Acetone and Water on TiO2(110): Competition for ... - ACS Publications

Langmuir 2005, 21, 3443-3450

3443

Acetone and Water on TiO2(110): Competition for Sites Michael A. Henderson* Interfacial Chemistry and Engineering Group, Pacific Northwest National Laboratory, P.O. Box 999, MS K8-93, Richland, Washington 99352 Received September 20, 2004. In Final Form: January 24, 2005 The competitive interaction between acetone and water for surface sites on TiO2(110) was examined using temperature programmed desorption (TPD). Two surface pretreatment methods were employed, one involving vacuum reduction of the surface by annealing at 850 K in ultrahigh vacuum (UHV) and another involving surface oxidation with molecular oxygen. In the former case, the surface possessed about 7% oxygen vacancy sites, and in the latter, reactive oxygen species (adatoms and molecules) were deposited on the surface as a result of oxidative filling of vacancy sites. On the 7% oxygen vacancy surface, excess water displaced all but about 20% of a saturated d6-acetone first layer to physisorbed desorption states, whereas about 40% of the first layer d6-acetone was stabilized on the oxidized surface against displacement by water through a reaction between oxygen and d6-acetone. The displacement of acetone on both surfaces is explained in terms of the relative desorption energies of each molecule on the clean surface and the role of intermolecular repulsions in shifting the respective desorption features to lower temperatures with increasing coverage. Although first layer water desorbs from TiO2(110) at slightly lower temperature (275 K) than submonolayer coverages of d6-acetone (340 K), intermolecular repulsions between d6-acetone molecules shift its leading edge for desorption to 170 K as the first layer is saturated. In contrast, the desorption leading edge for first layer water (with or without coadsorbed d6-acetone) shifted to no lower than 210 K as a function of increasing coverage. This small difference in the onsets for d6-acetone and water desorption resulted in the majority of d6-acetone being compressed into islands by water and displaced from the first layer at a lower temperature than that observed in the absence of coadsorbed water. On the oxidized surface, the species resulting from reaction of d6-acetone and oxygen was not influence by increasing water coverages. This species was stable up to 375 K (well past the first layer water TPD feature) where it decomposed mostly back to d6-acetone and atomic oxygen. These results are discussed in terms of the influence of water in inhibiting acetone photo-oxidation on TiO2 surfaces.

1. Introduction Competition between two or more chemical species often influences reaction rates and pathways in heterogeneous catalytic reaction rates. Site competitors can be reactants, solvents, intermediates, products, or spectators. Water is often in one or more of these roles in oxide-based catalysis because it is both a prevalent and strongly bound surface species.1 Of particular relevance to this work is the influence of water on photo-oxidation of organics on TiO2 materials.2-6 Water is thought to affect photo-oxidation reactions on TiO2 in two primary ways: it enhances reactivity by supplying OH groups for radical-initiated chemistry and it poisons reactivity by blocking key active sites. Detailed coadsorption studies involving model oxide surfaces can aid in understanding the competitive and cooperative interactions of water with other coadsorbates. In this paper, the competition between water and acetone for adsorption sites on TiO2 were probed using temperature programmed desorption (TPD) and the (110) crystal face of rutile TiO2. This work builds on a previous examination of the interaction of acetone with the O2oxidized and vacuum-reduced surfaces of TiO2(110)7 and, in concert with the accompanying paper to this one that * E-mail: [email protected]. (1) Henderson, M. A. Surf. Sci. Rep. 2002, 46, 1. (2) Lukaski, A. C.; Muggli, D. S. J. Catal. 2004, 223, 250. (3) Fu, X.; Zeltner, W. A.; Anderson, M. A. Stud. Surf. Sci. Catal. 1996, 103, 445. (4) Yamakata, A.; Ishibashi, T.-a.; Onishi, H. J. Phys. Chem. B 2003, 107, 9820. (5) Muggli, D. S.; Backes, M. J. J. Catal. 2002, 209, 105. (6) Coronado, J. M.; Zorn, M. E.; Tejedor-Tejedor, I.; Anderson, M. A. Appl. Catal., B 2003, 43, 329. (7) Henderson, M. A. J. Phys. Chem. B 2004, 108, 18932.

focuses on isotopic exchange between water and acetone,8 lays the groundwork for a detailed look at the photooxidation of acetone on TiO2(110).9 Without question, the TiO2(110) surface has emerged as the prototypical and most well-characterized oxide single-crystal surface.10 This surface has also demonstrated utility in understanding photochemical processes (see references in ref 10, as well as other recent examples11-14). Photocatalytic studies on TiO2 surfaces have shown that acetone photo-oxidation rates are inhibited when water is present in the reactant mixture,15-21 although several groups also indicate that some enhancement in the rate is achieved at low relative humidity (RH)6,16,18,22 and one group did not observe rate inhibition at high RH.6 The first step to understanding (8) Henderson, M. A. Langmuir 2005, 21, 3451. (9) Henderson, M. A., J. Phys. Chem. B, submitted for publication. (10) Diebold, U. Surf. Sci. Rep. 2003, 48, 53. (11) Idriss, H.; Legare, P.; Maire, G. Surf. Sci. 2002, 515, 413. (12) Henderson, M. A.; White, J. M.; Uetsuka, H.; Onishi, H. J. Am. Chem. Soc. 2003, 125, 14974. (13) White, J. M.; Szanyi, J.; Henderson, M. A. J. Phys. Chem. B 2003, 107, 9029. (14) Diwald, O.; Thompson, T. L.; Zubkov, T.; Goralski, E. G.; Walck, S. D.; Yates, J. T., Jr. J. Phys. Chem. B 2004, 108, 6004. (15) Pearl, J.; Ollis, D. F. J. Catal. 1992, 136, 554. (16) Chen, S.; Cheng, X.; Tao, Y.; Zhao, M. J. Chem. Technol. Biotechnol. 1998, 73, 264. (17) Raillard, C.; Hequet, V.; Le Cloirec, P.; Legrand, J. J. Photochem. Photobiol. A 2004, 163, 425. (18) Chang, C. P.; Chen, J. N.; Lu, M. C. J. Environ. Sci. Health A 2003, 38, 1131. (19) Vorontsov, A. V.; Kurkin, E. N.; Savinov, E. N. J. Catal. 1999, 186, 318. (20) Kim, S. B.; Hong, S. C. Appl. Catal., B 2002, 35, 305. (21) Choi, W.; Ko, J. Y.; Park, H.; Chung, J. S. Appl. Catal., B 2001, 31, 209-. (22) El-Maazawi, M.; Finken, A. N.; Nair, A. B.; Grassian, V. H. J. Catal. 2000, 191, 138.

10.1021/la0476579 CCC: $30.25 © 2005 American Chemical Society Published on Web 03/04/2005

3444

Langmuir, Vol. 21, No. 8, 2005

Figure 1. Mass 46 (bold trace) and mass 18 (dashed trace) TPD spectra from separate multilayer coverages of d6-acetone and H2O, respectively, on the reduced TiO2(110) surface.

the inhibition of acetone photo-oxidation by water is to examine how these two molecules interact on a welldefined oxide surface possessing a homogeneous surface site distribution. To better understand the collective interactions of water and acetone on TiO2(110), it is first important to consider how these molecules independently interact with the surface. The properties of water on TiO2(110) have been the subject of extensive experimental and theoretical efforts.1 While disagreements persist regarding the reactivity of water on the ideal TiO2(110) surface, UHV studies suggest that water does not dissociate without the influence of surface defects. Figure 1 shows a TPD spectrum from a multilayer coverage (∼3.5 ML (monolayer)) of water on the clean TiO2(110) surface possessing roughly 7% oxygen vacancy sites. Dissociative adsorption of water at oxygen vacancy sites yields in recombinative desorption of the resulting bridging OH groups in TPD at 500 K23,24 (not shown in Figure 1). The main desorption features from molecularly adsorbed water are at 160, 180, and 275 K, corresponding to desorption from the multilayer, from the second layer (water molecules hydrogenbonded to bridging O2-), and from the first layer (water molecules coordinated to Ti4+ sites).23,25 Water molecules that dissociate at step edges24 or through the influence of coadsorbed oxygen26 recombinatively desorb slightly above RT. The chemistry of acetone on TiO2(110) has only recently been examined.7 Desorption of low coverages from TiO2(110) possessing 7% oxygen vacancies peaked at about 345 K but shifted toward lower temperature as the (23) Hugenschmidt, M. B.; Gamble, L.; Campbell, C. T. Surf. Sci. 1994, 302, 329. (24) Henderson, M. A. Langmuir 1996, 12, 5093. (25) Henderson, M. A. Surf. Sci. 1996, 355, 151. (26) Epling, W. S.; Peden, C. H. F.; Henderson, M. A.; Diebold, U. Surf. Sci. 1998, 412/413, 333.

Henderson

coverage was increased. In the first layer regime, acetone molecules coordinate to Ti4+ sites in η1-configurations via lone pair electrons on the oxygen end of each molecule on the basis of HREELS measurements. This bonding configuration results in a situation in which intermolecular repulsions increased with increasing coverage because the individual molecular dipoles were registered to the surface in opposition to each other (i.e., in an ‘oxygen end down’ geometry). At first layer saturation, all available Ti4+ sites are occupied by acetone molecules. TPD from this saturated first layer yielded a broad desorption feature that spanned the range from ∼170 to 400 K, as shown in Figure 1. The width of the desorption feature arises from strong coverage dependence resulting from changes in the degree of intermolecular repulsion between neighboring molecules as the coverage was depleted by desorption. TPD of coverages above 1 ML also yielded a feature at 135 K corresponding to multilayer desorption. No decomposition products were observed on the reduced surface, and acetone showed no obvious propensity for adsorption at oxygen vacancy sites over nondefect sites. In contrast, about 7% of the acetone first layer decomposed if the TiO2(110) surface was pre-oxidized at 95 K with O2. About 0.02 ML of acetate was detected in TPD, with the remaining 0.05 ML amount of decomposition was not identified. Pre-oxidation of the surface also resulted in stabilization of about 0.25 ML acetone as the result of a reaction between acetone and oxygen to form an ‘acetoneoxygen’ surface complex. As will be shown below, the presence or absence of this acetone-oxygen interaction dictates how water influences acetone. 2. Experimental Section The base pressure of the UHV system used in this study was 2 × 10-10 Torr. The TiO2(110) crystal used in this study was obtained from First Reaction with dimensions of 10 mm × 10 mm × 1.5 mm. It was epi-polished on both sides to provide maximum thermal contact with the Au-foil-covered heating plate. Details on the cleaning and surface characterization are discussed elsewhere.7 The TiO2(110) surface possessing 7% oxygen vacancies was prepared by annealing the crystal in UHV at 850 K for 10 min prior to each experiment. The surface oxygen vacancy population was quantified using water TPD.24 In the oxidized surface case, the vacuum-annealed surface was exposed to g20 L of O2 at 95 K followed by flashing to RT before recooling for gas exposures. As discussed elsewhere,26-28 this O2 pretreatment method oxidizes vacancies and deposits reactive atomic and/or molecular oxygen species on the surface. Research-grade d6-acetone and H2O were obtained from Aldrich and were further purified using LN2 freeze-pumpthaw cycles. A calibrated directional doser was used for exposing either gas to the TiO2(110) surface. All gas exposures were performed with the crystal at 95 K. For dosing, the sample was positioned within 1 mm of the doser tube and the exposed area was approximated by the i.d. of the tube. Coverages were estimated from the total exposure of the gas emitted from the calibrated doser during the dose with the assumption that both d6-acetone and water adsorbed on TiO2(110) at 95 K with near unity sticking probabilities. In this study, coverages are expressed in units of ML where ‘1 ML’ is equivalent to the areal density five-coordinated Ti4+ cation sites on the ideal TiO2(110) surface (5.2 × 1014 molecules/cm2). The heating rate for all TPD measurements was 2 K/s.

3. Results and Discussion 3.1. Site Competition on the Reduced TiO2(110) Surface. Figure 2 shows the effect of water adsorption (27) Henderson, M. A.; Epling, W. S.; Perkins, C. L.; Peden, C. H. F.; Diebold, U. J. Phys. Chem. B 1999, 103, 5328. (28) Schaub, R.; Wahlstroem, E.; Ronnau, A.; Laegsgaard, E.; Stensgaard, I.; Besenbacher, F. Science 2003, 299, 377.

Acetone and Water on TiO2(110): Competition for Sites

Figure 2. Mass 46 (lower five traces) and mass 18 (inset) TPD spectra from various H2O coverages (0.02, 0.29, 0.58, 1.3, and 3.5 ML) adsorbed on ∼1 ML of d6-acetone on the reduced TiO2(110) surface. The upper mass 46 trace is from ∼1 ML of d6acetone adsorbed on 3.5 ML of H2O on the reduced TiO2(110) surface. Spectra are displaced vertically for clarity. The two vertical dashed lines mark the peak desorption temperatures of multilayer and first layer water on clean TiO2(110).

on the mass 46 TPD spectrum of ∼1 ML of d6-acetone from the TiO2(110) surface reduced by vacuum annealing to a level reflective of a 7% oxygen vacancy population (hereafter referred to as the ‘reduced’ surface). Two general effects were observed with increasing water coverage. First, increasing submonolayer amounts of water split the broad d6-acetone first layer desorption feature into two separate features, located above and below the dashed line at 275 K in Figure 2. The minimum between the two peaks extended to near the baseline after about 1 ML of H2O and became better defined at water coverages in excess of 1 ML. The d6-acetone desorption minimum coincided with the peak desorption temperature of first layer water (see inset to Figure 2 for the corresponding water TPD traces). Using the temperature of this minimum as a delineation point, the amount of d6-acetone that evolved in TPD above 275 K decreased from about 0.45 ML in the absence of water to 0.23 ML for water overlayers in excess of 1 ML. The d6-acetone TPD state that evolved at about 200 K with increasing water coverage is similar to the 175 K TPD state for d6-acetone coverages approaching first layer saturation on the clean surface (see bottom trace in Figure 2). This feature has been assigned to desorption of d6-acetone from a compressed d6-acetone first layer,7 and results from intermolecular repulsions between neighboring d6-acetone molecules that decrease the binding energy. The attenuation of the desorption rate (the high-temperature side of the 200 K TPD peak) resulted from desorption of sufficient d6-acetone in order to relieve repulsions in the first layer. This amounted to about 0.2 ML of the saturated d6-acetone first layer. This same effect is likely the source of the

Langmuir, Vol. 21, No. 8, 2005 3445

minimum in the d6-acetone desorption signal in the coadsorption experiment with water. The amount of d6acetone in the first layer after exposure to a multilayer coverage of water is reflected by the peak areas both in the 200 and 340 K features. During TPD, repulsions between water and d6-acetone in the saturated first layer cause some of the d6-acetone to desorb early, with the rate of this desorption attenuating as the onset of first layer water desorption (at about 225 K) is reached. Under UHV conditions, re-adsorption of desorbed water or d6-acetone is not favored. Therefore, the amount of d6-acetone in TPD above 275 K is primarily a function of the relative desorption rates of d6-acetone and water from the saturated first layer, although coverage dependences in these rates complicates the picture. Note from the inset of Figure 2 that the shape of the first layer water desorption feature at 275 K is relatively insensitive to the amount of d6acetone retained on the surface at that point in the TPD. The second layer peak at 180 K is also present at higher water coverages (although broader than on the clean surface, see Figure 1) indicating that water has access to bridging O2- sites. In contrast, the shape of the d6-acetone TPD profile is very sensitive to the amount of water, as discussed above. These behaviors suggest that water compresses d6-acetone into islands but the regions outside these acetone domains resembles water on the clean surface. The second effect of H2O coadsorption with ∼1 ML of d6-acetone on reduced TiO2(110) is that a significant amount of d6-acetone was displaced from the first layer to physisorption sites by even submonolayer coverages of H2O. This displacement is evident in the d6-acetone TPD spectrum by the growth of the desorption signal below 160 K. The amount of displaced d6-acetone increased with H2O coverages up to about 1.5 ML, and the characteristics of the d6-acetone desorption feature became more like those of multilayer desorption (compare with Figure 1). However, the desorption peak of displaced d6-acetone shifted to higher temperature for H2O coverages above 2 ML and evolved nearly coincident with the H2O multilayer desorption state at 160 K (labeled dashed line; see figure inset). On the basis of comparisons of TPD and FTIR spectra from acetone-adsorbed on films of amorphous solid water (ASW) and crystal ice (CI) recorded by Schaff and Roberts,29-31 the coincidence of the two desorption features is associated with d6-acetone molecules hydrogen-bonded to dangling O-H bonds at the ASW surface. In contrast, these authors have shown that the TPD spectrum of acetone adsorbed on CI resembles that of pure multilayer desorption (i.e., the coincident desorption with water is largely absent). This is because the majority of the dangling O-H bonds in the ASW surface reorient to form intermolecular hydrogen bonds with neighboring water molecules during the AWS-to-CI phase transition. These conclusions about acetone on water ice are echoed by other groups.32-35 In this study, water was adsorbed at 95 K, at which temperature ASW should form under UHV conditions on most solid surfaces (see ref 36 and references therein). As an additional test of these concepts, the upper trace of Figure 2 shows the result of switching the adsorption sequence (∼1 ML of d6-acetone adsorbed on (29) Schaff, J. E.; Roberts, J. T. Langmuir 1998, 14, 1478. (30) Schaff, J. E.; Roberts, J. T. J. Phys. Chem. 1996, 100, 14151. (31) Schaff, J. E.; Roberts, J. T. J. Phys. Chem. 1994, 98, 6900. (32) Mitlin, S.; Leung, K. T. Surf. Sci. 2002, 505, L227. (33) Picaud, S.; Hoang, P. N. M. J. Chem. Phys. 2000, 112, 9898. (34) Picaud, S.; Toubin, C.; Girardet, C. Surf. Sci. 2000, 454, 178. (35) Marinelli, F.; Allouche, A. Chem. Phys. 2001, 272, 137. (36) Dohna´lek, Z.; Kimmel, G. A.; Ciolli, R. L.; Stevenson, K. P.; Smith, R. S.; Kay, B. D. J. Chem. Phys. 2000, 112, 5932.

3446


3.5 ML of water). Although a significant amount of d6acetone found its way through the thin ice layer during the course of the TPD ramp, the spectrum in the 140-180 K range resembled those observed by Schaff and Roberts29-31 for acetone on much thicker ASW films. The difference between the 3.5 ML of H2O on 1 of ML d6-acetone and the 1 ML of d6-acetone on 3.5 ML of H2O cases is mainly in the temperature range between 140 and 180 K. (The two d6-acetone TPD spectra were nearly identical above 180 K.) While the difference below 180 K may result from different degrees of d6-acetone incorporation in water ice depending on which was on top, the similarity above 180 K is hard to rationalize apart from considering that a thermodynamically accessible partitioning of the two molecules exists in the first layer on TiO2(110). This partitioning was reached prior to the onset of either molecule’s multilayer desorption state. The possibility that d6-acetone might mix with water ice must also be considered. However, this does not appear to be a significant effect on the basis of previous UHV studies of acetone exposed to water ice surfaces.30,37 An interesting analogue exists between how acetone adsorbs on CI (where the coverage of dangling O-H bonds is minimized) and the d6-acetone TPD spectra of Figure 2. In the water coverage range below about 2 ML, displaced d6-acetone desorbed at temperatures below 140 K, consistent with typical multilayer desorption. However, displaced d6-acetone from water coverages in excess of 2 ML desorbed at slightly higher temperatures (150-160 K), consistent with these molecules being hydrogen-bonded to dangling O-H bonds on the ice surface. The water coverage regime between 1 and 2 ML coincides with development of the 180 K water desorption feature (see Figure 1) attributed to water molecules hydrogen-bonded to bridging O2- sites on the TiO2(110) surface.23,25 Absence of (CD3)2CdO‚‚‚H-OH interactions in the water coverage regime between 1 and 2 ML suggests that the majority of H atoms in second layer water are unavailable to interact with coadsorbed d6-acetone molecules, presumably because they instead are involved in hydrogen-bonding with the surface’s bridging O2- sites. This suggests that second layer water molecules lie flat on TiO2(110)23 possibly with the O end of each molecule tilted away from the surface. Figure 3 further illustrates the partitioning of d6-acetone and water on the reduced TiO2(110) using mass 46 TPD spectra from 3 ML H2O dosed on various precoverages of d6-acetone. The three-peak d6-acetone TPD spectrum (at 160, 200, and 340 K) discussed above was observed for all combinations of H2O and d6-acetone explored, even at d6acetone coverages below 0.1 ML. The inset to Figure 3 shows the fraction of d6-acetone that desorbed above 175 (circles) and 275 K (crosses). As discussed above, desorption of d6-acetone above 175 K is associated with molecules coordinated to Ti4+ sites, whereas the amount desorbing above 275 K reflects those that were retained on the surface after the repulsions of the packed d6-acetone + H2O first layer were relieved by the 200 K desorption of d6-acetone. Actual coverages can be had by multiplying the factions (y axis) by the total coverage (x axis), but for purposes of this discussion, fractions better illustrate the partitioning between d6-acetone and water. The fraction of d6-acetone desorbing from the surface above 175 K was relatively constant (at about 0.8) for d6-acetone coverages below about 0.3 ML despite the considerable amount of excess water present. Above this coverage, the fraction desorbing above 175 K decreased with increasing precoverage of (37) Hudson, P. K.; Zondlo, M. A.; Tolbert, M. A. J. Phys. Chem. A 2002, 106, 2882.

Henderson

Figure 3. Mass 46 TPD spectra from 3 ML of H2O adsorbed on various d6-acetone coverages (0.05, 0.09, 0.18, 0.28, 0.33, 0.60, 0.87, and 1.74 ML) on the reduced TiO2(110) surface. Spectra are displaced vertically for clarity. The two vertical dashed lines mark the peak desorption temperatures of multilayer and first layer water on clean TiO2(110). The inset shows mass 46 TPD peak area contributions from above 175 (circles) and 275 K (crosses) from the spectra shown.

d6-acetone but remained above about 0.3. (Note that the coverage of d6-acetone desorbing at 200 K was difficult to gauge for the highest d6-acetone precoverage used (∼1.7 ML) due to a significant pumping tail from the large multilayer peak.) The fraction of d6-acetone desorbing above 275 K mirrored that desorbing above 175 K. In both cases, there was a relatively sharp drop for coverages in excess of 0.3 ML d6-acetone with a postcoverage of 3 ML H2O. Examination of the low-temperature region of the spectra in Figure 3 indicates an enhancement in the amount of displaced d6-acetone that desorbed from a multilayer for d6-acetone precoverages of about 0.3 ML. Contribution from the dangling OH bond component was still evident in the d6-acetone spectra for the higher precoverages, but the growth was clearly at a lower temperature than that attributable to desorption from dangling OH bonds. The d6-acetone coverage in the 160 K TPD state saturated at about 0.1 ML. Assuming a 1:1 ratio between the d6-acetone coverage in this state and the dangling OH bond coverage leads to a similar coverage of the latter on the water ice surface grown in this manner. The influence of water on coadsorbed d6-acetone is perhaps more easily visualized with lower coverages of d6-acetone. Figure 4 shows d6-acetone TPD spectra (mass 46) from various H2O coverages adsorbed on 0.21 ML of d6-acetone on reduced TiO2(110) surface. Using this starting d6-acetone coverage, approximately 0.8 ML of water should be able to be added to the surface without one (or both) of the molecular species being displaced from the first layer. The bottom trace in Figure 4 reflects the amount of water adsorbed from background prior to adsorption of 0.21 ML of d6-acetone. A small portion of


Figure 4. Mass 46 TPD spectra from various H2O coverages (0.05, 0.18, 0.36, 0.60, 0.95, and 2.5 ML) adsorbed on 0.21 ML of d6-acetone on the reduced TiO2(110) surface. Spectra are displaced vertically for clarity. The two vertical dashed lines mark the peak desorption temperatures of multilayer and first layer water on clean TiO2(110).

the d6-acetone TPD feature shifted to about 250 K after addition of 0.18 ML of water. Further increases in the water postcoverage caused both the amount of destabilized d6-acetone and the degree of destabilization (reflected by the d6-acetone desorption onset) to increase. A 0.95 ML postcoverage of water resulted in the total d6-acetone + water coverage exceeding 1 ML, but no d6-acetone desorption was registered below 160 K. (As discussed above, the sharp feature at 200 K resulted from d6-acetone desorption in the highly compressed first layer and not from physisorbed acetone.) H2O TPD spectra (not shown) indicated that at this combination of d6-acetone and H2O (0.21 and 0.95 ML, respectively) a small amount of water desorbed at 180 K, the temperature corresponding to second layer water.23,25 With a much larger water coverage (2.5 ML), TPD shows that a considerable amount of d6acetone was displaced from the surface into the 160 K state. (Due to a dosing error, the amount of d6-acetone adsorbed in this case was 0.28 ML instead of 0.21 ML.) 3.2. Site Competition on the Oxidized TiO2(110) Surface. Results discussed to this point have dealt with coadsorption of d6-acetone and H2O on the reduced TiO2(110) surface. To consider how surface oxidation influences the interaction of d6-acetone and H2O on TiO2(110) surface, one must first come to an appreciation of the complexity with exposure of O2 to TiO2(110) surfaces possessing oxygen vacancies. Prior to 1998, researchers generally assumed that oxygen exposure to a TiO2(110) surface resulted in stoichiometric oxidation of vacancy sites (i.e., one O2 molecule filling two oxygen vacancies). Epling and co-workers26 were the first to show that this simple model for vacancy oxidation was not the case but that vacancies were filled in a 1:1 ratio with O2 resulting in oxygen

Langmuir, Vol. 21, No. 8, 2005 3447

Figure 5. Mass 46 TPD spectra from various H2O coverages (0.06, 0.28, 0.50, and 1.2 ML) adsorbed on 0.8 ML of d6-acetone on the oxidized TiO2(110) surface. Spectra are displaced vertically for clarity. The two vertical dashed lines mark the peak desorption temperatures of multilayer and first layer water on clean TiO2(110).

adatoms being left on the surface. Photodesorption,38-41 thermal desorption,27,41,42 and theoretical43-47 studies have shown that molecular O2 species also exist on the surface after low-temperature adsorption of O2 at oxygen vacancies on TiO2(110). More recently, STM results have shown that oxygen adatoms and molecules readily diffuse at low temperatures filling and even recreating vacancies.28,48 As it turns out, these oxygen species are key to stabilizing acetone on TiO2(110)7 and diminishing the influence of coadsorbed water displacement of acetone from the surface. Results in Figures 5 and 6 show that surface preoxidation has a significant influence on the competition between d6-acetone and water for sites on TiO2(110). Data in these figures complement those shown in Figures 2 and 4 for the reduced surface in terms of the starting (38) Rusu, C. N.; Yates, J. T., Jr. Langmuir 1997, 13, 4311. (39) Lu, G.; Linsebigler, A.; Yates, J. T., Jr. J. Chem. Phys. 1995, 102, 3005. (40) Lu, G.; Linsebigler, A.; Yates, J. T., Jr. J. Chem. Phys. 1995, 102, 4657. (41) Perkins, C. L.; Henderson, M. A. J. Phys. Chem. B 2001, 105, 3856. (42) Henderson, M. A.; Epling, W. S.; Peden, C. H. F.; Perkins, C. L. J. Phys. Chem. B 2003, 107, 534. (43) de Lara-Castells, M. P.; Krause, J. L. J. Chem. Phys. 2001, 115, 4798. (44) de Lara-Castells, M. P.; Krause, J. L. Chem. Phys. Lett. 2002, 354, 483. (45) Shu, C.; Sukumar, N.; Ursenbach, C. P. J. Chem. Phys. 1999, 110, 10539. (46) Rasmussen, M. D.; Molina, L. M.; Hammer, B. J. Chem. Phys. 2004, 120, 988. (47) Wu, X.; Selloni, A.; Lazzeri, M.; Nayak, S. K. Phys. Rev. B 2003, 68, 241402. (48) Wahlstroem, E.; Vestergaard, E. K.; Schaub, R.; Ronnau, A.; Vestergaard, M.; Laegsgaard, E.; Stensgaard, I.; Besenbacher, F. Science 2004, 303, 511.

3448


Figure 6. Mass 46 TPD spectra from various H2O coverages (0.03, 0.25, 0.47, 0.84, 1.2, 3.0, and 4.0 ML) adsorbed on 0.18 ML of d6-acetone on the oxidized TiO2(110) surface. Spectra are displaced vertically for clarity. The two vertical dashed lines mark the peak desorption temperatures of multilayer and first layer water on clean TiO2(110).

d6-acetone coverages. In the absence of water, preoxidation of the surface stabilizes about 0.25 ML of d6acetone to 375 K, which is about 40 K higher than observed on the reduced surface (compare the bottom d6-acetone desorption traces in Figures 5 and 6 with those in Figures 2 and 4). Even in the absence of coadsorbed water, TPD reveals a notable difference in how d6-acetone desorbs from the oxidized and reduced surfaces of TiO2(110). Whereas TPD spectra of d6-acetone coverages between 0 and 0.25 ML on the reduced surface show evidence of intermolecular repulsions, the same coverage range on the oxidized surface shows typical first-order desorption behavior without coverage dependence.7 The origin of this difference is linked to a reaction between η1-acetone species and oxygen to form an acetone-oxygen complex.7 Figure 5 shows d6-acetone TPD spectra (mass 46) from various postcoverages of water to ∼0.8 ML of d6-acetone adsorbed on oxidized TiO2(110). Although a similar series of changes to the d6-acetone TPD spectrum occurred on both the oxidized and reduced surfaces (e.g., development of a minimum in the desorption rate coincident with the first layer water state, displacement of first layer d6acetone to the multilayer, and development of a threepeak d6-acetone TPD spectrum), the amount of d6-acetone retained to higher temperature (i.e., above the first layer water TPD feature) was consistently greater for any given water coverage on the oxidized surface (Figure 5) compared to the reduced surface (Figure 2). For example, after a ∼1.2 ML of H2O coverage, the amount of d6-acetone that desorbed above 275 K (the first layer water desorption point) was nearly 50% greater for the oxidized surface (∼0.37 ML) than for the reduced surface (∼0.25 ML). The effect was more pronounced with lower initial d6-acetone

Henderson

Figure 7. Mass 46 TPD peak area data taken from Figures 3 and 6 showing the fraction of d6-acetone desorbing above 275 K as a function of water coverage for the oxidized (circles) and reduced (triangles) TiO2(110) surfaces. Dashed lines are from exponential fits to the data and are drawn to guide the reader.

coverages. Figure 6 shows d6-acetone TPD spectra with increasing postcoverages of water on 0.18 ML d6-acetone adsorbed on oxidized TiO2(110). Only a small amount (∼0.01 ML) of d6-acetone was destabilized in TPD by water coverages up to 0.5 ML, and the affected molecules desorbed coincident with the first layer water state at 275 K. Higher water coverages up to about 1.2 ML resulted in only slightly more destabilization of d6-acetone, with the amount of affected d6-acetone not exceeding 0.03 ML and the affected molecules still desorbing from the first layer (i.e., above 170 K). Displacement of d6-acetone to the multilayer (actually to dangling OH sites on the water ice surface) was observed with water coverages above 3 ML, but the amount of displacement was very small (below 0.01 ML). On the basis of TPD results in Figures 5 and 6, it appears that pre-oxidation of the TiO2(110) surface significantly limits the destabilizing influence of coadsorbed water on d6-acetone. Figure 7 illustrates this by plotting the fraction of d6-acetone (relative to the case with no water) retained in TPD above 275 K as a function of the amount of water coadsorbed with ∼0.2 ML d6-acetone on the oxidized (circles) and reduced (triangles) TiO2(110) surfaces. (Data for this plot were taken from Figures 3 and 6.) The ability of water to destabilize d6-acetone on the reduced surface is significant. An exponential fit to the data suggests the full impact of water in destabilizing d6-acetone is not reached after 2.5 ML of water where over 70% of the original d6-acetone coverage was shifted to lower temperature desorption states. In contrast, about 20% of the initial ∼0.2 ML of d6-acetone coverage on the oxidized surface was destabilized by water coverages up to 4 ML and the fit to the data suggests that additional destabilization is unlikely under UHV conditions. The question


Langmuir, Vol. 21, No. 8, 2005 3449

Figure 8. Schematic models depicting the effect of coadsorbed water on acetone adsorbed on the reduced (left) and oxidized (right) surfaces of TiO2(110).

now arises as to how pre-oxidation of the surface inhibits the destabilizing influence of water on coadsorbed d6acetone. 3.3. Models for Site Competition and Potential Impact on Acetone Photo-oxidation. Several groups have proposed that water inhibits photo-oxidation of acetone on TiO2 surfaces by blocking adsorption sites and displacing adsorbed acetone molecules.15-21 Additionally, Vorontsov and co-workers19 have shown that much of the inhibiting influence of water on acetone photo-oxidation over TiO2 was alleviated by increasing the catalyst temperature from 40 to >80 °C, possibly due to reducing the coverage of water on the catalyst surface. Water also appears to be important in promoting acetone photooxidation. Several groups have observed that small amounts of water enhance acetone photo-oxidation rates, presumably by supplying OH groups for the production of OH radicals that participate in the reaction.6,16,18,22 Results in this study show that coadsorbed water can have a significant impact on the surface coverage of acetone depending on the redox state of the TiO2 surface. On the reduced surface, the key issue appears to be the greater sensitivity to overall coverage for acetone desorption than for water desorption. Previous work on the clean TiO2(110) surface,25 as well as on other TiO2 surfaces (see references in ref 1), suggests only mild coverage dependence in the desorption characteristics of water. Although

gas-phase water has a significant dipole moment (1.85 D), the ability of water to adapt its adsorption structure, which in part includes hydrogen-bonding to neighbors and to the oxide surface, seems to mitigate the influence of dipole-dipole repulsions between neighboring water molecules that might otherwise lower the molecule’s desorption energy. In contrast, the acetone molecule adsorbs on TiO2(110) in a more restricted geometry (via an η1-configuration involving coordination through lone pair electrons on the oxygen atom) which more-or-less aligns the molecular dipole in opposition with its neighbors. Because of this and the magnitude of acetone gasphase dipole moment (2.88 D), a more pronounced influence of coverage is observed in the desorption properties of acetone on TiO2(110).7 The onset of acetone desorption is at about 170 K from a saturated first layer on TiO2(110), whereas the same for water is at about 210 K. Add to this the fact that the molecular dipole orientations of adsorbed water and acetone are the same (oxygen ends down), one expects a situation in which high coverages of water should destabilize acetone to a greater extent than the other way around. TPD data in Figures 2 and 3 show this to be the case. High water coverages displace most of the preadsorbed acetone into multilayer states, whereas water desorbs more-or-less unperturbed from what would be observed on the clean surface. Figure 8 (left side) illustrates the proposed model for this effect

3450


on reduced TiO2(110) surface. Water exposure to acetone on TiO2(110) causes acetone molecules to compress into high coverage islands and, depending on the amounts of both, may displace acetone from the first layer (not illustrated). Intermolecular repulsions within the acetone islands appear to be greater than those between water molecules because desorption of acetone is observed first in TPD (at about 200 K). By 200 K, all multilayer and second layer water has also desorbed in the UHV experiment (see Figure 1) so the tension between the water and acetone domains is not intensified by the addition of more water. In contrast, a sufficiently high water flux under non-UHV conditions should further deplete the acetone coverage by expanding the water domains against the acetone islands. However, in the UHV TPD experiment, the next desorption event is water at about 275 K, which then permits the remaining acetone to relax from the compressed island condition. In contrast, the oxidized TiO2(110) surface exhibits a greater resistance against acetone displacement by water. The explanation for this is the role of reactive oxygen species in stabilizing acetone on the surface in the form of an acetone-oxygen complex. Although the exact structure of this complex is not known, its C-O stretching frequency in HREELS is consistent with a bond midway between a single and double bond.7 A potential candidate with such a C-O bond stretch is ‘poly-acetone’,7 as illustrated in the top right drawing of Figure 8. Despite the assignment of this species being unresolved, its impact on acetone-water interactions is clear. Pre-oxidation of oxygen vacancies on the TiO2(110) surface results in an adsorbed form of acetone that is considerably less sensitive to displacement induced by coadsorbed water. Acetone molecules that are not stabilized by pre-oxidation exhibit similar displacement behavior as that observed on the reduced surface. Additionally, the oxygen-perturbed species has been shown to be considerably more photochemically reactive on TiO2(110) than the η1-acetone species.9 The interactions described above between acetone and water may also take place on other TiO2 surfaces under non-UHV conditions if similar relative binding energies and coverage dependences exist. For example, if conditions of temperature and relative humidity permit a saturation coverage of first layer water, then the residence times of adsorbed acetone molecules will be short due to intermolecular repulsions that favor desorption over retention.

Henderson

If relative humidity is initially low but increases due to water production, then initial conditions of high acetone coverage should dwindle to low coverages as the photooxidation reaction continues. Of course, these scenarios will also depend on the redox state of the surface. If the O2 partial pressure is low and surface electron trapping is inhibited (conditions that should limit O2 adsorption), then water should exhibit a greater tendency to displace acetone than if the O2 partial pressure is high and electron scavenging is fast, in which case the impact of water in displacing acetone should be minimized. Alternatively, surfaces with high concentrations of structural defects such as steps may also stabilize adsorbed acetone against water induced displacement much as pre-oxidation of the TiO2(110) surface does. 4. Conclusions Results in this study show that acetone and water compete for surface sites on the TiO2(110) surface. On vacuum-reduced surfaces that possess oxygen vacancy sites, η1-acetone (the dominant form of adsorbed acetone) is compressed into high coverage islands by coadsorbed water and is eventually displaced from first layer sites to the multilayer as the water coverage is increased. Preoxidization of oxygen vacancies on the surface results in reactive forms of adsorbed oxygen that stabilize acetone on the surface. Unreacted η1-acetone molecules are still compressed and displaced from the oxidized surface by water. These results suggest that under humid conditions acetone likely exists on TiO2 surfaces in high coverage islands compressed by larger domains of adsorbed water and hydroxyl, and that reactive oxygen species (such as those formed during UV photolysis) may stabilize acetone against the negative influence of water. Acknowledgment. This work was funded by the Office of Basic Energy Sciences, Divisions of Chemical and Materials Sciences. Pacific Northwest National Laboratory is a multiprogram national laboratory operated for the U.S. Department of Energy by the Battelle Memorial Institute under Contract No. DE-AC06-76RLO 1830. The research reported here was performed in the William R. Wiley Environmental Molecular Science Laboratory, a Department of Energy user facility funded by the Office of Biological and Environmental Research. LA0476579