J. Chem. SOC., Faraday Trans. I , 78, 3679-3689 CH(A 2A-X "n) and OH(A T+-X ") Chemiluminescent Radiation from q3P) + C6H6 Discharge-Fast-flow Mixtures BY FRANCISCO TABARES, VINCENTE SAEZ RABANOS AND ANGEL GONZ~LEZ UREGA* Departamento de Quimica Fisica, Facultad de Ciencias Quimicas, Universidad Complutense de Madrid, Ciudad Universitaria, Madrid-3, Spain Received 14th April, 1982 Using a discharge-fast-flow apparatus a study of the CH* and OH* chemiluminescent radiation produced in O(3P) + C,H, reactive mixtures has been carried out. CH* and OH* emission intensities centred at 43 15 and 3090 8, have been assigned to the CH(A2A-X2H) and OH(A2C+-X2H) electronic transitions, respectively, and monitored as a function of reactant concentrations for several temperatures. The following empirical relations for their emission intensities have been obtained : Z(CH*) = a,[B] exp (- b,[B]) and I(OH*) = a2[BI2 exp (- b,[B]) where [B] is the concentration of benzene and Q, and b, (i = 1,2) are empirical factors independent of [B].An absolute kinetic determination of the CH* chemiluminescent reaction, using the NO+O glow as a standard, has been determined, leading to an estimate of the quantum yield of d(CH*) = 0.06. 1. INTRODUCTION The major contribution of fast-flow discharge systems has been the measurement to high precision of the rate constants of elementary gas-phase reactions where atoms and radicals produced by means of radiofrequency or microwave discharge at low pressure are involved. Excellent reviews of fast-flow chemical reactions have been reported1* covering most of the literature.On the other hand modulation techniques3 constitute one of the simplest and most elegant methods of studying fast reactions. Its main use has been in the investigation of photoinduced additive polymerization (the rotating-sector method), although recently several a u t h o r ~ ~ - ~ have extended its use to atom-polyatom gas-phase reactions with organic compounds. Thus Hunziker4 has made a number of studies of the absorption spectra of triplet species formed by energy transfer from Hg(3P1) or Hg(3P0) by using rapidly modulated mercury resonance radiation. A similar modulation system has been used by Atkinson and Cvetanovic5 to study reactions of ground-state oxygen atoms produced by the 253.7 nm mercury- sensitized N20 decomposition in a flow system.In this study we describe a modulated-discharge-fast-flow reactor suitable for kinetic data determination of elementary reactions involved in more complicated processes, such as combustion, atmospheric and photochemical reactions, and we report the CH(A 2A-X "n) and OH(A 2C+-X "n) emissions observed in O(3P) + benzene mixtures. The kinetics and dynamics of reactions between oxygen atoms and hydrocarbons have been studied widely, and excellent reviews have been published.6 In particular, benzene as the hydrocarbon is of great interest in combustion processes and atmospheric chemistry, and this system has been studied by a variety of techniques. Thus, relative reaction rates for the O(3P) + benzene reaction have been determined using static 3679 119-23680 q 3 P ) + C6H6 DISCHARGE-FLOW MIXTURES photolysis techniques. 7 9 Also, absolute rate-constant measurements have been reported using pulse radioly~is,~ discharge-flow,1°-12 modulation phase-shifP? l3 and flash-photolysis-N02-chemiluminescence14 techniques.Recently the molecular-beam methodl5* l6 (and particularly the crossed molecular- beams investigation of Lee and coworkers16) has identified the first two major reaction pathways as (1) (2) the first channel being hydrogen elimination and the second oxygen addition leading to the formation of highly excited phenol. Except for the above-mentioned molecular-beam study, the identification of products in the O(3P)+C6H, system is complicated by the presence of viscous, non-volatile reaction products (probably polymeric in character).At the present, phenol, CO, C5H6 and C,H,O (but not CH) have been reported1'* 1 5 9 l6 among others as O+C,H, reaction products. CH emission from the A2A-X211 system has been observed mainly in oxyacetylene flames. On the other hand ultraviolet OH emission corresponding to the electronic A 2X+-X 211 transition has been observed in a variety of systems, such as hydrogen,17 hydrocarbon flames18 (particularly oxya~etylene,'~ shock tubes,20 in the reaction between ground-state 0 and H atoms at room temperature21 and also as a component in the spectra of discharge systems containing water vapour22 even in trace amounts. In the present paper we focus attention on trying to obtain maximum information on these chemiluminiscent processes in terms of band identification, dependence of the emission upon the reactant concentrations and temperature, and the absolute kinetic determination for these chemiluminiscent reactions using the NO + 0 glow as a These results are presented and discussed, with our main goal being to report such findings rather than to provide a detailed mechanism which would perhaps require more work, particularly directed toward identifying intermediates or precursors, because of the complexity of reactions occurring in the system.q 3 P j ) + C6H6 + C6H50 + H AHg8 = - 15.9 kcal m01-l q 3 P j ) + C,H, + C6H50H AHg8 = - 102.4 kcal mol-l 2. EXPERIMENTAL DISC H AR GE-G AS-FL 0 W RE ACTOR The apparatus2* is shown schematically in fig. 1. A Pyrex reactor (2 cm diameter, 50 cm length) is pumped by means of a strong mechanical pump producing a fast-flow velocity of several thousand cm3 s-l.The flow system is a conventional one with gas flows controlled by needle valves and using a calibrated 'floating ball' flow meter (Goring S.A.) to measure the organic compound. A typical benzene flow range for the present experiment is from 2.2 to 1 1.2 pmol s-l. The main gas flow (0.7-1.3 mmol s-l of Ar and 70-140 pmol s-l of 0,) passes through a modulated discharge cavity powered by a microwave unit (Electromedical Supplies, Microtom 200 MK2) operated at 100 W, in which the O(3P) atoms are produced before they enter the flow tube. The Pyrex reactor wall is coated with boric acid to minimize oxygen-atom recombination. Since the oxygen atoms in the present experiment are produced by modulated discharge any emission of products would appear modulated at the same frequency as the discharge.As the reactants are introduced into the reactor inlet the reaction between oxygen and the organic compound takes place, and any chemiluminescent product formed is detected by measuring its emission intensity. As shown in fig. 1 we monitored the signal via a monochromator (Jobin Ybon H20) coupled to a photomultiplier (RCA C3 1034) connected to a preamplifier/lock-in amplifier system (Keithley models 427 and 840, respectively) and then to an oscilloscope (Tetronix 5403) or strip-chart recorder (Leeds & Northrup, Spedomax W). After each emission band was identified we recorded its intensity as a function of reactant concentration and temperature, the latter being varied by a furnace allocated to the reactorF.TABARES, v. SAEZ RABANOS AND A. GONZALEZ URERA 368 1 MONOCROMATOR PHOTOMULTIPLIER BENZENE FLOWMETER - FIG. 1 .-Schematic diagram of the discharge-fast-flow reactor and monitored by several thermocouples placed inside the flow tube. Therefore, despite the fast-flow velocity (see below) of our gaseous mixture care was taken to insure both temperature homogeneity and that the observed temperature was that of the actual reactive mixture. FLO W-VE LOC I TY ME A S UREME N T S In the present experiment a time-of-flight determination of the flow velocity was adopted. In this method we produced the O(3P) atoms by a short (ca. 1 ms) discharge pulse and measured the appearance time of the NO,* fluorescence intensity produced by the reaction O+NO+M+NO,*+M when NO was introduced in a particular reactor inlet (dl).The same procedure was repeated further downstream (d,). Thus under our experimental conditions the difference in the appearance time A? can only be attributed to the gas velocity, 2, = Ad/At. Fig. 2 shows typical time-of-flight data giving a value tl = 4000 & 100 cm s-l. Apart from the experiments described in section 3.1, carried out 1.0 ms into the reaction, all the present experiments were carried out at 7.0 ms. t/ms FIG. 2.-Typical time-of-flight data for flow-velocity determination. MATERIALS Benzene, toluene, cyclohexane, cyclohexene and n-hexane (Merck, purity 99.5 %) were purified by low-temperature distillation and subsequent degassing at 77 K.All the gases were from S.E.O. and were of the following purity: Ar > 99.9%; 0, > 99.98%; NO > 99.9% and N,O > 99.9%. They were used directly.3682 o(3~) + C,H, D IS c HA R G E--F LO w M I x T u R ES 3. RESULTS AND DISCUSSION 3.1. OBSERVED EMISSIONS During the present experiments reactive mixtures of O(3P) +benzene were prepared and several emission bands, illustrated in table 1, were identified and recorded. Typical experimental conditions are also reported. For the benzene reaction a low-resolution spectrum of the most important bands is shown in fig. 3. The relative1 intense band CH(A 2A-X211) electronic transition, as reported in ref. (19c) and (19d). Although from 4240 to 4380 A, centred at 4315 A, was identified as the 4315 K band for the TABLE 1 .-EXPERIMENTAL CONDITIONS AND RELATIVE BAND INTENSITIES IN THE q 3 P ) + C,H, REACTION experimental conditions gas flow/pmol s-l total pressure/Torr temperature / K discharge frequency/Hz discharge power/W reaction time/ms photomultiplier voltage/V typical signal/pVa typical noise/pVa monocromator resolution/nm Ar 1300 0 2 140 'GH6 11.2 2.0 42.5 7.0 333 100 1800 ( 2 x 10-3)- 10-2 10-4 2 band intensities CH A2A-X211 4315 1 .ooo CH B2A-X211 3895 0.062 OH A2E-X211 3090 0.765 a The lower and higher values are shown, depending on the benzene concentration range. our resolution was not adequate to reveal the rotational distribution of the electronically excited CH(A 2A) species, the low shoulder observed around 4335 A, in agreement with more precise observations from the chemiluminescence of CH in the 0 + C2H, flame,lSd can be attributed to a P-branch with a rotational distribution characterized by Got = lOOO+ 100 K (estimated from a wavelength shift of AII x 20 A with respect to the main peak, i.e.the Q-branch). The 3060-3130 A band centred at 3090 A was ascribed to the OH(A 2Z+-X 211) transition as reported by Krishramachi and Broida in their study on the emission spectra of atomic-oxygen-acetylene flames.lgC A very low intense band around 3890 A observed in the 0(3P)+benzene reaction was attributed to the CH (B2A - X 2 n ) transition, as described e1~ewhere.l~~ The observed band intensities, corrected for the instrument's response, were normalized, with the intensity of the 43 15 A band of CH(A 2A-X ") taken as unity.These emissions (but with changes in intensity) were observed in the following cases: (a) when the 0 atoms were produced either by Ar/N,O discharge or by the N+NO + N 2 + 0 reaction, indicating that for the observed emissions no 0, is necessary; (b) when toluene was used instead of benzene, under the same experimental conditions.F. TABARES, V. SAEZ RABANOS AND A. G O N Z ~ L E Z U R E ~ A 3683 OH(A ’ 1 - X *n) 2 L I I 1 I I I I 1 300 305 310 315 ” 420 425 430 435 440 h/nm FIG. 3.-Low-resolution emission spectrum obtained in the O(3P) +C,H, reactive mixture under the experimental conditions of table 1 . No emission bands at 4315 and 3895 A, i.e. the A and B electronic states of CH, were observed when benzene was replaced by n-hexane, cyclohexane or cyclohexene.This could indicate that triply bonded hydrocarbons, formed only in the benzene reaction (see section 3.4), are needed as CH precursors in the complex reaction. 3.2. OH* AND CH* EMISSION INTENSITIES AS A FUNCTION OF BENZENE CONCENTRATION AT A FIXED INITIAL CONCENTRATION OF OXYGEN ATOMS FOR SEVERAL TEMPERATURES Fig. 4 and 5 show the benzene concentration dependence of both CH* and OH* emission intensities at a fixed initial concentration of oxygen atoms for several temperatures. These data are well represented by the following empirical relations : I(CH*) = a,[B] exp (- b,[B]) I(OH*) = a,[BI2 exp (- b2[B]) (2) where [B] is the benzene concentration and ai and bi ( i = 1,2) are parameters dependent on temperature but not benzene concentration. In fact this dependence, at least for the b parameters, shows Arrhenius-like behaviour (see fig.6), i.e. bi x Ai exp (- Ci/RT). From the slopes in fig. 6 the following Ci values were obtained: C, (CH* emission) = 4.4f0.3 kcal mol-l and Cz (OH* emission) = 2.1 kO.2 kcal mol-l. The overall temperature range was 328-474 K. Note that the present temperature depen- dence of the CH* emission is similar to the activation energy1’ for the corresponding initiation reaction of 0 + C,H, given by Ea = 4.4 f 0.5 kcal mol-l; this indicates that even though complex secondary reactions of 0 plus intermediates are occurring at least the CH* chemiluminescent radiation is limited by the primary reaction. on the CN(B2X - X 2 n ) emission intensity in mixtures of active Previous3684 q3P) + C6H6 D ISCH ARGE-FLO W MIXTURES I I 1 5 10 15 [ B ] /lo-'* rnol cm-3 FIG.4.-Logarithmic I/[B], i.e. CH* emission intensity divided by benzene concentration, [B], as a function of [B] for several temperatures: 0, T = 445 K; 0, T = 350 K; 0, T = 417 K; A, T = 383 K ; solid lines are smooth lines through the data. nitrogen and carbon tetrachloride have shown that empirical relationships similar to that of CH emission, i.e eqn (l), can reasonably be described by a complex mechanism where the initiation reaction is the rate-limiting step. All the present experiments were carried out at an initial concentration of oxygen atoms of [O], < 2 x mol ~ m - ~ . By adding a constant flow of NO to the 0 +benzene reactive mixture through the reactor inlet located in front of the monocromator/photomultiplier zone we measured the NO,* emission intensity due to the residual concentration of 0 atoms.Fig. 6 shows the CH* to NO,* emission intensity ratio as a function of benzene concentration. These data are quite well represented by a straight line with zero intercept, from which (under the present conditions) one may conclude that the CH* emission intensity is directly proportional to the product of benzene and oxygen concentrations, i.e. Z(CH*) oc [B] [O]. The same data representation (not shown in the figure) gives a more complicated dependence for the OH* intensity. This is perhaps an indication of a different reaction mechanism, as one might expect from the different value observed for its temperature dependence.3.3. KINETIC DATA DETERMINATION OF THE CH* A N D OH* CHEMILUMINESCENT REACTIONS. ESTIMATION OF QUANTUM YIELDS The kinetic rate constants of the chemiluminescent processes 0 + C6H6 + OH* + CH* +products + ow + CH + hv +productsF. TABARES, v. SAEZ RABANOS AND A. GONZALEZ UREGA 3685 0 [Bl/lO-'O mol ~ r n - ~ FIG. 5.-As fig. 4 but for OH* emission in the benzene reaction; the ordinate scale is now log I/[BI2: 0, T = 333 K; 0, T = 378 K; 0, T = 408 K; A, T = 438 K; ., T = 463 K. emission. Solid3686 q 3 P ) + C6H6 D I S C H A R G E-F L 0 W M I X T U R E S were estimated by using the NO+O glow as a standard reaction.23 First, and as suggested by Fontijn et aZ.,23 the unknown CH* glow and the NO+O glow were viewed under the same conditions (i.e.using the same apparatus, physical location, detector, etc.) and their respective intensities I NO,*) and I(CH*) were recorded over displayed in fig. 3. Next the ‘specific’ intensity I ( X ) was defined as the same wavelength range from 4240 to 4380 8, , corresponding to the CH* spectrum where [XI is the concentration of NO or C6H6 and [o] is the oxygen-atom concentration. To insure the vality of the present calculation care was taken to correct for small changes in the oxygen-atom concentration in both measurements. This was accom- plished by measuring the NO,* fluorescence intensity at 3, = 5500 A with and without C,H, and including this correction factor in our calculation. Then the rate constant of the two light-emitting reactions over the wavelength region A1-A2 (as indicated by Fontijn et al.) was calculated via the equation The value of k(NO,*) used here was the reported value of 6.4 x lo-’’ cm3 molecule-’ s-l, corrected by considerin the fractional area of the wavelength interval of interest (i.e.from 4240 to 4380 R ) over the whole NO,* glow spectrum extended from 3875 to ca. 14000 A. By using this procedure a k(CH*) value of 7.45 x lo9 cm3 mol-l s-l was obtained. Since the OH* glow lies outside the NO,* glow wavelength range we have normalized our OH* spectra to those of CH* under the same experimental conditions, their relative emission intensities being those shown in fig. 3. This simple scaling procedure shows that typically I(CH) > I(OH*), and could indicate (see fig. 3) a smaller quantum yield, i.e.&OH*) < q5(CH*), but care should be taken with this comparison since the OH* emission cannot be reduced (see below) to the ‘bimolecular’ scheme, i.e. I(OH*) z [B] [O], as the CH* emission does, and therefore it cannot be compared with the standard NO+O glow. In other words the respective quantum yield for the OH* emission is a complex function [see eqn (2)] of the concentrations of both benzene and atomic oxygen. Thus under our experimental conditions only q$(CH*) can be properly defined as #(CH*) = k(CH*)/k, where k is the rate constant for the primary reaction, i.e. O+C,H, --+ products. Now if one uses for k the mass-spectrometric value of Bonano et al.,ll obtained by monitoring the benzene concentration and given by 1.22 x loll cm3 mol-l s-l at T = 383 K, one obtains a fluorescence quantum yield of b(CH*) = 0.06, i.e.for every 100 benzene molecules 6 give fluorescence emission via the CH(A 2A-X 211) electronic transition. 3.4. POSSIBLE SOURCES FOR THE FORMATION OF CH* AND OH* In spite of the fact that our main goal, as mentioned before, is not to give a detailed mechanism for either OH* or CH* formation, the present experimental results together with previous studies2,? 27 provoke some comments on the most likely pathways for the formation of these radicals. These are as follows. CH* FORMATION A molecular-beam studyl59 l6 of the same reaction (0 + C6H6) has identified both CO and C5H6 (cyclopentadiene or 3-pent-1-yne) as products of the excited decom- position of phenol described in eqn (1). Indeed, the cyclopentadiene + oxygen-atom reaction has been shown2, to originate C,H, isomers, one of the major products beingF.TABARES, v. SAEZ RABANOS AND A. GONZALEZ UREAA 0.6 n E v) *-' .3 0.4 .f' 0" 5 v h 4 u 4' 0.2- 3687 - - I 1 [Bl/lO-'O mol cm-3 FIG. 7.-Z(CH*)/Z(NO,*) plotted against [B]. Both intensities are monitored simultaneously at 7 ms reaction time and at 340 K. Note the zero (within experimental error) intercept. but- 1 -yne, which upon reaction with oxygen atoms27 produces CH-C-CH, and CHECH. On the other hand if one assumes 3-pent-l-yne, rather than cyclopentadiene, as the C,H6 product of the C6H6 + 0 reaction, it also seems likely that acetylene is formed.27 Finally, evidence for the formation of CH* in O+C,H, mixtures has been given by Broida and Since all these intermediate reactions have rate constants faster than that of the primary reaction (0 + C6H6) we believe that under the present experimental conditions CH* could be formed via the sequence unsaturated hydro- carbons -+ acetylene -+ CH*; however, we reiterate that this can only be considered as a reasonable suggestion rather than a complete mechanism.OH* FORMATION OH* emission has also been observed in the O+C,H, reaction,lg and path CH*+O, -+ CO+OH* has been proposed as a major source of OH*. In the present study the OH* mechanism seems to differ from that of CH*, as shown by the following facts: (i) OH* emission is also observed when 0, is excluded and 0 produced from N,O; (ii) the empirical law found for its intensity evolution shows a different dependence on benzene concentration than in the case of CH*.4. CONCLUDING REMARKS The present study has focussed attention on the CH(A 2A-X ") and OH(A 2Z-X ") emissions observed in the O(3P) + benzene reactive mixtures produced in our modulated microwave-discharge-fast-flow reactor. We have drawn attention to the identification of bands, the dependence of the emission upon temperature and reactant concentrations and to absolute kinetic determinations for these chemiluminescent reactions by using the NO+O glow as standard. The main results of the present study are as follows. (a) In the discharge-flow q 3 P ) + C6H6 reaction, and under the present conditions, chemiluminescent radiation centred at 43 15 and 3090 A has been observed associated with the CH(A ,A-X ") and OH(A 2C+-X "n) electronic transitions.A very low intense band around 3890A observed in the benzene reaction is attributed to the CH(B2A-X211) transition. Both CH and OH emissions are present when the O(3P)3688 q3P) + C6H6 DISC H A RG E-F LO W MIXTURES atoms are produced either by an Ar/N,O discharge or by the N + NO -+ N, + O(3P) reaction, indicating that 0, for the observed emission is not essential. No CH* emission bands were observed when benzene was replaced by n-hexane, cyclohexane or cyclohexene. This is thought to indicate that unsaturated triple-bonded hydrocarbons are required as CH(A ,A) precursors. Also low-resolution data analysis has shown that the CH(A) state is formed with a rotational temperature of Tot = 1000 100 K. (b) Among the main findings are the empirical laws [eqn (1) and (2)] for the dependence of the CH* and OH* emission intensities upon temperature and reactant concentration.It seems that two different mechanisms are responsible for each chemiluminescent process. Whereas CH* formation appears to be limited by the initiation reaction (despite the presence of complex secondary and tertiary reactions) and shows the same temperature dependence as does the primary reaction, OH* formation seems to be a more complicated process in which subsequent steps with benzene participation could be equally important and be required to account for the observed emission behaviour. (c) A comparison of the chemiluminescent radiation with the standard NO + 0 glow discharge gave an absolute rate constant value of k(CH*) = 2.45 x lo9 cm3 mol-1 s-l, from which a quantum yield of @ = 0.06 was estimated, i.e. 6% of the benzene molecules give, upon reaction with atomic oxygen, fluorescence emission via the CH(A ,A-X ,lT) electronic transition; we consider this to be quite considerable since there are so many reactions occurring in our system. Although the OH* emission could not be reduced and compared with the NO+O glow discharge some indication was reported that, at least, its quantum yield (a complex function of both benzene and oxygen-atom concentrations) was not higher than that of CH*.Finally, we realize the complexity of the reaction system (under gas-phase conditions) being studied, and we reiterate that our main purpose has been to report several items of information concerning these chemiluminescence processes.More experimental and theoretical work is needed, in particular some direct identification of the assumed OH* and CH* precursors via (perhaps) laser-induced fluorescence. This type of study (to include modelling) is in progress in our laboratory. We thank the staff of the Universidad Complutense mechanics, glass and electronics workshop for their help. F.T. and V.S. acknowledge F.P.I. and I.N.A.P.E. fellowships. We also acknowledge the constructive criticism and valuable suggestions of the referees. J. Wolfrum, Atom Reactions, in Physical Chemistry: an Advanced Treatise, ed. H. Eyring (Academic Press, New York, 1975), vol. VI B. (a) A. A. Westemberg, Annu. Rev. Phys. Chem., 1973,24,77 (b) J. T. Herron and R.E. Huie, J. Phys. Chem. Ref. Data, 1973, 2, 467. L. F. Phillips, Prog. React. Kinet., 1975, 7, 83. H. E. Humzinker, IBM J. Res. Dev., 1971, 15, 10. (a) R. Atkinson and R. J. Cvetanovic, J. Chem. Phys., 1971, 55, 659; 1972, 56, 432; see also (6) R. Atkinson and J. N. Pitts, J . Phys. Chem., 1974, 78, 1780; 1975, 79, 295. (a) R. E. Huie and J. T. Herron, Prog. React. Kinet., 1975,8, 1 ; (b) M. C. Lin, Dynamics of Oxygen Atoms Reactions in Potential Energy Surfaces, ed. K. P. Lawley (Wiley, New York, 1980); (c) I. W. M. Smith, The Excited State in Chemical Physics (Wiley, New York, 1975). E. Grovenstein Jr and A. J. Mosher, J . Am. Chem. SOC., 1970, 92, 3810. I. Mani and M. C, Sauer Jr, Adv. Chem. Ser., 1968, 82, 142. 1, 28. 7 G. Bocock and R. J. Cvetanovic, Can.J. Chem., 1961, 39, 2436. lo L. I. Avramenko, R. V. Kolesnikova and G. I. Savinova, Isv. Akad. Nauk SSSR, Ser. Khim., 1965, l1 R. A. Bonamo, P. Kin, J. H. Lee and R. B. Timmons, J. Chem. Phys., 1972, 57, 1377.F. TABARES, v. SAEZ RABANOS AND A. GONZALEZ U R E ~ ~ A 3689 l2 D. Furuyama and N. Ebara, Znt. J. Chem. Kinet., 1975, 7, 289. l3 A. J. Colussi, D. L. Singleton, R. S. Irwin and R. J. Cvetanovic, J. Phys. Chem., 1975, 79, 1900. l4 R. Atkinson and J. N. Pitts Jr, Chem. Phys. Lett., 1979, 63, 485. l5 T. M. Sloane, J. Chem. Phys., 1977, 67, 2267. l6 S. J. Sibener, R. J. Buss, P. Casavecchia, T. Hirooka and Y. T. Lee, J. Chem. Phys., 1980, 72,4341. l 7 (a) M. Charton and A. G. Gaydon, Proc. R. SOC. London, Ser. A , 1958, 245, 84; ( b ) W. E. Kaskan, J . Chem. Phys., 1959, 31, 944; (c) H. P. Broida and K. E. Shuler, J. Chem. Phys., 1952, 20, 168. T. Carrington, J. Chem. Phys., 1957, 30, 887. l9 (a) A. G. Gaydon and H. G. Wolfhard Proc. R. SOC. London, Ser. A , 1951,208,63; ( b ) T. Carrington, J . Chem. Phys., 1959,30, 1087; (c) S. I. N. G. Krishnamachari and H. P. Broida, J. Chem. Phys., 1961, 34, 1709; ( d ) K. A. Quickert, J. Phys. Chem., 1972, 76, 825. 2o F. E. Belles and M. R. Lauver, J. Chem. Phys., 1964, 40, 415. 21 S. Ticktin, G. Spindler and H. 1. Schiff, Discuss. Faraday SOC., 1967, 44, 218. 22 A. Michel, 2. Naturforsch., Teil A , 1957, 12, 887. 23 A. Fontijn, C. B. Meyer and H. I. Schiff, J. Chem. Phys., 1964, 40, 64. 24 A preliminary description of the apparatus can be found in A. Gonzalez Ureiia, F. Tabares and 25 V. Saez Rabanos; F. Tabares and A. Gonzalez Ureiia, J. Photochem., 1982, 78, 301. 26 Kazumoto Nakamura and Sehchiro Koda, Int. J. Chem. Kinet., 1979, 9 67. 27 P. Herbrechtsmeier und H. Gg. Wagner, Ber. Bunsenges. Phys. Chem., 1975, 79, 461. V. Saez Rabanos, Physico-Chemical Hydrodynamics, Europ. Conf. Abs., 1980, 3F, 114. (PAPER 2/624)