Dynamics and Mode Specificity in OClO Photodissociation - American

Chemical Sciences Division, Lawrence Berkeley Loboratory and Department of Chemistry, University of .... The OClO was generated in situ by flowing a 5...
0 downloads 0 Views 503KB Size
5681

J. Phys. Chem. 1992,96, 5681-5684

Dynamics and Mode Specificity in OClO Photodissociation H. Floyd Davis and Yuan T. Lee* Chemical Sciences Division, Lawrence Berkeley Loboratory and Department of Chemistry, University of California, Berkeley, California 94720 (Received: March 25, 1992; In Final Form: May 18, 1992)

The photodissociation of OClO in a molecular beam was studied using photofragment translational energy spectroscopy at wavelengths between 350 and 475 nm. Although the dominant products are C10(X211)+ O(3P),we observe formation of C1(2P)+ 02(3Zg)and find strong evidence for a substantial yield of C1(2P) + 02('$).The total yield reaches a maximum of 3.9 f 0.8% near 404 nm and decreases to 99.8% of the photodissociation products are O(3P) C10(X211). We suggest that their observed C1+ signal7does not result from the C1+ O2channel but from photodissociation of the CIO(X211) product to C1+ 0 by the focused 235-237-nmprobe laser, followed by 2 1 REMPI detection of the C1 atom. This is not unlikely since C10 has a large absorption cross section ( u -2 X lo-'* cm2) at 235 nm.20 Photodissociation of CIO was discounted by Bishenden et al.7a since it should also result in C1 signal upon excitation of OClO near 366 nm, but this was not seen in the original spectrum. However, a more recent rein~estigation'~ shows that such a signal does exist. A second argument7, against photodissociation of C10(X211)to C1+ 0 was that some spin-orbit excited C1(2PI/2)was observed, whereas C10(A211) correlates to C1(2P3/2)+ O(ID) in the absence of perturbations. However, it is known that C10(A211) is predissociated,2'J2even at energies far below the threshold22for C1(2Py2) + O(ID). These predissociation products must be CI(2P) + O(3P). The early theoretical arguments2" predicting exclusive production of C1(2P3/2)only apply if the oxygen atom is O('D), and the dissociation is unperturbed. Since it is likely that neither requirement is satisfied, formation of spin-orbit excited C1(2P,12) from photodissociation of C10 cannot be ruled out. Previous discussion regarding OClO C1 O2 in the gas phase has centered on a photoisomerization mechanism involving an intermediate ClOO However, a mechanism formally involving isomerization of OClO to ClOO followed by simple bond rupture does not appear to be consistent with our observations. The large translational energy release that we observe in the recoiling C1+ O2is characteristic of a concerted unimolecular decomposition from a highly constrained transition state, followed by strong repulsion between the products. This transition state is probably best considered to be an OClO molecule with a strongly compressed bond angle and only some distortion from CZusymmetry. The OC10(2A2dBI)transition involves promotion of an electron into the 2bl orbital which is C1-0 antibonding and 0-0 bonding. This leads to a decrease in OClO bond angle from 1 17

+

+

0.30h

; I

8020r L

I

O

'

.

I

0

2

UI

,

I

I

1

3

2

1

0

i

1

CIO(V)

0

t

/

f

1

i

0.00

I

I

4

.

.

I

6

8

I

.

10

.

12

.

'

14

1 16

TRANS. ENERGY, E (kcallmol)

Figure 6. Translational energy distributions for CI(*P)+ O2 and OcP)

+ C10 channelsfrom the (5,1,0) level. The calculated maximum relative translational energies for production of internally excited diatomics are indicated.

confirmed by observation of the respective momentum matched 0 and O2fragments (Figure 5). Based on line width measurem e n t ~it, ~is known that the OClO dissociation lifetime is many piooseconds at these wavelengths. Thus, unless the laser was tuned very close to the origin' of a vibrational band, we found that rotating the laser polarization using a double fresnel rhomb had no effect on product intensities, since the parent molecule rotated at least several times before dissociating. All data reported here were obtained with the dye laser tuned -5 cm-I to the red of the bandhead positions reported in ref 1. Although our laser does not provide rotational resolution (AE 0.2cm-I), we see only minor effects on the shapes of the TOF or C1 + O2yields upon tuning across the rotational envelopes of given vibrational levels of OCIO. Since the shape of the TOF spectra and the C1+ O2 yields remained constant over a range of laser pulse energies (0.5-10 mJ/pulse), our signal results from photodissociation of OClO by absorption of one photon. The translational energy distributions for both channels from excitation of the (5,1,0)band are shown in Figure 6. They are based on data shown in Figures 4 and 5, as well as other data (not shown) obtained at other angles between the molecular beam and d e t e ~ t 0 r . lThe ~ fastest component of the C1+ O2channel, with E,,,,, up to -68 kcal/mol, must correspond to formation of ground-state 02(3ZJ. Based on the integrated area of the P(E) having E,,,,, > 43 kcal/mol, at least 20% of the O2is 32;. The dominant feature peaking at -33 kcal/mol strongly suggests a large yield of excited 02('Ag). As shown in Figure 4,the shapes of the CI + 02 times-of-flight (and 0 2 internal state distributions) are not strongly dependent on the initial OClO vibrational level. The C10 internal state distribution, on the other hand, is very sensitive to the initially prepared OClO state.13 Structure due to vibrational excitation of the C10(211) product is quite well resolved, particularly when its 0 atom recoil partner was monitored (Figure 5). Analysis is currently under way to extract approximate vibrational energy distributions as a function of OClO vibrational state.13 The branching ratios for formation of C1+ O2relative to C10 + 0 were calculated from the fits to the experimental data at m / e = 35 ( a + ) , accounting for the different Jacobian factors for the LAB C M transformation for each channel. One additional factor that is required in the analysis is the relative detection sensitivity for C10 and C1 a t m / e = 35 ( a + ) . This quantity was determined experimentally in a separate experiment. We monitored the equal yields of C1 and C10 from ClOCl photodissociation at 423 nm.13J4The fast peak in the m / e = 35 TQF resulted from C1, and the slower peak was from fragmentation of C10-their

-

-

+

+

+

+

-

+

5684 The Journal of Physical Chemistry, Vol. 96, No. 14, 199‘2 to -107’ and increase in C1-0 bond length from 1.47 to 1.62 Because of this large increase in C10 bond length, very large amplitude bending is required for concerted formation of C1 + 02.23 Based on ab initio ~ a l c u l a t i o n sthe ~ ~OC10(2B2)state has a strongly compressed bond angle of -90’. Thus, as first postulated by G ~ l e ?the ~ C1+ O2channel is probably facilitated by vibronic coupling with the OClO 2B2state. Vibronic coupling between the optically prepared 2A2state24and the strongly bent 2B2state can lead to close approach of the oxygen atoms in OClO and concerted unimolecular decomposition to C1 02.On the basis of orbital symmetry considerations, Gole has pointed out that the 2B2state of OClO correlates with C1(2P) + 02(1A).23 Indeed, our translational energy distributions strongly suggest that a large fraction of the O2products are in the IA state. Our observation that symmetric stretching + bending excitation promotes the C1 O2 channel strongly supports a gas-phase mechanism involving concerted decomposition from a transition-state geometry close to C, rather than from simple unimolecular dissociation of an isomerized C1-00 molecule. The region of the C1+ O2potential energy surface corresponding to the C1-00 isomer represents a very shallow minimum with Do(C1-00) = 4.76 f 0.49 kcal/m01~~ and Oclm l 10°.18923In the absence of a large torque on the O=O molecule, the newly born C1 atom will be near the C, axis of OC10, perpendicular to the O==O bond axis. Since the C1-02 interaction at this angle is repulsive,23the C1-02 “isomer” will not exist for much more than one vibrational period. Instead, strong product repulsion leads to a large C1 + O2recoil energy of up to -68 kcal/mol.

+

+

-

Conclusions We have clarified a number of questions regarding the formation of C1+ O2in the collision-freephotodissociation of OC10. Our results indicate that the yield is only appreciable (>1%) above 380 nm. For a given energy, the yield of C1+ O2is always slightly greater under those lines corresponding to excitation of the symmetric bending motion but is much smaller when exciting an asymmetric mode. The decrease in yield at shorter wavelengths or with asymmetric vibrational excitation likely results from faster predissociation to C10 + 0. Suggested gas-phase isomerization of OClO to C100, based on the earlier matrix isolation results of Arkell et a1.,18 appears to be inappropriate. Although some O2is formed in the ground electronic state, there is strong evidence for a substantial yield of 02(*A),the total reaching a maximum of 3.9 f 0.8% near 404 nm. Further details regarding the mode specificity in the dynamics, complete yields of C1 + 0 2 as a function of wavelength, and approximate C10 vibrational energy distributions will be presented in a forthcoming arti~1e.l~ Acknowledgment. We thank Prof. James Gole for very valuable discussions regarding OClO photodissociation. H.F.D. thanks Jim Myers for an improved version of C M L A B ~and acknowledges Dr.

Letters Albert Stolow for discussions regarding the photodissociation dynamics of triatomic molecules. H.F.D. also thanks NSERC (Canada) for a 1967 Science and Engineering Fellowship. This work was supported by the Director, Office of Energy Research, Office of Basic Energy Sciences, Chemical Sciences Division of the US. Department of Energy, under Contract No. DEAC03-76F00098. Some of the equipment used in this experiment was provided by the Office of Naval Research.

References and Notes (1) Richard, E. C.; Vaida, V. J . Chem. Phys. 1991, 94 (l), 153 and references therein. (2) Richard, E. C.; Wickham-Jones, C. T.; Vaida, V. J . Phys. Chem. 1989, 93,6346. (3) Richard, E. C.; Vaida, V. J . Chem. Phys. 1991, 94 (l), 163. (4) Ruhl, E.; Jefferson, A.; Vaida, V. J . Phys. Chem. 1990, 94, 2990. ( 5 ) (a) Vaida, V.; Solomon,S.; Richard, E. C.; Ruhl, E.; Jefferson, A. Nature 1989,342,405. (b) Vaida, V.; Richard, E. C.; Jefferson, A.; Cooper, L.A.; Flesch, R.; Ruhl, E. Ber. Bunsen-Ges. Phys. Chem. 1992, 96 (3), 391. (6) Lawrence, W. G.; Clemitshaw, K. C.; Apkarian, V. A. J. Geophys. Res. 1990, 95 (D11) 18, 591. (7) (a) Bishenden, E.; Haddock, J.; Donaldson, D. J. J . Phys. Chem. 1991, 95,2113. (b) Bishenden, E.; Haddock, J.; Donaldson, D. J. J . Phys. Chem., in press. (8) Glownia, J. H.; Misewich, J.; Sorokin, P. P. In Supercontinuum h e r s ; Alfano, R. R., Ed.; Springer: Berlin, 1990. (9) Hamill, P.; Toon,0. B. Phys. Today 1991,44 (12), 34 and references therein. (10) JANAF Thermochemical Tables, 3rd ed.; J . Phys. Chem. Ref.Data 1985, 14 (Suppl. 1). (11) (a) Wodtke, A. M.; Lee, Y. T. J. Phys. Chem. 1985,89,4744. (b) Minton, T. K.; Nathanson, G. M.; Lee, Y. T. J . Chem. Phys. 1987,86 (4), 1991. (12) Wahner, A.; Tyndall, G. S.; Ravishankara, A. R. J . Phys. Chem. 1987, 91, 2734. (13) Davis, H. F.; Lee,Y. T. To be published. (14) Renard, J. J.; Bolker, H. I. Chem. Rev. 1976, 76 (4), 487. (15) In the OClO experiments, we found that the C10 fragment TOF spectra measured at both C10’ and C1* had identical shapes. This indicates that the C10 fragmentation pattern to CI’ is not sensitive to its vibrational level for v’ = 1-5. (16) Fluorescence observed at longer wavelengths (ref 17) should be negligible at these energies. (17) Curl, R. F., Jr.; Abe, K.; Bissinger, J.; Bennett, C.; Tittel, F. K. J . Mol. Spectrosc.1973, 48, 72. (18) Arkell, A.; Schwager, I. J . Am. Chem. Soc. 1967, 89 (24), 5999. (19) Dum, R. C.; Richard, E. C.; Vaida, V.; Simon, J. D. J . Phys. Chem. 1991, 95, 6060. (20) Simon, F. G.; Schneider, W.; Moortgat, G. K.; Burrows, J. P. J . Photochem. Photobiol. A: Chem. 1990, 55, 1. (21) Clyne, M. A. A.; McDermid, I. S.; Curran, A. H. J . Photochem. 1976, 5 , 201. (22) (a) Durie, R. A,; Ramsay, D. A. Can. J . Phys. 1958, 36, 35. (b) Coxon, J. A.; Ramsay, D. A. Can. J. Phys. 1976, 54, 1034. (23) Gole, J. L. J . Phys. Chem. 1980, 84, 1333. (24) In the earlier literature, it was thought that a broad 2Al 2Bl absorption band may be hidden under the structured 2A2 ’B, spectrum. However, according to refs 3 and 12, this is unlikely and the absorption spectrum is attributed to excitation to the 2A2state. (25) Nicovich, J. M.; Kreutter, K. D.; Shackelford, C. J.; Wine, P. H. Chem. Phys. Lett. 1991, 179 (4). 367.

- -