Fabrication of an Fe-Doped SrTiO3 Photocatalyst with Enhanced Dinitrogen Photofixation Performance

Article abstract of DOI:10.1002/ejic.201900098

SrTiO₃ as semiconducting photocatalyst has been extensively investigated due to its band edges meeting the thermodynamic requirements for water splitting, but a few attention has been concentrated on its application in the NH₃ synthesis via N₂ photofixation process. Herein, Fe-doped SrTiO₃ (Fe_xSr_1–xTiO₃) products (0 ≤ x ≤ 0.20) were synthesized via a hydrothermal process followed by calcination at 700 °C. All Fe_xSr_1–xTiO₃ products (0.03 ≤ x ≤ 0.20) deliver an enhanced N₂ fixation ability, and Fe_xSr_1–xTiO₃ (x = 0.10) achieves the best NH₃ production activity of 30.1 μmol g^–1 h^–1, which is 3.2-hold higher than that of SrTiO₃ alone. Once the x value is higher than 0.10, Fe_xSr_1–xTiO₃ will transform into composites containing Fe-doped SrTiO₃ and α-Fe₂O₃, which acts as charge recombination sites, thus causes a decreased N₂ fixation activity. Further investigations demonstrate that the surface Fe³⁺-doped sites can not only chemisorb and activate N₂ molecules, but also promote the interfacial electron transfer from Fe-doped SrTiO₃ to N₂ molecules, and thus significantly improve the N₂ fixation ability. The present Fe-doped SrTiO₃ products exhibit characteristic features such as stable and efficient N₂ fixation ability as well as simultaneous realization of N₂ reduction and H₂O oxidation without co-catalyst, which are of significance in artificial photosynthesis with H₂O as electron and proton sources.

Full text of DOI:10.1002/ejic.201900098

A Journal of
Accepted Article
Title: Fabrication of Fe-doped SrTiO3 photocatalyst with enhanced
dinitrogen photofixation performance
Authors: Zihao Ying, Shengtao Chen, Tianyou Peng, Renjie Li, and
Jing Zhang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Inorg. Chem. 10.1002/ejic.201900098
Link to VoR: http://dx.doi.org/10.1002/ejic.201900098

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
Fabrication of Fe-doped SrTiO
photofixation performance
3
photocatalyst with enhanced dinitrogen
[
a]
[a]
[a]
[a]
[a]
Zihao Ying, Shengtao Chen, Tianyou Peng,* Renjie Li, Jing Zhang
Abstract: SrTiO
3
as semiconducting photocatalyst has been
2
because of the difficult adsorption and activation of N molecules
extensively investigated due to its band edges meeting the
thermodynamic requirements for water splitting, but a few attention
on photocatalysts. Therefore, developing new photocatalysts
with more efficient and stable performance is an important
challenge to the N
SrTiO , as a wide-bandgap semiconductor, has been applied
in the field of water splitting.[17] In 1983, SrTiO
was firstly used
as photocatalyst for N fixation, and the NH production activity
can be slightly improved by Ni doping.[18] Also, Nb-doped SrTiO
electrode loaded with Au or Ru nanoparticles as light harvesting
centers has been developed for visible-light-driven NH
2
photofixation technology.
has been concentrated on its application in the NH
3
synthesis via N
2
3
photofixation process. Herein, Fe-doped SrTiO
3
(Fe
x
Sr1-xTiO
3
)
3
products (0  x  0.20) were synthesized via a hydrothermal process
o
2
3
followed by calcination at 700 C. All Fe
x
Sr1-xTiO
3
products (0.03  x
3

0.20) deliver an enhanced N
2
fixation ability, and Fe
x
Sr1-xTiO
3
(x =
-
1
-1
0
.10) achieves the best NH
3
production activity of 30.1 μmol g h ,
3
which is 3.2-hold higher than that of SrTiO
is higher than 0.10, Fe will transform into composites
containing Fe-doped SrTiO and α-Fe , which acts as charge
recombination sites, thus causes a decreased N fixation activity.
Further investigations demonstrate that the surface Fe -doped sites
can not only chemisorb and activate N molecules, but also promote
the interfacial electron transfer from Fe-doped SrTiO to
molecules, and thus significantly improve the N fixation ability. The
present Fe-doped SrTiO products exhibit characteristic features
such as stable and efficient fixation ability as well as
simultaneous realization of N reduction and H O oxidation without
co-catalyst, which are of significance in artificial photosynthesis with
O as electron and proton sources.
3
alone. Once the x value
production by adding ethanol into the anodic compartment as
sacrificial donor.[19,20] Recently, it was reported that the surface
defects such as oxygen and/or nitrogen vacancies can activate
Sr1-xTiO
x 3
3
2 3
O
2
3+
N
2
and promote the interfacial electron transfer, and thus result
in the improvement of N
fixation ability.[7,8,21,22] Therefore, it can
be hypothesized that the above-mentioned Fe (in TiO [2]) or Ni
in SrTiO [18]) doped sites might also act as another surface
defects to play a similar effect on the improvement of N fixation
ability. More recently, Fe-doped graphitic carbon nitride (g-C
was reported to have an outstanding N fixation ability, and the
Fe3+ ions doped at interstitial sites can chemisorb and activate
molecules, then transfer the photogenerated electrons from
g-C to the adsorbed N
molecules.[15] However, there is no
investigation on the N fixation performance of Fe-doped SrTiO
2
2
2
3
N
2
(
3
2
2
3
3
4
N )
N
2
2
2
2
N
2
H
2
3
N
4
2
2
3
Introduction
to the best of our knowledge.
In this study, a series of Fe-doped SrTiO
3
(Fe
x
Sr1-xTiO
3
)
2
The Haber-Bosch process, a typical industrial dinitrogen (N )
fixation reaction, has sustained the modern agriculture and
provided 50% of the nitrogen atoms for production and life of
mankind. Nevertheless, the Haber-Bosch reactions cost 1-2%
annual global energy production capacity and generate more
products (0  x  0.20) were synthesized via a hydrothermal
process followed by calcination at 700 C. It was found that the
o
Fe
x
Sr1-xTiO
3
products with x > 0.10 will transform into
composites containing α-Fe O3 nanocrystals (NCs) and Fe-
2
than 300 million tons of CO
pressure conditions to activate and cleave the highly stable N≡N
triple bond of N fixation
molecules.[1] Therefore, artificial N
under mild conditions is of great significance because of the
reduction of input energy and no use of H gas. In 1977,
Schrauzer and Guth firstly reported TiO -based materials can
photofixate N into ammonia (NH ) under UV irradiation, and Fe
doping can enhance the photocatalytic reactivity of rutile TiO
2
since it needs high temperature and
, in which some Fe3+ ions replace Sr2+ sites of the
lattice. When used them as photocatalysts, those
doped SrTiO
cubic SrTiO
Fe
fixation ability compared to the single SrTiO
sacrificial reagent and metal co-catalyst. The effects of Fe
doping and -Fe NCs on the N fixation ability of the obtained
Fe products are discussed in detail.
3
3
2
2
x
Sr1-xTiO
3
products (0.03  x  0.20) deliver an improved N
without the aid of
2
3
2
2
2
O
3
2
2
3
x
Sr1-xTiO
3
[2]
2
.
[
3-6]
Thereafter, various materials such as metal oxides,
bismuth
metal sulfides,[11,12] and carbonaceous matters
have been used as photocatalysts for N fixation. Although
the above researches have made great advances, the NH
production activities of most photocatalysts are still unsatisfying
[
7-10]
oxyhalides,
Results and Discussion
[
13-16]
2
Crystal Phase and Composition Analyses
3
Figure 1 depicts the X-ray diffraction (XRD) patterns of the Fe-
doped SrTiO
hydrothermal process followed by calcination at 700°C. The
single SrTiO (Fe products with x = 0) exhibits obvious
3
(Fe
x
Sr1-xTiO
3
)
products derived from the
[
a] Mr. Z. Ying, S. Tao, Prof. Dr. T. Peng, Dr. R. Li, J. Zhang
College of Chemistry and Molecular Sciences, Wuhan University,
Wuhan 430072, P. R. China
3
x
Sr1-xTiO
3
diffraction peaks at 2θ = 22.8°, 32.4°, 40.0°, 46.5°, 52.4°, 57.8°
E-mail: typeng@whu.edu.cn
and 67.8° (Figure 1a), corresponding to the reflections of (100),
http://faculty.whu.edu.cn/show.jsp?lang=cn&n=Peng Tian You
Supporting information for this article is available on the WWW
under https://doi.org/10.1002/ejic.2019xxxxxx
(
(
110), (111), (200), (210), (211) and (220) planes of cubic SrTiO
JCPDS No. 73-0661),[23] respectively. The Fe
products
3
x
Sr1-xTiO
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
with x  0.10 show XRD patterns very similar to the single
SrTiO
3
as shown in Figure 1. From the element compositions
SrTiO
3
, while the product with x = 0.15 delivers very weak new
determined using X-ray fluorescence (XRF) technique and the
relative molar percentages of Fe, Sr, Ti and O atoms listed in
peaks, which intensities show an increasing trend along with the
increase of x value from 0.15 to 0.75 (Figure 1b). Especially,
Table 1, it can be seen that the Fe-atom molar ratios in Fe
TiO products are very similar to the respective addition
contents during the synthesis process. However, the O-atom
molar ratios in Fe products (0.03  x  0.10) are slightly
higher than the stoichiometric ratio (Table 1), inferring that some
low-valent Sr2+ ions might be substituted by high-valent Fe3+ ions,
and thus additional O2- ions might locate at the interstitial
positions so as to maintain the charge balance. Moreover, the
x
Sr1-
those Fe
obvious new peaks at 2θ = 33.2°, 35.6°, 40.8°, 49.5°, 54.1°,
2.3° and 64.0°, which can be assigned to the reflections of
104), (110), (114), (024), (116), (214) and (300) planes of α-
Fe
(JCPDS No. 89-2810),[24] respectively. The above results
demonstrate that those Fe products with x > 0.10 will
transform into composites containing α-Fe
SrTiO
x
Sr1-xTiO
3
products with x = 0.50 and 0.75 exhibit
x
3
6
x
Sr1-xTiO
3
(
2
O
3
x
Sr1-xTiO
3
2
O
3
and Fe-doped
3
.
obviously increased O-atom molar ratios of Fe
products with x  0.15 (Table 1) can be mainly due to the
formation of -Fe
x
Sr1-xTiO
3
2
3
O .
Table 1. Composition ratios of the obtained Fe
x
Sr1-xTiO
3
products.
Fe:Sr:Ti:O molar
Sr [mol%] Ti [mol%] O [mol%] ratio
x 3
Fe Sr1-xTiO
x value
Fe [mol%]
0
0
19.07
18.57
18.54
17.27
16.58
15.84
20.37
20.04
19.96
19.88
19.26
19.08
60.56
60.96
60.83
61.02
61.31
61.54
0:0.94:1:2.97
0.03
0.05
0.10
0.15
0.20
0.43
0.67
1.83
2.85
3.54
0.02:0.93:1:3.04
0.04:0.93:1:3.05
0.09:0.87:1:3.07
0.15:0.86:1:3.18
0.19:0.83:1:3.23
Also, the survey X-ray photoelectron spectra (XPS) indicate
that those Fe products have weak binding energy peak
x
Sr1-xTiO
3
of Fe element at 710 eV in addition to those of Sr, Ti and O
elements (Figure S2).[
29,30]
Moreover, no obvious binding energy
peak of N element can be observed in the range of 398-407
eV,[31] suggesting that those products have no N species, and
the following NH
The high-resolution Sr2p XPS spectra (Figure 2a) show that the
single SrTiO has two symmetrical peaks with binding energies
at 132.7 (Sr2p3/2) and 134.4 (Sr2p1/2) eV,[29] and their positions
for those Fe Sr1-xTiO products (0.03  x  0.10) shift ca. 0.22 eV
toward lower binding energy; whereas the two binding energy
peaks of Ti2p3/2 and Ti2p1/2 in those Fe products (0.03
x  0.10) only slightly shift toward lower binding energy
3 2
production is derived from N fixation reactions.
3
x
3
x
Sr1-xTiO
3
Figure 1. Typical XRD patterns of the obtained Fe
x
Sr1-xTiO
3
products with x =

0-0.10 (a) and x = 0.15-0.75 (b).
(Figure 2b).[ These results
29]
compared to the single SrTiO
3
imply that Fe3+ doping has some effects on the chemical
environments of Sr2+ and Ti4+ sites, and the more obvious shifts
of Sr2p peaks possibly can be due to some Fe3+ ions instead of
3
Compared to the single SrTiO , the (110) and (200) peaks of
Fe
angle accompanied by reduced intensities (Figure S1), implying
that some Fe3+ ions might dope into the cubic SrTiO
lattice to
form Fe solid solution. However, the slight shifts of
110) and (200) peaks terminated at x = 0.10, suggesting that
Fe3+ ions have very limited solubility in the SrTiO
lattice during
x
3
Sr1-xTiO products with x  0.10 slightly shift toward higher 2θ
Sr2+ sites of the SrTiO
binding energy peaks shift backward for the Fe
with x = 0.20, implying strong electron interactions existing
between the Fe-doped SrTiO and the newly formed -Fe
since this shift toward higher binding energy infers that some
electrons can transfer from Fe-doped SrTiO to -Fe , which
, and
lattice.[32] Nevertheless, those Sr2p
3
3
x
Sr1-xTiO
3
product
x
Sr1-xTiO
3
(
3
2 3
O
3
the present preparation process. It is reasonable by considering
that the radii (64.5 pm) of Fe3+ is much smaller than that (144
3
2 3
O
2
+
3+
leads to a decreased electron density of Fe-doped SrTiO
then the enhanced binding energy value of Sr2p.[33,34]
3
pm) of Sr , which leads to the difficulty in Fe ions instead of
Sr2+ sites of SrTiO directly,[25-28] and thus the excessive Fe3+
3
ions in Fe Sr1-xTiO products with x > 0.10 would generate as -
Fe to form composites containing α-Fe and Fe-doped
x
3
2
O
3
2 3
O
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
x 3 2 3
Figure 2. Regional XPS spectra Sr2p (a), Ti2p (b), O1s (c) and Fe2p (d) of the obtained Fe Sr1-xTiO products (x = 0-0.20) and the single -Fe O .
The O1s XPS spectra (Figure 2c) of the single SrTiO
3
can be
Fe
x
Sr1-xTiO
3
products with x > 0.10 would generate as -Fe
2 3
O
deconvoluted into two symmetrical peaks with binding energy
values at 530.8 and 529.1 eV, corresponding to the surface Ti-
to form composites containing α-Fe
2
O
3
and Fe-doped SrTiO ,
3
which is consistent with the above XRD analysis results (Figure
1). Moreover, there is no Fe2p3/2 binding energy peak of Fe2+
OH species and crystal lattice oxygen of the cubic SrTiO
respectively. As for those Fe products with x  0.10,
3
,[35]
ions at 709.6 eV, suggesting no Fe2+ species exist in the Fe
TiO
products.[37]
To further clarify the valence state of Fe, Mossbauer spectra
(Figure S3) of Fe products (x = 0.05, 0.10) were
x
Sr1-xTiO
3
x
Sr1-
the O1s peaks exhibit slight shift toward lower binding energy
value, implying the changed chemical environment of O atoms
after Fe doping. While the main O1s peak (Figure 2c) of the
x
3
x
Sr1-xTiO
3
Fe
x
Sr1-xTiO
3
product with x = 0.20 can be deconvoluted into two
obtained at room temperature. As seen, those Mossbauer
spectra exhibit typical doublet peaks with intensity ratios close to
1:1, and the fitting results demonstrate that the Fe valance
peaks with binding energies of 529.0 and 529.2 eV, which
correspond to the crystal lattice oxygen atoms of Fe-doped
SrTiO
Figure 2d) of the single -Fe
at 710.6 (Fe2p3/2) and 724.2 (Fe2p1/2) eV of Fe3+ ions with a
satellite peak at 718.4 eV,[37] and the Fe
Sr1-xTiO product with x
0.20 exhibits a similar Fe2p spectrum, while those products
3
and -Fe
2
O
3
,[36] respectively. The Fe2p XPS spectrum
presents binding energy peaks
states in Fe
ascribed to 3+.[38] It suggests that Fe3+ ions can dope into the
x 3
SrTiO lattice, and thus resulting in the formation of Fe Sr1-xTiO
x
3
Sr1-xTiO products (x = 0.05, 0.10) can be mostly
(
2 3
O
3
x
3
solid solutions, which is consistent with the above XRD and XPS
analysis results.
=
with x  0.10 show very low XPS signal due to the low Fe
concentration. To further investigate the detail information of
Fe2p XPS spectra, their deconvolution results are also shown in
Table 2. Positron lifetimes of the Fe
x
Sr1-xTiO
[ps]
396 ± 11 32 ± 2
3
products (x = 0, 0.05 and 0.10).
Sample
τ
1
[ps]
I
1
[%]
τ
2
I
2
[%] [ns] [%]
τ
3
I
3
Figure 2d. As seen, those Fe
x
Sr1-xTiO
3
products with x  0.10
x = 0
185 ± 4 64 ± 2
209 ± 6 63 ± 4
208 ± 3 60 ± 2
2.6 ± 0.2 1.8 ± 0.1
2.4 ± 0.2 2.2 ± 0.1
3.4 ± 0.1 2.3 ± 0.1
only present a single binding energy peak at 710.1 eV, which
can be due to the doped Fe3+ ions, while the Fe2p peak of
x = 0.05
x = 0.10
418 ± 7
445 ± 8
32 ± 4
38 ± 2
Fe
binding energy peaks at 710.1 and 710.7 eV, which can be due
to the Fe species in Fe and -Fe , respectively.
These results demonstrate that the excessive Fe3+ ions in those
x
3
Sr1-xTiO product with x = 0.20 can be deconvoluted into two
The above results indicate that some Fe3+ ions can dope into
the cubic SrTiO lattice, which would lead to the formation of
defect states, and thus positron annihilation spectrometry (PAS)
x
Sr1-xTiO
3
2 3
O
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
was used to quantify the existence and relative concentration of
defects via measuring the positron lifetime (Table 2).[39] As seen,
increase of x values. It implies that the enhanced surface
3+
defects can be mainly ascribed to the surface Fe -doped sites
the first lifetime (τ
the longer lifetime (τ
the longest one (τ ) is attributed to the positron annihilation in the
large voids present in the slot of nanoparticles.[39] Compared to
the single SrTiO , those Fe products with x = 0.05 and
.10 show obviously prolonged τ and τ values, inferring the
1
) of positron can be confirmed as bulk defects,
when the x value is enhanced from 0.05 to 0.10, which can
2
) can be assigned to surface defects, and
improve the reaction possibility for N
discussed below.
2
photoredcution as will be
3
3
x
Sr1-xTiO
3
Microstructure and Surface Area Analyses
0
1
2
3
The FESEM images exhibit that the single SrTiO has quasi-
obvious increases in bulk and surface defects due to the Fe3+
sphere-like morphology with smooth surface and particle sizes in
the range of 100-200 nm (Figure 4a). Although the Fe
product with x = 0.05 also shows quasi-sphere-like morphology
x
very similar to the single SrTiO (Figure 4b), those Fe Sr1-xTiO
3
doping, which then enhance the lifetime of positron in the Fe-
x
Sr1-xTiO
3
doped products.[40] Although those Fe
.05 and 0.10 have very similar τ values (contributed by bulk
defects), the Fe-doped product with x = 0.10 exhibits a longer τ
x
3
Sr1-xTiO products with x =
0
1
3
2
products with x = 0.10 and 0.20 display much smaller sizes with
the coexistence of ultrafine nanoparticles (Figure 4c and d). This
result suggests that the addition of Fe ions during the
preparation process would obviously change the particle’s
microstructure and morphology. The TEM image indicates that
value (contributed by surface defects) than that with x = 0.05. It
suggests that an enhanced Fe3+ doping level can contribute
more surface defects.
the single SrTiO
3
is mainly composed by quasi-sphere-like
nanoparticles with smooth surface (Figure 5a), and the HRTEM
image exhibits that the d-spacings of lattice fringes (as shown in
the white dot square in Figure 5b) are ca. 0.394, 0.281 and
0
.231 nm, similar to the theoretical values of (001), (110) and
111) facets of cubic SrTiO
,[43], respectively. Furthermore, the
Fast Fourier Transform (FFT) pattern displays obvious diffraction
spots of (001), (110) and (111) facets of cubic SrTiO , indicating
the high crystallinity of the single SrTiO
(
3
3
3
.
x 3
Figure 3. Typical EPR spectra of the obtained Fe Sr1-xTiO products (x = 0-
0.20) at room temperature.
The electron paramagnetic resonance (EPR) spectra (Figure
) indicate that the single SrTiO has very weak oxygen vacancy
) signal at g = 2.001,[9] while strong Fe3+ signals at g = 1.994
products (0.03  x 
signal implies that the single
is an oxygen-deficient product, which is consistent with
3
3
(V
O
can be observed from those Fe
.10).[41,42] The very weak V
SrTiO
x
Sr1-xTiO
3
0
O
3
the above XRF analysis result, from which it can be seen that
the O-atom molar ratio is lower than the stoichiometric ratio
(
Table 1); while the strong Fe3+ signal at g = 1.994 can be due to
the doped Fe3+ ions in Fe since Fe2+ ion has no EPR
signal. By considering that O-atom molar ratios of those Fe
TiO products (0.03  x  0.10) are slightly higher than the
stoichiometric ratio (Table 1), it can be concluded that Fe3
doping would not only reduce V sites but also cause additional
x
Sr1-xTiO
3
x 3
Figure 4. Typical FESEM images of the obtained Fe Sr1-xTiO products with x
x
Sr1-
= 0 (a), 0.05 (b), 0.10 (c), 0.20 (d).
x
3
+
x
3
The TEM image (Figure 5c) of Fe Sr1-xTiO product with x =
O
0.10 indicates that it is still mainly composed by quasi-sphere-
O2- ions located at the interstitial positions as mentioned above,
and therefore the gradually enhanced EPR signal intensities
like nanoparticles but with smaller sizes compared to the single
3
SrTiO . From the HRTEM image of the nanoparticle with large
along with the increases of x value of Fe
mainly contributed by the doped Fe3+ ions. Although very limited
Fe3+ ions can replace the Sr2+ sites of the cubic SrTiO
lattice
and those Fe products show very similar bulk defect
concentrations as mentioned above, the EPR signal intensities
and PAS τ values still show an increasing trend along with the
x
Sr1-xTiO
3
products are
size (Figure 5d), it can be found that the d-spacings (ca. 0.272
and 0.221 nm) of lattice fringes (as shown in the white dot
square in Figure 5d) are slightly lower than the d-spacings
(0.281 nm of (110) plane and 0.231 nm of (111) plane) of cubic
3
x
Sr1-xTiO
3
SrTiO
3
, and thus it could be assigned to the (110) and (111)
, respectively. Moreover, the FFT
2
planes of Fe-doped SrTiO
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
surfaces. Moreover, the addition of Fe3+ ions not only leads to
the formation of Fe-doped SrTiO , but also results in the
3
pattern (inset in the upper of Figure 5d) demonstrates that the
nanoparticles with large sizes has cubic crystal phase similar to
. It implies that Fe3+ ions with smaller radii
changes of microstructures and morphology of products. By
considering that the bulk Fe3+ doping amount is very limited, and
most doped Fe3+ ions would mainly distribute on the particle
surfaces, which might be one of the reasons that the particle
sizes of those Fe Sr1-xTiO products are obviously decreased
x
3
along with the increase of x value (Figure 4).
the single SrTiO
might replace the Sr sites, and thus lead to the decreases in d-
spacings of the cubic SrTiO , which is similar to the situation of
reported previously.[32] Similarly, the ultrafine
Cr-doped SrTiO
nanoparticle with much smaller size can also be assigned to Fe-
doped SrTiO (Figure 5e). Nevertheless, the Fe
3
3
3
3
x
Sr1-xTiO
3
product with x = 0.20 contains much less large particles but co-
existed with tremendous ultrafine nanoparticles (Figure 5f),
which is consistent the observation of FESEM image (Figure 4d).
As can be seen from the HRTEM images (Figure 5g), those
3
ultrafine nanoparticles still can be asssigned to Fe-doped SrTiO .
Moreover, the HRTEM image (Figure 5h) of the ultrafine
nanocrystals (NCs) shows that the d-spacing (ca. 0.268 nm) of
lattice fringes is very similar to that of the (104) plane of α-Fe
and the FFT pattern (inset in Figure 5h) demonstrates that it has
body-centred cubic crystal phase of α-Fe
,[44] which is in
2 3
O ,
2
O
3
accordance with the above FESEM and XRD results.
Figure 6. Typical liquid nitrogen adsorption-desorption isotherms (a) and
Barret-Joyner-Halenda (BJH) pore size distribution plots (b) of the obtained
x 3
Fe Sr1-xTiO products with different x values.
To further investigate the effects of Fe3+ ions on the micro-
structures, liquid adsorption-desorption experiments are
conducted and shown in Figure 6. All Fe products
exhibit type III isotherms with H3 hysteresis loops (Figure 6a),
which are similar to the single SrTiO . Especially, more obvious
jumps of N desorption branches starting at a high partial
pressure (P/P
products with 0.03  x  0.10, which can be due to the N
N
2
x
Sr1-xTiO
3
3
2
0
= 0.90) are observed from those Fe
x
Sr1-xTiO
3
2
condensation in the intercrystalline pores among the stacked
nanoparticles. Somewhat differently, the Fe product
with x = 0.20 delivers a more significant jump of N desorption
branche starting at a much lower partial pressure (P/P = 0.66).
x
Sr1-xTiO
3
Figure 5. Typical TEM (a, c, f) and HRTEM (b, d, e, g, h) images of the
2
x 3
obtained Fe Sr1-xTiO products with x = 0 (a, b), 0.10 (c, d, e), 0.20 (f, g, h).
0
It indicates that the stacked pore structures of Fe
product with x = 0.20 are different from that of Fe
products with 0.03  x  0.10. Similarly, those Fe
products with 0.03  x  0.10 display a slightly increasing trend
in the Brunauer-Emmett-Teller (BET) specific surface area (SBET
x
Sr1-xTiO
Sr1-xTiO
Sr1-xTiO
3
3
3
The above results indicate that Fe3+ ions with low contents
are mainly doped into the cubic SrTiO lattice and cause the
, and excessive Fe3+ ions in
decreases in the d-spacing of SrTiO
Fe products with x > 0.10 will transfrom into α-Fe
NCs, which located on the Fe-doped SrTiO nanoparticle
x
3
x
3
x
Sr1-xTiO
3
2 3
O
,
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
2
-1
from 9.9 to 13.9 m g ) along with the increase of x value from
reaction conditions for N
2
photoreduction were optimized under
2
-1
0
.03 to 0.10, which are higher than that (6.2 m g ) of the single
SrTiO (Figure 6a), while the Fe product with x = 0.20
exhibits a much higher SBET of 30.6 m g .
Xe-lamp irradiation without sacrificial reagent, and the
quantitative determination of NH production in water was
conducted by color reaction with Nessler’s reagent.[8] Control
experiments demonstrate that no NH was detected from the
irradiated suspension system when Ar flow replaces
suggesting the NH production are derived from N reduction,
and those Fe products has no N-species for NH
production. Under the optimized photoreaction conditions, the
single α-Fe shows no obvious NH production activity as
shown in Figure 7a, while the single SrTiO delivers very limited
NH
production activity (9.4 μmol g-1 h-1). All Fe
3
x
Sr1-xTiO
3
3
2
-1
Also, the Barret-Joyner-Halenda (BJH) pore size distribution
plots (Figure 6b) determined from the desorption branches
3
2
N ,
reveal that those Fe
x
Sr1-xTiO
3
products with 0  x  0.10 have an
3
2
increasing trend in the BJH desorption cumulation pore volumes
x
Sr1-xTiO
3
3
3
-1
(
V
De, from 0.017 to 0.086 cm g ) and a decreasing one in the
average pore diameters (from 45.7 to 28.3 nm); while the Fe
x
Sr1-
2
O
3
3
x
TiO
3
product with x = 0.20 has much higher pore volume (VDe
=
3
3
-1
0
.185 cm g ) and smaller pore diameter (22.5 nm). Since the
3
x
Sr1-xTiO
3
addition of Fe ions would obviously change the particle sizes
and its disturbutions of those Fe products with x  0.10
Figure 4a-c), while the Fe products with x > 0.10 will
transform into α-Fe /Fe-doped SrTiO composites with much
less large particles co-existed tremendous ultrafine
products (0.03  x  0.20) have better ability to convert N
2
to
, indicating that Fe3 doping can improve the N
+
fixation
x
Sr1-xTiO
3
NH
performance of SrTiO
With the increase of x value from 0.03 to 0.10, those Fe
3
2
(
x
Sr1-xTiO
3
3
.
2
O
3
3
x
Sr1-
x
TiO
3
products reveal an increasing N
2
fixation activity, and the
product with x = 0.10 achieves the highest N
production activity of 30.1
μmol g-1 h-1, which is 3.2 times higher than that of the single
SrTiO . Once the x value is larger than 0.10, the average NH
production activity shows
excessive Fe3+ ions in Fe
nanoparticles (Figure 5f), it would be reasonable to think that the
increasing SBET values can be related to those changes in
particle sizes and its disturbutions, which then lead to the
enhanced pore volumes and SBET values synergistically.
Fe
x
Sr1-xTiO
3
2
fixation ability with an average NH
3
3
3
a
decreasing trend. Since the
x
Sr1-xTiO
3
products (x > 0.10) would
NCs, it
production
product with x  0.10 might be due to the
Fe3+ doping, while the following decrease may be due to the
newly form -Fe NCs.
lead to the formation of composites containing α-Fe
can be concluded that the gradually enhanced NH
activity for Fe
2 3
O
3
x
Sr1-xTiO
3
2
O
3
Figure 8. Cycling photocatalytic behavior towards NH
3
production over the
Fe Sr1-xTiO product (x = 0.10) under Xe-lamp illumination.
x
3
As reported previously, the photocatalytic N
could be described as N reduction for NH production and H
oxidation for O
generation.[45] Therefore, we used a sealed
reactor connected with an on-line gas chromatograph to detect
the generated O in the reaction system, in which an automatic
2
fixation reaction
2
3
2
O
Figure 7. (a) Average NH
products (x = 0-0.20) under Xe-lamp illumination for 2 h. (b) Photocatalytic
/NH production amounts over the Fe Sr1-xTiO product (x = 0.10) in N and
3
production activity of the obtained Fe
x
Sr1-xTiO
3
2
O
2
3
x
3
2
2
Ar atmosphere under Xe-lamp illumination for 90 min.
sampling system was used to avoid the influence of O
Figure 7b depicts the time-dependent curves of
2
in air.
/NH
3
O
2
production amounts over the Fe
in the sealed reactor. As can be seen, both O
production amounts gradually increase along with prolonging the
x
Sr1-xTiO
3
product with x = 0.10
Photocatalytic Activity Analyses
2
and NH
3
x
3
By using those Fe Sr1-xTiO products as photocatalysts, the
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
irradiation time, and the average O
are 23.2 and 30.1 mol g-1 h-1, respectively. The generated
/NH molar ratio (0.77) is very closely to the theoretic
stoichiometric ratio (0.75) of the chemical reaction equation for
fixation reaction as shown in Figure 7b. It infers that the
present Fe product possesses suitable band edges to
implement N reduction and H O oxidation at the same time.
The control experiments (Figure 7b) demonstrate that the Ar
2
and NH
3
production rates
0.10) has the most efficient charge separation process among
those Fe-doped products, which is in good agreement with the
changing trend of photocatalytic performance shown in Figure
7a. The increasing charge separation efficiency might be
contributed by more Fe3+ ions doped at surface sites along with
enhancing the x value from 0.03 to 0.10 as mentioned above,
which can promote the photogenerated electron transfer of Fe-
O
2
3
N
2
x
Sr1-xTiO
3
2
2
doped SrTiO
3
,[15,47,48] while the decreasing charge one of Fe
products with x > 0.10 can be due to the newly formed -
3
NCs, which might act as charge recombination centers.
x
Sr1-
instead of N
reaction since there is no obvious O
confirms that the generated O and NH
oxidation and N
Fe . Moreover, the photoactivity of Fe
.10) is relatively stable during total 15 h irradiation for the
2
atmosphere can terminate the N
2
photofixation
x
TiO
3
2
and NH
3
generated. It
Fe
2
O
2
3
2
are originated from H O
2
reduction reactions caused by the photoexcited
(x =
x
Sr1-xTiO
3
x
Sr1-xTiO
3
0
successive 3 recycles (Figure 8), and the very similar XRD
patterns (Figure S4) and high-resolution XPS spectra (Figure S5)
of Fe
also demonstrate the present Fe
good photostability. The above features such as stable and
efficient N fixation ability are of significance in the utilization of
x
Sr1-xTiO
3
(x = 0.10) before and after 15 h photoreaction
x
Sr1-xTiO
3
product has relatively
2
the solar spectrum for artificial photosynthesis with H
2
O as
electron and proton sources.[45]
To further clarify the N source of the formed NH
labelling experiments are performed by using a commercially
, 15N isotopic
3
available 15N
MS study, the produced 15NH +
gas with purity of 99%. For facilitating the LC–
2
15
4
was transferred into N labeled
indophenol,[15,46] and its corresponding mass spectra are shwon
in Figure S6. As seen, a strong 15N labeled indophenol cation
mass spectroscopy signal can be observed at 201 m/z in the
LC-MS spectra of the indophenol derived from the photoreaction
system with 15N
atmosphere. This signal intensity is significantly
2
higher than the 14N/ N naturasl abundance ratio,[15] suggesting
15
that N
2
is the N source of those formed NH
fixation system.
3
molecules in the
present photocatalytic N
2
Figure 9. (a) Photoluminescence (PL) spectra of the Fe
0-0.20) with excitation wavelength of 280 nm; (b) photocurrent-time curves
of the obtained Fe Sr1-xTiO products (x = 0-0.20) under Xe-lamp illumination.
x 3
Sr1-xTiO products (x
Photoelectrochemcial and Surface Property Analyses
=
To explore the effects of Fe doping on the N
Fe products, their photoluminescence (PL) spectra and
photocurrent response curves are determined and shown in
Figure 9. As seen from Figure 9a, the Fe products
show new PL peak at shorter wavelength region in addition to
that of the single SrTiO , which can be due to the defect states
stemmed from Fe doping in the SrTiO lattice. Interestingly
2
fixation ability of
x
3
x
Sr1-xTiO
3
Nevertheless, the single SrTiO
intensity and higher photocurrent responses than those Fe
TiO (Figure 9), suggesting that the single SrTiO has the
lowest defect concentration and charge recombination efficiency
and Fe3+ doping would generate defect states in the SrTiO
lattice to act as recombination sites,[47] which is contrary to the
actual result that the single SrTiO shows very limited activity for
fixation (Figure 7a). One of the main reasons is that the
single SrTiO has larger particle sizes and lower surface area
than those Fe products (Figure 6), which will not
benefit to the adsorption and activation of N molecules since
the majority difficulty of N fixation is the activation of high-
3
shows much lower PL peak
x
Sr1-xTiO
3
x
Sr1-
x
3
3
3
3
3
enough, the PL peak intensity shows a decreasing trend with
enhancing the x value from 0.03 to 0.10, and then an increasing
one upon further enhancing the x value from 0.10 to 0.20.
3
N
2
x
3
Similarly, the photocurrent responses of Fe Sr1-xTiO products
3
show an increasing trend with enhancing the x value from 0.03
to 0.10, and then a decreasing one upon further enhancing the x
value (Figure 9b). These results suggest that the charge
separation efficiency exhibits an increasing trend along with the
increase of x value from 0.03 to 0.10, and then a decreasing one
x
Sr1-xTiO
3
2
2
energy N≡N bond. To further investigate this issue, in situ diffuse
reflectance Fourier transform infrared (in situ FTIR) spectra are
applied to identify the reaction intermediates. As shown in Figure
x
3
with further enhancing the x value, and the Fe Sr1-xTiO (x =
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
10a, the single SrTiO
3
after 15 min irradiation in N
2
and water
x
3
EPR signal at g = 1.994 of Fe Sr1-xTiO (x = 0.10) obviously
vapor atmosphere shows some absorption bands, in which the
two bands at 1685 and 1596 cm-1 can be assigned to the σ(N-H)
bending mode,[8] and the broad band at 1530 cm-1 can be
decreases after light irradiation (Figure 10b), suggesting that
some surface doped Fe3+ ions can trap the photogenerated
electrons of Fe-doped SrTiO
response. This evidence demonstrates that the Fe
0.10) could combine with N molecules through a chemisorption
process, and then transfer the electrons toward those adsorbed
molecules to weaken the N≡N triple bond for the following
formation.
To further investigate the N
processes, N temperature-programmed desorption (N
3
to generate Fe2+ ions with no EPR
attributed to the characteristic absorption NH
while the Fe (x = 0.10) after the irradiation time
displays several new absorption bands in addition to the
absorption bands of σ(N-H) bending mode and NH molecules.
3
molecules;[9,49]
x
Sr1-xTiO
3
(x =
x
Sr1-xTiO
3
2
3
N
2
Among which, the new broad band at 1301 cm-1 can be due to
the characteristic absorption of σ(O-H) bending mode,[8] and the
very strong band at 1630 cm-1 can be assigned to the
NH
3
2
chemisorption and activation
-TPD)
2
2
chemisorbed N
2
molecules, which is similar to the situation of
experiments are conducted and shown in Figure 11a. The single
Br-NT reported previously.[9,50]
SrTiO
only shows a strong N
desorption peak at 79 C with
o
Bi
5
O
7
3
2
o
no obvious TCD signal after 200 C, inferring its weak N
2
physisorption ability.[22,51,52] The Fe
products (x = 0.05
x
Sr1-xTiO
3
and 0.10) also show a strong N
2
desorption peak at slightly
enhanced temperature (90-95 oC) but with new low-intensity
peaks at 380 °C and 450 °C, which might be due to the
molecules.[
22,51]
Especially,
different chemisorption modes of N
the Fe
300 °C and 680 °C, implying that those surface Fe3+ doped
2
x
3
Sr1-xTiO product (x = 0.10) exhibits additional peaks at

sites could further promote the chemisorption of
molecules.[22,52] The above results indicate that the Fe
N
2
x
Sr1-xTiO
3
product with x =0.10 shows better adsorption and activation
ability of N
reactions.
2 2
molecules, which are beneficial for the N fixation
Figure 10. (a) In situ FTIR difference spectra recorded during the
photofixation over the Fe Sr1-xTiO products (x = 0, 0.10) before and after 15
min irradiation of Xe-lamp; (b) EPR spectra of the Fe Sr1-xTiO product (x =
.10) before and after 15 min irradiation of Xe-lamp.
N
2
x
3
x
3
0
2
Although N molecules themselves have no signal in FTIR
spectra because of their balanced construction, their even
balance can be broken through the formation of chemisorbed
species.[9,50] The above newly appeared broad bands at 1301
cm-1 (σ(O-H)) and 1630 cm-1 (chemisorbed N
Fe (x = 0.10) after light irradiation suggest that Fe-
doping can promote to the chemisorption of H O and N
molecules. It has reported that Fe3+ doping sites can not only
chemisorb but also activate the N molecules by promoting the
interfacial electron transfer from g-C to N molecules, thus
significantly improving the N
fixation ability.[15] The strong Fe3+
molecules) for the
2
x
Sr1-xTiO
3
2
2
2
Figure 11. (a) N
2
-TPD results of the Fe
x 3
Sr1-xTiO products (x = 0-0.1.0); (b)
3
N
4
2
DRS spectra of the Fe
x
Sr1-xTiO products (x = 0-0.20) and the single -Fe O .
3
2 3
2
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
Energy band structure and mechanism analyses
composites containing α-Fe
2
O
3
3
NCs and Fe-doped SrTiO ,
which would lead to the formation of type I heterojunction
according to their potential energy diagrams (Figure 12a), and
To further explore the different photocatalytic performance
x
3
among those Fe Sr1-xTiO products (x = 0-0.20), their energy
the photogenerated carriers of Fe-doped SrTiO
to those -Fe NCs, thus causing the decreased activity of N
fixation as shown in Figure 7a. Nevertheless, those Fe
products with x  0.10 have no such -Fe
charge recombination sites, and therefore deliver much better
ability for N fixation. As shown in Figure 12b, the best
photoactivity of Fe (x = 0.10) can be due to the suitable
band edges to implement N reduction and H O oxidation since
its ECB level is more negative than the potential of N reduction
to NH through the 6e reaction and its EVB is more positive than
the potential of H O oxidation to O (Figure 12b). The efficient N
fixation of this product might be due to the surface Fe3+ doped
sites, which can chemisorb and activate N molecules, and then
transfer the photogenerated electrons from Fe-doped SrTiO to
the adsorbed N molecules as mentioned above. Moreover, the
increasing N fixation activity along with the increase of x value
from 0.03 to 0.10 can be due to the increased surface Fe3+
3
would transfer
band structures are investigated. As can be seen from the UV-
vis diffuse reflectance absorption spectra (DRS) shown in Figure
2
O
3
2
3
x
Sr1-xTiO
1
1b, the single SrTiO
edges at 385 and 700 nm, corresponding to a bandgap
energy (E ) of ∼3.22 and ∼1.77 eV, which are comparable to the
reported values (~3.20 eV [52] for SrTiO and ∼1.90 eV [53] for -
Fe ). With the increase of x value, those Fe products
display slight redshift with the sample color changing from white
to light brown, and the E values of Fe product with x =
.10 can be calculated to be ∼2.67 eV from the onset absorption
edge (465 nm). Also, the flat-band potentials (Efb) of -Fe
SrTiO and Fe (x = 0.10) are determined from the Mott-
3
and -Fe
2 3
O show onset absorption
2
3
O NCs to act as
g
2
3
x
Sr1-xTiO
3
2
O
3
x
Sr1-xTiO
3
2
2
2
g
x
Sr1-xTiO
3
3
0
2
2
2
2
3
O ,
3
x
Sr1-xTiO
3
2
Schotty plots as shown in Figure S7. As seen, those Mott-
Schotty plots show positive slope, implying their n-type
semiconducting property. According to the intercepts with x-axis,
3
2
2
2 3 3 3
x
the Efb value of -Fe O , SrTiO and Fe Sr1-xTiO (x = 0.10) can
be determined to be -0.39, -0.50 and -0.45 V vs. NHE,[54]
respectively. Since the conduction band (CB) is usually 0.1 V
higher than Efb for n-type semiconductor, their ECB levels can be
calculated to be -0.49, -0.60 and -0.55 V,[44] respectively.
3
doped sites on the Fe-doped SrTiO .
Conclusions
Therefore, the valance band position (EVB) of -Fe
and Fe Sr1-xTiO (x = 0.10) can be calculated to be 1.28, 2.62
and 2.12 V according to the formula of E
respectively. Based on the above results, a potential energy
diagram and electron transfer path for Fe products’
photocatalytic process can be drawn as Figure 12.
2 3 3
O , SrTiO
x
3
In summary, we demonstrate that Fe-doped SrTiO
TiO ) products (0  x  0.20) are potential photocatalysts for
NH synthesis via N photofixation process without the aid of
sacrificial reagent or metal co-catalyst. Among which, those
Fe products with x > 0.10 transform into composites
containing Fe-doped SrTiO and α-Fe nanocrystals, which
might act as charge recombination sites, and thus causes a
decreased N photofixation activity. However, those Fe
products (0.03  x  0.10) reveal an increasing N photofixation
activity along with the increase of x value from 0.03 to 0.10, and
the Fe (x = 0.10) achieves the highest N photofixation
ability with an average NH
production activity of 30.1 μmol g-1 h-
1,
which is 3.2 times higher than that of the single SrTiO . Once
the x value is larger than 0.10, the average NH production
activity shows a decreasing trend. Moreover, the present Fe
TiO product possesses suitable band edges to implement N
reduction and oxidation at the same time. Further
investigations demonstrate that the surface Fe3+ doping sites
can not only chemisorb and activate the N molecules, but also
promote the interfacial electron transfer from Fe-doped SrTiO to
molecules, thus significantly improve the N photofixation
ability of SrTiO . The above features such as stable and efficient
photofixation ability are of significance in sunlight utilization
for artificial photosynthesis and renewable energy conversion
with H O as electron and proton sources.
3
x
(Fe Sr1-
g
= EVB – ECB,
x
3
3
2
x
Sr1-xTiO
3
x
Sr1-xTiO
3
3
2 3
O
2
x
Sr1-xTiO
3
2
x
Sr1-xTiO
3
2
3
3
3
x
Sr1-
x
3
2
2
H O
Figure 12. Electronic energy-level diagram of α-Fe
2
O
3
and Fe-doped SrTiO
production via N
3
2
2
as well as the proposed reaction mechanism of NH
photofixation process.
3
3
N
2
2
3
+
Compared with the one electron reduction process (N
e- → N H, E = -3.20 V vs. NHE) shown in Figure 12, the
multielectron reduction process (N
.27 V vs. NHE) may avoid the generation of high-energy
intermediate and decrease the thermodynamic barrier for NH
production.[45] As mentioned above, the excessive Fe3+ ions in
those Fe products with x > 0.10 would transform into
2
+ H +
N
2
2
+8H+ + 6e → 2NH
-
+
, E =

2
4
2
0
3
Experimental Section
x
Sr1-xTiO
3
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
Material Preparation: Fe
hydrothermal process followed by calcination at 700 °C. Typically,
Sr(NO and Fe(NO ·9H O with the stoichiometric ratio of Fe Sr1-xTiO
0  x  0.20) were dissolved into 2.5 mL of ethylene glycol (EG) pre-
x
Sr1-xTiO
3
products were prepared via
a
As for the preparation of work electrode, 1 mg of sample was added into
0.5 mL of Nafion ethanol solution (0.05wt%) under stirring and sonication
for 15 min, and then the resultant slurry was dripped on clean fluorine-
doped tin oxide (FTO) glass. After air-drying for at least 30 min, the film
electrode with 1.0 cm2 active area was readied for photocurrent. In a
3
)
2
3
)
3
2
x
3
(
o
4 9 4
heated to 85 C, and then Ti(C H O) (2.0 mmol) was dropped into the
solution to obtain a gel. After that, 30 mL of NaOH solution (5.0 M) was
added into the gel and gained a solution after vigorously stirring. The
resultant solution was put into a Teflon-lined stainless autoclave (volume
of 40 mL), and then heated up to 200 oC for 24 h. The resultant
precipitate was separated and washed with deionized water for 3-5 times
typical run, the three electrodes were immersed into a Na
(0.5 M). Before irradiation, the suspension was continuously purged by
to remove O , and then illuminated by light from 300 W Xe-lamp. For
2 4
SO solution
N
2
2
the measurement of flat-band potential (Efb), the conditions were almost
same as the test of transient photocurrent curve, in which the reference
o
θ
and dried in the vacuum oven at 70 C for 10 h. The as-prepared product
was then grinded and calcined at 700 oC for 4 h so as to remove the
electrode was changed to Ag/AgCl (Na
2
SO
4
solution (0.5 M), E = 0.198
V vs. NHE), and the Efb value was determined by the Mott-Schottky
remnant N species. For comparison, single -Fe
O
2 3
was prepared in the
relation:
same reaction condition by enhancing Fe source to 2.0 mmol without
adding Sr and Ti sources.
C-2 = (2/eεε
N )[V – Efb – kT/e]
0 d a
Material Characterization: The crystal phase of products were
characterized using a Bruker D8 advance X-ray diffractometer (XRD)
with Cu Kα irradiation (λ = 0.154178 nm) at 40 kV and 40 mA with a scan
rate of 1° min-1 in the range of 2 = 20°-70°. The morphology was
investigated with Zeiss-Sigma field emission scanning electron
microscope (FESEM), and high-resolution transmission electron
where, C is space charge layer capacitance, e is electron charge, ε is
dielectric constant, ε is permittivity of vacuum, N is electron donor
density, V is the applied potential, and Efb is flat band potential.
0
d
a
2
Photoactivity Measurement: The N photofixation experiment was
carried out in a home-made vessel with a 300 W Xe-lamp irradiation.
Generally, 15 mg of catalyst was dispersed into 150 mL of deionized
microscopic (HRTEM) observation was conducted on a LaB
6
JEM-2100
water. Before the irradiation,
suspension for 20 min to remove the dissolved
photocatalytic test, the vessel was cooled down to room temperature with
flow of cooling water. The NH production amount was
spectrophotometrically determined using Nessler’s reagent. All samples
have conducted blank test using Ar instead of N at the first 1 h, which
confirmed all color development reaction was stemmed from the N
fixation but not the false appearance by the photocatalyst itself. For the
determination of O production amount, the photoreaction was conducted
at a sealed reactor, then analyzed using an on-line gas chromatograph
N
2
gas flow was bubbled into the
(
HR) electron microscope (JEOL) working at 200 kV. UV-vis diffuse
2
O .
During the
reflectance absorption spectra (DRS) were recorded with a Shimadazu
UV-2550 spectrophotometer equipped with an integrating sphere using
BaSO
using a VG Multilab 2000 spectrometer (Thermo Electron) with Al Kα
source. Elemental compositions were detected using Bruker S4
Pioneer X-ray fluorescence (XRF) spectrometer with a Rh target and
without standard sample. Positron annihilation spectrometry (PAS) was
measured using a fast-fast coincidence system with time resolution of
a
3
4
as the reference. X-ray photoelectron spectra (XPS) was obtained
[
8]
2
a
2
2
220 ps and a 20 μCi 22Na positron source was sandwiched between two

(
Shimadazu GC-2010 Plus, BID detector, He as carrier gas).
pieces of identical sample. Electron paramagnetic resonance (EPR)
spectra were collected for power samples on a Bruker X-band A-200
spectrometer at room temperature. Mossbauer spectra were measured at
room temperature using a conventional Wissel MS-500 spectrometer
The 15N isotopic labeling experiments are performed by using a
commercially available 15N
gas with purity of 99%, which was
2
purchased from Wuhan Newradar Special Gas Co., Ltd. Typically, the
photoreactor containing photocatalyst aqueous suspension was
vacuumed and washed by pure Ar gas to eliminate the air and the
and then 15N
possible adsorbed NH , gas was bubbled into the
3 2
suspension until it filled with the isotopic labelled gas. The other
photoreaction conditions were the same as those for the photocatalytic
(
Germany) in transmission geometry with constant acceleration mode. A
5
7Co(Rh) source with activity of 25 mCi was used. The velocity calibration
was done with a room temperature α-Fe adsorber, and the Mossbauer
spectra were fitted by the software Recoil using Lorentzian Multiplet
Analysis.
14N
fixation reaction. The produced 15NH +
amount was estimated using
is too low to be detected
2
4
Photoluminescence (PL) spectra were carried out on a Hitachi Model
indophenol method since the mass of 15NH +
4
2
F-6800 fluorescence spectrophotometer. Liquid N adsorption-desorption
using liquid chromatograph-mass spectra (LC–MS) technique.[15,46]
Therefore, the sample for LC-MS analysis was prepared by using the
photoreaction solution (2.0 mL) reacted with NaOH solution (2.0 mL, 1.0
M) containing salicylic acid (5.0wt%) and sodium citrate (5.0wt%),
followed by addition of NaClO solution (1.0 mL, 0.05 M) and sodium
nitroferricyanide (0.2 mL, 1.0wt%). The LC-MS analysis was performed
on an LTQ Orbitrap Elite mass spectrometer (Thermo Fisher Scientific,
USA) coupled with an UltiMate 3000 UHPLC System (Thermo Fisher
Scientific, USA). The LC separation was performed under 40 °C on a
columns of a Waters Acquity UPLC BEH C18 (100 × 2.1 mm, 1.7 μm).
Analytes were eluted in buffer A (20% MeOH in water) and buffer B (1%
ammonium formate in water). The high resolution MS analysis was
performed under positive ion mode with full scan detection. The source
and ion transfer parameters applied were as follow: heater temperature,
isotherms at 77 K were measured by a Micrometrics ASAP 2460 system.
In situ diffuse reflectance infrared Fourier transform spectra (DRIFTS)
were recorded on
equipped with a liquid nitrogen cooled HgCdTe (MCT/A) detector, 4 cm
a Nicolet Is50 spectrometer (Thermo Electron)
-1
by using 64 scans. In general, appropriate amount photocatalyst was
uniformly placed in the center of designed reaction cell, then thoroughly
Ar purge and 200 oC heat treatment for 1 h, which was used to remove
all the air in the reaction cell. After the initial heat treatment of catalyst,
the water and N
small amount of water vapor using N
and IR signal was in situ collected through a MCT detector along with the
reaction. N temperature programmed desorption (N -TPD) curves were
determined by using Micromeritics Autochem 2. The sample was firstly
heated to 400 oC in flowing Ar for 1 h, and purge high pure N
for 1 h
under the room temperature, and then He purge the sample for 30 min to
decrease the physical adsorbed N on the surface. Finally, the sample
2
was pre-adsorbed on the surface through bubbling a
2
flow. At last, light was turned on
2
2
2
300 °C, capillary temperature, 350 °C, sheath gas flow, 35 arbitrary units;
auxiliary gas flow, 15 arbitrary units; spray voltage, 3.5 kV; capillary
2
6
o
o
-1
voltage, 35 V; S-lens RF level, 60%; automatic gain control, 1×10 ; the
heated to 800 C at a heating rate of 10 C cm under He flowing.
Transient photocurrent curve was obtained by using a CHI model
maximum injection time, 100 ms.
618C electrochemical analyzer and a standard three-electrode system, in
which graphite electrode, calomel normal electrode and sample electrode
acted as the counter electrode, reference electrode and work electrode.
Acknowledgements
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
[
[
[
[
24] D. E. Cox, W. J. Takei, R. C. Miller, G. Shirane, J. Phys. Chem. Solids
This work was supported by the National Natural Science
Foundation of China (21573166, 21271146, 20973128 and
1963, 23, 863-874.
25]T. Ishiguro, K. Tanaka, F. Marumo, M.G.M.U. Ismail, S. Hirano, S. Sōmiya,
Acta Crystallogr. B 1978, 34, 3346-3348.
26] S. Zhang, J. Liu, Y. Han, B. Chen, X. Li, Mat. Sci. Eng. B 2014, 110, 11-
20871096), the Science & Technology Program of Science,
Technology and Innovation Commission of Shenzhen
Municipality (JCYJ20180302153921190), the Fund for Creative
Research Groups of Hubei Province (2014CFA007), and the
Natural Science Foundation of Jiangsu Province (BK20151247),
China.
17.
27] F. Craciun, E. Dimitriu, M. Grigoras, N. Lupu, Appl. Phys. Lett. 2013, 102,
242903.
28] R. D. Shannon, Acta Crystallogr. 1976, 32, 751-767.
29] R. P. Vasquez, J. Electron Spectrosc. 1991, 56, 217-240.
30] V. I. Nefedov, Y. V. Salyn, G. Leonhardt, R. Scheibe, J. Electron
Spectrosc. 1977, 10, 121-124.
[
[
[
Keywords: Fe-doped SrTiO
3
•
Dinitrogen photofixation
•
[31] D. N. Hendrickson, J. M. Hollander, W. L. Jolly, Inorg. Chem. 1969, 8,
642-2647.
32] H. Yu, S. Ouyang, S. Yan, Z. Li, T. Yu, Z. Zou, J. Mater. Chem. 2011, 21,
1347-11351.
33] D. Jiang, J. Li, C. Xing, Z. Zhang, S. Meng, M. Chen, ACS Appl. Mater.
Inter. 2015, 7, 19234-19242.
34] S. Zhang, H. Yang, H. Gao, R. Cao, J. Huang, X. Xu, ACS Appl. Mater.
Inter. 2017, 9, 23635-23646.
35] M. Humayun, L. Xu, L. Zhou, Z. Zheng, Q. Fu, W. Luo, Nano Res. 2018,
2
Photocatalysis • N molecule activation
2
[
[
[
[
[
1
[
1] J. G. Chen, R. M. Crooks, L. C. Seefeldt, K. L. Bren, R. M. Bullock, M. Y.
Darensbourg, P. L. Holland, B. Hoffman, M. J. Janik, A. K. Jones, M. G.
Kanatzidis, P. King, K. M. Lancaster, S. V. Lymar, P. Pfromm, W. F.
Schneider, R. R. Schrock, Science 2018, 360, 6611.
[
[
2] G. N. Schrauzer, T. D. Guth, J. Am. Chem. Soc. 1977, 99, 7189-7193.
3] W. Zhao, J. Zhang, X. Zhu, M. Zhang, J. Tang, M. Tan, Y. Wang, Appl.
Catal. B Environ. 2014, 144, 468-477.
11, 6391-6404.
36]P. D. Schulze, S. L. Shaffer, R. L. Hance, D. L. Utley, J. Vac. Sci. Technol.
A 1983, 1, 97-99.
[
[
[
4] J. Yang, Y. Guo, R. Jiang, F. Qin, H. Zhang, W. Lu, J. Wang, J. C. Yu, J.
Am. Chem. Soc. 2018, 140, 8497-8508.
5] W. Zhao, H. Xi, M. Zhang, Y. Li, J. Chen, J. Zhang, X. Zhu, Chem.
Commun. 2015, 51, 4785-4788.
6] N. Zhang, A. Jalil, D. Wu, S. Chen, Y. Liu, C. Gao, W. Ye, Z. Qi, H. Ju, C.
Wang, X. Wu, L. Song, J. Zhu, Y. Xiong, J. Am. Chem. Soc. 2018, 140,
[
[
[
37] C. R. Brundle, T. J. Chuang, K. Wandelt, Surf. Sci. 1977, 68, 459-468.
38] M. Multani, Mat. Res. Bull. 1984, 19, 25-34.
39] Y. Zhao, Y. Zhao, G. I. N. Waterhouse, L. Zheng, X. Cao, F. Teng, L. Z.
Wu, C. H. Tung, D. O'Hare, T. Zhang, Adv. Mater. 2017, 29, 1703828.
40] D. J. Keeble, S. Wicklein, R. Dittmann, L. Ravelli, R. A. Mackie, W. Egger,
Phys. Rev. Lett. 2010, 105, 226102.
[
[
[
9434-9443.
[
[
[
7] H. Li, J. Shang, J. Shi, K. Zhao, L. Zhang, Nanoscale 2016, 8, 1986-1993.
8] H. Li, J. Shang, Z. Ai, L. Zhang, J. Am. Chem. Soc. 2015, 137, 6393-6399.
9] S. Wang, X. Hai, X. Ding, K. Chang, Y. Xiang, X. Meng, Z. Yang, H. Chen,
J. Ye, Adv. Mater. 2017, 29, 1701774.
41] K. Nagaveni, M. S. Hegde, G. Madras, J. Phys. Chem. B 2014, 108,
2
0204-20212.
42] T. Tong, J. Zhang, B. Tian, F. Chen, D. He, J. Hazard. Mater. 2008, 155,
72-579.
43] Y. Bi, M. F. Ehsan, Y. Huang, J. Jin, T. He, J. CO
44] C. Liu, G. Wu, J. Chen, K. Huang, W. Shi, New J. Chem. 2016, 40, 5198-
208.
5
[
[
[
[
[
[
[
[
[
[
[
[
10] Y. Bai, L. Ye, T. Chen, L. Wang, X. Shi, X. Zhang, D. Chen, ACS Appl.
Mater. Inter. 2016, 8, 27661-27668.
11] S. Sun, X. Li, W. Wang, L. Zhang, X. Sun, Appl. Catal. B Environ. 2017,
[
[
2
Util. 2015, 12, 43-48.
5
200, 323-329.
[
[
45] X. Chen, N. Li, Z. Kong, W. J. Ong, X. Zhao, Mater. Horiz. 2017, 5, 9-27.
46] C. D. Lv, C. S. Yan, G. Chen, Y. Ding, J. X. Sun, Y. S. Zhou, G. H. Yu,
Angew. Chem. Int. Ed. 2018, 57, 6073-6076.
12] L. Ye, C. Han, Z. Ma, Y. Leng, J. Li, X. Ji, D. Bi, H. Xie, Z. Huang, Chem.
Eng. J. 2017, 307, 311-318.
13] D. Zhu, L. Zhang, R. E. Ruther, R. J. Hamers, Nat. Mater. 2013, 12, 836-
[47] R. Panda, V. Rathore, M. Rathore, V. Shelke, N. Badera, L. S. S.
841.
Chandra, D. Jain, M. Gangrade, T. Shripati, V. Ganesan, Appl. Surf. Sci.
14] Y. Lu, Y. Yang, T. Zhang, Z. Ge, H. Chang, P. Xiao, Y. Xie, L. Hua, Q. Li,
H. Li, B. Ma, N. Guan, Y. Ma, Y. Chen, ACS Nano 2016, 10, 10507-10515.
15] S. Hu, X. Chen, Q. Li, F. Li, Z. Fan, H. Wang, Y. Wang, B. Zheng, G. Wu,
Appl. Catal. B Environ. 2017, 201, 58-69.
16] S. Hu, Y. Li, F. Li, Z. Fan, H. Ma, W. Li, X. Kang, ACS Sustain. Chem.
Eng. 2016, 4, 2269-2278.
17] M. S. Wrighton, A. B. Ellis, P. T. Wolczanski, D. L. Morse, H. B.
Abrahamson, D. S. Ginley, J. Am. Chem. Soc. 1976, 98, 2774-2779.
18] L. Qing-shui, D. Kazunari, N. Shuichi, O. Takaharu, T. Kenzi, Chem. Lett.
2
012, 258, 5086-5093.
48] K. Li, T. Y. Peng, Z. H. Ying, S. A. Song, J. Zhang, Appl. Catal. B Environ.
016, 180, 130-138.
49]F. Giraud, C. Geantet, N. Guilhaume, S. Gros, L. Porcheron, M. Kanniche,
D. Bianchi, J. Phys. Chem. C 2014, 118, 15664-15676.
[
[
[
2
50] K. Ding, A. W. Pierpont, W. W. Brennessel, G. Lukat-Rodgers, K. R.
Rodgers, T. R. Cundari, E. Bill, P. L. Holland, J. Am. Chem. Soc. 2009,
131, 9471-9472.
[51] C. Li, T. Wang, Z. J. Zhao, W. Yang, J. Li, A. Li, Z. Yang, G. A. Ozin, J.
1
983, 12, 321-324.
19] T. Oshikiri, K. Ueno, H. Misawa, Angew. Chem. Int. Ed. 2014, 53, 9802-
805.
20] T. Oshikiri, K. Ueno, H. Misawa, Angew. Chem. Int. Ed. 2016, 55, 3942-
946.
21] H. Hirakawa, M. Hashimoto, Y. Shiraishi, T. Hirai, J. Am. Chem. Soc.
017, 139, 10929-10936.
Gong, Angew. Chem. Int. Ed. 2018, 57, 5278-5282.
[
[
[
52] Y. Liang, D. A. Bonnell, Surf. Sci. 1994, 310, 128-134.
53] M. Sharon, B. M. Prasad, Sol. Energ. Mater. 1983, 8, 457-469.
54] S. S. Song, Y. Y. Guo, T. Y. Peng, J. Zhang, R. J. Li, RSC Adv. 2016, 6,
9
3
77366-77374.
2
[
[
22] G. Dong, W. Ho, C. Wang, J. Mater. Chem. A 2015, 3, 23435-23441.
23] B. Im, H. Jun, K. H. Lee, J. S. Lee, CrystEngComm 2011, 13, 7212-7215.
This article is protected by copyright. All rights reserved.

European Journal of Inorganic Chemistry
10.1002/ejic.201900098
FULL PAPER
Table of Contents
FULL PAPER
Fe-doped SrTiO
3
(Fe
x
Sr1-xTiO
3
) with x =
2
Photocatalytic N fixation*
0
.10 achieves the best N
2
photofixation
Zihao Ying, Shengtao Chen, Tianyou
Peng,* Renjie Li, Jing Zhang
activity in pure water because those
doped Fe3+ ions have good
chemisorption and activation abilities of
Page No. – Page No.
N
2
molecules, while the α-Fe
in Fe product with x > 0.10
would act as combination sites and
retard the N photofixation reaction.
2 3
O phase
x
Sr1-xTiO
3
Fabrication of Fe-doped SrTiO
photocatalyst with enhanced
dinitrogen photofixation performance
3
2
This article is protected by copyright. All rights reserved.