Photocatalytic Reactions on Model Single Crystal TiO2

0 downloads 0 Views 535KB Size Report
9. Photoreaction. Dark and photoreactions on metal oxides. A(g). A(a). D(a). D(g). TM ... CH3CH2OH ю 3 H2O ! 2 CO2 ..... TiO2 ю 2 UV photons ю CH3CH2OH !

3 Photocatalytic Reactions on Model Single Crystal TiO2 Surfaces G.I.N. Waterhouse1 and H. Idriss2,3 1

Department of Chemistry, University of Auckland, Auckland, New Zealand Department of Chemistry, University of Aberdeen, Aberdeen, UK 3 School of Engineering, Robert Gordon University, Schoolhill, Aberdeen, UK, Email: [email protected] 2

Light from the sun is by far the most abundant source of energy on earth. Yet, at present, less than 0.05% of the total power (15 000 GW annual) used by humans is generated from the sun (excluding solar heating, which contributes around 0.6%). The estimated practical and convertible power that the earth surface receives is equivalent to that provided by 600 000 nuclear reactors (one nuclear power plant generates, on average, 1 GW power) or about 40 times the present global need.1 One mode of solar energy utilization is the use of sunlight to generate energy carriers, such as hydrogen, from renewable sources (e.g., ethanol and water) using semiconductor photocatalysts. The photoassisted splitting of water into hydrogen and oxygen was first achieved by Fujushima and Honda [1], who showed that hydrogen and oxygen could be generated in an electrochemical cell containing a titania photoelectrode, provided an external bias was applied. Since that time, numerous researchers have explored ways of achieving direct water dissociation without the need for an external bias. Much work has been conducted, a large fraction of which is discussed in a recent review [2]. Among the many issues affecting direct water splitting is the need to separate hydrogen from oxygen and the relatively low hydrogen evolution rates so far achieved. These, in addition to the need for using UV light (>3eV) to excite TiO2 and other related materials, has been one of the main obstacles for practical applications. Many authors have sought modified photocatalysts which, unlike pure TiO2, respond to visible (sunlight) excitation, with limited success to date; see some of these materials in ref. [2]. 1

The total amount of sunlight reaching the earth surface is orders of magnitude higher than the quoted figure.

On Solar Hydrogen & Nanotechnology Edited by Lionel Vayssieres © 2009 John Wiley & Sons (Asia) Pte Ltd. ISBN: 978-0-470-82397-2

78

On Solar Hydrogen & Nanotechnology

Due to its widespread use as a photocatalyst, and its participation in many other processes and applications, including biomaterials, sensors and as catalyst support for transition metals and metal oxides, TiO2 powder and single-crystal surfaces have been studied extensively. The reactions over TiO2 single-crystal surfaces of chemical compounds with different functional groups, including carboxylic acids, alcohols, aldehydes, aromatics and alkynes, in addition to smaller molecules such as CO, NO, NO2, CO2, H2S and H2O, among others, have been studied in the last two decades [3–5]. Results of these studies are of some interest when dealing with photoreactions, in particular those related to adsorption geometry and energetics. The photoreactions of organic compounds over these single-crystal surfaces have received less attention, but interest in these reactions are increasing [6–10]. The most-studied photoreactions on rutile TiO2(110) single-crystal surfaces include those of ethanol [6], acetic acid [11], trimethyl acetic acid [8,9] and acetone [10]. Sporadic work has been conducted on anatase TiO2 single crystals, although, to date, no published work dealing with photoreactions is known. Before discussing some of these results, it is worthwhile giving a brief introduction on the surfaces of rutile TiO2. This will be followed by another brief introduction about the chemical and physical processes thought to be involved in photocatalytic reactions over a semiconductor.

3.1 TiO2 Single-Crystal Surfaces The surfaces of single crystals of rutile TiO2, such as the (001), (110) and (100), have been studied for decades as models for metal oxide. TiO2 exists in several different polymorphic forms, such as the rutile, anatase and brookite polymorphs. While brookite is not a common TiO2 phase, the anatase phase is often found in nature. In addition, because the anatase phase can be easily synthesized with high surface area many TiO2-based catalytic materials contain substantial amounts of the anatase phase [12]. However, the anatase phase is considerably less stable than the rutile phase; it can transform to the rutile phase at temperatures as low as 500 K. While this may not be very critical to some catalytic reactions, it has rendered their study by surface-science methods more complex, because high temperatures are required to prepare clean surfaces for chemical reactions. Some single-crystal surfaces of anatase TiO2 have been successfully studied by scanning tunneling microscopy (STM) [13,14] although most studies are conducted by computation modeling [15–17]. Surface energy is one of the main parameters determining the stability of a surface and possibly reactivity. One of the crucial factors determining surface stability is the coordination number of surface atoms. The closer the coordination number of a given surface atom to that of the bulk, the more stable the surface. In both the rutile and anatase structures the coordination number of Ti atoms in the bulk is six, and three for oxygen. Table 3.1 presents the computed surface energies for some low-Miller-index rutile TiO2 surfaces, while Figure 3.1 shows the most common surface structures. The (110) rutile surface contains alternating rows of Ti and O atoms five- and twofold coordinated, respectively. Ti atoms underneath the twofold coordinated oxygen atoms are sixfold coordinated. Because of the high coordination numbers of Ti atoms, this surface is the most stable, as evidenced by the value of its surface energy (Table 3.1). Because of its high stability, the (110) surface has received most of the attention in surface-science studies and has been often considered as a bench mark for metal oxides. The (100) surface is second most stable surface, where all surface Ti atoms are fivefold coordinated and all O atoms are twofold

79

Photocatalytic Reactions on Model Single Crystal TiO2 Surfaces Table 3.1

Computed surface energy by density functional theory (DFT) for rutile TiO2 surfaces [18].

Surface

(110) (100) (011) (011)b 1  1 (011)c 2  1 (001)

DFT methoda

Surface energy meV/(a.u.)2

J m2

15.6 19.6 24.4 19.3 7.4 28.9

0.89 1.12 1.39 1.10 0.42 1.65

LDA LDA LDA GGA GGA LDA

a

LDA: local density approximation; GGA: generalized gradient approximation. From [19]. c From [20]. b

coordinated. The (001) surface with fourfold coordinated Ti atoms is the least stable of the lowindex rutile surfaces. However, this surface is particularly interesting as it readily facets to the {011} or {114} surfaces at high temperature [21]. These facets are highly reactive, and the reactivity of a number of small molecules, including numerous oxygen- and nitrogen-containing compounds have been studied [22–24]. The exact structures of the (011) and (114) surfaces are still debated. Models derived from STM studies demonstrate that the (114) surface is composed of steps, each containing one fourfold-coordinated Ti atom. The shortest distance between two fourfold-coordinated Ti atoms is 0.65 nm (along the [110] direction) while the distance between two successive steps is equal to 0.30 nm along the [11 0] direction). This model agrees with Ti2p core level X-ray photoelectron spectroscopy (XPS) results, indicating that the surface is composed of stoichiometric TiO2 units [25]. On the other hand, low-energy electron diffraction (LEED) [21] and STM [28] data suggest that the stable (011) surface is a (2  1) reconstructed surface. Initially a model for the (011) surface containing terminal Ti ¼ O groups was proposed [26]. Based on surface X-ray diffraction, a new model has been recently introduced [20,27]. This model, unlike the previous one, does not contain Ti ¼ O structures.

Figure 3.1 Side view of the unreconstructed (bulk terminated) rutile TiO2 surfaces. The black circles represent Ti atoms, and the gray circles are O atoms. The coordination number (CN) for the Ti atoms is five for the (110) and (1 0 0) and four for the (001) structure.

80

On Solar Hydrogen & Nanotechnology

3.2 Photoreactions Over Semiconductor Surfaces Figure 3.2 is a schematic description of the complex processes involved in photocatalysis over a material composed of a metal deposited on a wide bandgap semiconductor. Excitation of a semiconductor, such as TiO2, with photons of appropriate energy produces hole–electron pairs, which may either recombine or react with appropriate adsorbed species. Photocatalysis with TiO2 is a well-established field, which exploits these phenomena and which has been extensively reviewed [28]. Despite the enormous number of published papers in this field (particularly in the context of environmental clean-up), many aspects of reaction mechanisms remain poorly understood and therefore catalyst optimization has proceeded in a largely empirical manner. Much higher hydrogen production rates are reported for the photocatalytic reforming of alcohols when compared to those for water. This is mainly because alcohols are very active hole scavengers. The loading of titania photocatalysts with metal particles dramatically enhances the rate of hydrogen evolution, presumed to be due to trapping of conduction-band electrons within the metal clusters. The ambient temperature reforming of methanol over a variety of metal-loaded titania photocatalysts can be described as [29] CH3 OH þ H2 O ! CO2 þ 3 H2

ð3:1Þ

However, methanol is made from fossil fuel (CO þ H2) and therefore contributes to increasing carbon dioxide emissions. This method is, in addition, largely nonefficient, because hydrogen is consumed to make methanol in the first place.

Dark and photoreactions on metal oxides Energy, eV

A(g) NHE

1, 8 A(a) CB, M x+ nd

A(a) 9 E BG hν > EBG

5

4, 5 3

TM 7 D(a) 9

1. Adsorption 2. Surface structure 3. Interface 4. Charge diffusion 5. Charge trapping 6. Particle size 7. Surface electronic states 8. Dark reactions 9. Photoreaction

VB, O 2p 2, 6

D(a) 1, 8 D(g)

Figure 3.2 A schematic of the different photoreaction/photoexcitation processes relevant to catalytic reactions. The figure presents a schematic of the main chemical and physical reactions involved on the surfaces of a semiconductor under photon irradiation (legend at the right-hand side of the figure). Large sphere: semiconductor, small sphere: transition metal (TM) such as Au, Ru or Rh; VB: valence band (mainly O2p in the case of TiO2); CB: conduction band (mainly Ti3d levels in TiO2) A: acceptor; D: donor; NHE: normal hydrogen electrode; EBG: Bandgap ¼ 3–3.2 eV for TiO2; (a): adsorbed; (g): gas phase.

Photocatalytic Reactions on Model Single Crystal TiO2 Surfaces

81

In contrast, ethanol is a biorenewable hydrogen source because it is obtained from crop fermentation, and the release of CO2 during the reaction to ethanol does not add to CO2 emissions because of the biological origin of ethanol [30]. CH3 CH2 OH þ 3 H2 O ! 2 CO2 þ 6 H2

ð3:2Þ

3.3 Ethanol Reactions Over TiO2(110) Surface There is no reported hydrogen production from ethanol or any other alcohol over single-crystal surfaces via photocatalysis. However, the photo-oxidation of ethanol has been studied over the rutile TiO2(110) surface. Ethanol is dissociatively adsorbed via its oxygen lone pair on fivefold coordinated Ti atoms to produce adsorbed ethoxide species (Figure 3.3). STM studies of the adsorption of ethanol on TiO2(110) demonstrated the presence of both alkoxides and surface hydroxyls [31], confirming the adsorption is dissociative. The adsorption on this surface is similar to that of an acid–base type, whereby the O atom of ethanol acts as a Lewis base and the surface Ti acts as a Lewis acid. Because of this strong interaction, the OH bond is broken, even though it is the strongest bond of the molecule (see bond energies of an ethanol molecule at the left-hand side of Figure 3.3). The adsorption energy of ethanol on this surface has not been computed yet, although that of methanol, at y ¼ 0.5, has been found equal to 100 and 120 kJ mol1 for the molecular and dissociated adsorptions, respectively [32]. Figure 3.4a at 0 minutes displays the XPS C1s spectra after saturation exposure of ethanol at 300 K; the surface coverage y was found to be 0.5 with respect to surface Ti atoms, and this

Figure 3.3 The left-hand side of the figure: top and side views of the (110) surface of the rutile TiO2. Small (blue) circles represent Ti atoms and large (red) circles represent O atoms. The right-hand side of the figure gives the bond energies of an ethanol molecule. A representation of the adsorbed ethoxide species on top of a fivefold coordinated Ti is also seen. The H resulting from the dissociative adsorption is presumably on top of the twofold coordinated surface O atoms (bridging O atoms), not shown. The atomic scale between ethanol and TiO2 is not unified. The very large circle (pink) represents the van der Waals dimension of an ethanol molecule. Data for this figure are provided from spectroscopic measurement. (Adapted with permission from P.M. Jayaweera et al., Photoreaction of ethanol on TiO2(110) single-crystal surface, Journal of Physical Chemistry C, 111(4), 2007. Ó 2007 American Chemical Society.)

82

On Solar Hydrogen & Nanotechnology

Time / min.

Intensity (arbitrary units)

10

0

UV + 10–6 Torr O2

ln

Ct

20

30

4×10–11 Torr Co

–0.1

–7

PH = 7×10

-CH2O- -RCOO-

1.0

20 min.

Torr

–8

Ct /C 0

-CH3

2

–0.2

65 min.

8×10 Torr 2×10–7 Torr 1×10–6 Torr

–0.3

0.9 0.8

0 min. 0.7

280

284 288 292 Binding Energy (eV)

8x10

–8

Torr

2x10

–7

Torr

296 0

20

40

60

Time (min) (b)

(a)

Figure 3.4 (a) XPS C1s of adsorbed ethanol on TiO2(110) at 300 K at saturation before (0 min) and after UV illumination (Hg lamp with main line at 360 nm) in the presence of 106 Torr O2 at different time intervals. (b) The peak area of the XPS C1s peak as a function of illumination time at different O2 partial pressures. The inset is the initial linear decrease of the computed peak areas at different partial pressures of O2. Included in the figure is an additional run under H2 (7  107 Torr) for comparison with the one in the absence of O2 (PO2 ¼ 4  1011 Torr is estimated as the upper value for oxygen pressure at a vacuum of 2  1010 Torr). (Adapted with permission from P.M. Jayaweera et al., Photoreaction of ethanol on TiO2(110) single-crystal surface, Journal of Physical Chemistry C, 111(4), 2007. Ó 2007 American Chemical Society.)

might be due to the van der Waals radius of the ethanol molecule, as depicted in the left-hand side of Figure 3.3. As seen in Figure 3.4, the C1s structure consists of two species, because the two carbon atoms of ethoxides are in different chemical environments (-CH3 and -CH2O-). Upon exposure to UV light in presence of 106 Torr O2, two main changes in the XPS spectra are observed [6]: (i) a slight decrease in the C1s peak area for both functional groups (-CH2Oand -CH3) with increasing irradiation time, and (ii) the formation of a peak at 290 eV, which is attributed to an RCOO(a) species. The absorption of a photon by a surface-bound ethoxide results in the transfer of electrons from the valence band (occupied states containing contributions from the O2p level) to the conduction band (empty states containing contributions from Ti3d level). The decrease in the ethoxide C1s peaks is due to its reaction with . O2 and/or . O, which form according to the following reactions: UV

TiO2 ! e þ h þ

ð3:3Þ

e þ O2 ! O2

ð3:4Þ

O2. þ CH3 CH2 OðaÞ ! HO2 þ CH3 CHO.ðadÞ ! þ OðaÞ or O2;ðgÞ ! CH3 COOðaÞ ! ! CO2 þ H2 O

ð3:5Þ

.

The two successive arrows ( ! ! ) indicate multiple reaction steps of unidentified intermediates.

Photocatalytic Reactions on Model Single Crystal TiO2 Surfaces

83

The change in ethoxide concentration (expressed as the ratio, Ct/C0, where Ct is the area of the two C1s peaks of the ethoxide species at time t and C0 is the area of the two peaks before irradiation) is plotted as a function of time (Figure 3.4b). These data are similar to those reported in Figure 3.4a, except each line is the result of a photoreaction conducted at a different O2 pressure. Assuming a simple exponential decay of adsorbed ethoxide with time, the timedependent concentration is written as: Ct ¼ C0 expðktÞ

or Ct ¼ C0 expðFQtÞ

and rearranging, the photoreaction cross-section, Q (in cm2) becomes   1 C0 Q ¼ ln Ft Ct

ð3:6Þ

ð3:7Þ

where k is a pseudo first-order rate constant, t is the illumination time and F is the UV flux. Q is found to be equal to 2  1018 cm2 at a pressure of 106 Torr O2. Some of the ethoxide species are converted to acetate (RCOO, where R ¼ CH3). In the photo-oxidation of ethanol, the reaction products are CO2 and water. Detailed studies of reaction intermediates on model metal oxides are scarce, however more work has been conducted on high-surface-area powders. The reaction intermediates on TiO2 have been studied by temperature programmed desorption (TPD) [33] and IR [34] and the formation of CH3CHO by the removal of an electron from the radical of Equation 3.8 has been reported CH3 CHOðaÞ þ h þ ! CH3 CHOðaÞ .

ð3:8Þ

The acetaldehyde radical may further react in a complex set of reactions, which are not well understood, to yield the final product CO2 [35].

3.4 Photocatalysis and Structure Sensitivity The steady-state photocatalytic reactions in ultra-high-vacuum conditions, using acetic acid and rutile TiO2 (001) single crystals, have been studied in considerable detail. The rutile TiO2 (001) surface reconstructs to two stable facets, depending on the annealing temperature: the low-temperature phase (011) and the high-temperature phase (114). Because photocatalysis directly depends on both bulk and surface properties, studying these single-crystal surfaces has helped in decoupling surface from bulk contributions. Acetic acid reacts photocatalytically at RT over TiO2 to give methane, ethane and CO2, as depicted by the two equations at the bottom of Figure 3.5. The left-hand side of the figure shows the formation of ethane (m/z 30), methane (m/z 16) and CO2 (m/z 44) upon UV illumination over the (011) reconstructed TiO2 surface. While CO2 and methane production is constant and tracks UV light exposure, that of ethane decreases with time. The reason has been found to be depletion of surface oxygen; that is why the surface oxygen atoms in the second equation of the figure are in bold as they are removed from the surface in the form of water. Co-feeding O2 in the gas phase with acetic acid was found to maintain ethane production, as it restores surface oxygen atoms. The right-hand side of the figure shows the activity towards ethane formation under UV light over the (011) and (114) reconstructed surfaces. It was found that the (011) reconstructed surface of rutile TiO2(001) is more active than the (114) reconstructed surface of the same crystal [36,37]. While the reasons for the different reactivity between the two surfaces

84

On Solar Hydrogen & Nanotechnology

7E-09

Pressure (torr)

2.0E-08 1.5E-08

44 m/z CO2

1.0E-08

16 m/z CH4

5.0E-09 0.0E+00 0:00:00

30 m/z C2H6

0:02:53

0:05:46

Peakjump ethane (torr)

2.5E-08

CH3CH3

4E-09 3E-09 2E-09

{114}-faceted surface

1E-09

1E-07

2E-07

3E-07

4E-07

5E-07

6E-07

Acetic acid pressure (torr)

Time (minutes)

CH3COOH(ads)

5E-09

0E+00 0E+00

0:08:38

{011}-faceted surface

6E-09

CO2(g)+ CH4(g)

2 CH3COOH(ads)+O(lattice)

2 CO2(g)+ CH3CH3(g)+ H2O(ads)

1 2

Figure 3.5 Structure sensitivity studied by photocatalytic reaction of rutile TiO2 surfaces: acetic acid photoreactions over TiO2(001) reconstructed surfaces ((011) and (114)). Acetic acid reacts under photons with the surface to form CO2 and methane or CO2, ethane and water. Reaction 2 consumes surface O atoms, while reaction 1 does not. The (011) surface is easier to reduce than the (114) surface and is therefore more active for reaction 2. Injecting O2 together with acetic acid regenerates surface O atoms and makes both surfaces behave similarly.

are not understood, the fact that both surfaces share the same bulk structures removes the bulk electron–hole recombination rate as a factor, and therefore different surface different electron–hole recombination rates would be one of the main contributing factors. One accurate method for monitoring adsorbate evolution on model surfaces is shown in Figure 3.6, where the signal of a dosed compound, in this case acetic acid, on a surface of a single crystal, TiO2(011) rutile, is monitored as a function of exposure to photon time, as well as gasphase oxygen pressure. From such measurements one can estimate the photoionization crosssection, as well as the reaction quantum yield, in a similar way to that presented for ethanol.

3.5 Hydrogen Production from Ethanol Over Au/TiO2 Catalysts Figure 3.7 presents some recent data for the photocatalyzed dehydrogenation of ethanol using 1.5 Au wt% on TiO2 Degussa P25. Degussa P25 is composed of about 80% anatase and 20% rutile and is one of the most photocatalytically active TiO2 materials known to date. The inset at the right-hand side of the figure is a TEM image, where anatase TiO2 particles are 30 nm in size, while the larger particles of 70–80 nm size are those of the rutile phase. The small dark particles of about 3 nm size are those of Au particles. The left-hand side of the figure presents XPS Au4f lines of Au particles of the catalyst, together with a comparison with an Au foil metal. The shift of the Au4f lines to lower binding energy with respect to that of the Au foil is due to negative charge transfer from the support to the metal. The atomic percentage of Au as determined from Au4f XPS is about 0.7%. Dark catalytic oxidation of CO to CO2 on Au/TiO2 has been shown to be most active when the particle size of Au is close to 3 nm [38,39]. Moreover, the upward shift of the Fermi level has been determined (with respect to the normal hydrogen electrode (NHE)) to increase with decreasing Au particle size, with a shift of 290 mV for 3 nm particles [40]. It is

85

Photocatalytic Reactions on Model Single Crystal TiO2 Surfaces

Acetic acid / TiO2(011)

1

285.5

Counts / s

Ct / C0

289.5

0.9

10–8

0.8

10–7

0.7

10–6

0.6 0

2000

4000

Time(s) Ct = C 0 exp(–k t), k rate constant in s–1 Ct = C 0 exp(–F Q t ) F, Flux of UV photon, Q, the photoionization cross-section. 292

288

284

Binding Energy (eV)

280

Q ethanol TiO2 (110) ≈ 6 X Q acetic acid TiO2(011) With OH radicals k ethanol ≈ 5 X Q acetic acid

Figure 3.6 XPS C1s following acetic acid adsorption on rutile TiO2 (011) surface. Acetates are formed from dissociative adsorption of acetic acid on a TiO2(011) surface. The left-hand side of the figure presents the signal of the two groups as studied by their XPS C1s signal. From monitoring the surface population with increasing photons and exposure time, as well as under different oxygen pressures, the intrinsic parameters of the reaction can be extracted. It was found that acetic acid photodecomposition is about six times slower than that of ethanol on TiO2(110). The second-order rate constant for ethanol and acetic acid reactions with OH radicals in the gas phase indicates similar numbers.

seen in the figure that after an initial (not well-understood) “induction” period, close to linear production of hydrogen occurs. The calculated quantum yield was found to be 0.06 (or 6%). Details of the reaction mechanism is not well understood, but involves in part a two-step electron transfer to make one mole of hydrogen per mole of ethanol and the associated production of acetaldehyde, as indicated below. Within a semiconductor (such as TiO2), upon excitation with photons of energy equal or higher than its bandgap (EBG), electron transfers from the valence band (VB) to the conduction band (CB) occur, consequently creating electron (e) and hole (h þ ) pairs. Photoexcitation : TiO2 þ 2 UV photons ! 2 e þ 2 h þ

ð3:9Þ

If we consider ethanol as the reactant, Equations 3.10–3.15 describe the formation of one hydrogen molecule per one ethanol molecule, using two photons. The presence of a transition metal (acting as an electron trap) decreases the electron–hole recombination rate (the reverse of Equation 3.9). If an adsorbed species (an acceptor such as H þ ) is at the interface M/TiO2, as depicted in Figure 3.2, it may react with the electron and is therefore reduced (H þ þ e ! 1 /2 H2). Similarly, regeneration of electrons in the VB may be possible by electron injection

86

On Solar Hydrogen & Nanotechnology

Figure 3.7 Hydrogen production from ethanol over 1.5 wt% Au/TiO2 P25 (80% anatase – 20% rutile). y-axis presents mol l1 of hydrogen evolved as a function of time (x-axis) from a solid–liquid batch reactor containing 50 mg of catalyst and 15 ml of ethanol (total reactor volume about 100 ml). Prior to reaction, the catalyst was reduced using hydrogen at 573 K overnight. Top inset: XPS Au4f of the catalyst and of gold foil. Inset bottom TEM of the catalyst. Au particles are seen as dark spots with an average diameter of 3 nm, larger particles (about 30 nm size) are anatase, while the largest particles (70–80 nm size) are rutile. Reaction temperature 315 K.

from another surface intermediate (CH3 CH2 OðaÞ ! CH3 CHO.ðaÞ þ H þ ). The latter reaction repeated a second time gives CH3CHO. Dissociative adsorption of ethanol on Ti–O sites: CH3 CH2 OH þ Ti-O ! CH3 CH2 OðTiÞ þ OH

ð3:10Þ

First electron reduction at the CB: OHðaÞ þ e ! OðaÞ þ 1=2 H2

ð3:11Þ

CH3 CH2 OðTiÞ þ h þ ! CH3 CHO ðTiÞ þ H þ

ð3:12Þ

First hole trapping at the VB: .

Second hole trapping at the VB: CH3 CHO ðTiÞ þ h þ ! CH3 CHOðgÞ þ Ti .

ð3:13Þ

Second electron reduction at the CB: H þ þ e ! 1=2 H2

ð3:14Þ

Photocatalytic Reactions on Model Single Crystal TiO2 Surfaces

87

Overall reaction: TiO2 þ 2 UV photons þ CH3 CH2 OH ! CH3 CHOðgÞ þ H2 ðgÞ

ð3:15Þ

For the above equation, the time intervals between two absorbed photons is considered too small compared to electron (hole) transfer reactions. This is true for large particles (bulk materials), but breaks down for nanosize particles, where large time intervals may prevent the formation of acetaldehyde, and other products can be formed [41].

3.6 Conclusions An overview of the photoreactions of organic compounds on TiO2 single-crystal surfaces has been presented. The photoreactions of ethanol and acetic acid were used as specific examples. The acetic acid photoreaction over reconstructed TiO2 (001) single crystal surfaces (the (011) and (114)) displays a structure-sensitive reaction with different yields attributed to changes in the electron–hole recombination rates. The ethanol photo-oxidation reaction yield on the rutile TiO2(110) surface has been shown to be a function of O2 pressure. The build up of surface acetates appears to slow the oxidation reaction with time, which is explained by the fact that the latter have a slower photo-oxidation reaction rate. The photoreduction of ethanol over transition metals and TiO2 can be successfully conducted. The example presented here taking 1.5 wt.% Au/TiO2 P25 shows that uniform 3 nm size gold particles supported on titania give a photoreaction quantum yield close to 10%, depending on the ethanol to water ratio.

References [1] Fujishima, A. and Honda, K. (1972) Electrochemical photolysis of water at a semiconductor electrode. Nature, 238, 37. [2] Kudo, A. and Miseki, Y. (2009) Heterogeneous photocatalyst materials for water splitting. Chemical Society Reviews, 38, 253. [3] Idriss, H. and Barteau, M.A. (2000) Active sites on oxides: from single crystals to catalysts. Advances in Catalysis, 45, 261. [4] Diebold, U. (2003) The surface science of titanium dioxide. Surface Science Reports, 48, 53. [5] Qiu, H., Idriss, H., Wang, Y., and W€oll, C. (2008) Carbon-carbon bond formation on model titanium oxide surfaces: identification of surface reaction intermediates by high-resolution electron energy loss spectroscopy. Journal of Physical Chemistry C, 112, 9828. [6] Jayaweera, P.M., Quah, E.L., and Idriss, H. (2007) Photoreaction of ethanol on TiO2(110) single-crystal surface. Journal of Physical Chemistry C, 111, 1764. [7] Henderson, M.A., White, J.M., Uetsukab, H., and Onishi, H. (2006) Selectivity changes during organic photooxidation on TiO2: Role of O2 pressure and organic coverage. Journal of Catalysis, 238, 153. [8] Lyubinetsky, I., Yu, Z.Q., and Henderson, M.A. (2007) Direct observation of adsorption evolution and bonding configuration of TMAA on TiO2(110). Journal of Physical Chemistry C, 111, 4342. [9] White, J.M., Szanyi, J., and Henderson, M.A. (2004) Thermal chemistry of trimethyl acetic acid on TiO2(110). Journal of Physical Chemistry B, 108, 3592. [10] Henderson, M.A. (2005) Photooxidation of acetone on TiO2(110): conversion to acetate via methyl radical ejection. Journal of Physical Chemistry B, 109, 12062. [11] Wilson, J.N. and Senanayake, S.D. (2004) Carbon coupling on titanium oxide at surface oxygen defects, H. Idriss. Surface Science, L231, 562. [12] Ni, M., Leung, M.K.H., Leung, D.Y.C., and Sumathy, K. (2007) A review and recent developments in photocatalytic water-splitting using TiO2 for hydrogen production. Renewable and Sustainable Energy Reviews, 11, 401.

88

On Solar Hydrogen & Nanotechnology

[13] Ruzycki, N., Herman, G.S., Boatner, L.A., and Diebold, U. (2003) Scanning tunneling microscopy study of the anatase (1 0 0) surface. Surface Science, 529, L239. [14] Tanner, R.E., Sasahara, A., Liang, Y., Altman, E.I., and Onishi, H. (2002) Formic acid adsorption on anatase TiO2(001)–(1  4) thin films studied by NC-AFM and STM. Journal of Physical Chemistry B, 106, 8211. [15] Lazzeri, M. and Selloni, A. (2001) Stress-driven reconstruction of an oxide surface: the anatase TiO2(001)-(1  4) surface. Physical Review Letters, 87, 266105. [16] Bonapasta, A.A. and Filippone, F. (2005) Photocatalytic reduction of oxygen molecules at the (100) TiO2 anatase surface. Surface Science, 577, 59. [17] Czekaj, I., Piazzesi, G., Kr€ocher, O., and Wokaun, A. (2006) DFT modeling of the hydrolysis of isocyanic acid over the TiO2 anatase (101) surface: Adsorption of HNCO species. Surface Science, 600, 5158. [18] Ramamoorthy, M. and Vanderbilt, A. (1994) First principles calculations of the energetic of stoichiometric TiO2 surfaces. Physical Review B-Condensed Matter, 49, 16721. [19] McGill, P.R. and Idriss, H. (2008) Plane wave density functional theory computation of the molecular adsorption of ammonia on the (011) surface of titanium dioxide. Langmuir, 24, 97. [20] Gong, X.-Q., Khorshidi, N., Stierle, A., Vonk, V., Ellinger, C., Dosch, H., Cheng, H., Selloni, A., He, Y., Dulub, O., and Diebold, U. (2009) The 2  1 reconstruction of the rutile TiO2(011) surface: A combined density functional theory, X-ray diffraction, and scanning tunneling microscopy study. Surface Science, 603, 138. [21] Firment, L.E. (1982) Thermal faceting of the rutile TiO2(001) surface. Surface Science, 116, 205. [22] Kim, K.S. and Barteau, M.A. (1990) Structure and composition requirements for deoxygenation, dehydration, and ketonization reactions of carboxylic acids on TiO2(001) single-crystal surfaces. Journal of Catalysis, 125, 353. [23] Idriss, H., Kim, K.S., and Barteau, M.A. (1991) Structure-activity and structure-selectivity relations for reactions of carboxylic acids on TiO2(001) surfaces. Studies in Surface Science and Catalysis, 64, 327. [24] Senanayake, S.D. and Idriss, H. (2006) Photo-catalyses and the origin of life. The synthesis of nucleotides from formamide on TiO2(001) single surfaces. Proceedings of the National Academy of Sciences of the United States of America, 103, 1194. [25] Ariga, H., Taniike, T., Morikawa, H., Tero, R., Kondoh, H., and Iwasawa, Y. (2008) Lattice-work structure of a TiO2(001) surface studied by STM, core-level spectroscopies and DFT calculations. Chemical Physics Letters, 454, 350. [26] Valentin, C., Tilocca, A., Selloni, A., Beck, T.J., Klust, A., Batzill, M., Losovyj, Y., and Diebold, U. (2005) Adsorption of water on reconstructed rutile TiO2(011)-(2  1): TiO double bonds and surface reactivity. Journal of the American Chemical Society, 127, 9895. [27] Torrelles, X., Cabailh, G., Lindsay, R., Bikondoa, O., Roy, J., Zegenhagen, J., Teobaldi, G., Hofer, W.A., and Thornton, G. (2009) Geometric structure of TiO2(011) (2  1). Physical Review Letters, 101, 185501. [28] Fujishima, A., Zhang, X., and Tryk, D.A. (2008) TiO2 photocatalysis and related surface phenomena. Surface Science Reports, 63, 515 and reference therein. [29] Al-Mazroai, L.S., Bowker, M., Davies, P., Dickinson, A., Greaves, J., James, D., and Millard, L.M. (2007) The photocatalytic reforming of methanol. Catalysis Today, 122, 46. [30] Yang, Y.Z., Chang, C.H., and Idriss, H. (2006) Photo-catalytic production of hydrogen from ethanol over M/TiO2 catalysts (M ¼ Pd, Pt or Rh). Applied Catalysis B-Environmental, 67, 217. [31] Zhang, Z., Bondarchuk, O., Kay, B.D., White, J.M., and Dohnalek, Z. (2007) Direct visualization of 2-butanol adsorption and dissociation on TiO2(110). The Journal of Physical Chemistry C, 111, 3021. [32] Bates, S.P., Kresse, G.K., and Gillan, M.J. (1998) The adsorption and dissociation of ROH molecules on TiO2(110). Surface Science, 409, 336. [33] Reztova, T., Chang, C.-H., Koresh, J., and Idriss, H. (1999) Dark and photoreactions of ethanol and acetaldehyde over TiO2/carbon molecular sieve fiber. Journal of Catalysis, 185, 223. [34] Yu, Z. and Chuang, S.S.C. (2007) In situ IR study of adsorbed species and photogenerated electrons during photocatalytic oxidation of ethanol on TiO2. Journal of Catalysis, 246, 118. [35] Zehr, R.T. and Henderson, M.A. (2008) Acetaldehyde photochemistry on TiO2(110). Surface Science, 602, 2238. [36] Wilson, J.N. and Idriss, H. (2002) Structure-sensitivity and photo-catalytic reactions of semiconductors. Effect of the last layer atomic arrangement. Journal of the American Chemical Society, 124, 11284. [37] Wilson, J.N. and Idriss, H. (2003) Effect of surface reconstruction of TiO2(001) single crystal on the photoreaction of acetic acid. Journal of Catalysis, 214, 46.

Photocatalytic Reactions on Model Single Crystal TiO2 Surfaces

89

[38] Goodman, D.W. and Lai, X. (2000) Structure–reactivity correlations for oxide-supported metal catalysts: new perspectives from STM. Journal of Molecular Catalysis A-Chemical, 162, 33. [39] Haruta, M. (1997) Size- and support-dependency in the catalysis of gold. Catalysis Today, 36, 153. [40] Kamat, P.V. (2008) Quantum dot solar cells. semiconductor nanocrystals as light harvesters. Journal of Physical Chemistry C, 112, 18737 and references therein. [41] Mller, B.R., Majoni, S., Memming, R., and Meissner, D. (1997) Particle size and surface chemistry in photoelectrochemical reactions at semiconductor particles. Journal of Physical Chemistry B, 101, 2501.

Suggest Documents