The gas phase tropospheric removal of fluoroaldehydes (CxF 2x+1CHO, x = 3, 4, 6)

Article abstract of DOI:10.1039/b703741b

The rate coefficient of the OH reaction with the perfluoroaldehydes C ₃F₇CHO and C₄F₉CHO have been determined in the temperature range 252-373 K using the pulsed laser photolysis-laser induced fluorescence (PLP-LIF) method: k_C3F7CHO+OH = (2.0 ± 0.6) × 10^-12 exp[-(369 ± 90)/T] and k_C4F9CHO+OH = (2.0 ± 0.5) × 10^-12 exp[-(356 ± 70)/T] cm³ molecule^-1 s^-1, corresponding to (5.8 ± 0.6) × 10^-13 and (6.1 ± 0.5) × 10^-13 cm³ molecule^-1 s ^-1, respectively, at 298 K. The UV absorption cross sections of these two aldehydes and CF₃(CF₂)₅CH₂CHO have been measured over the range 230-390 nm at 298 K and also at 328 K for CF₃(CF₂)₅CH₂CHO. The obtained results for C₃F₇CHO and C₄F₉CHO are in good agreement with two recent determinations but the maximum value of the absorption cross section for CF₃(CF₂)₅CH ₂CHO is over a factor of two lower than the single one recently published. The photolysis rates of C₃F₇CHO, C ₄F₉CHO and CF₃(CF₂)₅CHO have been measured under sunlight conditions in the EUPHORE simulation chamber in Valencia (Spain) at the beginning of June. The photolysis rates were, respectively, J_obs = (1.3 ± 0.6) × 10^-5, (1.9 ± 0.8) × 10^-5 and (0.6 ± 0.3) × 10 ^-5 s^-1. From the J_obs measurements and calculated photolysis rate J_calc, assuming a quantum yield of unity across the atmospheric range of absorption of the aldehydes, quantum yields J_obs/J_calc = (0.023 ± 0.012), (0.029 ± 0.015) and (0.046 ± 0.028) were derived for the photodissociation of C₃F₇CHO, C₄F₉CHO and CF ₃(CF₂)₅CHO, respectively. The atmospheric implication of the data obtained in this work is discussed. The main conclusion is that the major atmospheric removal pathway for fluoroaldehydes will be photolysis, which under low NO_x conditions, may be a source of fluorinated carboxylic acids in the troposphere. the Owner Societies.

Full text of DOI:10.1039/b703741b

View Article Online / Journal Homepage / Table of Contents for this issue
This paper is published as part of a PCCP Themed Issue on:
Synergies between Experimental and Theoretical
Studies of Gas Phase Reactions
Guest Edited by: Paul Seakins (Leeds) and Struan Robertson (Accelrys,
Cambridge and Leeds)
Published in issue 31, 2007 of PCCP
Image reproduced by permission of Dr Taatjes and Linda A.Houston for Sandia National
Laboratories from Phys.Chem.Chem.Phys., 2007, 9, 4315-4331
Other papers published in this issue include:
Direct detection of polyynes formation from the reaction of ethynyl radical (C
CH) and allene (CH C CH
2
H) with propyne (CH –
3
C
2
2
)
Fabien Goulay, David L. Osborn, Craig A. Taatjes, Peng Zou, Giovanni Meloni and
Stephen R. Leone, Phys. Chem. Chem. Phys., 2007
DOI: 10.1039/b614502g
Master equation methods for multiple well systems: application to the 1-,2-pentyl system
Struan H. Robertson, Michael J. Pilling, Luminita C. Jitariu and Ian H. Hillier, Phys. Chem. Chem. Phys.,
2007
DOI: 10.1039/b704736c
Ab initio methods for reactive potential surfaces
Lawrence B. Harding, Stephen J. Klippenstein and Ahren W. Jasper, Phys. Chem. Chem. Phys., 2007
DOI: 10.1039/b705390h
Visit the website for both cutting edge research papers and authoritative
review articles by leaders in a range of fields
www.rsc.org/pccp

PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
The gas phase tropospheric removal of fluoroaldehydes
CxF2x+1CHO, x = 3, 4, 6)w
(
a
a
a
b
c
G. Solignac, A. Mellouki,* G. Le Bras, Mu Yujing and H. Sidebottom
Received 12th March 2007, Accepted 18th April 2007
First published as an Advance Article on the web 21st May 2007
DOI: 10.1039/b703741b
The rate coefficient of the OH reaction with the perfluoroaldehydes C F CHO and C F CHO
4 9
3
7
have been determined in the temperature range 252–373 K using the pulsed laser photolysis–laser
ꢂ12
induced fluorescence (PLP-LIF) method: k 3 7
ꢂ12
C F CHO+OH = (2.0 ꢀ 0.6) ꢁ 10
exp[ꢂ(369 ꢀ 90)/T]
3
ꢂ1 ꢂ1
and k 4 9
C F CHO+OH = (2.0 ꢀ 0.5) ꢁ 10
exp[ꢂ(356 ꢀ 70)/T] cm molecule
s , corresponding
ꢂ13
ꢂ13
3
cm molecule s , respectively, at 298 K. The
ꢂ1 ꢂ1
to (5.8 ꢀ 0.6) ꢁ 10
and (6.1 ꢀ 0.5) ꢁ 10
UV absorption cross sections of these two aldehydes and CF
3
(CF
2
)
5
CH
2
CHO have been
(CF CH CHO. The
obtained results for C F CHO and C F CHO are in good agreement with two recent
measured over the range 230–390 nm at 298 K and also at 328 K for CF
3
2
)
5
2
3
7
4 9
determinations but the maximum value of the absorption cross section for CF (CF ) CH CHO is
2
3
2 5
over a factor of two lower than the single one recently published. The photolysis rates of
CHO, C CHO and CF (CF CHO have been measured under sunlight conditions in the
EUPHORE simulation chamber in Valencia (Spain) at the beginning of June. The photolysis
C
F
3 7
F
4 9
3
2 5
)
ꢂ5
ꢂ5
ꢂ5 ꢂ1
rates were, respectively, Jobs = (1.3 ꢀ 0.6) ꢁ 10 , (1.9 ꢀ 0.8) ꢁ 10 and (0.6 ꢀ 0.3) ꢁ 10
From the Jobs measurements and calculated photolysis rate Jcalc, assuming a quantum yield of
unity across the atmospheric range of absorption of the aldehydes, quantum yields Jobs/Jcalc
0.023 ꢀ 0.012), (0.029 ꢀ 0.015) and (0.046 ꢀ 0.028) were derived for the photodissociation of
C F CHO, C F CHO and CF (CF ) CHO, respectively. The atmospheric implication of the data
s .
=
(
3
7
4
9
3
2 5
obtained in this work is discussed. The main conclusion is that the major atmospheric removal
pathway for fluoroaldehydes will be photolysis, which under low NO
source of fluorinated carboxylic acids in the troposphere.
x
conditions, may be a
4
,5
Introduction
aldehydes under atmospheric conditions. Sellevag et al. have
investigated the photolysis of CF CHO and CHF CHO with
3
2
Long-chain fluorotelomers (FTOHs, C F2x+1CH CH OH),
x
2
2
natural sunlight and reported quite different photolytic atmo-
spheric lifetimes for the two compounds. A lower limit of 10 d
used in a variety of industrial products, have been suggested
as a potential source of persistent and toxic perfluorinated
was obtained for the photolysis lifetime of CF CHO whereas
3
x
carboxylic acids (PFCAs, C F2x+1C(O)OH, n = 6–12), which
the lifetime of CHF CHO was found to be about 10 h.
2
6
Chiappero et al. have recently determined the quantum yields
have been detected in the environment (e.g., ref. 1). PFCAs
may be produced in the gas phase from the perfluoroalde-
for photodissociation of CF CHO and C F CHO following
3 4 9
x
hydes, C F2x+1CHO, which have been shown to be oxidation
photolysis at 308 nm in 700 Torr of nitrogen, and esti-
mated from the data that these aldehydes will have atmospheric
lifetimes with respect to photolysis of less than a few days.
The present work was performed in order to provide a
better understanding of the relative importance of the loss of
fluoroaldehydes by reaction with OH, photolysis and uptake
in water droplets in the troposphere and hence to estimate the
potential importance of perfluoroaldehydes as a source of
PFCAs in the atmosphere. The reported measurements in-
clude rate coefficients for the reaction of OH radicals, photo-
lysis decay rates under natural sunlight conditions and UV
products of FTOHs under atmospheric conditions (e.g. ref. 2
and 3 and references therein). The yields of PFCAs will depend
on the relative importance of the various possible tropospheric
loss processes for the aldehydes. The oxidation of the alde-
hydes can be initiated by reaction with hydroxyl radicals,
photolysis or by uptake in water droplets. The rate of removal
by reaction with OH and the yield of fluorinated carboxylic
acids in these reactions has been reasonably established,
however, the importance of degradation by photolysis is
uncertain. Little data is available on the photolysis of fluoro-
3 9 4 9
absorption cross sections for C F CHO and C F CHO.
a
Institut de Combustion, Ae´rothermique, Re´activite´ et Environnement
(ICARE), CNRS, 1C Avenue de la Recherche Scientifique F-45071
cedex 02 Orle´ans France
Research Center for Eco-Environmental Sciences, Academia Sinica,
Beijing, 100085, China
b
c
Experimental
School of Chemistry and Chemical Biology, University College
Dublin, Ireland
Rate coefficients for OH radical reactions
The pulsed laser photolysis–laser induced fluorescence
(PLP-LIF) technique was used to measure the rate coefficient
w Electronic supplementary information (ESI) available: Absorption
cross sections. See DOI: 10.1039/b703741b
4
200 | Phys. Chem. Chem. Phys., 2007, 9, 4200–4210
This journal is ꢃc the Owner Societies 2007

of the reaction of OH with two perfluorinated aldehydes
FAL) C CHO and C CHO. This technique has been
concentrations of the fluoroaldehydes were calculated from
their mass flow rates, the temperature, and the pressure in the
reaction cell. All flow rates were measured with mass flow-
meters. The pressure in the cell was measured with a capaci-
tance manometer positioned at the cell entrance.
(
F
3 7
4 9
F
described in detail in previous papers from this laboratory,
7
hence only a brief review is given here. Hydroxyl radicals
were produced by photolysis of H O at l = 248 nm using a
2
2
KrF excimer laser. The temporal concentration–time profiles
of hydroxyl radicals were monitored at various reaction times
ranging from 10 ms to 10 ms by pulsed laser induced fluores-
cence. A Nd:YAG pumped, frequency-doubled dye laser was
used to excite the OH radical at l = 282 nm and its
fluorescence was detected by a photomutiplier, fitted with a
UV absorption cross section measurements
The UV absorption spectra of three perfluoroaldehydes
C F CHO, C F CHO and CF (CF ) CH CHO were deter-
3
7
4
9
3
2 5
2
mined in a 100 cm long double-jacketed Pyrex cell. An attempt
was made to measure the absorption cross sections of
CF (CF ) CHO but the low vapour pressure precluded accu-
3
09 nm narrow bandpass filter. The output pulse from the
3
2 5
photomultiplier was integrated for a preset period by a gated
charge integrator. Typically the fluorescence signals from 10 to
rate concentration measurements of this compound, and hence
the measured absorption cross section measurements were
15 different delay times from 100 probe laser shots were averaged
deemed to be unreliable. The optical set-up utilized a D lamp
2
to generate OH concentration–time profiles over at least three
lifetimes. Mixtures of FAL/H O in helium diluent were flowed
as the light source with a diode array detector. The Hg lines at
2
2
253.7, 313.2 and 365 nm from a low pressure Hg Penray lamp
and the 213.8 nm line from a Zn lamp were used for wave-
length calibration. The absorption measurements were made
in static and dynamic conditions over the wavelength range
230–390 nm with a resolution of 0.1 nm. The spectrum was
divided into five overlapping regions of about 40 nm in the
range 230–390 nm. A 0–10 Torr capacitance manometer was
used for pressure measurements.
slowly through the cell so that each photolysis/probe sequence
interrogated a fresh gas mixture and reaction products did not
build up in the cell. The experiments were conducted in the
temperature range of 252–373 K at around 100 Torr.
All kinetic experiments were carried out under pseudo-first-
o o
order conditions with [FAL] c [OH] . Under these condi-
tions, the OH concentration–time profiles followed the
pseudo-first-order rate law:
A series of experiments were carried out (typically 3–4
independent determinations were made at each pressure) to
establish the applicability of the Beer–Lambert law:
ꢂk0t
0
½
OHꢄ ¼ ½OHꢄ e
where k ¼ k½FALꢄ þ ko
ðIÞ
t
o
sðlÞ ¼ ꢂ ln½IðlÞ=IoðlÞꢄ=LC
where I and I are the light intensities for the filled and the
ðIIÞ
where k is the rate coefficient for the reaction of OH with the
FAL. The decay rate, k , is the first-order OH decay rate in the
absence of the FAL and is the sum of the reaction rate of OH
with the H precursor and the diffusion rate of OH out of
o
o
empty cells, respectively, L is the length of the absorption cell
and C is the concentration of the compound in the cell.
2
O
2
the detection zone. The concentrations of OH at various
reaction times (the delay time between the photolysis pulse
and the probe pulse) were determined by measuring the LIF
signal intensity at each delay time. Weighted (according to the
signal to noise ratio of the measured signal at a given time)
least square analyses were used to fit the OH concentra-
tion–time profiles to eqn (I) and hence to determine the values
Photolysis under sunlight conditions
Photolysis of the fluoroaldehydes by natural sunlight was
performed in a large outdoor simulation chamber in Valencia,
Spain, EUPHORE, (longitude = ꢂ0.51, latitude = 39.51)
during the month of June 2005. A detailed description of the
EUPHORE facility and the analytical instruments employed
0
8
of the rate coefficients k . The second order rate coefficients, k,
0
has previously been reported in the literature. It consists of a
3
200 m hemispherical chamber made of FEP foil with a
were obtained from the measured values of k at various
concentrations of FAL. The helium carrier gas (UHP certified
transmission of more than 75% of the solar radiation in the
wavelength range between 290 and 550 nm. The chamber is
equipped with a FTIR spectrometer coupled to a White-type
multi-path mirror system for in-situ analysis (optical path
to 499.9995% Alphagas) was used without purification. The
5
2 2
0 wt% H O solution obtained from Prolabo was concen-
trated by bubbling helium through the solution to remove
water for several days prior to use and constantly during the
course of the experiments. It was admitted into the reaction
cell by passing a small flow of helium through a glass bubbler
containing H O . The fluoroaldehydes were synthesized as
length 553.5 m). Infrared spectra were recorded every 10 min
ꢂ1
by co-adding 570 interferograms with a resolution of 1 cm
.
Analyses of the fluorinated compounds were also performed
using GC-MS, GC-ECD and SPME sampling (solid phase
2
2
described in the Materials section.
microextraction) followed by GC-MS. Ozone, CO and NO
x
The FAL to be studied was premixed with helium in a 10 L
glass light-tight bulb to form B1% mixture at a total pressure
of B800 Torr. The mixtures were analysed both by GC-FID
and UV absorption spectroscopy and did not show any
detectable impurities. All the gases were flowed into the
reactor through Teflon tubing. The gas mixture containing
the FAL, the photolytic precursor (H O ), and the bath gas
were analysed using specific analysers (Monitor Labs 9810,
Thermo Environment 48C, Monitor Labs 9841A and ECO-
Physics CLD770 AL ppt with PLC 760 photolytic converter).
Reactant and product concentrations were determined using
calibrated reference spectra. Photolysises of the fluoroalde-
hydes were studied in purified dry air (dewpoint less than
ꢂ20 1C), at ambient temperature and pressure. The large
volume of the chamber minimizes the effects of wall reactions.
To avoid contamination from outside air by diffusion into the
2
2
(
approximately 100 Torr of helium) were flowed through the
ꢂ1
cell with a linear velocity ranging from 5 to 20 cm s . The
This journal is ꢃc the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4200–4210 | 4201

chamber due to minor leaks in the system, the chamber was
kept at a slight overpressure relative to the outside air by
purified air inflation. This led to minor loss of the reactants
and photolysis products by leakage from the chamber during
less steel spoon. The mixture was kept at room temperature
under 1 atm pressure of helium for 1 h. The gas mixture in the
flask was transferred into the trap, which was cooled with
liquid nitrogen, using a controlled low flow of helium such that
the pressure after the trap was less than 8 Torr. The flask was
then refilled with helium to 1 atm pressure. The procedure was
repeated twice. About 1 mL of the colourless liquid samples of
the photolysis experiments. The inert tracer SF was added to
6
the chamber along with the reactants in order to measure the
loss by dilution.
In all experiments isoprene was added as a scavenger to
avoid loss of the fluoroaldehydes by reaction with OH radi-
cals, which may be generated in the system. Photolysis of the
aldehydes is expected to lead to the formation of fluoroalkyl
radicals, which in the presence of molecular oxygen will
x 3 2 5 2
C F2x+1CHO was obtained. CF (CF ) CH CHO (P&M-In-
vest, stated purity of 97%) and isoprene (Aldrich, stated purity
of 99%) were used without further purification.
eventually yield CF
3
O radicals. These radicals could also react
Results and discussion
with the aldehydes. The available rate coefficient data for the
9
reaction of CF O with alkenes indicates that these radicals
Rate coefficient measurements for the reaction of OH radicals
with C F CHO and C F CHO
3
3
7
4 9
will also be efficiently removed from the system by reaction
with isoprene.
Loss of OH radicals was monitored under pseudo-first-order
Under the present experimental conditions the fluoroalde-
hydes (FAL) are lost mainly by photolysis and dilution
through the following reactions:
conditions with [FAL] c [OH] for which the pseudo-first-
order rate coefficient is given by eqn (I). The OH radical
0
0
decays were plotted in the form ln[OH] versus time and
t
showed excellent linearity over at least three lifetimes (Fig.
0
FAL þ hn ! Products JobsðFALÞ
ð1Þ
ð2Þ
1
a). The pseudo-first-order rate coefficients, k = (k[FAL] +
0
k ), were obtained from the slopes of the plots. Possible
FAL ! dilution kdil
contributions to the measured rate coefficients from second-
ary reactions of OH with the products of the reaction of OH
were significantly reduced by using high [FAL]/[OH] ratios,
It is also possible that the aldehydes may be removed by
reaction at the chamber wall:
2
4
(
typically (10 –10 )), and low concentrations of OH radicals.
FAL ! wall loss kwall
SF6 was added to the reaction chamber to determine the
ð3Þ
The concentration ranges of the reactants were: [OH]
1
=
15
0
0
ꢂ3
(5–53) ꢁ 10 molecule cm and [FAL] = (0.7–2) ꢁ 10
0
ꢂ3
0
overall dilution rate coefficient kdil:
molecule cm . As expected, the values of k were indepen-
dent of the OH radical concentration. The rate coefficients
were also shown to be independent of variations in the flow
velocity through the reactor. The photolysis of the studied
aldehydes at 248 nm is expected to be of minor importance in
our experimental conditions considering their absorption
SF6 ! dilution kdil
ð4Þ
The temporal concentration–time profile for SF
6
is given by:
ln([SF
6 0 6 t 6 0 6 t
] /[SF ] ) = kdilt, where [SF ] and [SF ] are the initial
SF concentration and at time t, respectively. Thus, the
6
ꢂ21
2
apparent photolysis rate coefficient of the fluoroaldehydes,
Jmeas, can be obtained from the expression:
cross sections at this wavelength (s E 3 ꢁ 10
cm
ꢂ1
molecule ), the used laser fluence (i.e. 5 mJ) and the
1
aldehydes concentrations (up to 10 molecule cm ). Values
5
ꢂ3
lnð½FALꢄ =½FALꢄ Þ ¼ ðJobsðFALÞ þ kdil þ kwallÞt
ðIIIÞ
0
t
of k = (k [H O ] + kdiff) were determined to be in the
2
0
OH
2
ꢂ1
0
where Jmeas = Jobs(FAL) + kdil + kwall
range 50–200 s . Plots of (k –k ) versus the concentration of
0
Measured amounts of FAL and isoprene were introduced
C F CHO and C F CHO for the reaction of OH radicals
4 9
3
7
into the chambers with mixing ratios in the range B200 ppbv
were linear and the absolute rate coefficients, k, were derived
from the least-squares fit of the straight lines. Examples of
the data at various temperatures are given in Fig. 1b and c.
The experimental conditions and the measured values of
the rate coefficients at a total pressure of 100 Torr over the
temperature range 252–373 K are listed in Table 1. The
quoted errors for the rate coefficients, k, determined in
this work include 2s from the least-squares analysis and the
estimated systematic error of 5% due to uncertainties in
measured concentrations. The rate coefficients are shown
plotted in the conventional form of the Arrhenius equation,
for C
4
F
9
CHO and C
3
F
7
CHO, 98–144 ppbv for C
6
F
13CHO
and 200–300 ppbv for isoprene. Around 20 ppbv of SF
6
was
also added to the chamber at the start of the reaction. The
possibility of dark reactions was examined by starting the
analytical sampling at least 30 min before opening the cham-
ber to sunlight and hence the onset of photolysis.
Materials
C
3
F
7
CHO, C
from the corresponding hydrates C
,4,6). The system consisted of a 500 ml flask, a trap, a needle
4
F
9
CHO and CF
3
(CF
2
)
5
CHO were synthesized
x
F
2x+1CH(OH) (x =
2
3
k(T) = A exp(ꢂE /RT), in Fig. 2. The Arrhenius equations
a
valve, and a 10 Torr pressure gauge. It was first pumped and
heated to remove water adsorbed on the wall, then filled with
helium of high purity to more than 1 atm total pressure.
for the reactions of OH radicals with C F CHO and
3
7
3
C F CHO in cm molecule
ꢂ1 ꢂ1
s are:
4
9
ꢂ12
kC F CHOþOH ¼ ð2:0 ꢀ 0:6Þ ꢁ 10
exp½ꢂð369 ꢀ 90Þ=Tꢄ
exp½ꢂð356 ꢀ 70Þ=Tꢄ
3
7
Phosphorous pentoxide was added to the flask under a flow
ꢂ1
(
about 50 ml min ) of helium, followed by the addition of ca.
ꢂ12
kC F CHOþOH ¼ ð2:0 ꢀ 0:5Þ ꢁ 10
2
g of C F
x
CH(OH) and mixed with P O using a stain-
2
4 9
2x+1
2
5
4
202 | Phys. Chem. Chem. Phys., 2007, 9, 4200–4210
This journal is ꢃc the Owner Societies 2007

0
Fig. 1 (a) Examples of OH decays in the absence and presence of the fluoroaldehydes. (b) Plots of k –k
0
o
vs. [C
3
F
7
CHO] at different temperatures.
The lines represent the linear least-squares fitting. (c) Plots of k –k
o
vs. [C
4
F
9
CHO] at different temperatures. The lines represent the linear least-
squares fitting.
where the quoted errors are AslnA and sE/R for A and E/R,
respectively. At T = 298 K (in cm molecule s ):
coefficient for the reaction with acetaldehyde shows a slight
negative temperature dependence, whereas the reactions with
the fluoroaldehydes have small positive activation energies.
The major reaction pathway for the reaction of OH radicals
with aliphatic aldehydes is hydrogen abstraction from the
3
–1 –1
kðOH þ C3F7CHOÞ ¼ ð5:8 ꢀ 0:6Þ ꢁ 10ꢂ13
1
2
kðOH þ C4F9CHOÞ ¼ ð6:1 ꢀ 0:5Þ ꢁ 10ꢂ13
aldehydic group. Rayez et al. have suggested that the
decrease in the rate coefficient for the reaction of OH with
fluoroaldehydes compared to acetaldehyde is both a conse-
quence of the slight increase in the carbon–hydrogen bond
strength in the fluoroaldehydes and also due to destabilization
of the transition state for the reaction of OH with the
fluoroaldehydes by the electron withdrawing inductive effect
These results are in good agreement with recent relative rate
measurements, k(C F
x
CHO + OH) = (6.5 ꢀ 1.2) ꢁ
2x+1
ꢂ1 ꢂ1
ꢂ13
3
cm molecule
10
(x = 1, 3–4) and k(C F CHO
1
0
s
2
5
ꢂ13
3
cm molecule
ꢂ1 ꢂ1 11
+
OH) = (5.26 ꢀ 0.8) ꢁ 10
s
carried
out at (296 ꢀ 2) K and 700 Torr of N . The data also confirm
2
1
3
that the rate coefficients for these reactions do not depend on
the chain length of the fluorinated group, and hence can be
used to estimate the rate coefficients for the reaction of OH
with long chain fluorinated aldehydes. The rate coefficients
reported in this work are the first absolute rate measurements
on these systems and also give the temperature dependence of
the rate coefficients.
of the fluoroalkyl group. Smith and Ravishankara have
discussed the reaction of OH with carbonyl compounds in
terms of the initial formation of a weakly bound hydrogen-
bonded adduct between the OH radical and the carbonyl
group, which leads to an intramolecular H-atom transfer via
cyclic transition states. However, they concluded that for the
fairly rapid reaction of OH with aldehydes there is likely to be
virtually no barrier between the hydrogen-bonded adduct and
the reaction products. The results of the present study on the
reactions of OH with fluoroaldehydes show that the rate
The reported rate coefficients for the reaction of OH radicals
with the fluoroaldehydes C F
x
CHO are around a factor of
2x+1
4
0 lower than for reaction with acetaldehyde. The rate
3
This journal is ꢃc the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4200–4210 | 4203

1
4
Table 1 Experimental conditions and measured rate coefficients for
the reaction of OH + C
range 252–373 K
spheric behaviour of fluoroaldehydes by Wallington et al.,
ꢂ1 ꢂ1
x
F2x+1CHO (x = 3–4) in the temperature
ꢂ11
3
(k = 1.7 ꢁ 10
exp(ꢂ1000/T) cm molecule
s ). The use
of the new values will lead to some minor changes in the
calculated atmospheric lifetimes of fluoroaldehydes.
14
[PFAL]/10
molecule
cm
ꢂ13
cm
3
k ꢀ 2s/10
ꢂ1 ꢂ1
molecule s
ꢂ3
0
FAL
C F
T/K
k –k
o
UV-absorption cross sections
3
7
CHO
252
1.2–11.1
1–13.1
1–11.3
0.9–10
1–11.5
0.8–8.5
0.8–7.4
0.7–8
48–536
30–648
38–576
47–595
70–617
49–552
48–547
47–579
4.9 ꢀ 0.1
5.1 ꢀ 0.2
5.2 ꢀ 0.3
5.8 ꢀ 0.5
5.3 ꢀ 0.1
6.7 ꢀ 0.3
7.5 ꢀ 0.2
7.5 ꢀ 0.3
2
2
2
2
3
3
3
62
72
97
97
23
48
73
The UV spectra for C
x
F2x+1CHO (x = 3 and 4) and
CF (CF CH CHO are shown in Fig. 3 and 4, and the
3
2
)
5
2
derived absorption cross sections are provided in a table as
supplementary material.w The absorption cross sections were
measured over the wavelength range 230–390 nm with five
central wavelengths (250, 280, 310, 340, and 370 nm) and three
overlapping regions (about 10 nm for each overlapping re-
gion). The UV spectra of C F CHO, C F CHO and
C
4
F
9
CHO
253
1.3–15.3
1.5–14.9
1–11
1–12.2
1.3–11.8
0.7–11
0.9–9.6
0.8–9
58–726
70–744
52–609
31–702
43–712
37–703
58–680
45–650
33–905
4.8 ꢀ 0.2
5.1 ꢀ 0.2
5.8 ꢀ 0.3
6.1 ꢀ 0.3
6.4 ꢀ 0.2
6.4 ꢀ 0.3
7.1 ꢀ 0.1
7.6 ꢀ 0.2
7.9 ꢀ 0.1
2
2
2
2
3
3
3
3
63
73
94
99
23
45
59
73
3
7
4 9
CF
3 2 5 2
(CF ) CH CHO were determined for optical densities of
0
.1–2 at 10–12 different pressures in the range (1.4–8), (1.5–11)
and (0.6–2.0) Torr, respectively, using a static method at
room temperature. In addition, the UV spectrum of
3 2 5 2
CF (CF ) CH CHO was measured using a dynamic method
0.7–11.6
at room temperature and 328 K. The standard deviations of
the determinations were less than 10% for the wavelength
range 242–358 nm for C
3 7 4 9
F CHO and C F CHO and 255–
coefficients determined at 298 K and at 100 Torr of helium
using the PLP-LIF technique are in good agreement with the
rate coefficients determined at 298 K and 1 atmosphere of air
3
35 nm for CF (CF ) CH CHO. The relative error of the four
3
2 5
2
overlapping regions was less than 5% except for wavelengths
greater than 355 nm for C F CHO and C F CHO, and greater
3
7
4 9
1
0,11
from relative rate experiments.
Hence, it appears that the
than 335 nm for CF
3
(CF
2
)
5
2
CH CHO.
rate coefficients do not show a pressure dependence, at least
over the range 100–760 Torr. Also Arrhenius plots of the rate
coefficients determined in this work from 252–373 K showed
excellent linearity. Thus, formation of collision complexes
between the OH radical and the aldehydes appear to have
little effect on the dynamics of the reactions.
3 7 4 9
The maximum absorptions of C F CHO and C F CHO
occur at around 309 nm with standard deviations of the
absorption cross sections of less than 2% at this wavelength.
Uncertainties of 5% in the measured concentrations of the
aldehydes were added to the reported errors in the cross
sections. The maximum values determined in this work for
The Arrhenius expressions obtained in this work are different
from those used in the recent modeling study on the atmo-
the absorption cross sections of C
ꢂ20
3
F
7
CHO and C
ꢂ20
4
F
9
CHO are
ꢂ1
2
cm molecule
(
8.1 ꢀ 0.6) ꢁ 10
and (9.4 ꢀ 0.7) ꢁ 10
,
respectively, are in good agreement with those previously
1
reported by Hashikawa et al. and by Chiappero et al.,
5
6
1
which are also shown in Fig. 3. Hashikawa et al. and
5
6
Chiappero et al. have also reported the UV spectrum of
CF
3
CHO and C
2
F
5
CHO and it is concluded that the fluori-
2x+1CHO, all show similar absorption
nated aldehydes, C
x
F
characteristics having a broad absorption with a maximum
around 300–310 nm, which is due to the weak n - p*
1
6
transition of the CQO bond. The maximum absorption
cross sections determined in the present study for C CHO
and C F CHO are compared with those previously reported
3 7
F
4
9
for other fluorinated and hydrogenated aldehydes in Table
6
,15,17–21
2
.
Comparison of the wavelengths for the maximum
absorption cross sections of CF CHO and C F CHO shows a
2 5
3
red shift when a CF
CF
hydes C
length appears to have little effect on the wavelength of the
maximum absorption which is at 309 nm for both C CHO
and C CHO. The cross sections of the C to C fluoroalde-
2
group is added to CF
3
CHO, (300 nm for
3 2 5
CHO and 308 nm for C F CHO). For the longer alde-
x
F
2x+1CHO (x = 3 or 4) an increase in the chain
3 7
F
4
F
9
1
4
hydes at the maximum absorption increase gradually with the
increase in the length of the fluorinated chain, Table 2. It is
reasonable to suggest from the experimental absorption data
for C to C fluoroaldehydes that the absorption maxima for
Fig. 2 Plots of k vs. 1/T. The solid lines represent the two parameter
least-squares fits to the individual data points for each FAL. The error
bars of the individual point are 2s and do not include the estimated
systematic errors.
1
4
4
204 | Phys. Chem. Chem. Phys., 2007, 9, 4200–4210
This journal is ꢃc the Owner Societies 2007

1
5
6
Fig. 3 (a) UV-Vis absorption cross sections of C
3 7
F CHO from this work and those of Hashikawa et al. and Chiappero et al. (b) UV-Vis
15
6
4 9
absorption cross sections of C F CHO from this work and those of Hashikawa et al. and Chiappero et al.
fluoroaldehydes with chain lengths longer than C
4
will also be
chain length. The red shift of the maxima of the fluorinated
compared to the hydrogenated analogues could play a
significant role in their atmospheric photochemistry.
at approximately 310 nm. However, it is not possible to predict
whether the absorption cross sections of the longer chain
aldehydes will continue to increase with increasing chain
3 2 5 2
The UV absorption spectrum of CF (CF ) CH CHO was
length as observed along the series C
1
to C
4
.
investigated at room temperature and at 328 K. The observed
spectra at both temperatures were virtually identical and
showed an absorption maximum at 283 nm, Fig. 4a. The
absorption cross section at this wavelength was determined to
The maximum absorption of the fluorinated aldehydes is
shifted towards the red by about 15 nm compared to their
hydrogenated analogues, (smax(fluorinated) B310 nm and
smax(hydrogenated) B295 nm, Table 2). The maximum value
ꢂ20
2
ꢂ1
be (5.4 ꢀ 0.4) ꢁ 10
cm molecule . The quoted error
of the absorption cross section for CF CHO is slightly smaller
3
includes the standard deviation in the data of 3% together
with an uncertainty of 5% in the estimate of the concentra-
tions. The shape of the spectrum is similar to the fluoroalde-
than that for CH CHO, however values of smax(fluorinated)
3
2 4
are somewhat higher than smax(hydrogenated) for the C to C
compounds. The difference in the value of the maximum
absorption cross sections for the fluoroaldehydes and hydro-
genated aldehydes increases by approximately 10–40% in
x
hydes, C F2x+1CHO (x = 1–4), with the maximum
absorption shifted to a shorter wavelength and with a lower
absorption cross section. The stated purity of the sample of
going from C
2
to C
4
. Thus, while there is a gradual increase
3 2 5 2
CF (CF ) CH CHO was only 97%, and therefore a possible
in absorption with chain length for the fluoroaldehydes, the
absorption of the unsubstituted compounds remains almost
constant for the C to C aldehydes (Table 2). This observa-
contribution to the measured cross sections from highly
absorbing impurities could not be excluded. Chiappero
6
et al. have recently measured the UV spectrum of this
2
6
tion might indicate that the fluorinated aldehydes with chain
aldehyde, and while the shape and position of the absorption
maximum are similar to those found in this work, the reported
maximum value of the absorption cross section is at ca.
lengths longer than C may also have spectra similar to that of
4
C F CHO with little change in their absorption maxima with
4
9
6
CHO from this work and Chiappero et al. (b) The absorbance of C
Fig. 4 (a) UV-Vis absorption cross sections of C
6
F13CH
2
6
F
13CH
2
CHO at
283 nm (maximum) under different pressures (this work).
This journal is ꢃc the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4200–4210 | 4205

x x
Table 2 Maximum UV absorption of perfluorinated aldehydes (C F2x+1CHO) and their hydrogenated analogues (C H2x+1CHO)
C
C
x
F2x+1CHO
2
ꢂ1
x
x
H
2x+1CHO
l
max/nm
s/cm molecule
Reference
Hashikawa et al.
Chiappero et al.
ꢂ20
ꢂ20
15
1
CF
3
CHO
300
(3.10 ꢀ 0.18) ꢁ 10
ꢂ20
ꢂ20
6
2
.89 ꢁ 10
17
CH
C
3
CHO
CHO
290
308
4.86 ꢁ 10
Sander et al.
Hashikawa et al.
15
2
3
2
F
5
(6.25 ꢀ 0.36) ꢁ 10
ꢂ20
ꢂ20
6
5
.86 ꢁ 10
Chiappero et al.
Martinez et al.
18
C
C
2
3
H
5
CHO
CHO
290
309
5.88 ꢁ 10
ꢂ20
F
7
(8.1 ꢀ 0.6) ꢁ 10
This work
Hashikawa et al.
ꢂ20
15
(
8
8.96 ꢀ 0.52) ꢁ 10
ꢂ20
ꢂ20
6
.1 ꢁ 10
Chiappero et al.
Tadic et al.
19
C
C
3
4
H
7
CHO
CHO
294
309
6.03 ꢁ 10
ꢂ20
4
F
9
(9.4 ꢀ 0.7) ꢁ 10
This work
Hashikawa et al.
ꢂ20
15
(
9
10.9 ꢀ 0. 6) ꢁ 10
ꢂ20
ꢂ20
ꢂ20
6
.4 ꢁ 10
Chiappero et al.
Tadic et al.
19
C
C
C
4
5
6
H
H
H
9
CHO
11CHO
13CHO
294
294
6.14 ꢁ 10
6.08 ꢁ 10
20
5
6
a
Plagens et al.
Tang and Zhu
ꢂ20
a
( )
21
Between 290 and 300
Value at 295 nm: 6.62 ꢁ 10
Values given with 5 nm intervals.
ꢂ20
2
cm molecule .
ꢂ1
3
00 nm with a value of (13.3 ꢀ 1.5) ꢁ 10
401. The actinic flux was taken from Finlayson-Pitts and
2
2
This value is over a factor of two higher than that found in the
present work. The disagreement between the two studies can
probably be explained in terms of the difficulties in handling
this aldehyde due to its low vapour pressure and purity.
Pitts assuming a photolysis quantum yield for dissociation
of unity, and hence the values of Jcalc are the maximum
possible photolytic dissociation rates. The calculated values
ꢂ1
ꢂ4
ꢂ4
ꢂ4
of Jcalc are in (s ) 5.6 ꢁ 10 , 6.6 ꢁ 10 and 1.3 ꢁ 10 for
CHO, C CHO and CF (CF CH CHO, respectively.
6
Chiappero et al. were only able to determine the spectrum
C
3
F
7
4
F
9
3
2
)
5
2
at one pressure, (around 2 Torr), of the aldehyde, while in the
present work it was possible to determine the absorption cross
sections at ten different pressures over the range 0.6–2.0 Torr
by static and dynamic methods between room temperature
and 328 K. An example of the absorbance versus pressure plot
Photolysis of C
3 7 4 9 3 2 5
F CHO, C F CHO and CF (CF ) CHO under
sunlight conditions at EUPHORE
The photolysises were conducted at EUPHORE under summer
conditions in June 2005. Analysis of the fluorinated species were
made in situ by FTIR spectroscopy (using mainly the 1120–1280
4
6
is shown in Fig. 4b. Sellevag et al. and Chiappero et al. have
both measured the UV absorption spectrum of CF CH CHO
and found it is comparable to that of the unsubstituted
analogue CH CH CHO, but with a slightly lower value for
ꢂ1
3
2
cm band), and after sampling by gas chromatography em-
ploying both electron capture and mass spectrometric detectors.
Analysis was also performed by solid phase micro-extraction
3
2
the maximum absorption cross section at around 290 nm,
lower by E26%). The absorption data obtained in the
present study for CF (CF CH CHO indicates that the wave-
length and value of the maximum absorption cross section are
little changed from those in CF CH CHO, and are similar to
those reported for the unsubstituted aldehyde C 13CHO.
Thus, it appears that the absorption cross sections for fluoro-
aldehydes with general structure CF (CF CH CHO do not
(
SPME), followed by gas chromatography with mass spectro-
(
metric detection. The sunlight intensity was measured from the
3
2
)
5
2
photolysis frequency of NO (J(NO )). J(NO ) was calculated
2
2
2
from solar flux intensity measurements of the spectroradiometer
and reported data on the absorption cross section and quantum
3
2
6
H
efficiency for the photolysis of NO
reactants by dilution in the chamber was determined from the
decay of the inert tracer SF , which was monitored by FTIR
spectroscopy. To avoid interference from the reactions of OH
and CF O radicals with the aldehydes, isoprene was employed
as a scavenger for these species. The photolysis coefficient, Jobs
by sunlight. The loss of
2
3
2
)
x
2
6
change significantly with an increase in chain length. In
contrast, although the wavelength of maximum absorption
3
of fluoroaldehydes with structure CF (CF ) CHO is virtually
3 2 x
,
independent of chain length, the strength of absorption in-
creases markedly with chain length.
of the aldehydes was then derived from eqn (III) which includes
loss terms by photolysis, dilution and wall reaction:
The measured UV absorption cross sections enabled the
atmospheric photolysis frequency Jcalc of the flurorinated
aldehydes to be calculated using the expression:
lnð½FALꢄ =½FALꢄ Þ ¼ ðJobsðFALÞ þ kdil þ kwallÞt
ðIIIÞ
0
t
where Jmeas = Jobs(FAL) + kdil + kwall
R
Loss at the walls was found to be negligible in dark
reactions carried out prior to sunlight irradiation.
Jcalc ¼ sðlÞFðlÞIðlÞdl ðl4290nmÞ
ðIVÞ
where s(l), s(l) and I(l) are the absorption cross section, the
quantum yield of dissociation and the actinic flux at a wave-
length l. Calculations were carried over the wavelength range
Photolysis of C F CHO, C F CHO and CF (CF ) CHO
3
3
7
4
9
2 5
A single run for C
3 7 4 9
F CHO and two runs for each of C F CHO
2
90–360 nm with conditions similar to those of the experi-
and CF (CF ) CHO were conducted in order to measure their
3 2 5
ments performed at EUPHORE in June 2005: an albedo of
photolysis rates. The initial concentrations of the aldehydes
were around 200 ppb for C F CHO and C F CHO and
8
0%, at sea level, 401N latitude, June 1st for zenith angles of
3
7
4 9
4
206 | Phys. Chem. Chem. Phys., 2007, 9, 4200–4210
This journal is ꢃc the Owner Societies 2007

9
8–144 ppb for CF
3
(CF
2
)
5
CHO. The photolysis frequency of
ꢂ3 ꢂ1
losses of the aldehydes were considerably higher than the leak
rate, as measured by the loss of the inert tracer SF . It is likely
NO , J(NO
2
2
), was 9 ꢁ 10
s
in all the experiments. A
6
summary of the initial experimental conditions and the results
of the experiments, which were carried out in the middle of the
day, are given in Table 3. Fig. 5a, b and c show the concen-
tration–time profiles of the total decays of C F CHO,
that in situ analysis of the fluoroaldehydes by FTIR spectro-
scopy provides the most accurate measurement of the photo-
lytic removal of these compounds and difficulties uncounted in
the sampling techniques are more likely to lead to errors. The
reported values of the observed rate coefficients, kobs(FAL),
for photolytic loss of the fluoroaldehydes were derived from
the total measured loss rate coefficient, kmeas, obtained by the
loss of the aldehydes by the in situ FTIR method and corrected
for the rate coefficient for leakage in the particular experiment.
3
7
C
F
4 9
CHO and CF
ments at EUPHORE plotted in the form of eqn (III). The loss
of SF due to leakage from the chamber is also shown in the
3 2 5
(CF ) CHO during the photolysis experi-
6
plots. Large discrepancies were apparent between the results
obtained for the loss of the aldehydes when measured with the
various analytical techniques employed (Fig. 5 and Table 3).
This reflects the analytical difficulties in handling these types of
compounds. For the single run conducted on the photolysis of
C F CHO the data seems to be quite reproducible with three
6
The error limits are 2s and include the loss of SF and the
aldehyde:
JobsðC3F7CHOÞ ¼ ð1:3 ꢀ 0:6Þ ꢁ 10ꢂ5
JobsðC4F9CHOÞ ¼ ð1:9 ꢀ 0:8Þ ꢁ 10ꢂ5
JobsðCF3ðCF Þ CHOÞ ¼ ð0:6 ꢀ 0:3Þ ꢁ 10
s
ꢂ1
3
7
of the analytical techniques. However, the decay of this
compound when determined by the SPME/GC-MS method
was significantly higher than determined using the other
analytical procedures. In the photolysis experiments with
ꢂ1
s
ꢂ5 ꢂ1
s
2
5
4 9
C F CHO the removal rates obtained by FTIR spectroscopy
were similar in both experiments, while the rate of decay as
determined with the GC-ECD method in the first experiment
was slightly higher than that measured by FTIR spectroscopy.
If the values obtained employing the sampling techniques are
included, the errors reported for the Jobs values are considerably
higher. Several experiments were performed at EUPHORE to
investigate the photolysis of CF (CF ) CH CHO. The results
4 9
In the second photolysis experiment on C F CHO the three
3
2 5
2
sampling techniques, GC-ECD, GC-MS and SMPE/GC-MS
all showed reduced decay rates compared to that determined
by FTIR spectroscopy. Two studies on the photolysis of
CF (CF ) CHO were performed. The measured loss rate of
were not conclusive since the overall removal rate of the
aldehyde was only marginally faster than the measured leak
rate in the experiments.
3
2 5
The values of the photolytic rate coefficients determined for
the three fluoroaldehydes investigated in this work can be used
to derive the photolytic atmospheric lifetimes of these com-
pounds. The calculated lifetimes were obtained from the
expression tobs = 1/kobs(FAL) and gave: tobs(C F CHO) =
the aldehyde using FTIR spectroscopy in the first run was in
reasonable agreement with the results from the data obtained
from GC-ECD and GC-MS measurements, but the decay rate
determined by the SMPE/GC-MS method gave a significantly
higher loss rate. In the second experiment both the GC-MS
and SMPE/GC-MS results were much higher than the rate
determined by the GC-ECD sampling technique and also
higher than the loss of the aldehyde in the first photolysis
experiment.
3
7
(21 ꢀ 10) h, t (C F CHO) = (15 ꢀ 7) h and t (CF (CF ) -
obs
4
7
obs
3
2 5
CHO = (46 ꢀ 23) h. Hence, all three aldehydes investigated
have photolytic atmospheric lifetimes under the sunlight con-
ditions in which the experiments were performed of close to
one day. It is of some importance to note that the lifetime
It is clear from Fig. 5 that, despite the large errors in the
measurements of the total removal rates of the aldehydes in
the photolysises carried out at EUPHORE, the photolytic
of the inert tracer, SF , due to loss by leakage from the
6
EUPHORE chamber in the present series of experiments
varies over the range 19 to 60 d. Thus, the rate of photolytic
) = 9 ꢁ 10ꢂ3
s
ꢂ1
),
Table 3 Photolysis of C
3
F
7
CHO, C
4
F
experimental conditions and results (using isoprene as a trap for OH and CF
9
CHO and CF
3
(CF
2
)
5
CHO under sunlight conditions at EUPHORE (J(NO
2
12
ꢂ3
)
3
O [Isoprene] = 1.8 ꢁ 10 molecule cm
C
4
F
9
CHO CF (CF ) CHO
2 5
3
C
2
3
F
2 06 2005
7
CHO d
15 06 2005
16 06 2005
17 06 2005
20 06 2005
ꢂ3
12
12
12
12
12
[
Irradiation period
C
x
F
2x+1CHO]
0
/molecule cm
4.95 ꢁ 10
2.44 ꢁ 10
4.93 ꢁ 10
3.58 ꢁ 10
2.39 ꢁ 10
2h30
0.5 ꢀ 0.2
1h30
1.5 ꢀ 0.3
2h30
1.2 ꢀ 0.3
4 h
0.7 ꢀ 0.2
5 h
0.6 ꢀ 0.2
ꢂ5 ꢂ1
k
dilution/ꢁ10
s
ꢂ5 ꢂ1
Technique
FTIR
GC-ECD
J
meas/ꢁ10
s
a
1.8 ꢀ 0.2
1.9 ꢀ 0.1
3.3 ꢀ 0.6
3.3 ꢀ 0.5
1.6 ꢀ 0.1
2.4 ꢀ 0.5
1.9 ꢀ 0.2
1.3 ꢀ 0.2
1.3 ꢀ 0.1
3.6 ꢀ 1.2
1.7 ꢀ 0.4
4.3 ꢀ 0.2
1.4 ꢀ 0.1
3.1 ꢀ 1.2
2.3 ꢀ 0.5
d
a
SMPE/GC-MS
GC-MS
2.7 ꢀ 0.8 (1.8 ꢀ 0.3)
1.7 ꢀ 0.4
a
ꢂ5 ꢂ1 b
J
J
a
meas/ꢁ10
s
1.8 ꢀ 0.2
3.3 ꢀ 0.6
1.8 ꢀ 0.6
meas represents the averaged values obtained with different analytical techniques and the error limits are taken to encompass the
3.3 ꢀ 0.5
1.3 ꢀ 0.2
2.3 ꢀ 1.2
ꢂ5 ꢂ1 c
obs(FAL)/ꢁ10
s
1.3 ꢀ 0.6
2.1 ꢀ 0.8
0.6 ꢀ 0.3
b
Not used.
J
c
d
extremes of the 2s error of the individual determinations.
using only the 4 last data (see Fig. 5a).
Jobs(FAL) derived from Jmeas ꢂ kdilution using FTIR data only (see text). Obtained
This journal is ꢃc the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4200–4210 | 4207

Fig. 5 (a) Photolysis of C
3 7 4 9
F CHO, loss of the aldehyde versus photolysis time at EUPHORE. (b) Photolysis of C F CHO, loss of the aldehyde
versus photolysis time at EUPHORE. (c) Photolysis of CF
3
(CF
2
)
5
CHO, loss of the aldehyde versus photolysis time at EUPHORE.
removal of a compound in experiments carried out at the
EUPHORE facility would be within experimental error of the
leak rate for species having lifetimes longer than four to five
days and hence can not be determined with any degree of
accuracy using this facility. The atmospheric photolysis rates
of the fluoroaldehydes depend on the intensity and spectral
distribution of the incident sunlight. As a consequence, it is
useful to quote the photolysis rates relative to a conventional
The estimated quantum yields for photodissociation obtained,
3 7
seff = Jobs/Jcalc, were 0.023 for C F CHO, 0.029 for
C
4
F
9
3
2 5
CHO and 0.046 for CF (CF ) CHO. The errors asso-
ciated with these quantum yield values are largely those arising
from the measurements of Jobs and are probably about ꢀ0.01.
The effective quantum yield for photodissociation of
CF (CF ) CHO was estimated by assuming the absorption
3
2 5
cross sections for this compound were the same as those for
C F CHO, and hence the value of J was the same as that
measure of solar light intensity such as J(NO ). The average
2
4
9
calc
value of J(NO ) measured during the course of the present
2
determined for C F CHO.
4 9
ꢂ3 ꢂ1
studies was 9 ꢁ 10
s
and hence a ratio of Jobs(FAL)/
ꢂ3
The atmospheric lifetimes and quantum yields determined
ꢂ5
ꢂ3
J(NO
2
) B1.5 ꢁ 10 / 9 ꢁ 10 = 1.7 ꢁ 0 was obtained for
in this work for C
3
F
7
CHO, C
4
F
9
CHO and CF
3
(CF
2
)
5
CHO
the aldehydes investigated. The present series of photolysis
experiments on the fluoroaldehydes were carried out at
EUPHORE in the middle of the day, under relatively clear skies
can be compared with those previously reported for CF
3
CHO
4
also from studies carried out at EUPHORE by Sellevag et al.
In their photolysis experiments with CF
3
CHO the total re-
in June. The maximum photolysis rate occurs at solar noon in the
ꢂ2 ꢂ1
moval rate coefficient was close to the leak rate coefficient and
summer months where J(NO
2
) is typically about 1 ꢁ 10
s
.
an upper limit for the photolytic removal rate coefficient for
ꢂ6 ꢂ1
Under these conditions, Jobs(FAL) will be around one day.
CF CHO of approximately 1 ꢁ 10
s
with a quantum yield
3
From the experiments reported in this work it is possible to
derive the quantum yield for photodissociation of the alde-
hydes by sunlight from the ratio of the measured photolysis
rate coefficient, Jobs, to the maximum theoretical loss rate,
of o0.20 were derived from the data. Thus, a lower limit for
the lifetime of CF CHO under the sunlight conditions em-
3
ployed of around 10 d can be estimated from the results. This
result is surprising when compared with the results of the
present work on the longer chain fluoroaldehydes, where the
photolysis lifetimes were measured to be around one day, and
Jcalc. Jcalc was calculated from the actinic flux corresponding to
that at EUPHORE, the measured absorption cross sections of
the fluoroaldehydes and assuming a quantum yield of unity
across the atmospheric range of absorption of the aldehydes.
3
suggests that the photolytic loss of CF CHO in this study was
4
underestimated. Sellevag et al. have also measured the
4
208 | Phys. Chem. Chem. Phys., 2007, 9, 4200–4210
This journal is ꢃc the Owner Societies 2007

6
ꢂ3 25
yields an estimated
photolytic lifetime of CHF
2
CHO at EUPHORE and reported
ꢂ5 ꢂ1
concentration of 10 radicals cm
lifetime for C 2x+1CHO of around 20 d. The photolysis
a value of the photolytic rate coefficient of 2.82 ꢁ 10
s
x
F
with an estimated quantum yield for photodissociation of 0.30.
lifetime under atmospheric degradation of these aldehydes was
found to be close to one day and hence the dominant loss
process for the fluoroaldehydes in the gas phase will be by
photolysis. When a soluble species is formed fairly uniformly
in the atmosphere, which will be the case for short chain length
The photolysis lifetime for CHF CHO obtained from this work
2
of about 10 h is more in line with the lifetimes determined in the
6
present investigation on fluoroaldehydes. Chiappero et al. have
recently determined the quantum yields for photodissociation
of CF
lysis at 308 nm in 700 Torr of N
CF CHO), 0.08 (C CHO) and 0.04 (CF
used to estimate the rates of photolysis of CF CHO, C
CHO, C CHO, CF CH CHO and CF (CF CH
within the troposphere. The calculated photolytic lifetimes
for C F
3
CHO, C
4
F
9
CHO and CF
3
CH
2
CHO following photo-
. The values of 0.17
CH CHO) were
CHO,
CHO
x
C F2x+1CHO, the average lifetime for uptake into cloud
2
droplets is of the order of 15–20 d and hence this loss process
will only be a minor degradation pathway. It is concluded
from the available data that the major atmospheric removal
pathway for fluoroaldehydes will be by photolysis and that
(
3
F
4 9
3
2
3
2 5
F
C
3
F
7
F
4 9
3
2
3
2
)
5
2
x
under low NO conditions this pathway may be a source of
CHO were o2 d in reasonable agreement with
2x+1
fluorinated carboxylic acids in the troposphere.
x
the experimental photolysis lifetimes found in this study under
atmospheric conditions. The lifetimes estimated for
C F CH CHO of 10–15 d are in line with the value of
x 2x+1 2
Acknowledgements
approximately 10 d found experimentally at EUPHORE for
4
CF CH CHO by Sellevag et al.
3 2
We acknowledge The Telomer Research Program, the French
National Programme for Atmospheric Chemistry (LEFE-
´
n CEAM
CHAT) and EUROCHAMP for support. Fundacio
Conclusions and atmospheric implications
is supported by the Generalitat Valenciana and Fundacion
´
BANCAIXA.
x
Reaction of OH with C F2x+1CHO leads to the formation of
the C F2x+1C(O)O2 radical, which in the presence of NOx
x
produces the corresponding fluoroalkyl radical. Subsequent
stepwise degradation of these radicals eventually leads to the
References
generation of CF
2x+1C(O)O
source of fluorinated carboxylic acids. Wallington and co-
2
O. At low levels of NO
x
, the reaction of the
1 D. A. Ellis, J. W. Martin, S. A. Mabury, M. D. Hurley, M. P.
Sulbaek Andersen and T. J. Wallington, J. Environ. Sci. Technol.,
C
x
F
2
radicals with HO radicals can provide a
2
2
T. Kelly, V. Bossoutrot, I. Magneron, K. Wirtz, J. Treacy, A.
003, 37, 3816.
2
3
2
3
workers have also suggested that fluorinated carboxylic acids
may be generated by the reactions of C radicals with
Mellouki, H. Sidebottom and G. Le Bras, J. Phys. Chem. A, 2005,
1
09, 347.
x 2x+1 2
F O
M. D. Hurley, J. A. Misner, J. C. Ball, T. J. Wallington, D. A.
Ellis, J. W. Martin, S. A. Marbury and M. P. Sulbaek Andersen, J.
Phys. Chem. A, 2005, 109, 9816.
alkyl peroxy radicals. The molecular channel of this reaction
produces a fluorinated alcohol, which will eliminate HF to
form the corresponding acetyl fluoride. Uptake of these
fluorides into water droplets leads to acid formation. Photo-
4 S. R. Sellevag, T. Kelly, H. Sidebottom and C. J. Nielsen, Phys.
Chem. Chem. Phys., 2004, 6, 1243.
5
S. R. Sellevag, Y. Stenstrom, T. Helgaker and C. J. Nielsen, J.
Phys. Chem., 2005, 109, 3652.
¨
6 M. S. Chiappero, F. E. Malanca, G. A. Arguello, S. T. Wool-
lysis of C F
x
CHO in the atmosphere will produce C F2x+1
x
2x+1
radicals, which under low NO conditions generate acids as
x
dridge, M. D. Hurley, J. C. Ball, T. J. Wallington, R. L. Waterland
and R. C. Buck, J. Phys. Chem. A, 2006, 110, 11944.
outlined above. Thus, both the atmospheric reaction of
x x
C F2x+1CHO with OH and photolysis at low NO concentra-
7
A. Mellouki, S. Te
2
´
ton and G. Le Bras, Int. J. Chem. Kinet., 1995,
tions can lead to the formation of fluorinated carboxylic acids.
Short chain fluoroaldehydes are highly soluble in water and
following incorporation into water droplets will rapidly hy-
drate to form the corresponding hydrates, and in aqueous
media further oxidation will produce fluorinated acids. If the
acid is not formed prior to evaporation of the cloud droplet,
gaseous hydrates will be generated. Photolysis of the hydrates is
negligible and reaction with OH radicals leads to formation of
7, 791.
8 K. H. Becker, The European Photoreactor ‘‘EUPHORE’’, final
report, EC contract EV5V-CT92-0059, University of Wuppertal.
1
996.
9
C. Kelly, H. W. Sidebottom, J. Treacy and O. J. Nielsen, Chem.
Phys. Lett., 1994, 218, 29.
10 M. P. Sulbaek Andersen, O. J. Nielsen, M. D. Hurley, J. C. Ball, T.
J. Wallington, J. E. Stevens, J. W. Martin, D. A. Ellis and S. A.
Mabury, J. Phys. Chem. A, 2004, 108, 5189.
1 M. P. Sulbaek Andersen, M. D. Hurley, T. J. Wallington, J. C.
Ball, J. W. Martin, D. A. Ellis, S. A. Mabury and O. J. Nielsen,
Chem. Phys. Lett., 2003, 379, 28.
2 M. T. Rayez, J. C. Rayez, T. Berces and G. Lendvay, J. Phys.
Chem., 1993, 21, 5570.
3 I. Smith and A. R. Ravishankara, J. Phys. Chem. A, 2002, 106,
4798.
1
24
fluorinated acids. The long chain fluoroaldehydes are likely to
be relatively insoluble in cloud droplets but will behave like
1
1
1
x 2
surfactants and form the hydrates C F2x+1CH(OH) at the
surface of the water droplets. Evaporation of the droplets
followed by reaction of the gaseous hydrates with OH may
4 T. J. Wallington, M. D. Hurley, J. Xia, D. J. Wuebbles, S. Sillman,
A. Ito, J. E. Penner, D. A. Ellis, J. Martin, S. A. Mabury, O. J.
Nielsen and M. P. Sulbaek Andersen, Environ. Sci. Technol., 2006,
40, 924.
produce the acids C
suggested for the short chain fluorinated hydrates.
The atmospheric lifetimes of C 2x+1CHO due to removal
x
F2x+1C(O)OH in a similar manner to that
x
F
by reaction with OH radicals and by photolysis may be
estimated from the data obtained in this investigation. Using
15 Y. Hashikawa, M. Kawasaki, R. L. Waterland, M. D. Hurley, J.
C. Ball, T. J. Wallington, M. P. Subaek Andersen and O. J.
Nielsen, J. Fluorine Chem., 2004, 125, 1925.
16 J. G. Calvert and J. N. Pitts, Jr, Photochemistry, Wiley, New York,
1966.
the averaged rate coefficients k(OH + C F2x+1CHO) =
x
3
ꢂ1 ꢂ1
6
ꢁ 10ꢂ13 cm molecule
s and a global averaged OH
This journal is ꢃc the Owner Societies 2007
Phys. Chem. Chem. Phys., 2007, 9, 4200–4210 | 4209

1
7 S. P. Sander et al., Chemical Kinetics and Photochemical Data for
Use in Atmospheric Studies, JPL Publication, 2003, vol. 14, http://
jpldataeval.jpl.nasa.gov/.
22 B. J. Finlayson-Pitts and J. N. Pitts, Jr, Chemistry of the upper and
lower atmosphere, Academic Press, London, 2000.
23 D. A. Ellis, J. W. Martin, A. O. DeSilva, S. A. Mabury, M. D.
Hurley, M. P. Subaek Andersen and T. J. Wallington, Environ. Sci.
Technol., 2004, 38, 3316.
24 M. P. Sulbaek Andersen, A. Toft, O. J. Nielsen, M. D. Hurley, T.
J. Wallington, H. Chishima, K. Tonokura, S. A. Mabury, J. W.
Martin and D. A. Ellis, J. Phys. Chem. A, 2006, 110, 9854.
25 R. G. Prinn, J. Huang, R. F. Weiss, D. M. Cunnold, P. J. Fraser,
P. G. Simmonds, A. McCulloch, P. Salameh, S. O’Doherty, R. H.
J. Wang, L. Porter and B. R. Miller, Science, 2001, 292, 1882.
1
1
2
8 R. D. Martinez, A. A. Buitrago, N. W. Howel, C. H. Hearn and J.
A. Joens, Atmos. Environ., 1992, 26A, 785.
9 J. Tadic, I. Juranic and G. K. Moortgat, J. Photochem. Photobiol.,
A, 2001, 143, 169.
¨
0 H. Plagens, Broske, M. Spittler, L. Ruppert, I. Barnes and K. H.
Becker, Second Workshop of the EUROTRAC-2 Subproject Che-
mical Mechanism Development, Karlsruhe, Germany, 1998.
1 Y. Tang and L. Zhu, J. Phys. Chem. A, 2004, 108, 8307.
2
4
210 | Phys. Chem. Chem. Phys., 2007, 9, 4200–4210
This journal is ꢃc the Owner Societies 2007