J. Phys. Chem. 1985,89, 5381-5386 supported this hypothesis in light of his results on the thermolysis of the 9-decalyl tert-butoxy peresters. However, Bartlett's results also lend some credence to distinguishable cis and trans radicals, in that he reported different ratios of cis and trans hydroperoxides from high-pressure oxygen trapping of the radicals generated from the isomeric peresters! It is apparent from both ESR and product analysis that at least at low temperatures distinguishable radicals are obtained, negating a common planar radical intermediate. The small degree of cis to trans isomerization is probably best rationalized in terms of the reaction coordinate (Figure 2) which assumes retention by the radicals of the well-known ground-state energy ordering of the parent hydrocarbons.'0
In conclusion, we have shown that the bridgehead cis- and tranr-9-decalyl radicals are distorted from the normal tetrahedral geometry of the parent hydrocarbons. Although both radical sites are flattened they are not planar and therefore distinct cis and trans radicals result. This conclusion, which is in contrast to previous postulate^,^,^ suggests to us that the energy barrier to inversion is relatively low and is accessible at higher temperatures. Registry No. BOOB, 110-05-4; 9-decalyl radical, 52692-69-0; cisdecalin, 493-01-6; trans-decalin, 493-02-7. (10) E M , E. L. "Stereochemistry of Carbon Compounds"; McGraw-Hill: New York, 1962.
Dynamics of the Dissociative Adsorptlon of Hydrogen on Ni( 100) A. V. Hamza and R. J. Madix* Department of Chemical Engineering, Stanford University, Stanford, California 94305 (Received: June 10, 1985)
Nearly monoenergetic beams of hydrogen and deuterium were used to determine dissociative sticking probabilities for H2 and D2 on Ni( 100) at various energies. Variation of the surface temperature between 90 and 300 K had no effect on the dissociative sticking probability of H2at 3.6 and 5.8 Id/mol incident beam energy, indicating a direct mechanism of dissociation. A fourfold increase in the initial dissociative sticking Probability for H,from 0.2 to 0.8 was observed by increasing the translational kinetic energy from 0.7 to 7.0 kJ/mol. The initial dissociative sticking probability for D2 was slightly lower, increasing from 0.15 to 0.75 with increasing translational kinetic energy from 1.3 to 10.5 kJ/mol. The form of the increase with kinetic energy was explained by tunnelling through a low activation barrier, accounting as well for the high dissociation probability at low kinetic energies. The dissociative sticking probability decreased with hydrogen or deuterium adatom coverage at all energies. The decline in sticking probability with hydrogen coverage was fit to a s(0) = so( 1 - a6)" functional form. From this relationshipit was deduced that hydrogen adatoms block only single sites and that four vacant sites are needed for dissociation. The dissociative sticking probability for H, declined precipitously from 0.77 to 0.16 with oxygen adatom coverage from 0 to 5% of a monolayer at a translational energy of 9.6 kJ/mol.
1.0. Introduction Dissociative chemisorption may be the ratelimiting step in some surface chemical reactions. The difficulty of the chemisorption step is often attributed to an activation barrier to adsorption, yet examination of the existence and nature of these barriers has only recently been initiated. The hydrogen-nickel adsorption system has been the subject of considerable attention as a model for dissociative adsorption on metals' because of its importance in hydrogenation reactions. In his pioneering work van Willigen, measured the angular distribution of Hzmolecules associatively desorbing from a polycrystalline nickel sample by allowing hydrogen to permeate through the nickel sample. The deviation of the observed angular distribution from the Knudsen cosine law led van Willigen to propose a model for activated adsorption based on detailed balancing. van Willigen's activated adsorption model predicted a barrier of 16.8 kJ/mol for H2 adsorption on polycrystalline nickel. The direct observation of an activation barrier to the adsorption of H2 on epitaxially grown Ni( 11 1) was made by Palmer et al.3 The apparent activation energy was estimated to be 4 kJ/mol based on the increase in the desorbing H D flux with increasing oven temperatures for effusive hydrogen beams, and angular distributions for desorption more peaked than cosine were also reported. Recently, Rendulic, Winkler, and co-workersM have examined the adsorption of hydrogen on the low index (1) Wedler, G. "Chemisorption: An Experimental Approach"; Butterworths: London, 1976. (2) van Willigen, W. Phys. Lett. A 1968, 2 8 4 80. (3) Palmer, R. L.; Smith, J. N., Saltsburg, H.; OKeefe, D. R. J . Chem.
Phys. 1970, 53,
1666.
planes of nickel as a function of gas and surface temperature. On the Ni( 111) surface4 the sticking coefficient of H2 increased with gas temperature in agreement with Palmer et al.'s3 work. Howthe hydrogen sticking coefficient was nearly ever, on Ni( independent of the gas temperature; on the N i ( l l 0 ) surfaceSthe hydrogen sticking coefficient decreased slightly with gas and surface temperature. In all of these experimentse6 the gas impinging on the nickel surfaces had a Maxwellian velocity distribution. Balooch et al.' used a molecular beam nozzle source to examine the adsorption of hydrogen on copper surfaces. A threshold to H2 adsorption was observed as the translational kinetic energy of the incident beam was increased. At low energy dissociative adsorption probabilities of 2 or 3% were measured; at higher translational kinetic energies values of 10-1 5% were measured. In accordance with van Willigen's model of activated adsorption the reaction probability scaled with the translational kinetic energy normal to the surface, implying an adsorption barrier perpendicular to the surface. Contrary to van Willigen's model, however, a step function in the adsorption probability with energy was not observed. Balooch et al.' have proposed a distribution of barrier heights on the surface to account for t h e S-shaped nature of the (4) Steinruck, H. P.; Rendulic, K. D.; Winkler, A. Surf. Sci., to be published. . . .. .. -. ( 5 ) Luger, M.; Winkler, A.; Rendulic, K. D.; Steinruck, H. P. Proc. Symp. Surf. Sci., Obertraun 1985, 107. ( 6 ) Steinruck, H. P.; Rendulic, K. D.; Winkler, A. Proc. Symp. . Surf. . Sci., Obertraun 1985, 101. (7) Balooch, M.; Cardillo, M. J.; Miller, D. R.; Stickney, R. E. Surf. Sci. 1974, 46, 358.
0022-3654/85/2089-5381$01.50/0 0 1985 American Chemical Society
5382 The Journal of Physical Chemistry, Vol. 89, No. 25, 1985 sticking probability vs. energy. Also at odds with van Willigen’s model is the significant reaction probability at energies below the threshold and reaction probabilities much less than one at energies above the threshold. Comsa and co-workerss-’O have energy-analyzed hydrogen desorbing from polycrystalline Ni and Ni( 11 1) by permeation. While peaked (cosd 8; 3 < d < 5) angular flux distributions were observed, the variation iii energy with angle from the normal to the surface was inconsistent with van Willigen’s model. Comsa and Davidg introduced Yactivated adsorption with holes” in order to account for substantially all the discrepancies between their observations and van Willigen’s model, providing a nonactivated pathway for adsorption (holes) in the model. ”Holes” would also explain the relatively high sticking probabilities observed below the energy threshold by Balooch et al.’ Comsa and co-workersss10 and Stickney and co-workers’1 ~ 1 2have shown that sulfur impurities on polycrystalline Ni surfaces lead to more sharply peaked angular flux distributions, suggesting that impurities may raise the activation barrier. Hydrogen adsorption probabilities and adsorption kinetics for ambient hydrogen gas have been investigated on the low index planes of n i ~ k e l . ’ ~ - ’Initial ~ dissociative sticking probabilities between 0.1 and 0.6 have been reported for Ni(100).’3v143’6A molecular precursor to dissociative adsorption has been proposed for the Ni( 110) surface;’* however, adsorption kinetics on Ni( 100) seemed to indicate the adsorption process is direct.13 KO and Madix19 and Johnson and MadixZ0have studied the effects of carbon and sulfur adlayers repectively on the adsorption of H z on Ni(100). Both carbon and sulfur adlayers decrease the initial sticking probability for H2 as well as the peak temperature for H 2 desorption. Winkler and Rendulicls have investigated the effects of oxygen preadsorption on the sticking coefficients for H, on Ni(ll1) and Ni( 1IO). They have observed reduced initial sticking coefficients for H2 on oxygen-covered Ni( 1lo), but an initial increase with a subsequent decrease in initial sticking coefficient for H2 with oxygen coverage on Ni( 111). In this work we have initiated an investigation of the dynamical characteristics of the adsorption of hydrogen on the Ni( 100) surface. The magnitude of the barrier, the threshold behavior of the adsorption, and the effect of oxygen impurites on the barrier were examined. Changes in the adsorption kinetics with translational and rotational energy of the incident hydrogen beam were also probed.
2.0. Experimental Section The reactive scattering apparatus has been described in detail previously.21-22Briefly, a triply differentially pumped supersonic molecular beam source directed hydrogen at the Ni( 100) surface in the center of the ultrahigh vacuum (UHV) scattering chamber (base pressure