Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-Atom Protrusions

Article abstract of DOI:10.1016/j.chempr.2020.01.013

Electrocatalytic N₂ reduction reaction (NRR) offers a promising low-energy, sustainable ammonia-synthesizing alternative to Haber-Bosch reaction. One roadblock lying in access to high-performance ammonia electrosynthesis emanates from the unsatisfied ability of electrocatalysts to wreck N≡N bond. Here, we report that interfacial polarization is an efficient scenario to enhance N≡N fracture to boost electrocatalytic ammonia synthesis. As a proof-of-concept demonstration, protrusion-shaped Fe single-atom catalysts immobilized onto MoS₂ nanosheets engender electric fields to polarize N₂. The resultant interfacial polarization fields between Fe-MoS₂ and N₂ drive the injection of more electrons into N₂ antibonding orbitals in a fast manner, leading to a superior ammonia-evolving rate (36.1 ± 3.6 mmol g^?1 h^?1 or 97.5 ± 6 μg h^?1 cm^?2) at low applied potential. Similar phenomena are applicable in Co-MoS₂, Cu-MoS₂, Rh-MoS₂, or Ru-MoS₂, suggesting the potential universality of our interfacial polarization concept in upgrading wide-scope catalysis. Seeking a green, low-cost, sustainable approach to synthesize ammonia is crucial to society development and human living. A promising candidate is electrocatalytic nitrogen reduction. The insufficient ability of electrocatalysts to split the N≡N bond, however, limits the activity and selectivity. Using interfacial polarization as a conceptually novel strategy to promote N≡N disintegration, high-rate ammonia electrosynthesis up to 36.1 ± 3.6 mmol g^?1 h^?1 (97.5 ± 6 μg h^?1 cm^?2) is realized at a low applied potential (?0.2 V versus RHE). This work paves a new way toward replacing Haber-Bosch reaction with ambient ammonia electrosynthesis. Interfacial polarization is reported as a brand new, efficient, and generalizable strategy to accelerate electrocatalytic reduction of N₂ to ammonia. The polarization is established by using an electric field to polarize N₂. The electric field is triggered by protrusion-like single atoms anchored on MoS₂. The interfacial polarization accelerates electron transfer from single atoms to N₂ and thus promotes N₂ reduction. As a result, ammonia synthesis in an electrochemical flow cell proceeds at a high rate of 36.1 ± 3.6 mmol g^?1 h^?1.

Full text of DOI:10.1016/j.chempr.2020.01.013

Article
Accelerated Dinitrogen Electroreduction to
Ammonia via Interfacial Polarization Triggered
by Single-Atom Protrusions
Jie Li, Shang Chen, Fengjiao
Quan, Guangming Zhan, Falong
Jia, Zhihui Ai, Lizhi Zhang
zhanglz@mail.ccnu.edu.cn
HIGHLIGHTS
Interfacial polarization efficiently
enhances N2 electroreduction to
ammonia
Protrusion-like single-atom
catalysts on MoS2 trigger electric
field to polarize N2
Polarization field accelerates
electron injection to boost NhN
fracture
Ammonia electrosynthesis with
high rate and selectivity is
achieved
Interfacial polarization is reported as a brand new, efficient, and generalizable
strategy to accelerate electrocatalytic reduction of N2 to ammonia. The
polarization is established by using an electric field to polarize N2. The electric field
is triggered by protrusion-like single atoms anchored on MoS2. The interfacial
polarization accelerates electron transfer from single atoms to N2 and thus
promotes N2 reduction. As a result, ammonia synthesis in an electrochemical flow
cell proceeds at a high rate of 36.1 G 3.6 mmol gÀ1 hÀ1
.
Li et al., Chem 6, 1–17
April 9, 2020 ª 2020 Elsevier Inc.
https://doi.org/10.1016/j.chempr.2020.01.013

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
Article
Accelerated Dinitrogen Electroreduction
to Ammonia via Interfacial Polarization
Triggered by Single-Atom Protrusions
Jie Li,1,2 Shang Chen,1,2 Fengjiao Quan,1,2 Guangming Zhan,1,2 Falong Jia,1 Zhihui Ai,1
and Lizhi Zhang1,3,
*
SUMMARY
Electrocatalytic N2 reduction reaction (NRR) offers a promising low-energy, sus-
tainable ammonia-synthesizing alternative to Haber-Bosch reaction. One road-
block lying in access to high-performance ammonia electrosynthesis emanates
from the unsatisfied ability of electrocatalysts to wreck NhN bond. Here, we
report that interfacial polarization is an efficient scenario to enhance NhN frac-
ture to boost electrocatalytic ammonia synthesis. As a proof-of-concept demon-
stration, protrusion-shaped Fe single-atom catalysts immobilized onto MoS2
nanosheets engender electric fields to polarize N2. The resultant interfacial
polarization fields between Fe-MoS2 and N2 drive the injection of more elec-
trons into N2 antibonding orbitals in a fast manner, leading to a superior
The Bigger Picture
Seeking a green, low-cost,
sustainable approach to
synthesize ammonia is crucial to
society development and human
living. A promising candidate is
electrocatalytic nitrogen
reduction. The insufficient ability
of electrocatalysts to split the
NhN bond, however, limits the
activity and selectivity. Using
interfacial polarization as a
conceptually novel strategy to
promote NhN disintegration,
high-rate ammonia
ammonia-evolving rate (36.1 G 3.6 mmol gꢀ1 hꢀ1 or 97.5 G 6 mg hꢀ1 cmꢀ2
)
at low applied potential. Similar phenomena are applicable in Co-MoS2, Cu-
MoS2, Rh-MoS2, or Ru-MoS2, suggesting the potential universality of our inter-
facial polarization concept in upgrading wide-scope catalysis.
electrosynthesis up to 36.1 G
3.6 mmol gꢀ1 hꢀ1 (97.5 G 6 mg hꢀ1
cmꢀ2) is realized at a low applied
potential (ꢀ0.2 V versus RHE). This
work paves a new way toward
replacing Haber-Bosch reaction
with ambient ammonia
INTRODUCTION
Ammonia is a sustainable, carbon-free, high-energy fuel and also a building block for
agricultural fertilizer and industrial chemical manufacture.1 Electrocatalytic N2
reduction reaction (NRR) powered by renewable electricity epitomizes a safe,
scalable, low-cost, environmentally benign scenario for ammonia production.2–6
Unfortunately, the majority of NRRs under room temperature and atmospheric
pressure encounter performance handicaps with operation potential being prohib-
itive,7,8 selectivity far lower than that of hydrogen-evolving reaction (HER)9,10 and
electrosynthesis.
NH3 yield at a level of sub-mmol gꢀ1 hꢀ1 11–13
.
One roadblock lying in the access to high-performance ammonia electrosynthesis
emanates from N2 negative properties of formidable ionization potential (15.84
eV) and zero dipole moment;14 this renders NRR notoriously challenging because
electrocatalyst’s electrons amenable to attain and thereafter break NhN triple
bonds are sparse in this regard.
We took the view that interfacial polarization field, if triggered by extreme
electric field (EF) induced from high-curvature nano-textures to heavily polarized
NhN, could favorably drive electron migration from catalysts to polarized N2
by maneuvering driving force and attenuating interfacial energy barrier.15,16
Principally, the generation of EF is biased onto material morphologies including
heterojunction interfaces and curvature-rich configurations (typically vertex, corner,
or protrusion).16,17
Chem 6, 1–17, April 9, 2020 ª 2020 Elsevier Inc.
1

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
We posited that single-atom catalysts (SACs) isolated onto two-dimensional (2D)
ultrathin support could be an efficient enabler of N2 polarization and reduction.
Conceptually, such architectures have grand potentials to arouse EF because they
intrinsically own heterojunction interfaces and facilitate the formation of interfacial
protrusion-like shapes featuring high curvature.18,19 Structurally, SAC protrusions,
distinct from nanoparticles having slightly fastigiate profiles with low curvature,
are steep; can be anchored with high density because of the ultra-broad surface
of the ultrathin 2D host; and enable full use of every interfacial atom to construct
curvatures, all beneficial to stimulate EF at the maximum limit. Catalytically, SACs
bear prodigious catalytic activities by maximizing atom-utilization efficiency, and
their intrinsic size uniformity keep catalysis along exclusive trajectory, whereas
nanoparticles frequently suffer from size heterogeneity that leads to side reactions
and unsatisfactory product selectivity.20
We focused on transition-metal SACs (TMSACs) supported by atomically thin
molybdenum disulfide (MoS2). This could be rationalized by experiments and
simulations reported: (1) MoS2 are supports pervasively wielded to immobilize
SACs;21 (2) TMSACs bestow MoS2 with peculiar electronic structure of superior
activity for wide-scope catalytic applications;22,23 (3) the profuse electronic
configurations of TMSACs enables activation and elongation of chemical bonds in
adsorbed molecules to propagate interfacial polarization.24
Herein, we report superior NRR electrocatalysts based on protrusion-shaped
Fe SACs immobilized onto monolayer MoS2. Simulations and experiments jointly
show protrusions’ engagement enhances MoS2 EF by three orders of magnitude
and such a field polarizes N2 to trigger interfacial polarization that promotes
electron injection into NhN appreciably. Our Fe SACs showcases 31.6% G 2%
faradic efficiency and 36.1 G 3.6 mmol gꢀ1 hꢀ1 (97.5 G 6 mg hꢀ1 cmꢀ2) production
rate at ꢀ0.2 V versus reversible hydrogen electrode (RHE) in 0.1 M KCl in a flow cell,
comparable to that of most reported NRR electrocatalysts.
RESULTS
Our study began by isolating protrusion-shaped SACs of Fe onto monolayer MoS2
achieved via a facile and scalable hydrothermal grafting route (Figures S1, S2, S3,
S4, and S5). Hydrothermal processing reinforces protrusion-support interfacial
interplay, while the expansive surface area of monolayer MoS2 evades SAC aggrega-
tion and permits high-density dispersion. We term them as SACs-MoS2-Fe-Y, where
Y represents the SAC concentration (atom %).
Then, we sought to identify the protrusion-like shapes of Fe SACs by best-in-class
technologies of microscopes and synchrotron (Figures 1, S6, S7, S8, S9, S10, S11,
and S12). Atomic force microscopy (AFM) images demonstrate the SACs’ studding
increases MoS2 thicknesses by 0.15ꢁ0.25 nm, corresponding to the theoretical
heights of SAC protrusions (Figures 1A, 1F, 1K, and S6–S7). Experimental and
simulated aberration-corrected high-angle annular dark-field scanning transmission
1Key Laboratory of Pesticide & Chemical Biology
of Ministry of Education, Institute of
Environmental & Applied Chemistry, College of
Chemistry, Central China Normal University, 152
Luoyu Road, Wuhan 430079, P.R. China
electron microscopy (HAADF-STEM) images and electron energy loss spectroscopy
(EELS) elemental maps reveal SACs are straddled atop Mo, and their SAC concen-
trations are tunable by the dosage of Fe precursors (Figures 1B–1E, 1G–1J, 1L–
1O, and S8). The percentage of Fe SACs in the whole Fe species of SACs-MoS2-
Fe-2.0 is estimated to be about 98.2% (Figure 1M). The MoS2 support exhibits
a 2H semiconducting phase, a structure well-known to be inert for HER.25 The
dramatically enhanced brightness in SAC sites than in pure Mo sites, together
with side-view HAADF and EELS (Figure S9), directly identify that SACs exist, not
2These authors contributed equally
3Lead Contact
*Correspondence: zhanglz@mail.ccnu.edu.cn
https://doi.org/10.1016/j.chempr.2020.01.013
2
Chem 6, 1–17, April 9, 2020

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
Figure 1. Characterizations of the Protrusion-like Shape of Fe SACs Immobilized onto Monolayer MoS2
(A–Q) 3D topographic AFM (A, F, and K); sub-angstrom-resolution (B, G, and L); and atomic-resolution (C, H, and (M) HAADF-STEM images and EELS
elemental mapping images (D, I, and N); 3D topographic color-coded intensity images (E, J, and O) and EXAFS spectra (Q) of monolayer MoS2 (A–E);
SACs-MoS2-Fe-1.0 (F–J); and SACs-MoS2-Fe-2.0 (K–O). The inset in image (M) is the size distribution of Fe species. Scale bars in images (B), (G), and (L)
˚
represent 2 nm. Scale bars in images (C), (D), (H), (I), (M), and (N) represent 5 A. EDX spectrum of SACs-MoS2-Fe-2.0 (P).
(R–T) Side-view (R), top-view (S), and perspective-view (T) crystalline structures.
See also Figures S1–S12.
as interior dopants substituting for Mo of MoS2 skeletons but as external adatoms
conjugating with MoS2 to engender tridimensional protrusions, a SAC geometry
distinct from extensively reported metal-Nx- and defect-trapped-SACs usually
being planar. In contrast, Fe doping leads to the brightness darker than that of
Mo site (Figure S10). The Mo/S ratios in SACs-MoS2 relative to in MoS2, as calculated
by X-ray photoelectron spectroscopy (XPS) spectra, keep almost unchanged
(Figure S11), again indicating that Fe does not enter into the MoS2 lattices. En-
ergy-dispersive X-ray (EDX) spectrum further confirms the existence of Fe in our
SACs-MoS2-Fe-2.0 sample (Figure 1P). Experimental and simulated extended
X-ray absorption fine structure spectroscopy (EXAFS) reveal that the first shell of
the Fe atom in Fe/MoS2 has a coordination number of three with a mean Fe-S
˚
length of 2.23 A (Figure 1Q). This mechanistically unveils that protrusions are
formed by coordinating SA with three nearest neighboring S of Mo that was below
SAC (Figures 1R–1T and S12).
With protrusion-shaped SACs in hand, we next sought to canvass their electrocata-
lytic NRR using aqueous electrolyte containing inorganic salts under room
temperature and atmospheric pressure (Figures 2, S13, and S14). We detected
and quantified the ammonia product by ion chromatograph (IC) (Figure S15), which
Chem 6, 1–17, April 9, 2020
3

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
Figure 2. Performances of SACs-MoS2-Fe for Electrocatalytic NRR to NH3
(A) Faradic efficiencies and ammonia-evolving rates of SACs-MoS2-Fe-2.0 applied at ꢀ0.1, ꢀ0.2, ꢀ0.3, ꢀ0.4, and ꢀ0.5 V versus RHE in a flow cell. The
error bars represent the standard deviation of 3 independent measurements.
(B–D) The statistical results for the Faradaic efficiencies (B) and NRR rates (C and D) of 20 batches of SACs-MoS2-Fe-2.0 at ꢀ0.2 V versus RHE in flow cell.
(E) Performances of SACs-MoS2-Fe-2.0 under N2-Ar cycling with and without the potential of ꢀ0.2 V versus RHE in flow cell. The error bars represent the
standard deviation of 3 independent measurements.
(F) NMR spectra of the products of NRR catalyzed by SACs-MoS2-Fe-2.0 under Ar, 14N2, and 15N2 in flow cell.
(G) Charge-transfer ionization mass spectrum of the products of NRR catalyzed by SACs-MoS2-Fe-2.0 under 14N2 and 15N2 in flow cell.
(H) Performance comparison of SACs-MoS2-Fe-Y (Y = 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 atom %) at ꢀ0.2 V versus RHE in flow cell. The error bars in image (G)
represent the standard deviation of 3 independent measurements.
(I) Gibbs free energy diagram and optimized geometric structures of various intermediates generated by SACs-MoS2-Fe catalyzed NRR procedures.
(J and K) Arrhenius plots of the NRR partial current monitored (J) and activation energies (K) of SACs-MoS2-Fe-Y at ꢀ0.2 V versus RHE in flow cell.
See also Figures S13–S23.
was further complemented by mass spectrometer and nuclear magnetic resonance
(NMR) (Figures S16 and S17). We sprayed our catalyst onto a carbon gas-diffusion
layer as the working electrode and ran our electrocatalysis in a flow cell. In such
4
Chem 6, 1–17, April 9, 2020

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
electrochemical configuration, the long-distance gas-diffusion layers restrain HER
and localize the fed N2 around gas-solid-liquid interfaces and the mobile electro-
lytes quicken the mass transfer.26 We fed high-purity N2 into the cell (Figure S18).
Before reaching the cell where NRR took place, the fed N2 passed sequentially
through two scrubbing bottles, one filled with 0.1 M HCl and the other with 0.1 M
KOH. The two solutions could eliminate the effect of potential nitrogen-based
impurities. The breadth of applicability of our catalysts is broad, as NRR is operable
in plentiful inorganic electrolytes (Figure S19). These inorganic salts, before use,
were subject to high-temperature treatment to eliminate the possible nitrate or ni-
trite species.27 We achieved optimum selectivity and activity at ꢀ0.2 V versus RHE
in 0.1 M KCl (Figures 2A and S20). The reasons why our NRR worked best
in KCl lie in that K+ is known to promote N2 adsorption most efficiently, Clꢀ has
considerable ion and electron conductivities, acidic medium favors HER, and
alkaline electrolyte throttles proton supply and transfer. Moreover, electrocatalysis
operated at this potential, enjoys a best-level trade-off between N2 and hydrogen
adsorption energies, which maximally promotes NRR at the lowest cost of enhancing
HER (Figure S21). Our NRR started with an onset potential of ꢀ0.01 V versus RHE
(Figure S22). Among our samples, SACs-MoS2-Fe-2.0 with maximum Fe SAC
density exhibited champion performances, with a total current density of 1.38 G
0.12, faradic efficiency of 31.6% G 2%, and NH3-evolving rate of 36.1 G
3.6 mmol gꢀ1 hꢀ1 (97.5 G 6 mg hꢀ1 cmꢀ2), which we calculated from 20 batches of
samples (Figures 2B–2D and S23). These performance metrics exceed those of all
reported MoS2-based NRR electrocatalysts and are comparable to those of most
of the reported NRR electrocatalysts operated in H-type cell (Figure S24 and Table
S1). It is noteworthy that our performance was achieved at 1.38 G 0.12 mA cmꢀ2 cur-
rent density, while the current state-of-the-art NRR electrocatalysts attained
their best-performing activities usually at current densities less than 1 mA cmꢀ2
(Figure S25). We also prepared Au metal, Bi metal, C3N4, and nitrogen-doped
carbon for further comparison. These electrocatalysts, when proceeding in our
flow cell with the same operational parameters, displayed at least 7-fold lower per-
formances than SACs-MoS2-Fe-2.0 (Figure S26). Our SAC protrusions are structur-
ally robust and catalytically durable, as embodied by unobservable architectural
collapse, faint valence-state variation, trifling activity decay, and negligible Fe leach-
ing following 20 h of NRR (Figures S27, S28, S29, S30, and S31).
To validate the measured NH3 that originates from N2 electroreduction, we oper-
ated reactions under N2-Ar cycling in the presence and absence of applied potential
(Figures 2E and 2F). Without potential, we detected no ammonia in both cases of
bubbling N2 and argon, highlighting the necessity of electricity in yielding ammonia.
This observation, along with the fact that the synthetic precursors contain no nitro-
gen-based species, indicates that the working, counter, and reference electrodes,
as well as the reaction cell, do not cause ammonia contamination. With potential
applied, we still observed null ammonia under argon, excluding the contribution
of other possible electrocatalytic reactions. Only via the simultaneous participation
of electricity and N2 could ammonia be monitored. During this electrocatalytic
process, the sum of the NRR and HER faradic efficiencies was around 100% and
no hydrazine was detected (Figure S32), revealing the HER is the only side reaction.
To directly identify N2 as the exclusive nitrogen source, we conducted isotope label-
ing experiments using 15N2 as the fed gas (Figures 2F and 2G). When 14N2 was
bubbled into electrocatalytic cell, we traced 14NH3 with an IC peak at 4.86 min, a
triplet NMR peaks with a spacing of 52 Hz, and a striking m/z = 17 signal in mass
spectrum. In contrast, when utilizing 15N2 as the nitrogen source, the IC peak red-
Chem 6, 1–17, April 9, 2020
5

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
shifted to 4.98 min, the NMR showed doublet peaks with a spacing of 73 Hz, and the
mass spectrum was dominated by m/z = 18 peak, all pointing to 15NH3 signal. Under
non-electrocatalytic condition, bubbling 15N2 into the cell for 10 h generated no
15NH3 or 14N2. These isotope labeling experiments, as well as the above N2-Ar
cycling experiments, confirm the nitrogen source of the detected ammonia is the
fed N2.
Motivated by these superior performances, we sought to clarify the reactive sites.
For MoS2, vacancies and edges have been theoretically and experimentally demon-
strated to be active for HER.22 Our simulations also show these sites are more active
for HER than for NRR (Figures S33–S37). For comparison, we prepared sulfur-va-
cancy-rich MoS2, which exhibited excellent HER but poor NRR (Figure S34). SACs-
MoS2-Fe-2.0 and monolayered MoS2, despite the same amounts of vacancies
and edges, delivered a 240-fold performance difference. These results exclude
the possibility that vacancies and edges play dominant roles in our electrocatalytic
NRR. To directly assess the contribution of SAC, we theoretically simulated the
spatial distribution of N2 adsorption energies along (001) surface with Fe protrusion
as the center (Figure S37). The energy, only at the overhead region of the protrusion,
conveys the lowest value of ꢀ0.7 eV. The values mount dramatically when the
distances between the protrusion and adsorption site increase. The protrusion-can-
tered distribution in the N2 adsorption energy theoretically identifies the Fe
protrusion as the catalytically active center. To seek the experimental support, we
in situ probed Fe EXAFS spectra of SACs-MoS2-Fe-2.0 during NRR at ꢀ0.2 V versus
RHE (Figure S38). Upon electrocatalysis, SACs-MoS2-Fe-2.0 shows an additional
shoulder-like region, which could be assigned to the coordination between Fe pro-
trusion and nitrogen-based intermediates. This assignment could be further sup-
ported by the increase, in the Fe coordination number, from 3 to 4. These results
act as the solid experimental evidences of identifying Fe protrusion as the catalyti-
cally active center. We also examined the possible contribution of Mo and S by in
situ EXAFS and X-ray absorption near edge structure (XANES). The spectra of
SACs-MoS2-Fe-2.0, before and during NRR, show no appreciable changes. These
control experiments indicate Mo and S act as the spectators during the NRR.
We observed that the NRR performances depended closely on the SAC concentra-
tion (Figure 2H), lending further credence to the aforementioned attribution of
Fe SAC as the catalytic centers. For SACs-MoS2-Fe-Y (Y = 0.0, 0.5, 1.0, 1.5, and
2.0), their faradic efficiencies and ammonia production rates are both proportional
to SAC amounts. SACs-MoS2-Fe-2.0 showcases the best performances owing to
its highest SAC concentration (Figures S39, S40, S41, and S42). Overloading Fe to
a density exceeding 2 atom % converts gradually parts of SACs into nanoclusters
(Figure S39), which curtails SAC quantities and hence erodes performances. 3
atom % Fe deposition transforms the entire SACs into nanoclusters (Figure S41),
leading to performances slightly outperforming that of pure MoS2.
We then proceeded to trace the genesis of the SAC protrusion-dependent perfor-
mances. To understand how SACs catalyze NRR, we dissected reaction steps and
dynamics. Simulations indicate a monopolistic NRR trajectory of NhN / N=NH
/ N-NH2 / N-NH3 / N-H / N-H2 / 2NH3, all proceeding atop Fe protrusion
(Figure 2I). This is confirmed by in situ Fourier-transform infrared spectroscopy
(FTIR) with detected intermediates of N-H, NH4+, Fe-NH, and Fe-NH (Figures S43
and S44).28,29 Observation of highest energy barrier linking NhN with N=NH im-
plies the rate-limiting step is not N-Hx formation but NhN cleavage. As unveiled
by activation energy calculated as a function of Fe amount (Figures 2J and 2K),
6
Chem 6, 1–17, April 9, 2020

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
the rate augment can be attributable to the mitigation in energy barrier set for elec-
tron flow from SACs to N2.
Provoked by SAC-mediated decline in NRR interfacial energy barrier, we therefore
pursued whether SACs expedite NRR via interfacial polarization. The theoretical
feasibility is supported by the vast electrostatic potential difference between
SACs and N2 when a sufficient EF was applied (Figures 3A, S45, S46, and S47). Quan-
titative analysis of this polarization field gives a magnitude of 17.92 eV (Figure 3B),
outcompeting N2 ionization potential (15.84 eV). Electrons from SAC protrusions
are driven by this field, 36% to NhN and 60% to the zone between SACs and N2
(Figure 3C). Efficient N2 scission occurs as a consequence of electrons encompassing
and thus attacking N2 cadre. As comparison, for pure and nanocluster-interspersed
MoS2, rare electrons reach N2 owing to their respective curvature-free and curva-
ture-marginal features that generate insufficient polarization.
Bearing the theorem in mind that polarization field arose via polarizing molecule
through EF,30,31 we turned to explore the theoretical possibility that SAC protrusions
induces EF to trigger N2 polarization. We employed a finite-element numerical
simulation to map EF distribution and magnitude.32 Feeble EF appears in MoS2
owing to the planar, curvature-deficient in-plane landform (Figures 3D and S46). In
contrast, Fe SACs, benefiting from their intrinsically heterojunction interfaces,
high-curvature protrusions, and maximum atom-utilization efficiency, evoke con-
spicuous EF surrounding their vertices (Figure 3E). Higher density of protrusions
amplifies EF dispersion (Figure 3F). Consistent with simulations, experiments
from Kelvin probe AFM (KPFM) reveal a 256-fold EF enhancement, conferred by
the engagement of 2 atom % SACs (Figure 3G). With SAC density varying, the
enhancement factors of NRR partial current density (Fpcd) and EF intensity (Fefi
)
share the same trend. At each density, their values are close to each other. Further-
more, as profiled by DE calculations, a stronger EF induces the polarization with
higher magnitude.
To dive deeply into the process of SAC protrusion-triggered polarization, we
conducted ab initio molecular dynamics simulations. Increasing EF shortens Fe-N
bond between protrusion and N2 (Figure 3H). As the Fe-N bond is lengthened,
the interfacial barrier decreases largely to enable more electrons to be transferred
from protrusion to N2 (Figures 3I and 3J). Stronger EF enables shorter Fe-N bond,
lower interfacial barrier, and more electrons available to N2. The magnitude of
the dipole moment depends on the degree of N2 activation, which dictates the
energy required for splitting N2 (Figure 3K). Larger dipole moment indicates longer
NhN bond that leads to lower N2-splitting energy. The energy can be further
reduced by the enhancement in EF. Simultaneous increment in both EF and dipole
moment leads to the acceleration in the decrease of Fe-N length and in the rise of
NhN length (Figures 3L and 3M). In this case, more electrons from protrusion
can reach N2 at a much faster rate, such that high-concentration charge carriers
surround N2 and protrusion. This can lead to the prohibitively inhomogeneous
charge distribution that triggers a powerful interfacial polarization field. When the
dipole moment is fixed, the polarization field intensity scales linearly with the EF
intensity (Figure 3N). When the EF intensity is fixed, the polarization field intensity
scales linearly with the dipole moment magnitude (Figure 3O). The two linear
relationships indicate that the polarization field is co-governed by EF and dipole
moment. EF is the driving force for N2 polarization, while the prerequisite is N2
activation. To confirm this opinion, we utilized atomic layer deposition to synthesize
control sample having Fe SAC protrusions centering within a hexagonal ‘‘S-Mo-S-
Chem 6, 1–17, April 9, 2020
7

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
Figure 3. Origin of the Interfacial Polarization Field Triggered by Fe SAC Protrusions
(A–C) 3D topographic potential distribution images (A), electrostatic potentials (B), and charge density differences (C) of N2 adsorbed onto Fe SAC
protrusion immobilized on MoS2 (A1, B1, and C1), pure MoS2 (A2, B2, and C2), and Fe nanocluster immobilized on MoS2 (A3, B3, and C3). In image (C),
the yellow and blue isosurfaces represent charge accumulation and depletion in the space, respectively.
8
Chem 6, 1–17, April 9, 2020

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
Figure 3. Continued
(D–F) Computed EF (D1, E1, and F1), AFM images (D2, E2, and F2), and experimentally measured EF distribution (D3, E3, and F3) of monolayer MoS2 (D),
SACs-MoS2-Fe-1.0 (E), and SACs-MoS2-Fe-2.0 (F). Scale bars in images D2, E2, and F2 represent 50 nm.
(G) Fpcd and Fefi of SACs-MoS2-Fe as a function of Fe loading amount. Fpcd = JFe/JMoS2, where JFe and JMoS2 are the partial current densities of SACs-
MoS2-Fe and MoS2, respectively. Fefi = IFe/IMoS2, where IFe and IMoS2 are the EF intensities of SACs-MoS2-Fe and MoS2, respectively.
(H) Computed distribution function of the length of Fe-N bond between Fe protrusion and N2, when 0, 2,000, and 4,000 KV mꢀ1 of EFs are applied.
(I) Computed variation trend of the interfacial barrier between Fe protrusion and N2 as a function of the Fe-N bond length, when 0, 2,000, and 4,000 KV
mꢀ1 of EFs are applied.
(J) Computed distribution of electrostatic potentials under 0, 2,000, and 4,000 KV mꢀ1 of EFs. For all of the simulations shown in (H), (I), and (J); the dipole
moment is kept to be 1.8 D.
(K) Computed variation trend of N2 splitting energy as a function of N2 dipole moment, when 0, 2,000, and 4,000 KV mꢀ1 of EFs are applied.
(L) Computed variation trend of the Fe-N bond length as a function of the time after exerting EF.
(M) Computed variation trend of N2 bond length as a function of the time after electrons from Fe protrusions reached N2.
(N) Computed variation trend of polarization field intensity as a function of EF intensity, when the dipole moment is fixed to be 1.8 D.
(O) Computed variation trend of polarization field intensity as a function of dipole moment magnitude, when the EF is fixed to be 4,000 KV mꢀ1
.
See also Figures S45, S46, and S47.
Mo-S-Mo’’ sub-unit (Figure S48). This control sample possesses the same EF with
SACs-MoS2-Fe-2.0 but weaker N2 activation and thus lower performances (Figures
S49 and S50).
To confirm EF could polarize N2, we theoretically mapped the distribution of the
lowest unoccupied molecular orbital (LUMO) of NhN adsorbed onto SACs-MoS2-
Fe when exerting different magnitude of EF (Figures 4A–4C). With EF increasing,
parts of LUMOs transfer from Fe to N2. Also, the difference between negative-
and positive-phase LUMOs is dramatically widened, boding well for N2 polarization
by EF.
We, subsequently, sought to experimentations that are verifiable of the EF-medi-
ated polarization. Under no applied potential, Fe SACs enable blueshift in the
FTIR peak of N2 adsorption by the intrinsic activation ability (Figures 4D and
S51).33 When applied at ꢀ0.2 V versus RHE, further blueshift is observed, indicating
the evident contribution of EF to N2 activation. Increasing EF leads to stronger po-
larization that further lengthens N2 length, as evidenced by the further FTIR peak
blueshift.
Experimental evidence of interfacial polarization comes from the variation in
surface potential when environmental gas was switched from Ar to N2 (Figures 4E,
4F, S52, and S53).34 During this process, N2 is adsorbed, activated, and polarized
over SACs, building a polarization field to abstract electrons from catalyst to
elevate potential (Figure 4G). As a result of the polarization field, the rate V of
electron transfer from catalysts to N2 is enhanced. At peak SAC concentration,
the enhancement time of R reaches up to 1,252. The potential difference (DE)
of SACs-MoS2-Fe-2.0 under N2 and Ar is ca. 528.3 mV, 296 and 56 times those
of monolayer MoS2 and SACs-MoS2-Fe-3.0, respectively (Figure 4H). We observe
protrusion concentration-dependent DE, highlighting protrusions as the origin of
generating the polarization.
Finally, we sought to divulge how the interfacial polarization promoted N2 splitting. We
looked into time-resolved N2 adsorption, activation, and polarization under EF. The
˚
˚
NhN bond length in free N2 is 1.078 A but increases to 1.129 A when N2 is initially ad-
sorbed onto SACs (Figures 4I–4L), signaling the emergence of N2 activation. This in-
crease changes N2 dipole moment from original ‘‘0 D’’ to 0.59 D via bond-associated
charge redistribution and makes 36% of N2 p2p* orbital filled with electrons (Figures
4M–4T, and S54). The electron padding transforms the antibonding orbitals into
Chem 6, 1–17, April 9, 2020
9

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
Figure 4. Evidences for the Interfacial Polarization Field as the Descriptor of the NRR Performance
(A–C) Variation trends observed in the configurations (A and B) and percentages (C) of LUMO of free N2 and of N2 adsorbed onto SACs-MoS2-Fe under
EF of different magnitude. The pink and blue colors represent the negative- and positive-phase LUMOs.
(D) In situ FTIR of N2 adsorption of SACs-MoS2-Fe-2.0, SACs-MoS2-Fe-1.0, and monolayer MoS2 with and without the potential of ꢀ0.2 V versus RHE.
(E) The surface potentials measured under Ar and N2. Scale bars represent 100 nm.
(F) The surface potential differences (DE) as a function of N2-bubbling time (the initial environmental gas is Ar). DE = EN2 - EAr, where EN2 and EAr
represent the potentials measured under N2 and Ar, respectively.
(G and H) R (rate of electron transfer from SAC protrusion to N2) (G) and DE (H) as a function of Fe loading amount.
(I–T) NhN length (I–L), dipole moment (DM) (M), bader charges (N–P), variation trends in percentage (Pp2p*) of electron filling into N2 p2p* antibonding
orbit (Q), densities of states (DOSs) of Fe d and N2 p2p* (T-T), and when N2 is adsorbed, activated, and polarized by Fe SACs. Pp2p* = Aoverlap/Ap2p*
,
where Aoverlap is the overlapped area between the DOSs of Fe d and N2 p2p*, and Ap2p* is the area of DOS of N2 p2p*. DM = | N1 ꢀ N2 | 3 LN2 3 2.4, where
N1 ꢀ N2 represents the bader charge difference between the two nitrogen atoms in N2 and LN2 is the NhN length.
See also Figures S48–S54.
bonding ones, creating appreciable disequilibrium in band alignment.35 This disequilib-
rium renders NhN easily polarized by EF. Then, N2 polarization evolves gradually into a
˚
final, balanced state, where the length is further elongated to 1.182 A, with dipole
moment of 1.76 D and p2p* electron filling percentage of 76%. In this case, N2 splitting
efficiently proceeds via driving more electrons, at a fast rate, from Fe d-orbital to N2 p2p
*
antibonding orbital.
10
Chem 6, 1–17, April 9, 2020

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
To demonstrate the generalizability of the concept of protrusion-triggered polariza-
tion, we adopted the similar synthetic methodology to achieve MoS2-supported
SACs of Co, Cu, Rh, and Ru (Figure S55). These SACs still possess protrusion-like
shapes. They showcase enhanced EF and NRR performances compared with
monolayer MoS2. These results suggested the interfacial polarization could be
a general strategy to boost electrocatalytic NRR to ammonia. Despite these
advances, our flow-cell NRR performances are still limited by the poor N2 solubility
in the electrolytes and by the dissatisfactory interfacial design among the catalysts,
electrolytes, and N2.36
DISCUSSION
In closing, all the above simulations and experiments, taken together, persuasively
evidence that efficient electrocatalytic NRR to NH3 over SACs-MoS2-Fe proceeds
via accelerating and enhancing electron transportation from catalysts to N2,
which is enabled by interfacial polarization. The demonstration herein of SAC-trig-
gered interfacial polarization presents a promising, cost- and energy-effective
avenue breaking NhN, and this concept potentially would inspire innovative
ideas of enabling future ambient NH3 electrosynthesis of industrial interest, prob-
ably via designing stronger polarization. The phenomenon of triggering interfacial
polarization by curvature-rich surface and interface could function as a general
prescription that helps enhance chemistry, catalysis, and material communities’
accessibilities to industrial-level performances achieved in electrocatalytic water
splitting, O2 reduction, and CO2 reduction and beyond.
EXPERIMENTAL PROCEDURES
Methods
Preparation of Flower-like MoS2 Microspheres (Bulk MoS2)
Bulk MoS2 was prepared via
a modified hydrothermal method. 0.05 mM
Na2MoO4$2H2O and 0.75 mM thiourea were prepared, respectively, by dissolving
Na2MoO4$2H2O and thiourea in ultrapure water with resistivity of 18.00 MU$cm un-
der ultrasonication driven by an ultrasonic cell disrupter system. 20 mL of 0.05 mM
Na2MoO4$2H2O was then added drop by drop into 40 mL of 0.75 mM thiourea, and
this addition process lasted for 10 min. The mixture solution underwent vigorous
stirring and ultrasonication driven by an ultrasonic cell disrupter system for 1 h at
room temperature to give a transparent solution. The resulting mixture solution
was then poured into a 100 ml Teflon-lined stainless autoclave. The sealed autoclave
was subjected to heat treatment with an elevated rate of 1ꢂC per min, maintained
initially at 160ꢂC for 6 h and subsequently at 220ꢂC for 18 h both under autogenous
pressure and cooled to room temperature with a rate of 2ꢂC per min, all of which
were conducted under N2. The resulting precipitates were collected, washed
with ethanol and deionized water thoroughly, and finally dried at 50ꢂC under
vacuum. The obtained sample were flower-like MoS2 microspheres, which we term
bulk MoS2.
Preparation of Monolayer MoS2
Monolayer MoS2 was prepared via ultrasonic exfoliation of bulk MoS2 immersed in
the mixture of H2O and isopropanol. The volume ratio of H2O to isopropanol in
the mixed solvent used for liquid exfoliation was 1:4. In detail, 100 mg of bulk
MoS2 was added into 500 mL of the mixture solvent. Then the suspension solution
was sonicated by an ultrasonic cell disrupter system for 18 h under N2. During this
process the temperature of the suspension should be kept no more than 25ꢂC.
The unsuccessfully exfoliated nanosheets were extracted by centrifuging the
Chem 6, 1–17, April 9, 2020
11

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
dispersions at 10,000 rpm for 10 min and subsequently collecting the supernatants.
Centrifugation of the supernatants at 15,000 rpm for 30 min yielded the precipitates
that were MoS2 with monolayer percentage of ca. 62%. The collected precipitates
were further dispersed into 500 mL of the H2O/isopropanol (1:4 vol %) mixture
solvent and treated with the above procedures (including ultrasonication and centri-
fugation) for the second time. The percentage of monolayer MoS2 in the resultant
precipitates was >98%.
Preparation of Protrusion-Shaped Co, Cu, Fe, Rh, and Ru SACs Immobilized onto
Monolayer MoS2
SACs-MoS2-Fe was prepared via a facile, scalable hydrothermal grafting route.
0.15 g of monolayer MoS2, and a given amount of FeCl3 were ground together for
1 h in an agate mortar under condition of pure nitrogen. Then the mixtures were
added slowly into ultrapure water (bubbled with N2), sonicated by an ultrasonic
cell disrupter system for 1 h and continuously stirred for 12 h. The resulting mixture
solution was then poured into a 100 ml Teflon-lined stainless autoclave. The sealed
autoclave was subjected to heat treatment with an elevated rate of 1ꢂC per min,
maintained initially at 160ꢂC for 6 h and subsequently at 180ꢂC for 10 h under
autogenous pressure and cooled to room temperature with a rate of 2ꢂC per min,
all of which were conducted under N2. The resulting precipitates were collected,
washed with ethanol and deionized water thoroughly, and finally dried at 50ꢂC under
vacuum. The obtained samples were protrusion-shaped SACs immobilized onto
monolayer MoS2, which we term SACs-MoS2-Fe-Y, where Y (atom %) presents
the density of SACs. The total Fe content in SACs-MoS2-Fe-Y was measured by
inductively coupled plasma atomic emission spectroscopy (ICP-AES). When 0,
3.00, 6.00, 9.00, 12.00, 15.00, and 18.00 mg of FeCl3 were used, the Fe densities
were determined to be 0.0, 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 atom %, respectively.
For preparation of protrusion-shaped Co, Cu, Rh, and Ru, SACs immobilized
onto monolayer MoS2 we used CoCl2, CuCl2, RhCl3, and RuCl3 as the respective
precursors while keeping other conditions unchanged.
Atomic Layer Deposition-Assisted Preparation of Protrusion-Shaped Fe SACs
Immobilized onto Monolayer MoS2 (ALD-SACs-MoS2-Fe)
ALD-SACs-MoS2-Fe was prepared via atomic layer deposition (ALD, Savannah
S100 Atomic Layer Deposition System, Cambridge NanoTech) technology. This
technology adopted FeCl3 as precursors and N2 as the purging gas and carrier
gas. Monolayer MoS2 were placed in a container inside the ALD reactor chamber.
The deposition took place at the temperature of 200ꢂC. To enable a steady-state
flux of FeCl3 to the reactor the container placing FeCl3 was maintained at 75ꢂC.
Gas lines were set at 100ꢂC to avert the condensation of precursor. Each ALD cycle
proceeded via 1 s of FeCl3 pulse fist and then 30 s N2 purge. Tuning the number of
ALD cycles resulted in precise control in the density of Fe SACs on monolayer MoS2.
ALD-SACs-MoS2-Fe-2.0 were obtained by 10 ALD cycles.
Preparation of monolayer MoS2 with 0.6% (MoS2-SV-1) and 2.1% (MoS2-SV-2)
Sulfur Vacancies
Monolayer MoS2 with sulfur vacancies was prepared by processing monolayer MoS2
via the combination of plasma treatment, chemical etching, and H2 calcination.
10 mg of monolayer MoS2 was first subjected to plasma treatment for 1 h, then
etched by 5% H2O2 under 60ꢂC for 10 min, and subsequently calcined under H2 at
300ꢂC for 1 h. The resulting precipitates were collected, washed with ethanol and
deionized water thoroughly, and finally dried at 50ꢂC under argon. The obtained
sample were monolayer MoS2 with 0.6% sulfur vacancies, which we term MoS2-SV-
12
Chem 6, 1–17, April 9, 2020

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
1. For preparation of monolayer MoS2 with 2.1% sulfur vacancies, which we term
MoS2-SV-2, 10 mg of monolayer MoS2 was first subjected to plasma treatment
for 1 h, then etched by 10% H2O2 under 80ꢂC for 30 min, and subsequently calcined
under H2 at 300ꢂC for 3 h.
Characterization
The powder X-ray diffraction (XRD) patterns of the samples were recorded on a
Bruker D8 Advance diffractometer with monochromatized Cu Ka radiation (l =
0.15418 nm). The morphology of the samples was observed with a JEOL 6700-F
field-emission scanning electron microscope (FESEM) and a JEOL JSM-2010
high-resolution transmission electron microscopy (HRTEM). For the TEM sample
preparation, the powders were suspended in ethanol, then a drop of this suspension
was deposited onto a holey carbon film supported by a copper grid, and then dried
in air prior to the TEM examination. Chemical compositions were analyzed by EDX
attached to the FESEM instrument and XPS (Thermo Scientific ESCLAB 250Xi). All
binding energies were referenced to the C 1s peak (284.6 eV) arising from the
adventitious carbon. Tapping-mode AFM images were conducted using a DI
Innova Multimode SPM platform. During the sample preparation, we diluted the
concentrations of the solutions containing the nanosheets to a very low level. This
renders the nanosheets’ surface as flat as possible so as to avoid the wrinkling or
twisting. Before the AFM measurement, the samples were dried in a vacuum oven
and then subject to plasma treatment, both of which eliminate the surface-adsorbed
solvent molecules. When measuring AFM, we kept the scan rate at 0.2 Hz. Raman
measurements were carried out by a confocal laser micro-Raman spectrometer
(Thermo DXR Microscope, USA). The laser was 633 nm with a 5 mW.
Aberration-Corrected HAADF-STEM and EELS Elemental Mapping
HAADF-STEM images were acquired by a double-aberration-corrected JEOL
JEM-2200FS that operated at 200 keV. STEM has also been conducted by using
detectors including bright field (BF-STEM), medium angle annular dark field
(MAADF-STEM), and HAADF, all of which were implemented in JEOL JEM 2200-
FS equipped with double aberration correctors operating at 200 keV and a C5
corrected Nion UltraSTEM100 operating at 100 keV in UHV conditions (typical
pressures < 2 3 10ꢀ9 Torr). Improvement of image quality and suppression of noise
in the resultant images were enabled by a wide range of signal detection methods
including the collection of a plethora of successive images of the sample, which
were aligned relying on the SDSD plugin for Digital Micrograph to correct for
image drift. EELS elemental mapping has been carried out leveraging Gatan UHV
Enfina spectrometer and concurrently Nion UltraSTEM100. STEM-EELS was con-
ducted using a 1,340-channel detector outfitted with energy resolution of 0.2 eV/
channel that was comparable to the energy spread of the cold FEG electron gun.
Elemental mappings were undertaken by the integration of each point of the EELS
spectrum images through a window of ca. 50 eV above the Mo, S, and Fe EELS
edge onsets, following a background subtraction exploiting a power law model.
The EELS data collected were de-noised by adopting Principle Component Analysis.
Simulation of HAADF-STEM Image
Simulation of HAADF-STEM images of Monolayer MoS2 and SACs-MoS2-Fe pro-
ceeded via quantitative scanning transmission electron microscopy (QSTEM).29
The input parameters were set based on our experimental conditions, including
the probe size, the convergence angle, and the acceptance angle of the detector.
To enhance the contrast of the atoms, the medium-range annular-dark-field (ADF)
mode was selected instead of the high-angle ADF mode by proper adjustment of
Chem 6, 1–17, April 9, 2020
13

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
the camera length. The detector angle is at a pivot condition (with 20–60 mrad)
between the medium-range ADF mode and the high-angle ADF mode.
Electrocatalytic NRR Measurements in a Flow Cell
5 mg of catalysts were ultrasonically dispersed in the mixture of 2.1 mL ultrapure
water, 1.8 mL ethanol, and 0.24 mL Nafion (5 wt %) solution. An airbrush was used
to cast the catalyst ink onto gas diffusion electrode (GDE). The catalyst-coated
GDE was dried under vacuum before use. We determined the amount of catalysts
loaded onto GDE by calculating the difference in the weights of GDE before and
after catalyst loading. The loading amount was kept to be around 0.15 mg cm–2
.
IrO2 coated Ni foam was used as counter electrode. Ag/AgCl (with 3M KCl as the
filling solution) was used as reference electrode. Anion exchange membrane
was employed to separate anode and cathode chambers. 0.1 M KCl was used as
electrolytes to be circulated through both the anode and cathode chambers. The
rate of the electrolyte flow was kept at 40 ml minꢀ1. We use aqueous solution
containing KCl (99.999%, Sigma-Aldrich) as the electrolyte. KCl has a boiling point
of 1,420ꢂC, while the boiling point of KNO3 is 400ꢂC. Before use, KCl was annealed
under argon atmosphere at 500ꢂC for 4 h to eliminate the possible nitrate or
nitrite species. The electrolyte with a volume of 30 mL was circulated through the
electrochemical cell using peristaltic pumps with silicone Shore A50 tubing. 14N2
(or 15N2) gas was delivered into the cathodic compartment at a rate of 30 standard
cubic centimeters per min (s.c.c.m.). 14N2 was purchased from Wuhan Newradar
Special Gas Co., LTD., which provided an official certificate (Figure S18), demon-
strating the product comprises R 99.999 vol % 14N2 and the amount of impurities
[CH4, % 1 ppm (vol); O2, % 3 ppm; H2, % 1 ppm (vol); CO, % 1 ppm (vol); H2O,
% 3 ppm (vol); CO2, % 1 ppm (vol)] are negligible. 15N2 was supplied by Aladdin
with the purity higher than 98.50%. The impurities in the 15N2 are Ar, O2,
and CO2. Before reaching the cell where NRR takes place, 14N2 (or 15N2) passes
sequentially through two scrubbing bottles, one filled with 0.1 M HCl and the other
0.1 M KOH. The two solutions could eliminate the effect of potential nitrogen-based
impurities.
Detection and Quantification of the Produced Ammonia
We used an IC to detect and analyze the ammonia product. The IC generated time-
+
dependent spectra embodying reliable, distinguishable 14NH4
,
15NH4+, and
+
14ND4 characteristic peaks centered at 4.85 min, 4.98 min, and 5.26 min, respec-
tively. The dosage of the solution for acquiring such spectra spans 10 mL to 1 mL.
To quantify the ammonium, we prepared NH4Cl (analytical standard, Sigma-Aldrich)
with different concentrations and thus obtained a linear equation correlating the
ammonium concentration with the area of the peak at 4.85 min. After the reaction,
we mixed together the solution from the cathode chamber and the solution from
the tube used for circulating the catholyte. In this study, we took 20 mL of the solution
out of the after-reaction electrolyte for analysis. Based on the equation and the peak
area measured, we extracted the concentration of the ammonia generated during
our NRR.
We use NMR as a tool to complement IC. For the NMR measurement, the pH of the
after-reaction electrolyte was adjusted to 2 by the addition of HCl. 20% DMSO-d6
and maleic acid with the concentration of 1 mM were employed as the internal stan-
dard. The 1H NMR signal was recorded on a Bruker 400 MHz system.
The faradic efficiency for NRR (FENRR) was defined as the quantity of electric
charge used for synthesizing ammonium divided the total charge passed through
14
Chem 6, 1–17, April 9, 2020

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
the electrodes during the electrolysis. The total amount of ammonium produced was
measured by IC and NMR. Assuming three electrons were needed to produce one
ammonium molecule, the faradic efficiency can be calculated as follows: FENRR
=
(3 3 F 3 cNH4+ 3 V)/(18 3 Q), where F is the Faraday constant, cNH4+ is the measured
ammonium concentration, V is the volume of electrolyte, 18 is the molecular
weight of ammonium, and Q is the total charge passing through the work
electrode. The rate of ammonia formation was calculated using the following
equation: vNH3 = (cNH4+ 3 V)/(t 3 m), where t is the reaction time and m is the
catalyst mass.
For calculation of faradic efficiency for HER (FEHER), we used gas chromatograph
to quantify H2 evolved. Ar was utilized as the carrier gas. During the electrocatalysis,
N2 gas (the flow rate was monitored by mass flow controller) was delivered into
the cathodic compartment containing N2-saturated electrolyte and vented into
GC for analysis. A thermal conductivity detector (TCD) was employed to identify
H2 concentration. The partial current density for HER (JHER) was calculated using
the following equation: JHER = a 3 b 3 (nHER 3 F 3 P0)/(s 3 T) 3 (electrode area)ꢀ1
,
where a is the volume fraction of H2 product determined by GC referenced
to calibration curves from standard gas samples, b is H2 flow rate, nHER is the number
of electrons transferred, P0 = 101.3 kPa, s is the gas constant, and T = 273.15 K.
The HER faradic efficiency was calculated using the following equation: FEHER
=
(JHER/Jtotal) 3 100%, where Jtotal is the total current density passing the work
electrode.
For temperature-dependent measurements, the sealed glass cell was suspended
in a thermostatic silicone oil bath. The electrodes were passed through holes in
the lid with an O-ring slightly larger than the electrodes to minimize evaporative
losses. The effect of temperature on the performance of the catalysts is shown in
Figure 2. As expected, the kinetics of the N2 reduction are increased at elevated
temperatures, reflecting the temperature dependence of the chemical rate constant
which is approximately proportional to exp(ꢀDH*/kT), where DH* is the apparent
enthalpy of activation (hereafter simply termed as the activation energy), and k is
the Boltzmann constant. Specifically, the apparent electrochemical activation en-
ergy (Ea) for NRR can be determined using the Arrhenius relationship
vðlogikÞ
ðvð1=TÞ
DHꢃ
2:3
jn =
where ik is the kinetic current at n = ꢀ200 mV, T is the temperature, and R is the uni-
versal gas constant.
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at https://doi.org/10.1016/j.chempr.
2020.01.013.
ACKNOWLEDGMENTS
This work was financially supported by National Natural Science Funds for Distin-
guished Young Scholars of China (21425728), National Science Foundation of China
(51472100), and the 111 Project (B17019). This work has also benefited from Na-
tional Synchrotron Radiation Laboratory in Beijing and Shanghai for the characteriza-
tion. We thank the National Supercomputer Center in Jinan for providing high per-
formance computation. We thank Y. Pang for the help during the course of study. We
thank L. Cai and J. Shang for the help during the DFT simulation.
Chem 6, 1–17, April 9, 2020
15

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
AUTHOR CONTRIBUTIONS
J.L. and L.Z. supervised the project. J.L., S.C., and F.Q. designed and carried out the
experiments. G.Z. carried out the DFT simulations. F.J. and Z.A. contributed to data
analysis. J.L. and L.Z. wrote the paper. All the authors discussed results and provided
comments during the manuscript preparation.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: November 26, 2019
Revised: December 16, 2019
Accepted: January 17, 2020
Published: February 13, 2020
REFERENCES
1. Guo, C., Ran, J., Vasileff, A., and Qiao, S.Z.
(2018). Rational design of electrocatalysts and
photo(electro)catalysts for nitrogen reduction
to ammonia (NH3) under ambient conditions.
Energy Environ. Sci. 11, 45–56.
theoretical and experimental studies. Adv.
Mater. 30, e1800191.
19. Tao, H., Choi, C., Ding, L.-X., Jiang, Z., Han, Z.,
Jia, M., Fan, Q., Gao, Y., Wang, H., Robertson,
A.W., et al. (2019). Nitrogen fixation by Ru
single-atom electrocatalytic reduction. Chem
5, 204–214.
10. Bao, D., Zhang, Q., Meng, F.L., Zhong, H.X.,
Shi, M.M., Zhang, Y., Yan, J.M., Jiang, Q., and
Zhang, X.B. (2017). Electrochemical reduction
of N2 under ambient conditions for artificial N2
fixation and renewable energy storage using
N2/NH3 cycle. Adv. Mater. 29, 1604799.
2. Geng, Z., Liu, Y., Kong, X., Li, P., Li, K., Liu, Z.,
Du, J., Shu, M., Si, R., and Zeng, J. (2018).
Achieving a Record-High Yield Rate of
120.9 mgNH3 mgcat.ꢀ1 hꢀ1 for N2
20. Han, L., Liu, X., Chen, J., Lin, R., Liu, H., Lu, F.,
¨
Bak, S., Liang, Z., Zhao, S., Stavitski, E., et al.
(2019). Atomically dispersed molybdenum
catalysts for efficient ambient nitrogen fixation.
Angew. Chem. Int. Ed. Engl. 58, 2321–2325.
11. Chen, G.F., Cao, X., Wu, S., Zeng, X., Ding, L.X.,
Zhu, M., and Wang, H. (2017). Ammonia
electrosynthesis with high selectivity under
ambient conditions via a Li+ incorporation
strategy. J. Am. Chem. Soc. 139, 9771–9774.
Electrochemical Reduction over Ru single-
atom catalysts. Adv. Mater. 30, e1803498.
21. Li, H.S., Wang, S.S., Sawada, H., Han, G.G.D.,
Samuels, T., Allen, C.S., Kirkland, A.I.,
Grossman, J.C., and Warner, J.H. (2017).
Atomic structure and dynamics of single
platinum atom interactions with monolayer
MoS2. ACS Nano 11, 3392–3403.
3. Li, S.J., Bao, D., Shi, M.M., Wulan, B.R., Yan,
J.M., and Jiang, Q. (2017). Amorphizing of Au
nanoparticles by CeOx–RGO hybrid support
towards highly efficient electrocatalyst for N2
reduction under ambient conditions. Adv.
Mater. 29, 1700001.
12. Cao, N., Chen, Z., Zang, K., Xu, J., Zhong, J.,
Luo, J., Xu, X., and Zheng, G. (2019). Doping
strain induced bi-Ti3+ pairs for efficient N2
activation and electrocatalytic fixation. Nat.
Commun 10, 2877.
22. Deng, J., Li, H.B., Xiao, J.P., Tu, Y.C., Deng,
D.H., Yang, H.X., Tian, H.F., Li, J.Q., Ren, P.J.,
and Bao, X.H. (2015). Triggering the
4. Zheng, J., Lyu, Y., Qiao, M., Wang, R., Zhou, Y.,
Li, H., Chen, C., Li, Y., Zhou, H., Jiang, S.P., and
Wang, S. (2019). Photoelectrochemical
synthesis of ammonia on the aerophilic-
hydrophilic heterostructure with 37.8%
efficiency. Chem 5, 617–633.
13. Li, X., Li, T., Ma, Y., Wei, Q., Qiu, W., Guo, H.,
Shi, X., Zhang, P., Asiri, A.M., Chen, L., et al.
(2018). Boosted electrocatalytic N2 reduction
to NH3 by defect-rich MoS2 nanoflower. Adv.
Energy Mater. 8, 1801357.
electrocatalytic hydrogen evolution activity of
the inert two-dimensional MoS2 surface via
single-atom metal doping. Energy Environ. Sci.
8, 1594–1601.
5. Qin, Q., Heil, T., Antonietti, M., and Oschatz,
M. (2018). Single-site gold catalysts on
hierarchical N-doped porous noble carbon for
enhanced electrochemical reduction of
nitrogen. Small Methods 2, 1800202.
14. van der Ham, C.J.M., Koper, M.T.M., and
Hetterscheid, D.G.H. (2014). Challenges in
reduction of dinitrogen by proton and electron
transfer. Chem. Soc. Rev. 43, 5183–5191.
23. Liu, G., Robertson, A.W., Li, M.M., Kuo, W.C.H.,
Darby, M.T., Muhieddine, M.H., Lin, Y.C.,
Suenaga, K., Stamatakis, M., Warner, J.H., and
Tsang, S.C.E. (2017). MoS2 monolayer catalyst
doped with isolated Co atoms for the
hydrodeoxygenation reaction. Nat. Chem. 9,
810–816.
15. Kakekhani, A., and Ismail-Beigi, S. (2016).
Polarization-driven catalysis via ferroelectric
oxide surfaces. Phys. Chem. Chem. Phys. 18,
19676–19695.
6. Lin, Y.X., Zhang, S.N., Xue, Z.H., Zhang, J.J., Su,
H., Zhao, T.J., Zhai, G.Y., Li, X.H., Antonietti, M.,
and Chen, J.S. (2019). Boosting selective
nitrogen reduction to ammonia on electron-
deficient copper nanoparticles. Nat. Commun.
10, 4380.
24. Jiang, K., Siahrostami, S., Akey, A.J., Li, Y., Lu,
Z., Lattimer, J., Hu, Y., Stokes, C., Gangishetty,
M., Chen, G., et al. (2017). Transition-metal
single atoms in a graphene shell as active
centers for highly efficient artificial
16. Jiang, H.J., Hou, Z.H., and Luo, Y. (2017).
Unraveling the mechanism for the sharp-tip
enhanced electrocatalytic carbon dioxide
reduction: the kinetics decide. Angew. Chem.
Int. Ed. Engl. 56, 15617–15621.
7. Jin, H., Li, L., Liu, X., Tang, C., Xu, W., Chen, S.,
Song, L., Zheng, Y., and Qiao, S.Z. (2019).
Nitrogen vacancies on 2D layered W2N3: A
stable and efficient active site for nitrogen
reduction reaction. Adv. Mater. 31, 1902709.
photosynthesis. Chem 3, 950–960.
25. Lukowski, M.A., Daniel, A.S., Meng, F.,
Forticaux, A., Li, L.S., and Jin, S. (2013).
Enhanced hydrogen evolution
catalysis from chemically
17. Liu, D., Li, X., Chen, S., Yan, H., Wang, C., Wu,
C., Haleem, Y.A., Duan, S., Lu, J., Ge, B., et al.
(2019). Atomically dispersed platinum
supported on curved carbon supports for
efficient electrocatalytic hydrogen evolution.
Nat. Energy 4, 512–518.
8. Cheng, H., Ding, L.X., Chen, G.F., Zhang, L.,
Xue, J., and Wang, H. (2018). Molybdenum
carbide nanodots enable efficient
electrocatalytic nitrogen fixation under
ambient conditions. Adv. Mater. 30, e1803694.
exfoliated metallic MoS2 nanosheets.
J. Am. Chem. Soc. 135, 10274–
10277.
18. Hui, L., Xue, Y., Yu, H., Liu, Y., Fang, Y., Xing, C.,
Huang, B., and Li, Y. (2019). Highly efficient and
selective generation of ammonia and
9. Zhang, L., Ji, X., Ren, X., Ma, Y., Shi, X., Tian, Z.,
Asiri, A.M., Chen, L., Tang, B., and Sun, X.
(2018). Electrochemical ammonia synthesis via
nitrogen reduction reaction on a MoS2 catalyst:
26. Weekes, D.M., Salvatore, D.A., Reyes, A.,
Huang, A., and Berlinguette, C.P. (2018).
Electrolytic CO2 reduction in a flow cell. Acc.
Chem. Res. 51, 910–918.
hydrogen on a Graphdiyne-based catalyst.
J. Am. Chem. Soc. 141, 10677–10683.
16
Chem 6, 1–17, April 9, 2020

Please cite this article in press as: Li et al., Accelerated Dinitrogen Electroreduction to Ammonia via Interfacial Polarization Triggered by Single-
Atom Protrusions, Chem (2020), https://doi.org/10.1016/j.chempr.2020.01.013
27. Li, L., Tang, C., Yao, D., Zheng, Y., and Qiao,
S.-Z. (2019). Electrochemical nitrogen
reduction: identification and elimination of
contamination in electrolyte. ACS Energy Lett.
4, 2111–2116.
NaCl, H2O and SF6. J. Mol. Struct. 434,
229–237.
complex materials. Acc. Chem. Res. 43,
541–550.
31. Sowlati-Hashjin, S., and Matta, C.F. (2013). The
chemical bond in external electric fields:
energies, geometries, and vibrational Stark
shifts of diatomic molecules. J. Chem. Phys.
139, 144101.
34. Kitano, M., Inoue, Y., Yamazaki, Y., Hayashi, F.,
Kanbara, S., Matsuishi, S., Yokoyama, T., Kim,
S.W., Hara, M., and Hosono, H. (2012).
Ammonia synthesis using a stable electride as
an electron donor and reversible hydrogen
store. Nat. Chem. 4, 934–940.
28. Schimek, G.L., Kolis, J.W., and Long, G.J.
(1997). Metal hexaammine as a bulky cation:
structural and property studies of [M(NH3)6]
Cu8Sb3S13 (M = Mn, Fe, Ni) and [Fe(NH3)6]
AgES4 (E = As, Sb). Chem. Mater. 9, 2776–
2785.
32. Liu, M., Pang, Y., Zhang, B., De Luna, P.,
Voznyy, O., Xu, J., Zheng, X., Dinh, C.T., Fan,
F., Cao, C., et al. (2016). Enhanced
electrocatalytic CO2 reduction via field-
induced reagent concentration. Nature 537,
382–386.
35. Liu, J.C., Ma, X.L., Li, Y., Wang, Y.G., Xiao, H.,
and Li, J. (2018). Heterogeneous Fe3 single-
cluster catalyst for ammonia synthesis via an
associative mechanism. Nat. Commun. 9, 1610.
29. Yao, Y., Zhu, S.Q., Wang, H.J., Li, H., and Shao,
M.H. (2018). A spectroscopic study on the
nitrogen electrochemical reduction reaction on
gold and platinum surfaces. J. Am. Chem. Soc.
140, 1496–1501.
36. Li, A., Cao, Q., Zhou, G., Schmidt, B.V.K.J., Zhu,
W., Yuan, X., Huo, H., Gong, J., and Antonietti,
M. (2019). Three-phase photocatalysis for the
enhanced selectivity and activity of CO2
reduction on a hydrophobic surface. Angew.
Chem. Int. Ed. Engl. 58, 14549–14555.
33. Liscio, A., Palermo, V., and Samorı`, P. (2010).
Nanoscale quantitative measurement of the
potential of charged nanostructures by
electrostatic and Kelvin probe force
30. Delley, B. (1998). Vibrations and dissociation
of molecules in strong electric fields: N2,
microscopy: unraveling electronic processes in
Chem 6, 1–17, April 9, 2020
17