874
J . Phys. Chem. 1990, 94, 874-882
z =! o.30[:
t I cl
I
1
c3
c6
I
CIZ
A L K Y L CHAIN L E N G T H
Figure 7. Normalized deuterium modulation depth from alkylphenothiazines in vesicles for 10 min of irradiation with a WG320 filter versus the alkyl chain length. The symbols (0)and ( A ) show PCS0,Na ( n = 3, 6, 12) and C,PSO,Na ( m = 1, 12) in DODAC vesicles, while ( 0 ) and (A)show PC,SO,Na and C,PS03Na in DHP vesicles. The error bars indicate the average deviation of two or more determinations.
electrons from PC3S03Na can more easily escape across the positive vesicle interface than those from longer alkyl chain phenothiazines. This model is supported by the ESEM data. The normalized modulation depths30 are plotted against the alkyl chain length of A P in Figure 7 . The locations of PC$03Na ( n = 3 , 6 , and 12) (30) Szajdzinska-Pietek,E.; Maldonado, R.; Kevan, L.;Berr, S. S.;Jones, R. R. M. J . Phys. Chem. 1985.89,1547.
in DODAC vesicles are in a less hydrated region with increasing alkyl chain length consistent with being located deeper into the vesicle relative to the interface. The same trend is shown for CmPS03Na( m = 1, 12) with only two data points. In the DHP vesicles, the CmPS03Namolecules are located in a more hydrated region than the PC,,S03Na molecules. In addition, the hydration of PC$03Na increases with decreasing alkyl chain length. In spite of these differences in hydration, the photoionization yields are constant. In D H P vesicles with a negative surface charge, the photoionization yields of AP are not affected by the location of the AP or the position of the sulfonate group in the AP. This suggests that the photoionized electrons from AP are impeded from escaping from DHP vesicles by the negative surface charge, and therefore, the back-reaction from AP+ to AP is promoted. The vesicle surface charge seems to be more effective than the location of the sulfonate moiety for net photoionization. The effective factors in photoionization of alkylphenothiazinesulfonates in vesicles are the surface charge of the vesicles and the distance of the alkylphenothiazinesulfonates from the vesicle surface. The highest photoionization yield is obtained for the shortest phenothiazine moiety to interface distance and a positive vesicle interface charge.
Acknowledgment. This research was supported by the Division of Chemical Sciences, Office of Basic Energy Sciences, Office of Energy Research, US.Department of Energy. Registry No. C,,PSO,Na, 106049-51-8; C,PSO,Na, 124177-19-1; PC12S03Na,74339-97-2; PC6S03Na, 101199-39-7; PC3S03Na, 101199-38-6;Br(CH2),,CH3,143-15-7;Br(CH2),,Br, 3344-70-5; Br(CH,),Br, 629-03-8;Br(CH2),Br, 109-64-8; 1OH-phenothiazine, 92-84-2; 10-dodecyl-1OH-phenothiazine, 73487-07-7; IO-methyl-lOH-phenothiazine, 1207-72-3; lo-( 12-bromododecy1)-10H-phenothiazine, 8054836-3; 10-(6-bromohexyl)-1OH-phenothiazine, 101199-33-1 ; 10-(3bromopropyl)-10H-phenothiazine, 92357-95-4.
Photophysics and Photochemistry of Trb( 2,2'-bipyridyl)ruthenium( I I ) within the Layered Inorganic Solid Zirconium Phosphate Sulfophenylphosphonate Jorge L. C o b , Chao-Yeuh Yang, Abraham Clearfield,* and Charles R. Martin* Department of Chemistry, Texas A&M University, College Station, Texas 77843 (Received: May 30, 1989; In Final Form: July 31, 1989)
The photophysics and photochemistry of tris(2,2'-bipyridyl)ruthenium(II) (Ru(bpy),2+)adsorbed into the layered solid zirconium phosphate sulfophenylphosphonateare described. The decay kinetics of the metal complex are shown to depart from first-order behavior. Albery's model of dispersed kinetics, which assumes a continuous distribution of rate constants, is used to explain the decay kinetics. The oxidative quencher methylviologen (MV2+) is shown to react with Ru(bpy),2+in ZrPS via a combined dynamic and quasi static (sphere of action) quenching mechanism.
Introduction Much of the effort in solar energy research has focused on the use of electron-transfer reactions to convert and store solar energy. To meet this goal the energy-releasing back reaction has to be suppressed.'-3 Recently, several strategies for suppressing the back reaction and enhancing the charge separation efficiency have been devised.'-3 These new strategies make use of interfaces, surfaces, micelles, polyelectrolytes, and other heterogeneous microenvironments to obtain efficient charge separation. The layered zirconium phosphates4 are heterogeneous systems which could prove useful as media for solar energy conversion. Zirconium phosphates are acidic, inorganic, ion-exchange materials *Towhom correspondence should be addressed.
having a layered ~ t r u c t u r e .These ~ materials have potential applications as catalysts, ion exchangers, solid electrolytes, and hosts for various intercalants.5-7 Organic derivatives of a-zirconium phosphate (a-ZrP, which has the formula Zr(HP0,)2-H20) have (1) Connolly, J. S. Photochemical Conversion and Storage of Solar Energy; Academic: New York, 198 1. (2) Energy Resources Through Photochemistry and Catalysis; Gratzel, M., Ed.; Academic: New York, 1983. (3) Proceedings of the 6th International Conference on Photochemical Conversion and Storage of Solar Energy. N o w . J . Chem. 1987,II,79-207. (4)Inorganic Ion Exchange Materials; Clearfield, A., Ed.; CRC: Boca Raton, FL, 1982. ( 5 ) Alberti, G. Acc. Chem. Res. 1978,11, 163. (6)Intercalation Chemistry; Whittingham, M. S . , Jacobson, A. J., Eds.; Academic: New York, 1982. (7) Clearfield, A. Chem. Rev. 1988,88, 125.
0022-3654/90/2094-0874$02.50/00 1990 American Chemical Society
Photophysics and Photochemistry of Ru(bpy)32+ recently been synthesized. These new layered compounds usually have increased interlayer space, reactivity, and selectivity, relative to a-ZrP.*+ We have recently reported preliminary results of characterizations of one of these organic derivatives of a-ZrP, zirconium phosphate sulfophenylphosphonate (ZrPS).16 In those studies the luminescence probe ion tris(2,2’-bipyridyl)ruthenium(II) (Ru(bpy)32+)was used to obtain information about the microenvironment within this layered solid. We have recently investigated excited-state reactions between Ru(bpy),*+ and the quencher methylviologen (MV2+) within ZrPS. We have found that the layered structure of ZrPS affects the lifetime of the excited state of R ~ ( b p y ) ~ Albery’s ~+. model for dispersed kinetics in heterogeneous systems” was used to describe the Ru(bpy),2+ decay kinetics in ZrPS. We have also found that the excited-state quenching reaction of Ru(bpy)?+ with MV2+ in this system occurs via a combination of diffusional (dynamic/collisional) and sphere of action quenching. The results of these investigations are reported here.
Experimental Section Materials. Ru(bpy),C12dH20 was obtained from G. F. Smith dichloride and used as received. l,l’-Dimethyl-4,4’-bipyridinium hydrate (the cation is called methylviologen, MVZ+)was obtained from Aldrich and used as received. Water was either triply distilled or circulated through a Milli-Q water purification system (Millipore Corp.). All other reagents and solvents were of the highest grade available and were used without further purification. Procedures. The ZrPS, Zr(HP04)(03P-C6H4S03H),was prepared as described previously.Is Ru(bpy)?+ was incorporated (loaded) into ZrPS as described in our earlier study.I6 A typical procedure was as follows: 50 mg of the ZrPS was suspended in water; 6 mL of a 6.7 X lo4 M aqueous solution of Ru(bpy)?+ was added, and the mixture was shaken. Forty microliters of this mixture was then used to coat a quartz slide (Esco Products) and the solvent was allowed to evaporate overnight at room temperature. This procedure produced a thin (ca. 1 pm) film of Ru(bpy),2+-loaded ZrPS coated onto the quartz slide. The equilibrium water content of these films was ca. 5% (100 X wt of H 2 0 / w t of dry film). The Ru(bpy),2+ content was varied by varying the quantity of Ru(bpy),” added to the suspension. ZrPS films containing both R ~ ( b p y ) , ~and + MV2+ were obtained by adding measured amounts of both cations to the ZrPS suspension. In the quenching experiments, 2% of the ionic sites in the ZrPS were loaded with Ru(bpy),2+. This low level of loading was used to ensure that no self-quenching of the probe molecules occurred during these experiments. In this way, any observed quenching is the result of interactions between the probe and the quencher ions. Instrumentation. Steady-state emission spectra were obtained with a Spex Fluorolog 2 spectrofluorometer. The samples were excited with a 450-W xenon lamp. The excitation wavelength was 452 nm. The excitation and emission slits were set at 1.25 (8) Alberti, G.; Constantino, U.; Allulli, S.;Tomassini, N. J. Inorg. Nucl. Chem. 1978, 40, 1 1 13. (9) Dines, M. B.; DiGiacomo, P. M. Inorg. Chem. 1981, 20, 92. (IO) Dines, M. B.; DiGiacomo, P. M.; Callahan, K. P.; Griffith, P. C.; Lane, R. H.; Cooksey, R. E. In Chemically Modified Surfaces in Catalysis and Electrocatalysis; Miller, J. S., Ed.; ACS Symposium Series 192; American Chemical Society: Washington, DC, 1982; pp 223-240. ( I 1) DiGiacomo, P. M.; Dines, M. B. Polyhedron 1982, 1, 61. (12) Alberti, G.; Constantino, U. J . Mol. Catal. 1984, 27, 235. (13) Ferragina, C.; La Ginestra, A.; Massucci, M. A.; Patrono, P.; Tomlinson, A. A. G.J . Phys. Chem. 1985,89, 4762. (14) Alberti, G. In Recent Deuelopments in Ion Exchange; Williams, P. A,, Hudson, M. J., Eds.; Elsevier Applied Science: London, 1987; pp 233-248. ( I 5) Alberti, G.; Constantino, U.; Marmottini, F. In Recent Developments in Ion Exchange; Williams, P. A., Hudson, M. J., Eds.; Elsevier Applied Science: London, 1987; pp 249-256. (16) ColBn, J. L.; Yang, C.-Y.; Clearfield, A,; Martin, C. R. J . Phys. Chem. 1988, 92, 5777. (17) Albery, W. J.; Barlett, P. N.; Wilde, C. P.; Darwent, J . R. J . Am. Chem. SOC.1985, 107, 1854. (18) Yang, C.-Y.; Clearfield, A. React. Polym., Ion Exch., Sorbents 1987, 5, 1 3 .
The Journal of Physical Chemistry, Vol. 94, No. 2, 1990 875 mm, yielding a monochromator bandwidth of 4.6 nm. Luminescence was detected perpendicular to the incident radiation with a Hamamatsu R928 photomultiplier tube which was configured for photon counting. The Spex solid sample holder (front-face viewing geometry) was used. A Spex Datamate digitized and displayed the emission data. Electronic absorption spectra were recorded on a Perkin-Elmer Lambda 4B spectrophotometer. X-ray powder diffraction patterns were obtained with a Seifert-Scintag automated powder diffractometer (PAD 11) with a Ni-filtered Cu Ka radiation (A = 1.5418 A). Luminescence lifetimes measurements were obtained by using time-correlated single photon counting (TCSPC)’9-z0on an Edinburgh Instruments Model 199 fluorescence lifetime spectrometer. The excitation source was the 337-nm line of a coaxial thyratron-gated N2 flashlamp (Edinburgh Instruments Model 199F) (lo0 ns) the excitation pulse width (fwhm
I
L
0
% 0.000
0.200
0.400
0.600
[Mv2+]/M Figure 12. Sphere of action fit to the steady-state emission quenching data (assuming dispersed kinetics) in ZrPS. Vis the volume of the sphere of action, and N'is 6.02 X lozo.
than the sum of the radii of the reactants. For a situation where quenching has diffusional and static components the Perrin model can be combined with the diffusional Stern-Volmer model. The modified form of the Stern-Volmer equation which describes the diffusional plus sphere of action model is3'
Io/I = (1 + &[Q]) ex~([QIVN/1000)
(15)
where Vis the volume of the sphere, N is Avogadro's number, and Kd is the dynamic quenching Stern-Volmer constant. The dynamic portion of the quenching can be isolated by dividing the this division by raw Io/I data by the variable exp([Q] VN/lOOO); the exponential factor accounts for the proximity effect, and a linear Stern-Volmer plot should be obtained. This linear Stern-Volmer plot should be superimposable with the T,,/T Stern-Volmer plot obtained from lifetime measurements; if this is true this portion of the quenching is dynamic.37 Figure 12 shows the quenching data plotted in the form of the dynamic plus sphere of action model (eq 15). The plot shows that the division of the I o / I data by the exponential variable gives a curve collinear with the lifetime data; this collinearity means that the dynamic plus sphere of action model is correct. It is important to note that the only adjustable parameter used to fit this model is the volume of the sphere of action. A sphere of action with a radius of 10.8 A is obtained from the fit in Figure 12. The 10.8-A radius obtained from the data in Figure 12 is slightly larger than the sum of the geometric radii of R ~ ( b p y ) , ~ + and MV2+,which is I O A.54*55 The 10.8-Aradius is in excellent agreement with the literature value of 10.9 8, obtained by McLendon et al.54 for the R ~ ( b p y ) , ~ + - M v ~system + in rigid glycerol solution. Thus, the use of the Albery's dispersed kinetics model to analyze the lifetime data gives reasonable results for the quenching mechanism. In contrast, if the monoexponential decay model is used, the analysis gives an unreasonably large radius of the sphere of action of 12.8 A. Therefore, the dispersed kinetics model gives a more reasonable result than the exponential model for the radius of the sphere of action. The dispersed kinetics model not only fits the experimental decay curves better than the monoexponential decay model, but it gives more reasonable results for the quenching mechanism. The dispersed kinetics model is then the preferred model for the analysis of the quenching luminescence transients in ZrPS. From the diffusional component to the quenching mechanism we can calculate the diffusional bimolecular quenching rate constant k,. The Kd value is 9.83 M-' and the R ~ ( b p y ) , ~lifetime + in ZrPS is 806 ns. The k , value is calculated to be 1.2 X lo7 M-l s-I ( k 4 = &/To). The k, value obtained in ZrPS is smaller than the value obtained in water ((3-5) X IO8 M-' s-1),363Q7but it (54) Guarr, T.; McGuire, M.; Strauch, S.; McLendon, G. J . Am. Chem. Soc. 1983, 105, 616. ( 5 5 ) Stradowski, Cz.; Wolszczak, M. In Photochemistry and Photophysics of CFrdination Compounds; Yersin, H . , Vogler, A., Eds.; Springer-Verlag: Berlin, 1987; pp 125-1 28.
is larger than the value obtained in the layered clay hectorite by Ghosh and Bard (k, = 1.1 X lo6 M-' s-').,~The smaller k, value in ZrPS than in water indicates a slower movement of quencher ions through the interlayer space of ZrPS, which is also evident from the small D. The higher k, value in ZrPS than in hectorite indicates that, although the diffusion of ions in ZrPS is not very fast, it is faster than in hectorite. In their study of quenching in hectorite clay Ghosh and Bard36 claimed that the low k, value observed was due to segregation of the MV2+ and R ~ ( b p y ) , ~molecules + in different clay layers. In hectorite clay the emission kinetics of adsorbed R ~ ( b p y ) , ~ + in the presence of MV2+were equal to the kinetics in the absence of MVZ+. In ZrPS the possibility of segregation is ruled out since our emission decay profiles are distinctively different in the presence and absence of MV2+. The decay measurements prove that during the excited-state lifetime the quencher ions interact with the probe molecules. No spatial separation due to segregation occurs in the ZrPS layers. Assuming that the excited-state reaction is diffusion controlled, the Smoluchowski equation (eq 10) can be used, with the previously determined diffusion coefficient for R ~ ( b p y ) in ~ ~ZrPS, + to calculate the diffusion coefficient for MV2+ in ZrPS; a D of 1.6 X lo-@cm2 s-' is obtained. This D is 2 orders of magnitude cm2 s-I). Thus, larger than the D value for Ru(bpy)?+ (3.3 X the smaller MV2+ion has higher mobility through the interlayer space of ZrPS than Ru(bpy),2+. The rms distance (R,)travelled by MV2+during the R ~ ( b p y ) , ~ + excited-state lifetime can be calculated from the diffusion coefficient by using eq 1 1 ; a R, of 16.6 A is obtained. Compared to the R, of Ru(bpy):+ (2.3 A), the R, of MV2+indicates that during the lifetime of the excited state R ~ ( b p y ) , ~barely + moves, whereas MV2+ travels a longer distance. The difference in mobilities between R ~ ( b p y ) , ~and + MV2+ occurs because R ~ ( b p y ) , ~is+ a much bigger ion than MV2+. Ru(bpy):+ (a spherical ion, 14 8, in diameter) can accommodate itself between the layers of ZrPS, but its movement is restricted by the protruding phenylphosphonate groups in the structure of ZrPS. The space between the protruding phenylphosphonate groups is not big enough to facilitate diffusion of the R ~ ( b p y ) , ~ + ions. MV2+, being a smaller ion (1 3.4 A in length, 6.4 A in width, 3.4-4 8,in thickness46A7),can tumble around, reptate, and diffuse more easily within this crowded microenvironment than Ru(bpy)?+. The space between phenyl rings in ZrPS is large enough for MV2+ to move more easily through ZrPS than R ~ ( b p y ) , ~ + . Therefore, the difference in size of the ions with respect to the structure of ZrPS governs their mobility.
Conclusions The microenvironment within ZrPS restrains the movement of ions through the interlayer space, but diffusion leading to dynamic quenching reactions can occur. A model combining diffusional quenching and sphere of action quenching accounts for quenching of R ~ ( b p y ) , ~by + MV2+ in ZrPS. These results provide another example of the rich variety of excited-state reactions in heterogeneous systems. In the case of ZrPS the kinetics of the reactions reported here must be described by taking into account the heterogeneous nature of ZrPS. New derivatives of a-ZrP similar in structure to ZrPS can be synthesized with increased distance between the phenyl rings. This larger space should increase the mobility of ions through ZrPS. Furthermore, studies on layered systems with well-defined separation between an excited probe molecule and a quencher can provide insights into the distance dependence of electron transfer. Studies of the distance dependence of electron-transfer reactions in such systems are an extremely interesting and active area of research. We are currently pursuing these efforts and hope to report our results in the near future. (56) Gratzel, M.; Kiwi, J.; Kalyanasundaram, K. H e l a Chim. Acra 1978.
61, 2720.
(57) Darwent, J. R.; Kalyanasundaram, K. J . Chem. Soc., Faraday Tram. 2 1981, 77, 373.
882
J . Phys. Chem. 1990, 94, 882-885
Acknowledgment. Financial support for this work was provided by the Robert A. Welch Foundation, the Office of Naval Research, and the Air Force Office of Scientific Research. We gratefully acknowledge partial support of this research by the Regents of Texas A&M University through the AUF-sponsored Materials
Science and Engineering Program. J.L.C. acknowledges support from the Texas A&M University Minority Merit Fellowship. Registry No. MV2+, 4685-14-7; Ru(bpy)32+, 15158-62-0; ZrPS, 4685-14-7.
Valence Band Spectra of Acetylenic Moieties on Zinc Oxide: Acetylide versus Propargyl Formation J. M. Vohst and M. A. Barteau* Center for Catalytic Science and Technology, Department of Chemical Engineering, University of Delaware, Newark, Delaware 19716 (Received: June 5, 1989)
Acetylene adsorbs dissociatively on the ZnO(0001) surface, as on other metal oxides, to form acetylide intermediates, CCH. Propyne also dissociates via surface acid-base reactions; however, potential adsorbates include methylacetylide (CH,CC) and propargyl (CH2CCH)species. Ultraviolet photoelectron spectra (UPS) of the adsorbed intermediates formed from propyne and allene on the ZnO(0001) surface demonstrate that both react to form propargyl intermediates. UPS also provide useful fingerprints of o-bonded acetylides on both metal and metal oxide surfaces.
competing surface intermediates and products formed in the reIntroduction actions of individual Brernsted or Lewis acids. For example, while Zinc oxide has long served as a model material for the study higher aldehydes undergo nucleophilic attack by lattice oxygen of the surface acid-base chemistry of metal oxides. The pioneering on the (0001)-Zn polar surface of ZnO, the subsequent elimination efforts of Kokes and co-workers’ demonstrated that dissociative step preferentially removes alkyl rather than hydride ligands at adsorption of Brernsted acids was heterolytic in nature and occurred low temperature.16J5 The unusual selectivity of this reaction on Zn-0 cation-anion pair sites on the surface. The principal toward surface alkyl and formate over hydride and higher carevidence for this model was derived from infrared spectroscopy boxylate moieties appears to be due in part to the weakness of and from studies of the kinetics and selectivity of isotope-exchange the cy C-C relative to the a C-H bonds of these aldehydes and and isomerization reactions of unsaturated hydrocarbons.’-* Of in part to the affinity of the surface for the eliminated alkyl^.'^,^^ predictive value was the observation’ that B r ~ n s t e dacids with pK,’s less than 36 (e.g., propylene) dissociated to form carbanions on zinc oxide, while weaker acids did not adsorb dissociatively. This description of the surface chemistry of zinc oxide and of (1) Kokes, R. J. Intra-Sci. Chem. Rep. 1972, 6, 77. the requirement of ion-pair sites for acid-base reactions has held (2) Dent, A. L.; Kokes, R. J. J . Phys. Chem. 1969, 73, 3772. up quite well, both in its extension to other metal oxide^^*'^ and (3) Chang, C. C.; Kokes, R. J. J . Am. Chem. SOC.1970, 92, 7517. in studies of well-defined zinc oxide surfaces utilizing surface (4) Dent, A. L.; Kokes, R. J. J . Am. Chem. SOC.1970, 92, 1092. spectroscopic techniques.”-’4 Recent work from this laboratory (5) Chang, C. C.; Kokes, R. J. J . Catal. 1975,38,491. has shown that, of the two polar surfaces of zinc oxide, only the (6) Chang, C. C.; Kokes, R. J. J . Catal. 1973, 28, 92. (0001)-Zn polar surface presents accessible Zn-O site pairs and (7) Dent, A. L.; Kokes, R. J. J . Phys. Chem. 1971, 75, 487. is therefore active for dissociative adsorption of Brernsted acids. (8) Chang, C. C.; Conner, W. C.; Kokes, R. J. J . Phys. Chem. 1973, 77, The lattice oxygen ions exposed on this surface not only are good 1957. Brernsted bases but also act as Lewis bases in the nucleophilic (9) Gamone, E.; Stone, F. S. Proc. Int. Congr. Catal., 8th 1984, 3, 441. oxidation of carbonyl c o m p o ~ n d s . ~ IReactions ~ ~ ~ , ~ ~reported to (IO) Tanabe, K. In Catalysis Science and Technology; Anderson, J. R., Boudart, M., Eds.; Springer-Verlag: Berlin, 1980; Vol. I , p 232. date on the ion-pair sites of the Zn polar face of ZnO single (1 1) Vohs, J. M.; Barteau, M. A. Surf. Sci. 1986, 176, 91. crystals include the dissociative adsorption of alcohols,”J7-24 (12) Vohs, J. M.; Barteau, M. A. J . Phys. Chem. 1987, 91, 476. carboxylic a c i d ~ , ” J ~ *and ” ~ ~alkynes,I2 as well as oxidation of (13) Vohs, J. M.; Barteau, M. A. Surf. Sci. 1988, 201, 481. aldehydes”*’6q25and esters.I5 (14) Barteau, M. A,; Vohs, J. M. In Successful Design of Catalysts; Inui, There is a significant body of evidence, however, which suggests T., Ed.; Elsevier: Amsterdam, 1988; p 89. that this simple acid-base model of oxide activity may be inad(15) Vohs, J. M.; Barteau, M. A. Surf. Sci. 1988, 197, 109. equate to describe the subtleties of the selectivity of surface re(16) Vohs, J. M.; Barteau, M. A. J . Catal. 1988, 213, 497. actions. For example, if one considers the competitive adsorption (17) Vohs, J. M.; Barteau, M. A. Surf. Sci., in press. of Brernsted acids from the gas phase, their relative affinities for (18) Zwicker, G.;Jacobi, K.; Cunningham, J. Inr. J . Mass Spectrosc. Ion dissociative adsorption on both ZnO and MgO are better acProcesses 1984, 60, 213. counted for by their aqueous dissociation constants than by the (19) Akhter, S.; Lui, K.; Kung, H. H. J . Phys. Chem. 1985, 89, 1958. thermodynamics of heterolytic dissociation in the gas p h a ~ e . ~ ~ ~(20) ~ Cheng, W. H.; Akhter, S.; Kung, H. H. J . Catal. 1983, 82, 341. These observations have been explained in terms of a ”solvating” (21) Akhter, S.; Cheng, W . H.; Lui, K.; Kung, H. H. J . Catal. 1984, 85, 431. effect of the surface in stabilizing the conjugate base anions (22) Lui, K.; Vest, M.; Berlowitz, P.; Akhter, S.; Kung, H. H. J. Phys. produced by dissociative adsorption? In effect, the negative charge Chem. 1986, 90, 3138. on the conjugate base is reduced by its interaction with surface (23) Mokwa, W.; Kohl, D.; Heiland, G . Surf. Sci. 1982, 117, 659. cation sites. Such effects may also arise in the selectivity toward Present address: Department of Chemical Engineering, University of Pennsylvania, Philadelphia, PA 19104. * Author to whom correspondence should be addressed
0022-3654/90/2094-0882$02.50/0
(24) Berlowitz, P.; Kung, H. H. J . Am. Chem. SOC.1986, 108, 3532. (25) Vohs, .I.M.; Barteau, M. A. Langmuir 1989, 5, 965. (26) Spitz, R . N.; Barton, J. E.; Barteau, M. A,; Staley, R. H.; Sleight, A. W . J . Phys. Chem. 1986, 90, 4067.
0 1990 American Chemical Society
