Y. Cui et al. / Electrochimica Acta 228 (2017) 428–435
433
sulphur bonds and the low coordinate sulphur ion at surface [41].
The third peak at 168.2 eV is S-O interaction resulting from air
exposure [42]. Besides, XPS analysis of CNS catalyst exhibited an
elemental composition of Co: Ni: S close to 0.9: 0.1: 2, consistent
with the EDX result.
obtain the exchange current density normalized to the real surface
area (J0,real) of each electrode as shown in Table 1. J0,real value of CNS
ꢀ2
was 12.56 mA cm , almost 4 times more than a value of
ꢀ
2
3.36
m
A cm
2 2 4
presented by CoS in 0.5 M H SO . A higher
electrochemically active surface area in CNS indicated that a rapid
charge transfer between active sites and electrolyte was achieved.
To understand the effect of Ni doping on the catalytic activity of
cobalt for HER, DFT calculations was employed to investigate the
The electrocatalytic HER activities (Fig. 4) of the Co1-xNi
electrocatalysts with different amount of Ni doping deposited on
the glassy carbon electrode were performed in 0.5 M H SO
electrolyte continuously purged with H , as described in the
x 2
S
2
4
2
2
detailed HER pathways and process on CoS and CNS catalysts. The
experimental section. As a control experiment, a commercial Pt
catalyst (10% wt Pt) was also conducted and nearly zero onset
overpotential with high current density was achieved as shown in
Fig. 4a. Compared with the negligible current density from a bare
scheduling of proton in HER process mainly involved three steps
(in acid solution): proton adsorption (Fig. 5a), hydrogen atom
formation (Fig. 5b), H formation and desorption (Fig. 5c), which all
2
process occur on the edge of catalyst [43]. According to previous
reports, hydrogen evolution pathway was depended on the kinetic
energy barrier of the catalysts, including platinum [44]. Mean-
while, the sulfur atom on the edge of TMDs was recognized as HER
active site. Based on the well-known mechanism above, the kinetic
glassy carbon control, all of the Co1-xNi
lower onset overpotentials. Co0.9Ni0.1
value of 55 mV, suggesting the superior HER activity. Notably, the
overpotential required for the Co0.9Ni0.1 sample to produce
cathodic current densities of 10 mA cm is just ꢀ152 mV, which is
over 2 times larger than the current detected in pure CoS nanoball
catalyst and the other two synthesized doped samples (-165 and
198 mV, respectively). Moreover, the corresponding Tafel plots
Fig. 4b) were fitted to the Tafel equation, yielding Tafel slopes of
x
S
2
catalysts possess much
S
2
sample exhibits the lowest
S
2
ꢀ
2
2
energy barrier profiles of hydrogen evolution on CoS and CNS
2
were investigated by DFT (Fig. 5d). From Fig. 5d, the rate-
determining step is thought to be the combination of Hads with H+
ꢀ
2
(Step two). For pure CoS sample, the energy barrier for adsorption
(
3
was calculated to be 0.21 eV. Compared with pure catalyst, the
energy barrier of the doped sample (CNS) was found to be only
0.03 eV. Besides, the hydrogen generated on the surface of CNS
ꢀ1
0, 52, 59, 67, and 92 mV decade
for Pt, Co0.9Ni0.1
2
S (CNS),
Co0.95Ni0.05
S
2
, Co0.8Ni0.2
S
2
and pure CoS , respectively. The lowest
2
ꢀ
1
value (ꢃ 50 mV decade ) observed from CNS indicated a possible
Volmer-Heyrovsky reaction pathway with electrochemical de-
sorption of hydrogen as the rate-limiting step [43].
2
catalyst is much easier to release compared to CoS , which can be
attributed to the larger energy for desorption (Step three, ꢀ2.21 eV
vs. ꢀ2.15 eV). Therefore, based on the simulation results, the higher
catalytic activity achieved from the CNS catalyst than the undoped
To further investigate the electrode kinetics under HER process
of the as-prepared catalysts, electrochemical impedance spectros-
copy (EIS) measurements were performed as illustrated in Fig. 4c.
The inset of Fig. 4c shows equivalent circuit diagram for modeling
the measured impedance curves. Re is associated with the
electrolyte resistance, the first semicircle in the high frequency
region (Rs/CPE1) is associated with the migration and surface film
resistance, and the second semicircle in the moderate frequency
region (Rct/CPE2) is ascribed to the charge transfer resistance. Z
represents the Warburg impedance. The Nyquist plots reveal an
obvious decrease of the Rct from 189.7 (CoS ) to 92.2 (CNS),
W
V
2
V
which can be attributed to the homogeneous distribution of Ni
element. Stability in acid media is another important criterion to
evaluate an advanced electrocatalyst. For CNS sample, after 3000
cyclic voltammetric (CV) sweeps in 0.5 M H SO aqueous solution,
2 4
a slightly larger overpotential is displayed in Fig. 4d. The
remarkable durability indicates that the CNS (10% wt Ni doped
2
CoS ) catalyst is highly stable over the course of the cycling
investigation.
2
The exchange current density (J0,geo) of CoS based catalysts, an
inherent measure of HER activity, was deduced from the Tafel plots
ꢀ
2
(
Fig. 4b) as shown in Table 1. Obviously, the J0,geo of 4.98
m
A cm in
ꢀ2
CNS was higher than the values of 2.02
m
A cm for CoS sample.
2
Electrochemical double-layer capacitances capacitance (Cdl) be-
tween the catalysts and electrolyte can be used to evaluate the
effective surface area of various catalysts by cyclic voltammetry, as
ꢀ
2
shown in Fig. 4c. From Fig. 4d, the Cdl value of 23.2 mF cm for the
CNS was detected, remarkably higher than the undoped CoS (15.3
mF cm ), meanwhile comparable recently reported literature on
Co-based electrocatalysts for hydrogen production [22,42]. More-
over, Since Cdl is proportional to the real surface area, it was used to
Fig. 5. (a,b,c) Schematic reaction pathway of HER on sulfur atom of CNS catalyst
edge in acid environment. The yellow, red, green, and blue spheres in (a–c)
represent S, Ni, Co, and H atoms, respectively. (d) The kinetic energy barrier profiles
2
ꢀ2
2
of HER on CNS and pure CoS catalysts. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article.)
Table 1
2
Summary of the electrochemical properties of CoS and CNS electrodes.
ꢀ2
ꢀ1
ꢀ2
ꢀ2
ꢀ2
Sample
CoS
CNS.
h
(mV vs RHE) for J = –10 mA cm
Rct (
V
)
Tafel (mV decade
)
J
0,geo
(m
A cm
)
C
dl (mF cm
)
J
0,real
(
mA cm
)
2
152
197
189.7
92.2
92
52
2.02
4.98
15.3
23.2
3.36
12.56