Efficient photocatalytic reduction of dinitrogen to ammonia on bismuth monoxide quantum dots

Article abstract of DOI:10.1039/C6TA09275F

N₂ reduction to ammonia by solar light represents a green and sustainable ammonia synthesis approach which helps to suppress the global warming and energy crisis. However, conventional semiconductors usually suffer from low activity or poor stability, largely suppressing the application of this technology. Here, we report that bismuth monoxide (BiO) quantum dots with an average size of 2-5 nm exhibited efficient photocatalytic activity for ammonia synthesis under simulated solar light. A highly efficient ammonia synthesis rate of 1226 μmol g^?1 h^?1 is achieved without the assistance of any sacrificial agent or co-catalyst, which is about 1000 times higher than that using the traditional Fe-TiO₂ photocatalyst. Kinetic analysis reveals that the synergy of three low valence surface Bi(ii) species markedly enhances N₂ activation by electron donation, which finally resulted in the highly efficient N₂ photoreduction performance. This work will shed light on designing efficient and robust N₂ reduction photocatalysts.

Full text of DOI:10.1039/C6TA09275F

View Article Online
View Journal
Journal of
Materials Chemistry A
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: W. Wang, S. Sun,
Q. An, L. Zhang, J. Liu and W. A. Goddard, III, J. Mater. Chem. A, 2016, DOI: 10.1039/C6TA09275F.
This is an Accepted Manuscript, which has been through the
Volume
4 Number 1 7 January 2016 Pages 1–330
Royal Society of Chemistry peer review process and has been
accepted for publication.
Journal of
Materials Chemistry A
Accepted Manuscripts are published online shortly after
Materials for energy and sustainability
www.rsc.org/MaterialsA
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 2050-7488
PAPER
Kun Chang, Zhaorong Chang et al.
Bubble-template-assisted synthesis of hollow fullerene-like
MoS2 nanocages as a lithium ion battery anode material
rsc.li/materials-a

Page 1 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
Efficient Photocatalytic Reduction of Dinitrogen to Ammonia on
Bismuth Monoxide Quantum Dots
Songmei Sun1, Qi An2, Wenzhong Wang1*, Ling Zhang1, Jianjun Liu1*, William A.
Goddard III2
1State Key Laboratory of High Performance Ceramics and Superfine Microstructure,
Shanghai Institute of Ceramics, Chinese Academy of Sciences, Shanghai 200050, P.
R. China
2Division of Chemistry and Chemical Engineering, California Institute of Technology,
Pasadena, California 91125, United States
*Address correspondence to wzwang@mail.sic.ac.cn, jliu@mail.sic.ac.cn
Abstract
N2 reduction to ammonia by solar light represents a green and sustainable ammonia
synthesis approach which helps to suppress the global warming and energy crisis.
However, conventional semiconductors usually suffer from low activity or poor
stability, largely suppressing the application of this technology. Here, we report
bismuth monoxide (BiO) quantum dots with an average size of 2-5 nm exhibited
efficient photocatalytic activity for ammonia synthesis under simulated solar light. A
highly efficient ammonia synthesis rate of 1226 μmol/g·h is achieved without the
assistant of any sacrificial agent or co-catalyst, which is about 1000 times higher than
that of traditional Fe-TiO2 photocatalyst. Kinetic analysis reveals that cooperation of
three low valence surface Bi (II) species markedly enhances N2 activation by electron
donation, which finally resulted in the highly efficient N2 photoreduction
performance. This work will shed light on designing efficient and robust N2 reduction
photocatalysts.
1

Journal of Materials Chemistry A
Page 2 of 27
View Article Online
Introduction
DOI: 10.1039/C6TA09275F
The reduction of dinitrogen (N2) to ammonia is a requisite transformation for the life
growth and survival.1-4 Although over 78% of the atmosphere is composed of
dinitrogen molecules, they are not used by most of organisms due to strong N≡N
bond, nonpolarity and high ionization potential.5,6 At present, most of the required
nitrogen for human beings is still originated from the energy-intensive Haber–Bosch
ammonia synthesis process,7,8 which consumes 1-2% of the world’s energy supply
and generating more than 300 million tons of carbon dioxide annually.9-12 In contrast,
dinitrogen photoreduction to ammonia by solar light represents a green and
sustainable ammonia synthesis route without fossil fuel consumption and CO2
emissions.13-16 However, the state-of-the-art efficiencies for ammonia synthesis by this
approach are still far away from the industrial demand.17-24
Generally, the rate-determining step for N2 reduction is cleavage of the N≡N
bond.5,6 N2 activation by electron donation from some reducing active sites are
indispensable for the promotion of N≡N bond cleavage,25 but such reducing active
sites are rarely existed on most of the semiconductor photocatalysts. Up to the
present, the majority of the reported N2 photoreduction catalysts are only have defect
type activation centers,17-24 which resulted in the unsatisfied photocatalytic
performance because of their limited defect concentrations. Besides, defect states
often suffer from thermal instability and increased charge carrier recombination,
further prevents their practical application for artificial ammonia synthesis.
Consequently, developing an ideal photocatalytic material which has high
concentration of non-defect type activation centers with robust stability holds the key
to achieving a breakthrough in N2 photoreduction technology.
Using low valence metal species in low-valent metal oxide semiconductors for N2
2

Page 3 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
activation may be a viable strategy to achieve this important goal. Low-valent metals
in semiconductor solids have fewer coordination atoms than those in their high-valent
state and are hence chemically unsaturated to facilitate molecules chemisorption.26
Owing to the strong electron donation power, low-valent metal species such as Fe,
Mo, Zr, Ti etc., have been widely studied for N2 activation in organometallic
complexes,9,16,27-30 although a stoichiometric excess of strong reducing agents (such as
Na, K) and extra proton sources are generally required to further generate NH3 in
these reaction systems. Different from organometallic complexes, low-valent metal
species in semiconductor material may act as both the N2 activation and
hydrogenation centers without an extra reducing agents, which allows quick reaction
kinetics for N2 reduction, because the conduction band of a metal oxide
semiconductor was usually hybridized by the metal p, d orbitals,31 and photogenerated
electrons were mainly located on the metal sites. Moreover, different from defect-type
active sites, lattice low-valent metal species can refrain from becoming an
electron-hole recombination center and all of the surface low-valent metal sites could
contribute to the catalysis reactions theoretically. In these regards, low-valent metal
oxide semiconductors are expected to exhibit high photocatalytic activity for N2
photoreduction. To confirm this opinion, herein low valent bismuth monoxide (BiO)
was selected as an ideal model material to study the N2 photoreduction performance.
We focus on low valent bismuth (II) because it has high electron donation power and
empty 6d orbitals for N2 adsorption and activation. More importantly, bismuth based
materials have been widely studied with various catalytic applications.32-34
Experimental
Preparation of BiO. Firstly, sodium oleate (2.2 mmol) and Bi(NO3)3·5H2O (0.4
3

Journal of Materials Chemistry A
Page 4 of 27
View Article Online
DOI: 10.1039/C6TA09275F
mmol) were successively added to distilled water (20 mL). After vigorous stirring for
2 h, 20 mL water was added and the suspension was then transferred to a 50 mL
Teflon-lined autoclave, sealed and heated at 140 ºC for 17 h. The system was then
allowed to cool down to room temperature. The obtained solid products were
collected by centrifugation, washed with absolute ethanol three times, and then
freeze-dried for further characterization. For comparison, 0.2wt % Fe-doped TiO2 was
prepared according to a previous report by coprecipiation and calcination process at
500 ºC for 3h18. The XRD pattern of the Fe-doped TiO2 was shown in Fig. S1.
Characterization. The purity and the crystallinity of the as-prepared samples were
characterized by powder X-ray diffraction (XRD) on a Japan Rigaku Rotaflex
diffractometer using Cu Kα radiation while the voltage and electric current were held
at 40 kV and 100 mA. The transmission electron microscope (TEM) analyses were
performed by a JEOL JEM-2100F field emission electron microscope. X-ray
photoelectron spectroscopy (XPS) was carried out by irradiating the sample with a
320 μm diameter spot of monochromated aluminum Kα X-rays at 1486.6 eV under
ultrahigh vacuum conditions (performed on ESCALAB 250, Thermo Scientific Ltd.).
UV–vis diffuse reflectance spectrum (DRS) of the sample was measured using a
Hitachi UV-3010PC UV–vis spectrophotometer. Fourier transform infrared (FTIR)
spectrum of concentrated ammonia product was performed with a spectrophotometer
(Nicolet 380, Thermo, USA). Nitrogen temperature-programmed desorption (N2-TPD)
measurements were performed on a ChemiSorb 2750 instrument. Typically, 20 mg of
the sample, placed in a glass tube, was pretreated by a He gas flow at 150 °C for 2 h,
and then cooled down to 50 °C. The adsorption of N2 was performed in a 99.999% N2
gas flow for 2 h at 50°C. After purge by He gas, the sample was heated from 50 °C to
450 °C at a rate of 10 °C/min. The TPD signal was recorded by a thermal conductivity
4

Page 5 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
1
detector. All of the gas flow rates were set as 30 mL/min. H NMR spectra were
acquired on a Bruker AV500 NMR spectrometer.
Photocatalytic Test. Photocatalytic N2 reduction were performed under a 500 W Xe
lamp located approximately 10 cm from the sample. The reaction cell (capacity 600
mL) was made of Pyrex glass with a quartz window on top. For the atmospheric N2
fixation, 0.05 g of the as-prepared photocatalyst powder was dispersed in 200 mL
15
deionized water and then stirring under the simulated solar light irradiation. For N
isotopic labelling experiment, the reaction cell was enclosed by a quartz window on
15
14
top. N2 (50% N2, 50% N2) gas was slowly bubbled through the reaction vessel,
which contained 200 mL deionized water until that was saturated. Then the reaction
vessel was sealed and irradiated under the simulated solar light irradiation. During the
photocatalytic tests, the temperature of the reaction vessel was maintained at 25 ºC by
providing a flow of cooling water. The concentration of ammonia in the reactor
solution was measured using the indophenol blue method. The amount of evolved O2
was determined by using online gas chromatography.
Computational Methods. Periodic density functional theory (DFT) calculations with
spin-polarization were performed using the VASP code. The non-local exchange and
correlation energies were described using a generalized gradient approximation with
the Perdew–Burke–Ernzerhof (PBE) functional (PBE-GGA). The electron-ion
interactions were described by projector augmented wave (PAW) potential. The plane
wave-plane basis set with energy cutoff of 400 eV and 4×4×4 k-point
Monkhorst-Pack grid were used in our calculations. The convergence criterions for
electronic SCF and force were set as less than 1.0×10-6 eV and 0.01 eV/Å,
respectively.
Electrochemical measurements. Electrochemical measurements were performed on
5

Journal of Materials Chemistry A
Page 6 of 27
a CHI 660D electrochemical workstation (Shanghai Chenhua, China) using a standarVdiew Article Online
DOI: 10.1039/C6TA09275F
three-electrode cell with a working electrode, a platinum wire as counter electrode,
and a standard saturated calomel electrode (SCE) in saturated KCl as reference
electrode. The working electrodes were prepared by dip-coating: Briefly, 5 mg of
photocatalyst was suspended in 0.1 mL of ethanol in the presence of 1% Nafion to
produce slurry, which was then dipcoated onto a 2 cm × 1.5 cm FTO glass electrode
and drying at 25 °C.
Calculation of electron transfer number: For an adsorption-controlled and
irreversible electrode process, according to Laviron,35 Ep is defined by the following
equation:
Ep  E0  (RT/anF)ln(RTK0 / anF)  (RT / anF)ln v
where R is molar gas constant, T is thermodynamic temperature, F is faraday constant,
a is transfer coefficient, k0 is standard rate constant of the reaction, n is electron
transfer number involved in the rate-determining step, ν is scan rate(mV/S), and E0 is
formal potential. For an irreversible electrode process, a≈0.5.
Results and Discussion
We designed an oleate-modified hydrothermal synthesis method to obtain the desired
BiO material. Oleate ions played an important role in the formation of this
low-valence bismuth oxide compound, as shown in Fig. 1. Initially each Bi3+ ion
interacts electrostatically with three oleate ions to form Bi-oleate complexes, which
then aggregated into larger Bi-oleate particles driven by the hydrophobic effect of the
oleate tails. The presented extra oleate ions in the solution accumulated around the
Bi-oleate complex by hydrophobic carbon tails to form stable Bi-oleate micelles with
hydrophilic outer shell which dispersed homogeneously in the reaction solution. The
6

Page 7 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
BiO compound was nucleated through a dehydration and condensation reaction after
permeation of a small amount of OH－ into the Bi-oleate micelles. This process was
hindered by the strong steric effect from oleate which has long aliphatic carbon chain
to prevent polar ions to entry. Since the reductivity of unsaturated carbon–carbon
double bonds in oleate, the Bi (III) was gradually reduced to Bi (II) during reaction
with OH－to form low-valence BiO compound in the micelles. Without the addition of
oleate ions, Bi2O3 is the final hydrothermally synthesized product. The as-prepared
BiO compound exhibited quantum sized spherical-like morphology (quantum dots)
with average size of 2-5 nm (Fig. 2a). A high magnified TEM image (Fig. 2b) gives
more details of the microstructure, which revealed the single crystal nature of the BiO
particles. From a high resolution TEM (HRTEM) image in Fig. 2c, (003) and (100)
planes of BiO (space group R3m) with a d spacing of 0.325 and 0.336 nm were
clearly observed, which belong to the [010] zone axis of BiO. The schematic crystal
structure of BiO along the [010] direction (Fig. 2d) displayed the same orientation and
arrangement of (003) and (100) planes as that observed from Fig. 2c. The purity and
crystallinity of the BiO particles were investigated by X-ray diffraction pattern. As
shown in Fig. 2e, all of the diffraction peaks can be indexed to the hexagonal phase of
BiO (JCPDS No.27-0054). No other likely impurities, such as Bi2O3 or Bi(OH)3, were
detected. The broad diffraction peaks were ascribed to its quantum size which is
calculated to be about 3 nm in average based on the Scherrer equation. XPS spectrum
further confirmed this low-valent bismuth (II) monoxide compound. As shown in Fig.
2f, the binding energies of Bi 4f5/2 and Bi 4f7/2 peaks in original BiO surface mainly
located at 164.55 eV and 159.26 eV which is originated from the residual
surface-adsorbed Bi (III)-oleate precursor. However, in the case of an etched surface,
the binding energies of the Bi 4f5/2 and Bi 4f7/2 peaks mainly decreased to 162.35 eV
7

Journal of Materials Chemistry A
Page 8 of 27
View Article Online
and 156.95 eV, proving the existence of low-valent bismuth species in a cleaDnOeIr: 10B.1i0O39/C6TA09275F
surface. Optical absorption property of the prepared BiO was investigated by UV–vis
diffuse reflectance spectroscopy which exhibited an intense UV-vis light absorption at
wavelengths shorter than 430 nm (Fig. S2), indicating a band gap of 2.88 eV for the
BiO sample.
To study the photocatalytic N2 reduction property of the BiO catalyst, 0.05 g of the
BiO particles was dispersed in 200 mL of deionized water and then subsequently
irradiated under simulated solar light irradiation. Control experiments showed that
+
NH4 cannot be generated in the absence of BiO catalyst or solar light irradiation. In
contrast, simulated solar light irradiation resulted in continuous ammonia generation
by the BiO sample (Fig. 3a). After 24 h of irradiation, the BiO catalyst generated 7.4
+
mg/L of NH4 from pure water in air. In an acidic reaction solution (pH 3.8) which
could provide excess protons to decrease the kinetic barrier for N2 reduction, the
generated ammonia concentration by the BiO catalyst was increased to 20.6 mg/L
within 24 h, with an much improved ammonia synthesis rate of about 1226 μmol/g
(catalyst)·h, 1000 times higher than that of traditional Fe-TiO2 photocatalyst (Fig. 3a)
under the same conditions. When the reaction was performed in 15% aqueous
methanol, a known sacrificial electron donor, the ammonia synthesis rate was further
+
significantly improved, producing a total NH4 amount of 0.7 mmol after 24 h (Fig.
3b), equivalent to a turnover number of 3.18. This suggests that ammonia evolution
under these conditions is catalytic. The increase in the ammonia synthesis rate in
aqueous methanol indicates the BiO quantum dots are able to photooxidize methanol.
Under the same conditions, no ammonia generation was detected with the BiO
catalyst both in pure water and aqueous methanol in argon saturated solution,
indicating the ammonia in the reaction system was originated from N2 reduction. The
8

Page 9 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
stability of the BiO catalyst under acid conditions is also investigated. After 5
consecutive runs (120 h) for N2 reduction, the photocatalytic performance of the BiO
particles was well-maintained (Fig. 3c). The TEM image of the BiO catalyst after the
cycle of reaction also indicates the stability of the BiO catalyst (Fig. S3).
+
To further confirm the detected NH4 in our experiment was indeed originated from
15
N2 reduction, we performed an isotopic labelling study using 50 vol % N2 as the
+
purge gas. Infrared spectroscopy was used to characterize the NH4 product. The IR
+
+
spectra (Fig. 3d) indicate that both the 15NH4 and 14NH4 exist in the reaction solution
15
14
+
when using the 50% N2 labelled N2, while only NH4 is detected when using pure
+
14N2. The infrared absorption peak of 14NH4 was located around 1405 cm-1, while the
15
+
peak position for NH4 is at 1358 cm-1 which is in good agreement with the value
estimated according to the isotope effect (that is, 1405 cm-1×(14/15)1/2=1357 cm-1)25.
+
The isotope labelling study further proved that the NH4 originates from the
1
photocatalytic N2 reduction process. H NMR spectrum was also used for detecting
+
the 15NH4 specie in the obtained ammonia product. As shown in Fig. S4, both of
+
+
15NH4 and 14NH4 were observed in the NMR spectrum. The molar ratio of the
+
+
15NH4 to 14NH4 was about 2.7 : 3, which is close to the initial 15N2 content.
It is of significant importance to understand the underlying mechanism on the high
photocatalytic activity of BiO in reducing N2 in order to develop other efficient
photocatalysts. Quantum chemical calculations based on density functional theory
(DFT) were conducted to investigate the possible N2 activation and hydrogenation
process on BiO catalyst. According to the experimental observation from HRTEM
image, a presentative BiO {010} surface was selected to investigate the N2 activation
and hydrogenation performance. First, partial charge density nearby Fermi level was
calculated to analyze the possible interaction positions of BiO surfaces with N2. As
9

Journal of Materials Chemistry A
Page 10 of 27
shown in Fig. 4a and 4b, the valence bands of BiO near Fermi levels areDOmI: a10i.n10l3Vyi9e/wCA6rTtiAcl0e9O2n7li5nFe
generated by p-electrons of Bi atoms. The charge distribution on BiO {010} surface is
significantly localized and accumulated around Bi atoms, which indicated that N2
activation by electron donation may take place upon the Bi activation centers. The
extent of N2 activation and hydrogenation on the BiO surface can be reflected by the
change of N  N bond length. Fig. 4c presents N-N bond distance evolutions with
adding hydrogen atoms on the BiO {010} surfaces. For comparison, the
corresponding evolutions without catalyst are also plotted. As shown in Fig. 4c, the
BiO {010} surface has high catalytic activity on activation of N≡N bond in N2Hn
(n=0-3). The N-N bond length could be elongated from 1.09 Å to 1.12Å when it is
near the BiO {010} surface. The N-N bond length could be further increased along
with hydrogenation until three H atoms were hydrogenated, indicating that N≡N bond
may be broken after three H atoms were added by the BiO catalyst. Considering the
N2 photoreduction experiment was performed in water in this study, the influence of
water molecules on the N2 activation and hydrogenation was also investigated. The
calculated result indicates the catalytic activity of BiO was not changed with adding
H2O on surface. In addition, H2O molecules covering on surface can quickly
dissociate into H+ and OH － , providing enough hydrogen for ammonia generation.
Besides, the binding energy calculations reveal a weaker interaction between N2Hn
and water-covered BiO surface than that in dry surface (Fig. 4d). This indicates that
water on the BiO {010} surface is advantageous to improve N2 photoreduction
performance since hydrogenated N2Hn is easy to be desorbed from catalyst.
Temperature programmed desorption (TPD) investigation was also conducted to
visualize the adsorption and activation of N2 on the surface of the BiO catalyst. As
shown in Fig. S5, two peaks at 295 ºC and 330 ºC were clearly observed for N2
10

Page 11 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
desorption, which indicated the excellent N2 chemisorption on BiO catalyst.
Photoelectrochemical measurements further revealed the possible reasons for the
high performance of BiO catalyst. As shown in Fig. 5a, two well-defined cathodic
peaks around –0.7 V and -1.2 V appears in the cyclic voltammetry in N2 saturated 0.5
M Na2SO4 solution for BiO electrode in an acid solution (pH=3.8). The pH value of
the electrolyte was adjusted by dilute H2SO4. The lower reduction peak around -0.7 V
only arises after the electro-reduction process proceeding a certain time (Fig. 5a and
Fig. S6), indicating this reduction peak may be originated from further reduction of
intermediates during the photoelectrochemical process. Therefore the N2 reduction
peak around -1.2V (Ep) was adopted for the kinetic study under acid conditions. As
shown in Fig. 5a, the peak current densities increased with increasing the scan rates
(ν). The number of electron transfer (n) involved in the N2 reduction can be calculated
from the slope of the Ep-lnν (ESI). The value of n was calculated to be 2.93 at the
potential 20-100 mV, indicating a one-step three electron transfer N2 reduction
process may be existed on the BiO surface. As is well known, nitrogen fixation is a
proton-coupled reaction.36 Although the BiO surface could provide enough low valent
Bi (II) activation centers, excess H+ is important to realize a proton coupled
multi-electron reduction process. In a neutral reaction solution, the electron transfer
number was decreased to 1.34 based on the analysis of the reduction peak around -0.9
V (Fig. 5b), indicating N2 reduction on BiO catalyst in neutral solution may be
realized by six consecutive one-electron redox step, resulting in a much lower
ammonia synthesis rate. This is in line with the lower photocatalytic N2 reduction
activity of BiO at pH=7. Besides, a decreased cathodic current density (Fig. 5c) in
N2-saturated reaction system was observed when compared with that in air saturated
solution, which proved the larger amount of electrons transfer from the BiO catalyst
11

Journal of Materials Chemistry A
Page 12 of 27
to the N2 molecule in N2-saturated reaction system under the same conditions. Fig. 5Vdiew Article Online
DOI: 10.1039/C6TA09275F
is Mott−Schottky spectrum of the BiO sample, which is usually used for the analysis
of the flat band potential (Efb) of semiconductor electrodes.37 The positive slope of the
plot indicates the BiO catalyst is n-type semiconductor with electrons as the majority
charge carriers. The Efb value which was calculated from the intercept of the axis with
potential value was at −0.69 V vs SCE. For many n-type semiconductors, Efb is
considered to be about 0.1 V below the conduction band (Ecb).38 Based on this, the
estimated Ecb values of the BiO sample was -0.55 V vs. NHE. According to the
thermodynamic data of N2H3 reported by Bauer,39 the N2 reduction potential via N2 +
－
3H+ + 3e → N2H3 is not larger than −0.485 V vs. NHE. This indicates the
photogenerated electrons in the BiO catalyst could energetically reduce the N2
molecular via a proton-coupled three-electron reduction process. Besides, the valence
band potential of the BiO catalyst which estimated from the band gap and Ecb was also
energetically large enough (2.39 V vs. NHE) for water oxidation (1.23 V vs. NHE)
under acid condition, indicating water could efficiently act as the sacrificial electron
donor for ammonia generation. All of these studies proved the possibility of highly
efficient one-step three-electron N2 reduction on BiO catalyst, which may be realized
by cooperation of three adjacent low valent Bi (II) activation centers. This process is
difficult to occur in traditional N2 reduction photocatalyst which only has isolated
defect-type activation centers.
The crystal structure and the coordination environment of BiO compound may play
an important role for its highly efficient photocatalytic N2 reduction performance. As
shown in Fig. 2d, the BiO compound exhibits a hexagonal tunnel structure along the
[010] direction, which built up from three alternately arranged Bi groups and O
groups. The entrance diameter of this tunnel structure is about 4.4 ～ 4.7 Å, large
12

Page 13 of 27
Journal of Materials Chemistry A
enough for structurally filling one N2 molecular (diameter 3 Å). A careful obsDeOrvI: a10t.i1o03Vni9e/wCA6rTtiAcl0e9O2n7li5nFe
for local structure of BiO indicates BiO4 presents pyramidal unit with two OBiO
angles of 114.2o and 104.2o, far away from tetrahedral structure (Fig. S7). This
indicates that bismuth atoms may take an orbital hybridization of sp3d to form orbital
distribution of trigonal biyramid, leaving a lone pair of electrons in the Bi atom. The
N2 molecule may be stretched and activated by three alternatively arranged Bi atoms
through donating electrons to the empty 6d orbitals of Bi atom and accepting
electrons from the lone pairs of Bi atom to its three unoccupied anti-bonding orbitals
(π*2py, π*2pz and σ*2px), resulting in a 1N2-3Bi side-on bond structure (Fig. 6a). The
electron donations from ajacent three low valence bismuth ions will strongly weaken
the N≡N bond, which facilitates a one-step three electron N2 reduction process in a
successive photocatalytic process by proton and electron transfer (Fig. 6b).
Based on the above analysis, the possible N2 reduction reactions on BiO surface under
acid conditions were proposed as following.
BiOhv BiO(h  e- )
（1）
（2）
N2  3BiO(e ) 3H N2H3

N2H3  3BiO(e ) 5H 2NH4
N2H3 BiO(e )H N2H4
（3）
（4）
2N2H3 2N2  3H2
（5）
3H2O 6BiO(h ) 3/2O2  6H
（6）
First, simulated solar light excited electrons and holes were generated on the BiO
catalyst (equation 1). Then the photogenerated electrons may transfer to the surface
and participate in a one-step three electron N2 reduction process to obtain N2H3
13

Journal of Materials Chemistry A
Page 14 of 27
View Article Online
DOI: 10.1039/C6TA09275F
intermediate (equation 2) due to the possible 1N2-3Bi side-on bond structure. N2H3
+
intermediate will be further reduced into NH4 (equation 3) or N2H4 (equation 4) by a
subsequent one step three-electron or one-electron photocatalytic reduction process,
respectively. Otherwise, the N2H3 intermediate will be decomposed, because it is a
short-lived, endothermic compound.39,40 If the N2H3 intermediate is not further
reduced, it will decompose into N2 and H2 spontaneously (equation 5). In our study,
the reaction rates of equations 2, 3 are much higher than that of equation 5, giving the
+
continuous production of NH4 under the acid conditions. The amount of N2H4 in the
final product was only about 1.6 %, possibly because of the 1N2-3Bi side-on
coordinating mode of N2 which facilitate three electron reduction process (equations 2,
3). During the photocatalytic N2 reduction process, O2 was produced from water
oxidation by photogenerated holes (equation 6). The theoretical molar ratio of the
+
generated NH4 to O2 from the above equations is 1.33. In a closed reaction system
which filled with pure N2 and water, we observed a continuous O2 evolution (Fig. S8)
along with a prolonged irradiation time and a total amount of 117 μmol O2 was
+
+
obtained when generating 2.32 mg NH4 . The molar ratio of the generated NH4 to O2
is 1.41, which is close to the theoretical value. Besides, a tiny amount of H2 (2.4 μmol)
was also detected within 24 h in the closed reaction system, which may be originated
from the decomposition of trace amount of unconverted N2H3 intermediate.
Conclusions
In summary, bismuth monoxide quantum dots were successfully synthesized by a
facile hydrothermal synthesis method using oleate as capping agent. Different from
14

Page 15 of 27
Journal of Materials Chemistry A
the other previously reported N2 reduction photocatalysts which have limited anVdiew Article Online
DOI: 10.1039/C6TA09275F
randomly distributed defect-type N2 activation centers, the BiO catalyst has all of the
lattice bismuth species as potential activation centers. This characteristic makes the
as-synthesized BiO a highly efficient photocatalyst for solar N2 reduction to ammonia.
The ammonia synthesis rate is up to 1226 μmol/g·h without the assistant of any
sacrificial agent or co-catalyst, which is about 1000 times higher than that of
previously reported traditional Fe-TiO2 photocatalyst. More importantly, the
photocatalytic activity does not show any obvious deactivation even after 120 h.
Kinetic studies by electrochemical measurements and quantum chemical calculations
indicate the highly improved photocatalytic performance of BiO may be ascribed to
the cooperation of three low valent surface Bi (II) species for N2 activation. This study
may open up new opportunities for designing highly efficient and robust solar light
photocatalyst for artificial ammonia synthesis under ambient conditions.
Acknowledgment
This work was financially supported by the National Basic Research Program of
China (2013CB933200), National Natural Science Foundation of China (21671197,
51272269, 51272303, 51472260), and the research grant (16ZR1440800) from
Shanghai Science and Technology Commission. Prof. Guangyu Li from Shanghai
Institute of Organic Chemistry is gratefully acknowledged for measurement and
interpretation of the NMR data.
Supplementary Materials
Figure S1-S8.
References
15

Journal of Materials Chemistry A
Page 16 of 27
View Article Online
1. T. Rosswall, in Some Perspectives of the Major Biogeochemical Cycles (ed.DLOIi:k10e.1n0s3,9/C6TA09275F
G. E.) Ch. 2 (Wiley, New York, 1981).
2. J. B. Howard, D. C. Rees, Chem. Rev. 1996, 96, 2965−2982.
3. B. M. Hoffman, D. R. Dean, L. C. Seefeldt, Acc. Chem. Res. 2009, 42, 609–619.
4. R. Navarro-González, C. P. McKay, D. N. Mvondo, Nature 2001, 412, 61−64.
5. S. Gambarotta, J. Scott, Angew. Chem. Int. Ed. 2004, 43, 5298–5308.
6. A. J. Pool, E. Lobkovsky, P. J. Chirik, Nature 2004, 427, 527–530.
7. V. Smil, Enriching the Earth: Fritz Haber, Carl Bosch and the Transformation of
World Food Production (MIT Press, Cambridge, MA, 2001).
8 C. Sivasankar, S. Baskaran, M. Tamizmani, K. Ramakrishna, J. Organomet. Chem.
2014, 752, 44–58.
9. R. Schlögl, Angew. Chem. Int. Ed. 2003, 42, 2004–2008.
10. H. Bielawa, O. Hinrichsen, A. Birkner, M. Muhler, Angew. Chem. Int. Ed. 1999,
40, 1061–1063.
11. T. Shima, S. Hu, G. Luo, X. Kang, Y. Luo, Z. Hou, Science 2013, 340,
1549−1552.
12. Y. Tanabe, Y. Nishibayashi, Coordin. Chem. Rev. 2013, 257, 2551–2564.
13. G. N. Schrauzer, T. D. Guth, J. Am. Chem. Soc. 1977, 99, 7189–7193.
14. E. Endoh, J. K. Leland, A. J. Bard, J. Phys. Chem. 1986, 90, 6223−6226.
15. H. Li, J. Shang, Z. Ai, L. Zhang, J. Am. Chem. Soc. 2015, 137, 6393–6399.
16. H. Kisch, Angew. Chem. Int. Ed. 2013, 52, 812–847.
17. V. Augugliaro, A. Lauricella, L. Rizzuti, M. Schiavello, A. Sclafani, Int. J.
Hydrogen Energ. 1982, 7, 845−850.
18. J. Soria, J. C. Conesa, V. Augugliaro, L. Palmisano, M. Schiavello, A. Sclafani, J.
Phys. Chem. 1991, 95, 274−282.
16

Page 17 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
19. O. P. Linnik, H. Kisch, Mendeleev Commun. 2008, 18, 10–11.
20. L. Palmisano, V. Augugliaro, A. Sclafani, M. Schiavello, J. Phys. Chem. 1988, 92,
6710−6713.
21. K. Tennakone, S. Punchihewa, R. Tantrigoda, Sol. Energy Mater. 1989, 18,
217−221.
22. O. Rusina, A. Eremenko, G. Frank, H. P. Strunk, H. Kisch, Angew. Chem. Int. Ed.
2001, 40, 3993−3995.
23. W. Zhao, J. Zhang, X. Zhu, M. Zhang, J. Tang, M. Tan, Y. Wang, Appl. Catal. B:
Environ. 2014, 144, 468–477.
24. G. Dong, W. Hob, C. Wang, J. Mater. Chem. A 2015, 3, 23435–23441.
25. M. Kitano, Y. Inoue, Y. Yamazaki, F. Hayashi, S. Kanbara, S. Matsuishi, T.
Yokoyama, S. W. Kim, M. Hara, H. Hosono, Nature Chem. 2012, 4, 934−940.
26. G. Ertl, Angew. Chem. Int. Ed. 2008, 47, 3524–3535.
27. M. M. Rodriguez, E. Bill, W. W. Brennessel, P. L. Holland, Science 2011, 334,
780−783.
28. K. Arashiba, Y. Miyake, Y. Nishibayashi, Nature Chem. 2011, 3, 120−125.
29. B. A. MacKay, M. D. Fryzuk, Chem. Rev. 2004, 104, 385−401.
30. A. Caselli, E. Solari, R. Scopelliti, C. Floriani, N. Re, C. Rizzoli, A. Chiesi-Villa,
J. Am. Chem. Soc. 2000, 122, 3652−3670.
31. K. Maeda, T. Takata, M. Hara, N. Saito, Y. Inoue, H. Kobayashi, K. Domen, J.
Am. Chem. Soc. 2005, 127, 8286−8287.
32. S. Sun, W. Wang, D. Li, L. Zhang, ACS Catal. 2014, 4, 3498−3503.
33. J. Medina-Ramos, J. L. DiMeglio, J. Rosenthal, J. Am. Chem. Soc. 2014, 136,
8361−8367.
34. L. Liang, F. C. Lei, S. Gao, Y. F. Sun, X. C. Jiao, J. Wu, S. Qamar, Y. Xie, Angew.
17

Journal of Materials Chemistry A
Page 18 of 27
View Article Online
DOI: 10.1039/C6TA09275F
Chem. Int. Ed. 2015, 54, 13971−13974.
35. E. Laviron, J. Electroanal. Chem. Interfacial Electrochem. 1974, 52, 355-393.
36. L. Pospíšil, J. Bulíčková, M. Hromadová, M. Gál, S. Civiš, J. Cihelka, J. Tarábek,
Chem. Commun. 2007, 2270–2272.
37. A. Watanabe, H. Kozuka, J. Phys. Chem. B 2003, 107, 12713−12720.
38. Y. Matsumoto, J. Solid State Chem. 1996, 126, 227−234.
39. N. Bauer, J. Phys. Chem. 1960, 64, 833–837.
40. B. Ruscic, J. Berkowitz, J. Chem. Phys. 1991, 95, 4378−4384.
18

Page 19 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
Figure captions
Fig. 1 Schematic illustration of the whole synthesis procedure for the BiO quantum
dots.
Fig. 2 Characterizations for the hydrothermally synthesized BiO sample. a. Low
magnified TEM image of the BiO sample. b, High magnified TEM image. c, High
resolution TEM image. d, The related schematic atomic model of BiO, clearly
showing atomic configuration of Bi/O and hexagonal tunnel structure along the [010]
direction. e, XRD pattern of the as-prepared BiO sample. f, XPS spectrum of the
as-prepared BiO sample in the Bi 4f5/2 and Bi 4f7/2 binding energy region.
Fig. 3 Photocatalytic N2 reduction performance of the as-prepared BiO sample at
different conditions under simulated solar light irradiation. a, The generated ammonia
concentrations at different irradiation time in pure water, inset: the corresponding
ammonia synthesis rate. b, The generated ammonia concentrations at different
irradiation time in 15 vol% methanol solution. c, Photostability of the as-prepared
BiO sample for N2 reduction under simulated solar light at acid conditions (pH3.8). d,
+
Comparison of experimental infrared spectra of NH4 obtained after N2
photoreduction on BiO catalyst using 50% 15N2 labelled N2 and pure 14N2
respectively.
Fig. 4 Quantum chemical calculations of N2 activation and hydrogenation
performance of BiO catalyst based on density functional theory. (a, b) The calculated
charge density plot at valence band maximum (VBM) of BiO {010} surfaces, a: top
view, b: side view. The yellow and blue isosurfaces represent charge accumulation
and depletion in the space, respectively. c, The N-N bond distance evolutions with
adding hydrogen atoms on BiO {010} surface. d, The binding energy of N2Hn on BiO
{010} surface.
19

Journal of Materials Chemistry A
Page 20 of 27
View Article Online
DOI: 10.1039/C6TA09275F
Fig. 5 Kinetic study of N2 reduction on BiO catalyst by electrochemical
measurements in N2 saturated 0.5 M Na2SO4 under room temperature (25 ℃ ) and
simulated solar light irradiation. a, Cyclic voltammograms of BiO electrode at
pH=3.8. b, Cyclic voltammograms of BiO electrode at pH=7. Inset: the plot for the
reduction peak potential (Ep) vs. lnν (scan rate). c, Cathodic electrochemical scans on
the as-prepared BiO electrode under chopped simulated solar light irradiation. d,
Mott−Schottky plot of the as-prepared BiO electrode in 0.5 M Na2SO4 at pH=7.
Fig. 6 Possible pathway for N2 activation and hydrogenation on BiO catalyst. a,
Schematic electron donation from the lone pairs of three Bi atoms to the three
unoccupied anti-bonding orbitals (σ*2px, π*2py, and π*2pz) of N2 molecular, forming
a 1N2-3Bi(II) side-on bound structure. b, The whole N2 activation and reduction
process on BiO catalyst.
20

Page 21 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
Fig. 1
21

Journal of Materials Chemistry A
Page 22 of 27
View Article Online
DOI: 10.1039/C6TA09275F
b
a
e
c
d
f
d003
d100
Fig. 2
22

Page 23 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
a
b
c
d
Fig. 3
23

Journal of Materials Chemistry A
Page 24 of 27
View Article Online
DOI: 10.1039/C6TA09275F
b
a
c
d
Fig. 4
24

Page 25 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
b
a
a
c
d
Fig. 5
25

Journal of Materials Chemistry A
Page 26 of 27
View Article Online
a
DOI: 10.1039/C6TA09275F
b
Fig. 6
26

Page 27 of 27
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C6TA09275F
Table of contents
An efficient and sustainable production of ammonia from water and air under
simulated solar light is achieved by BiO photocatalyst.