies with the sum formula C12H11O5N. This is consistent with
the finding that no m/z 127 ion signal was observed in the OH-
cyclohexane experiment. Signal traces at m/z 174 may derive
from nitrooxy-cyclohexanecarbaldehyde. HNO2 loss from this
species would also produce an ion signal at m/z 127.
16
indicating that the SAR method overestimates the fractional
abundance of route B.
For cyclopropanecarbaldehyde and cyclobutanecarbaldehyde,
the individual contributions of characteristic reaction products
arising from route B (e.g., keto-dialdehydes or keto-nitrooxy-
aldehydes) are expected to be below the 1% signal-based yield
limit set for inclusion in the data analysis. An unambiguous
identification of such low-yield products goes beyond the
analytical capabilities of the PTR-MS technique.
The formation of large multifunctional compounds in the
reaction of OH radicals with cyclohexanecarbaldehyde, the
largest of the studied aldehydes, may lead to gas-to-particle
conversion. Particle formation was monitored using a differential
mobility particle analyzer. The aerosol yield was found to be
insignificant under the given experimental conditions.
For cyclopentanecarbaldehyde and cyclohexanecarbaldehyde,
the fractional contribution of route B is expected to be
substantial. However, the same product ion signals or even the
same products may be formed via both reaction routes. The
only product ion signal clearly attributable to route B was
observed in the cyclohexanecarbaldehyde experiment. The ion
signal at m/z 127 accounted for 1.4% carbon yield and has been
tentatively assigned to 2-keto-7-nitrooxy-heptanal.
The origin of the signal series m/z 87, m/z 101, m/z 115, and
m/z 129 observed in the oxidation of cyclopropanecarbaldehyde,
cyclobutanecarbaldehyde, cyclopentanecarbaldehyde, and cy-
clohexanecarbaldehyde has not been identified. Their contribu-
tion in terms of carbon yield is steadily increasing from 0.9%
for cyclopropanecarbaldehyde to 2.8% for cyclohexanecarbal-
dehyde. Thermal decomposition studies indicate that the signals
cannot be associated with NPCCAs or peroxyalkyl nitrate
species. However, a rapid decay was observed when additional
NO was added at the end of the experiment and UV lights were
off. Such decay would be expected only for NPCCAs or
peroxyalkyl nitrate species. The identity of these signals remains
unknown.
Conclusions
The rate coefficients for the gas phase reaction of NO3 and
OH radicals with C3-C6 cycloalkanecarbaldehydes have been
measured for atmospheric conditions. Quantitative product
analyses of the OH-cycloalkanecarbaldehyde reactions were
carried out using FT-IR spectroscopy and PTR-MS. The
measured carbon balance was in the range 89-97%.
The identified reaction products include mainly NPCCAs,
followed by cycloalkyl nitrates and cycloketones. Ring opening
produces carbonyl alkoxy radicals, CHO(CH2)nO , which de-
•
pending on the number of carbons of the aliphatic chain undergo
unimolecular decomposition (C-C fragmentation, observed only
for C3), reaction with O2, or isomerization (intramolecular H
shift). Isomerization prevails when the number of carbon atoms
5
1,54,58-60,65,66,71,76
is g4.
This is confirmed by the high carbon
yield of 3-hydroxy-propanal in the degradation of cyclobutan-
ecarbaldehyde. Unimolecular decomposition through -CHO
elimination was found to be the dominant reaction pathway for
•
â-carbonyl alkoxy radicals, CHOCH(O )CH2CH2CH2OH.
Remarkably large yields of CO and HCHO have been found
in the decomposition of cyclopropanecarbaldehyde and are only
partly attributable to secondary OH chemistry. CO and HCHO
formation from NPCCPr decomposition with the UV lights off
provided experimental evidence for the unimolecular decom-
This work demonstrates the applicability of PTR-MS for
quantitative detection of volatile organic reaction products which
cannot be achieved using other CI methods (e.g., API-MS) or
traditional analytical techniques unless authentic calibration
standards are used. However, PTR-MS is not a truly species-

5118 J. Phys. Chem. A, Vol. 109, No. 23, 2005
D’Anna et al.
specific method of analysis and unambiguous compound
identification is difficult.
(31) Hewitt, C. N.; Hayward, S.; Tani, A. J. EnViron. Monit. 2003, 5,
1.
(32) Jensen, N. R. V., J.; Karl, T.; Prazeller, P.; Wisthaler, A.; Larsen,
B. R.; Seufert, G.; Hjorth, J. Contributions: 1st International Conference
on Proton-Transfer Reaction Mass Spectrometry and Its Applications, 2003.
(33) De Gouw, J.; Warneke, C.; Karl, T.; Eerdekens, G.; van der Veen,
C.; Fall, R. Int. J. Mass Spectrom. 2003, 223-224, 365.
Acknowledgment. A.W. thanks the Verein zur F o¨ rderung
der wissenschaftlichen Ausbildung und T a¨ tigkeit von S u¨ dtirolern
an der Landesuniversit a¨ t Innsbruck for postdoctoral support. The
authors thank the anonymous referees for their extensive
comments and valuable suggestions.
(
34) Kwok, E. S. C.; Atkinson, R.; Arey, J. EnViron. Sci. Technol. 1995b,
2
9, 2467.
(35) Wisthaler, A. Unpublished results.
(36) Kames, J.; Schurath, U.; Flocke, F.; Volz-Thomas, A. J. Atmos.
Chem. 1993, 16, 349.
Supporting Information Available: Possible degradation
paths in the OH-initiated oxidation of cyclobutanecarbaldehyde,
cyclopentanecarbaldehyde, and cyclohexanecarbaldehyde; plot
of the time evolution for reagent and major products in the
cyclobutanecarbaldehyde-OH radical reaction; and PTR-MS
product ion signals, relative signal abundances, and sensitivities
of target compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
(
37) Hansel, A.; Wisthaler, A. Geophys. Res. Lett. 2000, 27, 895.
(38) Hatakeyama, S.; Tanonaka, T.; Weng, J.; Bandow, H.; Takagi, H.;
Akimoto, H. EnViron. Sci. Technol. 1985, 19, 935.
39) Hansel, A.; Singer, W.; Wisthaler, A.; Schwarzmann, M.; Lindinger,
W. Int. J. Mass Spectrom. Ion Processes 1997, 167/168, 697.
40) Noyes, W. A. Org. Synth. 1943, Coll. Vol. 2, 108.
(
(
(41) Taylor, W. D.; Allston, T. D.; Moscato, M. J.; Fazekas, G. B.;
Kozlowski, R.; Takacs, G. A. Int. J. Chem. Kinet. 1980, 12, 231.
(
(
(
42) Corey, E. J.; Suggs, J. W. Tetrahedron Lett. 1975, 2647.
43) Johnson, M. R.; Rickborn, B. Org. Synth. 1971, 51, 11.
44) York, D. Can. J. Phys 1966, 44, 1079.
References and Notes
(45) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr,
J. A.; Rossi, M. J.; Troe, J. J. Phys. Chem. Ref. Data 1999, 28, 191.
(
1) Guenther, A.; Hewitt, C. N.; Erickson, D.; Fall, R.; Geron, C.;
(46) Prinn, R. G.; Huang, J.; Weiss, R. F.; Cunnold, D. M.; Fraser, P.
Graedel, T.; Harley, P.; Klinger, L.; Lerdau, M.; McKay, W. A.; Pierce,
T.; Scholes, B.; Steinebrecher, R.; Tallamaraju, R.; Taylor, J.; Zimmermann,
P. J. Geophys. Res., [Atmos.] 1995, 100, 8873.
J.; Simmonds, P. G.; McCulloch, A.; Harth, C.; Salameh, P.; O’Doherty,
S.; Wang, R. H. J.; Porter, L.; Miller, B. R. Science 2001, 292, 1882.
(
47) Allan, B. J.; Plane, J. M. C.; Coe, H.; Shillito, J. J. Geophys. Res.,
(
(
2) Atkinson, R.; Aschmann, S. M. J. Atmos. Chem. 1993, 16, 337.
3) Bonn, B.; Schuster, G.; Moortgat, G. K. J. Phys. Chem. A 2002,
[
Atmos.] 2002, 107, ACH11/1.
(
48) Atkinson, R. J. Phys. Chem. Ref. Data 1991, 20, 459.
1
3
06, 2869.
(49) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr,
(
4) Calogirou, A.; Larsen, B. R.; Kotzias, D. Atmos. EnViron. 1999,
J. A.; Rossi, M. J.; Troe, J. J. Phys. Chem. Ref. Data 1997, 26, 521.
50) Ferenac, M. A.; Davis, A. J.; Holloway, A. S.; Dibble, T. S. J.
Phys. Chem. A 2003, 107, 63.
3, 1423.
(
(
5) Griffin, R. J.; Cocker, D. R., III.; Flagan, R. C.; Seinfeld, J. H. J.
Geophys. Res., [Atmos.] 1999, 104, 3555.
6) Hakola, H.; Arey, J.; Aschmann, S. M.; Atkinson, R. J. Atmos.
Chem. 1994, 18, 75.
7) Hoffmann, T.; Odum, J. R.; Bowman, F.; Collins, D.; Klockow,
D.; Flagan, R. C.; Seinfeld, J. H. J. Atmos. Chem. 1997, 26, 189.
8) Larsen, B. R.; Di Bella, D.; Glasius, M.; Winterhalter, R.; Jensen,
N. R.; Hjorth, J. J. Atmos. Chem. 2001, 38, 231.
(
51) Devolder, P. J. Photochem. Photobiol., A 2003, 157, 137.
(
(52) Mereau, R.; Rayez, M.-T.; Caralp, F.; Rayez, J.-C. Phys. Chem.
Chem. Phys. 2000, 2, 3765.
53) Meunier, N.; Doussin, J. F.; Chevallier, E.; Durand-Jolibois, R.;
Picquet-Varrault, B.; Carlier, P. Phys. Chem. Chem. Phys. 2003, 5, 4834.
54) Baldwin, A. C.; Barker, J. R.; Golden, D. M.; Hendry, D. G. J.
Phys. Chem. 1977, 81, 2483.
55) Orlando, J. J.; Tyndall, G. S.; Wallington, T. J. Chem. ReV. 2003,
03, 4657.
(
(
(
(
(
9) Jang, M.; Kamens, R. M. Atmos. EnViron. 1998, 33, 459.
(
(
10) Novakov, P., Penner, J. Nature 1993, 265, 823.
1
(11) Yu, J.; Cocker, D. R., III; Griffin, R. J.; Flagan, R. C.; Seinfeld, J.
(
(
56) Somnitz, H.; Zellner, R. Phys. Chem. Chem. Phys. 2000, 2, 1907.
57) Orlando, J. J.; Tyndall, G. S.; Vereecken, L.; Peeters, J. J. Phys.
H. J. Atmos. Chem. 1999, 34, 207.
12) Tyndall, G. S.; Orlando, J. J.; Wallington, T. J.; Hurley, M. D.;
Goto, M.; Kawasaki, M. Phys. Chem. Chem. Phys. 2002, 4, 2189.
13) Tuazon, E. C.; Aschmann, S. M.; Nguyen, M. V.; Atkinson, R.
Int. J. Chem. Kinet. 2003, 35, 415.
14) D’Anna, B.; Bakken, V.; Are Beukes, J.; Nielsen, C. J.; Brudnik,
K.; Jodkowski, J. T. Phys. Chem. Chem. Phys. 2003, 5, 1790.
15) Cameron, M.; Sivakumaran, V.; Dillon, T. J.; Crowley, J. N. Phys.
Chem. Chem. Phys. 2002, 4, 3628.
(
Chem. A 2000b, 104, 11578.
58) Jungkamp, T. P. W.; Smith, J. N.; Seinfeld, J. H. J. Phys. Chem.
A 1997, 101, 4392.
59) Hein, H.; Hoffmann, A.; Zellner, R. Phys. Chem. Chem. Phys. 1999,
1, 3743.
(
(
(
(
(
60) Eberhard, J.; Mueller, C.; Stocker, D. W.; Kerr, J. A. EnViron. Sci.
(
Technol. 1995, 29, 232.
(61) Orlando, J. J.; Tyndall, G. S.; Bilde, M.; Ferronato, C.; Wallington,
T. J.; Vereecken, L.; Peeters, J. J. Phys. Chem. A 1998, 102, 8116.
(62) Vereecken, L.; Peeters, J. J. Phys. Chem. A 1999, 103, 1768.
(63) Niki, H.; Maker, P. D.; Savage, C. M.; Breitenbach, L. P. Chem.
Phys. Lett. 1981, 80, 499.
(64) Baker, J.; Arey, J.; Atkinson, R. J. Phys. Chem. A 2004, 108, 7032.
(65) Johnson, D.; Cassanelli, P.; Cox, R. A. J. Phys. Chem. A 2004,
108, 519.
(66) Mereau, R.; Rayez, M.-T.; Caralp, F.; Rayez, J.-C. Phys. Chem.
Chem. Phys. 2003, 5, 4828.
(67) Houk, K. N.; Tucker, J. A.; Dorigo, A. E. Acc. Chem. Res. 1990,
23, 107.
(68) Francisco, C. G.; Freire, R.; Herrera, A. J.; Perez-Martin, I.; Suarez,
E. Org. Lett. 2002, 4, 1959.
(69) Henon, E.; Bohr, F.; Sokolowski-Gomez, N.; Caralp, F. Phys. Chem.
Chem. Phys. 2003, 5, 5431.
(70) Peeters, J.; Fantechi, G.; Vereecken, L. J. Atmos. Chem. 2004, 48,
59.
(71) Atkinson, R.; Arey, J. Chem. ReV. 2003, 103, 4605.
(72) IUPAC. http://www.iupac-kinetic.ch.cam.ac.uk/.
(73) Atkinson, R.; Aschmann, S. M.; Arey, J.; Shorees, B. J. Geophys.
Res., [Atmos.] 1992, 97, 6065.
(
16) Kwok, E. S. C.; Atkinson, R. Atmos. EnViron. 1995a, 29, 1685.
(17) Aschmann, S. M.; Chew, A. A.; Arey, J.; Atkinson, R. J. Phys.
Chem. A 1997, 101, 8042.
18) Hanson, D.; Orlando, J.; Noziere, B.; Kosciuch, E. Int. J. Mass
Spectrom. 2004, 239, 147.
(
(
(
19) McEwen, C. N.; Druliner, J. D. J. Am. Chem. Soc. 1988, 110, 8579.
20) Orlando, J. J.; Iraci, L. T.; Tyndall, G. S. J. Phys. Chem. A 2000a,
1
04, 5072.
21) Platz, J.; Sehested, J.; Nielsen, O. J.; Wallington, T. J. J. Phys.
Chem. A 1999, 103, 2688.
22) Rowley, D. M.; Lesclaux, R.; Lightfoot, P. D.; Noziere, B.;
Wallington, T. J.; Hurley, M. D. J. Phys. Chem. 1992, 96, 4889.
23) Rowley, D. M.; Lightfoot, P. D.; Lesclaux, R.; Wallington, T. J. J.
Chem. Soc., Faraday Trans. 1991, 87, 3221.
24) Rowley, D. M.; Lightfoot, P. D.; Lesclaux, R.; Wallington, T. J. J.
Chem. Soc., Faraday Trans. 1992, 88, 1369.
25) Takagi, H.; Washida, N.; Bandow, H.; Akimoto, H.; Okuda, M. J.
Phys. Chem. 1981, 85, 2701.
26) D’Anna, B.; Andresen, O.; Gefen, Z.; Nielsen, C. J. Phys. Chem.
Chem. Phys. 2001, 3, 3057.
27) Ballesteros, B.; Jensen, N. R.; Hjorth, J. J. Atmos. Chem. 2002,
3, 135.
(
(
(
(
(
(
(
4
(74) Atkinson, R.; Aschmann, S. M.; Carter, W. P. L.; Winer, A. M.;
Pitts, J. N., Jr. Int. J. Chem. Kinet. 1984, 16, 1085.
(75) Bradley, W. R.; Wyatt, S. E.; Wells, J. R.; Henley, M. V.; Graziano,
G. M. Int. J. Chem. Kinet. 2001, 33, 108.
(76) Vereecken, L.; Peeters, J. J. Chem. Phys. 2003, 119, 5159.
(77) Arey, J.; Aschmann, S. M.; Kwok, E. S. C.; Atkinson, R. J. Phys.
Chem. A 2001, 105, 1020.
(
(
28) Griffith, D. W. T. Appl. Spectrosc. 1996, 50, 59.
29) Rothman, L. S.; Rinsland, C. P.; Goldman, A.; Massie, S. T.;
Edwards, D. P.; Flaud, J. M.; Perrin, A.; Camy-peyret, C.; Dana, V.; Mandin,
J. Y.; Schroeder, J.; McCann, A.; Gamache, R. R.; Wattson, R. B.; Yoshino,
K.; Chance, K. V.; Jucks, K. W.; Brown, L. R.; Nemtchinov, V.; Varanasi,
P. J. Quant. Spectrosc. Radiat. Transfer 1998, 60, 665.
(30) Lindinger, W.; Jordan, A. Chem. Soc. ReV. 1998, 27, 347.
(78) Atkinson, R. J. Phys. Chem. Ref. Data 1997, 26, 215.