EPA-650/3-75-008
OXIDATION OF HALOCARBONS
by
J, P. Heicklen, E. Sanhueza, I. C, Hisatsune,
R. K. M. Jayanty, R. Simonaitis, L. A. Hull,
C. W, Blume, and E. Mathias
Center for Air Environment Studies
Pennsylvania State University
University Park, Pennsylvania 16802
Grant No, 800949
ROAP No. 26AAD-20
Program Element No. 1A1008
EPA Project Officer; Dr. Joseph Bufalini
Chemistry and Physics Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
U. S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
WASHINGTON, B.C. 20460
May 1975
-------
EPA REVIEW NOTICE
This report has been reviewed by the National Environmental Research
Center - Research. Triangle Park, Office of Research and Development,
EPA, and approved for publication. Approval does not ignify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environ-
mental Protection Agency, have been grouped into series. These broad
categories were established to facilitate further development and applica-
tion of environmental technology. Elimination of traditional grouping was
consciously planned to foster technology transfer and maximum interface
in related fields. These series are:
1. ENVIRONMENTAL HEALTH EFFECTS RESEARCH
2. ENVIRONMENTAL PROTECTION TECHNOLOGY"
3. ECOLOGICAL RESEARCH
4. ENVIRONMENTAL MONITORING
5. SOCIOECONOMIC ENVIRONMENTAL STUDIES
6. SCIENTIFIC AND TECHNICAL ASSESSMENT REPORTS
9. MISCELLANEOUS
This report has been assigned to the ECOLOGICAL RESEARCH series.
This series describes research on the effects of pollution on humans,
plant and animal species, and materials. Problems are assessed for
their long- and short-term influences. Investigations include formation,
transport, and pathway studies to determine the fate of pollutants and '
their effects. This work provides the technical basis for setting standards
to minimize undesirable changes in living organisms in the aquatic,
terrestrial, and atmospheric environments.
This document is available to the public for sale through the National
Technical Information Service, Springfield, Virginia 22161.
Publication No. EPA-650/3-75-008
-------
Previous Page Blank .
111
ABSTRACT
The gas-phase room-temperature oxidation of haloethylenes is reviewed.
In general oxidation has been carried out in five ways: 1) chlorine atom
initiation, 2) Hg 6(3P) sensitization, 3) reaction with 0(3P), 4) reaction
with 0(3P) in the presence of 02, and 5) reaction with 03.
The chlorine-atom initiated oxidation of CC12CC12, CHC1CC12, CH2CC12,
cis-CHClCHCl, trans-CHClCHCl, CF2CC12, CFC1CFC1 (mixed cis and trans),
CF2CFC1, and CjFi, proceed by a long chain free radical process. The major
products are the corresponding carbonyl chlorides containing 1 or 2 carbon
atoms. By contrast there is no chain process in CHC1CH2, For most of
the chloroethylenes, the chain length of the reaction exceeds 100 at suf-
ficiently high 02 pressures, and is independent of the absorbed light
intensity, Ia, or any of the reactant pressures.
The general mechanism of the long chain oxidation is (X = H, F, or Cl)
Cl + CX2CXC1 -* ClCX-jCXCl 2a
-*• CX2CXC12 2b
C1CX2CXC1 + 02 -»• C1CX2CXC102 3
CX2CXC12 + 02 + CXC12CX202 3'
2C1CX2CXC102 + 2C1CX2CXC10-+ 02 4a
-> (C1CX2CXC10)2 + 02 4b
2CXC12CX202 -> 2CXC12CX20 + 02 4a'
-> (CXC12CX20)2 + 02 4bf
C1CX2CXC102 + C1CX2CXC1 -»• (C1CX2CX2C10)2 5
C1CX2CXC10 -> C1CX2CX(0) + Cl 6a
-»• CXC10 + C1CX2 6b
CXC12CX20 -> CXC12CX(0) + X 6a'
-------
IV
•-»• CXC12 + CX20 6b!
CICXj + 02 -* CX20 + Cl + (1/2)02 7a
The chlorine atom attaches preferentially to the less chlorinated car-
bon atom of the chloroethylene. A long chain oxidation (>150) occurs when
the exothertnicity of either reaction 6a or 6b is greater than 11 kcal/mole.
For an exo'thermicity of 11 kcal/mole, a shorter chain length (=20) is in-
volved. In CHC1CH2, the radical produced is CH2C1CHC10, the exothermicity
of decay of this radical by any route is <11 kcal/mole, and the favored
decay route is by C-C bond cleavage which produces the terminating radical
CH2C1. The ejection of an H atom from CC1X2CXHO was never observed; decay
of CC1X2CH20 by any route was not observed.
The Hg 6(3P) sensitized oxidation of chloroethylenes leads to the same
free-radical long-chain process as observed with chlorine-atom initiation.
However the chain lengths are shorter and are proportional to the olefin
pressure in the Hg 6(3P) system. Furthermore CO is always a product of the
reaction. Thus the initiating and terminating steps are different in the
two systems.
For the chlorinated ethylenes, the Hg 6( P) system was studied for
C2C1,,, CyCl^E, and CCl^CHj, only. It was proposed that C2X2C1 radicals are
produced. These radicals can add 0^. The suggested initiation reaction was:
C2X2C102 + CX2CC1* -» C2X2C13 4- (CXO)2
The CO production could be associated with the termination step which is not
well understood. The overall reaction suggested was:
C2X2C102 -*• CO via termination
When 0(3P) reacts with haloethylenes, carbon-carbon double bond cleavage
can occur
0(3P) + CjjXi, -> CX2 + CX20 (or CO + X2)
-------
t'ne reaction can proceed through an excited intermediate
0(3P) + C2Xi, -»• CX2CX20*
or rearrangement, possibly followed by fragmentation, can occur
0(3P) + CzXk -*• CX3CXO (or CX3 + XCO)
For the haloethylenes studied, the relative importance of the three processes
are :
Hctl oe thy l_ene -
CC12CC12
CClzCClH
CC12CH2
cis-CHClCHCl
trans-CHClCHCl
CC1HCH?
C=C Cleavage
0,19
0.23
0.31
0.23
0.28
<0.25
0.85
0.80
0
Excited Molecule
0.81
0.77
0.55
0.73-0.77
0.68-0.72
>0.30
0.15
0.20
1.00
Rearrangement
0
0
0.14
£0.04
<0.04
0.09-0.34
0
0
0
CI'CICFCI (cis & trans)
C!;>CC12
In the presence of Oj, the diradical fragment produced along with the
carbonyl compound containing one carbon can oxidize to initiate the long
chain mono-free-radical process observed by either chlorine atom initiation
or Hg 6(?P) sensitization. If the diradical contains a chlorine atom, the
initiation step is
CXC1 4- 02 •> XO + C1CO
The chain length of the oxidation depends on the parameter [CC1XCX2 }/la1/i,
thus indicating additional radical-radical terminating steps.
The ozonolysis of haloethylenes proceeds by an entirely different
route than the above-mentioned oxidations. It is a chain oxidation, carried
by a diradical mechanism, which is inhibited in the presence of 02 . The
-------
VI
reaction rate law is first-order in both [03] and [CX2CX2] at high react-
ant pressures, but at low pressures the rate drops off faster than extra-
polated, and the rate law changes in a complex way. However in all cases
the major products of the reaction are the. corresponding 1- and 2-carbon
carbonyl products. For the less-substituted chloroethylenes, the
carbonyl products containing one carbon atom are the major, if not
exclusive, products,
The mechanism of the ozonolysis is complex and several paths may be
involved. The most often invoked mechanism starts with the cleavage of
the double bond,
c2Xit + 03 -»• cx2o + cxzo2
This reaction proceeds through the molozonide as an intermediate and occurs
in CaHn, and the higher hydrocarbon olefins as well as in CH2CHC1, However
.there is considerable evidence to suggest that either in addition to or in
place of the above reaction, the initiation can occur via
C2X^ + 03 £ 02X1*03 107
€2X^03 + C2X% J C%X803 108
djXaOs + 03 •* 2CX20 + 2CX202 109
The C2Xt,03 intermediate is probably the ir-complex. The route consisting
of reactions 107-109 is the dominant, if not exclusive route, to ozonolysis
in CHC1CHC1, CH2CG12, and CC12CC12. With either initiating mechanism, the
CX20a species carries the chain.
The chlorine-atom sensitized oxidation of CH2Cl2 gives CHC10 and
CC120 as products with respective quantum yields of 49 and 4.1 independent
of reaction conditions at 32°C. For the C1-™CH3C1 system the initial products
are HC1 and CHC10, the quantum yield of the latter being 2.0 under all
conditions. Thus CHC12 reacts with 02 similarly to CCl3s whereas CH2C1
-------
vii
reacts with 0% simlarly to CH3, (except that no alcohol is produced). In
both systems the CHC10 is removed by chlorine atoms in an hydrogen abstrac-
tion reaction with a rate coefficient of 7 x 108 JT"1 see'1 at 32°C.
The photolysis of CClt> at 213.9 nm is interpreted in terms of an
excited molecule mechanism which proceeds entirely by
•*• CC12 (singlet) + C12
at low pressures. At higher pressures CCli,* is quenched and CC12 pro-
duction is inhibited, though it may be (and probably is) replaced by
production of CC13 + Cl. For CFC13 and CFzCla photolysis at 213.9 nm,
the main, and probably exclusive, process is chlorine atom ejection
CFC13 (or CF2C12) + hV (213.9 nm) -»• Cl + CFC12 (or CF2C1)
The reaction of 0(1D) with the perhalomethanes leads mainly, if not
exclusively, to chlorine atom abstraction
0(JD) + CFnCU-n (n - 0-3) -»• CIO + CFnCl3_n
The rate coefficients for these reactions , relative to that for N20
(k - 2.2 x 10- 10 cmVsec) are
molecule CCU CFC13 CF2C12 CF3C1
relative k 2.1 1.5 1.2 0.52
This report was submitted in fulfillment of E.P.A. Grant No. R 800949
under the major sponsorship of the Environmental Protection Agency. Work
was completed as of February, 1975.
-------
Vlll
TABLE OF CONTENTS
Page
ABSTRACT iii
LIST OF FIGURES xi
LIST OF TABLES xiii
ACKNOWLEDGMENTS ....... ........... xiv
CONCLUSIONS xv
RECOMMENDATIONS ....... . ....... .xviii
INTRODUCTION. ........ .... .... 1
I. CHLOROETHYLENE OXIDATION. 2
Experimental 2
Photochemical Experittusnts . .............. 2
Materials 2
Procedure ............. .... 3
Ozone Experiments „ . 5
Cl-Atom Initiated Oxidation 7
Vinyl Chloride 16
03 Present 19
Fa-Initiated Oxidation . • 21
Hg 6(3P) Sensitized Oxidation .................... 21
Oz Absent. . 23
Oz Present 24
Reaction with 0(3P) Atoms . 29
Individual Substituted Ethylenes .... 30
CC12CC12. ............. ..... 30
CC12CHC1. ............. 31
CC12CH2 .............. , 31
-------
IX
Page
cis- and trans-CHClCHCl ..... 33
CHC1CH2 • . 34
CH2CH2 35
CH2CHBr . 36
CF2CF2 36
C2FnHi»-n(n = 1, 2 and 3) 36
CFC1CFC1 39
CF2eei2 39
CF2CFC1 40
Mechanism. 40
late Coefficient 43
Reaction with 0(3P) in the Presence of 02 46
Methylene Oxidation. ............ 48
Oxidation of CX2CX20* 49
Individual Molecules 49
C2CU 49
CHC1CC2.2 50
CHC1CHC1. ................. 50
CH2CC12 . 50
CFC1CFC1 50
CF2CC12 50
C2F1» 51
CHC1CH2 . .......... 52
Rate Law 53
Reactions with Ozone 54
Review of the Experimental Data. . ........... 56
-------
Page
................. ..... 56
65
............... 66
CHC1CC12. ..... o ....... . 69
CH2CC12 ..........,,.„ 69
CHC1CHC1. ........................... 70
CHC1CH2 . . , . » 77
Review of Ozonolysis Mechanism ....... ...... 83
.„..„....„. 83
91
...'.................... 92
CH2CC12 ................. . 95
CHC1CHC1. ........ 97
Discussion 100
II. OXIDATION OF CHLOROMETHANES 108
Experimental. ........ ...... 108
Photooxidation of the Perhalomethanes . 110
Photolysis of CCli» . , HO
Photolysis of CFC13 116
Photolysis of CF2C12 .......... 119
Reaction with 0(1D) Atoms ..... . 120
Chlorine-Atom Sensitiased Oxidation of CH2C12 and CH3C1. ........ 122
REFERENCES. ...... ..... . 125
LIST OF PUBLICATIONS. 132
-------
XX
LIST OF FIGURES D
Page
Fig,, 1 Log-log plot of ${CH2C1CC1(0)} vs. [02]/[C12] in the chlorine-
atom Initiated oxidation of CaHaCl at 31°C. From Sanhueza and
Heicklen (8) with permission of the American Chemical Society. ... 18
Fig. 2 Log-log plot of the ratio of the quantum yields of
and CClgO vs. [Os]/ [02] in the chlorine-atom sensitized oxidation
of C2Cli» by 02 and Q3 at 32°C. From Mathias et al (5) with
permission of the National Research Council of Canada. ....... 20
Fig. 3 Plot of (${CC13CC1(0)} -I- (1/2)*{CC120})/${CC12CC126} vs. [C2CU]/
[63] In the chlorine-atom sensitized oxidation of CzClt, by 02 and
03 at 32°C. From Mathias et al (5) with permission of the National
Research Council of Canada ........... . ..... .... 22
Fig. 4 Infrared spectra of primary and secondary ethylene ozonides at
liquid nitrogen temperature. In part from Hull et al (88) ..... 61
Fig. 5 Microwave structure of secondary ethylene ozonide. From data of
Gillies and Kuezkowski (85 , 86) ............ . ...... 64
Fig. 6 Time dependence of the composition of C^Ci^ ozonolysis reaction
at 24CC: [C2CU]0 = 6.9 Torr, [03]o » 4.1 Torr. From Mathias
et al (5) with permission of the National Research Council of
Canada ......... ..... ... ............. .68
Fig. 7 Fourth order kinetic plot of trans-CHClCHCl reaction with ozone
at 23°C. From Blume et al (71) ............ ....... 74
Fig. 8 First order kinetic plot of cls-CHClCHCl reaction with ozone at
23°C in N2 buffer. From Blume et al (71). . ..... . ...... 75
-------
Xll
Page
Fig. 9 First order kinetic plot of eis-CHClCHCl reaction with ozone
at 23°C in 02 buffer. From Blume et al (71), . . . . 76
Fig. 10 Ozone catalyzed isomerization of cis-CHClCHCl at 23°C, From
Blume et al (71) 79
Fig. 11 First order kinetic plot from the isomerization data for the
cis-CHClCHCl reaction at 23°C with 02 buffer. From Blume et al
(7D 80
Fig. 12 Infrared spectrum of vinyl chloride primary ozonide at liquid
nitrogen temperature. The absorption bands identified by
arrows are assigned to the more stable isomer of the ozonide,
The weak band at 1755 cm"1 is due to formyl chloride residue
still in the ozonide sample. From Hisatsune et al (92) .... 82
Fig. 13 Second order kinetic plot of vinyl chloride reaction with ozone
at 22°C. From Kolopajlo (94) .......... 84
Fig. 14 Plot of §{COC12} vs [N2] or [02] for CC1% photolysis at 213.9 nm
in the presence of 02 or 03 at 25°C. 0 [CCU] ^ 10 Torr in the
presence of 02, A [CCli*] ^ 10 Torr in the presence of 03,
• [CCU] ^ 50 Torr in the presence of 02 , A [CCU] ^ 50 Torr
in the presence of 63, All analyses by gas chromatography.
From Jayanty et al (111) Ill
-------
Xlll
• LIST OF TABLES
Page
Tattle I The chlorine-atom sensitized oxidation of C2Clit-nHn and
nFn .............. . ......... .... 8
Tatle II Bond energies (kcal/mole) in chloroethoxy radicals ....... 13
Table III Rate coefficient ratios from the Hg 6(3P) sensitized
oxidation of chloroethylenes at 30-32°C. ............ 28
Table IV Oxidation of CC1HCC12 . ....... ...... .......... 32
Table V Products and quantum yields in the reaction of 0(3P) with
chloro-, ehlorofluoro-, and f luoroethylenes . ..... ..... 38
Table VI Rate coefficient for the reaction of atomic oxygen with
haloethylenes at room temperature ................ 44
TabLe VII Relative reactivities of chloro- and chlorof luoroethylenes
with atom and radicals in the gas phase at room temperature. . . 47
labile VIII The reaction of chloroethylenes with 0(3P) in the presence
of 02 ............... ..... .......... 55
Table IX Kinetics of ethylene-ozone reaction in the presence of
excess 02 . ............ ... ............ 60
TabLe X Infrared spectra of primary and secondary ethylene ozonide ... 62
Table XI Kinetic data for the ozonolysis of 1,2-dichloroethylene (DCE) . . 78
Table XII Second order rate constants for the reactions of ozone
with haloethylenes in CCln solution at 25°C. .... ...... 94
Table XIII Elementary rate constants in the mechanism of ozonolysis of
1,1-dlchloroethylene at 25°C .................. 98
Table XIV Elementary rate constants in the mechanism of ozonolysis of
1,2-dichloroethylene at 23°C ............. ..... 101
Table XV Summary of measured and literature values of the rate coeffi-
cient: for O^D) reactions, k{X>, relative to that for N20,
k{N20} ...................... ....... 123
-------
XIV
ACKNOWLEDGMENT
The work reported herein which was done in our laboratory was done
with financial support from the Environmental Protection Agency under
Grant No, R800949 and the Center for Air Environment Studies at
Fenn State. Partial support for the studies on the perhalomethanes was
provided by the National Aeronautics and Space Administration through
Grant No, NGL-009-003 and the Atmospheric Sciences Section of the
National Science Foundation through Grant No. GA-42856.
-------
XV
CONCLUSIONS
Chloroethylenes oxidize by free radical attack to give one and two
carton carbonyl compounds and chlorine atoms. Consequently when chloro-
olefins are in the atmosphere they will generate chlorine atoms and oxidize
in a chain process, The carbonyl compounds are toxic and may hydrolyze to
give HC1, particularly in the respiratory tract. In addition HC1CO decom-
poses in the absence of H20 to give HC1 + CO and reacts with chlorine
atoms with a rate coefficient of 7 x 108 M"1 sec"1 at 32°C.
The chloroethylenes are attacked by 0(3P) atoms at a rate 0.1-1
that of 0(3P) attack on C^R^. There are two paths of importance. Most
of tie time for the non-fluorinated chloroethylenes, the reaction produces
an. excited adduct which may or may not react further in the chain process.
0(3P) + C2Xi> -> CX2CX20*
For the non-fluorinated chloroethylenes, double-bond cleavage occurs
19-31% of the time
0(3P) + C2X.4 -*• CX20 + CX2
The CXa diradical, which presumably is a triplet, can then react with 02
to generate mono-free radicals. For the fluorinated chloroethylenes, the
relative importance of the two processes is reversed. With CC12CH2,
CC1HCH2, and possibly cis- and trans-CHClCHCl, rearrangement occurs as a
minor process.
0(3P) + CX2CX2 -> CX3CXO (or CX3 + XCO)
The reactions of 03 with C2Clit, CC12CH2, cis- and trans-CHClCHCl, and
CHC1CH2 are not important in the atmosphere for two reasons: 1) These
reactions proceed by a long-chain diradical process in the absence of 02,
but 02 inhibits the chain and greatly reduces the rate of reaction.
2) la the Torr range, the reactions proceed with a rate law which is
-------
XVI
first-order in each reactant. However at lower pressures, the reactions
become second-order in chloroethylene concentration (and in some cases
also 03 concentration). Even in the regime .first order in both reactant
concentrations, the rate coefficient is smaller than for the Oa-CaHi*
reaction. With CHClCHa our preliminary studies, done in the Torr range,
have not shown any deviation from the rate law first-order in each reactant,
but the rate coefficient in the presence of 0% of 3.9 M"1 see"1 is too
small for this reaction to be of any importance, even if the rate law is
not modified at lower pressures.
The perhalomethanes are not chemically active in the troposphere.
However in the lower stratosphere they can be removed by at least two
processes: photooxidation and 0(ID) atom attack. The photooxidation of
CFCls and CFgCla at 213,9 nm proceeds mainly, if not entirely, by chlorine-
atom ejection
CFC13 (or CF2C12) + hv (213.9 nm) -* Cl + CFC12 (or CF2C1)
With CClit, however, the principal process expected for stratospheric
pressures is molecular chlorine elimination at 213.9 nm
CCln + hv (213.9 nm) -»- CC12' (singlet) + C12
The singlet CC12 produced reacted with CCl^ in our laboratory experiments,
but its fate in the stratosphere is not known. Probably it would react
with Os to give CC120 + 02 (singlet). At longer wavelengths, however,
(e.g. 253.7 nm) the photodissociation products may be CCls + Cl in the
stratosphere, since these are the higher-pressure products at longer
wavelengths (120).
The reaction of 0(1D) with the perhalomethanes is rapid and proceeds
mainly, if not entirely, by chlorine-atom abstraction
CFnGU-n (n - 0-3) 4- 0(1D) -> CFnCl3-n + CIO
-------
XVI1
The room temperature rate coefficients for these reactions, relative to
that for N20 (k = 2.2 x 10~10 an3/see) are
Molecule CCl^ CFC13 CF2C12 CF3C1
Relative k 2.1 1.5 1.2 0,52
'Hie lifetimes for removal of the perhalomethanes in the stratosphere
by tha two processes are given below at various altitudes.
CCU CFCla CF2C12 CF3C1
km
20
25
30
35
40
45
50
[O^D) ],a
molec/cc
0.25
1.5
7.5
25
87.5
175
150
yr
274
45.7
9.15
2.74
0.78
0.39
0.46
l{Phot},c
yr
6.60
0.647
0.117
0.032
0.014
0.008
0.006
TCOC'D)},"
384
64.0
12.8
3.84
1.10
0.55
0.64
l{Phot},c
yr
63.4
5.28
0.99
0 .26
0.106
0.063
0.045
T{00)},»
480
80.0
16.0
4.80
1.37
0.69
0.80
T{OCD)},'
1108
185
36.9
11.1
3.17
1.58
1.85
a) Average globed yearly value.
b) Lifetime for removal by O^D).
c) Reciprocal of the global average photodissociation coefficients given
by F. S. Rowland and M. J. Molina, Atomic Energy Commission Report
Na. 1974-1 (1974) "Chlorofluoromethanes in the Environment."
, which does not absorb strongly radiation >2000A is removed
mainly by 0(1D) atom attack, probably predominantly at 45 km. The
other three chlorinated methanes are removed principally by photo-
dissociation but removal of CF2C12 by O^D) at 25-45 km makes a
significant contribution.
-------
XVI11
RECOMMENDATIONS
In terms of problems of significance in urban atmospheres the
following studies need to be done:
1) The reactions of Cl, 0(3P), and HO with the haloethanes in the
presence of 02 should be studied.
2) For the molecules already studied, i.e. the chloroethylenes and the
halomethanes, the reactions with HO radicals in the presence of 02
should be examined. The products and mechanism of the reactions should
be determined and the rate coefficients obtained,
3) The fate of the halogenated one- and two-carbon atom carbonyl com-
pounds in the atmosphere should be determined. Their reactions with
Cl, 0(3P), HO, H20, H202, and NH3 should be studied.
4) Computer modelling should be done to determine the Influence of Cl
atoms on photochemical smog.
In terms of problems of significance in the stratosphere, the
photooxidation of the chloroalkanes and the carbonyl compounds should
be examined. Also more detailed wavelength studies should be made
with CClt* to determine its primary process as a function of wavelength
under stratospheric condtions,
Finally in terms of completing general scientific information on
these systems, the following studies would be useful:
1) The reactions of 03 with CC1HCH2 and CC12CC1H could be studied.
2) Temperature studies could be made on all the systems.
-------
INTRODUCTION
An earlier review (1) covered the literature on the gas-phase oxi-
dation of perhalocarbons, Since then concern has developed that the
chlorinated ethylenes, which are used as solvents and for other
incus trial use!( may be accumulating in urban atmospheres. Of particular
concern is vinyl chloride which has been shown to be a mild carcinogen.
Also there has developed recently a serious concern that some of the
coitmonly used chloromethanes may adversely affect the ozone concentra-
tion in the stratospheric layer surrounding our planet (2-4). Therefore
we undertook an, extensive study of the oxidation of ehloroethylenes and
chlorofluoromethanes with financial support from the Environmental Pro-
tection Agency through Grant No. R800949. The results of these studies
are presented here. For the sake of completion we also discuss the
pertinent experimental data and their interpretations from other
laboratories.
-------
I. CHLQEOETHYLENE OXIDATION
EXPERIMENTAL
Photochemical Experiments;
Most of the irradiation experiments were done in an infrared quartz
or PyreK cell with NaCl windows. The gases in the cell were irradiated by
radiation from a suitable Hg lamp which had passed through appropriate fil-
ters to provide the desired wavelength and intensity of radiation. The
reaction cell was situated in the sample beam of a Beckman IR-10 infrared
spectrometer for continual analysis of the reactants and products.
The experiments utilized conventional static photochemical techniques.
The working part of the vacuum system was greaseless, employing Teflon
stopcocks and greaseless joints with Viton 0-rings, For pressures <5 Torr,
a Consolidated Vacuum Corp. McLeod gauge was used whenever possible.
Pressures between 5 and 50 Torr were measured in a Wallace and Tiernan
absolute pressure indicator. Pressures larger than 50 Torr were measured
in a mercury manometer.
Materials; All gases were Matheson C.P. grade. When it was possible
they were purified by bulb to bulb distillation in the vacuum line and
degassed at -196°C before each run. The CaCli» was Baker Analyzed,
chlorine free, The fraction volatile at -21°C but condensable at -90°C
was used. It was degassed before each run from a trap at -90°C. The
CClaCClCO) was from Eastman Kodak Co. and was fractionated in the same
manner as CjCli* before use.
Chlorine (Matheson, 99*5% purity) was degassed thoroughly, and then
repeatedly exposed to KOH pellets to remove the HC1, The absence of HC1
was confirmed by infrared analysis. The chlorine was then slowly dis-
tilled into the storage vessel by keeping the KOH-H20-1KC1 mixture at as
-------
low a temperature as practical.
The C2HCla was Baker Analyzed, and the fraction volatile at -21°C
but condensable at =9Q°C was used. It was degassed before each run at
-90°G. The CCliHCClCO) was from Eastman Kodak Co. and was fractionated
in the same manner as CaHCla before use.
The CCl2CH2 was from J. T. Baker, and the fraction volatile at -80°C
but condensable at -130°C was used. It was degassed before each run at
-13()°C, Because it polymerized spontaneously in the storage vessel, it
was repurified periodically. For the 0$ experiments the CClaCHa was
obtained from the Fisher Scientific Co. and distilled at ~98°C before
use,
The cis- and trans-CHClCHCl were obtained from the Aldrich Chemical
Co. They were distilled at -50°C. Before use they were degassed at -80°C.
Infrared analysis indicated no impurity bands.
The CFC1CFC1 and CF2CC12 were from the Peninsular ChemResearch Co.
The fractions volatile at ~20°C, but condensable at -80°C, were used.
Infrared analysis showed no impurity peaks. The CFC1CFC1 was a mixture
of tie cis- and trans-isomers, and these could not be separated.
The CaHaCl was from the Matheson Co. Before use, it was distilled
at -30°C and collected and degassed at -196°C. Its infrared spectrum
showed no impurity peaks.
Proendure? The gases were saturated with mercury vapor at room tempera-
ture and mixed directly in the cell. For Hg-sensitized experiments
(directly or to produce 0(3P) from NiaO decomposition) a drop of Hg was
in the cell.
Three reaction cells were used. Cell No. 1 was a cylindrical quartz
vesse.l 5 cm in diameter and 10 cm long. Cells No. 2 and 3 were Pyrex
-------
T-shaped cells. For each cell the stem of the T had a 5 cm diameter
quartz window to permit entrance of the radiation. The top of the T had
two NaCl windows, each 5 cm in diameter for infrared analysis. The
windows were attached to the cells with Carter's epoxy cement. The
lengths of the tops of the T cells were 6,7 and 12.7 cm, respectively,
and were situated in the sample beam of a Beckman IR-10 infrared
spectrometer for continual analysis.
For Hg-sensitization experiments, either of the chloroethylene directly
or of H20 to produce 0(3P) atoms, irradiation was from a Hanovia .flat-
spiral, low-pressure mercury resonance lamp. Before entering the reaction
cell, the radiation passed through a Corning 9-54 filter to remove
o
radiation below 22QOA, When reduced intensities were desired, Corning
9-30 filters were inserted between the lamp and reaction vessel.
Actinometry was obtained from the Hg-photosensitized decomposition of
N20 in the presence of 1-2% C2Fif to scavenge the 0(3P) atoms produced.
For this system the quantum yield of Ng production §{^1 - 1,0. For
chlorine photolysis experiments, radiation was from a GE-H100-A4/T
medium pressure Hg lamp with, a Pyrex envelope. The radiation was
filtered through a 3.08 cm thick Corning 7-54 filter to remove visible
0
radiation and isolate the 3660A line of Hg.
Some experiments were done in a 10 cm long by 5 cm diameter quartz
cell with a pinhole bleed. The reaction mixture exited continuously,
prior to and during each photolysis, through the pinhole into 8 mm
Pyrex tubing which led through a second pinhole into a modified E.A.I.
160 quadrupole mass spectrometer for analysis. The leak rate of the
first pinhole was sufficiently small so that the pressure drop in the
reaction vessel was negligible during the time of the experiment.
-------
Whether infrared or mass spectral analysis was used, after some of
the experiments, the products were also analyzed by gas chromatography.
The gases noncondensable at -196°C (CO, Ng) were analyzed using an 8-
O
10 ft. long 5A molecular sieve column at 0°C with a He flow rate of 60-
100 cc/min. When Oa was present as a reactant gas an additional 10-ft.
long column was used to aid in separating the Na and 02. Gas chromatographic
analysis for CC'z was made utilizing a 24 ft. long column packed with Porapak
Q operating at 25°C and a Hj carrier gas flow rate of 60 cc/min.
Finally after completion of some of the runs analyzed continually by
infrared analysis, mass spectral analysis was performed, and vice versa.
Ozone Experiments;
For the experiments in which 03 was used, a mercury-free line was
nec€issary. The gases were mixed in a kinetic apparatus consisting of
two 100 cc bulbs connected by a T-stopcock (one that opens simultaneously
in three directions) to a stem adaptor which leads either through a stop-
cock to the vacuum line or through a stopcock to the optical cell. Sill-
cone grease was used in all the stopcocks. Two optical cells were used.
o
One '«?as a 10.0 cm long quartz cell for following 03 decay at 2537A in a
Gary 14 spectrometer. The other cell was *v 10 cm long and had NaCl,
KBr, or CsBr windows to permit monitoring olefin decay and phosgene -
acid chloride appearance by infrared spectroscopy. A Perkin-Elmer 112
infrared spectrometer with a CaF2 prism was used. All experiments were
performed at room temperature, about 25°C.
Before a run, the kinetic apparatus was evacuated for at least an
hour by an oil diffusion pump and then conditioned for 10 rain, with
about 5 Torr of Oa before final evacuation. The 03 was prepared from a
tesla coil discharge through 02, the excess Oa being removed at -196°C.
-------
A desired amount of freshly prepared 03 was then measured into the
cell with a Kel-F oil manometer. The half-life of the 03 in the
cell was about 8 hi, which was very much longer than the rest of
the procedure,
The kinetic apparatus then was removed from the vacuum line
and transferred to the appropriate spectrometer with the cell in
the optical path. The spectrometer was set at a fixed wavelength,
and the reaction was initiated by opening the T-stopcock first and
then opening and closing the stopcock to the cell. The reaction
was followed continuously for 4 or more half-lives and then inter-
mittently for 3 more to get an infinity point. The initial pressures
of the olefin and diluent gas in the cell were calculated using the
measured ratio of the volumes of the bulbs to the entire kinetic
apparatus. These were 0.485 for the infrared cell and 0.75 for
the quartz cell. Kinetic runs in which the 03 was added to the
olefin were performed in a similar manner by reversing the roles
of the Oa and olefin.
Runs for product analysis were performed in the infrared cell
in the manner described for the kinetic runs, except that the
entire region (4000-600 cm"1) was scanned on a Perkin-Elmer 521
grating infrared spectrometer. HC1 was identified from its known
band sequence (2750-3050 cm"1), but no quantitative estimates
could be made. Quantitative analysis for the olefin, CHC10,
CC13CC1(0), CC120, CH2C1CC1(0), and CO were made from their res-
pective bands. In experiments to measure Oa, the diluent gas was
a halocarbon (CCl^, CHjCla, or CFaCFCla) that was condensable
-------
at -196^0, The residual pressure of the gas noncondensable at
-IS'6°C was considered to be CO plus 02, and the 02 computed by
difference.
For sotne reactions in the C2Cli» system, used In conjunction
with the Beckman IR-10 instrument, the reactants were mixed by
freezing "into a cold finger on the Infrared cell and then per-
mitted to vaporize,
Cl-ATOM INITIATED OXIDATION
The chlorine-atom initiated oxidation of chloro- and chloro-
fluoroethylenes has been studied in our laboratory (5-9). Pre-
viously the chlorine atom initiated oxidation of C2Cli» (10-13),
C2HC13 (14-16), and CHC1CHC1 (17) had been studied. The oxidation
of <:C12CC12, CHC1CC12, CH2C12, cis-CHClCHCl, trans-CHClCHCl,
CF2CCl2, and CFC1CFC1 (mixed cis and trans) proceed by a long
chain free-radical process. The major products are the corres-
ponding acid chlorides containing 1 or 2 carbon atoms. By con-
trast there is no chain process in CHC1CH2. The oxidation
products and chain lengths are summarized in Table I, Also
included in Table I are some preliminary results (18) on the
chlorine-atom initiated oxidation of CF2CF2 and CF2CFC1. These
oxidations also involve long-chain reactions. For most of the
chic roe thy lenea", the chain length of the reaction exceeds 100
at sufficiently high 02 pressures, and Is independent of the
absorbed light intensity, Ia, or any of the reactant pressures.
-------
Table I
The chlorine-atom sensitized oxidation of C2Cl(f_IlHn and C2Cli,_nFI1
i b , , c
JS21 kji-
log k2
Compound
CC12CC12
CC12CHC1
CC12CH2
cis-
CC1HCC1H
trans~
CC1HCC1H
CClHCHa
CF2CC12
CFC1CFC11
CC1FCF2
CFjCFz
Oxidation Products (%)
CC13CC1(0) (75%), CC120 (25%)
CHC12CC1(0) (90%), CO and CC120
CH2C1CC1(0) (98%), CO and CC120
CHC10 (71%), CO (26%) and CC120 (3%)g
CHC10 (71%), CO (26%) and CC120 (3%)8
CHC10 (74%) and CO (25%)h
CC1F2CC1(0) (91%) CC120 (4%), CF20 (4%)
CC12FCF(0) CvlOO%)
CC1F2CF(0) (%95%)
CF20 (100%)
§{OX}a k2b
300f 1
200f ^10
172 >100
21.5 1
21.5 1
^2 >10
^85 >20
420 1
>1000 >10
0,250 1
ktb k6b
150 6.0
100 >6
86 >50
19 <50
19 <50
>22
210 >50
>20
<50
Ref.a
5,13
15,16
6
7
7
9
9
18
18
(M~lsec~1)
10.1
10.6
10.6
10.6
oo
a) ${OX} = -§{olefin}.
b) Calculated from product distribution.
c) At 30-32°C.
d) From Reference 23.
e) Reference for ksa/ksb-
f) At high 02 pressure.
g) In these olefins, geometrical isomerization of the starting olefin is an important process especially at
low total pressure.
h) At high [02]/[C12] ratios.
i) Equilibrium mixture of eis and trans isomers.
-------
The generalized mechanism which explained both the chlorinaticm and oxi-
dation was elucidated by Huybrechts et al (13, 15, 19). In their studies of
the photoehlorination of CC1HCC12 at 363 and 403°K, they found that small
amounts of oxygen inhibited the photochlorination, but that the reaction
proceeded further in the dark after irradiation was terminated (19). They
interpreted this after-effect to be due to the formation of a semi-stable
peroxide which decomposes in the dark on the walls of the reaction vessel to
reinitiate the chain chlorination,
Further studies of the photooxidation of CC1HCC12 (15) completed the
earlier work of Miiller and Schumacher (14) and showed that the principal oxida-
tion product was CC12HCC1(0) which accounted for >90% of the oxidation at
363°K and >82% of the oxidation at 403°K, Likewise they studied the chlorine-
sensitized photooxidation and the simultaneous oxygen-inhibited photochlorina-
tion of CC12CC12 and C2HC1$ at 353.5 and 373.4°K. Both systems produced
CpCls radicals as the chain carrier. The results of the two systems were
the same and nearly identical to those of Schumacher et al (12, 20); 85 ±
5% of tie oxidized CC12CC12 and CaHCl,? appeared as CCljCCl(O) and 15 ± 5%
as CC120. Trace quantities of CC1* (0.3%) and CC12CC120 (0.1%) were also
present. The quantum yield of oxidation, §{OX}, increased with the oxygen
pressure to an upper limiting value of about 300 for C2Cls radical oxidation
and about 200 for CHClaCCla oxidation, independent of absorbed intensity, Ia,
chlorocarbon pressure, C12 pressure, or added Ng pressure, Huybechts et al
(13, 15) emphasized the different light-intensity dependence in the quantum
yields for the oxygen inhibited chlorination (la"1'2) and the high 02~
pressure, limiting oxidation (intensity independent) . They showed that since
these two reactions are coupled they have common chain-breaking steps, which
must be bimolecular in radicals to explain the Ia~*/2 dependence of the
-------
10
quantum yield of chlorination. This led them to propose the following
general mechanism:
C12 + hv •*• 2C1 1
Cl + CX2CXC1 ->• C1CX2CXC1 2a
-»• CX2CXC12 2b
C1CX2CXC1 + 02 -* C1CX2CXC102 3
CX2CXC12 + O-i -* CXC12CX202 31
2C1CX2CXC10? -»• 2C1CX2CXC10 + 02 4a
-> (C1CX2CXC10)2 + 02 4b
2CXC12CX202 •* 2CXC12CX20 + 02 4a'
-»• (CXC12CX20)2 + 02 4bf
C1CX2CXC102 + C1CX2CXC1 -»• (C1CX2CX2C10)2 5
C1CX2CXC10 •*• C1CX2CX(0) + Cl 6a
••*• CXC10 + C1CX2 6b
CXC12CX20 •* CXC12CX(0) + X 6a'
•*• CXC12 + CX20 6b'
ClCXj. + Ot •* CX?0 + Cl + (1/2) 02 la
Reaction 7a, of course. Is not a fundamental, reaction, but must proceed
through several steps, which presumably are:
C1CX2 +(>;>-»• ClCXi.0?.
2C1CX202 -* 2C1CX20 + Of
C1CX20 •* CX20 + Cl
However there is an alternative to reaction 7a that could account for the
oxidation in which the CIO radical is an intermediate but in which C1CX20
is not:
C1CX2 + 02 -*• CXj.0 + CIO 8
CIO + CXgCXCl -*• CX2C1CXC10 9
-------
11
Ms.thias et al (5) tested the two possibilities. They examined the
chlorine-atom sensitized oxidation of CgCl^ in the preserfce of 03 to insure
that CIO was produced via the well established rapid reaction (21)
Cl + 03 •* CIO + 02 10
I 1
With Oj present, the epoxide, CClaCCliO, was produced. Since the epoxide
production depended on the ratio [C2Cli»]/t03], it was concluded that the
epoxide came from
+ CIO -»• CC12CC120 + Cl lla
No epoxrlde was produced in the absence of 03> so presumably CIO radicals
are absent, and reaction 7a is the correct representation of the oxidation
of CCla radicals.
From the experiments of Huybrechts et al (13, 15) and Mathias et al (5),
reaction 7a was established for CC13 radicals. In order to examine the oxi-
dation of partially chlorinated methyl radicals, SanhueEa and Heicklen (22)
examined the chlorine-atom sensitized oxidation of CHaClg and CH3C1 to
study the oxidation of CHClj and CHgCl, respectively. They found that CHC12
oxidized just like CCl^, but that CH^Cl oxidation did not generate the
chlorine atom. Presumably the oxidation of this radical is analogous to
that for CEj radicals.
C1CH2 + 0? + CHC10 via termination 7b
In the CClgCCla and CHCICCI? systems, the quantum yields of oxidation
products increased with the Qj pressure until upper limiting values were
reached. The mechanism predicts that if termination is exclusively by
reaction 5, then
{OX} - (%4^ '
-------
12
However if termination is exclusively by reaction 4b, then
t{OX} = 2ki,/k^b II
where ${OX) 5 -${olefin) = *{C1CX2CC1(0)} 4 (1/2)${CX20.} + (1/2)${CO}.
Eqn. I applies at low values of [02]2/Ia, whereas eqn. II applies at
high values of [0?]2/Ia.
In the oxidation of CHgCClj,, cis-CHClCHCl, trans-CHClCHCl and
CFC1CFC1 the quantum yields of the oxidation products are insensitive
to all the reaction parameters and the termination must be by reaction
4b exclusively; eqn, II always applies. However in the oxidation of
CCl?CH2, since CHaCl always is oxidized in a non-chain process,
reaction 6b' also can be a terminating step,
In the oxidation of CFiCCl^ there is one striking difference from
the results of the other chloro-and/or fluoroethylenes: ${CF2C1CC1(0)}
is reduced at high pressure, but fiCFjO} is not. Thus the details of
the mechanism contain some additional subtle deviation from the general
mechanism outlined above, A possible explanation is given in detail in
the original work (9).
In Table I the experimental results are summarized, and in Table II
the bond energies of the chloroethoxy radicals involved In the process of
oxidation of the chloroolefins are presented. From Tables I and II it
is possible to deduce the following information:
1) From the values of k^a/^-^b (obtained mainly from the distribution of
products) it is possible to conclude that the chlorine atom prefers to
attack the less chlorinated carbon atom. In the most unsymmetrical cases
(CCljCHj and CCl?CF?), the preference for the non-chlorinated carbon atom
is at least a factor of 20. The chloroolefins with one or three chlorine
atoms also show a high preference for substitution on the less chlorinated
-------
Table II
Bond energies (kcal/mole) in chloroethoxy radicals3
R0
CC13CC120
CC12HCC120
CC13CHC10
CC1H2CC120
CC13GH20
CC12HCHC10
CClH2CHeiO
CC12HCH20
D{C-H}U
—
—
6
—
17
6
2
14.8
D{C-C1}D
-17
-16
-4
-20.8
—
-5
-4
__
D{e-cr
-20
-13
-16
-11.6
11
-11
-8
7
D{C-C1} - D{C-C}
+3
—3
+12
-9.2
—
+6
+4
— _
ksa/ksb
6.0
>6.0
<10
>50
c
<50
<10
>_d
a) The values are mostly from Reference 24.
b) D{C-H} and D{C-Cl) represent the bond energies for loss of H or Cl, respectively, from the oxygen-
bearing carbon atom.
c) Neither of the products CC13CH(0) nor CH20 was observed experimentally (k2a/k2b £ 100).
d) Neither of the products CHC12CH(0) nor CH20 was observed experimentally.
-------
14
carbon atom. In particular this was demonstrated for CClaCHCl by Bertrand
et al (16) at 3576K, who found the preference for chlorine-atom addition to
the less chlorinated carbon atom to be at least 8 times greater than for
addition to the more chlorinated one. They did this by comparing the pro-
ducts of the reaction with those produced from the photochlorinated oxidation
of CH2ClCCla (to produce CCljCClH) and CHC12CHC12 (to produce CClzHCCU).
Of course the symmetrical chloroethylenes can show no preference and k2a and
kab are indistinguishable.
The inductive (I~) and mesomeric (M+) effects of the three substituent
atoms are F > Cl > H.
5- 5+
C1K 6+/ C1x- «"/
\)*-~N / \ ^ /
xc=c -c-—~c/
Cl/ ^H Cl/ ^H
Inductive Mesomeric
&~ &+
FxV~T/cl \, r/cl
(]•=£' XO~~(r
F/ ^Cl F-^ Cl
Inductive Mesomeric
If these effects dominated the chlorine-atom addition, then H and F sub-
stitution should give different results. However the chlorine atom always
prefers to add to the less chlorinated carbon atom. Thus we conclude that
steric effects must dominate the addition process.
2) A long-chain oxidation (> 150) occurs when the exothermieity of either
reaction 6a or 6b is greater than 11 kcal/mole. For an exothermieity of
11 kcal/mole, a relatively short-chain length (^ 20) is involved.
-------
15
3) In CHCICH?, the radical produced is CHjClCHCIO, and the exothermlcity
of decay of this radical by any route Is < 11 kcal/mole. The favored
route to decay (most exothermic) is by C-C cleavage which produces the
terminating radical CHjCl. Thus one cannot be certain that the parent
radical would, of itself, lead to short chains.
4) The reaction
G1CX2CXHO * C1CX2CX(0) + H
is always energetically less favorable than the cleavage of either the
C-C bond or the oxygenated carbon-chlorine bond. There was no evidence
that thi.s reaction occurred in any of the systems studied.
5) In radicals of the type CX^CHjO, all the decomposition routes are
sufficiently endothermic so that no decay products are observed. Thus in
the oxidation of CC1?CH2 no CH^O or CC13CH(0) was found as products, and
in the oxidation of CHClCHj, no CHCliCH(O) or CH20 was found as products,
6) When D{C-C1} - D{C-C} is >. 6 kcal/mole, almost all the chlorinated
ethoxy radical decomposition goes through reaction 611, When D{C-C1} -
D{C-C} •< -3 kcal/mole almost all the decomposition proceeds through
reaction 6a. For intermediate values of the bond energy difference, both
reaction paths are significant.
?) The chlorine-atom initiated oxidation of all the perfluorochloro-
olefins (CjF^-nCln) which have been studied gives long chains. No study
has been made for CClFCCla, but there is no reason to believe that its
oxidation will not proceed through a long-chain process.
8) In the mixed chlorofluoroethylenes, the products are almost entirely
(> 90%) the two-carbon acid chloride. Thus we would expect that D{G-C1} -
D{C-C} ' -3 in the ethoxy radical precursor. This observation can be com-
pared to the results for cis- and trans-CHClCHCl, where no two-carbon carbonyl
compounds were found. Thus the substitution of F for H either strengthens the
-------
16
C-C bond or weakens the oxygen-bearing carbon-chlorine bond or both.
9) In CaF^j the two-carbon carbonyl compound is missing. Thus we conclude
that the C-F bond is stronger than the C-C bond in the ethoxy radical,
whereas In CXsCXCIO the carbon-chlorine bond is weaker than C-C bond.
Vinyl Chloride;
The chlorine-atom sensitized oxidation of CHClCHg Is unique among the
chloroolefln oxidations for three reasons: 1) CO is produced as a major
initial product of the reaction, the ratio [CO]/[CHC10] being almost
independent of reaction parameters, 2) there is no chain at high values
of the ratio [02]/[Clj] and 3) at low values of [02]/[C12], there is a
long chain process which consumes Clz and produces CHaClCClCO) as the
principal chain product.
The production of CO as an initial product Is explained by a slight
extension of reaction 6
CH2C1CHC10 -* CH2C1 + CO + HC1 6c
where reaction 6c probably proceeds through an energetic CHC10 molecule
which always decomposes. Reaction 6b to produce CHC10 as a product still
occurs but represents that fraction of reaction 6 in which the CHC10 pro-
duced is stabilized. The same results were found in the CHC1CHC1 oxidation.
The lack of a chain reaction at high [02]/[C1?] pressures is the
result of the fact that reaction 6a does not occur and that reactions 6b
and 6c produce the terminating radical CHaCl. Presumably the termination
reaction 7b, occurs via the sequence of steps (22)
CH2C1 -1- 02 ->- CHaClOa
2CH2C102 -*• 2CH2C10 + 02
followed by
CH?.C102 + CH2C10 -> 2CHC10 + H20
-------
17
or
CH2C10 + 02 ->- CHC10 + HO 2
CH2C102 + H02 •*• CHC10 + H20 + 02
The production of CH2C1CC1(0) and the dependence of the results on
the Glj pressure represent findings not seen in any other chloroethylene
and vrhich were explained (8) by the competition:
CH2C1CHC1 + 02 -» CH2C1CHC102 3
CH2C1CHC1 + C12 •* CH2C1CC12 + HC1 12
where reaction 12 is then followed by oxidation to produce CH2C1CC1(0) as
it does in the CH2CC12 system. The competition between reactions 3 and 12
leads to the rate, law
»{CH2C1CC1(0)} -^ IH
since ultimately reaction 12 regenerates the chain. A log-log plot of
${CH2C1CC1(0)} vs [02]/[C12] is shown in Fig. 1. It is fitted reasonably
by a istraight line of slope - 1. The intercept yields a value for ki2/ka
= 9.5,
Confirmation that the chlorine-atom initiated oxidation of CC1HCH2
does riot lead to a chain process comes from the work of Bertrand et al (24)
who studied the chlorine-atom initiated oxidation of l,2-C2Hi,Clj> at 353°K
to produce CC1H2CC1H. Furthermore they showed that chloroethyl radicals
not chlorinated on the a- carbon do not lead to chain oxidations by examin-
ing the chlorine-atom initiated oxidation of C2H% (to produce CC1H2CH2)
and -CCljCHj (to produce CClsCHa). Earlier work (25) on the chlorine-
atom initiated oxidation of C2H6 had shown that C^s also does not enter
a ehai:a oxidation.
-------
18
50
10
o
o
o
-------
19
PS. Present;
The chlorine-atom initiated oxidation of C2Cl!» was studied in the presence
of 03, since the dark reaction for this system was very slow (5). The addi-
tion of 03 to the system introduced three major changes:
1) The ratio ${CC13CC1(0)}/${CCl20} dropped as the [03]/[02] ratio was
increased but was unaffected by changes in [OsJ/fCaCli,]. The effect of
the [03]/[02] ratio is seen in Fig. 2, There is considerable scatter in
the data, but at 32° the ratio drops from about 3.0 in the absence of 03 to
about 1.0 at [Oal/tOa] > 10. The data points at 24°C lie .below those at
32°C, as they do in the absence of 02. The shift in the ratio was attri-
buted to the production of CC13CC1(0) and CC120 via
C2C15 + 03 -> CC13CC1(0) + Cl + 02 13a
•* CC120 + CC1302 (or CC13 + 02) 13b
where i:he ratio k13a/kj3b is smaller than kga/^sb- It was argued (5) that
reactions 13a and 13b proceeded directly and not through energetic C2ClsO
radicals, since the thermal effect was known to move the product ratio in
the opposite direction (5).
2) The overall rate of the oxidation was reduced as the [03]/[C2Cl£,] ratio
was raised. This was attributed to the production of CIO radicals via
reaction 10 followed by the competition of CjClij and 0$ for CIO. Most
of the time that CIO reacted with 03, the chain is regenerated.
CIO + 03 -* Cl + 202 I4a
but occasionally termination might occur by
CIO + 03 •*• OC10 + 02 14b
followed by subsequent oxidation of OC10 to produce the observed product
C1207.
-------
iO i-
U
O
10
o
o
fO
U
O
-I
10
O A
'*
00
o
10
-2
10"
io
I, 24 ®C
O 32 SC
o
10'
so'
Figure 2:
Log-log plot of the ratio of the quantum yields of CClsCCICO) and CC120 vs. [03]/[02] in
the chlorine-atom sensitized oxidation of C2Clif by 62 and 63 at 32°C. From Mathias et al
(5) with permission of the National Research Council of Canada.
-------
21
3) fcciaCeiaO was produced, its quantum yield depending mainly on the
(03j ratio. This result suggested that it was produced in the Cl
interaction:
CIO + C2C14 -> CC12CC12
-------
rO
_JM
U
U
_«M
U
3-
o
f 3
OJ
^
A 2
O
o
U
o
o
TEMP =32* C
N/M
o < o.s
® O.I -I
O - > 1.0
O.I
0.2
0.3
0.4
0.5
0.6
[C2C!4]/[0:
10
0.7
Figure 3:
Plot of (${CC13CC1(0)} + (1/2)${CC120})/
-------
23
()|i Absent;
The Hg photosensitized decomposition of CC12CC12 (27) and CHClCClj. (28)
*rere studied. The results were similar in the two studies, The products
were Hg2Cl2 and polymeric material. The quantum yield of olefin loss,
-§{CX2CG12} (X = H, Cl), was ^ 1, independent of olefin pressure and
nearly independent of absorbed intensity, Ia. (-${CHC1CC12} appeared to
te between 1,5 and 2.0 at low Ia). In the CHC1CC12 study small amounts
of another unidentified product were found.
The results indicate that a long-chain polymerization of the olefin
is not involved, since -t{GX2GGl2} - 1.0. Double-bond cleavage can be
eliminated since c-CaCli, was not produced in the CC12CC12 system and
nixed ethylenes were not produced in the CHC1CG12 system. Molecular
elimination does not seem likely, and in fact the results with G2 present
eliminate that possibility as a major reaction path. It was concluded
that free radicals must have been produced by one of the following
processes:
Hg 6(SP) + CX2CC12 -* C2X2C1 + (1/2) Hg2ei2 15a
o:f
Hg 6(3P) + CX2CC12 •* Hg eC^S) + CX2CC12* 15b
Hg 6(;S) + CX2CCl2*-> C^XjCl + (1/2) Hg2Cl2 16
In the case of CHClCClj, the possibility also exists of producing C2Cl3
+ H + Hg 6(XS) as products, either directly or through the excited mole-
cule mechanism. Presumably the C2X2C1 radical dimerizes, and the resulting
1,, 3-butadiene polymerizes.
The above mechanism to produce free radicals is markedly different
than for the Hg-photosensitized decomposition of the fluoroethylenes or
C;;H».. Ethylene and the fluoroethylenes (except for C2Fi«) decompose by
-------
24
molecular elimination of E2 (,29) and HF (30) respectively, CzFi, (31-33),
and to a slight extent trifluoroethylene (30b), decompose by double-bond
cleavage ,
For CjFi, the mechanism that explained the results was:
Hg 6(3P;L) + CF2CF2 -> Hg 6(1S0) + (CF2CF2)n* 15b'
(CF2CF?)n* + 2 1CF2 17
(CF2eF;}n* 4- CF2CF2 -* (CF2CF2)0* + CFzCFs 18
(CF2CF2)0* (+ CF2CF2) -»• C!'ZCF2(+ CF2CF2) 19
followed by
2 1CF2 -* C2Fi, 20
^Fg + CaFit ->• c-~C3F6 21
where the superscript * represents an electronically excited state, the
subscripts n and o represent, respectively, molecules with either
sufficient or insufficient energy to dissociate, and *CFa is the singlet
CFz di radical.
The rate coefficients for the quenching of Hg 6(3P) by the olefins
have been measured. Relative to NaO they are 3.0 for CzCl^. (27), 4.1 for
^ (29), 0.35 for C2F!f (1), and 1.8 for GaH^ (29).
Present;
The Hg-photosensitized oxidation of C^F* has been reported in two
studies (33,34) and reviewed by Heicklen (1). A complete mechanism has
been presented and discussed in detail (34) . The products of the reaction
were cyclo-CsFs, CFjO and CF2CF2U (tetrafluoroethylene oxide).
The mechanism is very complex. However the oxidation products can
be explained as coming from the following reactions involving a diradical.
(C2F%)0* + 02 •* CF202 + 3CF2 22
)CF2 + C2F>, -* c-C3F6 21
-------
25
CF202 + C2F% •+ 2CF20 + 3CF3 . 23a
•+ CF20 + CF2CF21) 23b
JCF? + Oj ••> CFjOi 24
Here (CjF-Jo* is a vibrationally equilibrated electronically excited
C?Fj, molecule; 'CF2 and 3CF2 are the singlet and triplet of CF2 radical
respectively.
For the chloroolefins in the presence of Oj, a long-chain process occurs.
The major products are the same as for the chlorine-atom initiated oxidation of
the corresponding chloroolefin, and these products are produced in the same
ratio. There are two major differences for the two modes of initiating
the oxidation:
1) At high [OaJ/Ig1/2 the quantum yields are independent of the reactant
pressures and Ia for chlorine-atom initiation. For Hg 6(.3P) sensitization,
at: low [Oi]/[CX2CCl2] (to minimize removal of Hg 6(3P) by 02) the quantum
yields are independent of the 02 pressure and Ia» but they increase pro-
pcrtionately with the chloroolefin pressure.
2) CO is produced as an initial product from the Hg 6(3P) sensitization
but not from chlorine-atom initiation. (CO was found in the Cl + CHC1CC12 +
0« system (15), but presumably it is a decomposition product of CHC10 or
formed through energetic CHC10 as a precursor).
The conclusions from the above observations are that the same free-
radical chain process must occur in both systems, but that the initiation
mechanism must, be different. In order to explain the facts, Sanhueza and
Heicklen (6, 27, 28) utilized the following mechanism for the Hg 6(3P)-
se'tisitized oxidation:
Hg 6(*P) + CXiCCl2 -> CX2CC12* + Hg 6(1S) 15b
Hg 6(3P) + 02 -f 02* + Hg e^S) 25
-------
26
Oj.* + CXfcCCli •* CXiCCla* + 02 • 26
+ Hg 6C*S) •* C;X2C1 + (.1/2) Hg?,Cl?, 16
+ O-i •* C2,X2C102 27
•» CO via termination 28
C?XZC10? + CXzCClz -•* C2X2C13 + (CXO)2 29
In the hydrogenated ehloroolefins, the product yields decreased as
[02J/[CX2CC12] increased beyond a certain value (4.0 for CHC1CC12 and 1.8
for CHjCCla). However with CaCli*, there was no decrease in the product
yields even at [02 ^[C^Cli, ] = 22, in spite of the fact that quenching of
Hg 6(3P) by Oj is only slightly less efficient than quenching by CzCl.it.
Thus it was necessary to postulate that Hg 6(3P) sensitization leads to the
production of an excited olefin molecule, regardless of whether the olefin
I
or 02 quenched the Hg 6( P) atom. The same postulate was required in the
CjFs, system (1).
Presumably the free radical CiXzCl adds Og and this radical must
initiate the chain process, Since the chain process is proportional to
[CXj-CCi? ], and initiation must be via reaction (29), the main removal
process for C^X-jClO^ must be by some process represented by reaction 28,
Reaction 28 is, of course, not a fundamental process. In order for term-
ination ultimately to occur, another radical must be involved. Possibly
reaction 28 occurs on the wall of the reaction vessel. In any event it
must produce the excess CO that is observed as a product.
The mechanism leads to the following rate laws;
${COj =1,0 for CC12CC1? and CH2CC1? V
${CO) = 1.0 + <3>{CC1?0} for CHC1CC1-,. V
-*{CX2CC1?.} -1 = (k^g/k^bk^HezCU] VI
-------
2?
${00} should be higher for CHC1CC12 than for the other chloroolefins because
no CHC10 was found. Presumably it was formed "hot" and always decayed to
CO + HC1. For the C^Cl^ system, f{CO} was unity in good agreement with
expectation. In the CH2CC12 system, ${CO} was somewhat low (0,5 - 1.0),
bu1: this probably reflects experimental uncertainty since CO is a minor pro-
duct. However in the CHCICCI* system, ${CO} ~ 1 was larger than predicted
by eqn. V' by a factor of 4.6, and this extra CO has not been satisfactorily
explained.
Table III summarizes the rate coefficient data obtained for the three
olefins. The values of kjg/kzs are of the same order of magnitude in the
three systems; the variation that does exist does not follow any trend.
It is interesting to compare the above results with those of the photo-
oxidation of C^FS! (35) which produces CaFj, a radical analogous to
C2F3I + hv -*• C2F3 + I 30
In this system a small chain occurred which produced CFaO, CFIO, and
as major products and (CFO)2CF? and (CFO?)CF2(CFO) as minor products. The
latter products must involve CFO as a precursor and come from the oxidation
of CjjFj
C2Fj + 02 •* CF20 + FCO 31a
The main chain steps considered were
31b
CFIO 32a
-> C?.F3OI -I- C2FjO 32b
+ I -»• C^FjOI 33
Though Heicklen (35) did not consider them, additional chain steps are also
«.
possible through the iodine atom as chain carrier
I + C2F3I * CFjICFI 34
-------
28
Table III
Rate coefficient ratios from the Hg 6(3P) sensitized
oxidation of chloroethylenes at 30-32°C
^^_ : Value for
a
Ratio Units
None
Torr"1
CCl2CCl2a
1
0.029
CHClCCl2b
6.7
0.015
CH2CC12C
0.058
a) From Reference 27.
b) From Reference 28.
c) From Reference 6,
-------
29
CP2ICFI -f Q2 -* CFfclCFIO^ _ 35
2CF2ICFI02 -* 2CF2ICFIO + 0* 36
CF^ICPIO * Cj.F3OI + I 37a
0?
-*• CFIO + CF2I •+ CF20 + I 37b
In fact reaction 37a is needed to explain the high quantum yields (up to
10) of C2F3OI which were observed. Reactions 32b and 33 can only account
for f{ej>F3Ol} <. 2, It should be noticed that CaFsOI Is the enol type
isomer of CFtlCFCO), the molecule expected to be produced via reaction 37a
if i:he system were exactly analogous to those of the chlorinated ethylenes.
Thus this system contains many analogous features of the Hg 6(3P)
sensdtized oxidation of chloroethylenes. Reaction 31a is a termination
step corresponding to reaction 27 followed by reaction 28, and reaction 31b
followed by reaction 32a is a propagation step corresponding to reaction 27
followed by reaction 29.
REACTION WITH 0(3P) ATOMS
The oxygen atom might react with chloroethylenes by any of the following
paths: (X - H, F, Cl)
0(3P). + CXCieX* •*• CX?0 + CC1X 38a
•* CO + X* + CC1X 38b
-»• cxciex?o* 38c
-* CX}C1CX(0) • 38d
-* CC1X2 4 XCO 38e
Each of the reaction paths actually represents only one of two possibilities,
since the 0(3P) atom could attack either the more heavily or less heavily
chlorinated carbon atom. However we anticipate the results and show that
the 0(3P) generally attacks the less chlorinated carbon atom in the
chloroethylene.
-------
30
The diradical CC1X enters into one of two reactions.
2CC1X2 -» CC1XCC1X 39
CC1X + CC1XCX2 -» 6ciXCX2.CClX 40
The mono-free radicals either add to the chloroethylene or react with each
other via combination or disproportionation reactions.
The excited adduct formed in reaction 38c is a species which may
undergo many reactions:
1) It may polymerize either with itself or with the parent chloroethylene,
2) It might stabilize as the epoxide,
3) It might react with parent ethylene to give a. cyclopropane and a
carbonyl compound.
4) It might rearrange to an excited aldehyde which could be stabilized
or decompose to free radical products.
Individual Substituted Ethylenes;
Each of the substituted ethylenes behave slightly differently from
the others, and we now examine them individually.
CC1.2.CC1 ?: The mercury-sensitized photolysis of N?0 in the presence of
C;.C1, at 25CC yields N?, CCl^O, and polymer as exclusive products (36).
The absence of CCljCClgO and c-C3Cl6 as products indicates respectively,
that all the CaCl^O* polymerizes and that none of the CCla adds to C2Cli»;
reaction 40 is not operative in this system. Since HKCaCli,} "^ 1 the
polymer does not incorporate additional CCljCCla, and its formation must
come only from reaction 41,
CC1XCX20* ->• Polymer 41
The mechanism predicts that
*{CC120} - k383/^38 VII
lJ = 1 - k38a/2k9e VIII
-------
31
was found to be 0,19 independent of reaction conditions, jjp this
- -I'"'*,- ' ' . 'V\«*. <:^
Is the value of ksaa/kse- With this value, -tCCaClij} should be 0.9, which
is In agreement with the experimental results.
CCJLj.CHCl; The products of the reaction were CO, CHC13 and polymer (37).
./
The quantum yields of CO and CHC13 were 0.23 ± 0.01 and 0,14 ± 0,05
respectively. Thus the reaction channels involved in this system are
reactions 38b and 38c, with k38b/k3e = 0.23. Reaction 38b presumably
proceeds through an energetic CHC10 molecule which rapidly decomposed to
HC1 -f CO. No CC120 was produced and the cleavage products gave entirely
CC12, and no CHC1, diradicals . Most of the HC1 and CCla combine to form
CHClj, but some other products must also be formed to account for the
* ; *' >•
diffsrence in the CO and CHCls quantum yields.
Presumably CiHClsO* always polymerizes, since no epoxide or aldehyde
was found as products; and the polymerization must proceed without involving
additional C2HClj molecules through reaction 41 since ^the quantum yield of
CjiHC.*1. 3 disappearance was about 1.0.
The mechanism predicts that:
*{CO} - ${CHC13} = k98b/k38 IX .
-4{C2HC13} » 1.0 X
The results of the oxidation of CHC1CC12 in the three systems initiated
by Cl atoms, Hg 6(3P) and 0(3P) are summarized in Table IV. The three
systems give consistent findings,
CC_li>CHa! The reaction of 0(3P), prepared from the Hg photosensitization of
N20, with CC12CH2 was studied at 25°C (38). The products of the reaction
were CO, CH2C1CC1(0), polymer and another unidentified compound. The quantum
yieldB of CO and CH2C1CC1(0) were 0.35 and 0.06 respectively, independent of
reaction conditions.
-------
32
Table IV •
Oxidation of CClHCCla
System ${CO}/${CClgO}
Cl atom — 0.09
Hg 6(3P) 1.85 0,16
0(3P) 1.7 0.10
-------
33
Twelve possible reaction paths between 0(3P) and CC12CH2 were dis-
cus£;ed, the final conclusion gave the following mechanism as most likely (38)
4> = 0.31 0(3P) + CC12CH2 -*• CO + HC1 + CHC1 38b'
CC12CH20* 38c
= 0.06 -*• CH2C1CC1(0) 38d'
4> = 0.04 -> CHC12 + HCO 38e
CHC1CHC10* 38c
-* CH2C1CC1(0) 38d
followed by
CHC1CHC10* + CHC1CHC1 -> polymer 42
-------
34
CC1H + CHC1CHC1 •> Products 40'
2CHC1 > CHC1CHC1 39
The fractional importance of channels 38b and 38d are given by ${CO}
and ${CH2C1CC1(0)}, respectively. Thus k38b/k38 ~ 0.23 for the cis isomer
and 0.28 for the trans isomer. The ratio k38d/k38 £0,04.
CHCICH?; The reaction of 0(V) with CHC1CH2 gives as products CO, CH2C1CH(0),
CH3CC1(0), HC1, CHs,, and polymer (8). The quantum yields depend on the total
pressure (mainly N^O), and are given below for high and low pressure conditions.
Product $([NgO] * 400 Torr) $([N20] - 35 Torr)
CO 0.25 0.40
CH2C1CH(0) 0.40 0.25-0.35
CHiCCl(O) 0.09 0.07
The quantum yield of CHC1CH2 removal exceeds 1.0, and it as well as
${CH3CC1(0)} is independent of total pressure.
The reaction was explained by a scheme similar to that for the reaction
of 0(3P) with C2H<. (39) and which is substantially different from that for
the other chloroolefins.
0(3P) + C2H3C1 * C2H3C10* 38c
C2H3C10* + CH2 + CO + HC1 43a
* CH2C1CH(0)* 43b
->• CH3CC1(0)* 43b'
CH2C1CH(0)* * CH2C1 + HCO 44
CH2C1CH(0)* + M + CH2C1CH(0) + M 45
CH3C1CH(0)* -*• CH, + Cl + CO 44'
CH,CC1(0)* + M •+ CH3CC1(0) + M 45'
The excited intermediate, CaHsCIO* can decompose or rearrange to one of the
aldehydes which still contain the excess energy of reaction. If not
-------
35
deactivated they decompose to radical fragments. Thus at higher pressure
relatively more aldehydes and less CO are produced. The radical fragments
react with the C2H3C1 to form the polymeric material* Presumably some of
the CHj radicals abstract a hydrogen atom to give the small amount of CH^
produced. Reaction 43a has been included for completeness, but in fact
there is no evidence that it actually occurs. Apparently the CiHjCIO* is
never deactivated to the epoxide, since no epoxide was found.
CHgCH^; The reaction of 0(3P) with C2H% has been interpreted traditionally
(39) by the mechanism
0(3P) + CH2CH2 •* CH2CH20* 38c'
CH2CH20* •*• CH3CHQ* 43b"
CE2CH20* + M -> CH2CH20 + M 46
CH3CHO* •* CH3 + HCO 44"
CH,jCHO* + M •-»• CH3CHO 45"
Mos t of the products could be interpreted as coming from the free radical
fragments produced in reaction 44". The yields of both CH2CH20 and CH3CHO
Increased slightly with pressure indicating that they were produced, at
lea»t in part, from the pressure stabilization of the energetic intermediates.
Further evidence for this comes from the work at liquid N2 temperature
where the sole products were CH2CH20 and CH3CH(0) in a ratio of 1.2 (40).
More recent experiments at room temperature utilizing crossed beams or
a fast-flow reactor coupled to a photoionization mass spectrometer (41, 42)
hav« confirmed the presence of a small additional process, accounting for
5 % of the total reaction, to produce CH2CO + H2 directly as earlier
suggested by Cvetanovic (43, 44).
0(3P) + CH2CH2 -* CH2CO + H2 38f
-------
36
CHgCHBr; Slagle et al (45) have shown that three paths occur in the
reaction of 0(3P) + CH2CHBr:
0(3P) + CH2CHBr -»- CH3 + BrCO -> Br + CO 38a'
-> CH2Br + HCO 38a"
-*• CH2CO + HBr 38f'
Reactions 38a' and 38a'' may proceed through energetic intermediate adducts.
The relative importance of the three channels was 0.29, 0.51, and 0.21.
CFgCFg! Oxygen atoms react with CaF^ to produce. CF20 and c-C3Fe as
exclusive products (1). The reaction was studied by Saunders and Heicklen
(46) at temperatures of 23 and 125°C over a wide range of oxygen-atom con-
centrations and with C2Fi, pressures from 3-123 torr. The quantum yield of
CF20 production is 1.0 for all conditions.
The mechanism of reaction, is explained by
0(3P) + CF2CF2 -*• CF20 + CF2 38a
-*• C^O* 38c
with channel 38c occurring 15% of the time. The CF2 and C2Fi,0* are removed
as follows:
2CF2 -*• C2Fi» 39'
CF2 + C2F^ -> c-C3F6 40'
* -> CF20 + CF2 41'
C2Fi» -»• CF20 + c-C3F6 42'
C?FnH<>-n (n=l. 2 and 3); For the fluoroethylenes, there is not a complete
study of the reaction mechanism.
Mitchell and Simons (47) studied the reaction of CH2CF2 through flash
photolysis of N02 - olefin mixtures, and continuous photolysis of N02-
fluoroolefin mixtures. In the flash photolysis experiment ground state
-------
37
CF;> was monitored qualitatively from Its U.V. absorption bands (which persist for
> 60 msec.). Using CaFi, as a reference, Mitchell and Simons concluded that CF2 is
produced in large amounts in the GH2CF2 reaction when the [CH2CF2]/[N02]
ratio is high (to minimize the competition between N02 and CHgCFa for the
0(SIP) atom). :
The final products from continuous photolysis were: CO, C2Fi»t and
veiy small amounts of CF20. It was concluded that reaction occurs
via
0(.3P) + CH2CP2 -*• CH20 + CF2 38a
-> CF20 + CH2 38a'
and that CHgO was formed with sufficient vibrational energy to dissoci-
ate. Since no quantum yields were reported we can only tentatively outline
a mechanism similar to that proposed for the chloroethylenes
0(3P) + CH2CF2 -»• CF20 + CH2 38a'
-> CO + H2 + CF2 38b
2CF2 -*• C2F% 39'
It is interesting to point out that no CF2CH2CF2 was reported as a product.
For the same olefin both CF20 and CH20 were observed as products by
Huiis et al (48). However the authors pointed out that CF20 could be from
a secondary reaction with molecular oxygen in the reaction mixture.
In Table V.are the aldehydic products reported by Huie et al (48) in
the reaction of 0(3P) with fluoroethylenes, No quantitative estimates were
made.
Moss (49) reported that carbon monoxide is a primary product formed
in high yields in the reaction of CH2CHF with 0(3P).
-------
Table V
Products mid qaaatua yields is the reaction of 0(SP) with chloco-, dilorofluoro™, masi fluoroethylenes
Ha^oethylene
CCi2CClj
CCljCClH
OClaCH2
Cis-CClHCClH
trans-CClHCClH
CClHCHs
CF2CF2
CFjCFCl
CFC1CFC1
CF,CC12
CFiCFH
CFjCHz
CFzCHCl
CFjCFBr
CHjCHF
CFBCFH
Products
CCljO and Polyrcr
CO, CHC1, and Polyaer
CO, CHtClCCKO) and Polyaer
CO, HC1 and Polyncr
CO, HC1 and Polyaer
CHzClCH(O), CO,
CH,CC1(0) and Poly re r
CFzO and c-C3F6
CFtO
CFC10 and Polynec
CF20 and CCljCFjCClz
CF2O. CHFQ
CO, CZF», CF20
CFjO, CO, HC1
CFjO
CHFO, CHZ0 and CO
euro
_£.
0.19
0.23
0,35
0.23
0.28
•V0.4
1.0
1.0
0.80
1.0
—
._
_
_.
—
—
fc»«a/k«.
0.19
0
0
0
0
0
0.85
0.80
0
kj,b/k«.
0
0.23
0.31C
0.23
0.28
^.25
0
0
0
k,.e/k..
0.81
0.77
0.55
0.73-0.77
0.48-0.72
>0.30
0.15
0.20
1.00
ki.H/U..
0
0
0.06C
<0.04
<0.04
0.09-0. M
0
0
0
Kim/km
0
0
o.osd
0
0
0
0
0
Sanhueza and Heicklen (36)
Sanhueia and Hzlcklen (37)
Eantiusss ami Halehlen (38)
Sanhueza and Heicklen (7)
Sanhueza and Heicklen (7)
Sanhueza and Heickten (8)
Sanhueza and Heicklen
00
a) For ESSIE details see text.
b) Quantum yield of the principal oxygenated produce (Cirst listed in Products column) . $
c) These yields are for the abnormal reactions involving Cl atom nigration (aee text).
d) Total yield » 0.08. About 1/2 of the yield involves the nornal H atom nigration; and about 1/2 of the yield, the abnormal Cl atom migration (see textX
(where x m atb,c, or d) .
NOT REPRODUCIBLE
-------
39
CFCICFCI; Sanhueza and Heicklen (9) reported on the Hg-photosensitized
decomposition of N20 in the presence of an equilibrium mixture of cis- and
trans-CFClCFCl. The products were Na, CFC10, polymer, and an unidentified
compound. ${CFC10} was 'V- 0.80 independent of a factor of 6.7 change in
[CFCICFCI] and a factor of 14 change in Ia (at high I20 pressures). The
values for -${CFC1CFC1} showed some scatter, and they varied between 1.0
and 3.4, suggesting that more than one CFCICFCI is removed per 0(3P). The
unidentified product was probably cydo-(CFCl) 3 and its relative yield
showed no trend with changes in reaction conditions.
The reaction is most easily described by the mechanism:
0(3P) 4- CFCICFCI .-»• CFC10 + CFCl 38a
• , ..--»" CFC1CFC10* 38c
2CFC1 -»• CFCICFCI 39
CFCl + CFCICFCI -> cyclo-(CFCl) 3 40
CFC1CFC10* + CFCICFCI •* Polymer 42
with the ratio k38a/k38 » 0.80.
_CF^CC1,|_;- The reaction of 0(3P) with CF2CC12 gives CF20 and CC12CF2CC12,
both with quantum yields of about unity and with -${CF2CC12} = 2 invariant
to reaction conditions (9). The mechanism consistent with the other chloro-
oleilins was discarded for the following reasons:
1) No C2Cli» was found.
2) ^{CClaCFgCCla} should be pressure dependent and intensity dependent
unless reaction 39 never occurs.
3) In the presence of 02 (discussed in the next section) the long-chain
free-radical oxidation should occur and ${CF2C1CC1(0)} should approach 45.
In fact it never reaches 3.0.
-------
40
4. In the presence of 02, the long-chain oxidation should be a function of
[CCFaCCliJ/Ia1/2. In fact ${CF2C1CC1(0)} is independent of Ia, but depend-
ent on [CF2CC12]/[02]. Thus, Sanhueza and Heicklen (9) proposed the
following mechanism:
0(3P) + CF2CCl;i; -»• CC12CF20* 38c
CC12CF20* + CF2CC12 •+ CF20 + CC12CF2CC12 42'
This mechanism predicts that
${CF20} = ${CCl2CF2(ici2} = 1.0 XI
which conformed to the findings.
In Mitchell and Simons (47) system (discussed above) no CF2 was pro-
duced from CF2CC12 in their flash photolysis experiments and CC120 was the
main product from continuous photolysis. Tyerman (50), who looked for
ground-state CF2 by kinetic spectroscopy after long wavelength flash photo-
lysis of CF2CC12-N02--N2 mixtures reported that no CF2 diradical is released
from the reaction,
CFgCFCl; In the reaction of 0(3P) with CF2CFC1, Mitchell and Simons (47)
reported CF20 and CFC10 as products, the former being the more important.
In their experiments with 02 present, [CF20] - [CFC10], so it is possible
that the small amount of CFC10 detected in the absence of 02 really came
from the reaction with 02 produced in the reaction of 0(3P) with N02.
Preliminary results from our laboratory (18) show that the production
of CF20 has a quantum yield of 1.0, in agreement with Tyerman (50) who
reported that no CF2 was released from the reaction.
Mechanism;
The results of the various studies are summarized in Table V. Some
general comments are:
-------
41
1) There are three types of reactions which are most typified by the three
molecules C2Fi,, GaCls,, and C2Hi», For CjFs, the main result of Q(3P) attack
is double bond cleavage, reaction 38a; for CjClj,, the principal reaction
path is 38c to form CClzCClaO* which then polymerizes all the time; for
CiH^, the excited intermediate, CH2CH20 , is also formed, but it rearranges
to give CHaCHO or free-radical fragments. There is also some evidence that
these products are formed directly via reaction paths 38d and 38e. Mono
free radicals or the 2-carbon acid halide are never formed with C2Fi» or
Cad*.
For the f luoroethylenes , CHFCF2, CHFCHF, CH2CF2 , CHFCH2, CF2CFC1, and
CFjCFBr, the data are not quantitative. However no polymer, epoxide, 2-
caxbon carbonyl product, or products expected from mono-free-radicals were
found. Consequently we can assume that the principal reaction path is by
carbon-carbon double-bond cleavage, either reaction 38a or 38b.
C2Fi, and CFC1CFC1 react with 0(3P) primarily by the double-bond
cleavage reaction 38a, but some excited intermediate is produced by channel
38c. CFaCClg apparently reacts entirely by channel 38c. For CaFi, and
CFj CCli , the excited intermediate CX2CX20* always reacts with the parent
oleifin to give a short-chain polymerization (chain lengths < 10) .
With all the chloroolef ins the C=C double-bond cleavage paths ,
reactions 38a and 38b occur 19-31% of the time. The dominant path is
reaction 38c to produce CClXCXaO*. This molecule leads entirely to poly-
meiizerization without involving the parent olefin with CaCli, and CHC1CC12
CC1XCX20* -*• Polymer 41
CC1XCX20* lead« almost entirely to polymer for the CH2CC12 and cis- and
trans -CHC1CHC1 systems, either by reaction 41 or by incorporation of the
parent molecule
-------
42
CC1XCX20* + CC1XCX* •> Polymer 42
CClXCXjO* leads primarily to rearrangement for the CHClCHs and CzH^ systems,
Thus the 2-carbon carbonyl products and the products of mono-free-radical
reactions are observed. In CzH-k some ep oxide is also observed, but this
is the only ethylene for which the epoxide was ever reported,
2) It has been seen that except for C2Hit, no epoxide has ever been found.
Furthermore, except for CzHt,, CH2CHBr, .CHClCHz , to a minor extent CCl2CH2s
and cis- and trans -CHC1CHC1, no free-radical or rearrangement products were
found. This suggests that with the exception of CaH^, CH2CHBr, and CHC1CH2 ,
the excited CXaCXiO* intermediate has a a type bond with the oxygen atom
localized on one of the carbon atoms, presumably the one at the positive
end of the molecule,
\
Cl
,X
/
c— o
\
X
Thus this molecule has diradical character (from the spin conservation
rules, it should be a triplet) and reacts easily with the parent olefin
or with itself. Presumably for CaH^ and CHC1CH2 , the oxygen atom is more
centrally located, as has been postulated by Cvetanovie (39).
~ H 0 ,H~
\/\/
c-c
_/ y
A
H 0 R~
\/\/
\rt Vi
L.— L.
y v
Thus for these molecules probably the excited intermediates are the triplet
states of the corresponding epoxides.
3) For the unsymmetrical chloroethylenes, the 0(3P) atoms always pre-
ferentially attack the less chlorinated carbon atom. The same effect was
seen with chlorine atom attacks and the reason must be steric, rather than
due to inductive or mesomeric effects.
-------
43
In fluoroethylenes, since two carbonyl products are reported, both sides
art', attacked. However Mitchell and Simons (47) reported that in CF2CH2
the production of CFj was very low and that the main product was CO
(probably from CHaO* -> CO). Maybe, in this case the strong mesomeric
effect of the fluorine in the molecule is the explanation
6+
FN g H
C— */' + 0(3P) (electrophilic species)
/ \H
Moss (49) reported that CO is also produced in large amounts in the
reaction of CH2CHF with 0(3P). In the CF2CFBr-0(3P) system the main pro-
duct was CF20 (47) in agreement with the idea that steric effects dominate
in the addition. Haszeldine and Steele (51) concluded that atom or free-
radical attack on CFiCFCl occurs exclusively at the CF2 group.
Johari et al (52) in their paper on addition of CFa to chloroolefins
concluded that, "If the rate of attack at the «CF2 end of difluoroethylene
is assumed to be approximately the same as that for addition to the — CF2
end of chloro-2,2-difluoroethylene then the rate of attack at a CHC1 group
is estimated to be 103 to 10"* times slower than attack at the "GHz group."
Rate Coefficient;
The rate coefficient for many ethylenes have been measured at room
temperature. When the 0(3P) atoms are produced from Hg photosensitization
of N20, it is most convenient to measure the competition between two olefins
for the 0(3P) atom. From the variation in the product yields with relative
pressure, the relative rate coefficient can be obtained. If the rate
coefficient for one olefin is known, the other can be computed.
Rule et al (48) measured their rate coefficients using a discharge-flow
system coupled to a mass spectrometer by monitoring the decay of the olefin.
-------
44
Table VI •
Rate coefficient for the reaction, of atomic oxygen with
<3j
haloetiiylenes at room temperature
Olefin
CH2CHF
CH2CF2
cis-CHFCHF
trans-CHFCHF
CHFCHF
CHFGF2
CF2CF2
CH2CHC1
CH2CC12
eis-CHClCHCl
trans-CHClCHCl
CHC1CC12
CC12CC12
CF2CFC1
CF2CC12
CFC1CFC1
CH2CHBr
Source of
k/k{C^Hi») 0(3P) atom
0.51
0.38
0.43
0.22
0,32
0.54
0,52
0.57
1.0
1,0
1.6
1.0
1.0
0.47
0.11
0,10
0.10
0.51
0.67
0.20
0.78
1.0
microwave discharge
N20 + Hg*
microwave discharge
N20 + Hg*
N20 + Hg*
N20 + Hg*
microwave discharge
N20 + Hg*
N20 + Hg*
N02 + hv
N20 + Hg*
microwave discharge
N20 + Hg*
N20 + Hg*
N20 + Hg*
N20 + Hg*
N20 + Hg*
N02 + hv
N02 + hv
N20 + Hg*
crossed beams
microwave discharge
Reference
Huie et al (48)
Moss (49)
Huie et al (48)
Moss (49)
Moss (49)
Moss (49)
Huie et al (48)
Moss (49)
S aunders and He i cklen(46)
Tyerman (50)
Moss (49)
Huie et al (48)
Sanhueza and Heicklsn(38)
Sanhueza and Heicklen(7)
Sanhueza and Heicklen(7)
Sanhueza and Heicklen(37)
Sanhueza and Heicklen (36)
Tyerman (50)
Tyerman (50)
Sanhueza and Heicklen (9)
Slagle et al (45)
Huie et al (48)
) k{C2H%} = (4,0 ± 0.1) x
"1 at 25°G (53-59).
-------
45
In the technique used by Tyerman (50), ground state CF2 was monitored
by kinetic spectroscopy after the long wavelength photolysis of N02 (to
produce 0(3P)) in the presence of CzV>* and a competitive olefin diluted in
N2. The CFg was produced in the CK^-CaF^ reaction, and its diminution
in the competitive system gave a measure of the relative rate coefficient.
The results of the different studies are listed in Table VI. Rates
relative to Gall^ are reported. For CzB^ the room-temperature rate
coefficient is 4.0 ± 0.1 x 108 M""1 sec'1 (53-59). The rate coefficients
for CzHi,, CjFij, CHgCHCl, and CHjCCla are equal to each other and greater
than the rate coefficients for the other substituted ethylenes. The
partially fluorinated ethylenes have rate coefficient 1/3 - 2/3 that of
CaHij. The presence of chlorine on both carbon atoms (except for cis-
CHC1CHC1) drops the rate to about 0.1 that for C^Ki,.
Moss (49) pointed out that, "In considering the reactivities of atoms
or free radicals, it is usual to seek correlation with observed or calculated
properties of the reactant molecules. Successful correlation often provides
useful indication of the nature of the radical reactants and the main factor
controlling reactivity. The rates of reaction of 0(3P) with hydrocarbon
olefin correlates well with excitation energies and ionization potential of
the olefin (39). Since these properties show the ease with which an electron
may be removed or promoted from the ir-orbital of the ground-state molecule,
electrophilic behaviour of 0(3P) isindicated." These observations, and
others, have led, Cvetanovi£ to suggest that the transition state for the
reac.tion is a Tr-eomplex with the oxygen atom placed approximately centrally
between the carbon atoms forming the double bond.
Moss (49) measured the relative rate constant for the reaction of
oxygen atoms with the fluorinated ethylenes. The results (in Table VI)
were compared with data for other atoms and radicals with the same olefins,
-------
46
and briefly discussed in terms of the electronic changes produced in the
double bond by fluorine substitution. The 0(3P) reactivities showed no
correlation with the ionization potential.
The reactivities of the chlorinated ethylenes show a correlation
between the reactivities with 0(3P) and the ionization potentials. The
rate of reaction decreases (more chlorinated) as the ionization potential
decreases. However this correlation is in the opposite direction of that
if 0(3P) is an electrophilic species„ The results for oxygen atoms are
compared in Table VII with results for other atoms and radicals adding to
chloro- and chlorof luoroethylenes. Always C2Clif is the least reactive and
in a general way the inclusion of chlorine in the olefinic molecule decreases
the rate.
cis-CClHCClH reacts faster than the trans isomer in the 0(3P) reaction.
The significant difference for the rate coefficients for the two isomers
presumably reflects steric factor differences.
It is interesting to note that, for the reaction of 0(3P) with CHFCHF
and CvH8-2, the rate coefficient is also larger for the trans compounds
than for the cis compounds by respective factors of 1.7 (49) and 1,6 (60). .
However the reactions of CH30 listed in Table VII show that the cis isomer
reacts faster with CH30 than does the trans isomer.
REACTION WITH 0(3P) IN THE PRESENCE OF 02
The oxidation of the haLogenated ethylenes by 0(3P) atoms in the
presence of QI may proceed by three different routes:
1) A chain mechanism initiated by the oxidation of the substituted methylene,
CXj. This process is important for C2Cli,, CHCICCI^, CH2CCl£, cis- and trans-
CFC1CHC1, CFC1CFC1, and C^.
-------
Table VII
Relative reactivities of chloro- and chlorofluoroethylenes with atoms and radicals in the gas phase at room temperature'
Ethylene
0(3P)
CClj
CC12C12
CC12CHC1
CC12CH2
cis-CClHCClH
trans-CClHCClH
CC1HCH2
CH2CH2
CF2CP2
CF2CFC1
CF2CC12
CFC1CFC1
0.10
0,10
1.0
0.11
0.47
1.0
1.0*
1.0
0.51
0.67
0.20
<0.003
0.16
~\
\
V0.05
J
l.O8
—
—
—
0.06
—
ci
Chloroethylenes
0.18
0.40
1.28
0.95
1.0J
Chlorofluoroe thylenes
2.22
0.30s
0.79
0.72
1.17
0.90
Hg 6(3P)
1.7
2.3*
1.0-
0.20*
lonization ,
Potential, eV
9.34
9.48
9.83
9.65
9.64
10.00
10.66
10,. 11
9.84
9.65
a) A very complete table is given by Moss (49) for the fluoroethylenes.
b) For References see Table ¥.
c) Reference 52 at 150°C.
d) Reference 61.
e) Reference 63.
f) Reference 62.
g) Reference 23.
h) References 27, 29.
i) Reference 29.
j) Relative reactivity set at 1.0.
k) References 1, 29.
-------
48
2) The oxidation of the CX-jCXzO* intermediate. This process is Important
for CjF-., CF:;CClr, arid CHC1CH;-. >
3) The oxidation, of the mono-free-radical fragments, a process of Importance
in C^H. and CHC1CH2.
M&rhygiene OKI da tlon:
Dependent on the parent ethyi&ne, the 0(JP) atom can react with it to
produce any of the following methylenes; CC12, CH^, CFjj CC1H, CC1F, or
CFH. The spin conservation rules predict that these methylenes will be
produced in their triplet states and thus be reactive with Oj. This is to
be contrasted for the singlet earbene species, which have been shown to be
unteactive with 02 at room temperature for CHa (64), CdU (65), CFC1 (65),
and CFa (1) (singlet CFa reacts with 0^ at elevated temperatures to give
CF20 + 0(3P)).
All the evidence suggests that when CXz species are produced in the
0(3P)-CXaCX2 reaction, they are produced exclusively in the triplet state.
However the triplet methylenes react with 0? by three different routes
depending on the methylene involved.
1) The triplet species CCl* (36,37), CC1H (7) and CC1F (9) react with 02
as follows
3CC1X + Oi •* XO + C1CO 46
and the C1CO species can rapidly fall apart
C1CO •* Cl + CO
2) The triplet CF? adds to 02 (1)
3CF> + Oi- -> CFiOj 24
3) The triplet CH^ gives (43)
3CH2. + Oi ••» HCOOH 47a
HZ0 -f CO 47b
-------
49
Thus in the first case mono-free-radieal& are produced; in the second case,
dl radicals; and in the third case, stable products , The detailed fate of
triplet CFH with 02 is unknown, but Gordon and Lin (66) observed HF laser
emission from the reaction of CHF with 02, They attributed this product
to the formation of excited FCOOH which decomposes to give HF* + COa .
Thus, at least pare of the time, CHF oxidizes analogously to CH2 to give
molecular products directly,
The difference in the three reactions is probably energetics. In all
likelihood in all three cases the adduct CX402 is formed first. With CF^Oz
any rearrangement is endothermie and does not occur. For the other species
presumably they rearrange to XC^L, which decomposes to XO.+ XCO* Only in
the case of HCX..,. does stabilization occur. However the HCftL initially
^
fornfid on rearrangement contains excess energy, and apparently this energy
is sufficient for reaction 47b to proceed if the molecule is not deenergized,
Oxidation of CX;GX;.0*;
The reaction of CXjCX^O* with 0^ can proceed in two ways. The route
which prevails with CFiCF^O* is:
Of •*• CP.,CF,,0* * CFj-Oi •*• CF;.0
On ' the other haad, with CCl-jCF^O*, the process is
0,- •+ CCiJJP.O* -»• CF?0 f CiCO -h Ci
CHClCHiO* apparently can react by either route to produce the diradical or
monoradlcal products, respectively. For the other ethylenes, the oxidation
of 'CHClCHfcO* has not been elucidated, since it appears to be an unimportant
Individual Mo lecules ;
C^Cl, : A long-chain process is Involved which produces the same products as
in the chlorine-atom initiated reaction, and the ratio of CClsCCIO produced
-------
50
to CC120 produced is 2,0 at 25°C (36), similar to the ratio of 2.5 found in
the chlorine atom system. However the rate law is different, the quantum
yield of chlorinated product formation being proportional to [C^Cl^J/Ia ' «
In addition CO is formed with a quantum yield of 0.18, independent of con-
ditions. This value is identical to the CCl^O yield in the 0(3P)-C^Clit
system in the absence of 0(,
CHCICCI.?; FOE CHClCClj. the free-radical long-chain oxidation is observed,
as in the case of C ]. ${CF20} was
equal to 1.0 invariant to the reaction parameters (9). These results
-------
51
were interpreted by a mechanism analogous to that found for C2Fi» (1):
CCliCF20* 4- CF2eCl2 -»• CF20 4- 6ci7cF2fcci2 48
CC12CF20* + Oz -> CP20 + CIO + C1CO 49
Reaction 49 becomes the Initiating step for the chain reaction, and the
mechanism predicts *{CF2C1CC1(0)}/§{CO} in this system should equal
fCCFjClCClCO)} in the chlorine-atom initiated oxidation at high pressures.
The former quantity varies between 28 and 45, and the latter quantity is
about 45, so that the agreement is not too bad.
Reactions 48 and 49 must be simplifications of a much more complex
process since they predict ${CO} <= 1.0 and ^ICClCFitcij} «• 0 at low values
of [CF2eCl2]/[02], and *{CO} - 0 and *{eei2CF26ci2} - 1.0 at high values of
[CF2CCl2]/[02], contrary to the observations. Possibly CC12CF20* represents
several isomeric species, one of which always goes by reaction 48, one of
which always goes by reaction 49, and one or more which can proceed by either
route.
CzFfe" The reaction of oxygen atoms with CaFi, in the presence of Oz was
studied briefly by Saunders and Heicklen (46) at room temperature and in
more detail at 23 and 125°C by Heicklen and Knight (67). In addition to
CF?0 and c~C3F& (found in the absence of Og) the products included tetra-
f luoroethylene oxide (CF2CF20) . The results were reviewed and discussed
elsewhere (1). The results were explained by a • bi radical mechanism which
for the methylene is
-'CF2 + 0? -»• CF202 24
-»• 2CF20 + 3CFj 23a
-*• CF20 -I- CF2eF20 23b
2CF202 -»• 2CF20 + 02 50
and for the excited molecule mechanism is
C2F%0* + CaF^ •». c-C3F^ -I- CF20 42'
-------
52
C.F.O* 4 0* -* CFjO H- CFiOs 51
Vinyl chloride is unique among the chloroolefins and does not
oxidize like any of the higher homologs. Its oxidation follows more
nearly the pattern of C^F*, except that there is no chain in the 0(3P)-02-
CHC1CH? system; the products are CHC10, CO, HC1, and HCOOH (8).
Also a very surprising result occurs, namely no Ca carbonyl compounds
are produced. The Oj must intercept the intermediate in a scheme such as
CiH3C10* + 02 •* CHC10* + CHjOj 51a!
•*• HC1 + CO + CHj02 51b '
The CHiOi can either rearrange to HCOOH or decompose to CO + HzO
CH2Gj> -»• HCOOH 52a
-* CO 4 H20 52b
From the data it was difficult to assess the relative importance of the
products observed. However a reasonable designation for the initial quantum
yields was
${CO) 'v 0,6
With this assessment, k= a'/k^- ' '• 0.6 -and kgaa/^sa ^ 0»8,
There are two general race laws:
1) The dlradical mechanism which involves the OKidatlon of CXa and
CXj.CX^O*. If there is a chain (C2F»., CF^CCla) the chain lengths are
dependent on the ratio [CXgCXaJ/fOz ] , The rate law has been discussed in
detail in C^F^ elsewhere (1) .
2) The ntonoradical chain mechanism which appJlles to all the chloroolefins
studied, except CFaCClj and CHC1CH2 . The chain lengths depend on the para-
meter [CClXCXyJ/Ia'1/2 when the 0(3P) atom is generated in steady-state
-------
53
photolysis. Since the rate of the chain propagation step is proportional
to[CClXCXi], termination by a radical-radical mechanism is suggested in
which one radical is the chain carrier (i.e. Cl atoms) and the other radi-
cal must be one that is absent in the chlorine-atom initiated or Hg-
phocosensitized oxidations, since in those systems there is no intensity
dependence. The indicated reactions are
Cl + XCO -» C1X + CO 53
CIO + C1CO -* C120 + CO 54a
•* C12 + CO? 54b
The initiating reactions are;
3CC12 + Q2 + CIO 4- C1CO 46'
}CC1H + 0? + HO + C1CO 46"
3CC1P + 02 -* FO + C1CO 46" '
and the C1CO can decompose via
C1CO -> Cl -I- CO
The CIO and HO radicals react rapidly with the olefins to initiate the
chain. However the FO radical is apparently a terminating.radical, since
in the CFC1CFC1 system, the chain length was only one-half that expected
if FO propagated the chain.
If the termination is principally by reactions 53 and 54, which is
the case at low values of [CClXCXs]/la'/2, then the oxidation chain length
will be proportional to [CC1XCX2]/Ia1/2. On the other hand at high values
of [CClXCXzJ/Ia1/*, termination is principally by reactions 4b and 4b'.
The chain length should be independent of the reaction parameters, and
should be equal to that in the chlorine-atom initiated system multiplied
by the yield of CXC1 radicals produced in the primary step when 0(3P) reacts
with CC1XCX2 (1/2 that value for CFC1CFC1, since FO is not a propagating
radical). Thus
-------
54
*{OX)ao/*{OX}Ci - (k38a + k3sb)A38 - XII
where ${OX} is the upper limiting oxidation quantum yield at high [CClXCXg]/
la '•' '* in the 0(3P)-Oa-CClXCXi, system, and ${OX}C1 is the oxidation yield in
the chlorine-atom initiated oxidation. The right-hand side of eqn. XII also
can be obtained independently from the 0(3P)-CC1XCX2 system in the absence
of 0?. Thus eqn. XII relates, in one expression, the principal features of
the chlorine-atom initiated oxidation, the 0(3P) oxidation, and 0(3P)-Ot-
CXiClCX* oxidation.
Table VIII summarizes the results obtained for the upper limit long-
chain oxidation in the 0(3P)™02-CC1XCX2 system. The values of ^{OX}^/
${OX}QJ agree quite well with the values of (kjsa + k38b)/k38 obtained in
the absence of 0^..
REACTIONS WITH OZONE
Surprisingly, in spite of their commercial importance and possible
biological significance of some of the haloethylenes, relatively limited
kinetic studies have been carried out on the. ozonolysls reactions of these
compounds. The earliest work appears to be as recent as 1966 (68) and,
apparently, subsequent studies have originated only from Cvetanovic and
coworkers at the Canadian National Research Coxmcil (69) and from our
laboratory (5, 70, 71). The former group reported on the kinetics of
ozonolysis of various chloroethylenes in CC1H solution while our Investiga-
tions have dealt with the gas phase and the low temperature solid phase
reactions of several of the same chloroethylenes. Although some of our
studies have not yet been published, the lap or tan t conclusions resulting
from them will be reviewed here. Also, for the sake of completeness, the
ozono lysis data on ethylene Itself will be included in this review.
-------
Table VIII
the Reaction of chloroethylenes with 0(*P) in the presence of
[01efin]/IaI/2,
Olefin (Torr-sec)1/2
CCljCCl, 9.9 - 175
CCljCiCl 48 - 1150
CCljCHj 94 - 2000
cis-CHClCHCl 36.5 - 687
trans-CHClCHCl 32.5 - 638
CFC1CFC1 40.5 - 684
a) *{OX}Cl from Table I.
b) Values obtained when Of was absent
c) Has never reached under the actual
d) CO vas also produced in the chain.
e) 4>{CO} - 0.35 in the absence of 02.
[01efin]/Ial/!
Dependence *frnl
linear 0.18
less than linear d
less than linear 0.78
almost none d
almost none d
less than linear 0.80
(fable V).
e*perinental condition.
c --
77.0 0.38
55.0 0.32
7.0 0.32
7,0 0.32
160 0.76
(0.38 x 2)
k«a + k«*b
f-Mt
0.19
0.23
0.35
0.23
0.28
0.80
Reference
Sanhueza and Heicklen (36)
Sanhueza and Heicklen (37)
Sanhueza and Heicklen (38)
Sanhueza and Heicklen (7)
Sanhueza and Heicklen (7)
Sanhueza and Heicklen (9)
Ui
Ui
-------
56
Review of the Experimental Data
CgHi,; The reaction of ethylene with ozone has been studied in the
vapor phase under chemiluminescent conditions by Finlayson, et al (72, 73),
and under non-chemiluminescent conditions by several research groups (74-83).
Experimental data for the liquid phase reaction originate from the labora-
tories of Cvetanovic (69, 84) and of Kuczkowski (85-88). We have reported
previously on the reaction carried out in the solid phase at low temperatures
(88) and on the vapor phase decomposition of one of the relatively stable
reaction products obtained from the liquid and solid phase reactions (89).
In their study of the chemiluminescent reaction, Finlayson
et al, (72, 73) employed a flow system in which excess olefin reacted with
ozone (about 2 mole percent diluted in 02, N2, or He). The total pressures
were 2-10 Torr. Emission was seen from vibration-rotation bands of HO with
v<9. The emission was virtually identical to the Meinel bands seen from
the reaction of H with 03
H + 03 * 02 + HO* (v<9) 55
thus confirming that H atoms are produced under the experimental conditions
used. The emission yield was ^<10~? for the 9 ->• 3 transition per molecule
of reactant consumed at 4.6 Torr total pressure. Also seen was emission
from electronically excited CH20 ('A" ->• 'A.) and OH (A2Z ->• X2II), the yield
of the former emission being 'v 10" ; per molecule of reactant consumed.
The electronically excited HO emission was seen only in N2-buffered mixtures,
but the other two emissions were seen in either N2 or Oa buffered mixtures.
With the assumption of 1:1 reactant stoichiometry, the rate coefficient was
found to be 5 times larger in N2 than in 02 at 2-10 Torr total pressure and
reactant fractions of 03 "t 50 ppm and CaH* ^ 400 ppm. In 02 the rate
coefficient at room temperature was (1 ± 1) x 103 M"1 sec"1.
-------
57
Cadle and Schadt (74) appear to have been the first to study quantita-
tively the kinetics of the ethylene-ozone reaction. Infrared spectroscopy
was used to follow the decay of ozone, and the reactant pressures were in
the ::ange of 0,1 to 3 Torr, The consumption ratio [CgH^J/jOs] was reported
to vary between 1.9 and 3.2 and the products were not identified. The
initial rates, which were first order in each reactant, gave a second order
rate constant of 2.1 x 103 M~l sec"1. Evidently, the rate showed no depend-
ence on oxygen pressure (150 to 650 Torr) or on temperature (30° to 50°C) .
The second order kinetics was subsequently confirmed by Hanst et al
(75), by Bufalini and Altshuller (79), by DeMore (80), and by others (82, 83).
Bufalini and Altshuller (79) used in their work a 12-liter static reactor
kept at 25°C and under a dynamic condition a variable volume vessel (0.5 to
12 liter) with temperature kept between 30° and 100°C. Reactant concentra-
tions were in the parts per million range and air was used as diluent.
Ethylane was analyzed by gas chromatography while the iodide titration
method was used for ozone. Complete stoichiometry was not reported but the
consumption ratio [C2Hi,]/[03] was found to be near unity at low ethylene
concentrations and to increase to a limiting value of about 1.6 as the olefin
pressure was increased. Bufalini and Altshuller reported the experimental
Arrhenius frequency factor and activation energy to be 1.7 x 106 M"1 sec"1
and 4,2 ± 0.4 kcal/mole, respectively. At 256C, the latter parameters
correspond to a rate constant of 1.6 x 103 JM"1 sec"1.
Similar Arrhenius parameters were reported by DeMore (80) although his
reaction temperatures were in the range from -40° to -95°C. The rates in
this temperature range were still independent of the presence of oxygen,
and the consumption ratio [CgHi^/fOs] was 1.0 ± 0.3 in the absence of Og
and 1.2 ± 0.3 with Q2. DeMore also observed aerosol formation which was
reduce! by not using any diluent gas. However, infrared analysis apparently
-------
58
provided no information concerning the nature of this aerosol or of any
other reaction products„
More recently S.tedman et al (82) and Herron and Huie (83) have examined
the ozonolysis of ethylene at low reactaur. pressures, In the work of Stedman
et al., the reactant concentrations were in the parts per million range and
the total pressures were kept at one atmosphere. Only a single temperature
of 26 ± 2°C was used in this work, and the second-order rate constant was
found to be 0.93 x 103 M~J sec"' in either Q-? or N2 diluent. Herron and
Huie followed the reaction by mass spectroscopy In the temperature range of
-40 to 90°C. Ethylene pressures were below one Torr but kept about ten
times greater than the ozone pressuresc These authors observed with argon
carrier gas that nonreproducible results were obtained and the apparent
second-order rate constants were much greater than those obtained with
02 buffer gas. With oxygen at about 3 Torr, the resulting second-order
rate constants and their Arrhenius parameters were in close agreement with
values obtained by other investigators.
The difference in rate coefficient and mechanism in the presence of 02
found by Herron and Huie (83) and Finlayson et al (73) confirmed the earlier
report of Wei and Cvetanovic (78) who found that the ratio of olefin to
ozone consumed is unity in the absence of 0;> but between 1.4 and 2.0 in
its presence. Furthermore the relative rate coefficient (compared to the
i-C^He-Oa reaction) was different in the Oa and N? buffered systems (78).
Herron and Huie (83) also studied the ozonolysis of propylene and found
that its apparent second-order rate constant decreased by a factor of almost
two as the 62 pressure was increased to about one Torr. At higher Og
buffer gas pressures, the second-order rate constants :*£rnained constant
and at a value of 6.36 x 103 .H"1 sec"1 (25°C) which agreed with those
reported by earlier workers.
-------
59
Summary of the kinetic data for the ozonolysis of ethylene is presented
in Table IX. Experimental results obtained by Cvetanovie and coworkers
(76-78) are not included here since only relative rates of ethylene with
respect to other olefins were obtained. However, on the basis of analysis
by gs,s-liquid chromatography, Vrbaski and Cvetanovic (77) found that one
mole each of C2Hi( and 03 gave 0.25 mole of HCOOH, 0.019 mole of CH3CH05
and small amounts of other unknown products.
; The only attempt of a quantitative kinetic study of the ethylene-
ozone reaction irt the liquid phase was that by Williamson and Cvetanovic (84)
Carbon tetrachloride solution was used by these investigators, but due to
loss of olefin from the solution the kinetic results were inconclusive.
However, by assundng that the relative rates with respect to 1-hexene were
the same in the vapor and CClij solution, Williamson and Cvetanovic (69)
estimated the second-order rate constant for the ethylene-ozone reaction
in CC:U solution at 25°C to be about 2,4 x 101* M."1 sec"1.
Other liquid phase studies reported in the literature appear to deal
primarily with product identification for mechanistic purposes. Inert
solvents and reduced temperatures have been used in these studies in order
to minimize the decomposition of the reaction intermediates or products.
Under these experimental conditions, some higher molecular weight peroxides
are obtained but the major reaction product is the 1,2,4—trioxacyclopentane
(commonly called secondary ethylene ozonide or simply ethylene ozonide).
The infrared spectra of ethylene ozonide in the vapor phase at 30°C and in
the solid phase at liquid nitrogen temperature are shown in Figure 4. Our
vapor phase spectrum is essentially the same as that reported first by
Garvic, and Schubert (90), Band frequencies and their tentative assignments
are given in Table X. The complete microwave structure of ethylene ozonide
-------
Table IX
Kinetics of ethylene-ozone reaction in the presence of excess
Inves ti gators
Cadle and Schadt (74)
Bufalini
and Altshuler (79)
DeMore (80)
Stedman et al (82)
Species Temperature Reactants [CzH^/fOs] Arrhenius
Followed"* (°C) (Torr) Consumption E(kcal/mole)
03(IR) 30 to 50 0.1 to 3 2 to 3 0
03(KI)
C2H,.(GC) 30 to 100 ppm range 1 to 1.6 4,2 ± 0.4
03(UV) -95 to -40 2 to 20 J;° ± ^^ ^ ± Q^
03(NO) ,,
C2H,(GC) 26 PPm range
Parameters k at 25 °C
A(^~" sec" ) LM~ see" )
2.1 x 10 3 2.1 x 10 3
1.7 x 106 1.6 x 103
2,0 x 1Q6 0,79 x 10 3
0.93 x 103
Herron
and Huie (83)
Finlayson et al (73)
03(MS)
-40 to 90 ppm to 1
25
^0.5 x 10-3(03)
^4 x 1Q-3(C2H4)
5.1 ± 0,3 5.4 x 106 1,02 x 103
i n ^ tn3
-«_ • -w ^fc -t. v
a) IR = by infrared spectroscopy, KI = by KI titration, GC = by gas chromatography, UV = by ultraviolet
spectroscopy, NO = by nitric oxide chemiluminescence, MS = by mass spectroscopy.
-------
100
80
60
40
20
J I
4000 3000 1800 1600 1400 1200 1000 800 600 400
Frequency, cm""1
Figure 4: Infrared spectra of primary and secondary ethylene ozonides at liquid nitrogen
temperature. In part from Hull et al (88).
-------
62
Table X
Infrared spectra of primary and secondary ethylene ozonide0
Secondary Ozonide
Vapor (30°C) Solid (-190°C)
Primary Ozonide
Solid (-190°C)
2996 w
2974 s
2900 s
^1380 w
VL350 w
1260 w
957 vs
^933 m
798 m
698 w
^400 w
3050 w
2980 m
2910 s
2894 m
1646 w
^1480 w
1395 w
1350 m
^1207 w
1133 m
1082 vs
^1038 m
1212 m
1130 m
1060 vs
1020 s
932 vs
917 m
804 s
733 w
696 in
405 w
1390 w
1325 w
1214 w
1125 w
983 m
927 m
843 w
730 m
687 w
650 m
410 w
Tentative
Assignment
CH2 stretch
CH2 stretch
CH2 stretch
CH2 stretch
combination
CH2 deformation
CH2 deformation
CH2. deformation
CH2 twist
CH2 twist
difference band
CC stretch
CO stretch
CH2 wag
CH2 wag
CO stretch
CO stretch
00 stretch
CO stretch
CO stretch
00 stretch
00 stretch
ring bend
ring bend
ring bend
CH2 rock
CH2 rock
ring bend
ring bend
ring bend
a) Frequencies are in cm""1 unit.
b) A dash indicates overlap with the secondary ozonide.
-------
63
has been determined by Gillies and Kuczkowski (85, 86). The molecule has a
half-chair conformation (Cj point group) with the geometry as shown in
Figure 5, but no evidence of free or hindered ring pseudo-rotation was found.
This ozonide also has a dipole moment of 1.09 debye, and on the basis of
temperature dependence of its microwave line intensities a low frequency
fundamental at 2:00 ± 40 cm"1 was predicted by Gillies and Kuczkowski as a
possible ring bending vibrational mode. These investigators also carried
out the low temperature liquid phase ozonolysis reaction in the presence of
formaldehyde-180 and showed that the oxygen isotope appears exclusively in
the epoxy position of the ozonide (85-87). In addition, when ethylene-di
was used in the reaction, ethylene ozonide-do and two d^-species, all in
about; equal amounts, and smaller quantities of three types of da-species
were identified by microwave spectroscopy. Above room temperature, gaseous
ethylene ozonide decomposes slowly by a first-order process giving quanti-
tatively formaldehyde and formic acid as products (89). The first-order
rate constant has been determined in the temperature range of 46° to 85°C
to be k(sec~3) = 10!3-60 exp(-27.5 kcal/mole/RT) (89).
The reaction of ethylene with ozone in the solid phase has been studied
by Hull et al. (88), with infrared spectroscopy. As the reactants were
warmed from liquid nitrogen temperature to about -170°C, a new set of
infrared absorption bands appeared indicating the formation of one major
primary product. On further warming to temperatures near -100°C, this pri-
mary -product decayed into the known secondary ethylene ozonide. Vaporization
of tha reaction mixture and its spectroscopic analysis showed that the
secondary ethylene ozonide was the final major product with formaldehyde
and formic acid being minor products. Also, small amounts of polymeric
material remained on the low temperature infrared window. From the
-------
n
Figure 55 Microwave structure of secondary ethylene ozonide. From data of Gillies and Kuczkowski
(85, 86).
-------
65
frequencies of the absorption bands of the primary product and from the fact
that similar sets of bands were displayed by the initial products in reactions
of other olefins with ozone, the primary species of the ethylene-ozone solid
state reaction at low temperatures was identified as the 1,2,3-trioxacyclo-
pentane (primary ethylene ozonide) . The infrared spectrum of a 'solid sample
containing both the primary and the secondary ethylene ozonide is shown in
Fig ire 4. Frequencies of the primary ozonide bands are listed in Table X.
The most characteristic band in the spectrum of this ozonide at liquid
nitrogen temperature is the intense sharp band at 983 cm"1 which does not
overlap with bands of other species present in the reaction mixture.
CgFif- According to Heicklen C68) and later confirmed by
Gozr.o and Camaggi (91), only carbonyl fluoride and oxygen are observed as
proc.ucts of the vapor phase ozonolysis of tetrafluoroethylene at room
temperature. Two moles of carbonyl fluoride were obtained from each mole
of clefin, so the reaction stoichiometry is evidently
2C2F4 + 20 3 -»• 4CF20 + 02 56
The kinetics of reaction 56 at'25°C was studied by Heicklen for ozone and
olefin pressures in the range of 0.7 to 15 Torr and 0,2 to 6 Torr, respectively,
Initial rates, RiTCFaO}, were determined by following the infrared carbonyl
band of CF?0. At constant C2P^ pressure, RilCPaO} increased linearly with
increasing ozone pressure but became independent or even decreased at higher
pressures of ozone. The experimental data although limited were Interpreted
on the basis of the rate equation
k fcTOa
YTTT
, mi
for which Heicklen obtained at 25°C, k = 300 M"1 sec"1 and k1 > 9 x 10% M""1 .
The investigation of Gozzo and Catnaggi (91) was concerned primarily with
the reaction stoichiometry and product identification. Their vapor phase
work appeared to be limited to the confirmation of equation XIII and most
-------
66
of their studies were conducted with solutions- of inert halocarbon solvents
at 0°C. They employed a flow system with ozone in helium carrier gas and
determined the reaetant consumption and product formation in ndllimole/
hour. When the olefin was' in excess (reaetant ratio [C2Fi*]/[03] *v 50), the
major products were carbonyl fluoride and tetrafluoroethylene epoxlde.
Traces of parfluorocyclopropane also appeared but the reaction stoichlometry
was best represented by
2C2Fd, + 03 -* 2CF20 +
-------
67
before in many alkene and ozone solid phase reactions (88). No ozonides
were formed in the liquid phase either. Vaporization of the liquid phase
products gave CCl^O and CC13CC10 as major products, CClaCClaO as a minor
product, and traces of HCOOH and a high boiling polymer. The forma-
tion of the HCOOH was minimized by keeping our reaction vessel dry. The
stoichiometry of the liquid phase reaction could not be determined because
of poor oxygen mass balance. However, the olefin consumption appeared to be
accounted for by the CClaO and CClaCCIO yields, and more phosgene than acid
chloride was always produced.
The products of the gas phase reaction at 25°C were essentially the
same as those observed from the liquid phase study. However, traces of Cla,
COS and CQz were observed when the reaction was permitted to continue for
long periods. Once again, it was not possible to determine the reaction
stoic'iiometry but more products appeared to be formed when oxygen was used
as the buffer gas. The reaction was too slow, ozone loss occurred through
its o\m decomposition, the acid chloride slowly decayed with time, and the
strongest infrared band of the minor product epoxide was obscured by an
olefin infrared band. Figure 6 shows the variation in composition of a
typical gas phase reaction. The reaction was strongly inhibited by oxygen.
The initial rates, Ri{CCl20} or Rl{CCl3CCl(0)}, determined by infrared
spectroscopy, increased with olefin pressures but were not affected much by
nitrogen buffer gas or the initial ozone pressure (range of a factor of two).
A log-log plot of initial rates against olefin pressures (range of a factor
of five) gave a slope of 1.8, and the average value of RifCClaOl/RiCCClsCClO}
was I..3. With 02 buffer, the initial rate was decreased by a factor of at
leas t ten.
Williamson and Cvetanovic (69) found that the reaction rate in CCli,
solution at 25°C was first order in both [Os] and [CaCli,]. The concentration
-------
68
0
0
200
400 600 800 1000
REACTION TIME (min)
1200 1400
Figure 6: Time dependence of the composition of CjCli, ozonolysis reaction at 24°C:
[C2CU]0 - 6.9 Torr, [03]0 = 4.1 Torr. From Mathias et al (5) with per-
' mission of -the National Research Council of Canada,
-------
69
of olefin which was always In excess was varied In the range of one to five
millimole/liter, and the ozone ultraviolet band at 0.280 ym was used to
follow the rate,, Under these conditions the second-order rate constant was
1.0 M"1 sec"1,
CHClgClA: Only Williamson and Cvetsriuvic (69) studied this reaction.
In CCli, solution at 25°C, the rate was reported to be first order in each
reactant with the second-order rate constant being 3.6 M"1 sec"1. The
products of the reaction or their stoichiometries are not known.
CHzCClg; The products of the vapor phase reaction at 25 °C
have been identified by Hull et al (70) to be CC120, HCOOH, CH2C1CC1(0),
CO, COa, 02, HC1, and possibly water although it was never detected. The
yield of phosgene was always either comparable to or greater than,the yield
of formic acid, and the sum of the phosgene and acid chloride yields was
generally slightly less than the consumption of the olefin. Presumably,
the hydrolysis of the acid chloride led to the latter inequality. The con-
sumption ratio [olefin]/[Oj] was approximately unity when the reactant
press ures were comparable but this ratio approached two as the olefin was
made more in excess. With 02 buffer gas, the limiting consumption ratio was
near five for high excess olefin runs. On the other hand, the yield of
CC1*0 per mole of 0.3 with N^ buffer appeared to be independent of the
relative amounts of the reactants and varied in the range of 0.3 to 0.4,
The same mole ratio increased to near unity in oxygen buffer when the
olefin was made more excess.
In CCli, solutions at 25°C» Williamson and Cvetanovic (69) found that
the yield of phosgene per reactant olefin consumed was essentially quanti-
tative (GC analysis). However, the only other product observed was a white
solid which remained after solvent evaporation and which exploded violently
on attempt to collect the material. In the low temperature solid phase
-------
70
reaction (92) on the other hand, both CC120 and CH2C1CC1(0) have been identi-
fied. Infrared analysis of the solid reactants showed initially the reversible
formation of a ir-complex, and on further warming to about ~90°C only the
infrared bands due to phosgene and acid chloride appeared. Vaporization of
the reaction mixture showed unreacted olefin, CC120» CH2C1CC1(0), and HCGOH.
When olefin was in excess} the rate of ozonolysis of CH2CC12 at 25°C in
CC1<, solution was found to consume equal amounts of reactants, to be first
order in each reactant, and to have a rate constant of 22„! M"1 sec"1 (69),
*
The kinetics is more complex in the vapor phase at the same temperature for
Hull et al (70) determined the rate law to be
-d[03]/dt = -d[CH2CCl2]/dt - kXIV[CH2CCl2][Os]2 XIV
with Ng buffer gas, and
-d[0s]/dt = kxv[CH2CCl2][03] XV
when oxygen gas was used as buffer. In these studies the ozone pressure
was a Torr or less and the olefin was varied from 3 to 100 Torr. The
experimental values of the rate constants were kxiy = (2.4 ± 0,6) x 106
M~2 sec"1 and kXy - 2.2 ± 0.6 M"? sec"1. With 02 buffer, the second-order
rate constant for equation X? obtained from the decay of an olefin infrared
band or the combined rate of appearance of phosgene and acid chloride bands
was almost a factor of two greater. Also, the value of 2.2 M"1 sec"1 was
an average of kxv values which appeared to decrease systematically by a factor
of almost two as the olefin pressure was increased from 8 to 100 Torr. In
addition, with both N2 and 02 buffers, the initial rates appeared to be
somewhat faster than the rates predicted by the rate equations XI? and XV
when the olefin pressures were low.
GHCICHCI (DCE); The stoichiometry of the gas phase reaction
between cis or trans-DCE and ozone at 23°C has been established quantita-
tively by Blume et al (71) to be as given by reaction 57.
-------
71
2CHC1CHC1 + 203 -»- 4HCC10 + 02 57
The reactant pressures were measured directly, the oxygen concentrations
were determined by gas chromatography after completion of the reaction,
and the unstable formyl chloride (93) concentrations were established
sp-ectroscopically. Since the formyl chloride was known to decompose to
HC1 and CO with a half-life of about 10-20 minutes, the products of
reaction 57 were allowed to stand until the formyl chloride infrared bands
disappeared, From the infrared determination of the absolute concentra-
tion of CO and from the known reactant pressures and initial absorbance
of HCC10, the absorption coefficients of the HCC10 infrared bands were •
calculated. The two most intense bands of this molecule are the carbonyl
stretch at 1784 cm"1 and the CC1 stretch at 739 cm"1 (93). The decadic
abisorption coefficients of the R-branches of these bands were found to
be 0.0194 and 0.0129 Torr"1 cm"1, respectively (71). In addition to
HCC10, 02, HC1S and CO, traces of CClaO and HCOOH were observed in some
of the reaction mixtures. The latter two species evidently came from
hydrolysis reactions for it was possible to minimize their formation by
careful pumping of the reaction vessels.
The stoichlometry for the ozonolysis of cis-DCE in CCli* solution at
25CC was examined by Williamson and Cvetanovic (69). They determined the
consumption ratio [DCEj/tOs] to be one but were able to identify only one
product from the gas chromatographic analysis. This product was phosgene
and 0.18 mole of it was reported to be generated from each mole of DCE
consumed. These investigators reported, however, that with 50% completion
of the reaction three other GC-peaks appeared whose retention times were
-------
72
shorter than that for CC120 and whose relative peak areas changed with time.
In the study of the liquid phase (71) formed by allowing the solid reactants
to melt at reduced temperatures, HCC10 was observed as the major product
with only traces of HCOOH and CClaO. Small amounts of explosive clear liquid
also remained after evaporation of the liquid mixture. The decomposition
of HCC10 in the liquid phase was very much faster than the gas-phase rate
but HC1 and CO were still the products. The three unidentified GC-peaks
observed by Williamson and Cvetanovic may very well have been HCC10, HCl,
and CO. The reaction in the low-temperature solid phase gave essentially
the same products (92) as those observed in the liquid phase. Only a TT-complex
and no ozonides were observed as the solid reactants were allowed to warm slowly.
At temperatures above about -150°C, absorption bands due to solid HCC10 grew.
Formyl chloride began to sublime off the low temperature window at about -110°C.
Relatively simple kinetics was observed by Williamson and Cvetanovic (69)
for the ozonolysis of DCE in CCl^ solution at 25°C. The rate was first-order
in each reactant with the second-order rate constant being 35.7 M"1 sec"1 for
cis-DCE and 591 M"1 sec"1 for the trans-isomer. Thus, in CC^ solution at
25°C the reactivity toward ozone of trans-DCE is about seventeen times faster
than that of the cis-DCE which in turn reacts about six times faster than
does 1,1-DCE. The kinetics in the gas phase, on the other hand, was expected
to be complex since reaction 57 under excess olefin condition caused the
isomerization of the reactant in addition to giving the products formyl
chloride and oxygen (93). Subsequent studies by Blume et al (71) have
shown indeed that reaction 57 has an exceedingly complex kinetics.
Blume et al used infrared and ultraviolet spectroscopy to follow the
rates of reaction 57. Olefin pressures ranged from 0.2 to 40 Torr for cis-
DCE and from 0.3 to 80 Torr for trans-DCE. Ozone pressures were limited to
below about 7 Torr. Rates were also determined with the reactants buffered
-------
73
with N;. and 0? gas. It was found experimentally that the rates of reaction
57 satisfied the condition R •= -d[DCE]/dt=-d[03]/dt = +d[HCC10]/2dt and
could be expressed in the general form
R - kxvi [DCE]n[Q3]m XVI
where n and m had values of one or two depending on the pressure range of
each reactant. When the pressures of Oj and DCE were both of the order of
one Torr or less, it was found that n = m = 2. Figure 7 illustrates a
kinetic plot of such a reaction in which [trans-DCEj = [03] = 0.62 Torr so
that L/tO-,]' plotted against reaction time gave a straight line. Also, when
one or the other reaetant was in excess, second-order kinetic -plots were
obtained by following the reactant not in excess. However, in excess
ozone kinetic runs with [Ojj greater than about 3 Torr, n = 2 was still
satisfied but the fourth-order rate constant with m =*• 2 decreased as the
c^cne pressures were increased, Ir these kinetic runs more constant rate
coefficients were obtained by taking m - 1, Finally, when the olefin was
in excess, and [DCEJ was greater than about 4 Torr, both exponents became
n --- tn =• 1, Figure 8 shows a first-order plot of an excess cls-DCE reaction
in which the formation of HCC10 was followed. With Na buffer, the rates
were invariably faster at the beginning of the reaction as is apparent
in Fig.ure 8, but the rates soon followed first-order kinetics. Such Initial
deviations were absent when 0? buffer was used as illustrated in Figure 9.
Moreover, the second-order rate constants obtained from the final first-
order kinetic section of the Ni>-buffered reactions were the same within
experimental uncertainty limits as the second-order rate constants derived
from the 0; buffered reactions. Also, under all reactant pressure conditions,
rare constants obtained from reactions with Oj buffer were always less by as much
as a factor of ten than the initial rate constants from the Ns-buffered reactions,
-------
Irons -DCE]OS 0.68 Torr
L02]0 * 240.0 Torr
JL I 1 1
I I
0 10 20 30 40 50 60 70 80 90 100 HO
REACTION TIME, Sec
Figure 7: Fourth order kinetic plot of traas-CHClCECl reaction with ozone at 23°C.
From Blume et al (71),
-------
75
5.0
03]0=0.93Torr
N2]o s 59.9 Torr
o
u_
J, 0.5 -
o
u_
10
20 30 40 50
REACTION TIME, Sec
60 70
Figure 8: First order kinetic plot of cis-CHClCHCl reaction with
ozone at 23""'C in Nz buffer. From Blume et al (71).
-------
76
10.0
|2
o
.U-.
if
o.r
|ds-DCE]o= 20.12 Torr
0,
02
= 1.31 Torr
o = 56.4 Torr
20
40 60 80 100
REACTION TIME, Sec
120
Figure 9: First order kinetic plot, of cis-CHClCHCl reaction with
ozone at 23°C in 0? buffer, From Blume et al (71).
-------
77
Numerical values of the various experimental rate constants are summarized
in Table XI. There were considerable uncertainties due to the limited
-pressure ranges'in which the rates could be determined, but the trans-DCE
definitely reacted faster than did the cis-isomer.
It was already pointed out that the reaction under excess olefin
conditions led to some isomerization of the reactant olefin. Similar iso-
merizaticns were observed even in excess ozone runs provided the olefin
pressures were reasonably high. Figure 10 illustrates the experimental
results from such a case. Here, the pressure variations of the reactant
cis-DCE and the products trans-DCE and HCC10 were determined from three
separate experiments in which the reactant pressures were comparable. It
is evident from this figure that the isomerization reaction appears to be
faster than the oEcnolysis reaction. In excess cis-DCE with ozone pressures
of about one Torr, the yield of trans-isomer appeared to increase from 20%
to about 30% as the olefin pressure was increased from 6 to 40 Torr. On the
other hand, with a similar pressure of ozone only 3-4 Torr of cis-isomer was
formed from excess trans-DCE even though its pressures were varied from 6-20
Terr, Although it was not experimentally feasible to study the kinetics of
isomerization of the olefin, the isomerization rates appeared to be a measure
of the rates of ozone disappearance as illustrated in Figure 11. In this
mn, 1,72 Torr of trans-DCE and 4.58 Torr of cis-DCE were observed at the
end of the" reaction, so that 7.31 - 1.72 - 4.58 = 1.01 Torr of reactant olefin
was consumed while the initial pressure of Oj was 1.08 Torr.
CHCICH;: Vinyl chloride also was one of the chloroethylenes examined
by Williamson and Cvetanovic (69). They found that 1.2 mole of CH2CHC1 was
consumed for each mole of ozone during the ozonolysis in CClt, solution at
25~C, bu'i the only reaction product identified by gas chromatography was 0.06
-------
Table XI
Kinetic data for the ozonolysis of 1,2-diehloroethylene (DCE)
Solvent and
Temperature
CCli» solution
25°C
S2(02) gas
23°C
Concentration
[03]«[DCE] =
1 x 10~3 - 5 x 10*"3M
[DCE1 = [03]
<1 Torr
[DCE] < 1 Torr
[03] >^ 2 Torr
[DCE] > 3 Torr
Rate
Equation
l{03} - k[DCE][03]
R{DCE} - R{03} -
k[DCE]2[03]2
R{DCE} = k[DCE]2[03]
Rate
3.
Constants
cis 35.7 if1 sec"1
trans 591 IT1sec-1
cis 1,23 x 10 "M"'sec'1
(0.12 x 1011M-3sec-1)
trans 13.1 x lO^JT sec
(4.0 x lO^J
cis 2.4 x 107M-2sec~1
(0,146 x lO'M
trans 3.2 x 107tT2sec-1
R{03} = k[DCEj[03]
(0.59 x 10
cis 4.6 x
(0.37 x lO
trans 9.0 x 102M~~1 sec"1
(2.3 x K^M^
Source
Williamson and
Cvetanovic (69)
Blume et al (71)
GO
a) Rate constants enclosed in parentheses are from Oa buffered reactions.
-------
o DISAPPEARANCE OF cis-DCE
APPEARANCE AND
DISAPPEARANCE OF trons-DCE
APPEARANCE OF FORMYL CHLORIDE
[cis-DCE]o = 4.0 Torr
[03]0 = 4.7 Torr
CM> «J
[N2]0 • 140 Torr
VO
Figure 10:
10 20 30 40 50 60 70 80 90 100 110 120
REACTION TIME, Sec
Ozone catalyzed isomerization of cis-CHClCHCl at 23°C. From Blume et al (71)
-------
80
Cis-DCE]os7.3ITorr
0,
0 = 1.08 Torr
o
20
40 60 80 100
REACTION TIME, Sec
120 141
Figure 11: First: order kinetic plot from the isomerization
for the cis-CHClCHCl reaction at 23°C with 02 buffer
From Blume et al (71).
-------
81
mole of phosgene: per mole of olefin consumed. The reaction rate was observed
to be first ordesr in each reactant and to have a second-order rate constant
of 1.18 x 103 M"1 sec'1.
In the infrared spectroscopic study of vinyl chloride ozonolysis,
currently in progress in our laboratory, Kolopajlo (94) observed no CC120
among the products. Instead the primary products from both the gas and liquid
phase reactions were formic acid and formyl chloride. Furthermore, the
reaction stoichiometry appeared to be represented by
CH2CHC1 + 03 * HCOOH + HCC10 58
The reaction in the solid phase at low temperatures gave more informative
results (92). The 1030 cm'1 region where the olefin-ozone w-eomplexes absorb
(88) was obscured by the infrared bands of the reaction products and by an
olefin band, so the presence of a tr-complex in this case could not be veri-
fied. However, as the solid reactants were warmed to about -165°C, two sets
of nuw absorption bands started to appear. Repeated warming of the solid
sample to about -150°C caused the bands of both sets to grow at the same
rate, One set of bands was readily identified as belonging to formyl chloride
(93),, and this compound began to sublime off the low temperature infrared
window at about -120°C. The second set of bands which is illustrated in
Figure 12 has been assigned to the primary ozonlde of vinyl chloride. This
ozonj.de was found to be stable to about -55°C, above which it decomposed
irreversibly into formic acid, formyl chloride, and a somewhat volatile
polymer. Interestingly, the infrared spectrum of the latter polymer was
essentially the same as that of the peroxidic polymer observed in the de-
composition of ethylene primary ozonide (88).
The spectrum shown in Figure 12 has a strong resemblance to those of
prima.ry and secondary ethylene ozonides illustrated in Figure 4. Bands near
-------
100
80
0>
o
c
o
I
1 40
£
20h
00
4000
Figure 12:
1800 1600 1200
Frequency, cm"1
800 600 400 200
Infrared spectrum of vinyl chloride primary ozonide at liquid nitrogen temperature. The
absorption bands identified by arrows are assigned to the more stable isomer of the ozonide.
The weak band at 1755 cm"1 is due to formyl chloride residue still in the ozonide sample.
From Hisatsune et al (92).
-------
83
1000 cnT1 are presumably due to 0-0 and G-0 bond stretching modes, but there
are ;wo intense bands near 700 cm'1 where the C-C1 stretch band is expected.
Although these bands grew at the same rate as the primary ozonide was being
formed, during the decomposition their relative intensity ratios were no
longer constant. Thus, these bands evidently represent the expected two
isomers (95) of the vinyl chloride primary ozonide, but the assignment of
each peak to the axial C-C1 or the equatorial C-C1 stretch in the
puckered, five-membered tri-oxa ring is not apparent. Nevertheless, the
ozonide with the lower G-G1 stretch frequency appeared to be the more stable
isomer and other bands associated with this species are identified by arrows
in Figure 12.
The kinetics of the ozonolysis of vinyl chloride in the gas phase is
also under investigation in our laboratory (94). Preliminary studies have
shown that this reaction appears to be too fast for spectroscopic study
without oxygen buffer gas. With oxygen, however, the reaction is strongly
inhibited and its rate can be followed conveniently by ordinary spectroscopic
Instruments. The results from one such kinetic run are displayed in Figure 13,
Here the pressures of vinyl chloride and oxygen are similar, and a plot of
the inverse of vinyl chloride pressure is essentially a linear function of
time. Thus, the rate under these particular experimental conditions is first
order in each reactant. The resulting second-order rate constant is 3.9 M"1
sec"1 at 22°C which is three orders of magnitude smaller than the rate con-
stant for the same reaction in GCli, solution at 25°C in the absence of Og.
Review ofOzonolysis Mechanisms
CgHi^; The recent experimental data for the condensed phase reactions
of ethylene and ozone are still consistent with the Criegee mechanism (96)
of olefin ozonolysis, which can be represented by the following sequence
of reactions
-------
6.00
50
oo
150
200
800
Reaction Time, sec
Figure 13: Second order kinetic plot of vinyl chloride reaction with ozone at 22°C. From Kolopajlo (94),
-------
85
CH2CH2 + 03 -> 6H2CH2006 . 59
> H2C*00" + CH20 60
HaC+00~ + CH20 -»• fcH2OCH200 61
HgC+OCT •*• HCOOH 62
The low temperature infrared studies of Hull et al (88) have shown that the
first stable product formed by reaction 59 near-liquid nitrogen temperature
was the primary ozonide 1,2,3-trioxaeyelopentane. Further warming of the
solid sample to about ~100°C caused the primary ozonide to change smoothly
into the secondary ozonide which remained stable to room temperature. In
the vapor phase at temperatures- above 'about 50°C, the secondary ozonide was
observed to decompose (89) by a first-order process to give formalde-
hyde and formic acid. The formation of the zwitterion in reaction
60 was inferred by the small amounts of polymeric peroxides observed after
completion of each experiment. Minor amounts of HCOOH and CH20 were also
observed in the solid phase reaction. The HCOOH could come from reaction 62
or the formation of energetic secondary ozonide via reaction 61 followed by
decoirposltion prior to stabilization. The simultaneous formation of both
the primary and secondary ozonides during the initial warming sequence also
indicated that reaction 59 must be exothermic. The enthalpy change for the
corresponding reaction 59 with 1-butene has been estimated by O'Neal and
Blumstein (97) to be about -47 kcal/mole. Kuczkowski and coworkers (85-87)
have provided a convincing demonstration of reactions 60 and 61 in the liquid
phase ozonolysis by showing that the isotopic oxygen atom from the reactant
CHjO entered exclusively the epoxy position in the secondary ozonide, and
that no labelled oxygen entered the peroxy position as previously reported
(98, 99). Thus, it appears unnecessary in the present sequence of reactions
to invoke, as Story and coworkers (100, 101) have proposed in other ozonolysis
-------
86
studies, the Staudinger primary ozonide (102) as a precursor to the
0
H2C - CHZ
1,2,3-trioxaeyclopentane, or the following additional reaction paths (103)
for the formation of the secondary ozonide.
£I7cH?oo6+cH2o
,0—CH2
\
0
CH20
63
CH2CH2000+H2C~l"OCr
CH2—CH2
£H2OCH200
64
The low temperature spectroscopic studies of Hull et al (88), appear
to clarify one other aspect of reaction 59, and this concerns the precursor,
if any, to the primary ozonide. Vrbaski and. Cvetanovic (76) apparently were
the first to propose for the ozonolysis of an aliphatic double bond that a
Tt-complex may be formed in equilibrium with the reactants according to
reaction 65 and that the subsequent rearrangement of this complex by
reaction 66 was the source of the primary ozonide.
CHaCHa + 03 2 CH2CH2«03(ir) 65
66
Story et al (103) have also included type 65 and 66 reactions in their
ozonolysis mechanism but: not as a reversible step 65. Bailey, et al (104)
have described the 66 type reaction as a 1,3-dipolar cyclo-addition and have
included additional decay steps for the w-complex to account for the expoxides
and free radical products observed in many ozonolysis reactions . These
additional steps in the present case would be as follows,
CH2CH2*03(Tr) •*• CH2CH2°03(0) ->• epojcides 67
-------
87
CH2CH2'03(ir) £ CH2CH20,3 (free radical zwitterion) 68
"--* radical chain carrier
Carlss and Fliszar (105), on the other hand, proposed two parallel paths to
the 'primary ozonide formation, namely, the direct path reaction 59 and the
sequential reactions 65 and 66 in which the complex could be either a IT- or
a 0-oomplex,
A common feature in all these proposed mechanisms is that the precursor
of the primary ozonide is the ir-complex. Although such a it-complex was not
observed in the case of ethylene, the low temperature studies (88) revealed
their presence in all other olefin and ozone reaction systems and in toluene-
ozone: systems as well. It appears very probable that the negative results
with ethylene were not due to the real absence of such a complex but to the
temperature limitation in the low-temperature cell used in the experiments.
Thus, on the basis of the results from other olefins, one may conclude that
reaction 65 is correct and that the complex must be a charge-transfer type
TT-complex. However, in no instance were the ir-complexes of other simple
olefins observed to give the primary ozonides. Instead, they decomposed
reversibly to the original olefin and ozone. Hence, reaction 66 did not
appear to occur in the condensed phases and the formation of the primary
ozonide was by the direct reaction 59. Also, the products observed in these
condensed phase reactions indicated that reactions 67 and 68 were not important,
In summary, the ozonolysis of ethylene in the liquid and solid phases can be
described adequately by the mechanism consisting of reactions 59 through 62
and reaction 65.
The mechanism for the gas phase ozonolysis of ethylene, on the other
hand, still remains unclear primarily because of insufficient data and
because of the experimental difficulties in getting such data. For example,
-------
88
information on even the reaction products is not adequate, and species which
have been identified experimentally appear to be limited to HCOOH (77)»
CH3CHO (77), aerosol of unknown composition (80), and, under low-pressure
ehemiluminescent conditions (73), vlbrationally excited OH and electronically
excited CH20 and OH. However, more data on product analysis are available
for olefins of higher molecular weights, and on. the basis of such results
two mechanisms have been proposed. One mechanism, which is still essentially
the Criegee mechanism, suggests that reactions 59 and 60 occur rapidly and
subsequent reactions initiated particularly by CH.s.00 lead to the observed
products. Here, the species CH200 may react as a zwitterion (76, 106)or as
a diradlcal (73). For example, the diradical may add to the ethylene and
by a single or multiple steps lead to CHaO and the observed rearrangement
product CH3CHO. The zwitterion may rearrange into HCOOH or react with
oxygen (106) to produce hydroxyl and performate free radicals, both of which
can initiate other free radical reactions.
The second mechanism is that due to O'Neal and Blumstein (97), and it
was proposed principally to account for the energy requirements in chemi-
luminescent reactions and for the products such as a-diketones which are
difficult to explain by the Criegee mechanism. These authors suggested
that the primary ozonide formed in reaction 57 is in equilibrium with an
opened-ring diradical species which for ethylene would be as follows
CH2CH2000 £ »OCH2CH200« 69
The diradical may then dissociate to give the normal Criegee products of
reaction 60 or it may undergo an intramolecular a-hydrogen abstraction
reaction to give an a-keto hydroperoxide.
*OCH2CH200» •*• CH200 + CH20 70a
-»• OCHCH2OOH 70b
-------
89
The hydroperoxide may decompose tnto the normal ozonolysis products, HCOOH
and CH20, into water and glyoxal Ca-diketone), or produce an OH radical and
an oxy-free radical, O'Neal and Blurasteln also estimated the energetics of
reactions 70a and 70b and concluded that for ethylene and propylene the
diradical should decay mainly by step 70a while for 1-butene and other
olefins with greater internal degrees of freedom reaction 70b should
dominate. Consequently, secondary ozonides should not be the major products
from the latter ozonolysis reactions particularly at higher total pressures.
The low temperature infrared studies (88), on the other hand, revealed that
not only ethylene and propylene but 2-butenes also gave secondary ozonides.
Although O'Neal and Blumstein suggested that reaction .63. may be the source
of any unexpected secondary ozonides, the isotopic studies by Kuczkowski and
cowoi'kers (85-87) appear to rule out this possibility. It seems that the
estimates of energetics of these ozonolysis reactions may not be completely
valid or additional modifications of the reaction mechanism may be necessary,
A further shortcoming of the O'Neal-Blumstein mechanism is that internal
B-hydrogen abstraction by the diradical was suggested to explain the chemi-
luminescence of 03-olefin reactions. Of course in the C2Hi, system there is
no 6-hydrogen. Finlayson et al (73) modified the O'Neal-Blumstein mechanism
to suggest that electronically excited CH20 was produced by a-hydrogen
abstraction
0*
HC :CH2 ->• HCOOH +'CH20* 71
H O •
•V
but this route seems unlikely from both energetic and steric considerations.
The e:cperimental evidence clearly requires the presence of free H atoms, and
Finlayson et al (73) suggested two routes both of which are variations of
-------
90
sequential H atom splitting from the zwitterion
H2COO •->• HCOOH* ->• H + HCOO •> H + C02 72
Again, these routes seem unlikely to us, and we prefer hydrogen abstraction
by the single 0 atom in the diradical
0—H
H2C—CH •*• H2COH + HCOO -»• H + C02 73
.0—0
The proposed routes for electronically excited OH production were (73)
0 + H H- OH*
0 + HCOT •> OH* + CO
However, at this time the proposed routes to chemiluminescence must all be
considered to be speculative.
Until very recently (83, 73) the only investigation that indicated that
the second order kinetics of the gas phase ethylene ozonolysis was different
in the absence and presence of 02 was by Wei and Cvetanovic (78). Herron
and Huie (83) noted instead that the experimental results were nonrepreducible
and much larger second order rate constants were obtained when the reaction
was carried out in argon buffer gas. However, with added 02, rate constants
which agreed with earlier literature values (in the presence of 02) were
obtained. In the case of propylene, Herron and Huie showed that the second
order rate constant decreased by a factor of almost two as the 02 pressure
was increased to about one Torr and thereafter remained constant on further
increase in 02. This limiting rate constant was found to agree closely with
those reported by earlier investigators (See Table IX). On the basis of these
results and those reported by other investigators, Herron and Huie proposed
a schematic free radical mechanism for the ozonolysis reaction.
03 + C2H,t -»•• P* 74
P* •* Q + R + . . . 75
-------
91
P* 4- M -* P + M , 76
03 + (Q + R + . , .) •* products 77
02 + (Q + B, + ...)•* products 78
Here, P* was reported to be an adduct which is not necessarily formed initi-
ally but could, be formed by a subsequent rearrangement.
Cj^Fjt.: A simple mechanism involving an ozone-oleftn adduct was
praposed by Heicklen (1, 68) to account for the limited experimental data
available for this ozonolysis reaction.
79
03
+ C2F«» -* 4CF20 + 02 80
The nature of the initial adduct in reaction 79 was not specified but
reaction 80 was proposed originally (68) to be composite and to involve
intermediates such as CF2, (CF20)2, and e2Fi»0. Later (1), reaction 80
was represented by the following sequence of steps.
C2F<,03 + C2F% -»• C2F%0 + C2F^02 81
C2F%02 •> 2CF20 82
C?Fi,0 + 03 -> C2F^02 + 02 83
Whichever multiple steps reaction 80 may involve, the rate equation re-
sulting from reactions 79 and 80 is
R{CF20} . 4k7,kyfO,][C,F ]2 .
2
Comparison with the experimental rate equation gave k?g = 300 M_-1 sec"1,
kso/k-tg > 9 x 10"* M-:I and k79keo/k_79 > 3 x 107 M~2 sec"1 at 25°C. However,
thiu rate equation did riot account for the observation that the rate became
independent of [03] ur actually decreased at high ozone pressures.
Gozzo and Camaggi (91) included additional steps to the Criegee
mechanism to explain the observed solution phase reaction products i CF20,
-------
92
the epoxide CF2CF20, cyclo-C^Fe , and traces of secondary ozonide .
03 -* [C2F403] -*- CF20 + CF200 84
CF200 + C2F4 ->• CF2CF20 + CF20 85
2CF200 -*• 2CF20 + 02 86
CF200 ->• CF2 + 02. 87
CF2 + CaFi, ->- c-C3F6 88
However, these investigators also observed only CF20 in the gas phase
reaction, so the gas phase mechanism evidently consists of just reactions 84
and 86. Hence, the rate R{CF20} will be first order in each reactants a
result in accord with Heicklen's (1, 68) high pressure limit. The reaction
in the solution phase gave CF20 and CF2CF20 as the major products, so in this
case Gozzo and Camaggi considered only steps 84, 85, and 86, A steady-
state approximation for CF200 gives
d[CF20] = 2 + 2kBj>kB6_[_03_]_ XVIII
d[CF2CF2"5] ke5[C2F4]
2
with k8lfke5/k85 being about 4.
C^Cl^t.' On the basis of kinetic data from all the chloroethylenes
studied in CCli, solutions, Williamson and Cvetanovic (69) proposed the follow-
ing general mechanism for the ozonolysis of these olefins RR'.
RR' + 03 -»• product I 89
RR' +.03 ^ RR'»03 90
RR'»03 -»- product II 91
Here, reaction 89 was described as a one-step process giving product I
which does not return to the reactants. The complex formed in reaction 90
may return to the reactants with no geometric isomerization or it may decom-
pose irreversibly, according to reaction 91,, In these reactions the products
I and II are some intermediates of the reaction and not necessarily the final
products. With this mechanism, the observed second order rate constant was
-------
93
related to the elementary constants as follows:
k89 + (k9ok9j/k_90)/(l + k90/k-9o) , XIX
g
Although the experimental data were too limited to determine the relative
importance of the one and two-step terms in equation XIX, Williamson and
Cvetanovic suggested the possibility that reaction 89 may, lead, to the
primary ozonide while reactions 90 and 91 may correspond to the oxygen
transfer process observed, for examle, in the ozonolysis of ethylene where
CH3CHO was produced. We have already quoted the second order rate constants
obtai.ned by Williamson and Cvetanovic during the review of each haloe thy lene,
but these constants are summarized together in Table XII,.. . . . . .
A more elaborate mechanism was necessary to interpret the gas. .phase
results obtained by Mathias et al (5). The initial step in this mechanism,
shown below, gave the diradical Criegee product CC1200 which propagated a
chain reaction.
C2C1^ + 03 •> CC120 + CCl2Pq . , ; : 92 : ,
CC1200 + 03 •> CC120 + 202 93
CC1200 + C2Clv •* CCla002*C2c:U • 94
03 •* CCl200'C2Cllt'03 95a
-> CC1200 + C2C1V0 + 02 95b
* CC1200 + 2CC120 + C2C1(,0 96
In thd.s mechanism, CgCl^O formed in steps 95b and 96 represented both
the epoxide fcdaCClzd and the rearranged product CC13CC1(0). Application of
steady state approximations to the intermediates leads to the following
initial rates, RI{X}
- 2k92[03][C2Cli|] + 2 (9.9"Sa) [CaCl^ XX
kg 3kg 5
Ri{C2ClllO> - -*-** [C2C1<,]2 XXI
-------
94
Table XII
Second order rate constants for the reactions of ozone
with haloethylenes in CCli, solution at 25°Ca
Haloethylenes
CCl2CClt 1,0
CHC1CC12 3.6
CBaCCla 22.1
cis-CHClCHCl 35 . 7
trans-CHClCHCl 591
CH2CHC1 1,180
a) From Williamson and Cvetanovic (69).
-------
95
For long chains „ only the second term in equation XX is important so that the
litCClaO} should be second order in [C2Cli»] and independent of [03]. Experi-
mentally, the olefin order was about 1.8 and no 0$ dependence was found.
Also for long chains, li{CCl20}/Ri{C2Clj»0} = 2k9Sa/k95 = 1.3 so the branch-
ing ratio kgsb/kgsa = 0.54.
The ozonolysis reaction was inhibited by 02 » and in this case the chain
termination step was suggested to be as follows i
CC1200-C2C1^ + 02 -+ CClzOOCzCl^'Oa 97
CC1200«C2C1.»«02 + C2C1% -»• 3CC120 + C2Cli40 98
The rate law for high pressures of oxygen then becomes
R±{CCI20} = 4k92[03][C2Cllt] XXII
where the upper limit of ks2 was estimated to be 1.2 x 10"~2 M"1 sec""1.
This second order rate constant is two orders of magnitude smaller than
that obtained by Williamson and Cvetanovic for reaction 92 in CCli, solu-
tion, Since 'the average value of kgakg^/kgs from the nitrogen buffered gas-
phase reaction was O.IS^I"1 sec"1, the lower limit of the ratio kg
becones about 10.
CH2CCl2t The experimental data for the gas phase ozonolysis of
CH2CC12 were more extensive, and Hull et al (70) proposed the following
chain mechanism:
CH2CC12 + 03 •*• CH20 + CC1200 99a
-> CH200 + CC120 99b
CC1200 + 03 ^ CC120 + 202 93
CC1200 + GH2CC12 ->• CC1200»CH2CC12 100
CC1200»CH2CC12 + 03 -»- CC1200«CH2CC12*03 10 la
-»• CC1200 + CC120 + HCOOH IQlb
CC1200»CH2CC12 + 02 -»• CC1200'CH2CC12'02 102a . .
-------
96
-* 2CC120 + HCOOH 10 2b
CC1200«CH2CC12 -»- CC120 + CH2CC120 103
CH20 + 03 -»• 02 + HCOOH (or CO + H20) 104
CC1200'CH2CC12-03 + 2CH2CC12 •*- 3CH2CC120 + CC1200 105
CC1200'CH2CC12-02 + CH2CC12 -> 2CH2CC120 + CC1200 106
The Criegee dissociation of the initial ozone-olefin adduct can occur
in two ways, but on the basis of the reaction products, the stoichiometry,
and the dependence of the rates on oxygen, Hull et al proposed that reaction
99a was the dominant primary step. Following this step, the propagation of
the chain reaction is maintained by CC1200 through reactions 100, 101 ,
105, and 106. Reaction 104 was included to account for the absence of
CH20 among the products. The product CH2CC120 in steps 103, 105, and
106 was considered to be vibrationally excited and to be the source of the
rearranged acid chloride CH2C1CC1(0) and the products HC1 and CO. In the
absence of 02 and for long chains, if k^s « kioiEOal tne above mechanism
gives
-d[03]/dt = 2k99a[CH2CCl2][03]
ki 00 [CH2CC12 ])
k93[03] + kiook103[CH2CCl?J/k101[03]
At high CH2CC12 but low 03 pressures, equation XXIII reduces to
-d[03]/dt = (kcl9ak]Oi/kl03)[CH2CCl2][03]2 XXIV
while at high 03 but low CH2CC12 pressures it becomes
-d[03]/dt - (k99akii)o/k93)[CH2CCl2]2 XXV
since under all the experimental conditions k93[03] « ktoo [CH2CC12 ] .
The latter rate equation was proposed as the reason for the faster
rates observed initially in both N2 and 02 buffered reactions (See Figures
1 and 2 in Reference 70). For all pressure conditions, moreover, the
mechanism gives:
-------
97
d[CH2CCl2]/d[03] = 1 + 2k1Qia/ki0r XXVI
and
-d[CCl20]/d[09] - kjoib/kjoi XXVII
When 02 is present in excess and (kjo ibkioo/kioizHCI^CC^] » kgBfOa],
the predicted rates are
d[CC!20]/dt - -d[03]/dt = 2k99a[CH2CCl2][03] XXVIII
and the consumption ratio of the reactants becomes
- • - ',' . '
d[CH2CGl2]/d[03] « 1 + ki02a/k102b XXIX
Comparison of,the derived rate equations and various ratios of the rates
with those determined experimentally permitted Hull et al to evaluate the
elementary rate constants given in Table XIII. In the study of Williamson
and Cvetanovic (69), the olefin concentrations, which were always in excess
over the ozone concentration in the CC1»» solutions, were equivalent to 2 to
92 Torr range, and thus similar to the pressure range used by Hull et al in
Nj buffered gas phase studies. Therefore, if the rate observed by Williamson
and Cvetanovic corresponds to that of reaction 99a, then this reaction is
20 times faster in CCli» solution than in the N2 buffered gas phase.
CHCICHCI; The unusual changes in reaction order with, pressure,
the isomerization of the reactant,. and the inhibition of the rate by Og
suggested to Blume et al (71) that the mechanism of the ozonolysis of cis
or trans-diehloroethylene (DCE) was a very complex chain reaction. The
siirplest mechanism which accounted for the observed results except the
isomerization was proposed to be as follows:
R2 + 03 2 B203 107
B,a03 + R2 £ Ri*03 108
S.t»03 4- 03 -* 2RO + 2R02 109
R02 + R2 2 R302 110
-------
98
Table XIII
Elementary rate constants in the mechanism of
ozonolysis of 1,1-dichloroethylene at 25°C
Rate Constant Valuei ' Units
3 2.4 x 106 M-
0.6 None
0.4 None
tea "XL.9 x 10s M"1
1.1 M-*
4 None
2.1 x 106 M""1
a) From Hull et al. (68) „
-------
R~, 0,
R302 + 03 -> R305 % 3 4RO + ROg + 02 Ilia
•*• 3RD + 02 i lllb
03
R302 + 02 * R30% •* 3RO + 202' 112
Here, DCE i^ represented by R2 and R02 (t?he Criegee diradieal) is the chain
I
carrier. Reactions lllb and 112 are! the chain termination steps but
whichever step is operating the overall stoichiometry becomes the same as
that observed experimentally. Steady state approximations for the various
reaction intermediates allow the derivation of the following rate equations .
k_107(k-108 + kl99[03])
d[RO]/dt - 2RC109H1 + j^Mia +. fSinhHOll + 3kll2[0g]}.
kmbl03J •*• K-iizlOa J
The rate of disappearance of ozone -R{03} or of the olefin -R{R2} is
given by one-half the right-hand side of equation XXXI, and the expression
enclosed in braces provides the 02 dependence of the rate. With no 02,
this expression reduces to {4 -f 4(kn la/^mb)) while with excess 02 it
becomes simply (4). The R{109} coefficient In equation XXXI, on the other
hand, reduces to different expressions depending on the pressures of the
reactants. For low R2 and 03 pressures
R{109> = 07knek»fl3 [R2]2[03]2 XXXII
so the reaction rate becomes fourth order overall. If [R2] is high, then
R{109> » k107[R2][03] XXXIII
Finally, when [03] is high, the R{109} term reduces to
R{109} - (k107ki08/k-io7)[R2]2[03] XXXIV
The three rate laws XXXII-XXXIV correspond to the three limiting cases
observed experimentally, .
-------
100
By a computer fit of all the data, the pertinent ratios of rate coeffi-
cients for the mechanism consisting of reactions 107-112 were obtained, and
they are summarized in Table XIV. If the second order rate constants deter-
mined from the CCli^ solution by Williamson, and Cvetanovic (69) were the same
as k107, then the gas phase constants for cis and trans-DCE are, respectively,
about 0.62 and 0.25 of those in CCli^ solution. In CCli^ solution, the trans
isomer reacted 17 times faster than did the cis-DCE while in the gas phase
the trans isomer reacted only 6.7 times faster in the second order limit.
In comparison to these differences, Carles and Fliszar (105) found that the
ozonolysis of trans-2-pentene was just 1.5 times faster than the cis-isomer
reaction in CCli* solution at 0°C.
The mechanism proposed here also provides possible channels for the
isomerization of the DCE which was observed during the ozonolysis reaction.
If the products of reactions 107, 108, and/or 109 are noncyclic with loss of
carbon-carbon double bond character, then their decompositions to the reactants
will give an isomer different from the initial reactants. Unfortunately, the
ozonolysis rates of cis and trans-DCE were too similar in the gas phase and
the experimental data were not sufficient to make any quantitative deductions
concerning this isomerization reaction. However it appears that channels
-107, -108, and -110 are not sufficient to explain the isomerization results,
and other channels are needed.
Discussion
Although the available information, both experimental and mechanistic,
on each haloethylene reviewed here is not sufficient to derive a complete
general mechanism for the ozonolysis of simple olefins, two significant
characteristics of such reactions emerge when the entire data are examined
together. First, many of these ozonolysis reactions are inhibited by mole-
cular oxygen. Such inhibitions have been observed in the ozonolysis of
-------
101
Table XI?
Elementary rate constants in the mechanism of
3,
ozonolysis of 1,2-dichloroethylene at 23°C
Rate Constant
cis-Isomer
trans-Isomer
Units
k,!07
22
4.6 x 10"
3.2 x 10"
3.0
*2
148
4.6 x 10"
3.2 x 10*
1.0
^2
M~ s«
IT1
r1
None
None
-1
a) From Blune et al (71).
-------
102
C2e]U(5), CH2CC12(70), cis and trans-CHClCHCl(71), and CH2CHC1(94). This
inhibition has been reported now even for CH2CH2(73,83). Oxygen. Inhibition
was not reported in the ozonolysis of C2Fit(68), but in this case the oxygen
is a reaction product and the [03]o > [CaFi*] condition used in the gas phase
study would have made it difficult to observe this inhibitions since molecular
QZ was not added deliberately. Thus, it appears that oxygen inhibition may
be a general characteristic of the haloethylene ozonolysis reactions, and it
may be so even in the ozonolysis of other simple olefins. The mechanisms
of these reactions, therefore, presumably is a free radical type each in-
volving a biradical species.
The second significant characteristic of the ozonolysis reactions reviewed
here concerns the origin of the biradical species which caused the oxygen in-
hibition described above. In the low temperature infrared study of the
ozonolysis of CHgCHa(88), both the primary and secondary ozonides were detected,
Thus, the reaction in this case may be considered to proceed by the Criegee
type mechanism where the primary ozonide is the source of the biradical
species. The infrared study of the low-temperature ozonolysis of CHgCClj
and CHC1CHC1(92), on the other hand, showed no formation of primary ozonides
even though the reactions were taking place as evidenced by the appearance
of the infrared bands of phosgene and formyl chloride, respectively. The
source of the biradicals for these reactions m^ist, therefore, be other than
the primary ozonides. In the case of the low-temperature ozonolysis of
CH2CHC1(92), the formation of both formyl chloride and the primary ozonide
was observed in a temperature range where the decomposition rate of the latter
was negligible. Thus, two independent reaction paths appear to be available
here, one involving the primary ozonide and the other channel by-passing this
intermediate. These spectfoscopic observations indicate, in fact, that the
parallel path reaction mechanism proposed by Williamson and Cvetanovic (69)
-------
103
consisting of reaction steps 89-91 may be correct in principle and,
furthermore, it can now be modified to be consistent with the experimental
observations.
For a general olefin RR' the Criegee path of bir,adical formation is
given by reactions 113 and - 114. Reaction 113 is Irreversible and
Crie^ee Path
RR' + 03 -> RR'Ooi 113
RR'OOO •*• R02 + R'O 114a
-»• R»02 + RO 114b
RR'OOQ -*• Other products 115
leads to the primary ozonide which provides the blradical R02 and R*02, in
case of an unsymmetrleal olefin. In addition the trioxolane opens and decom-
poses to other mono free-radical and molecular products as proposed by O'Neal
and Blumstein (97), The second reaction channel, which we shall call the
TT-complex path of biradlcal formation, is based primarily on the requirements
provided by the kinetic study of els and trans-CHC1CHC1. This scheme consists
TT-Complex Path
RR' +'03 £ RR'03 107'
RR'Oj + RR1 £ R2R2'03 108'
R2R2'03 + 03 -»- 2R'0 + 2R02 109a'
-»• 2RO + 2R'02 109b'
o:: two reversible reactions 107' and 108' followed by an irreversible
si:ep 109't The products of reaction 107' may be the ir-complexes which have
been observed with many olefins at low temperatures (88,92). According
to the low temperature studies with cis-or trans-CHClCHCl, the reverse of
reaction 107' does not involve an isomerization but the reverse of reaction
108' may (92), Both the TT-complex and Criegee paths may be important in
the ozonolysis of CH2CHC1, but for the other chloroethylenes only the
-------
104
TT-complex appears necessary to account-for the experimental observations.
Thus we see that the 0$ reactions proceed primarily by different paths
for different ethylenes, similar to the observation for 0-atom reactions.
Earlier in this review we showed that there are three classes of 0-atom
reactions, one for C2Hi( and CH2CHC1, one for the fluoroethylenes except
CF2CC12, and one for the other chloroethylenes including CF2CCl2.
Likewise we find here that C2H^, CH2CHC1 (and the higher unhalogenated
olefins) form molozonides which decompose by the Criegee mechanism to give a
rate law first-order in both olefin and 0$ over the entire pressure range.
On the other hand, the higher chloroethylenes do not form molozonides and
react with 03 by a complex rate law which deviates from second order (first
order in each reactant) at low reactant pressures, indicating reversibility
of the initial reaction step. Furthermore CHC1CHC1 undergoes geometrical
isomerization, whereas the 2-butenes do not.
It is not clear from the data whether C^Pn ozonolysis fits into one of
these reaction classes or proceeds by a third scheme, as in 0-atom attack
on CaFit. No deviation was observed in the second-order rate law over the
range studied, but, by analogy with (\Fe-2 which did show the deviation,
Heicklen (68) interpreted his data in terms of a changing rate law. However
Heicklen believed that no CF202 diradicals were present because no CF2CF20
was observed as a product in the room-temperature gas-phase ozonolysis, yet
CFaCF26 was a product when CF202 was produced in the CzF^-Oz-O system
(46,67). If Heicklen's inference is correct then the ozonolysis of CzFn
must be different than that for either the olefins or the chloroethylenes.
One note of caution in this inference is that CFgOO may exist as either a
triplet or singlet species, which may react differently. Thus the triplet
CF200 (presumably produced in the 0-02-C2Fit system) might lead to
whereas the singlet CF200 (as expected in the 03~C2Fit system) might not
-------
105
.Lead to CF2CF20. However If CF200 is not produced the it-complex mechanism
might still explain the .results by adding the reaction
R2R2'03 + 03 •+ 2RO + 2R*0 + 202 109c'
Once the diradical R02 (or R'02) is produced, it can participate in a
chain process, an example of which is given by reaction 110' through 119.
Chain Propagation and Termination
R02 4- RR' -»• R2R'02 110'
BO 2 + 03 -»- RO + 202 116
RaR'02 •* RO 4- RR'O 117
RaR'O* + 03 -*• R2R«05 Ilia1
-> 2RO + R'O + 02 lllb1
R2R'05 + RR' ->• R3R2'05 118a
^ RR'O + RO 4- R'O + R02 H8b
R3R2'05 4 03 -> 2RO + 2R'0 + 02 + R02 119
Hare, RO and R'O are the carbonyl products while RR'O may be an epoxide or
a rearranged product such as an acid chloride. The biradical chain carrier
i(3 R02 but the oxygen inhibition of the rate is represented by reactions 112' ,
120, and 121.
Inhibition
R2R'02 4- 02 •* RzR'Oi, 112'
R2R'Oi» + 03 -*• 2RO 4- R'O + 202 120
R2R'0^ 4- RR' -»• RR'O 4- 2RO 4- R'O 121
For the chloroethylenes the mechanism consisting of reactions 107',
108', 109a', 109b', 110', 116, 117, Ilia', lllb', and 112' leads to the
generalized rate law
-d[RR'I - .2ak107'ki0a'r03][RR']2 ( k11Q'[RR'] _ ,
dt Ik_107'+k10ef o[RR!] u Bkiio ' [RR1 1 + ki,6 [02];
-------
106
except for stoichlometric factors which depend on the fates of R2R'Qs
(reactions 118 and 119) and Rj-R'Oj, (reactions 120 and 121). In eqn. XXXV s
the quantities a and g are defined by
a = (kiQ9a« + kiQ9bOEO«]/(k_jQ8' + (kj.a9a> + kiS9bf)[03])
1- B = killat[03]/(k111i[03] + ku2'[02] + ki17)
The rate law, eqn. XXXV » is adequate to give calculated rate equations which
are consistent with the experimental equations. For example, the gas phase
ozonolysis of C2Fi» can be interpreted on the basis of the ir-complex path.
Since the reaction products were 02 and CF20, reaction 109cf can replace
reactions 109 a1 and 109bf to obtain for the rate of formation of CF20
R{CF20} =
k_107'
which is independent of oxygen pressure. In the case of CgClif, the assump
tion of long chains with kn? and kmb' considered small compared to ku]
the rate laws in the absence of 02 become
ii{cci2o} = iJua (i + uiaa)[c cl jz -xxxvii
ki i6 ki IB
xxxvill
jLujii tci01t]
where CjClifO includes both CClaCC^O and CClsCClCO), and in the presence
of 02
Ri{CCl20} - 8kl07'[C2Cllt][03] XXXIX
Similarly, a long-chain process for the ozonolysis of CHC1CHC1 gives
-d[03]/dt - -d[CHC!CHCl]/dt - d[HCC10]/2dt «
4kio7tk10B«kio9'[CHClCHCl]2[032 _ _ fkiii'IOal +
oe1 + kiog'tOsl) + kio8'kio9'[CHClCHCl][03] mb'tOsl + lniiz'[0i]
XL
where in eqn. XL reaction 109cf is assumed to be negligible.
-,
-------
107
For the unsyiranetrical olefin CHaCCla the chain carrier was deduced to be
CClgOO rather than CHaOO, Furthermore, no formaldehyde was observed as a
product of the ozonolysis, so reaction 122
R'O + 03 -> HCOOH (or CO + H20) + 02 122
mtst be Included in the general mechanism. The use of steady state approxi-
mations on the various intermediates including R'O and the assumption of
long chains lead to the following rate equations ,
to>] + tll..[0l]}
-------
108
II. OXIDATION OF CHLOROMETHANES
EXPERIMENTAL
Mixtures of perhalomethane with 02 or 03 or both were irradiated in
a cylindrical quartz reaction cell 10 cm long and 5 cm in diameter. The
cell was attached to a conventional Hg-free glass vacuum line equipped
with Teflon stopcocks with Viton "0" rings. Extra dry grade Og from
the Matheson Co. was used without further purification. The Oa was
prepared from a Tesla coil discharge through Oz and was distilled at
87°K before use. The CClit was "Baker Analyzed" reagent and was purified
by distillation from a trap maintained at 2.10° to one at 177°K. The
CFC13, CF2C12, CF3C1, N20, and C02 were obtained from the Matheson Co.
and were purified by degassing at 77°K. In few experiments the CF2C12
was purified by distillation from a trap maintained at 87°K to one at
153°K. The perhalome thane pressure was measured with an H2SOit manometer,
and the Og, C02, and N20 pressures were measured with an alphatron
gauge calibrated against an H2S04 manometer. The 03 pressure was
measured spectrophotometrically at 253.7 nm and could be monitored
continuously.
The 213.9 nm radiation for the photolysis of the perhalomethanes
was provided by a Phillips (93109E) low-pressure zinc resonance lamp.
For the O^Dj-atom study, the 253.7 nm radiation was obtained from a
Hanovia "spiral" low pressure Hg resonance lamp. The 253.7 nm line was
isolated by passing the radiation through C12 gas and a Corning CS 7-54
filter before entering the reaction cell.
For the CClit system actinometry at both wavelengths was done by
measuring Hg production from HBr photolysis where the quantum yield of
H2 production, ${H2}, is 1.0 (107). For the chlorofluoromethanes, the
-------
109
actinometry for the photolysis experiments at 213,9 nm was done by
measuring the rate of Ng production from NgO photolysis. For this
system ${N2} = 1.41 (108). The actinometry at 253.7 nm was done by
either measuring the 03 removal in pure 03[-${03} «• 5.5 (109)], or by
measuring flNal in the photolysis of 03 in the presence of excess N20,
For the latter system *{N2} = 0.46 for thermally equilibrated 0(1D)
atoms and tCNa] =0,41 for 0(1D) atoms possessing excess translational
energy (108).
Analysis for COC12 was performed mainly by gas chromatography and
in a few experiments by infrared analysis. Chromatography was done with
a stainless steel column 10' x 1/4" containing 10% silicone oil (SP2100)
on 80/100 mesh Supelcoport (Supelco Inc., Beliefonte, Pa,). Analysis
for Cl2 was made in the photolysis experiments by chromatography in
the same column, as for COClz and by ultraviolet absorption spectroscopy
in .3. Gary 14 spectrometer. For the 0(1D) experiments the analysis for
Cl2 was made with a dual beam spectrophotometer (110) at 366.0 nm in
order to obtain greater sensitivity. It was assumed that the increase
in absorption at 366,0 nm was due entirely to Cl2. For CO analysis the
Q
column used was 10' x 1/4" containing 5A molecular sieves. For CaCle
analysis a flame ionization chromatograph was used equipped with a
10' x 1/4" column containing 31 SE 30 on Supelcoport.
Analysis for CFC10 and CF20 was made with a thermal conductivity
gas chromatograph equipped with a copper column (10* x 1/4") containing
silica gel. On this column the CFC10 and CFaO are quantitatively con-
verted to C02, (2), and it is actually the yield of COg that is measured,
O
For N£ analysis the column used was a 20' x 1/4" column containing 5A
molecular sieves. Analysis for Cla was made by UV absorption with a
Car}r 14 spectrophotometer.
-------
110
For the chlorine-atom initiated oxidations the experimental pro-
cedure was exactly the same as that described for the chloroethylenes .
The CH2C12 was Eastman Kodak Spectro ACS grade, and the fraction volatile
at -80° but condensable at -130° was used. The CH3C1 was from the
Matheson Co, and the fraction volatile at -80° but condensable at -196°
was used.
PHOTOOXIDATION OF THE PERHALOMETHANES
Photolysis of
When CCU is photolyzed at 25°C with 213.9 nm radiation in either
the presence of 02 or 03 the products are CClaO, C12 , and an unidentified
compound (111). At low total pressure, ${CC120} = 2.0, but this value
drops to 1.0 for [CCliJ ^ 50 Torr and [02] or [N2 ] = 700 Torr as shown
in Fig. 14. ${C12} is reasonably invariant, to pressure at ^ 1.3-1.4.
The results are interpreted in terms of an excited molecule mechanism
which proceeds entirely by
CCU* -> CC12 + C12 123 .
at low pressures, with singlet CC12 being produced. At higher pressures
CCU* is quenched and CC12 production is inhibited, though it may be
(and probably is) replaced by production of CC13 + Cl.
At low pressures the photolysis data for CCU in the presence of
02 are consistent with the scheme:
CCU + hv (213.9 nm) -»• CC12 + C12 123
CC12 + CCU •* 2CC13 124
CC13 + 02 -»• CC1302 125
2CC1302 -»• CC130 + 02 126
CC130 -»• CC120 + Cl 127
Cl -+ (1/2) C12 128
-------
Ill
100 200
300 400
[02] or JN2], torr
500 600 700
Figure 14: Plot of ${COC12} vs [N2] or [02] for CCU photolysis at 213,9 nm
in the presence of 02 or Oa at 25°G. 0 [CCU] ^ 10 Torr in the
presence of Og, A [CCU] 'v 10 Torr in the presence of Os,
• [CCU] 'V 50 Torr in the presence of 02» A [CCU] ^ 50 Torr
in the presence of Os. All analyses by gas chromatography.
From Jayanty et al (111).
-------
112
In the presence of 03 reactions 125, 126, and 128 must be replaced by:
CC13 + 03 -> CC130 + 02 129
Cl + 03 -> CIO + 02 10
2C10 -»- 2C1 + 02 130a
2C10 -»• C12 + 02 130b
Primary process 123 followed by reaction 124 is suggested because
the photolysis at low pressures leads to 2 molecules of phosgene per
photon absorbed and therefore 2 molecules of CCl^ must be removed per
photon. The only fragment which could decompose a second molecule of
CCli* appears to be CC12, Cl, CCla or any of the oxygenated radicals
are unlikely to react with CGli».
The CC12 produced would be expected to be in a singlet state, from
spin conservation rules. This is supported by the fact that the CCla
fragment does not react with 0%. Triplet CCla reacts readily with 02
to produce CO (36), but no CO was found in this system.
The fate of CC13 in the presence of 02 is given by reactions 125-
127 as first suggested by Huybrechts et al (13) and confirmed by Mathias
et al (5). The quantum yield of phosgene in the presence of 03 is the
same as in the presence of 0;?| consequently the reactions of CC13 with
02 and 03 must ultimately lead to a common precursor of COC12, Therefore
reaction 129 must be the principle reaction between CClj and 03. In the
presence of 03 the Cl atoms will be removed by reaction 10, (kio = 2 x 1Q"11
cm3/sec) (112). The CIO radicals produced in reaction 10 will be
removed by reactions 130a or 130b, depending upon the total pressure.
The bimolecular reaction of CIO radicals at low pressures (< 8 Torr argon)
is known to proceed exclusively by reaction 130a (113). At higher pressures
(_> 70 Torr argon) reaction 130b is the exclusive reaction (114, 115).
-------
113
In the present experiments reaction 130a could occur at the lowest
pressures used 010 Torr CCli,), though for experiments for which
*v» 50 Torr reaction 130a is negligible. The reaction
CIO + 03 ->- Cl + 202 or C102 + 02 14
can be neglected because it is slow.
At higher total pressures ${COC12} declines and reaches a value of
about 1,0 at 600-700 Torr N2 or 02 and 'v 50 Torr GCl^. The data is
shown graphically in Figure 14. A readily apparent explanation of this
pressure effect, which is consistent with all the data, is the partici-
pation of a relatively long-lived excited state of CCU. Thus the
following paths are possible:
CC1,, 4- hv •* CCli,*
CCU* + M -*• CCU
*»
CCli,* ->- CC12 + C12
or
eel.,* + M -»• ccii,**
CCli,* ->• CC12 + C12
CCJU** -*• CC13 + Cl
where * and ** are excited states of CC1«» and these could be different
electronic statess, or the same electronic state with different vlbrational
energies.
The difficulty of postulating a long lived excited state for CCU
is that spectral, studies of other halomethanes suggest that the broad
band observed from about 160-250 nm can be attributed to a n-0* trans-
ition which is not likely to lead to a stable excited state (116, 117).
However, the fact that the primary process appears to be molecular C12
elimination implies that the transition does not lead to a simple repul-
sive potential curve along the C-C1 bond reaction coordinate, but must
-------
114
involve considerable electronic rearrangement,
The most reasonable explanation of the pressure effect would be:
CC12 + CCU -*• CaCls*
C2C16* -*• 2CC13
C2C16* -*• M -> C2C16
but a careful search for C2Cl6 production was negative. The failure to
stabilize CgClg (providing it is formed) even at 1 a tin. 02 or N2 is
not impossible since the A factor for C2Cle decomposition is very large
(1017*7 sec"1) (118). Furthermore primary process 1 is exothermic by
^ 50 kcal/iuole at 213,9 nm, and the CCla may be produced with excess
energy,
In the mechanism, we have neglected the reaction of CCla with 02
at high 02 pressures, because addition of 02 has the same effect as the
addition of Nj. Consequently the reaction of CCla with 02 cannot compete
with reaction 124; the rate coefficient is <_ 10~13 em3/sec, and CC12
cannot be in its triplet state.
The first of the excited-state mechanisms predicts that ${COC12} =
${Clg} goes from 2 ->• 0 as [M] goes from 0 -»• °°; whereas the second
mechanism predicts that ${COC12} = §{Cl2) goes from 2 -»• 1 as [M] goes
from 0 -> oo, fhe highest total pressures used were not sufficiently
high to determine if the quantum yields of COC12 drop below 1, Thus
if either of the two mechanisms is operative the present data cannot
distinguish between them. However, since at longer wavelengths the
primary process
CCU + hv (%250-nm) •* CC13 + Cl 123'
becomes dominant, (119, 120) the second mechanism is more attractive,
because it provides .for the formation of CCls + Cl within the same
-------
115
electronic transition; * and ** now would refer to different vibratlonal
levels of the same electronic state.
The Cl-i quantum yield is substantially below 2 (1.2 - 1.4) at low
pressures and is insensitive to total pressure, contrary to expectation.
The-, reason for this is not known, but stable oxides of chlorine may have
been formed which were not detected. The previously mentioned product
observed in the U..V. spectrum of the reaction mixture does not" correspond
to that of any of the known chlorine oxides. Both mechanisms predict
that in the presence of 03 at low pressures -${03) should be 4 and
decline to either 3 or zero as M -*• °°. The data for 03 is very limited,
but it does show a slight downward trend with increasing pressure.
The present results can be compared to the only other study of
CCl^ photolysis at shorter wavelengths by Davis et al (120). This group
studied the photolysis at 253.7, 184.9, 147.0 and 106.7 nm. In that
study, using Br2 scavenging experiments, it was concluded that at 253.7
nm the dominant primary process is
CCU + hv (253.7 nm) -* CC13 + Cl 123'
in agreement with other studies (119). However at 184.9 nm process 123
becomes important with ${123} = 0.6 and {!23'} = 0.4. From the large
amounts of Br2 necessary to scavenge the CCla radicals it was concluded
that reaction 124 is very efficient, though no direct evidence for
reaction 124 was presented. CaClg formation via reaction 124 was
suggested but its presence was not determined.
The present results show that ${123} - 1 and
-------
116
The production of carbene in the photolysis of CCli, is not unique for
halomethanes. The photolysis of CH2l2 (121)„ CF2Br2 and CF2HBr (122)
have been shown to undergo molecular elimination reactions at around
200 nm.
Photolysis of CFC13
Very few studies of the photolysis of the chlorofluoromethanes
have been reported in the literature. Marsh and Heicklen (123) studied
the photolysis of CFC13 at 213.9 nm in the presence of NO and 02
scavengers and concluded that chlorine-atom ejection occurred with
a quantum efficiency of 1.0.
In our studies (1.24), the photolysis of CFC13 at 213.9 nm and 25°C
in the presence of 02 or 0;n gives CFC10 and C12 as products with
${CFC10} = 0.90 ± 0.15 and ${C12} = 0.50 - 0.63o In the 03 system,
-${03} increases from 2.75 at high total pressures to 4.6 at low total
pressures. These results support the claim of Marsh and Heicklen that
the dominant photochemical process is chlorine atom ejection with a
quantum efficiency near one.
The photolysis data for CFC13 in the presence of 02 is consistent
with the mechanism:
CFC13 + hv (213.9 nm) -> CFC12 + Cl 131a
-> CFC1 + C12 131b
CFC12 + 02 -> CFC1202 132
2CFC1202 -> 2CFC1.20 + 02 133
CFC120 -> CFC10 + Cl 134
'CFCl + CFC13 -> 2CFC12 135
3CFC1 + 02 -> Cl + CO + FO 136
2CFC102 -> 2CFC10 + 02 137
2C1 -> C12 138
-------
117
In the presence of 0$ reactions 132, 136, and 138 must be replaced by
CFC12 + 03 -*• CFClj.0 + 02 139
CFC1 + 03 ->• CFC10 + 02 140
Cl + 03 -> CIO + 02 10
2C10 -»• 2C1 + 02 130a
-*• C12 + 02 130b
In the mechanism the formation of F or FC1 in the primary process is
noc considered, since the bond energy of the C-F bond is much greater
than that of 'the C-G1 bond. Studies of other halomethanes show that
for the atom elimination process the bond broken is always the weakest
one (125), and for halomethanes containing more than one F atom the
stable CF2 radical is produced (117) , By analogy it seems likely that,
if carbene is produced from CFC12, it would be CFC1 rather than CCla-
First let us consider the reactions of CFC1. If the singlet CFC1,
1CIC1, is produced, as expected from spin conservation rules, then
reaction 135 would be expected by analogy to the 1Cl2-CClit system (111).
On the other hand, if triplet CFC1, 3CFC1, is produced it reacts with 02,
but not to give CFC10 (9). Thus the production of XCFC1 would tend to
promote CFC10 production, whereas production of 3CFC1 would diminish
CFC10 production. It is possible that both spin states are produced,
and their reaction processes just balance. However as we shall see
the results in the presence of 03 are inconsistent with reaction 131b
bei:ag an important process,
The reaction of CFC12 with 02 to give CFC10 can proceed via the
sequence of reactions 132-134 or via
CFC12 + 02 -*• CFC10 + CIO
-------
118
We favor the former path by analogy with the reaction of CC13 radicals
with 62 which has been shown to proceed by a sequence analogous to
reactions 132-134 (5, 13). In the presence of QZ the Cl atoms will
recombine, but in the presence of Oj they will be removed by reaction
10 (k10 = 2 x 1(T19 crnVsec) (112). The CIO radicals produced in
reaction 10 will be removed by either reactions 130a or 130b, depending
upon the total pressure. The bimolecular reaction of CIO radicals at
low pressures (<_ 8 Torr argon) is known to proceed exclusively by
reaction 130a (113), At higher pressures (>_ 70 Torr argon) reaction
130b Is the exclusive reaction (114, 115) , In the present experiments
reaction 130a could occur at the low CFC13 pressures (^ 10 Torr), but
for experiments in which the CFC13 pressure is ^ 50 Torr it should be
negligible. The reaction
CIO + 03 •* Cl + 202 or C102 + 02
can be neglected, because it is slow.
The mechanism in the presence of 63 requires that -${03} = 3 at
high pressures and -${03} _> 3 at low pressures if primary process 131a
is the exclusive reaction path. On the other hand if reaction 131b is
the exclusive reaction path, then one of three situations must occur;
1) all the CFC1 reacts with CFC13 and -${03} >_ 4, but ${CFC10> should
be 2.0,
2) all the CFC1 reacts with Q3; ${CFC10> and -${03} = 1,
3) the CFC1 reacts with both CFC13 and 03 (surely ^FCl will react with
both species) and there should be a dependence of -${03} and ${CFC10}
on the [CFC13]/[03] ratio..
The measured values of -0{03} range from 2.75 at high CFC13
pressures to 4.6 at low CFC13 pressures, but ${CFC10> = 0.90 ± 0.15
under all conditions. This clearly indicates that primary process la is
-------
119
dominant and that at low total pressure Ov 10 Torr) reaction 130a is
not negligible. Based on the mechanism the rate law for 03 removal is
-*{03} = 3 + 2k13oa/ki30b
Since at 10 Torr, -${03} = 4-4.5, then ki3oa/ki3Qb - 0.5-0.75.
The present conclusion that reaction 131a is the dominant primary
process Is consistent with the earlier work of Marsh and Heicklen (123)
who observed the formation of CFC12NO when CFCla was Irradiated in the
presence of NO indicating the formation of CFCla radicals in the primary
process.
If we ignore . reaction 131b , then the mechanism requires that
${CI'C10} = ${C12} in the presence of 02 or 03. ${CFC10} is 1 to within
the experimental uncertainty, but fCcia) < 1. The reason for the low
Gig yield is not known, but perhaps other chlorine oxides are formed.
The sane Clg deficiency was found in the CCli, photooxidation (111) and
there was some evidence for other unidentified products in that system.
In addition, f{CFC10} could be as low as 0.75, and the presence of
undetected products containing carbon and chlorine is also possible.
Photolysis of CFjCl^
As far as we know no photolysis studies of CFgCla have been pub-
lished, except for the qualitative observation that flash .photolysis
in tbe quartz O.V. produces a weak absorption due to the CFa radical
(127). In our studies with CFaCla (124), the photolysis was done at
213.3 nm only in the presence of Og . The products were CFaO and Clg
with quantum yields of 1.0 ±0.2 and 0.52-0.66, respectively.
The photolysis of CFgCla in the presence of 02 can be discussed in
term} of an entirely analogous mechanism to that for CFC13 (experiments
in the presence of 03 were not done, because of experimental difficulties)
-------
120
The question of interest is whether primary process 131a' or 131b' or both
are important,
CF2C12 + hv (213.9 nm) + CF2C1 + Cl 131a'
-* CF2 + C12 131b'
Since experiments in the presence of 63 could not be done, our data
cannot provide a definitive answer. The qualitative flash photolysis
experiments of Simons and Yarwood (126) in the quartz U.V.,, showed a
weak absorption due to CFa (by contrast CFgBr produced a strong absorp*-
tion); however it is not" clear whether this was due to the low primary
efficiency of the primary process or simply reflects the lower absorp-
tion coefficient of CF2C12, Nevertheless the participation of process
131b' to some extent is indicated.
Singlet CFj does not react with 62 at room temperature (1).
However it may react with CFaCla. Dependent on the fate of CF2S
${CF20) could vary anywhere from 0-2. In fact <£{CF20} = 1.0 ±0.1
invariant to conditions, exactly as would be expected if reaction 13la'
were the dominant process. Both reaction paths lead to the expectation
that ${CF20} = ${Cl2>. Again as was the case for CFCla, ffClz) <
§{CF20), but the reason for this is not known.
REACTION WITH 0(iD) ATOMS
The only report of the 0(1D)-CC1% reaction was by Meaburn et al
(127). They examined the gas-phase radiolysis of C02-02~CCli» mixtures
and concluded that singlet oxygen atoms react with CCli* to-give CIO
radicals. In our studies (111) we found the 0(JD) reaction with CCl^
at 25°C gives CC120 and C12 as the exclusive products. The 0(1D) was
produced from 03 photolysis at 253,7 nm. The quantum yields are
invariant to reaction conditions and are flCClaO} = 0.87 ± 0.2 and
-------
121.:
§{C12} = 1.1 ± 0,2. The 0$ consumption is the same, or slightly higher
th£;n in the absence of CCli». The three possible reaction paths are;
O^D) + CC1,, ->• CIO + CC13 141a
-»- CC120 + C12 141b
i
•»• 0(3P) + CC1% 141e
Reaction 141a was shown to be an important, and possibly the exclusive,
/
path, whereas reaction 141c Is unlmporjtant and proceeds < 20% of the
tiir,e. The overall reaction rate coefficient for reaction 141 was
measured by studying the decrease in $|cei20} in the presence of 02.
The rate coefficient for the 0(lD)~CCli» reaction relative to the
0( D)~0.2 reaction was found to be 4.0 with about a ± 10% uncertainty.
The reactions of 0(1D) atoms with the chlorofluoromethanes have
alsa not been extensively studied. Clerc (128) has observed CIO pro-
duction in the flash photolysis of 03-CF3C1 mixtures, indicating that
DC1!}) abstracts the Cl atom. Recently, since our work was completed,
a report by Pitts et al (129) was published which gives rate coefficients
for several chlorofluoromethanes, including CF2Cl2 and CFCla, obtained
by competitive methods relative to the reaction with N20.
In our studies (124) on the reactions of the chlorofluoromethanes
with 0(1D), prepared from the photolysis of 03 at 253.7 nm and 25°C,
the same products are obtained as in the photooxidation, and with the
same yields. The quantum yields of 03 removal are 5.7 ± I and 6.3 ± 1,
respectively for the CFC13 and CF2C12 systems. Thus the indicated
dominant reaction path is chlorine atom abstraction by 0(1D), with other
path« (0(1D) deactivation or direct molecular formation of products)
N
being negligible.
\
-------
122
Rate coefficients were obtained for the Q^D) reactions with 03,
CO2s CFC13, CF2Cl2, CF3C1, and COU relative to N20. The relative rate
coefficients are given in Table XV. The rate coefficients were also
measured for the first five gases in the presence of He to remove the
excess translational energy of the 0(1D) atom. Except for Os, the
same results were obtained in the presence and absence of He. However
for 03 in the presence of He, the relative rate coefficient was 1.6.
As a check on the reactivities of 0(1D) with the chlorofluoro-
methanes, the competition of 02, rather than N20S was studied in the
03-02-chlorofluoromethane system (124). By monitoring 03 decay the
relative rate coefficients for CFC13> CF2C12, and CCl^ relative to 02
were found to be 4.04, 2.78, and 5,3 respectively. These results are
consistent with those obtained from the N20 competition.
CHLORINE-ATOM SENSITIZED OXIDATION OF CH2C12 and CH3C1
Mixtures of C12, 02, and either CH2Cl2 or CH3C1 were irradiated
at 3655A and 32°C (22). The C12 photodissociates and CHC12 or CH2C1
radicals are produced via hydrogen abstraction from the corresponding
chlorinated methane. In the CH2C12 system, there is a long chain
process, the initial products are HCl, CHC10S CC120, and possibly CO.
${CHC10} = 49 and ${CC120} =4.1 independent of CH2Cl2 or 02 pressure
or la.
The main chain sequence appears to be analogous to that for
oxidation
C12 + hv -»• 2C1 1
Cl + CH2C12 -*• CHC12 + HCl 143
CHC12 + 02 ->CHC1202 144
2CHC1202 -> 2CHC120 + 02 ,l45a
-------
123
Table XV
Summary of measured and literature values of the rate coefficient
for 0(1D) reactions, k{x}, relative to that for N20, k{N20}
X
03
03
CO 2
CO 2
CPC13
CFC13
CF2C12
CF2C12
CF3C1
CF3C1
CCU
He,
Torr
-
400-500
-
400
_
400-550
-
72-420
-
500
—
k{x}/k{N20}
This work
2.5
1.6
0.65
0.65
1.5
1.5
1.2
^1.4
0.52
0.52a
2.1
k{X>/k{N20}
Literature
2-3 (109); 2.6 (130); 2.2 (131)
_
0.82 (131); 0.55 (132); 0.80 (133)
-
2.6 ± 0.5 (129)
-
2.4 ± 0.5 (129)
-
_
_
_
a) Based on one point.
-------
124
CHC120 -*• CHC10 + Cl 146
The two features that are not obvious are (1) how is CC120 produced?
and (2) how are the chains terminated?
The reactions of termination must involve two radicals. One of
these cannot be CHClaO, for then there would be an intensity dependence
on the quantum yield. Thus we propose
2CHC1202 -> (CHC120)2 + 02 145b
(CHC120)2 -+• CHC10 + CC120 + HC1 147a
-»• CC120 + H20 + CC12 147b
where reaction 147a produces CClaO by termination, but reaction 147b
gives CClaO by chain propogation since it is known (36) that CCiz
reacts with 02 via
CC12 + 02 ->• CIO + C1CO •*• Cl + CO 148
and the predominant fate of CIO is
2C10 ->- 2C1 + 02 149
Then part of the CO would come from reaction 148 and part from CHC10
decomposition.
In the CH3C1 system, the initial products are exclusively HC1 and
CHC10, the quantum yield of the latter being 2.0 independent of reaction
conditions. Thus in this system there is no chain at all, and the
mechanism is similar to that for CHs oxidation, except that no alcohol
is produced. In both systems the CHC10 produced is removed by chlorine
atom attack
Cl + CHC10 -* HC1 + C1CO •* Cl + CO 142
with km2 = 7 x 108 M'1 sec™1.
-------
125
REFERENCES
1. J. Helcklen, Adv. Photochem. . 7_, 57 (1969).
2. M. J. Molina and F, S, Rowland, Nature (London), 249_, 810 (1974).
3, R. J. Cicerone, R, S, Stolarski, and S. Walters, Science. 185. 1165 (1974).
4. R. J. Crutzen, Geophys . Res^. Lettr . .. !_, 205 (1974).
5. E. Mathias, E. Sanhueza, I. C. Hisatsune, and J. Heicklen, Can. J . Chem. ,
52, 3852 (1974).
6. E. Sanhueza and J. Heicklen, "The Chlorine-Atom Initiated and Hg(3Pj) -
Photosensitized Oxidation of CH2CCl2," Center for Air Environment Studies
Report No. 353-74, Penn State University (1974), J. Photochem.. in press.
7, E. Sanhueza and J. Heicklen, "The Oxidation of cis- and trans-CHClCHCl,"
Center for Air Environment Studies Report No. 364-74, Penn State University
(1974), int. . ............. J . ..... Chem . Kinej:ic£. in press.
8. E. Sanhueza and J. Heicklen, "The Oxidation of CHC1CH2," Center for Air
Environment Studies Report No. 371-74, Penn State University (1974),
J. Phys . Chem.^ in press .
9. E. Sanhueza and J. Heicklen, "The Oxidation of CFC1CFC1 and CF2eei2. ,"
Center for Air Environment Studies Report No. 365-74, Penn State University
(1974), Int.= =J. Chem. Kinetics^ in press.
10. R. G. Dickinson and J. A. Leermakers, J. Am. Chem. Soc. . 5,4. 3852 (1932).
11. R. G. Dickinson and J. L. Carrico, J. Am. Chem. Soc._. 56.. 1473 (1934).
12. C. Sehott and H. J. Schumacher, Z. Physik. Chem.. B49. 107 (1941).
13. G. Huybrechts, J. Olbregts, and K. Thomas, T r ans . Farad . Soc . . 63, 1647
~~~
14. K. L. Muller and H. J. Schumacher, Z. Physik. Chem.. B37. 365 (1937).
15. G. Huybrechts and L, Meyers, Trans_. Farad. Soc. . 6.2. 2191 (1966).
16. L. Bertrand, J. A. Franklin, P. Goldfinger, and G. Huybrechts, J. Phys. Chem...
21, 3926 (1968),
17. K. L. Muller and H. J, Schumacher, Z. Physik. Chem.. B35. 455 (1937).
18. E. Sanhueza, unpublished results, Penn State University (1974).
-------
126
19, G. Huybrechts, G. Martens, L. Meyers , J< Olbregts, and K. Thomas,
Trans. Farad. Soc.. 61., 1921 (1965) .
20. H. J. Schumacher and H. Thiirauf, j^^Phyj^ Chenu. A189 . 183 (1941).
21. R, T, Watson, Chemical Kinetics Information Center, National Bureau of
Standards, Report No, NBSIS. 74-516 (1974), "Chemical Kinetics Data Survey
VIII. Rate Constants of C10X of Atmospheric Interest."
22. E, Sanhueza and J, Heicklen, J. Phys. Chem.,. 79. 7 (1975).
23. J. A. Franklin, G. Huybrechts, and C. Cillien, Trans . Farad . Soc ... . 65_,
2094 (1969).
24. L. Bertrand, L. Exs teen-Meyers, J, A, Franklin, G, Huybrechts, and
J. Olbregts, Int. J. Chem. Kinetics. _3, 89 (1971).
25. P. Goldfinger, G, Huybrechts, G. Martens, L. Meyers, and J. Olbregts,
Trans. Farad. Soc.. 61, 1933 (1965).
26. W. T. Miller, Jr, and A. L. Dittman, J_.__Am._ Ghent. Soc . . 7.8, 2793 (1956).
27. E. Sanhueza and J. Heicklen, Can. J. Chem. . 5,2, 3863 (1974).
28. E. Sanhueza and J. Heicklen, "The Hg 6(3P) Photosensitized Oxidation of
CaClaH," Center for Air Environment Studies Report No. 321-73, Penn State
University (1973),
29. R. J. Cvetanovic, Prog. React. Kinetics, 2,, 39 (1964).
30. a) 0. P. Strausz0 R. J. Norstrom, D. Salahub, R. K. Gosani, H. E. Gunning,
and I. G. Csizmadia, J _._ L Am . _ Chem v Sjj_c._, 92., 6395 (1970); b) 0. P. Strausz,
private communication (1974),
31. B. Atkinson, Nature, 163, 291 (1949).
32. B. Atkinson, J. Chem. Soc_._, J-952, 2685.
33. J. Heicklen, V. Knight, and S. A. Greene, J_. Chem. Phys.. 42. 221 (1965).
34. J. Heicklen and V. Knight, J. Phys. Chem.. _7p_» 3901 (1966).
35. J. Heicklen, J. Phys. Chem.,, _7p_, 618 (1966).
36. E. Sanhueza and J. Heicklen, Can.i: ..... J,., Chem,_, 52. 3870 (1974).
37. E. Sanhueza and J. Heicklen, Int. ............ J, Chem. Kinetics. j6, 553 (1974).
38. E. Sanhueza and J. Heicklen, "The Reaction of 0(3P) with CC12CH2,"
Center for Air Environment Studies Report No. 360-74, Penn State University
(1974), J. Photochem. , in press.
-------
127
39, R. J, Cvetanovic, Adv. Phptpcheg.. !_, 115 (1963),
40. S.-I. Hirokami and R, J, CvetanoviS, Can._ J. Chem.. 51, 373 (1973).
41. R. J. Cvetanovic, Can. J._ Chem.. 33, 1684 (1955).
42. R, J. Cvetanovi£, J. Ghentv Phys.. _23, 1375 (1955).
43. J. R. Kanofsky and D, Gutman, Chem.Phys. Lettr.. 15, 236 (1972).
44. F. J. Pruss, Jr., I, R, Slagle, and D, Gutman, J. Phys. Chem..., 78. 663
(1974).
45, I. R. Slagle, D, Gutman, and J. R. Gilbert, Chem. Phys. Lettr., 26. Ill
(1974).
46,, D. Saunders and J. Heicklen, J. Am. Chem. Soc.. 87. 2088 (1965).
47, R. C. Mitchell and J, P. Simons, J. Chem.. SQC._. (B) 1968. 1005.
48, R. E. Huie, J. T. Herron and D. D, Davis, Int. J. Chem.Kinetics. j4, 521
(1972).
49. S. J. Moss, Trans. Farad. Soc.. jS7, 3503 (1971).
50. W. J. R. Xyerman, Trans._Farad. Soc.. 65, 163 (1969).
51. R. N. Has2;eldine and B. R. Steele, J^Chenu. Soc.,. 1954. 3447.
52. D. P, Johari, H. W, Sidebottom, J, M, Tedder,' and J. C. Walton, J. Chem. Soc..
(B) 1971. 95.
53. R. Atkinson and R. J. Cvetanovic, J. Chem. Phys.. 5,5, 659 (1971).
54. D. D, Davis, R. E, Huie, J. T. Herron, M. J. Kurylo, and W. Braun,
J. Chem. Phys.. 5j6, 4868 (1972).
55. F. Stuhl and H. Niki, J. Chem. Phys.. 5_7_, 5403 (1972).
56. M. J. Kurylo and R. E, Huie, J. Chem. Phys.. 5_8, 1258 (1973).
57. I. R. Slagle, F. J, Pruss, Jr., and D. Gutman, Int.J. Chern^ Kinetics. £,
111 (1974).
58. R. Atkinson and J. N. Pitts, Jr., Chem. Phys. Lettr.. 2.7., 467 (1974).
59. R. Atkinson and J. N. Pitts, Jr., J. Phys. Chem.. 2i> 1780 (1974).
60. W. A. McClenny, J. Chem. Phys.. 60, 793 (1974).
61. E. A. Lissi, private communication (1973).
62. R. F. Lake and H. Thompson, Proc. Roy. Soc., A^ ,315, 323 (1970).
-------
128
63. J, A, Kerr and M. H. Parsonage, "Evaluated Kinetic Data on Gas Phase
Addition Reactions," CRC Press, Cleveland, Ohio (1972), pp. 101-102,
64. W. Kirmse, "Carbene Chemistry," Academic Press (1964), pp. 20-21.
65. D. D. Davis, J. F. Schmidt, C. M. Neeley, and R. J. Hanrahan, J. Phys... Chem.
79_, 11 (1975).
66. R, J, Gordon and M, C, Lin, Chem. Phys,.. Lettr.. 22., 107 (1973).
67. J. Heicklen and ?. Knight, Jj,_Ph^s_.^Chem._t 70.* 3893 (1966).
68. J. Heicklen, J^Jhys. Chem.|t 70:, 477 and 4101 (1966).
69. D. G. Williamson and R, J. Cvetanovle", J. Am._ Chem. Soc.t 90, 4248 (1968).
70. L. A. Hull, I.C. Hisatsune, and J, Eleicklen, Can. J. Chem., 51., 1504,
(1973).
71. C. W. Blume, I. C, Hisatsune, and J, Heicklen, "Gas Phase Ozonolysis of cis-
and trans-Dichloroethylene," Center for Air Environment Studies Report No.
385-75, Penn State University (1975).
72. B. J. Finlayson, J. N. Pitts, Jr., and H. Akimoto, Chem. ^ _P hy_£. _ I,e rtr.,
12, 495 (1972).
73. B. J. Finlayson, J. N,, Pitts, Jr., and R. Atkinson, J..,.. Am. Chem. Soc., 96_,
5356 (1974).
74. R. D. Cadle and C. Schadt, J. Chem. Phys.. 21, 163 (1953).
75. P. L. Hanst, E. R. Stephens, W. E. Scott, and R. C. Doerr, Am. Chem. Soc,
Nat'l. Meeting, Symp. on Air Pollution Research, Division of Petroleum
Chemistry (Sept., 1959).
76. T. ¥rbaski and R, J. Cvetanovic, Can. J. Chem.. 38. 1053 (1960).
77. T. Vrbaski and R. J. Cvetanovic, Can. J. Chem., _38, 1063 (1960).
78. Y. K. Wei and R. J. Cvetanovic, C_an_._ J. Chem., 41, 913 (1963).
79. J. J. Bufalini and A. P. Altshuller, Can._ J. Chem.. 43, 2243 (1965).
80. W. B. DeMore, Int. J._ Chem. Kinetics. 1., 209 (1969).
81. R. Atkinson, B. J. Finlayson, and J. N. Pitts, Jr., J. Am. Chem. Soc.9
95_, 7592 (1973).
82. D. H. Stedman, C. H. Wu, and H. Niki, J_., Phys. Chem.. 7_7_, 2511 (1973).
83. J. T. Herron and R. E, Huie, J. PJiy_s^m£hem_.., 78, 2085 (1974).
84. D. G. Williamson and R. J. Cvetanovic, J. Am. Chem. Soc,., 9_0, 3668 (1968).
85. C. W. Gillies and R. L. Kuczkowski, J. Am. Chem.^Soc^, 94., 6337 (1972).
86,. C. W. Gillies and R. I,. Kuczkowski, J^ Am. Chem. Soc.. _94, 7609 (1972).
-------
. 129
117. C, W. Gillies, R. P. La t timer, and R. L. Kuczkowski, J. Am. Chem. Soc.,
96,, 1536 (1974),
£8. L. A. Hull, I, C, Hisatsune, and J. Heicklen, J. Am. Chen. Soc., 94.
4856 (1972).
89. L. A. Hull, I. C. Hisatsune, a*"i J. Heicklen, £f Phys . Chem., 76. 2659
(1972).
90. D. Garvln and C. Schubert, J L Phys f ; Chem. . 60. 807 (1956).
91. ^P. Gozzo and G, Cainaggi, Chim. Ind.. 50, 197 (1968).
92." I. C. Hisatsune, L. H, Kolopajlo, and J. Heicklen, unpublished work,
Penn State University (1975).
93. I. C. Hisatsune and J, Heicklen, Caxnv.J.rSpecj;jgyi..» ii> 77 (1973).
94. L. H, Kolopajlo, unpublished work, Penn State University (1975).
9i>. R. P. Lattimer, R. L, Kuczkoweki, and C. W. Gillies, J, Am. '..Chen, Soe. .
96., 348 (1974).
96. R. 'Criegee and G. Werner, Justus ^iebiffs Ann. Chem, . 546 . 9 (1949).
97. H. 1. O'Neal and C. BluMtein, Int^^ jTtV. -Clxeffi.... Kinetics. 5,, 397 (1973).
96;. C. E. Bishop and R. P. Story, J . A^'.^GhjiB.;... Ag&{_' S2.* I905 (1968).
99. P. R. Story, G. E. Bishop, J, R. Burgess, R. W. Murray, and R. D.
Youssefyeh, J. Am, j Chem. Sgc.. 90, 1907 (1968).
100. P. R. Story, J. A. Alford, J. R. Burgess, and W. C. Ray, J. Am, Chettt. Soc,.
93, 3042 (1971).
101. P. R. Story, J, A. Alford, W, C, Ray, and J. R. Burgess, JA, iAm.i Chem, Soc . .
91, 3044 (1971).
102, H. Staudinger, Chem. Ber . . 58, 1088 (1925).
103, P. R. Story, G. E. Bishop, J. R. Burgess, J. B, Olson, R. W. Murray, and
R. D. Youasefyeh in '"Oxidation of Organic Compounds, Vol. 3," Adv. Chem.
Series, 77,, American Chemical Society, Washington, IT. C., 1968, pp. 46-57.
104, P. S. Bailey, J, W. Ward, R. E. Hornish, and F, E. Potts, III, in "Ozone
Reactions with Organic Compounds," Adv, Chem. Series 112, American
Chemical Society, Washington, D. C., 1972, pp. 1-8.
105. J. Carles and S. Fliszar, in "Ozone Reactions with Organic Compounds,"
Adv. Chem. Series 112, American Chemical Society, Washington, D. C.,
1972, pp. 35-49.
106. H. Niki, E. S. Daby, and B. Weinstock in "Photochemical Smog and" Ozone
Reactions, " Adv. Chem. Series 113, American Chemical Society, Washington,
D. C., 1972, pp. 16-57.
-------
130
107. G. S. Forbes, I, E, Cline, and B, C, Bradshaw, J. Am. Chem. Soc.,
60, 1413 (1938),
108. R. Sinionaitis, R. I. Greenberg, and J. Heicklen, Intern. J. Chem.
Kinet.. 4., 497 (1972),
109. E. Lissi and J. Heicklen, J. Photochem._. 1_, 39 (1972).
110. R. Simonaitis and J. Heicklen, J. Phys_.:i Chem.. 78, 653 (1974).
111. 1. K. M. Jayanty, R. Simonaitis, and J. Heicklen, J. Photochem.,
in press.
112. R. T. Watson, National Bureau of Standards Report, NBSIR 74-516
"Chemical Kinetics Data Survey VIII" (1974).
113. M. A. A. Clyne and I. F. White, Trans. Farad. Soc., j>7., 2068 (1971).
114. G. Porter and F. J. Wright, Disc. Farad. Soc.. JL4, 23 (1953).
115. N, Basco and N. K. Dogra, Proc. Roy. Soc., AJ323. 29, 401 (1971).
116. J. Doucet, P. Sanvegeau and C. Sandorfy, J. Chem. Phys., _58_,
3708 (1973).
117. J. R. Majer and J. P. Simons, Ad vvPhoJ:ochem.a, _2, 137 (1964).
118. M. L. White and R. R. Kuntz, Internv J. Chem. Kinet.. 5_, 187 (1973).
119. J. Currie, J. Sidebottom and J. Tedder, Intern. J. Chem^^Kinet^.^ _§_,
481 (1974); earlier references are given, in this paper.
120. D. D. Davis, J. F. Schmidt, C. Neeley and R. J. Hanrahan, J_VrPlvys.
Chem., _7JL> 11 (1975).
121. G. W. G. Style, and J. C. Ward, J_._ Chem^Jioc^, 2125 (1952).
122. J. R. Majer and C. R. Patrick, .Nature., 192., 866 (1961).
123. D. Marsh and J. Heicklen, J. Phys.v Chem,. 69,. 4410 (1965).
124. R. K. M. Jayanty, R. Simonaitis, and J. Heicklen, submitted for
publication (1975).
125. J. N. Pitts, Jr., and J. G. Calvert, "Photochemistry" John Wiley &
Sons, Inc., New York, (1967) p. 522.
126. J. P. Simons and A..J. farwood, Trans A Farad.; Soc.. 5_7, 2167 (1961).
127. G, M. Meaburn, D. Perner, J. LeCalve, and M. Bourene, J. Phys.Chem.,
72, 3020 (1968).
-------
131
123. M, Clerc, Compt. Rend.. 260. 2189 (1965).
129. J. N. Pitts, Jr.s H, L. Sandoval and R. Atkinson, Chem. Phys. Letters,
29, 31 (1974).
130. c. S, Goldman, R, I. Greenberg and J. Heicklen, Intern. J. Chem.
Kinet.. .3, 501 (1971).
13],. D. Garvin and R. F. Hampson, National Bureau of Standards Report,
NBSIR 74--430 "Chemical Kinetics Data Survey VII" (1974).
132. p. M.Scott and R. J. Cvetanovic, J. Chem. Phys., 54., 1440 (1970).
133. G. Paraskevopoulos, K. F. Preston and R. J. Cvetanovic, J. Chem.
Phys.. 54, 3907 (1971).
-------
132
LIST OF PUBLICATIONS
1. L. A. Hull, I. C. Hisatsune, and J. Heicklen, Can. J. Chem., 51,
1504 (1973), "The Reactions of 03 with CC12CH2."
2. E. Sanhueza and J. Heicklen, Intern. T. Chem. Kinet . . 6_, 553 (1974).
"The Reaction of 0(3P) with C2HC13."
3. E. Mathias, E, Sanhueza, I, C. Hisatsune, and J. Heicklen, Canad. J.
Chem. , _52, 3852 (1974) "The Chlorine-Atom Sensitized Oxidation and
Ozonolysis of
4. E. Sanhueza and J. Heicklen, Canad. J. Chem.. 52, 3863, "The Hg 6(3P)-
Photosensitized Oxidation of
5. E. Sanhueza and J. Heicklen, Canad. J. Chem. , 5.2, 3870 (1974), "The
Reaction of 0(3P) with C2C1,,."
6. E. Sanhueza and J. Heicklen, J. Phys . Chem., 79_, 1 (1975), "The
Chlorine-Atom Sensitized Oxidation of CH2Cl2 and CH3C1.
7. E. Sanhueza and J. Heicklen, Intern. J. Chem. Kinet., in press,
"The Oxidation of CFC1CFC1 and CF2CC12."
8. E. Sanhueza and J. Heicklen, J. Photochem.,, in press, "The Reaction
of 0(3P) with CC12CH2."
9. E. Sanhueza and J. Heicklen, J . Photochem. , in press, "The Chlorine-
Atom Initiated and Hg 6( 3P] )-Photosensit:ized Oxidation of CH2CC12."
10. E. Sanhueza and J. Heicklen, Intern. J. Chem. Kinet., in press, "The
Oxidation of cis- and trans-CHClCHCl."
11. E. Sanhueza and J. Heicklen, J, Phys. Chem., in press, "The Oxidation
of c2H3ci." ;
12. E. Sanhueza and J. Heicklen, J. Photochem. , in press, "The Hg 6(3P)-
Sensitized Photooxidation of C2C13H."
13. Gary W. Blume, I. C. Hisatsune, and J. Heicklen, "Gas Phase Ozonolysis
of cis- and trans-Dichlorethylene," CAES Report No. 385-75, The
Pennsylvania State University.
14. R. K. M. Jayanty, R. Simonaitis, and J. Heicklen, J. Photochem. ,
in press (1975), "The Photolysis of CC1,, in the Presence of 02 or
03 at 213.9 nm, and the Reaction of O^D) with CCli,."
15. R. K. M. Jayanty, R. Simonaitis, and J. Heicklen, to be published,
"The Photolysis of Chlorofluoromethanes in the Presence of 02 or
03 at 213.9 nm and Their Reactions with 0(1D)."
-------
133
16. I.C. Hisatsune and J. Heicklen, Canad. J. Spectry... 18, 77,
(1973), "Infrared Spectrum of Formyl Chloride."
17. I.C, Hisatsune and J. Heicklen, Canad._J:.i Spectry,., JL8, 135,
(1973), "Are There Two Structural Isomers of Formic Acid?"
-------
134
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA--650/3-75-008
2.
4, TITLE AND SUBTITLE
The Oxidation of Halocarbons
3. RECIPIENT'S ACCESSION-NO,
5. REPORT DATE
May 1975 (date of approval)
B, PERFORMING ORGANIZATION CODE
7.AUTHOR(s) j. P> Heicklen, E. Sanhueza, I. C. Hisatsurie,
R, K. M. Jayanty, R. Slmonaitis, L. A, Hull,
C. W. Blume, and E. Mathias
8. PERFORMING ORGANIZATION REPORT NO.
9, PERFORMING ORG "\NIZATION NAME AND ADDRESS
Center for Air Environment Studies
226 Fenske Lab,
The Pennsylvania State University
University Park, Pa., 16802
10, PROGRAM ELEMENT NO.
1A1008, ROAP 26AAD-20
11. CONTRACT/GRANT NO.
Grant No. 800949
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Protection Agency
Research Triangle Park, N. C., 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final Report-6/1/72-5/31/74
i14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
Presented before the Division of Environmental Chemistry, American Chemical Society
April 9, 1975, 169th National Meeting, Philadelphia, Pa.
16. ABSTRACT
The gas-phase room-temperature oxidation of haloethylenes was studied. In
general oxidation has been carried out in five ways: 1) chlorine atom initiations
2) Hg 6(3P) sensitization, 3) reaction with 0(3P), 4) reaction with 0(3P) in the
presence of 02, and 5) reaction with 03. In the first four systems the major pro-
ducts are the corresponding carbonyl chlorides containing 1 or 2 carbon atoms, and
the reaction proceeds by a long-chain free radical process. With Os a diradical
chain Is involved which is inhibited by QZ.
Free radical attack of CHaCla or CHsCl in the presence of Oa gives carbonyl
halides, as does the photolysis of CClit, CClgFj and CC12F2 in the presence of 02
or Os. CClij and the chlorofluoromethanes react with 0(1D) via chlorine atom
abstraction in reactions with large rate coefficients which are nearly proportional
to the number of chlorine atoms in the chlorofluoromethane.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Halocarbons
Photochemistry
Radical Reactions
Gas Phase Kinetics
Spectroscopic Identification
b.IDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
S.-DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
151
20. SECLIRITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (3-73)
------- |