Synthesis of molybdenum-based two-dimensional materials

with liquid metal

Author:
Wang, Yifang
Publication Date:
2023
DOI:
https://doi.org/10.26190/unsworks/24651
License:
https://creativecommons.org/licenses/by/4.0/
Link to license to see what you are allowed to do with this resource.

Downloaded from http://hdl.handle.net/1959.4/100944 in https://

unsworks.unsw.edu.au on 2024-03-15
 Synthesis of molybdenum-based two-
dimensional materials with liquid metal

A thesis submitted to The University of New South Wales in partial fulfilment of the

degree of Doctor of Philosophy

By

Yifang Wang

School of Chemical Engineering

The University of New South Wales

September 2022
[This page was left intentionally blank]
 Thesis/Dissertation Sheet

Surname/Family Name : Wang

Given Name/s : Yifang
Abbreviation for degree as give in the University calendar : PhD
Faculty : Faculty of Engineering
School : School of Chemical Engineering
Synthesis of molybdenum-based two-
Thesis Title :
dimensional materials with liquid metal

Abstract
The enigmatic surface of gallium-based liquid metals that is ultra-active and smooth, offering opportunities for
synthesising and templating two-dimensional (2D) films. When the reactive surface is in contact with an appropriate
aqueous solution, the built-in potential at the aqueous solution-liquid metal interface can induce an interfacial reaction
and lead to the deposition of desired 2D materials onto the liquid metal surface. In this thesis, the author is involved in
research on molybdenum-based 2D materials synthesis at an aqueous solution-liquid metal interface.

In the first stage of this Ph.D. research, the author develops a technique of 2D molybdenum sulfide synthesis at the
surface of liquid metals. Utilising an aqueous solution of ammonium tetrathiomolybdate as a precursor, 2D
molybdenum sulfide layers of large area are deposited on the surface of liquid metals, which are transferrable onto
silicon wafers with a touch-transfer technique. Upon annealing, the resultant 2D molybdenum sulfide is of unit-cell
thickness and highly crystalline.

In the second stage, this Ph.D. research extends this technique to the deposition of molybdenum oxide patterns.
Uniform layer of molybdenum oxides, which are hydrated and amorphous, are templated onto liquid metal droplets
and later transferred. The as-synthesised molybdenum oxide layers can be selectively dehydrated and crystalised via
laser exposure, transforming them into conductive molybdenum dioxide patterns. The resultant conductive patterns
show optical and electronic responses to bio-stimuli, which can be used for bio-sensing.

In the third stage of this thesis, the liquid metal-aqueous solution interfacial reactions are applied for surface decoration
and band structure modulation of liquid metal-based particles. Utilising the built-in potential at the interfaces, a
secondary layer of molybdenum-based semiconductor is deposited on the surface of liquid metal particles, leading to
core-shell structures with different optical properties and band structures.

Collectively, the author demonstrates a new method for synthesising 2D materials utilising aqueous solution-liquid
metal interfacial reactions, which potentially establishes a new route for the room temperature deposition of 2D
materials at large lateral dimensions. The outcomes of this Ph.D. research offer significant possibilities for future
industrial uptakes.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents a non-exclusive licence to archive and to make
available (including to members of the public) my thesis or dissertation in whole or in part in the University libraries in
all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in
my thesis or dissertation, such as copyright and patent rights, subject to applicable law. I also retain the right to use all
or part of my thesis or dissertation in future works (such as articles or books).

19/09/2022

……………………………………………………… …… ……………………...…….…

Signature Date

The University recognizes that there may be exceptional circumstances requiring restrictions on copying or conditions
on use. Requests for restriction for a period of up to 2 years can be made when submitting the final copies of your
thesis to the UNSW Library. Requests for a longer period of restriction may be considered in exceptional circumstances
and require the approval of the Dean of Graduate Research.
[This page was left intentionally blank]
STATEMENT
[This page was left intentionally blank]
INCLUSION OF PUBLICATIONS STATEMENT
[This page was left intentionally blank]
Abstract

The enigmatic surface of gallium-based liquid metals that is ultra-active and smooth,

offering opportunities for synthesising and templating two-dimensional (2D) films. When

the reactive surface is in contact with an appropriate aqueous solution, the built-in

potential at the aqueous solution-liquid metal interface can induce an interfacial reaction

and lead to the deposition of desired 2D materials onto the liquid metal surface. In this

thesis, the author is involved in research on molybdenum-based 2D materials synthesis at

an aqueous solution-liquid metal interface.

In the first stage of this Ph.D. research, the author develops a technique of 2D

molybdenum sulfide synthesis at the surface of liquid metals. Utilising an aqueous

solution of ammonium tetrathiomolybdate as a precursor, 2D molybdenum sulfide layers

of large area are deposited on the surface of liquid metals, which are transferrable onto

silicon wafers with a touch-transfer technique. Upon annealing, the resultant 2D

molybdenum sulfide is of unit-cell thickness and highly crystalline.

In the second stage, this Ph.D. research extends this technique to the deposition of

molybdenum oxide patterns. Uniform layer of molybdenum oxides, which are hydrated

and amorphous, are templated onto liquid metal droplets and later transferred. The as-

synthesised molybdenum oxide layers can be selectively dehydrated and crystalised via

laser exposure, transforming them into conductive molybdenum dioxide patterns. The

resultant conductive patterns show optical and electronic responses to bio-stimuli, which

can be used for bio-sensing.

In the third stage of this thesis, the liquid metal-aqueous solution interfacial reactions are

applied for surface decoration and band structure modulation of liquid metal-based

particles. Utilising the built-in potential at the interfaces, a secondary layer of

i
molybdenum-based semiconductor is deposited on the surface of liquid metal particles,

leading to core-shell structures with different optical properties and band structures.

Collectively, the author demonstrates a new method for synthesising 2D materials

utilising aqueous solution-liquid metal interfacial reactions, which potentially establishes

a new route for the room temperature deposition of 2D materials at large lateral

dimensions. The outcomes of this Ph.D. research offer significant possibilities for future

industrial uptakes.

ii
Acknowledgement

The journey of getting a Ph.D. is so far the most incredible experience in my life. During

this journey, I received tremendous help and support. I acknowledge the Australian

Research Council (ARC) Laureate Fellowship grant (FL180100053), and the ARC Center

of Excellence Future Low-Energy Electronics Technologies (FLEET) (CE170100039)

for financially supporting the research projects included in this thesis. I acknowledge the

Australian Government Research Training Program (RTP) Scholarship and the Women

in FLEET, Ph.D. Top-up Scholarships for financially support my Ph.D.

First and most importantly, I would like to convey a sincere gratitude to my supervisor

Prof. Kourosh Kalantar-Zadeh for offering me this great opportunity and providing

enormous support and supervision. I’m not the brightest student, but he has always been

patient and supportive. He is an exceedingly knowledgeable researcher, and I learnt a lot

from him during the past three and a half years.

I would like to express heartfelt thanks to secondary supervisors, Dr. Mohammad Bagher

Ghasemian, Dr. Mohannad Mayyas and Dr. Jiong Yang for their supervision. Mohammad

helped a lot with my review paper and my last project. Without his help, I would not be

able to finish the last year of my Ph.D. I would like to thank my collaborators from other

institutes for their input of my research projects. Prof. Richard B. Kaner from the

University of California, Los Angles (UCLA), Dr. Yin Yao from Electron Microscope

Unit, UNSW, Dr. David Cortie from the University of Wollongong, and Dr. Aaron

Elbourne from RMIT University.

I would also like to thank my colleagues in the Centre for Advanced Solid and Liquid-

based Electronics and Optics (CASLEO) for their company and encouragement. We

iii
shared joy and laughter, as well as frustration when experiments failed. It was amazing

to work with you guys. This journey wouldn’t be the same without you.

I would like to thank my family, my father Mr. Jinghui Wang, my stepmother Ms. Li

Zeng, my little sister Ms. Lu Feng and my partner Dr. Aidong Tan. Although being

thousands of miles away, I can always feel the love and support you have for me. I want

to thank my late mother, Ms. Yafan Zhang, and my late grandfather Mr. Zhenxue Wang

for always having faith in me and being proud of me. I will carry you in my heart when I

attend my graduation ceremony.

I would also like to thank the Counselling and Psychological Services of UNSW and my

counsellor Dr. Juan Chen. Juan gave me enormous support and compassion that I

desperately need in those difficult times during this Ph.D., for which I’m truly grateful.

Juan, you are a real-life superhero because you are literally saving lives.

iv
List of publications

Publications contributing to this thesis

[1] Y. Wang, M. Mayyas, J. Yang, J. B. Tang, M. B. Ghasemian, J. L. Han, A, Elbourne,

T, Daeneke, R, B. Kaner, and K, Kalantar-Zadeh, Self-Deposition of 2D Molybdenum

Sulfides on Liquid Metals. Advanced Functional Materials, 31, 2005866 (2020)

[2] Y. Wang, M. Mayyas, J. Yang, M. B. Ghasemian, J. B. Tang, M. Mousavi, J. Han, M.

Ahmed, M. Baharfar, G. Mao, Y. Yao, D. Esrafilzadeh, D. Cortie, and K. Kalantar-Zadeh,

Liquid-Metal-Assisted Deposition and Patterning of Molybdenum Dioxide at Low

Temperature. ACS Applied Materials & Interfaces 13, 53181-53193 (2021)

[3] Y. Wang, M Baharfar, J. Yang, M. Mayyas, M. B. Ghasemian, K. Kalantar-Zadeh,

Liquid state of post transition metals for interfacial synthesis of two-dimensional

materials. Applied Physics Reviews 9, 021306 (2022)

[4] M. B. Ghasemian,† Y. Wang,† F.-M. Allioux, A. Zavabeti and K Kalantar-Zadeh,

Band structure modulation of gallium-based liquid metal particles via coating with

molybdenum oxide and oxysulfide (submitted)

Other publications by the author of this Ph.D. thesis with the theme of liquid metals

[1] H. Li, R. Abbasi, Y. Wang, F.-M. Allioux, P. Koshy, S. A Idrus-Saidi, M. A. Rahim, J.

Yang, M. Mousavi, J. Tang, M. B Ghasemian, R. Jalili, K. Kalantar-Zadeh, M. Mayyas,

Liquid metal-supported synthesis of cupric oxide, Journal of Material Chemistry C 8,

1656-1665 (2020)

[2] M. Mayyas, H. Li, P. Kumar, M. B. Ghasemian, J. Yang, Y. Wang, D. J. Lawes, J.

v
Han, M. G. Saborio, J. Tang, R. Jalili, S. H. Lee, W. K. Seong, S. P. Russo, D. Esrafilzadeh,

T. Daeneke, R. B. Kaner, R. S. Ruoff, K. Kalantar‐Zadeh, Liquid-Metal-Templated

Synthesis of 2D Graphitic Materials at Room Temperature, Advanced Materials 32,

2001997 (2020)

[3] M. Mayyas, M. Mousavi, M. B. Ghasemian, R. Abbasi, H. Li, M. J. Christoe, J. Han,

Y. Wang, C. Zhang, M. A. Rahim, J. Tang, J. Yang, D. Esrafilzadeh, R. Jalili, F.-M.

Allioux, A. P. O’Mullane, K. Kalantar-Zadeh, Pulsing Liquid Alloys for Nanomaterials

Synthesis, ACS Nano 14, 14070-14079 (2020)

[4] M. B Ghasemian, A. Zavabeti, R. Abbasi, P. V. Kumar, N. Syed, Y. Yao, J. Tang, Y.

Wang, A. Elbourne, J. Han, M. Mousavi, T. Daeneke, K Kalantar-Zadeh, Ultra-thin lead

oxide piezoelectric layers for reduced environmental contamination using a liquid metal-

based process, Journal of Material Chemistry A 8, 19434-19443 (2020)

[5] W. Xie, F.-M. Allioux, R, Namivandi-Zangeneh, M. B. Ghasemian, J. Han, M. A.

Rahim, J. Tang, J. Yang, M. Mousavi, M. Mayyas, Z. Cao, F. Centurion, M. J. Christoe,

C. Zhang, Y. Wang, S. Merhebi, M. Baharfar, G. Ng, D. Esrafilzadeh, C. Boyer, K.

Kalantar-Zade, Polydopamine Shell as a Ga3+ Reservoir for Triggering Gallium–Indium

Phase Separation in Eutectic Gallium–Indium Nanoalloys. ACS Nano 15, 16839–16850

(2021).

[6] M. Mousavi, M. B. Ghasemian, J. Han, Y. Wang, R. Abbasi, Y. Jiong, J. Tang, S. A.

Idrus-Saidi, X, Guan, M. J. Christoe, S. Merhebi, C. Zhang, J. Tang, R. Jalili, T. Daeneke,

T. Wu, K. Kalantar-Zadeh, Bismuth telluride topological insulator synthesized using

liquid metal alloys: Test of NO2 selective sensing. Applied Materials Today, 22, 100954

(2021).

[7] M. Mayyas, K. Khoshmanesh, P. Kumar, M. Mousavi, J. Tang, M. B. Ghasemian, J.

vi
Yang, Y. Wang, M. Baharfar, M. A. Rahim, W. Xie, F.-M. Allioux, R. Daiyan, R. Jalili,

D. Esrafilzadeh, K. Kalantar-Zadeh, Gallium-Based Liquid Metal Reaction Media for

Interfacial Precipitation of Bismuth Nanomaterials with Controlled Phases and

Morphologies Advanced Functional Materials, 32, 2108673 (2021)

[8] J. Han, M. Mayyas, J. Tang, M. Mousavi, S. A.Idrus-Saidi, S. Cai, Z. Cao, Y. Wang,

J. Tang, R. Jalili, A. P.O'Mullane, R. B. Kaner, K. Khoshmanesh, K. Kalantar-Zadeh.

Liquid metal enabled continuous flow reactor: a proof-of-concept, Matter, 5, 379-381

(2021)

vii
 Table of Contents

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

Acknowledgement ..........................................................................................................iii

List of publications .......................................................................................................... v

Chapter 1 Introduction ................................................................................................... 1

1.1 Introduction ............................................................................................................. 2

1.2 Reference ................................................................................................................. 7

Chapter 2 Literature review ........................................................................................ 12

2.1 Introduction ........................................................................................................... 13

2.2 Properties of post transition liquid metals ............................................................. 16

2.2.1 Origin of low melting points........................................................................... 16

2.2.2 Electrical conductivity .................................................................................... 20

2.2.3 Thermal conductivity ...................................................................................... 22

2.2.4 Surface-induced atomic layering .................................................................... 24

2.2.5 Surface oxidation of liquid metals .................................................................. 28

2.3 Interfacial synthesis of 2D materials with liquid metal ......................................... 32

2.3.1 2D materials synthesis via surface oxidation of liquid metals ....................... 32

2.3.2 2D materials synthesis with liquid metals as reducing agents ........................ 39

2.3.3 Deposition of 2D layers based on liquid metal processes .............................. 45

2.4 Application in devices ........................................................................................... 49

2.4.1 Optoelectronic devices .................................................................................... 50

2.4.2 Dielectric/ conducting units and bandgap manipulation ................................ 51

2.4.3 Piezoelectric devices ....................................................................................... 54

viii
 2.4.4 Sensors ............................................................................................................ 57

2.5 Conclusion ............................................................................................................. 58

2.6 Reference ............................................................................................................... 58

Chapter 3 Self-Deposition of 2D Molybdenum Sulfides on Liquid Metals ........... 74

3.1 Introduction ........................................................................................................... 75

3.2 Methods ................................................................................................................. 75

3.2.1 Materials ......................................................................................................... 75

3.2.2 Preparation of materials .................................................................................. 76

3.2.3 Self-deposition of molybdenum sulfides (MoSx) on EGaIn ........................... 76

3.2.4 Extraction and transfer of MoSx ..................................................................... 77

3.2.5 Annealing process ........................................................................................... 78

3.2.6 Electrochemical analysis ................................................................................ 79

3.2.7 Electrochemical deposition of MoSx .............................................................. 80

3.2.8 Morphological characterization ...................................................................... 80

3.2.9 Optical characterization .................................................................................. 81

3.3 Results and discussions ......................................................................................... 82

3.3.1 Self-deposition of ultrathin MoSx on liquid metal .......................................... 82

3.3.2 Interfacial assessment of liquid metal in aqueous precursor .......................... 84

3.3.3 Structure and morphology of 2D MoS2 sheet ................................................. 90

3.3.4 Thickness dependent optical properties of 2D MoS2 sheet ............................ 96

3.5 Conclusion ............................................................................................................. 99

3.6 References ........................................................................................................... 100

Chapter 4 Liquid-Metal-Assisted Deposition and Patterning of Molybdenum

Dioxide at Low Temperature ..................................................................................... 104

ix
 4.1 Introduction ......................................................................................................... 105

4.2 Methods ............................................................................................................... 105

4.2.1 Materials ....................................................................................................... 105

4.2.2 Synthesis of H2MoO3 Layer on EGaIn Surface ............................................ 106

4.2.3 Touch-transfer of H2MoO3 Layer onto Desired Substrates. ......................... 107

4.2.4 Thermal Annealing of H2MoO3 .................................................................... 108

4.2.5 Laser Annealing Process............................................................................... 109

4.2.6 Electrochemical Characterization. ................................................................ 109

4.2.7 Morphological and Structural Characterizations. ......................................... 110

4.2.8 Optical characterization ................................................................................ 110

4.2.9 Conductivity characterization ....................................................................... 111

4.2.10 Computational studies ................................................................................ 112

4.3 Results and discussion ......................................................................................... 113

4.3.1 Thickness-controlled deposition of H2MoO3 sheet on liquid metal ............. 113

4.3.2 Thermal annealing of H2MoO3: dehydration and crystallization ................. 120

4.3.3 Laser annealing: A direct laser writing of conductive MoO2 patterns ......... 123

4.3.4 Morphology properties of the MoO2 patterns ............................................... 130

4.3.5 Theoretical calculation: Impacts of oxygen vacancies and grain-boundaries on

band structure of 2D-MoO3 ................................................................................... 132

4.3.6 Peak-force tunnelling AFM measurement for conductivity assessment ...... 136

4.5 Conclusions ......................................................................................................... 141

4.6 References ........................................................................................................... 142

Chapter 5 Band structure modulation of gallium-based liquid metal particles via

coating with molybdenum oxide and oxysulfide ...................................................... 145

5.1 Introduction ......................................................................................................... 146

x
 5.2 Methods ............................................................................................................... 147

5.2.1 Materials ....................................................................................................... 147

5.2.2 Synthesis of EGaIn particles ......................................................................... 147

5.2.3 Synthesis of molybdenum sulfide decorated EGaIn particles (EGaIn@MoO xSy)

............................................................................................................................... 147

5.2.4 Synthesis of molybdenum oxide decorated EGaIn particles (EGaIn@MoO x)

............................................................................................................................... 148

5.2.5 Morphological and Structural Characterizations. ......................................... 148

5.2.6 Optical characterization ................................................................................ 149

5.3 Results and discussion ......................................................................................... 150

5.3.1 Decoration of EGaIn particles ...................................................................... 150

5.3.2 Morphology and chemical composition of EGaIn@MoO xSy, and

EGaIn@MoOx ....................................................................................................... 150

5.3.3 Band structures of EGaIn@MoOxSy, and EGaIn@MoOx ............................ 157

5.4 Conclusions ......................................................................................................... 159

5.5 References ........................................................................................................... 160

Chapter 6 Conclusions ................................................................................................ 163

6.1 Summary of the work .......................................................................................... 164

6.2 Publications ......................................................................................................... 166

6.2.1 Publications contributing to this thesis ......................................................... 166

6.2.3 Other co-authored publications with the theme of liquid metals .................. 167

6.3 Future research directions ................................................................................... 168

xi
Chapter 1 Introduction

1
1.1 Introduction

Several post-transition metals and their alloys offer low melting points at near room

temperature. The most well-known examples are EGaIn (eutectic alloy of gallium and

indium) and Galinstan (eutectic alloys of gallium, indium, and tin) with melting points at

15.4 and 13.2 °C, respectively.1 Such alloys are commonly referred to as liquid metals

since they take on a liquid state at room temperature. These liquids are fascinating alloys

as they offer fluidity, freely moving electrons and ions, as well as metallic thermal and

electrical conductivity.2-4 Additionally, gallium (Ga) is not as hazardous as its sister liquid

metal mercury.5 Other than these unique properties, liquid metals also extend intrinsic

characteristics that render them unique as reaction media for synthesising two-

dimensional (2D) materials.6-12 One of the key properties is that the surface of liquid metal

is atomically smooth which can naturally template 2D materials. Another advantage of

liquid metals is their non-polarized nature, which weakens the attractive forces between

the bulk metal and the interfacial 2D layers. Consequently, the synthesised 2D material

can be readily harvested with a mechanical delamination process.6-12

It was previously demonstrated that liquid metals can be used for synthesising self-

limiting oxide or sulfide metal compounds on their surface directly from the mother melt

by exposing them to oxic or sulfuric gaseous environments. Another possibility involves

adding a secondary metallic element into the liquid metal to make an alloy. If the

secondary metal wins the competition against the core metal, it emerges on the surface

and its oxide or sulfide can be harvested.13 However, this gaseous-surrounding-based

process only works if the secondary metal is soluble in the host liquid metal at low

temperatures. This means that metallic elements such as tungsten (W) and molybdenum

(Mo), which are not soluble in Ga, cannot be obtained using this method. 1

2
The author of this thesis hypothesizes that in this case, by bringing a precursor of these

metals to the surface of liquid metal, a designed reaction can take place to form the desired

2D compound. By the careful selection of liquid metals and precursors in their

surroundings, unique reactive interfacial environments can be created. These precursors

can be dissolved in a solvent to be brought onto this interface. The interfacial region

between the liquid metal and solvent can be destabilized to activate the precursors and

produce the target material, while the atomically smooth liquid metal surface serves as a

template.

Molybdenum disulfide (MoS2) is a layered-structured semiconductor with strong in-plane

bonding and weak inter-plane interaction.14, 15 When thinned down to nanometre or even

unit-cell thickness via mechanical exfoliating, these 2D MoS2 layers show intriguing

properties that are distinctively different from the bulk, which makes them valuable for

both fundamental research and practical applications.16, 17 Various methods for 2D MoS2

synthesis were developed including mechanical exfoliation, and various chemical or

physical deposition methods.18 Mechanical exfoliation can produce single crystal 2D

MoS2 at room temperature but with limited lateral dimensions (up to tens of micrometres)

and poor thickness-control.19 On the contrary, various deposition methods can also be

used for the deposition of 2D MoS2 while they still suffer from high-process temperatures

or poor morphology control.20 These deposition methods are also generally substrate

dependent. Valid approaches for synthesis of 2D MoS2 of large lateral dimensions on

arbitrary substrates are yet to be developed.

Another viable candidate for the interfacial gallium based liquid metal deposition process

is molybdenum oxide. Molybdenum oxide has been widely used in optics, electronics and

energy systems.21-23 Molybdenum oxide (MoOx) emerges in many different

stoichiometries. At one end of the stoichiometry range, it is a wide bandgap

3
semiconductor as fully oxidized MoO3 (x = 3), while at the other end as MoO2 (x = 2),

which offers metal-like conductivity.24, 25 Previous theoretical studies suggested that the

formation of the non-polar Mo−Mo metallic bonds in MoO 2 leads to a high conductivity

in this material.26 Therefore, MoO2 has been commonly used as a charge transport

medium in electronics,40 electrocatalysis26, 27 and energy storage devices.28, 29 Similar to

2D MoS2 mentioned above, 2D molybdenum oxides also call for more effective synthesis

methods for large area deposition for electronic and other applications.

In addition to synthesis of 2D materials in large area dimensions, the liquid metal-aqueous

solution interfacial reaction could also be used for surface decoration and band structure

modulation of liquid metal particles. When liquid metals are broken down into micro- or

nano-sized droplets, the significantly increased surface to volume ratio enable them for

various applications such as catalysis and sensing.30 Furthermore, decoration and

functionalization of the surface of liquid metal droplets could modulate their properties.

Taking advantage of the reactive surface of liquid metals, aqueous solution-liquid metal

interfacial deposition is an effective approach for such surface decorations. With careful

selection of the aqueous solution as the precursor, a layer of desired secondary material

can be deposited onto the surface of liquid metal droplets using a well-engineered

interfacial reaction. Upon surface deposition of materials with different bandgaps, the

band structure and optical properties of the liquid metal droplets can then be modulated.

Given the aforementioned research gaps, the author of this Ph.D. thesis endeavours to

achieve room-temperature deposition of molybdenum sulfides and oxides using liquid

metal-aqueous solution interfacial reactions, which is the core subject of this Ph.D.

research. The details of the research are presented in the chapters of this Ph.D. thesis as

follows.

4
Chapter 2 presents a comprehensive overview about liquid state of post-transition metals

for interfacial synthesis of 2D materials. The author discusses the key physical properties

of the liquid metals including the origin of their low melting points, and high thermal and

electrical conductivity. The author also illustrates their boundary induced layering and

oxidation, as essential traits for creating 2D films. Afterward, the interfacial synthesis of

2D materials is depicted with the discussing surface oxidation, reduction and exfoliation.

Different types of devices using liquid metal-induced 2D synthesis processes including

field effect transistors, optoelectronic devices, systems that use 2D dielectric and

conductive layers and piezoelectric devices are also presented.

The content of this chapter was published as review paper: Y. Wang, M Baharfar, J. Yang,

M. Mayyas, M. B. Ghasemian, K. Kalantar-Zadeh, Liquid state of post-transition metals

for interfacial synthesis of two-dimensional materials, Applied Physics Reviews 9,

021306 (2022).

Chapter 3 describes a self-deposition process of atomically thin molybdenum sulfide

sheets on the surface of liquid metals. In this process, EGaIn is applied as the reducing

agent and aqueous solution of ammonium tetrathiomolybdate [(NH 4)2MoS4] as the

precursor. It is verified by electrochemical assessment that the built-in potential at the

liquid metal-aqueous solution interface is sufficient to destabilize the precursor and leads

to interfacial deposition. A touch-transferring technic is developed which can transfer the

deposited layer onto any substrate. After a final annealing step, the obtained 2H MoS 2

layers are proven to be of unit-cell thickness, highly crystallinity, and thickness-

dependent optical properties.

The content of this chapter was published as a peer-reviewed paper: Y. Wang, M. Mayyas,

J. Yang, J. B. Tang, M. B. Ghasemian, J. L. Han, A, Elbourne, T, Daeneke, R, B. Kaner,

5
and K, Kalantar-Zadeh, Self-Deposition of 2D Molybdenum Sulfides on Liquid Metals.

Advanced Functional Materials, 31, 2005866 (2020).

In Chapter 4, the author designed a liquid metal-assisted deposition process of

molybdenum oxide sheet which can serve as a writing medium of conductive patterns.

Utilising the surface of liquid metal as an atomically smooth template and sodium

molybdate solution as precursor, a uniform sheet of hydrated molybdenum oxide sheet is

deposited on the surface of liquid metal. After transferring onto substrates, the hydrated

molybdenum oxide can be selectively transformed into crystalline and conductive

molybdenum dioxide via laser exposure. In this way, the transferred sheet can serve as a

writing medium that enables direct laser writing of conductive patterns. In addition, the

molybdenum dioxide patterns show optical and electrical response upon bio-stimuli,

which demonstrates their potentials is bio-sensing applications.

The content of this chapter was published as a peer-reviewed scientific paper as Y. Wang,

M. Mayyas, J. Yang, M. B. Ghasemian, J. B. Tang, M. Mousavi, J. Han, M. Ahmed, M.

Baharfar, G. Mao, Y. Yao, D. Esrafilzadeh, D. Cortie, and K. Kalantar-Zadeh, Liquid-

Metal-Assisted Deposition and Patterning of Molybdenum Dioxide at Low Temperature.

ACS Applied Materials & Interfaces 13, 53181-53193 (2021).

In Chapter 5, band structure modulation of gallium-based liquid metal particles is

demonstrated via coating with molybdenum oxide and sulfide. Utilising the surface of

EGaIn particles as a reactive medium, deposition of molybdenum oxide and sulfide shell

are induced, which leads to a core-shell structured microparticles with tunable band

structure and optical properties. The content of this chapter was included in a paper as M.

B. Ghasemian,† Y. Wang,† F.-M. Allioux, A. Zavabeti and K Kalantar-Zadeh, Band

structure modulation of gallium-based liquid metal particles via coating with

6
molybdenum oxide and oxysulfide. Submitted to Nanoscale, (2022).

Finally, Chapter 6 of this thesis concludes the outcomes achieved in this Ph.D. research

and provides recommendations for future research directions in this field.

1.2 Reference

1. Massalski, T. B.; Okamoto, H.; Subramanian, P.; Kacprzak, L.; Scott, W. W.,

Binary alloy phase diagrams. ASM International, 1986; Vol. 1.

2. Allioux, F.-M.; Ghasemian, M. B.; Xie, W.; O'Mullane, A. P.; Daeneke, T.;

Dickey, M. D.; Kalantar-Zadeh, K., Applications of liquid metals in

nanotechnology. Nanoscale Horiz. 2022, 7, 141-164.

3. Sun, X.; Yuan, B.; Sheng, L.; Rao, W.; Liu, J., Liquid metal enabled injectable

biomedical technologies and applications. Appl. Mater. Today 2020, 20, 100722.

4. Daeneke, T.; Khoshmanesh, K.; Mahmood, N.; De Castro, I. A.; Esrafilzadeh,

D.; Barrow, S.; Dickey, M.; Kalantar-Zadeh, K., Liquid metals: fundamentals

and applications in chemistry. Chem. Soc. Rev. 2018, 47 (11), 4073-4111.

5. Ivanoff, C. S.; Ivanoff, A. E.; Hottel, T. L., Gallium poisoning: A rare case report.

Food Chem. Toxicol 2012, 50 (2), 212-215.

6. Li, C.; Iqbal, M.; Lin, J.; Luo, X.; Jiang, B.; Malgras, V.; Wu, K. C.-W.; Kim,

J.; Yamauchi, Y., Electrochemical deposition: an advanced approach for

templated synthesis of nanoporous metal architectures. Acc. Chem. Res. 2018, 51

(8), 1764-1773.

7. Li, C.; Tan, H.; Lin, J.; Luo, X.; Wang, S.; You, J.; Kang, Y.-M.; Bando, Y.;

Yamauchi, Y.; Kim, J., Emerging Pt-based electrocatalysts with highly open

7
 nanoarchitectures for boosting oxygen reduction reaction. Nano Today 2018, 21,

91-105.

8. Jiang, B.; Guo, Y.; Kim, J.; Whitten, A. E.; Wood, K.; Kani, K.; Rowan, A.

E.; Henzie, J.; Yamauchi, Y., Mesoporous metallic iridium nanosheets. J. Am.

Chem. Soc. 2018, 140 (39), 12434-12441.

9. Li, C.; Iqbal, M.; Jiang, B.; Wang, Z.; Kim, J.; Nanjundan, A. K.; Whitten, A.

E.; Wood, K.; Yamauchi, Y., Pore-tuning to boost the electrocatalytic activity of

polymeric micelle-templated mesoporous Pd nanoparticles. Chem. Sci. 2019, 10

(14), 4054-4061.

10. Carey, B. J.; Ou, J. Z.; Clark, R. M.; Berean, K. J.; Zavabeti, A.; Chesman, A.

S.; Russo, S. P.; Lau, D. W.; Xu, Z.-Q.; Bao, Q., Wafer-scale two-dimensional

semiconductors from printed oxide skin of liquid metals. Nat. Comm. 2017, 8 (1),

14482.

11. Syed, N.; Zavabeti, A.; Messalea, K. A.; Della Gaspera, E.; Elbourne, A.;

Jannat, A.; Mohiuddin, M.; Zhang, B. Y.; Zheng, G.; Wang, L., Wafer-sized

ultrathin gallium and indium nitride nanosheets through the ammonolysis of liquid

metal derived oxides. J. Am. Chem. Soc. 2018, 141 (1), 104-108.

12. Wang, Y.; Baharfar, M.; Yang, J.; Mayyas, M.; Ghasemian, M. B.; Kalantar-

Zadeh, K., Liquid state of post-transition metals for interfacial synthesis of two-

dimensional materials. Appl. Phys. Rev. 2022, 9 (2), 021306.

13. Zavabeti, A.; Ou, J. Z.; Carey, B. J.; Syed, N.; Orrell-Trigg, R.; Mayes, E. L.

H.; Xu, C. L.; Kavehei, O.; O'Mullane, A. P.; Kaner, R. B.; Kalantar-Zadeh,

K.; Daeneke, T., A liquid metal reaction environment for the room-temperature

synthesis of atomically thin metal oxides. Science 2017, 358 (6361), 332-335.

8
14. Ghatak, S.; Pal, A. N.; Ghosh, A., Nature of electronic states in atomically thin

MoS2 field-effect transistors. ACS Nano 2011, 5 (10), 7707-7712.

15. Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A., Single-

layer MoS2 transistors. Nat. Nanotechnol. 2011, 6 (3), 147-150.

16. Ganatra, R.; Zhang, Q., Few-layer MoS2: a promising layered semiconductor.

ACS Nano 2014, 8 (5), 4074-4099.

17. Wang, Z.; Mi, B., Environmental applications of 2D molybdenum disulfide

(MoS2) nanosheets. Environ. Sci. Technol. 2017, 51 (15), 8229-8244.

18. Sun, J.; Li, X.; Guo, W.; Zhao, M.; Fan, X.; Dong, Y.; Xu, C.; Deng, J.; Fu,

Y., Synthesis methods of two-dimensional MoS2: a brief review. Crystals 2017, 7

(7), 198.

19. Li, H.; Wu, J.; Yin, Z.; Zhang, H., Preparation and applications of mechanically

exfoliated single-layer and multilayer MoS2 and WSe2 nanosheets. Acc. Chem.

Res. 2014, 47 (4), 1067-1075.

20. Liu, H.; Wong, S. L.; Chi, D., CVD growth of MoS2‐based two‐dimensional

materials. Chem. Vap. Depos. 2015, 21 (10-11-12), 241-259.

21. Hou, S.; Zhang, G.; Zeng, W.; Zhu, J.; Gong, F.; Li, F.; Duan, H., Hierarchical

core-shell structure of ZnO nanorod@ NiO/MoO2 composite nanosheet arrays for

high-performance supercapacitors. ACS Appl. Mater. Interfaces 2014, 6 (16),

13564-13570.

22. Petnikota, S.; Teo, K. W.; Chen, L.; Sim, A.; Marka, S. K.; Reddy, M. V.;

Srikanth, V.; Adams, S.; Chowdari, B., Exfoliated graphene oxide/MoO 2

composites as anode materials in lithium-ion batteries: an insight into intercalation

of Li and conversion mechanism of MoO2. ACS Appl. Mater. Interfaces 2016, 8

(17), 10884-10896.

9
23. Alsaif, M. M.; Field, M. R.; Murdoch, B. J.; Daeneke, T.; Latham, K.; Chrimes,

A. F.; Zoolfakar, A. S.; Russo, S. P.; Ou, J. Z.; Kalantar-Zadeh, K.,

Substoichiometric two-dimensional molybdenum oxide flakes: a plasmonic gas

sensing platform. Nanoscale 2014, 6 (21), 12780-12791.

24. De Castro, I. A.; Datta, R. S.; Ou, J. Z.; Castellanos‐Gomez, A.; Sriram, S.;

Daeneke, T.; Kalantar‐zadeh, K., Molybdenum oxides-from fundamentals to

functionality. Adv. Mater. 2017, 29 (40), 1701619.

25. Xing, K.; Xiang, Y.; Jiang, M.; Creedon, D. L.; Akhgar, G.; Yianni, S. A.;

Xiao, H.; Ley, L.; Stacey, A.; McCallum, J. C., MoO3 induces p-type surface

conductivity by surface transfer doping in diamond. Appl. Surf. Sci. 2020, 509,

144890.

26. Scanlon, D. O.; Watson, G. W.; Payne, D.; Atkinson, G.; Egdell, R.; Law, D.,

Theoretical and experimental study of the electronic structures of MoO 3 and

MoO2. J. Phys. Chem. C. 2010, 114 (10), 4636-4645.

27. Han, X.; Gerke, C. S.; Banerjee, S.; Zubair, M.; Jiang, J.; Bedford, N. M.;

Miller, E. M.; Thoi, V. S., Strategic design of MoO2 nanoparticles supported by

carbon nanowires for enhanced electrocatalytic nitrogen reduction. ACS Energy

Lett. 2020, 5 (10), 3237-3243.

28. Liu, C.; Luo, S.; Huang, H.; Zhai, Y.; Wang, Z., Direct growth of MoO2/reduced

graphene oxide hollow sphere composites as advanced anode materials for

potassium‐ion batteries. ChemSusChem 2019, 12 (4), 873-880.

29. Yang, J.-L.; Zhao, S.-X.; Lu, Y.-M.; Zeng, X.-T.; Lv, W.; Cao, G.-Z., In-situ

topochemical nitridation derivative MoO2-Mo2N binary nanobelts as

multifunctional interlayer for fast-kinetic Li-Sulfur batteries. Nano Energy 2020,

68, 104356.

10
30. Han, J. L.; Yang, J.; Tang, J. B.; Ghasemian, M. B.; Hubble, L. J.; Syed, N.;

Daeneke, T.; Kalantar-Zadeh, K., Liquid metals for tuning gas sensitive layers. J.

Mater. Chem. C 2019, 7 (21), 6375-6382.

11
Chapter 2 Literature review

12
2.1 Introduction

In the periodic table of elements, the metallic elements located between transition metals

and non-metallic metalloids are known by different names such as post-transition metals,

p-block metals, poor metals, and chemically weak metals. Several arrangements have

been introduced for the post transitions metals depending on where transition metals and

metalloids begin and end in the periodic table. Usually, post-transition metals are the

group 13–15 metals in periods 4–6 and sometimes group 11 and group 12 metals are also

assumed as post transition metals1, which are otherwise considered to be transition metals

or metalloids. However, all proposals for post transition metals include gallium, indium,

tin, thallium, lead, and bismuth, which are considered as the core members of the post

transition metals.2

Being located at the metal-nonmetal border, the crystalline structures of post transition

metals show covalent or directional bonding effects with more complexity than other

metallic elements. The reduced metallic characteristics in post transition metals are

dominantly attributed to the increasing nuclear charge across the table from left to right.

The accumulating nuclear charges are partially offset by the increasing number of

electrons, but extra electrons are not able to fully screen each successive increase in

nuclear charge. Therefore, the nuclear charge dominates.3 As such, the atomic radii

compress, ionization energies increase, fewer electrons become accessible for metallic

bonding4, and ions shrink and become more prone to covalency.5 These trends are more

obvious for post transition metals in periods 4-6 due to less screening of nuclear charges

by d10 and f14 electron configurations.6

Due to their electron configurations, post transition metals show many distinguished

features. In addition to their weak mechanical strength and poor ductility and malleability,

13
the most distinguished characteristic of post transition metals is their low melting points. 7

One of the core members of the post transition metals family, gallium, exhibits a melting

point just lower than 30 ℃,8 meaning it can melt in one’s palm. Upon the formation of

eutectic alloys, the melting points of most post transition metals can drop even further.

For example, eutectic alloys of gallium-indium (EGaIn) have a melting point of 15 ℃,

which takes on liquid form at room temperature.9 Members of post transition metals form

many eutectic alloys with melting points lower than 330 ℃, which is practically

accessible and easy to handle in safe laboratory environments.7 Furthermore, unlike other

low-melting-point metals, like alkali metals that are super-reactive in the presence of air

and moisture, or mercury which is highly toxic, post transitions liquid metals are

reasonably stable and safer to handle. These characteristics make post transition liquid

metals suitable for various applications like synthesis of materials, catalysis, and

fabrication of flexible optoelectronic devices.10-14

In addition to the intrinsic low melting points of post transition liquid metals, the interface

induced atomic layering renders them perfect media for facile synthesis of two-

dimensional (2D) materials.15 Theoretical calculations predict the existence of an

extremely flat layer of free electrons at the surfaces of liquid metals.16, 17 This surface

electron layer acts as a hard wall suppressing positional fluctuations of the near surface

ion cores, which leads to their positional ordering. The electron density profiles normal

to the surface are predicted to exhibit characteristic oscillations at the interface, which

decay to the bulk electron density after a few atomic distances.18, 19 This prediction was

verified by the observation of quasi-Bragg peaks in the X-ray reflectivity measurements

of many liquid metals.20

This surface layering phenomenon endows liquid metals with atomically flat surfaces,

which make perfect templates for 2D materials synthesis. Upon exposure to oxygen, a
14
thin layer of metal oxide emerges on the surface of liquid metal, which inherits its

atomically flat nature and has been proven to be ultrathin and uniform. 21 The surface
22
oxidation of liquid metals not only has significant effects on their physical properties,

but also provides routes for facile synthesis of 2D materials.23 Thermodynamic laws

govern the surface oxidation of liquid alloys, where the metal oxide with the lowest Gibbs

free energy dominates the surface. Thus, various metal oxide layers can be synthesised

on the surfaces of liquid metals with different compositions.24 The resultant surface oxide

layers are easily transferrable and could serve as precursors for other 2D metal

compounds upon post treatments.25-28 The surface layering phenomenon of liquid metals

was observed not only in vacuum or gaseous environments, but also when they are in

contact with aqueous solutions29, 30 or solids.31

The ultra-smooth surface of liquid metal could also serve at a reactive medium for liquid

metal-aqueous solution interfacial synthesis32-36, where the reaction environment at the

interface could be tuned by varying the composition, concentration or pH of the solution.

Liquid metal provides a super active surface for interfacial reactions that can take place

under catalytic or reductive conditions. In this case, precursors of certain materials can be

brought in contact with the surface, and subsequently, a 2D component is formed.

Due to the liquidity and non-polar nature of liquid metals, the synthesised surface layers

of 2D nature could be easily exfoliated from the bulk metal by mechanical or chemical

means.24, 37 In particular, such exfoliation techniques allow low temperature deposition

of 2D layers on various substrates.24, 32, 35, 38 When a surface layer comes to the vicinity

of a flat substrate of polar nature, the formation of van der Waals forces between the 2D

material and the substrate, and the minimal forces between the 2D layer and the liquid

core, allow for the transfer of the 2D material.28, 35 Various electronic and optical devices

15
have been developed using liquid metal process induced 2D materials as dielectric layers,

conductors, semiconductors, or piezo-responsive layers. 23

In this chapter, the author discusses the unique properties of post-transition metals based

liquid metals and comprehensively reviews the recent progress of 2D materials synthesis

via liquid metals processes. First, the author introduces the properties and features of these

liquid metals including their intrinsic low melting points, high thermal and electric

conductivities, surface layering phenomenon and discusses the fundamentals behind

them. Next, based on these unique features, synthesising techniques of various 2D

materials at the surface of liquid metals are introduced where liquid metals serve as either

precursor or reducing agents. Various extraction and transferring methods are introduced

which can deposit large area 2D materials. Then, applications of these 2D materials in

devices are summarized.

The content of this chapter was published as a peer-reviewed article in the Journal of

Applied Physics Reviews.39

2.2 Properties of post transition liquid metals

2.2.1 Origin of low melting points

The melting points of metals are affected by several factors such as crystal structures at

solid state, degrees of ionization and the population of electrons involved in the bonding. 10

When the melting point of metals40 are plotted against the group numbers for the three

long periods, a similar volcanic shape emerges which indicate a periodic property of

metals that is in correspondence with their electron and band structures (Figure. 2.1). To

16
some degrees, similar behaviours are found in other properties such as boiling points,

enthalpies of fusion, densities, and hardnesses.40

The initial increases in the strength of metallic bonding, as we move from group 1 to 6,

can be explained by the number of valence electrons that the metal can contribute to the

electron sea. The more electrons an atom loses, the larger is the charge of the positive ion

embedded in the electron sea and the greater is the electron probability density of the

electron sea itself. However, this trend cannot continue indefinitely. The more electrons

removed from an atom, the more energy it takes to remove the next electron. Eventually,

more energy is needed to remove an electron from a metal nucleus than is liberated by

placing it in the electron sea. The strength of the metallic bonding thus begins to level off

and eventually drops. In the region where the strength of the metallic bonding drops, due

to the poor screening effect of the d and f orbitals over the nuclear charge, covalent bonds

begin to emerge. For the group 12 elements (Zn, Cd, and Hg), low melting and boiling

points are attributed to their pseudo noble gas electron configurations since electrons

completely fill s and d shells. Therefore, valence electrons of these elements are relatively

inert and present only weak interactions.1, 41 Post transition metals present low melting

points but relatively high boiling points with broad temperature ranges in liquid state. In

the electronic configuration of post transition metals, d and p shells are completely and

partially filled, respectively.42 The unpaired electrons in p shell can generate covalent or

metallic bonds with significant covalent feature for the lighter elements, hence the

contributions of the reactive p shell electrons increase boiling points in the post transition

metals.43

17
Figure 2.1 Melting point of metals of fourth, fifth and sixth periods graphed against

group number that shows a volcanic feature.

In addition, the degree of delocalization of the valence electrons and the crystal structure

of solid phase determines the melting temperature of metals. Metallic bond occurs in both

solid and liquid states due to the low dimensionality, especially if they do not possess a

pseudo noble gas configuration.10

Yu and Kaviany carried out ab initio molecular dynamics (AIMD) simulations to explore

the atomic structures of liquid Ga, EGaIn, and eutectic alloys of gallium-indium-tin

(EGaInSn) and the effect of Ga alloying with Sn and In on their electrical, thermal, and

species transport properties.44 Their structures were investigated through the pair

correlation function, g(r), which provided close estimations of the atomic distribution,

average distance, and medium-range order in those alloys. Figure 2.2A presents the

estimated pair-correlation functions g(r) for Ga at both 310 K and 500 K and EGaIn, and

EGaInSn at 310 K.

The average probability of atoms' location at a given distance r from the designated atom

is understood by the variations of g(r). A relatively strong peak followed by weaker peaks
18
are seen in Figure 2.2A as a function of distance, which suggests the existence of strong

interactions in the first shell and reduced long range order. However, alloying of Ga with

Sn and In leads to a height reduction in the first peak and the shift of the g(r) to the right,

indicating the further distribution of atoms in that coordination shells. The shift of the g(r)

also shows the change in the average distance between atoms. By increasing the

temperature to 500 K, the first peak at g(r) shifted to a smaller distance due to distance

reduction between atoms in the first shell. However, it was expected, similar to alloying

with Sn and In, to shift to larger distance because of thermal expansion. This abnormal

peak shift and decrease in the inter-atomic distance can be explained by the redistribution

of the polyhedral clusters, which is a general characteristic for the metallic melts

containing various-size polyhedrals.45 Thus, the high-temperature structural behaviour

and alloying affect the first peak shifts in g(r) via different ways. Alloying, for instance

for EGaIn and EGaInSn productions, increases the average distance in the first shell due

to disordered structure, while the redistribution of the polyhedral clusters is the leading

factor at the high temperature. Displacing Ga atoms with larger In and Sn increases the

atomic volume and shifts the first peak positions to the right. Volume per atom

continuously increases to 19.01, 20.11, 20.99 Å3/atom for Ga, EGaIn, and EGaInSn,

respectively.

As shown in Figure 2.2B, in EGaInSn, the charge density is higher between Ga atoms

indicating a stronger force field among Ga atoms compared to In and Sn. Especially, the

charge density near the In atoms is significantly small which explains the dominant role

of In in reducing the electrical transport properties in Ga alloys containing In. In the case

of Hg, electrons in the low energy orbitals are strongly bound to the atom, which results

in a low melting temperature.41 Electrons distribution around the Sn and In atoms are

comparable to Hg. As such, EGaIn (288.65 K) and EGaInSn (283.65 K) show lower

19
melting points than Ga (303 K). The weak interactions of atoms induced by the In and Sn

in EGaInSn are more clearly illustrated in the isosurface charge density Figure 2.2C. The

study of partial pair distributions functions, gi(r) revealed that the inter-atomic distance

among Ga atoms in EGaInSn is similar to liquid Ga and smaller than those inter-atomic

distances involving the In and Sn atoms, as shown in Figure 2.2D. The larger atomic

volume of In and Sn leads to the disordered structure in EGaInSn by expanding the cell

volume and broadening their inter-atomic distances.

2.2.2 Electrical conductivity

Gallium in a liquid state renders a higher electrical conductivity than solid. The variation

of resistivity with temperature in Ga is linear. In fact, the conductivity of Ga-based liquid

metals is much higher than other liquids.46 Electrical and thermal conductivities in liquid

Ga-based alloys originate from their electron-rich bulk.47 AIMD simulations using Kubo-

Greenwood (K-G) and the Ziman-Faber (Z-F) formulations have been performed to

predict the electrical conductivity, σe, of liquid Ga, EGaIn, and EGaInSn as a function of

temperature. As shown in Figure 2.2E, the electrical conductivity of Ga (3.82×10 6 Ω-

1
m−1) is higher than EGaIn (3.28×106 Ω-1m−1) and EGaInSn (3.14×106 Ω-1m−1) at 310 K

temperature. This decrease arises from the disordered structures and weak inter-atomic

interactions due to the incorporation of Sn and In in EGaIn and EGaInSn (Figure 2.2B

and C). The effect of alloying on the electrical conductivity reduction of pure liquid

metals can be explained by Matthiessen’s rule. A size mismatch occurs in alloyed metals.

This size mismatch leads to a higher degree of structural disorder and weaker interatomic

interactions 48, 49, as shown in Figure 2.2B for Ga alloying with In and Sn.

20
Interestingly, Hg shows a lower electrical conductivity than Ga at room temperature

(1.04 106 Ω-1m-1). The electrical conductivity of Hg increases by alloying with most

metallic elements, such as In, Tl, Zn, and Pb.50 In contrast, electrical conductivity of other

metals is typically lowered by alloying according to Matthiessen's rule.

The electrical conductivity of pure liquid metals and alloys decreases with increasing

temperature. At higher temperatures, thermal motion of atoms and collision between

travelling electrons increase which leads to a less efficient charge transport due to the

scattering of the electrons out of their original paths.47 Electrical conductivity at a given

temperature can be calculated by σe = σeR/[1+ α(T-TR)]. In this equation, α is the

temperature coefficient of the electrical conductivity, σe and σeR are the electrical

conductivities at temperatures T and TR, respectively. Here R denotes the values for room

temperature.51

The solid-liquid phase transition has a significant effect on the electrical conductivity of

metals and alloys. The electrical conductivity exhibits a jump or discontinuity at the verge

of melting point and drops by a degree of ~1.5−2.3 for most of the normal metals just

below the melting point. This reduction at the transition stage is attributed to the more

disordered arrangement of the ions in the liquid state.52 However, certain abnormal metals

including Ga, Bi and Sb, which are rather poor conductors in the solid state, exhibit a different

trend and increase their conductivity upon melting.53 This anomalous behaviour is related

to anisotropy in the electrical conductivity of solid crystals of such elements. Single

crystals of metals such as Ga show higher anisotropism in their conducting properties

than those of other metals.54

21
2.2.3 Thermal conductivity

The thermal and electrical conductivities of metals are closely related, and both involve

the behaviour of free electrons. Similar to electrical conductivity, thermal conductivity is

anisotropic in the solid state.55 In the case of liquid metal, the total thermal conductivity,

kt, in liquid metals can be defined as the sum of molecular liquid thermal conductivity, kf,

and the electronic thermal conductivity, ke, where ke takes a dominant role.56 Unlike

electrical conductivities, thermal conductivities generally increase with temperature. The

electronic thermal conductivity (ke) and electrical (σe) conductivities are linked

quantitatively through the Wiedemann−Franz (W-F) law, ke/ σe =LT, where L and T are

Lorenz number and temperature, respectively. 57, 58

As shown in Figure 2.2G, the

experimentally measured thermal conductivity of Galinstan and its calculated electronic

thermal conductivity are graphed against temperature for comparison. 57

In contrast to liquid Hg with low thermal conductivity (8.5 W m-1K-1 at 292.15 K)59,liquid

Ga shows a higher thermal conductivity of approximately 29.3 W m -1K-1 at 302.15 K.60

Molecular liquid thermal conductivity kf of Ga is approximately 1.65 W m-1K-1 at the

temperature range of 293.3–348.3 K, which is higher than typical nonmetallic liquids due

to the low compressibility44 and high sound speeds24 in liquid metals kf is mostly

dominated by the Ga-Ga interactions, which is why the value of kf for liquid EGaIn and

EGaInSn is close to liquid Ga. Considering that ke is much larger than kf in liquid Ga and its

alloys, ke more strongly dictate kt variations in these liquid metals. The effect of Ga

alloying with In and Sn on kt follows the interpretation of the σe. As shown in Figure 2.2F,

the kt drops from that of Ga to lesser values for EGaIn and EGaInSn due to the more

disordered structures, which hinder the electrons transports and decrease electrical and

thermal conductivities. The thermal conductivity of Ga reduces upon alloying to 26.4 and

25.4 Wm-1K-1 (310 K) for EGaIn and EGaInSn, respectively.44

22
In another study61, the thermal conductivity of Al, Sn, Pb and Cu pure liquid metals and

Cu–Al, Ag–Ga, Ag–Ge, Cu–Pb, In–Mn, Ga–Ge and Sn–Bi bimetallic alloys were

obtained using the electrical resistivity, thermoelectric power data, and the relations

between the transport coefficients simplified using W–F law. The calculations for the

pure metals presented a close match with the experimental results. In terms of alloys, the

thermal conductivities correlate with their composition. As shown in Figure 2.2H, those

liquid metal alloys showed either a monotonically decreasing thermal conductivity with

the increasing concentration of polyvalent metals, or a minimum thermal conductivity at

around 20 at.% of the polyvalent metal.

23
Figure 2.2 (A) Predicted pair correlation function for liquid Ga, EGaIn, and EGaInSn at

310 K and for Ga also at 500 K. (B) predicted partial radial distribution function, gi(r) for

liquid EGaInSn at 310 K. (C) Contours of charge densities for liquid EGaInSn at 310 K.

(D) Structure of the liquid EGaInSn at 310 K. (E) the variations of electrical conductivity

with respect to temperature obtained for liquid Ga, EGaIn, and EGaInSn through K-G

and the Z-F treatments compared to some experimental results. (F) The predicted

variations of the total thermal conductivity with temperature for liquid Ga, EGaIn, and

EGaInSn and comparison with the experimental results.62 "Reprinted with permission

from Yu et al., J. Chem. Phys. 140, 064303 (2014). Copyright 2014, AIP Publishing.44

(G) Temperature dependence of thermal conductivity for liquid Galinstan. Open circles

and dashed lines indicate the experimental data and values obtained by the W-F law,

respectively. Filled and open triangles are literature values60 for comparison. Reprinted

with permission from Plevachuk et al., J. Chem. Eng. Data 59, 757 (2014). Copyright

2014, American Chemical Society.57 (H) Thermal conductivity plotted against

concentration for Sn–Bi, Ga–Ge, Ag–Ge, Cu–Al, In–Mn and Cu–Pb alloys . Reprint with

permission from Giordanengo et al., J. Non-Cryst. Solids 250, 377 (1999). Copyright 1999,

American Chemical Society. 61

2.2.4 Surface-induced atomic layering

Although the bulk of liquid metal is generally not structured (shows no long-range or

short-range order) due to its liquidity, a quasi-crystalline layer of nanometre thickness

tends to present on its surface, which is associated to ‘surface layering’ phenomenon. This

phenomenon in fact lays the very foundation that allows the implementation of liquid

metals, with their intrinsically atomically flat surfaces, for synthesising and templating

2D materials. The existence of these surface layers was first theoretically predicted in the
24
early 1980s.16, 17
The surface induced atomic layering of liquid metals was initially

observed in liquid gallium19 and mercury18 and was found later in other metals in liquid

states including In63 K64 Sn65, 66, as well as various liquid metal alloys.67 X-ray reflectivity

measurement and X-ray diffuse scattering was used for the characterization of the surface

electron density profile of liquid metals, whose experimental geometry setup is

schematically illustrated in Figure 2.3A. kin and kout are the wave vectors of the incident

and reflective beams, respectively, while q is the wave vector transfer that consists of the

surface-normal (qz) and in-plane (qxy) compounds. In X-ray reflectivity measurement,

incident angle and the reflection angle change simultaneously at the same rate. In

this case, the in-plain momentum transfer qxy remains zero, and the reflectivity can be

plotted against qz. However, in the X-ray diffuse scattering geometry, the incident angle

remains constant, while the detection angle varies within the y-z plane.

The X-ray reflectivity measurement is based on equation (2.1)68 which related the

reflectivity R( ) to the surface normal electron density profile e (z). In this expression,

RF ( ) is the theoretical Fresnel reflectivity of X-rays incident to an ideal flat surface, and

e
(∞) is the average electron density in the bulk. The local structure factor of the surface

(𝑞 ), according to the definition in equation (2.1), can be correlated to the in-plane-

averaged electron density profile e

(z) along the surface normal z axis based on the

Fourier transform. As shown in Figure 2.3B, the X-ray reflectivity measurement of liquid

gallium shows a distinct quasi-Bragg peak at qz=2.4 Å-1, which indicates a surface

induced atomic layering.20 The electron density profile inserted shows oscillations at the

gallium-vacuum interface decaying to the bulk, which agrees with the X-ray reflectivity

results. A similar phenomenon was observed in other liquid metals like Hg, Sn, and In.

25
 ∞ d〈 e (z)〉
1
R( ) ≈ RF ( ) dz eiqz z ≡ RF ( )| (𝑞 )| (2.1)
e
(∞) -∞ dz

18, 19
Surface layering was observed not only in liquid metals with high surface tension ,

but also in low-surface-tension liquid metals like potassium 64, whose surface tension is

around 100 N m-1 and comparable with that of water. The surface layering of liquid metal

could be attributed to the unique electronic structure of metallic liquids and the steep

variation of electron density across the liquid/vapor or liquid-vacuum interface. 69

Interestingly, surface layering of liquid metals was not only found when in contact with

vapor, but also with other liquids or flat solid surfaces. In 1997, surface structure of liquid

gallium was investigated by S. A. de Vries et al. when it is in contact with the (111)

surface of diamond.70 The measured X-ray reflectivity shows a broad oscillation which is

a signature of layerwise ordering at the interface. The liquid metal-aqueous solution

interface was also found to be layered. The X-ray reflectivity of liquid mercury (Hg) and

sodium fluoride (NaF) solution shows a peak at 2.2 Å−1 (Figure 2.3D), indicating a layered

structure of liquid Hg surface (Figure 2.3C).71 In addition, liquid metals’ surface structure

could be tuned by altering the composition and the concentration of the electrolyte or

applying bias upon the interface.29 This observation indicates that liquid metals exhibit

an atomically flat surface when in contact with other liquids, which lays the foundation

of the liquid metal-assisted 2D materials synthesis at liquid-liquid interface.

Other than surface layering, distinctive surface properties are observed in liquid alloys

like surface segregation72 and surface freezing.31 Surface segregation in liquid alloys

refers to the enrichment of one of the components at its surface, leading to different

chemical compositions between the surface and bulk. When the liquid metal is composed

with multiple components with different surface energies, Gibbs’ adsorption rule predicts

that the components with lower surface energy tend to segregate at the surface. In many
26
cases, the surface segregation coexists with surface layering in liquid alloys. For example,

in case of Ga–In,21 Ga–Sn,73Ga–Bi,74, 75 and Bi-Sn alloys76, the first surface layer is

dominated by the lower surface energy component (In, Sn or Bi) despite its low

concentration in bulk. In the following atomic layers, the bulk composition begins to

emerge. For other compositions of alloys, segregation into distinct crystal morphologies

can be observed during solidification. 77, 78 For some alloys, at a temperature slightly above

the melting point, the formation of a solid crystalline layer could be observed at their

surface, which is called surface freezing or surface crystallization. One of the examples

of surface freezing is seen in eutectic Au-Si alloys, which was verified by grazing

incidence X-ray diffraction measurements. 31

Figure 2.3 (A) Schematic illustration of the experimental geometry setup of X-ray

reflectivity measurement and diffuse X-ray scattering measurements. Reprinted with

permission from Shpyrko et al., Phys. Rev. B 67, 115405 (2003). Copyright 2003,

27
American Physical Society.64 (B) The X-ray reflectivity measurement of liquid gallium,

with the inset surface-normal electron density profile. Reprinted with permission from

Pershan et al., J. Appl. Phys. 116, 222201 (2014). Copyright 2014, AIP Publishing.20 (C)

Schematically illustration of the layered surface structure and electron density profile

normal to the interface. (D) X-ray reflectivity measurement of the interface between

liquid Hg and 0.01 M NaF solution at a potential of 0.28 V relative to the potential of zero

charge. The peak at 2.2 Å−1 indicates Hg surface layering. Profiles of the reflected beam

at different qz positions (1.6, 1.8, and 2.2 Å−1) are displayed in the inset. Republished with

permission from Murphy, et al., Nanoscale 8, 13859 (2016). Copyright 2016, Royal

Society of Chemistry.71

2.2.5 Surface oxidation of liquid metals

Liquid metals are highly reactive and rapidly oxidized in the presence of ambient oxygen

which is widely used for the fabrication of 2D metal oxides.15 Many metals follow a

similar oxidation behaviour at sufficiently low temperature with an extremely rapid initial

oxidation rate which drops to low and negligible values after a short time. The formation

of stable oxide films at low temperature was explained by Mott et al.79 stating a strong

field across the oxide film resulted from different contact potentials between metal and

adsorbed oxygen. In such cases, the electrons of the metal move from the inner core

through the tunnel effect and are transferred to the adsorbed oxygen. As a result, the

adsorbed oxygen is ionized, and an electrostatic potential (Mott potential) is created

between the oxide-metal and oxygen-oxide interfaces, as shown in Figure 2.4A.80 The

subsequently generated electric field significantly reduces the energy barrier for metal ion

diffusion through the oxide layer. As a result, the part of oxidation rate, determined by

28
the transfer of metal ions increases, promoting the growth of the oxide skin. Based on the

Cabrera-Mott model80, the equation 1/X = A B ln t (where X is the thickness of the oxide

film, t is the oxidation time, and A and B are the integration and rate constants,

respectively) holds the relation between the oxidation time and the thickness of the oxide

layer indicating that the oxide thickness increases with the oxidation time. The oxidation

rate drops with increasing oxide layer thickness and reaches a negligible rate at a critical

thickness, leading to the self-limiting surface oxidation phenomenon.

X-ray reflectivity measurements can be employed to investigate the low-temperature

oxidation kinetics of liquid metals and liquid alloys. 81 It was described previously that the

reflectivity measurements give information about the laterally averaged electron density

profile along the surface normal. The finite jump in the electron density at the metal-oxide

layer interface drives reflectivity oscillations and the layer thickness can be determined through

the oscillation frequency.

Self-limiting oxidation behaviour has been explored in liquid Ga by in situ X-ray

reflectivity measurements.82 As shown in Figure 2.4B, the position of the measured

reflectivity in different oxygen doses (expressed in Langmuirs 1L = 10-6 Torr s) is almost

fixed, indicating a constant thickness of oxide layers on liquid gallium. However, the

increasing intensity from the minimum oxygen dose to the maximum indicates the

expanding oxide coverage on the surface of liquid gallium. As Shown in Figure 2.4C, the

surface normal electron density profile, deduced from the X-ray reflectivity measurement

(bottom), agrees perfectly with the proposed atomic arrangement in the oxide layer. The

gallium oxide plane consists of alternating layers of Ga3+ and O2- ions, terminated by an

O2- layer, revealing a self-limiting oxide skin with a thickness of ~5 Å. In the absence of

any external perturbation, this dense and homogeneous shell perfectly isolates the bulk

liquid and prevents further oxidation.

29
Some liquid metals like Bi, Sn, In and Gd exhibit a non-self-limiting oxidation behaviour

where the oxide layer grows and thickens continuously as the formed oxide skin is

unstable, porous, and unable to prevent the liquid metal from further oxidation 83-86 The

oxidation kinetics and the change in the oxide thickness of Ga0.93Hg0.07 with time and

temperature were explored by in situ X-ray reflectivity technique and were compared

with Ga.81 It was found that the logarithmic growth behaviour can be observed for

ultrathin oxide films well below 40 Å. As shown in Figure 2.4D, the thicknesses of the

oxide layers grow with time at each temperature and are in agreement with the logarithmic

kinetics. The thickness of the oxide on liquid Ga exceeds that on Ga0.93 Hg0.07, (Figure

2.4E) due to a higher activation energy for the oxidation of the alloy.

In liquid metal alloys, selective oxidation occurs if the alloy is composed of metals with

different reactivities. In this case, the reactive metal preferentially oxidizes and dominates

the surface of the liquid alloy. According to the thermodynamic law, the surface of an

alloy is covered by metal oxides which cause the greatest decrease in the Gibbs free

energy.24 For example, the surfaces of gallium-based alloys such as EGaIn and Galinstan

are covered by Ga2O3 since gallium is the most reactive component in the alloy. In

contrast, competitive oxidation occurs if the metals are equally or similarly reactive, and

a mixed metal oxide is created on the surface of the liquid alloy, while the priority of the

surface oxidation may be determined by other factors such as surface energy. Therefore,

tuning the composition of liquid alloys regulates the composition of the formed oxide

skins. In addition, environmental conditions, such as ambient oxygen concentration, also

alter the outcomes.

30
Figure 2.4 (A) Two spherical configurations with different component layouts for the

oxidation of a metal particle in an oxidizing environment. Reprinted with permission from

Alexandre et al., Chem. Phys. Lett., 505, 47 (2011). Copyright 2011, from Elsevier.80 (B)

Reflectivity measurement for liquid-Ga surfaces dosed with an increasing amount of

oxygen. At the oxygen dose of 800 L, the sample temperature was increased from 25 to

300 ˚C without an observable change in the reflectivity. (C) Real-space models for the

proposed atomic arrangement in the oxide layer of Ga (top), and the corresponding

electron density profile with its separate components (bottom). Reprinted with permission

from Regan et al., Phys. Rev. B 55, 10786 (1997). Copyright 1997, from American

Physical Society.82 (D) The thickness of the oxide layer on Ga0.93Hg0.07 plotted against

time. (E) The growth rate of oxide film on liquid Ga compared to Ga 0.93Hg0.07 at similar

temperatures. Reprinted with permission from Plech et al., J. Condens. Matter Phys. 10,

971 (1998). Copyright 1998, from IOP Publishing, Ltd. 81

31
2.3 Interfacial synthesis of 2D materials with liquid metal

2.3.1 2D materials synthesis via surface oxidation of liquid metals

Surface oxidation of liquid metals offers unique opportunities for the synthesis of 2D

materials. When the surface oxide layer is stable and homogenous, it would passivate the

surface of liquid metals and hinder its further oxidation, which is described by the

Cabrera-Mott kinetics.79 When the surface oxidation does not follow the Cabrera-Mott

kinetics and the oxidation layer grows continuously, growth of the monolayer metal oxide

could be achieved by lowering the oxygen concentration in the ambient or shortening the

reaction time.87, 88 In the case of liquid alloys, competitive surface oxidation may occur,

and the alloy surface is then dominated by metal oxide whose formation is the most

thermodynamically favourable.24 Exposing liquid metals to other gaseous compounds

(like hydrogen sulfide) in the ambient could expend this method to other 2D metal

compounds other than oxides. The resultant 2D material could be easily exfoliated due to

the non-polarity and liquidity of liquid metals.28 Touch-printing24 or squeeze printing38

could transfer the uniform 2D materials at a large area on a secondary substrate, while

utilising bubbling or sonication could produce nanoflakes of the metal compound with

small lateral dimensions at high yield.

In many cases, the oxidation of liquid metals in air/oxygen follows the Cabrera-Mott

kinetics and exhibits fast initial oxidation rates, which rapidly decrease to negligible rates

due to the slow mass transfer through the dense metal oxide layer. Thus, the growth of

oxide layer is self-limiting, which results in ultrathin, uniform and smooth and

transferable metal oxides films at ease. The most typical example of Cabrera-Mott

oxidation is gallium, which was discussed in detail in the previous section. Liquid gallium

has been widely utilised for room temperature deposition of wide bandgap semiconductor

32
gallium oxide (Ga2O3).89, 90 (Figure 2.5A-C) Other than gallium oxide, the Cabrera-Mott

oxidation has been used for the synthesis of other metal oxides in a large area such as lead

oxide (PbO),91 and zinc oxide (ZnO).92 (Figure 2.5D-G)

Figure 2.5 Self-limiting growth of oxide surface layer on liquid metal for 2D-metal oxide

deposition (Ga2O3 and ZnO) (A) Large-area 2D Ga2O3 amorphous layer deposited on the

2-inch Si-SiO2 substrate. (B) Schematic diagram of the atomic model describing the

structural transformation of the 2D Ga2O3 amorphous layer to the 2D β-Ga2O3 crystal via

thermal annealing. (C) High-resolution transmission electron microscope (TEM) of the

2D crystalline Ga2O3 crystal image. The selected area electron diffraction (SAED) pattern

is inserted. Reprinted with permission from Li et al., Chem. Mater. 33, 4568 (2021).

Copyright 2021, American Chemical Society.89 (D) Optical microscopy image of

homogenous millimetre-scale ZnO sheet deposited on Si-SiO2 wafer (scale bar: 100 m).

33
(E) Atomic force microscope (AFM) image of a 0.6-nm-thick ZnO sheet (scale bar 500

nm) (F) AFM image of a 1.1-nm-thick ZnO sheet (scale bar: 4 m) (G) TEM image of a

ZnO sheet showing preferred growth direction in wurtzite phase of ZnO (scale bar: 1 nm).

Reprinted with permission from Mahmood et al., Mater. Today 44, 69 (2021). Copyright

2021, Elsevier.92

In other cases, such as the oxidation of liquid Sn, In and Bi, oxidation results in porous

or unstable oxide layer that’s unable to prevent further oxidation, and the surface oxide

layers grow continuously.87, 88, 93 In these cases, thickness and morphology of the planar

metal oxide vary with the reaction time and ambient oxygen concentration, while

monolayer and uniform metal oxide sheet could be achieved by precise control of these

reaction conditions. For example, when exposed to air, the surface of liquid tin goes

through a progressive colour change as the continuous oxidation process leads to surface

oxide colour due to optical reflections and absorption of different thicknesses and various

stoichiometries. Upon exposure to air, a monolayer tin oxide (SnO) sheet instantly forms

on the surface of molten tin, which goes through further oxidation into SnO 2, Sn2O3, and

Sn3O4 as the uniform sheet breaks into small islands. The breakage of the initial SnO

sheet results in exposure of molten tin surface underneath, and the oxidation process goes

on, leading to thicker films composed of tin oxide islands laying on top of each other.

Therefore, different reaction durations lead to tin oxide layers of different colours,

compositions, and morphologies (Figure 2.6A-E).93 A reduced oxygen concentration (10-

100 ppm) could significantly decrease the oxidation rate of molten tin and result in a

highly reproducible deposition of uniform and large area SnO (Figure 2.6F-I). 87 Similarly,

34
an atmosphere of reduced oxygen concentration (10-100 ppm) was utilised for the

synthesis of monolayer bismuth oxide (Bi2O3) synthesis (Figure 2.6J-M).88

Figure 2.6 Deposition of oxide tin oxide and bismuth oxide layer on liquid metal under

controlled oxygen concentration (e) (A) Schematics presentation of tin oxide layers of

different thickness, morphology and colour obtained by exposing molten tin to ambient

air at different times. (B-E) TEM images of fresh, yellow, pink, and grey tin oxides

obtained with different reaction times. Reprinted with permission from Atkin et al., Chem.

Commun. 54, 2102 (2018). Copyright 2018, Royal Society of Chemistry. 93 (F) TEM

image of a nanosheet synthesised in a continuously purged glovebox (N 2 atmosphere, O2

35
concentration typically between 10 and 100 ppm). (G) TEM image of a freestanding edge

and (H) High-resolution TEM image of a sample synthesised under reduced oxygen

conditions, inset features the Fast Fourier Transform (FFT) image of the High-resolution

TEM micrograph. The FFT image has been indexed to SnO. (I) AFM image of tin oxide

synthesised under reduced oxygen conditions. Step heights at the red and blue lines are

1.1 nm and 0.4 nm, respectively. Reprinted with permission from Daeneke et al., ACS

Nano 11, 10974 (2017). Copyright 2018, American Chemical Society. 87 (J) AFM

topography of touch printed Bi2O3 layer inside a controlled environment where oxygen

concentration is below 100 ppm. (K) TEM image of a Bi2O3 monolayer (synthesised

where oxygen concentration below 100 ppm) using a holey carbon TEM grid as a

substrate. (L) High-resolution TEM demonstrating the crystallinity of the ultrathin nano-

sheets of Bi2O3 with the planar spacing indicated. (M) The corresponding FFT image of

the area. Reprinted with permission from Messalea et al., Nanoscale 10, 15615 (2018).

Copyright 2018, Royal Society of Chemistry.88

For liquid metal alloys, selective oxidation occurs where the compound with higher

reactivity is preferentially oxidized, and its metal oxide dominates the surface of the liquid

alloy. It was governed by the thermodynamic law that the surface of an alloy is dominated

by metal oxide with the lowest Gibbs free energy. For example, the surfaces of the

gallium-based eutectic alloys such as EGaIn and Galinstan are covered by Ga2O3 since

its formation leads to a greater decrease in Gibbs free energy than indium or tin oxide.

Therefore, gallium-based eutectic alloys, such as EGaIn, could also be used for the

deposition of 2D-Ga2O3, and yield Ga2O3 with high purity and quality.90 Similarly, Al2O3,

Gd2O3 or HfO2 could be harnessed from the surface of gallium based alloys, even at bulk

concentrations as low as 1 wt. % of aluminium, gadolinium, or hafnium, since oxidation

36
of these metals have smaller Gibbs free energies than gallium (Figure 2.7). 24 These

resultant 2D metal oxides were proven to be of 0.5-1.1 nm thickness and highly

crystalline. It is noteworthy that this method offers the possibility of synthesising 2D

materials that are not intrinsically layered, and therefore, cannot be produced as

atomically thin nanosheets using conventional methods. In addition, by alloying a small

concentration of metals with gallium-based liquid metals, this room temperature

exfoliation strategy can be applied to metals with exceedingly high-melting points of

above 1000 °C like aluminium.

In the case of indium-tin alloy, the surface oxidation leads to a ternary compound that

contains both indium and tin oxides, which enables low temperature deposition of indium

tin oxides (ITOs), a commonly used transparent conductive oxide.38 Due to the solubility

of tin ions in indium oxide, the formation of ternary ITOs is more thermodynamically

favourable than pure In2O3 or SnO2. Squeeze printing of indium-tin alloys at room

temperature was shown to result in centimetre-scale deposition of ITO sheets on various

substrates including the flexible and heat-sensitive polymer substrates. The resultant ITO

sheets were highly transparent and remained in good conductivity upon elastic

deformation. It was also presented that multiple depositions of the ITO layers could be

achieved.

The surface oxidation of liquid metals offers highly versatile approaches for 2D materials

synthesis that are not limited to the original metal oxides. These as harvested 2D metal

oxide layers can be transformed via post treatments which expands this method to other

2D metal compounds.25 For example, touch printed gallium oxides were exposed to

phosphoric acid at elevated temperatures to produce highly crystalline 2D GaPO 4.28

Similarly, gallium oxide sheets can go through sulphuration or nitridation processes and

transform into 2D GaS94, Ga2S326 (Figure 2.8A-C) and GaN27 (Figure 2.8D-F),
37
respectively. Similarly, using liquid indium induced indium oxide (In 2O3) as precursor,

InN27 and In2OxS3-x 95 could be obtained.

Figure 2.7 (A) Gibbs free energy of formation for selected metal oxides. Oxides to the

right of the red dashed line are expected to dominate the surface of gallium based liquid

metals. (B) A cross-sectional diagram of a liquid metal droplet, with possible crystal

structures of thin layers of HfO2, Al2O3, and Gd2O3 as indicated. (C) TEM images, high

resolution TEM images, and SAED patterns of liquid alloy induced Ga2O3, HfO2, Al2O3,

and Gd2O3 sheets. Reprinted with permission from Zavabeti et al., Science 358, 332

(2017). Copyright 2017, The American Association for the Advancement of Science.24

38
In addition to post treatment of the initial oxide skin of liquid metal, 2D metal compounds

could be directly synthesised via exposing liquid metal to reactive gases other than

oxygen such as hydrogen sulfide (H2S), which further expend the applicability of this

method. For instance, 2D tin sulfide (SnS) could be synthesised by exposing liquid tin to

H2S gas.96, 97 (Figure 2.8G and H) Concentration of H2S (even at parts per million levels),

flow rate and temperature were carefully controlled to ensure the purity and uniformity

of the SnS sheets.

2.3.2 2D materials synthesis with liquid metals as reducing agents

It was discussed that surface oxidation of liquid metal provides a facile and versatile

approach for 2D materials synthesis. However, this route is limited to post transition metal

compounds, or compounds of metals that are soluble in gallium-based liquid metals at

low temperature, whose oxidation is thermodynamically favourable. An alternative

approach is to bring appropriate liquid precursors in contact with liquid metal surfaces.

In this case, with reactive liquid metals, as reducing agents, galvanic replacement

reactions occur at the interface, and deposit the desired materials on the surface of liquid

metal. Standard electrode potentials of redox couples can be used for designing the

interfacial redox system and predicting the possible interfacial products. In this way, a

39
Figure 2.8 Synthesis of metal sulfide and metal nitride via liquid metal processes. (A)

Schematic illustration of the sulphuration process of gallium oxide in a tube furnace. (B)

Optical microscopy image of gallium sulfide sheet on a Si-SiO2 wafer. (C) High-

resolution TEM image of the resultant Ga2S3 showing the d-spacing of 0.319 nm which

agrees with the (020) plane of 2D-Ga2S3. Reprinted with permission from Zavabeti et al.,

ACS Appl. Mater. Interfaces 2, 4665 (2019). Copyright 2019, American Chemical

Society.26 (D) Synthesis process of the 2D GaN nanosheet from 2D Ga 2O3 using

ammonolysis. (E) Optical microscopy image of the synthesised GaN on a Si-SiO 2 wafer.

(F) TEM micrograph of a GaN thin film. The top inserted represents the lattice fringes,

and the lower inserted shows the FFT patterns. Reprinted with permission from Syed et

al., J. Am. Chem. Soc. 141, 104 (2019). Copyright 2019, American Chemical Society.27

(G) Schematic depiction of the synthesis process, monolayer of SnS and its application

on a nanogenerator transducer. (H) The high-resolution TEM images of the SnS sheet

with the SAED pattern inserted. Reprinted with permission from Khan et al., Nat.

Commun. 11, 3449 (2020). Copyright 2020, Springer Nature 96

40
wider selection of materials could be templated at liquid metal surface, including

compounds of metals which are not soluble in Ga at low temperature 34, 35 or metals oxides

whose formation is not as thermodynamically preferable in comparison to gallium

oxide,33 or even non-metal-based materials like carbon.32, 98, 99 Due to the existence of

native oxide skin, liquid metal droplets need to be washed with acid or alkaline solutions

to expose their fresh and reactive surface first. In terms of extraction and transfer of the

resultant materials, similar touch transfer techniques could be used for extraction of thin

layers of the deposited materials. Alternatively, in some reaction systems, the deposited

nanosheets could delaminate themselves from the liquid metal surface and be collected

directly from the solution around the liquid metal droplet. It is noteworthy that the

morphologies of the deposited materials on the surface vary greatly with the reaction

systems, as well as the reaction conditions (temperature, pH) and duration of the reactions.

In many cases, short reaction time leads to initial growth of a uniform layer of target

materials of a few nanometres thick templated on the ultra-smooth surface of liquid metal.

With longer reaction time, it results in either continuous growth and thicker smooth

sheets34 or secondary growth and nanostructures.33

In an example, it has been demonstrated that aqueous solution of potassium permanganate

could be used for deposition of manganese oxide film on the surface of EGaIn droplet

(Figure 2.9).36 Permanganate anions are considered a strong oxidizer (E0[MnO42-/MnO2]

= 1.697 V versus standard hydrogen electrode (SHE). In contact with gallium based liquid

metal, the elemental gallium at the interface is oxidized into gallium cations and serves

as a reducing agent (E0[Ga3+/Ga0] = -0.529 V (vs. SHE) to transform permanganate anions

into solid MnO2 at the interface. Due to the non-polar nature of liquid metal and the fact

that the produced hydroxide ion can dissolve gallium oxide and hydroxide, the resultant

hydrated MnO2 sheets go through a spontaneous self-delamination. In another example,

41
the deposition of molybdenum dioxide (MoO2) on the surface of liquid metal with the

solution of sodium molybdate serves as a precursor which leads to deposition of hydrated

MoO2 layers on the surface of EGaIn droplets. In this system, the increased reaction time

leads to continuous growth and increased thickness of the resultant hydrated molybdenum

dioxide (H2MoO3) sheets that remain smooth and uniform. Utilising the touch-printing

process, the deposited sheets could be transferred onto various substrates. Here the

deposited amorphous molybdenum hydroxide sheets were dehydrated and crystallized via

thermal or laser annealing and transformed into conductive MoO 2.34

Other than galvanic replacement, surface of liquid metals can serve as templates for non-

redox depositions as well. For example, when copper sulfate (CuSO4) solution is used as

precursor, possible deposition products on liquid metal surface are cuprous oxides (Cu 2O)

and cupric oxide (CuO), which are products of galvanic replacement and non-redox

deposition, respectively.33 Ratio of two products in the deposited film varies with reaction

conditions like pH and concentration of the precursor, and careful control of these

conditions could result in deposition of pure CuO.

Figure 2.9 Liquid metal-assisted synthesis of 2D materials with aqueous solution as

precursors. (A) TEM image (B) High-resolution TEM image and (C) AFM image of the

resultant MnO2 nanosheets. Reprinted with permission from Ghasemian et al., Adv. Funct.

Mater. 29, 1901649 (2019). Copyright 2020, John Wiley and Sons. 36
42
Other than utilising the intrinsic built-in potential at the surface of liquid metal, an

external voltage could be applied to induce interfacial redox reactions. For instance, 2D

graphitic sheets were templated on the surface of liquid alloy by the voltage-induced

dissociation and reconstruction of simple organic precursors at the interfaces of liquid

metals.32

In addition to inducing redox reaction, an external voltage applied on the liquid metal

droplet can also induce changes in its surface tension according to Lippman’s equation

1
0
− = C (V − V0 )2 . where V is the applied potential, V0 in the intrinsic voltage at the
2

interface, C is the surface capacitance per area and 0

is the intrinsic surface tension when

no external voltage is applied. Therefore, applying an alternating voltage to a liquid metal

droplet immersed in electrolytic solution can initiate rhythmic pulses that mimic the

heartbeat, and be used for nano materials synthesis at the surface of liquid metal. (Figure

2.10A) For example, when rhythmic pulses were induced in a Ga-Zn droplet by a square

wave voltage signal, the oxidizing potential flattens the liquid metal droplet and helps

with the formation of zinc oxide materials on its surface, and the reducing potential

induces constriction of the liquid metal droplet and helps exfoliating zinc oxides from its

interface.100 In this way, the liquid metal droplet works as a continuous reaction medium

to produce zinc oxide nanoflakes. (Figure 2.10B and C)

Most recently, Mayyas et al. has observed that the modulation of the interfacial tension

of the liquid-liquid interfaces of liquid alloys (e.g., gallium–bismuth binary liquid alloy)

can trigger phase separation in the alloy where solute metals (i.e., bismuth) precipitate as

solid nanoclusters from the interface.101 Here, the interfacial tension was modulated via

electrocapillary, and the bismuth precipitation was triggered at cathodic potentials more

negative than -2.2 V versus saturated calomel electrode (SCE). Adjusting the electrolyte

43
chemistry could obtain different types of bismuth-based nanostructures including bismuth

metal nanostructures and several bismuth oxides such as monoclinic (α), tetragonal (β),

and body-centered cubic (γ) Bi2O3. Several morphologies were obtained including

nanotubular, atomically thin plates, and sea-urchin-shaped bismuth oxide. In particular,

smooth γ-Bi2O3 nanosheets were produced in a 0.10 M NaOH-1.0 M KCl aqueous system

at -3.0 V (vs. SCE) and 80 °C. (Figure 2.10D-H)

Figure 2.10 Synthesis of nanoflakes and nanosheets via external square wave potential-

induced pulsing of a liquid metal droplet. (A) GaSn droplet behaviour under a polarizing

voltage signal of ±5.0 V in 0.50 M NaOH. (B) Graphical representation of the ZnO

formation and its ejection from the surface of the Ga-Zn under a square wave potential.

(C) The high-resolution TEM image of the resultant ZnO nanoflakes obtained under a

square wave potential. Reprinted with permission from Mayyas et al., ACS Nano 14,

14070 (2020). Copyright 2020, American Chemical Society.100 (D, E) TEM images of

Bi2O3 nanosheets at different magnifications. (F) High-resolution TEM image featuring

the lattice planes (220) and (200) of γ-Bi2O3 (scale bar: 1.0 nm). (G) FFT pattern showing
44
the main crystallographic planes of the γ-Bi2O3 (scale bar: 5 1/nm). (H) AFM analysis

and thickness profile of the 2D γ-Bi2O3 (scale bar: 0.5 µm). Reprinted with permission

from Mayyas et al., Adv. Funct. Mater. 32, 2108673 (2021). Copyright 2020, John Wiley

and Sons.101

In summary, galvanic replacement and deposition at the surface of liquid metals provide

opportunities for facile synthesis of various nanomaterials, yet deposition of 2D materials

at liquid metal-precursor solution interface is still in its infancy. It should also be

considered that in some cases, the deposition on the liquid metal surface starts with crystal

nucleus and no uniform thin sheet forms.102 Although one could gain some degree of

morphology control via tunning the pH of precursor or other reaction conditions, the

mechanism of formation of different morphologies have yet to be studied, and precise

control of the morphology of deposited films is yet to be achieved in most of the systems,

which requires further exploration.

2.3.3 Deposition of 2D layers based on liquid metal processes

Due to the non-polar nature of liquid metals, the templated 2D materials on the surface

of liquid metals weakly adhere to the parent metal, which makes the exfoliation possible.

In the case of exfoliating 2D metal compounds synthesised at liquid metal-gas interfaces,

like metal oxide synthesised by exposing liquid metal to air, various exfoliation and

transfer methods have been developed. Squeeze-printing, touch-printing, or rolling the

liquid metal droplets on the substrates can lead to centimetre scales deposition of 2D

layers on smooth substrates, while bubbling gas into liquid metals can yield 2D flakes at

large quantities. In the case of 2D materials synthesised at the liquid metal-aqueous

45
solution interfaces, the resultant 2D materials either go through a self-exfoliation process

or could be touch-transferred.

Touch transferring process is commonly used for large area transferring of the metal oxide
24
layers onto desired substrates, which was schematically illustrated in Figure 2.11A.

When the smooth substrate is brought into contact with the metal oxide layer, the liquid

metal underneath acts as a soft cushion, and an atomic-scale contact between the metal

oxide layer and the substrate could be realized. The close contact induces accumulative

van der Waal interactions between the permanent dipoles of metal oxide layer and the

substrate, which leads to adhesion between them. On the other hand, due to the non-

polarity nature of liquid metal, only weak forces are expected between the metal and its

interfacial oxide layer, which means the bulk liquid metal could easily separate from the

metal oxide layers upon lift-off with the metal oxide layers remains on the substrates. It

is noteworthy that this touch-transfer technique is sensitive to morphology but material

independent. As long as the substrates and surface layers are smooth, which is favourable

for an atomic-scale contact over a large area34 between them, large area extraction and

transferring could be achieved on various substrates (Si wafer, quartz,

polydimethylsiloxane) and planar materials (Ga2O3, ZnO, PbO). For obtaining high

control and enhancing the quality of the harvested 2D layers, program-controlled piezo-

electric stages can be utilised for the touch-transfer process since they provide more

stability and reproducibility than human hands (Figure 2.11B). 35

Similar to the previously discussed touch-printing method, rolling a liquid metal droplet

(Figure 2.11C) on a substrate or squeeze printing (Figure 2.11D) allows deposition of the

metal oxide layer at centimetre scales. Squeeze printing involves placing a small liquid

metal droplet in between the desired substrate and squeezing them to expend the liquid

metal droplet in the middle. When the Cabrera-Mott kinetics governs the oxidation of the
46
liquid metal, the initial oxidation layer grows fast as the liquid metal droplet expends

upon squeezing, which allows a large area deposition of uniform and ultrathin metal oxide

layer. Utilising this method, conductive ITO was deposited on various substrates. 38

The aforementioned transfer processes are versatile and not limited to monolayer metal

oxides. Multiple touch-prints could precisely control the material thickness. 38

Additionally. Deposit one metal oxide sheet on the top of another can lead to 2D van der

Waals heterostructures.103 With patterning and functionalization of the substrate,

deposition of metal oxides on selective areas is possible, which can be utilised for

electronic device fabrication. Functionalization of the substrates, which tunes the surface

energy, offers control of the surface adhesion and transferring process. For example,

selective area deposition of Ga2O3 can be achieved by functionalizing the Si-SiO2 wafer

with per-fluorinated silanes at selective areas using standard photolithography. Silane

decoration of Si-SiO2 wafer is known to decrease its surface energy and make it less

adhesive. In this way, Ga2O3 could be selectively deposited onto the non-functionalized

SiO2 surface according to a photolithographic pattern. 94

An alternative approach to exfoliate surface oxide layers of liquid metal involves

bubbling air. This method can be used in applications such as catalysis and energy

storage, where a high yield of the products is needed. As shown in Figure 2.11E, a solvent

is placed on top of liquid metal. As the air is injected into liquid metal, an oxide layer

rapidly forms at the liquid metal-air interface. The air bubbles raise in the bulk liquid

metal due to gravity gradient, carrying the oxide layers through the liquid metal-solvent

interface, bursting and leaving the metal oxide sheets as a colloidal suspension. 24

Interestingly, the morphology and composition of metal oxide sheets can be tuned by

altering the pH or composition of the solvent at the top.104, 105

47
In some cases, where a reactive metal compound is alloyed with a room temperature

liquid metal, aqueous solvent can serve as an oxidizer leading to the exfoliation of the 2D

flakes by generating H2 gas. When Galinstan-aluminium alloy is placed in water,

aluminium reacts with water at the interface, producing aluminium hydroxide (Al(OH) 3)

and H2 gas.37 The resultant Al(OH)3 nanosheets are exfoliated from liquid metal surface

by the generated H2 and transformed into boehmite (γ-AlOOH).

Figure 2.11 Various extraction and transfer process of the oxide skins of liquid metals:

(A) schematic representation of the touch-transfer technique where a suitable substrate is

brought into contact with the oxide surface-bearing liquid metal droplet. Reprinted with

permission from Zavabeti et al., Science 358, 332 (2017). Copyright 2017, The American

48
Association for the Advancement of Science.24 (B) touch-transfer technique conducted

with a program-controlled piezoelectric stage. Reprinted with permission from Mayyas

et al., Adv. Mater. 32, 2001997 (2020). Copyright 2020, John Wiley and Sons.32 (C)

deposition of TeO2 by rolling a droplet of Te-Se alloy on the substrate. Reprinted with

permission from Zavabeti et al., Nat. Electron. 4, 277 (2021). Copyright 2017, Springer

Nature.106 (D) depositing ITO via squeeze-printing. Reprinted with permission from Datta

et al., Nat. Electron. 3, 51 (2020). Copyright 2020, Springer Nature.38 and (E) exfoliating

metal oxide layer via gas injection. (F) Photographs of the bubble carrying the metal oxide

surface bursting through the liquid metal-water interface. (G) An optical image of the

resultant metal-oxide sheets drop-casted onto a Si-SiO2 wafer (right). Reprinted with

permission from Zavabeti et al., Science 358, 332 (2017). Copyright 2017, The American

Association for the Advancement of Science.24

2.4 Application in devices

We learned that liquid metals provide new routes for low-temperature, facile synthesis of

uniform, large area, and ultrathin 2D materials. Using this approach, materials have been

synthesised in a stable 2D form that were otherwise considered challenging with other

methods.26 In this section, we summarize the utilisation of liquid metal-assisted

deposition of 2D materials in electronic and piezoelectric devices. liquid metals-induced

2D semiconductors were used as channel materials for field effect transistors (FETs),

photo-responsive materials in heterostructure photodetectors, or piezoelectric responsive

materials. Insulating and conductive 2D materials were also harnessed from the liquid

metal surface for dielectric and conductive units in electronic applications.

49
2.4.1 Optoelectronic devices

Post transition metal compounds harnessed from the surface of liquid metals are uniform,

large area, and easily transferable, making them suitable channel materials in thin-film

FETs. Those metal compounds include oxides that naturally form in ambient or controlled

environments, as well as post-treated sulfides or nitrides. Liquid metal compounds

provide the opportunity for both p- and n-type channel materials suitable for different

applications. Room temperature deposited p-type semiconducting -Ga2O3 sheets were

used for the fabrication of FETs that show on/off ratio exceeding 10 4 at 10 V back gate

voltage.107 (Figure 2.12A and B) Fu et al. reported a liquid gallium-induced 2D GaN with

high carrier mobility of 160 cm-2V-1s-1, based on which the FETs show an on/off ratio of

106.108

Liquid metal-induced 2D sheets can also serve as photo-responsive materials of

optoelectronic devices including FETs, and also two-terminals and heterostructural

photodetectors. (Figure 2.12C and D) With bandgaps ranging from deep UV to near-IR,

photodetectors made from 2D materials offer light detectivity in a broad range. Liquid

metal-induced indium oxide sheets were partially sulfurized in to In2O3-xSx, which is a n-

type semiconductor with carrier mobility as high as 44 cm2V-1s-1. The fabricated

phototransistors showed a photo-responsibility of 3400 AW -1 and a detectivity of

2.18×1013 at a wavelength of 285 nm and illumination intensity of 2 mWcm -2.95

Previously reported electronic devices based on liquid-metal induced 2D semiconductors

are summarized in Table 2.1.

50
2.4.2 Dielectric/ conducting units and bandgap manipulation

In addition to semiconductors, dielectric and conductive 2D materials can also be

synthesised via liquid metal surface-based processes. In electronic and optoelectronic

applications, dielectrics are generally necessary for gating and insulating purposes. The

pursuit of highly conductive and transparent layers lies in the core of fabricating displays

and touch screen appliances. Post transition liquid metals serve as reactive media for

diluting various metals and provide ultra-smooth interfaces for the formation of ultra-thin

metal compounds for the aforementioned applications.

The centimetre-scale synthesis of 2D ITO films was successfully reported using a liquid

metal-assisted process.38 (Figure 2.12E and F) For indium-tin liquid alloy, the formation

of ternary ITO is more thermodynamically favourable compared to pure In 2O3 and SnO2.

Squeeze printing indium-tin alloy allows centimetre-scale deposition of indium tin oxide

sheet on various substrates including the flexible polymer substrate. The deposited bilayer

ITO sheets offered 99.3% transparency and a sheet resistance as low as 5.4 kΩsq-1 and

provided high flexibility with minimal conductivity loss even after 1000 times cyclic

bending. A capacitive touch screen device operated on such centimetre-sized 2D ITO

nanosheets with confirmed performance.

Liquid metal assisted deposition technique can also be utilised for deposition of ultrathin

dielectric layers. Ultrathin insulating sheets of HfO2 or Sb2O3 were synthesised via

surface oxidation of liquid alloys, followed by a touch-transferring technique. 24, 109 A

gallium-based liquid alloy containing the Hf element was employed to deposit 0.5 nm

thick HfO2 sheets on desired substrates, which is the thinnest ever reported HfO2.24 A

peak force tunneling microscopy was utilised for conductivity measurement with Pt-

51
coated wafer as substrates. The deposited HfO2 sheet showed a high dielectric constant

of 39.

In another relevant work, Sb2O3 sheets were deposited by rolling an Sn-based liquid metal

droplet that contained a small amount of Sb element on the substrate. 109 The obtained α-

Sb2O3 exhibited high crystallinity and a wide band gap of ∼4.4 eV. The relative

permittivity assessment revealed a maximum dielectric constant of 84, while a breakdown

electric field of ∼10 MVcm-1 was observed (Figure 2.12G and H).

Deposition of 2D materials using the surface oxidation of liquid metal is not only facile,

but also highly versatile and tuneable. The composition and properties of the resulting 2D

materials could be easily tuned by varying the mixture of liquid alloy or deposition

conditions. For example, Bi-Sn liquid alloy was used for the deposition of Bi2O3 doped

SnO sheet.76 Since the formation of SnO is thermodynamically favourable, Bi

concentration was no more than approximately ~0.7 at% in the surface oxide layer even

when the Bi ratios in the bulk alloy were as high as 90%. By tunning the ratio of Bi in the

alloy, the doping concentration of Bi could be tunned between 0.02 at% to 0.73 at%, and

the bandgap of resultant metal oxide sheets varied between 4.2 to 2.7 eV.

52
Table 2.1 Summery of key features of electronic devices based on liquid-metal fabricated

2D semiconductors. Max.: Maximum value. Ave.: Average value.

Photo-
Band Carrier
Structure On/off responsivit Detectivity
Device Materials gap mobility
type ratio y D* (Jones)
(eV) (cm-2V-1s-1) -1
(R, AW )
104,
Max. 21.3
FET β‐Ga2O3 107 p-type 4.9 VG=10 - -
Ave. 5- 10
V
FET GaN 108 n-type NM 160 106 - -
Max. 232
106
FET β‐TeO2 p-type 3.7 Ave. 106 - -
146±42
102
FET Ga2S3 26 p-type 2.1 3.5 240 1010
VG=1V
Ave.
95
FET In2O3-xSx n-type 3.33 20.4±6.3 60 3.4×103 2.18×1013
Max. 44
Ave.
FET GaS 94 - 3.02 0.2 Ave. 6.4 -
170
FET SnO 87 p-type 4.2 0.7 20 - -
Two-terminal
Bi2O3 88 n-type 3.5 - - 400 1.1×1013
photodetector
Two-terminal Monolayer
p-type 1.81 35 - 9.2 × 102 1.09×109
photodetector SnS 97
Two-terminal Multilayer
p-type 1.41 35 - 3.51 × 103 6.83×1010
photodetector SnS 97
Heterostructure In2O3 103 n-type 3.65
37 - 1047 5×109
photodetector SnO 103
p-type 4.08
Heterostructure
Ga2O3 110 p-type 4.8 - - 44.6 3.4×1013
photodetector

53
2.4.3 Piezoelectric devices

By the thickness reduction within 2D scale, piezoelectricity might be induced in non-

piezoelectric materials. This feature in planar structures originates from the loss of

centrosymmetry, change in local polarization, and modulation of the free carrier

concentrations, and can also be tailored by surface modifications. 111 Out of the thirty-two

crystal sub-classes, in four main crystal classes of rectangular, cubic, hexagonal, and

oblique, the inversion symmetry can be broken by thinning down from that of bulk into

2D.112 Additionally, the broken inversion symmetry and piezoelectricity could exist in

layered materials with an odd number of layers due to the non-zero odd-rank tensor

properties.113 In addition, boundary changes were also demonstrated to enhance the

piezoelectricity by altering the symmetry.92

Many 2D materials with piezoelectric properties can be prepared feasibly through the

liquid metal approaches, while are not accessible using conventional methods. 111 Some

piezoelectric 2D materials, such as GaPO4, are not achievable by conventional exfoliation

but can be obtained in a two-step process of harvesting 2D Ga2O3 first and then

transforming it into GaPO4 via a vapor phase reaction. Employing piezoresponse force

microscopy (PFM) a vertical piezoelectric coefficient of 7.5 pm V −1 was measured for the

unit cell thick GaPO4 nanosheets.28 Liquid metal based synthesis has also been used for

preparing a range of semiconducting 2D piezoelectric materials for piezotronics

applications such as SnS,96 PbO 91 and ZnO 92. As demonstrated by Ghasemian et al. 91,

the health hazard and environmental pollution of lead-based piezoelectric materials, such

as Pb(Zr1−xTix)O3 (PZT), can be mitigated by size reduction to monolayer through the

liquid metal synthesis. PbO monolayers collected from the surface of molten Pb showed

vertical and lateral piezoresponses of 29.6 and 3.9 pmV-1, respectively, and a particular

device fabricated by these ultra-thin PbO nanosheets exhibited a significant voltage

54
output of ∼100 mV peak value, as presented in Figure 2.12I and J. The Pb content in a

device fabricated by this method is 104 times less than commercial PZT thin films of

comparable performance, while posing significantly less environmental contamination.

The maximum piezoelectricity in ultra-thin sheets has been reported for ultra-thin ZnO

sheets on SiO2 substrates obtained by a self-limiting approach at the metal-melt/air

interface.92 ZnO nanosheets as thin as 1.1 nm (corresponding 2.5 unit-cell) delivered an

out of plane (d33) value of 80 pmV-1 which is approximately 8 times larger than that of

the value for bulk ZnO. These findings demonstrated that the post transition liquid metals

can be used feasibly for creating large piezo-responsive 2D materials as building blocks

of devices to harvest the kinetic energy of various mechanical vibrations.

55
Figure 2.12 Liquid metal-based fabrication of 2D materials for electronic and

piezoelectric applications. (A). Schematic illustration of a β-Ga2O3 FET with Ag contacts.

(B) Ids–Vds output curves of β-Ga2O3 FET at different gate voltages (Vg). Reprinted with

permission from Lin et al., Phys. Status Solidi Rapid Res. Lett. 13, 1900271 (2019).

Copyright 2019, John Wiley and Sons.107 (C) Schematic illustration of the 2D touch-

transfer technique of SnO/In2O3 heterostructure from liquid metals. Crystal structure of

SnO exhibits trigonal symmetry while In2O3 sheets exhibit cubic crystal structure. (D)

The Ids–Vds characteristic curves of the 2D SnO/In2O3 photodetector on linear scales

under dark condition and light illumination (280, 365, and 455 nm). Inset: schematic

drawing of the SnO/In2O3 hetero-structural photodetector. Reprinted with permission

from Alsaif et al., Adv. Mater. Interfaces 6, 1900007 (2019). Copyright 2020, John Wiley

and Sons.103 (E) Schematic of an ITO and polymer-based flexible two-terminal resistive

test device, with the inset showing the ITO/polymer interface. (F) Schematic of the

resultant device. Reprinted with permission from Datta et al., Nat. Electron. 3, 51 (2020).

Copyright 2020, Springer Nature.38 (G) Schematics of conductive AFM measurement. (H)

I-V curve measured through a point on the α-Sb2O3 layer. Inset: data fitted to the Schottky

emission model showing a calculated dielectric constant of around 84. Reprinted with

permission from Messalea et al., ACS Nano 15, 16067 (2021). Copyright 2020, Royal

Society of Chemistry.109 (I) The optical images of the two-electrode device in which PbO

monolayer is embedded between two Au electrodes. (J) The output voltage response of

the two-electrode PbO device. Reprinted with permission from Mahmood et al., J. Mater.

Chem. A 8, 19434 (2020). Copyright 2020, American Chemical Society. 91

56
2.4.4 Sensors

A brief discussion is presented in this section on the development of sensors using the

liquid metal induced 2D processes. While full discussions regarding photoresistors and

phototransistors, and also piezoelectric sensors were presented as potential applications

for the synthesised 2D based devices, there are other possibilities that are offered for

developing sensors. One is regarding the development of different chemical sensors,

biosensors, and gas sensors using 2D materials derived from liquid metals.

In one of the studies, Wang et al. have shown that 2D graphene oxide flakes can be

reduced on the surface of liquid metals, forming a core-shell structure of 2D reduced

graphene oxide /liquid metal.114 Such a technique later was implemented by Baharfar et

al. to reduce graphene oxide flakes on liquid metal droplets for the development of

selective electrochemical biosensors for dopamine (DA) measurement, 115 providing a

modified electrochemical interface for separating voltametric peaks of DA, uric acid

(UA), and ascorbic acid (AA). This selective peak separation was attributed to the

repulsion of UA by negatively charged moieties on 2D sheets and reductive effect of AA

on coordinated Ga species. The method was practical, and they could establish the sensors

on paper substrates. Similarly, hexagonal molybdenum oxide (h-MoO 3) nanoflakes and

liquid metal-based core-shell structured nanomaterials were demonstrated for optical

sensing application.116 In this case, 2D flakes h-MoO3 were deposited on liquid metal

droplets, which induced reduction of h-MoO3 into sub-stoichiometric ratios and

formation of abundant oxygen vacancies. These induced events effectively increased the

concentration of charge carriers and provided a broad surface plasmon resonance (SPR)

well-suited for SERS-based sensing. In addition to biosensors, liquid metal-derived 2D

materials have been exploited for designing gas sensors. The liquid metal derived 2D SnO

sheets with tuneable level of Bi2O3 dopant, altering the bandgap, were applied for
57
selectivity sensing of gaseous species. 76 Upon the adsorption of H2 and NO2 on SnO

sheets doped with different levels of Bi2O3, different patterns of selective conductivity

were obtained.

2.5 Conclusion

In this chapter, the author discussed the fundamentals behind the intrinsic properties of

liquid metal including their low melting point, high thermal and electrical conductivities.

The author also demonstrated the surface layering and surface oxidation phenomenon of

liquid metals, which are essential for templating 2D materials. Afterwards, the author

presented an overview of liquid metal-assisted synthetic methods of 2D materials,

including the surface oxidation, reduction, and exfoliation of liquid metals. Finally,

application of liquid metal-induced 2D materials in various devices like field effect

transistors, optoelectronic devices and piezoelectric devices are also presented.

2.6 Reference

1. Jensen, W. B., The place of zinc, cadmium, and mercury in the periodic table. J.

Chem. Educ. 2003, 80 (8), 952.

2. Masterton, W. L.; Hurley, C. N., Chemistry: principles and reactions. Cengage

Learning: 2015.

3. Atkins, P.; De Paula, J., Physical chemistry for the life sciences. Oxford

University Press, USA: 2011.

4. Greenwood, N. N.; Earnshaw, A., Chemistry of the Elements. Elsevier: 2012.

58
5. Brown, T. L., The chemistry of metallic elements in the ionosphere and

mesosphere. Chem. Rev. 1973, 73 (6), 645-667.

6. Johnson, O., Role of f electrons in chemical binding. J. Chem. Educ. 1970, 47 (6),

431.

7. Lienert, T. J.; Babu, S. S.; Acoff, V. L.; Zhou, N. Y.; DeGuire, E.; Lampman,

S.; Marquard, E.; Musgrove, B.; Riley, B., ASM handbook. ASM International

Materials Park, Ohio: 2011; Vol. 3.

8. Sostman, H. E., Melting point of gallium as a temperature calibration standard.

Rev. Sci. Instrum. 1977, 48 (2), 127-130.

9. Shalom, S. B.; Kim, H. G.; Emuna, M.; Argaman, U.; Greenberg, Y.; Lee, J.;

Yahel, E.; Makov, G., Anomalous pressure dependent phase diagram of liquid

Ga–In alloys. J. Alloys Compd. 2020, 822, 153537.

10. Daeneke, T.; Khoshmanesh, K.; Mahmood, N.; de Castro, I. A.; Esrafilzadeh,

D.; Barrow, S. J.; Dickey, M. D.; Kalantar-Zadeh, K., Liquid metals:

fundamentals and applications in chemistry. Chem. Soc. Rev. 2018, 47 (11), 4073-

4111.

11. Allioux, F.-M.; Ghasemian, M. B.; Xie, W.; O'Mullane, A. P.; Daeneke, T.;

Dickey, M. D.; Kalantar-Zadeh, K., Applications of liquid metals in

nanotechnology. Nanoscale Horiz. 2022.7, 141-167

12. Jalili, A. R.; Satalov, A.; Nazari, S.; Rahmat Suryanto, B. H.; Sun, J.;

Ghasemian, M. B.; Mayyas, M.; Kandjani, A. E.; Sabri, Y. M.; Mayes, E., Liquid

Crystal-Mediated 3D Printing Process to Fabricate Nano-Ordered Layered

Structures. ACS Appl. Mater. Interfaces 2021, 13 (24), 28627-28638.

13. Allioux, F.-M.; Merhebi, S.; Ghasemian, M. B.; Tang, J.; Merenda, A.; Abbasi,

R.; Mayyas, M.; Daeneke, T.; O’Mullane, A. P.; Daiyan, R., Bi–Sn catalytic

59
 foam governed by nanometallurgy of liquid metals. Nano Lett. 2020, 20 (6), 4403-

4409.

14. Han, J.; Mayyas, M.; Tang, J.; Mousavi, M.; Idrus-Saidi, S. A.; Cai, S.; Cao,

Z.; Wang, Y.; Tang, J.; Jalili, R., Liquid metal enabled continuous flow reactor:

A proof-of-concept. Matter 2021, 4 (12), 4022-4041.

15. Zhao, S.; Zhang, J.; Fu, L., Liquid metals: A novel possibility of fabricating 2D

metal oxides. Adv. Mater. 2021, 33 (9), 2005544.

16. D’Evelyn, M. P.; Rice, S. A., A pseudoatom theory for the liquid–vapor interface

of simple metals: Computer simulation studies of sodium and cesium. J. Chem.

Phys. 1983, 78 (8), 5225-5249.

17. D’Evelyn, M. P.; Rice, S. A., A study of the liquid–vapor interface of mercury:

Computer simulation results. J. Chem. Phys. 1983, 78 (8), 5081-5095.

18. Magnussen, O.; Ocko, B.; Regan, M.; Penanen, K.; Pershan, P. S.; Deutsch, M.,

X-ray reflectivity measurements of surface layering in liquid mercury. Phys. Rev.

Lett. 1995, 74 (22), 4444.

19. Regan, M.; Kawamoto, E.; Lee, S.; Pershan, P. S.; Maskil, N.; Deutsch, M.;

Magnussen, O.; Ocko, B.; Berman, L., Surface layering in liquid gallium: An X-

ray reflectivity study. Phys. Rev. Lett. 1995, 75 (13), 2498.

20. Pershan, P. S., Review of the highlights of X-ray studies of liquid metal surfaces.

J. Appl. Phys. 2014, 116 (22), 222201.

21. Regan, M.; Pershan, P. S.; Magnussen, O.; Ocko, B.; Deutsch, M.; Berman, L.,

X-ray reflectivity studies of liquid metal and alloy surfaces. Phys. Rev. B 1997,

55 (23), 15874.

60
22. Xu, Q.; Oudalov, N.; Guo, Q. T.; Jaeger, H. M.; Brown, E., Effect of oxidation

on the mechanical properties of liquid gallium and eutectic gallium-indium. Phys.

Fluids 2012, 24 (6), 063101.

23. Akbari, M. K.; Verpoort, F.; Zhuiykov, S., State-of-the-art surface oxide

semiconductors of liquid metals: an emerging platform for development of

multifunctional two-dimensional materials. J. Mater. Chem. A 2021, 9, 34-73.

24. Zavabeti, A.; Ou, J. Z.; Carey, B. J.; Syed, N.; Orrell-Trigg, R.; Mayes, E. L.

H.; Xu, C. L.; Kavehei, O.; O'Mullane, A. P.; Kaner, R. B.; Kalantar-Zadeh,

K.; Daeneke, T., A liquid metal reaction environment for the room-temperature

synthesis of atomically thin metal oxides. Science 2017, 358 (6361), 332-335.

25. Zeng, M.; Chen, Y.; Zhang, E.; Li, J.; Mendes, R. G.; Sang, X.; Luo, S.; Ming,

W.; Fu, Y.; Du, M.-H., Molecular scaffold growth of two-dimensional, strong

interlayer-bonding-layered materials. CCS Chem. 2019, 1 (1), 117-127.

26. Alsaif, M. M.; Pillai, N.; Kuriakose, S.; Walia, S.; Jannat, A.; Xu, K.; Alkathiri,

T.; Mohiuddin, M.; Daeneke, T.; Kalantar-Zadeh, K., Atomically thin Ga2S3

from skin of liquid metals for electrical, optical, and sensing applications. ACS

Appl. Mater. Interfaces 2019, 2 (7), 4665-4672.

27. Syed, N.; Zavabeti, A.; Messalea, K. A.; Della Gaspera, E.; Elbourne, A.;

Jannat, A.; Mohiuddin, M.; Zhang, B. Y.; Zheng, G. L.; Wang, L.; Russo, S.

P.; Esrafilzadeh, D.; McConville, C. F.; Kalantar-Zadeh, K.; Daeneke, T.,

Wafer-sized ultrathin gallium and indium nitride nanosheets through the

ammonolysis of liquid metal derived oxides. J. Am. Chem. Soc. 2019, 141 (1),

104-108.

28. Syed, N.; Zavabeti, A.; Ou, J. Z.; Mohiuddin, M.; Pillai, N.; Carey, B. J.;

Zhang, B. Y.; Datta, R. S.; Jannat, A.; Haque, F.; Messalea, K. A.; Xu, C. L.;

61
 Russo, S. P.; McConville, C. F.; Daeneke, T.; Kalantar-Zadeh, K., Printing two-

dimensional gallium phosphate out of liquid metal. Nat. Commun. 2018, 9, 3618.

29. Duval, J. F.; Bera, S.; Michot, L. J.; Daillant, J.; Belloni, L.; Konovalov, O.;

Pontoni, D., X-ray reflectivity at polarized liquid-Hg–aqueous-electrolyte

interface: Challenging macroscopic approaches for ion-specificity issues. Phys.

Rev. Lett. 2012, 108 (20), 206102.

30. Elsen, A.; Murphy, B. M.; Ocko, B. M.; Tamam, L.; Deutsch, M.; Kuzmenko,

I.; Magnussen, O. M., Surface layering at the mercury-electrolyte interface. Phys.

Rev. Lett. 2010, 104 (10), 105501.

31. Shpyrko, O. G.; Streitel, R.; Balagurusamy, V. S.; Grigoriev, A. Y.; Deutsch,

M.; Ocko, B. M.; Meron, M.; Lin, B.; Pershan, P. S., Surface crystallization in

a liquid AuSi alloy. Science 2006, 313 (5783), 77-80.

32. Mayyas, M.; Li, H. Z.; Kumar, P.; Ghasemian, M. B.; Yang, J.; Wang, Y. F.;

Lawes, D. J.; Han, J. L.; Saborio, M. G.; Tang, J. B.; Jalili, R.; Lee, S. H.;

Seong, W. K.; Russo, S. P.; Esrafilzadeh, D.; Daeneke, T.; Kaner, R. B.; Ruoff,

R. S.; Kalantar-Zadeh, K., Liquid-metal-templated synthesis of 2D graphitic

materials at room temperature. Adv. Mater. 2020, 32 (29), 2001997.

33. Li, H. Z.; Abbasi, R.; Wang, Y. F.; Allioux, F. M.; Koshy, P.; Idrus-Saidi, S.

A.; Rahim, M. A.; Yang, J.; Mousavi, M.; Tang, J. B.; Ghasemian, M. B.;

Jalili, R.; Kalantar-Zadeh, K.; Mayyas, M., Liquid metal-supported synthesis of

cupric oxide. J. Mater. Chem. C 2020, 8 (5), 1656-1665.

34. Wang, Y. F.; Mayyas, M.; Yang, J.; Ghasemian, M. B.; Tang, J. B.; Mousavi,

M.; Han, J. L.; Ahmed, M.; Baharfar, M.; Mao, G. Z.; Yao, Y.; Esrafilzadeh,

D.; Cortie, D.; Kalantar-Zadeh, K., Liquid-Metal-Assisted Deposition and

62
 Patterning of Molybdenum Dioxide at Low Temperature. ACS Appl. Mater.

Interfaces 2021, 13 (44), 53181-53193.

35. Wang, Y. F.; Mayyas, M.; Yang, J.; Tang, J. B.; Ghasemian, M. B.; Han, J. L.;

Elbourne, A.; Daeneke, T.; Kaner, R. B.; Kalantar-Zadeh, K., Self-deposition of

2D molybdenum sulfides on liquid metals. Adv. Funct. Mater. 2020, 31 (3),

2005866.

36. Ghasemian, M. B.; Mayyas, M.; Idrus‐Saidi, S. A.; Jamal, M. A.; Yang, J.;

Mofarah, S. S.; Adabifiroozjaei, E.; Tang, J.; Syed, N.; O'Mullane, A. P., Self‐

limiting galvanic growth of MnO2 monolayers on a liquid metal—applied to

photocatalysis. Adv. Funct. Mater. 2019, 29 (36), 1901649.

37. Zavabeti, A.; Zhang, B. Y.; de Castro, I. A.; Ou, J. Z.; Carey, B. J.; Mohiuddin,

M.; Datta, R.; Xu, C.; Mouritz, A. P.; McConville, C. F., Green synthesis of

low‐dimensional aluminum oxide hydroxide and oxide using liquid metal reaction

media: ultrahigh flux membranes. Adv. Funct. Mater. 2018, 28 (44), 1804057.

38. Datta, R. S.; Syed, N.; Zavabeti, A.; Jannat, A.; Mohiuddin, M.;

Rokunuzzaman, M.; Zhang, B. Y.; Rahman, M. A.; Atkin, P.; Messalea, K. A.,

Flexible two-dimensional indium tin oxide fabricated using a liquid metal printing

technique. Nat. Electron. 2020, 3 (1), 51-58.

39. Wang, Y.; Baharfar, M.; Yang, J.; Mayyas, M.; Ghasemian, M. B.; Kalantar-

Zadeh, K., Liquid state of post-transition metals for interfacial synthesis of two-

dimensional materials. Appl. Phys. Rev. 2022, 9 (2), 021306.

40. Speight, J. G., Lange's handbook of chemistry. McGraw-Hill New York: 2005;

Vol. 1.

41. Norrby, L. J., Why is mercury liquid? Or, why do relativistic effects not get into

chemistry textbooks? J. Chem. Educ. 1991, 68 (2), 110.

63
42. Lide, D. R., CRC handbook of chemistry and physics. CRC press: 2004; Vol. 85.

43. Heine, V., Crystal structure of gallium metal. J. Phys. C: Solid State Phys. 1968,

1 (1), 222.

44. Yu, S.; Kaviany, M., Electrical, thermal, and species transport properties of liquid

eutectic Ga-In and Ga-In-Sn from first principles. J. Chem. Phys. 2014, 140 (6),

064303.

45. Lou, H.; Wang, X.; Cao, Q.; Zhang, D.; Zhang, J.; Hu, T.; Mao, H.-k.; Jiang,

J.-Z., Negative expansions of interatomic distances in metallic melts. Proc. Natl.

Acad. Sci. 2013, 110 (25), 10068-10072.

46. Dodd, C., The electrical resistance of liquid gallium in the neighbourhood of its

melting point. Proc. Phys. Soc. B 1950, 63 (9), 662.

47. Handschuh-Wang, S.; Stadler, F. J.; Zhou, X., Critical review on the physical

properties of gallium-based liquid metals and selected pathways for their

alteration. J. Phys. Chem. C 2021, 125 (37), 20113-20142.

48. Mujahid, M.; Engel, N.; Chia, E., Effect of alloying elements on the conductivity

of aluminum alloys. Scr. Metall. 1979, 13 (9), 887-893.

49. Amin, N. A. A. M.; Shnawah, D. A.; Said, S. M.; Sabri, M. F. M.; Arof, H.,

Effect of Ag content and the minor alloying element Fe on the electrical resistivity

of Sn–Ag–Cu solder alloy. J. Alloys Compd. 2014, 599, 114-120.

50. Evans, R., The resistivity and thermopower of liquid mercury and its alloys. J.

Phys. C: Solid State Phys. 1970, 3 (2S), S137.

51. Bakhtiyarov, S.; Overfelt, R., Electrical conductivity measurements in liquid

metals by rotational technique. J. Mater. Sci. 1999, 34 (5), 945-949.

64
52. Bakhtiyarov, S. I.; Overfelt, R. A.; Teodorescu, S. G., Electrical and thermal

conductivity of A319 and A356 aluminum alloys. J. Mater. Sci. 2001, 36 (19),

4643-4648.

53. Mott, N. F., The resistance of liquid metals. Proc. Math. Phys. 1934, 146 (857),

465-472.

54. Powell, R., The electrical resistivity of gallium and some other anisotropic

properties of this metal. Proc. R. Soc. A: Math. Phys. Eng. Sci. 1951, 209 (1099),

525-541.

55. Ho, C. Y.; Powell, R. W.; Liley, P. E., Thermal conductivity of the elements. J.

Phys. Chem. Ref. Data 1972, 1 (2), 279-421.

56. Kahn, A., Electrons and Phonons. The theory of transport phenomena in solids.

Science 1961, 133 (3447), 187-188.

57. Plevachuk, Y.; Sklyarchuk, V.; Eckert, S.; Gerbeth, G.; Novakovic, R.,

Thermophysical properties of the liquid Ga–In–Sn eutectic alloy. J. Chem. Eng.

Data 2014, 59 (3), 757-763.

58. Kittel, C., Introduction to solid state physics. Am. J. Phys. 1967, 35 (6), 547-548.

59. Peralta, M. V.; Dix, M.; Lesemann, M.; Wakeham, W., Thermal conductivity of

liquid mercury. High Temp. High Press. 2002, 34 (1), 35-39.

60. Prokhorenko, V. Y.; Roshchupkin, V. V.; Pokrasin, M. A.; Prokhorenko, S.;

Kotov, V., Liquid gallium: potential uses as a heat-transfer agent. High Temp.

2000, 38 (6), 954-968.

61. Giordanengo, B.; Benazzi, N.; Vinckel, J.; Gasser, J.; Roubi, L., Thermal

conductivity of liquid metals and metallic alloys. J. Non-Cryst. Solids 1999, 250,

377-383.

65
62. Braunsfurth, M.; Skeldon, A.; Juel, A.; Mullin, T.; Riley, D., Free convection in

liquid gallium. J. Fluid Mech. 1997, 342, 295-314.

63. Tostmann, H.; DiMasi, E.; Pershan, P. S.; Ocko, B.; Shpyrko, O.; Deutsch, M.,

Surface structure of liquid metals and the effect of capillary waves: X-ray studies

on liquid indium. Phys. Rev. B 1999, 59 (2), 783.

64. Shpyrko, O.; Huber, P.; Grigoriev, A.; Pershan, P.; Ocko, B.; Tostmann, H.;

Deutsch, M., X-ray study of the liquid potassium surface: Structure and capillary

wave excitations. Phys. Rev. B 2003, 67 (11), 115405.

65. Shpyrko, O. G.; Grigoriev, A. Y.; Steimer, C.; Pershan, P. S.; Lin, B.; Meron,

M.; Graber, T.; Gerbhardt, J.; Ocko, B.; Deutsch, M., Anomalous layering at the

liquid Sn surface. Phys. Rev. B 2004, 70 (22), 224206.

66. Pershan, P. S.; Stoltz, S.; Shpyrko, O. G.; Deutsch, M.; Balagurusamy, V. S.;

Meron, M.; Lin, B.; Streitel, R., Surface structure of liquid Bi and Sn: An X-ray

reflectivity study. Phys. Rev. Lett. 2009, 79 (11), 115417.

67. DiMasi, E.; Tostmann, H.; Shpyrko, O. G.; Deutsch, M.; Pershan, P. S.; Ocko,

B., Surface-induced order in liquid metals and binary alloys. J. Condens. Matter

Phys. 2000, 12 (8A), A209.

68. Pershan, P. S., Effects of thermal roughness on X-ray studies of liquid surfaces.

Colloids Surf. A: Physicochem. Eng. Asp. 2000, 171 (1-3), 149-157.

69. Shpyrko, O.; Fukuto, M.; Pershan, P.; Ocko, B.; Kuzmenko, I.; Gog, T.;

Deutsch, M., Surface layering of liquids: The role of surface tension. Phys. Rev.

B 2004, 69 (24), 245423.

70. Huisman, W. J.; Peters, J. F.; Zwanenburg, M. J.; de Vries, S. A.; Derry, T. E.;

Abernathy, D.; van der Veen, J. F., Layering of a liquid metal in contact with a

hard wall. Nature 1997, 390 (6658), 379-381.

66
71. Murphy, B.; Festersen, S.; Magnussen, O., The Atomic scale structure of liquid

metal–electrolyte interfaces. Nanoscale 2016, 8 (29), 13859-13866.

72. Bozack, M.; Bell, A.; Swanson, L., Influence of surface segregation on wetting

of liquid metal alloys. J. Phys. Chem. 1988, 92 (13), 3925-3934.

73. Lei, N.; Huang, Z.; Rice, S. A., Structure of the liquid–vapor interface of a Sn:Ga

alloy. J. Chem. Phys. 1997, 107 (10), 4051-4060.

74. Lei, N.; Huang, Z.; Rice, S. A., Surface segregation and layering in the liquid–

vapor interface of a dilute bismuth: gallium alloy. J. Chem. Phys. 1996, 104 (12),

4802-4805.

75. Tostmann, H.; DiMasi, E.; Shpyrko, O. G.; Pershan, P. S.; Ocko, B. M.;

Deutsch, M., II. Surface and interfacial phenomena: Surface phases in binary

liquid metal alloys: An X‐ray study. Ber. Bunsen-Ges. Phys. Chem. 1998, 102 (9),

1136-1141.

76. Ghasemian, M. B.; Zavabeti, A.; Mousavi, M.; Murdoch, B. J.; Christofferson,

A. J.; Meftahi, N.; Tang, J.; Han, J.; Jalili, R.; Allioux, F. M., Doping Process

of 2D Materials Based on the Selective Migration of Dopants to the Interface of

Liquid Metals. Adv. Mater. 2021, 33 (43), 2104793.

77. Tang, J.; Lambie, S.; Meftahi, N.; Christofferson, A. J.; Yang, J.; Ghasemian,

M. B.; Han, J.; Allioux, F.-M.; Rahim, M.; Mayyas, M., Unique surface patterns

emerging during solidification of liquid metal alloys. Nat. Nanotechnol. 2021, 16

(4), 431-439.

78. Tang, J.; Lambie, S.; Meftahi, N.; Christofferson, A. J.; Yang, J.; Han, J.;

Rahim, M.; Mayyas, M.; Ghasemian, M. B.; Allioux, F.-M., Oscillatory

bifurcation patterns initiated by seeded surface solidification of liquid metals. Nat.

Synth. 2022, 1, 158-169.

67
79. Cabrera, N.; Mott, N. F., Theory of the oxidation of metals. Rep. Prog. Phys. 1949,

12 (1), 163.

80. Ermoline, A.; Dreizin, E. L., Equations for the Cabrera–Mott kinetics of oxidation

for spherical nanoparticles. Chem. Phys. Lett. 2011, 505 (1-3), 47-50.

81. Plech, A.; Klemradt, U.; Metzger, H.; Peisl, J., In situ X-ray reflectivity study of

the oxidation kinetics of liquid gallium and the liquid alloy. J. Condens. Matter

Phys. 1998, 10 (5), 971.

82. Regan, M.; Tostmann, H.; Pershan, P. S.; Magnussen, O.; DiMasi, E.; Ocko,

B.; Deutsch, M., X-ray study of the oxidation of liquid-gallium surfaces. Phys.

Rev. B 1997, 55 (16), 10786.

83. Sangaletti, L.; Depero, L. E.; Allieri, B.; Pioselli, F.; Comini, E.; Sberveglieri,

G.; Zocchi, M., Oxidation of Sn thin films to SnO2. Micro-Raman mapping and

X-ray diffraction studies. J. Mater. Res. 1998, 13 (9), 2457-2460.

84. Ricci, E.; Lanata, T.; Giuranno, D.; Arato, E., The effective oxidation pressure

of indium-oxygen system. J. Mater. Sci. 2008, 43 (9), 2971-2977.

85. Xia, J.-y.; Tang, M.-t.; Cui, C.; Jin, S.-m.; Chen, Y.-m., Preparation of α-Bi 2O3

from bismuth powders through low-temperature oxidation. Trans. Nonferrous

Met. Soc. 2012, 22 (9), 2289-2294.

86. Niinistö, J.; Petrova, N.; Putkonen, M.; Niinistö, L.; Arstila, K.; Sajavaara, T.,

Gadolinium oxide thin films by atomic layer deposition. J. Cryst. Growth 2005,

285 (1-2), 191-200.

87. Daeneke, T.; Atkin, P.; Orrell-Trigg, R.; Zavabeti, A.; Ahmed, T.; Walia, S.;

Liu, M.; Tachibana, Y.; Javaid, M.; Greentree, A. D.; Russo, S. P.; Kaner, R.

B.; Kalantar-Zadeh, K., Wafer-scale synthesis of semiconducting SnO

68
 monolayers from interfacial oxide layers of metallic liquid tin. ACS Nano 2017,

11 (11), 10974-10983.

88. Messalea, K. A.; Carey, B. J.; Jannat, A.; Syed, N.; Mohiuddin, M.; Zhang, B.

Y.; Zavabeti, A.; Ahmed, T.; Mahmood, N.; Della Gaspera, E., Bi2O3

monolayers from elemental liquid bismuth. Nanoscale 2018, 10 (33), 15615-

15623.

89. Li, J.; Zhang, X. L.; Yang, B.; Zhang, C.; Xu, T. T.; Chen, L. X.; Yang, L.;

Jin, X.; Liu, B. D., Template approach to large-area non-layered Ga-group two-

dimensional crystals from printed skin of liquid gallium. Chem Mater 2021, 33

(12), 4568-4577.

90. Cooke, J.; Ghadbeigi, L.; Sun, R.; Bhattacharyya, A.; Wang, Y.; Scarpulla, M.

A.; Krishnamoorthy, S.; Sensale-Rodriguez, B., Synthesis and characterization

of large-area nanometer‐thin β-Ga2O3 films from oxide printing of liquid metal

gallium. Phys. Status Solidi A 2020, 217 (10), 1901007.

91. Ghasemian, M. B.; Zavabeti, A.; Abbasi, R.; Kumar, P. V.; Syed, N.; Yao, Y.;

Tang, J.; Wang, Y.; Elbourne, A.; Han, J., Ultra-thin lead oxide piezoelectric

layers for reduced environmental contamination using a liquid metal-based

process. J. Mater. Chem. A 2020, 8 (37), 19434-19443.

92. Mahmood, N.; Khan, H.; Tran, K.; Kuppe, P.; Zavabeti, A.; Atkin, P.;

Ghasemian, M. B.; Yang, J.; Xu, C.; Tawfik, S. A., Maximum piezoelectricity

in a few unit-cell thick planar ZnO-A liquid metal-based synthesis approach.

Mater. Today 2021, 44, 69-77.

93. Atkin, P.; Orrell-Trigg, R.; Zavabeti, A.; Mahmood, N.; Field, M.; Daeneke,

T.; Cole, I.; Kalantar-Zadeh, K., Evolution of 2D tin oxides on the surface of

molten tin. Chem. Commun. 2018, 54 (17), 2102-2105.

69
94. Carey, B. J.; Ou, J. Z.; Clark, R. M.; Berean, K. J.; Zavabeti, A.; Chesman, A.

S. R.; Russo, S. P.; Lau, D. W. M.; Xu, Z. Q.; Bao, Q. L.; Kevehei, O.; Gibson,

B. C.; Dickey, M. D.; Kaner, R. B.; Daeneke, T.; Kalantar-Zadeh, K., Wafer-

scale two-dimensional semiconductors from printed oxide skin of liquid metals.

Nat. Commun. 2017, 8, 14482.

95. Nguyen, C. K.; Low, M. X.; Zavabeti, A.; Jannat, A.; Murdoch, B. J.; Della

Gaspera, E.; Orrell-Trigg, R.; Walia, S.; Elbourne, A.; Truong, V. K., Ultrathin

oxysulfide semiconductors from liquid metal: a wet chemical approach. J. Mater.

Chem. C 2021, 9 (35), 11815-11826.

96. Khan, H.; Mahmood, N.; Zavabeti, A.; Elbourne, A.; Rahman, M. A.; Zhang,

B. Y.; Krishnamurthi, V.; Atkin, P.; Ghasemian, M. B.; Yang, J., Liquid metal-

based synthesis of high performance monolayer SnS piezoelectric

nanogenerators. Nat. Commun. 2020, 11, 3449.

97. Krishnamurthi, V.; Khan, H.; Ahmed, T.; Zavabeti, A.; Tawfik, S. A.; Jain, S.

K.; Spencer, M. J.; Balendhran, S.; Crozier, K. B.; Li, Z., Liquid‐metal

synthesized ultrathin SnS layers for high‐performance broadband photodetectors.

Adv. Mater. 2020, 32 (45), 2004247.

98. Tang, J.; Tang, J.; Mayyas, M.; Ghasemian, M. B.; Sun, J.; Rahim, M. A.;

Yang, J.; Han, J.; Lawes, D. J.; Jalili, R., Liquid‐metal‐enabled mechanical‐

energy‐induced CO2 conversion. Adv. Mater. 2022, 34, 2105789.

99. Allioux, F.-M.; Merhebi, S.; Tang, J.; Zhang, C.; Merenda, A.; Cai, S.;

Ghasemian, M. B.; Rahim, M. A.; Maghe, M.; Lim, S., Carbonization of low

thermal stability polymers at the interface of liquid metals. Carbon 2021, 171,

938-945.

70
100. Mayyas, M.; Mousavi, M.; Ghasemian, M. B.; Abbasi, R.; Li, H.; Christoe, M.

J.; Han, J.; Wang, Y.; Zhang, C.; Rahim, M. A., Pulsing liquid alloys for

nanomaterials synthesis. ACS Nano 2020, 14 (10), 14070-14079.

101. Mayyas, M.; Khoshmanesh, K.; Kumar, P.; Mousavi, M.; Tang, J.; Ghasemian,

M. B.; Yang, J.; Wang, Y.; Baharfar, M.; Rahim, M. A., Gallium‐based liquid

metal reaction media for interfacial precipitation of bismuth nanomaterials with

controlled phases and morphologies. Adv. Funct. Mater. 2021, 32 (8), 2108673.

102. Mousavi, M.; Ghasemian, M. B.; Han, J.; Wang, Y.; Abbasi, R.; Yang, J.;

Tang, J.; Idrus-Saidi, S. A.; Guan, X.; Christoe, M. J., Bismuth telluride

topological insulator synthesized using liquid metal alloys: Test of NO2 selective

sensing. Appl. Mater. Today 2021, 22, 100954.

103. Alsaif, M.; Kuriakose, S.; Walia, S.; Syed, N.; Jannat, A.; Zhang, B. Y.; Haque,

F.; Mohiuddin, M.; Alkathiri, T.; Pillai, N.; Daeneke, T.; Ou, J. Z.; Zavabeti,

A., 2D SnO/In2O3 van der Waals heterostructure photodetector based on printed

oxide skin of liquid metals. Adv. Mater. Interfaces 2019, 6 (7), 1900007.

104. Crawford, J.; Cowman, A.; O'Mullane, A. P., Synthesis of 2D cobalt oxide

nanosheets using a room temperature liquid metal. RSC Adv. 2020, 10 (49),

29181-29186.

105. Alkathiri, T.; Dhar, N.; Jannat, A.; Syed, N.; Mohiuddin, M.; Alsaif, M.; Datta,

R. S.; Messalea, K. A.; Zhang, B. Y.; Khan, M. W.; Elbourne, A.; Pillai, N.;

Ou, J. Z.; Zavabeti, A.; Daeneke, T., Atomically thin TiO2 nanosheets synthesized

using liquid metal chemistry. Chem. Commun. 2020, 56 (36), 4914-4917.

106. Zavabeti, A.; Aukarasereenont, P.; Tuohey, H.; Syed, N.; Jannat, A.; Elbourne,

A.; Messalea, K. A.; Zhang, B. Y.; Murdoch, B. J.; Partridge, J. G., High-

71
 mobility p-type semiconducting two-dimensional β-TeO2. Nat. Electron. 2021, 4

(4), 277-283.

107. Lin, J.; Li, Q.; Liu, T.-Y.; Cui, Y.; Zheng, H.; Liu, J., Printing of quasi‐2D

semiconducting β-Ga2O3 in constructing electronic devices via room‐temperature

liquid metal oxide skin. Phys. Status Solidi Rapid Res. Lett. 2019, 13 (9), 1900271.

108. Chen, Y.; Liu, K.; Liu, J.; Lv, T.; Wei, B.; Zhang, T.; Zeng, M.; Wang, Z.;

Fu, L., Growth of 2D GaN single crystals on liquid metals. J. Am. Chem. Soc.

2018, 140 (48), 16392-16395.

109. Messalea, K. A.; Syed, N.; Zavabeti, A.; Mohiuddin, M.; Jannat, A.;

Aukarasereenont, P.; Nguyen, C. K.; Low, M. X.; Walia, S.; Haas, B., High-k

2D Sb2O3 Made Using a Substrate-Independent and Low-Temperature Liquid-

Metal-Based Process. ACS Nano 2021, 15 (10), 16067-16075.

110. Li, Q.; Lin, J.; Liu, T.-Y.; Zhu, X.-Y.; Yao, W.-H.; Liu, J., Gas-mediated liquid

metal printing toward large-scale 2D semiconductors and ultraviolet

photodetector. npj 2D Mater. Appl. 2021, 5, 36.

111. Ghasemian, M. B.; Daeneke, T.; Shahrbabaki, Z.; Yang, J.; Kalantar-Zadeh, K.,

Peculiar piezoelectricity of atomically thin planar structures. Nanoscale 2020, 12

(5), 2875-2901.

112. Wu, T.; Zhang, H., Piezoelectricity in two‐dimensional materials. Angew. Chem.

Int. Ed. 2015, 54 (15), 4432-4434.

113. Nye, J. F., Physical properties of crystals: their representation by tensors and

matrices. Oxford university press: 1985.

114. Wang, Y.; Wang, S.; Chang, H.; Rao, W., Galvanic replacement of liquid

metal/reduced graphene oxide frameworks. Adv. Mater. Interfaces 2020, 7 (19),

2000626.

72
115. Baharfar, M.; Mayyas, M.; Rahbar, M.; Allioux, F.-M.; Tang, J.; Wang, Y.;

Cao, Z.; Centurion, F.; Jalili, R.; Liu, G., Exploring interfacial graphene oxide

reduction by liquid metals: application in selective biosensing. ACS Nano 2021,

15 (12), 19661-19671.

116. Alsaif, M. M.; Haque, F.; Alkathiri, T.; Krishnamurthi, V.; Walia, S.; Hu, Y.;

Jannat, A.; Mohiuddin, M.; Xu, K.; Khan, M. W., 3D visible‐light‐driven

plasmonic oxide frameworks deviated from liquid metal nanodroplets. Adv.

Funct. Mater. 2021, 31 (52), 2106397.

73
Chapter 3 Self-Deposition of 2D Molybdenum Sulfides on Liquid

Metals

74
3.1 Introduction

Two-dimensional (2D) transition metal dichalcogenides (TMDs) play increasingly

significant roles in research and future optoelectronics. However, the large-scale

deposition of 2D TMDs remains challenging due to sparse nucleation and substrate

dependency. Liquid metals can offer effective solutions to meet these challenges due to

their reactive, non-polarized, and templating properties.

In this chapter, the author shows the self-deposition of 2D molybdenum sulfide by

introducing a molybdenum precursor onto the surface of a eutectic alloy of gallium and

indium (EGaIn). A consecutive set of reactions and washing steps were carried out, using

a molybdenum precursor, to induce the self-deposition of ultra-thin molybdenum sulfide

on the surface of EGaIn. When the reactive species of the molybdenum sulfide precursor

are introduced onto the surface of EGaIn, they undergo a self-deposition process to

produce 2D molybdenum sulfide which are transferrable to any substrate. In this process,

EGaIn serves as an ultra-smooth template and reducing agent for the precursor to form

large-scale planar molybdenum sulfides, A highly crystalline 2H-MoS 2 is obtained after

a final annealing step.

The content of this chapter was published as a peer-reviewed article in the journal of

Advanced Functional Materials.1

3.2 Methods

3.2.1 Materials

All the chemicals were used as received without further purification. Gallium (round

shots, 99.9%), indium (beads, 99.9%) and tin (popcorn flakes, 99.9%) were purchased

75
from Rotometals, USA. Ammonium tetrathiomolybdate ((NH4)2MoS4, 99.97%) was

purchased from Sigma-Aldrich. Sodium hydroxide (NaOH, AR) pellets were obtained

from Chem-Supply Pty Ltd. Hydrochloric acid (HCl, 32%) was purchased from Chem-

Supply Pty Ltd. Potassium chloride (KCl, AR) was purchased from Chem-Supply Pty

Ltd. Gold wire (0.25 mm in diameter, 99.99%) was purchased from Sigma-Aldrich. Milli-

Q water (18.2 MΩ·cm at 25 °C) was used in all experiments.

3.2.2 Preparation of materials

Eutectic gallium indium alloy (EGaIn) was prepared in-house in order to avoid any

unknown additives of commercial EGaIn. A gallium (75.5 wt%) and indium (24.5 wt%)

metal mixture was melted at 180 °C on a hot plate. This temperature was well above the

melting point of indium (156.5 °C). Upon melting, the liquid alloy was stirred with a glass

rod to ensure homogeneous fusion and then allowed to cool down at room-temperature.

The EGaIn alloy was transferred from/into the glass vial using a glass Pasteur pipette.

3.2.3 Self-deposition of molybdenum sulfides (MoS x) on EGaIn

For the liquid metal-assisted synthesis of MoSx, an EGaIn liquid droplet (120 µL) was

placed in a plastic Petri dish (35 mm inner diameter, 36 mm outer diameter and 15 mm

height). The surface of the EGaIn liquid was cleaned from the passive oxide layer by

adding 5.0 mL of 5.0 M HCl directly onto the liquid metal droplet. The HCl instantly

removed the oxide layer and the liquid metal droplet turned from a dull and deformed

shape into a spherical and shiny ball. The HCl washing liquid was then removed and

replaced with 5.0 ml of a 1.92 mM (NH4)2MoS4 aqueous solution. The MoSx interfacial

76
layer formed immediately on the EGaIn droplet and the (NH4)2MoS4 solution was diluted

with 35 mL of Milli-Q water to halt the deposition process. The Petri dish was placed in

a larger container so that the overflow from the dilution step could be contained. The

solution in the vicinity of the MoSx-bearing EGaIn droplet was gently removed and

replaced with 5.0 mL of Milli-Q water. This was repeated another two times to ensure the

cleanliness of the MoSx-bearing EGaIn droplet. To ensure a successful transfer of the

uniform and large interfacial layer, extra care was taken while washing the EGaIn droplet

as any disturbance of the droplet could introduce microtears into the interfacial layer.

Afterwards, the MoSx-bearing droplet was left to dry naturally at room-temperature under

an N2 inert atmosphere.

3.2.4 Extraction and transfer of MoSx

The interfacial MoSx was extracted from the surface of the EGaIn droplet and transferred

to pre-cleaned substrates using a piezo-electric nano-positioning device (N-765

NEXACT precision z stage, 0.5 nm resolution, 25 N push/pull force, Physik Instrumente

(PI) GmbH & Co. KG). Various substrates (Si, SiO2(285 nm)/Si, and sapphire substrates)

were utilised to demonstrate the substrate independency of the process. The substrates

were sonicated in acetone, isopropanol and Milli-Q water for 10 min each and stored in

isopropanol. Prior to use, the substrates were rinsed in Milli-Q water, blow-dried with

nitrogen gas, and further cleaned with a Tergeo plasma cleaner-Pie Scientific, USA (15

W, 2.0 min).

The pre-cleaned substrate (1.0 cm × 1.0 cm) was fixed upside-down on a holder (2.0 cm

above the stage), and the MoSx-bearing EGaIn droplet in the Petri dish was placed on the

peizoelectric stage under the substrate. The piezo-electric stage then lifted the droplet

77
towards the substrate with a speed of 0.20 mm s 1 until they were in contact. The contact

was held for 5.0 s before the stage was brought down at the same speed, leaving the MoSx

layer behind on the substrate. Due to the van der Waals forces between the MoSx layer

and the solid substrate, the transferred MoS x layer adhered strongly to the substrate. After

the transfer, the substrate was removed from the holder and washed using an ultrasonic

bath in ethanol to eliminate any residual liquid metal. To achieve a successful transfer, it

was essential to completely dry the surfaces of the liquid metal droplet and the substrate.

The wafer, carrying the transferred 2D materials, was immersed in 50 mM NaOH for 30

s, to wash away any oxidized gallium residues, rinsed with Milli-Q water and blow-dried

with nitrogen gas.

3.2.5 Annealing process

A three-step annealing process was adapted to further crystallize and sulfurize the

resultant material into 2H-MoS2. A tube furnace (Thermo Scientific, TF55030C-1) with

a quartz tube (22 mm inner diameter, 25 mm outer diameter and 1.0 m length) was utilised

for annealing. In step one, the sample was annealed in a 5% H2 (balanced in N2)

atmosphere at 500 ˚C for 60 min for initial crystallization. In step two, the sample was

annealed in a 5% H2 (balanced in N2) and sulfur vapor at 750 C for 20 min. In step three,

the sample was annealed in ultra-pure argon and sulfur vapor at 850 C for 20 min to

further improve the quality of the 2D-MoS 2.

78
3.2.6 Electrochemical analysis

All the electrochemical processes were carried out using an electrochemical workstation

(CHI650E, CH Instruments Inc., USA).

The reference cyclic voltammetry (CV) was performed in a standard three-electrode cell,

with a gold-coated Si wafer (1.0 cm × 4.0 cm) as the working electrode, a spiral gold wire

as the counter electrode, and a saturated calomel electrode (SCE) as the reference. A glass

cell (25 mm inner diameter and 40 mm height) was used, and the electrolyte was an

aqueous solution of 1.92 mM (NH4)2MoS4 and 0.10 M KCl as the supporting electrolyte.

Different scan rates (25, 50, and 100 mV/s) were used at a resolution of 1.0 mA/V, with

a potential window from 1.3 V to 1.0 V.

The open circuit voltage (Voc) analysis of the EGaIn droplet was performed in a standard

three-electrode system with EGaIn as the working electrode. A plastic Petri dish (35 mm

inner diameter, 36 mm outer diameter and 15 mm height) was used as the cell, 0.10 M

KCl solution as the electrolyte, SCE as the reference and a spiral gold wire as the counter

electrode. An EGaIn droplet (120 µL) was used as the working electrode, and a gold wire

(inside a glass tube) was connected to the droplet by immersing one end of it into the

EGaIn droplet. The EGaIn electrode was washed with 5.0 mL of 5.0 M HCl which was

then removed and replaced with 0.10 M KCl solution as the supporting electrolyte for the

measurement. Prior to the measurement, the electrolyte was adjusted to the same pH as

the (NH4)2MoS4 precursor solution, monitored by a pH meter.

Electrochemical impedance spectroscopy (EIS) was carried out in the same three

electrode system as the Voc measurement, with 1.92 mM (NH4)2MoS4 and 0.10 M KCl as

the electrolyte. The frequency range was 1.0 Hz to 105 Hz, with 0.50 V used as the initial

voltage with a 5.0 mV amplitude. For the EIS measurement of the EGaIn droplet without

79
the self-deposited MoSx layer, 5.0 mL of 5.0 M HCl was used for cleaning the surface of

the EGaIn droplet before the measurement. The remnant of the HCl liquid was then

removed and replaced with the electrolyte. For the measurement of the MoS x-bearing

EGaIn droplet, MoSx was self-deposited onto an EGaIn electrode as described above, and

then the EIS measurement was performed. In order to make the deposition layer intact

during the measurement, a gold wire was connected to the droplet prior to the deposition.

3.2.7 Electrochemical deposition of MoSx

The electrochemical deposition of molybdenum sulfide was carried out by voltammetric

runs (scan rate 50 mV/s, 10 cycles) using the same three-electrode cell as described in the

Electrochemical analysis section.

3.2.8 Morphological characterization

Atomic force microscopy (AFM) images were collected using a Bruker dimension icon

SPM (USA) with ScanAsyst-Air probe (Bruker AFM probes, USA). Each sample was

scanned as prepared without further processing using the PeakForce Tapping mode. A

1
scan rate of 0.50 Hz and a resolution of 250 samples line was applied. A vacuum was

utilised to stabilize the specimens on the AFM stage.

High-resolution AFM images were obtained using a Cypher ES AFM (Oxford

Instrument, Asylum Research, Santa Barbara, CA, USA) in ambient room temperature

(25 °C) conditions using an amplitude modulated-AFM (AM-AFM). ArrowUHF

(NanoWorld, Switzerland, nominal spring constant kc = 6 N/m) were used for all images.

To minimize the imaging force, a setpoint ratio (Imaging Amplitude (A)/free amplitude

80
(A0)) of > 0.8 was maintained during imaging (unless otherwise stated). Cantilevers were

calibrated using the thermal spectrum method, in air, prior to use, and the lever sensitivity

was determined using force spectroscopy; the spring constant was resolved via the inverse

optical lever sensitivity (InVOLS) using force curve measurements on the hard surface.

A scanning electron microscopy (SEM, JEOL JSM-IT 500 HR) coupled with energy

dispersive X-ray spectroscopy (EDS, Bruker Silicon) was used for morphological and

elemental analysis of the electro-deposited materials.

A Philips CM200 field-emission high-resolution transmission electron microscopy

(TEM) system equipped with a GATAN ORIUS camera was employed for atomic-

resolution morphological and lattice structure analysis. Thermally robust silicon nitride

(Si3N4) grids (Ted Pella, 2158710) were employed for the TEM samples.

3.2.9 Optical characterization

X-ray photoelectron spectrometry (XPS) analysis was carried out using a Thermo

Scientific ESCALAB250i high-resolution XPS (UK) with monochromatic Al Kalpha soft

X-ray source (1486.68 eV), 2×10 9 mbar background vacuum, 90° photoelectron take-off

angle, 120 W power (13.8 KV × 8.7 mA), and 500 µm spot size.

A Renishaw inVia Raman microspectrometer (Gloucestershire, UK) equipped with 532

nm laser and WiRETM was utilised for Raman spectra collection. One second exposure

time, 200 accumulations with 1% laser power and 100 times magnification were used for

single point spectral acquisition. Raman mapping was obtained by raster scanning at 100

times magnification with a step-size of 0.3 μm, where a spectrum at each point was

collected with 5% laser power, 1 s exposure time and one accumulation.

81
3.3 Results and discussions

3.3.1 Self-deposition of ultrathin MoSx on liquid metal

The experimental process and mechanism by which the MoS x self-deposits on the surface

of EGaIn is schematically illustrated in Figure 3.1. In brief, prior to deposition, HCl

solution was used to wash the EGaIn droplet thereby removing the gallium oxide layer

and inducing intimate contact between the EGaIn and solvent for creating a reactive

interfacial environment. After washing with HCl, an aqueous solution of ammonium

tetrathiomolybdate [(NH4)2MoS4] was applied. Immediately, a uniform layer of

molybdenum sulfide formed on the interface and subsequently the surface was washed to

hinder any further growth. It has been established that for these Ga-based liquid metals,

when exposed to an oxic environment, a passivating oxide layer of gallium forms on their

surface (Figure 3.1A), as gallium oxide has a lower Gibb’s free energy than that of indium

oxide.2 For removing the oxide surface layer and exposing the surface of the liquid metal,

the droplet was washed with HCl (Figure 3.1B). EGaIn as a metal offers excessive

electrons on its surface and this attracts the H+ ions after washing (Figure 3.1C). The

surface of the EGaIn becomes saturated with electrons and reaches an equilibrium, where

no oxidation occurs at the interface between the liquid metal and the acidic aqueous

solution. This phenomenon is known to induce a surface potential on metals (Φm). The

surface potential of EGaIn is neutralized by counter-ions creating what is commonly

referred to as the electrical double-layer (EDL).3 Figure 3.1D features the different zones

of the EDL. In the inner Helmholtz layer, the H+ counter ions adhere directly to the surface

of the EGaIn droplet. The interfacial potential remains unbalanced in this region, and a

combination of co-ions and counter-ions is established in the outer Helmholtz layer to

reach the equilibrium potential (Φi).4 When (NH4)2MoS4 is added onto the EGaIn droplet,

the anions in the outer Helmholtz layer are substituted by the [MoS4]2- (Figure 3.1E).
82
These ions, in turn, react with the H+ in the inner Helmholtz layer according to Equation

(3.1)5 forming MoS3. Further reduction can occur in the inner Helmholtz layer by the

interfacial potential of Φm according to Equation (3.2) to form MoS2. The interfacial

product is collectively described here as MoSx. Weak interfacial forces hold the interfacial

MoSx film on the EGaIn surface as featured in Figure 3.1F. The EGaIn droplet provides

both the interfacial potential to destabilize the [MoS 4]2 and serves as an atomically flat

template, yielding a smooth and uniform 2D MoS x.

[MoS4 ]2- + 2H+ → MoS3 ↓ + H2 S↑ (3.1)

[MoS4 ]2- + 4H+ + 2e- → MoS2 ↓ + 2H2 S↑ (3.2)

Figure 3.1 Schematic illustration of MoSx formation on EGaIn. (A) The formation of a

gallium oxide passivating layer on the surface of liquid metal under an oxic environment.

83
(B) The removal of the oxide interfacial layer. (C) The build-up of surface charge on the

liquid metal. Here electrons saturate the liquid metal surface and system equilibrium is

reached. (D) Electron confinement and the Helmholtz layer formation on the surface of

liquid metal. The blue line is a hypothetical representation of the interfacial potential

(potential drop = Φm Φi). (E) Anionic substitution by [MoS4]2 in the outer Helmholtz

layer and subsequent reaction with the H+ in the inner Helmholtz layer. (F) The formation

of ultra-smooth interfacial film of 2D molybdenum sulfide.

3.3.2 Interfacial assessment of liquid metal in aqueous precursor

To further elucidate the mechanism by which MoSx deposits onto the EGaIn surface,

electrochemical assessment was employed. The electrochemical stability of the

(NH4)2MoS4 aqueous system was explored using a three-electrode system with a gold

working electrode at different scan rates as presented in Figure 3.2A. The anodic region,

at a potential more positive than 0.50 V, corresponds to the deposition of sulphur

saturated MoS3 (Equation 3.3). In the potential range from 0.56 to 1.3 V, a cathodic

region is observed, corresponding to both [MoS 4]2 deposition onto the gold electrode

(Equation 3.2) and reduction of the MoS3 (Equation 3.4).6, 7 The onset voltage for the

reduction of [MoS4]2 was determined to be 0.56 V. The chemically deposited material

during CV measurements was confirmed to be MoSx by SEM, EDS mapping, and XPS

(Figure 3.3 and 3.4). Optical microscopy images and Raman spectra of electrodeposited

and self-deposited MoSx are presented in Figure 3.5. It is observed that the self-deposited

MoSx layer on the EGaIn droplet is smooth and has no observable pinholes, while the

electrodeposited MoSx film on gold is rough and porous. Raman spectrum of

electrodeposited MoSx was collected directly from the MoSx-bearing gold-electrode after

84
the electrodeposition, and that of the self-deposited MoS x was collected directly from the

MoSx-bearing EGaIn droplet in order to avoid the effect of any signal from the underneath

1
substrates. The Raman peaks between 300 to 350 cm correspond to the Mo-S bond

vibrational modes, and the modes between 400 to 460 cm 1 are associated with the apical

sulfur atoms.8

As can be seen, the surface of the electrodeposited film onto gold is quite rough signifying

the importance of using the atomically smooth surface of EGaIn for avoiding such

nucleation formation.

[MoS ] → MoS ↓ + 1⁄8 S + 2e (3.3)

MoS + 1⁄8 S + 2e → [MoS ] (3.4)

In the second set of electrochemical experiments the working electrode was replace with

EGaIn to assess the Voc of a fresh EGaIn droplet, corresponding to the interfacial

potential. This experiment was conducted to determine whether the V oc is sufficient to

destabilize [MoS4]2 at the interface of EGaIn. As a reference, the Voc was acquired in an

aqueous solution without [MoS4]2 ions, with the same pH as the initial precursor

solution, over 1000 s as presented in Figure 3.2B. This Voc is approximately 0.80 V vs.

SCE. This value is considerably more negative than the onset potential of [MoS 4]2

reduction ( 0.56 V, as determined on a gold electrode), suggesting that the charge

accumulation and the interfacial potential between the interface of EGaIn and the aqueous

solution is sufficient to reduce the [MoS 4]2 species into an MoSx layer.

To initially confirm that the produced MoSx layer on EGaIn was present and of high

integrity, potentiostatic EIS was performed on the droplet with and without the self-

deposited MoSx layer. The presence of a uniform interfacial semiconducting MoSx layer

85
on the metal surface induces changes in the impedance of the electrochemical system. As

shown in Figures 3.2C and D, the impedance was higher for the EGaIn-MoS x than EGaIn

only, at all frequencies. A capacitive double layer appears to develop on the EGaIn droplet

at frequencies of >10 Hz. This capacitive component is not present in the EGaIn-MoS x

due to the interfacial layer reducing the charge transfer and thus the formation of a double

layer capacitance. Only at frequencies of <10 Hz, the EGaIn-MoSx shows a capacitive

component due to the transpassive effect. The drop in the capacitive component of an

EGaIn droplet at a frequency of <10 Hz can be attributed to the rapid formation of the

EDL and its saturation creating an electric barrier on the surface.

The Nyquist plots, Figure 3.2E, feature a higher series resistance (Rs) for EGaIn-MoSx

(92.6 Ω) compared to that of the EGaIn droplet (10.2 Ω), further suggesting that the MoS x

layer on EGaIn is uniform and without any pores.9, 10 This provides further evidence for

the templating advantage of EGaIn’s smooth interface. Additionally, from the

electrochemical assessments, it was seen that the intrinsic EDL of EGaIn can be utilised

for depositing MoSx, without applying any external stimulant.

To achieve the accurate exfoliation and transfer of the self-deposited MoSx layers from

the EGaIn surface onto a desired substrate, a program-controlled piezo-electric stage was

utilised. This program-controlled stage of 0.5 nm spatial steps is utilised to carry out the

touch-printing process (Figure 3.2F). The high-resolution stage allows the transfer of

large area layers (as large as several mm - Figure 3.6) and increases the stability and

reproducibility of the process. The transfer was conducted onto a variety of substrates

including SiO2 (285 nm)/Si and sapphire wafers (Figures 3.6 and 3.7) to show that

substrate independency of the technique. The transfer process is schematically illustrated

in Figure 3.2 G. Due to the non-polar nature of EGaIn, interactions between the metal and

the synthesised 2D materials on its surface is weak.11 Additionally, the absence of a long-
86
range orderly arrangement of atoms in liquid metal minimizes the cumulative atomic

forces over long distances.12, 13

Conversely, the forces between MoSx and oxygen

terminated substrates (such as SiO2 and sapphire) are stronger due to accumulative van

der Waals forces between the permanent dipoles.14 When the substrate is brought into

contact with the EGaIn, the 2D layer on its surface adheres to the substrate and separates

itself from the parent liquid metal. Residual liquid metal inclusions are sometimes found

on the transferred 2D layers that can be removed by an ultrasonic bath in ethanol.

Figure 3.2 Electrochemical characterization of the ((NH4)2MoS4) aqueous system and

schematic illustrations for the extraction of the MoSx layer. (A) CV profiles of

((NH4)2MoS4) aqueous solution acquired at different scan rates, scanned in the potential

window from 1.0 to 1.3 V (negative scan), Ei = 0.0 V and Ef = 0.0 V vs. SCE. (B) The

Voc measurement of the EGaIn droplet in an aqueous solution vs. time over 1000 s. (C

and D) A Bode plot of the EGaIn droplet with and without the MoS x interfacial layer;

scan conditions: 1.92 mM (NH4)2MoS4 and 0.10 M KCl as electrolyte, 0.50 V. (E) A

Nyquist plot of an EGaIn droplet with and without the MoS x interfacial layer, the recorded
87
values of series resistance (Rs) are 10.2 and 92.6 Ω, respectively; scan conditions: 1.92

mM (NH4)2MoS4 and 0.10 M KCl as electrolyte, 0.50 V. (F) Schematics of the

experimental set up for the extraction of an MoSx layer from the EGaIn surface. (G) A

cross-sectional schematic illustration of the extracting and transferring processes.

Figure 3.3 SEM image (top left) and EDS mapping results of Mo, S and O of the

electrodeposited MoSx (scale bars: 10 μm).

88
Figure 3.4 High-resolution XPS spectra of the Mo 3d and S 2p regions of the

electrodeposited MoSx.

Figure 3.5 Optical microscopy images of: (A) electrodeposited MoS x film on a gold

wafer and (B) MoSx self-deposited layer on an EGaIn droplet. (C) Raman spectra of both

the electrodeposited MoSx film on gold (black spectrum), and MoS x self-deposited layer

on EGaIn (red spectrum).

89
Figure 3.6 Optical microscopy image of the transferred MoSx layer on a Si/SiO2 wafer

that shows a large-area homogeneous layer with no pinholes.

Figure 3.7 Optical microscopy images of transferred MoSx layers onto different

substrates: Si (left) and sapphire (right).

3.3.3 Structure and morphology of 2D MoS2 sheet

After being transferred onto the substrate, the material appears to be a uniform sheet with

a thickness of ~1 nm as determined by AFM (Figure 3.8). The XPS analysis confirmed

that this sheet was MoSx. The high-resolution XPS spectra at the Mo 3d and S 2p regions

are shown in Figures 3.9. In the high-resolution XPS spectra of the S 2p region, one
90
doublet at 162 and 163.3 eV is observed, which corresponds to S 2 from molybdenum

sulfide.15 In the Mo 3d region, the peaks are fit to two doublets, representing two chemical

states of Mo and one single peak of S 2s at 226.2 eV. The doublet at 230.0 and 233.3 eV

can be attributed to Mo4+ 3d5/2 and Mo4+ 2d3/2 peaks, respectively. The doublet with higher

binding energies of 232.8 and 235.9 eV belongs to MoVI from MoO3, while the doublet

at 230.0 and 233.3 eV belongs to MoIV from molybdenum sulfide.15 Based on elemental

analysis, the stoichiometric ratio of MoIV to S in molybdenum sulfide is 1.82:1, indicating

the presence of substantial S-vacancies in the synthesised material. The XPS peaks are

slightly broader than that of single-crystal MoS 2, which can be attributed to the presence

of convoluted peaks from multiple oxidation states of the material.16 Overall, the product

of the self-deposition process is identified as S-deficient MoSx with MoO3 inclusions.

Figure 3.8 AFM image of the MoSx layer extracted from the surface of EGaIn onto a

SiO2/Si wafer. The height profile shows that the thickness is ~1.0 nm before annealing.

91
Figure 3.9 High-resolution XPS spectra of the Mo 3d and S 2p regions of the self-

deposited MoSx.

A TEM image and a selected area electron diffraction (SAED) pattern of a sample MoS x

sheet are shown in Figure 3.10. The translucent appearance of the sheet in the TEM image

suggests its thin nature in confirmation to AFM measurements, and the absence of any

diffraction dots or rings in the SAED pattern suggests low crystallinity of the product.

Therefore, annealing process was applied to obtain stoichiometric and crystalline 2D

MoS2 sheets. Characteristic Raman peaks of MoS 2 show up after annealing (the full range

of Raman spectra for the layers after each step are shown Figure 3.11).

92
Figure 3.10 TEM analysis of the self-deposited MoSx layer. (A) Optical microscopy

image of the transferred MoSx layer on a Si3N4 grid. (B) TEM image and (C) SAED

pattern of the self-deposited MoSx layer before annealing.

Figure 3.11 Raman analysis of the EGaIn-extracted MoSx layer on a SiO2/Si wafer at

different annealing stages. The folding of the layer created different thicknesses: one-

layer, few-layers and multilayer (thicker) 2H-MoS 2. In the optical microscopy image, the

crosses mark the points where the spectra were collected.

93
Figure 3.12A and B show the XPS of the sheets, after annealing, at Mo 3d and S 2p

regions. The doublet in Mo 3d region corresponds to the 3d 5/2 and 2d3/2 orbital of MoVI,

and the single peak is attributed to the 2s peak of divalent sulfide ions (S 2−). The doublet

in S2p region corresponds to the 2p1/2 and 2p3/2 orbital of S2−. The ratio of MoIV to S is 1:

2.03. As such, the XPS results confirm to be MoS2 sheets.16 AFM image (Figure 3.12C)

shows a smooth and uniform sheet. A line scan across the edge of the sheet presents a

reduced step height of ~0.7 nm after annealing, which is in line with the reported thickness

of MoS2 monolayers.17, 18 Atomic resolution AFM imaging (Figure 3.12E) presents a

repeated hexagonal pattern, which indicates a high crystallinity of the resultant sheets. In

order to gain further insight into the atomic structure of the synthesised nanosheets, high-

resolution TEM was conducted as presented in Figure 3.12D and F (presented at different

resolutions). Thermally robust silicon nitride (Si 3N4) grid was utilised here since copper

grid with lacy carbon could not sustain the high temperature annealing process. The hig-

resolution TEM image clearly reveals a honeycomb atomic arrangement with 2H-MoS 2.

The d-spacing of [1010] plane is 2.76 Å, and the d-spacing of [1120] plane is 1.59 Å,

which agrees with the reported structure of 2H-MoS2 crystal,19 as is shown in Figure

3.12G. The fast Fourier transform (FFT) pattern (Figure 3.12H) displays perfect

hexagonal symmetrical patterns, indicating the hexagonal lattice structure of MoS 2

crystals. TEM image of the folded edge of the MoS2 layer is presented in Figure 3.13.

The spacing between layers is ~0.66 nm, which is in agreement with the thickness of

monolayer MoS2.19

94
Figure 3.12 The high-resolution XPS of the crystalline MoS 2 sheets at Mo 3d (A) and S

2p (B) region. AFM and TEM characterizations of the 2D MoS2. (C) AFM image with

the inset representing the thickness of ~0.7 nm. (E) High-resolution AFM image. (D) and

(F) high resolution TEM image. (G) The representation of the crystal structure. (H) The

FFT pattern.

Figure 3.13 TEM image of a folded edge of a 2H-MoS2 thick layer after annealing.

95
3.3.4 Thickness dependent optical properties of 2D MoS2 sheet

To Raman spectra and photoluminescence (PL) characterization of the 2D-MoS 2 sheets

of different thicknesses were collected and are presented in Figure 3.14. An optical

microscopy image of a sample with different thicknesses is presented in Figure 3.14A. It

has been established that for few-layer MoS2 the frequency difference between the E12g

and A1g mode can be used for determining the number of layers.20 As shown in Figure

3.14B, Raman spectra of 2D-MoS2 present characteristic E12g and A1g modes, which

correspond to the in-plane optical vibrations of the Mo and S atoms and the out-of-plane

optical vibrations of the S atoms, respectively. With the number of layers increasing from

1 layer to 5, the E12g mode goes through a redshift from 388.1 to 385.3 cm 1
and the A1g

1
mode goes through a blueshift from 408.2 to 409.6 cm (Figure 3.14C). The difference

between the A1g and E12g modes increased from 20.1 cm 1 for the monolayer 2H-MoS2 to

1
24.2 cm for multi-layer 2H-MoS2 (Figure 3.14D), which is in agreement with the

reported value of chemical vapor deposited MoS 221-23 and indicates the highly crystalline

quality of the resultant 2D sheets. PL spectra of 2H-MoS2 have been reported to also be

a function of MoS2 thickness.24 As shown in Figure 3.14E, the PL of monolayer MoS2

presents two excitation peaks at 620 nm and 656 nm, which correspond to the B and A

excitations modes of MoS2.25 As the thickness increases, both PL peaks go through slight

blueshifts and these PL shifts agree well with that of past observations 25 and further

proved the ability of the liquid metal process in obtaining high quality 2H-MoS 2 from one

to several layers.

96
Figure 3.14 Raman spectroscopy and PL characterizations of the 2D-MoS 2: (A) Optical

microscopy image of the 2D MoS2. (B) Raman spectra of 2D MoS2 of different number

of layers. (C) A1g and E12g peak locations at different number of layers. (D) The difference

between peak locations of the A1g and E12g peaks for different number of layers. (E) PL

spectra of monolayer and bilayer 2D MoS2.

To further characterize the resultant 2H-MoS2 surface homogeneity at a large scale,

Raman mapping was carried out. An example is presented in Figure 3.15. The optical

microscopy image of a sample mapping region is shown in Figure 3.15A with the spatial

distribution of the Raman peak intensities of this image depicted in Figure 3.15B. As can

1
be seen in Figure 3.15B, the spatial map of peak intensity at 409 cm (the approximate

location of the A1g peak) matches perfectly with the optical microscopy image contrast of

the MoS2 sheet. The majority of the 2D sheet shows a similar distribution of intensity,
97
which indicts a high uniformity of the MoS2 over a large area. The folded area of the

material (point 3) shows higher intensity due to increased thickness.

The spatial maps of the peak location of the E12g and A1g modes are presented in Figures

3.15C and D, respectively, for which single point spectra collected from the marked

points are presented in Figure 3.16 as an example. The peak locations difference of the

E12g and A1g modes is approximately 20 cm 1, which indicates monolayer MoS2 over this

area. The Raman mapping, with homogeneous colour distribution on the images and

relatively invariant peak intensities, further demonstrate the quality and uniformity of the

large area 2H-MoS2 monolayers that can be obtained using the developed liquid metal-

based method.

Figure 3.15 Raman mapping of a 2H-MoS2 sheet. A single-point spectrum was collected

at the point marked by a cross and shown in the Figure 3.16. (A) Optical microscopy

image of the mapped region. (B) Intensity mapping at 409 cm 1. (C) E12g peak location

mapping. (D) A1g peak location mapping.

98
Figure 3.16 Single point spectrum collected from the marked point in the Raman

mapping analysis presented in Figure 3.15.

3.5 Conclusion

In conclusion, the author demonstrated the possibly of forming 2D molybdenum sulfide

on the surface of EGaIn. The EGaIn interface acted as a reactive and a templating media

simultaneously. The precursor, (NH4)2MoS4, was intrinsically destabilized and reduced

on the surface of EGaIn at room temperature to establish interfacial molybdenum sulfide.

With a series of electrochemical characterizations, the author defined the surface

reactions that constitute the formation of the interfacial 2D molybdenum compound. The

layer was readily exfoliated and transferred onto a desired substrate using a piezo stage

of nm resolution. The annealing steps were included to complete the process for obtaining

planar 2H-MoS2 of different thicknesses. The technique described presented a universal

procedure for the deposition of any 2D TMDs of large spatial dimensions that can be

adapted for large-scale production purposes, rivalling conventional deposition and

exfoliation methods for 2D TMDs, and could be applied for future

99
3.6 References

1. Wang, Y.; Mayyas, M.; Yang, J.; Tang, J.; Ghasemian, M. B.; Han, J.;

Elbourne, A.; Daeneke, T.; Kaner, R. B.; Kalantar‐Zadeh, K., Self‐Deposition of

2D Molybdenum Sulfides on Liquid Metals. Adv. Funct. Mater. 2021, 31 (3),

2005866.

2. Zavabeti, A.; Ou, J. Z.; Carey, B. J.; Syed, N.; Orrell-Trigg, R.; Mayes, E. L.;

Xu, C.; Kavehei, O.; O’Mullane, A. P.; Kaner, R. B., A liquid metal reaction

environment for the room-temperature synthesis of atomically thin metal oxides.

Science 2017, 358 (6361), 332-335.

3. Schmickler, W., Electrochemical theory: double layer. In Chemistry, Molecular

Sciences and Chemical Engineering, Elsevier Inc.: 2014.

4. Helmoltz, H., Ueber einige Gesetze der Verteilung elektrischer Ströme in

körperlichen Leitern, mit Anwendung auf die thierisch elektrischen Versuche.

Ann. phys. Chem. 1853, 3 (29), 353-377.

5. Quagraine, E.; Georgakaki, I.; Coucouvanis, D., Reactivity and kinetic studies of

(NH4)2(MoS4) in acidic aqueous solution: Possible relevance to the angiostatic

function of the MoS42- ligand. J. Inorg. Biochem. 2009, 103 (1), 143-155.

6. Ponomarev, E.; Neumann-Spallart, M.; Hodes, G.; Lévy-Clément, C.,

Electrochemical deposition of MoS2 thin films by reduction of

tetrathiomolybdate. Thin Solid Films 1996, 280 (1-2), 86-89.

7. Belanger, D.; Laperriére, G.; Marsan, B., The electrodeposition of amorphous

molybdenum sulfide. J. Electroanal. Chem. 1993, 347 (1-2), 165-183.

8. Daeneke, T.; Dahr, N.; Atkin, P.; Clark, R. M.; Harrison, C. J.; Brkljaca, R.;

Pillai, N.; Zhang, B. Y.; Zavabeti, A.; Ippolito, S. J., Surface water dependent

100
 properties of sulfur-rich molybdenum sulfides: electrolyteless gas phase water

splitting. ACS Nano 2017, 11 (7), 6782-6794.

9. Mayyas, M.; Sahajwalla, V., Carbon nano-sponge with enhanced electrochemical

properties: A new understanding of carbon activation. Chem. Eng. Sci. 2019, 358,

980-991.

10. Lyu, J.; Mayyas, M.; Salim, O.; Zhu, H.; Chu, D.; Joshi, R., Electrochemical

performance of hydrothermally synthesized rGO based electrodes. Mater. Today

Energy 2019, 13, 277-284.

11. Ghasemian, M. B.; Mayyas, M.; Idrus‐Saidi, S. A.; Jamal, M. A.; Yang, J.;

Mofarah, S. S.; Adabifiroozjaei, E.; Tang, J.; Syed, N.; O'Mullane, A. P., Self-

limiting galvanic growth of MnO2 monolayers on a liquid metal-applied to

photocatalysis. Adv. Funct. Mater. 2019, 29 (36), 1901649.

12. Pitzer, K. S., Corresponding states for perfect liquids. J. Chem. Phys. 1939, 7 (8),

583-590.

13. Rah, K.; Eu, B. C., The generic van der Waals equation of state and self-diffusion

coefficients of liquids. J. Chem. Phy. 2001, 115 (6), 2634-2640.

14. DelRio, F. W.; de Boer, M. P.; Knapp, J. A.; Reedy, E. D.; Clews, P. J.; Dunn,

M. L., The role of van der Waals forces in adhesion of micromachined surfaces.

Nat. Mater. 2005, 4 (8), 629-634.

15. Liu, Z.; Zhang, X.; Wang, B.; Xia, M.; Gao, S.; Liu, X.; Zavabeti, A.; Ou, J.

Z.; Kalantar-Zadeh, K.; Wang, Y., Amorphous MoSx-coated TiO2 nanotube

Arrays for enhanced electrocatalytic hydrogen evolution reaction. J. Phys. Chem.

C . 2018, 122 (24), 12589-12597.

16. Lee, Y. H.; Zhang, X. Q.; Zhang, W.; Chang, M. T.; Lin, C. T.; Chang, K. D.;

Yu, Y. C.; Wang, J. T. W.; Chang, C. S.; Li, L. J., Synthesis of large‐area MoS 2

101
 atomic layers with chemical vapor deposition. Adv. Mater. 2012, 24 (17), 2320-

2325.

17. Ghatak, S.; Pal, A. N.; Ghosh, A., Nature of electronic states in atomically thin

MoS2 field-effect transistors. ACS Nano 2011, 5 (10), 7707-7712.

18. Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A., Single-

layer MoS2 transistors. Nat. Nanotechnol. 2011, 6 (3), 147-150.

19. The Materials Project. https://materialsproject.org/materials/mp-2815/ (accessed

Aug, 29, 2022).

20. Li, H.; Zhang, Q.; Yap, C. C. R.; Tay, B. K.; Edwin, T. H. T.; Olivier, A.;

Baillargeat, D., From bulk to monolayer MoS2: evolution of Raman scattering.

Adv. Funct. Mater. 2012, 22 (7), 1385-1390.

21. Liu, H.; Chi, D., Dispersive growth and laser-induced rippling of large-area

singlelayer MoS2 nanosheets by CVD on c-plane sapphire substrate. Sci. Rep.

2015, 5, 11756.

22. Jeon, J.; Jang, S. K.; Jeon, S. M.; Yoo, G.; Jang, Y. H.; Park, J.-H.; Lee, S.,

Layer-controlled CVD growth of large-area two-dimensional MoS2 films.

Nanoscale 2015, 7 (5), 1688-1695.

23. Najmaei, S.; Liu, Z.; Zhou, W.; Zou, X.; Shi, G.; Lei, S.; Yakobson, B. I.;

Idrobo, J.-C.; Ajayan, P. M.; Lou, J., Vapour phase growth and grain boundary

structure of molybdenum disulphide atomic layers. Nat. Mater. 2013, 12 (8), 754-

759.

24. Eda, G.; Yamaguchi, H.; Voiry, D.; Fujita, T.; Chen, M.; Chhowalla, M.,

Photoluminescence from chemically exfoliated MoS2. Nano Lett. 2011, 11 (12),

5111-5116.

102
25. Dhakal, K. P.; Duong, D. L.; Lee, J.; Nam, H.; Kim, M.; Kan, M.; Lee, Y. H.;

Kim, J., Confocal absorption spectral imaging of MoS2: optical transitions

depending on the atomic thickness of intrinsic and chemically doped MoS 2.

Nanoscale 2014, 6 (21), 13028-13035.

103
Chapter 4 Liquid-Metal-Assisted Deposition and Patterning of

Molybdenum Dioxide at Low Temperature

104
4.1 Introduction

Molybdenum dioxide (MoO2), due to its near-metallic conductivity and surface

plasmonic properties, is a great material for electronics, energy storage devices and bio-

sensing. Yet to this day, room temperature synthesis of large area MoO 2 that allows

deposition on arbitrary substrates has remained a challenge. Gallium-based liquid alloys,

due to their reactive interfaces and specific solubility conditions, offer unique

opportunities for synthesising materials that can meet these challenges.

In this chapter, the author presents a substrate-independent liquid metal-based method for

the room temperature deposition and patterning of MoO2. By introducing a molybdate

precursor to the surrounding of a droplet made of eutectic alloy of gallium-indium in the

liquid state, a uniform layer of hydrated molybdenum oxide (H2MoO3) was achieved at

the interface. This layer was then transferred onto the desired substrate. Utilising this

transferred H2MoO3 layer, a laser-writing technique was developed which selectively

transforms this H2MoO3 into crystalline MoO2 and produces electrically conductive

MoO2 patterns at room temperature. The electrical conductivity and plasmonic properties

of the MoO2 were analysed and demonstrated.

The content of this chapter was published as an article in the Journal of ACS Applied

Materials and Interfaces.1

4.2 Methods

4.2.1 Materials

Gallium (round shots, 99.9%) and indium (beads, 99.9%) were purchased from

Rotometals, USA. Sodium molybdate dihydrate (Na 2MoO4·2H2O, 99.97%) and bovine

105
serum albumin (BSA, ≥ 98%) was purchased from Sigma-Aldrich. All other chemicals

were purchased from commercial vendors and used as received unless otherwise noted.

Milli-Q water (18.2 MΩ·cm at 25 °C) was used in all experiments.

Eutectic gallium–indium alloy (EGaIn) was prepared by grinding 75.5 wt% of gallium

with 24.5 wt% of indium using porcelain mortar and pestle until they are thoroughly

mixed. In order to ensure complete alloying, the metal mixture was heated at 180 °C and

stirred with a glass rod and then cooled down at room temperature. The EGaIn alloy was

then transferred into a glass vial using a glass Pasteur pipette and stored in a nitrogen-

filled glove box for later use.

4.2.2 Synthesis of H2MoO3 Layer on EGaIn Surface

EGaIn droplet was placed in a plastic Petri dish (35 mm inner diameter, 36 mm outer

diameter and 15 mm height). The passive oxide layer on the surface of EGaIn droplet was

removed by washing with hydrochloric acid (HCl) solution (5.0 mL, 5M). The HCl

instantly removed the oxide layer from the EGaIn droplet and exposed its shiny metallic

surface. The HCl washing liquid was removed with a glass Pasteur pipette and replaced

with the aqueous solution of sodium molybdate (7.0 mL, 0.20 M). The light-yellow

interfacial layer of H2MoO3 gradually formed on the EGaIn droplet. During the reaction,

any disturbance of the droplet was avoided to ensure the formation of a smooth and

uniform H2MoO3 layer. After the desired thickness of deposition layer was reached, the

surrounding solution was gently removed and replaced with 7.0 mL Milli-Q water. This

was repeated three times to remove any residual sodium molybdate from the H2MoO3-

bearing EGaIn droplet. Afterward, the Milli-Q water was removed and the EGaIn droplet

was left to naturally dry at room temperature.

106
An intact H2MoO3-bearing EGaIn droplet free from track and deformation is crucial for

the touch-transfer process. Therefore, extra care was taken during the reaction and

washing steps to avoid disturbance of the droplet. It is noteworthy that although larger

EGaIn droplets produce H2MoO3 of larger areas on their surfaces, they become disturbed

and deform more easily compared to the smaller droplets, which makes the touch-transfer

process more challenging with larger liquid metal droplets.. In general, touch-transferring

from EGaIn droplets of around 120 μL yields repeatable and reasonably large area and

H2MoO3 layers (approximately 7-8 mm in diameter).

4.2.3 Touch-transfer of H2MoO3 Layer onto Desired Substrates.

The extraction and transfer of the H2MoO3 layers were conducted utilising a program

controlled piezo-electric stage (N-765 NEXACT precision z stage, 0.50 nm resolution,

25 N push/pull force, Physik Instrumente (PI) GmbH & Co. KG), which was reported in

our previous work.22 The detail of the process is described as follows. Substrates (Si wafer

or quartz) were cleaned in ultrasonic bath of acetone, ethanol, and Milli-Q water (10 mins

each, sequentially). Sonicated substrates were rinsed with Milli-Q water, and then blow-

dried with nitrogen. Prior to use, substrates were further cleaned with a Tergeo plasma

cleaner (Pie Scientific, USA) at 15 W for 2.0 min to eliminate any possible organic

contaminant and improve the adhesion of the substrate surface. For touch-transfer, the

substrate was fixed upside down on a holder above the H2MoO3-bearing droplet, and the

droplet in Petri dish was placed on a program controlled piezo-electric stage. The stage

lifted the droplet up at a speed of 0.05 mm s -1 until it was in contact with the substrate.

The contact was held for 600 s before the droplet was brought down at the same speed.

The H2MoO3 layer adhered onto the substrate due to the accumulation of van der Waals'

107
forces between them. After the transfer, the substrate was removed from the holder and

sonicated in ethanol to remove any residuary EGaIn. Upon sonication in ethanol, more

than 95% of the EgaIn could be removed. The substrate, with the transferred materials on

it, was then blow-dried with nitrogen and stored in a nitrogen-filled glove box for future

process and characterization. In order to ensure the repeatability of the touch-transferring

process, the droplet needs to be completely dry prior to the transferring onto substrates,

the substrate used should be freshly plasma-treated, and there needs to be sufficient

contact time between the substrate and the droplet before the lift-off. In the case of

flexible substrate like polydimethylsiloxane (PDMS), freshly made PDMS film was

applied with the same touch-transferring process, and no plasma-treatment was applied

due to the intrinsic adhesive surface of PDMS.

4.2.4 Thermal Annealing of H2MoO3

The annealing process was conducted in a Linkam’s HFS probe stage (HFS600E-PB4,

Linkam Scientific Instruments Ltd.) Prior to annealing, the Linkam stage was purged with

ultra-pure argon for 10 mins. During the annealing process, a constant argon flow to the

Linkam stage was maintained. Samples were annealed non-isothermally at heating rate

10 ⁰C min-1. For thermal transformation investigations, the H2MoO3 layer was held at 50,

100, 150, 200 , 250, 300, 350, 400, and 450 ⁰C for 200 seconds, during which a single

point Raman spectrum was collected at each temperature.

108
4.2.5 Laser Annealing Process

Patterning conductive molybdenum oxide was carried out by scanning selective areas on

the H2MoO3 layer with a laser (λ = 532 nm, 36 mW, spot size around 1 μm) sourced by

inVia Raman microspectrometer (Gloucestershire, UK). Patterned areas were laser-

annealed using a step size of 0.40 μm and monitored through 50 times magnification

objective lens. The exposure dose was controlled by adjusting the incident laser power

and the exposure time. For large areas, 100% laser power and 0.01 s exposure time were

applied, and a focus tracking mode was utilised to ensure a constant laser focus over time.

4.2.6 Electrochemical Characterization.

The electrochemical investigations were carried out using an electrochemical workstation

(CHI650E, CH Instruments Inc., USA). Cyclic voltammograms were acquired, in a three-

electrode system, with a gold-coated Si wafer (8.0 mm × 30 mm) as the working

electrode, a gold wire (0.25 mm diameter) as the counter electrode, saturated calomel

electrode (SCE) electrode as the reference electrode, and a plastic Petri dish (35 mm inner

diameter, 36 mm outer diameter and 15 mm height) as the cell. The electrolyte was an

aqueous solution of sodium molybdate (Na2MoO4, 0.20 M) and potassium chloride (KCl,

0.10 M). The pH value of the electrolyte was adjusted to 4.0 using HCl. Initial and final

voltages were both set to 0.0 V, and the scan rate varied from 20 mV s -1 to 80 mV s-1. The

open-circuit voltage (Voc) measurement over time was conducted in the same cell with

0.10 M KCl solution (pH adjusted to 4.0 using HCl) as the electrolyte. For V oc

measurement, an EGaIn droplet (120 µL) was placed as working electrode. The EGaIn

droplet was connected to the potentiostat using a gold wire (0.25 mm diameter)

109
encapsulated in a glass tube with its end exposed and immersed into the EGaIn droplet.

Prior to measure the Voc, the EGaIn droplet was washed with 5.0 M HCl solution.

4.2.7 Morphological and Structural Characterizations.

Atomic force microscopy (AFM) images were collected using a Bruker dimension icon

SPM (USA) with a ScanAsyst-Air probe (Spring constant: 0.40 N m -1, Bruker AFM

probes, USA). PeakForce tapping mode was applied with a scan rate of 0.50 Hz and a

resolution of 512 sample per line. Auto control of feedback gain and peak-force setpoint

was utilised for an optimized tracking of the surface. Specimens were stabilized on the

AFM stage with vacuum. Roughness and depth distribution profiles were obtained with

Nanoscope Analysis software (Version 1.70, Bruker).

A Philips CM200 field-emission high-resolution transmission electron microscopy

(TEM) system equipped with energy-dispersive X-ray spectroscopy (EDS, Bruker SDD-

EDS) was employed for high-resolution TEM image acquisition and chemical

compositions analysis. Digital FEI Nova NanoLab 200 scanning electron microscopy

(SEM) system equipped with a high-resolution Focused Ion Beam (FIB) workstation was

employed to prepare samples for cross-sectional TEM analysis.

4.2.8 Optical characterization

Cyclic X-ray photoelectron spectroscopy (XPS) analysis was carried out using a Thermo

Scientific ESCALAB250i high-resolution XPS (UK) with monochromatic Al Kalpha soft

X-ray source (1486.68 eV). For spectral acquisition, 120 W power (13.8 kV × 8.7 mA),

2×10−9 mbar background vacuum, 90° photoelectron take-off angle, and 500 µm spot size

were applied. As a reference for calibration, C1s = 284.8 eV for adventitious hydrocarbon

110
was used. Avantage XPS software (from Avantage Software Ltd, United Kingdom) was

employed for fitting and deconvoluting the high-resolution XPS spectra.

A Renishaw inVia Raman microspectrometer (Gloucestershire, UK) equipped with 532

nm laser was utilised for obtaining Raman spectra. Spectra were collected with standard

confocality and recorded at static grating scan type with 1800 lines mm -1 grating, which

provided the spectral resolution of ~1.5 cm-1. One second exposure time, and 50 times

magnification were used for all single-point Raman measurements. To avoid laser-

induced modification of the sample property, 1.0% laser power (0.36 mW on sample),

with 200 accumulations, was applied unless otherwise noted.

For surface plasmonic measurements, transparent quartz substrates were utilised and the

same touch transferring, and laser annealing methods were applied. BAS solution (0.1 mg

mL-1) was prepared in phosphate-buffered saline buffer solution (0.1 M, pH=7.4). For

BSA treatment, the sample was soaked in freshly prepared BAS solution for 15 mins

before it was rinsed in Milli-Q water and blow dried with nitrogen. The optical absorption

spectra were obtained using an ultraviolet visible spectrometer (UV-Vis, Agilent

Cary5000). A blank quartz substrate was used for baseline correction.

4.2.9 Conductivity characterization

For preparing specimens for conductivity measurement, Si-SiO2 wafer with micro sized

gold pattern was used as the substrate to provide contact for the measurements. Gold

patterns were fabricated using conventional ultraviolet (UV, 365 nm) lithography, by

depositing 50 nm of gold on a 5 nm adhesion layer of chromium through a lift-off

photoresist (AZ1505) that was patterned using a maskless writing system (Heidelburg

111
Instrument Inc.). H2MoO3 layer was transferred onto the substrate using the same touch-

transfer process, and then MoO2 pattern was made using the same laser annealing process.

PeakForce Tunneling-AFM (PF-TUNA) measurement of the conductive patterns were

carried out using a Bruker dimension icon SPM (USA) with platinum coated conductive

tip (SCM-PIT-V2, Bruker AFM probes, USA). Sample was fixed on a metal disk with a

stick tap, and copper tape was used for providing the contact between the gold lead and

metal disk. Metal disk was stabilized on the AFM stage with vacuum during the

measurement. PeakForce tapping mode was applied with a scan rate of 0.50 Hz and a

resolution of 512 sample per line. Auto control of feedback gain and peak-force setpoint

was implied for optimized tracking of the surface. For open loop PF-TUNA measurement,

DC sample bias was 0.50 V, Current sensitivity is 20 pA V-1. For closed loop PF-TUNA

measurements, the feedback bias was turned on, and feedback setpoint was set to 10 pA,

feedback IGain was 15, and bias limit was between 0.0 and 2.0 V. For acquiring I-V

curves for points of interest, continuous ramp at each point was applied until the tip was

on the surface, and then I-V curve was collected at a scan rate of 0.20 V s -1.

4.2.10 Computational studies

Simulations and band-structure calculations were performed using density functional

theory (DFT) in the Vienna Ab Initio Simulation Package (VASP) version 5.4.4 68-70. The

calculations used the generalized gradient approximation (GGA) to the exchange-

correlation energy as implemented by Perdew et al. (GGA-PBE). The projector

augmented-wave method was used to represent the ionic cores, 68-70 and the number of

electrons treated as valence was 14 for Mo and 6 for O. Wave functions were represented

in a plane-wave basis and the cut-off energy was 400 eV. To treat the electronic

112
correlations in the Mo 3d orbital, the calculations included an additional Hubbard term

using the simplified scheme proposed by Dudarev et al.71 with a value of Ueff= 6.3 eV

chosen to reproduce the α-MoO3 bulk band-gap and be consistent with past work72 For

the crystalline 2D and 3D models, we used a dense q-point grid of 8×8×8 for the first

Brillouin zone sampling centred on the Γ-point. To simulate an oxygen vacancy, 2×1×2

supercells were constructed and a single oxygen was removed, initially at the O2 position

which has the lowest formation energy (in bulk). 72 For the 2D models, a 15 Å vacuum

slab was introduced to break the periodic boundary conditions in one direction, and 2D

layers were formed by cleaving at the van-der Waals gap of the bulk α-MoO3 crystal

structure which occurs at a specific (010) plane. Structures were relaxed until forces were

converged to better than 0.02 eV/Å and the total energy convergence threshold was 10 -6

eV. To simulate the polycrystalline nature of the 2D α-MoO3 surface layer, a grain

boundary model was constructed by matching the [100]||[001] directions of the 2D layer,

as this was indicated to be the best lattice match available for a coherent interface, and

then performing ionic relaxation using the same convergence criteria outlined above,

except with a single K-point at Γ. Calculations were performed on the GADI

supercomputer which is part of the Australian National Computer Infrastructure.

4.3 Results and discussion

4.3.1 Thickness-controlled deposition of H2MoO3 sheet on liquid metal

The deposition process of the H2MoO3 was achieved as schematically illustrated in Figure

4.1. Before adding the precursor solution, the EGaIn droplet was treated with 5.0 M HCl

solution to remove the gallium oxide passivating layer and to activate the EGaIn surface.

After the precursor solution of 0.20 M Na2MoO4 was added onto the EGaIn droplet, the

113
deposition process begins and a layer of H2MoO3 gradually forms on EGaIn surface as

schematically illustrated in Figure 4.2A. Upon the acid treatment, the EGaIn droplet

generates a galvanic surface potential that leads to the accumulation of electrons at the

interface which attracts cations (mostly H+), forming the electrical double layer (EDL).

The interfacial autogenous potential and excess electrons create an ideal environment to

drive reduction reactions. The MoO42- species, when introduced into the EDL, undergo a

reduction reaction at the interface as described by Equation 4.1, with the gallium at the

interface serving as reducer and offering electrons to Mo species (Equation 4.2). As a

reduction product, a uniform layer of H2MoO3 is then templated onto the atomically

smooth liquid-liquid interface between the EGaIn droplet and the precursor solution.

MoO42- +2e- +4H+ → H2 MoO3 +H2 O (4.1)

-
Ga→ Ga3+ +3e (4.2)

Figure 4.1. Schematic illustration of the synthesis process of H2MoO3.

To verify this reaction, cyclic voltammograms of the Na2MoO4 precursor solution were

acquired as shown in Figure 4.2B. In the cathodic region, the reduction peak at -0.24 V
114
is attributed to the reduction of the MoO42- species into H2MoO3.2 The onset potential for

the electrochemical transformation of MoO42- species into H2MoO3 was accordingly

determined to be -0.08 V as highlighted in Figure 4.2B. The oxidation peak at around

0.51 V corresponds to the oxidation of reduced molybdates. The galvanic potential

available on the acid activated EGaIn droplet was assessed with an V oc measurement in

an aqueous solution that simulates the same acidity of the precursor solution (Figure

4.2C). The measured surface potential of -0.4 V is significantly larger than that -0.08 V

which is the potential needed to trigger the reduction of MoO 42- species as determined

from the cyclic voltammograms in Figure 4.2A. This means that the surface potential of

EGaIn droplet is sufficient to reduce the MoO42- spontaneously and induce the template

deposition of H2MoO3 layer onto the liquid-liquid interface.

Figure 4.2 (A) Schematic illustration of H2MoO3 layer formation on EGaIn surface. (B)

Cyclic voltammograms of 0.20 M Na2MoO4 aqueous solution acquired at different scan

rates, scanned in the potential window from 0.80 to -0.40 V (negative scan), initial and

final voltages were both set to 0.0 V vs. SCE. (C) The galvanic surface potential of the

EGaIn droplet was measured over time in 0.10 M KCl (solution pH was adjusted to 4.0

using HCl).

115
The extraction and transfer of the deposited layer were conducted using a touch-transfer

technique as schematically presented in Figure 4.3 and detailed in Section 4.2.3. In order

to improve the reproducibility and stability of the transfer, a program controlled piezo-

electric stage with a nanometre spatial resolution was utilised. Without using a highly

controlled and high-resolution piezoelectric stage, the chance of failure to obtain high

quality and large area films, without any cracks or holes, would increase. In brief, a pre-

cleaned smooth substrate was brought into contact with the H2MoO3-bearing EGaIn

droplet. Atomic scale contact at large area could be achieved due to the ultra-smooth

surface of the H2MoO3 layer and the liquidity of the EGaIn underneath which acted as a

soft cushion. The non-polar nature of the EGaIn surface results in a weak interaction

between the surface layer and the parent metal. On the other hand, the cumulative van der

Waals forces between the smooth substrate and the layer on the surface of liquid metal

enabled the layer extraction onto the substrate upon lifting-off with the piezo-electric

stage. It is noteworthy that this touch-transferring technique is material independent, as

long as the surface of both the deposited material and the substrate is smooth enough for

the formation of the van der Waals force between them. Residual materials from EGaIn

that adhered on the transferred layer were removed by a sonication bath in ethanol later.

Figure 4.3 Schematic illustration of the touch-transferring process using a nano-

positioning device.

116
As shown in the optical images (Figure 4.4), the exfoliated and transferred layers feature

a uniform sheet of large area that is free of residual EGaIn. Various substrates including

a SiO2/Si rigid substrate and a flexible PDMS substrate are applied to prove the substrate-

independence of the touch-transfer technique.

Figure 4.4 Optical microscopy images of transferred H2MoO3 film on different

substrates.

AFM analysis further verified the transferred layer to be smooth and uniform (Figure 4.5).

It is noteworthy that the thickness of the layer can be controlled by the reaction duration

in a wide range from a few nanometres to ~100 nm (as per Figure 4.6 and 4.7).

117
Figure 4.5. An AFM image of H2MoO3 films.

Figure 4.6 AFM images of H2MoO3 films of different thickness.

118
Figure 4.7 The box chart presenting the thickness distribution of materials yield at

different reaction time (3 s, 2 mins, 5 mins, 20 mins, 1.0 h).

The chemical composition of the resultant H2MoO3 layer was analysed by XPS and

Raman spectroscopy. As shown in the high-resolution XPS depth profile of the as-

transferred H2MoO3 layer at Mo 3d region (Figure 4.8A), one doublet with the binding

energies of 229.4 and 232.9 eV could be observed, corresponding to Mo 4+ 3d5/2 and Mo4+

3d3/2 peaks, respectively.3 The peak locations show negligible changes over depth, which

means that the chemical state of molybdenum is identical at the surface and bulk of the

H2MoO3 layer. To avoid substrate interference, Raman spectrum of the as-deposited

H2MoO3 layer in Figure 4.8B was collected directly from a H2MoO3-bearing EGaIn

droplet (before the material was touch-transferred). The broad peak at 775 cm -1

corresponds to the stretching mode of triply coordinated oxygen-molybdenum bonds (vO-

Mo3). The small peak at 850 cm-1 corresponds to the stretching mode of a doubly

coordinated oxygen-molybdenum bond (vO-Mo2). The peak at around 920 cm-1

corresponds to the stretching vibration of the terminal oxygen molybdenum bond

119
(vMo=O).4-6 The peak group between 180 to 400 cm -1 agrees with the previous reported

hydrated molybdenum oxide.7

Figure 4.8 Chemical characterization of the as-deposited H2MoO3 sheet. (A) High-

resolution XPS depth profile of the as-transferred H2MoO3 layer at Mo 3d region and (B)

Raman spectrum of the H2MoO3 layer collected while on liquid metal droplet.

4.3.2 Thermal annealing of H2MoO3: dehydration and crystallization

The thermal transformation of the H2MoO3 was investigated via annealing the material

in an ambient controlled stage (Linkam, UK) under an inert atmosphere (Argon gas). As

can be seen in the optical microscopy images (Figure 4.9A), during the annealing, a few

blue dots started to emerge in the jade green H2MoO3 layer, which gradually expanded

throughout the entire layer. Such an observation could be associated with a phase

transformation and/or thickness change process that have been initiated at 350 ⁰C. To

verify this hypothesis, Raman spectra of the material were acquired before and after

thermal annealing. The Raman spectra of the H2MoO3 layer on SiO2/Si substrate before

and after annealing are presented in Figure 4.9B, black and red lines, respectively. To

investigate the temperature at which the thermal transformation occurs, in situ Raman

measurements were carried out during the thermal annealing process. As presented in
120
Figure 4.9C, the intensities of the two peak groups of the H2MoO3 gradually vanish in the

temperature range of 150 ⁰C to 350 ⁰C. When the sample is heated from 350 ⁰C to 400 ⁰C

(highlighted by the grey dashed rectangle), however, an abrupt change can be observed

in the Raman spectra, with three new peaks at 200, 355 and 735 cm -1 emerging. These

new peaks are in alignment with the characteristic peaks of MoO2, suggesting a phase

transformation process from H2MoO3 to MoO2 which starts at 350 ⁰C. The produced

MoO2 remains stable after heating up to 450 ⁰C as shown in Figure 4.9C.

Figure 4.9 Thermal transformation of H2MoO3 layer. (A) Optical microscopy images of

a H2MoO3 layer transforming into crystalline molybdenum in the temperature range

between 350 ºC and 400 ºC. (B) Raman spectra of the amorphous H 2MoO3 layer before

annealing and the crystalline MoO2 after annealing (on SiO2/Si substrate) and (C) in situ

Raman measurements showing the thermal transformation of H2MoO3.

121
The cross-section of the thermally annealed product was obtained using FIB milling

(Figure 4.10). A Pt protective layer was coated on the sample prior to the milling process.

The EDS mapping of the sample’s cross section shows a consistent distribution of Mo

and O sandwiched between the Si-SiO2 substrate and the Pt (Figure 4.11). The high-

resolution TEM in Figure 4.12 suggests a polycrystalline sample with a d-spacing of 3.4

Å corresponding to the (111) plane of the monoclinic MoO2 crystal.8

Figure 4.10 SEM-FIB process and the resultant sample for cross-sectional TEM analysis.

Figure 4.11 EDS mapping of the TEM sample (scale bar: 100 nm).

122
Figure 4.12 High-resolution TEM images of the thermo-annealed MoO2.

4.3.3 Laser annealing: A direct laser writing of conductive MoO2 patterns

Other than thermal annealing, the transformation of H2MoO3 into MoO2 could also be

induced via high-power laser exposure. This approach has the advantage of annealing

selective regions on the H2MoO3 at room temperature and creates conductive patterns of

MoO2 (that will be discussed later). As schematically presented in Figure 4.13A, a

scanning laser (λ = 532 nm, beam size of 1 µm) was utilised for transforming selected

regions of H2MoO3 layer into MoO2. The laser annealing at different powers and exposure

times were conducted for parameter optimization. As shown in Figure 4.14, laser power

of 18 mW led to the best results with strong MoO2 Raman signals. A higher laser power

of 36 mW resulted in weaker MoO2 Raman signals, which indicated excessive damage to

the material. It’s noteworthy that variation in exposure time resulted in no significant

difference in the annealing outcomes, which indicated that 1 ms exposure time was

sufficient for full phase transformation.

The chemical composition of the laser-patterned material was assessed with the high-

resolution XPS. Figure 4.13B presents the XPS depth profile of the laser-annealed

materials at the Mo 3d region. At the surface of the material (depth=0), two peaks at 232.9

123
eV and 236.1 eV are attributed to the Mo6+ 3d5/2 and Mo6+ 3d3/2 peaks are observed,9

suggesting the presence of a very thin layer of MoO3 on the surface. According to the

peak deconvolution results (Figure 4.15), the XPS signal also consists of two small peaks

at 235.2 and 231.1 eV, which correlates to a small amount of Mo with lower oxidation

state induced by oxygen vacancies. Beneath the MoO3 layer with high oxygen vacancy,

a significant shift toward lower binding energies, i.e., lower oxidation state of Mo, is

observed with increasing depth. At a depth from 5 to 85 nm, the Mo 3d 5/2 peak gradually

shifts from 229.5 to 229 eV, and the peak Mo4+ 3d3/2 gradually shifts from 232.7 to 232.3

eV, indicating the materials at this range to be slightly nonstoichiometric MoO 2. From

the XPS depth profile analysis, it is concluded that the produced material consists of

nonstoichiometric MoO2 covered by a thin layer of MoO3 with oxygen vacancies at the

surface.

Figure 4.13 (A) Schematic illustration of the laser-patterning process utilising Raman

laser (532 nm). (B) High-resolution XPS depth profile of the laser annealed molybdenum

oxide layer at Mo 3d region. The black dashed rectangle highlights the Mo6+ on the

surface, while the white dashed rectangles highlight the Mo4+ corresponding to MoO2

underneath.
124
Figure 4.14 Optical images and Raman spectra of materials annealed at different laser

powers and exposure times. (A-C) Optical images of materials acquired at 10%, 50%,

and 100 % laser powers, respectively. (D-F) Raman spectra of materials acquired at 10%,

50%, and 100 % laser powers, respectively. Characteristic peaks of MoO2 were marked

with stars.

Figure 4.15 Peak fitting of the high-resolution XPS of the surface of the laser-annealed

MoO2 film (depth 0- 1.8 nm).

125
Figure 4.16A presents an optical microscopy image of a diamond-shaped pattern. After

the laser annealing process, the sample colour transforms from jade green into a dark blue

as a result of change in both the thickness and chemical composition. The colour change

is in agreement with the colour change induced by the thermal transformation in Figure

4.9A (for comparable thicknesses). Raman analysis was utilised to further investigate the

chemical composition of the laser annealed sample. Raman spectra of the pristine

H2MoO3 layer (red spectrum) and the laser-annealed material (black spectrum) are

presented in Figure 4.16B. After laser annealing, the broad peaks of H2MoO3 (centred at

300 and 770 cm-1) film disappears, and new peaks of MoO2 at 200, 355 and 735 cm-1

emerge instead. The emerged Raman peaks observed in the annealed area are consistent

with those obtained by thermal annealing (Figure 4.9C). This further confirms that the

transformation of H2MoO3 to MoO2 can be achieved by the laser annealing technique

proposed here. It is noteworthy that due to the ultrathin nature of the surface MoO 3 layer,

no Raman signal from it is observed. 10, 11 It is well established that the most stable phase

of MoO3 is the α-MoO3. The other phases of MoO3 such as (β-phase, ε-phase) are

metastable.37 Hence, it is fair to deduce that the surface layer of MoO3 is likely to be alpha

phase.

Raman mapping of the diamond-shaped pattern was obtained to verify the compositional

distribution and uniformity. Raman intensities at 200, 355, 735 cm-1 (characteristic peaks

of MoO2) and 773 cm-1 (characteristic peak of H2MoO3) are highlighted using dotted lines

in Figure 4.16B and their spatial distributions are presented in Figure 4.16C-F,

respectively. The purple, green, and blue maps (Figure 4.16C-E) correspond to the spatial

distribution of MoO2 while the red map (Figure 4.16F) represents the distribution of

H2MoO3. The Raman maps at 200, 355 and 735 cm -1 feature strong intensity in the

diamond-shaped pattern indicating a uniform distribution of MoO 2. The maps suggest

126
that the laser annealing technique can selectively anneal desired areas without affecting

the surrounding materials.

Figure 4.16 (A) Optical microscopy image of molybdenum oxide sheet transferred onto

SiO2 substrate and annealed by exposure to a high-power laser. The darker diamond-

shaped area is the annealed material. (B) The Raman spectra of the molybdenum oxide

layer collected from non-annealed (red) and annealed (black) areas. Characteristic Raman

peaks of MoO2 at 200, 355 and 735 cm-1, marked by purple, green and blue dashed lines,

respectively, and the stretching mode of triply coordinated oxygen in the amorphous

H2MoO3 at 773 cm-1, marked by a red dashed line. Raman signals of Si substrate are

marked with yellow triangles. (C-F) Raman intensity mapping at 200, 355, 735, and 773

cm-1, respectively.

To further investigate the optical properties, the H2MoO3 layer was printed onto

transparent quartz substrates. Optical images and UV-Vis absorption spectra of the

127
printed H2MoO3 layer and MoO2 on these substrates were obtained. As can be seen in

Figure 4.17A, the optical images of the as transferred H2MoO3 layer features a pale-

yellow colour, which transforms into a dark blue colour upon laser annealing. As shown

in the UV-Vis absorption spectra in Figure 4.17B, the H2MoO3 layer shows little

absorption in the visible range. However, after laser annealing, a broad absorption band

at visible to near-infrared (NIR) region emerges, which could be attributed to the surface

plasmonic resonance effect induced by the collective oscillations of d-orbital free

electrons.12 It is known that BSA could be immobilized and adsorbed onto the

molybdenum oxide surface due to the electrostatic interactions and the collective van der

Waals’ forces between them. 13, 14 The negatively charged BSA repels the free electrons

on the surface and decreases its surface free electron intensity, which leads to the

reduction of the plasmonic resonance peak intensity.13, 14

To further investigate the

plasmonic properties of the laser annealed material, samples were treated with BSA and

the variation in its UV-vis absorption behaviour and electronic property was assessed. As

shown in Figure 4.17C, the plasmonic resonance peak at visible to NIR region

significantly reduced upon 15 mins treatment in a BSA solution. It is presented in Figure

4.18 that the material conductivity decreases due to the adsorption of BSA on the surface

of the film, which consequently reduced the thickness of the conductive channel and thus

increased the resistivity in response. Such observation suggests that the MoO 2 produced

here can be potentially utilised for plasmonic biosensing.

128
Figure 4.17 (A) Optical images of the pale yellow H2MoO3 layer and the dark blue laser

annealed molybdenum dioxide layer. (B) The UV-Vis absorption spectra of H 2MoO3

(yellow spectrum) and laser annealed molybdenum dioxide (black spectrum). A broad

absorption band centered at around 800 nm emerged upon laser annealing as highlighted

by a black arrow. (C) The UV-Vis spectra of the laser annealed molybdenum oxide before

(black spectrum) and after (purple spectrum) BSA treatment.

Figure 4.18 I-V curves of the MoO2 based device before and after the BSA treatment.

129
4.3.4 Morphology properties of the MoO2 patterns

To investigate the morphology variation induced by the laser annealing process, AFM

and SEM were employed. Figure 4.19A presents the AFM analysis of the laser annealed

area. As shown in the AFM height profile, the layer experiences a decrease in thickness

by 45 nm upon the laser exposure. High resolution AFM image of the laser annealed area

in Figure 4.20 shows a grainy feature. According to the roughness calculation and the

depth distribution profile (Figure 4.21), average roughness (Ra) of the material increases

about one order of magnitude (from 0.922 nm to 8.81 nm). The AFM analysis suggests

that the laser exposure process thins and roughens the layer. The SEM image of the laser

annealed area is presented in Figure 4.19B, and the zoomed-in images of the material

before and after laser annealing are shown in Figure 4.19C and D, respectively. The layer

before annealing appears to be smooth and uniform, while the annealed material presents

a rough and grainy surface. This observation is in alignment with the AFM results. High-

resolution TEM was utilised to assess the crystallinity and structure of the laser annealed

samples. As shown in Figure 4.19E, the laser annealed area is polycrystalline with distinct

crystallites and grain boundaries as highlighted by yellow dashed lines. In Figure 4.19F

and G, lattice spacings of 0.34 nm are highlighted, which correspond to the (1 11) and

(111) crystal planes of the monoclinic MoO2.8 Figure 4.19H presents the crystal structure

of the monoclinic MoO2 at orientations corresponding to the observed lattice fringes in

Figure 4.19F and G. It is concluded from the chemical and morphological analysis that

the laser exposure produced materials consists of grains of slightly nonstoichiometric

MoO2 in the bulk, and ultra-thin α-MoO3 on the surface, which is schematically presented

in Figure 4.19I.

130
Figure 4.19 Topological and structural characterization of the laser-annealed

molybdenum oxide. (A) The AFM image of the laser-annealed area. The inset is the

topography profile showing the thinning of the sample by 45 nm after laser annealing.

(B) SEM image of a diamond-shaped pattern of molybdenum dioxide produced by laser-

annealing. (C) SEM image of the non-annealed amorphous molybdenum hydroxide. (D)

SEM image of the laser-annealed molybdenum dioxide. (E) High-resolution TEM image

of the laser-annealed molybdenum dioxide layer showing grain boundaries as highlighted

by yellow dashed lines. The inset is the selected area electron diffraction (SAED) pattern

of the laser-annealed molybdenum oxide. (F, G) High-resolution TEM image of the laser-

annealed molybdenum dioxide layer, showing the d-spacing of the (1, 1, 1) and (1, 1, 1)

planes. (H) The representation of the crystal structure of monoclinic molybdenum

dioxide. (I) Schematic representation of the composition and grainy structure of the laser-

annealed molybdenum oxide layer.

131
Figure 4.20 High resolution AFM topography of the laser annealed sample.

Figure 4.21 Roughness calculations based on AFM imaging of a sample (A) before and

(B) after laser patterning.

4.3.5 Theoretical calculation: Impacts of oxygen vacancies and grain-boundaries on

band structure of 2D-MoO3

DFT calculations were carried out to assess the electronic structure of the laser produced

materials.15-17 A Hubbard correction was incorporated to accurately model the Mo 3d

orbitals, with the U parameter adopted from past work, with further details given in

section 4.2.10.18, 19 Figure 4.22A and B shows the unit cell structure and the electronic

band structure of bulk MoO2 which is semi metallic, and has a zero-gap feature between
132
conducting and valence bands at the Fermi level. The bulk, stoichiometric α-MoO3, on

the other hand, is a wide band gap insulator as concluded from the electronic band

structure calculations (Figure 4.23A and B). Pristine 2D α-MoO 3 as shown in Figure

4.22C, also presents a similar wide band gap structure as the bulk α-MoO3 (Figure 4.22D).

On the contrary, our DFT calculations show that oxygen vacancies (incorporated into the

O1/O2/O3 positions as highlighted in Figure 4.23C) in the 3D α-MoO3-x unit cell can

introduce defect states in the band structure (Figure 4.23D). Similar defect states were

observed in the band structure calculations for 2D MoO3-x with the oxygen vacancy in the

O3 position which is the energetically preferred site in 2D (Figure 4.22E and F). For any

type of oxygen vacancy, the defects lead to in-gap states in the calculations explaining

the finite conductivity20 in the laser-induced MoO3 surface layer. The presence of oxygen

vacancies in the thin MoO3 layer was also experimentally confirmed from the XPS

analysis in Figure 4.15.

Since the laser processed sample features a grainy structure and XPS showed that α-

MoO3 appears on the surface of the grains, the density of states (DOS) of the grain

boundaries of α-MoO3 were assessed. As observed from the DOS diagram in Figure

4.22G, the grain boundaries of α-MoO3 (the boundary structure in Figure 4.24A) can have

a narrower band gap owing to structural disorder, compared to the pristine bulk and 2D

α-MoO3. The band-tails from disorder can potentially provide conductive boundary states

between the grains. Furthermore, adding a vacancy in bridging O position of the coherent

[100]||[001] boundary (Figure 4.24B) increases the DOS at the Fermi level as presented

in Figure 4.23H, leading to more conductive boundaries. The DFT calculation results

suggest that the laser patterned molybdenum oxide is conductive in the bulk via the

underlying MoO2 and also conductive on the surface, at the grain boundaries, through the

boundary states and oxygen vacancies of the α-MoO 3-x surface layer.

133
Figure 4.22 First principal calculation results. (A) Crystal structure of monoclinic MoO 2

(B) Band structure of MoO2. (C) Crystal structure of 3D-MoO3 (D) Band structure of 2D-

MoO3 (E) Crystal structure of 2D-MoO3-x with an oxygen vacancy. (F) Band structure of

2D-MoO3-x with an oxygen vacancy. (G) DOS diagram of the grain boundary of 2D

MoO3. (H) DOS diagram of the grain boundary of 2D MoO3-x with an oxygen vacancy at

the bridging oxygen position.

134
Figure 4.23 (A and B) Crystal structure and band structure of 3D MoO 3. (C and D)

Crystal structure and band structure of 3D MoO3 with an oxygen vacancy.

Figure 4.24 (A) Structure of the modelled grain boundary of 2D MoO3. (B) Structure of

the modelled grain boundary of the 2D MoO3 with oxygen vacancy at bridging oxygen

position.

135
4.3.6 Peak-force tunnelling AFM measurement for conductivity assessment

To verify our hypothesis and the outcomes of the DFT calculations, the conductivity of

the MoO2 pattern was assessed. In this regard, peak force tunneling AFM (PF-TUNA)

was used for obtaining the conductivity of the patterned samples. In this case, the

specimens were fabricated by printing the H2MoO3 layer onto a gold-patterned Si/SiO2

substrate as schematically presented in Figure 4.25A. Conducting patterns of

molybdenum dioxide were then written by laser exposure between and on the gold

patterns. The gold pattern was connected to the AFM chuck to close the circuit. A constant

bias was applied between the sample and the conductive AFM tip, as schematically

presented in Figure 4.25B, and the current through the tip was measured as it scanned the

sample surface.

As shown in Figure 4.25C, a diamond shape pattern of molybdenum dioxide was written

on gold using a laser exposure time of 10 ms point -1. The corresponding AFM and PF-

TUNA measurements are presented in Figure 4.25D and E. The AFM topography

analysis presented in Figure 4.25D shows a thickness decrease by ~40 nm after annealing

(thickness profile is provided in Figure 4.26A). The TUNA current mapping (Figure

4.25E) shows a nearly homogeneous but discontinuous surface conductivity throughout

the pattern. As shown in the TUNA current profile (Figure 4.26B), multiple sharp peaks

of current signal can be clearly observed. No current signal was observed from the

surrounding non-annealed area, reflecting the insulating nature of the H 2MoO3 layer.

Single point I-V curves were collected from the laser-annealed area and non-laser-

annealed area as presented in Figure 4.25F. The I-V curve from laser-annealed area shows

a linear I-V characteristic, which proves the material to be conductive. The I-V curve

from non-laser-annealed areas, on the other hand, show insignificant current at the voltage

of 5V, which indicate the poor conductivity of the non-annealed H 2MoO3 layer.
136
Linear patterns (Figure 4.25G) were also written by laser annealing at different laser

exposure times (1, 10 and 100 ms point-1) between the gold patterns. As shown in Figure

4.25H, no significant morphological variations are observed except that the higher

exposure time at 100 ms point-1 leads to wider valleys compared to those of 1.0 and 10

ms point-1, which can also be observed in the thickness profile provided in Figure 4.26C.

The corresponding PF-TUNA current image (Figure 4.25I) and the current profile (Figure

4.26D) show that those patterns obtained at laser exposure time of 1 and 10 ms point -1 are

similar in terms of current magnitude and homogeneity, while the higher exposure time

of 100 ms point-1 led to a slightly higher current but less homogeneity due to laser

overexposure. In addition, given that the PF-TUNA measurement was carried out on the

materials in between the gold pattern, contact was provided entirely by the MoO2,

indicating the excellent electrical conductivity continuity for the product.

Despite the MoO2 is covered with a wide band gap material (α-MoO3), the conductivity

measurement suggests that the surface of the annealed material has conductive regions.

This can be a result of either boundary states, as a result of the increased DOS around the

Fermi level. This is induced by oxygen vacancies of α-MoO3 as predicted earlier by the

DFT calculations and is consistent with the surface-resolved XPS measurements.

Interestingly, the high-resolution PF-TUNA measurement (Figure 4.25J and K) reveals

conducting patterns that are consistent with the grain boundaries of the surface layer of

the α-MoO3 (as presented in Figure 4.25L, the overlapped image of the high-resolution

topography and TUNA current).

These results explain the discontinuous current signal in the PF-TUNA measurement in

Figure 4.25E where the discontinuous current signal originates from the grain boundaries

of the α-MoO3 due to boundary states and/or oxygen vacancies at the boundaries as

supported by our DFT calculations. Further high-resolution PF-TUNA measurements are

137
provided in Figure 4.27 confirming the reproducibility of these conductive boundaries. In

addition, the laser patterning and conductivity measurement were carried out on the

flexible PDMS substrate. In Figure 4.28, the I-V curve of the MoO 2 pattern on PDMS

shows a linear characteristic with good conductivity. This verified the compatibility of

the process with soft substrates such as PDMS.

Figure 4.25 The conductivity assessments of the laser-annealed patterns. (A) Schematic

illustration of the laser-patterning process utilising a focused beam laser at 532 nm. (B)

Schematic illustration of the open loop PF-TUNA measurement. (C) Optical microscopy

image of a diamond shaped MoO2 pattern made by laser annealing. (D) The AFM image

of the laser annealed pattern. (E) TUNA current channel of the open loop PF-TUNA

measurement of the laser annealed pattern. (F) I-V curve measured from points 1 and 2

(highlighted in Figure E) from laser-annealed area and non-laser-annealed area,

respectively. (G) Optical microscopy image of the linear MoO 2 pattern made by laser

annealing. (H) The AFM image of the linear MoO2 pattern. (I) TUNA current channel of
138
the open loop PF-TUNA measurement of the linear MoO 2 pattern. (J) High resolution

AFM image of the laser-annealed area, where grains of various sizes can be observed.

(K) TUNA current channel of the open loop PF-TUNA measurement (high resolution).

(L) The overlapped image of the high-resolution topography and TUNA current channel.

Figure 4.26 (A) Height profile of Figure 4.25D. (B) Current profile of Figure 4.25E. (C)

Height profile of Figure 4.25H. (D) Current profile of Figure 4.25I.

139
Figure 4.27 Further high-resolution PF-TUNA measurements. Left column: height

profiles. Middle column: current profiles. Right column: the overlapped images of

morphology and current signals.

140
Figure 4.28 The I-V curve of the laser annealed MoO2 on PDMS, with the optical

microscopy image of the laser annealed sample inserted.

4.5 Conclusions

The author demonstrated the deposition of large area and uniform H2MoO3 layers on

EGaIn droplet, where Na2MoO4 aqueous solution as the precursor and the built-in

potential at the EGaIn surface served as the driving force of the reaction and the smooth

surface templating. Due to the liquidity and non-polarity of EGaIn, the resultant H 2MoO3

layer could be transferred onto any desired substrate via a room temperature touch-

transferring process using a piezo-stage. The transferred hydrated and amorphous

molybdenum oxide layer could go through the dehydration and crystallization process by

both thermal annealing and also laser exposure at room temperature. These processes

transformed H2MoO3 into conductive MoO2. By exposing the H2MoO3 layer to an

optically focused laser beam, the H2MoO3 layer could be patterned into conductive MoO2

at selective areas, which was verified by Raman mapping and PF-TUNA. An atomically
141
thin surface layer of conductive MoO3-x could also be detected on the laser processed

layers. The high-resolution PF-TUNA measurements show higher conductivity at the

edges and boundaries of the grains, which according to the XPS analysis and DFT

calculations, could be attributed to oxygen-vacancy-introduced in-gap defect states, and

the band-tails along the grain boundaries from disordered structures. Additionally, the

MoO2 layer was shown to have plasmonic effects, and was employed for a proof-of-

principle biosensing demonstration. The results reported here establish a future protocol

for creating conductive patterns for electronic applications that can be all processed at

room temperature which can be incorporated onto any substrate with no limitation.

4.6 References

1. Wang, Y.; Mayyas, M.; Yang, J.; Ghasemian, M. B.; Tang, J.; Mousavi, M.;

Han, J.; Ahmed, M.; Baharfar, M.; Mao, G., Liquid-metal-assisted deposition

and patterning of molybdenum dioxide at low temperature. ACS Appl. Mater.

Interfaces 2021, 13 (44), 53181-53193.

2. Li, W.; Tian, L. P.; Huang, Q.; Li, H.; Chen, H.; Lian, X., Catalytic oxidation

of methanol on molybdate-modified platinum electrode in sulfuric acid solution.

J. Power Sources 2002, 104 (2), 281-288.

3. Spevack, P. A.; McIntyre, N., A Raman and XPS investigation of supported

molybdenum oxide thin films. 1. Calcination and reduction studies. J. Phy. Chem.

1993, 97 (42), 11020-11030.

4. Ajito, K.; Nagahara, L.; Tryk, D.; Hashimoto, K.; Fujishima, A., Study of the

photochromic properties of amorphous MoO3 films using Raman microscopy. J.

Phy. Chem. 1995, 99 (44), 16383-16388.

142
5. Eda, K., Raman spectra of hydrogen molybdenum bronze, H 0.30MoO3. J. Solid

State Chem. 1992, 98 (2), 350-357.

6. Gong, L.; Haur, S. C., On demand rapid patterning of colored amorphous

molybdenum oxide using a focused laser beam. J. Mater. Chem. C 2017, 5 (8),

2090-2097.

7. Vasilopoulou, M.; Douvas, A. M.; Georgiadou, D. G.; Palilis, L. C.; Kennou,

S.; Sygellou, L.; Soultati, A.; Kostis, I.; Papadimitropoulos, G.; Davazoglou,

D., The influence of hydrogenation and oxygen vacancies on molybdenum oxides

work function and gap states for application in organic optoelectronics. J. Am.

Chem. Soc. 2012, 134 (39), 16178-16187.

8. Brandt, B. G.; Skapski, A., A refinement of the crystal structure of molybdenum

dioxide. Acta Chem. Scand. 1967, 21 (3), 661-672.

9. Fleisch, T.; Mains, G., An XPS study of the UV reduction and photochromism of

MoO3 and WO3. J. Chem. Phys. 1982, 76 (2), 780-786.

10. Kalantar-Zadeh, K.; Tang, J. S.; Wang, M. S.; Wang, K. L.; Shailos, A.;

Galatsis, K.; Kojima, R.; Strong, V.; Lech, A.; Wlodarski, W.; Kaner, R. B.,

Synthesis of nanometre-thick MoO3 sheets. Nanoscale 2010, 2 (3), 429-433.

11. de Castro, I. A.; Datta, R. S.; Ou, J. Z.; Castellanos-Gomez, A.; Sriram, S.;

Daeneke, T.; Kalantar-Zadeh, K., Molybdenum oxides - from fundamentals to

functionality. Adv. Mater. 2017, 29 (40), 1701619.

12. Zhu, Y. P.; El-Demellawi, J. K.; Yin, J.; Lopatin, S.; Lei, Y. J.; Liu, Z. X.;

Miao, X. H.; Mohammed, O. F.; Alshareef, H. N., Unprecedented surface

plasmon modes in monoclinic MoO2 nanostructures. Adv. Mater. 2020, 32 (19),

1908392.

143
13. Alsaif, M. M. Y. A.; Latham, K.; Field, M. R.; Yao, D. D.; Medhekar, N. V.;

Beane, G. A.; Kaner, R. B.; Russo, S. P.; Ou, J. Z.; Kalantar-zadeh, K., Tunable

plasmon resonances in two-dimensional olybdenum oxide nanoflakes. Adv.

Mater. 2014, 26 (29), 3931-3937.

14. Alsaif, M. M. Y. A.; Field, M. R.; Murdoch, B. J.; Daeneke, T.; Latham, K.;

Chrimes, A. F.; Zoolfakar, A. S.; Russo, S. P.; Ou, J. Z.; Kalantar-Zadeh, K.,

Substoichiometric two-dimensional molybdenum oxide flakes: a plasmonic gas

sensing platform. Nanoscale 2014, 6 (21), 12780-12791.

15. Kresse, G.; Furthmüller, J., Efficient iterative schemes for ab initio total-energy

calculations using a plane-wave basis set. Phys. Rev. B 1996, 54 (16), 11169.

16. Kresse, G.; Furthmüller, J., Efficiency of ab-initio total energy calculations for

metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci.

1996, 6 (1), 15-50.

17. Kresse, G., Ab initio molecular dynamics for liquid metals. Journal of Non-

Crystalline Solids 1995, 192, 222-229.

18. Dudarev, S.; Botton, G.; Savrasov, S.; Humphreys, C.; Sutton, A., Electron-

energy-loss spectra and the structural stability of nickel oxide: An LSDA+ U

study. Phys. Rev. B 1998, 57 (3), 1505.

19. Lei, Y.-H.; Chen, Z.-X., DFT+ U study of properties of MoO3 and hydrogen

adsorption on MoO3 (010). J. Phy. Chem. C 2012, 116 (49), 25757-25764.

20. Peelaers, H.; Chabinyc, M.; Van de Walle, C., Controlling n-type doping in

MoO3. Chem Mater 2017, 29 (6), 2563-2567.

144
Chapter 5 Band structure modulation of gallium-based liquid metal

particles via coating with molybdenum oxide and oxysulfide

145
5.1 Introduction

Gallium-based liquid metals hold intriguing properties and can be utilised in various

applications. Upon micronizing and breaking down the liquid metal bulk into small

droplets, these droplets can be used for making flexible compounds with ideal electrical

properties that can be incorporated in catalysis, sensing, and biomedical applications.

Decoration of liquid metal particles’ surfaces is an effective way to modify their

properties and expand their applications into desired directions. Utilising the super active

surface of liquid metals, deposition of the secondary materials on their surfaces, induced

by interfacial reactions, is a viable way for surface decoration. By exposing the liquid

metal particles to appropriate precursor solutions, chemical reactions could be induced

which can lead to depositions of secondary materials on the surface of the EGaIn droplets

and modify their physical and chemical properties.

In this chapter, the author demonstrates the manipulation of the band structure of liquid

metal droplets via a liquid-liquid interfacial reaction. The surface decoration of

molybdenum oxide (MoOx) and molybdenum oxysulfide (MoOxSy) were chosen as

examples to show the possibility of surface decoration of liquid metal particles with thin

films of two different semiconductors. Molybdenum-based semiconductors were studied

as, after the synthesis, they remain on the surface of gallium-based alloy and do not

dissolve into the core of the droplets. The band structures of the liquid metal particles

before and after the decoration were investigated.

The content of this chapter was included in a paper titled ‘Band structure modulation of

gallium-based liquid metal particles via coating with molybdenum oxide and oxysulfide’,

which was submitted to Nanoscale for publication consideration.

146
5.2 Methods

5.2.1 Materials

All the chemicals were purchased from commercial sources and used as received without

further purification. Milli-Q water (18.2 MΩ·cm at 25 °C) was used in all experiments.

Eutectic gallium indium alloy (EGaIn) was prepared in-house by melting gallium (75.5

wt%) and indium (24.5 wt%) metal mixture at 100 °C on a hot plate and while stirring

with a glass rod until thoroughly mixed. The liquid alloy was then naturally cooled down

at room temperature, transferred from/into the glass vial, and stored for further use.

5.2.2 Synthesis of EGaIn particles

EGaInparticles were prepared as follows. 20 uL EGaIn was sonicated in 20 mL water for

15 mins with a probe sonicator (VCX 750, Sonics & Materials, Inc.) with a 5 mm diameter

probe to create EGaIn particles. After sonication, the suspension of EGaIn particles was

centrifuged (CR4000 centrifuge, Thermoline Scientic) at 10000 rpm for 12 min, and the

clear supernatant was removed. The remaining suspension of EGaIn particles was ready

for further synthesis steps.

5.2.3 Synthesis of molybdenum sulfide decorated EGaIn particles

(EGaIn@MoOxSy)

The aforementioned suspension of EGaIn particles was mixed with 18 mL saturated

aqueous solution of ammonium tetrathiomolybdate [(NH4)2MoS4] by bath sonication. An

aqueous solution of hydrochloric acid (HCl, 5M, 200 uL) was added to the solution to

accelerate the reaction. The mixture was magnetically stirred for 12 hours while the
147
molybdenum sulfide layer grows on the surface of EGaIn particles to form a core-shell

structure. The (NH4)2MoS4 solution slowly decomposes over time and such a process

slightly accelerates under light radiation.2 Therefore, freshly prepared precursor solutions

were applied, and the reaction was kept in dark. After the reaction, the resultant

EGaIn@MoOxSy particles were centrifuged and washed with water 3 times to remove

any residual solution of [(NH4)2MoS4], and the suspension of EGaIn@MoOxSy particles

was dried in a vacuum furnace at 50 degrees centigrade.

5.2.4 Synthesis of molybdenum oxide decorated EGaIn particles (EGaIn@MoO x)

Molybdenum oxide decorated EGaIn (EGaIn@MoOx) particles were prepared as follows.

The suspension of EGaIn particles was mixed with 18 mL aqueous solution of

Ammonium molybdate tetrahydrate [(NH4)6Mo7O24] (1.3 g in 40 mL) and HCl solution

(5M, 2.0 mL) by bath sonication. The mixture was magnetically stirred for 48 hours due

to its slower reaction rate while the molybdenum oxide layer grows on the surface of

EGaIn particles to form a core-shell structure. After the reaction, the resultant

EGaIn@MoOx particles were centrifuged and washed with water 3 times to remove any

residual solution of (NH4)6Mo7O24, and the suspension of EGaIn@MoOx particles was

dried in a vacuum furnace at 50 degrees centigrade.

5.2.5 Morphological and Structural Characterizations.

Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS)

system (JEOL JSM-IT500) were utilised to investigate the morphological features and

elemental distribution of the particles. The specimens were prepared by dispersing the

148
particles in water and drop-casting the suspension on conductive Si wafers. The electron

acceleration voltage was 15 V, and the images were captured at a speed of 20 seconds per

frame.

5.2.6 Optical characterization

X-ray photoelectron spectroscopy (XPS) analysis was carried out using a Thermo

Scientific ESCALAB250i high-resolution XPS (U.K.) with a monochromatic Al Kα soft

X-ray source (1486.68 eV). For spectral acquisition, 120 W power (13.8 kV × 8.7 mA),

2 × 10−9 mbar background vacuum, 90° photoelectron takeoff angle, and 500 μm spot

size were applied. C 1s = 284.8 eV for adventitious hydrocarbon was used as a reference

for calibration. Avantage XPS software (from Avantage Software Ltd., U.K.) was

employed for fitting and deconvoluting the high-resolution XPS spectra.

A Renishaw inVia Raman microspectrometer (Gloucestershire, U.K.) equipped with 532

nm laser was utilised for obtaining Raman spectra. Spectra were collected with standard

confocality and recorded at a static grating scan type with 1800 lines mm −1 grating

(spectral resolution of ∼1.5 cm−1 per point) One second exposure time, and 50 times

magnification 1.0% laser power (0.36 mW on the sample), and 600 accumulations were

applied for all Raman measurements.

The optical absorption spectra were obtained using an ultraviolet−visible spectrometer

(UV−vis, Agilent Cary5000). The specimens were prepared by drop-casting the particles

suspension on transparent quartz substrates. A blank quartz substrate was used for the

baseline correction.

149
5.3 Results and discussion

5.3.1 Decoration of EGaIn particles

Stoichiometric MoO3 and MoS2 and both are n-type semiconductors with bandgap of 3.2

eV1 and 1.1 eV, 2 respectively. The differences in bandgaps make them good choices for

observing the bandgap variations after the decorations. Aqueous solutions of ammonium

molybdate tetrahydrate [(NH4)6Mo7O24] and ammonium tetrathiomolybdate

[(NH4)2MoS4] were applied as precursors for MoOx and MoOxSy decoration layers,

respectively. Surface decoration of EGaIn particles was achieved via interfacial reaction

between liquid metal and aqueous precursors. The reactive and smooth surface of EGaIn

particles serves as a reaction media and deposition template, and the build-in potential at

the liquid metal-aqueous solution interface acts as a driving force that leads to the

deposition of a layer of secondary materials (in this case MoOxSy and MoOx) on the

surface of EGaIn droplet resulting core-shell structured liquid metal droplets

(EGaIn@MoOxSy, and EGaIn@MoOx).

5.3.2 Morphology and chemical composition of EGaIn@MoO xSy, and

EGaIn@MoOx

As shown in the SEM images in Figure 5.1A and B, the as-prepared EGaIn particles take

on a spherical shape with a smooth surface. EDS measurements and elemental mappings

(Figure 5.1C) show a uniform distribution of gallium and indium on the particles’ surface.

As shown in the SEM image in Figure 5.1D and E, the EGaIn@MoO xSy particles inherit

the spherical shape of EGaIn particles with a shell-like structure covering the liquid metal

particles forming a core-shell structure. As presented in the EDS elemental mapping of a

single particle in Figure 5.1F, uniform distribution of Mo and S were observed on the
150
shell of the particle, while gallium and indium were observed throughout the particle. In

particular, the particle shows a breakage of the shell and some leakage of the core

materials at its top-right corner. The leaked core materials show gallium and indium

elements but little Mo and S distribution, which verified that the particles consist of EGaIn

cores and MoOxSy -containing shells. Spherical structures with a somehow rougher

surface, in comparsion to EGaIn@MoOxSy, were observed for the EGaIn@MoOx (Figure

5.1G and H), however, as shown in Figure 5.1I, EGaIn@MoOx particles rendered uniform

distribution of Mo and O on the surface. According to the particle sizes distribution

histogram (Figure 5.2), the average diameter of the EGaIn particles is ~ 470 nm, which

increased to ~ 700 nm and ~ 650 nm after MoOxSy and MoOx decoration, respectively.

These results indicates that the thickness of the deposited layers is ~100 nm in both cases.

XPS surface analysis of EGaIn@MoOxSy showed the presence of gallium, molybdenum,

oxygen, and sulfur (Figure 5.3A). The peak centered at 20.9 eV (red line) is assigned to

3d orbitals of Ga3+, indicating the oxidization of metallic gallium during the deposition

process that led to Ga2O3 or GaOOH in the shell. The peak with a binding energy of 18.7

eV (blue solid line) corresponds to gallium in the metallic state (Ga0) indicating a small

amount of the exposed liquid metal cores due to broken shells. The peak at 24.1 eV

corresponds to the O 2s signals.3 The peak centered at 533.1 eV (O 1s region) corresponds

to the oxygen atoms in the surface absorbed ambient H2O molecules, which is consistent

with previous reports.4 The major peak centered at 531.7 eV agrees with the previously

reported data of metal hydroxide (Me-OH), attributed to GaOOH.5 The small peak located

at 530.9 eV agrees with the value of metal oxide, which in this case could be assigned to

a small amount of the Mo-O bonds.

151
Figure 5.1 SEM and EDS analysis that show the droplets’ morphologies and the special

distribution of Ga, In, Mo, O and Si elements. (A) SEM image of EGaIn particles (scale

bar 1 mm) (B, C) EDS mapping of EGaIn particles (scale bar 100 nm) (D) SEM image

of EGaIn@MoOxSy particles (scale bar 1 mm) (E, F) EDS mapping of EGaIn@MoO xSy

particles (scale bar 500 nm) (G) SEM image of EGaIn particles (scale bar 1 mm) (H, I)

EDS mapping of EGaIn@MoOx particles (scale bar 200 nm)

152
Figure 5.2 Histogram of size distributions of the EGaIn, EGaIn@MoOxSy, and

EGaIn@MoOx particles.

In addition to Ga and O, the XPS surface analysis for EGaIn@MoOxSy also showed the

presence of Mo and S. The peak deconvolution of Mo 3d region revealed two doublets

with a peak area ratio of roughly 3:2, and one single peak. The doublet peaks at 228.6 eV

and 231.9 eV (red lines) correspond to the 3d5/2 and 3d3/2 orbitals of Mo4+, respectively.

The doublet peaks at 229.6 eV and 232.8 eV (green solid lines) correspond to the 3d 5/2
153
and 3d3/2 orbitals of Mo5+, respectively. The single peak at 226.5 eV (blue solid line) is

attributed to the 2s orbitals of sulfur. These values are in agreement with the previous

reports of electrodeposited molybdenum sulfide sheets with the same ammonium

tetrathiomolybdate precursor.6 Two doublets with a peak area ratio of ~ 2:1 are related to

S 2p. The doublet at 163.2 eV and 164.5 eV (pink lines) is assigned to bridging S 22− and

apical S2−, while the doublet located at 162.3 eV and 163.5 eV (dark-yellow lines)

corresponds to unsaturated S2− and terminal S22−. The single peak centered at 161 eV

(purple line) is assigned to 3s orbitals of gallium.7 In addition, a stoichiometric ratio of

2.07:1 was found for S to Mo elements through the quantitative XPS elemental analysis,

indicating that there are substantial S-vacancies. Based on the XPS results, the observed

molybdenum sulfide component in the shell can be described as MoO xSy.

The XPS analysis and peak deconvolutions of EGaIn@MoOx at Ga 3d, O 1s, Mo 3d, S2p

regions, are presented in Figure 5.3B. Similar to EGaIn@MoOxSy, Ga3+ 3d and O 1s

peaks of metal hydroxide were observed at 20.9 eV and 531.7 eV, respectively, indicating

the formation of GaOOH in the shell of EGaIn@MoO x through a similar oxidation

process of gallium. The peak centered at 530.9 eV agrees with that of metal oxide and

belongs to molybdenum oxide. The XPS spectrum at the Mo 3d region also rendered two

3d doublets. The doublet at 235.1 eV and 231.9 eV (red lines) corresponds to the 3d 3/2

and 3d5/2 orbitals of Mo (VI), respectively,8 whilst the doublet at 233.3 eV and 230.0 eV

(green lines) stands for the 3d3/2 and 3d5/2 orbitals of Mo(V). These results confirmed that

the generated shell contains Mo6+ and Mo5+, and the obtained molybdenum oxides can be

described as MoOx. The single peak centered at 161 eV (purple line) in S 2p region is

assigned to 3s orbitals of gallium and no signal was detected for sulfur.

154
Figure 5.3 XPS and peak deconvolution of Ga 3d, O1s, Mo 3d, and S 2p regions for (A)

EGaIn@MoOxSy and (B) EGaIn@MoOx.

Raman spectroscopy was applied to further verify the chemical composition of the

resulting materials. Figure 5.4 presents the Raman spectra of EGaIn, EGaIn@MoOx and

EGaIn@MoOxSy particles. The Raman spectrum of EGaIn particles shows a peak at

around 710 cm-1, which is associated with gallium oxide.9 In the Raman spectrum of

EGaIn@MoOx particles, the signals between 750 and 1000 cm-1 belong to the Mo-O

bonds,10 which are attributed to the molybdenum oxide in EGaIn@MoOx particles. Peaks

in this region are broad and cannot be deconvoluted, which indicates the molybdenum

oxide to be highly amorphous. A distinctive GaOOH peak at around 710 cm -1 was also

observed in this spectrum due to the long deposition time of MoO x in the aqueous

solution. In the Raman spectrum of EGaIn@MoOxSy, the peaks between 150 cm-1 to 240

cm-1 are assigned to the vibration mode of Mo-Mo bonds (Mo-Mo).11,12 The peak group

between 280 cm-1 to 350 cm-1 belongs to the vibration mode of molybdenum sulfide

bonds (Mo-S). The broad peak at around 450 cm-1 is associated with the apical sulfur

atoms [ (Mo3- 3S)]. Peaks between 510-550 cm-1 belong to disulfide ligands.13 These
155
peaks indicate the presence of amorphous molybdenum sulfide in EGaIn@MoOxSy

particles. The small amount of molybdenum oxide compound can be understood by the

peaks between 750-1000 cm-1 that belong to Mo-O bonds, as verified earlier by the XPS

analysis. The absence of any distinctive sharp peak in this spectrum indicated the

deposited MoOxSy to be highly amorphous. In the case of EGaIn@MoOxSy particles, due

to the relatively short reaction time in the aqueous media, no distinctive GaOOH signal

was observed.

Figure 5.4 Raman spectra of EGaIn, EGaIn@MoOx, and EGaIn@MoOxSy particles

156
5.3.3 Band structures of EGaIn@MoOxSy, and EGaIn@MoOx

In order to establish the electronic band structure of the liquid metal particles, ultraviolet

photoelectron spectroscopy (UPS), XPS valence band measurements, and UV–vis

absorption spectroscopy were conducted (Figure 5.5, 5.6, and 5.7). The work function (F)

for the particles’ surface can be determined through the equation F = h - (EF - Ecutoff) 14,

where h is the energy of the photon, EF is the Fermi energy level which was calibrated

to zero by a clean gold surface, and the Ecutoff is the secondary electron cutoff energy at

the high binding energy region of the UPS spectra., The close work functions of 5.1, 5.1,

and 5.2 eV were estimated for EGaIn, EGaIn@MoOxSy, and EGaIn@MoOx particles,

respectively, As shown in Figure 5.6.

Locations of valance band maximum (EVBM) were determined by the linear extrapolation

of the measured valance band onsets subtracted to the Fermi level. As shown in Figure

5.5A, the distance between EVBM and the Fermi level of EGaIn particles is 3.2 eV prior

to surface decoration, which remains relatively unchanged upon MoO x decoration and

dropped to 2.4 eV after the deposition of MoOxSy, respectively. Accordingly, the

locations of EVBM respective to the vacuum for EGaIn, EGaIn@MoOx, and

EGaIn@MoOxSy particles were determined -8.3, -8.4, and -7.5 eV, respectively, as

presented in the band diagram in Figure5.5C.

UV-vis absorption spectroscopy was applied to investigate the optical properties of the

particles (Figure 5.5B). The optical band gap was calculated via Tauc analysis.15 Based

on the (ahv) = A (hv- Eg)1/2 equation, the square of the product between the absorption

coefficient, a, and the photon energy, hv, was plotted against the photon energy hv, and

the band gap (Eg) can be derived from the extrapolation of a linear fitting of the absorption

edge. As shown in Figure 5.7, the band gap of EGaIn particles was calculated to be around

157
4.0 eV, which agrees with the previous report.3 Upon the surface deposition of MoOx, an

enhanced absorption was observed from the 200 nm to 650 nm region in Figure 5.5B, and

the optical bandgap decreased to 3.4 eV. When the surface was decorated with MoO xSy,

the absorption of UV-Vis light was further enhanced throughout the spectrum, and the

optical bandgap dropped to 2.7 eV. Since the bandgap of semiconductors is defined as

the distance between the EVBM and ECBM (conductive band minimum), the ECBM locations

of EGaIn, EGaIn@MoOx, and EGaIn@MoOxSy particles are thereby 4.3, 5.0, and 4.8 eV

below the vacuum, respectively, as shown in Figure 5.5C. As demonstrated, the optical

properties and band structure of the EGaIn particles were modulated by surface

decoration of MoOxSy and MoOx via room-temperature liquid-liquid interfacial reaction.

Figure 5.5 (A) XPS valance band, (B) UV-vis absorption spectroscopy, and (C) band

diagram of EGaIn, EGaIn@MoOx, and EGaIn@ MoOxSy particles deduced from the

UPS, XPS valence band, and UV–vis absorption spectroscopy measurements.

158
Figure 5.6 UPS of EGaIn, EGaIn@MoOx, and EGaIn@MoOxSy particles

Figure 5.7 Tauc Plot of EGaIn, EGaIn@MoOx, and EGaIn@MoOxSy particles for

bandgap calculation.

5.4 Conclusions

In conclusion, EGaIn droplets were synthesised via sonication of EGaIn bulk in DI water,

and aqueous solution of [(NH4)2MoS4] and [(NH4)6Mo7O24] were used as precursors for

deposition of MoOxSy and MoOx on EGaIn droplets, respectively. Raman spectroscopy

and XPS analysis of the droplet confirmed the chemical composition of the deposition

layer, and SEM images verified its core-shell structure. Upon surface decoration of

MoOxSy and MoOx, the band structures of the liquid metal-based droplets were modulated.
159
This work demonstrates that the interfacial reaction between liquid metal and an

appropriate precursor solution was a feasible way to construct core-shell structures, and

the band structure of liquid metal droplets could be modulated by altering the composition

of the deposited shell materials.

5.5 References

1. De Castro, I. A.; Datta, R. S.; Ou, J. Z.; Castellanos‐Gomez, A.; Sriram, S.;

Daeneke, T.; Kalantar‐zadeh, K., Molybdenum oxides - from fundamentals to

functionality. Adv. Mater. 2017, 29 (40), 1701619.

2. Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.; Strano, M. S.,

Electronics and optoelectronics of two-dimensional transition metal

dichalcogenides. Nat. Nanotechnol. 2012, 7 (11), 699-712.

3. Han, J. L.; Yang, J.; Tang, J. B.; Ghasemian, M. B.; Hubble, L. J.; Syed, N.;

Daeneke, T.; Kalantar-Zadeh, K., Liquid metals for tuning gas sensitive layers. J.

Mater. Chem. C 2019, 7 (21), 6375-6382.

4. Junta-Rosso, J. L.; Hochella Jr, M. F., The chemistry of hematite 001 surfaces.

Geochim. Cosmochim. Acta 1996, 60 (2), 305-314.

5. Shakir, I.; Shahid, M.; Kang, D. J., MoO3 and Cu0.33MoO3 nanorods for

unprecedented UV/Visible light photocatalysis. Chem. Comm. 2010, 46 (24),

4324-4326.

6. Zhang, L.; Wu, L.; Li, J.; Lei, J., Electrodeposition of amorphous molybdenum

sulfide thin film for electrochemical hydrogen evolution reaction. BMC Chem.

2019, 13, 88.

160
7. Liu, Z.; Zhang, X.; Wang, B.; Xia, M.; Gao, S.; Liu, X.; Zavabeti, A.; Ou, J.

Z.; Kalantar-Zadeh, K.; Wang, Y., Amorphous MoSx-coated TiO2 nanotube

Arrays for enhanced electrocatalytic hydrogen evolution reaction. J. Phys. Chem.

C 2018, 122 (24), 12589-12597.

8. Xiang, D.; Han, C.; Zhang, J.; Chen, W., Gap states assisted MoO3 nanobelt

photodetector with wide spectrum response. Sci. Rep. 2014, 4, 4891.

9. Han, J.; Tang, J.; Idrus-Saidi, S. A.; Christoe, M. J.; O’Mullane, A. P.; Kalantar-

Zadeh, K., Exploring electrochemical extrusion of wires from liquid metals. ACS

Appl. Mater. Interfaces 2020, 12 (27), 31010-31020.

10. Gong, L.; Haur, S. C., On demand rapid patterning of colored amorphous

molybdenum oxide using a focused laser beam. J. Mater. Chem. C 2017, 5 (8),

2090-2097.

11. Escalera‐López, D.; Lou, Z.; Rees, N. V., Benchmarking the activity, stability,

and inherent electrochemistry of amorphous molybdenum sulfide for hydrogen

production. Adv. Energy Mater. 2019, 9 (8), 1802614.

12. Fominski, V.; Demin, M.; Nevolin, V.; Fominski, D.; Romanov, R.;

Gritskevich, M.; Smirnov, N., Reactive pulsed laser deposition of clustered-type

MoSx (x~ 2, 3, and 4) films and their solid lubricant properties at low temperature.

Nanomaterials 2020, 10 (4), 653.

13. Daeneke, T.; Dahr, N.; Atkin, P.; Clark, R. M.; Harrison, C. J.; Brkljaca, R.;

Pillai, N.; Zhang, B. Y.; Zavabeti, A.; Ippolito, S. J., Surface water dependent

properties of sulfur-rich molybdenum sulfides: electrolyteless gas phase water

splitting. ACS Nano 2017, 11 (7), 6782-6794.

161
14. Pansri, S.; Supruangnet, R.; Nakajima, H.; Rattanasuporn, S.; Noothongkaew,

S., Band offset determination of p-NiO/n-TiO2 heterojunctions for applications in

high-performance UV photodetectors. J. Mater. Sci. 2020, 55 (10), 4332-4344.

15. Tauc, J.; Grigorovici, R.; Vancu, A., Optical properties and electronic structure

of amorphous germanium. Phys. Status Solidi B 1966, 15 (2), 627-637.

162
Chapter 6 Conclusions

163
6.1 Summary of the work

This thesis focuses on expanding the possibility of two-dimensional (2D) materials

synthesis via gallium-based liquid metal processes. The surface of liquid metal, due to its

intrinsic ultra-smooth and reactive nature, makes a perfect template for 2D materials

deposition. When a liquid metal droplet is in an aqueous environment, the built-in

potential at the liquid metal-aqueous solution interface can serve as a driving force of the

redox reactions, leading to the destabilization of precursors and deposition of 2D

materials at the interface. In this way, with appropriate selections of aqueous solutions as

precursors, room-temperature deposition of desired 2D materials in a large area is

possible on the surface of liquid metal. In particular, this approach is applicable for

compounds of metals that are not dissolvable in gallium-based liquid metals at low

temperatures, which cannot be easily harnessed from the oxidation layer of liquid metals.

This synthetic route is highly feasible, since the chemical composition, morphology and

thickness of the deposited layers could be modulated with the chemical environment at

the interface.

In this thesis, molybdenum sulfide and molybdenum oxide were chosen as examples to

demonstrate the capability of this synthetic approach.

In chapter 2, the author presented a literature review about 2D materials synthesis via

liquid metal-based processes. First, the author discussed the intriguing properties and

features of post-transition liquid metals including their surface layering phenomenon,

which endows them with an intrinsically ultra-flat surface and lays the very foundation

of this research. Then, the author gave a comprehensive review of recent progress

regarding the synthesis and exfoliation of 2D materials on the surface of liquid metals. At

last, applications of these 2D materials in various devices were summarized.

164
In chapter 3, the author demonstrated the room-temperature synthesis of molybdenum

sulfide sheets through a self-deposition process at the liquid metal-aqueous precursor

interface. (NH4)2MoS4 solution was applied as an aqueous precursor and the surface of

EGaIn served as a reaction medium and template. It was verified by electrochemical

assessments that the built-in potential at the interface is sufficient to destabilize the

precursor and lead to the interfacial deposition of molybdenum sulfide sheets. The

deposited 2D sheets were shown to be extracted and transferred onto desired substrates

using a program-controlled piezoelectric stage. After annealing, the resultant 2D

materials were proved to be monolayer 2H-MoS2 with high crystallinity.

In chapter 4, the author expended the aforementioned synthetic route and used it for the

room-temperature deposition of molybdenum oxide. The as-synthesised materials were

insulting and hydrated molybdenum oxide sheets, which, under laser exposure, went

through a dehydration process and transformed into grains of crystalline MoO 2, a semi-

metal with good conductivity. As such, the as-deposited molybdenum oxide sheets could

serve as a writing medium that enables instant laser writing of conductive MoO 2 patterns.

In addition, high-resolution conductive atomic force microscopy (AFM) measurements

showed high conductivity at the boundaries of the grains, which according to the density

functional theory (DFT) calculations, could result from the oxygen-vacancy-introduced

defect states, and the band-tails along the grain boundaries from disordered structures.

In chapter 5, the author demonstrated band structure modulation of liquid metal particles

via surface decoration. As semiconductors with different band gaps, molybdenum oxide

and molybdenum sulfide were chosen as deposition materials to demonstrate the band

modulation. The liquid metal-aqueous solution interfacial reaction was applied for the

surface decoration of EGaIn particles, which led to core-shell structured liquid metal

165
particles. Upon decoration, the EGaIn particles showed significant changes in their

optical properties and electronic band structures.

6.2 Publications

The work presented in this thesis results in 3 first-authored and one co-first-authored peer-

reviewed journal papers. Additionally, the author of this thesis contributed and co-

authored another 8 journal papers during the course of her Ph.D. program.

6.2.1 Publications contributing to this thesis

[1] Y. Wang, M. Mayyas, J. Yang, J. B. Tang, M. B. Ghasemian, J. L. Han, A, Elbourne,

T, Daeneke, R, B. Kaner, and K, Kalantar-Zadeh, Self-Deposition of 2D Molybdenum

Sulfides on Liquid Metals. Advanced Functional Materials, 31, 2005866 (2020)

[2] Y. Wang, M. Mayyas, J. Yang, M. B. Ghasemian, J. B. Tang, M. Mousavi, J. Han, M.

Ahmed, M. Baharfar, G. Mao, Y. Yao, D. Esrafilzadeh, D. Cortie, and K. Kalantar-Zadeh,

Liquid-Metal-Assisted Deposition and Patterning of Molybdenum Dioxide at Low

Temperature. ACS Applied Materials & Interfaces 13, 53181-53193 (2021)

[3] Y. Wang, M Baharfar, J. Yang, M. Mayyas, M. B. Ghasemian, K. Kalantar-Zadeh,

Liquid state of post-transition metals for interfacial synthesis of two-dimensional

materials. Applied Physics Reviews 9, 021306 (2022)

[4] M. B. Ghasemian,† Y. Wang,† F.-M. Allioux, A. Zavabeti and K Kalantar-Zadeh,

Band structure modulation of gallium-based liquid metal particles via coating with

molybdenum oxide and oxysulfide (submitted)

166
6.2.3 Other co-authored publications with the theme of liquid metals

[1] H. Li, R. Abbasi, Y. Wang, F.-M. Allioux, P. Koshy, S. A Idrus-Saidi, M. A. Rahim, J.

Yang, M. Mousavi, J. Tang, M. B Ghasemian, R. Jalili, K. Kalantar-Zadeh, M. Mayyas,

Liquid metal-supported synthesis of cupric oxide, Journal of Material Chemistry C 8,

1656-1665 (2020)

[2] M. Mayyas, H. Li, P. Kumar, M. B. Ghasemian, J. Yang, Y. Wang, D. J. Lawes, J.

Han, M. G. Saborio, J. Tang, R. Jalili, S. H. Lee, W. K. Seong, S. P. Russo, D. Esrafilzadeh,

T. Daeneke, R. B. Kaner, R. S. Ruoff, K. Kalantar‐Zadeh, Liquid-Metal-Templated

Synthesis of 2D Graphitic Materials at Room Temperature, Advanced Materials 32,

2001997 (2020)

[3] M. Mayyas, M. Mousavi, M. B. Ghasemian, R. Abbasi, H. Li, M. J. Christoe, J. Han,

Y. Wang, C. Zhang, M. A. Rahim, J. Tang, J. Yang, D. Esrafilzadeh, R. Jalili, F.-M.

Allioux, A. P. O’Mullane, K. Kalantar-Zadeh, Pulsing Liquid Alloys for Nanomaterials

Synthesis, ACS Nano 14, 14070-14079 (2020)

[4] M. B Ghasemian, A. Zavabeti, R. Abbasi, P. V. Kumar, N. Syed, Y. Yao, J. Tang, Y.

Wang, A. Elbourne, J. Han, M. Mousavi, T. Daeneke, K Kalantar-Zadeh, Ultra-thin lead

oxide piezoelectric layers for reduced environmental contamination using a liquid metal-

based process, Journal of Material Chemistry A 8, 19434-19443 (2020)

[5] W. Xie, F.-M. Allioux, R, Namivandi-Zangeneh, M. B. Ghasemian, J. Han, M. A.

Rahim, J. Tang, J. Yang, M. Mousavi, M. Mayyas, Z. Cao, F. Centurion, M. J. Christoe,

C. Zhang, Y. Wang, S. Merhebi, M. Baharfar, G. Ng, D. Esrafilzadeh, C. Boyer, K.

Kalantar-Zade, Polydopamine Shell as a Ga3+ Reservoir for Triggering Gallium–Indium

Phase Separation in Eutectic Gallium–Indium Nanoalloys. ACS Nano 15, 16839–16850

(2021).
167
[6] M. Mousavi, M. B. Ghasemian, J. Han, Y. Wang, R. Abbasi, Y. Jiong, J. Tang, S. A.

Idrus-Saidi, X, Guan, M. J. Christoe, S. Merhebi, C. Zhang, J. Tang, R. Jalili, T. Daeneke,

T. Wu, K. Kalantar-Zadeh, Bismuth telluride topological insulator synthesized using

liquid metal alloys: Test of NO2 selective sensing. Applied Materials Today, 22, 100954

(2021).

[7] M. Mayyas, K. Khoshmanesh, P. Kumar, M. Mousavi, J. Tang, M. B. Ghasemian, J.

Yang, Y. Wang, M. Baharfar, M. A. Rahim, W. Xie, F.-M. Allioux, R. Daiyan, R. Jalili,

D. Esrafilzadeh, K. Kalantar-Zadeh, Gallium-Based Liquid Metal Reaction Media for

Interfacial Precipitation of Bismuth Nanomaterials with Controlled Phases and

Morphologies Advanced Functional Materials, 32, 2108673 (2021)

[8] J. Han, M. Mayyas, J. Tang, M. Mousavi, S. A.Idrus-Saidi, S. Cai, Z. Cao, Y. Wang,

J. Tang, R. Jalili, A. P.O'Mullane, R. B. Kaner, K. Khoshmanesh, K. Kalantar-Zadeh.

Liquid metal enabled continuous flow reactor: a proof-of-concept, Matter, 5, 379-381

(2021)

6.3 Future research directions

Spontaneous liquid-liquid interfacial reaction offers a promising route for 2D materials

synthesis via liquid metals. Despite numerous research in this field, this concept is still

facing many challenges while holding great potential. In many cases, the deposited 2D

materials are mixtures of different products and/or hydrated products, that require post-

treatments to obtain stable and crystalline 2D materials. Although electrochemical

assessment offers some insight into these interfacial reactions, mechanisms behind the

interfacial reaction and deposition of products still call for further investigations, and

168
precise control of the chemical compositions and morphologies of the resulting materials

is yet to be fully achieved.

The touch transfer process in aqueous or other solutions is still a challenge. Consecutive

reaction and transfer methods like squeeze printing and roll printing for centimetre-scale

deposition of 2D materials in aqueous environments are yet to be developed.

In addition, most of the existing research uses aqueous solutions, which limits the

operating temperature window to below 100 ℃. Organic solutions or ionic solutions with

higher boiling points can be used for expanding the applicable temperature window to

higher values. This also allows more choices in terms of liquid metals that can expand the

reaction system to more metals including In, Bi, and Sn as the base for the process.

169