Hydrogen evolution across nano-Schottky junctions at carbon supported MoS2 catalysts in biphasic liquid systems

Article abstract of DOI:10.1039/c2cc31398g

The activities of a series of MoS₂-based hydrogen evolution catalysts were studied by biphasic reactions monitored by UV/Vis spectroscopy. Carbon supported MoS₂ catalysts performed best due to an abundance of catalytic edge sites and strong electronic coupling of catalyst to support.

Full text of DOI:10.1039/c2cc31398g

View Online / Journal Homepage
Dynamic Article Links
ChemComm
Cite this: DOI: 10.1039/c2cc31398g
www.rsc.org/chemcomm
COMMUNICATION
Hydrogen evolution across nano-Schottky junctions at carbon supported
MoS2 catalysts in biphasic liquid systemsw
a
a
b
c
d
´
Peiyu Ge, Micheal D. Scanlon, Pekka Peljo, Xiaojun Bian, Heron Vubrel,
e
e
f
d
b
¨
Arlene O’Neill, Jonathan N. Coleman, Marco Cantoni, Xile Hu, Kyosti Kontturi,
BaoHong Liuc and Hubert H. Girault*a
Received 24th February 2012, Accepted 11th May 2012
DOI: 10.1039/c2cc31398g
The activities of a series of MoS2-based hydrogen evolution
catalysts were studied by biphasic reactions monitored by
UV/Vis spectroscopy. Carbon supported MoS2 catalysts
performed best due to an abundance of catalytic edge sites and
strong electronic coupling of catalyst to support.
However, in the absence of either substrate, identical experimental
procedures to the original syntheses produced much larger
coalesced MoS2 particles. The lower overpotentials for the
HER, for both electrocatalysts, were attributed to the intimate
interconnections between the conductive substrates and MoS2
NPs. This interaction allowed the formation of smaller and more
disperse MoS2 NPs, thereby unleashing an abundance of acces-
sible catalytic edge sites, and ensuring a robust electrical coupling
of MoS2 to the underlying carbon supports with rapid electron
transfer from the less-conducting MoS2 NPs to the electrode.3,4
The interface between two immiscible electrolyte solutions
(ITIES) represents a soft platform to develop new water splitting,
i.e. artificial photosynthetic, protocols.5 Protons can be reduced
at these defect free soft electrified interfaces to produce H2 in the
presence of organic reducing agents such as cobaltocene6 or
decamethylferrocene (DMFc)5,7 in the organic phase. Previously,
we have catalyzed this process by the in situ reduction of metallic
salts, forming adsorbed metallic NPs of Pt or Pd,7e or by
adsorbing MoS2 particles at the ITIES.7d
Molybdenum disulphide (MoS2) has been rigorously charac-
terized by experimental and computational means as an
alternative electrocatalyst to Pt-group metals towards the
Hydrogen Evolution Reaction (HER, i.e. 2H+ + 2eꢀ - H2).
MoS2 reduces protons at low overpotentials via under
co-ordinated sulphur edge sites, while its basal planes remain
catalytically inert.1 The primary factors that influence electro-
catalytic activity are catalyst morphology and electrical
conductivity. Nanocrystalline MoS2 is much more active than
catalytically inert bulk MoS2 due to the exposure of more
reactive sulphur edge sites.1 In turn, exfoliated MoS2, i.e. two-
dimensional MoS2 nanosheets,2 should also show enhanced
catalytic activity due to a further increase in these sites.
Recently, the syntheses of MoS2 on reduced graphene oxide
sheets and MoS2 on mesoporous carbon have been achieved
by the groups of Dai3 and Liu,4 respectively. Strong chemical
coupling and interactions between Mo precursors and func-
tional groups on graphene oxide or mesoporous carbon allowed
the nucleation and selective growth of highly disperse MoS2
nanoparticles (NPs), free of aggregation, on both substrates.
Herein, the relative abilities of nanocrystalline MoS2,
exfoliated MoS2, and MoS2 NPs grown on either graphene
or mesoporous carbon particles were investigated as hydrogen
evolution catalysts (HECs) at the ITIES (for synthesis and
characterisation details of each catalyst see Section S1, ESIw).
The lipophilic tetrakis (pentafluorophenyl)borate anion (TBꢀ)
in the aqueous phase (w) acts both as a proton pump, leading
to the extraction of HTB acid to the organic phase (o),
1,2-dichloroethane (1,2-DCE), and as a potential determining
ion to polarise the liquid–liquid interface.5
a Laboratoire d’Electrochimie Physique et Analytique (LEPA),
´
´
Ecole Polytechnique Federale de Lausanne (EPFL), Station 6,
CH-1015 Lausanne, Switzerland. E-mail: Hubert.Girault@epfl.ch;
Fax: +41 693 3667; Tel: +41 693 3145
H+,w + TBꢀ,w - HTBo
(1)
b Department of Chemistry, Aalto University, P.O. Box 16100, 00076,
Finland
The reaction between DMFc and aqueous protons catalyzed
by MoS2-based particles can be written as a proton transfer
reaction followed by its reduction in the organic phase7a,d
c Department of Chemistry, Institute of Biomedical Sciences,
Fudan University, Shanghai, 200433, P.R. China
d Laboratory of Inorganic Synthesis and Catalysis (LISC), EPFL,
BCH-3305, CH-1015 Lausanne, Switzerland
catalyst in 1;2ꢀDCE
e School of Physics and Centre for Research on Adaptive
Nanostructures and Nanodevices (CRANN), Trinity College Dublin,
D2, Ireland
DMFco þ Hþ;o
DMFcþ;o þ H2
ð2Þ
1
2
ƒƒƒƒƒƒƒƒƒƒ!
interfacial catalyst
or as an interfacial reaction
f Interdisciplinary Centre for Electron Microscopy (CIME), EPFL,
Station 12, CH-1015, Lausanne, Switzerland
DMFco þ Hþ;w
The concentrations of DMFc+ formed during biphasic reac-
tions in the presence and absence of different MoS2-based
DMFcþ;o þ H2
ð3Þ
1
2
ƒƒƒƒƒƒƒƒ!
w Electronic supplementary information (ESI) available: Synthesis and
characterisation of each MoS2-based catalyst; experimental details, dis-
cussion and figures for UV/Vis, zeta (z)-potential, cyclic voltammetry
experiments; MoS2 band calculations. See DOI: 10.1039/c2cc31398g
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun.

View Online
Fig. 1 Kinetics of the hydrogen evolution reaction (HER), in the
absence and presence of each of the MoS2-based catalysts, as followed
by monitoring changes in the UV/Vis absorbance (lmax = 779 nm) for
DMFc+. Fits (dotted lines) were obtained from the calculated rate
constants (k/sꢀ1, see Table 1). On the left, a photograph of the two-
phase reaction after 30 minutes stirring in an UV/Vis quartz cuvette
sealed with a Teflon cap. The clear aqueous phase contains 10 mM
LiTB and 100 mM HCl with 0.25 mM of MoS2 particles visible at the
interface; the green 1,2-DCE phase contains 2.5 mM DMFc+ post
reaction.
Fig. 2 Zeta (z)-potentials (mV) of nanocrystalline MoS2, exfoliated
MoS2, MoS2 on graphene and MoS2 on mesoporous carbon as a function
of pH in aqueous dispersions at concentrations of 0.7–1 mg per 5 ml.
charge of a particle for real systems is not possible. Instead,
the zeta (z)-potential, i.e. the electrostatic potential difference
between an average point at the hydrodynamic shear plane and
the bulk liquid, may be probed and represents a measure of the
surface charge.8 Herein, the z-potential represents a useful
qualitative probe to verify the relative abundances of sulphur
edge sites for each electrocatalyst. The z-potential of each of the
MoS2-based catalysts was analysed using a Nano ZS Zetasizer
(Malvern Instruments, U.K.) with irradiation (l = 633 nm)
from a He–Ne laser (Fig. 2, for experimental details and
additional discussion see Section S4, ESIw). Substantial negative
surface charges at pHs o 7 and shifts of the isoelectric points
(pIs) below pH 2 were observed, on moving from nano-
crystalline (pI E 3.1)9 to exfoliated MoS2 and on modifying
the surfaces of graphene and mesoporous carbon particles
with disperse islands of MoS2 (for z-potentials of the control
materials, see Fig. S5, ESIw). More negative z-potentials over
the pH range 2.5 to 11 for MoS2 NPs grown on mesoporous
carbon rather than on graphene particles infer the presence of
more MoS2 sulphur edge sites on mesoporous carbon, correlating
with its superior rate of reaction as seen in Fig. 1.
catalysts (0.25 mM) were monitored by observing changes of
the UV/Vis absorbance at 779 nm (Fig. 1, for experimental
details see Section S2, ESIw). The maximum stoichiometric
amount of DMFc+ (2.5 mM) limited by the initial DMFc
concentration (2.5 mM) was attained after 1800 s in the
presence of each catalyst. Slower kinetics in the absence of
catalyst preclude the reaction from proceeding to completion
during this time-frame. The reaction was found to be 1st order
with respect to DMFc (Fig. S2, ESIw). The rate constants
(Table 1) were calculated by fitting (Section S2, ESIw).
MoS2 on mesoporous carbon particles were found to be the
best catalyst for H2 evolution, increasing the rate of reaction
by a factor of 170. MoS2 on graphene particles showed a
40 times increase in reaction rate, and exfoliated MoS2 was
slightly more effective compared to nanocrystalline MoS2.
These trends in the relative performances of each of the
MoS2-based particles to act as HECs were further corroborated
by cyclic voltammetry (CV) studies at a potentiostatically
polarised ITIES. For experimental details and a thorough
discussion of the ion-transfer events refer to Section S3 (ESIw).
An increased abundance of exposed negatively charged
sulphur edge atoms, as discussed, is expected on moving from
nanocrystalline to exfoliated MoS2 and on modifying the
surfaces of graphene and mesoporous carbon particles with
disperse islands of MoS2. Directly measuring the surface
A reaction scheme outlining the biphasic mechanism of
proton reduction by DMFc in 1,2-DCE under anaerobic
conditions, with carbon supported MoS2 catalysts present
either at the interface or in the bulk, is illustrated in
Scheme 1 (see Section S5 (ESIw) for band calculations, redox
potentials, etc.). An interfacial Galvani potential difference
(Dwo f) is established due to the distribution of the lipophilic
anion, TBꢀ. Consequently, protons are pumped across the
interface and, in the presence of an MoS2 species, are adsorbed
at catalytic sulphur edge sites (denoted here as *). When the
semiconductor MoS2 is in contact with graphene (or, indeed,
mesoporous carbon) an insulated nano-Schottky junction is
formed, as highlighted in Scheme 1. For MoS2 on graphene
particles, electrons are transferred from graphene to the semi-
conductor until the Fermi levels are equalised. On transfer of
Table 1 Calculated apparent rate constants (k/sꢀ1) for the biphasic
reaction, and the rate constant relative to the experiment without any
catalyst (k/kno catalyst
)
Catalyst
k/sꢀ1
k/kno catalyst
No catalyst
0.00021
0.0030
0.0039
0.0080
0.036
1
14
19
38
Nanocrystalline MoS2
Exfoliated MoS2
MoS2 on graphene
MoS2 on mesoporous carbon
171
c
Chem. Commun.
This journal is The Royal Society of Chemistry 2012

View Online
as many multilayers clumped together as particles and not as
individual single-layer flakes. The conductive carbon supports
increase the HER rate in multiple ways. Firstly, these supports
act as nucleation sites for the formation of small and highly
dispersed MoS2 NPs, thus massively increasing the abundance
of catalytic edge sites in comparison to nanocrystalline MoS2
freely grown in solution.3,4 Secondly, their conductive nature
allow electron injection from DMFc to occur at any point on
the hybrid catalyst, i.e. not specifically at the MoS2 edge sites,
increasing the cross-section of reaction between DMFc and
the catalyst and hence also the reaction rate. The sizable
increase in rate for MoS2 on mesoporous carbon particles
over its graphene analogue suggests that the very large specific
surface area of mesoporous carbon allows huge loadings of small,
highly dispersed MoS2 NPs (hence more catalytic edge sites) than
possible on the surface of comparatively flat graphene.
In conclusion, MoS2 NPs nucleated and grown on carbon
supports, in particular mesoporous carbon, act as superior
electrocatalysts towards the HER. Their catalysis of the
biphasic reduction of protons at the liquid–liquid interface
offers new opportunities in energy research towards the
development of artificial photosynthetic systems.
Scheme 1 Biphasic mechanism of proton reduction to molecular
hydrogen in the presence of MoS2 on graphene catalyst. Dwo f is the
interfacial Galvani potential difference established by distribution of
the lipophilic anion, TBꢀ. VB and CB are the valence and conduction
bands of MoS2, respectively, Ef is the Fermi level of MoS2 and F is the
work function of graphene. The aqueous and organic phases are
coloured blue and red, respectively.
This work was financially supported by the SNF program
‘‘Solar fuels’’. The work in LSCI is supported by a ERC
starting grant under the European Community’s 7th Frame-
work Programme (FP7 2007-2013)/ERC Grant agreement no.
257096. P.P. and K.K. acknowledge financial support from the
Academy of Finland (Grant No. 133261).
an electron from graphene to MoS2, the Fermi level of
graphene is lowered. Inversely, since MoS2 becomes negatively
charged, the Fermi level moves up. Band bending occurs as a
result of the excess negative charge transferring from graphene
to MoS2. DMFc is insoluble in the aqueous phase and electron
injection can be considered to take place exclusively in
1,2-DCE. Electrons are injected from DMFc to a state in
graphene close to the Fermi level. Next, the electrons transfer
to an empty state in the conduction band of MoS2 and, finally,
electron transfer occurs to an adsorbed hydrogen atom at a
catalytic edge site (H*). This results in the release of molecular
hydrogen through one of the two processes: either 2H* -
H2 + 2* or H+ + eꢀ + H* - H2 + *.1a From a
thermodynamic point of view, only protons with sufficiently
positive reduction potentials may be reduced. Protons present
in the organic phase meet this criteria with a reduction
potential of 0.55 V.7a Energetically, aqueous protons may also
be reduced in the presence of a positive interfacial Galvani
potential difference greater than 0.55 V.
Notes and references
1 (a) B. Hinnemann, P. G. Moses, J. Bonde, K. P. Jørgensen,
J. H. Nielsen, S. Horch, I. Chorkendorff and J. K. Nørskov,
J. Am. Chem. Soc., 2005, 127, 5308–5309; (b) T. F. Jaramillo,
K. P. Jørgensen, J. Bonde, J. H. Nielsen, S. Horch and
I. Chorkendorff, Science, 2007, 317, 100–102; (c) J. Bonde,
P. G. Moses, T. F. Jaramillo, J. K. Norskov and I. Chorkendorff,
Faraday Discuss., 2009, 140, 219–231; (d) D. Merki, S. Fierro,
H. Vrubel and X. Hu, Chem. Sci., 2011, 2, 1262–1267.
2 J. N. Coleman, M. Lotya, A. O’Neill and S. D. Bergin, et al.,
Science, 2011, 331, 568–571.
3 Y. Li, H. Wang, L. Xie, Y. Liang, G. Hong and H. Dai, J. Am.
Chem. Soc., 2011, 133, 7296–7299.
4 X. Bian, J. Zhu, L. Liao, M. D. Scanlon, P. Y. Ge, C. Ji,
H. H. Girault and B. H. Liu, 2012, in preparation.
5 M. A. Mendez, R. Partovi-Nia, I. Hatay, B. Su, P. Ge, A. Olaya,
N. Younan, M. Hojeij and H. H. Girault, Phys. Chem. Chem. Phys.,
2010, 12, 15163–15171.
6 U. Koelle, P. P. Infelta and M. Graetzel, Inorg. Chem., 1988, 27,
879–883.
7 (a) I. Hatay, B. Su, F. Li, R. Partovi-Nia, H. Vrubel, X. Hu, M. Ersoz
and H. H. Girault, Angew. Chem., Int. Ed., 2009, 48, 5139–5142;
(b) B. Su, I. Hatay, P. Y. Ge, M. Mendez, C. Corminboeuf, Z. Samec,
M. Ersoz and H. H. Girault, Chem. Commun., 2010, 46, 2918–2919;
(c) B. Su, I. Hatay, F. Li, R. Partovi-Nia, M. A. Mendez, Z. Samec,
M. Ersoz and H. H. Girault, J. Electroanal. Chem., 2010, 639,
102–108; (d) I. Hatay, P. Y. Ge, H. Vrubel, X. Hu and
H. H. Girault, Energy Environ. Sci., 2011, 4, 4246–4251;
(e) J. J. Nieminen, I. Hatay, P. Ge, M. A. Mendez, L. Murtomaki
and H. H. Girault, Chem. Commun., 2011, 47, 5548–5550.
8 A. V. Delgado, F. Gonzalez-Caballero, R. J. Hunter, L. K. Koopal
and J. Lyklema, Pure Appl. Chem., 2005, 77, 1753–1805.
9 S. M. Ahmed, Electrochim. Acta, 1982, 27, 707–712.
The Gibbs energy of an adsorbed hydrogen (H*) on nano-
crystalline MoS2 edge sites is similar to that on Pt, i.e. close to
zero.1a Thus, protonation and subsequent H2 release occurs
at an accelerated rate on MoS2 compared to a mechanism
without catalyst.7c Exfoliated MoS2 is more reactive than
nanocrystalline MoS2, as more edge sites are exposed due to
its graphene-like sheet structure, though the increase in the
HER rate was not as substantial as expected (see Table 1).
This observation may be due to exfoliated MoS2 being present
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun.