A Photoluminescence Study Under High Vacuum Conditions

1 day ago - The interaction between hydrogen species and TiO2 surfaces is important because of its relevance to hydrogen production in the photocataly...
0 downloads 10 Views 965KB Size
Subscriber access provided by Stockholm University Library

C: Surfaces, Interfaces, Porous Materials, and Catalysis 2

Photoreduction of Hydrogen Cations on TiO and Its Impact on Surface Band Bending and the Charge Carrier Recombination Rate: A Photoluminescence Study Under High Vacuum Conditions Shiliang Ma, Zhen Zhang, and Ian Harrison J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b12624 • Publication Date (Web): 02 Apr 2018 Downloaded from http://pubs.acs.org on April 3, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

1 2 3 4

Photoreduction of Hydrogen Cations on TiO2 and its

5

Impact on Surface Band Bending and the Charge Carrier

6

Recombination Rate: A Photoluminescence Study Under

7

High Vacuum Conditions

8

Shiliang Ma1, Zhen Zhang2, Ian Harrison1, *

9

1.

10 11 12

Department of Chemistry, University of Virginia, McCormick Road, Charlottesville, Virginia 22904-4319, USA

2.

Department of Chemical Physics, University of Science and Technology of China, Hefei, Anhui, 230026, P. R. China

13

Corresponding Author

14

* To whom correspondence should be addressed. Email: [email protected]

15

Tel.: +1 (434) 924-3639

16

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

17

ABSTRACT

18

The interaction between hydrogen species and TiO2 surfaces is important because of its

19

relevance to hydrogen production in the photocatalytic splitting of water on TiO2. In this

20

study, the effect of the photocatalytic reduction of hydrogen cations, a half

21

photoelectrochemical reaction run under high vacuum conditions, on the surface band

22

bending of TiO2 was investigated by photoluminescence (PL) spectroscopy. Exposure of

23

TiO2 to molecular hydrogen had no effect on the PL emission from TiO2 but the PL

24

intensity increased under UV irradiation when TiO2 was exposed to hydrogen cations

25

produced by ionization of molecular hydrogen by a cold cathode pressure gauge. The PL

26

intensity increase, signaling a decrease in the upward band bending and increased

27

electron-hole radiative recombination rate, is caused by charge transfer during

28

photoreduction of hydrogen cations. Residual photoexcited holes left in TiO2 due to the

29

transfer of photoexcited electrons to hydrogen cations tend to accumulate at the TiO2

30

surface, balancing the originally trapped electrons at the TiO2 surface and thereby

31

lowering the original upward band bending. These unusual observations point out that

32

charge transfer during a photoelectrochemical reaction at a semiconductor interface alters

33

the surface band bending and photoexcited electron-hole recombination rate in ways that

34

are likely to impact the efficiency of photocatalytic devices.

2 ACS Paragon Plus Environment

Page 2 of 21

Page 3 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

35 36

INTRODUCTION Enhancing photoexcited electron-hole pair separation and suppressing their

37

recombination in TiO2 have attracted vast research interest due to their central roles in

38

improving the efficiency of photocatalytic hydrogen production, pollutant decomposition,

39

and photovoltaic cells based on TiO2.1-3 Though it has been rarely studied, tuning the

40

surface band bending of TiO2 to more efficiently promote electron-hole pair separation

41

and actively sweep desired charge carriers to the surfaces is a promising method to

42

improve photocatalytic activity. Upward band bending is usually observed at TiO2

43

surfaces because negative charges accumulate at its surfaces upon the equilibration of the

44

bulk and surface Fermi energies.4-5 The upward band bending towards the surface directs

45

photoexcited electrons and holes formed near the surface to migrate in opposite directions,

46

and thus improves charge separation.6-8 Sheppard et al.9 showed by surface photovoltage

47

measurements that surface band bending introduced by doping niobium at different

48

concentrations influences charge separation rates in TiO2 during illumination. Zhang et

49

al.7 reported that a 100-fold difference in the hole-mediated photodesorption rate of O2

50

from TiO2(110) surface can be achieved by modulating the band bending of TiO2 through

51

donor or acceptor molecule adsorption. In addition, the effects of doping various cations

52

into TiO2 on photocatalytic hydrogen production rates, attributed to the band bending

53

changes by Karakitsou et al.,10 indicates that band bending change is a physical process

54

broadly applicable rather than a molecule-site specific chemical process which is only

55

valid for certain molecules or sites. Thus, studying band bending may provide

56

fundamental insights to better understand and improve a broad range of photocatalytic

57

processes and devices involving TiO2.

58

Photoluminescence (PL) spectroscopy which observes the radiative

59

recombination of photoexcited charge carries (electron-hole pair) serves as a contactless

60

and nondestructive method to characterize the band bending change of TiO2 based on a

61

dead-layer model in which there is little PL emission from the band bending region

62

(depletion layer) underneath the surface where the local potential drives the photoexcited

63

electrons and holes in opposite directions.4-6, 11-13 Hence, the measured PL emission

64

originates mainly from the TiO2 bulk. Because the 320 nm UV irradiation only penetrates 3 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

65

~ 25 nm into the TiO2 sample,14 which is comparable to the length of the depletion layer,

66

PL intensity change due to the band bending change in TiO2 can be easily observed under

67

high vacuum conditions.5-6 An increase in upward band bending that increases the depth

68

of the depletion layer and reduces the dimensions of the flat-band bulk region where

69

radiative electron-hole pair recombination can take place, causes PL emission intensity to

70

decrease. Therefore, a decrease in PL intensity indicates an increase in the upward band

71

bending of TiO2 and vice versa.5

72

The interaction between hydrogen and TiO2 has been extensively studied due to

73

its relevance to the final step in hydrogen production through photocatalytic water

74

splitting. Although it was reported that H2 only interacts with TiO2 weakly through van

75

der Waals interactions at room temperature,15-17 heating TiO2 in H2 at elevated

76

temperature introduces a new visible light response to TiO2 and improves its

77

photoactivity. These high temperature H2/TiO2 changes were attributed to hydrogen atom

78

incorporation and subsequent OH and TiH formation in the TiO2 lattice.18-20 Studies have

79

also shown that at room temperature, atomic hydrogen can diffuse into the TiO2 lattice

80

and form OH and TiH species, during which the Ti3+ concentration is increased due to the

81

donation of electrons from hydrogen atoms, making TiO2 a more n-type doped

82

semiconductor.15, 21-23 A key reaction, of unsurpassed interest, between hydrogen and

83

TiO2 is the reduction of protons by photoexcited electrons to form hydrogen molecules.24

84

Nevertheless, little attention has been paid to the band bending of TiO2 that modulates the

85

electron-hole pair separation efficiency and which may change during photochemical

86

reaction because charge transfer is involved.5

87

In this study, we use PL spectroscopy to report that the upward band bending of

88

TiO2 is decreased due to charge transfer incurred during the photoreduction of hydrogen

89

cations at a TiO2 nanopowder sample held at 90 K in a high vacuum chamber. High

90

vacuum is employed to eliminate possible band bending changes caused by adsorption of

91

unknown species that might otherwise occur in air or aqueous solutions. The low sample

92

temperature rules out the possibility that any atomic hydrogen present might be able to

93

thermally diffuse into the lattice and thereby change the band bending. Similar PL

94

behavior observed during photoreduction of gas phase helium cations suggests that the 4 ACS Paragon Plus Environment

Page 4 of 21

Page 5 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

95

band bending change is a general physical outcome that will apply to other photocatalytic

96

processes at semiconductor interfaces where charge transfer is involved.

5 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

97 98 99

EXPERIMENTAL Experiments were conducted with the equipment illustrated in Fig. 1, which has been described elsewhere in detail.5-6 Briefly, the stainless-steel vacuum chamber of ~1.5 L

100

volume was evacuated with a turbopump followed by a mechanical pump to achieve high

101

vacuum with a base pressure of 5×10-9 Torr. The chamber pressure was measured by a

102

cold cathode gauge (MKS series 423 I-MAG) which also serves as a gas ionization

103

source as discussed later. A CaF2 window allowed both UV and visible light transmission

104

into the chamber for PL measurements. To prepare the sample, TiO2 nanoparticles were

105

pressed onto a tungsten grid (W-grid) with square openings of 0.22 mm to produce a

106

circular disk with a diameter of 7 mm and a thickness of ~ 0.095 mm. The tungsten grid

107

was put close to the CaF2 window by clamping it to an oxygen-free high-conductivity

108

(OFHC) copper sample holder which was mounted to a feedthrough. The temperature of

109

the sample was measured by a K-type thermocouple spot-welded onto the W-grid at the

110

periphery of the sample. The sample temperature could be precisely controlled within 0.1

111

K in the range of 82 – 1000 K by adjusting the combination of electrical heating across

112

the W-grid and liquid-nitrogen cooling conducted through the feedthrough. Research

113

grade (≥99.99%) O2, H2, and He gas bulbs were connected to the gas line.

114

A gas dosing control module with a trap volume of 51.2 cm3, whose pressure was

115

continuously monitored by a Baratron pressure gauge, was used to control the number of

116

molecules dosed into the chamber (Fig. 1). The number of molecules dosed can be

117

calculated using the ideal gas law based on the gas pressure reading and trap volume. The

118

trapped gas pressure can be controlled with a precision of 0.001 Torr by adjusting the

119

pressure control valve shown in Fig. 1. A quadrupole mass spectrometer (SRS RGA 200)

120

was installed between the chamber and the turbo pump in order to analyze the purity of

121

gases and to detect molecules desorbed from the sample surface.

122

PL of TiO2 was measured by a PerkinElmer FL 55S fluorescence spectrometer

123

equipped with a pulsed Xe light source and a photomultiplier capable of detecting light

124

from 200 – 900 nm. The excitation of TiO2 sample was by 320-nm UV irradiation with a

125

continuous intensity of ~ 1.5 × 1014 cm-2 s-1, which was selected by gratings inside the

6 ACS Paragon Plus Environment

Page 6 of 21

Page 7 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

126

spectrometer and an additional 320-nm band-pass filter. A 390-nm cutoff filter was used

127

to reduce the reflected UV from interfering the collected PL signal.

128

The TiO2 used in this study is P25 (Evonik Industries, formerly Degussa), a

129

nanopowder that is widely used in research as a reference standard. The TiO2 samples

130

were heated in ~ 0.50 Torr O2 at 600 K for 30 min and then pumped out for 10 min

131

before each set of experiments to remove adsorbates from the TiO2 surfaces and to make

132

their properties consistent. To study the effects of gases, a selected amount of gas

133

molecules was dosed into the chamber and isolated from the turbopump by closing the

134

gate valve. To study the effect of atomic hydrogen, a tungsten filament connected to the

135

chamber through an elbow was heated to 2300 K, which was measured by an optical

136

pyrometer (Pyrometer Instrument Co.), to dissociate molecular hydrogen into hydrogen

137

atoms.25

138 139

Figure 1. Configuration of the high vacuum system employed in this study.

7 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

140

RESULTS AND DISCUSSION

141 142

Figure 2. Effects of hydrogen dose on the PL of TiO2. (A) PL spectra of TiO2 collected at

143

different states. (B) PL peak intensity change caused by H2 dose (red) and chamber

144

pressure measured by cold cathode gauge (purple).

145

PL of P25. The PL spectra of P25 TiO2 depicted in Fig. 2A are collected at the three

146

different states marked in Fig. 2B. As discussed in other reports,5 the 390 nm and 780 nm

147

peaks, which are invariant between different scans, are due to reflection from the sample

148

and CaF2 window, and therefore can be neglected. The PL emission of TiO2 ranges from

149

450 nm to 650 nm with a peak at ~530 nm and only intensity changes were observed

150

within this study. Although the P25 TiO2 contains both anatase (~75%, 85 nm dia.) and

151

rutile (~25%, 25 nm dia.) phases,26 our previous work has shown that the PL emission of

152

P25 under our conditions comes exclusively from the anatase phase.5 Therefore, the

153

results presented here should be valid for the surface of anatase TiO2.

8 ACS Paragon Plus Environment

Page 8 of 21

Page 9 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

154

Effect of H2 dosing on PL of TiO2. Fig 2B shows the effect of dosing H2 on the PL

155

peak intensity of TiO2. Before dosing H2, the chamber pressure increased from 10-8 to 10-

156

7

157

shows no impact on the PL trend. However, it is clearly shown that the dose of 7.1 × 1016

158

H2 molecules causes PL intensity to increase rapidly. The fast pressure drop after the H2

159

dose is mainly due to the adsorption of H2 onto the inner chamber walls experiencing

160

liquid N2 cooling, while the gradual drop of pressure is due to the equilibration of gas

161

consumption due to ionization by the cold cathode gauge, leakage through the gate valve,

162

as well gas adsorption/desorption on the chamber walls. A comparison between the

163

chamber pressure and the PL increase rate suggests that the PL increase rate is correlated

164

with the chamber H2 pressure.

165

Torr due to the isolation of the chamber from the vacuum pumps. This pressure increase

TiO2 is an n-type semiconductor and naturally develops upward band bending

166

underneath its surfaces4-5, which can be treated as due to trapped negative charges on its

167

surfaces for simplicity. The upward band bending of TiO2 can be reduced when electron

168

donor molecules are adsorbed on its surfaces because the donor molecules which donate

169

electrons into the TiO2 lattice possess positive charges on TiO2 surfaces and therefore

170

balance the originally trapped negative charges which causes the upward band bending.

171

Similarly, the adsorption of electron acceptor molecules increases the upward band

172

bending of TiO2.4, 6-7 It might be argued that the increase of PL upon H2 dose should be

173

attributed to H2 acting as electron donor at the TiO2 surfaces. However, various studies

174

have shown that molecular hydrogen interacts only weakly with both anatase and rutile

175

TiO2.16-17, 27-28 Lin et al.16 reported only van der Waals interactions between H2 molecules

176

and anatase TiO2 such that the extent of charge transfer between H2 and TiO2 is

177

negligible. Consequently, the observed increase of PL intensity caused by H2 dosing on

178

TiO2 may be more complicated.

9 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

179 180

Figure 3. The role of the CCG in PL intensity increase caused by hydrogen dosing.

181

Role of Cold Cathode Gauge (CCG) as ionizer. Fig. 3 shows evidence that the

182

increase of PL emission is not solely due to the introduction of hydrogen molecules.

183

Instead, the CCG, which measures the chamber pressure, also plays an important role. As

184

depicted, switching off the CCG at 1100 s shows no impact on PL of TiO2 in vacuum.

185

However, when the CCG is turned off, the effect that H2 dosing increases PL emission

186

from TiO2 shown in Fig. 2 is absent, as revealed by the dose of 1.7 × 1016 H2 molecules

187

at ~ 2800 s in Fig. 3. The effect begins immediately after the CCG is switched on. These

188

results indicate that the PL increase due to H2 dosing is caused by species formed during

189

the interaction between the CCG and H2 molecules, instead of molecular hydrogen,

190

which is consistent with reports that molecular H2 interacts only weakly with TiO2.16-17,

191

27-28

192

TiO2 or hydrogen doped TiO2 for which a processing temperature above 600 K is usually

193

required.18-19, 29 Fig. 3 also shows that the PL intensity does not flatten out immediately

194

after pump out but continues to increase with gradually decreasing slope. This behavior is

195

mainly because the TiO2 nanoparticle sample is porous and therefore gaseous species can

196

be trapped by the pores.30 The overall PL increase from ~ 5 to ~ 50 (arb. unit) in Fig. 3

197

indicates that a 10-fold difference in the photoexcited electron-hole radiative

198

recombination rate is caused by interaction between TiO2 and the active hydrogen species

199

produced by the CCG.

It is worth noting that our finding at 90 K does not conflict with reports of black

10 ACS Paragon Plus Environment

Page 10 of 21

Page 11 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

200

The Journal of Physical Chemistry

The CCG operates at a voltage of ~ 5 kV, wherein electrons extracted from the

201

cathode are accelerated towards the anode. When gas molecules are present, the

202

accelerated electrons can strike and ionize the gas molecules. The ions formed are

203

collected by the cathode and the ion current provides a measure of the pressure. Based on

204

the working mechanism of the CCG, it is reasonable to assume that the active species

205

causing the PL increase are hydrogen cations formed during the ionization of H2

206

molecules.27 The fact that the CCG used to measure the chamber pressure also serves as

207

an ionization source of H2 molecules points to the attention that should be paid to the

208

possibility that a measurement process may influence an experiment.

209

One possible but unlikely interference to the above postulation is that oxygen, which

210

may adsorb on the walls of the CCG during sample pretreatment in oxygen, can desorb

211

by electron-stimulation from the CCG and cause changes to the PL of TiO2. To rule out

212

this possibility, the data shown in Fig. 3 was pretreated by 0.500 Torr of H2 for 20

213

minutes after cleaning the sample in oxygen at 600 K to exchange the oxygen adsorbed

214

on the inner walls of the chamber. Due to the adsorption of hydrogen to the inner walls of

215

the chamber, the starting chamber pressure in Fig. 3 (> 10-8 Torr) is higher than that of

216

Fig. 2 (< 10-8 Torr). This also causes a slower pressure drop rate and a significantly

217

higher chamber pressure before pump out (~ 10-5 Torr) than Fig. 2 (< 10-6 Torr). This

218

pretreatment by H2 significantly decreases the possibility that the PL increase is caused

219

by electron-stimulated oxygen desorption from the CCG. In addition, the result shown in

220

Fig. S2 that oxygen desorption decreases PL intensity of TiO2 completely rules out the

221

possible effect by oxygen since oxygen works in the reverse way.

11 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

222 223

Figure 4. UV dependence of PL increase caused by interaction between CCG and H2

224

molecules.

225

UV induced photoreaction. Fig. 4 illustrates the influence of UV irradiation on the

226

PL increase caused by active species produced during the interaction between the CCG

227

and H2 molecules, which are most likely to be hydrogen cations as discussed above. The

228

green curve in Fig. 4 shows that there is no PL increase without UV irradiation, even

229

though the CCG is turned on and active species contributing to the PL increase can be

230

produced. Instead, the purple curve collected under UV irradiation shows normal PL

231

increase as observed before. The indispensability of UV irradiation for PL to increase

232

elucidated in Fig. 4 indicates that the PL increase is caused by reaction between

233

photoexcited charges and the active hydrogen species generated by the CCG, which is

234

most likely to be hydrogen cations.

235 12 ACS Paragon Plus Environment

Page 12 of 21

Page 13 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

236 237

The Journal of Physical Chemistry

Figure 5. Effect of atomic hydrogen on PL development with the CCG off Ruling out possible effects of atomic hydrogen. Although the active species causing

238

PL to increase is most likely to be H2+ based on the working mechanism of the CCG,

239

which agrees well with the pressure dependent measurements shown in Fig S1, it is also

240

possible that atomic hydrogen (H·) may be produced by collision or ionization of H2

241

molecules by high energy electrons, or by collision of H2+ with H2 molecules.31-32 The

242

latter two processes allow for the production of several kinds of Hn+ (n=1, 2 or 3) cations.

243

It can be deduced that atomic hydrogen adsorption onto TiO2 should increase PL

244

intensity based on reports from the literature. Although it was reported that atomic

245

hydrogen can diffuse into TiO2 lattices, 21-22 due to its small diffusion coefficient at low

246

temperature,15, 21-22 it is reasonable to assume that hydrogen atoms adsorbed onto TiO2 at

247

90 K stay mainly on the surface. It was found that atomic hydrogen donates delocalized

248

electrons into the TiO2 lattice and thus atomic hydrogen acts as an electron donor.23

249

Consequently, atomic hydrogen adsorbed on TiO2 surfaces should be positively charged

250

after donating electronic charge into the TiO2 lattice, which will decrease the upward

251

band bending of TiO2 and increase PL. Hence, it is necessary to differentiate whether the

252

PL increase due to molecular hydrogen gas exposure with CCG on is caused by hydrogen

253

cations or atomic hydrogen.

254

Fig. 5 illustrates the effect of atomic hydrogen, which was introduced to the TiO2

255

sample by dissociating H2 through a hot tungsten filament at 2300 K,25 on the PL of TiO2.

256

The finding that atomic hydrogen exposure increases PL is consistent with the

257

expectations expressed above. But, unlike the effect observed in Fig. 4, atomic hydrogen

258

not only causes PL to increase under UV irradiation, but also causes PL to increase

259

without UV irradiation. The difference in the PL behavior between dosing atomic

260

hydrogen and the active species produced by molecular hydrogen with the CCG on

261

indicates that the active species produced by the CCG is unlikely to be atomic hydrogen.

262

Hence, we propose that the active species that causes the PL to increase during molecular

263

hydrogen dosing are hydrogen cations (Hn+, with n most likely 2 33) which are produced

264

by the ionization of hydrogen molecules by the CCG.

13 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

265 266

Figure 6. Schematic of processes occurring on TiO2 surface. The energy level diagrams

267

are for electrons where the direction of increasing energy is reversed for holes.

268

Band bending and charge recombination rate change model. A proposed

269

mechanism explaining the PL increase caused by hydrogen cations under UV irradiation

270

is illustrated in Fig. 6. Hydrogen cations are produced during the measurement of

271

chamber pressure through the CCG which ionizes hydrogen molecules. Under UV

272

irradiation photoexcited electron-hole pairs are generated in TiO2. Because the

273

photoexcited electrons are strong reducing agents, hydrogen cations that arrive at TiO2

274

surfaces are reduced by the photoexcited electrons (process 1). The photoexcited holes

275

left in the valence band tends to migrate toward the surface of TiO2 (process 2) under the

276

electric field in the band bending region (depletion layer). The accumulation of holes at

277

the surface of TiO2 balances the originally trapped negative charges under TiO2 surfaces

278

and thus causes the band bending to be less upward. The reduction of upward band

279

bending decreases the volume of the band bending (depletion) region and accordingly the

280

radiative recombination rate of photoexcited charge carriers increases as monitored by

281

the PL intensity. The overall photoreaction results in positive charge transfer from

282

hydrogen cations to TiO2 nanoparticles, making TiO2 particles positively charged.

283

Because the 320-nm UV penetration depth into TiO2 is only ~ 25 nm,14 the observed PL

284

only comes from the first layer of TiO2 particles, which are insulated from the tungsten

285

grid by thousands of layers of TiO2 particles. Hence, it is reasonable that the observed top

286

layer TiO2 can gain some excess positive charges without quasi-instantaneous discharge.

287

In addition, the fact that hydrogen cations cause no change to the PL of TiO2 in dark

288

without the assistance of photoexcited electrons suggests that the positive charge held by 14 ACS Paragon Plus Environment

Page 14 of 21

Page 15 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

289

the hydrogen cations cannot be directly injected into the valence band of TiO2 (process 3,

290

which cannot occur spontaneously) because the photoexcited holes are strong oxidants

291

and occupy an energy level higher than that of hydrogen cations, which can be inferred

292

from other studies in which H2 acts as hole scavenger.15

293 294

Figure 7. PL intensity of TiO2 decay after turning off UV irradiation (other figures are

295

with continuous UV irradiation except this one).

296

Quantitative analysis Although reaction between photoexcited electrons and

297

hydrogen cations will increase PL because of the associated band bending change as

298

discussed above, it is worth considering if consumption of photoexcited electrons by

299

reaction could substantially quench the PL intensity. As detailed below, the latter effect is

300

negligible. The number of photons incident on each TiO2 particle at the surface during

301

the period from 3500 s to 11000 s in Fig. 3, wherein the PL changes from ~ 4 to ~ 35

302

intensity units, is about 6.3 × 107 photons with the assumption that the particle size of the

303

PL-active, anatase particles in P25 is ~ 85 nm.26 Essentially all of the incident UV

304

photons are absorbed to produce e/h pair excitations given the 25 nm UV absorption

305

length of TiO2. According to the dead-layer PL model we developed earlier,5 it only

306

requires at most ~ 150 leftover holes per TiO2 particle to reduce the upward band bending

307

sufficiently to recover 88% of the maximum PL (appropriate for a flat band condition)

308

given a TiO2 volume defect density of 1018 cm-3. Therefore, in this study, the ratio of the

309

number of photoexcited electrons reacting with hydrogen cations to leave leftover holes 15 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

310

to the number of incident photons absorbed is at most ~ 0.00024%, which is negligible.

311

Non-radiative and radiative (ca. 0.25% quantum efficiency)34 e/h pair recombination

312

dominate the photoexcitation relaxation. Therefore, the side-effect of PL quenching

313

caused by reaction between photoexcited electron and hydrogen cations via consumption

314

of photoexcited electrons available for recombination, is negligible in this study.

315

If the number of photoexcited electron-hole pairs is much larger than the number

316

of inhomogeneously positioned charges establishing the band bending potential gradient,

317

a possible concern is whether the band bending could be electrostatically scrambled by

318

charge redistribution of the photoexcited electrons and holes over the TiO2 particle.

319

However, the short life time of photoexcited electron-hole pairs observed as ~ 17 µs in

320

Fig. 7 indicates that there are only ~ 0.15 photoexcited electron-hole pairs per TiO2

321

particle on average at any time during continuous UV irradiation as in Fig. 3. Sufficient

322

upward band bending to quench the PL to 12% of its maximal (flat band) value would be

323

established with 150 immobilized electrons at the surface and the counter positive

324

charges distributed across defects in the depletion zone at a defect density of 1018 cm-3.

325

Consequently, modulation of the TiO2 band bending by the steady-state photoexcited

326

charge density in our experiments should be negligible.

327 328

Figure 8. Effect of helium on PL intensity of TiO2. Influence of CCG on PL development.

329

Because the decrease in upward band bending, signaled by a PL increase, is

330

caused by positive charge transfer from the hydrogen cations into TiO2 through 16 ACS Paragon Plus Environment

Page 16 of 21

Page 17 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

331

photoinduced reduction, it is plausible that photoinduced reduction of other cations

332

should also induce a similar band bending change. Fig. 8 shows similar PL intensity

333

changes caused by exposure to helium gas with the CCG off or on. As depicted in Fig. 8,

334

neutral helium atoms have no impact on PL of TiO2 with the CCG off. But the PL

335

increases when the CCG is turned on and ions are formed. The finding that helium cation

336

exposure also causes PL of TiO2 to increase lends further support to the idea that the

337

active species in the hydrogen dosing PL experiments are hydrogen cations, rather than

338

hydrogen atoms, and that a band bending decrease is a general physical consequence of

339

cation photoreduction.

340

These unusual PL observations, enabled by low temperature, high vacuum, and a

341

TiO2 nanopowder sample not subject to prompt charge neutralization in its environment,

342

show that charge transfer during a photoelectrochemical reaction at a semiconductor

343

interface alters the photoexcited electron-hole pair recombination rate and surface band

344

bending in ways that are likely to impact the efficiency of photocatalytic devices.

345

CONCLUSION

346

By employing PL spectroscopy to monitor the photoexcited electron-hole radiative

347

recombination and the dead-layer model to analyze the band bending of TiO2, it was

348

found that the reduction of hydrogen cations by photoexcited electrons, which is a half

349

photoelectrochemical reaction achieved under high vacuum conditions, alters the surface

350

band bending and charge recombination rate of TiO2. Main findings include:

351

(1) Molecular hydrogen exposure had no effect on the band bending of TiO2 at 90 K.

352

(2) Molecular hydrogen could be ionized into hydrogen cations by the cold cathode

353

gauge measuring the chamber pressure and these cations could reach the TiO2

354

sample.

355

(3) The reduction of hydrogen cations by photoexcited electrons decreases the

356

upward band bending of TiO2, signaled by increased PL intensity (electron-hole

357

radiative recombination rate), because of the accumulation of positively charged

358

holes at TiO2 surfaces.

17 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

359

(4) The band bending and charge carrier recombination rate changes observed for

360

TiO2 due to photoreduction of both hydrogen and helium cations are physical

361

consequences of cation photoreduction at a semiconductor interface.

362

ASSOCIATED CONTENT

363

Supporting Information

364

Dependence of PL increase rate on H2 pressure (Fig. S1) and mean free path of H2+ in H2

365

(Table S1). The different effect of O2 (Fig. S2)

366

AUTHOR INFORMATION

367

Corresponding Authors

368

Ian Harrison: Email: [email protected]. Telephone: +1(434)924-3639.

369

ORCID

370

Shiliang Ma: 0000-0003-1478-2771

371

Ian Harrison: 0000-0002-3103-6247

372

Notes

373

The authors declare no competing financial interest

374

ACKNOWLEDGEMENTS

375

We gratefully acknowledge support of this work by the Department of Energy under

376

grant number DE-SC0002365 and the Army Research Office under grant number

377

W911NF-13-1-0101, as well AES fellowships for Shiliang Ma from AES Corp. and the

378

University of Virginia. Z. Zhang acknowledges support by NSFC (No. 21773047).

379

18 ACS Paragon Plus Environment

Page 18 of 21

Page 19 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

380

REFERENCES

381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424

1. Ni, M.; Leung, M. K. H.; Leung, D. Y. C.; Sumathy, K. A Review and Recent Developments in Photocatalytic Water-Splitting Using TiO2 for Hydrogen Production. Renewable Sustainable Energy Rev. 2007, 11, 401-425. 2. Hoffmann, M. R.; Martin, S. T.; Choi, W. Y.; Bahnemann, D. W. Environmental Applications of Semiconductor Photocatalysis. Chem. Rev. 1995, 95, 69-96. 3. Yan, Y.; Shi, W. D.; Yuan, Z.; He, S. G.; Li, D. M.; Meng, Q. B.; Ji, H. W.; Chen, C. C.; Ma, W. H.; Zhae, J. C. The Formation of Ti-H Species at Interface Is Lethal to the Efficiency of TiO2-Based Dye-Sensitized Devices. J. Am. Chem. Soc. 2017, 139, 20832089. 4. Zhang, Z.; Yates, J. T. Band Bending in Semiconductors: Chemical and Physical Consequences at Surfaces and Interfaces. Chem. Rev. 2012, 112, 5520-5551. 5. Ma, S. L.; Reish, M. E.; Zhang, Z.; Harrison, I.; Yates, J. T. Anatase-Selective Photoluminescence Spectroscopy of P25 TiO2 Nanoparticles: Different Effects of Oxygen Adsorption on the Band Bending of Anatase. J. Phys. Chem. C 2017, 121, 12631271. 6. Stevanovic, A.; Buttner, M.; Zhang, Z.; Yates, J. T. Photoluminescence of TiO2: Effect of UV Light and Adsorbed Molecules on Surface Band Structure. J. Am. Chem. Soc. 2012, 134, 324-332. 7. Zhang, Z.; Yates, J. T. Effect of Adsorbed Donor and Acceptor Molecules on Electron Stimulated Desorption: O2/TiO2(110). J. Phys. Chem. Lett. 2010, 1, 2185-2188. 8. Zhang, Z.; Yates, J. T. Electron-Mediated CO Oxidation on the TiO2(110) Surface During Electronic Excitation. J. Am. Chem. Soc. 2010, 132, 12804-12807. 9. Sheppard, L. R.; Dittrich, T.; Nowotny, M. K. The Impact of Niobium Surface Segregation on Charge Separation in Niobium-Doped Titanium Dioxide. J. Phys. Chem. C 2012, 116, 20923-20929. 10. Karakitsou, K. E.; Verykios, X. E. Effects of Altervalent Cation Doping of TiO2 on Its Performance as a Photocatalyst for Water Cleavage. J. Phys. Chem. 1993, 97, 1184-1189. 11. Hollingsworth, R. E.; Sites, J. R. Photo-Luminescence Dead Layer in P-Type InP. J. Appl. Phys. 1982, 53, 5357-5358. 12. Meyer, G. J.; Lisensky, G. C.; Ellis, A. B. Evidence for Adduct Formation at the Semiconductor Gas Interface - Photoluminescent Properties of Cadmium Selenide in the Presence of Amines. J. Am. Chem. Soc. 1988, 110, 4914-4918. 13. Anpo, M.; Tomonari, M.; Fox, M. A. Insitu Photoluminescence of TiO2 as a Probe of Photocatalytic Reactions. J. Phys. Chem. 1989, 93, 7300-7302. 14. Ghosh, A. K.; Maruska, H. P. Photoelectrolysis of Water in Sunlight with Sensitized Semiconductor Electrodes. J. Electrochem. Soc. 1977, 124, 1516-1522. 15. Berger, T.; Diwald, O.; Knozinger, E.; Napoli, F.; Chiesa, M.; Giamello, E. Hydrogen Activation at TiO2 Anatase Nanocrystals. Chem. Phys. 2007, 339, 138-145. 16. Lin, F.; Zhou, G.; Li, Z. Y.; Li, J.; Wu, J.; Duan, W. H. Molecular and Atomic Adsorption of Hydrogen on TiO2 Nanotubes: An Ab Initio Study. Chem. Phys. Lett. 2009, 475, 82-85. 17. Kunat, M.; Burghaus, U.; Woll, C. The Adsorption of Hydrogen on the Rutile TiO2(110) Surface. Phys. Chem. Chem. Phys. 2004, 6, 4203-4207. 19 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466

18. Chen, X. B.; Liu, L.; Yu, P. Y.; Mao, S. S. Increasing Solar Absorption for Photocatalysis with Black Hydrogenated Titanium Dioxide Nanocrystals. Science 2011, 331, 746-750. 19. Wang, B.; Shen, S. H.; Mao, S. S. Black TiO2 for Solar Hydrogen Conversion. J. Materiomics 2017, 3, 96-111. 20. Lavrov, E. V., Hydrogen Donor in Anatase TiO2. Phys Rev B 2016, 93, 045204. 21. Tao, J. G.; Cuan, Q.; Gong, X. Q.; Batzill, M. Diffusion and Reaction of Hydrogen on Rutile TiO2(011)-2x1: The Role of Surface Structure. J. Phys. Chem. C 2012, 116, 20438-20446. 22. Aschauer, U.; Selloni, A. Hydrogen Interaction with the Anatase TiO2(101) Surface. Phys. Chem. Chem. Phys. 2012, 14, 16595-16602. 23. Panayotov, D. A.; Yates, J. T. N-Type Doping of TiO2 with Atomic HydrogenObservation of the Production of Conduction Band Electrons by Infrared Spectroscopy. Chem. Phys. Lett. 2007, 436, 204-208. 24. Chen, X. B.; Shen, S. H.; Guo, L. J.; Mao, S. S. Semiconductor-Based Photocatalytic Hydrogen Generation. Chem. Rev. 2010, 110, 6503-6570. 25. Zheng, W. G.; Gallagher, A. Hydrogen Dissociation on High-Temperature Tungsten. Surf. Sci. 2006, 600, 2207-2213. 26. Ohno, T.; Sarukawa, K.; Tokieda, K.; Matsumura, M. Morphology of a TiO2 Photocatalyst (Degussa, P-25) Consisting of Anatase and Rutile Crystalline Phases. J. Catal. 2001, 203, 82-86. 27. Yang, Y.; Sushchikh, M.; Mills, G.; Metiu, H.; McFarland, E. Reactivity of TiO2 with Hydrogen and Deuterium. Appl. Surf. Sci. 2004, 229, 346-351. 28. Fukada, K.; Matsumoto, M.; Takeyasu, K.; Ogura, S.; Fukutani, K. Effects of Hydrogen on the Electronic State and Electric Conductivity of the Rutile TiO2(110) Surface. J. Phys. Soc. Jpn. 2015, 84, 064716. 29. An, H. R.; Park, S. Y.; Kim, H.; Lee, C. Y.; Choi, S.; Lee, S. C.; Seo, S.; Park, E. C.; Oh, Y. K.; Song, C. S. et al. Advanced Nanoporous TiO2 Photocatalysts by Hydrogen Plasma for Efficient Solar-Light Photocatalytic Application. Sci. Rep. 2016, 6, 29683 30. Stevanovic, A.; Yates, J. T. Probe of NH3 and CO Adsorption on the Very Outermost Surface of a Porous TiO2 Adsorbent Using Photoluminescence Spectroscopy. Langmuir 2012, 28, 5652-5659. 31. Padovani, M.; Galli, D.; Glassgold, A. E. Cosmic-Ray Ionization of Molecular Clouds. Astron. Astrophys. 2009, 501, 619-631. 32. Wobschall, D.; Fluegge, R. A.; Graham, J. R. Collision Cross Sections of Hydrogen and Other Ions as Determined by Ion Cyclotron Resonance. J. Chem. Phys. 1967, 47, 4091. 33. Bauer, N.; Beach, J. Y. Differences in Mass Spectra of H2 and D2. J. Chem. Phys. 1947, 15, 150-151. 34. Abazovic, N. D.; Comor, M. I.; Dramicanin, M. D.; Jovanovic, D. J.; Ahrenkiel, S. P.; Nedeljkovic, J. M. Photoluminescence of Anatase and Rutile TiO2 Particles. J Phys Chem B 2006, 110, 25366-25370.

467 468 20 ACS Paragon Plus Environment

Page 20 of 21

Page 21 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

469

The Journal of Physical Chemistry

TOC Graphic

470

21 ACS Paragon Plus Environment