Direct Growth Of Graphitic Carbon Layers On Fused Silica Substrate Using Nickel Thin Film Catalyst

(1)

Direct Growth Of

Graphitic Carbon Layers On Fused Silica Substrate Using

Nickel Thin Film Catalyst

Anusmita Addy

Master Thesis May 2017

Department of Physics and Mathematics

University of Eastern Finland

(2)

Anusmita Addy Direct Growth of Graphitic Carbon Layers, 88 pages University of Eastern Finland

Master’s Degree Programme in Photonics Supervisors Dr. Tommi Kaplas

Prof. Dr. Yuri Svirko

Abstract

In this thesis work the main aim is to synthesize graphitic layers directly on a dielectric substrate and characterize their electrical and optical properties. For direct synthesis, the chemical vapor deposition process is done at the temperature800^◦C on fused silica substrate developed with metallic nickel thin ﬁlm layer and further with a layer of negative tone resist polymer nLOF. The metallic nickel layer acts as a catalyst while the polymer nLOF is used as the carbon precursor during the process.

Post the CVD process, the nickel particles are etched out. Twenty such samples are prepared with diﬀerent thickness layers of nickel and nLOF. The samples are characterized by the Raman spectrometer and the scanning electron microscope. They showed results of successful synthesis of graphitic carbon layers. Also, the sheet resistance and absorbance of the samples are measured in order to determine the synthesized layers’ electrical and optical properties respectively.

Further, as an extension of the main work, the substrate with best synthesized graphitic carbon layers, corresponding to a particular nickel thin ﬁlm and resist nLOF layer thickness, is developed again and a micro-grating structure is fabri- cated by electron beam lithography and then, synthesis and characterization of the sample is done with the aforesaid method and processes respectively.

(3)

Preface

It is my immense pleasure to present my Master’s thesis onGraphitic Carbon Layers On Fused Silica Substrate Using Nickel Thin Film Catalyst. The sole purpose and motivation for this thesis work and my Master’s study in Photonics was to quench my thirst of pursuing higher academic career in this ﬁeld.

Here, I would like to express my deep sense of gratitude to my supervisor Dr.

Tommi Kaplas for his active supervision, hands-on training in the laboratories, timely guidance, and constant support and motivation at every step of this thesis work. I also forward my sincere thanks to Prof. Dr. Yuri Svirko, my second supervisor, for his valuable feedbacks and outlook during the progress of the work and encouragement to accomplish the thesis aims. I would also like to extend my regards to our course Coordinator Noora Heikkilä for providing all the important information, assistance and guidelines throughout the entire period from application to completion of the Master degree. I am obliged to the Department of Physics and Mathematics, University of Eastern Finland, for providing the requisite facilities for the thesis work and also for providing sufficient financial support during my entire Master’s studies.

I would further like to thank my colleague Marian Baah for her help in providing valuable literature, articles and suggestions for my thesis and also deﬁnitely, for sharing quite many light and crazy moments of laughter and chit-chats in between long working hours, without which the thesis would not stood by its present position. Special appreciation to my friends Atri, Rajannya and Tanmay, without whom staying alone away from home would not have been so easy, fun and enjoyable.

Warmest thanks to my seniors and all other persons who have directly or indirectly helped me count on them in whichever way possible during all ups and downs since September,2015 at Joensuu,Finland.

And, last but not the least, my heartfelt thanks to my husband Shiladitya, my parents and in-laws for their constant positivity, determination, kindness and uncon- ditional love during moments of breakdown and above all for being there throughout, at every point till the very end of my accomplishment of my goals.

Thank you all,

Joensuu, the 22th of May 2017 Anusmita Addy

(4)

Chapter I

Introduction

Be it metals like gold, silver, copper in pre-historic era or be it semiconductor like silicon in the nineteenth century, discovery of any new material have always sparked enormous curiosity, and research on them became the ground for all other discover- ies in science and technology. One of the earliest element discovered on earth and that which forms the basis of all living beings and organic compounds is the element carbon. Carbon has always fascinated researchers with its unique properties and it continues to do so till date. Carbon has the maximum number of allotropes in comparison to any other elements, with its crystalline forms diamond and graphite being the most important ones due to their unique properties [1]. However, it was only just a little more than a decade back, that its thinnest and strongest allotrope

‘graphene’, a monolayer of graphite, was isolated and discovered by Professor An- dre Geim and Dr Kostya Novoselov, in 2004 at the University of Manchester in England [2].

The history of scientiﬁc research in graphitic carbon goes back to 1859 when highly lamellar structure of thermally reduced graphite oxide was discovered by B. C.

Brodie [3, 4]. Next, in 1916, by powder diffraction method, the structure of graphite was first identified [5,6] and later determined in 1924 by single-crystal diffraction [7].

It was followed by the theoretical study of the band structure of graphite by Wallace in 1947 [8] and the very next year G. Reuss and F. Vogt published TEM images of a few layer graphite [9]. The term ‘graphene’, which is a combination of the word

‘graphite’ with the suﬃx ‘-ene’ meaning one, was ﬁrst coined by Hanns Peter Boehm to describe single-layer of carbon foils in the year 1962 [10,11]. Research on epitaxial growth of single layers of graphite on materials took stage from the 1970s [12] while to

(8)

obtain thin ﬁlms of graphite by mechanical exfoliation technique began in 1990s [13].

Eventually, what seemed impossible as made possible by Geim and Novoselov and hence they were awarded the Nobel Prize in 2010 [14]. The discovery indeed opened a new chapter in the ﬁeld of solid state and condensed matter physics while its unique electrical and optical properties gave new dimensions to the electronic and opto-electronic technology world.

Although high quality monolayer graphene was extracted and isolated, the method of mechanical exfoliation or the “scotch-tape” method used by Geim and Novoselov was not well suited for commercial applications. Hence research groups around the world focussed themselves with the task of synthesizing scalable yet high quality graphene that may be incorporated in large scale commercial applications of fu- ture electronic and optoelectronic devices [15]. Single to few layers graphene can be synthesized using diﬀerent techniques like micromechanical cleaving, desorption of silicon (Si) from silicon carbide (SiC), reduction of graphite oxide and chemical vapour deposition (CVD). Amongst all the prevalent method of fabrication, CVD is most widely implemented for its capability of large scale graphene production in industry [16]. But, the conventional CVD technique involves the synthesis of graphene on transition metal catalyst while practical applications demand the same on dielectric or semiconductor substrates. Hence, synthesized graphene layers need to be transferred and this is the ultimate stumbling block in the advancement of graphene-based technology. As it is understood, transferring an atom-thick layer ﬁlm without damaging or contaminating it is a very complicated process [17]. Here, it may be appropriate to mention that recent studies show that synthesis of few layers of graphene (FLG) or graphitic carbon layers have quite drawn attention because of comparable or in few cases more promising properties, along with being robust and easier handling than single layer of graphene (SLG) [18–26].

In this thesis, the main focus is based on synthesis of graphitic carbon layers directly on a dielectric substrate from a solid carbon precursor using catalytic metal thin ﬁlm. The structure of the thesis is divided into eight chapters. This chapter is followed by a brief theoretical background on carbon allotrope graphite and its 2D structure graphene along with review of their electrical and optical properties.

Simultaneously, the chapter also deals with the illustration of the two CVD synthesis methods- the most widespread transfer process and corresponding direct process used in this thesis work. In Chapter 3, a descriptive outline on the various process

(9)

setups and measurement devices is presented. The detailed experimental workings and consequent results obtained are put into Chapter 4 and Chapter 5 respectively.

Chapter 6 of the thesis deals with the fabrication of a self-assembled graphitic micro- grating structure with electron beam lithography and the direct method of synthesis used in this thesis work. It speciﬁcally comprises of a brief theory on gratings and lithography in optics, along with the experimental procedure and results. Next, the penultimate chapter provides the discussions on the thesis work, its experimental processes, results and probable applications while the thesis ends with a short conclusion in the last chapter.

(10)

Chapter II

Theoretical Background

Carbon is the sixth element in the periodic table with symbol C. It is non-metallic and its atom has four electrons in the outer shell, each of which covalently binds to the electron of other atoms. It has seven isotopes among which ¹²C and ¹³C are stable while others are radioactive [1]. Carbon in nature can be found in various forms known as its allotropes, having extremely diﬀerent properties from one another while its combination with other elements specially hydrogen forms the huge and complex branch of organic chemistry. Graphite which is one of the soft- est (with Mohs hardness scale 1 or 2), highly conducting material and opaque to visible light, diamond which on the other hand is the hardest-known (with Mohs hardness scale 10), electrical insulator and is transparent to visible light, while soot which is obtained as a result of any combustion process, are all carbon at the core.

Though carbon can exists in enormous number of forms, naturally it occurs widely as three basic allotropes- diamond and graphite as crystalline form and charcoal in amorphous form. These forms diﬀer from one another due the diﬀerent and unique arrangements of their basic constituent element, meaning that the carbon atomic bonds are distinct in every case. Thus, it is well understood that carbon in itself is a huge domain, however here, we are interested in one of its allotrope- graphite and its various forms.

2.1 Carbon Allotropes: From Graphite to Graphene

The six electrons in a carbon atom are arranged into two shells namely the K and the L shells with orbitals 1s, 2s and 2p. The two innermost electrons placed in the 1sorbital are strongly attached to the atomic nucleus and does not take part in any

(11)

chemical bonding. The other four electrons are arranged into 2s and 2p orbitals.

In the ground state, two electrons with opposite spins are placed in the 2s orbital while the other two are placed in the 2p orbital which is further divided into three sub-orbitals namely 2px, 2py and 2pz. The two electrons are individually placed in 2px and 2py with parallel spin while the other sub-orbital remains empty. In the ground state, elemental carbon is thus bivalent. However, when a carbon atom is excited, one electron from 2s jumps to the 2pz orbital. Hence, the orbitals 2s, 2px, 2py and 2pz each contain one electron and the orbitals are said to be hybridized.

The hybridized orbital is represented as spⁿ where n = 1,2,3 denotes the number of p sub-orbital taking part in the hybrid bond. The following ﬁgure 2.1 shows the electron conﬁguration of carbon atom in the ground and excited states.

Figure 2.1: Electron-orbital conﬁguration of a carbon atom.

On the basis of these hybridized states diﬀerent allotropes of carbon are formed.

The tetragonalsp³ bonds form diamond, diagonalsp bonds form types of fullerenes while the trigonalsp² bonds form the graphite [1].

2.1.1 Structure of Graphite and Graphene

Graphite is thus, an allotrope of carbon having sp² hybridized atoms. In these atoms, the three sp² hybrid orbitals lie in a plane at an angular diﬀerence of 120^◦ while the lone p_z orbital is oriented in a plane perpendicular to the former. Now,

(12)

in two neighbouring sp² hybridized carbon atoms onesp² orbital from each forms a σ bond and the pz orbital from each forms a π bond, together resulting in a strong σ−π double bond. Due to the strong C=C bond, the sp² orbitals cannot rotate and thus leads to a hexagonal planar lattice structure formation where each atom is surrounded by three adjacent ones. Also, the pz delocalized electrons give birth to highly polarized electrical conductivity within such materials [1, 27]. The ﬁgure 2.2 shows the lattice structure of the graphite.

Figure 2.2: Structure of graphite.

Graphene is ideally this single layer planar honeycomb structure ofsp²hybridized carbon atoms. Hence, graphite is an enormous number of stacked graphene layers.

The inter-atomic distance in the hybridized plane is 1.42 A, while the inter-planar^◦ distance is 3.34A which are linked by the weak van der Waal’s force [2, 27]. There-^◦ fore, from 3D graphite to 2D graphene to 1D nanotube to 0D fullerenes, all belong to the sp² hybridized carbon allotrope group ‘graphite’. Interestingly, all the above forms maybe obtained from the 2D graphene, which is why it is called the mother of all graphitic structures (see ﬁgure 2.3) [14].

(13)

Figure 2.3: Figure showing graphene as the mother of all graphitic structures: (a) wrapped up into 0D bucky balls, (b) rolled up to 1D nanotubes and (c) stacked up to 3D graphite (Copyrights 2007 Nature Publishing Group [14]).

2.1.2 Properties: Graphite to Graphene

Post its discovery, researchers emphasized themselves on studying the properties of graphene. In only a very short period however, the material’s properties managed to position itself in superlatives. Graphene at the same time is the thinnest, lightest yet strongest and almost impermeable material, it has the highest heat conductivity and may show superconductivity at room temperature [15]. Along with, it also has fascinating electrical and optical properties which are considered brieﬂy in the following sections.

2.1.2.1 Electrical Properties

Depending on how strongly the outer valence electrons are bonded to the nucleus in any atom, a material is classiﬁed as an insulator or metal. In metals, the valence electrons are very weakly bonded to the atomic nucleus and thus, can move readily inside the lattice. Since, electricity is the ﬂow of charge particles like electrons, hence

(14)

metals are very good conductors of electricity. The exact opposite case can describe the insulators while semi-conductors lie somewhere in between.

From the graphite structure discussed in the above section, it is seen that graphite has lone unhybridizedpz electron weakly bonded to the nucleus. These electrons can move freely along the basal plane of the crystal. Thus, graphite maybe considered as a semi-metal having high conductivity in the basal plane while in a plane normal to it, it acts as an insulator [1].

In case of graphene, which is a single layer of graphite, each lattice point has one free electron (unhybridized pz lone one) that can move freely within the crystal plane. The mean free path of the electrons is in the order of microns, resulting in high conductivity of the two-dimensional material. Intrinsic graphene is a zero band gap material where the top of the valence band touch the bottom of the conduction band at the Dirac points. The energy of the electrons/holes at these points is found to be equal to the Fermi energy and it also varies linearly with the momentum i.e.

it follows the Dirac E−k relation (see ﬁgure 2.4). Hence, at these points the charge carriers in a graphene behave like massless relativistic Dirac fermion particles [28].

This is particularly interesting because the visible and infrared frequency range lies near the Dirac points. On the other hand, if one compares with the graphite band structure, the E−k relation is non-linear because it is shifted from the Dirac points and lie at some intermediate position [8]. For a few layer graphene, the bands overlap at the vicinity of the Dirac points. This diﬀerence has particular signiﬁcance in the terahertz and far-infrared regions.

Figure 2.4: Figure showing graphene band structure(Copyrights 2012 IOP Publish- ing Ltd. [29]).

(15)

However, band gap is present in graphene with two or more layers. The value of the gap depends on the mutual interaction of the layers which is again dependent on the number of layers of graphene as well as the type of stacking of the layers [30]. The strong bonds between the planar and interplanar carbon molecules contributes to the high electronic and thermal stability of graphene. Hence, both single to multilayer graphene or graphitic carbon layers have high potential applications in electronics, sensors, memory chips and opto-electronic devices.

2.1.2.2 Optical Properties

The electronic properties of graphite and graphene contribute widely to their optical properties as well. Specially, the latter is found to possess quite fascinating optical characteristics. The linear optical properties are determined mostly by the transitions between the energy bands. In intrinsic monolayer graphene, there is no band gap and the energy bands are not parabolic. And uniquely, the point of inter- section of the bands coincide with the Fermi energy. As there is zero energy gap, it is optical absorption that predominates at all frequencies in the lower range of the spectrum [31]. The absorbance of monolayer graphene is found to be approximately 2.3% at visible frequencies which means that the transmittance of graphene is independent over the wide visible range of the spectrum [32]. Interestingly, it can be expressed by a simple formula combining natural constants as:

A =πα (2.1)

where α = e²/4π0c is known as the ﬁne structure constant. Here, 0 denotes the free space permittivity, e the electron charge,is the reduced Planck’s constant and c is the speed of light [31].

Depending on the purpose, this value can be manifested both as being high or low. It means that for application in photodetectors a few layers of graphene can be used due to high opacity [33] while for application of transparent electrodes with high conductivity a single layer can be implemented [34].

Optical properties of thin ﬁlms are usually described in terms of optical conductivity, which depends on temperature, frequency and carrier concentration [35].

However, in the visible spectrum the optical conductivity of graphene is given by the following equation as:

(16)

G = αc

4 (2.2)

From the equation 2.2 above it can be seen that the optical conductivity within the visible range is independent of the wavelength which signify that the optical measurable parameters transmittance and reﬂectance are constants too. For normal incidence, they are given respectively in equations 2.3 and 2.4 as (from [36]):

T =

1 + πα 2

₋2

(2.3)

R = π²α²T

4 (2.4)

The above equations are valid only if the carrier collision rate is much smaller than the frequency and spatial distribution. In order to generalize the optical properties of graphene over any frequency, one needs to study the dependence of the optical conductivity as a function of temperature, frequency and chemical potential. To obtain the same, graphene maybe considered as a semiconductor quantum well and consequent detail mathematical derivations can be performed [31, 35]. Also, the above equations are valid for suspended graphene as otherwise these are aﬀected by the substrate refractive index [37]. The transmittance of the graphene ﬁlm deposited on a substrate,at normal incidence is described by the Fresnel equation as [38]:

T_gs = 4n_s(1 + n_s+πα)⁻² (2.5) where n_s is the refractive index of the dielectric substrate. Similarly, the reﬂectance of a graphene coated substrate is given the Fresnel equation as:

R(h)_gs = r₁₂+ r₂₃exp(2iβ) 1 + r₁₂r₂₃exp(2iβ)

² (2.6)

where at normal incidence r₁₂= (1−n_g)/(1 + n_g) and r₂₃= (n_g −n_s)/n_g + n_s are reﬂection coeﬃcients of the vacuum-graphene and graphene-substrate interfaces respectively andβ = 2πn_gh/λ. As seen, the above equation is expressed as a function

(17)

of the graphene film thickness h. The refractive index of graphene denoted by n_g is well defined experimentally [39]. Following equation 2.5 and 2.6, experimental research gave results of absorbance 2.2% of graphene placed over quartz substrate at 550 nm [40]. The overall optical property of SLG to FLG is shown schematically in the figure 2.5 given below.

Figure 2.5: Figure showing (A) image of a 50 mm aperture partially covered by graphene mono- and bilayer. The scan proﬁle shows the intensity of transmitted white light along the yellow line. (Inset) The sample design: A 20 mm thick metal support structure having apertures 20, 30 and 50 mm in diameter with graphene placed over them. (B) Transmittance spectrum of single layer graphene represented (shown by open circles). It is seen that transmittance reduce slightly forλ <500 nm (probably caused by hydrocarbon contamination) [12]. The red line represents the transmittance given by equation 2.3, which is expected for two-dimensional Dirac fermions, whereas the green curve provides nonlinearity and triangular warping of graphene’s electronic spectrum. Standard error in measurements is shown by the grey area. (Inset) It shows the transmittance of white light as a function of number of graphene layers. The reduction in intensity by a factor πα is represented by corresponding dashed lines for each graphene layer added. (Copyright 2008 Science Express [41]).

(18)

2.1.3 Synthesis: Chemical Vapour Deposition

From the discovery of graphene, as mentioned in the introduction, quite many method of its synthesis have been developed over the past decade. Among them, the most significant methods are the mechanical exfoliation of graphite and the synthesis over metallic foil (mostly copper) by chemical vapor deposition followed by transfer of synthesized graphene layer/layers onto a substrate. In the first method, though the quality of graphene prepared is very high with almost no defects along with low complexity of the method, it has low reproducibility and lack of controllability. On the other hand, the second method is more commercially well pronounced and with minimum compromise on the quality of graphene, overcomes the drawbacks of the former method. However, the complexity of this method is very high and the main hindrance lies in transferring the graphene film over the substrate. Many a times the process of transferring becomes more difficult than the process of synthesising the graphene itself, resulting in damaging or destroying the synthesized layer/layers partially or completely. The following sections briefly outline the idea of the transfer process of graphene synthesis followed by the direct method of synthesis, which is focussed on overcoming the loopholes of the transfer method without compromising the quality of the graphene produced.

2.1.3.1 Transfer Method

In this section let us review brieﬂy about the CVD transfer technique for synthesis, which is so far the most established procedure to produce medium to high quality of the material. Mono to multilayer graphene are produced onto transition metal foil substrates such as nickel (Ni), copper (Cu), palladium (Pd), ruthenium (Ru) or iridium (Ir) [17]. The metal acts as catalyst and the choice depend on its capability to dehydrogenate the hydrocarbon precursor completely as well as solubility of carbon by the metal [42]. Also, it depends on the type of chemical vapour deposition process namely, either plasma enhanced CVD or thermal CVD. Plasma enhanced CVD was ﬁrst demonstrated by J. Wang et al. in the same year as the discovery of the material [43]. Later, thermal CVD was used to synthesize graphene and it became more prevalent and either copper or nickel is predominantly used as the catalytic metal. Early and widespread researches on graphene synthesis by thermal CVD was mostly done with gaseous precursors like methane (CH₄) or other similar

(19)

hydrocarbon along with carrier gases like hydrogen (H₂) or argon (Ar). One of the earliest thermal CVD process was demonstrated by Q. Yu et al. using Ni catalyst in 2008 [44]. But due to lack of control on the number of layers of synthesized graphene, the use of copper as catalyst took advancement. The ﬁrst large scale graphene synthesized on Cu foil was reported by X. Li et al. in 2009 [45]. Henceforth, copper due to its low carbon solubility, high crystallinity and low cost appealed most to researchers for the transfer CVD process.

The quality and thickness of graphene synthesized in the CVD process depends on various factors like thickness and crystallinity of the metal catalyst and substrate as well as heating and cooling rate, CVD process temperature and amount of precursor and carrier gases. As mentioned, in all these CVD processes the material is synthesized on the catalytic metal foil which needs to be transferred to a dielectric for all probable applications and this mostly incur defects and contamination in the otherwise high quality developed samples. Thus, a process to avoid the transfer step was required to be developed and eventually techniques to synthesize single to few layers of graphene directly on a dielectric substrate are reported in recent years [46–54].

2.1.3.2 Direct Method

Over the years, different techniques have been approached to overcome the diffi- culties of transferring synthesized graphene films, unaffected from metallic foils to dielectric substrates. However, simultaneously it has also been an area of significant interest to grow graphene on dielectric substrates directly. Here also, like the transfer CVD technique, copper and nickel are mostly used as the metal catalyst.

The basic idea of this direct CVD deposition process involves the partial melting and receding of the metal ﬁlms, pre-deposited on the dielectric substrate, followed by the pyrolysis of the precursor and graphitization process wherein the graphitic layers are formed both at the metal-air and metal-substrate interface. It so happens because the carbon molecules migrate through the blanks formed due to melting and receding, hence forming graphene layers directly on the dielectric. Along with, general graphitic layers are formed on the metal surface as well. This latter layer of graphitic ﬁlms and the residue metal layers are removed by wet etching procedure, thus retaining the graphitic layer only over the dielectric substrate.

Although the melting point of copper and nickel are respectively 1084^◦C and

(20)

1455^◦C, the usual CVD process temperature using the said metals are around 900−1000^◦C and 750−850^◦C respectively. This is because thin ﬁlm of any material shows diﬀerent unique physical properties compared to that of its bulk counterpart.

The thin metallic film in direct CVD synthesis, partially melt at the process temperature. In low dimensional system, thermal instability is caused due to higher surface atom vibrations compared to those present deep within the bulk material. This is called surface melting and the atoms present become free to migrate over the surface similar to liquid atoms. Hence, it is observed that the CVD process temperature is high enough to obtain surface melting of the catalytic metal used [55]. In case of copper, due to its low carbon solubility, the carbon cannot penetrate through the film directly but migrate over the liquified surface and only penetrate through areas forming grain boundaries due to molecular receding [48]. For nickel which has a high carbon solubility, however, at high growth temperatures carbon molecules get aggressively dissolved into the bulk material followed by appreciable amount of precipitated carbon to the surface upon cooling, resulting in formation of graphitic carbon layers [56].

In this work, direct deposition of graphitic layers was achieved by using nano- metrically thin nickel film pre deposited on a fused silica substrate. Also, instead of using conventional gaseous precursor for the CVD pyrolysis process, solidified thin film of a carbon-based polymer photoresist was used. The growth of graphitic layer thickness on the basis of pre-deposited precursor layer and catalytic metal thin film thickness was extensively studied.

Nickel was chosen over copper because it has been observed that the former has quite strong interaction with graphene along with a special case of lattice matched system [56]. The metal d-electrons and the π-orbitals of carbon interact chemically resulting in smaller separation (0.21 nm) between the two instead of 0.33 nm van der Waal’s gap between the graphitic layers [52, 56]. Thus, much physical attributes of the material are altered. A theoretical simulation on graphene over nickel was studied by K. M. Al-Shurman et al [57]. Also, the process temperature as mentioned above for nickel is lower than that of copper. Further detailed description of the experimental processes and results are provided in the following chapters.

(21)

Chapter III

Outline of Process Setups and Measurement Devices

In this chapter a brief outline of the setups and the devices used during the course of the thesis work is provided. The evaporator, spin coater and the CVD were the setups used during sample synthesis while the Raman spectrometer, SEM, 4- point probe, spectrophotometer and proﬁlometer are the devices used for sample characterization. The descriptions of the same are incorporated in the following sections:

3.1 Evaporator

Physical vapour deposition (PVD) comprises of a variety of technologies where any material released from a source gets deposited on a substrate either by mechanical, electromechanical or thermodynamic processes [58]. The most common methods for the PVD are thermal evaporation and sputtering. Thermal evaporation is the method in which a solid material is vaporised in a high vacuum environment onto any substrate producing a thin ﬁlm coating of the material having varied thickness range from microns to a few nanometers [59]. It is mostly done by using either of the two methods namely, electron beam evaporation or resistive evaporation. Sputtering, on the other hand, is the method in which the material to be deposited on a substrate is bombarded with high energy particles inside a vacuum chamber in presence of an inert gas, usually Ar [58]. In the present work, an evaporator has been used where resistive thermal evaporation of nickel was made on fused silica (FS) substrate.

Resistive evaporation is a process in which the material is heated to its evaporation point by using electrical energy. The vaporised material then travels to the substrate where they condense and nucleate together to form the thin ﬁlm coating.

(22)

For resistive evaporation coating, high level of vacuum is required mainly for two reasons. Firstly, it increases the mean free path of collision between the vapour molecules and gas molecules. The collision between the two is undesirable due to the fact that it may change the direction of the vapour molecules and thus adversely inﬂuence the coating process. For pressure as low as 10⁻⁵ mbar, the mean free path is more than 5 m, which again is usually much larger than the chamber dimensions.

As a result, for such low pressure it may be concluded that the resistive evaporation becomes highly directional and generally the material molecules travel in straight lines from the source to the substrate. Secondly, high vacuum decreases the amount of air particles greatly and consequently improves the ﬁlm purity [60].

Moreover, during this process, two crucial factors need to be taken care of namely, measurement of the ﬁlm thickness and control of ﬁlm deposition rate. For these pur- poses, in most thermal PVD systems, quartz oscillator crystals are suitably mounted inside the vacuum chamber in order to perceive deposition in real time and monitor both the thickness and deposition rate in a measurable way. The main working technique involves the detection of the drop in the oscillation frequency of the crystal as a result of change in its mass due to the deposition of the metal on it. The measurement system thus obtains the frequency data, which is converted to thickness data (both instantaneous rate and cumulated thickness) continuously by appropriate mathematical functions by an electronic instrument.

The procedure of PVD in this work involves the placement of the material, whose thin film is to be deposited, on a metal boat between two metal electrodes in the bottom of the chamber. The substrates are held inverted fixed by scotch tapes to a substrate-holder placed above the boat at an elevated level inside the chamber. Thus, the surfaces intended to be coated face downwards at the heated source material. High vacuum is created within the chamber. The material is heated to its melting temperature by passing optimum current through the electrodes and further increase of the current, and consequently the temperature, leads to evaporation of the material onto the substrate for thin film coating. The figure 3.1 below shows the schematic of a general evaporator.

(23)

Figure 3.1: Schematic of resistive thermal vacuum evaporator.

3.2 Spin-coater

Spin coating is another method to deposit thin films of volatile polymer solvents over any flat substrate [61]. The device used for this purpose is called a spin- coater or a spinner. For spin coating, the substrate is placed on an appropriate size substrate-holder and a small but sufficient amount of polymer solution is dropped on it carefully either manually by a dropper,pipette etc. or automatically by a dispense unit. Once the polymer solution is deposited, the spinner is accelerated very fast to the desired rotation speed and rotation speed typically varies in the range 1000−4000 rpm for usually 30−60 s. The basic principle for this type of thin film deposition is that when high speed rotation occurs the liquid spreads radially outward due to the centrifugal force and all the excess fluid falls off the edge of the substrate leaving thin film coated layer on the substrate. Also, the polymer is generally volatile, so simultaneous evaporation also takes place while rotation.

The final deposited film thickness depends on the spinning speed, viscosity and concentration of the solution and the solvent [62]. After spinning is complete, the substrate is usually baked at a certain temperature for the film to solidify and the temperature and time of baking depends on the polymer. The process of spin-coating and baking is shown in the figure 3.2 below.

(24)

Figure 3.2: Schematic of spin-coater and baker.

3.3 Chemical Vapour Deposition (CVD) Reactor

Chemical vapour deposition (CVD) is one of the many, yet most commercially pop- ular, chemical processes for thin ﬁlm deposition. It mostly deals with the formation of thin solid ﬁlms on a substrate by chemical reactions of one or more precursors.

CVD produces high-purity and high-performance solid materials on a large scale.

The classification of CVD techniques can be made based on different factors like source of energy input, pressure regime, growth mechanism etc. The classification however is interlinked [63].

On the basis of operating pressure, three broad classiﬁcations of the CVD are made, viz. atmospheric pressure CVD, low pressure CVD and ultra-high vacuum CVD. The low pressure CVD are further classiﬁed on the basis of energy input source. Plasma-enhanced CVD, photochemical vapour deposition and thermal CVD are the major energy input sources. The thermal CVD too can be of two types:

hot-wall reactor and cold-wall reactor. In our process, we have used the hot-wall thermal reactor at low pressure. A hot-wall reactor is essentially an isothermal surface within which the substrate is placed. Proper furnace design can lead to precise temperature control, achieved due to heating of the whole chamber. A major disadvantage in this technique lies in the fact that deposition occurs on the walls of the chamber along with that on the substrate. Thus, frequent cleaning is necessary to avoid contamination of the substrates. The CVD apparatus generally consists of the following basic components:

• Gas delivery system: To let in gaseous precursor and other reactive gases within the chamber.

• Reactor chamber: Chamber within which the substrates are placed and the deposition takes place.

(25)

• Energy Source: To provide energy within the chamber in order to facilitate the chemical reaction.

• Vacuum system: To remove all kinds of gaseous species that are undesired during the process and to remove every kind of volatile by-products formed as a result of the chemical reaction during the CVD process from inside the chamber.

Further, there can be diﬀerent types of chemical reactions that may occur within the CVD chamber namely, pyrolysis, reduction, oxidation, disportionation, reversible transfer and compound formation [64]. In a CVD process, various types of precursors maybe used. However, an ideal precursor should preferably possess high volatility and good thermal stability. It should decompose cleanly and controllably on the substrate without incorporation and produce stable by-products that can be readily removed from the reaction zone [65]. In our case, a solid thin ﬁlm layer of a polymer resist is used as the precursor material.

The CVD reactor used in our experiment comprises of a quartz tube and substrate holder because quartz has a very high melting point and is also chemically inert. The basic physiochemical steps in the overall CVD process involves the evaporation and pyrolysis of the precursor at high temperature in a low pressure hydrogen environment, followed by the mass transport and adsorption of the atoms to the substrate. Next film formation takes place by surface diffusion of the atoms and nucleation at the growth sites. The by-products of the decomposition are simultaneously desorbed and transported away from the reaction zone. The schematic of a general CVD unit is given below in figure 3.3.

Figure 3.3: Schematic of a CVD reactor.

(26)

3.4 Raman Spectrometer

Raman spectroscopy is one of the main and eﬃcient technique to determine the type, quality and property of a material, specially in case of carbon and its allotrope [66, 67]. The basic working principle of Raman spectroscopy is based on inelastic Raman scattering. Before we dwell into the details of Raman spectroscopy of graphene and graphitic carbon layers, let us brief out a little background on Raman spectroscopy in general and Raman scattering.

3.4.1 Raman Scattering

When incident light hits a medium, energy and direction of incident light may change. This is called scattering of light. Change in energy of light means change in frequency or wavelength of light. In the case of scattering of light, wavelength (or energy) change is typically given as reciprocal of wavelength or wavenumber in cm⁻¹ for convenience. If energy of light does not change, or the change is smaller than 10⁵ cm⁻¹, scattering of light is called Rayleigh scattering. If energy of light changes more than 1 cm⁻¹, scattering of light is called Raman scattering [68]. The diﬀer- ence between frequency of incident light and frequency of Raman scattered light is known as Raman shift or Raman frequency, which is typically presented using wavenumber. If wavenumber of incident light is set to zero, both positive and negative Raman shifts are observed [68]. These are called anti-Stokes and Stokes lines respectively. Absolute value of Raman shift is the same for a Stokes and anti-Stokes line [69]. However, anti-Stokes lines have lower intensity. Furthermore, Raman shift does not depend on the frequency of incident light [68]. Energy diﬀerence between incident light and Raman scattered light corresponds to molecular vibration mode.

However, not all of the modes of a molecule cause Raman scattering. Mode is Ra- man active, i.e. it causes Raman scattering, if it causes change in polarizability [69].

From quantum mechanical point of view molecule changes from one state to another through a virtual state in the case of Raman scattering. These virtual states do not exist in reality, they represent the case, where incident photon is annihilated and scattered photon is created at the same time. To distinguish Raman scattering from ﬂuorescence, process of Raman scattering takes one picosecond or less, whereas ﬂu- orescence involves absorbtion and emission and is considerably longer process [70].

The energy level diagram of the diﬀerent scattering processes are shown in ﬁgure 3.4.

(27)

3.4.2 Raman Spectroscopy

Vibrational modes and Raman shifts are diﬀerent for diﬀerent molecules, therefore Raman scattering can be used in spectroscopy. Raman spectrometer consists of light source, optical microscope, spectral dispersion and spectral acquisition components.

Usually laser is used as a light source, because it is highly monochromatic, coherent and polarized and can be directed and focused easily. Monochromatic light source is important, because absolute value of Raman shift depends on the wavelength of incident light. If light source is not very monochromatic, peaks in Raman spectra may not be that clear. The optical microscope focusses incident light, so that it is more monochromatic, and also adjusts intensity and polarization of incident light. Spec- tral dispersion and acquisition parts form the Raman scattering spectrum known as Raman spectrum [68]. Raman spectrum is generally obtained in the form of intensity as a function of Raman shift. Commonly intensity, frequency (Raman shift) and line shape and width are the spectral parameters seen in the spectrum. Ra- man spectra are different for different molecules, therefore sample’s molecules can be identified, provided Raman spectra of sample’s molecules are known. Polariza- tion of Raman scattered light depends on polarization of incident light and also on symmetry of sample’s molecules or crystal structure [68]. Raman spectroscopy is comparable to infrared spectroscopy. In infrared spectroscopy, intensities of incident and transmitted light are compared to find the wavelengths absorbed by the sample’s molecules. Infrared spectroscopy also gives information about molecules’

vibrations, yet it has many differences with Raman spectroscopy. Vibrational modes can be either Raman active, infrared active or both. Also some modes can be seen better in Raman spectrum than in infrared spectrum and vice versa. Water and glass do not interfere measurements in Raman spectroscopy, whereas they absorb infrared radiation strongly, thus causing problems in infrared spectroscopy. Another advantage in Raman spectroscopy is that Raman spectrum from one measurement can cover large range of frequencies. For instance, gratings and detectors have to be changed in infrared spectroscopy to achieve such large range of frequencies. Dis- advantages in Raman spectroscopy are possible fluorescence and local heating, and equipment of Raman spectroscopy is generally more expensive [71]. The schematic of a Raman spectrometer is shown in figure 3.5.

(28)

Figure 3.4: Energy level diagram showing Rayleigh scattering, Raman scattering and infrared absorption.

Figure 3.5: Schematic of a Raman spectrometer.

3.4.3 Raman Spectroscopy of Graphite and Graphene

Analysis of Raman spectra determine the quality, nature and number of layers of the graphitic structure. Within the wavenumber range 1000 cm⁻¹ to 3500 cm⁻¹, graphene or in general graphitic carbon layers give three peaks namely, the 2D, G and D peaks. For any carbon material possessingsp²hybridization bonds, intense 2D and

(29)

G peaks are observed at ∼2700 cm⁻¹ and at ∼1580 cm⁻¹ respectively [66]. The D or the defective peak, usually obtained at∼1350 cm⁻¹ gives a qualitative measure of the disorder in the material structure [66]. The position, shape and relative intensity of the peaks vary, giving quite significant idea about the crystallinity and number of layers of the graphitic structure [72]. The ratio of the D and G peaks specifically indicates the crystallinity of the graphene/graphite sample, while on the other hand the position of the 2D peak gives idea about the thickness of the graphitic layers [73, 74]. In the Raman spectrum, a wide G peak between 1500 cm⁻¹ and 1550 cm⁻¹ signifies amorphous form of carbon. Also, shift of the G peak towards 1600 cm⁻¹ from its original position at 1580 cm⁻¹determines the presence of nanosized graphitic flakes while a shift towards 1500 cm⁻¹ determines the presence of amorphous carbon [75]. Similarly, the widening and shifting of the 2D peak illustrates the number of layers of the graphitic material. The 2D peak is located at 2730 cm⁻¹ in bulk graphite [76], while in graphene monolayer it is located at 2680 cm⁻¹ [77]. The 2D peak is typically located near 2700 cm⁻¹ for a few layered graphene [74, 77]. Also, it may also be noted here that apart from the aforesaid three prominent peaks, minor peaks may also be observed at locations near 1600 cm⁻¹, 3250 cm⁻¹ and 2450 cm⁻¹ that arises due to doping and defects in the material [66].

3.5 Scanning Electron Microscope (SEM)

The scanning electron microscope is another important device in our experiment to characterize the synthesized samples. The SEM is much more powerful than an optical microscope. In terms of magniﬁcation, while an optical microscope images a sample a few hundred times, a SEM magniﬁes a sample to almost 300,000 times [78].

The general SEM consists of the following parts, namely, the electron gun, lens system comprising two condenser lenses and an objective lens, set of deflectors, electron detection systems and the sample chamber [79]. The entire system works in vacuum because otherwise the electron beam would constantly interfere with the air particles in the atmosphere [78]. The steady flow of electron necessary for the working of the SEM is generated by the electron gun which can either be thermionic or field emit- ting [78]. Electron gun is situated either at the top or the bottom of the SEM tube and eject the electrons and accelerates them with energy between 1−30 keV [78,79].

The next two segments comprising the lens and deﬂector systems focus and channel-

(30)

ize respectively, the electron beam towards the specific sample location. The final beam reaching the target sample usually passes through an electron probe having diameter in the range of 1−10 nm carrying a current of 1−100 pA [79]. Once the electron beam hits the sample kept in the sample chamber, the electrons interact in different ways with it. The electrons, called secondary electrons are scattered from the surface of the sample while others penetrate through it and scatter back from different depths of the sample. Such scattered electrons are called backscattered electrons. The former usually have energy less than 50 eV and depicts detailed image of the sample’s surface. On the other hand, the backscattered electrons possess a wide energy range from 50 eV to the energy of the incident electron beam. They give detailed image of the composition of the sample. The SEM has separate detectors for detecting these electrons individually [78,79]. Figure 3.6 shows schematic of a general SEM.

Figure 3.6: Schematic of the scanning electron microscope.

3.6 Four Point Probe Device

Resistivity is an inherent physical property of a material that determines how con- veniently electrical charges ﬂow through it. Resistivity (ρ) is measured in Ωm and

(31)

it is given by:

ρ= R·A

l (3.1)

where, R is the resistance of a bar of length l and cross-section area A. Also, we have from the Ohm’s law:

R = V

I (3.2)

where V is the voltage and A is the current. Combining equation 3.1 and 3.2, we get

ρ= V I ·A

l (3.3)

Now, sheet resistance, particularly gives the resistance of thin ﬁlms which has uniform micro to nano-order thickness. It is deﬁned as the resistance of a square of arbitrary dimension i.e. the sheet resistance of a material is numerically equal to the resistance of a square sample of the material [80]. Thus, equation 3.3 becomes:

ρsquare= V

I ·l (3.4)

and therefore the sheet resistance (R_s) is deﬁned as:

R_s = ρ l = V

I (3.5)

where R_s is the sheet resistance and its unit is Ω/sq.

It is important however, to measure the sheet resistance of samples of arbitrary shape and size and the four point probe is an established method to determine it directly. In this setup, as by the name, there are four equi-spaced probes whose spacing is much larger compared to the size and thickness of the film. The probes are pressed down onto the film and current passes through the outermost probes, which is connected to an external power source. The potential difference between the two inner probe is given by:

(32)

V = I·R_s· ln2

π (3.6)

whence,

R_s = k·V

I (3.7)

where k =π/ln2 denotes a correction factor constant and its value is approximately taken as 4.53. The four point probe setup schematic is shown in the ﬁgure 3.7 below.

Figure 3.7: Schematic of a four point probe setup.

3.7 Spectrophotometer

When light is incident on a material medium, a part of it gets reﬂected, absorbed or transmitted. The optical properties of any material can be quantitatively anal- ysed by the amount of light (from UV to IR) it absorbs, transmits and reﬂects.

Absorbance, transmittance and reflectance are defined by the ratio of the amount of light the material absorbs, transmits and reflects respectively,to the amount of incident light on it. Thus, if I₀ denotes the incident light, I_a the absorbed, I_r the reflected and I_t the transmitted light, then by the conservation of energy law, it can be written as

I₀ = I_a+ I_r+ I_t (3.8)

Spectrophotometers are instruments that are employed to measure the transmittance and reﬂectance of any sample directly as a function of wavelength. The

(33)

UV−VIS spectrophotometers works over the broad range of the ultraviolet and visible part of the electromagnetic spectrum roughly ranging from 200−400 nm in the UV region and 400−800 nm in the VIS region [81]. Any spectrophotometer usually consists of the following parts, namely, the source, a filter or monochromator, sample holder and detector (generally, phototubes or photomultipliers). The source provides the continuous radiant energy over the concerned spectrum region. The filter or monochromator scans the spectrum in narrow wavelength bands and the detector detects the incident, transmitted or reflected light intensities as applicable.

Spectrophotometers can either be of two types- single beam or double beam spectrophotometers. The main working of the double beam spectrophotometer deals with the splitting of the incident beam by a beam splitter and passing one to the photodetector directly as the reference and the other through the sample determin- ing the transmitted or reﬂected beam intensity. The ratio of the two intensities are calculated and the plot of transmittance or reﬂectance as a function of wavelength can be obtained in a connected computer device.

Figure 3.8: Schematic of a spectrophotometer during transmittance measurement.

Figure 3.9: Schematic of a spectrophotometer during reﬂectance measurement.

(34)

3.8 Proﬁlometer

Thickness as well as surface profile is one of the key characterization parameter of thin films. Profilometers are instrument that are used for the purpose and they are mainly of two types- stylus and optical. While the former is contact based, optical profilometers are non-contact profilometers [82]. Both the types have their respective advantages and disadvantages.

In this work, a stylus profilometer (schematic shown in figure 3.10 below)has been used to measure the thickness of the deposited films and synthesized layers on the samples. A stylus profilometer comprises mostly of two parts namely, the detector probe and the sample stage. For measurements, the probe physically moves along the surface to obtain the surface height. The probe determines the surface undulations with respect to a prior set reference. It is a high sensitive device and its resolution depends on the diameter of the needle tip [83]. Thus, as a result the stylus provides not so accurate results in case of very thin films in the order of a few nanometers (typically, less than 50 nm).

Figure 3.10: Schematic of a stylus proﬁlometer.

(35)

Chapter IV

Experimental Details

The experimental part of the scientiﬁc study maybe broadly classiﬁed into two parts namely, synthesis and characterization of samples. The detailed processes in both parts are given in the following sections:

4.1 Fabrication of Samples

The aim of the thesis work is to synthesize graphitic carbon layers directly on a dielectric substrate using metal thin ﬁlm layer as a catalyst. For the purpose, fused silica is used as the substrate and nickel as the metallic catalyst ﬁlm. The negative tone resist layer nLOF is being used as the carbon precursor.

Twenty samples with four different nickel layer thicknesses, each again with four different nLOF coating thicknesses were prepared to accomplish the experimental target. Among these samples, one set of four samples was taken as reference, in which there was no nickel layer deposition i.e. nickel layer thickness will be considered zero. The entire synthesis part maybe further divided into four subparts namely, deposition of nickel thin film, spin-coating of resist layer nLOF, synthesis of graphitic carbon layers and finally, etching of nickel particles, which are described in details in the following subsections:

4.1.1 Deposition of Nickel Thin Film

The ﬁrst step of synthesis dealt with the deposition of the nickel thin ﬁlm on the fused silica substrate in an evaporator. Circular disc-shaped FS substrates with 25 mm diameter and 0.5 mm thickness were used. The resistive thermal deposition process was performed by the Leybold Univex 300 vacuum evaporator. During each

(36)

evaporation process one set with four FS substrates were deposited with a specific nickel thin film thickness. Nickel pallets with 99.99% purity were used as source material for the deposition. Four such sets were made with nickel film thickness 5 nm, 10 nm, 15 nm and 20 nm. The aforesaid step was not done for the reference sample set having no nickel deposition.

During every evaporation process, vacuum of about 10⁻⁵mbar was created within the chamber by pumping for atleast 2 hour. The density, z-ratio and tooling factor were set to 8.91 g/cm², 0.331 and 90% respectively. Once vacuum was achieved, the evaporation and consequently, thin ﬁlm deposition process was done by passing a current of 6−8 mA with deposition rate no more than 4 A^◦/s, within the chamber.

The ﬁgures A.1 and A.2 in appendix A shows the evaporator setup and its various parts respectively.

4.1.2 Spin-coating of Resist Layer nLOF

In the second step of synthesis, the nickel deposited substrates were coated with negative tone resist nLOF in the spin-coater Headway Spinner PWM101D. Each sample set was coated with resist layer of thickness 50 nm, 100 nm, 350 nm and 1000 nm respectively. The negative resist AZ nLOF 2070 ( details see appendix E), of varied concentration (diluted with solvent AZ EBR) was used to prepare the samples. All the samples were spun for 1 min and then baked at 110^◦C for 2 min in order to dry out the film. To obtain the respective thickness, different concentration of the resist and spin rates were used which is shown in the table 4.1 below and figure A.3 in appendix A shows the spin-coater and the baking setup used.

Table I

Resist Layer Thickness (nm) Resist Concentration Spin-Rate (r.p.m)

50 1 : 10 4000

100 1 : 10 1000

350 1 : 4 1000

1000 1 : 1 3000

Table 4.1: Table showing the recipe used for spin-coating.

(37)

4.1.3 Synthesis of Graphitic Carbon Layers by Pyrolysis

Pyrolysis of the precursor constituted the main part of the synthesis process where the graphitic carbon layers were synthesized in a Carbolite CTF 12/75/700 CVD reactor (shown in appendix A figure A.4). During each CVD process, the chamber was pumped to vacuum of pressure 0 mbar for atleast 1 hour in order to remove all undesirable gaseous molecules. Post vacuum generation, the reactor was pro- grammed to ramp up to 700^◦C at 20^◦C/min and then further to 800^◦C at 10^◦C/min and pyrolysis was done for 10 min and eventually the program was terminated by ramping down to 700^◦C at 10^◦C/min. At 700^◦C, the device was dwelled on for another 5 min. The entire process was done at a constant flow of H₂ gas inside the reactor with corresponding pressure kept at 1 mbar. Once the CVD process ended, the sample inside the reactor was left to cool in the H₂ environment at presssure 10 mbar overnight. The CVD process-time plot is shown in the figure 4.1 below.

Figure 4.1: Plot showing the CVD process description.

In the CVD process, annealing is done from 100^◦C to 700^◦C. At around 700^◦C, the Ni molecules begin to recede. In order to achieve this more or less uniformly, the second ramp step is changed to 10^◦C/min as mentioned above. Further, during this stage, the precursor layer nLOF begins to evaporate simultaneously. Once the

(38)

reactor reach 800^◦C, pyrolysis of nLOF occur and the carbon atoms gets absorbed to the Ni molecules. Graphitization occurs during the ramping down of the CVD process to 700^◦C. The CVD process schematic is shown in the ﬁgure 4.2 below.

Figure 4.2: Schematic of the CVD process.

4.1.4 Etching of Nickel

During this process, each of the sample was dipped into a solution of 10 g copper sulphate (CuSO₄) , 50 ml hydrochloric acid (HCl) and 50 ml distilled water (H₂O), for exactly 1 min and then washed in distilled water and dried with compressed nitrogen (N₂) gas. As observed by an optical microscope, most of the nickel particles were removed from the samples.

4.2 Characterization of Samples

In this section, the processes performed for characterization of all the synthesized samples are described elaborately. The sample characterization are done in two parts, pre and post nickel etching of the samples. The following techniques are used to characterize each sample:

4.2.1 Measurement of Raman Spectrum

The main characterization of the samples was based upon the measurement of the Raman spectrum. The Raman spectrum of the sample was measured by Renishaw

(39)

inVia Raman microscope (see appendix A ﬁgure A.5) using laser of wavelength 514 nm. The samples were focussed by the 100X microscope objective. The measurement was made in the spectrum range 1000−3500 cm⁻¹with 30 second exposure time and 0.25 mW power, which is 5% of the total laser power. The 3.4Wire software was used to provide the measurement speciﬁcations to the the Raman unit and obtain the measured Raman spectrum.

4.2.2 Observation of SEM Images

The SEM images of all the samples was made with LEO 1550 Gemini scanning electron microscope. The setup was entirely controlled by the ZEISS SmartSEM program software. The samples were placed on the sample holder which was carefully placed on a motorized stage having five degrees of freedom in movement, namely X, Y, Z, tilt and rotation, within the sample chamber. The chamber was then pumped down to vacuum and the electron high tension (EHT) was kept at 3 kV. The samples are then focussed for sharp contrast and high resolution by first fixing the working distance(WD)and then by changing the magnification, focus, brightness, contrast and stigmation. All the samples are imaged at 100X, 1000X, 5000X and 10000X magnification. The figure A.6 in appendix A shows the SEM setup in our clean room.

4.2.3 Measurement of Sheet Resistance

The electrical properties of the synthesized samples were characterized by the measurement of the sheet resistance by the 4-point probe device named Signatone S-1160 Probe Station. The sample was placed on the measurement plate and the lever then pulled to bring down the four point probe on the sample. The source voltage was kept at 5 V. The current passing through the outermost probes and the the potential diﬀerence across the two inner probes were measured by separate multi-meters.

Such current-voltage combination was measured at five different positions for each sample. The sheet resistance is then calculated by equation 3.7. The four point probe setup is shown in figure A.7 in appendix A.

4.2.4 Measurement of Absorbance

The optical properties of the synthesized samples were characterized by absorbance measurement. For the same, both transmittance and reﬂectance for each sample was

(40)

measured by the spectrophotometer and consequently the absorbance was calculated from the measured values using the equation 4.1 given below in Matlab.

A = 1−(T + R) (4.1)

where A denotes the absorbance, T the transmittance and R the reﬂectance.

The spectrophotometer used during this work was the Perkin Elmer UV/VIS Spectrometer Lambda 18. The spectrophotometer was first calibrated with a bare FS substrate prior to both set of transmittance and reflectance measurements respectively, where the main scan parameters were given as scan length: 200−800 nm, scan interval: 1 nm and scan speed: 240 During the transmittance measurement, the synthesized samples were each placed one by one at the front window of the integrating sphere and during reflectance measurement, the same where placed at the back window of the integrating sphere. The spectrophotometer device is shown in figure A.8 and sample placement for the respective measurements is shown in figure A.9 in appendix A.

4.2.5 Measurement of Surface Proﬁle

The layer thicknesses of the deposited materials on the fused silica substrate and the resultant thickness of the graphitic carbon layers synthesized were measured by the Veeco Dektak 150 stylus proﬁlometer. It is shown in appendix A ﬁgure A.10. Since many of the measured layers were of the order of only a few nanometers (<50 nm) with high roughness, the measurement was not highly accurate and comprehendible.

Hence, rough approximate measurements are obtained with an error estimate of

±2/3 nm. The measurements were made in the hills and valleys mode over a scan length of 300 μm with scan duration of 20 second.

(41)

Chapter V

Results

In this chapter, the elaborate results of sample synthesis and characterization are presented sequentially. Characterization of samples was broadly done in two phases, namely pre-Nickel etch characterization and post-Nickel etch characterization. The following sections demonstrate the results obtained during this thesis work.

5.1 Sample Synthesis

The synthesized graphitic layers corresponding to the diﬀerent precursor nLOF layer is measured by the proﬁlometer. The following table 5.1 shows the developed graphitic layer thicknesses. As can be found out, the nLOF to graphitic layer thickness ratio approximately gives an average multiple factor of 6.

Table II

Resist Layer Graphitic Layer Sample Thickness (nm) Thickness (nm) Nomenclature

50 8 GrL8

100 15 GrL15

350 60 GrL60

1000 200 GrL200

Table 5.1: Table showing approximate thickness of synthesized graphitic carbon layers.