Modulation of the Reduction Potential of TiO2- x by Fluorination for Efficient and Selective CH4 Generation from CO2 Photoreduction

Article abstract of DOI:10.1021/acs.nanolett.8b00197

Photocatalytic reduction of CO₂ holds great promises for addressing both the environmental and energy issues that are facing the modern society. The major challenge of CO₂ photoreduction into fuels such as methane or methanol is the

Full text of DOI:10.1021/acs.nanolett.8b00197

Subscriber access provided by Kaohsiung Medical University
Modulation of the Reduction Potential of TiO2-x by Fluorination for
Efficient and Selective CH4 Generation from CO2 Photoreduction
Mingyang Xing, Yi Zhou, Chunyang Dong, Lejuan Cai, Lixi Zeng, Bin Shen,
Lihan Pan, Chencheng Dong, Yang Chai, Jinlong Zhang, and Yadong Yin
Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b00197 • Publication Date (Web): 27 Apr 2018
Downloaded from http://pubs.acs.org on April 27, 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 24
Nano Letters
1
2
3
4
5
6
7
8
9
Modulation of the Reduction Potential of TiO2ꢀx by
Fluorination for Efficient and Selective CH4
Generation from CO2 Photoreduction
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
†
†
§
†
†
Mingyang Xing,*,†,‡ Yi Zhou, Chunyang Dong, Lejuan Cai, Lixi Zeng, Bin Shen, Lihan Pan,
∥
Chencheng Dong,† Yang Chai,§ Jinlong Zhang,*,† and Yadong Yin*,‡
†Key Laboratory for Advanced Materials and Institute of Fine Chemicals, School of Chemistry
and Molecular Engineering, East China University of Science and Technology, 130 Meilong
Road, Shanghai 200237, P.R. China
‡Department of Chemistry, University of California Riverside, CA 92521, USA
§Department of Applied Physics, The Hong Kong Polytechnic University, Hung Hom, Kowloon,
Hong Kong, P. R. China
∥
School of Environment, Guangzhou Key Laboratory of Environmental Exposure and Health,
and Guangdong Key Laboratory of Environmental Pollution and Health, Jinan University,
Guangzhou 510632, P. R. China
AUTHOR INFORMATION
Corresponding Author
*Email: mingyangxing@ecust.edu.cn
ACS Paragon Plus Environment
1

Nano Letters
Page 2 of 24
1
2
3
4
5
6
7
8
9
*Email: jlzhang@ecust.edu.cn
*Email: yadong.yin@ucr.edu
ORCID
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
Mingyang Xing: 0000ꢀ0002ꢀ0518ꢀ2849
Jinlong Zhang: 0000ꢀ0002ꢀ1334ꢀ6436
Yadong Yin: 0000ꢀ0003ꢀ0218ꢀ3042
Author Contributions
Mingyang Xing, Yi Zhou and Chunyang Dong contributed equally to this work.
ABSTRACT: Photocatalytic reduction of CO2 holds great promises for addressing both the
environmental and energy issues that are facing the modern society. The major challenge of CO2
photoreduction into fuels such as methane or methanol is the low yield and poor selectivity.
Here, we report an effective strategy to enhance the reduction potential of photoꢀexcited
electrons by fluorination of mesoporous single crystals (MSCs) of reduced TiO2ꢀx. Density
functional theory (DFT) calculations and photoelectricity tests indicate that the Ti3+ impurity
level is upswept by fluorination, owing to the builtꢀin electric field constructed by the
substitutional F that replaces surface oxygen vacancies, which leads to the enhanced reduction
potential of photoꢀexcited electrons. As a result, the fluorination of the reduced TiO2ꢀx
dramatically increases the CH4 production yield by 13 times from 0.125 µmol/g·h to 1.63
µmol/g·h under solar light illumination, with the CH4 selectivity being improved from 25.7% to
ACS Paragon Plus Environment
2

Page 3 of 24
Nano Letters
1
2
3
4
5
6
85.8%. Our finding provides a metalꢀfree strategy for the selective CH4 generation from CO2
photoreduction.
7
8
9
KEYWORDS: Photocatalysis, semiconductors, mesoporous single crystal, fluorination, CO2
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
photoreduction
ACS Paragon Plus Environment
3

Nano Letters
Page 4 of 24
1
2
3
4
5
6
7
8
9
Photoreduction of CO2 into desirable chemicals such as methane or methanol has been widely
considered as one of the most promising approaches to reduce the excessive emission of
greenhouse gases and produce renewable energy that can replace the traditional fossil
feedstocks.1ꢀ5 Currently, much research effort on CO2 photoreduction has been focused on
photocatalysts based on TiO2,6ꢀ13 owing to its high photocatalytic activity and suitable redox
potential for CO2 reduction and H2O oxidation.14 It was found recently that the bandgap of TiO2
could be narrowed to better utilize the solar energy spectrum by employing reduced TiO2 (TiO2ꢀ
x) which are selfꢀdoped with Ti3+ and oxygen vacancies.15, 16 Although selfꢀdoping of Ti3+
narrows the bandgap of TiO2, it also weakens the reduction power of electrons, which greatly
slows the kinetics of the reduction reaction from CO2 to CH4 which involves eight electrons.14, 17
It is therefore highly desirable to develop new TiO2ꢀx photocatalysts with enhanced reduction
potential while maintaining the visible light response. While this represents a significant
challenge in the field, there has been some positive progress recently. For example, Zhang et al.
demonstrated that the presence of an electron acceptor such as Cl2 or O2 on the TiO2 surface
could cause upward band bending due to the formation of Helmholtz layer on the semiconductor
surface;18, 19 and Chen et al. found that the point defects such as Fꢀtype color centers in rutile can
also induce lattice relaxation and ionic polarization effects,20 which could induce the upward
band bending.21
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
Here, we report that significant improvement of the reduction potential of TiO2ꢀx
mesoporous single crystals (MSCs) could be achieved through a robust fluorination process,
while maintaining their visible light response. As indicated by density functional theory (DFT)
calculations, fluorination enables doping of F atoms in the lattice interval of TiO2ꢀx by
substituting oxygen vacancies. With its strong electronꢀwithdrawing property, fluorine atoms
ACS Paragon Plus Environment
4

Page 5 of 24
Nano Letters
1
2
3
4
5
6
7
8
9
attract electrons from neighboring Ti3+,18, 19 regenerating Ti4+ which carries more positive
charges. As a result, a builtꢀin electric field is generated between substitutional F and Ti3+,
leading to the upsweep of the Ti3+ impurity level, enhanced reduction potential of the
electrons,20, 21 and consequently significant improvements in the yield and selectivity of CH4
production in solarꢀlightꢀdriven CO2 photoreduction.
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 rutile TiO2ꢀx MSCs were prepared by using a hardꢀtemplate method, according to our
previous report.22 The asꢀprepared MSCs displayed threeꢀdimensional (3D) connected channels
and obvious blue color (Figure 1a and S1), implying the generation of Ti3+ in MSCs, which was
further confirmed by the low temperature EPR results shown in Figure 1b.23 Interestingly, after
fluorination by NH4F, the resulting FꢀMSCs samples showed almost unchanged crystal form
(Figure S2 and S3), color appearance (Figure 1a), and visible light absorption (Figure 1c)
compared with MSCs. However, the absorption shoulder of FꢀMSCs showed a significant blueꢀ
shift (Figure 1c). And the bandgap of MSCs also has a distinct increase from 3.01 eV to 3.10 eV
after the fluorination (inset of Figure 1c), implying that the Ti3+ impurity level in TiO2ꢀx has an
upward shift. Meanwhile, the Ti3+ signals of FꢀMSCs in the EPR spectra presented an obvious
decrease, further confirming the change of Ti3+ impurity level. The color of FꢀMSCs remained
very stable even upon air exposure for six months (Figure 1a).
ACS Paragon Plus Environment
5

Nano Letters
Page 6 of 24
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
Figure 1. (a) Photos of various MSCs samples. (b) EPR spectra of the samples on 77 K. (c) UV
visꢀDRS spectra and transformed Kubelka–Munk function against the photon energy plot (inset)
of the samples. (d) FESEM image of FꢀMSCs. (e) TEM image of FꢀMSCs. (f) Elemental
mapping images of FꢀMSCs, with red dots denoting element Ti, blue dots element O, and green
dots element F.
o
The TiO2ꢀx MSCs were calcined at 500 C in air for 30 min to produce a reference sample
(CalꢀMSCs), which changed its color to white (Figure 1a). In addition, the EPR signal of Ti3+
disappeared (Figure 1b), owing to the reꢀoxidation of Ti3+ in MSCs. As expected, the bandgap of
CalꢀMSCs increased to 3.12 eV, which was close to the value of FꢀMSCs (3.10 eV) but with a
white color appearance. We also fluorinated the CalꢀMSCs to produce a sample of FꢀCalꢀMSCs,
ACS Paragon Plus Environment
6

Page 7 of 24
Nano Letters
1
2
3
4
5
6
7
8
9
which however did not display any blueꢀshift of the light adsorption shoulder (Figure 1c). Its
bandgap (3.11 eV) also has no obvious increase compared with that of CalꢀMSCs (3.12 eV),
indicating that the conduction band (CB) of CalꢀMSCs cannot be elevated by fluorination, owing
to the absence of Ti3+ in CalꢀMSCs and FꢀCalꢀMSCs.21 XRD patterns of CalꢀMSCs and FꢀCalꢀ
MSCs are shown in Figure S2a to confirm that the phase of prepared samples remains the same
even after calcination. Many reports have demonstrated that doping with Ti3+ or oxygen vacancy
would induce the Fꢀtype color centers in a well crystallized TiO2 surface layer, which could
obviously enhance the visible light response.20, 24ꢀ28 The FꢀMSCs here could maintain the visible
light absorption while enhancing the reduction potential of the electrons. On the other hand, the
morphological structures of MSCs changed very little after fluorination, as shown in Figure 1d,
e. And the elemental mapping results in Figure 1f demonstrate the fluorine atoms have been
evenly dispersed in TiO2ꢀx.
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
ACS Paragon Plus Environment
7

Nano Letters
Page 8 of 24
1
2
3
4
5
6
7
8
Figure 2. (a) Valence band (VB) XPS spectra of MSCs before and after fluorination; (b) F1s
XPS spectra of blue MSCs before and after fluorination; and Mott–Schottky plots of (c) MSCs
and (d) FꢀMSCs.
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 XPS characterization and the MottꢀSchottky plots were used to investigate the change of
valence band (VB) and CB in FꢀMSCs (Figure 2), respectively. It is found that the fluorination
does not contribute to the additional valence bandꢀedge states of the Ti3+ doped MSCs, and the
VB of catalyst is fixed at 2.38 eV (Figure 2a). MottꢀSchottky plots give an answer for the
broadened bandgap of FꢀMSCs, that is, the CB of MSCs has elevated from ꢀ0.63 eV to ꢀ0.71 eV
(Figure 2c,d), which implies the strengthened reduction potential of TiO2ꢀx by fluorination. F 1s
XPS results indicate there is a new peak appearing at 684.6 eV after the fluorination (Figure 2b).
The peak at 683.7 eV is ascribed to the surface TiꢀF.8, 29 The new peak at a higher binding energy
is assigned to the substitutional F that replaces oxygen vacancy (as confirmed by the following
DFT calculations), which is caused by NH4Fꢀfluorination. A weak peak at 400.0 eV in the N 1s
XPS suggest a very low possibility of nitrogen doping in the TiO2ꢀx (Figure S4).29ꢀ31 In our
previous work,29 the generation of Ti3+ and the doping of F occurred at the same time by the oneꢀ
step vacuum activation method. As a result, there is a strong F1s XPS peak at 689.8 eV, which is
ascribed to the doped F in the lattice of TiO2. There is no XPS peak observed between the
surface ≡Ti–F species and doped F. More importantly, there is no blueꢀshift in light absorption
after the modification of TiO2 in the previous work. But in this case, the Ti3+ was introduced into
TiO2 firstly, and then the fluorination was carried out over the Ti3+ doped TiO2. A new F 1s XPS
peak appearing at 684.6 eV was observed after the fluorination, which is located between the
peaks of surface ≡Ti–F and doped F. Interestingly, the fluorination induces an obvious blueꢀshift
ACS Paragon Plus Environment
8

Page 9 of 24
Nano Letters
1
2
3
4
5
6
7
8
9
of light absorption. Hence, we can conclude that the fluorination over Ti3+ꢀTiO2 reported here is
very different from the conventional fluorination modification of TiO2, which could cause
changes in the band structure of Ti3+ꢀTiO2. On the other hand, when the Ti3+ was eliminated by
calcination (CalꢀMSCs), the subsequent fluorination treatment (FꢀCalꢀMSCs) would result
in no change of importance in the CB and VB locations (Figure S5). In fact, the tests of UVꢀDRS
and Mott–Schottky analysis were repeated for several times using samples of different batches to
eliminate the measurement error (Figure S6). These results indicate the presence of interaction
between Ti3+ and substitutional F in FꢀMSCs, which plays a key role in the elevation of CB.
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
ACS Paragon Plus Environment
9

Nano Letters
Page 10 of 24
1
2
3
4
5
6
7
8
9
Figure 3. (a) Calculated structures for rutile TiO2ꢀx (110) before (left) and after (right)
fluorination. For the construction of structure model, the concentration of Ti3+ before and after
fluorination was fixed at 3.7% and 2.7% respectively, according to the EPR and XPS results.
And the substitution fluorine concentration was fixed at 4.1% based on the XPS results. (bꢀd)
Calculated DOS of clean rutile TiO2, TiO2ꢀx, and FꢀTiO2ꢀx, and (eꢀg) the corresponding
amplification DOS (the magnetizations of TiO2, TiO2ꢀx, and FꢀTiO2ꢀx are 0, 4.0 and 3.0
respectively).
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
In order to investigate the interaction between substitutional F and Ti3+, we employed the
DFT calculations to study the effect of fluorination on the energy band structure of TiO2ꢀx. Some
previous DFT calculations have shown that fluorine doping could reduce the VB location of
TiO2,32, 33 which is very different from our results, probably because most of these studies
ignored the impact of the defects such as the oxygen vacancies in TiO2ꢀx. In our case, we
considered the oxygen vacancies (surface and bulk) while building the rutile (110) structure, as
shown in Figure 3a. The concentration of Ti3+ and F in the model was consistent with the actual
quantity obtained from the XPS and EPR results. Interestingly, when the fluorine atoms occupy
the position of surface oxygen vacancies, the corresponding results of DFT calculation are very
consistent with the experimental results. Compared with clean TiO2, the DOS of TiO2ꢀx shows an
obvious impurity level under the CB location (Figure 3e, f), which is induced by the introduction
of Ti3+. After the fluorination, the impurity level of FꢀTiO2ꢀx has a distinct upperꢀshift, as shown
in Figure 3f, g. Meanwhile, the magnetization of TiO2ꢀx decreases from 4.0 to 3.0 after the
fluorination, suggesting the decreased spin electrons and reduced Ti3+ concentration. After the
substitution of F for oxygen vacancy, an internal electric field was established between the
ACS Paragon Plus Environment
10

Page 11 of 24
Nano Letters
1
2
3
4
5
6
7
8
9
neighboring Ti3+ and substitutional F. The electron on Ti3+ would transfer to F to generate the
Ti4+ owing to the strong electronegativity of F atoms, which is responsible for the decrease of
Ti3+ concentration after the fluorination. In addition, we also have considered the effect of bulk F
(on the subsurface of TiO2, below 5 nm depth) for the construction of DFT model. As shown in
Figure S7, after the fluorination, the impurity level of FꢀTiO2ꢀx still exhibits a distinct upperꢀshift.
The result indicates that the bulk F plays an ignorable role for the upperꢀshift of Ti3+ levels in Fꢀ
MSCs. Hence, the DFT calculation results suggest that the F atoms for the substitution of oxygen
vacancies in the TiO2ꢀx can induce the elevation of impurity levels and the decrease of Ti3+,
which is very consistent with the Mott–Schottky plots and EPR results in Figure 2c, d and 1b.
Theoretically, there are two possibilities for the elevation of impurity levels in FꢀMSCs: one
is the overall upward mobility and the other the upsweep. The impedance, photocurrent and
photoluminescence (PL) emission spectra were measured to eliminate the possibility of overall
upward mobility of impurity levels, as seen in Figure S8. Otherwise, the conductivity, PL
intensity and photocurrent intensity should display an obvious reduction, due to the widening of
bandgap of TiO2ꢀx. In our case, the impedance and conductivity of FꢀMSCs showed no sign of
waning (Figure S8a), indicating efficient electron transfer in the single crystal. Interestingly, the
photocurrent intensity of FꢀMSCs was higher than that of MSCs (Figure S8c). Theoretically,
there are two possibilities for the enhancement of photocurrent intensity over FꢀMSCs: one is the
improvement of separation efficiency of photoꢀgenerated carriers, and the other is the increase of
the kinetic overꢀpotential (ꢁE) of electrons. However, the unchanged PL intensity of FꢀMSCs
indicated that this enhanced photocurrent intensity was not induced by the improvement of
electronꢀhole separation efficiency (Figure S8b). Hence, we can infer that the enhanced
photocurrent intensity of FꢀMSCs is induced by the increase in ꢁE, which further allows us to
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
ACS Paragon Plus Environment
11

Nano Letters
Page 12 of 24
1
2
3
4
5
6
7
8
9
conclude that the Ti3+ impurity level in FꢀMSCs is upswept by the fluorination. The enhanced
photocurrent intensity of FꢀMSCs indicated that the FꢀTiO2ꢀx has much more electrons escaping
from the impurity levels than the TiO2ꢀx, owing to the presence of polarization effect. However,
once we added a 10 mV bias voltage during the photocurrent measurement, the photocurrent
intensity of TiO2ꢀx changed to close to that of FꢀTiO2ꢀx (Figure S8d). Under the same external
voltage, the bending band of MSCs turned to be more flat than that of FꢀMSCs, owing to the
stronger band bending degree of FꢀMSCs than TiO2ꢀx. Hence, there were many electrons
escaping from the bending band over TiO2ꢀx under an external voltage, whose current intensity
equaled to the intensity of FꢀTiO2ꢀx. This result further confirmed the existence of strong bending
band in FꢀMSCs. Additionally, the very similar photocurrent between CalꢀMSCs and FꢀCalꢀ
MSCs suggested that the establishment of internal electric field was caused by the interaction
between Ti3+ and substitution fluorine (Figure S9). Doping of F in TiO2 induced a certain degree
of lattice distortion, which reduced the symmetry of the lattice. Meanwhile, the Ti3+ could
interact with the neighboring F to build an internal electric filed, owing to the strong
electronegativity of fluorine.21
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
ACS Paragon Plus Environment
12

Page 13 of 24
Nano Letters
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
Figure 4. (a) Time course evolution of CH4 production over different samples under the solar
light irradiation for 4 h (300 W Xenon lamp with an AM 1.5 filter, CO2 + H2O). (b) Selectivity
of photocatalytic reduction of CO2 and H2O vapor over MSCs and FꢀMSCs (Selectivity of CH4 =
[8n(CH4)]/[2n(CO) +8n(CH4) + 2n(H2)]×100%, Selectivity of CO = [2n(CO)]/[2n(CO)
+8n(CH4) + 2n(H2)]×100%).34, 35. (c) Time course evolution of CH4 production over FꢀMSCs
under continuous illumination for 16 h (300 W Xenon lamp with an AM 1.5 filter, CO2 + H2O).
(d) F 1s XPS spectra of FꢀMSCs before and after CO2 photoreduction (continuous illumination
for 16 h). (e) Scheme illustration of the difference of Ti3+ impurity levels between MSCs and Fꢀ
MSCs. ꢁE represents the kinetic overꢀpotential of the reduction process.
Internal electric fields in photocatalysts can arise from ferroelectric property, which is
induced by the polarization effect.21 With the introduction of F atoms (substitution for oxygen
ACS Paragon Plus Environment
13

Nano Letters
Page 14 of 24
1
2
3
4
5
6
7
8
9
vacancies) in the surface layer of TiO2ꢀx, some Ti3+ ions on Fꢀtype color centers are oxidized to
Ti4+ owing to the strong electronegativity of fluorine,18 which enhance the ionic polarization
effect between color centers and fluorine ions.20 Furthermore, a substitutional fluorine atom at
the crystal surface causes surface lattice deformation, which can destroy the centrosymmetry of
TiO2 crystal. The positive and negative charges with di erent centers of symmetry can also
enhance the polarization effect.21 As a result, the Ti3+ impurity level has an upsweep in FꢀMSCs.
Under solar light irradiation, some electrons on the Ti3+ impurity band are heated rapidly due to
more metallic nature of FꢀTiO2ꢀx after fluorination. The heated electrons have a higher “transient
energy” than the work function of the material; as a result, they can escape from the impurity
levels of FꢀTiO2ꢀx to take part in the CO2 reduction (detailed discussion in Figure S10).
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
Figure 5. (a) HRTEM image of FꢀPtꢀMSCs. (b) EPR spectra of fluorinated sample with and
without platinum on 77 K. (c) Time course evolution of CH4 production over different samples
in 8 h. (d) In situ FTIR spectra of CO adsorption over the samples under adsorption saturation
and evacuation by argon for 30 min (insert).
ACS Paragon Plus Environment
14

Page 15 of 24
Nano Letters
1
2
3
4
5
6
7
8
9
As expected, FꢀMSCs displayed significantly enhanced production of CH4 than MSCs or
Degussa P25 titania powders under solar light irradiation (Figure 4a), indicating its outstanding
advantages in CO2 deep photoreduction. First we have to exclude the effect of CO2 adsorption.
The CO2 adsorption capacity of different samples have been tested, as shown in Figure S11. We
can conclude that the CO2 adsorption plays no key role in the enhancement of CO2PR
performance over FꢀMSCs. The turnover number (TON) of different photocatalysts in CO2PR
was shown in Table S1 (details in Supporting Information). As a result, the TON of FꢀMSCs
(0.98) is much higher than that of MSCs (0.095). The low activities of CalꢀMSCs and FꢀCalꢀ
MSCs highlight the importance of Ti3+ selfꢀdoping in enhancing solar light harvesting for CH4
generation. It is worth mentioning that if argon was used as the reaction gas instead of CO2, no
CH4 could be obtained (Figure S12), indicating that CH4 is resulted from the reduction of CO2.7
In addition to CH4, CO was also generated during CO2 reduction (Figure S13). While CO
represented the primary product over other catalysts in previous reports,6, 9, 36 in our case CH4
was the main product when FꢀMSCs were used as the catalyst. In order to characterize in detail
the intermediates formed in the CO2PR reaction, the diffuse reflectance infrared Fourierꢀ
transform spectroscopy (DRIFTS) analysis was performed.37, 38 We can infer that the CO2PR
reaction on FꢀMSCs was initiated with the adsorption of CO2 molecules on the surface oxygen
vacancies. Afterwards, under light irradiation, the adsorbed CO2 molecules were activated,
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
2ꢀ
2ꢀ
ꢀ
generating negatively charged intermediate species such as mꢀCO3 , bꢀCO3 , and CO2 .
Subsequently, these intermediates underwent the proton coupled electron transfer (PCET) route
and generated CO or CH4 (detailed discussion in Figure S14). The direct photoreduction of CO
over FꢀMSCs was also presented in Figure S15. The CH4 yield over FꢀMSCs was also much
higher than that of MSCs, indicating that dominant CH4 production was due to the fast reduction
ACS Paragon Plus Environment
15

Nano Letters
Page 16 of 24
1
2
3
4
5
6
7
8
9
of CO to CH4. Our careful calculation suggests that the yield of CH4 can be increased 13 times
from 0.125 with MSCs to 1.63 µmol/g·h with FꢀMSCs, and the CH4 selectivity can be improved
from 25.7% to 85.8% (Figure 4b). Remarkably, FꢀMSCs still have a relatively stable CH4 yield
even after 10ꢀ16 h of continuous irradiation (Figure 4c and S16). Even after 16 h continuous
illumination, the substitutional F in FꢀMSCs showed almost no change (Figure 4d), indicating
the high stability of the fluorinated photocatalyst. Furthermore, the results of photocurrent, UVꢀ
DRS and MottꢀSchottky plots over the used FꢀMSCs confirm the stability of the internal electric
fields in FꢀMSCs (details in Figure S17). For TiO2ꢀx, the Ti3+ impurity level is approximately
located below the CB of 0.11 eV (Figure 2a,d), which is 0.39 eV higher than that of CO2/CH4
potential (ꢀ0.63 eV vs ꢀ0.24 eV). After fluorination, the kinetic overꢀpotential (ꢁE) increased
from 0.39 eV to 0.48 eV (Figure 4e), which indicated a stronger reduction potential of FꢀMSCs
than MSCs.14 In addition to the rutile MSCs, the anatase TiO2ꢀx after fluorination also exhibited a
distinct upperꢀshift of impurity levels (details of DFT in Figure S18) and an obvious
enhancement of CH4 generation under solar light irradiation (Figure S19), which demonstrated
the general applicability of the fluorination strategy in inducing strong reduction potential for
enhanced photocatalytic CO2 reduction. And the surface fluorination over anatase (without Ti3+)
cannot greatly improve the CO2 photoreduction for the CH4 generation. It is worth noting that the
fluorine modified anatase (FꢀTi3+ꢀAnatase: 5.57 µmol/h/g) shows a relatively higher CH4 yield
than most of the recently reported photocatalysts in CO2 reduction, as shown in Table S2.
The photocatalytic activity over FꢀMSCs with different F doping level was studied to reveal
the role of the concentration of Ti3+. With the increase of the concentration of F, the CH4
production yield over FꢀMSCs (m%) increased first and then decreased, as shown in Figure S20.
All the fluorination modified catalysts show an obviously enhanced CH4 production compared
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
ACS Paragon Plus Environment
16

Page 17 of 24
Nano Letters
1
2
3
4
5
6
7
8
9
with MSCs, which further confirms the modulation of the reduction potential of TiO2ꢀx by
fluorination for the efficient and selective CH4 generation from CO2 photoreduction. Seen from
the F 1s XPS spectra of Figure S21, compared with the peak of MSCs, the characteristic peaks of
FꢀMSCs (m%) are all obviously enhanced and moved to the higher binding energy, confirming
the introduction of substitutional F in MSCs. With the increase of the concentration of F from
4.7% to 8.5%, the relative intensity of the characteristic peak of surface TiꢀF (683.7 eV)
increased from 24.1% to 43.7%. Interestingly, the concentration of F has a great influence on the
Ti3+ content in MSCs, as shown in Figure S22. We have demonstrated that the afterꢀtreatment of
fluorination can decrease the Ti3+ concentration due to the substitution of F for oxygen vacancy
and the building of internal electric field. When the F amount was fixed at 4.7%, there was a
distinct decrease of Ti3+ in FꢀMSCs, compared with that of MSCs. It indicates that the F atoms
preferentially displace surface or subsurface oxygen vacancies and build an internal electric field
in the process of fluorination, resulting in a rapid decrease in Ti3+ concentration. With the
continued increase of F from 4.7% to 6.3%, the Ti3+ concentration starts to pick up, but still less
than the original value of MSCs. With the increase of F, the surface TiꢀF increased gradually,
which would induce some new defects appearing on the surface of the catalyst, owing to the
oxygen deficiency or interstitial fluorine induced by the excessive fluorination.39 As a result, the
Ti3+ increases with the increase of F from 4.7% to 6.3%. When the amount of F reaches 8.5%,
the newly generated Ti3+ EPR signal over FꢀMSCs has exceeded the original signal of MSCs.
The excessive Ti3+ is likely to become a new electronꢀhole recombination center, which is not
conducive to the occurrence of CO2 photoreduction. Therefore, the FꢀMSCs (8.5%) display a
decreased CH4 production yield compared with that of FꢀMSCs (5.5%). Despite all this, the
activity of all the fluorinated catalysts was much higher than that of blank samples.
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
ACS Paragon Plus Environment
17

Nano Letters
Page 18 of 24
1
2
3
4
5
6
7
8
9
To highlight another advantage of fluorination of TiO2ꢀx, we loaded Pt nanoparticles on the
surface of MSCs and followed by postꢀfluorination using a similar procedure for FꢀMSCs to
produce a control sample of FꢀPtꢀMSCs. The TGA and C1s XPS spectra of the samples before
and after modification indicate the absence of carbon impurities during the fluorination and Pt
loading treatment (Figure S23a,b). The UVꢀDRS spectrum of the MSCs treated in the
isopropanol (IPA) indicate that the solvothermal treatment (without NH4F) does not affect the
Ti3+ (Figure S23c). Analyses by TEM and HRTEM suggested that the Pt nanoparticles were
highly dispersed on the surface of MSCs (Figure 5a). After Pt loading, the Ti3+ concentration
was about the same as that of FꢀMSCs (Figure 5b), which ensured the formation of internal
electric field in FꢀPtꢀMSCs. As predicted, the FꢀPtꢀMSCs exhibited a significantly higher activity
for the CH4 generation from CO2 photoreduction than FꢀMSCs, PtꢀMSCs and other references
(Figure 5c). We have shown the original samples that treated in the same conditions, i.e. without
NH4F by the IPA only involved solvothermal method in the Figure 5c (MSCs (IPA only)). The
low activity of MSCs (IPA only) indicates the solvothermal treatment (without NH4F) cannot
enhance the CO2PR performance. The quantum yield (QY) of FꢀPtꢀMSCs was 0.038% (Table
S3). Most importantly, the FꢀPtꢀMSCs could keep a very stable CH4 generation rate even after
continuous illumination for 8 hours. However, the conventional PtꢀMSCs has an unstable activity
for the CH4 generation, owing to poisoning of the Pt surface by the adsorption of the byꢀproduct
of CO.40 To investigate the cause of the high stability of FꢀPtꢀMSCs, we have studied the
adsorption and desorption of CO on various samples using in situ infrared spectroscopy (Figure
5d). The peaks in the range of 2100~2200 cmꢀ1 could be ascribed to the characteristic adsorption
of CO molecules on TiO2,41, 42 and the small peak at about 2030 cmꢀ1 could be assigned to CO
adsorption on the Pt surface.43 After purging with Ar gas, only the characteristic peak of CO
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
ACS Paragon Plus Environment
18

Page 19 of 24
Nano Letters
1
2
3
4
5
6
7
8
9
adsorption on the Pt remained in the IR spectra (inset of Figure 5d). Moreover, the peak intensity
of FꢀPtꢀMSCs was very close to that of PtꢀMSCs, implying the fluorination could not decrease
the CO adsorption on Pt. In our case, the Pt NPs not only act as the electron trapping agent but
also become the active sites in CO2PR. According to our previous DFT calculations,44 the
hydrogenation of *CO and *COH require the highest energy barrier (the total free energy in
these two steps on Pt(111) is 0.74 eV), which become the rate limiting steps in CO2PR.
Considering the fact that the FꢀPtꢀMSCs not only showed the enhanced charge separation
efficiency (highest reacted electrons’ number in Table S4), but also possessed elevated CB level,
the FꢀPtꢀMSCs could provide sufficient electrons with strong reduction potential to promote the
deep reduction of *CO to generate CH4 in CO2PR. Hence, we believe that the enhanced
photocatalytic activity of FꢀPtꢀMSCs is not due to direct inhibition of Pt poisoning. Rather, it is
mainly caused by the deep reduction from CO to CH4, which decreases the standing time of CO
on Pt and thereby reduces the possibility of CO poisoning on Pt.
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
In order to reveal the key parameters that determine the products’ yield and selectivity in
CO2 photocatalytic reduction (CO2PR), the main products’ yield, reacted electrons’ number and
selectivity for CH4 over different photocatalysts in CO2PR are summarized in Table S4.
Apparently, both the FꢀMSCs and FꢀPtꢀMSCs show enhanced CH4 yield and selectivity; on the
contrary, the main product of MSCs and FꢀCalꢀMSCs is CO and the mixture of CO and H2,
respectively. Based on the MottꢀSchottky plots (Figure 2c,d and Figure S5b,c), there’s no doubt
that all of the above samples possessed suitable reduction potential which could trigger both the
CO2PR reaction and hydrogen evolution reaction (HER) (eqs. (4)~(6) in Supporting
Information) in thermodynamics.14 Therefore, the distinct difference of products yield acquired
by different photocatalysts was illustrated in both surface catalytic reaction and charge separation
ACS Paragon Plus Environment
19

Nano Letters
Page 20 of 24
1
2
3
4
5
6
7
8
9
efficiency. As for the MSCs and its fluorination derivatives, there exists abundant surface
oxygen vacancies (Figure 1b and Figure 5b). Hence, the CO2 molecules tend to be adsorbed and
activated on these oxygen vacancies45, 46 and the competitive hydrogen evolution reaction could
be greatly suppressed. However, after the calcination treatment in air, the oxygen vacancies of
CalꢀMSCs would vanish, in this case, the H2O molecules tend to be adsorbed and generate H2 in
the photocatalytic reaction, hence the FꢀCalꢀMSCs show the highest hydrogen yield (Table S4).
Thereafter, the surface catalytic reaction towards CH4 was determined by the charge separation
efficiency and enlarged reduction potential of the photocatalyst.47 The conduction band of Fꢀ
MSCs was elevated by the polarization effect of the lattice interval F, therefore, FꢀMSCs
possessed more negative reduction potential compared with MSCs, which could trigger the 8ꢀ
electron CH4 generation reaction. In addition, FꢀPtꢀMSCs with enhanced charge separation
efficiency, which could further improve the generation of CH4 and show the highest yield and
selectivity (Table S4). In sum, FꢀMSCs and FꢀPtꢀMSCs not only contain surface oxygen
vacancies but also display elevated reduction potential, thus could promote the CO2PR for
enhanced selective CH4 formation both thermodynamically and kinetically.
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
In conclusion, we have developed an effective fluorination strategy to enhance the reduction
potential of TiO2ꢀx while maintaining the efficient solar light harvesting. The interaction between
Ti3+ and substitutional fluorine contributed to the establishment of internal electric fields in TiO2ꢀ
x, which led to the upsweep of Ti3+ impurity level. The fluorinated TiO2ꢀx showed a high CH4
production yield and excellent selectivity for CO2 photoreduction under solar light irradiation.
Our research provides a novel metalꢀfree strategy for deep photoreduction of CO2 and efficient
generation of CH4, which is expected to have its high impact in addressing both environmental
and energy challenges.
ACS Paragon Plus Environment
20

Page 21 of 24
Nano Letters
1
2
3
4
5
6
ASSOCIATED CONTENT
7
8
9
Supporting Information.
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 Supporting Information is available free of charge on the ACS Publications website at DOI:
XXXX
Experimental procedures, summary of different catalytic systems, and additional
characterizations (file type, i.e., PDF)
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
This work was supported by the National Natural Science Foundation of China (21773062,
21577036, 21377038, 21677048, 5171101651), State Key Research Development Program of
China (2016YFA0204200), the Shanghai Education Development Foundation and Shanghai
Municipal Education Commission (16JC1401400), Shanghai Pujiang Program (17PJD011), and
the Fundamental Research Funds for the Central Universities (22A201514021).
Acknowledgment is also made to the Donors of the American Chemical Society Petroleum
Research Fund for partial support of this research (55904ꢀND10).
REFERENCES
(1) Han, N.; Wang, Y.; Ma, L.; Wen, J.; Li, J.; Zheng, H.; Nie, K.; Wang, X.; Zhao, F.; Li, Y.;
Fan, J.; Zhong, J.; Wu, T.; Miller, D. J.; Lu, J.; Lee, S.ꢀT.; Li, Y. Chem 2017, 3, 652ꢀ664.
(2) Rao, H.; Schmidt, L. C.; Bonin, J.; Robert, M. Nature 2017, 548, 74.
(3) Gao, S.; Gu, B.; Jiao, X.; Sun, Y.; Zu, X.; Yang, F.; Zhu, W.; Wang, C.; Feng, Z.; Ye, B.;
Xie, Y. J. Am. Chem. Soc. 2017, 139, 3438ꢀ3445.
(4) Li, K.; Peng, B.; Peng, T. ACS Catal. 2016, 6, 7485ꢀ7527.
(5) Chang, X.; Wang, T.; Gong, J. Energy Environ. Sci. 2016, 9, 2177ꢀ2196.
ACS Paragon Plus Environment
21

Nano Letters
Page 22 of 24
1
2
3
4
5
6
7
8
9
(6) Zhao, Y.; Chen, G.; Bian, T.; Zhou, C.; Waterhouse, G. I. N.; Wu, L.ꢀZ.; Tung, C.ꢀH.; Smith,
L. J.; O'Hare, D.; Zhang, T. Adv. Mater. 2015, 27, 7823ꢀ7823.
(7) Dong, C.; Xing, M.; Zhang, J. Mater. Horiz. 2016, 3, 608ꢀ612.
(8) Dong, C.; Xing, M.; Zhang, J. J. Phys. Chem. Lett. 2016, 7, 2962ꢀ2966.
(9) Ji, Y.; Luo, Y. J. Am. Chem. Soc. 2016, 138, 15896ꢀ15902.
(10) Abdellah, M.; ElꢀZohry, A. M.; Antila, L. J.; Windle, C. D.; Reisner, E.; Hammarström, L.
J. Am. Chem. Soc. 2017, 139, 1226ꢀ1232.
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
(11) Joo, J. B.; Dahl, M.; Li, N.; Zaera, F.; Yin, Y. Energy Environ. Sci. 2013, 6, 2082ꢀ2092.
(12) Liu, H.; Joo, J. B.; Dahl, M.; Fu, L.; Zeng, Z.; Yin, Y. Energy Environ. Sci. 2015, 8, 286ꢀ
296.
(13) He, Z.; Wen, L.; Wang, D.; Xue, Y.; Lu, Q.; Wu, C.; Chen, J.; Song, S. Energy Fuels 2014,
28, 3982ꢀ3993.
(14) Habisreutinger, S. N.; SchmidtꢀMende, L.; Stolarczyk, J. K. Angew. Chem., Int. Ed. 2013,
52, 7372ꢀ7408.
(15) Liu, L.; Chen, X. Chem. Rev. 2014, 114, 9890ꢀ9918.
(16) Yi, Q.; Zhou, Y.; Xing, M.; Zhang, J. Res. Chem. Intermed. 2016, 42, 4181ꢀ4189.
(17) Chang, X.; Wang, T.; Gong, J. Energy Environ. Sci. 2016, 9, 2177ꢀ2196.
(18) Zhang, Z.; Yates, J. T. J. Am. Chem. Soc. 2010, 132, 12804ꢀ12807.
(19) Zhang, Z.; Yates, J. T. J. Phys. Chem. Lett. 2010, 1, 2185ꢀ2188.
(20) Chen, J.; Lin, L. B.; Jing, F. Q. J. Phys. Chem. Solids 2001, 62, 1257ꢀ1262.
(21) Li, L.; Salvador, P. A.; Rohrer, G. S. Nanoscale 2014, 6, 24ꢀ42.
(22) Zhou, Y.; Yi, Q.; Xing, M.; Shang, L.; Zhang, T.; Zhang, J. Chem. Commun. 2016, 52,
1689ꢀ1692.
(23) Zuo, F.; Wang, L.; Wu, T.; Zhang, Z.; Borchardt, D.; Feng, P. J. Am. Chem. Soc. 2010, 132,
11856ꢀ11857.
(24) Brauer, J. I.; Szulczewski, G. J. Phys. Chem. B 2014, 118, 14188ꢀ14195.
(25) Kuznetsov, V. N.; Serpone, N. J. Phys. Chem. C 2009, 113, 15110ꢀ15123.
(26) Li, D.; Haneda, H.; Hishita, S.; Ohashi, N.; Labhsetwar, N. K. J. Fluorine Chem. 2005, 126,
69ꢀ77.
(27) Lee, K.; Kim, Q.; An, S.; An, J.; Kim, J.; Kim, B.; Jhe, W. Proc. Natl. Acad. Sci. U. S. A.
2014, 111, 5784ꢀ5789.
ACS Paragon Plus Environment
22

Page 23 of 24
Nano Letters
1
2
3
4
5
6
(28) Feng, C.; Wang, Y.; Jin, Z.; Zhang, J.; Zhang, S.; Wu, Z.; Zhang, Z. New J. Chem. 2008, 32,
1038ꢀ1047.
(29) Xing, M.; Zhang, J.; Qiu, B.; Tian, B.; Anpo, M.; Che, M. Small 2015, 11, 1920ꢀ1929.
(30) Xing, M.; Zhang, J.; Chen, F. Appl. Catal., B 2009, 89, 563ꢀ569.
(31) Spadavecchia, F.; Ceotto, M.; Lo Presti, L.; Aieta, C.; Biraghi, I.; Meroni, D.; Ardizzone,
S.; Cappelletti, G. Chin. J. Chem. 2014, 32, 1195ꢀ1213.
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
(32) Di Valentin, C.; Finazzi, E.; Pacchioni, G.; Selloni, A.; Livraghi, S.; Czoska, A. M.;
Paganini, M. C.; Giamello, E. Chem. Mater. 2008, 20, 3706ꢀ3714.
(33) Cui, Y.; Du, H.; Wen, L. Mater. Sci. Forum 2009, 620ꢀ622, 647ꢀ650.
(34) Neaţu, ꢂ.; MaciáꢀAgulló, J. A.; Concepción, P.; Garcia, H. J. Am. Chem. Soc. 2014, 136,
15969ꢀ15976.
(35) Zhai, Q.; Xie, S.; Fan, W.; Zhang, Q.; Wang, Y.; Deng, W.; Wang, Y. Angew. Chem., Int.
Ed. 2013, 52, 5776ꢀ5779.
(36) Wang, S.; Guan, B. Y.; Lu, Y.; Lou, X. W. D. J. Am. Chem. Soc. 2017, 139, 17305ꢀ17308.
(37) Long, R.; Li, Y.; Liu, Y.; Chen, S.; Zheng, X.; Gao, C.; He, C.; Chen, N.; Qi, Z.; Song, L.;
Jiang, J.; Zhu, J.; Xiong, Y. J. Am. Chem. Soc. 2017, 139, 4486ꢀ4492.
(38) Liu, L.; Jiang, Y.; Zhao, H.; Chen, J.; Cheng, J.; Yang, K.; Li, Y. ACS Catal. 2016, 6, 1097ꢀ
1108.
(39) Fang, W. Q.; Wang, X. L.; Zhang, H.; Jia, Y.; Huo, Z.; Li, Z.; Zhao, H.; Yang, H. G.; Yao,
X. J. Mater. Chem. A 2014, 2, 3513ꢀ3520.
(40) Wang, W.ꢀN.; An, W.ꢀJ.; Ramalingam, B.; Mukherjee, S.; Niedzwiedzki, D. M.;
Gangopadhyay, S.; Biswas, P. J. Am. Chem. Soc. 2012, 134, 11276ꢀ11281.
(41) Green, I. X.; Tang, W.; Neurock, M.; Yates, J. T. Science 2011, 333, 736ꢀ739.
(42) Derrouiche, S.; Gravejat, P.; Bianchi, D. J. Am. Chem. Soc. 2004, 126, 13010ꢀ13015.
(43) Qiao, B.; Wang, A.; Yang, X.; Allard, L. F.; Jiang, Z.; Cui, Y.; Liu, J.; Li, J.; Zhang, T. Nat.
Chem. 2011, 3, 634ꢀ641.
(44) Dong, C.; Lian, C.; Hu, S.; Deng, Z.; Gong, J.; Li, M.; Liu, H.; Xing, M.; Zhang, J. Nat.
Commun. 2018, 9, 1252.
(45) Liu, L.; Zhao, C.; Li, Y. J. Phys. Chem. C 2012, 116, 7904ꢀ7912.
(46) Lee, J.; Sorescu, D. C.; Deng, X. J. Am. Chem. Soc. 2011, 133, 10066ꢀ10069.
(47) Bae, K.ꢀL.; Kim, J.; Lim, C. K.; Nam, K. M.; Song, H. Nat. Commun. 2017, 8, 1156.
ACS Paragon Plus Environment
23

Nano Letters
Page 24 of 24
1
2
3
4
For TOC only
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
ACS Paragon Plus Environment
24