1. Rate Coefficients For Reactions With OH

JOURNAL OF GEOPHYSICAL RESEARCH. VOL. 96, NO. 03. PAGES 5001- 5011 , MARCH 20. 1991

Atmospheric Fate of Hydrofluoroethanes and Hydrofluorochloroethanes: 1. Rate Coefficients For Reactions With OH TOMASZ GIERCZAK 1 ' RANAJIT TALUKDAR 2 ' GHANSHYAM L. VAGHJIANI 3 ' EDWARD R. LOVEJOY ' AND A. 4

R. RAVISHANKARA

5

Aeronomy LBbor4tory, N4tionBI Oce~nic Bn_d Atmo~pheric Administr41ion ,Boulder, ColorBdo 4nd Cooper4tive Instit'IJ-te for ReseBrch an EnvaronmentBI Science~, Univer~ity of Color4do, Boulder

The rate coefficients for the reactions of OH with five halocarbons [CF3 CH 2 F (HFC 134a) CF3CHCIF (HCf'? 124), CF3CHCJ, (HCFC 123), CH 3CHF2 (HFC 152a), and CH 3 CF2 Ci (HCFC 142b )], w~ch are proposed as alternatives to chlorofluoromethanes, have been measured. A p~ed photolysis syst~ and a discharge flow apparatus were used to measure the rate co~ffic•~nts ~etween approx.unately 210 and 425 K. Use of the complementary techniques enabled Jdentifi~t1on of syste~at1c errors and minimization of these errors. The obtained values are compared w1th values previously measured by other groups. Thls data base is used in the subsequent paper to calculate the atmospheric lifetimes of the five compounds.

INTRODUCTION

Rowland and Molin a (197 5] postulated that emission of the chemically inert and man-made chlorofluorocarbons into the Earth's atmosphere could result in a reduction of the stratospheric oz.o ne density. This hypothesis is now widely accepted not only because of the sound physical and chemical bases on which it is founded, but also because of the recorded loss of ozone at high latitudes between the 1970s and today [Watson et al., 1988] and the appearance of the Antarctic ozone hole [Solomon, 1989]. Faced with this evidence many countries are restricting the production of chlorotluorocarbons (CFCs) and even an international agreement, the Montreal Protocol on substances that deplete the ozone layer [United Nations Environmental Programme, 1987), has been reached to curb CFC releases into the atmosphere. The recent finding oJ a very perturbed composition of the Arctic region in the winter/spring time [Fahey et al., 1990; Kawa et al., 1990; Brune et al., 1990] has also hastened the regulations. In addition to destroying ozone these CFCs contribute to the greenhouse warming of the Earth's surface (Ramanathan et al., 1985; Mitchel~ 1989) due to their accumulation in the atmosphere and because they have strong IR absorption bands where C02 is transparent. The greenhouse warming effect has added to tl1e urgency for action and it is clear that alternatives to CFCs will be used in th~ near future. 1_0n leave from Dept. of Chemistry, Warsaw University, Zwirki • 1 W1gury 101,02-089 Warsaw, POLAND. 2 On leave from Multidisciplinary Research Section Bhabha At~Jnic. Res~arch Centre, Bombay-400085, INDIA. ' Umvers1ty of Dayton Research Institute, Air Force Astronautics Laboratory, AL/LSCC, Edwards AFB, CA 93523. 4 Present address: Department of Chemistry, University of California, Berkeley CA 94720. . 5 Also affiliated with the Department of Chemistry and Bier chemistry, University of Colorado, Boulder, CO.

The alternatives to the cltlorofluorocarbons must have the correct physical and chemical properties to make them suitable refrigerants, foaming agents, solvents for cleaning electronics, etc. In addition, they should have a potential to destroy ozone that is less than those of CFCs. It has been found that the partially halogenated (with F and Cl) ethanes do have the appropriate physical and chemical properties to be substitutes for CFCs and can be degraded, at least partially, in the troposphere before they get to the stratosphere. The shorter tropospheric lifetime. also minimizes their contribution to the global warming problem. Therefore many of the partially chlorinated and fluorinated ethanes are being pursued as potential substitutes for CFCs. The shorter tropospheric lifetimes for the substitutes is due to the presence of the H atoms in them which can be abstracted by OH. Since these molecules are saturated alkanes, other oxidants such as 03 and N0 3 will not be effective in degrading these species. The reaction with OH initiates a sequence of reactions that remove the substitutes from the atmosphere. The substitutes are expected to have lifetimes which are short compared to CFC 11, 12, and 13 but not so short as to be local pollutants which generate ozone. Because of the shorter lifetimes the majority of them are removed in the troposphere. However, significant fractions of the emissions are expected to reach the stratosphere. To quantify this fraction, which relates to the potential for 03 destruction by the substitutes, the tropospheric lifetime needs to be well defined. Since the tropospheric lifetime is primarily controlled by their reaction with OH radicals, the rate coefficients for the reactions of OH with the substitutes as functions of temperature are needed. We report in this paper the rate coefficients for the reactions of OH with five halogen substituted ethanes,

(Rl)

H20 + CF3CH F

( HFC13••)

(R2) Copright 1991 by the American Geophysical Union. Paper number 90JD02736. 0148-{)227 /91/90JD-{)2735$05.00

OH + CF3CH2F OH + CF3CHC1F

-+

H 20

+ CF3CC1F

H20

+ CF3CCh

(HCFCn•)

(R3)

OH + CF3CHCl2 (HCFC123)

5001

;.•.

5002 (R4a)

G!ERCZAK ET AL.: OH RATE COEFFICIENTS

OH + CH3CH F2

---+

H 20

+ CH3 CF2

---+

H 20

+ CH2CF2Cl

(HCFCl>2•)

(R4b) (R5)

OH + CH3 CF2Cl (HCFCU2b)

between approximately 210 and 425 K. (HFC and HCFC stand for hydrofluorocarbon and hydrochlorofluorocarbon, respectively, which are the acronyms used in the chemical industry.) In the companion papers we present the UV absorption cross sections and the calculated atmospheric lifetimes (Orlando et ol., this issue] for the same five molecules. All these compounds have been studied earlier either at 298 K or as functions of temperature [DeMore et al., 1987]. A summary of the available data will be presented along with our results in the· section on results and discussion. The reason for carrying out the present study is to minimize the inaccuracies in the data base which will be used for assessing the ozone depletion potentials (ODPs) of these substitutes. The calculated ODPs will be the basis for selecting the substitutes. In this study we have employed two different experimental techniques, the pulsed photolysis and the flow tube, to measure the rate coefficients k1 through ks. These two methods are complementary and have been the main "direct" methods for measuring rate coefficients of radical-molecule reactions in recent years. It is hoped that by using two different methods, systematic errors would be exposed and avoided. This study was carried out concurrently with those of Liu et al. [1990], and we were aware of their results (and viceversa) when the experiments were completed. Also, these results were used in obtaining the recommended rate coefficients published in the World Meteorological Organization (WMO} report [1989]. EXPERIMENT AND DATA ANALYSIS

In both apparatuses used in this study the pseudo firstorder rate coefficient for the loss of OH is measured in an excess of the substituted ethanes by monitoring OH spectroscopically. The rate coefficients k1-ks are all less than 5 x10-14 cm3 molec- 1 s- 1 • Therefore the presence of impurities in the substituted ethanes could lead to erroneously large values for the 'measured rate coefficients and it is essential that very pure haloethanes are used. The identity and level of impurities in the samples used in the present study are listed in Table 1. All the five substituted ethanes were analyzed, using gas chromatography, and supplied by du Pont de Nemours and Company. These gases were used as supplied. The apparatus and the experimental procedures employed to carry out the measurements reported here have been in use in our laboratory for many years and are described in detail elsewhere [Stimpfle et al., 1979; Wang et al., 1987; Voghjiani and Ravishankara, 1989). Therefore only the aspects of the experiments imd data analysis essential for an understanding of the present measurements and the specific modifications made to carry out these experiments are described below. Discharge Flow-Laser Magnetic Resonance The discharge flow-laser magnetic resonance apparatus consisted of a 1.1 m long, 1" I.D., Pyrex flow tube fitted with

a Teflon sleeve on the inside to reduce wall loss of OH and temperature controlled to ± 1 K by flowing a temperature regulated fluid through the outer jacket. The concentration of OH in the flow tube was measured using laser magnetic resonance spectroscopy. The OH (_x2II 3/2• v=O, J = 5/2 ..- J 3/2) transition [Brown et al., 1981] was monitored using a magnetic field of 8.9 kG at 118.8-pm line from a far-infrared· CH 3 0H laser. The detection limit for OH in this system was approximately 5 x10 7 mole~ cm- 3 (S/N = 1) for one scan through the absorption line. The OH radicals were generated by the H + N02 reaction in an excess of N02 ([NOz]o/[H]. 20). A mixture of H 2 (1%) in high-purity helium was passed through amicrowave discharge to generate H atoms. A mixture of N02 (2%) in helium was added to the H atom flow in the side arm reactor. The HCFCs and HFCs were added through a movable injector which had its outer wall covered with a Teflon sleeve. The initial concentrations of reactants used in this study were (0.8-30) x10 10 molec cm- 3 of OH, (1.0 - 200) x10H molec cm- 3 of the substituted ethanes and (3- 20) xl0 16 molec cm- 3 of helium carrier gas. All measurements were made between 420 to 210 K. The concentrations of the reagents in the flow tube were measured using mass flow meters and pressure gauges. Electronic mass flowmeters, calibrated by measuring the time rate of change of the pressure in a known volume, were used for all flow measurements. A capacitance manometer was used for all pressure measurements. The system pressure was in the range of 1 to 5 torr. The linear gas flow velocities in the flow tube reactor ranged from 500 to 2000 em s- 1 • High purity helium (> 99.999%), passed through a liquid N2-cooled molecular sieve trap, was used as the carrier gas. The loss rate of OH was obtained by measuring the OH concentration at various reaction distances (or time). The temporal profile of OH is given by an exponential function,

=

=

[OH]o/(OH]t

= exp(k' t) =

exp(k' d/v)

=

where k' k;[X;] - kw, dis the reaction distance, vis the linear gas velocity in the reactor, k; is the bimolecular rate coefficient for the reaction OH + X; - + prod, and kw is the pseudo first-o~:der rate coefficient for the loss of OH on the outer wall of the injector. Plots of ln[OHJ versus d were made to obtain k'. Then k' was measured at various concentrations of X; to obtain k; from a linear least squares analysis of the k' versus [X;] data. Pulsed Photolysis-Pulsed Laser-Induced Fluorescence The pulsed photolysis apparatus consisted of a jacketed pyrex reaction vessel with an internal volume of....., 150 cm3 which was housed in a vacuum chamber. The reaction cell was maintained at a desired constant temperature (± 1 K) by circulating a thermostated fluid through its outer jacket. The temperature of gases in the cell was "measured by a chromel-alumel thermocouple located inside the reactor just above the volume where hydroxyl radicals were generated. A gas mixture containing the haloethane, the photolytic precursor (H 20, H2 0 2, or HN0 3 ) and a bath gas (approximately 100 torr of He or SFt!) were flowed through the cell with a linear velocity of 3 - 25 em s - 1 • All gas flow rates were measured by electronic mass flow meters. Flow meters were calibrated by measuring the time rate of change of pressure in a known volume. The composition of the gas mixture was calculated from measured mass flow rates and

~

..,.~

..

. :.. ' ·~ ..:.

5003

GJERCZAK ET AL.: OH RATE COEFFlCIENTS

TABLE 1. Level of Impurities in the HFCs and HCFCs Samples Used in the Present Experiments a.s Determined by duPont de Nemours Company via GLC Analyses of the Samples Molecule

Impurity ppmv

Molecule

CF3CHCh (123)

CF2Ch (Fl2) CF3CFCh {114a) CF3CHFCI {124) CF3CHF2 (125) CFaCH2Cl (133a) CF3CH3(143a) CF2CHCl (1122) unknown

401 9 2 41 3 85

CFC}a (Fll ) CHF3 {23) CF3CFCb (114a) CF3CH2Cl (133a) CF3CHzF (134a) CF2CCh (1112a) CF2CFCI (1113) CH2CCh (1130a) CF3CHFz (125) CF2CHF

< 10 10'. If other reactions, such as that between OH and the products of (R1) - (R5), were contributing significantly to the measured OH loss rate, deviations from the exponential temporal profile should become evident. H a significant fraction of the haloethane was photolyzed in the flash photolysis experiments, the concentration of the photoproducts could be much larger than that of OH. Hence one would still observe an exponential OH decay even though the reaction responsible for the observed OH loss is a secondary reaction. Also, in the flow tube, heterogeneous reactions could enhance the decay rate of OH when the gas phase reaction rate is very small. Therefore careful attention was paid to the temporal profile of OH and any deviation from the exponential decay was further investigated. Many experimental conditions were varied to see how the measured rate coefficient changed. In the flash photolysis experiments the photolysis energy, photolyte concentration, system pressure, and the gas flow rate through the cell were varied. In the flow tube studies the initial OH concentration, system pressure, and the linear gas flow velocity through the tube were varied. These variations provided valuable dues to possible complications. Since the experiments were carried out under pseudo first- order conditions in [OH], none of these variations should alter the measured bimolecular rate coefficients. Therefore only when these variations did not change the measured rate coefficient was the value deemed correct and is reported here. The parameters that were varied in measuring k, through k5 are described in the sections dealing with the specific rate coefficients. The major source of error in the reported rate coefficients was the knowledge of the concentration of the haloethane, the excess reagent. The measured mass flow rates and system pressure were used to determine its concentration.

Based on our estimated errors in these parameters, we believe that the error in haloethane concentration is ± 8% at the 95% confidence level. In general, the precision of the measured rate constants were less than the estimated error in the haloethane concentration. The precision was the error given by the linear least squares analysis of the k' versus [X;] data where k' was weighted according to the error in the analysis of the ln[OH) versus t (in the flash photolysis experiments) or d (in the flow tube experiments). All the errors reported in this paper are for 95% confidence limit, and are obtained by root-mean-square addition of the ± 8% estimated error with the measured precision. In many cases the temperature dependence of the bimolecular rate coefficient did not follow an Arrhenius behavior. In general, the Arrhenius plots exhibited a curvature such that the activation energy increased with increasing temperature. For atmospheric purposes only the lower temperature val ues are important and data below approximately 350 K were adequately fit by an Arrhenius expression. Such fits are shown in Tables 3, 5, 7, 9, and 11 where our data is compared with those from other investigations.

OH + CF3CH2F- H20 + CF3CHF, k1 Table 2 lists our data on measurement of k1 at various temperatures. The agreement between the values obtained by the pulsed photolysis and the flow tube methods at room temperature is excellent (better than 3%). Variation of the [OH]o by a factor of 8 in the pulsed photolysis experiments and a factor of 3 in the flow methods had no effect on the measured value of k1 • Variation of the pressure (1 to 5 torr) and linear gas flow velocity in the flow tube did not affect the measured value of the rate constant. Our average value of k"l at 298 K is (4.34 ± 0.35) ~10 - 15 cm 3 molec- 1 s- 1 where the quoted nncertainty is 2cr and includes estimated error. When k1 was measured in the flow tube at 255 K and below, it was found that the first order OH decay plots were curved and the k\ versus [CF3CH2F] plots showed negative intercepts which were beyond the 2cr error calculated from the k' versus [X;] data. These observations along with the difficulty in reproducing the data were attributed to heterogeneous effects and hence no values of k1 measured below 296 K using the flow tube are reported here. (The value of k1 measured in the flow tube at 255 K was within 10% of that in the flash photolysis system, and the [OH) profiles were reasonably linear. However, they were also rejected.) Of course, the rate coefficients measured down to 223 K using the pulsed photolysis method will not be affected by such heterogeneous problems and a.re reported in Table 2. All the measurements at temperatures greater than 298 K were carried out using the flow tube where there were no heterogeneous chemistry problems. Table 3 lists all the available data on k1 to date. The value of k1 at 298 K measured by Jeong et al. [1984] is nearly twice that measured by us. The room temperature value of Clyne and Holt [1979] (quoted by the authors as opposed to that calculated from their Arrhenius expression) is also SO% higher than our value. The values reported by Martin and Paraskevopoulos (1983] and Liu et ol. (1990] are in excellent agreement with ours. The possible reason for the discrepancy between our results and those of Clyne and Holt and Jeong et al. is the presence of impurities in the samples used by them. However, tl1e E/ R values reported by these three groups are very similar as shown in Table 3,

'•

5005

GIERCZAK ET AL.: OH RATE COEFFICIENTS

TABLE 2. Summary of Experimental Conditions and t.he Measured Values of l:1 as a Fwtction of Temperature, OH

Temp

Method

OH Source

K 450 425 404 376 324 298 297 296 294 273 243 223 Notes: a,

DF-LMR DF-LMR DF-LMR DF-LMR DF-LMR DF-LMR DF-LMR FP-LIF DF-LMR FP-LIF FP-LIF FP-LIF H + N02 --+

+ CF3CH2F(HFCt34a).....!... " H20 + CF3CHF

(OH)o X 10- 10 melee cm-3

[HFC134a) Range 101 ~ melee an- 3

0.6 7.7 a a 11.3 0.45 0.9 19.1 a a 13.6 1.0 a 8 1.1 a 6.1 1.0 14.1 3.03 a 5- 40 30 H20 18.7 a 1.4 H20 1.3 100 H20 3 80 0.8 H20 90 OH + NO; and b, Quoted error bars are

Number of Experiments

Buffer Gas/ Pressure, torr

kbi X 10 1!> b cm3 molec- 1 s- 1

3.8 1 31.8 25.0 5.1 1 20.7 6.4 1 14.1 8.4 1 9.0 7.70 1 4.50 18 3 16.7 2 4.35 6 4.26 600 He/100 9.6 1 4.57 2.76 610 1 He/100 1.55 550 1 He/100 0.990 520 1 He/100 2u and include estimated systematic errors.

± ± ± ± ± ± ± ± ± ± ± ±

2.6 2.0 1.8 1.1 0.69 0.37 0.38 0.29 0.33 0.27 0.13 0.085

TABLE 3. Comparison of Measurements of .l:1

8.44 ± 0.73 5.15 ± 0.58 6.9 ± 0.9& 5.18 ± 0.7 4.34 ± o.35b

A x1o12 an 3 molec- 1 s- 1

E/R ± (6E/R)

T

K

K

1.1 ± 0.11

1424±35

250-470

1800± 20 1990±280 1430± 60

294-429 270-400 223-324

!n

3.2 3.7 ± 1.5

Technique

DF-RF FP-RA DF-RF FP-RF

Reference

Jeong et o.l. [1984] M4rtin 4nd P4r4d:etJopoulol [1983] Clyne 4nd Holt [1979) Liu et 41. [1990]

DF-LMR FP-LIF

e value quote by the aut ors. ey not measure kt at 298 K. eir Arrhemus expression yie ds a value of 7.6 x 10- 1!> cm 3 molec- 1 s- 1 , while the correction of their 294 K value to 298 K yields 5.7 x 10- 1 ~ cm3 molec-1 s-1 ; b, Derived by using DF-LMR and PF-LIF data at 298,297, and 296 K. The 297 and 296 K were scaled to 298 K using an E/ R value of 1430 K.

where our results are presented along with those measured by other investigators. This observation suggests that if an impurity which reacts rapidly with OH at 298 K (and therefore has a small activation energy) was present in the samples used by Jeong et al. and Clyne and Holt, their E/ R value should be much lower- Jeong et al. mentioned that purification of their sample had no effect on the measured value of k1, but an analysis of the sample was not given. It is not clear as to why this discrepancy exists. Figure 1 shows the temperature dependence of k1 • It is clear that the Arrhenius plot is slightly curved. Such curvature over extended temperature regions is expected and is often observed [Cohen, 1989]. Jeong et al. have also reported such a curvature. The Arrhenius expression shown in Table 3 was derived from our data obtained between 223 and 324 K. '

OH + CF3CHClF--+ H20 + CF3CClF, k2 Table 4 lists our data on measurements of ~ at various temperatures. Variation of the initial OH concentration by a factor of 2 and the flash energy of the photolysis lamp

by a factor of 4 with water as the photolyte had no effect on the measured rate constant. However, at 243 K in 100 torr of He buffer gas the measured value of ~ increased by approximately 20% when the concentration of HN03, the photolyte precursor, was increased from 3 x10 14 to 3.2 xl0 1 ~ cm- 3. A few experiments at different concentrations of HNOa showed the measured value of~ to increase linearly with the (HN03). The value obtained by extrapolation of this data to zero HN03 concentration agreed with that mea· sured with H20 as the photolyte. It was suspected that the CF3CHCIF reactant was acting as a third body in enhancing the contribution of OH + HN03 + M reaction. (i.e., the loss rate of OH due to reaction with HNOa was not invariant with CF3CHCIF concentration, as assumed, to obtain k2 from k'2 versus [CF3CHClF] plot, but was increasing with increasing [CFaCHClF].) To check for this possibility, k, was measured in 100 torr of SF11 , an efficient third body (which does not quench OH fluorescence signal), to make the rate coefficient for OH + HN03 + M reaction approach koo and hence not change significantly with (CF3CHClF]. Indeed the measured value of~ did not change with [HNOa] . The ob-

, ·•·

~

GJERCZAK ET AL.:

5006

OH RATE COEFFICIENTS

10' r---------------------------------~

·-

•

10.

..(I)

10 2

0

.~

0.6.6

:J

Do

0

Cll

0

0

E

10'

M

E 0

II')

•

.-

0

:,I 10°

10" ~-----------T------------~----------~ 2 3 4 5

2

1000/T1(K)

Fig. 1. Plots In k versus 1/ T for the five compounds studied here. The compounds are marked on the figures next to the appropriate data. The straight lines are the fits of the to an Arrhenius expression over the temperature ranges indicated in the figure and the tables. The obtained values of E/ Rand A are given the tables. For visual clarity we have plotted k4 multiplied by 10 and ks divided by 10.

tained result agrees with that measured using H20 as the photolyte as well as that derived from extrapolation to zero [HN03] in He buffer gas. Our measured increase in the rate coefficient with [HN03] in He suggests that CF3CHClF is a better third body by a factor of approximately 6 than N2. This value has a large error bar but qualitatively explains our observation. Jolly et al. [1985] have contended that the OH + HN03 reaction does not exhibit an observable pressure dependence at 298 K. However, it should be noted that our experiments were carried out at lower temperatures and t he observed increases in the rate coeffieints are rather smalL The other possibility is that our sample of HN03 was contaminated with N02 which reacts with OH in a termolecular reaction. The same kind of pressure dependence could be expected if N02 was present. However, based on our previous experience with handling HN03, we do not believe that we had enough N0 2 to cause the observed variation with photolyte concentration. In any case we carried out all the k2 measurements at low [HN03] to avoid contribution from the third body induced OH + HN03 (or OR + N02) reaction. Variation of the system pressure from 50 to 100 torr of He at 243 K also had flO effect on the value of ~ that was measured. There were no indication of heterogeneous reactions as evidenced by exponential OR decays and zero intercepts (within the precision of t he fi ts) in k'2 versus [CF3CHCIF] plots in the flow tube measurements. The agreement between the two methods is excellent (~ 2%), yielding k2 (298 K) of (9.44 ± 0.75) x10- 1 s cm 3 molec- 1 s-1 where the error is 2q and includes estimated systematic errors. The temperature dependence of ~ is shown in Figure 1 and it clearly exhibits an Arrhenius behavior. There are two previous measurements of ~. one by Howard and Evenson [1974] at 296.K and the other by Watson et al. [1979] between 250 and 375 K which are listed in Table 5. Our Arrhenius parameters were calculated using data obtained between 210 and 347 K. The rate constant

TABLE 4. Summary of Experimental Conditions and the Measured Values of k2 as a Function of Temperature, OH + CF3CHCIF(HCFC124) ~ H20 + CF3CClF

Temp

Method

OH Source

DF-LMR DF-LMR DF-LMR FP-LIF FP-LIF DF-LMR FP-LIF FP-LIF DF-LMR FP-LIF FP-LIF FP-LIF DF-LMR DF-LMR FP-LIF DF-LMR H + N02 -

a a a HN03 H20 a HN03 H20

K

425 400 349 298 297 293.5 272 270 258 244 243 243 238 233 223 210 Notes: a,

(OH]o X 10- 10 molec cm- 3

[HCFC124] Range 101s molec cm- 3

6.4 0.56 7.2 0.59 7.6 0.92 1.5 27 2.4 29 13.8 1.8 2.4 18 2.6 26 15.9: a 1.9 3 .' H20 29 HN03 0.6-2.4 23 HNO~ 1-2.5 44 a 12.8 2.1 a 11.3 2.0 H20 2.4 39 6.1 a 3.4 OH +NO; and b, Quoted error bars are

4.4 5.7 4.6 178 184 7.9 217 199 9.5 213 238 271 10.6 10.3 224 15 2u

Number of Experiments 1 1 1 1 1 1 2

1 1 1 5 2

Buffer Gas/ Pressure, torr

He/100 He/100 He/100 He/100 He/100 He/100 SF6/100

1 1

1

He/100

2

and include estimated systematic erron;.

kbi X 10 1!> b cm3 molec- 1 s- 1

39.1 31.6 18.7 9.00 9.75 8.80 6.61 6.40 5.10 4.14 3.99 3.82 3.46 2.60 2.5 1.93

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

3.1 2.6 1.5 0.75 0.8 0.74 O.i2

0.52 0.42 0.34 0.41 0.38 0.28 0.22 0.2 0.16

,

.........

GIERCZAK ET AL.:

5007

OH RATE COEFFICIENTS

TABLE 5. Comparison of Measurements of k'2

15 k(298K) X 10 cm3 molec- 1 s- 1

A x1013 cm3 molec- 1 s- 1

E/R

± (!:>E/R)

T

I