electronic and photocatalytic properties of transition...

ELECTRONIC AND PHOTOCATALYTIC PROPERTIES OF TRANSITION METAL DECORATED MOLYBDENUM DISULFIDE

XINYING SHI

Nano and Molecular Systems Research Unit Faculty of Science University of Oulu University of Oulu Graduate School Finland

Academic Dissertation to be presented with the assent of the Faculty of Science, University of Oulu for public discussion in the Auditorium Saalastinsali, on September 12th, 2018, at 12 o’clock noon.

REPORT SERIES IN PHYSICAL SCIENCES Report No. 120OULU 2018 UNIVERSITY OF OULU

Opponent Prof. Peter Liljeroth, Aalto University School of Science, Finland Reviewers Prof. Toomas Plank, University of Tartu, Estonia Prof. Jian Wang, University of Saskatchewan, Canada Custos Doc. Wei Cao, University of Oulu, Finland ISBN 978-952-62-2007-9 (print) ISBN 978-952-62-2008-6 (electronic) ISSN 1239-4327 JUVENES PRINT Oulu 2018

i

Shi, Xinying: Electronic and photocatalytic properties of transition metal decorated molybdenum disulfide

Nano and Molecular Systems Research Unit Faculty of Science P.O. Box 3000 FI-90014 University of Oulu Finland

Abstract

This thesis is dedicated to realizations and physical understanding of electronic and photocatalytic properties after decorating transition metals to the semiconducting molybdenum disulfide. Synthesized via facile wet chemical methods, the MoS2-Au, MoS2-Au-Ni and MoS2-Ag-Ni composites were formed as binary or ternary compounds. The Au nanoparticles are stably joined to the MoS2 matrix without deteriorating layered structures of the host. After introducing the Au nanoglue as a common buffer, a metallic contact is reached between Ni and MoS2, and attributed to new electron migration channel via MoS2 edge contact. Adapting the Ag as the buffer element can attach the Ni to the basal plane of the MoS2 beside edge contact. The Ni-Ag-MoS2 composite effectively splits water under visible light irradiation and produce hydrogen. The excellent photocatalytic activity is attributed to effective charge migration through dangling bonds at the MoS2-Ag-Ni alloy interface and the activation of MoS2 basal planes.

Key words: inorganic layered crystals, molybdenum disulfide, band structure, metal-semiconductor contact, conductive atomic force microscopy, X-ray photoemission electron microscopy, X-ray photoelectron spectroscopy, photocatalysis, water splitting, hydrogen evolution

ii

Acknowledgements

The present work was carried out in the Nano and Molecular Systems Research Unit (NANOMO), University of Oulu. I would like to thank my principle supervisor Dr. Wei Cao and co-supervisor Prof. Marko Huttula for offering me the chance to accomplish my doctoral study. Special thanks are directed towards Wei for the firm guidance and frequent discussion during the four years. Being both a mentor and friend, his knowledge, patience, optimism and encouragement help to better my future direction and especially his prompt rescue that saved my life when I was evaluating the cleanness of Kuivasjärvi water.

This work has involved close collaborations and thus I would like to thank all the co-authors of the included papers. Especially, I thank Sergei Posysaev and Prof. Matti Alatalo in NANOMO and Prof. Meng Zhang from ECUST for their theoretical supports to my experimental work. Sami Saukko from the Center of Microscopy and Nanotechnology (CMNT) Oulu is gratefully thanked for the great help during the electrical and morphological characterizations. Dr. Yuran Niu and Alexei Zakharov from MAX IV Laboratory have helped a lot during the beamtime in Lund. Prof. Sebastiaan van Dijken and Dr. Diego López González from Aalto University are acknowledged for the VSM measurements and helpful discussion. Prof. Taohai Li from XTU helped the degradation tests and provided inspiring discussion about the photocatalytic experiments. Besides the co-authors, Santtu Heinilehto, Leena Palmu, Pasi Juntunen and Marcin Selent from CMNT are acknowledged for their help in various characterizations. Dr. Satu Ojala and Mika Huuhtanen helped a lot in the specific surface area tests. Paavo Turunen is thanked for his help in the laboratory work. Dr. Kari Jänkälä and Dr. Jack Lin are acknowledged for improving the language of the thesis.

I also want to express my gratitude to my Follow-up Group members, Prof. Timo Fabritius and Dr. Ville-Veikko Telkki, for their kindness and help that ensures the progress of my doctoral training plan. The staff in UniOGS is also acknowledged for the useful instructions during my doctoral study.

I especially thank my parents for providing all the support, understanding and encouragement. Without their support, I wouldn’t have come to this stage.

Oulu, July 2018 Xinying Shi

iii

List of original papers

The present thesis contains an introductory part and the following papers which will be referred in the text by their Roman numbers.

I. W. Cao, V. Pankratov, M. Huttula, X. Shi, S. Saukko, Z. Huang, M. Zhang. Gold nanoparticles on MoS2 layered crystal flakes. Materials Chemistry and Physics, 158, 89−95 (2015). DOI: 10.1016/j.matchemphys.2015.03.041

II. X. Shi, S. Posysaev, M. Huttula, V. Pankratov, J. Hoszowska, J.-Cl. Dousse, F. Zeeshan, Y. Niu, A. Zakharov, T. Li, O. Miroshnichenko, M. Zhang, X. Wang, Z. Huang, S. Saukko, D. L. González, S. van Dijken, M. Alatalo, W. Cao. Metallic contact between MoS2 and Ni via Au nanoglue. Small, 14, 1704526 (2018). DOI: 10.1002/smll.201704526

III. X. Shi, M. Huttula, V. Pankratov, J. Hoszowska, J.-Cl. Dousse, F. Zeeshan, Y. Niu, A. Zakharov, Z. Huang, G. Wang, S. Posysaev, O. Miroshnichenko, M. Alatalo, W. Cao. Quantification of bonded Ni atoms for Ni-MoS2 metallic contact through X-ray photoemission electron microscopy. Microscopy and Microanalysis, 24, 458−459 (2018). DOI: 10.1017/S1431927618014526

IV. X. Shi, M. Zhang, W. Cao, X. Wang, M. Huttula. Efficient photocatalytic hydrogen evolution via activated multilayer MoS2. Manuscript.

V. X. Shi, Z. Huang, M. Huttula, T. Li, S. Li, X. Wang, Y. Luo, M. Zhang, W. Cao. Introducing magnetism into 2D nonmagnetic inorganic layered crystals: a brief review from first-principles aspects. Crystals, 8, 24 (2018). DOI: 10.3390/cryst8010024

All the papers presented above are results of teamwork. In the Paper I, the author contributed to the synthesis and transmission electron microscopy measurements. In the Papers II, III and IV, he performed most of the experimental work and data analysis. He organized the structure of Paper V and collected the synthetic routes introduced therein. The author contributed to most of the writing work for the Papers II, III and IV. He is the corresponding author of the Paper III.

All the original publications are not included in the electronic version of the thesis.

In addition, the author has contributed to the following papers that are not included in the thesis.

iv

Y. Lu, X. Shi, Z. Huang, T. Li, M. Zhang, J. Czajkowski, T. Fabritius, M. Huttula, Wei Cao. Nanosecond laser coloration on stainless steel surface. Scientific Reports, 7, 7092 (2017). DOI: 10.1038/s41598-017-07373-8

H. Pauna, X. Shi, M. Huttula, E. Kokkonen, T. Li, Y. Luo, J. Lappalainen, M. Zhang, W. Cao. Evolution of lithium clusters to superatomic Li3O+. Applied Physics Letters, 111, 103901 (2017). DOI: 10.1063/1.5001700

M. Zhang, X. Shi, X. Wang, T. Li, M. Huttula, Y. Luo, W. Cao. Transition metal adsorbed-doped ZnO monolayer: 2D dilute magnetic semiconductor, magnetic mechanism, and beyond 2D. ACS Omega, 2, 1192−1197 (2017). DOI: 10.1021/acsomega.7b00093

Z. Huang, C. Cai, X. Shi, T. Li, M. Huttula, W. Cao. Leaf-mimicking polymers for hydrophobicity and high transmission haze. Proceedings of the Estonian Academy of Sciences, 66, 444−449 (2017). DOI: 10.3176/proc.2017.4.24

G. Wang, D. Pan, X. Shi, M. Huttula, W. Cao, Y. Huang. Tensile creep characterization and prediction of Zr-based metallic glass at high temperatures. Metals, 8, 457 (2018). DOI: 10.3390/met8060457

v

Abbreviations

AFM atomic force microscopy AQY apparent quantum yield C-AFM conductive atomic force microscopy CBM conduction band minimum CVD chemical vapor deposition DFT density functional theory DMS dilute magnetic semiconductor EELS electron energy loss spectroscopy FET field-effect transistor FWHM full width at half maximum GC gas chromatography HEP hydrogen evolution photocatalyst HER hydrogen evolution reaction ILC inorganic layered crystal I-V current-voltage JFET junction field-effect transistor LED light emitting diode M/S metal-semiconductor MOSFET metal-oxide-semiconductor field-effect transistor NEXAFS near edge X-ray absorption fine structure NHE normal hydrogen electrode NPs nanoparticles OEP oxygen evolution photocatalyst OER oxygen evolution reaction PL photoluminescence QE quantum efficiency SEM scanning electron microscopy SR synchrotron radiation SSA specific surface area TCD thermal conductivity detector TEM transmission electron microscopy TM transition metal TMDs transition metal dichalcogenides UV ultraviolet VBM valence band maximum VSM vibrating sample magnetometry

vi

XAS X-ray absorption spectroscopy XPEEM X-ray photoemission electron microscopy XPS X-ray photoelectron spectroscopy XRD X-ray diffraction

vii

Contents

Abstract ......................................................................................................................... i

Acknowledgements ...................................................................................................... ii

List of original papers ................................................................................................. iii

Abbreviations ............................................................................................................... v

Contents ..................................................................................................................... vii

1 Introduction .............................................................................................................. 1

2 Fundamentals of inorganic layered crystals ............................................................ 3 2.1 Chemical composition and crystal structures ........................................................ 3 2.2 Electronic structures of semiconductors ................................................................. 6

2.2.1 Direct and indirect band structures ............................................................... 7 2.2.2 Electronic band structures .............................................................................. 8

2.3 Tunability of molybdenum disulfide ....................................................................... 9 2.3.1 Transition metal doping ................................................................................. 9 2.3.2 Defect manipulations .................................................................................... 10 2.3.3 Heterogeneous layers ..................................................................................... 10 2.3.4 From nonmagnetic to magnetic ................................................................... 11

3 Spectroscopy techniques ......................................................................................... 13 3.1 Light-matter interaction .......................................................................................... 13

3.1.1 Photon absorption and system relaxation .................................................. 13 3.1.2 Scattering ......................................................................................................... 15 3.1.3 Beer-Lambert Law .......................................................................................... 16

3.2 X-ray photoemission electron microscopy ........................................................... 16 3.2.1 Synchrotron radiation ................................................................................... 16 3.2.2 X-ray photoemission electron microscopy ................................................. 17

3.3 X-ray photoelectron spectroscopy ......................................................................... 19 3.4 UV-visible spectroscopy .......................................................................................... 21

4 Sample preparations ............................................................................................... 23

viii

4.1 Synthetic methods .................................................................................................... 23 4.2 Noble metal nanoparticles ...................................................................................... 24

4.2.1 Size determination ......................................................................................... 24 4.2.2 Size engineering ............................................................................................. 24

5 Metal-semiconductor contacts ............................................................................... 27 5.1 Band alignment at metal-semiconductor interface ............................................. 27

5.1.1 Work function ............................................................................................... 27 5.1.2 Contact formation and the Schottky barrier .............................................. 29

5.2 Ohmic contact and charge injection ..................................................................... 31 5.3 Current-voltage relationships ................................................................................ 33

5.3.1 Field effect transistors ................................................................................... 33 5.3.2 Conductive atomic force microscopy ......................................................... 34

6 MoS2-based photocatalysts ..................................................................................... 37 6.1 Basic photocatalytic principles ............................................................................... 37 6.2 Photocatalyst design ................................................................................................ 39

6.2.1 Active sites of MoS2 ....................................................................................... 39 6.2.2 Enhancing photocatalytic ability ................................................................. 41

6.3 Evaluation of photocatalytic ability ....................................................................... 42 6.3.1 Hydrogen evolution efficiency ..................................................................... 42 6.3.2 Photocatalytic degradation ability ............................................................... 43

6.4 Gas chromatography ............................................................................................... 43

7 Research summary and discussion ......................................................................... 45 7.1 Morphologies............................................................................................................ 45 7.2 Evaluation of metal/semiconductor contacts ....................................................... 46 7.3 Photocatalytic applications ..................................................................................... 49 7.4 Stability of the synthesized composites ................................................................. 51

8 Conclusions and outlook ........................................................................................ 54

Bibliography ............................................................................................................... 55

1

Chapter 1 Introduction

Beyond Carbon and Silicon Carbon is everywhere in our planet, being the main component of organisms, forming the fossil fuels we rely on, and circulating between the biosphere and the environment. In its elemental form, graphite is a native mineral and known for its lubricant applications. Formed with intralayer sp2 hybridization and interlayer van der Waals interaction, individual graphitic layers can easily slide off each other and thus graphite is widely used as pencil cores and solid lubricants.

As a close relative to carbon, silicon has been the most important element in the semiconductor industry which supports our modern civilization. It is the second most-abundant element in the Earth’s crust (28%). With mature production techniques and low cost, the single crystal silicon has become the basis of central processing unit (CPU), memory storages, and debut the revolutionary fashion of the artificial intelligence (AI) applications. Silicon also dominates the market of commercial photovoltaic cells for solar energy conversion.

In the last decades, man-made elemental carbon materials have been focused on low-dimensional forms. Thanks to the layered structure of graphite, graphene was mechanically exfoliated from graphite, leading to a thorough revolution in carbon research [1,2]. As a two-dimensional and honeycomb-lattice carbon film, graphene is the building block of the other graphitic materials: wrapping up into fullerenes, rolling into carbon nanotubes, and stacking into graphite [3,4]. It exhibits extraordinary performance in electrical, optical and catalytic applications.

Subsequently, graphene has inspired studies of enormous analogues composed of inorganic elements. The inorganic layered crystals (ILC) are such composites with similar structural properties to graphene [5,6,7]. Thanks to intralayer and interlayer structures and species varieties, ILC materials have high structural hostability and electronic tunability. They cover a full range of electrical conduction species: insulators (h-BN), semiconductors (2H-phase), and metals (1T-phase). Among them, semiconducting ILCs have drawn great attention on semiconductor applications thanks to their structural and electronic properties. Compared to gapless graphene and bulk silicon, these abundant species are tunable in bandgap energies and thicknesses. Yet, more materials functionalities are expected than the graphene and silicon.

Inspired by advances led by graphene and aiming at the realm dominated by silicon, the emerging layered materials are expanding our knowledge and the materials

Chapter 1 Introduction

2

library. Enormous research has been done with novel ILCs, particularly the semiconducting molybdenum disulfide (MoS2). Due to its natural abundance, tunable electronic structures, and rich chemically reactive sites, MoS2 is considered as the most promising alternative to Si-based semiconductor industry and the novel candidate for catalytic applications.

Although monolayer MoS2 has been intensively studied experimentally and theoretically, few and multi-layer MoS2 is still an unexploited treasure for its tunable hostability, indirect bandgap, and the cheap price for possible industrial applications. The electrical contact between MoS2 and joined metals will be fabricated and evaluated in this thesis work. The photocatalytic property of the synthesized composites will be also explored.

Organization of the Thesis In this thesis, Chapter 2 introduces the basics of ILC materials, including the crystal structures, electronic characteristics, and the materials engineering strategies for MoS2 and other ILCs. It also briefly summarizes the theoretical and experimental routes to introduce magnetism to ILCs as discussed in Paper V. Chapter 3 introduces the light-matter interactions and the spectroscopic techniques employed in the thesis work. In Chapter 4, the sample synthesis and the engineering of noble metal nanoparticles are introduced. Chapter 5 demonstrates the formation and evaluation of metal-semiconductor contacts. The conductive atomic force microscopy technique is important for I-V measurements of nanoscale samples and used in Papers II and III. Chapter 6 focuses on the principles and applications of MoS2-based photocatalysts that are elaborated in Papers II and IV. Chapter 7 summarizes the results of the included papers (I ~ IV). The last chapter gives a brief conclusion and outlook for future work.

3

Chapter 2 Fundamentals of inorganic layered crystals

2.1 Chemical composition and crystal structures

Since the discovery of freestanding monocrystalline graphene in 2004, a paradoxical concept of an atomic crystal has been coined [2]. However, the concept is not only limited to the organic territory. The ILC materials are emerging as a huge group of non-carbon counterparts (Fig. 2.1). The ILC family contains several groups of layered materials according to their elemental constitution.

Figure 2.1 Library of inorganic layered crystals. The elements with green shadings are transition metals with confirmed ILC compounds, while the violet, red and orange shadings are nonmetals forming transition metal chalcogenides, halides, and oxides, respectively. Elements with blue shadings are 2D Xenes (X = Si, Ge, Sn, P, As, Sb). The data in this figure are summarized from references [8,9,10,11].

The ILCs can be metallic, semiconducting, or dielectric [12]. Thus, they are expected as alternatives to graphene or conventional silicon materials. Among them,

H He

Li Be B C N O F Ne

Na Mg Al Si P S Cl Ar

K Ca Sc Ti V Cr Mn Fe Co Ni Cu Zn Ga Ge As Se Br Kr

Rb Sr Y Zr Nb Mo Tc Ru Rh Pd Ag Cd In Sn Sb Te I Xe

Cs Ba La Hf Ta W Re Os Ir Pt Au Hg Tl Pb Bi Po At Rn

Fr Ra Ac Rf Db Sg Bh Hs Mt Ds Rg Cn Nh Fl Mc Lv Ts Og

Transition metal chalcogenides MoS2, MoSe2, MoTe2, WS2, WSe2, WTe2, ZrS2, ZrSe2, ReS2, NbS2, NbSe2, TaS2, TiS2, NiSe2, ZnSe, CdS, CdTe

Transition metal halidesCrCl3, CrBr3, CrI3, TiCl3, TiBr3, VCl3, VBr3, FeCl3, FeBr3, MoCl3, TcCl3, RuCl3, RhCl3, RhBr3, RhI3, IrCl3, IrBr3, IrI3,

TiCl2, TiBr2, TiI2, VCl2, VBr2, VI2, MnCl2, MnBr2, MnI2, FeCl2, FeBr2, FeI2, CoCl2, CoBr2, CoI2, NiCl2, NiBr2, NiI2, ZrCl2, ZrI2

Transition metal oxides MoO3, WO3, TiO2, MoO2, V2O5, TaO3, RuO2

Two dimensional Xenes Silicene, Germanene, Phosphorene, Stanene, Arsenene, Antimonene

Ternary compounds Cr2Ge2Te6, Cr2Si2Te6, Ni(OH)2

Others h-BN, …


4

transition metal dichalcogenides (TMDs) have received considerable attention. As a representative member, the MoS2 owns moderate and tunable semiconducting bandgaps and great natural abundance [12,13]. When a transition metal (TM) atom is connected to halogen anions (Cl, Br, I), another group of binary ILC materials is formed, known as transition metal dihalides or transition metal trihalides [8]. The aforementioned metal halides as well as ternary Cr2Ge2Te6 and Cr2Si2Te6 have recently gained considerable interest due to their possible intrinsic ferromagnetism at room temperature [14,15]. A number of transition metal oxides are also applicable to achieve few or single layer thickness [16]. Furthermore, single-layer sheets of group IV and V elements have posed potential electronic applications considering their 3D parent materials [17,18,19]. Another distinct member is hexagonal boron nitride (h-BN). It has been also given the nickname of white graphene due to its structural and electronic similarity to graphene/graphite in both 2D and 3D forms [20].

Transition metal dichalcogenides are generally marked as MX2, where M is a transition metal and X a chalcogen (S, Se, or Te). Take the MoS2 as an example. It is comprised of stacked S-Mo-S sandwiches through van der Waals interactions. Adjacent layers have a spacing of 6.5Å, typical for TMD species. Bulk MoS2 can be easily exfoliated down to few-layer or monolayer form by using mechanical exfoliation [2,21] or chemical intercalation [22,23] thanks to the weak interlayer van der Waals forces. Each MoS2 single layer consists of three atomic layers: an inner layer of Mo between two layers of S. The neighboring Mo and S atoms are covalently bonded. Such S-Mo-S layer is tiled with hexagonal atomic units from a top view (Fig. 2.2a,b) [24]. Altering the arrangements of the three layers lead to distinct MoS2 phases. Among them, the 2H-phase with trigonal prismatic symmetry is naturally available (molybdenite) and thermodynamically stable [25]. It corresponds to A-B-A stacking, where S atoms in the top and bottom layers are arranged in the same position (A) and situated on top of each other. Such coordination is kept for each S-Mo-S stack.

The 2H-phase can transform to the 1T-phase after sliding upper S-atoms with a distance of 1.82Å to occupy the hollow centers of the original 2H hexagons. The 1T-phase is arranged by A-B-C stacking (Fig. 2.2b,c). Such a change causes a semiconductor-to-metal transition in the system. Similar principles are applicable to most group VI TMDs except WTe2 [25]. This transition process is observed when bulk 2H-MoS2 precursor is chemically exfoliated into atomic thickness [26]. A recent study has recorded such phenomena in situ by using an aberration-corrected scanning transmission electron microscope (AC-STEM) [24]. In contrast to the 2H-phase, the 1T-phase is metastable and apt to transform back at either room temperature or at certain annealing temperature [27,28].

2.1 Chemical composition and crystal structures

5

Figure 2.2 Atomic structures of monolayer MoS2. (a) 2H-phase; (b) 1T-phase; (c) 2H to 1T phase transition due to the gliding of S plane. Figure adapted from ref. [24] with permission.

The above discussion mainly corresponds to a single S-Mo-S layer. However, when considering the vertical stacking of those layers, more structures can be created (Fig. 2.3). Here we use polytype to denote the few-layer and multi-layer TMDs with different stacking geometries. For example, there is only one layer in the repeating unit of the 1T type with tetragonal symmetry and octahedral coordination. In the case of the 2H polytype, a two-layer unit is repeated with hexagonal symmetry and trigonal prismatic coordination. The 3R type has the same polymorph as the 2H counterpart, but with distinct interlayer stacking orientations and a triple-layer repeat unit. A large number of such polymorphs and polytypes of TMDs have been verified or predicted [29,30].

2H 1T

S plane gliding (2H to 1T)

a b

c

A B A B CS

Mo

S’

3.16Å


6

Figure 2.3 Schematic illustrations of spatial stacking of 1T (a), 2H (b) and 3R (c) polytypes by X-M-X (X: chalcogen atom) sandwiches. Figure adapted from ref. [31] with permission.

2.2 Electronic structures of semiconductors

In addition to structural diversities, the ILCs also have unique electronic properties. Based on band theory, electronic states in solids can split into a series of quasi-continuous energy states forming an energy band. Gaps and overlaps exist among different energy bands. Semiconductors are characterized by a moderate energy gap separating the empty electronic states (conduction band, EC) from the occupied electronic states (valence band, EV). Such an energy gap (Eg) is also called bandgap. It leads to an electrical conductive ability that lies between conductors and insulators (Fig. 2.4).

Figure 2.4 Band diagrams of metals, semiconductors, and insulators. The dashed line represents the Fermi level (EF). EF is the chemical potential of electrons in a solid. It is the top of the collection of electron energy levels at absolute zero temperature while the 50% occupancy level at other temperatures.

Ene

rgy

Conduction band

Valence band

Metal Semiconductor Insulator

EF Eg

2.2 Electronic structures of semiconductors

7

2.2.1 Direct and indirect band structures

Silicon is by far the most widely used semiconductor with an energy gap of 1.12 eV at room temperature. It has an indirect band structure since its valence band maximum (VBM) and conduction band minimum (CBM) are located at different symmetry points in the Brillouin zone. On the other hand, for a direct bandgap semiconductor (e.g. GaAs), the VBM and CBM stay at the same symmetry point of the Brillouin zone. Due to these differences, the two band structures differ in electron excitation-recombination properties (see Fig. 2.5). In the case of a direct bandgap semiconductor, a photon with similar energy to the bandgap can be absorbed by the system directly, while exciting a valence band electron to the conduction band. The reverse process is also apt to happen. A photon is emitted when an electron in the conduction band falls back and recombines with a hole in the valence band. On the other hand, for electron excitation of indirect bandgap materials, simultaneous absorption of a photon and a phonon (i.e. field quantum of lattice vibration) is required for excitation. It lowers the efficiency of light absorption, and consequently the recombination process is deferred greatly in comparison to the case of direct gap materials.

Figure 2.5 Direct and indirect band structures and the related excitation-recombination processes.

Due to the aforementioned characteristics of band structures, semiconductors with a direct bandgap are featured in applications such as light emitting diodes (LEDs) and laser diodes. Applications of indirect bandgap materials are limited by the temperature dependency in the excitation process. Silicon, however, is an exception and establishes the current realm of field-effect transistors (FET) thanks to its moderate

Direct Indirect

k

E

Valence band Valence band

Directphotoexcitation

+ Photon

- Photon

Direct recombination

Electrons

Holes

Conduction band

Conduction band

+ Photon

+/- Phonon

- PhononEg

Eg- Photon


8

bandgap, native oxide layer, and abundance on earth. Also, a large recombination lifetime between electrons and holes makes Si the most popular materials for photovoltaic devices despite its relatively poor light absorption ability.

2.2.2 Electronic band structures

The direct and indirect band structures may be the intrinsic electronic features of the same materials species but with different dimensions. This happens exactly in inorganic layered crystals. As an example, bulk MoS2 has an indirect bandgap of 1.3 eV. The VBM is located at Γ point (the center of the Brillouin zone) while the CBM at a low-symmetry point (Fig. 2.6a). In the case of monolayer MoS2, on the other hand, both the VBM and CBM are positioned at the same K-point (the middle of an edge joining two rectangular faces), leading to the direct bandgap of ~1.9 eV (Fig. 2.6b).

Figure 2.6 Layer-dependent band structures of MoS2. (a) Bulk. (b) Monolayer. (c) Brillouin zone of monolayer 2H-MoS2. Brillouin zone is the Fourier transform of the minimum unit cell. (d) Layer number dependent bandgap. The electronic band structures are obtained from density functional theory (DFT) calculations. The red and blue circles indicate the VBM and CBM, respectively. Panels (a) and (b) are from [32] with permission and (d) is summarized from [5].

These band variations originate from the quantum confinement effect. Both the states of the valence and conduction bands at K-point are contributed mainly by localized Mo d orbitals, which are insensitive to interlayer coupling. Differently, states at Γ point contain considerable contribution from p orbitals of S atoms that are sensitive to the coupling. A detailed investigation shows that the bandgaps of thin MoS2 depend on the layer numbers (Fig. 2.6d). The bandgap is stretched when more layers are peeled off. It thus offers possibility to meet specific applications. For example, a monolayer of MoS2 exhibits electron mobility and optoelectronic performance that is comparable to graphene, while the multilayer form with indirect bandgap renders longer charge-hole recombination lifetime. Similar band gap trend also exists when altering chemical species in monolayer MX2 (M=Mo, W; X=S, Se, Te) materials. Bandgaps of both MoX2 and WX2 series decrease when changing the chalcogen atoms

3

2

1

0

-1

-2

-3

Energ

y (e

V)

M K

3

2

1

0

-1

-2

-3

Energ

y (e

V)

M K

a b

1 2 3 4 5 6

1.3

1.4

1.5

1.6

1.7

1.8

1.9B

and

gap

(eV

)

Layer Number

Bulk

c

d

2.3 Tunability of molybdenum disulfide

9

in the sequence of S, Se and Te. On the other hand, for MSe2 and MS2, substituting W by Mo reduces the energy gaps [33].

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

MoTe2

MoSe2

Ban

d G

ap

(eV

) MoS2

WS2

WSe2

WTe2

Figure 2.7 Bandgaps of typical monolayer TMDs computed by DFT method. The gap values are summarized from ref. [33].


Two-dimensional ILC materials can be prepared through various methods, including mechanical or chemical cleavage, chemical vapor deposition (CVD), and wet-chemical synthesis [ 34 ]. Typical ways to modify ILC materials are introducing foreign atoms/molecules, creating surface defects, or fabricating heterogeneous composites. In the following, these strategies are briefly introduced.

2.3.1 Transition metal doping

The doping technique stems from the conventional semiconductor industry. It refers to the introduction of small amount of impurities to a semiconductor matrix which changes the original crystal and electronic structures. For example, bringing group V or III dopant atoms into Si/Ge results in n-type or p-type semiconductors, respectively. Doping strategy may include the substitution of matrix atoms by dopants and also the physical adsorption and chemical bonding of foreign atoms/clusters/molecules onto or into the ILC hosts.

Doping shifts the position of the Fermi energy and may create additional energy levels for either electrons or holes into the gap between VBM and CBM. Consequently, the carrier density can be modulated. For example, introducing rhenium dopants into monolayer MoS2 results in Mo substitutions and Re adsorption. The Re dopants then locally increase the chemical affinity to impurity atoms. With the same doping method, the Au dopants are adsorbed on ILC surfaces. They induce mid-gap states and consequently give rise to p-type doping [35]. The Au nanoparticles (NPs) can also be introduced by wet-method [36]. They are selectively adsorbed on MoS2 edges or


10

surface-defective sites. A p-type electrical behavior is also observed accompanied by enhanced current on/off ratio but quenched mobility compared to the pristine MoS2 devices. It has also been observed that chemically doped Au NPs can raise the light absorption in the range of 530 ~ 670 nm while suppressing the surface plasmon resonance when the Au NPs are smaller than 10 nm [37 ]. It increases electrical conductivity, lowers thermal barriers for carrier transport [38], and activates MoS2 surface reactivity [39]. Since the Au NPs have high impacts on the chemical, optical, and electronic properties of the MoS2 matrix, a systematic study was performed in Paper I, focusing on the controls of morphological, chemical, and luminescent properties.

Besides gold, other noble metal dopants are also extensively employed as dopants or adatoms for monolayer and multilayer MoS2. The nanocrystallized Ag and Pt introduced via wet chemical synthesis improve either the photocatalytic or electrocatalytic abilities for water splitting [40,41]. Such increase of reactivity can also be achieved by sputtering deposition of Ni NPs [42]. The possibilities of doping with diverse TMs, e.g., Co, Fe, Mn, are verified by both experimental and theoretical work [43,44,45]. Those transition metal dopants are also likely to bring magnetic features to diamagnetic MoS2 matrix. In Papers II and IV, incorporation of Ni NPs as well as noble metals were experimentally realized through simple synthesis routes. This will be introduced in Section 4.1. The corresponding electronic and catalytic performances were discussed in these two papers.

2.3.2 Defect manipulations

Pristine MoS2 contains vacancies mainly at edge sites. The vacancies can induce localized defect states near the Fermi level leading to sharp transitions in both the in-plane and out-plane optical susceptibilities [46]. Inducing more S vacancies by argon plasma or hydrogen plasma can activate the inert basal surface for catalytic activity, especially the hydrogen evolution reactions [47,48]. More species of defects, e.g., S2 substituted by Mo, form the tilted grain boundaries containing 4−8 rings or S bridges. They are also predicted to be responsible for surface reactivity [49]. The defects at surface or even subsurface can be employed for epitaxial growth to join different ILC materials. Such growth process can produce hybrids without dangling bond interferences [50].

2.3.3 Heterogeneous layers

Incorporating other layered counterparts (e.g., graphene, CdS, TiO2, WS2) to MoS2 leads to various electronic and catalytic behaviors. For example, the relative orientations for a graphene-MoS2 bilayer heterogeneous junction give variable interplanar S-S distances and strongly affect the bandgap of MoS2 [51]. Combination of foreign ILC semiconducting layers typically modifies the separation and recombination of electron-hole pairs. This can be seen through the enhanced charge transport abilities in the covalently-bonded MoS2-CdS and MoS2-TiO2 heterostructures


11

[50, 52]. Being a popular strategy to facilitate the photocatalytic and electrocatalytic reactions, van der Waals heterojunctions are also designed for such a purpose [53].

However, poor contact qualities between hetero counterparts limit the applications of these heterostructures. To overcome the bottleneck, suitable buffer layers are therefore required. Introducing a thin BN layer helps to preserve the original structural and optoelectronic performances of MoS2 single layer for a MoS2-SiC heterostructure [ 54 , 55 ]. Noble metals are chemically compatible to many semiconducting materials. The Au nanoparticles provide pathways of charge transport between a BiVO4-SrTiO3 junction [56]. It may perform similar roles in a MoS2-based composite. Buffer layer insertion is more prevalent for MoS2-metal contacts by offering effective charge transfer at the interface [54]. This idea is a key point of the synthesis work in Papers II and IV where noble metals (Au, Ag) were adopted as the buffers.

In contrast to conventional heterostructures, the inherent 1T/2H junction is heterogeneous in atomic and electronic structures but homogeneous in chemical species [57,58]. There is no lattice mismatch in such junction, making good quality electrical contact at the interface. Such junctions are achievable through phase engineering or sequent CVD, but it still requires more efforts for simple and reliable experimental realizations.

2.3.4 From nonmagnetic to magnetic

Pristine MoS2 and most other ILCs are intrinsically nonmagnetic. Besides the structural and electronic adjustments, many of the aforementioned tailoring methods are also possible to introduce magnetism to MoS2. Incorporating the semiconducting features of ILCs may create dilute magnetic semiconductors (DMSs), which are considered as key components of the quantum computers [59]. Among them, doping transition metals to MoS2 has been a preferred route. The concept stems from the ferro- or ferrimagnetism residing on the TM impurities. For example, 3d and 4d transition metal dopants are predicted to induce tunable magnetic moments to MoS2 [44, 60 ]. Constructing zigzag edges is also possible to form magnetic MoS2 nanoribbons [61]. Creating S vacancies accompanied by phase or strain manipulations can also induce magnetic moments without deteriorating the stability of MoS2 [ 62 , 63 ]. Bringing magnetism into ILCs is a systematic work, yet, out of the main theme of the thesis. However, the latest progresses offered by first-principles predictions are presented in Paper V.

Despite the progress in theoretical predictions, experimentally introducing magnetism to ILCs are still demanding tasks. Doping via particle bombardments suffers from severe heat loads into the target, and the incoming dopants are apt to agglomerate on MoS2 slabs before reaching the inside layers [64]. In the case of wet chemical methods, reactions are more likely to happen on the MoS2 edges rather than the surfaces since defect sites are mostly at edge sites. In Paper II, magnetism has been realized in MoS2-Au-Ni composite via wet synthesis, but the introduced magnetic moments of Mo,


12

S and Au are localized in the interface region. In Paper V, synthetic routes were collected and reviewed for possible realizations of the ILC-based DMSs.

13

Chapter 3 Spectroscopy techniques

This chapter first introduces the physical background of light-matter interaction. In the following, several spectroscopy techniques involved in this thesis are introduced: synchrotron radiation (SR) based X-ray photoemission electron spectroscopy (XPEEM), X-ray photoelectron spectroscopy (XPS), and UV-visible spectroscopy.

3.1 Light-matter interaction

3.1.1 Photon absorption and system relaxation

Light is composed of discrete particles known as photons. According to the wave-particle duality, a photon can be described by its energy ( E ) and momentum ( p )

λν hchE == (3.1)

p

h=λ , (3.2)

where h is the Planck constant, c is the speed of light in vacuum, ν is the frequency, and λ is the wavelength.

Once a photon interacts with matter, either absorption or scattering process can happen. The absorption leads to annihilation of the photon. If the energy of the photon matches the energy difference between two states of the system, the photon can be absorbed, and the system is excited to an unoccupied state of higher energy (Fig. 3.1a, path A1). For a system A

*AhνA →+ , (3.3)

where A and *A represent the initial and excited states of the system. When a photon interacts with a bounded electron, the interaction primarily follows the electric dipole approximation and the process can be described by considering only the first term of expansion of the photon-matter interaction operator [65].

A photoionization process can happen if the photon energy is larger than the binding energy of an electron in the system


14

−+ +→+ eAhνA , (3.4)

where +A represents the ionized state. The electron is excited to the vacuum state and identified by its kinetic energy

bk EhE −= ν , (3.5)

where bE and kE are the binding energy and kinetic energy of the ionized electron (photoelectron).

Figure 3.1 Photon absorption and the system relaxation processes. (a) Photon absorption. A1 and A2 show two possible results: photoexcitation and photoionization. (b) Relaxation. B1 and B2 display two alternative relaxation routes: radioactive decay and normal Auger decay. Resonant Auger processes are not depicted in the figure.

Photoionization is the result of the photoelectric effect, and its probability depends on the photoionization cross section. The cross section varies for electrons of different elements, different electronic orbitals, and as a function of incident photon energy. It describes the number of photoelectrons per unit area and unit time produced by absorption of photons

( ) SN

N

StN

tN ⋅=⋅

=ph

e

ph

epe /

/σ , (3.6)

where peσ is photoionization cross section, eN and phN are the numbers of

photoelectrons and incident photons, t is time, and S is the photon flux area. The photoemission transition can be described using Fermi’s Golden Rule in the dipole approximation. The transition probability per unit time is

fif ifP ρπ 22 ><= V

, (3.7)

Core level

Occupied valence levels

Unoccupied levels

Ionization threshold

Photoelectron

hv

A1

A2

a b

hv

B1

B2

Auger electron

B2

3.1 Light-matter interaction

15

where ifP is the transition probability, >< if V is the matrix element between the

final and initial states, V is the dipole operator, and fρ is the density of final states. The photoionization cross section can be then expressed as

fph

ifx if

c

e

mP

Pρ

ωπσ 2

2

2

22 14 ><== V

. (3.8)

If a system in *A or +A states is unstable, subsequent relaxation processes occur (Fig. 3.1b). Regardless of vibrational and collisional relaxation, either a radiative or non-radiative transition can happen. The former involves a decay process of an electron dropping down from an excited orbital accompanied by a photon emission. Such radiative decay after photon absorption is sometimes called photoluminescence. It is classified into two types: the fluorescence where the total spin quantum number is unchanged and the phosphorescence where there is a change in the spin, resulting in a much longer lifetime of the excited state (10-3 ~ 102 s) [66].

The non-radiative case is usually called Auger decay [65]. Normal Auger processes can be schematically understood so that an electron at a higher energy orbital fills the core-hole and simultaneously another electron from an outer orbital is ejected into the continuum, known as an Auger electron.

3.1.2 Scattering

Light-matter scattering is classified into elastic (coherent) and inelastic (incoherent) scattering. Rayleigh scattering is the predominant type of elastic scattering where photons are scattered by particles that are much smaller than the photon wavelength. In such scattering, incident photons do not lose energy and atoms/molecules are not excited or ionized. The latter case is known as Compton scattering, where the photon is deflected by the electron and transfers part of its energy to the electron.

The atomic scattering factor (also named as atomic form factor) is used as a parameter of scattering amplitude of light by individual atoms. It is applicable when photon energies range from ultraviolet to hard X-rays. It is expressed as

210 ifff += , (3.9)

where 1f and 2f are components related to scattering and absorption. The scattering factor can be used to describe the refractive index with regards to the influence of scattering and absorption

( )21

20

211 iff

rnin a −⋅−=+−=

πλβδ , (3.10)

where δ and β are refractive index decrement and the absorption index respectively,

an is number of atoms, and 0r is the classical electron radius.


16

The radiation emitted during Rayleigh scattering has negligible interference by itself. However, when scattered by a periodic lattice, the interference of scattered light waves becomes important if the wavelength of incident light is comparable to the lattice distances. The phenomenon can be understood using Bragg’s description. If the reflection angle of the scattered light is the same as the incidence, the scattered light follows the superposition principle. A constructive interference would happen if the light path difference equals to integer number of the wavelength. The Bragg’s Law thus reads

λθ nd =sin2 , (3.11)

where d , θ , and n are lattice spacing, incident angle (between incident light and lattice surface), and an integer. The Bragg condition is the basis of X-ray diffraction (XRD) applications.

3.1.3 Beer-Lambert Law

Due to the attenuation by absorption and scattering, light transmitting through a substance will decrease in intensity. Such process follows the Beer-Lambert Law [67]

Xx eIeII ρ

μμ

−− ⋅=⋅= 00 , (3.12)

where 0I and I are the intensity of incident beam before and after passing through the materials, x is the length of light path, μ and ρ are the linear attenuation coefficient (cm-1) and density of the substance (g/cm-3), respectively, ρμ / is the mass attenuation coefficient (cm-2/g), and ρ⋅= xX is the mass thickness. This equation reveals an exponential drop of the transmitted intensity along the thickness of materials. The mass attenuation coefficient depends on the total cross section per atom ( totσ ) [68]

( )Rtot // Au ⋅= σρμ (3.13)

trippairincohcohpetot σσσσσσ ++++= , (3.14)

where u and RA are atomic mass unit and relative atomic mass, peσ , cohσ , incohσ ,

pairσ and tripσ are cross section components of photoionization, Rayleigh scattering, Compton scattering, and electron-positron production of nucleus and of atomic electrons, respectively.

3.2 X-ray photoemission electron microscopy

3.2.1 Synchrotron radiation

When the velocity of charged particles changes, electromagnetic waves are emitted. Based on this principle, synchrotron radiation is emitted by charged particles that are

3.2 X-ray photoemission electron microscopy

17

at relativistic velocities in curved orbits. In a modern SR facility, electrons produced by an electron gun are first accelerated through a linear accelerator and possibly a booster ring. After that, the electrons are injected into a polygonal storage ring. Bending magnets and insertion devices can be positioned at the bending and straight sections, respectively. Beamlines are tangentially guided out of the storage ring after insertion devices. Equipped with apertures, slits, mirrors, and monochromators, the beamlines focus and monochromatize the photon beam, tune the photon energy or the intensity of photon flux, etc.

Synchrotron radiation is advantageous in brilliance, tunability, collimation, and well-defined polarization in comparison to laboratory based X-ray sources such as X-ray tubes [69]. Such a high brilliance leads to much better precision in, for example, diffraction measurements. It can also unveil information from larger depth of the samples. The tunable energy also facilitates photon energy-dependent experiments.

3.2.2 X-ray photoemission electron microscopy

Each electron shell has a characteristic binding energy. Incident X-ray photons matching those energies (or above) will be dramatically absorbed, with core-level electrons ionized. X-ray absorption spectroscopy (XAS) records the photon intensity variation as a function of energy. Alternatively, the technique of X-ray photoemission electron microscopy (XPEEM) detects the emitted electrons. XPEEM technique aims to incorporate the merits of spectroscopy and microscopy (Fig. 3.2). By tuning the incident photon energy and focusing emitted electrons, it records the near edge X-ray absorption fine structure (NEXAFS) and the microscopic view simultaneously.

During XPEEM measurements, incident X-ray photons are tuned within an energy range. Strong electric field is applied between the objective lens and the sample. Thus, the electrons emitted by X-ray irradiation are accelerated. The electrons go through a series of lenses and finally reach a fluorescent screen with the image magnified. Due to the high voltage applied, the sample surface should be flat enough to avoid discharging. In practice, samples in powder form are typically deposited onto single crystalline Ge or Si substrates, where Ge is more favorable due to better electrical conductivity. Sample degassing and high vacuum are also required to carry out the measurement.


18

Figure 3.2 Schematic sketch of an XPEEM setup. The magenta lines show the electron paths.

XPEEM is a surface-sensitive technique. The incident SR beam and especially the ionized electrons are attenuated by the sample. Their intensity drops rapidly along the paths according to Beer-Lambert Law and the universal curve. By using a hard X-ray source, the detection depth can reach 50 nm, 10 times more than the case of soft X-ray application [70].

The experimental work in Papers II and III was performed at the PEEM endstation of Beamline I311 at MAX IV laboratory (now the MAXPEEM beamline in the new facility). The beamline combines spectroscopic photoelectron and low energy electron microscope in a single instrument. The photon energy can be tuned from 30 eV to 1200 eV.

XPEEM enables micro-regional spectroscopic analysis. In Fig. 3.3, by selecting micro- or nano-regions from different Ni nanoparticles, the obtained spectral results provide core-shell information of the particles. For example, an additional shoulder is found in the upper spectrum, which is attributed to surface oxidation. Further analysis is also possible from the morphological aspect between the two micro-regions.

Sample

Objective lens

Contrast aperture

Projector lens

Magnified image

Intermediate image

Detector

X-rays

3.3 X-ray photoelectron spectroscopy

19

845 850 855 860 865 870 875

Ni 2p1/2

Inte

nsity

(a

.u.)

Photon Energy (eV)

Ni 2p3/2

Figure 3.3 XPEEM Ni 2p spectra of (MoS2)86.7Ni8Au5.3 sample. The spectra are obtained from two different micro-regions as shown in the insets.

3.3 X-ray photoelectron spectroscopy

XPS technique employs monochromatized X-rays to get electrons out of the studied sample. The detector counts the emitted photoelectrons (or also Auger electrons). XPS measurements record the relationships between the electron counts per unit time and their kinetic energies. Basically it can, for example, identify elements and their chemical environments of a studied sample surface.

For solid materials, the binding energy of an electron is defined as the energy required to excite it from its original orbital to the Fermi level. In order to be captured by an electron analyzer, some additional energy is necessary to promote the electron further from the Fermi level to the vacuum level. The energy is equal to the work function of the material. Therefore, Eq. 3.5 is modified as follows:

ΦEhE −−= bk ν , (3.15)

where Φ is the work function of the sample to be measured. During XPS measurements, solid samples are electrically contacted to the sample holder. The Fermi level alignment after thermodynamic equilibrium ensures the following relation (Fig. 3.4):

spec'kk ΦEΦE +=+ , (3.16)


20

where specΦ is the work function of the XPS device and 'kE the measured kinetic

energy in practice. In most cases, the work function of the sample is unknown but specΦ is given. Thus, this serves an alternative way of measuring binding energies

spec'kb ΦEhE −−= ν . (3.17)

Figure 3.4 Energy band alignment in XPS measurements. The band bending is neglected in this figure for a generalized contact.

XPS is also highly surface sensitive. The outgoing electrons are decelerated by the sample where they come from. An effective detection depth of XPS device is typically defined as ( )Eλ3 , where ( )Eλ is inelastic mean free path of electrons. Such a depth is usually 1 ~ 10 nm for metals and inorganic compounds.

The elemental species can be identified by the binding energies of their core level electrons. Also, the binding energies are influenced by their chemical environment, e.g., the number, distance, and species of the neighboring atoms [71]. All these factors will alter the peak positions in the XPS spectra known as the chemical shift. Therefore, an elemental component with complex valences and chemical environment will present several peaks with certain chemical shifts. The measured spectrum is thus an incoherent sum of those peak components. In practice, different binding energies can be figured out for a certain element, ions with higher valences, or the same valence but bonded to neighbors with higher electronegativity.

An XPS spectrum peak is featured with its full width at half maximum (FWHM) besides the energy position and intensity. It represents the broadening of a peak, which is affected by both the intrinsic elemental species and instrumental resolutions. The

K

L

M

EF

E0

hv Eb

Ek

spec

Ek’

Detector

Sample XPS sample holder

Valence states

3.4 UV-visible spectroscopy

21

peak shape can be usually described by the convolution of Gaussian and Lorentzian distributions, known as the Voigt profile [72]. A fitting operation based on the Voigt profile can separate and identify the peak components that present specific valences and chemical environments. In the same spectrum, the relative quantities of each chemical species can be calculated from the ratio of the peak contents (the product of peak area and sensitivity factor).

3.4 UV-visible spectroscopy

UV-visible spectroscopy measures the absorption, reflectance and transmittance in the wavelength range from near-ultraviolet to near-infrared. For a liquid sample, the transmittance (T) is measured with a setup shown in Fig. 3.5. It is calculated based on Beer-Lambert Law

ττ

−−

=⋅== eI

eI

I

IT

0

0

0

, (3.18)

where 0I and I are the light intensity before and after passing through the liquid sample and τ is the optical depth. The absorbance (A) can be derived from either the transmittance or optical depth

10ln

1log10

τ=

=T

A . (3.19)

Similar principles are also applicable to solid film samples. The absorption coefficient (α) is used to evaluate the absorption of light at certain wavelength. It can also be calculated by Eq. 3.20 or by Eq. 3.21 specifically for liquid samples:

( )t

A

T

R

t

10ln1ln

1 2 ⋅=

−=α (3.20)

ct

A

⋅=α , (3.21)

where R and c are the reflectance and molar concentration, respectively and t is sample thickness in Eq. 3.20 or length of cuvettes in Eq. 3.21.


22

Figure 3.5 Absorption spectroscopy setups for liquid samples. In setup (a), light intensity before and after passing through the liquid sample is compared. If there are two channels (b), the intensity of light transmitting the reference cell (empty or filled with solvent) and the one passing through the sample cell are recorded.

Once we get the absorption coefficient, the optical bandgap can be extrapolated by a Tauc plot [73,74]. It has the relation

)()( g/1 Ehh m −∝ ννα , (3.22)

where gE and m are the optical bandgap and an exponent related to the optical

transition types (m = 2 for indirect semiconductors while m = 1/2 for direct ones).

SpectrometerLamp Reference cell

Sample cell

I0

I

Sample cell

I0I

t

SpectrometerLampa

b

23

Chapter 4 Sample preparations

4.1 Synthetic methods

The synthetic work of transition metal decorated MoS2 in Papers I ~ IV was performed with wet-chemical methods. Besides various synthetic methods introduced in Paper V, the thesis work is also progressed with special emphasis on the assistance by ultrasound.

An inert ambience is required in the case of wet chemical synthesis. HAuCl4 is a common precursor to obtain Au metal or compounds. It has strong oxidative ability so that it can easily react with reductants. In the HAuCl4/MoS2 aqueous system (Paper I), the synthetic reactions were performed under Ar or N2 ambience to eliminate the oxidization by O2 in air. A high-speed magnetic stirring was employed to disperse the reactants evenly and promote the reactions. The synthesis can proceed at room temperature while increased temperature would lead to higher yield of Au NPs with smaller size.

Situations become more intricate when both Ni NPs and Au or Ag reagents are put together with MoS2 aqueous suspension. Compared to Paper I, Paper II involves Ni NPs introduced to the HAuCl4/MoS2 system. Chemical reactions happen among the 3 reagents. Inert ambience is, however, still necessary to overcome the competitive oxidation by O2. Both stirring and moderate temperature are required to increase the possibilities of ternary collision among Ni, Au3+ and MoS2. In this case, a sonochemical environment was adopted to evenly mix the aqueous suspension and provide energy for the synthesis [75,76]. The synthesis was performed in an ultrasonic cleaner at 50 °C for 30 min. The sizes of obtained Au NPs are approximately in accordance with the ratio/size relationship reported in Paper I.

In contrast, AgNO3 solution was employed as the oxidative reagent instead of HAuCl4 for synthetic reactions in Paper IV. The Ag+ ion is less electronegative than the Au3+ ion. Thus, it is expected to facilitate a mild and long-term reaction to activate the MoS2 inert surface. In this case, the synthesis was performed under air ambience. Ultrasound source was also provided with prolonged reaction time.

By changing the proportions of HAuCl4 and MoS2, Au3+ are reduced and crystallized into different sizes. Such proportion effects can also make different photocatalytic abilities. This will be discussed in the following section.

Chapter 4 Sample preparations

24

4.2 Noble metal nanoparticles

Noble metals are stable species in composites. The size of noble metal NPs significantly affects the catalytic properties of the hosting ILCs [77]. The noble metal species and morphologies are also important when the NPs bridge ILCs and other metallic components. As a buffer layer, noble metal agglomerations (Au in Paper II and III, Ag in Paper IV) should have suitable sizes and distributions to form a channel for effective charge migration.

4.2.1 Size determination

The nanoparticle size can be analyzed through microscopy techniques, either scanning electron microscopy (SEM) or transmission electron microscopy (TEM). SEM technique is suitable for sample pretreatments, e.g., mechanical exfoliations to remove extra materials and get thinner layers. TEM is necessary in order to reach higher magnifications. Figure 4.1 shows Au NPs in different sizes decorated on MoS2 flakes.

Figure 4.1 TEM images of Au NPs on MoS2 multilayers with different molar ratios. (a) MoS2: HAuCl4 = 1:1. (b) MoS2: HAuCl4 = 15:1. Figures from Paper I.

Each Au NP may comprise a number of nanocrystals. It can be seen by the contrast information in Fig. 4.1b. However, quantitative analysis is possible with XRD determination. The mean crystal size along certain crystal orientation can be estimated by the Scherrer equation [78]

θβλ

cos

KD = , (4.1)

where K is the crystallite-shape factor (0.9 is normally used), λ is the wavelength of X-rays, β is the XRD peak width (FWHM), and θ is the Bragg angle. The Scherrer equation is applicable when crystal sizes are below 100 nm.

4.2.2 Size engineering

Au nanoparticle sizes can be tuned via altering reactant proportions [79] and reaction conditions. Figure 4.2a further proves such experimental routes. More HAuCl4

4.2 Noble metal nanoparticles

25

involved in the synthesis significantly raises the Au particle sizes. This figure also presents the discrepancies between crystal sizes and nanoparticle sizes. With high Au3+ content, each Au nanoparticle may comprise 5 ~ 20 single crystals. It is obvious that the mean crystal size is nearly unchanged. In Paper II, the mean crystal size of Au NPs in ternary (MoS2)92.9Au2.9Ni4.2 composite (molar ratio MoS2: Au ≈ 40:1) is around 20 nm. It suggests that the mean crystal size is independent on proportions while the particle size is tunable. In fact, the differences between nanoparticle size and mean crystal size bears the trace of nanoparticle formation kinetics. Au atoms are reduced from Au3+ and then form primary crystalline units (< 20 nm). Neighboring crystalline units agglomerate to form large spherical nanoparticles up to submicron scale.

0 5 10 15 20

10

100

Nanoparticle size (TEM) Mean crystal size (Scherrer equation)

Au

Dia

mte

r (n

m)

MoS2:HAuCl4 Molar Ratio

a

0 5 10 15 20

0

500

1000

1500

2000

PL

inte

nsity

(C

ount

s)

MoS2:HAuCl4 Molar Ratio

b

Figure 4.2 Proportion effect on Au sizes and optical properties. (a) Mean crystal sizes and nanoparticle sizes of Au agglomerations in MoS2-Au composites. (b) Photoluminescence intensity as the function of MoS2: HAuCl4 molar ratios. Figures from Paper I.

The relative contents of MoS2 and HAuCl4 also affect the photoluminescence (PL) property of the Au decorated MoS2 (Fig. 4.2b). The plasmonic effect of Au nanoparticles enhances the PL efficiency of MoS2. The PL intensity reaches the maximum when the two reactants have equal contents. Too high HAuCl4 content leads to larger Au agglomeration and MoS2 deterioration that inhibit the PL performance.

In the cases of ternary composites, size tunability of noble metals is also possible when the NPs act as buffer layers. For example, Au NP sizes can vary from 5 nm to more than 50 nm (Paper II). However, particular study is not included in Paper II or IV. Such study may be helpful to optimize the photocatalytic property in future work.

27

Chapter 5 Metal-semiconductor contacts

In practical electronic applications, semiconducting ILC components must be incorporated into a circuit via contacts with metallic electrodes. The contact between a metal electrode and ILC differs greatly from a contact between two metals, especially in the nanoscale. Typically, a metal-semiconductor (M/S) contact exhibits a non-linear current-voltage (I-V) relationship and a high electrical resistance, which is ascribed to a discontinuity on the electronic band structure of the interface [ 80 ]. Such a characteristic undermines the performance of semiconductor devices. It is thus the top priority to reduce the contact resistance and construct reliable metallic interface between a semiconductor and adjacent metal electrodes [81,82,83].

5.1 Band alignment at metal-semiconductor interface

5.1.1 Work function

At absolute zero, electrons of a metal fill the states under the Fermi level ( FE ), leaving all the states above the Fermi level empty. While at a certain temperature, electrons near

FE are promoted over FE by thermal excitation. The work function (Φ ) is the minimum energy required to completely extract an electron from a solid surface into vacuum. It can be quantified as the energy difference of an electron between the Fermi energy and the vacuum state energy ( 0E ) nearest to the surface:

F0F EEEeΦ −=−= φ , (5.1)

where e is the electron charge and φ is the electrostatic potential of the electron in the vacuum but near the solid surface.

The work function is a property specific to each solid surface. The surface of certain materials has nearly constant work function when the surface is under the same ambient conditions, e.g., surface cleanness, annealing treatment [84], etc. Figure 5.1 shows work functions of most solid metals. It demonstrates that metal work functions range mainly from 2 to 6 eV and vary periodically along with atomic numbers [85]. Though possessing relatively high work functions, Ni, Ag, and Au are still frequently used in electronics due to their high stability and excellent electrical conductivity. They are employed to modify MoS2 crystals in Papers I ~ IV.


28

0 10 20 30 40 50 60 70 80 90

2

3

4

5

6

Au

Ag

Wor

k F

unct

ion

(eV

)

Atomic Number

Ni

Figure 5.1 The dependence of work functions on atomic numbers. The dashed lines separate elements of different periods. Ni, Ag, and Au are marked in particular. Work function data in this figure are organized from ref. [86], where all values are measured from polycrystalline specimens. Note that the single crystalline planes may have slightly different work functions from the values shown in the figure [87].

Similarly, an energy difference between 0E and FE of a semiconductor is required to excite an electron from VBM to a vacuum state (Fig. 5.2). Since the Fermi level of a semiconductor depends on the doping levels, the work function is also correspondingly variable.

Figure 5.2 Band diagram of an n-doped semiconductor. χ is the electron affinity and

nE represents the energy difference between CBM and FE .

Valence band

Conduction band

Energy

EF

Eg

E0

CBM

χ

Φ

En

VBM

5.1 Band alignment at metal-semiconductor interface

29

5.1.2 Contact formation and the Schottky barrier

When two species of solid materials are joined, electrons of the one with lower work function are apt to move to the other with higher work function. Such phenomena exist in both metal-metal and M/S contact interfaces. Figure 5.3a shows a metal with higher work function ( mΦ ) than a semiconductor ( sΦ ). Before contact, the metal and semiconductor have equivalent vacuum level but different Fermi energy. Once they are electrically connected, electrons from the semiconductor tend to flow towards the metal part. Such a process leads to negatively charged metal surface and positively charged semiconductor surface. As a result, the Fermi levels of the metal and semiconductor will reach thermal equilibrium leaving the metal with reduced potential energy and a converse situation for the semiconductor [80] (Fig. 5.3b). The difference of contact potential ( φΔ ) is

q

ΦΦ mssmms

−=−==Δ φφφφ , (5.2)

where mφ , sφ , and msφ are the potentials of metal, semiconductor and the potential

difference at the M/S interface, respectively, mΦ and sΦ are the metal and semiconductor work functions, and q is the carrier charge.

Figure 5.3 Formation of a chemically bonded M/S contact (n-doped semiconductor,

mΦ > sΦ ). The symbol D refers to the distance between the metal and semiconductor.

When the metal and semiconductor are intimately contacted, more electrons are transferred to the metal side of the interface and leave more positive charges at the semiconductor side. These positive charges are confined in a space-charge region (also known as the depletion depth, dW ) [88]. Consequently, a space-charge region is formed, making the conduction and valence bands bended towards higher energy (Fig. 5.3c). In this case, the contact potential difference is defined by both the M/S interface and the depletion layer

dmsms φφφ +=−=Δ

q

ΦΦ, (5.3)

En

E0

EF,m

EC

EV

EF,s

m

s

E0

(a) Before contact (b) After contact(D >> atomic spacing)

(c) After contact(D ~ van der Waals radius)

(d) After contact(D ~ 0)

En

E0

EF

EC

EV

EF

m s

E0

D

qms

qd

Wd

E0

EF

EC

EV

EF

E0

En

m s

qB

Wd

qd

EnEF

qmsE0

EF

EC

EV

m s

E0

D


30

where dφ represents the potential difference across the depletion depth. The depletion depth without external bias can also be estimated as

2/1

Bd

D

sd

2

−=

e

Tk

eNW φε

, (5.4)

where sε , DN , dφ , Bk , and T are the static dielectric constant of the semiconductor, doping density, Boltzmann constant, the built-in potential difference of the M/S contact and the temperature, respectively. msφ is quite limited due to the small distance between metal and semiconductor parts. Therefore, the contact potential difference is predominantly contributed by dφ . If the distance is negligible, dφ equals to the total contact potential difference

dms φφ =−=Δ

q

ΦΦ. (5.5)

The barrier heights exist at both sides of the interface (Fig. 5.3d). At the semiconductor side, the barrier height is

smd ΦΦq −=φ (5.6)

and the barrier height at the metal side is calculated by the equation

χφ −== mBB ΦqΦ , (5.7)

where BΦ is known as the Schottky barrier that represents the energy required for an electron to move from the metal part to the semiconductor [80]. The above analysis primarily follows the Schottky-Mott model, which assumes that there is no interaction between the metal and semiconductor. Thus, the energy barrier height is only determined by the work function of the metal and the electron affinity of the semiconductor [89].

In the case of ILC semiconductors with limited thickness, such chemically bonded contact is usually achievable at the edge sites [37]. A metallized region can be formed in the contact interface. On the other hand, if a contact forms on the ILC’s basal plane, chemical bonding is unfavorable due to the lack of dangling bonds. In this case, the M/S contact is likely to be formed by van der Waals forces. There will be an additional tunnel barrier and flat-band region in the contact interface besides the Schottky barrier (Fig. 5.4) [90]. This typically happens when metal electrodes are lithographically deposited onto the ILC surface. Such a van der Waals gap increases both the height and width of the total energy barrier resulting in higher contact resistance.

5.2 Ohmic contact and charge injection

31

Figure 5.4 Band alignment of a physically contacted M/S interface (n-doped semiconductor, mΦ > sΦ ). BΦ and TΦ are the Schottky barrier and tunnel barrier, respectively.

By choosing metals with different work functions, three types of M/S contacts can be formed as depicted in Fig. 5.5. When mΦ < sΦ , electrons from the metal part can migrate freely to the semiconductor and accumulate in the conduction band confined within the interface region. Therefore, a negatively charged space-charge region is formed. The electrons thus encounter the minimum barrier when flowing in both directions. It is the preferred situation to form an Ohmic contact based on the simple Schottky-Mott model.

Figure 5.5 Three types of M/S contacts with n-doped semiconductors (Schottky-Mott model). (a) mΦ > sΦ ; (b) mΦ < sΦ ; (c) mΦ = sΦ . Tunnel barriers are neglected in all the three cases.

5.2 Ohmic contact and charge injection

Although the Schottky-Mott model provides a simple method to form an Ohmic contact by choosing a metal with smaller work function than the semiconductor,

EC

EV

EF

ΦB

ΦT

Metal Semiconductor

van der Waals gap

Flat band region

(a) Depletion (b) Accumulation

EF

EC

EV

ΦB

Wd

qφd

Metal

---EF

EC

EVMetal

-----

(c) Neutral

EF

EC

EVMetal

-----


32

experimental observations normally deviate from such an expectation. The measured Schottky barrier heights for a semiconductor with different metal electrodes are relatively independent on the metal work functions [ 91 ]. Experimental results demonstrate that the Schottky barrier height almost equals to two-thirds of the bandgap for an n-type semiconductor and one-third of gE for a p-type one.

Such phenomena are ascribed to the surface states of the semiconductors. Large density of states at semiconductor surface can form a space-charge region without contacting to a metal. It often pins the Fermi level of the semiconductor at certain energy states in the band gap (gap states) [81]. Once an M/S contact is formed, charge transfers across the contact will be accommodated in the gap states and thus the Fermi level will keep stationary at the semiconductor interface.

The Fermi level pinning makes it impractical to get an accumulation type of M/S contact. In other words, a depletion layer exists even when mΦ < sΦ . Manipulations of the doping levels become the primary method to achieve an Ohmic contact. Although the Schottky barrier height is relatively independent of the doping concentration, the barrier width can be reduced by doping [92]. It is beneficial to narrow the depletion width by using a heavily doped semiconductor. In this case, charge injection pathways can be modulated (Fig. 5.6).

Figure 5.6 Illustrations of charge injection mechanism for a depletion type of contact (n-doped).

For an intrinsic or lightly-doped semiconductor, electrons flow over the Schottky barrier by thermal excitation. Based on the thermionic emission theory, the current density can be estimated as:

−−⋅

−⋅⋅=

Tk

qV

Tk

qTAJ α

BB

B* exp1expφ

, (5.8)

where *A is the Richardson constant which is relevant to electrons’ effective mass, Tis temperature, α is a constant which equals to 2 and 2/3 for bulk and 2D

(a) Thermionic emissionLow doping

(b) Thermionic-field emissionMedium doping

(c) Field emissionHigh doping

EF

EC

EVMetal

---- EF

EC

EV

Metal

------ EF EC

EV

Metal

--------

5.3 Current-voltage relationships

33

semiconductors, respectively, q is the carrier charge, Bφ is the Schottky barrier potential, and V is the applied bias.

With medium doping concentrations, the barrier width is squeezed and thus both the thermionic emission and field emission contribute to the charge transfer. Heavily-doped semiconductors own narrow barriers and electrons can probably tunnel through the contact barrier directly. However, even in an absolute field-emission regime, the energy barrier can hardly be removed. For example, the electrical resistivity of 1T/2H-MoS2 contact is down to 0.2 Ω·mm (the lowest value ever reported for MoS2-based M/S contact) [58], which is still much larger than for most metals.

In order to construct a low-resistance Ohmic contact, controlling the doping concentration is a traditional way in semiconductor industry. It is usually carried out by a CVD procedure for high vacuum and temperature. The procedure creates an additional doped layer on the outer surface of the semiconductor. Straightforward fabrication by wet chemical method has been also widely studied recently (Section 4.1). Properly selected metallic counterparts and experimental parameters are favorable of obtaining metallic structures at M/S contacts. This method is employed in Papers II and IV, where ternary metallized interfaces are formed with greatly reduced contact resistance.


According to the thermionic emission theory (Eq. 5.8), current density depends on the Schottky barrier and the applied bias. A similar relation also exists in field-emission dominated occasions [88]. The current-voltage (I-V) relationship is of crucial importance to evaluate the electrical performance of semiconductor-based electronics. Electrical resistance and resistivity can be derived from I-V curves. Typically, there are two widely used techniques to perform the I-V measurements for microelectronics: field-effect transistor (FET) based and conductive atomic force microscopy (C-AFM) based measurements.

5.3.1 Field effect transistors

There are two main types of FET devices: the junction field-effect transistor (JFET) and metal-oxide-semiconductor field-effect transistor (MOSFET). The MoS2-based FET devices are commonly constructed in the latter type, comprising MoS2 monolayer or multilayers and the source, drain and gate metallic terminals. Typical monolayer MoS2 based FET devices are illustrated in Fig. 5.7 [13, 93].


34

Figure 5.7 Schematic diagrams of 2D MoS2-based FET devices. (a) Top-gated FET with a HfO2 dielectric layer and Au terminals. (b) Back-gated FET with Sc as the metallic terminals.

The terminal metals have smaller work functions than the MoS2 in both cases. In Fig. 5.7a, three terminals are fabricated by electron-beam lithography and vapor deposition. Such FET device reveals high mobility and on/off current ratio. However, the contact mode between MoS2 and Au terminals still remains elusive [93, 94 ], although edge contacts are specially designed and an annealing process has been done to lower the contact resistance. Generally speaking, it is not easy to get a chemically-bonded M/S contact via a vapor deposition process, and pure edge contacts are difficult to realize on the atomically thin slabs. According to the reported fabrication methods and microscopy images, the Au/MoS2 [13] and Sc/MoS2 [93] contacts are actually combinations of both edge and surface contacts. Apart from these issues, FET devices provide possibilities to study the I-V relationship.

With an FET device, the relationship of current (Ids) against drain-source voltage (Vds) can be obtained, and the electrical resistance can be calculated [95,96]. Meanwhile, the Schottky barrier height can be also estimated by extrapolating the En-Vgs curve to zero bias [81, 97 ]. However, deposited metal electrodes usually have widths at micrometer scale [98]. Constructing an FET structure is thus inconvenient in this thesis work due to even smaller sizes of both MoS2 and Ni than the electrode width. For example, Ni particles in the synthesized Ni-Ag/Au-MoS2 composites (Papers II ~ IV) have an average diameter of 200 nm and noble metal particles smaller than 50 nm. Conventionally deposited electrodes would cover the whole M/S contact region, incapable of forming a clean connection to either the Ni nanoparticles or the MoS2 flakes.

5.3.2 Conductive atomic force microscopy

Atomic force microscopy is a universal and convenient tool to study the surface topography of metallic, semiconducting, and insulating materials. By using a tip with conductive coating, AFM can also characterize electrical properties especially suitable

Substrate (Si)

SiO2 layer

MoS2

Dielectric layer (HfO2)

Drain(Au)

Top gate

Ids

Source(Au)

Vbg

Vtg

Vds

270nm

30nm

SiO2 layer

MoS2

Drain(Sc)

Ids

Source(Sc)

Vbg

Vds

100nm

Back gate (Si)

a b


35

for nano-scale samples [ 99 , 100 , 101 ]. The C-AFM technique incorporates high topographic resolution in tapping mode and high I-V sensitivity in contact mode or force mode. Typically, C-AFM can measure both the current map and the I-V curve [102]. A constant voltage is applied during the current imaging, and the tip maps the selected region with different current intensity. For the latter one, the tip is kept stationary and locates on certain position of the sample surface (lateral precision depends on the tip diameter), and the bias is varied by constant step [103]. The setup used for Papers II and IV is shown in Fig. 5.8.

Figure 5.8 Schematic illustration of C-AFM measurements. This figure shows Ni NPs are joined onto the surface of MoS2. In practice, appropriate Ni nanoparticles can be also those on the slope of multilayer MoS2 flakes.

During the I-V measurement, a constant contact force is applied to the AFM tip ensuring the reliability and the comparability of the results. From Fig. 5.7 and 5.8, it is clear that neither FET nor C-AFM method can directly give the specific interfacial resistance of an M/S contact. Instead, they determine the total resistance. For example, FET-based measurements give a total resistance consisting of contact resistance and channel resistance [104]. The C-AFM method (two-terminal method) measures a total resistance including the contact resistance of Tip-Ni, Ni-Au/Ag-MoS2, MoS2-Au film and also the intrinsic resistance of Ni, MoS2 and the AFM machine. Therefore, measurement of a control sample is always necessary.

The thermionic theory is also applicable for C-AFM data processing. In this regime, I-V curves with forward bias follow the relationship (modified from Eq. 5.8) [94]

( )

−−⋅

−⋅⋅⋅=

Tnk

IRVq

Tk

qTAAI

B

ttip

B

B2* expexpφ

, (5.9)

V

Ni

Ag

S

Mo


36

where A is the contact area ( 2tiprA π= ), *A is the Richardson constant (54.1 A·cm-2·K-

2 for bulk MoS2), T is temperature, q is the electron charge, Bφ is the Schottky barrier

potential, tipV is the applied bias, tR is the total resistance, and n is an ideality factor.

-6 -4 -2 0 2 4 6-300

-200

-100

0

100

200

300

400 MoS2-Au-Ni

MoS2-Ni

C-A

FM

Cur

ren

t (p

A)

Voltage (V)

a

0 1 2 3 4 5

0

20

40

60

80

100

Rc

Dro

p (

%)

Voltage (V)

b

Figure 5.9 C-AFM results. (a) I-V curves of MoS2-Au-Ni and MoS2-Ni contacts. (b) Resistance drop compared between the two types of contacts. Figures adapted from Paper II.

The I-V curves in Fig. 5.9a can be fitted following an exponential relationship. By taking the derivatives of the fitting plots, the total resistance values against applied voltage in both cases are achievable. Discrepancies between the two curves provide the information of the resistance drop (Fig. 5.9b).

Considering the term tIR in Eq. 5.9 is negligible with a small bias, the equation can be simplified, and then transformed as

( ) ( )Tnk

qV

Tk

qTAAI

B

tip

B

B2*lnln −−⋅⋅= φ. (5.10)

The Schottky barrier can thus also be roughly estimated. In addition to the I-V characteristics, investigation of the contact mechanism involves also the spectroscopy techniques such as the XPEEM (Paper II).

37

Chapter 6 MoS2-based photocatalysts

Besides the illumination, sunlight can be also utilized by means of, e.g., concentrated solar power and photovoltaic power systems. In 1972, photo-assisted water splitting was realized for the first time [ 105 ]. By using suitable catalysts, light of certain wavelengths can be the energy source to split water into H2 and O2. Conversion into H2, another clean and high-density energy source, is an alternative way of utilizing solar energy. This chapter briefly introduces the principles of photocatalysis, designs of photocatalysts, and the applications in photocatalytic degradation and hydrogen evolution from water.

6.1 Basic photocatalytic principles

Pure water is transparent to visible light, and thus it can be only decomposed under the irradiation of UV light or light with even shorter wavelengths [105]. Appropriate catalysts are needed to facilitate the splitting of water using visible light. A straightforward solution is adoption of semiconductors with bandgaps suitable for absorbing visible or UV light.

Light can be absorbed by a target sample. If the photon energy is larger than the bandgap of a semiconductor, electrons will be excited to the conduction band. The excited electrons and created holes then separate and migrate to the surface of the semiconductor and are involved in the redox reaction of water splitting, as shown in Fig. 6.1 [106,107]. An overall water splitting (Eq. 6.1) includes the oxygen evolution reaction (OER) and hydrogen evolution reaction (HER), where both O2 and H2 are generated. It is an uphill reaction that requires a Gibbs free energy of 237 kJ/mol [108],

222 O2

1HOH +→ . (6.1)

In a general case, Eq. 6.1 can be divided into two separate processes:

22 O2

12H2hOH +→+ ++ (6.2)

2H2e2H →+ −+ . (6.3)


38

Figure 6.1 One-step photocatalytic overall water splitting with a semiconductor photocatalyst. The potential values are based on the case of pH=0.

The minimum Gibbs free energy for an overall water splitting reaction (Eq. 6.1) requires a semiconductor bandgap of at least 1.23 eV [108] and it is the prerequisite for one-step photocatalytic water splitting. The semiconductor bandgap should also straddle the redox potentials of water. The CBM should be higher than the reduction potential of H+/H2, and VBM should be lower than the oxidation potential of O2/H2O. The redox potentials vary in different pH environments. For example, the H+/H2 potential is 0 and -0.41 versus the normal hydrogen electrode (NHE) at pH=0 and pH=7, respectively while the potential of O2/H2O is +1.23 and +0.82 in the two cases [107,108].

Figure 6.2 Two-step photocatalytic water splitting. The two semiconductors are combined through a solid-state mediator for electron transfer. The reduction and oxidation of water happen at either semiconductor. The dashed arrow denotes backward electron transfer pathways. The mediator can be either electrolyte solution or solid-state materials.

Alternatively, a two-step photocatalytic process is also possible where two semiconductors are involved [109]. In this case, one semiconductor is responsible for the reduction of H2O to H2 (hydrogen evolution photocatalyst, HEP) and the other facilitates the oxidation of H2O to O2 (oxygen evolution photocatalyst, OEP). Due to

Semiconductor

-

+

h+

e-

H+

H2

H2O

CBM

VBM

hνH+ + O2

Pot

entia

l (V

vs

NH

E)

at p

H=

7

-0.41(H+/H2)

0.82(O2/H2O)

OEP

-

+h+

H2O

CBM

VBM

hν

H+ + O2

Pot

entia

l (V

vs

NH

E)

at p

H=

7

-0.41(H+/H2)

0.82(O2/H2O)

HEP

-

+

e-

H+

H2

CBM

VBM

hν

e- e-

Electron mediator

6.2 Photocatalyst design

39

the combination of two semiconductor species, the CBM of OEP can be more positive than H+/H2 potential, and VBM of HEP more negative than O2/H2O potential (see Fig. 6.2). Therefore, such a scheme provides a larger driving force than the one-step case. More semiconductors are also applicable to be candidates of the catalyst-pair. Since the CBMs and VBMs of the two semiconductors usually position at different potentials, such two-step photocatalysis is named as Z-scheme process.

There are also disadvantages in the Z-scheme water splitting. In comparison to one-step reaction, more photons are consumed to produce the same amount of H2 and electron-hole pair recombination hinders the photocatalytic water splitting. Choosing suitable electron mediators is also a demanding task [110,111]. Either aqueous or solid-state mediators are possible, but both of them increase the complexity of the system and affect the catalyst stability.

In the case of photodegradation, the holes oxidize water to form H+ and OH· which can oxidize organic dyes while the photoexcited electrons reduce H+ to H· (Paper II). Oxygen is also possibly involved in the degradation in some cases [ 112 , 113 ]. Many organic substances can be degraded, such as methylene blue, rhodamine B, bisphenol A, etc.


6.2.1 Active sites of MoS2

In a photocatalytic HER process, Eq. 6.3 is actually an overall expression. Reducing water to H2 requires adsorption of hydrogen to active sites of catalysts in the first step [114,115]

*H*eH ⋅→++ −+ , (6.4)

where * represents an active site on the catalyst surface and *H ⋅ is a hydrogen radical adsorbed to an active site. Active sites provide places where electrons meet and react with hydrogen ions. The following step is the formation and release of H2 by either of the two pathways

*2H2H 2* +→⋅ (6.5)

*HeHH 2* +→++⋅ −+ . (6.6)

Active sites are involved in both the adsorption and desorption processes. Therefore, the amount and properties of active sites are crucial features of catalysts.

Both multilayer and monolayer MoS2 can be a good semiconductor host for photocatalyst fabrication. Its bandgap varies from 1.3 to 1.9 eV when thinned from bulk to single layer thus covering the water splitting potential at all thickness. The bandgap can also respond to sunlight in a wide wavelength range. The outer surface of MoS2


40

consists of various site species, e.g., vacancies, edges, and basal planes, as shown in Fig. 6.3. Among them, the S vacancy has the highest catalytic activity [116]. It appears mostly on the edges of both 2D and bulk MoS2 but less on bulk basal planes and seldom on basal planes of well-crystallized monolayers. Generally, one S vacancy leads to the rupture of two covalent bonds with Mo atoms [117] creating two dangling bonds. The newly formed dangling bonds are active for intimate chemical bonding or adsorption and beneficial to charge transfer [90].

Figure 6.3 Possible photocatalytic reaction sites on a multilayer MoS2 catalyst. Note that edge sites are mostly sulfided Mo-edges.

Edge sites were firstly found beneficial to hydrogen evolution [118]. As the dominant edge structures, the sulfided Mo-edges own intrinsic catalytic ability for HER [119]. The transition metal doped S-edges are also found active for photocatalysis [120]. Indeed, the edge sites are especially important for atomically thin MoS2 because dangling bonds exist only on the edges [90]. In the case of MoS2 2D nanoribbons, there are mainly two edge types: the zigzag and armchair edges. Zigzag nanoribbons bear metallic feature, therefore, they offer better electrical contacts with possible metallic dopants [61]. Thanks to the S vacancies, the zigzag ribbons are ferromagnetic, and their bandgaps can be tuned by applying external magnetic fields. Such bandgap tunability is promising for specific catalytic applications. Armchair edges are semiconducting, and the bandgaps are also tunable by altering the ribbon width.

However, basal planes of 2D MoS2 lack dangling bonds. This makes the basal planes inert for catalysis [121]. Slightly more dangling bonds exist on basal planes of multilayer or bulk MoS2, but those forms suffer from low specific surface area and smaller edge ratio. As a result, the catalytic ability of basal planes is also quite limited. Since the basal planes take a large ratio of the total surface area, activation of the inert planes will substantially improve the catalytic performance.

Mo

S

S vacancy

Edge

Basal plane


41

6.2.2 Enhancing photocatalytic ability

In principle, a promising semiconductor-based photocatalyst should own a suitable bandgap for wide sunlight response and capability for efficient electron-hole separation, migration, and consumption. Photocatalytic ability can be improved via two strategies: (1) manipulation of intrinsic crystal and electronic structures, and (2) alteration of reaction environment. The former involves the introduction of new active sites or activation of inert basal planes for prompt utilization of electrons and holes, while the latter normally requires additive electrolyte in the water/catalyst system.

New active sites can be introduced by the dispersion of metal nanoparticles on MoS2. As with Pt-decorated TiO2 catalyst, many transition metals or hetero-catalysts can be combined to MoS2 as co-catalysts. For example, Pt, Ru, and Ni accelerate HER process while Fe, Co, Ru, and Ni are beneficial to OER [107]. Graphene, CdS, TiO2, and other semiconductors are also popular co-catalysts [122,123]. Among them, Ni has been proven as a good promoter to increase catalytic ability with the possible locations on both MoS2 basal planes [124] and edges. Besides, the work function difference between Ni and MoS2 is favorable to electron migration from semiconductor to metal which in turn weakens the electron-hole recombination. However, the introduction of foreign metal nanoparticles needs either sophisticated processing (CVD, PVD) or buffer substance to join them to MoS2. Good chemical bonding and electrical connection are important for the catalytic property while van der Waals joints should be avoided [53]. By using Au or Ag as the buffers, Ni nanoparticles are successfully glued on the edges and basal planes of multilayer MoS2. See details in Papers II and IV. Due to the aqueous environment used during the catalyst synthesis and water splitting reaction, NiO is possible to be formed on the surface of nickel nanoparticles. This may not suppress the catalytic performance, because NiO is also found active for HER [107].

The above methods for incorporating co-catalysts are also possible to activate the MoS2 basal planes. Gold clusters introduced by thermal evaporation strongly interact with monolayer MoS2 and lead to enhancement of molecule adsorption and catalytic performance of basal planes [39]. In Paper IV, the surface contact between Ni nanoparticles and MoS2 has been verified. Compared to the edge contact mode in paper II, Ni decorated on MoS2 surface significantly reduces the M/S contact resistivity and enhances the photocatalytic efficiency. To the best of our knowledge, it is the first experimental realization of basal plane activation for multilayer MoS2 of several hundred-nanometers thick. Creation of S vacancies, which has been introduced in Chapter 2, can also increase the catalytic performance of the basal planes. Experimental work has been realized by argon or hydrogen plasma processing [47,48].

Alternatively, adding auxiliary additives to the H2O/MoS2 system can act as either sacrificial reagents or electron mediators for Z-scheme catalyzed reactions. The former case involves electrolytes serving as electron donors (hole scavengers) to promote the HER process. Lactic acid, ethanol, and triethanolamine (TEOA) are typically employed [125]. In the latter case, aqueous redox couples of IO3

-/I-, I3-/I- or Fe3

+/Fe2+ are popular


42

choices aiming at fast electron transportations [107]. Such electron mediators require either an acidic or basic environment. Therefore, the requirement of low or high pH environment is contradictory to pure water splitting. It is worth noting that the additive electrolytes are also necessities in the case of photoelectrochemical water splitting. They need to be replenished in long-term reactions.

6.3 Evaluation of photocatalytic ability

6.3.1 Hydrogen evolution efficiency

The photocatalytic efficiency for HER is typically described by quantum efficiency (QE) or the apparent quantum yield (AQY). The QE is calculated by the ratio of electron numbers involved in hydrogen evolution to the absorbed photon numbers per unit time [108]

%100photon

electron ×=N

NQE , (6.7)

where electronN and photonN are the number of photoexcited electrons involved in HER

and the number of absorbed photons per unit time, respectively. photonN can be roughly approximated by the attenuated light power due to absorption. Therefore, it is computed by the equation

λhcPP

N'

photon

−= , (6.8)

where P , 'P , and λ are power of light irradiating on the surface of H2O/catalyst aqueous system, power of light after passing through the aqueous system, and the wavelength of incident light, respectively. It is obvious that QE is an evaluation of HER efficiency under the incident light of a specific wavelength. It is not accurate to calculate QE with an average wavelength in case of polychromatic light.

The number of electrons contributing to HER can be inversely obtained from the amount of evolved hydrogen gas

t

NmN AH

electron2

2 ⋅⋅= , (6.9)

where 2Hm , t , and AN are the amount of evolved hydrogen gas in mole, reaction time,

and the Avogadro constant, respectively.

Quantum efficiency enables the comparison among various catalyst species, different light sources and reaction time. However, a more careful analysis should also include the amount of catalyst used. In this case, it is preferable to directly use the

6.4 Gas chromatography

43

amount of evolved H2 normalized by weight of the catalyst, light power, and reaction time. Such normalization makes the general comparison available among the results with different experimental conditions. The result can be also affected by the dispersion quality of the catalyst in water. Magnetic stirring is thus necessary to get evenly dispersed catalyst and enable a long-term reaction against catalyst aggregation.

In addition, the HER efficiency may vary greatly when changing the reaction cycle numbers, reaction time of one cycle, and the total reaction time. A convincing catalytic performance should incorporate experiments with both large number of cycles and a long-term continuous reaction.

6.3.2 Photocatalytic degradation ability

The photocatalytic ability to degrade organic pollutants is typically described by a removal curve or concentration ratio of C/C0, where C0 and C are the concentration before and after reaction. It presents the variation of pollutant concentration versus reaction time. Several control groups are also required for a complete experiment. A photocatalysis test without catalyst should be done to eliminate the effect of self-degradation [ 126 ]. A dark period should be performed for all tests before light irradiation in order to study the effect of adsorption. The degradation performance of photocatalysts does not only depend on their intrinsic ability but also on external conditions, e.g., power and wavelength of incident light, weight of catalysts, initial concentration of pollutants, effect of additive oxidants, etc. Thus, it is important to use the same conditions for efficiency evaluations. In addition, the products after degradation need to be analyzed. The measurement of total organic carbon gives brief information about the degradation of organic carbon into gaseous products.

The above discussion is mainly the direct quantification of photocatalytic HER efficiency and degradation ability. In the field dedicated to catalyst science, the specific surface area (SSA) is a popular concept as an indirect descriptor of possible catalytic performance [127]. It indicates the surface area per unit weight and the ability of adsorption. Generally speaking, high SSA allows more adsorption sites which are helpful for better catalytic ability. However, the SSA is highly dependent on the lateral size and vertical thickness of the catalysts. Samples in monolayer form have much higher SSA than bulk materials. Annealing and drying processes also affect SSA values. Therefore, SSA should be used along with other materials properties to evaluate the catalytic ability.

6.4 Gas chromatography

Gas chromatography (GC) is a common technique to identify gas species and analyze their concentrations. In HER applications, a GC machine can directly give the amount of evolved hydrogen gas by a more straightforward route than measuring current density. During the GC measurement, sample gas is transported by inert carrier gas and


44

then injected into a chromatographic column. The sample gas flows through the column and is then analyzed by a thermal conductivity detector (TCD). The detector responds to the difference in thermal conductivity between a reference cell (carrier gas only) and a measurement cell (carrier gas containing sample gas). Gas mixtures can be thus separated and identified.

In Paper IV, the H2 produced by HER was measured by an Agilent 490 Micro-GC where a Molsieve 5Å column was installed. During normal measurements (Fig. 6.4a), only H2 can be measured since the reaction was carried out under ambient air. Such experimental condition was designed based on the finding that HER efficiency was nearly unchanged under either air or Ar ambience. The O2 production is negligible compared to its original proportion in air. However, preliminary tests indeed showed simultaneous increase of H2 and O2 where the remaining air is partially replaced by Ar gas (Fig. 6.4b). It confirms the overall water splitting due to release of both H2 and O2.

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

-50

0

50

100

150

200

250N2

H2

Sig

nal

(m

V)

Time (min)

a

O2

0.2 0.3 0.4 0.5 0.6

N2

O2

Nor

mal

ize

d S

ign

al (

a.u

.)

Time (min)

Initial

H2

5 h

b

Figure 6.4 Gas quantification. (a) H2 concentration after 168 h reaction. (b) Normalized results where the HER was performed under the ambience of air/Ar mixture. The dashed plot is measured before HER and the plot in red is measured after 5h HER. The two spectra were normalized to have identical N2 peaks.

45

Chapter 7 Research summary and discussion

This chapter summarizes the main results of the papers included in this thesis. Through wet chemical synthesis, noble metals (Au, Ag) and nickel nanoparticles are joined to layered MoS2. The morphologies of introduced Au, Ag, and Ni NPs are controllable by altering reagent ratios and synthetic conditions. The MoS2-Au-Ni and MoS2-Ag-Ni composites are found to have low electrical resistivity at the contact interfaces. Effective charge transfer is also verified at the interface. The ternary composites exhibit enhanced photocatalytic ability for dye degradation and water splitting. The morphological, electronic, and photocatalytic properties are briefly presented in this chapter.

7.1 Morphologies

The main objective was to grow transition metal NPs onto layered MoS2 for tailored materials properties. The first progress was the introduction of Au NPs to MoS2 edges in Paper I. For commercial MoS2 nanopowder, most defects (active sites) are at the edge slopes. The electron-spectroscopic image in Fig. 7.1a shows that Au NPs are mainly grown at MoS2 edges and also a few on the surface or buried inside of MoS2 stacks.

In Fig. 7.1b, Au NPs can also grow to the Ni surface by the redox reaction between HAuCl4 and Ni NPs with the same synthetic method in Paper I. Such finding provides a promising route that Au can be a nanoglue for joining transition metal NPs to MoS2 multilayers. This idea was experimentally verified in Paper II (Fig. 7.1c, d). The Au NPs are crystallized with different sizes (50 nm in c and below 5 nm in d) and appear mainly on the MoS2 edges. The Ni NPs are bridged to MoS2 via the nanocrystallized Au or MoS2-Au-Ni ternary alloy. The chemically bonded MoS2-Ni composites survived from the mechanical exfoliation by the Scotch tape method.

Au NPs also appear at MoS2 surfaces (Fig. 7.1a, c) which suggests the possibility of basal plane activation. However, Au NPs get easily to larger sizes due to locally aggregated Au3+ NPs or increased reaction temperature. It weakens the ability of Au NPs to attract and adhere to either MoS2 basal plane or Ni NPs. Thus, no Ni was found on MoS2 basal plane in using the Au buffer. In Paper IV, the AgNO3 was used as the buffer reagent instead of HAuCl4. It is in favor of a mild redox reaction, leading to the reduced Ag NPs in uniform but small sizes. The Ni NPs are successfully decorated on both the edges and basal planes of MoS2 which can be seen from Fig. 7.1e and f.

46

Figure 7.1 Microscopic study of the synthesized MoS2-Ni, Au-Ni, MoS2-Au-Ni and MoS2-Ag-Ni composites. (a) Electron energy loss spectroscopy (EELS) image of MoS2-Ni sample. (b) TEM image of Au-Ni. (c, d) TEM-EDS images of MoS2-Au-Ni. (e, f) TEM-EDS images of MoS2-Ag-Ni. Panels (a) from Paper I, (c, d) from II and (e, f) from IV.

7.2 Evaluation of metal/semiconductor contacts

The Ni NPs with the sizes of 100 ~ 200 nm are combined with semiconducting multilayer MoS2. Such a metal-semiconductor regime could be used for electronic devices. However, clarification of the contact mode and estimation of electrical resistivity are critical for the nano-scale electronic devices.

The research on electrical resistivity was carried out in Papers II and IV for the contacts of MoS2-Au-Ni and MoS2-Ag-Ni, respectively. The contact resistivity is computed from I-V curves measured by C-AFM. A direct MoS2-Ni contact without noble metal buffer was also fabricated for comparison. The resistivity of all the three types of contacts drop along applied bias (Fig. 7.2a). However, such behaviors are caused by different contact properties. It should be noted that Schottky barriers exist in

100 nm

0.5 µm

500 nm

Mo Au Ni

25 nm

20 nm

Ni

100 nm

100 nm

100 nm

50 nm

200 nm 150 nm

Mo Ag Ni

a b

dc

fe

47

all the three contacts. The barriers will be decreased by applying higher voltages [83,97]. The direct Ni-MoS2 contact has very high resistivity due to the poor contact quality, i.e., high Schottky barrier. Although the barrier is lowered a bit at high voltages, the resistivity is still significant (46.3 mmΩ ⋅ at 5 V and 5.5 mmΩ ⋅ at 8 V).

0 2 4 6 8

1

10

100

1000 MoS2-Ni

MoS2-Au-Ni

MoS2-Ag-Ni

Res

istiv

ity (

Ω⋅m

m)

Voltage (V)

a

0 2 4 6 8

20

40

60

80

100

Re

sist

ivity

Dro

p (

%)

Voltage (V)

MoS2-Ag-Ni

MoS2-Au-Ni

b

Figure 7.2 Electrical resistivity of MoS2-Ni, MoS2-Au-Ni and MoS2-Ag-Ni contacts. (a) Resistivity of the three types of contacts. (b) Resistivity drop compared to direct Ni-MoS2 contact. Figures are summarized from Papers II and IV.

When the Ni and MoS2 are buffered by the Au nanoglue, the contact resistivity drops much more rapidly to 3.7 mmΩ ⋅ at 5 V (Fig. 7.2b). However, resistivity is still obvious at low voltages. The resistivity drop compared to direct Ni-MoS2 contact reaches more than 90% at 5 V. In contrast, the Ni-Ag-MoS2 contact exhibits low resistivity over the full voltage range (2.8 mmΩ ⋅ at 5 V and 0.5 mmΩ ⋅ at 8 V). The resistivity drop changes smoothly but is always over 90%. Such electrical performance suggests a highly reduced Schottky barrier at the presence of the Ag buffer.

The bonding at the contact interface between Ni and MoS2 is studied to reveal the physical mechanisms of M/S contact. However, the very small size of the joining interface demands simultaneous microscopic and spectroscopic determinations. For this reason, the SR-based XPEEM was employed.

48

Figure 7.3 XPEEM results. (a) SEM image of Ni-Au-MoS2 composites. Two interface regions and one bare Ni region are marked with green, red, and blue circles, respectively. (b) XAS spectra of the three regions in (a). (c) Zoomed-in image of the square region in (a). Two Ni/MoS2 interfaces are marked with dashed squares. A sequence of acentric circular regions is marked in each region. (d) Calculated percentages of bonded Ni atoms of the two interfaces at different distances to the interface. Panels (a, b) are adapted from Paper II and (c, d) from Paper III.

XPEEM technique allows the detection of total electron yield from a selected region. When comparing the Ni 2p L2,3 edge XAS spectra from the three regions in Fig. 7.3a, an unexpected peak at 861 eV was found where there is MoS2-Ni interface. On the contrary, such feature vanishes in the bare Ni spectrum. The additional peak is attributed to the Ni-S bonding in confined spaces [128].

In addition, the peak is more obvious from Region 1 where a large nodule is observed at the interface. The nodule was most probably formed during the sonochemical synthesis where a large number of Ni atoms are alloyed to MoS2 with the existence of Au buffer. Also, the peak intensity suggests the amount of bonded Ni atoms. The ratio of this peak to Ni 2p3/2 feature denotes the percentage of bonded Ni atoms (Fig. 7.3d). As expected, there are 50% ~ 60% of Ni atoms involved in the bonding formation, and the percentage falls with larger distance to the interface. This finding may be helpful to adjust the size of metal NPs for better M/S contacts.

100 nmInterface B

Ni

MoS2

Ni

200 nm

Region 1 Region 2Bare Ni

a

c

845 855 865 875

Region 1 Region 2 Bare Ni

Inte

nsity

(a.

u.)

Photon Energy (eV)

25 50 75 100 125 150

0.0

0.2

0.4

0.6

0.8 Region A Region B

Bon

ded

Ni A

tom

s (%

)

Distance to Interface (nm)

b

d

49

7.3 Photocatalytic applications

Following the realization of the MoS2-Au-Ni heterojunction, the photocatalytic ability of the fabricated metal/MoS2 composites was first investigated through degrading the MB (Fig. 7.4a). The ternary MoS2-Au-Ni composite shows better catalytic ability compared to pristine MoS2 and Au-decorated MoS2. During the degradation, the electron-hole pairs are created following the scheme in Fig. 6.1. The holes can oxidize water to ⋅OH which can subsequently react with organic dyes. On the other hand, the photogenerated electrons can easily migrate to Ni surface and reduce +H to ⋅H .

The hydrogen evolution reaction may undergo similar procedures as the dye degradation provided the ⋅H radicals are formed and further release H2 molecules. In Paper IV, the synthesized (MoS2)84Ag10Ni6 catalyst was employed for photocatalytic HER under visible light or sunlight. A time course HER is shown in Fig. 7.4b. It delivered an average HER rate of 73 μmol·g-1·W-1·h-1 during the first 6 reaction cycles. After 86 days of storage in water in ambient air, the photocatalyst preserves its activity without significant decay. The HER was performed in air ambience. Therefore, precise monitoring of the variation of O2 concentration is not applicable. However, a preliminary test under the ambience of air/Ar mixture (see Fig. 6.4b) shows a simultaneous increase of O2 and H2 referring to an overall water splitting. It should be mentioned that all the reaction cycles are performed without any sacrificial reagents. Such pure water splitting has been pursued for many years but restricted by relatively low HER efficiency. Compared to the results in Ref. [129], the HER efficiency in Paper IV has increased by 5890 times.

The photocatalysis is durable in time and versatile in ambient conditions. A 10-day continuous water splitting experiment resulted in a hydrogen production rate of 44 μmol·g-1·h-1·W-1 from pure water (Fig. 7.4c). It reaches as much as 63% of the short-term result and the decrease of reactivity probably comes from the accumulated H· radicals in liquid which hinders the reduction of the H+. The catalyst can also split natural water with promising efficiency, for example, producing H2 from river water as much as 74% of pure water’s efficiency.

50

Figure 7.4 Photocatalytic properties. (a) Photocatalytic degradation of MB under UV light. (b) Hydrogen evolution from pure water under visible light. (c) Long-term HER from natural water. (d) Decoloration of river and lake water simultaneously with the HER process. (e) HER under natural sunlight. Panel (a) from Paper II and (b ~ e) from Paper IV.

The decoloration of organics-contaminated water was also observed (Fig. 7.4d). It suggests the possibilities of simultaneous hydrogen production and water decontamination. However, the decoloration process consumes photoexcited electrons and holes, competing with the formation of hydrogen and thus yielding very low amount of hydrogen from densely colored water. Much less hydrogen is produced from river and lake water during the first day, which is probably due to the competing

51

decoloration process. The photocatalyst is also active under indoor sunlight (Fig. 7.4e). However, the tests with sunlight is uncontrollable due to cloudy and rainy weather.

The enhanced photocatalytic ability of MoS2-Au-Ni and MoS2-Ag-Ni composites is attributed to the charge transfer scheme between MoS2 and Ni. Specifically, the nanocrystallized Ag (111) face matches the MoS2 surface but the Au (111) counterpart does not. Therefore, the Au can only locate on the MoS2 edges while Ag can activate the basal planes along with the edge activation. Besides, the dangling bonds from Ni surface are found beneficial to charge transfer in the MoS2-Ag-Ni system. This can be also verified from Fig. 7.2 where the contact resistivity in the case of Ag buffer is much lower than the MoS2-Au-Ni.

Although the current work is dedicated to multilayer MoS2, the application to few-layer or even monolayer forms could also be possible. Besides synthesizing directly with thin-layers, mechanical or chemical exfoliation is also doable to get 2D samples after synthesis.

7.4 Stability of the synthesized composites

MoS2-based nanocomposites are usually stable under ambient conditions, but aging may happen if they are exposed simultaneously to radiation, moisture, and heat [130]. Hence, the stability information is discussed and presented here.

In Paper II, MoS2-Au-Ni composites were exposed to an electron beam twice for SEM/EDS characterization and synchrotron radiation for XPEEM measurement. There is no visible morphology variation from Fig. 7.5.

Figure 7.5 SEM images were taken before and after the SR-based XPEEM. (a) Before XPEEM tests, image taken on 21.01.2015. (b) After XPEEM tests, image taken on 25.02.2015. Figures from Paper II.

52

The chemical states stability was studied through XPS. Figures 7.6a and b are the Mo 3d and S 2p XPS spectra of commercial MoS2 and newly synthesized MoS2-Au-Ni composite. The Mo is partially oxidized by Au3+ during the synthesis. It can be seen that the relative contents of Mo4+(dioxide) and Mo6+ are larger than pure MoS2. But this does not tell about the stability of Mo and S elements in the synthesized ternary composites. In fact, the Mo and S species of the composites are stable during HER.

In Fig. 7.6c, XPS spectra of fresh MoS2-Ag-Ni and the one after HER are presented. Both Mo and S are chemically stable subject to the 40-hour photocatalytic reaction. Quantitative XPS analysis reveals negligible content variation, i.e., atomic percentages of Mo4+(MoS2), Mo4+(dioxide) and Mo6+ vary from 90.98%, 5.55%, 3.48% to 90.02%, 5.89% and 4.08%, respectively. There is no significant chemical-state change for Ag element either (Fig. 7.6d).

Figure 7.6 XPS study of Mo 3d, S 2p and Ag 3d. (a) MoS2. (b) Fresh MoS2-Au-Ni. (c) Fresh and used MoS2-Ag-Ni. (d) Ag 3d. Figures from Papers II and IV.

In contrast, Ni species do change after the photocatalytic process (Fig. 7.7). Nickel oxides are partially oxidized to Ni(OH)2 and NiOOH during the reduction of H2. According to the surface sensitive XPS results, there are more Ni(OH)2 and NiOOH formed on the surface of Ni NPs. But the thickness of the Ni(OH)2 and NiOOH layers

2 3 5 2 3 0 2 2 5 1 6 5 1 6 0

Inten

sity (

a.u.)

B i n d i n g E n e r g y ( e V )

F r e s h M o S 2 - A g - N i

M o S 2 - A g - N i a f t e r 4 0 h H E R

c

S 2 s

M o 3 d S 2 p

M o S 2 M o S 2

D i o x i d eM o 6 +

3 7 8 3 7 6 3 7 4 3 7 2 3 7 0 3 6 8 3 6 6 3 6 4

A g 3 d 3 / 2

Inten

sity (

a.u.)


F r e s h M o S 2 - A g - N i

M o S 2 - A g - N i a f t e r 4 0 h H E R

A g 3 d 5 / 2d

2 3 5 2 3 0 2 2 5 1 6 5 1 6 0

S 2 p 1 / 2

S 2 p 3 / 2

D i o x i d e

Inten

sity (

a.u.)


M o S 2 3 d 5 / 2

S 2 s

F r e s h M o S 2 - A u - N i

M o 6 +

b

2 3 5 2 3 0 2 2 5 1 6 5 1 6 0

S 2 p 1 / 2

S 2 p 3 / 2

D i o x i d eInten

sity (

a.u.)


M o S 2 3 d 5 / 2

S 2 s

C o m m e r c i a l M o S 2

M o 6 +

a

53

is unknown. From the application point of view, it may not be a problem for H2 production. There is no obvious drop of photocatalytic ability during long-term continuous HER (Fig. 7.4c,e). The catalysts remain efficient for HER even after being kept in water for 3 months (Fig. 7.4b).

Figure 7.7 Ni 2p XPS spectra of MoS2-Ag-Ni composites before and after HER. Figure from Paper IV.

However, further understanding of the variation of Ni species during HER is necessary. Such study demands future work involving the SR-based high-energy XPS or depth profile XPS techniques which are out of the scope of the included papers.

8 8 0 8 7 5 8 7 0 8 6 5 8 6 0 8 5 5 8 5 0

N i ( O H ) 2

N i 2 p

S a t e l l i t e sN i O O H

N i ON i m e t a l


F r e s h

A f t e r 4 0 h r e a c t i o nInten

sity (a

.u.)

54

Chapter 8 Conclusions and outlook

Molybdenum disulfide was experimentally manipulated by decoration of noble metals or other transition metal nanoparticles. The employed wet-chemical synthetic method, especially accompanied by an ultrasound source, is proven to be versatile for the introduction of transition metal nanoparticles to inorganic layered crystals. Adoption of a noble metal as a nanobuffer leads to stable and effective connection between MoS2 and transition metals. The electronic and catalytic properties of the synthesized composites have been comprehensively studied. The ternary composites were observed to have reduced Ni/MoS2 contact resistivity. Synchrotron radiation based X-ray photoemission electron spectroscopy was used to reveal the contact mechanism. The low electrical resistivity is ascribed to the efficient charge transfer at the metal-semiconductor interface where the Ni-S bonding has been experimentally verified.

Inspired by the photodegradation performance of MoS2-Au-Ni, the photocatalytic activity was then investigated, mainly focusing on the hydrogen evolution from water under visible light irradiation. The hydrogen production efficiency is very promising for future eco-friendly fuels considering the efficiency and reaction conditions. Such photocatalytic behavior probably arises from the activation of MoS2 basal planes. Previously such activation was only reported for monolayer MoS2 and most of the reports were just theoretical predictions. This thesis work may attract more attention to multilayer MoS2. It is much closer to industrial application.

Overall, this work provides a facile and low-cost way of tuning the physical and chemical properties of inorganic layered crystals or introducing new features to them. The synthesized MoS2-based composites have announced a new series of candidates for potential applications in both the electronics and hydrogen energy industry. It is worth noting that the thesis work received support from theoretical calculations. The incorporation of experimental and theoretical techniques serves deep insights of physical mechanisms in materials formation and functionalities along with materials performances in various domains.

55

Bibliography

[1] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva, A. A. Firsov. Electric field effect in atomically thin carbon films. Science 306, 666−669 (2004).

[2] K. S. Novoselov, D. Jiang, F. Schedin, T. J. Booth, V. V. Khotkevich, S. V. morozov, A. K. Geim. Two-dimensional atomic crystals. Proc. Natl. Acad. Sci. 102, 10451−10453 (2015).

[3] A. K. Geim, K. S. Novoselov. The rise of graphene. Nat. Mater. 6, 183−191 (2007).

[4] M. I. Katsnelson. Graphene: carbon in two dimensions. Mater. Today 10, 20−27 (2007).

[5] K. F. Mak, C. Lee, J. Hone, J. Shan, T. F. Heinz. Atomically thin MoS2: a new direct-gap semiconductor. Phys. Rev. Lett. 105, 136805 (2010).

[6] J. N. Coleman, M. Lotya, A. O’Neill, S. D. Bergin, P. J. King, U. Khan, K. Young, A. Gaucher, S. De, R. J. Smith, et al. Two-dimensional nanosheets produced by liquid exfoliation of layered materials. Science 331, 568−571 (2011).

[7] M. M. Benameur, B. Radisavljevic, J. S. Héron, S. Sahoo, H. Berger, A. Kis. Visibility of dichalcogenide nanolayers. Nanotechnology 22, 125706 (2011).

[8] M. A. McGuire. Crystal and magnetic structures in layered, transition metal dihalides and trihalides. Crystals 7, 121 (2017).

[9] A. K. Geim, I. V. Grigorieva. Van der Waals heterostructures. Nature 499, 419−425 (2013).

[10] P. Miró, M. Audiffred, T. Heine. An atlas of two-dimensional materials. Chem. Soc. Rev. 43, 6537−6554 (2014).

[11] S. Z. Butler, S. M. Hollen, L. Cao, Y. Cui, J. A. Gupta, et al. Progress, challenges, and opportunities in two-dimensional materials beyond graphene. ACS Nano 7, 2898−2926 (2013).

[12] M. Chhowalla, H. S. Shin, G. Eda, L.-J. Li, K. P. Loh, H. Zhang. The chemistry of two-dimensional layered transition metal dichalcogenide nanosheets. Nat. Chem. 5, 263−275 (2013).

[13] B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti, A. Kis. Single-layer MoS2 transistors. Nat. Nanotech. 6, 147−150 (2011).

[14] C. Gong, L. Li, Z. Li, H. Ji, A. Stern, et al. Discovery of intrinsic ferromagnetism in two-dimensional van der Waals crystals. Nature 546, 265−269 (2017).

56

[15] B. Huang, G. Clark, E. Navarro-Moratalla, D. R. Klein, R. Cheng, et al. Layer-

dependent ferromagnetism in a van der Waals crystal down to the monolayer limit. Nature 546, 270−273 (2017).

[16] M. Xu, T. Lian, M. Shi, H. Chen. Graphene-like two-dimensional materials. Chem. Rev. 113, 3766−3798 (2013).

[17] S. Cahangirov, M. Topsakal, E. Aktürk, H. Şahin, and S. Ciraci. Two- and one-dimensional honeycomb structures of silicon and germanium. Phys. Rev. Lett. 102, 236804 (2009).

[18] M. Akhtar, G. Anderson, R. Zhao, A. Alruqi, J. E. Mroczkowska, G. Sumanasekera, J. B. Jasinski. Recent advances in synthesis, properties, and applications of phosphorene. NPJ 2D Mater. Appl. 1, 5 (2017).

[19] S. Balendhran, S. Walia, H. Nili, S. Sriram, M. Bhaskaran. Elemental analogues of graphene: silicene, germanene, stanene, and phosphorene. Small 11, 640−652 (2015).

[20] L. Liu, Y. P. Feng, Z. X. Shen. Structural and electronic properties of h-BN. Phys. Rev. B 68, 104102 (2003).

[21] H. Li, J. Wu, Z. Y. Yin, H. Zhang. Preparation and applications of mechanically exfoliated single- and multi-layer MoS2 and WSe2 nanosheets. Acc. Chem. Res. 47, 1067−1075 (2014).

[22] M. B. Dines. Lithium intercalation via n-butyllithium of the layered transition metal dichalcogenides. Mater. Res. Bull. 10, 287−291 (1975).

[23] P. Joensen, R. F. Frindt, S. R. Morrison. Single-layer MoS2. Mater. Res. Bull. 21, 457−461 (1986).

[24] Y.-C. Lin, D. O. Dumcenco, Y.-S. Huang, K. Suenaga. Atomic mechanism of the semiconducting-to-metallic phase transition in single-layered MoS2. Nat. Nanotech. 9, 391−396 (2014).

[25] S. Manzeli, D. Ovchinnikov, D. Pasquier, O. V. Yazyev, A. Kis. 2D transition metal dichalcogenides. Nat. Rev. Mater. 2, 17033 (2017).

[26] L. F. Matthesis. Band structure of transition-metal-dichalcogenide layer compounds. Phys. Rev. B 8, 3719–3740 (1973).

[27] S. J. Sandoval, D. Yang, R. F. Frindt, J. C. Irwin. Raman study and lattice dynamics of single molecular layers of MoS2. Phys. Rev. B. 44, 3955–3962 (1991).

[28] G. Eda, H. Yamaguchi, D. Voiry, T. Fujita, M. Chen, M. Chhowalla. Photoluminescence from chemically exfoliated MoS2. Nano Lett. 11, 5111−5116 (2011).

[29] D. Voiry, A. Mohite, M. Chhowalla. Phase engineering of transition metal dichalcogenides. Chem. Soc. Rev. 44, 2702−2712 (2015).

57

[30] H. Schmidt, F. Giustiniano, G. Eda. Electronic transport properties of transition

metal dichalcogenide field-effect devices: surface and interface effects. Chem. Soc. Rev. 44, 7715−7736 (2015).

[31] Q. H. Wang, K. Kalantar-Zadeh, A. Kis, J. N. Coleman, Michael S. Strano. Electronics and optoelectronics of two-dimensional transition metal dichalcogenides. Nat. Nanotech. 7, 699−712 (2012).

[32] O. V. Yazyev, A. Kis. MoS2 and semiconductors in the flatland. Mater. Today 18, 20−30 (2015).

[33] Y. Ding, Y. Wang, J. Ni, L. Shi, S. Shi, W. Tang. First principles study of structural, vibrational and electronic properties of graphene-like MX2 (M=Mo, Nb, W, Ta; X=S, Se, Te) monolayers. Phys. B 406, 2254−2260 (2011).

[34] C. Tan, X. Cao, X. J. Wu, Q. He, J. Yang, X. Zhang, J. Chen, W. Zhao, S. Han, G.-H. Nam, M. Sindoro, H. Zhang. Recent advances in ultrathin two-dimensional nanomaterials. Chem. Rev. 117, 6225−6331 (2017).

[35] Y.-C. Lin, D. O. Dumcenco, H.-P. Komsa, Y. Niimi, A. V. Krasheninnikov, Y.-S. Huang, K. Suenaga. Properties of individual dopant atoms in single-layer MoS2: atomic structure, migration, and enhanced reactivity. Adv. Mater. 26, 2857−2861 (2014).

[36] Y. Shi, J.-K. Huang, L. Jin, Y.-T. Hsu, S. F. Yu, L.-J. Li, H. Y. Yang. Selective decoration of Au nanoparticles on monolayer MoS2 single crystals. Sci. Rep. 3, 1839 (2013).

[37] A. Y. Polyakov, L. Yadgarov, R. Popovitz-Biro, V. A. Lebedev, I. Pinkas, R. Rosentsveig, Y. Feldman, A. E. Goldt, E. A. Goodilin, R. Tenne. Decoration of WS2 nanotubes and fullerene-like MoS2 with gold nanoparticles. J. Phys. Chem. C 118, 2161−2169 (2014).

[38] T. S. Sreeprasad, P. Nguyen, N. Kim, V. Berry. Controlled, defect-guided, metal-nanoparticle incorporation onto MoS2 via chemical and microwave routes: electrical, thermal, and structural properties. Nano Lett. 13, 4434−4441 (2013).

[39] C. S. Merida, D. Le, E. M. Echeverria, et al. Gold dispersion and activation on the basal plane of single-layer MoS2. J. Phys. Chem. C 122, 267−273 (2018).

[40] A. J. Cheah, W. S. Chiu, P. S. Khiew, H. Nakajima, T. Saisopa, P. Songsiriritthigul, S. Radimane, M. A. A. Hamid. Facile synthesis of a Ag/MoS2 nanocomposite photocatalyst for enhanced visible-light driven hydrogen gas evolution. Catal. Sci. Technol. 5, 4133−4143 (2015).

[41] X. Huang, Z. Zeng, S. Bao, M. Wang, X. Qi, Z. Fan, H. Zhang. Solution-phase epitaxial growth of noble metal nanostructures on dispersible single-layer molybdenum disulfide nanosheets. Nat. Commun. 4, 1444 (2013).

[42] D. Escalera-López, Y. Niu, J. Yin, K. Cooke, N. V. Rees, R. E. Palmer. Enhancement of the hydrogen evolution reaction from Ni-MoS2 hybrid nanoclusters. ACS Catal. 6, 6008−6017 (2016).

58

[43] H. Wang, C. Tsai, D. Kong, K. Chan, F. Abild-Pedersen, J. K. Nørskov, Y. Cui.

Transition-metal doped edge sites in vertically aligned MoS2 catalysts for enhanced hydrogen evolution. Nano Res. 8, 566−575 (2015).

[44] M. Zhang, Z. Huang, X. Wang, H. Zhang, T. Li, Z. Wu, Y. Luo, W. Cao. Magnetic MoS2 pizzas and sandwiches with Mnn (n=1-4) cluster toppings and fillings: a first-principles investigation. Sci. Rep. 6, 19504 (2016).

[45] A. N. Andriotis, M. Menon. Tunable magnetic properties of transition metal doped MoS2. Phys. Rev. B 90, 125304 (2014).

[46] M. A. Khan, M. Erementchouk, J. Hendrickson, M. N. Leuenberger. Electronic and optical properties of vacancy defects in single-layer transition metal dichalcogenides. Phys. Rev. B 95, 245435 (2017).

[47] H. Li, C. Tsai, A. L. Koh, L. Cai, A. W. Contryman, A. H. Fragapane, J. Zhao, H. S. Han, H. C. Manoharan, F. Abild-Pedersen, J. K. Nørskov, X. Zheng. Activating and optimizing MoS2 basal planes for hydrogen evolution through the formation of strained sulphur vacancies. Nat. Mater. 15, 48−53 (2016).

[48] C.-C. Cheng, A.-Y. Lu, C.-C. Tseng, X. Yang, M. N. Hedhili, M.-C. Chen, K.-H. Wei, L.-J. Li. Activating basal-plane catalytic activity of two-dimensional MoS2 monolayer with remote hydrogen plasma. Nano Energy 30, 846−852 (2016).

[49] Y. Ouyang, C. Ling, Q. Chen, Z. Wang, L. Shi, J. Wang. Activating inert basal planes of MoS2 for hydrogen evolution reaction through the formation of different intrinsic defects. Chem. Mater. 28, 4390−4396 (2016).

[50] K. Zhang, J. K. Kim, B. Park, S. Qian, B. Jin, X. Sheng, H. Zeng, H. Shin, S. H. Oh, C.-L. Lee, J. H. Park. Defect-induced epitaxial growth for efficient solar hydrogen production. Nano Lett. 17, 6676−6683 (2017).

[51] A. Ebnonnasir, B. Narayanan, S. Kodambaka, C. V. Ciobanu. Tunable MoS2 bandgap in MoS2-graphene heterostructures. Appl. Phys. Lett. 105, 031603 (2014).

[52] L. Guo, Z. Yang, K. Marcus, Z. Li, B. Luo, L. Zhou, X. Wang, Y. Du, Y. Yang. MoS2/TiO2 heterostructures as nonmetal plasmonic photocatalysts for highly efficient hydrogen evolution. Energy Environ. Sci. 11, 106−114 (2018).

[53] J. Li, G. Zhan, Y. Yu, L. Zhang. Superior visible light hydrogen evolution of Janus bilayer junctions via atomic-level charge flow steering. Nat. Commun. 7, 11480 (2016).

[54] M. Farmanbar, G. Brocks. Controlling the Schottky barrier at MoS2/metal contacts by inserting a BN monolayer. Phys. Rev. B 91, 161304(R) (2015).

[55] M. K. L. Man, S. Deckoff-Jones, A. Winchester, G. Shi, G. Gupta, A. D. Mohite, S. Kar, E. Kioupakis, S. Talapatra, K. M. Dani. Protecting the properties of monolayer MoS2 on silicon based substrates with an atomically thin buffer. Sci. Rep. 6, 20890 (2016).

59

[56] Q. Wang, T. Hisatomi, Q. Jia, H. Tokudome, M. Zhong, C. Wang, Z. Pan, T.

Takata, M. Nakabayashi, N. Shibata, Y. Li, I. D. Sharp, A. Kudo, T. Yamada, K. Domen. Scalable water splitting on particulate photocatalyst sheets with a solar-to-hydrogen energy conversion efficiency exceeding 1%. Nat. Mater. 15, 611−615 (2016).

[57] D. Voiry, A. Goswami, R. Kappera, C. de Carvalho Castro e Silva, D. Kaplan, T. Fujita, M. Chen, T. Asefa, M. Chhowalla. Covalent functionalization of monolayered transition metal dichalcogenides by phase engineering. Nat. Chem. 7, 45−49 (2015).

[58] R. Kappera, D. Voiry, S. E. Yalcin, B. Branch, G. Gupta, A. D. Mohite, M. Chhowalla. Phase-engineered low-resistance contacts for ultrathin MoS2 transistors. Nat. Mater. 13, 1128−1134 (2014).

[59] T. Dietl. A ten-year perspective on dilute magnetic semiconductors and oxides. Nat. Mater. 9, 965−974 (2010).

[60] K. Dolui, I. Rungger, D. C. Pemmaraju, S. Sanvito. Possible doping strategies for MoS2 monolayers: An ab initio study. Phys. Rev. B 88, 075420 (2013).

[61] R. Ganatra, Q. Zhang. Few-layer MoS2: a promising layered semiconductor. ACS Nano 8, 4074−4099 (2014).

[62] L. Cai, J. He, Q. Liu, T. Yao, L. Chen, W. Yan, F. Hu, Y. Jiang, Y. Zhao, T. Hu, Z. Sun, S. Wei. Vacancy-induced ferromagnetism of MoS2 nanosheets. J. Am. Chem. Soc. 137, 2622−2627 (2015).

[63] A. Li, J. Pan, Z. Yang, L. Zhou, X. Xiong, F. Ouyang. Charge and strain induced magnetism in monolayer MoS2 with S vacancy. J. Magn. Magn. Mater. 451, 520–525 (2018).

[64] O. Ochedowski, K. Marinov, G. Wilbs, G. Keller, N. Scheuschner, D. Severin, M. Bender, J. Maultzsch, F. J. Tegude, M. Schleberger. Radiation hardness of graphene and MoS2 field effect devices against swift heavy ion irradiation. J. Appl. Phys. 113, 214306 (2016).

[65] C. J. Foot. Atomic Physics. (Oxford University Press, 2005).

[66] J. García Solé, L. E. Bausá, D. Jaque. An Introduction to the optical spectroscopy of inorganic solids. (Wiley, 2005).

[67] P. Atkins, J. de Paula. Atkins' Physical chemistry. (Oxford University Press, 2010). [68] J. H. Hubbell, S. M. Seltzer. Tables of X-ray mass attenuation coefficients and

mass energy-absorption coefficients. NISTIR 5632 (version 1.4, 2004).

[69] P. Willmott. An introduction to synchrotron radiation: techniques and applications. (Wiley, 2011).

[70] T. Wakita, T. Taniuchi, K. Ono, M. Suzuki, N. Kawamura, M. Takagaki, H. Miyagawa, F. Guo, T.Nakamura, T. Muro, H. Akinaga, T. Yokoya, M. Oshima,

60

K. Kobayashi. Hard X-ray photoelectron emission microscopy as tool for studying buried layers. J. Appl. Phys. 45, 1886 (2006).

[71] K. Siegbahn. Electron spectroscopy for chemical analysis (E.S.C.A.). Phil. Trans. Roy. Soc. Lond. A 268, 33−57 (1970).

[72] J. J. Olivero, R. L. Longbothum. Empirical fits to the Voigt line width: a brief review. J. Quant. Spectrosc. Radiat. Transfer 17, 233−236 (1977).

[73] J. Tauc. Optical properties and electronic structure of amorphous Ge and Si. Mater. Res. Bull. 3, 37−46 (1968).

[74] T. Y. Ko, A. Jeong, W. Kim, J. Lee, Y. Kim, J. E. Lee, G. H. Ryu, K. Park, D. Kim, Z. Lee, M. H. Lee, C. Lee, S. Ryu. On-stack two-dimensional conversion of MoS2 into MoO3. 2D Mater. 4, 014003 (2017).

[75] N. A. Dhas, A. Ekhtiarzadeh, K. S. Suslick. Sonochemical preparation of supported hydrodesulfurization catalysts. J. Am. Chem. Soc. 123, 8310−8316 (2001).

[76] K. S. Suslick, Y. Didenko, M. M. Fang, T. Hyeon, K. J. Kolbeck, W. B. McNamara III, M. M. Mdleleni, M. Wong. Acoustic cavitation and its chemical consequences. Phil. Trans. R. Soc. Lond. A 357, 335−353 (1999).

[77] I. Vamvasakis, B. Liu, G. S. Armatas. Size effects of platinum nanoparticles in the photocatalytic hydrogen production over 3D mesoporous networks of CdS and Pt nanojunctions. Adv. Func. Mater. 26, 8062−8071 (2016).

[78] U. Holzwarth, N. Gibson. The Scherrer equation versus the ‘Debye-Scherrer equation’. Nature Nanotech. 6, 534 (2011).

[79] X. Chu, G. Yao, A. T. S. Wee, X.-S. Wang. Size-tunable Au nanoparticles on MoS2(0001). Nanotechnology 23, 375603 (2012).

[80] R. T. Tung. The physics and chemistry of the Schottky barrier height. Appl. Phys. Rev. 1, 011304 (2014).

[81] A. Allain, J. Kang, K. Banerjee, A. Kis. Electrical contacts to two-dimensional semiconductors. Nat. Mater. 14, 1195−1205 (2015).

[82] F. Léonard, A. A. Talin. Electrical contacts to one- and two-dimensional nanomaterials. Nat. Nanotech. 6, 773−783 (2011).

[83] D. S. Schulman, A. J. Arnold, S. Das. Contact engineering for 2D materials and devices. Chem. Soc. Rev. 47, 3037−3058 (2018).

[84] W. S. Leong, C. T. Nai, J. T. L. Thong. What does annealing do to metal-graphene contacts. Nano Lett. 14, 3840−3847 (2014).

[85] H. L. Skriver, N. M. Rosengaard. Surface energy and work function of elemental metals. Phys. Rev. B 46, 7157−7168 (1992).

[86] H. B. Michaelson. The work function of the elements and its periodicity. J. Appl. Phys. 48, 4729 (1977).

61

[87] W. M. Haynes. CRC handbook of chemistry and physics. (CRC Press, New York,

2017).

[88] M. Grundmann. The physics of semiconductors: an introduction including devices and nanophysics. (Springer, 2006).

[89] R. T. Tung. Chemical bonding and Fermi level pinning at metal-semiconductor interfaces. Phys. Rev. Lett. 84, 6078−6081 (2000).

[90] J. Kang, W. Liu, D. Sarkar, D. Jena, K. Banerjee. Computational study of metal contacts to monolayer transition-metal dichalcogenide semiconductors. Phys. Rev. X 4, 031005, (2014).

[91] W. E. Spicer, I. Lindau, P. R. Skeath, C. Y. Su. The unified model for Schottky barrier formation and MOS interface states in 3–5 compounds. Appl. Surf. Sci. 9, 83−91 (1981).

[92] D. K. Schroder. Semiconductor material and device characterization. (Wiley, 2006).

[93] S. Das, H.-Y. Chen, A. V. Penumatcha, J. Appenzeller. High performance multilayer MoS2 transistors with scandium contacts. Nano Lett. 13, 100−105 (2013).

[94] F. Giannazzo, G. Fisichella, A. Piazza, S. D. Franco, I. P. Oliveri, S. Agnello, F. Roccaforte. Current injection from metal to MoS2 probed at nanoscale by conductive atomic force microscopy. Mater. Sci. Semicond. Process. 42, 174−178 (2016).

[95] L. Yang, K. Majumdar, H. Liu, Y. Du, H. Wu, M. Hatzistergos, P. Y. Hung, R. Tieckelmann, W. Tsai, C. Hobbs, P. D. Ye. Chloride molecular doping technique on 2D materials: WS2 and MoS2. Nano Lett. 14, 6275−6280 (2014).

[96] W. S. Leong, X. Luo, Y. Li, K. H. Khoo, S. Y. Quek, J. T. L. Thong. Low resistance metal contacts to MoS2 devices with nickel-etched-graphene electrodes. ACS Nano 9, 869−877 (2015).

[97] N. Kaushik, A. Nipane, F. Basheer, S. Dubey, S. Grover, M. M. Deshmukh, S. Lodha. Schottky barrier height for Au and Pd contacts to MoS2. Appl. Phys. Lett. 105, 113505 (2014).

[98] D. Lembke, S. Bertolazzi, A. Kis. Single-layer MoS2 electronics. Acc. Chem. Res. 48, 100−110 (2015).

[99] R. A. Oliver. Advances in AFM for the electrical characterization of semiconductors. Rep. Prog. Phys. 71, 076501 (2008).

[100] Y. Son, Q. H. Wang, J. A. Paulson, C.-J. Shih, A. G. Rajan, K. Tvrdy, S. Kim, B. Alfeeli, R. D. Braatz, M. S. Strano. Layer unmber dependence of MoS2 photoconductivity using photocurrent spectral atomic force microscopic imaging. ACS Nano 9, 2843−2855 (2015).

62

[101] S. Bertolazzi, J. Brivio, A. Radenovic, A. Kis, H. Wilson, L. Prisbrey, E. Minot, A.

Tselev, M. Philips, M. Viani, D. Walters, R. Proksch. Exploring flatland: AFM of mechanical and electrical properties of graphene, MoS2 and other low-dimensional materials. Microsc. Anal. 27, 21−24 (2013).

[102] J. Liu, A. Goswami, K. Jiang, F. Khan, S. Kim, R. McGee, Z. Li, Z. Hu, J. Lee, T. Thundat. Direct-current triboelectricity generation by a sliding Schottky nanocontact on MoS2 multilayers. Nat. Nanotech. 13, 112−116 (2018).

[103] J. Spradlin, S. Doǧan, J. Xie, R. Molnar, A. A. Baski, H. Morkoç. Investigation of forward and reverse current conduction in GaN films by conductive atomic force microscopy. Appl. Phys. Lett. 84, 4150 (2004).

[104] J. Kang, W. Liu, K. Banerjee. High-performance MoS2 transistors with low-resistance molybdenum contacts. Appl. Phys. Lett. 104, 093106 (2014).

[105] A. Fujishima, K. Honda. Electrochemical photolysis of water at a semiconductor electrode. Nature 238, 37−38 (1972).

[106] X. Chen, S. Shen, L. Guo, S. S. Mao. Semiconductor-based photocatalytic hydrogen generation. Chem. Rev. 110, 6503−6570 (2010).

[107] S. Chen, T. Takata, K. Domen. Particulate photocatalysts for overall water splitting. Nat. Rev. Mater. 2, 17050 (2017).

[108] T. Hisatomi, J. Kubota, K. Domen. Recent advances in semiconductors for photocatalytic and photoelectrochemical water splitting. Chem. Soc. Rev. 43, 7520−7535 (2014).

[109] K. Maeda, K. Domen. Photocatalytic water splitting: recent progress and future challenges. J. Phys. Chem. Lett. 1, 2655−2661 (2010).

[110] D. K. Lee, K.-S. Choi. Enhancing long-term photostability of BiVO4 photoanodes for solar water splitting by tuning electrolyte composition. Nat. Energy 3, 53−60 (2018).

[111] K. Tsuji, O. Tomita, M. Higashi, R. Abe. Manganese-substituted polyoxometalate as an effective shuttle redox mediator in Z-scheme water splitting under visible light. ChemSusChem 9, 2201−2208 (2016).

[112] C. Chen, W. Ma, J. Zhao. Semiconductor-mediated photodegradation of pollutants under visible-light irradiation. Chem. Soc. Rev. 39, 4206−4219 (2010).

[113] D. S. Bhatkhande, V. G. Pangarkar, A. A. Beenackers. Photocatalytic degradation for environmental applications – a review. J. Chem. Technol. Biotechnol. 77, 102−116 (2001).

[114] B. Hinnemann, P. G. Moses, J. Bonde, K. P. Jørgensen, J. H. Nielsen, S. Horch, I. Chorkendorff, J. K. Nørskov. Biomimetic hydrogen evolution: MoS2 nanoparticles as catalyst for hydrogen evolution. J. Am. Chem. Soc. 127, 5308−5309 (2005).

63

[115] Y. Liu, J. Wu, K. P. Hackenberg, J. Zhang, Y. M. Wang, Y. Yang, K. Keyshar, J.

Gu, T. Ogitsu, R. Vajtai, J. Lou, P. M. Ajayan, B. C. Wood, B. I. Yakobson. Self-optimizing, highly surface-active layered metal dichalcogenide catalysts for hydrogen evolution. Nat. Energy 2, 17127 (2017).

[116] G. Li, D. Zhang, Q. Qiao, Y. Yu, D. Peterson, A. Zafar, R. Kumar, S. Curtarolo, F. Hunte, S. Shannon, Y. Zhu, W. Yang, L. Cao. All the catalytic active sites of MoS2 for hydrogen evolution. J. Am. Chem. Soc. 138, 16632−16638 (2016).

[117] M. Lannoo. The role of dangling bonds in the properties of surfaces and interfaces of semiconductors. Revue Phys. Appl. 25, 887−894 (1990).

[118] T. F. Jaramillo, K. P. Jørgensen, J. Bonde, J. H. Nielsen, S. Horch, I. Chorkendorff. Identification of active edge sites for electrochemical H2 evolution from MoS2 nanocatalysts. Science 317, 100−102 (2007).

[119] M. V. Bollinger, J. V. Lauritsen, K. W. Jacobsen, J. K. Nørskov, S. Helveg, F. Besenbacher. One-dimensional metallic edge states in MoS2. Phys. Rev. Lett. 87, 196803 (2001).

[120] H. Wang, C. Tsai, D. Kong, K. Chan, F. Abild-Pedersen, J. K. Nørskov, Y. Cui. Transition-metal doped edge sites in vertically aligned MoS2 catalysts for enhanced hydrogen evolution. Nano Res. 8, 566−575 (2015).

[121] J. D. Benck, T. R. Hellstern, J. Kibsgaard, P. Chakthranont, T. F. Jaramillo. Catalyzing the hydrogen evolution reaction (HER) with molybdenum sulfide nanomaterials. ACS Catal. 4, 3957−3971 (2014).

[122] K. Chang, Z. Mei, T. Wang, Q. Kang, S. Ouyang, J. Ye. MoS2/graphene cocatalyst for efficient photocatalytic H2 evolution under visible light irradiation. ACS Nano 8, 7078−7087 (2014).

[123] Q. Lu, Y. Yu, Q. Ma, B. Chen, H. Zhang. 2D transition-metal-dichalcogenide-nanosheet-based composites for photocatalytic and electrocatalytic hydrogen evolution reactions. Adv. Mater. 28, 1917−1933 (2016).

[124] J. G. Kushmerick, P. S. Weiss. Mobile promoters on anisotropic catalysts: nickel on MoS2. J. Phys. Chem. B 102, 10094−10097 (1998).

[125] D. Voiry, J. Yang, M. Chhowalla. Recent strategies for improving the catalytic activity of 2D TMD nanosheets toward the hydrogen evolution reaction. Adv. Mater. 28, 6197−6206 (2016).

[126] P. Sellers, C. A. Kelly, J. W. M. Rudd, A. R. MacHutchon. Photodegradation of methylmercury in lakes. Nature 380, 694−697 (1996).

[127] N. Berntsen, T. Gutjahr, L. Loeffler, J. R. Gomm, R. Seshadri, W. Tremel. A solvothermal route to high-surface-area nanostructured MoS2. Chem. Mater. 15, 4498−4502 (2003).

64

[128] T. Kroll, R. Kraus, R. Schönfelder, V. Y. Aristov, O. V. Molodtsova, P. Hoffmann,

M. Knupfer. Transition metal phthalocyanines: insight into the electronic structure from soft x-ray spectroscopy. J. Chem. Phys. 137, 054306 (2012).

[129] K. Maeda, K. Teramura, D. Lu, T. Takata, N. Saito, Y. Inoue, K. Domen. Photocatalyst releasing hydrogen from water. Nature 440, 295 (2006).

[130] J. Gao, B. Li, J. Tan, P. Chow, T.-M. Lu, N. Koratkar. Aging of transition metal dichalcogenide monolayers. ACS Nano 10, 2628−2635 (2016).

electronic and photocatalytic properties of transition...

Documents