J . Phys. Chem. 1987, 91, 4603-4606
4603
Kinetics of the Reactions of SH Radicals with NOp and O2 R. A. Stachnik and M. J. Molina* Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California 91 109 (Received: March 5, 1987)
+
The S H NO2 reaction rate constant was measured by monitoring SH concentrations with long-path optical absorption at 324 nm in a laser photolysis system. At 298 K the rate constant was found to be (4.8 h 1.0) X lo-'' cm3 molecule-l s-l. An upper limit of 4 X cm3molecule-' s-' was inferred for the SH + Ozreaction, indicating that the atmospheric oxidation of the SH radical may occur via reaction with species such as NOz and 03.
Introduction Substantial effort has been directed in recent years toward understanding the chemistry of atmospheric sulfur compounds. Hydrogen sulfide, H2S, a significant form of biogenic sulfur emission,' is removed from the troposphere primarily by reaction with hydroxyl r a d i c a P H2S + OH SH + H2O
-
The atmospheric chemistry of the SH radical has been the subject of several recent kinetic and modeling s t u d i e ~ ; ~however, -'~ the details of the subsequent oxidation of this radical remain to be elucidated. Recent kinetic studies of the reaction SH + NO2 H S O NO (1)
-
+
agree that this is a rapid two-body process;8-12however, reported values for the bimolecular rate constant range from 2.4 X lo-" cm3 molecule-' s-'.I2 The cm3 molecule-l s-','I to 1.2 X react ions
SH
+ Oz
--
+ SO + SOz
OH
H
AH = -102 kJ mol-' AH = -223 kJ mol-'
(2a) (2b)
have not been observed. Evaluation of k2 is the key to determining whether oxidation of SH in the atmosphere is direct or is controlled by the subsequent chemistry of the H S O radical formed in reactions of SH with NO, and 0,. Upper limits for the rate constant cm3 molecule-' SI), imposed by Tiee et al.7 ( k , C 3.2 X cm3 molecule-' s-l), Fried1 et aL8 ( k , C Black9 ( k z C 4 X 1X cm3 molecule-' s-'), and Wang et a1.I0 ( k 2 < 1.5 X 1O-I' cm3 molecule-I s-l) do not eliminate reaction 2 as a significant removal process for tropospheric S H , as assumed in previous This paper reports the results of a flash photolysis/UV absorption study of the reactions of SH radicals with NO, and 0,.
Experimental Section Figure 1 is a schematic diagram of the apparatus. SH radicals were generated in this study by photodissociation of H2S with pulsed A r F excimer laser emission at 193 nm. Initial SH con~ , the H2S centrations were typically 5 X lo', molecule ~ m - and concentration was in the (1 5 2 ) X 1015molecule cm-3 range. The subsequent temporal decay of the SH concentration was measured by monitoring the transient absorption due to the (0,O) band of the SH(AZ X2) system near 324 nmm. The spectroscopic light source was a 75-W xenon high-pressure arc lamp. This emission was directed through a White-type multipass cell (6.4-m effective path length) and onto the entrance slits of a 0.45-m Czemy-Turner monochromator (McKee-Pedersen Model MP-1018B) equipped with a photomultiplier tube (EM1 type 9789 QA). Following amplification, the signal was recorded on a signal averager (Nicolet Model 1174) coupled to a computer (Data General Nova or IBM
-
*Mailing address: M/S 183-301, Jet Propulsion Laboratory, 4800 Oak Grove Drive, Pasadena, CA 91109.
0022-3654/87/2091-4603$01.50/0
PC-AT). Additional features of this apparatus have been described in a previous paper.I3 Kinetic measurements were made by summing the transient absorption signals due to SH, subsequent to typically 1024 photolyzing laser pulses, and fitting the result to an exponential decay function using linear and nonlinear least-squares algorithms. Initial absorptions were typically 5% and were followed for at least three l / e lifetimes. Gas mixtures of H,S, 02,N,, and NzO or CO, prepared by using calibrated mass-flow meters (Matheson Type 8141 and 8148) and capacitance manometers (MKS Baratron Type 220 and 170), were flowed a t rates sufficient to replace the reaction cell contents at approximately the laser pulse repetition rate of -1 Hz. Hydrogen sulfide (Matheson C.P. grade 99.55%), oxygen (Matheson U.H.P. grade 99.9%), and nitrogen (Matheson U.H.P. grade 99.9%) were used without further purification. A 10% mixture of NO, in 0, was prepared by mixing N O (Matheson C.P. 99.0%), purified by passage through an Ascarite column, with excess 0, in a 5-L bulb. The resulting NO, concentration was spectrophotometrically verified. Carbon monoxide (Matheson 99.99%) was purified by passage through a column of activated charcoal and iodine.I4
Results and Discussion Reaction of S H with NO2. Hydrogen sulfide was chosen as the SH photolytic precursor for its large absorption at 193 nm (u 7 X cm2) and high quantum yield of SH(V=O).'~ However, H2S participates in reactions which provide secondary sources or regeneration cycles for the SH radical. In the reaction
-
(1) Moller, D. Atmos. Enuiron. 1984, 18, 29. (2) Leu, M. T.; Smith, R. H. J . Phys. Chem. 1982, 86, 73. (3) Sze, N. D.; KO, M. K. W. Atmos. Enuiron. 1980, 4 , 1223. (4) Cox, R. A.; Sandalls, F. J. Atmos. Enuiron. 1974, 8, 1269. (5) Thiemens, M. W.; Schwartz, S. E., XI11 Informal Conference on Photochemistry, Clearwater, FL, 1978. (6) McElroy, M. B.; Wofsy, S . C.; Sze, N. D. Atmos. Enuiron. 1980.14, 159. (7) Tiee, J. J.; Wampler, F. B.; Oldenborg, R. C.; Rice, W. W. Chem. Phys. Lett. 1981, 82, 80. (8) Friel, R. R.; Brune, W. H.; Anderson, J. G. J . Phys. Chem. 1985, 89, 5505. (9) Black, G. J . Chem. Phys. 1984, 80, 1103. (10) Wang, N . S.; Lovejoy, E. R.; Howard, C. J., submitted to J. Phys. Chem. (1 1) Bulatov, V. P.;Kozliner, M. Z.; Sarkisov, 0. M. Khim. Fiz. 1984, 3, 1300.
(12) Schoenle, G.; Rahman, M. M.; Schindler, R. N. Ber. Bunsen-Ges. Phys. Chem. 1987, 91, 66. (13) Stachnik, R. A.; Molina, L. T.; Molina, M. J. J . Phys. Chem. 1986, 90, 2777. (14) Stedman, D. H.; Tammaro, D. A.; Branch, D. K.; Pearson, R. Anal. Chem. 1979, 51, 2340. (1 5) Okabe, H. Photochemistry of Small Molecules; Wiley-Interscience: New York. 1978.
0 1987 American Chemical Society
4604
The Journal of Physical Chemistry, Vol. 91, No. 17, 1987
Stachnik and Molina 1.51
I
I
I
I
,
ARflp ~
MONOCHROMATOR
~
FLOW OUT - - - - - - - - --
HERMOSTATED HOUSING
[NO21 (IOi4 molecules cmm3)
Figure 3. Plot and least-squares fit of pseudo-first-orderdecay rate, k', vs. [NO,] at 298 K in 30 Torr of 02,measured (open circles; leastsquares fit, solid line) and calculated (dotted line). The dashed line is a least-squares fit to four calculated points (closed circles). TABLE I: Numerical Simulation Reactions and Rate Constants
(3)
Figure 1. Schematic diagram of the flash photolysis/UV long-path ab-
sorption apparatus.
(6) SH
+
NO2
7
(14)
I
-
(5)
(8) (10) (11) (12) (13)
1.5 -
"0
(4)
(15)
1.0 -
(16)
T-
'3r
(17)
+
reaction H + HIS H2 + SH H + N 0 2 j O H + NO OH + H2S SH + H 2 0 H + O2 + M H02+ M OH + CO C02 + H SH + H 0 2 HSO + OH SH + SH S + HIS
--+ - + + + - + + + + - + + + -+
+
s+o2-+so+0
SO 0 2 SO2 0 0 O2 M 0, M
SH 03 HSO 0 2 HSO 0, SH O2 -OH SO 0 2 H02 + HO2 --+ H202 + 0 -+
2
k , cm3 moleculed s-I 7.2 x 10-13 1.1 x 10-10 4.7 X
1.o
2.0
3.0
k , , lo-" cm3S-I
4.0
Figure 2. Plot of pseudo-first-orderdecay rate, k', vs. [NO,] at 298 K in 100 Torr of 0, (open circles) and in 100 Torr of 02/630Torr of N2 (closed circles).
of S H with NO,, the H atom photofragment can react directly with H2S: H HZS H2 S H (3) or indirectly: H NO, OH N O (4)
+
+ OH + HZS
+
+
-
-+
+ S H + H20
(5)
The role of these processes was minimized by introducing molecular oxygen to sequester the H atom: H 0 2+M HO2 + M (6) The requirement that k 6 [ 0 2 ][MI >> k4[N02] over the range of NO, concentration employed ([NO,],,, = 4.0 X lOI4 molecule ~ m - necessitated ~) [O,] > 3 X lo1*molecule cm-3 and thus sets the minimum total pressure at approximately 100 Torr. SH secondary generation is also reduced by OH NO, + M HONO, M (7)
+
+
+
+
+
However, the effect is minor under the experimental conditions employed here. The bimolecular rate constant values, k l , for the reaction of S H radicals with NO, were determined by measuring a set of pseudc-first-order rate constants, k', corresponding to [NO,] from 0.1 X I O i 4 to 4.0 X lOI4 molecule ~ m - ~Under . the present experimental conditions, the equations In ([SH],/[SH]) = k't = (kI1[NO,] kd)t
+
3.0 f 0.8 3.5 f 0.4
2.4 f 0.2 6.7 f 1.0 12.0 f 2.0 4.8 f 1.0
technique" DF/LIF FP~LIF
FP/ICLA DF/LMR DF/MS FP/AB
8
2.7 X 10-I2
TABLE 11: Summary of Rate Constant Data for SH NO at 298 K
[NO2] (IOl4 molecules c m - 5
19 17 17 17 this work this work 17 17 17 8
5.5 X 10-32[M] 2.0 x 1043 1 X lo-" 4 x 10-1' 2.3 X 8.4 x 10-17 6.0 X 10-34[M] 3.2 X 10-l2 1 x 10-13
+
ti
ref 18
17
+ NO2
-+
HSO
press., Torr 2-8 He 29-300 He 100 Ar 1.0-1.1 He 2-5 He
9
30-730 0 2 / N 2
this work
ref 8 11 10 12
"DF = discharge flow; LIF = laser-induced fluorescence; ICLA = intracavity laser absorption (HSO followed); LMR = laser magnetic resonance; AB = long-path absorption; MS = mass spectrometry. apply and k" values were determined by linear least-squares analysis. Figure 2 is a plot of k' values vs. [NO,] with 100 Torr of O2(open circles) and with 100 Torr of O2and 630 Torr of N, present (closed circles). The close adherance to linearity and the agreement between the 100 and 730-Torr total pressure data sets indicate that sufficient 0, was present to avoid secondary generation of S H radicals. The slope of this plot yields k l = (4.8 f 1.0) X lo-" cm3 molecule-' s-I. The stated uncertainty includes both estimated systematic and random errors. In contrast, a k' vs. [NO,] plot of measurements with 30 Torr of O2present, Figure 3, plainly exhibits negative curvature consistent with SH secondary generation increasing with [NO,]. A linear fit to this curve yields the value 2.4 X lo-" cm3 molecule-' s-' for k l . However, the initial slope (4.7 X lo-" cm3 molecule-' s-I), obtained by considering the first four points, agrees with the measurements at higher 0, pressure. The results of a numerical model of the kinetics of this system are also shown in Figure 3. The model included reactions in Table I and used k l = 5.0 X lo-" cm3 molecule-' s-I. Although high concentrations of O2were employed throughout these measurements, secondary chemistry due to photolytically generated O3 is anticipated to be negligible. Absorption lines of the 0, Schumann-Runge system are sparse and weak in the vicinity of 193 nmI6 and the 0, present in the -1-m air path
The Journal of Physical Chemistry, Vol. 91, No. 17, 1987
Reactions of SH Radical with NO2 and O2
0
0.5 1.0 1.5 2.0 2.5 INITIAL SH ABSORPTION ('10)
01 0
3.0
Figure 4. Plot of [SH] decay rate vs. initial SH absorption in 730 Torr of O2with 10 Torr of CO (open circles) and without CO (closed circles).
between the laser and the reaction cell forms a gaseous filter which reduces the intensity of components of the ArF shape coincident with discrete Schumann-Runge features. A direct measurement of 0, yield was made by irradiating 730 Torr of O2 in a static cell. Approximately 9 mTorr of O3accumulated after 100 laser pulses with 180 mJ pulse-'. This corresponds to [O,] C 3 X 10l2 molecule for a single laser pulse. Table I1 summarizes the results of previous measurements of k l . The large discrepancies among these values are, evidently, not connected to experimental approach or total pressure range employed. The low-pressure discharge flow result of Fried1 et a1.8 is in reasonable agreement with the flash photolysis measurements by Black9 and with the indirect method of Bulatov et al.," but it is in sharp contrast to the discharge flow/LMR result of Wang et a1.I0 and the discharge flow/mass spectrometry value reported by Schoenle et a1.I2 Since hydrogen sulfide was the SH precursor in the former two studies, secondary generation processes may, in part, explain the current lack of consensus. Reaction ofSH with 4. As previously indicated, OH and SO radicals are energetically permissible products from a reaction of SH with O2 (reaction 2a). In the presence of H2S, SH can be. rapidly regenerated via reaction 5. Evidence for a slow reaction between SH and O2 to produce O H was sought by introducing carbon monoxide which removes OH via
OH
+ CO
-
COZ
+H
(8)
1
I
I
I
1
2.5 INITIAL SH ABSORPTION ('10)
0.5
1.0
1.5
4605
2.0
3.0
Figure 5. Plot of the difference between [SH] decay rates vs. initial S H absorption with 10 Torr of CO added and without CO in 730 Torr of 02. The intercept is 7 8 .
+
-
the rate of reaction 2b, SH O2 H + SO2: a conservative cm3 molecule-' s-l is obtained by upper bound for k2bof attributing the entire intercept to this reaction. The intercept is, however, most likely due to reaction with impurities; reaction 2b is surely much slower than reaction 2a in spite of the larger exothermicity, since it requires simultaneous breaking of the S-H and 0-0bonds with formation of two S-0 bonds. Similarly, this small intercept value indicates that the association reaction
-
SH
+0 2+M
--c
HSO2
+M
(9)
is also unimportant ( k 9 C cm6 molecule-2 s-l, assuming negligible back-reaction). A numerical model, simulating the reactions listed in Table I, verified the applicability of the linear extrapolation analysis to this system of primarily second-order reactions. The model also provided some additional kinetic information: a value of -4 X lo-'' cm3 molecule-' s-I for the SH self-reaction rate constant was found to produce good agreement with the observed decay rates and is consistent with the literature value.20 The slope of the experimental data shown in Figure 4 indicates a reaction between SH and another minor species which produces OH directly or by subsequent reaction of the products. Although S H reacts with 0, giving HSO, and HSO may react with O3to produce OH, these reaction ratess are insufficient to account for the observed SH regeneration. Another possible source of OH radicals is
Approximately 10 Torr of C O was required to ensure that ks[CO] >> k5[H2S]. The H atom product associates quickly with 02,and the reaction between SH and CO, analogous to reaction 8, is endothermic. Figure 4 displays the observed exponential decay rate of SH in the presence of 730 Torr of O2 alone and with 10 Torr of C O added vs. initial SH concentration from approximately 0.2 X lo1, to 3.0 X lo', molecule cm-,, obtained by varying the photolyzing laser fluence. Figure 5 presents the same data shown in Figure 4 plotted, however, as the difference between the two data sets. The dependence of the SH decay rate on [SH],,,t,al manifest in Figure 4 indicates clearly that SH self-reaction and reaction with other radicals or photolytically generated species are the dominant loss processes in this SH concentration regime. However, in the limit of [SH],,,t,al 0, second-order processes become negligible while first-order and pseudo-first-order contributions remain constant. A linear least-squares extrapolation yielded intercepts of 28 and 21 s-' for the measurements with and without 10 Torr of CO present, respectively. The difference between these intercept values, 7 s-I, represents an upper bound to k2a[02]and yields kza C 4 X cm3 molecule-' s-I. The limiting decay rate of 21 s-I is the sum of first-order and pseudo-first-order losses for SH including reactions with O2which do not yield OH radicals, reaction with impurities, and wall loss (which is negligible at 730 Torr total pressure). This intercept value provides independent, but limited, information Concerning
(17) DeMore, W. B.; Margitan, J. J.; Molina, M. J.; Watson, R. T.; Hampson, R. F.; Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R., "Chemical Kinetics and Photochemical Data for Use in Stratospheric Modeling", Evaluation No. 7; Jet Propulsion Laboratory: Pasadena, CA, 1985; JPL Publication.No. 85-37. (18) Kurylo, M. J.; Peterson, N. C.; Braun, W. J . Chem. Phys. 1971, 54,
(16) Ymhmo, K.; Freeman, D. E.; Esmond, J. R.; Parkinson, W. H. Planet. Space Sei. 1983, 31, 339.
74, 1280.
-
SH
+ H02
+
HSO
+ OH
AH = -105 kJ mol-I
-
(10)
The result of the numerical simulation using kIo lo-'' molecule-I cm3 s-l agreed well with the experimental data; however, this value is speculative and additional investigation of this interesting reaction is warranted. The upper bound for k2adetermined in this work is consistent with previous reports and reduces, by approximately a factor of 25, the measured upper limit for this rate constant. The significance for tropospheric sulfur chemistry is that, based on this result, direct irreversible oxidation of SH by O2 proceeds at a rate, at most, comparable to reaction with 03.
Conclusions Reactions of the SH radical with NO2 and with O2 were investigated by a flash photolysis/UV absorption technique. The rate constant for the reaction with NO2 ( k , = (4.8 f 1.0) X lo-''
943. (19) Bemand, P. P., Clyne, M. A. A. J . Chem. Soc., Faraday Trans. 2 1977, 73, 394, (20) Mihelic, D.; Schindler, R. N. Ber. Bunsen-Ges. Phys. Chem. 1970,
J . Phys. Chem. 1987, 91, 4606-4613
4606
cm3 molecule-’ s-I) is significantly larger than all but two of the previously reported values and is substantially below these two results which are 40% and 150% greater. N o evidence of reaction between SH and O2 was found. An upper bound for the rate constant was inferred which indicates that reaction with O2in the atmosphere is very slow. However, a smaller upper bound would be required to fully preclude a role for this atmospheric reaction. This upper limit determination reflects mainly the detection sensitivity for SH radicals of the
present technique and further investigation of this reaction is needed. Acknowledgment. We thank C. J. Howard for providing us with his results prior to publication. The research described was performed at the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. Registry No. SH, 13940-21-1; NO2, 10102-44-0; 02,7782-44-7.
luring Bifurcation and Stationary Patterns in the Ferroin-Catalyzed Belousov-Z habotinsky Reaction A. B. Rovinsky Institute of Biological Physics of the Academy of Sciences, Puschino, Moscow Region, 142292, USSR (Received: August 18, 1986; In Final Form: February 12, 1987)
A model of the bromate-ferroin-bromomalonicacid chemical system is studied. The model suggests that while the reaction running in a small well-mixed volume has a single, stable stationary state the same reagent spread in a thin layer may spontaneously form stationary periodic spatial patterns of either small or large amplitude, depending on conditions. The homogeneous state corresponding to the stable stationary state of an individual local element of the system may lose its stability because of diffusive coupling between spatially separated points. However, a situation is possible such that both the homogeneous and inhomogeneous(large amplitude) states are stable (bistability, despite the fact that the local system has the only stationary state). In this case, the periodic structure, being in contact with the homogeneous part of the reaction mixture, builds itself up to fill the whole vessel.
Many reaction-diffusion systems exhibit a rich variety of dynamic phenomena. The processes which are observed there can be divided into the two classes: traveling waves and stationary (Standing waves have also been f o ~ n d . ~ . ~ ) The traveling waves are often relaxational. That means that they are of large amplitude and characterized by several different time scales. Propagation of waves of this sort has the very remarkable feature of being in many respects almost independent of the actual mechanism of the underlying chemical or chemicallike process, Le., of the exact form of the kinetic nonlinearities and of the boundary conditions. Such relaxational systems can be described in terms of a rather simple model first suggested by Wiener and Ro~enblueth.~The principal feature of that model medium is that each of its local elements can be excited and immediately after the excitation it cannot be excited again during some refractory time, during which it relaxes to the original state. A most interesting class of standing patterns has been described by Turing.* H e discovered that a homogeneous medium of identical local elements, each being in the same stable steady state, can lose its stability and form a stationary spatialy periodic (1) Nicolis, G.; Prigogine, I. Self Organization in Non-Equilibrium Systems; Wiley-Interscience: New York, 1977. (2) Haken, H. Advanced Synergetics, Springer Series in Synergetics; Springer-Verlag: West Berlin, 1983. (3) Self-Organization. Auto-waues and Structures Far from Equilibrium, Krinsky, V. I., Ed.; Springer Series in Synerdetics; Springer-Verlag: West Berlin, 1984. (4) Non-equilibrium Dynamics in Chemical Systems, Pacault, A., Vidal, C., Eds.; Springer Series in Synergetics; Springer-Verlag: West Berlin, 1984. (5) Boiteux, A,; Ha,B. Ber. Bunsen-Ges.Phys. Chem. 1980,84,392-398. (6) Holmuhamedov, E. L. Eur. J . Biochem. 1986, 158, 543-546. (7) Wiener, N.; Rosenblueth, A. Arch. Inst. Cardiol. Mex. 1946, 16, 205-265. ( 8 ) Turing, A. Phil. Trans. R . SOC.London, Ser. B 1952, 2378, 31-12.
0022-3654/87/2091-4606$01.50/0
structure. The most striking point of the phenomenon is that it occurs despite the fact that the internal kinetic mechanism of the system opposes any deviation from the steady state, and diffusion opposes any deviation from homogeneity. The formation of Turing patterns is strongly dependent on the actual kinetic mechanism of the system and on the boundary conditions. Therefore studies of both waves and stationary patterns are equally important as they can provide a complementary information about the system. The best known example of an excitable reaction-diffusion system is the Belousov-Zhabotinsky reaction catalyzed by the ferroin/ferriin couple. A variety of wave patterns has been found and extensively investigated theree9 However, investigation of stationary structures is far less comprehensive and It is not yet clear whether hydrodynamic and surface effects always are involved in the formation of the patterns, although there is experimental evidence that in some cases they are (Showalter,” Agladze’*). Therefore, theoretical consideration of the phenomenon may provide a deeper insight into its nature and promote further experiments. The first step in this direction was made recently by Becker and Field.I3 They developed a “wind-diagram” method for a qualitative analysis of whether large-amplitude stationary patterns are possible. With this method they found sufficient conditions under which stationary patterns can be expected to develop in the Oregonator model and then demonstrated their existence computationally. Unfortunately, they found the stationary patterns ~
~~
~~~~
~
(9) Oscillations and Traveling Waves in Chemical Systems, Field, R. J., Burger, M., a s . ; Wiley-Interscience: New York, 1985. (10) Zhabotinsky, A. M.; Zaikin, A. N. J . Theor. Biol. 1973,40, 45-60. (1 1) Showalter, K. J. Chem. Phys. 1980, 73. 3735. (12) Agladze, K. I. An Investigation of Rotating Spiral Waues in a Chemical Actiue Medium; Akad. Sci. USSR: Puschino, 1983. (13) Becker, P. K.; Field, R. J. J . Phys. Chem. 1985, 89, 118-128
0 1987 American Chemical Societv
