Tommi Kaplas - UEF...Tommi Kaplas Synthesis and Optical Characterization of Ultrathin Carbon Films This work is devoted on experimental synthesis and optical characterization of ultrathin

Publications of the University of Eastern FinlandDissertations in Forestry and Natural Sciences

Publications of the University of Eastern Finland

Dissertations in Forestry and Natural Sciences

isbn 978-952-61-1303-6 (printed)

issnl 1798-5668

issn 1798-5668

isbn 978-952-61-1304-3 (pdf)

issn 1798-5676

Tommi Kaplas

Synthesis and Optical Characterization of Ultrathin Carbon Films

This work is devoted on experimental

synthesis and optical characterization

of ultrathin carbon films. Specifically

the main focus of this thesis is in two

carbon species with dominating sp2

hybridization. One of these materials is

pyrolytic carbon, which is well known,

amorphous carbon material but is only

very recently been synthetized as an

ultrathin film. Second investigated ma-

terial is crystalline, a single and a few

layered graphene, which has recently

drawn a lot of interest of scientific com-

munity. The synthesis is carried out by

using chemical vapor deposition (CVD)

techniques while the optical character-

ization is done by utilizing spectropho-

tometry and Raman spectroscopy.

disser

tation

s | 130 | To

mm

i Ka

plas | S

ynth

esis an

d O

ptica

l Ch

ara

cterizatio

n of U

ltrath

in C

arb

on

Film

s

Tommi KaplasSynthesis and Optical

Characterization of Ultrathin Carbon Films

TOMMI KAPLAS

Synthesis and OpticalCharacterization of

Ultrathin Carbon Films


No 130

Academic DissertationTo be presented by permission of the Faculty of Science and Forestry for public

examination in the Auditorium M100 in Metria Building at the University ofEastern Finland, Joensuu, on November, 29, 2013,

at 13 o’clock.

Department of Physics and Mathematics

Kopijyva Oy

Joensuu, 2013

Editor: Prof. Pertti Pasanen, Prof. Pekka Kilpelainen,

Prof. Kai Peiponen, Prof. Matti Vornanen

Distribution:

University of Eastern Finland Library / Sales of publications

[email protected]

http://www.uef.fi/kirjasto

ISBN: : 978-952-61-1303-6 (printed)

ISSNL: 1798-5668

ISSN: 1798-5668

ISBN: 978-952-61-1304-3 (pdf)

ISSNL: 1798-5668

ISSN: 1798-5676

Author’s address: University of Eastern FinlandDepartment of Physics and MathematicsP.O.Box 10180101 JOENSUUFINLANDemail: [email protected]

Supervisors: Professor Yuri Svirko, Ph.D.University of Eastern FinlandDepartment of Physics and MathematicsP.O.Box 10180101 JOENSUUFINLANDemail: [email protected]

Reviewers: Professor Werner Blau, Ph.D.Trinity College DublinDepartment of PhysicsP.O.Box 22 DublinIRELANDemail: [email protected]

Professor Alexander A. Balandin, Ph.D.University of CaliforniaDepartment of Electrical EngineeringP.O.Box92521-0425 CaliforniaUSAemail: [email protected]

Opponent: Professor Harri Lipsanen, Ph.D.Aalto UniversityDepartment of Micro- and NanosciencesP.O.Box02150 EspooFINLAND

Kopijyva Oy

Joensuu, 2013

Editor: Prof. Pertti Pasanen, Prof. Pekka Kilpelainen,

Prof. Kai Peiponen, Prof. Matti Vornanen

Distribution:

University of Eastern Finland Library / Sales of publications

[email protected]

http://www.uef.fi/kirjasto

ISBN: : 978-952-61-1303-6 (printed)

ISSNL: 1798-5668

ISSN: 1798-5668

ISBN: 978-952-61-1304-3 (pdf)

ISSNL: 1798-5668

ISSN: 1798-5676

Author’s address: University of Eastern FinlandDepartment of Physics and MathematicsP.O.Box 10180101 JOENSUUFINLANDemail: [email protected]

Supervisors: Professor Yuri Svirko, Ph.D.University of Eastern FinlandDepartment of Physics and MathematicsP.O.Box 10180101 JOENSUUFINLANDemail: [email protected]

Reviewers: Professor Werner Blau, Ph.D.Trinity College DublinDepartment of PhysicsP.O.Box 22 DublinIRELANDemail: [email protected]

Professor Alexander A. Balandin, Ph.D.University of CaliforniaDepartment of Electrical EngineeringP.O.Box92521-0425 CaliforniaUSAemail: [email protected]

Opponent: Professor Harri Lipsanen, Ph.D.Aalto UniversityDepartment of Micro- and NanosciencesP.O.Box02150 EspooFINLAND

ABSTRACT

This thesis is devoted to chemical vapor deposition (CVD) basedsynthesis and optical characterization of ultrathin carbon films. Abrief introduction into sp2 hybridization of carbon atoms, grapheneband structure and Raman and absorption spectra of sp2 carbonspecies is followed by the presentation of experimental results. Theperformed experiments aimed at the direct deposition grapheneand ultrathin pyrolytic carbon films on dielectric substrates. In par-ticular pyrolytic carbon films with thickness of 5-50 nm were fabri-cated in the fused silica substrates of arbitrary shaped and charac-terized using methods of the Raman spectroscopy and spectropho-tometry. Direct synthesis of the graphene on the silica substrateswere achieved by using pre-deposited sacrificial copper films. Inparticular it was demonstrated the self-assembling deposition ofa few layered graphene pattern on silica with submicron featuresize and submicron spatial resolution. These obtained results openinteresting opportunities for incorporation graphene elements intoSi/SiO2 photonic and optoelectronic components.

Universal Decimal Classification: 53.084.85, 535.3, 535.4, 681.7.02PACS Classification: 07.60.-j, 42.70.-a, 42.79.-e, 81.16.-cKeywords: graphene; graphite; pyrolytic carbon; Raman spectroscopy;spectrophotometry; copper melting; self-assembling

Preface

At first, I would like to thank my supervisor Professor Yuri Svirkofor his help and support that he has given me. Especially, I wouldlike to thank Yuri for the creative freedom that allowed me to findmy own paths in the jungle of science. Also I thank the Dean of thefaculty of Science and Forestry Professor Timo Jaaskelainen, Headof the Department of Physics and Mathematics Professor MarkkuKuittinen and Head of the Institute of the Photonics Professor PasiVahimaa for granting me this opportunity to work on such of aninteresting field of science.

Furthermore, I am indebted to my co-workers, especially SeppoHonkanen, Alexander Obraztsov, Elena Obraztsova, Polina Kuzhir,Sergey Maksimenko, Lasse Karvonen, Martti Silvennoinen, PetriStenberg and Dmitry Lyashenko for their support and guidance onmy road. Also I am greatly thankful for Hannele Karppinen, OlgaSvirko, Pertti Paakkonen, Tommi Itkonen, Timo Vahimaa and UntoPievilainen for all the help in practical problems.

Thank you the members of our ”NanoCarbon team” for adopt-ing me as one of your own and helping me every time and way youcould. Special thanks to Risto, Mikko, Jussi, Kalle, Kimmo, Timoand Joonas for nice and sometimes a bit extraordinary discussions.Also, all the colleagues working on the Physics department: Thankyou all for such of nice memories.

Finally, I want to express my gratitude to my parents Annikkiand Tuomo and of course my brother Teemu and sister Niina withher family for their love, support and understanding. Moreover, Iwant to thank all my friends, especially Otto, Heidi, Liisa, Niko,Jussi, Jyrki, Juho, Outi and Antti. Thank you.

Joensuu 5 November, 2013 Tommi Kaplas

ABSTRACT

This thesis is devoted to chemical vapor deposition (CVD) basedsynthesis and optical characterization of ultrathin carbon films. Abrief introduction into sp2 hybridization of carbon atoms, grapheneband structure and Raman and absorption spectra of sp2 carbonspecies is followed by the presentation of experimental results. Theperformed experiments aimed at the direct deposition grapheneand ultrathin pyrolytic carbon films on dielectric substrates. In par-ticular pyrolytic carbon films with thickness of 5-50 nm were fabri-cated in the fused silica substrates of arbitrary shaped and charac-terized using methods of the Raman spectroscopy and spectropho-tometry. Direct synthesis of the graphene on the silica substrateswere achieved by using pre-deposited sacrificial copper films. Inparticular it was demonstrated the self-assembling deposition ofa few layered graphene pattern on silica with submicron featuresize and submicron spatial resolution. These obtained results openinteresting opportunities for incorporation graphene elements intoSi/SiO2 photonic and optoelectronic components.

Universal Decimal Classification: 53.084.85, 535.3, 535.4, 681.7.02PACS Classification: 07.60.-j, 42.70.-a, 42.79.-e, 81.16.-cKeywords: graphene; graphite; pyrolytic carbon; Raman spectroscopy;spectrophotometry; copper melting; self-assembling

Preface

At first, I would like to thank my supervisor Professor Yuri Svirkofor his help and support that he has given me. Especially, I wouldlike to thank Yuri for the creative freedom that allowed me to findmy own paths in the jungle of science. Also I thank the Dean of thefaculty of Science and Forestry Professor Timo Jaaskelainen, Headof the Department of Physics and Mathematics Professor MarkkuKuittinen and Head of the Institute of the Photonics Professor PasiVahimaa for granting me this opportunity to work on such of aninteresting field of science.

Furthermore, I am indebted to my co-workers, especially SeppoHonkanen, Alexander Obraztsov, Elena Obraztsova, Polina Kuzhir,Sergey Maksimenko, Lasse Karvonen, Martti Silvennoinen, PetriStenberg and Dmitry Lyashenko for their support and guidance onmy road. Also I am greatly thankful for Hannele Karppinen, OlgaSvirko, Pertti Paakkonen, Tommi Itkonen, Timo Vahimaa and UntoPievilainen for all the help in practical problems.

Thank you the members of our ”NanoCarbon team” for adopt-ing me as one of your own and helping me every time and way youcould. Special thanks to Risto, Mikko, Jussi, Kalle, Kimmo, Timoand Joonas for nice and sometimes a bit extraordinary discussions.Also, all the colleagues working on the Physics department: Thankyou all for such of nice memories.

Finally, I want to express my gratitude to my parents Annikkiand Tuomo and of course my brother Teemu and sister Niina withher family for their love, support and understanding. Moreover, Iwant to thank all my friends, especially Otto, Heidi, Liisa, Niko,Jussi, Jyrki, Juho, Outi and Antti. Thank you.

Joensuu 5 November, 2013 Tommi Kaplas

LIST OF PUBLICATIONS

This thesis consists of the present review of the author’s work inthe field of optical interconnects and the following selection of theauthor’s publications:

I T. Kaplas, A. Zolotukhin and Y. Svirko, ”Thickness deter-mination of graphene on metal substrate by reflection spec-troscopy,” Optics Express 19, 17226–17231 (2011). (Reprintedwith permission from OSA)

II T. Kaplas and Y. Svirko, ”Direct deposition of semitransparentconducting pyrolytic carbon films,” Journal of Nanophotonics6(1), 061703–6 (2012). (Reprinted with permission from SPIE)

III T. Kaplas, D. Sharma, and Y. Svirko, ”Few-layer graphenesynthesis on a dielectric substrate,” Carbon 50(4), 1503–1509(2012). (Reprinted with permission from Elsevier)

IV T. Kaplas, M. Silvennoinen, K. Paivasaari and Y. Svirko, ”Self-assembled two-dimensional graphene grating on a dielectricsubstrate,” Appl. Phys. Lett. 102(21), 211603–3 (2013). (Reprin-ted with permission from AIP Publishing LLC)

V T. Kaplas and Y. Svirko, ”Self-assembled graphene on a di-electric micro- and nanostructures,” Carbon submitted in July(2013). (Reprinted with permission from Elsevier)

Throughout the overview, these papers will be referred to by Ro-man numerals.

In addition, the author has contributed to the works publishedin peer-reviewed journals in [1–11].

AUTHOR’S CONTRIBUTION

The publications selected in this dissertation are original researchpapers on synthesis and optical characterization of ultrathin nanocar-bon films.

The original idea of papers I–V were originally suggested bythe author. For the direct graphene deposition experiments in pa-pers III–V the author prepared the thin copper films. For the self-assembling demonstration in paper V author also fabricated the mi-cro and nanostructures. Carbon nanofilms in all papers were chem-ical vapor deposited by the author. Characterization by the reflec-tion and transmission spectroscopy in papers I and II, the electricalmeasurements in paper II, scanning electron microscopic imagingin papers III–V and the Raman spectroscopy in papers II–V werecarried out by the author.

The illustrations and texts for the manuscripts to papers I–IIand IV–V were mainly prepared by the author. In paper III authorhas contributed about one half to the overall writing process. Thepapers have been completed with significant co-operation with theco-authors.

LIST OF PUBLICATIONS

This thesis consists of the present review of the author’s work inthe field of optical interconnects and the following selection of theauthor’s publications:

I T. Kaplas, A. Zolotukhin and Y. Svirko, ”Thickness deter-mination of graphene on metal substrate by reflection spec-troscopy,” Optics Express 19, 17226–17231 (2011). (Reprintedwith permission from OSA)

II T. Kaplas and Y. Svirko, ”Direct deposition of semitransparentconducting pyrolytic carbon films,” Journal of Nanophotonics6(1), 061703–6 (2012). (Reprinted with permission from SPIE)

III T. Kaplas, D. Sharma, and Y. Svirko, ”Few-layer graphenesynthesis on a dielectric substrate,” Carbon 50(4), 1503–1509(2012). (Reprinted with permission from Elsevier)

IV T. Kaplas, M. Silvennoinen, K. Paivasaari and Y. Svirko, ”Self-assembled two-dimensional graphene grating on a dielectricsubstrate,” Appl. Phys. Lett. 102(21), 211603–3 (2013). (Reprin-ted with permission from AIP Publishing LLC)

V T. Kaplas and Y. Svirko, ”Self-assembled graphene on a di-electric micro- and nanostructures,” Carbon submitted in July(2013). (Reprinted with permission from Elsevier)

Throughout the overview, these papers will be referred to by Ro-man numerals.

In addition, the author has contributed to the works publishedin peer-reviewed journals in [1–11].

AUTHOR’S CONTRIBUTION

The publications selected in this dissertation are original researchpapers on synthesis and optical characterization of ultrathin nanocar-bon films.

The original idea of papers I–V were originally suggested bythe author. For the direct graphene deposition experiments in pa-pers III–V the author prepared the thin copper films. For the self-assembling demonstration in paper V author also fabricated the mi-cro and nanostructures. Carbon nanofilms in all papers were chem-ical vapor deposited by the author. Characterization by the reflec-tion and transmission spectroscopy in papers I and II, the electricalmeasurements in paper II, scanning electron microscopic imagingin papers III–V and the Raman spectroscopy in papers II–V werecarried out by the author.

The illustrations and texts for the manuscripts to papers I–IIand IV–V were mainly prepared by the author. In paper III authorhas contributed about one half to the overall writing process. Thepapers have been completed with significant co-operation with theco-authors.

Contents

1 INTRODUCTION 1

2 THE SP2 HYBRIDIZATION OF CARBON 52.1 Elemental carbon and C=C bonding . . . . . . . . . . 52.2 Graphene band structure . . . . . . . . . . . . . . . . . 8

3 OPTICAL SPECTROSCOPY OF GRAPHENE BASED THINFILMS 133.1 Raman scattering in graphene and pyrolytic carbon . 133.2 Spectrophotometry of nanocarbon films . . . . . . . . 17

4 FABRICATION OF ULTRATHIN NANOCARBON FILMSUSING CHEMICAL VAPOR DEPOSITION 214.1 Chemical vapor deposition of pyrolytic carbon . . . . 214.2 CVD graphene on a copper catalyst . . . . . . . . . . 23

4.2.1 Multilayered graphene on copper . . . . . . . 244.2.2 Graphene transfer from Cu catalyst to a di-

electric . . . . . . . . . . . . . . . . . . . . . . . 254.3 Direct graphene deposition on dielectric substrate . . 264.4 Experimental methods for a chemical vapor deposi-

tion of an ultrathin carbon films . . . . . . . . . . . . 29

5 OPTICAL CHARACTERIZATION OF CARBON FILMS 335.1 Raman spectroscopy of nanocarbon films . . . . . . . 33

5.1.1 Raman spectra of a graphene film transferredto a dielectric substrate . . . . . . . . . . . . . 33

5.1.2 Raman spectra of a graphene synthesized di-rectly on a dielectric substrate . . . . . . . . . 35

5.1.3 Raman spectrum of pyrolytic carbon . . . . . 365.2 Thickness determination of a nanocarbon film on a

copper substrate . . . . . . . . . . . . . . . . . . . . . . 38

Contents

1 INTRODUCTION 1

2 THE SP2 HYBRIDIZATION OF CARBON 52.1 Elemental carbon and C=C bonding . . . . . . . . . . 52.2 Graphene band structure . . . . . . . . . . . . . . . . . 8

3 OPTICAL SPECTROSCOPY OF GRAPHENE BASED THINFILMS 133.1 Raman scattering in graphene and pyrolytic carbon . 133.2 Spectrophotometry of nanocarbon films . . . . . . . . 17

4 FABRICATION OF ULTRATHIN NANOCARBON FILMSUSING CHEMICAL VAPOR DEPOSITION 214.1 Chemical vapor deposition of pyrolytic carbon . . . . 214.2 CVD graphene on a copper catalyst . . . . . . . . . . 23

4.2.1 Multilayered graphene on copper . . . . . . . 244.2.2 Graphene transfer from Cu catalyst to a di-

electric . . . . . . . . . . . . . . . . . . . . . . . 254.3 Direct graphene deposition on dielectric substrate . . 264.4 Experimental methods for a chemical vapor deposi-

tion of an ultrathin carbon films . . . . . . . . . . . . 29

5 OPTICAL CHARACTERIZATION OF CARBON FILMS 335.1 Raman spectroscopy of nanocarbon films . . . . . . . 33

5.1.1 Raman spectra of a graphene film transferredto a dielectric substrate . . . . . . . . . . . . . 33

5.1.2 Raman spectra of a graphene synthesized di-rectly on a dielectric substrate . . . . . . . . . 35

5.1.3 Raman spectrum of pyrolytic carbon . . . . . 365.2 Thickness determination of a nanocarbon film on a

copper substrate . . . . . . . . . . . . . . . . . . . . . . 38

5.3 Optically semitransparent, conducting pyrolytic car-bon film . . . . . . . . . . . . . . . . . . . . . . . . . . 40

6 GRAPHENE DEPOSITION AND SELF-ASSEMBLING ONA DIELECTRIC SUBSTRATE 436.1 Direct CVD of a few layered graphene on a dielectric 436.2 Self-assembled graphene grating with ten micron pe-

riodicity . . . . . . . . . . . . . . . . . . . . . . . . . . 456.3 Self-assembled graphene on a dielectric sub-micron

grating . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

7 CONCLUSIONS 51

REFERENCES 53

1 Introduction

The unique ability of carbon atoms to stabilize complicated net-works, which are the basis for the known forms of life, ensuresunique role of carbon in nature. Even elemental carbon demon-strates complicated behavior, forming a number of very differentstructures. In particular two forms of crystalline carbon, diamondand graphite, are known since ancient times and have drasticallydifferent physical and chemical properties. [12]

Such a difference is due to different bonding mechanisms of car-bon atoms in diamond and graphite. If carbon atoms are connectedwith σ single bonds one obtains diamond crystal that is transparentand electrically insulating material. However, if carbon atoms areconnected by σ−π double bonds then one obtains graphite that ab-sorbs light very efficiently and is electrically conducting. This workfocuses on ultrathin carbon films with dominating σ − π doublebonds, pyrolytic carbon and graphene, which are closely related tographite. [12]

The first reports of pyrolytic carbon (PyC), highly amorphoussp2 hybridized carbon allotrope, produced by chemical vapor de-position are found in 1880s when PyC was found to be suitable ma-terial for strengthen an incandescent lamp filament [12, 13]. It tooknearly 90 years until some interest was paid to the PyC synthesis us-ing methane based chemical vapor deposition (CVD) [14, 15]. Thiswas probably because theoretically the properties of PyC, whichconsists of intertwined graphene flakes with size of few nanome-ters, are difficult to predict due to its low crystallinity. However,recent experimental findings have opened interesting opportunitiesfor applications of ultrathin PyC films which are electrically con-ductive, chemically inert and mechanically strong.

Graphite consists of multiple layers stacked to one another byrelatively weak van-der-Waals forces [16]. Similarly to individualsheets in the paper stack one can remove and isolate only a mono-

Dissertations in Forestry and Natural Sciences No 130 1

5.3 Optically semitransparent, conducting pyrolytic car-bon film . . . . . . . . . . . . . . . . . . . . . . . . . . 40

6 GRAPHENE DEPOSITION AND SELF-ASSEMBLING ONA DIELECTRIC SUBSTRATE 436.1 Direct CVD of a few layered graphene on a dielectric 436.2 Self-assembled graphene grating with ten micron pe-

riodicity . . . . . . . . . . . . . . . . . . . . . . . . . . 456.3 Self-assembled graphene on a dielectric sub-micron

grating . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

7 CONCLUSIONS 51

REFERENCES 53

1 Introduction

The unique ability of carbon atoms to stabilize complicated net-works, which are the basis for the known forms of life, ensuresunique role of carbon in nature. Even elemental carbon demon-strates complicated behavior, forming a number of very differentstructures. In particular two forms of crystalline carbon, diamondand graphite, are known since ancient times and have drasticallydifferent physical and chemical properties. [12]

Such a difference is due to different bonding mechanisms of car-bon atoms in diamond and graphite. If carbon atoms are connectedwith σ single bonds one obtains diamond crystal that is transparentand electrically insulating material. However, if carbon atoms areconnected by σ−π double bonds then one obtains graphite that ab-sorbs light very efficiently and is electrically conducting. This workfocuses on ultrathin carbon films with dominating σ − π doublebonds, pyrolytic carbon and graphene, which are closely related tographite. [12]

The first reports of pyrolytic carbon (PyC), highly amorphoussp2 hybridized carbon allotrope, produced by chemical vapor de-position are found in 1880s when PyC was found to be suitable ma-terial for strengthen an incandescent lamp filament [12, 13]. It tooknearly 90 years until some interest was paid to the PyC synthesis us-ing methane based chemical vapor deposition (CVD) [14, 15]. Thiswas probably because theoretically the properties of PyC, whichconsists of intertwined graphene flakes with size of few nanome-ters, are difficult to predict due to its low crystallinity. However,recent experimental findings have opened interesting opportunitiesfor applications of ultrathin PyC films which are electrically con-ductive, chemically inert and mechanically strong.

Graphite consists of multiple layers stacked to one another byrelatively weak van-der-Waals forces [16]. Similarly to individualsheets in the paper stack one can remove and isolate only a mono-


Tommi Kaplas: Synthesis and optical characterization of ultrathin carbonfilms

layer of graphite [17]. This monolayer of carbon atoms organizedin the honeycomb lattice is called graphene. Experimentally theproperties of graphene were first time observed in 2004 althoughits band structure was calculated already in 1947 [18]. Despite theproperties of graphene have been theoretically predicted and de-scribed a long time ago, the experiment in 2004 was ground break-ing in two respects. First, it was a clear demonstration that twodimensional materials do exist. Secondly, it made bring into prac-tice a new material with unique electronic properties that is widelyrecognized to be the key element of future electronic and optoelec-tronic devices [19, 20].

Electrons in graphene have zero effective mass, and can travelfor micrometers without scattering at room temperature [21]. Elec-tron transport in graphene is described by a Dirac-like equation al-lowing one to investigate relativistic quantum phenomena in a stan-dard laboratory [21]. In photonics and optoelectronics, graphenecreates a new material platform for ultra-fast lasing [22], micro- andnanomechanics [23] and quantum electronics [24]. Thus, it can leadto commercial applications in optical communication, biomedicine,gas sensing, solar harvesting and ultra-fast electronics [19–21]. The-refore graphene could benefit science and technology both together.

The bottleneck of the graphene R&D lies currently in the fabri-cation [20]. Graphene is produced by several techniques includingmicromechanical cleaving, dissolving Si from SiC, chemical exfoli-ation and chemical vapor deposition (CVD) [20, 21]. The later hasbeen widely recognized as most prospective for practical imple-mentation because it can be employed to obtain very large graphenesamples [25].

However, in the conventional CVD technique, graphene is de-posited on the surface of the transition metal catalyst, while mostof the application requires graphene deposited on the dielectric orsemiconductor substrate [20]. The necessity of graphene transfer toa dielectric is the major difficulty in the development of graphenebased photonic and optoelectronic devices because transferring anatomically thin film inevitable causes deterioration of the graphene

2 Dissertations in Forestry and Natural Sciences No 130

Introduction

film quality [26]. The transferring process could be avoided by de-positing graphene directly on a dielectric substrate that is widelyrecognized as an eagerly awaited breakthrough should open a wayfor integration of graphene into silicon-based platform of modernphotonics and optoelectronics [20].

In this thesis, the focus is on the synthesis of the graphene basedmaterials and especially on direct graphene synthesis on a dielectricsubstrate. The presented results include a conventional graphenesynthesis on a metallic substrate (paper I) that follows a direct de-position of amorphous PyC film on a silica substrate (paper II).The development direct deposition of a few layered graphene on asilica substrate (paper III) was further improved by precise spatialcontrol (papers IV and V). Furthermore in this Thesis, the Ramanspectroscopy and spectrophotometry are used for monitor the syn-thetized material.

The background for understanding the material synthesis andoptical properties of graphene are introduced in the Chapters 2-4.The main results of the papers I-V are divided into two sectionsthat are the results on optical characterization (Chapter 5) and theresults on direct graphene deposition on a silica substrate (Chapter6).



layer of graphite [17]. This monolayer of carbon atoms organizedin the honeycomb lattice is called graphene. Experimentally theproperties of graphene were first time observed in 2004 althoughits band structure was calculated already in 1947 [18]. Despite theproperties of graphene have been theoretically predicted and de-scribed a long time ago, the experiment in 2004 was ground break-ing in two respects. First, it was a clear demonstration that twodimensional materials do exist. Secondly, it made bring into prac-tice a new material with unique electronic properties that is widelyrecognized to be the key element of future electronic and optoelec-tronic devices [19, 20].

Electrons in graphene have zero effective mass, and can travelfor micrometers without scattering at room temperature [21]. Elec-tron transport in graphene is described by a Dirac-like equation al-lowing one to investigate relativistic quantum phenomena in a stan-dard laboratory [21]. In photonics and optoelectronics, graphenecreates a new material platform for ultra-fast lasing [22], micro- andnanomechanics [23] and quantum electronics [24]. Thus, it can leadto commercial applications in optical communication, biomedicine,gas sensing, solar harvesting and ultra-fast electronics [19–21]. The-refore graphene could benefit science and technology both together.

The bottleneck of the graphene R&D lies currently in the fabri-cation [20]. Graphene is produced by several techniques includingmicromechanical cleaving, dissolving Si from SiC, chemical exfoli-ation and chemical vapor deposition (CVD) [20, 21]. The later hasbeen widely recognized as most prospective for practical imple-mentation because it can be employed to obtain very large graphenesamples [25].

However, in the conventional CVD technique, graphene is de-posited on the surface of the transition metal catalyst, while mostof the application requires graphene deposited on the dielectric orsemiconductor substrate [20]. The necessity of graphene transfer toa dielectric is the major difficulty in the development of graphenebased photonic and optoelectronic devices because transferring anatomically thin film inevitable causes deterioration of the graphene


Introduction

film quality [26]. The transferring process could be avoided by de-positing graphene directly on a dielectric substrate that is widelyrecognized as an eagerly awaited breakthrough should open a wayfor integration of graphene into silicon-based platform of modernphotonics and optoelectronics [20].

In this thesis, the focus is on the synthesis of the graphene basedmaterials and especially on direct graphene synthesis on a dielectricsubstrate. The presented results include a conventional graphenesynthesis on a metallic substrate (paper I) that follows a direct de-position of amorphous PyC film on a silica substrate (paper II).The development direct deposition of a few layered graphene on asilica substrate (paper III) was further improved by precise spatialcontrol (papers IV and V). Furthermore in this Thesis, the Ramanspectroscopy and spectrophotometry are used for monitor the syn-thetized material.

The background for understanding the material synthesis andoptical properties of graphene are introduced in the Chapters 2-4.The main results of the papers I-V are divided into two sectionsthat are the results on optical characterization (Chapter 5) and theresults on direct graphene deposition on a silica substrate (Chapter6).




2 The sp2 hybridization ofcarbon

The main focus of this work is graphitic thin film synthesis. In thischapter the basic concept of the graphitic C=C bonding mechanismis briefly discussed below.

2.1 ELEMENTAL CARBON AND C=C BONDING

In the periodic table, the elemental carbon (C) is located under thegroup 14, also referred as the carbon group, with an atomic num-ber of six. Thus the elemental carbon has six electrons, which aredivided into groups of two and four on the K and the L shells, re-spectively. The two electrons on K shell, are located at the atomic1s orbital with the electron spins up and down. These two elec-trons are not involved in carbon bonding but the remaining fourelectrons on the L shell are in the main role when the carbon bondsare formed [12].

In a ground state, the L shell of the carbon atom is formed bythe electrons at 2s and the 2p orbitals. More specifically, the 2porbital consist of three orhogonal suborbitals, namely 2px, 2py and2pz. The electron construction of the ground stated carbon atomis shown in Tab. 2.1. However, when a carbon atom becomes ex-cited, one electron from the 2s orbital is lifted to 2p orbital leadingto sp hybridization. Newly formed orbitals are called spn orbitals,where n = 1, 2, 3 and denote the number of p orbitals involved.In graphitic materials, one 2s orbital is hybridized with two 2p or-bitals, thus in these materials sp2 hybridization dominates. It isworth noting that for sp2 hybridized carbon, one electron on p or-bital does not mix with s orbital. Usually the lone p orbital is definesthe z direction while the sp2 orbitals are in x − y-plane. The elec-tron orbital configuration of sp2 hybridized carbon is shown Tab.




2 The sp2 hybridization ofcarbon

The main focus of this work is graphitic thin film synthesis. In thischapter the basic concept of the graphitic C=C bonding mechanismis briefly discussed below.

2.1 ELEMENTAL CARBON AND C=C BONDING

In the periodic table, the elemental carbon (C) is located under thegroup 14, also referred as the carbon group, with an atomic num-ber of six. Thus the elemental carbon has six electrons, which aredivided into groups of two and four on the K and the L shells, re-spectively. The two electrons on K shell, are located at the atomic1s orbital with the electron spins up and down. These two elec-trons are not involved in carbon bonding but the remaining fourelectrons on the L shell are in the main role when the carbon bondsare formed [12].

In a ground state, the L shell of the carbon atom is formed bythe electrons at 2s and the 2p orbitals. More specifically, the 2porbital consist of three orhogonal suborbitals, namely 2px, 2py and2pz. The electron construction of the ground stated carbon atomis shown in Tab. 2.1. However, when a carbon atom becomes ex-cited, one electron from the 2s orbital is lifted to 2p orbital leadingto sp hybridization. Newly formed orbitals are called spn orbitals,where n = 1, 2, 3 and denote the number of p orbitals involved.In graphitic materials, one 2s orbital is hybridized with two 2p or-bitals, thus in these materials sp2 hybridization dominates. It isworth noting that for sp2 hybridized carbon, one electron on p or-bital does not mix with s orbital. Usually the lone p orbital is definesthe z direction while the sp2 orbitals are in x − y-plane. The elec-tron orbital configuration of sp2 hybridized carbon is shown Tab.



Table 2.1: Formation of sp2 hybridized carbon. In ground state (1) only two electron outof six are on the 2p orbitals. When the carbon atom becomes excited (2) one electron from2s orbital is lifted to 2p orbital. During this process the 2s and 2p orbitals are mixed tosp2 orbitals (3).

1 ↑↓ ↑↓ ↑ ↑1s 2s 2px 2py 2pz

2 ↑↓ ↑ ↑ ↑ ↑1s 2s 2px 2py 2pz

3 ↑↓ ↑ ↑ ↑ ↑1s sp2 sp2 sp2 pz

2.1. [12]When two sp2 hybridized carbon atoms meet, they form a co-

valent σ-π double bond. The σ-bond is formed by overlapping oftwo sp2 orbitals, and it binds the two carbon atoms tightly together.The π bond is formed by overlapping of the pz electron orbitals ofan adjacent carbon atoms. The graphitic C=C double bond makesthe material mechanically strong. [12]

In the sp2 hybridized carbon, all the sp2 orbitals are in the sameplane with angular difference of 120 while the pz electron orbital isperpendicular to this plane (see Fig. 4.3) [12]. Because of the strongσ-π double bond, the two sp2 hybridized carbons carbon atomscannot rotate with respect to another. As a result an ensemble of thesp2 hybridized carbon will settle on the same plane. Moreover, sincethe angle between sp2 orbitals is 120, the increase of the numberof atoms will eventually lead to formation of the hexagonal planarlattice. Fig. 4.3(d) shows the building element of such a structure,the aromatic carbon ring consisting of six carbon atoms, which canbe considered as a smallest possible form of a graphite.

The π bond is formed by overlapping of an adjacent carbonatoms. The delocalized electrons at pz orbitals responsible for π

bonds gives rise to the electrical conductivity of the carbon ma-terials with sp2 hybridization. Furthermore, since the σ-π bondsappear in one plane, the conductivity is therefore strongly polar-


The sp2 hybridization of carbon

Figure 2.1: (a) The isolated p (pink) and s (blue) orbitals of a carbon atom. (b) A sp2

hybridized carbon atom with asymmetrical sp2 orbitals, which lay in the same plane andare oriented at 120 degrees with respect one another and symmetrical pz orbital perpen-dicular to sp2 orbitals. (c) Two sp2 hybridized carbon atoms forms a σ-π double bond.(d) Graphitic carbon ring contains six carbon atoms that form a regular hexagon boundedwith C=C bonds. The electrons at pz orbitals, which are responsible for the π bonds, aredelocalized.

ized. [12, 16, 27]Increasing the size of the aromatic structure by adding sp2 hy-

bridized carbon atoms, will lead to formation of a single plane ofcarbon atoms. In such a two-dimensional ensemble every atom issurrounded by three adjacent neighbors. This honeycomb mono-layer of carbon atoms is called graphene. When multiple graphenesheets are piled up we arrive at graphite (see Fig. 2.2). The distancebetween carbons in the plane is 1.42 A, while the spacing betweengraphene layers in graphite is 3.34 A. In three-dimensional graphitecrystal, the graphene layers are attached to each other by ratherweak van der Waals forces, which allow one to separate graphenelayers from graphite. [16, 17]

Graphene and graphite are carbon allotropes that belong to the”graphite family” which includes also fullerenes and carbon nan-otubes. The members of graphite family are presented in the Fig.



Table 2.1: Formation of sp2 hybridized carbon. In ground state (1) only two electron outof six are on the 2p orbitals. When the carbon atom becomes excited (2) one electron from2s orbital is lifted to 2p orbital. During this process the 2s and 2p orbitals are mixed tosp2 orbitals (3).

1 ↑↓ ↑↓ ↑ ↑1s 2s 2px 2py 2pz

2 ↑↓ ↑ ↑ ↑ ↑1s 2s 2px 2py 2pz

3 ↑↓ ↑ ↑ ↑ ↑1s sp2 sp2 sp2 pz

2.1. [12]When two sp2 hybridized carbon atoms meet, they form a co-

valent σ-π double bond. The σ-bond is formed by overlapping oftwo sp2 orbitals, and it binds the two carbon atoms tightly together.The π bond is formed by overlapping of the pz electron orbitals ofan adjacent carbon atoms. The graphitic C=C double bond makesthe material mechanically strong. [12]

In the sp2 hybridized carbon, all the sp2 orbitals are in the sameplane with angular difference of 120 while the pz electron orbital isperpendicular to this plane (see Fig. 4.3) [12]. Because of the strongσ-π double bond, the two sp2 hybridized carbons carbon atomscannot rotate with respect to another. As a result an ensemble of thesp2 hybridized carbon will settle on the same plane. Moreover, sincethe angle between sp2 orbitals is 120, the increase of the numberof atoms will eventually lead to formation of the hexagonal planarlattice. Fig. 4.3(d) shows the building element of such a structure,the aromatic carbon ring consisting of six carbon atoms, which canbe considered as a smallest possible form of a graphite.

The π bond is formed by overlapping of an adjacent carbonatoms. The delocalized electrons at pz orbitals responsible for π

bonds gives rise to the electrical conductivity of the carbon ma-terials with sp2 hybridization. Furthermore, since the σ-π bondsappear in one plane, the conductivity is therefore strongly polar-



Figure 2.1: (a) The isolated p (pink) and s (blue) orbitals of a carbon atom. (b) A sp2

hybridized carbon atom with asymmetrical sp2 orbitals, which lay in the same plane andare oriented at 120 degrees with respect one another and symmetrical pz orbital perpen-dicular to sp2 orbitals. (c) Two sp2 hybridized carbon atoms forms a σ-π double bond.(d) Graphitic carbon ring contains six carbon atoms that form a regular hexagon boundedwith C=C bonds. The electrons at pz orbitals, which are responsible for the π bonds, aredelocalized.

ized. [12, 16, 27]Increasing the size of the aromatic structure by adding sp2 hy-

bridized carbon atoms, will lead to formation of a single plane ofcarbon atoms. In such a two-dimensional ensemble every atom issurrounded by three adjacent neighbors. This honeycomb mono-layer of carbon atoms is called graphene. When multiple graphenesheets are piled up we arrive at graphite (see Fig. 2.2). The distancebetween carbons in the plane is 1.42 A, while the spacing betweengraphene layers in graphite is 3.34 A. In three-dimensional graphitecrystal, the graphene layers are attached to each other by ratherweak van der Waals forces, which allow one to separate graphenelayers from graphite. [16, 17]

Graphene and graphite are carbon allotropes that belong to the”graphite family” which includes also fullerenes and carbon nan-otubes. The members of graphite family are presented in the Fig.



Figure 2.2: Graphene is considered as two-dimensional material and it is the building blockfor (0D) fullerenes (1D) carbon nanotubes and (3D) graphite.

2.2. It is worth noting that zero-dimensional fullerene, one-dimensio-nal carbon nanotube and three-dimensional graphite can be recon-structed from graphene. [19, 28]

In this work, we are focused on the single layer graphene (SLG)and a few layer graphene (FLG), i.e. an ultrathin graphite crys-tal. We also consider synthesis and characterization of a pyrolyticcarbon (PyC) films. PyC is one of carbon allotropes with domi-nating sp2 bonding, yet it is not considered as a member of thegraphite family since it is strongly amorphous [12]. StructurallyPyC can be considered as an amorphous graphite consisting inter-twined graphene flakes of only few nanometers in size.

2.2 GRAPHENE BAND STRUCTURE

The optical properties of materials belonging to the graphite fam-ily originate from the electronic properties and in particular fromgraphene band structure [29, 30], which will be briefly introducedin this paragraph.

Because of the symmetrical hexagonal nature of a two-dimensio-nal graphene lattice, the unit cell of graphene is formed by twoatoms (A and B in the Fig. 2.3) with two unit cell vectors that are



Figure 2.3: (a) An unit cell of graphene consist only two atoms (A and B) with unit cellvectors a1,2. (b) The Bravais lattice is defined by the vectors b1,2. In the middle of thefirst Brillouin zone (blue hexagon) is Γ point and on corners there are K and K′ points.Between K and K′ points is M point.

a1,2 =a2(3,±

√3), (2.1)

where a ∼= 1.42 A is distance between two adjacent carbon atoms.The Bravais lattice of graphene is defined by vectors

b1,2 =2π

3a(1,±

√3). (2.2)

Fig. 2.3 shows a symmetrical sp2 hybridized carbon lattice and thefirst Brillouin zone with Γ, M and K points. [31]

By using the tight binding approximation, one can make a closeestimation for graphene band structure (presented in Fig. 2.4) [18,32,33]. In the approximation an ideal sp2 hybridized graphene crys-tal is formed by two atoms. Furthermore, since the high energyσ-bonds do not affect to the electronic structure at lower energies,it is reasonable to take into account only pz orbital [31]. In suchcase the electronic band structure of graphene can be presented byfollowing form:



Figure 2.2: Graphene is considered as two-dimensional material and it is the building blockfor (0D) fullerenes (1D) carbon nanotubes and (3D) graphite.

2.2. It is worth noting that zero-dimensional fullerene, one-dimensio-nal carbon nanotube and three-dimensional graphite can be recon-structed from graphene. [19, 28]

In this work, we are focused on the single layer graphene (SLG)and a few layer graphene (FLG), i.e. an ultrathin graphite crys-tal. We also consider synthesis and characterization of a pyrolyticcarbon (PyC) films. PyC is one of carbon allotropes with domi-nating sp2 bonding, yet it is not considered as a member of thegraphite family since it is strongly amorphous [12]. StructurallyPyC can be considered as an amorphous graphite consisting inter-twined graphene flakes of only few nanometers in size.

2.2 GRAPHENE BAND STRUCTURE

The optical properties of materials belonging to the graphite fam-ily originate from the electronic properties and in particular fromgraphene band structure [29, 30], which will be briefly introducedin this paragraph.

Because of the symmetrical hexagonal nature of a two-dimensio-nal graphene lattice, the unit cell of graphene is formed by twoatoms (A and B in the Fig. 2.3) with two unit cell vectors that are



Figure 2.3: (a) An unit cell of graphene consist only two atoms (A and B) with unit cellvectors a1,2. (b) The Bravais lattice is defined by the vectors b1,2. In the middle of thefirst Brillouin zone (blue hexagon) is Γ point and on corners there are K and K′ points.Between K and K′ points is M point.

a1,2 =a2(3,±

√3), (2.1)

where a ∼= 1.42 A is distance between two adjacent carbon atoms.The Bravais lattice of graphene is defined by vectors

b1,2 =2π

3a(1,±

√3). (2.2)

Fig. 2.3 shows a symmetrical sp2 hybridized carbon lattice and thefirst Brillouin zone with Γ, M and K points. [31]

By using the tight binding approximation, one can make a closeestimation for graphene band structure (presented in Fig. 2.4) [18,32,33]. In the approximation an ideal sp2 hybridized graphene crys-tal is formed by two atoms. Furthermore, since the high energyσ-bonds do not affect to the electronic structure at lower energies,it is reasonable to take into account only pz orbital [31]. In suchcase the electronic band structure of graphene can be presented byfollowing form:



Figure 2.4: The band structure of graphene. In the K and K′ points the valence andconductance bands touch and near these points the band structure is linear. Near M pointthe band structure changes to a nonlinear one.

E(k) = ±γ0

1 + 4 cos2

(√3

2aky

)+ 4 cos

(√3

2aky

)cos

(32

akx

),

(2.3)

where the signs ”+” and ”-” corresponds to the conductance andvalence band, respectively, and γ0 ∼= 2.8 eV is the ”hopping ener-gies” of nearest neighbor. [31]

In the Fig. 2.4 is presented the band structure calculated by us-ing Eq. 2.3. In the optical spectral range the crossing of the valenceand conductance band at Fermi energy near the K and K′ points ofthe Brilluen zone is the most interesting feature of graphene band-structure. At these points, the electron energy is a linear functionof the electron momentum, i.e. it is described by the Dirac equation

E(k) = hvF|k|, (2.4)

where the Fermi velocity is given by vF = 32h aγ0. Especially this is

interesting for optical frequencies because optical and infra red (IR)range lies close to Dirac points. [18, 31]

The gapless, linear band structure is a property of only a singlelayered graphene since the unit cell of graphene consist only two



carbon atoms. Unit cell in a graphite with more than one layerconsist of four carbon atoms. Therefore the band structure is inthe vicinity of the Dirac points is not linear and thus different fromthat of graphene (see Fig. 2.4) [34–38]. However, these differencesmanifest themselves for the THz and far-IR photons, while in thevisual and near-IR spectral range, the properties of SGL and FGLare nearly the same.



Figure 2.4: The band structure of graphene. In the K and K′ points the valence andconductance bands touch and near these points the band structure is linear. Near M pointthe band structure changes to a nonlinear one.

E(k) = ±γ0

1 + 4 cos2

(√3

2aky

)+ 4 cos

(√3

2aky

)cos

(32

akx

),

(2.3)

where the signs ”+” and ”-” corresponds to the conductance andvalence band, respectively, and γ0 ∼= 2.8 eV is the ”hopping ener-gies” of nearest neighbor. [31]

In the Fig. 2.4 is presented the band structure calculated by us-ing Eq. 2.3. In the optical spectral range the crossing of the valenceand conductance band at Fermi energy near the K and K′ points ofthe Brilluen zone is the most interesting feature of graphene band-structure. At these points, the electron energy is a linear functionof the electron momentum, i.e. it is described by the Dirac equation

E(k) = hvF|k|, (2.4)

where the Fermi velocity is given by vF = 32h aγ0. Especially this is

interesting for optical frequencies because optical and infra red (IR)range lies close to Dirac points. [18, 31]

The gapless, linear band structure is a property of only a singlelayered graphene since the unit cell of graphene consist only two



carbon atoms. Unit cell in a graphite with more than one layerconsist of four carbon atoms. Therefore the band structure is inthe vicinity of the Dirac points is not linear and thus different fromthat of graphene (see Fig. 2.4) [34–38]. However, these differencesmanifest themselves for the THz and far-IR photons, while in thevisual and near-IR spectral range, the properties of SGL and FGLare nearly the same.




3 Optical spectroscopy ofgraphene based thin films

The optical response of graphene and ultrathin carbon films aredetermined by the band structure. In graphene the linear and gap-less π band structure at Dirac points strongly influences the Ramanscattering and the linear refraction of light in graphene [29, 30, 39].Both these techniques are widely used for optical characterizationof various carbon materials.

3.1 RAMAN SCATTERING IN GRAPHENE AND PYROLYTICCARBON

Raman spectroscopy is a versatile tool that are conventionally usedfor recognizing different carbon allotropes [39–43]. The first Ra-man spectra of graphite were measured and reported more than40 years ago and recently Raman spectroscopy has become a ma-jor, nondestructive tool for probing graphitic materials [39, 44]. Inparticular, the Raman spectroscopy is conventionally used to iden-tify graphene [45, 46], and to determine number of graphene lay-ers [45, 47], doping level [48–50], strain [51, 52], electrical mobil-ity [53] and disordering of graphene lattice [54].

In this work we studied Raman spectra range in the wavenum-ber range from 1000 cm−1 to 3500 cm−1. At this the Raman spec-trum of pristine graphene is dominated by two peaks known as Gand 2D peaks at 1581 cm−1 and ∼2680 cm−1, respectively [39, 55].Furthermore, the magnitude of the so-called D peak at 1350 cm−1 isa quatitative measure of the graphene disorder [55]. By analyzingthe relative intensity, full width half maximum (FWHM) and posi-tion of these three peaks, it is possible to reveal information aboutthe crystallinity of the graphite material.




3 Optical spectroscopy ofgraphene based thin films

The optical response of graphene and ultrathin carbon films aredetermined by the band structure. In graphene the linear and gap-less π band structure at Dirac points strongly influences the Ramanscattering and the linear refraction of light in graphene [29, 30, 39].Both these techniques are widely used for optical characterizationof various carbon materials.

3.1 RAMAN SCATTERING IN GRAPHENE AND PYROLYTICCARBON

Raman spectroscopy is a versatile tool that are conventionally usedfor recognizing different carbon allotropes [39–43]. The first Ra-man spectra of graphite were measured and reported more than40 years ago and recently Raman spectroscopy has become a ma-jor, nondestructive tool for probing graphitic materials [39, 44]. Inparticular, the Raman spectroscopy is conventionally used to iden-tify graphene [45, 46], and to determine number of graphene lay-ers [45, 47], doping level [48–50], strain [51, 52], electrical mobil-ity [53] and disordering of graphene lattice [54].

In this work we studied Raman spectra range in the wavenum-ber range from 1000 cm−1 to 3500 cm−1. At this the Raman spec-trum of pristine graphene is dominated by two peaks known as Gand 2D peaks at 1581 cm−1 and ∼2680 cm−1, respectively [39, 55].Furthermore, the magnitude of the so-called D peak at 1350 cm−1 isa quatitative measure of the graphene disorder [55]. By analyzingthe relative intensity, full width half maximum (FWHM) and posi-tion of these three peaks, it is possible to reveal information aboutthe crystallinity of the graphite material.



Figure 3.1: a) The vibration direction of carbon atoms responsible for the G peak. b) Thebreathing mode of the disorder induced D peak.

The G peak is generated by the in-plane phonon (see Fig. 3.1 (a))at the Γ point and exist in the Raman spectrum of all sp2 hybridizedcarbon allotropes [56,57]. That is one can find the G peak also fromone-dimensional sp2 bonded carbon chains as well as from aromaticcarbon structures. This makes the G peak to be a good reference forthe D and 2D peaks.

The D peak originates from phonon at the K point and is thebreathing mode of six carbon atoms (see Fig. 3.1) [56]. In the pris-tine graphene, the breathing mode is suppressed and the D peak isactivated only when there are defects in the graphene lattice [56,58].Therefore, the magnitude of the D peak can be used as a quantita-tive measure of the crystallinity of a graphene sample.

The 2D peak does not need a defect for activation being orig-inated from scattering of photon on the pair of phonons with op-posite momenta [39, 40, 55]. The band structure of graphene willstrongly affect the 2D peak because it occures due to the interval-ley scattering of the Dirac cones at K and K′ [39]. Therefore, sincea monolayer graphene has different band structure compared to afew layered graphene the 2D peak can be used as a signature of themonolayer graphene.

Overall the relative magnitudes of D, G and 2D peaks in the Ra-man spectrum of graphene can be used as an indicator of the mate-rial quality [39,44]. Specifically, the ratio of D and G peaks is indica-tor of the crystallinity of the graphene sample, while the position of2D peak hints for the thickness of a few layered graphene [44,45,47].


Optical spectroscopy of graphene based thin films

Since the D peak requires the presence of a defect and the G peakis assumed to appear in all sp2 hybridized material, the ratio of theD and G peaks can be employed for quantitative characterizationof the crystallinity. If graphitic material does not consist any amor-phous carbon and sp3 bonds, one can use Tuinstra-Koening(TK)-relation (TK) to obtain the size of graphene crystallite La [44,56,59]:

La(nm) = C(λ)×(

I(D)

I(G)

)−1

, (3.1)

where the normalization coefficient C(λ) is a function of the exci-tation wavelength λ. The exact value for C(λ) is still under discus-sion and currently in literature two different values can be foundC(λ = 514nm) = 4.4 [60] or C(λ = 514nm) ≈ 16.7 [61]. It isworth noting that since the former and latter value differs aboutfour times the TK relation gives only a rough estimation for thesize of the graphene crystal.

In the Raman spectrum, an amorphous carbon manifest itself asa wide peak at between 1500 cm−1 and 1550 cm−1 [62]. If the ma-terial is highly amorphous (La ≤ 2 nm) and contains also traces ofan amorphous carbon, the La should be estimated by the followingrelation suggested by Ferrari and Robertson [56]:

L2a(nm) =

1C′(λ)

I(D)

I(G), (3.2)

where the normalization coefficient C′(λ) is 0.55 for 514 nm excita-tion wavelength [56].

Thus in the Raman spectrum for purely sp2 hybridized materi-als, I(D)/I(G) ∝ 1/La, while for the highly amorphous graphiteI(D)/I(G) ∝ L2

a.The G peak width and position gives information about the

graphite material crystallinity. Specifically, for a single crystallinegraphite, the full-width-half-maximum of the G peak (FWHMG) isas low as 12 cm−1, while for highly amorphous graphite the G peak



Figure 3.1: a) The vibration direction of carbon atoms responsible for the G peak. b) Thebreathing mode of the disorder induced D peak.

The G peak is generated by the in-plane phonon (see Fig. 3.1 (a))at the Γ point and exist in the Raman spectrum of all sp2 hybridizedcarbon allotropes [56,57]. That is one can find the G peak also fromone-dimensional sp2 bonded carbon chains as well as from aromaticcarbon structures. This makes the G peak to be a good reference forthe D and 2D peaks.

The D peak originates from phonon at the K point and is thebreathing mode of six carbon atoms (see Fig. 3.1) [56]. In the pris-tine graphene, the breathing mode is suppressed and the D peak isactivated only when there are defects in the graphene lattice [56,58].Therefore, the magnitude of the D peak can be used as a quantita-tive measure of the crystallinity of a graphene sample.

The 2D peak does not need a defect for activation being orig-inated from scattering of photon on the pair of phonons with op-posite momenta [39, 40, 55]. The band structure of graphene willstrongly affect the 2D peak because it occures due to the interval-ley scattering of the Dirac cones at K and K′ [39]. Therefore, sincea monolayer graphene has different band structure compared to afew layered graphene the 2D peak can be used as a signature of themonolayer graphene.

Overall the relative magnitudes of D, G and 2D peaks in the Ra-man spectrum of graphene can be used as an indicator of the mate-rial quality [39,44]. Specifically, the ratio of D and G peaks is indica-tor of the crystallinity of the graphene sample, while the position of2D peak hints for the thickness of a few layered graphene [44,45,47].



Since the D peak requires the presence of a defect and the G peakis assumed to appear in all sp2 hybridized material, the ratio of theD and G peaks can be employed for quantitative characterizationof the crystallinity. If graphitic material does not consist any amor-phous carbon and sp3 bonds, one can use Tuinstra-Koening(TK)-relation (TK) to obtain the size of graphene crystallite La [44,56,59]:

La(nm) = C(λ)×(

I(D)

I(G)

)−1

, (3.1)

where the normalization coefficient C(λ) is a function of the exci-tation wavelength λ. The exact value for C(λ) is still under discus-sion and currently in literature two different values can be foundC(λ = 514nm) = 4.4 [60] or C(λ = 514nm) ≈ 16.7 [61]. It isworth noting that since the former and latter value differs aboutfour times the TK relation gives only a rough estimation for thesize of the graphene crystal.

In the Raman spectrum, an amorphous carbon manifest itself asa wide peak at between 1500 cm−1 and 1550 cm−1 [62]. If the ma-terial is highly amorphous (La ≤ 2 nm) and contains also traces ofan amorphous carbon, the La should be estimated by the followingrelation suggested by Ferrari and Robertson [56]:

L2a(nm) =

1C′(λ)

I(D)

I(G), (3.2)

where the normalization coefficient C′(λ) is 0.55 for 514 nm excita-tion wavelength [56].

Thus in the Raman spectrum for purely sp2 hybridized materi-als, I(D)/I(G) ∝ 1/La, while for the highly amorphous graphiteI(D)/I(G) ∝ L2

a.The G peak width and position gives information about the

graphite material crystallinity. Specifically, for a single crystallinegraphite, the full-width-half-maximum of the G peak (FWHMG) isas low as 12 cm−1, while for highly amorphous graphite the G peak



Figure 3.2: a) The Raman spectra of graphite and graphene shows the main peak positionsof D, G and 2D at 1350 −1, 1581 −1 and 2700 −1. The most significant difference is the2D peak shifting and the 2D peak shape. Also the ratio of the I(G)/(2D) is noteworthy. b)Wide Raman peaks of PyC indicate a highly disordered graphitic material. Essentially theD, G and 2D peaks of graphite are found in PyC but are greatly broadened. Furthermorethere is a broad peak of amorphous carbon at 1500 −1 and disorder induced peak at 2930cm−1.

widens and the FWHMG can be more than 100 cm−1 [59]. More-over, the G peak can be shifted from its original position of 1581cm−1 towards 1600 cm−1 if the graphitic flakes are nanosized ortowards 1500 cm−1 if the material consist 20 % or more amorphouscarbon [59]. Although, the G peak widening or peak shifting onlygives a very rough estimation for the crystallinity it still backs upthe estimation obtained by Eq. 3.1 and 3.2.

The shifting and broadening of the 2D peak in a few layer grap-hene can be related to the number of graphene layers. In bulkgraphite, the 2D peak is located at ∼2730 cm−1 [42], while in thegraphene monolayer, the 2D peak is found at ∼2680 cm−1 [46].However, for a few layered graphene the 2D peak is typically lo-cated near 2700 cm−1 [45, 46]. The FWHM2D and the shape of 2Dpeak is associated also to the thickness of a few layered graphene[45, 46]. For a monolayer graphene, the 2D peak can be fitted asa single Lorentzian peak with FWHM2D of about 24 cm−1 [55],but for a multilayered structure four or more Lorentzian peaks areusually used for an accurate fitting and the FWHM2D is usuallybetween 45-60 cm−1 [55]. Typical examples of graphene, graphite



and PyC Raman spectra is shown in Fig. 3.2.The D, G and 2D peaks are not the only peaks in the graphitic

material. It is worth of mentioning that defects and doping willmanifest themselves as a shoulder in the G peak close to 1600cm−1 (sometimes referred as D′) and the combined D + D′ is seennear 3000 cm−1. The overtone of D′, i.e. the 2D′ is located at 3250cm-1 and does not require defect like the D′ does. Near 2450 cm−1

is found a D + D′′ peak that is raised by the same mechanism as itis suggested for 2D peak. [39]

3.2 SPECTROPHOTOMETRY OF NANOCARBON FILMS

The linear optical properties of suspended graphene is defined byits extraordinary band structure. Specifically, the linear gaplessband structure of graphene in the vicinity of K and K′ points re-sults in the independence of the absorption cross-section for pho-tons with energies lower than 2.5 eV (at light wavelength longerthan ∼500 nm) cross-section. When the photon energy is higherthan 2.5 eV the band structure changes due the M saddle pointwhich is shown as an absorption peak near 270 nm (4.6 eV) in thetransmission spectrum (see Fig. 3.3) [63].

At the photon energy below 2.5 eV, the optical absorbtion of asingle graphene sheet is defined by the fine structure constant α byAg = πα ≈ π/137 [30]. Thus a graphene monolayer absorbs about2.3 % of the incident light energy, which is extremely high sincegraphene sheet is only one atom thick. Furthermore, although theband structure of a few layered graphene is bit different in the Kand K′ points compared to a monolayer [64,65], the absorption of afew layered graphene film increases linearly as a function of layersnumber in the visual and near-IR spectral range.

When the photon energy is below 2.5 eV, the optical transmit-tance (T) and reflectance (R) of graphene are also independent onthe wavelength and can be expressed as T = (1 + 0.5πα)−2 ≈1 − πα and R = (0.5πα)2T ≈ 0 [30]. It is worth noting, however,that it is correct only for a suspended graphene since an underlying



Figure 3.2: a) The Raman spectra of graphite and graphene shows the main peak positionsof D, G and 2D at 1350 −1, 1581 −1 and 2700 −1. The most significant difference is the2D peak shifting and the 2D peak shape. Also the ratio of the I(G)/(2D) is noteworthy. b)Wide Raman peaks of PyC indicate a highly disordered graphitic material. Essentially theD, G and 2D peaks of graphite are found in PyC but are greatly broadened. Furthermorethere is a broad peak of amorphous carbon at 1500 −1 and disorder induced peak at 2930cm−1.

widens and the FWHMG can be more than 100 cm−1 [59]. More-over, the G peak can be shifted from its original position of 1581cm−1 towards 1600 cm−1 if the graphitic flakes are nanosized ortowards 1500 cm−1 if the material consist 20 % or more amorphouscarbon [59]. Although, the G peak widening or peak shifting onlygives a very rough estimation for the crystallinity it still backs upthe estimation obtained by Eq. 3.1 and 3.2.

The shifting and broadening of the 2D peak in a few layer grap-hene can be related to the number of graphene layers. In bulkgraphite, the 2D peak is located at ∼2730 cm−1 [42], while in thegraphene monolayer, the 2D peak is found at ∼2680 cm−1 [46].However, for a few layered graphene the 2D peak is typically lo-cated near 2700 cm−1 [45, 46]. The FWHM2D and the shape of 2Dpeak is associated also to the thickness of a few layered graphene[45, 46]. For a monolayer graphene, the 2D peak can be fitted asa single Lorentzian peak with FWHM2D of about 24 cm−1 [55],but for a multilayered structure four or more Lorentzian peaks areusually used for an accurate fitting and the FWHM2D is usuallybetween 45-60 cm−1 [55]. Typical examples of graphene, graphite



and PyC Raman spectra is shown in Fig. 3.2.The D, G and 2D peaks are not the only peaks in the graphitic

material. It is worth of mentioning that defects and doping willmanifest themselves as a shoulder in the G peak close to 1600cm−1 (sometimes referred as D′) and the combined D + D′ is seennear 3000 cm−1. The overtone of D′, i.e. the 2D′ is located at 3250cm-1 and does not require defect like the D′ does. Near 2450 cm−1

is found a D + D′′ peak that is raised by the same mechanism as itis suggested for 2D peak. [39]

3.2 SPECTROPHOTOMETRY OF NANOCARBON FILMS

The linear optical properties of suspended graphene is defined byits extraordinary band structure. Specifically, the linear gaplessband structure of graphene in the vicinity of K and K′ points re-sults in the independence of the absorption cross-section for pho-tons with energies lower than 2.5 eV (at light wavelength longerthan ∼500 nm) cross-section. When the photon energy is higherthan 2.5 eV the band structure changes due the M saddle pointwhich is shown as an absorption peak near 270 nm (4.6 eV) in thetransmission spectrum (see Fig. 3.3) [63].

At the photon energy below 2.5 eV, the optical absorbtion of asingle graphene sheet is defined by the fine structure constant α byAg = πα ≈ π/137 [30]. Thus a graphene monolayer absorbs about2.3 % of the incident light energy, which is extremely high sincegraphene sheet is only one atom thick. Furthermore, although theband structure of a few layered graphene is bit different in the Kand K′ points compared to a monolayer [64,65], the absorption of afew layered graphene film increases linearly as a function of layersnumber in the visual and near-IR spectral range.

When the photon energy is below 2.5 eV, the optical transmit-tance (T) and reflectance (R) of graphene are also independent onthe wavelength and can be expressed as T = (1 + 0.5πα)−2 ≈1 − πα and R = (0.5πα)2T ≈ 0 [30]. It is worth noting, however,that it is correct only for a suspended graphene since an underlying



Figure 3.3: Transmittance spectra of FLG and PyC films shows that the transmittanceof pyrolytic carbon resembles with some small differences. The M saddle point inducedabsorbtion peak is much wider in PyC and furthermore the absorption at ∼ 270 nm isnot as strong in PyC as it is in SLG and FLG. (∆T = TMAX − TMIN, where TMAX andTMIN are transmittances at 1500 nm and 270 nm wavelengths.)

substrate will affect the transmittance and reflectance [29].At the normal incidence, the graphene film deposited on a sub-

strate is described by the Fresnel equation [66]:

Tgs = 4ns(1 + ns + πα)−2, (3.3)

where the n is the refractive index of the dielectric substrate. Themeasured in the conventional experiment [25] differential transmit-tance of the bare and graphene coated substrate ∆T = (Ts − Tgs)/Ts

is strongly influenced by the ns. For example ∆T ≈ 1.9% pergraphene layer when graphene is deposited on the quartz substrate(ns ≈ 1.5). Similarly the differential reflectance ∆R is also deter-mined by ns [67],

∆R = 4Ag/(n2s − 1). (3.4)

In Eq. 3.4, the substrate for the graphene film is assumed tobe a thick and transparent. The Eq. 3.4 does not apply for metals



but since the refractive index of graphene (ng) is experimentallywell defined [68], the reflectance of a graphene coated substrate canbe expressed by Fresnel equation as a function of the carbon filmthickness h:

∆R = R(h)gs =

∣∣∣∣r12 + r23 exp(2iβ)1 + r12r23 exp(2iβ

∣∣∣∣2

, (3.5)

where at normal incidence r12 = (1 − ng)/(1 + ng) and r23 = (ng −ns)/(ng +ns) are reflection coefficients of the air-graphene and grap-hene substrate interfaces, respectively and β = 2πngh/λ. By usingEq. 3.5 and pre-described ng [68] one can define ∆R and the thick-ness of a few layered on an arbitrary substrate (see paper I).



Figure 3.3: Transmittance spectra of FLG and PyC films shows that the transmittanceof pyrolytic carbon resembles with some small differences. The M saddle point inducedabsorbtion peak is much wider in PyC and furthermore the absorption at ∼ 270 nm isnot as strong in PyC as it is in SLG and FLG. (∆T = TMAX − TMIN, where TMAX andTMIN are transmittances at 1500 nm and 270 nm wavelengths.)

substrate will affect the transmittance and reflectance [29].At the normal incidence, the graphene film deposited on a sub-

strate is described by the Fresnel equation [66]:

Tgs = 4ns(1 + ns + πα)−2, (3.3)

where the n is the refractive index of the dielectric substrate. Themeasured in the conventional experiment [25] differential transmit-tance of the bare and graphene coated substrate ∆T = (Ts − Tgs)/Ts

is strongly influenced by the ns. For example ∆T ≈ 1.9% pergraphene layer when graphene is deposited on the quartz substrate(ns ≈ 1.5). Similarly the differential reflectance ∆R is also deter-mined by ns [67],

∆R = 4Ag/(n2s − 1). (3.4)

In Eq. 3.4, the substrate for the graphene film is assumed tobe a thick and transparent. The Eq. 3.4 does not apply for metals



but since the refractive index of graphene (ng) is experimentallywell defined [68], the reflectance of a graphene coated substrate canbe expressed by Fresnel equation as a function of the carbon filmthickness h:

∆R = R(h)gs =

∣∣∣∣r12 + r23 exp(2iβ)1 + r12r23 exp(2iβ

∣∣∣∣2

, (3.5)

where at normal incidence r12 = (1 − ng)/(1 + ng) and r23 = (ng −ns)/(ng +ns) are reflection coefficients of the air-graphene and grap-hene substrate interfaces, respectively and β = 2πngh/λ. By usingEq. 3.5 and pre-described ng [68] one can define ∆R and the thick-ness of a few layered on an arbitrary substrate (see paper I).




4 Fabrication of ultrathinnanocarbon films usingchemical vapor deposition

The chemical vapor deposition (CVD) is the major route towardsfabrication of graphitic carbon materials including graphene andPyC films. In the CVD process, a gaseous precursor is decomposedat high temperature on the substrate surface to produce the desiredmaterial. In this chapter, we describe the methane based CVD ofPyC film and graphene.

4.1 CHEMICAL VAPOR DEPOSITION OF PYROLYTIC CAR-BON

A thin pyrolytic carbon (PyC) film can be produced by CVD di-rectly on a substrate surface providing it holds the 1000 C processtemperature [69–74]. Eventually the produced carbon film is highlyamorphous since the technique allows confocal and almost atom-ically smooth carbon coating with well defined thickness [73, 75].Furthermore, a PyC film can be deposited on any substrate (dielec-tric or metallic) opening an attractive route to making a conductingcarbon coating of arbitrary shaped substrates.

Generally in the CVD of PyC, one can define a homogenous anda heterogenous parts [69]. In the homogenous part of the processthe chemical reactions take place in gaseous atmosphere whereasin heterogenous part the molecular reaction takes place in the in-terface of the gas and substrate [76].

In this work, methane was employed as a gaseous carbon pre-cursor. The methane (CH4) molecule contains one carbon atombound to four hydrogen atoms by σ bonds. Due to the low reactiv-




4 Fabrication of ultrathinnanocarbon films usingchemical vapor deposition

The chemical vapor deposition (CVD) is the major route towardsfabrication of graphitic carbon materials including graphene andPyC films. In the CVD process, a gaseous precursor is decomposedat high temperature on the substrate surface to produce the desiredmaterial. In this chapter, we describe the methane based CVD ofPyC film and graphene.

4.1 CHEMICAL VAPOR DEPOSITION OF PYROLYTIC CAR-BON

A thin pyrolytic carbon (PyC) film can be produced by CVD di-rectly on a substrate surface providing it holds the 1000 C processtemperature [69–74]. Eventually the produced carbon film is highlyamorphous since the technique allows confocal and almost atom-ically smooth carbon coating with well defined thickness [73, 75].Furthermore, a PyC film can be deposited on any substrate (dielec-tric or metallic) opening an attractive route to making a conductingcarbon coating of arbitrary shaped substrates.

Generally in the CVD of PyC, one can define a homogenous anda heterogenous parts [69]. In the homogenous part of the processthe chemical reactions take place in gaseous atmosphere whereasin heterogenous part the molecular reaction takes place in the in-terface of the gas and substrate [76].

In this work, methane was employed as a gaseous carbon pre-cursor. The methane (CH4) molecule contains one carbon atombound to four hydrogen atoms by σ bonds. Due to the low reactiv-



ity of the methane in comparison to other hydrocarbon gasses, theCVD process temperatures are typically more than 900 C [69, 72].The main reaction sequence from methane to benzene-like aromaticstructures involves C2- and C4-species and can be as following:

CH4 → C2Hx → C4Hy → C6Hz, (4.1)

where x, y and z are the number of hydrogen atoms bound to car-bon atoms [72]. The effect of the unstable C3 and C5-species issupposed to play only minor role [72].

The evolving from C1 to C2- and C6-species are believed to behomogenous reactions and for further develop the heterogenousreactions are expected to take place on the substrate surface [69].While growing in size the ring-like carbon structures will settle ona substrate surface being attached to active sites. After attachmentthe carbon rings start evolving to more complex polycyclic hydro-carbons. Although the size of an individual polycyclic hydrocar-bons remains below 10 nm the small pyrocarbon clusters intertwinetogether forming a continuous carbon film [73].

Since the graphitic film grown on a substrate surface consist ofcrosslinked nanosize graphene clusters, the film can be expectedto be highly amorphous. However, surface of this amorphous filmis observed to be almost atomically smooth. Typically the averagesurface roughness of a prepared film (see paper II) on a silica sub-strate is in order of 1 nm. Although such a smoothness of the PyCfilms is still not understood, one may expect that it may be causedby the preferred orientation of the nanosize graphene clusters arealong the surface plane. Therefore, despite the PyC film is highlyamorphous, is very uniform and can be employed for nanoscalecoating material [75].

Because of high amorphicity of the PyC film the electron scatter-ing in the material is stronger than in graphene or graphite that isseen as higher sheet resistance [25, 73]. The strong electron scatter-ing in PyC suppresses the conductivity making this material hardlysuitable for transparent electrodes in optoelectronic devices that


Fabrication of ultrathin nanocarbon films using chemical vapordeposition

usually require high conductivity and low opacity [77]. However,these films can be used as lightweight chemically inert and me-chanically strong ultrathin conductive coatings for e.g. microwaveshielding [2–4].

4.2 CVD GRAPHENE ON A COPPER CATALYST

By using metal catalysts, the crystallinity of the CVD graphitic car-bon film can be increased from nanometer to millimeter scale [26].Among metals suitable for CVD of graphene are Iridium [78], plat-inum [79], ruthenium [80], palladium [81], nickel [82, 83] and cop-per [84]. The choice of a catalyst metal is crucial for the propertiesof the obtained SLG or FLG film.

The first experimental demonstrations for graphene CVD weremade using a nickel catalyst substrate [82, 83]. However, high car-bon solubility in nickel makes it difficult to control graphene layerthickness [85]. This difficulty can be removed when copper is usedas a catalyst in the CVD process. A low carbon solubility, conve-nient crystallinity and low cost have made copper the most attrac-tive material for graphene synthesis. Recently a thin copper foil hasbecome a standard catalyst for graphene CVD [85, 86].

The quality of graphene sheet that grows on the surface of acopper catalyst depends on the crystallinity of a copper substrate. Ithas been observed that the most and least convenient crystalline ori-entation for graphene synthesis on a copper substrate are Cu<111>,and Cu<100>, respectively [86]. In the <111> the copper atomsare packed densely forming a structure that gives a hexagonal sur-face structure preferable for the graphene synthesis while in <100>crystal plane, the copper atoms at the surface are arranged in asquare lattice, which is not well suited for the graphene crystalgrowth [86].

The synthesis of graphene on copper resembles the synthesis ofPyC but without homogenous process phase. In high, above 800 Ctemperature the copper catalysis becomes active and starts to assistmethane molecule decomposition efficiently enough for graphene



ity of the methane in comparison to other hydrocarbon gasses, theCVD process temperatures are typically more than 900 C [69, 72].The main reaction sequence from methane to benzene-like aromaticstructures involves C2- and C4-species and can be as following:

CH4 → C2Hx → C4Hy → C6Hz, (4.1)

where x, y and z are the number of hydrogen atoms bound to car-bon atoms [72]. The effect of the unstable C3 and C5-species issupposed to play only minor role [72].

The evolving from C1 to C2- and C6-species are believed to behomogenous reactions and for further develop the heterogenousreactions are expected to take place on the substrate surface [69].While growing in size the ring-like carbon structures will settle ona substrate surface being attached to active sites. After attachmentthe carbon rings start evolving to more complex polycyclic hydro-carbons. Although the size of an individual polycyclic hydrocar-bons remains below 10 nm the small pyrocarbon clusters intertwinetogether forming a continuous carbon film [73].

Since the graphitic film grown on a substrate surface consist ofcrosslinked nanosize graphene clusters, the film can be expectedto be highly amorphous. However, surface of this amorphous filmis observed to be almost atomically smooth. Typically the averagesurface roughness of a prepared film (see paper II) on a silica sub-strate is in order of 1 nm. Although such a smoothness of the PyCfilms is still not understood, one may expect that it may be causedby the preferred orientation of the nanosize graphene clusters arealong the surface plane. Therefore, despite the PyC film is highlyamorphous, is very uniform and can be employed for nanoscalecoating material [75].

Because of high amorphicity of the PyC film the electron scatter-ing in the material is stronger than in graphene or graphite that isseen as higher sheet resistance [25, 73]. The strong electron scatter-ing in PyC suppresses the conductivity making this material hardlysuitable for transparent electrodes in optoelectronic devices that



usually require high conductivity and low opacity [77]. However,these films can be used as lightweight chemically inert and me-chanically strong ultrathin conductive coatings for e.g. microwaveshielding [2–4].

4.2 CVD GRAPHENE ON A COPPER CATALYST

By using metal catalysts, the crystallinity of the CVD graphitic car-bon film can be increased from nanometer to millimeter scale [26].Among metals suitable for CVD of graphene are Iridium [78], plat-inum [79], ruthenium [80], palladium [81], nickel [82, 83] and cop-per [84]. The choice of a catalyst metal is crucial for the propertiesof the obtained SLG or FLG film.

The first experimental demonstrations for graphene CVD weremade using a nickel catalyst substrate [82, 83]. However, high car-bon solubility in nickel makes it difficult to control graphene layerthickness [85]. This difficulty can be removed when copper is usedas a catalyst in the CVD process. A low carbon solubility, conve-nient crystallinity and low cost have made copper the most attrac-tive material for graphene synthesis. Recently a thin copper foil hasbecome a standard catalyst for graphene CVD [85, 86].

The quality of graphene sheet that grows on the surface of acopper catalyst depends on the crystallinity of a copper substrate. Ithas been observed that the most and least convenient crystalline ori-entation for graphene synthesis on a copper substrate are Cu<111>,and Cu<100>, respectively [86]. In the <111> the copper atomsare packed densely forming a structure that gives a hexagonal sur-face structure preferable for the graphene synthesis while in <100>crystal plane, the copper atoms at the surface are arranged in asquare lattice, which is not well suited for the graphene crystalgrowth [86].

The synthesis of graphene on copper resembles the synthesis ofPyC but without homogenous process phase. In high, above 800 Ctemperature the copper catalysis becomes active and starts to assistmethane molecule decomposition efficiently enough for graphene



formation [87, 88]. On the surface of copper the carbon will attachto the copper while hydrogen will leave the surface [87]. Poly-cyclic carbon structures are formed at active sites on the coppersurface and start growing in size [89]. Eventually, when the adja-cent graphene cluster become big enough to reach one another, agraphene grain boundary is formed [26].

The copper foil that conventionally used for CVD graphene syn-thesis is polycrystalline and this affects to the quality of graphenesynthesized on its surface. Since the graphene flakes that are grownon copper grains with different crystal planes are connected toeach other with grain boundaries, the graphene film is continu-ous. However, the orientation of the graphene flakes is slightlydifferent, which makes graphene polycrystalline [90]. The size ofthe graphene crystallities grown on the copper surface is usually ofthe order of a few tens of micrometers [90], i.e. it is bigger thanthe light wavelength in visual and near IR region. Therefore CVDgraphene can be employed for investigation of various optical andoptoelectronic effects, while the fabrication of graphene on coppercatalysis CVD has opened an important route for inexpensive andhigh quality graphene synthesis.

4.2.1 Multilayered graphene on copper

A low carbon solubility in a copper substrate should lead into aformation of SLG film [85]. However, formation of FLG is alsopossible (see paper I and [91]) when so called template graphenegrowth takes place [92].

In the template graphene growth, the first layer of graphene isgrown catalytically on a copper surface at temperature of about1000 C. This high temperature is, however, high enough also forspontaneous methane decomposition [69]. Therefore, the hydrocar-bon decomposition in the CVD chamber takes place catalytically onsubstrate surface and also in the gas phase.

The decomposed hydrocarbon molecules start to form a newlayer of carbon on top of the first graphene layer. Although the



second graphene layer grows on the top of the graphene, usu-ally possesses lower crystallinity in comparison with the first layer[92, 93]. Increasing of the deposition time may result in formationof even more layers, however, the size of the graphene crystalli-ties should decrease from layer to layer, i.e. the crystallinity of thetopmost layer maybe as low as that of the amorphous pyrolytic car-bon [69, 92].

The amorphousness the template layers on a single graphenefilm are often considered as unwanted phenomena. In order toprevent formation of template layers, the SLG samples are preparedin low concentration of carbon precursor gas [26]. Also in orderto obtain single layer graphene on copper catalyst the unwantedcarbon layers can be etched by hydrogen after the graphitizationsequence [93].

4.2.2 Graphene transfer from Cu catalyst to a dielectric

Probably the most significant disadvantage of the CVD grapheneis the presence of the copper substrate. This is because in order tofully exploit the optical and electrical properties of graphene oneneed to transfer a nanometrically thin carbon film from copper to adielectric or semiconducting surface. There has been suggested anddemonstrated several different techniques to transfer the graphenefilm from Cu substrate to a dielectric [94]. In the paper I, we em-ployed the most conventional wet etching transfer technique de-scribed below [95].

At first, a supporting poly(methyl methacrylate) (PMMA) layeris deposited on top of graphene in order to prevent the grapheneto crack during the transfer. When graphene is coated with a pro-tective polymer layer, the underlying Cu substrate is removed bychemical etchant. Next the sample is rinsed in water and depositedon a dielectric substrate. A small droplet of liquid PMMA is de-posited on top of the PMMA/graphene/dielectric sample to makethe old PMMA layer relaxed. At the end the PMMA layer is re-moved by acetone or another solvent. [95]



formation [87, 88]. On the surface of copper the carbon will attachto the copper while hydrogen will leave the surface [87]. Poly-cyclic carbon structures are formed at active sites on the coppersurface and start growing in size [89]. Eventually, when the adja-cent graphene cluster become big enough to reach one another, agraphene grain boundary is formed [26].

The copper foil that conventionally used for CVD graphene syn-thesis is polycrystalline and this affects to the quality of graphenesynthesized on its surface. Since the graphene flakes that are grownon copper grains with different crystal planes are connected toeach other with grain boundaries, the graphene film is continu-ous. However, the orientation of the graphene flakes is slightlydifferent, which makes graphene polycrystalline [90]. The size ofthe graphene crystallities grown on the copper surface is usually ofthe order of a few tens of micrometers [90], i.e. it is bigger thanthe light wavelength in visual and near IR region. Therefore CVDgraphene can be employed for investigation of various optical andoptoelectronic effects, while the fabrication of graphene on coppercatalysis CVD has opened an important route for inexpensive andhigh quality graphene synthesis.

4.2.1 Multilayered graphene on copper

A low carbon solubility in a copper substrate should lead into aformation of SLG film [85]. However, formation of FLG is alsopossible (see paper I and [91]) when so called template graphenegrowth takes place [92].

In the template graphene growth, the first layer of graphene isgrown catalytically on a copper surface at temperature of about1000 C. This high temperature is, however, high enough also forspontaneous methane decomposition [69]. Therefore, the hydrocar-bon decomposition in the CVD chamber takes place catalytically onsubstrate surface and also in the gas phase.

The decomposed hydrocarbon molecules start to form a newlayer of carbon on top of the first graphene layer. Although the



second graphene layer grows on the top of the graphene, usu-ally possesses lower crystallinity in comparison with the first layer[92, 93]. Increasing of the deposition time may result in formationof even more layers, however, the size of the graphene crystalli-ties should decrease from layer to layer, i.e. the crystallinity of thetopmost layer maybe as low as that of the amorphous pyrolytic car-bon [69, 92].

The amorphousness the template layers on a single graphenefilm are often considered as unwanted phenomena. In order toprevent formation of template layers, the SLG samples are preparedin low concentration of carbon precursor gas [26]. Also in orderto obtain single layer graphene on copper catalyst the unwantedcarbon layers can be etched by hydrogen after the graphitizationsequence [93].

4.2.2 Graphene transfer from Cu catalyst to a dielectric

Probably the most significant disadvantage of the CVD grapheneis the presence of the copper substrate. This is because in order tofully exploit the optical and electrical properties of graphene oneneed to transfer a nanometrically thin carbon film from copper to adielectric or semiconducting surface. There has been suggested anddemonstrated several different techniques to transfer the graphenefilm from Cu substrate to a dielectric [94]. In the paper I, we em-ployed the most conventional wet etching transfer technique de-scribed below [95].

At first, a supporting poly(methyl methacrylate) (PMMA) layeris deposited on top of graphene in order to prevent the grapheneto crack during the transfer. When graphene is coated with a pro-tective polymer layer, the underlying Cu substrate is removed bychemical etchant. Next the sample is rinsed in water and depositedon a dielectric substrate. A small droplet of liquid PMMA is de-posited on top of the PMMA/graphene/dielectric sample to makethe old PMMA layer relaxed. At the end the PMMA layer is re-moved by acetone or another solvent. [95]



Figure 4.1: A graphene sheet (Gr) can be transferred from a copper substrate (a) to anarbitrary surface by depositing a polymer layer (b) on top of Gr-Cu sample. Next (c)the Cu is removed by wet etching and (d) the polymer+Gr structure is placed on finalsubstrate. (e) At the end the polymer layer is dissolved chemically.

It is worth noting that although graphene is mechanically verystrong, transferring of an atomically thin graphene film can be eas-ily damaged in the process of transfer. Furthermore, involvementof a number of chemical reagants in the process may affect to mor-phological and electronic properties of a graphene sample (e.g. toinduce structural defects and change doping level etc.). Thereforefinding techniques to deposit graphene directly on a dielectric is aneagerly awaited breakthrough necessary for fabrication graphenebased electronics and optoelectronics devices [26].

4.3 DIRECT GRAPHENE DEPOSITION ON DIELECTRIC SUB-STRATE

Although transferring techniques to move graphene from copperto dielectric substrate have been evolved recently, incorporationof graphene elements in to practical devices require synthesis ofgraphene directly on the dielectric or semiconductor substrate [26].This issue has attracted a widespread attention of nanocarbon com-munity and so far several promising methods have been suggested[96–101]. In this work the direct deposition of graphene film wasachieved by the assistance of a nanometrically thin copper film pre-deposited on a silica substrate.



Figure 4.2: For direct deposition a few hundred thick layer of copper is deposited on asilica substrate (a). (b) At high temperature the Cu film will partially melt forming grainboundaries. (c) When catalysis begins the carbon atoms migrate through Cu film usinggrain boundaries. (d) After the CVD process a few layered graphene film is formed oncopper-air and copper-silica interfaces. (e) By removing the topmost graphene layer andCu remains, one will obtain a few layered graphene on silica substrate.

The receding of the ultrathin copper film in the CVD chamberis governed by the surface melting that opens different routes forgraphene deposition [96, 98]. The first experimental demonstrationthat melting of thin copper film can be utilized for direct deposi-tion of graphene have been reported in 2010 by Ismach et al. [96].They suggested that graphene will grow on top of a thin copperfilm pre-deposited on a silica substrate similarly as it grows on aconventional copper foil. However, when the copper melts it startsto recede leaving graphene behind. By increasing the depositiontime with several hours it was possible to evaporate all the copperaway and obtain a graphene on a dielectric substrate.

Su et al. in 2011 demonstrated that the graphene grow on theCu-air interface as expected and also on the Cu-substrate interface[98]. This was explained by carbon migration through a copperfilm where the carbon atoms utilize the copper grain boundariesin order to penetrate through a thin copper film. After removingcopper remains by wet etching a few layered graphene is revealedbeneath the thin Cu film and one obtain again a graphene sampleon a dielectric substrate.

The fundamental reason for the formation of graphene to the



Figure 4.1: A graphene sheet (Gr) can be transferred from a copper substrate (a) to anarbitrary surface by depositing a polymer layer (b) on top of Gr-Cu sample. Next (c)the Cu is removed by wet etching and (d) the polymer+Gr structure is placed on finalsubstrate. (e) At the end the polymer layer is dissolved chemically.

It is worth noting that although graphene is mechanically verystrong, transferring of an atomically thin graphene film can be eas-ily damaged in the process of transfer. Furthermore, involvementof a number of chemical reagants in the process may affect to mor-phological and electronic properties of a graphene sample (e.g. toinduce structural defects and change doping level etc.). Thereforefinding techniques to deposit graphene directly on a dielectric is aneagerly awaited breakthrough necessary for fabrication graphenebased electronics and optoelectronics devices [26].

4.3 DIRECT GRAPHENE DEPOSITION ON DIELECTRIC SUB-STRATE

Although transferring techniques to move graphene from copperto dielectric substrate have been evolved recently, incorporationof graphene elements in to practical devices require synthesis ofgraphene directly on the dielectric or semiconductor substrate [26].This issue has attracted a widespread attention of nanocarbon com-munity and so far several promising methods have been suggested[96–101]. In this work the direct deposition of graphene film wasachieved by the assistance of a nanometrically thin copper film pre-deposited on a silica substrate.



Figure 4.2: For direct deposition a few hundred thick layer of copper is deposited on asilica substrate (a). (b) At high temperature the Cu film will partially melt forming grainboundaries. (c) When catalysis begins the carbon atoms migrate through Cu film usinggrain boundaries. (d) After the CVD process a few layered graphene film is formed oncopper-air and copper-silica interfaces. (e) By removing the topmost graphene layer andCu remains, one will obtain a few layered graphene on silica substrate.

The receding of the ultrathin copper film in the CVD chamberis governed by the surface melting that opens different routes forgraphene deposition [96, 98]. The first experimental demonstrationthat melting of thin copper film can be utilized for direct deposi-tion of graphene have been reported in 2010 by Ismach et al. [96].They suggested that graphene will grow on top of a thin copperfilm pre-deposited on a silica substrate similarly as it grows on aconventional copper foil. However, when the copper melts it startsto recede leaving graphene behind. By increasing the depositiontime with several hours it was possible to evaporate all the copperaway and obtain a graphene on a dielectric substrate.

Su et al. in 2011 demonstrated that the graphene grow on theCu-air interface as expected and also on the Cu-substrate interface[98]. This was explained by carbon migration through a copperfilm where the carbon atoms utilize the copper grain boundariesin order to penetrate through a thin copper film. After removingcopper remains by wet etching a few layered graphene is revealedbeneath the thin Cu film and one obtain again a graphene sampleon a dielectric substrate.

The fundamental reason for the formation of graphene to the



copper-silica interface originates from the physical properties ofthin copper film. More precisely the thin copper film will partiallymelt in the process temperature below the bulk melting temper-ature [102–104]. This thermal instability of low dimensional sys-tems is called surface melting and is explained by the surface atomvibrations that are much larger in magnitude compared to thoseatoms vibrating inside the bulk material. More importantly thesurface melting, sometimes associated with the Tammann temper-ature, is sufficient to make the surface atoms migrate freely [105].Free atom mobility at the surface leads to liquid-like behavior ofthe surface [106]. During the CVD process, the temperature near900 C is below the bulk copper melting temperature of 1084 C.However, it is high enough for surface melting of the copper layerwith thickness of a few hundreds nanometers.

It is expected that only the very surface of the copper film willmelt with a depth of few nanometers [102–104]. However, the sincethe Cu film is prepared by thermal evaporation, it composed ofdensely parked Cu nanoparticles, i.e. the film is intrinsically in-homogeneous. In particular there are points where disorder isstronger than the rest of the film (e.g. where the several nanopar-ticles are connected). Since all the nanoparticles have their ownsurface that melts separately the Cu film melting will start at thesepoints. The melted copper will recede towards still solid areasforming characteristic finger-like pattern on the substrate.

In the vicinity of 900 C the first graphene structures start toform to the copper-substrate interface [98]. Because of the low car-bon solubility of carbon in copper the carbon atoms cannot pene-trate the copper film directly but because of the melted surface thecarbon atoms migrate on liquefied copper [85,98]. On areas where agrain boundary is formed in the Cu film carbon atoms can migratethrough the film using the disordered and melted boundary [98].After successfully penetrating the copper film the carbon start toform graphene film to the copper-silica interface.

By described technique one can synthesize a few layered graphe-ne directly to a dielectric substrate after removing copper remains



[98]. Furthermore, since the copper film is in liquefied-like stateduring the process, it gives possibility to control how the copperwill settle on the dielectric substrate by changing the surface struc-ture of the substrate. In the chapter 6 we will discuss further howthe surface structure of the dielectric substrate will affect to the cop-per receding and how to make a few layered graphene samples byself-assembling technique (papers III–V).

4.4 EXPERIMENTAL METHODS FOR A CHEMICAL VAPORDEPOSITION OF AN ULTRATHIN CARBON FILMS

In this section, in order to outline the difference between synthesisof the PyC and graphene we will very briefly describe characteris-tics of the employed CVD setup.

The conventional hot wall CVD setup used in all of the exper-iments in papers I-V consist of computer controlled gas mass flowcontrollers and inlets for methane, hydrogen and nitrogen flows,tube oven (Carbolite CTF 12/75/700) as a heating element and vac-uum chamber with pressure meter and vacuum pump. Since theheating element cools down very slowly ( 1-5 C/min) the setupworks best with substrates that are not critical on cooling down rate.Furthermore the setup works in static atmosphere, i.e. there is nogas flow during the processes through the CVD chamber. Betweenthe cycles, when the gases are changed the chamber is vacumizedbefore injecting new gasses. This greatly reduces the gas consump-tion during the process.

For synthesizing FLG film on a copper (see the paper I) a cleancopper foil (purity 99,8 %, thickness 25 µm) is inserted into the vac-uum chamber and heated up to 1000 C temperature in hydrogenatmosphere. Usually the copper foil is annealed 30 min to 60 minin 1000 C in order to remove remains of copper oxide on coppersurface. During the annealing the copper will also re-crystallizeforming islands of monocrystalline copper with lateral diameterabout few tens of micrometer. After the annealing the gaseous at-mosphere was changed from hydrogen to hydrogen-methane gas



copper-silica interface originates from the physical properties ofthin copper film. More precisely the thin copper film will partiallymelt in the process temperature below the bulk melting temper-ature [102–104]. This thermal instability of low dimensional sys-tems is called surface melting and is explained by the surface atomvibrations that are much larger in magnitude compared to thoseatoms vibrating inside the bulk material. More importantly thesurface melting, sometimes associated with the Tammann temper-ature, is sufficient to make the surface atoms migrate freely [105].Free atom mobility at the surface leads to liquid-like behavior ofthe surface [106]. During the CVD process, the temperature near900 C is below the bulk copper melting temperature of 1084 C.However, it is high enough for surface melting of the copper layerwith thickness of a few hundreds nanometers.

It is expected that only the very surface of the copper film willmelt with a depth of few nanometers [102–104]. However, the sincethe Cu film is prepared by thermal evaporation, it composed ofdensely parked Cu nanoparticles, i.e. the film is intrinsically in-homogeneous. In particular there are points where disorder isstronger than the rest of the film (e.g. where the several nanopar-ticles are connected). Since all the nanoparticles have their ownsurface that melts separately the Cu film melting will start at thesepoints. The melted copper will recede towards still solid areasforming characteristic finger-like pattern on the substrate.

In the vicinity of 900 C the first graphene structures start toform to the copper-substrate interface [98]. Because of the low car-bon solubility of carbon in copper the carbon atoms cannot pene-trate the copper film directly but because of the melted surface thecarbon atoms migrate on liquefied copper [85,98]. On areas where agrain boundary is formed in the Cu film carbon atoms can migratethrough the film using the disordered and melted boundary [98].After successfully penetrating the copper film the carbon start toform graphene film to the copper-silica interface.

By described technique one can synthesize a few layered graphe-ne directly to a dielectric substrate after removing copper remains



[98]. Furthermore, since the copper film is in liquefied-like stateduring the process, it gives possibility to control how the copperwill settle on the dielectric substrate by changing the surface struc-ture of the substrate. In the chapter 6 we will discuss further howthe surface structure of the dielectric substrate will affect to the cop-per receding and how to make a few layered graphene samples byself-assembling technique (papers III–V).

4.4 EXPERIMENTAL METHODS FOR A CHEMICAL VAPORDEPOSITION OF AN ULTRATHIN CARBON FILMS

In this section, in order to outline the difference between synthesisof the PyC and graphene we will very briefly describe characteris-tics of the employed CVD setup.

The conventional hot wall CVD setup used in all of the exper-iments in papers I-V consist of computer controlled gas mass flowcontrollers and inlets for methane, hydrogen and nitrogen flows,tube oven (Carbolite CTF 12/75/700) as a heating element and vac-uum chamber with pressure meter and vacuum pump. Since theheating element cools down very slowly ( 1-5 C/min) the setupworks best with substrates that are not critical on cooling down rate.Furthermore the setup works in static atmosphere, i.e. there is nogas flow during the processes through the CVD chamber. Betweenthe cycles, when the gases are changed the chamber is vacumizedbefore injecting new gasses. This greatly reduces the gas consump-tion during the process.

For synthesizing FLG film on a copper (see the paper I) a cleancopper foil (purity 99,8 %, thickness 25 µm) is inserted into the vac-uum chamber and heated up to 1000 C temperature in hydrogenatmosphere. Usually the copper foil is annealed 30 min to 60 minin 1000 C in order to remove remains of copper oxide on coppersurface. During the annealing the copper will also re-crystallizeforming islands of monocrystalline copper with lateral diameterabout few tens of micrometer. After the annealing the gaseous at-mosphere was changed from hydrogen to hydrogen-methane gas



Figure 4.3: (a) A few layered graphene film on a 50 mm silica substrate. (b) By scanningelectron microscopy (SEM) one can notice small holes (circled) and vertical lines thatoriginate from copper substrate roughness. (c) A PyC synthesized on a silica substrate.(d) The PyC film is uniform throughout the substrate and almost atomically smooth butcan be scratched with scalpel.

mix (15 ml:15 ml with pressure of about 10 mBar). The graphitiza-tion lasted typically from 5 min to 20 min. It is worth noting thatthere has been no significant differences observed between short orlong graphitization time. In the paper I the hydrogen etching wasnot used and the received film is a multilayered graphene. For agraphene monolayer synthesis the hydrogen etching could be doneafter the graphitization period. This is done by removing carbonprecursor and injecting hydrogen to the CVD chamber for a fewminutes.

Transferring graphene from a copper foil to a dielectric sub-strate was done by using the methods described in more detailelsewhere [95]. A thin layer of PMMA was deposited on top ofgraphene/copper sample and the copper was then etched in ferricchloride (FeCl3). After etching the PMMA/graphene sample wasrinsed in the water, deposited on a silica substrate and dried with



mild 50 temperature. A new layer of PMMA was deposited ontop of the PMMA/graphene/silica in order to relax the old PMMAlayer. Finally the PMMA was removed from the sample surface byacetone.

For PyC film deposition the most significant difference in theprocess parameters, compared to the graphene synthesis, is thehigher amount of methane. In the paper II we reported the PyCfilms synthesis on a 0.5 mm thick silica substrate. In the pro-cess there was also some small changes done in comparison to thegraphene process. First the chamber was heated up to 700 C tem-perature in hydrogen atmosphere. At 700 C the hydrogen atmo-sphere is exchanged to hydrogen-methane gas mixture (see tablepaper II). After the gas exchange the chamber temperature is risenup to 1100 C for 5 min (with rate of 10 C/min) and cooled backdown to 700 C in 80 minutes. Next the chamber is cooled downto room temperature in hydrogen atmosphere. Since both sides ofthe sample are covered with PyC film, the back side of the sampleis cleansed with harsh oxygen plasma.

The thickness of the film can be controlled by the pressure andamount of methane concentration (parameters are found in the pa-per II). It is noteworthy that since the setup works in static atmo-sphere, therefore increasing the amount of methane concentrationwill affect to the pressure also. Second interesting point of the pro-cess is the dynamic process temperature ranges from 700 C to 1100C and back down to 700 C. The first decompositions of methaneare expected to happen close 700 C or 800 C. However, sincemethane is very stable molecule the decomposition rate of methaneis improved by increasing the temperature to 1100 C. Furthermorearomatic carbon structures are expected to decompose above 800 Cand aliphically (chain like) bonded carbon above 600 C. Thereforethe process was ended at 700 C.

For the direct graphene synthesis on a silica substrate the sam-ple was first coated with few hundred nanometers (100 nm - 300nm) thick copper film. The sample was loaded into the CVD cham-ber and heated to 700 C in hydrogen atmosphere. At 700 C the



Figure 4.3: (a) A few layered graphene film on a 50 mm silica substrate. (b) By scanningelectron microscopy (SEM) one can notice small holes (circled) and vertical lines thatoriginate from copper substrate roughness. (c) A PyC synthesized on a silica substrate.(d) The PyC film is uniform throughout the substrate and almost atomically smooth butcan be scratched with scalpel.

mix (15 ml:15 ml with pressure of about 10 mBar). The graphitiza-tion lasted typically from 5 min to 20 min. It is worth noting thatthere has been no significant differences observed between short orlong graphitization time. In the paper I the hydrogen etching wasnot used and the received film is a multilayered graphene. For agraphene monolayer synthesis the hydrogen etching could be doneafter the graphitization period. This is done by removing carbonprecursor and injecting hydrogen to the CVD chamber for a fewminutes.

Transferring graphene from a copper foil to a dielectric sub-strate was done by using the methods described in more detailelsewhere [95]. A thin layer of PMMA was deposited on top ofgraphene/copper sample and the copper was then etched in ferricchloride (FeCl3). After etching the PMMA/graphene sample wasrinsed in the water, deposited on a silica substrate and dried with



mild 50 temperature. A new layer of PMMA was deposited ontop of the PMMA/graphene/silica in order to relax the old PMMAlayer. Finally the PMMA was removed from the sample surface byacetone.

For PyC film deposition the most significant difference in theprocess parameters, compared to the graphene synthesis, is thehigher amount of methane. In the paper II we reported the PyCfilms synthesis on a 0.5 mm thick silica substrate. In the pro-cess there was also some small changes done in comparison to thegraphene process. First the chamber was heated up to 700 C tem-perature in hydrogen atmosphere. At 700 C the hydrogen atmo-sphere is exchanged to hydrogen-methane gas mixture (see tablepaper II). After the gas exchange the chamber temperature is risenup to 1100 C for 5 min (with rate of 10 C/min) and cooled backdown to 700 C in 80 minutes. Next the chamber is cooled downto room temperature in hydrogen atmosphere. Since both sides ofthe sample are covered with PyC film, the back side of the sampleis cleansed with harsh oxygen plasma.

The thickness of the film can be controlled by the pressure andamount of methane concentration (parameters are found in the pa-per II). It is noteworthy that since the setup works in static atmo-sphere, therefore increasing the amount of methane concentrationwill affect to the pressure also. Second interesting point of the pro-cess is the dynamic process temperature ranges from 700 C to 1100C and back down to 700 C. The first decompositions of methaneare expected to happen close 700 C or 800 C. However, sincemethane is very stable molecule the decomposition rate of methaneis improved by increasing the temperature to 1100 C. Furthermorearomatic carbon structures are expected to decompose above 800 Cand aliphically (chain like) bonded carbon above 600 C. Thereforethe process was ended at 700 C.

For the direct graphene synthesis on a silica substrate the sam-ple was first coated with few hundred nanometers (100 nm - 300nm) thick copper film. The sample was loaded into the CVD cham-ber and heated to 700 C in hydrogen atmosphere. At 700 C the



chamber was vacumized and injected with hydrogen-methane gasmix (15 ml: 15 ml with pressure of about 10 mBar). Next thechamber temperature was risen to 950 C for 5 min (with rate of10 C/min) and then cooled down to 700 C in about 60 minutes.The rest of the cooling was done in hydrogen atmosphere. As it canbe observed the process is very similar to PyC process. However,the temperature dynamics was different because we wanted to en-sure carbon availability during all the time when the carbon wasliquefied (the surface melting of thin copper film that takes place atabout 700 C [102–104]).

In a summary, graphene deposition on a copper foil, pyrolyticcarbon deposition on a silica substrate and direct graphene depo-sition on a silica substrate are can be performed using very similarCVD process. However, even minor changes in the process parame-ters can drastically affect the properties of the synthesized material.


5 Optical characterization ofcarbon films

Optical characterization is a nondestructive technique that can beused to probe the properties of ultrathin films. In this chapter wepresent results on the characterization of the ultrathin carbon filmsby Raman spectroscopy and spectrophotometry.

5.1 RAMAN SPECTROSCOPY OF NANOCARBON FILMS

All the Raman measurements described below were performed byusing the Renishaw inVia Raman Microscope at the excitation wave-length of 514 nm. Therefore the obtained results can be used forcomparative study of the fabricated films. Since at the high exci-tation intensity may cause some unwanted, heat induced intrinsiceffects in the ultrathin carbon films [46, 107, 108], the power levelof the excitation laser was kept low in order to avoid heat inducedeffects in the nanocarbon films.

5.1.1 Raman spectra of a graphene film transferred to a dielectricsubstrate

In the paper I the graphene sample was grown on a copper foil andit was transferred to a silica substrate as described in the chapter4.2.2. After the graphene film was transferred on the silica substratethe Raman spectrum was measured. A typical Raman spectrumof the graphene sample that was grown on a copper foil withouthydrogen etching is presented in the Fig. 5.1.

In the Raman spectrum reported in the paper I the ratio of the Dand G peak intensities indicates a graphene material with defects.However, since there are no traces of amorphous carbon (widenedpeak at 1500 cm−1), the Eq. 3.1 can be used to determine the size of



chamber was vacumized and injected with hydrogen-methane gasmix (15 ml: 15 ml with pressure of about 10 mBar). Next thechamber temperature was risen to 950 C for 5 min (with rate of10 C/min) and then cooled down to 700 C in about 60 minutes.The rest of the cooling was done in hydrogen atmosphere. As it canbe observed the process is very similar to PyC process. However,the temperature dynamics was different because we wanted to en-sure carbon availability during all the time when the carbon wasliquefied (the surface melting of thin copper film that takes place atabout 700 C [102–104]).

In a summary, graphene deposition on a copper foil, pyrolyticcarbon deposition on a silica substrate and direct graphene depo-sition on a silica substrate are can be performed using very similarCVD process. However, even minor changes in the process parame-ters can drastically affect the properties of the synthesized material.


5 Optical characterization ofcarbon films

Optical characterization is a nondestructive technique that can beused to probe the properties of ultrathin films. In this chapter wepresent results on the characterization of the ultrathin carbon filmsby Raman spectroscopy and spectrophotometry.

5.1 RAMAN SPECTROSCOPY OF NANOCARBON FILMS

All the Raman measurements described below were performed byusing the Renishaw inVia Raman Microscope at the excitation wave-length of 514 nm. Therefore the obtained results can be used forcomparative study of the fabricated films. Since at the high exci-tation intensity may cause some unwanted, heat induced intrinsiceffects in the ultrathin carbon films [46, 107, 108], the power levelof the excitation laser was kept low in order to avoid heat inducedeffects in the nanocarbon films.

5.1.1 Raman spectra of a graphene film transferred to a dielectricsubstrate

In the paper I the graphene sample was grown on a copper foil andit was transferred to a silica substrate as described in the chapter4.2.2. After the graphene film was transferred on the silica substratethe Raman spectrum was measured. A typical Raman spectrumof the graphene sample that was grown on a copper foil withouthydrogen etching is presented in the Fig. 5.1.

In the Raman spectrum reported in the paper I the ratio of the Dand G peak intensities indicates a graphene material with defects.However, since there are no traces of amorphous carbon (widenedpeak at 1500 cm−1), the Eq. 3.1 can be used to determine the size of



Figure 5.1: A typical Raman spectrum of a few layered graphene sample synthesized ona dielectric substrate. The disordering is seen as D and D′ peaks that are rather smallcompared to G. Furthermore, a narrow G peak indicate a rather well crystalline few layeredgraphene. The 2D peak is located at 2703 cm−1 and with FWHM2D of 50 cm−1 indicatinga few layered graphene material.

La. Typically the ratio of I(D)/I(G) in this sample was in the orderof 10 % - 20 %. According Eq. 3.1 the La was in range from 20 nmto 40 nm (C = 4.4) or 80 nm to 160 nm C = 16.6. From the otherhand the measured FWHMG ≈ 20 cm−1 corresponding roughly toLa ∼100 nm. The G-mode was also found to be shifted from 1582cm−1 to 1587 cm−1 which is a signature of the disorder [56]. Thissame disordering can be noticed from the small right side shoulderof the G peak (see Fig. 5.1). These disorders that are shown in thisspectrum are most probably caused by the additional carbon layersthat have grown on top of the first layer by template mechanism[92].

Furthermore the position of the 2D peak at 2703 cm−1 andFWHM2D ≈ 50 cm−1 in the paper I clearly indicates that the ma-terial was FLG. If the spectrum in the Fig. 5.1 is compared to aspectrum of FLG grown on Cu (Ref. [92]) the resemblance is obvi-ous.


Optical characterization of carbon films

Figure 5.2: A Raman spectrum measured from a few layered graphene sample grown bythe direct deposition technique. The disordering is revealed by D peak and also by theslight antisymmetry of the G peak. The 2D peak is located at 2710 cm−1 with FWHM2Dof 60 cm−1 which indicates that the material is a few layered graphene.

5.1.2 Raman spectra of a graphene synthesized directly on a di-electric substrate

Controlling of the number of graphene layers directly deposited ona dielectric substrate is more difficult task compared to the control-ling of the number of graphene layers grown on a copper. This isbecause copper protects the synthesis of the single graphene layer.So far there has not yet been presented a method to suppress for-mation of several graphene layers or remove disordering for directdeposition on dielectric.

In Fig. 5.2 there is presented Raman spectra of FLG (paper III)grown by direct deposition after the copper remains are removed.One can observe that there exist the disorder related D peak ataround 1350 cm−1 and no traces of amorphous carbon peak at 1500cm−1. Therefore one can conclude that the carbon material grownbeneath the copper film is solely sp2 bonded.

Interestingly the relation of the I(D)/I(G) is typically in thesame order compared to the graphene grown on a copper foil in



Figure 5.1: A typical Raman spectrum of a few layered graphene sample synthesized ona dielectric substrate. The disordering is seen as D and D′ peaks that are rather smallcompared to G. Furthermore, a narrow G peak indicate a rather well crystalline few layeredgraphene. The 2D peak is located at 2703 cm−1 and with FWHM2D of 50 cm−1 indicatinga few layered graphene material.

La. Typically the ratio of I(D)/I(G) in this sample was in the orderof 10 % - 20 %. According Eq. 3.1 the La was in range from 20 nmto 40 nm (C = 4.4) or 80 nm to 160 nm C = 16.6. From the otherhand the measured FWHMG ≈ 20 cm−1 corresponding roughly toLa ∼100 nm. The G-mode was also found to be shifted from 1582cm−1 to 1587 cm−1 which is a signature of the disorder [56]. Thissame disordering can be noticed from the small right side shoulderof the G peak (see Fig. 5.1). These disorders that are shown in thisspectrum are most probably caused by the additional carbon layersthat have grown on top of the first layer by template mechanism[92].

Furthermore the position of the 2D peak at 2703 cm−1 andFWHM2D ≈ 50 cm−1 in the paper I clearly indicates that the ma-terial was FLG. If the spectrum in the Fig. 5.1 is compared to aspectrum of FLG grown on Cu (Ref. [92]) the resemblance is obvi-ous.



Figure 5.2: A Raman spectrum measured from a few layered graphene sample grown bythe direct deposition technique. The disordering is revealed by D peak and also by theslight antisymmetry of the G peak. The 2D peak is located at 2710 cm−1 with FWHM2Dof 60 cm−1 which indicates that the material is a few layered graphene.

5.1.2 Raman spectra of a graphene synthesized directly on a di-electric substrate

Controlling of the number of graphene layers directly deposited ona dielectric substrate is more difficult task compared to the control-ling of the number of graphene layers grown on a copper. This isbecause copper protects the synthesis of the single graphene layer.So far there has not yet been presented a method to suppress for-mation of several graphene layers or remove disordering for directdeposition on dielectric.

In Fig. 5.2 there is presented Raman spectra of FLG (paper III)grown by direct deposition after the copper remains are removed.One can observe that there exist the disorder related D peak ataround 1350 cm−1 and no traces of amorphous carbon peak at 1500cm−1. Therefore one can conclude that the carbon material grownbeneath the copper film is solely sp2 bonded.

Interestingly the relation of the I(D)/I(G) is typically in thesame order compared to the graphene grown on a copper foil in



the paper I. The D peak in the graphene grown by direct deposi-tion technique on a silica substrate may originate from the rathergrainy structure that can be seen in the SEM image (Chapter 6, Fig.6.1). From this figure the size of the graphene grains seems to bein almost micron scale. However, the Eq. 3.1 shows that the mag-nitude of La is less than 100 nm. This may indicate that there aresmall intrinsic defects in the graphene lattice that cannot be seenwith SEM or that the multiple grain boundaries increase the inten-sity of D-peak. The former reason is more likely because measuredFWHMG ∼30 cm−1 10 nm < La < 50 nm. However, it is worthnoting that the G peak asymmetrical due to the presence of the D′

peak and therefore the size of the crystallite can be bigger than thatexpected from the width of the G peak.

The 2D peak of a few layered graphene made by the direct de-position also resembles that of a few layered graphene grown on acopper foil. Specifically, the 2D peak is located at 2710 cm−1 andthe FWHM2D is about 60 cm−1 indicating that the material is a fewlayered graphene rather than a monolayer graphene.

Although the electrical conductivity of a directly deposited grap-hene film is lower compared to that of graphene films prepared byusing transfer technique [26, 98], the Raman spectrum of a directlygrown graphene is very similar to the that of graphene grown on acopper foil. Therefore the direct deposition technique can be usedin the optics and optoelectronics applications of graphene. Further-more, by optimizing the direct deposition process the electronicproperties of graphene produced by this technique could be im-proved further.

5.1.3 Raman spectrum of pyrolytic carbon

Since the pyrolytic carbon films are deposited on a silica substratewithout a catalyst, it is expected that it is strongly disordered. Thisstrong disordering can of course be detected by the Raman spec-troscopy.

In the Fig. 5.3 is presented a Raman spectrum of a highly amor-



Figure 5.3: A Raman spectrum of a PyC sample. D and G peaks are very wide and anamorphous carbon related peak is located at 1540 cm−1.

phous pyrolytic carbon film that were synthesized and character-ized in the paper II. Once can observed that D and G dominatethe Raman spectrum. Since the D peak originates from the breath-ing mode of an aromatic carbon structure and G peak is found inpresence of sp2 carbon hybridization, the Raman spectrum givesan evidence that the PyC film consists of carbon rings, but is veryamorphous.

Furthermore, fitting of the measured Raman spectrum revealeda wide peak at 1540 cm−1 due to the presence of the amorphouscarbon in the film. In such case the graphene crystallite size can beestimated from Eq. 3.2 that gives ∼2 nm.

The 2D peak of PyC is almost completely suppressed becausethe honeycomb lattice is nearly completely disordered, howeverthere is very weak (in comparison to D and G) peak at 2730 cm−1

(see e.g. Fig. 3.2). Since the 2D peak originates from the dou-ble phonons in opposite phases at Dirac points, the absence of the2D peak in the Raman spectrum may indicate the presence of theasymmetrical carbon rings in the PyC film [39].



the paper I. The D peak in the graphene grown by direct deposi-tion technique on a silica substrate may originate from the rathergrainy structure that can be seen in the SEM image (Chapter 6, Fig.6.1). From this figure the size of the graphene grains seems to bein almost micron scale. However, the Eq. 3.1 shows that the mag-nitude of La is less than 100 nm. This may indicate that there aresmall intrinsic defects in the graphene lattice that cannot be seenwith SEM or that the multiple grain boundaries increase the inten-sity of D-peak. The former reason is more likely because measuredFWHMG ∼30 cm−1 10 nm < La < 50 nm. However, it is worthnoting that the G peak asymmetrical due to the presence of the D′

peak and therefore the size of the crystallite can be bigger than thatexpected from the width of the G peak.

The 2D peak of a few layered graphene made by the direct de-position also resembles that of a few layered graphene grown on acopper foil. Specifically, the 2D peak is located at 2710 cm−1 andthe FWHM2D is about 60 cm−1 indicating that the material is a fewlayered graphene rather than a monolayer graphene.

Although the electrical conductivity of a directly deposited grap-hene film is lower compared to that of graphene films prepared byusing transfer technique [26, 98], the Raman spectrum of a directlygrown graphene is very similar to the that of graphene grown on acopper foil. Therefore the direct deposition technique can be usedin the optics and optoelectronics applications of graphene. Further-more, by optimizing the direct deposition process the electronicproperties of graphene produced by this technique could be im-proved further.

5.1.3 Raman spectrum of pyrolytic carbon

Since the pyrolytic carbon films are deposited on a silica substratewithout a catalyst, it is expected that it is strongly disordered. Thisstrong disordering can of course be detected by the Raman spec-troscopy.

In the Fig. 5.3 is presented a Raman spectrum of a highly amor-



Figure 5.3: A Raman spectrum of a PyC sample. D and G peaks are very wide and anamorphous carbon related peak is located at 1540 cm−1.

phous pyrolytic carbon film that were synthesized and character-ized in the paper II. Once can observed that D and G dominatethe Raman spectrum. Since the D peak originates from the breath-ing mode of an aromatic carbon structure and G peak is found inpresence of sp2 carbon hybridization, the Raman spectrum givesan evidence that the PyC film consists of carbon rings, but is veryamorphous.

Furthermore, fitting of the measured Raman spectrum revealeda wide peak at 1540 cm−1 due to the presence of the amorphouscarbon in the film. In such case the graphene crystallite size can beestimated from Eq. 3.2 that gives ∼2 nm.

The 2D peak of PyC is almost completely suppressed becausethe honeycomb lattice is nearly completely disordered, howeverthere is very weak (in comparison to D and G) peak at 2730 cm−1

(see e.g. Fig. 3.2). Since the 2D peak originates from the dou-ble phonons in opposite phases at Dirac points, the absence of the2D peak in the Raman spectrum may indicate the presence of theasymmetrical carbon rings in the PyC film [39].



The above estimation of the graphene flake size indicates thatthe electrical conductivity of the PyC film should be lower than thatin graphene because of strong electron scattering at the grapheneflake grain boundaries.

5.2 THICKNESS DETERMINATION OF A NANOCARBON FILMON A COPPER SUBSTRATE

In the paper I the optical reflectance spectroscopy was used to re-veal the thickness of the nanographite film deposited on the topof copper substrate. Since SLG absorbs about 2.3 % of transmit-ted light, several graphene layers induce measurable changes inthe copper reflectivity. Such an opportunity is also supported bythe difference in the refractive index of graphene and copper thatis as high as 2 in the wavelength range from 600 nm to 800 nm.Both strong absorption and high refractive index contrast makes itpossible to measure the thickness of nanocarbon film on a coppersubstrate using optical reflectance spectroscopy.

The suggested technique is based on comparing the reflectanceof a bare and graphene coated substrate. Theoretically one graphenelayer should reduce the reflectance of a copper substrate by 1 % at600 nm. However in practice, measuring the reflectance change inthe reflectance is hampered by the copper oxide, which complexrefractive index is nearly the same as that of graphene in the wave-length region of 300 nm – 400 nm. Furthermore a few nanometerthick native copper oxide layers grows on a bare copper substratewithin minutes in normal atmospheric conditions (see paper I).

Since a graphene film prevents short term copper oxidation [109,110], the reflectivity of the graphene coated copper can be measuredas received after the CVD process. In order to measure the re-flectance of bare copper substrate, the copper oxide was removed byacetic acid from the reference sample [111]. Furthermore, since thelight absorption in the copper oxide is almost negligible at wave-lengths longer than 500 nm, the accuracy of the graphene thicknessmeasurement can be increased by measuring reflectivity in wave-



Figure 5.4: (a) Normalized, differential reflectance of a few layered graphene on a coppersubstrate shows film to contain two graphene layers. (b) After the carbon film is trans-ferred on a silica substrate the transmittance spectra shows the graphene film to containapproximately two and half layers.

length range from 500 nm to 800 nm.Experimentally the differential reflectance ∆R = R/R0 − 1, whe-

re R and R0 are reflectances of the graphene coated and refer-ence copper substrates, respectively, was measured by Perkin ElmerLambda-18 spectrophotometer with an integrating sphere. In orderto increase the signal-to-noise ration ∆R was averaged over 10 dif-ferent points of the sample. Results of the measurements are pre-sented in Fig. 5.4. From these data points one may conclude thatthe thickness of the graphene film was about two layers. For thecomparison the graphene film was transferred on a silica substrate



The above estimation of the graphene flake size indicates thatthe electrical conductivity of the PyC film should be lower than thatin graphene because of strong electron scattering at the grapheneflake grain boundaries.

5.2 THICKNESS DETERMINATION OF A NANOCARBON FILMON A COPPER SUBSTRATE

In the paper I the optical reflectance spectroscopy was used to re-veal the thickness of the nanographite film deposited on the topof copper substrate. Since SLG absorbs about 2.3 % of transmit-ted light, several graphene layers induce measurable changes inthe copper reflectivity. Such an opportunity is also supported bythe difference in the refractive index of graphene and copper thatis as high as 2 in the wavelength range from 600 nm to 800 nm.Both strong absorption and high refractive index contrast makes itpossible to measure the thickness of nanocarbon film on a coppersubstrate using optical reflectance spectroscopy.

The suggested technique is based on comparing the reflectanceof a bare and graphene coated substrate. Theoretically one graphenelayer should reduce the reflectance of a copper substrate by 1 % at600 nm. However in practice, measuring the reflectance change inthe reflectance is hampered by the copper oxide, which complexrefractive index is nearly the same as that of graphene in the wave-length region of 300 nm – 400 nm. Furthermore a few nanometerthick native copper oxide layers grows on a bare copper substratewithin minutes in normal atmospheric conditions (see paper I).

Since a graphene film prevents short term copper oxidation [109,110], the reflectivity of the graphene coated copper can be measuredas received after the CVD process. In order to measure the re-flectance of bare copper substrate, the copper oxide was removed byacetic acid from the reference sample [111]. Furthermore, since thelight absorption in the copper oxide is almost negligible at wave-lengths longer than 500 nm, the accuracy of the graphene thicknessmeasurement can be increased by measuring reflectivity in wave-



Figure 5.4: (a) Normalized, differential reflectance of a few layered graphene on a coppersubstrate shows film to contain two graphene layers. (b) After the carbon film is trans-ferred on a silica substrate the transmittance spectra shows the graphene film to containapproximately two and half layers.

length range from 500 nm to 800 nm.Experimentally the differential reflectance ∆R = R/R0 − 1, whe-

re R and R0 are reflectances of the graphene coated and refer-ence copper substrates, respectively, was measured by Perkin ElmerLambda-18 spectrophotometer with an integrating sphere. In orderto increase the signal-to-noise ration ∆R was averaged over 10 dif-ferent points of the sample. Results of the measurements are pre-sented in Fig. 5.4. From these data points one may conclude thatthe thickness of the graphene film was about two layers. For thecomparison the graphene film was transferred on a silica substrate



and the transmittance and Raman spectra was then measured. Thetransmittance curve of the graphene film indicates that the film is amultilayered sample consisting about 2.5 layers.

5.3 OPTICALLY SEMITRANSPARENT, CONDUCTING PYROLYTICCARBON FILM

Since Raman spectra show that the PyC consist mainly of sp2 bondedcarbon, one may expect that ultrathin PyC film can be highly elec-trically conductive. In order to characterize their electric propertieswe performed an optical transmittance and electrical conductivitymeasurement as a function of the film thickness (see paper II).

In the experiment, we measured the optical transmittance andthe sheet resistance of the PyC films. The optical transmittance wasmeasured by Perkin Elmer Lambda-9 spectrophotometer that wascalibrated with a bare silica substrate. The sheet resistance of PyCfilms was measured by conventional four point probe technique[112].

The optical transmittance spectrum of PyC films resembles thatFLG film, however one can observe from Fig. 5.5 that the absorptionpeak at ∼270 nm in PyC is much wider than that in graphene.Since this absorption peak originates from the M saddle point inthe graphene band structure, our experimental finding indicatesthat band structure in PyC is affected by the amorphous structureof the material. In the infrared region the transmittance of PyCfilms is very smooth and comparable to a crystalline graphene.

From the Fig. 5.5 it can be observed that transmittance of asilica substrate drops bellow 50 % when it is coated with a 20 nmthick PyC film. For a crystalline graphite the absorption coefficientα is about 5 × 105 cm−1, while for PyC the α = 4 × 105 cm−1 at800 nm wavelength. The 25 % smaller absorption coefficient of thePyC film can be caused by amorphous carbon, which was revealedby the Raman measurements. The amorphous carbon may increasethe film porosity in comparison with crystalline graphite.

The dependance of the transmittance on the sheet resistance of



Figure 5.5: (a) A few tens of nanometer thick PyC sample with different transmittancescan be easily synthesized directly on a silica substrate. (b) The prepared, amorphous PyCfilm is conductive but the sheet resistance is lower in comparison to crystalline graphenefilm.

the PyC film is shown in the Fig. 5.5. One can observe that for a filmwith 70 % transmittance the sheet resistance is about 2000 Ω/sq,while for a good quality SLG the sheet resistance on silica transmit-tance is about 98 % at the sheet resistance of ∼100 Ω/sq. The lowelectrical conductivity of a PyC film originates from its amorphousnature when small size of graphene crystallites and amorphous car-bon inclusion results in a strong electron scattering. As a resultat the same sheet resistance the transmittance of the PyC film is



and the transmittance and Raman spectra was then measured. Thetransmittance curve of the graphene film indicates that the film is amultilayered sample consisting about 2.5 layers.

5.3 OPTICALLY SEMITRANSPARENT, CONDUCTING PYROLYTICCARBON FILM

Since Raman spectra show that the PyC consist mainly of sp2 bondedcarbon, one may expect that ultrathin PyC film can be highly elec-trically conductive. In order to characterize their electric propertieswe performed an optical transmittance and electrical conductivitymeasurement as a function of the film thickness (see paper II).

In the experiment, we measured the optical transmittance andthe sheet resistance of the PyC films. The optical transmittance wasmeasured by Perkin Elmer Lambda-9 spectrophotometer that wascalibrated with a bare silica substrate. The sheet resistance of PyCfilms was measured by conventional four point probe technique[112].

The optical transmittance spectrum of PyC films resembles thatFLG film, however one can observe from Fig. 5.5 that the absorptionpeak at ∼270 nm in PyC is much wider than that in graphene.Since this absorption peak originates from the M saddle point inthe graphene band structure, our experimental finding indicatesthat band structure in PyC is affected by the amorphous structureof the material. In the infrared region the transmittance of PyCfilms is very smooth and comparable to a crystalline graphene.

From the Fig. 5.5 it can be observed that transmittance of asilica substrate drops bellow 50 % when it is coated with a 20 nmthick PyC film. For a crystalline graphite the absorption coefficientα is about 5 × 105 cm−1, while for PyC the α = 4 × 105 cm−1 at800 nm wavelength. The 25 % smaller absorption coefficient of thePyC film can be caused by amorphous carbon, which was revealedby the Raman measurements. The amorphous carbon may increasethe film porosity in comparison with crystalline graphite.

The dependance of the transmittance on the sheet resistance of



Figure 5.5: (a) A few tens of nanometer thick PyC sample with different transmittancescan be easily synthesized directly on a silica substrate. (b) The prepared, amorphous PyCfilm is conductive but the sheet resistance is lower in comparison to crystalline graphenefilm.

the PyC film is shown in the Fig. 5.5. One can observe that for a filmwith 70 % transmittance the sheet resistance is about 2000 Ω/sq,while for a good quality SLG the sheet resistance on silica transmit-tance is about 98 % at the sheet resistance of ∼100 Ω/sq. The lowelectrical conductivity of a PyC film originates from its amorphousnature when small size of graphene crystallites and amorphous car-bon inclusion results in a strong electron scattering. As a resultat the same sheet resistance the transmittance of the PyC film is



lower than that of graphene or film composed of carbon nanotubes(CNT) [25, 113].

However in contrast to graphene- or CNT-based coating, PyCfilms are inexpensive and can be grown directly on a dielectric sub-strate. The later makes further investigation of the properties andfabrication technology of PyC films especially interesting for mak-ing transparent electrodes in photonic and optoelectronic devices.


6 Graphene deposition andself-assembling on a dielectricsubstrate

Melting of a very thin copper film pre-deposited on a bare silicasubstrate takes place in much lower temperature compared to thebulk copper. Surface melting causes a thin copper film to liquefyand recede forming a few micron size copper droplets. However, byalternating the surface of the silica substrate it is possible to controlthe copper receding in the CVD chamber and hence the graphenedeposition. In this chapter, we will briefly describe the techniquethat allows one to make self-assembling few layer graphene struc-tures on dielectric substrate via controlled liquefying of the copperfilm.

6.1 DIRECT CVD OF A FEW LAYERED GRAPHENE ON A DI-ELECTRIC

In the paper III we present our results on the direct a FLG de-posited directly on a silica substrate with pre-deposited copper. Inthe experiment copper thin films with thicknesses of 100 nm, 200nm and 300 nm were deposited on a bare silica substrate by evap-oration. For the graphene growth we used the receipt describedin the chapter 4. Eventually the copper receding was observed todepend strongly on the original copper film thickness.

Figure 6.1 shows how the copper films of different thicknessesrecede in the CVD chamber. We observed that if the thickness ofthe pre-deposited copper film was 100 nm the copper melted andformed about micron size droplets. Increasing the film thickness to200 nm caused the film to form arbitrary ”finger” structures with



lower than that of graphene or film composed of carbon nanotubes(CNT) [25, 113].

However in contrast to graphene- or CNT-based coating, PyCfilms are inexpensive and can be grown directly on a dielectric sub-strate. The later makes further investigation of the properties andfabrication technology of PyC films especially interesting for mak-ing transparent electrodes in photonic and optoelectronic devices.


6 Graphene deposition andself-assembling on a dielectricsubstrate

Melting of a very thin copper film pre-deposited on a bare silicasubstrate takes place in much lower temperature compared to thebulk copper. Surface melting causes a thin copper film to liquefyand recede forming a few micron size copper droplets. However, byalternating the surface of the silica substrate it is possible to controlthe copper receding in the CVD chamber and hence the graphenedeposition. In this chapter, we will briefly describe the techniquethat allows one to make self-assembling few layer graphene struc-tures on dielectric substrate via controlled liquefying of the copperfilm.

6.1 DIRECT CVD OF A FEW LAYERED GRAPHENE ON A DI-ELECTRIC

In the paper III we present our results on the direct a FLG de-posited directly on a silica substrate with pre-deposited copper. Inthe experiment copper thin films with thicknesses of 100 nm, 200nm and 300 nm were deposited on a bare silica substrate by evap-oration. For the graphene growth we used the receipt describedin the chapter 4. Eventually the copper receding was observed todepend strongly on the original copper film thickness.

Figure 6.1 shows how the copper films of different thicknessesrecede in the CVD chamber. We observed that if the thickness ofthe pre-deposited copper film was 100 nm the copper melted andformed about micron size droplets. Increasing the film thickness to200 nm caused the film to form arbitrary ”finger” structures with



Figure 6.1: (a) After the CVD, a 100 nm thick Cu film is receded to a micron sized droplets.(b) Beneath the Cu droplets a few layered graphene is revealed. (c) By increasing the Cufilm thickness to 200 nm it is possible to obtain more uniform structures. (d) Yet still afew layered graphene is found on the Cu-silica interface. (e) With 300 nm thick Cu filmit is possible to obtain almost fully continuous Cu film after the CVD, (f) and still a fewlayered graphene is found beneath copper. (Graphene grown on substrate-air interface wasremoved by oxygen plasma before the Cu remains were etched off.)

feature size of about a few micrometers. The copper film with thick-ness of 300 nm and higher remained almost solid after the CVDprocess.

Removing the topmost graphene by oxygen plasma and copperremains by wet etching revealed few layered graphene synthesizedbeneath the copper fingers. As it can be seen in Fig. 6.1 the shapeof the graphene ribbons replicate shape of the copper fingers withvery high precision and even sub-micron details prevails the copperremoving process.


Graphene deposition and self-assembling on a dielectric substrate

Although from Raman spectra we can conclude that the opti-cal quality of directly deposited graphene is high, from SEM pic-tures the graphene film looks like consisting of several intertwinedgraphene ribbons. Probably because of these ribbons the conduc-tivity, measured by Su et al. [98] is also rather low in comparisonto a graphene sheet transferred to the silica substrate from copperfoil. Therefore the directly deposited graphene could work well forpohotonic devices, however its quality for should be improved forelectronic and optoelectronic applications.

Furthermore, we observed that when the film thickness was100 nm there was no graphene growth in between of the copperdroplets. However, when the film thickness was 200 nm or more,there was clear signatures of a few layered graphene between cop-per fingers as it was first reported by Ismach et al. [96]. Since agraphene film can be grown by ”receding technique”, described byIsmach, the missing graphene on 100 nm thick copper sample in-dicates that the copper receded before the graphene synthesis thatbegins at temperature around 850 C. But if the copper film thick-ness is 200 nm, or more the copper film still remains solid whengraphene deposition begins and starts to recede soon after that leav-ing some graphene between receded copper fingers.

In paper III we observed that a few layered graphene will growbeneath a thin copper film, copper film receding depends on thecopper film thickness and that a graphene beneath the copper fol-lowed very precisely the receding copper lines. Thus spatially con-trolled receding of copper can be employed to deposit grapheneonto prescribed positions on the silica substrate.

6.2 SELF-ASSEMBLED GRAPHENE GRATING WITH TEN MI-CRON PERIODICITY

Interaction of a liquid droplet with underlying solid surface (i.e.wettability) can be altered by modifying the properties of this sur-face. For example a silica is hydrophilic material and when a waterdroplet is placed on a bare silica substrate it will spread on a wide



Figure 6.1: (a) After the CVD, a 100 nm thick Cu film is receded to a micron sized droplets.(b) Beneath the Cu droplets a few layered graphene is revealed. (c) By increasing the Cufilm thickness to 200 nm it is possible to obtain more uniform structures. (d) Yet still afew layered graphene is found on the Cu-silica interface. (e) With 300 nm thick Cu filmit is possible to obtain almost fully continuous Cu film after the CVD, (f) and still a fewlayered graphene is found beneath copper. (Graphene grown on substrate-air interface wasremoved by oxygen plasma before the Cu remains were etched off.)

feature size of about a few micrometers. The copper film with thick-ness of 300 nm and higher remained almost solid after the CVDprocess.

Removing the topmost graphene by oxygen plasma and copperremains by wet etching revealed few layered graphene synthesizedbeneath the copper fingers. As it can be seen in Fig. 6.1 the shapeof the graphene ribbons replicate shape of the copper fingers withvery high precision and even sub-micron details prevails the copperremoving process.



Although from Raman spectra we can conclude that the opti-cal quality of directly deposited graphene is high, from SEM pic-tures the graphene film looks like consisting of several intertwinedgraphene ribbons. Probably because of these ribbons the conduc-tivity, measured by Su et al. [98] is also rather low in comparisonto a graphene sheet transferred to the silica substrate from copperfoil. Therefore the directly deposited graphene could work well forpohotonic devices, however its quality for should be improved forelectronic and optoelectronic applications.

Furthermore, we observed that when the film thickness was100 nm there was no graphene growth in between of the copperdroplets. However, when the film thickness was 200 nm or more,there was clear signatures of a few layered graphene between cop-per fingers as it was first reported by Ismach et al. [96]. Since agraphene film can be grown by ”receding technique”, described byIsmach, the missing graphene on 100 nm thick copper sample in-dicates that the copper receded before the graphene synthesis thatbegins at temperature around 850 C. But if the copper film thick-ness is 200 nm, or more the copper film still remains solid whengraphene deposition begins and starts to recede soon after that leav-ing some graphene between receded copper fingers.

In paper III we observed that a few layered graphene will growbeneath a thin copper film, copper film receding depends on thecopper film thickness and that a graphene beneath the copper fol-lowed very precisely the receding copper lines. Thus spatially con-trolled receding of copper can be employed to deposit grapheneonto prescribed positions on the silica substrate.

6.2 SELF-ASSEMBLED GRAPHENE GRATING WITH TEN MI-CRON PERIODICITY

Interaction of a liquid droplet with underlying solid surface (i.e.wettability) can be altered by modifying the properties of this sur-face. For example a silica is hydrophilic material and when a waterdroplet is placed on a bare silica substrate it will spread on a wide



Figure 6.2: (a) A silica sample, patterned with ablation technique is coated with 200 nmthick Cu film. In high process temperature Cu film starts to melt (b) and recede (c). (d)Near 900 C the graphitization will begin and carbon migrate beneath copper (e) resultinga few layered graphene beneath receded copper grating. (f) By removing copper remainsone obtains a few layered graphene grating grown directly on a silica substrate.

area. However, modifying of the silica surface by micro- or nanos-tructure the surface becomes hydrophobic and the droplet does notspread on the silica surface but it will keep its droplet shape. Simi-larly one may expect that when a silica substrate with pre-depositedcopper film is placed to the CVD chamber, droplets of liquefied cop-per films will behave differently on a bare and spatially modifiedsilica substrate. This, schematically described in Fig. 6.2, techniquemay open a way for spatially controlled graphene deposition.

In order to demonstrate this, we employed when silica substrateis modified with a few micron feature size (paper IV). In the experi-ment, a bare silica substrate surface was modified with femtosecondlaser ablation using Quantronix Integra -C 3.5 femtosecond laser(center wavelength of 790 nm, pulse length of 120 femtoseconds,and repetition rate of 1 kHz). Specifically, we made a matrix of holesin the substrate with the diameter of 4 µm and 10 µm periodicity,i.e. the spatial characteristics of the hole array were comparablewith feature size of copper fingers in the paper III. Before placinginto CVD chamber, the ablated substrate surface was cleaned outof quartz debris with buffered hydrogen fluoride and then coatedwith 200 nm thick copper film.



Figure 6.3: (a) Receding of 200 nm thick Cu film after CVD process on bare silica sub-strate causes formation of random copper islands. (b) Beneath copper islands a few layeredgraphene is found. (c) On ablation patterned substrate the copper recedes in well orderforming a two-dimensional grating. (d) Beneath copper grating a few layered graphenegrating is revealed.

During the CVD process (see chapter 4) the copper film liquefiesand recedes on areas with and without ablation patterning. Figure6.3 shows that copper deposited on non-ablated areas recedes form-ing a random, few micron sized copper islands. In contrary, in theablated areas, the copper melts and recedes on well ordered man-ner (see Fig. 6.3) by forming a two-dimensional grating. Removingthe remaining copper reveals graphene network that was depositedon the silica substrate beneath the copper. As it can be seen thedeposited graphene pattern precisely reproduces the shape of self-assembled copper network formed on the ablated substrate. Onecan observe from Fig. 6.2 that the Raman spectrum of the graphenedeposited beneath the copper is very similar to that of a few layeredgraphene grown on non-ablated areas.

Thus the receding of the melted copper can be controlled by



Figure 6.2: (a) A silica sample, patterned with ablation technique is coated with 200 nmthick Cu film. In high process temperature Cu film starts to melt (b) and recede (c). (d)Near 900 C the graphitization will begin and carbon migrate beneath copper (e) resultinga few layered graphene beneath receded copper grating. (f) By removing copper remainsone obtains a few layered graphene grating grown directly on a silica substrate.

area. However, modifying of the silica surface by micro- or nanos-tructure the surface becomes hydrophobic and the droplet does notspread on the silica surface but it will keep its droplet shape. Simi-larly one may expect that when a silica substrate with pre-depositedcopper film is placed to the CVD chamber, droplets of liquefied cop-per films will behave differently on a bare and spatially modifiedsilica substrate. This, schematically described in Fig. 6.2, techniquemay open a way for spatially controlled graphene deposition.

In order to demonstrate this, we employed when silica substrateis modified with a few micron feature size (paper IV). In the experi-ment, a bare silica substrate surface was modified with femtosecondlaser ablation using Quantronix Integra -C 3.5 femtosecond laser(center wavelength of 790 nm, pulse length of 120 femtoseconds,and repetition rate of 1 kHz). Specifically, we made a matrix of holesin the substrate with the diameter of 4 µm and 10 µm periodicity,i.e. the spatial characteristics of the hole array were comparablewith feature size of copper fingers in the paper III. Before placinginto CVD chamber, the ablated substrate surface was cleaned outof quartz debris with buffered hydrogen fluoride and then coatedwith 200 nm thick copper film.



Figure 6.3: (a) Receding of 200 nm thick Cu film after CVD process on bare silica sub-strate causes formation of random copper islands. (b) Beneath copper islands a few layeredgraphene is found. (c) On ablation patterned substrate the copper recedes in well orderforming a two-dimensional grating. (d) Beneath copper grating a few layered graphenegrating is revealed.

During the CVD process (see chapter 4) the copper film liquefiesand recedes on areas with and without ablation patterning. Figure6.3 shows that copper deposited on non-ablated areas recedes form-ing a random, few micron sized copper islands. In contrary, in theablated areas, the copper melts and recedes on well ordered man-ner (see Fig. 6.3) by forming a two-dimensional grating. Removingthe remaining copper reveals graphene network that was depositedon the silica substrate beneath the copper. As it can be seen thedeposited graphene pattern precisely reproduces the shape of self-assembled copper network formed on the ablated substrate. Onecan observe from Fig. 6.2 that the Raman spectrum of the graphenedeposited beneath the copper is very similar to that of a few layeredgraphene grown on non-ablated areas.

Thus the receding of the melted copper can be controlled by



Figure 6.4: (a) Grating coated with 180 nm thick copper film. (b) In high process temper-ature Cu will melt and start receding (c) to the grating valleys. The 900 C temperatureactivates catalysis and carbon migrate beneath Cu (d) forming a few layered grapheneto the Cu-silica interface. By removing Cu remains one obtains a few layered graphenegrating on a silica substrate.

modification of the substrate surface. The surface melting of a thincopper film will begin at the grain boundaries and other defects,which are concentrated at the boundaries of holes created by laserablation and than spreads towards non-ablated areas.

6.3 SELF-ASSEMBLED GRAPHENE ON A DIELECTRIC SUB-MICRON GRATING

For optical applications the periodicity of the graphene patternshould be reduced down to sub-micron scale. In order to controlcopper receding with such a spatial resolution the substrate can bemodified using electron beam lithography technique. The test wereperformed using three sets of 1D gratings with depths of 180 nm,330 nm and 480 nm. Moreover, every set contained gratings withperiods of 200 nm, 400 nm, 800 nm, 1200 nm, 1600 nm and 2000nm.

All the gratings were coated with 180 nm thick copper film andthe direct graphene CVD process was then executed. After the CVDprocess we found that on the bare substrate, the copper was re-ceded randomly forming conventional finger-like structures. Thecopper layer on top of grating areas was also receded but the grat-



Figure 6.5: (a) A grating with 180 nm depth and 2 µm periodicity is coated with copperfilm. (b) After the CVD copper is receded and formed grating lines. (c) Beneath coppera few layered graphene is found. (d) Raman spectra of a few layered graphene shows thematerial to be similar obtained by growth on a copper foil.

ing periodicity and depth strongly influenced how the copper wasreceded. The schematics of copper receding on grating structuresis presented in Fig. 6.4.

The most well ordered copper stripes were observed for 180nm deep grating structure with periods from 1.2 µm to 2 µm (seepaper V), where almost all the copper receded in between of thegrating lines. If the grating structure was too deep, the copperremained both between and inside grating valleys. Furthermore ifthe grating periodicity was less than 1.2 µm the copper recedingwas not well controlled but it formed small, micron size copperdroplets on the grating. Thus on can conclude that the gratingparameters has strong impact on the copper receding.

By removing the topmost layer of graphene grown on remain-ing copper by oxygen plasma and then removing copper remains by



Figure 6.4: (a) Grating coated with 180 nm thick copper film. (b) In high process temper-ature Cu will melt and start receding (c) to the grating valleys. The 900 C temperatureactivates catalysis and carbon migrate beneath Cu (d) forming a few layered grapheneto the Cu-silica interface. By removing Cu remains one obtains a few layered graphenegrating on a silica substrate.

modification of the substrate surface. The surface melting of a thincopper film will begin at the grain boundaries and other defects,which are concentrated at the boundaries of holes created by laserablation and than spreads towards non-ablated areas.

6.3 SELF-ASSEMBLED GRAPHENE ON A DIELECTRIC SUB-MICRON GRATING

For optical applications the periodicity of the graphene patternshould be reduced down to sub-micron scale. In order to controlcopper receding with such a spatial resolution the substrate can bemodified using electron beam lithography technique. The test wereperformed using three sets of 1D gratings with depths of 180 nm,330 nm and 480 nm. Moreover, every set contained gratings withperiods of 200 nm, 400 nm, 800 nm, 1200 nm, 1600 nm and 2000nm.

All the gratings were coated with 180 nm thick copper film andthe direct graphene CVD process was then executed. After the CVDprocess we found that on the bare substrate, the copper was re-ceded randomly forming conventional finger-like structures. Thecopper layer on top of grating areas was also receded but the grat-



Figure 6.5: (a) A grating with 180 nm depth and 2 µm periodicity is coated with copperfilm. (b) After the CVD copper is receded and formed grating lines. (c) Beneath coppera few layered graphene is found. (d) Raman spectra of a few layered graphene shows thematerial to be similar obtained by growth on a copper foil.

ing periodicity and depth strongly influenced how the copper wasreceded. The schematics of copper receding on grating structuresis presented in Fig. 6.4.

The most well ordered copper stripes were observed for 180nm deep grating structure with periods from 1.2 µm to 2 µm (seepaper V), where almost all the copper receded in between of thegrating lines. If the grating structure was too deep, the copperremained both between and inside grating valleys. Furthermore ifthe grating periodicity was less than 1.2 µm the copper recedingwas not well controlled but it formed small, micron size copperdroplets on the grating. Thus on can conclude that the gratingparameters has strong impact on the copper receding.

By removing the topmost layer of graphene grown on remain-ing copper by oxygen plasma and then removing copper remains by



FeCl3 a graphene ribbons were revealed beneath the copper. Simi-larly to the our previous experiments (papers III and IV) the shapeof the graphene pattern reproduces very precisely the shape of thethe receded copper up to very smallest details.

The Raman spectra of the graphene coated grating was foundto be very similar compared to previous direct deposition experi-ments. It is, however, noteworthy that since the linewidth of thelargest grating is one micron, the measured Raman signal is col-lected also from the borderlines of the graphene ribbons. That isin the measured Raman spectra, the magnitude of the disorder-induced D peak may be the higher than that it should be in reality.However, in Fig. 6.5 the intensity of the D peak is comparable whitthat in paper III, i.e. the borders of graphene ribbons does not makea significant contribution to the Raman signal.

Spatial resolution proposed direct deposition technique was ob-served to be better than one micron. Therefore we believe thatthis technique could be used in optical applications, especially innear infrared region. Furthermore, we believe that by optimizingthe grating structure and grating line profile the spatial resolutioncan be improved i.e. it can be employed for the direct fabricationgraphene nanopatterns on dielectric substrates.


7 Conclusions

The main focus of this thesis was CVD synthesis and optical charac-terization of ultrathin carbon films. These include CVD single- anda few layer graphene and the pyrolytic carbon films, which are com-posed on intertwined graphene ribbons with size of few nanome-ters. Especial attention was paid to direct deposition of grapheneand self-assembled graphene patterns on a dielectric substrate.

The obtained results of the can be summarized as the following:

• The change in the reflectrance of the copper foil after deposi-tion of a few graphene layers have been found to exceed 1 %.This allows us to propose a novel technique for in-situ mea-surement of the CVD graphene thickness using reflectancespectroscopy.

• A method was developed to synthesize an ultrathin PyC filmdirectly on a silica substrate using methane as carbon sourcein the CVD process. This well controllable catalyst free tech-nique can be employed for coating of a dielectric substrate ofany shape and size with conductive carbon layer with thick-ness from 5 to 50 nm. The sheet resistance of ultrathin PyCfilms was found to be lower than that of graphene due to elec-tron scattering on intertwined graphene ribbon that constitutePyC.

• Direct CVD of graphene on dielectric by using pre-depositedthin copper film was demonstrated. We found experimentallythat thickness of sacrificial copper layer can strongly affect thecopper receding. In particular if the copper layer thicknessesis smaller than 200 nm, CVD processing returns finger likecopper-graphene structures on the dielectric substrates, whileincreasing the film thickness up to 300 nm the copper laterremains nearly uniform. We show experimentally that in thestudied range of the copper layer thicknesses, CVD processing



FeCl3 a graphene ribbons were revealed beneath the copper. Simi-larly to the our previous experiments (papers III and IV) the shapeof the graphene pattern reproduces very precisely the shape of thethe receded copper up to very smallest details.

The Raman spectra of the graphene coated grating was foundto be very similar compared to previous direct deposition experi-ments. It is, however, noteworthy that since the linewidth of thelargest grating is one micron, the measured Raman signal is col-lected also from the borderlines of the graphene ribbons. That isin the measured Raman spectra, the magnitude of the disorder-induced D peak may be the higher than that it should be in reality.However, in Fig. 6.5 the intensity of the D peak is comparable whitthat in paper III, i.e. the borders of graphene ribbons does not makea significant contribution to the Raman signal.

Spatial resolution proposed direct deposition technique was ob-served to be better than one micron. Therefore we believe thatthis technique could be used in optical applications, especially innear infrared region. Furthermore, we believe that by optimizingthe grating structure and grating line profile the spatial resolutioncan be improved i.e. it can be employed for the direct fabricationgraphene nanopatterns on dielectric substrates.


7 Conclusions

The main focus of this thesis was CVD synthesis and optical charac-terization of ultrathin carbon films. These include CVD single- anda few layer graphene and the pyrolytic carbon films, which are com-posed on intertwined graphene ribbons with size of few nanome-ters. Especial attention was paid to direct deposition of grapheneand self-assembled graphene patterns on a dielectric substrate.

The obtained results of the can be summarized as the following:

• The change in the reflectrance of the copper foil after deposi-tion of a few graphene layers have been found to exceed 1 %.This allows us to propose a novel technique for in-situ mea-surement of the CVD graphene thickness using reflectancespectroscopy.

• A method was developed to synthesize an ultrathin PyC filmdirectly on a silica substrate using methane as carbon sourcein the CVD process. This well controllable catalyst free tech-nique can be employed for coating of a dielectric substrate ofany shape and size with conductive carbon layer with thick-ness from 5 to 50 nm. The sheet resistance of ultrathin PyCfilms was found to be lower than that of graphene due to elec-tron scattering on intertwined graphene ribbon that constitutePyC.

• Direct CVD of graphene on dielectric by using pre-depositedthin copper film was demonstrated. We found experimentallythat thickness of sacrificial copper layer can strongly affect thecopper receding. In particular if the copper layer thicknessesis smaller than 200 nm, CVD processing returns finger likecopper-graphene structures on the dielectric substrates, whileincreasing the film thickness up to 300 nm the copper laterremains nearly uniform. We show experimentally that in thestudied range of the copper layer thicknesses, CVD processing



results in the synthesis of the few layer graphene on copper-dielectric interface i.e. on a dielectric substrate beneath thecopper.

• We demonstrated that copper receding during the CVD pro-cess can be controlled by modification of the dielectric sub-strate surface. By using silica substrate with surface pro-filed using femtosecond laser ablation we demonstrated self-assembling of 2D graphene grating with period of several mi-crons.

• By pre-depositing sacrificial copper layer on dielectric sub-strates modified by e-beam lithography technique, we demon-strated direct synthesis of graphene on dielectric with spatialresolution of hundreds nanometers. The self-assembling ofperiodic graphene pattern on dielectric substrates is governedby the copper layer thickness as well as by the substrate mod-ification.

In this work we have developed a synthesis and characterizationtechniques that can be used in fabrication of ultrathin nanocarbonfilms for photonic, electronic and optoelectronic applications. Webelieve that proposed and demonstrated self-assembling graphenedeposition directly on a dielectric substrate can open novel routesfor monolithic integration of graphene and silicon photonics.


Bibliography

[1] P. Kuzhir, N. Volynets, S. Maksimenko, T. Kaplas, andY. Svirko, “Multilayered graphene in Ka-band nanoscale coat-ing for aerospace applications,” J. Nanosci and Nanotech 13,5864–5867 (2013).

[2] K. Batrakov, P. Kuzhir, S. Maksimenko, A. Paddubskaya,S. Voronovich, T. Kaplas, and Y. Svirko, “Enhanced mi-crowave shielding effectiveness of ultrathin pyrolytic carbonfilms,” Appl. Phys. Lett. 103, 073117–4 (2013).

[3] P. Kuzhir, A. Paddubskaya, S. Maksimenko, T. Kaplas, andY. Svirko, “Microwave absorption properties of pyrolytic car-bon nanofilm,” Nanoscale Res. Lett. 8, 1–6 (2013).

[4] P. Kuzhir, A. Paddubskaya, S. Maksimenko, T. Kaplas, andY. Svirko, “Transport and electromagnetic properties of ultra-thin pyrolytic carbon films,” J. Nanophot. 7, 073595 (2013).

[5] T. Ning, O. Hyvarinen, H. Pietarinen, T. Kaplas, M. Kaura-nen, and G. Genty, “Third-harmonic UV generation in siliconnitride nanostructures,” Opt. Express 21, 2012–2017 (2013).

[6] P. A. Obraztsov, T. Kaplas, S. V. Garnov, M. Kuwata-Gonokami, A. N. Obraztsov, and Y. Svirko, “All-optical con-trol of ultrafast photocurrents in unbiased graphene,” NaturePhot. Submitted in Nov. (2012).

[7] V. V. Zhurikhina, P. N. Brunkov, V. G. Melehin, T. Kaplas,Y. Svirko, V. V. Rutckaia, and A. Lipovskii, “Self-assembledsilver nanoislands formed on glass surface via out-diffusionfor multiple usages in SERS applications,” Nanoscale Res. Lett.7, 676 (2012).



results in the synthesis of the few layer graphene on copper-dielectric interface i.e. on a dielectric substrate beneath thecopper.

• We demonstrated that copper receding during the CVD pro-cess can be controlled by modification of the dielectric sub-strate surface. By using silica substrate with surface pro-filed using femtosecond laser ablation we demonstrated self-assembling of 2D graphene grating with period of several mi-crons.

• By pre-depositing sacrificial copper layer on dielectric sub-strates modified by e-beam lithography technique, we demon-strated direct synthesis of graphene on dielectric with spatialresolution of hundreds nanometers. The self-assembling ofperiodic graphene pattern on dielectric substrates is governedby the copper layer thickness as well as by the substrate mod-ification.

In this work we have developed a synthesis and characterizationtechniques that can be used in fabrication of ultrathin nanocarbonfilms for photonic, electronic and optoelectronic applications. Webelieve that proposed and demonstrated self-assembling graphenedeposition directly on a dielectric substrate can open novel routesfor monolithic integration of graphene and silicon photonics.


Bibliography

[1] P. Kuzhir, N. Volynets, S. Maksimenko, T. Kaplas, andY. Svirko, “Multilayered graphene in Ka-band nanoscale coat-ing for aerospace applications,” J. Nanosci and Nanotech 13,5864–5867 (2013).

[2] K. Batrakov, P. Kuzhir, S. Maksimenko, A. Paddubskaya,S. Voronovich, T. Kaplas, and Y. Svirko, “Enhanced mi-crowave shielding effectiveness of ultrathin pyrolytic carbonfilms,” Appl. Phys. Lett. 103, 073117–4 (2013).

[3] P. Kuzhir, A. Paddubskaya, S. Maksimenko, T. Kaplas, andY. Svirko, “Microwave absorption properties of pyrolytic car-bon nanofilm,” Nanoscale Res. Lett. 8, 1–6 (2013).

[4] P. Kuzhir, A. Paddubskaya, S. Maksimenko, T. Kaplas, andY. Svirko, “Transport and electromagnetic properties of ultra-thin pyrolytic carbon films,” J. Nanophot. 7, 073595 (2013).

[5] T. Ning, O. Hyvarinen, H. Pietarinen, T. Kaplas, M. Kaura-nen, and G. Genty, “Third-harmonic UV generation in siliconnitride nanostructures,” Opt. Express 21, 2012–2017 (2013).

[6] P. A. Obraztsov, T. Kaplas, S. V. Garnov, M. Kuwata-Gonokami, A. N. Obraztsov, and Y. Svirko, “All-optical con-trol of ultrafast photocurrents in unbiased graphene,” NaturePhot. Submitted in Nov. (2012).

[7] V. V. Zhurikhina, P. N. Brunkov, V. G. Melehin, T. Kaplas,Y. Svirko, V. V. Rutckaia, and A. Lipovskii, “Self-assembledsilver nanoislands formed on glass surface via out-diffusionfor multiple usages in SERS applications,” Nanoscale Res. Lett.7, 676 (2012).



[8] T. Kaplas, L. Karvonen, J. Ronn, M. Saleem, S. Kujala,S. Honkanen, , and Y. Svirko, “Nonlinear refraction in semi-transparent pyrolytic carbon films,” Opt. Mat. Express 2, 1822–1827 (2012).

[9] T. Ning, H. Pietarinen, O. Hyvarinen, R. Kumar, T. Kaplas,M. Kauranen, , and G. Genty, “Efficient second-harmonicgeneration in silicon nitride resonant waveguide gratings,”Opt. Lett. 37, 4269–4271 (2012).

[10] T. Kaplas, R. Ismagilov, A. Obraztsov, and Y. Svirko, “Char-acterization of nanographite films by specular gloss measure-ments,” J. Nanoelec. and Optoelec. 7, 54–59 (2012).

[11] T. Kaplas, K. Laitinen, T. Moilanen, Y. Tolonen, K. Albrecht,and R. Silvennoinen, “Optical sensing of parameters be-hind delivery of watercolour in alternative block materialsfor Japanese woodblock print making,” Opt. Rev. 17, 252–256(2010).

[12] H. Pierson, Handbook of Carbon, Graphite, Diamond andFullerenes: Properties, Processing and Applications (Noyes Publi-cation, New Jersey, 1993).

[13] W. E. Sawyer and A. Man, US Patent 229335 (, Cambridge,June 29, 1880).

[14] B. Eisenberg and H. Bliss, “Kinetics of methane pyrolysis,”Chem. Eng. Prog. Symp. Ser. 63, 3 (1967).

[15] H. B. Palmer, J. Lahaye, and K. C. Hou, “On the Kinetics andMechanism of the Thermal Decomposition of Methane in aFlow System,” J. Phys. Chem. 72, 348–353 (1968).

[16] J. C. Charlier, X. Gonze, and J. P. Michenaud, “Graphite Inter-planar Bonding: Electronic Delocalization and van der WaalsInteraction,” Europhys. Lett. 28, 403–408 (1994).


Bibliography

[17] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang,Y. Zhang, S. V. Dubonos, I. V. Grigorieva, and A. A. Firsov,“Electric Field Effect in Atomically Thin Carbon Films,” Sci-ence 306, 666–669 (2004).

[18] P. R. Wallace, “The Band Theory of Graphite,” Phys. Rev. 71,622–634 (1947).

[19] A. K. Geim and K. S. Novoselov, “The Rise of Graphene,”Nature Materials 6, 183–191 (2007).

[20] K. S. Novoselov, V. I. Fal’ko, L. Colombo, P. R. Gellert, M. G.Schwab, and K. Kim, “A Roadmap for Graphene,” Science315, 1379 (2007).

[21] F. Bonaccorso, T. H. Z. Sun, and A. C. Ferrari, “Graphene Pho-tonics and Optoelectronics,” Nature Phot. 4, 611–622 (2010).

[22] Z. Sun, T. Hasan, F. Torrisi, D. Popa, G. Privitera, F. Wang,F. Bonaccorso, D. M. Basko, and A. C. Ferrari, “GrapheneMode-Locked Ultrafast Laser,” ACS Nano 4, 803–810 (2010).

[23] J. S. Bunch, A. M. van der Zande, S. S. Verbridge, I. W.Frank, D. M. Tanenbaum, J. M. Parpia, H. G. Craighead, andP. L. McEuen, “Electromechanical Resonators from GrapheneSheets,” Science 315, 490–493 (2007).

[24] K. S. Novoselov, Z. Jiang, Y. Zhang, S. V. Morozov, H. L.Stormer, U. Zeitler, J. C. Maan, G. S. Boebinger, P. Kim, andA. K. Geim, “Room-Temperature Quantum Hall Effect inGraphene,” Science 315, 1379 (2007).

[25] S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, J. Balakr-ishnan, T. Lei, H. R. Kim, Y. I. Song, Y.-J. Kim, K. S. Kim,B. Ozyilmaz, J.-H. Ahn, B. H. Hong, and S. Iijima, “Roll-to-Roll Production of 30-Inch Graphene Films for TransparentElectrodes,” Nature Nanotech. 5, 574–578 (2010).



[8] T. Kaplas, L. Karvonen, J. Ronn, M. Saleem, S. Kujala,S. Honkanen, , and Y. Svirko, “Nonlinear refraction in semi-transparent pyrolytic carbon films,” Opt. Mat. Express 2, 1822–1827 (2012).

[9] T. Ning, H. Pietarinen, O. Hyvarinen, R. Kumar, T. Kaplas,M. Kauranen, , and G. Genty, “Efficient second-harmonicgeneration in silicon nitride resonant waveguide gratings,”Opt. Lett. 37, 4269–4271 (2012).

[10] T. Kaplas, R. Ismagilov, A. Obraztsov, and Y. Svirko, “Char-acterization of nanographite films by specular gloss measure-ments,” J. Nanoelec. and Optoelec. 7, 54–59 (2012).

[11] T. Kaplas, K. Laitinen, T. Moilanen, Y. Tolonen, K. Albrecht,and R. Silvennoinen, “Optical sensing of parameters be-hind delivery of watercolour in alternative block materialsfor Japanese woodblock print making,” Opt. Rev. 17, 252–256(2010).

[12] H. Pierson, Handbook of Carbon, Graphite, Diamond andFullerenes: Properties, Processing and Applications (Noyes Publi-cation, New Jersey, 1993).

[13] W. E. Sawyer and A. Man, US Patent 229335 (, Cambridge,June 29, 1880).

[14] B. Eisenberg and H. Bliss, “Kinetics of methane pyrolysis,”Chem. Eng. Prog. Symp. Ser. 63, 3 (1967).

[15] H. B. Palmer, J. Lahaye, and K. C. Hou, “On the Kinetics andMechanism of the Thermal Decomposition of Methane in aFlow System,” J. Phys. Chem. 72, 348–353 (1968).

[16] J. C. Charlier, X. Gonze, and J. P. Michenaud, “Graphite Inter-planar Bonding: Electronic Delocalization and van der WaalsInteraction,” Europhys. Lett. 28, 403–408 (1994).


Bibliography

[17] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang,Y. Zhang, S. V. Dubonos, I. V. Grigorieva, and A. A. Firsov,“Electric Field Effect in Atomically Thin Carbon Films,” Sci-ence 306, 666–669 (2004).

[18] P. R. Wallace, “The Band Theory of Graphite,” Phys. Rev. 71,622–634 (1947).

[19] A. K. Geim and K. S. Novoselov, “The Rise of Graphene,”Nature Materials 6, 183–191 (2007).

[20] K. S. Novoselov, V. I. Fal’ko, L. Colombo, P. R. Gellert, M. G.Schwab, and K. Kim, “A Roadmap for Graphene,” Science315, 1379 (2007).

[21] F. Bonaccorso, T. H. Z. Sun, and A. C. Ferrari, “Graphene Pho-tonics and Optoelectronics,” Nature Phot. 4, 611–622 (2010).

[22] Z. Sun, T. Hasan, F. Torrisi, D. Popa, G. Privitera, F. Wang,F. Bonaccorso, D. M. Basko, and A. C. Ferrari, “GrapheneMode-Locked Ultrafast Laser,” ACS Nano 4, 803–810 (2010).

[23] J. S. Bunch, A. M. van der Zande, S. S. Verbridge, I. W.Frank, D. M. Tanenbaum, J. M. Parpia, H. G. Craighead, andP. L. McEuen, “Electromechanical Resonators from GrapheneSheets,” Science 315, 490–493 (2007).

[24] K. S. Novoselov, Z. Jiang, Y. Zhang, S. V. Morozov, H. L.Stormer, U. Zeitler, J. C. Maan, G. S. Boebinger, P. Kim, andA. K. Geim, “Room-Temperature Quantum Hall Effect inGraphene,” Science 315, 1379 (2007).

[25] S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, J. Balakr-ishnan, T. Lei, H. R. Kim, Y. I. Song, Y.-J. Kim, K. S. Kim,B. Ozyilmaz, J.-H. Ahn, B. H. Hong, and S. Iijima, “Roll-to-Roll Production of 30-Inch Graphene Films for TransparentElectrodes,” Nature Nanotech. 5, 574–578 (2010).



[26] C. Mattevi, H. Kim, and M. Chhowalla, “A Review of Chemi-cal Vapour Deposition of Graphene on Copper,” J Mat. Chem.21, 3324–3334 (2011).

[27] E. Taft and H. Philipp, “Optical Properties of Graphite,” Phys.Rev. 138, A197–A202 (1965).

[28] Z. Ren, Y. Lan, and Y. Wang, Aligned Carbon Nanotubes:Physics, Concepts, Fabrication and Devices (Springer, Berlin,2013).

[29] L. A. Falkovsky, “Optical Properties of Graphene,” Journal ofPhysics: Conference Series 129, 012004–7 (2008).

[30] R. Nairv, P. Blake, A. Grigorenko, K. Novoselov, T. Booth,T. Stauber, N. Peres, and A. Geim, “Fine Structure ConstantDefines Visual Transparency of Graphene,” Science 320, 1308(2008).

[31] A. C. Neto, F. Guinea, N. Peres, K. Novoselov, and A. Geim,“The Electronic Properties of Graphene,” Reviews of ModernPhysics 81, 109–162 (2009).

[32] S. Reich, J. Maultzsch, C. Thomsen, and P. Ordejon, “Tight-binding description of graphene,” Physical Review B 66,035412–5 (2002).

[33] M. Sprinkle, D. Siegel, Y. Hu, J. Hicks, A. Tejeda, A. Taleb-Ibrahimi, P. L. Fevre, F. Bertran, S. Vizzini, H. Enriquez,S. Chiang, P. Soukiassian, C. Berger, W. A. de Heer, A. Lan-zara, and E. H. Conrad, “First Direct Observation of a NearlyIdeal Graphene Band Structure,” Phys. Rev. Lett. 103, 226803–4(2009).

[34] E. McCann, “Asymmetry Gap in the Electronic Band Struc-ture of Bilayer Graphene,” Phys. Rev. B 74, 161403–4 (2006).

[35] J. Slonczewski and P. Weiss, “Band Structure of Graphite,”Phys. Rev. 109, 272–279 (1958).


Bibliography

[36] J. McClure, “Band Structure of Graphite and de Haas-vanAlphen Effect,” Phys. Rev. 108, 612–618 (1957).

[37] M. Craciun, S. Russo, M. Yamamoto, J. Oostinga, A. Mor-purgo, and S. Tarucha, “Trilayer Graphene is a Semimetalwith a Gate-tunable Band Overlap,” Nature Nanotech. 4, 383–388 (2009).

[38] B. Partoens and F. Peeters, “From Graphene to Graphite: Elec-tronic Structure around the K point,” Phys. Rev. B 74, 075404–11 (2006).

[39] A. Ferrari and D. Basko, “Raman Spectroscopy as a VersatileTool For Studying the Properties of Graphene,” Nature Nan-otech. 8, 235–246 (2013).

[40] S. Reich and C. Thomsen, “Raman spectroscopy of graphite,”Phil. Trans. R. Soc. Lond. A 362, 2271–2288 (2004).

[41] M. Dresselhaus, G. Dresselhaus, R. Saito, and A. Jorio, “Ra-man Spectroscopy of Carbon Nanotubes,” Phys. Reports 409,235–246 (2005).

[42] Y. Wang, D. Alsmeyer, and R. McCreery, “Raman Spec-troscopy of Carbon Materials: Structural Basis of ObservedSpectra,” Chem. Mater. 2, 557–563 (1990).

[43] M. Dresselhaus, G. Dresselhaus, and P. Eklund, “Raman Scat-tering in Fullerenes,” J. Raman Spectroscopy 27, 351–371 (1996).

[44] F. Tuinstra and J. Koenig, “Raman Spectrum of Graphite,” J.Chem. Phys. 53, 1126–1130 (1970).

[45] A. Ferrari, J. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri,F. Mauri, S. Piscanec, D. Jiang, K. Novoselov, S. Roth, andA. Geim, “Raman Spectrum of Graphene and Graphene Lay-ers,” Phys. Rev. Lett. 97, 187401–4 (2006).

[46] D. Graf, F. Molitor, K. Ensslin, C. Stampfer, A. Jungen, C. Hi-erold, and L. Wirtz, “Spatially Resolved Raman Spectroscopy



[26] C. Mattevi, H. Kim, and M. Chhowalla, “A Review of Chemi-cal Vapour Deposition of Graphene on Copper,” J Mat. Chem.21, 3324–3334 (2011).

[27] E. Taft and H. Philipp, “Optical Properties of Graphite,” Phys.Rev. 138, A197–A202 (1965).

[28] Z. Ren, Y. Lan, and Y. Wang, Aligned Carbon Nanotubes:Physics, Concepts, Fabrication and Devices (Springer, Berlin,2013).

[29] L. A. Falkovsky, “Optical Properties of Graphene,” Journal ofPhysics: Conference Series 129, 012004–7 (2008).

[30] R. Nairv, P. Blake, A. Grigorenko, K. Novoselov, T. Booth,T. Stauber, N. Peres, and A. Geim, “Fine Structure ConstantDefines Visual Transparency of Graphene,” Science 320, 1308(2008).

[31] A. C. Neto, F. Guinea, N. Peres, K. Novoselov, and A. Geim,“The Electronic Properties of Graphene,” Reviews of ModernPhysics 81, 109–162 (2009).

[32] S. Reich, J. Maultzsch, C. Thomsen, and P. Ordejon, “Tight-binding description of graphene,” Physical Review B 66,035412–5 (2002).

[33] M. Sprinkle, D. Siegel, Y. Hu, J. Hicks, A. Tejeda, A. Taleb-Ibrahimi, P. L. Fevre, F. Bertran, S. Vizzini, H. Enriquez,S. Chiang, P. Soukiassian, C. Berger, W. A. de Heer, A. Lan-zara, and E. H. Conrad, “First Direct Observation of a NearlyIdeal Graphene Band Structure,” Phys. Rev. Lett. 103, 226803–4(2009).

[34] E. McCann, “Asymmetry Gap in the Electronic Band Struc-ture of Bilayer Graphene,” Phys. Rev. B 74, 161403–4 (2006).

[35] J. Slonczewski and P. Weiss, “Band Structure of Graphite,”Phys. Rev. 109, 272–279 (1958).


Bibliography

[36] J. McClure, “Band Structure of Graphite and de Haas-vanAlphen Effect,” Phys. Rev. 108, 612–618 (1957).

[37] M. Craciun, S. Russo, M. Yamamoto, J. Oostinga, A. Mor-purgo, and S. Tarucha, “Trilayer Graphene is a Semimetalwith a Gate-tunable Band Overlap,” Nature Nanotech. 4, 383–388 (2009).

[38] B. Partoens and F. Peeters, “From Graphene to Graphite: Elec-tronic Structure around the K point,” Phys. Rev. B 74, 075404–11 (2006).

[39] A. Ferrari and D. Basko, “Raman Spectroscopy as a VersatileTool For Studying the Properties of Graphene,” Nature Nan-otech. 8, 235–246 (2013).

[40] S. Reich and C. Thomsen, “Raman spectroscopy of graphite,”Phil. Trans. R. Soc. Lond. A 362, 2271–2288 (2004).

[41] M. Dresselhaus, G. Dresselhaus, R. Saito, and A. Jorio, “Ra-man Spectroscopy of Carbon Nanotubes,” Phys. Reports 409,235–246 (2005).

[42] Y. Wang, D. Alsmeyer, and R. McCreery, “Raman Spec-troscopy of Carbon Materials: Structural Basis of ObservedSpectra,” Chem. Mater. 2, 557–563 (1990).

[43] M. Dresselhaus, G. Dresselhaus, and P. Eklund, “Raman Scat-tering in Fullerenes,” J. Raman Spectroscopy 27, 351–371 (1996).

[44] F. Tuinstra and J. Koenig, “Raman Spectrum of Graphite,” J.Chem. Phys. 53, 1126–1130 (1970).

[45] A. Ferrari, J. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri,F. Mauri, S. Piscanec, D. Jiang, K. Novoselov, S. Roth, andA. Geim, “Raman Spectrum of Graphene and Graphene Lay-ers,” Phys. Rev. Lett. 97, 187401–4 (2006).

[46] D. Graf, F. Molitor, K. Ensslin, C. Stampfer, A. Jungen, C. Hi-erold, and L. Wirtz, “Spatially Resolved Raman Spectroscopy



of Single- and Few-Layer Graphene,” Nano Lett. 7, 238–242(2007).

[47] J. S. Park, A. Reina, R. Saito, J. Kong, G. Dresselhaus, andM. S. Dresselhaus, “G’ Band Raman Spectra of Single, Doubleand Triple Layer Graphene,” Carbon 47, 1303–1310 (2009).

[48] M. Kalbac, A. Reina-Cecco, H. Farhat, J. Kong, L. Kavan,and M. S. Dresselhaus, “The Influence of Strong Electronand Hole Doping on the Raman Intensity of Chemical Vapor-Deposition Graphene,” Nano Lett. 4, 6055–6063 (2010).

[49] C. Casiraghi, S. Pisana, K. S. Novoselov, A. K. Geim, andA. C. Ferrari, “Raman Fingerprint of Charged Impurities inGraphene,” Appl. Phys. Lett. 91, 233108–3 (2007).

[50] A. Das, S. Pisana, B. Chakraborty, S. Piscanec, S. K. Saha,U. V. Waghmare, K. S. Novoselov, H. R. Krishnamurthy,A. K. Geim, A. C. Ferrari, and A. K. Sood, “Monitor-ing Dopants by Raman Scattering in an ElectrochemicallyTop-Gated Graphene Transistor,” Nature Nanotech. 3, 210–215(2008).

[51] M. Huang, H. Yan, T. F. Heinz, and J. Hone, “Probing Strain-Induced Electronic Structure Change in Graphene by RamanSpectroscopy,” Nano Lett. 10, 4070–4079 (2010).

[52] M. Mohr, J. Maultzsch, and C. Thomsen, “Splitting of theRaman 2D Band of Graphene Subjected to Strain,” Phys. Rev.B 82, 201409–4 (2010).

[53] Z. H. Ni, L. A. Ponomarenko, R. R. Nair, R. Yang, S. Anissi-mova, I. V. Grigorieva, F. Schedin, P. Blake, Z. X. Shen, E. H.Hill, K. S. Novoselov, and A. K. Geim, “On Resonant Scatter-ers As a Factor Limiting Carrier Mobility in Graphene,” NanoLett. 10, 3868–3872 (2010).


Bibliography

[54] A. C. Ferrari, “Raman Spectroscopy of Graphene andGraphite: Disorder, Electron–Phonon Coupling, Doping andNonadiabatic Effects,” Solid State Comm. 143, 47–57 (2007).

[55] L. M. Malard, M. Pimenta, G. Dresselhaus, and M. Dressel-haus, “Raman Spectroscopy in Graphene,” Phys. Reports 473,51–87 (2009).

[56] A. C. Ferrari and J. Robertson, “Interpretation of Raman Spec-tra of Disordered and Amorphous Carbon,” Phys. Rev. B 61,14095–14107 (2000).

[57] A. C. Ferrari and J. Robertson, “Resonant Raman Spec-troscopy of Disordered, Amorphous, and Diamondlike car-bon,” Phys. Rev. B 64, 075414–4 (2001).

[58] C. Thomsen and S. Reich, “Double Resonant Raman Scatter-ing in Graphite,” Phys. Rev. Lett. 85, 5214–5217 (2000).

[59] A. C. Ferrari and J. Robertson, “Raman Spectroscopy ofAmorphous, Nanostructured, Diamond-like Carbon, andNanodiamond,” Phil. Trans. R. Soc. Lond. A 362, 2477–2512(2004).

[60] M. J. Matthews, M. A. Pimenta, G. Dresselhaus, M. S. Dres-selhaus, and M. Endo, “Origin of Dispersive Effects of theRaman D Band in Carbon Materials,” Phys. Rev. B 59, R6586–R6588 (1999).

[61] M. A. Pimenta, G. Dresselhaus, M. S. Dresselhaus, L. G. Can-cado, A. Jorio, and R. Saito, “Studying Disorder in Graphite-Based Systems by Raman Spectroscopy,” Phys. Chem. Chem.Phys. 9, 1276–1291 (2007).

[62] T. Jawhari, A. Roid, and J. Casado, “Raman SpectroscopicCharacterization of Some Commercially Available CarbonBlack Materials,” Carbon 33, 1561–1565 (1995).



of Single- and Few-Layer Graphene,” Nano Lett. 7, 238–242(2007).

[47] J. S. Park, A. Reina, R. Saito, J. Kong, G. Dresselhaus, andM. S. Dresselhaus, “G’ Band Raman Spectra of Single, Doubleand Triple Layer Graphene,” Carbon 47, 1303–1310 (2009).

[48] M. Kalbac, A. Reina-Cecco, H. Farhat, J. Kong, L. Kavan,and M. S. Dresselhaus, “The Influence of Strong Electronand Hole Doping on the Raman Intensity of Chemical Vapor-Deposition Graphene,” Nano Lett. 4, 6055–6063 (2010).

[49] C. Casiraghi, S. Pisana, K. S. Novoselov, A. K. Geim, andA. C. Ferrari, “Raman Fingerprint of Charged Impurities inGraphene,” Appl. Phys. Lett. 91, 233108–3 (2007).

[50] A. Das, S. Pisana, B. Chakraborty, S. Piscanec, S. K. Saha,U. V. Waghmare, K. S. Novoselov, H. R. Krishnamurthy,A. K. Geim, A. C. Ferrari, and A. K. Sood, “Monitor-ing Dopants by Raman Scattering in an ElectrochemicallyTop-Gated Graphene Transistor,” Nature Nanotech. 3, 210–215(2008).

[51] M. Huang, H. Yan, T. F. Heinz, and J. Hone, “Probing Strain-Induced Electronic Structure Change in Graphene by RamanSpectroscopy,” Nano Lett. 10, 4070–4079 (2010).

[52] M. Mohr, J. Maultzsch, and C. Thomsen, “Splitting of theRaman 2D Band of Graphene Subjected to Strain,” Phys. Rev.B 82, 201409–4 (2010).

[53] Z. H. Ni, L. A. Ponomarenko, R. R. Nair, R. Yang, S. Anissi-mova, I. V. Grigorieva, F. Schedin, P. Blake, Z. X. Shen, E. H.Hill, K. S. Novoselov, and A. K. Geim, “On Resonant Scatter-ers As a Factor Limiting Carrier Mobility in Graphene,” NanoLett. 10, 3868–3872 (2010).


Bibliography

[54] A. C. Ferrari, “Raman Spectroscopy of Graphene andGraphite: Disorder, Electron–Phonon Coupling, Doping andNonadiabatic Effects,” Solid State Comm. 143, 47–57 (2007).

[55] L. M. Malard, M. Pimenta, G. Dresselhaus, and M. Dressel-haus, “Raman Spectroscopy in Graphene,” Phys. Reports 473,51–87 (2009).

[56] A. C. Ferrari and J. Robertson, “Interpretation of Raman Spec-tra of Disordered and Amorphous Carbon,” Phys. Rev. B 61,14095–14107 (2000).

[57] A. C. Ferrari and J. Robertson, “Resonant Raman Spec-troscopy of Disordered, Amorphous, and Diamondlike car-bon,” Phys. Rev. B 64, 075414–4 (2001).

[58] C. Thomsen and S. Reich, “Double Resonant Raman Scatter-ing in Graphite,” Phys. Rev. Lett. 85, 5214–5217 (2000).

[59] A. C. Ferrari and J. Robertson, “Raman Spectroscopy ofAmorphous, Nanostructured, Diamond-like Carbon, andNanodiamond,” Phil. Trans. R. Soc. Lond. A 362, 2477–2512(2004).

[60] M. J. Matthews, M. A. Pimenta, G. Dresselhaus, M. S. Dres-selhaus, and M. Endo, “Origin of Dispersive Effects of theRaman D Band in Carbon Materials,” Phys. Rev. B 59, R6586–R6588 (1999).

[61] M. A. Pimenta, G. Dresselhaus, M. S. Dresselhaus, L. G. Can-cado, A. Jorio, and R. Saito, “Studying Disorder in Graphite-Based Systems by Raman Spectroscopy,” Phys. Chem. Chem.Phys. 9, 1276–1291 (2007).

[62] T. Jawhari, A. Roid, and J. Casado, “Raman SpectroscopicCharacterization of Some Commercially Available CarbonBlack Materials,” Carbon 33, 1561–1565 (1995).



[63] K. F. Mak, L. Ju, F. Wang, and T. F.Heinz, “Optical Sectroscopyof Graphene: From the Far Infrared to the Ultraviolet,” SolidState Comm. 152, 1341–1349 (2012).

[64] Z. Q. Li, E. A. Henriksen, Z. Jiang, Z. Hao, M. C. Martin,P. Kim, H. L. Stormer, and D. N. Basov, “Band StructureAsymmetry of Bilayer Graphene Revealed by Infrared Spec-troscopy,” Phys. Rev. Lett. 102, 037403–4 (2009).

[65] E. V. Castro, K. S. Novoselov, S. Morozov, N. M. R. Peres,J. M. B. L. dos Santos, J. Nilsson, F. Guinea, A. K. Geim,and A. H. C. Neto, “Band Structure Asymmetry of BilayerGraphene Revealed by Infrared Spectroscopy,” Phys. Rev. Lett.99, 216802–4 (2007).

[66] P. E. Gaskell, H. S. Skulason, C. Rodenchuk, and T. Szkopek,“Counting graphene layers on glass via optical reflection mi-croscopy,” Appl. Phys. Lett. 94, 143101–3 (2009).

[67] K. F. Mak, M. Y. Sfeir, Y. Wu, C. H. Lui, J. A. Misewich, andT. F. Heinz, “Measurement of the Optical Conductivity ofGraphene,” Phys. Rev. Lett. 101, 196405–4 (2008).

[68] F. J. Nelson, V. K. Kamineni, T. Zhang, E. S. Comfort, J. U. Lee,and A. C. Diebold, “Optical Properties of Large-Area Poly-crystalline Chemical Vapor Deposited Graphene by Spectro-scopic Ellipsometry,” Appl. Phys. Lett. 97, 253110–4 (2010).

[69] W. Benzinger, A. Becker, and K. J. Huttinger, “Chemistryand Kinetics of Chemical Vapour Deposition of PyrocarbonI: Fundamentals of Kinetics and Chemical Reaction Engineer-ing,” Carbon 34, 957–966 (1996).

[70] A. Becker and K. J. Huttinger, “Chemistry and Kinetics ofChemical Vapor Deposition of Pyrocarbon - II Pyrocarbon De-position From Ethylene, Acetylene and 1,3-Butadiene in theLow Temperature Regime,” Carbon 36, 177–199 (1998).


Bibliography

[71] A. Becker and K. J. Huttinger, “Chemistry and Kinetics ofChemical Vapor Deposition of Pyrocarbon - III PyrocarbonDeposition From Propylene and Benzene in the Low Temper-ature Regime,” Carbon 36, 201–211 (1998).

[72] A. Becker and K. J. Huttinger, “Chemistry and Kinetics ofChemical Vapor Deposition of Pyrocarbon - IV PyrocarbonDeposition From Methane in the Low Temperature Regime,”Carbon 36, 213–224 (1998).

[73] N. McEvoy, N. Peltekis, S. Kumar, E. Rezvani, H. Nolan, G. P.Keeley, W. J. Blau, and G. S. Duesberg, “Synthesis and Anal-ysis of Thin Conducting Pyrolytic Carbon Films,” Carbon 50,1216–1226 (2012).

[74] C. A. Taylor and W. K. Chiu, “Characterization of CVDcarbon films for hermetic optical fiber coatings,” Surface andCoatings Tech. 168, 1–11 (2003).

[75] A. P. Graham, G. Schindler, G. S. Duesberg, T. Lutz, andW. Weber, “An Investigation of the Electrical Properties ofPyrolytic Carbon in Reduced Dimensions: Vias and Wires,” J.Appl. Phys. 107, 114316–4 (2010).

[76] S. A. Campbell, The Science and Engineering of MicroelectronicFabrication (Oxford University Press, Oxford, 1996).

[77] S. De and J. N. Coleman, “Are There Fundamental Limi-tations on the Sheet Resistance and Transmittance of ThinGraphene Films?,” ACS Nano 4, 2713–2720 (2010).

[78] A. T. N’Diaye, J. Coraux, T. N. Plasa, C. Busse, and T. Michely,“Structure of epitaxial graphene on Ir(111),” New Journal ofPhysics 10, 1–16 (2008).

[79] L. Gao, W. Ren, H. Xu, L. Jin, Z. Wang, T. Ma, L.-P. Ma,Z. Zhang, Q. Fu, L.-M. Peng, X. Bao, and H.-M. Cheng,“Repeated Growth and Bubbling Rransfer of Graphene With



[63] K. F. Mak, L. Ju, F. Wang, and T. F.Heinz, “Optical Sectroscopyof Graphene: From the Far Infrared to the Ultraviolet,” SolidState Comm. 152, 1341–1349 (2012).

[64] Z. Q. Li, E. A. Henriksen, Z. Jiang, Z. Hao, M. C. Martin,P. Kim, H. L. Stormer, and D. N. Basov, “Band StructureAsymmetry of Bilayer Graphene Revealed by Infrared Spec-troscopy,” Phys. Rev. Lett. 102, 037403–4 (2009).

[65] E. V. Castro, K. S. Novoselov, S. Morozov, N. M. R. Peres,J. M. B. L. dos Santos, J. Nilsson, F. Guinea, A. K. Geim,and A. H. C. Neto, “Band Structure Asymmetry of BilayerGraphene Revealed by Infrared Spectroscopy,” Phys. Rev. Lett.99, 216802–4 (2007).

[66] P. E. Gaskell, H. S. Skulason, C. Rodenchuk, and T. Szkopek,“Counting graphene layers on glass via optical reflection mi-croscopy,” Appl. Phys. Lett. 94, 143101–3 (2009).

[67] K. F. Mak, M. Y. Sfeir, Y. Wu, C. H. Lui, J. A. Misewich, andT. F. Heinz, “Measurement of the Optical Conductivity ofGraphene,” Phys. Rev. Lett. 101, 196405–4 (2008).

[68] F. J. Nelson, V. K. Kamineni, T. Zhang, E. S. Comfort, J. U. Lee,and A. C. Diebold, “Optical Properties of Large-Area Poly-crystalline Chemical Vapor Deposited Graphene by Spectro-scopic Ellipsometry,” Appl. Phys. Lett. 97, 253110–4 (2010).

[69] W. Benzinger, A. Becker, and K. J. Huttinger, “Chemistryand Kinetics of Chemical Vapour Deposition of PyrocarbonI: Fundamentals of Kinetics and Chemical Reaction Engineer-ing,” Carbon 34, 957–966 (1996).

[70] A. Becker and K. J. Huttinger, “Chemistry and Kinetics ofChemical Vapor Deposition of Pyrocarbon - II Pyrocarbon De-position From Ethylene, Acetylene and 1,3-Butadiene in theLow Temperature Regime,” Carbon 36, 177–199 (1998).


Bibliography

[71] A. Becker and K. J. Huttinger, “Chemistry and Kinetics ofChemical Vapor Deposition of Pyrocarbon - III PyrocarbonDeposition From Propylene and Benzene in the Low Temper-ature Regime,” Carbon 36, 201–211 (1998).

[72] A. Becker and K. J. Huttinger, “Chemistry and Kinetics ofChemical Vapor Deposition of Pyrocarbon - IV PyrocarbonDeposition From Methane in the Low Temperature Regime,”Carbon 36, 213–224 (1998).

[73] N. McEvoy, N. Peltekis, S. Kumar, E. Rezvani, H. Nolan, G. P.Keeley, W. J. Blau, and G. S. Duesberg, “Synthesis and Anal-ysis of Thin Conducting Pyrolytic Carbon Films,” Carbon 50,1216–1226 (2012).

[74] C. A. Taylor and W. K. Chiu, “Characterization of CVDcarbon films for hermetic optical fiber coatings,” Surface andCoatings Tech. 168, 1–11 (2003).

[75] A. P. Graham, G. Schindler, G. S. Duesberg, T. Lutz, andW. Weber, “An Investigation of the Electrical Properties ofPyrolytic Carbon in Reduced Dimensions: Vias and Wires,” J.Appl. Phys. 107, 114316–4 (2010).

[76] S. A. Campbell, The Science and Engineering of MicroelectronicFabrication (Oxford University Press, Oxford, 1996).

[77] S. De and J. N. Coleman, “Are There Fundamental Limi-tations on the Sheet Resistance and Transmittance of ThinGraphene Films?,” ACS Nano 4, 2713–2720 (2010).

[78] A. T. N’Diaye, J. Coraux, T. N. Plasa, C. Busse, and T. Michely,“Structure of epitaxial graphene on Ir(111),” New Journal ofPhysics 10, 1–16 (2008).

[79] L. Gao, W. Ren, H. Xu, L. Jin, Z. Wang, T. Ma, L.-P. Ma,Z. Zhang, Q. Fu, L.-M. Peng, X. Bao, and H.-M. Cheng,“Repeated Growth and Bubbling Rransfer of Graphene With



Millimetre-Size Single-Crystal Grains Using Platinum,” Na-ture Comm. 3, 1–7 (2012).

[80] P. W. Sutter, J.-I. Flege, and E. A. Sutter, “Epitaxial Grapheneon Ruthenium,” Nature Mat. 7, 406–411 (2008).

[81] S.-Y. Kwon, C. V. Ciobanu, V. Petrova, V. B. Shenoy, J. Bareno,V. Gambin, I. Petrov, and S. Kodambaka, “Growth of Semi-conducting Graphene on Palladium,” Nano Lett. 9, 3985–3990(2009).

[82] K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim,J.-H. Ahn, P. Kim, J.-Y. Choi, and B. H. Hong, “Large-ScalePattern Growth of Graphene Films for Stretchable Transpar-ent Electrodes,” Nature 457, 706–710 (2009).

[83] A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, M. S.Dresselhaus, and J. Kong, “Large Area, Few-Layer GrapheneFilms on Arbitrary Substrates by Chemical Vapor Deposi-tion,” Nano Lett. 9, 30–35 (2009).

[84] X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Piner, A. Ve-lamakanni, I. Jung, E. Tutuc, S. K. Banerjee, L. Colombo, andR. S. Ruoff, “Large-Area Synthesis of High-Quality and Uni-form Graphene Films on Copper Foils,” Science 324, 1312–1314 (2009).

[85] X. Li, W. Cai, L. Colombo, and R. S. Ruoff, “Evolution ofGraphene Growth on Ni and Cu by Carbon Isotope Label-ing,” Nano Lett. 9, 4268–4272 (2009).

[86] L. Zhao, K. Rimb, H. Zhou, R. He, T. Heinz, A. Pinczuk,G. Flynn, and A. Pasupathy, “Influence of Copper Crys-tal Surface on the CVD Growth of Large Area MonolayerGraphene,” Solid State Comm. 151, 509–513 (2011).

[87] A. M. Lewis, B. Derby, and I. A. Kinloch, “Influence ofGas Phase Equilibria on the Chemical Vapor Deposition ofGraphene,” ACS Nano 7, 3104–3117 (2013).


Bibliography

[88] N. K. Memon, S. D. Tse, M. Chhowalla, and B. H. Kear, “Roleof Substrate, Temperature, and Hydrogen on the Flame Syn-thesis of Graphene Films,” Proc. Combustion Institute 34, 2163–2170 (2013).

[89] Z. Yan, J. Lin, Z. Peng, Z. Sun, Y. Zhu, L. Li, C. Xiang, E. L.Samuel, C. Kittrell, and J. M. Tour, “Toward the Synthesis ofWafer-Scale Single-Crystal Graphene on Copper Foils,” ACSNano 6, 9110–9117 (2012).

[90] C. M. Orofeoa, H. Hibino, K. Kawahara, Y. Ogawa, M. Tsuji,K. ichi Ikeda, S. Mizuno, and H. Ago, “Influence of Cu Metalon the Domain Structure and Carrier Mobility in Single-LayerGraphene,” Carbon 50, 2189–2196 (2012).

[91] W. Cai, Y. Zhu, X. Li, R. D. Piner, and R. S. Ruoff, “LargeArea Few-Layer Graphene/Graphite Films as TransparentThin Conducting Electrodes,” Appl. Phys. Lett. 95, 123115–4(2009).

[92] Y. Yao, Z. Li, Z. Lin, K.-S. Moon, J. Agar, and C. Wong, “Con-trolled Growth of Multilayer, Few-Layer, and Single-LayerGraphene on Metal Substrates,” J. Phys. Chem. C 115, 5232–5238 (2011).

[93] Y. Yao and C. ping Wong, “Monolayer Graphene GrowthUsing Additional Etching Process in Atmospheric PressureChemical Vapor Deposition,” Carbon 50, 5203–5209 (2012).

[94] J. Kang, D. Shin, S. Baea, and B. H. Hong, “Graphene Trans-fer: Key for Applications,” Nanoscale 4, 5527–5537 (2012).

[95] X. Li, Y. Zhu, W. Cai, M. Borysiak, B. Han, D. Chen, R. D.Piner, L. Colombo, and R. S. Ruoff, “Transfer of Large-AreaGraphene Films for High-Performance Transparent Conduc-tive Electrodes,” Nano Lett. 9, 4359–4363 (2009).

[96] A. Ismach, C. Druzgalski, S. Penwell, A. Schwartzberg,M. Zheng, A. Javey, J. Bokor, and Y. Zhang, “Direct Chemical



Millimetre-Size Single-Crystal Grains Using Platinum,” Na-ture Comm. 3, 1–7 (2012).

[80] P. W. Sutter, J.-I. Flege, and E. A. Sutter, “Epitaxial Grapheneon Ruthenium,” Nature Mat. 7, 406–411 (2008).

[81] S.-Y. Kwon, C. V. Ciobanu, V. Petrova, V. B. Shenoy, J. Bareno,V. Gambin, I. Petrov, and S. Kodambaka, “Growth of Semi-conducting Graphene on Palladium,” Nano Lett. 9, 3985–3990(2009).

[82] K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim,J.-H. Ahn, P. Kim, J.-Y. Choi, and B. H. Hong, “Large-ScalePattern Growth of Graphene Films for Stretchable Transpar-ent Electrodes,” Nature 457, 706–710 (2009).

[83] A. Reina, X. Jia, J. Ho, D. Nezich, H. Son, V. Bulovic, M. S.Dresselhaus, and J. Kong, “Large Area, Few-Layer GrapheneFilms on Arbitrary Substrates by Chemical Vapor Deposi-tion,” Nano Lett. 9, 30–35 (2009).

[84] X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, R. Piner, A. Ve-lamakanni, I. Jung, E. Tutuc, S. K. Banerjee, L. Colombo, andR. S. Ruoff, “Large-Area Synthesis of High-Quality and Uni-form Graphene Films on Copper Foils,” Science 324, 1312–1314 (2009).

[85] X. Li, W. Cai, L. Colombo, and R. S. Ruoff, “Evolution ofGraphene Growth on Ni and Cu by Carbon Isotope Label-ing,” Nano Lett. 9, 4268–4272 (2009).

[86] L. Zhao, K. Rimb, H. Zhou, R. He, T. Heinz, A. Pinczuk,G. Flynn, and A. Pasupathy, “Influence of Copper Crys-tal Surface on the CVD Growth of Large Area MonolayerGraphene,” Solid State Comm. 151, 509–513 (2011).

[87] A. M. Lewis, B. Derby, and I. A. Kinloch, “Influence ofGas Phase Equilibria on the Chemical Vapor Deposition ofGraphene,” ACS Nano 7, 3104–3117 (2013).


Bibliography

[88] N. K. Memon, S. D. Tse, M. Chhowalla, and B. H. Kear, “Roleof Substrate, Temperature, and Hydrogen on the Flame Syn-thesis of Graphene Films,” Proc. Combustion Institute 34, 2163–2170 (2013).

[89] Z. Yan, J. Lin, Z. Peng, Z. Sun, Y. Zhu, L. Li, C. Xiang, E. L.Samuel, C. Kittrell, and J. M. Tour, “Toward the Synthesis ofWafer-Scale Single-Crystal Graphene on Copper Foils,” ACSNano 6, 9110–9117 (2012).

[90] C. M. Orofeoa, H. Hibino, K. Kawahara, Y. Ogawa, M. Tsuji,K. ichi Ikeda, S. Mizuno, and H. Ago, “Influence of Cu Metalon the Domain Structure and Carrier Mobility in Single-LayerGraphene,” Carbon 50, 2189–2196 (2012).

[91] W. Cai, Y. Zhu, X. Li, R. D. Piner, and R. S. Ruoff, “LargeArea Few-Layer Graphene/Graphite Films as TransparentThin Conducting Electrodes,” Appl. Phys. Lett. 95, 123115–4(2009).

[92] Y. Yao, Z. Li, Z. Lin, K.-S. Moon, J. Agar, and C. Wong, “Con-trolled Growth of Multilayer, Few-Layer, and Single-LayerGraphene on Metal Substrates,” J. Phys. Chem. C 115, 5232–5238 (2011).

[93] Y. Yao and C. ping Wong, “Monolayer Graphene GrowthUsing Additional Etching Process in Atmospheric PressureChemical Vapor Deposition,” Carbon 50, 5203–5209 (2012).

[94] J. Kang, D. Shin, S. Baea, and B. H. Hong, “Graphene Trans-fer: Key for Applications,” Nanoscale 4, 5527–5537 (2012).

[95] X. Li, Y. Zhu, W. Cai, M. Borysiak, B. Han, D. Chen, R. D.Piner, L. Colombo, and R. S. Ruoff, “Transfer of Large-AreaGraphene Films for High-Performance Transparent Conduc-tive Electrodes,” Nano Lett. 9, 4359–4363 (2009).

[96] A. Ismach, C. Druzgalski, S. Penwell, A. Schwartzberg,M. Zheng, A. Javey, J. Bokor, and Y. Zhang, “Direct Chemical



Vapor Deposition of Graphene on Dielectric Surfaces,” NanoLett. 10, 1542–1548 (2010).

[97] M. H. Rummeli, A. Bachmatiuk, A. Scott, F. Borrnert,J. H. Warner, V. Hoffman, J.-H. Lin, G. Cuniberti, andB. Buchner, “Direct Low-Temperature Nanographene CVDSynthesis over a Dielectric Insulator,” ACS Nano 4, 4206–4210(2010).

[98] C.-Y. Su, A.-Y. Lu, C.-Y. Wu, Y.-T. Li, K.-K. Liu, W. Zhang,S.-Y. Lin, Z.-Y. Juang, Y.-L. Zhong, F.-R. Chen, and L.-J. Li,“Direct Formation of Wafer Scale Graphene Thin Layers onInsulating Substrates by Chemical Vapor Deposition,” NanoLett. 11, 3612–3616 (2011).

[99] L. Zhang, Z. Shi, Y. Wang, R. Yang, D. Shi, and G. Zhang,“Catalyst-Free Growth of Nanographene Films on VariousSubstrates,” Nano Research 4, 315–321 (2011).

[100] H.-J. Shin, W. M. Choi, S.-M. Yoon, G. H. Han, Y. S. Woo, E. S.Kim, S. J. Chae, X.-S. Li, A. Benayad, D. D. Loc, F. Gunes,Y. H. Lee, and J.-Y. Choi, “Transfer-Free Growth of Few-Layer Graphene by Self-Assembled Monolayers,” Adv. Mat.23, 4392–4397 (2011).

[101] J. Kwak, J. H. Chu, J.-K. Choi, S.-D. Park, H. Go, S. Y. Kim,K. Park, S.-D. Kim, Y.-W. Kim, E. Yoon, S. Kodambaka, andS.-Y. Kwon, “Near Room-Temperature Synthesis of Transfer-Free Graphene Films,” Nature Comm. 3, 1–7 (2012).

[102] D. G. Gromov and S. A. Gavrilov, “Manifestation of the Het-erogeneous Mechanism Upon Melting of Low-DimensionalSystems,” Physics of the Solid State 51, 2135–2144 (2009).

[103] D. Myers, Surfaces, Interfaces, and Colloids: Principles and Ap-plications (John Wiley & Sons, New York, 1999).

[104] C. S. Jayanathi, E. Tosatti, and L. Pietronero, “Surface Meltingof Copper,” Phys. Rev. B 31, 3456–3459 (1985).


[105] J. J. Spivey, S. K. Agarwal, H. Knozinger, and E. Taglauer,Catalysis (Royal Society of Chemistry, Cambridge, 1993).

[106] G. C. Kuczinski, A. E. Miller, and G. A. Sargent, Sinteringand Heterogeneous Catalysis: Material Science Research Vol. 46(Plenum Pub. Corp., New York, 1984).

[107] A. A. Balandin, S. Ghosh, W. Bao, I. Calizo, D. Teweldebrhan,F. Miao, and C. N. Lau, “Superior Thermal Conductivity ofSingle-Layer Graphene,” Nano Lett. 8, 902–907 (2008).

[108] I. Calizo, A. A. Balandin, W. Bao, F. Miao, and C. N. Lau,“Temperature Dependence of the Raman Spectra of Grapheneand Graphene Multilayers,” Nano Lett. 7, 2645–2649 (2007).

[109] S. Chen, L. Brown, M. Levendorf, W. Cai, S.-Y. Ju, J. Edge-worth, X. Li, C. W. Magnuson, A. Velamakanni, R. D. Piner,J. Kang, J. Park, and R. S. Ruoff, “Oxidation Resistance ofGraphene-Coated Cu and Cu/Ni Alloy,” ACS Nano 5, 1321–1327 (2011).

[110] M. Schriver, W. Regan, W. J. Gannett, A. M. Zaniewski, M. F.Crommie, and A. Zettl, “Graphene as a Long-Term MetalOxidation Barrier: Worse Than Nothing,” ACS Nano 7, 5763–5768 (2013).

[111] K. L. Chavez and D. W. Hess, “A Novel Method of EtchingCopper Oxide Using Acetic Acid,” Appl. Phys. Lett. 148, G640–G643 (2001).

[112] D. K. Schroder, Semiconductor Material and Device Characteri-zation (John Wiley & Sons, New Jersey, 2006).

[113] E. M. Doherty, S. De, P. E. Lyons, A. Shmeliov, P. N. Nirmalraj,V. Scardaci, J. Joimel, W. J. Blau, J. J. Boland, and J. N. Cole-man, “The spatial uniformity and electromechanical stabilityof transparent, conductive films of single walled nanotubes,”Carbon 47, 2466–2473 (2009).


Vapor Deposition of Graphene on Dielectric Surfaces,” NanoLett. 10, 1542–1548 (2010).

[97] M. H. Rummeli, A. Bachmatiuk, A. Scott, F. Borrnert,J. H. Warner, V. Hoffman, J.-H. Lin, G. Cuniberti, andB. Buchner, “Direct Low-Temperature Nanographene CVDSynthesis over a Dielectric Insulator,” ACS Nano 4, 4206–4210(2010).

[98] C.-Y. Su, A.-Y. Lu, C.-Y. Wu, Y.-T. Li, K.-K. Liu, W. Zhang,S.-Y. Lin, Z.-Y. Juang, Y.-L. Zhong, F.-R. Chen, and L.-J. Li,“Direct Formation of Wafer Scale Graphene Thin Layers onInsulating Substrates by Chemical Vapor Deposition,” NanoLett. 11, 3612–3616 (2011).

[99] L. Zhang, Z. Shi, Y. Wang, R. Yang, D. Shi, and G. Zhang,“Catalyst-Free Growth of Nanographene Films on VariousSubstrates,” Nano Research 4, 315–321 (2011).

[100] H.-J. Shin, W. M. Choi, S.-M. Yoon, G. H. Han, Y. S. Woo, E. S.Kim, S. J. Chae, X.-S. Li, A. Benayad, D. D. Loc, F. Gunes,Y. H. Lee, and J.-Y. Choi, “Transfer-Free Growth of Few-Layer Graphene by Self-Assembled Monolayers,” Adv. Mat.23, 4392–4397 (2011).

[101] J. Kwak, J. H. Chu, J.-K. Choi, S.-D. Park, H. Go, S. Y. Kim,K. Park, S.-D. Kim, Y.-W. Kim, E. Yoon, S. Kodambaka, andS.-Y. Kwon, “Near Room-Temperature Synthesis of Transfer-Free Graphene Films,” Nature Comm. 3, 1–7 (2012).

[102] D. G. Gromov and S. A. Gavrilov, “Manifestation of the Het-erogeneous Mechanism Upon Melting of Low-DimensionalSystems,” Physics of the Solid State 51, 2135–2144 (2009).

[103] D. Myers, Surfaces, Interfaces, and Colloids: Principles and Ap-plications (John Wiley & Sons, New York, 1999).

[104] C. S. Jayanathi, E. Tosatti, and L. Pietronero, “Surface Meltingof Copper,” Phys. Rev. B 31, 3456–3459 (1985).


[105] J. J. Spivey, S. K. Agarwal, H. Knozinger, and E. Taglauer,Catalysis (Royal Society of Chemistry, Cambridge, 1993).

[106] G. C. Kuczinski, A. E. Miller, and G. A. Sargent, Sinteringand Heterogeneous Catalysis: Material Science Research Vol. 46(Plenum Pub. Corp., New York, 1984).

[107] A. A. Balandin, S. Ghosh, W. Bao, I. Calizo, D. Teweldebrhan,F. Miao, and C. N. Lau, “Superior Thermal Conductivity ofSingle-Layer Graphene,” Nano Lett. 8, 902–907 (2008).

[108] I. Calizo, A. A. Balandin, W. Bao, F. Miao, and C. N. Lau,“Temperature Dependence of the Raman Spectra of Grapheneand Graphene Multilayers,” Nano Lett. 7, 2645–2649 (2007).

[109] S. Chen, L. Brown, M. Levendorf, W. Cai, S.-Y. Ju, J. Edge-worth, X. Li, C. W. Magnuson, A. Velamakanni, R. D. Piner,J. Kang, J. Park, and R. S. Ruoff, “Oxidation Resistance ofGraphene-Coated Cu and Cu/Ni Alloy,” ACS Nano 5, 1321–1327 (2011).

[110] M. Schriver, W. Regan, W. J. Gannett, A. M. Zaniewski, M. F.Crommie, and A. Zettl, “Graphene as a Long-Term MetalOxidation Barrier: Worse Than Nothing,” ACS Nano 7, 5763–5768 (2013).

[111] K. L. Chavez and D. W. Hess, “A Novel Method of EtchingCopper Oxide Using Acetic Acid,” Appl. Phys. Lett. 148, G640–G643 (2001).

[112] D. K. Schroder, Semiconductor Material and Device Characteri-zation (John Wiley & Sons, New Jersey, 2006).

[113] E. M. Doherty, S. De, P. E. Lyons, A. Shmeliov, P. N. Nirmalraj,V. Scardaci, J. Joimel, W. J. Blau, J. J. Boland, and J. N. Cole-man, “The spatial uniformity and electromechanical stabilityof transparent, conductive films of single walled nanotubes,”Carbon 47, 2466–2473 (2009).


Publications of the University of Eastern Finland

Dissertations in Forestry and Natural Sciences

isbn 978-952-61-1303-6 (printed)

issnl 1798-5668

issn 1798-5668

isbn 978-952-61-1304-3 (pdf)

issn 1798-5676

Tommi Kaplas

Synthesis and Optical Characterization of Ultrathin Carbon Films

This work is devoted on experimental

synthesis and optical characterization

of ultrathin carbon films. Specifically

the main focus of this thesis is in two

carbon species with dominating sp2

hybridization. One of these materials is

pyrolytic carbon, which is well known,

amorphous carbon material but is only

very recently been synthetized as an

ultrathin film. Second investigated ma-

terial is crystalline, a single and a few

layered graphene, which has recently

drawn a lot of interest of scientific com-

munity. The synthesis is carried out by

using chemical vapor deposition (CVD)

techniques while the optical character-

ization is done by utilizing spectropho-

tometry and Raman spectroscopy.

disser

tation

s | 130 | To

mm

i Ka

plas | S

ynth

esis an

d O

ptica

l Ch

ara

cterizatio

n of U

ltrath

in C

arb

on

Film

s

Tommi KaplasSynthesis and Optical

Characterization of Ultrathin Carbon Films

Documents

Tommi Kaplas - UEF...Tommi Kaplas Synthesis and Optical Characterization of Ultrathin Carbon Films This work is devoted on experimental synthesis and optical characterization of ultrathin