Dissociative adsorption of methane on Pt(111) induced by hyperthermal collisions

摘要

Faraday Discuss., 1993,96, 325-336 Dissociative Adsorption of Methane on Pt(ll1) induced by Hyperthermal Collisions Darren J. Oakes, Martin R. S. McCoustra and Michael A. Chesters School of Chemical Sciences, University of East Anglia, Norwich, UK NR4 7TJ Measurements have been made of the dissociative adsorption of methane on a Pt(ll1) surface at surface temperatures of 150 and 550 K using hyper- thermal beam techniques. Sticking coefficient measurements made at 550 K are in close agreement with existing literature data. Measurements made at 150 K extend our understanding of the dissociative sticking of methane into a new, low-temperature regime, Reflection-absorption infrared (RAIR) spec-tral measurements made at 150 K definitively identify the primary product of the dissociative adsorption as the methyl (CH,) moiety.The thermal evol- ution of this moiety has subsequently been followed by RAIRS and temperature-programmed desorption (TPD). The observation of the C, eth-ylidyne moiety at temperatures above 300 K provides clear evidence for the formation of C-C bonds during the thermal chemistry of the methyl moiety. We have studied the impact-induced dissociation of methane on Pt( 111) for two reasons, first to contribute to the understanding of the mechanism of C-H bond activation when alkanes interact with metal surfaces and secondly to follow the surface chemistry of the resulting adsorbed hydrocarbon species. There have been several studies of methane dissociation on metals involving purely thermal a~tivationl-~ and mixtures of hyperthermal and thermal activation.8-16 Early measurements of adsorption by thermal activation were interpreted in terms of the reac- tion proceeding across a potential-energy surface and involving a classical transition state.Experiments using hyperthermal beams and high surface temperatures showed that the sticking coeficient varied exponentially with the normal component of the inci- dent kinetic energy and also found large kinetic isotope effects. These results were inter- preted in terms of tunnelling of the dissociating H atom through an activation barrier. The tunnelling model has been modified to account for a substrate-temperature depen- dence of the sticking probability which becomes particularly apparent at incident ener- gies well below the barrier and at low substrate temperature^.'.^^ The experiments reported here extend measurements of sticking probability of methane on Pt(111) to low surface temperature (1 50 K) where the effect of surface vibra- tional activation would become small.We have been able to measure sticking probabil- ities at this low temperature by using a combination of Auger electron spectroscopy, to detect surface carbon concentrations, and thermal desorption spectroscopy, which mea- sures associative desorption of methane. The low-surface-temperature products of impact-induced dissociation of methane are generally assumed to be adsorbed methyl and adsorbed hydrogen. This has been con- firmed only in the case of methane dissociation on the Ni( 11 1) surface, by detection of adsorbed methyl by electron energy loss spectroscopy In this work, we detect the surface hydrocarbon species resulting from methane dissociation using RAIRS.The infrared spectrum of methyl adsorbed on Pt(ll1) is already known from 325 Dissociative Adsorption induced by Hyperthermal Collisions experiments using dissociation of methyl halides1 8-22 and direct adsorption of methyl radicals23 and the rich surface chemistry resulting from thermal activation of adsorbed methyl on Pt( 111) is already under investigation. We will show that the surface chemistry of adsorbed methyl generated by impact-induced dissociation is consistent with that observed for adsorbed methyl generated by other routes.Experimental All experiments were conducted in a stainless-steel ultra-high vacuum (UHV) chamber pumped by a liquid nitrogen trapped 9 in? oil diffusion pump to a base pressure of better than 1 x 10-l' mbar.? Details of the UHV system and its attendant hyper- thermal molecular beam system will be reported el~ewhere.~~.~~ Briefly, however, the chamber consists of two operating levels. The upper level is equipped with a hemispheri- cal electron energy analyser (VG CLAM 100) and electron gun for Auger electron spec- troscopy (AES) and an argon-ion sputter gun for sample cleaning. The lower experimental level, in addition to the infrared optics and hyperthermal molecular beam source, is equipped with rear view LEED optics and quadrupole mass spectrometers (VG Micromass Q7 and Micromass PC300D) for beam velocity and flux measurements and TPD measurements.A schematic representation of this level is given in Fig. 1. The hyperthermal molecular beam source consists of an alumina tube of 6 mm outside diameter and 4 mm inside diameter, laser drilled to provide a pinhole of ca. 40 pm diameter in the sealed end. Nozzle temperatures of between 300 and 1200 K are achieved by resistive heating of a length of tungsten wire wrapped around the alumina tube and coated with a temperature-resistant ceramic adhesive. The nozzle temperature is monitored by a chromel-alumel thermocouple contacted to this assembly by ceramic adhesive. The free-jet expansion from this source is skimmed with a 500 pm nickel skimmer and the resulting beam directed through two stages of differential pumping into the UHV chamber via a 4 mm collimating orifice positioned ca.100 mm from the crystal. With a typical gas composition of 0.5 to 1% of methane (BOC Ltd) in research grade helium (BOC Ltd) at source stagnation pressures of between 1000 and 1800 mbar, molecular beam U" FTlR -1 7)QMS / Y\spectrometer IR detector Fig. 1 Schematic diagram showing the optical arrangement on the molecular-beam level of our UHV chamber. The AES system, not shown, is at a higher level. t 1 in = 2.54 x lop2m. D. J. Oakes, M. R. S. McCoustra and M. A. Chesters time-of-flight measurements indicate substantial acceleration of the methane and result- ant gaussian-like velocity distributions characterised by velocity spreads (Av/u) of the order of 10 to 15%. This design of nozzle has allowed measurements to be made at methane kinetic energies of up to 100 kJ mol-' with this molecular beam system.However, at the higher energies, considerable quantities of methane pyrolysis products are observed in the beam. The measurements reported here have been made at kinetic energies of up to and including 85 kJ mol-'. At this and lower energies no evidence for methane pyrolysis products was found. RAIR spectra were recorded using a Mattson Sirius 100 FTIR spectrometer. The infrared beam from the interferometer was focused through a KBr window onto the platinum crystal at an incident angle of 84 & 6" to the surface normal, passed through a polariser after reflection, and collected by either a mercury cadmium telluride (MCT) or an indium antimonide (InSb) detector.The former enables spectra to be recorded in the region 4000-700 cm-',while the latter operates only at shorter wavelengths (5000-2000 cm -') but with a substantially higher sensitivity. Spectra presented here were obtained as a ratio of 1024 sample scans to 1024 background scans (clean sample) at a resolution of 4 cm-'. Each single-beam spectrum represents a total measurement time of 18 min. The Pt( 11 1) single crystal was mounted on a rotatable three-axis manipulator and could be resistively heated to above 1200 K and cooled to around 90 K. Surface tem- perature was measured by a chromel-alumel thermocouple spot-welded to the side of the crystal.Cleaning was afforded by cycles of argon-ion bombardment followed by annealing to 1000 K. Surface cleanness was monitored by AES and the RAIR spectrum of adsorbed carbon monoxide which is extremely sensitive to small levels of contami- nation. Auger spectroscopy was also used to measure quantities of carbon on the surface resulting from methane dissociation. The Auger signal was calibrated by measuring the spectrum of one monolayer of carbon monoxide at 100 K after 5 Lt exposure. The area under the carbon N(E) Auger peak for this adsorbed layer was taken to correspond to 0.55 monolayers of carbon.26 Auger spectra were recorded using an exciting electron- beam current of 0.1 pA into a spot diameter of ca.2 mm giving an electron flux of ca. 0.03 A mP2 which during the 400 s taken to record Auger spectra produces an electron beam dose of 12 C mP2. Comparison with electron-beam effects on hydrocarbon layers on Pt( 11 1) suggests that disruption of the adsorbed layer should be negligible.27 Thermal desorption spectra of methane were recorded using a quadrupole mass spectrometer equipped with an enclosed cross-beam source positioned on a rotary table within the UHV chamber. A line-of-sight tube affixed to the source inlet was used to enhance the selectivity of the system to material desorbing from the crystal. Methane 'TPD signals were measured by recording the intensities of both the m/z = 15 (CH,f) and m/z = 16 (CH:) peaks.Sample dosing was achieved by exposing the crystal to the hyperthermal beam for a set time. To initiate exposure, the tantalum beam flag was removed from the beam path. On completion of exposure, this flag was then returned to its beam blocking position. Exposure times, where mentioned, are accurate to better than 1 s. Beam-flux measure- ments were performed using a modified mass spectrometer employed as stagnation gauge and from a knowledge of the pumping speed and volume of the UHV system. Results and Discussion Ilynamical Measurements In this section, we report measurements of initial dissociative sticking coefficients, Sdgiss, of methane on Pt(ll1) at both 550 and 150 K over a range of incident kinetic energies t 1 L (Langmuir) = 1 x Torrs.Dissociative Adsorption induced by Hyperthermal Collisions normal to the surface (EN).Prior to our hyperthermal beam studies, it was confirmed that no AES or TPD evidence for dissociative sticking of methane on Pt( 111) at either of the two surface temperatures in our investigation could be found with an effusive, thermal (298 K) source of methane. At 550 K, AES measurements of carbon build-up on the substrate have been made as a function of beam-exposure time. A small fraction of these data, corresponding to a methane kinetic energy of 85 kJ mol-', is shown in Fig. 2. The spectra in derivative mode show the carbon peak growing above the two platinum peaks. The spectra were integrated twice to obtain the total peak area, carbon plus platinum, in the 210-310 eV , 1 1 ~ - 1 1 ~ ' 1 210 230 250 270 290 310 energyleV Fig.2 Series of differential AES spectra recorded following exposure of the Pt(ll1) surface at 550 K to a methane molecular beam of 85 kJ mol-' kinetic energy. The data show the growth of the carbon signal and attenuation of the platinum signal with increasing exposure. Exposure: (a)0, (b) 8, (c) 23, (d) 39 and (e) 190 L. Carbon coverage: (a) 0, (h) 0.13, (c) 0.25, (d) 0.38 and (e) 0.98 monolayers. D. J. Oakes, M. R. S. McCoustra and M. A. Chesters 329 range. There is significant attenuation of the platinum Auger peak intensity accompany- ing the growth in carbon signal and we have assumed that both of these effects vary linearly with carbon coverage in the submonolayer range.Carbon build-up curves for different values of EN are shown in Fig. 3. The initial rate of carbon build-up is very sensitive to ENas expected. The initial dissociative sticking coefficients, Sdgiss, of methane have been estimated from the initial slopes of the data presented in Fig. 3. These data are shown in Fig. 4. It is clear from the increase of over two orders of magnitude over this limited range of kinetic energies, that there is a substantial increase in the activation of methane on Pt( 11 1) as a function of kinetic energy normal to the surface. This dramatic translation activation is similar to that reported previously by Luntz and co-worker~~~~ and Schoofs et a1.l' for methane on Pt(ll1) and by Rettner et ul.," Hamza et al." and Ceyer and co-~orkers~~-'~for methane dissociation on W( 1 lo), Ir( 110) and Ni(l1 l), respectively.In agreement with the work of Luntz and co-w~rkers,~'~ our data show a curvature which is inconsistent with a pure exponential activation. In all, the agreement of our data with those of Luntz and Bethune is excellent as shown in Fig. 4. It should be noted, however, that the data of Luntz and Bethune reported in that figure refer to a somewhat higher surface temperature (ca.800 K) and, in part, the observed differences between our data and those of Luntz and Bethune can be attributed to this difference in substrate surface temperature. The surface-temperature dependence of Stis' has been reported to be approximately Arrhenius with an activation barrier of 16.7 kJ mol-' at E, = 40.4 kJmol-1.8.9 At this incident energy we can therefore estimate the ratio of Stss(K = 550 K)/S$"(K = 800 K) to be of the order of 0.32.At higher incident energies, the depen- dence of Spsis reported to be weakly linear. From data presented in ref. 8, we can estimate that the above ratio is increased to 0.88 for a kinetic energy of 122.5 kJ mo1-'. In our experiments, the nozzle temperature was varied from approximately 500 to 1000 .-If9 o.6/ rn 0 0 A $ 0.4 X:I . A AXo.2i0.0 oooo 0 25 50 75 100 125 150 175 exposure/ML Fig. 3 Carbon uptake curves as a function of methane kinetic energy as derived from data such as those in Fig.2. EN= I,85; U, 75;A,65; x, 55 and 0,45kJ mol-'. Dissociative Adsorption induced by Hyperthermal Collisions A 0 A 0 0 0 10" i'i'i'l' 40 50 60 70 a0 90 E,/kJ mot-' Fig. 4 Comparison of the sticking coefficients determined in this work, 0,T,= 550 K and those of Luntz and co-workers, 0,T,= 800 K8,' and Schoofs et UI.,'~A, T, = 500 K K in order to achieve the range of EN values measured. This will produce an enhance-ment in our sticking probability values at high EN,through the presence of vibrationally excited CH, in the beam,'-'' by about a factor of 2 over that which would be seen if the nozzle temperature had been kept constant at 500 K. As an alternative approach, hydro-gen seeding into the source gas could have been employed to provide the required range of energies at a constant nozzle temperature.This approach would, however, have com-plicated measurements made at low surface temperatures. Given these considerations, we feel confident that the small difference between our data obtained at a substrate-surface temperature of 550 K and those of Luntz and co-workers reported for 800 K in Fig. 4 can be explained simply on the basis of both the small surface-temperature depen-dence of the methane activation and the effect of varying nozzle temperature. The nature of the primary dissociation products and their subsequent reaction paths have not been established at this high surface temperature but it is reasonable to assume that, at low coverage, impact-induced dissociation is the rate-determining step in carbon build-up.At lower surface temperatures the primary dissociation products may be trapped as demonstrated by Ceyer and co-workers.' 3-16 We have used RAIRS to iden-tify surface species and a detailed account of the IR results is presented in the following section. Here we simply report that the only hydrocarbon fragment detected on the surface following impact-induced dissociation at a surface temperature of 150 K was the methyl moiety. At low surface coverages of methyl, thermal desorption results in the evolution of methane at ca. 270 K. This associative desorption would involve the chemi-sorbed hydrogen atoms originally dissociated from the methane molecule. This result is consistent with thermal desorption of surface methyl species generated on Pt(ll1) by other route^."-^^ AES measurements indicate that at low initial coverages of methyl, thermal desorption of methane leaves a clean platinum surface while at high methyl coverages significant amounts of carbon remain after heating to 300 K.The nature of the irreversibly adsorbed species will be discussed in the next section. Since dissociative D. J. Oakes, M. R. S. McCoustra and M. A. Chesters 33 1 adsorption of methane is reversible in the coverage range relevant to measurements of the initial sticking probability, we have used TDS to investigate the variation of with ENat a substrate temperature of 150 K. The data shown in Fig. 5 are typical of those obtained in these measurements and were obtained at EN= 85 kJ mol-’.The absolute coverage in these measurements was estimated by comparison of the uptake curve obtained by TPD with that obtained by AES as described above. The uptake curves at lower kinetic energies were determined only by TPD and the absolute coverage evaluated by comparison with the data from the 85 kJ mol-’ experiment. The S:iss were then estimated from the initial slopes of these absolute uptake curves. These data are summarised in Fig. 6. It is clear from the comparison in Fig. 6 that variation of Stis’ with EN at 150 K is strikingly different from that over the same energy range at 550 K. The data for 150 K show considerably more curvature than the higher temperature data.Moreover, it is clear from this comparison that Sdgiss,while weakly dependent on surface temperature at high EN, shows a considerably stronger dependence on surface temperature at lower kinetic energies. Are these new results consistent with the current framework of theoretical models for the dissociation of methane on transition metals? A number of models have been pro- posed based on the wide range of dynamical and kinetic measurements that have been carried out on this system. From thermal activation studies, various precursor models have been proposed to account for both the surface-temperature dependence and the gas-temperature dependence of Sdgiss. Typical of these models is that proposed by Winters’*2 from measurements of Stss for CH, on tungsten filaments over the surface- temperature range 600-2500 K. The observation of a dramatic increase in s$’’ with surface temperature and a large kinetic isotope effect, which was observed to decrease I--..-II.-II/ 150 200 250 300 350 JIK Fig.5 Series of methane TPD measurements made following exposure of the Pt(ll1) surface at 150 K to a methane molecular beam of 85 kJ mol-’ kinetic energy. The data show an increase in methane associative desorption with increasing exposure. Exposure: m, 1; 0,3; a,10; A, 25 and v 100 L. Dissociative Adsorption induced by Hyperthermal Collisions E,/kJ mol-10-2; 00 (I, lo4 7 10" I( 20 30 40 50 60 70 E,/kJ mol-' Fig. 6 (a) Summary of the sticking coefficient measurements in this work.m, Experimental data for T, = 550 K; e,for T, = 150 K. (b)The results of model calculations performed by Luntz and co-workers9.' for dissociation of methane on Pt(111) at the indicated surface temperatures using = .,the thermally assisted tunnelling model. T, 800 and .,0 K. D. J. Oakes, M. R. S. McCoustra and M. A. Chesters with increasing surface temperature, was taken as evidence for a mechanism based on tunnelling of H atoms through a barrier along the C-H stretching vibrational coordi- nate from vibrationally excited species in thermal equilibrium with the surface. At the lower surface temperatures, equilibration of the methane with the surface was assumed to occur through a precursor state. In contrast, at high surface temperatures where precursor concentrations would be too small to account for the measured sticking, equilibration was considered to occur through direct collision with the surface.Given recent measurements of the efficacy of surface-molecule vibrational energy transfer,28 this would seem highly unlikely. Such complex mixed mechanisms have also been invoked by a number of other workers to explain thermal activation under equilibrium conditions either by increasing the surface temperature or by increasing the gas tem- perat~re.~-~common factor in these studies is the observation of a large kinetic A isotope effect and the invoking of a tunnelling mechanism. Precursor-mediated disso-ciation is also invoked to explain the observed surface temperature dependence of Sdgiss.Molecular-beam studies of the dissociative adsorption of CH, on transition metals have offered greater control over experimental parameters such as methane kinetic energy and surface temperature. In the majority of these studies, SFsshas been found to exhibit a large, often apparently exponential, dependence on kinetic energy normal to the substrate surface. In addition, strong kinetic isotope effects have been observed. Some of these studies, invoke tunnelling in a direct dissociation mechanism to account for these observations.' While others' 3-'6 again describe an essentially direct disso- ciation mechanism, but from a molecule distorted by collision with the surface, a molec- ular 'splat '. However, Ceyer and co-workers concede that a tunnelling mechanism cannot be fully discounted by their measurements. A precursor model cannot, however, explain the observed dependence of Sdgisson EN.The evidence for surface-temperature dependence of Sfss in molecular-beam studies is, however, somewhat controversial. For example, in the case of CH, dissociation on Pt(11 l), Luntz and co-~orkers*~~ report clear evidence for a surface temperature depen- dence of Sdgisswhile Schoofs et al." do not. Our own measurements clearly indicate that such a temperature dependence does exist. It is not possible to account for such an observation within the simple framework of tunnelling through a rigid barrier in a direct dissociation as described above. Luntz and Harris have, however, proposed a more sophisticated mechanism which they describe as thermally assisted t~nnelling.~.'~ In this model, direct dissociation occurs on impact with the surface via a tunnelling mechanism.Tunnelling occurs through a barrier the height of which is effectively modulated with respect to the incoming methane molecules by the lattice vibrations of the substrate. The key factor in determining SFSsis therefore the phase of this oscillation with respect to the incoming methane molecule. Luntz and co-w~rkers~~'~ have reported wavepacket simu- lations within this model based on a simple reduced dimensionality potential-energy surface. The results of these calculations are consistent with a wide range of experimen- tal data, obtained at high surface temperatures.In addition, their model predicts that at high EN the surface-temperature dependence will be weak. In contrast, at low EN,a significant dependence on surface temperature will be observed. The trend predicted by the calculations compares favourably with the trend exhibited by our experimental results, as shown in Fig. 6. Identification of Adsorbed Hydrocarbon Species Impact-induced dissociation of methane on Pt(ll1) at 150 K produces an adsorbed species with the IR spectrum shown in Fig. 7(a) and (b). There is a single detectable absorption band in the C-H stretching region at 2885 cm-' and no bands detectable in the low-wavenumber region. The single C-H stretching band may be assigned the symmetric stretch of adsorbed methyl.The absence of a band near 2960 cm-' which Dissociative Adsorption induced by Hyperthermal Collisions II*''''lL I'11Illtl 3100 3000 2900 2800 2700 1400 1300 1200 1100 1000 wavenumber/cm-' wavenumber/cm-' Fig. 7 (a) RAIR spectrum, recorded using an InSb detector, following exposure of the Pt(ll1) surface at 150 K to a methane molecular beam of EN= 85 kJ mol-' for 40 min. (b)As (a) but using an MCT detector. (c) RAIR spectrum obtained by heating the adlayer in (b)to 300 K. could be assigned to the antisymmetric methyl stretch is evidence that the methyl group is oriented with its three-fold axis perpendicular to the metal surface. This spectrum has been reported for methyl on Pt(ll1) generated by dissociation of methyl and by adsorption of methyl radicals.23 The intensity of the C-H stretching band of the saturated layer of adsorbed methyl in Fig.7(a)and (b)is also similar to that produced by adsorption of methyl radicals.23 In our experiment, saturation was achieved by translating the crystal in stages so as to achieve exposure of the whole crystal surface to an 85 kJ mol-' methane beam. Experi- ments in which the crystal was positioned so that only its central area was exposed to the beam (ca.4 mm diameter) produced a spectrum five to six times less intense, which is consistent with the ratio of beam area to crystal surface area (10 mm diameter crystal). I L 11 1 I 8 1 ' 1 I 31'00 3000 2900 2800 2700 1400 1300 1200 1100 1000 waven u mber/cm -' wavenumber/cm-' Fig.8 (a) RAIR spectrum, recorded using an MCT detector, following exposure of the Pt(l11) surface at 400 K to a methane molecular beam EN= 85 kJ mol-' for 40 min. (b)RAIR spectrum of a saturated layer of ethylidyne produced by exposing the Pt(ll1) surface at 360 K to 6 L of ethene. D. J. Oakes, M. R. S. McCoustra and M. A. Chesters Heating the saturated methyl layer to 300 K produced the spectrum shown in Fig. 7(c) which is characteristic of the ethylidyne species. This result of C-C coupling has been reported previously resulting from decomposition of dimethylzinc over Pt( 1 1 1)29 or from reaction of surface methyl species generated by adsorption of methyl radicals.23 Impact-induced adsorption at a substrate temperature of 400 K also produces the ethyl- idyne species [Fig.8(a)], although presumably through the same intermediates as for adsorption at 150 K and subsequent heating. Comparison with the spectrum of a surface saturated with ethylidyne generated by ethene adsorption at 360 K, Fig. 8(b), shows that the C-C coupling route does not lead to a saturated surface. This is not surprising for the case of absorption of surface methyl at 150 K followed by heating, since desorption of methane has been shown to be a competing reaction. However, adsorption of methyl at 400 K might be expected to produce eventually a saturated ethylidyne layer unless there are other competing processes such as decomposition of surface methyl via CH, species to surface carbon. The detailed mechanism of decompo-sition of adsorbed methyl is still under investigation. Conclusions The impact-induced dissociation of methane on Pt( 11 1) at substrate temperatures of 550 and 150 K has been shown to follow a dependence on normal impact energy and sub- strate temperature which is in accord with the model of Luntz and co-w~rkers.~.~~ At 150 K, the model surface oscillator employed by Luntz and Harris (6= 160 cm-') would still be significantly thermally activated and so contribute to the sticking coeffi- cient of methane.At still lower temperatures, the effect of the surface oscillator will be further reduced and the sticking probability in the lower impact-energy range should continue to fall. A further important test of the model will be to extend sticking prob- ability measurements of CD, to low temperatures. Such measurements are underway in our laboratory.The surface products of methane dissociation on Pt( 1 1 1) have been identified directly using IR spectroscopy. At 150 K the expected surface methyl species was found to be the only hydrocarbon product of dissociation. Its IR spectrum is identical to that produced by adsorption of methyl radicals or dissociation of methyl halides. Subsequent heating of the adsorbed methyl leads, at low coverage, predominantly to methane desorption and, at high coverage, to C-C coupling and formation of ethylidyne. Again this behav- iour is qualitatively similar to that of methyl on platinum generated through other routes but with differences of detail which will be discussed in a later publication.We would like to thank the UK SERC for the award of an equipment grant and a studentship (to D.J.O.). References 1 H. F. Winters, J. Chem. Phys., 1975,62, 2454. 2 H. F. Winters, J. Chem. Phys., 1976,64,3495. 3 C. N. Stewart and G. Ehrlich, J. Chem. Phys., 1975,62,4672. 4 S. G. Brass and G. Ehrlich, Surf. Sci., 1987, 187, 21. 5 S. G. Brass and G. Ehrlich, J. Chem. Phys., 1987,87,4285. 6 S. G. Brass and G. Ehrlich, Phys. Rev. Lett., 1986,57, 2532. 7 W. H. Weinberg, J. Vac. Sci. Technof.A, 1992, 10, 2271. 8 A. C. Luntz and D. S. Bethune, J. Chem. Phys., 1989,90,1274. 9 J. Harris, J. Simon, A. C. Luntz, C. B. Mullins and C. T. Rettner, Phys. Rev. Lett., 1991, 67, 652. 10 G.R. Schoofs, C. R. Arumainayagam, M. C. McMaster and R. J. Madix, Surf Sci., 1989,215, I. 11 C. T. Rettner, H. E. Pfnur and D. J. Auerbach, Phys. Rev. Lett., 1985,54,2716. 12 A. V. Hamza, H-P. Steinruck and R. J. Madix, J. Chem. Phys., 1987,86,6506. 13 M. B. Lee, Q. Y. Yang, S. L. Tang and S. T. Ceyer, J. Chem. Phys., 1986,85, 1693. Dissociative Adsorption induced by Hyperthermal Collisions 14 S. T. Ceyer, J. D. Beckerle, M. B. Lee, S. L. Tang, Q. Y. Yang and M. A. Hines, J. Vac. Sci. Technol. A, 1987,5, 501. 15 M. B. Lee, Q. Y. Yang and S. T. Ceyer, J. Chem. Phys., 1987,87,2724. 16 S. T. Ceyer, Annu. Rev. Phys. Chem., 1988,39,479. 17 A. C. Luntz and J. Harris, Surf: Sci., 1991,258, 397. 18 M. A. Henderson, G. E. Mitchell and J. M.White, Surf. Sci., 1987, 184, L325. 19 Z.-M. Liu, S. Akhter, B. Roop and J. M. White, J. Am. Chem. Soc., 1988,110,8708. 20 G. Radhaknshnan, W. Stenzel, R. Hemmen, H. Conrad and A. M. Bradshaw, J. Chem. Phys., 1991,95, 3930. 21 F. Zaera, Langmuir, 1991,7, 1998. 22 F. Zaera, H. Hoffmann, J. Phys. Chem., 1991,95,6297. 23 D. H. Fairbrother, X. D. Peng, R. Viswanathan, P. C. Stair, M. Trenary and J. Fan, Surf. Sci., 1993,285, L455. 24 D. J. Oakes, PhD Thesis, University of East Anglia, in preparation. 25 D. J. Oakes, M. A. Chesters and M. R. S. McCoustra, Surf Sci., in preparation. 26 M. Tiishaus, E. Schweizer, P. Hollins and A. M. Bradshaw, J. Electron Spectrosc. Relat. Phenom., 1987, 44,305. 27 N. D. S. Canning, M. D. Baker and M. A. Chesters, SurF Sci., 1981, 111, 441. 28 P. L. Houston and R. P. Merrill, Chem. Rev., 1988,88,657. 29 D. J. Oakes, J. C. Wenger, M. A. Chesters and M. R. S. McCoustra, J. Electron Spectrosc. Relat. Phenom., in the press. Paper 3/03936F; Received 7th July, 1993