J. Phys. Chem. 1981, 85, 199-210
190
where k,[S] is the pseudo-first-order rate constant for reaction of a solute with OH and X the track recombination frequency which can be taken as 4.7 X lo8 s-l. This expression should describe the yield of a secondary intermediate that should be initially produced a t a particular solute concentration. It should be used with due caution in that, as pointed out above, loss can occur as a result of subsequent track processes. Equation 10 should provide an upper limit to the observable yield with the lower limit being given by the intercept, i.e., 5.2. In experiments such as measurements of equilibrium constants or extinction coefficients of intermediates produced by oxidation with OH it is desirable to make measurements a t as low solute concentrations as possible in order to minimize any uncertainty which could result from the last term in eq 10. The latter can, for example, contribute up to 7% a t M. Where measurements can be made only at appreciable concentrations some attempt should be made to explore the concentration dependence in order to assess the importance of the track scavenging processes. Acknowledgment. The authors thank Dr. Eberhard Janata for improvements in the optical monitoring equipment which have contributed significantly to the accuracy of the present measurements. They are also grateful to Mr. Terrence Deal for his skill in operating the accelerator as a precision instrument.
in and derived description of the time evolution of the OH population. For the NzO-saturated solutions time delays in the secondary reactions are of the order of 10 ns so that a description of direct observations can effectively be given only after about 20 ns. One does expect from eq 9 that as a result of intratrack processes the OH population should decay from a yield of 6.13 to 5.35 over the period from 20 ns to 1 ps. Separating this decay from the homogeneous second-order reactions of OH is, of course, not easy. We are currently attempting to use techniques similar to those described here to probe the component of OH produced directly from the water in order to describe this latter component more exactly and to provide a basis for comparison of conclusions from the scavenging studies with direct observations such as those by Jonah and Millere6 Because the lifetimes of the OH radicals produced directly will on average be somewhat shorter than those produced in the secondary processes the scavenging function for this direct component is expected to be shifted toward higher concentrations, i.e., the value for a should be somewhat lower. A general equation for the scavenging of OH radicals from NzO-saturated solutions can now be written as
Kinetics Study of the Pressure Dependence of the BrO
+ NO2 Reaction at 298 K
Stanley P. Sander,' Gary W. Ray, and Robert T. Watson Molecular Physics and Chemistry Section, Jet propulslon Laboratory, Callfornla Instnute of Technology, Pasadena, Callfornia 9 1 109 (Received: May 16, 1980; In Final Form: September 5, 1980)
-
-
The kinetics of the reaction BrO + NOz + M BrN03 + M were studied from 1to 700 torr at 298 K. The discharge-flow-mass-spectrometric (1-6 torr) and flash-photolysis-ultraviolet absorption (50-700 torr) techniques were used to monitor the pseudo-first-order decay of BrO (211) radicals in the presence of excess NOz. The reaction was found to be in the third-order region from 1 to 6 torr and in the falloff region between secondand third-order kinetics in the 50-700-torr pressure range. Estimates of the limiting third- and second-order rate constants, ko and k,, were determined by fitting the observed falloff curve (50-700 torr) to a theoretical expression developed by Troe and co-workers. The value of ko derived by this method was in good agreement with the third-order rate constant determined in this study using the technique of discharge flow-mass cm6 molecule-2 spectrometry. Based on these two independent determinations, a value of (5.0 f 1.0) x s-l is recommended for ko (M = N a at 298 K. The stratospheric implications of these measurements are discussed.
(1)(a) P.J. Crutzen, Q.J.R. Meteorol. SOC.,96,320 (1970);(b) H. S.Johnston, Science, 173, 517 (1971). (2) (a) R. S. Stolarski and R. J. Cicerone, Can. J. Chern., 52, 1610 (1974); (b) S. C.Wofsy and M. B. McElroy, ibid., 52, 1582 (1974);( c ) M. J. Molina and F. S. Rowland, Nature (London), 249, 810 (1974). (3)y. L.Yung, J. p, Pinto, R.T.Watson, and S. P. Sander, J. Atrnos. Sci., 37, 339 (1980).
also participate in ozdne destruction cycles. Thk net effect of these temporary reservoir species with respect to ozone depletion can be determined only by solving the coupled equations of atmospheric chemistry and transport. Doing so requires MXY+U"X kinetic data applicable atmospheric conditions of pressure and temperature.
0022-3654/81/2085-0199$01.00/0@ 1981 American Chemical Society
200
The Journal of Physical Chemistry, Vol. 85, No. 2, 7987
The rate constant for the formation of chlorine nitrate has been measured both in its third-order r e g i ~ n ~ (1-7 -~ torr total pressure) and in the falloff region between second- and third-order kinetics%" (25-600 torr). Discrepancies exist between rate measurements obtained by observing the overall rate of decay of C10 as in the discharge flash photolysis,"J2 and molecular modulation studies'O and a technique based on the thermal decomposition of chlorine n i t ~ a t e , leading ~,~ to the speculation8~9JzJ3 that several different ClN03 isomers may be formed. In contrast no rate data presently exist for the formation of bromine nitrate (reaction 1). In this study, BrO NOz M BrNO, + M (1)
+
+
-
we report the rate constant for reaction 1over the pressure ranges 1-6 torr (M = He, N,) and 50-700 torr (M = N2) at 298 K by using the discharge-flow-mass-spectrometric and flash-photolysis-ultraviolet-absorption techniques, respectively. Like the analogous process involving chlorine nitrate, the reaction is found to be in the third-order kinetic region between 1 and 6 torr and in the intermediate second-third-order kinetic region between 50 and 700 torr. Using theoretical methods developed recently by Troe and c o - ~ o r k e r s , ~it~was J ~ possible to use these falloff data to estimate the low- and high-pressure limiting rate constants, ko and k,. The value determined for ko was in good agreement with the experimental determination of ko performed by using the discharge-flow-mass-spectrometric technique. Because only BrO was observed in this study, the issue of isomer formation in reaction 1 could not be addressed.
Experimental Section
Discharge-Flow-Mass-Spectrometric S t u d y . The essential features of the discharge flow-mass spectrometric system are shown in Figure 1 and include a 2.5-cm diameter Pyrex flow tube with 10 fixed inlet jets spaced at 5-cm intervals, interfaced to a modular multistage (three or four) oil diffusion-pumped vacuum chamber.16 A three-stage configuration was employed in this study. Efficient collision-free sampling of the reaction mixture through a series of collinear pinholes (pinhole areas 10.03 cm2),and a low residual background pressure (less than torr), result in high sensitivity for both stable and labile species. Under reaction conditions, an ion source pressure of lo4 torr was used with a background pressure in the masstorr. The detection limit spectrometric chamber of for BrO ( m / e 97) using an Extranuclear quadrupole mass spectrometer operated with an electron energy of 22 eV (4) M. S. Zahniser, J. S. Chang, and F. Kaufman, J. Chem. Phys., 67, 997 (1976). (5) J. W. Birks, B. Shoemaker, T. J. Leck, R. A. Borders, and L. J. Hart, J . Chem. Phys., 66, 4591 (1977). (6) M. T. Leu, C. L. Lin, and W. B. DeMore, J . Phys. Chem., 81, 190 (1977). (7) R. Stimpfle, R. A. Perry, and C. J. Howard, manuscript in prepa-
ration. (8) H.-D. Knauth, Ber. Bunsenges. Phys. Chem., 82, 212 (1978). (9) G. Schonle, H.-D. Knauth, and R. N. Schindler, submitted for publication in J . Phys. Chem. (10) R. A. Cox and R. Lewis, J. Chem. SOC.,Faraday Trans. 1,75,2649 (1979). (11) V. Handwerk and R. Zellner, manuscript in preparation. (12) M. J. Molina, L. T. Molina, and T. Ishiwata, J . Phys. Chem., 84, 3100 (1980). (13) J. S. Chang, A. C. Baldwin, and D. M. Golden, J . Chem. Phys., 71, 2021 (1979). (14) K. Luther and J. Troe, presented at the 17th International Symposium on Combustion, Leeds, Aug. 1978. (15) J. Troe, J . Phys. Chem., 83, 114 (1979). (16) G. W. Ray, L. F. Keyser, and R. T. Watson, J. Phys. Chem., 84, 1674 (1980).
Sander et al.
i M P b CT 1 2d ABBREVIATIONS BN
= BARATRON = R b l I "bI"F
-.
-. .- - . ._. -
P"
CEM CT
= CHANNELTRON ELECTRON MULTIPLIER
CMiT
= COPPER W O O L TRAP
ob
=
CRYOSORPTION TRAP
O ~ OIFFUSION L PUMP = ELECTRON IMPACT IONSOURCE = ELECTROMETER = GATEVALVE MULTICHANNEL ANALYZER =
Ell EM GV MCA
i
NZ PC PI PMT QMF RE RI RL TI TTY
=
21
=
Nz111COOLEDBAFFLE
= PHOTON COUNTER =
PHOTOlONlZATlON ION SOURCE
= PHOTOMULTIPLIER TUBE = =
OUADRUPOLE MASS FILTER RECORDERS
= REAGENT INLET =
RESONANCE LAMP
= TITANIUM SUBLIMITOR = TELETYPE
ZEOLITE TRAP
Flgure 1. Schematic diagram of the discharge-flow-mass-spectrometric system.
without beam modulation and with analog signal processing is -lo9 molecule ~ m - ~All . experiments were carried out at 298 K, and the total pressure was varied from 1.5 to 6.0 torr (M = He) and from 1.5 to 4.0 torr (M = N2). Flow velocities were varied from 700 to 2000 cm S-1.
BrO radicals were produced by reacting atomic bromine with large excess concentration of ozone ([03]o/[Br]o1 103) (eq 2). The BrO radicals were produced close to the Br + O3 BrO + O2 (2)
-
k2(298 K) = 1:12 X
cm3 molecule-' s-l (ref 17)
bromine discharge, and the removal of BrO was monitored mass spectrometrically by adding a large excess of NO2 at each of the fixed inlet jets in turn. Atomic bromine was produced by flowing dilute Br2 in He mixtures through a 2450-MHz microwave discharge. In order to minimize potential kinetic complications from hydrogen and oxygen atoms, the discharge region was uncoated. The upper limit for impurity atom concentration was 5 x 1O'O molecule cm-3 when an uncoated discharge was used in comparison to an impurity atom concentration of N 1Ol2 molecule when a phosphoric acid coated discharge was used. The impurity atom concentrations were determined by using NO2 as a titrant.16 In the kinetics runs where N2 was used as the dominant diluent gas, experiments were performed to test for the presence of atomic nitrogen which could be produced by back-streaming of N2 into the discharge region, even though such back-diffusion would not be expected as the nitrogen was added 10 cm downstream of the discharge region. Nitric oxide was added through the fixed inlet jets and the N+ ( m / e 14) and the NO+ ( m / e 30) signals were monitored (see eq 3). No change was obN + NO N2 + 0 (3)
-
+
cm3 molecule-'^-^ (ref 17)
k3(298 K) = 3.4 X
served in either signal upon initiation of the microwave discharge which placed an upper limit on the atomic nitrogen concentration of 1Olo molecule ~ m - ~Initial . BrO radical concentrations were determined by establishing a known concentration of Br2, [Brz]O,and measuring the change in the Br2+ ( m / e 160) signal (AS) upon initiation
-
(17) D. L. Baulch, R. A. Cox, R. F. Hampson, Jr., J. A. Kerr, J. Troe, and R. T. Watson, J. Phys. Chem. Ref. Data, 9, 295 (1980).
Pressure Dependence of the BrO -k NO2 Reaction
of the discharge. As A S was monitored in the presence of excess O3 ([O,] N 3 x 1014-10 X 1014crn-7, the atomic bromine was rapidly removed close to the discharge region, thus minimizing the possibility of heterogeneous recombination of atomic bromine on the flow-tube walls. However, even a large error (i.e., 25%) in the determination of the initial BrO concentration would be unimportant as the experiments were performed by using pseudo-firstorder conditions ([NOz]o/[BrO]o1 1.1X lo3),and removal of BrO via the BrO BrO reaction is of negligible importance. Thus the BrO concentration can be derived from eq I, where So is the Brz+signal at m / e 160 with the dis[BrOIo = 2(AS[Brzlo/So) (1) charge off. In a separate set of experiments’* it has been shown that heterogeneous removal of BrO on the flow-tube along the walls was too slow to be observable, Le., 6% complete length of the flow tube. The BrO radical concentration was monitored at m / e 97 with an electron energy of 22 eV in order to optimize the BrO+ signal from BrO relative to the BrO+ signal from the dissociative ionization of BrN03. Flow-tube pressures and reagent pressures in the gas handling system were measured at two separate points, including the center of the reaction zone, which under the conditions of this study eliminated the need to correct for viscous pressure drop. All reagent flow rates, with the exception of NOz, were measured by calibrated Teledyne-Hasting Raydist mass flowmeters. In the case of the 03/He and 03/Nz mixtures, the flowmeter was bypassed after the flow-rate measurement, in order to avoid thermal decomposition of the 03. Ozone was generated by passing ultrahigh-purity oxygen through a laboratory ozonizer and was then deposited on a silica gel trap at dry-ice temperature. After impurity Oz was removed from the silica gel trap, the ozone was admitted to the flow tube close to the discharge region by flowing helium or nitrogen through the silica gel trap. With the exception of the helium which flowed through the discharge region (- 10% of the total flow rate) all of the helium or nitrogen carrier gas entered with the ozone. Ozone concentrations were determined mass spectrometrically by titrating the ozone with nitric oxide (reaction 4). Because reaction 4 is relatively slow, the titration was NO O3 NOz Oz (4)
+
+
K4(298K) = 1.8 X
-
+
cm3 molecule-’ s-’ (ref 17)
not taken to the end point. The O3was monitored at m / e 48, and the NOz, for which the mass spectrometer had been precalibrated under identical experimental conditions of flow-tube pressures and flow velocities, was monitored a t m / e 46. The O,/He and 03/N2 mixtures typically con1.25% ozone, resulting in O3 concentrations tained . ranging from 3 X 1014to 10 X 1014molecule ~ m - ~These O3 concentrations converted atomic bromine into BrO radicals (99% conversion) within 4-13 ms. Nitrogen dioxide was synthesized by mixing NO with a large excess concentration of O2 for sufficient time to ensure >99.99% conversion of NO to NOz. The NO2/ N204/Oz mixture was freezethaw degassed at dry-ice and liquid-nitrogen temperatures to remove all of the OF The purified product was a white solid, indicating the absence of NO. The purified N02/N204sample was placed in an insulated, jacketed Pyrex cylinder of known volume whose temperature was kept constant at 298 K by circulating water from a Haake constant-temperature bath. Oxygen
-
(18) G. W. Ray and R. T. Watson, manuscript in preparation.
The Journal of Physical Chemistry, Vol. 85, No. 2, 7987 201
was added to the NOz/Nz04mixture (10% by pressure) in order to suppress any formation of NO. It was essential to minimize the level of NO impurity because reaction 5 NO BrO NOz + Br (5)
-
+
k5(298 K) = 2.2
X
10-l1 cm3 molecule-’ s-’ (ref 17)
is rapid and could cause kinetic complications by removing BrO radicals (discussed in the Results section). The NOz flow rate was determined by using the method of monitoring the rate of pressure drop from a known volume a t a constant temperature. The NOz flow rate was calculated after allowing for the presence of the added oxygen and for the Nz04. The NOz/N204equilibrium constant reported by Blendlg was used to calculate the exact composition of the mixture. After passing through a needle valve which controlled the NOz/Nz04flow rate, the Nz04 rapidly dissociated into NOz and entered the flow tube through the fixed inlet jets. Flash Photolysis-Ultraviolet Absorption. The flashphotolysis-ultraviolet-absorption apparatus has been described in detail previously.20 The quartz reaction cell was operated in the continuously flowing mode with a flash interval equal to the cell residence time (15-30 s). As in our previous studies of BrO BrO radicals were monitored by their absorption at 339.0 nm, near the band head of the A2n(v’= 7) X211(u”= 0) transition. The absorption band at 339.0 nm was originally assigned22to the 4-0 transition of the A X system; however, the results from a recent study by Barnett et al.23has led to an upward renumbering of the u’levels by three. When a monochromator slit width of 100 pm and an absorption path length of 720 cm were used, the effective BrO cross section was previously determined to be (1.14 f 0.14) X cm2.21However, this parameter is not required for the data analysis because pseudo-first-order kinetics were employed. An analog multichannel analyzer was employed to average from 10 to 50 separate BrO decay profiles. Because of scattered light from the photolysis flash, detection of the BrO absorption signals could not take place until -50 ps after the flash. However, this period was usually short compared to the time scale of the BrO disappearance. The temperature of the reaction cell was maintained at 298 f 1 K by circulating methanol through the outer cell jacket from a constant-temperature circulator. Since pseudo-first-order conditions were employed ([NO210 >> [BrOIo), [BrO] obeyed the rate equation In ([BrOIo/[BrO]) = Izl[N02]t
-
-
Beer’s law was used to relate the absorption signal to [BrO] In (In (Io/I,J - In
(l0/4)) = kl[NOzlt
where It,,, I,, and Io are the transmitted light intensities immediately after the flash, at time t , and after many BrO half-lives, respectively. BrO radicals were produced by the reaction of oxygen atoms with a large excess concentration of molecular bromine (eq 6 and 7). Oxygen atoms were produced from 0 2
+ hv
-
20
(6)
(19) H. Blend, J. Chem. Phys., 53, 4497 (1970). (20) R. T. Watson, S. P. Sander, and Y. L. Yuna, J. Phys. Chem., 83, 2936 (1979). (21) S. P. Sander and R. T. Watson, manuscript in preparation. (22) R. A. Durie and D. A. Ramsay, Can. J. Phys., 36, 35 (1958). (23) M. Barnett, E. A. Cohen, and D. A. Ramsay, submitted for publication in Can. J. Phys.
202
The Journal of Physical Chemistry, Vol. 85, No. 2, 1981
0 + Br,
-
BrO
+ Br
Sander et ai.
(7)
k7(298 K) = 1.4 X lo-" cm3 molecule-' s-l (ref 24)
the photolysis of molecular oxygen in the wavelength region h >180 nm. A flash energy of 500 J dissociated only 0.0015% of the molecular oxygen, necessitating large 0, concentrations, typically around 1 x 1Ol8 molecule to produce BrO concentrations of 1 X 1013-2 X 1013molecule ~ m - At ~ . the lowest total pressures employed in this study (50-100 torr) where such high 0, concentrations would have contributed a nonnegligible third-body effect, O2was not added. Photolysis of the NOz present as a reactant provided a sufficient number of oxygen atoms via reaction 8 to produce BrO in concentrations less than -2 NO,
+ hv
-+
NO
+ O(3P)
(8)
X 1013 molecule ~ m - ~NzO . was used as an additional oxygen atom source in a few runs at 50 and 100 torr (eq 9). However, because NzO photolyzes relatively ineffiNzO + hv Nz + O('D) (9)
-
ciently, it was not a good source of atomic oxygen. The disadvantage of both of these oxygen atom precursors is that NO, which reacts rapidly with BrO via reaction 5, is formed as a byproduct. NO is formed directly from the photolysis of NO,, and in reaction 10a when NzO is added. NO production from
- +
O(lD) + N 2 0
2N0
klOa= 8.6 X lo-'' cm3 molecule-' O(lD) + N 2 0
N,
(loa) s-' (ref 17)
O2
(lob)
klOb = 7.4 X lo-" cm3 molecule-' s-' (ref 17) NO2 photolysis, both by the photolysis flash lamp and the malyzing lamp, was minimized by a series of measures. First, the lowest possible flash-lamp energy was used that was consistent with a given signal-to-noise ratio. This was typically less than 700 J per flash. Second, the flash-lamp filter cell was filled with a C12-Br, mixture, which a t equilibrium contained 60 torr of BrCl and 200 torr of Clz. BrCl has an absorption band centered at 370 nm. The BrCl and Clz absorbed a portion of the light which would otherwise have resulted in NO, photolysis. Photolysis also occurred in the spectroscopic analysis beam. This was suppressed by placing a 10-cm cell containing 200 torr of Br, in the analysis beam to filter out its near-UV spectral component. A manually operated shutter was also placed in the analytical beam path which was opened a few seconds before the photolysis flash and closed immediately thereafter. These measures reduced the NO, photolysis due to the flash lamp to 0.18% per flash a t 389-5 flash energy without reducing the efficiency of 0, photolysis and practically eliminated NO2 photolysis from the analysis lamp. The effect of photolytically produced NO is treated in the Results section. The ranges of reagent concentrations were (in molecules ~ m - as ~ )follows: [Br,] X 1.0-18.9; [O,] X 1.5-3.4; [N,O] X 10-17, 1.0-5.3; [NO,] X 0.34-8.6. [BrO], ranged from 0.5 X 1013to 4 X 1013molecules by changing [O,] or flash energy (300-700 J). This resulted in initial analytical beam absorptions of 4-28% with noise levels around 0.1 %. Molecular bromine concentrations were high enough to ensure that 99% of the atomic oxygen
(24) M. A. A. Clyne, P. B. Monkhouse, and L. W. Townsend, Int.J. Chern. Kinet., 8, 425 (1976).
was converted to BrO radicals within 50 p s after the photolytic flash. Characteristic times for BrO formation were always at least 10 times shorter than those for BrO loss by reaction 1. NOz-02 mixtures were made by reacting small amounts of NO with a large excess of 0, and dowing sufficient time for complete conversion. NO, mixtures were stored and transferred in darkened bulbs and lines. Reagents
The helium used as the carrier gas was Linde Ultrahigh Purity (UHP) grade (99.999%). It was passed through a zeolite trap cooled to 77 K before being admitted to the DF/MS flow tube. The nitrogen and oxygen were also Linde UHP grade. The chlorine (99.96%) and nitric oxide (99.0%) were obtained from Matheson. All of the above gases were used without further purification. Bromine was obtained from J. T. Baker (99.7%) and was purified by extensive freeze-thaw degassing followed by fractional distillation.
Results Discharge-Flow-Mass-Spectrometric Study. Table I summarizes the experimental conditions and results of the 97 experiments performed by using the DF/MS technique. Table I1 summarizes the initial reagent concentrations and results obtained at 4-torr total pressure of He. As mentioned earlier, this study was performed by using pseudo-first-order conditions, [NO,], >> [BrO], ([NO,], = 0.27 X 1015-6.43 X 1015~ m - [BrO], ~; = 0.76 X 10"-5.38 X 10" ~ ; = 0.5 X ~ m - [He] ~ ; = 0.5 X 10'7-2.0 X 10'' ~ m - [N,] 1017-1.3 x ~ m - ~giving ) , a range of initial stoichiometry of 0.11 X lo4-4.37 X lo4, with a typical value of 8.2 X lo3. Consequently, the BrO decay rates were analyzed by using the following equation: In ([BrO],/[BrO],) = k't = k 1 1 [ N 0 ~ l=~ t~ I I I [ N O ~ I ~ [ M I ~ where k ', kII, and kIIIare the pseudo-first-order, secondorder, and third-order rate constants, respectively. Each individual decay plot (In ([BrO],/[BrO],) vs. t) was corrected for the dissociative ionization of the BrNO, product molecule in the ion source: BrN03 + e- BrO+ + NOz 2e-
-
+
which contributes to the BrO+ signal being monitored. The uncorrected decay plots show negative curvature a t long reaction times when a significant fraction of the BrO has been converted into BrN03. The BrO+ signal from BrN03 was typically 5-7 % of that from an equal concentration of BrO. This was determined by measuring the BrO+ signal in the presence and the absence of a large concentration of NOz (sufficient to convert >99% of the BrO into BrN03). Figure 2 shows several typical decay plots for BrO with time (corrected for the dissociative ionization of BrNOJ. The decay of the BrO concentration with time was followed over a factor of 2-100 decrease, and typically greater than a factor of 10. The linearity verifies that the reaction is first order in [BrO]. The values of the first-order rate constants, k', which were obtained from plots similar to those shown in Figure 2, were corrected for axial and radial diffusion. The diffusion coefficient for wBrO at 298 K was calculated to be 0.58 atm cm2 s-' in He, and 0.14 atm cm2 s-l in NB. These values were derived, after allowing for the differences in molecular weights, from the data reported by Marrero and Masonz5for Kr in He and Kr in N,, re(25) T. W. Marrero annd A. Mason, J . Phys. Chern. Ref. Data, 1, 3 (1972).
Pressure Dependence of the BrO
+ NO, Reaction
The Journal of Physical Chemistry, Vol. 85, No. 2, 1981 203
TABLE I : Summary of the Discharge-Flow-Mass-Spectrometric Experimental Conditions and Results l o 14k11(mean), 10'4kII(slope), cm3 molecule-I cm3 molecule-' total press., no. of 10-'5[N0,], range, torr expt molecule cm-) k' range, s-' S-1 6-l 1.5 (He) 2.3 (He) 3.0 (He) 3.7 (He) 4.0 (He) 6.0 (He) 1.5 (N,)a 2.3 ( N , p 3.0 (N,? 4.0 (N,)a
9 9 6 12 13 8 11 10 9 10
0.90-6.43 0.92-2.39 0.40-2.57 0.48-4.18 0.51-3.79 0.43-2.90 0.3 7- 2.86 0.36-1.82 0.26-1.66 0.27-2.3 5
10.0-71.0 10.9-47.6 9.1-53.2 12.1-102 16.7-1 0 5 22.0-130 10.0-68.3 13.3-67.6 11.3-79.0 14.7-1 74
1.14 f 1.86 f 2.28 f 2.60 f 2.80 f 4.60 f 2.51 f 3.69 f 4.55 f 5.45 2
0.05 0.10 0.08 0.21 0.28 0.37 0.21 0.23 0.21 0.21
1.15 f 1.75 t 2.14 f 2.54 f 2.50 f 4.20 f 2.38 f 3.82 f 4.62 f 5.46 f
0.03 0.10 0.06 0.12 0.16 0.15 0.11 0.16 0.19 0.18
intercept, s-' -0.1 f 1.3 f 0.6 f 0.6 f 3.1 f 4.6 t 1.0 f -1.2 f -0.6 f -0.3 t
0.6 2.3 0.4 0.4 2.4 2.5 0.8 1.0 2.0 0.4
a In these experimental runs, < 10%of the total pressure was due to He. The observed k' values were first corrected for the contribution due to the BrO + NO, + M (M = He) reaction. The values of kII (M = N,) were then derived by using the corrected k' values.
TABLE 11: Summary of Experimental Data at 4-torr Total Pressure o f He (DF/MS)
lo-"
10-15 x
[NO,I o ,
molecule
x
[BrOI,, molecule 10- x cm-' [NO,],/[BrO],
0.51 0.69 0.83 0.90 1.10 1.44 1.28 1.83 2.47 2.01 2.24 2.62 3.79
1.45 5.35 1.46 5.35 5.38 5.35 1.46 1.45 5.38 3.33 3.33 1.46 1.45
k I ' ,s - l
16.7 21.0 22.9 24.8 29.5 36.5 37.7 46.0 55.8 57.6 61.9 83.9 105
0.35 0.13 0.57 0.1 7 0.20 0.27 0.88 1.26 0.46 0.60 0.67 1.79 2.61
5'0r 4.0 f
0 0
10
20
30 40 50 TIME lmillisecondri
MI
70
I
80
Figure 2. Pseudo-first-order decay of BrO radicals reacting with NOz at 298 K in the DF/MS system at a totalpessure of 3 torr of helium. [NO 1. (0) 2.72 X (0) 1.50 X 10' , (A) 4.26 X 10'' molecules cm-3. '
spectively. Corrections for axial diffusion were only required for the 1.5 torr He data, and the magnitudes of the corrections were typically less than 3%. Corrections for radial diffusion were only required for the data obtained at 3.5 and 4.0 torr in N2,and the magnitudes of the corrections were always less than 16%.26 It can be seen from (26) R. V. Poirier and R. W. Carr, J. Phys. Chem., 75,1593 (1971).
1 0 L 4 k I I ,cm3 molecule- ' s-'
3.29 3.04 2.75 2.76 2.68 2.53 2.94 2.51 2.26 2.87 2.76 3.20 2.76 2.80 t 0.28 (mean) 2.50 f 0.16 (slope)
1031k1n,cm6
molecule-'^-^ 2.57 2.38 2.13 2.14 2.08 1.96 2.30 1.95 1.75 2.24 2.16 2.50 2.16 2.18 t 0.23 (mean)
Figure 2 that the plots of In ([BrO],,/ [BrO],) vs. time exhibit a small negative intercept on the time axis. This negative intercept can probably be attributed to a mixing time that is several milliseconds. This is somewhat longer than normal ( [BrOl, ([NO210 = 0.34 X 1015-8,63 X 1015~ m - [BrOIo ~; = 0.48 X 1013-6.9X 1013~ m - ~ ; [N2]= 0.16 x 1019-2.1 x 1019~ m - ~giving ) , a range of initial stoichiometry of 34-580 with a typical value of 210. Consequently, the BrO decay rates were analyzed by using the equation In ([BrO],/[BrO],) = k’t = kI~[N02]t BrO decays were typically observed over two to five (27) M. A. A. Clyne and R.T.Watson, J. Chem. Soe., Faraday T r a m . 1, 71, 336 (1975).
The Journal of Physical Chemistry, Vol. 85, No. 2, 198 1 205 10
k
\
\
I
.\
L L J L L A
10
100
200
300
400
5W
600
700
TIME (micrasecondsi
Figure 5. Pseudo-first-order decay of BrO radicals reacting with NO2 at 298 K in the FP/UV system at a total pressure of 400 torr of N2 [NO3]: (A) 5.01 X lo“, (0) 1.50 X loi5,(0) 4.13 X loi5molecules cm- .
half-lives, depending on initial radical concentrations. Some typical BrO decays are shown in Figure 5. Plots of k’ vs. [NO,] a t each pressure were linear with small y intercepts. These plots show no evidence of a BrO removal process other than reaction with NO,. An obvious trend of k l with pressure is observed with the rate constant increasing by a factor of -5 as the pressure is increased from 50 to 700 torr of N,. A graph containing several k’ vs. [NO,] plots at different pressures is shown in Figure 6, illustrating that the reaction is first order in NOz and that the rate constant has a pressure dependence. The variation of k l with pressure is shown in Figure 7. Table IV contains the measured bimolecular rate constants derived both by averaging the values of k’/[NO,J for each run and by computing the weighted least-squares slopes of the k’ vs. [NO,] plots. While differences of up to 17% are observed between the rate constants calculated by the two methods, the average difference over the entire pressure range is only 7%, indicating that the differences are due only to experimental scatter. Indeed the standard deviations of the intercepts shown in Table IV are such that each of the intercepts is statistically meaningless. The preferred rate constants are obtained from the averages of the individual runs rather than the slopes. In the 50and 100-torr experiments, the runs which contain N,O have higher average rate constants than the ones which do not. This effect may be due to a higher NpO third-body efficiency or a small amount of NO formation from reaction loa. For these pressures, the final rate constant has been obtained by averaging the two sets of data. A number of potentially complicating reactions may take place which must be considered. These include the reactions of BrO with BrO and NO. The effect of reaction 12 on the BrO decay rate can, as in the mass-spectrometric study, be estimated by evaluating the ratio klp[BrO]/k1[NO,] for each kinetic run. The average value of this ratio for all runs is 0.015 when using [BrO] = [BrO], and 0.004 when using [BrO] = [Br0],/4, the “average value” of [BrO] for each run. The effect of reaction 12 is therefore negligible. Reaction 5 can occur as a result of NO being formed from three processes: photolysis of NOz (reaction 8), the
-
206
The Journal of Physical Chemistry, Vol. 85, No. 2, 1981
Sander et al.
TABLE IV : Summary of the Flash-Photolysis-Ultraviolet- Absorption Experimental Conditions and Results 10' 'kII(mean), 1012kII(slope), total press., no. of 10-'5[N0,], range, cm3 molecule-' cm3 molecule-' torr expt molecule cm-3 h' range, s - ' S-1 S-1
194 20 - 75 315 188 219 241 - 81 Experiments performed with NO, as the sole source of atomic oxygen resulted in a value of kII(mean) of (0.49t 0.08)X lo-" cm3 molecule-' s - ' , whereas experiments performed in the presence of N,O resulted in a value of kII(mean) of (0.60t 0.08)X l o - ' ' cm3 molecule-' s - ' . Combining both sets of data resulted in the value of (0.55t 0.10) X lo-" cm' 5w 1OO* 200 300 400 500 600 700
1.12-8.63 0.51-5.10 0.52-4.52 0.69-2.55 0.34-6.12 0.54-4.72 0.43-5.47 0.41-5.85
25 41 8 6 14 13 8 14
450-5900 420-5130 634-6680 1260-3990 705-10300 1280-14900 1230-31600 1200-20000
0.55 & 0.92f 1.36 f 1.60t 1.94 f 2.25 f 2.56 t 2.98*
0.10 0.16 0.20 0.15 0.33 0.38 0.24 0.31
0.58t 0.85f 1.37 f 1.37t 1.70 f 2.02 t 2.36t 2.98 f
0.03 0.05 0.16 0.11 0.14 0.29 0.13 0.08
intercept -
molecule-' s" for hII(mean)shown in the table. Experiments performed in the presence of N,O resulted in a value for kII(mean) of (0.98t 0.11) X c m 3 molecule-'s-', whereas experiments performed in the absence of N,O resulted in a value for hII(mean) of (0.86f 0.18)X 10'" cm3 molecule-' s-'. Combining both sets of data resulted in the value of (0.92t 0.16)X lo-'' cm3 molecule-'s-' for hII(mean) shown in the table. TABLE V: Summary of Experimental Data at 100-torr Total Pressure of N, (FP/UV)
x [Ozl,"
flash energy, J
molecule cmW3
10-15 x [Brzl, molecule
389 1.73* 9.94 389 1.82* 10.0 389 1.82* 10.0 389 0.37 10.9 529 0.0 18.9 389 0.37 10.9 389 0.53 10.0 389 1.73* 9.94 389 1.82* 10.0 389 0.17 11.4 389 0.37 10.9 691 0.37 10.9 941 0.37 10.9 200 0.48 10.9 527 0.0 18.9 389 1.82* 10.0 389 0.37 10.9 389 1.73* 9.94 529 0.0 18.9 691 0.37 10.9 389 0.37 10.9 389 0.37 10.9 941 0.37 10.9 200 0.37 10.9 389 1.82* 10.0 529 0.0 18.9 389 0.37 10.9 389 1.73* 9.94 389 0.53 10.0 691 0.37 10.9 389 0.37 10.9 389 1.82* 10.0 941 0.37 10.9 529 0.0 18.9 389 0.17 11.4 389 0.37 10.9 389 1.82* 10.0 389 1.73* 9.94 529 0.0 18.9 389 0.37 10.9 k(mean) = 0.92f 0.16x cm3 molecule-'s-' k(s1ope) = 0.85t 0.05x lo-', cm3 molecule-' s-'
10-13 x
10-15x
[BrOI,, molecule cm-3
PJ021, molecule cm-3
k'msd, S-'
0.80 0.79 0.91 1.3 0.67 0.90 1.0 1.0 1.0 1.0 1.3 3.0 5.3 0.48 0.95 1.1 1.4 1.1 0.97 4.4 1.5 1.6 6.9 0.51 1.3 1.2 1.5 1.2 1.3 1.4 3.5 1.5 6.7 1.0 1.1 2.4 2.1 2.0 0.97 2.5
0.511 0.590 0.950 1.13 1.30 1.31 1.40 1.51 1.58 1.65 1.70 1.70 1.75 1.77 1.93 2.23 2.27 2.54 2.55 2.84 2.84 2.89 2.95 2.98 3.05 3.33 3.40 3.56 3.85 3.97 3.97 3.98 4.11 4.32 4.34 4.54 4.58 4.60 4.89 5.10
667 447 771 1140 1140 1370 1760 1169 1330 1820 1610 1930 2170 2030 2110 1732 2170 1719 2750 3280 3000 3280 3780 3050 2197 3920 3070 2248 3720 3880 3640 3731 5420 4350 3480 5360 4888 4941 3950 4520
a N,O is used instead of 0, unless indicated by an (*). the final results.
k14 = 9.2
X
cm3 molecule-'^-^ (ref 17)
s-'
1O"k, cm3 molecule-' s - '
645 420 729 1090 1060 1310 1700 1100 1260 1750 1530 1680
1.26 0.712 0.767 0.965 0.815 1.00 1.21 0.728 0.797 1.06 0.900 0.988
b
1990 1990 1630 2060 1600 2590 3040 2860 3140
1.12 1.03 0.731 0.907 0.630 1.02 1.07 1.01 1.09
b
2980 2050 3710 2900 2070 3530 3530 3440 3530
1.00 0.672 1.11 0.853 0.581 0.917 0.889 0.866 0.887
b
4070 3260 5130 4510 4710 3630 4260
0.942 0.751 1.13 0.985 1.02 0.742 0.835
Experiments at high flash energy (941 J) were not averaged into
reaction of 0 with NO2 (reaction 14), and reaction (loa) 0 + NO2 NO + 0 2 (14) +
h',,,,
O(lD)
+ N20
-
2N0
(104
when N 2 0 is present. Of these reactions, NOz photolysis is most significant, followed in importance by reaction 14. Because of the low N 2 0 photodissociation rate in our
Pressure Dependence of the BrO
+ NO, Reaction
The Journal of Physical Chemistry, Vol. 85, No. 2, 198 1
207
A
16
I
2
1
4 IN021 i1015 molecule ~ r n - ~ )
5
3
6
0. 5
I
1.0 1.5 IMI 11019 malecule
-
d l
+
+
I
2.0
+
Figure 0. Pseudo-first-order rate constant, k', vs. [NO,] for the reaction BrO NO, N2 in the FP/UV system at 298 K. Total pressure: (0) 50, (0) 200, (A) 700 torr.
Figure 7. The reaction XO NO, M XNO, M. Effective bimolecular rate constant, kll,vs. concentration of N, for X = Br (this work (0)) and X = CI (data of Cox and Lewis" (A)). Solid lines are curve fits to the data using eq 1.
system, reaction 10a can be ignored. The value of k'obtained in each kinetic run was corrected for the amount of BrO which reacted with the NO formed in reactions 8 and 14. The NO formed by photodissociation was calculated by two methods. In the first, NO2-O2 mixtures were repeatedly photolyzed and the NO2 disappearance was measured. The percentage NO2 dissociation per flash was calculated by assuming a quantum yield of 2.0. In the second method, the NO, dissociation was calculated from measured values of [BrO], in runs whre NOz photolysis was the only source of oxygen atoms. Agreement between the two methods was good; however, the first method was assumed to be more accurate and was therefore used to compute the amount of NO formed. The value of 0.18% NO2 photolyzed per flash at 389 J was linearly scaled with flash energy. The average correction decreased monotonically from 8.8% at 50 torr to 1.6% at 700 torr, reflecting the increase in the ratio of k l / k 5 with pressure. At a flash energy of 700 J, -5% of the molecular bromine is dissociated, leading to initial atomic bromine concentrations of 0.5 X 1015-2 X 1015molecule ~ m - ~In. the presence of NOz, Br undergoes rapid catalytic recombination by the sequence of reactions Br + NOz + M BrNOz + M (15)
effect of changing the total pressure over an order of magnitude from 76 to 760 torr of N2, assuming that reaction 15 is third order up to 1 atm. The results indicate that at low pressure the maximum depletion of NO2 expected under worst-case conditions (low [NO,],, k16 = 2 X lo-'' cm3 molecule-' s-') is 5.5%. This increases to 34% at high pressure. Under more realistic conditions (k16 = 8 x lo-" cm3 molecule-'s-l) the NO, depletion drops to 1.5 and lo%, respectively. These results indicate that the formation of BrNO, should have had no discernible effect except at pressures exceeding 300-400 torr where the effect would be dependent on [NO2],. Because of this [NO,] dependence, the formation of BrN02, if it had been significant, should have manifested itself as curvature and a negative y intercept in the plots of k'vs. [NO,]. Since neither feature was observed, and the required rate constants are not known with sufficient accuracy to justify making such a small correction, the effect of BrN02 was ignored. Since reactions 15 and 16 occur on a rapid time scale (less than 100 p s to reach 95% of equilibrium conditions), bromine atoms are removed much more rapidly than if bimolecular recombination (eq 17) were the only Br + Br + M Br2 M (17)
+
+
Br
+ BrN02
-
Br2
+ NOz
(16)
-
k17 = 9.1 X
+
cm6 molecule-2 s-' (ref 30)
Reactions 15 and 16 pose a complication if a significant fraction of the NO2 becomes tied up in the form of BrN02. Rate constants for these reactions have not been measured; however, there are rate data for the X + NO2 M reaction (X = F, C1, I, and M = He, N2) and for the X + XN02 reaction (X = F, I).27928On this basis, we estimate kI5 = 2.2 X (M = He) and 4.4 X loe3' cm6molecule-2 s-l (M = N,) and k16 = 8.0 X lo-" cm3 molecule-' s-l with uncertainty factors of 2 for k15 and 4 for k16. Numerical simulations of reactions 15 and 16 were performed by using a range of conditions representative of the actual experiments and rate-constant uncertainties: [NO2], = 1 X iO'5-i0 x iO15 molecule ~ m -k15 ~ ,= 1.2 x 10-12-12 x cm3 molecule-' s-', and k16 = 2 X 10-11-8 X lo-" em3 molecule-'s-'. The variation in k15 also incorporates the
reaction. The rapid approach to equilibrium also ensures that any changes in the base-line analytical light absorption due to Br, and BrN0, will not interfere with the detection of BrO. Thermal decomposition of BrN03 (eq -1) can also affect BrN03 + M BrO + NO, + M (-1) the determination of k l if it occurs on the same time scale as the formation step. Although direct measurements of the gas-phase unimolecular decomposition rate have not been conducted for BrN03, data do exist for C10N02.8~9 If one assumes the same decomposition rate constants for ClONO, and BrN03, an upper limit for k-, at 700 torr is obtained by assuming that the decomposition is still in the second-order region a t this pressure. At 298 K and 700 torr the equivalent first-order rate constant is -0.0013 s-'. This compares with a value of 1200 s-l, the smallest
(28) NASA Panel for Data Evaluation, Pasadena, CA, 1979, Jet Propulsion Laboratory Report No. 79-27. (29) H. van den Bergh and J. Troe, J . Phys. Chem., 64, 736 (1976).
(30)J. A. Blake, R. J. Browne, and G. Burns, J . Chern. Phys., 53,3320 (1970).
+
-
208
The Journal of Physical Chemistry, Vol. 85, No. 2, 1981
first-order rate constant observed a t 700 torr for the association reaction. The effects of the thermal decomposition of BrN03 can therefore be ignored. Absorption of the analytical light by the product BrN03 does not pose a problem in the determination of kl. BrN03 absorbs very weakly at 339.0 nm; the ratio of the absorption cross sections for BrOZ1and BrN0331is 126. Even if the BrNO, absorption were appreciable, it can be that an absorber (BrN03) formed as a primary product of the reaction of the species being monitored (BrO) does not interfere with the measurement of k' as long as Io is evaluated at a time long compared with the BrO reaction time but short compared with the cell residence time. As indicated above, the system was operated in the continuously flowing mode with the cell contents being entirely replaced between flashes. Problems due to depletion of NOz and buildup of reaction products such as BrN03 and BrN02 were therefore not encountered.
Discussion Although there are no published measurements of kl, there are a number of studies of the analogous C10 + NOz reaction (reaction 11)both in the third-order region (1-7 C10 + NOz + M ClN03 M (11)
-
+
and the falloff region between second- and thirdorder kinetics."" Knauth and c o - w o r k e r ~found ~ ~ ~ little falloff (