Amorphous Molybdenum Sulfide on GrapheneâCarbon Nanotube

Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI

Article

Amorphous Molybdenum Sulfide on Graphene-Carbon Nanotube Hybrids as Highly Active Hydrogen Evolution Reaction Catalysts Kien-Cuong Pham, Yung-Huang Chang, David Mcphail, Cecilia Mattevi, Andrew Thye Shen Wee, and Daniel H.C. Chua ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.5b09690 • Publication Date (Web): 11 Feb 2016 Downloaded from http://pubs.acs.org on February 12, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Amorphous Molybdenum Sulfide on Graphene-Carbon Nanotube Hybrids as Highly Active Hydrogen Evolution Reaction Catalysts Kien-Cuong Pham a,b, Yung-Huang Chang c, David S. McPhail b, Cecilia Mattevi b, Andrew T.S. Wee a,d and Daniel H.C. Chua c,*

a

NUS Graduate School for Integrative Sciences and Engineering (NGS), National University of Singapore, 28 Medical Drive, Singapore 117456, Singapore

b

Department of Materials, Imperial College London, Exhibition Road, London, SW7 2AZ, UK

c

Department of Materials Science and Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117576, Singapore

d

Department of Physics, National University of Singapore, 2 Science Drive 3, Singapore 117542, Singapore

Email addresses: [email protected] (Kien-Cuong Pham); [email protected] (Yung-Huang Chang); [email protected] (David S. McPhail); [email protected] (Cecilia Mattevi); [email protected] (Andrew T.S. Wee); and [email protected] (Daniel H.C. Chua)

* Corresponding Author: Tel: +65-6516-8933, Fax: +65-6776-3604, Email address: [email protected] (Daniel H.C. Chua)

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 41

Abstract In this study, we report on the deposition of amorphous molybdenum sulfide (MoSx, with x ≈ 3) on a high specific surface area conductive support of Graphene-Carbon Nanotube hybrids (GCNT) as the Hydrogen Evolution Reaction (HER) catalysts. We found that the high surface area GCNT electrode could support the deposition of MoSx at much higher loadings compared with simple porous carbon paper or flat graphite paper. The morphological study showed that MoSx was successfully deposited on and was in good contact with the GCNT support. Other physical characterization techniques suggested the amorphous nature of the deposited MoSx. With a typical catalyst loading of 3 mg cm–2, an overpotential of 141 mV was required to obtain a current density of 10 mA cm–2. A Tafel slope of 41 mV decade–1 was demonstrated. Both measures placed the MoSx-deposited GCNT electrode among the best performing molybdenum sulfide-based HER catalysts reported to date. The electrode showed a good stability with only a 25 mV increase in overpotential required for a current density of 10 mA cm–2, after undergoing 500 potential sweeps with vigorous bubbling present. The current density obtained at –0.5 V vs. SHE (Standard Hydrogen Electrode potential) decreased less than 10% after the stability test. The deposition of MoSx on high specific surface area conductive electrodes demonstrated to be an efficient method to maximize the catalytic performance towards HER.

Keywords: Amorphous, Molybdenum Sulfide, Graphene, Carbon Nanotube, Hydrogen Evolution Reaction, HER, Catalyst


Page 3 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1.

Introduction As energy security and climate change have become among the greatest global

concerns, hydrogen economy has been proposed as a sustainable alternative to our current fossil fuel-based economy.

1, 2

In this ideal scenario, hydrogen functions as an effective

energy storage medium for renewable energy power sources, especially the intermittent ones such as solar or wind energy.

1, 3-5

Hydrogen can be produced directly from water through

processes such as photoelectrochemical water splitting or water electrolysis.

6

Both of these

electrochemical processes involve the Hydrogen Evolution Reaction (HER), for which catalysts are generally used to enhance the conversion efficiency. Currently, the most active HER catalysts are precious metal-based catalysts. 6 However, both the high cost and scarcity of these precious metals have been limiting their uses. Non-precious catalysts are therefore critical for a large scale implementation of the hydrogen economy. Molybdenum sulfide has recently emerged as a very promising class of non-precious HER catalyst with high catalytic activity and good stability in acidic electrolytes. The atomic basal plane of crystalline molybdenum disulfide (MoS2) is known to be inactive towards the HER.

7

However, after it was identified that the edge sites of crystalline MoS2 are

catalytically active towards the HER,

8, 9

much effort has been paid to engineer the fine

nanostructure of crystalline MoS2 to maximize the edge exposure. Notable examples include MoS2 grown on reduced graphene oxide, 10-13 MoS2/MoO3-x nanowires, 14 mesoporous double gyroid MoS2,

15

defect-rich MoS2 nanosheets,

16

edge oriented MoS2,

17-19

and others.

20-24

These studies have successfully demonstrated the high activity of the defect sites present along MoS2 edge planes. Nevertheless, even with these meticulous nanostructure engineering efforts, the active edge sites of crystalline MoS2 are hardly abundant enough for catalytic reactions, resulting in a low total electrode catalytic activity. On the other hand, amorphous molybdenum sulfide (MoSx) intrinsically possesses abundance of defect sites. 25 In contrast to



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 41

the edge sites of crystalline MoS2, the HER-active defect sites of MoSx are present throughout the entire surface of the catalyst and are readily available for catalytic reactions. The high catalytic activity towards HER of MoSx has been well demonstrated in the literature.

25-30

In

addition, MoSx can be synthesized easily at room temperature using wet chemical methods 27, 30

or electrodeposition methods. 6, 25 Molybdenum sulfide, however, intrinsically has low electronic conductivity. 21, 22, 31 The

slow electron transport in molybdenum sulfide may become a bottleneck and severely deteriorate the overall catalytic activity towards HER. 6, 24, 27, 32 Coupling molybdenum sulfide with highly conductive supports is an effective strategy to enhance the electron transport towards the active sites of catalyst. Examples include MoS2 deposited on reduced graphene oxide, 10-13 MoSx deposited on carbon nanotubes, 20, 27 MoSx deposited on graphene-protected Ni foam,

28

and so on. Besides, the HER reaction requires constant reactant supply and

product removal. The mass transport inside the electrode is critical for a high device performance, especially at high current density ranges. Graphene and reduced graphene oxide have two-dimensional morphology and have a tendency to re-stack. The re-stacking of graphene and reduced graphene oxide may hinder the mass transport inside the electrode. Ensuring both efficient electron transport and mass transport is important for an optimal electrode performance. In our previous study, we reported the fabrication of an integrated, binder-free, porous, high specific surface area and conductive electrode made of Graphene-Carbon Nanotube hybrids (GCNT) grown directly on Toray carbon paper.

33

In this nanostructure, CNTs were

grown directly on carbon paper, and subsequently, free-standing graphene was directly grown on the CNT scaffolds. The growth of graphene directly on the porous CNT scaffolds effectively eliminated the re-stacking issue of regular 2D graphene, and at the same time provided more secured anchor points for deposited catalyst materials than CNTs alone. The


Page 5 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


reported structure can be suitable for the deposition of MoSx. The GCNT hybrids can serve as high surface area and conductive supports for the deposition of MoSx. Carbon paper serves as both a mechanical support and a current collector. As the GCNT hybrids were grown directly on the carbon paper current collector, the electron transport from MoSx to the current collector is enhanced. The open and porous structures of both GCNT and carbon paper current collector allow efficient mass transport from both sides of the electrode. In this study, MoSx was deposited on the GCNT support using an electrochemical deposition method at room temperature. The electrodeposition method was chosen so that MoSx would be selectively deposited on electrically accessible sites of the electrode. The electronic transport towards the deposited MoSx would be therefore guaranteed. In contrast, if MoSx was deposited by other physical and chemical methods, part of the deposited MoSx might be located at electrically insulating sites (such as self-insulated sites of a thick MoSx film), and was therefore not catalytically active.

2.

Materials and methods

2.1. Materials Diammonium

disulfido(dithioxo)molybdenum

(also

known

as

ammonium

tetrathiomolybdate, (NH4)2MoS4, 99.97%) was purchased from Sigma-Aldrich. Sulfuric acid (H2SO4) was purchased from Merck. All materials were used as received without further purification. Deionized water (15 MΩ cm, Millipore) was used for all solution preparations. 2.2. Growth of the GCNT hybrids The GCNT hybrids were grown on carbon paper to form an integrated, binder-free, high surface area and conductive electrode (GCNT/CP) using the chemical vapor deposition methods. The details of growth process and fabricated structures have been previously



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

described elsewhere.

33

Page 6 of 41

To summarize, the GCNT hybrids were grown directly on Toray

carbon paper using a two-step procedure. In the first step, CNTs were grown directly on a catalyzed carbon paper using the thermal chemical vapor deposition technique. In the second step, free-standing graphene was grown directly on the CNT scaffolds without any additional catalyst, using the radio frequency plasma enhanced chemical vapor deposition technique. 2.3. Deposition of MoSx on the GCNT hybrids MoSx was deposited on the GCNT/CP electrode by an electrodeposition method based on the reported literature

26

with additional customizations. The deposition was performed in

a three-electrode electrochemical cell with the GCNT/CP electrode as the working electrode, a platinum foil as counter electrode, a Ag/AgCl (saturated KCl) reference electrode (EAg/AgCl (sat. KCl)

= 0.2 V vs. Standard Hydrogen Electrode potential (SHE)) and a 20 mM solution of

(NH4)2MoS4 in deionized water as electrolyte and MoSx precursor. The potentiostatic anodic deposition was performed by holding the working electrode potential at 0.7 V vs. SHE. The loading of MoSx catalyst was controlled by controlling total amount of charge transferred during the deposition process. 2.4. Reductive activation of the MoSx-deposited GCNT hybrids To activate the catalytic activity towards HER, the MoSx/GCNT/CP working electrode underwent 10 reductive linear potential sweeps performed from 0.2 V to –0.5 V vs. SHE in a second three-electrode electrochemical cell, with a platinum rod counter electrode, a Ag/AgCl (3 M KCl) reference electrode (EAg/AgCl

(3 M KCl)

= 0.21 V vs. SHE), and a 0.5 M H2SO4

electrolyte solution. The potential sweeps were performed at a scan rate of 5 mV s–1. In all subsequent discussions, MoSx/GCNT/CP in general refers to this activated condition of the electrode, except otherwise specified.


Page 7 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.5. Electrochemical activity measurements of MoSx for HER The catalytic activity measurements of the MoSx/GCNT/CP electrode were performed in the same three-electrode cell configuration as in the reductive activation step. The polarization curves were obtained by linear potential sweeps from 0.2 V to –0.5 V vs. SHE at a scan rate of 5 mV s–1. Electrochemical Impedance Spectroscopy (EIS) experiments were measured at –0.15 V vs. SHE in a frequency range of between 10 kHz and 0.05 Hz with an AC perturbation voltage of 5 mV. The cyclic voltammetry experiments were performed in a potential window of between 0 and 0.3 V vs. SHE at scan rates of 10, 20, 50 and 100 mV s–1. The stability test was performed using 500 potential sweeps from 0.2 V to –0.5 V vs. SHE at a scan rate of 100 mV s–1. For comparison purposes, MoSx was also deposited on a simple porous Toray carbon fiber paper and a flat graphite paper using the same procedure as for the MoSx/GCNT/CP. These electrodes underwent the same activation and polarization tests as for the MoSx/GCNT/CP. The polarization curves were also obtained for a plain carbon fiber paper electrode (without any MoSx deposited) and a platinum foil electrode for references. All electrochemical depositions and measurements were performed with an Autolab potentiostat/galvanostat (PGSTAT302N, Metrohm) equipped with a frequency response analyzer (FRA2.V10) and an analog scan generator (SCANGEN). All potentials were reported against SHE. A schematic illustration of the experimental setups used throughout the electrochemical deposition and measurements is shown in Fig S1 (Supporting Information). 2.6. Material characterization The morphology of the MoSx/GCNT/CP electrode was characterized with Field Emission Scanning Electron Microscopy (FE-SEM, JEOL JSM-6700F). The chemical analysis was obtained using X-ray Photoelectron Spectroscopy (XPS) (Omicron 7channeltron analyzer) with an Al Kα X-ray source (1486.7 eV photon energy, Omicron X-ray 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 41

twin anode source). Constant analyzer energy (CAE) modes of 50 eV and 20 eV pass energy were used for survey scan and high resolution scans, respectively. The crystallinity of the hybrids was assessed with Raman spectroscopy (Renishaw inVia Raman microscope, 532 nm laser). The specific surface area of the GCNT hybrids was measured with Brunauer–Emmett– Teller (BET) surface area and porosity analysis (ASAP 2020 Surface Area and Porosity Analyzer, Micromeritics) using N2 as the adsorbate gas.

3.

Results and Discussions

3.1. Deposition and reductive activation of the MoSx catalysts MoSx was deposited on conductive substrates using an anodic electrodeposition process performed at 0.7 V vs. SHE. Fig 1a shows the deposition current density profiles of these electrodes during the deposition of MoSx. The deposition current density of all three samples quickly decayed in the first few minutes and then slowly approached the limiting current density. The deposition of MoSx became slower with additional MoSx deposited. This may be due to the low electronic conductivity of MoSx that limited the current density as the deposited MoSx film thickened. It is evident from Fig 1a that the supported deposition current density increased with higher specific surface area electrodes. In the case of flat graphite electrode, the deposition current density quickly approached zero within five minutes. This quick current decay limited the practically maximum catalyst loading on a flat electrode. Whilst a similar trend happened to higher specific surface area electrodes, the limiting current densities were approached at longer deposition durations. In the case of a porous carbon fiber paper electrode, the deposition current density approached zero after approximately 40 minutes.


Page 9 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. MoSx catalysts preparation process. (a) Deposition current density j as a function of deposition time t during the electrodeposition of MoSx, with (inset) a magnified representation of the first five minutes of deposition. (b) Polarization curves (current density j as a function of working electrode potential E) during the reductive activation process of the MoSx/GCNT/CP-3 electrode, reported on iR-corrected potential and (inset) non-iR-corrected potential scales.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 41

With an even higher specific surface area, the GCNT/CP electrode supported a much higher deposition current density, and therefore much higher possible catalyst loadings. With similar deposition times of approximately 40 minutes, catalyst loadings of 0.15, 0.75 and 3 mg cm–2 were obtained on a flat graphite paper, a porous carbon fiber paper and a GCNT/CP electrode, respectively. The GCNT hybrids demonstrated that a high specific surface area electrode is essential for the deposition of a high catalyst loading. The high specific surface area of GCNT/CP can be demonstrated by the BET specific surface area and surface area enhancement factor (or “roughness factor”, defined as the ratio between the microscopic surface area and the projected geometric surface area of the electrode). The BET surface area and roughness factor of the porous carbon paper and the GCNT/CP electrode are provided in Table 1.

Table 1. The BET specific surface area and roughness factor of carbon paper and GCNT/CP electrodes. BET specific surface area m² g–1

Roughness factor (*)

0.26 31.5 Carbon paper (**) 1577 12.3 GCNT/CP (***) 214 ––– GCNT (*) Defined as the ratio between the microscopic surface area and the projected geometric surface area of the electrode. Areal density of carbon paper is measured at 12.1 mg cm–2. (**) Average BET specific surface area, based on the total weight of GCNT and Carbon paper. (***) Calculated BET specific surface area, based on the measured weight content (5.62 wt%) of GCNT hybrids in the GCNT/CP electrode.

As shown in Table 1, a simple carbon paper may sometimes be considered a high surface area electrode compared to a flat electrode geometry, with a roughness factor of 31.5. However, with the growth of GCNT hybrids on the carbon paper, the GCNT/CP electrode showed a tremendous enhancement in specific surface area, with a roughness factor of 1577, 10 ACS Paragon Plus Environment

Page 11 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


which was ≈ 50 times higher than that of a simple porous carbon paper. The enhancement in microscopic surface area is beneficial for a good dispersion of MoSx and efficient electron transport towards active sites of the MoSx catalyst. High specific surface area conductive supports such as carbon nanotubes and graphene have been effectively used to support electro-active materials for electrochemical applications. 10-13, 20, 27, 34, 35 After MoSx had been deposited, the MoSx catalysts underwent the reductive activation process. The activation was performed by 10 potential sweeps from 0.2 V to –0.5V vs. SHE. Fig 1b shows the current responses of the MoSx/GCNT/CP-3 electrode during the potential sweeps. MoSx/GCNT/CP-3 denotes a MoSx loading of 3 mg cm–2. The results were representative for all MoSx catalysts. During the first potential scan, whilst there was a strong reductive shoulder peak present between 0 V and approximately –0.2 V, no visible hydrogen evolution was observed in this potential range. We attributed this reductive shoulder peak to the chemical reduction of the catalyst itself. The exact activation mechanism and chemical reduction processes are, however, currently not fully understood. A similar reductive peak in the first potential scan was also reported in previous studies by Hu and co-workers. 6, 25, 26 The authors attributed the reductive peak in the first potential scan to a conversion of S22– to S2–. 25 The first hydrogen gas evolution event was visibly observed in the first potential sweep when the potential scan passed a potential of approximately –0.23 V (without Ohmic loss correction or “non-iR-corrected”), corresponding to approximately –0.18 V in Ohmic loss-corrected (“iR-corrected”) potential. In the subsequent potential scans, the shoulder peak became progressively less pronounced. Hydrogen evolution was observed earlier than in the first potential scan. Generally, within 10 reductive potential sweeps, the HER polarization curves stabilized and signified a completed activation of the MoSx catalysts. In fact, the reductive activation was nothing but the first 10 HER polarization curves before a stable and reproducible HER polarization curve could be obtained.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 41

3.2. Physical characterization of the activated MoSx/GCNT/CP electrode The morphological characterization of the GCNT/CP support and the activated MoSx/GCNT/CP electrode was obtained with FE-SEM and is shown in Fig 2. Fig 2a and 2b show the FE-SEM micrographs at medium and high magnifications of the GCNT/CP electrode prior to the deposition of MoSx. It is evident that the GCNT hybrids possessed a porous and fibrous morphology of CNTs with a high density of exposed graphitic edges of graphene. The hybrids offered a high density of exposed edges, which is not possible for ordinary CNTs where basal graphitic planes are exposed instead. The hybrids were uniform in size and morphology, with an average overall diameter of approximately 100 nm. The high specific surface area and porous nanostructure of the GCNT hybrids are beneficial for a good dispersion of MoSx and efficient mass transport inside the electrode. Fig 2c shows the FESEM micrograph of the activated MoSx/GCNT/CP electrode. MoSx was deposited on the GCNT scaffolds and thickened the fibrous morphology of the GCNT hybrids. MoSx was found to fully coat and was in intimate contact with the conductive GCNT support. The close contact between MoSx and the GCNT support is essential for efficient electron transport towards the active sites of MoSx to support electrochemical reactions. The total thickness of the MoSx/GCNT forest (excluding the thickness of porous carbon paper) was measured 50100 µm. For a better understanding of the hybrid nanostructures, a schematic illustration of the GCNT support and the MoSx/GCNT electrode is provided in Fig 2d.


Page 13 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Morphological characterization with FE-SEM of (a, b) the GCNT/CP support and (c) the activated MoSx/GCNT/CP electrode. (d) A schematic illustration of the GCNT/CP support and the MoSx/GCNT/CP electrode.

The chemical state of the MoSx/GCNT/CP electrode was assessed with XPS. Fig 3a shows the high resolution XPS scans in the Mo 3d and the S 2p regions of spectrum. The Mo 3d spectrum was deconvoluted into three peak sets, including a S 2s peak at the binding energy (BE) of 226.9 eV, a Mo 3d doublet at BE of 229.3 eV – 232.5 eV, and a Mo 3d



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 41

doublet at BE of 231.7 eV – 234.9 eV. Molybdenum was present in two chemical states, denoted as Mo(A) at the higher binding energy and Mo(B) at the lower binding energy. The lower binding energy of Mo(B) can be assigned to the oxidation state of Mo4+ as in MoS2 or MoS3. 30

26

The higher binding energy of Mo(A) can be attributed to the oxidation state of Mo5+

or of Mo4+ in molybdenum oxysulfide Mo(A)SxOy.

6, 26

The S 2p spectrum was

deconvoluted into two doublets at BE of 162.9 eV – 164.1 eV and 161.4 eV – 162.6 eV, denoted as S(A) and S(B), respectively. The S(A) 2p doublet can be attributed to apical S2– or bridging S22–. The S(B) 2p doublet can be attributed to basal plane S2– or terminal S22–.

36

The

chemical states of both Mo and S suggested that the deposited material can be described as mixed valence states of molybdenum sulfide. Using XPS spectra of MoS2 micro-crystals as references, the relative stoichiometry of Mo and S in the MoSx/GCNT/CP was determined as Mo : S = 1 : 3.02. The fact that both Mo and S appeared in multiple chemical states with broad XPS peaks suggested the deposited MoSx closely resembled amorphous MoS3. 36 In order to assess the crystallinity of the MoSx/GCNT/CP, the electrode and GCNT/CP support were examined with Raman spectroscopy. Fig 3b shows the Raman spectra of GCNT/CP support and MoSx/GCNT/CP electrode. The Raman spectrum of GCNT/CP showed typical characteristic peaks D, G, as well as second order peaks 2D, D + G, and 2D’, similar to our previously reported result.

33

As the deposited MoSx coated well the surface of

the GCNT/CP support, the graphitic characteristic peaks of GCNT were greatly suppressed in the Raman spectrum of MoSx/GCNT/CP electrode. The Raman spectrum of MoSx/GCNT/CP was dominated by additional bands from MoSx with Raman shift of between 150 cm–1 and 1000 cm–1. The Raman spectrum of MoSx did not show characteristic peaks E12g and A1g corresponding to crystalline MoS2, but instead showed broad band between 150 cm–1 and 1000 cm–1. The Raman spectrum resembled that of amorphous MoS3. 36, 37


Page 15 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Physical characterization of the activated MoSx/GCNT/CP electrode with (a) XPS high resolution scans in Mo 3d and S 2p regions, and (b) Raman spectroscopy, with inset shows the magnified view of the Raman spectrum between 150 and 1000 cm–1 Raman shift.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 41

3.3. Electrochemical activity of MoSx for HER The catalytic activity towards HER of MoSx was assessed from two perspectives, including: (1) total electrode activity (measured with polarization curve and EIS) and (2) intrinsic activity (or Turnover frequency (TOF), measured with cyclic voltammetry).

Total electrode activity of MoSx towards HER The total electrode activity of MoSx towards HER was assessed with polarization curves and EIS. Fig 4 shows the HER polarization curves of the catalysts. MoSx/flat graphite, MoSx/carbon paper and MoSx/GCNT/CP-3 electrodes were prepared as previously described. For additional comparisons, MoSx/GCNT/CP-1.5, MoSx/GCNT/CP-4.5 (denote MoSx loadings of 1.5 mg cm–2 and 4.5 mg cm–2, respectively), a plain carbon paper (without any catalysts deposited) and a Pt foil electrode were also prepared. Fig 4a shows the polarization curves of these electrodes reported on iR-corrected potential scale. The Ohmic losses were calculated based on the ohmic resistance extracted from EIS spectra of these electrodes. Detailed calculations will not be presented here.


Page 17 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. HER polarization curves of catalysts presented on (a) linear scale and (b) Tafel plot (electrode overpotential η as a function of current density j in log scale).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 41

As expected, the plain carbon paper electrode was inert towards HER, with nearly zero current density. On the other hand, the platinum foil electrode, being a precious metal catalyst, was highly active with an early onset of HER. Among the MoSx catalysts, it was evident that the MoSx/GCNT/CP electrodes showed much earlier onsets of HER than MoSx catalysts deposited on lower surface area electrodes, including carbon paper and flat graphite paper. Regardless of MoSx catalyst loadings of between 1.5 and 4.5 mg cm–2, the MoSx/GCNT/CP electrodes showed relatively similar performances and were all much better than the MoSx/flat graphite and MoSx/carbon paper electrodes. It is noteworthy that, at overpotentials of above 200 mV, the MoSx/GCNT/CP electrodes produced higher current density than the platinum foil electrode did. This is mainly due to the mass transport effect. It was visually observed that gas bubbles produced on the MoSx/GCNT/CP electrodes generally had sub-millimeter sizes and easily detached from the electrode surface. The phenomenon is probably due to the good hydrophilicity of the MoSx, especially after being deposited on a porous nanostructure of GCNT hybrids. On the other hand, the platinum foil produced large hydrogen bubbles that adhered relatively strongly on the surface of the platinum electrode and blocked part of the active electrode surface. For a better extraction of total electrode activity parameters, the polarization curves were presented in Tafel plots, as shown in Fig 4b. The red-colored segments denote the ranges in which the Tafel plot linear fittings were performed and Tafel parameters were extracted. From Fig 4, the following parameters were extracted for comparisons: overpotential required to reach a current density of 10 mA cm–2 (η10), Tafel slope and exchange current density j0. These parameters are summarized in Table 2.


Page 19 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Extracted parameters from the HER polarization curves and Tafel plots.

MoSx/flat graphite MoSx/carbon paper MoSx/GCNT/CP-1.5 MoSx/GCNT/CP-3 MoSx/GCNT/CP-4.5 Pt foil Degraded MoSx/GCNT/CP-3

Tafel slope mV decade–1

Exchange current density j0 mA cm–2

η10 mV

56.4 42.9 39.3 41.0 41.8 28.0 45.7

2.34 × 10–4 1.06 × 10–3 1.82 × 10–3 3.52 × 10–3 4.77 × 10–3 1.78 × 10–1 2.31 × 10–3

262 170 146 141 138 49 166

Table 3. A comparison of catalytic activity with representative molybdenum sulfide catalysts. (Adapted with permission from ref.

7

Copyright © 2014, American Chemical

Society) Reference

η10 mV

MoO3-MoS2 nanowires

14

254

Electrodeposited amorphous MoS3

25

242

1T – MoS2

38

207

Double gyroid MoS2

15

206

Wet-chemical amorphous MoSx

30

200

MoSx/Carbon paper

29

194

[Mo3S13] /Graphite Paper

39

174

LixMoS2/Carbon paper

40

168

MoS2/Reduced graphene oxide

10

154

MoSx/Piranha Carbon paper

29

152

7

149

This work

138

Li-MoS2/Carbon Paper

41

110

MoSx/N-CNT

27

110

This work

49

2–

2–

[Mo3S13] /anodized Graphite Paper MoSx/GCNT/CP-4.5

Pt foil



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 41

As shown in Table 2, the MoSx/flat graphite electrode showed a moderate performance with η10 = 262 mV. The required overpotential decreased significantly when MoSx was deposited on higher specific surface area electrodes, with η10 = 170 mV required for the MoSx/carbon paper electrode. The required overpotential decreased even further when MoSx was deposited on GCNT hybrids, with η10 of as low as 138 mV for the MoSx/GCNT/CP-4.5 electrode. This is among the lowest required overpotentials reported to date for molybdenum sulfide-based HER catalysts. A comparison of the required overpotential with the state of the art molybdenum sulfide catalysts is provided in Table 3. The required overpotential for MoSx/GCNT/CP decreased as the catalyst loading increased. However, the overpotential decreases were relatively small compared to the significant overpotential reduction over the MoSx catalysts deposited on low surface area electrodes. The differences implied that the enhanced catalytic activity of MoSx/GCNT/CP was beyond catalyst loading. This can be demonstrated further with the Tafel parameters. The exchange current density for all MoSx electrodes scaled almost linearly with the catalyst loading. However, there were large differences in Tafel slope for MoSx deposited on different substrates. With a low specific surface area electrode of a flat graphite paper, the Tafel slope was relatively high at 56.4 mV decade–1 even for a low catalyst loading of 0.15 mg cm–2. The Tafel slope reduced to 42.9 mV decade–1 for MoSx/carbon paper. The Tafel slope reduced even further to between 39.3 and 41.8 mV decade–1 for MoSx/GCNT/CP electrodes, even with much higher catalyst loadings of between 1.5 and 4.5 mg cm–2. In general, for a same substrate, the Tafel slope increases as the MoSx loading increases, due to the slow electron transport in MoSx. 6 The higher Tafel slopes even with low catalyst loadings on low specific surface area electrodes, such as carbon paper and flat graphite paper, are probably linked to the low electronic conductivity of MoSx. By depositing MoSx on a high specific surface area conductive support such as the GCNT hybrids, the electron transport towards the catalyst is enhanced. The electrode kinetics are


Page 21 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


therefore promoted. Whilst the MoSx/GCNT/CP also experienced some increase in the Tafel slope as the MoSx loading increased, the effect was relatively minimal. The Tafel slopes of between 39.3 and 41.8 mV decade–1 suggested that the HER reaction on MoSx/GCNT/CP followed the Volmer-Heyrovsky mechanism. 11 EIS is an important electrochemical technique that provides a wealth of information, including kinetic and surface information, of electrochemical reactions in general and HER in particular.

42

The enhanced catalytic activity of the MoSx/GCNT/CP electrodes over that of

the MoSx on low surface area electrodes can be demonstrated by significant reductions in the charge transfer resistance, as characterized by EIS and shown in Fig 5. Fig 5a and 5b show the EIS spectra of MoSx catalysts obtained at an overpotential of η = 150 mV, presented in Nyquist plot (Fig 5a) and Bode plots (Fig 5b). As evidenced from Fig 5a, the EIS spectra of all five MoSx catalyst samples were dominated by a single capacitive semicircle at medium frequency range, suggesting the catalytic reaction was characterized by a single time constant. The single time constant characteristic of the reaction was also evidenced by a single peak at frequency of f ≈ 1 Hz in the Bode phase plot (Fig 5b). Beside the dominant semicircle at medium frequency, the EIS spectra in Fig 5a also contained a minor capacitive semicircle at high frequency region. This minor semicircle was overpotential-independent, as will be evidenced in Fig 5c and 5d. This high frequency semicircle was thus attributed to non-Faradaic origins. Qualitatively, the EIS spectra of MoSx/GCNT/CP showed much smaller semicircles (and therefore higher catalytic activity) than the EIS spectra of MoSx/carbon paper and MoSx/flat graphite electrodes. Fig 5c and 5d show the EIS spectra of MoSx/GCNT/CP-3 electrode measured at three different overpotentials η of 100 mV, 150 mV and 200 mV. The high frequency semicircle was overpotential-independent, as evidenced by the overlap of the three EIS spectra at high frequency range in both Nyquist and Bode plots. As the overpotential increased from 100 mV



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 41

to 200 mV, the medium frequency semicircle shrank dramatically, indicating an increase in reaction rate. Similarly, the peak in Bode phase plot shifted to higher frequency, indicating a shorter reaction time constant as the overpotential increased.

Figure 5. EIS spectra of the MoSx catalysts measured at η = 150 mV, presented on (a) Nyquist plot and (b) Bode plots. EIS spectra of the MoSx/GCNT/CP-3 electrode measured at overpotentials η of 100, 150 and 200 mV, presented on (c) Nyquist plot and (d) Bode plots.


Page 23 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For quantitative comparisons of the charge transfer resistance, all EIS spectra in Fig 5 were fitted using the equivalent circuit presented in Fig 5a (inset). In this circuit, Lcab represents the parasitic inductance caused by connecting cables or measuring instrument, as evidenced by minor inductive responses at high frequency of ≈ 104 Hz. Rohm represents the ohmic resistance, which was mainly attributed to the uncompensated electrolyte resistance. RnF and CPE1 are the resistance and constant phase element corresponding to the minor overpotential-independent high frequency semicircle. These elements were of non-Faradaic origins, and were attributed to the distributed resistance/capacitance inside the electrode or the contact resistance/capacitance. Rct and CPEdl are the charge transfer resistance and the constant phase element representing the double layer capacitance of the electrode. The major fitting parameters for MoSx catalysts are summarized in Table 4.

Table 4. Equivalent circuit fitting parameters for MoSx catalysts.

MoSx/flat graphite MoSx/carbon paper MoSx/GCNT/CP-1.5 MoSx/GCNT/CP-3 MoSx/GCNT/CP-3 MoSx/GCNT/CP-3 MoSx/GCNT/CP-4.5

η mV

Rohm Ω

Rct Ω

CPEdl mF

τ s

150 150 150 100 150 200 150

1.06 1.20 1.21 1.17 1.17 1.12 1.07

1028 13.1 2.67 43.6 2.52 0.553 2.20

4.83 55.4 129 213 185 174 204

4.97 0.73 0.34 9.29 0.47 0.10 0.45

As presented in Table 4, the enhanced catalytic activity of MoSx/GCNT/CP electrodes was clearly demonstrated by the small Rct of between 2.20 Ω and 2.67 Ω (at η = 150 mV) compared to 13.1 Ω and 1028 Ω demonstrated by MoSx/carbon paper and MoSx/flat graphite, respectively. Similarly, the reaction time constant (τ = RctCPEdl) of MoSx/GCNT/CP electrodes at η = 150 mV was of between 0.34 s and 0.47 s, much smaller than 0.73 s and 4.97 s demonstrated by MoSx/carbon paper and MoSx/flat graphite, respectively. The Rct of 23 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 41

MoSx/GCNT/CP-3 decreased significantly from 43.6 Ω to 0.553 Ω as η increased from 100 mV to 200 mV. The reaction time constant τ of MoSx/GCNT/CP-3 decreased significantly from 9.29 s to 0.10 s as η increased from 100 mV to 200 mV. The total catalytic activity of catalyst is related to both the active electrochemical surface area and the characteristic reaction rate. The active electrochemical surface area is determined by both the amount of loaded catalyst and the material dispersion efficiency of the catalyst support. With the same catalyst support of GCNT/CP, MoSx/GCNT/CP catalysts had higher electrochemical surface area (demonstrated by CPEdl in Table 4) as the catalyst loading increased. The CPEdl value of MoSx/flat graphite, on the other hand, was much lower than that of MoSx/GCNT/CP catalysts. The significant difference in CPEdl in this case is related to the higher material dispersion efficiency of the high surface area catalyst support (GCNT/CP). The characteristic reaction rate of catalyst may be influenced by the electron transport rate. As demonstrated in Table 4, MoSx/flat graphite had a reaction time constant of ≈ 10 times higher (indicating a slower reaction rate) than that of the MoSx/GCNT/CP catalysts. The result may be explained by the slower electron transport due to the lower material dispersion efficiency in the MoSx/flat graphite catalyst.

Intrinsic activity of MoSx towards HER The per-site intrinsic catalytic activity of MoSx catalyst was assessed by a Turnover frequency (TOF) estimation. Specifically, the MoSx/GCNT/CP-3 electrode was used for the estimation. We followed an estimation process proposed by Jaramillo and co-workers.

30

To

summarize, the estimation steps were: (a)

Estimated a non-Faradaic capacitance of the MoSx/GCNT/CP-3 electrode.


Page 25 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(b)

Estimated a roughness factor of the MoSx/GCNT/CP-3 electrode. Roughness factor was estimated by the ratio between the non-Faradaic capacitance of the MoSx/GCNT/CP-3 electrode and that of a flat MoS2 standard electrode.

(c)

Estimated a surface site density of the MoSx/GCNT/CP-3 electrode by multiplying the roughness factor with the surface site density of the flat MoS2 standard electrode.

(d)

Estimated a per-site TOF of the MoSx/GCNT/CP-3 electrode as the total TOF at a current density of interest divided by the surface site density of the MoSx/GCNT/CP-3 electrode.

The capacitance of the MoSx/GCNT/CP electrode was estimated with cyclic voltammetry performed at various scan rates in a potential window of 0-0.3 V vs. SHE. The cyclic voltammograms are shown in Fig 6a. The cyclic voltammograms displayed pseudorectangular shapes without any obvious Faradaic peaks. However, the demonstrated capacitive behavior of the MoSx/GCNT/CP electrode was not a pure non-Faradaic response, as the current density did not scale linearly with the scan rate. It was suggested that the current density scaled linearly with scan rate in a pure non-Faradaic response, and scaled linearly with the square root of scan rate in a pure Faradaic response.

30, 43

The

MoSx/GCNT/CP electrode demonstrated a mixed behavior. In order to deconvolute the contributions from non-Faradaic and Faradaic processes to the total capacitive behavior, we followed a method proposed by Díaz and co-workers. 44 In their method, the current density j measured on the cyclic voltammograms can be described as a current density summation of two processes: j = k1v + k2v1/2, where v is the scan rate, k1 and k2 are scan rate-independent constants representing the contributions from non-Faradaic and Faradaic processes. By rearranging the equation to jv–1/2 = k1v1/2 + k2 and performing a linear fitting for jv–1/2 versus v1/2, the individual capacitive contributions can be estimated. The current density readings (a



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 41

half of the anodic-cathodic current density differences) at 0.15 V vs. SHE were extracted from the cyclic voltammograms. The linear fitting is shown in Fig 6b. By using this estimation method, a non-Faradaic capacitance of 0.204 F cm–2 was estimated. By using the reported value of 60 µF cm–2 for a flat MoS2 standard electrode,

30

a roughness factor of 3393 was

obtained. A surface site density of 4 × 1018 sites cm–2 was estimated. At a current density of 10 mA cm–2 (obtained at η = 141 mV), a per-site TOF of 7.8 × 10–3 s–1 was estimated. At η = 200 mV (corresponding to a current density of 182 mA cm–2), a per-site TOF of 0.14 s–1 was estimated. The per-site TOF at η = 200 mV is in the same order of magnitude with the reported value by Jaramillo.

30

The MoSx/GCNT/CP catalyst has a similar intrinsic per-site

activity to the reported literature.


Page 27 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6. Turnover frequency estimation for the MoSx/GCNT/CP-3 electrode. (a) Cyclic voltammograms measured in a potential window of 0-0.3 V vs. SHE at scan rates of 10, 20, 50 and 100 mV s–1, and (b) deconvolution of the MoSx/GCNT/CP-3 electrode capacitive behavior by a linear fitting of j v–1/2 versus v1/2.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 41

3.4. Electrochemical stability of MoSx Beside the electrocatalytic activity towards HER, electrochemical stability is also an important property of catalysts. The electrochemical stability of the MoSx/GCNT/CP catalyst was probed with 500 potential scans from 0.2 V to –0.5 V vs. SHE. A polarization curve was recorded for the degraded electrode. A comparison of the polarization curves for the MoSx/GCNT/CP-3 electrode before and after the degradation is shown in Fig 7. A modest increase of only 25 mV in η10 was demonstrated by the electrode after the stability test. Even with η10 = 166 mV, the degraded electrode is still considered a very active non-precious HER catalyst. Tafel plot analyses of the degraded electrode estimated a Tafel slope of 45.7 mV decade–1 and an exchange current density of 2.31 × 10–3 mA cm–2, as reported in Table 2. The changes in Tafel slope and exchange current density with respect to the original electrode’s parameters were relatively moderate. One of the possible contributing reasons for the increase in Tafel slope and the decrease in exchange current density is the mechanical detachment of MoSx from the GCNT support, due to the vigorous bubbling present during the stability test. The maximum current density obtained during the potential scan was approximately 300 mA cm–2, which is closer to the operation condition of water electrolyzers than to that of photoelectrochemical cells. In addition, ohmic loss is generally present in practical experiments. The relative degradation in a practical setting was much less significant. As shown in the inset of Fig 7, the current density obtained at a non-iR-corrected potential of – 0.5 V vs. SHE only decreased from 320 mA cm–2 to 290 mA cm–2 after the stability test. The decrease was less than 10%. We consider this an indication of a good stability demonstrated by the MoSx/GCNT/CP electrode.


Page 29 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7. Stability test of the MoSx/GCNT/CP-3 electrode, with polarization curves of the electrode before and after degradation presented on iR-corrected and (inset) non-iR-corrected potential scales.

3.5. Understanding of MoSx active sites through XPS characterization In order to have a better understanding of the chemical states present in MoSx at various steps during the electrochemical characterization procedure, XPS was performed on the MoSx/GCNT/CP-3 electrode at three different states, namely after electrodeposition, after reductive activation and after stability test. To facilitate semi-quantitative elemental analyses, XPS spectra of MoS2 micro-crystals were recorded to determine the instrument-specific relative sensitivity factors for Mo and S. The XPS spectra in Mo 3d and S 2p regions for four samples are shown in Fig 8.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 41

Figure 8. XPS high resolution scans in Mo 3d and S 2p regions of the MoS2 microcrystals and the MoSx/GCNT/CP-3 electrode after MoSx deposition, after reductive activation and after degradation.


Page 31 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In all samples, the XPS spectra in the Mo 3d region can be fitted with one or two sets of Mo 3d doublets, in addition to one S 2s peak. The Mo 3d doublet at higher binding energy is referred to as Mo(A) oxidation state. The Mo 3d doublet at lower binding energy is referred to as Mo(B) oxidation state. The Mo(B) state with Mo 3d5/2 peak at 229.3 ± 0.2 eV was assigned to Mo4+ oxidation state as in MoS2 or MoS3. 26 The Mo(A) state with Mo 3d5/2 peak at 231.6 ± 0.4 eV was assigned to Mo5+ oxidation state

30

or of Mo4+ in molybdenum oxysulfide

MoSxOy. 6, 26 Similarly, in all samples, the S 2p spectra can be fitted with two or three sets of S 2p doublets. The minor peak S(C) at the highest binding energy was assigned to the S6+ oxidation state from the surface oxidation of sample or residual H2SO4. 30 The main S 2p peak can be fitted with one or two doublets. The S 2p doublet at higher binding energy with S 2p3/2 peak at 162.9 ± 0.2 eV is referred to as S(A) state. The S 2p doublet at lower binding energy with S 2p3/2 peak at 161.5 ± 0.3 eV is referred to as S(B) state. The S(A) 2p doublet can be attributed to apical S2– or bridging S22–. The S(B) 2p doublet can be attributed to basal plane S2– or terminal S22–.

36

As expected from crystalline MoS2 spectra, the Mo 3d and S 2p main

peaks of MoS2 micro-crystals contained only Mo4+ and S(B) states. The present chemical states of MoSx/GCNT/CP remained nearly the same throughout the electrochemical tests, with the exception of a minor S(C) peak appeared after the degradation test. The XPS spectra were dominated by Mo(A), Mo(B), S(A) and S(B) mixed states. The relative composition of the present states, however, varied among different sample conditions. A detailed report of the relative stoichiometry for all four samples can be found in Table 5. Comparing the stoichiometry of the MoSx/GCNT/CP sample before and after reductive activation step, we found that the relative stoichiometry of Mo states remained nearly unchanged. The total S/Mo ratio, however, decreased from a sulfur-rich ratio of 3.35 to 3.02, which is close to the nominal formula of MoS3. This decrease in sulfur content came from the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 41

higher binding energy S(A) state. A small fraction of sulfur was lost in the activation process, probably in the form of soluble S2–.

Table 5. Detailed stoichiometry analysis of the present chemical states in the MoSx/GCNT/CP electrode after various stages of electrochemical measurement. Relative stoichiometry with respect to total Mo Stotal

S(A)

S(B)

S(C)

Mo(A)

Mo(B)

Experimentally determined formula (including unknown oxygen composition Oy)

Mo(B) S(B)1.87 S(C)0.13 Oy MoS2 2 (*) 0 (**) 1.87 0.13 0 (**) 1 (MoS2Oy) microcrystals (A) Mo 0.08 Mo(B)0.92 S(A)2.36 S(B)0.99 Oy MoSx/GCNT/CP 3.35 2.36 0.99 0 (**) 0.08 0.92 (MoS3.35Oy) after deposition MoSx/GCNT/CP Mo(A)0.10 Mo(B)0.90 S(A)2.04 S(B)0.99 Oy 3.02 2.04 0.99 0 (**) 0.10 0.90 after reductive (MoS3.02Oy) activation MoSx/GCNT/CP Mo(A)0.28 Mo(B)0.72 S(A)0.58 S(B)0.89 S(C)0.03 Oy 1.50 0.58 0.89 0.03 0.28 0.72 after (MoS1.50Oy) degradation (*) S/Mo = 2 by default, assuming a perfect stoichiometry of the reference sample. (**) Peaks are insignificant to warrant a fitting.

Comparing the stoichiometry of the MoSx/GCNT/CP sample before and after the degradation test, we found that Mo was partially oxidized to Mo(A) state, which constituted 28% of the total Mo content in degraded sample. The total S/Mo ratio decreased significantly from 3.02 to 1.50. The decrease once again mainly came from the S(A) state. The S(A) state appeared to be much less stable than the S(B) state during the continuous polarization sweeps. The low S/Mo stoichiometry and the enhanced Mo(A) state suggested that MoSx was probably partially converted to molybdenum oxysulfide MoSxOy during the stability test. Researchers sometimes used the S 2p XPS spectrum, especially the S(A) peak, to explain the catalytic activity of MoSx towards HER.

28, 45

However, we found that the S 2p XPS

spectrum may not truly reflect the possible catalytic activity of MoSx materials. Firstly, the MoSx/GCNT/CP electrode was not active towards HER until the first activation event. The 32 ACS Paragon Plus Environment

Page 33 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


XPS spectra of electrode before and after activation were, however, highly similar. The activation process may involve some chemical reduction (such as the conversion of S22– to S2– suggested by Hu and co-workers

25

) or bonding rearrangement. However, both S2– and S22–

can contribute to the S(A) and S(B) peak intensities, depending on their actual local bonding configurations. The conversion of sulfur chemical states, if present, may not necessarily be reflected on the changes of S 2p XPS spectrum. The possible conversion of S22– to S2– during the activation process therefore could not be unambiguously verified in this study. Secondly, we found that the S(A) peak intensity may not truly reflect the catalytic activity of MoSx. The decrease in sulfur content, especially from S(A), was much higher than the loss in the catalytic activity, such as from the exchange current density perspective. We suggest that the active sites may arise from both sulfur states S(A) and S(B), or more likely from the S(B) sites such as terminal S22– sites. Nevertheless, as previously mentioned, the S 2p peaks were not exclusively assigned to any specific structural sites. The S 2p spectrum alone, therefore, may not be a reliable prediction of the catalytic activity of MoSx towards HER.

4.

Conclusions In this study, we report the deposition of MoSx on a high specific surface area

conductive electrode made of GCNT hybrids as the HER catalyst. The deposition of MoSx on a high specific surface area electrode demonstrated to be an efficient method to increase the active loading of MoSx while keeping the Tafel slope close to 40 mV decade–1. The good electronic conductivity inside the electrode played a crucial role in the catalytic activity enhancement of MoSx. With this strategy, we reported one of the most active molybdenum sulfide-based HER catalysts to date. We also investigated the changes of MoSx chemical states after the three main steps of the electrochemical measurement process. We found that the S 2p XPS spectrum may not be a reliable prediction of the catalytic activity of MoSx. We



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 41

suggest that cautions should be made when one correlates the XPS spectrum to the catalytic activity of MoSx towards HER.

Associated Content Supporting Information Available: Schematic illustration of the experimental setups used for MoSx electrodeposition, reductive activation and electrochemical activity measurements. This material is available free of charge via the Internet at http://pubs.acs.org.

Acknowledgements K.C.P., D.H.C.C. and A.T.S.W. acknowledge the support from National University of Singapore (NUS) (Grants number: R284-000-123-112, R284-000-120-281 and R-144-000321-112). K.C.P. acknowledges NUS Graduate School for Integrative Sciences and Engineering for the NGS Scholarship. K.C.P. acknowledges Mr Wong How Kwong and Dr Zheng Minrui for their assistance with XPS and Raman spectroscopy, respectively.


Page 35 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References (1)

Barreto, L.; Makihira, A.; Riahi, K. The Hydrogen Economy in the 21st Century: A Sustainable Development Scenario. Int. J. Hydrogen Energy 2003, 28, 267-284.

(2)

Muradov, N. Z.; Veziroğlu, T. N. “Green” Path from Fossil-based to Hydrogen Economy: An Overview of Carbon-neutral Technologies. Int. J. Hydrogen Energy 2008, 33, 6804-6839.

(3)

Marbán, G.; Valdés-Solís, T. Towards the Hydrogen Economy? Int. J. Hydrogen Energy 2007, 32, 1625-1637.

(4)

Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis, N. S. Solar Water Splitting Cells. Chem. Rev. 2010, 110, 6446-6473.

(5)

Abbasi, T.; Abbasi, S. A. ‘Renewable’ Hydrogen: Prospects and Challenges. Renewable Sustainable Energy Rev. 2011, 15, 3034-3040.

(6)

Morales-Guio, C. G.; Hu, X. Amorphous Molybdenum Sulfides as Hydrogen Evolution Catalysts. Acc. Chem. Res. 2014, 47, 2671-2681.

(7)

Benck, J. D.; Hellstern, T. R.; Kibsgaard, J.; Chakthranont, P.; Jaramillo, T. F. Catalyzing the Hydrogen Evolution Reaction (HER) with Molybdenum Sulfide Nanomaterials. ACS Catal. 2014, 4, 3957-3971.

(8)

Hinnemann, B.; Moses, P. G.; Bonde, J.; Jørgensen, K. P.; Nielsen, J. H.; Horch, S.; Chorkendorff, I.; Nørskov, J. K. Biomimetic Hydrogen Evolution: MoS2 Nanoparticles as Catalyst for Hydrogen Evolution. J. Am. Chem. Soc. 2005, 127, 5308-5309.

(9)

Jaramillo, T. F.; Jørgensen, K. P.; Bonde, J.; Nielsen, J. H.; Horch, S.; Chorkendorff, I. Identification of Active Edge Sites for Electrochemical H2 Evolution from MoS2 Nanocatalysts. Science 2007, 317, 100-102.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(10)

Page 36 of 41

Li, Y.; Wang, H.; Xie, L.; Liang, Y.; Hong, G.; Dai, H. MoS2 Nanoparticles Grown on Graphene: An Advanced Catalyst for the Hydrogen Evolution Reaction. J. Am. Chem. Soc. 2011, 133, 7296-7299.

(11)

Liao, L.; Zhu, J.; Bian, X.; Zhu, L.; Scanlon, M. D.; Girault, H. H.; Liu, B. MoS2 Formed on Mesoporous Graphene as a Highly Active Catalyst for Hydrogen Evolution. Adv. Funct. Mater. 2013, 23, 5326-5333.

(12)

Zheng, X.; Xu, J.; Yan, K.; Wang, H.; Wang, Z.; Yang, S. Space-Confined Growth of MoS2 Nanosheets within Graphite: The Layered Hybrid of MoS2 and Graphene as an Active Catalyst for Hydrogen Evolution Reaction. Chem. Mater. 2014, 26, 2344-2353.

(13)

Zhou, W.; Zhou, K.; Hou, D.; Liu, X.; Li, G.; Sang, Y.; Liu, H.; Li, L.; Chen, S. Three-Dimensional Hierarchical Frameworks Based on MoS2 Nanosheets SelfAssembled on Graphene Oxide for Efficient Electrocatalytic Hydrogen Evolution. ACS Appl. Mater. Interfaces 2014, 6, 21534-21540.

(14)

Chen, Z.; Cummins, D.; Reinecke, B. N.; Clark, E.; Sunkara, M. K.; Jaramillo, T. F. Core–shell MoO3–MoS2 Nanowires for Hydrogen Evolution: A Functional Design for Electrocatalytic Materials. Nano Lett. 2011, 11, 4168-4175.

(15)

Kibsgaard, J.; Chen, Z.; Reinecke, B. N.; Jaramillo, T. F. Engineering the Surface Structure of MoS2 to Preferentially Expose Active Edge Sites for Electrocatalysis. Nat. Mater. 2012, 11, 963-969.

(16)

Xie, J.; Zhang, H.; Li, S.; Wang, R.; Sun, X.; Zhou, M.; Zhou, J.; Lou, X. W.; Xie, Y. Defect-Rich MoS2 Ultrathin Nanosheets with Additional Active Edge Sites for Enhanced Electrocatalytic Hydrogen Evolution. Adv. Mater. 2013, 25, 5807-5813.

(17)

Kong, D.; Wang, H.; Cha, J. J.; Pasta, M.; Koski, K. J.; Yao, J.; Cui, Y. Synthesis of MoS2 and MoSe2 Films with Vertically Aligned Layers. Nano Lett. 2013, 13, 13411347.


Page 37 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(18)

Yang, Y.; Fei, H.; Ruan, G.; Xiang, C.; Tour, J. M. Edge-Oriented MoS2 Nanoporous Films as Flexible Electrodes for Hydrogen Evolution Reactions and Supercapacitor Devices. Adv. Mater. 2014, 26, 8163-8168.

(19)

Lu, Z.; Zhu, W.; Yu, X.; Zhang, H.; Li, Y.; Sun, X.; Wang, X.; Wang, H.; Wang, J.; Luo, J.; Lei, X.; Jiang, L. Ultrahigh Hydrogen Evolution Performance of Under-Water “Superaerophobic” MoS2 Nanostructured Electrodes. Adv. Mater. 2014, 26, 26832687.

(20)

Li, X.; Zhu, G.; Kang, Q.; Huang, Z.-D.; Feng, X.; Li, Y.; Liu, R.; Ma, Y. Towards Free-Standing MoS2 Nanosheet Electrocatalysts Supported and Enhanced by N-doped CNT-Graphene Foam for Hydrogen Evolution Reaction. RSC Adv. 2015, 5, 5539655400.

(21)

Zhao, Y.; Xie, X.; Zhang, J.; Liu, H.; Ahn, H.-J.; Sun, K.; Wang, G. MoS2 Nanosheets Supported on 3D Graphene Aerogel as a Highly Efficient Catalyst for Hydrogen Evolution. Chem. - Eur. J. 2015, 21, 15908-15913.

(22)

Nikam, R. D.; Lu, A.-Y.; Sonawane, P. A.; Kumar, U. R.; Yadav, K.; Li, L.-J.; Chen, Y.-T. Three-Dimensional Heterostructures of MoS2 Nanosheets on Conducting MoO2 as an Efficient Electrocatalyst to Enhance Hydrogen Evolution Reaction. ACS Appl. Mater. Interfaces 2015, 7, 23328-23335.

(23)

Yan, Y.; Xia, B.; Ge, X.; Liu, Z.; Wang, J.-Y.; Wang, X. Ultrathin MoS2 Nanoplates with Rich Active Sites as Highly Efficient Catalyst for Hydrogen Evolution. ACS Appl. Mater. Interfaces 2013, 5, 12794-12798.

(24)

Ma, C.-B.; Qi, X.; Chen, B.; Bao, S.; Yin, Z.; Wu, X.-J.; Luo, Z.; Wei, J.; Zhang, H.L.; Zhang, H. MoS2 Nanoflower-Decorated Reduced Graphene Oxide Paper for HighPerformance Hydrogen Evolution Reaction. Nanoscale 2014, 6, 5624-5629.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(25)

Page 38 of 41

Merki, D.; Fierro, S.; Vrubel, H.; Hu, X. Amorphous Molybdenum Sulfide Films as Catalysts for Electrochemical Hydrogen Production in Water. Chem. Sci. 2011, 2, 1262-1267.

(26)

Vrubel, H.; Hu, X. Growth and Activation of an Amorphous Molybdenum Sulfide Hydrogen Evolving Catalyst. ACS Catal. 2013, 3, 2002-2011.

(27)

Li, D. J.; Maiti, U. N.; Lim, J.; Choi, D. S.; Lee, W. J.; Oh, Y.; Lee, G. Y.; Kim, S. O. Molybdenum Sulfide/N-Doped CNT Forest Hybrid Catalysts for High-Performance Hydrogen Evolution Reaction. Nano Lett. 2014, 14, 1228-1233.

(28)

Chang, Y.-H.; Lin, C.-T.; Chen, T.-Y.; Hsu, C.-L.; Lee, Y.-H.; Zhang, W.; Wei, K.H.; Li, L.-J. Highly Efficient Electrocatalytic Hydrogen Production by MoSx Grown on Graphene-Protected 3D Ni Foams. Adv. Mater. 2013, 25, 756-760.

(29)

Laursen, A. B.; Vesborg, P. C. K.; Chorkendorff, I. A High-Porosity Carbon Molybdenum Sulphide Composite with Enhanced Electrochemical Hydrogen Evolution and Stability. Chem. Commun. 2013, 49, 4965-4967.

(30)

Benck, J. D.; Chen, Z.; Kuritzky, L. Y.; Forman, A. J.; Jaramillo, T. F. Amorphous Molybdenum Sulfide Catalysts for Electrochemical Hydrogen Production: Insights into the Origin of their Catalytic Activity. ACS Catal. 2012, 2, 1916-1923.

(31)

Huang, K.-J.; Wang, L.; Zhang, J.-Z.; Wang, L.-L.; Mo, Y.-P. One-Step Preparation of Layered Molybdenum Disulfide/Multi-Walled Carbon Nanotube Composites for Enhanced Performance Supercapacitor. Energy 2014, 67, 234-240.

(32)

Vrubel, H.; Moehl, T.; Gratzel, M.; Hu, X. Revealing and Accelerating Slow Electron Transport in Amorphous Molybdenum Sulphide Particles for Hydrogen Evolution Reaction. Chem. Commun. 2013, 49, 8985-8987.


Page 39 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(33)

Pham, K.-C.; Chua, D. H. C.; McPhail, D. S.; Wee, A. T. S. The Direct Growth of Graphene-Carbon Nanotube Hybrids as Catalyst Support for High-Performance PEM Fuel Cells. ECS Electrochem. Lett. 2014, 3, F37-F40.

(34)

Lee, W. J.; Maiti, U. N.; Lee, J. M.; Lim, J.; Han, T. H.; Kim, S. O. Nitrogen-Doped Carbon Nanotubes and Graphene Composite Structures for Energy and Catalytic Applications. Chem. Commun. 2014, 50, 6818-6830.

(35)

Ghosh, D.; Kim, S. Chemically Modified Graphene Based Supercapacitors for Flexible and Miniature Devices. Electron. Mater. Lett. 2015, 11, 719-734.

(36)

Weber, T.; Muijsers, J. C.; Niemantsverdriet, J. W. Structure of Amorphous MoS3. J. Phys. Chem. 1995, 99, 9194-9200.

(37)

Chang, C. H.; Chan, S. S. Infrared and Raman Studies of Amorphous MoS3 and Poorly Crystalline MoS2. J. Catal. 1981, 72, 139-148.

(38)

Voiry, D.; Salehi, M.; Silva, R.; Fujita, T.; Chen, M.; Asefa, T.; Shenoy, V. B.; Eda, G.; Chhowalla, M. Conducting MoS2 Nanosheets as Catalysts for Hydrogen Evolution Reaction. Nano Lett. 2013, 13, 6222-6227.

(39)

Kibsgaard, J.; Jaramillo, T. F.; Besenbacher, F. Building an Appropriate Active-Site Motif into a Hydrogen-Evolution Catalyst with Thiomolybdate [Mo3S13]2− Clusters. Nat. Chem. 2014, 6, 248-253.

(40)

Wang, H.; Lu, Z.; Xu, S.; Kong, D.; Cha, J. J.; Zheng, G.; Hsu, P.-C.; Yan, K.; Bradshaw, D.; Prinz, F. B.; Cui, Y. Electrochemical Tuning of Vertically Aligned MoS2 Nanofilms and its Application in Improving Hydrogen Evolution Reaction. Proc. Natl. Acad. Sci. 2013, 110, 19701-19706.

(41)

Wang, H.; Lu, Z.; Kong, D.; Sun, J.; Hymel, T. M.; Cui, Y. Electrochemical Tuning of MoS2 Nanoparticles on Three-Dimensional Substrate for Efficient Hydrogen Evolution. ACS Nano 2014, 8, 4940-4947.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(42)

Page 40 of 41

Lasia, A. In Electrochemical Impedance Spectroscopy and its Applications, Springer New York: 2014; Chapter 7, pp 155-175.

(43)

Conway, B. E.; Birss, V.; Wojtowicz, J. The Role and Utilization of Pseudocapacitance for Energy Storage by Supercapacitors. J. Power Sources 1997, 66, 1-14.

(44)

León-Reyes, Á.; Epifani, M.; Chávez-Capilla, T.; Palma, J.; Díaz, R. Analysis of the Different Mechanisms of Electrochemical Energy Storage in Magnetite Nanoparticles. Int. J. Electrochem. Sci. 2014, 9, 3837-3845.

(45)

Hsu, C.-L.; Chang, Y.-H.; Chen, T.-Y.; Tseng, C.-C.; Wei, K.-H.; Li, L.-J. Enhancing the Electrocatalytic Water Splitting Efficiency for Amorphous MoSx. Int. J. Hydrogen Energy 2014, 39, 4788-4793.


Page 41 of 41

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents Graphic:


Amorphous Molybdenum Sulfide on GrapheneâCarbon Nanotube

Recommend Documents

Amorphous Molybdenum Sulfide on GrapheneâCarbon Nanotube