This journal is c the Owner Societies 2013 Phys. Chem. Chem. Phys., 2013, 15, 13897--13901 13897

Cite this: Phys. Chem.Chem.Phys.,2013,15, 13897

Imaging reactions of acetone with oxygen adatoms onpartially oxidized TiO2(110)

Yaobiao Xia,a Jingyun Ye,b Patrick Murray,c Amir Ali,a Qingfeng Geb andZhenrong Zhang*a

Understanding the interaction of O2 with ketones on metal oxide surfaces is important for the photo-

oxidation of toxic organic molecules. The consecutive reaction steps of acetone molecules with oxygen

adatoms (Oa’s) on partially oxidized TiO2(110) surfaces have been studied using high-resolution

scanning tunneling microscopy (STM) at 300 K. The sequential isothermal STM images reveal two types

of acetone–Oa species as a result of reactions of acetone with an oxygen adatom and a bridging bound

oxygen vacancy (VO). One such species is the Ti5c-bound acetone–Oa diolate formed from Ti5c-bound

acetone reacting with Oa. The diolate is mobile at 300 K and can assist the diffusion of surface Oa by

exchanging the acetone oxygen with the Oa. The second acetone–Oa species is the VO-bound acetone–Oa

complex formed from a VO-bound acetone reacting with an Oa located on the neighboring Ti row. The

VO-bound complex is stationary at 300 K. This species has not been reported previously.

Introduction

Indoor air pollutants including oxygenated organic compoundssuch as ketones, aldehydes and alcohols present an increasingenvironmental concern.1–6 Heterogeneous photocatalysis usingTiO2 is a promising strategy for remediating volatile organiccompounds. Thus, the chemistry and photochemistry of organiccompounds containing CQO groups have been widely studiedon various forms of TiO2.4–7

It is clear that the presence of surface oxygen is crucial incontrolling the reaction pathways in both thermal and photo-induced reactions taking place at the TiO2–gas phase inter-face.2,8–10 As a result, O2 on the TiO2(110) single crystal surfacehas been extensively studied to understand how molecularoxygen binds to titania surfaces and forms the various activespecies.11–20 Recently, the interaction of O2 with organics hasbeen studied on TiO2(110) surfaces to understand the role ofoxygen species in the reactions.21–24 Oxygen is often used as ascavenger of photo-excited electrons to prevent negative chargeaccumulation. Oxygen species can directly participate in theoxidation chemistry. For example oxygen adatoms thermallydissociate methanol to a photoactive methoxy species.25,26

An extensive series of studies by Henderson has shown thatthrough a thermal process, adsorbed oxygen converts the photo-inactive form of carbonyl molecules (e.g. acetone, acetaldehyde,butanone, and acetophenone) into photoactive carbonyl molecule–oxygen complex species on TiO2(110).7,27–35 Vibrational spectro-scopy, isotopic labeling studies and DFT calculations identified theconfiguration of these photoactive intermediates as Z2-diolatespecies. However, the atomic-level details of how the O2 activespecies interact with organic species are far from fully under-stood. Scanning tunneling microscopy (STM) has been veryhelpful in unraveling the atomic-level details of the interactionbetween O2 and defects on TiO2.21–24 STM studies of butyro-phonone36 and acetone22 on oxidized TiO2(110) show theformation of the ketone–O complex. However, the exactconfiguration of the complex is still not clear.

In our previous study, we reported the dynamic relationshipbetween acetone and surface bridging bonded oxygen vacancy(VO) defect sites on reduced TiO2(110) surfaces.37 In this study,we focus on the interaction of acetone with adsorbed oxygen onpartially oxidized TiO2(110) surfaces. High-resolution STM followedthe consecutive reaction steps of acetone with oxygen adatoms (Oa)and VO’s. We report here the direct imaging of two key acetone–oxygen species. One species is the Ti5c-bound acetone–Oa diolatewhich is reported to be critical in photoreaction and the otheris the VO-bound acetone–Oa complex which has not beenreported in the literature. By tracking species resulting from asequence of reactions, we show molecular-level details of theunderlying reaction mechanism.

a Department of Physics, Baylor University, One Bear Place #97316, Waco,

TX 76798, USA. E-mail: zhenrong_zhang@baylor.edub Department of Chemistry and Biochemistry, Southern Illinois University,

Carbondale, IL 62901, USAc Department of Physics, Arizona State University, Tempe, AZ 85287, USA

Received 19th April 2013,Accepted 17th June 2013

DOI: 10.1039/c3cp51695d

www.rsc.org/pccp

PCCP

PAPER

Publ

ishe

d on

18

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f Il

linoi

s at

Chi

cago

on

28/1

0/20

14 1

8:18

:48.

View Article OnlineView Journal | View Issue

13898 Phys. Chem. Chem. Phys., 2013, 15, 13897--13901 This journal is c the Owner Societies 2013

Experimental section

Experiments were performed in an ultrahigh vacuum chamber(base pressure o3 � 10�11 Torr) equipped with variable tempera-ture STM (SPECS), a quadrupole mass spectrometer (SRS) and anion gun (SPECS). The single crystal TiO2(110) surface (PrincetonScientific) was prepared by multiple cycles of sputtering (1–3 keV)and e-beam heating annealing (B900 K). The cleanness of thereduced surface is checked by STM before the surface is exposedto oxygen. Oxygen was dosed by backfilling the chamber. The STMtip was withdrawn about 1 mm from the surface during O2

exposure. The vacancy concentration and species coverage wereobtained by a direct counting from STM images and expressed inmonolayer (ML) units (1 ML corresponds to 5.2 � 1014 cm�2 Tiatoms). On prepared partially oxidized TiO2(110) surfaces, lessthan 0.005 ML of hydroxyl (OHb) is observed indicating that thebackground H2O adsorption is negligible. Acetone (EMD, 99.5+%)was cleaned by several freeze–pump–thaw cycles using liquidnitrogen prior to use and was introduced onto the partiallyoxidized TiO2(110) surfaces at the STM stage via a retractabletube doser containing a pinhole. All STM images (empty states)were collected at 300 K in a constant-current (o0.1 nA) mode atpositive sample bias voltages of 0.8–1.3 V. Images were processedusing WSxM software (Nanotech, freeware).38

Results

To probe the reaction of acetone with an oxygen adatom, weprepared the partially oxidized TiO2(110) surface where only afraction of the VO’s is converted to Oa. Fig. 1 shows snapshotsfrom a representative STM movie taken after acetone adsorbson the partially oxidized TiO2(110) surface with 0.07 ML of Oa

and 0.06 ML of VO at 300 K. In the empty state STM images ofTiO2(110), the low lying five coordinated Ti5c ions are imaged asprotrusions while the rows of bridging bonded oxygen (Ob) ions

are imaged as depressions. The VO site appears as the bridgingspot (highlighted by small white arrows in Fig. 1a) on the dark Ob

row. When O2 adsorbs on reduced TiO2(110) at 300 K, oxygendissociates on the surface via two dissociation channels formingeither oxygen adatom (Oa) pairs or single Oa’s.12,39,40 The Oa’sshow up as bright dots (blue arrows) on the Ti5c rows which aresurrounded by slightly darker regions. The distance between thetwo Oa’s of the Oa pair is two lattice constants at 300 K.12,18,19

After acetone adsorption at 300 K (Fig. 1a), two types ofbright acetone features are observed on Ob rows. One type(small circles) is VO-bound acetone which is symmetric with respectto the row. The assignment of the VO-bound acetone is based onthe adsorption sites and their mobility on clean TiO2(110).37 TheVO-bound acetone readily diffuses on the surface at 300 K via twodiffusion channels: one is diffusion along the Ob row and movingas an alkyl group, which heals the initial VO; another is diffusionfrom Ob row to the five coordinated Ti5c row and then movingalong the Ti5c row as an acetone, which leaves a VO behind. Thecombination of the two acetone diffusion channels mediates thediffusion of VO.37 Due to the small diffusion barrier (B0.25 eV)along the Ti row, the Ti5c-bound acetone molecules (Z1-acetone)diffuse faster than the STM scan speed (B80 nm s�1) and show upas occasional streaks on the STM images.

The other type (small square) is not symmetric and has afaint lobe protruding to the Ti row. In Fig. 1a, the faint lobeprotrudes from the upper-left side of the feature (small square) tothe Ti row. This type of bright feature is assigned as the VO-boundacetone–Oa complex. We call it the VO-bound complex from hereon. To explore the formation of the VO-bound complex, weacquired time lapsed images over extended periods of time(53 min) on the same area after acetone exposure. Fig. 1 showsthe formation of two VO-bound complexes. From Fig. 1a to b,the upper VO-bound acetone (upper small circle) diffuses to anunoccupied VO site (white arrow on the right, Fig. 1a) which hasa faint single Oa (upper blue arrow) located one lattice spacing

Fig. 1 Snapshots (a–c) from an STM movie taken on the partially oxidized surface after acetone exposure at 300 K. The ball model illustrates the conversion of theVO-bound acetone to the VO-bound acetone–Oa complex. (d) The line scan of the green line in image (c). Symbols in the STM images indicate Oa (blue arrow),VO (white arrow), VO-bound acetone (small circle), and VO-bound acetone–Oa complex (small square) respectively.

Paper PCCP

Publ

ishe

d on

18

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f Il

linoi

s at

Chi

cago

on

28/1

0/20

14 1

8:18

:48.

View Article Online

This journal is c the Owner Societies 2013 Phys. Chem. Chem. Phys., 2013, 15, 13897--13901 13899

downward on the neighboring Ti row. A VO appears where theacetone was originally located indicating that the whole acetonemolecule diffuses via the Ti row to the unoccupied VO as shown inthe ball model.37 The VO-bound acetone combined with nearby Oa

appears as a bright VO-bound complex (Fig. 1b, upper square). Inthe meantime, the bottom VO-bound acetone molecule (lower smallcircle) diffused back and forth several times through both thediffusion channels and ended one lattice constant downward fromFig. 1a to b. Compared to the VO-bound acetone, the VO-boundcomplex is also located on VO, but it has a small faint protrusionextended into the Ti row onto the neighboring Oa. Fig. 1c showsanother example of a VO-bound complex appearing at the VO

marked by the left white arrow in Fig. 1a. This VO has an Oa pairsitting on the neighboring Ti row with one of the Oa’s located onelattice spacing upward to the left of the VO.

To further illustrate the different appearances of the VO-boundacetone and the VO-bound complex, we have also included the linescan (dotted line in Fig. 1c) along the Ob row of such isolatedfeatures at a typical bias of 1.3 V. The apparent height of theVO-bound acetone (B0.18 nm) measured relative to the Ob row isslightly larger than that of the VO-bound complex (B0.15 nm). Forcomparison, the heights of VO and OHb are also examined andincluded on the line scan. The OHb feature is higher than VO asexpected at a typical bias voltage (1.0–1.5 V).39,41,42

The major difference between the VO-bound acetone and theVO-bound complex is their mobility. The VO-bound acetone readilydiffuses at 300 K with an average lifetime of 90 s.37 However, onceformed, the VO-bound complex becomes stationary. No diffusionis observed from three separate isothermal experiments with thetotal imaging time of B150 min. Interestingly, if the Oa is locatedtwo lattice spacings away from the VO-bound acetone as shownbelow, the formation of the VO-bound complex is not observed.

Fig. 2 shows several possible configurations for the observedVO-bound complex. One is that the VO-bound acetone rotates451 to maximize the attractive van-der-Waals (vdW) forcebetween the methyl group and Oa (configuration A). Thisattractive force may increase the diffusion barrier for VO-boundacetone to move from a VO site to the Ti row. On the reducedsurface, this is the rate limiting step (0.74 eV) for the diffusionof acetone via the Ti row since the diffusion barrier for acetonemotion along the Ti row is only 0.25 eV.37 Another possibility isthe formation of Z1-enolate (configuration B). In this case, theVO-bound acetone rotates 451 and transfers a hydrogen atomfrom the methyl group to Oa and forms the Z1-enolate whichbecomes a stable species. This is feasible as Oa has shownhigher reactivity than Ob due to an excess amount of chargeassociated with the Oa.19 On CeO2, stable Z1-enolate has beenobserved on reduced surfaces which involves the reaction ofacetone with VO and lattice oxygen.3 Many groups have observed

H/D exchange between acetone and OH groups on other oxide-based materials.5,43–47 The consensus mechanism involves for-mation of an acetone enolate species in which a hydrogen atomfrom one of the acetone’s methyl groups is transferred to the surfaceresulting in rehybridization of acetone to the enolate. We excludedthe possibility of Z2-bound acetone where the carbonyl carbon atomcoordinates with Oa (configuration C), because the distance betweenthe VO site and the Oa located one lattice spacing away on theneighboring Ti row is B4.2 Å. This distance is too far for the O–C–OZ2-configuration. In addition, Henderson et al. reported the adsorp-tion energy (Eads) of a similar configuration, Z2-bound acetone,where the carbonyl carbon atom coordinates to Oa located directlyacross the VO. The calculated Eads (�45 kJ mol�1) is much higherthan that (�72 kJ mol�1) of the Ti5c-bound acetone.35

Fig. 3 presents two subsequent images illustrating the for-mation of the second type of acetone–Oa complex—the Ti5c-boundacetone–Oa diolate. When an acetone molecule diffuses along a Tirow and passes an Oa, the Oa is observed to shift by one latticeconstant via a diolate intermediate. For example, from Fig. 3a to b,the VO-bound acetone (black dot, Fig. 3a) moves to the Ti rowleaving a vacancy (yellow dot, Fig. 3b) behind suggesting that theacetone moves along the Ti row upward as an acetone molecule.37

The acetone passes a single Oa and an Oa pair (blue dots, Fig. 3b)and stops on a single Oa appearing as a bright feature (red dot,Fig. 3b) which is centered on the Ti row. Compared to Fig. 3a, thepositions of the three Oa’s (blue dots, Fig. 3b) on the path of theacetone diffusion are shifted by one lattice constant downwardwhich is opposite to the diffusion direction (upward) of theacetone molecule. The bright feature located on Oa (red dot,Fig. 3b) is assigned as a Ti5c-bound diolate as proposed before.27

A large set of Oa diffusion events (144) were evaluated from threeseparate experiments. For every observed Oa diffusion event,either the formation of diolate or the diffusion of a VO–acetoneto a new VO was observed. It is worth noting that in Fig. 3a, thelower Oa (green dot, Fig. 3a) is located two lattice distances awayfrom the VO-bound acetone (black dot), thus the formation ofthe VO-bound complex is not observed due to the large distance(B6 Å) from the Oa to VO-bound acetone.

Fig. 2 The possible configurations of the VO-bound acetone–Oa complex.

Fig. 3 Two subsequent STM images (a and b) showing the diffusion of the Oa

and (c) the ball model of acetone assisted Oa diffusion via the formation of theTi5c-bound acetone–Oa diolate. Symbols in the STM images indicate Oa (greendots), shifted Oa (blue dots), VO-bound acetone (black dot), and Ti5c-boundacetone–Oa diolate (red dot) respectively.

PCCP Paper

Publ

ishe

d on

18

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f Il

linoi

s at

Chi

cago

on

28/1

0/20

14 1

8:18

:48.

View Article Online

13900 Phys. Chem. Chem. Phys., 2013, 15, 13897--13901 This journal is c the Owner Societies 2013

Fig. 3c illustrates our interpretation for the Oa diffusionprocess via the diolate intermediate. The VO-bound acetone movesto the Ti5c row and binds on Ti5c sites via the lone-pair electronson the carbonyl oxygen atom becoming a Z1-acetone species.27

The Z1-acetone species diffuses along the Ti5c row rapidly at300 K.37 When Z1-acetone reaches an Oa, the Oa coordinates toelectron-deficient acetone carbon forming Ti5c-bound diolatespecies (Z2-diolate). The diolate can dissociate back to an Oa

and an acetone. The acetone molecule continues to diffuse alongthe Ti row, either forward as a new acetone with the oxygen thatused to be the Oa, leaving behind its oxygen atom as a new Oa, orbackward in the form of original acetone. The former case resultsin the apparent motion of the Oa and the oxygen scrambling whilethe latter case leaves the Oa in its original position. Over 20isolated diolates were evaluated by overlaying the lattice grid onthe images to determine the position of diolate relative to Oa.Instead of centered on top of Oa, the diolate is centered betweenthe two Oa’s as expected. This Oa diffusion mechanism and theformation of the diolate are consistent with the mechanismproposed by Henderson and coworkers on the basis of isotopelabeling studies.27 A similar Oa diffusion mechanism has beenobserved when H2O interacts with Oa.48

A diolate can be observed on a single Oa as shown in Fig. 3and on an Oa pair as shown in Fig. 4. Unlike the diolate on asingle Oa, a majority (70%) of diolates formed on Oa pairsappear fuzzy (Fig. 4b). Here, an acetone moves downwardpassing an Oa (upper green dot, Fig. 4a). The Oa is shifted upby one lattice distance and acetone stops on an Oa pair (lowerpair of green dots, Fig. 4a). The distance between two Oa’s inthe pair is generally two lattice spacings apart.12,18,19 Theimages overlaid by the lattice grid (Fig. 4) show that the fuzzybright feature extends over four lattice positions. This indicatesthat the upper Oa of the Oa pair is shifted up one lattice spacingby exchanging oxygen with acetone when acetone moves down-ward and forms diolate with the lower Oa of the Oa pair. Theacetone molecule can diffuse back and forth between the twoOa’s and form diolates with either of the Oa’s which are nowthree lattice spacings apart. Therefore the fuzzy feature in ourexperiments represents the time average of acetone movingbetween the two Oa’s. Later, like the diolate on a single Oa, the

diolate on an Oa pair dissociates back to an Oa pair and anacetone. The acetone from the fuzzy diolate is observed to moveon and adsorb on a nearby VO (not shown). This indicates thatthe fuzzy diolate is not due to two acetone molecules adsorbingsimultaneously on an Oa pair. From two sets of experimentswith comparable concentrations of Oa pairs and single Oa, thetotal number of diolate species observed on Oa pairs is about4.5 (168/36) times of that of the diolate on the single Oa’s. Thisvalue is higher than the expected value of 2.0 from the simplestatistical probability of acetone encountering twice as manyOa’s from pairs as isolated Oa’s. This indicates that withanother Oa nearby, the lifetime of the diolate slightly increases.This might relate to the different charge distributions near thesingle Oa and the Oa pair since they are formed from twodifferent dissociation channels.

It should be noted that although the involvement of theacetone is observed, the diolate intermediates including bothon single Oa’s and Oa pairs are only directly observed in B40%(59/144) of the total number of Oa diffusion events. Thissuggests that the average lifetime of diolate is slightly shorterthan the data acquisition time (66 s per frame). Assuming thatthe diffusion process is Arrhenius-like, the hopping rate n isgiven as n = n0 exp(�DEB/kBT) where n0 is the attempt frequency,DEB is the activation energy barrier, kB is the Boltzmannconstant, and T is the temperature. The estimated lifetime(B1 min) corresponds to an energy barrier of 0.85–0.90 eV forthe diolate to dissociate back to Oa and Ti5c-bound acetone.A typical attempt frequency of B1013 s�1 is used in the estimate.This barrier is comparable to the two VO–acetone diffusionbarriers (0.74 eV and 0.82 eV) for the two diffusion channelsrespectively.37 Thus, the diffusion of VO–acetone and the conver-sion of the diolate to Oa and Ti5c-bound acetone are all observed onthe same surface at 300 K. At low coverages (below VO coverage) ofacetone and when the coverages of the VO and Oa are comparable,the populations of the diolate, VO-bound acetone and Ti5c-boundacetone reflect their relative adsorption energies (Eads). From ourexperiments, the coverage ratio of the diolate to VO-bound acetoneis B0.10 and no Ti5c-bound acetone is resolved. This indicatesEads’s of the diolate, VO-bound acetone and Ti5c-bound acetonefollow ETi-bound acetone

ads > Ediolateads > EVo-bound acetone

ads . It is reportedthat the calculated Eads of the diolate (�1.57 eV) is lower than thatof the Ti-bound acetone (�0.75 eV)35 which agrees with ourresults. However the energy difference between these two speciesindicates that the diolate is more stable than the VO-bound acetone(�1.24 eV)37 which is in contrast to our experiment results. Notethat these three calculated Eads’s are relative to different surfaces(the stoichiometric surface with an Oa for the diolate, the stoichio-metric surface for the Ti-bound acetone, and the reduced surfacefor VO-bound acetone) which could be the reason for the discrepancybetween the experiment and the calculation.

Conclusion

The consecutive reaction steps of acetone with Oa and VO onpartially oxidized TiO2(110) surfaces are imaged using STM at300 K. Acetone reacts with the surface to form two types of

Fig. 4 Two subsequent STM images overlaid with the lattice grid present theformation of a Ti5c-bound acetone–Oa diolate (the fuzzy bright feature) on an Oa

pair. Symbols in the STM images indicate Oa (green dots) and shifted Oa (bluedots) respectively.

Paper PCCP

Publ

ishe

d on

18

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f Il

linoi

s at

Chi

cago

on

28/1

0/20

14 1

8:18

:48.

View Article Online

This journal is c the Owner Societies 2013 Phys. Chem. Chem. Phys., 2013, 15, 13897--13901 13901

acetone–Oa complexes. A VO-bound acetone combines with anearby Oa appearing as a bright VO-bound acetone–Oa

complex. Once formed, the VO-bound acetone–Oa complexbecomes stationary. A Ti5c-bound acetone combines with an Oa

as a Ti5c-bound diolate which assists the diffusion of the surface Oa.The diolate can be observed both on a single Oa and on an Oa pair.The lifetimes of the Ti5c-bound acetone–Oa diolate, the VO-boundacetone and the Ti5c-bound acetone indicate that their adsorptionenergies follow ETi-bound acetone

ads > Ediolateads > EVo-bound acetone

ads .

Acknowledgements

Acknowledgement is made to The Donors of the AmericanChemical Society Petroleum Research Fund for the support ofthis research.

References

1 C. M. Schmidt, A. M. Buchbinder, E. Weitz and F. M. Geiger,J. Phys. Chem. A, 2007, 111, 13023.

2 E. Carter, A. F. Carley and D. M. Murphy, ChemPhysChem,2007, 8, 113.

3 S. D. Senanayake, W. O. Gordon, S. H. Overbury andD. R. Mullins, J. Phys. Chem. C, 2009, 113, 6208.

4 A. Mattsson and L. Osterlund, J. Phys. Chem. C, 2010,114, 14121.

5 M. El-Maazawi, A. Finken, A. Nair and V. Grassian, J. Catal.,2000, 191, 138.

6 M. D. Hernandez-Alonso, I. Tejedor-Tejedor, J. M. Coronado,M. A. Anderson and J. Soria, Catal. Today, 2009, 143, 364.

7 M. A. Henderson, J. Phys. Chem. B, 2005, 109, 12062.8 T. Tatsuma, S. Tachibana and A. Fujishima, J. Phys. Chem. B,

2001, 105, 6987.9 R. Nakamura and S. Sato, J. Phys. Chem. B, 2002, 106, 5893.

10 S. H. Szczepankiewicz, A. J. Colussi and M. R. Hoffmann,J. Phys. Chem. B, 2000, 104, 9842.

11 J. T. Yates, Surf. Sci., 2009, 603, 1605.12 S. Wendt, P. T. Sprunger, E. Lira, G. K. H. Madsen, Z. Li,

J. O. Hansen, J. Matthiesen, A. Blekinge-Rasmussen,E. Laegsgaard, B. Hammer and F. Besenbacher, Science,2008, 320, 1755.

13 C. L. Pang, O. Bikondoa, D. S. Humphrey, A. C. Papageorgiou,G. Cabailh, R. Ithnin, Q. Chen, C. A. Muryn, H. Onishi andG. Thornton, Nanotechnology, 2006, 17, 5397.

14 Y. Du, N. A. Deskins, Z. Zhang, Z. Dohnalek, M. Dupuis andI. Lyubinetsky, J. Phys. Chem. C, 2009, 113, 666.

15 N. G. Petrik and G. A. Kimmel, J. Phys. Chem. C, 2011, 115, 152.16 M. A. Henderson, W. S. Epling, C. L. Perkins, C. H. F. Peden

and U. Diebold, J. Phys. Chem. B, 1999, 103, 5328.17 P. Scheiber, A. Riss, M. Schmid, P. Varga and U. Diebold,

Phys. Rev. Lett., 2010, 105, 216101.18 E. Lira, J. ø. Hansen, P. Huo, R. Bechstein, P. Galliker,

E. Lægsgaard, B. Hammer, S. Wendt and F. Besenbacher,Surf. Sci., 2010, 604, 1945.

19 Y. Du, N. A. Deskins, Z. Zhang, Z. Dohnalek, M. Dupuis andI. Lyubinetsky, Phys. Chem. Chem. Phys., 2010, 12, 6337.

20 S. Tan, Y. Ji, Y. Zhao, A. Zhao, B. Wang, J. Yang andJ. G. Hou, J. Am. Chem. Soc., 2011, 133, 2002.

21 P. M. Jayaweera, E. L. Quah and H. Idriss, J. Phys. Chem. C,2007, 111, 1764.

22 M. Yasuo, A. Sasahara and H. Onishi, J. Phys. Chem. C, 2010,114, 14579.

23 D. Brinkley and T. Engel, J. Phys. Chem. B, 2000, 104, 9836.24 H. Idriss, P. Legare and G. Maire, Surf. Sci., 2002, 515, 413.25 M. Shen and M. A. Henderson, J. Phys. Chem. Lett., 2011,

2, 2707.26 K. P. Phillips, S. C. Jensen, M. Baron, S. C. Li and

C. M. Friend, J. Am. Chem. Soc., 2012, 135, 574.27 M. A. Henderson, J. Phys. Chem. B, 2004, 108, 18932.28 M. A. Henderson, J. Catal., 2008, 256, 287.29 M. A. Henderson, J. Phys. Chem. C, 2008, 112, 11433.30 M. A. Henderson, Surf. Sci., 2008, 602, 3188.31 R. T. Zehr and M. A. Henderson, Surf. Sci., 2008, 602, 2238.32 R. T. Zehr, N. A. Deskins and M. A. Henderson, J. Phys.

Chem. C, 2008, 114, 16900.33 R. T. Zehr and M. A. Henderson, Phys. Chem. Chem. Phys.,

2010, 12, 8085.34 T. Wang, D. A. Dixon and M. A. Henderson, J. Phys. Chem. B,

2010, 114, 14083.35 M. A. Henderson, N. A. Deskins, R. T. Zehr and M. Dupuis,

J. Catal., 2011, 279, 205.36 S. C. Jensen, A. Shank, R. J. Madix and C. M. Friend, ACS

Nano, 2012, 6, 2925.37 Y. Xia, B. Zhang, J. Ye, Q. Ge and Z. Zhang, J. Phys. Chem.

Lett., 2012, 3, 2970.38 I. Horcas, R. Fernandez, J. M. Gomez-Rodriguez,

J. Colchero, J. Gomez-Herrero and A. M. Baro, Rev. Sci.Instrum., 2007, 78, 013705.

39 O. Bikondoa, C. L. Pang, R. Ithnin, C. A. Muryn, H. Onishiand G. Thornton, Nat. Mater., 2006, 5, 189.

40 Y. Du, Z. Dohnalek and I. Lyubinetsky, J. Phys. Chem. C,2008, 112, 2649.

41 S. Wendt, R. Schaub, J. Matthiesen, E. K. Vestergaard,E. Wahlstrom, M. D. Rasmussen, P. Thostrup, L. M. Molina,E. Laegsgaard, I. Stensgaard, B. Hammer and F. Besenbacher,Surf. Sci., 2005, 598, 226.

42 Z. Zhang, O. Bondarchuk, B. D. Kay, J. M. White andZ. Dohnalek, J. Phys. Chem. B, 2006, 110, 21840.

43 S. Luo and J. L. Falconer, J. Catal., 1999, 185, 393.44 W. Xu and D. Raftery, J. Catal., 2001, 204, 110.45 M. I. Zaki, M. A. Hasan and L. Pasupulety, Langmuir, 2001,

17, 768.46 D. M. Griffiths and C. H. Rochester, J. Chem. Soc., Faraday

Trans. 1, 1978, 74, 403.47 M. A. Hasan, M. I. Zaki and L. Pasupulety, Appl. Catal., A,

2003, 243, 81.48 Y. Du, N. A. Deskins, Z. Zhang, Z. Dohnalek, M. Dupuis and

I. Lyubinetsky, Phys. Rev. Lett., 2009, 102, 096102.

PCCP Paper

Publ

ishe

d on

18

June

201

3. D

ownl

oade

d by

Uni

vers

ity o

f Il

linoi

s at

Chi

cago

on

28/1

0/20

14 1

8:18

:48.

View Article Online