摘要:

FARADAY DISCUSSIONS OF THE CHEMICAL SOCIETY NO. 84 1987 Dynamics of Elementary Gas-phase Reactions THE FARADAY DIVISION THE ROYAL SOCIETY OF CHEMISTRY LONDONOrgan king Com mi t tee Professor R. Grice ( Chairman) Dr M. S. Child Dr J. N. L. Connor Dr M. J. Pilling Professor I. W. M. Smith Professor J. P. Simons ISBN: 0-85186-807-X ISSN: 0301-7249 Printed in Great Britain by J. W. Arrowsmith Ltd, BristolA GENERAL DISCUSSION O N Dynamics of Elementary Gas-phase Reactions 14th, 15th and 16th September, 1987 A General Discussion on the Dynamics of Elementary Gas-phase Reactions was held at the University of Birmingham on 14th, 15th and 16th September, 1987. The President of the Faraday Division, Professor A. D. Buckingham, F.R.S., was in the chair: about 130 members of the Faraday Division and visitors from abroad attended the meeting.Among the overseas visitors were: Professor V. Aquilanti, Italy Dr M . Billy, France Professor M . T. Bowers, U.S.A. Dr N. Bras, France Professor P. R. Brooks, U.S.A. Mr H. Buchenau, Federal Republic of Germany Professor K. N . Chen, Republic of China Dr M . Costes, France Professor P. J. Dagdigian, U.S.A. Dr A. Dell Hammerich, Israel Dr G. Dorthe, r ance Dr T. H. Dur mg JT, U.S.A. Professor R. a u r e n , Federal Republic of Germany M r P. Elofson, Sweden Professor J. M . Farrar, U.S.A. Mr B. Girard, France Professor E. Gislason, U.S.A. Professor D. R. Herschbach, U.S.A. Dr H. Hippler, Federal Republic o$ Germany Professor L. Holmlid, Sweden Professor W. Jakubetz, Austria Mr S. K. Kim, U.S.A.Professor I . Koyano, Japan Dr P. Kuntz, Federal Republic of Germany Dr A. Kvaran, Iceland Professor A. Lagana, Italy Dr G. Lendvay, Hungary Professor S . R. Leone, U.S.A. Mr N. Markovic, Sweden Dr K. G. McKendrick, U.S.A. Dr J-M. Mestdagh, France Professor W. H . Miller, U.S.A. Professor C. Ottinger, Federal Republic of Professor P. D. Pacey, Canada Professor D. H. Parker, The Netherlands Dr G. Parlant, U.S.A. Professor J. C. Polanyi, Canada Mr G. Robinson, U.S.A. Mr H-G. Rubahn, Federal Republic oj’Germany Dr N . Sadeghi, France Professor G. C . Schatz, U.S.A. M r I . Schechter, Israel Mr S. Schlemmer, Federal Republic of Germany Professor D. W. Setser, U.S.A. Dr J. J. Sloan, Canada Professor H. Teitelbaum, Canada Professor P. J. van Tiggelen, Belgium Mr R.Timonen, Finland Professor D. G. Truhlar, U.S.A. Professor A. J . C. Varandas, Portugal Dr R. Vetter, France Dr J. Vigue, France Dr J-P. Visticot, France Mr K. Wagemann, Federal Republic of Dr J. Wanner, Federal Republic of Germany Dr J . Wolfrum, Federal Republic of Germany Professor R. N . Zare, U.S.A. Germ any GermanyA GENERAL DISCUSSION O N Dynamics of Elementary Gas-phase Reactions 14th, 15th and 16th September, 1987 A General Discussion on the Dynamics of Elementary Gas-phase Reactions was held at the University of Birmingham on 14th, 15th and 16th September, 1987. The President of the Faraday Division, Professor A. D. Buckingham, F.R.S., was in the chair: about 130 members of the Faraday Division and visitors from abroad attended the meeting. Among the overseas visitors were: Professor V.Aquilanti, Italy Dr M . Billy, France Professor M . T. Bowers, U.S.A. Dr N. Bras, France Professor P. R. Brooks, U.S.A. Mr H. Buchenau, Federal Republic of Germany Professor K. N . Chen, Republic of China Dr M . Costes, France Professor P. J. Dagdigian, U.S.A. Dr A. Dell Hammerich, Israel Dr G. Dorthe, r ance Dr T. H. Dur mg JT, U.S.A. Professor R. a u r e n , Federal Republic of Germany M r P. Elofson, Sweden Professor J. M . Farrar, U.S.A. Mr B. Girard, France Professor E. Gislason, U.S.A. Professor D. R. Herschbach, U.S.A. Dr H. Hippler, Federal Republic o$ Germany Professor L. Holmlid, Sweden Professor W. Jakubetz, Austria Mr S. K. Kim, U.S.A. Professor I . Koyano, Japan Dr P. Kuntz, Federal Republic of Germany Dr A. Kvaran, Iceland Professor A. Lagana, Italy Dr G.Lendvay, Hungary Professor S . R. Leone, U.S.A. Mr N. Markovic, Sweden Dr K. G. McKendrick, U.S.A. Dr J-M. Mestdagh, France Professor W. H . Miller, U.S.A. Professor C. Ottinger, Federal Republic of Professor P. D. Pacey, Canada Professor D. H. Parker, The Netherlands Dr G. Parlant, U.S.A. Professor J. C. Polanyi, Canada Mr G. Robinson, U.S.A. Mr H-G. Rubahn, Federal Republic oj’Germany Dr N . Sadeghi, France Professor G. C . Schatz, U.S.A. M r I . Schechter, Israel Mr S. Schlemmer, Federal Republic of Germany Professor D. W. Setser, U.S.A. Dr J. J. Sloan, Canada Professor H. Teitelbaum, Canada Professor P. J. van Tiggelen, Belgium Mr R. Timonen, Finland Professor D. G. Truhlar, U.S.A. Professor A. J . C. Varandas, Portugal Dr R.Vetter, France Dr J. Vigue, France Dr J-P. Visticot, France Mr K. Wagemann, Federal Republic of Dr J. Wanner, Federal Republic of Germany Dr J . Wolfrum, Federal Republic of Germany Professor R. N . Zare, U.S.A. Germ any GermanyCONTENTS 1 19 25 39 53 65 75 87 127 145 159 171 191 205 22 1 239 253 265 28 1 293 303 325 333 351 359 Spiers Memorial Lecture: The Dynamics of Elementary Reactions J. C. Polanyi General Discussion Dynamics of Endoergic Aromatic Substitution Reactions G. N. Robinson, R. E. Continetti and Y. T. Lee Product State Distributions from the Reaction O('P) + HBr K. G. McKendrick, D. J. Rakestraw and R. N. Zare Onset of Migration in the Reaction of Fluorine Atoms with Iodine Molecules N. C. Firth, N. W. Keane, D. J. Smith and R.Grice Laser-induced Fluorescence Study of the F+ I2 --+ IF+ I Reaction B. Girard, N. Billy, G. Gouedard and J. Vigue Dynamics of the Reactions of Aluminium Atoms studied with Pulsed Crossed Supersonic Molecular Beams M. Costes, C. Naulin, G. Dorthe, C. Vaucamps and G. Nouchi General Discussion Spin-Orbit Effects in Chemical Reactions. Investigation of Ground-state Products from Reactions of Ba(3D) M. L. Campbell and P. J. Dagdigian Reactive Scattering of Electronically Excited Alkali-metal Atoms with Molecules J. M. Mestagh, B. A. Balko, M. H. Covinsky, P. S. Weiss, M. F. Vernon, H. Schmidt and Y. T. Lee Angular Momentum Disposal in Atom Exchange Reactions S. K. Kim and D. R. Herschbach General Discussion Laser Studies on the Dynamics of Elementary Steps with Translationally and Vibrationally Excited Reactants J.Wolfrum Kinetics of Association and Dissociation in a Weakly Bound System: NO+ NOz N2O3 I. W. M. Smith and G. Yarwood The Dynamics of Electronically Excited States in the Rare-gas- Halogen Sys- tems R. J. Donovan, P. Greenhill, M. A. MacDonald, A. 3. Yencha, W. S. Hartree, K. Johnson, C. Jouvet, A. Kvaran and J. P. Simons Genera 1 Discussion Laser Probing of Product-state Distributions in Thermal-energy Ion- Molecule Reactions State-selected Charge-transfer and Rearrangement Reactions in Four-atom Ion- Molecule Systems Reactive Scattering from Double-minimum Potentials W. R. Creasy and J. M. Farrar Genera 1 Discussion Product Energy Disposal in Simple Thermal-energy Bimolecular Ion- Molecule Reactions Theoretical Studies of Non-adiabatic Processes in Ion- Molecule Collisions: Ar'('P,)+CO Chemical Reactions dominated by Long-range Intermolecular Forces D.C. Clary and J. P. Henshaw General Discussion State-to-state Chemistry with Fast Hydrogen Atoms. Reaction and Collisional Excitation in H + COz S. R. Leone and V. M. Bierbaum I. Koyano, K. Tanaka, T. Kato and s. Suzuki M. T. Bowers and M. Rincon E. A. Gislason, G. Parlant, P. Archirel and M. Sizun G. C. Schatz, M. S. Fitzcharles and L. B. Harding371 387 405 427 44 1 455 465 479 48 1 Y2 Calculations of Accurate Quantal-dynamical Reactive Scattering Transition Probabilities and their Us& to test Semiclassical Applications J. Z. H. Zhang, Y. Zhang, D. J. Kouri, B. C. Garrett, K. Haug, D. W. Schwenke and D. G. Truhlar Dynamics of Heavy + Light- Heavy Atom Transfer Reactions. The Reaction C1+ HCl - CIH + CI B. Amaee, J. N. L. Connor, J. C. Whitehead, W. Jakubetz and G. C. Schatz Genera 1 Discussion Insights into the Mechanisms of Chemical Reactions. Reaction Paths for Chemical Reactions The Reactive Flux Correlation Function for Collinear Reactions H + Hz, C1+ HCl and F+H2 General Discussion Closing Remarks D. R. Herschbach List of Posters Index o f Names T. H. Dunning Jr, E. Kraka and R. A. Eades J. W. Tromp and W. H. Miller

ISSN:0301-7249    

DOI:10.1039/DC98784FP001

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1987, 84, 19-24 GENERAL DISCUSSION Prof. J. P. Simons (University of Nottingham) said: 1s it possible that the photolytic behaviour of adsorbed molecules such as H2S could be drastically altered from that of the isolated molecule in the gas phase? The electronic absorption continuum accessed at 193 nm very likely involves excitation into Rydberg electronic states. These would be very sensitive to perturbation at the surface adsorption site in view of the diffuse spread of the electron density. Indeed the very large increase in the cross-section for photon capture in the adsorbed molecule suggests that its electronic excitation may be associated with charge transfer to (or from) particular sites on the surface. Prof. J. C. Polanyi (University of Toronto) replied: We have in fact found, as Prof.Simons conjectures, that the dynamics of the dissociation of H2S is markedly altered in the adsorbed At low coverage or at ‘high’ crystal temperature (ca. 150 K, at which temperature H2S has largely been desorbed) the H2S is likely to be preferentially adsorbed on defect sites. These have the effect, in 222 nm photolysis, of altering the major photolytic pathway from H2S -+ HS( u = 0) + H to H2S --* HS( u = 2) + S. Following the analysis of these differing pathways given by previous workers in the gas,’ we surmise that this may be due to altered H2S bond angle in adsorption at an active (i.e. defect) site. As regards the very large (103-104x) increase in cross-section for absorption of a photon to give photodissociation, this has been observed for OCS at low coverages but not for CH3Br.We would not wish to exclude the possibility of photoassisted electron transfer from defect sites leading to dissociative attachment; there can be little doubt that this will prove to be an important process for materials or regions having low work function. However, in the case of OCS photodissociation we have observed neutral CO and S photofragments in comparable yields. We have therefore suggested that defect sites may be involved in a different way; the ultraviolet (u.v.) electronically excites an electron trapped in a vacancy, and subsequently E --+ E (electronic-to-electronic) energy transfer excites OCS to a free state. Since E + E transfer can take place at long range, the surface acts as an extended receiver which channels photon energy into the adsorbate.The closest analogy to this process in the gas phase is ‘sensitized photolysis’, in which a species that strongly absorbs U.V. (such as mercury vapour) is mixed with a material to be photodissociated. The novelty here is that the sensitizer, being a solid substrate, extends in a two- or three-dimensional array. 1 E. B. D. Bourdon, P. Das, I. Harrison, J. C. Polanyi, J. Segner, C. D. Stanners, R. J. Williams and P. A. Young, Faraday Discuss. Chem. SOC., 1986, 82, 343. 2 I. Harrison, J. C. Polanyi and P. A. Young, J. Chem. Phys. in press. 3 G. N. A. Van Veen, K. A. Mohamed, T. Baller and A. E. DeVries, Chem. Phys., 1983, 74, 261. Prof. A. D. Buckingham ( University ofCarnbridge) said: In his fascinating Spiers Lecture, Prof.Polanyi mentioned the very large enhancement in the extinction coefficient that may be observed for molecules adsorbed on the surface of a crystal. It is possible that the surface may induce a large increase in intensity without significantly affecting the structure of the adsorbate. In Raman spectroscopy the surface-enhanced Raman scatter- ing (SERS) may be lo6 times as intense as that from an isolated molecule (for a recent review, see Chang’). A major cause of the enhancement of the absorption and scattering is the extra electric field at the active molecule due to its environment. As a simple example, consider two spherical molecules 1 and 2 of polarizabilities a l and a2, separated by a distance R. In an external electric field E, parallel to the line of centres the total 1920 General Discussion field strength at molecule 1 is and the polarizability of the pair is For an identical pair this reduces to If the external field is perpendicular to the line of centres, and the pair perpendicular polarizability is which, for an identical pair, becomes If 4a,a2RP6 (4mO)-’ is comparable to unity, the neighbouring molecule enhances the effective optical field parallel to the line of centres, thereby increasing the absorption coefficient and the scattering power.Near a resonance frequency, the polarizability a l undergoes its anomalous dispersion and may become very large,’ and enhancement may be expected. In the case of a solid substrate, it is necessary to sum eqn ( 1 ) and ( 2 ) over all neighbours 2, and for favourable geometries and polarizations, the enhancement may be large.1 R. K. Chang, Ber. Bunsenges. Phys. Chem., 1987, 91, 296. 2 M. Born and K. Huang, Dynamical Theory of Crystal Lattices (Oxford University Press, 1954), section 18. Prof. J. C . Polanyi (University of Toronto) replied: There are (as for SERS) an embarrass- ing number of sensible explanations for the large increase in photon-capture cross-section for photodissociation of OCS( ads). Prof. Buckingham’s proposal, based upon increased electric field strength due to charge polarization in (say) adjacent F-centre defect sites, represents a further possibility. The observation that this enhancement in cross-section applied to OCS but not to CH,Br might then be explained by a more favourable orientation of the transition moment in OCS relative to the polarization direction.It is perhaps surprising that the effect is so specific. Prof. R. N. Zare (Stanford University, CA) commented: Prof. Polanyi has presented new data on surface-assisted photochemistry which show that photofragments from one surface adsorbate molecule yield products by reaction with other such species on the surface. Should this be regarded as proceeding by some mechanism intermediate between those proposed by Eley and Rideal and by Langmuir and Hinshelwood? In theGeneral Discussion 21 Eley-Rideal mechanism the reaction is imagined to be direct attack of a gas-phase species on an adsorbate molecule, whereas in the Langmuir- Hinshelwood mechanism the reaction takes place via the encounter of two adsorbed species.If you agree that surface-assisted photochemistry is intermediate, which limiting reaction mechanism does it more closely resemble? Prof. J. C. Polanyi (University of Toronto) replied: I think Prof. Zare is quite right in attributing elements of the Langmuir-Hinshelwood (LH) and also the Eley-Rideal (ER) dynamics to surface-aligned photoreaction. The resemblance to the LH mechanism lies in the fact that the molecules involved are both present in the adsorbed state prior to the reactive event. However, in contrast to LH dynamics the molecules do not migrate across the surface in order to interact with one another. Instead a photofragment from one adsorbed molecule (e.g. H from an S-H bond at ca. 45" to the surface) passes sufficiently closely by an adjacent H,S(ads) that .abstraction (-+ H2) occurs.Since the attacking species is moving and the one under attack is fixed, there is a superficial resemblance to the ER mechanism. The difference is that the attacking radical comes from the surface rather than from the gas, and is constrained to move along preferred directions rather than randomly. I would not be inclined to say that this sort of event more closely resembles LH than ER, nor the reverse. Thesis and antithesis have achieved some sort of synthesis. Dr P. J. Kuntz (University of Berlin, Federal Republic of Germany) said: Prof. Polanyi mentioned that the HBr molecules on the surface could by virtue of the H and Br covalent radii, be oriented in only two ways, greatly restricting the range of impact parameters.This is a central point in his interpretation of the experimental measure- ments. Would it be possible to gain support for his ideas by performing the experiment under conditions where one would not expect such restricted collisions (e.g. by choosing another adsorbed molecule, such as HCl, or another substrate) in the hope of seeing a qualitatively different result? Prof. J. C. Polanyi (University of Toronto) replied: Regrettably the experiment to which Dr Kuntz refers was only a gedanken one. We found that if we made the assumption that HBr lies down on a LiF crystal surface up to high coverage, then the HBr molecules can librate around linear or bent arrangements. We used these patterns of alignment (as we could have used any other patterns) in a trajectory study, in order to explore the effect on the product attributes of restrictions in reagent angles and impact parameters.Our actual experiments as usual have not been as tidy as our gedanken ones. Comparing the surface aligned photoreaction in H2S: H + H2S( ads) - H2 + HS with that in HBr: H+HBr(ads) - H2+Br we do see qualitatively different outcomes; bimodal H2 translational energy distribution in the first case, and unimodal in the second. Clearly, much more remains to be done. An interesting variant on Dr Kuntz's proposal is to alter the angle to the surface at which one views the products of photoreaction, in the hope that this correlates to a detectable extent with the dynamics of the prior event. The case of H + H2S(ads) -+ H2 + HS may be a favourable one, since the bimodal H2 translational energy distribution could well arise from two distinctive molecular pathways giving rise to differing product angular distributions.In fact we do see a marked change in the relative importance of the two peaks in the H2 translational energy, as we change our viewing angle. This would not of course be the case for reaction of H coming from H2S which was being photolysed in a bulb (our laser is unpolarised).22 General Discussion Prof. R. Grice ( University ofMunchester) said: In his trajectory studies of surface reaction initiated by photodissociation, of an adsorbed HBr molecule, did Prof. Polanyi find any evidence of secondary collisions of nascent reaction products with the surface or other adsorbed molecules before the product leaves the surface? Prof. J.C. Polanyi (University of Toronto) replied: The potential-energy function that we used was restricted to the three atoms HBrH' (one H comes from photolysis of aligned HBr). We did not include the surface or further adsorbate molecules. As a result we can only conjecture about the importance of secondary encounters. The attacking H moved either parallel to the surface (i.e. 90" to the surface normal) or at somewhat less than 90", the distribution of angles of approach being governed by the libration of the HBr molecule from which the H was photoejected. For this (plausible) scenario collinear reaction scattered most H'Br product into angles <90°, hence for this case secondary encounters would not be expected to be important.For 'bent' reaction they would be more important. Experimentally, for the case of H2S we obtain a large yield of fast moving H2 photoproduct that appears to be the result of direct 'pick-up' reaction: H + H2S(ads) --+ H2(g) + HS. The product translational energy distribution is not degraded to low energy, as one would expect if secondary encounters were important. Mr G. N. Robinson (University of California, Berkeley, CA) said: What are the chances that Polanyi and co-workers are observing collisionally relaxed HD,f from the ion- molecule reaction H f + D2 --+ HDf (AH"= -110 kcal mol-')? Prof. J. C. Polanyi (University of Toronto) replied: Mr Robinson has touched on an important question. We have certainly considered ion-molecule reactions as a potential source of the observed HD; ion.' In order to be a candidate the process must depend in a very specific manner on the time delay between the firing of the first laser (the photolytic source of H atoms) and the second (the probe of the transition state).The background subtraction technique used in these investigations ensured that any measured signal depended on the time delay between two lasers and not on one laser alone. 'Single-laser' ion-molecule reactions such as can be ruled out. D;+H,S -+ HD;+HS Conceivable 'two-laser' ion-molecule reactions would be H+D;+M -+ HD;+M and Ht+D2+M -+ HD;+M (the ion HD; can be formed only if there is a third body, M present). Using a reasonable three-body ion-molecule association rate, we estimate that such a source of HD; would yield a concentration at m / e = 5 over seven orders of magnitude smaller than our measured signal.Preliminary photolysis-laser and probe-laser power studies are also consistent with transition-state spectroscopy of HD;. A linear photolysis laser power dependence was obtained for the HD; ion as was expected for a one-photon dissociation of H,S to yield atomic reagent H. A power dependence of l'.4 was obtained for the probe laser, consistent with two-photon resonantly enhanced ionization of HD:. Since the power dependence for the formation of D l is markedly different, namely 1 2 3 , we have a further reason to discount processes involving D i as the source of HD;. A similar argument applies to H+, since the formation of this ion would require at least three photons from a single laser.General Discussion 23 interqction ion dete Fig.1. The most positive identification of the observed HD,’ ion signal as stemming from H + D2 -+ HD: -+ HD+ D was the functional form of its time dependence, presented in the Spiers lecture and in our recent communication.’ 1 B. A. Collings, J. C. Polanyi, M. A. Smith, A. Stolow and A. W. Tarr, Phys. Rev. Lett., 1987, 59, 2551. Dr D. H. Parker (University of California, Santa Cruz, CA), Mr M. H. M. Janssen (University of Nijmegen, The Netherlands) and Dr. D. W. Chandler (Sandia Livermore Laboratories) said: Prof. Polanyi has advanced laser ionization as a powerful detection method for chemical reaction dynamics. Ion imaging is another dimension of laser ionization that should be applicable to cross-beam scattering studies for determination of the nascent product relative velocity vector, k’, as well as the alignment of product molecule rotational angular momentum.Removal of an electron from the nascent product species does not affect k’ in the absence of external electromagnetic fields or space-charge effects (i.e. too many ions in the focal volume). This can be exploited as follows: A laser is tuned to ionize state- selectively the chosen product in the interaction region. Once ‘tagged’ these ions propagate outwards in a pattern determined by k’, until a suitably large spatial distribu- tion is reached. A two-dimensional image of the pattern is then obtained by quickly applying an electric field to ‘crush’ the ions onto a flat, position-sensitive detector.Mass selective detection is possible by time-of-flight analysis. Ion imaging has been demonstrated recently in studies of a ‘half-collision’ system, the photodissociation of CH31,’ and CD31.’ Fig. 1 shows the apparatus and a sample result. We create an anisotropic k’ distribution of CD3 radicals by photodissociation of CD31 with a vertically polarized 266nm laser. Several nanoseconds later a second laser state-selectively ionizes CD3 to form CD:. For a given internal state of the CD, fragment two product channels are possible (formation of I or I*) thus two lobes of CD,’ ions propagate along the vertical axis. A few microseconds after ionization the pattern is flashed onto an 8 cm diameter microchannel plate/phosphor screen detector by application of a voltage to the repeller plate in order to create a visible image. A picture of the image produced by ionizing v = 2 of the v2 CD3 umbrella mode is also shown in fig. 1. The two channels are clearly separated and the (p, k ’ ) correlation ( p = 1.8)3 of the photodissociation event is obtained after a 2D-3D transformation.’ By placing a mask on the image and scanning the ionization laser, spectra of CD3 produced predominantly via either the I or I” channels can be measured. Anisotropic rotational24 General Discussion angular momentum distributions can also be probed by simply varying the plane of polarization of the ionization laser. By replacing the photodissociation laser with an atom beam a similar measurement of reactively scattered product molecules can be envisaged. Ion imaging provides vector property information along with internal state selectivity at essentially the same sensitivity as ‘normal’ laser ionization. 1 D. W. Chandler and P. L. Houston, J. Chem. Phys., 1987, 87, 1445. 2 J. W. Thoman Jr, D. W. Chandler, D. H. Parker and M. H. M. Janssen, Laser Chemistry, 1988, in press. 3 M. J. Dzvonik, S. C. Yang and R. Bersohn, J. Chem. Phys. 1974, 61, 4408.

ISSN:0301-7249    

DOI:10.1039/DC9878400019

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1987, 84, 25-37 Dynamics of Endoergic Aromatic Substitution Reactions Gary N. Robinson, Robert E. Continetti and Yuan T. Lee Materials and Chemical Science Division, Lawrence Berkeley Laboratory, and Department of Chemistry, University of California, Berkeley, CA 94720, U.S.A. The endoergic substitution reactions Br+ 0-, m-, p-CH3C6H4Cl + 0-, rn-, p-CH,C6H4Br+ C1 (AH: == 15 kcal mol-') have been studied using the crossed beams method in the collision energy range 20-30 kcalmol-I. o-Chlorotoluene was found to be more reactive than p-chlorotoluene at the highest energy but the reverse was true below 25 kcal mol-'. No reaction was observed for the meta isomer. An explana- tion for the lower reactivity of rn-chlorotoluene is offered in terms of possible features of the potential-energy surface.In all of the reactions observed, the products are largely forward scattered, indicating that the majority of collision complexes survive for less than one rotational period. This is understandable in light of the ca. 2 kcal mol-' endoergicity to Br addition that results from the loss in resonance stabilization energy. Very little of the energy available to the products of these reactions is channeled into transla- tion. The experimental product translational energy distributions and excita- tion functions suggest that, in those complexes that decompose through CI elimination, only a few vibrational degrees of freedom in the vicinity of the collision are involved in energy redistribution. Although homolytic, or free-radical, aromatic substitution reactions have been the subject of many kinetic studies,'-3 their detailed dynamics in both the liquid and gas phases are only partially understood.In the gas phase they proceed by addition of an atom or radical to an aromatic ring to form an activated cyclohexadienyl radical which subsequently decomposes through emission of another atom/radical. Questions relating to the dynamics of atomic addition to the ring, the features of the potential energy surface along the reaction coordinate, the extent of intramolecular vibrational energy redistribution prior to unimolecular decomposition and the relative importance of all of these factors in determining 'the energy dependence of the reactive cross-section can all be addressed by the method of crossed-beams scattering.Endoergic substitution reactions are particularly intriguing, since cleavage of the stronger bond is not the statistically favoured mode of decay of the chemically activated radical. Thus, observing the endoergic channel may offer some insight into factors other than the amount of phase space available to the activated radical and products that determine the course of the reaction. In the case of aromatic substitution reactions, one may, in addition, observe the effects of substituents on the reactivity of different sites on the aromatic ring. Since these effects are related to the energetics of adduct formation, they should be most pronounced in those reactions where the reagent atom/radical bonds weakly to the ring. In this respect Br is an ideal reagent.We have carried out a crossed-molecular-beam study of the endoergic substitution reactions (fig. 1 ) Br+ 0-, m-, p-CH3C6H4Cl + C1-k 0-, m-, p-CH3C6H5Br in the collision energy range of 20-30 kcal mol-'. The results of these experiments shed new light on the role of substituents in controlling the orientation of aromatic substitution and on the dynamics of these reactions in general. (A&* 15 kcal mol-') 2526 Endoergic Substitution Reactions 3 5 1 C7H7Br + CI -51 -10 I Fig. 1. Generalized reaction coordinate diagram. Shaded region indicates approximate collision energy range. Experimental The crossed-beam apparatus used in these experiments has been described previously.435 Two seeded, differentially pumped reagent beams cross at 90" in a vacuum chamber held at ca.lO-'Torr.t The products are detected with a triply differentially pumped mass spectrometric detector that rotates in the plane of the two beams. The bromine atom beam was generated by passing a mixture of Br, in rare gas through a resistively heated high density graphite oven designed in this laboratory by Valentini et aL6 The Br,-rare gas mixture is created by bubbling ca. 700 Torr of He, Ne or Ar through liquid bromine (reagent-grade, Fischer and Mallinkcrodt) at 0 "C [ p ( Br,) a 60 Torr]. The oven had a nozzle diameter of 0.14mm and was run at ca. 1380 "C. A conical graphite skimmer having an orifice diameter of 0.10 cm was positioned 0.76 cm from the nozzle. 90% of the BrZ dissociated into Br atoms, as determined from a direct measurement of Br/Br2 in the beam.The secondary molecular beam was formed by bubbling 450Torr of He through chlorotoluene heated to 60°C in a bath and expanding the mixture through a 0.21 mm diameter aperture nozzle. A stainless-steel skimmer with an orifice diameter of 0.66 mm was positioned 0.89cm from the nozzle. The source and feed line were heated with coaxial heating wire to a temperature of 200 "C. o- and p-Chlorotoluene (o-, p-CT) were purchased from MCB and m-CT from Aldrich. All of the compounds were used without further purification, except for the p-CT, which was distilled on a spinning band column. In order to reduce the background at the product mass, a liquid-nitrogen-cooled copper cold finger was placed against the differential wall inside the scattering chamber so that the detector would always face a cold surface during the angular scans.Product angular distributions were measured by modulating the CT beam with a 150 Hz tuning fork chopper and collecting data with the beam on and off using a dual channel scaler. Data were collected for ca. 6 min per angle. t 1 Torr = 101 325/760.Pa.G. N. Robinson, R. E. Continetti and Y. T. Lee 27 h v) U .- E - 5 - - - - - 0 ' 1 - I I I I I I 1 7 I I ' 10 20 30 40 50 60 70 20 30 40 50 60 70 I laboratory angle, @ / O laboratory angle, @/' 3 Fig. 2. 0- and p-BT laboratory angular distributions ( m / e = 170) normalized to constant reactant flux. Br beam is at 0". Solid lines are fits to data. Arrows indicate positions of centre-of-mass angles with collision energy decreasing from left to right.( a ) o-BT: E, = 31 .O; E, = 25.3; A E, = 20.9 kcal mol-'. (b) p-BT: 0 E, = 31.5; E, = 25.3; A E, = 21.4 kcal mol-'. In order to compute relative cross-sections for a given reaction at different collision energies, we scaled the product number density by the reactant flux, which is proportional to nBrncTurel, where ni = number density of beam i and ureI = relative velocity. Since the wide-angle Br elastic scattering cross-section does not change drastically as a function of energy, measuring Br on CT elastic scattering allows us to measure changes in this quantity. During each scan, the rn/e = 79 signal was monitored at three different LAB angles. The angles were all beyond the cutoff angle for elastic scattering of Br on He so the Br+ signal observed was from Br scattering on CT.The contribution of undissoci- ated Br2 to the rn/e = 79 signal was very small and was neglected. Relative values for nBrnCTurel derived from the Br elastic number density at angle of 16" are given in table 1. The velocities of the reactant beams were measured using the time-of-flight (TOF) technique. A 256-channel scaler interfaced with an LSI-11 computer accumulated the data. No TOF measurement was made for the rn-CT beam, but its velocity should be identical to that of o-CT since both have the same vapour pressure (22 Torr; p-CT = 21 Torr) at 60 "C. The peak beam velocities (in units of lo5 cm s-'), and speed ratios, S, are: Br/He: 1.85, S = 6.1; Br/Ne: 1.55, S = 6.9; Br/Ar: 1.29, S = 8.4; o-CT/He: 1.33, S = 11.3; p-CT/He: 1.35, S = 11.8.Product TOF were measured using the cross-correla- tion method.' Counting times were ca. 1 h per angle. Results and Analysis The 0- and p-bromotoluene (BT) substitution products were detected at m / e = 170 ( 79Br), however, the quadrupole mass spectrometer resolution was set sufficiently low to allow some of the *lBr product to be detected as well. The product angular distributions are shown in fig. 2. Elastic scattering of impurity in the p-CT beam contributed to background at rn/e = 170 near that beam. This was most problematic at a collision28 1.0 - Endoergic Substitution Reactions I I I 1 I I h - . h v) U ._ E 1.0 0.5 0.0 0.5 0.0 1.0 0.5 0.0 200 300 400 500 flight time/ps I I I I 100 200 360 400 500 flight time/ p s Fig.3. Time-of-flight spectra of p-BT, E, = 31.5 kcal mol-' ( m / e = 170), at five different angles. Solid lines are fits to data. energy, E,, of 21.4 kcal mol-', where the product signal level was lowest. At this energy the elastic scattering background was measured by substituting a properly diluted beam of Kr in Ar for the Br in Ar beam. It was then scaled to the product angular distribution at 74" and subtracted from it. At the peak of the E, = 31.0 kcal mol-' o-BT angular distribution, the product count rate was 20 Hz. The angular distributions reveal that, at all energies, the 0- and p - products are mostly forward scattered with respect to the centre-of-mass angle for the collision. Remarkably, no BT product was detected for the reaction Br+ rn-CT at a collision energy as high as 29.3 kcal mol-'.TOF spectra of p-BT, E, = 31.5, 21.4 kcal moll', and o-BT, E, = 25.3, 20.9 kcal mol-I, are presented in fig. 3 and 4. At each collision energy the TOF spectra at different angles have peak flight times close to that of the velocity of the centre-of-mass (fig. 7, see later), indicating that little energy is channeled into translation. The product angular distributions and TOF spectra were fitted using a forward convolution program' that starts with a separable form for the centre-of-mass (CM) reference frame product flux distribution, ICM( 8, E ' ) = T( 8 ) P ( E'), and generates labora-G. N. Robinson, R. E. Continetti and Y. T. Lee 29 10 0.5 0.0 1.0 h v) c1 .e c -e 0.5 h \.) v 2 0.0 1.0 0.5 0.0 160 260 360 460 260 360 460 5 flight time/ps flight time/ps 0 Fig.4. ( a ) TOF spectra of o-BT, E,= 25.3 kcal mol-'; (b) TOF spectra of p-BT, E, = 21.4 kcal mol-'; ( c ) TOF spectrum of o-BT, E, = 20.9 kcal mol-'. Solid lines are fits to data. tory (LAB) frame angular distributions and TOF spectra suitably averaged over the spread in relative velocities. T ( 8 ) , the CM frame angular distributions, is taken to be a sum of three Legendre polynomials whose coefficients are varied to optimize the fit. A RRK functional form is used for P ( E ' ) , the CM frame product translational energy distribution : where B is related to any barrier in the exit channel and E,,, is the total energy available to the products (E,-AH;). AH: was taken to be 15 kcal mol-' (see Discussion). The parameters p , q and B were optimized to give the best fit to the data.For a given experiment, the spread in beam velocities and intersection angles gives rise to a spread in relative velocities and hence in collision energies. A EJ E,( FWHM) = 30% for the reactions with Br seeded in He and =25% for the reactions with Br seeded in Ne and Ar. Each beam velocity and intersection angle permutation corresponds to a different kinematic configuration (Newton diagram) over which the calculated angular distribution and TOF fits must be averaged. The collision energies corresponding to P ( E ' ) = ( E ' - B ) P ( E,,, - E')'30 Endoergic Substitution Reactions the most probable kinematic configurations are listed in table 1. Since, for an endoergic reaction, the maximum translational energy of the products will depend strongly on E,, a P ( E ' ) with a unique value of E,,, was used for each kinematic configuration in the analysis.Also, since the cross-section is found to depend on collision energy, each kinematic configuration was weighted according to E,. Because of the large spread in E,, it was necessary to extrapolate the excitation function used in the weighting routine beyond the most probable experimental collision energies. This was done by taking the cross-section at E, = 15 kcal mol-' to be 0.0 and extrapolating linearly beyond the highest energy. The high-energy extrapolation has a marked affect on the fits to the E,= 31 kcal mol-'data, since it is used to determine the most probable collision energy and centre-of-mass angle.Linear extrapolation, however, appears to be the most unbiased approach to this problem. Although the data offer some latitude with regard to the exact form of the CM angular and energy distributions, they do place certain constraints on the fits. The best fits were obtained with T ( 8 ) distributions that peak at 0 and 180" with maxima at 0" (fig. 5). There is a range of acceptable values for the P ( E ' ) parameters [as there is for the coefficients of the Legendre polynomials that constitute T ( 8 ) ] , yet the average energy, ( E ' ) (table l), does not vary much within this range. Since the fits were relatively insensitive to the q parameter, which governs the curvature of the tail of the P ( E ' ) , this parameter was fixed for all the fits and the other parameters optimized. The resulting P ( E ' ) distributions (fig.6) peak between 0.0 and 1.2 kcal mol-', with the o-BT P ( E ' ) distributions peaking at lower energies and having slightly lower values of ( E ' ) than those for p-BT. The following changes in the P ( E ' ) for p-BT, E, = 25.3 kcal mol-', while not significantly affecting the fit, produced the indicated changes in ( E ' ) : *25% in q, stlo% in ( E ) ; k0.4 kcal mol-' in peak position, *2'/0 in ( E ' ) ; *2 kcal mol-' in endoergicity, ca. *15% in ( E ' ) . A CM frame product flux contour diagram for Br+ p-CT+p-BT+Cl, E, = 31.5 kcal mol-', is given in fig. 7. The overall quality of the fits justifies our use of a separable form for the CM flux distribution. The asymmetric CM angular distributions that we obtain indicate that the majority of 1-bromo- 1 -chloro-2-( -4-)-methylcyclohexa-2,4-dienyl (BCMC) complexes decompose in a time less than one rotational period.8b The p-BTCM angular distributions show more forward-backward symmetry at lower collision energies, suggesting that the lifetime of the BC4MC complex increases relative to its rotational period as E, decreases.We Table 1. Relevant experimental quantities reaction E," ( E ' ) S , (arb. units) nBrnCT~rel Br/ He + p-CT 31.5 5.0 0.7 1 0.98 Br/ Ne + p-CT 25.3 3.1 0.48 0.59 Br/Ar + p-CT 21.4 2.1 0.3 1 0.53 Br/ Ne + o-CT 25.3 2.9 0.45 0.54 Br/Ar+ o-CT 20.9 1.9 0.18 0.44 Br/ He + rn-CT CQ. 29.3 - 0.00 b Br/ He + o-CT 31.0 4.6 1 .oo 1 .oo All energies are in kcal mol-'; collision energies reflect cross-section weighting.' The rn-CT reaction was studied several weeks after the 0- and p-CT experiments were completed. The Br/ He + o-CT angular distribution was remeasured at this time, however. The o-BT and Br elastic signal levels indicated that the Br beam intensity was ca. 50% lower than during the earlier experiments; the o-BT signal-to- noise ratio had dropped by 20%. However, given the presence of elastic scattering background in the rn-CT experiment (ca. 2 Hz at 46"), it is doubtful that would have been able to see signal even if the Br beam were twice as intense.G. N. Robinson, R. E. Continetti and Y. T. Lee 31 CM angle, 8 / O CM angle, 8/" Fig. 5. CM frame product angular distributions for ( a ) o-BT and ( b ) p-BT. (-) o-BT: E, = 31.0; p-BT: E, = 31.5 kcal mol-'; (- - -) 0- andp-BT: E, = 25.3 kcal mol-'; (- .-) o-BT: E, = 20.9; p-BT: E, = 21.4 kcal mol-'. h Y 4 v Fig. 6. CM frame product translational energy distributions for ( a ) o-BT and ( b ) p-BT. (-) o-BT: E, = 31.0; p-BT: E,= 31.5 kcal mol-'; (- - -) o- and p-BT: E, = 25.3 kcal mol-'. (- . -) o-BT: E, = 20.9; p-BT: E, = 21.4 kcal mol-'. ( - .) Four-mode RRKM translational energy distribution.32 Endoergic Substitution Reactions Fig. 7. CM frame product flux contour diagram for p-BT, E, = 31.5 kcal mol -'. Scale given is for contour diagram. Scale of Newton diagram is half that of contours. Centre-of-mass velocity vector is represented by arrow between beam vectors. can estimate the rotational period of the BC4MC complex by assuming, for the sake of simplicity, that the Br atom collides perpendicular to the ring with an impact parameter of 0.9 A (the distance from the centre of mass of p-CT to the chlorinated carbon) and that the rotational angular momentum of the reagent is negligible.For the collision of Br with p-CT, E, = 31.5 kcal mol-', the magnitude of the orbital angular momentum, L, will be 160 A. The moment of inertia about the rotation axis of the complex is ca. 880 amu A', assuming that the halogenated carbon is sp3 hybridized, that the C-Br and C-C1 bond lengths are 2.0 and 1.7 A, respectively, and that the ring is undistorted. The rotational period, given by T , , ~ = 27rI/L, will therefore be ca. 5 ps in the present example. At E, = 21.4 kcal mol-', T , , ~ = 7 ps. If we calculate the approximate product orbital angular momentum, L', for the p-CT reaction, E, = 3 1.5 kcal mol-', using a relative velocity corresponding to ( E')p- HT = 5.0 kcal mol-' and an impact parameter of 0.1 A (the distance between the chlorinated carbon and the centre-of-mass of the complex, with the C-Cl bond perpendicular to the ring and the C-Br bond in the plane of the ring), we obtain IL'I == 6 h, far lower than the initial 160 h.It would take an average exit impact parameter of 2.7 A for the total angular momentum of the complex to be carried away as product orbital angular momentum. However, even if most of the angular momentum of the collision were carried away in rotation of the BT product, the rotational energy of the product would be small (only ca. 1 kcal mol-' for p-BT in the present example) because of its large moment of inertia ( I = 770 amu A2 for p-BT).The lack of a strong correlation between L and L' is the reason why the CM angular distributions do not peak more strongly in the forward and backward directions.8" The larger amount of sideways scattering for o-BT at 3 1 .O kcal mol-' could indicate an even weaker L to L' correlation in the Br + o-CT reaction at high collision energies. This may be due to the more complicated rotational motion of the asymmetric BC2MC complex. A fraction of the translational energy of BT must come from rotation of the complex at its exit transition state (TS). In the absence of extensive vibration-rotation couplingG. N. Robinson, R. E. Continetti and Y. T. Lee 4 / / / / / / / I I I I 16 20 25 30 .EJkcal mol-' 33 Fig. 8. Plot of relative cross-section us. collision energy. 0 p-BT; o-BT; 0 rn-BT. (-) Three-mode RRKM branching ratio curve. Normalized to equal (0.99)S,,,-BT at 31.5 kcal mol-'. (- - -) Six-mode RRKM branching ratio curve. Normalized to equal (O.97)Sr,,-,, at 3 1 .O kcal mol-.'. in the complex, the rotational energy at this TS will be ca. 1.2 kcal mol-' for BC4MC ( E , = 31.5 kcal mol-', C-CI bond perpendicular to the ring with a bond length of 2.6 A9). If this energy went entirely into relative motion of the products, p-BT would acquire only 0.2 kcal mol-' in translation. The rotational motion of BC2MC will, as noted above, be more complex. However, the fact that the o-BT P ( E ' ) distributions peak at slightly lower energies than those for p-BT could indicate that the ortho complex has a lower rotational energy at its exit TS than the para complex.Lastly, relative cross-sections, S,, were calculated at the most probable collision energies by integrating the CM frame product flux: S , = 2 ~ 1 ~ ~ 1; P ( E ' ) T( 6) sin 8 dE' do. The computed S , values were used to weight the collision energies used in the analysis. This procedure was repeated until the input and output values of S , agreed. Final values of S, as a function of E, are given in table 1 and are plotted in fig 8. A range of collision energies contributes to each value of S,, although we assign each to a single, most probable collision energy. This spread in E, is the dominant source of uncertainty in the derivation of S,. Another source, however, is the uncertainty in the form of the P ( E ' ) .By fixing the q parameter in the fits, we believe that we have eliminated this source of uncertainty in the relative cross-sections.34 E ndoergic Substitution React ions Discussion The endoergicities of the different isomeric reactions under study should not differ markedly from one another. The heats of formation of 0-, rn- and p-CT [AH&8(g)] are 3.8, 4.1 and 5.3 kcal mol-', respectively.'09'' We were able to find heat of formation data for the para isomer of bromotoluene (BT) only [AHLg8(g) = 13.0 kcal m ~ l " ] , ' ' ~ ' ~ but Szwarc's indicates that the C-Br bond dissociation energies in 0-, rn- and p-BT differ by only 0.6 kcal mol-'. Using the known values for AH&98 of Br, Cl,30 p-CT and p-BT, we calculate AH&8 = 10.1 kcal -mol-' for the reaction Br+p-CT-p-BT+ C1.This value strikes one a? being too low, considering that = 15 kcal mol-' for the reaction Br+ C6H5Cl -+ C6H5Br+ CLi4 In the absence of firm values for the heats of formation of the CT and BT isomers, we have used an endoergicity of 15 kcal mol-' for the present reactions. The energetics of Br addition to CT are, as far as we can tell, unknown. Ref. (15) gives AH = -8.8 kcal mol-' for Br+C2H4 + C2H4Br. The exothermicity of Br addition to benzene will be decreased by the loss in resonance stabilization energy ( E J that results from the disruption of the .rr-electron framework of the ring. In the case of H atom addition to benzene the loss in E,, will be CQ.1 1 kcal mol- ' . I 6 We conclude, therefore, that the BCMC radical will be unbound by ca. 2 kcal mol-' relative to reactants! Benson et a1." arrive at a similar value for the endothermicity of Br addition to benzotrifluoride. As a result, we do not expect there to be a potential minimum along the reaction coordinate corresponding to the BCMC complex. Based on the energetics for Br addition to CT, it is not surprising then that substitution occurs in less than one rotational period. We have calculated RRKM lifetimes, T R K K M , for the BCMC complex,20 using modified normal mode frequencies for toluene, and frequencies corresponding to C-Cl and C-Br stretching and Br-C-CI, C-C-Br and C-C-Cl bending modes.*l Including all 42 frequencies, TRRKM at E, = 30 kcal mol-' ( E * = 28 kcal mol-') is 0.02 ps, much lower than the estimated rota- tional period.TRRKM changes little as the collision energy is lowered, indicating that a quantitative comparison of the angular distribution data with the lifetime and rotational period calculations is not possible. It is interesting to note, however, that product angular distributions measured for the reactions Br + CH2CC12 - C1+ CH2CBrCl" are more symmetric about the centre-of-mass angle. In this reaction, adduct formation is indeed exoergic. Remarkably, very little of the energy available to the products ends up in translation. Apparently the vibrational modes of the aromatic ring act as a strong energy sink, with C1 elimination occurring only when sufficient energy has accumulated in the C-C1 bond.Yet, although intramolecular vibrational energy redistribution appears to be extensive prior to C-Cl bond rupture (there is certainly much in the current literature that indicates that, at these energies, it should be19) it does not seem likely that C1 elimination from BCMC is a statistical process involving energy sharing among a large number of vibrational degrees of freedom. We have calculated RRKM-AM P ( E ' ) distributions22 for BT, E, = 3 1 kcal mol- ', using different numbers of modes and a variety of values for the maximum centrifugal barrier, B,. Since the activation energies for C1 addition reactions are known to be very near zero,23 we have no reason to expect that there will be a barrier above the threshold for C1 elimination. Although a definitive comparison between the experimental and RRKM P( E ' ) distributions is not possible given the uncertainties in the fits, we obtain reasonable agreement between the experi- mental o-BT P ( E ' ) , E, = 3 1 .O kcal mol-', and a four-mode ( Y = 800-700 cm-') RRKM P( E ' ) with B,, = 0.028 (fig.6). [Using six modes ( Y = 900-700 cm-') gave a P ( E ' ) that fell too steeply.] Thus the data seem to indicate that only a limited number of degrees of freedom in the vicinity of the collision participate in energy sharing prior to C1 elimination.G. N. Robinson, R. E. Continetti and Y T. Lee 35 Such a mechanism is not unexpected. Endoergic substitution at collision energies not far from threshold must occur in a quasi-direct fashion or not occur at all since, as more vibrational modes participate in energy redistribution, the probability of C1 elimination, qcl, drops relative to qBr. This is due to the fact that qx is proportional to the density of states at the TS for X elimination.The smaller the number of active vibrational modes, the smaller the difference between the state densities for the exoergic and endoergic channels and the more C1 elimination will compete with Br elimination. For example, taking qx to be the microcanonical RRKM rate constant for X elimination, at E, = 30 kcal mol-' qBr/ qcl = 270 with 12 active modes ( Y = 800-300 cm-I), whereas with six active modes qBr/ qcl = 15. We attempted to detemine the relative importance of the Br elimination channel in both the p-CT reaction, E, = 21.4 kcal mol-I, and the o-CT reaction, E, = 31 .O kcal mol-', by measuring the TOF of CT from the channel BCMC --* Br + CT near the centre-of-mass angle.In both cases the TOF of non-reactively scattered CT obtained by substituting Kr for Br was very similar to that obtained with Br. This indicates that the Br addition cross-section is substantially smaller that the elastic/inelastic scattering cross-section. Yet, if qBr were indeed two orders of magnitude larger than qcl (as one would predict from a 12-mode RRKM calculation), we would have been able to see a substantial peak in the o-CT TOF spectrum corresponding to slow o-CT travelling at the velocity of the centre-of-mass, since the fast and slow components of the elastic scattering TOF spectrum were well resolved at E, = 31.0 kcal mol-'.Thus, our inability to observe slow o-CT provides additional (although indirect) evidence that only a few modes are active during the reaction. Further support for the reduced mode mechanism comes from an examination of the excitation functions. The measured relative cross-sections can be expressed as the product of the cross-section for forming the BCMC adduct and the relative probability of decomposition of the adduct through c1 elimination, s, = cradd[ qcl/(qcl + TBr)]. If cr,dd were constant over the energy range studied, and intramolecular energy randomi- zation were complete prior to atomic elimination, the quantity in brackets would be equivalent to the RRKM branching ratio, S R R K M . We have calculated S R R K M for the present system, using 35 ( v = 1600-200 cm-I), 12, 6 and 3 ( v = 800-700 cm-') modes. The 35- and 12-mode curves both rise steeply with energy and are essentially identical in slope.The 3- and 6-mode curves, scaled to Sr,p-BT and Sr,o-BT respectively, are plotted in fig. 8 alongside the experimental results. There is good qualitative agreement between the Slopes Of s ~ ~ ~ ~ ( 6 - m O d e ) and S r , o - ~ ~ , and between the Slopes Of s ~ ~ ~ ~ ( 3 - m O d e ) and Sr,p-BT* It is certainly plausible that the BC2MC collision complex has a larger number of active vibrational modes than BC4MC. We have already noted that the o-BT P ( E ' ) distributions have slightly lower values of ( E ' ) than those for p-BT. The reduced symmetry of the BC2MC complex may allow for enhanced vibrational energy redistribu- tion through state mixing.Coupling of the internal rotation of the methyl group to the ring vibrations is believed to be responsible for accelerated IVR in S, p - f l u o r ~ t o l u e n e . ' ~ ' ~ ~ ~ Although the barrier to methyl torsion is likely to be higher in o-CT than in p-CT,' the methyl group is closer to the collision site in the ortho isomer. The normalization factor used to scale S ~ ~ ~ ~ ( 6 - m o d e ) to Sr,o.BT is a factor of five higher than that used to scale s ~ ~ ~ ~ ( 3 - m O d e ) to Sr,p-BT, indicating, in the present context, that 0 , d d for Br+ o-CT is five times higher than for Br+p-CT. A higher addition cross-section for o-CT can be rationalized along the above lines. The greater number of active modes in BC2MC might serve to dissipate the energy of the collision better, allowing Br to add more readily to o-CT than to p-CT.There still remains the possibility that (T& changes with energy. If this were true, one could not attribute the energy dependence of S, solely to the statistical branching ratio. Using a semi-empirical potential-energy surface (PES) to calculate classical36 Endoergic Substitution Reactions trajectories, Hase et a1.26 found that the cross-section for H atom addition to C,H, varies with collision energy and that this variation is dependent on the shape of the entrance valley of the PES. Perhaps, then, the lower cross-section that we observe for o-CT substitution below E, = 25 kcal mol-' reflects a narrowing of the acceptance angle of the PES as a result of the presence of the methyl group? If this were the case, the methyl group would also raise the effective threshold for C1 elimination.One might imagine, therefore, that the o-BT excitation function could be modelled using a higher endoergic threshold and a smaller number of modes. We investigated this by calculating SRR,, using three modes and Eo = 17-20 kcal mol-'. Although the closest agreement with Sr,o-BT was obtained with Eo = 18 kcal mol-', the calculated branching ratio fell much more rapidly to zero than the experimental excitation function. If a reduced cross-section for Br addition at lower energies were included, the calculated excitation function would disagree even more sharply with the experimental result. Based on our experimental signal levels, the substitution cross-section for rn-CT must be at least a factor of 10 lower than for o-CT at E, = 31.0 kcal mol-'.It is well known that the methyl group is an ortho-para directing substituent in electrophilic (ionic and atomic) substitution reactions. This phenomenon is usually explained in terms of the electron donating capability of the methyl group, which stabilizes the o- and p- adducts by either increasing the 0- and p- frontier electron populations in the reactant molecule or lowering the total .;rr-electron energy of the 0- and p-adducts relative to the r n - a d d ~ c t . ~ ~ Considering the large excess of translational energy at E, = 31.0 kcal mol-', however, it is difficult to understand how the decreased stability of the BC3MC complex could cause the rn-CT cross-section to be so low.A possible explanation is that the reactivities of the isomers of CT are governed by the shape of the Br-CT PES rather than by fixed barriers. The increased electron populations ortho and para to the methyl group in CT could enhance the long-range attraction between Br and these sites, but again, this effect is unlikely to be strong at high collision energies. The shape of the potential in the exit valley might be the key.'* By microscopic reversibility, the reverse reactions, C1+ 0-, p-BT -+ Br + o,p-CT, must also be accelerated. A longer range attraction between C1 and 0- and p-BT will manifest itself in a more gradually sloping potential in the reverse endoergic direction. Alterna- tively, the lower 7r-electron energies of the BC(2,4)MC complexes could cause the o- and p- surfaces to rise more gradually.In either case, translational energy will be better able to promote the endoergic reaction. Classical trajectory studies on several different potential-energy surfaces lend support to these ideas. Polanyi et al.2xb have observed that translational energy is favoured over vibrational energy in endoergic reactions with a gradual ascent to the barrier crest. Likewise Chapman" has found that the curvatures of the Be+ HF- BeF+ H and NO+ O3 + N0,+02 surfaces have marked effects on the excitation functions and product energy distributions of these reactions. In conclusion, both limited intramolecular vibrational energy redistribution and the slope of the potential-energy surface along the reaction coordinate are likely to be responsible for the interesting dynamics that we observe for these substitution reactions.We thank Ms Anne Williamson for her assistance during these experiments and Prof. Sidney Benson, Prof. Andrew Streitweiser, Dr David Golden and Dr Gil Nathanson for helpful discussions. This work was supported by the Director, Office of Energy Research, Office of Basic Energy Sciences, Chemical Sciences Division of the U.S. Department of Energy under contract no. DE-AC03-76SF00098. References 1 G. H. Williams, Hornolytic Aromatic substitution (Pergamon, New York, 1960). 2 M. J. Perkins, in Free Radicals, ed. J . Kochi (Wiley, New York, 1973), vol. 2, pp. 231-271.G. N. Robinson, R. E. Continetti and Y.T. Lee 37 3 J. March, Advanced Organic Chemistry (Wiley, New York, 3rd edn, 1985) chap. 14. 4 Y. T. Lee, J. D. McDonald, P. R. LeBreton and D. R. Herschbach, Rev. Sci. Instr., 1969, 40, 1402. 5 R. K. Sparks, Ph. D. Thesis (University of California, Berkeley, California, 1979). 6 J. J. Valentini, M. J. Coggiola and Y. T. Lee, Rev. Sci. Instr., 1977, 48, 58. 7 R. J. Buss, Ph. D. Thesis (University of California, Berkeley, California, 1979). 8 ( a ) W. B. Miller, S. A. Safron and D. R. Herschbach, Discuss. Faraday Soc., 1967,44, 108; ( b ) G. A. 9 H. B. Schlegel and C. Sosa, J. Phys. Chem., 1984,88, 1141. Fisk, J. D. McDonald and D. R. Herschbach, Discuss. Faraday Soc., 1967, 44, 228. 10 D. R. Stull, E. F. Westrum Jr and G. C. Sinke, The Chemical Thermodynamics of Organic Compounds (Wiley, New York, 1969) 11 CRC Handbook of Chemistry and Physics (CRC, Cleveland, 64th edn, 1983).12 T. Holm, J. Organornet. Chem., 1973, 56, 87. 13 M. Szwarc and D. Williams, Proc. R. SOC. (London), Ser. A, 1953, 219, 353. 14 D. F. McMillan and D. M. Golden, Annu. Rev. Phys. Chem., 1982, 33, 493. 15 S. W. Benson and H. E. O’Neal, Kinetic Data on Gas Phase Unimolecular Reactions (NSRDS-NBS 16 D. G. L. James and R. D. Stuart, Trans. Faraday Soc., 1968, 64, 2752. 17 A. S. Rodgers, D. M. Golden and S. W. Benson, J. Am. Chem. Soc., 1967, 89, 4578. 18 G. N. Robinson, R. E. Continetti and Y. T. Lee, to be published. 19 ( a ) R. H. Page, Ph. D. Thesis (University of California, Berkeley, Berkeley, California, 1987); ( b ) K. V. Reddy, D. F. Heller and M. J. Berry, J. Chem. Phys., 76, 2814; ( c ) C. S. Parmenter and B. M. Stone, J. Chem. Phys., 1986, 84, 4710. 20 RRKM algorithm of W. L. Hase and D. L. Bunker, Quantum Chemistry Program Exchange, University of Indiana, Bloomington, Indiana. 21 ( a ) L. M. Sverdlov, M. A. Kovner and E. P. Krainov, Vibrational Spectra of Polyatomic Molecules (Wiley, New York, 1974); ( b ) A. Amano, 0. Horie and N. H. Hanh, Int. J. Chem. Kine?., 1976,8, 321. 22 S. A. Safron, N. D. Weinstein, D. R. Herschbach and J. C. Tully, Chem. Phys. Lett., 1972, 12, 564. 23 J. A. Kerr and M. J. Parsonage, Evaluated Kinetic Data on Gas Phase Addition Reactions (CRC, 24 D. B. Moss, C. S. Parmenter and G. E. Ewing, J. Chem. Phys., 1987, 86, 51. 25 K. Okuyama, N. Mikami and M. Ito, J. Phys. Chem., 1985,89, 5617. 26 W. L. Hase, D. M. Ludlow, R. J. Wolf and T. Schlick, J. Phys. Chem., 1981, 85, 958. 27 See, for example, L. Salem, The Molecular Orbital Theory of Conjugated Systems (W. Benjamin, New 28 ( a ) D. S. Perry, J. C. Polanyi and C. W. Wilson Jr, Chem. Phys., 1974, 3, 317; ( b ) J. C. Polanyi and 29 ( a ) H. Schor, S. Chapman, S. Green and R. N. Zare, J. Chem. Phys., 1978,69, 3790; ( b ) S. Chapman, 30 H. M. Rosenstock, K. Draxl, B. W. Steiner and J. T. Herron, J. Phys. Chem. Re$ Data, 1977,6, suppl. 1. 21, US Dept. of Commerce, Washington, DC, 1970). Cleveland, 1972). York, 1966). N. Sathyamurthy, Chem. Phys., 1978,33, 287. J. Chem. Phys., 1981, 74, 1001. Received 17th June, 1987

ISSN:0301-7249    

DOI:10.1039/DC9878400025

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1987, 84, 39-52 Product State Distributions from the Reaction O(3P) + HBr Kenneth G. McKendrick, David J. Rakestraw and Richard N. Zare* Department of Chemistry, Stanford University, Stanford, California 94305, U.S. A. An experimental investigation has been performed of the hydrogen atom abstraction reaction O(3P) + HBr -+ OH(X 211) + Br. A photolytic method was used to produce O(3P) in the presence of HBr, and laser-induced fluorescence detection employed to obtain the nascent vibrational, rotational and fine-structure state partitioning in the product OH. Distributions were found to be highly internally excited, with a strong vibrational inversion and substantial rotational excitation, extending to the limit of available energy. Non-statistical distributions of the fine-structure states were observed.The possibility of a subsidiary reactive channel producing spin- orbit excited Br(2P,,2) is proposed as an explanation for an apparent anomaly in the OH( u’’ = 1) rotational distribution. The dynamics generally appear consistent with the kinematic constraints imposed by the heavy + light-heavy mass combination. Considerable experimental and theoretical effort has been expended in attempting to understand the roles of kinematic constraints and potential-energy surfaces in governing reactivity and the partitioning of available energy in the products of chemical reactions. Hydrogen atom abstraction by a heavy atom from the hydride of a second heavy atom (or molecule) constitutes an important example of the class of reactions characterised by a heavy + right-heavy (H + LH‘) mass combination. In those exothermic H + LH’ reactions thought to proceed via direct abstraction, it has generally been found,’-I2 with some exceptions, 13-21 that kinematic constraints dominate over potential surface effects, causing substantial excitation of both rotational and vibrational degrees of freedom of the products. The majority of hydrogen atom abstraction reactions of ground-state atomic oxygen, O(’P), fall into the H + LH’ category, and are furthermore of immense practical import- ance because of their involvement in many combustion and atmospheric processes.Experimental investigations of the dynamics of the reactions of O(’P) with a wide variety of organic molecules have that a substantial proportion of the available energy is released as OH product vibration, but, uncharacteristically, that very little rotational excitation is imparted to OH. These results have been satisfactorily reproduced by quasiclassical trajectory ( QCT)20 and vibrationally adiabatic distorted- wave (VADW)” calculations, on a semiempirical London-Eyring-Polanyi-Sat0 (LEPS) potential-energy surface, constructed2’ to yield computed results consistent with certain limited aspects of the experimental data.No equivalent, detailed dynamical information has hitherto been reported for the reaction of O( ’ P ) with an inorganic hydride, the generally smaller reaction cross-section making such a study technically more challenging. Extensive kinetic investigations of the reactions with the hydrogen halides have been p e r f ~ r m e d , ~ ~ - ~ ~ including reaction with vibrationally excited HCl,25-29 and some information on vibrational branching in the OH product is available from infrared chemiluminescence measurements on 0 + HI,6 e.s.r.studies of O+ HBr,’ and laser-induced fluorescence (LIF) detection of the OH produced in the reaction of O( ’ P ) with vibrationally excited HC1.29 3940 Dynamics of the O(’P) + HBr Reaction Fig. 1. Schematic diagram of the apparatus for counter-propagating ‘pump-probe’ laser photolysis, laser-induced fluorescence experiments. The subject of this paper is an experimental investigation of the vibrational and rotational product state distributions in the exothermic reaction 0 ( 3 P ) + HBr -+ OH(X ’n) + Br, AH; = -61.5 kJ mol-’.An activation barrier of 14 kJ mol-’ has been deduced from kinetic leading to relatively slow reaction at room temperature [k(O+ HBr) = 3.4 x cm3 molecule-’ s-’1. A very strong vibrational inversion was observed in the OH product in the e.s.r. experiments of Spencer and Glass,’ but no rotational information was extractable, given the highly collisional flow-tube conditions. The experimental method of this study involves the generation of O(’P) by 355 nm photolysis of NO2 in the presence of HBr. The nascent OH product is detected in a state-specific fashion by excitation of LIF on the well known A ’C+-X *I3 transition,” at a short delay following the photolysis pulse. The principal question which will be addressed is whether the dynamics of this reaction are similar to those of hydrogen atom abstraction from organic molecules, or, conversely, more closely related to those of the majority of H + LH’ systems.An early collinear quantum-mechanical study of this system3’ must now be considered suspect in the light of subsequent criticism.8 Very recently, Persky and coworkers,’ in one of a series of theoretical papers on the dynamics of O(3P) +hydrogen halide reaction^,'.'^ have reported the results of a quasiclassical trajectory study on a semiempirical LEPS 0 + HBr surface, optimised to produce agreement with the available kinetic data. The predictions of this QCT study are compared with the new experimental results which we have obtained, and the validity of the proposed potential-energy surface discussed.Experimental A schematic diagram of the apparatus is shown in fig. 1, depicting the counter-propa- gating ‘pump-probe’ geometry of the experiment. A steadily flowing mixture of HBr and NO2 was maintained in the stainless-steel reaction chamber, exhausted by a partially throttled diffusion pump. The pressure was monitored by a capacitance manometerK . G. McKendrick, D. J. Rakestraw and R. N. Zare 41 (MKS Baratron 222BA, 0- 10 Torr,? absolute). The photolysis (‘pump’) beam consisted of 355nm light produced by third harmonic generation of the output of a Nd:YAG laser (Quantel 581), brought to a loose focus (spot-size ca. 2 mm) at the centre of the chamber. Pulse energies were typically 20-mJ. Laser-induced fluorescence was excited by tunable ultraviolet (‘probe’) radiation in the 280-360 nm region produced by a Nd:YAG (Quanta Ray DCR-2A) pumped dye laser (Quantel TDL-50) combination, with a spectral bandwidth of ca.0.2 cm-’. Both pump and probe lasers were operated at a repetition frequency of 20 Hz. The probe beam spot-size (5 mm) was arranged to be sufficiently greater than that of the pump beam to ensure the detection of all species produced at the short time delays between laser pulses employed in the experiments. Great care was taken to account fully for the effects of radiative saturation of the OH A-X t r a n ~ i t i o n : ~ ~ probe pulse energies were reduced by a combination of Glan-Taylor prisms (which allow variable attenuation while maintaining a fixed polarisation). ’These energies were monitored on a shot-to-shot basis via a partial reflection from a quartz plate, at near normal incidence, onto a power meter (Molectron 53-05).The laser beams propagated through internally baffled entrance and exit arms ( 1 m long), and all internal surfaces were coated with a matt black paint (Zuel Corporation, St Paul, MN), greatly reducing the level of scattered laser light. Fluorescence was collected in the vertical direction, perpendicular to the common axis of the laser beams, by a fused silica lens system (f= 5 cm), and imaged through an interference filter (chosen to transmit the desired vibronic band of the OH A-X system) onto the photocathode of a photomultiplier tube (Centronic 4283/8 1). Signals were captured during a gate of length 1.5 ps (corresponding to approximately twice the radiative lifetime of OH A *Z+), delayed by ca.20 ns from the probe pulse to discriminate against scattered laser light. The digitised data (LeCroy 2249SG ADC, CAMAC modular data bus) were passed to a microcomputer (IBM PC-XT) for storage and analysis. Gases used had the following stated purities: HBr (Matheson, >99.8%), HC1 (Matheson, >99.0%) and NO2 (Matheson, >99.5%). The NO2 reservoir was maintained at 0°C to ensure a stable backing pressure. Results Preliminary measurements were performed to identify conditions under which an authen- tic nascent product state distribution would be observed. The majority of these initial investigations probed the OH( v’’ = 1 ) rotational distribution, the signals being most intense for this vibrational level.Relaxation of this distribution could readily be observed as the product of the pressure, P, in the chamber and the delay time, At, between the photolysis and probe laser pulses was increased. As has previously been observed,33 the lowest rotational states were most rapidly relaxed, with the higher levels being relatively metastable. I t was found that the measured OH( v”= 1, J ” ) distribution was essentially invariant for values of PAt < 4 x Torr s. Data were generally collected at PAt = 2 x lo-* Torr s. Furthermore, it was also found that the distribution was insensi- tive to the total pressure, for a given value of PAt, in the pressure range 0-80 mTorr, indicating the absence of substantial variation in the fluorescence quenching rates of different rotational states in OH(A 2Z’) for the present colliding species (generally present in the ratio 1 : 1, HBr: NOz).(Such effects have previously been observed for certain quencher molecule^.'^) The OH A-X diagonal vibronic transitions may readily be radiatively saturated with the probe pulse energies available from the present dye-laser system. To determine the t 1 Torr = 101 325/760 Pa.42 Dynamics of the 0(3 P ) + HBr Reaction 11 12 13 14 I I I Rl 1 1 I S + 6 4 J d L 3 2 1 10 1 1 12 I I ’ Rz I 1 I 9 0 7 I 0 I I I 1 I I I 1 312.2 312.4 312.6 312.8 h, wavelength/ nm Fig. 2. LIF excitation spectrum of the OH A-X (1, 1) band from the reaction 0 + HBr. Fluores- cence was selectively detected on the (1,O) band. Conditions: 25 mTorr HBr, 25 mTorr NO2 ; pump pulse-to-probe pulse delay of 500 ns; probe pulse energy of 150 pJ.regime where the LIF signal would be linearly proportional to the probe pulse energy, measurements were made of the ratios of intensities of main branch lines to those of corresponding satellite branches (e.g. P2 and+P,,, or Q1 and Q21). These pairs of lines probe the same lower state, but with substantially different transition probabilities. From a knowledge of these transition probabilities, which have been accurately c a l c ~ l a t e d , ~ ~ the extent of any radiative saturation could be estimated.32 It was found that for excitation on the A *Z+, v’= 1-X 211, v ” = 1 band [henceforth denoted (v’, d’) = ( 1 , l)], pulse energies (10 pJ produced essentially linear res onse, in good agreement with expectations from the absolute transition probability3’ and the spatial and temporal characteristics of the probe beam.For the significantly weaker off-diagonal (0,l) and (1,2) transitions, pulse energies of <SO0 pJ were correspondingly found to be adequate for linear response. Measured Distributions OH(u”= 1,2) The OH( v” = 1 ) distribution was obtained by recording laser-excitation spectra (fluores- cence intensity as a function of probe laser wavelength) excited on the (1, 1 ) transition. Fluorescence was observed only on the ( 1 , O ) transition (through an interference filter centred at 280nm), eliminating any possible congestion of the spectrum from the excitation of (0,O) transitions lying in the same wavelength region, and reducing the level of observed scattered laser light.A representative spectrum, containing the R-branch region of the ( 1 , 1 ) band, is presented in fig. 2. Substantial rotational excitation is readily discerned, with well formed R,- and R2-branch bandheads. The maximum intensities occur in the returning, higher rotational lines. The rotational substate populations were obtained from analysis of this and other similar spectra, many of which were obtained at reduced probe pulseK . G. McKendrick, D. J. Rakestraw and R. N. Zare 43 1 0.5 0 1 2 3 4 5 6 7 8 9 10 11 12 N 13 14 15 16 Fig. 3. Relative rotational state populations in 2r13,2 (solid bars) and 2111,2 (hatched bars) fine structure components of OH(v”= 1) formed in the reaction O+ HBr. energies (see above), including data for all of the six main spectroscopic branches.The analysis incorporated appropriate corrections for variations in the probe pulse energy, account of the wavelength dependence of the interference filter transmission, and application of the small, well known polarisation corrections required when exciting LIF with a linearly polarised laser.36 (Derived populations were not sensitive to the relative directions of linear polarisation of the pump and probe lasers.) Fig. 3 shows the resulting distributions of the 2133,2 and 2131,2 states as a function of N” (the total rotational angular momentum quantum number excluding electron spin). The sharply peaked, highly excited character of the distribution is apparent. Similarly, the OH(X ‘II, u”= 2) rotational substate distribution was obtained by pumping the off-diagonal (1,2) transition.In this case, two possible detection schemes were exploited, using interference filters to detect selectively fluorescence on either the (1,O) or ( 1 , l ) bands. Signals on the ( 1 , O ) band were weaker than those on ( 1 , l ) , the radiative transition probability being smaller,35 compounded by a (purely instrumental) lower detection sensitivity. However, the spectrum is simplified by being uncongested by overlapping lines from other bands. Conversely, the detection scheme involving ( 1 , 2) excitation followed by ( 1 , 1 ) fluorescence detection results in simultaneous observa- tion of LIF excited from u”= 1 on the (0, 1 ) band, and emitted on the (0,O) band. The alternative approaches thus provide complementary information; a more complete distribution is obtained from detection of ( 1 , O ) fluorescence, but simultaneous observa- tion of fluorescence on the diagonal bands allows the vibrational branching ratio v” = 1/ U” = 2 to be determined, in addition to providing partial confirmation of the U” = 2 rotational distribution. Fig.4 shows a representative laser-excitation spectrum of the 348-350 nm region, with fluorescence observed on the diagonal bands. Lines originating from u”= 1 and v”=2 are indicated in the figure. Conditions were once again such that a nascent distribution is believed to have been observed. From this spectrum and similar data, including observations of the ( 1 , O ) fluorescence, the rotational substate populations in u”= 2 were derived (see fig.5). Corrections similar to those described for the u”= 1 data were applied. It can be seen that the U” = 2 rotational distribution is also substantially excited, but less so than that of v”= 1.44 1 7 0.5- 0- Dynamics of the O( P ) + HBr Reaction 100. 80. 60. 40- 20- 0- I I , 1 Q l I I 9 I I 1 Q2 5 6 7 b 13 14 15 16 17 12 13 14 15 16 17 I I I I I I 3 4 5 6 lo 9 a . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . J . . . . . . . . . . . . . . I I 34 9 350 wavelength/nm Fig. 4. LIF excitation spectrum of the OH A-X (0, 1) and (1,2) bands from the reaction O+ HBr. Fluorescence was observed on the (0,O) and (1, 1) diagonal bands. Transitions originating from u ” = 2 are explicitly indicated by dotted lines.Experimental conditions were: 50 mTorr HBr, 50 mTorr NO,; pump pulse-to-probe pulse delay of 200 ns; probe pulse energy of 700 pJ. I 4 N” - %/2 Fig. 5. Relative rotational state populations in the *n3,, fine structure component of OH( u” = 2) produced in the reaction O(3P) + HBr.K . G. McKendrick, D. J. Rakestraw and R. N. Zare 45 OH( d’ = 0) Attempts were also made to observe any possible production of OH( u” = 0) from the reaction 0 + HBr. Significant u” = 0 LIF signals were obtained by pumping and detecting on the (0,O) transition. However, these signals were found to be entirely ‘prompt’, with no kinetic accumulation (of the order of a few ps) of the OH population following the photolysis pulse, as had been observed for the higher vibrational states.Examination of the u”= 0 rotational state distribution revealed it to be identical to that reported” for OH produced in the photolysis of HONO at 355 nm. The magnitude of this signal could be substantially reduced by increasing the flow rate of gases through the reaction chamber while maintaining a constant pressure, once gain in contrast to the u” = 1 and 2 data, which were independent of the flow rate (except, as might be expected, at very low flows). It was concluded that OH( u” = 0) was arising through heterogeneous produc- tion and subsequent photolysis of HONO, Although previous reports3’ have indicated the absence of any significant production of higher vibrational states of OH in the 355 nm photolysis of HONO, a control experiment was performed to provide confirmation that the distributions ascribed to products of the O+ HBr reaction were not distorted by a contribution from HONO photolysis. Photolysis of a flowing mixture of HCl and NOz was found to generate OH( u” = 0) with a rotational state distribution identical to that from photolysis of HBr-NO,.The relative magnitudes of the OH( u” = 0) signals were taken as a measure of the HONO concentration, allowing an assessment to be made of the contribution of HONO photolysis to the observed OH( v”= 1 ) signal from HBr-NO,. No OH( u”= 1 ) was observable, above the residual noise level, following photolysis of HC1-NO2 mix- tures, constraining this estimate to be an upper limit. It was consequently concluded that HONO photolysis could not be responsible for a contribution of >5% to the population of the lowest rotational states of OH( u” = l ) , where such an effect would be expected to be most significant. Vibrational Branching Ratio As discussed above, measurements such as those displayed in fig.4 allow an estimate to be made of the vibrational branching ratio u”= l/u”= 2. Total populations in each of the vibrational states were calculated by summing over all rotational substates (each of the A-doublet components of each of the spin-orbit states). Those state populations not measured because of spectroscopic congestion were estimated by interpolation of the available data. Einstein coefficients used in obtaining the branching ratio were the experimentally determined values of Copeland et aZ.38 The statistical accuracy of the population measurements was relatively high, ca.lo%, and the major source of uncer- tainty in the branching ratio therefore arose from the quoted uncertain tie^^^ in the Einstein coefficients. The derived vibrational branching ratio was OH( u” = 1)/OH( u” = 2) = 9.4k3.1 at the 95% confidence level, assuming no other sources of systematic error. The significant background OH( u” = 0) population resulting from HONO photolysis prevented any rigorous assessment of the branching into u”= 0 of the OH product from 0 + HBr. However, the observed rotational distributions for u” = 1 and 2 exhibit a strong propensity for population of states which are approximately thermoneutral with respect to the reagents (as will be discussed further below). It therefore seems a reasonable assumption, as is supported by the results of the quasiclassical trajectory study of Persky and co-worker~,~ that any OH( u” = 0) produced in the reaction would occupy relatively high rotational states, peaking in the range N” = 14-18.The OH( u” = 0) rotational distribution from 355 nm photolysis of HONO is relatively ‘cold’, with very little population in these higher levels. We observed no significant secondary maximum in this distribution, and estimate, on the basis of the above assumption, that u”= 0 con- stitutes <lo% of the total population from the reaction O+ HBr. This observation is46 Dynamics of the 0(3 P ) + HBr Reaction consistent with the results of the previous e.s.r. study of Spencer and Glass,’ where >97% of the OH produced was estimated to have been formed in u”= 1.Fine Structure and A-Doublet Population Ratios The OH 211 ground state is subject to the well known spin-orbit and A-doublet splittings, resulting in four spin substates for each quantum number, N”. The lower and upper spin-orbit states are labelled ’11312 and ’II,,,, respectively. These levels are split by ca. 125 cm-’ at the lowest rotational state, but this splitting decreases quite rapidly with increasing N” as the character of the angular momentum coupling changes from Hund’s case ( a ) towards case (6). There has been considerable confusion and contradiction in the use of 7 ~ + and 7 ~ - as labels for A-doublet states. An exposition of the directional nature of the electronic charge distribution of A-doublets was recently presented by Andresen and R ~ t h e , ~ ’ but unfortunately did not resolve the residual contradictions of nomenclature in many prior and, indeed, subsequent papers. We have therefore chosen to adopt a notation whereby T( 1l.l) denotes states which, in the limit of large rotational angular momentum, J, have the unpaired p.rr electron lobe parallel to J ; correspondingly, m ( 1 .l ) refers to those states with the p.n lobe perpendicular to J, and hence in the plane of rotation of the molecule. For a E-II transition, such as in the present case for OH, ~(1I.l) states are probed by Q-branch lines, and ~(1.l) states by P- and R-branch lines. ~(1.l) populations of OH(u”= 1) presented in fig. 3, a preference of ca. 30% for the lower spin-orbit state was observed.A similar preference was found in the u” = 2 data. The spin-orbit ratio is obtained from data taken on branches with very similar line strengths and equivalent polarisation properties, and seems unlikely to be subject to systematic errors of this magnitude. The A-doublet ratio was less well defined. Near the peak of the u” = 1 distribution, with N” in the range 10-12, a reproducible, weak preference was found, with a maximum discrepancy of ca. 25% in favour of the ~(1.l) component at N”= 10. The ratio was found to vary with N”, being insignificantly different from unity for N”> 13 and N”< 7. (It is required that the states be equally populated in the limit of no nuclear rotation, since there can be no degree of electron alignment.) We hesitate to attach too great a significance to the observed weak preference, it being close to the level of systematic uncertainty in comparing data taken on branches with substantially different line strengths and polarisation properties.The u” = 2 data similarly showed little preference for either A-doublet component. As can be seen from the ’II3,2 r(1.l) and Discussion A diagram showing reagent and product state energetics in the O+HBr system is presented in fig. 6. Fine structure states in O(3P) and OH(X ’II) are not indicated: only the lowest-energy O(3P,) and OH(X 2113/2) states are shown. Channels leading to ground, 2P3/2, and excited, spin-orbit states of the bromine atom, separated by 3685 cm-’, are denoted by Br and Br*? respectively. In the photolysis of NO2, it is to be expected that a non-monoenergetic distribution of 0 atom velocities will be obtained, corresponding to a distribution over the internal product states of NO(X ’II).Two related studies of NO2 photodissociation,4”4’ at wavelengths close to 355 nm, are apparently in contradiction over the details of the energy relea~e,~’ but both demonstrate a relatively broad distribution of 0 atom velocities, extending to the upper limit imposed by total energy conservation. The energy available to be partitioned amongst the degrees of freedom of the products, EO,NO, may be expressed asEll---- . - O('P) + HBr K . G. McKendrick, D. J. Rakestraw and R. N. Zare OH(v"= 0) + Br OH(v"=l)+ Br" 1 - OH(v"=2)+Br - - - - 10 - - 10 - = 1 - OH(v"=O)+ Br OH(v"= 1) t Br - - 10 -5 =1 - - 47 Fig.6. Energetic diagram of reagent and product states in the O(3P)+ HBr + OH(X 'II) + Br system. 2P3/z and 2P1/2 bromine atom spin-orbit states are denoted by Br and Br", respectively. E, is the kinetically determined activation barrier. where E ( hv) = 28 190 cm-' for third harmonic Nd:YAG radiation, DG( NO,) = 25 132 ~ m - ' , ~ ' and Etherm( NO,) is the thermal translational and internal energy of NOz. Taking the limit where, in the photolysis, all the available energy appears as product translation, and assuming Etherm(N02) = : kT (translation) +$ kT (rotation) = 620 cm-', conservation of linear momentum constrains the translational energy of the 0 atom, Eo, to be Eo=i mou;= mNoEo,No/(mo+ mNo)=2395 cm-', (2) where mi and ui are the mass and velocity, respectively, of species i.Note that the production of electronically excited O(' 0) is energetically impossible for single-photon dissociation of NO2 at 355 nm. The absence of multiple-photon processes producing O ( ' 0 ) under the present conditions was verified by a failure to observe any OH when photolysing NO, in the presence of H2,42 or characteristically excited OH4' in the presence of organic species [which produced only the rotationally 'cold' distributions previously r e p ~ r t e d ' ~ - ' ~ for such systems in reactions with O(3P)]. In the centre-of-mass frame of the 0 + HBr collision, the collision energy, Ecoll, may be approximated by EcoIi = i P ( u i + 3 k T / m HBr) (3) where p is the reduced mass of O+ HBr and the second term in parentheses represents the average contribution from thermal motion of HBr.44 (In the present case, this term is almost negligible compared to u i . ) Using the limiting value for uo derived above, a value of 2050 cm-' is obtained for Ecoll.In addition to translational energy of the collision partners, the rotational energy of HBr is available to be partitioned between modes of the products. The QCT study of Persky and co-workers' suggests that reactivity is substantially promoted by HBr rotation, and the average rotational energy of reactive HBr molecules is correspondingly sig- nificantly higher than that of the Boltzmann ensemble of reagents. Taking the value of48 Dynamics of the 0(3 P ) + HBr Reaction 500 cm-’ predicted by the QCT calculations for the average rotational energy of reactive HBr at 300 K, and a reaction exothermicity of 5150 cm-’, an approximate limit to the total energy available to the products, relative to the lowest states of OH and Br, is deduced to be 7700cm-’. Consideration of the measured OH product distributions in the light of fig.6 and the discussion above clearly reveals that the majority of the energy available to the products appears as internal excitation of OH. The rotational distributions in Y” = 1 and 2 extend to N ” = 16 and 9, respectively, some 8300 and 8500 cm-’ above the lowest level of v’’ = 0. These energies exceed by several hundred cm-’ the approximate energetic limit derived above, suggesting a significant contribution to reaction from states in the high-energy Boltzmann tail of the HBr rotational distribution.The peaks in the Y” = 1 and 2 distributions at N”= 12 and N ” - 5 , respectively, correspond to energies of 6300 and 7500 cm-’. The O(3P) + HBr reaction therefore exhibits dynamical behaviour quite distinct from that of the previously studied reactions of O(3P) with organic species, in which low levels of rotational excitation of the products were ~ b s e r v e d . ’ ~ - ’ ~ In contrast, 0 + HBr displays properties characteristic of many H + LH’ systems, I-’’ in which the approximate propensity rule (where E,,,,, is the kinetic energy), is obeyed in the selection of dominant product channels. This effect is essentially kinematic in origin. Relatedly, it has also been proposed4 that further propensity rules: L( reactants) = L( products) J ( reactant diatomic) =: J ( product diatomic) ( 5 ) (6) (where L and J refer to orbital and rotational angular momentum, respective1y)would be obeyed if the reaction were uninfluenced by a controlling potential energy surface.Eqn ( 5 ) bears a close equivalence to eqn (4) in a direct H + LH’ system. It has, however, been found’,’2 that the conservation of rotational angular momentum expressed in eqn (6) is a relatively poor prediction, it being argued” that the product J is particularly sensitive to energy release controlled by the potential surface. The present experimental results do not allow a direct test of prediction (6), since the rotational distribution of reactive HBr molecules is unknown. However, only 0.2% of a 300 K sample of HBr occupies J = 12, the rotational angular momentum at the peak of the product OH( Y” = 1 ) distribution, suggesting that a very dramatic increase in the reaction cross-section with HBr rotation would be required if propensity rule (6) were to be valid.The above kinematic effects may alternatively be described using the vocabulary of potential-energy surfaces. In a mass-scaled coordinate system’ the H + LH’ mass combi- nation exhibits a very narrow skewing angle between the entrance and exit valleys (this angle is 15.4” for 0 + HBr). The channelling of energy into the internal states of the products is then interpreted2 as being the result of ‘corner-cutting’ trajectories, in which the much more rapid motion of the light compared to the heavy particles results in transfer of the light atom at relatively long range, before the equilibrium H---L internu- clear separation is attained in the entrance channel.As mentioned above, a quasiclassical trajectory study has recently been performed’ on a LEPS surface optimised to reproduce kinetic data for the 0 + HBr system. The surface derived has the barrier (of 13.2 kJ mol-’) in the entrance channel, but is pre- dominantly repulsive (the majority of the reaction exothermicity is released in the exit channel). The results of that study are not rigorously quantitatively comparable with those of the present experiments, since the calculations were performed for thermalised ensembles of reagents at various temperatures (200, 300 and 550 K). However, certainK . G. McKendrick, D. J. Rakestraw and R.N. Zare 0.5 0.1 0.05 49 0 8 0 O 0. 0 -8 0 . O 8 8 - 0 8 0 - d. I , a t I , , , I , E .- c) - 1 a a Fig. 7. Boltzmann plot of rotational state populations of OH( u” = 1 and 2) produced in the reaction O+ HBr. 0, v”= 1, 2113,2 spin-orbit component; 0, uf’= 1, ’nl,* spin-orbit component; a, uff = 2, 2113/2 spin-orbit component. enlightening comparisons are possible. The QCT studies correctly predict substantial vibrational and rotational excitation of OH. The predicted strong population inversion of v”= 1 over v”= 0, with 93% of OH in v”= 1 for 300 K reagents, is highly consistent with the failure to observe nascent OH( v”= 0) in this study and in the previous e.s.r. experiments of Spencer and Glasss The average vibrational energy of the products (excluding zero-point energy) was ca.40 kJ mol-’ (almost independent of reagent tem- perature) in the QCT study, compared to 47 kJ mol-’ in our experiments. Note, however, that no branching into u”= 2 was predicted, in disagreement with the experimental observations. Experimentally, the average rotational energy (of u”= 1 and 2) was 22 kJ mol-’, falling between the predicted values of 18 and 23 kJ mol-’ for reaction at temperatures of 300 and 550 K, respectively. It was also predicted that although the reaction cross-section does increase quite steeply with HBr rotation, the average rota- tional energy of the product OH is about three times greater than that of the reacting HBr molecules, consistent with the suggestion above that propensity rule (6) does not appear to be well obeyed in this system.No detailed product rotational state distributions were reported by Persky and coworkers. An analysis of our experimental data was performed according to the ‘surprisal’ f ~ r m a l i s m , ~ ~ in which measured distributions, Pexptl( v”, N ” ) , are compared with predic- ted prior distributions, P0(u”, N”), calculated solely on the basis of the density of available product states (neglecting angular momentum constraints). Plots of the sur- prisal function, I ( u ” = 1, N”) = -In [Pexpt,(u”= 1, N”)/P,(u”= 1, N”)] (8) were found to be highly non-linear. Such behaviour has been found elsewhere12 in a computational study of a similar H + LH’ system, C1+ HC1. Interestingly, it was found that I(u”= 1, N ” ) for OH exhibited two extrema, in contrast to the single extremum of the Cl+ HCI study.” This effect can also be discerned from fig.7, where our experimental rotational distributions have been presented as a semilogarithmic plot of populations divided by the rotational degeneracy, (2J”+ l ) ,50 Dynamics of the 0(3 P ) + HBr Reaction against the total internal energy. (This Boltzmann plot conveys information which is more limited but related to that of a surprisal plot, particularly at low internal energy where the density of available translational states is slowly varying, but has the advantage that the specification of an arbitrarily fixed total energy is avoided. The form of the surprisal of the more highly excited states is sensitive to this parameter, which is not well defined in our system.) The broad peak in fig.7 at ca. 6000cm-’ is characteristic of the type of behaviour also observed for Cl+ HC1.’’ It is the subsidiary maximum at the lowest rotational state of OH( u” = 1 ) which we consider anomalous. [Note that this is not the result of partial relaxation of the measured distribution. Careful investigations at reduced values of PA? (see Experimental section, above) verified that the observed distribution was collisionally unmodified. Further strong support for this assertion is that the OH(v”=2) data show no evidence for anomalous population of the lower rotational states.] We suggest that this observation may be associated with the opening of an additional channel to form bromine atoms in the upper spin-orbit state, 2P1/2. As fig. 6 shows, the OH( v” = 1 ) + Br* channel lies slightly (some 280 cm-’) above OH( v” = 2) + Br, and hence is energetically accessible for OH( u’’ = 1, N ” s 7).This suggestion seems consistent with the character of the inflection in the v ” = 1 data of fig. 7. Although we have not attempted to obtain any direct experimental evidence for the occurrence of a minor channel producing Br”, it is, in principle, an experimentally verifiable conjecture. It can also be seen from fig. 7 that, in the vicinity of the higher N” peak of the distribution, the 2113/2 spin-orbit state of OH( u” = 2 ) is preferentially populated over 2111/2, even when account is taken of the degeneracy differences (which are significant at low N ” ) and the data are plotted as a function of energy, rather than rotational quantum number.(No residual preference is apparent in the low N” data.) A similar effect to that for the higher N” levels was previously reported’” for the relatively unexcited distributions from reaction of O(3P) with organic systems. The argument presented by Andresen and LuntzI3 to explain this observation was that the reaction was partially electronically adiabatic on surfaces correlating spin-orbit states of O( P ) with those of the OH(X ’II) product. The predominance of the lower, *n3/2, state in OH was then essentially a consequence of the higher population of the lowest, 3P2 , state of the reagent oxygen atoms, in thermal equilibrium from a microwave discharge source. Unfortu- nately, the distribution over the spin-orbit states of O(’PP) produced by 355 nm photolysis of NOz is unknown, but it is rather intriguing, although possibly quite coincidental, that the OH spin-orbit ratio from the 0 + HBr reaction under the present conditions is quite similar to that of the previous studies of organic systems.The absence of any strong preference for either of the A-doublet substates is consistent with a direct, collinear mechanism of reaction. The possible slight discrepancy in favour of the n ( I J ) state, near the peak of the OH( v”= 1 ) distribution, noted above, would be compatible with a tendency for these states to be produced via a bent reactive geometry, with an incipient bonding interaction between the 0 and Br atoms aligning the unpaired electron lobe on the 0 atom in the 0-H-Br plane. Dynamical effects of this type, many of much greater magnitude, are well known in a number of other reactive systems.39 In the discussion of the dynamics above, we have implicitly discounted the involve- ment of the deep potential well on the singlet surface which corresponds to the bound intermediate, HOBr.A spin-forbidden insertion mechanism, whereby an electronically non-adiabatic curve-crossing from the initial triplet surface would allow access to the singlet intermediate, seems a less satisfactory explanation for the observed results, given the very significant OH vibrational inversion and the highly non-linear rotational surprisals. These departures from statistical behaviour are more extreme than in previous systems where a short-lived intermediate complex is thought to be involved [ e.g.O( ID) + HC146 and O(’ D ) + H2”].K . G. McKendrick, D. J. Rakestraw and R. N. Zare 51 Conclusion The 0 + HBr reaction can be well understood in terms of a direct abstraction mechanism for a H + LH’ system dominated by kinematic constraints. The highly inverted vibrational distribution results from an early barrier in the entrance channel, and ‘corner-cutting’ trajectories in which the H atom is transferred at relatively large internuclear distances. Substantial rotational excitation derives from repulsive interactions in the exit channel. Whilst the trajectory studies of Persky and co-worker~~ are capable of at least good qualitative prediction of the observed product distributions using a LEPS surface with a collinear minimum-energy pathway, it seems clear that high rotational excitation of the products must result from repulsive Br---H interactions acting off the line of centres in a bent 0-H-Br geometry.As has been discussed elsewhere,” the H + LH‘ mass combination channels this repulsive interaction very effectively into HL rotation, since the force acts on the L particle, with a long lever arm about the HL centre-of-mass. This suggests a possible explanation for the contrasting dynamics of 0 +organic systems compared to those of O+HBr, namely, that the O+organic systems may be subject to a stronger constraint towards collinearity, leading to much-reduced rotational excitation of the OH product. We express our special thanks to two groups of workers; R. A. Copeland, J.B. Jeffries and D.R. Crosley, and M. Trolier and J. R. Wiesenfeld, for making available unpublished OH transition probabilities. (K.G.McK.) is grateful to the S.E.R.C. for the provision of a Postdoctoral Research Fellowship. This work was supported by the U.S. National Science Foundation under grant number NSF CHE 84-07270. References 1 C. A. Parr, J. C. Polanyi and W. H. Wong, J. Chem. Phys., 1973, 58, 5 . 2 J. C. Polanyi and J. L. Schreiber, in Physical Chemistry, an Advanced Treatise, ed. W . Jost (Academic Press, New York, 1974), vol. 6A, chap. 6, pp. 383-487. 3 M. Baer, J. Chem. Phys., 1975, 62, 305. 4 K. Schulten and R. G. Gordon, J. Chem. Phys., 1976, 64, 2918. 5 J. E. Spencer and G. P. Glass, Int. J. Chem. Kinet., 1977, 11, 97. 6 B. S. Agrawalla, A. S. Manocha and D.W. Setser, J. Phys. Chem., 1981, 85, 2873. 7 A. Persky and M. Broida, J. Chem. Phys., 1984, 81, 4352. 8 P. L. Gertitschke, J. Manz, J. Romelt and H. H. R. Schorr, J. Chem. Phys., 1985, 83, 208. 9 M. Broida, M. Tamir and A. Persky, Chem. Phys., 1986, 110, 83. 10 A. Persky and H. Kornweitz, Chem. Phys. Lett., 1986, 127, 609. 1 1 P. M. Aker, D. J. Donaldson and J. J. Sloan, J. Phys. Chem., 1986, 90, 3110. 12 G. C. Schatz, B. Amaee and J. N. L. Connor, Chem. Phys. Lett., 1986, 132, 1 . 13 P. Andresen and A. C. Luntz, J. Chem. Phys., 1980, 72, 5842. 14 K. Kleinermanns and A. C. Luntz, J. Chem. Phys., 1982, 77, 3533. 15 K. Kleinermanns and A. C. Luntz, J. Chem. Phvs., 1982, 77, 3537. 16 K. Kleinermanns and A. C. Luntz, J. Chem. Phys., 1982, 77, 3774. 17 N. J. Dutton, I.W. Fletcher and J. C. Whitehead, Mol. Phys., 1984, 52, 475. 18 N. J. Dutton, I. W. Fletcher and J. C. Whitehead, J. Phys. Chem., 1985, 89, 569. 19 N. J. Barry, I. W. Fletcher and J. C. Whitehead, J. Phys. Chem., 1986, 90, 4911. 20 A. C. Luntz and P. Andresen, J. Chem. Phys., 1980, 72, 5851. 21 D. C. Clary, J. N. L. Connor and W. J. E. Southall, J. Chem. PhyJ., 1986, 84, 2620. 22 R. D. H. Brown and I. W. M. Smith, Znt. J. Chem. Kinet., 1975, 7, 301. 23 D. J. Singleton and R. J. Cvetanovic, Can. J. Chem., 1978, 56, 2934. 24 D. F. Nava, S. R. Bosco and L. J. Stief, J. Chem. Phys., 1983, 78, 2443. 25 D. Arnoldi and J. Wolfrum, Chem. Phys. Lett., 1974, 24, 234. 26 R. D. H. Brown, G. P. Glass and I. W. M. Smith, Chem. Phys. Lett., 1975, 32, 517. 27 Z. Karny, B. Katz and A. Szoke, Chem. Phys. Lett., 1975,35, 100. 28 R. G. Macdonald and C. B. Moore, J. Chem. Phys., 1978, 68, 513. 29 J. E. Butler, J. W. Hudgens, M. C. Lin and G. K. Smith, Chem. Phys. Lett., 1978, 58, 216. 30 G. H. Dieke and H. M. Crosswhite, J. Quant. Spectrosc. Radial. Transfer, 1962, 2, 97. 31 E. J. Shipsey, J. Chem. Phys., 1973, 58, 232.52 Dynamics of the O('P) + HBr Reaction 32 R. Altkorn and R. N. Zare, Annu. Rev. Phys. Chem., 1984, 35, 265. 33 J. T. Yardley, in Introduction to Molecular Energy Transfer (Academic Press, New York, 1980), chap. 34 R. A. Copeland, M. J. Dyer and D. R. Crosley, J. Chem. Phys., 1985, 82, 4022. 35 Rotational linestrengths were taken from: I. L. Chidsey and D. R. Crosley, J. Quant. Spectrosc. Radiat. Transfer, 1980, 23, 187 and W. L. Dimpfl and J. L. Kinsey, J. Quant. Spectrosc. Radiat. Transfer, 1979, 21, 233, except for off-diagonal bands for which values had not been reported, where the unpublished values of M. Troiler and J. R. Wiesenfeld were used. 10. 36 C. H. Greene and R. N. Zare, J. Chem. fhys., 1983, 78, 6741. 37 R. Vasudev, R. N. Zare and R. N. Dixon, J. Chem. Phys., 1984, 80, 4863. 38 R. A. Copeland, J. B. Jeffries and D. R. Crosley, Chem. Phys. Lett., in press. 39 P. Andresen and E. W. Rothe, J. Chem. Phys., 1985, 82, 3634. 40 G. E. Busch and K. R. Wilson, J. Chem. Phys., 1972, 56, 3626. 41 H. Zacharias, M. Geilhaupt, K. Meier and K. H. Welge, J. Chem. Phys., 1981, 74, 218. 42 C. B. Cleveland, G. M. Jursich, M. Trolier and J. R. Wiesenfeld, J. Chem. fhys., 1987, 86, 3253. 43 A. C. Luntz, J. Chem. Phys., 1980, 73, 1143. 44 E. E. Marinero, C. T. Rettner and R. N. Zare, J. Chem. Phys., 1984, 80, 4142. 45 R. D. Levine and R. B. Bernstein, Acc. Chem. Res., 1974, 7, 393. 46 A. C. Luntz, J. Chem. Phys., 1980, 73, 5393. Received 7 th May, 1987

ISSN:0301-7249    

DOI:10.1039/DC9878400039

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1987, $4, 53-63 Onset of Migration in the Reaction of Fluorine Atoms with Iodine Molecules Neil C. Firth, Norman W. Keane, David J. Smith and Roger Grice" Chemistry Department, University of Manchester, Manchester M13 9PL Reactive scattering of F atoms with I, molecules has been studied at an initial translational energy E = 40 kJ mol-', using a supersonic beam of F atoms seeded in He buffer gas generated by a high-pressure microwave discharge source. The source gives essentially complete dissociation of F2 precursor, so that the measured laboratory angular and velocity distributions of IF scattering arise only from the F+I, reaction. The centre-of-mass differential reaction cross-section peaks in the backward direction with a higher product translational energy than the forward scattering which shows a broad subsidiary maximum of relative height -0.6.This is in sharp contrast to the stripping dynamics exhibited by the F+Br, reaction under similar experimental conditions, and is attributed to the onset of migration in the F+ I, reaction dynamics. A forward-scattered component with very low product translational energy is attributed to reaction at large impact param- eters of I2 molecules strongly inclined to the plane of collision. This com- ponent corresponds to highly rotationally and vibrationally excited IF molecules observed in the laser-induced fluorescence measurements of Viguii and co-workers. The backward scattering is attributed to large-impact- parameter collisions with the I, molecules lying close to the plane of collision, in which the F atom orbits around the I atom to which it is initially bonded and then migrates to the other I atom by intervening in the extended I2 bond.The onset of migratory dynamics is attributed to the preferred bent geometry of the FI, intermediate and the existence of an attractive well in the exit valley of the potential-energy surface for this L+HH (light+ heavy-heavy) reaction. Reactive scattering measurements on the F+ IC1, CH31 reactions'-3 have shown short- lived complex dynamics with sharp forward and backward peaking. These results are consistent with measurements on the F2 + ICI, CHJ reaction^,^ which demonstrate the existence of stable FICI, FICH3 free-radical intermediates with dissociation energies which are approximately equal to the exoergicities of the F+ICl, CH31 reactions.The existence of the FII free radical has also been demon~trated,~ but in this case the dissociation energy is very much less than the exoergicity of the F+ I, reaction. So far only preliminary angular distribution measurements have been published"' for the F + I, reaction, although a study' of the closely related F+ Br, reaction has shown that stripping dynamics are followed in this case. Bent structures have been proposed for the preferred geometry of the reaction intermediates in both the F+ ICl' and Br, ' reactions. Monte- Carlo trajectory calculations8-'0 on the F+ I2 reaction have indicated that migration may play a significant role in the reaction dynamics, using a potential-energy surface with a shallow well in the exit valley and a collinear preferred geometry.Early laser-induced fluorescence measurements "*'* of the IF product vibrational state distribution were interpreted by Trick1 and Wanner" to indicate that iodine atoms were formed mainly in the spin-orbit excited I('PlI2) state. However, subsequent attempts"-I5 to detect excited I-atom products directly have failed to find the predicted yield. More recent laser-induced fluorescence measurements of the IF product vibra- tional-rotational state distributions for F + I2 I 6 * l 7 show an intriguing bimodal behaviour54 Migration in F + I2 Reaction Dynamics 1 .o 0.6 U .- $ Q ci 0.4 s 0.2 0.0 laboratory angle O/O Fig. 1. Laboratory angular distribution (number density) of IF reactive scattering from F+ I2 at an initial translational energy E = 40 kJ mol-'.(-) The overall fit of the Entemann kinematic analysis, (- - -) the contributions of the separate components. in the rotational state distributions for higher vibrational states. Here reactive scattering measurements using mass-spectrometric detection are reported for the reaction in order to help elucidate the underlying dynamical mechanism. The rate constant k = 2.6 x 10" dm3 mol-' s-l has been measured at 298 K in a discharge flow experiment.18 F+I2 --+ I F + I ( 1 ) Experimental The apparatus was the same as that employed in the F+ICl experiments' using a high-pressure microwave discharge source19 with an alumina discharge tube to give 395% dissociation of F2 molecules in He buffer gas.The F-atom beam velocity distribution was measured by a beam monitor mass spectrometer and peaked at vpk= 1950 m s-l with a full-width at half maximum intensity vwd = 550 m s-', corresponding to a Mach number A4 =: 4. The cross beam issued from a glass nozzle with a vapour pressure ==: 1 Torr* of 12, maintained by a reservoir at ca. 40 "C, seeded in ca. 45 Torr of N2 buffer gas. The I2 velocity distribution peaked at Upk =r 720 m s-' with a full-width at half maximum intensity Vwd = 230 m s-', corresponding to a Mach number =4. The velocity distributions of the I2 beam and the I F reactive scattering were both measured by the rotatable mass spectrometer detector using cross-correlation time-of-flight analy- sis.20 For all reactive scattering measurements, the faces of the beam source enclosures were screened from the detector by a copper cold shield cooled to liquid-nitrogen temperature in order to reduce the flux of I F background at very low laboratory velocity, which arises from reaction of scattered F atoms with I2 molecules absorbed on surfaces adjacent to the scattering centre.Results Angular distribution measurements of IF reactive scattering, made with the chopper disc rotating slowly, yield ca. 100 counts s-l against a background of ca. 50 counts s-'. The laboratory angular distribution of IF number density shown in fig. 1, was obtained *'F 1 Torr = 101 325/760 Pa.N. C. Firth, N. W. Keane, D. J. Smith and R. Grice 55 by repeated measurements with 60 s counting times. Despite the imposition of copper cold shielding over the faces of the beam-source enclosures, the appearance of IF background at very low laboratory velocity in the angular range close to the I2 beam was not entirely eliminated.The data points shown in fig. 1 were obtained when such remaining background contributions were removed by performing integrations over a restricted range of time channels in the observed IF time-of-flight distributions. The laboratory angular distribution thus obtained shows a peak at slightly smaller laboratory scattering angles than the direction of the laboratory centroid and a shoulder at laboratory scattering angles beyond the I2 beam. This contrasts sharply with the angular distribution obtained for the F+Br, reaction,' which shows a sharp peak close to the F-atom beam and very much lower intensity at wide laboratory scattering angles.Laboratory velocity distributions of IF flux, shown in fig. 2, were measured using integration times -1800 s to gain signal-to-noise ratios of ca. 20 at the peaks of the distributions. The numerical inversion procedure of Siska2' has been used to transform the laboratory data to centre-of-mass coordinates, yielding the contour map of the differential reaction cross- section shown in fig. 3. This shows scattering over the full angular range with a sharp peak in the backward direction at centre-of-mass scattering angle 6 = 180" and lower intensity in the forward direction 6 = 0". Kinematic analysis has also been performed using the stochastic method of Entemann," with the centre-of-mass differential cross-section expressed as the product of an angular function T ( 6 ) and a velocity function U(U).However the variation of the I F velocity distribution U ( u ) with scattering angle was so pronounced that it proved necessary to express the differential cross-section as a sum of two components: The resulting angular functions T( 6 ) and product translational energy distributions P ( E ' ) are shown in fig. 4. One component, shown by broken curves, yields scattering into the forward hemisphere with a product translational energy distribution which peaks at low energy with a long tail out to higher energy. The second component, shown by solid curves, yields scattering over the full angular range with a sharp peak in the backward direction 6 = 180" and a product translational energy distribution which peaks at higher energy and has a more symmetrical shape.The forward-scattered component has a lower intensity than the wide-angle component over the full range of centre-of-mass scattering angle in fig. 4. However, the separate contributions of each component to the laboratory angular distribution of fig. 1, shown by broken curves, indicate that the forward component makes an important contribution to the IF number density at laboratory scattering angles slightly less than that of the laboratory centroid 6,, = 80". The prominence of the forward component in this region arises from the kinematic enhancement of the I F laboratory intensity arising from low IF centre-of-mass recoil velocity.22 Integration of the differential cross-section weighted by sin 6 over the full range of centre-of-mass scattering angle and velocity Or= 2 7 ~ 1; ICm( 8, U ) sin 8 d6 du (3) shows that the forward component makes a relative yield of ca.12% compared with a yield of ca. 88% for the wide-angle component in the total reaction cross-section Or. The composite angular distribution and the product translational energy distributions in the forward and backward directions arising from the sum of the two components are shown in fig. 5. The angular distribution shows a peak in the backward direction and a subsidiary maximum of relative height of ca. 0.6 in the forward direction. The product translational energy distribution for scattering in the forward direction shows56 h la U .- c 4 0" w n a v * h U la .- E a X uz n la Y ." c +j v C.a 0" v % h c a X a U la .- Migration in F + I2 Reaction Dynamics 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 , 1 1 1 4 6 8 10 12 14 16 18 20 22 24 laboratory velocity/ 100 m s-l F l I 1 I I 1 I 1 I I I 1 : I I I I 1 I 1 I ----Q _ _ _ _ _ n x. 4 6 8 10 12 14 16 18 20 22 24 laboratory velocity/ 100 m s-l laboratory velocity/ 100 m s-' 1 l , , I ',',.; , I I I ,15 a a---+-----e---. 4 6 8 10 12 14 16 18 20 22 24 laboratory velocity/ 100 m s-l Fig. 2. Laboratory velocity distributions (flux density) of reactively scattered IF from F+ I2 at an initial translational energy E = 40 kJ mol-'. (- - -) The overall fit of the Entemann kinematic analysis. increased intensity at low translational energy compared with the backward scattering. The peak ELk and average ELv product translational energies for each component are listed in table 1 together with the initial translational energy E and the reaction exoergicity ADo calculated from the IF bond energy of Clyne and McDermidZ3 and the I z bond energy of Huber and H e r z b e ~ g .~ ~ The total reaction cross-section Qr= 85 A2 may be estimated at an initial translational energy E = 2.5 kJ mol-l from the room-temperature rate constant of Appelmann and Clyne.I8N. C. Firth, N. W. Keane, D. J. Smith and R. Grice \ goo 120° 57 Fig. 3. Polar contour map of IF flux from F+ I, as a function of centre-of-mass scattering angle 8 and velocity u obtained by the Siska method. Incident F-atom direction is denoted by 8 = O", incident I2 direction by 8 = 180".01 O E'lkJ mol-' Fig. 4. Component angular functions T ( 8) and translational energy distributions P ( E ' ) for F + I 2 at an initial translational energy E = 40 kJ mol-'. (-) The wide-angle component, (- - -) the forward component, (- - -) the distribution of initial translational energy. Discussion The differential reaction cross-section determined for F + I2 in the present experiments shows a remarkable contrast to the stripping cross-section previously determined' for F+ Br,. The forward scattering is severely depleted and has a lower translational energy than the wide-angle scattering which peaks in the backward direction. The vibrational- rotational state distributions of IF product determined by laser-induced fluorescencee / O E'/kJ mol-' Fig. 5. Composite angular distribution and translational energy distributions for forward scattering (-) and backward scattering (- - -) arising from the sum of both components for F+ I2 shown in fig.4. Table 1. Reaction energies (kJ mol-') for the most probable Newton diagram forward component wide-angle component E E ;k E ;V E ;I( E LV A Do 40 3 30 33 50 119 meas~rements'~~" for F+ I2 also exhibit two distinct components. The peak observed at the highest rotational levels close to the limit of energy conservation for the higher vibrational levels corresponds to a very low product translational energy and may therefore be identified with the forward-scattered component in the differential reaction cross-section. The plateau in the rotational state distributions for intermediate rotational levels of the intermediate and lower vibrational states corresponds to higher product translational energies and may therefore be identified with the wide-angle component in the differential reaction cross-section.It has been suggested from analysis of the reactive scattering from F+ Br, and F+ ICl' that the preferred geometry of the reaction intermediate becomes increasingly bent as the identity of the central halogen atom changes from Br to I, in accord with Walsh triatomic molecular-orbital t h e ~ r y . ' ~ The structure for the F+ I2 reactant transition state with an interbond angle p = 120°, shown in fig. 6, corresponds to that proposed for F+ICl' and is more strongly bent than that suggested for F+ Br, with p = 135". This structure is an asymmetric top with the principal moments of inertia, II , 1, , 13, about the x1 , x2, x3 principal axes shown in fig.6, and the bond lengths rIF, rI1 given in table 2. The values for the moments of inertia give a close approximation to a prolate symmetric top, with an asymmetry parameter K = -0.96. Fig. 6 shows the reactant transition state formed when the I, molecule axis lies perpendicular to the initial relative velocity and is inclined at an angle ca. 40" to the plane of collision. In this limiting case, the transition state is formed initially with the initial orbital angular momentum L lying in the plane of the transition state inclined at an angle ca. 55" to the dissociation axis. When the I? molecule axis is inclined at smallerN. C. Firth, N. W.Keane, D. J. Smith and R. Grice 59 Fig. 6. Reactant transition state for the F+12 reaction. Principal axes of the transition state are denoted, x, , x2, x3 with x3 perpendicular to the plane of the transition state. The initial orbital angular momentum L is perpendicular to the plane of collision. Table 2. Estimated structure for the reactant transition state with inter-bond angle @lo), bond lengths (A) and moments of inertia (amu A*) P TI F TI I 1, 1 2 13 120 3 .O 2.7 86 640 726 angles to the plane of collision, the transition state is formed initially with its plane inclined at larger angles to the initial orbital angular momentum L. 'If the reactant transition state is assumed to be a rigid asymmetric rotor, its classical equations of motion may be solved analytically.26 Differential cross-sections may be calculated27 from the precessional motion of such a rigid transition state as a function of the initial angle y between the I F bond and the plane of collision. Results of such calculations are shown in fig.7 for the extreme case of a strongly bent F+ I2 reactant transition state with p =90" and rIF=4.0 A. The lifetime of the reaction intermediate is assumed to be no longer than 0.5 rotational periods, so that precession carries the direction of IF scattering through only the forward hemisphere. Fig. 7 demonstrates that precessional motion arising from transition states initially perpendicular to the plane of collision y = 90" favours sideways scattering, while the scattering moves progressively forward when the plane of the transition state lies initially closer to the plane of collision.The differential cross-section for direct forward scattering which is enhanced27 by the form factor l/sin 8, is observed only for ye 10". The potential-energy surface for F+ I2 is strongly exoergic and is likely to be strongly attractive in the entrance valley, but has a shallow well of depth Eo = 13 kJ mol-' in the exit alley.^ The calculations shown in fig. 7 suggest that large-impact-parameter col- lisions, with the I2 molecule axis initially inclined at an angle ca. 40" to the plane of collision as shown in fig. 6 , will result in the reactant transition state precessing with its principal axis at an angle a =r 60" with respect to the initial orbital angular momentum L =z 220h. In these circumstances the component of angular momentum for rotation about the principal axis has a magnitude of ca.11Oh which will sustain the extension60 c .P 0.7- 2 0.6- 5 0.5- U U Migration in F+ I2 Reaction Dynamics 09 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 .CI CI 5 0.6 'C $ 0.5 U - - - - - - - - - - 1.0 - ( a ) 0.9 - 0.8 - .- 2 0.7 - -$ 0.6 - $ 0.5- U U 2 0.4 MI 0.3 0.2 0.1 0.0 - 1 2 20 40 60 80 100 120 140 160 180 0.8 5 0.3 ti 0.4 t 0.0 :k 20 ' -'. 1 80 $1" I 1 1.0- 0.9- 0.8 - 0.7 - OS- 0.5 - 100 120 140 160 180 O'O 20 1 0 0 0 O O O O 1, 7 40 GO 80 100 120 140 160 180 e l I 1 80 1 80 $1" 0 -80 100 120 140 160 180 $1 O Fig. 7. Angular distributions of reactive scattering (histograms) calculated for precession of a rigid reactant transition state with p = 90" and rIF = 4.0 A for scattering into the forward hemi- sphere, as a function of the initial angle y between the IF bond and the plane of collision.0, The angular distributions predicted for symmetric top precession with the same value of (cos2a'). y = ( a ) O , ( b ) 10, (c)30, ( d ) 6 0 , (e)90".N. C. Firth, N. W. Keane, D. J. Smith and R. Grice 61 of the I F bond for trajectories that involve orbiting on the attractive potential in the entrance valley of the potential-energy surface. Such precessional motion is thus likely to yield I F product molecules in highly excited vibrational-rotational states. The component of angular momentum for rotation of the principal axis has a magnitude of ca. 190 h, which will provoke dissociation of the I2 bond that is impeded by the potential-energy well in the exit valley of the potential-energy surface.The product translational energy for this dissociation may be estimated from the height of the centrifugal barrier in the exit valley B:, = pEL'2/(p'L2). (4) This gives == 8 kJ mol-l, which is rather greater than the peak product translational energy, Ekk = 3 kJ mol-', for the forward component of the reactive scattering. The angular distributions, shown in fig. 7, of scattering arising from precession of an asymmetric top2' indicate that such an initial configuration cannot yield scattering into the forward direction at 8 = 0". However the motion of the dissociating FII intermediate is certainly not rigid. Consequently, the I F product angular momentum J' may become oriented closer to the initial orbital angular momentum L without decrease in magnitude as the products separate, allowing scattering into the forward direction.Indeed, the forward-scattered component of the differential reaction cross-section shows only mild forward peaking in fig. 4, which is in accord with the angular distributions of fig. 7, when scattering arises from I2 molecule orientations which are initially inclined to the plane of collision. Large-impact-parameter collisions with I2 molecules lying in the plane of collision give rise to in plane motion of the reactant transition state. In this case sharp forward peaking is predicted for precession of a rigid asymmetric top as illustrated by the differential cross-section for y = 0" in fig. 7. Indeed the F+ Br2 reaction exhibits just such sharp peaking, but this is absent from the angular distributions for F+12 shown in fig.4 and 5 . Rather a sharp peak is observed in the backward direction associated with the higher product translational energy of the wide-angle component of the reactive scattering. This suggests that orbiting trajectories for in-plane I2 molecules may carry the F atom completely round the I atom to which it is initially bonded until it meets the other I atom of the I2 molecule. If the F atom is now able to intrude between the I atoms of the I2 molecule, repulsion will be induced, since the I-F-I configuration is less stable than the F-1-1 configuration of the reaction intermediate.25 Higher product translational energy will thereby be induced and, when the F atom remains with the I atom to which it has migrated, will result in scattering which peaks in the backward direction.Indeed, migratory trajectories of just this form have been observed in the Monte-Carlo study of the F+ I2 reaction by Fletcher and Whitehead.' The angular distribution for migratory trajectories, shown in fig. 1 of ref. (9), shows a striking resemblance to the angular distribution for the wide-angle component of the F+12 reactive scattering shown in fig. 4. A major proportion of trajectories are found to be migratory in the Monte-Carlo study,' in accord with the predominance of the wide-angle component of the reactive scattering found in the present experiments. The angular distribution for the wide-angle component shown in fig. 4 has a greater intensity for sideways scattering than has the angular distribution calculated9 for migratory trajectories and also exhibits a backward peak but no forward peak.However, the trajectory calculations were carried out on a potential-energy surface with a collinear preferred geometry. A surface with a bent preferred geometry would be expected' to modify the angular distribution in a way that would reduce these discrepancies. The peak in the rotational state distribution at the highest rotational energies for the higher I F vibrational levels observed in the laser-induced fluorescence study ' 6 q ' 7 of F+ 12, has been associated with the forward-scattered component. Indeed, the mechan- ism proposed for the forward scattering does predict high rotational and vibrational62 Migration in F + I2 Reaction Dynamics excitation and low product translational energy for the IF products.The plateau at intermediate rotational levels for intermediate and lower IF vibrational states 16,17 has been identified with the wide-angle component of the differential reaction cross-section. However, the laser-induced fluorescence studies ’ 6,17 indicate that a much higher propor- tion, ca. 30% of the total reaction cross-section, is associated with highly rotationally excited IF product than is indicated in the present study for the forward scattered component, ca. 12%. The laser-induced fluorescence experiments 16,’ were conducted at a lower initial translational energy, E = 10 kJ mol-’, than that of the reactive scattering measurements reported here, E = 40 kJ mol-’.The trajectory studies’ indicate that migration becomes an increasingly prominent feature of the reaction dynamics with increasing initial translational energy, so that this apparent discrepancy may arise from the effect of initial translational energy. The only previous example of migratory reaction dynamics which has been estab- lished experimentally are the reactions of H atoms with interhalogen molecules XY28-” where bimodal vibrational-rotational state distributions are observed for the lighter hydrogen halide HY product. These reactions all involve a light attacking atom reacting with a much heavier diatomic molecule on a strongly exoergic potential-energy surface, as is also the case for F+ 12. However, migration may be regarded as being provoked in the H+XY reactions by the greater exoergicity of the reaction forming the lighter HY product, while the potential-energy surface is more attractive for initial approach at the heavier end of the XY molecule.28 No such asymmetry arises in the F+ I2 reaction, where the participation of migratory trajectories arises only from the mobility of the light F atom compared with the ponderous motion of the I2 molecule on a strongly attractive potential-energy surface.This suggests that a small proportion of migratory trajectories may occur in many reactions which are mainly characterised by direct dynamics. Indeed, there is some indication that this might be the case for F+ Br,, ’ where the backward scattering exhibits increased translational energy in a differential reaction cross-section which otherwise exhibits stripping dynamics.The stripping reac- tions of alkali-metal atoms with halogen molecules31 have attractive potential-energy surfaces,32 which are similar to that of the F+ I2 reaction. Trajectory calculations3’ have suggested that migration may occur early in the entrance valley of the potential-energy surface rather than late in the exit valley as occurs for H + XY and F+ 12. The most favourable mass combination of Li + BrZ 34 continues to exhibit stripping dynamics, although Li+Cl, 34 shows evidence for a subsidiary backward peak with lower product translational energy. Indeed, calculations on the Li + F2 potential-energy surface3’ indicate that intervention of a migrating Li+ ion into the F, bond in the exit valley of the potential-energy surface may not invoke repulsion between the F atoms as is expected for F+ 12.Support of this work by the S.E.R.C. is gratefully acknowledged. References 1 N. C. Firth, D. J. Smith and R. Grice, Mol. Phys., 1987, 61, 859. 2 J. M. Farrar and Y. T. Lee, J. Chem. Phys., 1975, 63, 3639. 3 N. C. Firth, D. J. Smith and R. Grice, unpublished work. 4 J. J. Valentini, M. J. Coggiola and Y. T. Lee, Faraday Discuss. Chem. Soc., 1977, 62, 232. 5 C. F. Carter, M. R. Levy, K. B. Woodall and R. Grice, Furaduy Discuss. Chem. SOC., 1973,55,381; 385. 6 Y. C. Wong and Y. T. Lee, Faruday Discuss. Chem. SOC., 1973, 55, 383. 7 N. C. Firth and R. Grice, Mol. Phys., 1987, 60, 1273. 8 I. W. Fletcher and J. C. Whitehead, J. Chem.SOC., Faraday Trans. 2, 1982, 78, 1165. 9 I. W. Fletcher and J . C. Whitehead, J. Chem. SOC., Faraday Trans. 2, 1984, 80, 985. 10 M. I . Urrecha, F. Castano and J. Iturbe, J. Chem. SOC., Faruday Trans. 2, 1986, 82, 1077. 11 T. Trick1 and J. Wanner, J. Chem. Phys., 1983, 78, 6091.N. C. Firth, N. W. Keane, D. J. Smith and R. Grice 63 12 R. J. Donovan, D. P. Fernie, M. A. D. Fluendy, R. M. Glen, A. G. A. Rae and R. J. Wheeler, Chem. 13 B. S. Agrawalla, J. P. Singh and D. W. Setser, J. Chem. Phys., 1983, 79, 6416. 14 P. Das, J. Venkitachalam and R. Bersohn, J. Chem. Phys., 1984, 80, 4859. 15 H. Brunet, Ph. Chauvet, M. Mabru and L. Torchin, Chem. Phys. Lett., 1985, 117, 371. 16 N. Billy, G. Gouedard, B. Girard and J. ViguC, Proc. 6th Eur. Con$ on Dynamics of Molecular Collisions 17 B. Girard, N. Billy, G. Gouedard and J. ViguC, Faraday Discuss. Chem. Soc., 1987, 84, 65. 18 E. H. Appelman and M. A. A. Clyne, J. Chem. Soc., Faraday Trans. 1, 1975,71, 2072. 19 P. A. Gorry and R. Grice, J. Phys. E, 1979, 12, 857. 20 C. V. Nowikow and R. Grice, J. Phys. E, 1979, 12, 515. 21 P. E. Siska, J. Chem. Phys., 1973, 59, 6052. 22 E. A. Entemann and D. R. Herschbach, Discuss. Faraday Soc., 1967, 44, 289. 23 M. A. A. Clyne and I. S. McDermid, J. Chem. Soc., Faraday Trans. 2, 1978, 74, 1644. 24 K. P. Huber and G. Herzberg, Constantsfor Diatomic Molecules (Van Nostrand Reinhold, New York, 25 A. D. Walsh, J. Chem. Soc., 1953, 2266. 26 L. Landau and I. M. Lifschitz, Mechanics (Pergamon Press, Oxford, 3rd edn, 1976). 27 N. W. Keane and R. Grice, Mol. Phys., 1987, 61, 869. 28 J. C. Polanyi, J. L. Schreiber and W. J. Skrlac, Faraday Discuss Chem. Soc., 1979, 67, 66. 29 J. C. Polanyi and W. J. Skrlac, Chem. Phys., 1977, 23, 167. 30 D. Brandt and J. C. Polanyi, Chem. Phys., 1978, 35, 23; 1980,45, 65. 31 D. R. Herschbach, Adu. Chem. Phys., 1966, 10, 319. 32 G. G. Baht-Kurti, Mol. Phys., 1973, 25, 393. 33 P. J. Kuntz, E. M. Nemeth and J. C. Polanyi, J. Chem. Phys., 1969, 50, 4607; P. J. Kuntz, M. H. Mok 34 C . M. Shoieen, L. A. Gundel and R. R. Herm, J. Chem. Phys., 1976,65, 3223. Phys. Lett., 1980, 69, 572. and Half-Collisions (Molec VI, Aussois, 1986), p. 63. 1979). and J. C. Polanyi, J. Chem. Phys., 1969, 50, 4623. Received 6th May, 1987

ISSN:0301-7249    

DOI:10.1039/DC9878400053

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1987, 84, 65-73 Laser-induced Fluorescence Study of the F + I, --+ IF + I Reaction Bertrand Girard, Nicholas Billy, Gerard Gouedard and Jacques Vigue" Laboratoire de Spectroscopie Hertzienne de 1' ENS,? 24, Rue Lhomond, 75231 Paris Cedex 05, France' The F+ I2 4 IF+ I reaction has been studied in a crossed-beam experiment. The IF rovibrational distribution has been measured with the laser-induced fluorescence detection technique. Many new and important results concern- ing the energy disposal in this reaction are thus obtained. Moreover, a bimodal rotational distribution is observed in several vibrational levels: this phenomenon seems to be due to migratory collisions. In this paper we present a crossed-beam study of the F+ I2 -+ IF+ I reactive collision.The I F product has been detected by the laser-induced fluorescence (1.i.f.) technique. In order to obtain a well resolved spectrum, we used C.W. single-frequency dye lasers. Consequently very long spectral scans were made. In this way, the relative population of each rovibrational level of I F was measured accurately on a large range of vibrational and rotational levels. In the present report, we describe mainly the results of this study, many details appear in other The F + I2 + IF + I Reaction We have chosen to study this particular reaction as it belongs to a family already well studied, the halogen atom-halogen molecule exchange reactions. For various spectro- scopic reasons (in particular predissociation) I F is the interhalogen for which 1.i.f.detection can be most efficiently applied, therefore the F+ I2 reaction is a natural choice. However, having I2 as the reaction partner (rather than ICl or IBr), creates some experimental problems: the intense I2 fluorescence makes it difficult (or even impossible) to detect several vibrational levels of I F (2 < v < 7); the I2 + F2 chemiluminescence creates an intense background if the fluorine atomic beam contains F2 in an appreciable amount (as is the case in our experiment). On the contrary, the previous knowledge of this reaction is important and this stimulated our interest. reported results of crossed beam studies of this reaction (and there are two papers in this Discussion also). It was then shown that a large fraction (ca. 8 5 % ) of the energy available to products was channelled into internal energy (vibrational energy of IF or excitation energy of the iodine atom in its 2P1,2 state).In 1975, the rate constant of this reaction was measured and found to be large' [ k = (4.3 * 1.1) x lo-'' cm3 s-'1. Several 1.i.f. studies were then made.6-9 The most complete study' gave evidence of a bimodal vibrational distribution with a very intense peak at v = 0 and a slightly inverted distribution peaking at high v ( u = 17). The v = 0 peak was interpreted as indirect evidence of the production of iodine atoms and it was concluded that this was the main branch of this reaction. However, three experiments''-'* were then made to detect these 2P,,2 iodine atoms. In the 1973 Discussion of the Faraday Society, two t Unit6 associke au CNRS UA n"18.65L.I.F. Study of F+ Iz+ IF+ I Table 1. Main parameters of the fluorine and iodine molecular beams (the densities refer to the beam crossing region; their values, as well as the iodine beam velocity, were estimated, but the fluorine beam velocity distribution was measured by the time-of-flight technique) fluorine beam iodine beams supersonic supersonic quasi-eff usive source pressure/mbar 3 60 40 1 source temperature/'C 700 125 120 velocity/ cm s-' 1.02 x lo5 3 x lo4 2.2 x lo4 nozzle diameter/mm 0.2 0.2 0.8 density/g cmF3 1.5 x 10" 1.7 x 1013 6 x lot2 parallel speed ratio Si, 7.6 not measured not measured Although not in good agreement with each other, they all prove that the production of I 'P,,* is at most a very minor branch of this reaction.Experiment a1 The Crossed Beam Apparatus The two molecular beams cross inside a large vacuum chamber and the laser beam illuminates the beam-crossing region. The atomic fluorine beam is supersonic and is produced by thermal dissociation of F2 seeded in argon (10-9OOh) in a nickel oven. This beam passes through a skimmer and a collimator, both differentially pumped. The iodine molecular beam is produced by a pure iodine expansion. In order to have a large iodine density in the beam-crossing region, the beam source is located in the main chamber and is only weakly collimated. Table 1 gives the main characteristics of these two beams. The experiments with the blue laser (A < 495 nm) have been performed with a quasi-effusive iodine beam, also described in table 1.With the supersonic iodine beam, the blue laser excited an intense fluorescence signal that we attribute to iodine molecule dimers (12)2 and this signal was considerably reduced when the supersonic iodine beam was replaced by the quasi-effective beam. Lasers and Wavelength Measurement The lasers are home-made C.W. single-frequency dye lasersI3 pumped by krypton ion lasers. They allow long continuous scans (up to 10cm-'). This is very useful as we have scanned very large spectral ranges. The laser wavenumber is measured by a home-made Michelson type lambdameter,I4 with an accuracy of the order of cm-'. This accurate value of the wavenumber is very useful for the identification of the lines observed. The laser beam is carried from the lasers to the beam machine (in the next room) thanks to a set of multidielectric mirrors and it is weakly focussed on the beam-crossing region by a two-lens system.Several dyes were used as well as different pump laser wavelengths to cover the studied spectral ranges. Table 2 shows this information and the levels that are detected in each spectral region.B. Girard, N. Billy, G. Gouidard and J. Vigue 67 Table 2. The different spectral ranges, the dyes used and the vibrational levels of IF that can be detected wavelength range/nm 470-500 700-789 837-873 dye Coumarin 102 LD 700 HITC vibrational levels 091 (2) 8-16 13-20 v) * c 0 500( 1 I I I 13198 13200 wavenumber/cm-' Fig. 1. A typical laser-induced fluorescence spectrum of IF. The fluorescence signal (in counts per 0.2 s channel) is plotted as a function of the laser wavenumber.The 1.i.f. lines appear on the background due to the I, + F2 chemiluminescence. Fluorescence Detection The fluorescence light is collected by a two-lens system, filtered and detected by a photomultiplier with a bialkali-metal photocathode. The filter is used to reject the laser scattered light and also the anti-Stokes fluorescence emitted by iodine molecules excited by the i.r. laser. The combined effects of the cathode sensitivity and of the filter transmission select the following spectral ranges: 300-600 nm with the i.r. laser, 500- 630 nm with the blue laser. The photomultiplier pulses are treated by a standard photon counting electronics. Experimental Procedure and Data Treatment Typical scanning speed was 5 x lo-' cm-' s-' so that a typical 5 cm-' long scan was recorded in 103s.On the average 20 such scans were made every working day and scanning the i.r. regions took 35 days of experiment. During a scan, several quantities are measured and stored in the memory of a microcomputer, for further processing. The most important quantities are obviously the fluorescence intensity, as given by the number of counts per 0.2 s channel and the laser wavenumber. The 1.i.f. lines appear on the intense background of the chemiluminescence due to the I,+F2 reaction (the most intense unblended 1.i.f. line is ca. lo4 counts s - '68 L.I.F. Study of F+ Iz+ IF+ I Table 3. The different contributions to the available energya term average value r.m.s. deviation /eV molecule-' /eV molecule-' difference of binding energies AD, collision kinetic energy iodine internal energy fluorine atom fine structure energy total energy 1.226 0.111 0.016 0.01 1 1.364 0 0.01 8 0.02 1 0.02 1 0.035 a The internal energy of the iodine beam was estimated using reasonable values of the rotational and vibrational temperatures ( T,,, = 100 K, Tvib = 200 K). For each term, the table gives the average value and its r.m.s.deviation. All quantities are in eVmolecule-'. and the chemiluminescence ca. 4 x lo4 counts s-'). We used the chemiluminescent back- ground as a measure of the product of the two beam intensities and we divided the line intensity by the value of the background intensity. This gave an efficient way of correcting the slow drifts of the iodine beam intensity.Fig. 1 presents a typical spectrum. The positions and intensities of the lines were measured automatically by the computer. After a treatment briefly described in the next part, the population of a level UJ was deduced from the intensities of the 1.i.f. lines starting from this level. Results In this part, we discuss first the problem of the available energy, then we present separately the results obtained with the i.r. laser and with the blue laser. The Energy Available in the F + I2 Collision There are four contributions to this energy: the difference of binding energies AD,, the collision kinetic energy T and the internal energies of the two reactants (rotational and vibrational energies for 12, fine structure for F). Table 3 gives for each quantity its average value and its r.m.s.deviation. There is an important dispersion of the total energy. However, the exact importance of the F2P1,2 state is unknown, as is the reactivity of this state, Map of the Detected Levels We followed each band throughout the spectrum and thus deduced the assignments of all the observed lines (ca. lo4 lines were recorded and analysed). This was necessary as the existing spectro~copy'~ of the B-X system was not sufficiently accurate, especially at high J values, to give the assignments. A new spectroscopy has been derived and will be published soon.I6 Fig. 2 .shows the map of the detected levels. For each vibrational level, the last rotational level observed has an energy close to 1.40 eV. The small fluctuations which appear are perhaps significant, but may be due also to the fact that our ability to detect a low-intensity line depends on many factors (local density of lines, absolute intensity of the bands etc.).The energy limit is well inside our estimate of the total energy. We have not included here the results concerning the u = 0 level discussed later. Saturation of the Excitation Process We have shown" that under the experimental conditions used here the excitation process is in the coherent saturation regime. This saturation regime is well understood andB. Girard, N. Billy, G. Goubdard and J. Vigub .- 69 Fig. 2. The map of the observed levels in the VJ plane. The levels that we have detected are shown by the heavy lines. The energy limiting curves correspond to an internal energy of IF equal to 1.40 eV (curve I ) or 0.46 eV (curve I*), the difference being equal to the excitation energy of the iodine atom.experiments are in good agreement with theory. The most important results are the following. The 1.i.f. signal is a linear function of the laser electric field, i.e. the square root of the laser power, and also of the dipole moment of the excitation transition (in absence of saturation it should be quadratic in these two quantities). The 1.i.f. signal per molecule is not a simple function of the time T spent by the molecule in the laser beam: this signal is linear in T if r T < 1 and proportional to if r T >> 1, where r is the total decay rate of the excited state. This decay is equal to the radiative decay rate r r a d for predissociation free levels: r r a d is equal to about 1.2x lo5 s-’ l 8 and the average time (7) is equal to (2-4) x lop6 s, so that r ( T ) < 1 and it is a reasonable approximation to consider that the 1.i.f.signal is linear in T and that, consequently, this is a measure of the population of the level uJ in the beam-crossing region. However, the time (7) depends on the level uJ through the differential cross- section and the recoil energy. In this paper we have not taken this effect into account. Distribution of the Population over the Rovibrational Levels uJ From the integrated intensities of the 1.i.f. lines, we have deduced the populations. This calculation” requires in particular the knowledge of the excitation transition dipole moment ( u’lDBx ( T ) ~ U ’ ’ ) ~ .We have calculated this quantity (the index J means that both wavefunctions were corrected for centrifugal distortion) using the DBX ( r ) function measured by Trautmann et al.”70 L.I.E Study Of F+I,+IF+I t lo t g O 8 50 100 150 f 19 f l 8 4 l6 + J Fig. 3. The rovibrational distribution deduced from the 1.i.f. line intensities. For each vibrational level v, the population n(v, J ) is plotted as a function of J. The population unit is arbitrary, but the same for all v (including v = 0 in fig. 5 ) . Each tick on the horizontal axis corresponds to a J increment of 10. We have verified that the population thus calculated was independent of the transition used for its detection if we except the following problems. There were three sudden changes of sensitivity during the recording time, probably due to misalignments.By comparing the line intensities that were recorded twice (before and after each change), we easily corrected the effects of these changes. The B state is affected by predissociation. The predissociation threshold was known in the vibrational levels ZJ = 8-10,’’ and we found evidence of the thresholds in the levels ZJ = 4-9 through the reduction of the 1.i.f. signal. Obviously we did not use the intensities of the lines involving predissociated levels for the calculation of populations. Finally the measurements involving the weakest bands are more noisy and may differ from the measurements using intense bands by up to 30%. This is partly due to the imperfect knowledge of the dipole matrix elements, the relative error being larger for the weak bands.B.Girard, N. Billy, G. GouCdard and J. ViguC h a v E: 10 15 20 71 V Fig. 4. The vibrational population n ( u ) as a function of u. Our results are plotted in ( a ) and the curve is a guide for the eye. The results of Trick1 and Wanner' are plotted in ( b ) and they are noticeably different. In both cases the population unit is arbitrary. T X --Y X' x xx/ \ . X x\ f I I I I 1 0 10 20 30 40 50 J Fig. 5. The rotational distribution measured in the u = 0 level. The curve is a Boltzmann distribution fitted to the data.72 L.I.F. Study of F+ I2+ IF+ I Fig. 3 presents the resulting population n (v, J ) as a function of J for all the vibrational levels u = 8 to 19. The rotational distributions presented here are clearly bimodal for several vibrational levels, v = 9 to 15: there is a plateau and a peak at high J.These distributions can also be used to calculate integral properties, as the vibrational popula- tion n( u ) shown in fig. 4. All these results will be discussed in the final part of this paper. The D = 0 Level The results concerning this level are quite different in nature and they are described in detail in ref. (1 ). We have detected 1.i.f. lines arising from the v = 0 state. The populations thus measured are plotted as a function of J in fig. 5. This population is well represented by a Boltzmann distribution with a temperature T,,, = 263 f 8 K. We compared directly the populations of the v = 0 and v = 13 levels, and the population units are the same on fig.3 and 5. After integration over J, we get n( v = 0)/ n( v = 13) = 1.7 f 0.7 a result very different from the one found by Trickl and Wanner9 for the same quantity (103 f 22). Moreover we proved that this signal is an artifact; this was shown directly by modulating the fluorine beam at ca. 1.2 kHz. All the direct collision signals (the chemiluminescence of the F2 + I2 reaction, 1.i.f. signals of the high vibrational levels) were totally modulated, as the fluorine beam, while the 1.i.f. signal of the v=O level was not. This proves that this signal is due to a cold IF vapour which persists after the extinction of the fluorine beam, at least for a few hundred microseconds. The origin of this cold vapour is not clear. It is probably due to surface effects.The direct reactive signal in v = 0-1 has been found to be under the detection limit of our experiment. Discussion First, our results on the v = 0 levels prove that the results previously on the low v levels ( v = 0, 1) are not directly due to reaction products and that these levels represent only a minor branch of this reaction. The main features of the energy disposal in the F+I, reaction are now understood. The first crossed beam experiments"" showed that only a small fraction (ca. 15%) of the available energy appears as recoil energy of the products. It is now clear, thanks to the results concerning the direct detection of I 2P1/2,10-12 that the electronic energy is a negligible fraction of the available energy. This result is contrary to a prediction made by Dinur et uZ.:~' applying the information-theoretic approach of Levine and Bernstein2l to the interpretation of the experimental data of ref.(3) and (4), they predicted a value of the order of 0.24-0.30 for the 2P1/2-2P3/2 branching ratio, which is considerably larger than the observed value. If we neglect completely the electronic energy, the internal energy is shared by vibrational and rotational energies. From our data we may calculate the fractions of energy disposed in vibration and in rotation, and deduce the fraction in translation as the complementary part. These results appear in table 4. They are not very accurate as we ignore the contributions of the levels that have not been detected here. Neverthe- less, the recoil energy is in good agreement with previous result^."^ As shown in fig.4, the vibrational distribution presents a broad peak centred on v = 13. Trickl and Wanner9 also observed a peak at high v, but the shape and locations of the two peaks are noticeably different; it is not possible to establish what is due to a real phenomenon (due for instance to the different collision energies) and what is a consequence of their way of analysing unresolved spectra. Finally, the bimodal rotational distribution deserves a comment. Similar distributions have been observed by J. C . Polanyi and co-workers. The most striking example hasB. Girard, N. Billy, G. Goukdard and J. Viguk Table 4. The vibrational and rotational average energies deduced from the rovibrational distribution presented in fig. 3 (9 s v d 19) and the recoil energy deduced from the knowledge of the average total energy 73 fraction of the type of energy average energy/eV total energy rotational vibrational recoil 0.19 0.94 0.23 0.14 0.69 0.17 been obtained in the H + ICl- HCl+ I reaction.22 The bimodality has been interpreted in this case as a consequence of the existence of two different families of paths on the potential-energy surface:13 a direct path, in which the H atom interacts only with the CI end of the ICI molecule, and a migratory path in which the H atom interacts with the I end first and, after a more or less complex trajectory, reacts with CI.The existence of migratory collisions has been observed by Fletcher and Whitehead24 in their trajectory calculation of the F+I, reaction.However, migratory trajectories were found to be statistically insignificant in another calculation made by Urrecha et al.2s on another type of potential-energy surface. The dependence of the importance of migration on the potential-energy surface is not surprising, but it proves that in the absence of further information on the potential-energy surface, we cannot be conclusive on the origin of the observed bimodality. Obviously , one can imagine other explanations, for instance the existence of several reactive potential-energy surfaces. This work was supported by the GRECO 87 ‘Dynamique des Rkactions Molkculaires’ (CNRS). References 1 2 3 B. Girard, N. Billy, G. Gouedard and J . Vigue, Chem. Phys. Left., 1987, 136, 101. B. Girard, N. Billy, G . Gouedard and J .Vigui, to be published. C. F. Carter, M. R. Levy, K. B. Woodall and R. Grice, Discuss. Faraday Soc., 1973, 55, 381; 385. 4 Y. C. Wong and Y. T. Lee, Discuss. Furadaj. Soc., 1973, 55, 383. 5 E. H . Appelman and M. A. A. Clyne, J. Chem. Soc., Faraday Trans. 1, 1975, 71, 2072. 6 R. J . Donovan, D. P. Fernie, M. A. D. Fluendy, R. M. Glen, A. G. A. Rae and J. R. Wheeler. Chem. 7 J. R. Wheeler, Thesis (University of Edinburgh, 1982). 8 T. Trickl and J . Wanner, J. Chem. fhys., 1981, 74, 6508. 9 T. Trickl and J . Wanner, J. Chem. Phys., 1983, 78, 6091. Phys. Lett., 1980, 69, 472. 10 B. S. Agrawalla, J . P. Singh and D. W. Setser, J. Chem. Phys., 1983, 79, 6416. 11 P. Das, T. Venkitachalam and R. Bersohn, J. Chem. Phys., 1984, 80, 4859. 12 H. Brunet, Ph. Chauvet, M. Mabru and L. Torchin, Chem. Phys. Lett., 1985, 117, 371. 13 F. Biraben and P. Labastie, Opt. Cornmun., 1982, 41, 49. 14 J. Vigue and B. Girard, Rev. Phys. Appl., 1986, 21, 463 and references therein. 15 T. Trickl and J . Wanner, J. Mol. Spectrosc., 1984, 104, 174. 16 G. Gouedard, N . Billy, B. Girard and J . Vigue, to be published. 17 N. Billy, B. Girard, G. Gouedard and J . Vigue, Mol. Phys., 1987, 61, 65. 18 M. A. A. Clyne and I . S. McDermid, .I. Chem. Soc., Faraday Trans. 2, 1978, 74, 1644. 19 M. Trautmann, J. Wanner, S. K. Zhou and C. R. Vidal, J. Chem. Phys., 1985, 82, 693. 20 U. Dinur, R. Kosloff, R. D. Levine and M. J. Berry, Chem. Phys. Lett., 1975, 34, 199. 21 A. Ben Shaul, R. D. Levine and R. B. Bernstein, J. Chem. Phys., 1972, 57, 5427. 22 M. A. Nazar, J. C. Polanyi and W. J . Skrlac, Chem. Phys. Lett., 1974, 29, 473. 23 J. C. Polanyi, J. L. Schreiber and W. J. Skrlac, Discuss. Faraday Soc., 1979, 67, 66. 24 I . W. Fletcher and J . C. Whitehead, J. Chem. Soc., Faraday Trans. 2, 1982, 78, 1165. 25 I . Urrecha, F. Castaiio and J. Iturbe, J. Chem. Soc., Faraday Trans. 2, 1986, 82, 1077. Received 18th May, 1987

ISSN:0301-7249    

DOI:10.1039/DC9878400065

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1987, 84, 75-86 Dynamics of the Reactions of Aluminium Atoms studied with Pulsed Crossed Supersonic Molecular Beams Michel Costes," Christian Naulin, Gerard Dorthe, Claude Vaucamps and Guy Nouchi U.A. 348: Photophysique Photochimie Molkculaire and U.A. 283: Centre de Molkculaire Optique et Hertzienne Universite' de Bordeaux I, 33405 Talence, The reactive collisions of aluminium atoms with 02, C02 and SO2 have been studied with crossed pulsed supersonic molecular beams. Aluminium atoms were obtained from vaporization of an aluminium rod, using an excimer laser. They were probed, together with the product A10, by laser-induced fluorescence. Collision energy ranges were 0.08-0.49 eV for Al+ 02, 0.14-0.53 eV for A1 + C02 and 0.30-1.19 eV for A1 + SO2.The variation of A10 rovibrational distributions and reactive cross-sections with collision energy has been determined for each reaction. The value Dg (A10) = 5.26k0.03 eV has been found for the A10 dissociation energy. Physique France The reactive scattering studies of metal atoms have been hampered by the difficulty of obtaining beams of such atoms. For most studies only metal-oven effusive sources have been used. We recently demonstrated that intense pulsed supersonic beams of carbon or aluminium atoms obtained by laser vaporization of the corresponding solid at the exit of a pulsed nozzle could be successfully applied to reactive collision experiments. 1 - 3 Besides the well known advantages of supersonic beams over effusive sources, such as greater intensity and adjustable velocity from seeding in different carrier gases, interest in the method lies in its versatility, since the same design allows one to study the reactivity of a large number of metal atoms, including the most refractory ones.In the present paper we report a study of the dynamics of three reactions of aluminium atoms with oxidizers: A1(2PJ)+0,(X 'Zg> -+ AlO(X 'Z+)+O('PJ) Al('P,) + C02(X ' Z i ) -+ AlO(X *Z+) + CO(X 'Z+) A1(*PJ)+SO,(X ' A , ) -+ AlO(X 'Z+)+SO(X 'Z-). ( 1 ) ( 2 ) (3) The A1 + O2 reaction has been previously studied in a beam-gas configuration using an aluminium atom eff usive-oven source with and without velocity election,^^' thus providing an interesting test for comparison, whereas our work gives the first information on the Al+C02 and Al+SO, reactions.Experimental The instrument combines the techniques of pulsed crossed supersonic molecular beams, laser-induced fluorescence (1.i.f.) detection and atom generation by laser vaporization. Molecular-beam System The scattering chamber is a stainless-steel cylinder (400 mm i.d. and 500 mm long) evacuated by a single 8000 dm' s-' oil diffusion pump. The two pulsed valves of the 7576 Dynamics of the Reactions of Aluminium Atoms Table 1. Characteristics of the oxidizer beam gas mixturea u/m S-" S V / V (yo)' S t / p s d O2 : He = 0.12 : 0.88 02: Ar = 0.12: 0.88 ~ ~~~ 1360 15 24 660 17 24 C 0 2 : He = 0.18 : 0.82 1220 18 30 SO2: He = 0.10: 0.90 1280 20 30 a Source pressure P = 4 bar, except for SO2-He where P = 3 bar. ' Mean velocity at the peak intensity. Velocity spread (f.w.h.m.).Pulse width (f.w.h.m.) at the crossing point. Table 2. Characteristics of the A1 atom beam for some of the carrier-gas mixtures used' gas mixture' v/m s-' I' s u / v (Yo) st/ P s Ar Ar : Ne = 0.66 : 0.33 Ne Ne: He = 0.27 : 73 Ne : He = 0.09 : 0.91 He H2 810 176 17 20 890 200 15 16 1260 210 15 14 1630 620 12 7 1900 760 12 6.5 2140 830 11 4.3 3410 1200 10 3 a For symbols see table 1. 1.i.f. (arb. units). Source pressure P = 6 bar. A1 peak intensity, measured by Gentry and Giese design6 in its commercial version (Beam Dynamics) are operated with stiff I shape bars to achieve minimum pulse durations. The base pressure, which is low6 mbar with beams off, increases to <lop5 mbar with both valves in operation (uncorrected gauge response).The nozzle-crossing point distance is 80 mm for the oxidizer beam and 110 mm for the aluminium beam. The two beams crossing at right angles are collimated to 4 and 6.5" f.w.h.m., respectively. Fast ionization gauges (Beam Dynamics) are used to measure the time profiles, velocities and velocity distributions of the beams. These gauges also allow one to measure the attenuation of one beam by another, which remains < loo% in all experiments, thus ensuring reasonably single collision conditions. The characteristics of the beams are summarized in tables 1 and 2. L.I.F. Detection The reactant A1 and the product A10 are probed by 1.i.f. at the crossing point. A Nd-Yag pumped dye laser (Quantel YG481+TDLIV) is operated with 532nm pumping of rhodamine dyes.Radiations at shorter wavelengths are generated by stimulated Raman scattering in H2 and the desired anti-Stokes line is selected through a set of four Pellin-Broca prisms. The laser power is then lowered in a continuously variable attenuator (NRC 935-5) and measured at the exit of the collision chamber with a photodiode. The scattered laser light is reduced by extensive light baffling. The fluor- escence of the excited zone is imaged on the photocathode of a Hamamatsu R955 photomultiplier tube, the output of which is fed into a boxcar integrator (PAR models 115, 162, 165). When the dye laser is operated with a mixture of Rh590 and Rh610, rapid switching from A10 to A1 detection is possible. In that case, A10 is probed on theM. Costes, C. Naulin, G.Dorthe, C. Vaucamps and G. Nouchi 77 ( B 'X+ c+ X 'E')Au = +1 sequence, between 465 and 472 nm, by scanning over the first anti-Stokes wavelength range of the dye: the fluorescence is isolated with a coloured glass filter centred at 465 nm (8.5 nm f.w.h.m., T = 0.60). A1 atoms are detected on the 4s 2 S , / 2 ++ 3p ' P I / ' and 4s 2S1/2 ++ 3p 'P3/2 transitions at 394.4 and 396.2 nm situated in the second anti-Stokes wavelength range of the same dye: the fluorescence is attenuated through a filter having a transmission factor T = 3.5 x lop4 to avoid saturation of the p.m.t. Excitation spectra of A10( X 'X+) are recorded in the intermediate saturation regime keeping the mean laser power constant; this minimizes the pulse-to-pulse fluctuations in the fluorescence signal and simplifies the analysis of rovibrational distribution^.^ Optical pumping effects are expected to be small for vibrational levels u" = 0, 1,2, which give the main features of the observed spectra, as the corresponding Einstein coefficients in the A u = +1 sequence vary only by a factor of 1.34.738 Some experiments have been performed by excitation of AlO(X 2X+) on the Au = -1 sequence, between 508 and 515 nm, by scanning through the first anti-Stokes wavelength range of the Rh640 dye.Aluminium Beam Generation The aluminium beam is produced by the technique described in previous works of Smalley and co-workers for the generation of supersonic metal clusters."' However, as our goal is to produce atoms and not clusters, arrangement has the following features.First, we use an excimer laser (Lambda Physik EMGlOlE) operated at 249 nm with an unstable resonator, rather than a doubled Nd-Yag laser. A fraction of the output (ca. 7 mJ) is selected by spatial filtering and focussed to a 0.4 mm diameter circular spot on the rod (96% A1-4% Mg alloy). The combination of the 5 eV photon energy and the high fluence (5.5 x lo4 J m-2) ensures rupture of all the Al-A1 bonds in the ablated material. A higher production rate of atoms us. clusters is thus expected. Secondly the gas load of our valve (0.03 mbar dm-' per pulse) is ca. 20 times lower than that used by Smalley and co-workers for their pulsed valve. Cluster growth by three-body recombination is thus much lower. Furthermore, this latter effect is enhanced, as we use a short extension channel: the distance between the beam waist and the expansion into the vacuum is only 4 mm.Under such conditions the neutralization of ions is not complete. Most of these undesirable species are removed from the beam in the expansion region immediately after the nozzle, with an electric field of 2.5 x lo4 V m-'. The aluminium beam obtained is a mixture of the two spin-doublet components 2Pl/2 and 2P3/2, with the excited 'P3/' level ( E : = 0.014 eV) weighting for ca. 40-50% for all carrier-gas mixtures. Table 2 gives the variation of the velocity at the peak density for the different mixtures. However, the same buffer gas mixture can give beams with significantly different velocities depending on the fluence and the total energy deposited on the rod, the gas load of the valve and the time delay between the opening of the valve and the excimer laser pulse.Also given in table 2 is the variation of the peak density, which shows its growth with velocity: this is the result of the pulse narrowing in time as lighter seeding gas mixtures are employed. Procedure Experiments are performed with the whole system synchronized at 10 Hz repetition rate except for the oxidizer beam, which is triggered at 5 Hz. A10 excitation spectra are thus recorded as the result of the difference signal in consecutive shots with the oxidizer beam on and off. This procedure was found necessary as the aluminium beam also contains small amounts of A10, in the low rotational levels of u" = 0. The metal oxide production from such sources seems to be a general problem since it was also encountered in previous works.1o71' The tricky point in our own experiments is that the A10 production78 Dynamics of the Reactions of Aluminium Atoms I 1 I I 1 I 465 467 469 wavelength/ nm Fig.1. Excitation spectrum of AlO(X 2C+) from the A1 + O2 reaction at qrans = 0.29 eV: ( a ) experimental and ( b ) calculated. is found to be erratic, being zero for short periods of time. We relate this to the extraction of the third cavity round trip from the excimer unstable resonator, which results in a much higher energy density at the beam waist, and may thus be responsible for the complete dissociation of A10 by multiphoton absorption near the ablated surface. Owing to thermal drift in the laser, which modifies the cavity alignment, this part of the excimer beam cannot be extracted through the spatial filter with a constant efficiency.Inciden- tally, this observation strengthens our argument that the ablation gives predominantly A1 atoms rather than clusters. Results and Discussion Al+O,+ A10+O A10( X *X+) excitation spectra have been recorded at four relative translational energies: = 0.083, 0.19, 0.30 and 0.49 eV. Two of these are given in fig. 1 and 2. The quantitative information from such data is extracted by comparison with synthetic spectra, calculated with the molecular constants from Coxon and Naxakis.12 In these calculations the following distribution function is introduced to model the rotational populations: P ( v”, K”) cx (2K”+ 1)[ - G( v”) - F( K ” ) ] exp[-PF( K ” ) ] (4) where P(v”, K ” ) is the probability of populating the level v”, K” ( K ” = J”* 1/2, incor- porating the spin splitting), etot is the total energy available to the reaction products and G( v”) and F( K ” ) are the vibrational and rotational energies, respectively.The first part of the expression represents the prior expectation, given by the information-theoretic approach to the analysis of state-to-state reaction dynamics, and the exponential term the deviance to the prior expectation, defining the surprisal parameter P.I3M. Costes, C. Naulin, G. Dorthe, C. Vaucamps and G. Nouchi 79 1 I I I 1 I 465 467 469 wavelength/ nm Fig. 2. Excitation spectrum of AlO(X 2C+) from the Al+O, reaction at = 0.083 eV: ( a ) experimental and ( 6 ) calculated.At the lowest relative translational energy the experimental spectrum closely agrees with a synthetic spectrum computed for a completely statistical (both vibrational and rotational) distribution. No reasonable fit could be obtained when incorporating a value of p, even small, in the calculations. One interesting feature of a statistical-rotational distribution is that line intensities near the excitation limit can be markedly different with a slight variation in ctOt. This property is used to deduce the reaction exoergicity. The synthetic spectrum which gives the best fit (fig. 2) is calculated at Etrans = 0.083 eV with a reaction exoergicity = -0.145 eV for ground-state reactants A1(2P,,2) + 0 2 ( X 'Xi). The O2 internal energy is negligible for a supersonic nozzle beam, but the 0.014 eV energy difference between Al(2P3/2) and A1(2P1/2) cannot be neglected. The spectrum is thus simulated as a bimodal distribution wih equal weights of 2 P ! / 2 and P3,?.Synthetic spectra have been computed at several values of the total available energy ranging from 0.210 to 0.240 eV for ground-state reactants. Extreme values corresponding to = -0.130 and -0.160 eV result in a bad fit to the experimental spectrum (fig. 3). Among the different features, the (1 - O)P(53) line appears to be clearly underestimated at Aco= -0.130eV [fig. 3 ( b ) ] and overestimated at Aho= -0.160eV [fig. 3(c)]. An excitation limit of 0.228 f 0.01 eV for ground-state reactants is derived from these simula- tions. Introducing Etrans = 0.083 f 0.015 eV yields = -0.145 * 0.025 eV.This, com- bined with the O2 dissociation energy DE(0,) = 5.1 156 f 0.002 eV,I4 gives the A10 dissoci- ation energy DE(Al0) = 5.26 f 0.03 eV. Table 3 summarizes the results obtained for the A10 internal energy distributions. No significant difference can be seen in the variation of the fraction of energy going into A10 rotation (fR) and vibration (fv). They remain ca. 30 and 14'/0, respectively, for all the kinetic energies listed. However, there is a definite trend of deviance to the prior expectation of the rotational and vibrational distributions on increasing the collision energy. Table 3 also shows the decrease of the relative reactive cross-section, or, with80 Dynamics of the Reactions of Aluminium Atoms 468.5 1 46’9.5 nm wavelength/ nm Fig.3. Part of the excitation spectrum of AlO(X 2X+) from the Al+O, reaction at 0.083 eV: ( a : experimental, ( b ) calculated for A E ~ = -0.130 eV and ( c ) calculated for A E ~ = -0.160 eV. Table 3. Energy partitioning in A10 and reactive cross-section from the A1 + O2 - A10 + 0 reactior Pr“ V f f Po( v) (f”) 0.083 0 0 1 0.19 0 0 1 2 0.29 1.2 0 1 2 0.49 1.3 0 1 2 3 0.74 0.26 0.59 0.345 0.065 0.60 0.30 0.10 0.51 0.30 0.14 0.05 0.75 0.345 0.135 - 0.25 0.595 0.33 0.165 1.0k0.3 0.3 1 0.095 0.56 0.29 0.14 0.6 f 0.2 0.35 0.09 0.39 0.30 0.14 0.5 f 0.15 0.29 0.20 0.12 a Reduced surprisal parameter: pr = P [ qot - G( v”)]. Experimental vibrational distribution Prior vibrational distribution.M. Costes, C. Naulin, G. Dorthe, C. Vaucamps and G. Nouchi 81 qrans.This cross-section, derived from the total A10 population and A1 density, is not calculated for = 0.083 eV. This experiment was indeed performed with a different O2 beam (seeded in Ar) from the others, which employed the same 0, beam seeded in He. Reaction ( 1 ) was first studied by Dagdigian et al. in a beam-gas arrangement using an unselected effusive A1 source at an average collisional translational energy of = 0.13 eV.5 They deduced a value of the A10 dissociation energy DG(Al0) = 5.27 f 0.04 eV. They also concluded that the available energy was not entirely statistically distributed as vibrational levels above 21’’ = 1 appeared less populated than expected a priori. Very similar conclusions were given by a beam-gas study with a velocity-selected effusive source of Pasternack and Dagdigian.4 The A10 dissociation energy was confirmed [ D:(AlO) = 5.27 * 0.02 eV] from the translational energy threshold behaviour of the u” = 2 level.A direct mechanism, involving an attractive surface with mixed energy release channelled primarily into A10 rotation and translation, was suggested. Our results are in good agreement for the majority of points. The A10 dissociation energy is found to lie within the uncertainty range of the two previous studies. Very minor differences are found for the amount of energy channelled into product vibration and rotation, and for the vibrational distributions. However, there is a disagreement in the conclusion concerning a direct reaction mechanism. Our result at low translational energy, which shows randomization of the available energy among the different degrees of freedom, suggests the presence of a well in the potential-energy surface.The departure to statistical behaviour at higher energy can be easily understood by the decrease in the lifetime of the A102 complex. A definite choice is difficult to make, as we are discussing only slight differences in the deduced rovibrational distributions. Clearly we lack a knowledge of the angular distribution. However, the kinetic studies of Fontijn and coworkers, with a fast-flow reactor between 300 and 1700 K,15-” also suggest a complex mode mechanism. A1 + CO2 + A10 + CO This reaction is endoergic, thus exhibiting a translational energy threshold. The AlO(X 2E+, d’ = 0) band head begins to be detected at qrans = 0.17 f 0.02 eV.At this energy only the A1(2P3/2) state is presumably reactive. Spectra near threshold and at higher energy (fig. 4) show the growth of the rotational and vibrational excitations. The information contained in the spectra is again extracted with the aid of synthetic spectra. The same function [eqn (4)] is used to describe the rotational distributions. Although in this case it looses its information-theoretic significance, because no summa- tion over the CO density of states is carried out, this simple model has been kept since it gives excellent fits to the experimental spectra. The analysis of rotational excitation limits for the spectra obtained near threshold leads to a reaction endoergicity = 0.19 f 0.03 eV for ground-state reactants AI(’P,,,) + C 0 2 ( X ‘C,’).Incorporating the COz bond dissociation energy DE(C0-0) = 5.453 f 0.002 eV“ yields DP,(AlO) = 5.26 f 0.03 eV, in perfect agreement with the former determination. The results for the A10 internal energy partitioning are summarized in table 4. The fraction of total energy going into A10 rotation (fR) is ca. 32% at low energy and falls to 23% when the threshold for v ” = 1 is reached. The loss approximately corresponds to the fraction of energy (fv) going into vibration. A further increase of energy does not modify (fR) and (fv) significantly. Vibrational distributions exhibit minor differences from prior ones, the differences seeming greater for 0.30 eV than for 0.39 eV. However, calculated prior distributions are extremely sensitive to the assumed available energy when is close to the threshold for the production of a vibrational level.This is the case for 0.30 eV (threshold for d’ = 1 ) .32 Dynamics of the Reactions of Aluminium Atoms 46 5 467 469 wavelength/ nm Fig. 4. Excitation spectra of AlO(X 2X+) from the Al+CO, reaction at = 0.53 eV ( a ) and 0.20 eV ( b ) . Table 4. Energy partitioning in A10 and reactive cross-section from the A1 + C02 -+ A10 + CO reaction 0.20 1 0 0.235 1.2 0 0.265 0.7 0 0.30 2.5 0 1 0.39 3.5 0 1 0.53 3 0 1 2 1 0.87 0.13 0.87 0.13 0.81 0.16 0.03 1 0.34 0 0.18 f 0.05 1 0.32 0 0.28 * 0.09 1 0.3 5 0 0.40 f 0.13 0.992 0.23 0.1 1 0.51 k0.15 0.008 0.907 0.20 0.08 0.54k0.16 0.093 0.725 0.2 1 0.08 0.93 f 0.28 0.240 0.035 The variation of the relative reactive cross-section is plotted in fig.5 , with the following modelled functions: (Tr= (TO(Fth/Etrans)2’R(1 - &th/&trans)1-2’s ( 5 ) where uO is the maximum cross-section, &th the translational energy at threshold and s takes the value 4, 6 or for an ionic, van der Waals or hard-sphere potential, respec- tively.’’ A good agreement is obtained for s = 4 or 6 in the vicinity of the threshold when computing the curves for &th = 0.19 eV. However, the point at the highest energy seems to deviate significantly from the corresponding curves. This can perhaps be relatedM. Costes, C. Naulin, G. Dorthe, C. Vaucamps and G. Nouchi 83 1 0.E h v) U .C c ’ 0.6 < W d 0.4 0.2 Fig. 5. Relative reactive cross-section, u,, of the A1 + C 0 2 reaction as a function of the relative translational energy, 0, experimental values; (-) calculated (see text).to the unusual behaviour of the rate constant, which shows strong curvature and deviation from the Arrhenius form at high temperature.20 Clearly more information is needed in this energy region to see if there is a real change in the dynamics. A1 + SO2 ---+ A10 + SO The A1 + SO, reaction is also an endoergic process. The excitation of the AlO( X 2E+, zl” = 0) bandhead starts at = 0.36 f 0.04 eV. Spectra at two collision energies are given in fig. 6. Again the analysis of the rotational excitation limits for the spectra near threshold allows one to determine the reaction endoergicity: = 0.32 f 0.05 eV. However, this value combined with the SO, bond dissociation energy taken from the JANAF tables: ‘’ DG(S0-0) = 5.71 f 0.02 eV leads to Dz(Al0) = 5.39 f 0.07 eV.This latter value is com- pletely outside the uncertainty range of the two previous determinations. We are obviously in the same situation as Fontjin and Felder who, taking the JANAF value of SO2 as a reference, were obliged to increase the lower limit of DG(Al0) deduced in their high-temperature fast-flow reactor experiments. They found &( A10) > 5.29 eV from A1 + C02,20 but DE( AlO) > 5.46 eV from A1 + The two sets of independent experiments strongly suggest that DZ(S0-0) = 5.71 eV is too high. The highest value which can reconcile our three experiments is &(SO-0) = 5.65 eV, which coincides with a more recent value given by Okabe.22 Results on A10 energy partitioning are deduced as for the Al+CO, reaction, and are summarized in table 5.The amount of energy going into rotation is fairly constant, (fR) == 0.24, up to = 1.19 eV. The fraction going into vibration does not exceed (f,)=O.lO in the best case. The vibrational distribution remains statistical except at the highest collision energy sampled. = 0.66 eV, but drops to (fR) = 0.18 at84 Dynamics of the Reactions of Aluminium Atoms 465 467 469 wavelength/nm Fig. 6. Excitation spectra of AlO(X 2X+) from the Al+ SO2 reaction at = 1.19 eV ( a ) and 0.42 eV (b). Table 5. Energy partitioning in A10 and reactive cross-section from the A1 + SO2 4 A10 + SO reaction 0.42 2.5 0 0.47 2.5 0 1 0.52 2.5 0 1 0.565 2.5 0 1 0.6 1 2.5 0 1 2 0.66 2.5 0 1 2 1.19 4 0 1 2 3 4 5 1 0.95 0.05 0.9 1 0.09 0.84 0.16 0.79 0.17 0.04 0.75 0.22 0.03 0.61 0.275 0.095 0.02 - 1 0.26 0 0.19 * 0.06 0.98 0.25 0.04 0.41 k0.16 0.02 0.91 0.24 0.05 0.64*0.19 0.09 0.85 0.235 0.075 0.91 f 0.27 0.15 0.8 1 0.23 0.10 0.90 f 0.27 0.18 0.01 0.75 0.23 0.10 1.06 * 0.30 0.22 0.03 0.436 0.18 0.07 0.74 f 0.32 0.270 0.156 0.082 0.038 0.0142M.Costes, C. Naulin, G. Dorthe, C. Vaucamps and G. Nouchi 85 &translev Fig. 7. Relative reactive cross-section, ur, of the A1 + SO2 reaction as a function of the relative translational energy, etrans: 0, experimental values; (-) calculated (see text). The excitation function a , = f ( ~ ~ , ~ ~ ~ ) is given in fig. 7. The best fit to eqn (5) is obtained for a threshold value 6th = 0.40 eV for the ionic and van der Waals potentials. The decrease in the cross-section at the highest collision energy seems to favour s = 4, in agreement with the harpooning mechanism suggested for this reaction.2' References 1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 G.Dorthe, M. Costes, C. Naulin, J. Joussot-Dubien, C. Vaucamps and G. Nouchi, J. Chem. Phys., 1985, 83, 3171. M. Costes, G. Dorthe, B. Duguay, P. Halvick, J. Joussot-Dubien, C. Naulin, G. Nouchi, J. C. Rayez, M. T. Rayez and C. Vaucamps, in Recent Advances in Molecular Reaction Dynamics, ed. R. Vetter and J. Vigue (Editions du C.N.R.S., Paris, 1986), p. 97. M. Costes, C. Naulin, G. Dorthe, J. Marchais and C. Vaucamps, C.R. Acad. Sci. Paris, 1986, 11(303), 1279. L. Pasternack and P. J. Dagdigian, J. Chem. Phys., 1977, 67, 3854. P. J. Dagdigian, H. W. Cruse and R.N. Zare, J. Chem. Phys., 1975, 62, 1824, W. R. Gentry and C. F. Giese, Rev. Sci. Instrum., 1978, 49, 595. R. Alktorn and R. N. Zare, Annu. Rev. Phys. Chem., 1984, 35, 265. G. R. Hebert, R. W. Nicholls and C. Linton, J. Quant. Spectrosc. Radiar. Transfer, 1980, 23, 229. D. E. Powers, S. G. Hansen, M. E. Geusic, A. C. Pulu, J. R. Hopkins, T. G. Dietz, M. A. Duncan, P. R. R. Langridge-Smith and R. E. Smalley, J. Phys. Chem., 1982, 86, 2556. J. B. Hopkins, P. R. R. Langridge-Smith, M. D. Morse and R. E. Smalley, J. Chem. Phys., 1983,78, 1627. D. E. Powers, S. G. Hansen, M. E. Geusic, D. L. Michalopoulos and R. E. Smalley, J. Chem. Phys., 1983, 78, 2866. J. A. Coxon and S. Naxakis, J. Mol. Specrrosc., 1985, 111, 102. R. B. Bernstein, in Chemical Dynamics via Molecular Beam and Laser Techniques (Oxford University Press, New York, 1982), and references therein. K. P. Huber and G. Herzberg, in Molecular Spectra and Molecular Structure IV Constants of Diatomic Molecules (Van Nostrand Reinhold Company, New York, 1979). A. Fontijn, W. Felder and J. J. Houghton, Fifteenth Inr. Symp. Combustion (Combustion Institute, Pittsburgh, 1975), p. 775.86 Dynamics of the Reactions of Aluminium Atoms 16 A. Fontijn, W. Felder and J. J. Houghton, Sixteenth In?. Symp. Combustion (Combustion Institute, 17 A. Fontijn, Combust. Sci. Technol., 1986, 50, 151. 18 JANAF Thermochemical Tables, project directors, D. R. Stull and H. Prophet (N.B.S., Washington, 19 R. D. Levine and R. B. Bernstein, J. Chem. Phys., 1972, 56, 2281. 20 A. Fontijn and W. Felder, J. Chem. Phys., 1977, 67, 1561. 21 A. Fontijn and W. Felder, J. Chem. Phys., 1979, 71, 4854. 22 H. Okabe, J. Am. Chem. Soc., 1971, 93, 7095. Pittsburgh, 1977), p. 871. 2nd edn, 1971). Received 21sf May, 1987

ISSN:0301-7249    

DOI:10.1039/DC9878400075

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1987, 84, 87-126 GENERAL DISCUSSION Prof. J. N. Murrell ( University of Sussex) said: The experimental results presented by Robinson, Continetti and Lee are startling and if confirmed would have a substantial impact on our understanding of radical substitution reactions of aromatic molecules. For radical reactions in solution alkyl substituents have a very weak directing effect, and indeed in the paper it is recognised that both the endothermicity, and the energy of any intermediate complex, are little affected by the methyl group. The authors therefore attribute the non-reactivity of the meta isomer, and, as reported in the discussion, the non-reactivity of chlorobenzene itself, to subtleties in the shapes of the exit valleys of the reactions. If this were true we would be introducing new criteria into the theory of organic reactivity, and before going down that path it would seem wise to carry out other experiments, e.g.molecular beam studies of the reverse reactions. Mr G. N. Robinson, Mr R. E. Continetti and Prof. Y. T. Lee (University of California, Berkeley, CA) said: In reply to Prof. Murrell, our experimental results were reproduced several times and the difference in cross-section that we observed between the meta isomer and the ortho and para isomers is certainly real. Concerning any disagreements between our results and those of previous experiments on aromatic substitution in solution, we would simply observe that an experiment that measures relative rates of homolytic or ionic substitution in solution will not be sensitive to features of the reactive potential-energy surface in the way that a scattering experiment is.We are contemplating a series of experiments to investigate further the effects of substituents on the dynamics of aromatic substitution reactions. Prof. R. N. Zare (Stanford University, CA) said: I found this crossed-molecular-beam study of aromatic substitution reaction to be most fascinating and to raise a number of questions: ( 1 ) Was a study made of the reverse reaction, C1+ o-,m-,p-bromotoluene? What can your present results tell us to expect about this reverse reaction? (2) When fast Br atoms encounter o-,m-pchlorotoluene is HBr production observed? If not, is it understood why this reaction channel is unimportant? (3) In the Br + o-,p-chlorotoluene substitution reaction, the product angular distribu- tion and time-of-flight spectra are fit assuming that a limited number of oscillators in the complex are excited, i.e.participate in the energy sharing prior to C1 atom elimination. Is it known which oscillators are involved? Have studies been made with Br and fully or partially deuterated o-,p-chlorotoluene? (4) What role does the methyl group play in affecting the observed differences between ortho and para chlorotoluene? Are there both electronic and steric effects? Have corresponding studies been made on o-,p- chloroaniline and on o-,p-chlorophenol ? Mr G. N. Robinson, Mr R. E. Continetti and Prof. Y. T. Lee (University of California, Berkeley, CA) said in reply to Prof. Zare: ( 1 ) We did not investigate the reactions C1+ o-,m-,p-bromotoluene ---* Br+ o-,m-,p-chlorotoluene. On the basis of the reasoning proposed near the end of our paper, we believe that the reactions C1+ o-,p-bromotoluene will have higher cross- sections than Cl + rn-bromotoluene.Since A H o for the isomeric Br + chlorotoluene reactions are nearly the same, AGO and K,, are also likely to be the same. Therefore, if the rate constant for the reaction in the endoergic direction is large, the rate constant for the reverse, exoergic reaction should also be large. 8788 General Discussion (2) The H-atom abstraction reactions (Br + C7H,Cl + HBr + C7H6CI) were not studied because there would be substantial background at both product masses at all laboratory angles from inelastic/elastic scattering of the reagents.It is worth considering the energetics of these reactions, however. Abstraction of a ring H atom by Er is endoergic by 24 kcalmol-I. An additional barrier of ca. 10 kcal mol-' (which is not unlikely) would prevent this reaction from occurring at the highest collision energy of our experiments. Yet, even if H abstraction had the same energetic threshold as C1 substitution, direct abstraction would probably occur less readily than Cl substitution since the relative energy of a binary Br-H collision is 49 times smaller than that for Br+CT. Br addition and subsequent HBr elimination would be a more likely way to form HBr with a ring H atom.' Overall, direct abstraction of a benzylic H atom would be the more likely reaction to form HBr since it is exoergic by 2 kcal mol-I.The activation energy for the reaction Br+C6H5-CH3--+ HBr-tC6H5-CH, is known to be 7.2 kcal mol-1.2 (3) We have not studied the Br + chlorotoluene reactions with deuterated compounds although the lower frequency of the C-C-D bending modes might have an effect on the observed dynamics. We have studied the reactions Br + C,X,CI -+ C6X5Br + C1 (X = H, F).3 At a collision energy of ca. 30 kcal mol-' no substitution product is observed for X = H whereas substitution product is observed for X = F. We attribute this difference in reactivity to a larger addition cross-section for Br + C6F5Cl. Both the product transla- tional energy distributions and the excitation function for Br + C6F5Cl --+ C1-k C6F,Br suggest that a larger number of modes participate in energy sharing in this reaction than in the chlorotoluene reactions.In none of these reactions can we actually identify the modes involved and we do not maintain that there is a microcanonical equilibrium among a fixed number of modes prior to bond breakage. In fact IVR competes with bond breakage in these reactions. Our value for the number of active modes is then a relative measure of the extent to which energy is drained away from the reaction coordinate prior to bond breakage. A small point: in fitting our data we used parametrized functions for the c.m. frame flux distribution and assumed nothing about the extent of energy redistribution in the BCMC complexes. The best-fit P ( E ' ) for the Br + o-chlorotoluene reaction at 31.5 kcal mol-' was compared to RRKM-AM P ( E ' ) distributions that were calculated for different numbers of active modes. (4) It is certainly possible that factors other than those discussed in our paper contribute to the observed difference between the o-and p-bromotoluene excitation functions.The heats of formation of the BC2MC and BC4MC radicals, however, are unlikely to be different from one another. We base this conclusion on an assessment of the relative heats of formation of the (hypothetical) conjugated radicals formed when the halogenated carbon atom is removed from the ring, i.e. the 1-methylpentadienyl and 4-methylpentadienyl radicals. Although there appear to be no AH; data for these radicals, there are AH; values for the methylallyl and 2-methylallyl radical^.^ Within experimental error these values are the same (30.0 kcal mol-I).Subtle differences between the slopes of the Br+ o-CT and p-CT potential energy surfaces along the reaction coordinate cannot be ruled out however. The methyl group may indeed block Br addition to the chlorinated carbon of o-chlorotoluene at low collision energies causing Sr,o-bromotoluene to be lower than Sr,p-bromoto]uene for Ec<25 kcal mol-I. With the number of dissipative modes in the vicinity of the collision being larger for o-chlorotoluene, (Tadd( o-chlorotoluene) may be higher than a,,,(p-chlorotoluene) for Ec> 25 kcal mol-'. The increased sideways scattering that we observe for the o-chlorotoluene reaction at higher collision energies is consistent with wider angle Br + o-chlorotoluene collisions becoming reactive as the collision energy is raised.General Discussion b -50 -40 -30 -20 -10 0 10 20 30 89 laboratory angle, O / O Fig.1. CH31 ( m / e = 142) laboratory angular distribution from the reaction CH3 + CF31 + CH31 + CF3, E, = 12 kcal/rnol. Solid and dashed lines are fits to data using the solid and dashed line T( 6 ) distributions in fig. 2( a ) and the P ( E') in fig. 2( b). Radius of Newton circle represents the maximum CH31 recoil velocity in the c.m. frame. Finally, no studies have been carried out with OH and NH2 substituted compounds, although their ortho-para-directing effects should be greater than that for CH3. Future investigations of substituent effects are planned. 1 A. S. Rodgers, D. M. Golden and S.W. Benson, J. Am. Chem. SOC., 1967, 89, 4578. 2 H. R. Anderson Jr, H. A. Scheraga and E. R. VanArtsdalen, J. Chem. Phys., 1953, 21, 1258 3 G. N. Robinson, R. E. Continetti and Y. T. Lee, to be published. 4 D. F. McMillan and D. M. Golden, Annu. Rev. Phys. Chem., 1982, 33, 493. Mr G. N. Robinson, Dr G. M. Nathanson, Mr R. E. Continetti and Prof. Y. T. Lee (University of California, Berkeley, CA) said: We have carried out a crossed beam study of the iodine exchange reaction, CH3 + CF31 + CH31 + CF3 (-AH" == 0-2 kcal mol-I), at a collision energy of ca. 12 kcal mol-'. We generate a supersonic methyl radical beam in a manner similar to that of Grice and co-workers" by passing a mixture of ca. 2% di-t-butyl peroxide in He through a quartz nozzle (0.5 mm diameter) that is in contact with a resistively heated tantalum strip.The nozzle temperature is ca. 1000°C. A laboratory CHJ product angular distribution ( m / e = 142) is shown in fig. 1. The angle at which the distribution falls to zero for O>O" is uncertain because of the presence of background at these angles caused by in/elastically scattered impurity from the CF31 beam. Although it is certain that N(20")=0, it is likely that N(8")=0 (0,,=19"). The angular distributions and time-of-flight data (not presented) are fitted with the c.m. frame product angular and translational energy distributions shown in fig. 2. P ( E ' ) distributions having thresholds in the range 0-3 kcal mol-' can fit the data.90 General Discussion I I I I I l ( I I 0 2 4 6 8 10 12 14 16 E'/ kcal mol-' I,,,,,,,,,, 0 10 20 30 40 50 60 70 80 90 centre-of-mass angle, e l o Fig.2. ( a ) Centre-of-mass (c.m.) frame product relative translational energy distribution for CH3 + CF31 + CH31 + CF3, E, = 12 kcal mol-'. ( b ) CH,I c.m. frame angular distributions; see fig. 1. The high average product translational energy, ( E ' / EaVI)=O.66, and sharp backward scattering of the CH31 product indicates that a roughly collinear CH,-I--CF3 geometry is favoured. The fraction of available energy that is channelled into translation is greater than that for the CH, + IY + CHJ + Y (Y=Cl, Br, I ) reactions, where (E'/Eavl) = 0.3,' suggesting that the CH3 + CF31 potential-energy surface is more repulsive in its exitGeneral Discussion 91 valley than the CH3-I-Y surfaces. In addition, the vibrational modes of the CF3 group appear to play a very limited role in product energy partitioning for this reaction.This may be partly due to the similarity in structure of the CF3 groups in CF31 and CF;. Finally, these results are remarkably similar to those obtained by Davidson et a1.2 for the reaction D + CF31 + DI + CF3 where the DI product was strongly backward scattered and ( E ' / EavJ = 0.7. 1 ( a ) J. A. Logan, C. A. Mims, G. W. Stewart and J. Ross, J. Chem. Phys., 1976,64,1804; ( b ) L. C. Brown, J. C. Whitehead and R. Grice, Mol. Phys., 1976,31, 1069; ( c ) S. M. A. Hoffman, D. J. Smith, N. Bradshaw and R. Grice, Mol. Phys., 1986, 57, 1219. 2 F. E. Davidson, G. L. Duncan and R. Grice, Mol. Phys., 1981, 44, 1119. Prof. R.Grice ( University of Manchester) said: Does Mr Robinson conclude from the rebound distribution observed in his measurements on the CH, + CFJ exchange reaction that there is no significant well on the potential-energy surface corresponding to a stable CH3-I-CF3 radical intermediate? Mr G. N. Robinson, Dr G. M. Nathanson, Mr R. E. Continetti and Prof. Y. T. Lee ( University of California, Berkeley, CA) said: For polyatomic reactions, forward-back- ward symmetry in the product angular distribution and a product translational energy distribution that peaks near Okcalmol-' are usually taken as evidence of a bound collision complex. From a study of the reaction F+CH,I -+ IF CH,, Farrar and Lee' infer that CH3-I-F is bound by ca. 25 kcal mol-' with respect to F+CH,I. Likewise, reactive scattering studies of the reactions CH, + IY -+ CH31 + Yza and Y + CH31 + IY + CH32h (Y=Br, C1) suggest that the CH,-I-Cl and CH,-I-Br intermediates are slightly bound.Thus it appears that one can correlate the I-Y attraction on a CH3-I-Y potential-energy surfa.ce with the electronegativity of the Y atom. The low electronegativ- ity of the CF3 group relative to the halogen atoms and the large repulsive energy release and strong backward scattering of the CH,I product from the CH,+CF31 reaction suggest that CH3-I-CF3 is not bound by much energy, if at all. 1 J. M. Farrar and Y. T. Lee, J. Chem. Phys., 1975, 63, 3639. 2 ( a ) J. A. Logan, C. A. Mims, G. W. Stewart and J. Ross, J. Chem. Phys., 1976, 64, 1804; L. C. Brown, J. C. Whitehead and R. Grice, Mol.Phys., 1976, 31, 1069; S. M. A. Hoffman, D. J. Smith, N. Bradshaw and R. Grice, Mol. Phys., 1986, 57, 1219; ( b ) D. Krajnovich, 2 . Zhang, F. Huisken, Y. R. Shen and Y. T. Lee, Physics of Electronic and Atomic Collisions, ed. S . Datz (North-Holland, Amsterdam, 1982), p. 733; S. M. A. Hoffman, D. J. Smith, A. Gonzalez Urefia, T. A. Steele and R. Grice, Mol. Phys., 1984, 53, 1067. Prof. M. T. Bowers (University of California, Santa Barbara, CA) addressed Dr McKendrick. First, is the velocity distribution of the reactant O('P) atoms known? What effect would significant changes in this velocity have on the reaction dynamics? Secondly, have you considered calculating product rotational distributions using statistical methods simply by assuring rigorous conservation of the system angular momentum? This motion could well be decoupled from the clearly non-statistical product vibrational distribution.Mr H. Rieley (University of Bristol) said: McKendrick et al. have presented the internal energy distribution for product OH from the simple chemical reaction O('P) + HBr. They were able to infer the opening of a reactive channel leading to production of the electronically excited partner Br(2PP,,2) state, from an anomaly in the OH(X 2H, u"= 1) rotational distribution. I would like to add to the list of novel diagnostics, given by Prof. Polanyi in his Spiers Memorial lecture, and demonstrate how velocity-aligned Doppler spectroscopy (VADS) could be a powerful tool in the determi- nation of translational energy distributions of products from a bimolecular reaction (without the use of conventional time-of-flight techniques).Energy conservation may92 General Discussion then yield information about the distribution over energy states of a dark partner product, e.g. Br in the above reaction. Wittig et al. have pioneered the VADS technique measuring the recoil velocity distribution of photofragment H atoms from various precursors.’ By delaying the probe radiation following photolysis, fragments with velocity components perpendicular to the probe wave vector escape from the probe region and remain undetected. For a narrow velocity distribution the Doppler profile of the Lyman-a line of the H photofrag- ment collapses to leave two outer wings, representing the speed along the probe direction.It is necessary to use suitably collimated beams and work at sufficiently low pressures to ensure a collision-free environment over the timescale of the experiment. A structured Doppler profile obtained under these conditions directly reflects the population distribu- tion of the partner. This has been tested on the vibrational excitation of SH from photolysis of H2S.’ I wish to report here the first use of VADS on a species other than atomic hydrogen (namely OH), using laser-induced fluorescent (LIF) detection. Presented in fig. 3 are some results obtained from a study of the photodissociation dynamics of nitrous acid (HONO) at 355 nm.3 They show the evolution of a single OH(0,O) line profile with increasing delay between pump and probe lasers. The change in Doppler profile is quite dramatic as the width of the angular distribution of recoiling OH sampled by the probe becomes smaller. No further significant narrowing of the profile was seen for t > 800 ns. Resolution is limited by both geometrical and molecular factors including, of course, the desired internal energy distribution of the partner NO fragment.In the work of Wittig and coworkers on photofragment H from HBr, the presence of both spin-orbit states of Br was resolved in the Lyman-a Doppler profile using a conventional dye laser with intracavity etalon.* In our own study, photofragment OH has recoil velocities an order of magnitude smaller than those of H, and the overall widths of the OH profiles are correspondingly narrower. As a result, it was necessary to improve the resolution of the probe radiation.This was achieved using a pulse- amplified C.W. dye laser system to give a linewidth of ca. 100 MHz in the u.v., providing a negligible contribution to the overall lineshape. The long delay profile reveals a narrow distribution in OH translation and indicates a narrow distribution in NO internal energy. It is possible to detect the NO fragment, also in LIF, and probe the internal distribution directly; and this has been done. Fig. 4 shows rotational distributions obtained for NO(X 2KI, uf’ = 2 and 3). They were obtained by reduction of representative branches from the one-photon LIF excita- tion spectra of the NO[(A ’6+, v’= 0)-(X 211, u”= 2 and 3)] bands. Tunable U.V. at the required wavelength was generated by Raman-shifting, in H2, the doubled output of a Nd : YAG pumped dye laser.The distributions are approximately Gaussian in rotational quantum number, J: the v ” = 2 distribution peaks at ca. J = 26 with a half-width of ca. 15 and is the most highly populated, while that of u”= 3 peaks at ca. J = 21 with a similar half-width. NO in u” = 1 is partly obscured by photofragment NO from dissociation of NO2 at 355 nm. However, population in levels higher than those allowed by the available energy for this latter process may be attributed to NO from photodissociation of HONO, and this shows a similar J-dependence, peaking around J=30. Any nascent NO in v” = 0 is completely obscured by a background of thermal NO. No population in u” = 4 or 5 (the limit of available energy) was observed above the noise level.The main point to note is that each rotational distribution is narrow. As the peak value of J decreases with increasing vibrational quanta the distributions overlap significantly in total energy, resulting in a narrow distribution of internal energy overall. It is now possible to compare these two independent observations of NO internal energy. Fig. 5 is an enlargement of the 1 ps delay spectrum from fig. 3, and marked along the Doppler shift axis are the positions corresponding to the total energies of the peaks in the NO( or‘= 2 and 3) distributions. They agree favourably with the maxima in the observed OH translational energy distribution.General Discussion 93 i - 0 . 5 0 -0.25 0.00 0 . 2 5 0 . 5 0 Doppler shift/cm-’ Fig.3. Doppler profiles of the P,(2) line of OH[(A ’C+, u’=O)-(X 2rI, u ” = O)] at various delays between photolysis and probe. The delays are: ( a ) 0, ( b ) 200, ( c ) 600 and ( d ) 1000 ns, with photolysis at 355 nm. In general, VADS should be a useful diagnostic method when there is a narrow internal energy distribution in a heavy recoiling partner. Consider probing the OH product from the reaction: O( ’ P ) + HBr. The partner Br is heavy (mass=80 a.m.u.) and has no internal degrees of freedom apart from the spin-orbit splitting energy of 3685 cm-’. Therefore, at the threshold for production of the higher energy ’ state, the difference in Doppler shifts for a line in the OH( 1, 1 ) band will be ca. 0.24 cm-I. This may be resolvable even with a conventional dye-laser system.In photodissociation, the loss of resolution in a Doppler profile from VADS is partly due to the internal energy distribution in the recoiling partner, the thermal motion of the parent, and the finite size and profile94 General Discussion Fig. 4. Nascent NO rotational distributions, P ( J ) , from the photodissocation: 355 nm HONO(u,=2) - OH(X *n)+NO(X *11, u"). ( a ) v"= 2, R**(J). ( b ) u"= 3, Q * , / R , I ( J ) . of the laser beams. For a chemical reaction there may be further averaging due to a spread of reactive collision energies. The latter is the main source of complication for the reaction under discussion as the collision energy depends on the internal energy distribution of NO from the photodissociation of NOz, the Frecursor of O( ' P ) . However, the energy available from this process is ca.3000 cm-', and the range of reactive collision energies should not obscure the partner product distribution completely. It should therefore be possible to distinguish the spin-orbit states of product Br in the translational energy spectrum of OH and, perhaps more interestingly, correlate the Br spin-orbit branching ratio with OH internal energy states by performing VADS on different transitions. The feasibility of applying VADS to molecular products from a chemical reaction has been demonstrated. It is easy to envisage a number of simple chemical systems for which VADS could be used to determine cross-correlations between the quantum states of two molecular products. Indeed, a significant improvement may be gained over the experiment described above, by using a collimated molecular beam source, crossed at right-angles by the laser beams, to minimise any parallel velocity component from parent molecules, thereby further improving the energy resolution. 1 2.Xu, B. Koplitz, S. Buelow, D. Baugh and C. Wittig, Chem. Phys. Lett., 1986, 127, 534. 2 2. Xu, B. Koplitz and C. Wittig, J. Chem. Phys., 1987, 87, 1062. 3 H. Rieley, PhD Thesis (University of Bristol, 1988).General Discussion 95 - 0.50 -0.25 0.00 0.25 0.50 l Doppler shift/cm-’ Fig. 5. An enlargement of the 1 ps delay Doppler profile of the P,(2) transition: ( a ) and ( b ) are the Doppler shifts which, by energy conservation, correspond to the peaks in the rotational distributions of NO, u” = 2 and 3, respectively.Prof. S. R. Leone ( University of Colorado, CO) said: There are many different mechanisms that can lead to bimodal rotational distributions. We have seen several of these, and it is useful to formulate a more complete list. Different rotational distributions can be the result of different macroscopic product branches, such as suggested in the paper by Zare for the Br(2P,,2) and Br(’P,,,) states in the O+HBr reaction. Also possible are the direct versus migratory pathways (microscopic branches) referred to in the paper by Grice. We have learned many details of bimodal rotational distributions in the photofragmentation of bent versus linear states. In addition, it is important to consider the possibility that different rotational sub-branches could result from various spin-orbit states in the reagents.For example, the O( 3P2,,,0) states in 0 + HBr could lead to spin-orbit-selective reactivity and selective rotational population groups. In atoms such as O(3P2), the possibility also exists for orbital alignment effects, which may connect to specific rotation states. It has already been pointed out in the paper by Zare that the final OH spin-orbit states may be connected in a selective manner to the O( 3P2,,,0) reagent states. Tunnelling effects might also lead to different rotational states. Finally, in ion-molecule systems, potential surface hopping on the inward or outward parts of the trajectories may result in different rotational states from the varying topologies in the exit part of the trajectory.Prof. L. Holmlid and Dr P. A. Elofson (University of Giiteborg, Sweden) said: We wish to comment on the experiments on O( ’ P ) + HBr by McKendrick, Rakestraw and Zare. We have applied our statistical simulation algorithm to the calculation of the vibrational energy distributions for this reaction, using the experimental conditions in our calculations. The procedure we use simulates the reaction for each of a large number of colliding, reacting molecules and distributes the energy in a unimolecular (RRKM) fashion over the pertinent energy terms (‘boxes’). In this case, four boxes were used: molecular vibration and rotation, orbital motion and relative translation. Angular momentum and reactive flux are conserved strictly. Angular-momentum conservation restricts the number of available configurations at the transition states (centrifugal barriers) severely.In the present calculation, quantization of vibrational energy is96 General Discussion 0 20 40 J 0 20 40 J Fig. 6. Distributions of rotational quantum number J for u’ = 1 (a) and 2 (b). Triangles are experimental results from McKendrick et al. included also in the product channel, i.e. for OH in the form of a ‘window’ technique. In this way only simulated systems with vibrational energies within small limits around each vibrational energy level (ca. 1% of h o ) are allowed to move into product space. Systems that fail to meet this criterion go through a new energy distribution calculation. In this way, all complexes are forced into small energy bands around the quantum- mechanically allowed vibrational levels.Our result for the rotational J distributions, in fig. 6, match the experiments very well. At low Jrot and v’= 1, the discrepancy is probably due to Br* formation as concluded by Prof. Zare’s group. Rotational distributions of this type have been found for other systems as well. Our previous calculations192 on systems like O( ‘ D ) + H2, HCl show good agreement with experiments also in this respect. The rotational inversion is due to the angular-momentum restrictions, as shown by our results using this purely statistical theoretical description. In the case of vibrational energy descriptions, we can match the experiments by arbitrarily introducing a variation of the ‘window’ width with quantum number. Without such a procedure our calculations still give a large number of systems in IJ’ = 0, as seen in fig.7. We propose that the ‘window’ size variation is due to differences in the coupling of the orbital and rotational angular momenta at the transition state. For small v’, rotational energy is large, and the angular momenta are almost at right angles to each other, thus giving a small anisotropy of the potential between Br and OH and a small broadening of the OH vibrational levels. For large v’, the angular momenta become more parallel, the anisotropy in the Br-OH interaction increases and so does the broadening of the OH vibrational levels. We would like to conclude about this reaction as follows: ( 1 ) rotational distributions are statistical, even if they are inverted; ( 2 ) vibrational distributions may be statistical when the vibrational level ‘broadening’ proposed above is included; (3) even short-lived complexes like OHBr may decompose statistically.1 P. A. Elofson, K. Rynefors and L. Holmlid, Chem. Phys., 1985, 100, 39. 2 K. Rynefors, P. A. Elofson and L. Holmlid, Chem. Phys., 1985, 100, 53. Prof. D. W. Setser (Kansas State University, K S ) said: Zare and coworkers have measured the vibrational and rotational distribution for OH from the O( ‘ P ) + HBr reaction. They discuss the results in terms of the kinematic constraints imposed by the H + LH mass combination. They also suggest that the rather high rotational distribution found for OH(u=l and 2 ) could be associated with repulsive energy release from a bent O-H-Br configuration in the exit channel.We have utilized exoergic H abstraction reaction by O( ‘ P ) atoms to investigate’ the H + LH kinematic constraintsGeneral Discussion 0.0 0.0 10.0 20.0 Ek/ 10-20J 0.0 I r 10.0 20.0 E k/ 1 O-*'J 0.0 10.0 20.0 E &/ 1 O-*'J 0.0 0.0 10.0 20.0 E+/ 10-20J 10.0 20.0 E k/ 1 OP2OJ 10.0 20.0 E:/IO-~OJ 97 Fig. 7. Energy distributions for the OH products, [ ( a ) translational, ( b ) rotational and ( c ) vibrational] using a vibrational level width of 0.5% of hiw to the left. To the right, the level width has been varied to reach approximate agreement with experiments. Rotational J distributions for o' = 1 and 2 are unchanged during this process. to vibrational energy disposal; the results can be compared to F atom reactions for HI, GeH,, SiH, and SeH, as reagents.'.' Our experiments' were done in a fast flow reactor in which the product vibrational states were observed after ca.0.2 ms of reaction time. The 0 atoms were produced by microwave discharge in Ar. Laser-induced fluorescence was used to measure the ratio of the OH( u=O) to OH( u= 1 ) concentration, and infrared chemiluminescence was used to measure the relative populations of OH( u= 1, 2 and 3). The vibrational distributions are summarized in table 1, and (f,(OH)) is compared with (fJHF)). No information was obtained about rotational, spin-orbit or A-doublet populations of OH, since these levels exhibited Boltzmann distributions for our operating conditions of 0.5 Torr and 300 K. All OH vibrational distributions are inverted, with the OH( u, J ) populations extend- ing to the thermochemical limits.The vibrational energy disposal to OH(v) is very similar to that for HF(u) from F-atom reactions, even to the extent of having the same linear surprisal plots.3 This is true in spite of the much smaller (two orders of magnitude)98 General Discussion Table 1. Vibrational energy disposal summary O+HI 35.4 13 11 34 42 0.56 0.59 0 + GeH, 28.4 18 31 47 03 0.47 0.59 0 + SiH4 17.0 19 81 0.49 0.52 0 + SeH, 30.3 24 36 40 0.38 0.48 rate constants for the O-atom reactions. Thus the kinematic constraints to vibrational energy disposal imposed upon direct H abstraction by O('P) atoms is confirmed. We agree with Zare and co-workers in that the O( ' P ) reactions with H-containing molecules show no indication of a component proceeding via non-adiabatic pathways to a lower- energy singlet surface corresponding to bound singlet molecules.We interpreted the rather high OH(v=O)/OH(v=l) ratio from O+HI as possibly being the consequence of I( ' P l 1 2 ) formation.' If the higher-energy spin-orbit state is formed, energy constraints limit the OH product to v 5 1. The mechanism for I( formation could be V-E energy transfer in the exit channel, as advocated for Br( *Pl12) formation in the F+HBr In our work the ratio OH(v=O)/OH(v=l) was measured as a function of added reagent, and the ratio was invariant. Nevertheless, residual OH(u=O) emanating from the microwave discharge source of 0 atoms would make the measured ratio an upper limit to the true ratio from the chemical reaction.The rotational energy disposal from F atom reactions6" with HCI, HBr and HI can be examined for insight with respect to the 0 + HBr reaction. Each HF( u ) level tends to have approximately the statistically expected mean rotational energy when examined in terms of the reduced variable g R =fR/( 1 -fv), and the overall (gR(HF)) are 0.3-0.4.'.' Quasi-classical trajectory calculations for LEPS type surface having the lowest barrier for collinear geometry give qualitatively similar rotational energy disposal.' For the polyatomic hydride molecules, (fR( HF)) seems to increase as the reaction cross-section becomes larger, until the limit just mentioned for the diatomic molecules is reached.' Thus I would like to propose a question to Prof. Herschbach and Zare (and others) with respect to kinematic constraints to the rotational energy disposal for the H+LH mass combination. Will reduced reactive cross-sections (with increased barrier heights) make significant changes in the expected kinematic constraints to rotational energy disposal? Stated in more specific terms, is a bent configuration in the exit channel required to explain the rotational energy release for the O + HBr reaction vis-Ci-uis comparison to the F-atom reactions, which presumably proceed via collinear geometry? 1 B.S. Agrawalla and D. W. Setser, J. Chem. Phys., 1987, 86, 5421. 2 B. S. Agrawalla and D. W. Setser, J. Phys. Chem., 1986, 90, 2450. 3 B. S. Agrawalla and D. W. Setser, in Gas-phase Chemiluminescence and Chemi-ionization, ed. A. Fontijn 4 J.P. Sung and D. W. Setser, Chem. Phjx. Lett., 1977, 48, 413. 5 J. W. Hepburn, K. Liu, R. G. Macdonald, F. J. Northrup and J. C. Polanyi, J. Chem. Phys., 1981,75,3353. 6 K. Tamajake, D. W. Setser and J. P. Sung, J. Chem. Phys., 1980, 73, 2203. 7 P. Beadle, M. R. Dunn, N. B. H. Jonathan, J. P. Liddy and J. C. Naylor. J. Chem. SOC., Faraday Trans. (Elsevier, Amsterdam, 1985). 2, 1978, 74, 2170. Prof. D. R. Merschbach (Hurvard University, MA) (communicated): Prof. Setser invited comment on kinematic constraints for the rotational energy disposal in reactions having the heavy C light-heavy mass combination, such as O+ HBr. In particular, he asked ( 1 ) whether reduced reaction cross-sections resulting from high barriers might significantly alter the constraints and (2) whether a bent configuration in the exit valleyGeneral Discussion 99 is required to account for the high product rotational excitation found for 0 + HBr.He noted that this high excitation resembles that for F+HBr, which is presumed to react collinearly. I think the answer to ( 1 ) is yes. For instance, a high barrier might confine reaction to small impact parameters and thereby weaken the propensity for approxi- mately equal reactant and product orbital angular momenta. I think the answer to (2) is no in principle but quite likely yes in practice. The ‘in principle’ aspect is akin to the situation described in our paper, in which high product rotational excitation is fostered by antiparallel alignment of orbital and rotational momenta. For instance, suppose j can be neglected and an entrance barrier renders the typical I unusually small.Then even when the reaction involves no change in reduced mass, a large exit velocity or impact parameter can make the typical I’ substantially larger than I and thus yield high product rotation, j ’ = I - I’. The ‘in practice’ aspect is exemplified in several studies cited by McKendrick, Rakestraw and Zare and by Amaee, Connor, Whitehead, Jakubetz and Schatz in their papers. Also especially pertinent are recent trajectory calculations by Persky and Kornweitz.’ They compare results obtained for the C1+ HC1 reaction with three semi- empirical LEPS potential-energy surfaces, designated 1-111. For each, the preferred reaction configuration is collinear. The activation barrier is large; its height is the same for the collinear configuration (8.55 kcal mol-’) but increases more steeply for bent configurations as surface I + I1 -+ 111.Despite this, the reactive cone of acceptance increases considerably in the same order. That happens because surfaces I1 and par- ticularly I11 exert strong anisotropic attractive torques which pull the reactant molecule into a nearly collinear orientation even when the initial approach angle is unfavourable. As in previous studies, the product rotational excitation reflects repulsive torques in the exit valley. For surface I11 a nearly collinear configuration usually persists well into the exit trajectory, and the products emerge with only low rotational excitation. For surface I, although the critical reactive configurations are likewise nearly collinear, strongly bent configurations are attained early in typical exit trajectories.In these bent configurations, repulsion sets the light hydrogen atom spinning and produces high rotational excitation. Collinear reaction in the entrance valley and saddle-point regions is a property distinct from repulsive torques in the exit valley. Until we can reliably derive these forces from electronic structure, our diagnostic criteria remain provisional. However, in view of the special kinematics for heavy+light-heavy systems and the extensive trajectory studies, high product rotational excitation offers strong evidence that repulsion occurs in bent configurations early in the exit valley. This seems likely for the reactions of both F atoms and 0 atoms with hydrogen halides.1 A. Persky and H. Kornweitz, J. Phys. Chem., 1987, 91, 5496. Dr E. J. Kruus, Dr B. I. Niefer and Dr J. J. Sloan (NRC, Ottawa, Canada) said: We have measured the product state distributions from the corresponding reactions of O( ‘ D ) with HC1 and HBr, using a novel implementation of time-resolved Fourier transform spectroscopy (TRFTS) to observe the infrared emission spectra of the products. O( ’ D ) atoms were created by pulsed laser photolysis of ozone. The experiments were carried out at total reagent pressures (ozone and HCI or HBr) of a few mTorr. The submicrosecond time-resolution of the TRFTS technique permits the product emission spectra to be observed at approximately each gas-kinetic collision after the O( ‘ D ) atom is created.In this way, both the unrelaxed (initial) product energy distributions and the energy transfer pattern resulting from the collisional deactivation of the vibrationally excited product can be observed. Emission spectra from both OH and HCl created by the reaction, recorded at ca. 2, 10 and 30 gas-kinetic collisions after the ozone photolysis, are shown in fig. 8. From the measured spectral intensities and the known Einstein transition probabilities, the100 General Discussion O('D) + HCI 2300 2400 2500 2600 2700 2800 2900 3000 3iOO 3 i O O 3iOO 3400 3600 3600 wavenumber/cm-' Fig. 8. Infrared emission spectra from the unrelaxed OH( u ' ) and HCl( u ' ) products of the reaction: O(' D ) + HCl, recorded using fast time-resolved Fourier transform spectroscopy. rD = ( a ) 20, ( h ) 100 and (c) 300ps.ratio of the cross-section for reaction: O( ID) + HC1+ OH( v') + C1 to that for E-V energy transfer producing O( ' P ) + HC1( v'), is found to be 0.95-0.05. The initial vibra- tional distributions in both channels are strongly inverted; but they relax rapidly in subsequent collisions. Fig. 9 ( a ) and ( h ) show, respectively, the time dependences of the OH( u') and HC1( v') distributions. Details of these will be presented elsewhere.' The reaction and the E-V energy transfer process channel 64 and 62%, respectively, of the reaction exoergicity into product vibration. Very similar results were obtained for the O( 'D)+HBr reaction, which creates both OH and HBr with strong vibrational inversion as well.The singlet reactions occur on electronic surfaces which are topologically quite different from those of the triplet reactions. The O( 3P)/HX (X=C1 or Br) surfaces are monotonically exothermic, having no minima; whereas the singlet surfaces for both reactions studied in our work correlate with strongly bound intermediates of the form H-0-X. As reported in the study of McKendrick et al., the triplet reactions are probably simple hydrogen abstractions, having the expected dynamics for the heavy- light-heavy mass combination in direct reactions. If the singlet reactions are insertions, accessing the strongly bound intermediate, the strong inversions observed must be due to dynamical constraints limiting the energy randomization which would be expected to occur during the lifetime of the intermediate. Similar effects have been seen in theGeneral Discussion 101 0.4- 0 .3 - 0.2- 0.1- 0.0 I I I I 1 0 1 2 3 4 5 v level c .M w - 3 a a 0 1 2 3 4 5 6 v level Fig. 9. The time dependences of ( a ) the OH( v ' ) vibrational distribution created in the reaction: O('D)+ HCI --* OH(v') +Cl; and ( b ) the HCI created in the E-V energy transfer: O('D)+ HCI- HCl(v')+O(3P). (Z,,) is the average number of gas-kinetic collisions suffered by the product molecules at the time of observation: 0, 4; 0, 6; A, 13 and +, 18. case of the O( lD)/H2 reaction, for which it has been analogous H-0-H intermediate is too short to permit energy randomization. that the lifetime of the 1 E. J. Kruus, B. I. Niefer and J. J. Sloan, J. Chem. Phys., 1988, in press.2 P. M. Aker and J. J. Sloan, J. Chem. Phys., 1986, 85, 1412. 3 P. J. Kuntz, B. I . Niefer and J. J . Sloan, J. Chem. Phys., 1988, in press. Prof. D. G. Truhlar ( University of Minnesota, M N ) said: Based on earlier theoretical studies employing classical trajectories, McKendrick, Rakestraw and Zare present an interesting interpretation of the inverted product vibrational distribution for the reaction102 General Discussion 0 + HBr -+ OH + Br in terms of trajectories in which the H atom is transferred at relatively large internuclear distances. Another possibility is evident from our earlier semiclassical calculations on the similar exothermic reaction C1+ HBr + C1H + Br on a model poten- tial-energy surface.’ In that study we found that reaction under room-temperature thermal conditions was dominated by tunnelling, and the product population inversion arose naturally as a consequence of shorter tunnelling paths in the mass-scaled coordinate system for tunnelling into vibrationally excited product channels.Since the 0 + HBr -+ OH + Br reaction has a significant classical barrier, and since hydrogen-atom transfer reactions involving a significant barrier are usually dominated by tunnelling,’ I would suggest that this vibrationally assisted tunnelling mechanism probably controls the dynamics under room-temperature thermal conditions, and further interpretation of the details of the product distribution should include this mechanism, at least under such low-energy conditions. 1 B. C. Garrett, N. Abusalbi, D.J. Kouri and D. G. Truhlar, J. Chem. Phys., 1985, 83, 2252. 2 R. T. Skodje, D. G. Truhlar, and B. C. Garrett, J. Phys. Chem., 1982, 77, 5955; D. G. Truhlar and B. C . Garrett, Annu. Rev. Phys. Chem., 1984, 35, 159; J. Chim. Phys., 1987, 84, 365. Dr K. G. McKendrick (Stanford University, CA) replied: Prof. Bowers has inquired about the O( ’ P ) atom velocity spread following the 355 nm photolysis of NO2, and the effect that this might have on our observations. As briefly discussed in our paper, two previous studies have been reported of energy deposition in the fragments of NO2 photolysed at wavelengths close to 355 nm [347 and 337 nm; ref. (40) and (41) of our paper, respectively]. These studies, although not directly comparable given the slightly different wavelengths, are in contradiction over the partitioning of available energy between translational and internal degrees of freedom.{The time-of-flight measurements [ref. (40)] imply a greater proportion of energy appearing as fragment translation than measurements [ref. (41)] of the internal state populations of NO.} Although no equivalent measurements have been reported for photolysis at 355 nm, it is clear that in our experiment there will be a relatively broad distribution of O(’P) velocities and hence 0 + HBr collision energies, extending to the approximate energetic limit (ca. 25 kJ mol-I) derived in our paper. We also discuss in our paper some general aspects of the dynamics of reactions with the heavy + light-heavy (H + LH’) mass combination, the kinematic category to which 0 + HBr belongs.A number of generalisations have been derived from several indepen- dent studies. The propensity [expressed in eqn (4) of our paper] for translational energy to be approximately conserved during the reaction, which was reproduced in our quasi-classical trajectory (QCT) calculations on the O+ HBr system using the LEPS surface derived by Persky and coworkers, implies that a distribution of collision energies will have a relatively minor effect on the measured product internal state distributions. For example, in our QCT calculations, we find that on increasing the collision energy from 15 to 25 kJ mol-’ (corresponding to the range from slightly above the classical dynamical threshold to the limiting collision energy in our experiment) the average internal energy of the products increases by only 1.3 kJ mol-I.We therefore feel that the interpretation of our results is not seriously complicated by the distribution of O(’P) velocities resulting from their method of production. Mr Rieley has suggested an indirect method for the estimation of the branching ratio Br*/Br [ i.e. Br(’P,,2)/Br(2P,,2)] from the reaction O(’P) + HBr. It is, of course, possible to measure this quantity directly by spectroscopic means, e.g. by laser gain/absorption, multiphoton ionisation, or V.U.V. fluorescence. Using the last of these techniques, Polanyi and coworkers’ have determined the Br*/ Br ratio from the closely related reaction F+ HBr + FH + Br, finding ca. 6% branching into the spin-orbit excited state. As discussed in our paper, we have associated the low N” inflection in the reduced OH( V” = 1 ) rotational distribution with the channel leading to Br*.For the purposesGeneral Discussion 103 of estimating this branching ratio, we may assume that the shape of the rotational surprisal plot associated with a single channel is the approximately parabolic function obtained in a previous study of a similar H + LH' system, C1+ HCl (to be discussed later in this meeting2). A similar dependence was also found in our QCT calculations on 0 + HBr, and is consistent with the propensity rule (4) of our paper which predicts that the product state distribution is expected to peak sharply at an energy approximately isoenergetic with the reagents. The difference between the actual experimental surprisal and the extrapolated (postulated) parabolic plot corresponds to ca.6% 'extra' population in the low N" states of OH(u"= 1 ) . This value is interestingly similar to the 6% Br" branching in the F+ HBr system.' Prof. Leone has remarked on the different processes which can give rise to multimodal product state distributions, and in particular has raised the question of the role of the O( 'PJ ) reagent fine-structure state distribution in determining the product-state fine-structure partitioning. An argument of partial electronic adiabaticity has previously been presented by Andresen and Luntz [ref. (13) of our pa er] as an interpretation of the OH fine-structure state partitioning in the products of O( P ) reactions with organic molecules.However, by variation of the F*/F ratio in the F atom source, Polanyi and coworkers' were able to demonstrate that the Br*/Br ratio was not determined by the reagent fine-structure state distribution in the F+HBr system. It was argued that the Br* branching was controlled by non-adiabatic coupling between surfaces in the exit channel region. In a similar fashion, we have performed a second series of investigations of the 0 + HBr system using microwave discharge of molecular 0, as an alternative method for O('P) generation. An effusive jet of O('P) was crossed by a pulsed supersonic free jet of HBr seeded in a carrier gas (generally He). The O('PJ) fine-structure state distribution in the case of microwave discharge generation wiil be that of a thermalised sample (as a result of many downstream gas-phase and wall collisions).Unfortunately, the O('P,) distribution following 355 nm photolysis of NO, is unknown. However, it would be highly coincidental if the distributions from the two methods of production were to be identical. Nevertheless, we observe that the OH 2rI,,2/2r13,2 ratio is the same, within experimental error, in the complementary measurements. It seems therefore most likely that this aspect of the product-state partitioning is not determined by the reagent- state distributions. Similarly, it would be interesting to assess any dependence of the Br*/Br ratio on the method of 0-atom production. However, we were experimentally prevented from making this determination because, in the eff usive beam/ supersonic jet experiments, the low rotational state populations of OH were perturbed by a contribution from reaction with HBr van der Waals clusters formed in the free-jet expansion.' (Recall that we are presently only postulating the occurrence of Br" production by an indirect analysis of the OH rotational distribution.) An interesting extension to our work would clearly be to examine in more detail the dependence of the product fine-structure state partitioning on controlled reagent fine-structure populations.P I turn now to Prof. Holmlid, Bowers and Setser. There have been several comments concerning the relationship of our experimentally observed product state distributions to those expected from statistical considerations. We have not considered that a statistical approach to the interpretation of the data would be the most fruitful, primarily because of the very heavily inverted OH vibrational distribution (which is contrary to the concepts of statistical energy partitioning).There- fore, we have approached the problem from a dynamical standpoint, implicitly assuming that the product state attributes are deterministically related to the properties of the potential-energy surface describing the three-body interaction. The quasi-classical trajec- tory calculations which we have performed on the potential surface derived by Persky and coworkers, with initial conditions selected in the usual Monte Carlo fashion from104 General Discussion Table 2. Calculated and measured moments of the distributions exptl 0.51 0.24 (0.25) QCT 0.52 0.22 0.26 I60 -- 140-- g 120- .- a K 80.- 60- 40.- 0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 i” Fig.10. Rotational distribution in OH( zl” = 1) predicted by quasi-classical trajectory calculations on the reaction O(3P) + HBr, assuming a uniform collision energy distribution over the range 15-25 kJ mol-’. distributions designed to simulate our experiment, have yielded results in very satisfac- tory agreement with the experimental observations. The calculated moments of the distributions are particularly close to the measured values, as shown in table 2. The detailed distributions are in good qualitative but not perfect quantitative agree- ment: this is perhaps unsurprising for the vibrational distribution, where the quasi- classical ‘post hoc’ quantisation of the distribution is highly arbitrary, given that the product energy space spans only ca.two vibrational quanta. Although the QCT calcula- tions correctly predict the heavy vibrational inversion of OH( urr = 1 ) over OH( u” = 0), they fail to reproduce the experimentally observed 10% branching into u”=2. The calculated OH( u” = 1 ) rotational distribution (treating OH as a ‘C molecule) correctly shows a peak at relatively high j ” , but at 1 or 2 quanta below that of the experimental distribution, as shown in fig. 10 (cJ: fig. 3 of our paper). We therefore feel that a dynamical interpretation provides a satisfactory rationalisa- tion of both rotational and vibrational product attributes. It is interesting that Prof. Holmlid’s calculations, using a statistical algorithm which incorporates rigorously the conservation of total angular momentum, reproduce well the OH rotational distributions from the O+ HBr reaction.Perhaps this is indicative that this aspect of the product state partitioning (in this particular case) is controlled primarily by kinematic rather than authentic dynamical influences. Setser has further. commented on the similarity between the H atom abstraction reactions of F and 0 atoms with certain reagents, and questions the role of repulsive energy release in determining the product state distribu- tions in these systems. One might also consider related systems which share common kinematics, but which exhibit dramatically contrasting dynamics, e.g. the respectiveGeneral Discussion 105 rotationally 'hot' and 'cold' OH distributions formed in the reactions of O('P) with HBr and organic molecules.These differences can be rationalised in a dynamical picture3 in terms of the constraint towards collinearity in the controlling potential surface, which markedly affects the degree of rotational excitation of the products. Is a statistical model capable of reproducing such contrasting behaviour? More generally, what physical interpretation should be attached to the ability of statistical modelling to rationalise partially the observed product state attributes (rotational but not vibrational distributions)? Finally I reply to Dr Sloan and Prof. Truhlar. Dr Sloan has presented experimental results for the reactions of electronically excited O ( ' D ) with HBr (and HCl), and has remarked on the qualitative similarity of the product state energy partitioning in the reactions of O('P) and O( ' D ) with HX molecules.As he points out, the O( '0) reactions are considered to proceed on the ground-state singlet surface via transitory insertion of the 0 atom into the HX bond. Reaction is still direct, however, in the sense that the HOX intermediate survives for a time much shorter than that required for energy randomisation to take place and for corresponding 'statistical' (a term perhaps to be used rather cautiously) product state distributions to result. Whilst we are able to reproduce our observations by performing trajectory calculations on a surface which restricts reaction to direct abstraction, it is well known that it is not possible uniquely to invert product state distributions to recover the form of the potential surface.We are therefore not able to exclude rigorously the possibility that the O('P) + HBr reaction alternatively proceeds via O-atom insertion into the HBr bond, involving a non-adiabatic curve-crossing from the initial triplet surface. Insight into alternative mechanisms for these reactions would certainly be aided by the calculation of realistic ab initio potential-energy surfaces (ideally of both multi- plicities and for all fine structure states), a development which we would like to encourage strongly. Prof. Truhlar has commented on a possible role for tunnelling in this light-atom transfer reaction. Under the slightly superthermal conditions of our experiment, a significant proportion of collisions occur at energies above the classical barrier height, tending to reduce the importance of the contribution to reaction from tunnelling.More generally, the barrier height on a potential-energy surface derived from the comparison of classical calculations with experiment will be correspondingly underestimated to compensate for the extent to which tunnelling is important. It would be interesting to determine whether significant differences are predicted between the product state distri- butions of a more rigorous calculation, including the effects of tunnelling, and those of a quasi-classical calculation with an artificially reduced barrier height. 1 J. W. Hepburn, K. Liu, R. G. Macdonald, F. J. Northrup and J. C. Polanyi, J.Chem. Phys., 1981,75,3353. 2 B. Amaee, J. N. L. Connor, J. C. Whitehead, W. Jakubetz and G . C. Schatz, Farday Discu.ss. Chetn. Soc.,1987 84, 387. 3 Full details of further experimental investigations and QCT calculations will appear in a forthcoming, paper: K. G. McKendrick, D. J. Rakestraw and R. A. Zare, J. Phys. Chern., in press. Prof. P. J. Dagdigian (The Johns Hopkins University, Baltimore, MD) and Prof. M. H. Alexander (University of Maryland, M D ) said: There has been considerable confusion in the past about the symmetry properties of A doublet levels in I1 electronic states. Preferential population of A doublet components in collision phenomena can provide incisive information on the dynamics of these processes. Several years ago we' presented a detailed exposition of the reflection symmetry of the electronic wavefunction for molecules in n electronic states, concentrating in particular on '11 and 'n states.We showed the relationship between the total parity, i.e. the e and f labels, the spectroscopic branch employed and the reflection symmetry of the A doublet level probed. More recently we have extended this analysis in detail for 'Il electronic states.'106 General Discussion For one of the A doublet levels, the electronic wavefunction is symmetric with respect to reflection of the spatial coordinates in the plane of rotation in the high-J limit, while for the other it is antisymmetric. The analysis of the reflection symmetry of only the spatial coordinates of the electrons implies that the same measure of electronic asymmetry can be applied to wavefunctions in both the Hund’s case ( a ) limit, where the spin is coupled to the molecule-frame z-axis, and in the case (6) limit, where the spin is coupled to the space-frame Z-axis.This type of analysis is particularly appropriate to interpreta- tions of reactions and photodissociation processes in which arguments based on the evolution in space of the molecular orbitals of the precursor species are used to interpret the preferential production of a given A-doublet. At present, however, there is no convenient and unambiguous way to designate concisely these symmetry properties. In the literature, the two A doublet levels of a II electronic state are often designated by II+ and II-, but the meaning of these labels is not consistent in all papers.In their contribution to this Discussion, McKendrick, Rakestraw and Zare have proposed a new notation, n( llJ) and T ( L J ) , which describes the orientation of the unpaired p~ electron lobe with respect to the angular momentum vector J. In the former type of level, the pm lobe is parallel to J, and hence perpendicular to the plane of rotation in the high J limit, while for the latter, the p 7 ~ lobe is in the plane of rotation. Subsequent to the writing of the paper of McKendrick et al., we have had discussions with Prof. Zare on this question of the notation of A doublet levels and have agreed upon a notation which we believe is more descriptive for use by scattering-dynamicists. We propose the alternative notation II(A’) and II(A“).The label II(A’) designates A doublet levels whose electronic coordinates are symmetric with respect to reflection of the electronic spatial coordinates in the plane of rotation in the high-J limit, while II( A”) denotes those levels whose wavefunctions are antisymmetric. This notation has the advantage that it focuses upon the reflection symmetry of the A doublet level, which is the property of greatest relevance for collision dynamics. This convention can also be generalized to A states by writing A(A’) and A(A”). The notation ~ ( l l J ) and T ( I J ) could be interpreted as referring to the directionality of the electron distribution in the pn shell, or the expectation value of cos2+ - sin2+. Unfortunately, (cos*+ -sin2+) varies according to whether the p.n shell is less than or more than half filled, i.e.T ’ us. T’.’ The notation II(A’) and II(A”) is independent of the7 electron configuration of the state and applies equally, for example, to molecules in -n electronic states arising from a singly filled T orbital [CH(X’II) or NO(X’II)] and those in molecules with a T’ electron occupancy [OH(X’II) or CN(A2n)]. We would advocate the use of the notation II(A’) and II(A”) to designate the electronic symmetry of A doublet levels of a II electronic state when a concise description of this property is needed. 1 M. H. Alexander and P. J. Dagdigian, J. Chern. Phys., 1984, 80, 4325. 2 B. Pouilly, P. J. Dagdigian and M. H. Alexander, J. Chern. Phys., 1987, 87, 7118. Dr N. C. Firth and Prof. R. Grice (University of Manchester) said: The triplet potential-energy surfaces for the O( ’ P ) + HBr reaction have considerable formal similar- ity to those of the F+ HBr reaction’ as illustrated by the correlation diagram of fig. 1 1 for collinear O-H-Br configurations.The O(’P2) and O(’P,) spin multiplet states correlate with OH(2113/2) + Br(’P,,*) products via OHBr (’IT) intermediates which present only a modest barrier ca. 14 kJ mol-’ to reaction. The 0(’Pp,) and O(’P,) spin multiplet states correlate with the spin-orbit excited products OH(’II,,,) + Br(’P3/’) via OHBr(’C-) intermediates which are expected to present a much higher barrier to reaction. Consequently reaction via the lowest adiabatic surfaces is expected to yield products only in the ground spin-orbit states in analogy to the F+HBr reaction.’ However, calculations on the F + HF potential-energy surface using the diatomics-in-molecules method, including spin-orbit interaction,’ suggest that there will be a high probabilityGeneral Discussion 107 -OH(%,) + Fig.11. Correlation diagram for triplet states of the O(3P)+HBr reaction in the collinear configuration. of non-adiabatic transitions between the sZ= 1,O’ components of the surfaces arising from OHBI-(~II) and OHBr(’Z-) in the exit valley, owing to the small spin-orbit splitting of the OH radical. In the limit of strong non-adiabatic mixing for the R = l , Of com- ponents and no mixing for the R=2,0- components, a ratio 2: 1 is predicted for the populations of the OH(3113/2) and spin multiplet states. This accords not only with the measurements of McKendrick et al.3 but also with previous measurements of’ H-atom abstraction reactions of 0 atoms with hydrocarbon molecule^,^ where similar considerations apply to non-adiabatic transitions in the exit valley of the potential-energy surface.The measurements of Polanyi and coworkers’*2 show a small probability of non-adiabatic transitions to the excited Br(’PII2) spin multiplet state in the exit valley of the F+HBr potential-energy surface and the probability will be similarly small for O(’P) + HBr. The correlation diagram shown in fig. 11 has omitted any reference to the singlet HOBr molecule, which represents the lowest potential-energy surface for this system. The energy of this surface is unknown in the OHBr configuration and the possibility of transition to the singlet surface has been ignored. 1 J.W. Hepburn, K. Liu, R. G. McDonald, F. J. Northrup and J. C. Polanyi, J. Chem. Phvs., 1981,75,3353. 2 N. C. Firth and R. Grice, J. Chem. Soc., Faraday Trans. 2, 1987, 83, 1023. 3 K. G. McKendrick, D. J. Rakestraw and R. N. Zare, Faraday Discuss. Chem. Soc., 1987, 84, 39. 4 P. Andersen and A. C. Luntz, J. Chem. Phys., 1980, 72, 5842. Prof. A. Lagana (University of Perugia, Italy), Dr F. J. Basterrechea (University of Pais Vasco, Spain) and Prof. J. M. Alvariiio (University of Salamanca, Spain) said: Propensity rules tying reactant orbital angular momentum ( L ) to product rotational angular momentum ( J ‘ ) in halide exchanges for collisions of hydrogen halide molecules with an impinging alkali-metal atom are a distinctive feature of the light-heavy-light mass combination type.On these grounds the complexity of the computational procedure for the evaluation of the reactive cross-section of these reactions has been reduced by adopting an approach having built in a L + J’ switch.2 Following this approach quantum reduced dimensionality cross-sections for the H + BrH system have been ~alculated.~ Recent three-dimensional quantum calculations have confirmed the reliability of the pr~cedure.~ We have recently extended these techniques to the Li + HF reaction when performing a 10s calculation of the reactive cross-section.’ For this purpose we had to make the108 General Discussion 0 1 2 3 6 4 2 0 0 2 0 2 4 6 L/ loi4 amu A2 s-’ 4 6 Fig. 12. Correlation diagrams of L and J’ for the Li + MuF reaction calculated at a typical collision energy (15 kcal/mol) for Y = 0 ( a ) , 1 ( b ) and 2 ( c ) .mass of the central atom F ca. one hundred times heavier than its usual value. The factor 10 quoted in ref. ( 5 ) should read 100. Ordinate values of fig. 2 of the same reference should be multiplied by a factor of 12.75. In this way, reactant and product Jacobi coordinates are almost exactly exchangeable, and an analytical match of the entrance and exit arrangement channel wavefunctions can be applied at the dividing surface. A check for the influence of the arbitrary increase of the central atom mass on the Li + H F reactivity has been performed by comparing estimates of the quasiclassical cross-section obtained at different values of the F mass.In contrast to the blocking effect6 found for the collinearly dominated systems, when increasing the weight of the central atom for systems having a bent transition state we obtained a small enhancement of the global reactivity.’ Another important dynamical constraint of the model that needed to be checked before extending the light-heavy-light computational scheme to the Li + HF reaction was the validity of the L-+ J ’ exchange. For this purpose we have carried out classical trajectory calculations for the Li + HF reaction and its isotopic variants. For the lightest member of this family of reactions (Li + MuF) plots of L, J’ pairs associated with reactive trajectories are shown in fig. 12 for the ground and the first two excited vibrational states.Correlation between initial orbital and final rotational angular momenta is fairlyGeneral Discussion 109 good for v =O and excellent for v = 1 and 2. Similar results were obtained for heavier isotopes, although for a mass value equal to or three times larger than that of the hydrogen atom the L + J’ correlation becomes less satisfactory. For an extension of the light-heavy-light computational procedure to the Li + HF reaction, however, the L -+ J’ exchange need not be so tight: it needs only to be enforced at the switching point between entrance and exit channels. The fact that the Li + HF system globally obeys this dynamical constraint may give an opportunity for further simplification of the computational procedure. 1 K. G. McKendrick, D.J. Rakestraw and R. N. Zare, Furuday Discuss. Chern. SOC., 1987, 84, 39; 2 D. C. Clary, Mol. Phys., 1981, 44, 1083. 3 D. C. Clary, Chem. Phys., 1983, 81, 379. 4 Y. Zhang, J. Z. H. Zhang, D. J. Kouri, K. Haug, D. W. Schwenke and D. G. Truhlar, to be published. 5 A. Lagana; and E. Garcia, Chem. Phys. Lett., 1987, 139, 140. 6 J. Manz, and H. H. Schor, Chem. Phys. Lett., 1984, 107, 542. S. K. Kim, and D. R. Herschbach, Faruday Discuss. Chem. Soc., 1987, 84, 159. Dr T. Trickl (University of California, CA) and Dr J. Wanner (MPQ, Garching, Federal Republic of Germany) (communicated): Experiments on the F+ I2 reaction system exhibit the common phenomenon of visible IF chemiluminescence which varies in intensity depending upon the experimental conditions. Very weak visible IF( A - X ) and ( B - X ) emission was observed in connection with the F+ I2 reaction as studied in a single-stage molecular-beam apparatus with two crossed effusive nozzles.In an earlier communication’ we reported a first attempt at spectral resolution of the light emission associated with the F+12 system. There we attributed the spectrum to the reaction of F atoms with the trihalogen radical 12F. The existence of this light-emitting reaction step had been verified in a molecular-beam experiment by Kahler and Lee, yet without recording a spectrum. In their experiment the trihalogen radical was formed by reaction of supersonically seeded F2 above a distinct translational threshold energy of 17.6 kJ mol-’ according to: F2 + I2 -+ 12F+ F. This step, including the subsequent reac- tion of F+ 12F + IF+ IF, obeys bimolecular dynamics and hence proceeds as an elemen- tary reaction.* We were able to extend our earlier work with a more sensitive detection system.’ Hence at a collision energy of ca.3.2 kJ mol-’ it was possible to obtain a vibrational product state analysis of the IF(B) state. At flows of ca. 2 SCCM the overall chemiluminescence signal showed a linear dependence on the iodine flow. The spectrum of the reaction F+12F is depicted in fig. 13. Fig. 14 shows the corresponding IF(B) vibrational product state distribution. The population limit closely corresponds to the enthalpy according to AH; = -248.5 kJ mol ’. The distribution is characterized by a ‘vibrational temperature’ T, =: 800 K. The latter observation is essentially identical with results of Whitehead et al.for a low-pressure 12/F2 flame,4 where the reaction mechanism of Lee et al. is also anticipated. In our experiment the above reaction sequence is most likely induced by the ca. 5% undissociated F2, internally excited in the microwave discharge. 1 T. Trickl and J. Wanner, J. Chem. Phys., 1981 74, 6509. 2 C. C. Kahler and Y. T. Lee, J. Chem. Phys., 1980, 73, 5122. 3 The apparatus used is identical to the one described by M. Trautmann, J. Wanner, S. K. Zhou and C. R. Vidal, J. Chem. Phys., 1985, 82, 693. 4 D. Raybone, T. M. Watkinson and J. C. Whitehead, in Proc. NATO Advanced Research Workshop on Selectivity in Chemical Reactions, Bowness-omwindermere, Sept. 1987, ( Reidel, Dordrecht, in press). Mr G. N. Robinson (University of California, Berkeley, CA) said: Regarding the F+I, crossed-beam results of Firth et al.: (1) The relative contributions of the fast and slow P ( E ’ ) distributions to the IF time-of-flight spectra are not indicated in fig.2 of their paper. How were the relative magnitudes of the forward and backward components of the CM frame angular distribu- tion (fig. 4) determined?110 General Discussion + 1 1 1 1 0 (v z 0 0 c .- E t Y 0 a 0 u 9) 0 2 I- & 3 - ._ & u 0 cGenera 1 Discussion 1.0. c ." Y - 1 a a 111 I IF( B, v ' ) Fig. 14. The IF( B ) vibrational product state distribution in agreement with results of Whitehead et al. for a low-pressure 12/F2 flame confirming the work of Kahler and Lee.2 (2) HIF has two stable geometries with bond angles of 137"* and 87O.I' It is possible that F12 also has two stable geometries giving rise to different entrance valleys on the F+ I2 potential-energy surface.Could this be an explanation for the bimodal IF angular distributions? (3) If migration were indeed important, might not intrusion of the F atom between the I atoms of I2 promote forward and backward I F scattering as opposed to just backward scattering? 1 ( a ) R. J. Bartlett, L. Kahn and G. D. Purvis, J. Chem. Phys., 1982, 76, 731; ( b ) R. J. Bartlett, personal communication. Dr N. C. Firth, Mr N. W. Keane, Dr D. J. Smith and Prof. R. Grice (University of Manchester) said: Comparison between the laser-induced fluorescence measurements of IF vibrational-rotational state distributions for F+I, by Girard et al.' and the differential reaction cross-section determined by Firth et aL2 is not yet fully quantitative because these two experiments have been conducted at different initial translational energies and because the criteria of distinguishing two components in the reactive scattering are not identical.In order to enable a more precise description of the F+ I2 reaction dynamics to be formulated, further reactive scattering measurements have been undertaken using a supersonic beam of F atoms seeded in Ne buffer gas to give an initial translational energy E == 19 kJ mol-' close to that employed by Girard et al.' Preliminary kinematic analysis yields the centre-of-mass angular distribution and product translational energy distributions for I F scattering shown in fig.15. This shows a decrease in scattering into the backward direction which is in accord with the trajectory calcula- t i o n ~ ~ that predict a decrease in migratory trajectories at lower initial translational energy. The component of IF product scattering into the forward hemisphere with very low product translational energy has increased to a fraction ca. 0.15, which is in close agreement with the contribution of highly rotationally and vibrationally excited IF estimated by Girard et al.' The trajectory calculations3 on F+ I2 indicate that migratory trajectories do not give rise to precisely equal forward and backward scattering, and this is probably related to the weak transfer of angular momentum from motion of the light F atom to that of the heavy I atoms.The F atom is initially inhibited from finding a symmetrical location112 0.4 0.2 General Discussion .. .' 0.0 ~ .. I I I I I I I ' f I I 0 20 40 60 80 100 120 140 160 180 0 50 100 c.m. angle, @lo translational energy, E'/kJ mol-' Fig. 15. Composite angular distribution and translational energy distributions for forward scatter- ing (-) and backward scattering (- - -) for F+ I2 at an initial translational energy E = 19 kJ mol-'. between the two I atoms since this represents an energetically unfavourable configuration. For this reason the F atom was described as intruding rather than inserting into the I2 bond. In view of the extensive experimental data which have now been accumulated on F+ 12, it will be of great interest to see whether further trajectory calculations do indicate a bent preferred geometry for F12 and indeed to investigate the possibility of two stable geometries as just suggested by Mr.Robinson. 1 B. Girard, N. Billy, G. Gouedard and J. ViguC, Faraday Discuss. Chem. SOC., 1987, 84, 65. 2 N. C. Firth, N. W. Keane, D. J. Smith and R. Grice, Faraday Discuss. Chem. SOC., 1987, 84, 53. 3 I. W. Fletcher and J. C. Whitehead, J. Chem. SOC., Faraday Trans. 2, 1984, 80, 985. Dr J. C . Whitehead (University of Manchester) said: Firth et al.' and Girard et a1.2 have commented on the role of migration in the reaction F+ I2 and have referred to the trajectory studies of Fletcher and myself3 in which we first pointed out the importance of migration in this reaction. At the time of that study, experimental data of the quality and extent presented at this Discussion''2 were not available.We used an extended LEPS potential-energy surface with a well whose depth was adjusted to fit the known stability of F12 assuming a collinear geometry. Migration in which the F atom initially attacks one of the iodine atoms and then moves round to the other was found to be common at the energies of the experiments'72 occurring in > 50% of trajectories. The migratory trajectories were found to take place at higher impact parameters, and to have higher product recoil energy with more backward scattering, higher IF rotation (fig. 16) and lower IF vibration (fig. 17) than the direct trajectories in agreement with the interpretation of Firth et al.' In particular, note that the combination of direct trajectories giving forward scattering and an almost equal proportion of migratory trajectories that are backward scattered gives an overall appearance of symmetric forward-backward scattering that could be misinterpreted as indicating a long-lived collision complex where there is no evidence for statistical behaviour in this system.In fact it is possible that migration might occur in similar systems, e.g. 0 + I?, and the assumption of a long-lived collision complex might also be incorrect. It would be helpful to examine the rotational data of Girard et al.' and determine via Doppler profiles whether it is possible to associate the high J' component withGeneral Discussion 113 L IF rotational quantum number Fig. 16. The IF rotational state distribution for the reaction F+I, ( v =0, J = 20, EtranS= 8.4 kJ mol-') -+ IF(J') + I for the direct trajectories (D) and migratory trajectories (M).25 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 IF vibrational quantum number Fig. 17. The IF vibrational state distribution for the reaction F-t I 2 ( v = 0, J = 20, €,,,,,= 8.4 kJ mol-') -+ IF( u') + I for the direct trajectories (D) and migratory trajectories (M).114 General Discussion 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1.8 19 20 IF vibrational quantum number Fig. 18. The overall I F vibrational state distribution for the reaction F+ I2 ( v = 0, J = 20, E,,,,, = 8.4 kJ mol-’) -+ IF2( u ’ ) + I compared with the experimental results of Girard et d 2 (-) and of Trickl and Wanner5 (- - -).higher recoil velocity and the low J’ component with lower product recoil energy, and thereby confirm the characteristics of migratory collisions. The comparison of the overall computed IF( u’) vibrational distribution with the determinations of Girard et al. and Trickl and Wanner’ (fig. 18) shows that there is still a need to establish a definitive vibrational distribution. It is now clear that the LEPS surface used in the trajectories is incorrect in the geometry of the F12 intermediate, which should be bent,’ but that it is correct in having dominant long-range attraction. Given the improved data now available, we should be able to derive a surface of greater accuracy and hopefully improve agreement with experiment. 1 N. C. Firth, N. W. Keane, D. J. Smith and R.Grice, Faraday Discuss. Chem. SOC., 1987, 84, 53. 2 B. Girard, N. Billy, G. Gouedard and J. ViguC, Faraday Discuss. Chem. SOC., 1987, 84, 65. 3 I. W. Fletcher and J. C. Whitehead, J. Chem. SOC., Faraday Trans. 2, 1984, 80, 985. 4 J. J. Valentini, M. J. Coggiola and Y. T. Lee, J. Am. Chem. SOC., 1976, 98, 853. 5 T. Trickl and J. Wanner, J. Chem. Phys., 1983, 78, 6091. Mr B. Girard, Dr N. Billy, Dr G. Gouedard and Dr J. Vigue (Uniuersite/ Paris Vl, France) replied: In his comment, Dr Whitehead points out that the vibrational distribu- tion of IF produced by the F+12 reaction is not firmly established, as the results of Trickl and Wanner’ are not in good agreement with ours. We will show that this discrepancy can be explained as being due to the analysis of the laser-induced fluores- cence spectrum made by Trickl and Wanner.This analysis’ was based only on bandhead intensity measurements and was not sensitive to the high-J part of the rotational distribution. In order to give further support to this idea, we have calculated truncated vibrational densities from our data: J o nJ,= c n(u, J ) . J = OGenera 1 Discussion @ + + B o 115 h v 3 2 x + 0 (B 0 $ 10 15 20 U Fig. 19. Comparison between vibrational densities: + n ( u ) , Trickl and Wanner;' 0, r ~ ~ ~ ( u ) , this work; x, r ~ ~ ~ ( u ) , this work. We have taken J0=25 and 50, and fig. 19 presents plots of these quantities, n2,3(u) and n50( u ) as a function of v, together with the measurements of Trickl and Wanner. These three sets of points (which have been normalized at their maximum values) present exactly the same vibrational dependence.This proves directly that the difference of n ( u ) between our results4 and those of Trickl and Wanner' is entirely due to the high-J contribution to n( u). Whether this high-J part of the rotational distribution was produced in the experiment of Trickl and Wanner and neglected in the analysis of the spectrum, or not produced because the collision energy in their experiment is lower than in ours, is a point which remains to be settled. Finally, the conversion from density to flux remains to be made, in order to make a direct comparison with the results of trajectory calculations. This conversion requires some knowledge of the differential cross-sections that we want to get from a study of Doppler lineshapes.Furthermore, this information should reveal finer details on the bimodal character of the collision dynamics. 1 T. Trickl and J. Wanner J. Chem. Phys., 1983, 78, 6091. 2 J. Wanner, personal communication ( 1987). 3 By a reading of Dr Trickl's thesis we have found that there is a misprint in ref. (1); the value of n 4 B. Girard, N. Billy, G. GouCdard and J. ViguC Faraday Discuss. Chem. SOC., 1987, 84, 65. ( u = 18) should read 2.06 T 0.18. Mr K. Wagemann and Dr J. Wanner (MPQ, Garching, Federal RepubZic of Germany) and Prof. X. K. Zeng (Dalian Institute of Chemical Physics, China) said: In their paper as well as in a recent publication, ViguC'et al. give proof that almost all of the intense IF( u = 0) population so far observed in the laser-induced fluorescence studies of the F+ I2 reaction system does not result from the elementary gas-phase reaction but probably originates from a surface process releasing I F molecules with a sub-Doppler velocity profile.' This conclusion is confirmed in the paper by Grice and co-workers.2 In search of clarification we re-investigated the F+I, and F+ICl systems in an experimental variant in which the reaction was carried out in a concentric discharge-flow reactor at very low reagent flows, the iodides being diluted in He.The quartz flow tube was incorporated into the cryo-pumped and cold-shielded reaction vessel which was116 General Discussion I X /////////////I Ni -surface Fig. 20. Arrangement of discharge-flow reactor, surface and dye laser as used to confirm the existence of a surface reaction, generating IF( ZI = 0) molecules in reactions of F atoms with condensed IX ( X = I , C1).used earlier in connection with two effusive beams.3 In both reactions pronounced bimodal I F product state distributions were again obtained, with similar branching ratios as reported earlier.3 Under these conditions, where the products are presumably closer to translational equilibrium, the population distributions of levels ZI 3 2 as ascribed to the elementary gas-phase reaction were found to be essentially identical as reported in ref. (3). In addition, the flow tube in conjunction with a nickel surface mounted at a distance of 3 cm and maintained at a controlled temperature allows one to verify the participation of a surface reaction. The arrangement is depicted in fig.20. Nickel was chosen, since it is the material of the inner reaction vessel normally maintained at liquid-nitrogen temperature. Hence the flux of products emerging from the reactor together with the remaining reagents, mainly fluorine atoms and molecular iodine, was directed onto the surface and the I F density was monitored with laser-induced fluorescence. At elevated temperatures a strong signal increase by surface-enhanced generation of IF (u=O) molecules was noticed. For F 4 I2 this effect disappears at a characteristic threshold temperature of 7’’ = 200 K. Below this temperature (as seen in fig. 21) only those small amounts of IF(zI=O) molecules are detected which are already produced in the flow reactor.The rotational product state analysis of the surface-generated molecules shows that these are in equilibrium with the surface over the full range of temperatures used. In order to prove that the surface selectively enhances the u=O populations, the experiment was repeated by monitoring several high vibrational product levels. As shown in fig. 21 for v=14, for example, no signal enhancement was measured within the experimental error limits. Hence we confirm the existence of a surface reaction which releases equilibrated IF( ZI =,O) molecules into the gas phase. Since this effect disappears when the microwave discharge is switched off, we ascribe it, in agreement with the observations of Grice et al., to the reaction of F atoms with condensed iodine.General Discussion 117 1.0 0. E 0 I 1 1 I 100 200 300 400 surface temperature, T,/ K Fig.21. The generation of rotationally equilibrated IF(u = 0) molecules in the F+ I2 reaction at the surface exibits a distinct threshold behaviour (0). The small amount of products detected below the threshold temperature stems from the wall reaction at the flow tube exit. High vibrational I F product states are unaffected by the surface, as depicted, for example, for IF(u = 14) (0). T,,,/K= (1) 360, (2) 280, (3) 220, (4) 180 and ( 5 ) 270. We also found that this process depends linearly on the F atom flow. The same dependence for IF( v=O) was noticed in our crossed-beam experiments.3 Apart from the observation of sub-Doppler linewidths and the verification of single collision condi- tions, this behaviour led to the belief that the v=O branch of population originates from bimolecular gas-phase dynamics.Note that we also carried out an additional experiment in which a layer of I2 was first condensed onto the surface and subsequently exposed to the impinging F atoms. Here again IF( v =0) molecules, rotationally equilibrated with the surface, were released at the same threshold temperature as reported above. For the F+ ICl system there is a similar effect, which, however, is too large to be explained in terms of the small amount of residual I2 in IC1 ( CQ. 0.1 '/o '). The proof of a less efficient surface reaction of F atoms with condensed IC1 was obtained by observa- tion of a different, distinct threshold temperature T, = 170 K. The rotational analysis of the surface reaction products presents an unambiguous criterion as to where the v =O contributions in our product-state distributions originate.These locations are the warm walls close to the prereactor tip, and, in the crossed-beam experiments, the iodine nozzle maintained at 350 K. On the basis of the rotational distribution we can also determine the existence of real IF( v = 0) populations to be due to the gas-phase reactions. For the F+ I2 reaction118 General Discussion a rather small contribution was established in the flow-tube experiment. According to a rotational fit of the excitation spectrum this population corresponds to a temperature much greater than that of the prereactor walls, T = 270 K. On the other hand, for F+ ICl there is a population of the IF(v = 0) level at T, = 2750 K.This contribution is more significant as already noted in our crossed-beam work.3 1 B. Girard, N. Billy, G. GouCdard and J. ViguC, Furaday Discuss. Chem SOC., 1987, 84, 65; Chem. Phys. Lett., 1987, 136, 101. 2 N. C. Firth, N. W. Keane, D. J. Smith and R. Grice, Furaday Discuss. Chem. Soc., 1987, 84, 5 3 . 3 T. Trickl and J. Wanner, J. Chem. Phys., 1983, 78, 6091. Dr N. C. Firth, Dr D . J. Smith and Prof. R. Grice (University of Manchester) said: The description of the surface reaction of F atoms and I2 molecules is in accord with our previous experience when studying the F+ ICl and F+ I2 reactions''2 in a crossed- beam experiment using a mass spectrometer detector. During initial experiments I F was observed at a very low laboratory velocity, which was kinematically inconsistent with I F product scattering arising from reactive collisions in the scattering centre.Indeed this background signal was removed immediately upon flagging the F atom beam but decayed only slowly after flagging the ICl or I2 molecule beam. It was therefore concluded that this signal arose from the reaction of F atoms with I2 molecules adsorbed on warm surfaces lying within the field of view of the mass spectrometer detector. Consequently all such surfaces were clad with Cu shielding cooled to ca. 100 K, and this removed the background I F signal. The product translational energy distribution for the F+ ICl reaction' was compared with that predicted from the I F vibrational and rotational state distributions of Trickl and Wanner.3 Good agreement was obtained' for I F vibrational states other than v = 0 or 1 and it was concluded that the intensity of the laser-induced fluorescence signal observed by Trickl and Wanner for these levels arose from surface reaction which could not readily be distinguished from crossed-beam scattering in the absence of the kinematic discrimination provided by time-of-flight measurements.1 N. C. Firth, D. J. Smith and R. Grice, Mol. Phys., 1986, 61, 859. 2 N. C. Firth, N. W. Keane, D. J. Smith and R. Grice, Faruduy Discuss. Chem. SOC., 1987, 84, 53. 3 T. Trickl and J. Wanner, J. Chem. Phys., in press. Mr B. Girard, Dr N. Billy, Dr G. GouCdard and Dr J. Vigue (Universite' Paris VI, France) added: In their comment, Dr Wanner and Mr Wagemann have presented nice experimental results showing that the IF(v=O) molecules are due to a wall reaction.These confirm our result' that the IF( v = 0) molecules are not a direct product of the F+12 gas-phase reaction. We thus agree on this point and we wish to present further evidence of its correctness. The first piece of information comes from the Doppler lineshapes: the v = O lines were found to be very narrow in our experiment,' corresponding to an apparent temperature of <lo0 K. This value could not be explained by the wall temperature, as it remained unchanged even when all the walls were kept at room temperature. On the other hand, this feature is easily understood as due to the following collimation effect: the IF( v = 0) molecules are produced by wall reactions in the vicinity of the F beam and they can be detected only if they fly towards the scattering centre; this selects their velocity as almost parallel to the F beam and perpendicular to the laser beam.The second item of information comes from the fact that the v = 0 signal is strongly enhanced (typically by a factor of 5 ) if the F beam is flagged in the scattering chamber, before the scattering centre. In this case all the fluorine atoms diffuse to the walls where they react, whereas only a small fraction of the unflagged beam can strike a wall. We consider that these results prove clearly that the IF( v = 0) molecules observed are not due to the direct reaction of F+ I2 but to a wall reaction. We have tried to detectGeneral Discussion 119 u = 0 molecules due to the direct reaction.These efforts have been unsuccessful even when working at large J values (near J = 8 0 and 120), where the signal coming from the wall reaction is negligible. However, the upper limit we can fix to the density of u = 0 molecules is not very low: n ( u = 0 ) 6 n( u = 13). This is due to the fact that, if the iodine atom is in its *P3,* ground state, many levels can be populated in v = 0 (up to J==200), and there is a good deal of recoil energy available for the lower J values, implying very broad Doppler line profiles. 1 B. Girard, N . Billy, G. Gouidard and J. Vigui, Chern. Phys. Lett., 1987, 136, 101 Prof. A. Lagana (University of Perugia, Italy) and Dr L. A. M. Quintales and Prof. J. M. Alvarifio ((University of Salamanca, Spain) said: The dynamical behaviour of the H + ICl reaction has been taken as a model for rationalizing the reactivity of systems involving halogen-containing molecules.The H + ICl reaction was first investigated by Polanyi and co-workers.' In their study evidence was given for the possibility of obtaining the HCl product even when H first impinges and binds to the I atom. Recently we have reinvestigated this reaction using the same potential-energy surface and the same set of initial conditions. In addition, calculations were also carried out at higher collision energy ( T ) as well as at different initial orientations of the target diatom in order to provide evidence of stereospecific properties of this reaction. A basic feature of the H+ICl reactivity is the competition between C1 and I for forming the product hydrogen halide.Our results agree with the findings of Polanyi and co-workers of a strong preference of the reaction for the HI product at low collision energy. However, we also find that this bias is less pronounced at higher energies. This effect can be attributed mainly to a competition between two direct (collinearly domi- nated ) mechanisms. The typical behaviour of direct collinear reactive probabilities [ P ( E ) ] is illustrated in fig. 22. In the upper panel (for systems such as H+ I with no barrier to reaction) P( E ) is fully reactive at low collision energies as a result of the fact that the potential is attractive from long-range inward. At higher T values P ( E ) decreases. On the LEPS surface adopted for the H + ICl reaction, this can be understood in terms of back reflection from the wall opposite to the entrance channel.In fact, once the collision energy is large enough to allow reflected trajectories to return to the reactant asymptote, a further increase in 7' lowers the reactive probability (i.e. the complement to one of the fraction of back-reflected trajectories).* The lower panel shows the case of reactions having (like H + Clz) an early barrier. For systems having an early barrier to reaction there is first a reflection at low collision energy occurring at the bottleneck associated with the saddle. At higher collision energies P ( E ) increases to reach (possibly) full reactivity. At even higher T values the amount of energy in excess of the height of the saddle becomes larger, and back reflection can start to occur.Because of this, the tail of the P( E ) curve for these systems decreases, as in the case of reactions without a barrier. A more composite mechanism is followed by other reactive collisions.' If the reactive path goes through a situation in which insertion almost occurs, a non-negligible fraction of trajectories can spend a certain amount of time in finding the final exit. In this case, migration from the atom of the first attack to that forming the product diatom is likely to occur. Migration of H from I to C1 has been singled out in ref. ( 1 ) . The graphical study performed on our trajectories shows that at first H binds to I after being captured from its long-range attractive tail.Then, while IC1 stretches, H passes around the I atom, undergoing several hindered rotations (see fig. 23) till the moment it finds a location for which the barrier to react with C1 is small. Our calculations, however, have also shown the possibility that collisions starting on the C1 side lead to the formation of HI. In this case (fig.24) the hydrogen atom binds at first with the nearest atom (Cl), around which it performs a few oscillations. When it reaches a favourable location, capture by the iodine atom is possible.120 General Discussion N h Z rrl 0' a, 1.0,- 2 v I - - - - - -- \ I I I I I I I I I \ \ \ '\ 0 I N d E(arb. units) Fig. 22. Qualitative evolution with energy of reactive collinear probability for systems with a barrier to reaction (lower panel) and systems with no barrier (upper panel).Finally, a case less likely to occur is that of H orbiting for a while around the IC1 molecule before reacting with either I or C1. This type of floppy aggregate, stabilized by rotational trapping around the target molecule, has been found to be more effective in promoting the release of internal energy in non-reactive processes4 rather than those leading to reaction. 1 J. C. Polanyi, J. L. Schreiber and W. J. Skrlac, Faraday Discuss. Chem. SOC., 1979, 67, 66. 2 J. N. L. Connor and A. Lagana, Mol. Phys., 1979, 38, 657. 3 J. M. Alvariho and A Lagana, J. Phys. Chem., 1987,97, 5487. 4 J. P. Tosi, M. Ronchetti and A. Lagana, J. Chem. Phys., 1988, in press. Dr M. R. Levy (Newcastle-upon-Tyne Polytechnic) said: Prof.Polanyi' has referred to the novel technique of laser ablation for producing atomic beams, and Costes et al.' have admirably demonstrated its application to the study of aluminium atom reactions. As a further illustration of the potential of this approach, I would like to present some results from recent work on transition-metal atom reaction^.^ Experiments in this area have the potential to be particularly rewarding, since both the metal atoms and the diatomic oxide or halide product molecules have many low-lying electronic states (some metastable), so that a large number of potential surfaces may interact. Laser ablationGeneral Discussion 121 ? x -.' I I I - 8 -4 0 Fig. 23. A daisy plot of a reactive trajectory showing the migration of H from I to C1.The positions of H (O), I (A) and C1 (m) after identical intervals of time are labelled by the same number. ? x -6 - 3 0 3 x/A Fig. 24. As in fig. 23 for the migration of H from C1 to I. is especially appropriate in view of the high refractivity of the metals, which has restricted previous dynamical investigations to a relatively small n ~ m b e r . ~ My own studies have concerned chemiluminescent reactions of the type Mn + RO -+ MnO*( A 'Z+) + R (R = 0, N2, CO, NO). This represented a suitable starting point for studies of transition- metal atom reactions because of the relative simplicity of ( a ) the Mn electronic structure (fig. 25)5 and (6) the well known MnO(A %+-X %+) spectrum from ca. 450 to ca. 750 nm.6 Fig. 26 shows the experimental zrrangement. There are two important122 General Discussion 25 20 7 15 E m E: -.0 .- Y Y .C ; 10 5 0 3d54s2 3d?6S)4s4p 3 d 6 ( 5 0 ) 4 ~ Fig. 25. Low-lying electronic structure of the Mn atom.5 The a ‘DJ and a “DJ states are truly metastable. (a) A = 403.1, 403.3 and 403.5 nm; T = 50 ns. ( b ) A = 539.5 and 543.5 nm; T = 80 and 120 ps. differences from that used by Costes et a1.* First, a beam-gas configuration was employed; secondly, the metal target was ablated in the absence of any carrier gas from a pulsed valve. In this approach, devised by Kang and Beauchamp,’ the laser is focussed onto the target by a concave mirror placed in front of it, and the pulsed atomic beam passes out through a narrow hole in the centre of the mirror. The predominant ionic component in the beam is deflected away by a d.c.electric field. While this technique does not give the narrow velocity spread and high peak intensities achievable with pulsed valves, it does offer some advantages: it is relatively simple; the broad velocity spread is immedi- ately resolvable by time-of-flight analysis, so that excitation functions can be measured; the velocity distribution can be altered by adjusting the laser focussing; and the tem- perature of the laser-induced plasma is so high that the beam contains a substantial proportion of atoms in metastable states. In addition, in the case of Mn, the beam can be detected by the long-lived emission at ca. 540 nm from the P7/2,5,2 levels ( T = 83 f 10, 123 f 14 ps, re~pectively.~ The collimated beam and its reaction with oxidiser molecules were monitored optically at a point 283 mm from the target. Data were time-of-flight spectra of the beam number density (weighted by the exponential decay of the 8 P levels), and the corresponding time-dependent emissions produced when oxidiser was admitted to the scattering chamber.Between 64 and 1024 laser shots were averaged, depending on whatGeneral Discussion 123 I I I transient digit i ser trigger - - - - - - - I I I 1 I I I I I I I I I I I I I I I Fig. 26. Schematic of experimental arrangement. T, target; D, ion deflection plates. was being observed. Fig. 27 shows a typical time-profile of the metastable emission, P( t ) , together with the corresponding Mn flux distribution, N'( t ) (including the correc- tion for the metastable decay). The beam velocity ranges from > 2000 to < 20 000 m s-l, and the distribution in fig.27 can be characterised by a plasma temperature of ca. 77 000 K.3 The metal atom beam velocity determines the mean collision energy; and, for the bulk of the velocity range, the oxidiser velocity is negligible. MnO*(A 'X+) chemiluminescence was observed when 02, C 0 2 , N20 or NO, was admitted to the scattering chamber. The pressure dependence was linear up to ca. 0.020 Pa (ca. 1.5 x Torr), indicating first-order kinetics. Collision-induced atomic emission was also detected at ca. 403 nm, due to either the z ' P o + a 6S or z ' D o -+ a6D transitions of Mn. Fig. 28(a) shows typical time profiles, I ( t ) , of both the MnO" emission ( A > 550 nm) and the collision-induced Mn" emission from Mn + N20.As the radiative lifetimes of these emissions are short, I ( t ) measures flux and can be readily converted into a( E T ) , the relative cross-section at the translational energy ET corre- sponding to the time-of-flight t, by dividing by N ' ( t ) . The resulting excitation functions for the two emissions from Mn+ N,O are shown in fig. 2 8 ( 6 ) It is noticeable that the fall-off in the molecular emission is matched by the rise in the atomic signal. The arrows on the ET axis indicate the thresholds for Mn0*(A)698 and Mn*(z production from Mn(a ' S ) . Although there is some noise in the data, the figure strongly suggests that a ' S atoms, rather than a 6D or higher states, are the species responsible for both processes, and that the 403 nm emission is z 6P0-+ a 'S, rather than z ' D o - + a ' 0 .In fact the threshold for MnO* production indicates a barrier ca. 70 kJ mol-' above the endoergicity; the barrier for a 6 D atoms must therefore be even higher. This contrasts with Mn + O2 ,C02 and NOz ,3 where both MnO*(A) and Mn*(z ' P o ) derive, at least in part, from excited a 6D atoms, but where the MnO* threshold is equal to the endoergicity. However, in these cases too, the124 General Discussion z e a Fig. 27. Typical time profile of Mn*(8P) emission, P ( t ) , monitored at 283 mm from the target (-), together with the corresponding derived time-profile of the non-emitting Mn atom flux, " ( t ) (- -). reactivity seems largely restricted to Mn(a ' S , a ' D ) , despite the undoubted presence in the beam of even higher metastable states.Adiabatic correlation diagrams can go some way towards explaining the behaviour of these reactions3 In most cases they suggest that a ' S atoms should lead to MnO(X), whereas MnO*( A) should derive largely from a '0 atoms. Nonetheless, ground-state Mn atoms could still give MnO*( A) through intersystem crossing in the collision complex, if it lives long enough, or through overlap between the X - and A-state vibrational manifolds. The replacement of MnO* by Mn*(z ' P ) with increasing ET, in Mn+ N20, suggests that the collision complex lives long enough for crossover to occur between the surfaces corresponding to Mn(a ' S ) and Mn*(z ' P ) . Correlation diagrams, however, give little information about potential barriers. It is conceivable that the large excess barriers for reaction of both Mn(a 6S and a ' D ) with N20 arise from the high second ionisation potential of Mn; but this does not explain the different behaviour in the isoelectronic Mn + C02. Potential barriers have been noted previously in the reaction of a number of metals with N20, e.g. Na9 and Sn," but there have been relatively few comparative studies of N20 and C02 reactions (probably because the latter are mostly endothermic). In their paper, Costes et al.' show that Al+CO, has no barrier above the endothermicity; it would be interesting to learn whether Al+ N20, ca. 345 kJ mol-' exothermic,278 does have a barrier. The energy released should be enough to populate AlO(B 2Z+); however, a preliminary report elsewhere' ' indicated only weak A 'II emission under single-collision conditions.General Discussion 125 ET/kJ mol-* Fig. 28. ( a ) Typical time profiles of MnO*(A %+) (0) and Mn*(z 6Po - a 6 S or z ‘Do + a “0) (0) emission from Mn+ N,O; ( b ) excitation functions derived from the data. The arrows on the abscissa mark the thresholds for MnO*(A) and Mn*(z6P0) from Mn(a 6 S ) (ca. 19 and 297 kJ mol-’, respectively).126 General Discussion 1 J. C. Polanyi, Faraday Discuss. Chem. SOC., 1987, 84, 1. 2 M. Costes, C. Naulin, G. Dorthe, C. Vaucamps and G. Nouchi, Faraday Discuss. Chem SOC., 1987, 3 M. R. Levy, to be published; equipment loans from the SERC Central Laser Facility are gratefully 4 For a rather dated review, see M. R. Levy, Prog. React. Kinet., 1979, 10, 1. 5 J. Sugar and C. Corliss, J. Phys. Chem. Re$ Data, 1985, 14, suppl 2, 338 ff; S. M. Younger, J. R Fuhr, G. A. Martin and W. L. Wiese, J. Phys. Chem. Ref: Data, 1978, 7, 591ff. 6 J. M. Das Sarma, 2. Phys., 1959, 145, 98; K. C. Joshi, Spectrochim. Acta, 1962, 18, 625; R. M. Gordon and A. J. Merer, Can. J. Phys., 1980, 58, 642. 7 H. Kang and J. L. Beauchamp, J. Phys. Chem., 1985, 89, 3364. 8 G. Herzberg, Electronic Spectra of Polyatomic Molecules, and K. Huber and G. Herzberg, Constants of 9 J. Pfeifer and J. L. Gole, J. Chem. Phys., 1984, 80, 565. 84,75. acknowledged. Diatomic Molecules (Van Nostrand, London, 1966 and 1979). 10 J. R. Wiesenfeld and J. M. Yuen, Chem Phys Lett., 1976, 42, 293. 11 J. L. Gole and G. J. Green, cited in S. B. Oblath and J. L. Gole, Combust. Flame, 1980, 37, 293. Dr M. Costes, Dr C. Naulin and Dr G. Dorthe (University of Bordeaux, France) said: We have indeed performed some experiments on the exoergic Al+N20 reaction. However, we have not obtained quantitative results.' LIF signals of AlO(X 'E+) remained low even at a collision energy of 0.52 eV. We also failed to detect by LIF any electronically excited A10( A 211 ) product, by excitation on B - A transitions on the Av = +2 sequence near 580 nm and collecting the fluorescence of B - X transitions. Combined factors are certainly responsible for the low LIF signals: the spread of the products on the large manifold of accessible rovibrational states, the high recoil velocities which lower the densities in the laser-probed volume, and perhaps a lower reactive cross-section than for the other reactions. We have not made runs under 0.30 eV collision energy and thus cannot conclude about the existence of an energy barrier. This event- uality seems, however, very likely, as it is often the case for atom-nitrous oxide reactions. For example, we have observed an energy barrier for the exoergic C + N20 reaction.* 1 M. Costes, C . Naulin, G. Dorthe and G. Nouchi, in NATO Advanced Research Workshop on Selectivity in Chemical Reactions, ed. J. C . Whitehead (Reidel, Dordrecht, 1988). 2 M. Costes, G. Dorthe, B. Duguay, P. Halvick, J. Joussot-Dubien, C. Naulin, G. Nouchi, J. C. Rayez, M. T. Rayez and C. Vaucamps, .in Recent Advances in Molecular Reaction Dynamics, ed. R. Vetter and J. Vigue (C.N.R.S., Paris, 1986), p. 97.

ISSN:0301-7249    

DOI:10.1039/DC9878400087

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Furaduy Discuss. Chem. SOC., 1987, 84, 127-143 Spin-Orbit Effects in Chemical Reactions Investigation of Ground-state Products from Reactions of Ba(3D) Mark L. Campbell? and Paul J. Dagdigian* Department of Chemistry, The Johns Hopkins University, Baltimore, Maryland 2121 8, U. S. A. The dependence of the cross-sections for production of ground-state barium halide products on incident spin-orbit state has been determined by means of optical-pumping state selection for the reaction of metastable Ba(6sSd 3 D ) with HCI, CH3Cl, HBr and CH3Br. In addition, cross-sections for the metastable D level were related to those of the 3DJ multiplet by optical pumping on an intercombination line. For alkyl halide (RX) reactants, the spin-orbit dependence of the reactivity for the ground-state channel was substantial with an ordering J = 1 > J = 2 > J = 3.This is an opposite order- ing to that previously observed for the chemiluminescence channels in analogous reactions. The hydrogen halide reactions exhibited a varying spin-orbit dependence with vibrational level. For the most highly populated vibrational levels v, the spin-orbit dependence was comparable in sign and magnitude to that for RX reactants, while a significantly diminished variation of reactivity with incident J was observed for lower u. The variation of spin-orbit effect with product vibrational level is believed to be due to the dependence of the reaction dynamics on incident impact parameter. It has been demonstrated for a variety of atomic reactants that the chemical reactivity of different spin-orbit levels associated with an atomic term of non-zero spin and orbital angular momentum can vary significantly.’ Significant differences in reaction rates have been observed for atomic multiplets with large spin-orbit splittings, such as Hg(3P0), Sn(3P), Pb(3P) and the metastable 3P0 inert-gas atoms, as might be expected, but also for terms with small splittings, such as F(’P0) and the metastable triplet states of the alkaline-earth atoms. In most studies, the total rate of collisional removal was measured for the different spin-orbit levels; however, in some experiments it was possible to determine the spin-orbit dependence for specific product channels.In our laboratory, we have carried out a series of studies of spin-orbit effects in reactions of the alkaline-earth atoms, Ca(3P0),2-4 Sr(’PO),’ and Ba(3D).6 An optical- pumping state selection technique was employed to perturb the near-statistical spin-orbit populations of beams of these metastable terms emerging from a discharge atomic beam source.’ Chemiluminescence channels of a number of reactions of these atomic species with diatomic halogens, alkyl bromides and iodides, as well as with N20 and NO2, were investigated.With the exception of Ca(3P0) + SF, and Ba(’D) + N20, NO2, for which no spin-orbit dependence was observed, the chemiluminescence cross-sections were found to vary significantly with incident atomic spin-orbit level. In all cases, the level of highest energy [ J = 2 for Ca(’Po) and Sr(3P0) and J = 3 for Ba(’D)] possessed the largest cross-section. The cross-sections from the lowest-energy spin-orbit level [ J = 0 for 3P0, J = 1 for ’ D ] were typically 5 to 10 times smaller than for the highest-energy level.For several reactions [Ca(’P*) + C12 and Sr(3P0) + HBr, CH,Br2] it was possible by laser fluorescence detection to investigate the spin-orbit dependence of the reaction t Present address: Department of Chemistry, U.S. Naval Academy, Annapolis, MD 21402, U S A . 127128 Ground-state Products from Reactions of Ba( 3D) channel leading to ground-state prod~cts.~~* In marked contrast with the chemilumines- cence channels, the lowest-energy spin-orbit level ( J = 0) was found to possess the largest cross-sections. These experimental results have been interpreted with the help of a pseudo-quenching model calculation for Ca(3P') + C1, by Alexander.' In this quantum-mechanical treatment, reactive collisions were simulated in an atom-atom collision by quenching to a lower asymptote via an ionic-covalent curve crossing.While the calculated cross-sections displayed a strong sensitivity to the strength of the ionic- covalent coupling, these model calculations could qualitatively explain the experimental results. The spin-orbit effect arises from differences in the evolution of the asymptotic spin-orbit level onto the various electrostatic potential-energy surfaces. Despite the small Ca( 3P0) spin-orbit splitting, an adiabatic correlation argument for the spin-orbit dependence seems to be valid, wherein wavefunctions arising from the lower-energy incident spin-orbit levels selectively undergo charge transfer to an ionic surface.The opposite ordering of reactivity for the chemiluminescence channel vs. the ground-state product channel can further be explained as arising from a selective removal of flux at the outermost ionic-covalent curve crossing. In the present paper we report on the spin-orbit dependence in the ground-state reaction channel of the Ba(6s5d 3D) + HX, CH3X(X = C1, Br) reactions. For these reactions we were able to determine relative spin-orbit dependent reaction cross-sections for a range of BaX product vibrational levels v. It has also been possible to determine reaction cross-sections for the second metastable Ba level, 6s5d '0, relative to those for 3D. Our experimental results show that the spin-orbit dependence for the '0 manifold varies with v.These data allow us to gain information on the dependence of the spin-orbit effect upon impact parameter. Experimental These experiments were carried out in a beam-static gas scattering arrangement, as described in detail in previous publication^.^*^" A near-eff usive beam of metastable electronically excited (6s5d 3DJ and ID2) Ba atoms was produced in a discharge beam source'' and passed through a slit into a reaction chamber, to which gaseous reagents were introduced at pressures of up to 1 mTorr (as measured with a capacitance manometer). Laser-induced fluorescence signals due to barium halide reaction products were detected in a zone 1.6+0.1 cm beyond the collimating slit. A home-made grazing- incidence dye laser,' ' of bandwidth 0.15 cm-' (double grating arrangement 1 2 ) , which was pumped at 10 Hz by an excimer laser (Lambda Physik EMG53MSC), was employed for excitation.The fluorescence was collected with f/ 1.5 optics and imaged unfiltered onto a Hamamatsu R928 or EM1 9816 photomultiplier tube, whose output was fed into a boxcar integrator. Excitation spectra were acquired under computer control (DEC LSI-11/23), and the spectra were stored on magnetic diskettes for later analysis. Optical-pumping depletion with a single-mode C.W. dye laser (CR599-21 with rhodamine 6G dye) was employed for state-selection of the Ba" beam. The laser beam intersected the atomic beam at right-angles ca. 2 cm in front of the beam source; typical laser power was 90 mW in a 4 mm diameter beam.The C.W. laser wavelength was adjusted to the centre of a given atomic transition with the aid of a fluorescence detector ca. 50 cm downstream of the main fluorescence zone, as described in detail previously.6 A small fraction of the laser beam was directed to the auxiliary detector and the pump laser wavelength was optimized by maximizing the atomic fluorescence signal intensity. The ' D, level and individual DJ spin-orbit states were depleted by tuning the laser to the 5d6p 'P1 +-6s5d 'D2 line at 582.6nm and lines of the 5d6p3P"-6s5d ' D multiplet at 590.8 to 61 1.1 nm, respectively. Cross-sections for the ' Dz level were related to those for '0J by pumping on the 5d6p 'Fg +-6s5d 'D3 intercombination line at 580.6 nm. [See the Ba energy-level diagram given in ref.(6).] Because of the numberM. L. Campbell and P. J. Dagdigian 129 Table 1. Energies (in eV) for the Ba* + RX -+ BaX reactions studied HCl 0.05 f 0.09 0.03 0.08 0.16 f 0.09 1.32 f 0.09 HBr -0.05 f 0.10 0.03 0.14 0.12 f 0.10 1.28 f 0.10 CH3Cl 0.91 f 0.09 0.04 0.10 1.05 f 0.09 2.21 f 0.09 CH,Br 0.71 f 0.10 0.05 0.16 0.91 fO.10 2.07f0.10 a Dg (BaX) - Dg(RX). The latter were calculated from data in ref. (15). 'Initial energy of the RX reactant, calculated from the enthalpy at 298 K. 'Average initial relative translational energy, computed by convoluting the beam and target gas velocity distributions [see ref. (16)]; the former was derived from time-of-flight measurements [ref. (17)]. dEav,('S) and E,vl(3D) are the total energy available to the products for the 'S and 30 reactions, respectively: E,,, = ADo+ Ein,(RX)+ E:,,,,+ Ein,(Ba*).Ef,,,, has been taken to be the average reported in column 4. Eint(Ba*) = 0, 1.160, and 1.413 eV for the IS, 3D and 'D states, respectively [ref. (IS)]. of naturally occurring Ba isotopes, some of which have hyperfine sp1iting,l3 it was not possible to remove all the atoms in a particular level. The C.W. laser was tuned to deplete the most abundant mass 138 isotope (72% natural abundance14), and account was taken of incomplete depletion in the analysis of the data. The extraction of reaction cross-sections for individual atomic reagent states from the change of the product signal intensity upon optical-pumping depletion requires knowledge of the relative atomic populations for the various optical pumping conditions.These populations for Ba D2 and Dj were previously derived from laser-induced fluorescence measurements.6 The population of the ID2 state was related to those of 3DJ by utilizing the IF! + 3D3 intercombination line, since the upper state of this transition decays mainly to 'D2 rather than to the 'DJ states. The relative ID2 and 'DJ populations are given in table I of ref. (6); the lines employed in the present study are numbers 8-14 inclusive in that table. With no optical pumping, the ID2 population is 16% of the total metastable atoms present. Results The reactions of Ba(3D, ID) with HX and CH3X reactants, with X = C1 and Br, were chosen for study here, in part because these reactions yield little or no chemilumines- cence, which would interfere with laser fluorescence detection of ground-state products.Both hydrogen and methyl halide reactants have been investigated in order to see whether their slightly different reaction dynamics affect the dependence of reactivity on incident spin-orbit state. The energetics of the reactions studied are displayed in table 1. We have employed values obtained from a mass-spectrometric thermochemical study" for the barium halide dissociation energies: D:( BaCl) = 4.48 f 0.09 eV and D:( BaBr) = 3.7 1 f 0.10 eV. The latter value is essentially identical to that determined from a chemiluminescence study.16 It can be seen from table 1 that the Ba('S) reactions with the hydrogen halides are approximately thermoneutral, while the methyl halide reactions and all those involving metastable Ba( '0, D ) are substantially exothermic.Laser Fluorescence Excitation Spectra Fig. 1 compares BaBr laser fluorescence spectra of the C 'I13,,-X 'Z+ Av = 0 and - 1 sequences for the HBr reaction with ground-state Ba( 'S) reactant (source discharge130 Ground-state Products from Reactions of Ba( 3D) c2n3,,- x2c+ AV-0 AV=-1 10 0 R21 --%-r++ 10 0 R2+Q21 4 10 0 r I 1 I I I 1 I 1 I I I 1 I 520 524 528 51 6 laser wavelength/nm Fig. 1. Laser fluorescence spectrum of the BaBr C ’II312-X *Z+ Av = 0 and -1 sequences obtained for the ( a ) Ba(’S) + HBr and ( b ) Ba” = HBr reactions. The RZ1 and R2 + QZl are separately denoted. Lines due to excitation of atomic Ba* from the discharge source are indicated. off) and with metastable Ba(’D, 30) atoms present (source discharge on).The former is essentially identical (but with somewhat better spectral resolution) to spectra previously reported both from thermal beam-static gas and crossed beam configurations.20*21 The R2+Q2, and RZ, band heads are separately discernible in the spectra, as previously observed by Munakata et aL2’ in excitation spectra of BaBr product from the Ba + CH,Br reaction. For the Ba(’S)+HBr reaction, BaBr vibrational levels up to v = 19 are detectable in the excitation spectrum. The product vibrational distribution for this reaction has been derived Assuming the thermochemistry given pre- viously is correct, the high-v tail of the vibrational distribution must arise from collisions in the high translational energy tail.With the source discharge on, the BaBr fluorescence excitation spectrum becomes considerably more complex [see fig. 1( b ) ] , and BaBr vibrational bands of considerably higher Y” are discernible (0’’ a 30), consistent with the higher Ba( ’0, ’ 0) + HBr reaction exothermicity (om,, = 53). As has been noted previously by Schultz and Siegelz’ in aM. L. Campbell and P. J. Dagdigian I 131 11. il 1111 h V"=25 I I I I I I I 516 520 524 528 laser wavelength/nm Fig. 2. Laser fluorescence excitation spectrum for the Ba* + HBr reaction. The contribution due to the residual Ba('S) in the Ba* beam has been subtracted from the spectrum in fig. l ( b ) as described in the text. spectroscopic study of BaBr product from the homologous reaction series Ba( ' S ) + CH,Br,-,, the high-u" bands of a given Au sequence overlap the next sequence Av + 1.In our spectrum we observe a filling in of the fluorescence signal between the Av = 0 and Au = -1 sequences. One puzzling aspect of our Ba* + HBr spectrum is the presence of additional unassignable, but reproducible, features, particularly among the low-v" bands of both the Au = 0 and -1 sequences. It is tempting to ascribe these extra features to high-v" bands of another sequence. For instance, the v"= 50 band of the Au = -1 sequence falls at approximately the same wavelength as v" = 19 of the Au = 0 sequence." However, we have not attempted to make further assignments by extensive studies of other Ba + RBr reactions or by spectral simulations. Since the metastable conversion efficiency of our source is large, but not 100'/0,'~ the spectrum taken with the source discharge on includes some contribution from the Ba('S) + HBr reaction.To estimate this contribution, we monitored the 'S density by laser fluorescence excitation of the Ba 'Po + ' S resonance transition at 553.56 nm. We found that the fluorescence intensity on this line is reduced to 15% of its original value when the source discharge is turned on. This estimated conversion efficiency is similar to that measured in a previously reported characterization of our Ba* source." Accord- ingly, we have subtracted 15% of the discharge-off spectra from those taken with the source discharge on. The resulting corrected spectrum for Ba*+HBr is displayed in fig. 2. Because of the uncertainties in our BaBr spectra noted earlier, we have not attempted to extract product internal-energy distributions.The band heads of the Av = -1 sequence in fig. 2 are most prominent around u"= 28; however, this appears to be due to a coalescence of two separate series of band heads. In spite of the difficulties in analysing these spectra in detail, we may nevertheless surmise that a broad range of BaBr vibrational levels are populated by the Ba(3D, ' D ) + HBr reaction up to and beyond v = 30. Fig. 3 presents BaBr laser fluorescence spectra for the CH,Br reaction both without and with the source discharge on. The spectrum for the Ba('S) reaction is essentially identical to that previously reported by Munakata et ~ 1 . ~ ~ in a crossed-beam study at a132 Ground-state Products from Reactions of Ba( 3D) c2n,,2L X2E+ AV=O AV=-1 R21 20 10 0 20 10 0 - 20 10 0 R2+Q2, ---A+++ 20 10 0 I I I I I I I 516 520 524 528 laser wavelength/ nm Fig.3. Laser fluorescence spectrum of the BaBr C 2r13,2-X 'E+ Av = 0 and -1 sequences obtained for the Ba('S)+CH,Br ( a ) and Ba*+CH3Br ( b ) reactions. R2, and R2+Qz, band heads are separately denoted. Lines due to excitation of atomic Ba* from the discharge source are indicated. ( b ) The contribution due to residual Ba('S) in the Ba" beam has been subtracted, as described in the text. comparable average collision energy. In accord with the larger reaction exothermicity, the BaBr product contains somewhat higher vibrational excitation than for the Ba( ' S ) + HBr reaction. However, in contrast to the latter, the highest vibrational level observed ( u = 25) has considerably lower internal energy than the total available energy Eavl( ' S ) for the CH3Br reaction.The excitation spectrum for the CH3Br reaction, shown in fig. 3( b ) , is considerably more complex and indicates substantial product vibrational excita- tion, with a broad vibrational-state distribution. Again, as with the corresponding HBr reaction, we have not attempted to extract a quantitative product-state distribution from the fluorescence excitation spectrum. Considerable BaCl product vibrational excitation is also evident for the Ba* + HC1 reaction. Fig. 4 shows laser fluorescence spectra for the Ba( IS) and Ba" + HC1 reactions.Av=o 0 n M. L. Campbell and P. J. Dagdigian c 2n ,,2- x2z+ AV=-1 n 1 I I I I I I 133 I I I I I 524 528 532 laser wavelength/nm Fig.4. Laser fluorescence spectrum of the BaCl C 211,,2-X 'Z'+ Av = 0 and -1 sequences obtained for the Ba('S)+HCl ( a ) and Ba*+HCl ( b ) reactions. Lines due to excitation of atomic Ba* from the discharge source are indicated. ( b ) The contribution due to residual Ba('S) in the Ba* beam has been subtracted, as described in the text. The former reaction is known to yield relatively little BaCl vibrational mainly because of the small amount of energy available to the products (see table 1). Only a few bands of low v are visible in the spectrum in fig. 4 ( a ) . In contrast, there is considerable fluorescence intensity between the sequences in fig. 4 ( 6 ) for the Ba* reaction, indicative of considerable product vibrational excitation.It also appears that the vibrational state distribution is very broad for this reaction, as was observed for the Ba*+HBr reaction, although we have again not attempted to extract a quantitative product-state distribution from the spectra.134 Ground-state Products from Reactions of Ba( 3D) 50 40 I I I I 1 I I I 524 528 532 laser wavelength/nm Fig. 5. Laser fluorescence spectrum of the BaCl C 2111,2-X 2Xc+ Av = -1 sequence obtained for the Ba( IS) + CC14 ( a ) and Ba* + CH3Cl ( b ) reactions. Lines due to excitation of atomic Ba* from the discharge source are indicated. Fig. 5 shows a BaCl excitation spectrum over the same wavelength range for the Ba" + CH3Cl reaction. In this case, the corresponding ground state Ba( 'S) reaction does not yield any detectable BaCl fluorescence signal, in agreement with a previous study.24 Hence, the spectrum in fig.5(b) contains no contribution from Ba('S)+CH,Cl. For comparison, a laser fluorescence spectrum for the Ba( 'S) + CC14 reaction is presented in fig. 5(a). This spectrum is essentially identical (but with higher spectral resolution) to that previously reported by Schmidt et aZ.*' in an extensive study of this reaction. Their derived BaCl product vibrational-state distribution was very narrow, with a most probable o of 43 and significant population only for levels rtl5 units about this peak.M. L. Campbell and P. J. Dagdigian AwO - 5 0 Av=+l C2rI,,,- X 2 F I I I I Ba c 135 I I 1 1 51 2 51 8 laser wavelength/nm Fig.6. Laser fluorescence spectrum of the BaCl C 'n3,*-X *C+ Av = 0 and C 2111,2-X 'X+ Av = + 1 sequences obtained for the Ba('S) + HCl ( a ) and Ba* + CH3CI ( b ) reactions. The R, and Q, + RI2 band heads for low v are overlapped in the latter sequence. Lines due to excitation of atomic Ba" from the discharge source are indicated. Comparison of our CC14 spectrum in fig. 5 ( a ) with those for Ba" + HCI and CHICl in fig. 4( 6 ) and 5( b ) , respectively, suggest that the BaCl vibrational-state distribution for both of these reactions are very broad and extend out beyond v = 45. Finally, in fig. 6( 6 ) we present a BaCl fluorescence excitation spectrum of the Ba" + CH3CI reaction for the C 2113/2-X 'C+ Av = 0 and C 211,/2-X 'Z+ Av = + 1 sequences. This spectrum is compared to that for the Ba('S)+HCl reaction, which is given in fig.6 ( a ) ; the latter has been presented in the literature previously." Bands of the C zII,,2-X 'Z+ Av = + 1 sequence for low v severely overlapped each other, as discussed previously by Siege1136 Ground-state Products from Reactions of Ba( ,D) and Schultz;" however, the high-u bands occur to the blue. We thus see further evidence for high BaCl vibrational excitation in the Ba*+CH,Cl reaction by the presence of significant fluorescence intensity in fig. 6(b) to the blue of the low-u R, heads. Dependence on Reactant Ba 3D Spin-Orbit State The dependence of the reaction cross-sections for formation of various product BaX vibrational states from specific Ba D spin-orbit levels was determined by recording the change of the fluorescence signal at appropriate probe laser wavelengths as the incident Ba3DJ, 'D2 state distribution was altered by C.W.dye laser optical pumping. Two different derivations of the extraction of spin-orbit-dependent cross-sections from these data have been given in previous publications.27336 In this study we employed the same analysis as that used to determine the spin-orbit dependence for chemiluminescence channels in a number of Ba* reactions.6 We give an abbreviated version of the relevant equations here. With the C.W. optical pumping dye laser off, the laser fluorescence signal for detection of a given BaX product state can be written as where n.6' and nyff are the relative number densities of the Ba ' D and ,DJ states, respectively, at the collision zone. We have normalized these atomic populations so that their sum equals unity. The proportionality constant c includes such factors as the target gas and total Ba" number density, the probe laser power and detection sensitivity.The quantities aJ and aD represent the cross-sections for production of the detected BaX states from a specific reactant Ba 'DJ spin-orbit state and the ID2 state, respectively. Eqn (1) assumes that any possible contribution to the detected product from reaction with residual Ba(*S) in the beam has been subtracted from the observed signal. When the C.W. dye laser is tuned to a particular Ba* optical-pumping transition (denoted as line i), the product signal changes by a factor Ri. The observed signal in this case is given by Siax = RiSiFx or where nJ,i and nD,i are the appropriate relative Ba" number densities when line i is pumped.After dividing eqn (1) and (3) by cu3 and carrying out some algebra, we obtain the following final equations: ( Ri n 7" - n , ,i) ( a , / a,) + ( Ri n zff - n 2,i) ( a,/ a,) -I- ( Ri n "6' - n D,i) ( aD/ a3) = n3 ,i - Ri n ". ( 4) Eqn (4) is a set of equations whose number equals the number of optical pumping transitions, with the unknowns being the cross-section ratios ( a,/ a3), ( a'/ a3) and ( uD/ u 3 ) . In our earlier derivation' we referenced the spin-orbit-dependent cross-sections to the average over the incident, unpumped, spin-orbit state distribution. The present normalization to the cross-section for a specific reactant state (in this case the J = 3 spin-orbit level) is preferable since the reference cross-section is then not dependent upon experimental parameters, i.e.the spin-orbit state distribution. Since the number of Ba* lines optically pumped is greater than the number of unknowns, the solution of eqn (4) requires a linear regression procedure. However, theM. L. Campbell and P. J. Dagdigian 137 Table 2. Observed changes Ri in BaBr product fluorescence intensities upon Ba* optical pumping Ba" + HBr Ba" + CH,Br Pump transition 0" = 20 v" = 25 v" = 28 vtt = 30 vn = 25 v" = 30 3P:t3D, 0.97*0.02 0.93f0.02 0.91f0.02 0.86f0.02 0.93*0.02 0.93*0.02 'P: t ,D, 0.94 f 0.02 0.92 * 0.02 0.89 f 0.02 0.82 f 0.03 0.91 f 0.02 0.90 f 0.02 ,Po c 3 D 2 1.02f0.03 1.06f0.03 1.12f0.03 1.16k0.03 1.08f0.02 1.08f0.02 314 c3 D2 0.93 * 0.02 0.93 * 0.02 0.94 f 0.02 0.90 f 0.02 0.93 f 0.02 0.91 f 0.02 'P: t ,D3 1.08 * 0.02 1.05 f 0.02 1.08 f 0.02 1.11 f 0.02 1.07 f 0.03 1.09 f 0.03 'P: c ID2 0.89 f 0.02 0.91 f 0.02 0.91 f 0.02 0.95 f 0.02 0.92 f 0.02 0.94 f 0.02 F; c D3 0.95 f 0.02 0.93 f 0.02 0.94 f 0.02 0.88 f 0.02 0.89 * 0.02 0.89 f 0.02 Table 3.Observed changes Ri in BaCl product fluorescence intensities upon Ba" optical pumping 'P: c 3 D , 0.90st0.02 0.85f0.02 0.80f0.02 0.93f0.03 0.92f0.02 0.94f0.02 ' P ; t ,D, 0.86 f 0.02 0.81 f 0.02 0.74f 0.02 0.92 f 0.03 0.88 * 0.02 0.90 f 0.03 'P: t ,D2 1.10 f 0.03 1.18 f 0.03 1.26 f 0.03 1.10 f 0.03 1.08 * 0.03 1.07 f 0.03 ' P ; t ,D2 0.90* 0.02 0.87 f 0.02 0.88 f 0.02 0.91 f 0.03 0.90 f 0.02 0.90 f 0.03 3 P : c 3 D 3 1.08f0.02 1.15f0.03 1.15f0.03 1.13f0.03 l.llf0.02 1.11 f0.03 P: e D2 0.97 f 0.02 0.97 st 0.02 0.98 f 0.02 0.94 f 0.03 0.92 f 0.02 0.92 * 0.03 F , t D3 0.87 f 0.02 0.87 f 0.02 0.86 f 0.02 0.90 f 0.03 0.91 f 0.02 0.86 f 0.03 1 0 a Unassigned feature in the C 2111,2-X 'X+ Av = +1 sequence. present problem is more complicated than the usual least-squares fit,26 since there are uncertainties in both the 'dependent variable' ( y = n3,i - Rin;") and in the 'independent variables' x, = RinXff - n,,i ( a ! = 1,2, D ) .This type of fitting problem has been considered a number of times in the As discussed in detail previously,6 we have employed a simple iterative procedure given by Irvin and Quickenden3' wherein each pump transition is given a weighting of ( u ; + C , a t a ; ) - ' , where a, is the cross-section ratio (u,/u3) to be fitted.Initially, the a, in the weighting function are set to zero and successive iterations of eqn (4) use the a, values determined in the previous iteration until convergence is achieved (typically 2 or 3 iterations). The dependence of the reaction cross-section on the incident Ba(3D) spin-orbit state was investigated for a number of product BaX vibrational levels in the HBr, CH3Br, HCl and CH3Cl reactions. In most cases, the reagent gas pressure was maintained at <0.5 mTorr" in order to avoid collisional equilibration of the spin-orbit levels by intramultiplet mixing. In several of the reactions spin-orbit effects were also measured at higher pressures (up to 1.5 mTorr); no systematic change in the intensity ratios Ri were observed with pressure.Tables 2 and 3 present the observed ratios Ri [as defined by eqn (2)] of product fluorescence signals with the C.W. dye laser on and off for the various Ba" pumping transitions. The particular spectral features used to detect the different BaX vibrational levels are denoted by arrows in fig. 2-6. For the BaBr product from the Ba" + HBr and CH3Br reactions, bands in the C 211s,,z-X 2Ct Au = -1 sequence were employed for *1 Torr = 101 325/760 Pa.138 Ground-state Products from Reactions of Ba( 3D) Table 4. Observed changes R, in BaX product fluorescence intensities for low vibrational levels upon Ba" optical pumping Ba" + HBr transition VII = 5 utt= 10 V f t = 10 Ba* + HCI Pump 3 0 3 0 P , t 3 D , 1.00*0.04 0.97f0.04 1.06f0.07 P 2 t 3 D , 1.02*0.04 0.97f0.04 1.06*0.07 ' P : e 3D2 1.03 f 0.04 1.01 f 0.04 0.99 f 0.07 3 P ; e 3 D 2 1.00*0.04 1.01 f0.04 1.02k0.07 ' P ; c D3 1 .OO f 0.04 0.99 f 0.04 1.04 f 0.07 ' P : c ID2 1.06f0.04 1.08 k0.04 1.18 f 0.08 ' F ; e 3 D 3 1.14f0.05 1.02Zt0.04 1.18f0.08 " Contribution to the fluorescence signal from the Ba('S)+ HCI reaction subtracted out, as described in the text.Table 5. Relative reaction cross-sections" for formation of specific BaX product vibrational levels as a function of the reagent Ba atomic state reaction BaX state O ( ~ D ~ ) / ~ ( ~ D ~ ) ~ r ( ~ D ~ ) l o ( ' D ~ ) ~ ( ' D ) / u ( ~ D ~ ) Ba+ HBr utl = 20 u" = 25 utt = 28 u" = 30 Ba+ CH3Br u" = 25 u" = 30 Ba + HCl uJr = 30 utt = 35 u" = 40 u" = 40 VII = 45 Avi/* = + 1 Ba + CH3Cl 1.52 f 0.21 1.70 f 0.22 2.08 f 0.26 2.81 f0.35 1.83 f 0.23 1.99 f 0.26 2.26 f 0.28 3.92 f 0.44 4.13 f 0.56 2.13 f0.29 2.16 f 0.28 1.98 f 0.26 1.33k0.14 1.26 f 0.13 1.32f0.15 1.56 f 0.18 1.32 f0.14 1.45f0.15 1.48 f0.16 1.83 f 0.23 1.86 i 0.26 1.55 f 0.17 1.53 f0.17 1.50 f 0.16 0.84 f 0.14 0.73 f 0.13 0.77 f 0.14 0.46 f 0.14 0.60 f 0.13 0.55 f0.13 0.42 f 0.12 0.35 f 0.1 5 0.24 f 0.17 0.61 f0.14 0.64 f 0.13 0.48 f 0.13 Estimated uncertainties are la detection of vibrational levels v" ranging from 20 to 30.The BaCl C 2111,2-X 'X+ Av = -1 sequence was used for detection of levels v" from 30 to 45 in the Ba" + HC1 and CH3Cl reactions. For the latter, an unassigned feature in the BaCl C 211-X 'X+ Au = +1 sequence was also studied. It can be seen in tables 2 and 3 that the product fluorescence signals change significantly upon alteration of the Ba( 'DJ ) spin-orbit populations.The effect of optical pumping was also investigated for several low BaX vibrational levels in the Ba" + HBr and HC1 reactions. In these cases, it was necessary to correct for the Ba('S) contribution to the reaction product signals. These data, which were corrected in the same way that the Ba( IS) contribution was subtracted from the excitation spectra, are given in table 4. We find for these levels that changing the Ba(3DJ) spin-orbit population by optical pumping has a small effect on the product densities. The data in tables 2 and 3 were used with eqn (4) to derive spin-orbit-dependent cross-sections for formation of the various BaX product vibrational levels.The atomic Ba state populations nyff, nJ,j etc. were measured by laser fluorescence detection pre- viously in our study of spin-orbit effects in the chemiluminescence channels of otherM. L. Campbell and P. J. Dagdigian 139 Ba" reactions. These relative populations, which are listed in table 1 of ref. ( 6 ) , were employed in the present analysis. The derived spin-orbit-dependent cross-section ratios are presented in table 5 . It can be seen from table 5 that, at least for formation of the higher BaX vibrational levels, there is a significant dependence of reactivity on incident Ba('D,) spin-orbit level, with an ordering J = 1 > J = 2 > J = 3 and as much as a factor of 4 difference in the J = 1 vs. J = 3 cross-sections. The reaction cross-sections for the ' D2 level are also smaller than for any of the 'DJ levels.It should also be noted that the spin-orbit dependence varies with product vibrational level. This is evident for the Ba" + HBr and HCl reactions over the range of u" = 20 to 40. We have not derived relative spin-orbit- dependent cross-sections from the data in table 4 for the low vibrational levels because of the essentially negligible change of the fluorescence upon optical pumping for these levels. It is nevertheless clear that the spin-orbit dependence is at best very small and much less than for the higher product levels. For the CH'Br and CH3Cl reactions, the variation of the spin-orbit effect us. product vibrational level is within the estimated experimental uncertainties in the relative cross-sections. Discussion The present results on the relative cross-sections for formation of ground-state BaX product from the various atomic reactant spin-orbit levels for the Ba('D) + HX, CH3X (X=Cl, Br) reaction contrast sharply with our previous observations' on the chemiluminescence channels of a number of other Ba('D) reactions.For formation of ground-state product, the lowest-energy J = 1 spin-orbit level possesses the largest cross-sections, with successively smaller values for J = 2 and J = 3; however, an exactly opposite ordering of reactivity was found for the chemiluminescence channels.' Exactly analogous spin-orbit effects were also observed for Ca(3P0) and Sr('PPO) and have been interpreted with the aid of a pseudo-quenching model calculation of Alexander.' Because of the congested nature of these alkaline-earth monohalide chemilumines- cence spectra and our relatively low spectral resolution, we were able to study spin-orbit effects for production of diff erent excited product electronic states, but without resolution of individual vibrational levels.Our previous work on the ground-state product channel of several Ca(3P0) and Sr('Po) reactions was confined to a few wavelengths of the fluorescence excitation laser, without identification of the specific vibrational level detected. In the present experiment, we have been able to determine the spin-orbit effect for production of a wide range of BaX product vibrational levels. We have found for the Ba(' D) + HX reactions that the spin-orbit dependence varies from essentially no detectable difference on reactivity of the different spin-orbit levels for formation of low product vibrational levels to a factor of 3 to 4 variation in 3D, us.'D3 cross-sections for the highest vibrational levels studied ( u = 30 and 40 for the HCl and HBr reactions, respectively). The analogous reactions of ground-state Ba( ' S ) with hydrogen halides have been extensively studied previously, both experimentally*0~~*~3' and the~retically.'~ These reactions exemplify a kinematically constrained class involving a reactive encounter of a heavy atom (H) with a heavy-light molecule (H'L) to produce a heavy-heavy diatomic product (HH') and a light atom ( L).37338 One consequence of the conservation of angular momentum I is that the reagent orbital angular momentum is channelled into rotational angular momentum j ' of the diatomic product.Noda et ~ 1 . ' ~ have taken advantage of this constraint on the Ba+HI reaction to gain information on the impact parameter dependence of reactivity from measurement of the BaI product rotational state distribu- tion. While the opacity function could not be definitively determined because of the140 Ground-state Products from Reactions of Ba( 3D) width of the initial relative velocity distribution, they were able nonetheless to conclude that the specific opacity for BaI( = 8) was strongly peaked around a limited range of impact parameters. Siege1 and S~hultz’~ have carried out classical trajectory calculations on several model potential-energy surfaces for the Ba+HCl and HBr reactions.They do indeed find a very strong correlation between I and j ’ , as expected from kinematic considerations. Moreover, they also observe a correlation of the product vibrational energy with incident impact parameter. For LEPS surfaces, but not for hyperbolic map function (HMF) surfaces, the product recoil energy was found to be independent of product vibrational energy, and hence a constant for a given reaction. Noda and Zare3’ have invoked the assumption of constant product recoil energy (CPR) to develop a simple model for the kinematically constrained H + H’L reactions. This so-called CPR model predicts a bell-shaped product vibrational state distribution, wherein the specific opacity functions for low and high vibrational levels are sharply peaked at large and smail impact parameters, respectively.They show that this model can be used to interpret the product vibrational state distributions for the Ba+HF35 and HI34 reactions, as well as the less-constrained Ba + CH3Br22 and CF314’ reactions. In view of a similar kinematic constraint on the excited state Ba(3D) + HCl and HBr reactions, we may thus associate small and large incident impact parameters with the formation of products of high and low vibrational excitation, respectively. From our experimental results, we hence conclude that differences in the reactivity of the 3D, incident spin-orbit levels are present only for small and not large impact parameter collisons. This would suggest that in large impact parameter collisions the identity of the incident spin-prbit state is lost‘ before the reaction occurs, i.e.before the ionic- covalent crossing point is reached, whereas such mixing in the entrance channel does not occur in near head-on encounters of small impact parameter. In other words, large incident orbital angular momentum appears to induce non-adiabatic transitions in the entrance channel. We can put this explanation on a quantitative basis with the help of the pseudo- quenching model of Alexander.’ Spin-orbit effects in chemical reactions, as well as the simpler non-reactive intramultiplet mixing process, is most conveniently described in a Hund’s case ( e ) basis,41 wherein the good quantum numbers include 1, L, S, J and J (nuclear orbital, electronic orbital, electronic spin, total electronic and total overall angular momenta, respectively). While the Hamiltonian is diagonal in J and the elf symmetry: it may be further diagonalized for each value of the interparticle separation R to yield fully adiabatic potential curves.In the potential models employed by Alexander’ for Ca(3P0) + C12, adiabatic curves correlating with the J = 0 and 1 asymptotic P spin-orbit levels, but not for J = 2, undergo an ionic-to-covalent transition [see fig. 1 and 2 of ref. ( 9 ) ] . Thus, reaction from J = 2 requires, within this model, a non-adiabatic transition. As discussed before in connection with a semiclassical treatment of collisional intramultiplet m i x i r ~ g , ~ ~ . ~ ~ the strength of such non-adiabatic transitions are related to matrix elements of operators d/dR and d2/dR2, which are components of the kinetic- energy operator, between the non-adiabatic wavefunctions.The former first-derivative operator is usually the dominant term.43 If we express the adiabatic eigenvectors in terms of the case ( e ) functions, i.e. 3 0 InJM) = c An,Jl(R)IJ1JM) ( 5 ) J. 1 then the relevant non-adiabatic matrix elements can be expressed asM. L. Campbell and 0.1 I I I I I I 1 14 16 18 P. J. Dagdigian 141 14 16 18 Rlbohr Fig. 7. Dependence on interparticle separation of the non-adiabatic coupling matrix element G;,,, ( R ) for transitions between levels adiabatically correlating asymptotically to reagent atomic spin-orbit levels in the pseudoquenching model of Alexander, described in ref. (9) (potential 11).Results for two values of the total angular momentum J [ ( a ) 5 , ( b ) 2501 are given for the f-labelled symmetry levels. This matrix can be obtained by numerical differentiation of the matrix of eigenvectors. [The label M in eqn ( 5 ) denotes the space-fixed projection of 5.3 We present in fig. 7 values of G,,,”(R) for two values of J for the f-labelled parity levels between eigenvectors which correlate asymptotically to the Ca( 3P0) spin-orbit levels. These were calculated for one of the potential models considered by Alexander’ (his model 11). With this model, the ionic-covalent crossing occurs near 13 bohr. It should be noted there are 3 f-labelled case ( e ) eigenvectors for J = 2[ I = J, J * 21, while there is only one each for J=O and l[Z=J]. It can be seen that for J = 5, which is representative of a small impact parameter collision, G,,,J(R) is quite small for most pairs ( n ’ , n ) levels and is appreciable outside the region where charge transfer occurs only between two pairs of levels.By contrast, Gnf,”( R ) attains appreciable size between most covalent adiabatic potential curves for larger J, although the absolute magnitude of the largest non-adiabatic matrix element is somewhat reduced. Such a dependence of the non-adiabatic coupling on the total angular momentum, and hence impact parameter, has already been documented for the J = 0 to 1 intramultiplet transition in Ca(’P”) non-reactive collisions; in this case, mixing can occur only though the coupling of electronic and nuclear angular rn~rnenta.~’ The non-adiabatic matrix elements in fig.7 attain magnitudes comparable to those previously calculated for the Ca(3P0)-He system, in which significant collisional intramultiplet mixing This pseudo-quenching model thus strongly suggests the importance of the nuclear orbital angular momentum, and hence impact parameter, in determining the magnitude of the spin-orbit effect on reactivity. Of course, Alexander’s model’ is inadequate to describe Ba(.lD) + HX reactions accurately both because of the model’s atom-atom character and differences baetween the CaCl, model potentials employed therein and the correct BaHX potentials. With regard to the latter, even in a quasidiatomic approxi- mation, three covalent potential curves (Z, lI and A) will arise for an atom in a D state [e.g.Ba(’D)], as opposed to two for a P-state atom [e.g. Ca(’PO)]. Nevertheless, our142 Ground-state Products from Reactions of Ra( 3D) qualitative conclusions about the role of impact parameter should still be valid for real systems since the number of covalent potential-energy surfaces will almost always be less than the number of ionic surfaces. [See table V of ref. (6) for an enumeration of potential-energy surfaces for Ba reactions.] It would be very interesting to test these ideas quantitatively by further model calculations using realistic potentials. From an experimental point of view, the dependence of the spin-orbit effect upon chemical reactivity as a function of product vibrational state is expected to be manifest in other reactions and or in other product channels, such as chemiluminescence.We are indebted to Millard Alexander for conversations about the theoretical interpreta- tion of our results. Support for this work was provided by the National Science Foundation. References 1 P. J. Dagdigian and M. L. Campbell, Chem. Rev., 1987, 87, 1. 2 H-J. Yuh and P. J. Dagdigian, J. Chem. Phys., 1983, 79, 2086; 1984, 81, 2375. 2 N. Furio, M. L. Campbell and P. J. Dagdigian, J. Chem. Phys., 1986, 84, 4332. 4 M. L. Campbell, N. F. Furio and P. J. Dagdigian, Laser Chem., 1986, 6, 391. 5 M. L. Campbell and P. J. Dagdigian, J. Am. Chem. SOC., 1986, 108, 4701. 6 M. L. Campbell and P. J. Dagdigian, J. Chem. Phys., 1986, 85, 4453. 7 H.-J. Yuh and P. J. Dagdigian, Phys. Rev. A , 1983, 28, 63. 8 P. J. Dagdigian, in Gas-phase Chemiluminescence and Chemi-ionization, ed.A. Fontijn ( North-Holland, 9 M. H. Alexander, in Gas-phase Chemiluminescence and Chemi-ionization, ed. A. Fontijn ( North-Holland, Amsterdam, 1985), p. 203. Amsterdam, 1985), p. 221. 10 J. A. Irvin and P. J. Dagdigian, J. Chem. Phys., 1980, 73, 176. 11 M. G. Littman and H. J. Metcalf, Appl. Opt., 1978, 17, 2224. 12 M. G. Littman, Opt. Lett., 1978, 3, 138. 13 P. Grundevik, H. Lundberg, L. Nilson and G. Olson, 2. Phys. A , 1982, 306, 195. 14 American Institute ofPhysics Handbook, ed. D. E. Gray (McGraw-Hill, New York, 3rd edn, 19721, pp. 15 M. W. Chase Jr, C. A. Davies, J. R. Downey Jr, D. J. Frurip, R. A. Macdonald and A. N. Syverud, 16 R. C. Estler and R. N. Zare, Chem. Phys., 1978, 28, 253. 17 J. W. Cox and P. J. Dagdigian, J. Chem. Phys., 1983, 79, 5351. 18 C. E. Moore, Atomic Energy Levels, Natl. Bur. Stand. Ref. Data Ser., Natl. Bur. Stand. No. 35 (U.S.G.P.O., Washington, D.C., 1971), vol. 111. 19 D. L. Hildenbrand, J. Chem. Phys., 1977, 66, 3526. 20 H. W. Cruse, P. J. Dagdigian and R. N. Zare, Faraday Discuss, Chem. SOC., 1973, 55, 277. 21 A. Siegel and A. Schultz, J. Chem. Phys., 1980, 72, 6227. 22 T. Munakata, Y. Matsumi and T. Kasuga, J. Chem. Phys., 1983, 79, 1698. 23 A. Schultz and A. Siegel, J. Mol. Spectrosc., 1979, 77, 235. 24 R. W. Solarz and S. A. Johnson, J. Chem. Phys., 1979, 70, 3592. 25 W. Schmidt, A. Siegel and A. Schultz, Chem. Phys., 1976, 16, 161. 26 P. R. Bevington, Data Reduction and Error Analysis for the Physical Sciences (McGraw-Hill, New York, 27 A. Madansky, J. Am. Stat. Assoc., 1959, 54, 173. 28 M. O’Neill, I. G. Sinclair and F. J. Smith, Comput. J., 1969, 12, 52. 29 P. A. P. Moran, J. Multivar. Anal., 1971, I, 232. 30 D. R. Powell and J. R. Macdonald, Comput. J., 1972, 15, 148. 31 J. A. Irvin and T. I. Quickenden, J. Chem. Educ., 1983, 60, 71 1. 32 J. G. Pruett and R. N. Zare, 1. Chem. Phys., 1976,76, 1774. 33 A. Torres-Filho and J. G. Pruett, J. Chem. Phys., 1980, 72, 6736; 1982, 77, 740. 34 C. Noda, J. S. McKillop, M. A. Johnson, J. R. Waldeck and R. N. Zare, J. Chem. Phys., 1986,85, 856. 35 A. Gupta, D. S. Perry and R. N. Zare, J. Chem. Phys., 1980, 72, 6237. 36 A. Siegel and A. Schultz, J. Chem. Phys., 1982, 76, 4513. 37 D. R. Herschbach, Adv. Chem. Phys., 1966, 10, 319. 38 N. H. Hijazi and J. C. Polanyi, J. Chem. Phys., 1975, 63, 2249; Chem. Phys., 1975, 11, 1. 8-46. JANAF Thermochemical Tables, 3rd edn, J. Phys. Chem. Ref: Data, 1985, 14, suppl. 1. 1969).M. L. Campbell and P. J. Dagdigian 39. C. Noda and R. N. Zare, J. Chem. Phys., 1987, 86, 3968. 40 M. A. Johnson, J. Allison and R. N. Zare, J. Chem. Phys., 1986, 85, 5723. 41 G. Herzberg, Spectra of Diatomic Molecules (Van Nostrand, Princeton, 2nd edn, 1950). 42 E. E. Nikitin, J. Chem. Phys., 1965,43, 744; Adu. Chem. Phys., 1975, 28, 317. 43 M. H. Alexander, T. Orlikowski and J. E. Straub, Phys. Rev. A, 1983, 28, 73. 143 Received 1 1 th May, 1987

ISSN:0301-7249    

DOI:10.1039/DC9878400127

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1987, 84, 145-157 Reactive Scattering of Electronically Excited Alkali-metal Atoms with Molecules J. M. Mestdagh,"? B. A. Balko, M. H. Covinsky, P. S. Weiss,S M. F. Vernon,§ H. Schmidt7 and Y. T. Lee Materials and Molecular Research Division, Lawrence Berkeley Laboratory, and Department of Chemistry, University of California, Berkeley, California 94720, U.S.A. Representative families of excited alkali-metal reactions have been studied using a crossed-beam apparatus. For those alkali-metal-molecule systems in which reactions are also known for ground-state alkali metal and involve an early electron-transfer step, no large differences are observed in the reactivity as Na is excited. More interesting are the reactions with hydrogen halides (HCl); it was found that adding electronic energy into Na changes the reaction mechanism. Early electron transfer is responsible for Na(5S, 4 0 ) reactions, but not for Na(3P) reactions.Moreover, the NaCl product scattering is dominated by the HC1- repulsion in Na(5S, 4 0 ) reac- tions, and by the NaCl-H repulsion in the case of Na(3P). The reaction of Na with O2 is of particular interest since it was found to be state-specific. Only Na(4D) reacts, and the reaction requires restrictive constraints on the impact parameter and the reactants' relative orientation. The reaction with NO2 is even more complex, since Na(4D) leads to the formation of NaO by two different pathways. However, the identification of NaO as product in these reactions has yet to be confirmed. '' For a given elementary chemical reaction, specific forms of energy (i.e.translational, vibrational, rotational, and electronic) deposited in reactants are known to affect reactive scattering processes in different ways. Adding electronic energy to reactants is of special interest since in this way a relatively large amount of energy can be deposited in a single excitation step and, more importantly, the potential-energy surface on which the reaction is initiated can be selected by exciting reactants to states of different symmetries. The reactivity and reaction dynamics of such excited reactants often differ substantially from those of the corresponding ground-state molecules.' Reactive scattering of alkali-metal atoms with molecules has been studied extensively over the past two decades.* A wide variety of experimental techniques has been used to determine the dynamics of these reactions, as well as the effect of reactant translational, vibrational and rotational e~citation.~ Until recently the effect of electronic excitation remained The reactions of excited Na atoms with simple molecules have been systematically investigated in our laboratory using the crossed-molecular-beam m e t h ~ d .~ - ~ In this paper some examples of excited Na atom reactions will be presented to illustrate the effect of electronic excitation on sodium reactivity. t Permanent address: Service de Physique des Atomes et des Surfaces, CEN Saclay, 91 191 Gif sur Yvette $ Present address: AT & T, Bell Laboratories, f 1C-402, 600 Mountain Avenue, Murray Hill, NJ 07974, fi Present address: Department of Chemistry, Columbia University, New York, NY 10027, U.S.A.ll Permanent address: Braun A.G. Forschung, Frankfurter Str. 145, D6242, Kronberg, Federal Republic Cedex, France. U.S.A. of Germany. 145146 Reactive Scattering of Excited Alkali-metal Atoms molecular beam source Na beam secondary rotating molecular polarization detector I mass spectrometer b e a y rotator Fig. 1. Schematic diagram of the crossed-molecular-beam apparatus. First, systems are considered in which the reaction is exothermic and is initiated by an electron transfer from a sodium atom to a molecule; a representative example of this is the reaction In reactions ( l ) , n,l represents the electronic state of Na: ground-state 3 s or excited states [e.g.Na(3P)I. The AH: is for the reaction of Na(3S). The second type of reaction is where ground-state Na is known to react, but a simple electron-transfer model cannot be used to account for observed features of the collision dynamics. The chosen example is where the excited state n,l is 3P, 5s or 4 0 . The points which will be considered about processes ( 1 ) and (2) concern changes in the Na reactivity as well as changes in the reaction dynamics that are associated with depositing electronic energy into the Na. The remainder of the examples concern reactions which are substantially endothermic and energetically unfavourable for ground-state Na. The goal is to understand if and by what mechanism electronic excitation is able to turn on the reaction.One example that will be discussed is Na( n,l) + C1, --* NaCl+ C1, AH: = -40.4 kcal mol-'. ( 1 ) Na( n,l) + HCl -+ NaCl+ H AH: = 4.7 kcal mol-' (2) Na(n,l)+02 --* NaO+O, AH: = +58 kcal mol-'. (3) Na( n,l) + NO2 --* NaO + NO, (4) The somewhat more complicated process AH: = +9 kcal mol-' will be described in greater detail. Experiment and Analysis The experimental set-up has been described e l ~ e w h e r e . ~ - ~ . ~ Briefly, two supersonic beams and one or two lasers are crossed orthogonally under single-collision conditions. A mass-spectrometric detector rotates about the collision region in the plane defined by the two reactant beams as shown schematically in fig. 1. Several types of experimentsJ. M. Mestagh et al. 147 1 h m Y .- c $ 0.5 v - c M .* m 0 0 30 60 00 laboratory angle/" Fig.2. NaCl product angular distribution for the reaction Na(3S,3 P ) + C12 at 19 kcal mol-' collision energy. The solid lines are the best fit to the data generated from centre-of-mass angular and recoil energy distributions. 0, 3 P ; 0, 3s. are performed: (i) product angular distributions are measured by rotating the detector, (ii) product velocity distributions are measured by a time-of-flight technique, (iii) the dependence of the chemical reactivity on the nature of the excited state of sodium is investigated for Na(3P), Na(4D) and Na(SS), and (iv) the effect of collision energy is studied by changing the beam velocities. The experimental angular and velocity distributions are fitted using a forward convolution method which varies the assumed centre-of-mass angular and translational energy distributions and averages over the variation in experimental conditions.For cases in which the reaction cross-section is found to depend strongly on collision energy, the energy dependence of the reaction cross-section is specifically taken into account in the convolution process. Results Na(n,l) + C12 + NaCl + C1 Angular distributions of the NaCl products were measured for Na(3S,3P), at three nominal collision energies: 3, 6 , and 19 kcal mol-'. Typical results are shown in fig. 2 for 19 kcal mol-' collision energy. The results for reactions of Na(3S) agree with the general understanding of ground-state alkali-metal-halogen reactions.*l They show predominant forward scattering and narrow recoil energy distributions.The new feature brought out by fig. 2 is that no large difference is observed between the product angular distributions associated with the reactions of Na(3S) and Na(3P), even though the electronic excitation lowers the ionization potential of Na by 2.1 eV.148 Reactive Scattering of Excited Alkali-metal Atoms 0 Table 1. Features of the centre-of-mass distributions fitted to the laboratory angular distributions for Na(3S,3 P) + CI, reactions 1 i i ' + 1 I i l collision % of forward peak recoil Na level energy" scattering energy 4 3 W 4 3 S ) 3 s 6.0 3P 6.0 3s 19.0 3 P 19.0 83 76 85 82 0.6 1.2 0.8 1 .o 1.58 1.16 " Energies are in kcal mol '. Fig. 3. NaCl product angular distributions for the reaction Na(3S,3 P,5S,40) + HCI at 5.6 kcal mol-' collision energy.The solid lines are just for guiding the eyes. 0, 3 s ; 0, 3P; A, 5s; m, 4 0 . The predominance of forward scattering and narrow recoil energy distributions allows one to derive centre-of-mass distributions directly by fitting the laboratory angular distributions on the basis of an assumed centre-of-mass distribution." Best fits are shown in fig. 2 for a 19 kcalmol-' collision energy. A summary of features of the centre-of-mass distributions is found in table 1. Na( n,l) + HCl + NaCl + H This system was studied at collision energies of 3.4, 5.6 and 16.3 kcal mol-' for reactions with Na( 3S,3P,5S,4D). The laboratory angular distributions of the product NaCl are shown in fig. 3 for a 5.6 kcal mol-' collision energy. As can be seen, there are important changes in the Na reactivity when switching from Na(3S) to Na(4D).J.M. Mestagh et al. 149 n 0 30 60 90 laboratory angle/" Fig. 4. NaO product angular distribution for the reaction Na(4D) + O2 at 18 kcal mol-' collision energy. The solid lines are the best fit to the data generated from centre-of-mass angular and recoil energy distributions. These angular distributions, along with the measured laboratory velocity distributions of the product NaCl, allow one to extract a centre-of-mass NaCl flux distribution. As discussed in ref. (5), (7), (9), all the reactions of excited Na with HCl exhibit predominant backward scattering with low recoil energy; these characteristics are most apparent in the reactions of the two highest excited states, 5 s and 4 0 .Na(n,l)+O, -+ NaO+O Process (3) for Na (3S,3P,5S,4D) was studied over a wide range of collision energies: 8-22 kcal mol-'. The most interesting feature of the results is the state specificity; no reaction was found for Na( 3S,3P,5S). Backwards-scattered product was observed in the Na(4D) reactions when the average collision energy was >11 kcal m ~ l - ' . ~ The non-reactivity of Na(5S) is striking because process (3) has about the same exothermicity with either Na(5S) as the reactant or Na(4D) (36.9 us. 41.5 kcal mol-I). The high activation energy for the reaction of Na(4D) is also surprising in view of the large exothermicity of the reaction. Representative results are shown in fig. 4 and 5 for the reaction of Na(4D) at a collision energy of 18 kcal mol-'.Na(n,l) + NOz -+ NaO + NO Angular distributions for the Na+ NO, reaction were measured for Na(3S,3P,5S and 4 0 ) . The detected product in these experiments was the Naf that had fragmented from the NaO during electron bombardment; this means that all the recorded distributions have a Na elastic scattering contribution. No Na(3S) or Na(3P) reaction was observed above this background. The Na(5S) and Na(4D) states, however, do react with NOz,150 1 - n 2 0.5 c1 0 - Reactive Scattering of Excited Alkali-metal Atoms - recoil energy/ kcal mol-' 0 0.5 1 1.5 2 I ' I I I I 1 0 90 O / O 180 1 h w a, 0.5 - 0 Fig. 5. Best-fit centre-of-mass angular distribution (full line, left and lower scale) and recoil energy (dashed line, right and upper scale) of NaO in Na(4D) +O, reactions at 18 kcal mol-' collision energy./ I / / 0 30 60 90 laboratory scattering angle/" Fig. 6. NaO product angular distributions for the reaction Na(3S,3P,5S,4D) + NOz at 19 kcal mol-' collision energy. 0, 3P; ., 4 0 ; 0, 5s.J. M. Mestagh et al. 151 Na b+++H+l 1 x lo5 cm s-' Fig. 7. NaO product centre-of-mass contour plot for the reaction of Na(4D) + NOz at 19 kcal mol-I collision energy. as can be seen in fig. 6. The product angular distributions for both these states are not the same. The most obvious distinction is in the reaction cross-sections which are in a ratio of ca. 12 to 1 in favour of Na(4D). The products also scatter differently. The Na(4D) product is more backwardly scattered and has a narrower energy distribution, as evidenced by the sharper peak.The nominal collision energy was varied between 5.5 and 19 kcal mol-' so as to determine the Na(4D) + NOz reaction threshold. The measured speed ratios of the reactant beams, which ranged from 3.2 to 5.2, allowed the spread in these energies to be calculated while the signal levels in the measured angular distributions gave an estimate of the relative reaction cross-sections at each nominal energy. A cross-section vs. collision energy plot could then be constructed to match these observations. An adequate fit was obtained when a step function with a threshold of 20 kcal mol-' was selected. It must be mentioned, however, that this method is not very sensitive to the exact form of this function. Product velocity distributions were measured for the Na(4D) + NOz reaction at a collision energy of 19 kcal mol-'.The laboratory angular and velocity distributions were calculated and fitted using an assumed product recoil energy distribution [ P( E ) ] and centre-of-mass angular distribution [ T ( O)]. The best T ( 0) and P( E ) generated the contour map shown in fig. 7. Two different NaO products are apparent in fig. 7. One (product 1) is very similar to the NaO observed in process (3). It is backward-scattered, has very little energy in translation and has a very narrow distribution in velocity. The other type of NaO (product 2) is very different. It is forward-scattered, has much more energy in translation, and the velocity distribution is much broader. The relative cross-sections for formation152 Reactive Scattering of Excited Alkali-metal Atoms of each NaO have been estimated to be almost equal.However, product 1 is kinematically enhanced for detection in the laboratory reference frame, and is therefore detected preferentially in the present experiment. Discussion In the reaction of alkali-metal atoms with halogen-containing molecules, some essential features can be modelled by the electron transfer 'harpoon' mechanism. In this type of reaction, an ion-pair complex is formed early by transfer of the Na valence electron into the lowest unoccupied molecular orbital (LUMO) of the molecule. This electron transfer is normally assumed to be rapid compared to the nuclear motions. The negative ion is formed on a dissociative repulsive region of the potential. When the ion dissociates, the negative fragment combines with Na+ via coulombic forces to form the reaction product.* Reactions of Excited Na Atoms with Halogen Molecules The prominent feature seen in fig. 2 and table 1 is that no large difference appears in the collision dynamics and, consequently, in the reaction mechanism when Na is excited.This observation is an indication that an electron transfer of the Na electron to the LUMO of the molecule is the initial step.2 The shape of the LUMO is what determines the collision dynamics; it has been observed that the dynamics are very different when switching from halogen to organic halide molecules, but are only weakly sensitive to the identity of the alkali-metal reactant.2 This accounts for the similarity in collision dynamics for reactions with different Na states; exciting Na affects the valence electron orbital, but not, of course, the LUMO of the molecule.Experiments indicate that the total reactive cross-section of process ( 1 ) does not increase very much when Na is excited. This result is somewhat surprising. The higher the Na excitation, the lower the ionization potential; greater Na reactivity would then be expected, since the electron transfer initiating the reaction should be possible at larger Na-molecule distances. However, the shift to large interatomic distances of the crossing between covalent and ion-pair surfaces reduces the interaction between them. This makes the excited states of Na less reactive than they would otherwise be. Such a reduction of the coupling element between covalent and ion-pair surfaces has been demonstrated in a recent calculation using the DIM method on alkali-metal-Br, sur- faces." Apparently, these two factors, the earlier electron transfer and the lower coupling, compensate each other to some extent in these reactions. No large change is observed in the centre-of-mass recoil energies of NaCl produced in reaction ( 1 ) as Na is excited.The additional 48.4 kcal mol-' corresponding to excitation of Na to the 3 P level is thus reIeased into internal excitation of the product NaCl. This would suggest ion-pair formation occurring at a larger Na-C1, distance, which leads to the formation of more highly vibrationally excited NaCl. Reactions of Excited Na Atoms with Hydrogen Halides The dominant features of the excited Na+ HC1 reactions are the decreasing product recoil energy and the increasing reactive cross-section with increasing electronic energy.This can be observed directly in fig. 3 and has been checked quantitatively by considering the centre-of-mass product angular and velocity distribution. For the reaction of Na(5S, 4 0 ) most excess energy goes into the vibrational excitation of NaCl, in contrast to the reaction of Na(3P), where a large fraction of this energy goes into the translational energy of the products. Process (2) thus provides a situation where the reaction dynamics change as the electronic excitation of the reactant is increased.J. M. Mestagh et al. 153 Table 2. Covalent ion-pair non-adiabatic curve-crossing radii for various electronic states of Na for the Na+ HCl systema ionization crossink radius Na level potential/ eV /A 3s 5.1 3P 3.0 5s 1 .o 4 0 0.8 2.4 3.7 7.8 8.6 ~~ a The electron affinity of HCl is taken to be -0.82 eV.14 Reactions of ground-state alkali-metal atoms with hydrogen halides have never been adequately explained in terms of electron transfer ‘harpoon’ models.The electron transfer in these cases does not take place at a long distance because hydrogen halides have negative electron affinities.” However, excited Na atoms have much lower ioniz- ation potentials. This makes early electron transfer possible (see table 2), so such ‘harpoon’ models should provide a useful framework to discuss reactions of excited Na atoms. HCl molecules are known to be dissociated into C1- + H by low-energy elec- t r o n ~ , ~ ~ thus a dissociative electronic attachment to HCl is expected to initiate the reactions of Na( 3 P, 5 S, 4 0 ) . Let us consider the reaction of Na(5S,40) first.Based on the covalent and ion-pair curve crossing, the electron transfer initiating the reaction occurs at large Na-HCl separations. The H atom departs very quickly after the electron transfer. The Na’ and C1- ions are then left at a separation approximately equal to the electron jump radius. NaCl is thus formed with a large vibrational excitation, and the hydrogen atom is no longer present to release the excess energy. The early departure of H is thus able to explain the low product recoil energy found experimentally in reaction (2) for Na(5S, 4 0 ) . Turning to Na(3P), the electron transfer occurs at a smaller distance, which is less than the sum of van der Waals radii of HCl and Na, and the stages of electron transfer, H departure and NaCl formation, are probably not well separated.The dynamics of H atom departure in this reaction are no longer dominated by the dissociation of HCl-. When the electron transfer occurs at a small distance between’HCI and Na, the electronic configuration is closer to HCl--Na+. The repulsive departure between H and NaCl and the stronger coupling among all three atoms are thus responsible for a large fraction of the excess energy in reaction (2) for Na(3P) going into product recoil energy. When the electron is to transfer at a long distance, a model such as the direct interaction model with products distributed as in photodissociation (DIPR-DIP) should be able to describe the reaction dynamics well.14 This model characterizes the reaction by sequential two-body steps: electron transfer from Na to HCl, departure of H from C1-, and, finally, association of Na+ and C1- to form NaCl.The above discussion for process (2) suggests that the DIPR-DIP model should explain Na(5S,4D) reactions well, but not the reaction of Na(3P) with HCl. This has been confirmed more quantita- tively in ref. ( 5 ) , (7) and (9). Reactions of Excited Na Atoms with O2 Molecules The most prominent feature in reaction (3) is that Na(40) reacts with 02, but not Na(5S). Since the radiative decay of the 5s state populates the 4P,4S. . . levels, it is clear that these levels also do not react to produce NaO.154 Reactive Scattering of Excited Alkali-metal Atoms The strong back scattering of the NaO products with respect to the incoming Na atoms is evidence of a direct reaction (i.e.no long-lived collision complex is formed). The especially narrow NaO centre-of-mass angular range seen in fig. 5 suggests that there are restrictive constraints on the impact parameter and relative orientation required for the reaction to proceed. The narrow product recoil energy distribution in fig. 5 shows that very little of the excess energy of this reaction goes into translation. It has been proposed in ref. (6) that electronic excitation of NaO, 0 or both carries away most of the excess energy. Alexander has derived analytic forms for the lowest non-adiabatic potential surfaces of the Na+-Oy system.” Two reactive pathways have been proposed in this work as possibly involved in Na + O2 reactions: and Na+02 -+ Na+...O,(X’II,) -+ NaO+O ( 5 ) Na+O, -+ Na+-..0,(A21Z,) + NaO+O It was suggested in ref.(6) that pathway (5) is probably not involved in the reaction since this ‘harpoon’-like pathway would be unlikely to be associated with a large activation barrier as was observed experimentally for process (3). Recently, calculations have been performed where the reaction pathways (5) and (6) are included in a multiple ionic-covalent curve-crossing model. l 6 These calculations point out that pathway ( 5 ) is actually associated with much smaller reactive cross-sections than pathway (6). The competition of the inelastic channel Na(4D+ n,l) with the reactive channel along pathway (5) is one of the main reasons why (5) makes such a small contribution to the reaction.Consistent with experimental results, the cross-section associated with pathway (6) has a large threshold energy, and involves impact parameters < 1 A. This makes pathway (6) a likely candidate to explain process (3). It is worthwhile to recall that dissociative electronic attachment on 0, molecules giving 0.- + 0 fragments is a resonant process which proceeds through the same A *nu level of 0, as that involved in reactive pathway (6).” Reactions of Excited Na Atoms with NO2 Molecules Process (4) is different from all the others that have been discussed so far in that two products are formed from the reaction of Na(4D).One of these products, product 1, is very similar to the NaO formed in reaction (3). Both exhibit backward scattering and have a small, narrow recoil energy distribution. This suggests that, as with the oxygen reaction, product 1 formation is a direct reaction with certain strict impact-parameter and orientation requirements. The other product observed in the Na(4D)+N02 reaction, seen in fig. 7, is very different from product 1. This is most apparent in the angular distributions. Product 2 is forward scattered suggesting a direct reaction with a large impact parameter. The recoil energy distributions also differ. Product 2 is formed with a much broader distribution of recoil energies and the mean translational energy of this distribution is 5.5 kcal mol-’ larger than that of product 1.The differences between the two products formed in process (4) suggest that two different mechanisms are involved in their production. This is supported by the observa- tion that Na(5S) seems to react with NO, to produce product 2 but not 1, which can be seen from the Na(4D) and Na(5S) product angular distributions in fig. 6. What is most noticeable is the large difference in the reaction cross-sections. Since the Na(4D) and Na(5S) states are very close in energy (98.8 vus. 94.9 kcal mol- ’ ) and both put the reactions well above the reaction endothermicity and threshold barrier of 20 kcal mol-I, this difference in reactivity could be the result of different reaction pathways. This point is furthered by the different peak positions and widths in the two angular distributions.J. M.Mestagh et al. 155 Sholeen and Herm have studied the reaction Li(2S) + NO, which is similar to process (4) for ground-state Na.18 The reaction products they found share many of the charac- teristics of product 2. Their experiments showed the formation of LiO(X ’II) and LiO( A &), which were forward-scattered and exhibited a large, broadly distributed recoil energy (the peak energy was 45% of the reaction exothermicity or 3.8 kcal mol-I). They saw no low-translational-energy, backward-scattered product, although they did not try the experiment using electronically excited Li. Sholeen and Herm proposed a Li+...NO,(’B, or ‘Al) intermediate to explain their data. The Li+. - .NO;(’B,) ion-pair seems more plausible.It does not provide as deep a well in the reaction potential surface as the Li+. - .NO,( ‘ A , ) intermediate. Also, this choice could explain the large product recoil energies observed; NO;(3B1) results from the transfer of the Li electron to an antibonding orbital (26,7r) on NO2. Sholeen and Herm, however, did not choose one pathway over the other.I8 The mechanism proposed in both the Li + NO, and Na + O2 experiments is reaction through an excited ion-pair intermediate. Since one of the Na+ NO2 products is similar to that of the Na+02 and the other to that of the LiO, it seems possible to propose that process (4) proceeds through the following intermediates: Na+. - .NO,( ‘ A , ) (7) Na+...NO,(’B,) (8) Nat...NOy(lBl). (9) Based on the comparison of the two products found to those found by others, the Na+ NO, reaction might proceed via intermediate (9) to give product 1 and through intermediate (8) to form product 2.Since NO2 is a bent molecule, a near-collinear collision between the Na and the 0-N bond in the NO, would result in a forward-scattered product with a fairly large impact parameter. This approach geometry, then, is consistent with the observations for product 2 formation. Problems, however, arise when one attempts to determine the Na(4D)-N02 collision geometry necessary to form product 1. A small impact parameter, which was what was seen experimentally, would position the Na away from the oxygen with which it is supposed to be colliding. Also, one would expect forward scattering for this type of collision.Physically, how is the NaO of product 1 formed? One answer which must be considered is that NaO is not being formed. As mentioned in the experiment, Na’, not NaO’, is detected. The possibility thus exists that Rydberg sodium atoms are being produced. If this were true, the similarity between product 1 and the Na + 0, product would cast some doubt on the Na + O2 mechanistic assignment. Another explanation is that NaO, + N is forming. The Na(4D) + NO2 --* NaO, + N reaction channel is energetically possible (AH is ca. -10 kcal mol-I). If the Na approached the O2 along the C2” symmetry axis, one could expect the product to be backward-scattered as is observed. Experiments carried out thus far have not given a definitive answer to the question of what this backward-scattered product is.Future experiments are being planned to resolve this. Other Interesting Features The experiments reported here have been extended in order to study how rotating the laser polarization, and thus the excited Na(P or D ) orbital, affects the Na reactivity. Owing to space limitations, only a few of these results are summarized below [see ref. (7) and (9) for more details]. The Na(4D)+HCl reaction is enhanced when the Na(4D) orbital is aligned along the relative velocity vector. This behaviour is expected if the reaction proceeds through a long-range electron-transfer mechanism in collinear Na. - 421- H geometry. For the156 Reactive Scattering of Excited Alkali-metal Atoms Na(4D) + 0, system, the favourable alignment for the reaction changes with the scatter- ing angle, and corresponds to the Na(4D) orbital being perpendicular to the molecular axis Na-0-0.Finally, the polarization effects encountered in the Na(4D) + NOz reaction suggest that product 2 but not product 1 formation is affected by the Na(4D) orbital alignment. This reinforces the conclusion reached in the previous section that products 1 and 2 are formed by different reaction mechanisms. Conclusion Representative families of excited alkali-metal atom reactions have been studied using a crossed-beam apparatus, and the reactivity of various excited states of Na has been investigated. For those systems in which reactions are also known for the ground-state alkali metal and involve an early electron-transfer step (e.g. reactions of Na with C12), no large differences are observed in the reactivity as Na is excited.This can be understood in terms of two competing effects; the lower ionization potential of excited Na increases the Na reactivity, which is compensated by a lower coupling between the covalent and ion-pair non-adiabatic curves of the Na-halogen system. The dynamics of the reaction remain unaffected by the Na electronic excitation, since the reaction behaviour is entirely determined by the shape of the LUMO of the molecular reactant. Similar observations, which are not reported here, have been made for reaction of excited Na atoms with organic halide molecule^.^ More interesting are the reactions of Na atoms with hydrogen halides. For ground- state Na and Na(3P), the reaction does not proceed via early electron transfer, and the NaCl-H repulsion dominates the product scattering. The excitation of Na to the 5s and 4 0 levels changes the reaction mechanism. The reaction then proceeds via an early dissociative electron attachment of the Na valence electron on HCl, and the HCI- repulsion dominates the NaCl product scattering.The reactions of excited Na with oxygenated compounds are of particular interest. The reaction with O2 was found to be state-specific [only Na(4D) reacts with 02], and seems to involve electron transfer to O2 so as to form the excited state A *nu of O,, i.e. the state responsible for dissociative electronic attachment on 02. The reaction with NO2 is even more complex, since Na( 4 0 ) leads to the formation of NaO by two different pathways.It must be mentioned, however, that the identification of NaO as product in these reactions has yet to be confirmed. This work was supported by the Director, Office of Energy Research, Office of Basic Energy Sciences, Chemical Sciences Division of the U.S. Department of Energy under contract no. DE-AC03-76SF00098. B.A.B. acknowledges a National Science Foundation Graduate Fellowship, J.M.M. thanks the Centre National de la Recherche Scientifique (France) for financial support. Some of the lasers used were on loan from the San Francisco Laser Center supported by the NSF under grant no. CHE79-16250 awarded to the University of California at Berkeley in collaboration with Stanford University. References 1 R. J. Buss, P. Casavecchia, T. Hirooka, S. J. Siebener and Y. T. Lee, Chem. Phys. Lett., 1981,82, 386. 2 R. R. Herm, in Alkali Halide Vapors, ed. P. Davidovits and D. L. McFadden (Academic Press, New York, 1979), p. 189. 3 M. W. Geis, H. Dispert, T. L. Budzynski and P. R. Brooks, in State to State Chemistry, ed. P. R. Brooks and E. F. Hayes, ACS Symp. Ser. No 56 (American Chemical Society, Washington, DC, 1977), p. 103. 4 P. S. Weiss, J. M. Mestdagh, H. Schmidt, M. F. Vernon, M. H. Covinsky, B. A. Balko and Y. T. Lee, in Recent Advances in Molecular Reaction Dynamics, ed. R. Vetter and J. Vigue (1985). 5 M. F. Vernon, H. Schmidt, P. S. Weiss, M. H. Covinsky and Y. T. Lee, J. Chem. Phys., 1986 84, 5580.J. M. Mestagh et al. 157 6 H. Schmidt, P. S. Weiss, J. M. Mestdagh, M. H. Covinsky and Y. T. Lee, Chem. Phys. Lett., 1985, 118, 7 P. S. Weiss, J. M. Mestdagh, M. H. Covinsky, B. A. Balko and Y. T. Lee, to be published. 8 G. Rahmat, F. Spiegelmann, J. Verges and R. Vetter, Chem. Phys. Lett., 1987, 135, 459. 9 P. S. Weiss, Ph.D. Thesis (Lawrence Berkeley Laboratory, University of California, 1986). 539. 10 Y. T. Lee in Atomic and Molecular Beam Methods, ed. G. Scoles and U. Buck, Oxford University Press, New York, 1986. 11 E. M. Goldfield, E. A. Gislason and N. H. Sabelli, J. Chem. Phys., 1985,82, 3179; E. M. Goldfield, A. H. Kosmas, and E. A. Gislason, J. Chem. Phys., 1985, 82, 3191. 12 B. A. Blackwell, J. C. Polanyi and J. J. Sloan, Chem. Phys., 1978, 30, 299. 13 J. N. Bardsley and J. M. Wadehra, J. Chem. Phys., 1983, 78, 7227. 14 P. J. Kuntz, M. H. Mok and J. C. Polanyi, 1969, J. Chem. Phys., 1969, 50, 4623. 15 M. H. Alexander, J. Chem. Phys., 1978, 69, 3502. 16 J. M. Mestagh, D. Paillard and J. Berlande, J. Chem. Phys., submitted for publication. 17 D. S. Belic and R. I. Hall, J. Phys. B, 1981, 14, 365. 18 C. M. Sholeen and R. R. Herm, J. Chem. Phys., 1976, 64, 5261. Received 19th May, 1987

ISSN:0301-7249    

DOI:10.1039/DC9878400145

出版商:RSC    

年代:1987

数据来源: RSC