Interface confined hydrogen evolution reaction in zero valent metal nanoparticles-intercalated molybdenum disulfide

Abstract

Interface confined reactions, which can modulate the bonding of reactants with catalytic centres and influence the rate of the mass transport from bulk solution, have emerged as a viable strategy for achieving highly stable and selective catalysis. Here we demonstrate that 1T'-enriched lithiated molybdenum disulfide is a highly powerful reducing agent, which can be exploited for the in-situ reduction of metal ions within the inner planes of lithiated molybdenum disulfide to form a zero valent metal-intercalated molybdenum disulfide. The confinement of platinum nanoparticles within the molybdenum disulfide layered structure leads to enhanced hydrogen evolution reaction activity and stability compared to catalysts dispersed on carbon support. In particular, the inner platinum surface is accessible to charged species like proton and metal ions, while blocking poisoning by larger sized pollutants or neutral molecules. This points a way forward for using bulk intercalated compounds for energy related applications.

ARTICLE

Received 24 Aug 2016 | Accepted 11 Jan 2017 | Published 23 Feb 2017

Zhongxin Chen1,2,*, Kai Leng1,*, Xiaoxu Zhao1,2, Souradip Malkhandi1, Wei Tang1,3, Bingbing Tian1, Lei Dong4, Lirong Zheng5, Ming Lin3, Boon Siang Yeo1 & Kian Ping Loh1

Interface conned reactions, which can modulate the bonding of reactants with catalytic centres and inuence the rate of the mass transport from bulk solution, have emerged as a viable strategy for achieving highly stable and selective catalysis. Here we demonstrate that 1T0-enriched lithiated molybdenum disulde is a highly powerful reducing agent, which can be exploited for the in-situ reduction of metal ions within the inner planes of lithiated molybdenum disulde to form a zero valent metal-intercalated molybdenum disulde. The connement of platinum nanoparticles within the molybdenum disulde layered structure leads to enhanced hydrogen evolution reaction activity and stability compared to catalysts dispersed on carbon support. In particular, the inner platinum surface is accessible to charged species like proton and metal ions, while blocking poisoning by larger sized pollutants or neutral molecules. This points a way forward for using bulk intercalated compounds for energy related applications.

1 Department of Chemistry and Centre for Advanced 2D Materials (CA2DM), National University of Singapore, 3 Science Drive 3, Singapore 117543, Singapore. 2 NUS Graduate School for Integrative Sciences and Engineering, National University of Singapore, Centre for Life Sciences, #05-01, 28 Medical Drive, Singapore 117456, Singapore. 3 Institute of Materials Research and Engineering, 2 FusionopolisWay, Singapore 138634, Singapore. 4 Department of Macromolecular Science, Fudan University, 220 Handan Road, Shanghai 200433, China. 5 Beijing Synchrotron Radiation Facility (BSRF), Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049, China. * These authors contributed equally to this work. Correspondence and requests for materials should be addressed to K.P.L. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

DOI: 10.1038/ncomms14548 OPEN

Interface conned hydrogen evolution reaction in zero valent metal nanoparticles-intercalated molybdenum disulde

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14548

Two-dimensional (2D) molybdenum disulde (MoS2) crystal exhibits high catalytic activity for hydrogen evolution reaction (HER), thus it is being considered as a

low cost alternative to the commonly used platinum catalyst15. It is well known that the basal plane of MoS2 is relatively inert and only the edge sites turn over reactions. MoS2 can exist as the thermodynamically stable 2H phase (space group, P63/mmc) or the metastable 1T phase (space group, P 3m1). Bulk 2H-MoS2 is a semiconductor with an indirect band gap of B1.2 eV, while the 1T-MoS2 phase is metallic4. The 2H-to-1T0 MoS2 phase conversion, induced commonly by chemical intercalation with alkali metal ions, has been applied as a strategy to enhance HER activities recently1,6. The increased electron density in Mo 4d orbitals due to electron transfer from intercalated lithium increases the metallic character of the basal plane and improves catalytic activity. However the 1T0 phase is metastable with respect to the 2H phase due to its intrinsic reactivity, this is further compounded by the rapid hydration of Li in aqueous phase, thus any prospect of long term, stable operation is quickly offset by the short-lived nature of these systems3,7.

The use of 2D MoS2 nanosheets as catalysts requires tedious exfoliation processes, which constrains mass production and industrial applications. Although, it is possible to coat metal nanoparticles on exfoliated 2H-MoS2 to enhance its catalytic activity, such nanoparticles rapidly corrode under acidic HER environment and get leached, leading to the loss of activity2,8. Restacked MoS2 nanosheets intercalated with transition metal ions can be synthesized from single layer MoS2 dispersions by an ion exchange method, in which the cation (M2 ) neutralizes the negative charge of the MoS2 layer and the material restacks with alternating layers of MoS2 and M(OH)2. However, the chemical reduction of these metal ions requires the use of reducing agents and adds to the complexity of the process9. Alternatively, it is worth relooking at bulk MoS2. Due to its layered structure, bulk

MoS2 can serve as host for a large class of metal nanoparticles within its inner planes and protect them from corrosion9. An intriguing idea is whether the host and intercalant can operate synergistically to achieve highly stable HER catalysis with the added benet of a reduced loading of the noble catalyst, compared to coating on the bare, exposed surfaces of exfoliated 2D sheets.

A diffusion method relying on capillary action is primarily used to introduce nanoparticles into inner cavities of the host material. For example, the encapsulation of metal/metal oxide nanoparticles in carbon nanotubes, zeolites and metal organic frameworks produced enhanced catalytic performance towards alcohol oxidation10, FischerTropsch synthesis11 and asymmetric hydrogenation12. However, encapsulation (or so-called intercalation) in 2D, layered materials is challenging due to the strong van der Waals force between adjacent nanosheets. The narrow spacing between the layers means that only small, alkali metal ions (Li, Na or K ) can be electrochemically intercalated into bulk materials13,14. The insertion of larger cations like Co2 usually requires the exfoliation of MoS2 in water and their subsequent occulation in the presence of the cation intercalant to generate a restacked, sandwiched structure.

Herein, we demonstrate a simple, straightforward strategy to achieve zero-valent intercalation in bulk MoS2 by the in-situ reduction of noble metal ion precursors. Our method relies on the pre-intercalation of bulk MoS2 with Li ions to form 1T0 phase-engineered LixMoS2. Due to its enhanced electron density, we found that 1T0-LixMoS2 has a large reduction potential, and this property can be exploited to reduce a large class of transition metal ions. The enlarged interlayer spacing in LixMoS2 allows diffusional-ion exchange with the metal salt of interests, leading to the reduction of metal ion within the bulk of MoS2 and

removal of the intercalated Li as salt. The synergetic hostguest interaction in Pt-intercalated 2H MoS2 (abbreviated as Pt-MoS2) allows ultra-stable, long-term operation in HER with a reduced loading of Pt compared with commercial Pt/MoS2 catalyst.

ResultsStrong reducing power of 1T0-LixMoS2. 1T0-phase MoS2 is unstable against 2H conversion in water, thus any long-term application in aqueous phase will involve the 2H phase inevitably. Our working hypothesis is this: the enhanced reactivity of the 1T0 phase compared with the 2H phase can be exploited to reduce metal ions and make nanometal-intercalated MoS2 hybrids. To assess its electron reducing ability, the reduction of fullerene (C60)

is chosen as a model system15. C60 is an excellent electron acceptor, which can undergo distinct six successive reductions in solution. The B0.5 V separation between each reduction step of the C60 means that it is relatively easy to identify the reduction potential of a particular reducing agent16. In this regard, the chemical reduction, as well as the electrochemical reduction of C60 by 1T0-LixMoS2 is rst studied. As shown in Fig. 1a,b, in argon-protected environment, C60 can be readily reduced to C60

(fulleride) by LixMoS2, producing an obvious colour change from purple to orange. The existence of fulleride is conrmed using ultravioletvis spectroscopy, where four ngerprint peaks located at 923, 995, 1,033 and 1,075 nm, arising from allowed t1u-t1g

transitions and its vibrational ne structure, can be seen. The presence of fulleride also gives rise to a broad signal with a low g value (2,000) in the electron paramagnetic resonance (EPR) spectrum, which is due to the spin-orbit coupling effects of unquenched angular momentum in JahnTeller-distorted states of the (t1u)1 conguration15. We further compare LixMoS2 with a series of common reducing agents for the chemical reduction of C60, as shown in Supplementary Fig. 1. Similar to strong reducing agents like hydrazine ( 1.20 V versus RHE), sodium

borohydride ( 1.24 V versus RHE), 1T0-LixMoS2, is able to

spontaneously reduce C60 in the rst of the six-electron reduction reaction ( 0.68 V versus RHE) at room temperature, while weak

reducing agents and 2H-MoS2 cannot.

The reduction ability of 1T0-phase LixMoS2 is further examined by the electrochemical reduction of C60 (ref. 17). C60 is distinguished by a six-electron reduction process, which can be revealed by six distinct pairs of peaks in cyclic voltammetry and six peaks in differential pulse voltammetry in Supplementary Fig. 2. The modication of electrode with 1T0-phase LixMoS2 results in a positive shifts (B0.2 V) of these redox pairs together with a signicant enhancement in the current density in Fig. 1c. The rst redox pair (C60 fulleride) almost disappears due to the spontaneous pre-reduction of C60 in the presence of LixMoS2. Increasing the scan rate to 500 mV s 1 did not affect the intensity of the reduction peaks, which indicates fast electron transfer kinetics and excellent conductivity of LixMoS2 (Supplementary

Fig. 3)16. In contrast, 2H-MoS2 acts as a passive electrode in the electrochemical reduction of C60, its peak positions are similar to that of glassy carbon in Supplementary Table 1. On the basis of this study, we can infer that 1T0-LixMoS2 is a strong reducing agent with a reduction potential more negative than 0.7 V

versus RHE.

Ion exchange and intercalation of noble metals. The strong reduction ability of 1T0-LixMoS2 can be exploited for the synthesis of hybrid MoS2 nanocomposites consisting of zero-valent metal intercalants. As shown in Supplementary Fig. 4, we rst screen the rst two rows of transition metal precursors, including PtCl62 (1.48 V, versus RHE), AuCl4 (0.93 V), Pd2 (0.92 V),

Ag (0.80 V), Ru3 (0.70 V), Cu2 (0.34 V), Ni2 ( 0.25 V),

2 NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14548 ARTICLE

a

b c

LixMoS2

C Excess, r.t. under Ar

60

C60

Chemical

100 G

EPR spectroscopy

g = 2.000 (C60) 1T,LixMoS2

IV

III

2 A

V

VI

II

C60

C60

I II

III

IV

V

VI

2H, Bulk MoS2

300 600 900 1,200 1,500

Electrochemical

1 2 3

Wavelength (nm) Potential (V versus Fc/Fc+)

Figure 1 | Strong reduction ability of 1T0-LixMoS2 as evidenced by the spontaneous reduction of C60. (a) C60 reaction equation. (b) UVvis-NIR spectra and digital photos before/after LixMoS2 reduction. Inset: solution EPR spectrum of the generated C60 in AN/Toluene. (c) Differential pulse voltammetry curves of electrochemical reduction of C60 on 1T0-LixMoS2 and 2H-MoS2 in AN/Toluene.

Co2 ( 0.28 V) and Fe2 ( 0.44 V), with their standard

reduction potential indicated in the brackets. The reduction is performed in a non-aqueous solution (anhydrous THF or NMP depending on the solubility of precursors) to prevent the de-stabilization of the 1T0-phase18. In this anhydrous environment, the high-valent state metal ions diffuse into the layers of 1T0-phase LixMoS2 and are in-situ reduced to the zero-valent state. To compensate for the excess charges, Li and

Cl ions are completely leached from the MoS2 host after this ion exchange. It must be pointed out that bulk 2H-MoS2 and exfoliated MoS2 nanosheets cannot reduce PtCl62 to Pt(0)

(Supplementary Fig. 5); exfoliated MoS2 nanosheets only have very weak reduction ability due to the rapid destruction of the 1T0-phase after hydration.

LixMoS2 !

PtCl2 6 xLi xCl Pt 0

MoS2 !

Exchange PtMoS2 x Li Cl

1

According to the equation above, Pt-MoS2 is prepared from the intercalation of bulk MoS2 with lithium using n-BuLi, followed by subsequent redox reaction with sodium hexachloroplatinate. As illustrated in Fig. 2a, bulk, 1T0-phase LixMoS2 is directly used as reductant and host materials, which bypasses the tedious exfoliation process to form 2D and restacked MoS2. More importantly, Pt nanoparticles are sandwiched between adjacent MoS2 layers, thus protecting them against chemical corrosion and minimizing their dissolution and aggregation during long-term HER test19.

To assay for the spatial distribution of intercalated and reduced zero-valent Pt, time-of-ight secondary ion mass spectroscopy (ToF-SIMS) is used to acquire the 3D elemental distribution of Pt in MoS2 (ref. 20). As shown in Fig. 2b, Pt nanoparticles can be clearly seen on MoS2. At a longer sputtering time, we observe both Pt and MoS2 signals in the composition map of the deeper layers, which proves the successful intercalation of Pt at deeper regions of MoS2 single crystal (B200 nm). From the sputtering prole, Pt species can be detected at depths up to 400 nm and beyond. The SIMS proles of MoS2 and Pt signals in Fig. 2c suggest that the vertical distribution of Pt is quite homogeneous. Consistent with the kinetics of the intercalation process, a diffusion gradient exists for the Pt nanometals when we

analysed its lateral distribution on layers exfoliated from bulk Pt-MoS2 by the Scotch tape method. Pt signal can be detected by

EDS mapping only at the edge of MoS2 after 12 h of intercalation, while the signal can be detected homogeneously across the ake after 3 days, as seen in the FESEM and EDS images in Supplementary Figs 7 and 8. Pt-MoS2 has a well-ordered layered structure with an expanded layer spacing due to the intercalation of the metal nanoparticles18. Its bulk morphology is markedly different from the restacked nanosheets generated from exfoliation strategy. The selected area electron diffraction (SAED) pattern of Pt-MoS2 is distorted from the typical hexagonal pattern of bulk 2H-MoS2. The continuous ring with a measured d-spacing of 2.2 as determined by Fourier transform can be assigned to the {111} planes of Pt, while the appearance of two new rings strongly suggests the intercalation of Pt species into MoS2 host structure (Supplementary Fig. 9)8. The intercalation strategy was also veried to be successful for a wide range of transition metals, such as Au, Ru and Pd as veried by SEM and EDX (Supplementary Figs 7 and 8).

The change in interlayer spacing of MoS2 after intercalation by Pt was also veried by grazing incidence X-ray diffraction (GIXRD). As shown in Fig. 3, we can nd a broad peak at B2.46

(B3.6 nm) in 1D and 2D GIXRD for the MoS27Pt7MoS2 sandwiched structure, which is absent in the bulk sample. The d-spacing of MoS2 layer is calculated to be 0.63 nm from X-ray diffraction and the average size of Pt nanoparticles is found to be B2 nm from HRTEM images in Supplementary Fig. 6.

Accordingly, Pt-MoS2 has a repeat unit of 23 layers of MoS2 and 1 layer of Pt nanoparticles. In addition, we also observe a peak shift of the MoS2 {002} from 14.10 to 14.34 together with an increase in FWHM after Pt intercalation, which can be ascribed to the stress on the MoS2 layers induced by Pt nanoparticles in-between21,22. The existence of Pt nanoparticles in the inner layers of MoS2 is directly conrmed by the TEM images of exfoliated Pt-MoS2. As shown in Fig. 3d,e, Pt nanoparticles with sizes ranging from 1.5 to 3.5 nm are homogeneously dispersed on the MoS2 akes. The SAED pattern (Fig. 3f, as indicated by red arrow) and EDS spectrum (Fig. 3g) conrm the successful intercalation of Pt nanoparticles into the inner layers of MoS2 akes. The cross-section HAADF-STEM images, as well as the EELS mapping in Supplementary Fig. 6 also demonstrate the intercalation of Pt nanoparticles in between the MoS2 layers. The

NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14548

a

Li+ + Cl

M+

Exchange

Zero-valent

Intercalation

Bulk, 2H-MoS2

Metal nanoparticles

Top (0 s) Middle (300 s) Bottom (600 s) Side view, depth 400 nm

Intensity (a.u.)

Bulk, 1T-LixMoS2

MoS2

Lithium

Lithiation

by n-BuLi

2H, ternary metal-MoS2 (metal = Pt, Ru, Au, Pd)

b c

MoS2 MoS2

Pt

Pt

d

300 nm SiO2/Si

Si

MoS2

Pt

O

600

0 100 200 300 400 500

Sputter time (s)

Figure 2 | Homogeneous metal distribution in bulk MoS2. (a) Zero-valent intercalation of metal nanoparticles by an in-situ reduction strategy.(b) ToF-SIMS maps at the beginning (0 s), middle (300 s) and end (600 s) of the experiment. The false colours of blue and red corresponds to MoS2

(162 amu) and Pt (195 amu). (c) ToF-SIMS side view and (d) depth prole of the same region (etching depth is roughly 400 nm, as calculated by the step of 300 nm SiO2/Si wafer). Scale bar, 20 mm.

small-angle X-ray scattering (SAXS) proles in Supplementary Fig. 12 conrm a loosely stacked structure of Pt-MoS2 due to hydrogen bubbles-induced volume expansion during zero-valent intercalation (Supplementary Note 1).

Structure evolution of intercalated compounds. The phase conversion of 2H and 1T0 during in-situ reduction is followed by

X-ray photoelectron spectroscopy (XPS). As shown in Fig. 4a and Supplementary Fig. 10, chemical lithiation of bulk MoS2 converts it to the metallic 1T0-phase with the 1T0-related peaks at 228 and 231 eV for the spin orbit coupled Mo3d peaks, and 161.2 and 162.6 eV for the S2p peaks18. The Pt-intercalated MoS2 reverts to the 2H-phase after electrons are donated to Pt ion precursor and the exchange-diffusion of the Li ions. The intercalated Pt atoms are mostly at their zero-valence state as judged from the Pt4f XPS

spectrum in Fig. 4b. The absence of Pt4 species suggests the full reduction of sodium hexachloroplatinate by LixMoS2. The phase conversion from 1T0-LixMoS2 to 2H-Pt-MoS2 is also conrmed by the Raman spectra in Supplementary Fig. 11. The J1 to J3 phonon modes around 170300 cm 1 of the 1T0-phase MoS2 are clearly discernible in the lithiated and exfoliated materials, while absent in Pt-intercalated 2H-MoS2. The latter shows two main phonon peaks at 404 cm 1 and 380 cm 1, which are due to the

A1g and E12g modes of the 2H phase1,18.

To verify and quantify the diffusional substitution of Li ions by Pt ions, the elemental composition of the composite was analysed

by XPS (Supplementary Fig. 10) and ICP-OES (Supplementary Table 2). The residual lithium and chlorine percentage is o0.5 wt% in Pt-MoS2 and there are no Li1s or Cl2p peaks according to XPS analysis, which proves that most of the Li ions have been replaced by Pt. We also conrmed 480% lithium exchange using the Mohr titration of chloride23, where Li is assumed as the only counter ion of Cl in the supernatant of

Pt-MoS2 (Supplementary Fig. 13 and Supplementary Note 2).

The local bonding environment in Pt-MoS2 was investigated by extended X-ray absorption ne structure (EXAFS)7. The K-edge EXAFS spectrum for pristine MoS2, as shown in Fig. 4c, is consistent with the trigonal prismatic structure, with six MoS bonds of 1.93 and six MoMo bonds of 2.85 . These peaks are much weaker in the case of LixMoS2, which means the second shell of Mo atoms is disturbed; the decreased MoMo distance of2.67 also suggests the change in site symmetry due to trigonal prismatic (2H) to octahedral (1T0) phase change24. In the case of

Pt-MoS2, the K-edge EXAFS indicate structural rearrangement consistent with the 2H phase. We do not nd evidence of MoPt interaction. This suggests that the Pt atoms in MoS2 are probably existing as Pt nanoclusters, as opposed to Pt occupying vacancy sites7. Powder X-ray diffraction was also performed to examine the layered structure of MoS2. As presented in Fig. 4d, the strong {002} diffraction peak at 14.8o of bulk MoS2 shifts to lower position after Li intercalation (14.3o), which can be ascribed to the {001} of 1T0-phase of LixMoS2 (ref. 18). The intensity of this peak is reduced after Pt intercalation. The presence of

4 NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14548 ARTICLE

a b c

{002}

3.6 nm

Pt-MoS2

MoS2

{002}

{002}

0 5 10 15 20 2[afii9835] (deg)

d e

f

g

Mo, S

Pt Pt Pt Pt Mo Mo

0 5 10 15 20 Energy (keV)

Figure 3 | Evidence of Zero-Valent Intercalation. (a) GIXRD patterns of single crystal MoS2 and (b) Pt-MoS2; (c) Corresponding 1D spectra in the out-of-plane direction. (d,e) TEM images of the nanosheets exfoliated from bulk Pt-MoS2. Corresponding (f) SAED and (g) EDS spectrum of the exfoliated Pt-MoS2. Scale bar, d,e 50 nm; f 5 nm 1.

Intensity (a.u.)

a c

Mo-S

Mo-Mo

2H

Pt-MoS2

Pt-MoS2

Pt-MoS2

LixMoS2

LixMoS2

Intensity (a.u.)

12

9

6

3

0

0 2

1T

LixMoS2

MoS2

MoS2

MoS2

166

80 78 76 74 72 70 68

Mo foil

164 162 160

Binding energy (eV)

Binding energy (eV)

4 6 8 10

R (A)

b d

Pt-MoS2

Pt (0) PtO

Pt

*

Intensity (a.u.)

Intensity (a.u.)

002

JCPDS 37-1,492

10 20 30 40 50 60 70 80

2[afii9835] (deg)

Figure 4 | Phase transition process of metal intercalation. (a) XPS S2p spectra showing the evolution from 2H (bulk MoS2) to 1T0 (LixMoS2) to 2H (Pt-MoS2); (b) XPS Pt4f spectrum validating the zero-valent state of intercalated metals; (c) k2-weighted Mo K-edge EXAFS spectra of PtMoS2 in comparison to Mo foil, bulk MoS2 and LixMoS2; and (d) XRD patterns showing the fcc structure of intercalated Pt nanoparticles.

NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14548

Pt nanoparticles can be judged clearly from XRD peaks originating from the face-centered cubic (fcc) crystal phase of Pt, with two strong peaks at 39.4o {111}, 43.5o {200} and a weak peak at 67.5o {220}, respectively8. The peaks located at 32.8o and52.1o can be assigned to the {100} and {110} of MoS2.

HER performance in a conned space. To examine interface-conned catalytic reactions in the Pt-MoS2 hybrid, in which the catalysts are intercalated within the quasi-2D MoS2 layers, we chose HER as the model reaction because of the small hydrated radius and fast diffusion kinetics of proton in water, which has less mass diffusion limitation.

We measured the electrochemical HER performance using a three-electrodes cell and compared the performance with 40 wt% commercial Pt/C catalyst and exfoliated MoS2 nanosheets (Supplementary Figs 14 and 15). The content of Pt, Ru, Pd and Au in MoS2 are identied by ICP-OES as 10.4, 11.3, 24.5 and18.3 wt%, respectively. Figure 5a shows that these metal intercalated MoS2 are highly active toward HER. For instance,

Pt-MoS2 exhibits a negligible overpotential at current density of 10 mA cm 2, together with a very low charge transfer resistance (RCT) of o5 O in the electrochemical impedance spectroscopy (EIS) spectra (Fig. 5b). The facile kinetics of Pt-MoS2 is a major reason for its superior catalytic activity in HER (Supplementary Note 4)5,25,26. The RCT is much larger than the porosity resistance of the electrode (RP) and is overpotential-dependent in Supplementary Fig. 16, suggesting the domination of RCT

and a combination of VolmerTafel mechanism with the recombination of adsorbed H* as the rate determining step in HER reaction27,28. The growth of Pt within the conned layers of MoS2 may lead to a smaller particle size, leading to the excellent

HER performance. Importantly, it is observed that the long-term stability of these metal intercalated MoS2 far exceeds that of commercial Pt/C in Fig. 5c. For instance, there is no obvious decay in overpotential for Pt-MoS2 after 430 h continuous

operation at 50 mA cm 2, whereas commercial Pt/C is rapidly degraded during stability test as a result of Pt nanoparticles dissolution/aggregation, as well as carbon support corrosion (Supplementary Note 3)19. The sandwiching of Pt between the planes of MoS2 has a synergetic effect in enhancing the stability and kinetics of the hydrogen evolution reaction (HER)29. This is a key advantage compared with the conventional way of coating metal catalysts on the surface of nanosheets, in which these metal catalysts are invariably corroded in acidic environment and leached away, contributing to catalyst failure and instability of HER performance14. In contrast, our Pt intercalated-MoS2 system shows superior stability because of the synergetic host-intercalant effect (Supplementary Note 5). In addition, pressure-releasing channels in the form of cracks, voids and edges exist throughout the layered Pt-MoS2 structure and these prevent pressure build-up during hydrogen evolution (Supplementary Figs 18 and 19).

We adopt two common indicators here to further access the HER performance. The Tafel slope is the increase in overpotential required to produce a one-order magnitude rise in current density while the onset potential is the potential at which current density begins to fall steeply due to proton reduction30,31. We select a similar overpotential range (2545 mV) to calculate the Tafel slope for better comparison. As a benchmark catalyst, 40 wt% Pt/C shows a small Tafel slope of 22 mV dec 1 and an onset potential of 24 mV for hydrogen evolution. The Tafel slope and onset potential for Pt, Pd, Ru, Au-intercalated MoS2 are 25, 30, 61, 38 mV dec 1 and 20, 35, 38, 22 mV in Fig. 5d, respectively. The mass transport limitation is not signicant due to the fast diffusion kinetics of protons. The HER performance of various metal-intercalated MoS2 follows the theoretical prediction for noble metals (so-called Volcano shape) except for Au-MoS2 (ref. 5), suggesting that the active sites are on noble metal nanoparticles and MoS2 served as co-catalysts and support for noble metals. The hydrated Au precursor has an excellent solubility in THF, leading to a much better intercalation

a b

Current density (mA cm2 )

Overpotential (mV)

0

10

20

30

20

15

10

5

0

0 10

Overpotential (mV)Impedance (ohm)

40% Pt/C

Au-MoS

Pt-MoS

40% Pt/C

Ru-MoS

40

Pd-MoS

50

50

100

150

200

250

0 5 10 15 20 25 30 35Time (h) Log j (mA cm2)

0.20 0.15 0.10 0.05 0.00 0.05

Potential (V versus RHE)

Operate at 50 mA cm2

Test at 30 mV overpotentialRS RP RCT

CPE1 CPE2

20 30 40 50 Impedance R (ohm)

c d

0

50

61 mV dec

40

30 mV dec

25 mV dec

22 mV dec

38 mV dec

30

20

Pt-MoS

Ru-MoS

Pd-MoS

10

Au-MoS

0

0.5 0.5 1.0 1.5

0.0

Figure 5 | Superior HER performance of metal-intercalated MoS2 catalysts. (a) LSV curves, (b) Nyquist plots and EIS simulation model, (c) long-term stability test and (d) Tafel plots. All experiment were conducted in 0.5 M H2SO4 at room temperature.

6 NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14548 ARTICLE

a b

Pt-MoS20.05 mA cm2

0.20 mA cm2

Pt-MoS2

40 wt% Pt/C

0.4 0.6 0.8

0.10 mA cm2

0.02 mA cm2

CO

UPD 1

40 wt% Pt/C

CO

UPD 1

UPD 2

0.2

0.4 0.6 0.8 0.3 0.5 0.7

1.0

Potential (V versus RHE) Potential (V versus RHE)

c d

100

80

Saturation (%)

60

40

Cu2+

III

II

I

Pt/C Pt-MoS2

Anisotropic diffusion

0 200 400 600 800 1,000

Time (s)

Figure 6 | Catalysis mechanism of inner Pt nanoparticles. (a) CO stripping and (b) Cu UPD of 40 wt% Pt/C and Pt-MoS2, showing the different accessibility of inner Pt nanoparticles. (c) The saturation rate of the Cu monolayer deposition upon holding at a xed UPD potential, which indicates a much slower diffusion process in the case of Pt-MoS2 compared with Pt/C. (d) Schematics showing the anisotropic diffusion of CO gas and cupric ions into

Pt-MoS2.

efciency compared to Ru and Pd, which may be a possible reason for the enhanced activity of Au-MoS2. The similar trend in long-term stability also proves the validity of this zero-valent intercalation, while the relatively poor stability of Pd-MoS2 may be attributed to its surface oxidation and dissolution of its oxide in the acidic environment. Furthermore, the performance of Pt-MoS2 is compared with recently reported HER catalysts in

Supplementary Table 4, showing its superior performance. The advantage of zero-valent intercalation of MoS2 is that bulk MoS2 powder is processed directly to make the hybrid, bypassing the tedious exfoliation process of 2D MoS2 nanosheets. Thus we were able to demonstrate industrial scalability by fabricating a 5 5 cm2 catalyst-loaded water splitting membrane in

Supplementary Fig. 17. This is unprecedented in terms of research in 2D transition metal chalcogenide so far, which are often limited by the size of the exfoliated akes and difculty in spin-coating a continuous lms27.

One reason for the extra stability of Pt-MoS2 arises from the passivation effect enjoyed by the Pt nanoparticles sandwiched between the MoS2 layers. However, it is not known if the highly anisotropic diffusion needed to access the Pt in the inner layers will result in a lower catalytic efciency. To evaluate the electrochemical activity of these MoS2-sandwiched Pt catalysts, we carry out carbon monoxide (CO) voltammetric stripping and Cu underpotential deposition (UPD) (ref. 32). In stripping voltammetry, a saturated layer of CO is pre-adsorbed on the Pt catalyst followed by its subsequent oxidation in voltammetric scans. In UPD, Cu metal is deposited above the Nernst potential to form sub-to-monolayer Cu lms on Pt surface (Supplementary Note 6). The presence of the CO stripping peak or UPD peak allows us to assay the electrochemical active surface area (ECSA) of the Pt nanoparticles (Supplementary Figs 20 and 21)33. In addition, kinetic information can be obtained by observing the

saturation rate of the Cu monolayer deposition when implemented in the chronoamperometric mode (see Supporting information for detailed discussion on UPD and CO stripping). As presented in Fig. 6a,b, the benchmark Pt/C shows comparable ECSAs in CO stripping (0.62 cm2) and Cu UPD (0.58 cm2). The CO stripping peak at 0.81 V versus RHE arises primarily from its saturated adsorption on Pt {111} (ref. 34), while the two broad Cu UPD peaks correspond to two distinct redox reactions: an intermediate copper layer coadsorbed with sulfate ions (Cu UPD1), as well as the 1 1 monolayer deposition (Cu UPD2)

(ref. 35). However, the CO stripping peak is much weaker in the case of Pt-MoS2. Due to the nanosized Pt particles in the

MoS2-conned layers, we only observe one strong Cu UPD peak in Pt-MoS2, which is due to sulfate-coadsorbed copper layer. The ECSA as determined by CO stripping for Pt-MoS2 (0.16 cm2) is also much lower than that of Cu UPD (0.85 cm2), implying that the Pt sites in Pt-MoS2 are not accessible to CO adsorption, although these are accessible to cupric ions. The mass transport of Cu ions into the inner active surface of Pt-MoS2 can be electrically driven; however neutral CO molecules rely entirely on mass diffusion and the anisotropic diffusion path involving through the open edges of the MoS2 layers becomes rate-limiting (Supplementary Table 3).

Diffusion kinetics can be used to distinguish exposed Pt and trapped Pt in the electrode structure, it is investigated by holding the Cu UPD at a xed voltage (275 mV versus RHE). Fig. 6c and Supplementary Figs 2224 show that the current signal is rapidly saturated for Pt/C, whereas it takes a longer time to reach saturation for Pt/MoS2 due to the slower diffusion to reach the sandwiched Pt nanoparticles. Such intercalated structure is promising for electrocatalysis because the inner surface is accessible to charged particles like proton and metal ions, while blocking poisoning by larger sized pollutants or neutral molecules

NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14548

(Fig. 6d). Slow O2 or CO diffusion is benecial for long-term operation due to reduced corrosion or poisoning. Thus, the bulk, layered structure of Pt-MoS2 imparts more stability than its randomly restacked counterparts, contributing to excellent activity and stability in HER.

DiscussionThe direct intercalation of zero-valent metals in bulk, layered materials has rarely been achieved. Previous methods rely on the exfoliation of 2D sheets from bulk crystals, followed by ion-exchange of the Li with metal salts in solution and the re-stacking of the sheets, where upon the metal ions got intercalated9. Reduction to zero-valent metals requires additional chemical reduction steps, especially in the case of 2H-MoS2, which does not have inherently strong reducing power. As demonstrated in this work, the in-situ reduction of Pt precursors by making use of the highly reducing 1T0-LixMoS2 phase provides a convenient way of generating a zero valent-metal intercalated quasi-2D material. It involves the following steps: diffusion of Pt ions from solution into the inner planes of lithiated MoS2, and reduction therein of the Pt ions by the metallic 1T0 phase, followed by Pt nanoparticle growth under layer-conned condition. At the same time, Li /Cl ions will be ion exchanged from MoS2, while the latter undergoes phase conversion from the 1T0-phase to 2H-phase. The metal ion exchange proceeds readily when hydrated metal salts like Na2PtCl6 6H2O were used, since the

large hydration energies of Li favors its de-intercalation from MoS2 (ref. 36). In contrast, when anhydrous metal precursors (for example, K2PtCl4) were used, it was observed that large nanoparticles nucleate at the edge or surface of MoS2 akes due to its sluggish diffusion. The negative enthalpy of Pt reduction by 1T0 phase ( 571 kJ mol 1)37, along with the energy gain from

phase conversion from 1T0-phase to 2H phase (4 68 kJ mol 1

from reduction potential) and the large solvation energy gain of Li ( 101 kJ mol 1 in THF)38, contribute to the overall free

energy gain of the ion exchange process.

The conned space within the adjacent MoS2 layers provides an intriguing environment for catalysis15,39,40. First, the narrow space limits the growth of the Pt nanoparticles, and dispersion within the MoS2 layers prevents aggregation. Thus even at a reduced weight loading (10 wt%) of Pt in Pt-MoS2, it shows a higher ECSA value compared to benchmark (40 wt%) Pt/C. This is consistent with the larger electrochemical surface area of nanoparticles grown which are size-conned. For instance, Cameron et al.10 reported that zeolite-conned RuO2 nanoparticles have a very small particle size (1.30.2 nm). Importantly, such spatial restriction can also modify the accessibility and interaction between catalysts and reactants. Bao et al.41 observed strong deformations of transition metals within the CNT channels due to different electronic structures and spatial connement. This will lead to downshifted d-band states and weaker adsorption of CO, N2 and O2 molecules, which could also apply to Pt-MoS2. Due to the restricted mass diffusion of CO, Pt-MoS2 possess CO-tolerant properties, thus it can be potentially applied in methanol oxidation fuel cell where the CO poisoning of Pt catalyst is a serious problem.

In conclusion, we have demonstrated that 1T0 polymorph-enriched bulk LixMoS2 is a powerful reducing agent. As evidenced by the chemical and electrochemical reduction of C60 molecules, the reduction potential of 1T0-phase LixMoS2 is more negative than 0.7 V versus RHE. The reducing power can be exploited

to fabricate zero-valent metal intercalated 2H-MoS2 for a wide class of transition metals. Taking advantage of the enlarged interlayer spacing in 1T0-phase LixMoS2, the diffusional exchange of noble metals ions, followed by its in-situ reduction, can occur

within the layers of MoS2, giving rise to a unique hybrid quasi-2D system, that is, noble-metal intercalated MoS2. Although, there are plenty of reports on the use of 2D MoS2 in HER, 1T0-phase MoS2 is not stable for long-term practical applications while the performance of 2H-phase is not satisfactory and require defect engineering. The advantages of Pt-intercalated 2H-MoS2 include the size-restricted growth of Pt nanoparticles within MoS2, where the Pt remains highly active for HER, and the slow kinetics for poisoning species in conned space, which ensures that the long-term stability of this hybrid system exceeds that of commercial Pt/C catalyst. This study points to the possibility of a large class of zero valent metal-intercalated TMDs exhibiting synergetic properties in catalysis.

Methods

Synthesis of Pt-MoS2. Around 1 g of MoS2 powder was added to 20 ml 1.6M n-BuLi in hexane, and stirred at r.t. for 2 days in an argon-lled glove box. Excess amount of n-BuLi was removed by centrifugation. Then, 100 mg LixMoS2 was dispersed in 20 ml anhydrous THF and 33.7 mg Na2PtCl6 6H2O was added. The

mixture was sealed in a Teon-lined autoclave and kept at 80 C for 2 days. As-prepared Pt-MoS2 was thoroughly washed with THF, IPA, ethanol and water (Supplementary Methods).

Equipment. The following equipment were used: Raman (WITec Alpha 300R), SEM/EDS (Jeol JSM-6701F), AFM (Dimension Fast Scan), XPS (AXIS UltraDLD, monochromatic Al Ka), Electrochemistry (CHI 660E and Zahner Zenniumwith a three-electrodes cell, Supplementary Methods), STEM & EELS (Nion

UltraSTEM-100 with aberration-correction, 60 KV), TEM/EDS (FEI Titan, 80 kV), ToF-SIMS (ION-TOF SIMS5 with Bi and Cs beams, Supplementary Methods),

EXAFS (1W1B-XAFS beamline, BASF, Supplementary Methods), SAXS (SAXSess mc2), XRD/GIXRD (Bruker D8 & GADDS, Supplementary Methods).

Data availability. The authors declare that the data supporting the ndings of this study are available within this paper and its Supplementary information le,or from the corresponding authors.

References

1. Voiry, D. et al. Conducting MoS2 nanosheets as catalysts for hydrogen evolution reaction. Nano Lett. 13, 62226227 (2013).

2. Deng, J. et al. Triggering the electrocatalytic hydrogen evolution activity of the inert two-dimensional MoS2 surface via single-atom metal doping. Energy

Environ. Sci. 8, 15941601 (2015).3. Chou, S. S. et al. Understanding catalysis in a multiphasic two-dimensional transition metal dichalcogenide. Nat. Commun. 6, 8311 (2015).

4. Kong, D., Cha, J. J., Wang, H., Lee, H. R. & Cui, Y. First-row transition metal dichalcogenide catalysts for hydrogen evolution reaction. Energy Environ. Sci. 6, 35533558 (2013).

5. Zheng, Y., Jiao, Y., Jaroniec, M. & Qiao, S. Z. Advancing the electrochemistry of the hydrogen-evolution reaction through combining experiment and theory. Angew. Chem. Int. Ed. 54, 5265 (2015).

6. Voiry, D. et al. The role of electronic coupling between substrate and 2D MoS2 nanosheets in electrocatalytic production of hydrogen. Nat. Mater. 15, 10031009 (2016).

7. Dungey, K. E., Curtis, M. D. & Penner-Hahn, J. E. Structural characterization and thermal stability of MoS2 intercalation compounds. Chem. Mater. 10, 21522161 (1998).

8. Huang, X. et al. Solution-phase epitaxial growth of noble metal nanostructures on dispersible single-layer molybdenum disulde nanosheets. Nat. Commun. 4, 1444 (2013).

9. Benavente, E. Intercalation chemistry of molybdenum disulde. Coord. Chem. Rev. 224, 87109 (2002).

10. Zhan, B. Z. et al. Zeolite-conned nano RuO2: a green, selective, and efcient catalyst for aerobic alcohol oxidation. J. Am. Chem. Soc. 125, 21952199 (2003).

11. Chen, W., Fan, Z., Pan, X. & Bao, X. Effect of connement in carbon nanotubes on the activity of Fischer-Tropsch iron catalyst. J. Am. Chem. Soc. 130, 94149419 (2008).

12. Chen, Z., Guan, Z., Li, M., Yang, Q. & Li, C. Enhancement of the performance of a platinum nanocatalyst conned within carbon nanotubes for asymmetric hydrogenation. Angew. Chem. Int. Ed. 50, 49134917 (2011).

13. Luo, J., Gao, J., Wang, A. & Huang, J. Bulk nanostructured materials based on two-dimensional building blocks: a roadmap. ACS Nano 9, 94329436 (2015).

14. Deng, D. et al. Catalysis with two-dimensional materials and their heterostructures. Nat. Nanotechnol. 11, 218230 (2016).

8 NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms14548 ARTICLE

15. Reed, C. A. & Bolskar, R. D. Discrete fulleride anions and fullerenium cations. Chem. Rev. 100, 10751120 (2000).

16. Suzuki, T. et al. Redox properties of organofullerenes. J. Am. Chem. Soc. 116, 13591363 (1994).

17. Xie, Q., Perez-Cordero, E. & Echegoyen, L. Electrochemical detection of C606

and C706 : enhanced stability of fullerides in solution. J. Am. Chem. Soc. 114, 39783980 (1992).

18. Zheng, J. et al. High yield exfoliation of two-dimensional chalcogenides using sodium naphthalenide. Nat. Commun. 5, 2995 (2014).

19. Jiang, Z. Z., Wang, Z. B., Chu, Y. Y., Gu, D. M. & Yin, G. P. Carbon riveted microcapsule Pt/MWCNTs-TiO2 catalyst prepared by in situ carbonized glucose with ultrahigh stability for proton exchange membrane fuel cell. Energy Environ. Sci. 4, 25582566 (2011).

20. Liu, Y. et al. Thermal oxidation of WSe2 nanosheets adhered on SiO2/Si substrates. Nano Lett. 15, 49794984 (2015).

21. Robinson, I. & Harder, R. Coherent X-ray diffraction imaging of strain at the nanoscale. Nat. Mater. 8, 291298 (2009).

22. Welzel, U., Ligot, J., Lamparter, P., Vermeulenb, A. C. & Mittemeijer, E. J. Stress analysis of polycrystalline thin lms and surface regions by X-ray diffraction. J. Appl. Cryst. 38, 129 (2005).

23. Liu, P. et al. Photochemical route for synthesizing atomically dispersed palladium catalysts. Science 352, 797800 (2016).

24. Deng, J. et al. Highly active and durable non-precious-metal catalysts encapsulated in carbon nanotubes for hydrogen evolution reaction. Energy Environ. Sci. 7, 19191923 (2014).

25. Wang, H. et al. Electrochemical tuning of vertically aligned MoS2 nanolms and its application in improving hydrogen evolution reaction. Proc. Natl. Acad. Sci. USA 110, 1970119706 (2013).

26. Zheng, Y. et al. Hydrogen evolution by a metal-free electrocatalyst. Nat. Commun. 5, 3783 (2014).

27. Caban-Acevedo, M. et al. Efcient hydrogen evolution catalysis using ternary pyrite-type cobalt phosphosulphide. Nat. Mater. 14, 12451253 (2015).

28. Ma, L., Ting, L. R. L., Molinari, V., Giordano, C. & Yeo, B. S. Efcient hydrogen evolution reaction catalyzed by molybdenum carbide and molybdenum nitride nanocatalysts synthesized via the urea glass route. J. Mater. Chem. A 3, 83618368 (2015).

29. Li, J. S. et al. Coupled molybdenum carbide and reduced graphene oxide electrocatalysts for efcient hydrogen evolution. Nat. Commun. 7, 11204 (2016).

30. Chia, X., Ambrosi, A., Sofer, Z., Luxa, J. & Pumera, M. Catalytic and charge transfer properties of transition metal dichalcogenides arising from electrochemical pretreatment. ACS Nano 9, 51645179 (2015).

31. Ting, L. R. L. et al. Catalytic activities of sulfur atoms in amorphous molybdenum sulde for the electrochemical hydrogen evolution reaction. ACS Catal. 6, 861867 (2016).

32. Herrero, E., Buller, L. J. & Abruna, H. D. Underpotential deposition at single crystal surfaces of Au, Pt, Ag and other materials. Chem. Rev. 101, 18971930 (2001).

33. Chen, D. et al. Determining the active surface area for various platinum electrodes. Electrocatalysis 2, 207219 (2011).

34. Garca, G. & Koper, M. T. M. Stripping voltammetry of carbon monoxide oxidation on stepped platinum single-crystal electrodes in alkaline solution. Phys. Chem. Chem. Phys. 10, 38023811 (2008).

35. Schlaup, C. & Horch, S. Study of underpotential deposited Cu layers on Pt(111) and their stability against CO and CO2 in perchloric acid. Phys. Chem. Chem.

Phys. 15, 1965919664 (2013).36. Earnshaw, A. & Greenwood, N. N. in Chemistry of the Elements (eds Earnshaw, A, Greenwood, NN.) 2nd edn, 68106 (Butterworth-Heinemann, 1997).

37. Aurbach, D. & Weissman, I. in Nonaqueous electrochemistry (ed. Aurbach D) 152 (CRC Press, 1999).

38. Jarek, R. L. et al. Solvation of Li by acetone, THF, and diethyl ether in the gas phase and the ion molecule association mechanism. J. Phys. Chem. A 104,

22302237 (2000).39. Yao, Y. et al. Graphene cover-promoted metal-catalyzed reactions. Proc. Natl Acad. Sci. USA 111, 1702317028 (2014).

40. Zhang, F. et al. Tuning the redox activity of encapsulated metal clusters via the metallic and semiconducting character of carbon nanotubes. Proc. Natl Acad. Sci. USA 110, 1486114866 (2013).

41. Xiao, J., Pan, X., Guo, S., Ren, P. & Bao, X. Toward fundamentals of conned catalysis in carbon nanotubes. J. Am. Chem. Soc. 137, 477482 (2015).

Acknowledgements

We thank National Research Foundation, Primes Minister Ofce for support under the mid-sized research centre (CA2DM). We acknowledge Ms Zilu Niu and WinTech Nano-Technology for ToF-SIMS measurement, Dr Yonghua Du for sample preparation in the EXAFS experiment, Ms Qianqian Hu for her assistance with the XPS,Ms Guangrong Zhou for TEM cross-section sample preparation, Dr. Ping Yang for 1D GIXRD measurements and Prof Richard D. Webster, Dr Zheng Long Lim, Mr Hejian Zhang and Dr Guo-hong Ning for their contributions on EPR. Z.C. thanks the NGS Scholarship for support. K.L. appreciates scholarship support by Solar Energy Research Institute of Singapore (SERIS).

Author contributions

Z.C., K.L. and K.P.L. conceived the research and wrote the draft. Z.C. and K.L. synthesized the materials and performed the electrochemical measurements. Z.C. andX.Z. conducted TEM characterization and data analysis. S.M. and B.S.Y. performed the CO and Cu tests. W.T. and B.T. assisted in the materials characterization and data analysis. L.Z. performed the EXAFS measurement, M.L. performed the GIXRD measurement and L.D. acquired the SAXS patterns. All authors discussed and commented on this manuscript.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Chen, Z. et al. Interface conned hydrogen evolution reaction in zero valent metal nanoparticles-intercalated molybdenum disulde. Nat. Commun.8, 14548 doi: 10.1038/ncomms14548 (2017).

Publishers note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afliations.

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

r The Author(s) 2017

NATURE COMMUNICATIONS | 8:14548 | DOI: 10.1038/ncomms14548 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 9