Effect of structural defects on the electrical properties ...898103/FULLTEXT01.pdf · 1.1....
Transcript of Effect of structural defects on the electrical properties ...898103/FULLTEXT01.pdf · 1.1....
Determination of Electrical and Gas sensitivity Properties of
Graphene Sheets 1FA593: Project in Physics and Materials Science
(30 credits)
SYLVESTER WAMBUA MAKUMI
12/07/2015
Supervisor: Prof Klaus Leifer
Co-Supervisors:
Hu Li Ishtiaq Hassan Wani
Examiner: Prof Håkan Rensmo
i
Abstract
The understanding of the properties of graphene on a substrate is essential for the development
of the next generation electronic technology. Of equal importance is the understanding of ways
to modify its properties to suit specific application. Pure graphene has revealed high electrical
conductivity [35] however in the actual fabrication and application of graphene, defects are
unavoidable. We therefore need to understand the effect of defects on the properties of graphene
material and ways of modifying these properties. In this study, the modification of graphene’s
electrical properties using atomic scale structural defects has been studied. The relationship
between defect density and the electrical properties of graphene on SiO2 substrate have been
investigated using the PeakForce TUNA mode of MultiMode 8 AFM. NO2 gas sensing property
of graphene has been investigated using the Keithley 6430 source meter. It has been shown that
atomic scale structural defects lower the electrical conductivity of graphene sheet. For high
defect densities, graphene sheet did not show any conductivity. It has been shown that, the NO2
gas sensing property of graphene is good and can be increased by exposing graphene sheet to UV
light. A response of 12.56% has been achieved in N2 gas condition.
ii
LIST OF CONTENTS
Abstract i
Contents ii
1. NTRODUCTION 1
1.1. Background information 1
1.2. Electronic structure of graphene 2
1.3. Electrical properties of graphene 5
1.4. Gas sensing mechanism of graphene 6
1.5. Other Related Studies 8
1.5.1. Effect of structural defects on the electrical properties of graphene 8
1.5.2. Gas sensing properties of graphene 9
1.6. Atomic Force Microscope (AFM) 9
1.6.1. PeakForce TUNA 11
1.6.1.1. PF TUNA Imaging mode 12
1.6.1.2. I-V Spectroscopy mode 12
1.7. Basic understanding of Raman spectroscopy spectrum 13
2. EXPERIMENT I: ELECTRICAL PROPERTIES MEASUREMENT USING
PEAKFORCE TUNA 14
2.1. Sample Preparation 14
2.2. Raman spectroscopy studies 15
2.3. Experimental Details 17
3. EXPERIMENT III: TESTING THE GAS SENSING PROPERTY OF GRAPHENE 19
3.1. Experimental details 19
4. RESULTS AND DISCUSSION 20
4.1. Electrical properties of Silver contacts 20
4.2. Electrical properties of Graphene 21
4.2.1. Evaluation 27
4.3. Gas sensing properties of graphene 28
4.3.1. Evaluation 30
5. POSSIBLE APPLICATIONS OF RESULTS 31
iii
6. CONCLUSION 32
ACKNOWLEDGEMENT 32
References 33
APPENDIX I 39
Page 1 of 39
1. INTRODUCTION
1.1. Background information
To advance from the current semiconductor technology level dominated by silicon, there is
need for new materials with better controllable properties than silicon. Graphene has been
found to be a potential material to use in place of silicon. Suspended Graphene sheets exhibit
the highest carrier mobility[1], [2] but for application purpose, graphene need to be supported
by a substrate. Therefore, it is important to understand the properties of graphene sheet on a
substrate.
The graphene’s unique electronic properties and possible ways to modify it have drawn a
large interest. The possibility to modify the electronic property of graphene sheets by
chemical or physical means is a crucial step in the application of graphene in different
electronic devices and to make graphene a competitive material for future electronics. In the
past few years, many researchers have studied ways of modifying the graphene’s
properties[3]–[5].
The use of defects is one of the ways used to manipulate the mechanical and electrical
properties of graphene. Both the nature and amount of defects have a great effect on the
properties of graphene [6]. Defects can be seen as imperfections that can degrade the
electrical and mechanical properties of a material. However, structural defects in graphene
can be very useful in modifying the graphene properties for specific applications. Ion
irradiation using different ions[7]–[12]has been shown to be an effective methods of
introducing controlled amount of structural defect in a graphene material.
Page 2 of 39
In this study, Ga+ ions have been used to introduce atomic scale structural defects in graphene
sheet on SiO2 substrate. The effect of these defects on the electrical conductivity of graphene
has been studied using the PeakForce TUNA mode of MultiMode 8 AFM.
Studies have shown that graphene may offer an excellent sensing device in its application as
a sensor for different gases [13]–[15], radiation [16],biomolecules, and different chemicals
[17]–[19].Other studies have found reduced graphene oxide to be sensitive to toxic
vapors[20] water vapor[21].However, the gas sensitivity of graphene is not well studied and
there is need for more studies. In this study, the sensitivity property of graphene to NO2 has
been studied.
1.2. Electronic structure of graphene
Graphene is made up of carbon atoms arranged in honeycomb pattern. Carbon is the sixth
element of the periodic table. Its atom has six protons, A neutrons, and six electrons. If A = 6
and 7 result to the stable isotopes 12
C and 13
C, respectively, and A = 8 yield the radioactive
isotope 14
C. The isotope 12
C is the most common in nature. These 12
C take 99% of all carbon
atoms, 13
C and 14
C are only in traces[22]. At atomic ground state, the 6 electrons in a carbon
atom usually in 1s22s
22p
2 configuration, that is two electrons in the inner shell 1s orbital, two
in the 2s orbital and the other two in the 2p orbital. In the presence of other atoms, it is
favorable to excite one electron from the 2s to the third 2p orbital to form a covalent with the
other atoms. � hybridization involves a superposition of the 2s and two 2p orbitals to obtain
the planar � hybridization.
The carbon atoms in graphene condense in a honeycomb lattice due to this � hybridization.
Three valence electrons, in the carbon atom, form covalent C-C bonds in the x-y plane with
three neighboring carbon atoms. These form the �-bonds, which are responsible for the high
mechanical strength of graphene. There is one valence electron per atom which does not take
Page 3 of 39
part in the covalent �-bond formation. This remaining electron combines with its nearest
neighbors to form the orbital. These orbitals make graphene very conductive. Pure
graphene has one conduction electron per atom.
Graphene can be seen as a semiconductor without a bandgap. As electron traveling through a
semiconductor, it is influenced by the positively and negatively charged nuclei; there is an
internal electric field, the crystal field. When an external electric field is applied, the electrons
are influenced by two fields; the internal electric field and the external field due to the applied
external voltage. Thus, the total force on the electron will be
= �� + ���
The influence of the internal field is very complicated and it is determined by the electronic
properties of the material. These affect the current that flow in the material.
In crystal semiconductors, atoms are arranged periodically. This leads to periodic variation of
the potential energy. Therefore, the energy E as a function of k, E (k) is periodic because the
effective potential is periodic. The periodic zone picture of the E-k diagram is as shown in
figure 1 (a). The reduced zone, that is, the band structure for k values between − � and
� will
be as shown in figure 1 (b). The density of states in a band is very large but the number of
carrier concentration, which is responsible for electrical conductivity is very small. Therefore,
a small portion of the E-k diagram is important in the understanding of electrical conductivity
of a material. Only charge carriers close to the top of the valence band and those close to the
bottom of the conduction band participate in electrical conductivity of the material.
For each value there is a corresponding allowed energy value which is a solution of the
Schrödinger equation. values are discrete but the number is large such that a continuous
Page 4 of 39
line is drawn. These leads to the E-k diagram shown in figure 1. is the electron wave
vector, E is the energy, and is the Fermi energy.
(a) (b)
Fig. 1. The E-k diagram (a) Periodic zone picture and (b) reduced zone picture for ordinary semiconductor. (c)
The band dispersion near theK̅-point for ideal monolayer graphene and (d) the π-band near the K̅-point for
ideal bilayer graphene. is the electron wave vector.
There is two atoms per cell in a graphene structure. This leads to two ‘conical’ points per
Brillouin zone ( and ’) at which the band crossing occurs. A monolayer graphene shows
energy bands which have a linear dispersion which can be described by the equation =ħ � [23], where is the electron wave vector, it is the propagation vector, � = m/s is
the Fermi velocity, energy, and ħ is the reduced planks constant. This behavior is due to
symmetry considerations [24], [25]. On the other hand, a bilayer graphene show a parabolic
Ef Ef
(d) (c)
E(e
V)
k (A-1)
E (e
V)
k (A-1)
k(Å-1) k (Å-1)
Page 5 of 39
spectrum as shown in figure 1(d). The two systems form zero-gap semiconductors or a zero-
overlap semimetal. The lower valence �- and upper conduction �∗-bands meet at the �̅- and �′̅̅ ̅-points. These points, �̅ and �′̅̅ ̅, are the Dirac point. Their energy position (� ) is exactly
at the Fermi Level (� ). The conical spectrum in graphene results from the interaction of the
energy bands originating from the sub-lattices A and B. The quantum-mechanical hopping
between the sub lattices A and B give two energy bands whose intersection close to the edges
of the brillouin zone is a conical energy spectrum. Since the involved electrons belong to two
different sub-lattices, an energy gap is not opened. Hence, graphene is seen as a zero-gap
semiconductor or as a zero-overlap semimetal. Hence, graphene show unusual semi metallic
behavior.
1.3. Electrical properties of graphene
Graphene has revealed high electrical conductivity [2], [26]–[28]. Carrier mobility of up to
10,000 cm2/Vs was found for few layer graphene and up to 15,000 cm2/Vs at 300K for
multilayer graphene [28]. Electron in graphene are able to cover micrometer-long distances
without scattering even at room temperature [29]. Multilayer graphene has been found to be
able to sustain very high current densities of about × 7A/cm2 in ambient air [30]. The
estimated charge carrier velocity in graphene is high, about 106 m/s at the Dirac points. As a
result, graphene show electronic properties for a two dimensional gas of charged particles
which is described using relativistic Dirac equation instead of non-relativistic Schrödinger
equation. The charge carriers behave like massless Dirac fermions [31], [32]. In particular,
there is nothing relativistic about electrons moving around carbon atoms, when the electrons
interact with the periodic potential of graphene’s lattice new quasiparticles result. For low
energies E, these quasiparticles are precisely described by the Dirac equation with an
Page 6 of 39
effective speed of �� = ms-1
.These particles are called massless Dirac fermions and can
be seen as neutrinos with electron charge e or electrons without their rest mass mo[29].
Some semiconductors reveal charge mobility as high as ≈77,000 cm2 V
–1 s
–1 (Indium
antimonide (InSb)); however, these values are only achieved for un-doped bulk
semiconductors. For graphene, mobility remains high even at high charge carrier
concentration ( > cm2) in chemically and in electrically doped graphene [13].
In a defective graphene sheet, Charge transport usually through a combination of tunneling,
hopping, c-axis conduction across nano-holes, and percolation conduction pathways [4].
However, in a defect free graphene, c-axis transport is insignificant and charge transport
occurs through the graphene channel in the ab-plane [33]. c-axis transport is more significant
in a graphene sheet with nano-holes where charge carries diverge at the nano-holes’ walls [4].
If the graphene sheet is highly defective, charge carriers follow percolative conduction
pathways [4].Structural defects scatter charge carriers [3], [6], [34]–[37]. Therefore, the
presence of structural defects lower the electrical conductivity of graphene sheet as it has
been observed in this study.
1.4. Gas sensing mechanism of graphene
The operation principle of graphene gas-sensing device is based on the changes of its
electrical conductivity as a result of interaction of the graphene sheet and the gas molecules.
The gas molecules act as either electron donor or acceptors [13]. The transfer of electrons
from graphene sheet (or gas molecule) to the gas molecules (or graphene sheet) increases the
number of holes (or electrons) in the sheet. This increased charge carrier increase the
electrical conductivity of graphene. Therefore, by monitoring the changes in the conductivity
of graphene we can know the presence of gas molecules in the environment.
Page 7 of 39
Graphene is an excellent material to use as gas sensing devices because of the following
unique properties. First, it is very conductive; graphene show metallic conductivity thus, low
Johnson noise very low carrier densities[13], and a few extra electrons or holes can cause
significant changes in carrier concentration. This leads to changes in its conductivity.Second,
it is two-dimensional material; hence, a large surface area is exposed to interact with the gas
molecules. This increases the graphene sensitivity to gas molecule.
In this study, the NO2 gas sensing property of graphene has been investigated. Previous
studies have found NO2 molecules to be an electron acceptor which change graphene to p-
type [38], [39]. Electrons move from the graphene sheet to the NO2 molecules. This increases
holes concentration [13], [38], [40] in the graphene sheet. Therefore, absorption of NO2
molecules leads to an increase in the electrical conductivity of graphene.
The molecular doping strength is determined by the density of states (DOS) and the lowest
unoccupied and the highest occupied molecular orbitals (LUMO & HOMO) of the adsorbate.
The transfer of charge may occur because of the relative position in the DOS of the LUMO
and HOMO of the adsorbate. In case the HOMO is located above the Dirac point of
graphene, charge will be transferred from the adsorbate to graphene[39]. When the LUMO is
below the Dirac point, transfer of charge from graphene to the molecule will take place. The
transfers of charge are, to a small extend, determined by the mixing of the LUMO and
HOMO with graphene orbitals (hybridization). the Dirac point of graphene is above the
LUMO of NO2[39]. This leads to electron transfer from the graphene sheet into the NO2
molecule. Polarization of the NO2 on the graphene surface result to charges leading to gating
effect. This contributes to the increase in electrical conductivity of graphene. In addition,
some NO2 orbitals occur close to the Dirac point. This results to some charge transfer from
the NO2 molecule to graphene through orbital mixing.This transfer of charges increases the
conductivity of graphene as seen in this study.
Page 8 of 39
1.5. Other Related Studies
1.5.1. Effect of structural defects on the electrical properties of graphene
The electrical properties of defective graphene have been studied both theoretically and
experimentally.Conductivity of graphene can be increased by introducing vacancy defects[3].
They observed that, vacancy defects increase the conductivity of graphene sheets by more
than one order of magnitude. For small vacancy defects density, graphene near the vacancy
defects has metallic properties. However, for higher vacancy defect density, the defects act as
scattering centers, hence conductivity of graphene decrease with increasing defect density.
Intrinsic defects degrade the quality of graphene on SiO2 through complicated mechanism
rather than defect scattering only [41]. In their study, they have found that, the effect of
lattice defects on graphene conductivity become remarkable only when the defect density
exceed ~103/�m
2. They have noted that, the height of graphene on SiO2 substrate is
proportional to its defect densities.
Structural defects in graphene sheets cause intervalley [42] and intervalley scattering [36],
[42] that lower its electrical conductivity. The Fermi Velocity is lower near the defects as a
result of enhanced scattering[42]. Structural defects can also lead to a band-gap in a graphene
electronic structure[42].
In their study, graphene sheets on SiO2were irradiated by Ne+ and He
+ ions. They found that,
lattice defects greatly depress the minimum conductivity, even below / ℎ (the minimum
metallic value) and induce insulating temperature dependence of the conductivity. They have
noted that, defects induced by ion irradiation cause significant intervalley scattering which
results to a conductivity proportional to charge carrier density, with mobility decreasing as
the inverse of the ion dose. Point defects, grain boundaries, disorders, wrinkles, folds, cracks
Page 9 of 39
and tears scatter charge carriers hence increasing the electrical resistance of graphene sheets
[37], [43].
Thiyagarajan et al. [4] have studied the electrical transport of multilayer graphene (MLG)
after Ar plasma irradiation. They have observed that, the conductivity of MLG reduce with
defect inclusion and MLG is transformed from metallic to semiconductor. Defective MLG
show a metallic-to-semiconductor transition in the resistance as temperature changes.
There is no available study that has employed the PeakForce TUNA technique to study the
electrical properties of defected graphene on SiO2 substrate.
1.5.2. Gas sensing properties of graphene
Studies have shown that graphene may offer an excellent sensing device in its application as
a sensor for different gases [13], [14], [38], [44], [45], [46]. Chen et al. [44] showed that a
pristine graphene is sensitive to gas molecules at concentrations as low as 158 ppq at room
temperature. They have shown that, a continuous UV light illumination during gas detection
is able to make simple two-terminal graphene based sensor sensitive to as low as 158 ppq NO
molecules at room temperature. The ultra-sensitivity was also confirmed by the low-level
detection of NO2, NH3, N2O, O2, SO2, CO2, and H2O. Few layer graphene show good
sensitivity for NO2, NH3, H2O and CO [13]. Wu et al. [46] have use CVD graphene to sense
hydrogen gas in air with concentration between 0.0025 and 1%. NO2 molecules have been
found to be an electron acceptor, which change graphene to p-type [38]. In this study, the
NO2 gas sensing property of graphene has been studied.
1.6. Atomic Force Microscope (AFM)
The AFM involve a sharp probe that is scanned across the surface of a sample. While the tip
is being scanned on the sample surface, the interatomic forces between the surface of the
sample and the tip causes a vertical displacement of the tip and corresponding bending of the
Page 10 of 39
cantilever. This bending is governed by Hooke’s law, = , where is the force on the tip,
is the spring constant of the cantilever, and is the displacement of the cantilever. The
cantilever can be displaced by direct contact force, van der Waals forces, electric force,
magnetic force, etc. A laser beam is transmitted to the backside of the free end of the
cantilever, which reflects it to form a spot on a position-sensitive photo-detector (PSPD) as
shown in figure 2.
(a) (b)
Fig 2.(a)Schematic diagram of the basic principle of AFM (b) The z position of the probe (black curve), force-
time curve (blue curve) and current-time curve (brown curve) for one cycle in PF TUNA.
If the cantilever moves, the spot also move correspondingly. The detector then converts the
laser spot movement to electric signals, which is then provided to a computer for processing.
The AFM software interpretsthese signals, which give information about the topography of
the sample since the movement of the tip corresponds to the topography of the sample
surface.
The signal generated by the photodetectoris supplied to a feedback circuit andinto the z-
piezo. This is used to control the distance or force between the probe and the sample surface.
The AFM can be operated in different modes depending on the properties of a sample the
user wants to determine. Currents is passed through the tip to probe the electrical
Tip
Photodetector
Cantilever
Sample
Laser beam
Page 11 of 39
conductivity or transport properties of the underlying sample. This has been done in this
study using the PeakForceTunneling AFM (PF TUNA)
1.6.1. PeakForce TUNA
Like PeakForce QNM, PF-TUNA mode builds on PeakForce Tapping mode of the AFM.A
probe that is coated with an electrically conductive material is scanned across the surface of
the sample in PeakForce Tapping mode. The user applies a DC bias between the tip and the
sample. The TUNA module senses the current passing through the sample to the AFM tip. In
PeakForceTapping, the maximum force on the tip (peak force) is maintained at a constant
value for each individual cycle. By maintaining a constant tip-sample force, topographic and
current images are generated simultaneously.
A complete cycle involves a single approach and a single retract. In figure 2 (b) The top line
(black) is the Z-position of the probe/tip as a function of time, as it goes through one
period/cycle. As the tip approach and retract, the AFM senses the forces and current flowing
through the sample surface. This is given as force curves and current curves as shown in
figure 2 (b). The middle line (blue) shows the force measured by the probe during approach
and withdraws of the tip. The lower curve (brown) represents the detected current on the
sample surface for one cycle. For a frequency of 2 kHz, the time from A to E is 0.5ms. From
the current-time curve, the PF-TUNA algorithm extracts three measurements: Peak current,
TUNA/Cycle-average current, and Contact average current. The peak current is the current
detected when the AFM tip exerts maximum force on the sample surface (peak force): when
the z-scanner is at its lowest point (at point C in figure 2(b)). TUNA current is the average
current detected in one full tapping cycle (approach and retract), from point A to point E in
figure 2 (b). Contact current is the average current measured when the AFM tip is in contact
with the sample surface I.e. the current measured from point B to point D in figure 2 (b).
Page 12 of 39
TUNA current involves the current measured when the AFM tip is far away from the sample
and that measured while it is in contact with the sample surface. The AFM software extracts
this currents at every pixel and maps them into different current images, i.e. TUNA current,
Contact current, and Peak current images which are available through different AFM
channels.
1.6.1.1. PF TUNA imaging mode
In PF TUNA imaging mode, the current maps are presented at the same time with the
topography image and mechanical properties maps (Young’s modulus, adhesion force,
deformation and dissipation). The observed current can be used to give the local conductivity,
or electrical integrity of the sample. Offline analysis function (NanoScope Analysis) can be
used to calculate statistics of the electrical properties of different regions, to take cross
sections through the images showing the spatial distribution of the properties, and /or the
correlation among the topographic, mechanical, and electrical properties of the sample.
1.6.1.2. I-V Spectroscopy mode
Local I-V spectra can be measured in the PeakForce TUNA. To extract the I-V curves, first,
the imaging scan is discarded then the tip held in a fixed location and the sample bias is
ramped up and down either once or continuously. The feedback is turned to contact mode
where a constant deflection of the cantilever is maintained while ramping takes place. The
current passing across the sample is plotted against the applied DC bias. The AFM software
is able to record the average of multiple spectra or a single spectrum. The "point & shoot" can
also be used to extract the I-V curves. It is possible to manually select the specific sample
area
Page 13 of 39
of interest using this feature.In this study, both PF TUNA imaging mode and I-V
spectroscopy mode have been used to investigate the electrical properties of graphene sheets.
1.7. Basic understanding of Raman spectroscopy spectrum
Raman spectroscopy has been shown to be a very powerful technique to investigate the
disorder[47]–[50]and number of layers in a graphene sheet [51].
The Raman signals of a graphene sheet consists of a D peak (at 1350cm-1
), G peak (at 1584
cm-1
) and a 2D peak (at 2700cm-1
).The G and 2D peaks give information about the number of
layers [49], [51]–[55]. The increase in the number of graphene layers leads to an increase in
intensity of G band[54] and shifts the 2D peak towards higher energies[55]as illustrated in the
spectra in Figure 3.These spectra allow us to study the properties of graphene. The D peak
shows the presence of lattice defects [53], [56] which may include physical damage [48], [49]
or sp3 hybridization due to adatoms on the graphene surface [57].
Fig 3.The Raman spectra of one to four layers on SiO2/Si substrate. Retrieved from ref [51].
For low defect density IG does not change but ID and ID’ increase with defect density. The
intensity of the D peak reach a maximum at a certain defect density then it starts decreasing.
The intensity of the D’ peak remain constant and the peak merge with the G peak at high
defect densities [50]. The intensity ratio of the D and G peaks (ID/IG) gives information about
Page 14 of 39
the concentration of covalent defect sites [50], [58] in the graphene sheet[50]. In this study,
Raman spectroscopy has been used to determine the number of graphene layers and the
changes in structural defect density.
2. EXPERIMENT I:ELECTRICAL PROPERTIES MEASUREMENT USING
PEAKFORCE TUNA
2.1. Sample Preparation
To investigate the effect of nanoscale structural defects on the electrical properties of
graphene, thirteen (13) areas of graphene/SiO2 sample were irradiated with controlled
amounts of Gallium ion using the FIB. Each area was exposed to different amount of ion
dose to introduce deferent defect densities in these areas. The amount of ion dosage is
presentedin Table 1. Ion irradiation has been shown to be one of the best ways to introduce
controlled amounts of structural defects in graphene sheets [7], [59], [60].
In the ion irradiation process, energy is transferred[60] from the incident Ga+ ion to the
graphene sheet by interaction with electrons in the graphene layer and by direct collisions
with its carbon atoms. This interaction result to different types of defects[61] in the graphene
structure like the Stone-Wales (SW) defects, Multiple Vacancies, Single Vacancies (SV),
nano holes, foreign adatoms, carbon adatoms, etc. For ion irradiated graphene, vacancy
defects usually the majority defects [61]. Few Frenkel pairs (FP) and stone-wales defects may
also be present. Single layer graphene is influenced more by ion irradiation than bilayer and
multi-layer graphene[7]; for the same ion dose, single layer graphene show higher defect
density than bilayer and multilayer graphene. In this study, a single layer graphene has been
used. Therefore, high defect densities are expected to be present after ion irradiation. The
STM images (not shown in this report) of Ga+ ion irradiated graphene show vacancy defects
and nano-holes were created in the sheet after high ion dose exposure.
Page 15 of 39
Sample Number Ion dose(ion/cm2
)
0 0.000 x 1000
1 1.044 x 1011
2 5.219 x 1011
3 1.044 x 1012
4 2.088 x 1012
5 3.131 x 1012
6 4.175 x 1012
7 6.263 x 1012
8 1.044 x 1013
9 1.566 x 1013
10 2.609 x 1013
11 5.219 x 1013
12 1.044 x 1014
13 5.219 x 1014
Table 1.Area number and the corresponding ion dose(ion/cm2)
When low energy (few keV) ions are used to irradiate graphene sheets on SiO2, the projection
range of the implanted ions in the SiO2 is small (few nanometers). This affects the defect
release in graphene sheet [7]. In this study high energy Ga+ ions were used. It is expected
that the projection range is large i.e. the ions travel deeper below the SiO2 layer, the SiO2
substrate is almost unaffected by this implantation. This was proven in a previous studies on
the mechanical properties of ion irradiated SiO2[62]; it was observed that ion irradiation does
not affect the mechanical property of the SiO2 substrate.
2.2. Raman spectroscopy studies
The Raman spectroscopy has been used to study the nature of the sample before and after ion
irradiation. The Raman spectrum and an AFM image of un-defected graphene/SiO2 is shown
in figure 4. It was observed that, for the graphene sample used in this study, the intensity ratio
of 2D peak (2700 cm-1
) and G peak (1550 cm-1
) is three. This confirms that the graphene is a
single layer (see section 1.7 above). The D peak for graphene sample before irradiation was
Page 16 of 39
found to be very small as shown in figure 4(a). That shows that the level of defects is very
low.
(a) (b)
Fig 4. (a) Raman spectrum for un-defected graphene (before Ga+ ion irradiation) (b) AFM
Height image of graphene sheets before irradiation. Extracted from [62]
The AFM image in Figure 4 (b) showsgrain boundaries and foreign nanoparticles in the
sample surface. These defects are expected to affect the conductivity of the graphene sheet, as
we shall see later in this study.
Fig 5.Plot of ID/IG as a function of ion dosage in RT-graphene (black) and Cryo-graphene
(red). Extracted from [62]
G
D
D
Page 17 of 39
In Raman, the intensity ratio of D peak and G peak indicates the defect level. Normally this
can be well explained by the structure change in graphene lattice (See section 1.7). Figure 5
shows the relation of the intensity ratio of D peak and G peak in Raman spectroscopy as a
function of the applied ion dosage. When the defects increase, the ratio of ID/IG will also
increase until a peak, this is the typical transfer of graphene from single crystal to
polycrystalline. Then the ratio will drop, indicating the transfer from polycrystalline to
amorphous carbon [7]. In figure 5, both samples show this typical curve above, but Cryo-
graphene indicates a significant left-shift compared with RT-graphene, which means that at a
certain ion dosage, Cryo-graphene has higher defects density than RT-graphene. It is good to
note that in this study only graphene sample prepared in room temperature has been
investigated.
2.3. Experimental Details
The defected sample was then placed on a steel sample holder using a carbon tape. Silver
paste was applied on two sides (shown in figure 6) of the sample. These would connect the
graphene sheet to the steel sample holder for current to flow through the AFM stage, the
sample and to the AFM tip. AFM Peak Force TUNA was then used to map the current in the
samples when a DC sample bias of 2V, 3V, and 9V was applied. The current present in the
silver paste connecting graphene sheet to the steel sample holder was measured at a DC
sample bias of 2V. TUNA, contact, and peak current imageswere takenfor all the defected
areas of the graphene sheet in ambient conditions. The probe used in this experiment was
SCM-PIT with spring constant of 1-5N/m. The tip was coated with conductive platinum-
Iridium which has an electrical resistivity of ≈ �Ω. [63]. To estimate the amount of
current that can flow through the tip we assume a cylindrical shape of length = and
radius � = . From ohms law,
Page 18 of 39
= � ��2 [1]
Where is the voltage, � is current, and the resistivity. Substituting the values, we get a
current of 17.6mA. This shows that the tip is good for this study. The silver paste used is
large in dimensions and so it is expected that it have low electrical resistance as seen in this
study.
(a) (b)
Fig 6.Schematic diagram showing the set up in this experiment.(a) Top view of the graphene sample. (b) a cross
section of the set up. Silver paste has been applied on two points to connect the graphene layer to the steel
sample holder.
To understand the conductivity of the sample, I-V curves were taken for randomly chosen
points of the silver paste and un-defected graphene using the I-V Spectroscopy mode of the
AFM.
While scanning in PF TUNA, the cantilever oscillates at a rate of about 2kHz.This means
each single oscillation take 0.5ms. The AFM measures the current through the sample when
the tip is in contact with the sample (contact and peak current) and when not in contact(see
section 1.6.1 for more details on the currents mapped).
Page 19 of 39
Gas out Gas in
SiO2/Si
Graphen
e
Gold
SourceType equation here.Drain
3. EXPERIMENT II: TESTING THE GAS SENSING PROPERTY OF GRAPHENE
3.1. Experimental details
In this experiment, the gas sensing properties of graphene sheet was investigated. The
sample was first cleaned using ethanol and ultraviolet (UV)/ozone cleaning. Then it was
placed inside the gas-sensingchamber, which is mounted inside APS-4 Probe station. The gas
chamber had a total volume of 153.86 ml. It has a gas inlet and gas outlet. Both inlet and
outlet have same cross sectional area of 0.785mm2.Before gas sensing experiment, sample
was exposed to UV light.
(a)
(b) (c)
Fig. 7 (a) schematic diagram of the gas sensing set up. (b) The optical image and also the current-voltage curve
of the graphene device used.
The two gold (Au) electrodes were touched with needles/probes to form the source and drain.
A schematic diagram of the setup is shown in figure 7(a). The graphene device and its I-V
curve are also shown in figure 7(b) and (c). From the I-V curve we see that the graphene
sheet is Ohmic. The container was then closed tightly and N2 gas supplied to the container at
Au
Page 20 of 39
a flow rate of 10ml/s. The I-t measurements were taken in this N2 condition at room
temperature by applying a voltage of 0.5V. Current through the graphene sheet was recorded
for about 30 minutes then the N2 gas supply was stopped and NO2 gas, diluted in N2 gas,
supply started. This mixture of N2 and NO2 gas (85% and 15% respectively and its
concentration was 15ppm) was supplied at a flow rate of 40ml/s. The I-t measurements were
taken for a voltage of 0.5V. The gas detection of graphene was studied by monitoring the
current change upon exposure of the graphene sheet to the gas [64] using the Keithley 6430
source meter.
4. RESULTS AND DISCUSSION
4.1. Electrical properties of Silver contacts
The electrical conductivity of the silver contacts(see Figure 6) was studiedto establish the
current available to the graphene sheet. For the silver paste, current as high as 1.2nA was
measured for a DC sample bias of 2V.It has been observed that, high current was measured
on the steep parts of the samples but on the flat areas, measured current was almost zero (see
the figure in appendix I ). This can be understood by analyzing the tip-sample interaction
mechanism. Figure 8 illustrates a AFM tip as it is scanned along the sample surface.
(a) (b)
Fig. 8. (a) Schematic diagram of the AFM tip as it is scanned over a sample with slopes. (b) Show multiple PF
TUNA I-t curves.
AFM tip
Sample
Page 21 of 39
As the tip is scanned over a flat sample surface, only its sharp end interacts with the sample.
Very small current or even zero current was detected in that case. However, when the tip is
scanned over the slopes in the sample surface, the interaction involves the sides of the tip and
the sample surface, i.e. the tip touch the sample surface by its sides as shown in figure 8. This
has the effect of increasing the contact area between the sample surface and the AFM tip
hence higher current is expected to flow. This confirmed that the sharp part of the tip was not
as conductive as its sides. As a result of this effect, the graphene samples revealed low
currents on the flat surfaces and relatively higher currents on the grain boundaries. This led to
low average current for the entire scanned areas. The problem was observed even after
changing the probe to a new one. However, for comparison purpose the results obtained in
this study is reliable assuming the topography is uniform for all the areas.
The current present in the graphene sheet and the measured resistance is greatly affected by
the electrical properties of silver contacts. A resistance as low as . × − V/pA
( . × + Ω) was measured for silver paste contacting graphene sheet to the steel
sample holder.This resistance is expected to contribute to the resistance measured on the
graphene sheet.
4.2. Electrical properties of Graphene
The electrical property of graphene on SiO2 substrate has been investigated in this study
using the PeakForce TUNA mode of the AFM. The I-V curves obtained using the I-V
spectroscopy mode (see section 1.6.1.2) for un-defected graphene sheet on SiO2substrate
(Figure 9) show linear behavior indicating lack of significant Schottky barriers at the tip-
graphene interface. This observation is consistent with previous studies[65]; Studies have
shown that, graphene form Ohmic contacts with metals[65].From the I-V curves, the
resistance of the samples has been estimated (figure 9).
Page 22 of 39
From the relation,
� = � =
the resistivity of samples can be determined. Where is the length of the material travelled
by current, is the resistivity, is the width, and the thickness of the sheet. In the PF
TUNA experiment set up, the length L and width W are not well defined, thus making it
difficult to calculate the resistivity of the graphene sheet in this study.
dX/dY = 5.3535e-005 V/pA dX/dY = 5.2713e-004 V/pA dX/dY = 9.6606e-004 V/pA
Fig 9. I-V curves for different possition of the un-defected Graphene Sheet.The blue and red curves show the
approach and retract curves.
The graphene sample used in this study showed a lower electrical conductivity than expected.
This could have been due to a number of factors. First, as we have seen above(section 4.1),
the interaction of the sample surface and the AFM tip may have contributed to the lower
measured current in this study. Second, the graphene-metal contact resistance contribute to
the total resistance of graphene based devices [66], [67]. This may be due to reflection of
charge carrier at the interface and/or contaminations by impurities. The contact resistance
between graphene and silver and that between graphene and the AFM tip can affect the
amount of current measured by the AFM. Third, the electrical continuity in graphene sheet is
greatly influenced by the presence of cracks, and grain boundaries[68]. These defects cause
electrical discontinuity hence lower the electrical conductivity of the graphene sheet in this
study by increasing its electrical resistance. Electrical discontinuity makes some areas of the
graphene sheet show a lower current than expected as seen in this study.
Page 23 of 39
(a)
(b)
(c)
(d)
Fig. 10. PF TUNA contact curent images and their cross section graphs for a DC sample bias of 3V for
graphene samples (a) area 4, (b) area 7, (c) area 12, and (d) Area 13. Electrical conductivity decrease with
increase in structural defect density.
Page 24 of 39
13.0 13.3 13.5 13.8 14.0 14.3 14.5 14.8
0.00.0
0.3
0.50.5
0.8
1.01.0
1.3
1.51.5
1.8
2.02.0
TUNA Current
Contact Current
Peak Current
Cu
rre
nt(
pA
)
log(ion dose)
Nevertheless, the results obtained in this study was good in comparing the electrical
properties of graphene sheets with different defect density; since a single graphene sheet has
been used, it is expected that these factors, described above, are constant for all the areas. The
PeakForce TUNA images show a very good correlation between the sample topography and
current measurement.
Using the PeakForce TUNA imaging mode (described in section 1.6.1.1), TUNA, Contact,
and peak current measurements for all the areas have been taken and compared in figure 11
and 12. DC sample bias of 2V, 3V, and 9V have been used in this study. Figure 10 show the
contact current for different areas with different defect densities for an applied DC sample
bias of 3V. It is seen that the contact current decrease with increasing defect density.
However, higher current is detected on the grain boundaries than the other areas of graphene
sheet.
For an applied DC sample bias of 9V, a graph of current versus logarithm of ion dose
revealed that current decreased with defect density. However, the results were only possible
for areas 13 to 9, there after there was no current detected. It was speculated that the current
Graph 11. Graph of average current Vs area number for graphene samples using a DC sample bias of 9V.
Page 25 of 39
became too high and the probe coating melted or the tip coating could have been worn out
after several mappings.
Figures 11 and 12 give graphs of average current (taken over the whole current images)
against log (ion dose). It was found that, the conductivity of graphene sheets decreases with
increase in defect density. Conductivity throughout the sheet was not uniform; some areas
showed a lower conductivity than expected. This may be due to electrical discontinuity
caused by cracks and grain boundaries [68] observed in the graphene sheets. No current was
detected on area 13, which is highly defected(exposed to ion dose of 5.219 x 1014
ions/cm2).
Figure 10 (d) shows that the current in area 13 remain zero after a DC sample bias is applied.
This shows that structural defects are able to transform graphene sheet from a good electrical
conductor to a non-conductor. Previous studies have found that structural defects change
graphene from metallic to insulator [69]. This effect has been observed in this study.
(a) (b)
Fig 12. (a) Graph of average Current against log(ion dose) for a DC sample bias of 3V (b) Graph of Current(pA)
against log(ion dose) for a DC sample bias of 2V
These results are in agreement with previous studies [4], [36], [37], [41], [43]. The presence
of structural defect are anticipated to greatly influence the electronic properties of graphene
material such as the charge carrier mobility and hence its electrical conductance. Atomic
0.0 12.0 12.5 13.0 13.5 14.0 14.5
0
1
2
3
4
5
6
7
8
9
10
11
12
TUNA current
Contact current
Peak current
Curr
ent(
pA
)
log(ion dose)0.0 12.0 12.5 13.0 13.5 14.0 14.5
-1
0
1
2
3
4
5
6
TUNA Current
Contact Current
Peak Current
Cu
rre
nt(
pA
)
log(ion dose)
Page 26 of 39
scale structural defects scatter the charge carriers[3], [6], [35]–[37] hence increase the
electrical conductivity of the material.
(a) (b)
(c) (d)
Fig.13.Graphs of current against log(ion dose) (a) grain boundary current for a DC bias of 2V (b) grain
boundary current for a DC bias of 3V (c) Pristine graphene current for a DC bias of 2V (d) pristine graphene
current for a DC bias of 3V.
From figure 13, it is clear that, grain boundaries show higher electrical conductivity than the
pristine graphene. It has been observed that the grain boundary current decrease with increase
in defect density as shown in figure 13 (a) and (b). This shows that the grain boundaries are
affected by ion irradiation. The current measured on the surface of the pristine graphene was
relatively small. This may be because of the effects of the factors described in section 4.1
above. The interaction of AFM tip and the sample surface may influence the current
0.0 12.0 12.5 13.0 13.5 14.0 14.5
0
25
50
75
100
125
150
175
200
225
250 TUNA Current
Contact Current
Peak Current
Cu
rre
nt(
pA
)
log(ion dose)
0.0 12.0 12.5 13.0 13.5 14.0 14.5
0
25
50
75
100
125
150
175
200
225
250
275
300
TUNA Current
Contact Current
Peak Current
Cu
rre
nt(
pA
)
log(ion dose)
0.0 12.0 12.5 13.0 13.5 14.0 14.5
-10
-8
-5
-3
0
3
5
8
10
TUNA Current
Contact Current
Peak Current
Curr
ent(
pA
)
log(ion dose)
0.0 12.0 12.5 13.0 13.5 14.0 14.5
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
6
7
8
9
10
TUNA Current
Contact Current
Peak Current
Cu
rre
nt(
pA
)
log(ion dose)
Page 27 of 39
measured. These changes occur due to changes in the contact area caused by the topography
of the surfaces (see section 4.1 for details).
It has been observed from graphs 13 (c) and (d) that graphene sheet (free of grain boundary)
with low defect density has a higher electrical conductivity compared with that with higher
defect densities. This shows that structural defects increase the electrical resistance of
graphene sheet.
4.2.1. Evaluation
The PF TUNA mode of the AFM has been shown in this study to be a very powerful
technique in comparing the conductivity of different samples. Its capability to study the
electrical properties of highly conductive and hard samples has been tested in this study. The
challenge sets in when the current to be measured is high, the tip conductive coating could
melt and make it nonconductive. In addition, scanning the tip along a hard surface lead to
wearing out of the tip conductive coating and makes the tip nonconductive. This was a major
problem in this study; it was observed that the sharp part of the tip was less conductive (see
section 4.1).
However, it was possible to compare the electrical properties of different areas with different
defect density and also to get the I-V curves for graphene/SiO2. The I-t curves (Figure 8 (b))
were found to be in good shape. This show that the AFM was operating normally.The PF
TUNA has been shown to be a very effective method in mapping the electrical properties of a
material. From the height, mechanical properties, and current images, it is possible to analyze
the current properties of different areas of the sample and to compare the conductivity with
the topography of the samples. The relationship between structural defect density and the
electrical properties of graphene obtained in this study agree with previous studies [4], [36],
Page 28 of 39
[37], [41], [43].Therefore, the PeakForce TUNA technique used in this study is good and
reliable in comparingthe electrical properties of different materials.
4.3. Gas sensing properties of graphene
The NO2 gas sensing property of graphene has been investigated in N2 gas at room
temperature. Before the experiment, graphene sample was first cleaned using ethanol and
ultraviolet (UV)/ozone cleaningthen exposed to UV light. A reductionin the baseline
conductivity was observedafter exposure to UV light as shown in Figure 13 (b). Before
exposing graphene to UV light, graphene’s sensitivity to NO2 was insignificant. However,
after the exposure to UV light, it showed a good sensitivity to NO2 gas molecules. Figure
13(a) shows the response after exposure to NO2 gas molecules. It has been observed that,
after NO2 gas supply was started, the conductivity of graphene sheet remained fairly constant
for about 45 seconds. Within this time, it is expected that, the gas-sensing chamber be filled
by N2 gas that was then driven out and NO2 gas concentration increased from zero to a higher
concentration. After the 45 seconds, a rapid increase in current was observed. The current
increased at a rate of about 9.081nA/s. Figure 13 (c) illustrate a linear fit of the rapid change
in current after graphene was exposed to NO2 gas. Six minutes after the start of NO2 gas
supply, a gradual decrease in the rate of change of current was observed that follow
Boltzmann relation. This decrease may be due to some gas molecules leaving the graphene
surface while other molecules are being absorbed into the surface i.e. the rate at which the gas
molecules are leaving the graphene surface increase with time. A response of 12.56% has
been achieved upon exposure to NO2 gas for 46 minutes. The response time(the time taken
for the current to have a rise equal to three times the noise level) was found to be 275s (4.58
minutes).
Page 29 of 39
1700 1720 1740 1760 1780 1800 1820 1840
0.0002
0.0004
0.0006
0.0008
0.0010
0.0012
0.0014
0.0016
Cu
rre
nt(
A)
Time(s)
A
(a)
(b) (c)
(d) (e)
Fig 13(a)The I-t curve for NO2gas sensing response of graphene sheet. (b) Change in conductivity of graphene
after it was exposed to UV light. The baseline conductivity dropped after graphene sheet was exposed to UV
light (c) linear fit of the rapid change in current after graphene was exposed to NO2 gas (occurred 45s after NO2
gas supply was started). (d) Gradual decrease in the rate of change of current (occurred 260s(6minutes) after
NO2 gas supply/exposure start).(e) I-t curve showing exponential decay of current after NO2 gas supply was
stopped and N2 gas supply started.
NO2
N2
N2
4000 6000 8000 10000 12000
0.0000270
0.0000275
0.0000280
0.0000285
0.0000290
0.0000295
0.0000300
0.0000305
Data: Data1_A
Model: ExpDec1
Equation: y = A1*exp(-x/t1) + y0
Weighting:
y No weighting
Chi^2/DoF = 1.9691E-14
R^2 = 0.96215
y0 0.00003 ±2.388E-9
A1 0.00002 ±2.4503E-7
t1 1732.623 ±7.12133Cu
rre
nt (A
)
Time(s)
A
ExpDec1 fit of Data1_A
1500 2000 2500 3000 3500 4000
0.0000275
0.0000280
0.0000285
0.0000290
0.0000295
0.0000300
0.0000305
0.0000310
0.0000315
Data: Data4_A
Model: Boltzmann
Equation:
y = A2 + (A1-A2)/(1 + exp((x-x0)/dx))
Weighting:
y No weighting
Chi^2/DoF = 1.6938E-14
R^2 = 0.96976
A1 0.00002 ±3.446E-6
A2 0.00003 ±1.3804E-8
x0 295.09027 ±228.61916
dx 708.73677 ±24.75321
Cu
rre
nt (A
)
Time(s)
A
Boltzmann fit of Data4_A
1000 1050 1100 1150 1200 1250 1300 1350 1400
0.0000250
0.0000255
0.0000260
0.0000265
0.0000270
0.0000275
0.0000280
Y =1.55919E-5+9.081E-9 X
Cu
rre
nt (A
)
Time(s)
A
Polynomial Fit of Data1_A
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
24
25
26
27
28
29
30
31
32
Cu
rre
nt (u
A)
Time(s)
A
Page 30 of 39
The rise time (time taken for the current to rise to 90% of the maximum value) was found to
be 1850s (30.8 minutes). A recovery of 10.2% has been achieved in N2 gas condition. When a
graphene sheet is exposed to NO2 gas, electrons move from the graphene sheet to the gas
molecules, thus creating more holes; NO2 molecules act as electron acceptors[70]. This make
graphene sheet more electrically conductive, hence the current measured increase. Initially, it
is expected that, there are more NO2 gas molecules being absorbed into graphene surface than
the number of the gas molecules leaving the surface. This may lead to the rapid change in the
conductivity of graphene due to the high concentration of holes created in the graphene sheet
as a result of electron transfer from graphene to NO2 gas molecules. After some time of gas
supply, the system attains equilibrium and the rise stops.A saturation point (constant current)
is expected at which supply of more NO2 does not lead to a significant change in graphene’s
conductivity. The equilibrium may be because the rate at which gas molecules are leaving the
surface becomes equal to the rate at which gas molecules are being absorbed into the surface.
In this study, the saturation did not come out clearly. After NO2 gas supply was stopped and
N2 gas supply started the current decreased exponentially as shown in figure 13 (e)but could
not drop to the initial baseline current. This may be due to adsorbed gas molecules on the
surface.
4.3.1. Evaluation
The setup used in this experiment has been shown to be an effective way of testing the gas
sensing property of graphene. It is easy to control the conditions inside the container and
prevent contaminants. This setup can be developed to make a graphene based gas sensor. It
has been observed that, graphene’s electrical conductivity increase upon exposure to NO2 gas
molecules. The results obtained in this study are comparable to previous results[13], [38],
Page 31 of 39
[46], [70]; NO2 molecules increase the electrical conductivity of graphene sheet. Graphene
sheet is therefore a good material for use in NO2 gas sensors.
Chen et al. [44] found that, the conductance of graphene increase by about 1% when its
exposed to 40 ppt of NO2 for 5 min. They have estimated a detection limit (DL) of 2.06ppt.
Other researchers have achieved a DL as low as 5 ppm [14], [71] and even 1ppb [13]
detection of NO2 molecules by graphene. Compared with these results, the sensitivity
reported in this study is good. Though the detection limit has not been tested in this study, it
has been shown that graphene sheet can detect NO2 gas molecules at concentrations of
15ppm.
5. POSSIBLE APPLICATIONS OF RESULTS
Studies on graphene material have advanced speedily from fundamental physics theory to
application stage [72]–[80]. Graphene has been found to be a suitable material in designing
efficient, flexible and light electronics [72], flexible displays [79], batteries and
electrochemical capacitors [76], etc. There is need for electronic devices with lower power
consumption and of higher performance. These applications of graphene greatly rely on,
among others, the electrical conductivity of graphene. It is therefore important to understand
the electrical conductivity of graphene with structural defects since defects are unavoidable.
As shown in this study, electrical conductivity of graphene sheet decrease with increase in
defect density. If high electrical conductivity is needed, one must ensure the graphene
material used has minimum structural defects density possible. The finding in this study can
be used to engineer the electrical conductivity of graphene for specific applications.
Graphene can be used as a NO2 gas molecule sensor. It has been shown in this study that,
graphene has a good sensitivity to NO2 gas molecules. This result can be applied to design
Page 32 of 39
NO2 gas molecule sensors. Although more study is needed to establish how the responses due
to different types of gas molecules can be distinguished.
6. CONCLUSION
In this study, the electrical conductivity of Ga+ ion irradiated single-layer graphene sheet on
SiO2 substrate were studied using the PeakForce TUNA mode of the AFM. It has been shown
that, the electrical conductivity of graphene sheet decrease with increasing density of atomic
scale structural defects. Grain boundaries show higher electrical conductivity than the pristine
graphene sheet. The PeakForce TUNA has been shown to be a very powerful technique in
comparing the electrical conductivities of different samples. The results obtained in this study
offer useful information on the properties of graphene that can be used to engineer the
properties of graphene to suit specific applications; it gives an insight on the electrical
conductivity of defective graphene and its grain boundaries.
Finally, it has been shown that graphene has a good sensitivity to NO2 gas molecules. Its gas
sensitivity can be increased significantly by exposing it to UV light as it has been shown in
this study. A response of 12.56%was achieved after 46 minutes of exposure to NO2 gas
molecules at 15ppm. This shows that graphene is a good material to use as a NO2 molecules’
sensor.
ACKNOWLEDGEMENT
I greatly appreciate the work of my supervisor, Prof Klaus Leifer, my co-supervisors Hu Li
and Ishtiaq H. Wani, my examiner, Prof Håkan Rensmo and the whole ELMIN group. I thank
my sponsor Erasmus Mundu organization for their support.
Page 33 of 39
References
[1] X. Du, I. “ka hko, A. Barker, a d E. Y. A drei, Approa hi g allisti tra sport i suspe ded graphe e., Nature nanotechnology. 2008.
[2] K. I. Bolotin, K. J. Sikes, Z. Jiang, M. Klima, G. Fudenberg, J. Hone, P. Kim, and H. L. Stormer,
Ultrahigh ele tro o ilit i suspe ded graphe e, Solid State Commun., vol. 146, no. 9–10, pp. 351–355, Jun. 2008.
[3] S. H. M. Jafri, K. Carva, E. Widenkvist, T. Blom, B. Sanyal, J. Fransson, O. Eriksson, U. Jansson,
H. Grennberg, O. Karis, R. Quinlan, B. C. Holloway, and K. Leifer, Co du ti it e gi eeri g of graphe e defe t for atio , ol. , p. , .
[4] K. Thiyagarajan, A. Ananth, B. Saravanakumar, Y. S. Mok, and S.-J. Ki , Defe t-induced
metallic-to-se i o du ti g tra sitio i ultila er graphe e, RSC Adv., vol. 5, no. 22, pp.
16821–16827, Feb. 2015.
[5] Y.-B. Zhou, Z.-M. Liao, Y.-F. Wang, G. S. Duesberg, J. Xu, Q. Fu, X.-S. Wu, and D.-P. Yu, Io irradiatio i du ed stru tural a d ele tri al tra sitio i graphe e., J. Chem. Phys., 2010.
[6] F. Banhart, J. Kotakoski, a d A. V. Krashe i iko , “tru tural defe ts i graphe e, ACS
Nano. 2011.
[7] G. Co pag i i, F. Gia azzo, “. “o de, V. Rai eri, a d E. Ri i i, Io irradiatio a d defe t for atio i si gle la er graphe e, Carbon N. Y., vol. 47, no. 14, pp. 3201–3207, Nov. 2009.
[8] O. Lehti e , I. L. Tsai, R. Jalil, R. R. Nair, J. Kei o e , U. Kaiser, a d I. V. Grigorie a, No -
invasive transmission electron microscopy of vacancy defects in graphene produced by ion
irradiatio ., Nanoscale, vol. 6, no. 12, pp. 6569–76, Jun. 2014.
[9] C.-T. Pan, J. A. Hinks, Q. M. Ramasse, G. Greaves, U. Bangert, S. E. Donnelly, and S. J. Haigh,
I -situ observation and atomic resolution imaging of the ion irradiation induced
a orphisatio of graphe e, Sci. Rep., vol. 4, no. i, p. 6334, Jan. 2014.
[10] J. Kotakoski a d O. Lehti e , Na o a hi i g Graphe e ith Io Irradiatio , MRS Proc., vol.
1259, no. Md, 2010.
[11] L. Tapaszto, G. Dobrik, P. Nemes-Incze, G. Vertesy, P. Lambin, L. P. Biro, L. Tapasztó, G.
Dobrik, P. Nemes-I ze, G. Vertes , P. La i , a d L. P. Biró, Tu i g the ele tro i stru ture of graphe e io irradiatio , pp. –12, 2009.
[12] J. H. Kim, J. H. Hwang, J. Suh, S. Tongay, S. Kwon, C. C. Hwang, J. Wu, and J. Young Park,
Work fu tio e gi eeri g of single layer graphene by irradiation-i du ed defe ts, Appl.
Phys. Lett., vol. 103, no. 17, 2013.
[13] F. Schedin, A. K. Geim, S. V. Morozov, E. W. Hill, P. Blake, M. I. Katsnelson, and K. S.
No oselo , Dete tio of i di idual gas ole ules adsor ed o graphe e., Nat. Mater., vol.
6, no. 9, pp. 652–655, Jul. 2007.
Page 34 of 39
[14] J. D. Fo ler, M. J. Alle , V. C. Tu g, Y. Ya g, R. B. Ka er, a d B. H. Weiller, Pra ti al he i al se sors fro he i all deri ed graphe e, ACS Nano, vol. 3, no. 2, pp. 301–306, 2009.
[15] “. No iko , N. Le ede a, a d A. “atrapi ski, Ultrase siti e NO Gas “e sor Based o Epita ial Graphe e, ol. , .
[16] I. Jo a o i , E. Cazalas, I. Childres, A. Patil, O. Ko asi, a d Y. P. Che , Graphe e field effe t transistor-based dete tors for dete tio of io izi g radiatio , i 2013 3rd International
Conference on Advancements in Nuclear Instrumentation, Measurement Methods and their
Applications (ANIMMA), 2013, pp. 1–5.
[17] Y. Liu, X. Do g, a d P. Che , Biologi al a d he i al se sors ased o graphe e aterials, Chem. Soc. Rev., vol. 41, no. 6, p. 2283, Mar. 2012.
[18] Y. Oh o, K. Maehashi, a d K. Matsu oto, Che i al a d iologi al se si g appli atio s based on graphene field-effe t tra sistors, Biosens. Bioelectron., vol. 26, no. 4, pp. 1727–1730, 2010.
[19] R. Bogue, “. Wei, Y. Zha g, a d M. Zhou, Graphe e se sors: a re ie of re e t de elop e ts, Sens. Rev. Iss Sens. Rev., vol. 34, no. 23, pp. 233–238, 2014.
[20] J. T. Robinson, F. K. Perkins, E. S. Snow, Z. Wei, and P. E. “heeha , Redu ed graphe e o ide ole ular se sors, Nano Lett., 2008.
[21] I. Ju g, D. Diki , “. Park, W. Cai, “. L. Mielke, a d R. “. Ruoff, Effe t of Water Vapor o Ele tri al Properties of I di idual Redu ed Graphe e O ide “heets, J. Phys. Chem. C, vol.
112, no. 51, pp. 20264–20268, Dec. 2008.
[22] J.-N. Fu hs a d M. O. Goer ig, I trodu tio to the ph si al properties of graphe e, Lect.
Notes, vol. 472, no. 1860, pp. 8539–8539, 2008.
[23] V. Scenev, V. Dipl Phys Vitalij Scenev Dekan der Fakultät, E. Kulke, J.-H. Olbertz Gutachter, J.
P. Ra e, P. A dre Tur ha i , a d “. Ko arik, Ele tro i properties of graphe e a d other carbon- ased h rid aterials for fle i le ele tro i s, .
[24] M. I. Kats elso , Graphe e: ar o i t o di e sio s, Mater. Today, vol. 10, no. 1–2, pp.
20–27, 2007.
[25] P. R. Walla e a d P. R. W. A e, The Ba d Theor of Graphite, Phys. Rev., vol. 71, no. 1,
1947.
[26] C. Berger, Z. Song, T. Li, X. Li, A. Y. Ogbazghi, R. Feng, Z. Dai, N. Alexei, M. E. H. Conrad, P. N.
First, a d W. A. De Heer, Ultrathi epita ial graphite: D ele tro gas properties a d a route toward graphene- ased a oele tro i s, J. Phys. Chem. B, 2004.
[27] C. Berger, Z. Song, X. Li, X. Wu, N. Brown, C. Naud, D. Mayou, T. Li, J. Hass, A. N. Marchenkov,
E. H. Co rad, P. N. First, a d W. A. de Heer, Ele tro i o fi e e t a d ohere e i patter ed epita ial graphe e., Science, 2006.
Page 35 of 39
[28] K. S. Novoselov, A. K. Geim, S. V Morozov, D. Jiang, M. I. Katsnelson, I. V Grigorieva, S. V
Du o os, a d A. A. Firso , T o-di e sio al gas of assless Dira fer io s i graphe e., Nature, vol. 438, no. 7065, pp. 197–200, Nov. 2005.
[29] A. K. Gei a d K. “. No oselo , The rise of graphe e., Nat. Mater., vol. 6, no. 3, pp. 183–191, 2007.
[30] K. J. Lee, A. P. Cha drakasa , a d J. Ko g, Breakdo urre t de sit of CVD-grown
ultila er graphe e i ter o e ts, IEEE Electron Device Lett., vol. 32, no. 4, pp. 557–559,
Apr. 2011.
[31] J. C. Charlier, P. Eklu d, J. Zhu, a d a Ferrari, Ele tro a d Pho o Properties of Graphe e: Their Relatio ship ith Car o Na otu es, Carbon Nanotub., 2008.
[32] a. H. Castro Neto, F. Guinea, N. M. R. Peres, K. “. No oselo , a d a. K. Gei , The ele tro i properties of graphe e, Rev. Mod. Phys., vol. 81, no. 1, pp. 109–162, 2009.
[33] A. Venugopal, J. Chan, X. Li, C. W. Magnuson, W. P. Kirk, L. Colombo, R. S. Ruoff, and E. M.
Vogel, Effe ti e o ilit of single-layer graphene transistors as a function of channel
di e sio s, J. Appl. Phys., vol. 109, no. 10, pp. 1–5, 2011.
[34] C. Chen, S. Rosenblatt, K. I. Bolotin, W. Kalb, P. Kim, I. Kymissis, H. L. Stormer, T. F. Heinz, and
J. Ho e, Perfor a e of o olayer graphene nanomechanical resonators with electrical
readout., Nat. Nanotechnol., vol. 4, no. 12, pp. 861–867, 2009.
[35] “. Lee a d Z. Zho g, Na oele tro i ir uits ased o t o-dimensional atomic layer
r stals, Nanoscale, vol. 6, no. 22, pp. 13283–13300, Nov. 2014.
[36] J. H. Che , W. G. Culle , C. Ja g, M. “. Fuhrer, a d E. D. Willia s, Defe t s atteri g i graphe e, Phys. Rev. Lett., vol. 102, no. 23, pp. 1–4, Jun. 2009.
[37] S. Hussain, M. W. Iqbal, J. Park, M. Ahmad, J. Singh, J. Eom, and J. Ju g, Ph si al a d electrical properties of graphene grown under different hydrogen flow in low pressure
he i al apor depositio , Nanoscale Res. Lett., vol. 9, no. 1, p. 546, Jan. 2014.
[38] T. O. Wehling, K. S. Novoselov, S. V. Morozov, E. E. Vdovin, M. I. Katsnelson, A. K. Geim, and
A. I. Li hte stei , Mole ular dopi g of graphe e, Nano Lett., 2008.
[39] O. Lee aerts, B. Partoe s, a d F. M. Peeters, Adsorptio of H O, NH , CO, NO , a d NO o graphene: A first-pri iples stud , .
[40] R. Pear e, T. Iaki o , M. A dersso , L. Hult a , A. L. “petz, a d R. Yaki o a, Epita iall gro graphe e ased gas se sors for ultra se siti e NO dete tio , Sensors Actuators, B
Chem., vol. 155, no. 2, pp. 451–455, Jul. 2011.
[41] S. Wu, R. Yang, M. Cheng, W. Ya g, G. Xie, P. Che , D. “hi, a d G. Zha g, Defe t-enhanced
oupli g et ee graphe e a d “iO su strate, Appl. Phys. Lett., vol. 105, no. 6, p. 063113,
Aug. 2014.
Page 36 of 39
[42] H. Yan, C. C. Liu, K. K. Bai, X. Wang, M. Liu, W. Yan, L. Meng, Y. Zhang, Z. Liu, R. F. Dou, J. C.
Nie, Y. Yao, a d L. He, Ele tro i stru tures of graphe e la ers o a etal foil: The effe t of atomic-s ale defe ts, Appl. Phys. Lett., vol. 103, no. 14, pp. 1–5, 2013.
[43] O. V Yaz e a d “. G. Louie, Ele tro i tra sport i pol r stalli e graphe e., Nature
materials. 2010.
[44] G. Che , T. M. Paro a , a d A. R. Harut u a , “u -ppt gas detection with pristine
graphe e, Appl. Phys. Lett., vol. 101, no. 5, pp. 0–4, 2012.
[45] C. W. Chen, S. C. Hung, M. D. Yang, C. W. Yeh, C. H. Wu, G. C. Chi, F. Ren, and S. J. Pearton,
O ge se sors ade o ola er graphe e u der roo te perature, Appl. Phys. Lett.,
2011.
[46] W. Wu, Z. Liu, L. a. Jauregui, Q. Yu, R. Pillai, H. Cao, J. Bao, Y. P. Chen, and S.-“. “. Pei, Wafer-
scale synthesis of graphene by chemical vapor deposition and its application in hydrogen
se si g, Sensors Actuators, B Chem., vol. 150, no. 1, pp. 296–300, Sep. 2010.
[47] A. Jorio, E. Ferreira, a d L. Ca çado, Measuri g Disorder i Graphe e ith Ra a “pe tros op , Phys. Appl. Graphene - Exp., pp. 439–454, 2011.
[48] E. H. Martins Ferreira, M. V. O. Moutinho, F. Stavale, M. M. Lucchese, R. B. Capaz, C. A.
A hete, a d A. Jorio, E olutio of the Ra a spe tra fro si gle-, few-, and many-layer
graphene with increasing disorder, Phys. Rev. B, vol. 82, no. 12, p. 125429, Sep. 2010.
[49] L. G. Cançado, a. Jorio, E. H. M. Ferreira, F. Stavale, C. a. Achete, R. B. Capaz, M. V. O.
Mouti ho, A. Lo ardo, T. “. Kul ala, a d a. C. Ferrari, Qua tif i g defe ts i graphe e ia Ra a spe tros op at differe t e itatio e ergies, Nano Lett., vol. 11, no. 8, pp. 3190–3196, Aug. 2011.
[50] A. Eckmann, A. Felten, A. Mishchenko, L. Britnell, R. Krupke, K. S. Novoselov, and C. Casiraghi,
Pro i g the ature of defe ts i graphe e Ra a spe tros op ., Nano Lett., vol. 12, no.
8, pp. 3925–30, Aug. 2012.
[51] Y. ying Wang, Z. H. Ni, T. Yu, Z. X. Shen, H. M. Wang, Y. hong Wu, W. Chen, A. T. S. Wee, and
A. T. “he Wee, Ra a “tudies of Mo ola er Graphe e: The “u strate Effe t, J. Phys.
Chem. C, vol. 112, no. 29, pp. 10637–10640, Jul. 2008.
[52] A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K.
“. No oselo , “. Roth, a d A. K. Gei , Ra a spe tru of graphe e a d graphe e la ers, Phys. Rev. Lett., vol. 97, no. 18, pp. 1–5, Oct. 2006.
[53] A. C. Ferrari, Ra a spe tros op of graphe e a d graphite: Disorder, ele tro -phonon
oupli g, dopi g a d o adia ati effe ts, Solid State Commun., vol. 143, no. 1–2, pp. 47–57, Jul. 2007.
[54] U. Wurst auer, Graphe e o arious su strates, U i ersität Rege s urg, .
Page 37 of 39
[55] D. C. Elias, R. R. Nair, T. M. G. Mohiuddin, S. V. Morozov, P. Blake, M. P. Halsall, A. C. Ferrari,
D. W. Boukh alo , K. M. I., A. K. Gei , a d K. “. No oselo , Co trol of Graphe e’s Properties Re ersi le H droge atio : E ide e for Grapha e, ol. , pp. –613, 2009.
[56] A. Ferrari a d J. Ro ertso , I terpretatio of Ra a spe tra of disordered a d a orphous ar o , Phys. Rev. B, vol. 61, no. 20, pp. 14095–14107, May 2000.
[57] S. Niyogi, E. Bekyarova, M. E. Itkis, H. Zhang, J. Hicks, M. Sprinkle, C. Berger, C. N. Lau, W. A.
De, “. Ni ogi, E. Bek aro a, M. E. Itkis, H. Zha g, a d K. “hepperd, “pe tros op of Co ale tl Fu tio alized Graphe e, Nano Lett., vol. 10, pp. 4061–4066, 2010.
[58] Q. H. Wang, Z. Jin, K. K. Kim, A. J. Hilmer, G. L. C. Paulus, C.-J. Shih, M.-H. Ham, J. D. Sanchez-
Yamagishi, K. Watanabe, T. Taniguchi, J. Kong, P. Jarillo-Herrero, and M. S. Strano,
U dersta di g a d o trolli g the su strate effect on graphene electron-transfer chemistry
ia rea ti it i pri t lithograph , Nat. Chem., vol. 4, no. 9, pp. 724–732, 2012.
[59] L. Tapasztó, G. Dobrik, P. Nemes-I ze, G. Vertes , P. La i , a d L. P. Biró, Tu i g the electronic structure of graphe e io irradiatio , Phys. Rev. B, vol. 78, no. 23, p. 233407,
Dec. 2008.
[60] R. Marti azzo, “. Casolo, a d G. F. Ta tardi i, The effe t of ato i -scale defects and dopants
o graphe e ele tro i stru ture, p. , .
[61] O. Lehtinen, J. Kotakoski, A. V. Krasheninnikov, A. Tolvanen, K. Nordlund, and J. Keinonen,
Effe ts of io o ard e t o a t o-dimensional target: Atomistic simulations of graphene
irradiatio , Phys. Rev. B - Condens. Matter Mater. Phys., vol. 81, no. 15, pp. 1–4, Apr. 2010.
[62] “. W. Maku i, L. Hu, I. W. Hassa , H. Re s o, a d L. Klaus, Na o e ha i al Mappi g of Defe ted Graphe e B Multi ode , .
[63] ThePGMdata ase, Plati u - . % Iridiu , . [O li e]. A aila le: http://www.pgmdatabase.com/jmpgm/index.jsp?record=1064.
[64] H. H. Pu, S. H. Rhim, M. Gajdardziksa-Josifovska, C. J. Hirschmugl, M. Weinert, and J. H. Chen,
A statisti al ther od a i s odel for o ola er gas adsorptio o graphe e-based
aterials: i pli atio s for gas se si g appli atio s, RSC Adv., vol. 4, pp. 47481–47487, 2014.
[65] W. Liu, J. Wei, X. “u , a d H. Yu, A “tud o Graphe e—Metal Co ta t, Crystals, vol. 3, pp.
257–274, 2013.
[66] A. Ve ugopal, L. Colo o, a d E. M. Vogel, Co ta t resista e i fe a d ultila er graphene devices, Appl. Phys. Lett., vol. 96, no. 1, pp. 1–4, Jan. 2010.
[67] K. Nagashio, ) T Nishimura, K. Kita, A. Toriumi, T. Nishimura, K. Kita, A. Toriumi, ) T Nishimura,
K. Kita, a d A. Toriu i, Co ta t resisti it a d urre t flo path at etal/graphe e o ta t,
Appl. Phys. Lett., vol. 97, no. 14, p. 143514, Oct. 2010.
[68] J. D. Buron, F. Pizzocchero, B. S. Jessen, T. J. Booth, P. F. Nielsen, O. Hansen, M. Hilke, E.
White a , P. U. Jepse , P. Bøggild, a d D. H. Peterse , Ele tri all Co ti uous Graphe e
Page 38 of 39
from Single Crystal Copper Verified by Terahertz Conductance Spectroscopy and Micro Four-
Poi t Pro e, Nano Lett., vol. 14, no. 11, pp. 6348–6355, Nov. 2014.
[69] J. H. Bardarso , J. T orz dło, P. W. Brou er, a d C. W. J. Bee akker, O e-parameter scaling
at the dira poi t i graphe e, Phys. Rev. Lett., vol. 99, no. 10, pp. 1–4, 2007.
[70] Y. Hajati, T. Blom, S. H. M. Jafri, S. Haldar, S. Bhandary, M. Z. Shoushtari, O. Eriksson, B.
“a al, a d K. Leifer, I pro ed gas se si g a ti it i stru turall defe ted ila er graphe e, Nanotechnology, vol. 23, no. 50, p. 505501, 2012.
[71] G. Lu, L. E. O ola, a d J. Che , Gas dete tio usi g lo -temperature reduced graphene oxide
sheets, Appl. Phys. Lett., vol. 94, no. 8, pp. 1–4, 2009.
[72] G. Fiori, F. Bonaccorso, G. Iannaccone, T. Palacios, D. Neumaier, A. Seabaugh, S. K. Banerjee,
and L. Colombo, Ele tro i s ased o t o-di e sio al aterials, Nat. Nanotechnol., vol. 9,
no. 10, pp. 768–779, Oct. 2014.
[73] W. Ha , R. K. Ka aka i, M. G itra, a d J. Fa ia , Graphe e spi tro i s, Nat.
Nanotechnol., vol. 9, no. 10, pp. 794–807, 2014.
[74] F. H. L. Koppens, T. Mueller, P. Avouris, a C. Ferrari, M. S. Vitiello, and M. Polini,
Photodete tors ased o graphe e, other t o-di e sio al aterials a d h rid s ste s, Nat. Publ. Gr., vol. 9, no. 10, pp. 780–793, 2014.
[75] Null a d “. Böh , Graphe e agai st orrosio , Nat. Nanotechnol., vol. 9, no. 10, pp. 741–742, Oct. 2014.
[76] J. Liu, Chargi g graphe e for e erg , Nat. Nanotechnol., vol. 9, no. 10, pp. 739–741, Oct.
2014.
[77] E. J. “io hi, Graphe e i the sk a d e o d, Nat. Nanotechnol., vol. 9, no. 10, pp. 745–747,
Oct. 2014.
[78] F. Torrisi a d J. N. Cole a , Ele trif i g i ks ith D aterials, Nat. Nanotechnol., vol. 9,
no. 10, pp. 738–739, Oct. 2014.
[79] J.-H. Ah a d B. H. Ho g, Graphe e for displa s that e d., Nat. Nanotechnol., vol. 9, no.
10, pp. 737–8, Oct. 2014.
[80] K. Kostarelos a d K. “. No oselo , Graphe e de i es for life, Nat. Nanotechnol., vol. 9, no.
10, pp. 744–745, Oct. 2014.
[81] J. Liu, S. Safavi-Naei i, a d D. Ba , Fa ri atio a d easure e t of graphe e p–n junction
ith t o top gates.
Page 39 of 39
APPENDIX I:
PeakForce TUNA images of silver contact
(a)
(b)
(c)
(d)
Figure 15.AFM PeakForce TUNA images of silver on graphene layer for a DC sample bias of 2V. (a) and (d)
is the height image and its cross section graph. (b) and (e) is the TUNA current image and its cross section
graph. (c) and (g) show the contact current image and its cross section graph. (f) and (h) are the peak current
image and its cross section graph. High current values were measured on the slopes of the sample topography.