Atmospheric Chemistry of HFC-227ca: Spectrokinetic Investigation of the CF3CF2CF2O2 Radical, Its Reactions with NO and NO2, and the Atmospheric Fate of the CF3CF2CF2O Radical

Article abstract of DOI:10.1021/jp952769h

A pulse radiolysis technique was used to study the UV absorption spectrum of CF3CF2CF2O2 radicals, at 230 nm ? = (3.2 +/- 0.4)E-18 cm² molecule^-1.Rate constants for reactions of CF3CF2CF2O2 radicals with NO and NO2 were k₃ > 7E-12 cm³ molecule^-1 s^-1 and k₄ = (7.6 +/- 2.4)E-12 cm³ molecule^-1 s^-1, respectively.The rate constant for reaction of F atoms with CF3CF2CF2H was determined by using an absolute rate technique to be (3.6 +/- 1.3)E-13 cm³ molecule^-1 s^-1.The atmospheric fate of CF3CF2CF2O radicals is decomposition via C-C bond scission to give CF3CF2 radicals and CF2O.In one bar of SF6 at 296 K, CF3CF2CF2O radicals decompose with a rate > 1.5E5 s^-1.In their turn CF3CF2 radicals are converted into CF3CF2O radicals which also decompose via C-C bond scission.As part of this work a relative rate method was used to measure k(Cl + CF3CF2CF2H) = (3.4 +/- 0.7)E-16 and k(F + CF3CF2CF2H) = (3.2 +/- 0.8)E-13 cm³ molecule^-1 s^-1.The results are discussed with respect to the atmospheric chemistry of HFC-227ca.

Full text of DOI:10.1021/jp952769h

6572
J. Phys. Chem. 1996, 100, 6572-6579
Atmospheric Chemistry of HFC-227ca: Spectrokinetic Investigation of the CF3CF2CF2O2
Radical, Its Reactions with NO and NO2, and the Atmospheric Fate of the CF3CF2CF2O
Radical
Anders M. B. Giessing,† Anders Feilberg,† Trine E. Møgelberg, Jens Sehested,* and
Merete Bilde
Section for Chemical ReactiVity, EnVironmental Science and Technology Department,
Risø National Laboratory, DK-4000 Roskilde, Denmark
Timothy J. Wallington*
Ford Motor Company, Ford Research Laboratory, SRL-E3083,
Dearborn, P.O. Box 2053, Michigan 48121-2053
Ole J. Nielsen*
Ford Motor Company, Ford Forschungszentrum Aachen, Dennewartstrasse 25, D-52068 Aachen, Germany
ReceiVed: September 20, 1995; In Final Form: January 22, 1996X
A pulse radiolysis technique was used to study the UV absorption spectrum of CF3CF2CF2O2 radicals, at 230
nm σ ) (3.2 ( 0.4) × 10-18 cm2 molecule-1. Rate constants for reactions of CF3CF2CF2O2 radicals with
NO and NO2 were k3 > 7 × 10-12 cm3 molecule-1 s-1 and k4 ) (7.6 ( 2.4) × 10-12 cm3 molecule-1 s-1,
respectively. The rate constant for reaction of F atoms with CF3CF2CF2H was determined by using an absolute
rate technique to be (3.6 ( 1.3) × 10-13 cm3 molecule-1 s-1. The atmospheric fate of CF3CF2CF2O radicals
is decomposition via C-C bond scission to give CF3CF2 radicals and CF2O. In one bar of SF6 at 296 K,
CF3CF2CF2O radicals decompose with a rate > 1.5 × 105 s-1. In their turn CF3CF2 radicals are converted
into CF3CF2O radicals which also decompose via C-C bond scission. As part of this work a relative rate
method was used to measure k(Cl + CF3CF2CF2H) ) (3.4 ( 0.7) × 10-16 and k(F + CF3CF2CF2H) ) (3.2
( 0.8) × 10-13 cm3 molecule-1 s-1. The results are discussed with respect to the atmospheric chemistry of
HFC-227ca.
Introduction
OH radicals in the troposphere to produce a fluorinated alkyl
radical which will react rapidly with O2 to give CF3CF2CF2O2.
Recognition of the adverse impact of chlorofluorocarbons
(CFCs) on stratospheric ozone has led to an international effort
to replace CFCs with environmentally acceptable alternatives.1
Hydrofluorocarbons (HFCs) are one class of CFC replacements
used in refrigeration, air conditioning, foam blowing, and
cleaning applications. HFC-227ca (CF3CF2CF2H) is a potential
CFC replacement and has a chemical structure which is similar
to that of HFC-125 (CF3CF2H). HFC-125 is currently used in
blends with HFC-134a (CF3CH2F) and HFC-32 (CF2H2) as a
replacement for HCFC-22 (CHClF2). Global production of up
to 10 ktonnes per year of HFC-125 is expected2 while current
production of HFC-134a is approximately 150 ktonnes. There
are no available estimates for the future production of HFC-
227ca.
CF3CF2CF2H + OH f CF3CF2CF2 + H2O
CF3CF2CF2 + O2 + M f CF3CF2CF2O2 + M
(1)
(2)
The rate constant for reaction 1 is unknown; however, the
atmospheric lifetime of HFC-227ca can be estimated as 26 years
from the lifetime of CF3CF2H as discussed in a later section.
By analogy to other peroxy radicals,3 CF3CF2CF2O2 radicals
are expected to react with NO, NO2, HO2, or other peroxy
radicals in the atmosphere.
CF3CF2CF2O2 + NO f CF3CF2CF2O + NO2 (3a)
CF3CF2CF2O2 + NO + M f CF3CF2CF2ONO2 + M (3b)
The atmospheric chemistry of HFC-227ca is of interest for
two reasons. First, information concerning HFC-227ca is
needed to determine its environmental impact. Second, in light
of its structural similarity to HFC-125, information concerning
HFC-227ca sheds light on the atmospheric chemistry of HFC-
125 and indeed other HFCs. The main atmospheric loss
mechanism for CF3CF2CF2H is expected to be reaction with
CF3CF2CF2O2 + NO2 + M f CF3CF2CF2O2NO2 + M (4)
CF3CF2CF2O2 + HO2 f products
CF3CF2CF2O2 + R′O2 f products
(5)
(6)
* Authors to whom correspondence may be addressed.
† Present address: Roskilde University, Institute of Life Sciences and
Chemistry, P. O. Box 260, DK-4000 Roskilde, Denmark.
X Abstract published in AdVance ACS Abstracts, March 15, 1996.
A pulse radiolysis technique combined with UV-visible
absorption spectroscopy was used to determine the UV absorp-
tion spectrum of CF3CF2CF2O2 and to study the kinetics of
0022-3654/96/20100-6572$12.00/0 © 1996 American Chemical Society

Atmospheric Chemistry of HFC-227ca
J. Phys. Chem., Vol. 100, No. 16, 1996 6573
reactions 3, 4, and 6. In the case of reaction 6 we studied the
self reaction of the peroxy radical (R′ ) CF3CF2CF2). The fate
of the CF3CF2CF2O radical produced in reaction 3 was
determined by using a FTIR spectrometer coupled to a smog
chamber. The results are reported herein.
monitored to derive a value for the rate constant of its self-
reaction. Fourth, the kinetics of the reaction of the peroxy
radical with NO were determined by monitoring the kinetics of
NO2 formation following pulsed radiolysis of SF6/CF3CF2CF2H/
O2/NO mixtures. Fifth, similar experiments were performed
by using SF6/CF3CF2CF2H/O2/NO2 mixtures with the decay of
NO2 monitored to determine the kinetics of the reaction of CF3-
CF2CF2O2 radicals with NO2.
Experimental Section
FTIR Smog Chamber System. The FTIR (Fourier trans-
form infrared) system was interfaced to a 140-L Pyrex reactor.
Radicals were generated by the UV irradiation of mixtures of
13-30 mTorr of CF3CF2CF2H, 21-26 mTorr of NO, and 700-
715 mTorr of Cl2 in either 5 or 10 Torr total pressure of O2
diluent (760 Torr ) 1013 mbar). UV light was provided by
22 blacklamps, which emit light between 300 and 400 nm. This
light was used to produce Cl atoms from Cl2. Cl atoms were
used to initiate the reactions. The loss of reactants and the
formation of products were monitored by FTIR spectroscopy,
using an analyzing path length of 26 m and a resolution of 0.25
cm-1. Infrared spectra were derived from 32 co-added spectra.
CF3CF2CF2H and COF2 were monitored by using their char-
The two experimental systems used in the present work are
described in detail elsewhere.4,5 All experiments were per-
formed at 296 ( 2 K.
Pulse Radiolysis System. Radicals were generated by
irradiation of SF6/CF3CF2CF2H/O2 gas mixtures in a 1-L
stainless steel reaction cell with a 30-ns pulse of 2 MeV electrons
from a Febetron 705B field emission accelerator. SF6 was
always in great excess.
SF6 + 2 MeV e- f products
CF3CF2CF2H + F f CF3CF2CF2 + HF
CF3CF2CF2 + O2 + M f CF3CF2CF2O2 + M
(7)
(8)
(2)
acteristic features over the wavenumber range 700-2000 cm-1
.
Reference spectra were acquired by expanding known volumes
of reference materials into the reactor.
Transient absorptions were followed by multipassing the
output of a pulsed 150-W xenon arc lamp through the reaction
cell using internal White cell optics. Total path lengths of 80
and 120 cm were used. After exiting the cell, the light was
guided through a 1-m McPherson grating UV-vis monochro-
mator and detected with a Hamamatsu photomultiplier. The
spectral resolution was 0.8 nm. For the measurement of the
UV absorption spectrum of the peroxy radical CF3CF2CF2O2 a
Princeton Applied Research OMA-II diode array spectropho-
tometer was used in place of the photomultiplier. The system
consisted of the diode array, an image amplifier (type 1420-
1024HQ), a controller (type 1421), and a conventional PC
computer for handling and storage of the data. Wavelength
calibration was achieved by using a Hg pen ray lamp.
Reagent concentrations used were: SF6, 895-950 mbar; O2,
0-5 mbar; CF3CF2CF2H, 0-100 mbar; NO, 0.20-0.71 mbar;
and NO2, 0.17-1.00 mbar. All experiments were performed
at 296 K. Ultrahigh purity O2 was supplied by L’Air Liquide.
SF6 (99.9%) was supplied by Gerling and Holz. CF3CF2CF2H
(>97%) was obtained from PCR Inc. NO (99.8%) was obtained
from Messer Grieshem. NO2 (>98%) was obtained from Linde
Technische Gase. All reagents were used as received.
Prior to the present series of experiments the F atom yield
was determined by monitoring the transient absorbance at 260
nm due to methylperoxy radicals produced by pulse radiolysis
of SF6/CH4/O2 mixtures.6 Using a value of 3.18 × 10-18 cm2
molecule-1 for σ(CH3O2) at 260 nm7 the F atom yield was
determined to be (3.18 ( 0.31) × 1015 cm-3 at full radiolysis
dose and 1000 mbar of SF6. The quoted error on the F atom
yield includes both statistical (2 standard deviations) and
potential systematic errors associated with a 10% uncertainty
in σ(CH3O2). Errors are propagated by using conventional error
analysis methods and are (2 standard deviations if not otherwise
stated.
Results
Kinetics of the Reaction F + CF3CF2CF2H. The kinetics
of the reaction of F atoms with CF3CF2CF2H were studied
relative to the reaction of F atoms with CF3CCl2H by observing
the maximum absorbance following pulse radiolysis of mixtures
of 0-100 mbar of CF3CF2CF2H and 1-10 mbar of CF3CCl2H
with SF6 added to a total pressure of 1000 mbar. A relative
radiolysis dose of 0.32 was used. Under these circumstances
reactions 8 and 9 compete for F atoms:
F + CF3CF2CF2H f CF3CF2CF2 + HF
F + CF3CCl2H f CF3CCl2 + HF
(8)
(9)
The alkyl radical produced in reaction 9 (CF3CCl2) absorbs
strongly at 230 nm,8 whereas the radical produced in reaction
8 (CF3CF2CF2) absorbs weakly. The competition between
reactions 8 and 9 was studied by observing the dependence of
the maximum absorbance at 230 nm on the [CF3CF2CF2H]/
[CF3CCl2H] concentration ratio following pulsed radiolysis of
CF3CF2CF2H/CF3CCl2H/SF6 mixtures. In Figure 1 the maxi-
mum absorbance is plotted as a function of the [CF3CF2CF2H]/
[CF3CCl2H] concentration ratio. In the absence of CF3CF2CF2H
all F atoms are converted into CF3CCl2 radicals via reaction 9
and the transient absorption at 230 nm is substantial. When
CF3CF2CF2H is added, F atoms are scavenged by reaction 8
and the transient absorption attributed to CF3CCl2 radicals
decreases. The maximum absorbance decreases until the [CF3-
CF2CF2H]/[CF3CCl2H] ratio is approximately 25. Increasing
the ratio further has little impact on the maximum absorbance.
To obtain the rate constant ratio k8/k9 the following expression
was fitted to the data in Figure 1:
Five sets of experiments were performed. First, the rate
constant for reaction 8 was determined by observing the
dependence of the absorbance maximum on the [CF3CF2CF2H]/
[CF3CCl2H] concentration ratio following the pulse radiolysis
of SF6/CF3CF2CF2H/CF3CCl2H mixtures. Second, the diode
array was used to measure the UV absorption spectrum of the
CF3CF2CF2O2 radical following pulsed radiolysis of SF6/CF3-
CF2CF2H/O2 mixtures. Third, the decay of CF3CF2CF2O2 was
Amax ) {ACF CCl + ACF CF CF (k8/k9)[CF3CF2CF2H]/
3
2
3
2
2
[CF3CCl2H]}/{1 + (k8/k9)[CF3CF2CF2H]/[CF3CCl2H]}
Amax is the measured maximum absorbance, ACF CCl is the
3
2
expected absorbance if only CF3CCl2 radicals were produced,

6574 J. Phys. Chem., Vol. 100, No. 16, 1996
Giessing et al.
Figure 2. Transient absorption at 230 nm following the pulsed
radiolysis (full dose) of a mixture of 5 mbar of O2, 50 mbar of CF3-
CF2CF2H, and 945 mbar of SF6. An UV path length of 120 cm was
used.
Figure 1. Plot of maximum transient absorbance at 230 nm as a
function of the concentration ratio [CF3CF2CF2H]/[CF3CCl2H], fol-
lowing the radiolysis of mixtures of 0-100 mbar of CF3CF2CF2H, 10
mbar of CF3CCl2H, and 990 mbar of SF6. The UV path length was 80
cm and radiolysis dose was 32% of maximum. The smooth line and
the dotted lines are fit to the data. See text for details.
and ACF CF CF is the expected absorbance if only CF3CF2CF2
3
2
2
radicals were produced. The rate constant ratio k8/k9 was
derived by performing a three parameter fit to the data in Figure
1 in which the parameters ACF CCl , ACF CF CF , and k8/k9 were
3
2
3
2
2
varied simultaneously. The fit is shown as the solid line in
Figure 1 and was achieved with k8/k9 ) 0.30 ( 0.05. The dotted
lines in Figure 1 show the effects of varying k8/k9 by ( 0.05.
Using k9 ) (1.2 ( 0.4) × 10-12 9 gives k8 ) (3.6 ( 1.3) ×
10-13, in agreement with a value of k8 ) (3.2 ( 0.8) × 10-13
cm3 molecule-1 s-1 measured by using a relative rate technique
described in a subsequent section of the present paper.
Absorption Spectrum of CF3CF2CF2O2. Following the
pulse radiolysis (full dose) of mixtures of 5 mbar of O2, 50
mbar of CF3CF2CF2H, and 945 mbar of SF6 a rapid increase in
the absorption at 230 nm was observed followed by a slower
decay (see Figure 2). It seems reasonable to ascribe the UV
absorption resulting from radiolysis of SF6/CF3CF2CF2H/O2
mixtures to the formation of CF3CF2CF2O2 radicals. Before
measuring the absorption cross sections of the CF3CF2CF2O2
radical we need to consider possible interfering secondary
reactions.
Figure 3. Maximum transient absorbance as a function of radiolysis
dose at 230 nm following the pulse radiolysis of mixtures of 5 mbar
of O2, 50 mbar of CF3CF2CF2H, and 945 mbar of SF6. The solid line
is a linear fit to the low dose data, the dotted curve is a second order
regression of all the data to aid in visual inspection of the data trend.
TABLE 1: Measured UV Absorption Cross Sections
σ(CF3CF2CF2O2)
10-20 cm2
σ(CF3CF2CF2O2)
10-20 cm2
wavelength,
nm
wavelength,
nm
molecule-1
molecule-1
After the electron pulse reactions, (8) and (10) compete for
the available F atoms.
230
235
240
245
320
252
152
125
250
255
260
270
82
65
23
16
F + CF3CF2CF2H f CF3CF2CF2 + HF
F + O2 + M f FO2 + M
(8)
CF2CF2H, and 945 mbar of SF6 with the radiolysis dose varied
over an order of magnitude. In Figure 3 the maximum transient
absorbance observed at 230 nm using a path length of 120 cm
is plotted as a function of radiolysis dose. As seen from Figure
3, the maximum absorbance was linear up to 42% of maximum
radiolysis dose. At higher doses the observed absorbance was
smaller than expected from a linear extrapolation of the low
dose experiments. We ascribe the curvature to incomplete
conversion of F atoms into CF3CF2CF2O2 and FO2 radicals
caused by secondary radical-radical reactions, such as reactions
11-13. The solid line shown in Figure 3 is a linear least squares
fit of the low dose data. The slope is (0.515 ( 0.031).
Combining this result with three additional pieces of informa-
tion, (i) the F atom yield of (3.18 ( 0.31) × 1015 cm-3 (full
radiolysis dose and 1000 mbar of SF6), (ii) the conversion of F
atoms into 95% CF3CF2CF2O2 and 5% FO2 radicals, and (iii)
the absorption cross section for FO2 at 230 nm (σ230 ) 5.08 ×
10-18 cm2 molecule-1),6 we derive σ(CF3CF2CF2O2) at 230 nm
(10)
By use of k8 ) (3.6 ( 1.3) × 10-13 cm3 molecule-1 s-1 (see
previous section) and k10 ) (1.9 ( 0.3) × 10-13 cm3 molecule-1
s-1,6 it can be calculated that under our experimental conditions
95% of F atoms are converted into CF3CF2CF2O2 and 5% into
FO2 radicals.
We also need to consider possible interfering secondary
reactions such as
F + CF3CF2CF2 f products
F + CF3CF2CF2O2 f products
(11)
(12)
CF3CF2CF2O2 + CF3CF2CF2 f 2CF3CF2CF2O (13)
To check for these reactions, a set of experiments was
performed using mixtures of 5 mbar of O2, 50 mbar of CF3-

Atmospheric Chemistry of HFC-227ca
J. Phys. Chem., Vol. 100, No. 16, 1996 6575
Figure 5. Plot of 1/t1/2 for the self-reaction of CF3CF2CF2O2 as a
function of maximum absorbance at 250 nm.
Figure 4. Absorption spectrum of CF3CF2CF2O2 measured in this work
(solid line) compared to that of CF3CF2O2 (circles and dashed line).23
) (3.2 ( 0.4) × 10-18 cm2 molecule-1. A full spectrum of
CF3CF2CF2O2 absorption cross sections between 230 and 270
nm is shown in Figure 4. The spectrum was obtained by
recording the absorbance following pulse radiolysis (full dose)
of mixtures of 50 mbar of CF3CF2CF2H, 5 mbar of O2, and
945 mbar of SF6 using a diode array camera under the following
conditions: spectral resolution, 2 nm; gate, 20 µs; delay, 5 µs.
The observed absorbance values were corrected for FO2
absorbance by using the absorption cross sections measured by
Ellermann et al.6 and then scaled to σ230nm (CF3CF2CF2O2) )
formation of CF3CF2O2 radicals poses two problems in terms
of the interpretation of k14obs. First, since they absorb at the
monitoring wavelength of 250 nm (see Figure 4), the formation
of CF3CF2O2 radicals masks the loss of CF3CF2CF2O2 radicals.
Second, C2F5O2 radicals will react with CF3CF2CF2O2 radicals
hence complicating the kinetic analysis. At the present time
we are unable to correct k14obs for such complications and so
we are unable to derive a value for the true bimolecular rate
constant k14. The value of k14obs reported here is the observed
rate constant for decay of absorption at 250 nm.
Kinetics of the CF3CF2CF2O2 + NO Reaction. The
kinetics of reaction 3 were studied by measuring the transient
absorbance at 400 nm following pulse radiolysis of mixtures
of 100 mbar of CF3CF2CF2H, 5 mbar of O2, 0.20-0.71 mbar
of NO, and SF6 added to a total pressure of 1000 mbar.
3.2 × 10-18 cm2 molecule-1
.
The UV spectrum of CF3CF2CF2O2 measured in the present
work is compared to that for CF3CF2O2 in Figure 4. As
expected for such similar peroxy radicals their spectra are, within
the experimental uncertainties, indistinguishable.
Kinetics of the Self-Reaction of CF3CF2CF2O2. Pulsed
radiolysis of CF3CF2CF2H/O2/SF6 gas mixtures gave rise to an
increase in the transient absorbance at 250 nm caused by
formation of peroxy radicals. The maximum absorbance was
reached within a few microseconds after the electron pulse,
subsequently a slower decay in the transient absorbance was
observed. It seems reasonable to ascribe the decay in absor-
bance to the self reaction of the peroxy radicals.
CF3CF2CF2O2 + NO f CF3CF2CF2O + NO2 (3a)
CF3CF2CF2O2 + NO + M f CF3CF2CF2ONO2 + M (3b)
An increase in the transient absorbance following pulse
radiolysis was observed, and, since NO2 is known to absorb
significantly at 400 nm, we ascribe the observed increase in
absorbance to formation of NO2 in reaction 3a. This method
of measuring the kinetics of the reaction of peroxy radicals with
NO has been used extensively in our laboratory and is described
in detail elsewhere.10-13
CF3CF2CF2O2 + CF3CF2CF2O2 f products
(14)
The kinetic expression for reaction 14 is d[RO2]/dt )
-2k14obs[RO2]2, where k14obs is the observed second order rate
constant for reaction 14. The half-life of reaction 14 is defined
as t1/2 ) 1/(2k14obs[RO2]). Using Lambert-Beers law, we get
The transient absorbancies were fitted by using the first order
expression A(t) ) (Ainf - A0) exp(-k1stt) + Ainf with the fit
starting 3 µs after the electron pulse to allow time for formation
of the peroxy radicals. Figure 6A shows a typical transient
absorption, the smooth line is the first order fit with a pseudo-
first-order rate constant k1st ) 0.95 × 105 s-1. The bimolecular
rate constant was derived by plotting the pseudo-first-order rate
constants versus [NO]o as shown in Figure 7. Linear least
squares analysis of the data in Figure 7 gives k3 ) (8.1 ( 1.1)
× 10-12 cm3 molecule-1 s-1; the y-axis intercept is (0.9 ( 1.3)
× 105 s-1 and is not significant.
1/t1/2 ) 2.303Amax2k14obs/80σRO ). The half-life of reaction 14
2
can be obtained by fitting the following second order expression
to the observed decay in the transient absorbance: A(t) ) (A0
- Ainf)/(1 + 2k14obs(A0 - Ainf)t) + Ainf, where A(t) is the
measured absorbance at time t, A0 is the absorbance at time
zero and Ainf is the absorbance at infinite time. Since σRO has
2
been determined (see preceding section), k14obs can be derived
by plotting the reciprocal half-life versus Amax. This plot is
shown in Figure 5 with a linear least squares regression of the
data points. The different data points were obtained by using
different radiolysis doses. Figure 5 shows a plot of 1/t1/2 versus
the maximum absorbance attributable to CF3CF2CF2O2 radicals.
The yield of NO2 in these experiments, expressed in terms
of the initial F atom concentration, calculated by using σNO2
-
(400 nm)) 6 × 10-19 cm2 molecule-1 14
,
was 182 ( 16%.
Corrections were made for the consumption of F atoms by
reaction with O2 and NO, kF+NO ) 5.1 × 10-12 cm3 molecule-1
s-1 (average from refs 15 and 16). The fact that the NO2 yield
is greater than 100% can be explained by the formation of NO2
via reactions of NO with decomposition products of the alkoxy
radical CF3CF2CF2O (see subsequent section). The decomposi-
tion of CF3CF2CF2O leads to CF3CF2 radicals, which can add
The slope of the plot in Figure 5 equals 2k14obs2.303/(80σRO
-
2
(250)) and gives k14obs ) (1.0 ( 0.2) × 10-12 cm-3 molecule-1
s-1. The self reaction of CF3CF2CF2O2 radicals gives CF3CF2-
CF2O radicals, which, as we will show in a later section,
decompose rapidly to give CF3CF2 radicals and COF2. CF3-
CF2 radicals will add O2 to give CF3CF2O2 radicals. The

6576 J. Phys. Chem., Vol. 100, No. 16, 1996
Giessing et al.
Figure 8. Plot of k1st versus [NO2].
Kinetics of the CF3CF2CF2O2 + NO2 Reaction. The
kinetics of reaction 4 were studied by measuring the transient
absorbance at 400 nm following pulse radiolysis (53% of
maximum dose) of mixtures of 100 mbar of CF3CF2CF2H, 5
mbar of O2, and 0.17-1.0 mbar of NO2 with SF6 added to a
total pressure of 1000 mbar.
Figure 6. Transient absorbancies at 400 nm following pulsed radiolysis
of mixtures of 100 mbar of CF3CF2CF2H, 5 mbar of O2, 940 mbar of
SF6, and either 0.5 mbar of NO and 32% radiolysis dose (A) or 0.2
mbar of NO2 and 53% radiolysis dose (B). The UV path length was
120 cm.
CF3CF2CF2O2 + NO2 + M f CF3CF2CF2O2NO2 + M (4)
In these experiments a decay in absorbance was observed
following the radiolysis pulse. Since NO2 absorbs at 400 nm,
it seems reasonable to ascribe the decay to loss of NO2 through
reaction 4. This method of measuring the kinetics of the
reaction of peroxy radicals with NO2 has been used extensively
in our laboratory and is described in detail elsewhere.10,11,17,19
Figure 6B shows an example of the decrease in the transient
absorbance following pulse radiolysis (53% of maximum dose)
of 100 mbar of CF3CF2CF2H, 5 mbar of O2, and 0.2 mbar of
NO2, with SF6 added to a total pressure of 1000 mbar. The
transient absorbancies were fitted by using the expression [Abs]-
(t) ) [Abs](0) exp(-k1stt) + C where [Abs](t) + C and [Abs]-
(0) + C are the absorbance values at times t and 0, respectively,
due to NO2 and k1st ) k4[NO2]. The fit was started from 3 µs
after the electron pulse to allow time for formation of the peroxy
radicals. The resulting k1st values are plotted versus [NO2]o in
Figure 8. The line in Figure 8 is a linear least squares regression
of the data points, the slope gives k4 ) (7.6 ( 2.4) × 10-12
cm3 molecule-1 s-1. The y-axis intercept is not statistically
significant.
Figure 7. Plot of k1st versus [NO].
O2 to give another peroxy radical. CF3CF2O2 will react with
another NO in the system to give NO2. The product, CF3CF2O,
also falls apart to give CF3 radicals and CF2O. CF3 radicals
can add O2 and the following reaction of CF3O2 can then
produce another NO2. In spite of this complicated mechanism,
in all experiments the rise in the transient absorption at 400 nm
followed first order kinetics. It follows that the rate of
decomposition of CF3CF2CF2O radicals must be faster than the
Two reactions other than reaction 4 could be responsible for
loss of NO2:
F + NO2 + M f FNO2/FONO + M
CF3CF2CF2 + NO2 f products
(15)
(16)
In the presence of 100 mbar of CF3CF2CF2H, F atoms have
a lifetime of 1.3 µs. The absorption transients were fit from 3
µs after the radiolysis pulse reaction and hence reaction 15 will
not be a complication. Reactions 16 competes with reaction 2
for the available CF3CF2CF2 radicals. While the rate constant
for reaction 2 has not been measured, it is very likely to be
similar to that for the CF3 + O2 addition reaction, which has a
fastest pseudo-first-order rate measured, i.e., >1.5 × 105 s-1
.
The formation of NO2 from reactions which occur subsequent
to reaction 3 introduces a delay in the overall appearance of
NO2. The value of k3 ) (8.1 ( 1.1) × 10-12 determined in the
present work should be regarded as a lower limit, hence, k3 >
7.0 × 10-12 cm3 molecule-1 s-1. This value is consistent with
rate constants for other peroxy radicals derived from alkanes7
and HFCs with respect to reaction with NO. For CF3CF2O2, k
high pressure limit of 4.0 × 10-12 cm3 molecule-1 s-1 20
.
Hence,
the lifetime of CF3CF2CF2 radicals in the presence of 5 mbar
of O2 is 2 µs and conversion of CF3CF2CF2 into the peroxy
radical will be substantially complete by the start of the data
analysis at 3 µs after the radiolysis pulse. Complications caused
> (1.07 ( 0.15) × 10-11, for CF3CF2CFHO2, k > 8 × 10-12 17
,
for CF3CFHO2, k ) (1.31 ( 0.41) × 10-11 18
, and for CF3CH2O2,
k ) (1.2 ( 0.3) × 10-11 cm3 molecule-1 s-1 11
.

Atmospheric Chemistry of HFC-227ca
J. Phys. Chem., Vol. 100, No. 16, 1996 6577
Figure 9. Loss of CF3CF2CF2H versus those of CF3CF2H and CClF2-
CH3 when mixtures containing these compounds were exposed to Cl
atoms in 700 Torr of N2 diluent.
Figure 10. Loss of CF3CF2CF2H versus those of CF3CF2H and CF3-
CH2CF3 when mixtures containing these compounds were exposed to
F atoms in 700 Torr of N2 diluent.
by reaction 16 should manifest themselves as curves in the plot
of k1st versus [NO2]. Within the admittedly large data scatter
in Figure 8 there is no such trend and we believe it reasonable
to conclude that reaction 16 is not a significant complication.
The value of k4 determined in this work is entirely consistent
with the high pressure limiting rate constants for reactions
between other halogenated peroxy radicals and NO2 which lie
estimate that potential systematic errors associated with uncer-
tainties in the Cl and F reference rate constants could add
additional 15% and 20% uncertainty ranges for k18 and k8,
respectively. Propagating these additional uncertainties gives
values of k18 ) (3.2 ( 0.5) × 10-16, k18 ) (3.5 ( 0.6) × 10-16
,
k8 ) (3.2 ( 0.7) × 10-13, and k8 ) (3.1 ( 0.6) × 10-13 cm3
molecule-1 s-1. We choose to cite final values of k18 and k8
which are averages of those determined by using the different
reference compounds together with error limits which encom-
pass the extremes of the individual determinations. Hence, k18
) (3.4 ( 0.7) × 10-16 and k8 ) (3.2 ( 0.8) × 10-13 cm3
in the range (4.5-8.9) × 10-12 cm3 molecule-1 s-1 21
.
Relative Rate Studies of the Reactions of Cl and F Atoms
with CF3CF2CF2H. Prior to investigating the atmospheric fate
of CF3CF2CF2O radicals, a series of relative rate experiments
were performed using the FTIR system to investigate the kinetics
of reactions 18 and 8. The techniques used are described in
detail elsewhere.22
molecule-1 s-1
. Quoted errors reflect the accuracy of our
measurements. The value of k8 determined using the FTIR
technique is in agreement with the determination of k8 ) (3.6
( 1.3) × 10-13 cm3 molecule-1 s-1 from the pulse radiolysis
technique in the present work. There are no literature data
available for k18 or k8 with which to compare our results.
Study of the Atmospheric Fate of CF3CF2CF2O Radicals.
To determine the atmospheric fate of the alkoxy radical CF3-
CF2CF2O formed in reaction, 3 experiments were performed in
which mixtures of 13-30 mTorr of CF3CF2CF2H, 21-26 mTorr
of NO, and 700-715 mTorr of Cl2 in either 5 or 10 Torr total
pressure of O2 diluent were irradiated in the FTIR-smog
chamber system. Loss of HFC-227ca and product formation
were monitored by using FTIR spectroscopy. Only one carbon
containing product was observed; COF2. Figure 11 shows IR
spectra acquired before (A) and after (B) a 103 min irradiation
of a mixture of 29.3 mTorr of CF3CF2CF2H, 21.0 mTorr of
NO, and 697 mTorr of Cl2 in 10 Torr of O2. Panel C in Figure
11 shows the product spectrum derived by subtracting the IR
features attributable to CF3CF2CF2H from panel B. Comparison
of the product spectrum with a reference spectrum of COF2
given in panel D shows that COF2 is a major product. After
subtracting COF2 features there were no residual IR features
which could be attributable to any other carbon containing
products.
Photolysis of molecular halogen was used as a source of
halogen atoms.
Cl2 (or F2) + hν f 2Cl (or 2F)
Cl + CF3CF2CF2H f CF3CF2CF2 + HCl
F + CF3CF2CF2H f CF3CF2CF2 + HF
(17)
(18)
(8)
The rate of reaction 18 was measured relative to (19) and
(20). Reaction 8 was measured relative to (21) and (22).
Cl + CF3CHF2 (HFC-125) f products
Cl + CClF2CH3 (HCFC-142b) f products
(19)
(20)
F + CF3CHF2 (HFC-125) f CF3CF2 + HF
(21)
(22)
F + CF3CH2CF3 (HFC-236fa) f CF3CHCF3 + HF
The observed losses of CF3CF2CF2H versus those of reference
compounds in the presence of either Cl or F atoms are shown
in Figures 9 and 10, respectively. All experiments were
performed in 700 Torr of N2 diluent. Linear least squares
analysis gives k18/k19 ) 1.26 ( 0.06, k18/k20 ) 0.89 ( 0.05,
Figure 12 shows a plot of the formation of COF2 versus the
loss of CF3CF2CF2H observed in the present experiments.
Individual variation of the CF3CF2CF2H partial pressure and
the O2 diluent pressure by factors of 2 had no discernible impact
on the COF2 yield. Linear least squares analysis of the data in
Figure 11 gives a molar COF2 yield of 272 ( 15%; i.e., COF2
product accounts for 91 ( 5% of the loss of CF3CF2CF2H. The
shortfall of approximately 9% in the carbon balance may reflect
the presence of some undetected product or systematic uncer-
tainties in the determination of the COF2 yield or both.
k8/k21 ) 0.91 ( 0.05, and k8/k22)1.70 ( 0.10. Using k19
)
2.5 × 10-16 23
,
k20 ) 3.9 × 10-16 22
,
k21 ) 3.5 × 10-13,9 and k22
) 1.8 × 10-13 12 gives k18 ) (3.15 ( 0.15) × 10-16, k18
)
(3.47 ( 0.19) × 10-16, k8 ) (3.18 ( 0.17) × 10-13, and k8 )
(3.06 ( 0.64) × 10-13 cm3 molecule-1 s-1, respectively. We

6578 J. Phys. Chem., Vol. 100, No. 16, 1996
Giessing et al.
Figure 12. Yield of COF2 versus the loss of CF3CF2CF2H following
irradiation of CF3CF2CF2H/Cl2/NO mixtures in O2 diluent (see text for
details).
n-C3H7O2) tendency to form nitrates in their reactions with
NO.26
Implications for Atmospheric Chemistry. Like all HFCs,
the atmospheric lifetime of HFC-227ca will be determined by
its rate of reaction with OH radicals. To the best of our
knowledge there are no data available for the kinetics of the
reaction of OH radicals with HFC-227ca. In the present work
we show that HFC-227ca and HFC-125 have comparable
reactivities toward Cl and F atoms. This finding is not
unexpected when one considers the close similarity in the
molecular structure of these two HFCs. It seems reasonable to
expect that OH radicals will react with HFC-227ca at a rate
which is comparable to their reaction with HFC-125. Percival
et al.27 have reported a correlation between the reactivity of
HFCs towards OH radicals and the HFC ionization potential.
HFC-227ca and HFC-125 have very similar ionization potentials
(13.62 and 13.85 eV, respectively27), supporting the notion that
HFC-227ca and HFC-125 will have similar reactivities toward
OH radicals. Hence, the atmospheric lifetime of HFC-227ca
will be similar to that of HFC-125; approximately 26 years.28
Reaction of OH radicals with HFC-227ca in air gives CF3CF2-
CF2O2 radicals. In the present work it has been shown that in
the presence of NO, CF3CF2CF2O2 radicals are converted into
2 molecules of COF2 and a CF3O radical. In the atmosphere
CF3O radicals react with hydrocarbons and NO to give CF3OH
and COF2, respectively. The atmospheric fates of COF2 and
CF3OH are dominated by incorporation into cloud-rain-sea
water followed by hydrolysis to give HF and CO2.3
Figure 11. Infrared spectra acquired before (A) and after (B) irradiation
of a mixture of CF3CF2CF2H/Cl2/NO in O2 diluent (see text for details).
Panel C shows the result of subtracting features attributable to CF3-
CF2CF2H from panel B. Panel D is a reference spectrum of COF2.
The observed COF2 yield is consistent with the following
reactions occurring in the chamber:
CF3CF2CF2O2 + NO f CF3CF2CF2O + NO2 (3a)
CF3CF2CF2O + M f CF3CF2 + COF2 + M (23)
CF3CF2 + O2 + M f CF3CF2O2 + M
CF3CF2O2 + NO f CF3CF2O + NO2
CF3CF2O + M f CF3 + COF2 + M
CF3 + O2 + M f CF3O2 + M
(24)
(25a)
(26)
(27)
CF3O2 + NO f CF3O + NO2
(28a)
(29)
CF3O + NO f FNO + COF2
The observation of essentially complete conversion of CF3-
CF2CF2H into COF2 allows several interesting conclusions. First,
we can conclude that the atmospheric fate of the CF3CF2CF2O
radical is C-C bond scission (reaction 23). This result is
expected and is consistent with the determination from the pulse
radiolysis experiments that reaction 23 proceeds with a rate
greater than 1.5 × 105 s-1 at 296 K in 1 bar of SF6. Second,
we can confirm previous findings from this laboratory23 and
elsewhere24,25 that C-C bond scission is the atmospheric fate
of the alkoxy radical derived from HFC-125 (CF3CF2H). Third,
we conclude that nitrate formation (e.g., reaction 3b) in the
reaction of perfluoro alkyl peroxy radicals with NO is of minor
importance. If we assume that CF3O2, CF3CF2O2, and CF3-
CF2CF2O2 radicals all display the same tendency to form nitrates
in their reaction with NO, then the 91 ( 5% carbon balance
implies the nitrate yield in each of these reactions is <5%. This
finding is consistent with the behavior of the analogous alkyl
peroxy radicals which display little (e.g., 2% nitrate for
The present work represents the first study of the atmospheric
degradation of HFC-227ca. As expected, the behavior of HFC-
227ca is very similar to that of HFC-125 (CF3CF2H). The
reactivities of both HFCs toward F and Cl atoms are comparable.
The spectra of the peroxy radicals generated from both HFCs
and their reactivity of the peroxy radicals toward NO and NO2
are indistinguishable. Once initiated by OH radical attack a
sequence of rapid reactions occurs which results in the “unzip-
ping” of the molecule in which successive COF2 molecular
fragments are shed and a CF3 radical is formed. From the
viewpoint of assessing the environmental acceptability of HFC-
227ca as a CFC replacement, we must consider potential impact
in three areas: stratospheric ozone depletion, toxic/noxious
degradation product formation, and global warming. As with
all HFCs, HFC-227ca has no impact on stratospheric ozone.29
We show here that HFC-227ca is degraded into COF2 and CF3-
OH, neither of which can be considered toxic or noxious at the
concentrations expected in the environment. The global warm-

Atmospheric Chemistry of HFC-227ca
J. Phys. Chem., Vol. 100, No. 16, 1996 6579
(13) Sehested, J.; Nielsen, O. J.; Wallington, T. J. Chem. Phys. Lett.
1993, 213, 457.
(14) Zabel, F.; Kirchner, F.; Becker, K. H. Int. J. Chem. Kinet. 1988,
20, 827.
(15) Sehested, J. Int. J. Chem. Kinet. 1994, 26, 1023.
(16) Wallington, T. J.; Ellermann, T.; Nielsen, O. J.; Sehested, J. J. Phys.
Chem. 1994, 98, 2346.
(17) Møgelberg, T. E.; Feilberg, A.; Giessing A. M. B.; Sehested, J.;
Bilde, M.; Wallington, T. J.; Nielsen, O. J. J. Phys. Chem. 1995, 99,
17386.
ing potential of HFC-227ca is expected to be comparable to
that of HFC-125, which (assuming a 100 year time horizon) is
approximately 3200 times more effective at trapping infrared
radiation than the same mass of CO2.30 To place this value in
perspective it should be noted that CFC-11 and CFC-12 have
global warming potentials which over a 100 year time horizon
are factors of 4000 and 8500 larger than CO2, respectively.30
References and Notes
(18) Wallington, T. J.; Nielsen, O. J. Chem. Phys. Lett. 1991, 187, 33.
(19) Sehested, J. Int. J. Chem. Kinet. 1994, 26, 1023.
(20) Kaiser, E. W.; Wallington, T. J.; Hurley, M. D. Int. J. Chem. Kinet.
1995, 27, 205.
(21) Lightfoot, P. D.; Cox, R. A.; Crowley, J. N.; Destriau. M.; Hayman,
G. D.; Jenkin, K. E.; Moorgat, G. K.; Zabel, F. Atmos. EnViron. 1992, 26A,
1805.
(22) Wallington, T. J.; Hurley, M. D. Chem. Phys. Lett. 1992, 189, 437.
(23) Sehested, J.; Ellermann, T.; Nielsen, O. J.; Wallington, T. J.; Hurley,
M. D. Int. J. Chem. Kinet. 1993, 25, 701.
(24) Edney, E. O.; Driscoll, D. J. Int. J. Chem. Kinet. 1992, 24, 1067.
(25) Tuazon, E. C.; Atkinson, R. J. Atmos. Chem. 1993, 16, 301.
(26) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F.; Kerr, J.
A.; Troe, J. J. Phys. Chem. Ref. Data 1992, 21, 1125.
(27) Percival, C. J.; Marston, G.; Wayne, R. P. Atmos. EnViron. 1995,
29, 305.
(1) Alternative Fluorocarbon Environmental Acceptability Study, World
Meteorological Organization Global Ozone Research and Monitoring
Project, Report No. 20; Scientific Assessment of Stratospheric Ozone; 1989;
Vol. I and II.
(2) McCulloch, A. EnViron. Monit. Assess. 1994, 31, 167.
(3) Wallington, T. J.; Schneider, W. F.; Worsnop, D. R.; Nielsen, O.
J.; Sehested J.; Debruyn, W. J.; Shorter, J. A. EnViron. Sci. Technol. 1994,
28, 320.
(4) Wallington, T. J.; Japar, S. M. J. Atmos. Chem. 1989, 9, 399.
(5) Hansen, K. B.; Wilbrandt, R.; Pagsberg, P. ReV. Sci. Instrum. 1979,
50, 1532.
(6) Ellermann, T; Sehested, J; Nielsen, O. J.; Pagsberg, P.; Wallington,
T. J. Chem. Phys. Lett. 1994, 218, 287.
(7) Wallington, T. J.; Dagaut, P.; Kurylo, M. J. Chem. ReV. 1992, 92,
667.
(8) Wallington, T. J.; Ellermann, T.; Nielsen, O. J. Res. Chem. Intermed.
1994, 20, 265.
(9) Wallington, T. J.; Hurley, M. D.; Shi, J.; Maricq, M. M.; Sehested,
J.; Nielsen, O. J.; Ellermann, T. Int. J. Chem. Kinet. 1993, 25, 651.
(10) Møgelberg, T. E.; Nielsen, O. J.; Sehested, J.; Wallington, T. J.;
Hurley, M. D. J. Phys. Chem. 1995, 99, 5373.
(11) Nielsen, O. J.; Gamborg, E.; Sehested, J.; Wallington, T. J.; Hurley,
M. D. J. Phys. Chem. 1994, 98, 9518.
(12) Møgelberg, T. E.; Platz, J.; Nielsen, O. J.; Wallington, T. J. J. Phys.
Chem. 1995, 99, 5373.
(28) Derwent, R. G.; Volz-Thomas, A.; Prather, M. J., World Meteo-
rological Organization Global Ozone Research and Monitoring Project,
Report No. 20; Scientific Assessment of Stratospheric Ozone, 1989; Vol.
II, p 124.
(29) Wallington, T. J.; Schneider, W. F.; Sehested, J.; Nielsen, O. J.
Faraday Discuss. 1996, 100, 55.
(30) Intergovernmental Panel on Climate Change. Climate Change 1994;
Cambridge University Press: UK, 1995.
JP952769H