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