National Science Review
RESEARCH ARTICLE
1 Collaborative
Innovation Center of
Chemistry for Energy
Materials (iChEM),
State Key Laboratory
of Catalysis, Dalian
National Laboratory
for Clean Energy,
Dalian Institute of
Chemical Physics,
Chinese Academy of
Sciences, Dalian
116023, China;
2 Department of
Chemistry,
Collaborative
Innovation Center of
Chemistry for Energy
Materials (iChEM),
College of Chemistry
and Chemical
Engineering, Xiamen
University, Xiamen
361005, China;
3 Energy College,
University of Chinese
Academy of Sciences,
Beijing 100049, China
and 4 Department of
Materials, Imperial
College London,
London SW7 2AZ, UK
∗ Corresponding
authors. E-mails:
ftfan@dicp.ac.cn;
canli@dicp.ac.cn
Received 22 March
2023; Revised 9 May
2023; Accepted 5
June 2023
© The Author(s) 2023. Published by Oxford University Press on behalf of China Science Publishing & Media Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original
work is properly cited.
10: nwad166, 2023
https://doi.org/10.1093/nsr/nwad166
Advance access publication 6 June 2023
CHEMISTRY
Special Topic: Green Carbon Science—A Scientific Basis for Carbon Neutrality
Visualizing the role of applied voltage in non-metal
electrocatalysts
Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023
Ziyuan Wang1 ,2 , Jun Chen1 ,3 , Chenwei Ni1 ,3 , Wei Nie1 ,3 , Dongfeng Li1 ,3 , Na Ta1 ,
Deyun Zhang1 ,3 , Yimeng Sun1 ,3 , Fusai Sun1 ,3 , Qian Li1 ,3 , Yuran Li1 ,2 , Ruotian Chen1 ,
Tiankai Bu4 , Fengtao Fan1 , ∗ and Can Li1 , ∗
ABSTRACT
Understanding how applied voltage drives the electrocatalytic reaction at the nanoscale is a fundamental
scientific problem, particularly in non-metallic electrocatalysts, due to their low intrinsic carrier
concentration. Herein, using monolayer molybdenum disulfide (MoS2 ) as a model system of non-metallic
catalyst, the potential drops across the basal plane of MoS2 (Vsem ) and the electric double layer (Vedl )
are decoupled quantitatively as a function of applied voltage through in-situ surface potential microscopy.
We visualize the evolution of the band structure under liquid conditions and clarify the process of EF
keeping moving deep into Ec , revealing the formation process of the electrolyte gating effect. Additionally,
electron transfer (ET) imaging reveals that the basal plane exhibits high ET activity, consistent with the
results of surface potential measurements. The potential-dependent behavior of kf and ns in the ET reaction
are further decoupled based on the measurements of Vsem and Vedl . Comparing the ET and hydrogen
evolution reaction imaging results suggests that the low electrocatalytic activity of the basal plane is mainly
due to the absence of active sites, rather than its electron transfer ability. This study fills an experimental gap
in exploring driving forces for electrocatalysis at the nanoscale and addresses the long-standing issue of the
inability to decouple charge transfer from catalytic processes.
Keywords: non-metal electrocatalyst, surface potential, electron transfer, hydrogen evolution reaction,
mapping
INTRODUCTION regulates the electrochemical process differently for
metal and non-metal catalysts. For metal catalysts
Electrochemical reactions are a key strategy for re-
with a huge density of states (DOS) near the Fermi
alizing carbon neutrality and advancing the utiliza-
levels (EF ), the potential mainly drops within the
tion of clean energy through the production of
electric double layer (EDL), modulating the electro-
green hydrogen via water electrolysis with renewable
static potential (ϕ) and changing the reaction coor-
electricity. Unraveling the electrocatalytic mecha-
dinates directly [1]. While for non-metallic catalysts,
nism is of paramount importance to achieve a com-
due to the low electrical conductivity and complex
prehensive understanding of the process and de-
changes in chemical composition, the potential dis-
velop high-performance catalysts. Therefore, inves-
tribution across the solid–liquid interface becomes
tigating the electrocatalytic mechanism holds signif-
complicated [3–8]. A part of the potential drop may
icant promise for promoting a sustainable future. Ap-
drop across the non-metallic catalysts, while another
plying voltage is an effective method for changing
part of the potential drops across the EDL [6–8]. In
the electrochemical potential of an electrocatalyst
this case, both the chemical potential (μe ) and the
[1,2], thereby determining the direction and rate of
electrostatic potential (ϕ) are changed, resulting in
the reaction. However, the applied voltage (Vappl )
 Natl Sci Rev, 2023, Vol. 10, nwad166
the complex electrocatalytic mechanism. Due to the
wide variety of non-metal electrocatalysts, finding
the universal rules is a great challenge. For example, it
was reported that electrocatalytic oxygen evolution
reaction (OER) is driven by the chemical potential
(μe ) in an IrOx catalyst, rather than its electrostatic
potential [2,9]. While for M–N-doped carbon cata-
lysts, the electrocatalytic reaction is driven predomi-
nantly by the change in electrostatic potential across
the EDL [2,10,11]. If the potential distribution
across the solid–liquid interface can be directly mea-
sured, it will be very convenient for studying the elec-
trocatalytic mechanism of non-metallic catalysts.
Up until now, distinguishing the potential dis-
tribution between the non-metallic catalyst and the
EDL has still relied on complex theoretical calcula-
tions [3,4,12]. By using empirical formulas of capac-
itance, Bediako et al. showed that the potential dis-
tribution between twisted bilayer graphene and EDL
is a function of Vappl [8]. In the field of conventional
electrochemical technology, Mott-Schottky analysis
is a common method for calculating the flat band po-
tential of semiconductors. However, it cannot pro-
vide information on how the band structure moves,
which is related to the interfacial potential distri-
bution. Through I-V curve, the onset potential of
a low-DOS semiconductor at which the EF reaches
the band edge can be obtained [13]. However, the
potential drop across the semiconductor–electrolyte
interface still remains unknown, and hence it is not
possible to obtain the respective shift of Ec and EF .
The value of the total interface capacitance (Ci )
can be obtained by electrochemical impedance spec-
troscopy (EIS). However, the specific capacitance
of the non-metallic electrode and EDL cannot be
directly measured by EIS, which requires calcula-
tions based on the non-metallic structure model and
the double layer (Gouy–Chapman-Stern) model, re-
spectively [6]. In summary, there is a lack of in-situ
techniques for directly measuring the potential dis-
tribution between the non-metallic catalyst and the
EDL. Moreover, conventional electrochemical char-
acterization only provides the ensemble information
for electrode materials, neglecting the spatial hetero-
geneity in the electronic structures of catalysts [14–
20]. Therefore, a spatially resolved in-situ character-
ization technique is needed.
Scanning electrochemical potential microscopy
(SECPM), invented by Bard’s group, has been used
to measure EDL profiles on electrode surfaces,
demonstrating the feasibility of scanning probe tech-
niques for potential measurements in liquid [21].
Boettcher et al. developed a contact-based potential-
sensing electrochemical atomic force microscopy
technique and successfully mapped the photovolt-
age at nanoscopic semiconductor–catalyst interfaces
Page 2 of 10
[22,23]. This shows that changes in surface po-
tential of catalysts can be sensed through contact
mode. Based on these pioneering works, we fo-
cus on the key issues in electrocatalysis, to reveal
the working principle of electrocatalysts through in-
situ surface potential microscopy. Recently, the spa-
tial heterogeneity of local electrostatic potential at
Au nanoplate–solution interfaces was revealed un-
der open circuit potential (OCP) conditions [24].
However, it is still a challenge to figure out how the
electronic structure of non-metallic electrocatalysts
Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023
changes under applied voltage.
Semiconductors have been predicted to be non-
ideal catalysts due to their low intrinsic carrier con-
centration [25]. For example, it is widely believed
that the basal plane of MoS2 impedes charge trans-
fer in electrocatalytic reactions [26–31]. According
to classical electron transfer theories, the Schottky-
analogue junction is broken and becomes ohmic
once the EF is tuned into the bands, resulting in
the low conductivity of the semiconductor in liq-
uid [13,32]. However, it is difficult to explain the
high electrocatalytic activities of some semiconduc-
tor catalysts [13,33,34]. Liu et al. found that the ultra-
thin MoS2 can be modulated to be highly conductive
(‘on’) or insulating (‘off’), strongly correlating with
the hydrogen evolution reaction (HER) [13]. Mor-
purgo’s group also observed liquid-gating-induced
superconductivity in thin MoS2 crystals [35]. The
Y1.75 Co0.25 Ru2 O7−δ electrocatalyst was found to en-
hance the charge transfer step in the oxygen evolu-
tion reaction (OER) [36], while the β-Co(OH)2
electrocatalyst exhibited gradually increasing con-
ductivity after the onset potential of the OER [17].
The above results showed that a charge transport
pathway can be open under solution conditions, re-
flecting the dramatic change of the energy band
structure of the semiconductor. However, the under-
lying reason for the enhanced conductance is not
clear.
MoS2 is a representative material in both the
fields of electrocatalysis and semiconductor devices.
Studying how the voltage affects the band structure
and conductivity of MoS2 under solution conditions
is of great research significance. Herein, using MoS2
as a non-metallic model system, the potential distri-
bution between the basal plane of monolayer (ML)
MoS2 (Vsem ) and the electric double layer (Vedl )
were quantificationally decoupled as a function of
the applied voltage. We visualized how EF moves
into Ec of the basal plane of MoS2 , thereby realiz-
ing the transition of conductivity, which is the ori-
gin of the enhanced conductance in liquid. We pro-
vide a method for studying electrochemical driving
force at the nanoscale and demonstrate that only
when the electron transfer site and the chemical site
 Natl Sci Rev, 2023, Vol. 10, nwad166
Figure 1. Local surface potential measurement on ML MoS2 . (A) Schematic of
in-situ surface potential measurement set-up and potential drops across the electrode–
electrolyte interface for metal, semi-metal and semiconductor respectively. The poten-
tial of the tip (Vtip ) with respect to an Ag/AgCl reference electrode can be recorded
through a high impedance amplifier with negligible leakage current. The potential dis-
tribution between catalyst (Vsem ) and electrolyte (Vedl ) is determined by the series
of quantum capacitor (Cq ) and EDL capacitor (Cedl ), 1/Ctot = 1/Cq + 1/Cedl . (B) Surface
potential values of the basal plane of ML MoS2 , Au and solution were measured with
a tip in 0.1 M K2 SO4 . LSV was performed on the substrate (v = 10 mV/s, sample in-
terval = 0.05 V), while the tip recorded the surface potential relative to the Ag/AgCl
reference electrode.
are spatially coincident can the energy conversion
efficiency be maximized. This work has significant
importance in advancing our understanding of the
electrocatalytic mechanism, and provides theoreti-
cal guidance for the conversion of clean energy and
the achievement of carbon neutrality.
RESULTS AND DISCUSSION
Decoupling the potential drops across
the semiconductor (Vsem ) and the
electric double layer (Vedl )
Schematic illustrations of the home-built in-situ sur-
face potential microscope are shown in Fig. 1A and
Fig. S1. By combining a high impedance amplifier
(1 T) to the atomic force microscopy (AFM) po-
sitioning, the local surface potential Vs of the elec-
trode can be obtained. This is achieved by reading
the potential of the tip (Vtip ) in direct contact with
the electrode surface relative to the Ag/AgCl refer-
ence based on an assumption: the EF of the tip is con-
trolled by the solution and not by the semiconductor
in the ranges of applied voltage (the tip and the semi-
conductor are difficult to equilibrate, so the EF of the
tip does not follow the EF of the semiconductor, and
the potential of the tip (Vtip ) reflects the EDL poten-
tial at the surface of the semiconductor (Vs )).
Page 3 of 10
We take Vs (Vappl = 0) as a reference. The change
of potential drop over the EDL (Vedl ) can be read
directly by Vs (Vappl = 0)−Vs (Vappl = 0). By com-
bining with the change of applied voltage (Vappl ),
the change of potential drop across the electrocat-
alyst (Vsem ) can also be obtained by Vsem =
Vappl −Vedl . To eliminate the effect of redox re-
actions on the tip surface, which affect the potential
measurements, we used a solution that contains only
the supporting electrolytes (0.1 M K2 SO4 ) and stud-
ied the intrinsic properties of ML MoS2 at different
Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023
Vappl .
Due to the small Debye screening length, the
distance between the tip and the sample deter-
mines whether an accurate surface potential can
be measured. To demonstrate the feasibility of sur-
face potential measurement, a metal (Au) substrate
was used as a standard sample. The tip was rested
on the surface with an applied force of ∼29 nN,
slightly larger than that reported by Boettcher et al.
(∼25 nN) [21] but sufficient to prevent damage to
the sample. As shown in Fig. 1B, Vs faithfully tracked
Vappl , reflecting that the surface potential Vs is equal
to the applied voltage. This is consistent with the
common view for bulk metal electrodes [1,37] that
there is no potential drop inside the metal. In this sit-
uation, the surface electron concentration ns can be
considered a constant value and Vappl is applied to
increase the electric field intensity of EDL (Vappl
= Vedl ), as indicated by curve III of Fig. 1A. With
the same applied force (29 nN), the tip can be con-
sidered to be able to contact the surface of other sam-
ples. Then, the tip was lifted off ∼1 mm from the sur-
face and the potential of the solution (background)
was measured (Fig. 1B). It shows that the change in
solution potential (∼12 mV) is negligible because
the thickness of the EDL is less than 1 nm in a 0.1 M
K2 SO4 solution [38]. Therefore, the solution can be
considered a blank background and does not affect
the surface potential measurement.
Next, the surface potential of the basal plane of
ML MoS2 at different Vappl was monitored, as shown
in Fig. 1B. Under OCP conditions (0.14 V), there
was no apparent Vs change when the tip was trans-
ferred from the solution to the surface of MoS2 . The
result reflects that the solution has equilibrated the
EF of the tip and MoS2 before contact and the elec-
tron transfer between the tip and MoS2 can be ig-
nored. The result demonstrates that the introduction
of the nanotip did not change the initial structure of
MoS2 . When Vappl > −0.15 V (vs. Ag/AgCl), there
is no change in Vs . Vs starts to track Vappl when Vappl
< −0.15 V. The change in EDL potential (Vedl )
as a function of applied potential can be obtained
by Vs (Vappl = 0) −Vs (Vappl = 0), as shown in
Fig. 2A (blue circuit). Since a part of the potential
 Natl Sci Rev, 2023, Vol. 10, nwad166

A DOS B
Cq

Energy (eV)
CB
E
CB

-0.15 V CB EF
0.4 0.4
-eΔVappl
II I -eΔVedl
0.2 0.2 EF
Slow Fast

ΔVsem
ΔVedl
0.0 0.0

OCP ΔVappl
-0.2 -0.2

Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023

-0.4 -0.4
DOS
-0.4 -0.2 0.0 0.2 0.4
Potential (V vs. Ag/AgCl) DOS

C i ii iii iv
CB
EF 0.11 eV CB CB CB
D 4.0
-30
0.14 eV
EF
0.29 eV -20
3.0

I (nA)
EF
0.55 eV

n2D (1013 cm-2)

-10
EF
2.0 0
II I
-0.4 -0.2 0.0 0.2 0.4
Eg = 1.85 eV Potential (V vs. Ag/AgCl)
1.0

0.14 eV 0.0
VB
VB VB VB -0.4 -0.2 0.0 0.2 0.4
-0.4 V -0.15 V OCP (0.14 V) 0.4 V Potential (V vs. Ag/AgCl)

Figure 2. Decoupling of potential drops across MoS2 and EDL, and evolution of MoS2 band structure. (A) The decoupled
contribution of each potential drop (Vsem and Vedl ) across the ML MoS2 basal plane–electrolyte interface as a function
of Vappl . Vedl is obtained by Vs (Vappl = 0)−Vs (Vappl = 0), while Vsem can be obtained by Vappl −Vedl . A cartoon of
the DOS of the ML MoS2 basal plane is shown in the top panel of (A). The EF is tuned into the band edge of the conduction
band (Ec ) at −0.15 V, leading to an increase in local density of states at the EF . (B) The change of applied voltage (Vappl )
shifts the EF of the basal plane by −eVappl relative to its initial position. The potential change in EDL (Vedl ) shifts the
full band structure (including the conduction band (Ec ) and the valence band (Ev )) by −eVedl relative to its initial position.
The potential change in the semiconductor (Vsem ) corresponds to the relative movement rate between EF and Ec , shifting
the EF by −eVsem with respect to the Ec , which contributes to the change of surface conductance. (C) Energy diagrams for
the basal plane of ML MoS2 at different Vappl . (D) The ns -Vappl curve shows the switching effect of ns : the surface is turned
‘on’ with a high electron concentration under cathodic potential with an onset potential of approximately −0.15 V, while
remaining insulated under anodic potential. Inset of (D), in-situ electric conductivity measurement on the basal plane of ML
MoS2 in 0.1 M K2 SO4 . The AFM-SECM tip directly contacted the surface of the basal plane and recorded the conductivity
current I (SG/TC mode, lift = 0, ETip = 0 V (vs. Ag/AgCl), LSV was performed on the substrate at v = 10 mV/s).

drop over EDL is obtained, another part of the po-
tential drop across the basal plane of MoS2 can also
be given by Vsem = Vappl −Vedl (red circuit,
Fig. 2A). The results clearly show that when Vappl >
−0.15 V, the total potential is dropped across the ML
MoS2 (Vappl = Vsem ), as shown in Fig. 1A, po-
tential curve, type I, while the surface potential re-
mains constant. When Vappl is below −0.15 V, the
surface potential begins to change, resulting in a par-
tial drop of the potential across the solution (Vappl
= Vsem + Vedl , Fig. 1A, potential curve, type II).
Vedl increases as Vappl becomes more negative. Af-
ter decoupling the potential drops in the basal plane
(Vsem ) and the EDL (Vedl ), the evolution of the
band structure under electrode conditions can be
further revealed.
The evolution of the band structure under
electrode conditions
The evolution of the potential curve type I to type
II reflects a dramatic change in band structure of the
basal plane. As shown in Fig. 2B, the change in ap-
plied voltage (Vappl ) shifts EF of the basal plane by

Page 4 of 10
 Natl Sci Rev, 2023, Vol. 10, nwad166
−eVappl relative to its initial position [4,8], which
is considered as the movement rate of EF in the
energy coordinates. The potential change in EDL
(Vedl ) shifts the full band structure (including the
conduction band (Ec ) and the valence band (Ev ))
by −eVedl relative to its initial position [1], which
can be seen as the movement rate of Ec . The po-
tential change in the semiconductor (Vsem ) corre-
sponds to the relative rate difference between EF and
Ec , shifting EF by −eVsem with respect to Ec , which
contributes to the change in surface conductance
or ns [4,8]. The potential distribution between the
catalyst (Vsem ) and the electrolyte (Vedl ) is de-
termined by the series of semiconductor capacitors
(Csem ) and EDL capacitors (Cedl ) as shown in Fig. 1
A (bottom). Additionally, due to the smaller DOS of
2D semiconductors compared to bulk semiconduc-
tors, 2D semiconductor capacitors function as quan-
tum capacitors (Cq), where 1/Ctot = 1/Cq + 1/Cedl
[56,39].
When Vappl > −0.15 V, Vappl = −EF /e and
Vedl = −Ec /e = 0. Vsem = Vappl , indicating
that Cq  Cedl (that is, the EF is located in the band
gap with zero DOS near the EF ). In this case, Ec is
pinned at a fixed value in the energy coordinates, and
EF moves to the band edge of Ec at the maximum rate
(Fig. 2C, iv, iii and ⅱ). However, the transition from
type I to type II potential drop occurs at an onset
potential of approximately −0.15 V, indicating that
the EF is tuned into the bottom of the Ec at Vappl =
−0.15 V (Fig. 2C, ⅱ). This results in an increase in
Cq in the basal plane of ML MoS2 due to the large
DOS near EF , as shown in the top panel of Fig. 2B.
When Vappl < −0.15 V, Vappl = −EF /e, Vedl
= −Ec /e and Vsem = Vappl −Vedl . The Ec
starts to move. As Vappl becomes more negative, Ec
moves faster. In this situation, the relative rate be-
tween EF and Ec gradually decreases, as shown in
Fig. 2C, i. However, the difference in the movement
rate between EF and Ec still exists, which means that
the Fermi level keeps moving into the conduction
band. These results are inconsistent with the classical
theory that the Schottky-analogue junction is broken
once the EF is tuned into the conduction band, and
it cannot be further tuned deep in bands [32]. The
bandgap Egap = 1.85 eV of ML MoS2 can be deter-
mined from photoluminescence spectroscopy (PL),
depicted in Fig. S10C. Since the energy difference be-
tween EF and Ec determines the surface charge con-
centration (ns ), for 2D semiconductor materials, ns
can be calculated by (see Supplementary Data for
derivation details):
  
EC − EF
NS · = NC · ln 1 + exp − , (1)
kB T
Page 5 of 10
where Nc ∼8.6 × 1012 cm−2 is the effective den-
sity of states in the conduction band [40]. The con-
ductivity of the basal plane is constantly changing
since the relative positions of EF and Ec in the energy
coordinates change with the Vappl . Figure 2D plots
the relationship between ns and Vappl . The ns -Vappl
curve shows the switching effect of ns : the surface
is turned ‘on’ with high electron concentration (over
1 × 1013 e cm−2 in the basal plane of ML MoS2 , con-
sistent with the ‘liquid gating’ effect [13,35,41,42])
at Vappl below −0.15 V, and is turned ‘off’ with low
Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023
electron concentration (insulating) Vappl higher than
−0.15 V. The results show that the surface electron
concentration can be effectively tuned under elec-
trolyte conditions. Through the in-situ surface po-
tential measurement, we visualized the evolution of
the band structure under electrode conditions and
how the high conductivity on the semiconductor
surface occurs.
Moreover, it can be seen from the ns -Vappl curve
that the basal plane of ML MoS2 is nearly insulated
at the open circuit potential (∼0.14 V). It corre-
sponds to the depleted state of ML MoS2 with low
electron concentration in solution (the flat band po-
tential is approximately equal to −0.07 V, estimated
from a macroscopical Mott-Schottky measurement,
Fig. S9).
The switching effect of ns in the basal plane was
further proven by conductivity measurements [42–
44]. An in-situ local electric conductivity measure-
ment was performed, shown in the inset of Fig. 2D.
With the tip directly contacting the surface of the
basal plane (lift = 0, substrate generation/tip col-
lection (SG/TC) mode), a linear sweep voltam-
metry (LSV) curve was collected on the substrate
while the potential of the tip remained constant.
In this way, the conducting current I is recorded at
the solid–liquid interface (containing only the sup-
porting electrolytes to exclude the Faradic reaction).
It is shown that the basal plane of ML MoS2 be-
comes conductive (‘on’) once the Vappl is below the
onset potential of approximately −0.15 V. Other-
wise, the basal plane maintains an insulating state
(‘off’). In previous studies, the onset potential of the
conductivity of low DOS semiconductor materials
(2D WS2 , MoS2 , etc.) was regarded as the poten-
tial at which EF reaches the position of the band
edge of Ec [13,45], proving the accuracy of the sur-
face potential measurement. To date, how the band
structure of the semiconductor is transformed un-
der electrolyte conditions has not been clearly stud-
ied by conventional electrochemical methods. The
difficulty is in distinguishing the potential distribu-
tion between semiconductor and EDL, which still
depends on theoretical calculations. In our study, the
potential drops in the semiconductor (Vsem ) and
 Natl Sci Rev, 2023, Vol. 10, nwad166

Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023

Figure 3. ET imaging on the ML MoS2 electrocatalyst. SG/TC mode was used in this experiment. The Pt tip and substrate were used as dual-working elec-
trodes. Semi-ring graphite and Ag/AgCl (saturated KCl) were used as the counter electrode and reference electrode, respectively. Products ([Ru(NH3 )6 ]2+ )
are generated from the negatively biased substrate and oxidized to reactants ([Ru(NH3 )6 ]3+ ) by a positively biased tip. (A) ET images of ML MoS2 . ET
currents are superimposed on a 3D topography map. Inset, 2D topography map of ML MoS2 . Fixed height h = 25 nm, ESub = −0.5 V, ETip = 0.3 V (vs.
Ag/AgCl). The solution contains 5 mM [Ru(NH3 )6 ]Cl3 and 0.1 M KCl. P++ -Si was used as substrate. (B) Histograms of ET current distributions at different
sites. (C) Values of topography and ET current obtained along the dashed line in the inset of (A). (D) Spatially resolved LSV curves obtained on the edge
and basal plane of ML MoS2 for the [Ru(NH3 )6 ]3+ /[Ru(NH3 )6 ]2+ system (h = 25 nm, ETip = 0.3 V (vs. Ag/AgCl), v = 3 mV/s). We set the cathode current
to be positive and the anodic current to be negative. (E) The potential-dependent behavior of the rate constant kf and the surface electron concentration
ns in the ET reaction are further decoupled based on the measurements of Vsem and Vedl . (F) Schematic of the potential-dependent behavior of kf
and ns in (E).

the EDL (Vedl ) are decoupled by direct measure-
ments. Based on the potential values, the evolution
of band structure can be revealed under electrolyte
conditions.
The role of the applied voltage in
electrocatalytic reactions
Electron transfer (ET) and the subsequent for-
mation and rupture of chemical bonds (catalytic
reaction) are two fundamental processes in elec-
trocatalytic reactions [9]. However, it is difficult
to distinguish their contributions from each other
due to their convoluted nature. It becomes even
more challenging for non-metal electrocatalysts
with spatial heterogeneity in electronic structures
and catalytic centers [46,47].
To study how Vappl acts in an electrocatalytic reac-
tion, first we used outer-sphere redox pairs to decou-
ple the ET process from the electrocatalysis process
and figure out how the Vappl drives the ET process.
Atomic-force-microscope-based scanning electro-
chemical microscopy (AFM-SECM) [24,48,49]
(Fig. S11) was used to map the local electrochemical
activity of ML MoS2 electrocatalysts in SG/TC
mode [50,51]. A higher iTip reflects a higher local
product concentration (that is, higher local electro-
chemical activity) of the electrocatalysts. The outer-
sphere redox pairs with the reaction [Ru(NH3 )6 ]3+
+ e− → [Ru(NH3 )6 ]2+ were used to provide in-
formation about the ET [20]. The ET image clearly
shows that the basal plane of MoS2 has comparable
ET activity to that of the edge sites (Fig. 3A). The
ET currents in the back of the topographic image are
a bit higher than those in the front, due to the change
in local concentration of redox pairs near the surface
during the imaging process (scanning from back to
front in Fig. 3A). Histograms of ET current distri-
butions at different sites are shown in Fig. 3B. The
ET currents taken from the single scan line collected
in a short time are comparable. Figure 3C gives the
topography and ET current values obtained along
the same scan line (the dashed line in Fig. 3A). The
current values of the basal plane and the edge differ
by only 15 pA. The spatially resolved LSV curve

Page 6 of 10
 Natl Sci Rev, 2023, Vol. 10, nwad166
of the basal plane largely matches that of the edge
(Fig. 3D), indicating no significant difference in the
ET rate between basal plane and edges at different
biases. ET imaging shows high ET activity of the
basal plane, contrary to the traditional consensus
under ex-situ conditions [52,53]. It is reasonable to
speculate that the high ET activity may derive from
the aforementioned ‘liquid gating’ effect with high
ns . However, ET currents are jointly decided by both
the ns and rate constant kf (i ∝ ns and kf ) [1], that is
i = F · A · ns · kf · C0 ,
(2)
where F is Faraday constant, A is electrode area,
ns is surface electron concentration related to sur-
face electrical conductivity, and C0 is the reactant
([Ru(NH3 )6 ]Cl3 ) concentration near the surface
of the electrode. Here, the reverse reaction is neg-
ligible. The rate constant kf can be expressed as
k0 ·exp[−αf(Es –E0’ )], in which k0 is the standard
rate constant, Es is surface potentials of the elec-
trode, α is the transfer coefficient, f = RT F
, and E0’
is the formal potential. In a solution containing

5 mM [Ru(NH3 )6 ]3+ , the formal potential E0 , es-
timated as E1/2 of the metal substrate, is −0.14 V for
[Ru(NH3 )6 ]3+/2+ (Fig. S12B), which is close to the
Vappl at which MoS2 becomes semi-metallic. Only
when the voltage is applied to change the ns , rather
than kf for the basal plane, can the ‘liquid gating’ ef-
fect be responsible for the high ET activity. To date,
it is still unclear how much the ns contributes to the
ET rate in non-metal electrocatalysts.
Herein, the electronic structure of the basal
plane measured in supporting electrolytes (without
Faraday currents) was used to predict the factors
affecting the ET activity and give a meaningful refer-

ence. According to kf = k0 ·exp[−αf(Es –E0 )], since
k0 is unknown it is hard to obtain the specific value
of kf . However, kf at different Vappl can be compared
by kf (Es,1 )/kf (Es,2 ) = exp[−αf(Es,1 –Es,2 )]. Taking
kf at Vappl = 0.4 V as the reference point (marked as
kf,0 ), we can obtain kf /kf,0 at different Vappl . Similarly,
we can also have ns /ns,0 at different Vappl (the ref-
erence point ns,0 is ns at Vappl = 0.4 V). Combining
kf /kf,0 and ns /ns,0 , the potential-dependent behavior
of kf and ns can be obtained (Fig. 3E). As shown in
Fig. 3F, when Vappl is more positive than the formal

potential E0 (Vappl > −0.14 V), the electrochemical
potential (or EF ) of MoS2 is lower than that of the
[Ru(NH3 )6 ]3+/2+ redox pair and there is no driving
force to transfer electrons from MoS2 to the solu-
tion (ET reaction does not occur). In this voltage
range, all potential drops within the semiconduc-
tor increase the ns rapidly with a |Slope| ≈ 17 for
log(ns /ns,0 ). There are no potential drops within the
EDL, so the electric field intensity of EDL remains
Page 7 of 10
constant and kf does not change with |Slope| ≈ 0 for
log(kf /kf,0 ). In this voltage range, the effect of the
applied voltage can be considered as a ‘preparation’
for the ET reaction, transforming MoS2 from a semi-
conductor state to a semi-metallic state. When Vappl

is more negative than the formal potential E0 (Vappl
< −0.14 V), the electrochemical potential of MoS2
is higher than the [Ru(NH3 )6 ]3+/2+ redox pair,
creating a driving force for electrons to transfer from
MoS2 to solution, and the ET reaction begins to
occur. In this voltage range, the potential drop within
Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023
the EDL gradually increases. The Vappl begins to act
on the electric field intensity of EDL, changing the
value of kf . When Vappl < −0.3 V, the Vappl increases
kf rapidly with |Slope| ≈ 6 for log(kf /kf,0 ), while
the Vappl increases ns slowly with |Slope| ≈ 1 for
log(ns /ns,0 ). When the ET reaction occurs, the effect
of the applied voltage is to a greater extent to change
the kf value, prompting electrons to pass through
the solid–liquid interface for the ET reaction.
The performance of the ML MoS2 in an elec-
trocatalytic (inner-sphere) reaction was also inves-
tigated at the nanoscale. Figure 4A displays HER
current mapping superimposed on a 3D topogra-
phy map. The result shows that the high tip current
(iTip ) is collected at the edge sites, while the cur-
rent at the basal plane is weak. Histograms of HER
current distribution at different sites are shown in
Fig. 4B. Compared with the outer-sphere reaction,
the inner-sphere reaction is more sensitive to the
chemical properties of the active site. The HER cur-
rents at different edge sites vary greatly, ranging from
a few pA to hundreds of pA. The differences in HER
currents reflect the variation in the chemical proper-
ties of edge sites under in-situ conditions. Figure 4C
gives four current-topography plots along the dashed
lines in Fig. 4A, representing the edge sites with dif-
ferent HER activities. Despite differences in the cur-
rent, the topography of the edge corresponds exactly
to the maximum of the HER current. In addition, the
spatially resolved LSV method was also performed
for ML MoS2 (Fig. 4D). The basal plane is almost
catalytically inert over a wide potential range, while
the electrochemical currents at the edge increase ex-
ponentially with increasing negative bias. In a pre-
vious study, Norskov et al. found that the HER ac-
tivity of MoS2 is proportional to its edge length,
and thus hypothesized that the edges are the active
sites of the HER [15]. Our work provides direct evi-
dence for the structure–activity relationship of MoS2
through in-situ HER imaging. With the outer-sphere
and the inner-sphere reaction, the spatial mismatch
of charge transfer and chemical reaction processes
of ML MoS2 was clearly revealed at the nanoscale.
The applied voltage plays an essential role in the
electron transfer step, increasing the surface electron
 Natl Sci Rev, 2023, Vol. 10, nwad166

Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023

Figure 4. HER imaging on the ML MoS2 electrocatalyst. Products (H2 ) are generated from the negatively biased substrate and oxidized to reactants
(H+ ) by a positively biased tip. (A) HER images of ML MoS2 . HER currents are superimposed on a 3D topography map. Inset, 2D topography map of ML
MoS2 . Fixed height h = 25 nm, ESub = −0.5 V, ETip = 0.3 V (vs. Ag/AgCl). The solution contains 10 mM HClO4 and 0.1 M NaClO4 . p++ -Si were used
as substrate. (B) Histograms of HER current distributions at different sites. (C) Values of topography and HER current obtained along the dashed lines
in the inset of (A). (D) Spatially resolved LSV curves obtained on the edges and basal plane of ML MoS2 for the H+ /H2 system. The tip was biased at
0.3 V (vs. Ag/AgCl) at a fixed height, h = 25 nm, while LSV was performed on the substrate (v = 5 mV/s). We set the cathode current to be positive
and the anodic current to be negative. The dashed line in (D) is the fitting results according to the respective current values. (E) A model of ML MoS2
in the HER. (F) Comparison of the Gibbs free energies of the adsorbed H on the basal plane and edge of MoS2 .

concentration ns and the rate constant kf of the basal
plane. However, the large number of electrons reach-
ing the surface cannot participate in the chemical
reactions (Fig. 4E). The large Gibbs free energy of
the atomic hydrogen (GH* ) adsorbed on the basal
plane (1.96 eV) prevents chemical processes at the
solid–electrolyte interface (Fig. 4F). Therefore, the
low binding energy of H contributes to low HER
rates through the effect of high barriers, resulting in
low energy-conversion efficiency.
essential role in the ET process of the basal plane of
MoS2 , increasing the surface electron concentration
ns and the rate constant kf to different degrees under
different Vappl . However, the applied voltage cannot
be precisely applied to the electrocatalytic reaction
due to the spatial mismatch of charge transfer and re-
actant adsorption sites. This work paves the way for
the rational design of efficient non-metallic electro-
catalysts based on the understanding of how voltage
acts on non-metallic catalysts at the nanoscale.

CONCLUSIONS SUPPLEMENTARY DATA

Through in-situ surface potential measurements, the
potential distribution between the basal plane of ML
MoS2 (Vch ) and electrolyte (Vedl ) was decou-
pled quantitatively as a function of the applied volt-
age. We visualized the evolution of the band struc-
ture under electrode conditions and revealed the
process of conductivity transformation on the semi-
conductor surface. To clarify how the applied voltage
acts on the electrocatalytic reaction, the ET and cat-
alytic reaction processes on ML MoS2 were identi-
fied at the nanoscale through AFM-SECM mapping.
We clearly showed that the applied voltage plays an
Supplementary data are available at NSR online.
FUNDING
This work was supported by the Fundamental Research Funds for
the Central Universities (20720220011), the National Program
on Key Basic Research Project of China (2021YFA150 060 0),
the Fundamental Research Center of Artificial Photosynthesis
(FReCAP), the National Natural Science Foundation of China
(22088102), the Chinese Academy of Sciences Projects for Young
Scientists in Basic Research (YSBR-004) and the Dalian Institute
of Chemical Physics Innovation Foundation (DICPSZ201801).

Page 8 of 10
 Natl Sci Rev, 2023, Vol. 10, nwad166
AUTHOR CONTRIBUTIONS
Ziyuan Wang performed most of the experiments and wrote the
manuscript. Jun Chen and Chenwei Ni performed the theoret-
ical calculations. Wei Nie and Dongfeng Li analyzed the in-situ
surface potential data. Na Ta carried out the scanning transmis-
sion electron microscopy. Deyun Zhang performed scanning elec-
tron microscopy. Yimeng Sun, Ruotian Chen and Tiankai Bu an-
alyzed the in-situ electrochemical imaging data and revised the
manuscript. Fusai Sun, Qian Li and Yuran Li helped with the
in-situ electrochemical imaging measurement. Fengtao Fan and
Can Li proposed and supervised the project.
Conflict of interest statement. None declared.
REFERENCES
1. Bard AJ and Faulkner LR. Electrochemical Methods: Fundamen-
tals and Applications, 2nd edn. New York: Wiley, 2001.
2. Boettcher SW and Surendranath Y. Heterogeneous electrocatal-
ysis goes chemical. Nat Catal 2021; 4: 4–5.
3. Heller I, Kong J and Hendrik A et al. Individual single-walled
carbon nanotubes as nanoelectrodes for electrochemistry. Nano
Lett 2005; 5: 137–42.
4. Heller I, Kong J and Williams KA et al. Electrochemistry at
single-walled carbon nanotubes: the role of band structure and
quantum capacitance. J Am Chem Soc 2006; 128: 7353–9.
5. Li J and Burke PJ. Measurement of the combined quantum and
electrochemical capacitance of a carbon nanotube. Nat Com-
mun 2019; 10: 3598.
6. Li J, Pham PH and Zhou W et al. Carbon-nanotube–electrolyte
interface: quantum and electric double layer capacitance. ACS
Nano 2018; 12: 9763–74.
7. Waelder J, Vasquez R and Liu Y et al. A description of the
faradaic current in cyclic voltammetry of adsorbed redox species
on semiconductor electrodes. J Am Chem Soc 2022; 144: 6410–
9.
8. Yu Y, Zhang K and Parks H et al. Tunable angle-dependent elec-
trochemistry at twisted bilayer graphene with moiré flat bands.
Nat Chem 2022; 14: 267–73.
9. Nong HN, Falling LJ and Bergmann A et al. Key role of chemistry
versus bias in electrocatalytic oxygen evolution. Nature 2020;
587: 408–13.
10. Warburton RE, Hutchison P and Jackson MN et al. Interfa-
cial field-driven proton-coupled electron transfer at graphite-
conjugated organic acids. J Am Chem Soc 2020; 142: 20855–64.
11. Jackson MN and Surendranath Y. Molecular control of het-
erogeneous electrocatalysis through graphite conjugation. Acc
Chem Res 2019; 52: 3432–41.
12. Vasquez R, Waelder J and Liu Y et al. A Gauss’s law analysis
of redox active adsorbates on semiconductor electrodes: the
charging and faradaic currents are not independent. Proc Natl
Acad Sci USA 2022; 119: e2202395119.
13. He Y, He Q and Wang L et al. Self-gating in semiconductor elec-
trocatalysis. Nat Mater 2019; 18: 1098–104.
14. Gao X, Chen Y and Sun T et al. Karst landform-featured mono-
lithic electrode for water electrolysis in neutral media. Energy
Environ Sci 2020; 13: 174–82.
Page 9 of 10
15. Jaramillo TF, Jørgensen KP and Bonde J et al. Identification of
active edge sites for electrochemical H2 evolution from MoS2
nanocatalysts. Science 2007; 317: 100–2.
16. Mariano RG, Kang M and Wahab OJ et al. Microstructural origin
of locally enhanced CO2 electroreduction activity on gold. Nat
Mater 2021; 20: 1000–6.
17. Mefford JT, Akbashev AR and Kang M et al. Correlative
operando microscopy of oxygen evolution electrocatalysts.
Nature 2021; 593: 67–73.
18. Sun T, Wang D and Mirkin MV et al. Direct high-resolution map-
ping of electrocatalytic activity of semi-two-dimensional cata-
Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023
lysts with single-edge sensitivity. Proc Natl Acad Sci USA 2019;
116: 11618–23.
19. Zhu J, Wang ZC and Dai H et al. Boundary activated hydrogen
evolution reaction on monolayer MoS2 . Nat Commun 2019; 10:
1348.
20. Patten HV, Meadows KE and Hutton LA et al. Electrochemi-
cal mapping reveals direct correlation between heterogeneous
electron-transfer kinetics and local density of states in diamond
electrodes. Angew Chem Int Ed 2012; 51: 7002–6.
21. Hurth C, Li C and Bard AJ et al. Direct probing of electrical
double layers by scanning electrochemical potential microscopy.
J Phys Chem C 2007; 111: 4620–7.
22. Laskowski FAL, Oener SZ and Nellist MR et al. Nanoscale semi-
conductor/catalyst interfaces in photoelectrochemistry. Nat
Mater 2020; 19: 69–76.
23. Nellist MR, Laskowski FAL and Qiu J et al. Potential-sensing
electrochemical atomic force microscopy for in operando analy-
sis of water-splitting catalysts and interfaces. Nat Energy 2018;
3: 46–52.
24. Nie W, Zhu Q and Gao Y et al. Visualizing the spatial hetero-
geneity of electron transfer on a metallic nanoplate prism. Nano
Lett 2021; 21: 8901–9.
25. Franklin AD. Electrocatalysis on Non-Metallic Surfaces.
Gaithersburg: NBS Special Pub, 1976.
26. Yu H, Yu X and Chen Y et al. A strategy to synergistically increase
the number of active edge sites and the conductivity of MoS2
nanosheets for hydrogen evolution. Nanoscale 2015; 7: 8731–
8.
27. Voiry D, Salehi M and Silva R et al. Conducting MoS2
nanosheets as catalysts for hydrogen evolution reaction. Nano
Lett 2013; 13: 6222–7.
28. Xie J, Zhang J and Li S et al. Controllable disorder engineering
in oxygen-incorporated MoS2 ultrathin nanosheets for efficient
hydrogen evolution. J Am Chem Soc 2013; 135: 17881–8.
29. Zhang H, Yu L and Chen T et al. Surface modulation of hier-
archical MoS2 nanosheets by Ni single atoms for enhanced
electrocatalytic hydrogen evolution. Adv Funct Mater 2018; 28:
1807086.
30. Yousaf AB, Imran M and Farooq M et al. Synergistic effect of
Co–Ni co-bridging with MoS2 nanosheets for enhanced electro-
catalytic hydrogen evolution reactions. RSC Adv 2018; 8: 3374–
80.
31. Meng X, Yu L and Ma C et al. Three-dimensionally hierarchi-
cal MoS2 /graphene architecture for high-performance hydrogen
evolution reaction. Nano Energy 2019; 61: 611–6.
 Natl Sci Rev, 2023, Vol. 10, nwad166

32. Neamen DA. Semiconductor Physics and Devices: Basic Principles. New Delhi:
Tata McGraw Hill Publishing, 1992.
33. Burke MS, Kast MG and Trotochaud L et al. Cobalt–iron (oxy)hydroxide oxygen
evolution electrocatalysts: the role of structure and composition on activity,
stability, and mechanism. J Am Chem Soc 2015; 137: 3638–48.
34. Voiry D, Yang J and Chhowalla M et al. Recent strategies for improving the cat-
alytic activity of 2D TMD nanosheets toward the hydrogen evolution reaction.
Adv Mater 2016; 28: 6197–206.
35. Costanzo D, Jo S and Berger H et al. Gate-induced superconductivity in atomi-
cally thin MoS2 crystals. Nat Nanotechnol 2016; 11: 339–44.
36. Han N, Feng S and Liang Y et al. Achieving efficient electrocatalytic oxygen
44. Wadhwa P, Seol G and Petterson MK et al. Electrolyte-induced inversion layer
schottky junction solar cells. Nano Lett 2011; 11: 2419–23.
45. Braga D, Gutiérrez Lezama I and Berger H et al. Quantitative determination of
the band gap of WS2 with ambipolar ionic liquid-gated transistors. Nano Lett
2012; 12: 5218–23.
46. Li M, Ye KH and Qiu W et al. Heterogeneity between and within single hematite
nanorods as electrocatalysts for oxygen evolution reaction. J Am Chem Soc
2022; 144: 5247–52.
47. Mariano RG, McKelvey K and White HS et al. Selective increase in CO2 elec-
troreduction activity at grain-boundary surface terminations. Science 2017;
358: 1187–92.

Downloaded from https://academic.oup.com/nsr/article/10/9/nwad166/7191275 by guest on 25 November 2023

evolution in acidic media on yttrium ruthenate pyrochlore through cobalt incor-
poration. Adv Funct Mater 2023; 33: 2208399.
37. Chen S, Liu Y and Chen J et al. Heterogeneous electron transfer at nanoscopic
electrodes: importance of electronic structures and electric double layers.
Chem Soc Rev 2014; 43: 5372–86.
38. Scholz F. Electroanalytical Methods. Berlin: Springer, 2010.
39. Xia J, Chen F and Tao N et al. Measurement of the quantum capacitance of
graphene. Nat Nanotechnol 2009; 4: 505–9.
40. Furchi MM, Polyushkin DK and Pospischil A et al. Mechanisms of photocon-
ductivity in atomically thin MoS2 . Nano Lett 2014; 14: 6165–70.
41. Liu X, Li B and Li X et al. The critical role of electrolyte gating on the hydrogen
evolution performance of monolayer MoS2 . Nano Lett 2019; 19: 8118–24.
42. Saito Y, Nakamura Y and Bahramy MS et al. Superconductivity protected by
spin–valley locking in ion-gated MoS2 . Nat Phys 2016; 12: 144–9.
43. Garrido JA, Härtl A and Dankerl M et al. The surface conductivity at the dia-
mond/aqueous electrolyte interface. J Am Chem Soc 2008; 130: 4177–81.
48. Du HY, Huang YF and Wong D et al. Nanoscale redox mapping at the MoS2 -
liquid interface. Nat Commun 2021; 12: 1321.
49. Jiang J, Huang Z and Xiang C et al. Nanoelectrical and nanoelectrochem-
ical imaging of Pt/P-Si and Pt/p+ -Si electrodes. ChemSusChem 2017; 10:
4657–63.
50. Polcari D, Dauphin-Ducharme P and Mauzeroll J et al. Scanning electrochemi-
cal microscopy: a comprehensive review of experimental parameters from 1989
to 2015. Chem Rev 2016; 116: 13234–78.
51. Bollinger MV, Lauritsen JV and Jacobsen KW et al. One-dimensional metallic
edge states in MoS2 . Phys Rev Lett 2001; 87: 196803.
52. Wu D, Li X and Wu X et al. Uncovering edge states and electrical inhomo-
geneity in MoS2 field-effect transistors. Proc Natl Acad Sci USA 2016; 113:
8583–8.
53. Zhang C, Johnson A and Hsu C-L et al. Direct imaging of band profile in single
layer MoS2 on graphite: quasiparticle energy gap, metallic edge states, and
edge band bending. Nano Lett 2014; 14: 2443–7.

© The Author(s) 2023. Published by Oxford University Press on behalf of China Science Publishing & Media Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original
work is properly cited.

Page 10 of 10