Atmospheric chemistry of C2F5CHO: reaction with Cl atoms and OH radicals, IR spectrum of C2F5C(O)O2NO2

Article abstract of DOI:10.1016/j.cplett.2003.08.004

Smog chamber/FTIR techniques were used to measure k(Cl + C2F5CHO) = (1.96+/-0.28) x 10^-12 and k(OH + C2F5CHO) = (5.26+/-0.80) x 10^-13 cm³ molecule^-1 s^-1 in 700 Torr of N2 or air at 296+/-2 K. The Cl i

Full text of DOI:10.1016/j.cplett.2003.08.004

Chemical Physics Letters 379 (2003) 28–36
www.elsevier.com/locate/cplett
Atmospheric chemistry of C2F5CHO: reaction with Cl
atoms and OH radicals, IR spectrum of C2F5C(O)O2NO2
a
a,
a
M.P. Sulbaek Andersen a,c, M.D. Hurley , T.J. Wallington *, J.C. Ball ,
b
J.W. Martin , D.A. Ellis , S.A. Mabury , O.J. Nielsen
b
b
c
a
Ford Motor Company, P.O. Box 2053, Dearborn, MI 48121-2053, USA
b
Department of Chemistry, 80 St. George St., University of Toronto, Toronto, Ont., Canada M5S 3H6
Department of Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen, Denmark
c
Received 16 July 2003; in final form 6 August 2003
Published online: 3 September 2003
Abstract
Smog chamber/FTIR techniques were used to measure k(Cl + C2F5CHO) ¼ (1.96 ꢀ 0.28) ꢁ 10ꢂ12 and
k(OH + C2F5CHO) ¼ (5.26 ꢀ 0.80) ꢁ 10ꢂ13 cm3 moleculeꢂ1 sꢂ1 in 700 Torr of N2 or air at 296 ꢀ 2 K. The Cl initiated
oxidation of C2F5CHO in the presence of NO in air diluent gave COF2, CF3ONO2, and C2F5C(O)O2NO2 in molar
yields of 180 ꢀ 19%, 6.1 ꢀ 0.6%, and 1.0 ꢀ 0.3%, respectively. The IR spectrum for C2F5C(O)O2NO2 is reported. Re-
sults are discussed with respect to the potential for the atmospheric degradation of fluorinated aldehydes, CnF2nþ1CHO,
to contribute to the observed environmental burden of fluorinated carboxylic acids, CnF2nþ1CðOÞOH.
Ó 2003 Elsevier B.V. All rights reserved.
1. Introduction
aldehydes are formed during the atmospheric ox-
idation of HFCs, HCFCs, and fluorinated alco-
hols [2,3]. For example, C2F5CHO has been
identified as the primary oxidation product of
pentafluoropropanol (C2F5CH2OH) [4] and is a
likely oxidation product of C2F5CH3 (HFC-245cb)
and C2F5CH2Cl (HCFC-235ca) [2].
Long-chain perfluoroalkyl carboxylic acids
(PFCAs, CnF2nþ1COOH, where x ¼ 6–12) have
recently been observed in fish [5,6] and mammals
[7] in a variety of locations around the world.
PFCAs are not generally used directly in consumer
or industrial materials, other than in aqueous film
forming foams or as polymerization aids in
fluoropolymer manufacture [8]. Thermolysis of
fluoropolymers produces PFCAs, however the
Concerns regarding the environmental impact
of chlorofluorocarbon (CFC) release into the at-
mosphere have led to the replacement of CFCs in
industrial processes and consumer products. Hy-
drofluorocarbons (HFCs) and hydrochlorofluoro-
carbons (HCFCs) are two classes of compounds
which have gained widespread use as CFC sub-
stitutes. Polyfluorinated alcohols are currently
used in a variety of industrial products (paints,
coatings, polymers, adhesives, waxes, polishes,
electronic materials, and caulks [1]). Perfluorinated
* Corresponding author. Fax: +3133227044.
E-mail address: twalling@ford.com (T.J. Wallington).
0009-2614/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2003.08.004

M.P. Sulbaek Andersen et al. / Chemical Physics Letters 379 (2003) 28–36
29
magnitude of this source appears insufficient to ac-
count for the observed global environmental bur-
den of these compounds [9]. The PFCAs observed
in the environment are presumably the degrada-
tion products of precursor chemicals. However,
the identity of the percursors and mechanism by
which they are converted into PFCAs are unknown
at the present time.
perfluoropropionaldehyde hydrate (C2F5CH(OH)2)
into a heated flask (ꢃ60 °C) containing 25 g P2O5.
The mixture was heated while a slow (ꢃ20 cm3
minꢂ1) flow of nitrogen gas was passed across the
top of the hydrate/aldehyde solution and then
through a liquid nitrogen trap. C2F5CHO was
collected in the trap. IR analysis revealed the
presence of a small amount of C2H4 impurity in
the crude sample of C2F5CHO collected in the
trap. This is presumably attributable to ethyl
hemiacetal impurities in the hydrate sample, which
upon dehydration with P2O5 yields the aldehyde
and ethene. By freezing the sample at liquid ni-
trogen temperature, pumping on the sample and
then discarding the first fraction that evaporates as
the sample was allowed to warm to room tem-
perature we were able to remove all traces of C2H4
impurity.
Perfluorinated aldehydes of the general formula
CnF2nþ1CHO have a molecular formula which is
similar to that of perfluoroalkyl carboxylic acids.
Atmospheric oxidation of perfluorinated alde-
hydes offers a potential route to the formation of
PFCAs. Unfortunately, there is little information
available in the literature concerning the atmo-
spheric chemistry and degradation mechanism of
perfluorinated aldehydes. At the present time it is
not possible to assess the contribution, if any, of
the atmospheric oxidation of perfluorinated alde-
hydes to the observed budget of PFCAs. To im-
prove our understanding of the atmospheric
chemistry, and hence environmental impact, of
fluorinated organics in general and specifically the
ability of perfluorinated aldehydes to contribute to
the observed burden of PFCAs we have conducted
a study of the oxidation mechanism of C2F5CHO.
Chemical intuition, combined with our recent ex-
perience with long-chain fluorinated alcohols [10]
and acids [11] suggests that C2F5CHO will serve as
a useful model with which to understand the
degradation mechanism of longer chain perfluo-
roaldehydes.
2.2. Kinetic and mechanistic experiments
The apparatus and experimental techniques
used in this work have been described elsewhere
[12,13]. Experiments were performed in a 140-liter
Pyrex reactor interfaced to a Mattson Sirus 100
FTIR spectrometer. The reactor was surrounded
by 22 fluorescent blacklamps (GE F15T8-BL)
which were used to photochemically initiate the
experiments. Chlorine atoms were produced by
photolysis of molecular chlorine
Cl2 þ hm ! Cl þ Cl ð1Þ
OH radicals were produced by photolysis of
We have used smog chamber FTIR techniques
to study the kinetics of reactions of Cl atoms and
OH radicals with C2F5CHO, oxidation products
of Cl initiated oxidation of C2F5CHO, and IR
spectrum of C2F5C(O)O2NO2. Results are re-
ported herein and discussed with respect to the
atmospheric chemistry and environmental impact
of fluorinated organic compounds.
CH3ONO in the presence of NO in air:
CH3ONO þ hm ! CH3O þ NO
CH3O þ O2 ! HO2 þ HCHO
HO2 þ NO ! OH þ NO2
ð2Þ
ð3Þ
ð4Þ
Reactant and product concentrations were
monitored using in situ Fourier transform infrared
spectroscopy. IR spectra were derived from 32
coadded interferograms with a spectral resolution
of 0.25 cmꢂ1 and an analytical path length of
27.1 m. Calibrated reference spectra were acquired
by expanding known volumes of reference mate-
rials into the chamber. Analysis of the IR spec-
tra was achieved through a process of spectral
2. Experimental
2.1. Synthesis of C2F5CHO
Perfluoropropionaldehyde, C2F5CHO, was syn-
thesized by the drop wise addition of 10 g of

30
M.P. Sulbaek Andersen et al. / Chemical Physics Letters 379 (2003) 28–36
stripping in which small fractions of the reference
spectrum were subtracted incrementally from the
sample spectrum.
Experiments were performed at 296 ꢀ 2 K in
700 Torr of air diluent. All reactants were obtained
from commercial sources at purities >99%. Ultra-
high purity nitrogen and air diluent gases were
used as received. In smog chamber experiments
unwanted loss of reactants and products via pho-
tolysis and heterogeneous reactions has to be
considered. Control experiments were performed
in which product mixtures obtained after UV ir-
radiation of C2F5CHO/reference/CH3ONO/NO/
air mixtures were allowed to stand in the dark in
the chamber for 30 min. There was no observable
(<2%) loss of reactants or products. Samples of
C2F5CHO were subjected to UV irradiation in
700 Torr of air diluent for 15 min without any
observable loss (<1%), suggesting that photolysis
of C2F5CHO is not a significant complication in
the present work.
Fig. 1. Decay of C2F5CHO versus C2H5Cl (squares) or CH3Cl
(circles) in the presence of Cl atoms in 700 Torr of N2 diluent at
296 ꢀ 2 K.
Consistent results were obtained using both
references. We choose to cite a final value for k5,
which is the average of those, determined using
the two different reference compounds together
with error limits which encompass the extremes
of the individual determinations. Hence, k5 ¼
3. Results and discussion
3.1. Relative rate study of the Cl + C2F5CHO
reaction in 700 Torr of air
The kinetics of reaction (5) were measured rel-
ative to reactions (6) and (7):
ð1:96 ꢀ 0:28Þ ꢁ 10ꢂ12 cm3 moleculeꢂ1 sꢂ1
.
While there are no previous determinations of
k5, Wallington and Hurley [15] have reported
k(Cl + CF3CHO) ¼ (1.8 ꢀ 0.4) ꢁ 10ꢂ12 cm3 mole-
Cl þ C2F5CHO ! products
Cl þ C2H5Cl ! products
Cl þ CH3Cl ! products
ð5Þ
ð6Þ
ð7Þ
culeꢂ1
s
ꢂ1. This is indistinguishable from the
value of k(Cl + C2F5CHO) reported herein. In
contrast, Scollard et al. [16] have reported
k(Cl + CF3CHO) ¼ (2.70 ꢀ 0.11) ꢁ 10ꢂ12 cm3 mol-
eculeꢂ1 sꢂ1 which is substantially greater than the
value of k(Cl + C2F5CHO) measured in the present
work. It is of interest to note that in all cases
for which data are available (CF3H, C2F5H,
CF3CH2OH, C2F5CH2OH, CF3C(O)OH, and
C2F5C(O)OH [2,11]) the rate of H-atom abstrac-
tion by Cl atoms and OH radicals from CF3X
is equal to, or slower than, that from C2F5X,
where X is a hydrogen containing substituent. The
value of k(Cl + C2F5CHO) ¼ (1.96 ꢀ 0.28) ꢁ 10ꢂ12
measured in the present work is thus not con-
sistent with expectations based upon the value of
Reaction mixtures consisted of 5.9–18 mTorr of
C2F5CHO, 140–241 mTorr of Cl2, and 9.3–29
mTorr of either C2H5Cl or CH3Cl in 700 Torr of
N2. The observed loss of C2F5CHO versus that of
the reference compounds in the presence of Cl
atoms is shown in Fig. 1. Linear least-squares
analysis of the data in Fig. 1 gives k5=k6 ¼
0:258 ꢀ 0:021 and k5=k7 ¼ 3:86 ꢀ 0:28, quoted
uncertainties are two standard deviations from the
linear regressions. Using k6 ¼ 8:04 ꢁ 10ꢂ12 [14] and
k7 ¼ 4:8 ꢁ 10ꢂ13 [2] we derive k5 ¼ ð2:07 ꢀ 0:17Þ ꢁ
10ꢂ12 and (1.85 ꢀ 0.11) ꢁ 10ꢂ12 cm3 moleculeꢂ1 sꢂ1
.

M.P. Sulbaek Andersen et al. / Chemical Physics Letters 379 (2003) 28–36
31
k(Cl + CF3CHO) value reported by Scollard et al.
[16]. Wallington and Hurley [15] have argued
that loss of CF3CHO via reaction with CF3O
radicals caused Scollard et al. to overestimate
k(Cl + CF3CHO). The results from the present
work lend support to, but do not prove, this
argument.
strength upon fluorination (fluorination typically
leads to increased C–H bond strengths on adjacent
carbon atoms [18]).
3.2. Product study of the Cl + C2F5CHO reaction in
the presence of NOx in 700 Torr of air
Finally, it is worth noting that CF3CHO and
C2F5CHO are factors of 40 and 70 less reactive
towards Cl atoms than their non-fluorinated
counterparts CH3CHO and C2H5CHO [17]. Re-
action of Cl atoms with CH3CHO proceeds es-
sentially exclusively (>99%) via abstraction of the
aldehydic hydrogen [17]. Abstraction of the alde-
hydic hydrogen is also important in the reaction of
Cl atoms with C2H5CHO. It is clear that CF3 and
C2F5 groups have a strong deactivating effect on
the aldehydic functional group. This effect pre-
sumably reflects an increase in the C(O)–H bond
To investigate the mechanism of Cl atom initi-
ated oxidation of C2F5CHO in the presence of
NOx, mixtures containing 9 mTorr C2F5CHO, 148
mTorr Cl2 and 53 mTorr NO in 700 Torr of air
diluent were introduced into the reaction chamber
and irradiated using the UV blacklamps. Fig. 2
shows spectra acquired before (A) and after (B) a
165 s irradiation of the gas mixture described
above. The consumption of C2F5CHO was 36%.
Comparison of the IR features in Fig. 2B with the
reference spectra of COF2 and ClNO given in Figs.
2C and D shows the formation of these species.
Fig. 2. IR spectra obtained before (A) and after (B) a 110 s irradiation of a mixture of 9.3 mTorr C2F5CHO, 148 mTorr Cl2 and 53
mTorr NO in 700 Torr of air. The consumption of C2F5CHO was 36%. (C and D) Reference spectra of COF2 and ClNO, respectively.

32
M.P. Sulbaek Andersen et al. / Chemical Physics Letters 379 (2003) 28–36
C2F5CðOÞ þ O2 þ M ! C2F5CðOÞO2 þ M
ð8Þ
Product features attributable to CF3ONO2 and
C2F5C(O)O2NO2 were also detected.
Fig. 3 shows the observed formation of COF2,
CF3ONO2, and C2F5C(O)O2NO2 versus loss of
C2F5CHO following UV irradiation of C2F5CHO/
Cl2/NO mixtures in 700 Torr of air diluent. The
straight lines in Fig. 3 are linear least-squares
fits which give the following molar yields:
COF2, (180 ꢀ 20)%; CF3ONO2, (6.1 ꢀ 0.6)%; and
C2F5C(O)O2NO2, (1.0 ꢀ 0.3)%. Quoted errors are
two standard deviations from the regression
analysis. We estimate that potential systematic
errors associated with uncertainties in calibrations
of the reference spectra contribute an additional
10% uncertainty. Within the experimental
uncertainties, the observed formation of COF2,
CF3ONO2, and C2F5C(O)O2NO2 accounts for
100% of the loss of C2F5CHO.
C2F5CðOÞO2 þ NO ! C2F5CðOÞO þ NO2
C2F5CðOÞO ! C2F5 þ CO2
ð9Þ
ð10Þ
ð11Þ
ð12Þ
ð13Þ
ð14Þ
ð15Þ
ð16Þ
C2F5 þ O2 ! C2F5O2
C2F5O2 þ NO ! C2F5O þ NO2
C2F5O þ M ! CF3 þ COF2 þ M
CF3 þ O2 þ M ! CF3O2 þ M
CF3O2 þ NO ! CF3O þ NO2
CF3O þ NO ! COF2 þ FNO
NO2 is formed during the reaction sequence given
above. In the presence of NO2, C2F5C(O)O2NO2,
and CF3ONO2 will be formed via reactions (17)
and (18).
The simplest explanation for the observed
conversion of C2F5CHO into COF2 is that reac-
tion (5) is followed by (8)–(16):
C2F5CðOÞO2 þNO2 þM!C2F5CðOÞO2NO2 ð17Þ
C2F5CHO þ Cl ! C2F5CðOÞ þ HCl
ð5Þ
CF3O þ NO2 þ M ! CF3ONO2 þ M
ð18Þ
Evidence for the formation of C2F5C(O)OH
was sought in the product spectra but not found.
An upper limit for the molar formation of
C2F5C(O)OH of 2% was established.
3.3. Infrared spectrum of C2F5C(O)O2NO2
To record the IR spectrum of C2F5C(O)O2NO2
a mixture of 8.52 mTorr C2F5CHO, 228 mTorr Cl2
and 8.53 mTorr NO2 in 700 Torr of O2 diluent was
subjected to UV irradiation. The reaction of Cl
atoms with C2F5CHO in the presence of O2 leads
to the formation of C2F5C(O)O2 radicals. By
analogy to the behavior of other acetyl peroxy
radicals, it is expected that C2F5C(O)O2 radicals
will react rapidly with NO2 to give the peroxyni-
trate C2F5C(O)O2NO2
C2F5CHO þ Cl ! C2F5CðOÞ þ HCl
C2F5CðOÞ þ O2 ! C2F5CðOÞO2
ð5Þ
ð8Þ
Fig. 3. Yield of COF2 (circles), CF3ONO2 (triangles, values
multiplied by 10 for clarity) and C2F5C(O)O2NO2 (squares)
versus loss of C2F5CHO following Cl atom initiated oxidation
in the presence of NOx in 700 Torr of air.
C2F5CðOÞO2 þNO2 þM!C2F5CðOÞO2NO2 ð17Þ

M.P. Sulbaek Andersen et al. / Chemical Physics Letters 379 (2003) 28–36
33
Peroxy nitrates are thermally unstable and
decompose to reform acetyl peroxy radicals and
NO2. It seems reasonable to assume that
C2F5C(O)O2NO2 will decompose at a rate similar
to that of CF3C(O)O2NO2 and hence will have a
lifetime at 296 K of approximately 5 h [19]. In the
presence of excess NO2 thermal decomposition of
C2F5C(O)O2NO2 is masked by its reformation via
reaction (17). Following UV irradiation of
C2F5CHO/NO2/Cl2/O2 mixtures two carbon con-
taining products were observed; COF2 and a
product with IR features at 791, 970, 990, 1086,
1111, 1201, 1243, 1301, 1761, and 1849 cmꢂ1
whose spectrum is shown in Fig. 4A. The features
shown in Fig. 4A increased linearly with the loss of
C2F5CHO. The product features at 791, 1301,
1761, and 1849 cmꢂ1 are characteristic of the NO2
deformation, NO2 symmetric stretch, NO2 asym-
metric stretch, and CO stretching modes in acetyl
peroxynitrates and we ascribe this product to
C2F5C(O)O2NO2. We assign the spectrum in
Fig. 4A to C2F5C(O)O2NO2. For comparison
the spectrum of CH3C(O)O2NO2 is shown in
Fig. 4B. The inset in Fig. 4A shows a plot of the
observed formation of C2F5C(O)O2NO2 and
COF2 versus loss of C2F5CHO. Calibration of
C2F5C(O)O2NO2 was achieved by assuming that
reaction (5) leads to the formation of
C2F5C(O)O2NO2 and COF2 in a combined molar
yield of 100%.
The integrated band strength (1700–1800 cmꢂ1
)
of the NO2 asymmetric stretching feature in
C2F5C(O)OONO2 was (5.25 ꢀ 1.04) ꢁ 10ꢂ17 cm
moleculeꢂ1 (we attribute 20% uncertainty to the
integrated band strength). This value is indistin-
guishable from that of (5.14 ꢀ 0.10) ꢁ 10ꢂ17 cm
moleculeꢂ1 for the corresponding feature in
CH3C(O)O2NO2 [20].
3.4. Relative rate study of the OH + C2F5CHO
reaction in 700 Torr of air
The kinetics of reaction (19) were measured
relative to reactions (20) and (21):
C2F5CHO þ OH ! products
C2H2 þ OH ! products
C2H4 þ OH ! products
ð19Þ
ð20Þ
ð21Þ
Initial reactant concentrations were 10–34 mTorr
of C2F5CHO, 89–162 mTorr of CH3ONO, 15–31
mTorr NO, and either 3–4 mTorr of C2H2 or 9
mTorr of C2H4 in 700 Torr of air diluent. When
working with compounds such as C2F5CHO which
have modest reactivity towards OH radicals an
indirect variation of the relative rate technique can
be useful [13]. As discussed above, in the presence
of excess NO the oxidation of C2F5CHO gives
COF2 and CF3ONO2 in essentially quantitative
yield. COF2 and to a lesser extent CF3ONO2, have
intense characteristic IR features that are conve-
nient to monitor. In the present work the loss of
C2F5CHO was inferred by the formation of its
oxidation products COF2 and CF3ONO2.
Fig. 4. Infrared spectra of C2F5C(O)O2NO2 (A) and
CH3C(O)O2NO2 (PAN) (B). The inset in (A) shows the yield of
C2F5C(O)O2NO2 (squares) and COF2 (circles) versus loss of
C2F5CHO (see text for details).
The calculated loss of C2F5CHO versus that of
the reference compounds in the presence of OH
atoms is shown in Fig. 5. Linear least-squares

34
M.P. Sulbaek Andersen et al. / Chemical Physics Letters 379 (2003) 28–36
culeꢂ1 sꢂ1 [24] at 298 K. The reactivity of
C2F5CHO towards OH radicals measured in the
present work is indistinguishable from that re-
ported previously for CF3CHO. This observation
is consistent with the general pattern noted in
Section 3.1 that the rate of H-atom abstraction by
Cl atoms and OH radicals from CF3X is equal to,
or slower than, that from C2F5X.
4. Implications for atmospheric chemistry
The present work serves to improve our un-
derstanding of the atmospheric chemistry of
C2F5CHO and, by analogy, other perfluorinated
aldehydes of the general formula CnF2nþ1CHO. Cl
atoms and OH radicals react with C2F5CHO
with rate constants of (1.96 ꢀ 0.28) ꢁ 10ꢂ12 and
Fig. 5. Decay of C2F5CHO versus C2H2 (squares) or C2H4
(circles) in the presence of OH radicals in 700 Torr of air diluent
at 296 ꢀ 2 K.
(5.26 ꢀ 0.80) ꢁ 10ꢂ13 cm3 moleculeꢂ1
s
ꢂ1, respec-
tively. Reaction occurs at the aldehydic H-atom.
On progressing through the series of fluorinated
aldehydes of the general formula CnF2nþ1CHO, the
greatest change in chemical environment of the
aldehydic H-atom occurs on moving from
CF3CHO to C2F5CHO. Given the finding in the
present work that the reactivity of Cl atoms and
OH radicals towards C2F5CHO are indistinguish-
able from the corresponding reactions with
CF3CHO, it seems reasonable to generalize our
results to all members of the CnF2nþ1CHO series.
By taking the average of the results obtained
herein for C2F5CHO and literature data for
k(Cl + CF3CHO) [15] and k(OH + CF3CHO) [17]
we arrive at kðCl þ CnF2nþ1CHOÞ ¼ ð1:9 ꢀ 0:5Þ ꢁ
10ꢂ12 and kðOH þ CnF2nþ1CHOÞ ¼ ð5:6 ꢀ 1:6Þ ꢁ
analysis of the data in Fig. 5 gives k19=k20
and
¼
0:593 ꢀ 0:065
k19=k21 ¼ 0:0646 ꢀ 0:0037,
where the quoted uncertainties are two standard
deviations from the linear regressions. Using
k20 ¼ 8:45 ꢁ 10ꢂ13 [21] and k21 ¼ 8:52 ꢁ 10ꢂ12 [22]
cm3 moleculeꢂ1 sꢂ1 we derive k19 ¼ ð5:01 ꢀ 0:55Þ ꢁ
10ꢂ13 and k19 ¼ ð5:50 ꢀ 0:32Þ ꢁ 10ꢂ13 cm3 mole-
culeꢂ1 ꢂ1. We choose to quote a final value for k4
s
which is the average of those determined using the
two reference compounds together with error
limits which encompass the extremes of the indi-
vidual determinations, hence, k19 ¼ ð5:26 ꢀ 0:80Þꢁ
10ꢂ13 cm3 moleculeꢂ1 ꢂ1. There are no literature
s
data for k19 to compare with our results.
While there have been no previous studies of
the reactivity of OH radicals towards C2F5CHO,
we can compare our results with previous studies
10ꢂ13 cm3 moleculeꢂ1 ꢂ1. Quoted uncertainties
s
were chosen to encompass the extremes of the in-
dividual determinations.
ꢀ ꢀ
of the reactivity of CF3CHO. Dobe et al. [23]
The value of kðOH þ CnF2nþ1CHOÞ derived
above can be used to provide an estimate of the
atmospheric lifetime of CnF2nþ1CHO with respect
to reaction with OH radicals. Assuming an atmo-
spheric lifetime for CH3CCl3 with respect to re-
action with OH radicals of 5.7 years [25] and a rate
constant for the CH3CCl3 + OH reaction of
1.0 ꢁ 10ꢂ14 cm3 moleculeꢂ1 sꢂ1 [2] leads to an es-
timate of the atmospheric lifetime of CnF2nþ1CHO
of (1.0 ꢁ 10ꢂ14)/(5.6 ꢁ 10ꢂ13) ꢁ 5.7 ꢁ 365 ꢃ 40 days.
have measured k(OH + CF3CHO) ¼ (1.1 ꢀ 0.7) ꢁ
10ꢂ12 cm3 moleculeꢂ1 sꢂ1 using discharge flow
techniques while Scollard et al. [16] have
reported k(OH + CF3CHO) ¼ (6.5 ꢀ 0.5) ꢁ 10ꢂ13
and (5.4 ꢀ 1.2) ꢁ 10ꢂ13 cm3 moleculeꢂ1 sꢂ1 based
on pulsed laser photolysis resonance fluorescence
and relative rate experiments, respectively. The
IUPAC data evaluation committee recommends
k(OH + CF3CHO) ¼ (6.0 ꢀ 1.2) ꢁ 10ꢂ13 cm3 mole-

M.P. Sulbaek Andersen et al. / Chemical Physics Letters 379 (2003) 28–36
35
The optimal temperature for such a scaling
analysis is 272 K [26] (rather than 296 K used here)
but we do not have any data for kðOH þ
CnF2nþ1CHOÞ at 272 K. By analogy to other
fluorinated organics [25], the temperature depen-
dence of reaction (1) is expected to be similar to
that for reaction of OH radicals with CH3CCl3.
Hence, the use of 296 K rather than 272 K is not
expected to have any material impact on the esti-
mated atmospheric lifetime.
source of perfluorocarboxylic acids, CnF2nþ1CðOÞ
OH. At this point it is germane to note the exis-
tence of several other potential routes for conver-
sion of CnF2nþ1CHO into CnF2nþ1CðOÞOH in
the atmosphere. These include: (i) conversion of
CnF2nþ1CHO into the peroxy nitrate CnF2nþ1
CðOÞO2NO2 followed by hydrolysis of CnF2nþ1
CðOÞO2NO2 to give CnF2nþ1CðOÞOH, (ii) reaction
of CnF2nþ1CðOÞO2 with HO2 radicals to give
CnF2nþ1CðOÞOH and O3, and (iii) addition of water
to CnF2nþ1CHO to give the hydrate CnF2nþ1CH
As discussed elsewhere [16], photolysis is likely
to be of comparable importance to reaction with
OH as an atmospheric loss mechanism of
CnF2nþ1CHO. The published UV cross-section
data for C2F5CHO [27,28] suggests that this mol-
ecule has an absorption spectrum which is indis-
tinguishable from that of C2F5C(O)CF(CF3)2 [29].
C2F5C(O)CF(CF3)2 has an atmospheric lifetime
with respect to photolysis of 1–2 weeks [29]. It
seems likely that C2F5CHO will have a lifetime
with respect to photolysis of the order of a few
weeks. The absence of photodissociation quantum
yield data for C2F5CHO under atmospheric con-
ditions prevents a more quantitative assessment of
the photolysis rate at this time. Finally, there are
two interesting points to note with regard to the
atmospheric photolysis of CnF2nþ1CHO. First, it
has been reported that photolysis of C2F5CHO
using a high pressure Hg lamp gives a significant
yield of C2F5H [30]. Second, the intensity of the
UV spectra increase substantially with increasing
chain length for n ¼ 1, 2, and 3 [27,28]. Com-
pounds of the general formula CnF2nþ1H have high
global warming potentials. If photolysis is a major
atmospheric loss for CnF2nþ1CHO and if photoly-
sis produces CnF2nþ1H this could have important
environmental consequences.
The experiments conducted in the present work
show that the reaction of Cl atoms with C2F5CHO
in air in the presence of excess NO initiates a
sequence of reactions in which the molecule
‘‘unzips’’ to give COF2 fragments. It seems likely
that a similar fate awaits larger members of
the CnF2nþ1CHO series. No evidence for the for-
mationof perfluorocarboxylic acids wasobserved in
the present experiments. We conclude that the gas-
phase atmospheric oxidation of CnF2nþ1CHO in
the presence of excess NO is not a significant
ðOHÞ , reaction of the hydrate with OH radicals to
2
give the CnF2nþ1CðOHÞ radical which will probably
2
react with O2 to give CnF2nþ1 CðOÞOH and HO2.
These possibilities are under investigation in our
laboratories.
Acknowledgements
We thank Roc Carter (Ford Motor Company)
and Geoff Tyndall (National Center for Atmo-
spheric Research) for helpful discussions regarding
IR spectra. We thank Malisa Chiappero and
€
Gustavo Arguello (Universidad Nacional de
Cordoba, Argentina) for help calibrating the
ꢀ
CF3ONO2 spectrum. This research was funded, in
part, by an NSERC Strategic Grant and the
Danish Natural Science Research Council.
References
[1] E. Kissa, in: M.J. Schick, F.M. Fowkes (Eds.), Fluorinated
Surfactants, Synthesis, Properties, and Applications, Mar-
cel Dekker, Inc, New York, 1994.
[2] W.B. DeMore, S.P. Sander, D.M. Golden, R.F. Hampson,
M.J. Kurylo, C.J. Howard, A.R. Ravishankara, C.E.
Kolb, M.J. Molina, JPL Publ. (1997), 94–26.
[3] R. Atkinson, J. Phys. Chem. Ref. Data., Monograph 2,
(1989).
[4] V.C. Papadimitriou, A.V. Prosmitis, Y.G. Lararou, P.
Papagiannakopoulos, J. Phys. Chem. A 107 (2003) 3733.
[5] C.A. Moody, J.W. Martin, W.C. Kwan, D.C.G. Muir,
S.A. Mabury, Environ. Sci. Tech. 36 (2002) 545.
[6] C.A. Moody, W.C. Kwan, J.W. Martin, D.C.G. Muir,
S.A. Mabury, Anal. Chem. 73 (2001) 2200.
[7] J.W. Martin, M.M. Smithwick, B.M. Braune, P.F. Hoek-
stra, D.C.G. Muir, S.A. Mabury, Environ. Sci. Tech.,
submitted.

36
M.P. Sulbaek Andersen et al. / Chemical Physics Letters 379 (2003) 28–36
[20] N. Tsalkani, G. Toupance, Atmos. Environ. 23 (1989)
[8] ‘‘Determination of low levels of fluoropolymer polymeri-
zation aids – a guidance document’’, The Society of the
Plastics Industry, SPI Literature Catalogue #BZ-102, New
York, 2003.
1849.
[21] M. Sørensen, E.W. Kaiser, M.D. Hurley, T.J. Wallington,
O.J. Nielsen, Int. J. Chem. Kinet. 35 (2003) 191.
[22] J.G. Calvert, R. Atkinson, J.A. Kerr, S. Madronich, G.K.
Moortgat, T.J. Wallington, G. Yarwood, The Mechanism
of Atmospheric Oxidation of the Alkenes, Oxford Univer-
sity Press, Oxford, 2000.
[9] D.A. Ellis, S.A. Mabury, J.W. Martin, D.C.G. Muir,
Nature 412 (2001) 6844.
[10] D.A. Ellis, J.W. Martin, S.A. Mabury, M.D. Hurley, M.P.
Sulbaek Andersen, T.J. Wallington, Environ. Sci. Tech., in
press.
ꢀ ꢀ
[23] S. Dobe, L.A. Kachatryan, T. Berces, Ber. Bunsenges.
Phys. Chem. 93 (1989) 847.
[11] M.D. Hurley, M.P. Sulbaek Andersen, T.J. Wallington,
D.A. Ellis, J.W. Martin, S.A. Mabury, J. Phys. Chem. A,
submitted.
[24] R. Atkinson, D.L. Baulch, R.A. Cox, R.F. Hampson Jr.,
J.A. Kerr, M.J. Rossi, J. Troe, J. Phys. Chem. Ref. Data 26
(1997) 763.
[12] T.J. Wallington, S.M. Japar, J. Atmos. Chem. 9 (1989) 399.
[13] K. Takahashi, Y. Matsumi, T.J. Wallington, M.D. Hurley,
J. Geophys. Res. 107 (2002), ACH 4-1.
[25] R.G. Prinn, R.F. Weiss, B.R. Miller, J. Huang, F.N.
Alyea, D.M. Cunnold, P.J. Fraser, D.E. Hartley, P.G.
Simmonds, Science 269 (1995) 187.
[14] P.H. Wine, D.H. Semmes, J. Phys. Chem. 87 (1983)
3572.
[26] C.M. Spivakovsky, J.A. Logan, S.A. Montzka, Y.L.
Balkanski, M. Foreman-Fowler, D.B.A. Jones, L.W.
Horowitz, A.C. Fusco, C.A.M. Brenninkmeijer, M.J.
Prather, S.C. Wofsy, M.B. McElroy, J. Geophys. Res.
105 (2000) 8931.
[15] T.J. Wallington, M.D. Hurley, Int. J. Chem. Kinet. 25
(1993) 819.
[16] D.J. Scollard, J.J. Treacy, H.W. Sidebottom, C. Balestra-
ꢀ
Garcia, G. Laverdet, G. LeBras, H. MacLeod, S. Teton,
J. Phys. Chem. 97 (1993) 4683.
[27] G.O. Pritchard, G.H. Miller, J.K. Foote, Can. J. Chem. 40
(1962) 1830.
[17] R. Atkinson, D.L. Baulch, R.A. Cox, R.F. Hampson Jr.,
J.A. Kerr, M.J. Rossi, J. Troe, J. Phys. Chem. Ref. Data 29
(2000) 167.
[28] R.P. Borkowski, P. Ausloos, J. Am. Chem. Soc. 48 (1962)
4044.
[18] J.A. Kerr, CRC Handbook of Chemistry and Physics,
CRC Press, Boca Raton, 1993, 9–123.
[29] N. Taniguchi, T.J. Wallington, M.D. Hurley, A.G. Gus-
chin, L.T. Molina, M.J. Molina, J. Phys. Chem. A 107
(2003) 2674.
[19] T.J. Wallington, J. Sehested, O.J. Nielsen, Chem. Phys.
Lett. 226 (1994) 563.
[30] E.R. Morris, J.C.J. Thynne, J. Phys. Chem. 72 (1968) 3351.