Electron-Driven Heterogeneous Catalytic Synthesis of Ammonia: Current States and Perspective

Carbon Resources Conversion 1 (2018) 2–31
Contents lists available at ScienceDirect
Carbon Resources Conversion

CHINESE ROOTS
GLOBAL IMPACT
journal homepage: www.keaipublishing.com/en/journals/carbon-resources-conversion/
Electron-driven heterogeneous catalytic synthesis of ammonia: Current

states and perspective
Ke Wang, Daniel Smith, Ying Zheng ⇑
Institute for Materials and Processes, School of Engineering, University of Edinburgh, Edinburgh EH9 3FB, UK
a r t i c l e i n f o a b s t r a c t
Article history: Ammonia is the second most produced chemical worldwide that makes up 80% of nitrogen-based fertilis-
Received 3 April 2018 ers, which have supported approximately 27% of the world’s population over the last century. The Haber–
Revised 13 June 2018 Bosch process, which is the main process for producing ammonia, is extremely energy intensive and
Accepted 18 June 2018
consumes around 1% of the world’s energy. Additionally, it requires hydrogen gas as a reactant that is
Available online 20 June 2018
produced via steam reforming which emits carbon dioxide as a by-product. Over 500 million tonnes of
ammonia are produced per year via industrial processes which required 3–5% of total natural gas con-
sumption worldwide and also accounted for 2% global energy usage. Therefore, more sustainable pro-
cesses, such as electrocatalysis and photocatalysis, using electrons and the transfer of protons has
been investigated. This review covers the most state-of-the-art technologies used to produce ammonia
via electrocatalysis and photocatalysis by comparing different electrolyte systems and electrocatalysts
as well as discussing issues with these methods and possible solutions. In addition, substantial improve-
ments to electrocatalysts and photocatalysts as well as methods to prevent both the promotion of the
hydrogen evolution reaction and the decomposition of ammonia at higher temperatures are reviewed.
Challenges and perspectives are discussed.
Ó 2018 The Authors. Production and hosting by Elsevier B.V. on behalf of KeAi Communications Co., Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-
nd/4.0/).
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2. Basic principle and mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1. Ammonia synthesis on heterogeneous catalysts via thermocatalytic reaction: the Haber–Bosch process . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2. Electron-driven mechanism of ammonia synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.1. Photocatalytic synthesis of ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.2. Electrochemical synthesis of ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3. The quantitative measurement of ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3.1. Colorimetric method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3.2. Methods with selectivity or separation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3. Photocatalytic N2 fixation to ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1. Semiconductor engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1.1. Facet design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1.2. Vacancies design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.1.3. Doping design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.1.4. Structure design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2. Co-catalyst design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.2.1. Transition metal as a co-catalyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.2.2. Other as the co-catalyst . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.3. Plasmon-induced nitrogen fixation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4. Electrocatalytic N2 fixation to ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
⇑ Corresponding author.
E-mail address: y.zheng@ed.ac.uk (Y. Zheng).
https://doi.org/10.1016/j.crcon.2018.06.004
2588-9133/Ó 2018 The Authors. Production and hosting by Elsevier B.V. on behalf of KeAi Communications Co., Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
K. Wang et al. / Carbon Resources Conversion 1 (2018) 2–31 3
4.1. Aqueous electrolyte system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

4.1.1. Transition metal based electrocatalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.1.2. Transition metal free electrocatalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.2. Polymer electrolyte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3. Molten salt electrolyte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.4. Solid electrolyte. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.4.1. O2 conducting materials. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.4.2. Proton conducting materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.4.3. Composite electrolyte. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5. Summary and perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.1. Ammonia production rate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.1.1. Catalyst design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.1.2. Hole scavenger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.1.3. Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.1.4. Feedstock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.2. Selectivity and efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.2.1. Cell design and configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.2.2. Catalysts design and optimization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.2.3. Potential bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.2.4. Control water or proton supply for the cathode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1. Introduction The Haber–Bosch process has dominated the commercial pro-

duction of ammonia [5]. The reaction, N2 þ 3H2 ! 2NH3 ;
Nitrogen is an essential component of life and a crucial compo- DH298 k ¼ 92:2 kJ=mol, is an exothermic reaction with the
nent of proteins. It makes up 78% of the atmosphere. However, decrease of the total gas molecules from left to right side. Theoret-
atmospheric nitrogen is found as N„N with a triple bond. This ically speaking, the reaction favours the low temperature. High
has a strength of 945 kJ/mol at 298 K and is one of the strongest temperature is used to cleavage the triple bond of dinitrogen mole-
covalent bonds. As a result, significant energy must be supplied cule and to avoid the poisoning of a catalyst by the adsorbed N
to break the bond. atoms [6]. To balance the detrimental effects of high temperature
In nature, the main source of nitrogen that can be used by life on the reaction equilibrium, high pressure becomes vital to main-
known as reactive nitrogen that was synthesised from air by bac- tain a commercially profitable conversion. Although not included
teria to produce ammonia. Nitrogen fixation via the enzymes, the in the scope of this review, homogeneous catalytic ammonia syn-
so-called nitrogenases, is an incredibly complicated process. There thesis using designed molecules to activate the triple bonds in N2
are multiple different ones spread across bacterium. For one NH3 has also drawn considerable amounts of interests [7–9]. The
molecule synthesized, the FeMo nitrogenase requires 8 ATP mole- Haber–Bosch process typically runs in a range of 400–600 °C and
cules, which contain energy as high as 244 kJ mol1 in total. The 20–40 MPa with the presence of commercial Fe-based or Ru-
general reaction behind nitrogen fixation is [1]: based catalysts. Under such a harsh condition, the efficiency of a
single pass is still limited to 10–15%. In addition to serve as a feed-
N2 þ 8e þ 8Hþ ! 2NH3 þ 16ADP þ 16Pi þ H2 ð1Þ stock of fertilizer, ammonia is also promising in other industrial
sectors. Ammonia is proven to be an excellent fuel of fuel cells
The first significant impact that humans implemented on the [10,11] and applied in hydrogen storage [12].
nitrogen system was in the creation of agriculture. The implemen- Research of an alternative and sustainable way for ammonia
tation of crop rotation in the late 18th century dramatically synthesis has begun several decades ago [13]. In 1970s, Schrauzer
increased crop yields and was part of the cause of the industrial and Guth first applied UV light to excite titania and successfully
revolution. The subsequent development of the Haber–Bosch Pro- reduced the dinitrogen molecules with water to generate ammonia
cess in 1913 allowed humans to synthetically create reactive nitro- [13,14]. TiO2 doped with Fe, an active metal element in Haber–
gen which caused huge increases to crop yields. The amount of Bosch process, exhibits the capability of photocatalytic ammonia
ammonia produced has dramatically increased and demonstrated synthesis which was further confirmed by Augugliaro et al.
one of the most significant effects mankind has had on the planet. [15,16]. Later, FeTiO3 [17,18] ZnO, CdS, SrTiO3, GaP pure semicon-
To date, approximate 120 millions of tons of N2 is fixed each year ductors and those mixed with platinum black were further exam-
for production of ammonia alone [2] while over 500 million tonnes ined [19] and improved ammonia synthesis were achieved. Direct
of ammonia are produced per year via industrial processes which electrochemical synthesis of NH3 from the dinitrogen was firstly
required 3–5% of total natural gas consumption worldwide and reported by Sclafani et al. [20] based on Fe cathodic electrocata-
also accounted for 2% global energy usage [3]. As the main lysts with water chosen to act as the proton source. The pioneering
resources for the fertilizer, the ammonia production is essential attempts on reduction of dinitrogen to ammonia using sustainable
for supporting the modern agriculture and feeding people on this energy sources did not draw much attention until the second dec-
planet. In the production of ammonia, hydrogen is a key reactant. ade of 21st century when carbon emissions and sustainability
The world’s hydrogen production primarily comes from methane became one of the top priorities in science and engineering.
steam reforming which process emits huge amount of CO2 that There are several excellent reviews focusing on the N2 fixation
accounts for 1.6% of global emission per year [4]. Moreover, the and ammonia production [21,22], including plasma catalysts[23],
demanding on the nitrogen fertilizer is growing rapidly and esti- electrocatalysts [24–28], photocatalysts [29–31]. However, a com-
mated to show a two or three-folded increase in the second half prehensive understanding of the role of electron-driven in the
of the 21st century [3]. ammonia production and catalysts design are still lacking.
4 K. Wang et al. / Carbon Resources Conversion 1 (2018) 2–31
Recently, fixation of atmospheric N2 for ammonia synthesis via the towards ammonia synthesis with the electronic promoter, Cs
renewable energy sources have been well advanced. It is the [45,46]. A Mars–van Krevelen mechanism has been proposed on
appropriate timing to review the recent development. In this crit- the nitride catalysts for the thermocatalytic ammonia synthesis
ical review, an emphasis is given to the electron-driven conversion with the involvement of the lattice N in the reaction [47–49]:
from N2. It includes the theoretical and experimental mechanism
Co3 Mo3 Nx þ H ! Co3 Mo3 Nx H þ ð7Þ
investigation, reactor design and most importantly the catalysts
development and optimization. The comparison between different
Co3 Mo3 Nx H þ H ! Co3 Mo3 Nx H2 þ ð8Þ
electron-driven ammonia production methods and related cata-
lysts are intended to be conducted.
Co3 Mo3 Nx H2 þ H ! Co3 Mo3 Nx H3 þ ð9Þ
2. Basic principle and mechanism Co3 Mo3 Nx H3 ! Co3 Mo3 Nx1 þ NH3 ð10Þ
2.1. Ammonia synthesis on heterogeneous catalysts via 1

thermocatalytic reaction: the Haber–Bosch process Co3 Mo3 Nx1 þ N2 ! Co3 Mo3 Nx ð11Þ
2
Prior to investigation of the electron-driven mechanism of The lattice N vacancies (VN) become crucial in the N2 activation
ammonia synthesis, it is necessary to brief the thermocatalytic and the followed hydrogenation step on nitride catalysts. And VN
ammonia production. The ‘‘ammonia Fe” has been so successful also plays an important role in lowering the temperature required
that it has been a dominant catalyst for over 100 years in the for the thermocatalytic ammonia production [50,51]. A recent
commercial Haber–Bosch process. Although not completely under- research on ammonia production via the hydrogenation of alkali
stood by the researchers, the thermocatalytic ammonia production metal nitride indirectly supports the argument that ammonia
has reached general agreement on the reaction mechanism. could be synthesized via the reaction between the lattice N and
As shown in Fig. 1, the widely accepted dissociative pathway for hydrogen molecules with the formation of VN [52].
the Haber–Bosch process ammonia synthesis can be described
with the following equations and the asterisk mark represents 2.2. Electron-driven mechanism of ammonia synthesis
the active site on the catalysts surface [32]:
Synthesis of ammonia via the reduction of the dinitrogen mole-
H2 þ 2 ! 2 H ð2Þ cules is usually a strong endothermic reaction due to the high
energy requirement for cleavage the triple bond (9.44 eV).
N2 þ 2 ! 2 N ð3Þ When using H2 as the proton source, the overall reaction could
be described as following:
N þ H ! NH þ ð4Þ
N2 ðgÞ þ 3H2 ðgÞ ! 2NH3 ðgÞ ð12Þ

NH þ H ! NH2 þ ð5Þ
Considering the H2 production via steam methane reforming is
also an energy intensive process, H2O then became a promising
NH2 þ H ! NH3 þ ! 2 þ NH3 ð6Þ
proton source for the ammonia synthesis. One possible overall
It is well accepted that the dissociative adsorption of N2 on the reaction is listed below:
catalyst surface is the rate determining step (RDS) [33,34]. In order
3
to achieve the dissociation of H2 and especially N2, the catalysts, N2 ðgÞ þ 3H2 O ðlÞ ! 2NH3 ðgÞ þ O2 ðgÞ ð13Þ
2
metal clusters, shall donate the electrons to the anti-bonding orbi-
tal in the molecular orbitals (MO) of the adsorbate. Alkali metals In this reaction, 6 electrons along with 6 protons are required to
[35] and electrides, a new type of promotors [36] have been proven transfer, which makes the reaction kinetically difficult to happen.
to be effective in promoting thermal catalytic ammonia synthesis Fortunately, the energy barrier for each proton-coupled electron
via transferring the electrons to the adsorbates. And it is also transfer step is relatively low. A low driving potential is extremely
essential that the electron transfer should be moderate. The important for the photocatalytic process since there is no voltage
d-band theory, BEP (Brønsted–Evans–Polanyi) relationship and bias on catalysts like the electrocatalysts.
volcano plot have been also verified in the N2 hydrogenation
[37–41]. Fe, Os, Ru and their alloys fall within the window of the 2.2.1. Photocatalytic synthesis of ammonia
required moderate properties [4,5,42]. Although the thermody- Photocatalytic process could be divided into two relatively
namic calculation supports the formation of the meta-stable FeNx, independent redox reaction pairs: adsorbate reduction by photon
no bulk nitridation of Fe could be confirmed by Neutron diffraction induced electrons and adsorbate oxidation by the irradiation
experiments during the ammonia synthesis [43]. Irreducible oxides excited holes. The relative positions of the electrons or holes com-
[44] support could kinetically stabilize the defective a-Fe from the paring with the redox reaction potentials thermodynamically
nitridation. However, nitride materials are indeed active for the determine whether the reactions could proceed or not. For a con-
Haber–Bosch process. The XAMoAN (X = Fe, Co, Ni) catalysts have ventional semiconductor, the incident light with the energy hm lar-
been reported since 2000 and became the most active catalysts ger than the bandgap of the semiconductor could be absorbed to
excite an electron in the valence band (VB) to the conduction band
(CB) and left a hole in the VB: the generation of the electron-hole
pairs. Fig. 2 gives the VB and CB positions of commonly used semi-
conductors relative to the reaction potentials of nitrogen reduction
reaction (NRR) to ammonia and corresponding hydrogen evolution
reaction (HER) and oxygen evolution reaction (OER). From the
thermodynamic view, the higher position of the CB and lower posi-
Fig. 1. The mechanism for the ammonia synthesis via Haber–Bosch process on
tion of the VB will make the corresponding reduction and oxidation
heterostructure catalysts. Reproduced from Ref. [25], Copyright (2017), with reaction easier to happen. In electron-driven ammonia synthesis
permission from Elsevier. via the dinitrogen fixation with the presence of water, selectivity
Fig. 2. Band-edge positions for conventional semiconductors referring to the electrode potentials (vs. NHE when pH = 0) of the redox reduction of the dinitrogen molecules
with water and the corresponding reactions may involve in the process [30,54].
is one of the most essential challenges. Fig. 2 shows that N2 reduc- collective vibration, which is called the localized surface plasmonic
tion to ammonia require a much higher reduction potentials than resonance (LSPR). The decay of the LSPR could generate the hot
hydrogen evolution alone. It suggests that the ammonia synthesis electrons and holes in the metal [56]. When the electron energy
is thermodynamically unfavourable. Furthermore, reduction of is higher than the schottky barrier height (SBH) between the metal
nitrogen to ammonia requires 6 protons along with electrons and semiconductor, the electrons are moved to the CB of the
which makes it even more kinetically difficult comparing with semiconductor [57]. The LSPR may enhance the light adsorption
hydrogen evolution [53]. Both the theoretical and experimental along with the accompanied electron transfer makes the entire
results evidence that most of the electrons and protons consumed process even more complicated. Comprehensive investigations
are meant for the hydrogen production. are required to better understand the mechanism.
Although semiconductors alone can serve as photocatalysts, co- For the conventional photocatalysis, the adsorption of the reac-
catalysts usually play an essential role in enhancing the photocat- tants onto the active sites and further donates the photon induced
alytic performance. The metal clusters showing significant ability electrons into the adsorbates are the very first steps taking place on
in dissociative and associative adsorption of reactants are key to the surface of a catalyst. In the case of N2 photofixation, N2 has
achieve the desired activity. A simple schottky barrier model was been proven to be very inert and difficult to be adsorbed. Zhu
established to theoretically describe the interface between a semi- et al. [58] and Bandy et al. [59] proposed an alternative photocat-
conductor and active metal sites: if the work function of the metal alytic process to address the challenge. They reported the H-
is larger than the fermi energy level of the semiconductor, the elec- terminated diamond could directly induce the emission of electron
trons will transfer from the semiconductor to the metal in order to to the solution producing the ‘‘solvated electrons”. This new pro-
achieve the same chemical potential [55]. cess does not need the adsorption of N2 onto the catalyst surface
For several coinage metals (i.e. Au, Ag, Cu), the large density of and the ‘‘solvated electrons” directly react with the reactants.
free electrons can be excited by visible light and then generate the Fig. 3(a) shows the potentials of this unique process to drive
Fig. 3. (a) The comparisons between the potentials for several typical photocatalytic reactions with the CB and VB positions of the H-terminated B-doped diamond. (b) The
schematic illustration of the measurement of the ‘‘solvated electron” concentration. (c) The changes in the intensity of the Laser impulse which is caused by the adsorption of
the ‘‘solvated electrons”. Reproduced from Ref. [58], Copyright (2013) with permission of Springer Nature.
several conventional photocatalytic reactions including the reduc- Their calculation results indicated that, in the associative path-
tion of atmospheric N2. The existence of the ‘‘solvated electrons” is way, the addition of the first proton to the adsorbed dinitrogen
confirmed by the laser adsorption experiments, as shown in Fig. 3 atoms is the limiting step on a flat surface of transition metal. On
(b and c). A further mechanism study by Christianson et al. [60] a stepped surface, the reaction may proceed via a dissociative path-
implicated that this new N2 fixation approach included a H addi- way. As the ⁄NNH is stable, the addition of a second proton coupled
tion in the initial step and further sequential direct reduction by with electron transfer is considered to be the potential determin-
the ‘‘solvated electron” along with protonation. They also sug- ing step. The Volcano curve are proposed to estimate the most
gested the hydrogen shall be the main competitive products promising catalysts for the ammonia production. In Fig. 4, the met-
towards ammonia. als on the right sides with both stepped and flat surfaces exhibit
similar reaction mechanisms: addition of the first proton as the
2.2.2. Electrochemical synthesis of ammonia limiting step in the associative pathway, while cleavage of the tri-
Nørskov and Vegge et al. [53,61–63] have proposed a basic ple bond of dinitrogen is the determining step in the dissociative
model for the ammonia production for an electrochemical reduc- pathway. The metals on the left side share the identical RDS for
tion process. The Heyrovsky-type reaction is considered as the both the associative and dissociative pathways.
dominated reaction pathway since the Tafel-type is less effected
by the exerted potential and has a higher activation energy barrier Flat surface : NH þ Hþ þ e ! NH2 ð31Þ
than the former mechanism. The associative Heyrovsky mecha-
nism and dissociative Heyrovsky mechanism are widely accepted Stepped surfaces : NH2 þ Hþ þ e ! NH3 ð32Þ
and shown below.
Four different sections are separated by vertical lines (a, b, c and
(1) Associative Heyrovsky mechanism
d) in Fig. 4. The most energy favourable bonding species ⁄NH, ⁄NH2,
⁄
N2 þ $ N2 ð14Þ N2H, ⁄N2H2 locate in the first left section. Transition metals such as
Sc, Y, Ti and Zr bind N stronger than H are estimated to give a

N2 þ Hþ þ e ! N2 H ð15Þ higher selectivity towards ammonia rather hydrogen production.
On the top of the volcano curve are the metals (Rh, Ru) having

N2 H þ Hþ þ e ! NNH2 ðDistal pathwayÞ or ð16-aÞ the highest activity towards ammonia synthesis.
In order to achieve high Faradaic efficiency towards ammonia

N2 H þ Hþ þ e ! NHNHðAlternating pathway ½64 66Þ formation rather than hydrogen production, binding with N-
adatoms stronger than H-adatoms is recommended to select
ð16-bÞ [67]. Using ⁄N binding energy as the descriptor, Montoya et al.
[61] found that the limiting potentials for HER was consistently

NNH2 þ Hþ þ e ! N þ NH3 ðDistal pathwayÞ or ð17-aÞ less negative than the nitrogen reduction reaction (NRR), which
means the HER is always preferred in a solution with abundant sol-

NHNH þ Hþ þ e ! NHNH2 ðAlternating pathwayÞ ð17-bÞ vated protons. In their calculations, only the (1 1 1) and (2 1 1) sur-
faces are used to represent the catalysts, which could not cover all

N þ Hþ þ e ! NH ðDistal pathwayÞ or ð18-aÞ the conditions of practical catalysts. Howalt and Vegge et al.
[63,68] demonstrated that the ultrasmall Mo13 or Mo13Nx clusters

NHNH2 þ Hþ þ e ! NH2 NH2 ðAlternating pathwayÞ ð18-bÞ bonded strongly with N rather than H.
Associative adsorption of N2 is more preferred than the disso-

NH þ Hþ þ e ! NH2 ðDistal pathwayÞ or ð19-aÞ ciative adsorption in the low temperature electron driven NH3 syn-
thesis. However, ‘‘distal pathway mechanism” is not always

NH2 NH2 þ Hþ þ e ! NH2 þ NH3 ðAlternating pathwayÞ
ð19-bÞ

NH2 þ Hþ þ e ! NH3 ð20Þ

NH3 ! NH3 þ ð21Þ
(2) Dissociative Heyrovsky mechanism
2 þN2 ! 2 N ð22Þ
2 N þ Hþ þ e ! N þ NH ð23Þ

N þ NH þ Hþ þ e ! 2 NH ð24Þ
2 NH þ Hþ þ e ! NH þ NH2 ð25Þ

NH þ NH2 þ Hþ þ e ! 2 NH2 ð26Þ
2 NH2 þ Hþ þ e ! NH2 þ NH3 ð27Þ

Fig. 4. Volcano plots for the flat and stepped transition metal surfaces towards the
nitrogen reduction reaction producing ammonia via Heyrovsky step mechanism.
þ
NH2 þ NH3 þ H þ e ! 2 NH3 ð28Þ The solid and dotted lines represent the implementation with or without H-bond
correction respectively. The volcano plots are composed by the dissociative (solid
2 NH3 ! NH3 þ NH3 þ ð29Þ lines) and associative (dashed lines) mechanisms on the flat (black) and stepped
(red) surfaces. Reproduced from Ref. [62] with permission of the PCCP Owner

Societies. (For interpretation of the references to colour in this figure legend, the
NH3 þ NH3 þ ! 2NH3 þ 2 ð30Þ reader is referred to the web version of this article.)
preferred, from energy point of view, on the surfaces of catalysts. A Mars–van Krevelen mechanism pathway is considered specif-
Bao et al. [64] proposed ‘‘alternating hydrogenation mechanism” ically for nitride materials, on which surface the N vacancies (VN)
for electrocatalytic reduction of N2, which deviates from the path- are believed to play a critical role for N2 activation [73–75]. It is
way described by Eqs. (16)–(19), which is in agreement with the quite similar with the one we described in the Haber–Bosch pro-
Anderson et al.’s work [66] on the complex N2 fixation mimicking cess section. The reaction pathway involving the replenishment
the enzyme. Instead of the formation of ⁄N as an energy-favourable of VN can be described by the following equations:
step, they believe it is more likely to form ⁄NHNH2 which is more
3Hþ þ 3e ! 3 H ð33Þ
stable on the tetrahexahedral Au nanorod. The details will be
described and discussed in the later section. Additionally, the alter-
3 H ! VN þ NH3 ð34Þ
nating pathway is also believed very likely to take place on the
(0 1 0) surface of BiOCl, which finally leads to a higher ammonia
VN þ N2 ! N ð35Þ
production rate than the distal pathway on the (0 0 1) facets [65].
Höskuldsson et al. [69] conducted the computational screening
N þ Hþ þ e ! NH ð36Þ
on the transition metal oxide in rutile crystal structure. A volcano
plot has been presented and the (1 1 0) facet of NbO2 is proposed as
NH þ Hþ þ e ! NH2 ð37Þ
the most promising one considering the balance between the activ-
ity and selectivity (Fig. 5).
NH2 þ Hþ þ e ! NH3 ð38Þ
It is worth mentioning that not all researchers believe the disso-
ciative adsorption of N2 in electrochemical ammonia synthesis is
NH3 ! NH3 ð39Þ
not likely to happen. Kordi et al. [70] proposed the dissociative
pathway based on the observation that there is no hydrazine Considering on some surfaces, the replenishment of VN by N2
detected in the experiments. And it is also supported by the DFT has a high energy barrier, dimer N vacancies (V2N) have been pro-
calculation that N2 dissociation on the Ru surface is not energeti- posed to reduce the energy barrier for the N2 molecules to fill the
cally forbidden [71]. V2N on the catalysts surface. The adsorption of N2 to fill the VN is
TiO2 is still the benchmark model representing the transition not a limiting step for most situations.
metal oxides. Comer and Medford [72] conducted calculation on
3Hþ þ 3e ! 3 H ð40Þ
the rutile (1 1 0) surface with the Bayessian error estimation func-
tional (BEEF-vdWs). In agreement with Nørskov’s works, the acti-
3 H ! VN þ NH3 ð41Þ
vation energy of N2 triple bond cleavage is prohibitively high,
>8.5 eV. Because scission of the bond is not affected by the applied
VN þ Hþ þ e ! VN H ð42Þ
potential bias, the dissociative pathway is highly unlikely to be
carried out even with a high potential energy. A high energy
VN H þ Hþ þ e ! VN 2H ð43Þ
requirement of 3 eV is also found to be required for stabilization
of ⁄NH, which indicates any pathway with the involvement of
⁄

VN 2H þ Hþ þ e ! VN 3H ð44Þ
NH is unfavourable. Therefore dissociation of the NAN is likely
to take place via the species of ⁄H2NNH2 forming a ⁄NH2 and a
⁄ VN 3H ! V2N ð45Þ
NH3. The ‘‘alternating pathway” seems to be more reasonable than
the ‘‘distal pathway”. In photocatalytic reaction, the CB position
V2N þ N2 ! ð46Þ
and the incident light energy determines the reduction potential
in the reaction process. Even only considering the associative path- In summary, the ammonia production pathways on defect-free
way, the photocatalytic N2 reduction to ammonia is not favoured catalysts are generally divided into two categories: the associative
on the pure rutile (1 1 0) surface [73]. The O vacancies and other and dissociative pathway based on the Heyrovsky’s mechanism. At
defects are proven to be indispensable in the stabilization of the room temperature, the associative pathway is favourable due to
reaction intermediates. the difficulties in breaking the triple bonds of N2. Under the
associative adsorption mechanism, there are two pathways, the
so-called distal pathway and alternating pathway, which two
pathways are different with respect to the hydrogenation sites
for ⁄NAN dominated on different surfaces. For the catalysts espe-
cially the nitride catalysts containing intrinsic N vacancies, the
Mars-van Krevelen mechanism is proposed, which involves the
replenishment of the Nv or N2V which largely reduces the energy
requirement for activation of the dinitrogen atoms. The mecha-
nisms investigated under the thermocatalytic, photocatalytic or
electrochemical backgrounds are almost exchangeable, while it is
important to keep in mind the different basis on which the authors
deduce their conclusions.
2.3. The quantitative measurement of ammonia
It is a tough job to quantitatively detect ammonium ions or

ammonia molecules in solution. Due to the low efficiency of most
of the photocatalysts and electrocatalysts, ammonia produced is
frequently at low ppm level, which makes it extreme challenge
Fig. 5. Volcano plot towards the potential determining step in electrocatalytic
ammonia synthesis on metal oxide surfaces: the limiting potentials are plotted
to generate accurate and reproducible measurements. In general,
against the binding energy of *NNH species. Reproduced from Ref. [70]. Copyright there are 5 different kinds of measurement methods recommended
(2017) American Chemical Society. by ASTM [76,77], US EPA [78] (United States Environmental
Protection Agency) and APHA [79] (American Public Health Associ- In summary, it is appreciated that the colorimetric methods are
ation): (i) high-performance liquid chromatography (HPLC) or ion highly sensitive, accurate and facile in quantification of ammonia
chromatography (IC); (ii) ion selective electrode (ISE); (iii) col- in aqueous solution with known components and good lab prac-
orimetry using Nessler’s reagent; (iv) colorimetry using phenate tice. The IC method having a higher requirement on instrument
method and (v) titration method. The first 4 kinds are widely used do provide interference-free, more precise and reproducible quan-
in recent researches and the titration method can only be used titative results for research purposes. Retention time in chro-
when the NH3AN concentration higher than 5 ppm. Thus the titra- matography is also a powerful tool to identify the unknown
tion method is not discussed in this review. products of catalytic reactions. Considering the complicated chem-
ical compositions in the solution when hole scavengers are used in
the catalytic reaction, it is highly recommended to take chromatog-
2.3.1. Colorimetric method
raphy methods instead in order to gain reliable results. When no
The colorimetry using the Nessler’s reagent is a very facile
organic hole scavenger used in the catalytic system, more than
method can be largely reproduced in any modern chemistry lab,
one method is recommended to conduct for the quantitative
which makes it the most popular method in the recently published
measurements of ammonia in ultralow concentration with careful
works. The colorimetric Nesslerization reaction is described by the
considerations of possible inferences.
following equation: [80]
ð47Þ 3. Photocatalytic N2 fixation to ammonia
The main issue of colorimetric method is the complicated 3.1. Semiconductor engineering
chemical compositions of products generated by the photo- and
electrocatalytic reactions that could interfere the measurements. Many different semiconductors have been studied as photocat-
The interferences could be caused by glycine, urea, hydrazine, ami- alysts for the photocatalytic reduction of N2 to ammonia, including
nes, aldehydes, alcohols in different principles [76,81–83]. Cations TiO2 [91], Fe2O3 [92,93], Fe(O)OH [94], WOx [95], CuO [96], Sm2O3,
might also affect the results by increasing the turbidity under alka- V2O3 [97], Ti-exchanged Zeolite [98,99] and GaP[100]. Strategies
line conditions. As aforementioned, the hole scavengers are widely were also proposed to further improve the catalytic performance
used in photocatalytic ammonia production researches. The oxida- and relevant review is presented below.
tion products of hole scavengers and the dissolving ions generated
via catalysts [84] could severely influence the colorimetric reaction 3.1.1. Facet design
of Nessler’s reagents [83,85]. Pre-treatments sometimes are neces- Bismuth oxyhalide is one of the materials that can achieve
sary for the colorimetry using Nessler’s reagents. Additionally, it preferably exposed facets under well-regulated synthesis condi-
has been reported that keeping the variables (such as temperature, tions. Bai et al. [101] reported their controllable syntheses of Bi5O7I
reaction time etc) constant is a good practice to obtain accuracy with different dominant facets. The (0 0 1) facets have a much
and repeatable results [86]. Another concern is the environmental higher activity towards the N2 fixation to ammonia than the coun-
impacts of mercury generated due to the use of Nessler’s reagents. terpart (1 0 0) facets. The higher CB position of (0 0 1) facets is
Proper handling of liquid wastes is of great importance. Phenate attributed to the better activity towards N2 fixation. Two dimen-
method is another colorimetric approach to determine the ultra- sional (2D) materials are another example of facet design, which
low concentrations of ammonia in the water (linear response up are promising in photocatalytic nitrogen fixation. In Sun’s work
to a few ppm) [87].The mechanism of this method is straight for- [102], ultrathin MoS2 was synthesized where trions [103] (two
ward as well: the indophenol generated by the reaction of ammo- electrons combined with a hole) were generated under light
nia, phenol and hypochlorite under the catalysis of sodium irradiation (Fig. 6). The trions assist the simultaneous transfer
nitroprusside shows absorbance peak centred at around 640 nm. of six electrons so as to achieve a high ammonia synthesis of
Potential facts that interfere accuracy of the measurement include 325 lmol h1 g1 without using any sacrificial scavengers or noble
precipitation by addition of citrate and other ions potentially metal co-catalyst. The 6-electrons’ transfer allows high selectivity
increase turbidity in alkaline solution. to ammonia because its by-products receive less electron-protons
which result in unfavourable formation. Layered-double-
hydroxide (LDH) with ultrathin layered structure is another kind
2.3.2. Methods with selectivity or separation of 2D materials showing potentials in photo-fixation of nitrogen.
The ISE method is widely used because of its few interfering fac- Zhao et al. [104] successfully synthesized a series different kinds
tors, due to the high selectivity to ammonia and a large detection of LDH with general formula of MIIMIII–LDH (MII = Mg, Zn, Ni, Cu;
range of NH3AN from 30 ppb to several thousand ppm. Relatively MIII = Al, Cr) via the co-precipitation and fast nucleation method.
complicated operation procedures, slow response at ultralow con- Among the synthesized LDH, the CuCr–LDH ultrathin nanosheets have
centration and instrument drifts are the main drawbacks of this shown the most outstanding performance in the N2 photofixation: the
method. Chromatography method is based on the analysis with ammonia generation rate is 184.8 lmol L1 in water without using
different components of analyte separated in column before reach-
ing the detector, which makes it more reliable than other methods
in principle [88,89]. Fast development in IC instrument let the IC
method become a standard analysis method for ammonia and
many other ions in water, recognized by international organisa-
tions such as ASTM, US EPA and ISO. The automation realised by
modern instrumentation ensures the quantitative detection to be
more accurate, precise and much less time-consuming. It has been
reported by using the cation-exchange column and conductivity
detector the linear range for ammonia detection could reach 100
ppb to 80 ppm. RSD (residual standard deviation) for retention Fig. 6. Scheme on multi-electron transfer based on trion inducing the ammonia
time and peak area could be less than 0.2% and 1% respectively synthesis from dinitrogen molecules. Reproduced from Ref. [102], Copyright (2017),
[90]. with permission from Elsevier.
any organic hole scavenger. The best ammonia generation rate, Besides N vacancies, the vacancies of other elements are also
220.9 lmol L1 was observed at pH = 5.0. It is also highlighted in found to be important in ammonia production. The oxygen
their work that atmospheric N2 is completely reduced to the NH3 vacancy (VO) also exhibits significant importance in the bismuth
with a selectivity nearly reaching 100%. The superior performance oxyhalide as photocatalysts for nitrogen fixation with water
of ultrathin LDH is attributed to the defect-abundant structure, [65,110]. Li et al. observed negligible N2 fixation by the VO free
improved electronic structure and a metal-metal electron transfer. BiOBr due to the fact that the CB position of BiOBr is not high
Further decrease in the dimension of nanomaterials will move enough to directly reduce N2. Generation of VO enabled electrons
to the region of quantum dots, which is treated as 0-dimension. to be trapped and donated to the adsorbed N2 molecules. The pro-
Sun et al. [105] prepared the BiO quantum dots with an average moting function of VO is highly associated with the position of VO
size of 2–5 nm. A outstanding performance of 1226 lmol g1 h1 on the facets of BiOBr. It has been found the reaction pathways are
is reported without using any co-catalyst or organic hole scav- different on 0 0 1 and 0 1 0 facets as presented in Fig. 7. On the 0 0 1-
enger. It is believed that the low valence state of Bi(II) has great facets (Fig. 7(a)), VO enhances the end-on adsorption of N2 so as to
electron donating capability and facilitates N2 activation conse- promote the selectivity of NH3 generation. On the 0 1 0-facets
quentially. And a one-step 3 electron-proton transfer were pro- (Fig. 7(b)), the side-on adsorbed N2 prefers producing N2H4 at first
posed to take place on the BiO surface. and followed by a fast ammonia production rate comparing with
the 0 0 1-facets with VO. Ultrathin Bi5O7Br nanotubes with the
3.1.2. Vacancies design diameter of 5 nm were synthesized and VO was induced by light
The N vacancies in nitride materials play a key role in N2 fixa- excitation. Similar to other photocatalytic reaction, VO trapped
tion. Dong et al. [106] calcined the C3N4 in nitrogen flow and gen- the excited electrons and prolonged the lifetime of the active car-
erated N vacancies over C3N4, which were supported by the rier. Wang et al. [97] proposed the trapped electrons to be directly
enhanced chemical adsorption of N2 tested using the TPD method. inserted into the dinitrogen molecules chemisorbed on the VO site,
Li et al. [75] applied KOH to etch the structure of C3N4 and used which is essential for the higher reducing ability in nitrogen reduc-
methanol as both the proton source and hole scavenger. The KOH tion. A 1.38 mmol h1 g1 rate of ammonia production and 2.3%
etching resulted in formation of N vacancies that significantly quantum efficiency was achieved using pure water to scavenge
improved the chemical adsorption of N2. It was also stated that holes.
the KOH etched C3N4 follows a Mars-van Krevelen mechanism. Hirakawa et al. [111] reported their work on the N2 fixation to
Ma et al. [107] presented their C3N4 with microwave-induced N ammonia with water on a noble metal free TiO2 system. Their Ti3+
vacancies. The ammonium ions production rate increased by three- abundant catalysts demonstrated an impressive N2 fixation perfor-
fold. Both of HCl [108] and methanol [109] were also identified to mance without using any efficient sacrificial scavengers. VO is also
be effective in inducing the N vacancies and increasing the surface reported to have significant effect on N2 adsorption and activation
area. The HCl treated and methanol treated catalysts achieved in the Au/TiO2 photocatalysts. Li et al. [112] proposed an associated
13.4-folded and 27.2-folded increases in the ammonia production reaction mechanism (Fig. 8) in TiO2/Au/amorphous TiO2. VO pro-
rate respectively compared to the pristine g-C3N4. motes the adsorption of N2 and the protonation process before
the nitrogen-nitrogen triple bond is broken down via activation
of the N2 molecule and stabilization of the NxHy reaction interme-
diates, respectively.
Liu et al. [113] implemented comprehensive computational
study on the N2 fixation performance of the 0 0 1-facets of BiOBr.
Different from the previously published results, Liu et al. claimed
that VO primarily increased the band edge position and induced
excessive electrons which enhanced the formation of high-energy
reaction intermediates. The Bi-terminated (0 0 1) facets are sug-
gested to be a better candidate for the ammonia synthesis than
O-terminated (0 0 1) facets.
Hu et al. prepared a series of ternary metal sulfides in the forms
of Zn0.1Sn0.1Cd0.8S [114] and Mo0.1Ni0.1Cd0.8S [115] with varying
stoichiometric ratios of metals and different S vacancy concentra-
Fig. 7. (a) Distal reaction pathway for N2 fixation on Vo–BiOCl (0 0 1) facets. (b) tions. From Fig. 9, it is to observe a linear relationship between
Alternating pathway for N2 fixation on Vo–BiOCl (0 1 0) facets. Reproduced from Ref. the ammonia production rate and sulphur vacancy concentrations:
[65] with permission of The Royal Society of Chemistry. more S vacancies ensure a higher activity towards N2 fixation.
Fig. 8. TEM picture of TiO2/Au/amorphous-TiO2 nanostructure and the proposed associative N2 hydrogenation mechanism. Reproduced from Ref. [112] with permission,
Copyright (2018) Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.
Fig. 9. Nitrogen photofixation performance of (a) ZnxSnyCdzS and (b) MoxNiyCdzS with different S vacancy concentrations. Reproduced with permission [114,115] with
permission from The Royal Society of Chemistry.
Alloyed semiconductor photocatalysts are an amazingly compli- capable of adsorbing, activating N2 molecules and further transfer-
cated system so that the detailed mechanisms of photofixation of ring photon-induced electrons to adsorbates. A 13.5-folded
N2 have not been fully understood. Further investigations are increase on the NH3 production rate is achieved comparing to
required. the pristine C3N4. DFT calculation results suggested when a N2
molecule was adsorbed on the Fe3+ site, electron was transferred
3.1.3. Doping design to N2 and the Mulliken charge reaches 3.1. The excessive elec-
It was reported in 1988 by Palmisano et al. [116] that doping of trons are believed to be able to activate the triple bond of dinitro-
Cr and Fe to TiO2 was effective in promoting the ammonia produc- gen molecules. The DFT calculation on the Fe doped TiO2
tion rate under light irritation. Soria and Conesa [117] in 1991 and conducted by Comer and Medford [123] suggested the promoting
Martin et al. [118] in 1992 confirmed the promoting effects by the effects of the Fe dopants: significantly decrease the energy
Fe doped TiO2 and Cr dope TiO2 systems respectively. Mg2+ [119], required on the stabilization of ⁄NANH intermediates in the
Ce, and V [120] were also identified to be useful in promoting associative reaction pathway.
the catalytic activity of TiO2. Cu was doped onto C3N4 for N2 reduction as well. Hu et al. [124]
Extensive investigations on the role of dopant have been docu- took advantage of molten salts as the hard template generating
mented recently. Fe3+ doping was found to be effective in TiO2 by C3N4 with high surface area and Cu(I) doping. By using both exper-
Zhao et al. [121]. Fe3+ acted as the electron and hole trap to pro- imental and computational method, Cu was confirmed to be +1
mote the formation of the Fe2+ and Fe4+ intermediates, which fur- valence state. The Cu(I)AN formed became an active site for the
ther assisted the charge-separation in the photocatalytic process. adsorption of N2 and further donated the electron induced by pho-
The successive proton-coupled-electron-transfer mechanism is ton back to N2. Mo was found to be active in the nitrogenase
proposed for this reaction, as shown in Fig. 10. The nitrogen is first ammonia production [1]. Inspired by the natural process, Hao
associatively adsorbed on the TiO2 surface with an electron trans- et al. [125] synthesized the Bi2MoO6 catalysts that demonstrated
ferred from the semiconductor surface to form N2@NA. Similarly, impressive ammonia photocatalytic synthesis ability with N2 as
HN@NA, ÆHNANHA, H2NANHA, H2NANH2, NH2 and NH2OH are the feedstock. Comparing to Bi2WO6 that had the same crystal
formed on the surface as the intermediate species in the ammonia structure as Bi2MoO6, Hao et al. verified the essential catalytic
synthesis reaction. Finally, NH2 reacts with another proton pro- activity of Mo in the catalysts. It is worth mentioning that a fully
vided by water to generate NH3 which is further desorbed from coordinated Mo has little catalytic activity since Mo is fully covered
the catalyst surface. Although Fe3+ doping does facilitate the by O atoms in the octahedron. Only the edge sites of Mo atoms
ammonia synthesis, direct interaction between the reactive inter- with unsaturated coordination have the capability to activate N2.
mediates and Fe ions are not emphasized in their work. Hu et al. Theoretical calculation suggested a different novel system: doping
[122] also doped Fe3+ ions onto C3N4 by co-calcination of Fe of Al onto graphene. Tian et al. [126] presented their DFT calcula-
(NO3)3 and C3N4 precursors. The Fe dopant is believed to be tion results suggesting the Al site itself is inert in activating N2
Fig. 10. Schematic illustration of the successive proton-coupled-electron-transfer mechanism in photocatalytic N2 reduction to ammonia. Reproduced from Ref. [121],
Copyright (2014), with permission from Elsevier.
and the graphene shows some redox ability that suitable for the N2 [132]. In their work, Fe2O3 exhibited the highest activity comparing
reduction. Therefore a reasonable combination of the Al site with the Fe2O3/TiO2 or even the Pd/Fe2O3/TiO2 under UV–Vis irra-
embedded in the graphene would lead to a robust catalyst for N2 diation. Taking high specific surface area of Fe2O3/TiO2 and Pd/
fixation. Detailed investigation demonstrated that the Al atom sites Fe2O3–TiO2 into account, the excellent performance of single com-
acted as both the binding site and catalytic centre, while the gra- ponent Fe2O3 became even more surprising. It may primarily due
phene functioned as an electron buffer. The simple system success- to high activity of Fe(II) in Fe2O3 in N2 activation [4]. However,
fully mimics the nitrogen fixation process naturally taking place in the detailed mechanism remains unambiguous.
the nitrogenase enzyme. Non-metal element doping is another Z-scheme is a special photocatalyst structure design originally
common strategy in improving the photocatalytic performance, featured by the electric conductivity layer between two different
especially important in preparation of metal-free photocatalysts photo-active semiconductors as the electron mediators. The
taking into account of environmental friendly catalysts. Feng recombination of the electron-hole pairs could be effectively
et al. [127] implemented co-doping of oxygen and sulphur to retarded without compromising the reducing and oxidizing ability
C3N4 and successfully achieved the ammonium production rate [133]. In recent years, it was found the conductive layer is unnec-
of 2.18 mmol g1 h1. The co-doping of O and S into C3N4 not only essary if two semiconductors are carefully chosen to be compatible
elevated and CB and VB positions but also resulted in heterostruc- with their CB and VB positions. Liang et al. [134] proposed that
ture of C3N4 between doped and undoped elements. Its intrinsic W18O49/g-C3N4 converted N2 to NH+4 based on Z-scheme without
electron transport ability was then improved. The co-doping dou- using electron mediator. C3N4 is considered to be the active part
bled the catalyst activity. in the reaction and W18O49 is the light absorber that further
enhances the charge separation. Similar profitable effects of the
3.1.4. Structure design Z-scheme is reported in the Mg1.1Al0.3Fe0.2O1.7/C3N4 heterostruc-
The heterostructure has been verified to be an effective way to ture as well [135]. Cao et al. [136] designed the structure compris-
prolong the lifetime of photo-induced charge carrier by providing ing the Ga2O3 and C3N4 and used the 3,4-dihydroxybenzaldehyde
efficient charge separation. In 1983, Li et al. [128] reported their to functionalize the Ga2O3. The aromatic ring is believed to act as
pioneer work on building up the semiconductor/semiconductor the electron mediating layer and helps the formation of Z-
heterostructure for ammonia photocatalytic synthesis with N2 as scheme structure. Their study presents the mechanism of the
the feedstock. The BaTiO3 and SrTiO3 are proven to be active under ammonia synthesis with alcohol as the hole scavenger: O2 firstly
incident light. By adding RuO2 and NiO to the catalyst system a tri- oxidized to H2O2 further turns to OH, which further reacts with
component system was formed resulting in a 5-fold increase in methanol forming the CO
2 . It is CO2 that finally promotes N2 to
ammonia yield. Another early work presented by Tennakone ammonia.
et al. [129] generated the interface between Fe2O3 and TiO2 so that Functional group modification of a photocatalyst is one of the
approximately 3-fold increase in ammonia yield was achieved. strategies to promote catalyst selectivity. Cao et al. [137] devel-
Recently, Liu et al. [130] prepared the Ti3C2 MXene layered oped a new C3N4 synthesis method by collecting gaseous produc-
materials by etching the precursor Ti3AlC2 with HF. Further calci- tion of urea p thermal polymerization. The sample synthesized
nation with melamine introduced the heterostructure of TiO2@C/ via this method shows a high surface area and abundance of amine
g-C3N4 with a surface area of 49.5 m2 g1. Its abundant surface functional groups. The nitrogen fixation rate with the bulk C3N4 is
defects with good electron transport ability ensured the excellent almost doubled. Sun et al. [138] took advantage of phosphate mod-
NH3 synthesis performance which was as high as 250.6 lmol g1 ification method to improve the performance of the LaFeO3 in N2
h1. Zhang et al. [131] took advantage of the combination of photofixation. By immersing the LaFeO3 nanoparticles in the phos-
C3N4 and ZnMoCdS, the 13.5-fold and 1.75-fold increase in ammo- phoric acid and further calcining for 2 h at 500 °C, it was claimed

nia synthesis rate have been reached under the illumination of vis- that PO3+4 dispersed on the LaFeO3 surface and the bond La–PO3
ible light. Sulphur vacancies, heterostructure and complicated is formed. An optimal ammonium ions production performance
composition are attributed to the enhanced ammonia synthesis of 250 lmol g1 h1 is achieved under 500 W Xenon light irradia-
performance. Surprisingly different from the aforementioned tion, as shown in Fig. 11(a). A ‘‘Pull and Push” strategy is proposed
research, the unexpected detrimental effects of heterostructure in their work to illuminate the enhancement in the N2 activation
on the N2 photofixation was reported by Lashgari and Zeinalkhani and ammonia production (Fig. 11(b)). The Fe3+ is assumed to be
Fig. 11. (a) N2 photofixation performance of LaFeO3 with different phosphate modification degrees. (b) The scheme of proposed ‘‘pull and push” strategy in activation of N2.
Reproduced with permission from [138]. Copyright (2017) American Chemical Society.
the active sites for N2 adsorption as the ‘‘push” sites. The decreased due to the plausible oxidation of NH3 to NO 3 . Additional
FeALaAOAPAOH groups derived from phosphate modification analysis pointed out that hydrogen evolution competed with the
acted as Lewis acid sites to provide further ‘‘pull” sites in the N2 ammonia production. Noble metals including Ru, Rh, Pd and Pt
activation. The synergistic effects of the pull and push are thought were anticipated to be active in N2 activation [142]. The methanol
to be the reason for a strong chemical adsorption and efficient acti- is used as the hole scavenger. The activity towards N2 fixation is in
vation of N2. the order of Ru > Rh > Pd > Pt. Ranjit et al. also correlated the
A recently developed photocatalyst, which is named ‘‘chalco- metal–H bond strength to nitrogen fixation capability. Stronger
gels”, is quite promising and can mimic the nitrogenase behaviours metal-H bond is considered to have a higher ammonia yield.
in the N2 fixation process. Liu et al. and Banerjee [139,140] pre- Semiconductors other than the benchmark TiO2 have been used
pared the porous aerogel, ‘‘chalcogel”, via the metathesis reaction, as the supports and proved to be effective as well. Janet et al. [143]
which contains double-cubane [Mo2Fe6S8(SPh)3] and single- developed Pt/ZnO photocatalyst achieving the ammonia produc-
cubane (Fe4S4) or Sn2S6 clusters, as schematically presented in tion rate of 860 lmol h1 g1. It is worth mentioning that the
Fig. 12. Although the natural nitrogenase doesn’t need the light hydrogen production rate is less than half of the ammonia produc-
to realize the N2 fixation, the prepared chalcogels doesn’t react tion rate even though metals with most prominent HER activity
with N2 under dark condition. Another point needed to be empha- such as Pt were used as the co-catalyst. Ultra-small Ruthenium
sized is that water doesn’t act as an active proton source in the clusters were reported to be a highly efficient co-catalyst for
reaction. Pyridinium hydrochloride has to be added in the solution photofixation of nitrogen when it is supported by n-type gallium
to effectively provide protons needed for the NH3 production. In nitride nanowires [144]. The Ru nanoclusters synthesized via the
general, the ability to mimic the natural N2 fixation process has high temperature decomposition of Ru3(CO)12 in gas phase, shows
been advanced significantly. To date, however, the production of a mean diameter of 0.8 ± 0.4 nm. The Ru clusters are efficient in
ammonia from N2 is still very low. extracting electrons from the 3rg bonding orbitals of the adsorbed
N2 and back-donating the electrons to the antibonding orbital 1p⁄g
3.2. Co-catalyst design [145]. The Fermi energy level of Ru is lower than that of the n-type
GaN, therefore a negatively charged Ru cluster can be formed due
3.2.1. Transition metal as a co-catalyst to the spontaneous electron transfer. Under UV illumination, the
Transition metals functioned as a co-catalyst can promote pho- schottky barrier formed between the Ru and n-GaN can further
tocatalytic efficiency from two different aspects: enhancing the suppress the recombination and ensure a high gas phase ammonia
electron-hole separation and fine-tuning adsorption of key inter- synthesis.
mediates. Early work conducted by Rao et al. [141] compares the Graphene is another popular material in dispersing the active
photocatalytic N2 fixation abilities of Fe, Ru or Os supported on metal sites over photocatalysts, while the active metal nanoparti-
TiO2 under the irradiation of UV light. The catalytic activation is cles might migrate and aggregate during the reaction, which
found decreasing in the order of Ru > Fe > Os. The photocatalytic causes severe catalyst degradation. Lu et al. [146] prepared a
reduction of N2 was conducted in the aqueous solution with ascor- Fe/3D-garphene structure which is active for the ammonia produc-
bic acid as the hole scavenger. It was found that the ammonia con- tion using gaseous N2 and H2 as feedstocks. It was stated that Fe2O3
centration reached the maximum within 15 min and then with the assistance from the graphene support could give an
Fig. 12. (a) Schematic illustration of the chalcogels FeMoS–FeS–SnS containing the Mo2Fe6S8(SPh)3 and Fe4S4 clusters which are linked by Sn2S4
6 clusters. (b) Optical (c) SEM
(d) TEM images of the prepared chalcogels. (e) The quantitative analyses of the ammonia produced by the photocatalytic reaction on the chalcogels. (f) NH+4 production rates
of different chalcogels. Reproduced from [139], Copyright (2016) National Academy of Sciences.
Fig. 13. Mechanism proposed for the Fe/3D-graphene photocatalyst: (a) the role of graphene; (b) the role of Fe active sites. Reproduced from Ref. [146]. Copyright (2016),
American Chemical Society.
ammonia yield of 24 lmol g1 h1. The mechanism is illustrated in

Fig. 13. The graphene foam in the composite is believed to reach a
‘‘reverse saturation state” under illumination which induces ‘‘hot”
electrons in the graphene matrix [147]. The hot electrons further
transfer to Fe active centre or direct donate to N2 molecules. Fe acts
as the active sites are well accepted in the thermocatalytic N2 acti-
vation. The excessive electrons generated by the incident photons
could further enhance the back donation of electrons from Fe d-
orbitals to N2 p⁄ antibonding orbitals. The overall effect of external
light energy leads to significant reduction of the apparent activa-
tion energy from 160 kJ mol1 to 41 kJ mol1. Yang et al. [148]
introduced Al2O3 nanoparticles to the Fe/3D-graphene and con-
trolled the dispersion and morphology of Fe2O3. A higher NH3 pro- Fig. 14. Scheme of plasmon-induced ammonia synthesis equipment with two
duction rate and superior stability was reached. separated reaction chambers. Reproduced from [153] with permission. Copyright
(2014), WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.
3.2.2. Other as the co-catalyst

In Ye et al.’s work [149], Ni2P has been used as the co-catalyst type semiconductor support, which is the so-called the electron
supported on the Ce0.5Zn0.5S solid solution. The ammonia synthesis injection [152]. Oshikiri et al. [153] designed a reactor that parted
rate is even higher than the Pt/C3N4 benchmark catalyst. Addition the oxidation and reduction reaction in two separated chambers so
of different reaction radical quenchers to the solid solution during that the oxidation of the synthesized ammonia was hindered, as
the reaction led to a finding that improved electron-hole separa- shown in Fig. 14. They successfully demonstrated the capability
tion plays a key role in enhancing photocatalyst activity while of LSPR in driving the photocatalytic ammonia synthesis under vis-
quench of OH or O2 has little effect on the overall reaction rate. ible light. Au nanoparticles supported on (BiO)2CO3 nanodisks were
Recombination is still the most significant challenge in photocat- synthesized by Xiao et al. [154] via a facile hydrothermal method.
alytic N2 fixation. Black phosphorus is a novel 2D material which Under simulated 1 sun irradiation, a 2.5 times increase in ammonia
is intrinsically different from bulk phosphorus. Qiu et al.’s work production were observed due to deposition of Au NPs as the sen-
[150] demonstrated that the black phosphorus decorated on C3N4 sitizer. However, they also observed that the active sites disap-
nanosheets advanced photocatalytic efficiency of nitrogen fixation. peared after 1 h reaction and the reason is still ambiguous. Au
The 2D–2D coupling and formation of CAP covalent bonds are NP has also been deposited on silicon nanowires for enhancing
attributed to the enhancement of both photon excited electron- light irradiation response. Ali et al. [155] prepared well aligned
hole separation and electrons transfer between the 2D layers. Li p-type silicon nanowires via the reactive ion etching method. A
et al. [151] observed that carbon nanocrystals deposited on WO3- layer of gold nanoparticles were sputtered to the nanowires with
H2O can significantly (58-fold) improved ammonia evolution rate sizes ranging from 3 to 30 nm in diameter. It was also found that
of 205 lmol g1 h1. The carbon nanocrystals with unsaturated the ammonia production rate was in linear correlation with the
sites are believed to be active in weakening the N, N triple bonds nitrogen pressure used in the reaction system, which indicates
due to that the sp2 carbon could further participate in the the extremely low solubility of N2 in water is one of the factors
hybridization of chemisorbed N2 molecules. suppressing the overall efficiency of the photocatalytic N2 fixation.
Although using sodium sulfite as the hole scavenger, the quantum
3.3. Plasmon-induced nitrogen fixation efficiency is still very low, far away from practical application.
Utilization of visible light instead of UV illumination is of para- 4. Electrocatalytic N2 fixation to ammonia

mount importance in photocatalysis. Recently localized surface
plasmonic resonance (LSPR) has become one of the promising The electrochemically catalytic ammonia synthesis could be
strategies to realize the effective usage of visible light energy. LSPR classified into different categories by their electrolytes and work-
generates high energy electrons that can further transfer to the n- ing temperatures. We will focus on the aqueous electrolyte and
polymer electrolyte, since lower working temperature is always 4.1.1. Transition metal based electrocatalysts
preferred as the goals of our research (see Fig. 15). Bao et al. [64] reported the ammonia electrocatalytic produc-
tion at atmospheric pressure and ambient temperature using
4.1. Aqueous electrolyte system well-designed tetrahexahedral Au nanorods as the cathodic mate-
rials. As illustrated in Fig. 16(a), a Nafion membrane was used to
Aqueous electrolyte system is growing fast due to its mild reac- separate the dual-chamber electrolytic cell filled with 0.1 M KOH
tive condition: ambient pressure and temperature. Especially aqueous solution as the electrolyte. The redox half reactions take
water itself can act as proton sources, which is very appealing place on the cathode and anode can be expressed below:
because the usage of H2 is avoided.
Cathode 12H2 O þ 2N2 þ 12e ! 4NH3 þ 12OH ð48Þ
Anode 12OH ! 3O2 þ 6H2 O þ 12e ð49Þ
Overall 6H2 O þ 2N2 ! 4NH3 þ 3O2 ð50Þ

It is seen in Fig. 16(b) that the optimal potential to drive the
NRR reaction is around 0.2 V vs RHE (reversible hydrogen elec-
Fig. 15. The classification of the electrocatalytic system for the ammonia synthesis
trode). Under higher potential both ammonia production rate
based on the kinds of electrolytes. and Faradaic efficiency decreased dramatically, while hydrogen
Fig. 16. Electrocatalytic nitrogen reduction to ammonia using the tetrahexahedral Au nanorod. (a) Schematic diagram of the electrocatalytic N2 reduction setup using a dual-
chamber electrolytic cell separated by Nafion film. NH3 average yields and Faradaic efficiency versus the (b) potential bias and (d) temperature during the reaction. (c) The
stability tests of the electrocatalysts. (e) Proposed reaction pathway for Au electrocatalytic N2 reduction to ammonia based on the DFT calculation results. Reproduced with
permission [64]. Copyright (2016) WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.
production as the side reaction was much enhanced. At room tem- cessfully dispersed onto the TiO2 support (P25) and work as the
perature and ambient pressure, the average yields of the ammonia cathodic electrocatalysts. A new reduction process with the tannic
and hydrazine hydrate (the most common N containing side pro- acid acting both as the reductant and surfactant ensures the suc-
duct and important intermediate for NH3 production) are 1.648 cessful preparation of TiO2 supported subnanometer Au clusters.
and 0.102 lg cm2, respectively. The total Faradaic efficiency The ultrafine Au cluster gives a high NH3 yield of 21.4 lg h1
based on the two products reached 4.02%. Another important fac- mg1 at the potential bias of 0.2 V vs RHE. An impressively high
tor influencing the NRR reaction is temperature. Although varying Faradaic efficiency of 8.11% has been achieved, which is among the
in a relative small range of 5–65 °C, both NH3 production rate and highest values achieved in aqueous solution, especially the strong
Faradaic efficiency show monotonic increase with temperature acidic solution. The 1.6 times increase in the NH3 yield and as high
(Fig. 16(d)). The observation confirms that mass transfer is a limit- as 13.5% for the Faradaic efficiency have been measured when
ing factor in the electrocatalytic process. Based on the catalytic the reaction temperature reached 60 °C. Different from the tetra-
performance at different temperatures, the calculated activation hexahedral Au nanorods towards NRR in acidic aqueous electrolyte
energy for NRR using Au nanorods as the catalysts is only 13.704 [64], the subnanometer Au/TiO2 is proposed to follow the distal
kJ mol1, which is considerably smaller than Harber–Bosch process pathway in ammonia synthesis. Another example for the distin-
of 335 kJ mol1 without using any catalyst [159]. The authors also guished N2 fixation performance of ultrasmall Au cluster is
proposed the ‘‘alternating hydrogenation” pathway (Fig. 16(e)) for reported by Li et al. [161]. The amorphous Au nanoparticle well
NH3 production. The adsorbed ⁄N2H2 on the surface could be either dispersed on the ceria-reduced graphene oxide matrix achieves a
⁄
NNH2 or ⁄NHNH. It is revealed that the RDS is the addition of the high Faradaic efficiency of 10.10% with the yield rate of 8.3 lg
first proton to associatively adsorbed N2 with energy barrier higher h1 mg1. The amorphous or low-crystalline Au exhibits superior
than 2 eV. Fig. 16(f) presents the calculated results based on the activity to the high-crystalline counterpart. The higher defects
distal mechanism [62], as presented in Eq. (17). On the Au (2 1 0) and structural distortion are responsible for the enhancement.
and (3 1 0), the energy required to generate the ⁄N is quite high Since Au shows impressive Faradaic efficiency and activity
compared to the alternating hydrogenation pathway where the ⁄- towards electrocatalytic ammonia synthesis, it is of great impor-
NHNH2 and ⁄NH2NH2 are formed. Identification of hydrazine in tance to investigate the reaction pathways. In-situ FTIR is one of
the products serves as a further evidence that supports the pro- the most powerful techniques to experimentally explore it. Yao
posed mechanism. et al. [162] employed the surface-enhanced infrared absorption
A more recent work carried out by Shi et al. [160] employed a spectroscopy (SEIRAS) technique to investigate the intermediates
similar two-chamber electrolyte cell separated by a Nafion 211 generated on Au surface during the ammonia synthesis reaction.
membrane with 0.1 M HCl aqueous solution as the electrolyte. As presented in Fig. 17(a), the reaction pathway proposed is similar
The Au nanoclusters with an average diameter of 0.5 nm are suc- to the alternating hydrogenation pathway where hydrazine is an
Fig. 17. (a) Schematic illustration of the electrochemical ammonia production from N2 on the surface of Au. The alternating hydrogenation pathway is proposed in this work.
The FTIR spectra received during the linear potential scan from 0.4 to 0.5 V in N2-saturated KOH solution on (b) Au and (c) Pt surface. Reproduced with permission from Ref.
[162]. Copyright (2018) American Chemical Society.
important intermediate or a by-product. Since Pt is an optimal HER 0.35, 0.30 eV for ZrN, NbN and VN respectively. Both the associa-
catalyst, its surface was carefully characterized and compared with tive and dissociative replenishment mechanisms are thought to
Au. Comparing the FTIR spectra for Au (Fig. 17(b)) to that for Pt be able to happen on the transition metal nitride nanoclusters.
(Fig. 17(c)), one may notice that the main differences present in Besides nitrides, the carbides is another kind of promising electro-
the range of 1200–1500 cm1. The HANAH bending at 1453 catalysts for the ammonia synthesis from the dinitrogen mole-
cm1, ANH2 at 1298 cm1 and NAN at 1109 cm1 can be observed cules. Azofra et al. [170] presented their suggestion that the
on the Au surface but absent on the Pt surface. The results persua- MXene, a 2D metal carbide material could potentially be the robust
sively support the results that Au has a higher N2 adsorption than catalysts towards the ammonia synthesis. By doing the DFT calcu-
Pt. It may be due to the fact that N and H are competitive adsorbed lations on the adsorption of N2 molecules on the surfaces, it is
on metal active sites while Pt prefers the H adatoms. found the strong N2-philicity on these materials. Further detailed
In addition to Au, Rh, Os and Ru are also promising candidates reaction pathway investigations show that the transition metal
for the electrocatalytic ammonia synthesis [163]. Using the linear carbides with d2 electronic structure prefer the distal pathway
sweeping technique in 0.5 M H2SO4 aqueous solution, Kulger and the d3 and d4 ones intend to go along the alternating hydro-
et al. [164] compared the activities of Rh and Ru deposited on ran- genation mechanism. The Li et al. [171] also theoretically proves
domly structured Ti felts. The ammonia production rate is 8 times that the doping of Fe into the MoN2 monolayer shows ability fur-
higher on Ru than Rh. The DFT calculation conducted by Ishikawa ther reducing the energy barrier needed for the RDS: N2 adsorption
et al. [165] suggested that the activity sequence of these noble and replenishing VN.
metal were Ru > Os > Rh. Liu et al. [166] reported their work on There is another kind of electrolytic cell combining the gas
the synthesis and evaluation of the Rhodium nanosheets for N2 fix- phase chamber and liquid phase chamber together by using the
ation. A high NH3 yield rate of 13.88 lg h1 mg1 was achieved and polymer as the separator. Then the electrocatalytic N2 fixation
the robust HER activity of Rd nanosheets was attributed to its could run in the gas phase in the cathodic chamber on the gas/solid
unique structure and relatively low Faradaic efficiency (0.217%). interface and the proton could be provided by the electrocatalytic
For non-noble metals, Mo is a candidate showing great poten- water splitting in the aqueous electrolyte inside the anodic cham-
tial for ammonia electrocatalytic synthesis. Zhao et al. [167] calcu- ber. It is worth mentioning that we include this type of setup into
lated single Mo atoms dispersed on the BN monolayer for N2 the liquid electrolyte type while in some other research works or
fixation. The synergistic effects of B vacancies and Mo single atom reviews, it could be classified into the solid electrolyte section [26].
allow a voltage as low as 0.19 V to initiate the ammonia synthesis. As early as 2000, the Ru/C was reported by Kordali et al. [70] to
The enzymatic pathway with both N atoms of the dinitrogen mole- be active for the ambient condition ammonia synthesis. The setup
cule adsorbed on single Mo atom is believed to be the main path- used is a gas phase cathodic N2 reduction combined with the OER
way for the reaction. Mo film with micro-nano surface structures half reaction in the KOH aqueous solution on the surface of anode.
prepared by Yang et al. [168] also showed the ability to produce The ammonia has been successfully synthesized using N2 and
ammonia with atmospheric N2 as the feedstock. A yield of water at ambient temperature and pressure with the production
0.0309 nmol s1 cm2 is reached under a driving potential of rate of 1.3 lg h1 cm2. Recently Chen et al. [172] synthesized
0.49 V (vs RHE) and an optimal Faradaic efficiency of 0.72% under Fe2O3/CNT to act as the cathodic catalysts for ammonia synthesis
0.14 V (vs RHE). using N2 and water. A two-chamber cell comprising a gas phase
Transition metal nitride has also been researched for N2 fixa- flow reactor (cathodic) and a liquid phase electrolytic chamber
tion. Abghoui et al. [74,169] presented their comprehensive theo- (anodic) was used, which were separated by a Nafion 115 mem-
retical investigations on several transition metal nitrides. They brane as the PEM (Fig. 18). The anode materials are chosen with
proposed the RuN, WN, VN, NbN and CrN were the most promising Pt, the benchmark electrocatalysts for OER and electrolyte in the
nitride materials for ammonia synthesis. As described in the mech- anodic cell is KHCO3 aqueous solution. An optimal ammonia yield
anism section, the nitride materials tend to follow the Mars–van of 2.18 mg m2 h1 has been reached in the paper when 2.0 V (vs
Krevelen mechanism featured with the interaction of the N vacan- Ag/AgCl) is applied to the cell. A total Faradaic efficiency including
cies or dimer N vacancies with N2 molecules. They observed that both the NH3 and H2 production is as high as 95.1%, while if only
on specific surfaces of all the nitrides, the dissociative way for consider the NH3 it shall fall sharply to 0.025% at the same poten-
the replenishment of the N vacancies are thermodynamically pre- tial bias. It means the majority of the electrons consumed in the
ferred. The energy barrier for the dissociation of N2 under the inter- cell are used to generate hydrogen rather than ammonia. The same
action with nitrogen vacancies could be considerably small: 0.11, authors [173] also presented another work based on the Fe2O3/CNT
Fig. 18. (a) Schematic illustration of the electrocatalytic flow reactor for gas phase NRR with aqueous solution in the anode chamber providing protons. A PEM (proton
exchange membrane) is used as the separator. (b) The improved design of the same reactor by adding another gas diffusion layer. Reproduced with permission from Refs.
[172,173]. Copyright (2017) Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim and Copyright (2017) American Chemical Society.
with the Fe loading ranging from 5% to 45% investigated with the with the photocatalytic process. The two-chamber cell can effec-
similar electrocatalytic cell described above. And it was found that tively separate the ammonia produced from the oxidizing environ-
the ammonia synthesized in the cathodic chamber would back dif- ment, which reduces the negative effects of O2 in the ammonia
fuse to the anodic chamber through the PEM and being oxidized production and products distribution. In their work, the ammonia
into N2. The back diffuse phenomenon indeed causes the detrimen- production rate is 1.02 mmol g1 h1 using air as reactant, which is
tal effects on the ammonia production, therefore an additional gas only slightly less than 1.40 mmol g1 h1 using pure N2. The result
diffuse layer was added to the Nafion PEM on the gas phase cham- is expected and quite acceptable. Addition to carbon material,
ber side. The optimized design not only increases the ammonia metal-free BN is a promising catalyst. Although loading of metal
production rate by retarding the crossover of the ammonia, the nanoclusters will improve the performance, catalyst BN itself
selectivity (Faradaic efficiency) also increases due to the proton [167] is also quite active for the ammonia synthesis from N2 [175].
arriving rate has been effectively slowed down and fewer H2 is pro-
duced on the cathodic electrocatalysts surface. The final ammonia 4.2. Polymer electrolyte
yield reaches 1.06 102 nmol cm2 s1 with 0.164% Faradaic effi-
ciency, which are 158% and 571% higher than the old configuration The electrolytic cell using polymer electrolyte, different from
values. Different electrolytes including 0.5 M KOH, 0.5 M KHCO3, the cells using polymer membrane for separation, has two gas-
0.25 M K2SO4 and 0.25 M KHSO4 in the liquid chamber are investi- phase reaction chambers and it can be further classified into a pro-
gated to shed light on the electrolyte effects on the reaction. As it is ton conducting polymer cell and an anion conducting polymer cell
able to be expected, the KHSO4 acid solution could provide the pro- [28]. The Nafion is known as the proton exchange membrane
ton much more efficiently than the alkaline solution, which will (PEM) has long been used in the electrochemical setups such as
greatly facilitate the HER. Considering the HER and NRR are com- fuel cells and it still outperforms most of the alternatives [176].
petitive reactions on the catalysts, the NH3 yield rate is the lowest Transporting of anions, e.g. OH through polymer is an alternative
among all the 4 electrolytes used. We noticed that although the of PEM. The schematics of the cell assembly with polymer elec-
0.5 M KOH gives the best ammonia production rate, the difference trolyte is illustrated as in Fig. 19.
between the best and worst is not very significant. Another impor- Lan et al. [177] used the Nafion 211 membrane as the solid elec-
tant point is the non-linearity of the ammonia production rate trolyte for the ammonia synthesis from N2 and humidified H2/Ar
when higher potential-bias is applied to the electrocatalytic cell. mixture. Pt/C was chosen as the catalyst for the cell. The ammonia
When higher potential conducted, the ammonia exhibits a faster production rate was 31 lmol m2 s1 with a potential bias of 0.2 V
production rate during the first several hours. After that, the yields exerted between electrodes reaching a Faradaic efficiency of 2%.
decreases slowly under higher potential bias and finally falls to the It was also demonstrated that the ammonia could be directly syn-
same as the lower voltage one. Meanwhile, the overall current den- thesized using air and water instead of pure N2 or H2. It is also
sity remains the same, which means the selectivity towards NH3 noticed that when H2 is replaced by water in the anode side, a
drops. By using the operando XAFS and XPS ex situ characteriza- much higher voltage (1.6 V) is needed to allow a comparable
tion method, the valence state changes of the Fe in the catalysts ammonia production rate. It was assumed by the authors that a
is attributed to this phenomenon. When the higher potentials bias more efficient anodic catalyst was needed for the splitting of water
applied, the Fe in the Fe2O3 intends to be reduced to metallic sta- and supplying the protons. However, in N2 (Air) and H2 system,
tus, which is considered to be more active towards the HER rather further increase the potential will actually cause the decrease of
than NRR. the ammonia production. This phenomenon could be explained
by the low proton supplies in the cathodic part. Moreover, the Far-
4.1.2. Transition metal free electrocatalysts adaic efficiency is only about half when the water is used as the
Carbon as one of the most abundant and non-toxic materials proton sources rather than the hydrogen. Besides the noble metal,
have been extensively investigated with different allotropes and
microstructures. Liu et al. [174] presented their work on the MOF
(metal-organic framework) derived N doped porous carbon
towards the electrocatalytic NRR. The ZIF-8 (zeolite imidazolate
framework) was calcined 750 °C in N2/H2 atmosphere then etched
the Zn element by concentrated HCl. The metal-free N doped por-
ous carbon could be prepared and exhibit the microporous struc-
ture inherited from the MOF precursor with ultrahigh specific
surface area (896–1084.5 cm2 g1). The ammonia production rate
reported in their work reaches 1.40 mmol g1 h1 at the 0.9 V vs
RHE. The pyridinic and pyrrolic N are deemed as the active sites for
the ammonia production. Comparing with the electrocatalytic per-
formance run with Ar bubbling to remove any N2, the Faradaic effi-
ciency is 1.42% at 0.9 V vs RHE. Considering the electrolysis was
conducted in the 0.05 M H2SO4 solution which largely facilitates
the HER, the quite high current efficiency should be attributed to
the low HER activity of the N-doped carbon materials. While it is
noticed that the proton abundant electrolyte does promote the
production rates of both HER and NRR, which makes the improve-
ment of Faradaic efficiency an extremely challenging issue. Further
concerns about the feedstocks are also investigated in this work
since the replacing of the pure N2 with air could largely reduce
the cost for the practical industrial level of ammonia production.
Considering the continuous electrons provided by the power sup-
ply with potential bias, the electron scavenger role of O2 is less Fig. 19. The schematic illustration for the gas-phase electrodes assembly with ions-
detrimental to the electrocatalytic ammonia production comparing conducting polymer as electrolyte.
the oxides could also act as the electrocatalysts for low tempera- age of 0.25 V. The corresponding ammonia production rate is
ture ammonia synthesis. Xu et al. [178] reported a very high Fara- 0.028 nmol cm2 s1. Optimal ammonia yield is achieved by using
daic efficiency of 90.4% in ammonia synthesis at 80 °C from the N2 Pt/C with the cell potential bias of 0.5 V and the ammonia produc-
and H2. The SmFe0.7Cu0.1Ni0.2O3 was chosen as the cathode cata- tion rate was 0.04 nm cm2 s1. At this point, the Faradaic poten-
lysts, NiO–Ce0.8Sm0.2O2d as anode catalysts and Nafion as the tial falls dramatically to 0.01%, which is one order of magnitude
proton-conducting electrolyte. A high ammonia production rate smaller than the former Ir/C catalysts.
of 11.3 nmol cm2 s1 is reached by using the setup mentioned It is also of great interests to compare ammonia production per-
above. formance of conventional noble metal in both the proton-
Anions conducting polymer is another choice for the gas-phase conducting membrane and anion-conducting membrane elec-
electrochemical ammonia synthesis. Kong et al. [179] reported an trolyte. Nash et al. [181] prepared the cathode electrodes with Pt,
assembly of membrane electrodes with IrO2 as the anode catalysts Ir, Pd, Ru and Au supported by carbon respectively and evaluated
for water splitting and carbon fibre paper as cathode. c-Fe2O3 was their N2 electrocatalytic fixation with both proton exchange mem-
sprayed onto the carbon fibre. IrO2 is one of the most promising brane electrolyzer and hydroxide exchange membrane elec-
electrocatalysts towards OER. The membrane anode and cathode trolyzer, presented in Fig. 20(a and b). Noble metals supported
is assembled together with an anion exchange membrane as the on carbon performed differently in proton exchange membrane
electrolyte to form the flow gas phase reactor, as shown in (PEM) than hydroxide exchange membrane (HEM) (Fig. 20(c and
Fig. 20. The ammonia production rate in the flow gas phase electro- d)). The optimal ammonia production rate was achieved by the
catalytic chamber achieved a 3-fold increase (as high as 0.164 Ir/C with both the PEM and HEM configuration at the cell potential
nmol h1 cm2) than the conventional electrolytic cell using 0.1 of 0.2 V vs RHE: 2.09 103 nmol cm2 ECSA s
1
and 0.43 103
2 1
M KOH aqueous solution as the electrolyte. However, the Faradaic nmol cmECSA s . The Faradaic efficiencies were different. In HEM
efficiency drops from 1.9% in aqueous KOH solution to 0.044% in configuration, the highest selectivity was provided by Ru/C at
the membrane electrode assembly. The authors suggested that around 0.053%, while Au/C presents one order of magnitude higher
the lower pH value in the membrane electrode than the alkaline than the PEM optimum: 0.55%. The work has demonstrated that
aqueous solution was responsible for the significantly enhanced the alkaline electrolyte environment benefits reaction selectivity.
HER kinetics. On the other hand, acidic solution encourages higher kinetics for
A membrane electrolyte in gel form could also be used in the both NRR and HER. Another inspiring point obtained was that Au
gas-phase N2 and H2 reaction for the generation of ammonia. is a promising catalyst in enhancing selectivity. In HEM configura-
Sheets et al. [180] mixed polyacrylic acid homopolymer with tion, the Faradaic efficiency achieved by Au is much higher than its
KOH solution to prepare the gel which has the conductivity counterparts.
towards anions. It was proposed by the authors that the polymer
gel electrolyte will retard the water transport and suppress the 4.3. Molten salt electrolyte
HER, which is the main competitive side reaction to the N2 fixation.
However, as Pt/C and Ir/C were used as catalysts, low Faradaic effi- Electrolyte systems based on molten salts have been frequently
ciencies are expected due to their high activity to the HER. The reported with typical operating temperatures ranging from 300 to
highest Faradaic efficiency reaches 0.1% for Ir/C with the cell volt- 500 °C. Murikami et al. [182] have carried out multiple molten salt
Fig. 20. The (a) scheme (b) photograph of the membrane electrode assembly used for gas phase evaluation with anion conductive polymer electrolyte. The ammonia
production rate and corresponding Faradaic efficiency in the setup using the (c) proton-conducting polymer electrolyte and (d) anion-conducting polymer electrolyte.
Reproduced with permission from Refs. [179,181]. Copyright (2017) American Chemical Society.
electrolyte experiments. One example was to use a molten LiCl–

KCl–CsCl containing 0.5 mol% Li3N as a source of nitride ions with
nitrogen and hydrogen gas acting as reactants respectively.
Nitrides in the electrolyte were oxidised at the aluminium anode,
which then reacted with the hydrogen gas to form ammonia. The
nitride ions were replenished by nitrogen gas undergoing reduc-
tion at the porous nickel cathode. The proposed mechanism is
shown in Fig. 21 below. The ammonia production rate reached
3.33 108 mol s1 cm2 at 400 °C with a current efficiency of
72%. Methane was also examined to replace hydrogen gas with
the same molten salt electrolyte [183]. Methane has to be decom-
posed first to generate hydrogen which further reacts with the oxi-
dised nitride ions. The temperature for ammonia synthesis needs
to be 500 °C, lower than which methane decomposition to hydro- Fig. 22. Principle of ammonia synthesis via electrocatalysis from water and
gen gas is negligible so that the reaction to ammonia cannot be nitrogen gas. Reproduced with permission from Ref. [184]. Copyright (2007) The
proceeded. Electrochemical Society.
3 3
CH4 þ N3 ! NH3 þ C þ 3e ð51Þ
4 4
gas generated from water reacts with nitrogen gas to produce
Murikami et al. [184] employed the same molten salt but with a ammonia in a highly conductive molten electrolyte with the iron
boron-doped diamond (BDD) anode in synthesis of ammonia from based catalyst. The highest ammonia production rate was 1
water and nitrogen gas. It was proposed that it is the oxygen 108 mol s1 cm2 at 200 °C with a current efficiency of 35%. It
formed on a BDD anode in molten LiCl–KCl containing 1 mol% was also noted that dispersion of nano-iron oxides in the elec-
Li2O at 450 °C that facilitate nearly 100% current efficiency. Nitro- trolyte was not favourable as these particles coagulated with time.
gen gas was reduced at the cathode to nitride ions, which reacts As a result, the ammonia production reduced to 85% of the aver-
with water to form ammonia and oxide ions. These oxide ions were aged rate when it run over 4 h. It is also interesting to observe that
then oxidised to oxygen gas at the BDD anode. the synthesis rate can be restored by re-dispersion of the nano-
3 3 sized iron catalyst. The same research group further developed
H2 O þ N3 ! NH3 þ O2 ð52Þ an approach to immobilize nano-particles in the solution [187].
2 2
Fe2O3 nanoparticles are immobilized on activated carbon which
3 2 3 are dispersed into molten NaOH-KOH mixture. Ammonia was suc-
O ! O2 þ 3e ð53Þ cessfully generated from N2 and water. The cell worked at 250 °C
2 4
with a voltage of 1.55 V generated ammonia at a rate of
An ammonia production rate of 5.80 109 mol s1 cm2 at 8.27 nm s1 cm2. The optimal Faradaic efficiency of 13.7% was
300 °C was achieved. The current efficiency reached 80% at the reached under lower potential bias of 1.15 V. Further mechanism
beginning of the experiment but decreased to about 10% at the is illustrated by the DFT calculation presented by Nguyen et al. [188].
end. It was explained that nitrogen gas evolved at low potentials
so that nitride ions accumulated continuously at the cathode. On 4.4. Solid electrolyte
the contrary, the concentration of the oxide ions constantly
decreased due to oxygen evolution. In other words, the decreased 4.4.1. O2 conducting materials
current efficiency for nitrogen evolution directly resulted in the Ammonia synthesis is not only possible with the use of proton-
reduction in the current efficiency for ammonia synthesis. The conducting electrolytes, but with the use of oxygen-conducting
experiment was also repeated using a glassy carbon rod as an alter- ceramics also. Although certain materials develop either proton
native electrode [185]. Although a relatively higher ammonia pro- or electronic conductivity in the presence of water, it is possible
duction rate of 2 108 mol s1 cm2 at 300 °C was reported, the to remain O2 conductors. Few reports were documented. Skodra
current efficiency was still as low as 23%. An illustration of the and Stoukides [189] used an 8 mol% yttria-stabilised-zirconia,
experimental set up with the BDD anode is shown in Fig. 22. Y2O3–ZrO2 (YSZ) electrolyte with gaseous reactants of nitrogen
Licht et al. [186] reported an ammonia synthesis route using and steam fed to a cathode and helium to a silver anode. The cath-
molten hydroxide as an electrolyte, a nano-Fe2O3 catalyst with ode was made of a Pd layered with an industrial Ru–MgO catalyst
reactants of air and steam respectively. The reduced hydrogen coated on top as the cathode. The cathode was designed to ensure
it was both electrically conductive and being favourable for ammo-
nia synthesis. Although Ru catalyst is very active in terms of syn-
thesising ammonia, it is not a good electrical conductor so that it
is not suitable for being used alone as cathode. Conversely, silver
is an excellent conductor and binds well on YSZ. Thus, it is advan-
tageous to combine the two materials. The third material, Pd, was
used, as it is known as a good catalyst for water dissociation.
At the cathode, the steam and nitrogen reacts to form ammonia
with oxygen ions being produced additionally:
3H2 O þ N2 þ 6e ! 3O2 þ 2NH3 ð54Þ

These oxygen ions were then transferred to the silver anode
where oxygen gas is produced:
Fig. 21. Steps involved in the suggested mechanism of electrocatalytic ammonia 3
synthesis. Reproduced with permission from Ref. [182]. Copyright (2005) The 2O2 ! O2 þ 6e ð55Þ
Electrochemical Society.
2
The overall cell reaction is then just the combination of the

above two equations:
3
3H2 O þ N2 ! 2NH3 þ O2 ð56Þ
2
The experiment was carried out in the ranging from 500 °C to
700 °C. The ammonia production is negligible low at 500 °C but
increased by nearly two orders of magnitude at higher tempera-
tures but decreased when the temperature reached 700 °C. The
maximum rate of ammonia formation obtained was approximately
3.8 1013 mol s1 cm2 at a temperature of 650 °C. The poor
electrical conductivity of the industrial Ru catalyst is responsible
for the low production rate. Another issue presented with the cath-
ode design was that the protons formed from the dissociation of
water did not reach the Ru surface but combined each other to
form hydrogen gas at the silver layer. In spite of the challenges, Fig. 23. Perovskite structure where dark blue atoms are A-site atoms, light grey
the authors emphasized that an advantage with the oxygen- atoms are B-site atoms and dark grey as O atoms. Reproduce from Ref. [199] with
permission of The Royal Society of Chemistry. (For interpretation of the references
conducting cell was that pure oxygen gas could be produced as a to colour in this figure legend, the reader is referred to the web version of this
side product with the added bonus that it is already separated from article.)
the mixture of reactants.
4.4.2. Proton conducting materials protons through the ceramic electrolyte. It was observed that if the
Proton-conducting ceramics have been investigated for poten- pressure of nitrogen was increased by as much as 6 times it had a
tial applications such as hydrogen separation membranes, solid negligible effect on the ammonia production rate. In order to inves-
oxide fuel cells, hydrogen pumps and gas sensors [190,191]. The tigate the decomposition of ammonia, ammonia and the nitrogen-
most common proton sources are water and hydrogen. A widely helium mixture were fed to the reactor. 20% or less of the ammonia
recognised theory for the mechanism of proton derived from water introduced to the reactor underwent decomposition, which indi-
conducting in ceramics can be described as follows: steam was dis- cates that the initial conversion of protons to ammonia may be
sociative adsorbed on oxygen vacancies of the host lattice to pro- complete. Unreacted hydrogen found at the cathode is a result of
duce a proton and a hydroxyl group. The oxygen in the hydroxyl ammonia decomposition. Additionally, the authors reported that
group fills in the position of VO and the proton further bonds with the non-Faradaic enhancement parameter or Faradaic efficiency,
the O atoms inside the lattice forming another hydroxyl group K, ranged between 1.0 and 2.0 indicating a weak non-Faradaic
which could be represented by following equation: Electrochemical Modification of Catalytic Activity (NEMCA). Vaye-
nas et al. [197] defined Faradaic efficiency as the following
H2 O þ V O þ Olattice ! 2OH ð57Þ equation:
The two hydroxyl ions actually substitutes the O ions in the lat- Dr
tice, therefore they can be considered as two positively charged K¼ ; ð59Þ
ðI=2FÞ
‘‘protonic defects” [190]. The proton in the hydroxyl group could
‘‘jump” to another vicinal O ion so as to form the protons transport where the Dr is the increase in rate of hydrogen consumption and
in the oxide electrolyte. (I/2F) is the imposed flux of protons through the electrolyte. The
Similar to H2O as the proton source, the hydrogen itself could NEMCA effect is described as the possibility to enhance the rate of
also act the proton source in a relatively straightforward way a reaction by ‘pumping’ ions to or from the catalytic surface acting
[192]: as one or both of the electrodes [198]. The discovery of a weak
NEMCA effect in the experiment by Marnellos et al. [194] indicates
H2 þ 2Olattice ! 2OH þ 2e ð58Þ
that not all of the hydrogen must be supplied electrochemically for
Panagos et al. [193] investigated the thermodynamic con- the formation of ammonia. If the Faradaic efficiency were equal to
straints imposed on the ammonia synthesis system. Hydrogen or less than one, this would suggest all the hydrogen had to be sup-
and nitrogen gases were used as the reactants with a ceramic elec- plied electrochemically in the form of protons.
trolyte of Yb doped Sr-cerate. In order to compare to the solid elec- Li et al. [200] reported on the synthesis of ammonia using two
trolyte cell reactor (SECR), a conventional catalytic reactor (CCR) ceramic electrolytes of BaCe0.9Sm0.1O3d (BCS) and BaCe0.8Gd0.1-
was used as a reference to produce ammonia without the elec- Sm0.1O3d (BCGS) respectively. The motivation of use of the mate-
trolyte system. It was reported that the SECR achieved ammonia rials was that doped ABO3 oxides exhibit high proton and electrical
synthesis rates of several orders greater than the CCR at certain conductivities as well as impressive chemical stability over a wide
operating conditions. Pressure as high as 100 bar would be range of temperatures. Additionally, the doped BaCeO3 has the
required for CCR to synthesis ammonia at the same rate that the highest ionic conductivity and therefore was chosen as use as the
SECR would produce at atmospheric pressure. electrolyte in this system. Nitrogen and hydrogen gases were used
Marnellos et al. [194] employed a perovskite-structured as the reactants with an Ag–Pd paste used as the catalyst for both
SrCe0.95Yb0.05O3 (SCY) as the electrolyte with Pd electrodes to gen- the cathode and anode. Although both ceramic samples showed
erate ammonia. Perovskites are oxides with the structure ABO3 high proton conductivity, the BCGS electrolyte was found to be
with an A-site atom at the centre and a B-site atom in octahedral more efficient as an electrocatalyst for ammonia synthesis. This
coordination with oxygen atoms as shown in Fig. 23 [195,196]. was due to the BCGS sample showing much greater catalytic activ-
Protons are formed from hydrogen gas at the anode and then trans- ity than BCS. The highest rate of ammonia production reported for
ferred to the cathode where protons react with nitrogen of a 1.8% BCGS was 5.82 109 mol s1 cm2 whereas for BCS it was 5.23
nitrogen-helium mixture. A maximum observed rate of ammonia 109 mol s1 cm2 both at a temperature of 620 °C. A study by the
formed was reported to be 3.1 109 mol s1 cm2 at 570 °C with same research group [201] used BaCe0.8Gd0.2O3d to synthesize
a current efficiency of 75%. The challenge was the transportation of ammonia from nitrogen and hydrogen with Ag–Pd alloys as the
electrodes. It had previously been reported that Gd-doped barium

cerate has the highest electrical conductivity out of all the doped
BaCeO3 materials [201] and so this was the inspiration behind this
electrolyte choice. The maximum observed rate of ammonia for-
mation was 3.09 109 mol s1 cm2 at 480 °C. The area of the
electrode used was 1.13 cm2, which they believed was not large
enough to convert all of the hydrogen gas to ammonia. It was rec-
ommended to increase the area of the electrode in order to not
only increase the hydrogen gas conversion but to ensure the con-
version is carried out quantitatively.
Another type of ceramic materials that have gained a lot of
attention recently are doped-CeO2 due to their high ionic conduc-
tivity at relatively low temperatures [202]. Liu et al. [203] exam- Fig. 25. Scheme of the partial unit cell of the pyrochlore structure. Reproduced
ined the suitability of the fluorite-structured Ce0.8M0.2O2d (M = from Ref. [212]. Copyright (2004) Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.
La, Y, Gd, Sm) as the electrolyte at an intermediate temperature
range of 400–800 °C. The fluorite structure (Fig. 24) consists of a Zr). Ag–Pd electrodes were once again used with dry air as reac-
cation within a cube where the vertices are of anions with each full tants, compared to humidified hydrogen gas and nitrogen gas as
cube surrounded by empty ones and each empty cube by full ones reactants. Although La1.95Ca0.05Zr2O7 has a lower conductivity, its
[204]. This represents a stoichiometry of the fluorite structure of ammonia production rate reached as high as 2 109 mol s1
AX2. Similar to previous studies, Ag–Pd electrodes were used with cm2, greater than the rate (1.3 109 mol s1 cm2) obtained
the reactants being humidified hydrogen gas at the anode and over La1.95Ca0.05Ce2O7. Oxygen conduction is responsible for the
nitrogen gas at the cathode. The ammonia synthesis rates obtained low synthesis rate over La1.95Ca0.05Ce2O7.
at an optimal temperature of 650 °C. These rates matched well All the previously published reports concerning ceramic elec-
with the proton conductivity and increases from La to Sm. trolytes employed Ag–Pd electrodes for ammonia synthesis but
Li et al. [206] carried out a further study where they tested three Ouzounidou et al. [198] who used an industrial Fe based catalyst
different ceramic electrolytes. The electrolytes were Ba3(Ca1.18- as the cathode and an Ag paste as the anode. Experiments were
Nb1.82)O9 (BCN18), Ba3CaZr0.5Nb1.5O9 (BCZN) and Ba3Ca0.9Nd0.28- carried out in a single-chamber reactor cell (SCRC) and a double-
Nb1.82O9 (BCNN). The same electrodes and gaseous reactants as chamber reactor cell (DCRC). The experimental set-ups are shown
the authors’ previous studies were used [189,200]. Although all in Fig. 26. In the SCRC, the incoming gaseous reactants flow
ceramic samples showed excellent proton conductivity, BCNN gave through both the anode and cathode simultaneously. On the con-
the greatest ammonia production rate of 2.16 109 mol s1 cm2 trary in the DCRC, the hydrogen gas is oxidised to protons in one
at 620 °C. Again, it was noted that an increase in the area of the chamber, with nitrogen gas producing ammonia from the resulting
electrode resulted in increase of hydrogen gas conversion rate. protons in a second chamber. The electrolyte used was SrZr0.95-
Xie et al. [207] investigated the use of a ceramic material as the Y0.05O3-a (SZY) with nitrogen gas supplied to the cathode and
electrolyte which has an alternative structure with a general for- hydrogen gas to the anode. The obtained results were correlated
mula of A2B2O7. The structures are known as pyrochlores with an to the Temkin–Pyzhev equation, which is the most commonly used
A ion coordinated to eight oxygen atoms and a B ion to six oxygen relationship for ammonia production [213]. It was found that at
ions, illustrated in Fig. 25 [208]. The pyrochlore structure is an low nitrogen pressure, the ammonia synthesis rates were linearly
ordered defective fluorite structure. The types of ceramic materials associated with nitrogen pressure, indicating the first term of Eq.
have already been extensively researched [209,210] which proton- (56) is a dominate term.
conductivity is similar to that of perovskites but higher at rela-
tively low temperatures [211]. A La1.9Ca0.1Zr2O6.95 electrolyte " #/ " 2 #1/
was used with Ag–Pd paste as the electrodes [206]. The maximum ðP3H2 Þ ðP NH3 Þ
observed ammonia synthesis rate was 1.76 109 mol s1 cm2 at r ¼ kA PN2 kB ð60Þ
ðP2NH3 Þ ðP3H2 Þ
520 °C with a current efficiency obtained of more than 80%. In a
similar study, Wang et al. [211] studied electrocatalytically pro- Similarly, effects of temperature are substantial. The reaction
ducing ammonia using ceramic electrolytes of La2M2O7 (M = Ce, rate levelled when the temperature reached a certain point, 500
°C in this case, and beyond the temperature the rates became
decreased. Ouzounidou et al. [198] suggested that the activation
energy of kB is greater than that of kA. At temperatures higher than
500 °C, the decomposition rates of ammonia became faster than its
formation rate so that the overall reaction rate decreases. The max-
imum observed ammonia synthesis rate was 6.50 10–12 mol
s1 cm2 but the K was less than 3.0 indicating a weak NEMCA
effect. In addition, it was also noticed that the synthesis of ammo-
nia was sensitive to reaction conditions. Higher than 550 °C, the
rate of ammonia production dropped to zero. A possible explana-
tion for this was that any oxygen present might poison the indus-
trial catalysts as it has been reported even at minimal oxygen
concentrations the poisoning can occur [214]. The catalyst may
also be poisoned by water vapour mostly formed by the reaction
of oxygen and hydrogen at the cathode. Its poisoning effect is pro-
Fig. 24. Scheme of crystal structure of the fluorite. The red sphere represents the X- portional to (PH2O/PH2)0.5 and so is much less significant than the
site atoms and light yellow spheres the A-site atoms. Reproduced from Ref. [205]
with permission of The Royal Society of Chemistry. (For interpretation of the
effect caused by oxygen gas.
references to colour in this figure legend, the reader is referred to the web version of Yiokari et al. [215] implemented a ceramic electrolyte of
this article.) CaIn0.1Zr0.9O3a at a reactor pressure of 50 bar and a Fe-based
Fig. 26. Schematic diagrams of (a) single-chamber reactor and (b) the dual-chamber reactor. Reproduced from Ref. [198], Copyright (2006), with permission from Elsevier.
catalyst. This was the first study that not only exhibited the effec- numerous studies using composite electrolytes. They used the Fe3-
tiveness of NEMCA in a high-pressure reactor, but demonstrated Mo3N catalyst with silver as cathode. With a LiAlO2–carbonate
electrochemical promotion. Ag was used as the anode whilst the composite as an electrolyte, experimental investigation was car-
industrial Fe catalyst served as the cathode with hydrogen and ried out at three temperatures of 400 °C, 425 °C and 450 °C respec-
nitrogen as gaseous reactants. It was reported that the initial H2/ tively. The rate of ammonia formation increases to a maximum of
N2 ratios were highly associated with the rates of ammonia synthe- 1.88 1010 mol s1 cm2 at 425 °C and then decreases as temper-
sis. The highest measured rate was of 4.9 109 mols1 cm2 was atures are further increases. The corresponding current efficiency
achieved at a H2/N2 ratio of 0.67, 680% greater than the rates at this temperature was 1.84%, which was attributed to the rapid
obtained at the open-circuit (when the current is zero). As the hydrogen evolution reaction (HER). Amar et al. also repeated this
H2/N2 ratio is greater than 1.5, ammonia production is limited by experiment using Co3Mo3N catalyst and found an increase in both
the supply of protons, and beyond 2.3 the rate is virtually zero. This ammonia production rate and current efficiency with values of
suggests that the catalyst surface is completely covered by 3.27 1010 mol s1 cm2 and 3.83% at 450 °C [221]. The same
adsorbed hydrogen, which prevents the adsorption of nitrogen. research group also tested the same reactants with samaria-
The rate enhancement ratio was defined by Eq. (57) [216]: doped-ceria (SDC)–(Li/Na/K)2CO3 as a composite electrolyte and
r La0.6Sr0.4Fe0.8Cu0.2O3d–Ce0.8Sm0.2O2d (LSFCu–SDC) as a composite
q¼ ð61Þ cathode [222]. The maximum rate of ammonia formation increased
r0
ten folds, achieving 5.39 109 mol s1 cm2 at 450 °C.
where r0 and r are the ammonia synthesis rates before and after Lan et al. [223] developed a low-cost perovskite catalyst com-
the current is applied. The largest q of 13 was obtained at 0.33 pared to previously used Ru/MgO and Pt/C catalysts, for direct con-
H2/N2 ratio 0.67 in which strongly non-faradaic behaviour was version of wet air or nitrogen to ammonia. The catalyst used for
identified. both anode and cathode was La0.8Cs0.2Fe0.8Ni0.2O3d with a com-
posite electrolyte of Ce0.8Gd0.2O2d–(Li/Na/K)2CO3. The reported
4.4.3. Composite electrolyte rate of ammonia formation was 9.21 1011 mol s1 cm2 for
A composite electrolyte is comprised of one or more ionic con- wet air and 1.23 1010 mol s1 cm2 for wet nitrogen. Both rates
ducting phases with a second or third phase added to alter its ther- were two orders of magnitude greater than those given in Stou-
mal, mechanical or electrical properties. For example, LiAlO2 or kides et al. [189].
Sm2O3 doped CeO2 has shown to have carbonate-ion, oxygen-ion Amar et al. [224] also successfully produced ammonia using a
and proton conductivity under specific conditions, such as in the La0.6Sr0.4Fe0.8Cu0.2O3d–Ce0.8Gd0.18Ca0.02O2d (LSFCu–CGDC) com-
presence of hydrogen [217]. As a result, composite materials have posite catalysts as the cathode, Ce0.8Gd0.18Ca0.02O2d–(Li/Na/
been suggested as potential electrolytes for fuel cells as well as for K)2CO3 composite electrolyte and Sm0.5Sr0.5CoO3d–Ce0.8Gd0.18-
ammonia synthesis under a range of operating temperatures. Ca0.02O2d composite anode. Similar to their previous reports,
Wang et al. [218] developed a new doped ceria–Ca3(PO4)2– wet nitrogen was directly converted to ammonia with a maximum
K3PO4 composite electrolyte for reduction of nitrogen by natural rate of ammonia formation of 5 1011 mol s1 cm2 at 400 °C.
gas to ammonia. The ammonia synthesis mechanism follows what The current efficiency was just 0.10% which was attributed to the
was proposed by Iwahara et al. [219]. Natural gas is converted to fact that the competing HER is the dominant process. In the
hydrogen that is then broken down to protons which react with Haber–Bosch process, iron-based catalysts have chances to be poi-
nitrogen gas to form ammonia. soned by high concentrations of oxygen. Due to the ionic nature of
the catalysts used in this study, oxidization of catalysts can be
2CH4 ! C2 H6 þ H2 ð62Þ
avoided. Oxide ions formed at the cathode are transported through
the CGDC to the electrolyte and then to the anode surface where
2CH4 ! C2 H4 þ 2H2 ð63Þ
they are oxidised to oxygen gas. This therefore indicates no oxygen
gas is formed at the cathode side so that the catalysts have no
C2 H6 ! C2 H4 þ H2 ð64Þ
chance to be poisoned. Thus the rate of ammonia synthesis is not
The optimum ammonia production rate was found to be 6.95 expected to reduce with reaction time.
109 mol s1 cm2 at 650 °C. The experiment was tested over a A study carried out by Lan et al. [225] introduced BaO, a typical
temperature range of 400 °C–800 °C. At 650 °C, the maximum rate promoter of ammonia synthesis, to a doped SmFeO3 [80]. However,
of ammonia was identified. Amar et al. [220] have conducted the catalyst is unstable in wet air due to the reaction with water.
For that reason, barium was added to the A-site of the perovskite the photocatalytic process, organic hole scavengers are used and
oxide SmFeO3 which formed the catalyst Sm0.6Ba0.4Fe0.8Cu0.2O3d sometimes indispensable in many reported works (as shown in
to be used as both the cathode and anode. The calculated rates of Table 1). Indeed, organic hole scavengers could effectively improve
ammonia formation were 9.19 107 mol s1 cm2 and 1.53 the quantum efficiency and ammonia production rate. In Li et al.’s
1010 mol s1 cm2 for wet air and wet nitrogen respectively. The work [75] the very high apparent quantum efficiency of 21.5% was
authors noted that although rates of this magnitude are not great achieved using methanol as the solvent. They also pointed out that
enough for commercial use, the materials of use are of very low methanol acts as the proton sources as well. The drawbacks of
cost and if improved upon could produce ammonia on a large scale. using organic solvents are quite apparent. The extra costs and
the oxidation of the solvents, usually along with CO2 emissions,
disappoint the energy efficiency and environmental benignity of
5. Summary and perspective
methods [239]. How to maintain the good performance without
using any organic hole scavengers used to be and is still one of
Ammonia synthesis via the Haber–Bosch process has been
the main challenges in the photocatalysis. Future research aiming
developed for more than 100 years. However, its high energy con-
to eliminate organic hole scavengers is highly in demand in order
sumption, CO2 emission and requirement of huge amount of H2 can
to achieve a commercial-scaled production. Researches on the
be hardly overcome. It is the right time to develop new efficient
mechanism and economic feasibility are suggested to focus on in
approaches and catalysts towards the sustainable ammonia pro-
photocatalysis.
duction at low temperature and ambient pressure. In this review,
we have summarized the latest electron-driven ammonia produc-
5.1.3. Mechanism
tion research achievements as well as the remaining great chal-
Electron-driven reaction mechanism for ammonia synthesis is
lenges. The electrons generated by the materials under excitation
incredibly complex. The mechanism studies of electrocatalysts
or provided by the external sources can directly involve in the
heavily rely on the ground-state DFT calculation. However, much
proton-coupled electron transfer process of ammonia synthesis.
fewer calculation works aim to elucidate the excitation states
It is suggested that the electron density and ability of back donat-
properties. Methods such as MBPT (Many-body Perturbation The-
ing to the N2 is crucial in the adsorption and activation of N2
ory), TDDFT (Time-Domain Density Functional Theory) and corre-
[36,234–236]. The extra electrons received or generated will pro-
sponding NAMD (nonadiabatic molecular dynamics) are much
mote the reactions significantly. In terms of catalysis, the products
less used than those on the ground states. In case of photocatalysis,
yield and selectivity are the two main factors. Several concluding
the understanding of excitation states are indispensable supple-
remarks and suggestions are presented in the following section.
ments to the ground state calculation, which will facilitate us to
comprehend the sophisticated photocatalytic process [240–243].
5.1. Ammonia production rate Besides calculations, the in-situ spectroscopy technique such as
FTIR and EXAFS has been proven as excellent tools illustrating
5.1.1. Catalyst design the detailed mechanism. It is demanded more experimental mech-
The traditional methods in improving the photocatalyst or elec- anism studies could offer theoretical guidance on robust catalysts
trocatalyst activities are still effective in promoting the ammonia design.
synthesis from atmospheric N2. The approaches such as co-
catalysts, heterostructure design, defect engineering and electronic 5.1.4. Feedstock
structure modification, have been proven to be successful in pro- Instead of the conventional feedstock for ammonia synthesis -
moting the ammonia production activity and selectivity. The LSPR nitrogen and hydrogen gases, researches have been carried out to
metal has also demonstrated its significant capacity in driving use air as a feedstock to replace pure N2 in both photocatalytic
ammonia synthesis [153]. Although the LSPR phenomena pro- and electrocatalytic system. It has been dominantly stated that
motes the N2 fixation, it still suffers from the demand of high inci- the replacement by air can place a negative impact on the produc-
dent energy density and low reaction rate. A laser light source with tion rate of ammonia, while there are still a few reported cases that
the power density higher than 100 mW cm2 is often indispens- observed little effects of air to the reaction [17,174]. Hydrogen gas
able for a reliable and reasonable yield rate. Dedicated works on is traditionally produced via natural gas reforming with steam,
developing more active LSPR photocatalysis are highly demanded. which relies on fossil fuels. For the sake of sustainability, water
has been become attractive alternative proton source to hydrogen
5.1.2. Hole scavenger as it is economic and environmental friendly. The role of O2 is gen-
It is promising and economically feasible to use water or aque- eral believed as the electron and proton scavenger both in the pho-
ous solution system for ammonia synthesis. Low solubility of N2 in tocatalytic and electrocatalytic process through the oxygen
water presents a fundamental limitation for the catalytic process. reduction reaction (ORR) [177]:
Researchers have tried to use organics as the additive or solvent
1
to address this issue [227,228]. Different from electrocatalysts, 2Hþ þ O2 þ 2e ! H2 O ð65Þ
photocatalysts generate electrons via light irradiation which will 2
certainly accompany with the formation of holes as well. Holes However, oxygen gas may oxidize the produced ammonia
are anticipated to oxidize water or hydroxide anion to form O2. resulting in a lower production rate. The detrimental effect of oxy-
The OER reaction is crucial for the ammonia synthesis in the aque- gen gas could be worsen when a single-chamber cell was applied.
ous solution, especially alkaline, because it determines proton sup-
ply in the solution [237,238]. While the holes oxidize the 5.2. Selectivity and efficiency
ammonium cations to nitrate anions (a thermodynamically
favoured process), which will further cause the tricky selectivity Selectivity is a very important parameter in a catalytic reaction.
issue even worse. Since the OER reaction driven by the holes is In electrocatalytic processes, it can be assessed by Faradaic effi-
quite slow, there is a high possibility that the holes induced by ciency. Apparent quantum efficiency in photocatalytic processes
the photon accumulated in the semiconductor, which will largely partially represents the selectivity, while it is more appropriate
increase the recombination rates of electrons and holes. To avoid in measuring the capability of the photocatalysts in absorption
the main reason that has been blamed for the low efficiency of and conversion of the solar light energy.
Table 1
24
The summary of recent works on photocatalytic ammonia synthesis from N2 and their corresponding performances.
Light source Catalyst Reaction medium Temperature Products and yields Apparent Quantum Ref
efficiency
500 W Xe-lamp, 100 mW cm2 Fe2O3 789 mg L1 ethanol aqueous solution 298 K 2250 lmol L1 g1 h1 N/A [132]
200 W Hg lamp FeTiO3 Ethanol (75 vol%) aqueous solution Room NH+4 17 lmol/90 min N/A [17]
temperature
450 W Hg/Xe lamp, with IR cutoff Diamond In dual-compartment H-cell. Cathodic Room NH+4 6 lg/20 h N/A [58]
filter cell: pure water; anodic cell: Pt as the Temperature
anode with KI aqueous solution as the
hole scavenger
500 W Xenon lamp Amine functionalized ultrathin 4 wt.% methanol aqueous solution Room NH+4 60.5 lmol h1 N/A [137]
C3N4 Temperature
1 1
300 W Xenon lamp with AM 1.5G KOH modified C3N4 Methanol 25 °C NH3 3.632 mmol g h 21.5% @ 420 nm [75]
filter
500 W Xenon lamp with 420 Phosphate modified LaFeO3 Water N/A NH3 250 lmol g1 h1; N/A [138]
longpass filter N2H4 10.3 lmol g1 h1
150 W Xenon lamp, 100 mW cm2 FeMoS–FeS–SnS chalcogel 5 mM ascorbic acid (electron donor), 50 Ambient TON 17 in 96 h N/A [139]
mM pyridinium hydrochloride (proton temperature
donor) in 10 mL aqueous solution
K. Wang et al. / Carbon Resources Conversion 1 (2018) 2–31

150 W Xenon lamp, 100 mW cm2 Mo2Fe6S8–Sn2S 5 mM ascorbic acid (electron donor), 50 Ambient TON 8 in 72 h N/A [140]
chalcogel mM pyridinium hydrochloride (proton temperature
donor) in 10 mL aqueous solution
250 W sodium lamp Zn0.1Sn0.1Cd0.8S with S vacancies Ethanol (0.789 g L1) aqueous solution 30 °C NH+4 4.75 mg L1 h1 g1 N/A [114]
250 W sodium lamp Mo0.1Ni0.1Cd0.8S with s vacancies Ethanol (0.789 g L1) aqueous solution 30 °C NH+4 3.2 mg L1 h1 g1 N/A [115]
250 W sodium lamp Microwave treated C3N4 with N Ethanol (0.789 g L1) aqueous solution 30 °C NH+4 3.01 mg L1 h1 g1 N/A [107]
vacancies
250 W sodium lamp HCl treated C3N4 with N Ethanol (0.789 g L1) aqueous solution 30 °C NH+4 6.32 mg L1 h1 g1 N/A [108]
vacancies
250 W sodium lamp Methanol treated C3N4 sponge EDTA–2Na aqueous solution 30 °C NH+4 6.0 mg L1 h1 g1 N/A [109]
with N vacancies
250 W sodium lamp Fe3+ doped C3N4 Ethanol 298 K NH+4 5.40 mg L1 h1 g1 N/A [122]
250 W sodium lamp Cu(I) doped C3N4 Ethanol (0.789 g L1) aqueous solution N/A NH+4 8.8 mg L1 h1 g1 N/A [124]
500 W Xenon lamp with 420 nm O, S doped C3N4 Methanol aqueous solution 25 °C NH+4 2.18 mmol g1 h1 N/A [127]
UV cutoff filter
250 W Xenon lamp Fe/TiO2; Ru/TiO2; Os/TiO2 Aqueous solution, with ascorbic acid as 25 °C NH3: N/A [141]
hole scavenger (balanced with NaOH, pH Ru/TiO2 3.2 lmol L1/15 min;
= 5.8) Fe/TiO2 2.6 lmol L1/15 min;
Os/TiO2 0.7 lmol L1/15 min
150 W Xenon lamp Ru/TiO2; Rh/TiO2; Pd/TiO2; Pt/ Methanol aqueous solution 5% v/v N/A NH3 N/A [142]
TiO2 Ru/TiO2 17.3 lmol g1 h1; Rh/TiO2
12.6 lmol g1 h1; Pd/TiO2
11.8 lmol g1 h1; Pt/TiO2 4.8 lmol g1 h1
450 Hg Lamp Pt/ZnO Na–EDTA aqueous solution Room 860 lmol g1 h1 N/A [143]
temperature
500 W Xenon lamp, 420 nm cutoff Black phosphorus/C3N4 5 vol% methanol aqueous solution N/A NH3 347.5 lmol L1 h1 N/A [150]
filter
1 1
500 W Xenon lamp Carbon nanocrystal/WO3H2O Water 25 °C NH+4 205 lmol g h N/A [151]
300 W Xenon lamp, Ru cluster/GaN Gaseous H2/N2 v/v 3:1 10 °C NH3 120 lmol g1 h1 N/A [144]
500 W Hg Lamp Al–Fe/3D graphene Gaseous H2/N2 v/v 3:1, 20 mL min1 200 °C NH3 430 lg g1 h1 N/A [148]
500 W Hg Lamp Fe/3D Graphene Gaseous H2/N2 v/v 3:1, 20 mL min1 200 °C NH3 24 lmol g1 h1 N/A [146]
500 W Xenon lamp Ga2O3–DBD/g-C3N4 0.04 mM methanol aqueous solution N/A NH3 112.5 lmol L h1; N/A [136]
300 W Xenon lamp, 420 nm cutoff TiO2@C/g-C3N4 20 vol% methanol aqueous solution N/A NH3 250.6 lmol L h1 0.14% @ 420 nm [130]
filter
250 W sodium lamp g-C3N4/ZnMoCdS Ethanol (0.789 g L1) aqueous solution 298 K NH3 2.5 mg L1 h1 g1 N/A [131]
300 W Xenon lamp W18O49/C3N4 Ethanol (0.789 g L1) aqueous solution N/A NH+4 2.6 mg L1 h1 g1 N/A [134]
250 W sodium lamp Mg1.1Al0.3Fe0.2O1.7/C3N4 Ethanol (0.789 g L1) aqueous solution 30 °C NH+4 7.5 mg L1 h1 g1 N/A [135]
300 W Xenon lamp BiOBr Water 25 °C NH3 223.3 lmol h1 g1; 0.23% @ 420 nm [110]
2.4% for N2H4 and 1.9% for NO
3
5.2.1. Cell design and configuration
[156]
[125]
[149]
[154]
[157]
[112]
[155]
[158]
[101]
[105]
[102]
[104]
As presented in Table 2 one may notice that the Faradaic effi-
Ref
ciency ranges from 0.01% to 10% when aqueous solution was used.
If the ionic liquids, molten salts and solid electrolyte are used, the
@ 550 nm
@ 500 nm
corresponding Faradaic efficiency could reach one order of magni-
Apparent Quantum
@ 254
@ 425
0.44% @ 380 nm;
0.73% @ 500 nm;
tude higher. In some cases, it can achieve as high as 80% even H2O
4.32% @ 420 nm
0.25% @ 600 nm
5.1% @ 365 nm
2.3% @ 420 nm
0.1% @ 500 nm
1% @ 600 nm
rather than H2 being used as the proton provider [185]. A 90.4%
5 103%
1 103%
3 103%
3 103%
efficiency
Faradaic efficiency has been reported by Xu et al. [178] using

Nafion as the electrolyte, H2 as the proton source and the ammonia
nm;
nm;
N/A
N/A
N/A
N/A
production took place at 80 °C. The competitive HER reaction low-
ers the selectivity. If the cathode chamber is designed to be gas-
NH3 6.5 nmol h1 cm2 (ethanol as hole scavenger); phase that only needs a small amount of moisture, it could largely
suppress the HER process.
3 (0.3 lm L h
1
)
NH3 184.8 lmol L1 with 100% selectivity
5.2.2. Catalysts design and optimization

Ultrasmall metal clusters especially Au have great potentials in
0.73 nmol h1 cm2 (no scavenger)
promoting the selectivity towards ammonia synthesis. The reason

NH3 (111.5 lmol L h1); NO
could partially be due to the low activity of Au in the HER process.

Due to strong metal-support interactions (SMSI) [244,245], the
NH3 13.4 nmol cm2 h1
NH3 1226 lmol g1 h1
NH3 2685 lmol g1 h1

NH3 1.38 mmol h1 g1
NH3 38.2 lmol g1 h1
oxide support such as titania and ceria can function effectively

NH3 325 lmol g1 h1
NH3 13.3 mg m2 h1

Products and yields
for stabilizing the ultrafine metal clusters by forming half covering

1.3 mmol g1 h1
layering, which further ensures the excellent long-term stability of

NH3H2O; NO
3
the electrocatalysts. And more importantly, the SMSI could also

retard the H2 chemisorption. The ultrasmall cluster strategy was
confirmed with the DFT computational results by Howalt and
Vegge et al. [63,68]. Their calculations on the cube-octahedral
Mo13 nanocluster and N covered Mo13 prove that N atoms are
Temperature
temperature
temperature
temperature
temperature
adsorbed with priority over H atoms on the surface of the Mo nan-

ocluster via associative mechanism. The preferential adsorption of
Ambient
Ambient
Room
Room
333 K
N is the basis to achieve higher ammonia selectivity. Further com-

25 °C
25 °C
25 °C
20 °C
20 °C
N/A
N/A
prehensive understanding on the physical and chemical properties

of clusters remains ambiguous and needs more investigations.
Anodic chamber: 10 vol% alkaline aqueous
Another interesting strategy comes from the suggestion that

solution, pH = 13; cathodic chamber:
simultaneous 6-electron transfer will not produce any by-

Flowing H2/N2 (15 ml min1: 45 ml
150 ppm sodium sulphite aqueous
products but directly produce ammonia, resulted in high selectiv-

ity [102].
5.2.3. Potential bias

acidic solution pH = 2
Potential bias for electrocatalytic process is another key factor

Reaction medium
affecting the selectivity. In most of the practical experiments, the

potentials for electrochemical cell has an optimal value since
higher potential bias usually causes the drop in the Faradaic effi-
solution
min1)
Water
Water
Water
Water
Water
Water
Water
Water
Water
ciency [181]. To date it is widely accepted that the HER activity

increases much quicker than the NRR under higher voltage bias.
Reduction of nitrogen to ammonia is a proton-coupled electron
CuCr-layered-double-hydroxide
transfer process, which is positively affected by the external poten-

tial bias. Therefore with preferred N atom adsorption, the large-
TiO2/Au/amorphous TiO2
scale electron-driven ammonia synthesis with high selectivity will

become realized.
Au-NPs/Nb-SrTiO3
Au/p-type Si NWs
Bi5O7Br nanotube
BiO quantum dot
Ni2P/Cd0.5Zn0.5S
Ultrathin MoS2
5.2.4. Control water or proton supply for the cathode

Au/(BiO)2CO3
Au-Os/Cs2O
H-Bi2MoO6
Sheets et al. [64] proposed to use the anion conducting mem-

Catalyst
brane to limit water diffusion and to further suppress the HER.

Bi5O7I
Nash et al. [181] compared the noble metal as catalysts in the

membrane electrode assembly configuration using both proton
300 W Xenon lamp, 420 nm cutoff
and hydroxide-conducting membrane as the electrolyte. It is

300 W Xenon lamp, AM 1.5 filter
300 W Xenon lamp with 400 nm
300 W Xenon lamp, with 2 sun
Simulated light, 200 mW cm2,
proved that alkaline environment favours the Faradaic efficiency

than acidic environment. However, the ammonia production rate
Xenon lamp, 550–800 nm
in the alkaline environment is around one order of magnitude

450 nm cutoff filter
lower than the acidic one. Alternative way to control the proton
300 W Xenon lamp
500 W Xenon lamp
500 W Xenon lamp

300 W Xenon lamp
300 W Xenon lamp

300 W Xenon lamp
(100 mW cm2)
supply is to use organic solvents or adding organics to the aqueous

Table 1 (continued)
cutoff filter
electrolyte. It has been proposed by several research groups that

Light source
intensity
the addition of organic compounds can actually increase both the

filter
ammonia production rate and Faradaic efficiency [227,228,246].

It is attributed to the higher solubility of N2 in organic solvent
(usually an order of magnitude higher), a low supply of proton
Table 2
26
The summary of the electrocatalysts used in electrochemical ammonia production via N2 fixation and their corresponding performances.
Temperature (°C) Reactants Cathode Electrolyte Anode Products and yield Faradaic efficiency (%) Ref
1 2 (0.96 V vs Ag/AgCl)
90 N2 H2O Ru/C Gas phase/KOH aqueous solution Pt NH3 1.3 lg h cm 0.24 [70]
Room temperature N2 Tetrahexahedral 0.1 M KOH aqueous solution Graphite NH3 1.648 lg h1 cm2; N2H4H2O 0.102 lg h1 cm2 3.879 [64]
Au/carbon paper (0.2 V vs RHE)
20 N2 H2O Fe2O3/CNT Gas phase/KHCO3 aqueous Pt NH3 2.18 mg m2 h1 (2.0 V vs Ag/AgCl) 0.025 for NH3 only; [172]
solution 95.1 for NH3 + H2
20 N2 H2O Fe2O3/CNT Gas phase/0.5 M KOH aqueous Pt NH3 1.06 102 nmol cm2 s1 (2.0 V vs Ag/AgCl) 0.164 for NH3 [173]
solution
90 N2 H2O MOF (Fe) Gas phase/2 M KOH aqueous Pt wire NH3 2.12 10 nmol s1 cm2 (1.2 V vs Ag/AgCl) 1.43 [226]
solution
Room temperature N2 H2O N-doped porous carbon 0.05 M H2SO4 aqueous solution Pt foil NH3 1.40 mmol g1 h1 (0.9 V vs RHE) 1.42 [174]
Room temperature N2 H2O Au/TiO2 0.1 M HCl aqueous solution Pt plate NH3 21.4 lg h1 mg1 (0.2 V vs RHE) 8.11 [160]
Ambient temperature N2 H2O Rhodium nanosheets 0.1 M KOH aqueous solution Carbon rod NH3 3.82 lg h1 cm2 or 23.88 lg h1 mg1 (0.2 V vs RHE) 0.217 [166]
Room temperature N2 H2O Amorphous Au/CeOx– 0.1 M HCl aqueous solution Pt foil NH3 8.3 lg h1 cm2 or 21.4 lg h1 mg1 (0.2 V vs RHE) 10.10 [161]
RGO
Room temperature N2 H2O Mo nanofilm 0.01 M H2SO4 Pt plate NH3 0.0309 nmol s1 cm2 0.72% [168]
(0.49 V vs RHE) (0.29 V vs RHE)
Ambient temperature N2 H2O Porous Ni 10 mM H2SO4 aqueous solution/ Pt NH3 1.54 102 nmol s1 cm2 0.89 [227]
K. Wang et al. / Carbon Resources Conversion 1 (2018) 2–31

2-propanol (9:1 v/v)
1
25 N2 H2O Ni plate Ethylenediamine–LiCl (cathodic Glass carbon NH3 0.0358 nmol s c (1.8 V cell voltage) 17.2 [228]
cell)/0.05 M H2SO4 aqueous
solution (anodic cell)
50 Ethanol– Ni/Cu/Fe LiClO4 in THF Pt rNH3 4.90 nmol s1 cm2 58 [229]
N2
0–80 H2O–N2 Pt coated with Ppy film Li2SO4 with H+ solution Pt rNH3 0.368 nmol s1 cm2 - [230]
80 N2, H2 Pt/C, Ir/C, Pd/C, Ru/C, Proton exchange membrane Pt/C Highest ammonia production rate in both PEM and HEM is Ir/C i: Ru/C: 0.053 (PEM); Au/ [181]
Au/C (PEM) and hydroxide exchange 2.09 103 nmol cm2ECSA s
1
; 0.43 103 nmol cm2
ECSA s
1
C: 0.55 (HEM) (0.2 V
membrane (HEM) respectively (0.2 V cell voltage) cell voltage)
65 N2 H2O Fe2O3/carbon paper Anion exchange membrane IrO2 NH3 55.9 nmol h1cm2 (cell voltage 1.6 V) 0.044 [179]
(FAA-3, Fumatech)
2 1
30–60, 20 psi N2 H2 Pt/C or Ir/C Gel polymer electrolyte Pt/C or Ir/C NH3 0.04 nmol cm s (Pt/C) 0.01 (Pt/C) [180]
(polyacrylic acid homopolymer NH3 0.02 nmol cm2 s1 (Ir/C)Cell voltage 0.25 V, 0.5 V 0.1 (Ir/C)
and KOH)
80 H2–N2 SmFe0.7Cu0.1Ni0.2O3 Nafion NiO–Ce0.8Sm0.2O2d rNH3 11.3 nmol s1 cm2 (cell voltage 2 V) 90.4 [178]
(SCFN) (Ni–SDC)
20–100 H2–N2 Sm1.5Sr0.5NiO4 Nafion NiO–Ce0.8Sm0.2O2d rNH3 10.5 nmol s1 cm2 – [231]
(Ni–SDC)
1 2
20–100 H2–N2 Sm1.5Sr0.5NiO4 SPSF NiO–Ce0.8Sm0.2O2d rNH3 10.3 nmol s cm – [232]
(Ni-SDC)
1 2
400 H2O-N2 Al LiCl–KCl–CsCl with Li3N Porous Ni rNH3 33 nmol s cm 72 [233]
300 H2O–N2 Porous Ni LiCl–KCl–CsCl with Li3N BDD rNH3 5.80 nmol s1 cm2 80 [185]
200 N2–H2O Ni NaOH/KOH with nano- Ni rNH3 10 nmol s1 cm2 35 [186]
Fe2O3
250 N2 H2O Stainless steel NaOH–KOH molten salt with Ni NH3 8.27 nmol s1 cm2 (cell voltage 1.55 V) 13.7 (at 1.15 V) [187]
dispersed Fe2O3/activated
Carbon
500–700 He–N2/ Ru–MgO with Pd Y2O3–ZrO2 Ag rNH3 3.80 104 nmol s1 cm2 – [189]
H2O
1 2
570–750 H2–N2 Pd SrCe0.95Yb0.05O3 (SCY) Pd rNH3 3.10 nmol s cm 75 [194]
550–680 H2–N2 Ag–Pd Ba3Ca0.9Nd0.28Nb1.82O9 (BCNN) Ag–Pd rNH3 2.16 nmol s1 cm2 – [206]
460–560 H2–N2 Ag–Pd La1.9Ca0.1Zr2O6.95 Ag–Pd rNH3 1.76 nmol s1 cm2 80 [207]
450–600 H2–N2 Fe SrZr0.9Y0.1O3a (SZY) Ag rNH3 6.50 103 nmol s1 cm2 <1 [198]
440 H2–N2 Fe CaIn0.1Zr0.9O3a Ag rNH3 4.90 nmol s1 cm2 – [215]
400–450 H2–N2 Fe3Mo3N–Ag LiAlO2-carbonate Ag–Pd rNH3 0.188 nmol s1 cm2 1.84 [220]
400–450 H2–N2 Co3Mo3N–Ag LiAlO2-carbonate Ag–Pd rNH3 0.327 nmol s1 cm2 3.83 [221]
400–450 N2–H2O La0.8Cs0.2Fe0.8Ni0.2O3d Ce0.8Gd0.2O2d–(Li/Na/K)2CO3 La0.8Cs0.2Fe0.8Ni0.2O3d rNH3 0.123 nmol s1 cm2 0.55 [223]
375–450 N2/H2O– La0.6Sr0.4Fe0.8Cu0.2O3d– Ce0.8Gd0.18Ca0.02O2a–(Li/Na/ Sm0.5Sr0.5CoO3d– rNH3 0.05 nmol s1 cm2 0.39 [224]
Air Ce0.8Gd0.18Ca0.02O2a K)2CO3 Ce0.8Gd0.18Ca0.02O2a
by organic compounds and a thermodynamic preference to ammo- [21] N. Cherkasov, A.O. Ibhadon, P. Fitzpatrick, A review of the existing and
alternative methods for greener nitrogen fixation, Chem. Eng. Process. 90
nia rather than H2 in the organic electrolyte. It is worthwhile men-
(2015) 24–33.
tioning that, as water is used as the proton source, the following [22] P. Lanzafame, S. Abate, C. Ampelli, C. Genovese, R. Passalacqua, G. Centi, S.
side reaction might also take place along with the desired reaction Perathoner, Beyond solar fuels: renewable energy-driven chemistry,
with ammonia as the only product (Eq. (13)) and might cause a ChemSusChem 10 (2017) 4409–4419.
[23] J. Hong, S. Prawer, A.B. Murphy, Plasma catalysis as an alternative route for
selectivity issue: ammonia production: status, mechanisms, and prospects for progress, ACS
Sustainable Chem. Eng. 6 (2017) 15–31.
6 6 4
N2 ðgÞ þ H2 O ðlÞ ! NO ðgÞ þ NH3 ðgÞ ð66Þ [24] V. Kyriakou, I. Garagounis, E. Vasileiou, A. Vourros, M. Stoukides, Progress in
5 5 5 the Electrochemical Synthesis of Ammonia, Catal. Today 286 (2017) 2–13.
[25] M.A. Shipman, M.D. Symes, Recent progress towards the electrosynthesis of
According to the thermodynamic potential calculation results, ammonia from sustainable resources, Catal. Today 286 (2017) 57–68.
the DG at standard condition of reaction (66) is 3.90 eV with 4 [26] S. Giddey, S.P.S. Badwal, A. Kulkarni, Review of electrochemical ammonia
electron-proton pairs transfer involved in the reaction; the reac- production technologies and materials, Int. J. Hydrogen Energy 38 (2013)
14,576–14,594.
tion (13) requires the DG of 7.03 eV and 6 electron-proton pairs [27] N. Cao, G. Zheng, Aqueous electrocatalytic N2 reduction under ambient
transfer in total. The overall Gibbs energy is quite high though conditions, Nano Res. 11 (2018) 2992–3008.
the energy required for each electron transferred (potential) is [28] I. Garagounis, V. Kyriakou, A. Skodra, E. Vasileiou, M. Stoukides,
Electrochemical synthesis of ammonia in solid electrolyte cells, Front.
comparatively small: 0.975 V and 1.17 V for reaction (66) and Energy Res. 2 (2014).
(13) respectively [30]. [29] X.Z. Chen, N. Li, Z.Z. Kong, W.J. Ong, X.J. Zhao, Photocatalytic fixation of
nitrogen to ammonia: state-of-the-art advancements and future prospects,
Mater. Horiz. 5 (2018) 9–27.
Acknowledgements [30] A.J. Medford, M.C. Hatzell, Photon-driven nitrogen fixation: current progress,
thermodynamic considerations, and future outlook, ACS Catal. 7 (2017)
The authors are grateful for the financial support by the Univer- 2624–2643.
[31] C.X. Guo, J.R. Ran, A. Vasileff, S.Z. Qiao, Rational design of electrocatalysts and
sity of Edinburgh. photo(electro) catalysts for nitrogen reduction to ammonia (NH3) under
ambient conditions, Energy Environ. Sci. 11 (2018) 45–56.
References [32] G. Ertl, Science and catalysis—studies on the mechanism of ammonia
synthesis: the P. H. Emmett award address, Catal. Rev. 21 (2006) 201–223.
[33] G. Ertl, Primary Steps in Catalytic Synthesis of Ammonia, J. Vac. Sci. Technol. A
[1] B.M. Hoffman, D. Lukoyanov, Z.Y. Yang, D.R. Dean, L.C. Seefeldt, Mechanism of
1 (1983) 1247–1253.
nitrogen fixation by nitrogenase: the next stage, Chem. Rev. 114 (2014) 4041–
[34] G. Ertl, M. Huber, N. Thiele, Formation and Decomposition of Nitrides on Iron
4062.
Surfaces, Z Naturforsch A 34 (1979) 30–39.
[2] R.F. Service, Chemistry. New recipe produces ammonia from air, water, and
[35] K.I. Aika, Role of alkali promoter in ammonia synthesis over ruthenium
sunlight, Science 345 (2014) 610.
catalysts—Effect on reaction mechanism, Catal. Today 286 (2017) 14–20.
[3] J.W. Erisman, M.A. Sutton, J. Galloway, Z. Klimont, W. Winiwarter, How a
[36] M. Kitano, Y. Inoue, Y. Yamazaki, F. Hayashi, S. Kanbara, S. Matsuishi, T.
century of ammonia synthesis changed the world, Nat. Geosci. 1 (2008) 636–
Yokoyama, S.W. Kim, M. Hara, H. Hosono, Ammonia synthesis using a stable
639.
electride as an electron donor and reversible hydrogen store, Nat. Chem. 4
[4] H.Z. Liu, Ammonia synthesis catalyst 100 years: practice, enlightenment and
(2012) 934–940.
challenge, Chin. J. Catal. 35 (2014) 1619–1640.
[37] C.J. Jacobsen, S. Dahl, B.S. Clausen, S. Bahn, A. Logadottir, J.K. Nørskov, Catalyst
[5] R. Schlgl, Ammonia synthesis, Handbook of Heterogeneous Catalysis, 2008.
design by interpolation in the periodic table: bimetallic ammonia synthesis
[6] B. Hinnemann, J.K. Nørskov, Catalysis by enzymes: the biological ammonia
catalysts, J. Am. Chem. Soc. 123 (2001) 8404–8405.
synthesis, Top. Catal. 37 (2006) 55–70.
[38] J.K. Nørskov, P. Stoltze, Theoretical aspects of surface reactions, Surf. Sci. 189–
[7] M.A. Legare, G. Belanger-Chabot, R.D. Dewhurst, E. Welz, I. Krummenacher, B.
190 (1987) 91–105.
Engels, H. Braunschweig, Nitrogen fixation and reduction at boron, Science
[39] B. Hammer, J.K. Nørskov, Why gold Is the noblest of all the metals, Nature 376
359 (2018) 896–900.
(1995) 238–240.
[8] H. Tanaka, Y. Nishibayashi, K. Yoshizawa, Interplay between theory and
[40] A. Logadottir, T.H. Rod, J.K. Nørskov, B. Hammer, S. Dahl, C.J.H. Jacobsen, The
experiment for ammonia synthesis catalyzed by transition metal complexes,
Brønsted–Evans–Polanyi relation and the volcano plot for ammonia synthesis
Acc. Chem. Res. 49 (2016) 987–995.
over transition metal catalysts, J. Catal. 197 (2001) 229–231.
[9] I. Pappas, P.J. Chirik, Ammonia synthesis by hydrogenolysis of titanium-
[41] A. Hellman, E.J. Baerends, M. Biczysko, T. Bligaard, C.H. Christensen, D.C.
nitrogen bonds using proton coupled electron transfer, J. Am. Chem. Soc. 137
Clary, S. Dahl, R. van Harrevelt, K. Honkala, H. Jonsson, G.J. Kroes, M. Luppi, U.
(2015) 3498–3501.
Manthe, J.K. Nørskov, R.A. Olsen, J. Rossmeisl, E. Skulason, C.S. Tautermann, A.
[10] R. Lan, S. Tao, Ammonia as a suitable fuel for fuel cells, Front. Energy Res. 2
J. Varandas, J.K. Vincent, Predicting catalysis: understanding ammonia
(2014).
synthesis from first-principles calculations, J. Phys. Chem. B 110 (2006)
[11] A. Afif, N. Radenahmad, Q. Cheok, S. Shams, J.H. Kim, A.K. Azad, Ammonia-fed
17,719–17,735.
fuel cells: a comprehensive review, Renewable Sustainable Energy Rev. 60
[42] A. Ozaki, K. Aika, Catalysis-science and technology, Berlin, 1981.
(2016) 822–835.
[43] T. Kandemir, M.E. Schuster, A. Senyshyn, M. Behrens, R. Schlogl, The Haber–
[12] R. Lan, J.T.S. Irvine, S.W. Tao, Ammonia and related chemicals as potential
Bosch process revisited: on the real structure and stability of ‘‘ammonia
indirect hydrogen storage materials, Int. J. Hydrogen Energy 37 (2012) 1482–
iron” under working conditions, Angew. Chem., Int. Ed. 52 (2013) 12,723–
1494.
12,726.
[13] G.N. Schrauzer, T.D. Guth, Photocatalytic reactions. 1. Photolysis of water and
[44] H. Topsoe, J.A. Dumesic, M. Boudart, Alumina as a textural promoter of iron
photoreduction of nitrogen on titanium dioxide, J. Am. Chem. Soc. 99 (1977)
synthetic ammonia catalysts, J. Catal. 28 (1973) 477–488.
7189–7193.
[45] C.J.H. Jacobsen, Novel class of ammonia synthesis catalysts, Chem. Commun.
[14] G.N. Schrauzer, Photoreduction of nitrogen on TiO2 and TiO2-containing
1057–1058 (2000).
minerals, Energy Efficiency and Renewable Energy Through Nanotechnology,
[46] R. Kojima, K. Aika, Cobalt molybdenum bimetallic nitride catalysts for
Springer London, London, 2011, pp. 601–623.
ammonia synthesis, Chem. Lett. 29 (2000) 514–515.
[15] V. Augugliaro, A. Lauricella, L. Rizzuti, M. Schiavello, A. Sclafani, Conversion of
[47] J.S.J. Hargreaves, D. McKay, A comparison of the reactivity of lattice nitrogen
solar energy to chemical energy by photoassisted processes—I. Preliminary
in Co3Mo3N and Ni2Mo3N catalysts, J. Mol. Catal. A: Chem. 305 (2009) 125–
results on ammonia production over doped titanium dioxide catalysts in a
129.
fluidized bed reactor, Int. J. Hydrogen Energy 7 (1982) 845–849.
[48] C.D. Zeinalipour-Yazdi, J.S.J. Hargreaves, C.R.A. Catlow, Nitrogen activation in
[16] V. Augugliaro, F. Dalba, L. Rizzuti, M. Schiavello, A. Sclafani, Conversion of
a Mars–van Krevelen mechanism for ammonia synthesis on Co3Mo3N, J. Phys.
solar energy to chemical energy by photoassisted processes—II. Influence of
Chem. C 119 (2015) 28,368–28,376.
the iron content on the activity of doped titanium dioxide catalysts for
[49] S.M. Hunter, D.H. Gregory, J.S. Hargreaves, M. Richard, D. Duprez, N. Bion, A
ammonia photoproduction, Int. J. Hydrogen Energy 7 (1982) 851–855.
Study of 15N/14N isotopic exchange over cobalt molybdenum nitrides, ACS
[17] O. Rusina, O. Linnik, A. Eremenko, H. Kisch, Nitrogen photofixation on
Catal. 3 (2013) 1719–1725.
nanostructured iron titanate films, Chem. - Eur. J. 9 (2003) 561–565.
[50] C.D. Zeinalipour-Yazdi, J.S.J. Hargreaves, C.R.A. Catlow, DFT-D3 study of
[18] O. Linnik, H. Kisch, On the mechanism of nitrogen photofixation at
molecular N2 and H2 activation on Co3Mo3N surfaces, J. Phys. Chem. C 120
nanostructured iron titanate films, Photochem. Photobiol. Sci. 5 (2006)
(2016) 21,390–21,398.
938–942.
[51] C.D. Zeinalipour-Yazdi, J.S.J. Hargreaves, C.R.A. Catlow, Low-T mechanisms of
[19] H. Miyama, N. Fujii, Y. Nagae, Heterogeneous photocatalytic synthesis of
ammonia synthesis on Co3Mo3N, J. Phys. Chem. C 122 (2018) 6078–6082.
ammonia from water and nitrogen, Chem. Phys. Lett. 74 (1980) 523–524.
[52] R. Michalsky, A.M. Avram, B.A. Peterson, P.H. Pfromm, A.A. Peterson, Chemical
[20] A. Sclafani, V. Augugliaro, M. Schiavello, Dinitrogen electrochemical reduction
looping of metal nitride catalysts: low-pressure ammonia synthesis for
to ammonia over iron cathode in aqueous-medium, J. Electrochem. Soc. 130
energy storage, Chem. Sci. 6 (2015) 3965–3974.
(1983) 734–735.
[53] A.R. Singh, B.A. Rohr, J.A. Schwalbe, M. Cargnello, K. Chan, T.F. Jaramillo, I. [82] K. Grasshoff, H. Johannsen, A new sensitive and direct method for the
Chorkendorff, J.K. Nørskov, Electrochemical ammonia synthesis—the automatic determination of ammonia in sea water, ICES J. Mar. Sci. 34 (1972)
selectivity challenge, ACS Catal. 7 (2016) 706–709. 516–521.
[54] J.L. White, M.F. Baruch, J.E. Pander Iii, Y. Hu, I.C. Fortmeyer, J.E. Park, T. Zhang, [83] X. Gao, Y. Wen, D. Qu, L. An, S. Luan, W. Jiang, X. Zong, X. Liu, Z. Sun,
K. Liao, J. Gu, Y. Yan, T.W. Shaw, E. Abelev, A.B. Bocarsly, Light-driven Interference effect of alcohol on Nessler’s reagent in photocatalytic nitrogen
heterogeneous reduction of carbon dioxide: photocatalysts and fixation, ACS Sustainable Chem. Eng. 6 (2018) 5342–5348.
photoelectrodes, Chem. Rev. 115 (2015) 12,888–12,935. [84] R. Chen, C.J. Yang, W.Z. Cai, H.Y. Wang, J.W. Miao, L.P. Zhang, S.L. Chen, B. Liu,
[55] Z. Zhang, J.T. Yates Jr., Band bending in semiconductors: chemical and Use of platinum as the counter electrode to study the activity of nonprecious
physical consequences at surfaces and interfaces, Chem. Rev. 112 (2012) metal catalysts for the hydrogen evolution reaction, ACS Energy Lett. 2 (2017)
5520–5551. 1070–1075.
[56] S. Linic, P. Christopher, D.B. Ingram, Plasmonic-metal nanostructures for [85] S.H. Yuen, A.G. Pollard, The determination of nitrogen in agricultural
efficient conversion of solar to chemical energy, Nat. Mater. 10 (2011) 911– materials by the nessler reagent. I.—preparation of the reagent, J. Sci. Food
921. Agric. 3 (1952) 441–447.
[57] T. Tatsuma, H. Nishi, T. Ishida, Plasmon-induced charge separation: chemistry [86] J. Thompson, G. Morrison, Determination of organic nitrogen. control of
and wide applications, Chem. Sci. 8 (2017) 3325–3337. variables in the use of nessler’s reagent, Anal. Chem. 23 (1951) 1153–1157.
[58] D. Zhu, L. Zhang, R.E. Ruther, R.J. Hamers, Photo-illuminated diamond as a [87] P.L. Searle, The berthelot or indophenol reaction and its use in the analytical-
solid-state source of solvated electrons in water for nitrogen reduction, Nat. chemistry of nitrogen - a review, Analyst 109 (1984) 549–568.
Mater. 12 (2013) 836–841. [88] R. Michalski, Ion chromatography applications in wastewater analysis,
[59] J.A. Bandy, D. Zhu, R.J. Hamers, Photocatalytic reduction of nitrogen to Separations 5 (2018).
ammonia on diamond thin films grown on metallic substrates, Diamond [89] S.B. Butt, M. Riaz, Determination of cations and anions in environmental
Relat. Mater. 64 (2016) 34–41. samples by HPLC: review, J. Liq. Chromatogr. Relat. Technol. 32 (2009) 1045–
[60] J.R. Christianson, D. Zhu, R.J. Hamers, J.R. Schmidt, Mechanism of N2 1064.
reduction to NH3 by aqueous solvated electrons, J. Phys. Chem. B 118 [90] D.H. Thomas, M. Rey, P.E. Jackson, Determination of inorganic cations and
(2014) 195–203. ammonium in environmental waters by ion chromatography with a high-
[61] J.H. Montoya, C. Tsai, A. Vojvodic, J.K. Nørskov, The challenge of capacity cation-exchange column, J. Chromatogr. A 956 (2002) 181–186.
electrochemical ammonia synthesis: a new perspective on the role of [91] S. Bourgeois, D. Diakite, M. Perdereau, A study of TiO2 powders as a support
nitrogen scaling relations, ChemSusChem 8 (2015) 2180–2186. for the photochemical-synthesis of ammonia, React. Solids 6 (1988) 95–104.
[62] E. Skulason, T. Bligaard, S. Gudmundsdottir, F. Studt, J. Rossmeisl, F. Abild- [92] M.M. Khader, N.N. Lichtin, G.H. Vurens, M. Salmeron, G.A. Somorjai,
Pedersen, T. Vegge, H. Jonsson, J.K. Nørskov, A theoretical evaluation of Photoassisted catalytic dissociation of water and reduction of nitrogen to
possible transition metal electro-catalysts for N2 reduction, Phys. Chem. ammonia on partially reduced ferric oxide, Langmuir 3 (1987) 303–304.
Chem. Phys. 14 (2012) 1235–1245. [93] K. Tennakone, S. Wickramanayake, C.A.N. Fernando, O.A. Ileperuma, S.
[63] J.G. Howalt, T. Bligaard, J. Rossmeisl, T. Vegge, DFT based study of transition Punchihewa, Photocatalytic nitrogen reduction using visible-light, J Chem
metal nano-clusters for electrochemical NH3 production, Phys. Chem. Chem. Soc Chem. Comm. (1987) 1078–1080.
Phys. 15 (2013) 7785–7795. [94] K. Tennakone, J.M.S. Bandara, C.T.K. Thaminimulla, W.D.W. Jayatilake, U.S.
[64] D. Bao, Q. Zhang, F.L. Meng, H.X. Zhong, M.M. Shi, Y. Zhang, J.M. Yan, Q. Jiang, Ketipearachchi, O.A. Ileperuma, M.K.A. Priyadarshana, Photoreduction of
X.B. Zhang, Electrochemical reduction of N2 under ambient conditions for dinitrogen to ammonia by ultrafine particles of iron hydroxide oxide (Fe(O)
artificial N2 fixation and renewable energy storage using N2/NH3 cycle, Adv. OH) formed by photohydrolysis of iron(II) bicarbonate, Langmuir 7 (1991)
Mater. 29 (2017). 2166–2168.
[65] H. Li, J. Shang, J. Shi, K. Zhao, L. Zhang, Facet-dependent solar ammonia [95] E. Endoh, J.K. Leland, A.J. Bard, Heterogeneous photoreduction of nitrogen to
synthesis of BiOCl nanosheets via a proton-assisted electron transfer ammonia on tungsten-oxide, J. Phys. Chem. 90 (1986) 6223–6226.
pathway, Nanoscale 8 (2016) 1986–1993. [96] K. Tennakone, S. Punchihewa, R. Tantrigoda, Nitrogen photoreduction with
[66] J.S. Anderson, J. Rittle, J.C. Peters, Catalytic conversion of nitrogen to ammonia cuprous chloride coated hydrous cuprous-oxide, Sol. Energy Mater. 18 (1989)
by an iron model complex, Nature 501 (2013) 84–87. 217–221.
[67] J. Greeley, Theoretical heterogeneous catalysis: scaling relationships and [97] K. Tennakone, C.T.K. Thaminimulla, W.C.B. Kiridena, Nitrogen photoreduction
computational catalyst design, Annu. Rev. Chem. Biomol. Eng. 7 (2016) 605– by coprecipitated hydrous oxides of samarium(III) and vanadium(III),
635. Langmuir 9 (1993) 723–726.
[68] J.G. Howalt, T. Vegge, Electrochemical ammonia production on [98] F. Khan, P.L. Yue, L. Rizzuti, V. Augugliaro, M. Schiavello, Photochemical-
molybdenum nitride nanoclusters, Phys. Chem. Chem. Phys. 15 (2013) synthesis of ammonia over zeolites, J Chem Soc Chem. Comm. (1981) 1049–
20,957–20,965. 1050.
[69] A.B. Hoskuldsson, Y. Abghoui, A.B. Gunnarsdottir, E. Skulason, Computational [99] F. Khan, P. Yue, L. Rizzuti, V. Augugilaro, A. Brucato, Photoassisted water
screening of rutile oxides for electrochemical ammonia formation, ACS cleavage and nitrogen-fixation over titanium-exchanged zeolites, Ind. Eng.
Sustainable Chem. Eng. 5 (2017) 10,327–10,333. Chem. Prod. Res. Dev. 22 (1983) 238–241.
[70] V. Kordali, G. Kyriacou, C. Lambrou, Electrochemical synthesis of ammonia at [100] C.R. Dickson, A.J. Nozik, Nitrogen fixation via photoenhanced reduction on p-
atmospheric pressure and low temperature in a solid polymer electrolyte cell, gallium phosphide electrodes, J. Am. Chem. Soc. 100 (1978) 8007–8009.
Chem. Commun. 1673–1674 (2000). [101] Y. Bai, L. Ye, T. Chen, L. Wang, X. Shi, X. Zhang, D. Chen, Facet-dependent
[71] T.H. Rod, A. Logadottir, J.K. Nørskov, Ammonia synthesis at low temperatures, photocatalytic N2 fixation of bismuth-rich Bi5O7I nanosheets, ACS Appl.
J. Chem. Phys. 112 (2000) 5343–5347. Mater. Interfaces 0–0 (2016).
[72] A. Vojvodic, A.J. Medford, F. Studt, F. Abild-Pedersen, T.S. Khan, T. Bligaard, J.K. [102] S. Sun, X. Li, W. Wang, L. Zhang, X. Sun, Photocatalytic robust solar energy
Nørskov, Exploring the limits: a low-pressure, low-temperature Haber-Bosch reduction of dinitrogen to ammonia on ultrathin MoS2, Appl. Catal., B 200
process, Chem. Phys. Lett. 598 (2014) 108–112. (2017) 323–329.
[73] Y. Abghoui, A.L. Garden, V.F. Hlynsson, S. Bjorgvinsdottir, H. Olafsdottir, E. [103] C.H. Lui, A.J. Frenzel, D.V. Pilon, Y.H. Lee, X. Ling, G.M. Akselrod, J. Kong, N.
Skulason, Enabling electrochemical reduction of nitrogen to ammonia at Gedik, Trion-induced negative photoconductivity in monolayer MoS2, Phys.
ambient conditions through rational catalyst design, Phys. Chem. Chem. Phys. Rev. Lett. 113 (2014) 166,801.
17 (2015) 4909–4918. [104] Y. Zhao, Y. Zhao, G.I.N. Waterhouse, L. Zheng, X. Cao, F. Teng, L.Z. Wu, C.H.
[74] Y. Abghoui, A.L. Garden, J.G. Howalt, T. Vegge, E. Skúlason, Electroreduction of Tung, D. O’Hare, T. Zhang, Layered-double-hydroxide nanosheets as efficient
N2 to ammonia at ambient conditions on mononitrides of Zr, Nb, Cr, and V: a visible-light-driven photocatalysts for dinitrogen fixation, Adv. Mater. 29
DFT guide for experiments, ACS Catal. 6 (2015) 635–646. (2017) 1,703,828.
[75] X. Li, X. Sun, L. Zhang, S. Sun, W. Wang, Efficient photocatalytic fixation of N2 [105] S.M. Sun, Q. An, W.Z. Wang, L. Zhang, J.J. Liu, W.A. Goddard, Efficient
by KOH-treated g-C3N4, J. Mater. Chem. A 6 (2018) 3005–3011. photocatalytic reduction of dinitrogen to ammonia on bismuth monoxide
[76] ASTM D1426–15, Standard Test Methods for Ammonia Nitrogen In Water, quantum dots, J. Mater. Chem. A 5 (2017) 201–209.
ASTM International, West Conshohocken, PA, 2015, www.astm.org. [106] G. Dong, W. Ho, C. Wang, Selective photocatalytic N2 fixation dependent on g-
[77] ASTM D6919–17 Standard Test Method for Determination of Dissolved Alkali C3N4 induced by nitrogen vacancies, J. Mater. Chem. A 3 (2015) 23,435–
and Alkaline Earth Cations and Ammonium in Water and Wastewater by Ion 23,441.
Chromatography, ASTM International, West Conshohocken, PA, 2017, www. [107] H.Q. Ma, Z.Y. Shi, S. Li, N. Liu, Large-scale production of graphitic carbon
astm.org. nitride with outstanding nitrogen photofixation ability via a convenient
[78] US, EPA Method 300.7 Dissolved sodium, ammonium, potassium, magnesium microwave treatment, Appl. Surf. Sci. 379 (2016) 309–315.
and calcium in wet deposition by chemically suppressed ion [108] H.Q. Ma, Z.Y. Shi, Q. Li, S. Li, Preparation of graphitic carbon nitride with large
chromatography. US, EPA, Cincinnati, OH. 1986. specific surface area and outstanding N-2 photofixation ability via a dissolve-
[79] Standard method for examination of water and wastewater, American Public regrowth process, J. Phys. Chem. Solids 99 (2016) 51–58.
Health Association (APHA) (2005) 20th edn. APHA, AWWA, WPCF, [109] G. Wu, Y. Gao, B.H. Zheng, Template-free method for synthesizing sponge-
Washington. like graphitic carbon nitride with a large surface area and outstanding
[80] L. Zhou, C.E. Boyd, Comparison of Nessler, phenate, salicylate and ion nitrogen photofixation ability induced by nitrogen vacancies, Ceram. Int. 42
selective electrode procedures for determination of total ammonia nitrogen (2016) 6985–6992.
in aquaculture, Aquaculture 450 (2016) 187–193. [110] H. Li, J. Shang, Z. Ai, L. Zhang, Efficient visible light nitrogen fixation with
[81] N.T. Crosby, Determination of ammonia by the Nessler method in waters BiOBr nanosheets of oxygen vacancies on the exposed 001 facets, J. Am.
containing hydrazine, Analyst 93 (1968). Chem. Soc. 137 (2015) 6393–6399.
[111] H. Hirakawa, M. Hashimoto, Y. Shiraishi, T. Hirai, Photocatalytic conversion of photocatalyst with aromatic rings as electron mediators for visible-light
nitrogen to ammonia with water on surface oxygen vacancies of titanium photocatalytic nitrogen fixation, Appl. Catal., B 218 (2017) 600–610.
dioxide, J. Am. Chem. Soc. 139 (2017) 10,929–10,936. [137] S.H. Cao, H. Chen, F. Jiang, X. Wang, Nitrogen photofixation by ultrathin
[112] C. Li, T. Wang, Z.J. Zhao, W. Yang, J. Li, A. Li, Z. Yang, G.A. Ozin, J. Gong, amine-functionalized graphitic carbon nitride nanosheets as a gaseous
Promoted fixation of molecular nitrogen with surface oxygen vacancies on product from thermal polymerization of urea, Appl. Catal., B 224 (2018)
plasmon-enhanced TiO2 photoelectrodes, Angew. Chem., Int. Ed. (2018). 222–229.
[113] H. Liu, Z. Fang, Y. Su, Y. Suo, S. Huang, Y. Zhang, K. Ding, Different atomic [138] X. Sun, D. Jiang, L. Zhang, S.M. Sun, W.Z. Wang, Enhanced Nitrogen
terminations affect the photocatalytic nitrogen fixation of bismuth Photofixation over LaFeO3 via Acid Treatment, ACS Sustainable Chem. Eng.
oxybromide: a first principles study, Chem., – Asian J. 13 (2018) 799–808. 5 (2017) 9965–9971.
[114] S.Z. Hu, X. Chen, Q. Li, Y.F. Zhao, W. Mao, Effect of sulfur vacancies on the [139] J. Liu, M.S. Kelley, W. Wu, A. Banerjee, A.P. Douvalis, J. Wu, Y. Zhang, G.C.
nitrogen photofixation performance of ternary metal sulfide photocatalysts, Schatz, M.G. Kanatzidis, Nitrogenase-mimic iron-containing chalcogels for
Catal. Sci. Technol. 6 (2016) 5884–5890. photochemical reduction of dinitrogen to ammonia, Proc. Natl. Acad. Sci. U. S.
[115] Y.H. Cao, S.Z. Hu, F.Y. Li, Z.P. Fan, J. Bai, G. Lu, Q. Wang, Photofixation of A. 113 (2016) 5530–5535.
atmospheric nitrogen to ammonia with a novel ternary metal sulfide catalyst [140] A. Banerjee, B.D. Yuhas, E.A. Margulies, Y. Zhang, Y. Shim, M.R. Wasielewski,
under visible light, RSC Adv. 6 (2016) 49,862–49,867. M.G. Kanatzidis, Photochemical nitrogen conversion to ammonia in ambient
[116] L. Palmisano, V. Augugliaro, A. Sclafani, M. Schiavello, Activity of chromium- conditions with FeMoS-chalcogels, J. Am. Chem. Soc. 137 (2015) 2030–2034.
ion-doped titania for the dinitrogen photoreduction to ammonia and for the [141] N.N. Rao, S. Dube, P.Natarajan Manjubala, Photocatalytic reduction of
phenol photodegradation, J. Phys. Chem. 92 (1988) 6710–6713. nitrogen over (Fe, Ru or Os)/TiO2 catalysts, Appl. Catal., B 5 (1994) 33–42.
[117] J. Soria, J.C. Conesa, V. Augugliaro, L. Palmisano, M. Schiavello, A. Sclafani, [142] K.T. Ranjit, T.K. Varadarajan, B. Viswanathan, Photocatalytic reduction of
Dinitrogen photoreduction to ammonia over titanium dioxide powders dinitrogen to ammonia over noble-metal-loaded TiO2, J Photoch Photobio A
doped with ferric ions, J. Phys. Chem. 95 (1991) 274–282. 96 (1996) 181–185.
[118] C. Martin, I. Martin, V. Rives, L. Palmisano, M. Schiavello, Structural and [143] C.M. Janet, S. Navaladian, B. Viswanathan, T.K. Varadarajan, R.P. Viswanath,
surface characterization of the polycrystalline system CrxOy TiO2 employed Heterogeneous Wet Chemical Synthesis of Superlattice-Type Hierarchical
for photoreduction of dinitrogen and photodegradation of phenol, J. Catal. ZnO Architectures for Concurrent H2 Production and N2 Reduction, J. Phys.
134 (1992) 434–444. Chem. C 114 (2010) 2622–2632.
[119] O.A. Ileperuma, K. Tennakone, W.D.D.P. Dissanayake, Photocatalytic behavior [144] L. Li, Y. Wang, S. Vanka, X. Mu, Z. Mi, C.J. Li, Nitrogen photofixation over III-
of metal doped titanium-dioxide - studies on the photochemical-synthesis of nitride nanowires assisted by ruthenium clusters of low atomicity, Angew.
ammonia on Mg/TiO2 catalyst systems, Appl. Catal. 62 (1990) L1–L5. Chem., Int. Ed. 56 (2017) 8701–8705.
[120] O.A. Ileperuma, C.T.K. Thaminimulla, W.C.B. Kiridena, Photoreduction of N2 to [145] C.N.R. Rao, G.R. Rao, Nature of Nitrogen Adsorbed on Transition-Metal
NH3 and H2O to H2 on metal doped TiO2 catalysts, M = Ce, V, Sol. Energy Surfaces as Revealed by Electron-Spectroscopy and Cognate Techniques, Surf.
Mater. Sol. Cells 28 (1993) 335–343. Sci. Rep. 13 (1991) 221–263.
[121] W.R. Zhao, J. Zhang, X. Zhu, M. Zhang, J. Tang, M. Tan, Y. Wang, [146] Y. Lu, Y. Yang, T. Zhang, Z. Ge, H. Chang, P. Xiao, Y. Xie, L. Hua, Q. Li, H. Li, B.
Enhanced nitrogen photofixation on Fe-doped TiO2 with highly exposed Ma, N. Guan, Y. Ma, Y. Chen, Photoprompted hot electrons from bulk cross-
(101) facets in the presence of ethanol as scavenger, Appl. Catal., B 144 (2014) linked graphene materials and their efficient catalysis for atmospheric
468–477. ammonia synthesis, ACS Nano 10 (2016) 10,507–10,515.
[122] S. Hu, X. Chen, Q. Li, F. Li, Z. Fan, H. Wang, Y. Wang, B. Zheng, G. Wu, Fe3+ [147] S.A. Jensen, Z. Mics, I. Ivanov, H.S. Varol, D. Turchinovich, F.H. Koppens, M.
doping promoted N2 photofixation ability of honeycombed graphitic carbon Bonn, K.J. Tielrooij, Competing ultrafast energy relaxation pathways in
nitride: the experimental and density functional theory simulation analysis, photoexcited graphene, Nano Lett. 14 (2014) 5839–5845.
Appl. Catal., B 201 (2017) 58–69. [148] Y. Yang, T. Zhang, Z. Ge, Y. Lu, H. Chang, P. Xiao, R. Zhao, Y. Ma, Y. Chen, Highly
[123] B.M. Comer, A.J. Medford, Analysis of Photocatalytic Nitrogen Fixation on enhanced stability and efficiency for atmospheric ammonia photocatalysis by
Rutile TiO2(110), ACS Sustainable Chem. Eng. 6 (2018) 4648–4660. hot electrons from a graphene composite catalyst with Al2O3, Carbon 124
[124] S. Hu, X. Qu, J. Bai, P. Li, Q. Li, F. Wang, L. Song, Effect of Cu(I)–N Active Sites on (2017) 72–78.
the N2 Photofixation Ability over Flowerlike Copper-Doped g-C3N4 Prepared [149] L. Ye, C. Han, Z. Ma, Y. Leng, J. Li, X. Ji, D. Bi, H. Xie, Z. Huang, Ni2P loading on
via a Novel Molten Salt-Assisted Microwave Process: The Experimental and Cd0.5Zn0.5S solid solution for exceptional photocatalytic nitrogen fixation
Density Functional Theory Simulation Analysis, ACS Sustainable Chem. Eng. 5 under visible light, Chem. Eng. J. 307 (2017) 311–318.
(2017) 6863–6872. [150] P.X. Qiu, C.M. Xu, N. Zhou, H. Chen, F. Jiang, Metal-free black phosphorus
[125] Y. Hao, X. Dong, S. Zhai, H. Ma, X. Wang, X. Zhang, Hydrogenated bismuth nanosheets-decorated graphitic carbon nitride nanosheets with C-P bonds for
molybdate nanoframe for efficient sunlight-driven nitrogen fixation from air, excellent photocatalytic nitrogen fixation, Appl. Catal., B 221 (2018) 27–35.
Chem. - Eur. J. 22 (2016) 18,722–18,728. [151] X. Li, W. Wang, D. Jiang, S. Sun, L. Zhang, X. Sun, Efficient solar-driven
[126] Y.H. Tian, S. Hu, X. Sheng, Y. Duan, J. Jakowski, B.G. Sumpter, J. Huang, Non- nitrogen fixation over carbon-tungstic-acid hybrids, Chem. - Eur. J. 22 (2016)
transition-metal catalytic system for N2 reduction to NH3: a density 13,819–13,822.
functional theory study of al-doped graphene, J. Phys. Chem. Lett. 9 (2018) [152] X. Shi, K. Ueno, N. Takabayashi, H. Misawa, Plasmon-enhanced photocurrent
570–576. generation and water oxidation with a gold nanoisland-loaded titanium
[127] X. Feng, H. Chen, F. Jiang, X. Wang, Enhanced visible-light photocatalytic dioxide photoelectrode, J. Phys. Chem. C 117 (2012) 2494–2499.
nitrogen fixation over semicrystalline graphitic carbon nitride: Oxygen and [153] T. Oshikiri, K. Ueno, H. Misawa, Plasmon-induced ammonia synthesis
sulfur co-doping for crystal and electronic structure modulation, J. Colloid. through nitrogen photofixation with visible light irradiation, Angew. Chem.,
Interface Sci. 509 (2018) 298–306. Int. Ed. 53 (2014) 9802–9805.
[128] Q.S. Li, K. Domen, S. Naito, T. Onishi, K. Tamaru, Photocatalytic synthesis and [154] C.L. Xiao, H. Hu, X.Y. Zhang, D.R. MacFarlane, Nanostructured Gold/Bismutite
photodecomposition of ammonia over SrTiO3 and BaTiO3 based catalysts, Hybrid Heterocatalysts for Plasmon-Enhanced Photosynthesis of Ammonia,
Chem. Lett. 12 (1983) 321–324. ACS Sustainable Chem. Eng. 5 (2017) 10,858–10,863.
[129] K. Tennakone, C.A.N. Fernando, S. Wickramanayake, M.W.P. Damayanthi, L.H. [155] M. Ali, F. Zhou, K. Chen, C. Kotzur, C. Xiao, L. Bourgeois, X. Zhang, D.R.
K. Silva, W. Wijeratne, O.A. Illeperuma, S. Punchihewa, Photocatalytic MacFarlane, Nanostructured photoelectrochemical solar cell for nitrogen
Reduction of Nitrogen to Ammonia with Coprecipitated Fe(III) and Ti(IV) reduction using plasmon-enhanced black silicon, Nat. Commun. 7 (2016)
Hydrous Oxides, Sol. Energy Mater. 17 (1988) 47–53. 11,335.
[130] Q.X. Liu, L.H. Ai, J. Jiang, MXene-derived TiO2@C/g-C3N4 heterojunctions for [156] S. Wang, X. Hai, X. Ding, K. Chang, Y. Xiang, X. Meng, Z. Yang, H. Chen, J. Ye,
highly efficient nitrogen photofixation, J. Mater. Chem. A 6 (2018) 4102– Light-switchable oxygen vacancies in ultrafine Bi5O7Br nanotubes for
4110. boosting solar-driven nitrogen fixation in pure water, Adv. Mater. 29 (2017).
[131] Q. Zhang, S. Hu, Z. Fan, D. Liu, Y. Zhao, H. Ma, F. Li, Preparation of g-C3N4/ [157] T. Oshikiri, K. Ueno, H. Misawa, Selective dinitrogen conversion to ammonia
ZnMoCdS hybrid heterojunction catalyst with outstanding nitrogen using water and visible light through plasmon-induced charge separation,
photofixation performance under visible light via hydrothermal post- Angew. Chem., Int. Ed. 55 (2016) 3942–3946.
treatment, Dalton Trans. 45 (2016) 3497–3505. [158] H.S. Zeng, S. Terazono, T. Tanuma, A novel catalyst for ammonia synthesis at
[132] M. Lashgari, P. Zeinalkhani, Photocatalytic N2 conversion to ammonia using ambient temperature and pressure: Visible light responsive photocatalyst
efficient nanostructured solar-energy-materials in aqueous media: A novel using localized surface plasmon resonance, Catal. Commun. 59 (2015) 40–44.
hydrogenation strategy and basic understanding of the phenomenon, Appl. [159] N.D. Spencer, R.C. Schoonmaker, G.A. Somorjai, Iron single-crystals as
Catal., A 529 (2017) 91–97. ammonia-synthesis catalysts - effect of surface-structure on catalyst
[133] H. Tada, T. Mitsui, T. Kiyonaga, T. Akita, K. Tanaka, All-solid-state Z-scheme in activity, J. Catal. 74 (1982) 129–135.
CdS-Au-TiO2 three-component nanojunction system, Nat. Mater. 5 (2006) [160] M.M. Shi, D. Bao, B.R. Wulan, Y.H. Li, Y.F. Zhang, J.M. Yan, Q. Jiang, Au sub-
782–786. nanoclusters on TiO2 toward highly efficient and selective electrocatalyst for
[134] H. Liang, H. Zou, S. Hu, Preparation of the W18O49/g-C3N4 heterojunction N2 conversion to NH3 at ambient conditions, Adv. Mater. 29 (2017).
catalyst with full-spectrum-driven photocatalytic N2 photofixation ability [161] S.J. Li, D. Bao, M.M. Shi, B.R. Wulan, J.M. Yan, Q. Jiang, Amorphizing of Au
from the UV to near infrared region, New J. Chem. 41 (2017) 8920–8926. nanoparticles by CeOx–RGO Hybrid support towards highly efficient
[135] Y. Wang, W. Wei, M. Li, S. Hu, J. Zhang, R. Feng, In situ construction of Z- electrocatalyst for N2 reduction under ambient conditions, Adv. Mater. 29
scheme g-C3N4/Mg1.1Al0.3Fe0.2O1.7 nanorod heterostructures with high N2 (2017).
photofixation ability under visible light, RSC Adv. 7 (2017) 18099–18107. [162] Y. Yao, S. Zhu, H. Wang, H. Li, M. Shao, A spectroscopic study on the nitrogen
[136] S. Cao, N. Zhou, F. Gao, H. Chen, F. Jiang, All-solid-state Z-scheme 3,4- electrochemical reduction reaction on gold and platinum surfaces, J. Am.
dihydroxybenzaldehyde-functionalized Ga2O3/graphitic carbon nitride Chem. Soc. 140 (2018) 1496–1501.
[163] S. Back, Y. Jung, On the mechanism of electrochemical ammonia synthesis on [190] K.D. Kreuer, Proton-conducting oxides, Annu. Rev. Mater. Res. 33 (2003) 333–
the Ru catalyst, Phys. Chem. Chem. Phys. 18 (2016) 9161–9166. 359.
[164] K. Kugler, M. Luhn, J.A. Schramm, K. Rahimi, M. Wessling, Galvanic deposition [191] J.W. Phair, S.P.S. Badwal, Review of proton conductors for hydrogen
of Rh and Ru on randomly structured Ti felts for the electrochemical NH3 separation, Ionics 12 (2006) 103–115.
synthesis, Phys. Chem. Chem. Phys. 17 (2015) 3768–3782. [192] T. Norby, M. Wideroe, R. Glockner, Y. Larring, Hydrogen in oxides, Dalton
[165] A. Ishikawa, T. Doi, H. Nakai, Catalytic performance of Ru, Os, and Rh Trans. 3012–3018 (2004).
nanoparticles for ammonia synthesis: a density functional theory analysis, J. [193] E. Panagos, I. Voudouris, M. Stoukides, Modelling of equilibrium limited
Catal. 357 (2018) 213–222. hydrogenation reactions carried out in H+ conducting solid oxide membrane
[166] H.M. Liu, S.H. Han, Y. Zhao, Y.Y. Zhu, X.L. Tian, J.H. Zeng, J.X. Jiang, B.Y. Xia, Y. reactors, Chem. Eng. Sci. 51 (1996) 3175–3180.
Chen, Surfactant-free atomically ultrathin rhodium nanosheet [194] G. Marnellos, S. Zisekas, M. Stoukides, Synthesis of ammonia at atmospheric
nanoassemblies for efficient nitrogen electroreduction, J. Mater. Chem. A 6 pressure with the use of solid state proton conductors, J. Catal. 193 (2000)
(2018) 3211–3217. 80–87.
[167] J. Zhao, Z. Chen, Single Mo Atom Supported on Defective Boron Nitride [195] R.S. Roth, Classification of Perovskite and Other ABO3-Type Compounds, J Res
Monolayer as an Efficient Electrocatalyst for Nitrogen Fixation: A Nat Bur Stand 58 (1957) 75–88.
Computational Study, J. Am. Chem. Soc. 139 (2017) 12,480–12,487. [196] J. Sunarso, S.S. Hashim, N. Zhu, W. Zhou, Perovskite oxides applications in
[168] D. Yang, T. Chen, Z. Wang, Electrochemical reduction of aqueous nitrogen high temperature oxygen separation, solid oxide fuel cell and membrane
(N2) at a low overpotential on (110)-oriented Mo nanofilm, J. Mater. Chem. A reactor: A review, Prog. Energy Combust. Sci. 61 (2017) 57–77.
5 (2017) 18,967–18,971. [197] C.G. Vayenas, S. Ladas, S. Bebelis, I.V. Yentekakis, S. Neophytides, J. Yi, C.
[169] Y. Abghoui, E. Skulason, Computational predictions of catalytic activity of Karavasilis, C. Pliangos, Electrochemical promotion in catalysis - non-faradaic
zincblende (110) surfaces of metal nitrides for electrochemical ammonia electrochemical modification of catalytic activity, Electrochim. Acta 39
synthesis, J. Phys. Chem. C 121 (2017) 6141–6151. (1994) 1849–1855.
[170] L.M. Azofra, N. Li, D.R. MacFarlane, C. Sun, Promising prospects for 2D d2–d4 [198] M. Ouzounidou, A. Skodra, C. Kokkofitis, M. Stoukides, Catalytic and
M3C2 transition metal carbides (MXenes) in N2 capture and conversion into electrocatalytic synthesis of NH3 in a H+ conducting cell by using an
ammonia, Energy Environ. Sci. 9 (2016) 2545–2549. industrial Fe catalyst, Solid State Ionics 178 (2007) 153–159.
[171] Q.Y. Li, L.Z. He, C.H. Sun, X.W. Zhang, Computational Study of MoN2 [199] K. Zhang, J. Sunarso, Z.P. Shao, W. Zhou, C.H. Sun, S.B. Wang, S.M. Liu,
Monolayer as Electrochemical Catalysts for Nitrogen Reduction, J. Phys. Research progress and materials selection guidelines on mixed conducting
Chem. C 121 (2017) 27,563–27,568. perovskite-type ceramic membranes for oxygen production, RSC Adv. 1
[172] S. Chen, S. Perathoner, C. Ampelli, C. Mebrahtu, D. Su, G. Centi, (2011) 1661–1676.
Electrocatalytic synthesis of ammonia at room temperature and [200] Z. Li, R. Liu, J. Wang, Z. Xu, Y. Xie, B. Wang, Preparation of double-doped
atmospheric pressure from water and nitrogen on a carbon-nanotube- BaCeO3 and its application in the synthesis of ammonia at atmospheric
based electrocatalyst, Angew. Chem., Int. Ed. 56 (2017) 2699–2703. pressure, Sci. Technol. Adv. Mater. 8 (2016) 566–570.
[173] S. Chen, S. Perathoner, C. Ampelli, C. Mebrahtu, D. Su, G. Centi, Room- [201] Z.J. Li, R.Q. Liu, J.D. Wang, Y.H. Xie, F. Yue, Preparation of BaCe0.8Gd0.2O3d by
temperature electrocatalytic synthesis of NH3 from H2O and N2 in a gas– the citrate method and its application in the synthesis of ammonia at
liquid–solid three-phase reactor, ACS Sustainable Chem. Eng. 5 (2017) 7393– atmospheric pressure, J. Solid State Electrochem. 9 (2004) 201–204.
7400. [202] H. Inaba, H. Tagawa, Ceria-based solid electrolytes – review, Solid State Ionics
[174] Y. Liu, Y. Su, X. Quan, X. Fan, S. Chen, H. Yu, H. Zhao, Y. Zhang, J. Zhao, Facile 83 (1996) 1–16.
Ammonia Synthesis from Electrocatalytic N2 Reduction under Ambient [203] R. Liu, Y. Xie, J. Wang, Z. Li, B. Wang, Synthesis of ammonia at atmospheric
Conditions on N-Doped Porous Carbon, ACS Catal. 8 (2018) 1186–1191. pressure with Ce0.8M0.2O2d (M=La, Y, Gd, Sm) and their proton conduction at
[175] C.V.S. Kumar, V. Subramanian, Can boron antisites of BNNTs be an efficient intermediate temperature, Solid State Ionics 177 (2006) 73–76.
metal-free catalyst for nitrogen fixation? – A DFT investigation, Phys. Chem. [204] A. Navrotsky, Thermodynamics of solid electrolytes and related oxide
Chem. Phys. 19 (2017) 15,377–15,387. ceramics based on the fluorite structure, J. Mater. Chem. 20 (2010) 10,577–
[176] A. Kraytsberg, Y. Ein-Eli, Review of advanced materials for proton exchange 10,587.
membrane fuel cells, Energy Fuels 28 (2014) 7303–7330. [205] A. Chroneos, B. Yildiz, A. Tarancon, D. Parfitt, J.A. Kilner, Oxygen diffusion in
[177] R. Lan, J.T. Irvine, S. Tao, Synthesis of ammonia directly from air and water at solid oxide fuel cell cathode and electrolyte materials: mechanistic insights
ambient temperature and pressure, Sci. Rep. 3 (2013) 1145. from atomistic simulations, Energy Environ. Sci. 4 (2011) 2774–2789.
[178] G. Xu, R. Liu, J. Wang, Electrochemical synthesis of ammonia using a cell with [206] Z. Li, R. Liu, Y. Xie, S. Feng, J. Wang, A novel method for preparation of doped
a Nafion membrane and SmFe0.7Cu0.3xNixO3 (x = 0–0.3) cathode at Ba3(Ca1.18Nb1.82)O9d: Application to ammonia synthesis at atmospheric
atmospheric pressure and lower temperature, Sci. China, Ser. B: Chem. 52 pressure, Solid State Ionics 176 (2005) 1063–1066.
(2009) 1171–1175. [207] Y.H. Xie, J.E. Wang, R.Q. Liu, X.T. Su, Z.P. Sun, Z.J. Li, Preparation of
[179] J. Kong, A. Lim, C. Yoon, J.H. Jang, H.C. Ham, J. Han, S. Nam, D. Kim, Y.E. Sung, J. La1.9Ca0.1Zr2O6.95 with pyrochlore structure and its application in synthesis
Choi, H.S. Park, Electrochemical Synthesis of NH3 at Low Temperature and of ammonia at atmospheric pressure, Solid State Ionics 168 (2004) 117–121.
Atmospheric Pressure Using a c-Fe2O3 Catalyst, ACS Sustainable Chem. Eng. 5 [208] M. Glerup, O.F. Nielsen, F.W. Poulsen, The Structural Transformation from the
(2017) 10,986–10,995. Pyrochlore Structure, A2B2O7, to the Fluorite Structure, AO2, Studied by
[180] B.L. Sheets, G.G. Botte, Electrochemical nitrogen reduction to ammonia under Raman Spectroscopy and Defect Chemistry Modeling, J. Solid State Chem. 160
mild conditions enabled by a polymer gel electrolyte, Chem. Commun. 54 (2001) 25–32.
(2018) 4250–4253. [209] P.J. Wilde, C.R.A. Catlow, Molecular dynamics study of the effect of doping
[181] J. Nash, X. Yang, J. Anibal, J.H. Wang, Y.S. Yan, B.J. Xu, Electrochemical and disorder on diffusion in gadolinium zirconate, Solid State Ionics 112
nitrogen reduction reaction on noble metal catalysts in proton and hydroxide (1998) 185–195.
exchange membrane electrolyzers, J. Electrochem. Soc. 164 (2017) F1712– [210] T. Omata, K. Okuda, S. Tsugimoto, S. Otsuka-Matsuo-Yao, Water and
F1716. hydrogen evolution properties and protonic conducting behaviors of Ca2+-
[182] T. Murakami, T. Nishikiori, T. Nohira, Y. Ito, Investigation of anodic reaction of doped La2Zr2O7 with a pyrochlore structure, Solid State Ionics 104 (1997)
electrolytic ammonia synthesis in molten salts under atmospheric pressure, J. 249–258.
Electrochem. Soc. 152 (2005) D75–D78. [211] J.-D. Wang, Y.-H. Xie, Z.-F. Zhang, R.-Q. Liu, Z.-J. Li, Protonic conduction in
[183] P. Nikolaidis, A. Poullikkas, A comparative overview of hydrogen Ca2+-doped La2M2O7 (M=Ce, Zr) with its application to ammonia synthesis
production processes, Renewable Sustainable Energy Rev. 67 (2017) 597– electrochemically, Mater. Res. Bull. 40 (2005) 1294–1302.
611. [212] C.R. Stanek, L. Minervini, R.W. Grimes, Nonstoichiometry in A2B2O7
[184] T. Murakami, T. Nohira, Y. Araki, T. Goto, R. Hagiwara, Y.H. Ogata, Electrolytic Pyrochlores, J. Am. Ceram. Soc. 85 (2004) 2792–2798.
synthesis of ammonia from water and nitrogen under atmospheric pressure [213] M. Temkin, V. Pyzhev, Kinetics of ammonia synthesis on promoted iron
using a boron-doped diamond electrode as a nonconsumable anode, catalysts, Acta Physicochim. URSS 12 (1940) 327–356.
Electrochem. Solid-State Lett. 10 (2007) E4–E6. [214] C.N. Satterfield, Heterogeneous catalysis in practice. McGraw-Hill, 1980.
[185] T. Murakami, T. Nohira, T. Goto, Y.H. Ogata, Y. Ito, Electrolytic ammonia [215] C.G. Yiokari, G.E. Pitselis, D.G. Polydoros, A.D. Katsaounis, C.G. Vayenas, High-
synthesis from water and nitrogen gas in molten salt under atmospheric pressure electrochemical promotion of ammonia synthesis over an industrial
pressure, Electrochim. Acta 50 (2005) 5423–5426. iron catalyst, J. Phys. Chem. A 104 (2000) 10,600–10,602.
[186] S. Licht, B. Cui, B. Wang, F.F. Li, J. Lau, S. Liu, Ammonia synthesis. Ammonia [216] J. Nicole, D. Tsiplakides, S. Wodiunig, C. Comninellis, Activation of catalyst for
synthesis by N2 and steam electrolysis in molten hydroxide suspensions of gas-phase combustion by electrochemical pretreatment, J. Electrochem. Soc.
nanoscale Fe2O3, Science 345 (2014) 637–640. 144 (1997) L312–L314.
[187] B. Cui, J. Zhang, S. Liu, X. Liu, W. Xiang, L. Liu, H. Xin, M.J. Lefler, S. Licht, [217] J. Di, M. Chen, C. Wang, J. Zheng, L. Fan, B. Zhu, Samarium doped ceria–(Li/
Electrochemical synthesis of ammonia directly from N2 and water over iron- Na)2CO3 composite electrolyte and its electrochemical properties in low
based catalysts supported on activated carbon, Green Chem. 19 (2017) 298– temperature solid oxide fuel cell, J. Power Sources 195 (2010) 4695–4699.
304. [218] B.H. Wang, J.D. Wang, R. Liu, Y.H. Xie, Z.J. Li, Synthesis of ammonia from
[188] M.T. Nguyen, N. Seriani, R. Gebauer, Nitrogen electrochemically reduced to natural gas at atmospheric pressure with doped ceria–Ca3(PO4), 2–K3PO4
ammonia with hematite: density-functional insights, Phys. Chem. Chem. composite electrolyte and its proton conductivity at intermediate
Phys. 17 (2015) 14,317–14,322. temperature, J. Solid State Electrochem. 11 (2005) 27–31.
[189] A. Skodra, M. Stoukides, Electrocatalytic synthesis of ammonia from steam [219] H. Iwahara, Y. Asakura, K. Katahira, M. Tanaka, Prospect of hydrogen
and nitrogen at atmospheric pressure, Solid State Ionics 180 (2009) 1332– technology using proton-conducting ceramics, Solid State Ionics 168 (2004)
1336. 299–310.
[220] I.A. Amar, R. Lan, C.T.G. Petit, S. Tao, Electrochemical Synthesis of Ammonia [234] M.C. Tsai, U. Ship, I.C. Bassignana, J. Küppers, G. Ertl, A vibrational
Using Fe3Mo3N Catalyst and Carbonate-Oxide Composite Electrolyte, Int. J. spectroscopy study on the interaction of N2 with clean and K-promoted Fe
Electrochem. Sci. 10 (2015) 3757–3766. (111) surfaces: p-bonded dinitrogen as precursor for dissociation, Surf. Sci.
[221] I.A. Amar, R. Lan, C.T.G. Petit, S. Tao, Electrochemical Synthesis of Ammonia 155 (1985) 387–399.
Based on Co3Mo3N Catalyst and LiAlO2–(Li, Na, K)2CO3 Composite [235] H.P. Wang, J.T. Yates, Infrared spectroscopic study of molecular nitrogen
Electrolyte, Electrocatalysis 6 (2014) 286–294. chemisorption on rhodium surfaces, J. Phys. Chem. 88 (1984) 852–856.
[222] I.A. Amar, C.T.G. Petit, L. Zhang, R. Lan, P.J. Skabara, S. Tao, Electrochemical [236] G. Ertl, S.B. Lee, M. Weiss, Adsorption of nitrogen on potassium promoted Fe
synthesis of ammonia based on doped-ceria-carbonate composite electrolyte (111) and (100) surfaces, Surf. Sci. 114 (1982) 527–545.
and perovskite cathode, Solid State Ionics 201 (2011) 94–100. [237] S. Anantharaj, S.R. Ede, K. Sakthikumar, K. Karthick, S. Mishra, S. Kundu,
[223] R. Lan, K.A. Alkhazmi, I.A. Amar, S. Tao, Synthesis of ammonia directly from Recent trends and perspectives in electrochemical water splitting with an
wet air using new perovskite oxide La0.8Cs0.2Fe0.8Ni0.2O3-d as catalyst, emphasis on sulfide, selenide, and phosphide catalysts of Fe Co, and Ni: a
Electrochim. Acta 123 (2014) 582–587. review, ACS Catal. 6 (2016) 8069–8097.
[224] I.A. Amar, R. Lan, S. Tao, Electrochemical Synthesis of Ammonia Directly from [238] Y. Guo, D. Guo, F. Ye, K. Wang, Z. Shi, Synthesis of lawn-like NiS2 nanowires
Wet N2 Using La0.6Sr0.4Fe0.8Cu0.2O3– -Ce0.8Gd0.18Ca0.02O2- Composite Catalyst, on carbon fiber paper as bifunctional electrode for water splitting, Int. J.
J. Electrochem. Soc. 161 (2014) H350–H354. Hydrogen Energy 42 (2017) 17,038–17,048.
[225] R. Lan, K.A. Alkhazmi, I.A. Amar, S. Tao, Synthesis of ammonia directly from [239] P.V. Kamat, S. Jin, Semiconductor photocatalysis: ‘‘Tell us the complete
wet air using Sm0.6Ba0.4Fe0.8Cu0.2O3-d as the catalyst, Faraday Discuss. 182 story!”, ACS Energy Lett. 3 (2018) 622–623.
(2015) 353–363. [240] M. Muuronen, S.M. Parker, E. Berardo, A. Le, M.A. Zwijnenburg, F. Furche,
[226] X. Zhao, F. Yin, N. Liu, G. Li, T. Fan, B. Chen, Highly efficient metal–organic- Mechanism of photocatalytic water oxidation on small TiO2 nanoparticles,
framework catalysts for electrochemical synthesis of ammonia from N2 (air) Chem. Sci. 8 (2017) 2179–2183.
and water at low temperature and ambient pressure, J. Mater. Sci. 52 (2017) [241] A. Kazaryan, R. van Santen, E.J. Baerends, Light-induced water splitting by
10,175–10,185. titanium-tetrahydroxide: a computational study, Phys. Chem. Chem. Phys. 17
[227] K. Kim, N. Lee, C.Y. Yoo, J.N. Kim, H.C. Yoon, J.I. Han, Communication- (2015) 20,308–20,321.
Electrochemical Reduction of Nitrogen to Ammonia in 2-Propanol under [242] X. Zhang, A. Bieberle-Hutter, Modeling and simulations in
Ambient Temperature and Pressure, J. Electrochem. Soc. 163 (2016) F610– photoelectrochemical water oxidation: From single Level to multiscale
F612. modeling, ChemSusChem 9 (2016) 1223–1242.
[228] K. Kim, C.Y. Yoo, J.N. Kim, H.C. Yoon, J.I. Han, Electrochemical synthesis of [243] K.K. Ghuman, L.B. Hoch, P. Szymanski, J.Y. Loh, N.P. Kherani, M.A. El-Sayed, G.
ammonia from water and nitrogen in ethylenediamine under ambient A. Ozin, C.V. Singh, Photoexcited surface frustrated lewis pairs for
temperature and pressure, J. Electrochem. Soc. 163 (2016) F1523–F1526. heterogeneous photocatalytic CO2 reduction, J. Am. Chem. Soc. 138 (2016)
[229] A. Tsuneto, A. Kudo, T. Sakata, Lithium-Mediated Electrochemical Reduction 1206–1214.
of High-Pressure N2 to Nh3, J. Electroanal. Chem. 367 (1994) 183–188. [244] J.A. Farmer, C.T. Campbell, Ceria maintains smaller metal catalyst particles by
[230] F. Koleli, D.B. Kayan, Low overpotential reduction of dinitrogen to ammonia strong metal-support bonding, Science 329 (2010) 933–936.
in aqueous media, J. Electroanal. Chem. 638 (2010) 119–122. [245] H. Tang, Y. Su, B. Zhang, A.F. Lee, M.A. Isaacs, K. Wilson, L. Li, Y. Ren, J. Huang,
[231] G. Xu, R. Liu, Sm1.5Sr0.5MO4 (M=Ni Co, Fe) cathode catalysts for ammonia M. Haruta, B. Qiao, X. Liu, C. Jin, D. Su, J. Wang, T. Zhang, Classical strong
synthesis at atmospheric pressure and low temperature, Chin. J. Chem. 27 metal-support interactions between gold nanoparticles and titanium dioxide,
(2009) 677–680. Sci. Adv. 3 (2017) e1700231.
[232] R.Q. Liu, G.C. Xu, Comparison of electrochemical synthesis of ammonia by [246] B.M. Lindley, A.M. Appel, K. Krogh-Jespersen, J.M. Mayer, A.J.M. Miller,
using sulfonated polysulfone and nafion membrane with Sm1.5Sr0.5NiO4, Evaluating the thermodynamics of electrocatalytic N2 reduction in
Chin. J. Chem. 28 (2010) 139–142. acetonitrile, ACS Energy Lett. 1 (2016) 698–704.
[233] T. Murakami, T. Nohira, Y.H. Ogata, Y. Ito, Electrolytic ammonia synthesis in
molten salts under atmospheric pressure using methane as a hydrogen
source, Electrochem. Solid-State Lett. 8 (2005) D12–D14.

Electron-Driven Heterogeneous Catalytic Synthesis of Ammonia: Current States and Perspective

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Electron-Driven Heterogeneous Catalytic Synthesis of Ammonia: Current States and Perspective

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Electron-Driven Heterogeneous Catalytic Synthesis of Ammonia: Current States and Perspective

Uploaded by

Copyright:

Available Formats

Carbon Resources Conversion 1 (2018) 2–31

Contents lists available at ScienceDirect

Carbon Resources Conversion

Electron-driven heterogeneous catalytic synthesis of ammonia: Current

4.1. Aqueous electrolyte system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

1. Introduction The Haber–Bosch process has dominated the commercial pro-

2.1. Ammonia synthesis on heterogeneous catalysts via 1

2 NH2 þ Hþ þ e ! NH2 þ NH3 ð27Þ

2.3. The quantitative measurement of ammonia

It is a tough job to quantitatively detect ammonium ions or

ð47Þ 3. Photocatalytic N2 fixation to ammonia

ammonia yield of 24 lmol g1 h1. The mechanism is illustrated in

3.2.2. Other as the co-catalyst

Utilization of visible light instead of UV illumination is of para- 4. Electrocatalytic N2 fixation to ammonia

Anode 12OH ! 3O2 þ 6H2 O þ 12e ð49Þ

Overall 6H2 O þ 2N2 ! 4NH3 þ 3O2 ð50Þ

electrolyte experiments. One example was to use a molten LiCl–

3H2 O þ N2 þ 6e ! 3O2 þ 2NH3 ð54Þ

The overall cell reaction is then just the combination of the

electrodes. It had previously been reported that Gd-doped barium

K. Wang et al. / Carbon Resources Conversion 1 (2018) 2–31

5.2.1. Cell design and configuration

Faradaic efficiency has been reported by Xu et al. [178] using

NH3 184.8 lmol L1 with 100% selectivity

5.2.2. Catalysts design and optimization

promoting the selectivity towards ammonia synthesis. The reason

could partially be due to the low activity of Au in the HER process.

NH3 2685 lmol g1 h1

NH3 38.2 lmol g1 h1

oxide support such as titania and ceria can function effectively

NH3 13.3 mg m2 h1

for stabilizing the ultrafine metal clusters by forming half covering

layering, which further ensures the excellent long-term stability of

the electrocatalysts. And more importantly, the SMSI could also

adsorbed with priority over H atoms on the surface of the Mo nan-

N is the basis to achieve higher ammonia selectivity. Further com-

prehensive understanding on the physical and chemical properties

Another interesting strategy comes from the suggestion that

simultaneous 6-electron transfer will not produce any by-

products but directly produce ammonia, resulted in high selectiv-

5.2.3. Potential bias

Potential bias for electrocatalytic process is another key factor

affecting the selectivity. In most of the practical experiments, the

ciency [181]. To date it is widely accepted that the HER activity

transfer process, which is positively affected by the external poten-

scale electron-driven ammonia synthesis with high selectivity will

BiO quantum dot

5.2.4. Control water or proton supply for the cathode

Sheets et al. [64] proposed to use the anion conducting mem-

brane to limit water diffusion and to further suppress the HER.

Nash et al. [181] compared the noble metal as catalysts in the

and hydroxide-conducting membrane as the electrolyte. It is

300 W Xenon lamp, with 2 sun

Simulated light, 200 mW cm2,

proved that alkaline environment favours the Faradaic efficiency

in the alkaline environment is around one order of magnitude

500 W Xenon lamp

500 W Xenon lamp