U.S. DEPMTMENT OF COMMERCE
National Ttchnical literatim Stnto
PB-284 066
Study of Removal
Processes for Halogenated
Air Pollutants
III Research Inst, Chicago,
Prepared for •
Environmental Sciences Research Lab, Research Triangle Park, N C
Jun 78
-------
Environmental Protection
Agency
Laboratory
"Research Triangle Park NC 27711
Research and Development
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine 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 (STAR)
7. InteVagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series. This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences. Investigations include formation, transport, and pathway studies to deter-
mine 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 through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED
FROM THE BEST COPY FURNISHED US BY
THE SPONSORING AGENCY. ALTHOUGH IT
IS RECOGNIZED THAT CERTAIN PORTIONS
ARE ILLEGIBLE, IT IS BEING RELEASED
IN THE INTEREST OF MAKING AVAILABLE
AS MUCH INFORMATION AS POSSIBLE.
-------
TECHNICAL REPORT DATA
(I'lt-asc read luuriK-lionx on the reverse before completing)
I. III I'OHI NO. 2.
EPA-600/3-78-058
.1. in LI AMU sum n 1.1
STUDY OF REMOVAL PROCESSES FOR HALOGENATED
AIR POLLUTANTS
7. AUTHORIS)
A. Snelson, R. Butler, and F. Jarke
6. REPORT DATE
June 1978
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
TIT Research Institute
Chicago, Illinois 60616
10. PROGRAM ELEMENT NO.
1AA603
11. CONTRACT/GRANT NO.
R803805-02-2
t
12. SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory- RTF,
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park. North Carolina 27711
13. TYPE OP REPORT AND PERIOD COVERED
" ,'"• I -
7/7 — • /77
' ' '
14. Si
ING AGENCY
EPA/60Q/Qq
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The fate of halocarbons released into the atmosphere from anthropogenic sources
and their potential impact on the stratospheric ozone layer has been studied.
Experimental kinetic data have been obtained for the reaction of OH radicals with
Cl,
C1-, CtCCl
C H Br , and CO. Bestvalues
en assessea and tropospheric life-
CH3C1, CHt C12, CHC1 , 2 , j -, ,
for the OH + halocarbon rate constants have bee
times calculated.
The kinetics of the homogeneous gas phase hydrolysis of the secondary anthro-
pogenic halocarbon pollutants, CClgCOCl, CC12HCOC1, CC1H COC1, and COC12 have been
studied. The rate studies show that removal by gas phase hydrolysis would have
half lives in excess of 100 years. Some preliminary heterogeneous rain out studies
have been made on CC13COC1, which though not conclusive, suggest that rain out
would probably be an effective sink for its removal but that the uncertainties
are such that this mechanism requires further study.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air pollution
Halohydrocarbons
Reaction kinetics
Chemical reactions
Chemical radicals
Half life
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
h.IDENTIFIERS/OPEN ENDED TERMS
Removal processes
19. SECURITY CLASS (This Report)
UNCLASSIFIED
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
c. COSATI Field/Group
13B
07C
07D
18H
21. NO
22. PRICb
EPA Form 2220-1 (»-73)
I
-------
EPA-600/3-78-058
June 1978
STUDY OF REMOVAL PROCESSES FOR HALOGENATED AIR POLLUTANTS
by
A. Snelson
R. Butler
F. Jarke
IIT Research Institute
Chicago, Illinois 60616
Grant No. R803805-02-2
Project Officer
P. L. Hanst
Atmospheric Chemistry and Physics Division
Environmental. Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
\OJ
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences
Research Laboratory, U. S, Environmental Protection Agency, and
approved for publication. Approval does not signify that the -
contents necessarily reflect the views and policies of the H
U. S. Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
-------
ABSTRACT
This study is concerned with the fate of some chlorinated
compounds, released into the atmosphere from anthropogenic
sources, and their potential impact on the stratospheric ozone
layer. The study may be divided conveniently into three parts:
(1) An investigation of hydroxyl radical reaction kinetics
with CH3C1, CH2G12, CHC13> C2H5C1, C2H4C12> CH3CC13, C2C14,
C2H,Br2 and CO, under conditions closely resembling those found
in the atmosphere, using a competitive rate method. Best values
for the OH + halocarbon rate constants have been assessed based
on the presently derived data and that existing in the litera-
ture. Tropospheric lifetimes have been calculated, which, with
the possible exception of CH3CC13> indicate that none of these
compounds will enter the stratosphere as such in significant
amounts. CH3CC13 does not at present represent a problem, but
could, if anthropogenic releases kept increasing give reason for
concern. The previously unsuspected pressure dependence of the
OH + CO reaction in air-like mixtures has been established.
This knowledge will be of great value in better understanding the
chemistry of polluted atmospheres.
i
(2) An investigation of the homogeneous gas phase hydrolysis
kinetics of the secondary anthropogenic halocarbon pollutants
CCloCOCl, CCl-HCOCl, CC1H9COC1 and COC17. These data were ob-
tained using a static reactor in the temperature range 470-620 K.
The derived rate constants were used to determine tropospheric
half lives for these species against hydrolysis, In all case
values in excess of 100 yrs resulted indicating that in the
absence of other removal mechanisms, large quantities of these
materials would enter the stratosphere.
iii
-------
(3) An investigation of the heterogeneous "rain out" rate-of. .
the secondary anthropogenic halocarbon pollutants noted above,
A simple model for the atmospheric "rain out" of these species
was constructed based on gas collision kinetic theory. This
model requires the determination of an effective rain drop-acid
chloride sticking coefficient. Some preliminary experimental
data were obtained for CCUCOCl which indicated that "rain-out"
would probably be an effective removal mechanism for this species
from the atmosphere. However, the preliminary nature of the data,
the surprisingly low value of the sticking coefficient obtained,
and the admitted fact that insufficient experimentation was
made to determine the true effective sticking coefficient for
atmospheric application, strongly suggest that further studies
should be made before firm conclusions can be obtained with
respect to the effectiveness of urain-out" as an atmospheric
removal mechanism for these species.
Recommendations for future studies and the fate of these
halocarbon acid chlorides are made.
This report was submitted in fulfillment of Grant No.
R803805-02-2 by IIT Research Institute under the sponsorship of
the U. S. Environmental Protection Agency. This report covers
the period July 1, 1975 to December 31, 1977, and the work was
completed December 31, 1977.
-------
CONTENTS
Abstract iii
Acknowledgment v^
1. Introduction 1
2. Conclusions 8
3. Recommendations 10
4. Pressure Dependence of the CO + OH Rate
Constant in 02 + N2 Mixtures 12
Introduction 12
Experimental 13
Results 18
5. Rate Constants for the Reactions of OH With
CHiCl, CHoClo, C2H5C1, C1CH2CH2C.1, CH3CC13,
C2H4Br2, CHCTU and C2C14 in the Presence of
02 + N2 26
Kinetic Results Using the Static Reactor . 26
Introduction 26
Experimental 27
Results and Discussion . 28
Kinetic Results Using the Dynamic Reactor, 40
Introduction 40
Apparatus 43
Experimental Method 43
Results 46
Discussion of Results . 53
Implications With Respect to Atmospheric
Lifetimes of the Kinetics of the OH +
Halocarbon Reactions 56
Introduction 56
Evaluation of the Kinetic Data ... 56
Tropospheric OH Radical Concentration 57
Halocarbon Tropospheric Lifetimes . . 58
6. Kinetics of the Homogeneous Gas Phase
. Hydrolysis of CC13COC1, CHCUCOC1,
CH2C1COC1 and COC12 .... 7 60
Introduction 60
Preliminary Studies - Static Reactor ... 60
Preliminary Studies - Dynamic Reactor . . 63
Experiments With the Large Static
Reactor 66
Experimental 66
Results and Discussion 73
-------
Results and Discussion 73
Implications With Respect to Atmospheric
Lifetime of the Homogeneous Gas Phase
Hydrolysis Data 84
7. Kinetics of the Heterogeneous Hydrolysis or
"Rain-Out" of the Acid Chlorides 88
Introduction 88
The Simple Rain-Out Model 89
Experimental 91
Results and Discussion 95
Implication of "Rain Drop" Experiments
With Respect to Atmospheric Wash Out of the
Acid Chloride . 99
References 103
vi
-------
ACKNOWLEDGMENTS
The following IITRI personnel have made contributions to
the program, Dr. A. Snelson, Dr. I. J. Solomon, Dr. M. Lustig,
Mr. R. Butler, Mr. F. Jarke and Mr. L. Presta. Dr. D, Gutman
of IIT provided valuable advice with respect to the OH radical
kinetic studies, particularly those of Mr. F. Jarke, which were
made in part fulfillment of a Ph.D. degree.
vli
-------
SECTION 1
INTRODUCTION
Interest in the fate of halogen compounds introduced into
the atmosphere from anthropogenic and natural sources has been
stimulated by their theorized potential threat to the stratos-
1-5
pheric ozone layer. Initially the major concern was directed
at the "Freons," chlorofluoromethanes widely used as aerosol
propellants and refrigerants. These compounds apparently have
no atmospheric sink, they diffuse to the higher regions of the
atmosphere (stratosphere) where they are dissociated by short
wavelength radiation (190-220 nm) to yield atomic chlorine and
fluorine. The major problem lies with the atomic chlorine which
can take part in the catalytic decomposition of ozone, via the
series of reactions:
Cl + 03 + CIO + 02
03 + hv ->• 02 + 0 Net 20~ = 302
CIO + 0 •» Cl +' 02
Although t;here are some sinks for the chlorine in the stratos-
pheric (e.g., HC1) its removal in this form by normal diffusion
process is slow. If there are large fluxes of chlorine bearing
compounds to the stratosphere, there appears to be a real danger
that their accumulation could result in a significant reduction
in the earth's ozone layer. Eventual reduction of the ozone
layer by some 10-15% has been postulated within 20-30 years if
the scale of present emissions of Freons continue. The actual
consequence of such an ozone reduction would be to increase the
amount of UVB radiation (290-310 nm) the earth receives. The
-------
precise result of this increased UV radiation on life is not
clearly defined but increases in skin cancer, plant life -damage
and possible adverse chemical effects have been suggested.
As noted above, the initial concern was with the Freons,
however, it was shortly thereafter that the potential danger
associated with anthropogenic releases of halocarbons into the
atmosphere was also realized, In an Environmental Protection
Agency study, the size and nature of the anthropogenic
emissions of halogen compounds was tabulated, A summary of
these data is presented in Table 1. It was pointed out that the
only known sink for most of these compounds in the troposphere -•
is attack by OH radicals which exist in the atmosphere at
C ^O Q
a concentration of 2-6 x 10 molecules cm" . If the reactions
with OH are fast enough, then most of the compounds will be
destroyed before they have chance to diffuse into the
stratosphere and suffer photolysis with subsequent effect on
/
the ozone layer.
Table 2 shows the state of knowledge at that time with
respect to potential destruction of the halocarbons in the
atmosphere and the expected degradation products. Estimated
half-lives for these compounds were based on the then known
rate constants for the halocarbon + OH reactions or on estimated
values. In general OH + unsaturated halocarbon reaction rates
are sufficiently rapid (bimolecular rate constants usually
-12 3 -1 -1
> 5 x 10 cm molecules sec . A notable exception is
-13 3 -1 .iIO.lI
CC12CC12 for which K ^2 x 10 cnT molecule L sec L )
that tropospheric lifetimes are « 1 year. Rate constant data
for the saturated halocarbons + OH reactions, were not, with
'" . '
the exceptions, of the Freons, too well defined. Their estimated
,•.. *-v
lifetimes in many cases lay in the 0.1-1 year region, and gave
reason for concern since tropospheric lifetimes of ~10 years
could result in a possibly significant transport of the halo- .
carbons into the stratosphere.
-------
Table 1
ESTIMATED EMISSIONS OF HALOGENATED COMPOUNDS INTO THE ATMOSPHERE
C impound
1
Fluorocarbon- 1 1
Fluor ocarbon- 12
Fluorocarbon-22
Fluorocarbon- 113
Fluorocarbon-114
Carbon tetra-
chloride
Chloroform
Ethyl chloride^
Ethylene dichloride
Methyl chloride
Methyl chloroform
.
Methylene chloride
Perchloroethylenc
Trichloroethylene
Vinyl chloride
World
production
in 1973,
millions
of pounds
670
980
270
110
100
2090
496
1210
26,400
880
900
935
1650
1540
15,600
Estimated Emissions into
Principal Uses
(and percent of
total production)
Aerosol propellant (782)
Aerosol propellant (47%)
Refrigerant (34%)
Refrigerant (66%)
Solvent (85%)
Propellant (9 IX)
Refrigerant (9X)
Production oC fluoro-
carbona (88%)
Production of fluoro-
carbons (902)
Produce tetraethyl
lead (85%)
Produce vinyl
chloride (78%)
Produce silicones (43%)
and tetramethyl lead (38%)
Metal cleaning (70%)
and degreasing
Paint remover (40%)
Dry cleaning (65%)
Metal cleaning (86%)
Produce polyvinyl
chloride (89%)
the atmosphere
World total,
millions of U
pounds
600
740
120
100
70
88
12
29
1250
11
835
760
1370
1390
774
in 1973
.S. total,
percent
56
50
50
50
50
50
50
55
35
60
60
55
45
30
25
-------
Table 2
ATMOSPHERIC CHEMISTRY SUMMARY
Worldwide
!»" Vf n Estimated
„ , chlorine, , , ,. , ._
Compound name _ half-life,
and formula 10 Ibs years
Perchloro- 1.17 0.01
ethylene,
c2ci4
Trichloro- 1.12 0.001
ethylene,
C2HC13
Ethylene di- 0.90 0.3
chloride
Methyl chloro- 0.74 1.1
form, CH CC1
Methylene 0.63 0.3
chloride
CH2C12
Vinyl chloride 0.48 0.0001
Fluor ocarbon-11, 0.46 100
CCljF
Type of Reaction
Photooxidation
initiated at
double bond
Photooxidation
initiated at
double bond
Photooxidat ion
initiated by OH
abstracting H
Photooxidation
initiated by OH
abstracting H
Photooxidation
initiated by
OH abstracting H
Photooxidation -
initiated at
double bond
Stratospheric
photo-dissociation
Tropospheric chlorinated
reaction products
(and approximate percent
of chlorine)
Trichloro acetyl
chloride (80%)
Phosgene (10%)
Hydrogen chloride (10%)
Dichloro acetyl
chloride (70%)
Phosgene (12%)
Formyl chloride (8%)
Hydrogen chloride (10%)
Mono chloro acetyl
chloride (100%)
Trichloro-
acetaldehyde (100%)
Phosgene (100%)
Formyl chloride (50%)
Hydrogen chloride (50%)
Fluorocarbon-11 (100%)
Chlorine
in product,
109 Ibs
0.93
0.12
0.12
0.79
0.13
0.09
0.11
0.90
0.74
0.63
0.24
0.24
0.46
-------
Table 2 (cont.)
Compound name
and formula
Fluorocarbon- 1 2 ,
Carbon tetra-
chloride, CC14
Fluorocarbon- 2 2 ,
CHC1F2
Ethyl chloride,
Chloroform,
CHC13
Worldwide
release in
1973 of
chlorine,
109 Ibs
0.4.4
0.084
0.050
0.015
0.011
Estimated
half-life,
- years
100
100
0.3
0.3
0.2
Type of Reaction
Stratospheric
photo-dissociation
Stratospheric
photo-dissociation
Photooxidation
initiated by OH
abstracting H
Photooxidat ion
initiated by OH
abstracting H
Photooxidation
initiated by OH
Tropospheric chlorinated
reaction products
(and approximate percent
of chlorine)
Fluorocarbon-12 (100%)
Carbon tetrachloride (100%)
Carbonyl fluoride &
chlorine monoxide (100%)
Fonnyl chloride (100%)
Phosgene (67%)
Chlorine monoxide (33%)
- Chlorine
in product,
109 Ibs
0.46
0.084
0.050
0.015
0.008
0.003
Methyl chloride, 0.007
0.4 Photooxidation
initiated by OH
abstracting H
Fonnyl chloride (100%)
0.007
-------
Another difficulty with the tropospheric lifetime estima--**
tion stems from the requirement that the average atmospheric OH
radical concentration must be known, since t^ (tropospheric
lifetime due to OH radical attack) - EKCOH]av] where k is the
rate constant for the reaction, halocarbon + OH, and COH]
is the average COH] concentration. At the time the EPA report^
was compiled, COH] «2 x 10 molecules cm was generally
thought to be the correct order of magnitude. More recently
5 -3°
this value has been lowered to 2r-6 x 10 molecules cm which,
of course, has the effect of increasing tropospheric halocarbon
lifetimes over those previously estimated. f;
Although OH radicals are capable of reacting with halocar-
bons this does not automatically remove them from the
troposphere. In Table 2, the resulting expected degradation
products from OH attack are also tabulated. For the chlorine
only containing compounds, the major products are the mono, dif
*• ' >$ f f
and trichloroacetyl chlorides, trichloroacetaldehyde, phosgene,
formyl chloride and hydrogen chloride. In the troposphere,
formyl chloride is almost certainly rapidly hydrolyzed to formic
acid and HC1, and as such, are purged from the atmosphere by
rain. The fate of the other degradation products in the ;i t
atmosphere was not so certain. The mono, di and triacetyl-
chlorides and phosgene were believed to hydrolyze to the
corresponding chloroacetic acids and hydrochloric acid via
homogeneous gas phase hydrolysis with atmospheric water vapor.
The acids would then presumably be washed out of the atmosphere.
° *•
The purpose of the present investigation was to resolve
some of the above noted uncertainties with respect to the fate
of halogenated hydrocarbons in the troposphere. Specifically,
the program had two tasks.
The first task was to determine-rate constants for the ?.£•.
reactions of OH radicals with the following saturated halo-
carbons; CH-jCl (methyl chloride), CH2C12 (methylene chloride),
-------
CHC13 (chloroform), CH3CH2C1 (ethyl chloride), CH2C1CH2C1
(ethylene dtchloride) and CH3CC13 (methyl chloroform), During
the course of the program data appeared which suggested that
rate data also be obtained for CCl2"CCl2 (perchloroethylene) and
C^BrCtUBr (ethylene dibromide) , and also for the pressure
dependence of the OH + CO reaction. From the derived rate data,
tropospheric lifetimes were to be estimated and the possible
implications with respect to the stratospheric ozone problem
assessed.
The second task was to determine homogeneous gas phase
hydrolysis rates for mono, di and trichloroacetyl chlorides
and phosgene. Again, from the derived data, tropospheric life-
times were to be estimated and the possible implications with
respect to the stratospheric ozone problem assessed. During the
course of the program, it became obvious that the heterogeneous
rate of removal of these compounds should also be investigated
and some preliminary data were obtained,
The remainder of this report describes the work performed,
and conclusions derived. Recommendations for future studies to
clarify still existing uncertainties have been made.
-------
SECTION 2
CONCLUSIONS
The results of the experimental studies of the kinetics of
the OH radical reaction with C1UC1, CH5C19, CHC1-, C9HcCl,
w «t> A O *• J
CH^ClCl^Cl, and CH3CC13 lead to the conclusion, that, with the
exception of some possible reservations concerning future
increases in CHoCClo production and release to the atmosphere,
the halocarbons listed will not make any significant contribution
to the chlorine burden in the stratosphere compared to that made
by the Freons 11 and 12. However, this refers only to the parent
halocarbon. As pointed out in the E.P.A, assessment of the
overall halocarbon - stratospheric ozone problem, it is also
necessary to consider the degradation products of the halocar-
bons after attack by hydroxyl radicals and subsequent oxidation
in the atmosphere.
The results obtained from the study of the CO + OH reaction
in 02 + N2 mixtures further confirm the contention12'13'1^'15
that the CO -I- OH reaction is pressure dependent. In addition,
they indicate that the change from the low to the high pressure
limiting value occurs in the range 100 < P < 300 torr. From
-13 3
the available data, a value of k, = 2.69 + 0.40 x 10 cm
-1-1
molecule sec is suggested for the high pressure rate
constant. The derived value of k(OH + iso C/Hin) - 1,59 + 0.18
-12 3 -1-1 ^10 -
x 10 cm molecule sec obtained in the present study is
not in very good agreement with the two other reported literature
values. The reasons for this are not clear and the absolute
value of this rate constant at the present time must be
considered not well defined. % .,„•
-------
From the results of the experimental studies of the
homogeneous gas phase hydrolysis of CC1H9COC1, CCUHCOC1,
Lm 4U
CCloCOCl, and COCl2i the tropospheric and stratospheric life-
times of those compounds were determined. Using the usual
formula, halflives of all four compounds were determined to
Q
be > 20 yrs in the troposphere, and > 10, yrs in the stratos-
phere. Although there is a considerable uncertainty in the
calculated lifetimes, a result of having to extrapolate values
for the rate constants to considerably lower temperatures than
that at which they were determined, there is no doubt that the
lifetimes of these species due to homogeneous gas phase hydrolysis
in the troposphere are indeed substantial and that unless other
mechanisms are available for their removal there is no
question that these species will all find their way into the
stratosphere in large amounts. Kinetic data were not obtained
for CC12HCOC1 due to experimental difficulties but it is
reasonable to assume that this compound will probably have a
lifetime in the same range as the mono and trichloroacetyl
chlorides. There is also no question that in the stratosphere,
homogeneous gas phase hydrolysis will not occur except on a
time scale which is absurdly large.
The study of the heterogeneous hydrolysis or "rain out" of
CC1H2COC1, CC12HCOC1, and CC13COC1 were preliminary in nature,
however, in view of the uncertainties, and the distinct possi-
bility that "rain out" may not be a very efficient removal
mechanism for these compounds from the atmosphere, it is
concluded that until further experimentation is made to determine
reliable values of sticking coefficients for all these com-
pounds, it is not possible to reliably assess the potential
impact of these compounds on the stratospheric ozone layer,
other than to note it could be substantial, since these com-
pounds are formed and are present in the atmosphere at consi-
derably larger levels than the Freons 11 and 12.
-------
SECTION 3
RECOMMENDATIONS
As a result of the. present program it is not clear whether
"rain out" is an efficient mechanism for the removal of the
secondary halocarbon pollutants'from the lower troposphere'!
which would prevent these compounds reaching the stratosphere.
The compounds, namely mono, di and trichloroacetyl chlorides,
phosgene and trichloroacetaldehyde, have a flux into the atmos-
phere which has the potential of carrying up to 5 times as much
chlorine into the stratosphere as the Freons 11 and 12. •**;•.
Preliminary "rain out" experiments for these compounds in
the laboratory have been made, but only for CC1-COC1 with any
precision. Even here further experimentation is required to be
certain that the derived sticking coefficients is applicable to
the real atmospheric situation. The present data suggest the
distinct possibility that "rain-out" for these compounds may not
be a very efficient removal mechanism Because of this possi-
bility, further study of the "rain out" problem is required.
The model used to date in these studies is quite simple and has
the advantage that relatively few experimental parameters^need
to be investigated in order to derive a realistic value for the
sticking coefficient to be applied to the atmospheric situation.
Once realistic tropospheric sticking coefficients have been
obtained the efficiency of the "rain out" process for these
:. ; ' *Vt-
compounds can be easily calculated, and their potential threat
to the stratospheric ozone assessed. Should the "rain out"
process be found to be inefficient, then a study of ice particle-
10
-------
acid chloride interactions should also be investigated to
determine their potential as atmospheric scavengers.
A further advantage of such a study would be the insight
it would provide into the whole question of the efficiency of
atmospheric "rain out" with respect to the removal of atmos-
pheric trace gases. This aspect of the study would be particu-
larly valuable since at the present time more and more concern
is being directed at the fate of potential toxic substances in
the environment, Finally, it would also be highly desirable to
make tropospheric mixing ratio determinations for these species
since, as of now, only limited data have been obtained for
phosgene and none are available for any of the other com-
pounds .
Should the above study demonstrate that "rain out" is not
efficient removal process for some, or all of these compounds,
then alternative mechanisms for their removal should be investi-
gated. One obvious possibility might be OH radical attack on
the hydrogen containing compounds.
Finally, the results of the present study showing the
pressure dependence of the OH + CO reaction in the atmosphere
also raises another problem. The temperature dependence of the
OH 4- CO reaction in it low pressure limit is known to be zero,
but it is not known if it is zero in the high pressure regime.
The latter would be of importance in the lower atmosphere with
respect to its effect on the OH radical concentration and to
enable better mathematical modeling predictions to be made.
The temperature dependence of the OH + CO reaction should be
determined in its high pressure region.
11
-------
SECTION 4
PRESSURE DEPENDENCE OF THE CO +• OH RATE CONSTANT
IN 02 + N2 MIXTURES
INTRODUCTION
The kinetics of the gas phase reaction
OH + CO ->• H + C02
have been the subject of many experimental investigations.
The reaction is of interest because of its significance in
combustion processes and atmospheric reactions. In addition,
because the rate constant was thought to be well established
it has been used as a reference reaction in experimental
determinations of hydroxyl radical reaction rates using a
18
competitive reaction rate method. This latter method was
being used in this laboratory to determine hydroxyl radical
reaction rates with halogenated hydrocarbons (Section 5 of this
report). During the course of this investigation, results of,
several experimental studies were published ' ' strongly
indicating the reaction, OH + CO -> H + COp to be pressure
dependent in the range 100-760 torr, and that the nature of
the third body also influenced the rate. Prior to these find-
ings the OH + CO reaction rate was considered to be pressure
independent. The reasons for this have been excellently
14 15
summarized recently, ' and will not be repeated here. In
two of the more recent pressure dependent studies referred to
above, air (or a mixture closely similar) was used as the make
13
up gas and the rate determined at a total pressure of 760 torr,
12
-------
700 torr 5 and, 100 torr, At 100 torr, k was determined as
1.4 x 10" cm molecule" sec*" , in good agreement with the
•I O
-------
Figure 1. Static Photochemical Reactor, Showing Inlet and
Outlet Ports and Gas Sampling Storage Volumes
14
-------
glass valves were connected to the top of the reactor. Two
of these were used for filling the reactor with &2®2 vaPor
premixed reaction gases. The third, connected to a capillary
tube projecting about 8 cm into the reactor was connected to
2
four storage volumes of about 3 cm each. The latter could be
evacuated and filled with gas samples from the reactor for
subsequent analysis.
The entire reactor assembly sat in the center of a
Rayonet Model RPR-1000 Photochemical Reactor. Irradiation
was provided by 15 RPR lamps with maximum intensity centered at
soooX.
The reactor was connected to a mercury free vacuum and gas
handling line which is shown in Figure 2 along with the rest
of the experimental apparatus. Pumping was provided by a
mechanical oil pump and a Varian ion pump. Pressures in the 10"
torr range were routinely attained, The line was constructed
of stainless steel tubing with the exception of the section
used for handling the H^O,* This section was constructed
of glass, with all glass and teflon valves. Three large volume
vessels were used for mixing and storing reaction gases. A
Wallace and Tiernan absolute pressure gauge calibrated against
the vapor pressure of several liquids was used to measure
pressures when making up reaction gas mixtures. Hydrogen
peroxide pressures were measured with a silicone oil manometer.
The latter was calibrated against a Barytron gauge, and also
compared with oil density measurements.
Connected to both vacuum line and reactor via a gas
sampling loop was a Varian Model 2700 Trace Gas Analyzer
for the analysis of carbon dioxide samples from the reactor.
o
The sampling loop had a volume of 8,9 cm . It could be opened
to the vacuum line, the reactor gas storage volumes or
isolated for injection into the chromatograph. Prior to filling,
the sample loop was first evacuated and then opened to the gas
15
-------
Figure 2. The St. ••cocher.iical Rt. . • v
The P'i ical Reactoi, Gas Chro;nat o^,raph,
Gas Sto • Mottles and Vacui
L6
-------
source of interest. Injection was accomplished via two three-
way ball valves which allowed the carrier gas to flush the sam-
ple into the chromatograph . For C02 analysis, the chromato-
graph was fitted with a 10* x' 1/8" stainless steel column packed
with 80/100 mesh For op ak Q. Carrier gas flow (Matheson U.K. P.
3 "1
Helium) was set at 45 cm min , with the analysis run iso-
thermally at 50°C. This resulted in a C02 retention time of
about 2 minutes. The helium-ionization detector used on the
trace gas analyzer has great sensitivity (~10~ moles of C02
could be detected in an air sample). However, the detector
response was non-linear and sensitive to coluifin pretreatment .
Accordingly, just prior to each experiment, the instrument was
calibrated (Matheson Calibration gas, 498 ppm (v) C05 in air)
~ —9
in the range expected in the kinetic experiment (10 to 10
mole C02) .
Gas mixtures of 02. N2, CO, and 02, N2, CO and isobutane
were prepared using standard vacuum line P,V.T. techniques.
Linde Zero Grade oxygen and nitrogen were used without further
treatment. Carbon monoxide, Matheson research grade purity
(min. 99,99%), was passed through a liquid nitrogen cold trap
prior to use to remove possible metal carbonyls. Isobutane,
99% mole minimum purity from Phillips Petroleum Company, was
used without further treatment.
90% H202 from F.M.C. was used in the experiments. It was
analyzed by titration with a standard solution of KMnO^. A
degassed sample was stored in the vacuum line at liquid N2
temperatures prior to use . All parts of the apparatus
exposed to hydrogen peroxide were passivated using standard
procedures with a 25% solution of sulf uric acid. Prior
to each experiment hydrogen peroxide vapor at a pressure of
about 1 torr was left in the reactor for conditioning purposes .
It was then evacuated in preparation for the experiment.
17
-------
H202 vapor was the first component introduceU.into the
reactor. The vapor was obtained from a thermostated H202
liquid sample maintained at a preselected temperature between
21°C and 24°C. The reactor was exposed to this vapor for two
minutes and the pressure recorded, From the latter, the amount
1 Q
of #2®2 an<1 H2° vaP°r in t^ie r«actor was calculated. Follow-
ing this, the other gaseous components (which had been premixed
and stored) were added until the desired reactor total pressure
was attained. Approximately 18 minutes after initial H202
introduction, two gas samples were extracted from the reactor
and analyzed for C02 (present as a minor trace impurity in the
oxygen and nitrogen). Thirty minutes after initial H202
introduction, irradiation was initiated. The first four gas ^
samples for analysis were extracted during the first 2-1/2
minutes of the experiment and retained in the storage volumes
above the reactor. These four samples were then analyzed, and
as an empty sample volume became available, a new sample was
••* • • . •
taken. A total of eight gas samples were extracted from the
reactor in a time period of from 10*12 minutes.
RESULTS
The basic equation used to calculate the rate constant k,
1 ft 9 ft •'•
for the reaction CO + OH * C02 + H was as follows:10' u
k
1
CisoC4H10]
,-1
[isoCH]
^ >"10-
where,
L • Initial rate of C02 formation with CO present
in the reaction mixture and isobutane absent
M - Initial rate of C02 formation with both CO and
isobutane present In the reaction mixture.
18
-------
k2 and ko are the rate constants for the hydroxyl radical
reaction with isobutane and hydrogen peroxide respectively.
The reaction conditions under which this equation applies
with respect to relative concentrations of the various reactants,
18 20
H2°2:02 and H2°2:C0' have been def^ned ln earlier studies '
and were satisfied in the present investigation. From the
experimental data, CQ2 production versus time was plotted
graphically using a computer best fit for the data points.
In all experiments the C02 production rates were almost, but
not quite linear, showing a slight fall off with time as the
experiment progressed. In order to take this slight curvature
into account, the data were Mbest fitted" using a 2nd degree
polynomial and the initial rate determined from the resulting
regression equation.
In order to derive values for k« as a function of pressure,
it was necessary to assign values to k2 and k^.
ka, the rate constant for the H«00 + OH reaction, has been
21 22
the subject of two reviews ' and three fairly recent determi-
nations.18'20'23 The latter gave k3-1.2 + 0.3x ICf12,
6.0 ± 0.4 x 10"13 and 8.4 + 1.7 x 10~13 cm3 molecule"1 sec"1
and present little reason to change the earlier recommended
.,1 o o . i i 22
value, k3 • 8.0 + 4.0 x 10 cm molecule" sec . Two
values are available for k2, the rate constant of the OH + iso
C4HlO reaction- A flash photolysis study bv Greiner gave
k7 = 2.3•+ 0.2 x 10" cm molecule" sec" and a competitive
• -13 3 -1 -1
method using CO + OH (k, - 1.5 x 10 cm molecule sec )
•L ,.12 3
as a reference reaction yielded k0 - 3.5 + 0.8 x 10 cm
i -i 1 ft —
molecule" sec" , These two values do not quite agree within
the stated experimental errors, the reasons for this not being
obvious. Because of the uncertainty in k2, it was felt prefer-
able to determine this quantity using CO + OH as the reference
reaction at a total pressure of 100 torr. At this pressure, the
CO + OH rate constant has its Hlow pressure value" of 1.50 x
19
-------
.. J.. J]
In Table 3"the data used to evaluate k0 are presented, A
•>.! 9 "i 1 1
value of 1.59 + 0.18 x 10~ cm molecule sec" was obtained,
where the error limits represents the standard deviation of the
experimental data points. This value is lower than either of
the two previously determined rate constants for the reaction.
• » ••'
The lack of agreement is most severe with respect to Gors'e and
Volman's determination, which was made using essentially the
same technique. The discrepancy is difficult to account for,
but could be due to an impurity in the isobutane sample.- In ?'•
Greiner's determination, a correction was made to the experi-
mentally determined value of k« to take into account the
destruction of OH by isobutyl radicals. This correction,
amounting to approximately -10% of the experimental value, was
-10 3
based on an assumed value for k(OH + C/HQ) of.1.7 x 10 cm
-1-1
molecule sec and an assumed concentration of OH radicals of
-11 -3
5 x 10 molecule cm , The lower value of k« determined
here might indicate that Greiner underestimated the correction
factor due to the OH + C/Hg reaction.
With the above determined value of k^, the rate constants
for the CO 4 OH reaction were determined as a function of
total system pressure from the data presented in Table 3. Again,
the quoted errors represents the standard deviation of the
experimental data points. In Figure 3, the data are presented graphi-
cally, and for comparison the values obtained by Cox and
Calvert in oxygen - nitrogen mixtures are also shown. The
error limits associated with our data are relatively large
10-1570 but unfortunately this is an inherent difficulty With
the method. These data strongly suggest that the high pressure
limit of the rate constant is achieved at 300 torr, and that
the change from the low to the high pressure value occurs within
the 100-300 torr pressure range.
20
-------
Table 3
SUMMARY OF THE EXPERIMENTAL DATA TOR THE REACTION OH + CO
100 TORR TOTAL PRESSURE
without iso-C.H,ft
4 10
with lso-C.H,n
Average
H2°2
1.01
1.06
System Composition (Torr)
H20 CO iso-CAH1()
1.65 14.6 0.00
1.73 14.6 1.24
°2 N2
32.8 49.9
32.8 48.6
-1 -1 9
Initial O>2 Production Rates (mole-liter -sec x 10 J
without iso-C.H, A
4 10
with
(3 expts): 8.18, 7,89, 7.91
(3 expts): 4.79, 5.09, 4.64
K- 1.50 + 0.24 x 10"13 cm.3-mol«cule"1-sec.~1
200 TORR TOTAL PRESSURE
Average System Composition (Torr)
without iso-C.H, n
4 10
with iso-C,Hin
1.05 1.71 29.6
1.06 1.73 29.6
0.00
2.51
66.6 101.0
66.6 98.5
Initial O>2 Product Rates (mole-liter"1-sec.'"1 x 109)
without iso-C.H .
with iso-C.H.-
4 10
(4 expts): 9.04, 9.35, 8.95, 9.29
(4 expts): 5.32, 5.48, 5.17, 5.57
1 - 1.65 0.16 x 10"13 cm.3-molecule~1-sec."1
21
-------
Table 3 (cont.)
300 TORR TOTAL PRESSURE
Average System
H2°2 H2°
without iao-e.H,n 1.31 2.12
4 10 t.
with iao-C.H.. 1.34 2.17
Composition (Torr)
CO i«°-C4Hl0. 02 N2
23.0 0.00 51.8 • 221.7
23.0 3.78 51.8 217.9
-1 -1 9
Initial CO. Production Rates (mole-liter -sec. x 10 )
without iao-C H
with iso-C.H_rt
4 10
(3 expts): 10.46, 10.82, 10.56
(4 expts): 5.93, 5.38, 6.05, 5.74
K - 2.66 0.32 x 10~13 cm. ^molecules-sec.""1
400 TORR TOTAL PRESSURE
Average System
H.O. H.O
without iso-C.H,A
4 10
with iso-C.H,A
2 I
1.13
1.15
I
1.84
1.87
Composition (Torr)
i
59.6
59.6
4 J
0.00
5.06 .
.u z
134.0
134.0
2
203.4
198.3
-1 -1 9
Initial CO. Production Rates (mole-liter -sec. x 10 )
without iso-C.H -
with iso-C^H 0
(4 expts): 10.90, 10.95, 10.08, 10.21
(4 expts): 7.33, 7.06, 6.34, 6.67
K, - 2.35 + 0.45 x 10~13 cm.^molecules-sec."1
22
-------
Table 3 (cont.)
without iao-C4H.0
with iso-C,H,ft
500
Average
H2°2
1.21
1.23
TORR TOTAL PRESSURE
System Composition (Torr)
H.O CO iso-C.H,-
2 4 10
1.94 38.6 0.00
1.97 38.6 6.33
P2 N,
86.8 371.5
86.8 365.1
-1 -1 9
Initial CO. Production Rates (mole-liter -sec. x 10 )
without iso-C4Hl0 (4 expts): 10.08, 10.05, 10.14, 10.23
(4 expts): 4.93, 5.59, 4.90. 5.55
with iso-C4Hl0
" 2.54 + 0.38 x 10*13 cm^-molecule^-sec."1
600 TORR TOTAL PRESSURE
Average System Composition (Torr)
without iso-C4H1Q 1.19 1.93 46.4 0.00
with iso-C.Hin 1.16 1.88 46.4 7.60
104.3 446.2
104.3 438.7
-1 -1 9
Initial C02 Production Rates (mole-liter -sec. x 10 )
without l8o-C4H1Q (3 expts): 10.40, 9.94, 9.62
with i»o-C4Hl0
(3 expts): 5.73, 5.28, 5.35
K - 2.93 0.36 x 10"13 cm.3 molecules-sec."1
23
-------
3.DU
^ 325
i
TJ
S 3.00
©
V)
i
~£ 2 75
u
1 2.5O
10
o 2.25
to 2
-is o 2.OO
X
c 1.75
o
5 1.50
t»
1 1.25
I.OO
c
^
—
-
—
— •
- _
1 1
^
^
l l
/•
J
I
T
r 4
\ _L T
1 r1]
1 I
— *-
i i i i i i i i i i
I |OO 2OO 3OO 4OO 5OO 6OO 7OO 8OO
Totol Pressure ( Torr }
Figure 3. Pressure Dependence of the Rate Constant for the Reaction CO 4- OH -*• H
+ Present Data, ACox EtalO) and D Calvert
-------
Taking a simple average of all the "high pressure" values
(P > 300 torr) given in Figure 3, k, - 2.69 + 0.23 x 10~13 cm3
— -1 ««T ~~
molecule" sec" . Heicklen's high pressure data for k, were not
included in Figure 3 since they were obtained in a hydrogen
atmosphere, which apparently has an even greater effect on the
20
reaction rate. Both the Cox and Calvert high pressure values
for k, are in substantial agreement with ours, though the value
-13 3
given by Calvert would be lowered from 2.97 to 2.07 x 10 cm
-1 -1 -12 3 -1
molecule sec if our value of k0 = 1.59 x 10 cm molecule
-1
sec had been used instead of Greiner's in their calculation
of kt.
Further support for the proposed high pressure value of
-13 3 -1 -1
k^ = 2.69 + 0.23 x 10 cm molecule sec comes from a
determination in this laboratory of the rate constant for the
OH + CH-jCl reaction using the CO + OH reaction as the reference
reaction (see Section 5 of this report). In an oxygen-nitrogen
mixture at 400 torr, k(OH + CH3C1) was determined at 3.1 +
0.7 x 10"14 cm3 molecule"1 sec"1 with k, = 2.69 x 10"13 cm3
-1-1
molecule sec . The former value compares well with reported
literature values of 3.6,10 4.2,25 4.626 and 5.9512 x 10"14 cm3
molecule" sec" determined using different methods.
25
-------
SECTION 5
RATE CONSTANTS FOR THE REACTIONS OF OH WITH CH-Cl, CH9C1,,,
C2H5C1, C1CH2CH2C1, CH3CC13, C2H4Br2, CHC13 arid C^ljr £•
IN THE PRESENCE OF 02 + N£
KINETIC RESULTS USING THE STATIC REACTOR
.,'• *'••"••'
Introduction !
The possibility that anthropogenic releases of halogenated
hydrocarbons can find their way into the stratosphere and be
involved in a catalytic ozone destruction cycle has been well
2 27
publicized. ' In the troposphere these compounds are suscept-
ible to attack by OH radicals, and this mode of reaction could
act as a natural sink for these species, preventing them from
diffusing into the stratosphere. In order to determine the"
likely efficiency of such a sink it is necessary to have kinetic
data for the halocarbon-hydroxyl radical reaction rate. To date
several groups of workers have been studying these
Q ID 95 9ft 9Q "\f\ ^1 90
reactions,*' '•'**•"' but with one exception, all were
made in the presence of inert gases, rather than oxygen and
nitrogen mixtures as occurs in the atmosphere. Although it is
not likely that these reactions are sensitive to the presence
of different "third bodies," greater confidence could be placed
in the rate date when applied to atmospheric chemistry, if the
"third body" independence were established. For this reason,
the present study was initiated. A competitive rate method was
chosen in which the halocarbon competes with CO for available OH
2
radicals. The rate of formation of C02 in presence of halo-
carbon can then be related to the rate of OH attack on the
halocarbon.
26
-------
Experimental
The details of the experimental equipment and techniques
have been described previously in Section 4 of this report.
For this reason, only those details peculiar to the present
study will be presented. Photolysis of H202 (^90%) obtained
from F.M.C. was used as the OH radical source, as in the pre-
vious study. The halocarbons, CH3C1 and C^HcCl from Matheson,
were used without further purification except for vacuum line
degassing. Chromatographic analyses indicated purities of
>99.99 and 99%, respectively. CH3CC13, CH2C1CH2C1, and
C2H^Br2 (Eastman Kodak), CH2C12 and CHC13 (Fisher), and
C2C1/ (Baker) were all treated chemically to remove any
unsaturated impurities. This, as experiments indicated here
9 10
and elsewhere was particularly necessary for CHoCCl^. Puri-
fication was accomplished as follows; Bromine, 1-2%, was added
to the halocarbon and the mixture allowed to stand overnight.
A solution of sodium thiosulfate was used to remove the bromine.
This was followed by washing with water, a solution of sodium
bicarbonate and finally water again. The halocarbon was dried
over calcium chloride, carefully distilled and stored in light
resistant bottles. Gas chromatographic analysis indicated the
following purities; CH3CC13 -99.4%, C2H2C12 ^99.9%, CH2C12
^99.5%, C2H4Br2 ^99.5%, CHC13 ^99.9% and C2C14 -99.0%. Where
possible, the amount of halocarbon added to the reactor was
optimized to give the best precision in determining the rate
constant. However, for CHoCCl3 and C2H,Br2 the vapor pressures
of the compounds were insufficient to enable the best conditions
to be realized.
Gas mixtures of 02, N<,, CO,, H202 and halocarbons were pre-
pared using standard vacuum line techniques as described
previously in Section 4 of this report. Linde Zero Grade 02 and
N2 were used without further treatment. Carbon monoxide,
Matheson research grade purity (min 99.99%) was passed through
27
-------
a liquid nitrogen cold trap to remove impurities prior to use.
In all cases the total pressure in the reactor was adjusted to
400 mm Hg and the reaction run at 29.5°C, Photolysis of* the
hydrogen peroxide was made in a Rayonet Model RPR-1000 Photo-
chemical Reactor using lamps with maximum intensity centered at
o
3000A. Chromatogra"phy was used to monitor C02 production.
Results and Discussion
Under the conditions used in this study the following
reactions were considered as important in interpreting the
18
system's kinetics;
H202 + hv • 20H 2kx I
OH + H0 - H0 + H0
2
H02 + H02 22 2 3
OH + CO =* C02 + H k4
OH + RX' « RX' + H20 k5
H + 02 + M = H02 + M k?
RX' + 02 - nC02 + ? kg
RX' is the radical formed due to hydrogen abstraction by OH on
RX. With the exception of kg, the reaction sequence is identi-
cal to that used by Volman. In the present study it was
necessary to include kg since sometimes the experimental data
clearly indicated that the halocarbon radical was also producing
C02 in the system.
The above sequence of reactions may be treated in the usual
way and the following sets of equations derived:
Case I with CO but no
dCCOoJ
dt 4
k/rOHHCO]
28
-------
At steady state—
OH
la CH2023 - k2COHKH202l * k4COH3CCO] = 0
2k, I CH9093
1 * 2 2
ss kCH03 + k CC01
Case II with CO + RX and kg
dCC09]
• z - M = kCOH3CC03
At steady state—
k5COH]CRX3 » 0
2k, I CH0003
COHJ - 1 a z z
ss k2CH2023 + k4CC03
M
k2CH2°2] + k4CCO:i + k5CRX:i
29
COH3CC03 -
-------
Case III with CO + RX and kQ # 0—
•o
dCC03
M - k4COH]CCO]
At steady state--
k^OHHRX] = 0
2k,I [H~09]
COH]sg = k [H o ] Vk
-sr^- = k5 [OH][RX] - kg
k,[OHHRX]
]e_ = \ rn -, —
ss kg LC^J
M = k4i:OHJCCO] + nk5 COH3CRX]
M = [OH] (k4CCO] + nk5 [RX])
+ nk5[RX])
M = k2[H202] + k4[CO] + k
Case IV with RX no CO and kg #0—
d[C09:l
—3^— - N « nkgl'RX' li:02 I
30
-------
At steady state—
] - k2COHKH2023 - kgCOHlCRXD - 0
2k..I CH9093
[OH3 -
'as k2CH2023 +
= k5COH]CRX3 - k8CRX'3C023
r
o =
SS
krCOH3CRX3
N - nk [OH][RX3
2k1IaCH2023nk5CRX]
k2CH2°2] k5 [RX]
If these equations are manipulated to eliminate 2k, I , the
following results are obtained:
k[H0] k,CCO]
+ L> - when k = °
k2CH2023 (L-M+N) + k.CCO] (L-M)
k5 " -- (RX) (M>N) - when k8
Nk4CC03 k2r.H2023 + k5 LRX'J
and n = RX ^ki;H0l + k
31
-------
The symbols have the following meanings:
L = Initial C02 production rate with only CO in the system.
M = Initial C02 production rate with CO + RX in the system.
N * Initial CQ2 production rate with RX but no CO in the system.
In the above equations the following values were used for the
100 -i *
rate constant; k0 - 8 x 10" cm molecule" sec"> and, from
-13 3 -1 -1
Section 4 of this report, k, = 2.69 x 10 cm molecule sec ,
In the well behaved experiments, (k« = 0), C0« production rates
were almost, but not quite linear, showing a slight fall off
with with time as the experiment progressed. To take the slight
curvature into account, the data were "best fitted" using a,2nd
degree polynominal and the initial rate determined from the
resulting regression equation. In those experiments in which
kg was not zero, the C0« production rate curves showed a slightly
more marked curvature with time. These curves were "best
fitted" to an experimental equation of the type y = ae x + C.
The results of the various experimental determinations are shown
in Table 4. The quoted error limits represent the standard
deviations of the experimental data points.
'CH3C1--
Methyl chloride and methyl chloroform were both well
behaved systems in that no CO- was produced on photolysis of
RX + H202 + 02, or from photolysis of RX + 02. For methyl
chloride k(OH + CH-jCl) = 3.1 + 0.8 x 10~14 cm3 molecule"1 sec"1
which compares well..with values of 4.4 + 0.5 x 10" , '
3.6 + 0.8 x 10~14 ' and 4.6 + 0.2 x 10"14 ' cm3 molecule"1
sec obtained in low pressure experiments in the absence of
00. The agreement is not very good with the value reported by
29
Cox, determined in an air mixture at 1 atm. pressure of
8.5 x 10 cm molecule sec" . However, the latter value
29
was only considered good to within a factor of two.
32
-------
Table 4
SUMMARY OF EXPERIMENTAL DATA FOR THE REACTIONS
OH + HALOCARBONS
Methyl Chloride CH3C1
Average System Composition (Torr)
H202 H20 CO CH3C1 J)z N2
without CHjCl 0.93 1.61 29.8 0.0 134.2 233.5
with CH3C1 0.96 1.59 29.8 59.6 134.1 174.0
Initial C02 Production Rates (mole-liter"'sec"l x 109)
without CH3C1 (6 expts) = 0.62, 10.07, 9.47, 9.85, 9.77, 10.25
with CH3C1 (3 expts) - 8.06, 7.89, 8.47
ks=3.1+0.8x 10"1M cm3 molecule"1 second"1
Methyl Chloroform CH3CC13
Average System Composition (Torr)
H;0?. H20 CO CH3CC13 0? N?
without CH,CC131.09 1.76 29.8 0.0 134.0 233.4
with CHsCClsl.09 1.77 29.8 29.8 134.0 203.5
Initial C02 Production Rates (mole liter"1 sec"1 x 109)
without CH3CC13 (6 expts) = 16.0, 16.0, 18.0, 16.7, 14.3- 15.4
with CH3CC13 (6 expts) = 13.5, 16.2, 13.6, 15.1, 16.2, 15.6
k5 D 2.1 x 10"l "* cm3 molecule"1 second 1
k < 5.7 x 10"^ cm3 molecule"1 second"1
33
-------
Table 4 (cont.)
Ethyl Chloride C2H5C1
Average System Composition (Torr)
without C2HSC1
with C2HSC1 and CO
with C2HSC1, no CO
H202
0.00
1.00
0.97
H20
1.63
1.63
1.59
CO
29.8
29.8
0.0
C2H5C]
0.0
29.8
29.8
L 0.2
134.1 '
134.1
134.1
_^2
233.5
203.7
233.5
Initial C02 Production rates (mole liter"1 sec"1 x 109)
without C2HSC1
with C2H5C1 and CO
with CjH5Cl, no CO
(6 expts) - 18.9, 17.2, 14.9, 17.6, 17.3, 17.1
(4 expts) - 14.3, 13.3, 15.2, 13.8
(4 expts) = 11.6, 10.2, 12.4, 11.5
k, - 4.4 + 2.5 x 10"
nk, « 2.8 + 1.5 x 10"
13 cm3 molecule"1 second"1
13 cm3 molecule"1 second"1
1,2-Dichloroethane CH2C1CH2C1
Average System Composition (Torr)
H202 H20 CO CH2C1CH2C1 02
without CH2C1CH2C1 1.02 1.68 29.8 0.0
with CH2C1CH2C1 and CO 1.02 1.67 29.8 29.8
with CH2ClCli2Cl, no CO 1.00 1.64 0.0 29.8
233.4
203.6
238.5
Initial C02 Production Rates (mole liter"1 sec"1 x 109)
without CH2C1CH2C1 (5 expts) = 9.9, 9.7, 10.7, 10.7, 11.1
with CH2C1CH2C1 and CO (6 expts) = 9.7, 12.2, 11.8, 11.7, 9.9, 9:5
with CH2C1CH2C1, no CO (5 expts) = 8.5, 9.0, 9.4, 8.5
k
k
nk
nk
5 «= 6.5 x 10" ^ cm3 molecule"1 second"1
s£ 2.9 x 10]]13 cm3 molecule"1 second]]1
5= 7.1 x 10"'" cm3 molecule"1 second"1
.s <_ 2.4 x 10"13 cm3 molecule"1 second"1
34
-------
Table 4 (cont.)
Methylene Chloride CH2C12
Average System Composition (Torr)
H202 H20 CO CH2C12 02 N2
without CH2C12 0.9"3 1.52 29.8 0.0 134.2 233.6
with CH2C12 and CO 0.93 1.52 29.8 59.6 134.2 174.0
with CH2C12 no CO 0.94 1.54 0.0 59.6 134.2 203.7
Initial C02 Production Rates (mole liter"1 sec"1 x 109)
without CH2C12 (6 expts) - 9.62, 10.07, 9.47, 9.85, 9.77, 10.25
with CH2C12 and CO (4 expts) = 8.54, 8.33, 9.20, 8.08
with CH2C12 no CO (2 expts) - 0.81, 1.08
k5 - 2.7 + 1.0 x 10" 1H cm3 molecule"1 second"1
nk5 - 3.5 ± 1.1 x 10"l5 cm3 molecule"1 second"1
Chloroform CHC13
CO,, was produced from the reaction CHCU 4- OH + 02 and the photolysis
of CHC13 + O,. The second calculation method was used and corrected
for photolysis.
Average System Composition (Torr)
H2°2 "V^_ CO
Without CHC13 1.16 1.90 29.8
With CHC13 and CO 1.14 1.87 29.8
With CHC13, no CO 1.16 1.89 0.0
Photolysis of CHC13 0.00 0.00 0.0
Initial CO, Production Rates x 10 (mole-liter" -second )
Without CHC13 (2 expts.) - 10.24, 8.03
With CHC13 and CO (2 expts.) - 9.71, 9.05
With CHC13, no CO (2 expts.) - 4.69, 3.08
Photolysis of CHC13 (2 expts.) - 0.59, 0.16
k- - 1.4 x lO" cm3-molecule" -second"
k5 <. 5.0 x 10"14 cm3-molecule" -second"
n - 0.74
These results were calculated after making a correction for photolysis
in the CHC13 + 02 system.
35
-------
32 '
It has recently been suggested that one of the channels
for reaction of CH2C1, formed by the attack of OH on CH3C1, in
air, is as follows; CH2C1 '+ 02 •*• CH2C102 followed by either,
2CH2C102 •*- 2CH2C10 + 02 and or CH2C102 + H02 -»• CH2C10 + OH.
The present results indicate that the latter reaction channel
cannot be very important, since if it were, it would result in
k(OH + CH^Cl) obtained in our system being lower than the
reported values.
CH3CC13--
The precision of the data for methyl chloroform was not as
good as for methyl chloride. This was largely due to the
relatively low vapor pressure of CH~CC1~ preventing an optimum
-J J I Vi
pressure of the gas being used in the reaction mixture. An <•"
unfortunate characteristic of the competitive rate method requires
that for optimum precision L = 2M using the nomenclature de-
fined earlier. For reactions, (OH + RX) , with relatively low
rates compared to CO + OH, optimum precision requires relatively
high pressures of RX. Because of this difficulty, the rate
constant for k(OH + CH3CC13) = 2.1 x 10~14 cm3 molecule"1 sec"1
is quoted as the most probable value, with a limit of k£ 5.7 x
10 cm molecule sec . These values are in accord with f
-14.29 -14'31
those in the literature of 1.64 + 0.2 x 10 2.19 x 10 ,
1.5 ±0.3 x 10"149'10 and 2.8 x 10"1429 cm3 molecule"1 sec"1.
The next three compounds, C^Cl, CH2C1CH2C1 and CH2C12,
were not well behaved in so much that C02 was generated in the
reactions RX + H202 + 02 on photolysis, though not in the
reaction RX + 02 on photolysis. Experiments were therefore made
to determine initial C02 production rates for RX + H202 + 02 on
photolysis, in addition to the usual required C0« production rate
determinations.
36
-------
C2H5C1--
4.4 + 2.5 x ICf13 cm3
The value of k(OH + C0H,-C1)
-1-1
molecule sec calculated from equation 2, is in good agree-
ment with the only other literature reference of 3.9 + 0.7 x
-13 3 -1 -1 31 ~~
10 cm molecule sec .
C1CH2CH2C1--
The precision of the results for 1,2 dichloroethane, were
not very good, largely due to the similar reaction rates in the
presence and absence of RX in the reaction mixture. A most
)ba
-1
T / ^ 1
probable value of k(OH + ClUClCtCl) = 6.5 x 10 cm molecule"
sec was derived with an upper limit of k(OH + CHC1CH9C1)
-13 "* -1 1
£2.9 x 10 cm" molecule sec . The latter is in satisfactory
agreement with the only other literature value, of
k(OH + CH9C1CH9C1) = 2.2 + 0.5 x 10"13 cm3 molecule"1 sec"1.
4-> £• "" *
CH2C12--
The C02 production rates in the methylene chloride system,
CH2C12 + OH + 02 showed a pronounced upward curvature after
about 2-3 minutes of reaction indicating the possibility of a
chain reaction. However, the initial C02 production rate was
only about 10% of that in the CH2C12 + OH + 02 + CO reaction.
The rate was considerably less than in the C2H,-C1 and C H^C12
experiments, and the C02 versus time plots for the CHpC-L + OH
+ 02 + CO system, only showed a very slight upward curvature.
The derived rate constant k(OH + CH0Cl0> = 2.7 + 1.0 x 10"14
3 -1-1 ~~
cm molecule sec was not in good agreement with^literature
-13 25 -139,10
values of 1.45 + 0.2 x 10 , 1.55 + 0.34 x 10 and
-13 3~ -1 -1 ~
1.04 x 10 cm molecule sec , the latter being obtained
in an experiment with air at 1 atm pressure. It is not obvious
why the present result differs from the literature values.
Impurities were detected chromatographically in our CH2C12
sample prior to purification, but were all absent after treat-
ment. In two of the literature studies, CH^C purities were
37
-------
Q in 95 "
assigned at 99 . 93% ' iu and 99.9% and it seems unlikely that
-0.1% of a likely impurity could account for the order of
magnitude difference in reaction rates. A possible explanation
lies in the reaction mechanism involved. The primary OH attack
on CH9C1, is that of H abstraction, CH«Cl, + OH '•»• CHC10 + H0dV
L- £* » t* £• £» Zt
In the presence of oxygen, the CHC10 radical can react with
32
02 to presumably form CHC1202 which then decomposes as;
CHC1202 -»• COG12 + HO. Alternatively one could write, CHC12 +
02 - COC12 + OH directly, for which it is easily possible to
show the reaction is highly exothermic. Previously it had
been postulated that CHC10 decomposes to CIO and HCOG1,
though this interpretation caused some difficulties since
no HCOC1 was observed in the experimental system. If the
reaction channel CHC12 + 02 •* COC12 + OH, is a significant
path for the destruction of CHCl^, then indeed this would
result in our method giving a low value for k(OH -f CH^Clj) ,
since the CH^Cl^ is effectively not competing with CO for
the available OH. Alternatively if CHC1907 •»• HCOC1 + CIO is • "'
£ £ .
an important channel, then a low value could also result if the
reaction CO + CIO '•*• C02 + Cl takes place readily. To date, there
do not appear to be any data on this reaction.
In those reactions in which C02 is produced due to RX + OH
+ 02 •*• C02 + product, it is possible to calculate values of n
for the process from the expression;
Nk^CCOJ k9[H909;i + k,CRX]
n = *
k5URXJ
where the symbols have the values defined earlier. Values so
obtained are as follows; 0.63, 1.09 and 0.13 for-^HeCl,
C2H^C12 and C^C^, respectively. That n for the C2H^Cl2
reaction is >1 simply implies that it is possible for both car-
bon atoms to be converted to C02 in the reaction, though within
38
-------
the precision of the data the value of 1,09 is hardly signifi-
cant. Recently, a study of the reaction products formed when
C2HcCl and C^H^C^ are subject to H atom abstraction in air
has been made (private communication from P. L. Hanst and J. W.
Spence) . (X^ was found to constitute ~60 and 50 mole %t
respectively of the product for 02^01 and^H^C^, in good
agreement with the values reported here.
Detailed experiments were not made on CHClo, C9Cl,, and
•J £m *T
C^H/ Br£ for a variety of reasons .
CHC13«
Preliminary experiments made with this compound indicated
several difficulties. Irradiation of the system, CHClo + 0^
alone resulted in the production of CC^, with an initial rate
about 10% of that in the CHC1- + OH + CO + 09 experiments.
-------
can be assigned to these results.
C2CV~
Preliminary experiments with this system indicated that
large amounts of C02 were produced from the system C2C1, + OH
+ 02- Since fairly comprehensive data were already available
for the C2C1, + OH reaction it was decided not to pursue the
study further. *
C2H4Br2-
The system C2H,Br2 + 02 produced substantial amounts of
C02 on irradiation. For this reason no attempt was made to
study the C2H,Br2 + OH reaction rate.
KINETIC RESULTS USING THE DYNAMIC REACTOR
Introduction
The primary goal of this phase of the project was to
determine the rates of reaction of various halocarbons with
hydroxyl radicals. A competitive rates method was used as
described in the previous section in which the rate of an unknown
reaction is determined by comparison with a well known reaction
in which it is in competition. The main difference in the pre-
sent case was the use of a flow, rather than a static reactor
with mass spectrometric rather than gas chromatographic
detection of products.
The source of hydroxyl radicals was the photolytic
o
decomposition by 2537A light of hydrogen peroxide;
H000 + hv 20H
/ z.
33
This reaction has been extensively studied by Volman and
produces a clean source of hydroxyl radicals in the electronic
ground state without the complication of any other interferring
radicals or excited molecules. The competing reactions are:
40
-------
' k
OH + CO
and
OH + CO — i— > C02 + H
OH + RX — =-> P2
where RX is a chlorinated hydrocarbon and P2 is an unidentified
product. The requirements for use of this competitive rate
technique are that the OH radicals must react only with the
CO and RX and that no other sources of C02 are present.
This latter requirement played an important role in some of the
studies.
The rates of formation of the products of the two reactions
are given by:
dt -C02
= k, CC01COH]
di:p9j
-dF- = RP2 ' k2
Taking the ratio of these two rates;
RC02 k
R k
P2
l CCOJCOHJ
2 [RXKOH]
[RXJ = k_l
CCOJ ko
If R*rn is defined as the rate of formation of C09 in the
v \J f\ **
absence of competing reactant, then;
41
-------
RC02 " R*C02
defines the rate of formation of C02 in the presence of^a com-
peting reactant, R«. Rearranging 5,
RP2 " R*C02 " RC02
and substituting into 4, we have;
[R2J
CCOJ
4
iv co2 ixco2
Rco2
or
R*C02
Equation 7 is the form of the rate expression used to compare the
two rates, k-. and k2.
The experimental method is as follows. A flow of gases ,
containing H202> CO and a carrier gas (air, 02 or He) is set up.
The H202 and CO concentrations may be determined absolutely
or simply as known gas flow rates. The 2537A lamps are turned
on and the rate of formation of C02, R*co » determined for a
given flow rate of reactants. Without changing any of the
initial gas flow rates, a small amount of chlorinated hydrocarbon
o
is added to the reactant gas stream. The 2537A lamps are again
turned on and the new rate of formation of C00 determined, Rrr. .
/ \j\Jn
The flow rate of chlorinated hydrocarbon may be changed and the
rate of formation of C02 determined for this new condition. This
42
-------
procedure is repeated for the desired number of points.
Each set of data points yields a ratio of the rate constants,
^1/^2 versus RX concentration. Since k, is well known, k« can
be accurately determined.
Apparatus
A diagram of the apparatus used in these experiments is
shown in Figure 4. The carrier gas was either helium, nitrogen,
oxygen or air. All were tried with various degrees of success.
Matheson Research Grade CO was added directly to the carrier
gas flow. Part of the carrier gas was diverted and bubbled
through a reservoir of ^0^ immersed in a thermostated bath
held slightly below ambient temperature to prevent condensation.
The two flows were recombined prior to entering the quartz
reactor. A pyrex mixing tube allowed the gases to be trans-
ported to the lower end of ±he quartz reactor which was
positioned in the center of an RPR-16 Rayonet Photochemical
Reactor with 16, 2537A mercury lamps. The gases entered the
reactor where they were irradiated. The reacted gases then
entered a second manifold, shown in Figure 5. A 9" capillary
tube was used to sample these reacted gases into the first
differentially pumped stage of the mass spectrometer. The
gases then passed through two stages of differential pumping be-
fore entering the ion region of an Extranuclear quadrupole mass
spectrometer using electron impact ionization with phase
sensitive detection using a Daley detector.
Experimental Method
A typical experiment was conducted in the following manner.
First, a known flow of CO, H^O^ and carrier gas was established.
Once the flow of gases had stabilized, the photolysis lamps were
turned on and the amount of C02 formed determined. The latter
was not determined absolutely but as a linear pen deflection on
a chart recorder with the mass spectrometer locked on mass 44.
43
-------
- Lamps
Metering
Valves
Calibrated
Flow Meters
M
CO
Calibrated Flow Meter
Helium
Figure 4. Gas Flow and Reactor System
-------
Ui
Quadrupole
Mass Spectrometer
Icnisation
Source -*
Vacuum
Pump
Ion Detector
Phase Sensitive
Detector Electronics
Vacuum
Pump
Photochemical
Reactor
Figure 5. Mass Spectrometric Gas Sampling System
-------
This was the value of R*CQ in equation 7.
The lamps were then turned off, and RX was added to the
reactant gas stream. After a period to allow for equilibrium,
the lamps were turned on and the formation of CO 2 again measured.
This was the value RrA in equation 7. Values of R*rn /Rrn
c.u« ^ • WJft ^2
were determined for various flows of RX in equation 7 with GO
kept constant.
Results
Calibration Technique--
During the development of the flow reactor system, a paper
by Heicklen, et. al. was published indicating a pressure
dependence for the reaction OH + CO. Literature values for
the OH + CO, rate constant, mainly determined in the low
pressure region, were felt not to be trustworthy for the
atmospheric pressure conditions used in the present study. An
independent determination of the reaction under our experimental
conditions was deemed necessary. The reaction OH + CnH, had
been well studied; it appears to be a simple hydrogen
abstraction reaction and had been shown to be pressure
34
independent. This reaction was therefore used to determine
a suitable value for the rate constant of the CO + OH reaction
at atmospheric pressure. A plot of these data is presented in
Figure 6. The value of the rate constant for the OH + CO
reaction was found to be;
kOH + CO = 3<1 - °'5 x 10~13 cm3/molecule-sec
This result supports Heicklein's assertion that the OH + CO
reaction is indeed pressure dependent. This value was used
throughout the flow reactor studies.
46
-------
2.0
1.5
1.0
R*CO /RCO
LCD]
O 0.26 ml/sec
D 0.32 ml/sec
A 0.56 ml /sec
I i i i I I J I I
0.25
0.50
:c2H6]/rco]
0.75
Figure 6. Plot of Data for OH + Ethane. Reaction at 760 Torr and 298°C
-------
OH + CH3C1--
The first reaction studied was the reaction of OH + CH-Cl.
The data obtained is plotted in Figure 7. The rate constant
determined for this reaction was
1 /
~
The temperature dependence of this reaction over the range
10 to 90°C was also investigated. An Arrhenius plot of the data
is shown in Figure 8 and the temperature dependent rate constant
was determined to be;
-12 -2111 3
kOH + CH Cl " 1>68 x 10 e — RT~ cm /roolecule.sec
*j
in good agreement with the two previously reported values.
OH + CH3CC13--
A purified sample of methyl chloroform was used in this
investigation. The data obtained for this reaction is plotted
in Figure 9. The rate constant obtained was;
I/O *
r>t = 1-6 + 0.2 x 10" cm /molecule . sec
nUln —
in good agreement with the literature values .
OH + CHC13--
An inhibitor free sample of chloroform was obtained and
used in this study. The data obtained is shown in Figure 10.
The rate constant;
-13 3
NDH + CHCl = 1- 16 + 0. 15 x 10 cm /molecule. sec
is in good agreement with existing literature values.
48
-------
z.q
R*co2/Rco2
(0.15+ 0.01)rCH3cn/LC
-------
-30-
--31
© This work
£ Pttts
rn Davis
-32
lill
. I
1 1 t I
2.0
2,5
3.0
1000/T
3.5
4.0
Figure 8. Temperature Dependence of the Reaction OH + CH3C1 •
-------
1.1
l.tf
R*CO /RCO = (0.053*0.003):CH3CC13]/[CO] + (1.004^0.003)
fCO] - 0.26 ml/sec
i , . .
I I
0.5
1.0
rcH3cci3]/rco;
1.5
2.0
Figure 9. Plot of Data for OH + Methyl chloroform. Reaction at 760 Torr and 298°C
-------
2.Or
R*CQ /Rco = (0.376-0.005) :CHC13:/:CO- + (0.960* 0.005)
[CO]
CD 0.26 ml/sec
u3 0^31 ml/sec
Cn
K)
i.o-
t i i
0.5
1.0
[CHC13]/[CO]
ii r_._ i
1.5
2.0
Figure 10. Plot of Data for OH + CHCl^, Reaction at 760 Torr and 298 C
-------
OH + CH2C12--
Methylene chloride was the last compound investigated by
the flow reactor technique. The data obtained is plotted in
Figure 11. The value obtained for the rate constant;
-13 3
knit , ru r-i = 1.46 + 0.19 x 10 cm /molecule, sec
Un T L>n^Li.Ln —
is in fair agreement with previous literature values.
Discussion of Results
Completed Systems—
As can be seen from Table 5, the data obtained on the
systems studied are in excellent agreement with previous
literature values. The contribution made by this study was to
confirm the reliability of these rate constants under conditions
more closely resembling tropospheric conditions.
Problems with the Technique--
The flow reactor technique was designed to be simple and
yet yield pertinent data regarding the tropospheric reactivity
of various halocarbons with hydroxyl radicals. This was the
first time an attempt to carry out these reactions was made in
the presence of air at 760 torr and ambient temperatures. How-
ever, as with many such systems, side reactions occurred which
prevented the use of this technique for the measurement of the
kinetics of all the halocarbons desired. Previously it was stated
that a requirement of the successful application of this
technique was that no other sources of C02 should exist.
Unfortunately, the conditions under which these reactions were
conducted resulted in the production of large amounts of C02
from the reaction of OH with some of the halocarbons. Only
in the four systems reported was the production of C02 by
unwanted reaction low enough or completely absent that the
technique could be used satisfactorily. All other systems
53
-------
R*co /RCQ = (0.472^0.008)[CH2C12]/[CO] + (1.02±0.01)
rCCn = o,26 ml/sec
lilt
1 I i I ii I
till
1.0
2.0
[CH2C12]/[CO]
3.0
4.0
Figure 11. Plot of Data for OH + Methylene Chloride. 760 Torr and 298 C
-------
Table 5
S15WARY OF RATE DATA FOR SELECTED HALOCARBOSS
Ul
Ul
k x 10 era Ax 10"" cm
Compound Molecule sec aolecule sec
CH,C1 4.4 + 0.5 i.l
3 -
3.6 4 0.8
4.3 1.34
8.5
3.1 + 0.8
4.6 + 0.7 1.68
CH2C12 14.5+2.0
15.5 + 3.4
11.3 ± 0.6 4.27
10.4
2.7 4- l(d)
CHC1. 10.1 + 1.5
16.8
10.4 + 1.2 4.69
I5-0
11.6 + 1.5
CjH.Cl 39+7.0
44 + 25
CH_C1CH_C1 22. 0 •+• 5.0
6.5 (£29)
CHjCClj 1.5 + 0.3
1.59 + 0.16 3.72
2.19 + 0.27 1.95
2.8
2.1 (<5.7)
1.6 + 0.2
a) IITRI Static Reactor b) IITRI Flow Reactor
d> This value is known to be incorrect and was not
considered in deriving k (selected).
k298°K (selected) k265°K (selected)
E Cal. Source x 10 cm aolecule . sec x 10 co Molecule sec
2700 Pitts25
Howard
2181 Davis30
Cox29 i.; 2.6
HIRI***
2211 HTRI(t)
Pitts
Howard
2174 David 13.8 9.0
Cox
IITRI
Howard
Cox
2254 David 11.1 7.3
IITRI
IITRi(b>
Howard 40.2 26.3
IITRI
Howard 19.1 12.5
IITRI
Howard
1627 David
32
1331 Kaufman 1.7 1.2
Cox
IITRI
IITRI(b>
-------
produced more CO,-, than the reference reaction OH + CO. This
revelation was disappointing since it precluded further use of
this technique .
IMPLICATIONS WITH RESPECT TO ATMOSPHERIC LIFETIMES
OF THE KINETICS OF THE OH + HALOCARBON REACTIONS
Introduction
The overall objective of the preceding kinetic studies is
to allow estimates to be made of the likely tropospheric life-
times of the halocarbon species studied. In making such a
calculation it is assumed that the only sink for these species
in the troposphere is attack by OH radicals. To date, this is
the only known sink. The tropospheric lifetime for the halo-
carbon species is given by;
xe (tropospheric lifetime) = ^
In this expression k. is the bimolecular rate constant for the
halocarbon + hydroxyl reaction, RX. + OH ->• Products, and
lOHj . is the average OH concentration. Both these quantities
are averaged over their geographic, altitude, diurnal and'
seasonal variation. Of these latter variables , the only one
with an effect on the rate constant k. is the temperature
variation with altitude. The best values to be assigned to k,
and LOHJ will now be considered.
ave
Evaluation of the Kinetic Data
In Table 5 all the currently available data on the rate
constants for the RX. + OH reactions studied in this investi-
gation are presented. A value of k (selected) was obtained by
weighting the various values in proportion to their assigned
error. The temperature dependence of the rate constant has
been determined for CH-jCl, CH2C12 , CHC13 and CH-jCCl-j. For the
purpose of the calculations a weighted average tropospheric
56
-------
temperature 265°K is assumed, and the rate constants at 265 K
were calculated based on the quoted activation energies. For
the halocarbons for which the temperature dependency of the rate
constant was not known, an activation energy of 2.07 kcal was
assumed based on a simple average of the values shown in
Table 5.
Tropospheric OH Radical Concentration
The value to be used for COH3Q,yo is somewhat uncertain.
^WG
Several attempts to measure OH radical concentrations in the
35 37
troposphere have been reported. Measured concentrations
f\ — *^
have ranged from 16 to < 4 x 10 molecule cm , the lower
detection limit of the experimental method. In one case, a
8 -3
high value of ^1.5 x 10 molecule cm was reported. Prior
to these measurements, estimates of the tropospheric OH radical
concentration had been made based largely on the then known
chemistry of the atmosphere in terms of sources and sinks for
the trace gas species. Values in the range 3-6 x 10 molecule
- 3
cm were quoted for yearly average concentrations. More
recently lower values have been proposed; ~7 x 10 molecule
—3 38
cm based on the atmospheric CO distribution (this estimate
was made before the pressure dependence of the CO + OH reaction
was established which would have the effect of decreasing
5 -3
further this estimate) and 2-6 x 10 molecule cm based on
8 39
the halocarbon sources, sinks and atmospheric abundances. '
In summary, although two direct measurements have been made
of OH radical atmospheric concentrations in which relatively
high values of [OH] were reported, there appears to be a body
of evidence developing which strongly suggests that concentra-
5 -3
tions in the range 2-6 x 10 molecule cm are probably close
to reality and will be assumed in the following calculations.
57
-------
Halocarbon Tropospheric Lifetimes
The results of the tropospheric lifetime calculations based
on the above rate constants and OH radical concentrations are
shown in Table 6. It is instructive to compare the computed
lifetime in Table 6 with those obtained from measurements of .„.
CH^CClo and CHC1~ emissions, and present atmospheric loading.
Values of 7.2 + 1.2 and 1.7 4- 0.4 years, respectively, were
~ ~~ 8 39
obtained for CH-jCCl-j and CHC13. ' y These values are in fair ;
agreement with those calculated in Table 6 suggesting an average
5 -3
OH radical concentration of the order of ~4 x 10 molecule cm
38 39
in good agreement with the latest estimates, '
Generally if the tropospheric lifetime of a chemical species
is 10 years or less, its flux into the stratosphere is reduced
to 10% or less of the total ground level emissions. With the -••
possible exception of CHUCClo, it appears that none of the
•J J ••
halocarbons listed in Table 5 will contribute significantly to
the stratospheric chlorine burden as intact halocarbons.. Even
for CH~CClo, at its present levels in the atmosphere, 98.8 + 9.7
j j ~
ppt (v/v),8 it is doubtful that it could contribute any more
chlorine to the stratosphere than CHQC1 which is present, largely
8
from natural sources in the troposphere, at 713 + 51 ppt (v/v) .
However, if emissions of CHoCCl^, which it seems have been
established as entirely anthropogenic in origin, were to continue
to increase, this compound could possibly become a cause for
concern.
58
-------
Table 6
CALCULATED TROPOSPHERIC LIFETIMES
vo
Compound
CH3C1
CH2C12
CHC13
C2H5C1
CH2C1CH2C1
CH0CC10
k (265°K)xl014
3 -1 -1
cm molecule sec
2.6
9.0
7.3
26.3
12.5
1.2
Lifetime
[OH]=2xl05
-3 •
molecule cm
6.1
1.8
2.2
0.6
1.3
13.2
Yrs
[OH]= 6x10
molecule cm
2.0
0.6
0.7
0.2
0.4
4.4
5
-3
-------
SECTION 6
KINETICS OF THE HOMOGENEOUS GAS PHASE HYDROLYSIS
OF CC13COC1, CHC12COC1, CH2C1COC1 and COC12
INTRODUCTION
When this study was initiated there were no kinetic data
in the literature on the homogeneous gas phase hydrolysis of the
three chloroacetyl chlorides. Data were available for the
liquid phase hydrolysis of CH-COC!40 in 25% H«0 + 75% acetone
•3 £t
mixtures which indicated that this reaction was quite fast.
A pseudo first order rate constant of ^25°c = ^'^ x ^"®~ sec"
was reported. The homogeneous gas phase hydrolysis of phosgene
has been reported in the temperature range 220-420°C. At
300°C, k = 10 liter mole" min~ , with an activation energy
of 12.02 kcal mole . The data were obtained in a high tempera-
ture infrared cell, but no error limits were given.
An experimental program was therefore undertaken to obtain
homogeneous gas phase hydrolysis data for the four acid
chlorides. Initially, it was thought that this would be a rela-
ively simple task, but as experience showed, this was not to be
the case for a variety of reasons. A brief resume of the early
unsuccessful studies will be presented, followed by a more
detailed account of the final experimental determination.
PRELIMINARY STUDIES - STATIC REACTOR
Initially an attempt was made to study the hydrolysis
reactions in a static gas phase infrared cell and use IR to
monitor the reactant and product concentrations. A few crude
experiments were tried with a commercially available gas
60
-------
cell. The initial results seemed encouraging though there were
obvious indications that the acetyl-chlorides were attacking
the buna-N, 0-rings in the cell.
A static IR gas cell reactor and suitable thermostated
enclosure were built. The latter was capable of temperatures
up to -1200°C. The gas cell is shown in Figure 12 and was
constructed with the idea of minimizing possible attack by
the acetyl chlorides on the cell materials.
Preliminary tests showed that the acetyl chlorides were
attacking the Viton 0-rings and the greased stopcocks. Several
different 0-ring materials were tried without success.
Although Teflon 0-rings would have been unreactive towards the
acid chlorides they were not tried since it was thought they
would not provide good sealing characteristics between the pyrex
cell body and the relatively soft AgCl windows. To eliminate
this problem the 0-rings were removed and the window - pyrex
seal made using glyptal wax. The grease lubricated glass stop-
cocks were replaced by all glass - teflon valves. With these
modifications the acetyl chlorides could be maintained in the
cell indefinitely.
Experiments were tried on the H20 (g) + CCl^H COC1 (g)
reaction. Initially experiments were made at 25 C for periods
of about 90 minutes with reactant concentrations being measured.
The reaction appeared to be proceeding at a convenient rate,
however, it was not possible to fit the data to any simple
rate equation. Experiments were tried at 65°C and 125°C and
it was found that the reaction rate actually decreased with
temperature. This behavior strongly suggested that heterogeneous
surface reactions were complicating the situation. This con-
clusion was further substantiated when glass beads were added to
the reactor to deliberately increase the surface area and the
reaction rate was found to increase almost in direct proportion
to the added surface area.
61
-------
ON
fO
Ball Joint To Connect Reaction
Cell To Vacuum
AgCI NaCI
Window Window
Stainless Steel
Retaining Plate
Viton O Rings
Threaded Tie
Rod To Pressurize
Retaining Plates
Teflon
Spacer
Stainless Steel
Spacer
Figure 12.
Static Infrared Reaction Cell to Study the Gas Phase Hydrolyses of
COC12 and CClnH3_nCOCl (n = 1, 2 or 3)
-------
From these studies it was possible to conclude that
the homogeneous gas phase hydrolysis of CC19HCOC1 at ambient
1 1
temperature was slow, k £ 0.06 liter mole" sec" and could
not be measured in the static IR cell. It was further noted
that if the homogeneous gas phase hydrolysis rate for CC^HCOCl
was of the same order as that of COC10, a value of k ~10
-1-1 ~~
liter mole sec could be expected at ambient temperature.
PRELIMINARY STUDIES - DYNAMIC REACTOR
Because of the above difficulties, a dynamic flow reactor
system was built in which reaction temperatures up to 500°C
and residence times of up to 20-45 minutes were attainable.
The system is shown schematically in Figure 13.
Nitrogen was used as the carrier gas rather than air to
avoid possible oxidation reactions with the acid chlorides
at the elevated reaction temperature. Two independent N~
carrier gas flow systems were used to obtain gas streams
containing water vapor and the acid chloride vapors. To obtain
a given gaseous concentration of the reactants, the carrier
gas was passed through two thermostated bubblers (temperatures
good to + 0.1°C) containing either the acid chlorides or water.
The two gas streams, saturated with the reactants at a given
bubbler temperature could either be passed directly into the IR
cell (this was the same cell used in the static measurements)
to monitor the initial reactant concentrations or through the
pyrex reactor and then into the IR cell to monitor the reactant
concentration after the reaction. The gases from the IR cell
were passed through a wet test meter so that the total gas flow
through the reactor, and hence contact time in the reactor,
could be determined. All valves in the system exposed to the
acyl halides were of glass-teflon construction. Where necessary
the carrier gas reactant lines were heated.
63
-------
N2 Cilinders
Bubblers For Saturating
N2 Gas With CCIn H3_nCOCI
( n» 1,28 3 )
Needle
Valves
Flow
Meters
\
Thermostated
Oil Baths
i
^
1
r~~
K i
*• i
i
i
i
Thermostated
Oven O8 To 50O8C
Thermostated
Oven I25*C
(\
\
1 '
(
(
1
'-[
1
1
1
J
• ta
i
f~"
i
, 1
M
\*~~ ~
\
. ,\
v\
Coiled 1
Reactkx
^ Valves
3yrex
i Cell
"~l
I
. .'
1 <
j
l~'
|
'•1
1 |
L_.
I.F
Ce
Wet Tes
Flow Met
1
1 '
* '»
it r
i
To Ati
Bubblers For Saturating
N2 Gas With H2O
Figure 13. * Dynamic Reactor System to Study the Gas Phase Hydrolyses of COC1« and
CClnH3_nCOCl (n = 1, 2 or 3) 2
-------
The reactor initially had a volume of about 600 cc,
though later this was increased to 3000 cc, and was of 18 ram
ID pyrex tubing wrapped in a spiral with a diameter of about
6". The thermostated oven was capable of temperatures up to
500°C and could be controlled to + 0.5°C. The IR cell was
thermostated at 125 + 0.25°C. Under the conditions used in
these experiments, with a gas flow through the system of 30 to
500 cc minute, the amount of reaction occurring in the heated
carrier gas reactant lines and within the IR cell was negligible
compared to that occurring within the reactor proper.
The COClo/ N + H00/ N reaction was studied between 295 and
2(g) 2 (g)
450 C in the reactor. The rate data appeared to be dependent
on which reactant was in excess, and two temperature dependent
equations for the rate constant were derived;
—8120 -1 ~1
For excess H^O, k « 101 e £* liter mole i sec and
For excess COC10, k - 19.8 e "7£2° liter mole"1 sec"1
Next, the reaction CCl^HCOClx v + ^0, v was investigated
between 215-325°C. The temperature dependent equation was as
follows:
k - 24.3 e () liter mole"1 sec"1
Some preliminary investigations were also made with
CC1H2COC1 + H20 and CCl-jCOCl + H20, and the initial results
indicated similar behavior as in the CCl^HCOCl + H^O system.
At this point experimentation was stopped for the following
reasons .
Data precision was not good. This was due to at least to
two problems. The full temperature range of the reactor could
not be used due to decomposition of the acetyl chlorides at
65
-------
elevated temperature as evidenced by carbon deposits on the
walls of the reactor. All of these compounds started to
decompose in the 290-325°C temperature range, This limitation
on the usable temperature range necessitated extending the
contact time by reducing flow rates, and difficulty was
experienced in controlling and measuring these rates accurately.
In the two systems studied in detail, COC10 and CH0C1COC1, the
-1 -1
pre-exponential factors were
-------
Hg
Monometer
y
5 Liter
Reaction Flask
To Vacuum
Pump
By pas*
Capillary
Tubing
HtO Bubbler
And
Thermostoted Bath
Oven
Acid Chloride Bubbler
And
Thermos to ted Bath
Supply
Figure 14. Schematic of the Reactor System Used to Study the Gas Phase Hydrolysis
of the Chloroacetylchlorides
-------
was mounted in a thermostated Labline oven, capable of operating
up to a temperature of 350 C and stable to +2°C.
Connected to the reactor via the filling and evacuation
port was the reactor gas delivery system. This system consisted
of a vacuum pump for evacuation of the reactor, a mercury
manometer for pressure measurements, and two thermostated
bubblers, controlled to + 0.1°C, for gas saturation. Both
water vapor and the acetyl chloride vapors were delivered by
flowing Linde prepurified grade N« through the thermostated
bubblers and into the reactor. Flow rates for the N« carrier
gas were measured using calibrated Matheson flow meters. ,COC12
was delivered by bleeding a small amount of COC12 into the N2
stream. The COC19 flow was also measured using a calibrated
t.
Matheson flow meter. CC1H2COC1 and CC13COC1 were obtained from
Eastman, CC1«HCOC1 was obtained from Fluka AG, and COC19 was
£* £*
obtained from Matheson. The acetyl chlorides were used without
further purification. COC12 was liquified with a dry ice-acetone
slush and "pumped on" to remove impurities. H20 was laboratory
distilled.
The reactor, at the selected temperature, was filled by
removing the sampling port septum and flowing the reactant gases
into the reactor and out via the sampling port. This process
continued for about 1/2 hour. The flows were then stopped, the
reactor sealed by replacing the sampling septum, and sampling
begun.
The acetyl chlorides were quantitated using a Packard Model
427 gas chromatograph with a flame ionization detector. This
was fitted with a 3 ft, 2 mm I.D. glass column packed with 10%
SE-30 on Chromosorb WHP, 80/100 mesh. The flow rate was set at
30 ml/min. The injector port and detector were held at 150°C
and the oven run isothermally at 80°C for CC1H2COC1 and
CC12HCOC1, and at 100°C for CCl-jCOCl. Sample injection was
accomplished with a Precision Sampling Series A-2 syringe fitted
68
-------
with a 12" "side port" needle. The long sampling needle was
necessary in order to remove samples from the center of the
reactor. Calibration of the gas chromatograph for the acetyl
chlorides was accomplished by injecting gaseous samples of known
volume and concentration. These calibration gases were prepared
by saturating a slow flowing nitrogen gas stream through a liquid
bubbler maintained at a known constant temperature (usually 0°C).
Vapor pressure data were not available in the literature for any
of the acetyl halides, and so these data were generated using
standard vacuum line techniques. The resulting vapor pressure-
temperature curves are shown in Figures 15, 16 and 17.
In the experiments involving COC12, direct monitoring of
the phosgene was not made due to difficulties in reactant
sampling and calibration of the chromatograph. For the 1^0 +
COC19 -»• C09 + 2HC1 reaction, C00 was monitored instead on a
*L £• jL
Hewlett-Packard Model 5710A gas chromatograph with a thermal
conductivity detector. This was used in conjunction with a 6 ft,
1/4" O.D. stainless steel column packed with Chromosorb 102,
80/100 mesh. The helium carrier gas flow rate was set at 40
ml/min; the injection port at 100°C; the detector at 200°C; the
detector current at 200 mA, and the oven was run isothermally
at 70°C. A premixed gas from Matheson containing 0.0498% C02
in air was used for calibration.
Initially H^O was to be monitored using gas chromatography,
but high humidity coupled with the tendency of H20 to stick to
the syringe wall made analysis difficult and time consuming.
As an alternative, the efficiency of the 1^0 bubbler in the gas
delivery system was determined and the concentration of 1^0 in
the reactor was calculated from relative flow rates of the two
reactant gas streams on filling. The bubbler efficiency was
determined by careful use of the gas chromatograph.
69
-------
(OOOr-
100
o
a
o
10
P(Torr)- ( 2.59X I0a)exp (-4826/T)
I 1 i J J _L I I I I I
2.6 2.8 3.0 3.2 3.4 3.6 3.8
TH x)09 (oK)
Figure 15, Vapor Pressure Curve for Chloroacetyl Chloride
70
-------
1000 r
100
o
a
o
10
P(Torr)-(l.80X I0a) txp(-47!6/T)
I I I I I I II II I I I I
2.6 2.8
3.0 3.2
THX|0*(»
3.4 3.6 3.8
Figure 16. Vapor Pressure Curve for Dichloroacetyl Chloride
71
-------
lOOOr
100 -
o
a
o
2.8
3.0 3.2 3.4
T"'x 10s CK)
3.6 3.8
Figure 17. Vapor Pressure Curve for Trichloroacetyl Chloride
72
-------
The same chromatograph using the,same column and conditions
as that described for COj, was used for l^O analyses with the
exception that the column was run isothermally at 100°C. The
chromatograph (was iiinjestedcwifch several samples of N^ saturated
with H^O, with each injection followed by several injections of
N2 alone. In this way, the immediate•H«0 background of the
syringe could be determined and the chromatograph was calibrated.
Using the same procedure, samples of N2 + H^O were taken from
the gas delivery bubbler with the bubbler flow rate the same
as that used in an experiment. From these data, the bubbler
efficiency was calculated at 65%.
In a typical experiment when filling the reactor, the flow
rate through the H^O bubbler was set at 700 ml/min, and the
flow through the acetyl chloride bubbler was set at 50 ml/min.
In the 1^0 -4- COC12 experiments, a COCl^ flow of approximately
15 ml/min was bled into a N« stream of about 275 ml/min.
Depending on reaction rates, experiments usually lasted
from 24 to 100 hrs with a sampling frequency of 20 to 60
minutes (nighttime excluded).
Results and Discussion
In order to interpret kinetically the data from the
hydrolysis experiments on the acetyl chlorides and phosgene,
the following three reactions were considered as possibly
occurring;
A + B -»• C + D kx
A + B + C + D k2
A -»• C + E k3
where A = acetyl chloride or phosgene
B = H20
C = HC1
D = Chlorinated acetic acid for the acetyl halides
and C02 for phosgene
73
-------
E = unknown.
Reaction 1 represents the assumed bimolecular hydrolysis
reaction. That this reaction is second order in reactants has
41
been established for COC12 +• HLO by the Russian workers. For
the acetyl halides, the reaction was simply assumed to be second
order. Reaction 2, the reverse of reaction 1, was included
since for the acetyl halides the experimental data indicated that
an equilibrium was being established under the conditions being
used. Finally, reaction 3 was included, since there was evidence
•••'t
for the mono and di-chloroacetyl chlorides, at the highest
-------
Case II: Reactions #1 and #2 proceeding, A monitored, initial
B measured.
kr(A) (B) - k2 (C) (D)
Concentrations
A B CD
at time » 0 a b 0 0
at time = t x (b-a+x) (a-x) (a-x)
k^x) (b-a+x) - k2(a-x)2
Take several points along A vs t curve
and obtain a best fit to Z » Ax + By
where: Z - dA/dt
B - k2
x » (x) (b-a+x)
y - -(a-x)2
Case III: Reactions #1, #2, and #3 proceeding, A monitored,
initial B measured, k~ known from separate
experiment
- k2(C)(D) + k3(A)
Concentrations
A
a
B
b
C
0
D
0
E
0
at time = 0
i
at time = t x (b-a+x+y) (a-x) (a-x-y) y
75
-------
~k2(a-x)(a-x~y) + k3(x)
dA/dt + k,(x)(b-a+x) - k9(a-x)2 + kq(x)
___ JL _____ ____ _____ *•*
Solve as for Case II using y » 0 at all points with
the following Equation
- k3(x) = k^xXb-a+x+y) - k2(a-x) (a-x-y)
Using the calculated ki and k2, calculate all values for
y and solve again. Repeat until k, and k9 are obtained
with desired precision.
In practice only Cases II and III were used for the water
vapor plus acetyl chloride reactions. The existence of the
reverse reaction (Reaction 2) was always considered to be a
possibility. In those experiments where it was negligible, k2
approximated to zero. Unfortunately, even when reaction 2 was
significant it was not possible to calculate its value with good
precision. In order to apply the above equations to the
observed experimental data, reactant concentration versus time
was plotted. A computer was used to obtain the best smooth line
fit of the experimental data. A variety of mathematical forms
were tried and an equation of the form y = aex + c was found
satisfactory in all cases. Two typical hydro lysis -time curves
are shown in Figures 18 and 19. The preceding equation was then
used as a basis for calculation of the rate constants using the
equations presented earlier. In Table 7 the experimental data
are presented.
Only Case I was considered for the reaction of COCl^ + H20
-> CC-2 + 2HC1. The measured concentrations of CC^ were plotted
against time and fitted to the equation y = ae + c. Since
76
-------
5.0
4.5
•» 4.O
• 3.5
o
S
^ 3.0
* 2.5
o
8
2.0
1.5
1.0
0.5
0.0
_L
I
I
_L
I
_L
I
I
_L
_L
20XI03 40XI03 6OXIO3 8OX|OS IOOX|OS I2OX|QS
Reaction Time (Seconds)
I40XI03 160XI03
Figure 18. Experimental Data for the Homogeneous Gas Phase Hydrolysis of
CC1H2COC1 + H20 In N2 at 517°K
-------
oo
O.O
0 20XIO3 40xl63 6OXIO* 8OXIQ3 lOOXlO3 I2QXJ03 !4OXfOs I60X|O5 I8OXIO3
Reaction Time (Seconds)
Figure 19. Experimental Data for the Homogeneous Gas Phase Hydrolysis of
COC1 + H0 In N at 544°K
-------
Table 7
EXPERIMENTAL CONDITIONS AND RATE CONSTANTS FOR THE HYDROLYSIS EXPERIMENTS
Kxpt.
_._./.—
1
3
6
8
*5
Temp.
(°K)
' f
533
480
508
517
530
Initial CC1H2COC1
(mole-liter-1)
1.45 x 10"5
2.25 x 10"5
2.10 x 10"5
4.53 x 10"5
2.52 x 10"5
* Thin Is A measurement of k3 which
k.. was
Expt.
1
19
20
23
•^
27
k .. WUH
Expt.
31
32
33
34
37
38
39
40
41
CClHjCOCl + H20 in N
Initial H20
(mole-liter-1)
3.23 x 10"4
3.75 x 10"4
3.51 x 10~4
3.71 x 10"4
2
kl
(liter-mole" l-sec'l)
0.0628
0.00576
0.0204
0.0292
Significant U2?
Yes
No
No
Small
K. - 3.45 x 10" second"
was used in the calculation of kj in experiment 1.
Insignificant at lower temperatures.
2£'
* /
473
503
S53
M9
VH
1 us 1 gn 1 f
Temp.
(°K)
561
533
580
592
611
' 544
592
569
619
Initial CC13COC1
(mole-liter-1)
5.41 x 10~5
3.13 x 10"5
1.10 x 10"5
3.U x 10~5
1.84 x 10"5
CCljCOCl + H20 in li
Initial H20
(mole-liter-li
5.21 x 10~4
4.33 x 10"4
3.82 x 10"4
4.14 x 10~4
4.32 x 10"4
I2
(liter-mole-1-sec"1)
0.00940
0.0264
0.182
0.0436
0.0611
Significant k2?
No
Small
Yes
Small
Yes
Icnnt at the temperatures of these experiments
Initial COClo
(mole- 1 lter-1)
2.17 x 10"4
1.41 x 10~4
2.60 x 10~4
2.47 x 10~4
1.98 x 10"4
1.43 x 10"4
1.77 x 10~4
1.77 x 10"4
1.59 x 10"4
COC12 + H20 In N2
Initial H20
(mole-liter-1)
3.26 x 10"4
3.70 x 10~4
3.18 x 10"4
2.99 x 10"4
2.41 x 10"4
2.70 x 10"4
2.83 x 10~4
3.26 x 10"4
2.77 x 10"4
kl-l -1
(liter-mole -sec )
0.0205
0.0153
0.0465
0.0460
0.0972
0.0260
0.0421
0.0241
0.108
79
-------
H?0 was always in excess, the difference between the concentra-
tion of COj at time = °° and time = 0 is equal to the initial
concentration of COC12. Also, the production of C02 should
exactly correspond to the reduction of COC12. This allowed k,
to be calculated by monitoring .C02 and by knowing the initial
concentration of H20.
CC1H2COC1--
The CC1H2COC1 results are based on four experiments in the
temperature range 480°C to 533°K. k2 was significant only in
the experiment at 533°K. Also at 533°K there was some thermal
decomposition of CC1H2COC1 and this was measured by running an
experiment with only CC1H2COC1 in the reactor and no H20. In
this experiment, HC1 was seen chromatographically as a probable
product of the thermal decomposition. At 517°K thermal decomposi-
tion was insignificant and was ignored.
A qualitative experiment was performed in order to confirm
the existence of the back reaction (2) at the higher temperature.
N2 was passed through a U-tube containing CCIHLCOOH and fed into
the reactor together with HC1 gas. The reactor oven was set at
528°C. CCIFUCOCI was found to be produced at a significant rate,
thus verifying the existence of the back reaction. From the four
experiments, the Arrhenius expression for the vapor phase reaction
of CC1H2COC1 + H20 was as follows:
k = (1.14 x 108) exp (-22,630 + 780/RT) liter-mole"1-sec"1
The data points for the temperature dependence are shown in
Figure 20.
CC12HCOC1--
Hydrolysis experiments for CC12HCOC1 in the temperature
range 476° to 546°K indicated that very little hydrolysis of
the compound was occurring and that chemical equilibrium in
the system must have been close to the starting conditions.
80
-------
00
«
o
1.0
-1.5
-2.0
-2.5
-3.0
-3.5
-4.0
-4.5
-5.0
-5.5
-6.0
-6.5
55O*K
525 *K
50O »K
1.80
I
I
I
I
475 *K
r~
I
1.84 1.88 1.92 I 96 2.OO
T"'x 10s ( T in *K)
2.04
2.O8
2.12
Figure 20. Arrhenius Plot for the Gas Phase Reaction CCIH^COCI + H20 In N2 at a
Total Pressure of 760 cm Hg
-------
The latter in terms of reactant concentrations were similar to
those used with CClflUCOCl. Because of the small extent of
reaction, the experimental precision did not warrant any attempt
to fit the data to a given kinetic scheme. Attempts to study
the reaction outside the above temperature limits either
resulted in reactant decomposition or on excessively slow
reaction.
CC13COC1--
The results for CCloCOCl are based on five experiments in
the temperature range 437° to 553°K. k2 was significant in the
two experiments at 533°K and 553°C, and was insignificant in
the lower temperature experiments. In all cases, k« was never
important enough to have a significant effect on the value ,of
k-, (i.e., k, could be calculated using either Case I or Case II
with little effect on the final answer). Thermal decomposition
of CCloCOCl was not significant in the temperature range of
these experiments. The Arrhenius expression for CCloCOCl + H20
is as follows:
kL = (2.54 x 106) exp (-18,350 + 1,750/RT)
The data points for the temperature dependence are shown in
Figure 21.
t.
coci2--
The results for COC12 + H20 are based on a total of 9 ex-
periments in a temperature range of 533 to 619 K. These data
were treated by the Case I equation, since thermodynamically,
the back reaction is known to be unimportant. Thermal
decomposition was not important in the temperature range of the
experiments. The Arrhenius expression for COC12 4- H20 is as
follows:
82
-------
00
-I.O
- -1.5
~ -2.0
o
E -3.0
fc
5 -3.5
-4-0
c -4.5
-5.0
55O-K
-5.5
525-K
50O-K
I I I I II I I
1.80
J I
1.84
1.88 1.92 1.96 2.OO
T"'x|0s (T in -K)
2.04
475 *K
2.O8
2.12
Figure 21. Arrheritus Plot for the Gas Phase Reaction CClgCOCl + H20 In N2 at a
Total Pressure of 760 cm Hg
-------
, kx = (9192) exp (-14,200 + 2100/RT) liter-mole~1-see~1
The data points for the temperature dependence are shown in
Figure 22. The present result is in satisfactory agreement with
the expression obtained by Gaisinovich and Ketov of:
k^ - (6250) exp (-12,020/RT) liter-mole mole" -sec**
IMPLICATIONS WITH RESPECT TO ATMOSPHERIC LIFETIME OF
THE HOMOGENEOUS GAS PHASE HYDROLYSIS DATA
The primary object of the present study was to determine
the tropospheric and stratospheric lifetimes of the acetyl
chlorides and phosgene due to homogeneous gas phase hydrolysis.
16 '
Using the usual formula these lifetimes are tabulated for
the assumed concentrations and temperatures in Table 8.
Although there is a considerable uncertainty in the calculated
lifetimes, a result of having to extrapolate values for the rate
constants to considerably lower temperatures than that at which
they were determined, there is no doubt that the lifetimes of
these species due to homogeneous gas phase hydrolysis in the
troposphere are indeed substantial and that unless other
mechanisms are available for their removal there is no question
that these species will all find their way into the stratosphere
in large amounts. Kinetic data were not obtained for CC^HCOCl
due to difficulties noted earlier but it is reasonable to assume
that this compound will probably have a lifetime in the same
range as the mono and trichloroacetyl chlorides. There is also
no question that in the stratosphere, homogeneous gas phase
hydrolysis will not occur except on a time scale which is
absurdly large.
The present data leave no doubt that the homogeneous gas
phase hydrolysis of these halides is not a viable sink for these
compounds in the atmosphere. For this reason some preliminary
84
-------
620eK
t
•o
o
u
oo
Ol
-2.2
-2.4
-2.6
-2.8
-3.0
-3.2
-3.4
-3.6
-3.8
-4.O
-4.2
-4.4
1.60
6OO-K
I
580'K
l
56O*K
54O*K
1.64
L68
1.72 1.76
T~'x IOS{ T in
I.8O
1.84
1.88
1.92
Figure 22. Arrhenius Plot for the Gas Phase Reaction COC12 + H20 In N2 at a
Total Pressure .of 760 cm Hg
-------
Table 8
CALCULATED ATMOSPHERIC LIFETIMES FOR CC13COC1, CClH2COCl AND COC12
At Sea Level
T = 298'K
Atm. P » 760 Torr
Acyl Chloride or Phosgene = 1 ppb (v)
HJD Vapor =10 Torr
t 1/2 (CC1H2COC1) - 13,100 yrs. (range - 7.600 to 22,680 yrs.)
t 1/2 (CC13COC1) - 455 yrs. (range = 130 to 1595 yrs.)
t 1/2 (COC12) = 113 yrs. (range = 20 to 630 yrs.)
In the Stratosphere
T - 220°K
Atm. P - 50 Torr
Acyl Chloride or Phosgene = 1 ppb (v)
H20 Vapor = 1 ppm (v)
t ,,, (CC1H0COC1) = 1.4xl015 yrs. (5.0 x 1014 to 3.8 x 1015 yrs.)
'^ 1 ? 11 11
t j/2 (CC1-COC1) = 4.0 xlO L yrs. (4.0 x lO11 to 4.0 x 10 J yrs.)
t II2 (COC12) =8.2 xlO10 yrs. (4.2 x 109 to 1.6 x 1012 yrs.)
86
-------
experimental studies were made on the heterogeneous hydrolysis
or "rain-out" of these compounds in the atmosphere since this
mode of removal could represent another possible sink for these
compounds in the atmosphere. These experiments will be
described in Section 7 of this report.
87
-------
SECTION 7
KINETICS OF THE HETEROGENEOUS HYDROLYSIS OR "RAIN-OUT"
OF THE ACID CHLORIDES
INTRODUCTION
In Section 6 of this report experimentation showed that the
homogeneous gas phase hydrolysis rate of CC1H2COC1, CCl^HCOCl,
CCl-jCOCl and COCl^ to the acids in the atmosphere was not an
effective removal process and that these species would have
long atmospheric residence times as the acid chlorides and
would certainly, if no other destruction process were present,
find their way into the stratosphere and there add to the total
chlorine burden.
It was therefore necessary to consider other mechanisms by
which these compounds could be removed from the troposphere.
One obvious possibility is the heterogeneous hydrolysis or
"rain-out" of these species in the lower troposphere. There
has been much effort spent on the problem of precipitation
scavenging rates for trace gases in the atmosphere. The under-
standing of the problem is complicated by a number of factors
including reversible sorption behavior, liquid phase mixing
and chemical reaction. A rather thorough treatment of the
43
problem has been given by Hales, largely in terms of the
macroscopic properties of the system, e.g., diffusion coefficients,
functional solubility relationships, Henry's Law constant,
individual mass transfer coefficients, overall mass transfer
coefficients, properties of the liquid film-gas interface etc.
Unfortunately application of the theory to specific cases is
difficult because in general the physical properties required
88
-------
to apply the theory are not available, hence its practical
utilization is limited.
In view of the above difficulties, a somewhat different and
simplified approach to the problem of rain-out has been taken
based on the kinetic theory of gases. With this approach a
single "wash out parameter" or effectively a sticking coeffi-
cient is determined experimentally as a function.of a few
experimental variables. Based on the experimental value of the
sticking coefficient an estimate is made of the likely efficiency
of the "rain out" process for the acetyl chloride removal.
THE SIMPLE RAIN-OUT MODEL
All the aeetyl halides and phosgene undergo hydrolysis in
contact with water. There have been several studies in the
past on the rate of hydrolysis of phosgene. In all of
these, the phosgene was dissolved in water miscible organic
solvents, water added, and the hydrolysis rate determined.
Under the conditions of the experiments a 1st order rate
constant at 25 C for phosgene hydrolysis was obtained,
—2 -1
k = 1.3 x 10 sec . Data for the chloroacetyl chlorides are
not available, but for acetyl chloride at 25°C a 1st order
-1 -1 47
hydrolysis rate of 8.6 x 10 sec was reported. In the case
of phosgene there, was no evidence that the HC1 formed in the
hydrolysis reaction acted catalytically. There were indications
48
that the reaction is base catalyzed.
In terms of the present problem it is not obvious how these
rate data can readily be used to determine rain out rates. The
above rate data were determined under conditions different from
those likely .to be applicable to the gas phase - rain drop
interaction occurring in the atmosphere.
In order, therefore, to make some meaningful estimate of
rain out rates with a model amenable to simple experimental
application the following approach was taken. It was assumed
89
-------
that the rate of removal of the acetyl chlorides from the gas
phase by -contact with a rain drop will be proportional to the
total number of collisions of the acid chloride molecules with
the water droplet while the latter is traversing an atmosphere
containing the acid chloride vapor. From gas kinetic theory,
the following expressions may be derived easily:
k,
or
ALAcidJ » k1 N(^r)'5 A At
where
k1 = effectively a sticking coefficient, i.e., the
fraction of total collision between the acid
chloride molecules and the water droplet surface
which result in reaction
_o
N = No. of acid chloride molecules cm
k = Boltzmann's Constant 1.38 x 10 erg deg"1
T = Absolute temperature
m = Mass of one molecule of acid chloride =
MW/6.0225 x 1023
_2
A = Surface area of one drop in cm
At = Time in seconds for the drop to traverse the gas
phase containing the acid chloride vapor
Al AcidJ = No. molecules of hydrolyzed acid chloride (free acid)
formed per drop- of water in a time, At.
Kinetically the above model is equivalent to a 1st order
reaction, with the rate of hydrolysis being independent of the
water concentration. At the expected low concentration of
these trace species in the atmosphere this would appear to be a
90
-------
reasonable model. A further assumption of the model is that
collisions resulting in hydrolysis also result in the hydrolyzed
molecule remaining in the water droplet and not re-evaporating.
At the expected low concentration of the acetyl chloride species
in the atmosphere, and therefore low droplet concentration,
Raoult's Law would predict very low vapor pressures for the
dissolved acid. An experimental program was initiated to obtain
values of k', the sticking or accommodation coefficient, for the
organic acid chlorides.
EXPERIMENTAL
A schematic diagram of the simple experimental apparatus
used to determine the sticking coefficients of the acid chlorides
is shown in Figure 23. Droplets of distilled water were formed
on a glass dropper (1 ml volumetric pipette in most experiments)
at the rate of about 20-30 per minute. These droplets were
allowed to fall through a known gaseous atmosphere and collected
in suitable pyrex glass collector tubes. The collector tubes
were cooled with liquid nitrogen to minimize splashing and keep
the vapor pressure of the water essentially zero. Droplet size
(assumed spherical) was determined by weighing a known number
of droplets. With a given dropper, drop size appeared to be
reproducible to <+!%.
A pyrex tube with a 25 mm ID was used to contain the known
gaseous environment through which the droplets were allowed to
fall. It had three sections joined by ball joints. A long
central section (~.160 cm in most experiments and ~80 cm in a
few expts.) and two shorter sections, one containing the liquid
droplet former and the second the collector assembly.
Known gaseous atmospheres were introduced into the dropper
column as shown in Figure 23. Gas streams A and B consisting of
pure He (Linde Zero Grade) and in a few experiments Air (Linde
Zero Grade) were metered into the column at known flow rates
(stream A ^ 240 cc/min and stream B ~_ 50 cc/min). The purpose
91
-------
v£>
N>
I ML Volu-
metric piper
6OO ML
Flask with
Distilled
Water
Acid Chloride Bub-
blers And Therm-
ostated Bath
^J
\-J
\J
Rotating Col lector
With Removable
Gloss Collecting
Tubes
Figure 23. Schematic of "the Acid Chlorides-Water Droplets Experiment
-------
of these gas flows was to prevent diffusion of the acid chloride
vapors into the collecting and droplet dispensing systems. Gas
stream C contained the acid chloride vapor at a known pressure,
generated by saturation bubblers in a thermostated bath.
In most experiments the thermostated bath was held at 0°C, on
a few occasions temperatures of ~_ 20°C was tried. Gas flows of
^200 cc/min" were used. All three flow meters were calibrated
against a wet test meter. The vapor pressure of the acid
chloride in the dropper column was calculated from the known
vapor pressure temperature curves presented in the previous
section, and the relative gas flow rates. Prior to every set
of experiments, careful tests were made to ensure that contami-
nation of the droplet dispensing and the collecting systems
with the acid chlorides was not occurring. Prior to collecting
drops for analyses, the gas was set flowing through the system
to allow equilibration and surface effects to be saturated.
Analysis of the water droplets for the acid chlorides was
made chromatographically. Initially a Varian Model 1400
chromatograph equipped with a flame ionization detector was
used. A 6 ft glass column packed with 80/100 mesh Chromosorb
101 was used to separate the water from the acids. This column
had to be preconditioned prior to use by making several injections
with pure water. Much difficulty was experienced with this unit
and analytical precision was not very good. For this reason
the analyses were finally made using a Hewlett Packard Model
5840A Gas Chromatograph fitted with a flame ionization detector
and a 6 ft glass column packed with 60/80 mesh Carbopack B with
3% Carbowax 20M treated with 0.57. H-jPO^. Due to lack of time
this instrument was only used for analyses of CCl-COOH, though
it was suitable for all the chloroacids. It was necessary to
precondition the column with water prior to use.
In Figure 24 a calibration curve is shown for CCl^COOH which
was made using solutions of known concentrations of the acid and
93
-------
21
18
15
o
- 12
X
o
c
o
Cone. -(-I.200XIO'IS)(A)2+(4.672XIO"8)(A) 4- {1.651 X JO"8}
I i i it l f I t I I l I I l t t ii
5OOO IOOOO I5OOO 2OOOO 25OOO 30OOO 35OOO 4OOOO 45OOO 50OOO
Response < Area Units)
Figure 24. CC13COOH Calibration Curve
-------
^1 yl injections. The acid was "Baker Analyzed" reagent grade.
With this arrangement, the lower detection limit corresponds to
an aqueous solution of 5 x 10 molar. In experiments, solution
"•3 "»4
concentrations in the 1 x 10 to 5 x 10 M range were usually
encountered.
Contact times of the drops in the column reactor were cal-
culated based on the usual measuration formula. No attempt was
made to account for gas viscosity or change in droplet shape.
Because of the relatively short drop length employed, terminal
velocity of the drop was not approached. (The radius of the
drops used in the present investigation was of the order of
0.2 cm. These would have an atmosphere terminal velocity of
-1 49
about 900 cm sec . This compares with the value calculated
for the longer drop experiments of ~580 cm sec" .)
RESULTS AND DISCUSSION
Lack of time, partly due to difficulties in developing a
good analytical technique for the dissolved acid chlorides and
also due to instrumental problems, prevented a detailed study of
the "rain-out" kinetics of these species from being, made. A
summary of the data obtained is presented in Table 9.
The first three sets of experimental data on the mono, di
and trichloroacetyl chlorides (Expts. 1, 2, and 3) were obtained
i
using the Varian chromatograph for analyses of the dissolved
acids. The accuracy of these data is uncertain. The results
shown are the simple averages of two separate determinations on
each of the acid chlorides. Due to chromatographic difficulties
it is possible that these results could be in error by a factor
of 5-10. Because of this, it is doubtful if much significance
can be attached to the lower values of the sticking coefficient
reported for the mono and di-chloroacetyl chlorides compared to
the higher value for trichloroacetyl chloride.
95
-------
Table 9
SUMMARY OF DATA OBTAINED IN THE RAIN-OUT EXPERIMENTS
Expt
No.
1
2
3
4
5
6
7
8
9
10
Compound
CClH2COCla
CC12HCOC13
cci3cocia
cci3coci
cci3coci
cci3cocib
cci coci
cci3cocic
cci3coci
CCI COCI
Carrier
Gas
He
He
He
He
He
N2
Air
He
He
He
Pram Hg
Acetyl
Chloride
8.3
8.2
5.4
5.4
2.1
2.1
2.1
2.1
2.1
2.1
Ambient
Temp. °C
20
20
20
20
20
20
18
20
20
20
Drop Surface
2
Area (cm )
0.6729
0.6729
0.6729
0.6729
0.6729
0.6729
0.6729
0.6729
0.4408
0.4408
Contact
Time (sec)
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.47
Sticking Coefficient k'
2.82 x 10~4
2.23 x 10~4
8.07 x 10~4
6.69 + 0.17
1.10 + 0.03
8.27 x 10~5
7.50 + 0.06
7.85 + 0.06
1.38 + 0.23
9.92 ± 0.90
x 10~5
x 10~4
x ID'5
K 10~5
x 10~4
x 10~5
a) Results obtained with Varian chromatograph, all other data obtained with
Hewlett Packard chromatograph.
b) Result from one experiment only.
-4
c) Water droplets formed from an aqueous solution 2.68 x 10 M in CCl^COOH
-------
The remaining experiments shown in Table 9 (Expt. Nos.
4 through 10) were made using the Hewlett Packard gas chromato-
graph for the chlorinated acetic acid analyses. Because of lack
of time, data were only obtained for trichloroacetyl chloride.
However, the analytical procedure was reliable. Generally,
three or four separate experimental determinations were made in
deriving the values of the sticking coefficients presented in
Table 9 and the errors there shown are the standard deviations
of the separate experiments.
There is poor agreement between Expts. 3 and 4 shown in
Table 9 which were run under identical conditions but with
different chromatographic analyses as noted earlier. These
results confirmed the previous suspicions on the unreliability
of the earlier experiments. Before returning to Expt. 4, the
results of Expts. 5 through 8 will be considered since these
were all made with the same pressure of trichloroacetyl chloride
in the vapor phase.
Expt. 5 will be used as a base from which to discuss the
other results. Under the conditions of the experiment a stick-
ing coefficient of 1.10 + 0.03 x 10 was obtained. For a
reactive system like the present one, the apparent observed
sticking coefficient seems rather low. Part of the reason for
this is probably due to diffusion limited gas transport to the
drop. In the proposed simplified model this type of effort is
not taken into account. It appears reasonable to support that
the drop rapidly depletes a region of the gas phase around
itself of the trichloroacetyl chloride, and that further pick-up
by the drop of the reactive chloride is a time dependent diffu-
sion controlled process.
If in fact the process is diffusion controlled then to
some extent, depending on the importance of the diffusion
process, the sticking coefficient obtained in air mixtures
rather than helium mixtures should be smaller since rates of
97
-------
diffusion of the acetyl chloride in air will be lower than in
helium. This is exactly the trend which is observed k'(air) *
7.5 + 0.06 x 10"5 compared to k'(helium) = 1.10 x 10 . The
relative rates of diffusion of the trichloroacetyl chloride in
helium and air is not known, but a factor of at least 3 is
probable. The fact that the apparent sticking coefficient
only decreases by about 25% in air compared to helium suggests
that the diffusion limitations on the scavenging process are of
some significance, though not of overriding importance.
Diffusional processes are relatively slow compared to gas-
surface collision processes. In the present experimental
arrangement diffusion limited processes are over emphasized.
The drop starts out with zero velocity and attains a maximum
value of ^580 cm second. This compares to the situation
existing in the atmosphere where a drop of comparable size
is travelling at its terminal velocity of ^900 cm sec" for most
of its lifetime.
Expt. 8 was run in essentially the same way as Expt. 5
except the water droplets were formed from an aqueous solution
containing 2.683 x 10 M CCl-COCl. This latter concentration
corresponded to typical values found in the rain drops in
Expt. 5. The sticking coefficient of 7.85 + 0.06 x 10"5,
Expt. 8, was lower than that when no chloroacetyl chloride was
present initially. This result strongly suggests th6*'S ticking
coefficient does depend on the concentration of the •chlbroacetyl
chloride in the rain drop, at least at the concentration used
in these experiments. In the atmosphere the chloroacetyl
chlorides are present at very low concentrations, and rain drop
concentrations will probably not attain such high values as
used here. Consequently, this may not be a very important
effect in the real life rain out situation. (As shown later,
-8
rain drop concentrations are probably
-------
sticking coefficient observed in Expt. 4 k' = 6.69 + 17 x 10~5
with P (acetyl chloride) = 5.4 mmHg (measured droplet average
concentration ^5 x 10 M).
In Expt. 9 the drop size was reduced, the surface area
2 2 •
being diminished to 0.4608 cm from that of 0.6729 cm used in
the previous experiments. Within the error limits of the data
k1 = 1.38 + 0.23 x 106 and k1 - 1.10 + 0.03 x 106 for the
smaller and larger drop sizes, respectively, there is no
significant difference in the sticking coefficients.
In Expt. 10 the contact time was reduced from 0.6 seconds
used in all the previous experiments, to 0.47 seconds. The
value of the sticking coefficient at k1 =» 0.92 + 0.90 x 10 is
barely significantly different from that obtained with the
longer contact time, of 1.38 + 0.23 x 10" . In view of a
possible 10% error in the calculated contact times it would be
premature to attach too much significance to this possible
difference.
t<
Finally, and not shown in Table 9, tests were made to
determine if any significant loss of the dissolved acid was
occurring during the drop period. Droplets 2.683 x 10 M in
CCl-jCOOH were allowed to drop through an atmosphere of pure
helium with a contact time of 0.6 seconds. No discernible
differences in acid concentrations before and after the
experiment were found.
IMPLICATION OF "RAIN DROP" EXPERIMENTS WITH RESPECT TO
ATMOSPHERIC WASH OUT OF THE ACID CHLORIDE
i '
Although the above data from the "rain out" experiment must
be regarded as somewhat tentative in nature, lack of time pre-
venting an in-depth study of the phenomenon being made, it is
instructive to make a preliminary estimate of atmospheric wash
out rates ,for CCl-jCOCl.
99
-------
A value of 7.5 x 10~ was obtained for the sticking
coefficient of CCloCOCl in air under the conditions of the
experiment. Compared to the atmospheric situation, this value
may be somewhat too low, due to the relatively high concentration
of acid in the drop, and'somewhat too large due to an over emphasis
"4
of a diffusional contribution to the observed sticking coeffi-
cient. In the absence of any data at this time to correct these
effects, it will be assumed that they contribute equal but
opposite effects, and that the value, k' » 7.5 x 10 , for the
sticking coefficient, is probably not too much in error for the
proposed atmospheric application.
In order to make the wash out estimate the following
conditions will be assumed:
CCl-jCOCl mixing ratio « 0.1 ppb (v)
Rain drops only exist below an altitude of 5 km
All the rain drops have a diameter of 2 mm
An average atmospheric pressure of 60 cmHg and
temperature of 282°K
-I49
Rain drop terminal velocity of 900 cm sec
-2 52
An average yearly rainfall of 100 g cm
With the above assumptions, it is found that 1.16 x 10
-2 -1
molecules cm yr of CCl-jCOCl would be purged from the atmos-
phere. This quantity may be compared with the number of
molecules of CC1QCOC1 in a column of air 5 km high, at a concen-
j 15 _o
tration of 0.1 ppb (v), of 1.03 x 10 molecules cm . Thus a
compound with a sticking coefficient of ~7 .5 x 10 has an atmos-
pheric half life of about 5 months towards wash out by rain in
the above simplified, though not unrealistic model. In the above
calculation the temperature coefficient of the sticking coeffi-
cient has been ignored. If the temperature dependence of the
process is similar to that of the heterogeneous hydrolysis of
100
-------
CH3COC1 and COC12> the above half life would probably be
about 1 year.
These preliminary findings on the rain out rate for tri-
chloroacetyl chloride give reason for concern for the following
reasons:
Although the value obtained for the sticking coefficient
of trichloroacetyl chloride must be regarded as a pre-
liminary quantity it is difficult to imagine that it is
in serious error. It would therefore seem that effec-
tive low values of sticking coefficient might be rather
common for trace gas - rain droplet type interactions.
The present value obtained for CCloCOCl, could be in
error by a factor of 2 to 3, which if in the wrong
direction would result in a half life approaching
3 years. This would probably not result in a signifi-
cant amount of CC13COC1 entering the stratosphere.
R. J. Ciceone, private communication, suggests that
a half life of about 5 years could result in a
significant quantity of a material like CCl-jCOCl
entering the stratosphere, but admits that atmos-
pheric wash out phenomenon are not well understood.
He did point out that at the present time, measured
Cl and OC1 stratospheric concentration levels are a
factor of about 10 higher than current models pre-
dict and that some source of chlorine appears
present which is currently not recognizable.
CCloCOCl is usually regarded as a reactive species
towards hydrolysis as are the mono and dichloroacetyl
chlorides. Existing data on liquid phase hydrolysis
AA
rates for CH-jCOCl and COC12 suggest that the latter
is a.factor of 6 less reactive than the acetyl
chloride. If a similar reduction in sticking coeffi-
cient fpr COC12 occurred relative to that of
101
-------
then its atmospheric half life could fall in the
2-18 year region. Trichloroacetaldehyde, another de-
gradation product of the chlorocarbons, due to OH
radical attack, could also have a low sticking coeffi-
cient, since in this case no reaction with the water
droplet:, other than by solvation forces, occurs„
Tempering the above estimates, it must be admitted
that at the present time it is not known whether
chemical reactivity or none-reactivity is associated
generally with larger or smaller sticking coefficients;
or if indeed there is any correlation at all.
The data presented in this section on rain out rates are
preliminary in nature. Assuming the validity of the results,
rain out, in the case of CC1-COC1, would probably be a fairly
efficient removal process. If the sticking coefficients for
the mono, di and trichloroacetyl chlorides follow the same
trend as the gas phase homogeneous hydrolysis rate constants,
smaller sticking coefficients for the former two compounds,
compared to CCloCOCl would be expected. On the same basis
COC12 could possibly have larger sticking coefficient than
CCl^COCl, though heterogeneous hydrolysis data would suggest
the opposite effect. At the present time it is not known
whether such trends apply or not. In addition to the above
compounds the wash out rate for trichloroacetaldehyde should
be known since this compound is produced in amounts comparable
to the acetyl chlorides. This compound does not, of course,
hydrolyze chemically in water, and the value of its sticking
coefficient is not known.
102
-------
REFERENCES
1. Molina, M. J. and F. S. Rowland, Nature 249, 810 (1974).
2. Rowland, F. S. and M. J. Molina, Rev. Geophys. Space.
Phys. 13, 1 (1975).
3. Molina, M. J. and F. S. Rowland, Geophys. Res. Letts.
1, 309 (1974).
4. Wofsy, S. C. and M. B. McElroy, Can. J. Chem. 52, 1582
(1974).
5. Stolarski, R. S. and R. J. Cicerone, Can. J. Chem. 52,
1610 (1974).
6. Proceedings of the Fourth Conference on the Climatic
Impact Assessment Program Ed. by T. M. Hard and A. J.
Broderick, U. S. Dept. Transportation Report No.
DOT-TSC-OST-75-38 (1975).
7. Altshuller, A. P. and P. L. Hanst, Report on the Problem
of Halogenated Air Pollutants and Stratospheric Ozone
EPA-600/9-75-008, 1975.
8. Singh, H. B., Geophysical Res. Letts. 4, 101 (1977).
9. Howard, C. J., J. Chem. Phys. 64, 197 (1976).
10. Howard, C. J., J. Chem. Phys. 64, 4303 (1976).
11. Altshuller, A. P. and P. L. Hanst, Report on the Problem
of Halogenated Air Pollutants and Stratospheric Ozone,
EPA Report No. 600/9-75-008, December 1975.
12. Ung, A. Y. M. and R. A. Bock, Can. J. Chem. 42, 753 (1964)
13. Cox, R. A., R. G. Derwert and P. M. Holt, J. Chem. Soc.
Fara. Trans. I, 72, 2031 (1976).
14. Sie, B. K. T. , R. Simonaitis and J. Heicklen, Intemat.
J. Chem! Kinetics 8, 85 (1976).
103
-------
15. Chan, W. H., W. M. Uselman, J. G. Calvert and J, H. Shaw,
Chetn. Phys. Letts. 45, 240 (1977).
16. Benson, S. W., The Foundation of Chemical Kinetics, McGraw
Hill Brook Company, Inc., New York 1960.
17. Singh, H. B. , Geophys. Research ;Letter 4, 101 (1977),
18. Gorse, R. A. and D. H. Volman, J. Photochemistry 1, 1
(1972/73).
19. Hydrogen Peroxide, W. C. Schumb, C. N. Satterfield and R.
L. Wentworth, A.C.S. Monograph Series, Reinhold Publ. Co.,
New York, pp. 221-227 (1955).
20. Meagher, J. F. and J. Heicklen, J. Photochem. 3, 344
(1974/75).
21. Drysdalel D. D. and A. C. Lloyd, Combustion Rev. 4, 156
(1970).
22. Hock, W., K. Hoyermann and H. G. Wagner, Internat. J. Chem.
Kinetics Sym. NOI 329 (1975).
23. Bauch, D. L. , D. D. Drysdale, D. G. Home and A. C. Lloyd,
Evaluated Kinetic Data for High Temperature Reactions,
Vol. 1 CRC, International Science Series, Cleveland,
Ohio, 1972.
24. Greiner, N. R., J. Chem. Phys. 53, 1070 (1970).
25. Perry, R. A., R. Atkinson and J. N. Pitts Jr., J. Chem.
Phys. 64, 1618 (1976).
26. Davies, D. D., G. Machado, B. Conway, T. Oh and R. Watson,
J. Chem. Phys. 64, 197 (1976).
27. Cicerone, R. J. , R. S. Stolarski and S. Walters, Science
185, 1165 (1976).
28. Atkinson, R., D. A. Hansen and J. N. Pitts, J. Chem. Phys,
63, 1703 (1975).
29. Cox, R. A., R. G. Derwent, A. E. J. Eggleton and J. E.
Lovelocjk, Atm. Env. 10, 305 (1976).
30. Watson, R. T., G. Machado, B. Conaway, S. Wagner and D. D.
Davis, J. Phys. Chem. 81, 256 (1977).
104
-------
31. Chang, J. S. and F. Kaufman, J. Chem. Phys. 66, 4989
(1977).
32. Spence, J. W., P. L. Hanst and B. W. Gray Jr., J. Air
Poll. Contr. Assoc. 26, 996 (1977).
33. Volman, D. H., J. Chem. Phys. 17, 947 (1949).
34. Overend, R. P., G. Paraskevopoulos and R. J. Cuetanovic,
Can. J. Chem. 53, 337 (1975).
35. Wang, C. C. and L. J. Davies, Phys. Rev. Letts. 32, 349
(1974).
36. Wang, C. C., L. J. Davies, C. H. Win, S. Japor, H. Niki and
B. Weinstock, Science 189, 800 (1975). Int. L. Chem.
Kinetics 8, 765 (1976).
37. Perner, D., D. H. Ehhalt, H. W. Putz, U. Platt, E. P. Roth
and A. Volz, Geophys. Res. Letts. 3, 466 (1976).
38. Warneck, P., Planet. Space Sci. 23, 1507 (1975).
39. Lovelock, J. E., Nature 267, 32 (1977).
40. Swain, G. G. and C. B. Scott, J. Amer. Chem. Soc. 75, 246
(1953).
41. Gaisinovich, M. S. and A. N. Ketov, Russ. J. Inorg. Chem.
14, 1218 (1969) (English Translation).
42. JANAF Thermochemical Tables, Dow Chemical Co., Midland,
Michigan.
43. Hales, J. M., Atmos. Envir. 6, 635 (1972).
44. Velasco, J. Rodriguez, Anales. Fis. Y. Quim. 37, 254
(1941), 34, 603 (1936), 34, 905 (1936).
45. Bohme, H., Ber. 74B, 248 (1941).
46. Rona, R., Z. Exp. Med. 13, 16 (1921).
47. Swain, C. G. and C. B. Scott, J. Amer. Chem. Soc. 75, 246
(1953).
48. French Patent No. 42,779 (1933).
105
-------
49. Fletcher, N. H., The Physics of Rainclouds, Cambridge
Univ. Press, 1969.
50. Loeb, L. B., The Kinetic Theory of Gases, Dover Publica-
tions, Inc., New York, 1961.
51. Handbook of Geophysics, United States Air Force,
Macmillan Co., New York, 1960.
52. Hantel, M. and H. Langholz, J. Atmosph. Sci. 34, 713 (1977)
53. The Natural Stratosphere of 1974 CIAP Monograph No. 1,
Sept. 1975, DOT-TST-75-51.
106
------- |